id
string | text
string | source
string | added
string | created
string | metadata
dict |
|---|---|---|---|---|---|
PMC1065019
|
See related commentary, <http://ccforum.com/content/8/5/319>, See related letter, <http://ccforum.com/content/8/5/397>
Introduction
============
Tracheostomy is usually performed in patients who need prolonged mechanical ventilation, frequent suctioning of bronchopulmonary toilet or have obstruction of the upper airway. The percutaneous tracheostomy is a minimally invasive, effective and reliable procedure and has become the alternative to surgical tracheostomy \[[@B1]\]. Almost all percutaneous procedures in The Netherlands are performed with one of the three following techniques: guide wire dilating forceps (GWDF) tracheostomy, Ciaglia Blue Rhino (CBR) tracheostomy, and sequential dilation tracheostomy (classic Ciaglia) \[[@B2]\]. We have extensive experience with the first two techniques \[[@B3],[@B4]\]. This study is a sequel to our previous reports. Several studies have compared different percutaneous techniques \[[@B5]-[@B12]\], but because CBR is relatively new, a comparison with GWDF has been made only twice in two small prospective cohorts \[[@B5],[@B12]\]. The strength of the present study is the large group of patients, so the incidence of relevant complications is more meaningful.
The aim of this study was to compare GWDF and CBR. The study not only focuses on the immediate peri-operative complications but also describes the long-term sequelae of both techniques.
Methods
=======
This is a retrospective analysis of all patients who underwent percutaneous tracheostomy in the University Medical Centre Nijmegen between March 1997 and April 2003. We compared the two historic data sets that we have published previously \[[@B3],[@B4]\], but we specifically focused on the precise definition of early complications and long-term sequelae. Between March 1997 and February 2000 we performed percutaneous tracheostomy on 171 patients, using the GWDF technique. Between March 2000 and April 2003 we performed percutaneous tracheostomy on a further 171 patients, using the CBR technique. Indications, contra-indications and technique for percutaneous tracheostomy are standardised \[[@B3],[@B4]\]. Patients or family gave informed consent before the procedure. Ethical approval from the institution\'s medical ethical committee was not obtained because the standard of care was provided and no other experimental treatments were introduced. Published data cannot be reduced to a single recognisable patient. All data were recorded prospectively on pre-designed forms. \'Procedure time\' was defined as the time from incision to successful placement of the cannula. A \'peri-operative complication\' was defined as a complication related to the procedure and occurring during or within 24 hours of the procedure. Postoperative complications were divided into \'complications while cannulated\' and \'late complications\'. A \'complication while cannulated\' was defined as a complication occurring in the period between 24 hours after the procedure until removal of the cannula. A \'late complication\' was defined as a complication occurring after removal of the cannula up to a follow-up of 3 years. Complications were divided into minor and major (see Tables [1](#T1){ref-type="table"}, [2](#T2){ref-type="table"}, [3](#T3){ref-type="table"}). Moreover, complications were classified as procedure-specific and procedure-non-specific. Hypotension was defined as a systolic blood pressure of less than 90 mmHg. Hypoxaemia was defined as an arterial oxygen saturation of less than 90%. It was considered minor when lasting less than 5 min, and major when lasting 5 min or longer. Information regarding late complications was obtained by structured interviews with patients who were decannulated successfully. Patients or close relatives were asked about voice changes, dyspnoea, stridor, pain, and cosmetic problems. Patients were also asked to grade specific problems as absent, minor or major.
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Peri-operative complications
:::
Complication GWDF (*n*= 171) CBR (*n*= 171) *P* Conversion to surgical tracheostomy
------------------------------------ ----------------- ---------------- ----- ------------------------------------- -------- --- ---
No complications 128 74.9 100 58.5 \<0.01
Minor complications
Procedure-specific
Bleeding (local pressure) 11 6.4 24 14.0 0.04
Difficult dilation 0 23 13.5 \<0.01
Difficult procedure 6 3.5 7 4.1 NS
Subcutaneous emphysema 2 1.2 2 1.2 NS
Cannula insertion difficult 0 3 1.8 NS
Air leakage cuff 0 2 1.2 NS
Procedure-non-specific
Puncture endotracheal tube 9 5.3 8 4.7 NS
Puncture posterior tracheal wall 4 2.3 2 1.2 NS
Accidental detubation 1 0.6 3 1.8 NS
Hypotension 1 0.6 2 1.2 NS
Total 34 19.9 75 43.9 \<0.01
Major complications
Procedure-specific
Bleeding (exploration) 6 3.5 4 2.3 NS 2
Bleeding (life-threatening) 1 0.6 1 0.6 NS
Fausse route 2 1.2 1 0.6 NS 1
Oesophageal perforation 1 0.6 0 NS 1
Cannula insertion impossible 3 1.8 0 NS 3
Pneumothorax 0 3 1.8 NS 1
Total 13 7.6 9 5.3 NS
^a^Some patients had more than one complication. CBR, Ciaglia Blue Rhino; GWDF, guide wire dilating forceps; NS, not significant.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Complications while cannulated
:::
Complication GWDF CBR *P*
---------------------------------- ------ ------ ----- ------ ----
Surgical tracheostomy 6 2
Lost to follow-up 1 0
Available for analysis 164 169
No complications 139 84.8 138 81.7 NS
Minor complications
Bleeding (local pressure) 15 9.1 14 8.3 NS
Infection 4 2.4 6 3.6 NS
Granulation tissue around stoma 1 0.6 1 0.6 NS
Pain from stoma 1 0.6 0 NS
Tracheal oedema 0 1 0.6 NS
Subcutaneous emphysema 0 1 0.6 NS
Dyspnoea 0 1 0.6 NS
Total 21 12.8 24 14.2 NS
Major complications
Bleeding (exploration) 0 2 1.2 NS
Bleeding (life-threatening) 0 0 NS
Stridor (with empty cuff) 2 1.2 0 NS
Cardiopulmonary resuscitation 1 0.6 0 NS
Cannula obstruction 1 0.6 3 1.8 NS
Hypoxaemia 0 2 1.2 NS
Total 4 2.4 7 4.1 NS
CBR, Ciaglia Blue Rhino; GWDF, guide wire dilating forceps; NS, not significant.
:::
All data were analysed with Statistical Product and Service Solutions (SPSS) version 11.0. All variables were checked for normal distribution. Data are given as means ± SD or medians. Continuous variables were compared with Student\'s *t*-test or the Mann--Whitney test as appropriate. Bonferroni\'s correction for multiple comparisons was used. Categorisable variables were compared with the χ^2^test. A cut-off level of *P*\< 0.05 was accepted as statistically significant.
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Late Complications
:::
Complication GWDF CBR P
--------------------------- ------ ------ ----- ------ --------
Surgical tracheostomy 6 2
Lost to follow up 5 6
Still cannulated 0 3
Deceased 53 60
Available for analysis 107 100
No complications 86 80.2 73 73.0 NS
Minor Complications
Voice 9 8.5 22 22.0 \<0.01
Cosmetic problems 10 9.4 2 2.0 0.04
Pain 0 2 2.0 NS
Total minor complications 19 17.9 26 26.0 NS
Major complications
Stridor 2 1.9 1 1.0 NS
CBR, Ciaglia Blue Rhino; GWDF, guide wire dilating forceps; NS, not significant.
:::
Results
=======
Demographic data are shown in Table [4](#T4){ref-type="table"}. The procedure was successful in 165 of 171 patients (96.5%) in the GWDF group and in 169 of 171 patients (98.8%) in the CBR group. Most tracheostomies were performed by an intensivist or a fellow (under supervision). More procedures were performed by a fellow in the CBR group than in the GWDF group (51 versus 27, respectively; *P*\< 0.01).
::: {#T4 .table-wrap}
Table 4
::: {.caption}
######
Demographic data
:::
Parameter GWDF (*n*= 171) CBR (*n*= 171) *P*
-------------------------------- ----------------- ---------------- ----- --------- ------ ----- ------
Age (years) 57.5 18.2 62 57.5 18.4 62 NS
Male/Female 99/72 114/157 NS
Endotracheal intubation (days) 16.9 12.2 14 20.3 12.3 18 0.03
Procedure time (min) 9.1 8.3 5.0 10.8 10.5 7.0 NS
Cannulation time (days) 38.4 63.4 24 29.6 39.8 18 NS
Time in ICU (days) 39.4 29.8 33 44.1 38.3 34 NS
CBR, Ciaglia Blue Rhino; GWDF, guide wire dilating forceps; ICU, intensive care unit; NS, not significant.
:::
Peri-operative complications
----------------------------
Peri-operative complications are described in Table [1](#T1){ref-type="table"}. In total, there were 47 peri-operative complications in 43 patients in the GWDF group, and 84 peri-operative complications in 71 patients in the CBR group (*P*\< 0.05). This difference is explained by a greater number of difficult dilations (*P*\< 0.01) and minor bleedings with the CBR technique. After the introduction of a Crile\'s forceps for blunt dissection of the pretracheal tissues preceding CBR, the procedure became much easier. In the GWDF group, 13 patients (7.6%) had a major complication, compared with 9 patients (5.3%) in the CBR group. All these major peri-operative complications were procedure-specific. One life-threatening bleeding in the GWDF group led to severe hypoxia at the end of the procedure. After removal of the cannula, large blood clots were suctioned from the trachea. There was no significant difference in the number of patients in whom conversion to a surgical tracheostomy was necessary. In the GWDF group, six patients underwent conversion to a surgical tracheostomy: one patient had a major venous bleeding after dilation of the trachea and the cannula could not be inserted. In another patient, arterial blood was aspirated and the procedure was terminated. In two patients, the trachea was difficult to locate, resulting in hypoxaemia and hypercapnia. In one patient the guide wire was placed correctly but the cannula perforated the posterior tracheal wall and entered the oesophagus. Surgical exploration confirmed rupture of the oesophagus, and the tracheo-oesophageal wall was immediately repaired. The post-operative course was uneventful. In the last patient the distance between skin and trachea was too large for the insertion of a cannula. In the CBR group two patients underwent surgical tracheostomy: in one patient the trachea was difficult to locate, and the cannula was placed pretracheally as a result of guide wire kinking. Another patient developed major bleeding and tension pneumothorax several hours after the procedure. After immediate drainage with a chest tube, surgical exploration showed that the tracheostomy tube had perforated the cricothyroid membrane. No deaths were seen after either procedure.
Complications while cannulated
------------------------------
In total, 164 GWDF and 169 CBR patients were analysed for complications while cannulated (Table [2](#T2){ref-type="table"}). Four major complications (2.4%) occurred in the GWDF group, and seven major complications (4.1%) in the CBR group. One patient in the GWDF group had an obstruction of the cannula by a mucous plug, leading to a cardiorespiratory arrest. Another patient sustained a cardiorespiratory arrest shortly after decannulation, possibly due to aspiration. Both patients were resuscitated successfully. Three patients in the CBR group had an obstruction of the cannula: one of them died on his first day on the ward, possibly owing to an obstructive blood clot in the cannula. The second patient had a mucous plug causing severe hypoxaemia. He received a minitracheotomy through the old tracheostomy opening. The third patient with an obstructed cannula was found in bed on the ward, having a respiratory arrest. The inner cannula, which was obstructed by a blood clot, was removed. The patient recovered uneventfully.
Late complications
------------------
Of 164 patients in the GWDF group, 53 (32.3%) died with the cannula in place or within 1 week after decannulation, and five patients were lost to follow-up. One hundred and seven GWDF patients (62.6%) were decannulated successfully and analysed for late complications (Table [3](#T3){ref-type="table"}). Of 169 CBR patients, 60 (35.5%) died with the cannula in place or within 1 week of decannulation, six patients were lost to follow-up, and three patients had the cannula still *in situ*. Finally, 100 CBR patients (58.5%) were analysed for late complications. There was no significant difference between both groups with regard to total late complications. All patients with voice problems were given the opportunity to consult an ENT specialist. None of these had an objective laryngeal abnormality explaining their voice problems. Patients with cosmetic problems relating to the tracheostomy scar were offered specialist consultation. Six GWDF patients underwent scar revision. Three patients developed a severe stridor after decannulation. In the GWDF group, an 83-year-old woman had tracheal stenosis and was treated with an endotracheal stent, and an 80-year-old woman was treated with laser for a granuloma just above the tracheostomy opening. In the CBR group, an 18-year-old man suffered from severe tracheal stenosis. He had a tracheal stent placed initially, but because of recurrence of the stenosis, a tracheal resection was necessary. The patient recovered uneventfully.
Discussion
==========
In this study we have compared two different techniques of percutaneous tracheostomy, GWDF and CBR. Both techniques are frequently used in The Netherlands and are replacing the surgical technique \[[@B2]\]. This study showed no significant differences in clinically relevant complications between the two techniques. This is in agreement with two other studies comparing these techniques \[[@B5],[@B12]\]. Although the total number of complications in the two groups in the study of Ambesh and colleagues was not significantly different, the authors noticed an increased rate of minor peri-operative bleeding in the GDWF group \[[@B5]\]. This was balanced by an increase in the number of patients with one or more tracheal ring fractures in the CBR group (30%). The increase in major peri-operative bleeding with the GDWF technique might be explained by the poorly controllable dilation with the forceps \[[@B9]\]. Although the study of Añón and colleagues did not find any significant differences, in three of 26 patients in the GWDF group there was an inability to insert the cannula \[[@B12]\].
Several other studies comparing sequential dilation (classic Ciaglia) and CBR \[[@B6],[@B8]\], and comparing sequential dilation and GWDF \[[@B7],[@B9]-[@B11]\], have been described in the literature. Van Heurn and colleagues concluded that sequential dilation and GWDF are both reliable but that sequential dilation has fewer early complications than GWDF \[[@B7]\]. Nates and colleagues also preferred sequential dilation to the GWDF technique, because of fewer surgical complications, less peri-operative and postoperative bleeding, and easier use \[[@B9]\]. Añón and colleagues found a comparable complication rate, but the procedural time of the GWDF method was significantly shorter \[[@B10]\]. Unfortunately, comparing these studies is difficult because complications were not defined uniformly.
In our study, a major complication while cannulated was obstruction of the cannula, which occurred in four patients. These figures correspond to the prevalence of cannula obstruction in the literature (0.3--3.5%) \[[@B13]-[@B15]\]. Strict adherence to nursing protocols and a low threshold for cleaning the inner cannula should be the standard of care in the intensive care unit. An outreach team from the intensive care unit should visit patients, discharged to the general ward with a cannula in place, on a daily basis.
There are only few data available concerning late complications of percutaneous tracheostomy. Unfortunately, many confounders might be present, such as the disease process itself, the duration of endotracheal intubation, and other treatments in the intensive care unit (such as sedation or physical therapy). Moreover, both patients and caregivers often interpret late complications subjectively. The total number of late complications in our study was not significantly different between the two groups. Subjective voice changes and hoarseness were more frequent in the CBR group (*P*\< 0.01). An explanation might be the longer mean endotracheal intubation time, because this is possibly the most important cause of voice problems. With sequential dilation tracheostomy, the incidence of voice problems ranges between 0% and 21% \[[@B16]-[@B22]\]. More patients in the GWDF group complained of cosmetic problems. Only a few studies have mentioned cosmetic complaints, but differences of opinion between patient and caregiver are frequent \[[@B23]\]. In each group in our study, one patient developed a critical symptomatic tracheal stenosis. More patients might have had an asymptomatic tracheal stenosis, but because no additional diagnostic tests such as computed tomography or magnetic resonance imaging scans were performed, the actual incidence is unknown. Several studies have incriminated the GWDF technique as a cause of tracheal stenosis, but no studies with the CBR have been described. The incidence varied from 0% to 63% \[[@B18],[@B23]-[@B27]\]. Most of these tracheal stenoses were asymptomatic.
Several factors might decrease the strength of our conclusions. First, the study used historical data sets with a sequential design; a time bias is therefore possible. As experience with percutaneous tracheostomy increases, the number of complications will decrease, even if another technique is used, although in our study this might well have been balanced by the fact that over time more fellows performed the procedure. Second, scoring of the peri-operative complications by different physicians might be variable because of different interpretations. Despite these shortcomings, we conclude from our study that, although the CBR technique has more minor peri-operative complications, the two techniques are comparable. More prospective, randomised studies are required to compare these different tracheostomy techniques adequately. We are currently conducting a prospective, randomised study in which we compare GWDF and CBR tracheostomies; we are specifically looking for the occurrence of precisely defined early and late complications. The occurrence of tracheal stenosis will be analysed using the forced oscillation technique and magnetic resonance imaging.
Key messages
============
• GWDF and CBR tracheostomy seem equally reliable.
• Major peri-operative complications occur in 5.3--7.6% of patients.
• Late complications are rare
Competing interests
===================
None declared.
Abbreviations
=============
CBR = Ciaglia Blue Rhino; GWDF = guide wire dilating forceps.
|
PubMed Central
|
2024-06-05T03:55:54.819238
|
2004-7-5
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065019/",
"journal": "Crit Care. 2004 Jul 5; 8(5):R299-R305",
"authors": [
{
"first": "Bernard G",
"last": "Fikkers"
},
{
"first": "Marieke",
"last": "Staatsen"
},
{
"first": "Sabine GGF",
"last": "Lardenoije"
},
{
"first": "Frank JA",
"last": "van den Hoogen"
},
{
"first": "Johannes G",
"last": "van der Hoeven"
}
]
}
|
PMC1065024
|
See related commentary, <http://ccforum.com/content/8/5/322>
Introduction
============
Patients with multiple trauma often require mechanical ventilation for prolonged periods because of their inability to protect their airways, persistence of excessive secretions, and inadequacy of spontaneous ventilation \[[@B1]\]. Tracheostomy plays an integral role in the airway management of such patients, but its timing remains subject to considerable practice variation \[[@B2]\]. The decision to proceed to tracheostomy is often made only if the patient could not be extubated within 10--14 days or more \[[@B3]\]. In 1989, the American College of Chest Physicians Consensus Statement on Artificial Airways in Patients Receiving Mechanical Ventilation considered translaryngeal intubation to be the preferred technique for patients requiring up to 10 days of mechanical ventilation \[[@B4]\]. For those with anticipated need for artificial airway for more than 21 days, tracheostomy was recommended. For all other patients, the decision regarding the timing of tracheostomy was left to daily assessment and physician preference. Such practice was based on earlier reports showing high tracheal stenosis rates with tracheostomy as compared with endotracheal intubation \[[@B5],[@B6]\]. For example, one study reported in 1981 \[[@B6]\] found an incidence of tracheal stenosis after tracheostomy of 65%, as compared with 19% after endotracheal intubation. The authors of that study concluded that tracheostomy for patients requiring an artificial airway for periods as long as 3 weeks could not be recommended. However, the incidence of tracheal stenosis has decreased substantially with recognition of its aetiology and improvements in tracheostomy materials, design and management \[[@B7]\], particularly with the use of high-volume, low-pressure cuffs. Also, the complications associated with prolonged endotracheal intubation are increasingly being recognized, including injury to the larynx and trachea, and patient discomfort. In addition, endotracheal intubation often requires the administration of systemic sedation, with attendant complications. Finally, the incidence of ventilator-associated pneumonia is related directly to the duration of mechanical ventilation \[[@B8]\] -- a complication that carries significant morbidity and mortality \[[@B9]\].
One of the under-appreciated consequences of delaying tracheostomy is prolonged mechanical ventilation and intensive care unit (ICU) stay. Notably, the large body of literature addressing local complications of tracheostomy contrasts with the paucity of reports on the advantages of this procedure, especially its impact on resource utilization. This contrast may have encouraged practitioners to consider alternatives to tracheostomy. The aim of the present study is to examine the impact of early tracheostomy on resource utilization in ICU trauma patients. This examination is followed by a review of the existing literature in this area.
Methods
=======
Settings
--------
The study was performed at a major tertiary care trauma centre in Riyadh, Saudi Arabia. The 600-bed hospital has a 21-bed medical/surgical ICU staffed by full-time, on-site intensivists 24 hours a day and 7 days a week. Our department has nine consultant intensivists, all of whom are certified in critical care. The hospital has a designated trauma service, including a consultant surgeon, available 24 hours a day. Medical care in the ICU is provided by the ICU team, with the trauma team being responsible for surgical aspects of care. Ventilatory management, and decisions regarding extubation or tracheostomy and discharge from the ICU are made primarily by the ICU team. All percutaneous tracheostomies are performed at the bedside by the ICU team.
Data collection
---------------
We have maintained a prospective database including all consecutive ICU patients admitted since March 1999. For the present study we extracted data on all consecutive patients admitted to the ICU over a 5-year period (March 1999 to February 2004) with new trauma and who underwent tracheostomy during their ICU stay. We excluded patients with history of previous trauma but admitted to the ICU for other reasons, readmissions to the ICU and trauma referrals from other hospitals. Data were collected on demographics and admission severity of illness, estimated using the Acute Physiology and Chronic Health Evaluation (APACHE) II \[[@B10]\], Simplified Acute Physiology Score II \[[@B11]\], postresuscitation Glasgow Coma Score (GCS) and Injury Severity Score (ISS) \[[@B12],[@B13]\]. We documented the presence of injuries to brain, maxillofacial bones, chest, abdominal organs, spinal cord and pelvis/lower extremities. We documented whether an extubation trial was given before tracheostomy. The type of tracheostomy procedure (surgical versus percutaneous) was recorded. The number of days from initiation of ventilation to tracheostomy, from admission to tracheostomy, from tracheostomy to weaning, from tracheostomy to discharge from ICU, the duration of mechanical ventilation, ICU length of stay (LOS) and hospital LOS were all calculated. All these durations were calculated as the number of calendar days, with the day of admission being considered day 0. ICU and hospital mortality rates were documented.
We stratified patients into two groups: the early tracheostomy group, in which tracheostomy was performed within the first 7 days of initiation of mechanical ventilation; and the late tracheostomy group, in which tracheostomy was performed after 7 days. Prolonged ICU stay was defined as ICU stay in excess of 14 days.
Statistical analysis
--------------------
Minitab for Windows, release 12.1 (Minitab Inc., State College, PA, USA), was used for statistical analysis. Continuous variables are expressed as means ± standard error of the mean, and were compared using t-tests. Medians and interquartile ranges are also given. Categorical variables are expressed as absolute and relative frequencies, and were compared using χ^2^tests. Linear correlation was performed to test for associations between the duration from initiation of mechanical ventilation to tracheostomy and ICU LOS. To assess further the impact of delayed tracheostomy on ICU LOS, univariate and multivariate analyses were performed to examine whether delayed tracheostomy is an independent predictor of prolonged ICU stay. Results of prediction are expressed as odds ratios (ORs) and 95% confidence intervals (CIs). P ≤ 0.05 were considered statistically significant.
Results
=======
Baseline patient characteristics
--------------------------------
Table [1](#T1){ref-type="table"} summarizes the patients\' characteristics at baseline. During the period of study there were 653 trauma admissions to the ICU. The number of patients who required tracheostomy was 136 (21%); 29 patients had tracheostomy within 7 days of mechanical ventilation and the remaining 107 underwent tracheostomy after 7 days. Comparison of demographic data between the two groups revealed no significant differences with regard to age, sex, APACHE II score, Simplified Acute Physiology Score II or ISS. GCS was slightly lower in the early tracheostomy group (5.2 ± 0.5 versus 6.5 ± 0.4; *P*= 0.04). There was no significant difference in the presence of head, chest, abdominal, or pelvic injuries between the groups. Maxillofacial injuries were more common in patients who received early tracheostomy (34% versus 16%; *P*= 0.03) whereas spinal cord injuries were less common (3% versus 16%; *P*= 0.08). The proportions of percutaneous and surgical tracheostomies were not different between the early and late groups.
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Baseline patient characteristics
:::
Tracheostomy ≤ 7 days Tracheostomy \>7 days *P*
-------------------------------------- ----------------------- ----------------------- ------
Number 29 107
Age (years) 33 ± 3 31 ± 1 0.5
Male sex (%) 26 (90%) 98 (92%) 0.75
APACHE II score 20 ± 1 19 ± 1 0.35
SAPS II score 42 ± 2 39 ± 1 0.36
ISS score 33 ± 2 34 ± 1 0.79
GCS score 5.2 ± 0.5 6.5 ± 0.4 0.04
Type of injury (*n*\[%\])
Head 20 (69%) 66 (62%) 0.47
Maxillofacial 10 (34%) 17 (16%) 0.03
Chest 11 (38%) 51 (48%) 0.35
Abdomen 3 (10%) 14 (13%) 0.69
Spinal cord 1 (3%) 17 (16%) 0.08
Pelvic/lower extremities 10 (34%) 40 (37%) 0.77
Percutaneous tracheostomy (*n*\[%\]) 21 (72%) 75 (70%) 0.81
Values are expressed as mean ± standard error of the mean, where appropriate. APACHE, Acute Physiology and Chornic Health Evaluation; GCS, Glasgow Coma Scale; ISS, Injury Severity Score; SAPS, Simplified Acute Physiology Score.
:::
Tracheostomy timing and main outcomes
-------------------------------------
Table [2](#T2){ref-type="table"} shows tracheostomy timing data and main outcomes. Extubation trials were performed in 22% of patients with late tracheostomy as compared with 3% of those with early tracheostomy (*P*= 0.019). After placement of the tracheostomy, both groups were weaned off mechanical ventilation and discharged from the ICU after similar periods. Early tracheostomy was associated with a significantly shorter duration of mechanical ventilation (9.6 ± 1.2 days versus 18.7 ± 1.3 days; *P*\< 0.0001) and shorter ICU LOS (10.9 ± 1.2 days versus 21.0 ± 1.3 days; *P*\< 0.0001). Hospital LOS, ICU mortality and hospital mortality were not different between the two groups.
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Main findings
:::
Tracheostomy ≤7 days Tracheostomy \>7 days *P*
------------------------------------------------- ------------------------ ------------------------- ----------
Ventilation days before tracheostomy 4.6 ± 0.5 (6, 2.5--7) 13.9 ± 0.5 (13, 10--16) \<0.0001
Days from ICU admission to tracheostomy 4.6 ± 0.5 (6, 2.5--7) 14.1 ± 0.5 (13, 11--17) \<0.0001
Number (%) of patients with extubation trials 1 (3%) 24 (22%) 0.019
Days from tracheostomy to weaning 4.9 ± 1.2 (2, 1--7) 4.9 ± 1.1 (1, 1--4) 1.0
Days from tracheostomy to ICU discharge 6.3 ± 1.3 (4, 2--8.5) 6.9 ± 1.1 (3, 2--7) 0.72
Total duration of mechanical ventilation (days) 9.6 ± 1.2 (8, 6--13) 18.7 ± 1.3 (15, 12--20) \<0.0001
ICU LOS (days) 10.9 ± 1.2 (10, 7--14) 21.0 ± 1.3 (17, 14--23) \<0.0001
Hospital LOS (days) 101 ± 19 (68, 33--139) 105 ± 7 (83, 54--136) 0.84
ICU mortality (*n*\[%\]) 1 (3%) 1 (1%) NS
Hospital mortality (*n*\[%\]) 5 (17%) 15 (14%) 0.66
Values are expressed as mean ± standard error of the mean (median, interquartile range), where appropriate. ICU, intensive care unit; LOS, length of stay.
:::
Figure [1](#F1){ref-type="fig"} shows the distribution of patients by timing of tracheostomy and the mean ICU LOS for patients, stratified by timing of tracheostomy. There was a direct correlation between the timing of tracheostomy and mean ICU LOS (r = 0.91; *P*\< 0.001). Figures [2](#F2){ref-type="fig"} and [3](#F3){ref-type="fig"} show Kaplan--Meier curves of the duration of mechanical ventilation and ICU LOS in the two groups. Similarly, both the duration of mechanical ventilation and ICU LOS were significantly shorter in the early tracheostomy group (log rank *P*value \< 0.001 for both).
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Distribution of patients by timing of tracheostomy and corresponding intensive care unit (ICU) length of stay (LOS). There was a direct correlation between timing of tracheostomy and mean ICU LOS (r = 0.91; *P*\< 0.001).
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Kaplan--Meier curves of the duration of mechanical ventilation in early and late tracheostomy groups. Early tracheostomy was associated with a significantly shorter duration of mechanical ventilation.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Kaplan--Meier curves of intensive care unit (ICU) length of stay (LOS) in early and late tracheostomy groups. Early tracheostomy was associated with a significantly shorter ICU LOS.
:::
:::
Using univariate analysis the following factors were found to be associated with prolonged ICU stay (\>14 days): late tracheostomy (OR 7.7, 95% CI 3.0--19.9; *P*\< 0.001), spinal cord injury (OR 6.1, 95% CI 1.3--27.7; *P*= 0.019) and extubation trials (OR 3.1, 95% CI 1.1--8.7; *P*= 0.037). The presence of head injury was a significant negative predictor of prolonged ICU stay (OR 0.5, 95% CI 0.2--1; *P*= 0.047), as was the presence of maxillofacial bone injuries (OR 0.4, 95% CI 0.2--1.01; *P*= 0.052). APACHE II score, ISS and GCS score exhibited no significant association with prolonged ICU stay. Using multivariate analysis, late tracheostomy (OR 6.9, 95% CI 2.6--18.1; *P*\< 0.001) and, to a much lesser extent, spinal cord injury (OR 4.7, 95% CI 0.99--22.6; *P*= 0.052) emerged as independent predictors of prolonged ICU stay.
Discussion
==========
In our study we found that early tracheostomy in trauma ICU patients was associated with a significant reduction in the duration of mechanical ventilation and ICU LOS without affecting patient outcome. Weaning patients from mechanical ventilation and discharge occurred shortly and in similar periods after tracheostomy in both groups, suggesting that tracheostomy was a critical factor in weaning and discharge. We also found that late tracheostomy was an independent predictor of prolonged ICU stay.
The study also showed that tracheostomy was more likely to be performed early in patients with maxillofacial fractures, reflecting the need for this procedure for airway management. In patients with spinal cord injury tracheostomy was more likely to be performed late because many of these patients had to undergo surgical spinal fixation before tracheostomy. In such cases, the surgeons preferred to wait until the surgical wound in anterior spinal fusion was healed before performing the tracheostomy. Patients with early tracheostomy had lower GCS, reflecting the common practice of performing tracheostomies earlier in patients with low GCS while delaying tracheostomy in patients with higher GCS in case extubation becomes possible.
The very low mortality seen in the patients we studied may be explained by selection of proper candidates for tracheostomy, excluding those patients who were unlikely to survive. Hospital LOS in these patients was prolonged, reflecting their severe injuries that required lengthy rehabilitation periods. The very limited rehabilitation facilities meant that the patients had to undergo rehabilitation while they were hospitalized, prolonging further the hospital LOS.
Table [3](#T3){ref-type="table"} summarizes studies that examined the impact of early tracheostomy on resource utilization \[[@B2],[@B3],[@B14]-[@B18]\]. All of these studies, except one \[[@B2]\], found reduction in the duration of mechanical ventilation, ICU LOS and/or hospital LOS. Some of these studies found reduction in ventilator-associated pneumonia or colonization incidence. Some of the studies \[[@B3],[@B14]-[@B16],[@B18]\] were retrospective, and all found a positive impact of early tracheostomy on duration of mechanical ventilation, ICU LOS, hospital LOS, or pneumonia rates. The study by Rodriguez and coworkers \[[@B17]\] was a prospective randomized trial in which patients were assigned to early tracheostomy (≤7 days) if they were admitted on an odd day and to late tracheostomy if admitted on an even day. That study found a reduction in duration of mechanical ventilation, ICU LOS and hospital LOS. Sugerman and coworkers \[[@B2]\] conducted a \'multicenter\' randomized trial in five centres involving patients with head trauma, nonhead trauma and no trauma. Those investigators randomized patients on days 3--5 to receive tracheostomy or to continue with translaryngeal intubation. A second randomization for patients who remained intubated was performed on days 10--14. Those authors found no differences in ICU LOS or frequency of pneumonia between early and late tracheostomy. However, the study had several limitations. Out of the five participating centres, only one completed the study. Out of 157 eligible patients, only 112 completed the study because of physician bias and incomplete information. Only 14 patients entered the second randomization. That report illustrates the difficulty in performing studies that challenge widely accepted beliefs. Reviewing these studies also illustrates the lack of consensus regarding the definition of early tracheostomy, with different cutoff points used ranging between 3 and 14 days.
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Literature review
:::
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Ref. Type of study Number of patients Reason for admission Timing of tracheostomy Main outcomes
---------- ------------------------------------ -------------------- ----------------------------------------------------- -------------------------------------- --------------------------------------------------------------------------------------------------------
\[3\] Retrospective 101 Blunt multiple trauma Early tracheostomy ≤4 days\ ↓Duration of MV, ↓incidence of nosocomial pneumonia
Late Tracheostomy \>4 days
\[14\] Retrospective 31 Head trauma Early tracheostomy ≥7 days\ ↓Duration of MV, ↓hospital LOS, ↓ICU LOS
Late tracheostomy \>7 days
\[15\] Retrospective 118 Multiple trauma Early tracheostomy ≤3 days\ ↓Incidence of pneumonia
Intermediate tracheostomy 4--7 days\
Late tracheostomy \>7 days
\[18\] Retrospective 157 Blunt trauma Early tracheostomy ≤6 days\ ↓Duration of MV, ↓ICU LOS, ↓hospital LOS, ↓hospital charges
Late tracheostomy \>6 days
\[16\] Retrospective 30 Neurosurgical (CVA, head injury, trauma, infection) Early tracheostomy ≤7 days\ ↓Duration of MV, ↓incidence of colonization, ↓faster recovery from pneumonia
Late tracheostomy \>7 days
\[17\] Prospective randomized 106 Multiple trauma Early tracheostomy ≤7 days\ ↓Duration of MV, ↓ICU LOS, ↓hospital LOS, ↓pneumonia if tracheostomy was performed earlier than 3 days
Late tracheostomy \>7 days
\[2\]^a^ Prospective randomized multicentre 157 eligible\ Head-trauma, Nonhead trauma, no trauma First randomization: 3--5 days\ No difference in ICU LOS, frequency of pneumonia, or death
patients Second randomization: 10--14
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
^a^Of five participating centres, only one completed the study; of 157 eligible patients, only 112 completed the study because of physician bias and incomplete information; and only 14 patients entered the second randomization. ICU, intensive care unit; LOS, length of stay; MV, mechanical ventilation.
:::
Strengths of our study include prospective data collection ensuring complete data and the relatively large number of patients. However, data extraction and analysis was retrospective. Because the database was not designed specifically to examine tracheostomy practices, certain issues were not documented, such as when the decision for tracheostomy was made and how different intensivists and surgeons varied in their timing of tracheostomy. In addition, the study was observational and was conducted from one centre. A large multicentre randomized controlled trial in which patients are randomized to early versus late tracheostomy would be the ideal way to test the impact of procedure timing on resource utilization.
In summary, the present study, in addition to the existing literature, indicates that early tracheostomy is associated with reduced ICU LOS. Adopting a standardized strategy may help in improving resource utilization. In addition, there is an urgent need for a multicentre randomized controlled trial to assess the most appropriate timing for tracheostomy.
Key messages
============
• Early tracheostomy in trauma ICU patients was associated with shorter duration of mechanical ventilation and ICU LOS without affecting ICU or hospital outcomes.
• There was a direct correlation between timing of tracheostomy and ICU LOS.
• Using multivariate analysis, late tracheostomy emerged as an independent predictor of prolonged ICU LOS.
Competing interests
===================
None declared.
Abbreviations
=============
APACHE = Acute Physiology and Chronic Health Evaluation; CI = confidence interval; ICU = intensive care unit; ISS = Injury Severity Score; GCS = Glasgow Coma Score; LOS = length of stay; OR = odds ratio.
|
PubMed Central
|
2024-06-05T03:55:54.823186
|
2004-8-23
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065024/",
"journal": "Crit Care. 2004 Aug 23; 8(5):R347-R352",
"authors": [
{
"first": "Yaseen",
"last": "Arabi"
},
{
"first": "Samir",
"last": "Haddad"
},
{
"first": "Nehad",
"last": "Shirawi"
},
{
"first": "Abdullah",
"last": "Al Shimemeri"
}
]
}
|
PMC1065025
|
Introduction
============
Of drowning and near drowning victims who are younger than 20 years, 63--68% are 0--5 years old \[[@B1],[@B2]\]. Of submersion events in the age group 1--4 years, 56% occurred in artificial pools \[[@B3]\]. Death from drowning is the second leading cause of accidental death in children \[[@B4]\], and one-third of all survivors have neurological damage \[[@B4]\]. Hypothermia frequently accompanies submersion accidents, especially in children with a relatively large ratio of surface area to body mass \[[@B3]\]. Mild hypothermia (32--34°C) reduces oxygen consumption by 7% per 1°C decrease in temperature, and reduces cerebral blood flow and cerebral intracranial pressure \[[@B5]-[@B7]\]. Temperature under 28°C leads to cardiocirculatory depression and finally cardiac arrest \[[@B3]\]. Hypoxaemia and capillary leak develop due to apnoea, regardless of whether aspiration occurs \[[@B3]\]. The degree of cerebral protection that can be expected due to hypothermia depends, among other factors, on the amount of time that elapses before induction of mild hypothermia \[[@B1],[@B3],[@B6]\]. Induced mild hypothermia for cerebral protection after near drowning accidents has yielded controversial results in terms of mortality and neurological outcome \[[@B1],[@B3],[@B8]\]. However, induced mild hypothermia after cardiac arrest has led to improved neurological results, whereas life-threatening complications such as infections and resultant sepsis may counter these neurological benefits \[[@B9]\].
We report here a case of twins who both suffered near drowning, but with different post-trauma treatment and different neurological outcomes.
Case report
===========
The twins (a girl and boy, aged 2 years and 3 months old) were found lifeless by their father in the neighbours\' garden pond. It was early spring, and the toddlers had been unattended for at least 10 min. Bystander cardiopulmonary resuscitation (CPR) was performed. The emergency doctor could not palpate any pulse, the children were hypothermic, and the pupils were dilated and pupil reflexes absent. Both children had aspirated. Under CPR the children exhibited pulseless bradycardia on the electrocardiogram.
The girl
--------
The girl was transported to a university hospital. Admission parameters are presented in Table [1](#T1){ref-type="table"}. After rewarming to 32°C and successful CPR, 180 min after admission to the hospital, haemodynamic stability was achieved with adrenaline (epinephrine) infusion and the child was admitted to the intensive care unit (ICU). The pupils were slightly dilated with reaction to light and the corneal reflex was absent. Cranial computed tomography (CT; Fig. [1](#F1){ref-type="fig"}), done 7 hours after admission, revealed cerebral oedema; this was regressive, as indicated by cranial CT obtained 3 days later. Mannitol therapy and prolonged mild hypothermia (32--34°C) were begun the day of the accident. Repeated fundoscopy did not show signs of papillary congestion. Intracranial pressure was not monitored. Under sedation with fentanyl and midazolam to a Ramsay level of 6 and controlled mild hyperventilation (arterial CO~2~tension 30--35 torr), mild hypothermia was continued and reduced gradually (0.5°C/8 hours). Seventy-two hours after the accident the child was normothermic without development of rebound hyperthermia. After rewarming the pupils were tight and reflexes present. Catecholamine therapy on admission to the ICU was switched to dobutamine and dopamine infusion. To achieve a mean arterial pressure greater than 70 mmHg, noradrenaline (norepinephrine) infusion had to be added.
Under pressure controlled ventilation the oxygenation index improved initially and the inspiratory oxygen fraction could be reduced to 0.3 over the first 48 hours after the accident. However, 72 hours after the accident oxygenation deteriorated. The initial CT of the thorax had shown infiltrations in the basal dorsal thorax after aspiration (Fig. [2](#F2){ref-type="fig"}). The following chest X-ray films revealed increasing bilateral infiltrations of the lung (Fig. [3](#F3){ref-type="fig"}). After 3 days in the ICU, sepsis with multiple organ failure developed (acute respiratory dysfunction syndrome \[ARDS\] with an oxygenation index of 109 torr, circulatory failure requiring catecholamines, liver dysfunction with increased transferases and reduced prothrombin time, and disseminated intravascular coagulopathy). Substitution of blood products was necessary. Acute renal failure did not develop. Antibiotic treatment was started (ceftazidime for *Pseudomonas aerguinosa*in the tracheal aspirate, vancomycin for *Enterococcus faecium*at the central venous catheter tip).
Under differentiated pressure controlled ventilation, oxygenation did not improve. Because it was unclear at this time whether extracorporeal membrane oxygenation would be required, 5 days after the accident the child was transferred by helicopter to another university hospital because of limited capacity at our hospital. Under high-frequency oscillatory ventilation and nitric oxide inhalation, oxygenation improved and extracorporeal membrane oxygenation was not necessary. Conventional pressure controlled ventilation could be restored 7 days after the accident, and at the same time the multiple organ failure improved. Sedation was reduced and the girl was extubated 11 days after the accident, with no neurological deficit. Twenty-three days after the accident she was transferred to the community hospital where her brother was initially hospitalized, and she was discharged 1 day later completely restored to health.
The boy
-------
The brother was transported to a community hospital. Admission parameters are presented in Table [1](#T1){ref-type="table"}. Haemodynamic stability was achieved 150 min after admission to the hospital under dopamine and dobutamine therapy. The pupils were slightly dilated with reaction to light and the corneal reflex was present. He was rewarmed and normothermia was achieved 5 hours after admission. Continuous catecholamine therapy was stopped 4 days after the accident. The boy was sedated with fentanyl and midazolam, and ventilated to achieve normocapnia using a pressure-controlled mode. With improvement in oxygenation, he was extubated 6 days after the accident. The initial chest X-ray films showed bilateral infiltrations of the lung as a sign of aspiration pneumonia, which improved within the next few days. Liver and kidney function remained normal.
After the end of sedation, an apallic syndrome with extension posturing developed. The initial cranial CT obtained 36 hours after admission was normal, and fundoscopy did not show signs of papillary congestion. A cranial CT obtained 32 days after the accident showed marked expansion of the internal and external cerebral fluid interspaces with marked cerebral atrophy. At discharge from hospital, 41 days after the accident, the little boy remained in an apallic state, with flexion and extension posturing.
Discussion
==========
We present a case of twin toddlers with different neurological outcomes after near drowning with severe hypothermia and protracted cardiac arrest. Hypothermia at the scene has yielded controversial results with respect to cerebral protection. Factors such as time to achieve hypothermia (e.g. water temperature), the degree of hypothermia, the time of submersion, and other effects such as cardiocirculatory depression or arrest have various influences on the cerebral protection conferred \[[@B3],[@B8]\]. Some of these factors are unclear in this case report. The two institutional approaches to management of the twins were optimal because both hospitals have paediatric departments with paediatric ICUs. In addition, the community hospital is a training hospital and part of the university hospital.
Lavelle and Shaw \[[@B8]\] described three patients with body temperature under 28°C on arrival at the emergency department. All three patients had a good neurological outcome, but they fell into icy water. The use of prolonged or induced mild hypothermia for cerebral protection after near drowning has yielded controversial results \[[@B4],[@B8]\]. Bohn and coworkers \[[@B6]\] reported on 40 children aged under 15 years who suffered severe near drowning accidents with submersion time longer than 5 min and need for CPR. Twenty-four children were treated with hypothermia (30--33°C) for 24--36 hours, and 14 survived but three of these children had permanent neurological damage. Sixteen children were kept normothermic, and 13 survived but four had permanent neurological damage. Nussbaum and Maggi \[[@B10]\] investigated 31 children aged under 6 years who had undergone near drowning and were in a flaccid state of coma. All children were treated with hypothermia (32--34°C) for 48 hours (half of them received additional barbiturate therapy). Twelve children recovered completely, 12 children had brain damage and seven died.
Two recently published studies, conducted in patients who had suffered out-of-hospital cardiac arrest, compared induced mild hypothermia for 12--24 hours with normothermic management \[[@B7],[@B9]\]; they found that a significantly greater percentage of patients in the groups treated with mild hypothermia had good neurological outcomes. In patients affected by brain injury with a Glasgow Coma Scale score from 3 to 8, induced mild hypothermia for 24--48 hours yielded controversial findings \[[@B11],[@B12]\]. In these patients hypothermia on admission correlated with poor outcome, suggesting that spontaneous hypothermia may be a result of major brain injury \[[@B11]\].
In the present case report, hypothermia on the scene and on admission was probably the result of external factors such as water and air temperature and the children\'s age, suggesting cerebral protection from hypothermia. Up until the arrival of the twins at hospital, the treatment was identical. The boy was passively warmed to achieve normothermia, and the girl underwent prolonged (72 hours) mild hypothermia (32--34°C). The different neurological outcomes could have been influenced by these different treatments. However, some factors remain uncertain. For example, was the boy the first to go into the water, with resulting longer submersion and hypoxaemia times? How effective was bystander CPR in the two children? Was the time to achieve hypothermia the same in both children? Excluding bystander CPR, the remaining factors are considered strong predictors of outcome after near drowning \[[@B1],[@B3],[@B8]\]. The girl developed ARDS and septic shock, whereas the boy recovered from aspiration pneumonia without further complications.
There is concern that prolonged mild hypothermia has adverse effects on cardiac and lung function, coagulation and the immune system \[[@B3],[@B5],[@B7]\]. In a series of 41 patients with submersion injury (temperature on admission \>32°C, no induced mild hypothermia), 32% developed pneumonia and one person ARDS \[[@B8]\]. Significantly higher infection rates, predominantly pneumonia, were described in patients treated with induced mild hypothermia as compared with patients treated under normothermic conditions \[[@B5],[@B13],[@B14]\]. However, other investigations evaluating patients following out-of-hospital CPR and with brain damage did not identify any differences in the incidence of infection between normothermic and hypothermic groups treated. just for 12--24 hours \[[@B7],[@B9],[@B15]\]. It seems posssible that the duration of mild hypothermia has an impact in the incidence of infection and sepsis. Among the 41 normothermic patients described by Lavelle and Shaw \[[@B8]\], after submersion 14% developed sepsis. In experimental animal models it was shown that hypothermia under 29°C leads to a reduced neutrophil response to endotoxin \[[@B16]\]. Leukocytopenia has been described to be significantly more frequent in patients with induced mild hypothermia \[[@B13],[@B14]\].
The girl was highly catecholamine dependent in the first 7 days after the accident. It has been reported that, in patients with mild hypothermia, significantly higher doses of catecholamines are required in comparison with normothermic patients after acute brain injury \[[@B11]\]. Vasopressor requirements have been described as having a significant impact on outcome \[[@B2]\]. The rate of other organ dysfunctions (liver, kidney) has also been found to be significantly higher in patients under induced mild hypothermia. The girl also developed transient liver dysfunction. Together with sepsis syndrome, coagulopathy developed. Disturbances of this system with resultant bleeding complications are known to occur during therapy with mild hypothermia \[[@B5],[@B13],[@B14],[@B17]\].
Conclusion
==========
Although the twin toddlers experienced a near drowning accident together, the outcomes in terms of neurological status and postinjury complications were completely different. One of the factors that possibly influenced the different postinjury courses might have been prolonged mild hypothermia.
Key messages
============
• Two twin toddlers suffered hypothermic near drowning with protracted cardiac arrest and aspiration.
• The girl was treated with mild hypothermia and developed acute respiratory dysfunction syndrome and sepsis, but recovered without neurological deficits.
• The boy was treated under normothermic conditions and developed an apallic syndrome.
• One of the factors that possibly influenced the different postinjury course might have been prolonged mild hypothermia.
Competing interests
===================
None declared.
Abbreviations
=============
ARDS = acute respiratory dysfunction syndrome; CPR = cardiopulmonary resuscitation; CT = computed tomography; ICU = intensive care unit.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
The girl: cranial computed tomography, done 7 hours after admission, showing cerebral oedema.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
The girl: computed tomography of the thorax, done shortly after admission to hospital, showing infiltrations in the basal dorsal thorax following aspiration.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
The girl: chest X-ray film, done after admission to hospital, showing increasing bilateral infiltrations of the lung, done after admission.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Parameters at the scene and on admission in the twins
:::
Site/parameter Girl Boy
--------------------------------------- ---------- ----------
At the scene
Time to bystander CPR (min) \> 10 \> 10
Pulseless bradycardia under CPR Yes Yes
On admission
Pupils dilated, nonreactive to light Yes Yes
Corneal reflex Negative Negative
Temperature on admission (°C) \<28 \<27
pH on admission 6.63 7.00
Arterial CO~2~tension (torr) 38 36
BE on admission (mmol/l) -27 -17
Oxygenation index (torr) 75 65
CPR time (min) 180 120
BE, base excess; CPR, cardiopulmonary resuscitation.
:::
|
PubMed Central
|
2024-06-05T03:55:54.825469
|
2004-9-2
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065025/",
"journal": "Crit Care. 2004 Sep 2; 8(5):R353-R357",
"authors": [
{
"first": "Ortrud Vargas",
"last": "Hein"
},
{
"first": "Andreas",
"last": "Triltsch"
},
{
"first": "Christoph",
"last": "von Buch"
},
{
"first": "Wolfgang J",
"last": "Kox"
},
{
"first": "Claudia",
"last": "Spies"
}
]
}
|
PMC1065049
|
See related commentary <http://ccforum.com/content/8/6/416>
Introduction
============
Critical illness polyneuropathy (CIP) and myopathy are neuromuscular disorders that occur in critically ill patients \[[@B1],[@B2]\]. Clinical features often consist of difficulty in weaning from mechanical ventilation, tetraparesis and muscle wasting of the limbs, with tendon reflexes absent or markedly decreased. Regarding the causes of these disorders, it has been hypothesized that systemic inflammatory response syndrome (SIRS) and sepsis, with their impact on the body\'s defence system, may be involved \[[@B3],[@B4]\]. Associations with drugs such as neuromuscular blocking agents, steroids and catecholamines have also been suggested \[[@B3],[@B5],[@B6]\]. In addition, other factors such as malnutrition, underlying disease, immobility and antibiotics have been considered \[[@B7]\].
Patients with weakness acquired in the intensive care unit (ICU) are often sedated and mechanically ventilated, and have unreliable sensory and motor examinations, and so diagnosis can be quite difficult because of the prolonged sedation. \[[@B8]\] Electromyography is useful for identifying and localizing a lesion to a particular component of the motor unit. Using electromyography, Bolton \[[@B9]\] and Zochodne \[[@B10]\] and their colleagues were the first to identify one of the major neuromuscular causes of neuromuscular weakness in acute ill patients, namely CIP \[[@B9],[@B10]\]. However, use of electromyography and clinical examination without obtaining neuromuscular biopsy findings may lead to some patients being diagnosed as having CIP only \[[@B2],[@B11]\]. Therefore, such examinations may distract attention from other neuromuscular disorders that occur in the critically ill, thus leading to identification of phenomena of neuromuscular junction blockade or critical illness myopathy, or a combined picture involving both syndromes.
Because of this, and because of the lack of studies evaluating neuromuscular disorders in ICUs using a combination of histopathological (nerve and muscle) and electromyographic studies, we conducted the present study in which we monitored critically ill patients prospectively following complicated cardiovascular surgery. After inclusion in the study, serial clinical, neuromuscular biopsy and electromyographic analyses were systematically conducted. Using this method we were able to diagnose neuromuscular disorders; to identify the predominant neurogenic or myogenic pattern, or a combined picture involving both lesions; and to compare these findings with functional outcome.
Methods
=======
Hospital
--------
This prospective study was carried out in the ICU of a teaching hospital with an 11-bed surgical unit. Annually, 650 critically ill patients are admitted following cardiac surgery using cardiopulmonary bypass.
Patients
--------
From 1998 to 2002, 15 patients who had a complicated course (with one or more organ dysfunctions) following cardiovascular surgery were prospectively included after prolonged mechanical ventilation (\>15 days) associated with tetraparesis and failure to wean. Peripheral electromyographic analysis and neuromuscular biopsy were performed in all of these patients to determine the cause of limb weakness and diaphragm dysfunction. The medical committee of the hospital approved the study, and informed consent was obtained from the relatives of the patients.
Excluded were those patients who were suspected of having pre-existing polyneuropathy because they had a diagnosis of chronic diabetes mellitus, alcohol abuse, HIV infection, or end-stage renal disease (associated with chronic haemodialysis), or had used neurotoxic medication. Patients with acute or chronic spinal cord lesion, myasthenia gravis, or Guillain--Barré syndrome were also excluded.
After the patients had been enrolled, clinical examination was performed daily during their stay in the ICU, and we assessed motor deficit, muscle wasting, sensory loss and tendon reflexes.
In these patients undergoing cardiac surgery, at ICU admission the Euroscore was caculated to evaluate risk for postoperative mortality \[[@B12]\]. This prognostic scoring system was developed in Europe for use in patients undergoing cardiac surgery. The score is calculated by simple arithmetic (additive model) using risk factors that were found to be robust in predicting postoperative mortality, such as age, sex, emergency surgery, preoperative left ventricular dysfunction and type of surgery. Organ failure score and presence of SIRS were noted, according to criteria presented by Bone and coworkers \[[@B13]\], at admission and during the clinical course \[[@B7]\]. The diagnosis of SIRS required the presence of two or more of the following criteria \[[@B13]\]: body temperature \>38°C or \<36°C; heart rate \>90 beats/min; tachypnoea \>20 breaths/min; and hyperventilation, as indicated by an arterial carbon dioxide tension \<32 mmHg, leucocyte count \>12 g/l or \<4 g/l, or presence of \>10% immature neutrophils.
Use and dosage of neuromuscular blocking agents and intravenous corticosteroids were registered daily.
Electrophysiological monitoring and neuromuscular biopsies were performed as soon as the neuromuscular disorder was recognized after the end of prolonged sedation associated with mechanical ventilation (\>15 days after the onset of mechanical ventilation). Failure to wean from mechanical ventilation was characterized by inability to extubate or inefficient spontaneous ventilation through tracheostomy. Peripheral weakness affected both proximal and distal muscle groups and was defined as failure to move against gravity.
Electromyography
----------------
Nerve conduction studies were performed with Nicolet Viking IV apparatus via percutaneous stimulation and surface recording. Quantitative concentric needle electromyography was performed in distal and proximal muscles of upper extremities (such as deltoid, muscle interosseus I, and muscle abductor pollicis brevis) and lower extremities (such as tibialis anterior and flexor hallucis brevis). Electroneurography included upper (median or ulnar) nerves and lower (peroneal or sural) nerves. In patients who could not exercise, repetitive stimulation at 20--30 Hz was also given. Diaphragmatic and phrenic nerve studies were not performed.
Electromyography reports were analyzed by the same doctor and categorized as axonal polyneuropathy or myopathy, or a combination thereof. Patients were diagnosed as having axonal sensorimotor polyneuropathy if electrodiagnostic studies revealed very low amplitude or absent sensory responses and low motor amplitudes with normal or mildly reduced conduction velocities. Patients were diagnosed as having myopathy in the setting of low or normal motor amplitudes, with relatively normal sensory responses. Short duration of motor unit potentials with normal or early recruitment with or without fibrillation potentials, or fibrillation potentials and either no firing or polyphasic motor unit potentials of normal duration were also considered to reflect myopathy.
Combined muscle and nerve biopsies
----------------------------------
Muscular biopsies were obtained in all 15 patients, from skeletal muscle specimens of the vastus lateralis or anterior tibialis muscles, under local anaesthesia (lidocaine 1% 100--150 mg) if necessary; these patients did not have thombocytopenia or coagulopathy. Muscle samples were first snap frozen in isopentane precooled in liquid nitrogen and stored at -80°C until examination. For routine histology, the samples were placed in formaldehyde fixative and paraffin embedded. For conventional transmission electron microscopy, specimens were fixed in 2.5% glutaralehyde in 0.1 mol/l phosphate-buffered saline, postfixed with 1% osmic acid and embedded in araldite. Semithin resin sections were stained using toluidine blue. Ultrathin sections were double stained with uranylacetate. Histoenzymology was performed in serial transverse cryostat sections (6 µm thick), stained using routine histochemical methods \[[@B14]\].
Sensitive nerve biopsies were obtained from sural, peroneal nerves or the sensory branch of the musculocutaneous nerve (in the distal third of the leg). Nerve samples were fixed in 2.5% glutaraldehyde in 0.1 mol/l phosphate-buffered saline, postfixed with 1% osmic acid and embedded in araldite. Semithin resin transverse and longitudinal sections were stained using haematein and eosin, solochrome blue and paraphenylene diamine. Ultrathin sections were double stained with uranylacetate and citrate. Histopathological reports and original slides were reviewed by two medical experts who were blinded to the electromyographic findings.
Functional outcome
------------------
Follow-up data were available in all 15 patients. The end-points were death or time to ambulation without assistance. The maximal duration of follow up was 12 months. Findings were not analyzed statistically because of the relatively small numbers included in the various groups.
Results
=======
Clinical features
-----------------
A total of 25 patients, suffering in most cases from sepsis or SIRS following cardiac surgery, and who were undergoing long-term mechanical ventilation (\>15 days), were enrolled. Ten patients were excluded because of previous alcoholic liver disease or end-stage renal failure (*n*= 4) or a previous history of neuromuscular disease (*n*= 6).
The patients\' median age was 53 years (range 33--82 years), and the median Euroscore was 7 (range 1--20). All patients presented with at least one episode of sepsis or a systemic inflammatory response. Multiorgan dysfunction was diagnosed in 10 patients, with a median Multiple Organ Dysfunction Score of 3 (range 1--4; Table [1](#T1){ref-type="table"}). The most common organ system failure was cardiovascular, and more than 86% of patients fulfilled criteria for heart failure. Postoperative renal dysfunction requiring continuous venovenous haemofiltration was diagnosed in nine patients (60%). The severity of weakness and variability in reflex abnormality were noted. Patients were noted to be weak 15--40 days after ICU admission. Because patients received neuromuscular blocking agents and sedatives, the exact time of onset of weakness was usually not possible to determine. Of the patients studied, 50% presented with asymmetric tetraparesia, predominantly involving the legs. The other 50% had global tetraparesia with much reduced muscle tone (Table [1](#T1){ref-type="table"}). Five patients were areflexic, seven had hyporeflexia and three had normal tendon reflexes. All patients grimaced in response to painful stimuli, but sensory testing was initially unreliable in most. One patient was transiently encephalopathic.
Table [2](#T2){ref-type="table"} summarizes clinical diagnoses and medications affecting the neuromuscular system in these patients with neuromuscular weakness syndromes. Patients in all groups received muscle relaxants; only two were used in our unit during the study period -- vecuronium and atracurium. The groups included patients who received large doses of muscle relaxants as well as patients who received none. Muscle relaxants were used in 66% of the patients included in the study; 20% of these received muscle relaxants for less than 24 hours. Six patients received intravenous corticoids; four of these patients had undergone transplants and the other two patients were asthmatic.
Electrodiagnostic testing
-------------------------
The spectrum of neuromuscular causes of weakness, along with electromyographic localization and neuromuscular biopsy findings, is summarized in Table [2](#T2){ref-type="table"}. The two most common causes of weakness in these patients were polyneuropathy or myopathy in isolation. Six patients presented with neuropathy, three of whom had sensory motor neuropathy characterized by reduced sensory and motor action potential amplitudes. Six other patients presented with acute myopathy, characterized by sensory and motor action potentials in the normal range or partially reduced. Three patients presented with CIP associated with myopathy. Motor unit potentials were usually reduced, with pathological fibrillation potentials associated.
Muscle and nerve pathology
--------------------------
A range of histopathological abnormalities was identified in the neuromuscular biopsies from the 15 patients (Table [2](#T2){ref-type="table"}). The most common muscle abnormality was diffuse atrophy of fibre types I and II (14 out of 15 patients) associated with acute necrosis (12 out of 15 patients; Figs [1](#F1){ref-type="fig"} and [2](#F2){ref-type="fig"}). There were abnormalities in the nerve biopsy from most patients (Fig. [3](#F3){ref-type="fig"}; except those with acute myopathy); these included axonal degeneration and demyelinating lesions. In some patients (patients 5, 7 and 8), analysis of neuromuscular biopsy revealed myopathic lesions associated with necrotic and atrophic fibres. This was associated with a huge reduction in sensory or motor action potential amplitude. Ten patients received neuromuscular blocking agents, five of whom exhibited an acute myopathy and the other five a peripheral neuropathy. However, two patients who did not receive any muscle relaxant developed a neuropathic pattern (with or without myopathic lesions). All patients who received corticosteroids exhibited a myopathic pattern (with or without axonopathic lesions) but never an exclusively axonopathic pattern.
The overall mortality rate was 40%, with a further patient dying on the ward after ICU discharge, giving a hospital mortality of 46%. Among the eight survivors at 12 months, two patients were not ambulatory (patients 2 and 8; Table [3](#T3){ref-type="table"}). Compound muscle action potentials (CMAPs) were undetectable in these two patients on electrophysiological examination (Table [2](#T2){ref-type="table"}).
Discussion
==========
Although acute neuromuscular weakness appears prevalent among patients on prolonged mechanical ventilation, few prospective studies have been reported that include both electrophysiological and histological patterns \[[@B5],[@B15],[@B16]\]. Furthermore, nerve biopsy findings are absent, even in recent prospective studies \[[@B5]\]. Our study is among the first to perform electromyography and obtain neuromuscular biopsies prospectively for all patients included. The neurophysiological abnormalities identified were of three types, namely CIP alone, acute myopathy and mixed neurogenic and myogenic disturbances, and they developed in a group of long-term mechanically ventilated patients who had undergone cardiovascular surgery. This type of severe and disturbing complication following cardiac surgery has been described previously \[[@B6],[@B17]\], and led to an evaluation of hypothetical risk factors for such neurological disorders.
It is also widely believed that the development of critical illness neuropathy is invariably associated with multiple organ failure, sepsis and SIRS \[[@B3],[@B6],[@B8],[@B9]\]. Thus, CIP probably represents an organ failure caused by sepsis and SIRS, presumably as a result of the same basic mechanisms that lead to multiple organ dysunction, including inflammation, thrombosis, apoptosis and oxidant injury \[[@B18]\]. In the present study, however, peripheral neurological changes occurred in a few patients who did not fulfill accepted objective criteria for sepsis or single organ failure. This observation was previously reported in four series of patients with respiratory failure \[[@B7],[@B15],[@B19],[@B20]\]. As might be expected, in the present study of long-term mechanically ventilated patients following cardiovascular surgery, sepsis and multiple organ failure were common but did not seem to be a prerequisite for the development of acute neuromuscular weakness, the cause of which remains unclear \[[@B18]\].
It was previously suggested that use of neuromuscular relaxants is associated with neuromuscular disorders in the ICU \[[@B21]-[@B24]\]. Possible mechanisms include persistent effects of these drugs or their active metabolites, pharmacological denervation hastening muscle atrophy, or association of these drugs with intravenous corticosteroids or aminoglycosides \[[@B22],[@B24]\]. Vecuronium and its steroid components were also implicated as a cause of weakness \[[@B25]\]. In our study atracurium was also administered in patients with prolonged neuromuscular weakness, although it has no steroidal component and there is no accumulation of this molecule in the event of kidney or liver failure. However, we are unable to conclude that neuromuscular relaxants predispose to the development of neuromuscular disorders in general, or any type of neuromuscular disease in particular, because of the lack of a control group. Other observational studies failed to identify neuromuscular relaxants as possible additional risk factors \[[@B20],[@B31]\].
The link between use of aminoglycosides and CIP was previously reported \[[@B26],[@B27]\]. In the present study 66% of patients presenting with CIP received intravenous aminoglycosides, but only five patients out of 15 received intravenous aminoglycosides. Therefore, intravenous use of aminoglycosides may be another measure of severity of sepsis and multiple organ failure.
Steroid administration appears to be associated with muscular lesions, regardless of association with neuropathy. All study patients who received high doses of corticosteroids (cumulative equivalent dose \>1000 mg methylprednisolone) exhibited a myopathic pattern (with or without associated axonopathic lesion) but never an exclusive axonopathic pattern. This finding supports the deleterious effect of corticosteroids predominantly on the muscles, as suggested by several studies \[[@B5],[@B28]-[@B32]\].
Electromyographic abnormalities and neuromuscular histology patterns were concordant in 13 patients out of 15. In the two discordant cases, electromyographic examination failed to show any myogenic component, whereas muscle histology suggested severe myopathy with necrotic fibres and vacuolization zones. This lack of complete agreement between neurophysiological testing and muscle histology has already been noted by Coakley and colleagues \[[@B20]\]. Some authors have also indicated that it could be difficult to differentiate myopathy from axonal motor neuropathy via electromyographic analysis alone, especially in unconscious patients \[[@B2],[@B5]\]. Thus, in the absence of systematic muscle biopsy, some patients can be misdiagnosed with myopathy when motor axonopathy is present and, as shown in the present study, it is conceivable that some patients have both myopathy and neuropathy \[[@B4]\]. The combination of neurophysiological testing with neuromuscular histology could therefore help in the precise identification of the type and severity of neuromuscular weakness, and may lead to a better understanding of the causes and consequences of neuromuscular weakness \[[@B5]\].
There was a high mortality rate in patients with acute neuromuscular weakness. Patients who died did so as a result of their underlying diseases and not from neuromuscular affection. These findings are similar to those from previous studies that reported on neuromuscular abnormalities \[[@B4],[@B33]\]. In fact, the mortality rate and functional prognosis were similar between survivors with acute myopathy and those with neuropathy. However, of the three patients with combined neurological and muscular lesions, two died and the third survived but with severe functional disability.
Survivors from critical illness have sustained impairments in physical function and health status, even after 1 year of recovery \[[@B34]-[@B36]\]. In our study, of the eight survivors at 12 months only two were not ambulatory (patients 2 and 8). These were the only two survivors in whom CMAPs were undetectable on electrophysiological examination. This electrophysiological finding attests to the severity of the axonopathy and may be a predictor of prolonged functional disability. However, because of the relatively small number of patients in this group, additional studies are necessary to confirm this clinical finding.
Certain limitations of present study are worthy of mention. This prospective study evaluated electrophysiological and histological neuromuscular patterns in 15 ICU patients with prolonged mechanical ventilation after a complicated course following cardiovascular surgery. However, because the lack of control group and the small numbers of patients included, we were not able to determine precisely the risk factors for each pathology. Furthermore, as in previous studies \[[@B20],[@B37]\], we were unable to determine exactly the time of onset of weakness. Therefore, electrophysiological and histological analyses were not performed at the same time point for all patients.
Conclusion
==========
Neurophysiological and neuromuscular histological abnormalities associated with acute neuromuscular weakness were identified in mechanically ventilated patients in the ICU who had undergone cardiovascular surgery. Such patients are assumed to present with reversible neurological damage, although in a proportion this damage could be irreversible. Among survivors the absence of CMAPs on electrophysiological examination could suggest prolonged functional disability. Because of the lack of sensitivity of clinical examination in such patients, combined electromyographic diagnosis and neuromuscular abnormalities on histology could help to identify the type and severity of neuromuscular weakness, and the functional prognosis.
Key messages
============
• In patients undergoing neuromuscular weakness syndrome following cardiovascular surgery, the combination of electromyographic evaluation and neuromuscular histological abnormalities could help identify type and severity of these neuromuscular weakness, in turn helping to evaluate more precisely the patient\'s functional prognosis.
Competing interests
===================
None declared.
Abbreviations
=============
CIP = critical illness polyneuropathy; CMAP = compound muscle action potential; ICU = intensive care unit; SIRS = systemic inflammatory response syndrome.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Cryostat cross-section of a biopsy specimen from patient 10 from the musculus vastus lateralis with acute necrotizing myopathy. Necrotic vacuolated and regenerating muscle fibres are present. Endomysial connective tissue is increased. Frozen section, stained with haematein and eosin, magnification 200×.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Electron microscopic changes in muscular fibres associated with a pronounced myosin heavy chain depletion. Near total loss of thick filaments is seen in the A band (barr = 2 µm).
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Cross-section of a nerve biopsy specimen from patient 4 exhibiting severe axonal neuropathy. There is loss of myelinated nerve fibres. Some large myelinated fibres show degenerating myelin ovoids. Secondary demyelination is seen in rare nerve fibres. Clusters of Schwann cells without nerve fibres are increased. Resin section, stained with paraphenylene diamine, magnification 115×.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Summary of diagnoses and medications affecting the neuromuscular system in 15 patients with neuromuscular weakness syndromes
:::
Patient Age (years) Sex Euroscore Primary disease Surgery Duration of MV (days) Complications SIRS Sepsis MOD score Neurological presentation^a^ Neuromuscular blocking medication Corticosteroids
--------- ------------- ----- ----------- --------------------------------------- ------------------- ----------------------- ------------------------------------------ ------ -------- ----------- ------------------------------------- ---------------------------------------- --------------------------------------- ----------------------------
1 59 M 3 Coronary artery disease CABG 70 Pneumonia, kidney failure Yes Yes 2 Flaccid TP, stupor Atracurium 1830 mg, vecuronium 72 mg Tobramycin 240 mg, gentamycin 1120 mg None
2 44 M 1 Obesity, asthma CABG 52 Pneumonia Yes Yes 1 Flaccid TP Atracurium 1980 mg, vecuronium 1640 mg None Methylprednisolone 240 mg
3 50 F 7 Active endocarditis VR 24 RV failure, kidney failure Yes No 2 Flaccid TP None None None
4 76 M 6 Coronary artery disease, hypertension CABG 51 ARDS, septic shock Yes Yes 4 Flaccid TP None Amikacin 3200 mg None
5 67 M 14 Coronary artery disease CABG 40 ARDS, kidney failure Yes Yes 3 Flaccid TP Atracurium 1380 mg None None
6 33 M 0 Bicuspid aorta Ross procedure 25 Haemorrhagic shock, kidney failure, ARDS Yes Yes 4 Flaccid TP, distal amyotrophy Atracurium 3000 mg None None
7 76 F 8 Hypertension, endocarditis VR 163 Septic shock, kidney failure Yes Yes 3 Flaccid TP None Gentamycin 1000 mg None
8 60 M 6 Coronary artery disease HT 36 Cardiac arrest, pneumonia Yes Yes 2 Flaccid TP, diffuse amyotrophy None None Methylprednisolone 2000 mg
9 60 M 20 Coronary artery disease HT 120 Mucormycosis, pneumonia, kidney failure Yes Yes 4 Flaccid TP, diffuse amyotrophy, ROT Atracurium 300 mg Gentamycin 600 mg Methylprednisolone 3000 mg
10 53 M 6 Coronary artery disease HT 54 Pneumonia, kidney failure Yes Yes 4 Flaccid TP Atracurium 100 mg Tobramycin 820 mg Methylprednisolone 8460 mg
11 57 M 9 Cardiomyopathy HT 28 RV failure, pneumonia, kidney failure Yes Yes 3 Flaccid TP, ROT None None Methylprednisolone 3000 mg
12 26 F 6 None RV blast 64 Cardiac arrest, encephalopathy Yes No 3 Flaccid TP, ROT Atracurium 11,000 mg None None
13 77 F 10 Mitral regurgitation, endocarditis VR 45 Septic shock, kidney failure Yes Yes 3 Flaccid TP Atracurium 630 mg None None
14 77 F 11 Hypertension, asthma Aortic dissection 34 Pneumonia, kidney failure Yes Yes 2 Flaccid TP, ROT Atracurium 7300 mg None Methylprednisolone 1120 mg
15 82 F 10 Hypertension Aortic dissection 36 Pneumonia, kidney failure Yes Yes 2 Flaccid TP, ROT Atracurium 3300 mg None None
^a^Neurological presentation after stopping sedation (approximately 72 hours). CABG, coronary artery bypass grafting; HT, heart transplantation; MOD, multiple organ dysfunction; MV, mechanical ventilation; PP, paraparesia; TP, tetraparesia; ROT, tendon reflexes abolished; RV, right ventricular; SIRS, systemic inflammatory response syndrome; VR, valve replacement.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Electrodiagnostic and histopathological findings in 15 patients with neuromuscular weakness syndromes
:::
Patient Type^a^ Electromyography Muscular biopsy Nerve biopsy
--------- --------- ------------------ ----------------- -------------- ---- ---- ---- ----------------------- --------- ----------------------- ------- ------ ---- ---- ---- ----
1 N \+ \+ Peroneal 46/ulnar 68 1/0.1 I/II \+
2 N \+ \+ Peroneal 0/median 0 0/0 Sural 0/radial 0 0/0 II \+ \+
3 N \+ \+ \+ Peroneal 33 0.1 Peroneal 41 10 I \+ \+
4 N \+ \+ Peroneal 26/median 37 1/1 Peroneal 35/median 38 2/1 I/II \+ \+
5 N \+ \+ \+ Peroneal 41 3 Peroneal 33 4 I/II \+ \+ \+
6 N \+ \+ \+ Peroneal 40 2 Peroneal 36/radial 48 7/48 I/II \+ \+ \+
7 NM \+ \+ Peroneal 0 0 Peroneal 0 0 I/II \+ \+ \+
8 NM \+ \+ Peroneal 0 0 Peroneal 0 0 I/II \+ \+ \+
9 NM \+ \+ \+ Peroneal 38 0.1 Peroneal 35/radial 48 3/8 I/II \+ \+ \+
10 M \+ \+ \+ \+ Peroneal 36/median 34 0.8/0.8 Peroneal 52/radial 56 7/16 II \+ \+ No No
11 M \+ Peroneal 42 1.6 Peroneal 58 12 II \+ No No
12 M \+ \+ Peroneal 35 1.2 Peroneal 43/radial 43 20/22 I/II \+ No No
13 M \+ Peroneal 39 0.8 Peroneal 42 24 II \+ No No
14 M \+ \+ \+ \+ Peroneal 43 1.6 Peroneal 44 12 II \+ \+ No No
15 M \+ \+ Peroneal 40 1.0 Peroneal 42 7 II \+ No No
^a^Type of neurological lesion. ^b^Recruitment (muscular contraction): 0, absent; 1, simple; 2, weak intermediate; 4, rich intermediate. Recruitment data not available for patients 4, 8 and 12. CMAP, compound muscle action potential; CV, conduction velocity; FP, fibrillation potentials; LP, slow potentials of denervation; M, myopathy; N, neuropathy; NM, neuromyopathy.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Clinical outcome in 15 patients with neuromuscular weakness syndrome
:::
Patient Type^a^ ICU stay (days) Clinical outcome Ambulatory activity without assistance in \<3 months Ambulatory activity without assistance in 3--12 months Nonambulatory Mechanical ventilation^b^
--------- --------- ----------------- ------------------------------------------------------ ------------------------------------------------------ -------------------------------------------------------- --------------- ---------------------------
1 N 86 TP, death (day 120) No No Yes No
2 N 80 Slight proximal TP, normal tendon reflexes No No Yes No
3 N 25 Complete recovery with normal tendon reflexes Yes Yes No No
4 N 51 TP with marked atrophies, death (day 51) No No Yes Yes
5 N 44 Complete recovery of legs, distal spasticity of arms Yes Yes No No
6 N 45 Slight proximal TP Yes Yes No No
7 NM 281 TP with marked atrophies, death (day 281) No No Yes Yes
8 NM 76 TP with marked atrophies No No Yes No
9 NM 105 TP with marked atrophies, death (day 105) No No Yes Yes
10 M 204 TP with marked atrophies, death (day 204) No No Yes Yes
11 M 57 Slight left upper MP Yes Yes No No
12 M 91 Slight proximal TP No Yes No No
13 M 45 TP with marked atrophies, death (day 45) No No Yes Yes
14 M 56 Complete recovery with normal tendon reflexes Yes Yes No No
15 M 36 TP with marked atrophies, death (day 36) No No Yes No
^a^Type of neurological lesion. ^b^Weaning from mechanical ventilation at the end of the intensive care unit (ICU) stay. M, myopathy; MP, monoparesia; N, neuropathy; NM, neuromyopathy; TP, tetraparesia.
:::
|
PubMed Central
|
2024-06-05T03:55:54.827050
|
2004-9-10
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065049/",
"journal": "Crit Care. 2004 Sep 10; 8(6):R358-R366",
"authors": [
{
"first": "François",
"last": "Kerbaul"
},
{
"first": "Muriel",
"last": "Brousse"
},
{
"first": "Frédéric",
"last": "Collart"
},
{
"first": "Jean-François",
"last": "Pellissier"
},
{
"first": "Denis",
"last": "Planche"
},
{
"first": "Carla",
"last": "Fernandez"
},
{
"first": "François",
"last": "Gouin"
},
{
"first": "Catherine",
"last": "Guidon"
}
]
}
|
PMC1065052
|
See related commentary <http://ccforum.com/content/8/6/422>
Introduction
============
Knowledge of the occurrence of and determinants of critical illness is important for establishing its burden and the risk factors for acquisition to guide wise allocation of limited healthcare and research resources. Population-based cohort studies that strictly include all episodes of disease occurring in residents of a geographically defined region are commonly accepted as the optimal design for such purposes \[[@B1]-[@B3]\]. However, these designs have rarely been used in the critical care medical literature \[[@B4]-[@B6]\]. Studies attempting to evaluate the distribution and determinants of critical illness typically have been case series reported from academic tertiary care referral hospitals \[[@B7]-[@B9]\]. Multicentered studies that include intensive care units (ICUs) in different regions and/or countries have less commonly been performed \[[@B10]-[@B13]\].
A major limitation to these institution-based studies is that if the population at risk is unknown, then incidence rates may not be calculated. Furthermore, if these studies focus on tertiary care centers and fail to include critically ill patients admitted to ICUs in other hospitals, a biased assessment of disease occurrence and severity may occur \[[@B14]\]. This may still be problematic even if all ICUs in a defined geographic region are included if investigators do not exclude patients nonresident in that base population from analysis \[[@B15]\]. Although referral bias has been shown to be of major importance in a number of disease conditions \[[@B16]-[@B20]\], its importance in the ICU has only been systematically assessed in one study reported from a single medical ICU in a tertiary care university hospital \[[@B21]\].
The importance of excluding patients external to the base population in observational studies in the critically ill has not been well defined. Furthermore, few population-based studies have been conducted among the critically ill and none in the English-language literature have assessed the overall burden and risk factors associated with ICU admission. The objective of this study was to evaluate the impact of inclusion of nonresidents in population-based studies on the occurrence of, on the risk factors for, and on the outcomes of ICU admission.
Materials and methods
=====================
Study population
----------------
The Calgary Health Region (CHR) administers all medical and surgical acute care to the residents of the cities of Calgary and Airdrie, and to approximately 20 nearby small towns, villages, and hamlets (2001 population, 958,610). In April 2003 the CHR was expanded to include the adjacent mountain parks and Wheatland regions \[[@B22]\]. All tertiary care services are provided by the CHR with the only exception being liver, heart, or lung transplantation, where patients are referred to the provincial program in Edmonton. The acute care institutions within the CHR also serve as referral centers for other communities in southern Alberta and the neighboring provinces of British Columbia and Saskatchewan.
All adult ICUs within the CHR are closed units staffed by fully trained intensivists, and they are administered by the Department of Critical Care Medicine, University of Calgary and the CHR. These ICUs currently include a 14-bed cardiovascular surgery intensive care unit (CVICU) and a 22-bed multidisciplinary ICU that serves as the regional trauma and neurosurgical referral center at the Foothills Medical Centre, a 12-bed multidisciplinary ICU at the Peter Lougheed Centre that is also the vascular surgery referral center, and a 10-bed multidisciplinary ICU at the Rockyview General Hospital. All patients 18 years and older admitted to an adult multidisciplinary ICU or the CVICU in the CHR between 1 May 1999 and 30 April 2003 were included. Ethics approval was obtained from the Conjoint Health Research Ethics Board at the University of Calgary and the CHR.
Protocol
--------
The study utilized a population-based surveillance cohort design with linkage of data collected from regional critical care and administrative databases. Demographic data, clinical data, basic laboratory data, and scoring data were obtained from all patients admitted to ICUs in the CHR in a consistent manner across all sites using the ICU Tracer database, as previously described \[[@B23],[@B24]\]. Patients were classified as CHR residents or nonresidents using data from the CHR Data Warehouse (a regional administrative database), where regional residents are flagged if their home address is within the geographical boundaries of the CHR.
Severity of illness at admission was assessed using the Acute Physiology and Chronic Health Evaluation (APACHE) II score, and the intensity of care was assessed using the Therapeutic Intervention Scoring System score \[[@B25],[@B26]\]. Shock was defined as a mean arterial pressure \< 60 mmHg on the first day of admission to the ICU or requirement for a vasopressor infusion. The diagnosis of systemic inflammatory response syndrome (SIRS) was based on a modification of consensus criteria and required at least two of the following; heart rate \> 90/min, respiratory rate \> 20/min, temperature \< 36°C or \> 38°C, or white blood cell count \< 4 × 10^9^/l or \> 12 × 10^9^/l \[[@B15]\]. A surgical patient was any patient recorded as having an operative diagnosis or any patient admitted from the trauma ward or directly from the operating room.
Statistical analysis
--------------------
Analysis was performed using Stata version 8.0 (Stata Corp, College Station, TX, USA). With the exception of calculating SIRS criteria, where missing values were treated as normal, missing data were not replaced and a reduced number (*n*) reported where they ocurred. Only first ICU presentations were analyzed from patients with multiple ICU admissions. Normally or near-normally distributed variables were reported as means ± standard deviations and non-normally distributed variables were reported as medians with interquartile ranges (IQRs). Means were compared using the Student *t*test and medians were compared using the Mann--Whitney *U*test. Differences in proportions among categorical data were assessed using Fisher\'s exact test. Incidence rates were calculated using regional denominator data and compared as previously described \[[@B2]\]. Levels of significance were not *a priori*adjusted for multiple testing, and a two-sided *P*\< 0.05 was considered significant for all comparisons.
A multivariable logistic regression model was developed to assess independent risk factors for death. The initial model included clinically suspected variables and those identified as potentially important predictors, including CHR residency, multidisciplinary ICU admission as compared with CVICU admission, the presence of SIRS, shock, hypothermia, age, gender, surgical diagnosis, and APACHE II and Therapeutic Intervention Scoring System scores. Backward stepwise variable elimination was then performed to develop the final model. The final model discrimination was assessed using the area under the receiver operator curve and calibration using the Hosmer--Lemeshow goodness-of-fit test.
Results
=======
During the 4-year study period 12,193 adult patients had a total of 13,638 admissions to CHR ICUs; 4509 were surgical admissions for less than 48 hours. Overall 7767 (63.7%) patients were classified as CHR residents, for an incidence of ICU admission of 263.7 per 100,000 per year. Both the quarterly and yearly numbers of admissions were stable over the study. More than one-third (4426) of patients were nonresident in the CHR (incidence not able to be calculated) and were primarily (3424 patients) from other health regions in Alberta, 705 patients were from British Columbia, 121 were from Saskatchewan, 135 were from other Canadian provinces and territories, and 41 were international residents. Among the four study ICUs there were 4715 admissions to the CVICU, 3584 to Foothills Medical Centre ICU, 2144 to Peter Lougheed Centre ICU, and 1750 to Rockyview General Hospital ICU, of which 2587 (54.9%), 2264 (63.2%), 1541 (71.9%), and 1375 (78.6%) were CHR residents, respectively. A significant proportional difference in admission rates for CHR and non-CHR residents was observed (*P*\< 0.001) between each of the ICUs.
Demographic features
--------------------
The overall median age (IQR) was 64.6 years (50.6--74.0 years) and 7819 patients (64.1%) were male. Although the overall median age of CHR and non-CHR residents was not different, in the subgroup of patients aged 85 years and older patients were nearly twice as likely to be CHR residents (relative risk \[RR\], 1.80; 95% confidence interval \[CI\], 1.43--2.27; *P*\< 0.0001). There was a gender difference associated with residency status as non-CHR residents were significantly more likely to be male as compared with CHR residents (67.9% versus 62.0%; *P*\< 0.0001). Age-specific and gender-specific population incidence rates were established for the population-based cohort as shown in Fig. [1](#F1){ref-type="fig"}. Males were at significant increased risk for ICU admission as compared with females (330.5 per 100,000 versus 198.2 per 100,000; RR, 1.67; 95% CI, 1.59--1.74; *P*\< 0.0001), and this was consistent observed among all age groups (Fig. [1](#F1){ref-type="fig"}). Increasing risk was associated with incrementally advancing age up to the age of 85 years, where a decrease in incidence was then observed (Fig. [1](#F1){ref-type="fig"}). As compared with younger individuals, those aged 65 years and older were at substantially increased risk of admission to an ICU (1719.9 per 100,000 versus 238.7 per 100,000; RR, 7.21; 95% CI, 6.95--7.47; *P*\< 0.0001).
Clinical features
-----------------
Although the magnitudes of differences were small, a number of clinical features were significantly different among CHR and non-CHR residents, as presented in Table [1](#T1){ref-type="table"}. In general, non-CHR residents had more markers of increased severity as compared with CHR residents (Table [1](#T1){ref-type="table"}). No difference was observed between CHR and non-CHR residents in the occurrence of SIRS, although overall 90% (11,020) of patients fulfilled criteria.
Outcomes
--------
The overall medians of ICU length of stay and hospital length of stay were 1.9 (IQR, 1--3.9) and 11 (IQR, 6--21), respectively. No significant differences were observed between CHR and non-CHR residents with respect to length of stay. In total, 1443 (11.8%) patients died in the ICU and a further 667 died during that hospitalization, for an overall inhospital case fatality rate of 17.3%. There was a significant effect of CHR residency on case fatality; CHR residents were much more likely to die in the ICU (1016 \[13.1%\] versus 427 \[9.6%\]; RR, 1.36; 95% CI, 1.22--1.51; *P*\< 0.0001) and in hospital (1547 \[20%\] versus 563 \[12.7%\]; RR, 1.57; 95% CI, 1.43--1.71; *P*\< 0.0001) as compared with non-CHR resident patients. A multivariable logistic regression model (*n*= 11,569) was developed that had good fit (*P*= 0.4) and discrimination (area under receiver operator curve = 0.83). As presented in Table [2](#T2){ref-type="table"}, CHR residency status was independently associated with inhospital death.
Discussion
==========
This study describes the occurrence of, the demographic risk factors for, and the outcome associated with ICU admission in a large nonselected North American population. Although it is notable that the annual incidence of ICU admission is reported, it is of greater interest that demographic risk groups in the population that were at increased risk for admission to an ICU were defined. Not surprisingly, older age and male gender were associated with an increased need for ICU admission. This may be at least partly due to a higher rate of comorbid conditions, such as smoking or alcohol use, or other high-risk behaviors or activities among males as compared with females \[[@B4],[@B5]\]. The population-based cohort design is an excellent method for defining the actual magnitude of such risks \[[@B2]\]. However, detailed information on each of the patient\'s comorbidities was not available for all patients in this study, and as a result the risk factor analysis was limited to the evaluation of demographic features alone. The actual burden of disease requiring ICU admission in an entire population was established in a minimally biased fashion in this study. Such accurate information on the degree of human suffering and death related to critical illness is important to potentially support continued or increased funding of clinical ICUs and critical care medical research.
This study demonstrates that inclusion of nonresidents of a base population may have a major impact on biasing the results of studies in the ICU. When nonresidents of the CHR were included in this study, the occurrence of ICU admission in the CHR was overestimated by more than 50%. This observation is consistent with previous studies in the CHR and elsewhere in noncritically ill specific populations \[[@B2],[@B16],[@B19],[@B27]\]. On the other hand, it is highly unlikely that a significant number of CHR residents requiring ICU admission were missed in this study. This is because all multidisciplinary and cardiovascular surgical ICUs in the CHR were included in surveillance and that, with the exception of acute liver, heart, and lung transplantation, patients are rarely referred out of the CHR for provision of healthcare. Furthermore, the CHR is relatively geographically isolated, with the closest tertiary care center to the CHR in Edmonton approximately 300 km away. Therefore, with the exception of the small number of CHR residents who may have required ICU admission while traveling, it seems unlikely that a substantial number of CHR residents requiring ICU admission would have been lost to analysis in this study.
A number of statistically significant differences in the clinical features between CHR and non-CHR residents (Table [1](#T1){ref-type="table"}) were observed, and with the exception of fever these would remain significant even if a conservative correction for multiple statistical comparisons such as the Bonferroni method were used. However, although statistically significant, the magnitudes of these differences are small and may not be of practical clinical difference. On the other hand, there was a dramatic effect of residency status on the outcome of patients admitted to ICUs in the CHR. The observation of a lower mortality among non-CHR patients is in contrast to the recent hospital-based study reported by Rosenberg and colleagues, although the definition of \'referral\' was different in their study \[[@B21]\]. The reason why non-CHR patients were at lower risk for inhospital mortality is unexplained by the present study data, especially given that they appeared in general to be somewhat sicker on average than CHR residents (Table [1](#T1){ref-type="table"}). The possibility exists that non-CHR patients may have died after transfer back to their \'home\' health region hospitals and have therefore not been captured in the study inhospital mortality. This would explain the apparent lower inhospital case fatality rate among non-CHR residents but is only speculation. Of note, there were no significant pair-wise interactions between residency status and each of the other variables in the multivariable model. This study demonstrates that if nonresidents of a base population are included in studies of patients admitted to ICUs, gross errors in the determination of occurrence and outcomes may occur.
The results of this study raise concerns regarding the generalization of results obtained from hospital-based reviews or population-based studies where nonresidents are included. However, this may not always be of major practical significance depending on specific study objectives. For example, a hospital-based study defining the outcome of a certain patient population such as transplant patients may be generalizable to other transplant centers because transplant recipients are nearly always managed at academic tertiary care referral institutions \[[@B28]\]. Generalization of results to other populations may therefore not be necessary. Similarly, population-based studies that strictly exclude nonresidents may not always be necessary for providing important information to guide allocation of health resources at regional levels. For example, Manns and colleagues conducted an economic evaluation of activated protein C for severe sepsis using clinical information from such a \'population-based\' cohort in the CHR \[[@B29]\]. Although they included non-CHR residents, their results should be widely generalizable to other centers in North America and worldwide because typically patients requiring this therapy are admitted in tertiary care ICUs that are composed of a substantial number of referral patients. It should be recognized, however, that although studies that suffer from such selection bias may provide useful clinical information, results should not be generalized to unlike patient cohorts, and rarely, if ever, to the population as a whole.
There are some limitations to this study that merit discussion. First, the CHR may have a different socioeconomic and demographic profile as compared with other regions, and this may influence the validity of generalizing results to other populations. One advantage, however, is that since this study was population-based, age and gender standardization against a reference population may be performed to facilitate comparison among different regions. This has been demonstrated to be of significant value in other population-based studies conducted in the United States \[[@B3]\]. Second, although the data were collected in a uniform fashion at each of the regional ICUs and much of this was directly linked from bedside monitors, systematic manual auditing of the information was not performed. However, previous work has suggested a high degree of accuracy \[[@B24]\]. Third, the need for admission to an ICU in this study was determined by the attending intensivist and not on some predefined objective criteria. This may be important for generalization to other centers that use different criteria for ICU admission. For example, patients admitted to Canadian ICUs tend to be sicker than those admitted to American ICUs, although adjustment according to APACHE II scores is possible \[[@B30]\]. Fourth, we did not have adequate admission data to further define patients into more refined subgroups for analysis. Finally, it is possible that some case patients were missed by our study surveillance as a result of care external to the CHR. However, given the comprehensiveness of the critical care system in the CHR and its relative geographic isolation, this would be only expected to have a minor effect on the study findings.
Conclusion
==========
This study demonstrates the adverse effect of inclusion of nonresident patients of the base population on the determination of occurrence and outcome in studies of patients admitted to ICUs. Further well-designed, population-based studies in other regions that exclude nonresidents of the base population are required to better define the distribution and determinants of ICU admission internationally.
Key messages
============
• This population-based cohort study included all adults admitted to CHR multidisciplinary and cardiovascular surgical ICUs during a 4-year period. The effect of inclusion of non-residents in the study was evaluated.
• Failure to exclude non-residents would lead to an overestimation of the incidence of ICU admission by more than 50%. A number of clinical features were significantly different between resident and non-resident patients; most notably, the in-hospital mortality rate was much lower in the non-resident cohort.
• This study supports that non-resident patients should be strictly excluded from population-based studies.
Competing interests
===================
The authors declare that they have no competing intrests.
Abbreviations
=============
APACHE = Acute Physiology and Chronic Health Evaluation; CHR = Calgary Health Region; CI = confidence interval; CVICU = cardiovascular surgery intensive care unit; ICU = intensive care unit; IQR = interquartile range; RR = relative risk; SIRS = systemic inflammatory response syndrome.
Acknowledgement
===============
This work was supported in part by a grant from the Canadian Intensive Care Foundation.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Age-specific and gender-specific population incidence of intensive care unit admission in Calgary Health Region, Alberta, Canada.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Statistically significant different clinical features of Calgary Health Region residents and nonresidents admitted to intensive care units, Alberta, Canada
:::
Characteristic Calgary Health Region resident Non-Calgary Health Region resident Total *P*value
--------------------------------------------------- -------------------------------- ------------------------------------ ----------------------------- -----------
APACHE II score (mean ± standard deviation) 24.90 ± 8.71 (*n*= 7704) 25.46 ± 8.16 (*n*= 4381) 25.10 ± 8.52 (*n*= 12,085) \< 0.001
APACHE II score = 25 3556/7704 (53.8%) 1818/4381 (58.5%) 5374/12,085 (55.5%) \< 0.0001
TISS score (mean ± standard deviation) 43.52 ± 18.78 (*n*= 7411) 48.94 ± 17.99 (*n*= 4260) 45.50 ± 18.67 (*n*= 11,671) \< 0.0001
Surgical patient 4600/7754 (59%) 3254/4420 (74%) 7854/12,174 (65%) \< 0.0001
Fever = 37.8°C 3683/7604 (48.4%) 2200/4343 (50.7%) 5883/11,947 (49.2%) 0.02
Hypothermia \< 35°C 1413/7610 (18.6%) 950/4347 (21.9%) 2363/11,957 (19.8%) \< 0.0001
Shock 4551/7767 (58.6%) 2731/4426 (61.7%) 7282/12,193 (59.7%) \< 0.001
Tachycardia = 100/min 5059/7698 (65.7%) 2729/4381 (62.3%) 7788/12,079 (64.5%) \< 0.001
Median (interquartile range) respiratory rate/min 26 (20--33) 24 (18--30) 25 (19--32) \<0.0001
APACHE, Acute Physiology and Chronic Health Evaluation; TISS, Therapeutic Intervention Scoring System.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Multivariable logistic regression modeling of risk factors for inhospital death among patients admitted to intensive care units in the Calgary Health Region, Alberta, Canada
:::
Variable Odds ratio 95% Confidence interval^a^
-------------------------------- ------------ ----------------------------
APACHE II score = 25 3.06 2.68--3.50
TISS score = 45 2.02 1.75--2.34
Age = 65 years 1.95 1.73--2.20
Hypothermia \< 35°C 1.99 1.71--2.32
Shock 1.66 1.46--1.88
Calgary Health Region resident 1.36 1.20--1.55
Noncardiac surgery^b^ 0.57 0.50--0.64
Cardiac surgery^b^ 0.02 0.02--0.03
APACHE, Acute Physiology and Chronic Health Evaluation; TISS, Therapeutic Intervention Scoring System.
^a^*P*\< 0.0001 for all variables.
^b^As compared with medical diagnosis as the reference group.
:::
|
PubMed Central
|
2024-06-05T03:55:54.831107
|
2004-10-15
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065052/",
"journal": "Crit Care. 2004 Oct 15; 8(6):R431-R436",
"authors": [
{
"first": "Kevin B",
"last": "Laupland"
}
]
}
|
PMC1065053
|
See related commentary <http://ccforum.com/content/8/6/414>
Introduction
============
Biological systems are complex systems; specifically, they are systems that are spatially and temporally complex, built from a dynamic web of interconnected feedback loops marked by interdependence, pleiotropy and redundancy. Complex systems have properties that cannot wholly be understood by understanding the parts of the system \[[@B1]\]. The properties of the system are distinct from the properties of the parts, and they depend on the integrity of the whole; the systemic properties vanish when the system breaks apart, whereas the properties of the parts are maintained. Illness, which presents with varying severity, stability and duration, represents a systemic functional alteration in the human organism. Although illness may occasionally be due to a specific singular deficit (e.g. cystic fibrosis), this discussion relates to illnesses characterized by systemic changes that are secondary to multiple deficits, which differ from patient to patient, with varied temporal courses, diverse contributing events and heterogeneous genetic contributions. However, all factors contribute to a physiological alteration that is recognizable as a systemic illness. Multiple organ dysfunction syndrome represents the ultimate multisystem illness, really representing a common end-stage pathway of inflammation, infection, dysfunctional host response and organ failure in critically ill patients, and frequently leading to death \[[@B2]\]. Although multiple organ dysfunction syndrome provides a useful starting point for discussion regarding complex systems and variability analysis \[[@B3]\], the application of variability analysis to other disease states is readily apparent and exciting.
Life is composed of and characterized by rhythms. Abnormal rhythms are associated with illness and can even be involved in its pathogenesis; they have been termed \'dynamical diseases\' \[[@B4]\]. Measuring the absolute value of a clinical parameter such as heart rate yields highly significant, clinically useful information. However, evaluating heart rate variability (HRV) provides additionally useful clinical information, which is, in fact, more valuable than heart rate alone, particularly when heart rate is within normal limits. Indeed, as is demonstrated below, there is nothing \'static\' about homeostasis. Akin to the concept of homeorrhesis (dynamic stability) introduced by CH Waddington, homeokinesis describes \'the ability of an organism functioning in a variable external environment to maintain a highly organized internal environment, fluctuating within acceptable limits by dissipating energy in a far from equilibrium state\' \[[@B5]\].
Clinicians have long recognized that alterations in physiological rhythms are associated with disease. The human eye is an excellent pattern recognition device, which is capable of complex interpretation of ECGs and electroencephalograms (EEGs) \[[@B6]\], and physicians make use of this skill on a daily basis. However, more sophisticated analysis of variability provides a measure of the integrity of the underlying system that produces the dynamics. As the spatial and temporal organization of a complex system define its very nature, changes in the patterns of interconnection (connectivity) and patterns of variation over time (variability) contain valuable information about the state of the overall system, representing an important means with which to prognosticate and treat our patients \[[@B3]\]. As clinicians, our goal is to make use of this observation in order to improve patient care. This technology of variability analysis is particularly valuable in the intensive care unit (ICU), where patients are critically ill and numerous parameters are routinely measured continuously. The intensivist is poised to marshal the science of variability analysis, becoming a \'dynamicist\' \[[@B6]\], to measure and characterize the variability of physiological signals in an attempt to understand the information locked in the \'homeokinetic code\' \[[@B7]\], and thus contribute to a breakthrough in our ability to treat critically ill patients.
The focus of this review and analysis is the measurement and characterization of variability, a science that has undergone considerable growth in the past two decades. The development of mathematical techniques with a theoretical basis in chaos theory and nonlinear dynamics has provided us with greater ability to discern meaningful distinctions between biological signals from clinically distinct groups of patients. The science of variability analysis has developed from a close collaboration between mathematicians, physicists and clinicians. As such, the techniques for measuring variability sometimes represent a bewildering morass of equations and terminology. Each technique represents a unique and distinct means of characterizing a series of data in time. The principal objectives of this review are as follows: to present a concise summary, including definition, interpretation, advantages, limitations and calculation of the principal techniques for performing variability analysis; to discuss the interpretation and application of this technology; and to propose how this information may improve patient care. Although the majority of the discussion relates to the analysis of HRV because is it readily and accurately measured on an ECG, the techniques are applicable to any biological time signal. Two tables are included to facilitate review of the techniques for characterizing variability (Table [1](#T1){ref-type="table"}) and the evidence for altered variability in illness (Table [2](#T2){ref-type="table"}).
Science of variability analysis
===============================
Sampling
--------
The analysis of patterns of change over time or variability is performed on a series of data collected continuously or semicontinuously over time. For example, a heart rate tracing may be converted to a time series of intervals between consecutive heart beats (measured as R--R\' intervals on an ECG). The same may be done with inter-breath intervals, albeit not as easily. When there is no intrinsic rhythm such as a heart or respiratory rate, sampling a signal occurs in discrete time intervals (e.g. serum concentrations of a hormone measured every few minutes). In order to reconstruct the underlying signal without error, one must respect the Nyquist Theorem, which states that the sampling frequency must be at least twice the highest frequency of the signal being sampled.
Stationarity
------------
Stationarity defines a limitation in techniques designed to characterize variability. It requires that statistical properties such as mean and standard deviation of the signal remain the same throughout the period of recording, regardless of measurement epoch. Stationarity does not preclude variability, but it provides boundaries for variability such that variability does not change with time or duration of measurement. If this requirement is not met, as is the case with most if not all biological signals when physiological and/or pathophysiological conditions change, then the impact of trends with change on the mean of the data set must be considered in the interpretation of the variability analysis. The relative importance of stationarity to individual techniques of variability analysis is addressed below.
Artifact
--------
Variability analysis should be performed on data that are free from artifact, with a minimal noise:signal ratio. Noise is measurement error, or imprecision secondary to measurement technology. Often present in patient monitoring, artifact must be removed, often by visual inspection of the raw data. For example, in the evaluation of HRV the presence of premature atrial and/or ventricular beats require that the data be removed, and appropriate interpolation be performed without compromising the integrity of the variability analysis. Several techniques, such as a Poincaré Plot of the difference between consecutive data points, have been developed to facilitate automated identification and removal of artifact \[[@B8]-[@B10]\]. Different techniques are more or less sensitive to artifact, which again is addressed below.
Standardized technique
----------------------
Various factors alter variability measurement. For example, standing or head-up tilt (increased sympathetic activity) and deep breathing (increased respiratory rate induced HRV) will alter HRV indices in healthy individuals. With deference to Heisenberg, experimental design should take into account that the process of measurement may alter the intrinsic variation. An important component of standardized technique is the duration of measurement for analysis. For example, indices of HRV may be calculated following a duration of 15 min or 24 hours. In general terms, it is inappropriate to compare variability analysis from widely disparate durations of measurements \[[@B11]\]. More specifically, the impact of duration of measurement varies in relation to individual analysis technique, and is discussed below.
Time domain analysis
====================
Definition
----------
Time series analysis represents the simplest means of evaluating variability, identifying measures of variation over time such as standard deviation and range. For example, quantitative time series analysis is performed on heart rate by evaluating a series of intervals between consecutive normal sinus QRS complexes (normal--normal, or NN or RR\' interval) on an ECG over time. In addition, a visual representation of data collected as a time series may be obtained by plotting a frequency distribution, plotting the number of occurrences of values in selected ranges of values or bins.
Calculation
-----------
Mathematically, standard deviation is equal to the square root of variance; and variance is equal to the sum of the squares of difference from the mean, divided by the number of degrees of freedom. Evaluating HRV, the standard deviation of a series of NN intervals (SDNN) represents a coarse quantification of overall variability. As a measure of global variation, standard deviation is altered by the duration of measurement; longer series will have greater SDNN. Thus, SDNN can be calculated for short periods between 30 s and 5 min and used as a measure of short-term variability, or calculated for long periods (24 hours) as a measure of long-term variation \[[@B12]\]. Because it is inappropriate to compare SDNNs from recordings of different duration, standardized duration of recording has also been suggested \[[@B11]\].
Various permutations of measurement of standard deviation, in an effort to isolate short-term, high frequency fluctuations from longer term variation, are possible. For example, SDANN (standard deviation of the average NN interval calculated over 5-min intervals within the entire period of recording) is a measure of longer term variation because the beat-to-beat variation is removed by the averaging process. In contrast, the following variables were devised as a measure of short-term variation: RMSSD (square root of the mean squared differences of consecutive NN intervals), NN50 (number of pairs of adjacent NN intervals differing by more than 50 ms), and pNN50 (proportion of NN intervals differing by more than 50 ms = NN50 divided by total number of NN intervals). These measures of high frequency variation are interrelated; however, RMSSD has been recommended because of superior statistical properties \[[@B11]\]. The conventional 50 ms used in the NN50 and pNN50 measurements represents an arbitrary cutoff, and is only one member of a general pNNx family of statistics; in fact, a threshold of 20 ms may demonstrate superior discrimination between physiological and pathological HRV \[[@B13]\].
In order to characterize a frequency distribution, it may be fitted to a normal distribution, or rather a log-normal distribution -- one in which the log of the variable in question is normally distributed. The skewness or degree of symmetry may be calculated, with positive and negative values indicating distributions with a right-sided tail and a left-sided tail, respectively. Kurtosis may also be calculated to identify the peakedness of the distribution; positive kurtosis (leptokurtic) indicates a sharp peak with long tails, and negative kurtosis (platykurtic) indicates a flatter distribution.
Interpretation and clinical application
---------------------------------------
Time domain analysis involves the statistical evaluation of data expressed as a series in time. Clinical evaluation of time domain measures of HRV have been extensive, using overall standard deviation (SDNN) to measure global variation, standard deviation of 5-min averages (SDANN) to evaluate long-term variation, and the square root of mean squared differences of consecutive NN intervals (RMSSD) to measure short-term variation. An abridged review of an extensive literature suggests that diminished overall HRV measured with time domain analysis portends poorer prognosis and/or increased mortality risk in patients with coronary artery disease \[[@B14],[@B15]\], dilated cardiomyopathy \[[@B16]\], congestive heart failure \[[@B17],[@B18]\] and postinfarction patients \[[@B19]-[@B23]\], in addition to elderly patients \[[@B24]\]. Time domain HRV analysis has been used to compare ß-blocker therapies postinfarction \[[@B25]\], to evaluate percutaneous coronary interventions \[[@B26],[@B27]\], to predict arrhythmias \[[@B28]\] and to select patients for specific antiarrhythmic therapies \[[@B29]\], which are a few examples of a vast body of literature that is well reviewed elsewhere \[[@B30],[@B31]\].
Time series of parameters derived from biological systems are known to follow log-normal frequency distributions, and deviations from the log-normal distribution have been proposed to offer a means with which to characterize illness \[[@B32]\]. For example, in paediatric ICU patients with organ dysfunction, HRV evaluated using a frequency distribution (plotting frequency of occurrence of differences from the mean) revealed a reduction in HRV and a shift in the frequency distribution to the left with increasing organ failure; these changes improved in surviving patients and were refractory in nonsurvivors \[[@B33]\]. The authors utilized a technique that was initially described in the evaluation of airway impedance variability, demonstrating increased variability in asthma patients characterized by altered frequency distribution \[[@B5]\].
Advantages and limitations
--------------------------
Statistical measures of variability are easy to compute and provide valuable prognostic information about patients. Frequency distributions also offer an accurate, visual representation of the data, although the analysis may be sensitive to the arbitrary number of bins chosen to represent the data. Time domain measures are susceptible to bias secondary to nonstationary signals. A potential confounding factor in characterizing variability with standard deviation is the increase in baseline heart rate that may accompany diminished HRV indices. The clinical significance of this distinction is unclear, because the prognostic significance of altered SDNN or SDANN remains clinically useful. A more condemning limitation of time domain measures is that they do not reliably distinguish between distinct biological signals. There are many potential examples of data series with identical means and standard deviations but with very different underlying rhythms \[[@B34]\]. Therefore, additional, more sophisticated methods of variability analysis are necessary to characterize and differentiate physiological signals. It is nonetheless encouraging that, using rather crude statistical measures of variability, it is possible to derive clinically useful information.
Frequency domain analysis
=========================
Definition
----------
Physiological data collected as a series in time, as with any time series, may be considered a sum of sinusoidal oscillations with distinct frequencies. Conversion from a time domain to frequency domain analysis is made possible with a mathematical transformation developed almost two centuries ago (1807) by the French mathematician Jean-Babtiste-Joseph Fourier (1768--1830). Other transforms exist (e.g. wavelet, Hilbert), but Fourier was first and his transformation is used most commonly. The amplitude of each sine and cosine wave determines its contribution to the biological signal; frequency domain analysis displays the contributions of each sine wave as a function of its frequency. Facilitated by computerized data harvest and computation, the result of converting data from time series to frequency analysis is termed spectral analysis because it provides an evaluation of the power (amplitude) of the contributing frequencies to the underlying signal.
Calculation
-----------
The clinician should note that the power spectrum is simply a different representation of the same time series data, and the transformation may be made from time to frequency and back again. It is not necessary for the clinician to know how to perform power spectral density analysis using the fast Fourier transformation because computers can do so quickly and reliably, calculating a weighted sum of sinusoidal waves, with different amplitudes and frequencies. This provides an analysis of the relative contributions of different frequencies to the overall variation in a particular data series. Interpretation of the analysis must factor in the assumptions inherent to this calculation, namely stationarity and periodicity. Note that the square of the contribution of each frequency is the power of that frequency to the total spectrum, and the total power of spectral analysis (area under the curve of the power spectrum) is equal to the variance described above (they are different representations of the same measure) \[[@B11]\]. The fast Fourier transform or analysis (see Appendix 1) represents a nonparametric calculation because it provides an evaluation of the contribution of all frequencies, not discrete or preselected frequencies.
Interpretation and clinical application
---------------------------------------
Spectral analysis of heart rate was first performed by Sayers \[[@B35]\]. It was subsequently used to document the contributions of the sympathetic, parasympathetic and renin--angiotensin systems to the heart rate power spectrum, which introduced frequency domain analysis as a sensitive, quantitative and noninvasive means for evaluating the integrity of cardiovascular control systems \[[@B36]\]. Spectral analysis has been utilized to evaluate and quantify cardiovascular and electroencephalographic variability in numerous disease states, and is perceived as an important tool in clinical medicine \[[@B37]\].
The power spectral density function or power spectrum provides a characteristic representation of the contributing frequencies to an underlying signal. By identifying and measuring the area of distinct peaks on the power spectrum, it is possible to derive quantitative connotation to facilitate comparison between individuals and groups. In 2--5 min recordings, spectral analysis reveals three principal peaks, identified by convention with the following ranges: very low frequency (VLF; frequency = 0.04 Hz \[cycles/s\], cycle length \>25 s), low frequency (LF; frequency 0.04--0.15 Hz, cycle length \>6 s) and high frequency (HF; frequency 0.15--0.4 Hz, cycle length 2.5--6 s). In 24 hour recordings VLF is further subdivided into VLF (frequency 0.003--0.04 Hz) and ultralow frequency (ULF; frequency = 0.003 Hz, cycle length \>5 hours) \[[@B11]\]. Correlations between time and frequency measures have also been demonstrated, for example in healthy newborns \[[@B38]\] and in cardiac patients following myocardial infarction \[[@B39]\].
Numerous factors in health and disease have an impact on the amplitude and area of each peak (or frequency range) on the HRV power spectrum. Akselrod and coworkers \[[@B36]\] first demonstrated the contributions of sympathetic and parasympathetic nervous activity and the renin--angiotensin system to frequency specific alterations in the HRV power spectrum in dogs. Several authors have evaluated and reviewed the relationship between the autonomic nervous system and spectral analysis of HRV \[[@B40]-[@B44]\]. Although autonomic regulation is clearly a significant regulator of the HRV power spectrum, evidence demonstrates a lack of concordance with direct evaluation of sympathetic tone, for example in patients with heart failure \[[@B45]\], and reviews increasingly conclude that HRV is generated by multiple physiological factors, not just autonomic tone \[[@B46],[@B47]\].
In interpreting the significance of the HRV power spectrum, investigators initially focused on peaks because of a presumed relationship with a single cardiovascular control mechanism leading to rhythmic oscillations; however, others documented nonrhythmic (no peak) fluctuations in both heart rate and blood pressure variability, indicating the need to analyze broadband power \[[@B48]\]. Thus, the calculation of HF, LF, VLF and ULF using the ranges listed above serve to facilitate data reporting and comparison, but they are nonetheless arbitrary ranges with diverse physiological input. A recent review of HRV \[[@B47]\] documented the evidence that ULF reflects changes secondary to the circadian rhythm, VLF is affected by temperature regulation and humoral systems, LF is sensitive to cardiac sympathetic and parasympathetic nerve activity, and HF is synchronized to respiratory rhythms, primarily related to vagal innervation.
What does spectral analysis of HRV tell us about our patients? Despite nonspecific pathophysiological mechanisms, there is ample evidence that the frequency contributions to HRV are altered in illness states, and that the degree of alteration correlates with illness severity. It is illustrative that alterations in the spectral HRV analysis related to illness severity have been demonstrated from hypovolaemia \[[@B49]\] to heart failure \[[@B50]-[@B52]\], from hypertension \[[@B53],[@B54]\] to coronary artery disease \[[@B55],[@B56]\], and from angina \[[@B57]\] to myocardial infarction \[[@B58]\], in addition to chronic renal failure \[[@B59]\], autonomic neuropathy secondary to diabetes mellitus \[[@B60]\], depth of anaesthesia \[[@B61]\] and more. Spectral analysis of HRV has been applied in the ICU. For example, using spectral HRV and blood pressure variability analyses in consecutive patients admitted to an ICU, increasing total and LF HRV power were associated with recovery and survival, whereas progressive decreases in HRV were associated with deterioration and death \[[@B62]\]. In separate investigations involving patients in the emergency room \[[@B63]\] or admitted to an ICU after 48 hours \[[@B64]\], decreased total, LF and LF/HF HRV was not only present in patients with sepsis but also correlated with subsequent illness severity, organ dysfunction and mortality. Several reviews discuss the application of HRV spectral analysis to the critically ill patient \[[@B65]-[@B68]\]. Thus, alterations in spectral analysis correlate with severity of illness, a finding consistently reported in cardiac and noncardiac illness states, providing the clinician with a means with which to gauge prognosis and determine efficacy of intervention.
Advantages and limitations
--------------------------
In order to derive a valid and meaningful analysis using a fast Fourier transform and frequency domain analysis, the assumptions of stationarity and periodicity must be fulfilled. The signal must be periodic, namely it is a signal that is comprised of oscillations repeating in time, with positive and negative alterations \[[@B69]\]. In the interpretation of experimental data, periodic behaviour may or may not exist when evaluating alterations in spectral power in response to intervention. The assumption of stationarity may also be violated with prolonged signal recording. Changes in posture, level of activity and sleep patterns will alter the LF and HF components of spectral analysis \[[@B70]\]. Spectral analysis is more sensitive to the presence of artifact and/or ectopy than time domain statistical methods. In addition, given that different types of Holter monitors may yield altered LF signals \[[@B71]\], it is essential to ensure that the sampling frequency of the monitor used to read QRS complexes does not contribute to error in the variability analysis \[[@B11],[@B72]\]. Thus, the performance and interpretation of spectral analysis must incorporate these limitations. Recommendations based upon the stationarity assumption include the following \[[@B11]\]: short-term and long-term spectral analyses must be distinguished; long-term spectral analyses are felt to represent averages of the alterations present in shorter term recordings and may hide information; traditional statistical tests should be used to test for stationarity when performing spectral analysis; and physiological mechanisms that are known to influence HRV throughout the period of recording must be controlled.
Time spectrum analysis
----------------------
Another means to address the stationarity assumption inherent in the Fourier transform is to evaluate the power spectral density function for short periods of time when stationarity is assumed to be present, and subsequently follow the evolution of the power spectrum over time \[[@B73]\]. This combined time varying spectral analysis allows the continuous evaluation of change in variability over time. One can use sequential spectral approach \[[@B74]\], Wavelet analysis \[[@B75]\], the Wigner-Ville technique or Walsh transforms, all of which provide an analysis of frequency alteration over time, which is useful in clinical applications \[[@B37]\]. For example, time frequency analysis has demonstrated increased LF HRV power during waking hours (considered primarily a marker of sympathetic tone) and increased HF HRV during sleep (thought to be related to respiratory fluctuations secondary to vagal tone) \[[@B70]\]. The authors hypothesized that observations of increased cardiovascular events occurring during waking hours may be secondary to sudden increases in sympathetic activity. However, spectral analysis should not be the only form of variability analysis because there are patterns of variation that are present across the frequency spectrum, involving long-range organization and complexity.
Power law
=========
Definition
----------
Power law behaviour describes the dynamics of widely disparate phenomena, from earthquakes, solar flares and stock market fluctuations to avalanches. These dynamics are thought to arise from the system itself; indeed, the theory of self-organized criticality has been suggested to represent a universal organizing principle in biology \[[@B76]\]. It is illustrative to discuss the frequency distribution of earthquakes. A plot of the log of the power of earthquakes (i.e. the Richter scale) against the log of the frequency of their occurrence reveals a straight line with negative slope of -1. Thus, the probability of an earthquake may be determined for a given magnitude, occurring in a given region over a period of time, providing a measure of earthquake risk. In areas of increased earthquake activity, the line is shifted to the right, but the straight line relationship (and the slope) remains unchanged. Thus, the vertical distance between the straight line log--log frequency distributions or the intercept provides a measure of the difference in probabilities of an earthquake of all magnitudes between the two regions. Power law behaviour in physics, ecology, evolution, epidemics and neurobiology has also been described and reviewed \[[@B77]\].
Power laws describe dynamics that have a similar pattern at different scales, namely they are \'scale invariant\'. As we shall see, detrended fluctuation analysis (DFA) is also a technique that characterizes the pattern of variation across multiple scales of measurement. A power law describes a time series with many small variations, and fewer and fewer larger variations; and the pattern of variation is statistically similar regardless of the size of the variation. Magnifying or shrinking the scale of the signal reveals the same relationship that defines the dynamics of the signal, analogous to the self-similarity seen in a multitude of spatial structures found in biology \[[@B78]\]. This scale invariant self-similar nature is a property of fractals, which are geometric structures pioneered and investigated by Benoit Mandelbrot \[[@B79]\]. Akin to a coastline, fractals represent structures that have no fixed length; their length increases with increased precision (magnification) of measurement, a property that confers a noninteger dimension to all fractals. In the case of a coastline, the fractal dimension lies between 1 (a perfectly straight coastline) and 2 (an infinitely irregular coastline). With respect to time series, the pattern of variation appears the same at different scales (i.e. magnification of the pattern reveals the same pattern) \[[@B78]\]. This is often referred to as fractal scaling. Of principal interest to clinicians and scientists is that one can measure the long range correlations that are present in a series of data and, as we shall see, measure the alterations present in states of illness.
Calculation
-----------
As with frequency domain analysis (discussed above), the first step in the evaluation of the power law is the calculation of the power spectrum. This calculation, based on the fast Fourier transform (defined above), yields the frequency components of a series in time. By plotting a log--log representation of the power spectrum (log power versus log frequency), a straight line is obtained with a slope of approximately -1. As the frequency increases, the size of the variation drops by the same factor, and this patterns exists across many scales of frequency and variation, within a range consistent with system size and signal duration. Mathematically, power law behaviour is scale invariant; if a variable x is replaced by Ax\', where A is a constant, then the fundamental power law relationship remains unaltered. A straight line is fitted using linear regression, and the slope and intercept are obtained (see Appendix 1).
Interpretation and clinical implications
----------------------------------------
Power law behaviour has been observed for numerous physiological parameters and, relevant to clinicians, a change in intercept and slope is both present and prognostic in illness. Power law behaviour describes fluctuations in heart rate (first noted by Kobayashi and Musha \[[@B80]\]), foetal respiratory rate in lambs \[[@B81]\], movement of cells \[[@B82]\] and more. Power laws in pulmonary physiology were recently reviewed \[[@B83]\], noting a link between fractal temporal structure and fractal spatial anatomy. Alterations in the heart rate power law relationship (decreased or more negative slope) are present with ageing in healthy humans \[[@B84]\] as well as in patients with coronary artery disease \[[@B85]\]. Illness also confers changes in heart rate power law relationship. In over 700 patients with a recent myocardial infarction, as compared with age-matched control individuals, a steeper (more negative slope) power law slope was the best predictor of mortality evaluated \[[@B86]\]. In a random sample of 347 healthy individuals aged 65 years or older, a steep slope in the power law regression line (ß \< -1.5) was the best univariate predictor of all-cause mortality, with an odds ratio for mortality at 10 years of 7.9 (95% confidence interval 3.7--17.0; *P*\< 0.0001) \[[@B87]\]. Furthermore, only power law slope and a history of congestive heart failure were multivariate predictors of mortality in this cohort. Thus, changes in both slope and intercept have been documented to provide prognostic information in diverse patient populations.
Given that power law analysis is performed by plotting the log of spectral power versus the log of frequency using data derived from spectral analysis, what is the relationship between the two methods of characterizing variability? Although derived using the same data, the two methods assess different characteristics of signals. Spectral analysis measures the relative importance or contribution of specific frequencies to the underlying signal, whereas power law analysis attempts to determine the nature of correlations across the frequency spectrum. These analyses may have distinct and complementary clinical significance; for example, investigations of multiple HRV indices in patients following myocardial infarction \[[@B86]\] and in paediatric ICU patients \[[@B33]\] found that the slope of the power law had superior ability to predict mortality and organ failure, respectively, as compared with traditional spectral analysis.
Limitation
----------
Because determining power law behaviour requires spectral analysis, namely the determination of the frequency components of the underlying signal, the technique becomes problematic when applied to nonstationary signals. This limitation makes it difficult to draw conclusions regarding the mechanisms that underlie the alteration in dynamics observed in different patient groups. In addition, because power law behaviour measures the correlation between a large range of frequencies, it requires prolonged recording to achieve statistical validity. Nonetheless, as with the time and frequency domain analysis, valid clinical distinctions based on power law analysis have been demonstrated.
Specifically addressing the problem of nonstationarity, there is a problem in differentiating variations in a series of data that arise as an epiphenomenon of environmental stimuli (such as the effect of change in posture on heart rate dynamics) from variations that intrinsically arise from the dynamics of a complex nonlinear system \[[@B88],[@B89]\]. Both lead to a nonstationary variations but nonetheless represent clinically distinct phenomena. The subsequent technique was developed to address this issue.
Detrended fluctuation analysis
==============================
Definition
----------
Introduced by Peng and coworkers \[[@B90]\], DFA was developed specifically to distinguish between intrinsic fluctuations generated by complex systems and those caused by external or environmental stimuli acting on the system \[[@B88]\]. Variations that arise because of extrinsic stimuli are presumed to cause a local effect, whereas variations due to the intrinsic dynamics of the system are presumed to exhibit long-range correlation. DFA is a second measure of scale invariant behaviour because it evaluates trends of all sizes, trends that exhibit fractal properties (similar patterns of variation across multiple time scales). A component of the DFA calculation involves the subtraction of local trends (more likely related to external stimuli) in order to address the correlations that are caused by nonstationarity, and to help quantify the character of long-range fractal correlation representing the intrinsic nature of the system.
Calculation
-----------
The calculation of DFA involves several steps (see Appendix 1). The analysis is performed on a time series, for example the intervals between consecutive heartbeats, with the total number of beats equal to N. First, the average value for all N values is calculated. Second, a new (integrated) series of data (also from 1 to N) is calculated by summing the differences between the average value and each individual value. This new series of values represents an evaluation of trends; for example, if the difference between individual NN intervals and the average NN interval remains positive (i.e. the interval between heartbeats is longer than the average interbeat interval), then the heartbeat is persistently slower than the mean, and the integrated series will increase. This trend series of data displays fractal, or scaling behaviour, and the following calculation is performed to quantify this behaviour. In this third step, the trend series is separated into equal boxes of length n, where n = N/(total number of boxes); and in each box the local trend is calculated (a linear representation of the trend function in that box using the least squares method). Fourth, the trend series is locally \'detrended\' by subtracting the local trend in each box, and the root mean square of this integrated, detrended series is calculated, called F(n). Finally, it is possible to graph the relationship between F(n) and n. Scaling or fractal correlation is present if the data is linear on a graph of log F(n) versus log(n). The slope of the graph has been termed a, the scaling exponent. A single scaling exponent represents the limit as N and n approach infinity; however, applicable to real life data sets, the linear relationship between log F(n) and log n has been noted to be distinct for small n (n \< 11) and large n (11 \< n \> 10,000), yielding two lines with two slopes, labelled the scaling exponents a~1~and a~2~, respectively. For a more detailed description, see Appendix 1; excellent descriptions of the calculation of DFA may be found elsewhere \[[@B34],[@B88]\].
Interpretation and clinical applications
----------------------------------------
DFA offers clinicians the advantage of a means to investigate long range correlations within a biological signal due to the intrinsic properties of the system producing the signal, rather than external stimuli unrelated to the \'health\' of the system. In addition, the calculation is based on the entire data set and is \'scale free\', offering greater potential to distinguish biological signals based on scale specific measures \[[@B91]\]. Theoretically, the scaling exponent will vary from 0.5 (random numbers) to 1.5 (random walk), but physiological signals yield scaling exponents close to 1. A scaling exponent greater than 1.0 indicates a loss in long range scaling behaviour and a pathological alteration in the underlying system \[[@B88]\]. The technique was initially applied to detect long range correlations in DNA sequences \[[@B90]\] but has been increasingly applied to biological time signals.
As with other techniques of variability analysis, DFA has been used to evaluate cardiovascular variation. Elderly individuals \[[@B92]\], patients with heart disease \[[@B93]\] and asymptomatic relatives of patients with dilated cardiomyopathy who have enlarged left ventricles \[[@B94]\] all exhibit a loss of \'fractal scaling\'. To date, a~1~has demonstrated greater clinical discrimination of distinct heart rate data sets, as compared with a~2~\[[@B88],[@B94]\]. For example, a~1~provided the best means of distinguishing patients with stable angina from age-matched control individuals; however, the correlation did not extend to angiographical severity of coronary artery disease \[[@B95]\]. In a retrospective evaluation of 2 hour ambulatory ECG recordings in the Framingham Heart Study \[[@B96]\], DFA was found to carry additional prognostic information that was not provided by traditional time and frequency domain measures. In a retrospective comparison between 24 hour HRV analysis using several techniques in patients post-myocardial infarction with or without inducible ventricular tachyarrhythmia \[[@B97]\], a decrease in the scaling exponent a~1~was the strongest predictor of risk for ventricular arrhythmia. DFA was superior to spectral analysis in the analysis of HRV alteration in patients with sleep apnoea \[[@B98]\]. In a prospective, multicentre evaluation of HRV post-myocardial infarction, reduced short-term scaling exponent (a~1~\< 0.65) was the single best predictor of subsequent mortality \[[@B99]\]. In patients who had undergone coronary artery bypass surgery, reduced short-term scaling exponent in the postoperative period was the best predictor of a longer ICU stay, as compared with other HRV measures \[[@B100]\]. Thus, alteration in DFA scaling exponent (both increased and decreased) of heart rate fluctuation provides additional diagnostic and prognostic information that appears independent of time and frequency domain analysis.
In addition to cardiovascular variation, DFA has increasingly been applied to investigate other systems. Alterations in the scaling exponent of respiratory variation (inter-breath intervals) have been noted in elderly individuals \[[@B101]\]; and the finding of long-range correlations in breath--breath end-tidal carbon dioxide and oxygen fluctuations in healthy infants introduce novel avenues for investigation of respiratory illness \[[@B102]\]. Remarkably, the scaling properties of temperature measurements (every 10 min for 30 hours) are altered in association with ageing \[[@B103]\]. In addition, DFA provides meaningful information on EEG signals and has been utilized to distinguish normal individuals from stroke patients \[[@B104],[@B105]\].
Advantages and limitations
--------------------------
The principal advantage to DFA is the lack of confounding due to nonstationary data. DFA is readily calculated using a computer algorithm available through a cooperative academic internet resource, Physionet <http://www.physionet.org>\[[@B106]\]. Although DFA represents a novel technological development in the science of variability analysis and has proven clinical significance, whether it offers information distinct from traditional spectral analysis is debated \[[@B107]\]. Data requirements are greater than with other techniques and have been suggested to include at least 8000 data points, as noted by empirical observations \[[@B88]\]. It is inappropriate to simply \'run\' the DFA algorithm blindly on data sets; for example, a clear shift in the state of the cardiovascular system (e.g. spontaneous atrial fibrillation) would prohibit meaningful DFA interpretation. Finally, although appealing in order to simplify clinical comparison, the calculation of two scaling exponents (one for small and one for large n) represents a somewhat arbitrary manipulation of the results of the analysis. The assumption that the same scaling pattern is present throughout the signal remains flawed, and therefore techniques without this assumption are being developed and are referred to as multifractal analysis.
Multifractal analysis
---------------------
DFA is a monofractal technique, in that the assumption is that the same scaling property is present throughout the entire signal. Multifractal techniques provide multiple, possibly infinite exponents, such that the analysis produces a spectrum rather than a discrete value. For example, wavelet analysis is a multifractal analysis technique similar to DFA, which is capable of distinguishing the heart rate dynamics of patients with congestive heart failure from healthy control individuals \[[@B34]\]; a full discussion of multifractality of biological signals can be found elsewhere \[[@B108]\]. A separate technique recently introduced by Echeverría and colleagues \[[@B109]\] utilizes an a--ß filter (a technique imported from real-time radar tracking technology) to characterize heart rate fluctuations. Those authors suggested that this representation provides a superior means of identifying clinically distinct signals, and in order to demonstrate this they evaluated both theoretically and experimentally derived data sets. It remains unclear whether the added complexity and theoretical advantages of these techniques will afford consistent clinically significant improvements in the ability to distinguish physiological from pathological rhythms.
Entropy analysis
================
Definition
----------
Entropy is a measure of disorder or randomness, as embodied in the Second Law of Thermodynamics, namely the entropy of a system tends toward a maximum. In other words, states tend to evolve from ordered statistically unlikely configurations to configurations that are less ordered and statistically more probable. For example, a smoke ring (ordered configuration) diffuses into the air (random configuration); the spontaneous reverse occurrence is statistically improbable to the point of impossibility. Entropy is the measure of disorder or randomness. Related to time series analysis, approximate entropy (ApEn) provides a measure of the degree of irregularity or randomness within a series of data. It is closely related to Kolmogorov entropy, which is a measure of the rate of generation of new information \[[@B110]\]. ApEn was pioneered by Pincus \[[@B111]\] as a measure of system complexity; smaller values indicate greater regularity, and greater values convey more disorder, randomness and system complexity.
Calculation
-----------
In order to measure the degree of regularity of a series of data (of length N), the data series is evaluated for patterns that recur. This is performed by evaluating data sequences of length m, and determining the likelihood that other runs in the data set of the same length m are similar (within a specified tolerance r); thus two parameters, m and r, must be fixed to calculate ApEn. Once the frequency of occurrence of repetitive runs is calculated, a measure of their prevalence (negative average natural logarithm of the conditional probability) is found. ApEn then measures the difference between the logarithmic frequencies of similar runs of length m and runs with the length m+1. Small values of ApEn indicate regularity, given that the prevalence of repetitive patterns of length m and m+1 do not differ significantly and their difference is small. A derivation is included in Appendix 1, and a more comprehensive description of ApEn may be found elsewhere \[[@B112]-[@B114]\].
Interpretation and clinical application
---------------------------------------
ApEn is representative of the rate of generation of new information within a biological signal because it provides a measure of the degree of irregularity or disorder within the signal. As such, it has been used as a measure of the underlying \'complexity\' of the system producing the dynamics \[[@B111],[@B112],[@B115]\]. The clinical value of a measure of \'complexity\' is potentially enormous because complexity appears to be lost in the presence of illness \[[@B114],[@B116],[@B117]\] (discussed in greater detail below).
As with other means of characterizing biological signals, ApEn has been most extensively studied in the evaluation of heart rate dynamics. Heart rate becomes more orderly with age and in men, exhibiting decreased ApEn \[[@B118]\]. Heart rate ApEn has demonstrated the capacity to predict atrial arrhythmias, including spontaneous \[[@B119]\] and postoperative atrial fibrillation after cardiac surgery \[[@B120]\], and to differentiate ventricular arrhythmias \[[@B121]\]. Heart rate ApEn is decreased in infants with aborted sudden infant death syndrome \[[@B122]\]; among adults, postoperative patients with ventricular dysfunction \[[@B123]\] and healthy individuals infused with endotoxin \[[@B124]\] exhibit reduced heart rate ApEn.
Because ApEn may be applied to short, noisy data sets, it was applied to assess the variation of parameters in which frequent sampling is more difficult (e.g. a blood test is necessary) and a paucity of data exists. This was most apparent in the evaluation of endocrine variability, as demonstrated in the following investigations. By applying ApEn to measurements of growth hormone (GH) every 5 min for 24 hours in healthy control individuals and patients with acromegaly, reduced orderliness (i.e. increased ApEn) was observed in acromegaly \[[@B125]\]; and normalization of GH ApEn values was demonstrated after pituitary surgery for acromegaly \[[@B126]\]. Increased disorderliness has been observed in insulin secretion in healthy elderly individuals as compared with young control individuals (insulin measured every minute for 150 min) \[[@B127]\], and in first-degree relatives of patients with non-insulin-dependent diabetes mellitus (insulin measured every minute for about 75 min) \[[@B128]\]. ApEn of adrenocorticotrophic hormone, GH, prolactin and cortisol levels (sampled every 10 min for 24 hours) is altered in patients with Cushing\'s disease \[[@B129],[@B130]\]. Finally, altered dynamics of parathyroid hormone pulsatile secretion has been demonstrated in osteoperosis and hyperparathyroidism \[[@B131]\].
ApEn has also been used to evaluate neurological, respiratory and, recently, temperature variability. ApEn offers a means of assessing the depth of anaesthesia \[[@B132]-[@B134]\], and ApEn of tidal volume respiratory rate has been evaluated in patients with respiratory failure weaning from mechanical ventilation \[[@B135]\]. Alterations in respiratory variability are present in psychiatric illness; for example, increased entropy of respiration has been observed in patients with panic disorder \[[@B136]\]. Comparing chest wall movement and EEG activity in healthy individuals, sleep (stage IV) produced more regular breathing and more regular EEG activity \[[@B137]\]. Finally, demonstrating the remarkable potential and novel applications of variability analysis, ApEn of temperature measurements (every 10 min for 30 hours) revealed increased regularity and decreased complexity associated with age \[[@B103]\].
Advantages and limitations
--------------------------
ApEn statistics may be calculated for relatively short series of data, a principal advantage in their application to biological signals. Referring to both theoretical analysis and clinical applications, Pincus and Golberger \[[@B112]\] concluded that m = 2 and r = 10--25% of the standard deviation of all the N values, and an N value of 10^m^, or preferably 30^m^, will yield statistically reliable and reproducible results (i.e. 100--900 data points). Pincus \[[@B114]\] also reported that ApEn is applicable to any system with at least 50 data points. In contrast to time domain measures of variability, which are independent of the sequence of the data set, ApEn required an evaluation of vectors representing consecutive data points, and thus the order of the data is integral to the calculation of ApEn and must be preserved during data harvest. Significant noise or nonstationary data compromise meaningful interpretation of ApEn \[[@B113]\]; therefore, it should not be used as the only means to measure signal characteristics.
Sample and multiscale entropy
-----------------------------
An inherent bias within the ApEn calculation exists because the algorithm counts similar sequences to a given sequence of length m, including counting the sequence itself (to avoid the natural logarithm of 0 within the calculations). As a result, ApEn can be sensitive to the size of the data set, giving inappropriately low values when the total number of data points is low; this, and a lack of consistency in differentiating signals when m and r are altered, have led to the development of a new family of statistics named sample entropy (SampEn), in which self-matches are excluded in the analysis \[[@B110]\]. SampEn has the advantage of being less dependent on the length of the data series in question, and has been applied to heart rate fluctuations in the paediatric ICU \[[@B138]\]. Finally, because both ApEn and SampEn are noted to evaluate differences between sequences of length m and m+1, they evaluate regularity on one scale only, the shortest one, and ignore other scales. Thus, given the temporal complexity of biological signals on multiple scales, a novel technique, multiscale entropy, was developed as a more robust measure of complexity \[[@B139]\]. Initial investigations of multiscale entropy have been promising \[[@B140]\], but comprehensive evaluation remains to be performed.
Summary and discussion of variability techniques
================================================
The preceding sections highlight the considerable range of techniques that have been developed to characterize biological signals. Each with distinct theoretical background and significance, they contribute complementary information regarding signal characteristics. Time domain measures of variation represent an evaluation of overall, short-term or long-term variation, and are clinically proven as a means of identifying clinically significant alterations in biological signals, in particular with cardiovascular variability. Frequency domain analysis also has prognostic value, and has been useful in demonstrating the importance of sympathovagal balance in regulating HF and LF cardiovascular oscillations. Power law analysis contributes an analysis of fractal, long range correlations, allowing distinction between physiological and pathological signals with the slope and intercept of the power law. DFA also represents a means of detecting long range correlations, and is less bound by the stationarity assumption inherent to the other techniques. By measuring the degree to which sequences of data repeat themselves within a signal, ApEn provides a measure of signal irregularity, related to the rate of production of new information. Although techniques have shown consistent prognostic capacity, prediction of mortality is not the sole virtue of HRV analysis; separate techniques also may clarify mechanisms of disease \[[@B141]\]. Attempts to characterize biological signals should incorporate the \'toolkit\' of techniques discussed in this review as well as the publication of raw data and code to facilitate comparison and development of this still young, exciting science \[[@B117]\].
Interpretation and significance of altered variability
------------------------------------------------------
Following this review of the technology of variability analysis, the meaning of altered variability in biological signals must be addressed. A synthesis of the multiple but consistent theories regarding the significance of altered variability is presented to assist in the clinical application of this novel technology. A leading investigator within this field, Goldberger \[[@B142]\] proposed that increased regularity of signals represents a \'decomplexification\' of illness, citing numerous examples of illness states with increased regularity of rhythms. For example, Cheyne--Stokes respiration, Parkinsonian gait, loss of EEG variability, preterminal cardiac oscillations, neutrophil count in chronic myelogenous leukaemia and fever in Hodgkin\'s disease all exhibit periodic, more regular variation in the dynamics of disease states \[[@B142]\]. Given that scale invariance is believed to be a central organizing principle of physiological structure and function, breakdown in this scale invariant, fractal behaviour, leads to uncorrelated randomness or more predictable behaviour, both representing a pathological alteration to the underlying system \[[@B78],[@B84]\]. Thus, health is characterized by \'organized variability\' and disease is defined by decomplexification, increased regularity and reduction in variability.
In contrast to the \'decomplexification\' hypothesis, Vaillancourt and Newell \[[@B143]\] noted increased complexity and increased approximate entropy in several disease states, including acromegaly and Cushing\'s disease, and hypothesized that disease may manifest with increased or decreased complexity, depending on the underlying dimension of the intrinsic dynamic (e.g. oscillating versus fixed point). In a rebuttal, Goldberger \[[@B142]\] noted that increased complexity demonstrated by lower entropy (specifically ApEn) requires corroboration by other techniques, given potential problems with using ApEn as the only technique to assess variability. A rebuttal to the rebuttal (all published concurrently) \[[@B144]\] noted that others accept the fundamental premise that increased and decreased variability occur in disease.
In addition to the discussion regarding complexity, increased short-term variation in airway calibre in patients with asthma is observed, and reproduced experimentally with activation of airway smooth muscle with inhaled methacholine \[[@B5]\]. Given that smooth muscle activation is associated with increased metabolic rate, energy dissipation and an increased likelihood of statistically unlikely airway configurations, Macklem\'s hypothesis states that asthma is a disease of higher energy dissipation, greater distance from thermodynamic equilibrium, lower entropy and greater variation \[[@B5]\]. This suggests that health is defined by a certain distance from thermodynamic equilibrium; too close (decreased variation, too little energy dissipation, low entropy) or too far (increased variation and energy dissipation, high entropy) both represent pathological alterations.
The science of complex systems is intimately related to variability analysis. Taking a broad systems based interpretation, the human organism is a complex system or, more accurately, it is a complex system of complex systems. The host response to sepsis, shock, or trauma is an example of a biological complex system that is readily apparent to intensivists \[[@B3]\]. Every complex system has \'emergent\' properties, which define its very nature and function, including the presence of health versus illness. Variability or patterns of change over time (in addition to connectivity or patterns of interconnection over space) represent technology with which to evaluate the emergent properties of a complex system, which may be physiological or pathological \[[@B3]\]. It is possible to conceive complex systemic host response in a phase space of variability parameters, in which health represents stable \'holes\' in space, exhibiting marked systemic stability accompanied by specific patterns of variability (and connectivity). Illness represents an alteration from health, separate \'holes\' with distinct patterns of variability. Often, it takes a major insult to change a stable healthy state to an illness state, which may have varying degrees of stability. It is within this complex systems conception of health and illness that the clinical utility of variability analysis may be appreciated.
How can variability analysis improve outcome in the intensive care unit?
------------------------------------------------------------------------
What does variability analysis offer that conventional monitoring does not? What is the clinical utility of this technology? We propose that multi-system continuous variability analysis offers the intensivist a unique monitoring tool that is capable of improving prognostication and directing therapeutic intervention. Intuitively, there is additional information in this analysis. Variability analysis tracks specific patterns of change in individual parameters over time (akin to calculating the first derivative or velocity in calculus). Monitoring patterns of change in variability continuously over time offers an additional dimension of analysis (akin to a second derivative evaluation or acceleration). Just as monitoring individual system variability offers an evaluation of the underlying individual system producing those dynamics, evaluating multisystem variability provides an evaluation of the whole, namely the systemic host response. By using variability analysis at different time points or, more powerfully, continuously over time, it is theoretically possible to track the \'system state\' over time. Then, by selecting patients according to pathological patterns of variability and pursuing interventions with a therapeutic response or physiological alteration in variability, we hypothesize that outcomes in critically ill patients may be improved.
Why does this individualized variability directed therapy offer exciting clinical potential? First, as the host response is a complex system, response to intervention in individual patients is unpredictable, although response to an intervention may be statistically beneficial for a cohort of patients. Thus, only by evaluating the response to intervention in individual patients can it be ascertained that the intervention is beneficial in those patients. Interventions that have not proven beneficial for the \'average\' patient may still be beneficial in selected individual patients, in whom pathological variability is both present and improved by therapy. In summary, continuous, individualized, variability directed, goal directed therapeutic intervention has numerous theoretical advantages over conventional epidemiological cohort analysis evaluating response to a single intervention given to a heterogeneous population of patients. This technology is well suited to the ICU, in which real-time, continuous, digital physiological data acquisition (including waveform analysis) has been demonstrated \[[@B145]-[@B147]\]. Unresolved questions include whether, how and when is it possible to convert pathological to physiological variability, to prod our patients from illness to health. Answering these questions will determine the impact variability analysis has on ICU patient outcome.
Conclusion
==========
The science of analyzing biological signals has undergone tremendous growth over the past decade, with the development of advanced computational methods that characterize the variation, oscillation, complexity and regularity of signals. These methods were developed in response to theoretical limitations of the others; however, all appear to have clinical significance. There is no consensus that any single technique is the single best means of characterizing and differentiating biological signals; rather, investigators agree that multiple techniques should be performed simultaneously to facilitate comparison between methods, techniques and studies. Variability analysis represents a novel means to evaluate and treat individual patients, suggesting a shift from epidemiological analytical investigation to continuous individualized variability analysis. Existing literature documents the clinical value of measuring variability to provide diagnostic, prognostic and pathophysiological information; future research must utilize this technology to improve care and the outcomes of our patients.
Key messages
============
• A complex systems paradigm provides insights regarding research and treatment of critically ill patients.
• Variability analysis is the science of measuring the degree and character of patterns of variation of a time-series of a biologic parameter, in order to evaluate the state of the underlying complex system responsible for the biologic signal.
• Using techniques that measure overall variation, frequency contribution, scale-invariant variation and degree of disorder, altered variability in consistently present in illness states, and the degree of alteration provides a measure of prognosis.
• Using continuous multiogan variability analysis (CMVA), we hypothesize that goal-directed variability-directed therapeutic intervention will improve outcome and reduce mortality in critically ill patients, a novel individualized systems approach that complements analytical basic science and epidemiologic population science.
Appendix 1: techniques of variability analysis
==============================================
Variability analysis
--------------------
The description of means to characterize and differentiate biological signals, or sequences of data in time produced by biological systems, is referred to as \'variability analysis\'. For example, a heart rate recording may be considered a series of intervals between consecutive heart beats, referred to as NN intervals (interval between consecutive normal sinus beats) or RR intervals (interval between consecutive R waves on an ECG). With the goal of providing a single means of characterizing a whole series of data, the following techniques were developed to perform variability analysis and applied to clinical data sets.
Time domain analysis
--------------------
Considered the simplest means of measuring variability, time domain analysis involves performing a statistical analysis of data expressed as a sequence in time. For example, SDNN (the standard deviation of NN intervals) has been used as a measure of HRV; greater variation yeilds higher standard deviation. Standard deviation is the square root of the average of the squared differences from the mean. SDANN (standard deviation of the average NN interval calculated over 5-min intervals within the entire period of recording) is a measure of longer term variation because the averaging process removes beat-to-beat variations. In contrast, the following variables were devised as a measure of short-term variation: RMSSD (square root of the mean squared differences of consecutive NN intervals), NN50 (number of pairs of adjacent NN intervals differing by more than 50 ms), and pNN50 (proportion of NN intervals differing by more than 50 ms = NN50 divided by total number of NN intervals).
Frequency domain analysis
-------------------------
Physiological data collected as a series in time may be considered a sum of rhythmic oscillations with distinct frequencies. Conversion from time domain to frequency domain analysis is performed most commonly using the Fourier transform, which decomposes the signal into a series of sine and cosine waves with frequencies that are multiples of the fundamental frequency (reciprocal of the time length to the input data record); the fast Fourier transform is a discrete Fourier transform that reduces the number of computations. The result of the Fourier transform is a complex number (a number multiplied by the square root of -1) for each frequency, the square of which is considered the spectral power of that frequency. The whole process is called spectral analysis, because it provides an evaluation of the spectral power (amplitude) of the contributing frequencies of an underlying signal.
Power law analysis
------------------
Power law behaviour may be described by the following equation:
F(x) = ax^ß^
Where a and ß are constants. Taking the logarithm of both sides, a straight line (graph log f \[x\] versus log x) with slope ß and intercept log a is revealed:
Log f(x) = log (ax^ß^) = log a + log x^ß^= log a + ß log x
Thus, power law behaviour is scale invariant; if a variable x is replaced by Ax\', where A is a constant, then the fundamental power law relationship remains unaltered. If dynamics follow a power law, a log--log representation of the power spectrum (log power versus log frequency) reveals a straight line, always within a defined range consistent with the size and duration of the system. The straight line is fitted using linear regression, and the slope ß and intercept can readily be obtained. When ß = -1, the dynamics are described as 1/f noise. Power law behaviour describes the dynamics of widely disparate phenomena, including heart rate fluctuations, inter-breath intervals, earthquakes, solar flares, stock market fluctuations, and avalanches.
Detrended fluctuation analysis
------------------------------
Variations that arise because of extrinsic stimuli are presumed to cause a local effect, whereas variations due to the intrinsic dynamics of the system are presumed to exhibit long range correlation. DFA attempts to quantify the presence or absence of long range scale-invariant (fractal) correlation.
The first step in the technique to calculate DFA is to map a biological signal, such as a series of heart beats, to an integrated series. The integrated series is calculated by the sum of the differences between individual inter-beat intervals represented as NN~i~and the average interbeat interval for the whole data set, equal to NN~ave~.
y(k) = S~i\ =\ 1~^N^(NN~i~- NN~ave~)
This series y(k) represents an evaluation of trends; for example, if the difference NN~i~- NN~ave~remains negative (heart beat is persistently faster than the mean), then y(k) increases as k increases. This trend function y(k) is then separated into equal boxes of length n, where n = N/(total number of boxes). In each box, the local trend y~n~(k) is calculated as a linear representation of the function y(k) in that box using the least squares method. Least squares analysis involves the principle of optimization of the estimate based on minimizing the sum of the squared differences from the values predicted by the model. The series y(k) is then \'detrended\' by subtracting the local trend y~n~(k). The root mean square of this integrated and detrended series is represented by the following:
F(n) = v (1/N S~k\ =\ 1~^N^\[y(k)^2^- y~n~(k)^2^\])
By performing this analysis for all values of n, it is possible to calculate the relationship between F(n) and n. Scaling or fractal correlation is present if the data is linear on a graph of log F(n) versus log(n). The slope of the graph has been termed a, the scaling exponent, which will vary from 0.5 (white noise or uncorrelated random data) to 1.5 (Brownian noise or integrated white noise or random walk). When a = 1, behaviour corresponds to the 1/f noise. As a increases above 1 to 1.5, behaviour is no longer determined by a power law. Because the linear relationship between log F(n) and log(n) appears to have two distinct linear segments, one for small (n \< 11) and large n (n \> 11), the slopes of both lines are calculated separately and termed a~1~and a~2~, respectively; repeatedly, a~1~has proven superior to a~2~in terms of prognostic ability.
Approximate entropy
-------------------
ApEn is a measure of \'irregularity\'; smaller values indicate a greater chance that a set of data will be followed by similar data (regularity), and a greater value for ApEn signifies a lesser chance of similar data being repeated (irregularity). To calculate ApEn of a series of data, the data series is evaluated for patterns that recur. This is performed by evaluating data sequences or runs of length m, and determining the likelihood that other runs of length m are similar, within a tolerance r. Thus, two parameters, m and r, must be fixed to calculate ApEn. Increased regularity is associated with illness.
The following is a description of the calculation of ApEn. Given any sequence of data points u(i) from i = 1 to N, it is possible to define vector sequences x(i), which consist of length m and are made up of consecutive u(i), specifically defined by the following:
x(i) = (u \[i\], u \[i + 1\], \... u \[i + m - 1\])
In order to estimate the frequency that vectors x(i) repeat themselves throughout the data set within a tolerance r, the distance d(x \[i\],x \[j\]) is defined as the maximum difference between the scalar components x(i) and x(j). Explicitly, two vectors x(i) and x(j) are \'similar\' within the tolerance or filter r (i.e. d(x \[i\],x \[j\]) = r) if the difference between any two values for u(i) and u(j) within runs of length m are less than r (i.e. \|u(i + k) - u(j+k)\| = r for 0 = k = m). Subsequently, C~i~^m^(r) is defined as the frequency of occurrence of similar runs m within the tolerance r:
C~i~^m^(r) = (number of j such that d(x \[i\],x \[j\]) = r)/(N - m - 1), where j = (N - m - 1)
Taking the natural logarithm of C~i~^m^(r), F^m^(r) is defined as the average of ln C~i~^m^(r):
F^m^(r) = S~i~ln C~i~^m^(r)/(N - m - 1), where S~i~is a sum from I = 1 to (N - m - 1)
F^m^(r) is a measure of the prevalence of repetitive patterns of length m within the filter r.
Finally, approximate entropy, or ApEn(m,r,N), is defined as the natural logarithm of the relative prevalence of repetitive patterns of length m as compared with those of length m + 1:
ApEn(m,r,N) = F^m^(r) - F^m+1^(r)
Thus, ApEn(m,r,N) measures the logarithmic frequency that similar runs (within the filter r) of length m also remain similar when the length of the run is increased by 1. Thus, small values of ApEn indicate regularity, given that increasing run length m by 1 does not decrease the value of F^m^(r) significantly (i.e. regularity connotes that F^m^\[r\] \~ F^m+1^\[r\]). ApEn(m,r,N) is expressed as a difference, but in essence it represents a ratio; note that F^m^(r) is a logarithm of the averaged C~i~^m^(r), and the ratio of logarithms is equivalent to their difference.
Competing interests
===================
None declared.
Abbreviations
=============
ApEn = approximate entropy; DFA = detrended fluctuation analysis; EEG = electroencephalogram; GH = growth hormone; HF = high frequency; HRV = heart rate variability; ICU = intensive care unit; LF = low frequency; NN50 = number of pairs of adjacent NN intervals differing by more than 50 ms; pNN50 = proportion of NN intervals differing by more than 50 ms; RMSSD = square root of the mean squared differences of consecutive NN intervals; SampEn = sample entropy; SDANN = standard deviation of the average NN interval calculated over 5 min intervals within the entire period of recording; SDNN = standard deviation of a series of NN intervals; ULF = ultralow frequency; VLF = very low frequency.
Acknowledgements
================
The authors would like to thank John Marshall, Paul Hébert, Farid Shamji, Donna Maziak, Sudhir Sundaresan, John Seely and Kathy Patterson for their valuable contributions, feedback and support.
Figures and Tables
==================
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Techniques to characterize variability
:::
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Variability analysis Description Advantages Limitations Output variables
------------------------------------ ----------------------------------------------------------------------------------------- -------------------------------------------------------------------------------------------------------------------------------------------------------- ---------------------------------------------------------------------------------------------------------------- --------------------------------------------------------------------------------------------------------------------------------------------------
Time domain Statistical calculations of consecutive intervals Simple, easy to calculate; proven clinically useful; gross distinction of high and low frequency variations Sensitive to artifact; requires stationarity; fails to discriminate distinct signals SD, RMSDD Specific to HRV: SDANN, pNNx
Frequency distribution (plot number of observations falling in selected ranges or bins) Visual representation of data; can fit to normal or log-normal distribution Lacks widespread clinical application; arbitrary number of bins Skewness (measures symmetry): positive (right tail) versus negative (left) Kurtosis (measures peakedness): flatter top (\<0) versus peaked (\>0)
Frequency domain Frequency spectrum representation (spectral analysis) Visual and quantitative representation of frequency contribution to waveform; useful to evaluate relationship to mechanisms; widespread HRV evaluation Requires stationarity and periodicity for validity; sensitive to artifact; altered by posture, sleep, activity Total power (area under curve) Specific to HRV: ULF (\<0.003 Hz), VLF (0.003--0.04 Hz), LF (0.04--0.15 Hz), HF (0.15--0.4 Hz)\
Time spectrum analysis
Scale invariant (fractal) analysis Power law: log power versus log frequency Ubiquitous biologic application; characterization of signal with single linear relationship; enables prognostication Requires stationarity and periodicity; requires large datasets Slope of power law Intercept of power law
DFA Identifies intrinsic variations 2°system (versus external stimuli), does not require stationarity Requires large datasets (\>8000 patients) Scaling exponent a~1~(n \< 11)\
Scaling exponent a~2~(n \> 11)\
a--ß filter
Entropy Measures the degree of disorder (information or complexity) Unique representation of data; requires fewest data points (100--900 patients) Needs to be complemented by other techniques ApEN SampEN Multi-scale entropy
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
ApEn, approximate entropy; DFA, detrended fluctuation analysis; HF, high frequency; HRV, heart rate variability; LF, low frequency; pNNx, proportion greater than x ms; RMSDD, root mean square of standard deviation; SampEn, sample entropy; SD, standard deviation; SDANN, standard deviation of 5 min averages; ULF, ultralow frequency; VLF, very low frequency.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Evidence for altered patterns of variability in illness states
:::
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Variability analysis Cardiac Respiratory Neurological Miscellaneous Critical care
---------------------- ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ -------------------------------------------------------------------------------------------------------------------------------------------- ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- ----------------------------------------------------------------------------------------------------------------------- --------------------------------------------------------------------------------------------------------------------------------------------
Time domain ?HRV ??mortality risk in elderly, CAD, post-MI, CHF and dilated cardiomyopathy \[14--24\] Altered frequency distribution of airway impedance in asthma \[5\]\
Altered respiratory variability (?kurtosis) in sleep apnoea \[148\]
Frequency domain Altered spectral HRV analysis?illness severity in cardiac disease (CHF \[50--52\], hypertension \[53,54\], CAD \[55,56\], angina \[57\], MI \[58\]) and noncardiac disease (hypovolaemia \[49\], chronic renal failure \[59\], diabetes mellitus \[60\], anaesthesia \[61\]) ?Total HRV, ?LF and ?LF/HF HRV following trauma \[149\], sepsis and septic shock in the ICU \[62,64,68,150,151\] and in ER patients \[63\]
Power law analysis Altered HRV power law (?HRV left shift and steeper slope) with age \[84\], CAD \[85\] and post-MI \[86\] ?Respiratory variability (right shift) in patients with asthma \[7\] ?Variability of foetal breathing with maternal alcohol intake \[152\] Altered variability in gait analysis \[153--155\] and postural control \[156\] with ageing and neurological disease Altered variability of mood?psychiatric illness \[157--159\] Haematological: altered leucocyte dynamics \[160,161\] observed in haematological disorders (e.g. cyclic neutropenia) Altered HRV power law (?HRV left shift)??mortality risk in paediatric ICU patients \[33\]
DFA Altered DFA scaling exponent?age \[92\], heart disease \[93--96\], post-ACBP \[100\], prearrhythmias \[97\], patients with sleep apnoea \[98\], and ?mortality risk post-MI \[99\] Altered respiratory variability (?DFA scaling exponent)?age\[101\] Temperature: altered temperature measurements?age\[103\] ?Heart rate DFA scaling exponent?septic shock\[162\] and procedures\[61\] in paediatric ICU patients
Entropy ?HR ApEn?age \[118\], ventricular dysfunction \[123\], occurs prior to arrhythmias \[119--121\] Greater respiratory irregularity in patients with panic disorder \[136\] Altered EEG entropy with anaesthesia\[132,163,164\] Endocrine: ?ApEn of GH \[125,126\], insulin \[127,128\], ACTH, GH, PRL \[129,130\], PTH \[131\]?age and/or illness ?HR ApEn?healthy individuals infused with endotoxin \[124\] ?TV ApEn in respiratory failure \[135\]
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
?, decreased; ?, increased; ?, is associated with; ACBP, aorto--coronary bypass procedure; ACTH, adrenocorticotrophic hormone; ApEn, approximate entropy; CAD, coronary artery disease; CHF, congestive heart failure; DFA, detrended fluctuation analysis; EEG, electroencephalogram; ER, emergency room; GH, growth hormone; HF, high frequency; HRV, heart rate variability; ICU, intensive care unit; LF, low frequency; MI, myocardial infarction; PRL, prolactin; PTH, parathyroid hormone; TV, tidal volume.
:::
|
PubMed Central
|
2024-06-05T03:55:54.833211
|
2004-9-22
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065053/",
"journal": "Crit Care. 2004 Sep 22; 8(6):R367-R384",
"authors": [
{
"first": "Andrew JE",
"last": "Seely"
},
{
"first": "Peter T",
"last": "Macklem"
}
]
}
|
PMC1065055
|
Introduction
============
Pesticide self-poisoning is a major clinical problem in many parts of the world \[[@B1],[@B2]\], probably killing about 300,000 people every year \[[@B3],[@B4]\]. Although most deaths occur in rural areas of the developing world \[[@B2]\], pesticide poisoning is also a problem in industrialized countries, where it may account for a significant proportion of the deaths from self-poisoning that do occur \[[@B5],[@B6]\].
The case fatality for self-poisoning in the developing world is commonly 10--20%, but for particular pesticides it may be as high as 50--70% \[[@B2]\]. This contrasts with the less than 0.3% case fatality ratio normally found for self-poisoning from all causes in Western countries. The causes of the high case fatality are multifactorial but include the high toxicity of locally available poisons, difficulties in transporting patients across long distances to hospital, the paucity of health care workers compared with the large numbers of patients, and the lack of facilities, antidotes, and training for the management of pesticide-poisoned patients \[[@B2],[@B4]\].
The problem is compounded by a lack of proven interventions with which to develop treatment protocols. In 2002 we set up a cohort study in the North Central Province of Sri Lanka that sought to follow 10,000 acutely self-poisoned patients prospectively. So far, over 6000 patients have been recruited, of whom more than 3000 have ingested pesticides. All patients are rapidly resuscitated on admission to hospital and stabilised according to a standard protocol.
Basic pharmacology and animal work suggests that early antagonism of pesticide toxicity should be associated with better outcomes \[[@B7],[@B8]\]. Although there are few studies on the subject, there is some evidence that patients in the developing world often die soon after admission (\[[@B9]\], and CGS Rao, unpublished data). The rapid and effective stabilisation and treatment of pesticide-poisoned patients on their admission should reduce the number of early deaths, improve the prognosis for surviving patients over the next few days, and reduce the number and severity of long-term sequelae.
Organophosphorus and carbamate pesticide poisioning
===================================================
This paper presents the protocol that we use to treat organophosphorus (OP)-poisoned or carbamate-poisoned patients on admission, based on our clinical experience and the best available evidence (see [Additional file 1](#S1){ref-type="supplementary-material"} ). It focuses on intentional ingestion of pesticides because such patients are more often severely poisoned than those with accidental or occupational exposure. We have not used any of the published severity poisoning scales because none have been independently validated. More importantly, pesticide-poisoned patients are unstable and a mildly poisoned patient can rapidly become very ill. An initial severity score suggesting a mild poisoning might allow doctors to relax with unfortunate results, as recognised by the IPCS/EC/EAPCCT poison severity score, which is designed only to be used retrospectively \[[@B10]\].
Poisoning with other pesticides
===============================
We concentrate here on OPs and carbamate pesticides because OPs in particular are responsible for most pesticide deaths across Asia \[[@B2],[@B11]-[@B13]\]. In addition, careful administration of oxygen, atropine and mechanical ventilation offers the opportunity to make a significant difference in outcome. However, the protocol can be adapted for the resuscitation of patients poisoned with other pesticides. Readers are referred to textbooks of clinical toxicology for details of subsequent treatment.
Initial assessment of the unconscious patient
=============================================
Initial assessment involves checking airway, breathing and circulation. As part of this process, provide high-flow oxygen if available and ensure a patent airway through the placement of a Guedel airway or access.
Place the patient in the left lateral position, ideally in a head-down position, to reduce the risk of aspiration. Extension of the neck in this position helps to keep the airway patent.
Watch out for convulsions and treat with intravascular (IV) diazepam immediately if they do occur. Record a baseline Glasgow Coma Score to help with subsequent monitoring of the patient\'s condition. If available, affix a pulse oximeter.
Does the patient require atropine? Recognition of OP/carbamate poisoning
========================================================================
Next, assess whether the patient requires atropine. Textbooks list many features of the cholinergic syndrome \[[@B14],[@B15]\]. However, we use five in routine assessment: miosis, excessive sweating, poor air entry into the lungs due to bronchorrhoea and bronchospasm, bradycardia, and hypotension.
Severely OP- or carbamate-poisoned patients are typically covered with sweat, and have small pinpoint pupils and laboured breathing (often with marked bronchorrhoea and wheeze). The presence of pinpoint pupils and excessive sweat suggests that the patient has taken an OP or carbamate and requires atropine. The heart rate may be slowed, but normal or even fast heart rates are common.
If none of these signs are present, then the patient does not yet have clinical cholinergic poisoning and does not require atropine. However, it is possible that these signs will occur later, for example as a pro-poison (thion) OP is converted to the active oxon form, as a fat-soluble OP such as fenthion leaches out of fat stores into the blood, or if the patient has presented soon after the ingestion. Careful observation is required to look for the development of cholinergic signs.
Loading with atropine and IV fluids
===================================
Dose of atropine
----------------
For an unconscious patient, give atropine 1.8--3 mg (three to five 0.6 mg vials) rapidly IV into a fast-flowing IV drip. Although it is preferable that oxygen is given early to all ill patients, do not delay giving atropine if oxygen is unavailable. Because atropine dries secretions and reduces bronchospasm, its administration will improve patient oxygenation. There is no good evidence that giving atropine to a cyanosed patient causes harm.
Atropine takes only a few minutes to work. During the 5 min after atropine administration, record three other signs of cholinergic poisoning against which atropine dosing will be titrated (Table [1](#T1){ref-type="table"}): (1) air entry into lungs; (2) blood pressure; (3) heart rate.
There is no need to do this before atropine is given, because pinpoint pupils and sweating in a region where these pesticides are common are sufficient to indicate OP/carbamate poisoning and trigger the decision to give atropine.
If the clinical presentation is not clear, administer atropine 0.6--1 mg. A marked increase in heart rate (more than 20--25 beats/min) and flushing of the skin suggest that the patient does not have significant cholinergic poisoning and further atropine is not required.
Giving fluids
-------------
While waiting for the atropine to have effect, ensure that the two IV drips have been set up (one for fluid and drugs, the other for atropine). Give 500--1000 ml (10--20 ml/kg) of normal saline over 10--20 min.
Assess whether enough atropine has been given -- is the patient atropinised?
----------------------------------------------------------------------------
Three to five minutes after giving atropine, check the five markers of cholinergic poisoning (Table [2](#T2){ref-type="table"}). Mark them on an OP/carbamate observation sheet (Table [1](#T1){ref-type="table"}). A uniform improvement in most of the five parameters is required, not improvements in just one. However, the most important parameters are air entry on chest auscultation, heart rate, and blood pressure.
Pupil dilatation is sometimes delayed. Because patients do not die from constricted pupils, and the other parameters may improve more rapidly, it is reasonable to wait for the pupils to dilate. Check frequently and carefully that the other parameters are improving.
When all the parameters are satisfactory, the patient has received enough atropine and is \'atropinised\'.
Continuation of bolus atropine loading to reach atropinisation
--------------------------------------------------------------
If after 3--5 min a consistent improvement across the five parameters has not occurred, then more atropine is required. Double the dose, and continue to double each time that there is no response \[[@B16],[@B17]\] (Table [1](#T1){ref-type="table"}). Do not simply repeat the initial dose of atropine. Some patients need tens or hundreds of mg of atropine, so repeating 3 mg doses will mean that it may take hours to give sufficient atropine \[[@B16]\]. Severely ill patients will be dead by this point -- atropinise the patient as quickly as possible.
Beware of pupils that do not dilate because pesticide has been splashed into them directly, and lung crepitations that are due to aspiration of the pesticide rather than the systemic effects of the pesticide. Generalised wheeze may be a better sign of under-atropinization in a patient who has aspirated pesticide.
Atropine treatment after atropinization
---------------------------------------
Once atropinised (with clear lungs, adequate heart rate \[more than 80 beats/min\] and blood pressure \[more than 80 mmHg systolic with good urine output\], dry skin, and pupils no longer pinpoint), set up an infusion using one of the two IV cannulae. This should keep the blood atropine concentration in the therapeutic range, reducing fluctuation compared with repeated bolus doses.
In the infusion, try giving 10--20% of the total amount of atropine that was required to load the patient every hour. If very large doses (more than 30 mg) were initially required, then less can be used. Larger doses may be required if oximes are not available. It is rare that an infusion rate greater than 3--5 mg/hour is necessary. Such high rates require frequent review and reduction as necessary.
Observation of the patient
==========================
Review the patient and assess the five parameters every 15 min or so to see whether the atropine infusion rate is adequate. As atropinisation is lost, with for example recurrence of bronchospasm or bradycardia, give further boluses of atropine until they disappear, and increase the infusion rate (Table [1](#T1){ref-type="table"}).
Once the parameters have settled, see the patient at least hourly for the first 6 hours to check that the atropine infusion rate is sufficient and that there are no signs of atropine toxicity. As the required dose of atropine falls, observation for recurrence of cholinergic features can be done less often (every 2--3 hours). However, regular observation is still required to spot patients at risk of, and going into, respiratory failure.
Atropine toxicity
=================
Excess atropine causes agitation, confusion, urinary retention, hyperthermia, bowel ileus and tachycardia \[[@B15]\]. During regular observation for signs of overtreatment, check for the features given in Table [3](#T3){ref-type="table"}.
The presence of all three suggests that too much atropine is being given. Stop the atropine infusion. Check again after 30 min to see whether the features of toxicity have settled. If not, continue to review every 30 min or so. When they do settle, restart at 70--80% of the previous rate. The patient should then be seen frequently to ensure that the new infusion rate has reduced the signs of atropine toxicity without permitting the reappearance of cholinergic signs.
Do not follow heart rate and pupil size because they can be fast or slow, and big or small, respectively, depending on the balance of nicotinic and muscarinic features. Tachycardia also occurs with rapid administration of oximes and with pneumonia, hypovolaemia, hypoxia, and alcohol withdrawal, and is not a contraindication to giving atropine.
Catheterise unconscious patients soon after resuscitation is completed. Look for urinary retention in an agitated confused patient; agitation may settle after insertion of the catheter.
Care of the airway
==================
If a pesticide-poisoned patient is unconscious, place an endotracheal (ET) tube at this point even if a Guedel airway is working well, to minimise the risk of aspiration and to facilitate respiratory care if there is deterioration.
Use diazepam to keep the patient sedated and tolerant of the ET tube. Because patients are often unstable during the first 6--12 hours, it may be better to sedate the patients to keep their ET tube in position if they start to waken with the atropine and the first dose of oxime.
Active cooling and sedation
===========================
Hyperthermia is a serious complication in hot and humid wards. A febrile patient should receive the minimum amount of atropine needed to control muscarinic signs, sedation if there is excessive agitation and muscle activity, and active cooling. Lay a towel soaked with water over the patient\'s chest and place in a fan\'s airflow. Cold water soaked towels can also be placed at points of maximum heat loss (for example axillae, groins).
Reduce agitation with diazepam 10 mg given by slow IV push, repeated as necessary in an adult, up to 30--40 mg per 24 hours. Tying a non-sedated agitated patient to the bed is associated with complications, including death. Such patients struggle against their bonds and generate excess body heat, which may result in hyperthermic cardiac arrest.
Diazepam is preferred over haloperidol because large doses of haloperidol may be required in patients receiving atropine. Haloperidol is also non-sedating, associated with disturbances of central thermoregulation and prolongation of the QT interval, and pro-convulsant. Diazepam may also have other advantages because animal studies suggest that it reduces damage to the central nervous system \[[@B18]\] and diminishes central respiratory failure \[[@B8]\].
Confirmation of exposure to cholinergic compounds
=================================================
Confirmation of poisoning by anti-cholinesterase pesticides can be sought by measuring butyrylcholinesterase and/or red-cell acetylcholinesterase activity. However, such assays cannot be performed in the ward. Furthermore, emergency therapy should be determined by the patient\'s clinical features, not by knowledge of the ingested poison.
Treatment of the resuscitated and stable patient -- should gastric decontamination be performed?
================================================================================================
Consider the need for gastric decontamination once the patient has been stabilised. Do not perform gastric decontamination until the patient is stable and, if necessary, intubated.
Ipecac is contraindicated in pesticide-poisoned patients. The effectiveness of both gastric lavage and activated charcoal is unknown.
Gastric lavage
--------------
Consider lavage only if a patient has taken a highly toxic pesticide and arrives at hospital within 1--2 hours. It can be given to calm patients who have given explicit consent to the procedure or to unconscious intubated patients. Its use in agitated non-compliant patients or un-intubated drowsy or unconscious patients risks major complications including death.
Pass a nasogastric tube to decompress the stomach and to suck out its contents. If patients have been previously given forced emesis, their stomach may well be already filled with fluid.
If a decision is made to give lavage, after aspirating the stomach contents give water or normal saline in lots of 300 ml through a nasogastric tube. Larger volumes of fluid may push the poison into the small bowel. There is no reason to use a large-bore oro-gastric lavage tube for liquid poisons unless food blocks the nasogastric tube. Take off 300 ml before giving a further two or three 300 ml aliquots, otherwise the stomach may become distended, allowing fluid to pass into the small bowel or causing the patient to vomit. Measure the amount of fluid taken off to ensure that fluid is not left in the stomach.
Activated charcoal
------------------
A dose of activated charcoal can be left in the stomach at the end of the lavage. There is currently no evidence that either single-dose or multiple-dose regimens of activated charcoal result in clinical benefit after pesticide poisoning.
Oximes and other therapies
==========================
The clinical benefit of oximes for OP pesticide poisoning is not clear, being limited by the type of OP, poison load, time to start of therapy, and dose of oxime \[[@B19],[@B20]\]. Current World Health Organisation guidelines recommend giving a 30 mg/kg loading dose of pralidoxime over 10--20 min, followed by a continuous infusion of 8--10 mg/kg per hour until clinical recovery (for example 12--24 hours after atropine is no longer required or the patient is extubated) or 7 days, whichever is later \[[@B20],[@B21]\]. Where obidoxime is available, a loading dose of 250 mg is followed by an infusion giving 750 mg every 24 hours \[[@B20]\]. Too rapid administration will result in vomiting, tachycardia and hypertension (especially diastolic hypertension).
Oximes are not recommended for carbamate poisoning.
The role of hydrocortisone and antibiotic treatment after aspiration is not known. Aspiration of pesticide and stomach contents initially causes a chemical pneumonitis and not pneumonia \[[@B22]\]. It is unknown whether pneumonitis benefits from steroids. Pneumonia is diagnosed if the fever persists for more than 48 hours or there is focal consolidation on X-ray. Earlier use of antibiotics risks antibiotic-associated diarrhoea.
Alcohol co-ingestion requires assessment of blood sugar levels and vitamin B supplementation.
Care after the first few hours
==============================
General observation
-------------------
OP/carbamate-poisoned patients are unstable and require regular observation to pick up changes in their general condition and their atropine requirements. Consider repeated doses of diazepam to keep the patient calm and settled.
If facilities permit, give patients a general anaesthetic, and intubate and mechanically ventilate them. This should reduce the number of deaths from respiratory complications.
Observation for impending respiratory failure and recurring cholinergic crises
------------------------------------------------------------------------------
Watch for early signs of intermediate syndrome in OP-poisoned patients. Weakness of neck flexion is common: the patient has difficulty lifting their head off the pillow; subsequent signs include the use of accessory muscles of respiration, nasal flaring, tachypnoea, sweating, cranial nerve palsies and proximal muscle weakness in the limbs with retained distal muscle strength.
Not all patients with neck weakness will develop the full intermediate syndrome requiring intubation and ventilation, but such patients are at risk and should be seen regularly. Measure tidal or minute volume and blood gases, if available. A locally agreed value should act as a trigger for prophylactic sedation and intubation, followed as necessary by ventilation.
Recurrence of toxicity, requiring atropine therapy, commonly occurs after poisoning with fat-soluble OPs, such as fenthion, that leak out of fat over days and even weeks. Recurring cholinergic crises may occur with little notice.
Conclusions
===========
Medical management of severe cholinergic pesticide poisoning is difficult, with high mortality. Some patients will die no matter how well managed. However, careful resuscitation with appropriate use of antidotes, followed by good supportive care and observation, should minimise the number of deaths in the period after admission to hospital.
Key messages
============
• Initial treatment of OP/carbamate pesticide poisoned patients involves the standard ABC of resuscitation.
• Since most deaths occur from respiratory failure, airway protection and ventilatory support is essential.
• Atropine can be given in an individualised dosing regimen to stabilise the patient.
• Careful observation probably saves many lives.
• Decontamination should only be done after the patient is fully stabilised, and not directly on admission.
Competing interests
===================
The authors declare that they have no competing interests.
Abbreviations
=============
ET = endotracheal; IV = intravascular; OP = organophosphorus.
Supplementary Material
======================
::: {.caption}
###### Additional file 1
Evidence for the protocol.
:::
::: {.caption}
######
Click here for file
:::
Acknowledgements
================
We thank the doctors of the Ox-Col Poisoning Study Team for their excellent work, patient care, and feedback about the protocol; the Directors, and medical and nursing staff of the study hospitals for their help with the study; and Surjit Singh and Alison Moffat for their critical review. ME is a Wellcome Trust Career Development Fellow; funded by grant GR063560MA from the Wellcome Trust\'s Tropical Interest Group. The South Asian Clinical Toxicology Research Collaboration is funded by the Wellcome Trust/National Health and Medical Research Council International Collaborative Research Grant GR071669MA.
Figures and Tables
==================
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
An observation chart recording the initial atropinisation of an organophosphorus-poisoned patient
:::
*Initials XX* *Study number Axxxx* *Date of arrival xx/xx/xx*
--------------- ---------------------- ---------------------------- ---------------- ---------- -------- ----- ---- ---- ---------- ----------
22.30 ***52*** ***Creps+*** ***Pinpoint*** ***No*** 90/60 I No No 2.4 mg
22.35 ***60*** ***Creps+*** ***Pinpoint*** ***No*** 90/60 I No No 4.8 mg
22.40 82 **+/-** ***Pinpoint*** Yes 110/60 N No No 4 mg
22.50 100 ***Wheeze*** 2 mm Yes \- D No No 2 mg
23.00 105 Clear 3 mm Yes \- D No No 2 mg/h Infusion
23.15 105 Clear 3--4 mm Yes \- D No No 2 mg/h Infusion
23.32 102 Clear 3--4 mm Yes \- D No No 2 mg/h Infusion
00.30 98 Clear 3--4 mm Yes 110/60 D No No 2 mg/h Infusion
01.30 85 Clear 3--4 mm Yes \- D No No 2 mg/h Infusion
02.30 ***72*** ***Wheeze*** 3--4 mm Yes \- N/D No No 2 mg
02.35 96 Clear 3--4 mm Yes \- D No No 2.4 mg/h Infusion
02.45 98 Clear 3--4 mm Yes \- D No No 2.4 mg/h Infusion
04.00 102 Clear 3--4 mm Yes \- D No No 2.4 mg/h Infusion
Atropinisation was reached at 23.00, 30 min after the first atropine dose was given; a total of 13.4 mg of atropine was required. After 10 min, doubling doses were no longer used because there was a clear response to therapy with the pulse climbing above 80 beats/min and the chest sounding better. After a further 1.5 hours, the pulse rate started to drop but it was not until it had dropped below 80 beats/min and wheeze had become audible in the chest that another 2 mg bolus was given to atropinise the patient again. The atropine infusion rate was also increased and the patient remained stable for the next few hours.
**A/D/N/I, absent/decreased/normal/increased**; creps, crepitations; syst. BP, systolic blood pressure. **Clinical features in bold type indicate that atropine is required. Dashes indicate that no BP reading was taken.**
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Target end-points for atropine therapy
:::
--------------------------------------------
Clear chest on auscultation with no wheeze
Heart rate \>80 beats/min
Pupils no longer pinpoint
Dry axillae
Systolic blood pressure \>80 mmHg
--------------------------------------------
Notes:
1\. The aim of atropine therapy is to clear the chest and reach the end-points for all five parameters.
2\. There is no need to aim for a heart rate of 120--140 beats/min. This suggests atropine toxicity rather than simple reversal of cholinergic poisoning. Such high heart rates will cause particularly severe complications in older patients with pre-existing cardiac disease -- myocardial infarctions may result. However, tachycardias are also caused by hypoxia, agitation, alcohol withdrawal, pneumonia, hypovolaemia, and fast oxime administration. Tachycardias are not a contraindication for atropine if other features suggest under-atropinisation.
3\. Aspiration will commonly result in focal crepitations. Attempt to distinguish such crepitations from the more general crepitations of bronchorrhoea.
4\. Splashes of organophosphorus into the eye will produce intense miosis that may not respond to atropine therapy. However, symmetrical miosis is likely to be due to systemic effects of the ingested pesticide.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Markers used to assess atropine toxicity
:::
-----------------------------------------
Confusion
Pyrexia
Absent bowel sounds (Urinary retention)
-----------------------------------------
Notes:
Many factors can cause confusion and pyrexia. However, confusion and/or pyrexia in the absence of bowel sounds suggests that they are due to atropine toxicity and will respond to a reduction in the rate of atropine administration.
Alcohol withdrawal, requiring benzodiazepine therapy, must be considered in poisoned patients who are confused.
Control pyrexia as soon as possible; conditions causing pyrexia include agitation from alcohol withdrawal or atropine toxicity, atropine-induced failure to sweat, and high ambient temperature. Active cooling of the patient with fan and water-soaked towels must be a priority because they are at risk of hyperthermia-induced cardiac arrest. Most ill patients will be catheterised after resuscitation to observe urinary output. Urinary retention can therefore not then be used as a marker of toxicity.
:::
|
PubMed Central
|
2024-06-05T03:55:54.841001
|
2004-9-22
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065055/",
"journal": "Crit Care. 2004 Sep 22; 8(6):R391-R397",
"authors": [
{
"first": "Michael",
"last": "Eddleston"
},
{
"first": "Andrew",
"last": "Dawson"
},
{
"first": "Lakshman",
"last": "Karalliedde"
},
{
"first": "Wasantha",
"last": "Dissanayake"
},
{
"first": "Ariyasena",
"last": "Hittarage"
},
{
"first": "Shifa",
"last": "Azher"
},
{
"first": "Nick A",
"last": "Buckley"
}
]
}
|
PMC1065056
|
Introduction
============
Three investigators have reported that tidal volume (*V*~T~) in children is inaccurate when measured at the ventilator, even when effective *V*~T~is used \[[@B1]-[@B3]\]. Cannon and colleagues \[[@B1]\] studied 98 infants and children and found a significant discrepancy between expiratory *V*~T~measured at the ventilator and that measured with a pneumotachometer.
Calculation of the effective *V*~T~did not alter this discrepancy. Castle and colleagues \[[@B2]\] studied 56 intubated children and found that exhaled *V*~T~displayed by the Servo 300 significantly overestimated *V*~T~measured at the airway by between 2% and 91%. After correcting for gas compression, effective *V*~T~overestimated true *V*~T~by as much as 29% in older children but underestimated the true *V*~T~by up to 64% in the smallest infants. Neve and colleagues \[[@B3]\] studied 27 infants and found that *V*~T~was overestimated by the ventilator in comparison with *V*~T~measured at the Y piece. None of these investigators controlled for endotracheal tube (ETT) leakage, which is more of a problem in children than in adults because of the use of uncuffed ETTs.
Accurate measurement of *V*~T~is increasingly important because the Acute Respiratory Distress Syndrome (ARDS) Network investigators have shown that the use of a low effective *V*~T~leads to decreased mortality in their patient population \[[@B4]\]. The effective *V*~T~goal in their ventilator protocol was 6 ml/kg but could be reduced to as low as 4 ml/kg if the plateau pressure was above 30 cmH~2~O. At such low *V*~T~values, accurate measurement is imperative to prevent atelectasis and subsequent ineffective minute ventilation.
Clinically, there are three methods to estimate delivered *V*~T~: first, direct measurement at the expiratory limb of the ventilator; second, direct measurement at the ETT with a pneumotachometer; and third, indirect calculation of effective *V*~T~by using set *V*~T~minus calculated compressible volume lost in the ventilator circuit \[[@B5]\]. The principle of Boyle\'s law (the volume of gas decreases as the absolute pressure exerted by the gas increases, and vice versa) is used to calculate the compressible volume in ventilator circuits.
How effective *V*~T~compares with *V*~T~measured at the airway has not been rigorously tested. Using *V*~T~measured at the ETT as the gold standard, we used three test lung models in a controlled laboratory setting to evaluate the accuracy of ventilator measured *V*~T~and effective *V*~T~under conditions of poor lung compliance, with and without ETT leakage, across a range of simulated patient sizes. We proposed that the discrepancy between effective *V*~T~and *V*~T~measured at the ETT in children was due mainly to ETT leakage around uncuffed ETTs, and that in situations with minimal ETT leakage there would be minimal difference between the effective *V*~T~and *V*~T~measured at the airway.
Materials and methods
=====================
Experimental conditions
-----------------------
A Servo 300 ventilator (Siemens-Elema, Solna, Sweden) in the SIMV volume control mode was used. A pressure differential pneumotachometer (CO~2~SMO Plus; Novametrix Medical Systems, Wallingford, CT) was used between the ventilator and ETT connection. The temperature of the humidifier was set at 37°C. A heated disposable respiratory circuit (Allegiance Healthcare Corporation, McGaw Park, IL) was used. We tested the compliance of the circuit to ensure that it was stable across a range of conditions. To do this, we first set the ventilator on the following: inspiratory time of 1.3 s, positive end-expiratory pressure (PEEP) of 0, respiratory set frequency of 6 breaths per minute, and a pause time of 15%. *V*~T~was increased by increments of 50 ml and the plateau pressure was recorded from the ventilator with the patient outlet occluded. No component other than the humidifier was added to the circuit \[[@B6]\]. A linear relationship was found, with no change of the circuit compliance at high airway pressure.
In the pediatric and infant models, a valve distal to the ETT was used to adjust volume leaks of 0%, 10%, 20%, and 30%. A shown in Fig. [1](#F1){ref-type="fig"}, a separate pneumotachometer (NVM-1; Thermo Respiratory Group, Palm Springs, CA) was used for independent measurement of the percentage of ETT leakage.
The Servo 300 was used for all test conditions. To control for differences between the ventilators, we tested each set of experimental conditions on three different ventilators. The CO~2~SMO Plus respiratory mechanics monitor was used to measure the *V*~T~at the ETT. This monitor measures flow with a fixed-orifice differential pressure pneumotachometer located at the ETT. Respired gas flowing through the flow sensor produces a small pressure decrease across the two tubes connected to the sensor. This pressure decrease is transmitted through the tubing sensor to a differential pressure transducer inside the monitor and is correlated with flow according to a factory-stored calibration. The pressure transducer is automatically \'zeroed\' to correct for changes in ambient temperature. Data are filtered and sampled at 100 Hz. The monitor continuously displays a range of ventilatory variables, including both *V*~T~and airway pressures. Three CO~2~SMO Plus sensors are available: neonatal, pediatric, and adult. The manufacturer recommends that the choice of sensor be based on various criteria: first, the diameter of the tracheal tube; second, the patient\'s age; third, the expected flow/volume range; and fourth, the acceptable levels of dead space and resistance. Table [1](#T1){ref-type="table"} lists the experimental conditions for all lung models. Before data collection, all ventilators, respiratory mechanics monitors, and tachometers used in this study were calibrated in accordance with the manufacturer\'s recommendation.
To ensure that different ventilators and monitors did not influence the results, all data were repeated three times, each time with a different Servo 300 ventilator and a different CO~2~SMO Plus monitor.
### Adult lung model
A TTL™ adult test lung (Vent Aid; Michigan Instruments Inc., Grand Rapids, MI) was used. This device has two separate lungs, each with a functional residual capacity (FRC) of 900 ml. The lung compliance can be adjusted by moving a spring up and down with a compliance ranging from 10 to 150 ml/cmH~2~O per lung. Each lung is tested before use to assess for leakage. Lung--thorax compliance levels were set at 10, 20, 40, 60, 100, and 150 ml/cmH~2~O.
### Pediatric lung model
A TTL™ adult test single lung was used with the FRC adjusted to give 30 ml/kg by displacing the extra volume with water-filled bags. Lung--thorax compliance levels were set at 5, 10, 20, 40 and 60, ml/cmH~2~O.
### Infant lung model
An infant lung simulator (D.B&M products, Redlands, CA) was used. The model has three different preset compliances of 1, 3, and 10 ml/cmH~2~O.
Data recording
--------------
Data were recorded from both the ventilator light-emitting diode display and the CO~2~SMO Plus monitor display by a single observer. Variables recorded were inspired *V*~T~, expired *V*~T~, peak inspiratory pressure (PIP), PEEP, and plateau pressure. Effective *V*~T~was calculated from the following equation \[[@B2]\]: set inspired *V*~T~- \[circuit compliance × (PIP - PEEP)\].
Analysis
--------
The major outcome variable was the calculated difference between the effective *V*~T~and the exhaled *V*~T~measured either at the ventilator or at the ETT in each experiment. For each set of test conditions (Table [1](#T1){ref-type="table"}) we used the mean of the three replicate measurements and also give the highest and lowest values. *V*~T~was adjusted for the simulated weights and expressed as ml/kg. We determined a priori that the difference between the *V*~T~values would be considered excessive if it exceeded 10% of the 6 ml/kg goal (0.6 ml/kg).
Results
=======
Test lung models
----------------
As shown in Fig. [2](#F2){ref-type="fig"}, for the adult, pediatric, and infant models with no ETT leak, the difference between *V*~T~measured at the ETT and at the ventilator increased with decreasing lung compliance. *V*~T~measured at the ventilator was always higher than that measured at the ETT. The ventilator measurement overestimated *V*~T~by more than 10% (0.6 ml/kg) as lung compliance dropped to moderately low values and the difference exceeded 20% (1.8 ml/kg) at the lowest lung compliances in each model. The standard deviation of the difference was 0--0.2 ml/kg for all sets of measurements.
In all models, in the absence of ETT leakage the difference between effective *V*~T~and *V*~T~measured at the ETT was less than 10% across the range of lung compliances with a standard deviation of 0--0.2 ml/kg for all sets of measurements. As shown in Fig. [3](#F3){ref-type="fig"}, however, the agreement between effective *V*~T~and *V*~T~measured at the ETT was poor when a 20% and 30% simulated ETT leak was added in the infant and pediatric test lung models. Under these conditions, the effective *V*~T~was at least 10% higher than that measured at the ETT for all simulated conditions, and the standard deviation was 0.1--0.4 ml/kg for all sets of measurements.
Discussion
==========
Using well-controlled experimental conditions, we showed that in the absence of ETT leakage, effective *V*~T~approximated the *V*~T~measured at the ETT in the test lung even when lung compliance was poor. As expected, exhaled *V*~T~measured at the ventilator became increasingly inaccurate with poor lung compliance. In the presence of ETT leakage, effective *V*~T~overestimated the *V*~T~measured at the ETT by at least 0.6 ml/kg. It is clear that in the presence of ETT leakage, effective *V*~T~is inaccurate and *V*~T~is most accurately estimated at the airway.
We used an *in vitro*model to manipulate experimental conditions while controlling for all other variables. Accurate measurement of *V*~T~is essential when a low-*V*~T~strategy is used to protect injured lungs as is recommended by the recent ARDS Network study \[[@B4]\]. In the adult lung model, we manipulated the compliance to simulate the lung compliance quartiles reported in the ARDSNet study \[[@B4]\]. Our findings have clinical implications. In agreement with other investigators \[[@B1]-[@B3]\], we found that unadjusted *V*~T~measured at the ventilator is highly inaccurate. We found this inaccuracy to increase markedly when lung compliance was abnormal. This means that dual-control automated ventilator modes (for example volume support or pressure-regulated volume control) that make adjustments based on *V*~T~measured at the ventilator might ineffectively ventilate patients with poor lung compliance. Automated ventilator modes should be used with care in critically ill children.
We support the current recommendations of previous investigators \[[@B1]-[@B3]\] that *V*~T~should be measured at the ETT in critically ill children receiving mechanical ventilator support. These investigators emphasized the need to measure *V*~T~at the ETT for all children; they did not control for the presence of uncuffed ETTs in their studies or evaluate the effect of leakage. Significant loss of *V*~T~occurs when both ETT leakage and poor lung compliance are present. Although the *V*~T~measured at the ETT may underestimate the actual *V*~T~being delivered in this situation, it is still the best estimation of the tidal volume delivered to the lung. Use of cuffed ETTs to minimize ETT leakage may lead to more accurate measurement of *V*~T~when lung compliance is poor \[[@B7]\]. When ETT leakage is 20% or greater, Main and colleagues \[[@B8]\] reported inconsistent tidal volume delivery and gross overestimation of respiratory compliance and resistance in children.
When ETT leakage is minimal, it seems from our simulated lung models that calculation of effective *V*~T~would give similar readings to *V*~T~measured at the airway, even in small patients. This could potentially negate the need for the addition of sensors at the airway and their associated increase in airway resistance for small ETTs \[[@B2]\]. Unfortunately, ETT leakage is dynamic and dependent on head position. Unless a simple, accurate and continuous means of measuring ETT leakage is available, it is safest to measure *V*~T~at the airway in all mechanically ventilated children. Future studies of *V*~T~measurement accuracy in mechanically ventilated children should control for the degree of ETT leakage.
Key messages
============
• Previous investigators have emphasized the need to measure tidal volume at the endotracheal tube for all mechanically ventilated children.
• When endotracheal leakage is minimal, it would appear from this study using simulated lung models that calculation of effective tidal volume would give similar readings to tidal volume measured at the airway, even in small patients.
• Future studies of tidal volume measurement accuracy in mechanically ventilated children should control for the degree of endotracheal tube leakage.
Competing interests
===================
None declared.
Abbreviations
=============
ARDS = acute respiratory distress syndrome; ETT = endotracheal tube; FRC = functional residual capacity; PEEP = positive end-expiratory pressure; PIP = peak inspiratory pressure; *V*~T~= tidal volume.
Acknowledegments
================
This study was funded by Novametrix Medical Systems and ARAMCO.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Schematic diagram demonstrating the placement of CO~2~SMO and NMV pnueumotachometers in infant and pediatric models.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Effect of decreasing lung compliance on the difference between effective tidal volume and tidal volume at the endotracheal tube (ETT) in the infant, pediatric, and adult test lungs with no leak around the ETT.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Effect of decreasing lung compliance on the difference between effective tidal volume and tidal volume at the endotracheal tube (ETT) in the infant and pediatric test lung models with 20% and 30% simulated ETT leakage.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Experimental conditions for test lung model
:::
Parameter Infant Pediatric Adult
--------------------------------- ---------- ----------- ------- ----- ----- ----- -----
Simulated weight (kg) 4 7 10 20 31 50 70
ETT internal diameter (mm) 3.0 3.5 4.0 5.0 6.5 7.0 7.5
Tidal volume at 6 ml/kg (ml) 24 42 60 120 186 300 420
PEEP (cmH~2~O) 5 5 5 5 5 5 5
Rate per minute 20 20 20 20 20 12 12
Inspiratory time (s) 1 1 1 1 1 1.2 1.2
FiO~2~(%) 21 21 21 21 21 21 21
Circuit compliance (ml/cmH~2~O) 1 1 1 1.5 1.5 2.9 2.9
Servo 300 set range Neonatal Pediatric Adult
CO~2~SMO Plus sensor Neonatal Pediatric Adult
ETT, endotracheal tube; FiO~2~, fraction of inspired oxygen; PEEP, positive end-expiratory pressure.
:::
|
PubMed Central
|
2024-06-05T03:55:54.843844
|
2004-10-6
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065056/",
"journal": "Crit Care. 2004 Oct 6; 8(6):R398-R402",
"authors": [
{
"first": "Sami I",
"last": "Al-Majed"
},
{
"first": "John E",
"last": "Thompson"
},
{
"first": "Kenneth F",
"last": "Watson"
},
{
"first": "Adrienne G",
"last": "Randolph"
}
]
}
|
PMC1065057
|
Introduction
============
Maxillofacial and major neck surgery has a considerable risk for postoperative laryngo-pharyngeal oedema and airway obstruction due to surgical manipulation or haematoma \[[@B1]\]. When patients undergoing these operations develop laryngeal oedema or airway obstruction and require reintubation after they have been extubated, reintubation may be very difficult or impossible through laryngoscopy because of the characteristics of these operations such as mandibular fixation with an archbar or as a result of anatomical changes. Extubation of a patient with risk factors for difficult tracheal reintubation is approached with concern, even in the experienced hands of the anaesthesiologist and critical care physician. Although all of the criteria used to predict successful extubation are generally satisfactory before extubation, none predict the adequacy of the airway once the endotracheal tube (ETT) has been removed \[[@B2]\].
Hence, acute respiratory distress can develop after extubation and mandate emergency tracheal reintubation. Mask ventilation and tracheal intubation may be difficult or impossible. Considerable time and an experienced physician are needed to secure a difficult airway with the use of alternative methods such as fibre-optic bronchoscope, retrograde\] intubation or cricothyroidotomy. Re-establishing the airway in these patients can be extremely challenging, and often results in considerable morbidity and mortality \[[@B3]\]. In the study by Loudermilk and colleagues \[[@B2]\], the advantages of the use of a pediatric airway exchange catheter (PAEC) inserted before tracheal extubation of adult patients with a known or expected difficult airway were well shown. However, the routine use of PAEC as a rescue for reintubation after maxillofacial surgery has not been reported.
The aim of this study was to determine the usefulness of routinely inserting the PAEC before tracheal extubation of adult patients undergoing major maxillofacial or neck surgery (Fig. [1](#F1){ref-type="fig"}).
Methods
=======
Patients
--------
Thirty-six patients admitted to our intensive care unit (ICU) after maxillofacial or major neck surgery between January 2001 and May 2002 were routinely extubated with the use of a no. 11 PAEC (Cook Critical Care, Bloomington, IN), with the approval of the Institutional Review Board. Patients included in the study consisted of 13 post-operative patients with maxillofacial trauma, 14 patients who had undergone neck surgery (5 with hugely enlarged thyroid gland or tumor and 9 with larynx or tongue cancer), and 9 patients who had undergone maxillofacial cancer surgery.
Written consent for publication of the photos of the patients was obtained.
Technique
---------
A no. 11 PAEC is 83 cm in length and has a 4 mm external diameter and a 2.3 mm internal diameter with a hollow lumen. It is semi-rigid and made of radio-opaque polyurethane; there are six sideports in the distal 3 cm of the catheter. The patients were extubated when they became conscious and had normal body temperature and normal blood gases with an inspired oxygen concentration (FiO~2~) of 0.4, a positive end expiratory pressure of less than 5 cmH~2~O and pressure support of less than 8 cmH~2~O. In addition, the haemodynamic status of the patients had to be stable before the decision to extubate was made. The PAEC was carefully inserted through the existing ETT before extubation, avoiding carinal irritation by placing it at the same depth as the ETT tip (20--22 cm orally or 27--30 cm nasally). The PAEC was not inserted against a resistance. After the ETT had been removed and the PAEC had been secured, humidified oxygen with a low flow of 1--2 l/min was insufflated via the lumen of the PAEC. Signs of respiratory failure and tolerance were also assessed. The PAEC was removed when it became clinically apparent that the need for tracheal reintubation was unlikely. We considered the ability of patients to manage secretions including cough and swallow functions in making the decision about extubation of the PAEC. A stable O~2~saturation and the extent of surgery were also important factors in this decision. The timing of removal of the PAEC was therefore different depending on various characteristics of patients and surgery. When patients failed to respond to tracheal extubation, the PAEC was used to facilitate the reintubation.
Results
=======
Twenty-eight patients (77.8%) were men, and 8 (22.2%) were women. Ages ranged from 19 to 76 years, with a mean age of 52.6 ± 10.8 (all results are means ± standard deviation) years. An oral ETT was in place in 18 patients (50%) and a nasal ETT in 18 (50%). All patients had a cuff leak test before tracheal extubation. The median duration of endotracheal intubation after the operations was 1.2 days (range 2 hours to 10 days). After tracheal extubation with the PAEC, 4 of 36 patients (11.1%) required reintubation (Table [1](#T1){ref-type="table"}). The reintubation of these four patients, who are discussed in detail as case reports below, was achieved over the PAEC and was easily accomplished on the first attempt without the need of an alternative method. We used the assistance of laryngoscope during the reintubation of two patients in whom the PAEC had been inserted orotracheally. In the other 32 patients who did not require reintubation, the PAEC was kept in the trachea for between 4 and 24 hours (mean 10.4 ± 4.2 (all results are means ± standard deviation) hours) and none of them required reintubation after the PAEC had been removed. Thirty-one patients had nasogastric tubes at the same time. The PAEC was well tolerated in 34 of 36 patients (94.4%). Two patients tried to remove the PAEC; they were therefore sedated for a few hours. We did not give any sedative drugs to the patients who could tolerate the PAEC. No adverse events were observed while the PAEC was being kept in the trachea.
Case 1
------
The reason for reintubation of this male patient, who had undergone radical neck surgery for cancer and had been intubated easily by direct laryngoscopy before the operation, was excessive surgical bleeding and haematoma, which developed 2 hours after extubation. The patient was immediately taken to the operating room. He could not be ventilated effectively by bag-valve-mask during the induction of anaesthesia (fentanyl 2 μg/kg, propofol 2 mg/kg, vecronium 0.1 mg/kg) and his oxygen saturation decreased to 85%. He was reintubated orally over the PAEC with the assistance of a laryngoscope within a few seconds by using an 8 mm ETT. During observation with a laryngoscope, reintubation of this patient by direct laryngoscopy was thought to be nearly impossible because the glottis could not be seen as a result of the anatomic abnormality caused by haematoma. He was extubated again using the PAEC 24 hours after his second operation; the PAEC was removed again 6 hours after insertion.
Case 2
------
The second patient (a male), who had also undergone neck surgery (unilateral dissection), was intubated with difficulty using a Fasttrach (intubating laryngeal mask airway) because of anatomical abnormalities, which developed as a result of previous operations and radiotherapy. He was extubated 4 hours after the operation in accordance with the criteria mentioned above. However, he required emergency reintubation 18 hours after extubation because he suffered acute respiratory distress following aspiration and bronchospasm. We found out from the history obtained from his relatives that the patient had already had a swallowing disorder before the operation and suffered from aspiration. Thus, we prolonged the presence of the PAEC. Reintubation of this hypoxic patient was urgently achieved over the PAEC with the assistance of a laryngoscope using a 7.5 mm ETT under sedation and neuromuscular relaxation. During laryngoscopic observation we could not see the glottis. In this patient, a surgical tracheotomy was performed later because of recurrent aspiration and the need for tracheal suction.
Case 3
------
This patient (a female) was admitted to the ICU after she operation for maxillofacial trauma. She had been intubated nasally by direct laryngoscopy using a Magill forceps; she could not open her mouth after the operation because of inter-maxillary fixation (Fig. [2](#F2){ref-type="fig"}). Six hours after extubation her arterial CO~2~pressure increased, and she became confused as a results of hypoxaemia. She was reintubated nasally with a 7 mm ETT over the PAEC, with intravenous midazolam 0.05 mg/kg and fentanyl 1 μg/kg, without cutting the archbar. She was extubated with the use of the PAEC 2 days after her reintubation, and the PAEC was left in place for 8 hours. She did not need intubation again after the PAEC had been removed.
Case 4
------
The fourth patient, a male, underwent maxillofacial reconstructive surgery for cancer after he had been intubated nasally over a flexible bronchoscope because of anatomical abnormalities in the oral route. He became hypoxic 4 hours after his extubation and required immediate reintubation. A serious pharyngeo-laryngeal oedema was thought to be the reason for hypoxia. His reintubation was easily achieved over the PAEC, with intravenous midazolam 0.05 mg/kg. He was extubated with the use of the PAEC 3 days after reintubation, and the PAEC was left in place for 6 hours. He did not require intubation again after removal of the PAEC.
Discussion
==========
During the perioperative period, serious respiratory events due to inadequate airway management can develop, which can cause severe brain damage or death. Rosenstock and colleagues \[[@B4]\] reported that 60 of 284 complaints filed at the National Board of Patients\' Complaints in Denmark over a period of 4 years were associated with perioperative respiratory complications, 50% of which resulted in death. Adverse outcomes associated with respiratory events constituted the single largest class of injury in the American Society of Anesthesiology Closed Claims Study (522 of 1541 cases; 34%). Death or brain damage occurred in 85% of these cases. Three mechanisms of injury accounted for three-quarters of the adverse respiratory events: inadequate ventilation (38%), oesophageal intubation (18%) and difficult tracheal intubation (17%) \[[@B5]\].
In previous studies, reintubation rates of 5--19% have been reported in surgical ICU patients \[[@B6]-[@B8]\]. In our study, 11% of the patients required reintubation because of surgical bleeding, pharyngo-laryngeal oedema, aspiration, and inability to manage secretions. The reintubation risk of our study patients was higher than general ICU patients because they had high risks in terms of airway obstruction due to surgical manipulation. Patients who are expected to have a difficult airway may remain intubated longer than necessary, simply for fear of the inability to reintubate. Before the use of the PAEC in our clinic, we usually restricted extubation of patients who had undergone maxillofacial surgery and were at risk of difficult reintubation to the daytime, when experienced physicians were available, rather than during the night, to provide safer conditions. Prolonged tracheal intubation not only increases the risk of complications but is also expensive because it requires respiratory therapy and more extensive monitoring \[[@B9]\].
The PAEC is a long, flexible and hollow tube designed to facilitate the exchange of an *in situ*ETT. The primary use of the PAEC (adult size, 16--18 F) has been as a tube exchanger in the critical care setting. It has been also used before the extubation of patients with a known difficult airway \[[@B10]\]. In the study of Loudermilk and colleagues \[[@B2]\], the use of the PAEC in 40 patients with risk factors for difficult reintubation, including a history of previous difficult intubation, airway edema secondary to surgical manipulation or volume resuscitation, morbid obesity, and an immobilized or unstable cervical spine, was well described. They reported that 3 of 40 patients (8%) had been easily reintubated with the use of PAEC, which is a reintubation rate similar to our results. Although our findings are similar to those in the study of Loudermilk, our study population consisted of a specific surgery group and we used PAEC as a routine procedure in this group without considering whether the patients had previously been intubated with difficulty.
Various methods have been used to facilitate the reintubation of these patients such as a fibre-optic bronchoscope \[[@B11]\], rigid ETT guides \[[@B12]\] and retrograde intubation. When all of these methods fail, an urgent cricothyroidotomy or tracheotomy may be the only solution. The PAEC offers several advantages over these alternatives: first, it provides a method for the continuous administration of oxygen; second, it can be used as a stylet for tracheal reintubation; and third, it provides a method of ventilating the patient (jet ventilation) \[[@B13]\].
In patients whose reintubation was considered a risk and who were known to present difficult tracheal reintubation, elective tracheotomy has even been performed in many institutions \[[@B2]\]. Besides, there have been many cases reported who have undergone tracheotomy because of airway obstruction or other respiratory pathologies after neck surgery \[[@B14],[@B15]\]. Intraoperative tracheotomy is a safe route to secure the airway in the postoperative period in patients undergoing maxillofacial or major neck surgery. However, tracheotomy is a considerably invasive method and can lead to serious complications including bleeding, pneumothorax, infection and tracheal stenosis. Furthermore, only about 10% of the patients undergoing maxillofacial or neck surgery require reintubation after their operations, and most of these patients can be extubated later. This means that performing the tracheotomy routinely is not necessary in most of these cases. However, sometimes tracheotomy can be unavoidable in a selected group, especially when the patients are expected to need prolonged mechanical ventilation or are at great risk of reintubation because of severe airway obstruction. Thus, both methods can be considered depending on patient characteristics. In addition to the operative factors, the patients should be meticulously evaluated before the operation in terms of respiratory capacity, neurological status and co-morbid factors. However, there are no strict criteria for a decision on tracheotomy or a trial extubation. For example, our case 2 would have benefited from an intraoperatively performed tracheotomy. Fortunately, we were able to reintubate this patient easily over the PAEC, and then decide to perform the tracheotomy.
Although the PAEC is rigid enough to facilitate tracheal reintubation, not all patients\' tracheas may be easily reintubated. Forceful insertion of the ETT should be avoided to minimize trauma to vital airway structures and to avoid kinking the PAEC. Direct laryngoscopy may also relieve the obstruction and identify its cause. We also used the assistance of the laryngoscope during the reintubation of two patients over the PAEC both to facilitate the intubation and to evaluate the anatomical structure of the upper airway with regard to the possibility of laryngoscopy. Gentle rotation of the ETT while trying to insert it may release the tip \[[@B16]\]. The PAEC should never be inserted against a resistance. Although the tip of the PAEC is rounded and blunt, perforations of the tracheo-bronchial tree during the insertion of these catheters have been reported \[[@B17],[@B18]\].
In a study of patients requiring tracheal reintubation, 87% (34 of 39) required reintubation within the first 4 hours after extubation \[[@B19]\]. In our series, one patient required reintubation 2 h after extubation, two reintubations occurred within 6 h and the other 18 hours after extubation. This finding shows that the need for reintubation later than 4 hours after extubation is not rare. As it is impossible to know at what time patients may develop respiratory distress, the timing of removal of the PAEC can be decided only on an individual basis. We have no data to determine the optimal period for which the PAEC should be left in place. Potential complications of the prolonged use of PAEC are airway trauma and aspiration caused by incomplete glottal closure. One of our patients who underwent neck surgery and required reintubation after 18 hours of extubation aspirated before the reintubation, but this patient had already had swallowing dysfunction due to radiotherapy before the operation. We therefore considered that the aspiration was not associated with PAEC only, although it could have contributed to the development of aspiration. Besides, the presence of the PAEC in the trachea can cause the retention of tracheal secretion by inhibiting effective coughing, especially in patients with chronic pulmonary disease, smokers, or patients who stayed immobile for a long time before surgery. In these conditions, the PAEC should be left in place for as short a duration as possible.
Conclusion
==========
The routine use of a PAEC in patients who have undergone maxillofacial or major neck surgery facilitates reintubation when necessary, and can be a life-saving method. It allows a safer trial of tracheal extubation and therefore can shorten the duration of intubation. We suggest that after these surgical procedures a PAEC be used routinely before tracheal extubation because it is difficult to predict which patients will require reintubation.
Key messages
============
• The PAEC is a long, flexable and hollow tube designed to facilitate the exchange of an in-situ endotracheal tube.
• The routine use of the PAEC in patients who underwent maxillofacial or major neck surgery facilitates the reintubation when necessary, and can be a life-saving method.
Competing interests
===================
The authors declare that they have no competing interests.
Abbreviations
=============
ETT = endotracheal tube; ICU = intensive care unit; PAEC = pediatric airway exchange catheter
Acknowledgement
===============
This study was supported by the Akdeniz University Scientific Research Unit, Antalya, Turkey
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
A patient who had undergone maxillofacial reconstructive surgery was extubated with the use of a pediatric airway exchange catheter (PAEC) in the intensive care unit. The PAEC was left in place for 6 hours, and the patient did not require reintubation after the PAEC had been removed.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
A patient who underwent maxillofacial surgery due to trauma, presented as case 3 in the text. She was extubated with the use of the pediatric airway exchange catheter (PAEC), and required reintubation after 6 hours of extubation. This was easily achieved over the PAEC without cutting the archbar.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Demographic data, duration of use of endotracheal tube and pediatric airway exchange catheter, and reintubation ratio
:::
Parameter Value (*n*= 36)
-------------------------------------------- ----------------------------
Sex (F/M) 28/8 (77.8%/22.2%)
Age (years) 52.6 ± 10.8 (range 19--76)
Pathology
Maxillofacial trauma 13 (36.1%)
Neck surgery 14 (38.8%)
Maxillofacial cancer surgery 9 (25.0%)
Endotracheal tube, oral/nasal 18/18 (50%/50%)
Duration of endotracheal intubation (days) 2.8 ± 1.6 (range 0.1--10)
Reintubation ratio 4/36 (11.1%)
Duration of PAEC^a^(h) 10.4 ± 4.2 (range 4--24)
PAEC, pediatric airway exchange catheter.
^a^In 32 patients who did not require reintubation.
:::
|
PubMed Central
|
2024-06-05T03:55:54.845436
|
2004-9-22
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065057/",
"journal": "Crit Care. 2004 Sep 22; 8(6):R385-R390",
"authors": [
{
"first": "Levent",
"last": "Dosemeci"
},
{
"first": "Murat",
"last": "Yilmaz"
},
{
"first": "Arif",
"last": "Yegin"
},
{
"first": "Melike",
"last": "Cengiz"
},
{
"first": "Atilla",
"last": "Ramazanoglu"
}
]
}
|
PMC1065058
|
Introduction
============
Interprofessional tensions can threaten the delivery of quality health care in a hospital setting. Such tensions have been documented in several clinical domains including internal medicine \[[@B1]-[@B3]\], pediatric wards \[[@B4],[@B5]\], the operating room \[[@B6]-[@B8]\] and the intensive care unit (ICU) \[[@B9]\]. The ICU in particular is a nexus for interspecialty tensions because of its pivotal role in the care of the hospital\'s most critically ill patients and in the management of critical care resources. Within the hospital community, the ICU exists at the high-stakes intersection of emergency, surgery, internal medicine and palliative care, an intersection where the patient care resources are expensive, in scarce supply and a source of intense competition.
Repeated calls have been made for improved collaboration, communication, congruence and equity within health care teams as ways of improving quality of care and protecting patient safety. Current notions of team-building advocate increasing flexibility in team structure, abolishing hierarchies and cultivating shared decision making \[[@B10]-[@B16]\]. Although these are important concepts, they can reflect a naive sense of the team as a unified entity rather than as a collection of individuals with distinct professional identities based on different models of care, skills, economic circumstances and political agendas.
To foster optimal team function, we first need to understand better the forces governing the interactions between professions (for example, nurses and physicians) and between specialties (for example, the ICU team and external consultants) as they work together in an environment charged with professional, temporal and financial tensions. Previous work by our research group has described team dynamics in the ICU \[[@B9]\]. We found that the level of collaboration or conflict within the ICU team, and between the ICU and other specialties, fluctuated on the basis of six key catalysts: authority, education, patient needs, knowledge, resources and time. These findings provided insight into the divisive forces present even in high-functioning teams, and alerted us to the strategies that team members enact as they seek to balance individual needs with team goals.
We also found that \'team\', in the ICU, is not a unified body but rather is a complex and fluid entity composed of core and expanded groups. Membership in these groups is continually negotiated on the basis of relative professional roles, immediate needs and tacit \'rules of play\'. In essence, to become empowered actors in the ICU, team members must progress beyond learning procedural steps to understanding the rules of the game: who has power on the team, how is that power commodified, how is it accessed, and in what circumstances is it applied? Understanding these rules can be the difference between knowing how to make something happen in principle (for example, ordering an X-ray) and being able to make it happen in practice (for example, getting an X-ray *now*).
Understanding the rules of the game is also essential if team members are to move beyond thinking as individuals to begin thinking as part of a team.
The purpose of this study was to describe these tacit \'rules of the game\'. We sought to determine how power is commodified and exchanged by ICU team members in their daily interactions as they work to achieve clinical goals, delineate professional boundaries, and problem solve around complex system issues.
Methods
=======
In a follow-up to 4 months of ethnographic non-participant observations (phase 1, detailed methods and results previously reported \[[@B9]\]), seven 1-hour focus groups were conducted with ICU team members in two urban teaching hospitals in Toronto, Canada. Two hospitals were included because the participating intensivists and residents divide their time between the sites and because differences in the settings (for example, case types and case loads, nurse staffing patterns and hospital cultures) might affect team communication and collaboration.
A semi-structured question script was derived to pursue recurrent patterns identified in the observational data. Participants consisted of a sample of four nursing focus groups (*n* = 27), two resident groups (*n* = 6 or 10 available individuals) and one intensivist group (*n* = 4 of 8 available individuals). Residents and intensivists constituted a convenience sample of individuals who were able to accommodate the time for the focus group discussion. Within the nursing group, purposeful sampling was used to ensure some range in years of ICU experience and age in this population \[[@B17]\]. The sample was selected through consultation with the nurse managers of the units. The number of focus groups conducted was determined through theoretical sampling, in which data collection occurred alongside preliminary analysis, and collection ceased when no new themes were arising from the focus group discussions \[[@B18]\]. The study received institutional ethics approval, and informed consent was obtained from all participants.
Focus group interviews were audio-recorded, anonymized and transcribed with standard linguistic conventions to yield about 140 pages of transcription for analysis. In the grounded theory tradition \[[@B19]\], transcripts were read iteratively by four researchers and were analyzed for emergent themes as well as for the themes identified in the analysis of the observational data. Both open coding (identification of primary themes) and axial coding (analysis of relationships among themes) were conducted. The combined expertise of the four analysts was essential to the coding process: one researcher was an intensivist experienced in qualitative research, one was an expert in team discourse, and the remaining two had conducted the observations in the first phase of the study. Emergent themes were revised and refined through the constant comparison of instances from the data set both by individual researchers and in a series of weekly 2-hour meetings during which the analysts compared interpretive memos and discussed relationships between categories. Discrepancies were given particular attention to ensure the validity of the analysis: they were considered by consulting specific instances in the transcripts, discussing their relationship to established themes, and reaching consensus as a group \[[@B20]\].
Results
=======
The phase 1 observation data provided insight into three areas: the shifting notion of team, the fluctuating levels of collaboration and tension on the team, and the catalysts underlying such fluctuations (previously reported) \[[@B9]\]. Thematic analysis of the focus group data extended our understanding of these three areas, in particular revealing team members\' perceptions of the mechanisms by which collaboration is achieved or undermined. Two dominant mechanisms were recurrently described and were categorized in our analysis as \'the perception of ownership\' and \'the process of trade\'. The findings reported here describe these mechanisms as revealed by the focus group data and supported by the observational data; implications for team collaboration and conflict are emphasized.
Perception of ownership
-----------------------
This category included references to team members\' perceived ownership of valued constructs or commodities, including specialized knowledge, technical skills, equipment, clinical territory and even the patient himself or herself. These constructs and commodities formed the basis of negotiation or exchange during interprofessional interactions. The title of \'ownership\' rather than the more traditional concept of \'role\' was selected to reflect the participants\' emphasis on possession.
Ownership was perceived as both collective (for example, ownership by the ICU team) and individual (for example, ownership by a nurse or by nursing as a profession). Shared perception of collective ownership was portrayed by participants as the foundation of the group\'s identity. It promoted collaboration between members of the ICU team and was often established by contrast with those outside the core team such as surgeons, internists, or nurses from the wards. For example, nurses explained the team\'s collective ownership of the patient in contrast to interlopers from outside the unit:
\'*We don\'t negotiate in the ICU because we are ultimately responsible for the patient, so there is no negotiating when you are in charge of that patient*\' (Nurse FG1).
Individual ownership was also a dominant issue and included instances where team members recognized their own or others\' possession of valued commodities. For instance, respiratory therapists acted in a proprietary manner regarding the ventilator, and this ownership was recognized and respected by other team members. One resident acknowledged that:
\'*The RTs\' role is probably essential, because, uh, as a medicine resident, we don\'t know much about the ventilators \... we don\'t have the time to learn the specifics that they know, so they contribute in areas that we-- --we can\'t\...*\' (Resident FG1).
In cases like this, the recognition of others\' possession of knowledge and skills is part of the smooth collaborative functioning of the team. However, individual ownership can also create interdisciplinary tension when team members feel that their ownership of particular knowledge and skills is not recognized:
Nurse: \'*And we\'re the ones who do keep track because we\'re there 24 hours a day. It\'ll be like: \"Well order a blood culture\", well we did one just yesterday. Or \"Order a thyroid test.\" They just did them 2 days ago. You know?*\' (Nurse FG4).
In both observations and focus group data, the designation of ownership was a complex mechanism and frequently a site of tension. In some cases, the allocation of ownership was defended by a particular group and in others, chafed at:
Intensivist: \'*At the end of the day the staff \[intensivist\] is the bottom line. I mean for better or for worse. I am not necessarily saying that it\'s the right thing but \... the amount of control you relinquish is really wholly dependent on how strong you feel these other members of the team are*\' (Intensivist FG1).
Nurse (describing a situation at morning rounds): *\'The staff intensivist asked the nurse, are there any issues, any concerns for the patient going to the floor?\" The nurse started up, and she was talking about blood pressure issues. The staff intensivist interrupts to say, \"Oh well, that\'s a medical issue. No, I mean specifically a nursing issue. So shot her down immediately*\' (Nurse FG2).
The staff intensivist in the first example asserts his ultimate responsibility for patient care. In the latter example, however, the knowledge designated as nursing territory by the intensivist was perceived by the nurse as inappropriately constrained, signaling a conflict between the two professional domains.
Although the recognition of others\' ownership of commodities frequently facilitated smooth team function, it also served as a provocation for usurpation and theft. For instance, nurses reported situations in which residents sought nursing knowledge but later portrayed that knowledge as their own:
\'*They rely on our notes and our talking to them in the morning to give them the physical assessment of the patient but then they totally disregard you when it comes to rounds as part of the team as though they\'ve done this assessment themselves and nothing you say is worthwhile*\' (Nurse FG4).
Participants\' discussions of ownership illustrated key problems on an interprofessional team, problems that revolve around respecting the interface between individual and collective knowledge and the balance between individual and collective responsibility.
Process of trade
----------------
This second category captured instances in which team members traded valued commodities as they negotiated their collaborative work. Such trade commonly involved concrete, physical commodities, including equipment and resources, and abstract, social commodities, including respect, goodwill and knowledge.
The trade of scarce physical resources was a catalyst for tension on the team. In many cases, this tension was amplified by its recurrence and by the infuriating smallness of some of the issues under debate:
Nurse: \'*I\'ll give you an example: I need a pump because my patient\'s blood pressure is dropping and some nurse is hoarding all of them and saying she needs it too. And I say, \"I don\'t think you need it\", so I just yank it out and get it because I know this is just a regular drip*\' (Nurse FG1).
Trade in such mundane resources was a commonplace ritual as team members negotiated to locate the items required for everyday patient care. In other cases, tension was amplified by the critical importance of the resources. Trade in beds, for example, was fraught with tension, particularly for trainees:
Resident: \'*There is always a shortage of nurses and they\'re always closing beds and we \[trainees\] sort of have to bear the brunt \... and get caught in a bed war*\' (Resident FG1).
Nurse: \'*\[There was\] a new resident on call and the ER calls him, he accepts the patient. And then after he accepts the patient he comes to me to say, \"Well, we have a patient\", and I say, \"No, you don\'t do that. You ask me first, do we have any beds?\" Things like that. They\'re learning the rules*\' (Nurse FG2).
As the latter example illustrates, the trade in physical resources is governed by implicit, social rules, such as who can authorize a trade. Trainees frequently had difficulty in recognizing and negotiating these implicit rules.
Alongside the trade of concrete resources was trade in more abstract commodities. For the nursing group, the most dominant currency for trade was \'respect\', which they described themselves expecting in return for information, knowledge, resources and goodwill. The failure of other team members to present the currency of respect was often met with revenge strategies in the form of an embargo of trade. For instance, a nurse might refrain from offering her knowledge if appropriate respect was not proffered first:
Nurse: \'*\[Consultants to the ICU should\] introduce themselves, to say what service they\'re from, and to ask some questions about the patient as you\'re the primary caregiver. And \... then they would learn so much more and it would save a lot of time, instead of digging through all this information \... they\'re flipping, flipping, trying to find bloodwork, but they\'re not asking me, so I\'m not going to help, you know? You find it yourself*\' (Nurse FG2).
Such trade of knowledge for goodwill occurred not only among team members but also between the ICU team and consulting teams. This critical sort of trade was recognized and discussed by all team members in the study. Failure to engage in such trade could mean that \'a good team approach was lost\' (Nurse FG2). It could also be seriously detrimental to an individual team member\'s success. For instance, residents expressed that
\'*Your name can be ruined or made on one \... encounter, so \... you have to be very careful, because if you create one enemy you can end up having a tough time with a lot of people, and if they love you, then they love you mostly for whatever the time that you\'re here \... so it\'s a bit of a social game; you have to be careful*\' (Resident FG1).
The process of trade was a constant and at times difficult social game with potentially long-term consequences. The constancy of trade caused it to be a source of accumulated tension and perceived historical injustices, with a single trade event causing a ripple effect that might impact other patients, other team members, other hospital services, or other events later in time. For instance, based on experience, one nurse asserted that
\'*When you want to transfer a patient in a hurry there will be an obstruction there \... you know there will be excuses. You know sometimes we feel like they\'re \[ward nurses\] prolonging it \... so I say, \"Well, I\'ll call housekeeping for you.\" Of course they don\'t like that\...*\' (Nurse FG1).
The environmental tensions endemic to the ICU served to make the successful negotiation of trade more difficult but also more essential. As one staff intensivist put it:
\'\... *we deal with a lot of conflict and you have to learn how to control yourself and how to become adept at conflict resolution. And not through intimidation and humiliation of the colleagues you have but honestly listening to them and trying to understand where they are coming from and trying to be respectful of them although \... that is tough sometimes when you are not feeling particularly patient or magnanimous towards these folks that you are talking with and, you know, you are tired, you\'re sleep deprived \... and you may be getting hassled from all sorts of people because of resource issues*\' (Intensivist FG1).
Discussion
==========
Our data depict team collaboration in a decidedly non-romanticized manner. The notion of team collaboration as rooted in the ownership and trade of commodities presents a stark contrast -- and a strong challenge -- to the established literature on creating medical teams, which emphasizes mutual support and shared goals, and minimizes competition and contest. What our participants describe as underlying \'rules\' in the daily negotiation of individual and team activity, the literature has tended to portray as \'barriers\' to teamwork \[[@B16]\]. Recent ethnographic studies of health professional teams suggest that the traditional conception of a stable, unified team does not account for the daily workings of teams in complex environments \[[@B6]-[@B9]\]. Further, this current research should caution us that adherence to the traditional ideal of \'team\' may, in fact, constrain us from recognizing and promoting the functional mechanisms of group effort in the health care domain.
As our results demonstrate, the forces of ownership and trade have a central role in the daily negotiations that constitute teamwork in the ICU setting. When these forces are ignored -- that is, perceived ownership is not attended to, or one commodity is not offered in trade for another -- tensions accumulate and collaboration becomes sluggish. When these forces are accommodated -- for example, competition for ownership of resources is anticipated, or requests are accompanied by offers of trade -- the team members navigate their competing interests more smoothly to act effectively together.
From a sociological perspective, this is common sense. There are sound theoretical reasons for these rituals of ownership and trade, the most basic of which is that the \'team\' is not a unified entity but rather a compilation of individuals with distinct professional identities: intensivist, nurse, respiratory therapist, resident, and so on. These professional identities are based in distinct models of care, different skill sets, diverse economic circumstances and competitive political agendas.
A useful way of theorizing the construct of professional identity, particularly when diverse professions come in contact with one another, can be found in the theory of social structuration \[[@B21]\]. In this theory, professions or organizations are conceptualized as social systems, in which each professional\'s role is determined by its position in relation to others and by its access to certain commodities. These commodities include access to material resources (\'economic capital\'), access to levels of information (\'cultural capital\') and access to social connections and acknowledged forms of expertise (\'social capital\'). Structuration theory is especially useful because it recognizes that individuals both within a profession (such as nursing) and between professions (such as nursing and critical care medicine) are in the constant process of attempting to distinguish themselves and their profession and thus acquire more \'capital\' so as to promote their ability to act (\'agency\') \[[@B22]\]. This notion of a profession and of an interprofessional team as a contested space is important, as it moves beyond a simplified notion of \'community\' as a group with shared values \[[@B23],[@B24]\] and allows us to theorize about important tensions in the formation of professional identity and the interaction between multiple professions. Acknowledging these tensions enables us to understand the way in which teams sustain the delicate balance between achieving a shared goal and competing for agency and status in the interprofessional setting.
The forces of ownership and trade are products of the contested relations on an interprofessional team. The point is not to stamp out these forces or to overcome them, but rather to articulate their role in team collaboration, so that they can be more strategically harnessed by team members and, as a consequence, smooth team functioning can be promoted. Handled adeptly, these forces allow members of a team to get necessary clinical work done, even in the chaos of competing ambitions and interests that is the ICU team. As one nurse put it: \'It may be construed that you are demanding, but then if you don\'t demand sometimes you don\'t get it; it\'s just a matter of strategy\' (Nurse FG1).
Limitations
===========
This study is constrained by the design decisions underpinning it. Findings may reflect the attitudes of a subset of ICU team members, for instance those more interested in exploring these topics. Generalizability is not the goal of grounded theory research, which seeks instead to produce rich descriptions and theoretical explanations of situated processes. However, the explanatory utility of these findings may be explored and enhanced in future research in different centers or other interprofessional health care team contexts.
Conclusions
===========
It is time that our understanding of team collaboration moved beyond the rhetoric of cooperation, and towards a more authentic depiction of the skills required to function in the competitive setting of the interprofessional health care team. Our intention is not to suggest a new rhetoric (of economics), but rather to shift our attention from idealized or abstracted depictions of teamwork, towards a grounded understanding of how collaboration is accomplished in daily practice. Knowing about perceptions of ownership, valued commodities and the rules of trade allows team members to shape outcomes and persuade people, to anticipate reactions and deflect obstructions, and to achieve individual goals while maintaining team cohesion. Efforts to improve teamwork must reflect such authentic, everyday \'rules of the game\' if they are to affect how work gets done on health care teams in complex settings such as the ICU.
These findings suggest educational implications relevant both to trainees and practising intensivists. In most training programs, professionalism and collaboration are part of an implicit, ad hoc curriculum largely consisting of role modeling and trial and error. As medical schools respond to recent calls to ensure competence in domains such as communication and collaboration \[[@B25]\], an understanding of authentic collaborative practice is essential to inform evidence-based curricula. For practising intensivists who may experience tension and difficulty in some team situations, understanding the rules of the game may assist them to analyze and improve their collaborative practice and, it is hoped, to improve the quality of care they provide to critically ill patients.
Contributors
============
All authors contributed to the design, conduct, analysis and interpretation of the research reported. LL and LH were Co-Principal Investigators and led the conceptual design of the study. SE and CE assisted with data collection and analysis, and with manuscript preparation.
Key messages
============
• In the daily negotiations that constitute inter-professional ICU teamwork, the ownership and trade of valued commodities play a central role.
• When ownership and trade are appreciated and handled well, team members are able to anticipate reactions, deflect obstructions, and achieve individual goals while maintaining team cohesion.
• Articulation of such authentic \"rules of the game\" is essential to the development of evidence-based curricula in collaborative practice.
Competing interests
===================
None declared.
Abbreviations
=============
ICU = intensive care unit.
Acknowledgement
===============
This research was supported by the University of Toronto Faculty of Medicine Dean\'s Fund.
|
PubMed Central
|
2024-06-05T03:55:54.847289
|
2004-10-8
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065058/",
"journal": "Crit Care. 2004 Oct 8; 8(6):R403-R408",
"authors": [
{
"first": "Lorelei",
"last": "Lingard"
},
{
"first": "Sherry",
"last": "Espin"
},
{
"first": "Cathy",
"last": "Evans"
},
{
"first": "Laura",
"last": "Hawryluck"
}
]
}
|
PMC1065059
|
See related commentary <http://ccforum.com/content/8/6/435>
Introduction
============
Sepsis is a major cause of death worldwide, with a large impact on mortality in the intensive care unit (ICU). It has been estimated that every day about 1400 patients die in ICUs as a result of sepsis \[[@B1]\].
Recent progress in sepsis research has been able to improve the knowledge about the basic pathophysiological processes of sepsis. However, in daily ICU practice it remains difficult to identify and treat sepsis, and its related conditions, adequately. Concerns remain about the lack of consistent definitions and understanding about sepsis among the global medical community \[[@B2],[@B3]\]. The American College of Chest Physicians and the Society of Critical Care Medicine (ACCP/SCCM) proposed a definition of sepsis and related syndromes in 1991 \[[@B4]\]. Although these definitions were based on expert opinion, the recommendations have not found unequivocal acceptance. However, these definitions have since been used for research purposes investigating new therapeutic modalities, in essentially all intervention trials.
To be able to diagnose and treat sepsis better it is important not only to improve knowledge about definitions and pathophysiology, but also to gain more insight into specialists\' perception of, and attitude towards, the current diagnosis and treatment of sepsis. This knowledge is important for the development of strategies to improve consensus in defining sepsis criteria among the intensive care society. Moreover, the introduction of intensivists supporting critical care units has been shown to be associated with improved survival of septic patients \[[@B5],[@B6]\]. Agreement among intensivists, as separate clinical specialists, in terms of their diagnosis of sepsis therefore also needs to be clarified.
Our hypothesis was that although there is good awareness among physicians involved in treating septic patients, a fragmented view of the definitions of sepsis is present. To investigate these hypotheses an international survey was conducted among intensivists and other specialists involved in the diagnosis and treatment of sepsis.
Materials and methods
=====================
In an international survey 1058 physicians were interviewed for this study; they were interviewed after a random selection of 1100 physicians in Europe and the USA. Of these, 756 physicians were interviewed in France (*n*= 150), Germany (*n*= 155), Italy (*n*= 150), Spain (*n*= 151) and the UK (*n*= 150). A further 302 physicians were interviewed in the USA. In each country equal numbers of intensive care and other specialists were interviewed. The specialist physicians included anaesthesiologists, cardiologists, endocrinologists, internists, nephrologists, pulmonologists, surgeons and emergency room physicians. The intensivists had to spend 50% or more of their time treating adults in the ICU, had to treat on average five or more ICU patients per month, had to treat two or more adult sepsis patients per month on average, and had to have worked for 2 years or more in the ICU. Otherwise they were classified as other physicians. The other specialists were also involved in the treatment of patients with sepsis, although on a less regular basis (fewer patients). They had to spend 10% or less of their time treating adult patients in the ICU and had to have been in practice for at least 2 years. It was intended that physicians spending between 10% and 50% of their time in the ICU should be excluded, but no physicians fulfilled this exclusion criterion. The study was conducted from November to December 2000. A recent study has shown a reduced mortality in patients with septic shock \[[@B7]\]. However, it was performed before the results of the present study were available. The survey was performed by telephone interview using trained staff of Yankelovich Partners. We list the questions asked in additional file [1](#S1){ref-type="supplementary-material"}. All questions were grouped into three categories based on a model describing behaviour framework \[[@B8]\]. To implement sepsis definition guidelines effectively, first the physician\'s awareness of the problem should be raised, then agreement on the problem should be reached and finally the ability to implement the definition guidelines should be present.
Statistics
----------
The data for this study are presented as means ± SEM or as percentages. Data were analysed with Student\'s *t*-test or χ^2^testing. *P*\< 0.05 was considered statistically significant.
The margin of error for the total group of physicians in this study was 3.0%, on the basis of the combined error values of all questions combined.
Results
=======
Respondent profile
------------------
Most physicians (83%) were male with an average age of 44.2 ± 0.3 years. The majority (57%) of these physicians were working in a non-teaching hospital. There was no difference between the intensivists interviewed and the other physicians with respect to gender, age distribution, percentage working in teaching hospitals, and percentage of practice based in hospital (Table [1](#T1){ref-type="table"}). The intensivists worked on an average 77.2 ± 0.95% of their time in the ICU. The number of adult patients treated in the ICU per month by the intensivists was 60 ± 3; of these 16.5 ± 0.9 were septic patients. The intensivists had worked for 11.6 ± 0.3 years after residency on the ICU. Of the other physicians, interviewed 120 (23%) were anaesthesiologists, 26 (5%) cardiologists, 26 (5%) endocrinologists, 83 (16%) internists, 18 (3%) nephrologists, 48 (9%) pulmonologists, 32 (6%) surgeons, 119 (23%) emergency room physicians, and 57 (11%) oncologists. These physicians worked 4.0 ± 0.3% of the time on the ICU and had been 13.5 ± 0.4 years in practice since residency.
Awareness of the problem of sepsis
----------------------------------
Three-quarters (767 of 1058) of all interviewed physicians agreed (strongly or somewhat) that sepsis is a leading cause of mortality compared with other conditions in intensive care. Of the intensivists, 78% considered sepsis as the leading cause in comparison with 67% of other physicians (*P*\< 0.0001). Nine in ten (934 of 1058) physicians agreed (strongly or somewhat) that sepsis is a significant financial burden on the health care system in their country. Among all physicians, 88% (937 of 1058) considered sepsis among the most challenging conditions that a doctor can treat. Two in five physicians (420 of 1058) had the impression that the incidence rate of sepsis has increased \'steadily\' to \'dramatically\' over the past 5 years, whereas 48% said that it remains stable. Two-thirds (285 of 420) thought that this increase is either \'extremely serious\' or \'very serious\'. Of the physicians surveyed, 77% reported the following major factors involved in this increase: an increased resistance of bacteria to antibiotics, an increased number of immuno-compromised patients, and a higher survival chance of post-surgical patients and patients with serious pathology.
A majority (656 of 1058, 62%) of physicians believed that their definition of sepsis is commonly accepted within their speciality. More than four in five (905 of 1058, 86%) physicians agreed (strongly or somewhat) that the symptoms of sepsis can easily be misattributed to other conditions. There was concern (ranging from \'somewhat\' to \'extremely concerned\') about the lack of a common definition for sepsis in 67% (708 of 1058) of the physicians. Of the physicians who were concerned about the lack of a common definition, 83% (199 of 708) stated that it is at least somewhat likely that the diagnosis of sepsis is missed. This figure was 53% (29 of 350) for the physicians who were not concerned about the lack of a common definition for sepsis. Although physicians are divided over whether the lack of a common definition for sepsis hinders proper diagnosis, they are not divided over whether a common definition would be a significant step towards better treatment.
Agreement on definitions of sepsis
----------------------------------
In general, physicians\' definitions of sepsis were fragmented. When defining sepsis, only 22% (114 of 529) of the intensivists and 5% (26 of 529) of the other physicians gave the definition of the ACCP/SCCM consensus statement (*P*\< 0.0001). Fewer than one-fifth (17%) of the physicians agreed on any one definition for sepsis, and six different definitions were mentioned by at least 1 in 10 physicians. This was not different between intensivists and other physicians. Moreover, physicians were divided as to whether sepsis is a systemic response (46%, 490 of 1058) as opposed to a syndrome (36%, 380 of 1058). One in ten physicians (103 of 1058), of both the intensivists and the other physicians, said that sepsis is a disease.
Among physicians, 71% (751 of 1058) said that fever is a sign or symptom that must be present to diagnose sepsis rather than any other factor. Aside from fever, no one symptom was listed by a majority of physicians as a sign or symptom that must be present to diagnose sepsis. Tachycardia was only cited by 29%, leukocytosis or leukopenia by 20%, hypothermia by 14%, and tachypnoea by 9% of physicians.
Ability to diagnose sepsis and communicate about sepsis
-------------------------------------------------------
Four in five physicians (911 of 1058) agreed (strongly or somewhat) that patients need better monitoring to diagnose sepsis at the earliest possible stage. In addition, 84% (890 of 1058) agreed (strongly or somewhat) that patients are often treated too late to reverse the onset of sepsis. According to the physicians, 46% of sepsis deaths are recorded as death by other diseases rather than death by sepsis. Bacterial culture results ranked as the most effective method for diagnosing sepsis by physicians; 80% found bacterial cultures either \'extremely\' or \'very effective\'. The second most effective method for diagnosing sepsis was haemodynamic monitoring. A significantly greater percentage of intensivists (74%, 393 of 1058) than the other physicians (66%, 350 of 1058) ranked haemodynamic monitoring as either extremely or very effective (*P*= 0.002) for diagnosing sepsis. Two-thirds (65%, 684 of 1058) of physicians agreed that a physical examination of symptoms is an effective method.
When speaking to the patients\' relatives, 81% (858 of 1058) of physicians agreed that communicating a diagnosis of sepsis to the families of patients with sepsis is difficult. Therefore, more than four in five (85%, 899 of 1058) physicians said that they describe sepsis to patients\' relatives as a complication arising from an underlying condition, as opposed to 10% who said they describe the diagnosis as sepsis.
Discussion
==========
In the present age of intensive care, sepsis remains responsible for a considerable number of deaths in critically ill patients. This disease has a major impact on both health care and society resources. Despite an increased understanding of sepsis, so far no information has been presented about physicians\' perception and knowledge of sepsis. This international survey was therefore conducted among physicians involved in treating septic patients.
One of the main findings of this study is that there is a general awareness of the importance and impact of sepsis among the physicians interviewed. A vast majority of physicians consider sepsis a leading cause of mortality. Moreover, the physicians agree that sepsis is a commonly encountered condition with an increasing incidence. Two recent reviews summarised the published studies on the incidence and mortality rates reported for sepsis. In a review by Brun-Buisson \[[@B9]\], 25% of patients on the ICU develop sepsis, with incidence rates varying from 45 in 1000 hospital admissions to 494 in 1000 ICU admissions. In a review by Matot, sepsis occurred with a mean frequency of 22.4% \[[@B1]\]. In both reviews a clear division between definitions of sepsis and severe sepsis or septic shock was used. In the review by Brun-Buisson an additional 10--15% of patients developed septic shock \[[@B9]\]. In practice, however, a majority of physicians agree that it is at least somewhat likely that the diagnosis of sepsis is being missed frequently.
One of the remarkable findings of this study is the lack of agreement on the definition of sepsis. A new set of definitions was proposed by the consensus conference of the ACCP/SCCM in 1992 \[[@B4]\] to improve the bedside recognition of sepsis, to permit early intervention and to differentiate infectious from non-infectious conditions. However, only a small percentage of physicians report the ACCP/SCCM criteria for the definition of sepsis. Not more than one-fifth agree on any one definition. This is consistent with the fact that a majority of physicians were concerned that there is no common definition of sepsis and a large proportion of physicians (for non-intensive care physicians even 41%) believe that other physicians within their speciality define sepsis differently from themselves. This perceived lack of a common definition might also explain why a significant number of physicians believe that sepsis is missed as a diagnosis. Indeed, the recommendations from the International Sepsis Forum recognise that in the past different definitions of sepsis were used interchangeably, which led to confusion \[[@B10]\].
When looking at the precise criteria that must be present according to the physicians interviewed, a wide variety of signs and symptoms were given. The one factor most frequently quoted was fever; the second most frequent answer was hypotension. This is of interest, given the fact that intensivists, in this survey, considered themselves extremely knowledgeable about the definition of sepsis and in the distinction between sepsis, severe sepsis and septic shock. Both the use of only one criterion and the use of hypotension are not at all consistent with the consensus definitions established in 1992 \[[@B4]\]. This misunderstanding with regard to the consensus criteria is consistent with the perception, among most physicians surveyed, of a lack of clear definitions for sepsis.
The lack of agreement on the definitions of sepsis criteria has an influence on the ability of physicians to diagnose and communicate about sepsis. The physicians in this survey were not content about the diagnostic tools they have for the diagnosis of sepsis. Most physicians agreed that better monitoring tools are needed to diagnose sepsis at the earliest possible time. Although a large percentage of physicians surveyed considered bacterial cultures and haemodynamic monitoring very effective for diagnosing sepsis, they also reported a high degree of interest in the investigation of other, more sensitive tools.
Another aspect of this survey was the differences found between intensivists and other specialists with less involvement in ICU care, indicating a difference in patient numbers with sepsis. Recent studies investigated the effects of specialised ICU staffing on outcome \[\[[@B5],[@B6],[@B11]\], 12\]. The results of these studies suggested that the presence of intensive care physician staffing is associated with a decreased length of ICU stay and with decreased costs, complications and mortality. However, it remained relatively unclear whether the institution of specialised ICU staffing had its effects on agreement, awareness and ability to diagnose sepsis. This survey showed that in general the intensivist seems to be more aware of issues involved for critically ill patients with sepsis. More intensivists consider sepsis a leading cause of mortality and a significant financial burden on the health care system. Moreover, they more frequently have the impression that the incidence is increasing. However, although awareness seems to be higher in specialised ICU staff, agreement on the definitions of sepsis is just as scattered as with non-ICU specialists. As a consequence the ability of intensivists and other specialists to diagnose sepsis is more or less comparable. Moreover, the ability of physicians to communicate the diagnosis of sepsis to the patients\' relatives is equally problematic. Two conclusions can be drawn from this survey, despite the limitations of a telephone survey. First, many doctors cannot define sepsis in accordance with the previously published consensus criteria. Second, sepsis is perceived as a leading cause of death in ICUs. The incidence of sepsis is high, and in addition physicians believe that the diagnosis of sepsis is often missed. This survey lends support to the idea that definitions of sepsis should be reviewed and that education is required, for both physicians and the public, for a better standardisation of clinicians\' definition and diagnosis of sepsis.
Key messages
============
• The current awareness of physicians concerning the impact which sepsis has on resources is widespread.
• Physicians are concerned that lack of agreement on the definitions of sepsis may lead to underestimating of the incidence of sepsis.
• The lack of agreement on the definitions of sepsis criteria has its influence on the ability of the physicians to diagnose and communicate about sepsis.
Competing interests
===================
The author(s) declare that they have no competing interests.
Abbreviations
=============
ACCP = American College of Chest Physicians; ICU = intensive care unit; SCCM = Society of Critical Care Medicine.
Supplementary Material
======================
::: {.caption}
###### Additional File 1
A PDF file containing a list of questions from the international sepsis survey.
:::
::: {.caption}
######
Click here for file
:::
Figures and Tables
==================
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Respondent demographics
:::
Respondent profile Intensivists Other specialists *P*
------------------------------------------ -------------- ------------------- -----
Number 529 529
Gender (% female) 14 20 0.2
Age, years (mean ± SEM) 43.8 ± 0.4 44.6 ± 0.4 0.7
Working in teaching hospital (%) 43 42 0.5
Percentage of practice based in hospital
Less than 30% 1 6
30--50% 2 3
50--70% 8 5
More than 70% 88 85
Unknown 1 1 0.4
Comparison of respondent demographics was by χ^2^or Student\'s *t*-test.
:::
|
PubMed Central
|
2024-06-05T03:55:54.850041
|
2004-10-14
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065059/",
"journal": "Crit Care. 2004 Oct 14; 8(6):R409-R413",
"authors": [
{
"first": "Martijn",
"last": "Poeze"
},
{
"first": "Graham",
"last": "Ramsay"
},
{
"first": "Herwig",
"last": "Gerlach"
},
{
"first": "Francesca",
"last": "Rubulotta"
},
{
"first": "Mitchel",
"last": "Levy"
}
]
}
|
PMC1065063
|
See related commentary <http://ccforum.com/content/8/6/425>
Introduction
============
The incidence of nosocomial pneumonia in mechanically ventilated patients ranges from 9% to 68%, and mortality rates range from 33% to 71% \[[@B1],[@B2]\]. In the EPIC (European Prevalence of Infection in the Intensive Care) study \[[@B3]\], ventilator-associated pneumonia (VAP) was the most frequent infection acquired in the intensive care unit (ICU), accounting for 45% of all infections in European ICUs.
The diagnosis of VAP is a challenge for the clinician because the presentation is variable, and other causes of fever and chest infiltrates may occur in these patients. Clinical/radiological evaluations provide the only criteria that permit timely diagnosis. Early institution of adequate antibiotic therapy is associated with decreased mortality, at least in the more severely ill patients. Culture results are currently used to guide adjustment or withdrawal of antibiotic therapy rather than to decide whether to treat. The practice of changing therapy with culture results has resulted in reduced consumption of antibiotics. Conversely, studies have shown that over-treatment with antibiotics may select organisms such as *Pseudomonas aeruginosa*and *Acinetobacter calcoaceticus*\[[@B4],[@B5]\].
The value of endotracheal aspirates for diagnosing VAP is controversial, but there is a growing body of evidence showing an important role for these cultures. Recent studies have consistently shown that outcome in VAP may not be influenced by whether cultures are obtained by bronchoscopy or from tracheal aspirates collected at the bedside. Furthermore, a cost-effectiveness analysis \[[@B6]\] strongly supported the employment of tracheal aspirates in the management of VAP.
Although the use of tracheal aspirates in VAP management is increasing, there are few data regarding the usefulness of quantitative as opposed to qualitative cultures. Some studies \[[@B7],[@B8]\] suggested that quantitative cultures should be used in order to avoid false-positive results, but little is known about the sensitivity and specificity of quantitative culture findings in severely ill patients who have previously received broad-spectrum antibiotics.
We conducted a prospective follow up of severely ill patients in a general ICU with a high rate of antibiotic use in order to evaluate the value of quantification of bacterial colonies in tracheal aspirates for diagnosing VAP.
Methods
=======
Study protocol
--------------
This study was conducted between March 2000 and January 2001 in a 28 adult bed medical/surgical critical care unit at the Hospital Israelita Albert Einstein -- a major referral tertiary care centre. The ethics committee of our institution granted approval for this investigation.
During the study period, every Monday morning all patients under mechanical ventilation for at least 48 hours were examined to determine whether they had VAP by three well trained intensivists and a respiratory therapist. We chose to evaluate all ventilated patients irrespective of the presence of VAP because on Mondays we routinely perform surveillance cultures of tracheal aspirates (in a search for multidrug-resistant pathogens and to determine contact precautions for such situations). We also aimed to include both patients with and without VAP based on clinical and radiological criteria.
The diagnosis of VAP was confirmed if there was agreement between two of the three physicians using clinical/radiological criteria. On the same day, the respiratory therapist also provided a description of the appearance (purulence) of the tracheal secretions. Endotracheal secretions were collected using a standard procedure and endotracheal aspirates samples were sent for qualitative and quantitative culture. The research team was blind to culture results, but the physicians were aware of the patients\' antibiotic consumption when they were evaluated.
Clinical characteristics were recorded at every evaluation (not just at enrolment in the study).
Diagnosis of ventilator-associated pneumonia
--------------------------------------------
For the purposes of the present study, VAP was diagnosed when a patient on mechanical ventilation for at least 48 hours developed a new or progressive pulmonary infiltrate on the chest radiograph in association with at least two of the following findings: râles or dullness to percussion on chest examination; new onset of purulent sputum or change in sputum character; decrease of at least 10% in arterial oxygen tension/fractional inspired oxygen ratio; leucocytes in excess of 12,000/mm^3^or under 4000/mm^3^; positive blood cultures or pleural effusion cultures; and axilar temperature greater than 37.8°C or under 36.0°C in the absence of antipyretic treatment (excluding another site of infection).
Tracheobronchial aspirate samples and microbiological processing
----------------------------------------------------------------
Tracheobronchial secretions were collected by the respiratory therapist, following specimen collection guidelines, after tracheal instillation of 5 ml saline. The specimens were sent to the laboratory and cultivated within 1 hour of collection. A dilution of the tracheal aspirate was prepared and inoculated with a calibrated loop on chocolate agar and MacConkey agar. After overnight incubation in appropriate conditions, the plates were interpreted according to quantification of growth \[[@B9],[@B10]\]. Qualitative cultures were considered positive when the growth of any micro-organism occurred and quantitative cultures were considered positive when the growth of 10^5^colony-forming units (cfu)/ml or more was observed. Sensitivity, specificity, positive predictive value and negative predictive values for qualitative and quantitative (10^5^cfu/ml and 10^6^cfu/ml) cultures from tracheal aspirates were calculated according to standard formulae. All samples were collected on the day of clinical and radiological evaluation.
Results
=======
A total of 106 patients were prospectively evaluated during the study period. The mean age (± standard error) was 66.6 ± 18.3 years. A total of 88 patients (83.0%) were male and 18 (17.0%) were female. The mean Acute Physiology and Chronic Health Evaluation II score was 20.1 ± 6.5. Medical patients constituted the majority (60.38%) compared with surgical patients (39.62%; Table [1](#T1){ref-type="table"}). Among medical patients, 30 (28.2%) were neurological and 21 (19.8%) were cancer patients.
In these 106 patients, a total of 314 clinical evaluations were conducted and endothracheal aspirates collected, corresponding to 42.3 ± 36.5 days (mean ± standard error) of mechanical ventilation. In 95 of these evaluations the radiological or laboratory investigations for VAP were incomplete at the time of clinical evaluation, and so these evaluations were excluded. Therefore, a total of 219 evaluations in 106 patients were included in the analysis.
Thirty-eight (17.4%) evaluations were classified as \'with VAP\' in 33 patients and 181 (82.6%) were classified as \'without VAP\' in 73 patients (Table [2](#T2){ref-type="table"}). The overall concordance between the first two observers for a diagnosis of VAP in the total population was high (94%). Within the VAP group, the overall concordance between the first two observers was 86.9%.
Qualitative and quantitative analyses
-------------------------------------
For qualitative analysis, among all 219 evaluations, 168 (76.7%) yielded cultures that were positive for at least one agent. In the VAP group, 31 of the 38 evaluations yielded positive cultures (81.6%). Thus, the sensitivity of qualitative cultures of tracheal aspirates was 81% and the specificity was 23%. The likelihood ratio for a positive test was 1.05 and the likelihood ratio for a negative one was 0.83. The positive predictive value was 18% and the negative predictive value was 86%.
For quantitative analysis, among the 219 evaluations, 117 had = 10^5^cfu/ml in tracheal secretions (53.4%) and 49 had = 10^6^cfu/ml (22.4%). In the VAP group, 25 of the 38 evaluations had = 10^5^cfu/ml (65.8%) and 10 of them had = 10^6^cfu/ml (26.3%). Thus, for 10^5^cfu/ml the sensitivity was 65% and the specificity was 48%. The likelihood ratio of a positive test was 1.25 and the likelihood ratio of a negative test was 0.73. The positive predictive value was 21% and the negative predictive value was 87%. For 10^6^cfu/ml the sensitivity was 26% and the specificity was 78%. The likelihood ratio of a positive test was 1.18 and the likelihood ratio of a negative test was 0.95. The positive predictive value was 20% and the negative predictive value was 83% (Table [3](#T3){ref-type="table"}).
In the VAP group leucocytosis was present in 26 evaluations (68.4%) and fever in 24 (63.1%), and purulent endotracheal secretions were observed by the therapist in 22 (57.8%) evaluations. In four evaluations only (10.5%) was blood culture positive for the same agent as was isolated in endotracheal secretions (Table [4](#T4){ref-type="table"}).
Overall, in 96.8% of evaluations patients were receiving at least one antibiotic. Prescription of antibiotics for three or more days before data collection was high (86.7%). The most frequently administered antibiotics were glycopeptides (49.7%), antifungals (42.4%), third-generation cephalosporins (39.2%), or carbapenem (34.2%; Table [5](#T5){ref-type="table"}).
Considering all VAP episodes, the most frequently isolated agents were *Staphylococcus aureus*(15.7%), *P. aeruginosa*(15.7%) and *Acinetobacter baumanii*(7.3%). Fungi accounted for 13.3% of all agents isolated. In 18.4% of evaluations in the VAP group, no agent was recovered from the endotracheal aspirates (Table [6](#T6){ref-type="table"}).
Clinical observations
---------------------
Considering the population as a whole, in 59 evaluations (26.9%) patients had a tracheostomy. Stress ulcer prophylaxis was present at 210 of the 219 evaluations (96%), with H~2~-receptor blockers in 58.4%, proton pump inhibitors in 36.5% and sucralfate in 0.9%. Sepsis was diagnosed in 46 (21%) evaluations.
Among the 38 evaluations classified as positive for VAP, tracheostomy was present in ten (26.3%). Previous lung disease was observed in six (15.7%) events. Ulcer prophylaxis was present in 100% of evaluations, with H~2~-receptor blockers in 22 (57.8%) and proton pump inhibitors in 16 (42.2%). Sepsis was diagnosed in 14 (36.8%) evaluations.
Other clinical characteristics are listed in Table [2](#T2){ref-type="table"}. A total of 31 (29.2%) patients died during their hospitalization: 11 (33.3%) of the 33 patients in the VAP group and 20 (27.3%) of the 73 patients without VAP (not significant).
Discussion
==========
VAP is the most frequent type of infection in ICU patients in Europe and Latin America (almost half of all nosocomial infections) \[[@B3]\] and ranks second in US ICUs \[[@B11]\]. The attributable mortality is higher in medical than in surgical patients, and rates vary according to the case mix and aetiological agent \[[@B12]\].
Inadequate or delayed antimicrobial treatment in VAP is an established independent predictor of death \[[@B13]\]. According to published data, changing an initial empirical treatment based on subsequent culture results may have either a beneficial effect (in terms of mortality, less antibiotic use, less days on antibiotics) \[[@B14]\] or no effect in more severely ill patients \[[@B15]\]. For this reason, efforts must be directed at choosing adequate empirical treatment as early as possible, which may be accomplished with a high degree of suspicion and adequate guidelines based on local antibacterial susceptibilities. In addition, adhering to ideal pharmacological principles (choosing continuous as opposed to intermittent administration, adjustment for renal and hepatic failures), reducing dosages when appropriate, and shortening the duration of treatment are presently standard of care for VAP.
In order to avoid any delay in instituting antibiotic treatment, reliable diagnostic methods should be employed. Despite their variable sensitivity and specificity \[[@B16]\], clinical/radiological findings may currently be considered the best option, although rapid tests, such as the percentage of infected leucocytes on bronchial specimens, are promising in that they can provide rapid confirmation \[[@B17]\]. Culture results for bronchial or tracheal samples may be available late in the course of an episode of VAP and should not be used to decide whether to treat, especially in patients who are severely ill. On the other hand, culture results should be used to adjust (narrow or extend antibiotic spectrum) or withdraw empirical treatment -- a practice that has been shown to be beneficial, with no increase in mortality, and that directs medical staff to seek other unsuspected foci of infection \[[@B18]\].
Although bronchoscopic samples increase the degree of confidence that a diagnosis of VAP is correct \[[@B14]\], endotracheal aspirates, despite their lack of consistency as a diagnostic tool \[[@B19]\], are widely employed in the management of VAP. Recent small trials have consistently shown that there is no advantage of using bronchoscopic methods over relying on tracheal aspirate cultures when mortality is an end-point \[[@B6],[@B20],[@B21]\]. Reduced costs and similar outcomes were reported using either quantitative or qualitative tracheal aspirates for guiding or deciding to interrupt antibiotic treatment for VAP \[[@B6]\]. This may be due to the high correlation between tracheal aspirates (both quantitative and qualitative) and bronchoscopic cultures when presence of VAP is highly probable \[[@B21],[@B22]\]. However, the above-mentioned studies did not determine the value of quantification of micro-organisms in tracheal aspirate samples as compared with qualitative assessment.
Quantification of micro-organisms in biological samples for the purpose of diagnosing infectious conditions is widely used, particularly for nosocomial infections. Regarding respiratory infections, bronchoscopic samples have established cutoff values (10^4^cfu/ml for bronchoalveolar lavage \[BAL\] fluid and 10^3^cfu/ml for protected brush specimen \[PBS\]) for improving diagnostic performance. On the other hand, use of these cutoff values has yielded conflicting results, and previous antibiotic treatment has great impact on these values. Souweine and coworkers \[[@B23]\] showed that the standard cutoff values of BAL and PSB would have to be lowered to 10^3^cfu/ml and 10^2^cfu/ml to retain diagnostic accuracy where antibiotics were previously administered, mainly when they are given in the preceding 24 hours.
Only a small number of studies have evaluated the role of quantitative endotracheal cultures in the diagnosis of VAP. Albert and coworkers \[[@B24]\], studying 20 ventilated patients and using clinical/radiological parameters, found the threshold of 10^5^cfu/ml to have a sensitivity of 81%, specificity of 65%, positive predictive value of 55% and negative predictive value of 55%. In that study different cutoff values were not tested to evaluate the real usefulness of quantification. Jourdain and coworkers \[[@B25]\] studied a group of 57 patients with presumed VAP, 19 (33%) of whom were confirmed by PSB sample with more than 10^3^cfu/ml. Using quantification in this population, those investigators showed that the sensitivity of the test reduced considerably from 86% to 43% whereas specificity increased from 52% to 95% when a cutoff of 10^3^cfu/ml was compared with one of 10^7^cfu/ml. No data regarding previous use of antibiotics were available to explain the decreased sensitivity.
We conducted a prospective follow up of severely ill patients with a high rate of antimicrobial use prior to diagnosis of VAP. Not surprisingly, the most frequent agents recovered were multidrug-resistant agents, such as methicillin-resistant *S. aureus*, *P. aeruginosa*and *Acinetobacter*spp.
We found different levels of sensitivity (81%, 65%, 26%) and specificity (23%, 48%, 78%) for qualitative and quantitative (cutoffs 10^5^cfu/ml and 10^6^cfu/ml) findings, respectively, as was expected. However, the positive (18%, 21%, 20%) and negative (86%, 87%, 83%) predictive values obtained were very similar.
Our data reveal sensitivity values for tracheal aspirates similar to those observed in the above-mentioned studies, although specificity values were lower. According to our data, use of the cutoff value 10^5^cfu/ml reduced the sensitivity of the test to levels too low to be useful in clinical practice, bearing in mind the proposed role of tracheal aspirates to guide antibiotic withdrawal or modification. Moreover, quantification did not improve predictive values for the purposes of diagnosing VAP at the time when a suspected case was evaluated.
Patient characteristics may have an impact on the accuracy of diagnostic tests. Although there is broad correlation between the number of bacterial colonies in biological samples and the occurrence of infection as opposed to colonization, the exact bacterial count cannot be predicted in highly ill patients, for whom a lower inoculum may be sufficient for disease development. This has been observed for catheter-related infections in severely ill patients in a surgical ICU \[[@B26]\], in which true catheter-related bacteraemia was reported with fewer than 15 cfu on catheter tips. In our patient population there was a significant proportion of patients with renal failure, diabetes, cancer and sepsis -- conditions that are known to be associated with immunosuppression.
These decreased sensitivity values may also be explained by antimicrobial use. More than 95% of the patients studied were receiving antibiotics when the sample was collected for analysis, and the majority of them were broad-spectrum antibiotics (almost 50% had received glycopeptides and 35% carbapenems). About 80% had received them for longer than 72 hours. Decreased accuracy of quantification with samples obtained by bronchoscopy was reported by Soweine and coworkers \[[@B23]\]. BAL and PSB had significantly less sensitivity when the procedure was performed within 24 hours of antibiotic use than when antibiotics had not been given for longer than 72 hours. The impact of antibiotic use may be greater for tracheal aspirates, irrespective of the timing of administration; this may be due to the higher concentration of the antibiotic in upper tract secretions, although this point requires further investigation.
Our study has a number of limitations. While we attempted to achieve a high degree of certainty in clinical/radiological parameters, with the participation of three experienced ICU physicians (with a high degree of correlation between them), no \'gold standard\' technique was employed, such as bronchoscopic samples (although it remains controversial whether bronchoscopy samples can be regarded as the gold standard for VAP). Because of the low specificity of clinical judgement, we must consider the fact that we are studying a population in which VAP rate is over-estimated. This is supported by the rate of 18.4% of VAP diagnoses with a negative tracheal aspirate finding and a 13.3% rate of fungal isolates, which only rarely can be considered true causative agents. Thus, it is possible that we have false-positive rate of at least 31.7%, although technical problems with specimen collection cannot be ruled out. The virtual absence of a gold standard for VAP makes study designs that address the issue of diagnostic tests difficult. In accordance with our study design, we evaluated all patients with mechanical ventilation every week, irrespective of clinical suspicion of VAP. This strategy may have beneficial effects because we included in the same population patients who were likely and those who were unlikely to have definite VAP, but increasing the possibility of false-positive cases.
Other study designs use populations selected because clinical/radiological judgement suggest the presence of VAP. In these studies, the control cases (no VAP) are defined as having negative bronchoscopic cultures, based on predetermined cutoff values. In these situations, problems with the lesser sensitivity of bronchoscopic samples in patients on antibiotics, and even the intrinsically low sensitivity of this diagnostic strategy when compared with histological criteria \[[@B27]\], increase the likelihood of including false-negative control individuals. In other words, with our study design we might have overestimated VAP, as compared with underestimating it with conventional study designs. For this reason we think that there is no ideal design for such studies, and studies that rely solely upon clinical/radiological parameters should not systematically be discarded. Furthermore, the use of bronchoscopy in our hospital is unreliable, as it may be in a large number of general ICUs.
Tracheal aspirates have a definite role to play in the management of VAP, but only when correlated with clinical findings \[[@B28]\]. The use of quantitative results may be associated with under-diagnosis of VAP, leading to inappropriate changes to antibiotic regimens and, in some cases, antibiotic delay or withdrawal.
Conclusion
==========
The severely ill and those who have previously received courses of broad-spectrum antibiotics -- a population whose number is expected to increase in modern ICUs -- may be targeted for use of qualitative findings rather than quantitative cultures of tracheal secretions for VAP management. Quantitative results may add costs and workload (in our laboratory it is five times more time consuming) and may then be of limited value in this group of patients, although enhanced specificity may be beneficial in terms of avoiding unnecessary treatment. In selected groups of severely ill patients, quantitative cultures of tracheal aspirates should not replace qualitative cultures for confirmation of diagnosis or management of antibiotic therapy.
Key messages
============
• Quantitative cultures of tracheal aspirates have increased specificity compared with qualitative analysis for diagnosis of VAP.
• The sensitivity values for quantitative cultures of tracheal aspirates are significantly lower than those for qualitative cultures for VAP diagnosis in severely ill patients receiving prior antibiotics.
• Quantitative cultures of tracheal aspirates should not replace qualitative cultures for the purpose of confirming a clinical diagnosis of VAP or adjusting antimicrobial therapy.
Competing interests
===================
The authors declare that they have no competing interests.
Abbreviations
=============
BAL = bronchoalveolar lavage; cfu = colony-forming unit; ICU = intensive care unit; PSB = protected specimen brush; VAP = ventilator-associated pneumonia.
Figures and Tables
==================
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Demographic data of the patients investigated
:::
Parameter Value
--------------------------------- -------------
Number of patients 106
Age (years) 66.6 ± 18.3
Ratio of males to females (*n*) 88/18
APACHE II score 20.1 ± 6.5
Clinical category (*n*\[%\])
Medical 64 (60.3%)
Cancer 21 (19.8%)
Neurological 30 (28.2%)
Surgical 42 (39.7%)
APACHE, Acute Physiology and Chronic Health Evaluation.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Clinical characteristics of the patients in the events investigated.
:::
Parameter Total evaluations (%) Evaluations in VAP group (%)
-------------------------------------------------- ----------------------- ------------------------------
Number of evaluations 219 (100%) 38 (100%)
Tracheostomy (*n*\[%\]) 59 (26.9%) 10 (26.3%)
Atelectasis^a^(*n*\[%\]) 13 (6.0%) 6 (15.7%)
Lung edema^a^(*n*\[%\]) 35 (16.0%) 9 (23.6%)
Lung contusion^a^(*n*\[%\]) 6 (2.7%) 1 (2.6%)
Pleural effusion^a^(*n*\[%\]) 21 (9,5%) 8 (21.0%)
Previous lung disease (*n*\[%\]) 30 (13.6%) 6 (15.7%)
COPD 16 (7.3%) 3 (7.8%)
Cancer 11 (5.0%) 2 (5.2%)
Asthma 2 (0.9%) 1 (2.6%)
Pulmonary fibrosis 1 (0.4%) None
Broncoaspiration (*n*\[%\]) 12 (5.4%) 3 (7.8%)
Sepsis (*n*\[%\]) 46 (21.0%) 14 (36.8%)
ARDS (*n*\[%\]) 12 (5,4%) 3 (7.8%)
Renal failure (*n*\[%\]; creatinine \>2.0 mg/dl) 91 (41.5%) 20 (52.6%)
Diabetes (*n*\[%\]) 35 (16.0%) 5 (13.1%)
Chemotherapy (*n*\[%\]) 13 (6.0%) 2 (5.2%)
Radiotherapy (*n*\[%\]) 2 (0.9%) None
Immunossupressants drugs (*n*\[%\]) 5 (2.2%) 1 (2.6%)
AIDS (*n*\[%\]) 3 (1.3%) None
Renal transplantation (*n*\[%\]) 5 (2.2%) 1 (2.6%)
Abdominal surgery (*n*\[%\]) 32 (14.6%) 7 (18.4%)
Multiple trauma (*n*\[%\]) 21 (9.5%) 3 (7.9%)
Neuromuscular blocking agents (*n*\[%\]) 7 (3.1%) 7 (18.4%)
Central venous line (*n*\[%\]) 215 (98.0%) 38 (100%)
Intracranial pressure monitoring (*n*\[%\]) 20 (9.1%) 1 (2.6%)
^a^According to clinical judgement. ARDS, acute respiratory distress syndrome; COPD, chronic obstructive pulmonary disease; VAP, ventilator-associated pneumonia.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Qualitative and quantitative analysis
:::
Parameter Qualitative Quantitative
----------------------------------- ------------- -------------- ------
Sensitivity 81% 65% 26%
Specificity 23% 48% 78%
Positive predictive value 18% 21% 20%
Negative predictive value 86% 87% 83%
Likelihood ratio of positive test 1.05 1.25 1.18
Likelihood ratio of negative test 0.83 0.73 0.95
cfu, colony-forming units.
:::
::: {#T4 .table-wrap}
Table 4
::: {.caption}
######
Diagnostic criteria for ventilator-associated pneumonia in order of occurrence
:::
Diagnostic criteria *n*(%)
------------------------------------------------------ ------------
Leukocytosis 26 (68.4%)
Fever 24 (63.1%)
Purulent tracheal secretion 22 (57.8%)
Decrease of at least 10% in PaO~2~/FiO~2~ratio 16 (42.1%)
Rales or dullness to percussion on chest examination 9 (23.6%)
Leucopenia 4 (10.5%)
Blood positive cultures 4 (10.5%)
Hypothermia 2 (5.2%)
FiO~2~, fractional inspired oxygen; PaO~2~, arterial oxygen tension.
:::
::: {#T5 .table-wrap}
Table 5
::: {.caption}
######
Prescription of antimicrobials in all the events studied
:::
Class of antimicrobial Evaluations (*n*\[%\])
--------------------------------- ------------------------
Glycopeptide 109 (49.7%)
Antifungical 93 (42.4%)
Third generation cephalosporin 86 (39.2%)
Carbapenem 75 (34.2%)
Clindamicin 34 (15.5%)
Quinolone 31 (14.1%)
Metronidazol 26 (11.8%)
Fourth generation cephalosporin 23 (10.5%)
Macrolides 23 (10.5%)
Total 212 (96.8%)
:::
::: {#T6 .table-wrap}
Table 6
::: {.caption}
######
Infectious agents isolated in the evaluations of patients with ventilator-associated pneumonia
:::
Aetiological agents Evaluations (*n*\[%\])
--------------------------- ------------------------
Gram-negative bacteria 15 (39.4%)
Gram-positive bacteria 11 (28.9%)
Negative cultures 7 (18.4%)
One or more agents 6 (15.7%)
Fungus 5 (13.3%)
Isolated agents
*Staphylococcus aureus* 6 (15.7%)
*Pseudomonas aeruginosa* 6 (15.7%)
*Acinetobacter baumanii* 3 (7.8%)
:::
|
PubMed Central
|
2024-06-05T03:55:54.851997
|
2004-10-14
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065063/",
"journal": "Crit Care. 2004 Oct 14; 8(6):R422-R430",
"authors": [
{
"first": "Luis Fernando Aranha",
"last": "Camargo"
},
{
"first": "Fernando Vinícius",
"last": "De Marco"
},
{
"first": "Carmen Sílvia Valente",
"last": "Barbas"
},
{
"first": "Cristiane",
"last": "Hoelz"
},
{
"first": "Marco Aurélio Scarpinella",
"last": "Bueno"
},
{
"first": "Milton",
"last": "Rodrigues Jr"
},
{
"first": "Verônica Moreira",
"last": "Amado"
},
{
"first": "Raquel",
"last": "Caserta"
},
{
"first": "Marinês Dalla Valle",
"last": "Martino"
},
{
"first": "Jacyr",
"last": "Pasternak"
},
{
"first": "Elias",
"last": "Knobel"
}
]
}
|
PMC1065064
|
Introduction
============
The rate of expansion of medical knowledge is increasing rapidly, and it is frequently difficult for clinicians to keep abreast of important new literature. For example, several recently published randomized controlled trials in critical care have demonstrated mortality benefits \[[@B1]-[@B5]\], but uptake of new knowledge into clinical practice is often delayed \[[@B6]-[@B8]\]. Improving access to this knowledge base at the point of care may lead to better clinical decision making, which could improve patient outcome, reduce costs and optimize bed utilization \[[@B9]\]. In critical care, rapid access to medical reference information may be particularly important in facilitating timely management decisions and avoiding errors \[[@B10]\].
Computing technology can allow point-of-care access to up-to-date medical reference material \[[@B11]\]. A study evaluating a mobile computerized cart to make evidence available to clinicians in an internal medicine setting \[[@B12]\] demonstrated that evidence-based medicine was more likely to be incorporated into patient care when the computerized system was used. Because of their portability, handheld devices may be more practical tools for disseminating knowledge to the point of care. Despite the popularity of handheld devices in medicine, few studies have evaluated the usefulness of this technology \[[@B13]\]. Before widespread dissemination of this type of technology can be encouraged, its impact must be thoroughly evaluated \[[@B14]\]. In the present study we evaluated whether it would be feasible and effective to provide updateable reference information from a central academic centre to handheld computers used by critical care specialists in community hospitals.
Methods
=======
Study design, participants and setting
--------------------------------------
A total of 17 intensivists at four community hospital intensive care units (ICUs) in the Greater Toronto Area participated in the present prospective interventional study.
Intervention
------------
After training, each physician was equipped with a handheld computing device (Palm M505; Palm Inc., Milpitas, CA, USA) loaded with medical reference material pertinent to the critical care physician. This information included a customized critical care information handbook (\'Critical Care\'), which was previously developed for use by residents and physicians at our centre ([Additional file 1](#S1){ref-type="supplementary-material"}). Commercially available medical reference software was also incorporated, namely PEPID ED (PEPID LLC, Skokie, IL, USA) and MedCalc <http://medcalc.med-ia.net/>.
The handheld devices were able to receive literature updates on a regular basis, using customized software (IqSync; Infiniq Software, Mississauga, Ontario, Canada), which accessed an internet-based server using either a connection via desktop computer or infrared data transfer to a telephone modem (Fig. [1](#F1){ref-type="fig"}). New information was sent to the handheld devices and appeared in a file called \'What\'s New\'. These updates, provided every 2--3 weeks, comprised brief reviews of relevant new literature including a short summary, a commentary and the article abstract.
All handheld devices were equipped with backup software that allowed the content to be rapidly restored in the event of a hardware failure (BackupBuddy VFS; Blue Nomad Software, Redwood City, CA, USA). The devices were also equipped with software capable of generating a log of the applications used (AppUsage; Benc Software Production, Slavonski Brod, Croatia).
Between September and November 2002 the handheld devices were distributed to participating physicians, at which time they each received a 1-hour training session on the use of the handheld device and the internet link (Fig. [2](#F2){ref-type="fig"}). After training, the participants were able to utilize the devices in clinical practice for 12 months. We provided 24-hour support by telephone and e-mail, with a website for independent review.
Outcome measures
----------------
### Feasibility
Feasibility of the system was assessed by tracking physicians\' use of the handheld device and tracking their access of the individual handheld applications during the study period. Physicians who updated their handheld computers at least once a month for 6 months were identified as \'regular users\'. A qualitative assessment of the system was achieved through surveys and focus group methodology. Participants completed surveys at baseline to identify their prior familiarity with handheld devices, and at the end of the study period to evaluate subjectively the handheld reference system and the individual handheld applications. Survey data were scored on a 7-point scale, in which \'poor\' scored 1 and \'excellent\' scored 7. An independent company (The NRC+Picker Group, Markham, Canada) conducted the focus group evaluations at the end of the intervention period, to determine the perceived utility of the information system. Each hospital physician group participated in one focus group meeting.
### Information access
Information sources that physicians accessed to make clinical decisions were evaluated during simulated patient care scenarios, completed in the physicians\' own ICU utilizing a computerized patient simulator (SimMan; Laerdal Medical Corporation, Wappingers Falls, NY, USA). Each physician completed one scenario before the handheld device was introduced (baseline scenario) and one at the end of the intervention period (final scenario), when the handheld device could be used (Fig. [2](#F2){ref-type="fig"}). A small pool of five scenarios with equivalent complexity was developed, such that physicians would likely need to access information sources in order to make management decisions. The scenarios involved unusual but important conditions, namely thyroid storm, myasthenia gravis, methanol toxicity, malaria and methemoglobinaemia. They were allocated to study participants in such a way as to avoid participants from the same site receiving the same scenario at the same time point, and to avoid repetition of scenarios among individual participants. Each scenario concluded with the physician writing admission orders for the simulated patient.
During the scenarios we tracked all medical reference sources utilized by the physicians, who were encouraged to use a \'think aloud\' process \[[@B15]\]. An audiovisual recording was made of the scenarios for later analysis, and when the handheld was used real-time screen capture was incorporated into the recording ([Additional file 2](#S2){ref-type="supplementary-material"}). This allowed us to document which handheld applications were accessed, the time taken to access information and the time taken to complete the scenario. We developed an objective scoring system for the admission orders generated at each scenario. The admission orders were assigned a score (range 0--100) by a critical care physician (SM) and critical care pharmacist (LB), who were blinded as to whether the physician used the handheld device. The scenario-specific scoring system allocated points for all necessary diagnostic and therapeutic interventions, weighted according to relative importance. Negative points were given for potentially harmful orders.
Data analysis
-------------
Data are presented as median and interquartile range (IQR), and permutation tests were used for comparisons because numbers were small and not normally distributed. The differences between the final and baseline admission order scores and the time to completion of scenarios were calculated for each participant. A two-sample permutation test was used to compare these differences between the group of physicians who chose to use the handheld in the final scenario and those who did not use the device. Admission order scores obtained for each of the five scenarios were compared. Outcomes were considered statistically significant at α \< 0.05. The SAS System for Window version 8.2 (SAS Institute, Inc., Cary, NC, USA) was used for all analyses.
Focus groups were recorded, transcribed verbatim and subsequently analyzed. Themes were identified and unique perspectives on key issues noted \[[@B16]\].
Results
=======
Feasibility
-----------
The handheld information system functioned well during the study period. Tracking of the deployment of handhelds identified 10 regular users (59%), four physicians (23%) who used the system variably and three physicians (18%) who never used their handheld device. The regular users accessed the personal information management applications more commonly (median 32.8 times/month, IQR 28.3--126.8) than the medical software (median 9/month, IQR 3.7--13.7; *P*= 0.028), although significant variation was noted (Table [1](#T1){ref-type="table"}).
Baseline survey data identified that, of the 17 critical care physicians participating, 12 (71%) had previous experience with handheld devices (nine had used the Palm operating system, and three had used Windows CE) for a median duration of 1 year (range 1 month to 3.8 years). Seven participants (41%) reported using handhelds for accessing medical information before the study. Of the 16 final survey respondents, seven (44%) felt that the handheld system had had a positive impact on their clinical practice. The handheld medical applications (Critical Care, What\'s New, Medcalc and PEPID) received similar ratings, with overall evaluation scores ranging from 4.1 to 5.3 on the 7-point scale.
Four focus group meetings, involving a total of 13 participants (76%), identified the benefits and barriers to use of handhelds for information access, and made suggestions for improvement (Table [2](#T2){ref-type="table"}). The overall impression of participants was that there is a role for handhelds for mobile information access, but that in situations away from the bedside other electronic media such as desktop computers were preferable.
Information access
------------------
Not all study physicians were able to participate in the simulated clinical scenarios on the pre-assigned day. Fourteen physicians (82.3%) participated in the baseline scenarios and 13 (76.5%) in the final scenarios. Information sources utilized during the baseline scenarios included the internet (50% of participants; e.g. Medline searches and electronic textbooks), textbooks (43%), telephoning colleagues, the ICU pharmacist or Poison Control Centre (71%), and other sources such as pocket guides (21%). In the final scenarios, the handheld device was used as the primary source of information by eight participants (62%; Table [3](#T3){ref-type="table"}). Of 14 information searches on the handheld device, 11 searches (79%) were successful and the median time to access information was 19 s (IQR 15--40 s). The information sources of those participants not using the handheld device were similar to those in the baseline surveys (Table [3](#T3){ref-type="table"}). Analysis of the time to completion of the clinical scenarios demonstrated no significant difference between those physicians who used the handheld and those who did not (12.92 min, IQR 10.73--16.62 min versus 15.5 min, IQR 12.85--22.72 min, respectively).
Physicians who did not use their handheld device in the final clinical scenarios had similar scores to their baseline scenario scores (median 60.0, IQR 40.0--60.0 versus 58.0, IQR 44.5--70.5, respectively). In contrast, an improvement in the final scenario score as compared with the baseline score was noted for those participants who chose to use the handheld device (median 66.0, IQR 52.5--74.5 versus 44.8, IQR 30.5--54.5, respectively; *P*= 0.018; Fig. [3](#F3){ref-type="fig"}). When scores recorded for each of the five clinical scenarios were compared, no significant difference was noted, reducing the likelihood that scenario assignment influenced outcomes.
Discussion
==========
This study demonstrates the feasibility of using an electronic knowledge translation system to provide high quality, regularly updated medical reference information from a central academic centre to multiple peripheral users. User acceptance of this technology was not uniform, with just over half of the participants using their handheld devices to access information on a regular basis. Nevertheless, the availability of point-of-care access to information may have improved the quality of clinical decision-making.
Although mobile computing devices have potential beneficial roles to play in clinical medicine, few publications describe formal evaluation of this technology \[[@B13]\]. Because the present study was an early hypothesis-generating evaluation of this technology, multiple quantitative and qualitative outcomes were measured. We generated novel data on the use of handheld devices in a clinical situation, but the study has several limitations. The number of physicians involved was relatively small, with a significant proportion not utilizing the technology. The allocation of clinical scenarios was not randomized, because they were allocated predominantly to avoid using the same scenario at the same site and time point. However, the analysis performed compared participants who used the handheld with those who did not; because it was not known which participant would use the handheld at the time of allocation of scenarios, potential bias was minimized. Furthermore, the scenarios appeared to be equivalent in difficulty because no difference was noted when scores for the individual scenarios were compared. A confounding factor in the study was the outbreak of SARS (severe acute respiratory syndrome) from March to May 2003, which had a significant impact on the study ICUs \[[@B17]\]. Participants were advised to avoid using their handhelds during patient contact because of the potential to transmit infection, and this affected continuity of the study. Had we not encountered this event, utilization might have been higher.
The lack of universal acceptance of this technology is not surprising and may be due to a number of factors, including inadequate training and the lack of familiarity with the technology \[[@B18]\]. Training is essential when introducing handheld computing technology \[[@B19],[@B20]\] and, although all users underwent a training programme, the surveys and focus groups indicated a need for improvement. Familiarity with handhelds is increasing, with 33% of all Canadian physicians and 53% of under 35-year-olds using these devices in 2003, but these levels of utilization remain relatively low when compared with use of the internet, at 88% \[[@B21]\]. Increasing familiarity with the technology will probably increase acceptance of such a system. Other potential barriers to use of the handheld system may be addressed by the rapidly developing technology, including improved screen resolution, ease of data entry and wireless connectivity. Acceptance may be increased through the development of an all-in-one package on the handheld, allowing additional functionality such as decision support, billing, electronic prescribing and communication.
The study demonstrated the potential role of an updateable handheld information system for knowledge translation in critical care. Rapid access to current clinical guidelines may be a valuable component of a comprehensive solution to reducing error and improving efficiency. Information access may be most beneficial in areas without full-time critical care physicians, particularly given the current imbalance between demand and supply with critical care physicians, which is expected to worsen \[[@B9],[@B10]\]. Recent recommendations highlight the importance of leveraging information technology to standardize practice and promote efficiency in critical care \[[@B10]\]. Handheld information access alone is unlikely to change clinical practice, but it should be considered a component of an electronic knowledge translation system. In many situations other media, such as desktop or tablet computers, may be preferable for information access.
Although the study was carried out in a critical care environment, such a system is probably applicable to other specialties in which clinicians are mobile and may not have ready access to a desktop computer (for example, anaesthesia, emergency medicine, home care). This study provides insight into the potential impact of this technology in improving health care outcomes \[[@B14]\]. Nevertheless, further study that builds on our findings is essential to determine how these new technologies can best be incorporated into the patient care setting.
Conclusion
==========
A handheld computer system is feasible as a means of providing point-of-care access to medical reference material in the ICU. During this study acceptance of this system was variable, and improved training and more advanced technology may be required to overcome some of the barriers we identified. In clinical simulations, use of such a system appeared to improve clinical decision-making.
Key messages
============
• This study demonstrated that an updateable handheld computer information resource is a feasible means for providing point-of-care access to medical reference information in the ICU.
• Acceptance of this system was variable and may be improved by enhanced training and newer technological innovations.
• In clinical simulations, this system appeared to improve clinical decision making.
Competing interests
===================
The author(s) declare that they have no competing intrests.
Author contributions
====================
Stephen Lapinsky, Randy Wax and Thomas Stewart were responsible for study design. Stephen Lapinsky, Randy Wax, Randy Showalter and Carlos Martinez implemented the handheld system and collected study data. Sangeeta Mehta and Lisa Burry were responsible for data collection and interpretation. Stephen Lapinsky and David Hallet analyzed the data. The manuscript was written by Stephen Lapinsky, Randy Showalter and Thomas Stewart, with all authors participating in revisions and giving approval to the final draft for submission for publication.
Abbreviations
=============
ICU = intensive care unit; IQR = interquartile range.
Supplementary Material
======================
::: {.caption}
###### Additional File 1
Quicktime movie (video clip) providing a brief overview of the content of the handheld \'Critical Care\' handbook, which is used as one of the medical reference sources in the present study.
:::
::: {.caption}
######
Click here for file
:::
::: {.caption}
###### Additional File 2
Quicktime movie (video clip) demonstrating a clinical simulation scenario, using the patient simulator Sim-Man. The physician can be seen accessing the handheld device, and utilization of the various information resources can be tracked.
:::
::: {.caption}
######
Click here for file
:::
Acknowledgements
================
We acknowledge the contributions of the intensive care physicians from William Osler Health Centre (Brampton Memorial Campus), Scarborough General Hospital, North York General Hospital and Trillium Health Centre (Mississauga). This study would not have been possible without the financial support of The Change Foundation of the Ontario Hospital Association (grant 01011) and Bayer Canada Inc. We acknowledge the technical support provided by Infiniq Software (<http://www.infiniq.com>; Mississauga, Ontario, Canada) and would like to thank Dr Arthur Slutsky and Dr Allan Detsky for reviewing the manuscript and providing valuable comments.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
The internet-based data transfer system. Updated information is downloaded to the handheld device from a study server. Connection to the internet can take place via hardwire synchronization with a desktop computer or using infrared (IR) data transmission to a dial-up modem. ISP, internet service provider.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
The study time course.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Comparison of scores for admission orders generated during the baseline and final clinical scenarios. Solid lines connect baseline and final scenario scores of participants who used the handheld device in the final scenario, and dotted lines connect scores of participants who did not use the handheld device (solid circles = scenarios where handheld was not used; open circles = scenarios where the handheld device used). A significant improvement was noted in scores in the handheld group as compared with the nonhandheld group (*P*= 0.018).
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Prospective tracking of the utilization of handheld applications
:::
-------------------------------------------------------------------------
Handheld application\ Number of accesses/month
---------------------------------- -------------------------- -----------
Median IQR
Personal information management
Date book 11.7 1.6--47.7
Address book 8.9 1.5--48.7
To Do List 9.8 4.0--17.7
Note Pad 6.0 2.3--11.3
Memo Pad 0.4 0--4.0
Medical information
iSilo (Critical Care, Whats New) 3.0 1.5--5.6
Med Calc 0.9 0.4--1.3
PEPID 0.2 0--4.2
-------------------------------------------------------------------------
Data were collected from 10 participants who used their handheld devices on a regular basis (i.e. updated their handheld device at least monthly for 6 months) IQR, interquartile range.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Major themes identified during focus group discussions
:::
Theme Details
--------------------------------------------- ------------------------------------------------------------------------
Benefits of handheld system Small size and portability
Pharmaceutical information
Literature updates
Preferences for information content Require more specialty (critical care)-specific content
Require more practical treatment-based information
Prefer all content in a single application
Barriers to the use of handhelds Small text fonts for reading
Technical problems, predominantly battery discharge
Inability to access information rapidly:
Inadequate search engine
Unfamiliarity with layout of content
Errors during text entry using handwriting recognition
Prefer \'all-in-one\' solution (e.g. pager, e-mail, physician billing)
Comparison with other information resources Desktop computer often preferable
Preferred desktop information resources
PubMed (Medline literature search)
Google (internet search engine)
UpToDate (electronic textbook)
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Evaluation of information sources used during the final clinical scenarios
:::
Resources Handheld used^a^(*n*= 8) Handheld not used^a^(*n*= 5)
------------------------------ -------------------------- ------------------------------
Nonhandheld resources
UpToDate^b^ 1 3
Textbook 0 1
Pharmacy/Poison Control 3 2
Telephone consult 3 3
Mean resources per scenario 0.88 1.8
Handheld resources
PEPID 11 0
Critical care 2 0
Other 1 0
Mean resources per scenario 1.75 0
^a^The decision to use the handheld device was at the discretion of the individual physician.
^b^UpToDate electronic textbook <http://www.uptodate.com>.
:::
|
PubMed Central
|
2024-06-05T03:55:54.855089
|
2004-10-14
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065064/",
"journal": "Crit Care. 2004 Oct 14; 8(6):R414-R421",
"authors": [
{
"first": "Stephen E",
"last": "Lapinsky"
},
{
"first": "Randy",
"last": "Wax"
},
{
"first": "Randy",
"last": "Showalter"
},
{
"first": "J Carlos",
"last": "Martinez-Motta"
},
{
"first": "David",
"last": "Hallett"
},
{
"first": "Sangeeta",
"last": "Mehta"
},
{
"first": "Lisa",
"last": "Burry"
},
{
"first": "Thomas E",
"last": "Stewart"
}
]
}
|
PMC1065065
|
Introduction
============
Sepsis is defined as the systemic response to infection \[[@B1],[@B2]\]. The deleterious effects of bacterial invasion of body tissues results from the combined actions of enzymes and toxins produced by the micro-organisms themselves, and the actions of endogenous cells in response to the infectious process. Despite advances in supportive care, mortality rates in patients with severe sepsis continue to exceed 30%. During sepsis vasoactive arachidonic acid metabolites of the cyclo-oxygenase (COX) pathway are released. In particular, thromboxane A~2~and prostacyclin have been found to be elevated in sepsis \[[@B3],[@B4]\]. Thromboxane A~2~has been associated with bronchoconstriction, vasocontriction and platelet aggregation \[[@B3]\]. Prostacyclin, the predominant eicosanoid generated by activated endothelial cells, is a powerful vasodilator and antagonist of thrombosis \[[@B3]\]. Prostaglandin (PG)E~2~is among the most potent and inducible of the prostanoids that are produced in states of inflammation. Specifically, there is evidence to support roles for PGE~2~as a mediator of sepsis-induced immunosuppression, an inhibitor of proinflammatory cytokine expression from monocytes, and an inducer of IL-10 production \[[@B5]-[@B7]\]. Conversely, PGE~2~has been shown to mediate detrimental effects in sepsis, including vasodilation and increased vascular permeability \[[@B8]\]. In addition, its role as a mediator in fever induction and augmentation of pain is well established \[[@B9]\]. Several studies \[[@B10]-[@B12]\] conducted in endotoxin-challenged animals have found beneficial effects of nonselective COX inhibitors. These beneficial effects were felt to be mediated, in part, by mitigation of pathophysiological events in sepsis induced by PGs.
COX exists as two isoforms -- COX-1 and COX-2. The former is constitutively expressed, whereas COX-2 is expressed at low levels in most normal resting cells. Marked upregulation of COX-2 occurs in synoviocytes, macrophages and endothelial cells during stress and in inflammatory conditions such as sepsis. COX-2 expression is induced by a number of cytokines, including tumour necrosis factor (TNF) and IL-1, mitogens and growth factors, lipopolysaccharide (LPS), and other inflammatory stimuli \[[@B13]\]. Recent studies \[[@B14],[@B15]\] provided evidence suggesting that selective COX-2 inhibitors have significant advantages over their nonselective counterparts. The specific benefits of COX-2 inhibitors include decreased gastrointestinal toxicity and bleeding \[[@B14],[@B16]\].
As with other nonsteroidal anti-inflammatory drugs (NSAIDs), lornoxicam inhibits PG synthesis via inhibition of COX, but it does not inhibit 5-lipoxygenase. The ratio of inhibitory potency of human COX-1 to COX-2 for lornoxicam is 0.6 \[[@B17]\]. Lornoxicam was reported to be 100-fold more potent than tenoxicam in inhibiting PGD~2~formation in rat polymorphonuclear leucocytes *in vitro*, and it was more active than indomethacin and piroxicam in preventing arachidonic acid induced lethality in mice *in vivo*\[[@B17]\]. Lornoxicam also inhibited the formation of nitric oxide in RAW264.7 mouse macrophages stimulated with endotoxin, indicating an effect on inducible nitric oxide synthase \[[@B18]\]. It also exhibited marked inhibitory properties on endotoxin-induced IL-6 formation in THP1 monocytes, with less activity on TNF and IL-1β. It appears that lornoxicam, in addition to markedly inhibiting COX and inducible nitric oxide synthase, has a moderate effect on the formation of proinflammatory cytokines \[[@B19]\].
The purpose of the present study was to evaluate the effects of intravenous lornoxicam on serum cytokine levels, haemodynamic and biochemical parameters, and outcomes in humans with severe sepsis.
Methods
=======
Patient population and study design
-----------------------------------
The regional committee on medical research ethics approved the study. Written informed consent was obtained, directly from the patients wherever possible or from the next of kin. Critically ill patients with bacteriologically documented infections were included in the study as soon as they met at least two of the following criteria for sepsis, as defined by the American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference Committee \[[@B2]\]: temperature \>38°C or \<36°C; heart rate \>90 beats/min; respiratory rate \>20 breaths/min or arterial carbon dioxide tension \<32 mmHg; and leucocyte count \>12 × 10^9^cells/l or \<4 × 10^9^cells/l. In addition, at least one of following conditions was required: hypoxaemia (arterial oxygen tension/fractional inspired oxygen ratio \<250); oliguria (urine output \<0.5 ml/kg body weight for 2 hours); lactic acidosis (lactate concentration \>2 mmol/l); thrombocytopaenia (platelet count \<100 × 10^9^/l); and a recent change in mental status without sedation. Patients who were younger than 18 years, had known or suspected hypersensitivity to COX inhibitors, or had received a COX inhibitor within 12 hours (or aspirin within 24 hours) were enrolled in another experimental protocol (not part of the present study), or were excluded if consent could not be obtained. Also excluded were patients with known or suspected brain death; those with advanced acute or chronic renal or hepatic failure; those who had received potent immunosuppressive drugs; those with gastrointestinal bleeding; those who were pregnant; and those with a known irreversible underlying disease, such as end-stage neoplasm.
The Acute Physiology and Chronic Health Evaluation (APACHE) II score \[[@B20]\] and Sepsis-related (or Sequential) Organ Failure Assessment (SOFA) score \[[@B21]\] (Table [1](#T1){ref-type="table"}) were employed to determine the initial severity of illness.
If required, patients underwent surgical procedures before the start of the study. No invasive surgery was performed during the 72-hour study period. All patients were ventilated in volume-controlled mode (Puritan Bennett 7200; Carlsbad, CA) and received continuous analgesic sedation with midazolam and fentanyl. Ventilator settings, level of positive end-expiratory pressure and fractional inspired oxygen were kept constant during intravenous administration of lornoxicam or placebo. Antibiotic treatment was adjusted according to the results of bacteriological culture, such as blood culture or culture of samples taken from different body sites. In all participants fluid replacement was administered to maintain central venous pressure between 4 and 8 mmHg. No inotropic agent was administered during the study. Those patients who met the criteria for severe sepsis presented above were enrolled in the study within 8 hours of intensive care unit (ICU) admission.
Protocol
--------
Randomization was done using a computer-steered permuted block design. The study was planned prospective, randomized, double blind, and placebo controlled. In order to perform the study in a double-blind manner, drug solution was administered to all patients by a nurse who had no knowledge of the study protocol, and follow up was done by an anaesthetist who also had no knowledge of the study protocol. Twenty patients received lornoxicam 8 mg (Xefo; Abdi Ýbrahim, Istanbul, Turkey), administered intravenously every 12 hours for a total of six doses. In the placebo group, also including 20 patients, saline was administered using the same volume and dosing regimen.
Measurements
------------
All patients had arterial catheters placed (Abbott Transpac^®^IV; Abbott, Sligo, Ireland) and central venous catheters placed via subclavian (Certofix trio V 720 7F×8\"; Braun, Melsungen, Germany). Arterial blood samples were simultaneously withdrawn for measurements of pH, partial oxygen tension, partial carbon dioxide tension and arterial oxygen saturation (Medica Easy BloodGas; Massachusetts, USA). Central venous pressure, mean arterial pressure, heart rate and naso-opharyngeal temperature were continuously monitored (Space Labs Inc., Redmond, WA, USA). All measurements were obtained at baseline (before the start of the study) and again at 24, 48 and 72 hour after the start of infusion. Lactate, platelets, leucocytes, bilirubin, alanine aminotransferase and creatinine were determined at the same times (Vitalab Flexor, Dieren, The Netherlands).
TNF-α, IL-1β, IL-2 receptor, IL-6 and IL-8 levels were measured at the same times. Venous blood was collected into a 10 ml sterile plain tube (without anticoagulant) before administration of any medications and stored at -20°C. Before assay, all samples were thawed to room temperature and mixed by gentle swirling or inversion. All sera were assayed on the same day to avoid interassay variation. TNF-α, IL-1, IL-2 receptor, IL-6 and IL-8 levels were measured using a solid-phase, two-site chemiluminescent enzyme immunometric assay method (Immulite TNF-α, Immulite IL-1β, Immulite IL-2 receptor, IL-6 Immulite and IL-8 Immulite; EURO/DPC, Llanberis, UK). The antibodies used in this procedure have no known cross-reactivities with other cytokines. The intra-assay and interassay coefficients of variation, respectively, for this procedure were as follows: for IL-1β, 2.8--4.9% and 4.8--9.1%; for IL-2 receptor, 2.9--3.7% and 6.1--8.1%; for IL-6, 3.6--6.2% and 5.4--9.6%; for IL-8, 3.6--3.8% and 5.2--7.4%; and for TNF-α, 2.6--3.6% and 4.0--6.5%. The lowest detectable limits of IL-1β, IL-2 receptor, IL-6, IL-8 and TNF-α were 1.5 pg/ml, 5 U/ml, 5 pg/ml, 2 pg/ml and 1.7 pg/ml, respectively.
The duration of mechanical ventilation was recorded. Survival was defined as being alive at hospital discharge.
Statistical analysis
--------------------
Repeated measures analysis of variance was used to evaluate the differences between and within groups from baseline. In the case of statistical significance, groups were tested by independent sample t-test to determine which difference was significant. Data are expressed as mean ± standard deviation. *P*\< 0.05 was considered statistically significant.
Results
=======
Patient characteristics
-----------------------
Clinical and demographic characteristics of the patients are listed in Table [2](#T2){ref-type="table"}. Of the 40 patients included, 20 received intravenous lornoxicam and 20 received placebo. Fifteen patients had septic shock on admission (seven \[35%\] in the lornoxicam group and eight \[40%\] in the placebo group) and died while in the ICU. Baseline APACHE II scores (17.10 ± 3.58 and 18 ± 3.72 in the lornoxicam and placebo groups, respectively) and SOFA scores (5.90 ± 1.72 and 6.20 ± 2.2) were similar in the two groups (*P*\> 0.05). SOFA scores at 24 hours (5.50 ± 1.52 and 6.1 ± 1.2 in the lornoxicam and placebo groups, respectively), 48 hours (5.60 ± 1.6 and 6.0 ± 1.3) and 72 hours (5.72 ± 1.4 and 6.1 ± 1.6) were also similar (*P*\> 0.05). Infection was documented in all patients.
Haemodynamic parameters and oxygen transport variables
------------------------------------------------------
There were no significant differences between groups with respect to pH, partial oxygen tension, partial carbon dioxide tension, arterial oxygen tension/inspired fractional oxygen ratio and arterial oxygen saturation (*P*\> 0.05). No significant changes in mean arterial pressure and heart rate were found in either group (Table [3](#T3){ref-type="table"}). There were no significant differences between groups in biochemical parameters (Table [4](#T4){ref-type="table"}; *P*\> 0.05).
Outcomes
--------
Outcomes are listed in Table [2](#T2){ref-type="table"}. In the ICU, the overall mortality rates were 35% (seven patients out of 20) in the lornoxicam group and 40% (eight patients out of 20) in the placebo group (*P*\> 0.05). All of those who died did so while they were being mechanically ventilated. In the lornoxicam and placebo groups the mean durations of ventilation were 6.1 ± 2.4 and 5.8 ± 3.1 days, respectively (*P*\> 0.05). The length of ICU stay in lornoxicam treated survivors was not significantly different from that of placebo treated survivors (10.2 ± 7.1 versus 9.2 ± 8.4 days; *P*\> 0.05).
Plasma cytokine levels
----------------------
TNF-α, IL-1β, IL-2 receptor, IL-6 and IL-8 levels remained unchanged during the study (Table [5](#T5){ref-type="table"}).
Side effects
------------
Intravenous lornoxicam was well tolerated by all patients, and no side effects were noted during or after administration of lornoxicam.
Discussion
==========
Systemic inflammatory response leading to postoperative organ dysfunction and sepsis remains a formidable clinical challenge and carries a significant risk for mortality. Sepsis and septic shock remain major causes of death in ICUs. A number of studies have examined the role of nonselective COX inhibitors both in animal models of sepsis and in patients with sepsis syndrome. Several studies \[[@B10]-[@B12]\] demonstrated beneficial effects of nonselective COX inhibition, predominantly in endotoxin-treated animals. However, subsequent studies \[[@B22],[@B23]\] examining the role played by NSAIDs, particularly ibuprofen, in human sepsis trials have been disappointing. The present study was therefore conducted to determine whether COX inhibition is upregulated early after the onset of severe sepsis, and if so whether COX inhibition prevents the occurrence of septic shock.
The arachidonic acid pathway is highly activated in macrophages, monocytes and other inflammatory cells, resulting in the formation of eicosonoids. PGs are involved in all phases of the inflammatory process, including fever and pain reactions, as well as in a large number of physiological functions, including intestinal motility, platelet aggregation, vascular tone, renal function and gastric secretion, among others. Two COX isoforms have been identified: COX-1 and COX-2. The former is a constitutive enzyme that is expressed in many cells as a house-keeping enzyme and stimulates homeostatic production of PGs. COX-2 is an inducible form of the enzyme that is expressed at the onset of inflammation by many cell types that are involved in the inflammatory response. NSAIDs act mainly through COX inhibitors, thus preventing the formation of proinflammatory prostanoids. Lornoxicam, a new member of the oxicam class of NSAIDs, inhibits PG synthesis via inhibition of COX, but it does not inhibit 5-lipoxygenase. Lornoxicam is at least 10 times more potent as an anti-inflammatory agent than piroxicam, and 12 times more potent as an analgesic than tenoxicam \[[@B17],[@B19]\].
The primary pharmacological action of NSAIDs is, of course, to decrease the formation of PGs and thromboxanes by inhibiting COX, a key enzyme in the biochemical pathway that leads to formation of these potent mediators \[[@B24]\]. Accordingly, products of the COX pathway, sometimes referred to as \'prostanoids\', have been implicated in the pathogenesis of the deleterious systemic consequences of serious infection and/or endotoxaemia. In addition, the toxic effects of TNF (thought to be one of the primary cytokines responsible for LPS-induced lethality) can be ameliorated by treating mice or rats with NSAIDs such as indomethacin or ibuprofen \[[@B25]\]. NSAIDs have been shown to increase release cytokines (TNF, IL-6, or IL-8) by stimulated mononuclear cells *in vitro*\[[@B26],[@B27]\].
Complications of sepsis have been related to an intense host response based on a delicate equilibrium between various proinflammatory and anti-inflammatory mediators \[[@B28]\]. Overwhelming production of proinflammatory cytokines, such as TNF-α, IL-1β, IL-2 receptor, IL-6 and IL-8, may induce biochemical and cellular alterations either directly or by orchestrating secondary inflammatory pathways.
Reddyl and coworkers \[[@B5]\] evaluated the effect of pretreatment with NS-398, a highly selective COX-2 inhibitor, on survival and inflammatory mediator production in two models of sepsis in mice (LPS challenge and peritonitis induced by caecal ligation and puncture \[CLP\]). They found that selective inhibition of COX-2 resulted in improvement in early survival in murine endotoxaemia but not in a more physiologically relevant model of abdominal sepsis (CLP). The early improvement in survival in endotoxin-challenged animals was not attributable to changes in inflammatory cytokine expression or organ-specific neutrophil sequestration. Pretreatment with NS-398 failed to improve long-term survival in either of the models studied, although in the endotoxaemia model administration of the COX-2 inhibitor had a modest salutary effect on early mortality. In addition, although treatment with NS-398 blocked LPS-induced increases in the circulating levels of immunoreactive PGE~2~, injection of the COX-2 inhibitor did not modulate plasma concentrations of TNF or the CXC chemokine KC.
Knoferl and coworkers \[[@B29]\] also evaluated the effect of pretreatment with NS-398, that trauma/haemorrhage results in activation of Kupffer cells to release inflammatory mediators and it leads to immunosuppression. *In vitro*production of IL-6 by Kupffer cells after CLP was significantly reduced by *in vivo*NS-398 treatment. However, NS-398 had no effect on TNF-α levels *in vivo*or *in vitro*. Strong and coworkers \[[@B12]\] showed that administration of NS-398 for 24 hours after trauma improved survival when mice were subjected to CLP and puncture 7 days later. It is noteworthy that NS-398 exhibited protective effects in two models of sepsis characterized by infection in the setting of trauma-induced immunosuppression, whereas the drug was largely ineffective when sepsis was induced in immunocompetent animals. Dallal and coworkers \[[@B30]\] demonstrated that T-cell suppression during neonatal sepsis is accompanied by a decrease in IL-2 production. Such suppression was ameliorated by COX-2 inhibitor, suggesting a role for PGE~2~in suppressed T-cell-mediated immune function in neonatal sepsis. Arons and colleagues \[[@B22]\] compared the clinical and physiological characteristics of febrile septic patients with those of hypothermic septic patients, and compared plasma levels of cytokines TNF-α and IL-6 and thromboxane B~2~and prostacyclin between hypothermic septic patients and febrile patients. They administered ibuprofen but found that this drug had no effect on cytokine levels.
Reddyl and coworkers \[[@B5]\] indicated that pharmacological inhibition of COX-2 has only very modest effects on outcome in experimental sepsis or endotoxaemia. Because these findings are discrepant with respect to those obtained with isoform nonselective agents, it is regrettable that those investigators did not include a \'positive control\' arm in their studies to evaluate the effects of treatment with an agent such as indomethacin or ibuprofen in their laboratory\'s models of sepsis. In our study we did not observe any significant changes in systemic cytokine levels during NSAID administration in humans with severe sepsis. Cytokine levels in plasma do not necessarily reflect local synthesis of cytokines by cells. Many cells have surface receptors for these cytokines with high binding properties, and target cells and soluble receptors trap cytokines. Thus, cytokines released at the local level may remain undetected in plasma. In the present study we found plasma cytokine levels to remain unchanged over a period of 72 hours.
Wang and coworkers \[[@B31]\] conducted a study to determine whether inhibition of PGI~2~synthesis prevents the hyperdynamic response in early sepsis in animals. Those investigators found that inhibition of PGI~2~production did not prevent the hyperdynamic and hypercardiovascular responses during early sepsis; hence, mediators other than PGI~2~appear to play a major role in producing the hyperdynamic response under such conditions. Fox and colleagues \[[@B32]\] postulated that the attenuated pulmonary and systemic vascular contractility observed in sepsis was secondary to the release of vasodilator PGs. They used the COX inhibitor meclofenamate to inhibit PG synthesis in a model of hyperdynamic sepsis, and found that meclofenamate had no effect on either the pulmonary or systemic response to phenylephrine infusion in septic animals. However, Wanecek and coworkers \[[@B11]\] demonstrated that endotoxin-induced pulmonary hypertension in the pig can be prevented with a combination of the nonpeptide mixed endothelin receptor antagonist bosentan and the COX inhibitor diclofenac. They found that the combination of bosentan and diclofenac induced systemic and pulmonary vasodilatation. During endotoxin shock, this drug combination efficiently counteracted pulmonary hypertension and improved cardiac performance, and splenic and renal blood flows. These favourable circulatory effects might have resulted in a reduction in both sympathetic nervous system activation and metabolic acidosis. In the present study we found that lornoxicam had no effect on the cardiovascular and pulmonary systems in severe sepsis in humans, but our study was designed to assess the effects of lornoxicam treatment given before septic shock but after systemic inflammatory response syndrome. For this reason we identified no serious cardiovascular and pulmonary system problems in the patients studied.
Arons and coworkers \[[@B22]\] compared clinical and physiological characteristics of febrile septic patients with those in hypothermic septic patients, and compared plasma levels of cytokines TNF-α and IL-6, and thomboxane B~2~and prostacyclin between hypothermic septic patients and febrile patients. Those investigators found that ibuprofen treatment had a positive impact on vital signs, organ failure and mortality in hypothermic septic patients, and concluded that ibuprofen could substantially decrease mortality in this selected group of septic patients. In our study we found that lornoxicam had no effect on vital signs and mortality in patients with severe sepsis. The overall ICU mortality rate was 37.5% (15 patients out of 40) in total, and these deaths were all attributable to septic shock. However, all of the patients died after completion of the study.
Lornoxicam has been shown to produce less gastric toxicity than its nonselective counterparts. This may be especially important in critically ill patients, who are at significantly greater risk for developing gastric ulceration. In addition, the lack of inhibitory effect on platelet function, which occurs with the use of COX-2 selective compounds, may decrease the incidence of bleeding complications \[[@B17],[@B19]\]. In the present study we did not identify any lornoxicam related adverse effects.
In summary, we found that intravenous lornoxicam had no effect on haemodynamic and biochemical parameters, cytokine levels, or patient outcomes in severe sepsis. Selective inhibition of COX-2 in sepsis requires further study. However, the findings reported here, indicating that lornoxicam lacks benefit in patients with severe sepsis, are disappointing.
Key messages
============
• Administration of intravenous lornoxicam appeared to confer no benefit in patients with severe sepsis.
Competing interests
===================
The author(s) declare that they have no competing interests.
Abbreviations
=============
APACHE = Acute Physiology and Chronic Health Evaluation; CLP = caecal ligation and puncture; COX = cyclo-oxygenase; ICU = intensive care unit; IL = interleukin; LPS = lipopolysaccharide; NSAID = nonsteroidal anti-infllammatory drug; PG = prostaglandin; SOFA = Sepsis-related (Sequential) Organ Failure Assessment; TNF = tumour necrosis factor.
Figures and Tables
==================
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Sepsis-related (or Sequential) Organ Failure Assessment (SOFA) scores
:::
Parameter SOFA score
-------------------------------------------- ---------------- ------------------- ---------------------------------------- ------------------------------------------------------------------------------------ -------------------------------------------------------
Respiration (PaO~2~/FiO~2~ratio) \>400 ≤ 400 ≤ 300 ≤ 200 with respiratory support ≤ 100
Coagulation (platelets × 10^3^/mm^3^ \>150 ≤ 150 ≤ 100 ≤ 50 ≤ 20
Liver (bilirubin \[mg/dl (μmol/l)\]) \<1.2 (\<20) 1.2--1.9 (20--32) 2.0--5.9 (33--101) 6.0--11.9 (102--204) \>12.0 (\>204)
Cardiovascular (hypotension) No hypotension MAP \<70 mmHg Dopamine ≤ 5 or dobutamine at any dose Dopamine\>5 or adrenaline (epinephrine) ≤ 0.1 noradrenaline (norepinephrine) ≤ 0.1 Dopamine \>15 or adrenaline \>0.1 noradrenaline \>0.1
Central nervous system (GCS score) 15 19--14 10--12 6--9 \<6
Renal (creatine \[mg/dl\] or urine output) \<1.2 1.2--1.9 2.0--3.4 3.5--4.9 or \<500 ml/day \>5 or \<200 ml/day
FiO~2~, fractional inspired oxygen; GCS, Glasgow Coma Scale; MAP, mean arterial pressure;
PaO~2~, arterial oxygen tension.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Demographic and clinical characteristics of lornoxicam treated and placebo patients
:::
Characteristic Lornoxicam group (*n*= 20) Placebo group (*n*= 20)
---------------------------- ---------------------------- -------------------------
Age (years \[range\]) 49 (19--87) 51 (20--89)
Sex (male/female) 13/7 9/11
Source of infection
Respiratory 15 17
Gastrointestinal 2 1
Blood 2 1
Urinary tract 1 1
APACHE II score^a^ 17.10 ± 3.58 18 ± 3.72
SOFA score^a^ 5.90 ± 1.72 6.20 ± 2.2
Duration of ventilation^a^ 6.1 ± 2.4 5.8 ± 3.1
Length of stay^a^ 10.2 ± 7.1 9.2 ± 8.4
Mortality rate (%) 35 40
There were no significant differences between the groups. ^a^Values are expressed as mean ± standard deviation. APACHE, Acute Physiology and Chronic Health Evaluation; SOFA, Sepsis-related (or Sequential) Organ Failure Assessment.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Haemodynamic, oxygen and temperature variables
:::
Parameter Baseline Hours after the start of infusion
------------------------------- ------------- ----------------------------------- -------------- --------------
Heart rate (beats/min)
Lornoxicam 98 ± 24 99 ± 24 99 ± 22 96 ± 22
Placebo 95 ± 17 100 ± 23 98 ± 23 97 ± 22
Mean arterial pressure (mmHg)
Lornoxicam 91 ± 16 92 ± 13 89 ± 13 93 ± 13
Placebo 91 ± 18 93 ± 14 91 ± 12 89 ± 11
Arterial pH
Lornoxicam 7.34 ± 0.08 7.36 ± 0.07 7.33 ± 0.10 7.36 ± 0.06
Placebo 7.35 ± 0.07 7.37 ± 0.07 7.34 ± 0.07 7.37 ± 0.05
PaCO~2~(torr)
Lornoxicam 34.8 ± 12.1 34.6 ± 11.8 35.00 ± 9.2 36.54 ± 11.1
Placebo 32.6 ± 10 33.5 ± 10.2 36.3 ± 10.13 34.33 ± 12
PaO~2~/FiO~2~ratio (torr)
Lornoxicam 182 ± 68 186 ± 56 188 ± 64 189 ± 76
Placebo 184 ± 76 187 ± 45 181 ± 68 185 ± 68
SaO~2~(%)
Lornoxicam 96.1 ± 3 96.1 ± 3.1 95.9 ± 3.8 95.9 ± 4.2
Placebo 96.9 ± 3.1 96.0 ± 3.2 96.0 ± 2.8 95.8 ± 3.9
Temperature (°C)
Lornoxicam 37.8 ± 0.75 37.2 ± 0.6 37.8 ± 0.5 37.9 ± 0.4
Placebo 37.6 ± 0.57 37.8 ± 0.4 37.6 ± 0.6 37.8 ± 0.5
No significant differences were found between groups. Data are expressed as mean ± standard deviation. FiO~2~, fractional oxygen tension; PaCO~2~, arterial carbon dioxide tension; PaO~2~, arterial oxygen tension; SaO~2~, arterial oxygen saturation.
:::
::: {#T4 .table-wrap}
Table 4
::: {.caption}
######
Biochemical parameters
:::
Parameter Baseline Hours after the start of infusion
--------------------------------- -------------- ----------------------------------- -------------- --------------
Lactate (mg/dl)
Lornoxicam 25.2 ± 4.1 25.0 ± 3.9 25.9 ± 5.2 26.2 ± 3.7
Placebo 26.3 ± 3.8 26.7 ± 2.9 27.0 ± 3.1 26.9 ± 4.8
Platelets (×10^9^/l)
Lornoxicam 192.9 ± 16.5 193.5 ± 15.0 178.1 ± 15.4 182.9 ± 15.6
Placebo 188.5 ± 14.4 190.8 ± 15.1 188.7 ± 13.3 190.5 ± 16.7
Leucocytes (×10^9^/l)
Lornoxicam 14 ± 8.3 15.8 ± 8.3 16.0 ± 7.8 15.9 ± 6.2
Placebo 13.5 ± 6.6 14.7 ± 4.7 15.7 ± 6.7 14.8 ± 7.4
Bilirubin (mg/dl)
Lornoxicam 0.89 ± 0.38 0.90 ± 0.45 0.93 ± 0.33 0.90 ± 0.34
Placebo 0.90 ± 0.62 0.91 ± 0.38 0.92 ± 0.28 0.91 ± 0.36
Alanine aminotransferase (IU/l)
Lornoxicam 35.4 ± 5.5 36.0 ± 10.8 35.1 ± 8.6 36.7 ± 8.4
Placebo 35.4 ± 7.4 35.5 ± 5.2 36.2 ± 6.0 37.4 ± 4.9
Creatinine (mg/dl)
Lornoxicam 1.13 ± 0.96 1.15 ± 0.85 1.2 ± 0.3 1.28 ± 0.8
Placebo 1.01 ± 0.91 1.1 ± 0.65 1.08 ± 0.7 1.1 ± 0.8
No significant differences were found between groups. Data are expressed as mean ± standard deviation.
:::
::: {#T5 .table-wrap}
Table 5
::: {.caption}
######
Cytokine levels
:::
Cytokine Baseline Hours after the start of infusion
---------------------- ------------- ----------------------------------- -------------- --------------
TNF-α (pg/ml)
Lornoxicam 25.7 ± 15 27.4 ± 16 26.68 ± 12.9 28.1 ± 16.8
Placebo 24.6 ± 18 25.4 ± 16.9 25.55 ± 11.8 26.3 ± 14.2
Il-1β (pg/ml)
Lornoxicam 6.4 ± 3.8 6.48 ± 6.07 6.21 ± 3.26 6.57 ± 1.8
Placebo 6.35 ± 1.4 6.2 ± 4.30 6.34 ± 4.2 6.19 ± 2.7
IL-2 receptor (U/ml)
Lornoxicam 1950 ± 1266 1890 ± 1150 1929 ± 1027 2050 ± 1100
Placebo 2270 ± 1110 2179 ± 1005 2300 ± 1190 2268 ± 1000
IL-6 (pg/ml)
Lornoxicam 100.6 ± 58 105.2 ± 54 108.5 ± 47 104.8 ± 38
Placebo 114.5 ± 385 115.6 ± 48 113.6 ± 51 116 ± 28
IL-8 (pg/ml)
Lornoxicam 171.50 ± 35 171.6 ± 19.3 172.1 ± 12.6 168.9 ± 11.3
Placebo 169.55 ± 27 168.3 ± 18.4 169.8 ± 18.2 171.0 ± 18.2
No significant differences were found between groups. Data are expressed as mean ± standard deviation. IL, interleukin; TNF, tumour necrosis factor.
:::
|
PubMed Central
|
2024-06-05T03:55:54.857879
|
2004-10-27
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065065/",
"journal": "Crit Care. 2004 Oct 27; 8(6):R474-R482",
"authors": [
{
"first": "Dilek",
"last": "Memiş"
},
{
"first": "Beyhan",
"last": "Karamanlıoğlu"
},
{
"first": "Alparslan",
"last": "Turan"
},
{
"first": "Onur",
"last": "Koyuncu"
},
{
"first": "Zafer",
"last": "Pamukçu"
}
]
}
|
PMC1065066
|
See related commentary <http://ccforum.com/content/8/6/427>
Introduction
============
Until recently hyperglycaemia after surgery was considered to be a benign phenomenon. However, in a landmark study, van den Berghe and coworkers \[[@B1]\] showed that treating transient postoperative hyperglycaemia with intensive insulin therapy in a surgical intensive care unit (ICU) dramatically reduces mortality and morbidity. Strict glucose control (target range between 4.4 mmol/l and 6.1 mmol/l) was responsible for a reduction in both ICU and in-hospital mortality, which was primarily attributed to the prevention of septic complications \[[@B1]\].
The population studied by van den Berghe and coworkers was diverse but consisted primarily of patients who underwent cardiac surgery (63%). Others have found beneficial effects of intensive insulin therapy in patients with obstructive vascular disease such as acute myocardial infarction and acute stroke, and in those who have undergone cardiovascular bypass surgery \[[@B2]-[@B7]\]. Strict glucose control is relatively time consuming for ICU personnel because of frequent glucose monitoring, and it may be hazardous because of the risk for hypoglycaemia. It is therefore important to determine which patient groups in the ICU are likely to benefit most or least from aggressively correcting hyperglycaemia.
We investigated whether postoperative hyperglycaemia is a risk factor for postoperative infections and prolonged in-hospital stay in a cohort of patients undergoing resection for adenocarcinoma of the oesophagus (i.e. patients with a low prevalence of risk factors for insulin resistance and cardiovascular disease but who are subject to great postoperative stress).
Methods
=======
Patients
--------
A total of 220 consecutive patients with adenocarcinoma of the oesophagus from two university hospitals in Amsterdam and Rotterdam were included in a previously reported randomized clinical trial investigating differences in short-term and long-term morbidity and mortality between two surgical approaches for resection of oesophageal adenocarcinoma \[[@B8]\]. Classification into American Society of Anesthesiologists (ASA) class 1 or 2 was a requirement for eligibility in that study.
Only patients included in Amsterdam were included in the present analysis (*n*= 160), because glucose values were taken only in a small proportion of the Rotterdam patients. In nine cases oesophageal resection was cancelled peroperatively because of distant dissemination of tumour, leaving 151 patients for this analysis.
Data collection
---------------
Glucose values were automatically determined with each arterial blood sample test (Ciba Corning 865; Chiron Diagnostics, Medford, MA, USA), and were collected retrospectively from laboratory reports. Forced expiratory volume in 1 s (FEV~1~) expressed as percentage of the predicted value corrected for age and sex, and patient height (to calculate body mass index \[BMI\]) were collected retrospectively from preoperative lung function reports. Insulin use in the first 48 hours after surgery was determined retrospectively from ICU charts. In the prospective cohort patients were visited at least twice a week by one of the investigators to score postoperative complications. Postoperative infections were defined as signs of infection and positive culture \[[@B9]\]. History of cardiovascular disease, hypertension, weight loss, ASA class, postoperative occurrence of left ventricular failure and length of hospital stay were determined prospectively \[[@B8]\].
Patients were allowed to eat as they wished until 24 hours before surgery. Patients with more than 10% weight loss in the year preceding surgery received preoperative enteral tube feeding. Postoperatively, all patients received continuous tube feeding through a needle jejunostomy, starting 12--24 hours postoperatively, with 25 ml/hour tube feeding containing immunomodulatory nutrients (Impact^®^; Novartis, Basel, Switserland). As a general rule, patients received 30 ml glucose 5% intravenously during the first 48 hours after surgery and patients were treated with insulin when glucose values exceeded 12 mmol/l.
Statistical analysis
--------------------
For each patient the mean postoperative glucose concentration was calculated using all available glucose measurements obtained until 48 hours postoperatively. For further analysis, mean postoperative glucose concentrations were divided into quartiles because of nonparametric distribution.
Univariate regression analysis was used to select parameters associated with infectious complications and length of hospital stay. Parameters with *P*\< 0.1 in univariate regression analysis were examined in multivariate analysis to define the independent contribution of each possible risk factor \[[@B10]\]. Postoperative glucose concentrations were automatically selected for multivariate analysis because it was the main aim of the study to determine their relationship with outcome. Logistic regression analysis was used for infectious complications, and linear regression analysis was used for length of stay. Because of nonparametric distribution, length of stay data were logarithmically transformed before regression analysis.
Parameters included in the analysis
-----------------------------------
Age, amount of preoperative weight loss, BMI and FEV~1~were entered into regression analyses as continuous variables. Postoperative glucose levels, insulin use within 48 hours after surgery, type of surgical procedure, sex, ASA class, history of hypertension, coronary artery disease, cardiac valve disease or arrhythmia, clinical staging of the tumour and presence of diabetes mellitus were entered as categorical variables.
Results
=======
Preoperative characteristics are summarized in Table [1](#T1){ref-type="table"}. At least one postoperative glucose value could be retraced in 150 out of 151 cases (99%; median 7 glucose values per patient; range 1--21). A glucose level greater than 6.1 mmol/l was found in 97% of patients. During the first 48 hours after surgery, insulin was administrated to four patients with known diabetes mellitus and to five patients without diabetes mellitus, but insulin administration could not be retraced in one patient with known diabetes mellitus. At least one infectious complication occurred in 55 patients (36%) and more than one infection occurred in 15 patients (9.9%). Pneumonia occurred in 44 patients, wound infection in 15, urinary tract infection in six and sepsis in seven. Patients were admitted to the ICU for a median duration of 3 days (range \<24 hours to 71 days). The median length of stay was 16 days (range 9--154 days). The incidences of postoperative left ventricular failure (*n*= 13; 8.6%) and in-hospital death (*n*= 5; 3.3%) were too low to allow for regression analysis.
Postoperative glucose levels and postoperative infections
---------------------------------------------------------
According to univariate regression analysis, no association was found between postoperative glucose levels and infectious complications (*P*= 0.21; Fig. [1a](#F1){ref-type="fig"}) or between insulin administration and infectious complications (*P*= 0.37; odds ratio \[OR\] 0.5, 95% confidence interval \[CI\] 0.1--2.4). Parameters associated with postoperative infections in univariate regression analysis were history of cardiac valve disease or arrhythmia (*P*= 0.026; OR 11.5, 95% CI 1.35--98.2), FEV~1~per 10% increase (*P*= 0.021; OR 0.78, 95% CI 0.63--0.96; OR per 10% of expected FEV~1~), age per 10 years (*P*= 0.069; OR 1.39, 95% CI 0.98--1.97) and duration of surgery per hour (*P*= 0.059; OR 1.23, 95% CI 0.99--1.52). In the subgroup of patients with an ICU stay in excess of 5 days, there was no association between postoperative hyperglycaemia and infection (*P*= 0.9 for trend; *P*= 0.8 by ?^2^analysis). Also in multivariate analysis, postoperative hyperglycaemia was not found to be a predictor of postoperative infection (*P*= 0.28; OR 1.21, 95% CI 0.86--1.72; Table [2](#T2){ref-type="table"}). Also, patients with at least one glucose value in excess of 10 mmol/l were not at greater risk for infections (data not shown).
Postoperative glucose levels and length of stay
-----------------------------------------------
In univariate analysis, a positive association was found between postoperative hyperglycaemia and length of hospital stay (P \< 0.001; ß = 0.053; standard error \[SE\] of ß = 0.014), but not with insulin administration (*P*= 0.5; ß = -0.56; SE of ß = 0.7). Other parameters associated with length of in-hospital stay were duration of surgery (*P*\< 0.001; ß = 0.050; SE of ß = 0.010), transthoracic procedure (*P*\< 0.001; ß = 0.119, SE of ß = 0.032), BMI (*P*= 0.036; ß = 0.013; SE of ß = 0.006) and history of cardiac valve disease or arrhythmia (*P*= 0.103; ß = 0.130; SE of ß = 0.079). After correction for these variables in multivariate analysis, mean postoperative glucose concentration was found not to be an independent risk factor for prolonged hospital stay (*P*= 0.20; Table [3](#T3){ref-type="table"}). Adding duration of ICU stay greater than 5 days as an interaction term was not statistically significant (*P*= 0.12).
Discussion
==========
In a cohort of patients undergoing highly invasive surgery for oesophageal cancer, we found that postoperative hyperglycaemia was present in almost all patients but that it was not associated with increased incidence of postoperative infections and length of hospital stay.
Van den Berghe and coworkers \[[@B1]\] found that lowering postoperative hyperglycaemia with intensive insulin therapy significantly decreased morbidity and mortality in postoperative patients. *Post hoc*analysis revealed that both administration of insulin and, possibly to a greater degree, lower glucose levels contributed to better outcome \[[@B11]\]. However, it is unclear how the effect of intensive insulin therapy in surgical intensive care patients can be explained and which patient groups benefit most from intensive insulin therapy. We propose the following explanation for the seemingly contradictory findings of our study.
The population evaluated in the study by van den Berghe and coworkers \[[@B1]\] consisted mainly of patients undergoing cardiovascular surgery. Transient or \'stress induced\' hyperglycaemia was previously reported to be associated with a poor prognosis, primarily in patients with obstructive vascular disease such as those with acute myocardial infarction and acute stroke, and in those who have undergone cardiovascular bypass surgery and peripheral vascular surgery \[[@B12]-[@B16]\]. Few patients in our cohort suffered from (cardio)vascular disease because ASA class 1 or 2 was a prerequisite for inclusion in the study, and only 11% had a history of coronary artery disease. It could thus be hypothesized that, in a population with little vascular disease, high postoperative glucose levels are not associated with poor outcome.
In response to surgery, both plasma glucose levels and free fatty acid (FFA) levels rise \[[@B17]\]. Pathophysiological mechanisms that may explain the relationship between stress induced hypermetabolism and poor outcome in patients with cardiovascular disease include the following: toxic effects of elevated FFA levels on the ischaemic myocardium \[[@B18]\]; elevated FFA levels and hyperglycaemia causing QT prolongation \[[@B19]\]; hyperglycaemia attenuating ischaemic preconditioning \[[@B20]\]; and hyperglycaemia causing reduced collateral coronary perfusion \[[@B21]\]. Haemodynamic effects of glucose and insulin may also play an important role in the pathophysiology of stress induced hypermetabolism. Hyperglycaemia has vasoconstrictive effects \[[@B22]\], which may aggravate tissue ischaemia, particularly in patients with obstructive vascular disease. Insulin has been reported to have vasodilatory effects, and part of the beneficial effect of intensive insulin therapy may be explained by increasing tissue perfusion \[[@B23]\].
Our data do not exclude the possibility that intensive insulin therapy or glucose--insulin--potassium infusions may still be beneficial in this particular subgroup of patients. The benefits of intensive insulin therapy may not solely be attributed to lowering hyperglycaemia, but may be mediated by the effect of insulin on protein and lipid metabolism, independent of its effects on glucose metabolism. In patients with sepsis and cancer, lower levels of insulin are needed to restore lipid levels than glucose levels \[[@B24]\]. Similarly, depleted protein storage and severe surgical stress after oesophageal resection may impair the immune response postoperatively and thus increase the risk for postoperative infection \[[@B25]\], which may be ameliorated by insulin. However, the administration of insulin was not associated with lower infection risk in our cohort.
A shortcoming of the present study is that the number of glucose measurements taken in each patient was not standardized, because of the study\'s retrospective design. For some patients more glucose measurements were available than for others, and this may have influenced our results. However, glucose measurements were taken randomly with each arterial blood gas analysis, and because mean postoperative glucose levels were used, the relative weight of incidental extreme values was diminished. A strength of our cohort is its homogeneity. It represents a unique group of patients with high postoperative stress and a low frequency of risk factors for obstructive vascular disease.
Conclusion
==========
Despite the limitations associated with the retrospective analysis of a prospective study, our data indicate that early postoperative hyperglycaemia is more likely to be a risk marker than a risk factor in a patient group encountering severe surgical stress but with a low prevalence of cardiovascular disease. We therefore suggest that the value of intensive insulin therapy, which is time consuming and potentially hazardous, needs further investigation in this particular patient group.
Key messages
============
• Postoperative hyperglycaemia after oesophagectomy was not found to be associated with postoperative infection risk.
• Postoperative hyperglycaemia after oesophagectomy was found to be associated with longer duration of postoperative stay. However, when corrected for possible confounders, postoperative hyperglycaemia was not found to be an independent risk factor for longer duration of stay.
• Strict glycaemic control may not be beneficial for patients after oesophagectomy.
Competing interests
===================
The author(s) declare that they have no competing interests.
Author contributions
====================
TMV participated in the design of the study, data collection, data analysis and writing of the manuscript. JHDV participated in data analysis and writing of the manuscript. JBH participated in the design of the study, data collection, data analysis and writing of the manuscript. FH participated in the design of the study and writing of the manuscript. JJvL participated in the design of the study, data collection and writing of the manuscript. JBLH participated in the design of the study, writing of the manuscript and coordinated the study.
Abbreviations
=============
ASA = American Society of Anesthesiologists; BMI = body mass index; CI = confidence interval; FEV~1~= forced expiratory volume in 1 s; FFA = free fatty acid; ICU = intensive care unit; OR = odds ratio; SE = standard error.
Acknowledgements
================
The authors gratefully thank Michiel Berenschot for assisting with data collection and Glaxo Smith Kline, The Netherlands, for providing financial support for this study.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Percentage of **(a)**infections and **(b)**median length of hospital stay per glucose quartile: first quartile 5.2--7.4 mmol/l, second quartile 7.5--8.2 mmol/l, third quartile 8.3--9.2 mmol/l, and fourth quartile 9.3--17.2 mmol/l. The error bars in panel b represent the interquartile range.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Other possible risk factors for infection and length of stay
:::
Preoperative parameter Value Missing values (%)
----------------------------------------------------------- ------------- --------------------
Age in years (mean ± SD) 62.4 ± 10.0 0
Male sex (*n*\[%\]) 127 (84.1) 0
ASA 0
Class I (*n*\[%\]) 47 (31.1)
Class II (*n*\[%\]) 104 (68.9)
History of cardiac valve disease or arrhythmia (*n*\[%\]) 7 (4.6) 0.7
History of hypertension (*n*\[%\]) 21 (13.9) 0.7
History of coronary artery disease (*n*\[%\]) 16 (10.6) 0.7
BMI (mean ± SD) 25.4 ± 3.3 25.8
Percentage of expected FEV~1~(mean ± SD) 101 ± 18 18.5
Diabetes mellitus (*n*\[%\]) 9 (6.0) 0.7
Preoperative weight loss (kg; mean ± SD) 5.3 ± 6.6 7.9
Clinical staging of tumor 1.3
Stage I (*n*\[%\]) 18 (11.9)
Stage II (*n*\[%\]) 68 (45.0)
Stage III (*n*\[%\]) 58 (38.4)
Stage IV (*n*\[%\]) 5 (3.3)
Allocated to transthoracic procedure (*n*\[%\]) 78 (51.7%) 0
Duration of surgery (hours; mean ± SD) 5.3 (1.6) 0.7
Insulin use within 48 hours after surgery (*n*\[%\]) 9 (6.0%) 9.2
A total of 151 patients were included. ASA, American Society of Anesthesiologists; BMI, body mass index; FEV~1~, forced expiratory volume in 1 s; SD, standard deviation.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Multivariate analysis of infectious complications
:::
Prognostic variable OR (95% CI) *P*
------------------------------------------------ ------------------- -------
FEV~1~(per 10% of expected FEV~1~) 0.79 (0.63--0.99) 0.045
History of cardiac valve disease or arrhythmia 7.30 (0.78--68.3) 0.081
Duration of surgery 1.27 (0.98--1.64) 0.069
Age per 10 years 1.36 (0.90--2.07) 0.142
Mean postoperative glucose 1.21 (0.86--1.72) 0.279
CI, confidence interval; FEV1, forced expiratory volume in 1 s; OR, odds ratio.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Multivariate analysis of length of stay
:::
Prognostic variable ß SE of ß *P*
------------------------------------------------ ------- --------- -------
Duration of surgery 0.062 0.021 0.004
BMI 0.010 0.006 0.072
Mean postoperative glucose 0.024 0.018 0.195
History of cardiac valve disease or arrhythmia 0.058 0.091 0.527
Transhiatal procedure 0.034 0.065 0.599
BMI, body mass index; SE, standard error.
:::
|
PubMed Central
|
2024-06-05T03:55:54.861985
|
2004-10-18
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065066/",
"journal": "Crit Care. 2004 Oct 18; 8(6):R437-R442",
"authors": [
{
"first": "Titia M",
"last": "Vriesendorp"
},
{
"first": "J Hans",
"last": "DeVries"
},
{
"first": "Jan BF",
"last": "Hulscher"
},
{
"first": "Frits",
"last": "Holleman"
},
{
"first": "Jan J",
"last": "van Lanschot"
},
{
"first": "Joost BL",
"last": "Hoekstra"
}
]
}
|
PMC1065067
|
Introduction
============
Bacterial infection is an important cause of mortality and morbidity in newborns and critically ill paediatric patients \[[@B1],[@B2]\]. The high risks associated with untreated infection and the lack of accurate clinical or laboratory prediction methods result in a low threshold for initiating empirical antibiotic therapy. In neonatal and paediatric intensive care, antibiotic therapy is used in as many as 80% of patients, with an average of about 50% \[[@B3]\]. Only a minority of treated patients suffer from true infection. The majority receive antibiotics for 48--72 hours because clinical signs suggest possible infection and laboratory parameters are unable to rule out infection. In otherwise healthy newborns, this practice causes prolonged separation from the mother and increased the costs of care \[[@B4],[@B5]\]. The high prevalence of unnecessary antibiotic therapy augments the risk for selecting resistant bacterial strains. Despite liberal antibiotic prescription, in some patients sepsis is not diagnosed until they have progressed to serious conditions such as septic shock.
Several groups have suggested that measurement of cytokines may be done to facilitate early diagnosis \[[@B6]-[@B8]\]. We previously reported diagnostic test accuracy studies in which we derived a prediction model based on the measurement of plasma levels of granulocyte colony-stimulating factor (G-CSF) and IL-8, and tracheal aspirate levels of G-CSF \[[@B9],[@B10]\]. If plasma cytokine concentrations rise above pre-specified thresholds, then serious bacterial bloodstream infection is highly likely. Gram-negative sepsis is practically excluded if plasma levels remain low. Although plasma measurements assist in ruling out life-threatening sepsis, localized infections such as ventilator-associated pneumonia \[[@B11]\] cannot be diagnosed on the basis of blood derived cytokine concentrations. However, we previously showed tracheal aspirate levels of G-CSF to assist in diagnosing ventilator-associated pneumonia \[[@B10]\], which is the most frequent reason for prescribing antibiotics in our unit \[[@B3]\]. We recently conducted validation studies for plasma measurements of IL-8 and G-CSF and tracheal aspirate levels of G-CSF, employing a new laboratory method that allows simultaneous determination of parameters from 50 μl blood or tracheal aspirate. We refined the fluorescent bead-based immunoassay to reduce the assay turnaround time from 4.5 hours to 2 hours, rendering it suitable for routine clinical use.
To assess the clinical efficacy of the new diagnostic measures, we suggest that a randomized controlled trial be conducted comparing two management strategies. The control strategy will consist of routine management, with the exception that physicians are requested to provide a probability estimate for the presence of bacterial infection whenever a diagnostic work up (blood cultures or tracheal aspirate culture) is ordered. The intervention strategy will consist of cytokine measurement from the sample and provision of a result based post-test probability within a few hours after sample collection. The null hypothesis states that the management arms will not differ with respect to antibiotic utilization rate, measured as the number of days on systemic antibiotic treatment per 1000 days of hospitalization. The secondary null hypothesis states that the arms will not differ with respect to costs associated with hospital acquired septic shock.
Methods
=======
Design
------
The study is a multicentre randomized controlled trial comparing a new diagnostic treatment strategy for diagnosing bacterial infection versus standard care in critically ill newborns and children. During a 16-week period in 2003 we conducted a pilot study, which tested the intervention and data collection procedures, and led to modifications to the study design. The pilot study is outlined in detail below. In brief, physicians provide pre-test probabilities whenever they order a diagnostic work up for sepsis or ventilator-associated pneumonia (microbiological cultures). This includes any prescription of antibiotics. In the intervention arm, physicians are provided with cytokine results and the updated post-test probability. In the control arm no information is given.
Eligibility criteria for participants
-------------------------------------
All patients admitted to the interdisciplinary neonatal or paediatric intensive care unit (ICU) of the Children\'s Hospital of Zurich are eligible. Patients who are referred to other wards within 24 hours after admission will be excluded from data analysis, because in these patients the decision to stop antibiotic treatment is no longer the responsibility of participating intensivists.
Setting
-------
The participating university hospital is the tertiary referral centre for Eastern and Southern Switzerland, and serves a population of approximately 3 million. The Department of Neonatology and Pediatric Intensive Care at the University Children\'s Hospital of Zurich contributes patients from its two ICUs, named unit A and unit B. Both units have average occupancy of 8--10 beds. The two units admit between 900 and 1000 patients annually, with the number of hospitalization days amounting to 5500 each year. The patient population in unit A includes infants of extremely low birth weight referred from other hospitals, critically ill children and adolescent patients, trauma victims and high-risk surgical patients. Unit B predominantly cares for infants and children who have undergone cardiac surgery.
Intervention and controls
-------------------------
### Control strategy
For patients randomized to the control arm, if the physician orders microbial cultures then they are obliged to document their best estimate of the probability that the patient has sepsis or pneumonia on two logarithmic visual-analogue scales (range 0--100%). This documentation is mandatory and must be marked on the laboratory form (Fig. [1](#F1){ref-type="fig"}). In the control arm blood or tracheal aspirate specimens are not analyzed; thus, physicians do not receive any information beyond routinely available data. Under the control strategy antibiotic treatment is managed according to current recommendations (cessation of therapy after 48 hours provided that blood cultures remain negative).
### Intervention strategy
If patients are randomized to the intervention arm, then physicians are also obliged to document their best estimate of the probability that the patient has sepsis or pneumonia, again on two logarithmic visual-analogue scales (range 0--100%). This documentation is again mandatory and must be marked on the laboratory form (Fig. [1](#F1){ref-type="fig"}). In the intervention arm, blood or tracheal aspirate specimens are analyzed and results are returned to the unit before 1 p.m. Physicians receive the raw cytokine values as well as the calculated likelihood ratio and the post-test probability (Fig. [2](#F2){ref-type="fig"}). This information is provided in addition to routinely available data. Provided that the available post-test probability indicates absence of infection, physicians are encouraged to stop antibiotic treatment. It is suggested that antimicrobial therapy be continued if the post-test probability indicates infection. It is important to note that the protocol provides only \'suggestions\', and that the final decision regarding therapy is left to the discretion of the responsible clinician. This is similar to clinical routine, in which diagnostic results may suggest alterations to treatment decisions but they do not dictate treatment.
Randomization
-------------
The units of randomization are calendar days. Randomization is generated through pre-specified assignment of 15 working days/month as intervention days. Physicians remain blinded to the allocation roster. Thirty minutes after the deadline for delivery of samples to the laboratory (10 a.m.), physicians are informed about the randomization status (control or intervention) of the day. In this way, physicians are able to adjust their decision making while they await test results if they so wish.
Data collection
---------------
Routine sepsis work up includes collection of blood cultures, other microbial specimens where appropriate, and measurements of white blood cell count, including differential and plasma levels of C-reactive protein. Routine surveillance for ventilator-associated pneumonia comprises microbiological examination of the tracheal aspirate, including cultures. As described above, physicians must provide two probability estimates, one for the presence of sepsis and one for pneumonia, whenever they order a sepsis or pneumonia work up. This ensures that clinicians state their estimate before knowledge of the test result. These estimates (pre-test probabilities) are integrated with cytokine concentrations derived from likelihood ratios for sepsis or pneumonia using Bayes\' theorem. The algorithms for calculating post-test probabilities are presented in Table [1](#T1){ref-type="table"}. A study nurse records clinical data for both groups on the day preceding collection of culture specimens and on the following 6 days (Fig. [3](#F3){ref-type="fig"}). We will collect data on mortality, but this will not be included as a study outcome because of low mortality rates and the intended study size. Further data are collected from the hospital\'s database. This database contains all physician\'s reports, patient baseline data, routine laboratory results, pharmacology data, costs per patient and day of specific medications (e.g. fresh frozen plasma), and staff allocation.
Cytokine measurement
--------------------
Blood samples are collected until 10 a.m. in EDTA-containing vacutainers. Immediately thereafter they are centrifuged at 3000 rpm for 10 min and plasma removed for cytokine analysis. Tracheal aspirate samples are obtained through the endotracheal tube using a sterile suction system (Medinorm AG, Quierschied, Germany). Samples are centrifuged at 10,000 rpm for 5 min and cell free supernatant removed for analysis. Cytokine concentrations (tracheal aspirate and plasma) are simultaneously determined using fluorescent latex beads linked to monoclonal antibodies (R&D Systems, Abington, UK) marked after incubation and coupling with a second phycoerythrin monoclonal antibody (sandwich technique) (R&D Systems, Abington, UK). Final measurement and analysis is done on a Cytomics™ FC 500 Series analyzer (Beckman Coulter Inc., Fullerton, CA, USA).
Pilot study
-----------
During a 16-week period in 2003, we conducted a pilot study in both paediatric ICUs. During this period, cytokine concentrations were available daily for all hospitalized patients so that the new laboratory marker could be implemented as part of routine diagnostic decision making. Clinical data were collected from all hospitalized patients for each day of their ICU stay by one of the investigators (TH). Several teaching sessions both for physicians and nurses were held to enhance the implementation process.
The pilot study revealed that the diagnostic test performance (combined likelihood ratio derived from plasma levels of IL-8 and G-CSF; receiver operating characteristic \[ROC\] 0.88) was similar to that of a published study (ROC 0.85) \[[@B9]\]. However, because clinicians were certain about the presence or absence of infection in half of the episodes, potentially clinically useful test results were found in fewer than a third of all episodes. Thus, we designed the randomized controlled trial as a test to rule-in or rule-out suspected infection only.
Objectives and hypotheses
-------------------------
Our objectives are to achieve a clinically relevant reduction in overall antibiotic use and to reduce treatment costs caused by delayed diagnosis of nosocomial infection. In this study we will test the hypothesis that routine surveillance by determination of cytokine levels in plasma and tracheal aspirates will allow safe discontinuation of antibiotic therapy within 24 hours if the proposed laboratory prediction model indicates absence of infection. We regard a reduction in antibiotic exposure by 15% to be a clinically relevant effect. The second hypothesis we will test is whether early diagnosis reduces the morbidity and costs associated with hospital acquired infection. Ascertaining relevant indicators of morbidity and costs in all patients with culture proven bloodstream infection will operationalize this.
Measures of outcome
-------------------
The primary outcome measure is the rate of systemic antibiotic use per 1000 days of hospitalization (see details under Sample size calculation and statistical considerations). Secondary outcome measures are as follows (for all episodes of hospital acquired infection with positive blood cultures for the first 7 days following initiation of antibiotics after adjusting for important possible patient confounders): number of days free from mechanical ventilation (an indicator of respiratory failure); number of days free from inotropic support (an indicator of circulatory failure); costs for specific expensive medications (e.g. fresh frozen plasma); and nurse allocation (an indicator of treatment intensity).
Sample size calculation and statistical considerations
------------------------------------------------------
At present, in the ICU antibiotic therapy is employed in 40% of patients, which represents a decline from our original survey conducted in 1998 (up to 80% of all patients) \[[@B3]\]. The expected reduction in antibiotic usage is 10--25%, with a clinically relevant reduction considered to be any reduction in excess of 10%. The minimum number of days of hospitalization in each arm required to detect a 10% reduction with a type I error under 5% and a power of 80% is 2300. The expected follow up rate is in excess of 90%. Because the unit of randomization is days and not individuals, an unknown intracluster (intraday) correlation coefficient must be considered. The standard χ^2^statistic, which assumes independence of individuals, may not be applicable. We may be forced to acknowledge the nested nature of the data (clustered randomized controlled trial) by using test statistics based on the generalized linear mixed model \[[@B12]\]. To safeguard against insufficient power we believe that the sample size must be increased to 25%, leading to a required accrual of 3000 hospitalization days per arm. Given the size of the participating units, this translates to a study duration of 24 months.
All analyses will be carried out on an intention-to-treat basis. This means that any antibiotic treatment course will be allocated according to the randomization status of the day on which the decision to withhold or to continue had to be made. This requires us to perform three subgroup analyses: antibiotic prescription prevalence according to the day\'s randomization status; antibiotic free days following the 4 days after any microbiological work up; and antibiotic free days during the week following any initiation of antibiotics.
Stopping rules
--------------
Twelve months after initiating the trial, we will conduct an interim analysis at a two-sided *P*\< 0.01 level. If the results indicate no trend toward a change (increase or reduction) in antibiotic treatment (curtailment from 48 to 24 hours) in prophylactic empirical therapy, and if there is no trend at the *P*\< 0.1 level toward improved secondary outcomes, then the trial will be discontinued. The interim analysis implies that the result of the final analysis should be considered significant if *P*\< 0.04.
Discussion
==========
A variety of publications report excellent diagnostic performance of new markers of infection \[[@B13],[@B14]\]. However, a theoretically useful test may not necessarily provide clinically useful information. Most test accuracy studies derive their results from a subgroup of potentially eligible patients who satisfy unanimously accepted criteria for acceptance as cases or controls. Unfortunately, this practice suffers from the potential overestimation of the test accuracy \[[@B15]\] and, even more importantly, it disregards any clinical information that is available apart from that pertaining to the test under question.
In this randomized controlled trial we wish to assess the clinical efficacy of an innovative diagnostic procedure for the diagnosis of bacterial infection in newborns and critically ill children. It will evaluate whether this strategy results in a clinically relevant reduction in overall antibiotic usage, and whether the strategy is cost-effective by reducing treatment costs caused by delayed diagnosis of nosocomial infection.
One of the possible limitations of the study is the required extended study duration of 24 months. It is conceivable that experience gained from patients in the intervention arm or other factors attributable to the conduct of the study (for example increased awareness by physicians because of more conscious decision making) will also affect the control arm. This might lead to an altered prescription pattern in the control group, which would reduce our ability to find a significant difference between the study arms.
If the new test proves efficacious in clinical practice and is cost-effective, then it may become established as a routine marker of infection in this specific setting.
Author\'s contributions
=======================
JF initiated the project and is the principal investigator. JF, TH, SH, DN and OB participated in the design of the study. JF and TH wrote the protocol. TH carried out the pilot study under supervision of JF. TH implemented the project into clinical routine. JF will carry out statistical analyses. All authors read and approved the final manuscript.
Key messages
============
• Test accuracy should be evaluated prospectively with integrated bedside clinical information.
• The presented design of this ongoing RCT addresses these demands and shall test whether an innovative diagnostic procedure results in a relevant reduction in unnecessary antibiotic utilization and whether this new strategy proves to be cost effective.
Competing interests
===================
The author(s) declare that they have no competing interests.
Abbreviations
=============
G-CSF = granulocyte colony-stimulating factor; ICU = intensive care unit; IL = interleukin; ROC = receiver operating characteristic.
Acknowledgements
================
We thank Adrian Urwyler (Institute of Behavioural Sciences, ETH Zurich) for technical assistance and development of the refined cytokine assay. Our sources of funding include the Chance for the Critically Ill Child Foundation, Zurich, Switzerland (Stiftung Chance für das kritisch kranke Kind) and Bonizzi-Theler Foundation, Lucerne, Switzerland.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Order form for cytokine analysis. Physicians must enter date, time and material for microbiological examination. If antibiotic treatment is started or if a previously ordered treatment is changed, then the reason for this change must be checked in one of the boxes provided. Physicians must indicate their estimate of the likelihood of sepsis and ventilator associated pneumonia on the logarithmic visual-analogue scale. (The final form will be in German.)
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Result form. Results are presented in three ways: raw cytokine concentrations in pg/μl; cytokine concentration derived likelihood ratios for the presence of sepsis or pneumonia; and post-test probabilities of the presence of sepsis or pneumonia. G-CSF, granulocyte colony-stimulating factor; IL, interleukin. (The final form will be in German.)
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Clinical data record form. A trained study nurse collects all relevant clinical data for the day before and until 6 days after collection of blood and/or tracheal aspirate for microbiological examination. ICU, intensive care unit. (The final form will be in German.)
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Equations for calculating post-test probabilities
:::
Algorithm/parameters Equations
----------------------------- -------------------------------------------------------------
Algorithm for sepsis
Pre-test probability~s~ = prevalence of sepsis
Pre-test odds~s~ = pre-test probability~s~/(1 - pre-test probability~s~)
Likelihood ratio~s~ = exp(-8.2 + 0.85 × Ln \[G-CSF~p~\] + 0.7 × Ln \[IL-8~p~\])
Post-test odds~s~ = likelihood ratio~s~× pre-test odds~s~
Post-test probability~s~ = post-test odds~s~/(1 + post-test odds~s~)
Algorithm for pneumonia
Pre-test probability~vap~ = prevalence of pneumonia
Pre-test odds~vap~ = pre-test probability~vap~/(1 - pre-test probability~vap~)
Likelihood ratio~vap~ = exp(-6.8 + 1.0 × Ln \[G-CSF~t~\])
Post-test odds~vap~ = likelihood ratio~vap~× pre-test odds~vap~
Post-test probability~vap~ = post-test odds~vap~/(1 + post-test odds~vap~)
The concentrations of granulocyte colony-stimulating factor (G-CSF) and interleukin (IL)-8 used in the above equations are in pg/μl. Definitions of subscript abbreviations: p, plasma; s, sepsis; t, tracheal aspirate; vap, ventilator associated pneumonia.
:::
|
PubMed Central
|
2024-06-05T03:55:54.863831
|
2004-10-19
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065067/",
"journal": "Crit Care. 2004 Oct 19; 8(6):R443-R450",
"authors": [
{
"first": "Thomas",
"last": "Horisberger"
},
{
"first": "Stephan",
"last": "Harbarth"
},
{
"first": "David",
"last": "Nadal"
},
{
"first": "Oskar",
"last": "Baenziger"
},
{
"first": "Joachim E",
"last": "Fischer"
}
]
}
|
PMC1065068
|
Introduction
============
Bacterial meningitis and ventriculitis remain the most frequent complication in neurosurgery. Diagnosis is based almost exclusively on biochemical and bacteriological analysis of cerebrospinal fluid (CSF) withdrawn either by puncture in the lumbar space or through an external drain located either in the lumbar or ventricular space. It is established that CSF infection is strongly suspected in the presence of a positive CSF culture and/or of a CSF : serum glucose ratio of less than 0.6 and/or of a CSF leukocyte count of more than 11/mm^3^in the lumbar space \[[@B1]\].
Differences in the composition of ventricular and lumbar CSF, based on single pairs of CSF samples, were previously described \[[@B2]-[@B4]\]. These studies showed a rostrocaudal gradient of leukocytes and protein and an inverse gradient of glucose in the first CSF withdrawn in patients with a confirmed diagnosis of meningitis. However, the time course of a ventricular-to-lumbar gradient of leukocytes, glucose and protein, during the occurrence and the relief of meningitis, remains unknown.
Here we describe a patient who developed, after surgery for a chordoma of the clivus, three episodes of recurrent meningitis due to *Pseudomonas aeruginosa*O12. The last two episodes were monitored by daily biochemical and bacteriological analysis of CSF withdrawn in parallel from the lumbar space and ventricles by external lumbar drainage (ELD) and external ventricular drainage (EVD).
Case report
===========
A 20-year-old Caucasian man with no medical history was admitted for elective surgery of a chordoma that extended from the sphenoidal sinus to the anterior face of C2. The first surgical step consisted of a subtotal removal of the tumour by a transfrontal approach. An EVD was inserted at day 1 (D1) because of the appearance of hydrocephalia.
At D10, the second approach consisted in a transoral resection of the tumour with a reconstruction of the pharyngeal wall with skin taken from the arm. However, the wall was not totally occlusive, with a continuous CSF leak into the pharyngeal cavity. Seven days later (D17), the patient developed meningitis with fever and a white blood cell count of 13,800/mm^3^. CSF withdrawn through the ventricular drain showed CSF leukocytes at 830/mm^3^, a CSF protein concentration of 0.99 g/l and a CSF glucose concentration of 3 mmol/l (for a glycaemia of 6 mmol/l). A *Ps. aeruginosa*O12 resistant to almost all antibiotics except ceftazidime and polymyxin B, similar to that repeatedly found in the oral cavity, grew in CSF culture. It was therefore decided to replace the ventricular drain with another in the controlateral hemisphere for two purposes: first, to withdraw CSF to reduce CSF leakage by the fistula, and second, to perform a biochemical and bacteriological analysis. Antibiotherapy was started with intravenous (i.v.) ceftazidime (6 g/day for 2 days, followed by 8 g/day for 25 days) combined with amikacin and polymyxin B both in the ventricles.
A clear improvement in the meningitis allowed us to perform the third and last approach (at D42): an occipito-cervical fixation procedure with EVD removed. Three days later (D45), the patient developed a new episode of hydrocephalia. It was therefore decided to introduce ELD rather than EVD.
Twelve days later (D57), the patient developed a second episode of meningitis: fever, lumbar CSF leukocytes at 14,000/mm^3^, a CSF protein concentration of 1.88 g/l and a glucose concentration of 0.9 mmol/l (for a glycaemia of 6 mmol/l). A CSF culture found the same bacteria as in the first episode of meningitis. This second episode was considered to be related to the persistent pharyngeal fistula. The ELD was replaced with a new one and EVD was added because of the suspicion of an additional obstruction in the 4th ventricle related to post-surgical oedema. Meningitis was treated with an increasing dose of i.v. ciprofloxacin (from D61 to D95: 1.2 g continuously over 24 hours for 4 days, followed by 2.4 g over 24 hours for a further 31 days) and i.v. cefepim (from D61 to D95: 4 × 2 g/day for 4 days to 4 × 3 g/day for a further 31 days; see below for the inhibitory minimal concentration and the plasma and CSF concentrations of antibiotics) and amikacine and polymyxin B colistine both administered directly into the ventricles.
The third episode of meningitis appeared at D66 with identification of the same *Ps. aeruginosa*O12 in CSF culture, increased CSF protein and decreased CSF glucose levels in both ELD and EVD. Antibiotics were kept constant and, despite negative cultures, ELD and EVD were replaced with new drains. Interestingly, since this last episode of meningitis, the pharyngeal fistula disappeared, which indicated the end of pharyngeal contamination of CSF. The patient improved rapidly and was discharged home at D108. No further episode of meningitis during the next 3 years, nor any toxic effect related to the high doses of antibiotics, was observed. It is noteworthy that repetitive cerebral computed tomography scans showed no empyema.
Figure [1](#F1){ref-type="fig"} shows the time course of the following parameters: leukocyte counts, glucose and protein concentrations, measured in parallel in CSF from EVD and ELD, for 17 days (D57 to D73) corresponding to the second and third episodes of meningitis. Figure [1](#F1){ref-type="fig"} shows strikingly that the leukocytes and the protein concentration were always higher and the glucose concentration was always lower in ELD than in EVD. Interestingly, the highest ventriculo-lumbar CSF gradients in leukocytes, protein and glucose concentration were present at the very acute phase of meningitis, when *Ps. aeruginosa*O12 was present in the meningeal cavity.
Our case also showed that the increase in the amount of antibiotics given did increase their concentration in plasma but not in CSF. Indeed, i.v. cefepim was increased from 8 to 12 g/day and i.v. ciprofloxacin from 1.2 to 2.4 g/day from D64 to D95. This induced a persistent increase in plasma cefepim concentration from 46 μ g/ml to more than 60 μg/ml and plasma ciprofloxacin concentration from 0.2 μg/ml to more than 1.0 μg/ml. However, only a transient increase in cefepim concentration (D63, 7 μg/ml; D73, 15 μg/ml; D81 and D95, less than 9 μg/ml) and no increase in ciprofloxacin concentration (0.4--0.5 μg/ml from D63 to 95) were seen in lumbar and ventricular CSF. It is noteworthy that the inhibitory minimal concentrations of cefepim and ciprofloxacin for *Ps. aeruginosa*O12 were 16 and 0.25 mg/ml, respectively.
Discussion
==========
Our case report followed ventriculo-lumbar CSF gradients in leukocytes, protein and glucose concentration during two episodes of post-operative recurrent meningitis due to *Ps. aeruginosa*O12. It showed the presence of a rostrocaudal gradient of leukocytes and protein and an inverse gradient of glucose. This confirmed previous work that showed greater leukocytes and protein concentration in lumbar than in ventricular CSF in patients with a central neural system infection, mostly after neurosurgery \[[@B2],[@B4]\]. However, patients from those studies each had only one pair (ventricular and lumbar) of measurements within a 24-hour interval \[[@B2]\] and glucose concentration in CSF was measured in only six patients \[[@B4]\].
We extend previous studies by showing that the greatest amplitude of ventricular-to-lumbar gradients for all measured parameters (leukocytes, protein and glucose concentration) were seen during the very acute phase of meningitis, when bacteria were present in the meningeal cavity. The mechanisms of such ventricular-to-lumbar gradients are unknown. Our data strongly suggest a compartmentalization of meningeal inflammation in the ventricular and lumbar area. Indeed, similar bacteria, here *Ps. aeroginusa*O12, in similar quantities, seemed to induce a greater alteration of meningeal permeability with greater leukocyte and protein concentrations and a lower glucose concentration in the lumbar than the ventricular CSF region. Although still debatable, the decrease in glucose concentration in CSF seems to be less related to a \'leukocyte-induced glucose consumption\' but rather to a meningeal shift of glucose metabolism to anaerobic glycolysis, as indicated by the concomitant increase in CSF lactate concentration and/or a decrease in meningeal glucose transport \[[@B5]\]; the latter is probably directly related to the degree of meningeal inflammation. An alternative explanation of the existence of a rostrocaudal gradient of leukocytes is that leukocytes from ventricular CSF might fall by gravity to lumbar CSF. However, as explained above, a greater concentration of leukocytes cannot by itself explain a greater protein concentration and a lower glucose concentration in lumbar CSF. Accordingly, our study suggests that meningeal inflammation was greater in the lumbar than the ventricular region in our patient with CSF infection due to a pharyngeal fistula.
Recurrent meningitis led us to increase the antibiotic dosage to achieve a better concentration in CSF \[[@B6]\]. Surprisingly, only a transient increase in CSF cefepim concentration and no change in CSF ciprofloxacine concentration were observed despite a more than 50% increase in plasma concentrations of both antibiotics. The transient increase in cefepim in CSF paralleled that of protein in CSF and could be related to the transient alteration in meningeal permeability.
In summary, this case report shows that the maximal rostrocaudal gradient of leukocytes, protein and glucose was seen in the very acute phase of meningitis. This strongly suggests a greater alteration in the meningeal barrier and very probably a greater meningeal inflammation in the lumbar than the ventricular regions.
Key messages
============
• The paper describes a patient that developed, after surgery for a chordoma of the clivus, three episodes of recurrent meningitis due to *Ps. aeruginosa* O12.
• Episodes were monitored by biochemical and bacteriological daily analysis of CSF withdrawn in parallel from lumbar space and ventricles by external lumbar and ventricular damamge.
• We observed a permanent ventricular-to-lumbar CSF gradients of leukocytes, protein and glucose that increased during the acute phase of meningitis, with the greatest amplitude observed when bacteria was present in both ventricular and lumbar CSF.
• This may suggest a greater extent of meningeal inflammation in lumbar than in ventricular region
Abbreviations
=============
CSF = cerebrospinal fluid; D1, day of insertion of EVD; ELD = external lumbar drainage; EVD = external ventricular drainage; ICU = intensive care unit; i.v. = intravenous.
Competing interests
===================
The author(s) declare that they have no competing interests.
Author\'s contributions
=======================
WN and AM coordinated the data analysis and drafted the manuscript. JM, LR and J-F T participated in bacteriological analysis. A-C L and BG participated in analysis of clinical data. DP helped to draft the manuscript. All authors read and approved the final manuscript
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Time course of the ventricular-to-lumbar gradient of cerebrospinal fluid leukocyte, glucose and protein concentrations in cerebrospinal fluid. The arrows represent days of positive cerebrospinal fluid culture.
:::
:::
|
PubMed Central
|
2024-06-05T03:55:54.865866
|
2004-10-27
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065068/",
"journal": "Crit Care. 2004 Oct 27; 8(6):R491-R494",
"authors": [
{
"first": "Walid",
"last": "Naija"
},
{
"first": "Joaquim",
"last": "Matéo"
},
{
"first": "Laurent",
"last": "Raskine"
},
{
"first": "Jean-François",
"last": "Timsit"
},
{
"first": "Anne-Claire",
"last": "Lukascewicz"
},
{
"first": "Bernard",
"last": "George"
},
{
"first": "Didier",
"last": "Payen"
},
{
"first": "Alexandre",
"last": "Mebazaa"
}
]
}
|
PMC1065070
|
Introduction
============
Acute lung injury (ALI) of septic and non-septic origin is a frequent cause of mortality in critically ill patients. During ALI, the inflammatory process in the lungs may increase the microvascular pressure and permeability, resulting in an accumulation of extravascular lung water (EVLW) and development of pulmonary oedema \[[@B1]\]. However, it is difficult to estimate the amount of oedema fluid at the bedside. Clinical examination, chest radiography and blood gases have proven to be of limited value in quantifying pulmonary oedema \[[@B1]-[@B3]\]. Several techniques to assess EVLW have therefore been developed.
Among the various methods for measurement of EVLW, thermo-dye dilution has been used most frequently \[[@B4]-[@B8]\]. In animal models of lung oedema, this method has been evaluated by comparison with postmortem gravimetry, which is supposed to be the \'gold standard\' of EVLW measurements \[[@B7]-[@B9]\]. In critically ill patients, fluid management guided by thermo-dye measured EVLW was associated with improved clinical outcome \[[@B10]\]. Hence, EVLW has been suggested to play a role as an independent predictor of the prognosis and course of illness \[[@B6],[@B8],[@B10]\]. However, the thermo-dye dilution method is relatively time consuming, cumbersome and expensive. For these reasons, the method has not gained general acceptance \[[@B4],[@B5],[@B7]\].
Use of a technique based on injection of a single thermo-indicator that can be detected using an indwelling arterial catheter was an appealing concept. Recent experimental and clinical studies have shown that EVLW assessed by single thermodilution (ST) exhibits good reproducibility and close agreement with the thermo-dye double indicator technique \[[@B11],[@B12]\]. The ST method is simpler to apply, less invasive and more cost effective; all of these factors make it more suitable for use at the bedside. However, to date, this new method has been sparsely evaluated against gravimetry \[[@B13],[@B14]\], and further validation is needed.
Thus, the aim of the present study was to evaluate the accuracy of the ST technique by comparing it with that of postmortem gravimetry (EVLW~G~) in conscious sheep, in which ALI was induced either by lipopolysaccharide (LPS) or by oleic acid (OA). Both of these models of ALI are reproducible and have been extensively described \[[@B7],[@B9],[@B11],[@B15],[@B16]\].
Methods
=======
Surgical preparation and measurements
-------------------------------------
The study was approved by the Norwegian Experimental Animal Board and conducted in compliance with the European Convention on Animal Care. Eighteen yearling sheep weighing 27.5 ± 0.4 kg were instrumented, as a modification to previously described techniques \[[@B16]-[@B19]\], by inserting introducers into the left external jugular vein and common carotid artery. After 1--4 days of recovery, sheep were placed in an experimental pen. A thermodilution catheter (131HF7; Edwards Life Sciences, Irvine, CA, USA) was introduced into the pulmonary artery and a 4-Fr thermistor-tipped catheter (PV2014L16; Pulsion Medical Systems, Munich, Germany) into the carotid artery. The catheters were connected to pressure transducers (Transpac^®^III \[Abbott, North Chicago, IL, USA\] and PV8115 \[Pulsion Medical Systems\], respectively).
Mean pulmonary arterial pressure (PAP), pulmonary arterial occlusion pressure (PAOP) and right atrial pressure (RAP) were displayed on a 565A Patient Data Monitor (Kone, Espoo, Finland) and recorded on a Gould Polygraph (Gould Instruments, Cleveland, OH, USA). Heart rate, mean systemic arterial pressure, cardiac index (CI), systemic vascular resistance index, extravascular lung water index (EVLWI) assessed using the single thermodilution technique (EVLWI~ST~), pulmonary vascular permeability index (PVPI), global end-diastolic volume (GEDV) index (GEDVI), intrathoracic blood volume (ITBV) index (ITBVI) and blood temperature were determined at 1-hour intervals using a PiCCO*plus*monitor (Pulsion Medical Systems). Every value reported here is the mean of three consecutive measurements, each consisting of a 10 ml bolus of ice-cold 5% dextrose injected into the right atrium randomly during the respiratory cycle.
To estimate EVLW we used the following formula \[[@B12]\]: EVLW~ST~(ml) = ITTV - ITBV (where ITTV is the intrathoracic thermal volume). During clinical application of ST by means of the PiCCO monitor, ITBV is calculated as 1.25 × GEDV, the coefficient 1.25 being derived from critically ill patients \[[@B12]\]. However, in our previous investigations in sheep \[[@B17]-[@B19]\], in which ITBV was measured directly using the thermal-dye dilution technique, we found the coefficient to be 1.34 \[[@B14]\]. Thus, in the present study we used the corrected values of ITBVI, EVLWI~ST~and PVPI, based on the following equation: ITBVI = 1.34 × GEDVI.
Blood samples were drawn from the systemic arterial (a) and pulmonary arterial (v) lines and analyzed every two hours for blood gases and haemoglobin (Rapid 860; Chiron Diagnostics Corporation, East Walpole, MA, USA). The pulmonary vascular resistance index (PVRI), venous admixture (Qs/Qt), oxygen delivery index (DO~2~I) and oxygen consumption index were calculated as described previously \[[@B16],[@B19],[@B20]\].
Experimental protocol
---------------------
After establishing a stable baseline at time 0 hours, awake and spontaneously breathing sheep were randomly assigned to three experimental groups: a sham operated group (*n*= 4); a LPS group (*n*= 7), receiving an intravenous infusion of *Escherichia coli*O26:B6 LPS (Sigma Chemical, St. Louis, MO, USA) at 15 ng/kg per min for 6 hours; and an OA group (*n*= 7), in which sheep were subjected to an intravenous infusion of OA (Sigma Chemical) 0.06 ml/kg mixed with the animal\'s blood. The duration of the infusion of OA was 30 min.
During the experiment, all animals received a continuous infusion (5 ml/kg per hour) of Ringer\'s lactate, aiming to maintain intravascular volume at baseline levels. After the last measurements, at 2 hours in the OA group and at 6 hours in the sham-operated and the LPS groups, the sheep were anaesthetized and killed with a lethal dose of potassium chloride. Then, postmortem EVLWI (EVLWI~G~) was determined by gravimetry, as previously described \[[@B21]-[@B24]\].
Statistical analysis
--------------------
For each continuous variable, normality was checked using the Kholmogorov-Smirnov test. Data are expressed as mean ± standard error of the mean, and assessed by analysis of variance followed by Scheffe\'s test or test of contrasts, when appropriate. To evaluate the relationship between EVLWI~ST~and EVLWI~G~, we used linear regression and Bland-Altman analysis. *P*\< 0.05 was considered statistically significant.
Results
=======
All animals survived until the end of the experiments. At baseline no significant differences were found between groups, as shown in Figs [1](#F1){ref-type="fig"} and [2](#F2){ref-type="fig"}, and Tables [1](#T1){ref-type="table"} and [2](#T2){ref-type="table"}. In the sham-operated sheep, all variables remained unchanged throughout the study.
Haemodynamic and extravascular lung water measurements
------------------------------------------------------
Figure [1](#F1){ref-type="fig"} and Table [1](#T1){ref-type="table"} show that LPS and OA induced marked increments in PAP and PVRI, peaking at 1 hour and subsequently decreasing gradually to values significantly above the respective baselines and the corresponding values in the sham-operated group. PAOP and RAP also rose in both the LPS and the OA groups (*P*\< 0.05; data not shown). In parallel, LPS increased EVLWI~ST~transiently by 20--35% (*P*\< 0.05; Fig. [1](#F1){ref-type="fig"}). After OA administration, EVLWI~ST~rose to a maximum of 84% above baseline (*P*\< 0.01). At the end of the experiment, EVLWI~ST~in the OA group had increased by 6.4 ml/kg and 9.3 ml/kg relative to the LPS and the sham-operated groups, corresponding to increments of 54% and 104%, respectively (*P*\< 0.05). PVPI increased by 40% after LPS administration and by 90% after OA (*P*\< 0.05; Fig. [1](#F1){ref-type="fig"}). GEDVI and ITBVI varied within 10--15% of baseline with no intergroup differences. As shown in Table [1](#T1){ref-type="table"}, LPS caused tachycardia and a rise in CI accompanied by a slight increase in mean arterial pressure whereas systemic vascular resistance index decreased (*P*\< 0.05). In contrast, in the OA group CI declined and systemic vascular resistance index increased relative to baseline (*P*\< 0.05).
Oxygenation and gas exchange
----------------------------
LPS caused significant increments in mixed venous oxygen saturation, DO~2~I and Qs/Qt (Fig. [2](#F2){ref-type="fig"}). OA decreased both arterial and venous oxygenation and reduced DO~2~I (*P*\< 0.05). Oxygen consumption index did not change significantly (not shown). LPS caused a transient reduction in arterial carbon dioxide tension and a rise in pH (*P*\< 0.05; Table [2](#T2){ref-type="table"}). After OA, pH decreased (*P*= 0.04). The haemoglobin concentration as well as the body temperature rose only in the LPS group (*P*\< 0.05).
Linear regression and Bland-Altman analysis
-------------------------------------------
As shown in Fig. [3](#F3){ref-type="fig"}, the regression analysis between EVLW~ST~and postmortem EVLW~G~yielded the following relation: EVLWI~ST~= 1.30 × EVLW~G~+ 2.32 (*n*= 18, *r*= 0.85, *P*\< 0.0001). Notably, the mean EVLWI~ST~at the end of experiment was higher than EVLWI~G~: 13.6 ± 1.1 ml/kg versus 8.7 ± 0.7 ml/kg (*P*= 0.0005). Ranges of EVLWI~ST~and EVLWI~G~values were 7.5--21.0 ml/kg and 4.9--14.5 ml/kg. According to the Bland-Altman analysis, the mean difference between EVLWI~ST~and EVLWI~G~was 4.91 ml/kg, with upper and lower limits of agreement (± 2 standard deviations) of +9.99 ml/kg and -0.17 ml/kg, respectively (Fig. [4](#F4){ref-type="fig"}). The difference between methods increased with increasing values of mean EVLWI (*n*= 18, *r*= 0.64; *P*= 0.005); the regression line equation was as follows: EVLWI~ST~- EVLWI~G~= 0.89 × (\[EVLWI~ST~+ EVLWI~G~\]/2) + 6.82.
Postmortem gravimetry
---------------------
As shown in Fig. [5](#F5){ref-type="fig"}, EVLWI~G~in the OA group increased by 4.7 ml/kg and 5.6 ml/kg relative to the LPS and the sham-operated groups, amounting to increments by 65% and 90%, respectively (*P*= 0.001).
Discussion
==========
The present findings confirm that, in sheep, EVLW measured using the single transpulmonary thermodilution technique correlates closely with EVLW determined using postmortem gravimetry. However, EVLWI~ST~overestimates EVLWI~G~, with the degree of overestimation increasing with the severity of ALI.
A number of experimental and clinical studies focused on the potential role of EVLW as a guide to diagnosis and treatment of critically ill patients \[[@B3],[@B6]-[@B14],[@B25],[@B26]\]. During pulmonary oedema, accumulation of EVLW occurs before any changes take place in blood gases, chest radiogram and, ultimately, pressure variables. In addition, the latter variables are nonspecific diagnostic tools that are influenced by a variety of factors \[[@B2],[@B4],[@B5],[@B8]\]. Thus, Boussat and coworkers \[[@B3]\] recently demonstrated that, in sepsis induced ALI, commonly used filling pressures such as PAOP and RAP are poor indicators of pulmonary oedema. Rather than those measures, they recommended direct measurement of EVLW. Consistent with this, we found that EVLW, in contrast to RAP, correlates with markers of lung injury in human septic shock \[[@B26]\]. Victims of ALI, regardless of pathogenesis, have a significantly higher EVLW than do other patients \[[@B6],[@B26]\]. Hence, measurement of EVLW supports the diagnosis and may even improve clinical outcomes when used cautiously in combination with treatment protocols that are known to hasten the resolution of pulmonary oedema \[[@B10],[@B25]\].
Instrumented awake sheep represents a stable experimental model for measuring cardiopulmonary variables, as demonstrated in the sham-operated group in the present study as well as by other investigators \[[@B15],[@B27]\]. The model can be used to assess different interventions during ALI.
Consistent with previous investigators \[[@B15],[@B17],[@B27]\], we observed that infusion of LPS and OA caused pulmonary hypertension, increased EVLW and impaired gas exchange. Despite increments in PAP, PAOP and PVRI, both ITBV and GEDV remained constant whereas PVPI (an index of microvascular permeability, calculated as the ratio of EVLW to pulmonary blood volume) increased significantly. Thus, the haemodynamic responses to LPS and OA are not purely hydrostatic but may also manifest as noncardiogenic permeability pulmonary oedema \[[@B13],[@B15]-[@B18],[@B27],[@B28]\].
In the present study lung oedema was significantly more severe in the OA group than in the LPS group, which is consistent with the findings of other investigators \[[@B29]\]. In fact, OA causes acute haemorrhagic alveolitis, which may lead to acute endothelial and alveolar necrosis and a severe proteinaceous oedema \[[@B30]\]. In contrast, the LPS-induced ALI is initiated by accumulation of granulocytes and lymphocytes in the pulmonary microcirculation that results in more moderate damage to endothelial cells and lung oedema \[[@B31]\].
Lung injury in the LPS group was accompanied by a hyperdynamic circulatory state, which was manifested by systemic vasodilation and increments in CI and DO~2~I toward the end of the experiment. In contrast, in the OA group we observed cardiac depression and systemic vasoconstriction. This is consistent with previous investigations of LPS and OA \[[@B18],[@B27],[@B30],[@B32]\]. Thus, ovine models exhibit a scatter of cardiopulmonary changes from normal in the sham-operated group to mild or moderate ALI in endotoxaemic sheep and moderate to severe ALI in animals subjected to OA.
The significant correlation of EVLWI~ST~and EVLWI~G~observed in the present study is consistent with findings of Katzenelson and coworkers \[[@B13]\], who validated EVLWI~ST~versus postmortem gravimetry in dogs \[[@B13]\]. However, those investigators did not specifically assess the relationship between EVLWI~ST~and EVLWI~G~in sepsis-induced ALI. In addition, their study was performed in anaesthetized and mechanically ventilated animals; hence, further investigation of the correlation in a conscious state was required. Recently, ST has been evaluated against the thermo-dye dilution method in both experimental and clinical settings \[[@B11],[@B12]\]. The studies revealed a close agreement between the techniques. Thus, we believe that injection of cold saline can provide valuable information about the EVLW content and the severity of pulmonary oedema.
During ALI, both ST and postmortem gravimetry demonstrated similar relative increases in EVLWI as compared with sham-operated animals. However, we noticed that ST overestimates the absolute values of EVLWI compared with the gravimetric technique -- a discrepancy that increased with progression of pulmonary oedema. This finding could be accounted for by heat exchange of the thermal indicator with extravascular intrathoracic structures, such as the walls of the large vessels and the myocardium, and by recirculation of the indicator \[[@B8]\]. In addition, the coefficients for calculation of EVLWI~ST~and ITBV may vary with weight and age, as well as between animal species \[[@B11]\]. Consequently, in the experimental setting EVLWI~ST~requires a specific correction. In the present study we replaced the coefficient 1.25 used in humans in the ITBVI equation (i.e. ITBVI = 1.25 × GEDVI) with the recalculated \'ovine\' coefficient 1.34 \[[@B14]\], which is based on 426 measurements in 48 animals \[[@B17]-[@B19]\].
In contrast to ST, the thermo-dye dilution technique runs the risk of underestimating EVLW in comparison with gravimetry \[[@B4]\]. This underestimation increases during ALI caused by instillation of hydrochloric acid into the airways, and has been explained by redistribution of pulmonary blood flow away from the oedematous areas. The redistribution is thought to prevent indicator diffusion and consequently to prevent detection of oedema \[[@B7]\]. In addition, detection of EVLW by thermo-dye dilution can be impaired by changes in CI as well as by positive end-expiratory pressure during mechanical ventilation \[[@B8],[@B28]\].
Compared with other techniques for assessment of EVLW, ST may underestimate EVLW during pulmonary oedema due to intratracheal instillation of saline, although it is an accurate method in normal lungs \[[@B33]\]. However, intratracheal instillation of saline can also be criticized because a proportion of the fluid is rapidly absorbed and obscured from detection \[[@B34]\].
Notably, the use of postmortem gravimetry as the reference method for evaluating pulmonary oedema also has limitations \[[@B21],[@B33]\]. For example, the method only allows one measurement and is therefore of no use in following variations over time. The application of gravimetry is limited almost exclusively to experimental studies. The comparison of gravimetric measurement with results of other techniques for determination of EVLW can be influenced by the duration from death to removal of the lungs and by pathophysiological changes in the lungs after cardiac arrest. Thus, the gravimetric technique can underestimate the real value of EVLWI because of partial reabsorption of fluid before excision of the lungs.
Conclusion
==========
The determination of EVLW by ST in sheep correlates closely with gravimetric measurements over a wide range of changes, and thus it may potentially be of benefit in quantifying lung oedema in critically ill patients. However, compared with postmortem gravimetry, single transpulmonary thermodilution overestimates the absolute values of EVLW. Thus, further studies are warranted to evaluate the accuracy of this method for managing ALI in humans.
Key messages
============
• In sheep, extravascular lung water assessed by transpulmonary single thermodilution correlates closely with gravimetric measurements over a wide range of changes.
• Despite a moderate overestimation of the extravascular lung water content compared with post-mortem gravimetry, single thermodilution can be a useful tool for assessment of pulmonary oedema during ALI.
Competing interests
===================
This study was supported by Helse Nord (Norway), project number 4001.721.132; departmental funds, the Department of Anesthesiology, University Hospital of North Norway; and Pulsion Medical Systems (Germany).
Author contributions
====================
MYK participated in the design of study, performed statistical analysis, and drafted the manuscript. VVK participated in the design of study, performed statistical analysis, and prepared the figures. VVK and KW participated in the design of study. LJB participated in the design of study and provided coordination. All authors read and approved the final manuscript.
Abbreviations
=============
ALI = acute lung injury; CI = cardiac index; DO~2~I = oxygen delivery index; EVLW = extravascular lung water; EVLWI = extravascular lung water index; GEDV = global end-diastolic volume; GEDVI = global end-diastolic volume index; ITBV = intrathoracic blood volume; ITBVI = intrathoracic blood volume index; LPS = lipopolysaccharide; OA = oleic acid; PAOP = pulmonary arterial occlusion pressure; PAP = pulmonary arterial pressure; PVPI = pulmonary vascular permeability index; PVRI = pulmonary vascular resistance index; Qs/Qt = venous admixture; RAP = right atrial pressure; ST = single thermodilution.
Acknowledgements
================
The authors are grateful to Professor Anton Hauge for critical review of the manuscript and Mrs Alexandra Saab Bjertnaes, MBA, for linguistic advice.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Changes in pulmonary haemodynamics and extravascular lung water in sheep. Data are expressed as mean ± standard error of the mean. \**P*\< 0.05, LPS versus sham-operated group; ^†^*P*\< 0.05, OA versus sham-operated group; ^‡^*P*\< 0.05, LPS versus OA group; ^§^*P*\< 0.05, versus t = 0 hours in LPS group; ^ll^*P*\< 0.05 versus t = 0 hours in OA group. EVLWI~ST~= extravascular lung water index measured by single thermodilution; LPS = lipopolysaccharide; OA = oleic acid; PAP = pulmonary arterial pressure; PVPI = pulmonary vascular permeability index; Sham = sham-operated group.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Changes in oxygenation variables in sheep. Data are expressed as mean ± standard error of the mean. \**P*\< 0.05, LPS versus sham-operated group; ^†^*P*\< 0.05, OA versus sham-operated group; ^‡^*P*\< 0.05, LPS versus OA group; ^§^*P*\< 0.05, versus t = 0 hours in LPS group; ^ll^*P*\< 0.05, versus t = 0 hours in OA group. DO~2~I = oxygen delivery index; LPS = lipopolysaccharide; OA = oleic acid; Qs/Qt = venous admixture; SaO~2~= arterial oxygen saturation; Sham = sham-operated; SvO~2~= venous oxygen saturation.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Linear regression analysis between extravascular lung water index (EVLWI) as determined by transpulmonary single thermodilution (EVLWI~ST~) and postmortem gravimetry (EVLWI~G~) in sheep. EVLWI~ST~= 1.30 × EVLW~G~+ 2.32 (*n*= 18, *r*= 0.85, *P*\< 0.0001). Line of identity is dashed; 95% confidence intervals are indicated by solid lines. LPS, lipopolysaccharide; OA, oleic acid; Sham, sham-operated.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Bland-Altman plot for the extravascular lung water index (EVLWI) measured using transpulmonary single thermodilution (EVLWI~ST~) and postmortem gravimetry (EVLWI~G~) in sheep. The x-axis shows the mean of EVLWI measurements by single thermodilution and gravimetry. The y-axis shows the difference between the methods. The bold line indicates the value for the mean difference between EVLWI~ST~and EVLWI~G~(bias), and each dashed line indicates two standard deviations (SDs). Mean difference EVLWI~ST~- EVLWI~G~= 4.91 ml/kg (SD 2.54 ml/kg).
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Gravimetric extravascular lung water index (EVLWI~G~) in sheep. Data are expressed as mean ± standard error of the mean. ^†^*P*\< 0.05, OA versus sham-operated group; ^‡^*P*\< 0.05, LPS versus OA group. LPS = lipopolysaccharide; OA = oleic acid; Sham = sham-operated group.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Haemodynamics during acute lung injury in sheep
:::
Parameter Group Time point (hours)
----------------------------- ------- -------------------- ------------------ --------------- -------------- --------------- --------------- ----------------
PVRI (dyne·s/cm^5^per m^2^) Sham 117 ± 14 131 ± 19 141 ± 10 145 ± 11 114 ± 21 133 ± 13 151 ± 16
LPS 115 ± 6 284 ± 20\*^†^ 240 ± 21\*^†^ 198 ± 31^†^ 193 ± 26\*^†^ 199 ± 23\*^†^ 182 ± 16\*^†^
OA 103 ± 9 351 ± 64^‡§^ 300 ± 45^‡§^ \- \- \- \-
GEDVI (ml/m^2^) Sham 570 ± 46 601 ± 68 572 ± 43 566 ± 12 661 ± 74 607 ± 67 655 ± 60
LPS 571 ± 23 620 ± 57 564 ± 32 579 ± 42 598 ± 38 624 ± 42 615 ± 37
OA 646 ± 38 629 ± 60 590 ± 55 \- \- \- \-
ITBVI (ml/m^2^) Sham 764 ± 62 806 ± 91 766 ± 58 759 ± 16 886 ± 99 813 ± 90 878 ± 81
LPS 765 ± 30 831 ± 76 756 ± 43 776 ± 57 801 ± 51 836 ± 57 825 ± 49
OA 866 ± 51 912 ± 42 790 ± 74 \- \- \- \-
HR (beats/min) Sham 106 ± 6 104 ± 8 96 ± 7 91 ± 5 99 ± 6 98 ± 11 97 ± 5
LPS 96 ± 4 122 ± 6^†^ 109 ± 6 109 ± 4\*^†^ 109 ± 8 122 ± 4\*^†^ 130 ± 5\*^†^
OA 111 ± 5 104 ± 13 102 ± 13 \- \- \- \-
CI (l/min per m^2^) Sham 5.7 ± 0.3 5.5 ± 0.3 5.2 ± 0.3 5.1 ± 0.2 5.4 ± 0.4 5.2 ± 0.3 5.3 ± 0.3
LPS 5.7 ± 0.1 7.3 ± 0.5\*^†ll^ 5.9 ± 0.2 5.8 ± 0.3 5.6 ± 0.4 6.2 ± 0.3\* 6.8 ± 0.2\*^†^
OA 6.1 ± 0.3 4.5 ± 0.3^§^ 4.6 ± 0.5 \- \- \- \-
MAP (mmHg) Sham 102 ± 5 101 ± 6 101 ± 6 101 ± 5 102 ± 5 102 ± 5 101 ± 4
LPS 94 ± 4 104 ± 5^†^ 105 ± 3^†^ 106 ± 4^†^ 105 ± 3^†^ 104 ± 5 100 ± 6
OA 94 ± 4 101 ± 2 104 ± 4^§^ \- \- \- \-
SVRI (dyne·s/cm^5^per m^2^) Sham 1453 ± 101 1496 ± 125 1589 ± 138 1536 ± 105 1579 ± 109 1607 ± 119 1681 ± 169
LPS 1410 ± 81 1128 ± 102^†^ 1352 ± 62 1515 ± 107 1428 ± 101 1308 ± 79 1126 ± 58\*^†^
OA 1266 ± 63 1847 ± 368 1670 ± 147^§^ \- \- \- \-
Data are expressed as mean ± standard error of the mean. \**P*\< 0.05, LPS versus sham-operated group; ^†^*P*\< 0.05, versus t = 0 hours in LPS group; ^‡^*P*\< 0.05, OA versus sham-operated group; ^§^*P*\< 0.05, versus t = 0 hours in OA group; ^ll^*P*\< 0.05, LPS versus OA group. CI, cardiac index; GEDVI, global end-diastolic volume index; HR, heart rate; ITBVI, intrathoracic blood volume index; LPS, lipopolysaccharide; MAP, mean arterial pressure; OA, oleic acid; PVRI, pulmonary vascular resistance index; Sham, sham-operated; SVRI, systemic vascular resistance index.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Gas exchange during acute lung injury in sheep
:::
Parameter Group Time point (hours)
------------------------ ------- -------------------- ----------------- ----------------- -----------------
pHa Sham 7.52 ± 0.03 7.53 ± 0.02 7.50 ± 0.02 7.50 ± 0.02
LPS 7.48 ± 0.01 7.50 ± 0.02 7.55 ± 0.02\* 7.53 ± 0.02
OA 7.50 ± 0.01 7.44 ± 0.03^†^ \- \-
PaCO~2~(mmHg) Sham 38.7 ± 2.7 36.4 ± 2.1 36.4 ± 1.6 37.4 ± 1.0
LPS 39.7 ± 1.4 38.8 ± 1.5 32.6 ± 0.6\*^‡^ 33.1 ± 1.0\*^‡^
OA 36.4 ± 1.1 42.5 ± 3.7 \- \-
Haemoglobin (g/dl) Sham 10.7 ± 0.9 10.1 ± 0.6 10.2 ± 0.5 10.3 ± 0.5
LPS 10.4 ± 0.6 10.4 ± 0.6 11.0 ± 0.7\* 10.9 ± 0.6
OA 10.3 ± 0.4 10.7 ± 0.5 \- \-
Blood temperature (°C) Sham 39.3 ± 0.1 39.2 ± 0.1 39.3 ± 0.1 39.3 ± 0.1
LPS 39.3 ± 0.1 40.0 ± 0.1\*^‡^ 41.3 ± 0.1\*^‡^ 41.0 ± 0.1\*^‡^
OA 39.5 ± 0.1 39.6 ± 0.1 \- \-
Data are expressed as means ± standard error of the mean. \**P*\< 0.05, versus t = 0 hours in LPS group; ^†^*P*\< 0.05, OA versus sham operated group; ^‡^*P*\< 0.05, LPS versus sham-operated group. LPS, lipopolysaccharide; OA, oleic acid; PaCO~2~, arterial carbon dioxide tension; Sham, sham-operated.
:::
|
PubMed Central
|
2024-06-05T03:55:54.867279
|
2004-10-19
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065070/",
"journal": "Crit Care. 2004 Oct 19; 8(6):R451-R458",
"authors": [
{
"first": "Mikhail Y",
"last": "Kirov"
},
{
"first": "Vsevolod V",
"last": "Kuzkov"
},
{
"first": "Vladimir N",
"last": "Kuklin"
},
{
"first": "Kristine",
"last": "Waerhaug"
},
{
"first": "Lars J",
"last": "Bjertnaes"
}
]
}
|
PMC1065072
|
See related commentary <http://ccforum.com/content/8/6/433>
Introduction
============
Being admitted to an intensive care unit (ICU) can be considered a stressful life event, the reason for admission being a critical or even life-threatening condition. The ICU stay itself may also be stressful. Some patients report vivid recollections \[[@B1]-[@B3]\] whereas others have a poor or even no recollection at all of their stay on the ICU. No recollection at all of the ICU study ranges from 23% to 38% among postsurgical patients \[[@B4]\]. Various authors have reported that patients had unpleasant recollections after a stay on an ICU. Patients recalled discomfort arising from anxiety, pain, thirst, sleeplessness, disorientation, shortness of breath, inability to move, painful medical interventions and the presence of an endotracheal tube \[[@B5]\]. Turner and coworkers \[[@B6]\] specifically mentioned arterial blood gas sampling and endotracheal suctioning.
However, recollection of discomfort during the ICU stay is inseparably connected to the quality of recollection itself: events considered stressful at the time may not be remembered; conversely, recollections of stressful events may not be based on actual experiences. Jones and coworkers \[[@B7]\] investigated patients\' estimation of the duration of their ICU stay in order to evaluate the accuracy of their memories. The patients\' recall of events was generally poor, and 41% of them felt that they had been confused at some time during their stay in the ICU. To our knowledge, there is no literature investigating whether the recollection of discomfort is related to the accuracy of recollection of facts as such, and for what sources of discomfort this holds true. The purpose of this study was to describe the incidence of discomfort reported by ICU patients, the sources of their discomfort, the factual recollection of ICU patients and ward patients, and determinants of the recollection of discomfort in ICU patients.
Methods
=======
Consecutive ICU patients, who were older than 18 years and who had undergone intubation for longer than 24 hours, were included in the study. During mechanical ventilation patients received sedation by continuous infusion of midazolam (range 1--4 mg/hour) and fentanyl (range 50--150 µg/hour), with the degree of sedation given depending on their clinical requirements. The patients participated in a study comparing routine endotracheal suctioning with minimally invasive airway suctioning. The study was approved by the medical ethics committee of the University Hospital. The Acute Physiology Age and Chronic Health Evaluation (APACHE) II score was used to quantify the severity of illness \[[@B8]\] and was recorded on the day of admission to the ICU.
All ICU patients participated in a structured in-person interview, using a standardized questionnaire, within three days after discharge from the ICU to the ward. The reference group consisted of postsurgical ward patients, matched for age and sex. Data from the reference group were obtained in a structured telephone interview conducted within three days after discharge from hospital. In the questionnaire, all patients were asked to give answers to 14 questions concerning the ICU environment (lighting, timing of ward rounds, number of fellow ICU patients), the nursing staff (uniform, male/female) and personal care (clothing, position of intravenous drip, washing and toilet activities).
Patients from the ICU group were asked whether they remembered any discomfort during their stay on the ICU, and if they did then they were asked to specify the sources of discomfort that they remembered.
The questions regarding recollection of facts were first asked as open questions. Two points were given for each correct answer to these open questions. Patients who were unable to answer the open questions were presented with four multiple choice answers. One point was given for each correct answer to the multiple choice questions. Summation of the points resulted in a total score for factual recollection, providing an indication of the level of factual recollection. The range for the total score was 0--28 points.
Statistical analysis
--------------------
SPSS version 10 (SPSS Inc., Chicago, IL, USA) was used to perform all analyses. To assess the reliability of the questionnaire, a Cronbach\'s alpha was calculated. Differences between the ICU group and the reference group were analyzed using the ?^2^test for categorical variables and the t-test for normally distributed intervals or ratio scale variables. Differences between patients who recalled discomfort and those who recalled no discomfort were analyzed using the ?^2^test in case of categorical variables, the Mann--Whitney test for ordinal variables and the t-test for normally distributed intervals or ratio scale variables such as age. To analyze potential determinants of discomfort, logistic regression was performed. The presence or absence of discomfort was entered as the dependent variable, and independent variables were as follows: age, sex, APACHE II score (only in ICU patients), length of stay in the ICU or ward, factual recollection score and duration of tracheal intubation. Correlation coefficients between factual recollection score and age were calculated using a Spearman\'s test for categorical variables.
From the logistic regression analysis, odds ratios (ORs) were calculated for all independent variables in the equation. The OR expresses the odds in the group with the condition relative to the other group without the condition. To an extent, the OR can be considered a measure of relative risk. An OR greater than 1 indicates a higher risk and an OR below 1 indicates a lower risk in the group with the condition relative to the group without the condition.
Results
=======
A total of 125 patients discharged from the ICU were included in this study. Two patients were unable to respond to the questions. Patient characteristics are summarized in Table [1](#T1){ref-type="table"}. In the population studied the prevalence of any discomfort recalled after discharge from the ICU was 54% (*n*= 66). The sources of discomfort identified by these 66 patients are summarized in Table [2](#T2){ref-type="table"}. Six patients were disorientated at the time of the interview, but were able to recall discomfort.
The median (min--max) factual recollection score was 15 (0--28) in the ICU patients and 25 (19--28) in the reference group; the difference between the groups was highly significant (*P*\< 0.001). Analyses of reliability of the questionnaire for the ICU patients revealed a Cronbach\'s alpha of 0.86, indicating high reliability. Items of factual recollection by ICU patients and the reference group, in descending order of being identified correctly, are listed in Table [3](#T3){ref-type="table"}.
ICU patient characteristics are summarized in Table [4](#T4){ref-type="table"} separately for the group that recalled any discomfort and the group that did not recall any discomfort. Significant differences were found between the two groups in factual recollection, age and duration of intubation.
Logistic regression analysis of determinants of recollection of discomfort confirmed that factual recollection was indeed an independent factor in predicting recollection of discomfort. The calculated OR was 1.1 (*P*\< 0.001), with a correct percentage in regression analysis of 68%.
This implies that the risk for recalling discomfort was 1.1 times higher for each factual recollection point. Age also was a determinant of recollection of discomfort. The calculated OR was 0.97 (*P*= 0.006; correct percentage in regression analysis 66%). This implies that the risk for recalling discomfort was lower by a factor of 0.97 for each year of advancing age. The duration of intubation appeared not to be independently related to recollection of discomfort.
Factual recollection appears to be inversely related to age. Analysis of the relationship between factual recollection score and age in the ICU group revealed that the correlation coefficient was -0.352 (*P*\< 0.001); in the reference group it was -0.327 (*P*= 0.023; Fig. [1](#F1){ref-type="fig"}).
Finally, recollection of pain appeared to be related to age (OR 0.936, *P*= 0.002; correct percentage in regression analysis 94%). This implies that younger patients reported more recollection of discomfort in the form of pain.
Discussion
==========
The results of the present study show that a considerable proportion (54%) of patients discharged from the ICU had a recollection of discomfort during their stay in the ICU. The presence of an endotracheal tube, medical interventions, noise and experiences of hallucination were among the sources of discomfort most frequently reported. To our knowledge, this study is the first to evaluate the association between recollection of discomfort and intact factual recollection. In a study conducted by Rose and coworkers \[[@B9]\] in 50 patients, 60% remembered endotracheal suctioning and 52% remembered extubation as unpleasant experiences. In a study by Turner and coworkers \[[@B6]\], arterial blood gas sampling and tracheal suctioning were recalled by 48% and 44% of the patients. Although those two studies did not investigate the prevalence of discomfort *per se*, we conclude that their findings are similar to ours, in that discomfort was recalled by 54% of ICU patients.
Within the context of ICU patients\' recollections, a memory of an (stressful) event raises the question of whether this recollection is based on reality or fantasy/imagination. In the present study we found the degree of factual recollection to be an important determinant of discomfort, in the sense that more discomfort was reported by those with better factual recollection. Each item of factual recollection that was scored correctly increased slightly the risk for recollection of discomfort. Factual recollection and recollection of discomfort therefore appeared to be related.
In an ICU many factors contribute to impairment in memory: critical illness itself, the use of benzodiazepines and opioids, and the common occurrence of delirious states. When a patient\'s health is improving or when sedative agents are reduced below effective levels, patients tend to remember more regarding factors, mostly unpleasant, in the ICU. Jones and coworkers \[[@B10]\] described many causes of amnesia during severe illness, including large dosages of sedative medication and withdrawal syndromes. Because levels of sedation strongly influence the function of memory, a weak point in our study is that no sedation score was recorded to enable us to evaluate the effects of sedatives on patient recollection. It should also be noted that we did not look for objective signs of postdischarge psychological distress or examine their relationship to memories of stressful events, either real or perceived. We merely wished to improve our understanding of discomfort by taking into account the confounding role of memory.
The presence of an endotracheal tube, medical activities, and noise and bustle were the sources of discomfort remembered most frequently (Table [2](#T2){ref-type="table"}). This finding is comparable with those of other studies. In a group of 68 ventilated medical patients, Turner and coworkers \[[@B6]\] found a prevalence of recollection of endotracheal suctioning of 44%, and in 26 mainly surgical patients those investigators found a prevalence of recollection of endotracheal suctioning of 47% \[[@B11]\]. In a mixed surgical/medical group of cardiac patients (*n*= 50), Rose and colleagues \[[@B9]\] found a 60% prevalence of recollection of endotracheal suctioning during the ICU stay.
The reason for discomfort relating to the endotracheal tube may be endotracheal suctioning. While intubated, patients are regularly suctioned via the endotracheal tube in order to maintain airway patency. The strong mechanical stimuli resulting from endotracheal suctioning may explain why the endotracheal tube is remembered as a prominent source of discomfort. In a previous study \[[@B12]\], we investigated recollection of endotracheal suctioning with two methods of suctioning: routine endotracheal suctioning and minimally invasive airway suctioning. In the case of routine endotracheal suctioning, a 49 cm suction catheter was passed into the lower airways. With minimally invasive airway suctioning the suction catheter did not enter the lower airways and suctioning was limited to the endotracheal tube. A significantly lower prevalence of recollection of airway suctioning was found in the minimally invasive airway suctioning group (20%) than in the routine endotracheal suctioning group (41%; *P*\< 0.001). Our findings show that discomfort resulting from the endotracheal tube and its handling can be reduced by changing the procedure.
Hallucinations were another source of discomfort. In the total ICU patient group (*n*= 123), 24 (20%; 95% confidence interval 13--23%) of patients experienced hallucinations. This finding is comparable with that of an earlier and smaller study conducted by Holland and coworkers \[[@B2]\], who found that 10% of patients reported hallucinations. In a more recent study, Ely and colleagues \[[@B13]\] found that 81.7% of ICU patients developed delirium at some stage in their ICU stay. Delirium was an important variable, contributing as an independent predictor to higher 6-month mortality and longer hospital stay. Delirium was defined as \'a disturbance in consciousness characterized by an acute onset and fluctuating course of impaired cognitive functioning so that a patient\'s ability to receive, process, store and recall information is strikingly impaired\'. Clearly, the presence of delirium by this definition does not imply the presence of hallucinations. The exact percentage of patients who recalled hallucinations was not stated in the report by Ely and coworkers.
In studies conducted by Puntillo \[[@B14]\] and Holland and coworkers \[[@B2]\], pain was reported as a source of discomfort as well. In a post-cardiac surgery population (*n*= 24), Puntillo \[[@B14]\] described awareness of pain during the ICU period as a significant problem. Holland and coworkers \[[@B2]\] reported that, in a group of postsurgery patients (*n*= 21), 71% had a recollection of pain. In our study of mainly surgical ICU patients, only 12% indicated that pain was a source of discomfort. Differences in type of sedation and pain medication, number of patients, inclusion criteria and type of questionnaire used are possible explanations for the low recollection of pain in the present study as compared with previous ones.
A standardized score to assess recollection in this type of patient was lacking at the time our study was performed. We developed a factual recollection questionnaire that may represent a reliable new tool for acquiring information regarding recollection of facts in post-ICU patients. Analysis of reliability revealed a high Cronbach\'s alpha, and the descriptive data of our score showed a significant difference between ICU patients and the reference group. These findings are hardly surprising in view of the considerable differences between groups in severity of illness and consumption of hypnotics and sedatives. Further studies are needed to determine the validity and reliability of this instrument. Jones and coworkers \[[@B15]\] have since proposed a similar tool (Intensive Care Unit Memory tool), which has been validated in a number of settings \[[@B4],[@B16]\].
Both good factual recollection and younger age increased the risk for discomfort. Factual recollection and age were inversely associated with each other, but this association was weak. The association of increasing age with reduction in memory function is widely recognized \[[@B17],[@B18]\].
Although factual recollection and recollection of discomfort appear to be related, increasing the level of sedation is not necessarily the best way to prevent discomfort. Not only will deep sedation lead to increased length of stay in the ICU and prolonged ventilator dependency \[[@B19]\] but it may also have an adverse effect on the rate of post-traumatic stress disorder experienced by patients after their discharge from the ICU \[[@B10]\]. It has been proposed by various authors that factual recollection helps to offset the emotional impact of delusional memories \[[@B10],[@B19]\] and may actually help to avoid adverse psychological outcomes in this type of patient. The development of drugs that can eliminate the emotional impact of stressful events in the ICU, while preserving mental clarity and memory, might offer the best way to avoid long-term psychological distress. Meticulous treatment of delusional states will also contribute to this end.
Conclusion
==========
In a series of patients discharged from the ICU, 54% recalled discomfort. The most frequent sources of discomfort cited were presence of an endotracheal tube, hallucinations and medical interventions. The median factual recollection score for ICU patients was significantly lower than the median factual recollection score for ward patients who had not been in an ICU environment. Younger patients were at greater risk for remembering pain as source of discomfort. Patients with better factual recollection had greater recollection of discomfort. Factual recollection and age were inversely related, but this relationship was weak.
Discomfort thus appears to be a serious problem for patients in an ICU environment. Its prevalence is probably underestimated because retrospective assessment of the degree of discomfort when the patient has been discharged from the ICU is seriously handicapped by global or partial amnesia, caused by critical illness, delusional states and the use of drugs. However, the fact that discomfort is not always remembered does not imply that the patient has not suffered during his or her stay in the ICU. Reduction in discomfort should remain a focus of attention for both researchers and clinicians caring for critically ill patients.
Key messages
============
• Discomfort is a serious problem; 54% of ICU patients experienced discomfort.
• Endotracheal tube, hallucinations and medical interventions were cited as sources of discomfort.
• Patients with a higher factual recollection have greater recollection of discomfort.
Abbreviations
=============
APACHE = Acute Physiology and Chronic Health Evaluation; ICU = intensive care unit; OR = odds ratio.
Competing interests
===================
The author(s) declare that they have no competing interests.
Author\'s contributions
=======================
JvdL designed the study, performed data collection, data entry, statistical analysis and wrote the manuscript. CvdS, BL, BD and JZ participated in the design of the study. CvdS, BL, JG and JZ participated in the statistical analysis and writing the manuscript.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Scatterplot of factual recollection by age in intensive care unit (ICU) patients and the reference group (Control).
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Patient characteristics
:::
Characteristic ICU group (*n*= 123) Reference group (*n*= 48) *P*
---------------------------------------------- ---------------------- --------------------------- -------
Age (mean ± standard deviation) 61.5 ± 16 60.2 ± 16 0.617
Male sex (%) 71 65 0.435
APACHE II score (median \[min--max\]) 11 (2--26) NA \-
Type of patient: trauma/medical/surgical (%) 8/7/85 13/4/83 0.537
ICU stay in days (median \[min--max\]) 6.5 (2--133) NA \-
Ward stay in days (median \[min--max\]) NA 10 (3--53) \-
APACHE, Acute Physiology and Chronic Health Evaluation; ICU, intensive care unit; NA, not applicable.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Sources of discomfort in intensive care unit patients (*n*= 66)
:::
Source of discomfort \%^a^
---------------------- -------
Endotracheal tube 42
Hallucinations 32
Medical activities 29
Noise and bustle 14
Pain 12
Thirst 9
Inability to talk 9
Shortness of breath 6
Being afraid 6
^a^Because patients could list more than one source of discomfort, the summation of percentages exceeds 100%.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Items of factual recollection by intensive care unit patients and the reference group
:::
Correct (%) Incorrect (%) Don\'t know (%)
------------------------------------------- ------------- --------------- ----------------- ---- ---- ----
Type of patients\' clothing^†^ 68 100 12 0 20 0
Gender of nursing staff^†^ 66 98 7 2 27 0
Place of intravenous access^†^ 65 98 11 0 24 2
Color of staff uniform^†^ 62 98 14 2 24 0
Number of fellow patients\* 62 71 8 17 30 12
Type of personal hygiene^†^ 62 98 7 2 31 0
Logo on staff uniform^†^ 55 88 5 0 40 12
Type of lighting† 54 96 12 4 34 0
Reason inability to talk^†^ 50 94 24 6 26 0
Time of personal hygiene^†^ 48 100 3 0 49 0
Toilet visits† 42 100 32 0 26 0
Alternative headstand positions of bed^†^ 42 92 18 6 40 2
Type of food received^†^ 23 100 54 0 23 0
Time of ward round^†^ 11 98 34 0 55 2
\**P*\< 0.05 and ^†^*P*\< 0.005, intensive care unit patients versus reference group, by ?^2^test.
:::
::: {#T4 .table-wrap}
Table 4
::: {.caption}
######
Characteristics of ICU patients with and those without recollection of discomfort
:::
Characteristic Discomfort (*n*= 66) No discomfort (*n*= 57) *P*
---------------------------------------------------- ---------------------- ------------------------- ---------
Age (years; mean ± standard deviation) 59 ± 17 65 ± 14 0.004
Male sex (%) 65 77 0.143
APACHE II score (median \[min--max\]) 12 (2--26) 11 (5--24) 0.171
Duration of intubation (days; median \[min--max\]) 5 (2--35) 3 (1--57) 0.001
Factual recollection score (median \[min--max\]) 18 (0--28) 11 (0--24) \<0.001
APACHE, Acute Physiology and Chronic Health Evaluation.
:::
|
PubMed Central
|
2024-06-05T03:55:54.869918
|
2004-10-28
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065072/",
"journal": "Crit Care. 2004 Oct 28; 8(6):R467-R473",
"authors": [
{
"first": "Johannes P",
"last": "van de Leur"
},
{
"first": "Cees P",
"last": "van der Schans"
},
{
"first": "Bert G",
"last": "Loef"
},
{
"first": "Betto G",
"last": "Deelman"
},
{
"first": "Jan HB",
"last": "Geertzen"
},
{
"first": "Jan H",
"last": "Zwaveling"
}
]
}
|
PMC1065073
|
Introduction
============
Neonatal, pediatric, and adult extracorporeal membrane oxygenation (ECMO), using venoarterial or venovenous modes, have been practised for over 3 decades \[[@B1]-[@B5]\]. These modes of ECMO are known to activate the inflammatory cascade \[[@B6],[@B7]\], but the long-term cardiopulmonary outcome (10--15 years follow-up period) and neurodevelopmental outcome (at age 5 years) are relatively comparable to those in control individuals \[[@B8]-[@B10]\]. Patients who now receive ECMO therapy may also be different from patients treated in the 1980s and early 1990s because the alternative therapies have improved \[[@B11]\]. A search for safer modes of bypass therapy, including arteriovenous (AV)-ECMO, is warranted because of the cardiovascular and cerebral autoregulatory complications that are common during ECMO operations \[[@B12],[@B13]\]. This new mode of ECMO therapy may have some advantages over conventional venoarterial ECMO or venovenous ECMO techniques because the AV-ECMO technique appears simpler and may involve fewer operational complications \[[@B14]\].
The first investigators to conduct AV-ECMO trials, Kolobow and coworkers \[[@B15]\] studied eight normal and conscious lambs (age 1--8 days) for periods up to 96 hours. They described reductions in hemoglobin concentrations during AV-ECMO therapy, showing some mild postmortem pulmonary pathology in a few cases. In a later study, those investigators \[[@B16]\] also designed a carbon dioxide membrane lung, which was used to reduce ventilation in spontaneously breathing or sedated animals subjected to controlled mechanical ventilation. They suggested that a carbon dioxide membrane lung could ideally be operated in an AV mode without using a pump.
The AV shunt of the AV-ECMO circuit requires adequate blood flow from the systemic circulation, which may require an increase in cardiac output (CO). Animal models of AV-ECMO without acute lung injury (ALI) show clinically acceptable cardiorespiratory stability \[[@B17]-[@B21]\], whereas models with ALI usually require inotropic and fluid support \[[@B13],[@B22]-[@B26]\]. Conrad and coworkers \[[@B27]\], following a series of preclinical studies \[[@B14],[@B23]-[@B25]\], evaluated the safety and efficacy of AV-ECMO therapy in a phase I clinical study. They treated eight patients (five males and three females, aged 21--67 years), who had acute respiratory failure and hypercapnia, with AV-ECMO over a 72-hour period. They found no significant changes in hemodynamic variables, whereas arterial carbon dioxide tension (PaCO~2~) was significantly reduced from 90.8 ± 7.5 mmHg to 51.8 ± 3.1 mmHg after 2 hours of AV-EMCO therapy \[[@B23]\]. At the same time, minute ventilation was reduced from a baseline of 6.92 ± 1.64 l/min to 3.00 ± 0.53 l/min.
AV-ECMO technique applied in the presence of ALI requires reasonable hemodynamic stability to permit an extracorporeal AV shunt sufficient for carbon dioxide clearance. Recently, we demonstrated that lambs with normal lungs are able to maintain effective CO and provide efficient ventilator support with a relatively moderate AV shunt of 15% \[[@B17]\]. The aim of the present study was to determine the cardiovascular support needed to maintain hemodynamic stability and the minute ventilation needed to maintain normocapnia in lambs subjected to severe ALI and treated with AV-ECMO (AV shunt flow of up to 15%) or conventional mechanical ventilation (CMV; AV shunt flow of 0%).
Methods
=======
Surgical procedures
-------------------
The experimental protocol for this study was approved by the Institutional Animal Care and Use Committee of the Mount Sinai Hospital Research Institute (Miami Beach, FL, USA). Seventeen lambs (aged 2--6 weeks, weight 3.6--12.7 Kg) and their ewes were transported to the laboratory at least 3 days before the experiments began. On the day of an experiment, an intravenous line was established, and anesthesia was induced (initial dose 50 mg/kg ketamine intravenously) and maintained throughout the experiment (5 mg/kg per hour intravenous ketamine). A 2% xylocaine solution was used to provide local anesthesia at the incision sites. A while after induction of anesthesia (30--45 min), a tracheotomy was performed and the lambs were connected to a ventilator (Adult Star Infrasonics, Inc., San Diego, CA, USA) at a fractional inspired oxygen (FiO~2~) of 1.0. Animals were then paralyzed with an intravenous bolus of 1.0 mg/kg vecuronium bromide, followed by 0.1 mg/kg per hour.
To establish an ECMO circuit, one internal jugular vein and one carotid artery were cannulated using neonatal ECMO catheters (Medtronic Bio-Medicus, Inc., Eden Prairie, MN, USA). A femoral vein was then cannulated using a 5 Fr Swan--Ganz catheter (Baxter Health Care Co., Critical Care Division, Irvine, CA, USA) for periodic measurement of CO employing the thermodilution technique (Oximetrix-3, CO Computer; Abbott Critical Care System, North Chicago, IL, USA) and for continuous recording of the mean pulmonary artery pressure (PAP). A femoral artery was cannulated for continuous monitoring of the mean arterial pressure (MAP; Datascope 2001; Datascope Co., Paramus, NJ, USA) as well as periodic blood sampling for gas analyses. A bolus of 200 U/kg heparin was administered intravenously, followed by a maintenance infusion of 200 U/kg per hour. Normothermia (38 ± 0.5°C) was maintained throughout the experiments. Lactated Ringer\'s solution (5 ml/kg per hour) was provided for fluid replacement.
Procedures before injury
------------------------
One hour after the completion of all invasive procedures, pre-ALI baselines were determined for all investigated variables. Arterial blood samples, corrected for body temperature, were measured using a blood gas analyzer (ABL-30; Radiometer, Copenhagen, Denmark). The same samples were used to measure arterial hemoglobin concentration and hemoglobin--oxygen saturation (Hb-O~2~) using a hemoximeter (OSM-3; Radiometer). CO was determined by the thermodilution technique using the indwelling Swan--Ganz catheter and a CO computer (Oximetrix-3; Abbot Critical Care System). Minute ventilation was measured using a neonatal respiratory monitor (Bicore Neonatal Respiratory System, Model CP-100; Bicore, Irvine, CA, USA). The ventilator tidal volume was set at 7 ml/kg body weight and positive end-expiratory pressure was set at 4 cmH~2~O. The peak inspiratory pressure was maintained below 30 cmH~2~O. Because arterial hypercapnia may affect the cardiovascular system \[[@B28]\], maximizing the ability of the heart to drive the AV shunt, we elected to maintain the PaCO~2~between 30 and 45 mmHg, rather than allowing permissive hypercapnia to occur.
Acute lung injury model
-----------------------
To establish a model of severe ALI, in a preliminary study we used the above surgical procedures without AV lines in two lambs. This was accomplished with three consecutive saline lavages (5 ml/kg saline for each). The third lung lavage was followed by an intratracheal instillation of a single dose of 2.5 ml/kg 0.1 N HCl. This procedure resulted in substantial increases in the alveolar--arterial oxygen gradient and an average 60% increase in PAP with relatively stable CO over an 8-hour study period (Fig. [1](#F1){ref-type="fig"}). Saline lavage followed by tracheal instillation of HCl was used in all animals administered CMV and AV-ECMO therapy. This combination may result in surfactant deficiency (caused by the saline lavage), and cellular injury and edema (caused by pulmonary exposure to acid).
Post-acute lung injury procedures
---------------------------------
In our ALI model significant arterial hypercapnia developed (data not presented), which was adjusted to relative normocapnia by changes in the respiratory frequency. Based on our preliminary results in the ALI model, we allowed a 90-min interval before determination of a postinjury baseline in order to stabilize gas exchange and hemodynamic parameters. During this recovery period, arterial blood gases were determined every 15 min. A postinjury baseline for all variables was then determined (time 0). At this stage, lambs were consecutively assigned either to continued CMV treatment or to AV-ECMO plus CMV therapy.
### Group I
These lambs received continuous CMV support during a 6-hour study period with a closed AV shunt (*n*= 8). All hemodynamic, and arterial and venous mixed blood gas exchange variables were recorded every 2 hours. The oxygen content of both arterial and mixed venous blood was determined for calculation of oxygen consumption as a product of oxygen delivery (the difference between arterial oxygen content and mixed venous blood oxygen content) and CO (Fick\'s equation). Oxygen extraction was calculated using the differences between the measured values of arterial Hb-O~2~and venous Hb-O~2~saturation. After completion of the study period the lambs were euthanized by lethal dose of pentobarbital (100 mg/kg intravenously).
### Group II
In this treatment group a set of baseline values were obtained during CMV with a closed AV shunt (*n*= 9). Subsequently, lambs were subjected to 6 hours of AV-ECMO plus CMV (AV-ECMO therapy) with a maximum AV shunt of 15% (calculated from CO measured during postinjury baseline). The AV-ECMO circuit was established using a hollow fiber oxygenator (Minimax; Medtronic, Inc. Minneapolis, MN, USA) primed with fresh maternal blood (150--200 ml). To test the efficiency of AV-ECMO as compared with that of CMV in terms of carbon dioxide clearance, we attempted to maintain relative normocapnia in both groups. This required changes in minute ventilation that were achieved by modifying the respiratory rate while maintaining peak inspiratory pressure below 30 cmH~2~O. To control the flow rate through the AV shunt, a clamp was placed on the arterial side of the AV-ECMO circuit and the flow was continuously measured (Medical Volume Flow Meter; Transonic Systems Inc., Ithaca, NY, USA).
Carbon dioxide clearance during an AV-ECMO operation is dependent on the gas flow through the oxygenator. The efficacy of carbon dioxide removal and oxygenation of the Minimax hollow fiber oxygenator were previously studied in our laboratory using 15% AV shunt during stepwise decreases in minute ventilation and oxygenation with gas flow of 1 l/min \[[@B17]\]. This gas flow was approximately four times the maximum blood flow through the AV shunt and maintained normocapnia with a 50% reduction in minute ventilation \[[@B17]\]. In the present study, the oxygenator\'s gas flow was kept constant at 1 l/min of 100% oxygen and was controlled by an in-line gas regulator (Servo pressure limited system; Hudson RCI, Temecula, CA, USA). To ensure proper performance of the oxygenators during AV-ECMO therapy, the post-oxygenator partial oxygen tension and partial carbon dioxide tension were measured at 2 and 6 hours during the study period.
Resuscitative measures
----------------------
The outcome measures in our study were the degree of cardiovascular support needed to maintain hemodynamic stability and the minute ventilation needed to maintain normocapnia during both CMV and AV-ECMO therapy. A number of resuscitative measures were used to maintain hemodynamic stability during both CMV and AV-ECMO trials. These included the following: boluses of 10 ml/kg per hour of lactated Ringer\'s solution, which were provided if MAP fell below 60 mmHg; infusion of dopamine (5 μg/kg per min) and epinephrine (adrenaline; 0.5--2 μg/kg per min) to maintain MAP above 60 mmHg, given if this MAP was not achieved with fluid resuscitation; and 1 mEq/kg sodium bicarbonate, which was given if the base excess was below --5 mmol/l despite institution of other resuscitative measures. The end-point for resuscitation was deemed to have occurred when all of the above measures failed and the MAP fell below 30 mmHg for a period of 15 min. This cutoff point was selected empirically because below this level of MAP the AV-ECMO animals could not maintain an AV shunt of over 5% of baseline CO.
Statistical analyses
--------------------
All values are expressed as mean ± standard deviation. Differences in specific variables after establishment of postsurgery baseline (60 min after completion of surgery) and post-ALI baseline (90 min after injury), both within the same group at different times and between the CMV and AV-ECMO groups, were evaluated using two-tailed unpaired t-tests. Data from the surviving lambs in the same group over the 6-hour study period were evaluated using analysis of variance (ANOVA), followed by Dunnett multiple comparisons test. For this analysis, we used the postinjury baselines in each variable as controls. Differences in each parameter among the surviving lambs in CMV and AV-ECMO groups and a group of nonsurvivors in the AV-ECMO category were evaluated using ANOVA, followed by Bonferroni multiple comparison test for comparable time periods. The use of resuscitative measures (lactated Ringer\'s, dopamine, epinephrine and bicarbonate) in all lambs after time zero and in the surviving lambs in the CMV and AV-ECMO groups, as well as mortality (death before completion of the 6-hour study period), were compared using Fisher\'s exact test. All resuscitative measures before baseline (time zero) were excluded from data analyses. *P*\< 0.05 was considered statistically significant.
Results
=======
Pre- and post-acute lung injury baselines
-----------------------------------------
These data were collected in all animals (survivors and nonsurvivors) before assignment to the CMV or the AV-ECMO groups (Table [1](#T1){ref-type="table"}). No significant differences were found between the preinjury values of lambs that were later randomized to CMV and AV-ECMO groups. After ALI, all lambs required significant increases in minute ventilation in order to achieve relative normocapnia (Table [1](#T1){ref-type="table"}). Comparison of postinjury PaCO~2~and pH between the two treatment groups revealed statistically significant differences in favor of the AV-ECMO group (Table [1](#T1){ref-type="table"}). ALI created a arterial oxygen tension (PaO~2~)/FiO~2~ratio of less than 200 (also representing PaO~2~) in both groups. After ALI, the PAP was significantly increased by approximately 50% in both groups. There were no significant differences in the postinjury baselines of MAP, PAP, and CO between the groups. The average body weight, measured before surgical procedures, was not significantly different between lambs consecutively assigned to CMV and those that were assigned to AV-ECMO (6.3 ± 1.7 kg versus 8.5 ± 2.8 kg, respectively). However, the four surviving lambs in the AV-ECMO group had significantly greater body weight than the five nonsurviving lambs (11.0 ± 2.2 kg versus 6.5 ± 1.3 kg; *P*\< 0.05, by two-tailed unpaired t-test).
Conventional mechanical ventilatory support versus arteriovenous extracorporeal membrane oxygenation therapy
------------------------------------------------------------------------------------------------------------
The data presented in Tables [2](#T2){ref-type="table"} and [3](#T3){ref-type="table"}, and Figs [1](#F1){ref-type="fig"} and [2](#F2){ref-type="fig"} are from the surviving lambs only. Six out of eight lambs (75%) in the CMV group and four out of nine lambs (44%) in the AV-ECMO group survived the 6-hour study period after ALI. Three of the five nonsurviving lambs in the AV-ECMO group died within 45--90 min and two others died after 4 hours, despite a combination of resuscitative measures. On average, the surviving lambs in both groups had stable CO and MAP during the 6-hour study period (Tables [2](#T2){ref-type="table"} and [3](#T3){ref-type="table"}). The four surviving lambs in the AV-ECMO group were able to maintain CO and MAP with varying degrees of hemodynamic support. This also allowed for a relatively stable AV shunt flow (14.8 ± 0.4% of the CO, measured at 0, 2, 4, and 6 hours) and a significant reduction of 25--30% in minute ventilation, as compared with the CMV group (Fig. [2](#F2){ref-type="fig"}).
There were no significant differences between the PaCO~2~in CMV and AV-ECMO treated lambs during the study period, but the alveolar--arterial oxygen gradient was consistently higher in the AV-ECMO group (Fig. [3](#F3){ref-type="fig"}). The last measurements of MAP, PAP, and PaO~2~, which were obtained in four out of the five nonsurviving lambs in the AV-ECMO group, were 33.5 ± 9.3, 36.0 ± 6.3, and 53.7 ± 9.2 mmHg, respectively. These values were significantly lower than those recorded in the surviving lambs in either the AV-ECMO or the CMV group (Tables [2](#T2){ref-type="table"} and [3](#T3){ref-type="table"}; ANOVA followed by Bonferroni multiple comparison test). Gas exchange of the oxygenators remained stable within the 6 hours of the study period. For example, the postoxygenator partial oxygen tension was 282 ± 8 mmHg and 282 ± 7 mmHg at 2 and 6 hours, respectively, and the postoxygenator partial carbon dioxide tension was 19.7 ± 5.1 mmHg and 21.0 ± 5.0 mmHg at 2 and 6 hours of AV-ECMO therapy.
Hemodynamic stability
---------------------
Analysis of the use of resuscitative measures as indicators of hemodynamic stability between the CMV and AV-ECMO groups revealed that significantly more lambs in the AV-ECMO group (including survivors and nonsurvivors) were resuscitated than in the CMV group (Table [4](#T4){ref-type="table"}; *P*\< 0.001, Fisher\'s exact test). However, there was no significant difference in \'mortality\' between AV-ECMO and CMV groups within the 6-hour period of study (*P*\> 0.05, Fisher\'s exact test).
Discussion
==========
The cardiovascular effects of AV-ECMO have been studied in adult and neonatal animal models \[[@B14]-[@B26]\]. It has been suggested that the resistance of the membrane oxygenator, hemodynamic stability, and the number, size and length of the conducting cannula, as well as the viscosity of the blood, will all affect the exogenous flow rate \[[@B22]\]. In the present study we utilized a low resistance membrane oxygenator, minimized the length of the conducting cannulae, and attempted to maintain MAP above 60 mmHg by using various resuscitative measures (Table [4](#T4){ref-type="table"}). These measures in the AV-ECMO group failed to sustain hemodynamic stability in five out of nine lambs (56%), whereas the survivors (44%) were able to maintain normocapnia with a maximum of 30% reduction in minute ventilation over a 6-hour period of study (Fig. [2](#F2){ref-type="fig"}). The latter implies that AV-ECMO therapy, providing an AV shunt flow of up to 15% of the CO, may be able to reduce ventilator-induced lung injury in hypercapnic respiratory failure. However, in acute respiratory failure or acute respiratory distress syndrome with high intrapulmonary right-to-left shunt, extracorporeal blood flow in the range of 5--15% of CO may not be sufficient to provide adequate arterial oxygenation.
The reasons for the relatively poor performance of AV-ECMO therapy in our lamb model, as compared with the findings of studies conducted in adult animals \[[@B14],[@B23],[@B25]\], may be related to a number of factors. These possibilities are considered below.
First, differences between our model and other experimental models of ALI could account for differences between our findings and those of other studies. The present model may create a noncardiogenic pulmonary edema, which could be associated with loss of intravascular volume. Such conditions may require prolonged fluid and positive inotropic treatments to support a sufficient AV shunt flow. In comparison, Zwischenberger and coworkers \[[@B6],[@B25]\] used an adult sheep model, in which acute respiratory distress syndrome was induced by smoke inhalation and 40% third degree burns. Sheep were then ventilated for 2 days before randomization to CMV and AV-ECMO (AV shunt of 11--14%) groups for a period of 7 days. There were no deaths in the AV-ECMO group (*n*= 8), as compared with only three survivors in the CMV group (*n*= 8). That model \[[@B6],[@B25]\] demonstrates that perhaps a longer period of CMV support is needed to achieve relative cardiovascular stability before subjecting animals with severe ALI to the additional stress of an AV shunt.
How may a short recovery period after ALI affect hemodynamic stability during an AV-ECMO operation? ALI leads to the release of a variety of bioactive materials, including proinflammatory cytokines and reactive oxygen species \[[@B29]\]. The addition of an ECMO circuit to animals with ALI is known to stimulate the generation of inflammatory mediators, leading to further deterioration in cardiovascular function \[[@B6],[@B7],[@B16]\]. Zwischenberger and coworkers \[[@B6]\] studied the pathophysiology of ovine smoke inhalation lung injury after a relatively short recovery interval of 6 hours during both conventional ECMO therapy and CMV in female sheep. Those investigators demonstrated that animals treated with smoke and ECMO had significantly increased circulating thromboxane B~2~levels and oxygen free radical activity, and a significant increase in lung wet:dry weight ratios. They suggested that an ECMO operation could potentiate the pathophysiology of smoke inhalation injury and lead to initial deterioration in native lung function \[[@B6]\]. Therefore, despite the simplicity of AV-ECMO procedures, as compared with conventional ECMO \[[@B14],[@B30]\], it could be still subject to free radical generation because of presence of the membrane oxygenator. Thus, the addition of an AV shunt after ALI may further compromise the cardiovascular system.
A second factor that could account for the discrepancy between our findings and those of other investigators is that the AV shunt opening in our study led to a mortality rate in the smaller lambs, resulting in a difference between the body weights of the surviving lambs in two groups. This implies that smaller (and presumably younger) lambs with ALI could be more vulnerable to the presence of an AV shunt than relatively larger or older animals. Thus, studies concerning the safety and efficacy of neonatal AV-ECMO therapy should use animals with a narrow age range (1--7 days in lambs).
The third factor is whether the ALI in the CMV and AV-ECMO therapy groups was equal in severity. Whether the severity of ALI was different between the groups may be indirectly evaluated by comparing the indices of pre- and post-injury gas exchange. Our data indicate that pulmonary performance before starting AV-ECMO therapy was comparable with that observed in the CMV group (Table [1](#T1){ref-type="table"}). The degree of lung injury was not significantly worsened during the 6-hour study period, as judged by lack of significant changes in alveolar--arterial oxygen gradient in the surviving lambs subjected to CMV or AV-ECMO therapy (Fig. [3](#F3){ref-type="fig"}).
Study limitations
-----------------
The outcome measures in this study were the degree of hemodynamic stability and the minute ventilation required to maintain relative normocapnia, while comparing CMV support with AV-ECMO therapy. Our study was not designed to evaluate mortality as an ultimate clinical outcome. A greater number of lambs would have been required to demonstrate significant differences in mortality between the CMV and AV-ECMO groups. However, the more than 50% mortality rate in the AV-ECMO group may raise questions about the clinical and/or statistical significance of our findings. Technically, we failed to use a narrow range of age and body weight in our lambs. However, the average body weights in lambs consecutively randomized to CMV support and AV-ECMO therapy were not significantly different (Table [1](#T1){ref-type="table"}).
Conclusion
==========
Our study indicates that cardiovascular support is required to maintain hemodynamic stability during application of AV-ECMO therapy in lambs with severe ALI. In this model, AV-ECMO therapy with continuous cardiovascular support and an AV shunt flow of 15% of CO can provide a maximum 30% reduction in minute ventilation. We suggest that AV-ECMO with cardiovascular support \[[@B30]\] could be suitable for use in ALI of mild severity, in which permissive hypercapnia is not an acceptable treatment \[[@B28],[@B31]\].
Key messages
============
• Continuous hemodynamic support is required during AV-ECMO in lambs subjected to severe ALI.
• By using a shunt flow of up to 15% of CO, AV extracorporeal therapy in lambs with severe ALI can reduce minute ventilation by 25--30%.
• Neonatal patients with severe ALI and hemodynamic instability may not be suitable candidates for AV-EMCO therapy.
Abbreviations
=============
ALI = acute lung injury; ANOVA = analysis of variance; AV = arteriovenous; ECMO = extracorporeal membrane oxygenation; CMV = conventional mechanical ventilation; CO = cardiac output; FiO~2~= fractional inspired oxygen; Hb-O~2~= hemoglobin--oxygen saturation; MAP = mean arterial pressure; PaCO~2~= arterial carbon dioxide tension; PaO~2~= arterial oxygen tension; PAP = pulmonary artery pressure.
Competing interests
===================
The author(s) declare that they have no competing interests.
Author\'s contributions
=======================
BRT, JBS and DT completed the proposal writing and experimental design. DT and BRT participated in research coordination, data analysis and presentation. JG, HF, YM, and JLO conducted all experimental aspects of the study. BRT, DT, JBS, and JW prepared the manuscript.
Acknowledgment
==============
This study was supported, in part, by a Research Grant from Miami Children\'s Hospital Foundation to Jeffrey B Sussmane, MD, FAAP, FCCM, and by the Alex Simberg Fund for Critical Care Medicine.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Changes in the average alveolar--arterial oxygen (A-a O~2~) gradient, pulmonary artery pressure (PAP), and cardiac output in two lambs after three separate lavages and intratracheal instillation of 2.5 ml/kg of 0.1 N HCl (fractional inspired oxygen 0.6). Time -- 1 hour indicates baseline values before induction of acute lung injury (ALI). Data were periodically collected, starting 90 min after ALI procedures.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Comparisons between the minute ventilations (calculated per kg body weight) required to maintain normocapnia in lung-injured lambs subjected to conventional mechanical ventilation (CMV) or arteriovenous (AV)-extracorporeal membrane oxygenation (ECMO) with shunt flow of 15% of baseline cardiac output. Analysis of variance (ANOVA) followed by Dunnett multiple comparisons test was used to compare the preinjury level of minute ventilation in each group with subsequent measurements. ANOVA followed by Bonferroni test was used to compare CMV and AV-ECMO therapies at different time periods during the study period. Values are expressed as mean ± standard deviation. \**P*\< 0.05. ALI, acute lung injury.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Changes in the alveolar--arterial oxygen (A-a O~2~) gradient in six lambs subjected to continued conventional mechanical ventilation (CMV) support and four lambs subjected to arteriovenous (AV)-extracorporeal membrane oxygenation (ECMO) therapy with a maximum shunt flow of 15%, up to 6 hours after establishment of acute lung injury (ALI). A-a O~2~after ALI was consistently higher with AV-ECMO therapy than with CMV support. These differences became statistically significant at 4--6 hours, indicating higher deterioration in lung performance in the AV-ECMO group (repeated measures of analysis of variance followed by Dunnett multiple comparisons test, using the postinjury baseline in each group as controls).
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Comparison of cardiorespiratory variables before and after induction of lung injury
:::
CMV (*n*= 8) AV-ECMO (*n*= 9)
--------------------------- -------------- ------------------ -------------- ------------------
Minute volume (ml/kg/min) 408 ± 79 640 ± 144\* 382 ± 109 561 ± 187\*
PaO~2~(mmHg) 389 ± 128 113 ± 85\*\*\* 393 ± 131 179 ± 86\*\*\*
PaCO~2~(mmHg) 37.1 ± 5.1 40.9 ± 3.1 36.7 ± 1.9 35.1 ± 3^†^
Arterial pH 7.311 ± 0.05 7.263 ± 0.04 7.349 ± 0.05 7.344 ± 0.03^††^
HCO~3~^-^(mmol/l) 17.7 ± 3.2 17.4 ± 1.7 19.5 ± 2.8 18.3 ± 2.1
Arterial Hb-O~2~(%) 99.8 ± 0.3 89.6 ± 11 99.8 ± 0.3 95.1 ± 11
O~2~-extraction (%) 28.0 ± 5.7 35.9 ± 4.9\* 24.7 ± 8.1 35.7 ± 9.3\*
Hb (g/dl) 9.4 ± 1.5 10.3 ± 2.0 8.1 ± 2.0 8.7 ± 2.3
MAP (mmHg) 84.2 ± 12.1 88.0 ± 10.9 96.7 ± 9.0 92.0 ± 14
PAP (mmHg) 13.1 ± 4.1 20.1 ± 5.7 14.5 ± 5.4 21.5 ± 4.9\*
CO (ml/kg per min) 185 ± 23 164 ± 53 181 ± 69 173 ± 56
VO~2~(ml/kg per min) 5.7 ± 2.2 8.7 ± 2.7\* 5.4 ± 1.5 7.8 ± 2.5\*
Body weight (kg)
All lambs 6.3 ± 1.7 \- 8.5 ± 2.8 \-
Surviving lambs \- 6.5 ± 1.3 \- 11.0 ± 2.2^†^
Range 3.6--9.2 \- 5.0--12.7 \-
Comparison of cardiorespiratory variables before and after induction of lung injury in surviving and nonsurviving lambs subjected to conventional mechanical ventilation (CMV) or arteriovenous (AV)-extracorporeal membrane oxygenation (ECMO) therapy. Values are expressed as mean ± standard deviation. \**P*\< 0.05, \*\**P*\< 0.01, \*\*\**P*\< 0.001, pre-injury baseline versus post-injury baseline in the same group. ^†^*P*\< 0.05, ^††^*P*\< 0.01, pre-injury or post-injury baselines: CMV versus AV-ECMO. CO, cardiac output; Hb-O~2~, hemoglobin--oxygen saturation; MAP, mean arterial pressure; PaCO~2~, arterial carbon dioxide tension; PaO~2~, arterial oxygen tension; PAP, pulmonary artery pressure; VO~2~, oxygen consumption.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Hemodynamics and oxygen consumption
:::
Study period (hours after establishment of acute lung injury)
------------------------------------ --------------------------------------------------------------- --------------- ---------------- ----------------
Minute ventilation (ml/kg per min)
CMV 624 ± 189 550 ± 152 590 ± 151 608 ± 192
AV-ECMO 397 ± 96 214 ± 83^†^\* 222 ± 128^†^\* 256 ± 155^†^\*
MAP (mmHg)
CMV 89.6 ± 12.2 81.5 ± 10.1 83.8 ± 17.0 82.5 ± 9.0
AV-ECMO 92.5 ± 10.6 90.0 ± 16.2 73.5 ± 25.0 74.5 ± 40
PAP (mmHg)
CMV 20.8 ± 6.4 23.1 ± 7.7 22.6 ± 3.7 18.6 ± 6.8
AV-ECMO 19.0 ± 5.7 24.7 ± 7.5\* 26.2 ± 7.8\* 26.2 ± 9.3\*
Cardiac output (ml/kg per min)
CMV 164 ± 45 165 ± 28 189 ± 62 194 ± 54
AV-ECMO 144 ± 56 177 ± 30 143 ± 47 184 ± 54
Oxygen consumption (ml/kg per min)
CMV 9.6 ± 2.4 9.2 ± 1.5 10.1 ± 3.4 8.9 ± 3.7
AV-ECMO 7.1 ± 1.7 9.7 ± 2.3 8.1 ± 1.3 8.2 ± 3.7
Hemodynamics and oxygen consumption in lung injured lambs (surviving) supported by conventional mechanical ventilation (CMV; *n*= 6) or arteriovenous (AV)-extracorporeal membrane oxygenation (ECMO; *n*= 4) during a 6-hour period of study. Values are expressed as mean ± standard deviation. \**P*\< 0.05, baseline (time 0) versus 2, 4, and 6 hours of study by repeated measures analysis of variance (ANOVA) followed by Dunnett multiple comparisons test. ^†^*P*\< 0.05, CMV versus AV-ECMO groups; ANOVA followed by Bonferoni multiple comparisons test.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Gas exchage variables
:::
Study period (hours after establishment of acute lung injury)
--------------------- --------------------------------------------------------------- -------------- -------------- --------------
PaO~2~(mmHg)
CMV 131 ± 90 207 ± 171 231 ± 175 221 ± 189
AV-ECMO 174 ± 100 95 ± 29 77 ± 17\* 94 ± 76
PaCO~2~(mmHg)
CMV 41.3 ± 3.1 43.4 ± 8.3 38.9 ± 10.3 37.0 ± 7.1
AV-ECMO 34.8 ± 2.3\*\* 35.1 ± 8.5 37.1 ± 7.8 37.8 ± 5.7
pH
CMV 7.263 ± 0.05 7.237 ± 0.05 7.286 ± 0.10 7.289 ± 0.08
AV-ECMO 7.322 ± 0.04\* 7.277 ± 0.15 7.235 ± 0.10 7.207 ± 0.14
HCO~3~^-^(mmol/l)
CMV 17.0 ± 3.1 16.8 ± 3.2 16.7 ± 2.7 15.9 ± 3.3
AV-ECMO 17.2 ± 2.4 15.7 ± 2.3 15.7 ± 5.4 14.6 ± 4.6
Arterial Hb-O~2~(%)
CMV 91.8 ± 8.1 89.2 ± 14.9 94.0 ± 4.9 88.9 ± 18.4
AV-ECMO 90.7 ± 16.5 89.0 ± 9.9 84.4 ± 7.0 74.0 ± 24.8
O~2~extraction (%)
CMV 42.4 ± 15.6 37.6 ± 9.2 47.2 ± 11.5 41.2 ± 6.9
AV-ECMO 41.6 ± 11.9 37.3 ± 14.2 49.3 ± 21.0 43.3 ± 21.7
Gas exchage variables in lung injured lambs (surviving) supported by conventional mechanical ventilation (CMV; *n*= 6) or arteriovenous (AV)-extracorporeal membrane oxygenation (ECMO; *n*= 4) during a 6-hour period of study. No significant differences were found when comparing baselines (time 0) with 2, 4, and 6 hours of study by repeated measures analysis of variance (ANOVA) followed by Dunnett multiple comparisons test. Values are expressed as mean ± standard deviation. \**P*\< 0.05, \*\**P*\< 0.01, CMV versus AV-ECMO group; ANOVA followed by Bonferroni multiple comparisons test. Hb-O~2~, hemoglobin--oxygen saturation; PaCO~2~, arterial carbon dioxide tension; PaO~2~, arterial oxygen tension.
:::
::: {#T4 .table-wrap}
Table 4
::: {.caption}
######
Numbers of lambs undergoing various resuscitative measures
:::
Type of support Group *P*^a^
----------------------------------------------------------- ------------------------------------------ ------------------------------------------------------------------ -------
Lactated Ringer\'s (10 ml/kg) 2 8 0.015
Epinephrine (0.5--2 μg/kg per min) 1 6 0.049
Dopamine (5 μg/kg per min) 1 6 0.049
Bicarbonate (1 mEq/kg bolus) 1 6 0.049
Surviving/nonsurviving 6/2 4/5 0.333
Total number of resuscitative measures in surviving lambs 2 (*n*= 6) 12 (*n*= 4) 0.001
Cause of death Prolonged hypotension with MAP \<30 mmHg Prolonged hypotension with MAP \<30 mmHg and AV shunt \<5% of CO
Comparison of various resuscitative measures after acute lung injury in surviving and nonsurviving lambs subjected to conventional mechanical ventilation (CMV) with closed arteriovenous (AV) shunt or CMV with AV-extracorporeal membrane oxygenation (ECMO) using up to 15% AV shunt. ^a^*P*values derived using Fisher\'s exact test. CO, cardfiac output; MAP, mean arterial pressure.
:::
|
PubMed Central
|
2024-06-05T03:55:54.872573
|
2004-10-28
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065073/",
"journal": "Crit Care. 2004 Oct 28; 8(6):R495-R503",
"authors": [
{
"first": "Balagangadhar R",
"last": "Totapally"
},
{
"first": "Jeffrey B",
"last": "Sussmane"
},
{
"first": "Dan",
"last": "Torbati"
},
{
"first": "Javier",
"last": "Gelvez"
},
{
"first": "Harun",
"last": "Fakioglu"
},
{
"first": "Yongming",
"last": "Mao"
},
{
"first": "Jose L",
"last": "Olarte"
},
{
"first": "Jack",
"last": "Wolfsdorf"
}
]
}
|
PMC1065074
|
See related commentary <http://ccforum.com/content/8/6/437>
Introduction
============
The majority of mechanically ventilated patients in the intensive care unit (ICU) require sedation to reduce their anxiety and to increase their tolerance of the tracheal tube and mechanical ventilation. The choice of sedative drugs and the way in which they are administered may have an important impact on patient outcome and cost of care \[[@B1]\]. Excessively deep sedation will prolong ventilator dependence and length of stay in the ICU, which can be avoided by careful monitoring and interruption to sedative infusions \[[@B2]\]. Differentiation between adequate comfort and excessive sedation requires the use of clinically relevant sedation scales; however, these are not suitable for application during deep sedation or muscle relaxation. Other methods to assess the level of sedation in the clinical setting are therefore needed.
Growing knowledge of the depressive effects of sedative drugs on the central nervous system has led to increasing interest in a possible correlation between neurophysiological indices and the level of sedation. The most commonly used neurophysiological indices in the assessment of sedation are electroencephalogram (EEG) and auditory evoked potentials (AEPs), which measure different aspects of brain functioning. The evoked potentials show whether the central nervous system responds systematically to an auditory stimulus, and they may thus be considered a direct measure of the responsiveness of the brain. In contrast, the EEG signal, if not associated with a sensory stimulus, will only reflect the ongoing background electrical activity of the brain. In other words, if the patient is not stimulated and the level of sedation is measured using indices derived from the EEG signal, then it can be speculated that those indices may only be used as predictors of whether the patient will actually react to a given stimulus, but they provide no measure of responsiveness. AEPs may therefore provide a more accurate tool with which to assess the level of sedation.
Within the AEPs, the middle-latency AEPs (10--50 ms after the stimulus) are mainly evoked by the physical features of the auditory stimulus. Their presence establishes the integrity of the afferent auditory pathway and confirms that basic auditory signal processing is taking place in the primary auditory cortex (Fig. [1a](#F1){ref-type="fig"}). The long-latency AEPs, or event-related potentials (ERPs; \>50 ms after the stimulus), result from deeper processing of the auditory stimulus and are generated by areas of cortex at and beyond the primary projection area. ERPs may therefore be better indicators of the effect of sedative drugs on the mental state than are middle-latency AEPs.
The most prominent ERP component is N100, which appears about 100 ms after the onset of stimulus and reflects the simultaneous activation of several different brain regions, indicating detection of a change in acoustic surroundings (Fig. [1b](#F1){ref-type="fig"}) \[[@B3]\]. Another ERP component, namely mismatch negativity (MMN), is elicited by infrequently presented stimuli that differ in some physical dimension from the standard stimuli and reflects the brain\'s automatic auditory change detection mechanism, which depends on the integrity of auditory sensory memory (Fig. [1c](#F1){ref-type="fig"}) \[[@B4]\]. Appearance of MMN indicates that several brain regions are activated simultaneously. The fact that MMN reflects widespread brain activation may explain why sedative drug induced changes in the MMN have been shown to be a better marker of mental state than are the respective changes in the middle-latency AEPs \[[@B5]\].
ERPs have exhibited graded changes with increasing doses of sedative drugs in volunteers and surgical patients \[[@B6],[@B7]\], but to date only few data are available concerning the use of ERPs for monitoring sedation level in the ICU. Despite the known superiority of ERP parameters over EEG parameters for monitoring sedation level, in this preliminary pilot study we hypothesized that both ERPs and EEG may be used to assess the level of sedation in a heterogeneous group of neurologically intact intensive care patients.
Methods
=======
The study protocol was approved by the local ethical committee and written informed consent was obtained from each patient or from the next of kin.
We measured EEG and ERPs in a heterogeneous group (*n*= 19; 13 males and six females; age 65 ± 11 years) of mechanically ventilated patients presenting with a range of surgical and medical conditions requiring intensive care but with no known organic brain dysfunction (Table [1](#T1){ref-type="table"}). Patients who were known to have impaired hearing were excluded from the study. Sedation was administered following the modified Brook protocol \[[@B1]\]. Repeated midazolam boluses were initially used to induce and maintain sedation. If the obtained sedation level was still considered inadequate, then propofol infusion was begun and midazolam administration discontinued. The optimal depth of sedation for each patient was determined on clinical grounds, independent of the study, and was assessed using the sedation--agitation scale (SAS; Table [2](#T2){ref-type="table"}) \[[@B8]\].
At the time of the first EEG and ERP recordings, patients were receiving propofol sedation (infusion rate 1.91 ± 0.88 mg/kg per hour) and the duration of the infusion had exceeded 8 hours in all patients (31 ± 29 hours). Discontinuation of sedation was then considered necessary so that the patients could be weaned from the ventilator or so that their neurological status could be evaluated. Propofol infusion was interrupted, and the measurements were repeated once the sedation had subsided and the patients were able to follow commands (i.e. to open their eyes and squeeze their hand). Apart from propofol, no sedative drugs other than opioids were allowed during the 8 hours preceding the measurements or during the study period (Table [1](#T1){ref-type="table"}).
Electroencephalogram and event-related potential recording
----------------------------------------------------------
The EEG signal was recorded using Ag/AgCl electrodes placed on the scalp according to the international 10--20 system. Two electrode locations (frontal \[Fz\] and central \[Cz\]) were used. Both electrodes were referred to the right mastoid, and the electrode--skin impedances were kept below 5 kO. The EEG signal was amplified and digitized continuously at 279 Hz using EMMA (ERP measuring machine; developed and custom-made in the Department of Clinical Neurophysiology, Kuopio University Hospital, Kuopio, Finland).
Background EEG was recorded for 5 min during sleep and/or while the patients lay motionless with their eyes closed. Auditory stimulation was then set to \'on\' so that ERPs could be recorded. The stimulation was applied according to an oddball paradigm, which consisted of 85% standard (800 Hz) and 15% deviant (560 Hz) stimuli, with an interstimulus interval of 1 s. The duration of each stimulus was 84 ms, including 7 ms rise and fall times. Altogether 600 stimuli were delivered through earphones to the right ear for each measurement, corresponding to a recording time of about 10 min. The stimulus intensity was set at 75 dB.
Electroencephalogram analysis
-----------------------------
The background EEG, measured before auditory stimulation, was band pass filtered using a finite impulse response-type filter employing cutoff frequencies of 0.5 and 32 Hz (Matlab, version 6.12; The Mathworks Inc., Natick, MA, USA). Then, the filtered EEG signal (5 min long) was cut into 5 s epochs with 50% overlap. Serious artifacts were excluded by checking the maximum amplitude for each epoch; if the amplitude was greater than 100 µV then the epoch was excluded. The appropriateness of artifact rejection was manually confirmed.
For each EEG epoch, first the root mean squared (RMS) total power was calculated. Then, the epoch was subjected to power spectral density estimation, using Welsh\'s averaged periodogram method \[[@B9]\], and the spectral edge frequency 95% (SEF95) was computed from the power spectral density using a frequency range of 0.5--32 Hz. The mean of the RMS and SEF95 values of the accepted epochs were then individually computed.
Event-related potential analysis
--------------------------------
The EEG signal recorded during the auditory stimulation was first filtered using a finite impulse response-type filter using cutoff frequencies of 1 and 20 Hz, and then transformed to epochs from -100 ms to +900 ms relative to the onset of each stimulus. After removing artifactual epochs (rejection level ± 100 µV), the individual responses to standard and deviant stimuli were averaged. The N100 component was defined as a maximum negative deflection appearing 80--150 ms from the stimulus onset. The amplitude and the latency of the prominent N100 components in response to standard stimuli were manually scored with respect to the pre-stimulus baseline. The MMN was obtained by subtracting first the waveform elicited by the standard stimuli from the one resulting from the deviant stimuli. The MMN was then computed from the difference curve (deviant standard) as the mean amplitude between 100 and 250 ms \[[@B10]\].
Statistical analysis
--------------------
We carried out exploratory analyses to determine which EEG and ERP parameters changed significantly in response to interruption to sedation. For this purpose, Wilcoxon signed rank test (nonparametric paired sample test) was applied to the N100 amplitude and latency values (in response to standard stimuli), MMN, RMS power and SEF95 values measured before and after interruption to sedation. Moreover, Kruskal--Wallis test (nonparametric counterpart of one-way analysis of variance) was used to test whether the ERP and the EEG parameters differed among the sedation levels present during propofol infusion. The effect of the total duration of propofol infusion on the studied parameters was assessed using Spearman\'s correlation coefficient. The recording channels Fz and Cz were studied separately. Data are expressed as mean ± standard deviation, unless otherwise indicated. All statistical analyses were done using the SPSS software (SPSS for Windows, version 11.0; SPAA Inc., Chicago, IL, USA). *P*\< 0.05 was considered statistically significant.
Results
=======
During propofol infusion the sedation level for each patient was determined on clinical grounds. It varied from deep sedation (SAS score 2) to light sedation (SAS score 4). All patients were responsive and cooperative (SAS score 4) within 30 min after discontinuation of propofol. Weaning and extubation were successful in 10 patients, whereas sedation was electively restarted in the remaining nine patients.
Of the ERP recordings, 2% and 5% were discarded as artifact during and after sedation, respectively. Accordingly, 8% and 20% of the background EEG recordings were discarded.
Effect of interruption to propofol infusion
-------------------------------------------
The EEG parameters (RMS power and SEF95) and ERP parameters (N100 and MMN) measured before and after interruption to sedation did not differ between those patients who proceeded to weaning and extubation and those in whom sedation was restarted. The RMS power increased after interruption to sedation (Fz and Cz, *P*\< 0.05; Fig. [2a,2b](#F2){ref-type="fig"}), whereas the SEF95 values exhibited only a tendency toward a decrease (not significant; Fig. [2c,2d](#F2){ref-type="fig"}). The amplitude of the N100 component (in response to standard stimuli) increased at both frontal (Fz, *P*\< 0.05) and central recording sites (Fig. [3a,3b](#F3){ref-type="fig"}). The latency of the N100 component (in response to standard stimuli) and the MMN did not change in response to interruption to propofol infusion. The MMN mean amplitude, which should be a negative value while awake, exhibited both positive and negative values after sedation had subsided (Fig. [3c,3d](#F3){ref-type="fig"}).
Effect of sedation level
------------------------
During propofol infusion, seven patients were deeply sedated (SAS score 2), seven patients were moderately sedated (SAS score 3) and five patients were lightly sedated (SAS score 4). The level of sedation did not influence EEG parameters. The amplitude of the N100 component (in response to standard stimuli) differed between sedation levels (Fz and Cz, *P*\< 0.05), in contrast to N100 latency and MMN (Fig. [3](#F3){ref-type="fig"}). Both negative and positive MMN mean amplitudes were obtained independently of sedation level (Fig. [3c,3d](#F3){ref-type="fig"}).
Patient characteristics and duration or rate of propofol infusion did not differ among sedation level groups.
Effect of propofol infusion duration
------------------------------------
None of the ERP and EEG parameters correlated with the total duration of propofol infusion.
Discussion
==========
ERPs have exhibited graded changes with increasing doses of sedative drugs in volunteers and surgical patients \[[@B6],[@B7]\], but to date no parallel studies have been conducted in severely ill patients. We assessed ERPs together with EEG parameters in a heterogeneous group of intensive care patients under sedation with propofol. The range of doses of sedative and analgesic drugs varied widely, but despite this our preliminary data suggest an association between clinical level of sedation and neurophysiological parameters. Our main findings were that the amplitude of the standard N100 component differed among the sedation levels during propofol sedation, and that the amplitude of the standard N100 in the frontal area as well as the RMS power increased in response to interruption to propofol infusion.
We selected RMS power and SEF95 to describe the changes in the EEG spectrum related to the interruption to propofol infusion. The RMS power represents the total power of the signal and the SEF95 is the frequency below which 95% of the power in the EEG spectrum resides. Sedative doses of propofol have been shown to produce an increase in total, delta and beta activity in the EEG signal, especially in the Cz and Fz regions \[[@B11]-[@B13]\]. In our study the total power of the EEG signal was inversely related to sedation, increasing after interruption to propofol infusion. However, the SEF95 decreased in many patients under the same circumstances. This suggests that awakening was not paralleled by a prominent increase in the high frequency range, probably due to the decrease in beta activity related to interruption to propofol infusion. Administration of opioids might also have markedly modified the EEG pattern as compared with that observed during isolated propofol infusion.
Identifiable ERPs may indicate an increased risk for auditory perception during general anaesthesia \[[@B14],[@B15]\] and a positive outcome in coma patients \[[@B16],[@B17]\]. During propofol sedation, the N100 component has been reported to decrease in amplitude and to delay in latency as compared with recording before the beginning of propofol infusion \[[@B5]\]. As sedation subsides, the opposite (amplitude increase and latency shortening) has been observed in surgical patients recovering from postoperative propofol sedation \[[@B7]\]. In the present study the N100 amplitude recovered similarly as the level of sedation subsided, although the amplitude values were markedly smaller than those of the surgical patients both during sedation and after sedation had subsided. Moreover, the MMN exhibited a large inter-individual variability and many patients had a positive MMN mean amplitude (Fig. [3c,3d](#F3){ref-type="fig"}), suggesting that MMN was not present or could not be reliably measured. In our earlier study conducted in surgical patients \[[@B7]\], the MMN was present at comparable sedation levels.
The small N100 amplitude and the absence of the MMN could have resulted from the use of medication other than propofol and opioids during the study period. We cannot exclude the presence of some level of sedative potentiation or side effects resulting from this medication, which might have affected the results. In all patients benzodiazepines were discontinued for a minimum of 8 hours before measurements were taken. However, some degree of residual sedative effect due to potentially impaired metabolism might have influenced our findings. Clifford and Buchman \[[@B18]\] reported that the combination of benzodiazepine and fentanyl affected information processing in response to novel and standard stimuli in a different manner than the combination of propofol and fentanyl in intensive care patients. Nevertheless, both of these drug combinations globally reduced the amplitudes of the responses to all stimuli as the sedative drug dose increased, in a manner similar to that in our study. We also speculate that, because of the short time allowed after propofol discontinuation, patients were still under influence of this drug during the later measurements. Thus, ERP parameters might not have had enough time to recover, even if the patients were awake and able to follow simple commands (SAS score 4). We did not study the effect of opioids on the ERPs in more detail because subanaesthetic doses of fentanyl \[[@B19]\] and remifentanil \[[@B20]\] have been shown not to attenuate the N100 component.
In the intensive care setting, EEG parameters and ERPs are influenced not only by the administration of sedative drugs but also by the underlying illness, which may cause considerable changes in functioning of the sensory pathways \[[@B21]\]. Diagnosis and reason for intensive care varied considerably in our population. We excluded patients with known organic brain dysfunction from the study, but it is possible that some of the patients suffered from mild subclinical neurological deficits. However, because all patients woke up and were able to follow commands, we believe that possible brain dysfunction did not have a significant effect on our results. Moreover, no differences could be found in neurophysiological parameters between extubated patients and those whose sedation was continued electively.
The statistical methods we applied deserve comment. We conducted exploratory analyses to determine which EEG and ERP parameters changed significantly because of interruption to sedation. Performing multiple comparisons, as we did, is known to increase the risk for type I error (i.e. obtaining significant differences by chance). However, because of both the exploratory nature of our analysis and the controversy concerning the Bonferroni method, we opted not to use this adjustment \[[@B22],[@B23]\]. Furthermore, the heterogeneity of our patient group limits the power of statistical analysis. To overcome this limitation, we presented individual data points and used statistical analysis only to show trends in our findings.
Conclusion
==========
In a group of intensive care patients, with heterogeneous diagnosis and reasons for intensive care, assessment of the level of sedation using spectral EEG alone may not be sufficiently accurate. Concomitant use of ERPs, especially the N100 component, which requires widespread activity and functional integrity of the brain, may provide better distinction between sedation levels. Neurophysiological methods may thus be useful complements to clinical sedation scales in the monitoring of sedation status over time in intensive care patients under controlled sedative drug administration.
Key messages
============
• The EGG alone may not be sufficiently accurate in the assessment of sedation levels in intensive care unit patients.
• Concomitant use of ERPs, especially the N100 potential, may help to differentiate between sedation levels.
• Neurophysiological methods may offer a complement to clinical sedation scales in neurologically intact intensive care patients.
Abbreviations
=============
AEP = auditory evoked potential; Cz = central region; EEG = electroencephalogram; ERP = event-related potential; Fz = frontal region; ICU = intensive care unit; MMN = mismatch negativity; RMS = root mean squared; SAS = sedation--agitation scale; SEF95 = spectral edge frequency 95%.
Competing interests
===================
The author(s) declare that they have no competing interests.
Author\'s contributions
=======================
HYP, SN, IK, JP and ER participated in the interpretation of the results and writing of the manuscript. HYP and SN performed data collection, data entry and statistical analysis.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
**(a)**The middle-latency auditory evoked potential (MLAEP) components Na, Pa, and Nb appear 10--50 ms after the onset of auditory stimulus. **(b)**N100 is the most prominent event-related potential (ERP) component. The thick line is the N100 for standard stimuli (N100 S) and the thin line is the N100 for deviant stimuli (N100 D). (c) The mismatch negativity (MMN) curve is obtained as a difference curve N100 D--N100 S. The MMN is the negative area under the curve between 100 and 250 ms.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Average and individual root mean squared (RMS) power and spectral edge frequency 95% (SEF95) values during and after discontinuation of propofol infusion in the **(a, c)**frontal (Fz) and **(b, d)**central (Cz) regions. Lines connect values obtained from the same patient; black squares with vertical lines indicate the mean ± standard deviation. Individual sedation levels obtained with the Sedation--Agitation Scale (SAS): white spheres: SAS 4, gray spheres: SAS 3, black spheres: SAS2. \*Significantly different from \'propofol on\'.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Average and individual N100 standard amplitude and mismatch negativity (MMN) values during and after discontinuation of propofol infusion in the **(a, c)**frontal (Fz) and **(b, d)**central (Cz) regions. Lines connect values obtained from the same patient; black squares with vertical lines indicate the mean ± standard deviation. Individual sedation levels obtained with the Sedation--Agitation Scale (SAS): white spheres: SAS 4, gray spheres: SAS 3, black spheres: SAS2. \*Significantly different from \'propofol on\'.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Demographic data, duration and rate of propofol infusion at the time of measurements
:::
Patient number/sex Age (years) Length (cm) Weight (kg) Diagnosis Propofol infusion^a^(mg/kg per hour) Duration of infusion^b^(hours) Opioids^c^
-------------------- ------------- ------------- ------------- ------------------------------------------------------ -------------------------------------- -------------------------------- -----------------------------------
1/M 59 180 86 Thoracic aorta dissection 1.63 31 Oxycodon 10 mg
2/M 53 180 130 Acute myocardial infarction 0.62 11 Fentanyl 0.100 mg/hour
3/M 47 173 68 Pneumonia and sepsis (streptococcal pneumonia) 1.18 13 Oxycodon 3 mg
4/F 47 170 68 Multitrauma (renal rupture, pelvic fracture) 1.76 20 Oxycodon 10 mg
5/M 61 169 96 Ruptured abdominal aortic aneurysm 2.08 19 Fentanyl 0.100 mg/hour
6/M 76 174 73 Acute myocardial infarction and peritonitis 2.47 66 Fentanyl 0.100 mg/hour
7/F 72 160 70 Acute lung injury and status post-AVR+CABG 1.71 10 Oxycodon 5 mg
8/M 66 176 90 Acute lung injury and status post-CABG 4.44 46 Fentanyl 0.150 mg/hour
9/M 68 162 79 Wound infection post-CABG 3.04 112 Oxycodon 35 mg
10/M 64 164 65 Peritonitis and septic shock 1.85 13 Fentanyl 0.200 mg
11/F 70 162 89 Acute myocarial infarction and status post-CABG 0.67 16 Oxacodon 5 mg
12/M 83 167 65 Peritonitis 2.15 19 Fentanyl 0.075 mg
13/M 72 176 96 Sternal dehiscence post-CABG 1.46 14 Oxycodon 36 mg
14/M 76 183 77 Acute myocarial infarction and pulmonary haemorrhage 1.04 18 Oxycodon 10 mg
15/F 71 162 58 Wound infection post-CABG+AVR 1.72 19 Fentanyl 0.150 mg, oxycodon 29 mg
16/M 77 167 73 Acute respiratory distress syndrome 2.74 69 Fentanyl 0.825 mg, oxycodon 15 mg
17/F 50 170 75 Low cardiac output (status post-CABG) 1.60 86 Fentanyl 0.150 mg/hour
18/F 59 165 60 Acute lung injury and septic shock 1.67 12 Oxycodon 3 mg
19/M 71 170 80 Acute myocarial infarction and pulmonary oedema 1.50 14 Oxycodon 18 mg
Mean 65 170 79 \- 1.91 31 \-
SD 11 7 16 \- 0.88 29 \-
^a^Rate of propofol infusion at the time of measurements. ^b^Number of hours of continuous propofol infusion before measurements. ^c^Opioid medication administered during the 12 hours before (total intravenous bolus) and/or during the measurements (infusion rate). AVR, aortic valve replacement; CABG, coronary artery bypass graft.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
The Sedation--Agitation Scale
:::
Score Clinical status
------- ----------------------
7 Dangerous agitation
6 Very agitated
5 Agitated
4 Calm and cooperative
3 Sedated
2 Very sedated
1 Unarousable
Data from Riker and coworkers \[8\].
:::
|
PubMed Central
|
2024-06-05T03:55:54.876361
|
2004-10-22
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065074/",
"journal": "Crit Care. 2004 Oct 22; 8(6):R483-R490",
"authors": [
{
"first": "Heidi",
"last": "Yppärilä"
},
{
"first": "Silvia",
"last": "Nunes"
},
{
"first": "Ilkka",
"last": "Korhonen"
},
{
"first": "Juhani",
"last": "Partanen"
},
{
"first": "Esko",
"last": "Ruokonen"
}
]
}
|
PMC1065077
|
See related commentary <http://ccforum.com/content/8/6/430>
Introduction
============
Morbidity and mortality after surgery for severe acute pancreatitis (SAP) remain considerable, despite the introduction of new strategies to reduce infectious complications \[[@B1],[@B2]\], such as antibiotic prophylaxis, early enteral nutrition \[[@B3]\], and the recognition of complications such as abdominal compartment syndrome in severely ill patients \[[@B4]\].
There is limited evidence in the literature that postponing surgery beyond the initial phase of the disease leads to improved survival. Mier and colleagues \[[@B5]\] randomized 36 patients to early versus late surgery, and stopped the study after an interim analysis showed that patients operated on early had a higher mortality. This finding has been confirmed by others in retrospective studies. Hungness and colleagues \[[@B6]\] found a trend toward an increased mortality in 14 of 26 patients who were operated on within the first two weeks of diagnosis. Hartwig and colleagues \[[@B7]\] found in a review of 62 surgically treated patients that those operated on within three days had a higher mortality rate (53% versus 22%, *P*= 0.02). In contrast, Fernández-del Castillo and colleagues \[[@B8]\] found a similar mortality rate in their patients when either operated on early or later than 6 weeks after admission. There are conflicting data on the impact of timing of surgery on mortality, and the different definitions used for early surgery, ranging from three days to six weeks, makes comparing the data in the literature difficult.
All studies that reported increased mortality in patients undergoing early surgery included patients operated on for a range of indications (such as absence of clinical improvement after 3--5 days, persistent pancreatitis, infected necrosis, pancreatic abscess and sepsis syndrome) at different stages of the disease. It is not clear to what extent the severity of illness at the moment of surgery or the microbiological status of the necrosis were confounding factors and were a bias in finding increased mortality rates for early surgery.
In this paper we report our study on the impact of the timing of surgical intervention and perioperative factors (severity of organ dysfunction and microbiological status of the necrosis) on mortality in patients undergoing surgery for SAP.
Materials and methods
=====================
Data collection
---------------
We retrospectively (January 1994 to March 2003) analyzed all patients admitted with SAP to the intensive care unit (ICU) of the Ghent University Hospital, a tertiary referral centre with a total of 1060 beds. SAP was defined in accordance with the criteria described by the International Symposium on Acute Pancreatitis \[[@B9]\]. Patients were identified from the hospital registry with the use of the International Classification of Diseases (ICD-9-CM) code for acute pancreatitis\' (577.0). Preoperative data collected included age, sex, etiology, use of antibiotics, C-reactive protein level, Ranson score and Acute Physiology And Chronic Health Evaluation (APACHE) II score \[[@B10]\] on admission. Time to the first surgical intervention, the severity of organ dysfunction at the day of the first surgical intervention (as assessed by the sepsis-related organ failure assessment (SOFA) score \[[@B11]\]), length of stay in the ICU and in the hospital, and mortality were retrieved from the patient\'s file.
The occurrence of organ dysfunction during the ICU stay was recorded, and organ dysfunction was defined as follows (based on a score of 2 or more in the SOFA scoring system): (1) cardiovascular dysfunction was defined as hypotension requiring vasoactive medication; (2) renal dsyfunction, serum creatinine above 2.0 mg/dl; (3) respiratory dyscfunction, the need for mechanical ventilation or a PaO~2~/FIO~2~ratio of less than 300.
Microbiological data collected included peroperative cultures from the initial surgical intervention, and fine-needle aspirates (FNAs), when available. Infected pancreatic necrosis was defined as the presence of microorganisms in cultures obtained at the first operation or in cultures of a FNA of the pancreatic necrosis without previous surgery; consequently, sterile pancreatic necrosis was defined as negative cultures from intraoperative cultures, independently of infections occurring later in the course of the disease.
Mortality was defined as in-hospital mortality.
The study was approved by the local ethical committee.
Study design
------------
Patients treated surgically early in the course of the disease were compared with patients who underwent delayed surgical intervention. Early surgery was defined as surgery within 12 days of diagnosis, as described in the prospective trial by Mier and colleagues \[[@B5]\]. Furthermore, we compared patients with sterile pancreatic necrosis with patients with infected necrosis, and survivors with non-survivors, using univariate and multivariate analysis techniques.
Patient management
------------------
All patients were admitted to the ICU before or after surgical treatment and were treated by the same surgical team. The use of antibiotic prophylaxis was left to the discretion of the attending ICU physician. Enteral nutrition was started as early as possible. Computed tomography (CT) scanning and FNA of the pancreatic necrosis was performed on an individual patient base, namely when the clinical condition of the patient was suggestive of infection of the pancreatic necrosis. Indications for surgery were a documented infection of pancreatic necrosis (as evidenced by positive cultures from FNA), a deterioration of the clinical condition of the patient, unresolving pancreatitis or suspected pancreatic infection without proof on FNA or CT scan. Surgical intervention consisted of necrosectomy through a midline laparotomy as described by Beger and colleagues \[[@B12]\]. The pancreas was debrided using blunt dissection, and two to four large-calibre drains were inserted in the retroperitoneum. Continuous postoperative lavage of the retroperitoneum was started initially at a rate of 500--1000 ml/h, and progressively decreased, on the basis of the general condition of the patient, inflammatory parameters (C-reactive protein), and the macroscopic aspect of the drain effluent.
Statistical analysis
--------------------
Statistical analysis was performed with SPSS for Windows 11.0.1^®^(SPSS, Chicago, IL, USA). Continuous variables were compared by using Student\'s *t*-test or the Mann--Whitney *U*-test where appropriate. Categorical data were compared with the ?^2^or Fisher Exact test. A double-sided *P*value of less than 0.05 was considered statistically significant. Parameters found to be different in survivors and non-survivors in univariate analysis with a *P*value of 0.25 or less were entered in a logistic regression model with mortality as the dependent variable, to identify factors available at the moment of surgery that were independently associated with mortality.
Results
=======
Patients
--------
Of 124 patients with SAP, 56 (35 male, 21 female) were treated surgically. The mean age of the patients was 56 years (SD 13.5). The cause of the pancreatitis was biliary tract stones in 19 patients (33.9%), alcohol in 21 (37.5%), trauma in 6 (10.7%), hyperlipemia in 1 (1.8%) and idiopathic in 9 (16.1%). Thirty-nine patients (69.6%) were referred from other hospitals; for three patients the first surgical intervention was performed in the referring hospital 1 day before referral (*n*= 2) or on the day of referral (*n*= 1).
Early versus late surgical intervention
---------------------------------------
Twenty-two patients (39.2%) were operated on within the first 12 days of diagnosis of pancreatitis (median 5 days, interquartile range 3--9), and 34 (60.8%) later than 12 days after admission (median 20 days, interquartile range 17--31).
Age and gender distribution were comparable in both groups (Table [1](#T1){ref-type="table"}). Disease severity, assessed by Ranson and APACHE II scores on admission in these patients, was not different; neither was the SOFA score at the day of surgery. Indications for surgery in patients operated on early were different from those operated on later in the course of the disease. In patients operated on early, deterioration of multiple organ dysfunction syndrome (MODS) was the indication for surgical intervention in 41% of the patients, compared with 9% in the late surgery group. Overall, the length of stay in the hospital was significantly longer for the patients who underwent surgery late in the course of the disease, even after censoring the patients who died in both groups. Duration of ICU stay was not different. There was a trend toward a higher mortality in the early surgery group (55% versus 29%, *P*= 0.06).
Microbiological status of necrosis and mortality
------------------------------------------------
In 26 (46.4%) patients, intraoperative cultures confirmed the diagnosis of infected pancreatic necrosis. Microorganisms isolated from the necrosis are listed in Table [2](#T2){ref-type="table"}. Gram-negative and Gram-positive microorganisms were present in comparable numbers (38.9%); seven patients had fungal infections at the first operation. In 10 patients more than one organism was isolated. Thirty of 56 patients (54%) had sterile pancreatic necrosis at the moment of the first surgical intervention. Patient characteristics, severity of disease, and the timing of surgery were not different in patients with sterile or infected pancreatic necrosis (Table [3](#T3){ref-type="table"}). There was a trend toward a higher occurrence rate of organ failure in patients with sterile pancreatic necrosis, and mortality was significantly higher in patients with sterile necrosis (57%) than in patients with infected necrosis (19%) (*P*= 0.004).
Especially in patients undergoing early surgery, mortality was significantly higher in patients with sterile pancreatic necrosis (85% versus 11%, *P*= 0.001) In the patients who underwent delayed surgery, there was no difference in mortality between patients with sterile pancreatic necrosis and those with infected pancreatic necrosis (35% in patients with sterile pancreatic necrosis and 23% in patients with infected pancreatic necrosis, *P*= 0.71).
Factors influencing outcome after surgical intervention
-------------------------------------------------------
Overall mortality in our patients was 39.2% (22 of 56 patients). Table [4](#T4){ref-type="table"} summarizes differences between survivors and non-survivors. In univariate analysis, patients who died were older, had higher APACHE II scores on admission, higher SOFA scores on the day of surgery, more often sterile necrosis, and more often organ dysfunction during their ICU stay, and were operated more often because of MODS. There was also a trend toward earlier surgical intervention in patients who died.
The following variables were entered in a logistic regression analysis: age, SOFA score on the day of surgery, the presence of sterile pancreatic necrosis at surgery, and interval from diagnosis to surgical intervention as a continuous variable. SOFA score at the day of surgery was preferred to APACHE II score on admission and deteriorating MODS as an indication for surgery, because it better describes the severity of illness at the moment of surgery, and the difference in univariate analysis was more significant. In multivariate analysis, only age, SOFA score at the moment of surgery, and the presence of sterile necrosis were associated with mortality (Table [5](#T5){ref-type="table"}).
Discussion
==========
It has been suggested that postponing surgery beyond the initial phase of the disease leads to improved survival \[[@B5]-[@B7]\]. In this analysis of 56 patients undergoing surgery because of SAP, we found that disease severity at the moment of surgery, age, and the presence of sterile necrosis, but not early surgery, determined mortality. The trend toward an increased mortality in patients operated on within 12 days of diagnosis, found in univariate analysis, was apparently confounded by perioperative factors.
Disease severity at admission has long been recognized as an important factor determining outcome in patients with SAP, irrespective of surgical intervention. So far, Ranson score at admission and C-reactive protein levels at 48 hours \[[@B13],[@B14]\] have proven to be the best predictors of disease severity; more recently, the APACHE II score \[[@B15]\] and determination of the individual Ranson parameters \[[@B16]\] at 48 hours showed improved predictive value compared with admission scores.
In patients undergoing surgery for SAP, perioperative organ dysfunction affects outcome. Connor and colleagues \[[@B17]\] reported that a high postoperative APACHE II score was the only factor associated with mortality in a group of moderately ill patients (initial APACHE II score 9) undergoing pancreatic necrosectomy. Hungness and colleagues \[[@B6]\] reported higher organ failure scores and more advanced age in patients who died after surgery for SAP. The present study further confirms these findings.
The reason for this increased mortality is not clear. Surgical intervention by itself in the early phase of the disease is a possible explanation for the high mortality rate in patients undergoing early surgery, and has been suggested by several authors \[[@B6],[@B7]\], but the evidence for this is indirect. It seems plausible that in the early stage of the disease, when there is peripancreatic and retroperitoneal inflammation, surgery is often difficult, with increased blood loss. Patients with severe organ dysfunction might also be more prone to other complications that could arise from the surgical intervention, such as gastrointestinal ischemia or blood loss. Another possibility is that in these patients other complications -- that have only recently been recognized -- were involved, and were left untreated for too long. Intra-abdominal hypertension and abdominal compartment syndrome are increasingly described in patients with SAP \[[@B4],[@B18]\], and can lead to multiple organ dysfunction. Other problems such as relative adrenal insufficiency \[[@B19]\], which is increasingly recognized in patients with septic shock \[[@B20]\] or high-risk surgical patients \[[@B21]\], might be involved.
The fact that sterile necrosis is a risk factor for mortality in patients undergoing surgery is an important finding. At first sight this might be in sharp contradiction of the fact that infected pancreatic necrosis has been associated with increased mortality in several studies. It should be kept in mind that this undoubtedly is true for patients with SAP as a whole, and that this analysis included only patients who were operated on.
The findings of the present study are in line with the current concept that patients with sterile pancreatic necrosis do not need surgery, although this is still advocated by some experts in the field. Several authors have reported mortality rates below or about 10% when managing these patients non-operatively \[[@B22]-[@B24]\]. Le Mée and colleagues reported that, in most of their patients, organ dysfunction was reversible if necrosis remained sterile \[[@B25]\]. These and our results suggest that in patients with suspected infection of the pancreatic necrosis, the presence of microorganisms should be actively sought with ultrasound-guided or CT scan-guided FNA before surgical intervention is considered \[[@B26]\].
Our study could not reproduce the negative impact of early surgery on outcome after adjustment for other factors that were associated with increased mortality in univariate analysis. The often-used strategy to postpone surgery in patients with SAP is based on limited data. Mier and colleagues \[[@B5]\] randomized 36 patients with SAP to early (within 48--72 hours) versus late surgery (later than 12 days). Mortality in the group that underwent early debridement was 56%, 3.4-fold that in the control group, a result that halted the trial.
This finding has also been reported by other investigators, but the definition of early surgery should be carefully considered, because the use of different time frames makes it very difficult to compare the evidence available in the literature. Fernandez-del Castillo and colleagues \[[@B8]\] analyzed 64 patients operated on with a technique of closed packing, and found that mortality in patients operated on within the first six weeks after onset of the disease was not different from mortality in patients operated on later than six weeks. This study included patients with pancreatic abscesses, a disease that has a different clinical course and prognosis from that of patients who require surgery for infected pancreatic necrosis. Patient selection and the definition of early surgery make it very difficult to compare this study with ours.
Hartwig and colleagues \[[@B7]\] found a significantly higher mortality in patients operated on within 72 hours (53% versus 22%, *P*= 0.02) in 136 patients treated between 1980 and 1997, about half of them surgically. During the study period, indications for surgery gradually shifted from a lack of clinical improvement after 2--3 days to a suspicion of infected necrosis, resulting in patients being operated on later, and lower mortality rates. Over all, operating less, and if necessary, as late as possible, markedly improved outcome.
From the data available in the literature, the advice to postpone surgery by default beyond the first 2--3 weeks seems to be based on unblinded, unadjusted, or retrospective analyses. A similar process has been observed with the use of prophylactic antibiotics. The use of these became widespread on the basis of limited evidence, but the benefit could not be demonstrated in a controlled randomized trial \[[@B27]\]. Although we agree that there are several pathophysiological considerations in deferring surgery, such as those described above, we did not find any evidence that the timing of surgery by itself influenced outcome.
Conclusion
==========
Our data suggest that not the timing of the surgical intervention, but rather perioperative factors, determine mortality in critically ill patients undergoing necrosectomy for SAP. We found that mortality was associated with greater age, increasing severity of organ dysfunction, as expressed by the SOFA score at the moment of surgery, and the presence of sterile necrosis. In future studies on the effect of timing of surgery, the severity of organ dysfunction and microbiological status at surgery should be evaluated as possible confounding variables.
Key messages
============
• In a series of 56 patients who were treated surigically for severe acute pancreatitis, no effect of the timing of surgery was found if perioperative factors such as severity of illness and microbiological status of the necrosis were considered.
Abbreviations
=============
APACHE = Acute Physiology And Chronic Health Evaluation; CT = computed tomography; FNA = fine-needle aspirate; ICU = intensive care unit; MODS = multiple organ dysfunction syndrome; SAP = severe acute pancreatitis; SOFA = sepsis-related organ failure assessment.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
JDW, UH and FC were responsible for the conception and design of the study. JDW and SB acquired a substantial portion of the data. JDW, EH and DV performed the analysis and interpretation of data. JDW and DV drafted the manuscript. FC, PP, BDH, JDC and EH undertook critical revision of the manuscript for important intellectual content. EH and SB were responsible for statistical expertise. FC performed supervision and took overall responsibility for all aspects of the project or study. All authors read and approved the final manuscript.
Acknowledgements
================
This paper was presented in part at the 16th annual congress of the European Society of Intensive Care Medicine, Amsterdam, The Netherlands, 5--8 October 2003. The study was supported by a Clinical Doctoral Grant of the Fund for Scientific Research -- Flanders (Belgium) (F.W.O.-Vlaanderen).
Figures and Tables
==================
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Characteristics, indications for surgery and outcome of patients operated on for severe acute pancreatitis (*n*= 56)
:::
Parameter Early surgery^a^(*n*= 22) Late surgery^a^(*n*= 34) *P*
------------------------------------------------------------------------- --------------------------- -------------------------- ---------
Patient characteristics
Age, years (mean ± SD) 54 ± 14.8 56 ± 12.8 0.58
Male sex 16 (72.7%) 19 (55.9%) 0.20
APACHE II score (mean ± SD) 22 ± 12.1 19 ± 8.9 0.47
Ranson score (mean ± SD) 6.2 ± 2.46 5.8 ± 1.80 0.59
Sterile necrosis at first surgical intervention 13 (59.1%) 17 (50%) 0.50
Interval from diagnosis to surgery, days (median and IQR) 5 (3--9) 20 (17--31) \<0.001
SOFA score at surgical intervention (median and IQR) 4 (2--8) 4 (2--8) 0.78
Indications for surgery
Documented infection of pancreatic necrosis 5 (22.7%) 14 (41.2%) 0.25
Deteriorating clinical condition 9 (40.9%) 3 (8.8%) 0.007
Unresolving pancreatitis or suspected infection of pancreatic necrosis 8 (36.4%) 17 (50%) 0.41
Outcome
LOS in ICU, days (median and IQR) 14 (5--33) 14 (6--35) 0.75
LOS in hospital, days (median and IQR) 29 (15--58) 87 (54--106) \<0.001
LOS in ICU in hospital survivors, days (median and IQR)^b^ 16 (4--46) 12 (5--31) 0.92
LOS in hospital in hospital survivors, days (median and IQR)^b^ 44 (30--107) 88 (60--106) 0.034
Mortality 12 (54.5%) 10 (29.4%) 0.06
APACHE II score, Acute Physiology And Chronic Health Evaluation II score; ICU, intensive care unit; IQR, interquartile range; LOS, length of stay; SOFA, sepsis-related organ failure assessment.
^a^Early surgery was defined as surgery within the first 12 days after admission.
^b^Early surgery, *n*= 10; late surgery, *n*= 24.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Microorganisms (*n*= 36) isolated from 26 patients with infected pancreatic necrosis
:::
Microorganism *n*
----------------------------- -----
Gram-positive bacteria 14
*Staphyloccus epidermidis* 4
*Staphylococcus aureus* 4
Enterococci 6
Gram-negative bacteria 14
*Escherichia coli* 7
*Enterobacter aerogenes* 4
*Pseudomonas*spp. 1
*Stenotrophomonas*sp. 1
*Proteus mirabilis* 1
Fungi 7
*Candida*spp. 7
Anaerobes 1
Bacteroides 1
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Comparison of patients with infected and sterile pancreatic necrosis
:::
Characteristic Infected pancreatic necrosis (*n*= 26) Sterile pancreatic necrosis (*n*= 30) *P*
----------------------------------------------------------- ---------------------------------------- --------------------------------------- -------
Age, years (mean ± SD) 56 ± 13.6 55 ± 13.7 0.75
Male gender 18 (78%) 17 (57%) 0.33
APACHE II score (mean ± SD) 20 ± 9.7 21 ± 11.1 0.71
CRP at admission, mg/dl (mean ± SD) 17 ± 14.7 16 ± 15.8 0.87
Ranson score (mean ± SD) 6.2 ± 1.7 5.7 ± 2.3 0.50
Organ failure
Respiratory insufficiency 19 (7%) 24 (80%) 0.54
Acute renal failure 15 (58%) 24 (80%) 0.07
Cardiovascular failure 21 (81%) 20 (67%) 0.12
Interval from diagnosis to surgery, days (median and IQR) 14 (6--24) 18 (6--31) 0.36
LOS in ICU, days (median and IQR) 14 (5--32) 12 (5--30) 0.86
LOS in hospital, days (median and IQR) 68 (44--90) 54 (19--97) 0.36
Mortality 5 (19%) 17 (57%) 0.004
APACHE II score, Acute Physiology And Chronic Health Evaluation II score; CRP, C-reactive protein; ICU, intensive care unit; IQR, interquartile range; LOS, length of stay.
:::
::: {#T4 .table-wrap}
Table 4
::: {.caption}
######
Comparison of survivors and non-survivors (*n*= 56)
:::
Characteristic Non-survivors (*n*= 22) Survivors (*n*= 34) *P*
----------------------------------------------------------- ------------------------- --------------------- ---------
Age, years (mean ± SD) 62 ± 12.0 51 ± 12.8 0.002
APACHE II score (mean ± SD) 25 ± 8.5 18 ± 10.6 0.019
Ranson score (mean ± SD) 6.3 ± 2.03 5.7 ± 2.09 0.36
Male gender 15 (68%) 20 (59%) 0.48
CRP at admission, mg/dl (mean ± SD) 176 ± 162.3 170 ± 142.2 0.90
Organ dysfunction
Respiratory insufficiency 22 (100%) 21 (62%) 0.001
Acute renal failure 22 (100%) 17 (50%) \<0.001
Cardiovascular failure 20 (91%) 12 (35%) \<0.001
Sterile necrosis 17 (77%) 13 (38.2%) 0.004
SOFA score at surgery (median and IQR) 4 (9--13) 2 (2--7) 0.005
MODS as indication for surgery 8 (36%) 4 (12%) 0.045
Early surgical intervention 12 (55%) 10 (29%) 0.06
Interval from diagnosis to surgery, days (median and IQR) 11 (4--22) 18 (12--29) 0.09
APACHE II score, Acute Physiology And Chronic Health Evaluation II score; CRP, C-reactive protein; IQR, interquartile range; MODS, multiple organ dysfunction syndrome; SOFA, sepsis-related organ failure assessment.
:::
::: {#T5 .table-wrap}
Table 5
::: {.caption}
######
Multivariate analysis
:::
Variable *P* OR 95% CI
------------------------------------ ------- -------- ----------------
Sterile necrosis 0.012 13.704 1.778--105.602
SOFA score at surgery (per point) 0.009 1.351 1.076--1.695
Age (per year older) 0.004 1.124 1.037--1.218
Interval from diagnosis to surgery 0.868 1.006 0.939--1.078
CI, confidence interval; OR, odds ratio; SOFA, sepsis-related organ failure assessment.
:::
|
PubMed Central
|
2024-06-05T03:55:54.879181
|
2004-11-2
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065077/",
"journal": "Crit Care. 2004 Nov 2; 8(6):R504-R511",
"authors": [
{
"first": "Jan J",
"last": "De Waele"
},
{
"first": "Eric",
"last": "Hoste"
},
{
"first": "Stijn I",
"last": "Blot"
},
{
"first": "Uwe",
"last": "Hesse"
},
{
"first": "Piet",
"last": "Pattyn"
},
{
"first": "Bernard",
"last": "de Hemptinne"
},
{
"first": "Johan",
"last": "Decruyenaere"
},
{
"first": "Dirk",
"last": "Vogelaers"
},
{
"first": "Francis",
"last": "Colardyn"
}
]
}
|
PMC1065097
|
See related commentary <http://ccforum.com/content/9/1/25>
Introduction
============
Most paediatric intensive care unit (ICU) patients need sedative and analgesic drugs during mechanical ventilation \[[@B1]\]. Sedatives and analgesics are given to improve comfort, to reduce pain, to facilitate aggressive ICU therapy (i.e. mechanical ventilation or insertion of intravascular lines) and to avoid accidental removal of medical devices. Most clinicians give sedatives and analgesics according to their professional experience and the patient\'s estimated need for sedation. Inadequate monitoring of sedation and analgesia may contribute to adverse outcomes and complications \[[@B2]\].
Only a few clinical scores have been validated for estimating the level of sedation in paediatric ICU patients. The best evaluated score is the COMFORT score \[[@B3]\]. The COMFORT score consists of eight categories and can evaluate a child\'s behaviour and physiological responses to discomfort, fear and pain in approximately 2 min. The COMFORT score is age independent because age-adapted physiological parameters are used. Apart from the examination of muscular tone, calculation of the COMFORT score does not require any stimulation of the patient. The COMFORT score can be divided into three groups. A score of 8--16 points corresponds to deep sedation, 17--26 indicates light sedation and 27--40 indicates inadequate sedation \[[@B4]\]. Common concerns about clinical sedation scales are that the employed parameters are susceptible to subjective interpretation and that information about the level of sedation can only be obtained intermittently. Deeper levels of sedation are difficult to assess using clinical sedation scores. Sufficient information cannot be obtained regarding over-sedation, which is associated with adverse outcomes, prolonged ICU stay and increased costs \[[@B5]-[@B7]\]. Additionally, patients treated with muscle relaxants cannot be evaluated using clinical rating scores. In paediatric ICU patients, neuromuscular blocking agents are used in approximately 6--16% of ventilatory support days \[[@B8]\]. A point of criticism regarding the COMFORT score is that physiological parameters such as haemodynamic indices and heart rate, which contribute to the score (Table [1](#T1){ref-type="table"}), can be influenced by ICU therapy. Therefore, objective tools with which to measure the level of sedation are urgently needed so that over-sedation can be avoided and the level of sedation adjusted if muscle relaxants are given.
With this in mind, data obtained continuously using nonstimulating methods such as bispectral index (BIS) may have benefits in comparison with clinical monitoring of sedation. BIS is a processed electroencephalographic (EEG) parameter that provides a measure of sedative levels on a relative scale \[[@B9]-[@B11]\]. For various agents (e.g. propofol and midazolam) it has been shown that the BIS may correlate with dose-dependent levels of anaesthesia \[[@B10],[@B12]-[@B14]\] and ICU sedation \[[@B15]-[@B18]\]. Level of sedation and changes in memory function correlated well with BIS in volunteers \[[@B19]\]. The BIS monitor was initially designed to measure the level of consciousness in adults during anaesthesia. In paediatric patients age-specific changes in EEG activity could complicate interpretation of BIS measurements \[[@B20]\]. Nevertheless, the first reports of use of BIS in paediatric anaesthesia were promising \[[@B21]-[@B23]\]. Recently, McDermott and coworkers \[[@B24]\] investigated the use of BIS during sedation in children undergoing elective diagnostic or therapeutic procedures. Those investigators found good correlation between BIS and the University of Michigan Sedation Scale. In a paediatric intensive care unit (PICU) setting, four groups found a good to moderate correlation between BIS and clinical scores \[[@B25]-[@B28]\]. Unfortunately, those studies used clinical scores that have not been evaluated in paediatric patients \[[@B25],[@B27]\], the investigators were not sufficiently blinded to the BIS results \[[@B25],[@B26],[@B28]\], or uncontrolled repeated measurements in one patient were included in the study \[[@B25],[@B27],[@B28]\].
The objective of the present study was to evaluate use of the BIS to monitor levels of sedation in paediatric ICU patients in a blinded and prospective manner, and to compare it with use of the COMFORT score. To avoid bias by repeating analysis of data from the same patient, each patient was included only once in the statistical analysis. In addition, we wished to focus on the age of the children and on quality of EEG data.
Methods
=======
Following agreement from the local ethics committee and once informed consent from the legal guardians (parents) had been obtained, we studied 50 paediatric patients admitted to our PICU. All included patients met primary inclusion (age \<18 years, requirement for mechanical ventilation) and exclusion criteria (brain trauma, any kind of end-stage disease, use of muscle relaxants or persistent postoperative relaxation according to train-of-four monitoring, intractable agitation). After enrolment patients were excluded if correct attachment of electrodes was impossible, impedances did not comply with the quality requirements of the manufacturer (\<10 kΩ, signal quality index \>0.8, less than threefold deviation between electrode impedances), or recording could not be repeated at least twice. The sedation and/or analgesic regimens were not controlled in the study and were administered in accordance with standard operating procedures in the PICU.
After PICU admission, the study was started when all examinations on admission were completed and all sources of irritation were eliminated. Electrodes were placed at F7, F8 and Fp (reference), and one electrode was placed beside Fp (ground) in accordance with the international ten-twenty system \[[@B29]\]. Electrode sites were abraded using conventional alcohol swabs. Paediatric silver--silver--chloride self-adhesive ECG electrodes were applied (Blue Sensor Neonatal™ Medicotest, Friedberg, Germany). An Aspect A-1000™ monitor (software version 3.12; Aspect Medical Systems, Natick, MA, USA) was used to compute the BIS. As recommended by the manufacturer, electrode impedances were kept below 10 kΩ, deviations among electrode impedances less than threefold, and signal quality index above 0.8.
Group assignment
----------------
Patients were assigned *post hoc*into groups according to their level of sedation as measured using the COMFORT scale (deep sedation, light sedation, inadequate sedation) and into age groups (≤ 6 months and \> 6 months of age). This age limit was used because in children younger than 6 months old \[[@B30]\] synchronization of the EEG is limited. Synchronization is among the columns of the BIS algorithm \[[@B11]\].
Study protocol
--------------
The study protocol is summarized in Fig. [1](#F1){ref-type="fig"}. After PICU admission, the first recording was started immediately after all necessary manipulations (examinations, laboratory specimens) were completed. If the data were of good quality (see above), then BIS values were sampled for 1 min and the median was calculated. A second investigator, who was blinded to the BIS results, assessed the patient clinically using the COMFORT scale \[[@B3]\]. Every patient was assessed three times, with a minimal interval of 1 hour between measurements. The set with the best impedance values was chosen for calculation of further statistics. To interpret the significance of the data quality, we also calculated the correlation between BIS and COMFORT score using the data couples with the poorest starting impedances. For each patient the following were also recorded: age, sex, medical diagnosis, Pediatric Risk of Mortality III score \[[@B31]\], medications administered, blood gases and temperature.
Statistical analysis
--------------------
Correlations between the BIS and the COMFORT score were calculated for each group using the Spearman\'s rank order correlation coefficient. Because BIS and COMFORT score were classified on an ordinal scale, regression analysis and confidence intervals could not be calculated. Therefore, we calculated the coefficient of determination (r^2^).
The ability of BIS to discriminate between sedation levels as classified using the COMFORT scale was tested using \'receiver operating characteristic\' (ROC) statistics. The cut-off point was determined at the point of greatest sensitivity and specificity for discrimination. A logistic regression model using the BIS was developed to predict sedation levels of the patients (deep versus light sedation) in accordance with the COMFORT classification.
Data are expressed as mean ± standard deviation in the case of a normal distribution and interval scaling level, and as median (range) if the data were not distributed normally or in case of ordinal scaling level. The χ^2^and Fisher\'s exact tests were used to analyze categorical data and Mann--Whitney U-test for data on at least ordinal level. Data analysis was performed using SPSS statistical software (version 8.0; SPSS Inc., Chicago, IL, USA) and Microsoft Excel with Analyze It™ (version 1.48; Analyze-it Software Ltd, Leeds, UK) modification.
Results
=======
A total of 53 paediatric intensive care patients met our primary inclusion criteria. Three patients were excluded because of inability to achieve lead impedances with less than threefold deviance with the first set. In 10 patients recording could not be repeated at least twice because extubation was performed during the period of study. Finally, we enrolled 40 PICU patients into the study. According to COMFORT scoring no child was inadequately sedated, resulting in two COMFORT groups (i.e. deep sedation and light sedation). Patient characteristics are presented in Table [2](#T2){ref-type="table"}. On comparing the sedation groups, we found that the parameters diagnosis (*P*= 0.162), Pediatric Risk of Mortality III score (*P*= 0.891), sex (*P*= 0.770) and age (*P*= 0.716) did not differ significantly between groups. We also found no significant differences in arterial carbon dioxide tension (*P*= 0.750), body temperature (*P*= 0.879) and type of medication (benzodiazepines, *P*\> 0.99; opioids, *P*= 0.650; propofol, *P*= 0.286; ketamine, *P*\> 0.99).
For those data couples with the best impedance levels (3.3 ± 1.8 kΩ, range 0.3--7.8 kΩ), BIS and COMFORT score correlated significantly for all patients (*n*= 40, *P*= 0.001; Spearman\'s rho: r = 0.651, r^2^= 0.42; Fig. [2](#F2){ref-type="fig"}) and for patients without ketamine (n = 38, *P*= 0.001; Spearman\'s rho: r = 0.668; r^2^= 0.45). The correlation between BIS and COMFORT for data couples with the worst impedances (5.1 ± 2.2 kΩ, range 1.9--9.9 kΩ) was poor (*P*= 0.05; Spearman\'s rho: r = 0.387, r^2^= 0.15). All further results reported are for the data couples with the best impedance levels.
Our results showed a significant correlation (*n*= 29, *P*= 0.003; Spearman\'s rho: r = 0.525, r^2^= 0.28) for deeply sedated patients (COMFORT score 8--16) and for patients who had not received ketamine (*n*= 27, *P*= 0.002; Spearman\'s rho: r = 0.565, r^2^= 0.32) whereas no correlation was found in the group with light sedation (COMFORT score 17--26; *n*= 11, *P*= 0.956; Spearman\'s rho: r = 0.019, r^2^\< 0.01).
ROC analysis identified the BIS index level that distinguished best between deep and light sedation (groups classified by COMFORT scores). The calculated cut-off point between the groups was at a BIS of 83, which had a sensitivity of 75.9% and a specificity of 81.8% (area under the curve 0.834, 95% confidence interval 0.699--0.968; Fig. [3](#F3){ref-type="fig"}).
The logistic regression model using BIS as an explanatory variable was able to predict sedation according to COMFORT score correctly for 80% of the children. This total percentage is derived from correct predictions in 90% of deeply sedated patients (COMFORT score \<17, *n*= 29) and in 55% of lightly sedated patients (COMFORT score 17--26, *n*= 11).
In patients younger than 6 months (*n*= 21, range 0.7--5.7 months) BIS and COMFORT score correlated significantly better (*P*\< 0.001 in the younger group versus *P*= 0.041 in the older group; *n*= 19, range 6.3 months--16 years) and with a higher correlation coefficient (Spearman\'s rho: in younger patients r = 0.781, r^2^= 0.61 versus in older children r = 0.473, r^2^= 0.22; Fig. [4](#F4){ref-type="fig"}).
The correlation was slightly better in older children when the two patients who had received ketamine were excluded from the analysis (Spearman\'s rho: *n*= 17, *P*= 0.030; r = 0.527, r^2^= 0.28).
Discussion
==========
Most sedation scoring systems use responses to stimuli \[[@B32]\] and/or patient appearance and physiological variables \[[@B3]\] to estimate the level of sedation. These scores must be interpreted subjectively or, in case of physiological parameters, can be influenced by ICU therapy. In contrast to clinical scoring systems, the BIS system generates information continuously and objectively. The BIS monitor was developed to assess intraoperative depth of anaesthesia and to avoid awareness in adults. Data from studies that were not blinded sufficiently \[[@B25],[@B26],[@B28]\] or that employed scores that have not been evaluated in the PICU setting \[[@B25],[@B27]\] suggested a moderate correlation between BIS and clinical scoring systems.
The aim of this prospective, blinded study was to determine whether BIS is a useful tool for assessing the level of sedation in critically ill paediatric patients. For statistical calculation only one data set (BIS versus COMFORT score) per patient was evaluated. Therefore, bias introduced by including multiple observations from one patient was avoided. In the study we compared BIS with the COMFORT score, which was previously validated in the PICU setting \[[@B3],[@B4]\]. The study indicated that there was a moderate correlation between BIS and corresponding COMFORT scores (r^2^= 0.42). Subanalysis revealed a distinctly better coefficient of determination during deep sedation (r^2^= 0.28) than with light sedation (r^2^\< 0.01) for the assessment tools. This was confirmed by the binary logistic regression findings, which revealed a good ability of BIS to predict deep levels of sedation (90%). According to COMFORT score, BIS could predict this level among lightly sedated children in only 55% of the cases. The overall percentage of correct prediction was 80%. Using ROC analysis we found a BIS value of 83 to distinguish best between light and deep sedation as assessed using the COMFORT score. Probably because of the exclusion of agitated children, we did not observe under-sedation in the study. We were therefore unable to differentiate further between lightly sedated and under-sedated children. This could be interpreted as an investigational bias. Furthermore, poor EEG data quality could cause a large number of movement artifacts, resulting in low signal quality, and we cannot exclude the possibility that movement artifacts contributed to the lower coefficient of determination for lighter sedated children than for children under deeper sedation. This was reported in adult settings \[[@B33],[@B34]\].
Our data support the findings of previous studies \[[@B25]-[@B28]\]. Berkenbosch and colleagues \[[@B25]\] compared BIS with three simultaneously measured clinical sedation scores (Ramsay Sedation Score \[RSS\], Tracheal Suctioning Score, and Pediatric Intensive Care Unit Sedation Score) in paediatric patients (age 5.7 ± 6.1 years, range 1 month--20 years). None of the studied scores was clinically validated for use in paediatric ICU patients. The BIS monitor correlated moderately with clinically assessed sedation levels (r^2^= 0.12, 0.08 and 0.21, respectively). BIS was found to differentiate reliably between adequate and inadequate sedation (cut-off BIS 70), but it was relatively insensitive in differentiating between adequate and over-sedation (cut-off BIS 50). Critical aspects in the study conducted by Berkenbosch and coworkers are that different individuals performed the sedation assessments, which might have resulted in considerable interobserver variability. The nurses assessing the level of sedation were not formally blinded to the BIS results. Multiple sets of data derived from single patients were included in same analysis, which might have influenced the results. Twenty-four patients were included in the study, but measurements were repeated 18 ± 14 times per patient.
Crain and colleagues \[[@B26]\] also used the COMFORT score to estimate the level of sedation. Those investigators studied 31 patients (age 53 ± 11 months, median 25 months; no range presented) and selected each patient\'s lowest and highest BIS measurement for further investigation. The direct coefficient of determination between BIS and COMFORT score was only moderate (r^2^= 0.26), which reflected their finding that some patients exhibited good correlation whereas others did not. After grouping BIS results into four levels of sedation, a high coefficient of determination (r^2^= 0.89) with the COMFORT score resulted. Grouping of BIS values was conducted according to results formerly derived from adult data, but it is not proven whether this procedure is appropriate in children.
Aneja and coworkers \[[@B27]\] studied 24 patients without neuromuscular blockade, comparing BIS with RSS. They found a high and significant coefficient of determination (r^2^= 0.77; age 6.3 ± 2.9 years, range 1--16 years). The calculated cut-off point for distinguishing between over-sedation (RSS 6) and comfortable sedation (RSS 2--5) by ROC analysis was a BIS level of 42. Under-sedation (RSS 1) was found at a BIS level in excess of 76. Another component of that study dealt with patients receiving neuromuscular blockade (age 8.4 ± 3.7 years, range 0.5--19 years). According to BIS, the authors observed a significant number of patients suffering from inadequate sedation, which would not have been detectable by clinical investigation. Limitations of the study were the inclusion of multiple observations per patient and that the RSS has not been validated for use in paediatric patients, as mentioned above.
Courtman and coworkers \[[@B28]\] recently compared BIS and COMFORT score in critically ill children (mean age 3.9 ± 4.5 years). Those investigators found a moderate coefficient of determination (r^2^= 0.26) between BIS and COMFORT score in 25 neurologically normal children and a weak coefficient of determination (r^2^= 0.06) in 15 children who were classified as being neurologically abnormal. In that study BIS could discriminate between light and deep levels of sedation. In conformance with the findings of Berkenbosch and coworkers \[[@B25]\], Courtman and colleagues found that BIS was unable to discriminate between deep and very deep levels of sedation. The significance of this study is also limited by the inclusion of multiple observations per patient.
In our study the quality of EEG impedance appeared to have a major impact on correlation between BIS and COMFORT score. Although within the limits recommended by the manufacturer, we found only a low coefficient of determination (r^2^= 0.15) between COMFORT score and corresponding BIS in case of higher impedances (5.1 ± 2.2 kΩ). This emphasizes the importance of good data quality. On the basis of our experience, we would advise use of impedance values of less than 5 kΩ.
Unexpectedly, correlation between the methods was better in children younger than 6 months. Patient basic data do not explain this finding. The impact of age on BIS is still debated, with divergent findings reported in the anaesthesia literature \[[@B21],[@B35]\]. Davidson and coworkers \[[@B35]\] compared BIS with the corresponding consciousness level during emergence from anaesthesia in a prospective, blinded manner in children (≥1 year old) and infants (\<1 year old) undergoing elective circumcision. BIS increased significantly as sevoflurane concentrations decreased in children, but a similar relationship was not demonstrated in infants. Adult EEG sensors were used in that study for all patients, which could partly account for the difference in findings between that study and ours. There are no additional data in the paediatric ICU literature.
Because different scoring systems are used for clinical estimation of the level of sedation, it is difficult to compare our findings with those of other investigators. The end-points of sedation are neither defined consistently nor comparable between the studies mentioned above. In our study COMFORT scores corresponded to a wide range of BIS values. In other words, the BIS index does not always reflect the expected clinical/subjective level of sedation. This observation is in agreement with the experience of other groups who compared neurophysiological parameters (i.e. BIS, somatosensory evoked magnetic fields, evoked potentials) with clinical sedation scores in adult patients \[[@B36]-[@B38]\] and with the other paediatric ICU studies \[[@B25]-[@B28]\]. This could be related to the fact that neurophysiological parameters and clinical sedation scores measure different attributes. BIS automatically matches certain EEG patterns to clinical states that are found in adult volunteers under sedation or anaesthesia. In contrast, clinical scores are used to summarize the investigator\'s impression of whether the patient is comfortable in the ICU setting. An extreme example would be the patient who is awake with a high BIS score but who appears to be completely comfortable. The limitations of clinical scores in estimating the level of sedation have been discussed broadly \[[@B32]\]. In addition to these limitations, in the case of the COMFORT score there is a specific limitation caused by the inclusion of physiological data (heart rate and blood pressure). When physiological parameters are used in a heterogeneous population such as paediatric ICU patients, it is difficult to define reference data. If changes in physiological parameters caused by sedatives are to be interpreted, then a standardized sedation regimen and co-medication are needed. In our opinion this is not the case in the ICU setting. The majority of children included in the present study were postoperative cardiac surgical patients. Most of them had reduced cardiac function and required catecholamines. In this group of patients in particular, haemodynamic variability could not be attributed solely to sedative drugs.
The COMFORT score cannot distinguish between very deep stages of sedation, whereas with neurophysiological methods such as the BIS this might be possible. EEG measurements often fail to provide correct values with very light or absent sedation because motor activity and skeletal muscle tone increase. However, this situation will never fall within the domain of electrophysiological monitoring because clinical estimation will be sufficient in unsedated or only lightly sedated patients in the ICU setting.
Data from the investigations cited above and our data suggest that the BIS monitor might be useful in the case of moderate to deep levels of sedation. Our group found that BIS levels below 83 had good correlation with clinical estimation. Berkenbosch and coworkers \[[@B25]\] suggested that a moderate level of sedation could be achieved at BIS values between 50 and 70, and deep sedation at levels below 50. This corresponds well with the data presented by Aneja and coworkers \[[@B27]\], who found that a BIS above 76 indicated inadequate, light sedation and that a BIS below 46 marked very deep sedation or over-sedation. Whether BIS can detect over-sedation remains controversial. Berkenbosch and coworkers \[[@B25]\] and Courtman and colleagues \[[@B28]\] found that BIS could not discriminate between deep and very deep levels of sedation. In our opinion, this finding could indicate that clinical scores are not useful for identifying very deep levels of sedation.
Few data exist for paralyzed PICU patients. In this situation clinical scores are not applicable because they need muscular activity to be present to rate the level of sedation. Aneja and coworkers \[[@B27]\] stated that the RSS and bedside nurse assessments are inadequate for monitoring the depth of sedation in paralyzed children, and concluded that BIS is a useful adjunct for assessing sedation in paralyzed patients. In our opinion use of electrophysiological monitoring tools such as the BIS is imperative in paralyzed patients to prevent levels of sedation that are too light or too deep.
Furthermore, the terms \'under-sedation\' and \'over-sedation\' should be used carefully because the level of sedation required depends on the individual patient\'s needs. Bearing this in mind, clinician should adapt the level of sedation to the demand of the patient. When deeper levels of sedation are needed, BIS can help by avoiding undesired levels of sedation. In situations when muscular paralysis is necessary, BIS could make a valuable contribution because established clinical tools fail to measure the level of sedation in this setting.
Conclusion
==========
Our data indicate that an impedance level below 5 kΩ is needed for valid interpretation of BIS values in PICUs. At such impedance levels the BIS index correlates well with the COMFORT sedation score, in particular in children with deeper levels of sedation. Discrimination between light and moderate sedation with high sensitivity and specificity was possible at a BIS level of 83. The BIS monitor provides continuous measurements of the level of sedation without having an impact on patient comfort. It may be a useful adjunct in clinical routine and may be especially helpful in certain situations when clinical estimation fails (e.g. if muscular paralysis is necessary).
Key messages
============
• In daily clinical practice of paediatric ICU therapy, the BIS monitor provides continuous measurements of sedation level without having an impact on the patient\'s comfort.
• In certain situations when clinical estimations fail (i.e. if muscular paralysis is necessary), electrophysiological monitoring tools such as BIS should be considered imperative to prevent inadequately light or inadequately deep sedation.
• In the case of deep sedation, BIS correlated satisfactory with the COMFORT score results if low EEG impedances were guaranteed.
Abbreviations
=============
BIS = bispectral index; EEG = electroencephalography; ICU = intensive care unit; PICU = paediatric intensive care unit; ROC = receiver operating characteristic; RSS = Ramsay Sedation Score.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
All authors contributed to the design, conduct, analysis and interpretation of the research reported. CS and WK were principal investigators and led the conceptual design of the study. MS and GN assisted with data collection and analysis, and with manuscript preparation.
Acknowledgements
================
We would like to thank Prof. Gaedicke (Department of Pediatrics, Campus Charité Mitte Universitätsmedizin Berlin) for his help with the preparation of the manuscript.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Study design. BIS, bispectral index; PICU, paediatric intensive care unit.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Correlation of bispectral index (BIS) and COMFORT score at different levels of sedation.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Receiver operating characteristic (ROC) analysis: cut-off point between light and deep sedation. BIS, bispectral index.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Correlation of bispectral index (BIS) and COMFORT score in patients younger and older than 6 months.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
The COMFORT scale
:::
Alertness Calmness/agitation Respirator response Physical movement Blood pressure (MAP) Heart rate Muscle tone Facial expression Points
----------------------- -------------------- ------------------------------------------------------------------- --------------------------------------------- ----------------------------------------------------- --------------------------------------------- --------------------------------------------------------- -------------------------------------------- --------
Deeply asleep Calm No coughing and no spontaneous respiration No movement Below baseline Below baseline Totally relaxed; no tone Totally relaxed 1
Lightly asleep Slightly anxious Spontaneous respiration with little or no response to ventilation Occasional, slight movement Consistently at baseline Consistently at baseline Reduced Normal; no facial tension evident 2
Drowsy Anxious Occasional cough or resistance to ventilator Frequent, slight movement Infrequent elevations of 15% or more (1--3/observ.) Infrequent elevations of 15% or more (1--3) Normal Tension evident in some facial muscles 3
Fully awake and alert Very anxious Actively breathes against respirator or coughs regularly Vigorous movement limited to extremities Frequent elevations of 15% or more (\>3/observ.) Frequent elevations of 15% or more (\>3) Increase tone and flexion of fingers and toes Tension evidence throughout facial muscles 4
Hyper alert Panicky Fights ventilator, coughing or choking Vigorous movement, including torso and head Sustained elevation ≥15% Sustained elevation ≥15% Extreme muscle rigidity and flexion of fingers and toes Facial muscles contorting and gromacing 5
MAP, mean arterial pressure. Data from Ambuel and coworkers \[3\].
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Patient characteristics
:::
Characteristic/parameter All patients Deeply sedated Lightly sedated *P*
----------------------------- -------------------------------- ----------------------------------- -------------------------------- ---------
Number 40 29 11 \-
Age (median \[range\]) 5.6 months (21 days--16 years) 5.7 months (1.5 months--13 years) 5.1 months (21 days--16 years) 0.716
Sex (*n*\[%\])
Male 21 (52) 16 (55) 5 (45) 0.770
Female 19 (47) 13 (45) 6 (55)
Diagnosis (*n*\[%\])
Cardiac 34 (85) 23 (79) 11 (100) 0.162
Gastrointestinal 2 (5) 2 (7) 0 (0)
Other 4 (10) 4 (14) 0 (0)
PRISM III score (mean ± SD) 7.22 ± 5.29 7.34 ± 5.78 6.91 ± 3.96 0.891
Medication (*n*\[%\])
Benzodiazepines 34 (85) 25 (86) 9 (82) ≥ 0.999
Opioids 33 (83) 23 (79) 10 (91) 0.650
Propofol 13 (32) 11 (38) 2 (18) 0.286
Ketamine 2 (5) 2 (7) 0 (0) ≥ 0.999
PaCO~2~(kPa; mean ± SD) 5.1 ± 0.8 5.1 ± 0.9 4.9 ± 0.8 0.750
Temperature (°C; mean ± SD) 37.4 ± 0.7 37.4 ± 0.6 37.4 ± 0.7 0.879
*P*values are given for the comparison of deeply sedated versus lightly sedated. PRISM, Pediatric Risk of Mortality III score, PaCO~2~, arterial carbon dioxide tension; SD, standard deviation;
:::
|
PubMed Central
|
2024-06-05T03:55:54.882349
|
2004-11-10
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065097/",
"journal": "Crit Care. 2005 Nov 10; 9(1):R9-R17",
"authors": [
{
"first": "Andreas E",
"last": "Triltsch"
},
{
"first": "Grit",
"last": "Nestmann"
},
{
"first": "Helmut",
"last": "Orawa"
},
{
"first": "Maryam",
"last": "Moshirzadeh"
},
{
"first": "Michael",
"last": "Sander"
},
{
"first": "Joachim",
"last": "Große"
},
{
"first": "Arka",
"last": "Genähr"
},
{
"first": "Wolfgang",
"last": "Konertz"
},
{
"first": "Claudia D",
"last": "Spies"
}
]
}
|
PMC1065100
|
See related commentary <http://ccforum.com/content/9/1/27>
Introduction
============
Ischemia--reperfusion injury in lung tissue is a common problem in medical practice, with sometimes severe consequences such as acute respiratory distress syndrome (ARDS) and a high mortality rate for the patient. Some causes of warm ischemia--reperfusion injury are cardiopulmonary bypass during cardiac surgery and pulmonary sleeve resection. In contrast, lung transplantation is the main example of partial cold ischemia--reperfusion injury.
Neutrophils are known to be one of the cell types responsible for tissue damage in many ways. First, they are able to deliver toxic radicals that damage pulmonary endothelium directly or indirectly by activating caspase-3, which results in apoptosis \[[@B1],[@B2]\]. Second, they can damage pulmonary endothelium and parenchyma by delivering elastase and other proteases \[[@B3]\]. Third, the cell membrane of activated neutrophils becomes rigid and adhesion between neutrophils and endothelial adhesion molecules occurs, resulting in sequestration and a \'no-reflow phenomenon\' \[[@B4],[@B5]\].
The role of neutrophils in pulmonary ischemia--reperfusion injury has also been investigated in experiments in which neutrophil depletion was induced and by the inhibition of tissue infiltration. The role of the neutrophil is currently still controversial \[[@B6]-[@B9]\].
The role of macrophages has been investigated in several transplantation models \[[@B3],[@B10],[@B11]\]. Eppinger and colleagues have specified chemical mediators of reperfusion injury by using antibodies against cytokines. Although some mediators seemed to be required during the early phase of ischemia--reperfusion injury, only tumor necrosis factor-α (TNF-α) is involved in the evolution of late ischemia--reperfusion injury. These cytokines are released from activated macrophages probably as a result of acute lung reperfusion \[[@B10]\]. These results suggest a role for macrophages in the early reperfusion phase and a role for activated and recruited neutrophils in the late reperfusion phase \[[@B3]\]. Currently, the role of lymphocytes in ischemia--reperfusion injury remains unclear. Qayumi and colleagues concluded that upregulation of MHC II on peripheral lymphocytes is related to the degree of damage caused by ischemia--reperfusion \[[@B12]\].
Apoptosis, necrosis and alveolar oedema, representing alveolar permeability, are morphological changes of ischemia--reperfusion-induced lung injury. Fischer and colleagues were the first to describe apoptosis of specifically type II alveolar pneumocytes resulting from pulmonary ischemia--reperfusion in a human lung transplantation study \[[@B13]\].
In summary, little is known about the role of neutrophils, T cells and macrophages in ischemia--reperfusion injury. In preparation for studies investigating the specific role of infiltrating cells, the aim of this study was to specify the type of infiltrating cells and their sequence after 1 hour of warm ischemia followed by 30 min to 4 hours of reperfusion in a model of acute lung injury, which was defined by quantifying apoptosis and alveolar oedema.
Materials and methods
=====================
Animals
-------
Male inbred Wistar rats (mean weight 225 g), obtained from Iffa Credo (Brussels, Belgium), were used for all experiments. Animals were treated in accordance with the Animal Welfare Act and the *Guide for the Care and Use of Laboratory Animals*(NIH Publication 86-23, revised 1985). The rats were transported in sterile conditions, housed in suspended mesh-wired cages and fed *ad libitum*with a standard pellet diet (standard rat chow; Hope Farms, Woerden, The Netherlands). The Ethical Committee of the University of Antwerp approved the experimental protocols.
Study design
------------
Fifty rats were randomized into eight groups. Five groups underwent 1 hour of warm lung ischemia followed by 30 min, 1, 2, 3 and 4 hours of reperfusion, respectively (*n*= 7 in each group). One sham group underwent the identical surgical procedure without ischemia--reperfusion (*n*= 4). To find out whether adhesion of the inflammatory cells had occurred, the lungs in one extra group were flushed with 6% buffered hetastarch after 1 hour of ischemia and 4 hours of isolated lung perfusion (*n*= 7) \[[@B14]\]. This group was compared with a sham group, which was also flushed (*n*= 4) (Fig. [1](#F1){ref-type="fig"}).
Induction of ischemia--reperfusion
----------------------------------
Anesthesia was induced by 4% isoflurane in a mixture of oxygen (O~2~) and nitrous oxide (N~2~O) in a ratio of 1:3 for 4 min. Intubation was performed with a 16-gauge Insyte-W catheter using translaryngeal illumination in accordance with the technique described by Hendriks \[[@B14]\]. After the rats had been connected to the ventilator, the N~2~O : O~2~ratio was set to 1:1 and the concentration of isoflurane was titrated to 0.5--1.5% according to muscle relaxation, heart rate and pupil size. To prevent thrombosis in lung vasculature during ischemia, 100 IU/kg heparin was infused into the left femoral vein 5 min before the left lung hilum was clamped. After left posterolateral thoracotomy through the fourth intercostal space, a rib retractor was placed to luxate the left lung anteriorly.
Ischemia was induced by clamping the left lung hilum with two occluding curved microvascular clamps (Kleinert-Kurz WK65145) without further dissection. One clamp was placed in a cranial--caudal direction and the other clamp was placed laterally in the opposite direction. In a separate experiment, four rats received intravenous and bronchial injection of methylene blue solution to test the vascular and bronchial occlusion obtained by the microvascular clamps. Complete vascular and bronchial occlusion was achieved. To simulate physiological circumstances, the thoracotomy incision was closed in layers after the introduction of a 16-gauge catheter connected to a 50 ml syringe into the left chest cavity.
When animals recovered, the chest tube and endotracheal tubes were removed. Ten minutes before reperfusion, anesthesia was induced and rats underwent a left thoracotomy with the use of the same incision as described above. Reperfusion occurred on removal of the clamps. The left thoracotomy was closed as described above. Ten minutes before the end of reperfusion time anesthesia was induced and the rat underwent a left thoracotomy for the third time followed by left pneumonectomy. Ten seconds before the rat was killed, maximal inflation of the left lung was achieved by occlusion of the expiratory ventilation cannula for 3 s to prevent inter-animal variation of inflation of the left lung. After reperfusion all rats underwent intramuscular injection of tramadol for pain control.
To prevent cooling, rats were placed on a warm-water pad during the operation and under a heating light during both ischemia and reperfusion. Rectal temperature was measured before clamping of the left lung hilum and before killing and was held constantly between 36.8 and 37.4°C.
Rats in the sham group underwent an identical surgical procedure except for clamping the left lung hilum. Rats in this group were killed 1 h after anterior luxation of the left lung.
Flush procedure
---------------
To study cellular adhesion to the endothelium, lungs of one more group were flushed after 4 hours of isolated lung perfusion with buffered starch. This procedure has been extensively described previously \[[@B15],[@B16]\]. In brief, after ischemia--reperfusion, the pulmonary artery and vein were clamped with curved microvascular clamps. A 16-gauge angiocatheter was placed through the chest wall. A PE-10 perfusion catheter (Clay Adams, Parsippany, NJ, USA) was introduced into the chest through the angiocatheter and secured by a 4/0 silk suture after insertion into the pulmonary artery. Perfusate (6% buffered starch) was delivered through this catheter for 4 min at 0.5 ml/min. In addition, a pulmonary venotomy was performed to discard the venous effluent.
Killing and tissue storage
--------------------------
At killing, the left lung was taken out of the rat and cut caudal--cranially into four pieces. The lateral sample was fixed in metacarn for 4 hours at room temperature (23°C) and stored in 70% ethanol at 4°C. Directly after killing, the weight of the medial sample was measured and the sample was put into an oven at 65°C for 5 days to assess the wet : dry ratio as a parameter for lung oedema. The middle samples were fixed in chloroform calcium for 90 min at room temperature and then stored in buffer (10 ml of distilled water, 1 g of CaCl~2~, 0.121 M cacodylate) at 4°C until further processing. Killing was performed by a cut down of the superior caval vein.
Sample processing
-----------------
Tissue samples for light-microscopic investigations were dehydrated with propan-2-ol, cleared with toluene and embedded in paraffin wax. Sections 4 μm thick were stained with hematoxylin/eosin stain (H&E) for neutrophil count. Immunohistochemistry was applied for macrophage and T cell visualization. After deparaffination, endogenous peroxidase was blocked by incubation in 0.9% H~2~O~2~for 15 min. The sections were incubated overnight with CD-3-specific antibody 1F4 (Pharmingen, Becton Dickinson, Erembodegem, Belgium) or with antibody ED-1 (Serotec, Diagnostic Products Cooperation, Humbeek, Belgium) directed against lysosomal membrane glycoprotein on macrophages. Incubation for 30 min with secondary biotinylated horse anti-mouse antibody (Vector, Burlingame, CA, USA) was followed by incubation for 1 hour with peroxidase-labeled avidin--biotin complex (Vector). The slides were developed in 3,3-diaminobenzidine with 0.03% H~2~O~2~or 3-amino-9-ethylcarbazole (AEC) with 0.006% H~2~O~2~for 30 min. Finally, counterstaining was performed in methyl green and Haemaluin Carazzi to reveal T cells and macrophages, respectively.
Light-microscopy investigation
------------------------------
All slides were evaluated in random order. The first field was chosen at random and the next fields in accordance with a standard pattern. Neutrophils were counted in 20 fields per slide (0.95 mm^2^per slide, magnification ×500). Macrophages were counted in 30 fields per slide (5.65 mm^2^per slide, magnification ×250). T cells were counted in 20 fields per slide (1.54 mm^2^per slide, magnification ×400). Apoptosis was determined by terminal deoxynucleotidyl transferase-mediated (TdT) dUTP nick end labelling (TUNEL) staining. Deparaffinization was performed as described above. After decalcification with 3% citrate dissolved for 1 hour at 37°C, sections were incubated with TdT (Roche, Brussels, Belgium) in combination with fluorescein isothiocyanate-labelled dUTP nucleotides (AP Biotech, Roosendaal, The Netherlands) for 1 hour at room temperature. Furthermore, incubation with anti-fluorescein isothiocyanate (Dako, Glostrup, Denmark) peroxidase was performed followed by subsequent washes and the specimens were stained in AEC and counterstained with Haemaluin Carazzi. Only cells with TUNEL-positive nuclear and no cytoplasmic staining were considered to be apoptotic. Cells containing positive cytoplasmic staining were not counted. TUNEL-positive fragments closely ordered in a group were defined as apoptotic bodies. Apoptotic bodies and cells were both counted in 20 fields per slide (0.23 mm^2^per slide, magnification ×800). The occurrence of necrosis was investigated in H&E by a pathologist (EvM) who did not have any knowledge of details of the study. Oedema was twice assessed blindly at H&E and was graded, ranging from mild, moderate to severe. Mild oedema was defined as no to slight exudation within the alveolar space (Fig. [2a](#F2){ref-type="fig"}). Severe oedema was defined as easily recognizable full exudation in the alveolar space (Fig. [2b](#F2){ref-type="fig"}); moderate oedema was defined as being between mild and severe.
Statistics
----------
All statistics were performed with SPSS 9.0 for Windows. Cellular infiltrates and apoptosis were evaluated statistically with the Kolmogorov--Smirnov test to confirm normal distribution. Analysis of variance and Student\'s *t*-test were applied to compare data obtained from the different reperfusion periods with the sham groups. Graded oedema frequencies were analyzed with the χ^2^test by comparison of the reperfusion groups with the sham groups. Statistical significance was accepted at *P*\< 0.05.
Results
=======
Cellular infiltrations
----------------------
### Neutrophils (H&E, magnification ×500)
A significant increase in neutrophils was observed after 30 min to 4 hours reperfusion compared with the sham group (*P*\< 0.01) (Fig. [3](#F3){ref-type="fig"}). After 4 hours of reperfusion followed by flushing, significantly more neutrophils were counted than in the flushed sham group (*P*= 0.003), whereas no significant difference was observed compared with 4 hours of reperfusion without flushing (*P*= 0.10).
### Macrophages (ED-1, magnification ×250)
Significantly more macrophages were counted after 30 min of reperfusion (*P*= 0.0002), 1 hour (*P*= 0.004) and 2 hours (*P*= 0.007) of reperfusion compared with the sham group (Fig. [4](#F4){ref-type="fig"}). A significant decrease was observed after 1 hour of reperfusion compared with 30 min of reperfusion (*P*= 0.01). After 3 hours (*P*= 0.06) and 4 hours (*P*= 0.61) of reperfusion no significant increase in macrophages was observed compared with the sham group.
### T cells (1F4, magnification ×400)
A fourfold increase of T cells was observed after 30 min of reperfusion (*P*= 0.0002) compared with the sham group (Fig. [5](#F5){ref-type="fig"}). This increase was also significant after 1 hour of reperfusion (*P*= 0.004). From 2 hours to 4 hours no significant increase was observed.
Injury evaluation
-----------------
### Apoptosis (TUNEL, magnification ×800) and necrosis (H&E)
Significantly more apoptotic cells were seen after 1 hour (*P*= 0.03), 2 hours (*P*= 0.01), 3 hours (*P*= 0.04) and 4 hours (*P*= 0.00004) of reperfusion (Fig. [6](#F6){ref-type="fig"}). The number of apoptotic bodies was significantly higher after 4 hours of reperfusion (*P*= 0.0006).
Necrosis was not observed in any group.
### Oedema (H&E)
Histological examination showed significantly more alveolar oedema after 30 min, 2, 3 and 4 hours of reperfusion (*P*\< 0.0001) compared with the sham group (Fig. [7a](#F7){ref-type="fig"}). However, after 1 hour of reperfusion, oedema was not significantly increased compared with the sham group. The wet : dry ratio was significantly increased in all groups (30 min, *P*\< 0.05; 2 hours, *P*\< 0.01; 3 hours, *P*\< 0.001; 4 hours, *P*\< 0.01) except for 1 hour of reperfusion (Fig. [7b](#F7){ref-type="fig"}).
Discussion
==========
In this study a significant increase in neutrophils was observed after 1 hour of warm ischemia followed by 30 min to 4 hours of reperfusion. A first peak was shown after 30 min of reperfusion and a second peak after 3 hours of reperfusion. Furthermore, after 4 hours of reperfusion, significantly more neutrophils were observed after pulmonary artery flushing than in the flushed sham group. This resulted in flushing of cells that did not adhere to the endothelium. These results suggest activation and adhesion of neutrophils to the endothelium. Our observations are partly in contrast with results of Eppinger and colleagues, who showed a bimodal pattern of lung injury after 90 min of warm ischemia, with a first peak after 30 min of reperfusion and a second peak after 4 hours of reperfusion \[[@B17]\]. In their report, myeloperoxidase activity, representing neutrophil sequestration, diminished during the reperfusion time course. Neutrophil depletion did not have a protective effect on microvascular permeability after 30 min of reperfusion but the authors did show a protective effect after 4 hours, suggesting an early neutrophil-independent phase and a late neutrophil-dependent phase \[[@B17]\]. The observation of late neutrophil-dependent lung injury is indirectly related to our observation that significantly more neutrophils were counted after flushing of non-adhesive cells, suggesting activation of these cells.
The role of macrophages has been investigated only in transplantation models \[[@B3],[@B10],[@B11]\]. Our data show significantly more macrophages after 30 min to 2 hours of reperfusion, which is in accordance with data from Eppinger. Using the permeability index Eppinger showed an attenuation of reperfusion injury using antibodies against monocyte chemoattractant protein-1, TNF-α and interferon-γ, suggesting that reduced early reperfusion injury is probably due to suppression of macrophage function \[[@B10]\]. A recent report by Maxey and colleagues confirmed the central role of macrophages in early reperfusion injury. They demonstrated significantly less lung injury in TNF-α-deficient mice after 1 hour of ischemia and 1 hour of reperfusion, suggesting that TNF-α is a key initiating factor in acute lung injury \[[@B18]\].
Fiser has made a distinction between the role of donor macrophages on the one hand and the role of recipient macrophages on the other. Activation of donor macrophages could be the initial consequence of ischemia and early reperfusion. In reaction to activation, donor macrophages deliver cytokines, chemotactic agents and proteolytic enzymes responsible for early reperfusion injury \[[@B3],[@B11]\]. Subsequently, early lung injury activates the inflammatory mechanisms of the recipient \[[@B10]\].
Beside augmentation of neutrophils and macrophages, our study also showed a fourfold (*P*= 0.0002) increase in T cells after 30 min to 1 hour of reperfusion, followed by a rapid attenuation. Because of the short duration of reperfusion it is unlikely that local proliferation of lymphocytes occurred, suggesting that chemotaxis is responsible for these observations. However, it is not clear that activation of these cells happened because of the rapid attenuation after 2 hours of reperfusion. This finding implies that the early augmentation of lymphocytes is just a non-specific inflammatory reaction on early reperfusion injury.
The role of T cells was investigated recently in a model of mouse lung perfusion with fresh blood \[[@B19]\]. The interaction between allogenic blood lymphocytes and vascular endothelial cells is correlated with high expression of mRNA of both adhesion molecules and TNF-α in the perfused lung, suggesting that antigen-dependent activation of lymphocytes had occurred \[[@B19]\].
To our knowledge the present study is the first to show apoptosis in the absence of necrosis in lung tissue after warm ischemia--reperfusion. An explanation for the absence of necrosis after 4 hours of reperfusion might derive from the length of reperfusion. Experiments with longer reperfusion periods will be necessary to confirm this hypothesis.
The number of apoptotic bodies is significantly increased after 1--4 hours of reperfusion, whereas the number of apoptotic cells is significantly increased after 4 hours of reperfusion. The tendency of apoptosis to increase is in accordance with observations of Fischer and colleagues in a human transplantation study with 1--5 hours of cold ischemia that showed significant increases in the number of apoptotic cells after reperfusion, in a time-dependent manner \[[@B13]\]. In particular, alveolar type II pneumocytes seemed to be apoptotic \[[@B13]\]. Stammberger and colleagues reported a peak of apoptotic cells after 18 hours of cold ischemia and 2 hours of reperfusion followed by a quick decrease in apoptotic cells as a function of reperfusion time \[[@B20]\]. The rapid attenuation of apoptotic cells is probably due to the occurrence of apoptosis after 6--12 hours of preservation and especially necrosis after 18--24 hours of preservation as described by Fischer and colleagues \[[@B21]\]. Furthermore, an inverse correlation of the occurrence of necrosis with oxygenation was shown, implying the necessity of preventing necrosis \[[@B21]\].
This study showed an identical pattern of alveolar oedema in a function of time by using a histological examination (H&E) and assessment by wet : dry ratio. An important increase of alveolar oedema was observed after 30 min, 2, 3 and 4 hours of reperfusion. We do not have an explanation for the absence of significant oedema after 1 hour of reperfusion. However, a bimodal pattern of lung injury reported by Eppinger and colleagues \[[@B17]\] is confirmed by our results. Using the vascular permeability of ^125^I-labeled bovine serum albumin, Eppinger and colleagues showed an increased presence of serum albumin in bronchoalveolar lavage after 90 min of warm ischemia followed by a first peak after 30 min of reperfusion and a second peak after 4 hours of reperfusion, indicative of damage to the normal vascular/airway barrier \[[@B17]\].
Pulmonary ischemia results histologically in alveolar oedema due to changing permeability at the blood/air barrier after only 30 min of reperfusion. Apoptotic cells appear after 4 hours of reperfusion in a warm model of ischemia--reperfusion and after 6--9 hours of reperfusion in a transplantation model, whereas necrosis is observed after 18--24 hours of reperfusion related to an inverse correlation with oxygenation \[[@B21]\]. It may be noticed that these observations are related to a clinical feature known as ARDS. Clinical ARDS is characterized by acute hypoxemic respiratory failure due to non-cardiogenic pulmonary oedema caused by increased permeability of the alveolar capillary barrier, resulting in mortality ranging from 35% to 44% \[[@B22]\]. On the basis of the results of this study, research has to be focused on how cellular infiltrates are involved in the occurrence of ARDS and in what manner intervention might diminish the damaging effect of pulmonary ischemia--reperfusion.
Conclusion
==========
This study has shown a significant increase in neutrophils after 30 min to 4 hours of reperfusion as well as after reperfusion followed by flushing. Macrophages doubled in number in lung tissue after ischemia--reperfusion. A fourfold increase in T cells in lung tissue after 1 hour of warm ischemia and 30 min of reperfusion was observed. Furthermore, apoptosis in the total absence of necrosis was shown together with important alveolar oedema.
Key messages
============
• Significant early increase of T-cells macrophages and neutrophils after 1 hour of ischemia and 4 hours of reperfusion
• Significant late increase of neutrophils after 1 hour of ischemia and 4 hours of reperfusion.
• Significant apoptosis and lung oedema in the absence of necrosis after 1 hour of ischemia and 4 hours of reperfusion.
Abbreviations
=============
AEC = 3-amino-9-ethylcarbazole; ARDS = acute respiratory distress syndrome; H&E = hematoxylin/eosin stain; TNF = tumor necrosis factor; TUNEL = TdT-mediated dUTP nick end labelling.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
BVP and JH performed all surgical procedures under the supervision of PVS. BVP and VP performed histological analyses of the lung specimens under the supervision of EvM and MDB. VP also performed statistical analyses. BVP drafted the manuscript and was advised by JK. All authors read and approved the final manuscript.
Acknowledgements
================
We thank S Dauwe for processing and staining all the tissue samples, D De Weerdt for layout assistance and A Van Laer for technical assistance during all experiments.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Experimental setting.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Mild **(a)**and severe **(b)**alveolar oedema after 1 hour of warm pulmonary ischemia followed by 4 hours of reperfusion; hematoxylin/eosin stain.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Neutrophil infiltration after 1 hour of warm pulmonary ischemia followed by 30 min to 4 hours of reperfusion. Results are expressed as neutrophils/mm^2^and are means ± SD. \**P*\< 0.01; \*\**P*\< 0.001.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Macrophage infiltration after 1 hour of warm pulmonary ischemia followed by 30 min to 4 hours of reperfusion. Results are expressed as macrophages/mm^2^and are means ± SD. \**P*\< 0.01; \*\**P*\< 0.001.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
T cell infiltration after 1 hour of warm pulmonary ischemia followed by 30 min to 4 hours of reperfusion. Results are expressed as T cells/mm^2^and are means ± SD. \**P*\< 0.01; \*\**P*\< 0.001.
:::
:::
::: {#F6 .fig}
Figure 6
::: {.caption}
######
Apoptotic cells and bodies after 1 hour of warm pulmonary ischemia followed by 30 min to 4 hours of reperfusion. Results are expressed as apoptotic cells and bodies/mm^2^and are means ± SD. \**P*\< 0.05; \*\**P*\< 0.001.
:::
:::
::: {#F7 .fig}
Figure 7
::: {.caption}
######
Alveolar oedema after 1 hour of warm pulmonary ischemia followed by 30 min to 4 hours of reperfusion.**(a)**Histological assessment of alveolar oedema in H&E. **(b)**Wet : dry ratio. \**P*\< 0.05; \*\**P*\< 0.01; \*\*\**P*\< 0.001.
:::
:::
|
PubMed Central
|
2024-06-05T03:55:54.885000
|
2004-11-10
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065100/",
"journal": "Crit Care. 2005 Nov 10; 9(1):R1-R8",
"authors": [
{
"first": "Bart P",
"last": "Van Putte"
},
{
"first": "Jozef",
"last": "Kesecioglu"
},
{
"first": "Jeroen MH",
"last": "Hendriks"
},
{
"first": "Veerle P",
"last": "Persy"
},
{
"first": "Erik",
"last": "van Marck"
},
{
"first": "Paul EY",
"last": "Van Schil"
},
{
"first": "Marc E",
"last": "De Broe"
}
]
}
|
PMC1065104
|
Introduction
============
Patients with severe sepsis often manifest symptoms of encephalopathy. Acute alterations in mental status, which occur fairly frequently in septic patients, have been shown to be associated with poor prognosis \[[@B1]\]. However, not much is known about the exact mechanism of brain injury in sepsis. Studies have suggested that septic encephalopathy might involve a disturbance of plasma and brain neutral amino acid transport across the blood--brain barrier (BBB), similar to those seen in porto-systemic encephalopathy. This process has been related to the breakdown of the BBB because patients with septic encephalopathy have high protein levels in the cerebrospinal fluid \[[@B2]\]. Recently, derangements in the BBB causing perivascular edema have been demonstrated in sepsis-induced pigs \[[@B3]\].
Protective effects of magnesium sulfate (MgSO~4~) against BBB breakdown after severe insulin-induced hypoglycemia have been reported in animals \[[@B4]\]. Similar effects of magnesium on BBB were also evident in a diffuse traumatic brain injury model in rats \[[@B5]-[@B7]\].
In summary, MgSO~4~was shown to have a protective effect on BBB integrity in multiple experimental models. We hypothesized that MgSO~4~will also protect against BBB derangements observed in sepsis and tested the hypothesis in a rat model of sepsis induced by an intraperitoneally inserted infected fibrin--thrombin clot.
Methods
=======
One hundred and twenty-six male Sprague--Dawley rats weighing 320--440 g were used in this study. Rats were purchased from the Institute for Experimental Research and Application (Istanbul Medical Faculty), and were cared for before and during all stages of the experimental protocol in compliance with the applicable institutional guidelines and regulations of the Institute for Experimental Medicine Research and Application.
Rats were prepared for surgery under anesthesia with intramuscular 100 μg/g ketamine (Parke-Davis, Morris Plains, NJ, USA) and 20 μg/g xylazine hydrochloride Rompun 2% (Bayer, Munich, Germany) and allowed to breathe spontaneously. The loss of corneal reflex and no movement in response to a painful stimulus confirmed maintenance of adequate anesthesia for the experimental procedure. The rats were subsequently randomized into one of four groups: sham control (C), sham control MgSO~4~-treated (C-Mg), septic (S) and septic with MgSO~4~(S-Mg).
Intraperitoneal sepsis was induced with the infected fibrin--thrombin clot model described by Mathiak and colleagues \[[@B8]\]. Fibrin--thrombin clots were formed by adding 2 ml of 1% sterile fibrinogen solution, 1 ml of a bacterial suspension (1.8 × 10^9^colony-forming units/ml \[infected\] or vehicle \[sterile 0.9% NaCl\]) and 160 μl (100 units/ml) of sterile human thrombin to a 5 ml syringe. The resulting clot was then incubated at room temperature for 30 min before implantation into the abdominal cavity. The *Escherichia coli*strain was isolated from an intra-abdominal collection from a patient with secondary peritonitis. The bacteria were inoculated into a brain heart infusion broth (DIFCO Laboratories, Detroit, MI, USA) and incubated overnight at 35°C. The count of *E. coli*was adjusted to 1.8 × 10^9^colony-forming units/ml with McFarland standard 6. After making a 0.5 cm midline abdominal incision, the peritoneum was opened and the prepared clot was injected into the peritoneal cavity directly from the syringe. Sham-operated rats had a sterile clot injected into their peritoneal cavity. To examine the effect of magnesium in septic and sham-operated rats, a dose of 750 μmol/kg MgSO~4~was given intramuscularly immediately after surgery. Control groups for both infected and sham-operated rats were injected with an equal volume of saline.
After surgery, the animals were given 50 μl/g per hour of saline subcutaneously and were allowed to wake up while breathing spontaneously. They were returned to their cages and were allowed free access to water. Those rats surviving for 24 hours after the surgery were anesthetized and decapitated for the investigation of brain tissue specific gravity (SG) and BBB integrity.
We used the method described by Mikawa and colleagues \[[@B9]\] to determine BBB integrity by Evans blue (EB) dye. EB dye (4 ml/kg, 2%) was administered intravenously and allowed to circulate for 60 min. The animals were then perfused with saline through the left ventricle at a pressure of 110 mmHg until colorless fluid was obtained from the right atrium. Afterwards, the brains were removed and dissected. Each hemisphere was weighed and the samples were then homogenized in 3.5 ml phosphate-buffered saline and vortex-mixed for 2 min after the addition of 2.5 ml of 60% trichloroacetic acid to precipitate protein. The samples were then cooled for 30 min and centrifuged for 30 min at 1000 r.p.m. The absorbance of the supernatants for EB dye was measured at 610 nm with a spectrophotometer. EB dye content is expressed as μg/mg of brain tissue against a standard curve.
The method defined by Marmarou and colleagues was used for the determination of SG \[[@B10]\]. We obtained 1 mm^3^samples taken from the right and left hemispheres of each animal. Samples were placed into linear density gradient columns of kerosene and bromobenzene. A calibration curve was determined for each column by using anhydrous K~2~SO~4~solutions of known SG (1.045, 1.040, 1.035 and 1.025). Brain tissue SG values were subsequently determined with this calibration curve.
Another set of experiments were performed for hemodynamic measurements and plasma magnesium level analysis. These rats were allocated into four parallel experimental groups with identical procedures. Right femoral artery catheterization was performed under general anesthesia for blood pressure monitoring and blood sampling. Blood samples (0.5 ml) were taken for the determination of plasma magnesium levels at baseline (T0) and 24 hours (T24) after the induction of sepsis, and an equal volume of saline was given. Mean arterial pressure was recorded at baseline and 2, 3, 4, 8, 12 and 24 hours after the surgical procedure. Four of 12 rats in group S and 3 of 11 rats in group S-mg died within 24 hours of the induction of sepsis. Data for these rats were excluded from the study. We continued to enter rats with a balanced randomization sequence until we had eight surviving rats for each group.
Statistical analysis
--------------------
The results are expressed as means ± standard deviation. EB dye content, brain tissue SG, serum magnesium levels, mean arterial pressures and heart rates were compared among four groups with a Kruskal--Wallis analysis of variance followed by Dunn\'s multiple comparisons test. A Mann--Whitney *U*-test and a Friedman nonparametric repeated-measures test were used for within-group comparisons. Paired serum magnesium levels were compared within each group by using a Wilcoxon signed rank test. Mortality rate was compared between septic groups receiving and not receiving magnesium with a χ^2^test. A probability (*P*) of less than 0.05 was considered significant.
Results
=======
Thirteen of 29 rats in group S and 10 of 26 rats in group S-Mg died within 24 hours after the induction of sepsis, whereas all of the rats in groups C and C-Mg survived. The mortality rate was not statistically different between septic rats receiving and not receiving magnesium (in the experimental groups, χ^2^= 0.229, *P*= 0.632; in the monitoring groups, χ^2^= 0.100, *P*= 0.752). Both groups of septic rats appeared ill as demonstrated by exudates around nose and eyes, tachypnea and decreased spontaneous movement. Sham-operated rats seemed grossly normal and were active within their cages.
Changes in mean arterial pressure are summarized in Figure [1](#F1){ref-type="fig"}. A significant decrease was observed 2 hours after the induction of sepsis in groups S and S-Mg. No further changes in blood pressures were observed with the administration of magnesium in the control and sepsis groups.
Plasma magnesium levels were comparable between groups at baseline (Table [1](#T1){ref-type="table"}). A significant decrease in plasma magnesium levels was observed 24 hours after the induction of sepsis. An intramuscular dose of 750 μmol/kg MgSO~4~maintained the baseline plasma magnesium levels in magnesium-treated septic rats.
Quantitative estimation of the EB dye revealed that sepsis significantly increased BBB permeability as measured by EB extravasations into brain tissue. In the S-Mg group, BBB permeability was significantly decreased in comparison with the S group (Table [2](#T2){ref-type="table"}).
The SG of both hemispheres taken from sepsis-induced rats were significantly less than the sham-operated rats, indicating the formation of brain edema after the induction of sepsis (Table [3](#T3){ref-type="table"}). Brain tissue SG measurements in the magnesium-treated septic rats were significantly higher than in the untreated sepsis group. Within-group comparisons indicated no difference between the right and left hemispheres.
Discussion
==========
The results of the present study demonstrate that treatment with magnesium immediately after experimental sepsis attenuated BBB permeability and the extent of brain edema formation.
Alterations of BBB permeability with subsequent brain edema formation are common features of septic encephalopathy. Several hypotheses for the pathogenesis of septic encephalopathy have been discussed in the literature: metabolic derangement, direct bacterial invasion of the central nervous system, the effect of endotoxin on the brain, or altered cerebral macrocirculation and microcirculation \[[@B11]-[@B16]\]. Recent evidence implicates the changes in the BBB permeability that favor brain edema formation in the pathophysiology of septic encephalopathy \[[@B3],[@B17]\]. In our model the BBB permeability defect induced by sepsis, as demonstrated by the EB dye extravasation technique, is consistent with previous reports demonstrating a loss of BBB integrity as a result of a septic challenge; however, the change in the SG representing brain tissue edema formation was relatively minor. Although the small change in SG that we obtained in the sepsis group reached statistical significance, indicating some amount of edema formation with the induction of sepsis, it is not possible to relate the edema formation to the disturbed integrity of the BBB.
Our results are consistent with previous reports on the integrity of the BBB and the role of a permeability defect in the formation of cerebral edema using other models of cerebral damage \[[@B18]-[@B20]\]. In our previous experimental study we evaluated the effects of magnesium on brain edema formation and BBB breakdown after closed-head trauma in rats \[[@B5]\]. Our results of BBB breakdown by the measurement of EB dye extravasation were comparable with those that we obtained in our sepsis model; however, the changes in SG were higher in the traumatic brain injury model than in our sepsis model. This might be explained by the different mechanisms causing BBB breakdown and edema formation in trauma and sepsis. The discrepancy between brain edema and BBB permeability defect in sepsis might also indicate a low grade of permeability defect due to the complex cascade of sepsis, which is not enough to create edema as such in trauma. Another possible explanation might be that the quantitative determination of BBB permeability defect by EB dye extravasation is more sensitive than the SG method for determining brain edema.
Other methods have been used to determine BBB damage in septic encephalopathy. In rodents with sepsis, colloidal iron dioxide \[[@B21]\], ^14^C-labelled amino acids \[[@B22]\] and ^125^I-labelled albumin \[[@B23]\] have been shown to pass from the circulation into the brain parenchyma in a similar manner to that seen in portosystemic encephalopathy. However, there is no evidence in the literature to suggest that this damage is related to edema formation in sepsis. Most recently, morphologic changes have been showed in the frontal cortex of a pig model of sepsis \[[@B3]\]. Fecal peritonitis resulted in severe perimicrovessel edema that was associated with swelling and rupture of astrocyte endfeet. Although this was suggested as evidence for the breakdown of the BBB, the ultrastructure of intercellular tight junctions seemed morphologically intact in pigs with sepsis. The authors have suggested that some other mechanism might be involved in the formation of edema. It is not known whether edema formation is related to BBB breakdown or other factors in sepsis. The exact mechanism and the relation between BBB breakdown and edema formation in sepsis-induced brain injury need to be further evaluated by more sensitive methods.
A major finding of the present study is that magnesium administration attenuates the increase in BBB permeability and edema formation. The exact mechanism of magnesium\'s beneficial effect on the integrity of the BBB is unclear. However, magnesium can affect many aspects of the mediator cascade that can cause a permeability defect in the BBB. Alternatively, magnesium can act directly on the BBB. Magnesium\'s cytoprotective effect to reduce the profound breakdown of the BBB was first demonstrated in a rat model of severe insulin-induced hypoglycemia \[[@B4]\]. In this study it was speculated that magnesium might exert this effect through suppression of the endothelial cells. It was suggested that even before magnesium reaches the brain site it interacts with the endothelial cells forming the BBB and inhibits their activation \[[@B4],[@B24],[@B25]\]. To our knowledge, the present data are the first to show the positive effects of magnesium on sepsis-induced BBB permeability changes. Although this might have clinical significance, a contrary suggestion could be that increasing the integrity of the BBB might also have negative effects in terms of antibiotic emergence when the clinical situation is complicated with encephalitis or meningitis. In our model of intra-abdominal sepsis, the cultures of brain specimens taken after the experiment were all sterile (data not shown).
One of the major pitfalls in the interpretation of the data was the difficulty of establishing a dose response for magnesium. In our present study the dose and the timing of magnesium administration were chosen with reference to our previous experiments on traumatic brain injury \[[@B5]\]. This dose of magnesium was determined as an optimum dose showing the best neurologic outcome in a traumatic brain injury model \[[@B26]\]. Plasma magnesium levels decreased significantly with the induction of sepsis and returned to nearly control levels with the dose of magnesium that we administered. However, it is known that the plasma magnesium level does not represent tissue magnesium content, and the lack of correlation between plasma magnesium and total body magnesium content in healthy subjects has already been reported \[[@B27]\]. More recently, it was demonstrated that free magnesium levels in brain tissue is a sensitive method that reflects magnesium homeostasis in a traumatic brain injury model \[[@B28]\]. Although we do not know to what extent the plasma magnesium levels represent brain tissue levels in the present study, our data show that significant beneficial effects are achievable with the dose administered. However, future studies will be needed to establish a dose response by measuring free magnesium levels in brain tissue for the effects of magnesium therapy in sepsis-induced brain injury.
Conclusion
==========
This investigation shows that sepsis increases BBB permeability and leads to the formation of brain edema in septic rats. Magnesium administration attenuated the increased BBB permeability and caused a reduction in brain edema formation in our rat model of intraperitoneal sepsis. The precise mechanisms and the pharmacodynamics of magnesium administration in sepsis-induced brain injury need further investigation.
Key messages
============
• Sepsis causes BBB permeability defect.
• Magnesium attenuates the increased BBB permeability associated with sepsis.
Abbreviations
=============
BBB = blood--brain barrier; EB = Evans blue; SG = specific gravity.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
All authors were responsible for study design and implementation of the experiment. Study data were collected by TE, DA and FE. Results were analyzed by FE and TE. The manuscript was written by FE and TE; all authors participated in revisions and gave approval to the final draft for submission for publication.
Acknowledgements
================
We thank Riyan Disci for statistical advice.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Hemodynamic data. Groups: sham control (C, *n*= 8), sham control MgSO~4~-treated (C-Mg, *n*= 8), septic (S, *n*= 8) and septic MgSO~4~-treated (S-Mg, *n*= 8). Mean arterial pressures compared among four groups using a Kruskal--Wallis analysis of variance followed by Dunn\'s multiple comparisons test. ^a^Septic versus sham control, *P*\< 0.05. ^b^Septic versus sham control MgSO~4~-treated, *P*\< 0.05. ^c^At 2 hours after the induction of sepsis versus baseline value (in the septic group), *P*\< 0.01. ^d^Septic MgSO~4~-treated versus sham control MgSO~4~-treated, *P*\< 0.01. ^e^At 2 hours after the induction of sepsis versus baseline value (in the septic MgSO~4~-treated group), *P*\< 0.05. A Friedman nonparametric repeated-measures test was used for within-group comparisons.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Plasma magnesium concentrations
:::
Measurement Group (*n*) T0 T24 *P*
---------------- ------------- ------------- ------------------ --------
Plasma Mg (mM) C (8) 1.11 ± 0.05 1.10 ± 0.05 NS
S (8) 1.09 ± 0.05 0.89 ± 0.06^a,b^ 0.0078
C-Mg (8) 1.10 ± 0.06 1.29 ± 0.06 0.0078
S-Mg (8) 1.13 ± 0.03 1.01 ± 0.08^c^ 0.0156
KW 2.708 26.863
d.f. 3 3
*P* \>0.05 \<0.0001
Abbreviations: d.f., degrees of freedom; KW, Kruskal--Wallis test statistic; NS, not significant; *P*, approximate χ^2^*P*value; T0, basal measurement; T24, measurement at 24 hours. Groups: sham control (C), sham control MgSO~4~-treated (C-Mg), septic (S) and septic MgSO~4~-treated (S-Mg). Data are expressed as means ± standard deviation.
Dunn\'s multiple comparisons test: ^a^septic versus sham control, *P*\< 0.05; ^b^septic versus sham control MgSO~4~-treated, *P*\< 0.001; ^c^septic MgSO~4~-treated versus sham control MgSO~4~-treated, *P*\< 0.01. Paired serum magnesium levels were compared within groups using a Wilcoxon signed rank test. Two-tailed *P*values are shown in the last column.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Assessment of blood--brain barrier permeability by Evans blue dye content in brain tissue
:::
Measurement Group (*n*) Left hemisphere Right hemisphere P
--------------- ------------- ----------------------- ----------------------- ------
EB dye (μg/g) C (8) 0.00160 ± 0.0003 0.00145 ± 0.0003 0.33
S (8) 0.00466 ± 0.0002^a,b^ 0.00641 ± 0.0003^c,d^ 0.13
C-Mg (8) 0.00135 ± 0.0002 0.00145 ± 0.0003 0.44
S-Mg (8) 0.00218 ± 0.0005 0.00199 ± 0.0007 0.57
KW 19.720 23.039
d.f. 3 3
*P* \< 0.001 \< 0.0001
Abbreviations: d.f., degrees of freedom; EB, Evans blue; KW, Kruskal--Wallis test statistic; *P*, approximate χ^2^*P*value. Groups: sham control (C), sham control MgSO~4~-treated (C-Mg), septic (S) and septic MgSO~4~-treated (S-Mg). Data are expressed as means ± standard deviation.
Dunn\'s multiple comparisons test: ^a^septic versus sham control, *P*\< 0.01; ^b^septic versus sham control MgSO~4~-treated, *P*\< 0.001; ^c^septic versus sham control, *P*\< 0.001; ^d^septic versus sham control MgSO~4~-treated, *P*\< 0.001. A Mann--Whitney test was used for within-group comparisons. Two-tailed *P*values are shown in the last column.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Assessment of edema by specific gravity of brain tissue
:::
Measurement Group (*n*) Left hemisphere Right hemisphere *P*
------------- ------------- ---------------------- ---------------------- ------
SG C (8) 1.0444 ± 0.0001 1.0443 ± 0.0002 0.24
S (8) 1.0429 ± 0.0009^a,b^ 1.0424 ± 0.0012^c,d^ 0.44
C-Mg (8) 1.0444 ± 0.0002 1.0444 ± 0.0001 0.44
S-Mg (8) 1.0438 ± 0.0007 1.0439 ± 0.0004^e^ 0.24
KW 18.831 24.724
d.f. 3 3
*P* \< 0.001 \<0.0001
Abbreviations: d.f., degrees of freedom; KW, Kruskal--Wallis test statistic; *P*, approximate χ^2^*P*value; SG, specific gravity. Groups: sham control (C), sham control MgSO~4~-treated (C-Mg), septic (S) and septic MgSO~4~-treated (S-Mg). Data are expressed as means ± standard deviation.
Dunn\'s multiple comparisons test: ^a^septic versus sham control, *P*\< 0.001; ^b^septic versus sham control MgSO~4~-treated, *P*\< 0.01; ^c^septic versus sham control, *P*\< 0.01; ^d^septic versus sham control MgSO~4~-treated, *P*\< 0.001; ^e^septic MgSO~4~-treated versus sham control MgSO~4~-treated, *P*\< 0.05. A Mann--Whitney test was used for within-group comparisons. Two-tailed *P*values are shown in the last column.
:::
|
PubMed Central
|
2024-06-05T03:55:54.887107
|
2004-11-23
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065104/",
"journal": "Crit Care. 2005 Nov 23; 9(1):R18-R23",
"authors": [
{
"first": "Figen",
"last": "Esen"
},
{
"first": "Tulin",
"last": "Erdem"
},
{
"first": "Damla",
"last": "Aktan"
},
{
"first": "Mukadder",
"last": "Orhan"
},
{
"first": "Mehmet",
"last": "Kaya"
},
{
"first": "Haluk",
"last": "Eraksoy"
},
{
"first": "Nahit",
"last": "Cakar"
},
{
"first": "Lutfi",
"last": "Telci"
}
]
}
|
PMC1065106
|
Introduction
============
Midazolam is an intravenous sedative that is commonly used during ventilation in critical illness. It is often regarded as the sedative of choice in the intensive care unit (ICU). According to the findings of our recent electronic survey (93% respondents) \[[@B1]\], midazolam is still routinely used in the UK as a sedative in ICUs.
When used as a single dose, midazolam\'s pharmacological characteristics appear favourable, with a rapid onset of action and a short plasma elimination half-life. Midazolam is 94--98% bound to plasma albumin and has a volume of distribution of 1.7 l/kg in healthy individuals \[[@B2]\]. It is extensively metabolized first via cytochromes p450, 3A4 and 2B6 to 1-hydroxymidazolam, before undergoing glucuronidation to form 1-hydroxymidazolam glucuronide (1-OHMG), which has sedative properties and is excreted in the urine \[[@B3],[@B4]\]. A wide interpatient variability in the pharmacokinetic properties of midazolam in critically ill patients with multiple organ failure has been reported \[[@B5]\], which can lead to prolonged sedation after midazolam therapy is stopped. However, there are other important causes of patients being \'slow to wake\'; of these, it is most important to identify severe neurological damage. Patients with multiple organ failure are at high risk for neurological damage because they frequently have episodes of hypotension and dysrhythmia, and may have significant coagulopathy during the course of their critical illness.
We suspected that some patients in our ICU, particularly those with renal impairment, were becoming over-sedated with midazolam and the active metabolite 1-OHMG, and that this was complicating the neurological assessment of \'slow to wake\' patients. We previously developed a rapid assay for measuring midazolam and its glucuronide metabolite simultaneously \[[@B1]\]. This short report describes the usefulness of this assay for identifying midazolam over-sedation and its potential use as a predictor of eventual neurological recovery.
Methods
=======
The assay was available for clinical application in the ICU. To differentiate between midazolam over-sedation and neurological damage, consultant intensivists requested detection of midazolam and 1-OHMG in serum. This request was normally made during the morning ICU ward round. The patients studied were those who had received intravenous midazolam therapy by continuous infusion either before (e.g. in operating theatres) or during the course of their ICU admission, and who were \'slow to wake\' and in whom there was clinical suspicion of neurological damage.
Arterial blood (2 ml) was collected from each patient via an *in situ*arterial catheter. The time of sample collection and the midazolam administration history, including cessation time, were recorded. A specific assay utilizing high-performance liquid chromatography coupled to mass spectrometric detection was used for simultaneous detection and quantification of midazolam and 1-OHMG \[[@B1]\]. Mass spectrometry allowed identification of midazolam and 1-OHMG individually based on their isotopic patterns. The studies were performed on the basis of clinical need, and in all cases they were requested by the consultant intensivist, normally during the morning ward round. The quantified serum level of midazolam and 1-OHMG could be reported to the medical team after a minimum of 2 hours so that they could consider the findings in their decisions regarding further clinical intervention. In practice, morning requests were available for interpretation by the evening round.
Unit characteristics
--------------------
The ICU at Guy\'s and St. Thomas\' National Health Service Trust is a 30-bed, level 3 unit that serves a mixture of medical, surgical, trauma, oncology and haematology patients. It has an average of 100 admissions per calendar month. For the year from March 2003 to April 2004, the mean Acute Physiology and Chronic Health Evaluation II score (day 1) was 18.5 ± 7.3, with a hospital mortality of 32.5% and a median length of stay of 5 days (variance 189.5, maximum 246).
Patient characteristics
-----------------------
All patients appeared to be deeply sedated at the time that the sample was taken, with a Glasgow Coma Scale score of less than 5. They were considered \'slow to wake\' from either a pharmacological and neurological cause if, in the absence of a focal neurological deficit, consciousness did not return within 36 hours of stopping sedation. Patients were deemed to have regained consciousness if they both opened their eyes and moved their limbs in response to commands.
Studies were conducted in 26 patients who had received midazolam sedation therapy by continuous intravenous infusion and in whom neurological damage was considered clinically possible (e.g. a hypoxic event was noted during cardiac surgery). The mean age of these patients was 63 ± 16 years, and the median time from cessation of midazolam therapy to serum collection was 67 hours (range 36--146 hours). The median daily midazolam dose was 4 mg/hour (range 2--20 mg/hour). The reasons for ICU admission are described in Table [1](#T1){ref-type="table"}.
We followed the clinical outcomes of these patients in terms of mortality, neurological recovery and neurological damage on discharge. If no midazolam or 1-OHMG was detected, then a series of standard clinical and diagnostic tests was undertaken to determine whether neurological damage was likely. These included the response to painful stimuli and computed tomography of the head. In patients in whom midazolam or 1-OHMG was detected, tests were deferred until either the patients awoke or levels became undetectable.
Results
=======
Midazolam and/or 1-OHMG were detected in the serum of 13 of the 26 patients (referred to as the midazolam-positive group). Of these 13 patients, 10 made a full neurological recovery; nine of these patients were discharged from the ICU and one later died as a result of critical illness but with intact neurological function. The remaining three patients died without regaining consciousness as a result of neurological damage.
In contrast, neurological damage was observed in 10 of the remaining 13 patients who had no detectable serum concentrations of midazolam and/or 1-OHMG (midazolam-negative group). Midazolam-positive patients were significantly less likely to have experienced neurological damage (χ^2^test \[degrees of freedom = 1\]: *P*\< 0.002).
Twelve of the midazolam-positive patients had serum midazolam concentrations between 16 and 650 ng/ml, with a median value of 30 ng/ml, whereas the remaining patient\'s level exceeded the upper limit of the assay (3000 ng/ml). 1-OHMG was detected at a mean of 6800 ± 3432 ng/ml (range 3121--11,525 ng/ml) in the serum of six of the 13 midazolam-positive patients. All six of these patients exhibited a degree of renal impairment (defined as serum creatinine \>130 μmol/l; Table [1](#T1){ref-type="table"}), four of whom required renal replacement therapy in the form of continuous venovenous haemofiltration (employing an ultrafiltration rate of between 1500 and 3000 ml/hour). 1-OHMG was not detected in any of the midazolam-negative patients.
Of the 13 midazolam-negative patients, eight died without regaining consciousness as a result of neurological damage, and two were discharged from the ICU with significant neurological impairment and required prolonged neurological rehabilitation. None of these 10 patients had responded appropriately to painful stimuli when in the ICU. In addition, in seven of these patients structural neurological damage was detected by computed tomography scan. Only three out of 13 patients in this group of midazolam-negative patients left the ICU with no neurological deficit.
Other sedative and opiate agents
--------------------------------
Out of 26 patients, 15 were administered fentanyl by continuous intravenous infusion at a dosage between 0 and 300 μg/hour. In the 15 patients the fentanyl infusion was ceased at a minimum of 56 hours and a maximum of 120 hours before sample collection. In 25 of the 26 patients we could find no documented evidence of administration of sedative and opiate agents for a minimum of 36 hours before serum sample collection. The remaining patient, in the midazolam-negative group, was receiving 30 mg/day of the sedating antihistamine chlorphenamine; this was one of the three patients who were discharged from the ICU with neurological function intact.
Discussion
==========
In this study, midazolam with or without 1-OHMG was detected in half of the \'slow to wake\' patients, in whom testing was requested after a mean time from therapy cessation of 3 days. In one patient, in whom there was no record of midazolam administration in the ICU, a level of 200 ng/ml was recorded. It later transpired that a large dose of midazolam had been administered in the operating theatre more than 96 hours earlier. Detection of 1-OHMG in renal impairment confirmed that 1-OHMG accumulates in the presence of renal failure. Furthermore, its presence in high serum concentrations (3121--11,525 ng/ml) in the face of midazolam levels below the therapeutic range, normally quoted in the critically ill of 100--1000 ng/ml \[[@B5]\], while the patient remained deeply sedated concurs with earlier reports \[[@B3],[@B4]\] that 1-OHMG has a sedative effect and contributes to prolonged sedation in renal impairment. Other investigators have reported the presence of 1-OHMG in the absence of midazolam \[[@B3],[@B4]\], but we did not observe this and suspect that it was because the assay we used is able to detect very low concentrations of midazolam.
Our findings suggest that serum levels of midazolam and/or 1-OHMG in \'slow to wake\' patients may be used to aid differentiation between prolonged sedation and neurological damage. Patients found to be midazolam positive using this rapid assay were significantly less likely to have suffered neurological damage. Correct discrimination between neurological damage and prolonged sedation was made for 20 out of 26 patients, indicating a high degree of accuracy. Clearly, the possibility that midazolam-positive patients also have neurological damage remains and must be excluded if these patients do not awaken when serum concentrations of benzodiadepines have fallen to undetectable levels. Additionally, in the midazolam-negative group three patients were discharged with neurological function intact. This of course does not exclude a neurological cause of the coma that had fully resolved on discharge. One patient was receiving the sedating antihistamine chlorphenamine (30 mg/day intravenously) and did not regain full consciousness until it was stopped. In the remaining two patients no other clinical cause of the coma was apparent.
The only other agent used routinely in these patients that could have significantly contributed to their reduced level of consciousness was the intravenous opiate fentanyl. Although fentanyl is known to accumulate in critical illness \[[@B6]\], we could find no evidence of accumulation for longer than 36 hours \[[@B7]\], and, because our group of patients had not received the drug for more than 2 days before sampling, it was not thought to contribute to the patients being \'slow to wake\'.
Arguably, the most important finding is that over three-quarters of the \'slow to wake\' patients with no detectable serum midazolam/1-OHMG either died or were discharged from the ICU with profound neurological damage, whereas more than three-quarters of those with detectable midazolam/1-OHMG went on to make a full recovery. This observation suggests that prolonged sedation occurs after midazolam therapy and that it can be difficult to differentiate this from neurological damage in the acutely ill patient. The exclusion of midazolam or its metabolite 1-OHMG should be confirmed either by assay detection, as we describe, or by using the short-acting benzodiazepine antagonist flumazenil before a formal diagnosis of neurological damage is made. There are reports \[[@B3],[@B8]\] in the literature of successful reversal of benzodiazepine sedation in critical illness using flumazenil, but we rarely use it in our unit because we find it to be nonspecific, short acting and able to induce seizures \[[@B9]\].
We recommend that use of alternatives to midazolam be considered in this patient group whenever possible, and that if its use is considered essential then steps should be taken to exclude the continuing presence of the drug or its metabolite before an opinion regarding neurological damage is formed.
These findings have led to a change in prescribing practice in our ICU. We no longer use midazolam for sedation, and our sedation policy is now based on administering propofol or lorazepam. This view is also supported by the Society of Critical Care Medicine\'s most recently published guidelines \[[@B10]\], which recommend use of lorazepam for sedating most patients via intermittent or continuous infusion and use of propofol for short-term sedation, and that midazolam be reserved for rapid control of agitated patients and for short-term sedation. As a consequence, we were unable to conduct a more formal study of midazolam\'s role in over-sedation or extend the study to a larger group of patients.
Conclusion
==========
The results of this investigation confirm that prolonged sedation from midazolam or 1-OHMG should always be considered in the differential diagnosis of neurological damage in critically ill patients who have received midazolam. This can be accurately detected using the assay method described. The sedative effects of midazolam/1-OHMG can persist for days after stopping administration of midazolam. If prolonged sedation can be excluded in these patients, then it is highly likely that neurological damage has occurred.
Key messages
============
• In some patients midazolam is metabolized to its glucuronide, which has sedative properties.
• Prolonged sedation resulting from this metabolite should be considered when making a differential diagnosis of neurological damage in \'slow to wake\' patients.
• Measurement of midzolam and its metabolite in slow to wake patients will aid the differential diagnosis in these patients.
Abbreviations
=============
ICU = intensive care unit; 1-OHMG = 1-hydroxymidazolam glucuronide.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
All authors participated in the study design, interpretation of results and manuscript preparation. CMK also performed data collection and analyses.
Acknowledgements
================
This work was supported by the Special Trustees for St. Thomas\' Hospital. We thank the members of the UK Clinical Pharmacy Association Critical Care Group for participating in the electronic survey, Dr Jonathan Edgeworth for his helpful comments and Myra Wiseman for her statistical advice.
Figures and Tables
==================
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Patient characteristics and outcome
:::
Characteristics Midazolam-positive group (*n*= 13) Midazolam-negative group (*n*= 13) Statistics
----------------------------------------------------------------------------------------- ------------------------------------ ------------------------------------ ----------------
Admission diagnosis
Cardiothoracic surgery 8 6
Severe sepsis 0 4
Cardiorespiratory arrest 2 1
General surgery 2 0
Pancreatitis 1 0
Acute asthma 0 1
Perforated duodenal ulcer 0 1
APACHE II score (day 1; mean ± standard deviation) 19.1 ± 6.6 19.4 ± 7.1
Renal Impairment (serum creatinine \>130 μmol/l or receiving renal replacement therapy) 5 4
Neurological function intact 10 3 *P*\< 0.002^a^
Neurological damage 3 10 *P*\< 0.002^a^
All-cause mortality 4 (31%) 8 (62%)
Median midazolam dose mg/hour (range; 24 hours before cessation) 4 (2--20) 3.5 (2--15) NS^b^
Median time (range) from midazolam cessation (hrs) 66 (36--120) 68 (36--146) NS^b^
^a^χ^2^test. ^b^Mann--Whitney U-test. APACHE, Acute Physiology and Chronic Health Evaluation.
:::
|
PubMed Central
|
2024-06-05T03:55:54.888856
|
2004-12-14
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065106/",
"journal": "Crit Care. 2005 Dec 14; 9(1):R32-R36",
"authors": [
{
"first": "Catherine A",
"last": "McKenzie"
},
{
"first": "William",
"last": "McKinnon"
},
{
"first": "Declan P",
"last": "Naughton"
},
{
"first": "David",
"last": "Treacher"
},
{
"first": "Graham",
"last": "Davies"
},
{
"first": "Gary J",
"last": "Phillips"
},
{
"first": "Philip J",
"last": "Hilton"
}
]
}
|
PMC1065107
|
See related commentary <http://ccforum.com/content/9/1/31>
Introduction
============
In mechanically ventilated patients the frequency of post-extubation stridor is estimated to range between 4% and 22% \[[@B1]-[@B3]\]. Post-extubation stridor is usually due to laryngeal edema or decreased cross-sectional area of trachea, although vocal-cord dysfunction and overdose of sedative drugs may be also the cause. Nevertheless, this complication may result in emergency re-intubation in rather difficult circumstances with increased morbidity and mortality. The cuff-leak test has been proposed as a simple method of predicting the occurrence of this complication \[[@B4]-[@B7]\]. This test consists of deflating the balloon cuff of the endotracheal tube to assess the air leak around the tube during expiration by measuring the expiratory tidal volume with and without a deflated cuff \[[@B4]-[@B6]\]. A relatively large difference between these two values indicates that the cross-sectional area of the tracheal and/or upper airways is large enough to render the occurrence of post-extubation stridor, and therefore the possibility of re-intubation due to airway obstruction, unlikely \[[@B4]-[@B7]\]. Obviously the cuff-leak test is not useful if vocal cord dysfunction or overdose of sedative drugs is the cause of post-extubation stridor.
Typically the cuff-leak test is performed during volume control ventilation (using a tidal volume of 10 ml/kg) by deflating the cuff, whereas the expired tidal volume is measured a few breaths later \[[@B4]-[@B7]\]. The leak is calculated as the difference between the expiratory tidal volume with and without a deflated cuff \[[@B4]-[@B7]\]. However, because most ventilators in the intensive care unit do not compensate for leaks, it is possible that during inspiration with a deflated cuff a portion of the total amount of the predetermined volume given by the ventilator may leak around the endotracheal tube. In this case, the difference between expiratory tidal volume with and without a deflated cuff represents a total leak consisting of an inspiratory and an expiratory component. This total leak may depend on various factors such as the cross-sectional area around the endotracheal tube, inspiratory flow and respiratory system mechanics, which may affect either the inspiratory component or the expiratory component or both, therefore contributing to the poor performance of the cuff-leak test in identifying patients with post-extubation stridor, reported by some studies \[[@B8]\]. The aims of this physiological study were, first, to examine the effects of various variables, such as cross-sectional area around the endotracheal tube, inspiratory flow and respiratory system mechanics on total leak, and second, to compare the total leak with that obtained when the inspiratory component was eliminated, leaving only the expiratory leak. The inspiratory leak was eliminated by deflating the cuff at end-inspiration, a manoeuvre that guarantees that the ventilator delivers all the predetermined gas volume into the lung.
Methods
=======
Clinical study
--------------
Fifteen mechanically ventilated patients (aged 65 ± 19 years \[mean ± SD\]; seven males, eight females) were prospectively studied. All were orotracheally intubated (low-pressure cuff endotracheal tube, diameter 8.0 ± 0.5 mm, tube length 28 ± 1 mm), hemodynamically stable without vasoactive drugs, lightly sedated with propofol and with a PaO~2~/F~i~O~2~of more than 250 mmHg. The study was approved by the Hospital Ethics Committee, and informed consent was obtained from the patients or their families.
Flow (*V*\') at the airway opening was measured with a heated pneumotachograph (model 3700; Hans-Rudolf, Kansas City, KS, USA) and a differential pressure transducer (Micro-Switch 140PC; Honeywell Ltd, Montreal, Ontario, Canada), both placed between the endotracheal tube and the Y-piece of the ventilator. Flow was electronically integrated to provide volume. Airway pressure (*P*~aw~; Micro-Switch 140PC; Honeywell Ltd) was measured from a side port between the pneumotachograph and the endotracheal tube. Each signal was sampled at 150 Hz (Windaq Instruments Inc., Akron, OH, USA) and stored on a computer disk for later analysis.
Initially the patients were placed on volume control mode (Puritan-Bennett 840, Lenexa, KS, USA) with no flow compensation, heavily sedated (propofol--fentanyl) to achieve a Ramsay scale of 6 and paralyzed with *cis*-atracurium. Inactivity of respiratory muscles was confirmed with the use of standard criteria \[[@B9]\]. Tidal volume (*V*~T~) was set to 10 ml/kg given with a constant inspiratory flow rate of 1 litre/s. No end-inspiratory pause was applied. External positive end-expiratory pressure (PEEP) was set to zero while ventilator frequency was adjusted such as to achieve zero intrinsic PEEP, confirmed by end-expiratory occlusion \[[@B10]\].
When the patients were stable on volume control, the (baseline) expiratory *V*~T~was measured by averaging five consecutive breaths (*V*~T,baseline~). The absence of a leak was verified by an end-inspiratory occlusion of 10 s and observing a constant *P*~aw~after 3 s of occlusion. Thereafter, the cuff-leak test was performed randomly, either using the conventional method or by deflating the cuff at the end of a 3 s end-inspiratory pause. The conventional method consisted of balloon cuff deflation and measuring the expiratory tidal volume four breaths later (*V*~T,defl~). Five such trials were performed to obtain an average value of *V*~T,defl~. The difference between *V*~T,baseline~and *V*~T,defl~was defined as the cuff-leak volume obtained by the conventional method (Leak~conv~). When the cuff was deflated at the end of the end-inspiratory pause only the following expiratory tidal volume was measured (*V*~T,pause~). Again five such trials were performed. The difference between *V*~T,baseline~and *V*~T,pause~was defined as the cuff-leak volume obtained by deflating the cuff during end-inspiratory pause (Leak~pause~).
The mechanics of the respiratory system were measured by using the occlusion technique \[[@B10]-[@B12]\]. In each patient at least five breaths with a satisfactory plateau were analyzed and the mean values were reported. Respiratory system static inflation end-inspiratory compliance (*C*~rs~), minimum (*R*~int~) and maximum (*R*~rs~) resistance of the respiratory system and the difference between *R*~rs~and *R*~int~(Δ*R*) were computed according to standard formulas and procedures \[[@B11],[@B12]\].
In all patients ΔLeak was calculated as the difference between Leak~conv~and Leak~pause~. Assuming that the difference between peak inspiratory *P*~aw~(Δ*P*~aw,peak~) between methods was entirely due to different end-inspiratory lung volume, the predicted ΔLeak was calculated by the product of Δ*P*~aw,peak~and *C*~rs~.
Lung model study
----------------
To examine the effects of various variables on cuff-leak volume measurement, a two-chamber test lung (Michigan Instruments Inc., Grand Rapids, MI, USA) was used \[[@B13]\]. Each chamber was connected to a common tube representing the trachea by a tube with varying resistance. The compliance of each chamber was also variable. The two chambers were connected to a ventilator (Puritan-Bennett 840) via a cuffed endotracheal tube 8 mm in diameter inserted into the common tube. Small plastic bands were inserted between the endotracheal tube and the common tube to create controlled leaks when the balloon cuff was deflated. Two levels of leak were created, simulating two different cross-sectional areas around the endotracheal tube (large and small). The cross-sectional area around the endotracheal tube was quantified by cuff deflation during the end-inspiratory pause time and observation of the rate of pressure drop when an inspired tidal volume of l litre was used and total model compliance was 50 ml/cmH~2~O. The rate of pressure decrease was about 10 and 5 cmH~2~O/s with large and small cross-sectional areas, respectively. The absence of leak with the cuff inflated was confirmed by end-inspiratory occlusion and demonstration of a constant plateau *P*~aw~.
*V*~T~was set at 0.6 litre (given with constant flow rate) and external PEEP to zero throughout. Ventilator frequency was adjusted so that no dynamic hyperinflation was observed. The absence of dynamic hyperinflation was verified by end-expiratory occlusion and no intrinsic PEEP demonstration \[[@B10]\]. Two protocols were performed. In the first (protocol A), the effects of inspiratory flow (*V*\'~I~) on cuff-leak volume measurement as well as the interaction between *V*\'~I~, cross-sectional area around the endotracheal tube and model mechanics were studied. At small and large cross-sectional area around the endotracheal tube and three combinations of model mechanics, representing normal (model airway resistance, *R*= 8 cmH~2~O/litre per second; model airway compliance, *C*= 50 ml/cmH~2~O), restrictive (*R*= 8 cmH~2~O/litre per second, *C*= 20 ml/cmH~2~O) and obstructive pattern (*R*= 16 cmH~2~O/litre per second, *C*= 100 ml/cmH~2~O), *V*\'~I~was varied between 0.6 and 1 litre/s and cuff-leak volume was measured either by the conventional method or by deflating the cuff at the end of a 3 s end-inspiratory pause as described above. The effects of model mechanics on cuff-leak volume were further studied in a separate protocol (protocol B). At a constant cross-sectional area around the endotracheal tube (large) and an inspiratory flow of 0.6, each method of cuff-leak volume measurement was studied at three levels of *R*and *C*, resulting in nine combinations of system mechanics (*R*= 8, 16 and 32 cmH~2~O/litre per second and *C*= 20, 50 and 100 ml/cmH~2~O). Similarly to protocol A, at each combination of model mechanics the cuff-leak volume was measured either by the conventional method or by deflating the cuff at the end of a 3 s end-inspiratory pause.
Data were analyzed with a paired *t*-test and a multi-factorial analysis of variance for repeated measurements, where appropriate. When the *F*value was significant, Tukey\'s test was used to identify significant differences. Linear regression analysis was performed with the least-squares method. *P*\< 0.05 was considered statistically significant. Data are expressed as means ± SD. In the lung model study, means ± SD for the variables were determined from a total of 10 measurements.
Results
=======
Clinical study
--------------
Baseline ventilator settings and respiratory mechanics are shown in Table [1](#T1){ref-type="table"}. When the cuff remained deflated throughout the respiratory cycle, *P*~aw,peak~of the analyzed breaths (24.0 ± 6.6 cmH~2~O) was significantly lower than that of the breath in which the cuff was deflated at the end of the inspiratory pause (26.7 ± 7.1 cmH~2~O); the mean Δ*P*~aw,peak~averaged 2.6 ± 2.6 cmH~2~O (range 0.5--8.2 cmH~2~O). As expected, *P*~aw,peak~of the breaths in which the cuff was deflated at the end of the inspiratory pause was similar to the corresponding value of the baseline. In all patients Leak~conv~was higher than Leak~pause~, averaging 188 ± 159 ml (32 ± 25% of *V*~T,baseline~) and 61 ± 75 ml (10 ± 12% of *V*~T,baseline~), respectively (*P*\< 0.05; Fig. [1](#F1){ref-type="fig"}). There was a significant linear relationship between Leak~conv~and Leak~pause~(*y*= - 12.3 + 0.39*x*, *r*= 0.84, *P*\< 0.05; Fig. [1](#F1){ref-type="fig"}). The observed ΔLeak averaged 127 ± 105 ml. There was a significant linear relationship between Δ*P*~aw,peak~and the observed ΔLeak (*y*= 64.8 + 26.2*x*, *r*= 0.66, *P*\< 0.05) and between the predicted and observed ΔLeak (*y*= 13.14 + 0.73*x*, *r*= 0.69, *P*\< 0.05). There was no relationship between observed ΔLeak and respiratory system mechanics (*R*~int~, *R*~rs~, Δ*R*and *C*~rs~), the time constant of the respiratory system and *V*~T,baseline~.
Model study
-----------
### Protocol A
For a given condition, Leak~conv~was significantly higher than Leak~pause~(Table [2](#T2){ref-type="table"}). For a given cross-sectional area, and independently of model mechanics, Leak~pause~was not affected by *V*\'~I~, whereas Leak~conv~increased significantly with decreasing *V*\'~I~(Table [2](#T2){ref-type="table"}). Independently of the cross-sectional area around the endotracheal tube with simulated restrictive respiratory system disease and at a *V*\'~I~of 0.6 litre/s, Leak~conv~was significantly higher than the corresponding values with simulated normal mechanics and obstructive respiratory system disease. ΔLeak increased significantly with decreasing *V*\'~I~and increasing the size of the cross-sectional area around the endotracheal tube (Fig. [2](#F2){ref-type="fig"}). The effect of *V*\'~I~on ΔLeak was significantly higher with simulated restrictive respiratory system disease and large cross-sectional area around the endotracheal tube (Fig. [2](#F2){ref-type="fig"}).
### Protocol B
Similarly to protocol A, and independently of model mechanics, Leak~conv~was significantly higher than Leak~pause~(Table [3](#T3){ref-type="table"}). For a given *R*, Leak~conv~increased significantly with decreasing *C*, whereas Leak~pause~remained constant. For a given *C*, Leak~pause~and Leak~conv~tended to increase slightly with the highest resistance, the difference being significant only for Leak~pause~. ΔLeak was not affected by model resistance, whereas it increased significantly with decreasing compliance (Fig. [3](#F3){ref-type="fig"}).
Discussion
==========
The main findings of this study were as follows. First, because in mechanically ventilated patients the expiratory leak volume is about 30% of the sum of inspiratory and expiratory leaks (total leak), the inspiratory leak significantly affected the results of the cuff-leak test. Second, the cross-sectional area around the endotracheal tube is not the only determinant of cuff-leak test. Third, respiratory system compliance and inspiratory flow affect the test significantly, mainly through an effect on the inspiratory component. Fourth, the expiratory component is slightly influenced by respiratory system resistance.
To avoid the confounding factors of respiratory muscle activity and dynamic hyperinflation on the calculation of cuff-leak volume, the patients were paralyzed and ventilated with settings that permitted the respiratory system to reach passive functional residual capacity at the end of expiration. Similarly, in the lung model the ventilator settings were such that dynamic hyperinflation was not observed. Therefore, for a given experimental condition the inspired tidal volume entirely determined the total expired volume. Finally, contrary to other studies \[[@B5]\], cuff-leak volume was measured by comparing the expired tidal volume with and without a deflated cuff. In this case the difference between inspired and expired tidal volume due to gas exchange and the different temperature and humidity of inspired and expired gas were not an issue.
By deflating the cuff at the end of the inspiratory pause we guaranteed that the ventilator delivered all of the predetermined gas volume into the lung, as indicated by the similar peak *P*~aw~between the breaths used to calculate the cuff-leak volume. Because inactivity of respiratory muscles and absence of dynamic hyperinflation were ensured, any difference in expired volume with and without a deflated cuff should be entirely due to gas leak around the endotracheal tube during expiration (pause cuff leak). In contrast, when the cuff-leak volume was measured with the conventional method, a fraction of gas volume delivered by the ventilator might leak around the endotracheal tube during inspiration. In that case the measured cuff-leak volume is the total leak consisting of an inspiratory and expiratory component. The design of this study did not permit us to measure with accuracy the inspiratory leak. This is because pause cuff leak is not similar to expiratory leak obtained with the conventional method because end-inspiratory lung volume and thus elastic recoil pressure at the beginning of expiration differ substantially between the two methods of cuff leak determination. The pause cuff leak should be higher than the expiratory component of the total leak, because end inspiratory lung volume and elastic recoil pressure were considerably higher when pause cuff leak was obtained.
Both in clinical and model study the cuff-leak volume determined with the conventional method (Leak~conv~) was always higher than that obtained by cuff deflation at end-inspiratory pause, which eliminated the inspiratory component of total leak (Leak~pause~). It follows that the inspiratory component is an important determinant of the cuff-leak test. It is of interest to note that in patients Leak~conv~was about threefold Leak~pause~whatever the amount of the total leak.
In Protocol A of the lung model study, for a given cross-sectional area, the system mechanics and inspiratory flow considerably affected Leak~conv~; Leak~conv~increased significantly with decreasing compliance and inspiratory flow. In contrast, neither system compliance nor inspiratory flow influenced Leak~pause~, which remained relatively constant. As a result ΔLeak increased significantly with decreasing compliance and inspiratory flow. The constancy of Leak~pause~suggested that the expiratory component of the total leak was also unaffected by changes in system compliance and inspiratory flow. It follows that respiratory system compliance and inspiratory flow have an important impact on cuff-leak test, mainly through an effect on the inspiratory component. The increased inspiratory leak with decreasing system compliance is predictable because the stiffness of the respiratory system causes a greater fraction of inspiratory flow to deviate to atmosphere though the free space between the endotracheal tube and the trachea. Similarly, the increased inspiratory leak with low inspiratory flow was also expected. The free space between the endotracheal tube and trachea represents a low-resistance pathway and, because for a given tidal volume low inspiratory flow is associated with longer inspiratory time, the inspiratory leak should increase, a situation resembling that of bronchopleural fistula in which high inspiratory flows are recommended so as to reduce the amount of air leaking through the fistula \[[@B14]\]. Thus the cuff-leak volume calculated by the conventional method does not solely reflect the cross-sectional area of the trachea and/or the upper airways but is influenced by other factors such as respiratory system mechanics and inspiratory flow.
In protocol B of the lung model study, a slight increase in cuff-leak volume at the highest resistance value was observed with both methods. As a result, ΔLeak was not influenced by model resistance, indicating that system resistance affected mainly the expiratory component of the total leak. Although the factors underlying the above increase are not clear, the flow velocity profile during expiration could account for these findings. Nevertheless the difference was relatively small (less than 25 ml or less than 4% of *V*~T~), making the clinical significance of this finding questionable. Furthermore the increase in expiratory leak was observed at very high values of resistance that preclude the weaning process, making the performance of the cuff-leak test clinically irrelevant.
We should note that in patients the cuff leak was determined at the relatively high constant inspiratory flow of 1 litre/s. Although the effect of flow was not studied in our patients, the model study indicates that overestimation should be higher at low flow. Nevertheless, high inspiratory flow is recommended in patients with obstructive lung disease ventilated on volume control so as to reduce dynamic hyperinflation \[[@B15]\].
In contrast with the model study, in the clinical study there was no relationship between observed ΔLeak and respiratory system mechanics (*R*~int~, *R*~rs~, Δ*R*and *C*~rs~), the time constant of the respiratory system and *V*~T,baseline~. Differences in cross-sectional area of the trachea and upper airways between patients might obscure any relationship between these variables and ΔLeak.
Studies suggest that leak volume, as obtained by the conventional method, may predict the occurrence of post-extubation stridor and might thus identify the subset of patients at risk of re-intubation due to upper airway obstruction \[[@B4],[@B5],[@B7]\]. However, the cut-off point of leak volume differed substantially between studies. In addition, the positive predictive value was quite low, indicating that the results of the cuff-leak test should not be used to postpone the extubation but might be particularly useful to exclude significant laryngeal edema \[[@B4],[@B5],[@B7],[@B16]\]. In contrast, other authors concluded that the cuff-leak test is inaccurate \[[@B8]\]. Indeed, a cuff-leak volume (measured conventionally) of more than 300 ml has been observed in three patients who developed post-extubation stridor after cardiac surgery \[[@B8]\]. Although these different results between studies might be due to the populations studied, our study indicates that the respiratory system mechanics and inspiratory flow, factors influencing the inspiratory leak that were not taken into account, might to some extent contribute to the poor performance of the cuff-leak test.
A measured conventional cuff-leak volume of less than 15.5% \[[@B4]\], 12% \[[@B7]\] or 10% of predetermined *V*~T~\[[@B6]\] has been used to identify patients at risk for post-extubation stridor. In our study with the conventional method, 5 of 15 patients had a cuff-leak volume less than 15.5% of predetermined *V*~T~, whereas with the pause method 11 patients demonstrated true cuff-leak volume less than this threshold (10 patients had a cuff-leak volume less than 12%). The purpose and design of our study were such that they did not permit us to examine whether by eliminating the inspiratory leak it would be possible to improve the predictive value of the cuff-leak test. The number of patients was small and the cuff-leak volume was not determined on the day of extubation, but the patients were examined under highly controlled conditions. The aim of the study was not to propose a new method of cuff leak determination but to examine factors affecting the total cuff-leak volume obtained by the conventional method. Our results clearly showed that the cuff-leak test (particularly its inspiratory component) is influenced by factors other than the cross-sectional area of the trachea and/or the upper airways and thus the above-mentioned cut-off points of cuff-leak volume should be re-evaluated.
Conclusion
==========
Our study has shown that the cross-sectional area around the endotracheal tube is not the only determinant of the cuff-leak test. Respiratory system mechanics and inspiratory flow are other important determinants of the cuff-leak test, mainly through an effect on the inspiratory component of the total leak, complicating its interpretation.
Key messages
============
• Cross-sectional area around the endotracheal tube is not the only determinant of the cuff leak test.
• Respiratory system mechanics and inspiratory flow are the other important determinants of the cuff leak test, complicating its interpretation.
Abbreviations
=============
*C*= model airway compliance; *C*~rs~= end-inspiratory static compliance of the respiratory system (ml/cmH~2~O); ΔLeak = difference between Leak~conv~and Leak~pause~; Δ*P*~aw,peak~= difference between peak inspiratory *P*~aw~between methods; Δ*R*= difference between *R*~rs~and *R*~int~; Leak~conv~= cuff-leak volume obtained by the conventional method; Leak~pause~= cuff-leak volume obtained when the cuff was deflated at the end of the end-inspiratory pause; *P*~aw~= airway pressure; PEEP = positive end-expiratory pressure; *R*= model airway resistance; *R*~int~= minimum resistance of the respiratory system; *R*~rs~= maximum resistance of the respiratory system; *V*\' = flow at the airway opening; *V*\'~I~= inspiratory flow; *V*~T~= expired tidal volume; *V*~T,baseline~= expiratory *V*~T~measured by averaging five consecutive breaths; *V*~T,defl~= expiratory *V*~T~measured when cuff was deflated; *V*~T,pause~= expiratory tidal volume measured at the end of the end-inspiratory pause.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
GP designed the study and performed the statistics. CA collected the data from patients and from the model. EM and EK participated in data collection. DG designed the study, evaluated the data and drafted the manuscript. All authors read and approved the final manuscript.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Clinical study. Individual cuff-leak volume was measured when the cuff remained deflated both during inspiration and expiration (conventional method, Leak~conv~) and when the cuff was deflated at the end of 3 s of inspiratory pause (Leak~pause~). Notice that in all patients Leak~conv~is higher than Leak~pause~. Solid line, line of identity; broken line, regression line.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Lung model study, protocol I. ΔLeak (difference between Leak~conv~and Leak~pause~) is shown at given inspiratory flow (*V*\'~I~) as a function of cross-sectional area around the endotracheal tube in a simulated model of respiratory system disease. Filled circles, large cross-sectional area; open circles, small cross-sectional area. \*, Significantly different from the corresponding value at *V*\'~I~= 1 litre/s. ^+^, Significantly different from the corresponding value at *V*\'~I~= 0.8 litre/s. ^&^, Significantly different from the corresponding value for simulated restrictive respiratory system disease. ^\#^, Significantly different from the corresponding value for simulated normal respiratory system.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Lung model study, protocol II. ΔLeak (difference between Leak~conv~and Leak~pause~) is shown at constant inspiratory flow as a function of respiratory system mechanics in a simulated model of constant cross-sectional area around the endotracheal tube. *R*, model airway resistance (cmH~2~O/litre per second); *C*, model compliance (ml/cmH~2~O). \*, Significantly different from the corresponding value at *C*= 100 ml/cmH~2~O. ^+^, Significantly different from the corresponding value at *C*= 50 ml/cmH~2~O.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Baseline ventilator settings and patients\' respiratory system mechanics
:::
No. *V*~T~ Fr *C*~rs~ *R*~int~ *R*~rs~
------ -------- ------ --------- ---------- ---------
1 0.68 12.8 47.5 12.6 17.1
2 0.64 13.0 27.2 8.4 12.4
3 0.61 8.1 63.2 13.1 20.4
4 0.70 7.1 57.8 14.1 17.5
5 0.46 7.1 63.8 14.9 17.2
6 0.68 14.9 30.5 11.0 15.5
7 0.52 14.5 30.9 10.8 15.0
8 0.58 11.6 28.0 13.8 18.5
9 0.62 8.5 51.2 13.4 15.1
10 0.60 13.5 32.2 13.3 17.3
11 0.66 9.8 56.1 8.5 12.8
12 0.58 10.4 17.7 9.6 22.8
13 0.51 13.0 37.6 9.0 14.0
14 0.56 16.0 43.9 6.7 13.3
15 0.49 11.8 36.4 10.7 13.6
Mean 0.59 11.5 41.6 11.3 16.2
SD 0.07 2.9 14.4 2.5 2.9
*C*~rs~, end-inspiratory static compliance of the respiratory system (ml/cmH~2~O); Fr, ventilator frequency (breaths/min); *R*~int~and *R*~rs~, minimum and maximum inspiratory resistance (cmH~2~O/l per second), respectively; *V*~T~, tidal volume (litres).
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Model study: protocol A
:::
Parameter Normal pattern Restrictive pattern Obstructive pattern
------------------ ---------------- --------------------- --------------------- ---------- ------------ ------------- ---------- --------- --------------
Large area
Leak~pause~(ml) 191 ± 7 196 ± 6 190 ± 4 190 ± 13 190 ± 15 190 ± 6 196 ± 5 185 ± 6 187 ± 6
Leak~conv~(ml) 298 ± 6 315 ± 3^a^ 339 ± 4^ab^ 303 ± 6 330 ± 2^a^ 358 ± 2^ab^ 308 ± 7 309 ± 5 320 ± 10^ab^
Small area
Leak~pause~(ml) 146 ± 2 135 ± 5 135 ± 4 147 ± 8 148 ± 12 137 ± 4 146 ± 9 139 ± 6 141 ± 11
Leak~conv~(ml) 239 ± 7 228 ± 3 244 ± 4^ab^ 249 ± 10 243 ± 4 269 ± 7^ab^ 243 ± 14 234 ± 4 254 ± 6^ab^
Results are means ± SD. *V*\', constant inspiratory flow (litre/s); Leak~conv~, cuff-leak volume measured when the cuff remained deflated during both inspiration and expiration; Leak~pause~, cuff-leak volume measured when the cuff was deflated at the end of 3 s of inspiratory pause.
^a^Significantly different from the corresponding value at *V*\'~I~= 1 litre/s.
^b^Significantly different from the corresponding value at *V*\'~I~= 0.8 litre/s.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Model study: protocol B
:::
Parameter *R*= 8 *R*= 16 *R*= 32
----------------- ------------- --------- --------- ------------- ---------- --------- -------------- --------- -------------
Leak~pause~(ml) 96 ± 9 99 ± 6 96 ± 9 105 ± 10 103 ± 11 110 ± 8 123 ± 12^c^ 115 ± 9 118 ± 12^c^
Leak~conv~(ml) 275 ± 11^a^ 257 ± 9 245 ± 8 278 ± 6^ab^ 261 ± 10 253 ± 9 287 ± 13^ab^ 268 ± 7 255 ± 6
Results are means ± SD. *C*, model compliance (ml/cmH~2~O); Leak~conv~, cuff-leak volume measured when the cuff remained deflated during both inspiration and expiration; Leak~pause~, cuff-leak volume measured when the cuff was deflated at the end of 3 s of inspiratory pause; *R*, model resistance (cmH~2~O/litre per second).
^a^Significantly different from the corresponding value at *C*= 100 ml/cmH~2~O.
^b^Significantly different from the corresponding value at *C*= 50 ml/cmH~2~O.
^c^Significantly different from the corresponding value at *R*= 8 cmH~2~O/litre per second.
:::
|
PubMed Central
|
2024-06-05T03:55:54.890772
|
2004-11-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065107/",
"journal": "Crit Care. 2005 Nov 29; 9(1):R24-R31",
"authors": [
{
"first": "George",
"last": "Prinianakis"
},
{
"first": "Christina",
"last": "Alexopoulou"
},
{
"first": "Eutichis",
"last": "Mamidakis"
},
{
"first": "Eumorfia",
"last": "Kondili"
},
{
"first": "Dimitris",
"last": "Georgopoulos"
}
]
}
|
PMC1065109
|
Introduction
============
Affecting about 700,000 people annually, sepsis accounts for 210,000 deaths each year in the USA, and both of these figures are likely to increase \[[@B1],[@B2]\]. Sepsis is not an homogenous disease; rather, it is a complex clinical syndrome with distinct immunological features \[[@B3],[@B4]\]. The ambiguity of clinical findings and unclear risk stratification in sepsis have been major problems in sepsis intervention trials \[[@B5]\]. The effectiveness of anti-inflammatory treatment correlates with risk for death and severity of disease \[[@B6]\]. Thus, the prognosis of a septic patient may contribute significantly to the success of any intervention \[[@B5]\]. Within this context, there is need for biomarkers to tackle the challenges of sepsis monitoring and treatment \[[@B7]\].
Members of the natriuretic peptide family are established markers of congestive heart failure \[[@B8]-[@B10]\]. Defending against hypertension and salt and water retention, they antagonize the renin--angiotensin--aldosterone system, including effects on renal tubule sodium reabsorption, vascular tone and cell growth. Atrial natriuretic peptide (ANP) is predominantly produced in the atrium of the heart and comprises 98% of natriuretic peptides in the circulation \[[@B11]\]. Recently, both ANP and pro-ANP have attracted interest as new markers in the field of sepsis \[[@B12]-[@B16]\]. Mature ANP is derived from carboxyl-terminal amino acids 99--126 of the prohormone (pro-ANP), which is 126 amino acids in length \[[@B11]\]. The amino-terminal portion of pro-ANP (termed NT-pro-ANP, or pro-ANP~1--98~) is secreted at the same molar ratio as ANP. Because it has a much longer half-life than has mature ANP, it has been suggested that pro-ANP~1--98~is a more reliable analyte \[[@B17]\]. However, results from various competitive immunoassays and high-performance liquid chromatography analyses indicate that pro-ANP~1--98~may be subject to further fragmentation \[[@B18],[@B19]\]. Consequently, sandwich immunoassays for pro-ANP~1--98~might underestimate actual levels of pro-ANP, and immunoassays for measurement of mid-regional pro-ANP may have an advantage \[[@B20]\].
In the present study we aimed to evaluate the prognostic value of mid-regional pro-ANP levels in a well defined cohort of medical intensive care unit (ICU) patients as compared with those of other biomarkers (i.e. IL-6, C-reactive protein \[CRP\] and procalcitonin \[PCT\]) and a physiological score (Acute Physiology and Chronic Health Evaluation \[APACHE\] II).
Methods
=======
Patients
--------
In the present study we evaluated plasma samples from a cohort of 101 consecutive critically ill patients admitted to the medical ICU of the University Hospital of Basel, Switzerland. The primary end-point of this study was the prognostic value of endocrine dysfunction in critically ill patients (\'PEDCRIP\' study). The characteristics of the study population, study design, diagnostic criteria and levels of various markers of inflammation and infection were reported in detail elsewhere \[[@B21]-[@B24]\]. Briefly, over a 9-month period 101 consecutive patients, including neutropenic and immunosuppressed patients, admitted to the medical ICU were included. Patients were followed until hospital discharge or death.
Data were collected on admission (i.e. during the first 24 hours), on day 2, and on the day of discharge from the ICU or on the day of death. At these time points (a total of 276 plasma samples), patients were either very sick or in a stable condition and ready for discharge to a medical ward, respectively. In patients who died within 24 hours after admission, only data from admission were collected (*n*= 5). Vital signs, clinical status and severity of disease parameters (APACHE II score) were assessed daily. The APACHE II score was calculated by means of maximal daily deviations of 12 physiological variables from normal plus correction for age and various chronic illnesses. A pulmonary artery catheter was not routinely inserted. When feasible, consent was obtained from conscious patients before enrolment; otherwise, consent was obtained from the next of kin. The study protocol had been granted approval by the hospital institute\'s ethical review board.
Patients were classified at the time of blood collection into those with systemic inflammatory response syndrome (SIRS), sepsis, severe sepsis and septic shock, which were defined according to international criteria \[[@B25],[@B26]\]. Infection was diagnosed according to standardized criteria or, in case of uncertainty, by an infectious disease specialist. This was done retrospectively on the basis of review of complete patient charts, results of microbiological cultures, chest radiographs and, when available, autopsy reports. An isolated micro-organism was considered to be pathogenic if it was identified within a 24-hour period before or after the onset of the systemic response. Colonization with bacteria (e.g. in a patient with a bladder catheter but without leucocyturia) or positive blood cultures at autopsy were disregarded. Microbiological tests were requested and antibiotic therapy prescribed by physicians on duty according to the usual practice, without interference from the research team.
Although optimal fluid resuscitation was done in the initial treatment phase in all patients, 31% of septic patients needed additional treatment with intravenous noradrenaline (norepinephrine). The mean dose of noradrenaline on admission was 8.7 ± 12.1 μg/min, on day 2 it was 10.1 ± 10.9 μg/min and on the day of discharge/death it was 47.2 ± 35.2 μg/min (*P*\< 0.001). Nonsurvivors from severe sepsis and septic shock needed higher doses of noradrenaline than did survivors (5.7 ± 7.8 μg/min versus 30.5 ± 28.1 μg/min; *P*\< 0.001). Overall, 23 of the 101 patients died (22.8%). The majority of patients who died suffered from multiple organ failure (56.5%), defined as failure of two or more vital organs.
Assays
------
Results of the routine blood analyses (i.e. complete blood count, serum chemistry including CRP, blood gas analyses) were known and recorded. Blood was obtained from an indwelling arterial or venous catheter. Plasma was separated from the blood samples at the time of blood draw and frozen at -70°C until assayed. Measurement was done in a blinded manner as a batch analysis.
Mid-regional pro-ANP (epitopes covering amino acids 53--90) was detected in EDTA plasma from all patients with a new sandwich immunoassay (BRAHMS Seristra^®^LIA; BRAHMS AG, Hennigsdorf/Berlin, Germany), as described in detail elsewhere \[[@B20]\]. As a modification to the published assay, the calibration was changed from a synthetic peptide to pro-ANP in human serum. This modification to the initial description increased the precision and dynamic (i.e. signal to noise ratio) of the assay, and allowed measurement of pro-ANP in serum and plasma (with EDTA, heparin, or citrate). Briefly, patient samples (1:40 dilution of 5 μl plasma in incubation buffer) or standards were added in duplicate to antibody-coated tubes (directed at pro-ANP peptide 73--90) and incubated for 30 min at room temperature. After five washings with 1 ml washing buffer, 200 μl tracer was added, containing acridinium ester-labelled anti-pro-ANP antibody (directed at peptide 53--72), followed by 30 min incubation at room temperature. Tubes were washed three times with 1 ml washing buffer, and detection was performed in a luminometer (1 s detection time per sample). Relative light units of the chemiluminescence assay were expressed in pmol/L pro-ANP, as calculated from a calibration curve (4--1800 pmol/l) that was included in every analytical run. The lower detection limit of the assay is 4.3 pmol/l and the functional sensitivity of the assay (interassay coefficient of variation \<20 %) is 11 pmol/L pro-ANP. The 97.5th percentile in 325 healthy individuals was 163.9 pmol/l (median 45 pmol/l), with no difference between sexes \[[@B20]\].
PCT was measured using the LUMITest^®^PCT (BRAHMS AG), following the manufacturer\'s instructions. CRP was determined using en enzyme immunoassay (EMIT; Merck Diagnostica, Zurich, Switzerland). A serum level greater than 5 mg/l was considered abnormally elevated. Serum IL-6 concentrations were measured using a commercially available quantitative sandwich enzyme immunoassay (Pelikine Compact™; CLB, Amsterdam, The Netherlands), with a limit of detection at 0.6 ng/l.
Statistical analysis
--------------------
Data in the text are expressed as mean ± standard deviation. Frequency comparison was done by χ^2^test. Two-group comparisons were performed using the Mann--Whitney U-test. For multigroup comparisons, Kruskal--Wallis one-way analysis of variance was used with Dunn\'s post-test evaluation. Levels that were nondetectable were assigned a value equal to the lower limit of detection for the assay. All testing was two-tailed, and *P*\< 0.05 was considered statistically significant. Correlation analyses were performed by using Spearman rank correlation.
Results
=======
Descriptive characteristics of the patients
-------------------------------------------
The mean age of the 101 patients (55 men and 46 women) included in the study was 57 ± 15 years (range 23--86 years) and the mean APACHE II score on admission was 22 ± 8. The median length of stay in the medical ICU was 4 days (range 0.2--60 days) and the mortality rate was 23%. More detailed baseline characteristics of the study population are described elsewhere \[[@B21]\]; however, to allow better understanding of the study results, the principal diagnoses of patients are summarized in Table [1](#T1){ref-type="table"} and the sites of infection in Table [2](#T2){ref-type="table"}. Sepsis was diagnosed in 58% of patients (on admission in 53 patients; five additional patients developed sepsis during their stay in the ICU). The principal site of infection was the lung (Table [2](#T2){ref-type="table"}). In 38 (66%) of the 58 patients with infections, the responsible micro-organism was identified and 14 patients (24%) had bacteraemia. There was no difference in mortality between patients with and those without infection. Of the 53 patients admitted with sepsis, severe sepsis, or septic shock, 13 (25%) died; 10 (21%) of the 48 patients without infection on admission died.
Pro-atrial natriuretic peptide and severity of the disease
----------------------------------------------------------
Figure [1a](#F1){ref-type="fig"} shows the distribution of pro-ANP values according to severity of infection (i.e. SIRS, sepsis, severe sepsis and septic shock) and serum PCT concentrations. Depending on the clinical severity of the infection, pro-ANP values exhibited a gradual increase from the group with SIRS to the group with septic shock (*P*\< 0.001). Similarly, circulating pro-ANP levels showed a similar gradual increase when categorized based on PCT levels (Fig. [1b](#F1){ref-type="fig"}).
Post-test analysis revealed a significant difference (*P*\< 0.001) between patients without SIRS, SIRS, sepsis and severe sepsis as compared with patients with septic shock. There was no significant difference between patients with severe sepsis and those with septic shock. Accordingly, patients with PCT levels greater than 10 ng/ml and greater than 1 ng/ml had significantly higher pro-ANP levels than did patients with PCT levels of 0.5--1 ng/ml and under 0.5 ng/ml (*P*\< 0.001).
Pro-ANP levels correlated with serum IL-6 levels (r = 0.22; *P*\< 0.001), and with serum and urine osmolarity (r = 0.55 and r = -0.43, respectively; *P*\< 0.001), but not with serum sodium (r = 0.03; not significant) and only weakly with urine sodium concentrations (r = -0.17; *P*\< 0.01).
Pro-atrial natriuretic peptide and outcomes in patients with sepsis, severe sepsis and septic shock
---------------------------------------------------------------------------------------------------
Figure [2](#F2){ref-type="fig"} shows all pro-ANP values in survivors and nonsurvivors with sepsis, severe sepsis or septic shock, measured during their stay in the ICU. Thereby, patients were grouped by clinical diagnosis of sepsis according to international guidelines (panels a and c) or by circulating PCT level in excess of 1 ng/ml (panels b and d). The median pro-ANP value in the nonsurvivors was significantly greater than in the survivors, independent of grouping used. This difference in pro-ANP values was clear on the first day of admission to the ICU (*P*\< 0.001). In contrast, the difference between the survivors and nonsurvivors on the first day of admission was not significant for PCT (*P*= 0.38 and *P*= 0.05, respectively), CRP, or IL-6 (data not shown for CRP and IL-6). Similarly, in patients without infections pro-ANP values were not higher in nonsurvivors than in survivors (all time points: 197.2 ± 361.5 pmol/l versus 226.0 ± 183.4 pmol/l, *P*= 0.7; on admission: 221.5 ± 209.7 pmol/l versus 161.3 ± 132.1 pmol/l, *P*= 0.3).
To define an optimal decision threshold for pro-ANP values in septic patients, we performed receiver operating characteristic (ROC) plot analysis, including only data from patients with sepsis, severe sepsis, or septic shock obtained within the first 48 hours after admission to the ICU. Sensitivity was calculated among those patients who did not survive sepsis, and specificity was assessed among those patients who were discharged from the ICU. For comparison, the same ROC plot analysis was performed with CRP, PCT, IL-6 and APACHE II score. Table [3](#T3){ref-type="table"} shows the area under the ROC curve (AUC) for all parameters, including the 95% confidence interval. The AUC for pro-ANP was 0.88, which was significantly higher than the AUCs for PCT and CRP, and similar to the AUC for the APACHE II score (0.86). ROC curves are shown in Fig. [3](#F3){ref-type="fig"}. Again, patients were grouped by clinical diagnosis of sepsis according to international guidelines (panel a) or by circulating PCT levels in excess of 1 ng/ml (panel b), yielding comparable results.
The optimal threshold for pro-ANP was 530 pmol/l. At this cut-off, the sensitivity for correct prediction of death in the ICU was 86.7% and the specificity was 88.6%. Considering a prevalence of 33% for death in the ICU as a result of sepsis, the positive predictive value (PPV) of pro-ANP was 72.2% with a negative predictive value (NPV) of 95.1%. None of CRP, PCT, or IL-6 had similarly high values for sensitivity, specificity, PPV and NPV.
The APACHE II score was also predictive for prognosis but yielded lower values as compared with pro-ANP. At an APACHE II threshold of 30, the sensitivity was 73.3% and the specificity was 95.6% (PPV = 84.6 %, NPV = 91.5 %). At a cut-off of 25, which was recommended by the US Food and Drug Administration for the use of Xigris^®^, sensitivity was 80.0%, specificity 75.6%, PPV 48.0% and NPV 91.4%.
Because PPV and NPV are dependent on the prevalence of the disease, Table [4](#T4){ref-type="table"} shows the relative likelihood with the prevalence independent likelihood ratio for different cut-offs.
Discussion
==========
ANP and pro-ANP are markers for congestive heart failure \[[@B8]-[@B10]\], but their pathophysiological and prognostic significance in severe sepsis and septic shock is not yet understood. In the present study we found a significant increase in mid-regional pro-ANP in the plasma of sepsis patients as compared with patients without sepsis and healthy individuals. This increase was most marked in those patients with sepsis who did not survive their disease. Importantly, on the first day of admission to the ICU, pro-ANP, but not other markers of infection and inflammation such as CRP and PCT, were significantly increased in nonsurvivors as compared with survivors, suggesting that pro-ANP levels represent a new and valuable prognostic tool in patients with sepsis. At a threshold of 530 pmol/l, pro-ANP had a sensitivity of 86.7% for death in the ICU with sepsis, with a specificity of 88.6%; these figures were not reached by any of the other tested biomarkers.
As is generally recommended, we diagnosed sepsis, severe sepsis and septic shock using well defined and widely accepted clinical guidelines \[[@B25],[@B26]\]. However, true gold standards for the diagnosis of infections do not exist, and clinical classification of critically ill patients is not 100% certain despite the use of these guidelines, not only in sepsis trials but also in routine bedside use \[[@B27],[@B28]\]. An ideal sepsis marker should permit early diagnosis, should provide information about the course of disease, and should help to differentiate bacterial from noninfectious and viral causes of systemic inflammation. It was shown that PCT has some of these features and is helpful in diagnosing septic conditions \[[@B29]-[@B31]\]. Therefore, we also classified pro-ANP levels according to circulating PCT levels, which are not subject to the uncertainty associated with clinical sepsis definitions. Importantly, the prognostic value of pro-ANP was similar independent of the classification system used, which suggests that our findings are reproducible. Thus, pro-ANP is of prognostic value in critically ill septic patients, in contrast to PCT, which is predominantly a diagnostic parameter.
The first observations that ANP may play a role during endotoxic shock came from animal studies in which ANP was elevated within 2--6 hours after lipopolysaccharide injection \[[@B32],[@B33]\]. Subsequent studies in critically ill humans showed an association of ANP with various cardiac physiological parameters \[[@B34],[@B35]\].
The use of different assays might be responsible for part of the inconsistency in reported findings over recent years. Whereas Berendes and coworkers \[[@B14]\] found no association of ANP values with severity of the disease or mortality in critically ill patients, Hartemink and coworkers \[[@B13]\] found a strong association of ANP levels with myocardial depression in septic shock and with lethal outcome in 14 patients. A similar association of cardiac depression in septic shock was described for NT-pro-ANP in 17 patients \[[@B12]\].
Unfortunately, a limitation of our study is that cardiac indices were not routinely assessed by pulmonary artery catheter. Therefore, the precise mechanisms of pro-ANP release in patients with sepsis remain unknown. Nevertheless, a cardiac origin of natriuretic peptides makes an association with septic cardiac dysfunction likely. In addition, apart from volume overload, osmolarity rather than sodium concentration is associated with pro-ANP release, as suggested by regression analyses in our patients. Based on our findings and recent reports in the literature \[[@B36]\], in critically ill patients increased levels of natriuretic peptide are not specific for decompensated heart failure. In this context, the increase in ANP levels in septic shock may be potentiated by IL-6 elevation \[[@B15]\]. A recent study in meningococcal sepsis provided conclusive evidence that IL-6 is directly involved in myocardial depression \[[@B37]\]. Accordingly, in the present study IL-6 levels were correlated with pro-ANP levels, albeit relatively weakly. IL-6 had lower value in terms of outcome prediction than did mid-regional pro-ANP, which may be due to differences in the half-life of the molecules. The half-lives of both IL-6 and mature ANP are short, and measurement of those markers in septic patients does not allow a direct conclusion to be drawn regarding the level of production. We recently developed a sandwich immunoassay for the detection of a mid-regional fraction of pro-ANP in plasma \[[@B20]\]. This fragment has a much longer half-life in plasma, and because it is produced in equimolar concentrations to the mature hormone, it mirrors true production of ANP. Furthermore, it is possible that mid-regional pro-ANP exerts a physiological effect on its own, as is described for other fragments of NT-pro-ANP \[[@B38]\] and fragments of other prohormones, such as pro-adrenomedullin amino-terminal 20 peptide \[[@B39]\].
Measurement for ANP or fragments of NT-pro-ANP is potentially influenced by other factors, such as sex, age and kidney function, as is discussed elsewhere for brain-type natriuretic peptides \[[@B40],[@B41]\]. Indeed, we observed a significant correlation of circulating pro-ANP levels with serum osmolarity and creatinine. Measurements in nonseptic patients with kidney failure revealed mostly normal pro-ANP values, and it is therefore possible that the observed elevation in pro-ANP and creatinine in this study is a result of kidney failure related to sepsis.
Sepsis is a complex syndrome, and the immunological and biochemical situation may vary considerably between individual patients \[[@B3],[@B4]\]. In the past almost all intervention trials failed to show any benefit from therapy for sepsis, and sepsis intervention has been termed the \'graveyard for pharmaceutical companies\' \[[@B7],[@B42]\]. Reasons for this may be found in immunological heterogeneity and insufficient patient stratification in those trials \[[@B5]\]. The need for markers that permit better stratification of patients with different stages of sepsis is underlined by the ongoing discussion concerning recombinant human activated protein C (drotrecogin alpha; Xigris^®^) \[[@B42]-[@B45]\]. The US Food and Drug Administration approved recombinant human activated protein C only for those patients with an APACHE II score in excess of 24, and thus only for those patients with the greatest risk for dying \[[@B28],[@B46],[@B47]\]. The APACHE II score -- a complex algorithm -- was not originally developed for individual outcome prediction in sepsis patients \[[@B48]\]. Despite its limitations, outcome predictors such as the extensively evaluated APACHE II score are helpful in identifying those septic patients who are at high risk for death and who are more likely to benefit from intervention \[[@B6]\]. In the present study the prognostic value of pro-ANP levels was comparable to that of APACHE II score. Importantly, mid-regional pro-ANP it is easier to determine than a physiological score and mirrors distinct pathophysiological changes that occur in sepsis.
Conclusion
==========
In septic patients, we found that APACHE II score and mid-regional pro-ANP level on admission to a medical ICU had similar ability to predict outcome. The results of our study are novel and of interest because they may help to improve stratification of septic patients. Our findings are descriptive in nature and warrant validation in future prospective studies, including measurement of cardiac indices or evaluating patients who have undergone surgery. If our findings are confirmed, then mid-regional pro-ANP might become a new and useful additional prognostic marker for individual risk assessment in sepsis, and may represent a helpful tool for patient stratification in future intervention trials.
Key messages
============
• In septic patients mid-regional pro-ANP levels on admission to a medical ICU had a similar ability to predict outcome as did the APACHE II score.
• Pro-ANP levels appear to be a useful tool for individual risk assessment in septic patients and for stratification of high risk patients in future intervention trials.
• Because our findings are descriptive in nature, further prospective studies are warranted to validate our results.
Abbreviations
=============
ANP = atrial natriuretic peptide; APACHE = Acute Physiology and Chronic Health Evaluation; AUC = area under the curve; CRP = C-reactive protein; ICU = intensive care unit; IL = interleukin; NPV = negative predictive value; NT = amino terminal; PCT = procalcitonin; PPV = positive predictive value; ROC = receiver operating characteristic; SIRS = systemic inflammatory resonse syndrome.
Competing interests
===================
NG, JS and AB are employees of BRAHMS AG, the manufacturer of the pro-ANP assay (BRAHMS Seristra^®^LIA; BRAHMS AG, Hennigsdorf/Berlin, Germany). BM has served as a consultant and received payments from BRAHMS AG to attend meetings related to the trial and for travel expenses, speaking engagements and research.
Authors\' contributions
=======================
BM conceived the study, collected the data, drafted the protocol and supervised the writing of the manuscript. NGM, JS and AB were involved in assay development. NGM and MCC conducted statistical analyses and wrote the report. All authors read and approved the final manuscript.
Acknowledgements
================
The authors wish to thank Dr Barbara Thomas for helpful discussion, the Laboratory of Chemical Pathology of the University Hospital Basel, and Professor Peter Huber, Dr Marc A Viollier, Uwe Zingler, Frank Bonconseil and Margret Schröder for excellent technical assistance.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Pro-atrial natriuretic peptide (ANP) according to severity of disease and circulating procalcitonin (PCT) levels. All patient data were grouped according to **(a)**the severity of the disease following consensus criteria (\'no SIRS\',\' SIRS\', \'sepsis\', \'severe sepsis\' and \'septic shock\') or **(b)**circulating PCT concentrations. Data from all time points (i.e. on admission, day 2, day of discharge and death) are displayed. Solid lines denote median values, boxes represent 25th to 75th percentiles and whiskers indicate the range. ANOVA, analysis of variance.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Pro-atrial natriuretic peptide (ANP) and procalcitonin (PCT) levels in surviving as compared with nonsurviving patients. Data from the patients on admission are shown. Patients were grouped **(a, c)**by clinical diagnosis of sepsis according to international guidelines or **(b, d)**by circulating PCT levels in excess of 1 ng/ml. Solid lines denote median values, boxes represent 25th to 75th percentiles and whiskers indicate the range.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Receiver operating characteristic plot analysis of different biomarkers with respect to outcome prediction of sepsis. Patient data on admission were grouped by **(a)**clinical diagnosis of sepsis according to international guidelines or by **(b)**circulating procalcitonin (PCT) levels in excess of 1 ng/ml. Sensitivity was calculated in nonsurvivors, and specificity in survivors. APACHE, Acute Physiology and Chronic Health Evaluation; CRP, C-reactive protein; PCT, procalcitonin.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Clinical diagnoses of the patients
:::
Diagnosis Details Number of patients^a^
---------------- ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -----------------------
Respiratory Pneumonia (33), chronic obstructive pulmonary disease (14), acute asthma (3), bronchial carcinoma (3), pneumothorax (3), pharyngeal obstruction (2), toxic pulmonary oedema (2), Wegener\'s granulomatosis (1) 61
Cardiovascular Myocardial infarction (12), heart failure (11), pulmonary embolism (2) haemorrhagic shock (1) 26
Abdominal Gastrointestinal bleeding (7), abdominal infection (6), urinary tract infection (5), acute renal failure (3), hepatic coma (3) 24
Cerebral Ischaemic stroke (5), subarachnoid (4) or intracerebral (3) haemorrhage, seizures (3), suicidal intoxication with sedatives (5), cavernous sinus thrombosis (1) 21
Others Leukaemia (7), postoperative (6), diabetic coma (3), other infections (3) 19
^a^One patient can have more than one diagnosis, and so the total exceeds the absolute number of patients (*n*= 101).
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Site of infection and microbiology
:::
Site of infection Details Number of patients^a^
-------------------------------------------------------------------- -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- -----------------------
Lung *Streptococcus pneumoniae*(6), *Pseudomonas aeruginosa*(5), *Haemophilus influenzae*(3), *Streptococcus pyogenes*(3), *Staphylococcus aureus*(3), *Klebsiella pneumoniae*(2), *Escherichia coli*(2), *Enterobacter*spp. (2), *Streptococcus salivarius*(1), *Legionella pneumophilia*(1), unknown (16) 44
Urinary tract *Escherichia coli*(5), *Pseudomonas aeruginosa*(1) 6
Abdominal (gastrointestinal tract, liver, bile duct, and pancreas) *Clostridium difficile*-associated colitis (1), unknown (4) 5
Others Meningococcal meningitis (1), sepsis with *Torulopsis glabrata*(1), malaria with *Plasmodium falciparum*(1) 3
^a^An infection was diagnosed in 58% of the patients (on admission in 53 patients; five additional patients developed sepsis during their stay in the medical intensive care unit).
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Area under the curve of receiver operating characteristic plot analysis
:::
Parameter AUC 95% CI *P*(versus pro-ANP)
----------- ------ ------------ ---------------------
Pro-ANP 0.88 0.77--0.95 \-
APACHE II 0.86 0.74--0.93 0.79
IL-6 0.79 0.66--0.88 0.34
PCT 0.67 0.53--0.78 0.027
CRP 0.51 0.38--0.64 \< 0.001
ANP, atrial natriuretic peptide; APACHE, Acute Physiology and Chronic Health Evaluation; AUC, area under the curve; CI, confidence interval; CRP, C-reactive protein; PCT, procalcitonin.
:::
::: {#T4 .table-wrap}
Table 4
::: {.caption}
######
Sensitivity, specificity, positive likelihood ratio, negative likelihood ratio and odds ratio at different cut-off levels of pro-ANP
:::
Pro-ANP cut-off (pmol/l) Sensitivity (95% CI) Specificity (95% CI) LR^+^ LR^-^ Odds ratio (95% CI)
-------------------------- ---------------------- ---------------------- ------- ------- ---------------------
350 93% (68--99%) 77% (62--88%) 4.1 0.09 47.6 (5.5--408)
530 86% (59--98%) 88% (75--96%) 7.6 0.15 50.7 (8.7--293)
700 73% (44--92%) 93% (81--98%) 10.7 0.29 37.5 (7.3--193)
ANP, atrial natriuretic peptide; CI, confidence interval; LR^-^, negative likelihood ratio; LR^+^, positive likelihood ratio;
:::
|
PubMed Central
|
2024-06-05T03:55:54.893527
|
2004-12-17
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065109/",
"journal": "Crit Care. 2005 Dec 17; 9(1):R37-R45",
"authors": [
{
"first": "Nils G",
"last": "Morgenthaler"
},
{
"first": "Joachim",
"last": "Struck"
},
{
"first": "Mirjam",
"last": "Christ-Crain"
},
{
"first": "Andreas",
"last": "Bergmann"
},
{
"first": "Beat",
"last": "Müller"
}
]
}
|
PMC1065112
|
Introduction
============
Tracheostomy is among the most frequently performed procedures in critically ill patients, being done in about 24% of patients in medical intensive care units (ICUs) \[[@B1]\]. The most common indication for tracheostomy in the ICU is need for prolonged mechanical ventilation \[[@B2],[@B3]\]. Tracheostomy has several advantages over endotracheal intubation, including lower airway resistance, smaller dead space, less movement of the tube within the trachea, greater patient comfort and more efficient suction \[[@B4],[@B5]\]. Although recent studies have suggested that tracheostomy can be a safe procedure in the ICU \[[@B6],[@B7]\], tracheostomy has also been found to lead to serious complications, including tracheal stenosis, increased bacterial colonization and haemorrhage \[[@B8],[@B9]\]. Many critically ill patients\' families have been hesitant in authorizing tracheostomy because of cosmetic issues and speech problems.
Because there are no definitive guidelines available, the timing of tracheostomy depends on clinical conditions, physician judgement and communication with families. The judgement of the attending physician can be influenced by the patients\' likelihood of extubation, life expectancy and other clinical conditions, including haemodynamic status, oxygenation, consciousness level and ability to protect the airway. There is little consensus on the timing of tracheostomy. In the 1989 American College of Chest Physicians (ACCP) Consensus Conference on Artificial Airways in Patients Receiving Mechanical Ventilation \[[@B10]\], it was concluded that the appropriate duration of translaryngeal intubation could not be defined. It was suggested that if the anticipated need for mechanical ventilation is longer than 21 days then tracheostomy is preferable. For mechanical ventilation that is anticipated to last between 10 and 21 days, the decision was left to the physician, and daily assessment was recommended. Recent ACCP guidelines \[[@B11]\] suggest that tracheostomy should be considered after an initial period of stabilization on the ventilator, when it becomes apparent that the patient will require prolonged ventilator assistance.
Maziak and coworkers \[[@B12]\] reviewed five reports on the timing of tracheostomy and concluded that there was insufficient evidence to conclude that the timing of tracheostomy alters the duration of mechanical ventilation. However, there is still a lack of data on the relationship between the timing of tracheostomy and weaning from mechanical ventilation for patients in the medical ICU. Therefore, we investigated the timing of tracheostomy and other factors that might influence weaning from mechanical ventilation and outcomes of patients admitted to the medical ICU.
Methods
=======
Patients
--------
Over a period of 36 months (from July 1998 to June 2001), all adult patients admitted to the medical ICU of National Taiwan University Hospital -- a 1500-bed tertiary medical centre that accommodates tracheostomy within the ICU -- were considered for inclusion in the study. Patients were excluded if the tracheostomy was performed in an emergency setting because of difficulties with the airway or other causes. Tracheostomy was performed using standard surgical techniques at bedside in the ICU, and no patients underwent percutaneous tracheostomy. The timing of tracheostomy depended on the attending physician\'s decision. Indications to initiate an attempt to wean a patient from mechanical ventilation included stable haemodynamic status, improved oxygenation (arterial oxygen tension \[PaO~2~\]/fractional inspired oxygen \[FiO~2~\] ratio \>150), controlled infection and lack of need for further intervention. The weaning process was begun with synchronized intermittent mandatory ventilation with pressure support. Then, patients underwent continuous positive airway pressure with pressure support, or intermittent T-piece for a spontaneous breathing trial when clinical conditions improved. Successful weaning was defined as weaning from mechanical ventilation for more than 72 hours. Patients were transferred to long-term care settings once tracheostomy and the weaning process were completed if there was no other active clinical disease.
Data collection
---------------
The indications for intubation were defined as any major problem(s) that necessitated intubation. The underlying disease of the patients, including diabetes mellitus, hypertension, congestive heart failure, chronic renal insufficiency, chronic obstructive pulmonary disease and malignant disease with lung metastasis, were ascertained through chart reviews. Medical records were analyzed for age, sex, underlying disease and cause of intubation, Acute Physiology and Chronic Health Evaluation (APACHE) II score \[[@B13]\], duration of mechanical ventilation, complications of tracheostomy, pneumonia after tracheostomy, length of ICU stay, and mortality in the ICU and hospital. APACHE II scores were calculated using clinical data, which were available from the first 24 hours of intensive care. Clinical data within 72 hours before tracheostomy, including PaO~2~/FiO~2~ratio, peripheral white blood cell (WBC) counts, haemoglobin, creatinine and albumin, were also recorded and analyzed. Old age was defined as age above 65 years. Anaemia was defined as haemoglobin below 10 g/dl, and leucocytosis was defined as a WBC count above 11,000/ μl before tracheostomy. Renal insufficiency was defined as creatinine above 1.5 mg/dl, and poor oxygenation as PaO~2~/FiO~2~ratio below 250.
Complications of tracheostomy, including bleeding, air leakage, pneumothorax, subcutaneous emphysema, cardiopulmonary arrest, dislodgement of the tube, obstruction, tracheal stenosis, granuloma, tracheo-oesophageal fistula and tracheomalacia, were recorded. Complications that occurred within 7 days after tracheostomy were defined as early complications; those occurring later were considered late complications. Severity of bleeding after tracheostomy was classified as follows: minor if there was only minimal blood clot over the wound or if new onset bloody sputum was noted on the next day of the tracheostomy; moderate if bleeding needed external compression and component therapy or surgical management; and massive if the bleeding resulted in obvious haemodynamic change. The clinical definition of post-tracheostomy pneumonia used was as follows \[[@B14]\]: new and persistent radiographic opacity found after the tracheostomy had been removed and within 48 hours into the weaning period; positive sputum culture; and three of body temperature above 38°C, WBC count above 15,000/μl, increased airway secretions, or worsening gas exchange.
Statistical analysis
--------------------
Values are expressed as mean ± standard deviation (continuous variables) or as a percentage of the group from which they were derived (categorical variables). Only variables with complete data were analyzed in the study. Differences in the groups, including sex, underlying diseases and associated medical conditions, indications for intubation, occurrence of post-tracheostomy pneumonia, successful weaning and mortality, were analyzed using χ^2^test. Other variables, including age, sex, APACHE II score, the length of ICU stay, PaO~2~/FiO~2~ratio, peripheral WBC count, haemoglobin, albumin and weaning period, were analyzed by an independent t-test. The correlations between the intubation period and the length of ICU stay were analyzed using a Pearson bivariate correlation test. The correlations between successful weaning and potentially influential factors, including old age, sex, presence of comorbidities, indications for intubation, leucocytosis, anaemia, thrombocytopenia, renal insufficiency, poor oxygenation, post-tracheostomy pneumonia and timing of tracheostomy, were analyzed using the Kaplan--Meier method with a log rank test. Censoring was performed for those patients who died during mechanical ventilation. A Cox regression model was applied for multivariate analysis with variables that were significantly associated with successful weaning in the univariate analysis. *P*\< 0.05 was considered statistically significant.
Results
=======
Clinical characteristics
------------------------
From July 1998 through June 2001, a total of 167 patients who underwent tracheostomy in the medical ICU were included in the study. Four patients were excluded because of emergent tracheostomy due to difficult airway (*n*= 3) or laryngeal oedema (*n*= 1). Thus, 163 patients were included (93 male and 70 female; mean age 70 years, range 19--104 years; Table [1](#T1){ref-type="table"}). The indications for intubation in the 163 patients were classified into four categories: pulmonary (*n*= 107), infectious (*n*= 18), neurological (*n*= 28) and circulatory (*n*= 10) disease. The most common cause of intubation was pneumonia with respiratory failure (*n*= 81 \[73%\]). The mean APACHE II score within the first 24 hours after ICU admission was 20.0 ± 7.2. The mean duration of intubation was 18.5 ± 10.9 days (range 1--62 days).
Complications
-------------
The most common early complication of tracheostomy was bleeding (moderate bleeding in 11 \[6.7%\] and minor bleeding in 46 \[28.2%\]), followed by subcutaneous emphysema (3 \[1.8%\]; in two this occurred together with bleeding and in one it occurred together with air leakage) and obstruction (3 \[1.8%\]). The most common late complication was bleeding (4 \[2.5%\]), followed by air leakage (3 \[1.8%\]) and tracheal stenosis (2 \[1.2%\]). The incidence of complications did not differ significantly between the successful weaning and failure-to-wean groups (early complications: 38.5% versus 37.6%, *P*= 1.0; late complications: 6.4% versus 9.4%, *P*= 0.6). No patient died during the procedure operation or because of complications of tracheostomy.
Timing of tracheostomy and outcomes
-----------------------------------
The patients were divided in two groups according to weaning outcome. Seventy-eight patients were successfully weaned from mechanical ventilation, and 85 patients failed to wean. The clinical characteristics, including sex, age, APACHE II score and previous comorbid conditions, were similar between the groups (Table [1](#T1){ref-type="table"}). The most frequent reason for intubation was pulmonary disease (107 \[65.6%\]), followed by neurological disease (28 \[17.2%\]). The indications for intubation in the two groups were also similar, except that more neurological disease was noted in the successful weaning group (Table [2](#T2){ref-type="table"}). Hypoalbuminaemia, anaemia, leucocytosis and impaired gas exchange were noted before tracheostomy. Pre-tracheostomy albumin, creatinine and haemoglobin levels were similar between groups, but the failure-to-wean group was noted to have higher WBC counts (*P*= 0.05), lower platelet counts (*P*= 0.005) and poor PaO~2~/FiO~2~ratio (*P*= 0.003; Table [3](#T3){ref-type="table"}). After tracheostomy, 109 patients (66.9%) developed nosocomial pneumonia. The average number of post-tracheostomy ventilator days was 27.3. Higher rates of post-tracheostomy pneumonia (*P*= 0.05) and longer post-tracheostomy mechanical ventilation periods (*P*= 0.001) were noted in the failure-to-wean group (Table [4](#T4){ref-type="table"}). Shorter intubation periods (*P*= 0.02), length of ICU stay (*P*= 0.001) and post-tracheostomy ICU stay (*P*= 0.005) were noted in the successful weaning group (Table [4](#T4){ref-type="table"}). The overall ICU mortality was around 19%.
ICU mortality is summarized in Fig. [1](#F1){ref-type="fig"}. Regarding the relationship of timing of tracheostomy to successful weaning, an intubation period in excess of 21 days was associated with decreased rate of successful weaning (31.5% versus 56%, *P*= 0.004) and increased ICU mortality (27.8% versus 14.7%, *P*= 0.057). The intubation period exhibited a correlation with length of ICU stay in the successful weaning group (r = 0.70, *P*\< 0.001; Fig. [2](#F2){ref-type="fig"}). We used day 21 as a cut-off point to define early and late trachostomy, in accordance with the clinical observations summarized in Fig. [1](#F1){ref-type="fig"}. Early tracheostomy was defined as tracheostomy performed within 21 days after intubation (*n*= 110); late tracheostomy was defined as tracheostomy performed later than this (*n*= 53). The early tracheostomy patient group had a higher rate of successful weaning (56.4% versus 30.2%, *P*= 0.002) and lower ICU mortality (14.5% versus 28.3%, *P*= 0.05), but there were no differences between early and late tracheostomy groups in terms of hospital mortality (44.5% versus 54.7%, *P*= 0.25) or occurrence of nosocomial pneumonia during the weaning period (43.6% versus 60.4%, *P*= 0.06). The patients who underwent early tracheostomy also had shorter post-tracheostomy ICU stays (10.8 versus 14.2 days, *P*= 0.04) and weaning periods (19.0 versus 44.3 days, *P*\< 0.001).
In univariate analysis using the Kaplan--Meier method with log-rank test, reasons for intubation (pulmonary disease \[*P*= 0.03\] and lack of neurological disease \[*P*\< 0.01\]), thrombocytopenia (*P*= 0.03), poor oxygenation before tracheostomy (*P*\< 0.001), post-tracheostomy pneumonia during the weaning period (*P*\< 0.001) and late tracheostomy (*P*\< 0.001) were correlated with lower rates of successful weaning. A Cox regression model applied to the multivariate analysis showed that late tracheostomy, poor oxygenation and post-tracheostomy pneumonia during the weaning period were independent predictors of unsuccessful weaning (Fig. [3](#F3){ref-type="fig"}).
Discussion
==========
The present study demonstrated that patients who underwent tracheostomy and failed to wean from mechanical ventilation had longer intubation periods before tracheostomy. Timing of tracheostomy was correlated with length of ICU stay in the successful weaning group.
The type of ICU may also have an impact on the timing of tracheostomy. In surgical ICUs most patients do not have chronic lung disease or severe lung injury. These patients usually undergo tracheostomy early if they underwent a major surgical procedure and failed to extubate within several days after the operation. Previous studies \[[@B15]-[@B18]\] conducted in surgical ICUs have shown that tracheostomy performed within 1 week after intubation may be beneficial in lowering rates of pneumonia, and in shortening the duration of mechanical ventilation and length of ICU stay. However, other studies reported a higher incidence of ventilator-associated pneumonia \[[@B19],[@B20]\] and longer length of ICU stay \[[@B21]\] in association with tracheostomy. In a neurological ICU, tracheostomy is usually performed if there is a depressed level of consciousness and poor ability to protect the airway. A recent study \[[@B22]\] demonstrated that early tracheostomy in patients in a medical ICU shortened the length of hospital stay and lowered hospital costs. The present study demonstrated that late tracheostomy may predispose to failure to wean and ICU mortality, especially when the intubation period is longer than 3 weeks. We also found that the duration of intubation before tracheostomy was correlated with length of ICU stay in patients who weaned successfully.
There were no obvious differences in terms of age, sex, APACHE II score, or underlying disease between the successful weaning and failure-to-wean groups, except for more neurological disease in the successful weaning group. However, in the 3 days before tracheostomy, higher WBC count, lower platelet count and lower PaO~2~/FiO~2~ratio were noted in the failure-to-wean group. These observations suggest that leucocytosis, low platelet count and severity of respiratory failure before tracheostomy might have had a greater impact on outcome than initial presentation at ICU admission.
A longer intubation period was noted in those patients who failed to wean, indicating that, like the pre-tracheostomy conditions mentioned above, late tracheostomy may predispose to poor weaning outcome. A prolonged intubation period may impair the local barrier and bronchial hygiene, increasing the risk for bacterial colonization. Also, it may result in a higher rate of post-tracheostomy pneumonia -- an association that was found in the failure-to-wean group. Ely and coworkers \[[@B23]\] demonstrated that prolonged intubation with mechanical ventilation was associated with increased hospital mortality and was independent of severity of illness. In the present study we found that prolonged intubation was associated with prolonged ICU stay. Delaying tracheostomy might not have been beneficial in these patients.
Reasons for intubation, poor pre-tracheostomy conditions, prolonged intubation and post-tracheostomy pneumonia were found to influence ventilator weaning in univariate analysis. However, in multivariate analysis we found that only late tracheostomy, pre-tracheostomy poor oxygenation and post-tracheostomy pneumonia during the weaning period were independent predictors of unsuccessful weaning. This finding suggests that timing of tracheostomy has an impact on ventilator weaning, as well as other clinical events. The 1989 ACCP Consensus Conference on Artificial Airways in Patients Receiving Mechanical Ventilation \[[@B10]\] suggested that tracheostomy is preferable if the anticipated need for mechanical ventilation is for more than 21 days. Recent ACCP guidelines \[[@B11]\] encourage early tracheostomy after patient stabilization if the patient needs prolonged mechanical ventilation. Our data support the suggestion of the earlier ACCP guidelines \[[@B10]\] that, when tracheostomy is performed more than 3 weeks after intubation, rates of ICU mortality and failure to wean increase.
The incidence of complications in adults who have undergone tracheostomy varies from 6% to 51% \[[@B4],[@B24],[@B25]\]. In the present study, the early complication rate was 38% and the late complication rate was 8% during hospitalization. The major early complication was minor to moderate bleeding from surgical wounds, which did not cause obvious clinical deterioration. We found tracheostomy to be a relatively safe procedure for airway management in patients who needed prolonged mechanical ventilation.
There are some limitations to the study. This retrospective study lacks baseline pulmonary function data before tracheostomy, which might have influenced the duration of weaning. Poor patient condition on admission to the medical ICU might have influenced the decision to perform a tracheostomy late.
Conclusion
==========
In this study we found that performance of tracheostomy more than 21 days after intubation was associated with prolonged weaning periods and low rates of successful weaning. It might also result in prolonged ICU stay. If one waits longer than 21 days, then it may be better to forego tracheostomy altogether.
Key messages
============
• We found that performance of tracheostomy more than 21 days after intubation was associated with prolonged weaning periods and low rates of weaning.
• Late tracheostomy might also result in prolonged ICU stay; if one waits longer than 21 days, then it may be better to forego tracheostomy altogether.
Abbreviations
=============
ACCP = American College of Chest Physicians; APACHE = Acute Physiology and Chronic Health Evaluation; FiO~2~= fractional inspired oxygen; ICU = intensive care unit; PaO~2~= arterial oxygen tension; WBC = white blood cell.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
CLH participated in the study design and drafted the manuscript. KYC conceived the study, participated in its design and helped to draft the manuscript. JSJ, CJY and PCY participated in study design.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
**The relationship of weaning rates, ICU mortality and durations of intubation.(a)**Rate of successful weaning in patients who underwent tracheostomy after different durations of intubation. The rate of successful weaning declined when patients underwent tracheostomy after 21 days of intubation. **(b)**Intensive care unit (ICU) mortality rates in patients who underwent tracheostomy after different durations of intubation. The ICU mortality rates increased when the patients underwent tracheostomy after 21 days of intubation.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Correlation of intubation period and the length of intensive care unit (ICU) stay in patients who weaned successfully.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
**Survival curves of independent predictors of weaning failure.(a)**Difference in rates of successful weaning between patients who underwent tracheostomy within 21 days (dotted line) and those who underwent tracheostomy later than 21 days (solid line; *P*\< 0.001). **(b)**Difference in rates of successful weaning between patients with an arterial oxygen tension (PaO~2~)/fractional inspired oxygen (FiO~2~) ratio \> 250 (dotted line) and those with a PaO~2~/FiO~2~ratio \< 250 (solid line; *P*\< 0.001) before tracheostomy. **(c)**Difference in rates of successful weaning between the patients with post-tracheostomy pneumonia (solid line) and those without post-tracheostomy pneumonia (dotted line; *P*\< 0.001)
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Demographic and clinical characteristics
:::
Characteristics Total (*n*= 163) Successful weaning (*n*= 78) Failure to wean (*n*= 85) *P*
---------------------------- ------------------ ------------------------------ --------------------------- ------
Age 70.3 ± 15.1 68.6 ± 15.4 71.8 ± 14.8 0.2
Sex (male/female) 93/70 49/29 44/41 0.2
APACHE II score 20.0 ± 7.2 19.6 ± 7.1 21.3 ± 7.6 0.1
Comorbid conditions
Hypertension 63 (38.7%) 34 (43.6%) 29 (34.1%) 0.3
Diabetes mellitus 50 (30.7%) 27 (34.6%) 23 (27.1%) 0.3
COPD 34 (20.9%) 12 (15.4%) 22 (25.9%) 0.1
Malignancy 34 (20.9%) 14 (17.9%) 20 (23.5%) 0.4
Congestive heart failure 32 (19.6%) 19 (24.4%) 13 (15.3%) 0.2
Renal insufficiency 32 (19.6%) 15 (19.2%) 17 (20.0%) 1.0
Stroke 25 (15.3%) 16 (20.5%) 9 (10.6%) 0.09
Autoimmune disease 12 (7.4%) 5 (6.4%) 7 (8.2%) 0.8
Cancer metastatic to lung 8 (4.9%) 1 (0.01%) 7 (4.5%) 0.07
Shown are demographic data for 163 critically ill patients who underwent tracheostomy, and differences between patients who weaned successfully and those who failed to wean. APACHE, Acute Physiology and Chronic Health Evaluation; COPD, chronic obstructive pulmonary disease.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Reasons for intubation
:::
Reason for intubation Total (*n*= 163) Successful weaning (*n*= 78) Failure to wean (*n*= 85) *P*
----------------------- ------------------ ------------------------------ --------------------------- -------
Pulmonary disease 107 46 (59.0%) 61 (71.8%) 0.1
Infectious disease 18 7 (9.0%) 11 (12.9%) 0.5
Neurological disease 28 20 (25.6%) 8 (9.4%) 0.007
Circulatory disease 10 5 (6.4%) 5 (5.9%) 1.0
Shown are the reasons for intubation of the 163 patients who underwent tracheostomy, and differences between patients who weaned successfully and those who failed to wean.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Pre-tracheostomy conditions
:::
Parameter Total (*n*= 163) Successful weaning (*n*= 78) Failure to wean (*n*= 85) *P*
---------------------------- ------------------ ------------------------------ --------------------------- -------
Albumin (g/dl) 2.7 ± 0.5 2.7 ± 0.4 2.6 ± 0.5 0.2
AST (U/l) 44.2 ± 49.6 47.5 ± 63.2 41.2 ± 32.5 0.4
Creatinine (mg/dl) 1.5 ± 1.7 1.4 ± 1.8 1.6 ± 1.6 0.6
Haemoglobin (g/dl) 10.5 ± 1.3 10.6 ± 1.3 10.4 ± 1.2 0.4
WBC count (cells/μl) 11993 ± 5474 11110 ± 4570 12803 ± 6104 0.05
Platelet (cell × 10^3^/μl) 192.2 ± 113.7 217.9 ± 116.3 168.7 ± 106.4 0.005
PaO~2~/FiO~2~ratio 261.5 ± 93.6 284.5 ± 85.5 240.7 ± 96.1 0.003
pH 7.4 ± 0.1 7.4 ± 0.1 7.4 ± 0.1 0.2
Paco~2~(mmHg) 40.9 ± 11.3 39.5 ± 7.5 42.2 ± 13.8 0.1
Shown are the pre-tracheostomy conditions in the 163 patients who underwent tracheostomy, and differences between patients who weaned successfully and those who failed to wean. AST, aspartate aminotransferase; CNS, central nervous system; FiO~2~, fractional inspired oxygen; PaO~2~, arterial oxygen tension; PaCO~2~, arterial carbon dioxide tension; WBC, white blood cell.
:::
::: {#T4 .table-wrap}
Table 4
::: {.caption}
######
Outcomes after tracheostomy
:::
Outcomes Total (*n*= 163) Successful weaning (*n*= 78) Failure to wean (*n*= 85) *P*
------------------------------------ ------------------ ------------------------------ --------------------------- ---------
ICU mortality (*n*\[%\]) 31 (19.0%) 2 (2.6%) 29 (34.1%) \<0.001
In-hospital mortality (*n*\[%\]) 78 (47.9%) 15 (19.2%) 63 (74.1%) \<0.001
Intubation period (days) 18.5 ± 10.9 16.3 ± 10.5 20.4 ± 10.9 0.02
Overall ICU stay (days) 29.7 ± 15.8 25.4 ± 13.8 33.7 ± 16.4 0.001
Post-tracheostomy ICU stay (days) 11.9 ± 9.9 9.7 ± 7.6 14.0 ± 11.3 0.005
Post-tracheostomy MV period (days) 27.3 ± 40.7 10.5 ± 12.6 42.6 ± 50.4 0.001
Post-tracheostomy pneumonia 109 (66.9%) 46 (59.0%) 63 (74.1%) 0.05
Shown are the outcomes after tracheostomy in the 163 patients who underwent tracheostomy, and differences between patients who weaned successfully and those who failed to wean. ICU, intensive care unit; MV, mechanical ventilation.
:::
|
PubMed Central
|
2024-06-05T03:55:54.896706
|
2004-12-23
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065112/",
"journal": "Crit Care. 2005 Dec 23; 9(1):R46-R52",
"authors": [
{
"first": "Chia-Lin",
"last": "Hsu"
},
{
"first": "Kuan-Yu",
"last": "Chen"
},
{
"first": "Chia-Hsuin",
"last": "Chang"
},
{
"first": "Jih-Shuin",
"last": "Jerng"
},
{
"first": "Chong-Jen",
"last": "Yu"
},
{
"first": "Pan-Chyr",
"last": "Yang"
}
]
}
|
PMC1065114
|
See related commentary <http://ccforum.com/content/9/1/29>
Introduction
============
Nosocomial pneumonia due to multidrug-resistant Gram-negative bacteria, such as certain *Pseudomonas aeruginosa*and *Acinetobacter baumannii*strains, is among the most serious complications that occur in the intensive care unit (ICU) setting. Mortality, morbidity and health care costs are substantially increased by this type of infection \[[@B1]-[@B3]\]. Increasing rates of resistance among Gram-negative bacteria to most classes of antimicrobial agents have frequently led to clinical failure of currently employed therapies. Lack of development and introduction into clinical practice of new antibiotics to combat multiresistant Gram-negative bacteria have stimulated renewed interest in the use of the older antibiotic colistin.
Outcomes in patients with ventilator-associated pneumonia (VAP) due to multidrug-resistant Gram-negative bacteria are poor \[[@B1]\]. Intravenous colistin was recently used to treat such infections. Notably, a recent study \[[@B4]\] compared intravenous colistin (21 patients) with imipenem (14 patients) in the treatment of VAP due to multidrug-resistant *A baumannii*. Mortality rates were similar: 61.9% among patients treated with intravenous colistin and 64.2% among patients treated with imipenem. In patients with cystic fibrosis, aerosolized colistin has successfully been used to treat acute pulmonary exacerbations of infection or initial colonization with *P aeruginosa*strains \[[@B5],[@B6]\]. However, there is a lack of data regarding the use of aerosolized colistin in patients without cystic fibrosis. A few reports have indicated that aerosolized colistin may be a beneficial additional therapeutic intervention in the management of nosocomial pneumonia (whether ventilator associated or not) \[[@B7]-[@B10]\]. In addition, a few old reports of the use of aerosolized polymyxin B yielded controversial results. Feeley and coworkers \[[@B11]\] reported that use of polymyxin B aerosol in seriously ill patients is associated with increased incidence of pneumonia due to polymyxin-resistant organisms. However, Klastersky and colleagues \[[@B12]\] found endotracheal administration of polymyxin B plus aminosidin to be a useful alternative regimen to endotracheal gentamicin for the prevention of lung infections.
We present data from our recent experience with aerosolized colistin for the treatment of pneumonia due to multidrug-resistant Gram-negative bacteria in eight ICU patients.
Methods
=======
Design of the study and patient population
------------------------------------------
Patients who received colistin (Colomycin^®^, Forest Laboratories, Kent, UK, or Colistin^®^, Norma, Athens, Greece) for treatment of infections with multidrug-resistant Gram-negative bacteria from 1 October 2000 to 31 January 2004 at \'Henry Dunant\' Hospital (a 450-bed tertiary care centre in Athens, Greece) were identified from the pharmacy electronic database. Medical records, specifically nursing records of medication administration, were retrospectively reviewed for all patients in order to identify those who received aerosolized colistin. One milligram of the colistin formulations used is approximately equal to 12,500 IU (Forest Laboratories, Kent) or 13,333 IU (Norma, Athens). Administration of aerosolized colistin for the treatment of nosocomial pneumonia due to Gram-negative bacteria, and review of patients\' charts were approved by the institutional review board of the hospital.
Data collection and entry
-------------------------
Data for several variables, including demographic and clinical information, as well as the results of laboratory and imaging tests (chest radiography or computed tomography of the thorax), were collected from the medical records of patients receiving aerosolized colistin. All available results of renal function tests (creatinine, urea, creatinine clearance, urinalysis), liver function tests (serum glutamate-pyruvate transaminase, serum glutamic-oxaloacetic transaminase, alkaline phosphatase, γ-glutamyltransferase, bilirubin), creatine phosphokinase and arterial blood gases were recorded during the course of colistin treatment and at hospital discharge.
Microbiological testing
-----------------------
All causative micro-organisms were identified using routine microbiological methods. Susceptibility testing was done using both the disk diffusion method and an automated broth microdilution method (Vitek II; bioMerieux, Hazelwood, MO, USA). (The breakpoints were those defined by the National Committee for Clinical Laboratory Standards \[[@B13],[@B14]\].) Susceptibility to colistin was tested by means of the disk diffusion method using a 10 μg colistin disk (Oxoid, Basingstoke, UK); isolates were considered sensitive if the inhibition zone was ≥ 11 mm. Intermediate sensitivity of isolated Gram-negative pathogens to antimicrobial agents was considered resistance. Multidrug-resistant was defined as resistance of the isolate to five antipseudomonal classes of antimicrobial agents (i.e. antipseudomonal penicillins, cephalosporins, carbapenems, monobactams, quinolones, colistin and aminoglycosides). An isolate was defined as colistin-only sensitive if it was resistant to all antipseudomonal agents except colistin.
Definition of pneumonia
-----------------------
Diagnosis of pneumonia required two or more serial chest radiographs with at least one of the following: new or progressive and persistent infiltrate, consolidation, cavitation, or pleural effusion. In addition, patients were required to have had fever \>38°C with no other recognized cause or an abnormal white blood cell count (leucopenia \[\<4000 white blood cells/mm^3^\] or leucocytosis \[≥ 12,000 white blood cells/mm^3^\]), and at least two of the following: new onset of purulent sputum, change in the character of sputum, increased respiratory secretions, or increased requirement for suctioning; new onset or worsening of cough, or dyspnoea or tachypnoea; rales or bronchial breath sounds; or worsening gas exchange. Pneumonia was considered to be ventilator associated (VAP) when its onset occurred 48 hours after the initiation of mechanical ventilation, and was judged not to have been incubating before the initiation of mechanical ventilation \[[@B15]\].
Definition of outcome
---------------------
The definition of positive outcome (cure or improvement) of pneumonia was based on clinical (fever defervescence, resolution or partial resolution of presenting symptoms and signs of pneumonia, decrease in suctioning requirements), radiological (decrease or disappearance of presenting findings on chest x-ray), and laboratory findings (improvement in arterial blood gases, or normalization of white blood cell count and C-reactive protein).
Results
=======
From 1 October 2000 through 31 January 2004, 152 patients received treatment with intravenous colistin for infections with multidrug-resistant Gram-negative bacteria. Eight out of 152 patients were identified as having received aerosolized colistin for the management of Gram-negative nosocomial pneumonia. Table [1](#T1){ref-type="table"} describes the demographic and clinical features of these patients, including comorbidities, responsible pathogen(s) and susceptibility of the pathogen(s) to commonly tested antimicrobial agents, as well as the outcome of the infection and of the patient.
The mean age of the patients was 59.6 years and most of them were male (six out of eight). All patients had been admitted to the ICU, with a mean Acute Physiology and Chronic Health Evaluation II scores on the day of ICU admission and on day 1 of aerosolized colistin administration of 14.6 and 17.1, respectively. During the preceding 3 months, three patients had been hospitalized in the same or another unit. All patients had received other antimicrobial regimens before aerosolized colistin was initiated. In addition, three patients received immunosuppressive treatment (steroids) and four received immunoglobulin therapy during their hospitalization.
The responsible pathogens in the eight cases of nosocomial pneumonia were *Acinetobacter baumannii*(seven out of eight) and *P aeruginosa*(one out of eight) strains. Only in one case was a second strain isolated from the same culture specimen, and it was found to be methicillin-resistant *Staphylococcus aureus*. Half of the isolated pathogens were sensitive only to colistin; the rest were multidrug-resistant strains.
All patients received mechanical ventilatory support for a mean of 19.4 days. Colistin was prepared for nebulization; 1 or 2 million IU colistin was diluted in 2 or 4 ml sterile normal saline 0.9%, respectively. In patients undergoing mechanical ventilation aerosolized colistin was delivered by means of the Siemens Servo Ventilator 300 (Siemens-Elma AB, Solna, Sweden). In spontaneously breathing patients colistin was administered as follows: 1,000,000 IU were added to 4 ml normal saline and the solution was nebulized with 8 l/min oxygen flow and inhaled via a face mask. This technique of administration of aerosolized medication is commonly used worldwide for the administration of bronchodilators in nebulized form. The daily dose of aerosolized colistin ranged from 1.5 to 6 million IU divided into three or four doses, and the duration of administration ranged from 3 to 32 days (mean 10.5 days). No strictly uniform dosing strategy for aerosolized colistin was applied, and differences in regimen reflect the differing approaches of the individual attending physicians. In addition, seven out of eight patients received concomitant intravenous treatment with colistin or other antimicrobial agents with activity against Gram-negative bacteria, such as β lactams, quinolones and aminoglycosides. Only one patient received aerosolized colistin as monotherapy; she had received intravenous colistin therapy before aerosolized colistin for 7 days and continued to receive the intravenous therapy after the end of aerosolized therapy (for 32 days).
The pneumonia was observed to respond to treatment in seven out of eight patients who received supplemental therapy with aerosolized colistin. Four episodes of pneumonia were cured and three were improved at the end of treatment. Only one out of the eight patients who received aerosolized colistin for the treatment of multidrug-resistant Gram-negative pneumonia deteriorated and finally died. He was a 50-year-old multiple trauma patient, who was admitted to the ICU with fractures located at C4--C5, haemothorax and functional dissection of the spinal cord due to a car accident. His past medical history was noteworthy for arterial hypertension, Wolff--Parkinson--White syndrome, chronic renal insufficiency due to polycystic kidney disease and ankylosing spondylitis, for which he was receiving steroid therapy. During his prolonged hospitalization in the ICU, the patient developed pneumonia due to multidrug-resistant *A baumannii*, requiring intubation. His clinical condition became complicated by sepsis syndrome due to an infection caused by a colistin-only sensitive *P aeruginosa*strain, which was unresponsive to administered antimicrobial treatment. On day 95 of his hospitalization in the ICU, he died from septic shock and multiple organ failure.
Follow-up cultures were available for five out of eight patients. In four of them the responsible pathogen was eradicated, and in one case the pathogen persisted in repeated specimen cultures; this patient died. Superinfection with Gram-positive micro-organisms or yeasts was not observed. No Gram-negative bacterium developed resistance to colistin in subsequent specimen cultures during or at the end of aerosolized treatment.
Administration of aerosolized colistin was well tolerated by all patients. During treatment, all patients were closely monitored for possible respiratory adverse reactions, but none of them experienced chest tightness, bronchoconstriction, or apnoea. Only two patients, who had history of chronic obstructive pulmonary disease, received concurrent treatment with inhaled β~2~agonist. Only in the patient who died did renal function worsen (baseline serum creatinine increased by 1.4 mg/dl) during aerosolized colistin treatment. This patient, as mentioned above, had a history of polycystic kidney disease and chronic renal failure, and died from septic shock and multiple organ failure. No deterioration in renal function was observed in the other seven patients during colistin treatment. One patient had baseline serum creatinine levels of 5.4 mg/dl, and at the end of colistin treatment serum creatinine had decreased to 4.5 mg/dl. That particular patient was already receiving haemodialysis treatment before the initiation of intravenous or aerosolized colistin.
Of 152 patients who received treatment with intravenous colistin for infections with multidrug-resistant Gram-negative bacteria during the period of study, 55 had received less than 72 hours of intravenous colistin and were excluded from all analyses. Medical records were not available for three patients; in addition, one patient was in the hospital during data collection. Thus, 93 patients were further analyzed. Forty-five of these patients received intravenous colistin for the treatment of nosocomial pneumonia due to Gram-negative bacteria. Survival and clinical cure rates for the infection were better, although not statistically significantly so, in patients with pneumonia who received additional aerosolized colistin than in patients who received only intravenous colistin treatment (survival: 7/8 patients \[87.5%\] versus 34/45 patients \[75.6%\], *P*= 0.41; clinical cure: 7/8 patients \[87.5%\] versus 30/45 patients \[66.7%\], *P*= 0.67).
Discussion
==========
Aerosolized colistin may be an effective adjunctive intervention for the treatment of nosocomial pneumonia due to multidrug-resistant Gram-negative bacteria in patients without cystic fibrosis. Colistin and polymyxin E are old antibiotics; colistin was almost abandoned for many years because of its reported nephrotoxicity and neurotoxicity. This medication was reintroduced into clinical practice just a few years ago, and this resulted mainly from increased resistance rates among Gram-negative bacteria, especially in the ICU setting, and the absence of new and effective alternative therapeutic options \[[@B16]-[@B18]\].
The idea of using colistin or polymyxin B (which belongs to the same group of antibiotics, and has similar antimicrobial spectrum, usage indications and toxicities as colistin) in the nebulized form for the management of pneumonia due to Gram-negative bacteria is not new. In 1963, Pino and coworkers \[[@B19]\] used aerosolized colistin in patients with pulmonary suppurations. A few years later, Marschke and Sarauw \[[@B20]\] reported two cases of pneumonia due to *P aeruginosa*strains in patients with underlying bronchiectasis and chronic bronchitis, in which polymyxin B was given by inhalation. Both patients experienced dyspnoea due to airway obstruction. Recently, aerosolized colistin was used successfully to treat and prevent pneumonia caused by *P aeruginosa*in patients with human immunodeficiency syndrome and in patients with nosocomial pneumonia and tracheobronchitis \[[@B21]-[@B23]\].
There is extensive experience with administration of aerosolized colistin to patients with cystic fibrosis, in whom this type of treatment is used to prevent or treat lung infections with *P aeruginosa*strains. Notably, studies found that nebulized colistin reduced the number of relapses of lung infections and subsequently the decline in lung function among patients with cystic fibrosis \[[@B24]-[@B27]\].
The pharmacokinetic properties and dosing strategies of aerosolized colistin are not well defined. Whether the various forms of colistin used for inhalation therapy (e.g. dry powder formulation for inhalation, colistin solutions for nebulization) or the different types of nebulizing systems influence the effectiveness and safety of colistin remains to be determined \[[@B28]-[@B31]\].
Adverse effects of aerosolized colistin or polymyxin B are a major concern; potential adverse effects include bronchoconstriction, chest tightness and apnoea due to neuromuscular blockade. One study conducted in 58 children with cystic fibrosis who received nebulized colistin for the treatment of lung infections \[[@B32]\] reported that 20 of them experienced a decrease in forced expiratory volume in 1 s by greater than 10% from baseline. In addition, another study \[[@B33]\] found that 35 out of 46 adult patients with cystic fibrosis who also received nebulized colistin for lung infection developed chest tightness. However, treatment with inhaled β~2~agonists before the initiation of aerosolized colistin was able to prevent the development of such side effects in the respiratory system. Another significant concern regarding the use of aerosolized colistin for the treatment of nosocomial pneumonia is dissemination of multidrug-resistant bacteria through nebulizer devices \[[@B34],[@B35]\]. However, this potential problem could be eliminated by strict use of appropriate infection control guidelines by medical and nursing hospital staff.
Our study is not without limitations. It is a small case series and is of a retrospective design. In addition, there is no control group of patients receiving treatment with only intravenous antimicrobial agents. Furthermore, some of the patients also received intravenous treatment with other antimicrobial agents, which might have influenced the outcomes.
Two major risks are arising from the wide use of colistin: the emergence of Gram-negative bacteria, such as *P aeruginosa*and *A baumannii*, resistant to colistin; and an increase of infections due to Gram-positive and Gram-negative pathogens, such as *Proteus*and *Serratia*spp., inherently resistant to colistin. Consequently, there is an urgent need to restrict the use of colistin use in order to minimize these risks.
Conclusion
==========
Inhaled colistin may be beneficial in the treatment of nosocomial pneumonia (ventilator associated or not) due to multidrug-resistant, Gram-negative bacteria. However, the severity of these infections in the ICU setting means that treatment just with aerosolized colistin is unlikely to be sufficient. This is in contrast to therapeutic strategies employed in patients with cystic fibrosis, in which initial lung colonization with *P aeruginosa*strains is commonly treated with aerosolized colistin alone. Randomized controlled trials studying the possible additional benefits and risks associated with use of nebulized colistin, as an adjunct to intravenous antimicrobial treatment, in patients with pneumonia due to multidrug-resistant Gram-negative bacteria are urgently needed.
Key messages
============
• Aerosolized administration of colistin is a promising adjunctive therapy for management of patients with pneumonia (whether ventilator associated or not) due to multiresistant Gram-negative bacteria
• Aerosolized colistin was safe in this group of patients.
• There is an urgent need for randomized controlled trials examining the efficacy and safety of aerosolized colistin for the management of patients with nosocomial pneumonia.
Abbreviations
=============
ICU = intensive care unit; VAP = ventilator-associated pneumonia.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
AM and MEF conceived the study. SKK, ZM, KR and AMK collected data. All authors contributed to the writing and preparation of the manuscript.
Figures and Tables
==================
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Demographics, clinical features, responsible pathogens, and outcomes of patients treated with aerosolized colistin
:::
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Characteristic Patient
-------------------------------------------------------------------------------------------------- --------------------------------------------------------------- -------------------------------------------------------------------------------------------- --------------------------------------------------------------------------------------------------------------------------------------------------------- -------------------------------------------------------------------------------------------------------------------------------------------------------------- -------------------------------------------------------------------------------------------------------------- --------------------------------------------------------- -------------------------------------------------------------------------------------------------------------------------------------------------------- ------------------------------------------------------------
Medical history Fatty liver, arterial hypertension Smoking, arterial hypertension, pulmonary oedema, heart attack, mild chronic renal failure Liver hamartoma, chronic obstructive pulmonary disease, urinary incontinence, hypothyroidism, Sjögren\'s syndrome, excised left frontal lobe meningioma Catarract, cholosteatoma, arterial hypertension, urinary tract infection 3 weeks before admission Wolff--Parkinson--White syndrome, chronic renal failure (polycystic kidney disease), ankylosing spondylitis Smoking, obesity, chronic obstructive pulmonary disease Arterial hypertension, chronic renal dysfunction (creatinine clearance 75--80 ml/min), adenoma of hypophysis, epileptic seizures, cerebral haemorrhage Arterial hypertension, cerebral arteriovenous malformation
Reason for admission Stomach lymphoma Acute myocardial infarction Epileptic seizures Fever, headache Multitrauma patient, C4--C5 fractures due to car accident, functional dissection of spinal cord, haemothorax Oesophageal perforation Adenoma of hypophysis, cerebral haemorrhage Pneumonia, sleep apnoea syndrome, cerebral haemorrhage
Discharge diagnosis Stomach lymphoma, nosocomial pneumonia Acute myocardial infarction, nosocomial pneumonia Postsurgical intracranial haematoma, pulmonary embolism, inferior vena cava filter placement Pneumococcal meningitis, hydrocephalus, pulmonary embolism, pneumonia, urinary tract infection Septic shock, multiple organ failure Mediastinitis Pneumonia Pneumonia, sleep apnoea syndrome, cerebral haemorrhage
APACHE II score on ICU admission 14 17 17 9 12 17 19 12
APACHE II score on first day of colistin treatment 10 29 19 8 19 20 18 14
Surgery during hospitalization Liver biopsy, partial gastrectomy Coronary artery bypass surgery Drainage of postsurgical haematoma of left frontal lobe, inferior vena cava filter placement Endoscopic ethmoidectomy, surgical drainage of the frontal and maxillary sinuses Spinal arthrodesis surgery (C5--T1) Surgical repair of oesophageal perforation Excision of pituitary adenoma Embolization of arteriovenous malformation
Duration of mechanical ventilation (days) 10 16 5 18 65 25 8 8
Time from ICU admission to develop the infection for which aerosolized colistin was given (days) 8 1 7 22 24 1 7 5
Site of infection Pneumonia (VAP) Pneumonia, urinary tract infection Bacteraemia, pneumonia (VAP) Pneumonia (VAP) Pneumonia (VAP) Pneumonia Pneumonia (VAP) Pneumonia (VAP)
Isolated micro-organism (source) *Acinetobacter baumannii*(BAL) *A baumannii*(bronchial secretions) *A baumannii*(blood), *A baumannii*(bronchial secretions) *Pseudomonas aeruginosa*(bronchial secretions) *A baumannii*(bronchial secretions) *A baumannii*(BAL) *A baumannii*(bronchial secretions) *A baumannii*(bronchial secretions)
Susceptibility of the isolated pathogen MDR (sensitive to colistin and gentamycin) COS COS COS MDR (sensitive to colistin and gentamycin) COS MDR (sensitive to colistin and gentamycin) MDR (sensitive to colistin and gentamycin)
Duration/dosage of nebulized colistin 6 days/1 million IU q8 h 13 days/1 million IU q8 h 10 days/0.5 million IU q8 h 5 days/1.5 million IU q8 h 7 days/2 million IU q8 h 3 days/1 million IU q8 h 8 days/0.5 million IU q6 h 19 days/1 million IU q8 h
Duration/dosage of concomitant intravenous antibiotic treatment Colistin: 2 days/3 million IU q8 h, 6 days/2 million IU q8 h\ Colistin: 14 days/1 million IU q8 h\ Colistin: 26 days/3 million IU q8 h\ She received intravenous colistin before nebulized treatment (7 days/1 million IU q8 h) and after the end of nebulized treatment (32 days/1 million IU q8 h) Tobramycin: 7 days/80 mg q24 h\ Colistin: 14 days/2 million IU q8 h\ Colistin: 8 days/2 million IU q8 h\ Meropenem: 27 days/2 g q8 h\
Levofloxacin: 3 days/500 mg q24 h\ Meropenem: 12 days/1 g q12 h Meropenem: 26 days/2 g q8 h Aztreonam: 3 days/1 g q8 h Meropenem: 15 days/2 g q8 h\ Meropenem: 4 days/2 g q8 h Gentamicin: 27 days/80 mg q8 h
Co-trimoxazole: 4 days/3 ampules q8 h\ Gentamicin: 8 days/80 mg q8 h
Ciprofloxacin 4 days/400 mg q12 h
Duration of hospitalization (days) 17 16 41 234 94 25 36 40
Duration of ICU stay (days) 11 16 21 62 95 25 13 20
Outcome of infection Cure Improvement Cure Improvement Deterioration Improvement Cure Cure
Outcome of patient Discharge Discharge Discharge Discharge Death Discharge Discharge Discharge
Serum creatinine value (mg/dl) on the first day of aerosolized colistin administration 1.1 5.2 1 0.4 2.4 0.6 0.8 0.8
Serum creatinine value (mg/dl) at the end of aerosolized colistin administration 0.8 4.5 0.9 0.5 3.8 0.5 0.7 0,6
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
APACHE, Acute Physiology and Chronic Health Evaluation; BAL, bronchoalveolar lavage; COS, colistin-only-sensitive; ICU, intensive care unit; MDR, multidrug-resistant; VAP, ventilator-associated pneumonia.
:::
|
PubMed Central
|
2024-06-05T03:55:54.899347
|
2005-1-6
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065114/",
"journal": "Crit Care. 2005 Jan 6; 9(1):R53-R59",
"authors": [
{
"first": "Argyris",
"last": "Michalopoulos"
},
{
"first": "Sofia K",
"last": "Kasiakou"
},
{
"first": "Zefi",
"last": "Mastora"
},
{
"first": "Kostas",
"last": "Rellos"
},
{
"first": "Anastasios M",
"last": "Kapaskelis"
},
{
"first": "Matthew E",
"last": "Falagas"
}
]
}
|
PMC1065265
|
Introduction {#s1}
============
Many pathogenic and nonpathogenic bacteria produce type IV pili (Tfp), among them, Neisseria gonorrhoeae, *N. meningitidis, Pseudomonas aeruginosa, Legionella pneumophila,* enteropathogenic and enterohemorrhagic *Escherichia coli,* and Vibrio cholerae \[[@pbio-0030100-b01]\]*.* Tfp are fimbriate organelles that play a crucial role in the interaction of the bacterium with its environment, as evidenced by their requirement for motility \[[@pbio-0030100-b02]\], biofilm formation \[[@pbio-0030100-b03],[@pbio-0030100-b04]\], and horizontal gene transfer \[[@pbio-0030100-b05],[@pbio-0030100-b06],[@pbio-0030100-b07]\]. These appendages also promote bacterial attachment to host cells and contribute to virulence \[[@pbio-0030100-b08],[@pbio-0030100-b09],[@pbio-0030100-b10],[@pbio-0030100-b11],[@pbio-0030100-b12]\].
Recent evidence has shown that the gonococcal Tfp can physically retract---a process that underlies twitching motility \[[@pbio-0030100-b13]\] (i.e., the ability of the bacterium to move on solid surfaces \[[@pbio-0030100-b14]\]). It is now generally believed that twitching motility occurs via extension, substrate tethering, and retraction of the pilus filament. Two inner membrane/cytoplasmic ATPases, PilF and PilT, take part in these activities. PilF mediates pilus assembly, as *pilF* mutants produce pilin subunits but are not piliated \[[@pbio-0030100-b15]\]. PilT is involved in pilus disassembly, as *pilT* mutants are piliated but cannot retract their pili \[[@pbio-0030100-b13],[@pbio-0030100-b16]\]. Neither mutant is motile.
Pilus retraction allows gonococci to form organized microbial communities on the cell surface and on synthetic substrates (S. Lee and M. S., unpublished data), via both specific and nonspecific interactions. During attachment to host cells, microcolonies stimulate the formation of cortical plaques---structures in the cell cortex containing high concentrations of transmembrane receptors, nonreceptor tyrosine kinases and their anchors, and components of the cortical cytoskeleton \[[@pbio-0030100-b10],[@pbio-0030100-b17]\]. Though *pilT* mutants adhere normally to both synthetic surfaces and epithelial cells, they form disordered microcolonies, fail to induce cortical plaques, and are less invasive than their wild-type (wt) parent strain \[[@pbio-0030100-b17]\].
Retraction of a single gonococcal pilus can exert forces up to 80--100 pN on its substrate \[[@pbio-0030100-b13],[@pbio-0030100-b18]\]. Forces of lesser magnitude can elongate the membrane into microvillus-like structures \[[@pbio-0030100-b19],[@pbio-0030100-b20]\], promote cytoskeleton rearrangements and protein clustering \[[@pbio-0030100-b21],[@pbio-0030100-b22]\], induce calcium fluxes \[[@pbio-0030100-b23],[@pbio-0030100-b24]\], and alter gene expression \[[@pbio-0030100-b25],[@pbio-0030100-b26],[@pbio-0030100-b27],[@pbio-0030100-b28]\]. Pilus retraction has therefore been speculated to induce host cell signaling by exerting mechanical tension on the membrane \[[@pbio-0030100-b17]\]. Indirect support for a mechanical signaling hypothesis comes from observations that *pilT* mutants, unlike wt piliated strains, can neither trigger cortical plaque formation \[[@pbio-0030100-b10]\] nor activate PI-3 kinase (S. Lee and M. S., unpublished data), a member of a mechanical stress--activated pathway. Moreover, a *pilT* mutant induces an attenuated calcium flux in epithelial cells, as compared to infection with wt gonococci (P. Ayala and M. S., unpublished data). Here we provide further evidence that pilus retraction acts as a mechanical stimulus by activating mechanical stress--signaling pathways that alter epithelial cell gene expression and generate a cytoprotective environment within the host cell.
Results {#s2}
=======
Pilus Retraction Enhances the Expression of Cell Stress/Survival Genes {#s2a}
----------------------------------------------------------------------
We used microarrays to examine the transcriptional profiles of T84 human colorectal epithelial cells infected with retraction-proficient (N400) or retraction-deficient (N400*pilT*) gonococci for 3 h. Infection with N400 or N400*pilT* induced transcriptional changes in the same genes. Contrary to expectations, no genes responded uniquely to infection with either strain. Instead, infection with *pilT* affected the level of expression of a small subset of infection-responsive genes. To segregate the genes responding to pilus retraction, a wt to *pilT* fold-change expression ratio (W/P) was calculated for each infection-regulated gene. This method identified, out of approximately 300 infection-regulated genes, 69 probe sets (representing 52 genes) whose expression appeared to be enhanced by pilus retraction ([Figure 1](#pbio-0030100-g001){ref-type="fig"}).
::: {#pbio-0030100-g001 .fig}
Figure 1
::: {.caption}
###### Infection-Regulated, Retraction-Enhanced Epithelial Cell Genes
Wt and *pilT* values represent the mean fold-change in the transcript level of each gene in infected cells compared to uninfected cells (*n* = 2). W/P values represent the degree of enhancement of gene expression resulting from pilus retraction and are the result of dividing the wt fold-change value by the *pilT* fold-change value from two independent experiments. The *p*-value for each gene represents the statistical significance of the difference in its expression level (as determined by Cyber-T analysis) between wt and *pilT*. The color code assigned to each gene represents its degree of response to infection as expressed by its fold-change value, W/P, and *p*-value.
:::
:::
To confirm the microarray results, real-time quantitative RT-PCR was initially performed on two infection-regulated genes, *DUSP5* and *ADM*. According to our microarray data, *DUSP5* expression was enhanced by pilus retraction (W/P = 1.63), and *ADM* expression was not (W/P ≈ 1.0). RT-PCR results corroborated the microarray analysis, as *DUSP5* transcript levels were significantly higher in N400-infected cells than N400*pilT*-infected cells, whereas *ADM* transcript levels were similar in both sets of cells ([Figure 2](#pbio-0030100-g002){ref-type="fig"}A). Ten additional genes predicted to respond to retraction and five additional genes predicted to be not affected by retraction were similarly analyzed by real-time quantitative RT-PCR ([Figure 2](#pbio-0030100-g002){ref-type="fig"}B). In every case, the presumptive positives yielded W/P ratios of 1.5 or more, whereas the presumptive negatives yielded W/P ratios of approximately 1.0.
::: {#pbio-0030100-g002 .fig}
Figure 2
::: {.caption}
###### Real-Time Quantitative RT-PCR Verification of Microarray Results and Initial Characterization of Retraction-Enhanced Genes
\(A) Microarray (top panels) and real-time quantitative RT-PCR (bottom panels) expression profiles of ADM and DUSP5 in uninfected cells (UI), N400-infected cells (WT), and N400*pilT*-infected cells*(pilT)*. Microarray data are shown as box-plots (*n* = 3). RT-PCR data are plotted as triplicate samples from one representative experiment.
\(B) Real-time quantitative RT-PCR verification of retraction-enhanced expression of selected genes. Data are expressed as average W/P (±SEM, *n* = 3). Genes with a W/P statistically greater than 1.0 (*p* \< 0.05) are denoted with an asterisk.
\(C) Grouping of retraction-enhanced genes according to function, based on published reports (see [Table S1](#st001){ref-type="supplementary-material"}). Some genes have multiple functions and thus appear in more than one group.
\(D) Genes in this study that are known to be induced by environmental stress, mechanical stress, or MAPK signaling (see [Table S1](#st001){ref-type="supplementary-material"}).
:::
:::
The identification of genes whose expression is enhanced by pilus retraction raised the question of whether these genes share a common regulatory pathway or perform similar functions. The majority of genes whose expression is enhanced by retraction are involved in the cell stress response and survival ([Figure 2](#pbio-0030100-g002){ref-type="fig"}C). Over half of these can be induced by environmental or other cellular stresses, and a striking number can be induced specifically by mechanical stress ([Figure 2](#pbio-0030100-g002){ref-type="fig"}D). Importantly, the majority of the genes from both groups can also be induced by mitogen-activated protein kinases (MAPKs; [Figure 2](#pbio-0030100-g002){ref-type="fig"}D) (For literature citations, see [Table S1](#st001){ref-type="supplementary-material"}.) These results indicate that pilus retraction may enhance infection-induced gene expression through the MAPK pathway.
ERK, JNK, and P38 MAPK Are Activated by Infection and Enhanced by Pilus Retraction {#s2b}
----------------------------------------------------------------------------------
The MAPK cascades are well known for their involvement in the stress response, including the response to bacterial infection. Previous studies have shown that JNK is activated in N. gonorrhoeae--infected HeLa, Chang, and phagocytic cells \[[@pbio-0030100-b29],[@pbio-0030100-b30]\], and MAPK signaling is induced in conjunctival cells by N. meningitidis (R. Bonnah and M. S., unpublished data). To study the role of MAPK signaling in retraction-enhanced gene expression, we first determined which of these pathways are activated in infected T84 cells.
Compared to resting cells ([Figure 3](#pbio-0030100-g003){ref-type="fig"}A, left panel), the addition of medium alone slightly increased the levels of ERK-p, JNK-p, and P38-p ([Figure 3](#pbio-0030100-g003){ref-type="fig"}A, UI), but levels of each phosphorylated kinase returned to baseline after 90 min. Infection with N400 dramatically increased the levels of all three activated kinases by 60 min post-infection ([Figure 3](#pbio-0030100-g003){ref-type="fig"}A, WT). Densitometric analysis of immunoblots from two independent experiments is shown in [Figure 3](#pbio-0030100-g003){ref-type="fig"}B. ERK-p levels were elevated throughout the course of infection, with only a slight decrease in phosphorylation visible by 3 h post-infection. In contrast, P38-p and JNK-p levels peaked between 60 and 90 min post-infection and dropped noticeably by 3 h post-infection.
::: {#pbio-0030100-g003 .fig}
Figure 3
::: {.caption}
###### Levels of Activated MAPK in Infected Cells and Their Involvement in Retraction-Enhanced Gene Expression
\(A) Representative immunoblot showing ERK-p, P38-p, and JNK-p, in uninfected cells before (0 h) and after medium change (UI), or in cells infected with N400 (WT) or N400*pilT (pilT).* Total P38 protein levels in each sample served as the internal control (bottom lanes).
(B and C) ERK-p JNK-p, and P38-p levels over time in cells infected with N400 and N400*pilT,* respectively. Immunoblots from (A) were analyzed by densitometry, and levels of activated kinase from infected cells were normalized to that from uninfected cells (UI). Values represent mean normalized protein levels (±SEM, *n* = 2). Solid markers indicate a significant difference between wt and *pilT*-induced MAPK phosphorylation at that time point (*p* \< 0.05); thus, ERK-p is significant at 60, 90, and 180 min; JNK-p is significant at 60 min; and P38-p is significant at 60 and 90 min.
\(D) Representative immunoblot showing ERK-p, MAPKAPK2-p, and c-Jun-p in cells preincubated with vehicle (DMSO) or MAPK inhibitors and infected for 90 min with N400 (WT) or left untreated (UI). Total P38 protein levels in each sample served as the internal control (bottom lanes).
\(E) Real-time quantitative RT-PCR analysis of the effect of MAPK inhibitors on the expression of retraction-responsive genes. Light bars indicate cells infected with N400 in the presence of vehicle (DMSO); dark bars indicate cells infected with N400 in the presence of MAPK inhibitors. Values represent the fold-change (±SEM, *n* = 2) in transcript levels compared to uninfected, DMSO treated control. A significant difference in expression between the two conditions is denoted by an asterisk (*p* \< 0.1).
:::
:::
We next examined MAPK phosphorylation in T84 cells infected with N400*pilT* to determine whether kinase activation was influenced by pilus retraction. Low levels of all three activated MAPKs were detected in N400*pilT*-infected cells only after 90 min of infection ([Figure 3](#pbio-0030100-g003){ref-type="fig"}A, PT). Densitometric analysis of immunoblots from two independent experiments is shown in [Figure 3](#pbio-0030100-g003){ref-type="fig"}C. Although the kinetics of MAPK activation appear to be different in wt- and *pilT*-infected cells, a firm conclusion cannot be drawn from these results, given the delayed onset of activation and the low levels of phosphorylation of each enzyme. Taken together, these results demonstrate that infection by piliated gonococci activates all three MAPK pathways and that pilus retraction enhances this activation.
MAPK Signaling Is a Mediator of Retraction-Dependent Enhancement of Gene Expression {#s2c}
-----------------------------------------------------------------------------------
We next determined whether MAPK signaling regulates the expression of retraction-enhanced genes. T84 cells were preincubated with vehicle or MAPK inhibitors SB203588, U0126, and SP600125, and assessed for ERK, P38, and JNK activation by immunoblotting for ERK-p, MAPKAPK2-p, and c-Jun-p, respectively. MAPK inhibitors dramatically reduced the levels of all three activated kinases in both uninfected and N400-infected cells ([Figure 3](#pbio-0030100-g003){ref-type="fig"}D). They also significantly reduced the transcript levels of four of the five retraction-responsive genes in N400-infected cells, as judged by real-time quantitative RT-PCR ([Figure 3](#pbio-0030100-g003){ref-type="fig"}E). In contrast, the inhibitors did not affect the transcript levels of genes with a W/P of approximately 1.0. Interestingly, *cyr61* expression was unaltered by MAPK inhibitors. This gene was shown by microarray (W/P = 2.15) and RT-PCR analysis (W/P = 1.86) to respond to retraction. These results implicate MAPK signaling in the regulation of some, but not all, retraction-responsive genes. They indicate that other pathways also influence the response of genes to pilus retraction.
Mechanical Stress Activates MAPK Signaling and Upregulates Retraction-Responsive Genes {#s2d}
--------------------------------------------------------------------------------------
A significant number of the retraction-responsive genes are known to be induced specifically by mechanical strain on the cell membrane. Although substantial force is generated by pilus retraction in vitro, this force has not yet been demonstrated to influence host responses to infection. To examine this issue, we determined whether artificial mechanical force on the epithelial cell membrane could mimic retraction-induced MAPK activation and retraction-enhanced gene expression. To generate mechanical stress in a manner similar to that of pilus retraction, a modified magnet-based force assay was used \[[@pbio-0030100-b31]\]. Magnetic beads were coated with crude pili preparations (CPPs) from piliated gonococci and added to T84 cells ([Figure 4](#pbio-0030100-g004){ref-type="fig"}A). Within 30 min, small clusters of approximately two to ten beads attached to the cells, with each cell containing two to three clusters of beads (data not shown). Cell monolayers were then placed 10 mm beneath the magnet. At this distance, the magnet generates an upward force of 4 pN per bead ([Figure 4](#pbio-0030100-g004){ref-type="fig"}B), or approximately 20--100 pN per cell.
::: {#pbio-0030100-g004 .fig}
Figure 4
::: {.caption}
###### Artificial Force Triggers MAPK Phosphorylation and Induces the Expression of Retraction-Enhanced Genes
\(A) Representation of the magnet/magnetic bead assay.
\(B) Average force generated on one bead as a function of magnet distance from the culture dish. Data represent the forces calculated from four identical magnets (±SEM). All subsequent assays were performed using a magnet distance of 10 mm, which corresponds to a force of 4 pN per bead (dotted line).
\(C) Magnet-induced clustering of actin beneath magnetic beads. CPP-coated beads were seeded onto T84 cells and exposed to the magnet for 1 h (top panels) or left untreated (no magnet, bottom panels). Differential interference contrast images (left panels) reveal the location of the beads; phalloidin staining (middle panels) shows the presence of actin at the same site. Right panels show the two previous images merged.
\(D) Representative immunoblot of ERK-p, JNK-p, and P38-p in cells seeded with CPP-coated beads, BSA-coated beads, or no beads, and exposed to the magnet for 15 or 30 min. Total P38 protein levels in each sample served as the internal control (bottom panels).
\(E) Quantitation of ERK-p, JNK-p, and P38-p signals by densitometry from the representative immunoblot shown in (D), normalized to the no-bead control. Solid lines indicate signals from cells exposed to membrane-coated beads; dotted lines indicate signals from cells exposed to BSA-coated beads.
\(F) Real-time quantitative RT-PCR analysis of the transcript levels of selected genes in cells seeded with CPP beads and exposed to the magnet for 3 h. Data represent the average fold-change (±SEM, *n* = 2) compared to a no-magnet control. A significant difference in expression on force induction is denoted by an asterisk (*p* \< 0.1).
:::
:::
T84 cells seeded with CPP-coated beads and exposed to the magnet were first examined for the presence of actin recruitment into cortical plaques (see Introduction). The clustering of actin near these beads would indicate that the magnetic force was sufficient to mimic pilus retraction forces from the bacterial microcolony. In the presence of magnetic force, actin concentrated in the cell cortex around membrane-coated beads ([Figure 4](#pbio-0030100-g004){ref-type="fig"}C, top panel). In contrast, actin did not cluster with the beads in the absence of the magnet ([Figure 4](#pbio-0030100-g004){ref-type="fig"}C, bottom panel). Thus, the force generated by this magnet system was sufficient to recruit actin to the site of the attached beads.
We next determined whether magnetic forces applied to CPP-coated beads were sufficient to activate MAPK and alter gene expression. The levels of all three phosphokinases were slightly reduced when the magnetic field was applied to cells incubated with medium alone ([Figure 4](#pbio-0030100-g004){ref-type="fig"}D, no beads). Levels of each phosphorylated kinase from bead-treated samples ([Figure 4](#pbio-0030100-g004){ref-type="fig"}D, CPP) were normalized to those from the no-bead samples to account for the effect of the magnet alone on MAPK phosphorylation. Following normalization, increased levels of all three phosphokinases are evident within the short time course ([Figure 4](#pbio-0030100-g004){ref-type="fig"}E, solid lines).
In parallel experiments, cells were seeded with bovine serum albumin (BSA)-coated beads and exposed to the magnet. Under these conditions, less force was applied to the cells, as fewer bead clusters attached to the cells, and each cluster contained only two to three beads on average (data not shown). Again, levels of each phosphorylated kinase from bead-treated samples ([Figure 4](#pbio-0030100-g004){ref-type="fig"}D, BSA beads) were normalized to those from the no-bead samples to account for the effect of the magnet alone on MAPK phosphorylation. Despite lower forces, BSA-coated beads also activated ERK, JNK, and P38 ([Figure 4](#pbio-0030100-g004){ref-type="fig"}E, dashed lines). Interestingly, force-induced activation of both JNK and ERK was higher in cells treated with BSA-coated beads. This can most likely be attributed to the fact that BSA-coated beads, unlike CPP beads, induce no MAPK activation in the absence of force (data not shown). Thus when force-induced MAPK activation is calculated, the CPP-coated beads are normalized to a higher level of "background" activation than are the BSA-coated beads. The observation that force induction via both CPP- and BSA-coated beads can induce these signals strongly indicates that activation of MAPK cascades is, in part, a response to stress forces on the membrane rather than to force mediated through specific adhesin--receptor contacts between the bacterium and the host.
To examine the effect of mechanical stress on gene expression changes, cells seeded with CPP-coated beads were exposed to magnetic force for 3 h, and gene expression levels were analyzed by real-time quantitative RT-PCR. Transcript levels were expressed as the ratio of signals from magnet-stimulated cells to those from cells not subjected to magnetic force. All three "enhanced" genes tested, *EGR1, DTR,* and *DUSP5,* were upregulated in cells exposed to magnetic force ([Figure 4](#pbio-0030100-g004){ref-type="fig"}F). In contrast, neither *ADM* (W/P ≈ 1.0) nor *cyr61* (which did not respond to MAPK inhibitors; see [Figure 3](#pbio-0030100-g003){ref-type="fig"}E) was affected by the magnet.
In this and the previous experiment, magnet-induced changes were of lower magnitude than those induced by infection. The most plausible explanation for this difference is that pilus retraction from a microcolony likely generates greater force than a magnet acting on a small cluster of beads. In our magnet assay, an average force of 20--100 pN was placed on each cell. During an infection, each *pilus* can induce this amount of force. Thus, if there are 10--100 bacteria per microcolony, and each bacterium expressed 10 pili (a conservative estimate), pilus retraction from a single microcolony could place forces of 10^4^--10^5^ pN on the cell. Nonetheless, our method of artificial force application did indeed activate all three MAPK cascades and increased the expression level of each gene examined by approximately 1.5-fold. (Note that a minimum 1.5-fold change in expression level was found to accurately identify retraction-responsive genes in the microarray experiment.) Together, these results demonstrate that retraction-enhanced MAPK activation and gene expression changes can be replicated by artificial force.
Pilus Retraction Mediates Host Cell Cytoprotection {#s2e}
--------------------------------------------------
Many of the retraction-responsive genes are known to protect cells from apoptosis and from a variety of cellular stresses. Moreover, prolonged ERK activation accompanied by transient JNK and P38 activation (as observed in a wt infection; see [Figure 3](#pbio-0030100-g003){ref-type="fig"}A and [3](#pbio-0030100-g003){ref-type="fig"}B) is hypothesized to mediate cytoprotection \[[@pbio-0030100-b32],[@pbio-0030100-b33],[@pbio-0030100-b34],[@pbio-0030100-b35]\]. We therefore investigated whether pilus retraction was involved in determining cell fate by assaying infected cells for cleaved poly(ADP-ribose) polymerase (PARP) and cleaved caspase 8. PARP is a 116-kDa nuclear protein that mediates DNA repair in response to cell stress and is required to maintain cell viability \[[@pbio-0030100-b36],[@pbio-0030100-b37]\]. During programmed cell death, the protein is cleaved by caspase 3 or caspase 7, a terminal step in the caspase cascade \[[@pbio-0030100-b38],[@pbio-0030100-b39]\]. Caspase 8, however, is an initiator caspase that is upstream of caspase 3, caspase 7, and PARP, and represents an earlier event in the apoptosis cascade. Thus, increased levels of cleaved PARP or caspase 8 indicate that a cell is undergoing apoptosis.
Cells infected with N400 for 6 h contained lower levels of both cleaved PARP and cleaved caspase 8 than did uninfected cells ([Figure 5](#pbio-0030100-g005){ref-type="fig"}A). In contrast, N400*pilT*-infected cells had higher levels of cleaved PARP and cleaved caspase 8 than did both uninfected and wt-infected cells. These results indicate that piliated gonococci that cannot retract pili induce low levels of programmed cell death in a culture. In contrast, gonococci capable of retracting their pili lower the tendency for cells to enter the apoptosis pathway.
::: {#pbio-0030100-g005 .fig}
Figure 5
::: {.caption}
###### Pilus Retraction during Bacterial Attachment Promotes Host Cell Cytoprotection
\(A) Levels of cleaved PARP and cleaved caspase 8 in T84 cells infected for 6 h with N400 (WT) or N400*pilT (pilT),* normalized to cleaved PARP or cleaved caspase 8 levels in uninfected cells.
\(B) Levels of cleaved PARP and cleaved caspase 8 in T84 cells infected with N400 (WT) or N400*pilT (pilT)* or left uninfected (UI) for 4 h, then incubated with STS (1 μM) for an additional 4 h to induce apoptosis. A significant difference from uninfected cells is denoted by two asterisks (*p* \< 0.05).
\(C) Cleaved PARP and cleaved caspase 8 levels in cells exposed to magnetic force. T84 cells were seeded with CPP- or BSA-coated beads and exposed to the magnet for 2 h or were left unexposed, then incubated with STS (1 μM) for an additional 4 h away from the magnet. The cleaved PARP and cleaved caspase 8 level in cells without beads and not exposed to magnetic force is arbitrarily assigned a value of 1.0, and all other treatments are expressed relative to this value. For all experiments, cleaved protein levels were quantified by densitometry of immunoblot signals. Values represent the mean levels of cleaved target (±SEM) from two independent experiments. A significant difference from untreated cells is denoted by two asterisks (*p* \< 0.05) or by a single asterisk (*p* \< 0.1).
:::
:::
We next determined whether this cytoprotective effect of pilus retraction was sufficient to protect cells from staurosporine (STS)-induced apoptosis. STS is a cell-permeant protein kinase inhibitor that induces apoptosis at micromolar concentrations \[[@pbio-0030100-b40],[@pbio-0030100-b41]\]. Infection of urethral epithelium with N. gonorrhoeae was recently reported to protect these cells from STS-induced apoptosis \[[@pbio-0030100-b42]\]. Both N400 and N400*pilT* infection protected T84 cells from STS-induced apoptosis, as compared to uninfected cells ([Figure 5](#pbio-0030100-g005){ref-type="fig"}B). However, cleaved PARP and cleaved caspase 8 levels in N400*pilT*-infected cells were higher than in wt-infected cells, indicating that pilus retraction enhances protection from STS-induced apoptosis.
Finally, we examined whether this retraction-enhanced cytoprotection is specifically mediated by mechanical force. In the absence of force, CPP-coated beads provided moderate protection from STS-induced apoptosis, demonstrated by lower cleaved PARP and cleaved caspase 8 levels than the no-bead cell control ([Figure 5](#pbio-0030100-g005){ref-type="fig"}C). This result is similar to that seen in *pilT*-infected cells and suggests that components in the bacterial membrane are sufficient to protect against STS-induced apoptosis. Cells seeded with CPP-coated beads and exposed to the magnetic field had still lower cleaved PARP and cleaved caspase 8 levels, consistent with data from wt-infected cells ([Figure 5](#pbio-0030100-g005){ref-type="fig"}B). BSA-coated beads did not protect against STS-induced apoptosis in the absence of magnetic force. However, when force was applied to these cells, the level of cleaved PARP and cleaved caspase 8 was reduced nearly to the value observed for membrane-coated beads in the presence of the magnet. Together, these data indicate that nonspecific membrane tension is capable of protecting the host cell against apoptosis.
Discussion {#s3}
==========
Retraction of the N. gonorrhoeae Tfp during bacterial attachment elicits host cell signaling cascades essential for the establishment of intimate attachment and promotion of bacterial invasion \[[@pbio-0030100-b17]\]. We tested the hypothesis that Tfp retraction induces changes in epithelial cell gene expression during bacterial attachment. Pilus retraction, per se, did not regulate a unique set of genes. Rather, retraction enhanced the expression of a small subset of infection-regulated genes (see [Figure 1](#pbio-0030100-g001){ref-type="fig"}), many of which are known to respond specifically to mechanical stress and to be induced by the MAPK cascade. We confirmed that wt bacteria activated MAPKs ERK, JNK, and P38 at a higher level than the *pilT* mutant. Moreover, MAPK inhibitors lowered the expression level of all but one retraction-responsive gene selected for further examination (see [Figure 3](#pbio-0030100-g003){ref-type="fig"}). These results strongly indicate that MAPK signaling plays a major role in the enhancement of gene expression by pilus retraction.
Importantly, artificial force placed on the cell membrane using magnets and magnetic beads can replicate the gene expression changes and MAPK activation observed using wt bacteria, indicating that pilus retraction may induce these events via mechanical force. Although the total force produced by pilus retraction within a bacterial microcolony is not known, we estimate that it is on the order of 10^4^--10^5^ pN, based on 100 pN per retraction event, approximately 10 pili per bacteria, and roughly 10--100 bacteria per microcolony. In comparison, this amount of force is equivalent to that applied to integrin complexes in the periodontal ligament by a human bite \[[@pbio-0030100-b31]\]. Retraction forces from a microcolony could therefore be physiologically relevant.
We cannot exclude the possibility that pilus retraction enhances these signaling events by mechanisms independent of membrane tension (i.e., through secondary receptor engagement or via an inherent difference in the pilus structure/composition between wt and *pilT* bacteria). Our data with CPP-coated beads strongly argue against these possibilities, however. The CPP preps used for bead coating were from wt cultures, and thus were identical. In addition, the magnet pulled the beads upward. This should pull the bead farther from the cell surface, making secondary receptor engagement less likely. The possibility remains, however, that pilus differences or secondary receptor engagement may act in concert with membrane tension to generate the higher levels of MAPK activation and gene expression changes seen with infection. Further research is needed to examine this possibility. Nonetheless, we are confident that force plays at least some role in the signaling events identified through this work.
We have begun to assess the biological functions of enhanced gene expression and MAPK activation during gonococcal infection. ERK, JNK, and P38 play a role in determining cell survival during stress and entry into the apoptosis pathway \[[@pbio-0030100-b32]\]. Moreover, nearly half of the identified retraction-enhanced genes are known to be involved in cell cycle/survival signaling. We show that cells infected with wt bacteria have lower levels of cleaved PARP and cleaved caspase 8 than do uninfected and *pilT*-infected cells. Pilus retraction is therefore predicted to enhance the ability of the cell to withstand apoptosis-inducing signals generated by infection. Indeed, cells infected with wt bacteria withstood STS-induced apoptosis better than uninfected cells and cells infected with *pilT*.
The effect of N. gonorrhoeae infection on cell fate has been a long-standing controversy. The neisserial porin has been reported to protect cells from apoptosis \[[@pbio-0030100-b43]\] as well as to induce programmed cell death \[[@pbio-0030100-b44]\]. These conflicting observations are likely a result of differences in experimental systems and bacterial strains. We believe that our results may clarify the issue of N. gonorrhoeae and programmed cell death, through the identification of another bacterial factor (i.e., pilus retraction) involved in such a response.
A number of factors influence the ability of the cell to withstand apoptosis, including the signaling cascades that are activated and the degree and duration of the activation of these cascades \[[@pbio-0030100-b32]\]. They also include the virulence genes expressed by the infecting bacteria. The bacterial strains used for previous studies on N. gonorrhoeae and apoptosis differed in their piliation state and their ability to invade the host cell. Our results indicate that piliated bacteria, in the absence of pilus retraction, slightly increase the tendency of the infected cell to undergo apoptosis (see [Figure 5](#pbio-0030100-g005){ref-type="fig"}A). However, these bacteria are still able to moderately protect infected cells from STS-induced apoptosis, indicating that a certain level of cytoprotection is provided by other bacterial factors. In contrast, bacteria that can retract their pili, and thus presumably induce mechanical stress on the host-cell membrane, strongly mediate pro-survival signaling.
The influence of mechanical stress on apoptosis has been studied in some detail. Importantly, such studies indicate that different stress patterns result in different cellular outcomes. Extended, repetitive mechanical force increases the expression of genes encoding cytoprotective heat shock proteins and lowers the number of apoptotic cells in a culture \[[@pbio-0030100-b45]\]. Suppression of apoptosis requires permanent membrane tension or rhythmic, pulsatile forces \[[@pbio-0030100-b46]\], which are thought to allow the cell to adapt to new environmental conditions. Retraction events in N. gonorrhoeae generate strong, pulsatile forces every 1--20 s \[[@pbio-0030100-b13]\]. The nature of the pilus retraction force may therefore be the key to counteracting infection-induced apoptosis. Our data strongly indicate that pilus retraction from a microcolony is capable of stimulating mechanoprotective signals.
In light of the results presented here and elsewhere, we propose a model to explain how pilus retraction by N. gonorrhoeae influences survival signaling in the infected cell ([Figure 6](#pbio-0030100-g006){ref-type="fig"}). Initial contact between the bacterium and the epithelial cell activates MAP kinases and alters gene expression at a low level. The cell senses "stress" from the infection, the degree of which varies depending on the metabolic state of the cell and the constellation of virulence factors expressed by the infecting strain. As a result of this stress, the cell is poised to enter the apoptosis pathway. In the absence of pilus retraction and membrane tension, the low levels of activated MAP kinases may or may not be enough to counteract this stress. As the infection proceeds, microcolonies are formed. Pilus retraction from microcolonies is hypothesized to exert stress forces on the membrane, amplifying the levels of activated MAPK, enhancing the transcription of infection-induced genes, and possibly activating other as-yet-unidentified pathways. The end result is the enhanced stimulation of pro-survival pathways and an overriding of pro-apoptotic stress signals. In other words, the fate of the infected cell is decided by the type of signaling networks induced by infection and the extent of activation of these networks. Pilus retraction tips the balance in favor of cell survival.
::: {#pbio-0030100-g006 .fig}
Figure 6
::: {.caption}
###### Model of the Role of Pilus Retraction in Promoting a Cytoprotective Environment during Gonococcal Infection of an Epithelial Cell
\(1) Initial contact between the bacterium and host cell activates low levels of MAPK, and transcription of infection-induced genes. This level of signaling may or may not be able to protect the cell from apoptosis; thus, the host cell "teeters" on the edge of life and death.
\(2) As the infection proceeds, microcolonies of gonococci are formed, and more pili are locally available to retract.
\(3) Pilus retraction amplifies MAPK activation, which in turn enhances the transcription of mechanical stress--induced genes.
\(4) Pilus retraction may also stimulate other pathways that mediate gene expression and survival signaling. Overall signaling events tip the balance in favor of cell survival.
:::
:::
We have used a tissue culture system to study the interplay between pilus retraction, host cell signaling, and gene expression during the attachment phase of N. gonorrhoeae infection. How these interactions may affect the disease in vivo remains to be clarified. Our results make teleologic sense when the bacterial life cycle and gonococcal disease are taken into consideration. N. gonorrhoeae does not survive on fomites and has no intermediate host. Transmission depends on person-to-person spread. Simple mucosal gonorrhea infections can be mild, and inflammatory responses begin days after exposure \[[@pbio-0030100-b47]\]. Moreover, a significant number of infected individuals carry gonococci without overt symptoms of disease \[[@pbio-0030100-b47],[@pbio-0030100-b48]\]. Indeed, the ability of the bacterium to survive as a species requires a relatively healthy host. Our model for pilus retraction is consistent with these considerations.
Materials and Methods {#s4}
=====================
{#s4a}
### Reagents {#s4a1}
Antibodies to PARP, caspase 8, c-Jun, phospho-c-Jun (Ser63), P44/42 MAPK, phospho-p44/42 MAPK (Thr202/Tyr204), phospho-MAPKAPK2 (Thr334), p38 MAPK, phospho-p38 MAPK (Thr108/Tyr182), SAPK/JNK, and phospho-SAPK/JNK (Thr183/Tyr185) were purchased from Cell Signaling Technology (Beverly, Massachusetts, United States). MAPK inhibitors SB203588, U0126, and SP600125 were purchased from Calbiochem (San Diego, California, United States) and used at a final concentration of 10 μM unless otherwise stated. STS was purchased from Cell Signaling Technology and used at a final concentration of 1 μM to induce apoptosis. Neodymium iron boron (NdFeB) magnets (Eneflux Armtek Magnetics, Bethpage, New York, United States) measured 2 in. in diameter by 1 in. thick and were grade 30 (MGOe).
### Cell lines, bacterial strains, and infections {#s4a2}
T84 human colonic epidermoid cells (American Type Culture Collection, Manassas, Virginia, United States) were maintained in DMEM-F-12 plus 5% heat-inactivated, filter-sterilized fetal bovine serum at 37 °C and 5% CO~2~. For all experiments, cells were seeded into 35-mm dishes and allowed to become confluent before infection. N. gonorrhoeae strains N400 and N400*pilT* \[[@pbio-0030100-b49]\] were used for all infections and were maintained on GCB agar plus Kellogg\'s supplements at 37 °C and 5% CO~2~. Piliation and Opa phenotypes were monitored by colony morphology. Only piliated, Opa^−^ bacteria were used. For infection experiments, bacteria were resuspended in GCB liquid medium and added to the epithelial cells at a multiplicity of infection of 50.
### RNA isolation and microarray analysis {#s4a3}
T84 cells were infected with N400 or N400*pilT* or treated with GCB medium alone for 3 h. For RNA isolation, labeling, and microarray hybridization procedures, see [Protocol S1](#sd001){ref-type="supplementary-material"}. Comparative analysis was performed using MAS 5.0 algorithms to determine fold-change values between uninfected and infected samples from the same experiment, with uninfected samples representing the baseline. Statistical analysis was performed on natural-log transformed data using Cyber-T (<http://visitor.ics.uci.edu/genex/cybert/>). Subsequent data analysis was performed using Excel (Microsoft, Redmond, Washington, United States) and GeneSpring version 4.0 (Silicon Genetics, Redwood City, California, United States). Genes with a "presence call" *p*-value of less then 0.1 across all chips were eliminated from analysis, as were genes that were given a "no change" call across all samples. A gene was identified as differentially regulated if the fold-change was greater than ±1.5 in at least two out of three experiments. "Enhanced" genes were identified by calculating the ratio of the fold-change for the wt-infected cells to the fold-change for the *pilT*-infected cells (W/P). Gene expression was considered to be enhanced by pilus retraction if the W/P, averaged from at least two out of three individual experiments, was greater than 1.5, and the individual W/P from each experiment was greater than1.25.
### Real-time RT-PCR analysis {#s4a4}
One microgram of total RNA (as isolated above) was reverse-transcribed to generate cDNA, using the iScript cDNA synthesis kit (Bio-Rad, Hercules, California, United States). As a control, parallel samples were run in which reverse transcriptase was omitted from the reaction mixture. Quantitative real-time PCR was performed using an ABI PRISM 7000 Sequence Detection System (Applied Biosystems, Foster City, California, United States). Amplification was carried out using TaqMan master mix (Applied Biosystems), and pre-designed TaqMan probes (Assays on Demand, Applied Biosystems) according to the manufacturer\'s instructions. Assay numbers are given in [Table 1](#pbio-0030100-t001){ref-type="table"}. Reactions were performed in triplicate in a 20-μl volume, with the following cycle parameters: 95 °C/10 min enzyme activation, 95 °C/15 s, 60 °C/1 min for 40 cycles. Data analysis was performed using the comparative Ct method (Applied Biosystems) to determine relative expression levels.
::: {#pbio-0030100-t001 .table-wrap}
Table 1
::: {.caption}
###### Assays on Demand (TaqMan Probes and Primers) Used for Real-Time Quantitative RT-PCR in This Study
:::
:::
### Immunoblotting {#s4a5}
T84 cells were infected with N400 or N400*pilT* or treated with GCB medium alone for specified times. Following infection, cells were lysed with 150 μl of 1× SDS lysis buffer (62.5 mM Tris-HCl \[pH 6.8\], 2% w/v SDS, 10% glycerol, 50 mM DTT, 0.1% w/v bromophenol blue), scraped into Eppendorf tubes, vortexed for 15 s, and immediately stored at −20 °C. For PARP and caspase 8 assays, samples were incubated with 150 μl of cell lysis buffer (20 mM Tris \[pH 7.5\], 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 0.5% NP40, 2.5 mM sodium pyrophosphate, 1 mM β-glycerolphosphate, 1 mM Na~3~VO~4~, 1 μg/ml Leupeptin) for 20 min on ice, followed by a 15-s sonication. Samples were boiled for 5 min at 100 °C, then separated by SDS 8% polyacrylamide gels and transferred onto nitrocellulose sheets. Membranes were probed with the specified antibodies following the manufacturer\'s protocol.
### CPPs and bead coating {#s4a6}
N. gonorrhoeae CPPs were generated from piliated, Opa^−^ gonococci. Bacteria were scraped from overnight cultures (grown on plates) into HBSS and vortexed for 2 min, followed by centrifugation at 14,000*g* for 5 min. Supernatants were removed, quantitated by spectrophotometric analysis, and stored at −80 °C until use. Pili preparations were assayed for the presence of pili via indirect immunofluorescence microscopy and immunoblot, using anti-pilin antibody (data not shown). Bio-Mag Plus carboxy-modified paramagnetic microspheres (Bangs Laboratories, Fishers, Indiana, United States), were activated with 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide, hydrochloride (EDAC), and incubated with piliated N. gonorrhoeae CPPs or BSA as per the manufacturer\'s instructions. Bead coating was confirmed by immunoblotting using antibodies to BSA (ICN Biomedicals, Irvine, California, United States) and pilin (antibody SM1; data not shown).
### Immunofluorescence microscopy {#s4a7}
T84 cells were grown on coverslips to 50% confluency and incubated with either BSA-coated or CMP-coated magnetic beads for 15 min. Unbound beads were washed off and the magnet placed at a distance of 10 mm from the cell surface for 1 h. The medium was then aspirated, and the cells fixed for 15 min at room temperature in 4% paraformaldehyde. Cells were blocked and permeabilized in isotonic PBS containing BSA (3%, w/v) and saponin (0.02% w/v) for 1 h at room temperature, followed by staining with Alexa-Fluor 594 phalloidin (Molecular Probes, Eugene, Oregon, United States) at 1:1,000 for 30 min. Samples were rinsed extensively in PBS before mounting in Fluoromount-G (Fisher Scientific, Hampton, New Hampshire, United States). Images were obtained with a Deltavision Restoration Microscope (Applied Precision, Issaquah, Washington, United States) fitted with a Nikon (Tokyo, Japan) 60× oil-immersion objective and processed at a Silicon Graphics (Mountain View, California, United States) workstation with accompanying API software. The images were subsequently exported to Adobe Photoshop (version 7.0) and Adobe Illustrator (version 11.0) (Adobe Systems, San Jose, California, United States) for manuscript preparation.
### Calculation of magnetic force {#s4a8}
To quantify the amount of force that the magnet exerts per magnetic bead, the change-in-mass method \[[@pbio-0030100-b31]\] was used. Briefly, the mass of a known number of dry beads (0.12 g) was measured on an electronic balance in the presence and absence of the magnet. Given the mean bead diameter of 1.5 μm and the bead density of 2.5 × 10^3^ kg/m^3^ (Bangs Laboratories), the number of beads in this sample was calculated to be 1.2 × 10^10^. The change in mass of the beads in the presence of the magnet was entered into the equation---force = Δmass × acceleration (with acceleration being equal to gravity, or 9.81 m/s^2^)---to give a value for the force. Change-in-mass measurements were taken at varying distances from the magnet to determine force as a function of distance (see [Figure 4](#pbio-0030100-g004){ref-type="fig"}B).
### Magnetic force experiments {#s4a9}
T84 cells were grown to confluency in 35-mm culture dishes. Before assay, the cells were incubated with prewarmed, serum-free medium for 2 h. Cells were then incubated for 30 min with medium alone, or with CPP- or BSA-coated beads diluted in the same medium. Cells were then washed with fresh, serum-free medium to remove unbound beads. Magnets were placed at a distance of 10 mm from the bottom of the tissue culture dish, and the dishes were incubated for the specified time at 37 °C, 5% CO~2~. The samples were then processed for RNA isolation or SDS-PAGE, as described above. Control samples were treated in parallel but were not exposed to the magnet.
### Statistics {#s4a10}
Statistical analysis was performed using standard *t*-test analysis with SPSS version 11.0 (SPSS, Chicago, Illinois, United States) unless otherwise stated.
Supporting Information {#s5}
======================
Protocol S1
::: {.caption}
###### MIAME Checklist
(42 KB DOC).
:::
::: {.caption}
######
Click here for additional data file.
:::
Table S1
::: {.caption}
###### Supplementary References
(409 KB DOC).
:::
::: {.caption}
######
Click here for additional data file.
:::
Accession Numbers {#s5a3}
-----------------
The GenBank (<http://www.ncbi.nlm.nih.gov/Genbank/>) accession numbers for the genes and gene products discussed in this paper are *ADM* (D14874), *cyr61* (Y11307), DTR (M60278), *DUSP5* (U15932), EGR1 (X52541), PilF (U32588), and PilT (S72391).
We wish to thank S. W. Lee, J. Larson, and A. Friedrich for their thoughtful suggestions and careful reading of the manuscript. We also wish to thank the Affymetrix Microarray Core (OHSU Gene Microarray Shared Resource) for performing RNA labeling and hybridization. This work was supported in part by National Institutes of Health grant RO1-AI049973 awarded to MS, and National Institutes of Health grant T32-AI07472 awarded to HLH.
**Competing interests.** The authors have declared that no competing interests exist.
**Author contributions.** HLH and MS conceived and designed the experiments. HLH performed the experiments and analyzed the data. MG and MS contributed reagents/materials/analysis tools. HLH and MS wrote the paper.
Citation: Howie HL, Glogauer M, So M (2005) The N. gonorrhoeae type IV pilus stimulates mechanosensitive pathways and cytoprotection through a *pilT*-dependent mechanism. PLoS Biol 3(4): e100.
BSA
: bovine serum albumin
CPP
: crude pili preparation
MAPK
: mitogen-activated protein kinase
PARP
: poly(ADP-ribose) polymerase
STS
: staurosporine
Tfp
: type IV pili
W/P
: wild-type to *pilT* fold-change expression ratio
wt
: wild-type
|
PubMed Central
|
2024-06-05T03:55:54.903088
|
2005-3-22
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065265/",
"journal": "PLoS Biol. 2005 Apr 22; 3(4):e100",
"authors": [
{
"first": "Heather L",
"last": "Howie"
},
{
"first": "Michael",
"last": "Glogauer"
},
{
"first": "Magdalene",
"last": "So"
}
]
}
|
PMC1065266
|
Introduction {#s1}
============
Latent herpesviruses can persist as episomes in quiescent cells without viral protein expression. However, the gamma-herpesviruses are characteristically latent in memory lymphocytes, which intermittently divide. Viral genomes must therefore be replicated and segregated between daughter cells in step with cellular mitosis. This requires a viral episome maintenance protein, creating a potential target for the immune recognition of latently infected cells. A glycine--alanine repeat in the Epstein--Barr virus (EBV) episome maintenance protein, EBNA-1 \[[@pbio-0030120-b01]\], inhibits its degradation \[[@pbio-0030120-b02]\] and translation \[[@pbio-0030120-b03]\] such that EBNA-1 epitopes are poorly presented to CD8^+^ T lymphocytes (CTLs) \[[@pbio-0030120-b04],[@pbio-0030120-b05],[@pbio-0030120-b06]\]. However, the in vivo effectiveness of the glycine--alanine repeat and its quantitative contribution to host colonisation remain unknown. In vitro studies have suggested that abortive EBNA-1 translation products still provide sufficient epitopes for EBNA-1-specific CTLs to recognise latently infected B cells \[[@pbio-0030120-b07],[@pbio-0030120-b08],[@pbio-0030120-b09]\].
In vivo immune function is difficult to analyse directly with human viruses. However, the murid pathogen murine gamma-herpesvirus-68 (MHV-68) affords an opportunity to manipulate a gamma-herpesvirus in its natural host. MHV-68 is a gamma-2-herpesvirus \[[@pbio-0030120-b10]\], more closely related to the Kaposi\'s sarcoma--associated herpesvirus than to EBV \[[@pbio-0030120-b11]\], but clear functional parallels exist between all three viruses. Like EBV, MHV-68 causes an acute infectious mononucleosis-like illness, associated with a massive expansion of latently infected germinal centre B cells, and it persists in memory B cells \[[@pbio-0030120-b12],[@pbio-0030120-b13],[@pbio-0030120-b14]\]. The episome maintenance protein of gamma-2-herpesviruses is encoded by ORF73 \[[@pbio-0030120-b15]\]. Just as EBNA-1-deficient EBV \[[@pbio-0030120-b16]\] and ORF73-deficient Kaposi\'s sarcoma--associated herpesvirus \[[@pbio-0030120-b17]\] fail to maintain latency in vitro, ORF73-deficient MHV-68 has a profound latency deficit in vivo \[[@pbio-0030120-b18],[@pbio-0030120-b19]\].
At least two MHV-68 gene products inhibit CTL recognition of latently infected cells. The M3 chemokine-binding protein \[[@pbio-0030120-b20],[@pbio-0030120-b21]\] is abundantly secreted by lytic virus \[[@pbio-0030120-b22]\]. This co-exists with latent virus in infected lymphoid tissue \[[@pbio-0030120-b14],[@pbio-0030120-b23],[@pbio-0030120-b24]\]. Bystander protection by M3 \[[@pbio-0030120-b25]\] probably explains why M3 disruption causes mainly a reduction in MHV-68 latency amplification \[[@pbio-0030120-b26],[@pbio-0030120-b27]\]. Protection by M3 may be quite context-dependent \[[@pbio-0030120-b28]\] and probably functions best in acute infection, before M3-specific immunity is established. A second MHV-68 CTL evasion protein, K3, degrades major histocompatibility complex (MHC) class I heavy chains \[[@pbio-0030120-b29]\] and TAP (transporter associated with antigen processing) \[[@pbio-0030120-b30]\]. K3 is transcribed in latency as well as in the viral lytic cycle, and again protects against CTLs during latency amplification \[[@pbio-0030120-b31]\]. However, K3 is not expressed in all forms of latency: it is not detectable in the MHV-68-infected S11 tumor line \[[@pbio-0030120-b32]\] or in persistently infected, B cell--deficient mice \[[@pbio-0030120-b33]\]. Notably, ORF73 disruption causes a much more profound latency deficit than does K3 disruption, implying that immune evasion by K3 is not important in every cell expressing ORF73. These restrictions on M3 and K3 function imply that further immune evasion mechanisms contribute to MHV-68 latency.
Selective EBNA-1 expression occurs in normal EBV persistence \[[@pbio-0030120-b34],[@pbio-0030120-b35]\] and is characteristic of EBV-associated Burkitt\'s lymphoma \[[@pbio-0030120-b36]\]. The gamma-2-herpesviruses presumably implement an equivalent program of selective ORF73 expression. We have used MHV-68 to determine the in vivo importance of avoiding epitope presentation during episome maintenance. We first established that the MHV-68 ORF73 is equivalent to EBNA-1 in reducing the presentation of an MHC class I--restricted epitope linked to it in *cis.* We then modified the ORF73 transcript to bypass this evasion, and used the mutant virus to define the consequences of epitope presentation for viral fitness. Our analysis of MHV-68 implies that if MHC class I--restricted viral epitope presentation occurs during gamma-herpesvirus episome maintenance, even a relatively small CTL response can very effectively clear latent infection.
Results {#s2}
=======
Limited Presentation of a CD8^+^ T Cell Epitope Linked in *cis* to ORF73 {#s2a}
------------------------------------------------------------------------
As with EBNA-1, CTL epitopes in the MHV-68 ORF73 have been hard to find, perhaps reflecting limited ORF73 entry into the MHC class I antigen-processing pathway. To determine whether ORF73 is similar to EBNA-1 in its resistance to MHC class I--restricted antigen presentation, we introduced the H2-K^b^-restricted SIINFEKL epitope of ovalbumin (OVA) near the ORF73 C-terminus (73-SC) or N-terminus (73-SN). L929-K^b^ cells transfected with 73-SC or 73-SN were poorly recognised by the SIINFEKL-specific hybridoma, B3Z ([Figure 1](#pbio-0030120-g001){ref-type="fig"}A), suggesting poor antigen processing. There was no evidence for ORF73 inhibiting SIINFEKL presentation from co-transfected OVA ([Figure 1](#pbio-0030120-g001){ref-type="fig"}B). The apparent immune evasion therefore acted in *cis* rather than in *trans,* and it was not due to any ORF73 toxicity.
::: {#pbio-0030120-g001 .fig}
Figure 1
::: {.caption}
###### Inhibition of MHC Class I--Restricted Epitope Presentation by Physical Linkage to ORF73
\(A) The SIINFEKL epitope of OVA was introduced into ORF73 near either its N-terminus (ORF73-NC) or its C-terminus (ORF73-SC). Both ORF73 derivatives were cloned into the pcDNA3 expression vector and compared with OVA in the same vector for their capacity to stimulate the SIINFEKL-specific T cell hybridoma B3Z after transfection into L929-K^b^ cells. After 48 h, beta-galactosidase production was assayed by cell lysis in the presence of chlorophenol-red-beta-[D]{.smallcaps}-galactoside and reading absorbance at 595 nm. nil, vector only.
\(B) L929-K^b^ cells were co-transfected with OVA plus the plasmid indicated. C1--C4 are control plasmids, expressing MHV-68 ORFs 19, 30, 31, and 35, respectively. K3 degrades MHC class I heavy chains and m152 is a murine cytomegalovirus gene that retains MHC class I molecules in the endoplasmic reticulum. Net absorbance = A~595~ with co-transfection − A~595~ with untransfected cells (\<0.02).
\(C) Hybrids of OVA and ORF73-SC were made to identify regions of ORF73 that inhibited SIINFEKL presentation. Responses are expressed as 100(A~595~ with plasmid − A~595~ with untransfected)/(A~595~ with OVA transfection − A~595~ with untransfected). nil, vector only. Mean ± standard deviation (SD) values of triplicate cultures are shown. Each graph is representative of at least three separate experiments. In at least one experiment, equal transfection efficiency was confirmed by co-transfecting a GFP expression plasmid and checking fluorescence under ultraviolet illumination.
\(D) ORF73 was fused to the C-terminus of the OVA coding sequence in pcDNA3. C-terminal deletions were then made as shown. Each construct was transfected into L929-H2-K^b^ cells. The shaded area in (D--G) highlights a region of ORF73 that appeared to be important for inhibiting epitope presentation.
\(E) N-terminal ORF73 truncations were generated by PCR and fused in frame to amino acid 325 of OVA. Each construct was transfected into L929-H2-K^b^ cells and assayed for SIINFEKL presentation as in (D).
\(F) Progressive truncations of ORF73-SN were assayed for their capacity to present the SIINFEKL epitope to B3Z cells after transfection into L929-K^b^ cells. Selective presentation from the ORF73-SN-PstI construct was confirmed in multiple experiments, including independent plasmid preparations.
\(G) PCR-generated C-terminal truncations of ORF73-SN were assayed for SIINFEKL presentation after transfection of L929-H2-K^b^ cells. Deletions across the area identified as important for inhibiting epitope presentation in (D--E) again improved epitope presentation.
:::
:::
To identify possible contributions of the N-terminal and C-terminal regions of ORF73 to its poor epitope presentation, we made hybrids of OVA (amino acid residues 1--325) and 73-SC using a shared PstI site ([Figure 1](#pbio-0030120-g001){ref-type="fig"}C). N-terminal ORF73 (amino acid residues 1--150) diminished SIINFEKL presentation from OVA relatively little, suggesting that any inhibitory segment was C-terminal of its PstI site (amino acid residues 151--314). With N-terminal OVA, SIINFEKL presentation from 73-SC remained low.
A Key Region of ORF73 for *cis-*Acting Immune Evasion {#s2b}
-----------------------------------------------------
We adopted two strategies to localise an inhibitory segment C-terminal of the ORF73 PstI site. First, we fused the entire ORF73 coding sequence to the C-terminus of OVA, thereby inhibiting SIINFEKL presentation from OVA. C-terminal truncation of the fusion protein ([Figure 1](#pbio-0030120-g001){ref-type="fig"}D) up to the ORF73 HinDIII site (residue 278) then had little effect, but truncation up the ORF73 KpnI site (residue 206) restored SIINFEKL presentation somewhat. N-terminal truncation ([Figure 1](#pbio-0030120-g001){ref-type="fig"}E) to ORF73 residue 166 (OVA-73C) did not compromise the inhibition, but further truncation to residue 186 (OVA-73B) and then residue 206 (OVA-73A) progressively improved SIINFEKL presentation. These data were consistent with the region of ORF73 encoding approximately residues 170--220 (the shaded region in [Figure 1](#pbio-0030120-g001){ref-type="fig"}D--[1](#pbio-0030120-g001){ref-type="fig"}G) reducing antigen presentation.
We also truncated 73-SN from its C-terminus and looked for presentation of its N-terminal SIINFEKL epitope ([Figure 1](#pbio-0030120-g001){ref-type="fig"}F). With SIINFEKL in this context, the same C-terminal truncations as in [Figure 1](#pbio-0030120-g001){ref-type="fig"}D (amino acid residues 1--204) gave no epitope presentation, presumably because SIINFEKL was less efficiently processed from 73-SN than it was from OVA. However, a further truncation up to the ORF73 PstI site (residue 150) dramatically improved SIINFEKL presentation. PCR-generated C-terminal truncations of 73-SN ([Figure 1](#pbio-0030120-g001){ref-type="fig"}G) supported the idea of a region just upstream of the ORF73 KpnI site (ORF73-SN-B) limiting the presentation of SIINFEKL from 73-SN. These results were therefore consistent with those shown in [Figure 1](#pbio-0030120-g001){ref-type="fig"}D and [1](#pbio-0030120-g001){ref-type="fig"}E.
The Effects of ORF73 on the Turnover of Linked OVA {#s2c}
--------------------------------------------------
The inhibition of EBNA-1 epitope presentation by its glycine--alanine repeat has been attributed principally to reduced protein synthesis and secondarily to reduced protein degradation \[[@pbio-0030120-b03]\]. We therefore analysed the effect of ORF73 on OVA turnover using constructs equivalent to those in [Figure 1](#pbio-0030120-g001){ref-type="fig"}E, except that we removed the signal sequence of OVA to avoid any protein secretion (SOVA, [Figure 2](#pbio-0030120-g002){ref-type="fig"}A). All constructs were cloned into pcDNA3 and transfected into 293T cells. We observed a hierarchy of SOVA/ORF73 antigen presentation ([Figure 2](#pbio-0030120-g002){ref-type="fig"}B) similar to that seen with the OVA/ORF73 hybrids: amino acids 206--314 of ORF73 (SOVA-73A) reduced somewhat SIINFEKL presentation from SOVA; amino acids 166--314 (SOVA-73C) reduced it further; including an additional 40 amino acids of ORF73 (SOVA-73E) gave no additional inhibition.
::: {#pbio-0030120-g002 .fig}
Figure 2
::: {.caption}
###### Inhibition of Epitope Presentation by ORF73 Fusion to OVA Correlates with Reduced Translation of the Fusion Protein
\(A) N-terminal ORF73 truncations equivalent to those in [Figure 1](#pbio-0030120-g001){ref-type="fig"}E were fused in frame to the C-terminus of OVA amino acids 41--325, thereby removing both the OVA signal sequence and the ORF73 nuclear localisation signal.
\(B) Serial dilutions of an expression plasmid containing each fusion gene were transfected into L929-H2-K^b^ cells as in [Figure 1](#pbio-0030120-g001){ref-type="fig"}. SIINFEKL presentation was assayed using beta-galactosidase production from the B3Z hybridoma.
\(C) Equivalent transfected cells were immunoblotted with an anti-OVA rabbit serum (OVA). Fusion products are indicated by arrowheads where visible. Parallel immunoblots for neomycin phosphotransferase II (NPT), which is expressed from a different promoter of the same plasmid (pcDNA3), were used to control for transfection efficiency. The endogenous neomycin phosphotransferase II expressed by 293T cells was not visible at this exposure. One of three equivalent experiments is shown.
\(D) Forty-eight hours after transfection with the constructs indicated, 293T cells were pulse-labelled (P) for 30 min with ^35^S-cysteine/methionine, followed by a 2-h chase (C) with excess unlabelled cysteine/methionine. OVA derivatives were then immunoprecipitated with an OVA-specific rabbit serum and resolved by SDS-PAGE. The specific bands corresponding to each fusion protein are indicated by arrowheads. The graph shows densitometry readings for each band.
\(E) 293T cells transfected with selected fusion proteins were labelled for a variable period (15--120 min) as indicated. OVA derivatives were then immunoprecipitated and analysed as in (D). Arrowheads show the predicted position of the relevant fusion proteins for the 120-min label samples.
\(F) Either SOVA-ORF73A or SOVA was transfected into 293T cells. Forty-eight hours later the cells were pulse-labelled (P) for 15 min with ^35^S-cysteine/methionine, followed by a 15-min (C1), 45-min (C2), and 105-min (C3) chase with excess unlabelled cysteine/methionine. This was done in the presence or absence of 100 μM lactacystin. The graph shows densitometry readings for each specific band.
:::
:::
Steady-state protein levels, determined by immunoblotting transfected cell lysates ([Figure 2](#pbio-0030120-g002){ref-type="fig"}C), were greatest with SOVA, followed by SOVA-73A, and least with SOVA-73D-E. Parallel immunoblots for neomycin phosphotransferase II, expressed from a different promoter of the same plasmid, showed no significant variation in signal, arguing against an effect of ORF73 on cell viability or transfection efficiency. Pulse-chase metabolic labelling of transfected 293T cells and immunoprecipitation with an OVA-specific antiserum ([Figure 2](#pbio-0030120-g002){ref-type="fig"}D) showed that all the ORF73/SOVA fusions were more stable than SOVA alone, and that their labelling was reduced as more ORF73 sequence was attached. The differences in protein synthesis rate ([Figure 2](#pbio-0030120-g002){ref-type="fig"}D) correlated with steady-state protein levels (see [Figure 1](#pbio-0030120-g001){ref-type="fig"}C). Using a variable labelling window ([Figure 2](#pbio-0030120-g002){ref-type="fig"}E), SOVA and SOVA-73A were detectable after a 15-min pulse, whereas SOVA-73E was hard to discern even after a 120-min pulse.
The apparent stability conferred by ORF73 on SOVA ([Figure 2](#pbio-0030120-g002){ref-type="fig"}D) was confirmed by further analysis of SOVA-73A, the fusion protein that labelled most efficiently ([Figure 2](#pbio-0030120-g002){ref-type="fig"}F). After a 20-min pulse, most labelled SOVA was lost over the next 2 h, whereas the labelled SOVA-73A was relatively well maintained. Proteasome inhibition with lactacystin partially stabilised SOVA, but it remained less stable than SOVA-73A. Thus, the stability afforded by amino acids 206--314 of ORF73 appeared to extend beyond protection against proteasome-mediated degradation. Overall, the *cis-*acting immune evasion of ORF73 appeared functionally similar to that of EBNA-1, in that reduced epitope presentation was associated with reduced protein synthesis and reduced protein degradation. In contrast to EBNA-1, these functions were mediated by distinct regions of ORF73. The key region for inhibiting epitope presentation (see [Figure 1](#pbio-0030120-g001){ref-type="fig"}D--[1](#pbio-0030120-g001){ref-type="fig"}G) corresponded to that responsible for reducing protein synthesis.
An MHV-68 Mutant That Lacks *cis-*Acting Immune Evasion {#s2d}
-------------------------------------------------------
The region of ORF73 responsible for *cis-*acting immune evasion---that encoding amino acids 170--220---shows considerable amino acid homology to both EBNA-1 and the Kaposi\'s sarcoma--associated herpesvirus ORF73 \[[@pbio-0030120-b37]\]. Any mutagenesis was therefore likely to compromise other ORF73 functions. Also, altering this region would change the steady-state levels of ORF73, with likely toxic effects in latently infected cells. In order to keep ORF73 function intact, therefore, we bypassed the *cis-*acting immune evasion not by mutating the ORF73 protein, but by modifying its mRNA. Thus, we inserted an internal ribosome entry site (IRES) just downstream of the ORF73 coding region and used this to co-express either green fluorescent protein (GFP) or three tandem MHC class I--binding peptides ([Figure 3](#pbio-0030120-g003){ref-type="fig"}A).
::: {#pbio-0030120-g003 .fig}
Figure 3
::: {.caption}
###### Modification of the MHV-68 Genome to Overcome *cis-*Acting Immune Evasion by ORF73
\(A) An IRES element was inserted just downstream of ORF73, between its stop codon and that of M11. This allowed either three tandem CD8^+^ T cell epitopes (EPI) or GFP to be translated from the ORF73 mRNA.
\(B) DNA from BAC-cloned viral genomes (BAC) or virus-infected cells (VIR) was digested with NcoI, electrophoresed, transferred to nylon membranes, and blotted with a probe corresponding to the BamHI-G genomic fragment shown in (A). The predicted bands for WT virus were 1,021 bp, 3,121 bp, and 4,630 bp. The IRES-GFP insert introduced an NcoI site such that the WT 3,121-bp band was cut into 2,975-bp and 1,466-bp fragments. The NcoI site was lost from the IRES-EPI insert, such that the WT 3,121-bp band became a 3,861-bp band.
\(C) BHK-21 cells were infected (0.01 PFU/cell) with WT, GFP, or EPI viruses as indicated. Plaque titres of cell cultures are shown with time after infection.
\(D) H2^b^ MEF-1 cells or L929-K^b^ cells were left uninfected (UI) or infected for 2 h with MHV-68 expressing either OVA under a strong lytic promoter (OVA) or the SIINFEKL epitope of OVA as part of the ORF73-IRES-EPI construct (EPI). B3Z cells were then added, and 18 h later their beta-galactosidase response was assayed using chlorophenol-red-beta-[D]{.smallcaps}-galactoside substrate. Mean ± SD values of triplicate cultures are shown. The data are from one or two equivalent experiments.
\(E) A20-syndecan-1 cells were infected (20 PFU/cell) with GFP^−^ WT virus, WT virus with an HCMV IE1 promoter-driven GFP expression cassette (HCMV IE1-GFP), or with the ORF73-IRES-GFP virus. The numbers indicate the percentage of total cells in the gated region (GFP^+^). Expression from the HCMV IE1 promoter is probably limited to lytic infection, whereas ORF73 is expressed in latency.
:::
:::
Southern blots confirmed the predicted genomic structure of the ORF73-IRES-epitope (EPI) and ORF73-IRES-GFP viruses ([Figure 3](#pbio-0030120-g003){ref-type="fig"}B). Both mutants showed unimpaired growth in vitro ([Figure 3](#pbio-0030120-g003){ref-type="fig"}C). Infection of H2-K^b^ fibroblasts with the EPI virus established that its SIINFEKL epitope could be processed and presented ([Figure 3](#pbio-0030120-g003){ref-type="fig"}D). MHV-68 expressing OVA from an intergenic expression cassette under the control of an ectopic viral M3 promoter (MHV-OVA), which shows high-level lytic cycle OVA production (data not shown), was tested in parallel. In murine embryonic fibroblast (MEF)-1 cells, which support MHV-68 lytic replication, MHV-OVA showed better SIINFEKL presentation than did the EPI virus. In L929 cells, which support viral entry into the lytic cycle relatively poorly \[[@pbio-0030120-b38]\], the EPI virus gave better SIINFEKL presentation than did MHV-OVA. These data were consistent with the EPI virus presenting SIINFEKL in latency.
The GFP mutant provided further evidence that the IRES constructs were expressed in latency. Although MHV-68 is predominantly latent in B cells in vivo, it appears to infect B cells poorly in vitro. This may reflect that efficient infection by MHV-68 virions requires cell-surface glycosaminoglycans \[[@pbio-0030120-b39]\]. We therefore enhanced infection of the A20 B cell line by transducing it with a retroviral vector expressing the extracellular domain of syndecan-1, a major carrier of cell-surface glycosaminoglycans \[[@pbio-0030120-b40]\], linked to the transmembrane and cytoplasmic domains of H2-D^b^. This form of syndecan-1 resists proteolytic cleavage. GFP expressed from a human cytomegalovirus IE-1 promoter \[[@pbio-0030120-b41]\] gave green fluorescence in relatively few A20-syndecan-1 cells. Although the ORF73-IRES-GFP virus gave much weaker fluorescence, many more cells were positive ([Figure 3](#pbio-0030120-g003){ref-type="fig"}E). These data were consistent with B cell infection being predominantly latent, and with gene expression from the IRES constructs in latency.
Epitope Presentation during Episome Maintenance Leads to a Severe In Vivo Latency Deficit {#s2e}
-----------------------------------------------------------------------------------------
We tested the capacity of the EPI virus to replicate in vivo by intranasal infection of C57BL/6J mice ([Figure 4](#pbio-0030120-g004){ref-type="fig"}). There was no difference between wild-type (WT) and EPI viruses in lytic replication in lung epithelial cells or in seeding latent virus to the spleen ([Figure 4](#pbio-0030120-g004){ref-type="fig"}A). However, by 14 d after infection, when WT virus had reached its peak latent load, the titre of EPI virus was drastically reduced ([Figure 4](#pbio-0030120-g004){ref-type="fig"}B). In agreement with the reduced number of infectious centres, the EPI virus genome load was low ([Figure 4](#pbio-0030120-g004){ref-type="fig"}C) and there was little virus-driven B cell activation, T cell activation, or Vbeta4^+^CD8^+^ T cell expansion ([Figure 4](#pbio-0030120-g004){ref-type="fig"}D). The GFP control virus showed no such deficit, so it was not the IRES element that compromised ORF73 function. An independently derived EPI mutant showed a similar in vivo latency deficit, and reverting the EPI mutation restored latency establishment to normal levels ([Figure 4](#pbio-0030120-g004){ref-type="fig"}E). The latency deficit was therefore due specifically to the expression of a poly-epitope construct downstream of ORF73.
::: {#pbio-0030120-g004 .fig}
Figure 4
::: {.caption}
###### Replication of the IRES-EPI Virus In Vivo
\(A) Six days after intranasal infection with WT or EPI viruses as indicated, infectious virus in lungs was titred by plaque assay (left panel) and infectious plus latent virus in spleens was titred by infectious centre assay (right panel). Each point shows an individual mouse. Pre-formed, infectious virus was undetectable in equivalent, freeze-thawed spleen samples, so the infectious centres represent latent virus.
\(B) By 14 d post-infection, infectious centre titres were much lower with the EPI virus than with WT. The GFP control virus is shown for comparison. This difference was preserved at day 19 post-infection, indicating that the EPI virus was not merely delayed in host colonisation.
\(C) DNA was extracted from spleens and its viral genome content quantitated by real-time PCR. Genome loads broadly reflected the infectious centre titres, indicating that the viral load was reduced rather than the efficiency of ex vivo reactivation.
\(D) As a further measure of host colonisation, we measured B cell activation (CD69 expression on CD19^+^ B cells) at 14 d post-infection and CD8^+^ T cell activation (loss of CD62L expression) at 19 d post-infection. We also measured the day 19 expansion of the Vbeta4^+^CD8^+^ T cell subset that is characteristic of MHV-68-associated infectious mononucleosis. All these measures correlated closely with the viral latent load in lymphoid tissue and were markedly reduced with the EPI virus compared to WT or GFP. GFP expression was undetectable in ex vivo B cells after infection with the GFP virus (data not shown).
\(E) C57BL/6J mice were infected intranasally with WT virus, the EPI mutant, an independently derived EPI mutant (EPI-IND), or a revertant of the EPI virus (EPI-REV). Splenic infectious centres were then measured 13 and 17 d post-infection. The dashed line shows the lower limit of assay sensitivity.
:::
:::
Antigen-Specific Immune Responses to the EPI Virus {#s2f}
--------------------------------------------------
The EPI virus was notably controlled without a need for the massive T cell activation that characterises MHV-68- or EBV-associated infectious mononucleosis ([Figure 4](#pbio-0030120-g004){ref-type="fig"}D). We measured virus-specific immune responses ([Figure 5](#pbio-0030120-g005){ref-type="fig"}) to gain some idea of what effector response might be responsible for the latency amplification deficit. At 13 d post-infection, ELISPOT assays ([Figure 5](#pbio-0030120-g005){ref-type="fig"}A) showed a low CD4^+^ T cell response to the EPI virus compared to WT. A similar reduction in CD4^+^ T cell response is seen with MHV-68 specifically made to be latency deficient \[[@pbio-0030120-b42]\], presumably because lytic reactivation after latency amplification normally provides a large CD4^+^ T cell stimulus. Virus-specific serum antibody titres were marginally higher in the EPI-virus-infected mice ([Figure 5](#pbio-0030120-g005){ref-type="fig"}B). CD8^+^ T cell responses to immunodominant MHV-68 lytic epitopes (p56 and p79) were comparable between WT and EPI viruses at day 13 post-infection ([Figure 5](#pbio-0030120-g005){ref-type="fig"}A and [5](#pbio-0030120-g005){ref-type="fig"}C). Thus, the EPI virus was most likely being cleared by CTLs directed against an ORF73-associated epitope.
::: {#pbio-0030120-g005 .fig}
Figure 5
::: {.caption}
###### Antigen-Specific Immune Responses to the IRES-EPI Virus
\(A) CD8^+^ and CD4^+^ T cell responses were measured by interferon-gamma ELISPOT assay 13 d post-infection. The response to virus-exposed targets (VIR) is mediated by CD4^+^ T cells; the response to the p56, p79, and SIINFEKL (OVA) peptides is mediated by CD8^+^ T cells \[[@pbio-0030120-b62]\]. The mean number of spots with untreated targets was subtracted from the number of spots with each specific target. There was a response to the OVA peptide in the IRES-epitope construct, but not to the ASNENMETM peptide (NP). Mean ± SD values of five mice per group are shown.
\(B) Total and MHV-68 virion-specific serum IgG responses were measured by ELISA at 18 d post-infection. "Naive" indicates age-matched, uninfected controls. Mean ± SD absorbance values of four mouse sera per group are shown.
\(C) Spleen cells were stimulated for 5 h in the presence of Brefeldin A plus the peptide indicated and then stained for cell-surface CD8 and intracellular interferon-gamma. The percentage of interferon-gamma^+^ CD8^+^ cells without peptide was subtracted from the value with peptide to give the specific response. Mean ± SD values of five mice per group are shown.
:::
:::
There was no ASNENMETM-specific response to the EPI virus ([Figure 5](#pbio-0030120-g005){ref-type="fig"}A and [5](#pbio-0030120-g005){ref-type="fig"}C). Also, there was no evidence of an enhanced response to the p79 epitope, which was present both in the IRES-epitope construct and in its native context in ORF61 \[[@pbio-0030120-b43]\]. These epitopes were probably not processed from the polytope construct, since transfecting it as an expression plasmid into H2-K^b^- or H2-D^b^-expressing L929 cells stimulated the SIINFEKL-specific hybridoma B3Z but not T cell hybridomas specific for ASNENMETM or p79 (data not shown). In contrast to the lack of response to ASNENMETM, there was a clear response to the SIINFEKL epitope co-expressed with ORF73 ([Figure 5](#pbio-0030120-g005){ref-type="fig"}A and [5](#pbio-0030120-g005){ref-type="fig"}C), which was also presented in vitro (see [Figure 3](#pbio-0030120-g003){ref-type="fig"}D). By 18 d post-infection ([Figure 5](#pbio-0030120-g005){ref-type="fig"}C), the CD8^+^ T cell response to WT infection had made its characteristic shift in immunodominance from the p56 epitope associated with epithelial infection to the p79 epitope associated with B cell infection \[[@pbio-0030120-b43]\]. This did not occur with the EPI virus, presumably because the number of latently infected B cells remained low. Thus, it seemed likely that SIINFEKL-specific CTLs eliminated the EPI virus.
Attenuation of the EPI Virus Is H2-Type-Restricted {#s2g}
--------------------------------------------------
As all of the CTL epitopes in the IRES-epitope construct were H2^b^-restricted, a major prediction was that the EPI virus would not be attenuated in H2^d^ mice. This was found to be the case ([Figure 6](#pbio-0030120-g006){ref-type="fig"}). In H2^d^ BALB/c mice, the EPI virus attained infectious centre titres in the spleen equivalent to WT virus ([Figure 6](#pbio-0030120-g006){ref-type="fig"}A). B cell activation ([Figure 6](#pbio-0030120-g006){ref-type="fig"}B), splenomegaly ([Figure 6](#pbio-0030120-g006){ref-type="fig"}C), and viral genome load ([Figure 6](#pbio-0030120-g006){ref-type="fig"}D) were also normal. We further assayed latency by in situ hybridization for the expression of viral tRNA homologues in splenic germinal centres ([Figure 6](#pbio-0030120-g006){ref-type="fig"}E). These are expressed at high levels in MHV-68-infected lymphoid tissue and provide an additional marker of latency establishment \[[@pbio-0030120-b44]\]. The EPI virus showed no viral tRNA^+^ cells in C57BL/6J mice and normal numbers in BALB/c mice. These data supported the idea that the attenuation of the EPI virus was due to the expression of a CTL target from the ORF73 mRNA.
::: {#pbio-0030120-g006 .fig}
Figure 6
::: {.caption}
###### Normal EPI Virus Replication in Non-H2^b^ Mice
\(A) Infectious centre titres in individual spleens were determined 14 d after intranasal infection of C57BL/6J (H2^b^) or BALB/c (H2^d^) mice with WT or EPI virus.
\(B) CD69 expression on splenic B cells was measured by flow cytometry 14 d post-infection. B cells from uninfected mice were less than 5% CD69^+^.
\(C) The weights of individual spleens are shown 14 d post-infection, with spleens of age-matched, uninfected mice (UI) for comparison.
\(D) Viral genome loads in individual mice were determined by real-time PCR at 14 d post-infection.
\(E) Viral tRNA expression in infected germinal centres was visualised by in situ hybridization with a digoxigenin-labelled riboprobe specific for tRNAs 1--4. Representative follicles of at least five sections per mouse and three mice per group are shown. tRNA^+^ follicles were abundant with WT virus and with the EPI virus in BALB/c mice, but were not seen with the EPI virus in C57BL/6J mice.
\(F) BALB/c or C57BL/6J mice were infected intranasally with MHV-68 expressing OVA from an intergenic expression cassette (OVA) or with WT virus. The extent of lymphoid colonisation was determined by infectious centre assay of spleens 12 and 15 d post-infection. Mean ± SEM titres of five mice per group are shown. In contrast to the EPI virus, the OVA virus showed no defect in host colonisation.
:::
:::
In contrast to the EPI virus, MHV-OVA showed no significant attenuation in either BALB/c or C57BL/6J mice compared to WT virus ([Figure 6](#pbio-0030120-g006){ref-type="fig"}F). MHV-OVA expresses the SIINFEKL epitope at high levels during lytic infection (see [Figure 3](#pbio-0030120-g003){ref-type="fig"}D). Thus, SIINFEKL expression during episome maintenance, when epitopes are not normally presented, was catastrophic for the virus, whereas SIINFEKL expression outside of this context, when MHV-68 does not rely on limiting epitope presentation for its survival, had little effect.
Attenuation of the EPI Virus Is CD8^+^ T Cell--Dependent {#s2h}
--------------------------------------------------------
The MHC class I restriction of the EPI virus\'s latency deficit and its association with anti-SIINFEKL immunity implied that CD8^+^ T cells were eliminating latently infected cells. This was confirmed by rescuing the EPI virus with CD8^+^ T cell depletion ([Figure 7](#pbio-0030120-g007){ref-type="fig"}). Thus, in C57BL/6J mice treated with an anti-CD8 monoclonal antibody, the EPI virus achieved WT levels of B cell activation ([Figure 7](#pbio-0030120-g007){ref-type="fig"}A), viral genome load ([Figure 7](#pbio-0030120-g007){ref-type="fig"}B), and infectious centres ([Figure 7](#pbio-0030120-g007){ref-type="fig"}C). The ORF73 CTL evasion (see [Figure 1](#pbio-0030120-g001){ref-type="fig"}) that was bypassed in the EPI virus was therefore essential for in vivo episome maintenance.
::: {#pbio-0030120-g007 .fig}
Figure 7
::: {.caption}
###### Rescue of the EPI Virus by CD8^+^ T Cell Depletion
Mice were left undepleted (UD) or depleted of CD8^+^ T cells (CD8^−^) by an initial intravenous injection of mAb YTS169 2 d before infection, followed by intraperitoneal injections of the same antibody every 2--3 d up to the time of sampling. Infection was by intranasal inoculation of either WT or EPI viruses.
\(A) Depletion was 95%--99% complete as assessed by flow cytometry of spleen cells. CD69 expression on splenic B cells was measured 13 d post-infection.
\(B) Genome loads were measured 13 d post-infection by real-time PCR. Each point shows an individual mouse.
\(C) The infectious centre titres of individual mice at 13 d post-infection are shown for one of two equivalent experiments. The titres of pre-formed, infectious virus in freeze-thawed spleens were less than 5% of the infectious centre titres, so even after CD8^+^ T cell depletion, the infectious centre assay essentially measured latent virus. By 13 d post-infection, the lungs of both immunocompetent and CD8^+^ T cell--depleted mice were clear of infectious virus.
\(D) In situ hybridization for viral tRNA expression in splenic germinal centres is shown 13 d after infection of CD8^+^ T cell--depleted or undepleted mice, infected with either EPI or WT virus. Spleens of two representative mice are shown in each case.
:::
:::
Discussion {#s3}
==========
The MHV-68 ORF73 lacks the glycine--alanine repeat of EBNA-1 but still conferred poor presentation on a linked CTL epitope. Thus, despite different means, MHV-68 and EBV have arrived at a similar end of inhibiting epitope presentation during episome maintenance. Neither molecular mechanism is fully understood, but both seem to rely primarily on limiting protein synthesis. A large proportion of CTL epitopes are derived from abortive translation events \[[@pbio-0030120-b45]\]. Understanding the mechanism of *cis-*acting CTL evasion therefore means understanding the major source of abortive translation events, whether damaged RNA, ribosomal errors, or protein misfolding. Since the key evasion regions of EBNA-1 and ORF73 are located centrally, they are translated only after potential N-terminal epitopes, and so may exert their inhibitory effects prior to translation, as RNA. Our aim here was to ask what *cis-*acting evasion contributes to the fitness of a gamma-herpesvirus. Inserting an IRES element downstream of ORF73 allowed us to bypass *cis-*acting immune evasion in MHV-68. CTLs then wiped out latency. We conclude that avoiding epitope presentation during episome maintenance is fundamental to gamma-herpesvirus survival.
As yet, no endogenous CTL epitopes to our knowledge have been described for the MHV-68 ORF73. This may reflect its *cis-*acting immune evasion in the same way that EBV infection was initially thought not to elicit EBNA-1-specific CTLs \[[@pbio-0030120-b46],[@pbio-0030120-b47]\]. However, EBNA-1 epitopes can be presented by cross-priming \[[@pbio-0030120-b05]\]. It seems likely that MHV-68 will elicit ORF73-specific CTLs by a similar route. Certainly there is no lack of predicted cleavable, MHC class I--binding peptides in ORF73; for example, H2-D^b^, FSSTHPYTL; H2-K^b^, QCVTYYLL; H2-D^d^ and H2-K^d^, KYQGMRRHL; and H-L^d^, APPSPDVDV. Thus, evasion must occur at the level of endogenous ORF73 presentation.
The effectiveness of immune evasion is inevitably context-dependent. Defining its impact on host colonisation therefore requires natural thresholds of in vivo antigen presentation. The results are not always predictable. For example, MHV-68 transcribes its K3 gene in the lytic cycle as well as in latency, but a lack of K3 has no discernable impact on primary lytic infection, only on latency amplification \[[@pbio-0030120-b31]\]. The recognition of EBNA-1 during latency III \[[@pbio-0030120-b07],[@pbio-0030120-b08],[@pbio-0030120-b09]\] does not necessarily imply EBNA-1 recognition during latency I, when autoregulation \[[@pbio-0030120-b48]\] and a cell cycle dependence \[[@pbio-0030120-b49]\] of the Qp promoter reduce EBNA-1 transcription. Our results with MHV-68 suggest that the EBNA-1 glycine--alanine repeat is a key component of in vivo EBV persistence. Of course MHV-68 is not EBV, and it is possible that the expression of a strong MHC class I--binding peptide exaggerated somewhat the potential of ORF73-specific CTLs to control infection. However, a clear message is that *cis-*acting CTL evasion is an important feature of the gamma-herpesvirus lifecycle.
Latency-associated *trans*-acting CTL evasion comes into play during the MHV-68 growth program, when rapid cell division probably raises ORF73 production above a level that can be disguised by *cis-*acting evasion, and additional viral gene products are expressed. This *trans-*acting evasion allows latency amplification to progress despite evidence of a CTL response to at least one viral growth program antigen \[[@pbio-0030120-b32]\]. However, *trans*-acting evasion alone was insufficient for even an initial amplification of MHV-68 latency. SIINFEKL production from the EPI virus did not simply compromise K3 function, since a complete loss of K3 typically gives a 1-log reduction in infectious centres with relatively little effect on the viral genome load \[[@pbio-0030120-b31]\], whereas the EPI virus showed a 3- to 4-log infectious centre deficit and a severely reduced genome load. *cis*-acting immune evasion therefore operated in a distinct setting relatively early in latency establishment. Multiple patterns of both EBV \[[@pbio-0030120-b50],[@pbio-0030120-b51]\] and MHV-68 \[[@pbio-0030120-b14]\] latent gene expression occur in acutely infected lymphoid tissue. Notably, EBV implements EBNA-1-only latency even during acute infectious mononucleosis \[[@pbio-0030120-b35]\]. Thus, the extreme dependence of MHV-68 on *cis-*acting evasion probably reflects early B cell entry into "ORF73-only" latency.
The rather modest SIINFEKL-specific response to the EPI virus contrasted with the massive CTL activation stimulated by WT virus. For example, in one experiment WT virus progressed from 4.7 × 10^3^ infectious centres per spleen at day 10 of infection to 6.4 × 10^4^ at day 14 (means of five mice), while the percentage of CD8^+^ T cells expressing CD69 increased from 10.7% to 20.8%. Over the same time, the EPI virus infectious centres fell from 7.3 × 10^2^ to 4.6 × 10^1^, with 6.9% and 5.9%, respectively, of CD8^+^ T cells expressing CD69. Indeed the numerous lytic antigen-specific CTLs stimulated by EBV \[[@pbio-0030120-b52]\] and MHV-68 \[[@pbio-0030120-b43]\] infections imply an immune response failure, since latently infected B cells proliferate and progress to lytic gene expression without hindrance by latent antigen-specific CTLs. It is crucial in persistent viral infections for the immune system to attack appropriate targets. End-stage cells may stimulate large T cell responses, but the control of infection depends more on overcoming immune evasion. A major challenge in vaccinating against complex pathogens is to direct the immune system against the key, self-renewing population that maintains the parasite load.
Materials and Methods {#s4}
=====================
{#s4a}
### Mice {#s4a1}
C57BL/6J and BALB/c mice (Harlan Olac, Bicester, United Kingdom) were kept in Cambridge University animal facilities in accordance with United Kingdom Home Office guidelines (project licence 80/1579). Mice were infected intranasally with 2 × 10^4^ plaque-forming units (PFU) of MHV-68 under brief halothane anaesthesia. T cell subset depletion was by intravenous and then intraperitoneal injection of purified mAb YTS169 \[[@pbio-0030120-b53]\].
### Cell lines {#s4a2}
BHK-21 cells, MEF-1 cells, NIH-3T3-CRE cells \[[@pbio-0030120-b31]\], A20 cells, L929 cells transfected with H2-K^b^ \[[@pbio-0030120-b54]\], and the B3Z T cell hybridoma \[55 were all grown in DMEM, supplemented with 2 mM glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 10% fetal calf serum (complete DMEM). A20-syndecan-1 cells were made by retroviral transduction of A20 cells with a vector expressing the extracellular domain of murine syndecan-1 linked to the transmembrane and cytoplasmic domains of H2-D^b^, and will be described in detail elsewhere. MEFs were harvested at 13 d of gestation and were grown in complete DMEM with 50 μM 2-mercaptoethanol.
### Plasmids {#s4a3}
We amplified ORF73 by PCR (Hi-Fidelity PCR kit, Roche Diagnostics, Lewes, United Kingdom), including EcoRI and SalI restriction sites at its respective 5′ and 3′ ends, and cloned the product into the EcoRI and SalI sites of pSP73 (Promega, Chilworth, United Kingdom) to make pSP73-ORF73. To introduce the SIINFEKL epitope of OVA \[[@pbio-0030120-b56]\] near the 3′ end of ORF73, we digested pSP73-ORF73 with HinDIII and dephosphorylated it (P. borealis alkaline phosphatase, Roche Diagnostics). Two complementary oligonucleotides (5′- AGCTAGTATAATCAACTTTGAAAAACTGCT and 5′- AGCTAGCAGTTTTTCAAAGTTGATTATACT) (Sigma-Genosys, Cambridge, United Kingdom) were then heated, annealed, phosphorylated, and ligated into the HinDIII site (T4 DNA ligase, New England Biolabs, Hitchin, United Kingdom). Thus, amino acid residues 277--283 of ORF73 (QASGTQH) were changed to QASIINFEKLLASGTQH (ORF73-SC). Oligonucleotide insertion was confirmed by DNA sequencing. To insert the SIINFEKL coding sequence in the 5′ end of ORF73, we amplified ORF73 by PCR, pairing the 5′ primer , containing a 5′ EcoRI restriction site (underlined) upstream and a HinDIII site (double-underlined) downstream of the ORF73 start codon (bold), with a 3′ primer downstream of the ORF73 BstEII site, and containing an XhoI site. This PCR product was cloned into the EcoRI and XhoI sites of pSP73. The complementary oligonucleotides 5′- AGCTAGTATAATCAACTTTGAAAAACTGAC and 5′- AGCTGTCAGTTTTTCAAAGTTGATTATACT were then inserted into the HinDIII site, changing amino acid residues 1--6 of ORF73 from MPTSPP to MQASIINFEKLTASPP (ORF73-SN). The modified 5′ end of ORF73 was then subcloned as an EcoRI/BstEII fragment into pSP73-ORF73, thereby reconstituting the gene with its 3′ HinDIII site intact. Each form of ORF73 was then cloned into the pcDNA3 mammalian expression vector (Invitrogen, Carlsbad, California, United States). We made 3′ deletions of pcDNA3-ORF73-SN by digesting it with HinDIII or KpnI, each of which cuts within ORF73 and within the pcDNA3 polylinker 5′ of its EcoRI site. The N-terminal ORF73 fragment was then gel-purified and ligated into a new pcDNA3 vector. We generated a 3′ PstI deletion by digestion with PstI, gel purification, and ligation of the vector back to itself, since PstI cuts downstream of the pcDNA3 XhoI site.
We subcloned the N-terminal 325 amino acid residues of OVA as an EcoRI/XhoI fragment from pMSCV-OVA-IRES-GFP \[[@pbio-0030120-b57]\] into pSP73. Hybrids of 5′ ORF73 and 3′ OVA (which contains the SIINFEKL epitope), or 5′ OVA and 3′ ORF73 (with its SIINFEKL insert), were made by cutting each at a unique internal PstI site and swapping an in-frame 3′ PstI/XhoI fragment between them. Each form of ORF73/OVA was then subcloned as an EcoRI/XhoI fragment into the EcoRI and XhoI sites of pcDNA3. To fuse the ORF73 coding sequence to the C-terminus of OVA, we PCR-cloned the N-terminal 325 amino acid residues of OVA without a stop codon into the EcoRI and XhoI sites of pcDNA3 and ligated in PCR-cloned ORF73 as an XhoI/ApaI fragment, downstream of and in frame with the OVA coding sequence. 3′ HinDIII and KpnI truncations of this construct were generated as above. We also generated 5′ ORF73 truncations as XhoI/ApaI-digested PCR products, starting at amino acid residue 126 (OVA-73E), 146 (OVA-73D), 166 (OVA-73C), 186 (OVA-73B), or 206 (OVA-73A), and fused these to OVA 1--325 in pcDNA3. We also cloned OVA lacking its signal sequence, with translation starting at its methionine residue 41 (SOVA), and made the same fusions with N-terminal ORF73 truncations A--E.
### Recombinant viruses {#s4a4}
The MHV-68 M11 and ORF73 coding sequences (genomic co-ordinates 103418--103933 and 104868--103924, respectively) overlap by 10 bp at their 3′ ends \[[@pbio-0030120-b11]\]. We therefore duplicated this overlap to generate an insertion site between them. Thus, we PCR-cloned genomic co-ordinates 131171--103933, including XhoI and SalI restriction sites at the respective 5′ and 3′ ends, and cloned the fragment into the XhoI and SalI sites of pSP73-ORF73 (see Plasmids, above) to make pSP73-ORF73-M11, with the M11 and ORF73 coding sequences each complete and separated by a SalI site. The M11/ORF73 genomic overlap (the stop codons of ORF73 on the noncoding strand and M11 on the coding strand are underlined) was thus changed to TTTATGTC GTCGACTTATGTCTGAG. We then generated a poly-epitope construct downstream of an encephalomyocarditis IRES for insertion into the SalI site. We first inserted the adenovirus E19K leader sequence as two complementary oligonucleotides (5′- AATTGACCACCATGAGGTACATGATTTTAGGCTTGCTCGCCCTTGCGGAGTCTGCAGCGCGAATTCAGATCTCTCGAGTGAT and 5′- TCGAATCACTCGAGAGATCTGAATTCGCGCTGCAGACTGCCGCAAGGGCGAGCAAGCCTAAAATCATGTACCTCATGGTGGTC) into the EcoRI and XhoI sites of pMSCV-IRES-NEO \[[@pbio-0030120-b57]\]. Two complementary oligonucleotides encoding the peptide sequence MTSINFVKIASNENMETMSIINFEKL (5′- AATTCCTACCACCATGACCAGTATCAACTTTGTGAAGATAGCTTCCAATGAAAACATGGAGACTATGAGTATAATCAACTTTGAAAAACTGTGAC and 5′- TCGAGTCACAGTTTTTCAAAGTTGATTATACTCATAGTCTCCATGTTTTCATTGGAAGCTATCTTCACAAAGTTGATACTGGTCATGGTGGTAG) were then inserted into the EcoRI and XhoI sites of the pMSCV-NEO-leader construct. TSINFVKI is an H2-K^b^-restricted epitope from the MHV-68 ORF61 \[[@pbio-0030120-b43]\]; ASNENMETM is an H2-D^b^-restricted epitope from the influenza A/PR/8/34 nucleoprotein \[[@pbio-0030120-b58]\]. We amplified the leader-epitope construct by PCR, using the primers 5′- CCCCCATGGCCAGGTACATGATTTTAGGCTTGCTC and 5′- CCCGTCGACTCACAGTTTTTCAAAGTTGATTATACT, thereby adding a 5′ NcoI site and replacing the 3′ XhoI site with a 3′ SalI site, and cloned it into the NcoI and SalI sites of pMSCV-IRES-GFP \[[@pbio-0030120-b59]\]. Thus, the GFP coding sequence downstream of the IRES was replaced by the leader-epitope construct. We used the 3′ SalI site and an XhoI site just 5′ of the IRES to excise an IRES-leader-epitope XhoI/SalI fragment and cloned it into the SalI site of pSP73-ORF73-M11. We also subcloned an XhoI/SalI IRES-GFP fragment from pMSCV-IRES-GFP into the SalI site of pSP73-ORF73-M11 to make a control virus. Each ORF73-IRES construct was then subcloned into a larger genomic fragment for recombination into the MHV-68 genome. To do this, we used a BamHI-G genomic fragment \[[@pbio-0030120-b10]\] (genomic co-ordinates 101653--106902), cloned into pACYC184 (New England Biolabs) lacking a BspHI site \[[@pbio-0030120-b60]\]. The ORF73-IRES constructs and pACYC184-BamHI-G were digested with BstEII (genomic co-ordinate 104379) and BspHI (genomic co-ordinate 103750). Because BspHI is blocked by methylation, we used plasmids derived from Dam^−^ *E. coli.* Finally, the mutant BamHI-G fragments were subcloned into the BAC mutagenesis shuttle vector pST76K-SR. Rec A--mediated recombination into the MHV-68 BAC was then carried out as previously described \[[@pbio-0030120-b41]\]. Sequence analysis revealed that the E19K leader sequence had been mutated during cloning, destroying the NcoI site and changing the start of the coding sequence downstream of the IRES from MARYMILG to MILG. Since this change was unlikely to prevent epitope presentation, no attempt was made to correct it. The ORF73-IRES-epitope BAC was subsequently reverted using an unmutated BamHI-G clone in pST76K-SR. MHV-OVA was generated by cloning OVA cDNA into EcoRI and XhoI sites of an ORF57/ORF58 intergenic expression cassette, driven by an ecotopic MHV-68 M3 promoter \[[@pbio-0030120-b61]\]. This virus will be described in more detail elsewhere. All BACs were reconstituted into infectious virus by transfecting 5 μg of BAC DNA into BHK-21 cells with Fugene-6 (Roche Diagnostics). The BAC cassette was removed by serial viral passage through NIH-3T3-CRE cells. Virus stocks were grown and titred on BHK-21 cells.
### Virus assays {#s4a5}
Infectious virus in freeze-thawed lung and spleen homogenates was plaque assayed on MEFs. Latent plus pre-formed virus in spleens was assayed on MEFs by explant culture of single-cell suspensions \[[@pbio-0030120-b39]\]. Cells expressing viral tRNAs 1--4 were detected by in situ hybridization of formalin-fixed, paraffin-embedded spleen cell sections, using a digoxigenin-labelled riboprobe transcribed from pEH1.4 \[[@pbio-0030120-b44]\]. Hybridized probe was detected with alkaline phosphatase-conjugated anti-digoxigenin Fab fragments (Boehringer Ingelheim, Bracknell, United Kingdom) according to the manufacturer\'s instructions. The viral genome load in individual spleens was measured by real-time PCR. DNA was extracted (Wizard genomic DNA purification kit, Promega) and a portion of the MK3 ORF (genomic co-ordinates 24832--25071) amplified by PCR from 10 ng of each sample (Rotor Gene 3000, Corbett Research, Cambridge, United Kingdom). PCR products were quantitated with Sybr green (Invitrogen) and compared with a standard curve of cloned MK3 template, serially diluted in uninfected cellular DNA and amplified in parallel. The MK3 copy number was calculated from the cycle number at which the Sybr green signal crossed a set threshold on the standard curve. Amplified products were distinguished from paired primers by melting curve analysis, and the correct size of the amplified products was confirmed by electrophoresis and staining with ethidium bromide.
### Southern blotting {#s4a6}
Viral DNA was isolated from infected BHK-21 cells by alkaline lysis \[[@pbio-0030120-b39]\], digested with NcoI, electrophoresed on a 0.8% agarose gel, and transferred to positively charged nylon membranes (Roche Diagnostics). A ^32^P-dCTP-labelled probe (APBiotech, Amersham, United Kingdom) was generated from the BamHI-G genomic fragment by random primer extension (Nonaprimer kit, Qbiogene, Bingham, United Kingdom) according to the manufacturer\'s instructions. Membranes were hybridized with probe (65 °C, 18 h), washed to a stringency of 0.2× SSC with 0.1% SDS, and exposed to X-ray film.
### Metabolic labelling and immunoprecipitation {#s4a7}
Cells were metabolically pulse-labelled with ^35^S-cysteine/methionine (APBiotech) and chased with 1 mM unlabelled cysteine and methionine \[[@pbio-0030120-b29]\]. Labelled cells were lysed on ice for 30 min in 50 mM Tris-Cl (pH 7.4), 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 1 mM PMSF, plus Complete protease inhibitors (Roche Diagnostics). Cell debris and nuclei were removed by centrifugation (13,000 × *g,* 15 min). Lysates were precleared with rabbit anti-actin whole serum and formalin-fixed S. aureus (Sigma Chemical, Poole, United Kingdom), and then again with protein A-sepharose. OVA was precipitated with rabbit anti-OVA serum (Abcam, Cambridge, United Kingdom) followed by protein A-sepharose. Beads were washed five times in 1% Triton X-100 buffer. Samples were dissociated (95 °C, 2 min) in Laemmli\'s buffer prior to SDS-PAGE. Gels were fixed, dried, and exposed to X-ray film.
### Immunoblotting {#s4a8}
Cells were lysed as for immune precipitations (above). Post-nuclear lysates were denatured (95 °C, 2 min) in Laemmli\'s buffer, separated by SDS-PAGE, and transferred to polyvinylidene difluoride membranes. Membranes were blocked in PBS/0.1% Tween-20/10% non-milk fat and probed with rabbit anti-OVA serum or rabbit anti-neomycin phosphotransferase II serum (Upstate, Milton Keynes, United Kingdom), followed by horseradish-peroxidase-coupled donkey anti-rabbit IgG pAb (APBiotech) and ECL substrate development.
### Antigen presentation assays {#s4a9}
L929-K^b^ cells (1--2 × 10^5^/well in 24-well plates) were transfected with 1 μg of plasmid using Fugene-6. Forty-eight hours later, B3Z cells (5 × 10^5^) were added to each well. B3Z is an H2-K^b^-restricted, SIINFEKL-specific T cell hybridoma that produces beta-galactosidase in response to T cell receptor ligation \[[@pbio-0030120-b55]\]. For virus infections, L929-K^b^ cells or MEF-1 cells (5 × 10^5^/well) were infected for 2 h and washed once before adding B3Z T cell hybridoma cells (5 × 10^5^/well). After a further 18 h, the cells were washed once in PBS and lysed in PBS/5 mM MgCl~2~/1% NP-40/0.15 μM chlorophenol-red-beta-[D]{.smallcaps}-galactoside (Merck Biosciences, Nottingham, United Kingdom) to assay beta-galactosidase activity. After 2--4 h at 37 °C, the absorbance at 595 nm was read on a Bio-Rad (Hercules, California, United States) Benchmark microplate reader.
### ELISA and ELISPOT assays {#s4a10}
For IFN-γ ELISPOT assays \[[@pbio-0030120-b62]\], duplicate dilutions of effector cells were incubated with 3 × 10^5^ naive irradiated syngeneic spleen cells in nitrocellulose-bottomed 96-well plates (Millipore Corporation, Bedford, Massachusetts, United States) coated with rat anti-mouse IFN-γ mAb (BD-Pharmingen, San Diego, California, United States). The naive spleen cells were either (1) untreated, (2) pulsed with 1 μM AGPHNDMEI (p56), 1 μM TSINFVKI (p79), 1 μM ASNENMETM, or 1 μM SIINFEKL peptide, or (3) infected with WT MHV-68 (2 PFU/cell). After 48 h culture at 37 °C in complete RPMI/50 μM 2-mercaptoethanol/10 U/ml human recombinant IL-2, captured IFN-γ was detected with a further, biotinylated rat anti-mouse IFN-γ mAb (BD-Pharmingen), followed by streptavidin-alkaline phosphatase (Dako Cytomation, Ely, United Kingdom) and 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium substrate.
Virus-specific and total serum IgG levels were measured by ELISA. Maxisorp ELISA plates (Nalge Nunc, Rochester, New York, United States) were coated overnight with either affinity-purified goat anti-mouse IgG sera (Sigma Chemical) or 0.05% Triton X-100--disrupted MHV-68 \[[@pbio-0030120-b63]\]. After incubation with 2-fold serum dilutions, bound murine IgG was detected with alkaline phosphatase-conjugated goat anti-mouse IgG--Fcγ serum and nitrophenylphosphate substrate (Sigma Chemical). Absorbance was read at 405 nm.
### Flow cytometry {#s4a11}
A20-syndecan-1 cells infected with GFP^+^ viruses were trysinized, washed in PBS, and analysed directly for green channel fluorescence. Spleens were disrupted into single-cell suspensions, washed in PBS/0.1% BSA/0.01% azide, and incubated for 15 min on ice with 5% mouse serum/5% rat serum and anti-CD16/32 mAb. Specific staining (1 h, 4 °C) was with fluorescein-isothiocyanate-coupled anti-CD69 and phycoerythrin-coupled anti-CD19 (BD-Pharmingen), or tricolour-coupled anti-CD8 (Caltag Laboratories, Burlingame, California, United States), Fluorescein-isothiocyanate-coupled anti--T cell receptor Vbeta4, and phycoerythrin-coupled anti-CD62L (BD-Pharmingen). For intracellular cytokine staining, spleen cells (5 × 10^5^--1 × 10^6^ in 200 μl of complete RPMI/50 μM 2-mercaptoethanol/10 U/ml human recombinant IL-2/10 μg/ml Brefeldin A) were stimulated (5 h, 37 °C) with 1 μM ASNENMETM, 1 μM AGPHNDMEI, 1 μM TSINFVKI, or 1 μM SIINFEKL peptides, or left without peptide. All cells were then washed in PBS/10 μg/ml Brefeldin A, blocked with anti-CD16/32, stained with tricolour-conjugated anti-CD8 plus fluorescein-isothiocyanate-conjugated anti-I-A^b^ (1 h, 4 °C), washed twice, fixed in 2% paraformaldehyde (30 min, 4 °C), washed once, permeabilized with 0.5% saponin, washed once, stained with phycoerythrin-coupled anti-interferon-gamma (BD-Pharmingen), and washed twice. All cells were analysed on a FACS Calibur using Cellquest software (Becton-Dickinson, Oxford, United Kingdom). I-A^b^ staining was used to exclude B cells and myeloid cells. Data were graphed with FCSPress v1.3 ([www.fcspress.com](www.fcspress.com)).
Jenny Phillips kindly provided mAb YTS169. We thank Stacey Efstathiou and Gabrielle Belz for helpful discussions. This work was supported by the United Kingdom Medical Research Council (grants G9800943 and G9901295). NJB is supported by a Wellcome Trust studentship. PGS is an Academy of Medical Sciences/Medical Research Council clinician scientist (G108/462).
**Competing interests.** The authors have declared that no competing interests exist.
**Author contributions.** PGS conceived and designed the experiments. NJB, JSM, and PGS performed the experiments. PGS analysed the data and wrote the paper.
Citation: Bennett NJ, May JS, Stevenson PG (2005) Gamma-herpesvirus latency requires T cell evasion during episome maintenance. PLoS Biol 3(4): e120.
CTL
: CD8^+^ T lymphocyte
EBV
: Epstein-Barr virus
EPI
: ORF73-IRES-epitope
GFP
: green fluorescent protein
IRES
: internal ribosome entry site
MEF
: murine embryonic fibroblast
MHC
: major histocompatibility complex
MHV-68
: murine gamma-herpesvirus-68
OVA
: ovalbumin
PFU
: plaque-forming units
SD
: standard deviation
WT
: wild-type
|
PubMed Central
|
2024-06-05T03:55:54.906532
|
2005-3-22
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065266/",
"journal": "PLoS Biol. 2005 Apr 22; 3(4):e120",
"authors": [
{
"first": "Neil J",
"last": "Bennett"
},
{
"first": "Janet S",
"last": "May"
},
{
"first": "Philip G",
"last": "Stevenson"
}
]
}
|
PMC1065267
|
Introduction {#s1}
============
The identification of tumor-specific molecular markers is a powerful tool in cancer diagnostics, and the targeting of tumor-specific pathways is the best hope for developing nontoxic and efficient anticancer therapies. Targeting of cancer cells relies on the development of molecular beacons, suited for in vivo applications, that are endowed with the required affinity, specificity, and favorable pharmacokinetic properties.
With the systematic evolution of ligands by exponential enrichment (SELEX) technology \[[@pbio-0030123-b01],[@pbio-0030123-b02]\], specific macromolecular ligands---aptamers---can be generated by screening very large pools of oligonucleotides containing regions of random base composition with reiterated cycles of enrichment and amplification. At each cycle, the individual oligonucleotides with affinity for the desired target are kept, those with affinity for the sham target are rejected, and the population is enriched in oligonucleotides that distinguish between sham and real target. Aptamers that recognize a wide variety of targets, from small molecules to proteins and nucleic acids, and from cultured cells to whole organisms, have been described \[[@pbio-0030123-b03],[@pbio-0030123-b04],[@pbio-0030123-b05],[@pbio-0030123-b06],[@pbio-0030123-b07],[@pbio-0030123-b08],[@pbio-0030123-b09],[@pbio-0030123-b10]\]. These oligonucleotides generally meet the requirements for in vivo diagnostic and/or therapeutic applications: Besides their good specificity and affinity, they are poorly immunogenic, and the SELEX technology can now accept chemically modified nucleotides for improved stability in biological fluids \[[@pbio-0030123-b11]\]. Conspicuously, less than fifteen years after the first applications of the technique, several lead compounds, including an anti-vascular endothelial growth factor aptamer \[[@pbio-0030123-b12]\], are currently under clinical trials \[[@pbio-0030123-b13]\].
Receptor tyrosine kinases (RTKs) are involved in a variety of signaling processes that regulate cell growth and proliferation and in several cancers \[[@pbio-0030123-b14]\]. RTKs are privileged targets for cancer therapy, which is underscored by the promising outcome of clinical trials with small molecules or antibody inhibitors \[[@pbio-0030123-b14]\]. In the present study, we validated a general strategy to target transmembrane receptors by SELEX. The RET (rearranged during transfection) RTK is physiologically stimulated by any member of the glial cell line-derived neurotrophic factor (GDNF) family \[[@pbio-0030123-b15],[@pbio-0030123-b16]\]. Germline mutations in the *RET* gene are responsible for constitutive activation of the receptor and for inheritance of multiple endocrine neoplasia (MEN) type 2A and 2B syndromes and of familial medullary thyroid carcinoma \[[@pbio-0030123-b17],[@pbio-0030123-b18],[@pbio-0030123-b19],[@pbio-0030123-b20]\].
Mutations in the extracellular domain of RET, responsible for MEN2A syndrome, lead to constitutive dimerization of two mutated RET molecules. Conversely, a single point mutation, within the RET catalytic domain, that causes the MEN2B syndrome, involves an intramolecular mechanism to convert *RET* into a dominant transforming gene. Therefore, RET constitutes a model system of choice \[[@pbio-0030123-b20]\], in that the transforming mutations located in the extracellular domain simplify the issue of intracellular accessibility for a molecule targeting the receptor mutated in the extracellular domain (in its monomeric or dimeric form) and might provide alternative models (e.g., RET with mutations of the 2B kind) for controls or to elucidate the mode of target recognition.
Here we adopted a whole-cell SELEX strategy to target RET in a complex environment that is expected to expose a native protein to the selection procedure, thus best mimicking in vivo conditions. We obtained aptamers that not only recognize the extracellular domain of RET, but also block RET downstream signaling and subsequent molecular and cellular events. The fact that aptamers with antioncogenic activity were isolated in the absence of a specific selective pressure suggests that our method could be used to identify active macromolecules with potential therapeutic interest against other transmembrane receptors.
Results {#s2}
=======
A library of 2′-fluoropyrimidine (2′F-Py), nuclease-resistant RNAs was subjected to a differential SELEX protocol against intact cells expressing different forms of the human *RET* oncogene ([Figure 1](#pbio-0030123-g001){ref-type="fig"}). For the selection step, PC12 cells were used that express the human RET^C634Y^ mutant receptor (PC12/MEN2A). RET^C634Y^ is mutated in the extracellular domain and forms spontaneously active homodimers on the cell surface, which induces biochemical and morphological changes that mirror the RET-dependent human pheochromocytoma phenotype of MEN2 syndromes \[[@pbio-0030123-b21]\]. The counterselection necessary to avoid selecting for aptamers that nonspecifically recognized the cell surface included a first step against parental PC12 cells in order to eliminate nonspecific binders of the PC12 cell surface, followed by a second counterselection step against PC12/MEN2B cells that expressed an allele of RET (RET^M918T^) mutated in the intracellular tyrosine kinase domain. PC12/MEN2B and PC12/MEN2A cells have a similar morphology, but the extracellular domain of the RET^M918T^ receptor is identical to the wild type and, in the absence of the ligand and co-receptor, remains monomeric. This step was originally aimed at selecting aptamers that recognize specifically the dimeric form of the extracellular domain.
::: {#pbio-0030123-g001 .fig}
10.1371/journal.pbio.0030123.g001
Figure 1
::: {.caption}
###### Schematic Protocol for the Selection of PC12/MEN2A Cell-Specific Aptamers
A pool of 2′F-Py RNAs was incubated with suspended parental PC12 cells (Counterselection 1). Unbound sequences were recovered by centrifugation and incubated with adherent PC12/MEN2B cells (Counterselection 2). Unbound sequences in the supernatant were recovered and incubated with adherent PC12/MEN2A cells for the selection step (Selection). Unbound sequences were discarded by several washings, and bound sequences were recovered by phenol extraction. Sequences enriched by the selection step were amplified by RT-PCR and in vitro transcription before a new cycle of selection.
:::
:::
After 15 rounds of selection, the pool of remaining sequences bound PC12/MEN 2A cells in a saturable manner with an apparent Kd approximating 100 nM. From this pool, 67 sequences were cloned and analyzed. Two individual sequences (D14 and D12) dominated the selection and constituted together more than 50% of the clones, four other sequences represented together 25% of the clones, and eight sequences were present only once. As is often the case for a selection against a complex target \[[@pbio-0030123-b07],[@pbio-0030123-b22]\] (and in contrast to in vitro SELEX on purified proteins) we found almost no similarity among sequences, except for clones D24 and D4, which shared common sequence motifs and structure prediction ([Figure 2](#pbio-0030123-g002){ref-type="fig"}A).
::: {#pbio-0030123-g002 .fig}
10.1371/journal.pbio.0030123.g002
Figure 2
::: {.caption}
###### Predicted Structure and Association Constants of D4 and D24
\(A) Comparison of a secondary structure prediction for the D4 and D24 aptamers**.** Structures were predicted using MFOLD software version 3.1 (available at <http://www.bioinfo.rpi.edu/applications/mfold/>).
\(B) Binding curve of the D4 aptamer on PC12/MEN2A. D4 was ^32^P-radiolabeled and incubated at different concentrations on cell monolayers. The background binding value for a D4 scrambled sequence is subtracted from every data point. Scatchard analysis (inset) was used for the evaluation of the binding constant.
\(C) Binding of the ^32^P-labeled D4 aptamer to several cell lines expressing (or not) human RET. Binding was performed on the cell lines indicated in the same condition at 50 nM, and the results are expressed relative to the background binding detected with the starting pool of sequences used for selection. Expression of RET could not be detected by Western blot in HeLa, NBTII, PC12wt and NIH3T3 cells, whereas PC12/MEN2A and NIH3T3/MEN2A express RET^C634Y^ and PC12/MEN2B and NIH3T3/MEN2B express RET^M918T^.
:::
:::
We assessed binding to PC12/MEN2A cells of all individual aptamers that were found more than once and also of some unique sequences (including D4 and D24). Several sequences bound PC12/MEN2A cells with apparent Kd values ranging from 30 to 70 nM ([Figure 2](#pbio-0030123-g002){ref-type="fig"}B and unpublished data), but not parental PC12, rat-derived bladder carcinoma (NBTII), or human cervical carcinoma (HeLa) cells ([Figure 2](#pbio-0030123-g002){ref-type="fig"}C and unpublished data). As a first attempt to deconvolute the complex pool of winning aptamers, we first produced a recombinant fragment of RET, EC-RET^C634Y^ \[[@pbio-0030123-b23]\], but all attempts to identify in the winning pool aptamers binding to EC-RET^C634Y^ were fruitless. Likewise, SELEX against this purified EC-RET^C634Y^ protein gave rise to aptamers unable to recognize the PC12/MEN2A cells, suggesting that they did not bind to the RET protein present in its native conformation on the cell surface. Consequently, we screened the winning pool of aptamers for the ability to interfere with the biological activity of RET. To this end, we used an in vitro cell system in which we assessed the capability of each aptamer to inhibit RET^C634Y^ autophosphorylation and receptor-dependent downstream signaling. Mutant RET^C634Y^, expressed in PC12/MEN2A cells, forms homodimers on the cell surface that cause constitutive activation of its tyrosine kinase activity \[[@pbio-0030123-b24]\] and induce several downstream signaling cascades, including the activation of extracellular signal-regulated protein kinase (ERK) \[[@pbio-0030123-b25]\]. As previously reported \[[@pbio-0030123-b25]\], levels of phosphorylated RET and ERK were constitutively high in untreated PC12/MEN2A cells due to the presence of the active RET^C634Y^ allele. Surprisingly, some of the tested aptamers inhibited RET^C634Y^ and ERK phosphorylation, compared to the control starting pool and to the other aptamers ([Figure 3](#pbio-0030123-g003){ref-type="fig"}A and unpublished data). In all experiments, inhibition of phosphorylation was more rapid and quantitative for ERK than for RET^C634Y^. We believe that this is due to a different sensitivity to changes in RET tyrosine kinase activity of the two processes and/or to differences in the half-lives of the phosphorylated forms of the two proteins \[[@pbio-0030123-b26]\]. In a dose-response experiment ([Figure 3](#pbio-0030123-g003){ref-type="fig"}B, left panel), the best inhibitor, D4, was effective at a concentration of 200 nM to inhibit RET^C634Y^ autophosphorylation up to 70% and to drastically reduce ERK phosphorylation. Time-activity studies showed that the treatment of PC12/MEN2A cells at 200 nM for 1 h was sufficient to significantly inhibit RET^C634Y^ autophosphorylation and to drastically reduce ERK phosphorylation ([Figure 3](#pbio-0030123-g003){ref-type="fig"}B, right panel).
::: {#pbio-0030123-g003 .fig}
10.1371/journal.pbio.0030123.g003
Figure 3
::: {.caption}
###### Effect of Selected Aptamers on RET^C634Y^ Activity
\(A) PC12/MEN2A cells were either left untreated or treated for 16 h with 150 nM of the indicated RNA aptamer, or the starting RNA pool (pool). Cell lysates were immunoblotted with anti-(phospho)-ERK (pErk), then stripped and reprobed with anti-ERK (Erk) to confirm equal loading. Values below the blots indicate signal levels relative to untreated controls.
\(B) PC12**/**MEN2A cells were treated for 1 h with increasing amounts of D4 (left blots) or with 200 nM D4 for the indicated incubation times (right blots). Cell lysates were immunoblotted with anti-(Tyr-phosphorylated)-RET (pRet) or anti-(phospho)-ERK (pErk) antibodies, as indicated. To confirm equal loading the filters were stripped and reprobed with anti-RET (Ret) or anti-ERK (Erk) antibodies, respectively.
In (A) and (B), "C" indicates mock-treated cells. Quantitations were done on the sum of the two RET- or ERK- specific bands, and values are expressed relative to the control, arbitrarily set to 1. Standard deviations are indicated (*n* = 4).
:::
:::
Comparison of the predicted structures of D4 and of the related clone D24 ([Figure 2](#pbio-0030123-g002){ref-type="fig"}A) suggests that a conserved stem-internal loop-stem is crucial for binding. Consistently, we found that replacing the apical loop with a stable tetraloop (UUGC) or deleting nucleotides not included in the conserved structure did not significantly affect binding of D4 to PC12/MEN2A cells (unpublished data). However, only the full-length D4 inhibits RET^C634Y^ signaling, demonstrating that binding is necessary but not sufficient for inhibition. A 2′F-Py RNA oligonucleotide of identical composition but with a scrambled sequence (D4Sc) was ineffective for both binding and inhibition.
The D4 aptamer bound to PC12/MEN2A with an estimated apparent Kd of 35 ± 3 nM ([Figure 2](#pbio-0030123-g002){ref-type="fig"}B), but also to PC12/MEN2B cells ([Figure 2](#pbio-0030123-g002){ref-type="fig"}C and unpublished data), suggesting that one of the counterselection steps employed in the SELEX procedure was ineffective in this case. The D4 aptamer bound to transfected NIH3T3 cells expressing at similar levels the two mutant forms (RET^C634Y^ and RET^M918T^) of the RET receptor (NIH/MEN2A and NIH/MEN2B, respectively \[[Figure 2](#pbio-0030123-g002){ref-type="fig"}C; see also below\]). Binding was dependent on expression of human RET, as D4 did not recognize parental untransfected PC12, NIH3T3 cells, or other cell lines, including rat NBTII, human HeLa cells, and mouse MN1 ([Figure 2](#pbio-0030123-g002){ref-type="fig"}C and unpublished data). Interestingly, the latter, a mouse motor neuron-neuroblastoma fusion cell line, expresses the mouse RET^wt^, suggesting some species-specificity in RET recognition by D4. Finally, D4 bound a human neuroblastoma cell line (SK-N-BE) that naturally expresses endogenous RET (L. Cerchia et al., personal communication). Consistently with what was observed for the pool of winning aptamers, D4 was unable to bind the purified EC-RET^C634Y^ protein (unpublished data), thus supporting the specificity for the membrane-bound RET.
We next determined whether D4 could inhibit wild-type RET. Cells from a PC12-derived cell line expressing the human wild-type RET (PC12/wt) were stimulated with a mixture containing GDNF and soluble GDNF family receptor α1 (GFRα1), and either treated with the D4 aptamer or with the starting pool of 2′F-Py RNA as a negative control. As shown in [Figure 4](#pbio-0030123-g004){ref-type="fig"}A, the D4 aptamer, but not the control RNA pool, strongly inhibited GDNF-induced phosphorylation of RET (left panel) and of the downstream effector ERK (middle panel). A similar inhibitory effect was observed in PC12-α1/wt cells, a PC12-derived cell line that stably expresses both human RET and GFRα1 (unpublished data). In contrast, D4 was inactive in inhibiting the signaling triggered by the unrelated nerve growth factor (NGF) receptor tyrosine kinase TrkA, thus indicating that D4-induced inhibition of ERK phosphorylation was specific for RET intracellular signaling ([Figure 4](#pbio-0030123-g004){ref-type="fig"}A, right pane)
::: {#pbio-0030123-g004 .fig}
10.1371/journal.pbio.0030123.g004
Figure 4
::: {.caption}
###### D4 Aptamer Inhibits RET^wt^ but Not RET^M918T^Activity
\(A) PC12/wt cells were treated for 10 min with GDNF (50 ng/ml) and soluble GFRα1 (1.6 nM), or 5 min with NGF (100 ng/ml), together with 200 nM of either the D4 aptamer or the starting RNA pool. "C\*" indicates cells treated with GDNF and GFRα1 in the absence of aptamer.
\(B) PC12/MEN2B cells were starved for 6 h and then treated for 1 h with 200 nM D4 or the starting RNA pool. Cell lysates were immunoblotted with anti-(Tyr-phosphorylated)-RET or anti-(phospho)-ERK antibodies, as indicated (see [Figure 3](#pbio-0030123-g003){ref-type="fig"} legend).
In (A) and (B), "C" indicates mock-treated cells. Quantitations were done as in [Figure 3](#pbio-0030123-g003){ref-type="fig"}, and relative abundances are expressed relative to controls, arbitrarily set to 1. Standard deviations are indicated (*n* = 4).
:::
:::
Although the D4 aptamer binds PC12/MEN2B cells, treating these cells with 200 nM D4 for 1 h ([Figure 4](#pbio-0030123-g004){ref-type="fig"}B) or longer, or at higher D4 concentrations (unpublished data), did not interfere with signaling due to the monomeric RET^M918T^. This further confirms that inhibition of ERK phosphorylation is not a nonspecific effect of exposing the cells to the D4 aptamer. The kinase and the biological activities of RET^M918T^, although constitutive, are responsive to GDNF stimulation in the presence of GFRα1 \[[@pbio-0030123-b27],[@pbio-0030123-b28]\]. Similarly to the inhibition of RET^wt^ activity, the treatment of PC12/MEN2B cells by D4 abolished the GDNF-dependent overstimulation of RET and ERK phosphorylation (unpublished data). These data strongly suggest that D4 inhibits exclusively the dimerization-dependent RET activation.
We then searched for phenotypic effects of D4 on RET-dependent cell differentiation and transformation. First we measured neurite outgrowth in PC12-α1/wt cells following GDNF stimulation. As shown in [Figure 5](#pbio-0030123-g005){ref-type="fig"}, cells extended long neurite-like processes in response to a 48-h exposure to GDNF ([Figure 5](#pbio-0030123-g005){ref-type="fig"}B) with respect to the nonstimulated control cells ([Figure 5](#pbio-0030123-g005){ref-type="fig"}A). Treatment of the cells with the D4 aptamer ([Figure 5](#pbio-0030123-g005){ref-type="fig"}C), but not with the D4Sc scrambled control ([Figure 5](#pbio-0030123-g005){ref-type="fig"}D), significantly decreased the proportion of neurite outgrowth ([Figure 5](#pbio-0030123-g005){ref-type="fig"}E). To biochemically monitor differentiation, we determined the levels of the nerve growth factor-inducible protein (VGF) in cell extracts following 48 h of treatment. *VGF* is an early gene that is rapidly induced by both NGF and GDNF in PC12 cells \[[@pbio-0030123-b29]\]. As expected, in GDNF-treated cells, VGF expression was stimulated and, consistent with the phenotypic effects reported above, treatment with D4, but not with D4Sc, kept the VGF levels close to basal ([Figure 5](#pbio-0030123-g005){ref-type="fig"}F).
::: {#pbio-0030123-g005 .fig}
10.1371/journal.pbio.0030123.g005
Figure 5
::: {.caption}
###### D4 Aptamer Inhibits the GDNF-Induced Differentiation of PC12-α1/wt Cells
Cells were either left unstimulated (A), stimulated with GDNF (B), or with GDNF together with D4 or D4Sc (C and D, respectively). Following 48 h of GDNF treatment, the percentage of neurite outgrowth was calculated. The data represent the average of three independent experiments and are expressed as percentage of neurite-bearing cells/total cells analyzed (E). Following 48 h of treatment, cells were lysed and proteins immunoblotted with anti-VGF antibodies. Equal loading was confirmed by immunoblotting with anti-ERK antibodies as indicated (F).
:::
:::
Upon expression of either RET^C634Y^ or RET^M918T^, NIH3T3 cells show drastic changes in their morphology \[[@pbio-0030123-b24]\]. We treated NIH/MEN2A and NIH/MEN2B cells stably expressing the RET mutants with D4 for 72 h, and analyzed the morphological changes induced by the aptamer. As shown in [Figure 6](#pbio-0030123-g006){ref-type="fig"}, NIH/MEN2A and NIH/MEN2B cells have a spindle shape, long protrusions, and a highly refractive appearance ([Figure 6](#pbio-0030123-g006){ref-type="fig"}B and [6](#pbio-0030123-g006){ref-type="fig"}E, respectively). As expected, D4-treated NIH/MEN2A cells ([Figure 6](#pbio-0030123-g006){ref-type="fig"}C) reverted to a flat and polygonal morphology similar to the parental NIH3T3, whereas no morphological changes were observed in NIH/MEN2B ([Figure 6](#pbio-0030123-g006){ref-type="fig"}F), which is consistent with the notion that constitutive signaling from RET^C634Y^, but not from RET^M918^, is inhibited by D4. On the other hand, treatment with D4Sc had no effects on any cell line ([Figure 6](#pbio-0030123-g006){ref-type="fig"}D and unpublished data).
::: {#pbio-0030123-g006 .fig}
10.1371/journal.pbio.0030123.g006
Figure 6
::: {.caption}
###### D4 Aptamer Reverts the Transformed Morphology of NIH/MEN2A Cells
NIH3T3-derived cell lines were either left untreated (A, B, and E) or treated with D4 (C and F) or D4Sc (D), and the cells were maintained in culture for 72 h. Each experiment was repeated a minimum of three times.
:::
:::
Discussion {#s3}
==========
RTKs are involved in a variety of signaling pathways that affect cell growth and differentiation. Targeting specifically RTKs holds potential for dissecting the molecular mechanisms of receptor function, but also for diagnosis and therapeutics of cancer \[[@pbio-0030123-b14]\].
Here we employed a modified SELEX procedure to target the RET RTK, and we obtained nuclease-resistant RNA ligands capable of binding and inhibiting the protein on the cell surface. Aptamers against recombinant heregulin 3 (HER3) RTK have been recently isolated and shown to inhibit the heregulin-induced activation of the HER3/HER2 dimer \[[@pbio-0030123-b30]\]. However, finding the most efficient binders and inhibitors is likely to generally rely on the recognition of the target protein in its native state.
In the case of transmembrane receptors, whole-cell SELEX offers the advantage of selecting molecules capable of recognizing the target protein in its natural glycosylation state and presented in its physiological environment. An important drawback of this strategy is the lack of knowledge of the identity and abundance of the effective targets and the possibility that unwanted aptamers may dominate the selection, preventing the emergence of the molecules of interest. However, the abundance of the target protein and an appropriate selection scheme might provide sufficient selective pressure to favor the wanted aptamers \[[@pbio-0030123-b10]\].
The D4 aptamer binds to different cell types, provided that human RET is expressed on the cell surface, and specifically inhibits both RET and ERK phosphorylation, strongly suggesting that RET is the bona fide target of D4. Interestingly, aptamers isolated by whole-cell SELEX were unable to bind purified EC-RET^C634Y^ and, conversely, aptamers coming from the selection with purified EC-RET^C634Y^ were unable to bind the membrane-bound RET. Thus, it is likely that D4 binding is dependent on the association of RET with the cellular membrane, which might reflect changes in the receptor\'s conformation/modification state or, alternatively, might imply unidentified molecular components interacting with RET at the cell surface. This latter possibility is supported by a recent report demonstrating that the presence of heparan sulfate glycosaminoglycan on the cell surface is required for RET-dependent GDNF intracellular signaling \[[@pbio-0030123-b31]\].
Our interpretation of the D4 aptamer\'s mode of action relies upon three observations: (1) D4 binds with similar affinities to cells expressing RET in a monomeric or dimeric form; (2) D4 inhibits dimerization-dependent RET activation, as a consequence either of GDNF stimulation of RET^wt^ or RET^M918T^ or of constitutive dimerization of the RET^C634Y^ mutant; and (3) D4 does not inhibit a monomeric form of RET that is constitutively activated by a mutation in the intracellular kinase domain (RET^M918T^). These results taken together are compatible with the notion that D4 acts by interfering with the formation of a stable, active RET dimer, regardless of whether dimerization is caused by the formation of the RET/GDNF/GFRα1 complex or by the direct interaction of two mutated RET^C634Y^ proteins. This might occur either by D4 binding to monomeric RET, which would impede subsequent formation of the dimer, or by binding directly to the dimer.
Differential whole-cell SELEX strategies (this work; see also \[[@pbio-0030123-b05],[@pbio-0030123-b07],[@pbio-0030123-b08],[@pbio-0030123-b10]\]) can be employed to identify new markers on the surface of a given cell type, define the specificity of a cellular state, and/or allow in vivo targeting for diagnostic and therapeutic applications. The identification of lead compounds by reiterated affinity selection on living cells appears crucial when the molecular target is a membrane-bound or large transmembrane protein for which the conformation is frequently dictated by the interaction with other molecules, including membrane constituents \[[@pbio-0030123-b31]\]. Given that several of these proteins, as transmembrane receptors, integrins, and adhesion molecules, are involved in cell proliferation, apoptosis, and differentiation, aptamers for these targets could be promising prognostic tools in human therapy for widespread, devastating diseases such as cancer and neurodegeneration.
Materials and Methods {#s4}
=====================
{#s4a}
### Cell culture and immunoblot analysis {#s4a1}
Growth conditions for PC12 cells and derived cell lines were previously described \[[@pbio-0030123-b32]\]. NIH/MEN2A and NIH/MEN2B cells were obtained from NIH3T3 cells stably transfected with vectors expressing human RET^C634Y^ and RET^M918T^. To assess the effects of aptamers on RET activity, cells (160,000 cells per 3.5-cm plate) were serum-starved for 2 h and then treated with the indicated amount of RNA aptamers or the starting RNA pool after a short denaturation-renaturation step. When indicated, 2.5S NGF (Upstate Biotechnology, Lake Placid), GDNF (Promega), or recombinant rat GFRα1-Fc chimera (R&D Systems, Minneapolis, Minnesota, United States) were added to the culture medium. Cell extracts and immunoblotting analysis were performed as described \[[@pbio-0030123-b23]\]. The primary antibodies used were anti-RET (C-19), anti-VGF (R-15), and anti-ERK1 (C-16) (all three, Santa Cruz Biotechnology, Santa Cruz, California, United States); and anti-(Tyr-phosphorylated) RET and anti-phospho-44/42 MAP kinase (also indicated as anti-\[phospho\]-ERK) monoclonal antibodies (E10) (both from Cell Signaling, Beverly, Massachusetts, United States). Four independent experiments were performed.
### Cell transformation and neurite outgrowth bioassay {#s4a2}
PC12-α1/wt or NIH3T3 cells were plated at equal density on 12-well culture plates. Aptamers were added at 3 μM final concentration to the growth medium. To ensure the continuous presence of a concentration of at least 200 nM, this treatment was renewed every 24 h, which takes into account the half-life of the D4 aptamer in 10% serum (approximately 6 h, unpublished data). At least 15 random fields were photographed every 24 h with a phase-contrast light microscope. To evaluate the effects of D4 on cell differentiation, cells were pretreated for 6 h with 400 nM D4 or D4Sc and then incubated with 50 ng/ml GDNF together with 3 μM of the appropriate aptamer (see above). At 24 and 48 h of GDNF stimulation, 50 cells per frame were counted and scored as having neurites or not. A neurite was operationally defined as a process outgrowth with a length more than twice the diameter of cell body.
### Ex vivo SELEX {#s4a3}
The SELEX cycle was performed essentially as described \[[@pbio-0030123-b33]\]. Transcription was performed in the presence of 1 mM 2′F-Py and a mutant form of T7 RNA polymerase (T7^Y639F^, kind gift of R. Souza) \[[@pbio-0030123-b11]\] was used to improve yields. 2′F-Py RNAs were used because of their increased resistance to degradation by seric nucleases. The complexity of the starting pool was roughly 10^14^. 2′F-Py RNAs (1--5 nmol) were heated at 85 °C for 5 min in 3 ml of RPMI 1640, snap-cooled on ice for 2 min, and allowed to warm up to 37 °C before incubation with the cells. Two counterselection steps were performed per cycle. To avoid selecting for aptamers nonspecifically recognizing the cell surface, the pool was first incubated for 30 min at 37 °C with 10^7^ PC12 cells, and unbound sequences were recovered by centrifugation. These were subsequently incubated with 10^7^ adherent PC12/MEN2B cells, expressing a human RET receptor mutated in the intracellular domain (RET^M918T^), and unbound sequences were recovered for the selection phase. This step was meant to select sequences recognizing specifically the human RET receptor mutated in the extracellular domain (RET^C634Y^) expressed on PC12/MEN2A cells. The recovered sequences were incubated with 10^7^ adherent PC12/MEN2A cells for 30 min at 37 °C in the presence of nonspecific competitor RNA (total yeast RNA) and recovered after several washings with 5 ml of RPMI by total RNA extraction (Extract-All, Eurobio, Les Ulis, France).
During the selection process, we progressively increased the selective pressure by increasing the number of washings (from one for the first cycle up to five for the last three cycles) and the amount of nonspecific RNA competitor (100 μg/ml in the last three cycles), and by decreasing the incubation time (from 30 to 15 min from round 5) and the number of cells exposed to the aptamers (5 × 10^6^ in the last three cycles). To follow the evolution of the pool we monitored the appearance of four-base restriction sites in the population, which reveals the emergence of distinct families in the population \[[@pbio-0030123-b34]\]. After 15 rounds of selection, sequences were cloned with TOPO-TA cloning kit (Invitrogen, Carlsbad, California, United States) and analyzed.
### Binding experiments {#s4a4}
Binding of individual aptamers (or the starting pool as a control) to PC12 cells and derivatives was performed in 24-well plates in triplicate with 5′-^32^P-labeled RNA. 10^5^ cells per well were incubated with various concentrations of individual aptamers in 200 μl of RPMI for 10 min at 37 °C in the presence of 100 μg/ml polyinosine as a nonspecific competitor. After extensive washings (5 × 500 μl of RPMI), bound sequences were recovered in 350 μl of SDS 0.6%, and the amount of radioactivity recovered was normalized to the number of cells by measuring the protein content of each well. Binding of individual sequences to different cell lines was performed in the same condition at 50 nM only.
For the binding curve of D4 to PC12/MEN2A cells (see [Figure 2](#pbio-0030123-g002){ref-type="fig"}B), nonspecific binding was assessed using a 5′-^32^P-labeled naive pool of 2′F-RNAs (i.e., the starting pool of the selection), and the background values obtained were subtracted from the values obtained with the D4 aptamer. Apparent Kd values for each aptamers were determined by Scatchard analysis according to the equation
where \[T\]~tot~ represents the total target concentration.
Supporting Information {#s5}
======================
Accession Numbers {#s5a1}
-----------------
The Swiss-Prot (<http://www.ebi.ac.uk/swissprot/>) accession numbers for the proteins discussed in this paper are ERK (P27361), GDNF (P39905), GFRα1 (P56159), NGF (P01138), RET RTK (P07949), TrkA (P04629), and VGF (P20156).
This work was supported by the European Union contract QLG1--2000--00562 (Oligonucleotide Ligands Imaging, OLIM ), the European Molecular Imaging Laboratory (EMIL) network, the CNRS, the Association por la Recherche contre le Cancer (grant 3527) and the MIUR-FIRB (Ministero dell\'Istruzione, dell\'Università e della Ricerca Fondo per gli Investimenti della Ricerca di Base) grant RBNE0155LB. FD was supported by a Commissariat à l\'Energie Atomique (CEA) fellowship. We wish to thank M. Buckingham, E. Brody, M. S. Carlomagno, L. Di Giamberardino, C. Ibanez, C. Mann, S. Tajbakhsh and J.J. Toulmé for critical reading of the manuscript and fruitful discussions, and R. Souza for the gift of a T7^Y639F^ RNA polymerase-expressing plasmid.
2′F-Py
: 2′-fluoropyrimidine
ERK
: extracellular signal-regulated protein kinase
GDNF
: glial cell line-derived neurotrophic factor
GFR
: GDNF family receptor α1
HER
: heregulin
MEN
: multiple endocrine neoplasia
NGF
: nerve growth factor
RET
: rearranged during transfection
RTK
: receptor tyrosine kinase
SELEX
: systematic evolution of ligands by exponential enrichment
VGF
: nerve growth factor-inducible protein
[^1]: LC, FD, BT, VdF, and DL conceived and designed the experiments. LC, FD, CP, JB, YA, and KG performed the experiments. LC, FD, BT, VdF, and DL analyzed the data. VdF and DL contributed reagents/materials/analysis tools. BT, VdF, and DL wrote the paper.
[^2]: A patent application was filed covering the D4 aptamer and its use in diagnostic and therapeutics of cancer.
|
PubMed Central
|
2024-06-05T03:55:54.910956
|
2005-3-22
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065267/",
"journal": "PLoS Biol. 2005 Apr 22; 3(4):e123",
"authors": [
{
"first": "Laura",
"last": "Cerchia"
},
{
"first": "Frédéric",
"last": "Ducongé"
},
{
"first": "Carine",
"last": "Pestourie"
},
{
"first": "Jocelyne",
"last": "Boulay"
},
{
"first": "Youssef",
"last": "Aissouni"
},
{
"first": "Karine",
"last": "Gombert"
},
{
"first": "Bertrand",
"last": "Tavitian"
},
{
"first": "Vittorio",
"last": "de Franciscis"
},
{
"first": "Domenico",
"last": "Libri"
}
]
}
|
PMC1065268
|
From *The Worst Case Scenario Survival Handbook*---with handy entries like "How to escape from killer bees" and "How to escape from quicksand"---to *The Zombie Survival Guide: Complete Protection from the Living Dead*, survival guides are one of the latest publishing fads.
If there was a market for it, a survival guide for bacteria might include topics like "How to use your pili to keep your host from going apoptotic." A host\'s cells can respond to a bacterial infection with apoptosis, or programmed cell death. For bacteria that pass directly from host to host, this can pose a problem. If the bacteria are highly virulent and induce too much cell death, they could take down their host before they\'re able to jump ship, thus hurting the bacteria\'s chances of survival in the long run.
Earlier studies suggested that bacteria can use their pili, finger-like appendages that many bear on their surface, to pull on a host\'s cell membranes and thus influence the cell\'s behavior. But these studies, which looked at mutant bacteria that could not retract their pili, did not examine the matter of how the bacteria coax their hosts to stay alive.
Now, in *PLoS Biology*, a group of researchers present more direct evidence that bacteria can induce changes in hosts\' gene expression---and possibly keep the host cells alive longer---through tiny tugs on cell membranes. The study, led by Magdalene So, examined gene activity in human epithelial cells infected with Neisseria gonorrhoeae, the bacteria responsible for the sexually transmitted disease gonorrhea.
By comparing cells infected with normal N. gonorrhoeae to those infected with a mutant strain with defective pili, the researchers found a subset of 52 host genes that had higher activity when the host was infected with the normal bacteria, suggesting that the pulls of the pili were responsible. They also ran a key control experiment with an artificial mechanical pull on the host cell membrane. By coating magnetic beads with a preparation of bacterial pili, the beads attached themselves to the cell membranes. Then, in the presence of a magnetic field, the beads tugged on the cell membrane, approximating the effects on gene expression during infection with normal bacteria.
Thus, the mechanical tugs seem responsible for triggering a signaling cascade in the host cells, which ultimately affects the host\'s gene expression. Many of the genes that increased in activity due to the tugs were already known to regulate apoptosis and cellular response to stress, including mechanical strain on the membrane. Also, a majority of these genes were known to be induced by a family of proteins called mitogen-activated protein kinases, or MAPKs. The researchers showed that blocking MAPKs reduced the activity of several of the genes that are usually enhanced by infection with the normal bacteria. Also, they found that cells infected with the bacteria tended to survive treatment with staurosporine, a chemical that normally induces apoptosis.
Overall, the group\'s findings support previous speculations that some bacteria influence gene expression and the fate of cells in their hosts by tugging on the host cells\' membranes with their pili. For bacteria like N. gonorrhoeae that pass directly from host to host, the researchers argue, it would be in a bacterium\'s interest to help keep its host alive. And bacteria appear to do this with the help of their pili.
|
PubMed Central
|
2024-06-05T03:55:54.914051
|
2005-3-22
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065268/",
"journal": "PLoS Biol. 2005 Apr 22; 3(4):e140",
"authors": []
}
|
PMC1065269
|
For most of human history, cancer has been incurable. But with the invention of anesthesia in the mid-19th century, surgeons were able to remove some forms of cancer surgically. Radiotherapy arrived next, soon after the discovery of X rays in 1896. Chemotherapy, now a mainstay of cancer treatment, did not arrive until the mid-1940s, when nitrogen mustard, an alkylating agent related to the mustard gas used in the two World Wars, was developed as an anticancer agent. Unfortunately, although cancer cells are hypersensitive to the effects of alkylating agents---molecules that introduce lethal changes into the cell\'s DNA---normal cells are also targeted by them, although less damage is caused because normal cells typically divide slower than cancer cells. Using chemotherapy based on alkylating agents to treat cancer is like using a sledgehammer to crack a nut. But with improved knowledge about how cancer cells differ from normal cells, chemotherapeutics are now being designed that hit only cancer cells.
Many of these new chemotherapeutics target protein receptors called tyrosine kinases. These receptors, which sit on the cell surface, normally stimulate intracellular pathways that control proliferation and other cellular functions in response to growth factors. In tumors, these receptors often have mutations that allow them to become active without growth factor binding, which results in the uncontrolled proliferation that is characteristic of cancer cells. For instance, mutations in the RET receptor tyrosine kinase are responsible for multiple endocrine neoplasia (MEN) type 2 syndromes. Whereas external stimulation by a growth factor is normally needed before two RET molecules can bind together (a process called dimerization) to activate intracellular signaling cascades, in MEN type 2A, a mutation in the RET receptor tyrosine kinase provokes (or induces) dimerization without external stimulation.
In recent years, several proteins and various small synthetic chemicals have been designed that specifically inhibit the activity of mutated receptor tyrosine kinases and show anticancer activity. Domenico Libri and colleagues are now working on another class of molecules, called aptamers, that have potential as anticancer drugs. Aptamers---single-stranded nucleic acid molecules that are 50--100 bases long and can be selected for their ability to bind directly and tightly to specific proteins---are less likely to be targeted and destroyed by the body\'s natural defenses than some other types of potential therapeutic molecules.
To find an aptamer able to recognize the RET receptor kinase within a cellular membrane environment, the researchers used whole-cell SELEX (systematic evolution of ligands by exponential enrichment), a process in which large pools of oligonucleotides are enriched for molecules that can distinguish between a real and sham target. First, they incubated a large pool of RNAs with PC12 cells, a rat cell line not expressing RET, to remove sequences binding non-specifically to the PC12 cell surface. Unbound sequences were recovered and applied to PC12 cells expressing human RET with the MEN type 2A mutation that causes dimerization. This time, bound sequences were retained, and the whole selection process was repeated another 14 times to select for aptamers that recognize the dimeric form of the RET extracellular domain.[](#pbio-0030147-g001){ref-type="fig"}
::: {#pbio-0030147-g001 .fig}
::: {.caption}
###### A newly synthesized molecule, D4, inhibits cellular differentiation
:::
:::
Of the 67 sequences pulled out of the final pool of RNAs, the researchers found one sequence, D4, that not only bound the extracellular domain of RET but also blocked RET downstream signaling events and subsequent cellular and molecular changes. The researchers suggest that D4 blocks the dimerization-dependent activation of RET---whether it\'s induced by its physiological signaling molecule or by an activating mutation---and suggest that their method can be used to identify macromolecules with potential therapeutic effects against other transmembrane receptors involved in tumorigenesis, particularly since the whole-cell SELEX approach should efficiently select aptamers that recognize these receptors as they are found on the surface of tumor cells.
|
PubMed Central
|
2024-06-05T03:55:54.914683
|
2005-3-22
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065269/",
"journal": "PLoS Biol. 2005 Apr 22; 3(4):e147",
"authors": []
}
|
PMC1065270
|
For a virus to survive, it must elude the ever vigilant immune sentinels of its host. A latent virus can escape immune detection if it resides in nondividing cells and doesn\'t produce any proteins. No viral proteins means no red flags for immune cells. If the virus targets one of the many cell types that rarely divide, it\'s relatively safe while latent. But some viruses, like the gamma-herpesvirus, infect B cells of the immune system, which occasionally divide. The gamma-herpesvirus genome persists as circular pieces of DNA called episomes. When an infected B cell divides, the latent gamma-herpes virus episome must replicate and segregate into daughter cells along with the cell\'s genome. Viral replication and segregation requires the services of a protein called the episome maintenance protein---a potentially recognizable target for immune cells.
Gamma-herpesviruses, including Epstein-Barr virus (EBV) and Kaposi\'s sarcoma--associated herpesvirus (KSHV), can induce uncontrolled lymphocyte (immune cell) proliferation and result in lymphoma, Hodgkin\'s disease, and Kaposi\'s sarcoma. These diseases arise from the persistent latent infections that take hold after initial infections are controlled by immune defenses. The episome maintenance protein produced by EBV, called EBNA-1, harbors an amino acid element in its epitope---the region that binds to a T cell and triggers an immune response---that helps the viral protein evade the killer T cells that could destroy it. Lab studies show that the amino acid element limits EBNA-1\'s interaction with T cells by inhibiting synthesis and, to a lesser degree, degradation of the protein. How this evasive action works or helps the virus in a living organism is not entirely clear. But if T cells aren\'t presented with bits of viral protein, they have no way of knowing the virus is present.
In a new study, Neil Bennett, Janet May, and Philip Stevenson explore this question by studying virus--host interactions in mice infected with the murine gamma-herpesvirus-68 (MHV-68). Though MHV-68 infects mice, it behaves similarly to EBV and KSHV infections in humans, producing an acute mononucleosis-like illness and a pervasive pool of latently infected B cells. The episome maintenance protein in MHV-68 and KSHV is called ORF73. None of the viruses can maintain latent infections with deficient episome maintenance proteins.
Stevenson and colleagues first demonstrated that ORF73 limits T cell recognition and then identified a key region responsible for immune evasion by modifying different regions of the viral protein. In the next round of experiments, the authors asked how the viral protein manages this feat. They discovered that ORF73 limits T cell recognition much like EBNA-1 does, by reducing synthesis and degradation of the protein. One region strongly associated with inhibiting epitope presentation to killer T cells corresponded to reduced protein synthesis. When the authors modified the ORF73 transcript to circumvent T cell evasion, the T cells "wiped out" latent virus. These results indicate that avoiding epitope presentation during episome maintenance is key to the virus\'s survival.[](#pbio-0030149-g001){ref-type="fig"}
::: {#pbio-0030149-g001 .fig}
::: {.caption}
###### MHV-68 virions emerging from infected cells
:::
:::
Interestingly, the MHV-68 episome maintenance protein mediates immune evasion even though it lacks the amino acid element that does the job for EBV. Future studies will have to determine the responsible MHV-68 epitope and the mechanisms that engineer immune avoidance. Since a majority of epitopes that killer T cells recognize come from aborted translation events, it may be that evasive action is taken at the RNA transcript stage, before RNA is translated into protein. Evading killer T cells, the authors argue, is key to the survival of the gamma-herpesvirus. By figuring out just how evasion occurs, scientists can identify a promising target for controlling infection.
|
PubMed Central
|
2024-06-05T03:55:54.915466
|
2005-3-22
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065270/",
"journal": "PLoS Biol. 2005 Apr 22; 3(4):e149",
"authors": []
}
|
PMC1065312
|
Introduction
============
Vasospastic symptoms occur in about 30% of patients with primary fibromyalgia (FM) \[[@B1]\]. These patients present with Raynaud\'s phenomenon and intolerance to cold \[[@B2]\]. Frodin and colleagues, using nailfold capillaroscopy in FM patients, found slight morphological changes, such as moderate enlargement of capillary loops and variations in calibre \[[@B3]\]. Jeschonneck and colleagues showed decreased microcirculatory blood flow above tender points in FM patients \[[@B4]\]. About 60% to 90% of systemic scleroderma (SSc) patients have Raynaud\'s phenomenon \[[@B5]\]. In patients with SSc other workers, using videomicroscopy with sodium fluorescein, have found typical changes of the nailfold capillaries, characterised by reduced capillary density, giant capillaries, avascular fields, microhaemorrhages, and disturbance of diffusion \[[@B6]\]. Furthermore, in rheumatoid arthritis (RA), peripheral malperfusion and vasculitis occur \[[@B7]\], resulting in skin ulcers, neuropathy, necrosis, or gangrene.
Our aim in this preliminary study was to investigate capillary abnormalities and blood flow by two independent objective methods, capillary microscopy and laser Doppler fluxmetry, to obtain evidence of disturbed microcirculation in FM patients.
Materials and methods
=====================
The study group consisted of 10 women (54.0 ± 3.7 years of age) from the Outpatient Department in the Department of Rheumatology, University Hospital, Zurich, with primary FM classified in accordance with the criteria of the American College of Rheumatology \[[@B8]\]. The controls were three groups (*n*= 10 in each group) of age-matched women who were healthy (negative controls) or who had symptomatic RA or SSc (positive controls). The subjects were studied using laser Doppler fluxmetry and capillary microscopy.
None of the subjects had vasculitis. The RA patients did not have Raynaud\'s phenomenon and only 2 of the 10 had hypertension as a possible risk factor for small-vessel disease. Although the SSc patients did not have vasculitis, all except one presented with a typical Raynaud\'s syndrome, with a mean Medsger score \[[@B9]\] of 2.2 ± 1.03. This score describes how advanced the disease is (Table [1](#T1){ref-type="table"}); a low scores (0 or 1) indicates no or mild SSc, and a high score (4) indicates the end stage. Smokers and patients treated with nitrate or Ca^2+^-channel blockers were excluded from the study. The study was approved by the local ethical committee of the University Hospital, Zurich.
Capillary microscopy
--------------------
The morphology of nailfold capillaries has been studied by intravital capillaroscopy (Leica, Glattbrugg, Switzerland) at a magnification of 50× \[[@B10]\]. The room temperature was maintained between 22°C and 24°C. Patients were examined in a sitting position after a resting time of at least 20 minutes. The capillaries were evaluated in accordance with the criteria of the German Association of Angiology \[[@B11]\]. The following changes were analysed: density of capillaries (normal 7--16 capillaries per millimetre), microhaemorrhages, dilatation of capillaries, giant capillaries, and \'irregular formations\' (that is, instances where capillaries were arranged in clusters with gaps in between). A dilatation was considered to be present when the arteriolar limb of the capillary loop was thicker than 50 µm and the venous limb was thicker than 20 µm, and giant capillaries were defined as those having an apex diameter of over 50 µm \[[@B11]\].
Laser Doppler fluxmetry
-----------------------
Skin blood flow was measured in supine subjects at the lateral epicondyle (typical FM tender point \[[@B8]\]), in fingertips II and III (that is, of the forefinger and middle finger) and in the lower arm (control point) using the laser Doppler technique (PeriFlux PF3; Perimed, Järfälla, Sweden), as described elsewhere \[[@B12]\]. A blood-pressure cuff was positioned on the upper arm and standard laser Doppler probes for skin blood flow measurements were attached to the epicondyle, fingertips II and III, and the lower arm. The resting flow was recorded 5 minutes before the cuff pressure was inflated to a suprasystolic level for 3 minutes. After release of the cuff pressure, reactive hyperaemia was recorded at the four defined areas. The time to peak flow and type of peak were evaluated in each group \[[@B13]\]. Peak flow corresponds to the highest flow value after release of the cuff.
Four types of reactive hyperaemia were identified \[[@B13]\] (Fig. [1](#F1){ref-type="fig"}). In type A, the first peak is within 23 seconds after cuff release and is followed by a second, smaller, wave. In type B, the amplitude of the second wave is greater than that of the first; the first peak is characterised by a fast dilatation of the myogen-activated arterioles and small arteries with a concomitant increase of the vessel tonus. Both types A and B are biphasic and are classified as \'normal\', because so far they have been predominantly found in healthy subjects \[[@B13]\]. They also show the same characteristics in therapy and there is so far no proof that one type predisposes to a certain illness. Type C is monophasic; the fast initial component of the muscular reaction is absent. In type D, postocclusive reactive hyperaemia is missing. Types C and D are pathological reactions.
The Mann--Whitney *U*test was used for statistical comparison of the groups. *P*values \< 0.05 were considered to be statistically significant. Means ± standard deviations are given.
Results
=======
Capillary microscopy
--------------------
The density of capillaries per millimetre in FM patients (9.92 ± 0.19) was significantly lower than in controls (11.31 ± 0.34) (*P*\< 0.001) but still within the normal range. Four or more capillary dilatations were detected in 2 of the 10 FM patients (Fig. [2](#F2){ref-type="fig"}), and one to three dilatations per nail fold were detected in 6 of the10 FM patients. No microhaemorrhages, giant capillaries, or avascular fields were detected in FM patients.
The number of capillaries in patients with SSc (6.21 ± 1.03) was significantly lower than in healthy controls (Fig. [3](#F3){ref-type="fig"}) and significantly more microhaemorrhages were found in SSc patients (8 of 10) than in controls or in FM or RA patients (P \< 0.01). Giant capillaries were detected only in SSc patients.
Laser Doppler fluxmetry
-----------------------
The time to peak blood flow at the lateral epicondyle was significantly longer in FM (7 ± 0.5 s) and SSc (7 ± 0.91 s) patients than in healthy controls (4 ± 0.34 s) (Fig. [4](#F4){ref-type="fig"}). In SSc patients, the time to the peak in the second finger (7.5 ± 1.22 s) was significantly longer than in FM patients (5 ± 0.27 s) and healthy controls (4.5 ± 0.58 s) and also in the third finger (7.5 ± 0.67 s, 5 ± 0.3 s, and 4.5 ± 0.17 s, respectively) (Fig. [4](#F4){ref-type="fig"}). In RA patients, the time to the peak in the second finger (7 ± 1.15 s) was significantly longer than in FM patients (5 ± 0.27 s) and healthy controls (4.5 ± 0.58 s). In the lateral epicondyle, both FM and SSc patients had longer times to peak than the healthy controls (Fig. [4](#F4){ref-type="fig"}).
All of the FM patients showed a type-B hyperaemic response in the lower-arm and epicondyle measurements (Fig. [5](#F5){ref-type="fig"}). The monophasic, type-C response was seen at the lateral epicondyle in 6 of 10 RA patients and 6 of 10 SSc patients. One patient with SSc and one with RA showed no postocclusive reaction (type D) either in the lower arm or at the lateral epicondyle.
In measurements made in the fingers, FM patients showed the postocclusive, type-B response in fingers II and III (Fig. [6a,6b](#F6){ref-type="fig"}), except for a type-A response in finger II in one patient. This was significantly different from the response in healthy controls (*P*\< 0.01). The monophasic, type-C response was found in some patients with SSc and RA. In one patient with SSc, a type-D response was observed in all four fingers.
Discussion
==========
Patients with FM and SSc present functional as well as morphological changes in microcirculation, but the diseases are distinguishable by the severity of morphological pathologies of capillaries in SSc. Specific capillary abnormalities are present in patients with SSc and have a high predictive value \[[@B5]\]. These changes and the irregular formations in SSc may be due to microinfarcts \[[@B14]\]. Our results show that the density of capillaries in patients with FM is still normal but lower than in healthy controls. In an earlier study, however, it was reported that the number of capillaries is decreased \[[@B15]\]. Morphological abnormalities and vascular malfunction (spasms) therefore have to be discussed as possible reasons for this decreased number of capillaries. In our study, the main finding is a longer time to peak flow (reactive hyperaemia) after occlusion in FM patients than in the healthy controls. An earlier study \[[@B13]\] showed that 80% of healthy persons have a biphasic type of reactive hyperaemia, such as type A or B (see Fig. [1](#F1){ref-type="fig"}). Apart from one type-A response in the second fingertip, all of the patients with FM whom we studied were recorded as having a type-B response. This type of response is classified as \'normal\', but it lacks the first, fast, myogen-activated peak. The explanation may be that FM patients have a reduced primary muscular vessel reaction whereas -- and this is important -- the second wave is normal.
The small number of patients in this preliminary study may be a limitation in that we may have coincidentally recorded 10 FM patients with a type-B response to occlusion of the blood flow. Further studies with more patients are needed to confirm this finding.
The missing fast component of reactive hyperaemia in our FM population is presumably due to a higher sympathetic tonus, resulting in increased vasoconstriction; this increased vasoconstriction would explain both the significantly increased time to peak, especially at the lateral epicondyle -- which is a tender point in FM -- and the reduced density of vessels. Earlier workers \[[@B4]\] advanced the idea that psychological and physical situations of stress might have an impact on this system. Local ischaemia, which may result from these proposed mechanisms in more advanced stages of the disease, could be a possible explanation for Raynaud\'s phenomenon in a fraction of FM patients. This local ischaemia results in an influence on spinal and supraspinal structures with sympathetic and motor efferences.
SSc patients showed also a prolonged time to peak flow at all points studied. It is well known that endothelial changes are present in SSc \[[@B14]\].
Both morphological and symptomatic disturbances in microcirculation can occur in FM but occur most often in SSc \[[@B16]\]. Functional changes have already been observed in FM, SSc, and RA \[[@B7],[@B17]\]. However, the changes in FM support the hypothesis of increased sympathetic activity, and hence a functional hyperexcitability of the sympathetic nervous system \[[@B4]\].
Conclusion
==========
We have shown that functional changes of the microcirculation are present in patients with FM and this finding may be important for new treatment options in FM. A possible therapeutic strategy could be selective suppression of the sympathetic tone or the undertaking of symptomatic measures to activate the microcirculation, such as active and passive physical methods.
Because the unblinded design was a weakness of this preliminary study, blinded studies with more patients are needed to confirm our findings.
Abbreviations
=============
FM = primary fibromyalgia; RA = rheumatoid arthritis; SSc = systemic scleroderma.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
SM recorded measurements for all subjects and performed the capillary microscopy and laser fluxmetry. AF helped with the capillary microscopy. BA-V helped with the laser fluxmetry. UKF and RK discussed the methods and the results of the measurements. DU gave advice with respect to the study design and manuscript. HS developed the study and supervised the work of SM. All authors read and approved the final manuscript.
Acknowledgements
================
The authors wish to thank Leanne Pobjoy for her help in preparing the manuscript.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Curves depicting the four types of reactive hyperaemia measured by laser Doppler fluxmetry after occlusion of blood flow with a blood-pressure cuff for 3 minutes (occlusion marked by vertical dashed lines). (Reproduced from reference \[13\] with permission.)
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Example of capillary microscopy of the nail fold of a 44-year-old FM patient. The number of capillaries per millimetre (6) is reduced and capillary dilatations and irregular formations of the capillaries are present.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Number of capillaries (mean ± standard deviation) per millimetre in the nail fold as seen on capillary microscopy. \**P*\< 0.05 in comparison with healthy controls (Co); \*\**P*\< 0.01 in comparison with Co and with primary fibromyalgia (FM) or rheumatoid arthritis (RA) patients; \*\*\**P*\< 0.001 in comparison with Co.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Time (s) (mean ± standard deviation) to peak capillary flow after occlusion of blood flow with a blood-pressure cuff on the upper arm for 3 minutes. Measurements were made using laser Doppler fluxmetry. In primary fibromyalgia (FM) and systemic scleroderma (SSc) patients, the time to peak in the lateral epicondyle was longer than in healthy controls (Co). In SSc patients, the times to peak in the second and third fingertips were longer than those in FM patients and healthy controls. \**P*\< 0.001 in comparison with Co; °*P*\< 0.05 in comparison with Co and with FM patients; ^+^*P*\< 0.05 in comparison with Co and with FM or rheumatoid arthritis (RA) patients.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Histograms showing types of peak flow as measured using laser Doppler fluxmetry \[13\]. **(a)**In the lower arm and **(b)**at the lateral epicondyle \[13\] of patients with systemic scleroderma (SSc), rheumatoid arthritis (RA), or primary fibromyalgia (FM) and in healthy controls (Co). \**P*\< 0.05, \*\**P*\< 0.01, in comparison with Co.
:::
:::
::: {#F6 .fig}
Figure 6
::: {.caption}
######
Histograms showing types of peak flow as measured using laser Doppler fluxmetry \[13\]. In the **(a)**second and **(b)**third fingertips of patients with systemic scleroderma (SSc), rheumatoid arthritis (RA), or primary fibromyalgia (FM) and in healthy controls (Co). °*P*\< 0.05 in comparison with Co; §*P*\< 0.05 in comparison with Co and FM; \**P*\< 0.05 in comparison with Co, FM and RA; \*\**P*\< 0.01 in comparison with Co.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Characteristics of patients and controls studied.
:::
Group, and subject no. Risk factor Raynaud\'s phenomenon Medsger score^a^
-------------------------------------------- -------------- ----------------------- ------------------
*Patients with fibromyalgia*
1 Obesity \-
2 \- \-
3 \- \-
4 \- X
5 \- X
6 Obesity \-
7 \- \-
8 \- \-
9 \- \-
10 \- \-
*Controls with systemic scleroderma (SSc)*
1 \- X 4
2 \- X 3
3 \- X 2
4 \- \- 1
5 \- X 2
6 Obesity X 2
7 Ex-smoker X 1
8 Ex-smoker X 3
9 \- X 1
10 \- X 3
*Controls with rheumatoid arthritis*
1 Obesity \-
2 Hypertension \-
3 \- \-
4 \- \-
5 \- \-
6 Obesity \-
7 Hypertension \-
8 \- \-
9 \- \-
10 \- \-
*Healthy controls*
1 Obesity \-
2 Ex-smoker \-
3 \- \-
4 \- \-
5 \- \-
6 Ex-smoker \-
7 Obesity \-
8 \- \-
9 \- \-
10 \- \-
^a^Assigned to SSc patients only; indicates severity of the disease, from 0 to 4 (none to end stage) \[9\]. -, Not present; X, present.
:::
|
PubMed Central
|
2024-06-05T03:55:54.916008
|
2004-12-10
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065312/",
"journal": "Arthritis Res Ther. 2005 Dec 10; 7(2):R209-R216",
"authors": [
{
"first": "Susanne",
"last": "Morf"
},
{
"first": "Beatrice",
"last": "Amann-Vesti"
},
{
"first": "Adrian",
"last": "Forster"
},
{
"first": "Ulrich K",
"last": "Franzeck"
},
{
"first": "Renate",
"last": "Koppensteiner"
},
{
"first": "Daniel",
"last": "Uebelhart"
},
{
"first": "Haiko",
"last": "Sprott"
}
]
}
|
PMC1065313
|
Introduction
============
Plasmacytoid dendritic cells (pDCs) are a distinct population of antigen-presenting cells (APCs) with the capacity for potent antigen-presenting function and production of large amounts of cytokines, including tumor necrosis factor (TNF)-α and IFN-α. Human pDCs can be identified by cell-surface expression of MHC molecules, the α-chain of the IL-3 receptor (CD123), and the presence of blood dendritic-cell (DC) antigens known as BDCA2 and BDCA4 in a proportion of cells \[[@B1]\]. In comparison with CD11c^+^myeloid DCs, pDCs display a distinct set of chemokine and Toll-like receptors \[[@B2]-[@B4]\]. In response to viruses and CpG DNA, pDCs become activated to produce IFN-α and their APC function is enhanced \[[@B5]-[@B8]\]. While pDCs were first demonstrated in the T-cell areas of lymph nodes \[[@B5],[@B9]\], precursors of this DC population have been isolated from several sources, including normal peripheral blood (PB), thymus, fetal liver, and bone marrow \[[@B10]\]. Although they do not reside in normal peripheral tissues, pDCs have been shown to infiltrate certain inflamed tissues and tumor sites, including the skin in psoriasis and lupus, the cerebrospinal fluid in multiple sclerosis, and melanoma and ovarian carcinoma \[[@B11]-[@B15]\]. While pDCs play an important effector role in viral disease, being the major producers of IFN-α and having a primary role in innate immunity, there is also evidence that they may play an immunoregulatory role, through the induction of Th2 (T helper 2)-type cytokines \[[@B9],[@B16]-[@B18]\].
The synovial autoimmune reaction of rheumatoid arthritis (RA) is characterized by lymphocyte, macrophage, and DC infiltration that can progress to the development of lymphoid tissue in established disease \[[@B19]-[@B21]\]. DCs are likely to contribute to the formation and maintenance of such organized lymphoid tissue and antigen presentation in RA and other autoimmune lesions \[[@B22]-[@B24]\]. We have previously shown that the effector site in RA synovial tissue is enriched in differentiated myeloid DCs, which express CD33, CD11c, MHC and costimulatory molecules, and nuclear RelB \[[@B21],[@B25]\]. Translocation of RelB to the nucleus of myeloid DCs is associated with APC function, particularly through increased expression of MHC molecules CD86 and CD40 \[[@B26]\].
The proinflammatory cytokines TNF-α and IL-1β are key contributors to the inflammatory cytokine cascade in RA \[[@B27],[@B28]\]. This relates to a number of actions, but activation of the endothelium by TNF-α is particularly important in cellular recruitment to the synovium \[[@B29]-[@B31]\]. Since RA is characterized by endothelial activation, leukocyte recruitment, and the development of high endothelial venules, we hypothesized that pDCs would be enriched in inflamed but not normal synovium. Since the functional role of pDCs in disease pathogenesis is only partly understood, we also wished to address whether these cells represent a population distinct from the described nuclear RelB^+^synovial DCs, and whether they may contribute as APCs or inflammatory cells in RA \[[@B21]\].
Materials and methods
=====================
Patients and controls
---------------------
Thirty patients who fulfilled the American College of Rheumatology criteria for RA were included \[[@B32]\]. Of these, 10 provided synovial fluid (SF) samples and 27 provided PB samples. Of the 30 patients, 80% were seropositive, 62% were female, and 73% were taking at least one disease-modifying antirheumatic drug or low-dose prednisone or both. Synovial tissue was obtained at arthroscopy from seven patients with RA, of whom three were untreated and four were taking at least one disease-modifying antirheumatic drug and low-dose prednisone. The duration of disease ranged from 0.5 to 18 years. In addition, we studied synovial tissue from four healthy individuals with nonspecific knee pain undergoing arthroscopy, one patient who had had psoriatic arthritis for 8 years, and one patient who had had ankylosing spondylitis for 30 years. Each patient with spondyloarthropathy was taking sulfasalazine. No patient in the study was taking biologics. Synovial tissue was provided by Dr Malcolm Smith (Repatriation Hospital, Adelaide, Australia). PB buffy coats prepared from 30 healthy donors were obtained from the Red Cross Blood Transfusion Service (Brisbane, QLD, Australia). The study was approved by the Research Ethics Committee of the Princess Alexandra Hospital.
Culture medium and cell isolation
---------------------------------
All cells were cultured in RPMI 1640 (Gibco, Life Technologies, Mulgrave, VIC, Australia) supplemented with 10% FCS (CSL Ltd, Parkville, VIC, Australia), 0.3 mg/ml L-glutamine (Trace Biosciences, Castle Hill, NSW, Australia), 0.12 mg/ml benzylpenicillin (CSL), and 10 μg/ml gentamicin (Delta West, Pharmacia and Upjohn, Spring Hill, QLD, Australia). The monoclonal antibodies used in this study include FITC, phycoerythrin (PE), and purified anti-CD11c, CD14-PerCP, PE, biotinylated and purified anti-CD123, CD86-FITC (all from BD Pharmingen, San Diego, CA, USA), BDCA2-FITC (Miltenyi Biotech, San Francisco, CA, USA), HLA-DR-biotin (Coulter Immunotech, Fullerton, CA, USA), CD40-FITC (Biolegend, San Diego, CA, USA), CD80-FITC (Cymbus Biotech, Chandlers Ford, Hants, UK), CD68 (Kp-1, DAKO, Carpinteria, CA, USA), RelB (C-19, Santa Cruz Biotech, Santa Cruz, CA, USA), and biotinylated *Ulex europaeus*agglutinin I (Vector Laboratories, Burlingame, CA, USA).
Mononuclear cells were prepared from normal or RA PB or RA SF by density gradient centrifugation over Ficoll-Paque (Pharmacia Biotech, Uppsala, Sweden) as described elsewhere \[[@B33]\]. T cells were purified from PB mononuclear cells by passing the cells over a nylon wool column, followed by immunomagnetic depletion of remaining monocytes, DCs, B cells, and NK (natural killer) cells using monoclonal antibodies against CD14, CD16, CD19, CD56, and HLA-DR (all from BD Pharmingen), followed by goat antimouse immunoglobulin magnetic beads, then passage through a strong magnetic field (MACS, Miltenyi Biotech), and collection of the unbound fraction. On analysis by flow cytometry, the unbound fraction routinely contained 95--98% CD3^+^T cells. DC-enriched non-T cells were produced by immunomagnetic depletion of T, B, and NK cells from non-T cells, by incubation with monoclonal antibodies against CD19, CD16, CD56, and CD3.
Flow cytometric analysis and selection of cells by cell sorting
---------------------------------------------------------------
To enumerate CD123^+^and CD11c^+^subsets of DCs, mononuclear cells from normal PB or RA SF were stained for four-colour flow cytometry as described elsewhere \[[@B33]\], using monoclonal antibodies against CD14-PECy5, CD11c-FITC, CD123-PE, and HLA-DR-APC. Live CD14^-^HLA-DR^+^mononuclear cells were gated for analysis. Subset percentages are expressed as percentage of total mononuclear cells. Listmode data were analyzed using Winlist 2.0 software (Verity Software House, Topsham, ME, USA). For sorting, PB or SF DC-enriched non-T cells stained with the same four markers were sorted using the Moflo flow cytometer (DAKO), gating on CD14^-^HLA-DR^+^and either CD123^+^CD11c^-^or CD11c^+^CD123^-^cells, respectively. For phenotypic analysis, mononuclear cells from PB or SF were stained with CD14-PECy5, CD123-PE or CD11c-PE, HLA-DR-APC, and either a fourth monoclonal antibody or isotype control monoclonal antibody conjugated with FITC. DCs were gated as described above.
Electron microscopy
-------------------
Electron microscopy of freshly sorted cells was carried out as described elsewhere \[[@B5]\]. After fixation in 2.5% glutaraldehyde in phosphate-buffered saline, the cells were post-fixed with an aqueous solution of 1% OsO~4~containing 1.5% K~4~Fe(CN)~6~. Subsequently, the specimens were dehydrated in an alcohol series and embedded into epon. Ultrathin sections (50 nm) were contrasted with lead citrate and uranyl acetate and studied with a CM100 electron microscope (Philips, Eindhoven, The Netherlands).
Mixed lymphocyte reactions and cytokine analysis
------------------------------------------------
Various numbers of sorted PB or RA SF DCs were incubated with 10^5^allogeneic PB T cells in triplicate wells for 5 days, as described elsewhere \[[@B33]\]. Supernatants were removed from some cultures and \[^3^H\]thymidine (1 μCi/well, ICN Biochemicals) was added to the remainder for the final 18 h. Cells were harvested onto glass-fiber filter mats and the incorporation of \[^3^H\]thymidine was determined by liquid scintillation spectroscopy (Packard Topcount, Packard Instrument Co, Meriden, CT, USA). IFN-γ, IL-10, and TNF-α were measured in supernatants by ELISA using OptEIA ELISA kits (BD Pharmingen).
Immunohistochemistry
--------------------
Frozen or paraffin-embedded sections of synovial tissue from patients with untreated active RA were obtained by arthroscopic biopsy and supplied by Malcolm Smith (Repatriation Hospital, Adelaide, Australia). Normal synovial tissue was obtained at arthroscopy from patients undergoing arthroscopy for nonspecific knee pain and in whom no abnormality was found. After fixation with acetone, sections were stained with anti-CD11c or anti-CD123 using an immunoperoxidase technique, and revealed with diaminobenzidine (brown). Frozen sections were double-stained with *U. europaeus*agglutinin I (Ulex), a lectin that specifically binds endothelial cells (fast red), and anti-CD123 (brown), using a double, immunoperoxidase--immunoalkaline phosphatase technique as described elsewhere \[[@B34]\]. Formalin-fixed paraffin-embedded sections were antigen-retrieved in 10 mM citrate buffer at pH6 in an autoclave, then stained with anti-CD123 (diaminobenzidine, brown) alone, or in combination with anti-RelB (BCIP, DAKO, purple). Sections were counterstained with hematoxylin except when they had been double-stained for CD123 and RelB and were photographed using a transmitted-light microscope (Leitz Diaplan, Leica, Germany). To quantitate infiltration by CD123^+^DCs, the number of CD123^+^Ulex^-^cells was counted in sections double-stained with CD123 and Ulex. Cells were counted in each of the entire sections from three patients and three normal controls at high power, and for each biopsy this number was corrected for the area of the section to obtain the number per mm^2^. To quantitate infiltration by CD11c^+^cells, the number of these cells was counted in three high-power fields of the synovial sublining in sections from three patients and three normal controls.
Statistical analysis
--------------------
Differences were analyzed using unpaired Student\'s *t*-tests.
Results
=======
CD123^+^nuclear RelB^-^DCs are located in perivascular regions of RA synovial tissue
------------------------------------------------------------------------------------
We have previously shown that synovial tissue in RA and spondyloarthropathy is enriched in differentiated myeloid DCs that express CD33, CD11c, MHC class II, costimulatory molecules, and nuclear RelB \[[@B21],[@B25]\]. Translocation of RelB to the nucleus is associated with maturation and APC function of myeloid DCs \[[@B26]\]. These nuclear RelB^+^DCs are absent in normal synovial tissue and are rare in RA SF \[[@B21],[@B23]\]. To determine whether RA synovial tissue was infiltrated by CD123^+^pDCs in addition to CD11c^+^myeloid cells, frozen synovial tissue sections, either normal or from patients with RA or spondyloarthropathy, were stained with CD11c or CD123. CD11c^+^cells were found both in the lining layer and adjacent to vessels in the sublining of normal synovial tissue. In contrast, CD123 only stained endothelial cells in the normal tissue (Fig. [1a,1b](#F1){ref-type="fig"}). In RA synovial tissue, CD11c again stained cells adjacent to vessels, now within lymphoid aggregates in the sublining. A population of CD123^+^cells with dendritic appearance was also stained adjacent to CD123^+^blood vessels in RA (Fig. [1c,1d](#F1){ref-type="fig"},). Cells expressing TNF-α in RA synovial tissue were found in a similar location in serial sections (data not shown), as demonstrated previously \[[@B35]\]. To confirm the perivascular CD123^+^cells in synovial tissue, normal and RA synovial tissue were double-stained with the endothelial cell marker Ulex agglutinin (red) and with CD123 (brown). Whereas all CD123^+^structures in normal synovial tissue colocalized with Ulex agglutinin (orange), single-stained CD123^+^cells (brown) were located in perivascular lymphoid aggregates and within the lumen of occasional blood vessels in RA synovial tissue (Fig. [1e,1f](#F1){ref-type="fig"}). These CD123^+^cells are similar in appearance to those previously demonstrated as CD123^+^pDCs in human tonsil, in that they are smaller than CD11c^+^myeloid DCs, with shorter dendritic processes, and cell clusters gave the appearance of locally proliferating cells (Fig. [1g](#F1){ref-type="fig"}) \[[@B5],[@B36]\]. While some macrophages can express CD123, there was no colocalization in synovial tissue of CD123 and CD68 (data not shown). However, aside from the dendritic morphology, we cannot exclude that some of the CD123^+^cells stained are mast cells \[[@B37]\]. To determine whether CD123^+^cells in synovial tissue were also nuclear RelB^+^, formalin-fixed tissue was double-stained for RelB and CD123 without hematoxylin counterstaining. No CD123^+^cells had translocated RelB to the nucleus, although some expressed cytoplasmic RelB (Fig. [1h,1i](#F1){ref-type="fig"}). In contrast, nuclear staining of RelB was evident in adjacent CD123^-^cells (Fig. [1h](#F1){ref-type="fig"}, arrows). All patients with RA showed similar infiltration by pDCs and no differences in the cell numbers or location were noted between patients with RA or spondyloarthropathy (data not shown).
We quantitated pDCs in normal or RA synovial tissue by counting CD123^+^Ulex^-^cells in synovial tissue sections from patients with RA or normal controls stained with CD123 and Ulex as shown in Fig. [1](#F1){ref-type="fig"}. Whereas no pDCs infiltrated the normal tissue, approximately 22 pDCs per mm^2^were identified within the RA tissue (Fig. [2](#F2){ref-type="fig"}). This number is similar to the number of nuclear RelB^+^differentiated DCs identified previously in RA synovial tissue \[[@B38]\]. In contrast, CD11c^+^cells infiltrated both normal and RA synovial tissue, with significantly larger numbers in RA (*P*\< 0.05) (Fig. [2](#F2){ref-type="fig"}). We conclude that the CD123^+^cell population is most likely a pDC population that infiltrates RA and spondyloarthropathy but not normal synovial tissue and that it is distinct from the described nuclear RelB^+^DCs \[[@B21],[@B36],[@B39]\]. CD11c^+^cells comprise immature and differentiated myeloid DCs as well as monocytes \[[@B1],[@B34]\]. Differentiated nuclear RelB^+^DCs are found within the CD11c^+^DC population in RA and other inflammatory arthritides but not in normal synovial tissue \[[@B21]\].
CD11c^+^and CD123^+^DCs in RA SF
--------------------------------
Workers in our laboratory have previously shown that RA SF is enriched in CD11c^+^CD33^bright^CD14^-^myeloid DCs with efficient APC function \[[@B25],[@B40]\]. However, when freshly isolated, only a small proportion of SF CD33^bright^CD14^-^DCs have translocated RelB to the nucleus. RA and normal PB mononuclear cells contain similar proportions of CD33^bright^CD14^-^DCs \[[@B25]\]. To examine plasmacytoid and myeloid DCs in parallel, we compared RA SF with RA and healthy, control PB for the proportion of CD123^+^and CD11c^+^HLA-DR^+^CD14^-^DCs. After purification of mononuclear cells from either normal or RA PB or RA SF by gradient centrifugation, cells were stained with CD123-PE, CD11c-FITC, CD14-PECy5, and HLA-DR-APC. Polymorphonuclear cells were excluded on the basis of forward and side light-scatter. Since basophils and monocytes can also express CD123, potential CD123^+^non-DCs were excluded by gating CD14^-^HLA-DR^+^cells \[[@B10]\]. By four-color analysis, CD14^-^HLA-DR^+^CD123^+^and CD11c^+^DC populations could be distinguished (Fig. [3](#F3){ref-type="fig"}). The percentages of CD123^+^CD11c^-^pDCs in RA PB and normal PB were low and did not differ from each other. This observation contrasts with the reduction in pDCs observed in blood from patients with systemic lupus erythematosus \[[@B41]\]. CD11c^+^CD123^-^myeloid DCs were more common than CD123^+^DCs in patient and control blood (*P*\< 0.005), in keeping with previous studies of normal PB \[[@B1]\]. RA SF contained a significantly greater percentage of CD11c^+^DCs than normal or RA PB (*P*\< 0.005) -- in accord with previous studies using the markers CD33 and CD14 \[[@B40]\]. The proportion of CD11c+ DCs in RA SF was higher than that of RA SF CD123+ DCs (*P*\< 0.05). Although the difference was small, the percentage of CD123^+^DCs in RA SF was higher than in RA or control PB (*P*\< 0.05). The data show that CD123^+^DCs are present in RA SF, and that the ratio of CD11c^+^to CD123^+^DCs is similar in RA SF to that in normal or RA PB (approximately 10:1). In RA synovial tissue, mature myeloid nuclear RelB^+^and CD123^+^DCs have infiltrated perivascular lymphoid aggregates in similar numbers. Previously, similar numbers of immature and mature myeloid DCs were identified in RA synovial tissue \[[@B42]\]. Thus pDCs make up about 30% of DCs within RA synovial tissue. The present and previously published data, taken together, show that both pDCs and myeloid DCs are recruited to RA synovium, with an enrichment of pDCs in synovial tissue relative to blood or SF.
CD123^+^PB DCs are immature whereas SF pDCs show signs of activation
--------------------------------------------------------------------
In normal PB, pDCs circulate as precursors with the potential for recruitment into tissues in response to chemokines \[[@B2],[@B43]\]. These precursors exhibit a characteristic plasmacytoid morphology on electron microscopy, a cell-surface phenotype characterized by expression of the BDCA2 antigen, by low levels of costimulatory molecule expression, and by the potential for IFN-α production in response to viral or immunostimulatory CpG DNA motifs \[[@B1]\]. We therefore analyzed the characteristics of sorted RA SF CD123^+^DCs and compared them with control PB CD123^+^DCs. On electron microscopic examination, freshly sorted PB and SF CD123^+^DCs appeared similar, with a smooth surface and abundant rough endoplasmic reticulum in the cytoplasm. The nucleus was nonlobulated and abundant in euchromatin and contained a distinct nucleolus (Fig. [4](#F4){ref-type="fig"}). CD11c^+^PB DCs were morphologically distinct from the CD123^+^pDCs, with a lobulated nucleus and some phagocytic vesicles. CD11c^+^DCs from SF showed more membrane ruffling and phagocytic activity than those from PB (Fig. [4](#F4){ref-type="fig"}). Thus SF CD123^+^DCs morphologically resemble CD123^+^DCs in PB, whereas CD11c^+^SF DCs display a greater level of ruffling and phagocytic activity, consistent with their enhanced level of activation, than CD11c^+^circulating precursors \[[@B21]\].
On four-color flow cytometric analysis, gated RA SF CD14^-^HLA-DR^+^CD123^+^DCs expressed low levels of CD40, CD80, and CD86. All or the majority of SF CD123^+^DCs expressed the BDCA2 marker of immature pDC precursors \[[@B1]\]. This cell-surface phenotype closely resembles that of control PB CD123^+^DC precursors, although BDCA2 was consistently expressed at high levels only by a subset of CD123^+^HLA-DR^+^cells in PB (Fig. [4b](#F4){ref-type="fig"}). No PB or SF cells expressed the DC differentiation marker CD83 (data not shown). However, SF CD123^+^and CD11c^+^DCs expressed higher levels of cell-surface HLA-DR than the corresponding cells in PB, suggesting some cellular activation within the SF environment \[[@B5],[@B10],[@B44]\]. Thus CD123^+^pDCs comprise a small proportion of RA SF mononuclear cells, which are predominantly immature but show some evidence of activation *in situ*. These observations regarding phenotype and PB and SF numbers are consistent with findings in two recent studies \[[@B39],[@B45]\].
CD123^+^and CD11c^+^SF DCs are efficient APCs
---------------------------------------------
We have previously shown that freshly isolated CD33^bright^CD14^-^CD11c^+^SF DCs efficiently stimulate resting T cells in allogeneic mixed lymphocyte reactions \[[@B21]\]. In contrast, whereas freshly isolated CD11c^+^PB DCs are efficient APCs in mixed lymphocyte reactions, CD123^+^PB DCs usually require prior activation in the presence of IL-3 and CD154 for acquisition of APC function in this assay. To analyse the functional capability of RA SF DCs, CD11c^+^and CD123^+^DCs were sorted from either normal PB or RA SF and incubated with freshly isolated normal allogeneic PB T cells. Freshly isolated PB CD11c^+^but not CD123^+^DCs efficiently stimulated allogeneic T-cell proliferation and IFN-γ and IL-10 production in mixed lymphocyte reactions. Addition of IL-3 made no difference to the T-cell proliferation in response to CD123^+^DCs (data not shown), suggesting that death of the APCs was not responsible. In contrast, both freshly isolated CD11c^+^and CD123^+^SF DCs efficiently stimulated proliferation and IFN-γ and IL-10 production by resting normal allogeneic T cells (Fig. [5](#F5){ref-type="fig"}). A recent study demonstrated the capacity of RA SF to inhibit pDC differentiation *in vitro*\[[@B39]\]. The current studies are consistent, in that SF pDCs showed only some evidence of activation *in situ*, but once incubated in mixed lymphocyte reactions in the absence of SF they displayed enhanced APC function relative to that of PB pDCs. Whereas stimulation of mixed lymphocyte reactions either by CD11c^+^or by CD123^+^PB DCs resulted in little TNF-α production, stimulation by either of these DCs from RA SF resulted in high levels of TNF-α secretion (Fig. [5](#F5){ref-type="fig"}). The data indicate that pDCs have the capacity for enhanced APC function relative to PB pDCs once removed from the RA SF environment. Furthermore, at the time of antigen presentation by SF DCs to T cells, production of a number of cytokines by either T cells or DCs may be stimulated, including TNF-α, and this appears to be a characteristic of RA synovial DCs rather than the subtype of stimulating DCs.
Discussion
==========
Ongoing inflammation in RA involves positive feedback loops between activated T cells, B cells, DCs, macrophages, and their products, with destructive consequences for parenchymal cells. Clinical and animal data indicate that effector-site DCs play an important proinflammatory role in the perpetuation of autoimmune disease and contribute to the lymph-node-like organization of that tissue \[[@B22],[@B46]\]. This role may be effected by local antigen presentation to CD4^+^and CD8^+^effector cells, but DC cytokine and chemokine secretion are also important \[[@B47],[@B48]\]. TNF-α and IL-1β are important downstream proinflammatory and destructive cytokines in RA for somatic cells, whose release is promoted by activation of macrophages. IL-10 is highly expressed in RA, and IFN-γ is an important T-cell effector cytokine \[[@B49],[@B50]\].
In the current studies, we show that, in addition to the previously described population of nuclear RelB^+^DCs, a further population of nuclear RelB^-^CD123^+^pDCs is located in perivascular regions of RA but not normal synovial tissue sublining. Moreover, pDCs were located within blood vessels, and both DC populations were observed in perivascular areas in which cells producing TNF-α were colocated \[[@B35]\]. Adherence of CD123^+^and CD11c^+^DCs to TNF-α-activated endothelium was higher than to resting endothelium *in vitro*(data not shown). TNF-α plays an important role in the recruitment of other leukocytes to RA synovial tissue \[[@B29]\], and this most likely pertains to the recruitment of pDCs to RA but not normal synovial tissue through expression of adhesion molecules such as intercellular adhesion molecule (ICAM)-1, CD62-E, and CD62-P and interaction with their ligands on pDCs \[[@B9],[@B11],[@B51]-[@B54]\]. Furthermore, TNF-α up-regulates synthesis of chemokines by endothelial cells \[[@B55]\]. During experimentally elicited allergic rhinitis, CD123^+^HLA-DR^+^pDCs have been shown to be recruited to human nasal mucosa \[[@B11]\].
The gene for MxA is specifically induced by IFN-α and therefore identifies a population of activated pDCs. In contrast, BDCA2 is a marker of immature pDCs. MxA^+^pDCs have previously been demonstrated in involved lupus skin and inflamed tonsil \[[@B13]\]. In RA synovial tissue, BDCA2 was shown to stain fewer cells than CD123 or MxA, suggesting differentiation *in situ*of a large proportion of pDCs into cells with a capacity for production of IFN-α and other cytokines. Together, the current and previous studies demonstrate recruitment of pDCs to normal lymphoid organs as well as inflammatory sites, with local differentiation, but no recruitment to normal peripheral tissues. In contrast, CD11c^+^myeloid precursors populate normal resting tissues, as shown here, but additional CD11c^+^myeloid cellular recruitment takes place at inflammatory sites, where RelB nuclear translocation takes place \[[@B21],[@B56]\]. We have previously shown that, like synovial pDCs, CD123^+^DCs in the T-cell area of human tonsil are also nuclear RelB^-^\[[@B23]\]. The data suggest either that activation of pDCs is not associated with nuclear translocation and transcriptional activity of RelB or that conditions in tonsil and synovium do not induce sufficient RelB translocation for detection by immunohistochemistry \[[@B26],[@B57]\]. As preliminary studies *in vitro*demonstrate induction of RelB in PB pDCs after stimulation with lipopolysaccharide and CpG, and reduced production of IFN-α by pDCs in RelB-deficient mice, it is likely that RelB activation does accompany pDC activation. However, RelB translocation might be quantitatively reduced or RelB might be more rapidly degraded in the nucleus of pDCs than of myeloid DCs in inflamed tissues \[[@B58]\].
Of relevance to the RA inflammatory lesion, stimulation of blood pDC precursors with signals including CD154, influenza virus, or CpG oligonucleotides induces production of large amounts of cytokines, including IFN-α, IFN-β, and TNF-α; induction of DC differentiation; and stimulation of APC function \[[@B3],[@B39],[@B44],[@B59]\]. Although inhibitory effects of SF on DC function, and thus on T-cell proliferation and cytokine production, are confirmed here \[[@B21],[@B39],[@B60]\], factors in the RA SF environment, such as IL-3 and CD154, may be sufficient to precondition the SF pDCs for efficient APC function *ex vivo*\[[@B36],[@B44]\]. IFN-γ, IL-10, and TNF-α were produced in mixed lymphocyte reactions stimulated by myeloid or pDCs derived from SF but not PB, potentially by DCs or by T cells or both. In the tissue, as a result of antigen presentation by myeloid DCs or pDCs, key effector cytokines may be produced in perivascular areas in RA, located strategically close to endothelial cells, as well as incoming leukocytes. It is not known whether pDCs are capable, like myeloid DCs, of migration from synovial tissue to draining lymph nodes. However, it seems probable that pDCs conditioned by local IL-3 and CD154, or even viral or bacterial products transported to the synovium, predominantly play local proinflammatory and antigen-presenting roles, through secretion of cytokines such as IFN-α and possibly TNF-α \[[@B61],[@B62]\].
Conclusion
==========
pDCs are recruited to RA synovial tissue and comprise an APC population distinct from the previously described nuclear RelB^+^synovial DCs. The APC function of pDCs is greater in SF than in PB. Activated pDCs and interacting T cells may contribute significantly to the inflammatory environment in RA.
Abbreviations
=============
APC = antigen-presenting cell; DC = dendritic cell; ELISA = enzyme-linked immunosorbent assay; FCS = fetal calf serum; FITC = fluorescein isothiocyanate; IFN = interferon; IL = interleukin; NK = natural killer; PB = peripheral blood; pDC = plasmacytoid DC; PE = phycoerythrin; RA = rheumatoid arthritis; SF = synovial fluid; TNF = tumor necrosis factor.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
LC, RT, LF, and PP conceived the experiments and LC, AB, LS, JP, and LF carried them out. LC, RT, and LF wrote the manuscript.
Acknowledgements
================
We thank Malcolm Smith (Repatriation Hospital, Adelaide, South Australia) for providing synovial tissue. This research was supported by grant 210237 from the National Health and Medical Research Council of Australia and by a grant-in-aid from the Princess Alexandra Hospital Foundation. Dr Thomas and Ms Smith were supported by the Arthritis Foundation of Queensland.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Nuclear RelB^-^CD123^+^plasmacytoid dendritic cells (pDCs) are located in close association with cells expressing tumor necrosis factor (TNF)-α in lymphoid aggregates of rheumatoid arthritis (RA) synovial tissue. Sections of frozen normal human synovial tissue **(a, b, e)**or synovial tissue from a patient with untreated active RA **(c, f, g)**or formalin-fixed sections from a patient with active RA **(d, h, i)**were stained with anti-CD11c (brown, **a, c**), anti-CD123 (brown, **b, d**), or anti-TNF-α (brown, **g**) using an immunoperoxidase technique. For double staining, sections were stained with Ulex (red) and anti-CD123 (brown, **e, f, g**) or with RelB (purple) and CD123 (brown, **h, i**) using a double, immunoperoxidase--immunoalkaline phosphatase technique. All sections were counterstained with hematoxylin (blue) except h and i, in which the nucleus of CD123^+^cells appears as a hole. The thick arrow in d identifies a blood vessel. Thin arrows denote representative CD123^+^perivascular DCs (d), representative double-stained vessels (e), a CD123^+^cell within a blood vessel (f), and nuclei stained by RelB (h). Data are representative of at least three separate RA donors in individual experiments. Scale bars represent 20 μm.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
CD123^+^dendritic cells (DCs) and CD11c^+^cells are enriched in rheumatoid arthritis (RA) synovial tissue (ST). CD123^+^DCs were quantitated by counting the number of CD123^+^Ulex^-^cells in sections double-stained with CD123 and Ulex. Cells were counted in each of the entire sections from three patients and three normal controls at high power, and for each biopsy this number was corrected for the area of the section to obtain the number/mm^2^. To quantitate CD11c^+^cellular infiltration, the number of CD11c^+^cells was counted in three high-power fields of the synovial sublining in sections from three patients and three normal controls. Data represent means ± standard error of the mean.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
CD11c^+^dendritic cells (DCs) and CD123^+^DCs are present in rheumatoid arthritis (RA) synovial fluid (SF). After purification of mononuclear cells from either normal or RA peripheral blood (PB) or RA SF, cells were stained with CD123-PE, CD11c FITC, CD14-PECy5, and HLA-DR-APC. Live HLA-DR^+^CD14^-^mononuclear cells were gated; polymorphonuclear cells were excluded on the basis of forward and side light-scatter. Scatter plots depict percentage of total PB or SF cells expressing the indicated markers for all donors. APC, antigen-presenting cell; PBMC, peripheral blood mononuclear cells; PE, phycoerythrin; SFMC, synovial fluid mononuclear cells.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Morphology and cell-surface phenotype of dendritic cell (DC) subpopulations. **(a)**Freshly isolated CD11c^+^CD123^-^DCs and CD123^+^CD11c^-^DCs were sorted from either normal peripheral blood (PB) or rheumatoid arthritis (RA) synovial fluid (SF) according to the gating strategy outlined in the legend to Fig. 3 and were then prepared for transmission electron microscopy. Micrographs are representative of three separate donors. Scale bars represent 20 μm. **(b)**Normal PB or RA SF mononuclear cells were stained with CD123-PE, CD14-PECy5, HLA-DR-APC, and either isotype control or the depicted marker labelled with FITC. Live HLA-DR^+^CD14^-^CD123^+^cells or HLA-DR^+^CD14^-^CD11c^+^cells were gated; polymorphonuclear cells were excluded on the basis of forward and side light-scatter. The expression of isotype control (dotted lines) and indicated markers (continuous lines) by the gated cells are depicted. Data are from three donors of normal PB or RA SF. APC, antigen-presenting cell; PE, phycoerythrin.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
CD123^+^and CD11c^+^synovial fluid (SF) dendritic cells (DCs) are efficient antigen-presenting cells (APCs) and induce secretion of tumor necrosis factor α. Freshly sorted CD123^+^CD11c^-^or CD11c^+^CD123^-^DCs from either normal peripheral blood (PB) **(a--d)**or rheumatoid arthritis (RA) SF **(e--h)**were incubated with 10^5^purified normal allogeneic T cells. T-cell proliferation was measured by \[^3^H\]thymidine incorporation of triplicate wells after 60 hours (a, e). Unstimulated T-cell proliferation was routinely \<500 cpm. Data represent means ± standard error of the mean of triplicate wells and are from three individual PB and SF donors. The concentrations of IFNγ, IL-10, and tumor necrosis factor (TNF)-α were measured in supernatants of allogeneic mixed lymphocyte reactions stimulated by sorted PB **(b--d)**or SF DCs **(f--h)**. Data are means of duplicate samples and are from two separate SF donors.
:::
:::
|
PubMed Central
|
2024-06-05T03:55:54.917791
|
2005-1-11
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065313/",
"journal": "Arthritis Res Ther. 2005 Jan 11; 7(2):R230-R240",
"authors": [
{
"first": "Lois L",
"last": "Cavanagh"
},
{
"first": "Amanda",
"last": "Boyce"
},
{
"first": "Louise",
"last": "Smith"
},
{
"first": "Jagadish",
"last": "Padmanabha"
},
{
"first": "Luis",
"last": "Filgueira"
},
{
"first": "Peter",
"last": "Pietschmann"
},
{
"first": "Ranjeny",
"last": "Thomas"
}
]
}
|
PMC1065314
|
Introduction
============
Systemic sclerosis (SSc), also known as scleroderma, is a connective tissue disease of unknown aetiology. Possibly an autoimmune disorder, it is accompanied in the vast majority of cases by the presence of antinuclear antibodies \[[@B1]\]. SSc may affect virtually any organ of the body, including skin, gastrointestinal tract, lungs, heart, kidneys, and musculoskeletal system. Altered connective tissue metabolism can cause either localized or diffuse thickening of the skin, while inflammation is associated with endothelial damage. Clinically, microvascular disturbance, teleangiectasia, Raynaud\'s phenomenon, polyarthralgia and polyarthritis, as well as oesophageal hypomobility, visceral muscolaris mucosa damage and pulmonary fibrosis, have been described \[[@B2]\].
The mechanisms leading to endothelial damage, inflammation and fibrosis are unclear. Reactive oxygen species in neutrophils may increase the extent of inflammation and fibrosis during the respiratory burst and could be involved in endothelial damage \[[@B3]\]. The endothelial cells of microvessels are deficient in the synthesis of catalase, which provides natural defence against superoxide damage, and appear to be particularly susceptible to superoxide injury during reperfusion \[[@B4]\].
Adenosine is an important endogenous regulator of neutrophil functioning. It is released intracellularly and modulates neutrophil activity by interacting with specific surface receptors \[[@B5]\]. Distinct adenosine receptor (AR) subtypes A~1~, A~2A~, A~2B~and A~3~have been identified and their functions characterized in neutrophils. Specifically, activation of A~1~ARs enhances chemotaxis, phagocytosis and adherence \[[@B6],[@B7]\]; A~2A~ARs inhibit reactive oxygen species generation, phagocytosis and adherence \[[@B8]-[@B10]\]; and A~2A~and A~3~ARs inhibit neutrophil degranulation \[[@B11]-[@B14]\]. Adenosine has been shown to prevent the release of vascular endothelial growth factor from neutrophils via A~2B~AR activation \[[@B15]\]. Because activation of ARs reduces both immune and inflammatory responses, adenosine release has been hypothesized to be a possible mechanism of cell self-protection from activated neutrophils \[[@B5]\]. An increase in adenosine deaminase activity has been described in patients with SSc, suggesting an alteration in adenosine control mechanisms in this disease \[[@B16],[@B17]\].
In the present study we analyzed A~2A~and A~2B~AR subtypes in neutrophils from patients affected by SSc by means of expression analysis, radioligand binding assays and functional studies.
Methods
=======
Chemicals and reagents
----------------------
Bacitracine, benzamidine, trypsin inhibitor, sodium orthovanadate, Nonidet P-40, SDS, phenylsulfonylfluoride, aprotinin and adenosine deaminase (ADA) were purchased from Sigma (St. Louis, MO, USA). Unlabelled AR agonists/antagonists and the anti-β-actin antibody were supplied by RBI/Sigma (St. Louis, MO, USA). \[^3^H\]CGS~21680~(CGS~21680~= \[2-p-(2-carbowyethyl)phenylethylamino\]-5\'-N-ethylcarboxamidoadenosine), \[^3^H\]NECA (NECA = 5\'-N-ethylcarboxamidoadenosine), and \[^32^P\]α-ATP were supplied by NEN Life Sciences (Köln, Germany). Electrophoresis reagents were purchased from BioRad (Munchen, Germany). A~2A~AR and A~2B~AR antibodies were supplied by Alpha Diagnostic (San Antonio, TX, USA). All other chemicals were from standard commercial sources.
Patients
--------
Twenty-six patients affected by SSc were included in the study (22 women and 4 men; mean age ± standard deviation 53.0 ± 11.3 years). They all fulfilled standard criteria of the American College of Rheumatology for SSc. Sixteen patients were anticentromere antibody positive and four were SCL-70 positive. Limited symptoms of disease, involving skin thickness alterations to the face, hands and feet, were present in 18 patients (mean disease duration \<5 years, skin score range \[according to the modified Rodnan total skin thickness score\] 10--21). Diffuse symptoms with more extensive skin involvement were present in eight patients (mean disease duration \<5 years, total skin thickness score range 27--30). The activity score \[[@B18]\] varied between 0.5 and 3.5 and the severity score \[[@B19]\] between 2 and 6. The erythrocyte sedimentation rate was 24 ± 23 mm/hour (mean ± standard deviation).
Control samples were obtained from 26 healthy volunteers, who were similar to the patients included in the study in terms of sex distribution and age (20 women and 6 men; mean age ± standard deviation 49.0 ± 9.2 years). Informed consent to participate in the study was obtained from all individuals.
Sample collection and neutrophil preparation
--------------------------------------------
Venous blood (20 ml) was drawn between 08:00 and 09:00 a.m. from fasting individuals by antecubital venipuncture, collected in heparinized (10 IU/L) plastic tubes and processed immediately. Neutrophils were isolated following the Boyum method \[[@B20]\] with some modifications.
Western blot analysis
---------------------
Neutrophils were lysed in RIPA buffer (150 mmol/l NaCl, 50 mmol/l Tris-HCl, pH 8, 0.5% sodium deoxhycolate, 1% Nonidet P-40, 1 mmol/l phenylsulfonylfluoride, 10 μg/ml aprotinin, 100 μmol/l sodium orthovanadate) for 1 hour at 4°C. After centrifugation at 15,000 *g*for 30 min, soluble fractions were assayed for protein content using BioRad protein assay. Equivalent amounts of proteins (50 μg/sample) were analyzed by SDS-PAGE, using 10% (weight/vol) polyacrylamide resolving gels. Protein bands were transferred to nitrocellulose and probed with 0.1 μg/ml rabbit anti-human A~2A~AR or A~2B~AR antibodies.
A~2A~AR antibody is an affinity-purified rabbit polyclonal antibody raised against a peptide mapping to the carboxyl-terminus of A~2A~AR. It specifically reacts with human, bovine, rat and pig A~2A~receptors and does not cross-react with A~1~, A~2B~, or A~3~AR subtypes. A~2B~AR antibody is an affinity-purified rabbit polyclonal antibody raised against a region that corresponds to the second extracellular domain of A~2B~AR of human origin.
After washing, membranes were incubated with anti-rabbit secondary antibody conjugated to horseradish peroxidase for 2 hours at room temperature, and bands were visualized by chemiluminescence, in accordance with the manufacturer\'s instructions (Sigma-Aldrich). Membranes were re-probed with an anti-β-actin antibody for normalization.
Binding assay
-------------
For membrane preparation, cells were washed twice with 10 mmol/l Tris-HCl buffer, pH 7.4, containing 10 mmol/l MgCl~2~, in the presence of protease inhibitors (200 μg/ml bacitracine, 160 μg/ml benzamidine, 20 μg/ml trypsin inhibitor \[T1\]) and centrifuged at 48,000 *g*for 15 min at 4°C. Pellets were diluted in 20 volumes of T1 buffer, treated with ADA (2 IU/ml) for 60 min at 37°C to remove endogenous adenosine, and washed twice with 50 mmol/l Tris-HCl buffer, pH 7.4, containing 10 mmol/l MgCl~2~(T2).
A~2A~AR binding assay was performed by using a specific radiolabelled A~2A~AR agonist, namely \[^3^H\]CGS~21680~. Aliquots of neutrophil membranes (0.2--0.3 mg protein) were incubated with different \[^3^H\]CGS~21680~concentrations (5--30 nmol/l) in a final volume of 250 μl of T2 buffer. Nonspecific binding was determined in the presence of 100 μmol/l NECA. After 90 min incubation at 25°C, the binding reaction was terminated by vacuum filtration through Whatman GF/C glass fibre filters (Whatman, Maidstone, UK), accompanied by three washes with ice-cold T2 buffer (4 ml). A~2A~AR specificity was evaluated through competition experiments, using different AR ligands.
A~2B~AR binding assay was performed using 20 nmol/l \[^3^H\]NECA in the presence of 50 nmol/l cyclopentyladenosine (CPA) and 100 nmol/l SCH~58261~(SCH~58261~= 5-amino-7-\[phenylethyl\]-2-\[2-furyl\]-pyrazolo \[4,3-e\]-1,2,4-triazolo \[1,5-c\]pyrimidine) to prevent \[^3^H\]NECA binding to A~1~and A~2A~ARs, respectively \[[@B21]\]. Scatchard analysis was performed on competition experiments carried out in the presence of unlabelled NECA at concentrations ranging from 50 nmol/l to 2 mmol/l. Aliquots of neutrophil membranes (0.2--0.4 mg proteins) were incubated in a final volume of 250 μl T2 buffer. Nonspecific binding was evaluated in the presence of 100 μmol/l NECA. After 90 min incubation at 0°C, the reaction was terminated either by vacuum filtration through Whatman GF/C glass fibre filters, accompanied by three washes with ice-cold T2 buffer (4 ml), or by centrifugation at 2900 *g*for 15 min at 4°C. A~2B~AR specificity was evaluated through competition experiments, using different AR ligands.
Adenylyl cyclase assay
----------------------
Neutrophils were homogenized in buffer solution containing 10 mmol/l Hepes, 1 mmol/l EGTA and 10 mmol/l NaCl~2~, and then centrifuged at 46,500 *g*for 20 min at 4°C. Pellets were resuspended in 10 volumes of 10 mmol/l Hepes, containing protease inhibitors (200 μg/ml bacitracine and 160 μg/ml benzamidine), incubated for 30 min at 30°C with 2 U/ml ADA, and centrifuged. Adenylyl cyclase (AC) activity was measured as described by Salomon \[[@B22]\] and Johnson and Salomon \[[@B23]\], with some modifications. NECA-mediated stimulation of AC activity was assessed by incubating aliquots of membranes with increasing NECA concentrations from 0.01 nmol/l to 10 μmol/l. The reaction was started by adding membrane aliquots (10--50 μg proteins/tube), conducted for 15 min at 24°C, and then stopped by transferring samples on ice and adding 500 μl ice-cold stop solution (120 mmol/l zinc acetate, 144 mmol/l Na~2~CO~3~). The stop solution contained \[^3^H\]cAMP (10,000--15,000 cpm/sample) to monitor column recovery. Newly formed ZnCO~3~allowed precipitation of residual ATP, discarded through centrifugation at 2700 *g*for 8 min. Supernatants containing both \[^32^P\]α-cAMP and \[^3^H\]cAMP were further purified by double-step Dowex-Alumina chromatography and counted by means of a β-counter (Packard Tricarb 1600; Perkin Elmer, Wellesley, MA, USA).
To evaluate A~2B~AR mediated cAMP accumulation, the reaction was carried out in the presence of selective A~2A~antagonist SCH~58261~at a concentration (100 nmol/l) able to block A~2A~receptors completely \[[@B21]\].
Data and statistical analysis
-----------------------------
Affinity constant values (Kd) and maximum number of binding sites (B~max~) were calculated using the nonlinear multipurpose curve-fitting computer program Graph-Pad Prism The 50% inhibitory concentration values were calculated using the same program and converted to Ki values through the Cheng and Prusoff equation.
A GS-670-BIO-RAD imaging densitometer was used for semiquantitative analysis of immunoblots. Partial F test (*P*\< 0.01) was used to determine binding data with the best fit to a one-site or two-site model. Differences in binding parameters between SSc patients and control individuals were evaluated by one-way analysis of variance.
Results
=======
In both control and SSc neutrophils, Western blot analysis identified two specific immunoreactive bands of 45 kDa and 50 kDa, corresponding to A~2A~and A~2B~ARs, respectively (Fig. [1](#F1){ref-type="fig"}). This confirmed the presence of both AR subtypes in human neutrophils.
To characterize ARs, binding assays were conducted in neutrophil membrane fractions. SSc patients were randomly divided into two subgroups in order to obtain large amounts of protein, as required by the experiments.
The selective A~2A~AR agonist \[^3^H\]CGS~21680~identified a homogenous population of binding sites in control individuals. Kd and B~max~values were 25 ± 1.3 nmol/l and 35 ± 2.4 fmol/mg protein, respectively (Fig. [2](#F2){ref-type="fig"}). Competition experiments using \[^3^H\]CGS~21680~in combination with a variety of A~2A~ligands revealed a pharmacological profile typical for A~2A~ARs (R-PIA \[R-N6-phenylisopropyladenosine\] \> teofilline \> NECA \> SCH~58261~; data not shown). Scatchard analysis for SSc neutrophils revealed no significant differences in Kd and B~max~between patients (mean values: Kd = 23 ± 1.8 nmol/l, B~max~= 40 ± 3.2 fmol/mg protein) and healthy control individuals (*P*\> 0.05; Fig. [2](#F2){ref-type="fig"}), suggesting that no alteration in A~2A~binding sites occurs in SSc. In agreement with this, densitometric analysis of immunoblots showed no significant changes in A~2A~AR immunoreactive bands in SSc neutrophils relative to controls (optical density: 0.11 ± 0.03 for patients versus 0.15 ± 0.02 for controls).
A~2B~AR binding sites were identified using \[^3^H\]NECA as radioligand in the presence of 50 nmol/l CPA and 100 nmol/l SCH~58261~, to prevent nonspecific binding to A~1~and A~2A~AR subtypes. We performed competition experiments using a wide range (50 nmol/l to 2 mmol/l) of \[^3^H\]NECA concentrations to allow the identification of A~2B~AR low-affinity binding sites. Data analysis revealed that the one-site model produced a significantly better fit than the two-site model (*P*\< 0.05), both in control and SSc neutrophils. In our experimental conditions, control neutrophils exhibited the presence of low-affinity binding sites with Kd and B~max~values of 476 ± 34 nmol/l and 3696 ± 210 fmol/mg, respectively (Fig. [3](#F3){ref-type="fig"}). Competition experiments using \[^3^H\]NECA in combination with a variety of AR ligands revealed a pharmacological profile typical for A~2B~ARs (R-PIA \> teofilline \> SCH~58261~= MRS1220 \> DPCPX \> 2Cl-adenosine \> NECA \> MRS1706; Table [1](#T1){ref-type="table"}). Scatchard analysis for SSc neutrophils showed no significant differences in Kd and B~max~between the two subgroups of patients. However, a significant alteration in B~max~was found relative to controls, whereas Kd values remained unaltered. Overall, mean values for Kd and B~max~in SSc were 469 ± 35 nmol/l and 1292 ± 98 fmol/mg protein, respectively (*P*\< 0.05; Fig. [3](#F3){ref-type="fig"}). Moreover, experiments conducted in individual patients using a concentration of NECA of 500 nmol/l showed similar specific binding values (expressed as fmol/mg protein), confirming the homogeneity of A~2B~AR sites between SSc subgroups (Fig. [4](#F4){ref-type="fig"}). The alteration in A~2B~AR levels in SSc patients was confirmed by immunoblotting assay. Densitometric analysis of immunoreactive bands showed a reduction in A~2B~expression in SSc patients (optical density 0.22 ± 0.04) as compared with controls (optical density 0.40 ± 0.06; *P*\< 0.05; Fig. [1](#F1){ref-type="fig"}).
Functional coupling of A~2B~ARs to stimulatory G proteins in neutrophil membranes was assessed by evaluating the effects of the agonist NECA (in the presence of 100 nmol/l SCH~58261~) on AC activity. NECA stimulated AC activity in a concentration dependent manner. Dose-response curves revealed significant differences between SSc patients (EC~50~= 373 ± 26 nmol/l; E~max~= 35 ± 2.9%) and controls (EC~50~= 165 ± 9.3 nmol/l; E~max~= 43 ± 3.2%), suggesting an alteration in A~2B~AR responsiveness in SSc (Fig. [5](#F5){ref-type="fig"}).
Discussion
==========
In the present study we analyzed A~2A~and A~2B~AR subtypes in neutrophils of patients affected by SSc, by means of Western blot, radioligand binding techniques and functional studies. This is the first report of use of Western blot analysis to confirm the presence of A~2A~and A~2B~ARs in human neutrophils.
A~2A~and A~2B~AR equilibrium binding parameters were measured using radioligand binding assays. Scatchard analysis of \[^3^H\]CGS~21680~saturation binding to A~2A~AR showed no significant difference in B~max~or Kd between SSc neutrophils and controls, suggesting that the A~2A~AR subtype remained unaltered in SSc. Conversely, when A~2B~AR was analyzed a reduction in B~max~(65%) was observed, with no significant change in Kd values.
A~2B~ARs are known to be low-affinity adenosine binding sites. Competition experiments using a variety of A~2B~AR agonists and antagonists revealed a pharmacological profile typical of A~2B~ARs, which is consistent with studies conducted in transfected cell models. Our findings represent the first characterization of A~2B~ARs in neutrophils with binding experiments.
In order to analyze a population of nonhomogenous patients and to evaluate the impact of the disease on A~2~ARs, SSc patients were randomly divided into two subgroups. No difference was found when the two groups were compared, suggesting that different degrees of disease severity and activity had no impact on the assays, but that the disease *per se*is required to modulate levels and functioning of A~2B~receptors.
Functional studies were performed to investigate whether the decrease in level of A~2B~ARs was accompanied by alterations in receptor responsiveness. An evaluation of the ability of NECA to increase AC activity revealed functional coupling of A~2B~receptors to G proteins. In SSc patients a significant reduction (by more than 50%) in NECA potency was observed, without any effect on agonist efficacy.
Our findings suggest that a selective reduction in A~2B~AR levels and responsiveness occurred in SSc. Alterations in the expression and functionality of A~2B~ARs (low-affinity ARs) in patients with SSc may be responsible for the increase in free oxygen radicals, and consequent oxidative damage, that characterizes SSc. This would account for impaired control of hypoxic and inflammatory processes.
In neutrophils it has long been known that adenosine and its analogues inhibit O~2~^-^generation, phagocytosis and cell adherence by occupying specific A~2~ARs. Because hypoxia, ischaemia and inflammation can stimulate adenosine production, A~2~AR regulation has been postulated to be a self-protective mechanism for cells from activated neutrophils \[[@B24]\]. Eltzschig and coworkers \[[@B25]\] reported that A~2B~ARs are selectively upregulated in endothelial cells by hypoxia (more than fivefold increase in mRNA), which is associated with ATP hydrolysis and release of adenosine. Taken together, these findings show some coordination between AR transcription and nucleoside signalling at the vascular interface during hypoxia. We might speculate that chronic inflammatory conditions in SSc patients impaired regulatory mechanisms mediated by the anti-inflammatory effects of adenosine via A~2B~AR activation. In addition, it was reported by Visser and coworkers \[[@B26]\] that increases in cAMP in activated neutrophils play an anti-inflammatory role. The reduced activation of cAMP we observed in SSc patients might be correlated with the inability of these patients to control the inflammatory process.
It was no surprise to find an alteration in adenosinergic system responsiveness in SSc. In fact, adenosine produces a constellation of responses, including anti-inflammatory actions and vasodilatation, mediated through interactions with high-affinity receptor subtype A~2A~and low-affinity receptor subtype A~2B~. Moreover, in SSc and related disorders, alterations in adenosine metabolism have been suggested. Indeed, purine analogue 2-chlorodeoxyadenosine, which is utilized for the treatment of such chronic disorders \[[@B27],[@B28]\], appears to reduce the number of abnormal fibroblasts.
A~2B~ARs were initially thought to be of lesser physiological relevance because of their relatively low affinity for adenosine, and it was only recently that important functions attributable to A~2B~ARs were discovered. A pivotal role for them was postulated in inflammatory pathological conditions, when adenosine is released at high levels (up to the micromolar range). In light of our findings, a closer examination of A~2B~AR functions may be valuable because of the potential therapeutic importance of these receptors as targets for treatment with selective agents.
Conclusion
==========
Our findings demonstrated a reduction in A~2~low-affinity (A~2B~) AR density and functioning in neutrophils of patients affected by SSc, suggesting an alteration in adenosinergic system responsiveness. This reduction could relate to the increased production of free oxygen radicals and consequent oxidative damage that characterize SSc, highlighting an impairment in the ability of neutrophils to control hypoxia and inflammation.
No differences between two randomly selected subgroups of SSc patients were found, thus suggesting that different degrees of disease severity and activity had no impact on the degree of A~2B~AR reduction. Consequently, the functional status of A~2B~ARs may be considered a marker of the disease, making it worthwhile to characterize a larger cohort of patients, including their closest relatives and patients with early SSc.
Abbreviations
=============
AC = adenylyl cyclase; ADA = adenosine deaminase; AR = adenosine receptor; B~max~= maximum number of binding sites; CGS~21680~= (2-p-\[2-carbowyethyl\]pheylethylamino)-5\'N-ethylcarboxamidoadenosine; CPA = cyclopentyladenosine; Kd = affinity constant; NECA = 5\'-N-ethylcarboxamidoadenosine; R-PIA = R-N6-phenylisopropyladenosine; SCH~58261~= 5-amino-7-(phenylethyl)-2-(2-furyl)-pyrazolo(4,3-e)-1,2,4-triazolo(1,5-c)pyrimidine; SSc = systemic sclerosis.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
LB organized the study design and recruited the patients. LT carried out the binding experiments and statistical analysis. AR participated in the immunoblotting experiments and helped to draft the manuscript. FdF participated in the collection of human samples. AL participated in the coordination of the study and helped with problem solving. SB participated in the coordination of the study and in planning the manuscript. CM participated in the coordination of the study and designed the AC assay. All authors read and approved the final manuscript.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Immunoblotting analysis of A~2A~and A~2B~adenosine receptors (ARs) from systemic sclerosis (SSc) neutrophils and controls. Cells obtained from 26 healthy volunteers and 26 SSc patients were lysed as described in the Methods section. Equal amounts of protein (50 μg) were separated on polyacrylamide gel, blotted and probed with 0.1 μg/ml rabbit anti-human A~2A~AR or A~2B~AR antibodies. Immunoreactive bands were visualized according to electrogenerated chemiluminescence protocol. A~2A~and A~2B~AR antibodies recognized immunoreactive bands of 45 kDa and 50 kDa, respectively. **(a)**Representative experiment performed on neutrophils from one healthy volunteer and one SSc patient. **(b)**Densitometric analysis of A~2A~and A~2B~AR immunoreactive bands from 26 healthy volunteers and 26 SSc patients. Graph bars: mean ± standard error of band density, normalized to β-actin. White bars are controls; grey bars are SSc patients.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Representative Scatchard plot of \[^3^H\]CGS~21680~saturation binding data. Empty circles indicate neutrophil membranes from healthy volunteers (affinity constant \[Kd\] = 25 ± 1.3 nmol/l; maximum number of binding sites \[B~max~\] = 35 ± 2.4 fmol/mg); filled circles indicate neutrophil membranes from systemic sclerosis (SSc) patients overall (Kd = 23 ± 1.8 nmol/l; B~max~= 40 ± 3.2 fmol/mg). Assays were performed in triplicate.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Representative Scatchard plot of \[^3^H\]NECA saturation binding data. Competition binding experiments were performed, incubating aliquots of neutrophil membranes with 20 nmol/l \[^3^H\]NECA and different NECA concentrations (50 nmol/l to 2 mmol/l), in the presence of 50 nmol/l CPA and 100 nmol/l SCH~58261~. Empty circles indicate neutrophil membranes from healthy volunteers (affinity constant \[Kd\] = 476 ± 34 nmol/l, maximum number of binding sites \[B~max~\] = 3696 ± 210 fmol/mg); filled circles indicate neutrophil membranes from systemic sclerosis (SSc) patients overall (Kd = 469 ± 35 nmol/l, B~max~= 1292 ± 98 fmol/mg). Assays were performed in triplicate.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
A~2B~adenosine receptor binding experiments performed in individual patients using NECA at 500 nmol/l concentration. Neutrophils were obtained from healthy volunteers (*n*= 26) and systemic sclerosis (SSc) patients (*n*= 26). Horizontal lines indicate the mean values.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
A~2B~adenosine receptor (AR)-mediated stimulation of adenylyl cyclase activity in control (empty circles) and systemic sclerosis (SSc; filled circles) neutrophil membranes. Membranes were incubated with different NECA concentrations (ranging from 10 nmol/l to 100 μmol/l) and the activity of adenylyl cyclase, expressed as pmol/min per mg protein, was evaluated. Values are expressed as mean ± standard error of three indipendent experiments. EC~50~values were 165 ± 9.3 for control versus 373 ± 26 nmol/l for SSc.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Specificity of \[^3^H\]NECA binding to A~2B~adenosine receptors in control neutrophil membranes
:::
\[^3^H\]NECA Ki (μmol/l)
---------------- --------------------------
NECA 0.315 ± 0.028
2 Cl-adenosine 0.954 ± 0.600
R-PIA 1000 ± 86
SCH~58261~ \>10
Teofilline 47 ± 3.5
MRS1706 0.005 ± 0.0003
DPCPX 2 ± 0.12
MRS1220 \>10
Competition experiments were performed, incubating aliquots of neutrophil membranes with 20 nmol/l \[^3^H\]NECA (plus 50 nmol/l CPA and 100 nmol/l SCH~58261~) in the presence of increasing ligand concentrations. Ki values are expressed as mean ± SEM of three separate experiments. Ki values were calculated from IC~50~values (concentration of drug causing 50% inhibition of specific binding) using the Cheng and Prusoff equation.
:::
|
PubMed Central
|
2024-06-05T03:55:54.920230
|
2004-12-10
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065314/",
"journal": "Arthritis Res Ther. 2005 Dec 10; 7(2):R189-R195",
"authors": [
{
"first": "Laura",
"last": "Bazzichi"
},
{
"first": "Letizia",
"last": "Trincavelli"
},
{
"first": "Alessandra",
"last": "Rossi"
},
{
"first": "Francesca",
"last": "De Feo"
},
{
"first": "Antonio",
"last": "Lucacchini"
},
{
"first": "Stefano",
"last": "Bombardieri"
},
{
"first": "Claudia",
"last": "Martini"
}
]
}
|
PMC1065315
|
Introduction
============
The completion of the human and mouse genome sequencing programs and the subsequent annotation of previously unidentified genes have opened a new epoch in biology and biomedical sciences. The genetic information greatly facilitated the discovery of novel disease-related genes and the mapping of signature genes for early diagnosis. More specifically, polynucleotide or oligonucleotide arrays have been applied in both human and experimentally induced disease conditions to determine characteristic expression patterns of signature genes.
In an inflammatory disease such as rheumatoid arthritis (RA), the gene expression profile is extremely complex owing to the diversity of cell types involved in the pathology and the polygenic character of the autoimmune disease \[[@B1]-[@B5]\]. The overall picture of molecular interactions in an inflamed joint, deduced from gene expression studies in both RA and its corresponding animal models, involves proteins participating in immunity, inflammation, apoptosis, proliferation, cellular transformation and cell differentiation, and other processes \[[@B3]-[@B8]\]. Several studies analyzed the patterns of gene expression in peripheral blood or synovial fluid mononuclear cells, and in the inflamed synovium of human patients \[[@B1],[@B3]-[@B5],[@B7],[@B9]-[@B11]\]. However, the genetic heterogeneity of the human population is a serious obstacle to the correct interpretation of data in gene expression studies. Animal models of RA can facilitate the interpretation of genome-wide gene expression by providing genetic and clinical homogeneity, and an opportunity to monitor the onset and progression of the disease \[[@B12]-[@B20]\]. DNA microarray technology was successfully applied to inflamed paws of mice or rats systemically immunized with arthritogenic compounds to induce arthritis \[[@B6],[@B21]-[@B23]\]. Despite the usefulness of the information provided by these studies, the early gene expression events at the site of inflammation (joint and synovium) and the mechanisms of disease initiation remain unknown.
Systemic immunization of genetically susceptible BALB/c mice with human cartilage proteoglycan aggrecan (PG) induces PG-specific immune responses that then trigger inflammation in peripheral joints \[[@B13],[@B19]\]. PG-induced arthritis (PGIA) is a murine model which bears many similarities to RA as indicated by clinical assessments, radiographic analyses, various laboratory and functional tests, and by histopathologic studies of diarthrodial joints \[[@B13],[@B19],[@B24],[@B25]\]. Moreover, genome-wide screening studies identified multiple genomic loci in PGIA \[[@B20],[@B26]-[@B29]\] that are syntenic with those described in RA \[[@B25]\]. Both RA and PGIA are polygenic autoimmune diseases with a major permissive role of the MHC, although non-MHC genes account for a significant portion of the genetic susceptibility. PGIA can be successfully transferred into naive BALB/c or syngeneic severe combined immunodeficient (SCID) mice either with unseparated spleen cells or with antigen (PG)-stimulated T lymphocytes from arthritic donor BALB/c mice \[[@B30]-[@B32]\].
In the present study, we adoptively transferred the disease (PGIA) into syngeneic BALB/c^SCID^mice lacking functional T and B cells. SCID mice carry a natural mutation that prevents the V(D)J recombination in B and T lymphocytes, resulting in a failure to generate functional immunoglobulins and T cell receptors \[[@B33],[@B34]\]. Consequently, adoptively transferred arthritis in BALB/c^SCID^mice is an ideal model in which activated lymphocytes of arthritic donor BALB/c mice migrate and interact with the intact innate immunity environment in the joints of BALB/c^SCID^mice. The gene expression profiles in normal, pre-arthritic and arthritic joints of the recipient BALB/c^SCID^mice were determined by using DNA microarray technology (Affymetrix). Although a significant number of genes were differentially expressed in joints with acute and chronic arthritis, in this study we focused on early genes whose expression occurred before the onset of clinical symptoms.
Methods
=======
Animals, antigen and immunization
---------------------------------
The use of human cartilage from joint replacement surgeries for antigen isolation was approved by the Institutional Review Board, and all animal experiments were approved by the Institutional Animal Care and Use Committee. Female BALB/c mice at the age of 24--26 weeks (National Cancer Institute, Kingston Colony, New York, USA) were injected intraperitoneally with 100 μg of cartilage PG (measured as protein) emulsified in dimethyldioctadecylammonium bromide (DDA) adjuvant (Sigma-Aldrich, St Louis, Missouri, USA). The use of adjuvant DDA allowed us to avoid the harmful effects of oil and bacterial proteins present in Freund\'s adjuvants \[[@B35],[@B36]\]. Booster injections of the same doses of PG with DDA were given on days 21 and 42. BALB/c mice develop swelling and redness of one or more limbs 7--10 days after the second or third injection with PG in adjuvant \[[@B25]\]. Arthritis was assessed daily, and inflammation was scored from grade 0 to grade 4 for each paw \[[@B13],[@B36],[@B37]\]. Female SCID mice of the BALB/c background (NCI/NCrC.B-17-scid/scid; henceforth BALB/c^SCID^) were used for adoptive cell transfer. BALB/c^SCID^mice were purchased from the National Cancer Institute and maintained under germ-free conditions.
Stimulation of lymphocytes *in vitro*, and adoptive transfer of arthritis
-------------------------------------------------------------------------
To ensure uniformity and reproducibility of disease transfer, donor spleen cells were isolated from arthritic BALB/c mice within 1--2 weeks after the onset of inflammation. At least two paws of donor BALB/c mice were arthritic, and the cumulative inflammation score (for four paws) was in the range 5--8. Spleen cells of arthritic BALB/c mice were collected and cultured in six-well plates (2.5 × 10^6^cells/ml) with cartilage PG (50 μg/ml) for 4 days in Dulbecco\'s modified Eagle\'s medium supplemented with 5% fetal bovine serum (HyClone Laboratories, Logan, Utah, USA). After stimulation *in vitro*for 4 days with cartilage PG, non-adherent cells were collected, and live cells (lymphocytes) were separated on Lympholyte-M (Cedarlane, Ontario, Canada). Finally, 2 × 10^7^lymphocytes were injected intraperitoneally on days 0 and 7 into recipient BALB/c^SCID^mice as described \[[@B32]\].
A standard scoring system used for primary arthritis was applied to the assessment of disease severity in BALB/c^SCID^mice \[[@B24],[@B37]\]. Typically, one to four paws became inflamed simultaneously 3--5 days after the second cell transfer, and the rest of the peripheral joints became inflamed within 2--4 days after the onset of the first symptoms. BALB/c^SCID^mice were scored twice daily, and were killed as soon as the inflamed paw reached an individual arthritis score of 2, but not later than 24 hours after the onset of arthritis. This paw was designated as acute arthritic (AA), and contralateral or ipsilateral paws that were not inflamed at that time were used as pre-arthritic (PA) samples. The PA joints did not show evidence of inflammation on histopathological examination, although thickening of the synovial lining in small joints was observed occasionally (data not shown). Several arthritic BALB/c^SCID^mice were scored daily and were killed 8--10 days after disease onset. These joint samples represented subacute-chronically arthritic (CA) samples. In addition to PA, AA and CA experimental conditions, paws of naive non-immunized BALB/c^SCID^mice were used as \'absolutely negative\' (control naive; AN) samples for RNA isolation and subsequent hybridization. Each sample represented RNA pooled from four paws of two mice.
Probe preparation
-----------------
Synthesis and biotinylation of cRNA and hybridization were performed in accordance with the manufacturer\'s instructions (Affymetrix, Santa Clara, California, USA). In brief, total RNA was isolated from normal or inflamed paws of mice by using TRIzol reagent (Invitrogen, Gaithersburg, Maryland, USA) with additional purification on RNeasy columns (Qiagen, Valencia, California, USA). RNA quality was confirmed by spectrophotometry and electrophoresis on formaldehyde gels \[[@B38]\]. Double-stranded complementary DNA was synthesized with the T7-dT24 primer incorporating a T7 RNA polymerase promoter. Biotinylated cRNA was prepared with the Enzo BioArray High Yield RNA Transcript Labeling Kit (Enzo Diagnostics, Inc., Farmingdale, New York, USA) and hybridized to the murine genome Affymetrix U74v2 chip set, which included three DNA chips, MG\_U74Av2, MG\_U74Bv2 and MG\_U74Cv2, interrogating more than 36,000 genes that represented essentially the entire mouse genome \[[@B39]-[@B42]\]. Fluorescent hybridization signals were developed with phycoerythrin-conjugated streptavidin and were further enhanced with fluorescently labeled anti-streptavidin antibodies. DNA chips were scanned to obtain quantitative gene expression levels. DNA chip hybridization, Fluidics Station operations, scanning, and preliminary data management were performed in accordance with Affymetrix protocols as described previously \[[@B43],[@B44]\].
Microarray analysis
-------------------
Fluorescent intensity data from Affymetrix Microarray Suite version 5 were exported as CEL files and imported into DNA-Chip Analyzer version 1.3 \[[@B45]\]. Data were normalized, and expression values, based on the perfect match/mismatch (PM/MM) model, were calculated for each DNA chip. All chips were examined for the image spikes, chip and gene outliers. Exported expression values for each DNA chip were combined into a single file (three chips × four experimental conditions × three to five replicates), and imported back to DNA-Chip Analyzer; the resulting data were normalized by using an array with median probe intensity.
For the pairwise comparison of experimental conditions, signals were filtered by using several criteria. Gene expression was considered above the background if it showed the signal on most chips (more than 50%; that is, for three replicates, the gene should be detectable on at least two chips; for five replicates, the gene should be present on at least three DNA chips). Fold changes for gene expression were calculated when any of three following criteria were met: (1) the gene was present in the experimental condition but absent in the basal condition; (2) the gene was present in the basal condition but absent in the experimental condition; (3) the gene was present in both basal condition and experimental conditions. Student\'s *t*-test was used to determine the statistical significance of the difference in gene expression between basal and experimental conditions (*P*\< 0.05 was taken as significant). An additional cut-off threshold of twofold change in gene expression (either upregulation or downregulation) was used to characterize a gene as being differentially regulated (for example, a negative twofold value corresponded to a twofold downregulation). The Fisher exact test (implemented by us in Visual Basic code for MS Excel 2000) and the Mann--Whitney *U*-test (SPSS, Chicago, Illinois, USA) were used to verify non-paired Student\'s *t*-test calculations of the probability of gene expression differences in pairwise comparisons. Finally, the false discovery rate was established with 500 permutations for each pairwise comparison to estimate the proportion of false-positive genes.
To characterize gene expression patterns, hierarchical gene clustering was performed with a DNA-Chip Analyzer program \[[@B45],[@B46]\]. The algorithm was based on the distance between two genes defined as 1 - *r*, where *r*is the Pearson correlation coefficient between the standardized expression values of the two genes across the samples used. To characterize functional relationships between differentially expressed genes, Gene Ontology terms classification \[[@B47]\], incorporated in DNA-Chip Analyzer, was performed \[[@B48]\]. The significance level for a functional cluster was set at *P*\< 0.05, and the minimum size of a cluster was three genes. Venn diagram calculations were performed in Visual Basic code for MS Excel 2000 to analyze overlapping of sets of genes differentially expressed in the samples at different phases of arthritis.
Results
=======
The major goal of the present study was to find and characterize early signature genes whose expressions were different (at least twofold change in the threshold level) and statistically significant (*P*\< 0.05) between experimental groups at different phases of joint inflammation. The induction of arthritis in BALB/c^SCID^mice was a multi-step process. First, donor BALB/c mice were immunized with cartilage PG to induce arthritis. Second, spleen cells from acutely arthritic (AA) donor mice were stimulated *in vitro*with cartilage PG, and live lymphocytes were isolated on a Lympholyte-M density gradient. Third, these antigen-stimulated donor lymphocytes were injected into BALB/c^SCID^mice. For gene expression profiling during the time course of the adoptively transferred arthritis, RNA was isolated from pre-arthritic paws (PA) and diseased paws (AA and CA) (Table [1](#T1){ref-type="table"}). In addition, RNA was isolated from normal paws of naive BALB/c^SCID^mice and served as a baseline non-arthritic control condition (AN). Three pairwise comparisons were performed: PA versus AN, AA versus AN, and CA versus AN (hereafter denoted as PA/AN, AA/AN and CA/AN).
Each experimental condition was reproduced three to five times (RNA isolation, probe preparation, and independent hybridizations), and each replicate contained RNA samples pooled from a total of four paws of two arthritic animals. When the number of replicates is low and the distribution of data in the general population is basically unknown, the applicability of Student\'s *t*-test is questionable. We therefore analyzed data by using both Student\'s *t*-test and the Fisher exact test, in which the first approach requires normal data distribution, whereas the second test does not have this requirement \[[@B45],[@B49],[@B50]\]. Setting the significance level for the difference between groups at *P*\< 0.05 and no threshold for the fold change in expression, 1805 genes passed the Fisher exact test and 1752 genes passed the DNA-Chip Analyzer Student\'s *t*-test \[[@B45]\] for the PA/AN comparison. In AA/AN pairwise comparisons, 3676 genes passed the Fisher exact test and 3305 genes passed Student\'s *t*-test. Concluding that Student\'s *t*-test provided similar results and was even more conservative than the Fisher exact test, we employed the former for all further analyses.
Effect of the numbers of replicates on data variability
-------------------------------------------------------
Being aware of the importance of data reproducibility, we determined the optimal number of arrays to be included in experimental design by monitoring the convergence of variance for gene expression signals in five replicates representing the condition AA. For each replicate, we pooled equal amounts of quality-controlled RNA samples, isolated from two inflamed paws of two BALB/c^SCID^mice that had been identically treated (in terms of the number of donor cells and antigen stimulation) and had similar disease onset and severity. A total of five replicates represented 20 paws of 10 arthritic mice. We used the coefficient of variation (CV) to measure data variability. The CV for each gene on the chip and the mean CV for the entire probe set were calculated. Mean CV reached a plateau when the number of replicates increased beyond three (Fig. [1](#F1){ref-type="fig"}, experimental condition AA) and there was no significant change afterwards. Therefore, for all other experimental conditions, we used three replicates representing three independent hybridization experiments of three RNA samples isolated from six paws. Mean CV after sampling of the three repeats ranged between 0.21 and 0.25 for all experimental conditions.
Arthritis \'signature\' genes in pre-inflamed joints
----------------------------------------------------
Paws of naive BALB/c^SCID^mice and still non-inflamed (PA) paws were clinically normal with no sign of inflammation, and comparison of these two experimental conditions (PA/AN) identified a relatively small number of differentially expressed genes. Only 37 of the 36,000 screened genes were differentially expressed (that is, showed greater than a ± twofold change relative to threshold level), of which 11 genes were over the ± threefold threshold, and seven genes changed beyond ± fivefold (Fig. [2](#F2){ref-type="fig"}). The seven genes with the most significant change in expression levels encoded chemokine CC motif receptor 5 (*Ccr5*), chemokine CXC motif ligand 1 (*Cxcl1*), interferon-γ-inducible protein (*Ifi47*), membrane-spanning 4-domains subfamily A member 6C (*Ms4a6c*), tumor necrosis factor-α-induced protein 6 (*Tnfip6*), T cell receptor β variable 13 (*Tcrbv13*), and Terf1-interacting nuclear factor2 (*Tinf2*) (Table [2](#T2){ref-type="table"}). Although the upregulation of *Tcrbv13*, *Tgtp*and interferon-induced genes might indicate the appearance of antigen-specific T cells in the synovium (Table [2](#T2){ref-type="table"}), the significant upregulation of *Tnfip6*suggests the activation of an anti-inflammatory cascade \[[@B51]\]. Thus, gene expression related to pro-inflammatory and anti-inflammatory events can be detected even before the migration of inflammatory leukocytes into the joints.
To characterize major biological functions in context with the initiation phase of the disease, we assigned the 37 early genes (Table [2](#T2){ref-type="table"}, [Additional file 1](#S1){ref-type="supplementary-material"}) to separate groups according to the corresponding protein functions and Gene Ontology classification \[[@B47],[@B48]\]. We found that differentially expressed genes in PA joints were related to immune responses, chemokine activity (including chemotaxis), cell adhesion, proteolysis regulation, inflammation and wounding, cytokines, and cytoskeletal activity (Fig. [3](#F3){ref-type="fig"}, yellow circles). All clustered genes were upregulated at the pre-inflamed phase of arthritis.
Gene expression profile in acute and chronic arthritis
------------------------------------------------------
To monitor the progression of disease, we analyzed genes that were differentially expressed in paws with acute and chronic joint inflammation. Both AA and CA experimental conditions were associated with the activity of a large number of genes: 256 genes were upregulated and 21 were downregulated in acute arthritis (AA/AN comparison), and 201 genes were upregulated and 217 were downregulated in chronic inflammation (CA/AN) (Fig. [2](#F2){ref-type="fig"}, Additional files [2](#S2){ref-type="supplementary-material"} and [3](#S3){ref-type="supplementary-material"}). A Venn diagram summarizes the relationships between gene sets that were differentially expressed at different phases of the disease. Only 15 genes were differentially expressed in all three phases of the disease (PA, AA, and CA), 25 genes were differentially expressed both at the PA phase and during acute inflammation, 127 genes were active both in acute and chronic phases, and 17 transcripts shared a common expression pattern in pre-inflamed and chronically inflamed joints (Fig. [2](#F2){ref-type="fig"}).
Using Gene Ontology terms for the functional classification of genes differentially expressed in acute and chronic arthritis \[[@B47]\], dozens of cell signaling pathways and gene clusters were identified. By further filtering of functional clusters, and by combining clusters encoding proteins with similar functions, we found that the acute and chronic phases of the disease can be comprehensively described by the differential expression of 15 macro-clusters (Fig. [3](#F3){ref-type="fig"}). Six clusters were found in all three phases of inflammation; they were related to immune response, chemokine activity, cytokines, inflammation and wounding, cell adhesion, and proteolysis regulation. The most abundantly represented genes in inflamed joints were those involved in immune responses: 51 genes in AA and 25 genes in CA. These genes were upregulated as much as 31-fold (group average) in acute arthritis and 15-fold in chronic arthritis (Fig. [3](#F3){ref-type="fig"}). Cytokine and chemokine genes demonstrated the highest overexpression levels: about 64-fold in acute and 28-fold in chronic arthritis, where both groups included more than a dozen genes. Proteolysis-regulating genes (proteases and their inhibitors) were highly represented at the acute phase (45 genes), but were less abundant in chronic arthritis (19 genes). Extracellular matrix-related genes, mostly relevant to tissue repair and healing, were more abundant in chronic than acute disease. Some functional clusters were phase-specific, such as lysosome, antigen presentation, scavenger receptors, immunoglobulin binding, and complement cascade; these genes were preferentially expressed in acute joint inflammation. Suppression of genes related to the respiratory chain complex was specific to chronic inflammation (Fig. [3](#F3){ref-type="fig"}).
Hierarchical clustering of arthritis phase-specific genes
---------------------------------------------------------
To identify genes whose expression might be specific for the actual phase of arthritis, and to combine transcripts by the pattern of their expression through all disease phases, we applied a hierarchical clustering technique \[[@B46]\]. Genes that were specific for pairwise comparisons (PA/AN, AA/AN, and CA/AN) were combined into one single file (excluding redundant genes); the merged set included 507 genes. Hierarchical clustering was performed for all experimental conditions studied (AN, PA, AA, and CA), and four major gene clusters were identified, each with a distinct expression pattern (Fig. [4](#F4){ref-type="fig"}, clusters I--IV). Using further classification analysis with Gene Ontology terms, to examine the functions of genes inside each cluster, we identified genes encoding proteins whose biological functions were the most relevant to arthritis development and progression.
Cluster I contained genes with major functions in collagen turnover and tissue repair; the expression of these genes reached a peak in chronically inflamed joints.
Cluster II was the largest cluster including about half of all phase-specific genes (Fig. [4](#F4){ref-type="fig"}). The cluster included genes with roles in immune, inflammatory and stress responses, extracellular matrix formation, cell growth, and receptor activity. The expression of cluster II genes reached a peak at the acute phase of joint inflammation.
Transcription of genes in clusters III and IV gradually decreased during disease progression (Fig. [4](#F4){ref-type="fig"}). These genes were mostly related to cytoskeleton remodeling, the formation of cell junctions, and the production of structural molecules such as desmin, β-3 laminin, envoplakin, and dystonin (for a detailed gene list see Additional files [1](#S1){ref-type="supplementary-material"}, [2](#S2){ref-type="supplementary-material"}, [3](#S3){ref-type="supplementary-material"}). Genes associated with early arthritis (Table [2](#T2){ref-type="table"}) were found in clusters III and IV, further underlining the importance of cell adhesion and cytoskeleton remodeling during the initiation phase of arthritis.
Expression patterns of early arthritis genes
--------------------------------------------
Hierarchical clustering of a large number of phase-specific genes (*n*= 507) (Fig. [4](#F4){ref-type="fig"}) obscured the expression pattern of a relatively small number (*n*= 37) of early arthritis genes (Table [2](#T2){ref-type="table"}). A separate hierarchical clustering was therefore performed for these 37 early genes, and the levels of expression were monitored at later phases of the disease. Six distinct expression patterns were identified (Fig. [5](#F5){ref-type="fig"}, clusters A--F) using this approach. Clusters A--D contained early arthritis genes whose transcription increased as the disease progressed, reaching a peak in the pre-inflamed joint or during inflammation. Cluster A included genes that coded for variable parts of the T cell receptor, together with genes related to cytoskeleton reorganization such as Rho interacting protein 3, myosin, and β-actin (reviewed in \[[@B52],[@B53]\]). Cluster A genes were at the peak of their expression in the PA joint. However, most early arthritis genes in clusters C and D showed an expression peak later, at the acute phase of inflammation (Fig. [5](#F5){ref-type="fig"}), and encoded chemokine receptors (*Ccr2*and *Ccr5*) and chemokine ligands (*Cxcl1*, *Ccl2*, *Ccl7*, and *Ccl9*). Clusters C and D also included interferon-activating genes *Ifi203*, *Ifi47*, and *Ifigtp*, and cell differentiation antigens such as CD48 and CD53.
Hierarchical clusters E and F contained four genes whose expression was downregulated in the pre-inflamed joint but returned to a \'normal\' level (as expressed in naive paws) during arthritis progression. Clusters E and F included genes encoding Terf1-interacting nuclear factor 2, tissue inhibitor of metalloproteinase 1, makorin, and DNA clone *4833424O15*with unknown function (Table [2](#T2){ref-type="table"} and Fig. [5](#F5){ref-type="fig"}).
Discussion
==========
This study describes genome-wide gene activity taking place in mouse joints during three major phases of autoimmune arthritis: initiation, acute inflammation, and chronic inflammation. Spleen cells from PG-immunized arthritic BALB/c mice were used to transfer the disease into non-immunized syngeneic SCID mice \[[@B30],[@B32]\]. This adoptive transfer system minimized the individual differences that are typical in primary arthritis (induced by systemic immunization), and also excluded antigen-independent stimulation of the immune system by the adjuvant. Additional benefits of the cell transfer included a decrease in the time needed for arthritis development, and uniformity and synchronization of joint inflammation in recipient mice \[[@B32]\].
Two major criteria were used to select genes that might be important for arthritis development: (1) significant differences in expression levels between experimental groups and (2) the fold change in expression levels. When only the first criterion was applied, genome-wide analysis identified a large number of genes whose expression was significantly (*P*\< 0.05) different between any pair of the experimental conditions compared. Irrespective of the statistics used (either unpaired Student\'s *t*-test, the Fisher exact test or the Mann--Whitney *U*-test), the number of differentially expressed genes was found to represent about 5--10% of the entire mouse genome. We further \'filtered\' these genes by using a cut-off threshold set at twofold change of expression, because this threshold could reflect a physiologically important change in gene activity, and a twofold change exceeded the average CV for all pairwise comparisons. Decreasing the number of \'false positive\' genes by application of these two filtering procedures proved to be an effective technique for the identification of genes that are likely to be involved in arthritis development.
The present study indicates that the number of differentially expressed genes increases with the progression of the disease. At the initiation phase, when no clinical symptoms of inflammation were yet detected, only 37 genes were upregulated or downregulated. However, a differential expression of 277 genes was observed at the acute phase, and chronic inflammation was characterized by the differential activity of 418 genes. Interestingly, most early arthritis signature genes (27 of 37) remained upregulated or downregulated in inflamed joints (Fig. [2](#F2){ref-type="fig"}). A different set of genes was also involved in acute inflammation. At the chronic phase, less than half of AA-specific genes (127 of 277) were differentially expressed, and another half was CA-specific. A very limited number of transcripts (*n*= 15) remained upregulated or downregulated in all three phases of arthritis.
Activated T cells must be present in the peripheral blood of recipient BALB/c^SCID^mice after the transfer, but donor lymphocytes can be detected in joints as early as 3--5 days after the second transfer \[[@B32]\]. In earlier studies \[[@B31]\], and in control experiments (data not shown), using fluorescein-labeled or isotope-labeled donor lymphocytes, only very few cells were found in joints during the first week of transfer, and a second cell transfer was needed to induce a significant influx of lymphocytes into the joints and cause subsequent inflammation. In this study, we detected overexpression of a T cell-specific GTPase (*Tgtp*) and T cell receptor β (*Tcrbv13*) in still non-inflamed (pre-arthritic) paws of recipient BALB/c^SCID^mice as early as 3--5 days after the second injection, indicating the presence of donor BALB/c lymphocytes. Thus, the initiation and development of arthritis in adoptively transferred PGIA must depend on cooperation between adaptive immunity cells (represented by donor BALB/c lymphocytes) and cells of innate immunity (represented by non-lymphoid cells in the recipient BALB/c^SCID^mice). Analysis of the cellular and tissue specificity of gene expression, using public gene expression databases \[[@B54]-[@B56]\], indicated that genes encoding CD48 (*Cd48*), membrane-spanning 4A6B and 4A6C (*Ms4a6b*and *Ms4a6c*), epidermal growth factor-like receptor-like protein 1 (*Emr1*), and interferon-induced 47 kDa protein (*Ifi47*) were most probably originating from donor lymphoid cells, whereas other early arthritis genes (Table [2](#T2){ref-type="table"}) were related to the activation of the innate immune system (represented by macrophages, dendritic cells, and cells of myeloid lineage) of recipient BALB/c^SCID^mice.
Transcriptional control of gene activity is only one component of the complex cellular regulatory pathways. In other words, the functional activity of a protein depends on several factors such as interaction with other proteins, phosphorylation/dephosphorylation, subcellular compartmentalization, and other post-translational modifications. All of these factors might be involved in the regulation of interactions between the donor lymphocytes and the synovial/joint cells of recipient mice that lack an adaptive immune system. The list of genes we present in this study is rather short; that is, it includes only genes profoundly affected during arthritis initiation and progression at the level of transcription. Genes and proteins that are under subtle regulatory pressure, or are controlled by non-genetic mechanisms such as protein phosphorylation and other post-translational events, could not be detected and analyzed in this study. The development of new proteomics assays, and the synthesis of existing knowledge in cellular signaling pathways with information provided by gene expression studies, will be necessary to build up a complete arthritis-related regulatory network and to unravel the mechanisms involved in the development and progression of autoimmune arthritis.
Conclusions
===========
The development and progression of a complex polygenic autoimmune disease such as RA are controlled by hundreds or thousands of genes, in addition to the MHC. Despite the relatively high incidence of RA in the human population, only a few studies have applied gene array methods to the monitoring of disease progression and efficacy of treatment, or to predicting the prognosis of the disease. The major obstacles in the human studies are the relatively late diagnosis of RA, the large variety of cell types (cells of the immune system and of synovial joints) involved in autoimmune arthritic processes, and the extreme genetic heterogeneity of the human population. The present study applied an adoptively transferred murine model of RA and a microarray approach to detect differentially expressed, disease-related signature genes in PA (still non-inflamed) joints, days before the clinical symptoms or histopathological abnormalities of joint inflammation could be observed.
However, the detection of early arthritis signature genes in joints can be done only in an experimental system in which particular joints have already been affected before the inflammatory symptoms can be identified. To make this experimental system uniform, that is, to exclude individual variations, we adoptively transferred antigen (PG)-specific lymphocytes (representing cells of adaptive immunity) from primarily arthritic mice into syngeneic SCID mice, which lack an adaptive immune system. In this highly synchronized and uniform system we were able to detect differentially expressed genes in still non-inflamed paws of arthritis-\'prone\' animals. We identified a relatively small number of mostly upregulated early arthritis signature genes (known to be involved in arthritic processes and/or autoimmunity), some of which were expressed at even higher levels in the acute phase of arthritis. These early arthritis signature genes, originating from donor cells, indicated the involvement of adaptive immunity, whereas the innate immunity genes were differentially expressed by cells of the recipients.
The early signature genes, together with those that were differentially expressed in the acute (277 genes) and chronic (418 genes) phase of arthritis, are listed in the Additional files. Although many of these differentially expressed genes, detected either in the acute phase or during the progression of the disease, have been implicated in inflammation or autoimmunity, the list contains a significant number of differentially expressed genes whose function, or association with arthritis, is unknown at present.
Abbreviations
=============
AA = acutely arthritic; AN = absolutely negative (control naive); CA = chronically arthritic; CV = coefficient of variation; DDA = dimethyldioctadecylammonium bromide; PA = pre-arthritic; PG = cartilage proteoglycan aggrecan; PGIA = PG-induced arthritis; RA = rheumatoid arthritis; SCID = severe combined immunodeficient.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
VAA performed essentially all statistical analyses and put together the draft version of the results and figures. CV isolated all RNA samples, prepared biotinylated samples and was involved in Affymetrix hybridization experiments; he also performed preliminary clustering experiments with GeneSpring version 6.2 (not included in this paper). AH performed all *in vitro*stimulation and adoptive transfer experiments, and assessed arthritis three or four times a day together with KM, who was also involved in all phases of the experimental processes and in the finalization of the manuscript. EGB controlled Affymetrix hybridization and scanning experiments, managed preliminary data analysis and finalized the manuscript. TTG designed experiments, controlled all experimental steps, data analysis, and finalized the manuscript. All authors read and approved the final manuscript.
Supplementary Material
======================
::: {.caption}
###### Additional File 1
A table (Excel file) that lists all information about the 37 genes differentially expressed (more than twofold level) in pre-arthritic joints/paws, when compared with the same genes expressed in normal (naive) joints/paws of BALB/c^SCID^mice (PA/AN comparison). The pre-arthritic (still non-inflamed) joints were collected within 24--48 hours after the onset of inflammatory symptoms in BALB/c^SCID^mice with adoptively transferred PGIA. Acutely inflamed paws from these arthritic BALB/c^SCID^mice were used as acute arthritic (AA) samples (the list of gene expression profiles is provided in [Additional file 2](#S2){ref-type="supplementary-material"}.)
:::
::: {.caption}
######
Click here for file
:::
::: {.caption}
###### Additional File 2
This file contains information about 256 upregulated and 21 downregulated genes in acutely arthritic joints/paws in five independent hybridization experiments. All genes with twofold or higher expression levels are listed.
:::
::: {.caption}
######
Click here for file
:::
::: {.caption}
###### Additional File 3
This file includes 201 upregulated and 217 downregulated genes in subacute/chronic phase of arthritis (8--12 days after onset) with the corresponding information. The gene expression levels (twofold or higher) are compared with those expressed in normal joints of naive (absolutely negative) BALB/c^SCID^mice.
:::
::: {.caption}
######
Click here for file
:::
Acknowledgements
================
We thank Dr Kira Adaricheva for algorithmic and mathematical support during data analysis. We are indebted to David George and Yonghong Zhang for Affymetrix fluidic station operation and data management. This research was supported in part by grants AR40310, AR45652, and AR51163 from the National Institutes of Health, and the JO Galante MD DSc endowment fund.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Average coefficient of variation with increasing number of replicates of gene expression experiments. Data represent results obtained with RNA from normal paws of naive BALB/c^SCID^mice (AN), clinically normal pre-arthritic paws (PA), acutely arthritic paws (AA) and chronically inflamed paws (CA).
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Fold change distribution for genes differentially expressed in pre-inflamed joints, in paws with acute and chronic arthritis, in comparison with gene expression in normal paws of naive BALB/c^SCID^mice. Values indicate the number of genes that fall in the given range of expression. Negative numbers for expression levels indicate downregulation (e.g. a negative twofold change corresponds to downregulation to 0.5-fold). Spikes at ± 5-fold expression change represent the extremes of histogram when combining all genes with differential expression level greater than ± 5-fold. The Venn diagram (bottom) indicates the number of overlapping genes that were differentially expressed in pre-inflamed and arthritic joints.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Gene activities at different phases of arthritis progression. All clusters identified in pre-inflamed joints (PA/AN comparison, yellow circles), acute arthritis (AA/AN, red circles), and chronic arthritic paws (CA/AN, blue circles) are indicated by the number of genes in the cluster (circle diameter represents cluster size) and the average fold change of gene expression (logarithmic horizontal scale). The size of the cluster varies from 3 genes (\'complement cascade\' cluster) in pre-inflamed joints to 51 genes (\'immune response\' cluster) in acute arthritis. AN, normal paws of naive BALB/c^SCID^mice; PA, clinically normal pre-arthritic paws; AA, acutely arthritic paws; CA, chronically inflamed paws.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Signature gene clusters at different phases of autoimmune arthritis. Hierarchical clustering was performed for genes whose expression significantly differed when paws of naive mice (AN) were compared with those in the pre-arthritic (PA), acute (AA), or chronic (CA) phases of arthritis. The total number of genes (*n*= 507) is less than the sum of the phase-specific genes because of partial overlap (Fig. 2). Rows represent individual genes; columns represent individual expression values for each gene at the indicated phase of arthritis. The major biological activities, specific for each cluster, were examined by using functional clustering of genes. This analysis yielded four different expression patterns (clusters I--IV). Upregulated genes are shown in red, downregulated genes in blue.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Hierarchical clustering (left) and expression patterns (A--F) for 37 early arthritis genes (listed in Table 2) differentially expressed in pre-inflamed (PA) joints of recipient BALB/c^SCID^mice. Gene expression was compared with normal paws (AN) of naive BALB/c^SCID^mice (PA/AN comparison, with a cut-off threshold at twofold change). The expression profiles of these 37 signature genes are shown for each phase of the disease (PA, acute \[AA\], or chronic \[CA\]) and also in normal paws.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Experimental groups used for adoptive disease transfer and differential expression analysis
:::
Group RNA source Treatment Days after injection No. of animals
------- --------------------------------------------------- --------------- ---------------------- ----------------
AN Naive control (absolute negative) BALB/c^SCID^paw None N/A 3
PA Normal paw from arthritic BALB/c^SCID^mouse Cell transfer 6 3
AA Acute arthritic paw of BALB/c^SCID^mouse Cell transfer 6 5
CA Chronically arthritic paw of BALB/c^SCID^mouse Cell transfer 12--14 3
Group AN represents naive BALB/c^SCID^mice that received no cells. Experimental groups PA, AA, and CA received antigen-stimulated lymphocytes from arthritic BALB/c donor mice. RNA was isolated from four paws of two mice at the indicated number of days after injection, and pooled.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Array-based expression values of upregulated or downregulated genes in pre-inflamed joint
:::
Affy ID Description Gene Mean AN expression AN presence call Mean PA expression PA presence call Fold Cluster
--------------- ------------------------------------------------ ----------------- -------------------- ------------------ -------------------- ------------------ ------- ---------
161968\_f\_at **Chemokine (CC motif) receptor 5** *Ccr5* 1 A 57.1 P 57.1 D
95349\_g\_at Chemokine (CXC motif) ligand 1 *Cxcl1* 1 A 55.9 P 55.9 D
104750\_at **Interferon-γ inducible protein** *Ifi47* 0.69 A 15.69 P 22.7 D
130509\_at Membrane-spanning 4-domains member A6C *Ms4a6c* 1.17 A 10.27 P 8.78 D
93106\_i\_at T-cell receptor beta, variable 13 *Tcrbv13* 1.86 A 10.34 P 5.55 A
98474\_r\_at Tumor necrosis factor-α induced protein 6 *Tnfaip6* 3.61 A 19.97 P 5.54 D
94761\_at **Chemokine (CC motif) ligand 7** *Ccl7* 24.47 A 115.33 P 4.71 D
101578\_f\_at Actin, β, cytoplasmic *Actg* 188.33 P 859.16 P 4.56 A
102736\_at **Chemokine (CC motif) ligand 2** *Ccl2* 29.89 P 125.48 P 4.20 D
95121\_at Polymerase (DNA-directed) ε 4 p12 *Pole4* 8.95 P 26.13 P 2.92 C
93397\_at **Chemokine (CC) receptor 2** *Ccr2* 93.67 P 259.21 P 2.77 D
102906\_at **T cell-specific GTPase** *Tgtp* 24.27 A 66.03 P 2.72 D
97322\_at Membrane-spanning 4-domains member A6B *Ms4a6b* 9.43 A 25.16 P 2.67 D
93514\_at Myosin, light polypeptide 3 *Myl3* 29.28 A 77.42 P 2.64 A
103089\_at CD48 antigen *Cd48* 34.17 A 85.82 P 2.51 C
103507\_at **EGF-like hormone receptor-like sequence 1** *Emr1* 59.04 A 145.95 P 2.47 D
96764\_at **Interferon-inducible GTPase** *Ifigtp* 66.38 A 157.88 P 2.38 D
102326\_at Neutrophil cytosolic factor 2 *Ncf2* 23.67 P 55.88 P 2.36 D
104388\_at **Chemokine (CC motif) ligand 9** *Ccl9* 246.56 P 568.47 P 2.31 D
93321\_at Interferon-activated gene 203 *Ifi203* 43.58 P 99.27 P 2.28 C
94085\_at Proteoglycan, secretory granule *Prg* 329.68 P 747.2 P 2.27 D
92762\_at **C-type lectin, superfamily member 6** *Clecsf6* 35.73 A 80.64 P 2.26 D
93136\_at Dermatan sulphate proteoglycan 3 *Dspg3* 30.23 A 67.77 P 2.24 B
101753\_s\_at P lysozyme structural *Lzps* 635.78 P 1398.72 P 2.20 D
94939\_at CD53 antigen *Cd53* 112.16 P 243.17 P 2.17 D
94958\_at RIKEN cDNA 1110013L07 gene *1110013L07Rik* 24.91 A 53.7 P 2.16 A
162066\_f\_at Rho interacting protein 3 *Rip3* 19.15 A 40.54 P 2.12 A
93039\_at RIKEN cDNA 1190003P12 gene *1190003P12Rik* 36.44 P 77.03 P 2.11 A
101048\_at Protein tyrosine phosphatase, receptor type, C *Ptprc* 138.7 A 289.19 P 2.09 D
92217\_s\_at Glycoprotein 49 B *Gp49b* 72.11 P 148.01 P 2.05 D
93869\_s\_at **Hematopoietic-specific A1-d protein** *Bcl2a1a* 49.43 A 100.11 P 2.03 D
103989\_at RIKEN cDNA 4432417F03 gene *4432417F03Rik* 35.66 A 72.18 P 2.02 B
160611\_at **Cytochrome P450 polypeptide 4v3** *Cyp4v3* 82 P 164.59 P 2.01 B
162107\_r\_at **Tissue inhibitor of metalloproteinase 1** *Timp1* 9.64 P 4.39 A -2.20 E
164493\_i\_at **Makorin, ring finger protein, 1** *Mkrn1* 8.23 P 3.21 A -2.56 F
167637\_i\_at RIKEN cDNA 4833424O15 gene *4833424O15Rik* 8.19 P 2.64 A -3.11 F
167950\_r\_at Terf1 (TRF1)-interacting nuclear factor 2 *Tinf2* 8.82 P 1.65 P -5.35 E
Affy ID, unique Affymetrix probe set identifier. Description, gene description. Gene, gene abbreviation. Mean AN expression, average expression value in basal experimental condition of clinically normal paws of naive severe combined immunodeficient mice without cell transfer. Mean PA expression, average expression value in pre-arthritic joints. Presence call, average presence call for gene in AN or PA experimental condition: P, transcript was actually present in the majority of samples; A, transcript was actually absent in the majority of samples. Fold, fold change in gene expression in PA joint compared with AN basal expression. Cluster, cluster designation from Fig. 5. Difference in expression was significant by Mann--Whitney *U*-test, *P*\< 0.05. Differential expression for listed genes was either greater than twofold overexpression or less than twofold downregulation (negative values). Genes that were differentially expressed in both pre-inflamed paws and *in vitro*-stimulated lymphocytes used for cell transfer are shown in bold type.
:::
|
PubMed Central
|
2024-06-05T03:55:54.922814
|
2004-12-14
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065315/",
"journal": "Arthritis Res Ther. 2005 Dec 14; 7(2):R196-R207",
"authors": [
{
"first": "Vyacheslav A",
"last": "Adarichev"
},
{
"first": "Csaba",
"last": "Vermes"
},
{
"first": "Anita",
"last": "Hanyecz"
},
{
"first": "Katalin",
"last": "Mikecz"
},
{
"first": "Eric G",
"last": "Bremer"
},
{
"first": "Tibor T",
"last": "Glant"
}
]
}
|
PMC1065316
|
Introduction
============
Rheumatoid arthritis (RA) is a chronic inflammatory condition that affects multiple joints, and it results in the accumulation of leukocytes within the synovial tissue (ST) and synovial fluid (SF). The inflammatory infiltrate consists predominantly of B lymphocytes, T lymphocytes and macrophages in the ST, whereas neutrophils are mainly found in the SF. The lymphocyte infiltration is organized in lymphoid-like microstructures in just under 50% of the RA patients; however, the patients present germinal centre reactions in only 20% of cases \[[@B1]\]. The pathogenesis of the RA is still largely unknown but leukocytes and their products play an important role in the development of inflammation, joint destruction and pain \[[@B2],[@B3]\]. The attraction of leukocytes into the joints is controlled by chemokines, a family of small chemotactic cytokine-like molecules that act as potent mediators of inflammation \[[@B4]\].
Chemokine activity is dependent on the presence of and interaction with chemokine receptors on the leukocyte surface. Indeed, chemokines and their receptors are involved together in the development and perpetuation of inflammation \[[@B5]\]. *In vitro*and *in vivo*experiments have indicated that blocking chemokines or their receptors could potentially provide an effective treatment of inflammatory diseases \[[@B5],[@B6]\]. The 19 receptors so far identified belong to a super-family of G-protein-coupled receptors with seven transmembrane domains \[[@B7]\]. Chemokine receptors have a regulatory effect on the maturation and traffic of leukocytes, and they are implicated in several disease states \[[@B8]\]. There have been several reports on chemokine receptor expression on T cells from RA ST, RA SF and RA peripheral blood (PB) \[[@B9]-[@B13]\]. The expression of some chemokine receptors on monocytes/macrophages, dendritic cells and neutrophils has also been reported \[[@B14]-[@B17]\], and the importance of the role of chemokine receptors in RA is emerging \[[@B18],[@B19]\].
CXCR5 is a chemokine receptor highly expressed in recirculating B cells, in subsets of CD4^+^and CD8^+^T cells and monocytes \[[@B20],[@B21]\]. It also has been identified on B-cell infiltrates in Sjogren\'s syndrome \[[@B22],[@B23]\]. CXCR5 is involved in the immune-system homeostasis and in lymphoid organogenesis \[[@B24]\]. Several morphological and functional studies suggest that lymphoid neogenesis takes place in RA \[[@B1],[@B25],[@B26]\]. Furthermore, an important disturbance of follicle and germinal centre formation in the spleen and Peyer\'s patches is observed in CXCR5-deficient mice \[[@B27]\]. CXCL13, the unique ligand of CXCR5, is also involved in follicular homing, as observed in CXCL13-deficient mice \[[@B28]\].
In view of the role of chemokine receptors in leukocyte traffic, the aim of the present study was to compare their expression in inflamed and non-inflamed tissue to shed light on which chemokine receptors may be involved in the recruitment and retention of leukocytes in ST. We examined chemokine receptor expression in ST taken from RA and non-RA patients using microarray technology, RT-PCR and immunohistochemistry. The microarray and RT-PCR experiments demonstrated the differential expression of *CXCR5*, and immunohistochemistry showed that this receptor is expressed in B-cell and T-cell infiltrates, on macrophages and blood vessels. Our study identifies CXCR5 as a potentially interesting therapeutic target in RA and points to the use of antagonists to this receptor as a treatment strategy in the disease.
Materials and methods
=====================
Tissue and cell source
----------------------
Tissue samples were obtained from patients with RA (*n*= 8) who fulfilled the American Rheumatism Association criteria for RA (Table [1](#T1){ref-type="table"}). The patients\' mean age was 59 ± 14.8 years with a male to female ratio of 1:8. The disease duration of six out of eight RA patients was over 10 years. ST was taken from these subjects at the time of total knee replacement. Non-RA patients (*n*= 9) had knee joint symptoms for suspected articular cartilage or meniscal damage (Table [1](#T1){ref-type="table"}). Their mean age was 47.6 ± 6.8 years with a male to female ratio of 8:1. Except for one patient, the non-RA patients had knee complaints for 1 year or less. ST biopsies were obtained from these patients at the time of arthroscopy. All samples were taken with informed consent and ethical approval. The ST samples were taken from the suprapatellar pouch and the medial gutter, which is reported to provide representative sampling of synovial membrane pathology \[[@B29]\]. Synovia were cut into approximately 4 mm^3^pieces and were either snap frozen in isopentane and stored in liquid nitrogen or formalin fixed and paraffin embedded.
Monocytes/macrophages were isolated from the PB and SF of another four RA patients (Table [1](#T1){ref-type="table"}). In brief, the blood and hyaluronidase-treated SF were centrifuged over a ficoll cushion (Amersham Biosciences, Chalfont St Giles, UK). The macrophages were isolated from the buffy coat layer (lymphocytes, macrophages) by adherence onto a glass dish.
RNA extraction
--------------
Total RNA was extracted from frozen blocks of synovia or from isolated monocytes/macrophages using TRIreagent solution (Sigma, Poole, UK) according to the manufacturer\'s recommendation. The quantity recovered was determined by spectrophotometry and the integrity was assessed by gel electrophoresis. For microarray experiments, equal quantities (7 μg) of RNA from each RA or non-RA patient were pooled and the messenger RNA was extracted using the mRNA GeneElute Kit (Sigma). The quantity recovered was determined by fluorometry using SYBR Green II (Molecular Probes, Leiden, The Netherlands). RNA had to be pooled since only small biopsies could be obtained from non-RA patients.
Microarray technology
---------------------
The panorama human cytokine gene array (Sigma-Genosys, Pampisford, UK) was used. This array contains 375 different cDNAs including 16 chemokine receptors and 33 chemokines, each printed in duplicate on nylon membranes.
The probe labelling and hybridization were carried out according to the manufacturer\'s instructions. Briefly, ^33^P-radiolabelled cDNA probes were prepared from 0.5 μg mRNA (see earlier) using human cytokine cDNA labelling primers (Sigma-Genosys) and AMV reverse transcriptase at 42°C, and were purified on a Sephadex^®^G-25 spin column (Sigma-Genosys). The arrays were hybridized for 17--18 hours at 65°C, washed and subjected to autoradiography for various lengths of time using Kodak BioMax MR X-ray film.
The intensity of hybridization signals was quantified using the ArrayVision, version 6.0, software (Imaging Research Inc., Haverhill, UK). The intensity of each spot was corrected for background levels using the \'corners between spots\' (set to 3 pixels) with or without \'segmentation\' protocols, and were normalized for differences in labelling using the average values of seven housekeeping genes: *β~2~-microglobulin*, *β-actin*, *cyclophilin A*, *glyceraldehyde-3-phosphate dehydrogenase*, *HLA-A 0201 heavy chain*, *human hypoxanthine phosphoribosyl transferase*, and *α-tubulin*. The remaining two housekeeping genes, *L19*and *transferrin R*, were excluded because of signal saturation and differential expression, respectively. The software performs the normalization automatically.
Reverse transcription-polymerase chain reaction
-----------------------------------------------
Total RNA aliquots from individual patients were reverse transcribed using oligo(dT~18~) (MWG Biotech, Ebersberg, Germany) and MMLV reverse transcriptase (Promega, Southampton, UK) at 37°C for 1 hour. The reactions were terminated at 70°C for 10 min and were diluted to 80 μl with H~2~O. For two of the non-RA patients no more RNA was available for RT-PCR following microarray analysis.
The PCR reactions were normalized against the ribosomal RNA *L27*using specific primers (MWG Biotech) (Table [2](#T2){ref-type="table"}). Appropriate cDNA dilutions were used subsequently for the RT-PCR reactions using specific primers for *CXCR5*(MWG Biotech) (Table [2](#T2){ref-type="table"}). Specific primers were designed from the published sequences. The number of cycles and the annealing temperature were optimized for each primer pair. The RT-PCR conditions were one cycle at 94°C for 3 min, 57°C for 1 min and 72°C for 1 min, *X*cycles at 94°C for 1 min, 57°C for 1 min and 72°C for 1 min, and one cycle at 94°C for 1 min, 57°C for 1 min and 72°C for 10 min. *X*equals 34 cycles for *CXCR5*and 24 cycles for *L27*.
Immunohistochemistry for CXCR5
------------------------------
The ST from the patients that had been examined at the transcription level was also available for protein expression analysis. Paraffin embedded sections were cut 4 μm thick on 3-aminopropyltriethoxysilane-coated slides. Sections were deparaffinized and rehydrated before blocking endogenous peroxidase activity with H~2~O~2~(0.3%) in methanol. The slides were rinsed with PBS and incubated with normal serum (1:67 in PBS) for 10 min before applying anti-human CXCR5 monoclonal antibody (1:666; R&D, Abingdon, UK) and the respective IgG control (Dako, Ely, UK). The sections were rinsed with PBS and incubated with biotinylated secondary antibody. The antibody binding was detected using reagents in the Vectastain ABC Elite kit (Vector, Peterborough, UK) and the chromogen 3,3\'-diaminobenzidine (DAB) (Vector). Sections were rinsed and counter stained in Mayer\'s haematoxylin.
B cells and macrophages were localized using anti-human CD20 antibodies (1:100; Dako) and CD68 antibodies (clone PG-M1, 1:75; Dako), respectively. CD20 required antigen demasking by 15 min microwaving in citrate buffer (pH 6.0), but H~2~O~2~treatment was not necessary. CD68 antigen was demasked using 0.05% pronase in Tris-buffered saline (pH 7.2) for 10 min.
Double immunohistochemistry was performed with anti-human CD3 rabbit monoclonal antibodies (Labvision) and CXCR5 antibodies. The slides were deparaffinized, rehydrated and microwaved for 15 min in citrate buffer pH 6.0 before being treated with H~2~O~2~in methanol. The slides were incubated with 2.5% normal swine serum for 20 min before applying CD3 diluted 1:60 in 2.5% serum for 30 min. The sections were rinsed with PBS and were incubated with swine anti-rabbit antibody linked to alkaline phosphatase (1:40; Dako). CD3 binding was detected using Vector Red substrate (Vector). Sections were rinsed and were either counter stained in Methyl Green (Vector) or subjected to a second round of immunohistochemistry. CXCR5 was used as for single immunohistochemistry except that blocking and antibody dilutions were made in 2.5% normal horse serum and CXCR5 was revealed with DAB-Nickel (Vector). No counter stain was performed for double immunohistochemistry sections.
Results
=======
Patients and tissue selection
-----------------------------
Synovia were obtained from knee joints as this allowed the use of arthroscopic samples of non-RA (normal) as controls instead of osteoarthritic tissue, which can show more enhanced inflammatory changes. The histology of H&E-stained RA synovial sections demonstrated classic signs of inflammation. Mononuclear cell infiltrates were visible in seven out of eight patients and consisted of aggregate structures; one of these seven patients also contained more germinal-like centre structures. In addition one patient revealed a diffuse infiltration. The synovium of the non-RA patients showed minimal signs of inflammation. In eight out of nine patients no mononuclear infiltrates were observed, and in one case only a small infiltrate was seen. No thickening of the intima was observed in the non-RA compared with the RA samples.
Microarray analysis of chemokine receptor expression
----------------------------------------------------
To allow rapid preliminary screening of a large number of chemokines and their receptors in RA ST and non-RA ST, chemokine expression was investigated using microarrays.
A pair of human cytokine microarrays including 16 chemokine receptors and 33 chemokines was hybridized with labelled cDNA probes prepared from mRNAs obtained from RA and non-RA pools of synovial RNA. Figure [1](#F1){ref-type="fig"} shows the results of hybridization of the RA and non-RA probes to the array membranes. To reduce the bias that could be introduced during the quantification, arrays showing very similar signals for the housekeeping genes were chosen and only non-saturated and non-regulated signals/genes were used for normalization. The intensity of each spot was corrected for background levels. The analysis step was repeated eight times for each pair of autoradiogram.
Of the 16 chemokine receptors present, the expression of 12 chemokine receptors was visible on the RA microarrays. These were *CCR1*, *CCR2a*, *CCR5*, *CCR7*, *CCR9*, *CX3CR1*, *CXCR1*, *CXCR2*, *CXCR4*, *CXCR5*, *CXCR6*(*STRL33*) and *Bob*(Table [3](#T3){ref-type="table"}). Expression of the same receptors could be observed on the non-RA membranes with the exception of *Bob*, *CCR7*and *CCR9*. *Bob*/*GPR15*is an orphan receptor that is a coreceptor for human and simian immunodeficiency viruses, and its expression in the RA synovium is a novel observation that might be worthy of further investigation. The detection of *CCR7*and *CCR9*in RA was only possible after extended exposure times, but at the time points used for quantification no regulation was demonstrated. Four chemokine receptors (*CCR2b*, *CCR3*, *CCR4*and *CCR6*) could not be detected in RA samples or non-RA samples under our conditions.
The most obvious differences between RA samples and non-RA samples were for the chemokine receptors *CXCR5*and *CXCR2*, and to a lesser extent *CXCR4*, which gave stronger signals in RA samples (Fig. [1](#F1){ref-type="fig"}). In order to quantify the differential expression of these receptors the densities of autoradiographic spots were measured using ArrayVision software (Table [3](#T3){ref-type="table"}). The criteria we set for a gene to be considered as upregulated or downregulated were a RA/non-RA ratio higher than 3 or lower than 0.3, respectively, and a 95% confidence interval below 10% (criteria as \[[@B30]\]). In the present study the expression of *CXCR5*and *CXCR4*was 22.6 ± 0.7-fold higher and 3.5 ± 0.1-fold higher in RA tissue than in non-RA tissue, respectively. These results indicated that, of the chemokine receptors studied, *CXCR5*was the most upregulated in RA (Table [3](#T3){ref-type="table"}). The upregulation of *CXCR2*could not be calculated for mathematical reasons because the signal intensity of *CXCR2*in non-RA tissue after correction for the background was zero. *CXCR2*was only visible on the non-RA autoradiogram upon prolonged exposure, at which point the housekeeping genes were saturated and were therefore unsuitable for quantification purposes.
Out of the 33 chemokines present on the arrays, 29 of these ligands were visible on the RA membranes and 21 on the non-RA membranes (Fig. [1](#F1){ref-type="fig"} and Table [3](#T3){ref-type="table"}). These included *CXCL13*, *CXCL12*, *CXCL8*, *CXCL1-3*and *CXCL5*, which are ligands for the chemokine receptors CXCR5, CXCR4 and CXCR2. Several chemokines were visible on RA microarrays but not on non-RA microarrays (namely *CXCL13*, *CCL21*and *CCL24*), suggesting that these genes might be induced in the inflamed synovium. In contrast, there were no chemokine signals that were present on non-RA membranes and were absent on RA membranes. Where chemokine signals were detectable on RA and non-RA microarrays, it was possible to quantify the degree of upregulation or downregulation using the criteria described earlier for chemokine receptors. Of these chemokines, the following showed upregulation: *CCL18*(4.5 ± 0.4-fold increase), *CXCL9*(3.6 ± 0.1-fold increase), *CXCL5*(3.5 ± 0.3-fold increase), and *CXCL8*(3.3 ± 0.2-fold increase). No chemokines displayed a downregulation with a RA/non-RA ratio less than 0.3. The upregulation of *CXCL9*in RA synovia is in agreement with the only microarray study of RA synovia, in which this chemokine was also shown to be increased \[[@B31]\]. In our study only five chemokines (*CXCL7*, *CCL13*, *CCL20*, *CCL17*and *CCL25*) could not be detected at all, whether in RA or non-RA samples.
The rapid screening of several genes at once made array technology a very attractive method. Its use has revealed disadvantages, however, including the requirement for large amounts of RNA (which are not always available from human tissue biopsies), a susceptibility to experimental variability and a lack of standard optimum methods for statistical analysis \[[@B32]\]. Arrays also present the risk of cross-hybridization leading to false positive or negative results \[[@B31]\]. However, the array approach remains a valuable tool if the samples can be pooled and if it is used in conjunction with alternative methods such as RT-PCR.
RT-PCR analysis of *CXCR5*
--------------------------
To confirm the array results and to examine individual patients, RT-PCR was performed on the total RNA from each patient sample (Fig. [2](#F2){ref-type="fig"}). PCR primers were run through the BLAST program (available through the UK MRC HGMP-RC website: <http://www.hgmp.ac.uk>) to ensure the gene specificity of the RT-PCR results and to exclude the possibility of cross-hybridization with other genes. Overall, CXCR5 RNA was more abundant in RA patients than in non-RA patients, confirming the microarray data. *CXCR5*expression was detected in the synovia of all eight RA patients and showed some degree of patient-to-patient variation. The difference in *CXCR5*expression between RA patients and non-RA patients was unlikely to be due to differences in the relative amount of cDNA produced by different RT reactions since the PCR reactions were normalized using the ribosomal gene *L27*. RT-PCR showed that the difference between RA patients and non-RA patients was less marked for *CXCR2*and *CXCR4*than for *CXCR5*(data not shown).
Immunohistochemistry
--------------------
To identify the cell types expressing *CXCR5*, and since RNA expression and protein expression do not always correlate, the protein expression of this receptor and three specific cell markers (CD20, CD3 and CD68) was investigated by immunohistochemistry of paraffin-embedded sections.
Seven out of eight RA patients presented substantial lymphoid follicles in their synovia. The specific cell markers CD20 and CD3 confirmed the presence of B cells and T cells, respectively, in these infiltrates. In every RA patient where lymphoid follicles occurred, CXCR5^+^cells were always present in these structures; this indicates a correlation between the expression of CXCR5 and the occurrence of lymphoid follicles. Serial sections indicated that CXCR5 was expressed by CD20^+^B cells (Fig. [3a,3c](#F3){ref-type="fig"}).
It was not possible to colocalize CXCR5 and CD3 in serial sections, so a double-label immunohistochemistry technique was developed. Sections were treated with anti-CD3 followed by alkaline phosphatase and Vector red substrate. Anti-CXCR5 was added to the same sections, and the colour developed using peroxidase and DAB-Nickel. CD3 expression alone gave a light red colour (Fig. [3e](#F3){ref-type="fig"}) and CXCR5 expression alone produced a grey--black colour (Fig. [3f](#F3){ref-type="fig"}). Where these two proteins colocalized a dark red colour was obtained (Fig. [3f](#F3){ref-type="fig"}). Using this technique it was evident that in the RA synovium there was a population of CD3^+^T cells that expressed CXCR5 (Fig. [3f](#F3){ref-type="fig"}). These were localized exclusively in lymphoid follicles in the synovia of five out of the eight RA patients. The patient with diffuse infiltration was negative for CXCR5^+^/CD3^+^cells. Serial sections treated with anti-CXCR5 and the macrophage marker anti-CD68 suggested that CXCR5^+^cells in the intima included macrophages (Fig. [4a,4b](#F4){ref-type="fig"}). The endothelial cells of synovial postcapillary venules were positive for CXCR5 in the RA synovium (Fig. [4e](#F4){ref-type="fig"}).
In non-RA tissue, CXCR5 was localized in the intima and endothelial cells (Fig. [5](#F5){ref-type="fig"}). Intimal cells were widely positive for CXCR5 and serial sections indicated that these included CD68^+^macrophage-like cells (Fig. [5a,5b](#F5){ref-type="fig"}). No lymphocytic infiltrates were present in these synovial samples due to their non-inflamed nature. Sections treated with CD20 and CD3 antibodies were negative, showing that no B cells or T cells could be detected. In non-RA tissue and RA tissue, fibroblasts were negative for CXCR5, as were neutrophils in RA synovia, indicating selectivity in the cell types expressing this receptor.
For all immunohistochemistry experiments in this study, the use of isotype-matched immunoglobulin controls or sera instead of primary antibodies resulted in negative staining of RA sections and non-RA sections (Figs [3b,3d,3g,3h](#F3){ref-type="fig"}, [4c,4d,4f](#F4){ref-type="fig"} and [5c,5d,5f](#F5){ref-type="fig"}).
RT-PCR on isolated RA monocytes/macrophages
-------------------------------------------
To further investigate whether macrophages themselves are producing CXCR5 and to confirm the results of immunohistochemistry, we performed RT-PCRs on monocytes/macrophages isolated from the PB and SF of four additional RA patients (Fig. [6](#F6){ref-type="fig"}). *CXCR5*was strongly expressed in all four samples and there was little difference between PB and SF.
Discussion
==========
The major finding of the present study is that CXCR5 is upregulated in the RA synovium. The cells expressing this chemokine receptor are B lymphocytes, T lymphocytes, macrophages and endothelial cells. The increased numbers of B lymphocytes, T lymphocytes and macrophages producing CXCR5 in the RA synovium are probably responsible for the increased expression of the receptor in this chronically inflamed tissue. The majority (seven out of eight) of the RA synovia included in this study contained substantial lymphoid aggregates but only one out of nine non-RA patients presented a very small infiltrate. These cell aggregates contained CD20^+^B cells that expressed CXCR5. The expression of CXCR5 has been reported in mature B cells and secondary lymphoid organs but as far as the authors are aware this is the first report of a chemokine receptor expressed by B cells in the RA synovium and its ectopic lymphoid structures.
Our findings are particularly interesting in view of the functional role of B cells in RA. This includes autoantibody production, antigen presentation, a role in lymphoid follicle and germinal centre formation, and the promising results of the anti-CD20 treatment in RA \[[@B33],[@B34]\]. The microarrays showed that the mRNA for CXCL13, the only known ligand of CXCR5, was present in the RA synovium and not in the non-RA synovium. Furthermore, other reports have shown a CXCL13 message in RA synovia, together with its protein that localizes to follicular dendritic cells, endothelial cells and synovial fibroblasts, suggesting that these cells produce the chemokine \[[@B1],[@B25]\]. Taken together with our data, this indicates that CXCR5 on B cells may be important in the recruitment of these cells into the RA synovium, in addition to their positioning and retention within the synovial infiltrates. In this regard, the role of CXCR5 on B cells in secondary lymphoid organs has been well documented \[[@B35],[@B36]\]. CXCR5 guides B cells into the B-cell follicles and also directly promotes the recruitment of these cells into Peyer\'s patches via high endothelial venules \[[@B27],[@B28],[@B37],[@B38]\]. In addition CXCR5-deficient mice exhibit impaired development of lymph nodes and Peyer\'s patches, and the tissue architecture of these organs is severely disturbed showing a lack of B-cell follicles \[[@B27],[@B28]\].
Our double immunohistochemistry data indicate that there is a population of CXCR5^+^CD3^+^T cells present in the RA synovium. CXCR5^+^T cells have been shown in secondary lymphoid tissue where some of these cells localize within germinal centres \[[@B20],[@B39]\], and it is proposed that CXCR5 enables them to enter B-cell follicles guided by CXCL13 \[[@B36]\]. Within these follicles they may provide B-cell help and have therefore been named follicular B helper T cells, since purified human tonsillar CD4^+^CXCR5^+^T cells efficiently stimulate the production of immunoglobulins by B cells \[[@B39],[@B40]\]. These follicular B helper T cells are CD57^+^, whereas the majority of the CXCR5^+^T cells that are present in interfollicular and T-cell areas of the lymphoid tissue are CD57^-^and are poor B-cell helpers \[[@B41]\]. Since lymphoid neogenesis occurs in the RA synovium it is possible that the CXCR5 expression on T cells as shown in the present study is involved in the positioning of these cells within the synovium and in providing B-cell help, although further studies are required to characterize this synovial T-cell population. Whether the two populations of CXCR5^+^CD57^+^and CXCR5^+^CD57^-^T cells are present in the RA synovium and what their role could be is still unknown. However, CD57^+^T cells are reported to be present in the RA synovium and SF, where levels of this marker are elevated compared with controls \[[@B42],[@B43]\]. Furthermore, an involvement of CD57^+^T cells has been shown in disease activity of RA \[[@B44]\].
Immunohistochemical experiments indicated that CD68^+^cells in the synovial intima express CXCR5. Intimal cells comprise two cell types: macrophage-like cells and fibroblast-like cells. In RA the former macrophage-like cells are numerous, comprising up to 80% of this cell layer \[[@B45]\]. It has been reported that in the RA synovium anti-CD68 reacts strongly with intimal macrophages, but fibroblasts also show some reactivity with this antibody \[[@B45]\]. Therefore, since macrophages are abundant in the RA intima and because of their strong reactivity with anti-CD68, it is likely that intimal macrophages are positive for CXCR5. In the normal non-RA intima, macrophages are positive for CD68 and fibroblasts are negative, making it more certain that macrophages express CXCR5 in this cell layer \[[@B45]\]. Consequently, RT-PCR was performed to verify that RA macrophages/monocytes can express *CXCR5*. The RT-PCR did indeed demonstrate CXCR5 mRNA in macrophages from RA SF, as well as PB monocytes from the same RA patients. Since the CXCR5 mRNA is expressed at similar levels in RA PB and RA SF it is suggested that the contribution of monocytes/macrophages to the upregulation of CXCR5 in the RA synovium is due to their increased number, rather than due to an increased abundance of CXCR5 transcripts per cell. CXCR5 mRNA has also been found in normal human PB monocytes by RT-PCR \[[@B21]\]. Studies by ourselves and other workers have shown that monocytes/macrophages express several other CXC chemokine receptors in RA, including CXCR1, CXCR2 and CXCR4 \[[@B15],[@B16],[@B46]\]. Furthermore, RA monocytes/macrophages express CC chemokine receptors such as CCR1, CCR2, CCR3 and CCR5 \[[@B14]\], illustrating their broad profile of chemokine receptor expression.
Endothelial cells are another cell type expressing CXCR5 in the synovium. There have been several reports of endothelial cells in the RA synovium expressing chemokine receptors, including CXCR3 and CXCR4, in addition to the Duffy antigen that is a non-signalling chemokine receptor \[[@B18],[@B47]-[@B49]\]. In the RA synovium there is increased turnover of blood vessels with enhanced formation of new blood vessels together with enhanced vascular regression \[[@B50],[@B51]\]. These mechanisms are regulated by the balance of angiogenic and angiostatic factors, and these factors include chemokines. Some chemokines are angiogenic (e.g. CXCL8, CXCL12, CCL1 and CCL2) and other chemokines are angiostatic (e.g. CXCL9 and CXCL10), and activation of their respective chemokine receptors results in the stimulation of or inhibition of endothelial cell proliferation \[[@B47],[@B52]-[@B57]\]. CXCL13 has been shown to have an angiostatic function, inhibiting the angiogenic effects of FGF-2 on human umbilical vein endothelial cells \[[@B58]\]. In addition, the presence of CXCR5 in a variety of cultured human endothelial cells -- from umbilical and saphenous veins, for example -- may mediate the angiostatic effects of CXCL13 \[[@B58],[@B59]\]. Our data showing the presence of CXCR5 on endothelial cells in the synovium and the presence of its ligand in this tissue \[[@B1],[@B25]\]suggest that CXCR5 may play an angiostatic role in RA pathophysiology, although the angiostatic effects of CXCL13 could potentially be acting through CXCR3, which is also expressed by the RA synovial endothelium \[[@B48],[@B60]\].
In the present study mRNA for other chemokine receptors were detected in the RA synovium, such as CXCR1, CXCR2, CXCR4, Bob, CCR1, CCR2a, CCR7, CCR9 and CX3CR1 (CXCR3 was not on the microarray). All of these showed variable degrees of increased mRNA expression in RA, although the upregulation was less compared with that of CXCR5. Several previous reports have shown the expression of chemokine receptors by leukocytes from RA joints. These have included CCR4--CCR6, CXCR3, CXCR4 and CX3CR1 by T lymphocytes \[[@B9],[@B12],[@B13],[@B19]\] and CCR1--CCR5 and CXCR1--CXCR4 by monocytes/macrophages \[[@B14]-[@B16],[@B18]\]. Such reports mainly focused on selected cell types and certain chemokine receptors. The present study took a different approach. Ours was primarily a whole-tissue study examining the mRNA expression of a wide range of chemokine receptors in RA and control synovia. While our study is in general accord with previous reports, differences may in part be due to the RA ST used. This tissue was highly infiltrated and, in all but one sample, had extensive lymphoid follicles bearing resemblance to those of secondary lymphoid organs. This feature may be responsible for the particular upregulation of the constitutive chemokine receptor CXCR5. In addition, our RA patients had long-standing disease (Table [1](#T1){ref-type="table"}) and the patient sample may also have influenced the types of chemokine receptors expressed.
Conclusion
==========
Our study demonstrates the expression of CXCR5 on B cells, on T cells, on monocytes/macrophages and on endothelial cells in the RA synovium. The expression of a marker shared by cells that are known to play a central role in the process of chronic inflammation is of particular interest and suggests that targeting CXCR5 could provide a powerful new treatment for RA.
Abbreviations
=============
DAB = 3,3\'-diaminobenzidine; H&E = haematoxylin and eosin; PB = peripheral blood; PBS = phosphate-buffered saline; PCR = polymerase chain reaction; RA = rheumatoid arthritis; RT = reverse transcription; SF = synovial fluid; ST = synovial tissue.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
CS carried out the microarray work, the RT-PCR and the double immunohistochemisty, and drafted the manuscript. AH carried out the single colour immunohistochemistry. AB isolated the peripheral blood and synovial fluid monocytes, and isolated the RNA after adhesion. BA participated in the design of the study. CB and MS collaborated on the study or coordinated the collection of samples in Birmingham, and contributed to the writing of the manuscript. JM conceived the study, and participated in its design and in the writing the manuscript. All authors read and approved the final manuscript.
Acknowledgements
================
The authors are indebted to the patients who kindly agreed to take part in this study. They thank Mr C McGeoch, Mr D Rees, Mr L van Niekerk and Mr S White and the theatre teams for their help in obtaining synovial tissue. They are also very grateful to P Evans, M Pritchard and N Harness for their histological expertise and to J Menage for helpful immunohistochemistry discussion. Finally, the authors thankfully acknowledge the Henry Smith Charity, Droitwich Medical Trust Ltd and the Orthopaedic Institute Ltd for their financial support.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Microarray analysis of chemokine and chemokine receptor expression in the rheumatoid arthritis (RA) and non-RA synovia. A pair of human cytokine array membranes were hybridized to ^33^P-labelled cDNA probes prepared from pools of **(a)**RA mRNA (*n*= 8) and **(b)**non-RA mRNA (*n*= 9). The membranes were washed and autoradiographed. **(c)**The position of the 33 chemokines (C), the 16 chemokine receptors (CR), the nine positive control \'housekeeping genes\' (HKG) and the six negative controls (NC). Each gene was printed in duplicate. The star indicates the position of the genes *CXCR1*, *CXCR2*, *CXCR4*and *CXCR5*(reading top to bottom) and shows their differential expression in RA and non-RA synovia. Exposure time was 7 days and 14 days for (a) and (b), respectively.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
RT-PCR on rheumatoid arthritis and non-rheumatoid arthritis synovial tissue. CXCR5 RT-PCR products were separated on 0.8% agarose gels and stained with ethidium bromide. The reactions were performed on the individual RNA samples that were applied to the microarray membranes. The ribosomal RNA L27 was employed to normalize the amount of RNA used in each reaction.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Immunohistochemistry of CXCR5 in lymphoid cell aggregates of rheumatoid arthritis synovia. Sections of rheumatoid synovium were treated with **(a)**CXCR5 antibody or **(b)**isotype control. Serial sections were treated with **(c)**anti-CD20 as a marker of B lymphocytes or **(d)**isotype control. Arrows indicate B lymphocytes expressing CXCR5. **(e)**Rheumatoid synovium treated with the T-cell marker anti-CD3 followed by alkaline phosphatase and Vector red substrate (methyl green counterstain). T cells stain a light red colour. **(f)**Serial section from the same synovial sample as (e) treated with anti-CD3, alkaline phosphatase and Vector red followed by anti-CXCR5, peroxidase and 3,3\'-diaminobenzidine (DAB)-Nickel substrate (no counterstain used). T cells that express CXCR5 are stained dark red whereas cells expressing CXCR5 alone are grey--black in colour. **(g)**Control for (e), in which isotype-matched rabbit immunoglobulin (Ig) was used instead of anti-CD3. **(h)**Control for (f), in which isotype-matched rabbit and mouse Ig were applied instead of CD3 and CXCR5 antibodies (no counterstain used). Unless stated otherwise, DAB substrate was used. (a), (c) and (e)--(h) Original magnification, 420 ×; isotype controls (b) and (d) original magnification, 280 ×.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Immunohistochemistry of CXCR5 in the intima and postcapillary venules in rheumatoid arthritis synovia. **(a)**CD68^+^cells in the intima. **(b)**Serial section to (a) stained for CXCR5. Note the colocalization of CXCR5 and CD68 to the same group of cells. **(c)**and **(d)**Sections from the same region as (a) and (b), treated with isotype-matched control immunoglobulin instead of CD68 and CXCR5 antibodies, respectively. **(e)**Postcapillary venule positive for CXCR5 within a lymphoid aggregate. Labelling was revealed using 3,3\'-diaminobenzidine substrate. **(f)**Isotype control for (e). (a), (b), (e) and (f) Original magnification, 420 ×; (c) and (d) original magnification, 350 ×.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Immunohistochemistry of CXCR5 in non-rheumatoid arthritis synovia. **(a)**CD68 staining in the intimal layer. **(b)**Serial section to (a) treated with anti-CXCR5, showing that CXCR5^+^cells in the intimal layer included those also positive for CD68. **(c)**and **(d)**Sections from the same region as (a) and (b), treated with isotype-matched control immunoglobulin instead of CD68 and CXCR5 antibodies, respectively. **(e)**Subintimal postcapillary venule stains for CXCR5 expression (arrow). **(f)**Isotype-matched control for (e). Labelling was revealed using 3,3\'-diaminobenzidine substrate. (a), (b), (e) and (f) Original magnification, 420 ×; (c) and (d) original magnification, 350 ×.
:::
:::
::: {#F6 .fig}
Figure 6
::: {.caption}
######
RT-PCR on monocytes/macrophages from peripheral blood (PB) and synovial fluid (SF). CXCR5 RT-PCR products were separated on 0.8% agarose gels and stained with ethidium bromide. The reactions were performed on four additional rheumatoid arthritis patients. The ribosomal RNA L27 was employed to normalize the amount of RNA used in each reaction.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Details of rheumatoid arthritis (RA) and non-RA patients
:::
Patient (sex, age \[years\]) Diagnosis/pathology Disease duration (years) Medication
------------------------------ ---------------------------------------------- -------------------------- ----------------------------------
1 (male, 69) RA 37 Auranofin, NSAID
2 (female, 41) RA 23 NSAID
3 (female, 51) RA 10 NSAID, analgesic
4 (female, 79) RA 4 Methotrexate, NSAID, steroid
5 (female, 70) RA 8 Penicillamine, steroid, NSAID
6 (female, 63) RA + secondary osteoarthritis 39 Methotrexate, steroid, analgesic
7 (female, 33) juvenile chronic arthritis 20 NSAID
8 (female, 66) RA 38 NSAID, analgesic
1 (female, 50) Articular cartilage damage \<1 NSAID, analgesic
2 (male, 44) Meniscal tear \<1 \-
3 (male, 42) Meniscal tear and articular cartilage damage 4 \-
4 (male, 60) Meniscal degeneration \<1 \-
5 (male, 52) Articular cartilage degeneration 1 \-
6 (male, 46) Meniscal tear \<1 \-
7 (male, 53) Meniscal degeneration \<1 \-
8 (male, 38) Meniscal tear 1 Steroid
9 (male, 43) Articular cartilage degeneration \<1 \-
1 (female, 91) RA \<1 Analgesic
2 (female, 56) RA \<1 NSAID, steroid, methotrexate
3 (male, 67) RA \-
4 (male, 67) RA 4 Analgesic, NSAID, methotrexate
Synovia were obtained from eight RA patients and nine non-RA patients.
Monocytes/macrophages from peripheral blood/synovial fluid were obtained from the last four patients. NSAID, non-steroidal anti-inflammatory drug.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Sequences of the primers used for RT-PCR
:::
mRNA Product Sequence Size (base pairs) Accession number
------- --------- ------------------------------------ ------------------- ------------------
L27 Forward 5\'-GACGCAAAGCTGTCATCGTG-3\' 344 BC007273
Reverse 5\'-GCAGTTTCTGGAAGAACCAC-3\'
CXCR5 Forward 5\'-TGA CCT GAG GAA GCG TGA AG-3\' 639 NM001716
Reverse 5\'-CGT GAA GAC ACT CTC ACG TG-3\'
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Chemokine and chemokine receptor expression data analysis
:::
Gene RA Non-RA Regulation (ratio RA/non-RA)
----------------------------- ---------- ---------- ------------------------------
**Receptors**
Array column 3 (Fig. 1, C3)
CCR1 0.050 0.023 Up (2.2 ± 0.2)
CCR2a 0.031 0.012 ○● Up (2.7 ± 0.2)
CCR2b 0.000 ○ 0.000 ○ Not visible (NA)
CCR3 0.000 ○ 0.000 ○ Not visible (NA)
CCR4 0.003 ○ 0.000 ○ Not visible (NA)
CCR5 0.042 0.000 ○● Up (NA)
CCR6 0.001 ○ 0.000 ○ Not visible (NA)
CCR7 0.022 ○● 0.011 ○ Not visible (2.5 ± 0.9)
CCR9 0.001 ○● 0.000 ○ Not visible (NA)
CX3CR1 0.028 ○● 0.019 ○● Not visible (1.5 ± 0.1)
CXCR1 0.036 ○● 0.008 ○● Not visible (9.6 ± 5.6)
CXCR2 0.651 0.000 ○● Up (NA)
CXCR4 0.190 0.055 Up (3.5 ± 0.1)
CXCR5 1.328 0.059 Up (22.6 ± 0.7)
CXCR6 (STRL33) 0.016 0.025 Not visible (0.7 ± 0.0)
Bob 0.034 0.000 ○ Up (NA)
**Chemokines**
Array column 1 (Fig. 1, C1)
CCL21 (6Ckine) 0.030 0.019 ○ Up (1.6 ± 0.1)
CXCL13 (BLC/BCA-1) 0.052 0.020 ○ Up (2.8 ± 0.4)
CXCL10 (IP-10) 0.031 0.019 ○● Up (1.6 ± 0.1)
CXCL5 (ENA-78) 0.050 0.015 ○● Up (3.5 ± 0.3)
CCL11 (eotaxin) 0.050 0.052 Not visible (1.0 ± 0.0)
CCL24 (eotaxin-2) 0.032 0.013 ○ Up (2.5 ± 0.3)
CX3CL1 (fractalkine) 0.030 ○● 0.012 ○ Not visible (2.5 ± 0.2)
CXCL1 (GRO-α) 0.090 0.057 Up (1.6 ± 0.0)
CXCL2 (GRO-β) 0.151 0.101 Up (1.5 ± 0.0)
CXCL3 (GRO-γ) 0.101 0.055 Up (1.8 ± 0.0)
CCL14 (HCC-1) 0.202 0.296 Down (0.7 ± 0.0)
CCL16 (HCC-4) 0.006 ○● 0.000 ○ Not visible (NA)
CCL1 (I-309) 0.007 ○● 0.000 ○ Not visible (NA)
CXCL8 (IL-8) 0.086 0.026 Up (3.3 ± 0.2)
CXCL7 (LDGF) 0.005 ○ 0.000 ○ Not visible (NA)
CCL15 (MIP-1δ) 0.009 ○● 0.008 ○ Not visible (11 ± 13)
XCL1 (lymphotactin) 0.048 0.031 Up (1.6 ± 0.0)
CCL2 (MCP-1) 0.482 0.545 Not visible (0.9 ± 0.0)
CCL8 (MCP-2) 0.053 0.062 Not visible(0.9 ± 0.0)
CCL7 (MCP-3) 0.066 0.047 Up (1.4 ± 0.1)
CCL13 (MCP-4) 0.000 ○ 0.002 ○ Not visible (0.0)
CCL22 (MDC) 0.000 ○● 0.008 ○ Not visible (0.0)
Array column 2 (Fig. 1, C2)
Midkine 0.299 0.155 Up (1.9 ± 0.0)
CXCL9 (MIG) 0.166 0.047 Up (3.6 ± 0.1)
CCL3 (MIP-1α) 0.040 0.042 Not visible (1.0 ± 0.0)
CCL4 (MIP-1β) 0.023 0.018 ○● Up (1.3 ± 0.2)
CCL20 (MIP-3α) 0.000 ○ 0.013 ○ Not visible(0.0)
CCL19 (MIP-3β) 0.023 0.042 Not visible (0.6 ± 0.0)
CCL23 (MPIF-1) 0.039 0.042 Up (0.9 ± 0.0)
CCL18 (PARC) 0.146 0.033 Up (4.5 ± 0.4)
CCL5 (RANTES) 0.037 0.031 Up (1.2 ± 0.1)
CXCL12 (SDF-1) 0.409 0.573 Down (0.7 ± 0.0)
CCL17 (TARC) 0.000 ○ 0.000 ○ Not visible (NA)
CCL25 (TECK) 0.000 ○ 0.000 ○ Not visible (NA)
Following hybridization to labelled mRNA extracted from rheumatoid arthritis (RA) and non-RA synovia, a pair of array membranes was autoradiographed for varying lengths of time. The autoradiograms were scanned and analysed with the ArrayVision software (version 6.0; Imaging Research Inc., Haverhill, UK). For each RA/non-RA pair the housekeeping genes on the membranes showed very similar intensities, were not saturated and were used to normalize the data. The analysis measured the \'volume\' of each spot (i.e. the density value of each spot multiplied by its area). The background was measured using the \'corners between spots\' protocol of the software and was deducted from the \'volumes\'. The ratio of RA synovia versus non-RA synovia was also calculated for each spot. The analysis was repeated eight times for each pair of autoradiograms, providing 16 values for each gene (each gene is spotted in duplicate) on each pair. Figures in the columns RA, non-RA and ratio RA/non-RA represent the average of 16 values. For each average ratio the 95% confidence level was calculated, and the results presented are those from the autoradiogram pair giving the smallest variation. ○, spot was not visible by eye on the corresponding autoradiogram; ●, spot was visible after prolonged exposure. The mRNA regulation of RA versus non-RA as observed by eye at the time point used for quantification is indicated by not visible, up or down. NA, ratio could not be calculated due to the presence of zero values. The recent systematic nomenclature of chemokines is used, with the former names in parentheses. The order of the genes presented is the same as that appearing on the microarray in Fig. 1.
:::
|
PubMed Central
|
2024-06-05T03:55:54.927184
|
2004-12-16
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065316/",
"journal": "Arthritis Res Ther. 2005 Dec 16; 7(2):R217-R229",
"authors": [
{
"first": "Caroline",
"last": "Schmutz"
},
{
"first": "Alison",
"last": "Hulme"
},
{
"first": "Angela",
"last": "Burman"
},
{
"first": "Mike",
"last": "Salmon"
},
{
"first": "Brian",
"last": "Ashton"
},
{
"first": "Christopher",
"last": "Buckley"
},
{
"first": "Jim",
"last": "Middleton"
}
]
}
|
PMC1065317
|
Introduction
============
Tumour necrosis factor (TNF)-α has been found to play a central role in the pathogenesis of rheumatoid arthritis (RA), which has led to the rational development of novel drug therapies that neutralize the deleterious effects of this cytokine \[[@B1],[@B2]\]. Several studies have shown dramatic therapeutic effects of anti-TNF-α antibodies, both in experimental collagen-induced arthritis and in the treatment of inflammatory diseases such as rheumatoid arthritis (RA) \[[@B3]-[@B5]\], psoriatic arthritis \[[@B6]\], juvenile rheumatoid arthritis \[[@B7]\] and Crohn\'s disease \[[@B8]\]. The role played by phagocytic cells in the pathogenesis of these inflammatory diseases \[[@B9]-[@B11]\] and the capacity of TNF-α to prime and/or activate phagocytic cells \[[@B12]\] suggest, at least in part, that downregulation of phagocyte activity may be involved in the mechanism of action of anti-TNF-α therapy \[[@B9]\].
There is increasing evidence that inhibition of TNF-α may be associated with the development of adverse consequences such as carcinogenesis, autoimmune disorders and, importantly, infectious diseases caused by Gram-positive and Gram-negative bacteria, mycobacteria and fungi (for review, see Olsen and Stein \[[@B2]\]). Again, the role played by TNF-α in the activation of phagocytic cells and the involvement of these cells in the host defence against infections suggest that impairment in phagocytic cell activity may heighten the risk for infection during TNF-α neutralization \[[@B13]\].
Few data have been reported on the effect of anti-TNF-α therapy on neutrophil function *ex vivo*. Decreased influx of neutrophils in inflamed joints was reported by Taylor and coworkers \[[@B14]\] in RA patients treated with infliximab (a chimeric anti-TNF-α mAb) and by Den Broeder and coworkers \[[@B15]\] in patients treated with adalimumab (a fully human anti-TNF-α mAb). However, no significant impairment in *ex vivo*neutrophil function was observed in RA patients treated with etanercept (a soluble human p75 TNF receptor) \[[@B16]\] or with adalimumab \[[@B15]\].
In this work we evaluated certain phenotypic and functional aspects of neutrophils obtained from RA patients during treatment with adalimumab. To this end, chemotaxis, phagocytosis and reactive oxygen species (ROS) production were assessed in peripheral blood neutrophils, together with membrane expression of CD11b and CD69 -- two functionally different activation molecules \[[@B17]\].
Methods
=======
Reagents
--------
The anti-CD69 mAb (IgG~2a~, clone HP-4B3) was obtained from Calbiochem (La Jolla, CA, USA). The anti-CD11b was OKM1 (mouse IgG~2~isotype; Ortho Diagnostics, Raritan, NJ, USA). FITC-conjugated goat anti-mouse IgG was from Immunotech SA (Marseille, France). Irrelevant class-matched mAbs were used as controls for nonspecific binding (Becton Dickinson, San Jose, CA, USA). Lymphoprep gradient (density 1.077 g/ml) was purchased from Nyegaard (Oslo, Norway). RPMI 1640 was obtained from HyClone Laboratories (Logan, UT, USA). Bovine serum albumin (BSA), *N*-formyl-methionyl-leucyl-phenylalanine (FMLP), phorbol 12-myristate 13-acetate (PMA), lucigenin (bis-*N*-methylacridinum nitrate) and zymosan A were from Sigma Chemical Company (St. Louis, MO, USA).
Patients
--------
Peripheral blood samples were collected from10 selected and consenting RA patients who satisfied the American College of Rheumatology 1987 criteria \[[@B18]\], who had active disease (defined as a disease activity score 28 \> 3.2) \[[@B19]\], and who were enrolled in a European open-label, multicentre, multinational phase IIIb study (the Adalimumab Research in Active RA \[ReAct\] study \[[@B20]\]). The study was approved by the ethical committee of the Ospedale L Sacco (Milan, Italy). The mean age of the patients was 61.4 years (range 40--83 years); eight were rheumatoid factor positive and two were rheumatoid factor negative. Three months before and during the study, all patients received hydroxychloroquine (200 mg twice daily), methotrexate intramuscularly (7.5--15 mg/week), and no more than 10 mg/day prednisone. Adalimumab was administered subcutaneously every other week (40 mg). Peripheral blood samples were obtained before anti-TNF-α therapy and immediately before administration of adalimumab at weeks 2, 6 and 12. Controls were 20 healthy individual who were matched to the patients with respect to age and sex.
*Ex vivo*neutrophil function
----------------------------
Peripheral blood neutrophils were obtained by density gradient centrifugation (Lymphoprep) \[[@B21]\]. The purified cells consisted of a more than 95% pure population of viable neutrophils, as assessed by morphology and trypan blue exclusion test.
Neutrophil chemotaxis was evaluated using a modified Boyden chamber assay, with blind well chambers and 3 μm micropore filters (Millipore, Bedford, MA, USA) \[[@B22]\]. Briefly, 200 μl of the cell suspension, containing 2.5 × 10^6^neutrophils/ml in RPMI1640 + 0.4% BSA were layered on top of the filter, and the lower compartment was filled with 200 μl of the chemotactic factor (see below). Following incubation at 37°C for 90 min in a humidified atmosphere with 5% carbon dioxide, the filters were fixed with ethanol and stained with haematoxylin--eosin. The chemotactic response was then determined by evaluating the number of cells × high power field that had migrated through the entire thickness of the filter; triplicate chambers were used in each experiment and five fields were examined in each filter. In all cases the person scoring the assay had no knowledge of the experimental grouping. The chemoattractants were zymosan-activated serum (1 mg/ml for 30 min at 37°C) at a 10% (vol/vol) final dilution in RPMI 1640, and the synthetic peptide FMLP at 10^-8^mol/l final concentration.
Phagocytosis was evaluated using C3-coated zymosan (C3Zy) as particles for uptake \[[@B23]\]. C3Zy was prepared incubating zymosan in normal human serum (5 mg/ml) for 30 min at 37°C followed by extensive washing. The neutrophil suspension (200 μl) was incubated with C3Zy (cell to particle ratio, 1:5) for 30 min at 37°C in a shaking water bath. Cytocentrifuge slides of the mixtures were then immediately prepared and stained with May Grunwald--Giemsa. The number of particles ingested per cell (phagocytic index \[PI\]) were established by direct light microscopy (1000 × magnification) of at least 200 cells. In all cases the person scoring the assay had no knowledge of the experimental grouping.
Lucigenin-amplified chemiluminescence was used to evaluate production of ROS by neutrophils \[[@B23]\]. For the measurement of chemiluminescence, 1 × 10^5^neutrophils were mixed in 3 ml polystyrene vials with 5 × 10^-5^mol/l lucigenin in a final volume of 700 μl. The vials were placed in the Luminometer 1251 (LKB Wallace, Turku, Finland) in the dark and allowed to equilibrate for 5 min at 37°C with intermittent shaking previously to record the background of the light output in mV. PMA (final concentration 5 ng/ml) was added with an appropriate dispenser (1291; LKB Wallace) and chemiluminescence was recorded continuously. Background counts were subtracted from the values obtained after neutrophil stimulation.
Levels of neutrophil membrane expression of CD69 and CD11b were evaluated as previously reported \[[@B23]\]. Briefly, 2 × 10^5^neutrophils were washed in phosphate-buffered saline (PBS) and resuspended with 100 μl PBS containing 0.1% NaN~3~, 10% human AB serum (to prevent nonspecific binding of mAb to Fc receptors) and predetermined saturating concentrations of the anti-CD69 or anti-CD11b mAbs. After incubation for 60 min at 4°C the cells were washed twice with PBS/NaN~3~/0.1% BSA and the pellets were resuspended in 100 μl of the same buffer containing FITC-conjugated goat anti-mouse IgG in a saturating concentration and incubated for 30 min at 4°C. The cells were then washed twice in PBS and resuspended in 0.5 ml of ice-cold 2% paraformaldehyde in PBS (pH 7.2). The percentage of neutrophils positive for CD69 or CD11b was quantified within 24 hours on a FACSscan flow cytometer (Becton Dickinson). A relative measure of antigen expression was obtained using the mean fluorescence intensity (MFI), converted from log to linear scale, after subtraction of the cells\' autofluorescence and the fluorescence of cells incubated with irrelevant isotype control mAbs.
Statistical analysis
--------------------
Data are expressed as mean ± standard error of the mean. Statistical analysis was performed using the Student\'s t-test for unpaired or paired data as appropriate. *P*\< 0.05 was considered statistically significant.
Results
=======
The chemotactic activity of neutrophils obtained from RA patients at baseline was significantly impaired as compared with that in neutrophils from control individuals; the defect was evident both using zymosan-activated serum (*P*\< 0.001; Fig. [1](#F1){ref-type="fig"}) and FMLP (*P*\< 0.02; Fig. [2](#F2){ref-type="fig"}) as chemoattractant. Two weeks after the start of therapy with adalimumab, the neutrophil chemotactic responsiveness was significantly improved (Figs [1](#F1){ref-type="fig"} and [2](#F2){ref-type="fig"}), with no differences between patients and control individuals. The improvement was evident and persistent during anti-TNF-α therapy at weeks 6 and 12 (Figs [1](#F1){ref-type="fig"} and [2](#F2){ref-type="fig"}).
The phagocytic capacity of neutrophils was similar between control individuals (PI 0.99 ± 0.03) and RA patients at baseline (PI 1.19 ± 0.32), and no changes were observed during anti-TNF-α therapy (week 2: 1.11 ± 0.03; week 6: 1.17 ± 0.09; week 12: 1.03 ± 0.04). The CD11b molecule was spontaneously expressed on more than 90% of neutrophils both in control individuals and in RA patients before and during anti-TNF-α therapy (data not shown). The level of both spontaneous and FMLP-induced CD11b membrane expression (MFI) was also similar between controls (MFI for spontaneous: 155.3 ± 3.7; MFI for FMLP-induced: 591.3 ± 13.9) and RA patients at baseline (MFI for spontaneous: 159.2 ± 8.5; MFI for FMLP-induced: 558.7 ± 27.1), as well as during adalimumab therapy (MFI for spontaneous, week 2: 166.3 ± 12.2; MFI for spontaneous, week 6: 161.0 ± 16.7; MFI for spontaneous, week 12: 154.4 ± 14.9; MFI for FMLP-induced, week 2: 503.6 ± 33.1; MFI for FMLP-induced, week 6: 547.8 ± 27.7; MFI for FMLP-induced, week 12: 610.2 ± 41.8).
Both spontaneous and PMA-induced production of ROS by RA neutrophils was slightly increased at baseline as compared with controls (*P*\< 0.05) and the differences persisted at all time points examined during adalimumab therapy (Fig. [3](#F3){ref-type="fig"}).
Although control neutrophils stained with anti-CD69 mAb yielded very low fluorescence, just above that of unstained cells (%CD69^+^cells: 1.3 ± 0.5; MFI: 1.0 ± 0.3), CD69 was significantly expressed on neutrophils from RA patients at baseline (%CD69^+^cells: 22.8 ± 5.4; MFI: 7.6 ± 1.4; *P*\< 0.001 versus controls; Fig. [4](#F4){ref-type="fig"}). As shown in Fig. [4](#F4){ref-type="fig"}, a significant inhibition of CD69 expression on RA neutrophils was induced by adalimumab therapy; the inhibition was already evident at week 2 after the start of therapy (%CD69^+^cells: 5.5 ± 0.9; MFI: 2.6 ± 0.6; *P*\< 0.01 versus RA baseline) but it was complete at weeks 6 and 12, when no differences were observed between RA patients and control individuals (Fig. [4](#F4){ref-type="fig"}).
Discussion
==========
The first aim of the study was to determine whether anti-TNF-α therapy could downregulate neutrophil function, thus reducing the antimicrobial host defence in patients with RA. Our *ex vivo*functional assays do not support this possibility. In fact, we demonstrated that TNF-α neutralization in RA patients did not modify neutrophil activities such as phagocytosis, which were normal at baseline, or ROS production, which was slightly increased at baseline. In agreement with previous studies \[[@B24],[@B25]\], we found impaired chemotaxis of neutrophils from RA patients toward two different chemoattractants. Unexpectedly, TNF-α neutralization induced complete reversal of the neutrophil chemotactic defect. Various mechanisms may account for the defective neutrophil migration in RA patients, such as saturation of membrane receptors with immune complexes \[[@B25]\], cytokine (TNF-α)-induced desensitization \[[@B26]-[@B28]\] and drug-induced cell toxicity \[[@B29]-[@B33]\]. Of particular relevance are the observations that TNF-α-primed neutrophils are less responsive to chemoattractants \[[@B26]-[@B28]\] and are more susceptible to the inhibitory effect of methotrexate on chemotaxis \[[@B31]\]. Because circulating TNF-α has been demonstrated in RA patients \[[@B34]\], it is possible that anti-TNF-α therapy improves neutrophil migration by removing the deleterious effect exerted by soluble and/or membrane bound TNF-α on these cells.
The second aim of the study was to determine whether downregulation of phagocyte activities are involved in the anti-inflammatory activity of anti-TNF-α therapy. The lack of activity on phagocytosis, ROS production or CD11b membrane expression, and the improved migration of neutrophils did not implicate neutrophils as targets of the therapeutic effect of anti-TNF-α. The improved chemotactic responsiveness we observed in patients during adalimumab therapy does not explain the decreased influx of neutrophils into synovial joints previously observed in RA patients during anti-TNF-α therapy \[[@B14],[@B15]\]. However, there is evidence that anti-TNF-α mAbs downregulate the expression of cytokine-inducible adhesion molecules on endothelial cells \[[@B35],[@B36]\]. The decreased activation of endothelial cells in the synovial microvasculature, rather than a defective neutrophil migration, could be responsible for the decreased homing of neutrophils to the inflamed joints.
We recently found that both synovial fluid and peripheral blood neutrophils from RA patients have increased membrane expression of CD69 \[[@B37]\], and this observation was confirmed in the present study. This activation molecule is not constitutively expressed on neutrophils but it may be induced on these cells *in vitro*by several cytokines, such as granulocyte--macrophage colony-stimulating factor, interferon-γ and interferon-α \[[@B23],[@B38]\]. Although a specific ligand for this molecule has not been identified, a role for CD69 in the pathogenesis of RA was previously suggested by Laffon and coworkers \[[@B39]\], who found that CD69^+^T lymphocytes were detectable at high levels in synovial fluid and synovial membrane from RA patients and correlated with disease activity. Furthermore, Murata and coworkers \[[@B40]\] recently reported that CD69-null mice were protected from collagen-induced arthritis, and that transfer of neutrophils from wild-type mice could restore arthritis in these animals. These data suggested a crucial role for CD69^+^neutrophils in the pathogenesis of arthritis and implicate the molecule as a possible therapeutic target for human arthritis. In the present study we observed that CD69 was downregulated (or inhibited) on neutrophils from RA patients during adalimumab therapy. The mechanism underlying this inhibition is not clear because, in our experience, TNF-α *per se*is not an inducer of CD69 on neutrophils. However, it is possible that other and as yet undefined CD69 inducers are indirectly inhibited by TNF-α neutralization. In agreement with our data, Moore and coworkers \[[@B41]\] recently reported decreased CD69 expression on natural killer cells obtained from mice treated with anti-TNF-α.
Conclusion
==========
In this study we found that administration of the anti-TNF-α mAb adalimumab to patients with RA does not interfere with the neutrophil activities that are required to maintain an adequate antimicrobial host defence capacity. On the other hand, the inhibitory activity of the mAb on CD69 membrane expression on neutrophils indicates that these cells are among the possible targets of anti-TNF-α activity in RA, and may provide an insight into a new and interesting mechanism of action of anti-TNF-α mAbs in the control of inflammatory arthritis.
Abbreviations
=============
BSA = bovine serum albumin; C3Zy = C3-coated zymosan; FMLP = *N*-formyl-methionyl-leucyl-phenylalanine; mAb = monoclonal antibody; MFI = mean fluorescence intensity; PBS = phosphate buffered saline; PI = phagocytic index; PMA = phorbol 12-myristate 13-acetate; RA = rheumatoid arthritis; ROS = reactive oxygen species; TNF = tumour necrosis factor.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
FC conceived the study, participated in conducting neutrophil functional assays and drafted the manuscript. FM conducted the neutrophil functional assays. PB conducted the immunofluorescence assays. PS-P participated in study design and coordination, and helped to select patients. FA helped with monitoring patients before and during the study. MC participated in coordination of the study. All authors read and approved the final manuscript. AD helped to perform statistical analysis.
Acknowledgments
===============
This work was supported by research funds FIRST 2003 (University of Milan) and by research funds \'Ricerca Corrente 2002\' Ospedale Maggiore IRCCS, Milan, Italy. We thank Abbott Laboratories and Abbott SpA for their funding of the ReAct study.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Effect of adalimumab on neutrophil chemotaxis. Peripheral blood neutrophils were purified from 20 controls and 10 patients with rheumatoid arthritis (RA) before (baseline) and during therapy with adalimumab at weeks 2 (w2), 6 (w6) and 12 (w12). Values represent the number of cells migrated × high power field (hpf) using zymosan-activated serum (ZAS) as chemoattractant. The dotted lines indicate the mean values.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Neutrophils were obtained as described in the legend to Figure 1 and then tested for chemotactic responsiveness toward the chemoattractant *N*-formyl-methionyl-leucyl-phenylalanine (FMLP). The dotted lines indicate the mean values. RA, rheumatoid arthritis.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Effect of adalimumab on neutrophil chemiluminescence production. Neutrophils were obtained as described in the legend to Fig. 1 and then tested for chemiluminescence (CL) production in resting conditions (spontaneous CL) or in response to 5 ng/ml phorbol 12-myristate 13-acetate (PMA-induced CL). Results are expressed as mean ± standard error of the mean of peak CL values. \**P*\< 0.05 versus control individuals. RA, rheumatoid arthritis.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Modulation of CD69 membrane expression on neutrophils by adalimumab. Neutrophils, obtained as described in the legend to Fig. 1, were labelled with anti-CD69 mAb by indirect immunofluorescence. Results are expressed as mean ± standard error of the mean of percentage of positive cells (% CD69-positive cells) and as mean fluorescence intensity (MFI) corrected for nonspecific staining. \**P*\< 0.001 versus controls; °*P*\< 0.01 versus rheumatoid arthritis (RA) baseline
:::
:::
|
PubMed Central
|
2024-06-05T03:55:54.931647
|
2005-1-10
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065317/",
"journal": "Arthritis Res Ther. 2005 Jan 10; 7(2):R250-R255",
"authors": [
{
"first": "Franco",
"last": "Capsoni"
},
{
"first": "Piercarlo",
"last": "Sarzi-Puttini"
},
{
"first": "Fabiola",
"last": "Atzeni"
},
{
"first": "Francesca",
"last": "Minonzio"
},
{
"first": "Paola",
"last": "Bonara"
},
{
"first": "Andrea",
"last": "Doria"
},
{
"first": "Mario",
"last": "Carrabba"
}
]
}
|
PMC1065318
|
Introduction
============
Osteoarthritis is the most common disabling condition of humans in the western world. Although osteoarthritis is mainly a disease and functional loss of the articular cartilage covering the joint surfaces, it is clearly the cells that are the active players during the disease process \[[@B1]\]. Whatever pleomorphisms the cellular reaction patterns display at first sight during the osteoarthritic disease process, they can be basically summarized in three categories (reviewed in \[[@B2]\]). First, the chondrocytes can degenerate or proliferate. Second, chondrocytes can activate or deactivate their synthetic anabolic or catabolic matrix-degrading activity by increasing or decreasing anabolic or catabolic gene expression. Last, chondrocytes can undergo phenotypic modulations implicating an overall severely altered gene expression profile of the cells in the diseased tissue. In fact, several distinct phenotypes of chondrocytes are known to occur *in vitro*, *in vivo*during fetal development and potentially also in the disease process itself, but new markers are required for the more accurate characterization of cellular behavior \[[@B3]\]. This will allow further analysis of the underlying pathology to develop therapeutic approaches that could delay, stop, or even reverse cartilage degeneration.
In many laboratories single and multiple gene analyses have been performed on normal and osteoarthritic cartilage specimens; however, a global overview of disease-associated changes is not available. This highlights the need for establishing a broader gene expression profile of osteoarthritic chondrocytes by modern screening technologies so as to characterize more properly the cellular events and regulatory pathways directly involved in cartilage destruction. In the present study, we designed a custom-made oligonucleotide-based microarray to screen for differentially expressed genes in normal and osteoarthritic cartilage specimens. We found that *Tob1*, a gene involved in cell cycle regulation and cell quiescence \[[@B4],[@B5]\], was significantly repressed in osteoarthritic chondrocytes. This was confirmed by quantitative polymerase chain reaction (qPCR) and further analyzed in adult articular chondrocytes *in vitro*and *in vivo*.
Materials and methods
=====================
Donors for mRNA expression analysis
-----------------------------------
For the study of mRNA expression levels within the tissue, cartilage from human femoral condyles of normal knee joints was used. Normal articular cartilage (*n*~qPCR~= 10, age range 45--88 years, mean age 64.1 years; *n*~array~= 9, age range 37--83 years, mean age 59 years) was obtained from donors at autopsy, within 48 hours of death. Osteoarthritic cartilage samples from late-stage osteoarthritic joint disease were obtained from patients undergoing total knee replacement surgery (*n*~qPCR~= 15, age range 63--85 years, mean age 74.5 years; *n*~array~= 10, age range 57 to 84 years, mean age 76 years). The cartilage was frozen in liquid nitrogen immediately after removal and stored at -80°C until required for RNA isolation.
Cartilage was considered to be normal according to a macroscopic scoring system of the opened joint: this mainly included normal synovial membrane, normal synovial fluid, no significant overall softening or surface fibrillation (except on the tibial plateau, which is basically found in all specimens depending on age). The Mankin\'s grade of histological plugs taken was less than 3. Osteoarthritic cases fulfilled the criteria published by the American College of Rheumatology \[[@B6]\]. Cases of rheumatoid origin were excluded from the study.
Isolation of primary human articular chondrocytes; stimulation with interleukin (IL)-1β and bone morphogenetic protein (BMP)-7
------------------------------------------------------------------------------------------------------------------------------
Normal human knee articular cartilage was obtained from six normal cases at autopsy within 48 hours of death. Cartilage pieces were finely chopped and chondrocytes were isolated enzymatically as described previously \[[@B7]\]. Chondrocytes were either plated in high-density monolayer cultures or cultured in alginate beads. Cultures were maintained for 48 hours in serum-free Dulbecco\'s modified Eagle\'s medium/F12 medium (Gibco BRL, Eggstein, Germany) supplemented with 1% penicillin/streptomycin solution (Gibco BRL) and 50 μg/ml ascorbate (Sigma, Taufkirchen, Germany) and 10% fetal calf serum (Biochrom, Berlin, Germany).
After 48 hours, primary (non-passaged) chondrocytes were stimulated with 1 ng/ml IL-1β (R&D System, Minneapolis, MN, USA) in DMEM/F12 medium, 100 ng/ml recombinant human BMP-7 (Stryker Biotech, Hopkinton, MA, USA) or cultivated in medium alone for 24 hours with no medium change afterwards. The same experiments were performed in parallel in the presence and in the absence of 10% fetal calf serum. At the end of the culturing/stimulation period the cells were washed in sterile phosphate-buffered saline (PBS), lysed in 350 μl of lysate RLT buffer/10^6^cells and stored at -80°C.
Culture of HCS-2/8 cells
------------------------
The human HCS-2/8 chondrosarcoma cell line (around passage 50--55) \[[@B8],[@B9]\] was cultured in DMEM (PAA, Linz, Austria) supplemented with 20% fetal bovine serum (Gibco BRL) and with 50 μg/ml ascorbate (Sigma) in a humidified atmosphere of 5% CO~2~at 37°C as described \[[@B9]\]. Cells were seeded at 10^5^cells/cm^2^and grown for 3 days to obtain subconfluent stage cultures, at 2 × 10^5^/cm^2^and cultured for 7 days to obtain confluent stage cultures, and at 6 × 10^5^/cm^2^and grown for 10 days for over-confluent stage cultures.
RNA isolation from articular cartilage and isolated articular chondrocytes
--------------------------------------------------------------------------
Total RNA from both cartilage tissue and isolated chondrocytes was isolated as described previously \[[@B10],[@B11]\]. The quality of total RNA samples was checked by agarose-gel electrophoresis and with the Bioanalyzer RNA 6000 Nano assay (Agilent, Waldbronn, Germany).
Construction of the SensiChip cartilage microarray
--------------------------------------------------
The SensiChip technology is a two-color microarray platform using the Planar Wave Guide technology for microarray detection \[[@B12]\], which increases signal-to-noise ratios and thereby the sensitivity of hybridization experiments. The arrays were spotted in duplicate with 70-mer oligonucleotides representing the 3\' untranslated region (UTR) of about 340 human cartilage-relevant genes, whereas one single gene was represented by one 70-mer oligonucleotide.
Expression profiling with the SensiChip two-color DNA-microarray platform
-------------------------------------------------------------------------
Total RNA (250 ng) from osteoarthritic cartilage (10 samples) and pooled normal cartilage was amplified and labeled with Cy3-UTP and Cy5-UTP respectively (Amersham Pharmacia) using the MessageAmp aRNA kit (Ambion). After clean-up of the complementary RNA (cRNA) with the RNeasy kit (Qiagen), 5 μg of Cy3-labeled cRNA from osteoarthritic cartilage was mixed with 5 μg of Cy5-labeled cRNA from pooled normal cartilage. cRNA was fragmented by incubation with 40 mM Tris-acetate, pH 8.1, 100 mM potassium acetate, 30 mM magnesium acetate for 15 min at 95°C and desalted with a Microcon YM-10 concentrator (Millipore). Mixed Cy-dye labeled cRNA samples (600 ng) were hybridized for 16 hours on a SensiChip microarray (Qiagen) spotted in duplicate with 70-mer oligonucleotides representing the 3\' UTR of selected genes. The gene-specific oligonucleotide sequences were designed by Operon by using GenBank accession numbers and proprietary algorithms. After washing steps performed in accordance with the manufacturer\'s standard protocol, arrays were scanned with the SensiChip Reader. The resulting array images were analyzed with SensiChip View 2.1 software (Qiagen) to quantify gene-specific signal intensities.
For quality control of RNA labeling and hybridization efficiency, oligonucleotides representing human housekeeping genes, negative and external bacterial spiking controls were also included. These sequences were prelabeled with fluorescent Cy3 and Cy5 dyes, and mixed in different concentrations into the hybridization solutions containing the labeled cRNA samples from human cartilage.
Expression data analysis
------------------------
All microarray scans were inspected visually and checked for quality on the basis of the performance of negative, housekeeping and externally added Cy3/Cy5-prelabeled spiking controls. Raw signal intensities from each scan were imported into the gene expression analysis software Resolver version 4.0 (Rosetta Biosoftware, Seattle, WA, USA). The software employs an error-modeling approach for the analysis of microarray data \[[@B13]\]. An error model specific for the SensiChip microarray platform was designed by Rosetta Biosoftware based on expression data from repeated hybridizations of the same RNA material to determine the variation of signal intensities. A complete description of the statistical methods used is available in the technology section of the Rosetta Biosoftware website <http://www.rosettabio.com/tech/default.htm>.
All scans were pre-processed and normalized with the SensiChip error model to calculate *P*values and error bars for every gene expression profile. The *P*value represented the probability that an observed gene regulation was due to a measurement error. Gene regulation was considered as statistically significant if the calculated *P*value was below a threshold of 0.05. For normalization of expression data, the average brightness of the Cy3 and Cy5 channels respectively was used that was calculated from spots within a range from 30% to 85% of the signal intensity distribution of all spots. Scans from multiple experiments (replicates) were combined by averaging expression data with an error-weighted algorithm (also described in the statistical methods document available on the Rosetta Biosoftware website).
Real-time quantitative PCR using TaqMan technology
--------------------------------------------------
Real-time PCR was used to detect human *Tob1*, *collagen type II*, *Ki-67*, *matrix metalloproteinase (MMP)-13*and *glyceraldehyde-3-phosphate dehydrogenase*mRNA expression levels in human articular cartilage RNA samples. The primers (MWG Biotech, Ebersberg, Germany) and TaqMan probes (Eurogentec, Liège, Belgium) were designed using Primer Express™ software (Perkin Elmer). To be able to obtain quantifiable results for all genes, specific standard curves using sequence-specific control probes were performed in parallel to the analyses. Thus, for each gene a gene-specific cDNA fragment was amplified by the gene-specific primers (Table [1](#T1){ref-type="table"}) and cloned into pGEM T Easy (Promega, Mannheim, Germany) or pCRII TOPO (Invitrogen, Karlsruhe, Germany). The cloned amplification product was sequenced to confirm correct cloning. Cloned standard probes were amplified with the plasmid amplification kit (Qiagen), linearized and used after careful estimation of the concentration (gel electrophoresis, photometry, and a fluorimetric assay for deoxyribonucleic acids (Picogreen; Molecular Probes, Eugene, OR, USA)). For the standard curves concentrations of 10, 100, 1000, 10,000, 100,000, and 1,000,000 molecules per assay were used (all in triplicate).
For the analyses of the different genes, a separate master mixture was made up for each of the primer pairs and contained a final concentration of 200 μM NTPs, 600 nM Roxbuffer and 100 nM TaqMan probe. For all genes the final reaction mixture contained, besides cDNA and 1 U polymerase (Eurogentec), forward and reverse primers, the corresponding probes, and MgCl~2~at concentrations given in Table [1](#T1){ref-type="table"}. All experiments were performed in triplicate.
Immunofluorescence
------------------
Immunofluorescence studies were performed on paraformaldehyde-fixed paraffin-embedded specimens of normal (*n*= 5) and osteoarthritic (*n*= 5) articular cartilage. Sections were first incubated with the primary antibodies overnight, then with biotin-labeled goat anti-mouse antibodies (Dianova, Hamburg, Germany) and then with peroxidase-labeled streptavidin (Dianova). Subsequently, the tyramide amplification system (PerkinElmer, Boston, MA, USA) was used for signal amplification. Finally, the signals were detected with Cy5-labeled streptavidin (Dianova). Nuclear staining was again performed with 4,6-diamidino-2-phenylindole. The sections were evaluated by a (fluorescence) microscope (Olympus AX70) and photographed digitally.
To obtain optimal staining results various enzymatic pretreatments were tested, including hyaluronidase (Boehringer, Mannheim, Germany; 2 mg/ml in PBS pH 5 for 60 min at 37°C), pronase (Sigma, Deisenhofen, Germany; 2 mg/ml in PBS pH 7.3 for 60 min at 37°C), and bacterial protease XXIV (Sigma; 0.02 mg/ml; PBS pH 7.3 for 60 min at 37°C). Finally, the mouse monoclonal antibodies against Tob1 (Assay Designs, Ann Arbor, MI, USA) were used at a dilution of 1:20 without pretreatment of the sections.
Amplification and cloning of *Tob1*cDNA
---------------------------------------
RNA was isolated from differentiated ADTC5 cells (Ricken Library) in accordance with the extraction method with Trizol^®^(Invitrogen) and reverse-transcribed into cDNA with SuperScript II™ reverse transcriptase (Invitrogen) by following the manufacturer\'s recommendation.
PCR amplification of a 607 base pair *Tob1*cDNA fragment (nucleotides 402--1008 of the sequence in GenBank accession no. NM\_009427) was performed with gene-specific primers (forward, 5\'-GGAGCCCCCAGGTGTTCATGC-3\'; reverse, 5\'-CTCGTTGAGGCCTCCGTAGG-3\') by a standard method, and amplification products were cloned into pCR^®^-BluntII-TOPO^®^vector (Invitrogen).
*In situ*hybridization
----------------------
*In situ*hybridization of sectioned appendicular skeleton from newborn mice was performed with digoxigenin-labeled antisense riboprobes transcribed from the *Tob1*cDNA fragment. Hindlegs of newborn mice were fixed overnight in 4% paraformaldehyde resolved in PBS. After stepwise transfer through solutions with increasing ethanol concentration, the specimens were incubated in xylene and finally embedded in paraffin wax.
For *in situ*hybridization, paraffin-embedded samples were cut into slices 7 μm thick and mounted on microscope slides. The sections were hybridized with digoxigenin-11-UTP-labeled antisense riboprobes, which were transcribed with T7 RNA polymerase from the *Tob1*cDNA fragment cloned into pCR^®^-BluntII-TOPO^®^(Invitrogen), after linearization of the plasmid with *Bam*HI.
*In situ*hybridization was performed as described by Dietz and colleagues \[[@B14]\]. After detection of hybridization products, the sections were mounted under coverslips in Kaiser\'s glycerol gelatin (Merck) and photographed under a Zeiss Axioplan 2 microscope.
Results
=======
Construction of the SensiChip cartilage microarray
--------------------------------------------------
A microarray covering 340 human cartilage relevant genes was constructed, where one single gene was represented by one 70-mer oligonucleotide (Fig. [1a](#F1){ref-type="fig"}). Most genes were selected from the literature and have important roles in anabolic or catabolic pathways during osteoarthritis (for example, cartilage matrix proteins such as collagens, relevant degrading enzymes such as MMPs and aggrecanases, and genes from important catabolic \[IL-1, tumor necrosis factor-α\] and anabolic \[BMP, transforming growth factor-β\] signaling pathways).
Gene expression analysis: differentially expressed genes
--------------------------------------------------------
Total RNAs from 10 late-stage osteoarthritic cartilage samples were hybridized separately against a pool of mixed total RNAs from nine normal cartilage donors on the customized SensiChip microarrays. Merging of expression profiles obtained from all 10 late-stage osteoarthritic cartilage samples used for hybridizations resulted in about 200 significantly regulated genes that were differentially expressed between normal and osteoarthritic cartilage, with *P*\< 0.01 (Fig. [2](#F2){ref-type="fig"} and Table [2](#T2){ref-type="table"}; the whole data set is in [Additional file 1](#S1){ref-type="supplementary-material"}).
*Tob1*is repressed in osteoarthritic chondrocytes
-------------------------------------------------
One of the differentially expressed genes was the human transducer of ERBB2,1 (*Tob1*; GenBank accession no. NM\_005749). *Tob1*was transcriptionally downregulated in all 10 human osteoarthritic cartilage samples to, on average, one-sixth (Fig. [1](#F1){ref-type="fig"}). Corresponding *P*values were less than 0.05 for all human OA samples.
Confirmation of *Tob1*expression and regulation by (quantitative) PCR and immunostaining in normal and osteoarthritic articular cartilage
-----------------------------------------------------------------------------------------------------------------------------------------
Conventional PCR confirmed the expression of *Tob1*, both in normal (*n*= 3) and osteoarthritic (*n*= 3) chondrocytes, with a weaker signal detected in the osteoarthritic samples (Fig. [3a](#F3){ref-type="fig"}). To validate and quantify differential regulation of *Tob1*, qPCR was performed on a set of normal (*n*= 10) and osteoarthritic (*n*= 15) samples. These experiments confirmed both its expression in normal articular cartilage and a highly significant decrease in *Tob1*transcript levels in osteoarthritic samples (7.8-fold; *P*\< 0.001; Fig. [3b](#F3){ref-type="fig"}).
Immunolocalization with monoclonal antibodies against Tob1 showed the presence of *Tob1*protein in normal (*n*= 5) and osteoarthritic (*n*= 8) articular chondrocytes (Fig. [3c,d](#F3){ref-type="fig"}). A somewhat weaker staining was observed in the osteoarthritic specimens than in the normal specimens, but this was not quantifiable because of the immunostaining technology used.
Correlation of *Tob1*expression to markers for chondrocyte anabolism, catabolism, and proliferation
---------------------------------------------------------------------------------------------------
Next we examined whether *Tob1*gene expression levels were correlated with the expression of marker genes of cell proliferation (*Ki-67*) and anabolic (*collagen type II*) and catabolic (*MMP-13*) activation of articular chondrocytes. This analysis showed highly significant correlations between these genes in osteoarthritic compared with normal chondrocytes (Fig. [4](#F4){ref-type="fig"}).
Expression of *Tob1*in articular chondrocytes *in vitro*
--------------------------------------------------------
*Tob1*was expressed in isolated human adult articular chondrocytes *in vitro*. The mRNA expression levels of *Tob1 in vitro*were comparable to those of osteoarthritic chondrocytes *in situ*and were therefore significantly lower than those of normal chondrocytes *in situ*(oligo-array, Fig. [1d](#F1){ref-type="fig"}; qPCR, Fig. [3b](#F3){ref-type="fig"}). It is noteworthy that *Tob1*was more strongly expressed in cells cultured without serum than with it. No significant regulation of *Tob1*was found by two major anabolic (BMP-7) and catabolic (IL-1β) mediators in adult articular cartilage in cultured chondrocytes *in vitro*(data not shown).
Expression of *Tob1*in the fetal growth plate and during chondrocyte differentiation *in vitro*
-----------------------------------------------------------------------------------------------
*In situ*hybridization on mouse fetal growth plate cartilage was performed to assess differential expression in the different cartilage zones. This showed that the expression of *Tob1*was concentrated in the hypertrophic zone (zone of terminal differentiation and cessation of proliferation). Cells of the resting and proliferating zones (that is, areas of proliferation and matrix synthesis) showed no or very much weaker staining (Fig. [3e](#F3){ref-type="fig"}). In addition, osteoblasts were positive (not shown).
Expression profiling in HCS-2/8 cells, which are known to show a more differentiated phenotype in high-density cultures than when cultured in subconfluent or confluent status \[[@B15]\] showed an inverse relationship between *Tob1*expression and the proliferation marker *Ki-67*(Fig. [5](#F5){ref-type="fig"}).
Discussion
==========
Differential gene expression analysis, as performed by us on normal and osteoarthritic chondrocytes, reveals long lists of differentially expressed genes of potential interest for furthering the understanding, diagnosis and/or modulation of osteoarthritis. The genes identified might be interesting with regard to any of these three aspects, but careful validation is needed to confirm the relevance of the findings obtained. In this regard, three levels of validation have to be achieved: (1) technical validation of screening results, (2) functional validation of the gene *in situ*or *in vitro*, and finally (3) establishment of relevance of the gene for the (physiology and/or) pathophysiology of the tissue.
In our oligonucleotide-based array screen we detected many known differentially expressed genes. Thus, many marker genes behaved as expected from previous investigations: stromelysin I (*MMP-3*) \[[@B7]\] and the cartilage transcription factor *SOX9*\[[@B16]\] were significantly downregulated, whereas many constituents of the extracellular matrix were significantly upregulated (*collagen types II*\[[@B17]\], *III*\[[@B18]\], *VI*\[[@B19]\], *COMP*\[[@B20]\], and *fibronectin*\[[@B21]\]). Further, MMP-13, the major collagenase of osteoarthritic cartilage \[[@B22],[@B23]\], was induced \[[@B7]\]). Taken together, these findings validated this gene array technology as a reliable tool for identifying differentially expressed genes. In addition, many genes previously unknown to be differentially regulated in osteoarthritic cartilage were detected.
Among the new differentially expressed genes we identified *Tob1*as being significantly downregulated in osteoarthritic compared with normal articular chondrocytes. For technical validation (validation level I), this was confirmed by conventional and quantitative PCR at a very high significance level. Immunostaining provided additional evidence of the presence of Tob1 in normal and osteoarthritic chondrocytes.
*Tob1*, originally identified as binding partner of Erb (\'transducer of Erb\' \[[@B24]\]), is a member of a larger family of proteins, which share common protein domains and are known to exert anti-proliferative and phenotype-stabilizing effects on various cell types including osteoblasts (\[[@B24],[@B25]\]; reviewed in \[[@B4]\] and \[[@B5]\]).
Thus, to obtain insights into the functional activity of *Tob1*in articular cartilage (validation level II), we correlated Tob1 expression with the expression of the *Ki-67*antigen, a well-established gene expressed only by cells in the proliferation phase \[[@B26]\]. We found a highly significant inverse correlation between *Tob1*expression and proliferative activity of chondrocytes. It is noteworthy that after isolation from the articular matrix *Tob1*was also repressed in normal articular chondrocytes *in vitro*. This might well reflect the fact that adult articular chondrocytes show an increased proliferative activity and also enhanced anabolic \[[@B27]\] and catabolic activity \[[@B7]\] after removal from the tissue. The fact that cells cultured with serum *in vitro*showed even lower *Tob1*expression levels than cultures without serum further supports this notion, because serum is known to increase proliferation of chondrocytes *in vitro*\[[@B28]\]. In addition, the chondrocytic cell line HCS-2/8 showed an inverse relationship between proliferative activity and cell differentiation on the one hand and Tob1 expression on the other. Interestingly, fetal chondrocytes *in situ*selectively express Tob1 in the hypertrophic zone, which is in contrast to other zones where no proliferative activity is seen \[[@B25]\]. This indicates that *Tob1*expression in chondrocytes is inversely related to proliferation in a similar way to that seen in T cells \[[@B29]\]. Another basic effect of *Tob1*is also observed in chondrocytes: a repression of *Tob1*is needed before activation of otherwise quiescent T cells \[[@B29],[@B30]\]; similarly, there is a clearcut inverse correlation between (anabolic and catabolic) chondrocyte activity and Tob1 expression.
In many respects the downregulation of *Tob1*fits well into the scenarios taking place during osteoarthritis (validation level III), which suggests that *Tob1*is a potential key molecule of cell phenotype regulation in osteoarthritic chondrocytes. Thus, in osteoarthritic cartilage an increase in proliferation \[[@B31]-[@B35]\] is found, whereas hardly any proliferative activity exists in normal articular adult cartilage \[[@B31],[@B32]\]. These cells seem to be G0-arrested, quiescent and phenotypically stable, in other words exactly the cell type that would be expected to express high levels of *Tob1*\[[@B4],[@B29]\]. It is noteworthy that both phenotypic instability \[[@B36]\] and anabolic activation \[[@B17]\] are key features of osteoarthritic chondrocytes, fitting well to the downregulation of *Tob1*.
*Tob1*seems in many circumstances and, in particular, in skeletal cells to interact with the BMP pathway \[[@B37]\]. Tob1-knockout mice develop osteopetrosis due to a lack of inhibition of BMP-stimulated bone growth \[[@B37]\]. In addition, overexpression of *Tob1*reduces BMP2 signaling \[[@B38]\]. Although in *Tob1*-knockout mice no specific \'hyperplastic\' cartilage phenotype was obvious, BMP-2 and BMP-7 are reported to have important functions in cartilage homeostasis \[[@B39],[@B40]\]. The presence of Tob1 could therefore explain why, despite the presence of BMPs within articular cartilage \[[@B39]\], normal chondrocytes show only very low anabolic activity. By the same argument, osteoarthritic chondrocytes BMPs might have much more anabolic potential, a feature recently suggested in studies *in vitro*\[[@B27]\].
In sum, our study provides for the first time compelling evidence of the expression and presence of Tob1 as a new intracellular mediator in adult articular chondrocytes and its downregulation in the osteoarthritic disease process. Tob1 fits well functionally with the cellular biological changes found in this condition such as proliferation, activation and the loss of a differentiated phenotype. Our data, together with the knowledge from other cellular systems in the literature, suggest that *Tob1*is a key molecule in the scenario of cellular alterations of osteoarthritis.
Conclusions
===========
Oligonucleotide-based microarray analysis was used to screen for differences in gene expression levels in between normal and osteoarthritic chondrocytes. Among other genes, *Tob1*was identified as being significantly downregulated in osteoarthritic chondrocytes. Correlative gene expression studies on cellular features such as cell proliferation, cell activation and the loss of a differentiated phenotype suggest that downregulation of *Tob1*expression might be an important aspect of cellular processes in osteoarthritic cartilage degeneration.
Abbreviations
=============
BMP = bone morphogenetic protein; cDNA = complementary DNA; cRNA = complementary RNA; IL = interleukin; MMP = matrix metalloproteinase; PBS = phosphate-buffered saline; PCR = polymerase chain reaction; qPCR = quantitative polymerase chain reaction; UTR = untranslated region.
Competing interests
===================
The author(s) declare that they have no competing interests. MG, JS, UD, and EB are all employed by Sanofi-Aventis as research scientists. The publication is a result of a scientific collaboration between industry and the other academic authors. The protein Tob1 is not pursued as a project within the osteoarthritis portfolio of Sanofi-Aventis; therefore the industry-affiliated authors have stated that they and the company have no competing interests.
Authors\' contributions
=======================
MG performed the gene expression analysis. JS cultured the HCS-2/8 cells and contributed to the bioinformatic analysis of obtained data sets. JH performed the collection and processing of human material (including RNA isolation). UD performed the *in situ*hybridization analysis. MT contributed the HCS-2/8 cell line. EB participated in the design of the study and coordinated the gene expression experiments including the bioinformatic analysis. TA wrote most of the manuscript and participated in the design of the study. His group contributed the TaqMan, conventional PCR and immunohistochemical analyses (together with JH). All authors contributed to writing and correcting the manuscript and have approved the final version.
Supplementary Material
======================
::: {.caption}
###### Additional File 1
An Excel file that contains details of 200 significantly regulated genes that were differentially expressed between normal and osteoarthritic cartilage with *P*values \<0.01.
:::
::: {.caption}
######
Click here for file
:::
Acknowledgements
================
We thank Klaus Lindauer PhD (Sanofi-Aventis Pharma, Frankfurt, FRG) for bioinformatic support, and Chris Barnes PhD (Sanofi-Aventis, Frankfurt) for a critical reading of the manuscript. We are grateful for excellent technical support by Mrs Anke Nehlen, Freya Boggasch, and Beatrice Schumann. We acknowledge the kind gift of recombinant BMP-7 by Stryker Biotech, Hopkinton, MA (DC Rueger). This work was supported by the German Ministry of Research (grant 01GG9824).
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
*Tob1*expression in normal and osteoarthritic cartilage (oligonucleotide array experiments). **(a)**Area of one customized SensiChip microarray illustrating the 70-mer oligonucleotide spots that represent the 3\'-untranslated region of human *Tob1*and demonstrate its differential expression between normal and one late-stage osteoarthritic cartilage. **(b)**Trend plot demonstrating the transcriptional downregulation of human *Tob1*in all RNA samples from cartilage of late-stage osteoarthritis patients used for SensiChip hybridization experiments. The logarithmic ratio of differential *Tob1*expression calculated by the software Resolver is plotted against the corresponding osteoarthritic patient sample used for expression profiling. Error bars indicate standard deviations of ratios. *P*values for the ratios of all 10 osteoarthritic samples were less than 0.05. **(c)**Transcriptional downregulation of human *Tob1*in late-stage osteoarthritic cartilage samples from 10 human donors. All *Tob1*ratios were calculated by the gene expression analysis software Resolver. *P*values for the ratios of all 10 osteoarthritic samples were less than 0.05. **(d)**Plot of average *Tob1*signal intensities from independent SensiChip microarray hybridizations using RNA samples from normal cartilage, late-stage osteoarthritic cartilage and cultured primary human chondrocytes. *Tob1*signal intensities from independent SensiChip hybridizations of RNA samples from pooled normal cartilage (10 hybridization experiments), 10 late-stage osteoarthritis patients (10 hybridization experiments) and 5 different cell culture samples of proliferating primary human chondrocytes (5 hybridization experiments) were merged respectively and are plotted as average *Tob1*signal intensities. Error bars indicate standard deviations of *Tob1*signal intensities.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Plot of normalized logarithmic expression signal intensities RNAs from late osteoarthritic against the intensities for normal cartilage. RNA samples from 10 late-stage osteoarthritis patients were hybridized in comparison with normal cartilage (pool of nine donors) on SensiChip microarrays. After normalization, expression data were merged and corresponding signal intensities of late-stage osteoarthritis patients and normal cartilage were plotted against each other. Several differentially expressed marker genes (*P*\< 0.01) are highlighted and diagonal lines indicate a twofold regulation. Error bars show standard deviations of ratios.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
*Tob1*expression in fetal, normal and osteoarthritic cartilage (PCR, immunostaining, *in situ*hybridization). **(a)**Conventional PCR demonstrates the expression of *Tob1*in normal and (at a reduced level) in osteoarthritic cartilage samples (lanes 1 and 9, molecular weight standards; lanes 2--4, normal cartilages; lanes 5--7, osteoarthritic cartilages; lane 8, negative control). In all experiments the RNA was directly from the tissue (without isolation of cells before isolation of RNA). **(b)**Quantitative real-time PCR analysis for mRNA expression levels of *Tob1*in normal (*n*= 10) and peripheral (pOA, *n*= 8) and central (cOA, *n*= 7) osteoarthritic cartilage as well as normal adult articular chondrocytes cultured with (*n*= 6) and without (*n*= 3) serum. Results are shown as ratios to glyceraldehyde-3-phosphate dehydrogenase. **(c, d)**Immunolocalization of Tob1 in human normal (c) and osteoarthritic (d) articular cartilage (in both the middle and upper deep zones of the cartilage are shown). **(e)**mRNA expression analysis of *Tob1*in fetal growth plate cartilage of mice, with the use of *in situ*hybridization: detectable expression levels are restricted to the hypertrophic zone (and osteoblasts).
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Comparative analysis of mRNA expression levels of collagen type II (*Col2*), *Ki-67*, and *MMP-13*relative to *Tob1*in normal and osteoarthritic chondrocytes.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Comparative mRNA analysis for *Tob1*and proliferation associated *Ki-67*mRNA expression in chondrocytic HCS-2/8 cells. HCS-2/8 cells were cultured in sub-confluent, confluent, and over-confluent conditions and the *Tob1*mRNA levels determined by qPCR (shown are the ratios to glyceraldehyde-3-phosphate dehydrogenase (GADPH)).
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Sequences of primers and probes for quantitative real-time polymerase chain reaction
:::
Gene GenBank accession no. Primers (5\'→3\') Conc. (nM) Probe (5\'→3\') MgCl~2~(mM)
---------- ----------------------- ----------------------------------- ------------ --------------------------------- -------------
*GAPDH* NM\_002046 Forward: GAAGGTGAAGGTCGGAGTC 50 CAAGCTTCCCGTTCTCAGCC 5.5
Reverse: GAAGATGGTGATGGGATTTC 900
*TOB1* NM\_005749 Forward: TCTGCTGCTGTAAGCCCTACCT 300 CGGTCCACTCAGCCTTTAACCTTTACCACT 6.5
Reverse: TTCATTTTGGTAGAGCCGAACTT 900
*Ki-67* NM\_002417 Forward: CAGTGATCAACGCCGTAGGTC 900 CTTCCAGCAGCAAATCTCAGACAGAGGTTC 6.0
Reverse: TCGGCTGATAGACACTCTCTTTTG 900
*COL2A1* NM\_001844 Forward:CAACACTGCCAACGTCCAGAT 50 ACCTTCCTACGCCTGCTGTCCACG 5.5
Reverse: CTGCTTCGTCCAGATAGGCAAT 300
*MMP-13* NM\_002427 Forward: TCCTCTTCTTGAGCTGGACTCATT 900 TCCTCAGACAAATCATCTTCATCACCACCAC 7.0
Reverse: CGCTCTGCAAACTGGAGGTC 50
*GAPDH*, glyceraldehyde-3-phosphate dehydrogenase; *MMP*, matrix metalloproteinase; *COL2A1*, collagen type II (alpha 1 chain); *TOB1*, Transducer of ERBB2.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Table showing genes which were upregulated or downregulated in osteoarthritic chondrocytes (changes in mRNA expression levels \>2-fold; *P*\< 0.01)
:::
Downregulated genes GenBank accession no. OA:N OA N Upregulated genes GenBank accession no. OA:N OA N
------------------------- ----------------------- ------ ------- ------- ---------------------- ----------------------- ------- ------- -------
*CHI3L1* NM\_001276 0.02 0.001 0.05 *COMP* NM\_000095 2.07 19.93 10.68
*Follistatin* NM\_006350 0.06 0.01 0.22 *BGLAP* X53698 2.18 0.13 0.07
*APOD* NM\_001647 0.08 1.10 15.68 *TIMP1* X03124 2.20 12.18 6.34
*EDR2* NM\_004427 0.09 0.09 0.86 *ARGBP2* AF049884 2.25 0.11 0.04
*MMP3* X05232 0.09 1.04 10.28 *FGF18* AF075292 2.35 0.40 0.16
*Tob1* NM\_005749 0.14 0.12 0.96 *Fibronectin* M10905 2.36 19.08 9.46
*SLC3A2* NM\_002394 0.19 0.04 0.22 *Frizzled homolog 1* NM\_003505 2.37 0.05 0.03
*MAPK7* NM\_002749 0.22 0.01 0.05 *Choriolysin h 2* AA829685 2.38 0.06 0.03
*SOX9* Z46629 0.26 0.29 1.14 *CRTL1* NM\_001884 2.43 0.76 0.31
*SERPING1* NM\_000062 0.27 0.76 3.27 *MKP-L* NM\_007026 2.57 0.32 0.11
*TGFBR3* NM\_003243 0.27 0.05 0.15 *THBS3* NM\_007112 2.59 0.11 0.04
*FRZB* NM\_001463 0.27 0.23 1.43 *ADAMTS2* NM\_014244 2.68 0.06 0.02
*MRS3/4* AF327402 0.28 0.12 0.43 *ADAMTS1* NM\_006988 2.70 0.12 0.05
*UBC* M26880 0.31 3.29 11.04 *CHM-I* NM\_007015 2.88 0.36 0.09
*Integrin α5* NM\_002205 0.35 0.36 1.17 *MMP-13* X75308 2.88 0.13 0.06
*GNB2L1* NM\_006098 0.38 1.15 2.39 *COL6A1* X15880 2.89 9.55 2.94
*NF-κB\_p65* Q04206 0.40 0.04 0.10 *COL11A1* J04177 3.84 0.52 0.16
*Pim-1* Z58595 0.41 0.05 0.15 *TNFAIP6* NM\_007115 4.73 0.33 0.07
*NCK1* NM\_006153 0.41 0.01 0.04 *Thrombospondin 2* NM\_003247 5.01 0.13 0.03
*GSS* NM\_000178 0.43 0.03 0.06 *COL2A1* X16468 5.67 11.62 2.20
*HLA-B* M81798 0.43 1.43 3.50 *SPP1* NM\_000582 5.78 1.96 0.38
*ICAM1* NM\_000201 0.44 0.09 0.26 *CKTSF1B1* NM\_013372 5.89 0.06 0.01
*TG-interacting factor* NM\_003244 0.47 0.05 0.10 *COL6A3* NM\_004369 8.70 3.49 0.45
*Phosphomannomutase 1* U86070 0.47 0.11 0.26 *COL1A2* X55525 14.25 4.21 0.33
*DLX5* NM\_005221 0.47 0.04 0.08 *TGFBI* NM\_000358 14.67 2.41 0.18
*Biglycan* NM\_001711 0.49 0.21 0.50 *COL3A1* X14420 31.33 15.33 0.52
TP53 NM\_000546 0.50 0.14 0.34
N, mean of mRNA expression levels in the normal cartilage samples (in arbitrary units); OA, mean of mRNA expression levels in the osteoarthritic cartilage samples (in arbitrary units); OA:N, ratio of osteoarthritic to normal.
:::
|
PubMed Central
|
2024-06-05T03:55:54.933214
|
2005-1-11
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065318/",
"journal": "Arthritis Res Ther. 2005 Jan 11; 7(2):R274-R284",
"authors": [
{
"first": "Mathias",
"last": "Gebauer"
},
{
"first": "Joachim",
"last": "Saas"
},
{
"first": "Jochen",
"last": "Haag"
},
{
"first": "Uwe",
"last": "Dietz"
},
{
"first": "Masaharu",
"last": "Takigawa"
},
{
"first": "Eckart",
"last": "Bartnik"
},
{
"first": "Thomas",
"last": "Aigner"
}
]
}
|
PMC1065319
|
Introduction
============
Juvenile rheumatoid arthritis (JRA) is the most common chronic arthritic condition of childhood, encompassing pauciarticular, polyarticular, and systemic-onset disease subtypes. JRA is typically considered autoimmune in etiology, with characteristic T-cell abnormalities and chronic synovitis. The extent of synovitis may range from minimal to severe, and vary in terms of number of joints involved, with systemic onset disease typically associated with the greatest morbidity. JRA is probably a collection of diseases with complex overlapping etiologies, with each subtype influenced by multiple genetic susceptibility loci and mediated by environmental effects \[[@B1]\]. The MHC on the short arm of chromosome 6 has been intensively analyzed, and associations with both HLA and non-HLA genes have been reported. Genetic associations with MHC alleles have been documented primarily within the HLA class II region, but also with certain class I alleles. These associations are largely JRA subtype and age specific \[[@B2]\], and are strongest for pauciarticular and polyarticular disease \[[@B1]\]. For systemic onset JRA (SoJRA), associations with *HLA-B8*, *HLA-Bw35*\[[@B3],[@B4]\] and *HLA-DR4*\[[@B3],[@B5]\] have been observed, whereas *HLA-DPB1\*0401*was reported to have a protective effect in one Caucasian population \[[@B6]\]. Associations with *HLA-DRB1\*0401*and *HLA-DRB1\*0405*have been reported in a Japanese population \[[@B7]\]. Most of these associations have not been replicated.
In the present study we targeted the tapasin gene (*TPSN*), which is in the class II region of the MHC, 180 kilobases centromeric of HLA-DP. The tapasin protein is necessary for the proper assembly and peptide-presenting function of HLA class I molecules \[[@B8]\]. The *TPSN*gene has a polymorphism in exon 4 that results in a nonconservative amino acid substitution of Arg/Thr at amino acid 260 (ref SNP ID: rs2071888) \[[@B9],[@B10]\]. Three intronic polymorphisms of *TPSN*have also been described, none of which appear to be in linkage disequilibrium (LD) with HLA class I alleles or the extended *HLA-A1*, *HLA-B8*, *HLA-DR3*haplotype, in at least one healthy Caucasian population \[[@B11]\]. Furthermore, using a large UK Caucasian sample, Ahmad and coworkers \[[@B12]\] recently reported that *TPSN*polymorphisms are not in LD with more telomeric MHC haplotypes. In the present study we report an association between the exon 4 *TPSN*polymorphism and susceptibility to SoJRA, involving the *TPSN*allele Arg260 (01 allele).
Methods
=======
The study cohort included 88 SoJRA affected families recruited in Cincinnati (US cohort) and 74 simplex (with one affected offspring) SoJRA families identified by the British Paediatric Rheumatology Study Group (UK cohort). Unaffected siblings were available for analysis in the US but not in the UK cohort. An additional 113 healthy unrelated control individuals, primarily Caucasians from the Midwest and New England, resembled the SoJRA-affected families in terms ethnicity and served as the control population. Ethics approvals were obtained from the participating institutions, and informed consent was obtained from parents and/or children.
All affected children met American College of Rheumatology criteria for a diagnosis of JRA; they were subgrouped as pauciarticular, polyarticular, or SoJRA. Genomic DNA was purified from peripheral blood cells by standard techniques and analyzed for *TPSN*alleles (Arg260/01 and Thr260/02) by polymerase chain reaction and restriction site enzyme digestion. Briefly, a 298-base-pair fragment of exon 4 of the *TPSN*gene containing the polymorphism was amplified and then digested with *Bfa*I, which recognizes the 01 allele, and *Sfc*I, which recognizes the 02 allele. The primers used were Tsn 479 forward (5\'-CCC ACC CTC TAC CCC TGG A-3\') and Tsn 641 reverse (5\'-CAG CAC CTG GGT AAG GGA CA-3\'). HLA types were determined for a subgroup of the participants using DNA-based low-resolution methodology (Geno-Vision Inc., Exton, PA, USA), and serologically using standard typing sera and microcytotoxicity assays.
Preliminary association analysis was conducted by χ^2^testing on contingency tables comparing the three genotypic frequencies between cases and control individuals to yield a χ^2^with two degrees of freedom. Family-based association analysis was performed using the transmission disequilibrium test (TDT). The TDT \[[@B13],[@B14]\] is a family-based association test that compares within a cohort the number of times a particular parental allele is transmitted to an affected offspring versus the number of times it is not transmitted. To allow inclusion of families with missing data for a single parent, the Transmit program <http://www-gene.cimr.cam.ac.uk/clayton/software/>\[[@B15]\], which uses population allele frequencies to weight the possible parental genotypes, was used for the TDT analysis. In the Transmit program, genotypes of unaffected siblings (or siblings whose disease status is unknown) are used to infer parental genotypes, thus increasing the power to detect association. We applied this test first to the combined US and the UK data and then to each set separately. The significance level of the combined results was also calculated using Fisher\'s method of combining *P*values for two independent analyses that test the same hypothesis. LD between *TPSN*and the HLA region (limited to the Cincinnati cohort) was evaluated using the EH program <http://linkage.rockefeller.edu/ott/eh.htm>\[[@B16],[@B17]\]. The Geno-Pdt test in the PdT 5.1 program <http://wwwchg.duhs.duke.edu/software/pdt.html> was also used as a test for association and linkage in the US SoJRA-affected families \[[@B18]-[@B20]\].
Results
=======
The distribution of tapasin genotypes among SoJRA-affected children was compared with that in their healthy siblings, as well as with that in unrelated healthy control individuals using a two degrees of freedom χ^2^test in the US cohort (Table [1](#T1){ref-type="table"}). The differences did not reach statistical significance. The allelic frequencies of tapasin in the independent cohorts from the USA (Table [1](#T1){ref-type="table"}) were in Hardy--Weinberg equilibrium (healthy individuals from the USA: χ^2^= 0.3049, *P*= 0.58; SoJRA-affected individuals from the USA: χ^2^= 2.004, *P*= 0.156). In the UK data only SoJRA-affected individuals were available for Hardy--Weinberg equilibrium testing, and the result was borderline (χ^2^= 5.26, *P*= 0.02).
We then tested for evidence of linkage of *TPSN*to JRA by applying the family-based TDT only to the cohorts of affected individuals for whom parental and sibling information was available. These families included 82 US SoJRA familes (389 individuals, including family members) and 74 UK SoJRA families. The TDT test, as implemented in the Transmit program, detected the preferential transmission of the *TPSN*01 allele in the UK and US SoJRA families (*n*= 156; χ^2^= 4.2, 1 degree of freedom \[df\]; *P*= 0.04; Table [2](#T2){ref-type="table"}). When the US SoJRA cohort was analyzed alone, this preferential transmission for the *TPSN*allele 01 was even more significant (χ^2^= 6.0, 1 df; *P*= 0.01; Table [2](#T2){ref-type="table"}).
However, SoJRA families from the UK alone as a subgroup failed to show significant preference for *TPSN*allele 01 transmission (χ^2^= 0.075, 1 df; *P*= 0.78; Table [2](#T2){ref-type="table"}). When the SoJRA UK cohort was analyzed, it was recognized that information for one of the parents was missing in 22 families (29.7% of the total). There was no information from unaffected siblings of the probands for these UK families, which is necessary for the Transmit program to narrow down the range of the possible parental genotypes. When one parent is missing, the Transmit program assigns and weights possible haplotypes to the missing parent using the information from the known parent and the siblings of the proband, and averages the probability of all transmissions to the proband. Because there were no data regarding the genotypes of unaffected siblings for the UK families, the Transmit program was unable to infer parental genotypes and thus had less power to detect the preferential transmission of the 260Arg allele. In the US SoJRA-affected cohort parental information was missing for 32 of the 82 families (39%), but for 12 of these (14.6%) there was information regarding unaffected siblings. Although 24.4% of the cohort was unavailable for TDT calculation, 62 families (75.6%) were available.
We combined the two *P*values using Fisher\'s method \[[@B21]\] to obtain a χ^2^with four degrees of freedom and found the combination of the two analyses to be significant (χ^2^= 9.7, 4 df; *P*= 0.05). Overall, when the two cohorts were analyzed together the tapasin 260Arg allele was transmitted more often than the 260Thr allele, suggesting association and linkage between the *TPSN*polymorphism and SoJRA. In order to include information from the unaffected siblings for association testing, we also applied an alternative association test to the US SoJRA population -- the pedigree disequilibrium test (PDT) \[[@B18],[@B19]\]. A new version of PDT, the genotype-based association test for pedigrees (Genotype-PDT), was applied to the data. Genotype-PDT \[[@B20]\] tests for linkage and underlying patterns of association at the genotypic level. It is more conservative and has lesser type 1 error when compared with the TDT test implemented in the Transmit program. Genotype-PDT also uses information from affected individuals, unaffected siblings, and their nuclear families. Therefore, we were only able to apply this test to the US SoJRA cohort. The genotype-PDT test revealed association and linkage of the tapasin 260Arg allele with SoJRA at the genotypic level (χ^2^= 6.727, 1 df; *P*= 0.034) in the US SoJRA cohort.
Furthermore, we wished to control for possible transmission distortion of tapasin 260Arg allele in SoJRA-affected families. This allele could also be preferentially transmitted to the unaffected siblings of the SoJRA-affected individuals from their parents, and our statistical significance could be falsely inflated because of allele-specific segregation bias (altered transmission of an allele independent of its role in disease). We therefore applied the TDT test to the unaffected siblings from the US SoJRA cohort. In contrast to the affected siblings, there was no significant preference toward 260Arg allele transmission to healthy siblings, suggesting no segregation bias (χ^2^= 1.043, 1 df; *P*= 0.3; Table [2](#T2){ref-type="table"}). These data provide evidence of a genotypic association and linkage between the *TPSN*260Arg allele and susceptibility to SoJRA.
Discussion
==========
Although JRA is the most common rheumatologic disease in childhood, the SoJRA subtype comprises less then 20% of cases and is a rare disease. In the past, because of the small sample sizes, studies conducted by single centers failed to establish strong genetic associations. The present study was therefore done with collaboration between two different centers. These two centers recruited mainly Caucasian families with one SoJRA-affected offspring. An association of the *TPSN*allele 260Arg with SoJRA was detected when both cohorts were analyzed and in the US cohort by itself. In the UK cohort the statistical analysis did not reveal a significant association. This discrepancy may be due to the limited sample size for the UK data, the different ethnic backgrounds of the two cohorts, and/or gene--environment interactions.
In general, it is suggested that studies using independent controls are more powerful than those using related (family-based) controls, but they may be biased if cases and controls have different ethnic backgrounds because of population stratification. Family-based control studies are less powerful because of overmatching, but they are robust to population stratification. In the present study we used family-based control association tests, which allowed us to analyze SoJRA family cohorts recruited by two different centers (US and UK).
Statistical programs designed to test genetic linkage based on TDTs (i.e. linkage in the presence of association) calculate the transmission of alleles from heterozygous parents to affected individuals. In the absence of one parent, the family becomes uninformative regarding single nucleotide polymorphisms and cannot be included in the analysis. This decreases the sample size, thus reducing the power to detect association or genetic linkage in rare diseases. Recently, programs such as Transmit and PDT have become available that are designed to calculate the possible genotypes of the missing parent from unaffected children or other family members such as grandparents. However, in cohorts consisting of simplex families (mother, father and the affected child), which do not have unaffected siblings or grandparents, and when there are families with missing parents, these programs are unable to achieve as much power to detect genetic association. The TDT test in the Transmit program was unable to infer the missing parental information (22 families, 29% of the data) from the UK cohort, which decreased the sample size to 52 families (71% of the cohort). In contrast, in the US SoJRA cohort the presence of unaffected siblings made 62 families (75.6% of the cohort) available for testing and increased the power to detect linkage in the presence of association.
In order to detect whether the group of children with earlier age at onset of SoJRA (\<6 years at onset) is in association with *TPSN*allele 260Arg, both cohorts were dichotomized by age at disease onset and analyzed using the Transmit program. *TPSN*allele 260Arg was still preferentially transmitted in both of the age onset groups but there was no statistical significance at the 5% level. When the age at onset groups \<6 years and ≥6 years were pooled together from US and UK cohorts and analyzed, there was still no statistically significant association with *TPSN*allele 260Arg.
The other possible reason for the lack of significant linkage in the UK cohort when analyzed alone might be ethnic difference, with a different polymorphism associated with SoJRA and a different disease frequency. Although both cohorts consisted of Caucasians, there might still have been ethnic differences between them. Therefore, it could be that the association of *TPSN*allele 260Arg with SoJRA in the US population is due to LD, with different SoJRA susceptibility alleles located on chromosome 6 being due to differing recombination processes between the US and UK Caucasian populations.
Because weak associations between SoJRA and *HLA-DR*alleles \[[@B5]-[@B7]\] have previously been noted, we compared the available class II allele frequencies, including *HLA-DR*, *HLA-DP*and *HLA-DQ*, in SoJRA patients from the US cohort (*n*= 69) with healthy control individuals (*n*= 66). No statistically significant differences were found (data not shown). However, it is worth noting that we detected a small trend toward a lower *HLA-DPB\*0401*frequency in SoJRA patients (28%) as compared with healthy control individuals (35%), which is consistent with a previous report \[[@B6]\] that suggested a possible protective role for this particular HLA allele in SoJRA. Furthermore, we applied a test for LD (using the EH program) to assess LD between *TPSN*alleles and the HLA alleles in SoJRA patients. We included 34 SoJRA-affected children from the US cohort and 38 healthy individuals for whom HLA typing data were available. There was no evidence for LD between the *TPSN*and any of the HLA alleles with the SoJRA-affected individuals or the healthy individuals. Because HLA typing was not available for all patients in the US and UK SoJRA patient groups, these calculations were done using very small sample sizes, and so the possibility of LD between the *TPSN*and HLA loci in these groups cannot be completely eliminated.
The landmark cytokines that contribute directly to the clinical features or autoimmune process of SoJRA, namely IL-6, tumor necrosis factor (TNF)-α, and IL-1, are also known to be important regulators of apoptosis. SoJRA\'s characteristic clinical and laboratory features, such as fever, skin rash, hypergammaglobulinemia, hypoalbuminemia, elevated erythrocyte sedimentation rate, and fibrinogen levels, may all be explained by cytokine-activated inflammatory and/or immune responses. Elevated blood level of IL-6 in SoJRA is known to correlate with fever episodes \[[@B22],[@B23]\]. Some of these cytokines were evaluated for their associations with SoJRA. The non-MHC cytokine gene polymorphisms that have been associated with SoJRA are the IL-6 5\' flanking polymorphism \[[@B24]\], the TNF-α 5\' flanking polymorphism \[[@B25]\], and the macrophage migration inhibition factor polymorphism \[[@B26]\]. Recent cytotoxicity studies also implicate natural killer cell dysfunction in this process \[[@B27]\].
Functional differences between the tapasin proteins encoded by the two alleles (Arg versus Thr at 260) have not, to our knowledge, been described. Given what is known about the function of tapasin, it is conceivable that polymorphisms might affect the quality or quantity of peptides presented by class I molecules, thereby influencing the immune response. It is also worth noting that the *TPSN*gene is separated from *Daxx*, an effector of Fas ligand and transforming growth factor-β mediated apoptosis \[[@B28],[@B29]\] by only a single gene (*BING2*). Apoptosis plays a key role in regulating the immune response in part by balancing excess cellular proliferation, and several of the key cytokines that have been implicated in the pathogenesis of SoJRA, such as TNF-β, IL-6, and IL-1, are known to influence apoptotic pathways. Thus, it is perhaps more tempting to speculate that the *TPSN*01 polymorphism (TPSN 260Arg) associated with SoJRA might be in LD with another susceptibility allele in a gene such as *Daxx*(or other genes in the region that play roles in apoptosis). Furthermore, the *TPSN*260Arg allele might be part of a haplotype in the HLA region that contributes to susceptibility to SoJRA. It will be important to examine additional SoJRA populations to determine whether TPSN is associated with disease. If so then further genetic studies of this region, including LD testing and exploration of candidate gene alleles in the region, may be of considerable interest.
Conclusion
==========
In conclusion, our studies support the existence of a weak association, possibly due to a linked gene in the region, between the *TPSN*01 allele and susceptibility to SoJRA.
Abbreviations
=============
df = degrees of freedom; IL = interleukin; JRA = juvenile rheumatoid arthritis; LD = linkage disequilibrium; MHC = major histocompatibility complex; PDT = pedigree disequilibrium test; SoJRA = systemic onset juvenile rheumatoid arthritis; TDT = transmission disequilibrium test; TNF = tumor necrosis factor.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
HB carried out the molecular genetic study in US, genotyped the US cohort, did the statistical analysis of both US and UK cohorts, and drafted the manuscript. MF carried out the molecular study in UK and participated in drafting the manuscript. MT confirmed the genotypes of US SoJRA patients. SDT participated in the coordination of the study and drafting the manuscript. NAT genotyped the UK SoJRA cohort. PW coordinated the UK study and participated in drafting the manuscript for the UK cohort.
Acknowledgements
================
We acknowledge the British Paediatric Rheumatology Study Group for their contributions to this study. Hulya Bukulmez, MD, is supported in part by NRSA/NHLBI, T32HL07567 and NRSA, NIAMS, T32AR07594. Jane Olson is supported in part by USPHS grants HG01577 from the NCHGR and RR03655 from the NCRR. Robert Elston is supported in part by grants GM28356 from NIGMS and RR 03655 from NCRR. Monica Tsoras, Susan Thompson, David Glass and Robert Colbert are supported by NIH/NIAMS N01AR42218, P60AR47784, P30AR47363, and AR41677. The work conducted in the UK was funded by the Arthritis Research Campaign (AR47363-02).
Figures and Tables
==================
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Tapasin genotypic distributions in systemic onset juvenile rheumatoid arthritis families
:::
Tapasin genotypes US healthy control individuals US healthy siblings of SoJRA US SoJRA UK SoJRA UK and US SoJRA
--------------------- -------------------------------- ------------------------------ ------------ ------------ -----------------
0101 35 (31.0%) 26 (28.0%) 49 (36.6%) 23 (30.0%) 72 (34.4%)
0102 56 (49.5%) 50 (53.7%) 63 (47.0%) 45 (60.0%) 108 (51.6%)
0202 22 (19.5%) 17 (18.3%) 22 (16.4%) 7 (9.2%) 29 (14.0%)
01 allele frequency 0.56 0.55 0.6 0.6 0.6
02 allele frequency 0.44 0.45 0.4 0.4 0.4
Total 113 93 134 76 209
Tapasin 01 allele is Arg260, and 02 is Thr260. SoJRA, systemic onset juvenile rheumatoid arthritis.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Transmission disequilibrium test analysis of the US and UK systemic onset juvenile rheumatoid arthritis families using Transmit
:::
Arg260 transmission Thr260 transmission
------------------------ ----- --------------------- --------------------- ----- --------- ------- --------
US only 82 99 (88.0) 65 (75.9) 6.0\* 0.01\*
US unaffected siblings 79 82 (77.5) 76 (80.5) 1.04 0.3
UK only 74 91 (89.0) 57 (58.0) 0.08 0.78
UK and US together 156 190 (177.8) 122 (134.2) 4.2\* 0.04\*
\*Significant findings. *n*, number of informative families for transmission; UK, British Paediatric Rheumatology Study Group cohort; US, Cincinnati cohort.
:::
|
PubMed Central
|
2024-06-05T03:55:54.936920
|
2005-1-11
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065319/",
"journal": "Arthritis Res Ther. 2005 Jan 11; 7(2):R285-R290",
"authors": [
{
"first": "Hulya",
"last": "Bukulmez"
},
{
"first": "Mark",
"last": "Fife"
},
{
"first": "Monica",
"last": "Tsoras"
},
{
"first": "Susan D",
"last": "Thompson"
},
{
"first": "Natalie A",
"last": "Twine"
},
{
"first": "Patricia",
"last": "Woo"
},
{
"first": "Jane M",
"last": "Olson"
},
{
"first": "Robert C",
"last": "Elston"
},
{
"first": "David N",
"last": "Glass"
},
{
"first": "Robert A",
"last": "Colbert"
}
]
}
|
PMC1065320
|
Introduction
============
The trafficking and accumulation of immunocompetent cells are essential components in the pathophysiology of the inflammatory processes. A number of recent data suggest that most of these events are regulated by chemokines, a superfamily of 8--10 kDa molecules that has been divided into four branches (C, CC, CXC, and CXXXC) according to variations in a shared cysteine \[[@B1],[@B2]\]. The current roster approaches more than 50 related proteins. Structural variations of chemokines have been associated with differences in their ability to regulate the trafficking of immune cells during inflammatory disorders. The biological activity of chemokines is mediated by seven-transmembrane-domain, G-protein-coupled receptors classified as C, CC, CXC, or CXXXC chemokine receptors according to the type of chemokine bound. Chemokine receptors are constitutively expressed on some cells, whereas they are inducible on others \[[@B3]\].
Three CXC chemokines (IP-10/CXCL10, Mig/CXCL9, and I-TAC/CXCL11) that are produced in response to IFNγ allow for the accumulation of activated lymphocytes by interacting with a specific receptor (CXCR3) \[[@B2]\]. Although the interactions of chemokine receptors are often characterized by considerable promiscuity, CXCR3 is selective in the recruitment of Th1 cells, B cells, and NK (natural killer) cells but not of nonlymphoid cells. Juvenile idiopathic arthritis (JIA) is characterized by chronic inflammation of the synovium in multiple joints. Early studies of the synovial membrane in JIA have shown the presence of a dense infiltrate of activated T cells clustered around activated dendritic cells, suggesting that lymphocyte recruitment is crucial in the pathogenesis of the disease \[[@B4],[@B5]\]. There is also strong evidence of an up-regulation of IFNγ expression in synovial tissue relative to that in peripheral blood of patients with JIA \[[@B6],[@B7]\], indicating a Th1 type polarization of local inflammatory response. Taken together, these data suggest that lymphocyte-specific CXC chemokines could be involved in the mechanisms promoting the development of inflammatory events in JIA patients.
In this study, using immunohistochemical and molecular studies of tissue sections and flow cytometry evaluation of cells recovered from synovial fluid, we evaluated the role of CXCR3/CXCL10 interactions in the regulation of T-cell migration into the joints of patients with JIA. We have demonstrated the presence of IP-10/CXCL10 in the synovial tissue and its release into the synovial fluid, where it exerts chemotactic activity toward activated CXCR3^+^T cells. Taken together, our data suggest that the local production of CXCL10 is involved in the pathophysiology of JIA-associated inflammatory processes.
Materials and methods
=====================
Study populations
-----------------
We analyzed synovial tissue from nine patients with oligoarticular JIA who were undergoing arthroscopic synovectomy. All the patients fulfilled the revised criteria for JIA according to the International League of Associations for Rheumatology (ILAR) classification \[[@B8]\] and were managed at the Pediatric Rheumatology Unit of Padua University. The procedure was performed in the case of persistently inflamed joints that did not respond either to systemic anti-inflammatory therapy or to intra-articular steroid injections. In all these patients, gadolinium-enhanced MRI showed marked thickening of the synovial membrane throughout the joint. The patients\' mean age at onset of the disease was 70.6 months (range 34--156); the average disease duration at synovectomy was 29.5 months (range 2--60).
As controls, three synovial tissue specimens obtained from children with noninflammatory arthropathy were analyzed by immunochemistry. These subjects had presented with either hexadactylism, bone dysplasia, or bone fracture.
Paired samples of peripheral blood (PB) and synovial fluid (SF) from 20 consecutive patients undergoing intra-articular steroid injection were examined. These patients\' mean age at onset of the disease was 77 months (range 13--264) and the mean disease duration was 17 months (range 2--108). Patients who were having systemic anti-inflammatory treatment at the time were excluded from the study.
Since the local ethics committee was not established yet at the beginning of the study, institutional review board approval was not requested, but informed consent was obtained from the parents of all the children included in this study.
Phenotypic evaluation of lymphocytes from peripheral blood and synovial fluid
-----------------------------------------------------------------------------
The commercially available conjugated or unconjugated monoclonal antibodies used were from the Becton Dickinson (Sunnyvale, CA, USA) and PharMingen (San Diego, CA, USA) series and included CD3, CD4, CD8, CD45R0, CD45RA, and isotype-matched controls. Fluorescein-isothiocyanate-labelled mouse antihuman CXCR3 (R&D Systems Inc, Minneapolis, MN, USA) was also used, and the frequency of PB and SF cells positive for this reagent was determined by flow cytometry as previously reported \[[@B9]\]. Specifically, cells were scored using a FACSCalibur analyzer (Becton Dickinson) and data were processed using the Macintosh CELLQuest software program (Becton Dickinson).
For immunofluorescence analysis, control mouse IgG~1~and IgG~2a~were obtained from Becton Dickinson.
Chemotactic activity of synovial fluid
--------------------------------------
The CXCR3-positive cell line 300-19 (kindly provided by Dr B Moser, Theodor-Kocher Institute, University of Bern, Switzerland) was used to evaluate the chemotactic activity of SF. The cells were grown in RPMI 1640 medium supplemented with 1% glutamine, 5% human serum, 1% kanamycin, and 100 U/ml human recombinant IL-2. Cells were periodically expanded by restimulation with phytohemagglutinin (1 μg/ml) in the presence of irradiated blood mononuclear cells (10:1 ratio of feeder cells : 300-19 cells) and were used for experiments after a culture period of 10 to 14 days.
Cell migration was measured in a 48-well modified Boyden chamber (AC48, Neuro Probe Inc, Gaithersburg, MD, USA). The chamber contains two sections. Chemotactic stimuli were loaded in the bottom section, and cells were put into the top compartment. Polyvinylpyrrolidone-free polycarbonate membranes with 3- to 5-μm pores and coated with fibronectin were placed between the two chamber parts. Only the bottom face of filters was pretreated with fibronectin; this treatment maximizes attachment of migrating cells to filters, increasing their adherence. SF samples or control medium (30 μl) was added to the bottom wells, and 50 μl of 300-19 cells resuspended in RPMI 1640 medium was added to the top wells. The chamber was incubated at 37°C with 5% CO~2~for 2 hours. The membranes were then removed, washed with PBS on the upper side, fixed, and stained with DiffQuik (Dade AG, Düdingen, Switzerland). Cells were counted in three fields per well at magnification ×800. All assays were performed in triplicate. In blocking experiments, cell suspensions were preincubated before chemotaxis assay for 30 min at 4°C with antihuman IP-10 antibodies at 20 μg/ml. In a few experiments, T cells purified from SF were evaluated for their migratory capability in response to CXCL10 (20 ng/ml and 200 ng/ml, R&D Systems).
Data are expressed as a migration index, which is the ratio between the number of migrating cells in the presence of the stimulus and that in medium alone.
Immunohistochemical analysis
----------------------------
Expression of CXCR3 and CXCL10 was detected by immunohistochemistry with anti-CXCR3 and anti-IP-10 antibodies, respectively (R&D Systems). Paraffin-embedded sections (4 μm thick) from patients and controls were used for immunostaining with the standard avidin--biotin complex method (Vectastain ABC kit; Vector Laboratories, Burlingame, CA, USA), as previously described \[[@B10]\].
Briefly, for the microwave antigen-retrieval procedure, slides were placed in a 2-L glass beaker containing 0.01 mol/L citrate buffer, pH 5.9, and microwaved at full power (800 W for 5 min, three times) before cooling and equilibration in PBS.
To neutralize endogenous peroxidase activity, we pretreated slides with 3% hydrogen peroxide for 5 min. Primary antibodies were applied at a concentrations of 1:100 for both antibodies (anti-hCXCR3 monoclonal antibody and anti-hIP-10/CXCL10 polyclonal antibody) for 1 hour in a humidified chamber at 37°C. Immunoreactivity was detected using biotinylated secondary antibodies (1:50 rabbit antigoat and 1:1000 goat antirabbit antibodies diluted in PBS--bovine serum albumin buffer) incubated for 45 min, followed by a 30-min incubation with avidin--peroxidase (1:200) and visualized by a 7-min incubation with the use of 0.1% 3,3\'-diaminobenzidene tetrahydrochloride as the chromogen. Thereafter the slides were rinsed and washed with PBS for 5 min, and the sections were counterstained with Mayer\'s hematoxylin. The last steps were performed at room temperature. Control slides were incubated with Tris-buffered saline containing isotype-matched antibodies instead of the primary antibody; they were invariably negative (data not shown). The intensity of antibody staining was classified as strong, moderate, weak, and negative. Parallel control slides were prepared either lacking primary antibody or lacking primary and secondary antibodies, or were stained with normal sera to control for background reactivity.
Immunohistochemistry for the characterization of inflammatory infiltrate, endothelial cells, and synovial cells was carried out using the following monoclonal antibodies CD45 (1:20), CD45RO (1:100), CD20 (1:100), CD68 (1:50), CD4 (1:100), CD8 (1:100), CD31 (1:30) (all from Dako Glostrup, Denmark), and cytokeratin--CAM 5.2 (1:1 Becton Dickinson). The immunoreaction products were developed using the avidin--biotin--peroxidase complex method as described above.
Confocal microscopy
-------------------
In order to evaluate the expression of CXCL10 by synovial macrophages, confocal microscopy experiments were performed in three patients with JIA. Paraffined sections were prepared for immunofluorescent labelling. Briefly, primary antibodies against CD68 and IP-10 (diluted 1:50 and 1:1 00, respectively, in PBS with 5 g/L bovine serum albumin and 1 g/L gelatin, respectively) and secondary antibodies (goat antimouse IgG and donkey antigoat IgG) conjugated with Texas red or Alexa 488 (Sigma, Milan, Italy) were used. Double labelling using both antibodies on the same section was performed. Primary antibodies and secondary antibodies were incubated for 1 hour at room temperature. Nuclear staining was carried out with DAPI (4\' 6-diamidino-2-phenyindole; Sigma) in PBS. Slides were stored at 4°C and analyzed within 24 hours. As a control, the primary antibody was omitted.
Immunofluorescence was observed with a Leica TCS SL spectral confocal and multiphoton system (Leica, Heidelberg, Germany). We used an argon laser at 488 nm in combination with a helium neon laser at 543 nm to excite the green (CD68) and red (IP-10) fluorochromes simultaneously. Emitted fluorescence was detected with a 505--530-nm band-pass filter for the green signal and a 560-nm long-pass filter for the red signal.
RT-PCR
------
RNA was extracted from the tissues using TRIzol reagent (Invitrogen, San Giuliano Milanese, Milan, Italy). The concentration of RNA was estimated by spectrophotometer. The RNA was treated with DNase I (Invitrogen) to remove any genomic DNA that might contaminate the RNA preparations. Complementary DNA (cDNA) was prepared using a synthesis kit (SuperScript II DNA Preamplification System; Invitrogen). A cDNA reaction mixture from 0.1 μg of RNA was used for DNA amplification by PCR. A typical amplification reaction included 2 units of *Taq*polymerase (Takara, Shiga, Japan), 20 pmol of sense and antisense oligonucleotide primers, and 200 μM each of dATP, dCTP, dGTP, and dTTP. Amplification was carried out for 30 cycles: 1 min at 92°C, 1 min at 55°C, and 1 min at 72°C. The amplified DNA was electrophoresed on a 2% agarose gel (Invitrogen), stained with ethidium bromide, visualized under ultraviolet light, and photographed.
The primer sequences used were as follows: for glyceraldehyde-3-phosphate dehydrogenase (GAPDH), 5\'-TCC-ATG-ACA-ACT-TTG-GTA-TCG-3\' (sense) and 5\'-GTC-GCT-GTT-GAA-GTC-AGA-GGA-3\' (antisense); for CXCR3, 5\'-TTG-ACC-GCT-ACC-TGA-ACA-TA-3\' (sense) and 5\'-ACG-TCT-ACC-CTG-CTT-TCT-CG-3\'. The expected sizes for the cDNA amplicons were as follows: 376 bp for GAPDH, 377 bp for IP-10, and 456 bp for CXCR3. All assays were performed in triplicate.
The number of cycles (30) was chosen to ensure that the amount of products synthesized was proportional to the amount of specific mRNA in the original preparation.
After PCR amplification, PCR products (15 μl) were subjected to electrophoresis on 2% agarose gels containing 0.03 μg/ml ethidium bromide. The quantification of transcript level was carried out by scanning photographs of gels and analyzing the area under peaks, using Quantity one Biorad software. Levels of mRNA expression were normalized by calculating them as a percentage of 3GAPDH mRNA expression levels \[[@B11]\]. The band intensity for 3GAPDH did not differ significantly between experiments.
Statistical analysis
--------------------
Data were analyzed with the assistance of the Statistical Analysis System. Data are expressed as means ± standard deviation. Mean values were compared using the ANOVA test.
Results
=======
Immunohistochemical analysis of the expression and cellular distribution of CXCL10 in the synovial membrane during JIA
----------------------------------------------------------------------------------------------------------------------
Immunohistochemical analysis was used to investigate the pattern of expression of this chemokine in synovial membranes from nine children with JIA and three age-matched controls. All the JIA synovial tissues showed moderate or strong staining for CXCL10 (Table [1](#T1){ref-type="table"}). As shown in Fig. [1a](#F1){ref-type="fig"} and, at higher magnification, in Fig. [1b](#F1){ref-type="fig"}, CXCL10 was demonstrated on the surface of three types of cells, specifically macrophages, epithelial cells, and endothelial cells, as determined by cell morphology. Most of the IP-10-expressing cells were macrophages. Matched controls revealed no CXCL10 staining (Fig. [1c,d](#F1){ref-type="fig"}). In order to verify whether macrophages express CXCL10 morphology, data were confirmed by the use of confocal microscopy. As shown in Fig. [2](#F2){ref-type="fig"}, double staining with CD68 and CXCL10 clearly demonstrated that CD68^+^macrophages showed an intense coexpression of the chemokine.
CXCL10 is present in synovial fluid from patients with JIA and mediates chemotactic activity
--------------------------------------------------------------------------------------------
To evaluate if CXCL10 is released into the SF and is capable of inducing T-cell migration, the chemotactic activity of supernatants from the SF of four patients with JIA was tested on a T-cell clone expressing high levels of CXCR3 (300-19). As shown in Fig. [3](#F3){ref-type="fig"}, SF of all the patients we studied exerted significant chemotactic activity on the CXCR3^+^T-cell line. The addition of an anti-CXCL10 neutralizing antibody (α CXCL10) but not of a control antibody inhibited chemotactic activities, suggesting the presence of IP-10/CXCL10 in SF and its responsibility in the chemotaxis of CXCR3^+^cells. In a second set of experiments, T cells purified from SF exhibited a definite migratory capability per se, which was significantly enhanced in response to CXCL10. Two representative experiments are represented in Fig. [4](#F4){ref-type="fig"}.
Immunohistochemical and flow cytometry analysis of the expression of CXCR3 by PB, SF, and synovial-tissue T lymphocytes in JIA
------------------------------------------------------------------------------------------------------------------------------
The possibility that CXCL10 in synovial fluid and membrane might account for the recruitment of CXCR3^+^T lymphocytes from the bloodstream to the synovium was investigated by immunohistochemical analysis of the expression of this chemokine receptor. All the JIA patients showed CXCR3-expressing lymphocytes infiltrating the synovium, with strong or moderate staining intensities (see Table [1](#T1){ref-type="table"}). Particularly strongly stained cells were observed close to the perivascular area (as in Fig. [5a,b](#F5){ref-type="fig"}, showing two different magnifications of the same slide). In a few cases, a follicular pattern of strongly marked lymphocytes was visible close to the luminal space (Fig. [6](#F6){ref-type="fig"}). The control synovial tissues revealed no CXCR3 staining (Fig. [5c,d](#F5){ref-type="fig"}).
Densitometric analysis showed that CXCR3 mRNA levels were significantly higher in patients with JIA than in controls (CXCR3:GADPH ratio 2.25 ± 1.8 vs 0.6 ± 0.49, *P*\< 0.05) (Fig. [7](#F7){ref-type="fig"}).
Flow cytometry analysis confirmed the selective recruitment of CXCR3^+^lymphocytes into the synovium. We analyzed paired samples of PB and SF from 20 children with JIA, and in 18 of these patients, T lymphocytes isolated from the SF showed greater expression of CXCR3 with than did those from PB, both in terms of percentage of positive cells and of the MFI (*P*= 0.01) (Table [2](#T2){ref-type="table"}). Flow cytometry profiles for one representative patient are shown in Fig. [8](#F8){ref-type="fig"}. Taken together, these results strongly suggest a role for the CXCL10 released into the synovial compartment in the accumulation of its selective CXCR3-receptor expressing T cells.
Discussion
==========
JIA is characterized by a persistent accumulation in the synovial membrane of T lymphocytes most of which express surface markers indicative of activation, such as CD45RO, and a type-1 cytokine profile \[[@B4],[@B5]\]. The cellular infiltrate is defined largely by the composition of locally produced chemokines as well as by the diversity of circulating leukocytes expressing the relevant receptors. Our principal findings are that in JIA, CXCL10/IP-10 is strongly expressed in synovial membranes and is released into synovial fluid (SF), where it exerts a definite chemotactic activity on CXCR3^+^T-cell clones and on T cells purified from SF; and that there is an accumulation of CXCR3 expressing T lymphocytes from the bloodstream to the synovial fluid and membrane. These findings suggest a role for CXCL10 in the mechanism of T-cell activation and recruitment into the inflamed synovium.
The high expression of CXCR3 by T cells retrieved from the synovia of patients with JIA might be considered a by-product of the *in vivo*cell hyperactivity of the tissue T-cell compartment in this disease. In fact, recent data clearly indicate that CXCR3 and its ligands become functional on recently activated T cells \[[@B12]\]. After antigenic challenge or in response to stimulation through the T-cell receptor (TCR), T cells express CXCR3, respond with chemotaxis to CXCR3 ligands, and produce IFNγ. Furthermore, in the presence of persistent antigenic stimulations, CXCR3 expression is maintained and poised for rapid up-regulation with reactivation. We and other authors have previously shown that CXCR3/CXCL10 interaction is involved in the pathogenesis of other Th1-mediated processes, such as Crohn\'s disease and sarcoidosis \[[@B13],[@B14]\]. A similar sequence of events could take place in the synovia of children with JIA. In fact, as previously reported \[[@B15]\], the evaluation of the molecular organization of the TCR revealed that T cells proliferating in children with JIA show a preferential usage of definite TCR gene regions, indicating an ordered immune response in which a specific TCR has been triggered and CXCR3 expression is induced \[[@B16]\].
CXCL10 was expressed by macrophages in synovial membrane of patients with JIA but not of controls. This finding suggests that CXCL10 is part of the matrix of cytokines that regulates the accessory activity of macrophages at sites of inflammatory lesions in the synovial microenvironment. Since large amounts of type 1 inflammatory cytokines, such as IFNγ, tumor necrosis factor α, IL-15, and IL-18, have been detected in JIA synovium \[[@B7]\], it is likely that these cytokines act in concert, sustaining the local proinflammatory responses and up-regulating CXCL10 expression. In turn, since CXCL10 is known to be capable of up-regulating cytokine synthesis in human Th1 cells, it is likely that macrophage-derived chemokines as IL-18 and IL-15 could participate in the maintenance of the default Th1/Tc1 polarization seen during JIA inflammation. It should be noted that, as shown in Fig. [2](#F2){ref-type="fig"}, anti-IP-10 almost completely inhibited the migration of the CXCR3^+^300-19 T cells in response to synovial fluid. Given the ability of I-ITAC/CXCL11 and Mig/CXCL10 to favor T-cell recruitment \[[@B17]\], we are currently investigating whether this non-ERL chemokine may influence entry of T cells into the JIA synovia.
It remains to be established whether synovial endothelial cells express CXCL10 (Fig. [1a](#F1){ref-type="fig"}). In a previous report it has been shown that human umbilical-vein-derived endothelial cell monolayers stimulated with IFNγ and tumor necrosis factor α produce IP-10/CXCL10, retaining it on their surface, and that this leads to a rapid adhesion of T lymphocytes. This effect was drastically reduced by anti-CXCR3 monoclonal antibody \[[@B18]\]. Furthermore, it is known that unstimulated human umbilical-vein-derived endothelial cells are able to retain IP-10 added exogenously, through binding to cell-surface proteoglycans \[[@B19]\]. Finally, recent data have definitively demonstrated that human endothelial cells may express a previously unrecognized receptor for CXC chemokines named CXCR3B and derived from an alternative splicing of the CXCR3 gene \[[@B20]\]. This receptor shows higher affinity for CXCL10 than classic CXCR3, mediates the inhibition of endothelial-cell growth, and accounts for the known angiostatic capability of CXCL10. Thus, it is possible that nonspecific binding of IP-10 may be responsible for the CXCL10 positivity we observed on endothelial cells. Further studies are in progress to determine whether synovial endothelial cells express CXCR3B *in vivo*and, if this be the case, to determine the putative role of CXCR3B/IP-10 interactions on the balance of angiogenic/angiostatic events in the JIA synovia.
Previous studies on chemokines and their receptors in modulating the recruitment of leukocytes at the sites of inflammation suggested that targeting these molecules with engineered agents might have therapeutic utility in down-modulating inflammatory responses. Results of CXCR3 or IP-10/CXCL10 blockade have already been reported in animal models. Recently, some authors have shown a rapid and marked improvement of adjuvant-induced arthritis in rats treated with IP-10 DNA vaccine \[[@B21]\]. Moreover, anti-mCXCR3 neutralizing antibodies were found to inhibit Th1 lymphocyte recruitment to peripheral inflammatory sites in a mouse model \[[@B22]\]. Further studies are needed in animal models to explore the therapeutic potential of CXCR3- or CXCL10-antagonists, with the ultimate goal of offering new clues for immune intervention in Th1-mediated diseases such as JIA and rheumatoid arthritis.
Conclusion
==========
Our results provide the first evidence of the functional role of CXCR3/CXCL10 interactions in mediating recruitment of T cells at sites of synovial inflammation in JIA. An in-depth molecular study of mechanisms regulating overexpression of CXCR3/CXCL10 might help in defining the role of these molecules in synovial inflammatory responses, offering new insights into elements controlling the immune response within joints.
Abbreviations
=============
cDNA = complementary DNA; GAPDH = glyceraldehyde-3-phosphate dehydrogenase; IFNγ = interferon γ ; IL = interleukin; JIA = juvenile idiopathic arthritis; PB = peripheral blood; PBS = phosphate-buffered saline; PCR = polymerase chain reaction; RT-PCR = reverse transcriptase PCR; SF = synovial fluid; TCR = T-cell receptor; Th1 = T helper cell type 1.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
GM conceived and coordinated the study and drafted the manuscript. FZ participated in the design of the study. FC performed the immunohistochemistry and helped to draft the manuscript. MB and MF carried out the chemotaxis. AC performed the flow cytometry experiments. MV participated in the immunohistochemistry. FZ participated in the design of the study. CA conceived the study and helped in the draft of the manuscript. All authors read and approved the final manuscript
Acknowledgements
================
This work was supported by a grant from the Regione Veneto (Venice, Italy) and COFIN MIUR 2002 (No. 2002068787002).
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
IP-10/CXCL10 expression in the synovium of a patient with juvenile idiopathic arthritis. Few inflammatory cells showing moderate staining; original magnification ×50 **(a)**, ×100 **(b)**. Negative staining in control patient: panoramic view **(c)**(original magnification ×25) and particular view **(d)**(original magnification ×50).
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Expression of IP-10/CXCL10 in the synovium of a representative patient with JIA. Immunofluorencence confocal laser scanning microscopy indicates the presence of chemokine IP-10 (red) **(a)**; **(b)**the same cells are shown to be synovial macrophages, as they are marked with CD68 (green). **(c)**The co-localization of IP10 and CD68 by macrophages (brown) is clearly visible. Original magnification ×1000.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Chemotactic activity of 300-19 cells in the presence of synovial fluid alone (grey bar), synovial fluid with an anti-CXCL10 neutralizing antibody (αCXCL10) (black bar), and synovial fluid with a control antibody (white bar) from four representative patients with juvenile idiopathic arthritis.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Chemotactic activity migration indices of T cells from synovial fluids of two representative patients with juvenile idiopathic arthritis in the presence of RPMI 1640 medium alone or medium containing CXCL10 at 20 ng/ml or at 200 ng/ml.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
CXCR3 expression in the synovium of a patient with juvenile idiopathic arthritis. Note the marked staining of inflammatory cell infiltrate in the perivascular area \[original magnification ×50 **(a)**, ×100 **(b)**\]. Negative staining in control patient: panoramic view **(c)**(original magnification ×25) and particular view **(d)**(original magnification ×50).
:::
:::
::: {#F6 .fig}
Figure 6
::: {.caption}
######
CXCR3 expression in juvenile idiopathic arthritis synovium. A follicular pattern of strongly marked lymphocytes is visible close to the lumen surface. Original magnification ×25.
:::
:::
::: {#F7 .fig}
Figure 7
::: {.caption}
######
Semiquantitative RT-PCR determination of CXCR3 expression in patients and controls. Unnumbered frame: DNA marker. Representative results of agarose-gel electrophoresis of RT-PCR products of CXCR3 mRNA (456 bp) and glyceraldehyde-3-phosphate dehydrogenase (234 bp) for nine patients (frames 1--9) and three controls (frames 10--12).
:::
:::
::: {#F8 .fig}
Figure 8
::: {.caption}
######
Flow cytometry profile of CXCR3 expression in peripheral blood (PB) and synovial fluid (SF) lymphocytes from patient 3 and a control subject.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
CXCR3 and CXCL10 expression in patients with juvenile idiopathic arthritis and controls
:::
Subject no. Sex Age at onset (months) CXCR3 CXCL10
------------- ----- ----------------------- ------- --------
Patients
1 F 54 ++ ++
2 F 34 +++ +++
3 F 84 +++ ++
4 F 70 +++ ++
5 F 65 +++ ++
6 F 156 +++ ++
7 F 141 +++ +++
8 M 70 +++ ++
9 F 42 +++ +++
Controls
1 F 72 \+ \+
2 F 36 \+ \+
3 M 24 \-- \+
+++, strong; ++, moderate; +, weak; \--, negative.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
CXCR3 expression in peripheral blood (PB) and synovial fluid (SF) lymphocytes in five representative patients with juvenile idiopathic arthritis
:::
Mean fluorescence of CXCR3^a^
--- ------------------------------- ------- ------
1 35.79 55.8 36.5
2 23.02 77.13 36
3 20.27 48.88 22.4
4 15.84 27.75 34
5 16.44 20.59 20.2
^a^*P*≤ 0.001 in every case. ^b^On the Kolmogorov--Smirnov test; D/s values \>10, and *P*values \<0.05 were considered significant. D/s is calculated as a function of the number of data; it ranged from 0.5 to 100 and is a measure of the significance of the difference between two distributions.
:::
|
PubMed Central
|
2024-06-05T03:55:54.939615
|
2005-1-7
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065320/",
"journal": "Arthritis Res Ther. 2005 Jan 7; 7(2):R241-R249",
"authors": [
{
"first": "Georgia",
"last": "Martini"
},
{
"first": "Francesco",
"last": "Zulian"
},
{
"first": "Fiorella",
"last": "Calabrese"
},
{
"first": "Marta",
"last": "Bortoli"
},
{
"first": "Monica",
"last": "Facco"
},
{
"first": "Anna",
"last": "Cabrelle"
},
{
"first": "Marialuisa",
"last": "Valente"
},
{
"first": "Franco",
"last": "Zacchello"
},
{
"first": "Carlo",
"last": "Agostini"
}
]
}
|
PMC1065321
|
Introduction
============
Rheumatoid arthritis (RA) is a common chronic inflammatory and destructive arthropathy that consumes substantial personal, social, and economic costs. The synovial membrane in patients with RA is characterized by hyperplasia, by increased vascularity, and by an infiltration of inflammatory cells, including activated macrophages \[[@B1]\]. Activated macrophages presenting in large numbers of arthritic joints play an active role in RA \[[@B2]\] and other inflammatory diseases \[[@B3]\] by producing cytokines that drive subsequent inflammatory reaction.
Folate receptor (FR) is a 38-kDa glycosyl phosphatidylinositol-anchored protein that binds the vitamin folic acid with high affinity (\< 1 nM) \[[@B4],[@B5]\]. With the exception of the kidney and the placenta, normal tissues express low or undetectable levels of FR \[[@B4]\]. Previously it has been reported that FR has three isoforms: FR-α, FR-β, and FR-γ. Among them, FR-β, a nonepithelial isoform of FR, is expressed on activated synovial macrophages but not on resting synovial macrophages \[[@B6]\]. Folate derivatization might therefore be exploited to target activated macrophages involved in inflammatory joint disease. Turk and colleagues \[[@B7],[@B8]\] have recently used folate-^99m^Tc for assaying the participation of activated macrophages in an adjuvant-induced arthritis model, and have shown that folate-^99m^Tc selectively targets activated macrophages. This suggests that folate-linked imaging agents warrant further scrutiny as possible tools for evaluating arthritis.
A newly synthesized folic acid and near-infrared fluorochrome conjugate (NIR2-folate) was recently used as a FR-targeting imaging probe *in vivo*\[[@B9],[@B10]\]. Fluorescence in the near-infrared spectrum (700--900 nm) was used for *in vivo*imaging because it allows efficient photon migration through the tissues and has minimal autofluorescence \[[@B11]\]. The use of near-infrared fluorescent (NIRF) *in vivo*imaging probes has been shown to significantly enhance tumor detection \[[@B12]-[@B15]\], to facilitate identification of small preneoplastic lesions \[[@B16]\], and to allow objective assessment of new therapeutic paradigms \[[@B17]\] in animal studies. The NIRF imaging technology has recently been extended to arthritic studies. *In vivo*NIRF imaging of arthritis in experimental animals was demonstrated using a protease-sensitive probe and NIRF-labeled antibody \[[@B18]-[@B21]\]. The goal of the present study is to determine whether a fairly abundant FR on activated macrophages in the arthritic inflammatory process could serve as a target for NIRF-enhanced optical imaging.
Materials and methods
=====================
Imaging probe
-------------
The folate-targeting optical probe NIR2-folate, consisting of a near-infrared fluorochrome (NIR2) and folic acid, was synthesized and characterized as previously described \[[@B9],[@B10]\]. Briefly, folic acid was first reacted with 2,2\'-(ethylenedioxy) bis(ethylamine) using di-isopropylcarbodimide as the coupling agent in dimethyl sulfoxide. The *N*-hydroxysuccinimide-activated ester of NIR2 \[[@B22]\] was then coupled with the amino-derivatized folic acid. The final conjugate was purified by C-18 reverse-phase HPLC and confirmed by mass spectroscopic analysis. The NIR2-folate has an excitation wavelength maximum at 662 nm and an emission wavelength maximum at 686 nm.
Animal preparation and arthritis models
---------------------------------------
All animal studies were approved by the Institutional Animal Care Committee. Carbon dioxide inhalation was used for euthanasia. C57BL/6 mice (Jackson Laboratory, Bar Harbor, ME, USA) weighing 19--21 g, 12 weeks old, were handled in accordance with government guidelines. Lipopolysaccharide (LPS) intra-articular injection and KRN transgenic mice serum transfer served as two mice arthritis models in this study.
The LPS induction arthritis model was achieved according to published protocols \[[@B23],[@B24]\]. Mice were anesthetized with ketamine (90 mg/kg) and xylazine (10 mg/kg) intraperitoneally, and then LPS (Sigma, St Louis, MO, USA), 10 μg in 20 μl saline, was injected intra-articularly into the right ankle joint through the Achilles tendon using a 30-gauge needle. As a control, the same volume of normal saline was injected in the opposite ankle joint of the same animal.
The KRN transgenic mice were a gift from Dr D Mathis and Dr C Benoist (Joslin Diabetes Center, Boston, MA, USA). Blood was obtained from arthritic adult KRN mice, and the sera containing arthritogenic autoantibodies were pooled \[[@B18],[@B25],[@B26]\]. One hundred micoliters of KRN mice serum were intravenously injected into healthy C57BL/6 mice, and the NIR2-folate probe was then given at different time points after serum transfer to detect early inflammatory changes.
Experimental groups
-------------------
In the LPS induction model, the three experimental groups of animals were injected intravenously with NIR2-folate probe (2 nmol per animal, *n*= 12), with free NIR2 (2 nmol per animal, *n*= 5), or with 600-fold of folic acid (1200 nmol per animal) 5 min prior to NIR2-folate probe injection (2 nmol per animal, *n*= 5) to demonstrate the competition effect of free folic acid against the probe. In the KRN serum transfer model, four animals were intravenously injected with 100 μl KRN serum and the NIR2-folate probe was given 24 hours (*n*= 1) or 96 hours (*n*= 3) after serum transfer.
*In vivo*NIRF reflectance imaging and lesion assessment
-------------------------------------------------------
All animals were imaged in a prone position using a home-built NIRF reflectance imaging system, which has been described elsewhere \[[@B27]\]. For fluorescence acquisition, a 615--645 nm excitation filter and a 680--720 nm emission filter (Omega Optical, Brattleboro, VT, USA) were used. Images were analyzed using commercially available software (Digital Science 1D software; Kodak, Rochester, NY, USA). Following data acquisition, postprocessing and visualization were performed using the in-house program CMIR Image. The enhancement ratio of the inflamed joint was used to demonstrate the effectiveness of the probe, which was defined by the fluorescence signal intensity (SI) at the affected ankle joint divided by the fluorescence SI at the opposite ankle joint. NIRF images were acquired preinjection and postinjection at different time points.
Histology, immunohistochemistry, and immunofluorescent microscopy assessment
----------------------------------------------------------------------------
Ankles were excised and fixed in phosphate-buffered formalin for 24 hours, and were subsequently decalcified in 10% EDTA for 48 hours, paraffin embedded, cut into 8-μm sections, and stained with H&E. Immunohistochemistry was performed using an anti-activated macrophage antibody \[[@B28]\] (Mac-3, 1:500 dilution, rat anti-mouse monoclonal antibody; BD Biosciences, San Diego, CA, USA) and a goat anti-human folate receptor polyclonal antibody (sc-16387, 1:100 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA), revealed with biotinylated rabbit anti-rat and donkey anti-goat secondary antibodies (1:250 dilution; Santa Cruz Biotechnology). The staining procedure was performed with a modified avidin--biotin--peroxidase complex technique. The slides were visualized with a chromogen of diaminobenzidine (Vectastain; Vector Laboratories, Burlingame, CA, USA). Sections were counterstained with hematoxylin (Vector Laboratories). Positive immunoreactions appeared as dark brown staining on a blue background. Control sections were processed identically but with incubation of the nonspecific isotype immunoglobulin (Vector Laboratories).
Immunofluorescence staining was performed using Mac-3 rat anti-mouse monoclonal antibody (1:500 dilution) and FITC-conjugated anti-rat secondary antibody (1:250 dilution; Vector Laboratories). The inflamed ankles were cut into 10-μm thick slices using a Leica CM 1900 cryotome (Leica, Bannockburn, IL, USA). Slices were analyzed using an inverted epifluorescence microscope (Axiovert; Zeiss, Thorn-Wood, NY, USA). FITC and Cy5.5 channels were used for Mac-3 and NIR2-folate fluorescence signal detection. A cooled CCD camera (Sensys; Photometrics, Tucson, AZ, USA) adapted with a bandpass filter was used for image capture, and IPLab software (Scanlytics, Fairfax, VA, USA) was used for image analysis.
Statistical analysis
--------------------
Data are presented as the mean and standard error of the mean. Statistical analysis of the fluorescence SI and the enhancement ratio between different groups was conducted using a two-tailed paired Student *t*test. The paired Student *t*test was used for analyzing the SI difference between bilateral ankles in the same mouse. *P*\< 0.05 was considered to indicate a statistically significant difference. All statistics were analyzed using Stata 7.0 (Stata, College Station, TX, USA) for Windows (Microsoft, Redmond, WA, USA).
Results
=======
Establishment of a LPS-induced arthritis model
----------------------------------------------
Progressive discoloration and swelling of the ankle joints was noted 24 hours after LPS intra-articular injection. Abundant polymorphonuclear cell infiltration was noted in the synovial lining layer and the subsynovial adipose tissue in histologic sections 48 hours after LPS injection. Immunohistochemistry revealed Mac-3-positive and FR-positive cells scattered among polymorphonuclear cells and subsynovial tissues in adjacent tissue sections (Fig. [1](#F1){ref-type="fig"}). These findings indicate that arthritis can be induced by LPS, and that the presence of active macrophages within inflammatory tissues can be used as a target for the NIR2-folate probe.
NIRF imaging of a LPS-induced mice arthritis model
--------------------------------------------------
The NIR2-folate probe was injected 48 hours after LPS induction (*n*= 12). The fluorescence SI of the inflamed joints was significantly higher than the opposite ankle joint at 2 min, and 12, 24, 48, and 72 hours after probe injection (468 ± 51 arbitrary units \[AU\] versus 303 ± 33 AU, 400 ± 31 AU versus 181 ± 18 AU, 310 ± 18 AU versus 137 ± 8 AU, 209 ± 14 AU versus 111 ± 7 AU, and 144 ± 14 AU versus 80 ± 4 AU; *P*\< 0.001 in all sets) (Fig. [2](#F2){ref-type="fig"}). There was no significant difference in the preinjection fluorescence SI in bilateral ankle joints (85 ± 6 AU versus 82 ± 7 AU, *P*\> 0.05).
The average enhancement ratio of the inflamed joint was up to 2.3-fold in the first 12 and 24 hours after probe injection, and remained at 1.8-fold 72 hours after probe injection (Fig. [3](#F3){ref-type="fig"}). In comparison, the NIR2-free dye group (*n*= 5) showed a persistent lower enhancement ratio than the probe injection group at all time points (Fig. [3](#F3){ref-type="fig"}). The average enhancement ratios of the inflamed ankles in the NIR2-free dye group and the NIR2-folate group at 24-hour, 48-hour, and 72-hour time points were 1.6 ± 0.1 versus 2.3 ± 0.1, 1.3 ± 0.1 versus 1.9 ± 0.1, and 1.3 ± 0.03 versus 1.8 ± 0.1 (*P*\< 0.05), respectively. To understand the possible mechanism, folic acid was used to compete with the probe. In the folic acid competition study (*n*= 5), 600-fold folic acid (1.2 μm per animal) was given intravenously 5 min before the NIR2-folate injection. The enhancement ratio of the arthritic joint in the folic acid competition group was significantly lower than that of the NIR2-folate injection group (1.1 ± 0.1 versus 1.6 ± 0.1, *P*\< 0.05).
Colocalization of NIRF signal with Mac-3 immunofluorescence
-----------------------------------------------------------
Immunofluorescence of the LPS-treated arthritic joint showed scattered Mac-3-positive cells in the inflammatory tissues in the FITC channel (Fig. [4a](#F4){ref-type="fig"}), whereas NIR2-folate uptake cells were seen in the near-infrared channel using an inverted epifluorescence microscope (Fig. [4b](#F4){ref-type="fig"}). In the superimposed image (Fig. [4c](#F4){ref-type="fig"}), the Mac-3-positive cells colocalized well with NIR2-folate uptake cells.
Establishment of a KRN serum transfer mice arthritis model
----------------------------------------------------------
There was no visible swelling or discoloration at peripheral joints in the first 2 days after KRN serum transfer. Progressive discoloration and swelling of the peripheral joints was noted 3 days after serum transfer in sick KRN mice (Fig. [5a](#F5){ref-type="fig"}). In histological sections, Mac-3-positive cells intermingled among polymorphonuclear cells, and pannus formation was noted in the affected joints (Fig. [5b,c](#F5){ref-type="fig"}).
NIRF imaging of a KRN serum transfer mice arthritic model
---------------------------------------------------------
NIR2-folate was first given intravenously 4 days after KRN mice serum transfer. At this time point, discoloration and swelling of the affected peripheral joints was clearly observed (Fig. [5a](#F5){ref-type="fig"}). An intense fluorescence signal was found in peripheral joints (Fig. [5d](#F5){ref-type="fig"}). The NIR signal of the affected joints was 1.5-fold to 3.5-fold (average, 2.4-fold) higher than that of the unaffected joints.
To evaluate its ability for early detection of the inflammatory process, NIR2-folate was then given intravenously at a much earlier time point -- 24 hours after serum transfer. No gross swelling or discoloration at peripheral joints could be observed (Fig. [6a](#F6){ref-type="fig"}). Six hours after the NIR2-folate probe injection (30 hours after serum transfer), however, the NIRF reflectance imaging showed a 1.8-fold increase in the fluorescence signal at the right wrist joint as compared with the opposite site (Fig. [6b](#F6){ref-type="fig"}). The correlated histology showed an increased amount of inflammatory cells at the affected joint compared with the opposite wrist (Fig. [6c,d](#F6){ref-type="fig"}). Abundant Mac-3-positive cell infiltration at the right wrist joint region was also revealed by immunohistochemistry (Fig. [6e](#F6){ref-type="fig"}).
Discussion
==========
Activated macrophages are thought to be intimately involved in the pathogenesis of RA by directly destroying articular tissue, secreting metalloproteinases, and attracting or activating other immune cells via the release of cytokines \[[@B2],[@B29]\]. The quantitation of activated macrophages in joint tissues might consequently be of diagnostic value because activated macrophage content correlates well with articular destruction and poor disease prognosis in humans \[[@B2],[@B30]\]. Because FR expression may coincide with macrophage activation \[[@B6]\], we hypothesized that arthritic joints could be imaged using folate-derivatized fluorescent imaging agents. The present studies demonstrated that the folate-targeted NIRF probe can indeed selectively target activated macrophages *in vivo*, and that folate-linked imaging agents can facilitate the noninvasive analysis of inflammatory activity *in situ*.
Two different animal arthritis models were used in this study. The LPS induction model was established by intra-articular injection of LPS, which induces transient synoviocyte hyperplasia and polymorphonuclear cell infiltration \[[@B23],[@B24],[@B31],[@B32]\]. The advantage of the LPS induction model is that the opposite ankle joint could be used as an internal control, thus demonstrating the effectiveness of the probe in statistical analysis. The entity of this model, however, is a bacterial toxin-induced arthritis that resembles pyogenic arthritis instead of RA. The second model was established by transferring serum of sick KRN mice into healthy B6 mice, which induces synovial polymorphonuclear cells and macrophage infiltration by arthritogenic immunoglobulins \[[@B18],[@B26],[@B33]\]. The KRN serum transferred model resembles human RA because both are chronic symmetric joint diseases with pannus formation and destructive bone and cartilage erosion, predominantly of the distal joints.
The enhancement ratio of inflamed joints in the LPS model was slightly increased in the NIR2-free dye injection group during the first 24 hours after NIR2 injection. This might be due to nonspecific phagocytosis by activated macrophages, or due to NIR2-free dyes pooled at the interstitial space because of increased vascular permeability at the inflammation tissues. However, the enhancement ratio of the inflammatory joints in the NIR2-folate injection group was significantly higher than that of NIR2 injection group, which was more prominent 48 hours after injection (Fig. [3](#F3){ref-type="fig"}). Most of the NIR2-free dye began to be washed out from the inflamed joints, but NIR2-folate remained at the inflamed joints 72 hours after injection. The data indicate that the NIR2-folate probe has significant advantages over nonspecific fluorochromes for *in vivo*imaging, the latter often being used for nontargeted image enhancement \[[@B34],[@B35]\].
Histological colocalization of the infiltrated Mac-3-positive and FR-positive cells was found to correlate well in the inflammatory tissues (Fig. [1](#F1){ref-type="fig"}). The NIR2-folate uptake cells colocalized with Mac-3-positive cells using fluorescence microscopy (Fig. [4](#F4){ref-type="fig"}), which indicates that uptake of folate conjugates at inflammatory joints is mediated by activated macrophages. In addition, the *in vivo*competition study confirmed that free folate was able to compete with the NIR2-folate probe for FR binding. The average enhancement ratio of arthritic joints in the folic acid competition group was significantly lower than in the NIR2-folate group postadministration. The results support the fluorescent probe uptake being receptor dependent.
Another important finding of this study is the potential of applying this technique in early assessment of RA. Our results indicate that the folate-linked NIR fluorescence probe could detect mild inflammatory changes as early as 30 hours after arthritogenic antibody transfer, before any morphological changes can be observed. A sensitive imaging modality for assessment of early events in RA could provide valuable information for diagnosis and treatment \[[@B36]\].
^99m^Tc-folate has recently been used to assay the participation of activated macrophages in adjuvant-induced arthritis mice models using gamma scintigraphy as the imaging modality \[[@B7]\]. In contrast, optical imaging is a noninvasive method and does not depend on radiolabeled contrast agents such as those in nuclear medicine; there is thus no exposure of the patient to ionizing radiation. The present hindrance of optical imaging is that tissue penetration of light in living tissue may attenuate the SI. The near-infrared fluorescence probe allows the most efficient photon migration through the tissues \[[@B11]\]. In addition, there is less soft tissue around peripheral joints, which gives the near-infrared optical imaging a competitive role in the diagnosis of peripheral joint disease, especially in detection of early arthritis or assessment of treatment effects.
Conclusions
===========
The results indicate that it is feasible to image the activated macrophage status in inflamed joints *in vivo*at an early stage. The FR-targeting probe not only offers better assessment at early stages in inflammatory disease, but also improves the evaluation of future anti-inflammatory treatments. This technique may therefore represent a step toward the level of molecular diagnosis of arthritis.
Abbreviations
=============
AU = arbitrary units; FITC = fluorescein isothiocyanate; FR = folate receptor; H&E = hematoxylin and eosin; HPLC = high-performance liquid chromatography; LPS = lipopolysaccharide; NIRF = near-infrared fluorescent; RA = rheumatoid arthritis; SI = signal intensity.
Competing interests
===================
The author(s) declare that there are no competing interests.
Authors\' contributions
=======================
WC and CT participated in all experimental design, data collection and analysis, and drafted the manuscript. UM participated in the KRN experiments and drafted the manuscript. RW participated in the design and helped to draft the manuscript. All authors read and approved the final manuscript.
Acknowledgements
================
This research was supported in part by National Institutes of Health grants P01-A154904, P50 CA86355 and R24 CA92782. WTC was supported by the Taipei City Government.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Immunoperoxidase staining of **(a)**Mac-3 and **(b)**folate receptor (FR) at an arthritic ankle 72 hours after lipopolysaccharide induction. The Mac-3-positive and FR-positive cells morphologically correlated well in adjacent tissue sections. Magnification, 400 ×.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
*In vivo*near-infrared fluorescent (NIRF) imaging of inflammatory joints in the lipopolysaccharide (LPS) induction model. The NIR2-folate probe was intravenously injected 2 days after LPS intra-articular injection. **(a)**White-light images obtained 48 hours after intra-articular LPS injection at the right ankle joint; soft tissue swelling was noted at the affected joint. **(b)**NIRF images obtained 24 hours after NIR2-folate injection. Note the strong fluorescence signal in the LPS-treated ankle compared with the opposite control side (enhancement ratio = 2.31). **(c)**A merged NIRF signal with a white-light image showing specific increased fluorescence signal intensity at the affected joint. **(d)**H&E-stain section of the right ankle joint showing abundant inflammatory cell infiltration at subsynovial tissues. Original magnification, 100 ×. **(e)**NIRF images of a longitudinal section of the LPS-treated ankles. Pseudo-color coding was used to demonstrate the stronger fluorescence signal surrounding the ankle joint.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Enhancement ratio of lipopolysaccharide (LPS)-treated inflamed ankles in NIR2-folate (*n*= 12) and NIR2-free dye (*n*= 5) injection groups at different time points. A significantly higher enhancement ratio was noted in the NIR2-folate injection group at 24-hour, 48-hour and 72-hour time points (*P*\< 0.05).
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Colocalization of Mac-3-positive cells and NIR2-folate uptake cells in a lipopolysaccharide-induced arthritic ankle 48 hours after NIR2-folate injection. **(a)**Immunofluorescence staining for activated macrophage revealed in the FITC channel. **(b)**NIR2-folate uptake cells are revealed in the near-infrared fluorescent channel. **(c)**Superimposed image shows Mac-3-positive cells colocalized well with NIR2-folate cellular origins. **(d)**A negative control without primary antibody. Original magnification, 400 ×.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Establishment of the KRN serum transfer model. **(a)**Discoloration and swelling (arrow) of the right third proximal interphalangeal joint is noted in a healthy C57BL/6 mouse 4 days after KRN serum transfer. **(b)**Near-infrared fluorescent imaging of the right paw showed increase fluorescence signal intensity at the inflammatory joint (enhancement ratio = 1.9). **(c)**Correlated H&E-stain section showed abundant inflammatory cells infiltration with pannus-like formation. Original magnification, 100 ×. **(d)**Immunoperoxidase staining of Mac-3. Mac-3-positive cell infiltration among polymorphonuclear cells was noted in the pannus. Original magnification, 400 ×.
:::
:::
::: {#F6 .fig}
Figure 6
::: {.caption}
######
Early detection (30 hours after KRN serum transfer) of the inflammatory joint by NIR2-folate. **(a)**White-light image showed no remarkable swelling at bilateral paws. **(b)**Merged near-infrared fluorescent signal with a white-light image showed increase fluorescence signal intensity at the dorsal aspect of the right wrist, which has a 1.8-fold increase compared with the left wrist. **(c)**H&E-stain histology of the right wrist showed polymorphonuclear cell infiltration at the dorsal aspect of the right wrist (arrow). Magnification, 20 × (400 ×, insert). **(d)**Histology of the left wrist showed no remarkable inflammatory cell infiltration. Magnification, 20 × (400 ×, insert). **(e)**Immunohistochemistry of the right wrist showed Mac-3-positive cell infiltration at subsynovial tissues. Magnification, 400 ×.
:::
:::
|
PubMed Central
|
2024-06-05T03:55:54.942330
|
2005-1-14
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065321/",
"journal": "Arthritis Res Ther. 2005 Jan 14; 7(2):R310-R317",
"authors": [
{
"first": "Wei-Tsung",
"last": "Chen"
},
{
"first": "Umar",
"last": "Mahmood"
},
{
"first": "Ralph",
"last": "Weissleder"
},
{
"first": "Ching-Hsuan",
"last": "Tung"
}
]
}
|
PMC1065322
|
Introduction
============
Rheumatoid arthritis (RA) is the most common autoimmune disease in humans, affecting 1% of the population in western countries. Histologically, RA is characterized by hyperplasia and infiltration of the synovial membrane with mononuclear cells, development of an aggressive tissue called pannus and secretion of proteases, which are responsible for the destruction of articular cartilage and adjacent bone. It is well established that macrophages and synovial fibroblasts are effector cells of joint destruction, and it is presumed that autoreactive CD4^+^T cells are involved in their activation \[[@B1]\]. There is now a large body of evidence that, in rodents, regulatory T cells (T~reg~) actively control the activation of autoreactive T cells and thus maintain immunological self-tolerance. Apart from adaptive T~reg~cells, which can be induced by antigen-specific stimulation of conventional peripheral T cells under tolerogenic conditions (for review \[[@B2]\]), there is no doubt that naturally occurring T~reg~cells exist in healthy mice as well as in humans and rats, and these are characterized by constitutive expression of CD25 \[[@B3]-[@B5]\]. Absence of these cells *in vivo*results in a multi-organ autoimmune syndrome \[[@B3],[@B6]\]. These CD4^+^CD25^+^T~reg~cells leave the thymus as committed \'professional\' suppressor T cells \[[@B7]-[@B9]\], proliferate in the periphery, and acquire an effector/memory-like phenotype \[[@B10]\]. In unmanipulated mice, T~reg~cells can also be found in the CD25^-^compartment, based on the expression of the integrin α~E~β~7~\[[@B10],[@B11]\], possibly reflecting differences in developmental stages of these cells.
The exact role played by naturally occurring CD4^+^CD25^+^T~reg~cells in the pathogenesis of arthritis remains controversial. Arthritis is part of the autoimmune syndrome induced by transfer of CD25-depleted splenocytes into lymphopenic hosts \[[@B3]\], and CD4^+^CD25^+^cells are protective in collagen-induced arthritis \[[@B12]\]. However, Bardos and coworkers \[[@B13]\] ruled out a role for naturally occurring CD4^+^CD25^+^T~reg~cells in proteoglycan-induced arthritis.
To clarify this issue, we used the antigen-induced arthritis (AIA) model. AIA is a Tcell-dependent experimental arthritis that is induced by intra-articular injection of antigen (methylated bovine serum albumin \[mBSA\]) into knee joints of preimmunized mice \[[@B14],[@B15]\]. This results in an acute inflammatory reaction, which is characterized by exudation of neutrophils and fibrin, which later proceeds to a chronic arthritis with synovial hyperplasia, infiltration of mononuclear cells, and cartilage and bone destruction -- histopathological changes similar to those that occur in RA. Autoimmune responses against cartilage constituents such as collagen types I and II and proteoglycans are involved in rendering the disease chronic \[[@B16],[@B17]\]. Beyond the 100% incidence of arthritis, another major advantage of the AIA model is that the time point of induction of arthritis is known, allowing manipulation of CD4^+^CD25^+^T~reg~cell number *in vivo*at defined stages in the disease. Using depletion of CD25-expressing cells or transfer of CD4^+^CD25^+^cells, in the present study we demonstrated that T~reg~cells modulate the onset of AIA but are ineffective at later stages, calling into question their value as a new therapeutic approach to established chronic arthritis.
Methods
=======
Animals, arthritis induction and assessment
-------------------------------------------
For all animal experiments, female C57Bl/6 mice (Charles River, Sulzfeld, Germany; age range 6--10 weeks) were used. Animals were kept under standard conditions, fed a standard diet and given free access to water. All animal studies were approved by the government commission for animal protection.
At 21 and 14 days before arthritis induction, mice were subcutaneously injected with 100 μg mBSA (Sigma, Deisenhofen, Germany), emulsified in complete Freund\'s adjuvant (Sigma) supplemented to 2 mg/ml heat-killed *Mycobacterium tuberculosis*(strain H37RA; Becton Dickinson \[BD\], Heidelberg, Germany). Simultaneously, mice received 5 × 10^8^heat-inactivated *Bordetella pertussis*(Chiron-Behring, Marburg, Germany) intraperitoneally. Arthritis was induced by intra-articular injection of 100 μg mBSA in 25 μl phosphate-buffered saline (PBS) into the right knee joint cavity.
Arthritis severity was monitored by measurement of lateral joint diameter using a vernier caliper (Oditest, Kroeplin Längenmesstechnik, Schlüchtern, Germany). Histological severity of arthritis was scored in a blinded manner by two investigators (PKP and MG) in frontal knee joint sections, stained with haematoxylin and eosin and prepared as described previously \[[@B14]\]. Briefly, at least four sections per knee joint were semiquantitatively examined on a 0--3 point scale for each of the following: extent of synovial hyperplasia, mononuclear infiltration, cartilage destruction and pannus formation.
Antibodies and reagents
-----------------------
The following antibodies were grown and purified from the culture supernatants in our laboratory: anti-CD25 (PC61), anti-CD3 (145 2C11), anti-CD4-FITC and FITC-labelled anti-CD4-F(ab) (GK1.5), anti-CD8 (TIB105), anti-CD28 (37.51) and anti-Mac-1 (M1/70). The following antibodies and secondary reagents were purchased from BD Pharmingen (Heidelberg, Germany): PE-Cy5-labelled anti-CD4 (H129.9), biotinylated anti-α~E~β~7~(M290), biotinylated anti-CD25 (7D4), allophycocyanine or FITC-conjugated anti-CD25 (PC61), streptavidin-allophycocyanine and streptavidin-PE, and matched antibody pairs for ELISPOT analysis of IFN-γ (R4-6A2 and biotinylated XMG1.2) and IL-4 (BVD4 1D11 and biotinylated BVD6-24G2) production.
*In vivo*depletion
------------------
Mice were injected with 0.5 mg purified anti-CD25 antibody (PC61) 4 and 2 days before intra-articular antigen injection. Polyclonal rat IgG, purified from normal rat serum, was used as control. The degree of depletion was determined by fluorescence-activated cell sorting, using a non-cross-reactive biotin-labelled anti-CD25, FITC-labelled anti-CD4 and streptavidin-conjugated allophycocyanine. Measurement was performed using FACSCalibur^®^(BD) and data were analyzed using WinMDI <http://www.scripps.edu>.
Preparation, pre-activation and transfer of regulatory T cells
--------------------------------------------------------------
Pooled spleen and lymph node cells from naive C57Bl/6 donors or, if indicated, from immunized mice were incubated with anti-CD4-FITC (clone GK1.5) and anti-CD25-biotin (clone 7D4; BD). CD4^+^T cells were isolated using an anti-FITC-Multisort-Kit (Miltenyi Biotech, Bergisch-Gladbach, Germany) in accordance with the manufacturer\'s instructions. CD4^+^T cells were sorted into CD25^-^and CD25^+^cells using anti-biotin MicroBeads (Miltenyi Biotech). Purity was greater than 92% for CD4^+^CD25^-^and greater than 80% for CD4^+^CD25^+^cells.
CD25-expressing and α~E~β~7~-expressing subsets were sorted by FACS. Briefly, pooled spleen and lymph node cells from naive mice were stained with anti-CD25-FITC, anti-α~E~β~7~-biotin and streptavidin-PE. The stained cells were enriched with anti-FITC and anti-PE MicroBeads, using the AutoMACS separation unit (Miltenyi Biotech). Thereafter, the cells were sorted into subsets according to their expression of CD25 or α~E~β~7~using a FACSDiVa cell sorter (BD). The purity was 90--95%, as determined by FACS.
For activation, cells were cultured for 24--72 hours in the presence of plate-bound anti-CD3 (3 μg/ml), anti-CD28 (10 μg/ml) and rhIL-2 (100 U/ml; Chiron, Ratingen, Germany) in RPMI 1640 containing 10% fetal calf serum (FCS; Gibco, Karlsruhe, Germany). Thereafter, cells were washed with PBS and transferred intravenously via lateral tail vein into mice at the time point of AIA induction or at later time points when indicated.
Delayed-type hypersensitivity reaction
--------------------------------------
Seven days after arthritis induction, mice were challenged by intradermal injection into their ears of 5 μg mBSA in 10 μl PBS. Ear thickness was measured before injection and 24 and 48 hours later using a vernier caliper (Kroeplin).
Proliferation assay and ELISPOT analysis
----------------------------------------
Single cell suspension from spleens and lymph nodes (inguinal, popliteal, axillary) were cultured at a density of 1 × 10^6^/ml in RPMI 1640, containing 10% FCS, 2 mmol/l L-glutamine, 10 mmol/l Hepes, 1 mmol/l sodium pyruvate, 0.5 μmol/l 2-mercaptoethanol and antibiotics (100 U/ml penicillin, 0.1 mg/ml streptomycin; all from Gibco) in the presence of medium alone or 25 μg/ml mBSA for 72 hours in 96-well tissue culture plates (Greiner Bio One, Nürtingen, Germany). Cells were pulsed with 0.5 μCi \[^3^H\]thymidine (Amersham-Buchler, Braunschweig, Germany) for the last 18 hours of culture. Thereafter, cells were harvested onto 96-well glass fibre filters (Packard Bioscience, Groningen, The Netherlands), and \[^3^H\]thymidine incorporation was measured with a scintillation counter (Top-Count; Packard Bioscience).
For ELISPOT analysis, PVDF-membrane 96-well microplates (Millipore, Eschborn, Germany) were coated overnight at 4°C with the primary antibody diluted in sterile PBS. After washing, plates were blocked for 2 hours with RPMI 1640 containing 10% FCS. Thereafter 2 × 10^5^(IL-2 and IFN-γ) or 1 × 10^6^(IL-4) cells were cultured in duplicate wells for 24 (IL-2 and IFN-γ) or 48 hours (IL-4). After washing again plates were incubated overnight at 4°C with the secondary antibody diluted in PBS/1% BSA. Extravidin--alkaline phosphatase conjugate (1:30,000 in PBS/1% BSA) and BCIP/NBT solution (bromochloroindolyle phophate/nitroblue tetrazolium; both from Sigma) were used for spot development. The number of spots was quantified using a KS-ELISPOT-Reader (Carl Zeiss, Oberkochen, Germany).
Determination of serum IgG by ELISA
-----------------------------------
Microplates (96-well; Greiner Bio One) were coated with antigen (0.125 μg/ml mBSA), collagen type I (from rat tail tendon) and type II (10 μg/ml), and proteoglycans (10 μg/ml both from bovine cartilage) and left overnight, as described previously \[[@B14]\]. After washing, plates were incubated with serially diluted serum samples and the amount of bound IgG was determined using anti-mouse IgG-peroxidase conjugate (ICN, Eschwege, Germany) and ortho-phenylendiamine (Sigma) as substrate. Extinction was measured at 492 nm against 620 nm with an ELISA reader (Tecan, Crailsheim, Germany).
Cell transfer for *in vivo*homing assay
---------------------------------------
For *in vivo*homing assay, cells were sorted with a modified protocol and labelled with ^111^indium, as described elsewhere \[[@B10]\]. Briefly, CD4^+^cells were enriched by negative selection. Enriched CD4^+^T cells were stained with FITC-conjugated anti-CD4-F(ab) and anti-CD25-allophycocyanine and sorted into CD4^+^CD25^+^or CD4^+^CD25^-^cells by FACS (BD). Cells were labelled with ^111^In (Indiumoxin; Amersham-Buchler) for 20 min at room temperature; 1 × 10^6^labeled cells were injected intravenously, and 24 hours later mice were killed and the distribution of radioactivity in various organs and the rest of the body was measured in a γ-counter (Wallac Counter, Turku, Finnland).
Alternatively, a proportion of these cells was labelled with 5,6-carboxyfluorescein diacetate succinimidyl ester (CFSE) by incubation with 5 μmol/l CFSE (Molecular Probes, Leiden, The Netherlands) in RPMI 1640 for 5 min at room temperature. After washing, 1 × 10^6^cells were injected intravenously. Twenty-four hours later single cell suspensions were prepared from the draining and nondraining peripheral and mesenteric lymph nodes, the spleen and the peripheral blood, and stained with anti-CD4 and analyzed by FACS. Dead cells were excluded using propidiumiodide.
Statistical analysis
--------------------
Data are expressed as mean ± standard error of mean, unless otherwise indicated. Experimental groups were tested for statistically significant differences with the Mann--Whitney U-test using SPSS 10.0 (SPSS Inc, Chicago, IL, USA).
Results
=======
Depletion of CD25-expressing cells exacerbates antigen-induced arthritis
------------------------------------------------------------------------
Mice were injected intraperitoneally with 0.5 mg anti-CD25 (PC61) 4 and 2 days before induction of arthritis (i.e. 19 and 17 days after first immunization). Depletion of CD25-expressing cells was confirmed using FACS at the time of AIA induction (day 0) using an antibody that recognizes a different epitope on the CD25 molecule. In the PC61-treated group there was a 70.9 ± 11.4% (*n*= 3) reduction in CD4^+^CD25^+^cells in the spleens as compared with control mice injected with rat IgG (Fig. [1](#F1){ref-type="fig"}). Of note, the anti-CD25 treatment almost completely depleted cells with high expression of CD25, which are considered T~reg~cells, in contrast to CD4^+^T cells with low or intermediate levels of CD25 expression.
After intra-articular antigen injection, knee joint swelling of the CD25-depleted mice was significantly greater from day 3 onward than in the control group injected with rat IgG (Fig. [2a](#F2){ref-type="fig"}). Histological examination of knee joint sections 14 days after AIA induction revealed increased hyperplasia and infiltration of the synovial membrane, as well as increased articular damage in those animals (Fig. [2b--d](#F2){ref-type="fig"}). In summary, this indicates a marked exacerbation of AIA by depletion of CD25-expressing cells.
Increased cellular and humoral immune responsiveness in CD25-depleted mice
--------------------------------------------------------------------------
To assess how *in vivo*cellular immune responses against mBSA are influenced by depletion of CD25-expressing cells, delayed-type hypersensitivity (DTH) reaction against the same antigen was tested by intradermal injection of mBSA into the ears of mice at day 7 after induction of AIA. Anti-CD25 treated mice mounted a significantly stronger DTH response than did rat IgG-treated controls (Fig. [3a](#F3){ref-type="fig"}).
For analysis of the cellular immune responses *ex vivo*, draining lymph node cells of arthritic animals were harvested 14 days after AIA induction, restimulated with mBSA, and analyzed for proliferative response and cytokine production. As expected from the increased DTH reaction, the proliferative response to mBSA was significantly increased in cells from CD25-depleted mice as compared with that in rat IgG-treated controls (Fig. [3b](#F3){ref-type="fig"}). Importantly, even without antigenic stimulation the lymph node cells from CD25-depleted mice proliferated fourfold as much as cells from mice treated with control IgG. These data imply that a substantial proportion of the T-cell compartment is still activated 14 days after intra-articular antigen challenge in the absence of T~reg~cells.
Compatible with these findings is that the production of cytokines in response to mBSA was greater in CD25-depleted mice. Importantly, both T-helper-1 (IFN-γ) and T-helper-2 (IL-4) responses were aggravated by depletion of T~reg~cells, indicating that both types of response are subject to suppression by T~reg~cells (Fig. [3c](#F3){ref-type="fig"}). Again, cytokine secretion from T~reg~-depleted animals was increased even without antigenic stimulus. In accordance with this, serum levels of IgG directed against mBSA as well as levels of the cartilage-specific autoantigens collagen type I, collagen type II and proteoglycans, were found to be increased in CD25-depleted mice (Fig. [3d](#F3){ref-type="fig"}).
Taken together, these data clearly demonstrate that CD4^+^CD25^+^T~reg~cells regulate the severity of arthritis by limiting the cellular and humoral immune responses against the inducing antigen mBSA as well as some arthritis-related autoantigens.
Transfer of CD4^+^CD25^+^cells
------------------------------
To further characterize the suppressive potential of CD4^+^CD25^+^T~reg~cells, we performed cell transfer studies. In a first set of experiments we transferred T~reg~cells freshly isolated from naive (Fig. [4a](#F4){ref-type="fig"}) or mBSA/CFA immunized (Fig. [4b](#F4){ref-type="fig"}) mice into mBSA-immunized recipients at the time of intra-articular antigen challenge (day 0). With this protocol, a slight decrease in the severity of clinical arthritis (knee joint swelling) could be induced. Accordingly, the histological severity of AIA was also found to be reduced, albeit not statistically significantly (Fig. [4a, b](#F4){ref-type="fig"}).
It is known that T~reg~cells must be activated via their T-cell receptor to exert their suppressive function. Because we were unable to use antigen-specific (i.e. T-cell receptor transgenic) T~reg~cells, we opted to pre-activate the CD4^+^CD25^+^cells by *in vitro*culture in the presence of anti-CD3, anti-CD28 and IL-2 in order to increase their suppressive potential. Transfer of 1 × 10^6^pre-activated cells significantly suppressed both knee joint swelling and histological arthritis score (Fig. [4c](#F4){ref-type="fig"}). This effective suppression of AIA development was a consistent finding in different experiments, even with the use of lower cell numbers (for instance 2 × 10^5^cells; data not shown).
In the next step, we attempted to cure established arthritis by transfer of T~reg~cells. Surprisingly, 1 × 10^6^pre-activated CD4^+^CD25^+^cells had no influence on either knee joint swelling or histological arthritis score when transferred at day 1 (Fig. [5a](#F5){ref-type="fig"}) or day 7 (Fig. [5b](#F5){ref-type="fig"}) after induction of arthritis. Also, the transfer of 1 × 10^6^pre-activated α~E~β~7~-expressing T~reg~cells, which are highly effective in preventing AIA \[[@B10]\], had no effect on disease at this time point (Fig. [5c](#F5){ref-type="fig"}).
Taken together, our data demonstrate that T~reg~cells can inhibit arthritis development when transferred at the time of arthritis induction. However, we were unable to demonstrate any therapeutic effect of T~reg~cell transfer (in numbers that are effective in prevention) when performed after disease onset.
Transferred CD4^+^CD25^+^T~reg~cells do not suppress humoral or cellular immune responses
-----------------------------------------------------------------------------------------
Because CD25-depletion caused a substantial increase in both cellular and humoral immunoreactivity against mBSA, we examined whether transfer of CD4^+^CD25^+^T~reg~cells can suppress these responses. Neither DTH reactivity against mBSA (analyzed 7 days after AIA induction; Fig. [6a](#F6){ref-type="fig"}) nor mBSA-induced proliferation (Fig. [6b](#F6){ref-type="fig"}) and cytokine production by draining lymph node cells (Fig. [6c](#F6){ref-type="fig"}) at day 14 after induction of AIA were found to be suppressed in the recipients of 1 × 10^6^pre-activated CD4^+^CD25^+^cells. Thus, transfer of T~reg~cells into immunized animals does not eliminate or induce functional modification to the previously primed mBSA-specific immune response. In contrast, transfer of CD4^+^CD25^-^cells did significantly enhance the proliferation as well as the cytokine production in the recipients. Accordingly, serum levels of IgG directed against mBSA and the cartilage-specific autoantigens collagen type I and type II, and proteoglycans were also not significantly diminished in T~reg~cell recipients compared with the saline-treated control group. Recipients of CD4^+^CD25^-^cells had higher levels of IgGs (Fig. [6d](#F6){ref-type="fig"}).
Homing properties of CD4^+^CD25^+^T~reg~cells
---------------------------------------------
Because the mechanism of suppression of T~reg~cells *in vitro*is cell contact dependent, localization of cells might be important for their regulatory activity. Therefore, we investigated the migration behaviour of CD4^+^CD25^+^and CD4^+^CD25^-^cells *in vivo*. For these experiments CD4^+^cells were enriched by negative selection and sorted by FACS into CD4^+^CD25^+^and CD4^+^CD25^-^populations with preferential use of F(ab)-fragments or antibodies, which do not interfere with migration *in vivo*. Cells were labelled with ^111^In and injected intravenously into AIA mice 7 days after induction of arthritis. After 24 hours radioactivity was measured in different organs.
Compared with CD4^+^CD25^-^cells, CD4^+^CD25^+^T~reg~cells were less abundant in secondary lymphoid organs such as lymph nodes and spleen. Thus, CD4^+^CD25^+^cells recirculate through these organs less than do CD4^+^CD25^-^cells. In the liver, more radioactivity was recovered in recipients of CD4^+^CD25^+^cells as compared with CD4^+^CD25^-^cells. Importantly, CD4^+^CD25^+^cells also had a significantly better capacity to enter the inflamed joint than did CD4^+^CD25^-^cells (Fig. [7a](#F7){ref-type="fig"}). The level of radioactivity detected in the arthritic joints was low but similar to levels found in transfer experiments with effector T cells \[[@B18]\]. As a control, some mice were injected with CFSE-labelled cells. FACS analysis of the secondary lymphoid organs revealed the presence of viable cells 24 hours after transfer and excluded the possibility that the difference in migration pattern is due to leakage of radioactivity (Fig. [7b](#F7){ref-type="fig"}). The migration behaviour of CD4^+^CD25^+^T~reg~cells does reflect their more activated phenotype, and their ability to enter inflamed joints makes it possible that they act directly at the site of inflammation.
Discussion
==========
Our findings provide clear evidence that CD4^+^CD25^+^T~reg~cells are critical for regulating the severity of AIA in mice. We showed this by manipulating the T~reg~cell numbers using two different approaches: depletion of CD25-expressing cells and transfer of purified CD4^+^CD25^+^T~reg~cells. It is important to stress that we depleted CD25-expressing cells in the interval between immunization and AIA induction, because CD25-depletion before immunization profoundly increases the resulting humoral and cellular immune responses \[[@B3],[@B12]\]. These data are consistent with studies conducted in collagen-induced arthritis; however, in these experiments CD25-expressing cells were depleted before immunization with collagen type II, and the resulting more severe arthritis could be interpreted as the result of stronger immunization state \[[@B12]\]. With our experimental design, we were able to examine the effect of T~reg~cells in ongoing joint inflammation directly. Because CD25 is also expressed on activated conventional T cells, it could be assumed that injection of an anti-CD25 antibody would deplete not only T~reg~cells but also effector T cells, but the exacerbated AIA in CD25-depleted mice argues against such a depletion of effector T cells. Accordingly, in control experiments lymph node cells from CD25-depleted mice isolated at the time of induction of AIA were able to mount a similar anti-mBSA response *in vitro*as compared with control mice (data not shown). Furthermore, CD4^+^CD25^+^cells isolated from immunized donors can suppress development of AIA (Fig. [4b](#F4){ref-type="fig"}). Taken together, these data imply that the CD25^+^compartment in immunized mice largely consists of T~reg~cells.
AIA induction in CD25-depleted mice resulted in a much more severe arthritis in the acute and chronic stages of disease. We recently showed, with the use of a depleting anti-CD4 antibody, that this acute stage of AIA is already under the control of T cells \[[@B15]\]. Nevertheless, early AIA is dominated by cells of the innate immune system \[[@B19]\], and the exacerbation of arthritis in CD25-depleted mice could be due to a lack of suppression of these cells by T~reg~cells. In accordance with this view, CD4^+^CD25^+^T~reg~cells are able to suppress innate immune cells in a model of bacteria-induced colitis \[[@B20]\].
In later stages exacerbated arthritis in CD25-depleted mice is accompanied by increased mBSA-specific proliferation and IgG production. This enhanced responsiveness emerged during arthritis development and is due to sustained T cell activation. Such prolonged T cell activation in the absence of CD4^+^CD25^+^cells has also been described in other disease models \[[@B21]\] and is probably the cause of the increased AIA severity. Moreover, the PC61 antibody used in our study has a half-life of approximately 3 weeks *in vivo*(Sutmuller R, personal communication), which makes it possible that T~reg~cell function is not only impaired by depletion but also by blockade of IL-2 binding to CD25 by the PC61 antibody. IL-2 or IL-2 signalling via CD25 has been shown to be critical to the regulatory action of T~reg~cells \[[@B22],[@B23]\]. Also, activation-induced cell death of pathogenic T cells, which is regulated by IL-2, could be impaired by withdrawal of IL-2 signalling and therefore contribute to the observed high levels of cellular immune responses in our study \[[@B24]\].
The fact that depletion of CD4^+^CD25^+^T~reg~cells enhances the immune response against the foreign antigen mBSA clearly demonstrates that their suppressive effect is not strictly limited to autoreactive T cells. Taking into consideration that T~reg~cells are also critically involved in the control of immune responses against pathogens \[[@B25],[@B26]\], their physiological function is not just to prevent autoimmunity but also to control the extent of inflammatory reactions in order to prevent tissue damage to the host. Further support for the influence of CD4^+^CD25^+^T~reg~cells on arthritis development came from the transfer experiments. When transferred at the time of induction of AIA, CD4^+^CD25^+^cells were able to ameliorate ongoing disease. Analysis of the recipients did not reveal a remarkable long-lasting suppression of systemic mBSA-specific immune reactions. Thus, prevention of AIA appears to be possible without inducing anergy or abrogating previously induced T-cell effector functions \[[@B27]\]. In contrast to this, transferred CD4^+^CD25^-^cells significantly enhance cell-mediated and humoral immune responses.
Furthermore, the homing data presented here demonstrate that CD4^+^CD25^+^cells can migrate into the arthritic knee joint. Functional T~reg~cells have repeatedly been found within such effector sites and/or draining lymph nodes, for instance in tolerated allografts \[[@B28]\], in Langerhans islets and pancreatic lymph nodes in inflammation-induced diabetes \[[@B29]\], in chronically inflamed skin in a *Leishmania*infection model \[[@B30]\], and in the mucosa and mesenteric lymph nodes in inflammatory colitis in severe combined immunodeficient (SCID) mice \[[@B31]\]. Interestingly, two recent papers \[[@B32],[@B33]\] reported an accumulation of functional T~reg~cells in the inflamed joints of patients with RA, juvenile arthritis and other rheumatic diseases.
It is most likely that the transferred CD4^+^CD25^+^T~reg~cells act in the draining lymph node as well as in the inflamed tissue. Within such a scenario, it could be possible that T~reg~cells inhibit the activation of effector T cells and their subsequent migration to the joints. Such a mechanism was recently speculated in modulation of virally induced immunopathology by T cells \[[@B26]\].
Huehn and colleagues \[[@B11]\] recently demonstrated that CD4^+^CD25^+^T~reg~cells can be divided into subsets based on the expression of the integrin α~E~β~7~. Moreover, this marker identifies CD25^-^T~reg~cells \[[@B34]\]. Both α~E~β~7~-expressing subsets had better capacity to reach the inflamed joint and to prevent arthritis in the AIA model, as compared with α~E~β~7~^-^T~reg~cells \[[@B10]\]. Thus, suppression at the site of inflammation is also an important part of the activity of T~reg~cells. How this effect is mediated is unclear but an involvement of IL-10 or transforming growth factor-β is possible \[[@B20],[@B35],[@B36]\].
If these hypotheses are correct, then they could explain why the transfer of T~reg~cells after arthritis induction is not effective. On the one hand, transfer of T~reg~cells 24 hours after intra-articular antigen challenge might be too late to inhibit activation of effector T cells and their migration to the joint. Indeed, T-cell activation is an early event in AIA because CD4^+^T cell depletion ameliorates the acute stage of the model \[[@B15]\]. On the other hand, it could be possible that the suppressive function of regulatory T cells is switched off under the inflammatory conditions present in the inflamed tissue by factors such as IL-6 or glucocorticoid-induced tumor necrosis factor family-related gene (GITR) and GITR-ligand interactions, abrogating the suppressive effect of T~reg~cells \[[@B37]\]. With this in mind, it could be interesting to investigate whether the accumulated T~reg~cells in patients with arthritis function properly *in vivo*and whether these patients could really benefit from a therapeutic enhancement of T~reg~function, as suggested by some enthusiastic investigators in this field.
In this regard, data on the curative effects of T~reg~cells in experimental disease models are conflicting. To best of our knowledge, a curative effect of CD4^+^CD25^+^T~reg~cells has only been demonstrated in the colitis model induced by transfer of CD45RB^high^T cells into SCID mice \[[@B31],[@B38]\]. In contrast, other authors were unable to demonstrate such an inhibitory effect of T~reg~cells on SCID colitis when they were transferred 1 week after administration of pathogenic CD45RB^high^T cells \[[@B39]\]. Because arthritis in the AIA model has a hyperacute onset, it could be assumed that the time window for an ameliorative effect of T~reg~cell transfer ends very shortly after intra-articular injection of antigen. However, further studies on the role of T~reg~cells in other arthritis models are clearly needed to clarify whether enhancement in T~reg~cell function might be beneficial in experimental arthritis and perhaps in human disease.
Conclusion
==========
Our data show that T~reg~cells are critically involved in the control of immune responses that are responsible for the pathogenesis of chronic arthritis. Transfer of such cells can modulate the severity of ongoing inflammatory arthritis but they cannot suppress established disease. Thus, timing of T~reg~cell transfer for therapeutic purposes is of considerable importance.
Abbreviations
=============
AIA = antigen-induced arthritis; CFSE = 5,6-carboxyfluorescein diacetate succinimidyl ester; DTH = delayed-type hypersensitivity; ELISA = enzyme-linked immunosorbent assay; FACS = fluorescence-activated cell sorting; FCS = fetal calf serum; IFN = interferon; IL = interleukin; mBSA = methylated bovine serum albumin; PBS = phosphate-buffered saline; RA = rheumatoid arthritis; SCID = severe combined immunodeficient; T~reg~= regulatory T cell.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
OF purified the anti-CD25 hybridoma and purified the monoclonal antibodies from the supernatant; planned and conducted all animal experiments, including ELISA and ELISPOT analysis; and drafted the manuscript. PKP and MG scored the histological changes in arthritic joints. KS, JH and AH conducted the migration experiments, as well as the α~E~β~7~transfer experiments. RB supervised the project and participated together with AS and AR in the design of the study and its coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.
Acknowledgments
===============
We thank T Kaiser and K Raba for FACS sorting; M Schinz and A Kaufmann for help with ELISPOT; H Börner, C Hüttich and R Stöckigt for their excellent technical assistance; and KW Pratt and D Szczawinska for critical comments on the manuscript. This work was supported by the Kompetenznetz Rheuma (Grant 01 GI 0344), Deutsche Forschungsgemeinschaft (Grant Br 1372/5-1) and the Interdisciplinary Center for Clinical Research (IZKF) Jena.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Depletion of CD25-expressing cells by anti-CD25 treatment. Mice immunized with methylated bovine serum albumin (mBSA) were injected intraperitoneally with 0.5 mg PC61 (anti-CD25) or rat IgG as control 4 and 2 days before arthritis induction. Representative example for flow-cytometric assessment of depletion in spleen cells, using a non-cross-reactive anti-CD25 antibody (7D4) at the time of arthritis induction (day 0).
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Clinical and histological severity of antigen-induced arthritis (AIA) in CD25-depleted mice. **(a)**Knee joint swelling (difference in mediolateral joint diameters of arthritic minus nonarthritic knee joints) during the time course of arthritis was higher in CD25-depleted mice. **(b)**Haematoxylin and eosin stained frontal knee joint sections were scored on a 0--3 point scale at day 14 of AIA for each of the following: severity of synovial hyperplasia, cellular infiltration, cartilage destruction and pannus formation. A score for inflammatory changes (Inf) was calculated by adding the points for synovial hyperplasia and infiltration, and for joint destruction (Dest) by adding the points for cartilage damage and pannus formation. Total arthritis score (Score) was calculated by adding scores for inflammatory changes and joint destruction, giving a maximal AIA score of 12 points. Representative photomicrographs of **(c)**a control (rat IgG-injected) and **(d)**a knee joint from an anti-CD25-treated mouse. Ten animals were included in each group in two independent experiments. \**P*\< 0.05, \*\**P*\< 0.01, \*\*\**P*\< 0.001, versus control.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Analysis of *in vivo*and *ex vivo*immune responses in CD25-depleted mice. **(a)***In vivo*delayed-type hypersensitivity (DTH) response against methylated bovine serum albumin (mBSA) as a marker for cellular immune response was measured as the increase in ear thickness after intradermal antigen challenge on day 7 of antigen-induced arthritis (AIA). **(b)**Proliferation, measured as \[^3^H\]thymidine incorporation of unstimulated (unst) or mBSA-stimulated (mBSA) draining lymph node cells at day 14 of AIA. **(c)**Cytokine production was measured with ELISPOT. **(d)**Serum levels of IgG against mBSA, collagen type I, collagen type II and cartilage proteoglycans were measured using ELISA after 14 days of AIA. Proliferation, DTH reaction and serum IgG titres were tested in 10 animals per group; cytokine production was measured in six animals per group. Data are from one out of two similar experiments. \**P*\< 0.05, \*\**P*\< 0.01, \*\*\**P*\< 0.001, versus control.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Modulation of antigen-induced arthritis (AIA) by transfer of regulatory T cells (T~reg~cells). Amelioration of clinical and histological severity of AIA by transfer of 2 × 10^6^CD4^+^CD25^+^cells freshly isolated from **(a)**naive or **(b)**immunized mice at the time of AIA induction (day 0; *n*= 6 per group). **(c)**Transfer of 1 × 10^6^*in vitro*pre-activated cells at the time of AIA induction (*n*= 6). ^\#^*P*\< 0.05, ^\#\#^*P*\< 0.01 for CD4^+^CD25^+^versus CD4^+^CD25^-^; ^+^*P*\< 0.05, ^++^*P*\< 0.01 for CD4^+^CD25^+^versus phosphate-buffered saline. Dest, joint destruction; Inf, inflammatory changes; Score; total arthritis score
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Transfer of regulatory T cells (T~reg~cells) cannot cure established arthritis. Pre-activated CD4^+^CD25^+^cells (1 × 10^6^) were transferred on **(a)**day 1 or **(b)**day 7 of antigen-induced arthritis (AIA). Arthritis severity was monitored by measurement of knee joint swelling and by histological assessment 14 days after cell transfer (*n*= 6--7 per group). **(c)**Also, 1 × 10^6^pre-activated α~E~β~7~-expressing T~reg~cells have no curative effect in AIA (*n*= 8 per group).
:::
:::
::: {#F6 .fig}
Figure 6
::: {.caption}
######
There is no suppression of cellular or humoral methylated bovine serum albumin (mBSA)-specific immunity with transfer of T~reg~cells. Pre-activated CD4^+^CD25^+^cells (1 × 10^6^) were transferred at the time of antigen-induced arthritis (AIA) induction. **(a)**Delayed-type hypersensitivity (DTH) reactivity against mBSA *in vivo*was tested 7 days later by an intradermal antigen challenge into the ears. **(b)**Antigen-specific proliferation (\[^3^H\]thymidine incorporation) and **(c)**cytokine production (ELISPOT) of draining lymph node cells was measured 14 days after AIA induction. **(d)**Serum levels of IgG against mBSA, collagen type I, collagen type II and cartilage proteoglycans were measured with ELISA after 14 days of AIA. Proliferation, DTH reaction, cytokine production, and serum IgG titres were tested in six animals per group. ^\#^*P*\< 0.05 for CD4^+^CD25^+^versus CD4^+^CD25^-^; \**P*\< 0.05 for CD4^+^CD25^-^versus phosphate-buffered saline.
:::
:::
::: {#F7 .fig}
Figure 7
::: {.caption}
######
Migration behaviour of regulatory T cells (T~reg~cells). **(a)**CD4^+^CD25^+^and CD4^+^CD25^-^cells were purified by fluorescence-activated cell sorting (FACS) and labelled with ^111^In. Cells (1 × 10^6^) were injected intravenously into antigen-induced arthritis (AIA) mice at day 7. After 24 hours radioactivity in isolated organs and the rest of the body was determined using a γ-counter. Thereafter, the total radioactivity recovered per animal was calculated by adding the counts of the organs and the rest of the body. **(a)**The proportion of radioactivity found in the isolated organs is shown here as a percentage of total recovered radioactivity (*n*= 6; mean ± standard error of the mean; one representative out of two independent experiments; \*\**P*\< 0.01). **(b)**FACS-purified cells were labelled with 5,6-carboxyfluorescein diacetate succinimidyl ester (CFSE) and injected intravenously. After 24 hours single-cell suspensions from draining lymph node (dLN), nondraining peripheral lymph node (pLN), mesenteric lymph node (mLN), spleen and peripheral blood lymphocytes (PBL) were analyzed by FACS. The percentage of CFSE^+^cells of the total CD4^+^cells was measured. Histogram plots are gated on CD4^+^cells after propidium--iodide exclusion of dead cells (*n*= 3 per group). Higher numbers of CFSE^high^cells are found in the secondary lymphoid organs in the recipients of CD4^+^CD25^-^cells.
:::
:::
|
PubMed Central
|
2024-06-05T03:55:54.944405
|
2005-1-11
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065322/",
"journal": "Arthritis Res Ther. 2005 Jan 11; 7(2):R291-R301",
"authors": [
{
"first": "Oliver",
"last": "Frey"
},
{
"first": "Peter K",
"last": "Petrow"
},
{
"first": "Mieczyslaw",
"last": "Gajda"
},
{
"first": "Kerstin",
"last": "Siegmund"
},
{
"first": "Jochen",
"last": "Huehn"
},
{
"first": "Alexander",
"last": "Scheffold"
},
{
"first": "Alf",
"last": "Hamann"
},
{
"first": "Andreas",
"last": "Radbruch"
},
{
"first": "Rolf",
"last": "Bräuer"
}
]
}
|
PMC1065323
|
Introduction
============
Migration and accumulation of memory T cells in the synovium is a critical step in the pathogenesis of chronic arthritides \[[@B1]-[@B3]\]. Chemokines are a large family of small secreted proteins (8--15 kDa) that control lymphocyte trafficking in physiological and pathological processes. The evaluation of type and distribution of chemokines and their receptors in the synovium is therefore crucial to an understanding of the mechanisms of synovial T cell recruitment.
From a functional point of view, chemokines can be broadly classified into two groups: inflammatory and homeostatic \[[@B4]\]. The inflammatory chemokines are induced by proinflammatory stimuli and control the migration of leukocytes to the site of inflammation. CCR5 and CXCR3 are classical examples of receptors for inflammatory chemokines \[[@B5]\]. The homeostatic chemokines regulate the basal traffic of lymphocytes and other leukocytes through peripheral lymphoid tissues. CCR7 is an example of a receptor for homeostatic chemokines. CCR7 and its ligands (CCL19 and CCL21) have also been shown to have a pivotal role in the development and maintenance of secondary lymphoid organ microarchitecture \[[@B4],[@B5]\]. Recently, the CCR7 chemokine receptor has been identified as an important marker of memory T cell differentiation. It has been proposed that CCR7^+^memory T cells represent a pool of \'central\' memory T cells homing to lymph nodes, where they undergo further differentiation into CCR7^-^memory T cells, which migrate to the peripheral tissues to perform their effector functions \[[@B6]\].
However, this model has been disputed by other investigators \[[@B7],[@B8]\] and CCR7^+^naive and memory T lymphocytes have been detected in both normal and inflamed human tissues \[[@B9]\]. Previous studies have shown that Th1-polarised \[[@B10],[@B11]\], CCR5^+^and CXCR3^+^lymphocytes are enriched in synovial inflammatory infiltrates and in synovial fluid (SF) lymphocytes from patients with adult rheumatoid arthritis (RA) \[[@B12],[@B13]\] and juvenile idiopathic arthritis (JIA) \[[@B14]-[@B16]\]. CCR5 and CXCR3 ligands, namely RANTES (or CCL5) and macrophage inhibitory protein-1α (MIP-1α, or CCL3), and interferon-inducible protein-10 (IP-10, or CXCL10) and ITA-C (CXCL11), respectively, have also been detected in rheumatoid synovium \[[@B17]\].
Limited information is available on CCR7 expression in synovial lymphocytes from patients with chronic arthritis. Naive CD45RA^+^T cells with a CCR7 phenotype have been found to infiltrate the synovial tissue in patients with RA \[[@B16]\]. The CCR7 ligands CCL19 and CCL21 have been detected in endothelial cells and in the perivascular infiltrate in RA synovium, suggesting their potential involvement in lymphoid neogenesis that occurs in inflamed synovial tissue \[[@B18]-[@B20]\].
No information is so far available on the expression of CCR7 in memory T cells homing to the synovial microenvironment in relation to expression of the inflammatory chemokine receptors CCR5 and CXCR3.
In this study we therefore investigated the expression of CCR7, CCR5 and CXCR3 on SF and peripheral blood (PB) memory CD4^+^T cells from patients with JIA, chemotaxis of the latter cells to the ligands of these receptors, and the distribution of cells positive for CCR7, CCR5 and CXCR3 in the inflamed synovium.
Methods
=======
Patients
--------
Immunophenotypic and functional characterisation of freshly isolated PB and/or SF lymphocytes was performed in a total of 25 patients with JIA (14 female, 9 male) undergoing therapeutic arthrocentesis. According to ILAR Durban classification criteria \[[@B21]\], 15 patients had persistent oligoarticular JIA (pOJIA), 6 had extended oligoarticular JIA (eOJIA) (which means a total of five or more joints involved after the first 6 months of disease and therefore a polyarticular course) and 4 had rheumatoid factor (RF)-negative polyarticular JIA. Several clinical (number of active joints, number of joints with limited range of motion, and physician global assessment of overall disease activity) and laboratory parameters (erythrocyte sedimentation rate, C-reactive protein, white blood cell and platelet counts, and hemoglobin serum concentration) of disease activity were recorded, together with the ongoing treatment, at the time of the study.
Paired serum and SF samples from 28 additional patients with JIA (16 with pOJIA, 6 with eOJIA, 4 with RF negative polyarticular JIA, and 2 with systemic JIA) were tested for CCL21 concentrations. The clinical characteristics of patients with JIA and the ongoing treatment at the time of the study are reported in Tables [1](#T1){ref-type="table"} and [2](#T2){ref-type="table"}. For each patient, SF was collected at the time of intra-articular steroid injection. Paired serum sample was obtained, with permission, on the occasion of concomitant routine venipuncture. Both SF and sera were stored at -80°C immediately after centrifugation. A previous steroid injection into the same joint in the previous 6 months was considered to be an exclusion criterion.
Peripheral blood and/or sera from 15 age-matched healthy subjects attending our clinic for routinary pre-operative examinations for minor surgical procedures were used as controls. Synovial tissue from six patients (two with pOJIA, one with eOJIA and three with RF-negative polyarticular JIA) was obtained, with permission, at the time of synoviectomy.
Samples were taken from patients and healthy controls, and stored after parental permission in accordance with the informed consent approved by the ethical committee of the \'G. Gaslini\' Institute.
Cell preparation and flow cytometry
-----------------------------------
PB and SF mononuclear cells (MNC) were isolated from heparinised blood and SF samples by Ficoll--Hypaque (Sigma, St Louis, MO, USA) density gradient centrifugation. Cells were washed, resuspended in complete medium (RPMI 1640 with L-glutamine, penicillin/streptomycin, nonessential amino acids and 10% fetal bovine serum; Sigma) and depleted of adherent cells by adherence to plastic for 1 h at 37°C in 5% CO~2~. To analyse the expression of CCR7 on CD4^+^memory T cells in SF and PB MNC, cells were triple-stained with CD45RO-TC (Caltag, Burlingame, CA, USA), CD4--FITC (BD Biosciences, San Jose, CA, USA) and anti-CCR7--PE (BD Pharmingen, San Diego, CA, USA) monoclonal antibodies (mAbs) and analysed by flow cytometry (CellQuest software and FACScan; BD Biosciences). CCR7 expression was evaluated by gating on the CD45RO^+^CD4^+^lymphocyte population. CD45RO^+^cells were purified from PB and SF MNC by negative selection with a CD45RA mAb (Caltag) and goat anti-mouse IgG-coated magnetic beads (Immunotech, Marseille, France), in accordance with the manufacturer\'s instructions. Recovered cells were 95% enriched for CD45RO^+^cells.
CCR5 or CXCR3 expression was investigated by three-colour staining of freshly isolated SF and PB CD45RO^+^cells with fluorescein isothiocyanate (FITC)-conjugated CD4 (BD Biosciences), anti-CCR7--phycoerythrin (PE) and anti-CCR5--CyChrome mAbs (BD Pharmingen) or CD4-TC (where TC stands for Tri-color), anti-CCR7--PE and anti-CXCR3--FITC (R&D System, Minneapolis, MN, USA), respectively, gating on the CD4^+^CCR7^+^and CD4^+^CCR7^-^lymphocyte populations.
For interferon (IFN)-γ intracellular staining, freshly purified SF CD45RO^+^cells (10^6^) were incubated for 5 hours in the presence of phorbol 12-myristate 13-acetate (20 ng/ml; Sigma), the calcium ionophore A-23187 (250 ng/ml; Sigma) and brefeldin-A (5 μg/ml; Sigma). Cells were washed in phosphate-buffered saline (PBS) with 1% fetal calf serum (staining buffer) and surface stained with CD4--TC (Caltag) and anti-CCR7--PE (BD Pharmingen) mAbs for 30 min at 4°C in the dark. Cells were washed in staining buffer and fixed in 4% paraformaldehyde for 20 min at 4°C in the dark. Afterwards, the cells were washed twice with permeabilisation buffer (PBS containing 1% fetal calf serum and 0.1% saponin \[Sigma\]) and stained with FITC-conjugated mAbs against human IFN-γ (Caltag) for 30 min at 4°C in the dark. Cells were then washed in staining buffer and analysed by flow cytometry, gating on the CD4^+^CCR7^+^and CD4^+^CCR7^-^lymphocyte subsets. Although stimulation with phorbol 12-myristate 13-acetate and calcium ionophore downregulates the intensity of CD4 and CCR7 expression, the proportion of cells positive for each marker was similar before and after stimulation.
Isotype matched, PE-, FITC-, TC- and CyChrome-conjugated mAbs of irrelevant specificity were tested as negative controls in all of the above experiments. The results of flow cytometry experiments were expressed as percentage positive cells or as mean fluorescence intensity; that is, the staining intensity of a test mAb minus that of an isotype-matched, irrelevant control mAb. The threshold for calculating the percentage positive cells was based on the maximum staining obtained with irrelevant isotype-matched mAb, used at the same concentration as the test mAb. Negative cells were defined such that less than 1% of cells stained positive with control mAbs. Cells labelled with test antibody that were brighter than those stained with isotypic control antibody were defined as positive. Mean fluorescence intensities of the isotype control and of test mAbs were used to evaluate whether the differences between the peaks of cells were statistically significant with respect to the control. The Kolmogorov--Smirnov test for the analysis of histograms was used, in accordance with the CellQuest software user\'s guide. Differences between paired PB and SF MNC of patients with JIA on the one hand, and PB MNC of healthy controls on the other, were evaluated by the Kruskal--Wallis analysis of variance (ANOVA) test and the Wilcoxon rank test.
Chemotactic assays
------------------
Migration assays were performed in 24 transwell plates (pore size 5 μm, polycarbonate membrane; Costar, Cambridge, MA, USA). Freshly purified SF CD45RO^+^cells (5 × 10^5^) were dispensed in the upper chamber in 100 μl, and 600 μl of different chemokines at 100 ng/ml (R&D System) or medium alone was added to the lower chamber. Migration was performed in migration medium (RPMI 1640, 0.1% bovine serum albumin; Sigma). Plates were incubated for 2 hours at 37°C. After removal of the transwell inserts, cells from the lower compartments were collected. Furthermore, 0.5 ml of 5 mM EDTA was added to the lower chamber for 15 min at 37°C to detach adherent cells from the bottom of the wells. Detached cells were pooled with the previously collected cell suspensions and counted by staining with trypan blue. To evaluate the percentage of migrated CD4^+^lymphocytes, cell suspensions were double-stained with CD4--PE and CD3--FITC mAbs (BD Biosciences) before and after migration and analysed by flow cytometry. The percentage input was calculated as follows: 100 × (cells migrated to chemokine/total cell number). Differences between cells that migrated to a given chemokine and the same cells that migrated in medium alone were calculated with non-parametric Wilcoxon rank test.
CCL21 serum and SF concentrations
---------------------------------
Forty-three sera (15 from controls) and 28 SFs were tested for CCL21 by an enzyme-linked immunosorbent assay kit from R&D System (Minneapolis, USA), in accordance with the instructions of the manufacturer.
Serum levels of CCL21 were compared in three groups of patients (12 patients with JIA with a polyarticular course, 16 patients with JIA with an oligoarticular course and 15 healthy controls) with the use of the non-parametric Kruskal--Wallis ANOVA test. Correlations between all the variables considered were evaluated with the non-parametric Spearman rank test. Differences between paired serum and SF chemokine concentrations were evaluated by the Wilcoxon rank test.
Immunohistochemical studies
---------------------------
Tissue specimens with sizes between 5 and 12 mm were treated for single and double immunohistochemical stainings with a standard technique as reported previously \[[@B22]\]. In brief, all specimens were fixed in 4% formalin for 24 hours, then dehydrated and embedded in paraffin. Sections 4 μm thick were layered on polylysine-coated slides. Slides were deparaffinised in xylene, and rehydrated in a descending sequence of ethanol concentrations (100--70%).
Three different immunohistochemical techniques, namely alkaline phosphatase--anti-alkaline phosphatase (APAAP) for CCR7, avidin--biotin complex for CD21, and indirect immunoperoxidase (CD3, CD4, CD45RO, CD20, CCR5, CXCR3, CCL19 and CCL21), were performed after 30 min of warming in an oven in citrate buffer, pH 6, with subsequent inhibition of endogenous peroxidase. For single staining, tissue sections were incubated overnight at 4°C with the anti-CCR7 murine mAb, clone 2H4 (Pharmingen). Incubation of tissue sections with anti-CCL21 goat antiserum (R&D), anti-CCR5, clone 2D7 (Pharmingen), anti-CXCR3, clone 1C6 (Pharmingen), anti-CD3 (Dako, Glostrup, Denmark), anti-CD4, clone 4B12 (Neomarkers, Fremont, CA, USA), anti-CD20, clone L26 (Dako) and anti-CD45RO, clone UCHL1 (Menarini, Firenze, Italy) and anti-CD31 clone JC70A (Dako) was performed overnight at 4°C.
Sections were subsequently reacted for 30 min at room temperature (20--25°C) with (1) anti-mouse Ig antibody conjugated to peroxidase-labelled dextran polymer (EnVision; Dako) for CD3, CD45RO, CCR5 and CCR7 stainings, (2) anti-goat secondary biotinylated antibody, followed by high-sensitivity streptavidin--horseradish peroxidase conjugate for CCL21 determination (Cell and Tissue Staining kit; R&D), and (3) APAAP-conjugated rabbit anti-mouse Ig (1:25 dilution; Dako) antibody for CCR7 determination. The chromogenic diaminobenzidine substrate (Dako) was applied for 10 min. All washings were performed by incubating the sections in PBS. For CCR7 determination the alkaline phosphatase reaction was performed with a medium containing Tris-HCl buffer pH 8.2, naphthol AS-TR salt (Sigma) and levamisole (Sigma), for 20 min at 98°C. Slides were counterstained with Mayer\'s haematoxylin. For double CCR7/CCL21 staining, the sections were subjected to peroxidase reaction with goat CCL21 and were washed three times in Tris-buffered saline. Subsequently, the APAAP technique (see above) was applied with the mouse CCR7 Ab (at room temperature, for 3 hours). The secondary reagents were applied for 30 min each. For CCR7 and CCL21, a reactive lymph node from a 10-year-old boy was considered as positive control. Reactions in the absence of primary antibody and with irrelevant antibodies of the same isotypes (anti-cytomegalovirus, clones DDG9 and CCH2; Dako) were performed as negative controls.
Slides were evaluated on two different occasions by two blinded observers (MG and AG) and an expert pathologist (CG). Each specimen was evaluated for the pattern of lymphocyte infiltration in three different categories: (1) aggregates of T cells (CD4) and B cells (CD20) with germinal centre (GC)-like reaction (presence of CD21-positive cells), (2) aggregates of T and B cells without GC-like reaction, and (3) diffuse lymphocytic infiltrate without lymphoid organisation \[[@B20]\]. For each sample a semiquantitative score for the overall degree of T lymphocyte infiltration (CD3) was used (range 0--3). For the assessment of chemokine receptor expression each sample was subjected to microscopical analysis of: (1) the lining layer and sublining zone of the superficial subintima \[[@B23]\]; (2) perivascular infiltrates of the sublining layer without lymphoid organisation; (3) aggregates of T and B cells. Because CCR7, CXCR3 and CCR5 can be expressed by several cell types (lymphocytes, dendritic cells, B cells and plasma cells) \[[@B18],[@B24]\], only areas characterised by a clear lymphocyte infiltration (as defined by anti-CD3 and anti-CD4 positivity) were taken into consideration. The following semiquantitative global score was based on a visual inspection of four different high-power fields (40×) at each level: absent (-, no positive cells per high-power field), weakly positive (+, 1--10 positive cells per high-power field), moderately positive (++, 10--20 positive cells per high-power field), and strongly positive (+++, more than 20 positive cells per high-power field). The assignment of each sample to one of the above categories was based on the predominant pattern observed. Minor differences between the observers were resolved by mutual agreement. Intra observer and interobserver variability was less than 5%.
Results
=======
Phenotypic and functional characterisation of CCR7^+^and CCR7^-^CD4^+^memory T cells isolated from SF
-----------------------------------------------------------------------------------------------------
Expression of CCR7 on CD4^+^memory T cells from the PB and SF of 10 patients with JIA was investigated by three-colour immunofluorescence analysis and compared with that detected on the same PB cell subset from eight age-matched healthy controls.
The heterogeneity test between the three subgroups was highly significant (Kruskal--Wallis ANOVA test, *P*= 0.0001). At *post hoc*analysis, in the PB from patients with JIA, the percentage of CCR7^+^cells in the CD4^+^CD45RO^+^subpopulation (median 65.5%, range 50--90%) was significantly lower than in PB from controls (median 76%, range 73--89%, *P*= 0.03, Mann--Whitney *U*-test). A further decrease in CCR7^+^cells was observed in memory CD4^+^cells isolated from SF (median 41.2%, range 12--59%) in comparison with paired PB.
Thus, even a variable proportion of SF memory CD4^+^T cells are positive for CCR7; this subpopulation is clearly enriched in CCR7^-^cells in comparison with paired PB.
Next we investigated the expression of CCR5, CXCR3 and IFN-γ in SF CD4^+^CD45RO^+^CCR7^+^and CCR7^-^cells from 10 consecutive patients and compared it with that detected in paired PB from 5 of these patients and in the PB of 5 healthy controls. To this end, purified CD45RO^+^cells were stained with anti-CD4, anti-CCR7 and respectively, anti-CCR5, CXCR3 or IFN-γ mAbs in three-colour immunofluorescence.
In SF, CCR5^+^cells were found to be enriched in the CD4^+^CD45RO^+^CCR7^-^lymphocyte subset (median 85%, range 74--99%) as compared to the CD4^+^CD45RO^+^CCR7^+^cell fraction (median 65%, range 46--84%, *P*= 0.005; Wilcoxon test; not shown). These data were in line with our previous observation of a higher expression of CCR7 in \'early\' CD27^+^memory T cells and a prevalent CCR7^-^CCR5^+^phenotype in \'effector\' CD27^-^T cells in SF from patients with JIA \[[@B25]\]. The median percentage of CD4^+^, CD45RO^+^, IFN-γ positive cells was 27% (range 21--46%) for CCR7^+^cells and 40% (range 24--69%) for CCR7^-^cells (*P*= 0.005) (Fig. [1a,c](#F1){ref-type="fig"}), as assessed by intracellular staining. Accordingly, the mean fluorescence intensity for IFN-γ was lower for CCR7^+^cells (median 136, range 84--184) than for paired CCR7^-^CD4^+^memory T cells (median 190, range 127--307, *P*= 0.005).
CXCR3 was highly expressed on both CCR7^+^and CCR7^-^subsets of SF memory CD4^+^T cells. However, in all patients with JIA, SF CCR7^+^memory CD4^+^T cells showed a higher expression of CXCR3 (median 86%, range 74--93%) than the CCR7^-^counterpart (median 76%, range 62--85, *P*= 0.005) (Fig. [1b,d](#F1){ref-type="fig"}). In comparison with SF, PB of patients with JIA showed a lower expression of CXCR3 in both CCR7^+^(median 38.5%, range 24--55%) and CCR7^-^(median 27%, range 17--40%) CD4^+^memory T cells. A similar expression was also found in circulating CCR7^+^(median 40%, range 32--55%) and CCR7^-^(median 17%, range 13--45%) CD4^+^memory T cells from age-matched healthy controls.
Taken together, these results show that the CD4^+^CD45RO^+^CCR7^-^subpopulation is enriched in \'effector\' CCR5 and IFN-γ expressing cells, whereas the CD4^+^CD45RO^+^CCR7^+^subpopulation shows a lower expression of CCR5 and IFN-γ and a higher degree of coexpression with CXCR3.
Different localisation of CCR7, CXCR3 and CCR5 positive cells in synovial tissue
--------------------------------------------------------------------------------
We next addressed the following questions: (1) is CCR7 expressed in synovial tissue, (2) how does its expression correlate with the pattern of lymphocytic infiltration, and (3) how is CCR7 expression related to that of CXCR3 and CCR5, two Th1-associated chemokine receptors? To this end, synovial tissues obtained at synoviectomy from six patients with JIA were analysed for the expression of CCR7, CXCR3 and CCR5 in areas characterised by a clear lymphocyte infiltration (Table [3](#T3){ref-type="table"}).
Different patterns in the synovial inflammatory infiltrate were observed in the individual patients (Table [3](#T3){ref-type="table"}). One patient (no. 6) showed T and B cell aggregates with the presence of a GC reaction, as demonstrated by the presence of CD21^+^follicular dendritic cells \[[@B20]\]. In three patients (nos 3, 4 and 5) clusters of T and B cell aggregates in the absence of follicular dendritic cells were observed \[[@B20]\]. Two patients (nos 1 and 2) displayed diffuse lymphocytic infiltrates as perivascular aggregates in the sublining layer or scattered throughout the synovium up to the lining layer \[[@B20]\].
CCR7-positive cells were detected both in cases showing a diffuse lymphocytic infiltrate (Fig. [2a,b](#F2){ref-type="fig"}) and in those displaying a more organised lymphoid structure (Fig. [2c--e](#F2){ref-type="fig"}). In the former, CCR7 expression was detected mostly in the perivascular lymphocytic infiltrates of the sublining layer (Fig. [2b,o](#F2){ref-type="fig"}) and, only occasionally, in scattered cells in the sublining zone of the superficial subintima (see also Fig. [4c](#F4){ref-type="fig"} below). In the latter, CCR7-positive cells were localised inside and around lymphoid aggregates (Fig. [2e](#F2){ref-type="fig"}).
Because naive CD45RA^+^T cells with a CCR7^+^phenotype had also previously been found to infiltrate the synovial tissue from patients with RA \[[@B19]\], serial sections were stained with CCR7 and CD45RO antibodies. A clear positivity for CCR7 was detected in lymphocytic infiltrates staining heavily for CD45RO (not shown).
CXCR3 was abundantly expressed in all lymphocyte-infiltrated areas examined (Table [3](#T3){ref-type="table"}). In fact, CXCR3-positive cells were detected in lymphoid aggregates (Fig. [2f](#F2){ref-type="fig"}) and in perivascular infiltrates of sublining layer (Table [3](#T3){ref-type="table"}). In many areas, CXCR3 and CCR7 displayed a similar pattern of tissue distribution, especially at the level of lymphocyte aggregates (Fig. [2e,f](#F2){ref-type="fig"}).
Conversely, CCR5-positive cells were detected mainly in the lining layer and in the sublining zone of the superficial subintima and, to a smaller extent, in the perivascular infiltrates of the sublining layer (Fig. [2p](#F2){ref-type="fig"}) and in the T and B cell aggregates (Fig. [2n](#F2){ref-type="fig"}) (Table [3](#T3){ref-type="table"}).
Altogether, even if a certain degree of co-localisation of the two chemokine receptors was found (Table [3](#T3){ref-type="table"}), CCR5 positive cells showed a substantially different tissue distribution from that of CCR7, either in T and B cell aggregates (Fig. [2e,g,l,n](#F2){ref-type="fig"}) or in diffuse lymphocytic infiltrates (Fig. [2o,p](#F2){ref-type="fig"}). Conversely, a variable degree of co-localisation was found for CCR5 and CXCR3 at the level of the sublining and lining layer (Table [3](#T3){ref-type="table"}).
Chemotaxis of SF CD4^+^memory T cells to inflammatory and homeostatic chemokines
--------------------------------------------------------------------------------
In further experiments, chemotaxis of freshly isolated SF memory T cells in response to CCR7, CCR5 and CXCR3 ligands was investigated, and migrated CD4^+^CD45RO^+^T cells were detected by flow cytometry.
Chemotactic assays were performed with SF CD45RO^+^cells isolated from eight patients with JIA (five with pOJIA, three with eOJIA) and tested in the presence or absence of two inflammatory chemokines that bind to CCR5 (CCL3) and CXCR3 (CXCL11), respectively, and of homeostatic chemokines binding to CCR7 (CCL21 and CCL19).
CD4^+^CD45RO^+^T cells migrated significantly to both CCL3 and CXCL11 (*P*= 0.02 for both chemokines). Similar responses were observed when CCL19 was tested (*P*= 0.02). Chemotaxis of CD4^+^memory T cells to CCL21 approached but did not reach statistical significance (*P*= 0.1) (Fig. [3](#F3){ref-type="fig"}). The latter finding might be related to the limited number of the samples tested.
In the patients studied, the variability of chemotaxis of SF CD4^+^memory T cells did not show any significant correlation with disease form, degree of disease activity and treatment at the moment of sampling.
Expression of CCL21 in SF and synovial tissue
---------------------------------------------
To gain further insight into the relevance of the interactions between CCR7 and its ligand CCL21 *in vivo*, sera and SF CCL21 concentrations were tested in 28 consecutive patients with JIA and in 15 healthy controls.
The heterogeneity test between the three subgroups was highly significant (Kruskal--Wallis ANOVA test, *P*= 0.0045). Concentrations of CCL21 were significantly higher in SF (median 1769.5 pg/ml, range 110--25,556 pg/l) than in paired sera from patients with JIA (median 268 pg/ml, range 57.6--5146.9 pg/ml, *P*\< 0.0001; Wilcoxon test; Fig. [4a](#F4){ref-type="fig"}).
A strong correlation was found between paired serum and SF CCL21 concentrations (*r*= 0.91, *P*= 0.001; Spearman\'s test).
No significant difference was observed in CCL21 serum concentrations between patients with JIA with oligoarticular course (median 229.2 pg/ml, range 67--3948 pg/ml), patients with JIA with polyarticular course (median 378 pg/ml, range 65--5146 pg/ml) and age-matched healthy controls (median 282.2 pg/ml, range 76--2349 pg/ml, *P*= 0.3; Kruskal--Wallis ANOVA test). Similarly, no significant difference was found in SF CCL21 concentrations between patients with JIA with an oligoarticular course and patients with a polyarticular course (*P*= 0.52; Mann--Whitney *U*-test).
Finally, no significant correlation was found between CCL21 serum concentrations and several clinical and laboratory parameters of disease activity in patients with JIA (see the Methods section; not shown).
The expression of CCL21 was also analysed in synovial tissues by immunohistochemistry. CCL21 was detected in all specimens. In the samples characterised by lymphoid organisation, CCL21 staining was observed in the perivascular lymphocytic aggregates and in the vascular endothelium within follicular structures, a pattern reminiscent of that observed on staining for CCR7 \[[@B19]\]. A similar pattern was detected in tissues showing a diffuse lymphocytic infiltration (Fig. [4b](#F4){ref-type="fig"}). Moreover, a clear-cut expression of CCL21 was also observed in flat wall vessels of the superficial subintima of the sublining layer (Fig. [4c,d](#F4){ref-type="fig"}) \[[@B23]\].
Discussion
==========
In this study we have investigated the role of CCR7 in the recruitment of CD4^+^memory T cells into the inflamed joints of patients with JIA, and attempted the functional and anatomical dissection of these cells according to their expression of CCR7, CXCR3, CCR5 and IFN-γ. We detected two populations of SF CD4^+^memory T cells: the CD4^+^CD45RO^+^CCR7^-^subset, which was enriched in \'effector\' CCR5 and IFN-γ positive cells, and the CD4^+^CD45RO^+^CCR7^+^subset, which was less well represented and showed higher CXCR3 coexpression. SF CD4^+^memory T cells displayed chemotactic activity to both inflammatory and homeostatic chemokines representing the physiological ligands of these receptors.
Of the three chemokine receptors studied, CXCR3 proved to be the most widely expressed in synovial tissue, with a clear distribution both in lymphoid aggregates and in perivascular infiltrates of sublining layer and in the lining layer.
Conversely, CCR7-positive and CCR5-positive cells in the synovial tissue displayed a different distribution, showing an even higher differentiation in their expression in respect to SF. In fact, CCR7^+^cells were detected in synovial tissues irrespective of the pattern of lymphoid organisation and were localised mainly in lymphoid aggregates and in perivascular infiltrates of the sublining layer. Notably, CCL21, the CCR7 ligand, was found in the SF as well as in perivascular lymphocytic aggregates and in the vascular endothelium of follicular structures.
In contrast, CCR5^+^cells were detected mainly in the lining layer and in the sublining zone of the superficial subintima and, to a smaller extent, in the perivascular infiltrates of sublining layer and in the T and B cell aggregates.
These findings in synovial tissue are in line with the results of the phenotypic characterisation of SF CCR7^+^and CCR7^-^memory CD4^+^T cells performed in the present study and with previous observations showing a variable degree of coexpression of CXCR3 and CCR5 on T cells isolated from inflamed tissues \[[@B14],[@B26],[@B27]\].
To our knowledge, this is the first demonstration of a different anatomical localisation of cells positive for CCR7, CCR5 and CXCR3 infiltrating the inflamed synovium; this finding may have functional implications for the intra-tissue migration of T cells.
During the past decade several studies have focused on the capacity of memory T cells to differentiate in the context of inflamed tissues. Many of these studies used a member of the tumour necrosis factor receptor family, CD27, to distinguish recently activated CD27^+^from \'effector\' CD27^-^memory CD4^+^T cells \[[@B28]\]. Notably, a clear enrichment of the latter subpopulation has been found in SF of patients with RA and JIA \[[@B29],[@B30]\]. In a recent study we showed that CD27^+^memory T cells in SF of patients with JIA expressed CCR7 more highly than CCR5, whereas CD27^-^T cells displayed a prevalent CCR7^-^CCR5^+^phenotype \[[@B25]\]. Notably, the immunohistochemical characterisation of rheumatoid synovial tissue in adult RA has shown a prevalent localisation of CD4^+^CD27^+^T cells in the perivascular lymphocytic aggregates, with a relative increase in CD27^-^T cells in diffuse lymphocytic infiltrates \[[@B31]\].
Thus, it is conceivable that the functional and phenotypic characterisation of CCR7^+^and CCR7^-^memory CD4^+^T cells and the different tissue distribution between CCR7 and CCR5 found in the present study might reflect the same behaviour already observed for CD27^+^and CD27^-^memory T cells, yielding more insight into the migratory properties of memory T cells into and within the synovial tissue.
The partial overlap of CCL21 and CCR7 expression in the inflamed synovium might suggest that the CCR7/CCL21 system, probably in synergy with CXCR3 and its ligands, is involved in the recruitment of memory T cells, as already shown for naive T cells \[[@B19]\]. However, the possibility cannot be ruled out that CCR7 expression in CD4^+^memory T cells isolated from SF was upregulated after the reactivation of these cells at the site of inflammation \[[@B32]\].
CCL21, together with other homeostatic chemokines such as CXCL13, has been shown to have a fundamental function in the development of secondary lymphoid organs by interacting with CCR7 \[[@B4],[@B33]\]. Mice whose CCR7 or CCL21 genes have been knocked out exhibit marked deficiencies in the structural and cellular composition of lymph nodes \[[@B34]\].
A sequence of events similar to that taking place in lymph node organogenesis is supposed to be involved in the development of organised lymphoid structures in inflamed tissues, such as the rheumatoid synovium \[[@B18]-[@B20]\]. Indeed, up to 20% of synovial tissue biopsies from patients with RA show the typical features of the GC reaction. Other patients show aggregates of T and B cells in the absence of an evident follicular organisation \[[@B20]\], whereas in more than 50% of synovial tissue samples from RA \[[@B20]\] and a considerable proportion of patients with JIA (M Gattorno, unpublished data), diffuse T and B lymphocytic infiltrates in the absence of aggregates or follicular structures are observed. Interestingly, in the individual patients with RA, the pattern of lymphocytic infiltration was found to persist unaltered over time, and showed similar features in all biopsies taken from different joints at the same type \[[@B20]\].
In our study, both CCR7 and its ligand CCL21 were found to be abundantly expressed in synovial biopsies, irrespective of the pattern of lymphoid infiltration. These observations support the hypothesis that CCR7 and its ligands have a direct function in the recruitment of memory T cells to the inflamed synovium, one that is independent of their ability to organise in lymphoid structures.
In this respect, the recent demonstration of different regulation of CCL21 in lymphoid and non-lymphoid tissues is noteworthy. Lymphotoxin-α directs the formation of lymph nodes and Peyer\'s patches through the induction of adhesion molecules and the production of chemokines, including CCL21, by the mesenchymal organiser cells during the early developmental steps \[[@B4],[@B35]\]. Lymphotoxin-α-deficient mice show a marked impairment of lymphoid organisation in secondary lymphoid organs, but normal recruitment of naive and memory T cells to peripheral inflamed tissue through the CCR7/CCL21 system \[[@B36]\]. CCL21 and CCR7 might therefore either regulate lymphoid neogenesis by a lymphotoxin-dependent mechanism or recruit T cells to the inflamed tissues by a lymphotoxin-independent mechanism.
Taken together, our findings suggest that CCR7^+^memory T cells can be directly recruited, with the possible contribution of other chemokines such as CXCR3 ligands, to the synovium, where they undergo further differentiation leading to the downregulation of CCR7 from the cell surface and the concomitant upregulation of CCR5. This differentiation might be driven either by antigen-dependent or antigen-independent mechanisms. In fact, cytokines produced in the synovial microenvironment (namely interleukin-7 and interleukin-15) might allow the proliferation, expansion and differentiation of CCR7^+^memory T cells into effector cells, marked by the downregulation of CCR7, the upregulation of CCR5 and the production of IFN-γ \[[@B37]\]. In this model, CCR5 could represent the major chemokine receptor used for CD4^+^memory T cell locomotion within the inflamed tissue, according to a step-by-step navigation model through different chemoattractant gradients \[[@B38]\]. In contrast, the enrichment of CCR7^-^CCR5^+^cells infiltrating the lining and sublining layer could be also related to the presence of other relevant effector cells, such as the granzyme B^+^cytotoxic cells \[[@B39]\].
Conclusion
==========
The present study delineates a coordinated pattern of expression of homeostatic and inflammatory chemokines in the inflamed synovium, with potential implications for the mechanisms regulating the intra-tissue migration and local differentiation of inflammatory cells.
Abbreviations
=============
ANOVA = analysis of variance; APAAP = alkaline phosphatase--anti-alkaline phosphatase; eOJIA = extended oligoarticular JIA; FITC = fluorescein isothiocyanate; GC = germinal centre; IFN = interferon; JIA = juvenile idiopathic arthritis; mAb = monoclonal antibody; MNC = mononuclear cells; PB = peripheral blood; PBS = phosphate-buffered saline; pOJIA = persistent oligoarticular JIA; RA = rheumatoid arthritis; RF = rheumatoid factor; SF = synovial fluid; TC = Tri color.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
MG conceived and coordinated the study, performed patients\' selection and wrote the manuscript. IP, AM and VP participated in the study design and helped to draft the manuscript. IP, FM, SC and FF performed the cytofluorimetric analysis and chemotaxis studies. AU performed the enzyme-linked immunosorbent assay for CCL21 determination in sera and SF, and helped to draft the manuscript. AG, AF and CG performed the immunohistochemical analysis of synovial tissue. All authors read and approved the final manuscript.
Acknowledgements
================
Part of this work was funded by Italian Ministry of Health (Ricerca Corrente) and Italian Multiple Sclerosis Society.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Expression of CXCR3 and interferon (IFN)-γ by (SF) CCR7^+^and CCR7^-^memory CD4^+^cells from synovial fluid. IFN-γ expression was investigated by three-colour staining of freshly isolated SF CD45RO^+^cells with CD4--fluorescein isothiocyanate (FITC), anti-CCR7--phycoerythrin (PE) and anti-CCR5--CyChrome monoclonal antibodies (mAbs) or CD4--TC (where TC stands for Tri-color), anti-CCR7--PE and anti-IFN-γ mAbs, respectively; CXCR3 expression was investigated by triple staining with CD4--TC, anti-CCR7--PE and anti-CXCR3--FITC, as described in the Methods section. Subsequently, cytofluorimetric analysis was performed by gating on the CD4^+^CCR7^+^and CD4^+^CCR7^-^lymphocyte subsets. Data are expressed as percentages of positive cells or/and mean fluorescence intensity. **(a, b)**Expression of IFN-γ (a) and CXCR3 (b) by SF CCR7^+^and CCR7^-^memory CD4^+^cells from 10 patients with juvenile idiopathic arthritis (JIA). Boxes contain values falling between the 25th and 75th centiles; whiskers show lines that extend from the boxes represent the highest and lowest values for each subgroup. Differences between paired SF mononuclear cells were evaluated by the Wilcoxon rank test. **(c, d)**Dot plots show the cytofluorimetric analysis IFN-γ (c) and CXCR3 (d) expression by the gated CD4^+^CCR7^+^(gate 1) and CD4^+^CCR7^-^(gate 2) cell populations in three representative patients with JIA.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Expression of CD4, CD20, CD45RO, CCR7, CCR5 and CXCR3 in synovial tissue obtained from patients with juvenile idiopathic arthritis (JIA) after synoviectomy. **(a, b)**Presence of CCR7^+^cells (red) in the sublining layer of a synovial tissue characterised by a diffuse lymphocytic infiltrate (scattered CD4^+^cells, brown) from a 14-year-old girl with antinuclear antibody-positive (ANA^+^) oligoarticular JIA (no. 1, Table 3) (Magnification × 20). (**c--n)**Serial stainings with CD20, CD4, CCR7, CXCR3 and CCR5 monoclonal antibody in synovial tissue from a 12-year-old girl with ANA^+^oligoarticular JIA (no. 4, Table 3). The distribution of cells positive for CCR7, CCR5 and CXCR3 is shown in two different areas containing T and B cell aggregates (magnification × 10). **(o, p)**Different expression of CCR7 (red) and CCR5 (brown) in synovial membrane infiltrate (patient no. 1, Table 3) (magnification × 4.5). CCR7^+^cells are observed exclusively in the perivascular lymphocytic infiltrate of the deep sublining layer (open rectangle). Conversely, CCR5-positive cells are prevalently observed at the level of the lining layer and superficial subintima (\*) and in the perivascular infiltrates of the sublining layer (\*\*).
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Chemotactic activity of CD4 memory T cells from the synovial fluid of eight patients with juvenile idiopathic arthritis to inflammatory (CXCL11 and CCL3) and homeostatic (CCL19, CCL21) chemokines. Results are expressed as the percentage of migrated cells in the total cell input (see also the Methods section).
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Expression of CCL21 in synovial fluid and tissue. **(a)**CCL21 concentrations in sera from 15 age-matched healthy controls, paired sera (Sera) and synovial fluids (SF) from 28 patients with juvenile idiopathic arthritis (JIA). Lines represent median values. Boxes contain values falling between the 25th and 75th centiles; whiskers show lines that extend from the boxes represent the highest and lowest values for each subgroup. The heterogeneity test among the three subgroups was highly significant (Kruskal--Wallis analysis of variance test, *P*= 0.0045). At *post hoc*analysis, differences between paired sera and SF were evaluated by the Wilcoxon rank test. Difference between JIA sera and healthy controls were evaluated by the Mann--Whitney [*U*-test]{.underline}. **(b)**Expression of CCL21 in perivascular aggregates and vascular endothelium in synovial tissue with diffuse lymphocytic infiltration from 10-year-old girl with persistent oligoarticular JIA. **(c, d)**Double staining with anti-CCR7 (red) and anti-CCL21 (brown) monoclonal antibodies at different magnifications (×10 and ×40, respectively) shows CCL21 expression by endothelial cells of vessels located in the sublining zone of the superficial subintima.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Clinical and laboratory features of patients with juvenile idiopathic arthritis at the time of phenotypic and functional studies of peripheral blood and synovial fluid lymphocytes
:::
Course No. of patients Age (years) Disease duration (years) No. of joints with active/limited range of motion PGI ESR (mm/h) Treatment; *n*
---------------- ----------------- ------------------ -------------------------- --------------------------------------------------- ------------- -------------- -------------------
Polyarticular 10 10.9 (3.3--16.1) 3.2 (0.5--11.2) 6 (1--16)/12.2 (1--29) 8.3 (6--10) 65 (23--131) NSAID, MTX; 7
NSAID, CS, MTX; 2
NSAID alone; 1
Oligoarticular 15 8.5 (3--17.9) 2.4 (0.3--14) 1.3 (1--3)/1.5 (1--3) 6.3 (5--10) 26.5 (7--90) NSAID alone; 11
NSAID, MTX; 2
Nil; 2
Results are means (ranges in parentheses).
CS, corticosteroids; ESR, erythrocyte sedimentation rate; MTX, methotrexate; NSAID, non-steroidal anti-inflammatory drugs; PGI, physician global index.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Clinical and laboratory features of patients with juvenile idiopathic arthritis at the time of determination of CCL21 in sera and synovial fluid
:::
Course No. of patients Age (years) Disease duration (years) No. of joints with active/limited range of motion PGI ESR (mm/h) Treatment; *n*
---------------- ----------------- ------------------ -------------------------- --------------------------------------------------- ------------- -------------- -------------------
Polyarticular 12 11.5 (4.3--14.3) 3.4 (0.5--1.2) 5.3 (1--13)/18.3 (1--22) 7.8 (5--10) 78 (24--137) NSAID, MTX; 8
NSAID, CS, MTX; 2
NSAID, CS; 1
NSAID alone; 1
Oligoarticular 16 7.5 (2.1--15.9) 2.4 (0.3--14) 1.8 (1--4)/1.9 (1--4) 6.9 (5--10) 31.5 (5--65) NSAID alone; 11
NSAID, MTX; 2
Nil; 3
Results are means (ranges in parentheses). See also \'Patients\' in the Methods section.
CS, corticosteroids; ESR, erythrocyte sedimentation rate; MTX, methotrexate; NSAID, non-steroidal anti-inflammatory drugs; PGI, physician global index.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Distribution of chemokine receptors of synovial tissues from patients with juvenile idiopathic arthritis
:::
Patient no. Age (year) Form Pattern CD3 CCR7 CXCR3 CCR5
------------- ------------ ------------ --------- ----- ------ ------- ------ ----- ----- ----- ----- ---- ----
1 14 Oligo per. Diffuse +++ \+ +++ NP +++ +++ NP +++ ++ NP
2 19 Poly RF- Diffuse \+ \- ++ NP \+ +++ NP \+ \- NP
3 10 Oligo per. T-B ++ \- \+ ++ \+ ++ ++ ++ \+ \-
4 12 Oligo ext. T-B +++ ++ +++ +++ ++ +++ +++ ++ \+ ++
5 7 Poly RF- T-B ++ \- \- \+ ++ +++ +++ +++ \+ \+
6 15 Poly RF- GC +++ \- \+ ++ +++ +++ +++ +++ \+ \-
Oligo per., persistent oligoarticular; oligo ext., extended oligoarticular; Poly RF-, polyarticular rheumatoid factor-negative; NP, not present; T-B, aggregates of T and B cells; GC, T and B cell aggregates with germinal centre (GC)-like reaction. Scoring: -, absence of positive cells in high-power field; +, 1--10 positive cells per high-power field; ++, 10--20 positive cells per high-power field; +++, more than 20 positive cells per high-power field (see also the Methods section). Lining, lining layer and superficial subintima; sublining, perivascular infiltrates in sublining layer; aggregates, T and B cell aggregates with or without GC-like reaction. All evaluations were performed in areas characterised by a clear lymphocyte (anti-CD3 and anti-CD4 positive cells) infiltration (see also the Methods section and Fig. 3).
:::
|
PubMed Central
|
2024-06-05T03:55:54.947602
|
2005-1-12
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065323/",
"journal": "Arthritis Res Ther. 2005 Jan 12; 7(2):R256-R267",
"authors": [
{
"first": "Marco",
"last": "Gattorno"
},
{
"first": "Ignazia",
"last": "Prigione"
},
{
"first": "Fabio",
"last": "Morandi"
},
{
"first": "Andrea",
"last": "Gregorio"
},
{
"first": "Sabrina",
"last": "Chiesa"
},
{
"first": "Francesca",
"last": "Ferlito"
},
{
"first": "Anna",
"last": "Favre"
},
{
"first": "Antonio",
"last": "Uccelli"
},
{
"first": "Claudio",
"last": "Gambini"
},
{
"first": "Alberto",
"last": "Martini"
},
{
"first": "Vito",
"last": "Pistoia"
}
]
}
|
PMC1065324
|
Introduction
============
Sjögren\'s syndrome (SS) is characterized by diffuse chronic inflammation of exocrine glands, which leads to symptoms and complaints referred to as \'sicca syndrome\' \[[@B1]\]. No single instrumental or laboratory parameter is available for the diagnosis of SS, which relies instead on the evaluation of multiple clinical, serological, functional, and morphological parameters \[[@B2]\], such as those proposed and validated by a group of investigators sponsored by the European Community (now the European Union) \[[@B3],[@B4]\] and recently revised by the American-European Consensus Group (AECG) \[[@B5]\]. The presence of chronic inflammatory infiltrates in lip salivary glands, as assessed with minor salivary gland biopsy (MSGB), is one of the parameters included in most criteria sets proposed for SS classification \[[@B3],[@B5]-[@B9]\]. Salivary gland inflammation is assessed by scoring the degree of infiltration according to the method of Greenspan and Daniels, who defined the focus score (FS) as the number of inflammatory infiltrates of at least 50 cells present in 4 mm^2^of gland surface unit \[[@B10],[@B11]\]. Different criteria sets consider as positive a FS ≥ 1 or FS ≥ 2 \[[@B3],[@B9]\]. Although the methodology of sampling, processing, and examining MSGBs has been standardized \[[@B10],[@B11]\], the reproducibility of the routine histopathological evaluation in the diagnosis of SS at different section levels within the same biopsy specimen has been recently challenged \[[@B12],[@B13]\]. To avoid any bias that might therefore arise, the examination of multiple levels of tissue has been recommended, to maximize the number of foci, the glandular area, and the technical quality of the material, although the number of sections required has not yet been standardized \[[@B12]\].
In this study, we tried to standardize a protocol for histopathological MSGB evaluation in which the FS is assessed by examining a larger area of the biopsy tissue, and we investigated how the FS obtained affects the number of patients classified as having SS, as compared with the routine method, using the classification criteria recently proposed by the AECG \[[@B5]\]. The diagnostic accuracy of the test was validated against the clinical re-evaluation of the patients performed by two experienced rheumatologists after at least 1 year of follow-up.
Materials and methods
=====================
Selection criteria
------------------
We retrospectively studied a consecutive series of patients thoroughly investigated at our hospital between 1998 and 2002 for suspected primary SS, including a follow-up of at least 1 year after the diagnostic evaluation. Patients with secondary SS or who had been diagnosed by biopsy as having nonspecific inflammation, fibrosis, and atrophy of the gland were excluded \[[@B10]-[@B12]\]. Less-than-optimal tissue area (biopsy section area less than 4 mm^2^) was not considered a criterion for exclusion, provided that at least one normotrophic glandular lobule had been sampled.
Baseline clinical and histopathological evaluation
--------------------------------------------------
All patients had undergone thorough clinical and instrumental evaluation \[[@B3],[@B4]\], including MSGB performed as suggested by Daniels \[[@B11]\]. The diagnosis of SS was established for all patients according to the classification criteria proposed by the AECG \[[@B5]\]. MSBG samples were fixed in formalin, processed, and embedded in paraffin according to standardized laboratory methods. Baseline histopathological slides containing 4-μm-thick sections stained with hematoxylin and eosin were reviewed by a pathologist, blinded to clinical and laboratory data, who recorded for each patient the number of glands, the sample surface area, the presence of alterations suggestive of nonspecific sialoadenitis, and the baseline FS \[[@B10],[@B11]\]. The lymphocytic focus and the focus score were defined according to Greenspan and Daniels \[[@B10],[@B11]\]. In individual biopsies, lobules with acinar atrophy and diffuse fibrosis were excluded from diagnostic evaluation. The histological parameter was considered as negative in the absence of any inflammatory infiltrate (FS = 0) and in the presence of less than 1 focus per 4 mm^2^(0 \< FS \< 1) \[[@B5]\]; the presence of one or more foci per 4 mm^2^was considered positive when the adjacent glandular parenchyma was histologically normal. We further classified patients with a positive FS into two groups, those with fewer than two foci per 4 mm^2^(1 ≤ FS \< 2) and those with two or more (FS ≥ 2). The area of the biopsy sections was assessed with video-assisted morphometric software capable of measuring the area of delineated surfaces (ImageDB System, Casti Imaging, Cazzago di Pianiga, Italy). The comparison of automated and manual area measurements of a smaller series of MSGB sections did not show a significant difference (data not shown). This prompted us to choose the automated system to simplify the examination of the large number of samples involved in the study.
Serial histopathological re-evaluation
--------------------------------------
Sample blocks were recut at two additional levels, about 200 and 400 μm deeper than the original section. Sections 4 μm thick corresponding to these levels were collected on separate slides and stained with hematoxylin and eosin. Considering that an infiltrate of 50 lymphocytes in our section had a mean diameter of 50 μm, we assumed that the interposition of 200 μm between the evaluated sections was enough to ensure that the FS recorded at each level was independent of the other two and that if the same focus was present in two section levels, the focus itself was large enough to justify repeated scoring. The two new sections were blindly examined by the same pathologist, who again recorded the area and the focus score for each level. For each patient, the total number of foci at all three levels and the total surface area measured at all levels were used to calculate a cumulative FS (cFS) for the three sections.
Reclassification of patients
----------------------------
The cFS obtained after re-evaluation was entered in the AECG criteria set \[[@B5]\], to obtain a re-classification of each patient. To compare the diagnostic performance of the original classification and the reclassification, a \'gold standard\' was needed independent of the AECG criteria set. We adopted as reference standard the opinion of experienced clinicians, analogously to what had been done by the European Community Study Group on Diagnostic Criteria for Sjögren\'s Syndrome when SS and control patients were selected to validate the proposed criteria \[[@B3]-[@B5]\]. Briefly, three experienced rheumatologists, blinded to the results of the histopathological re-evaluation, performed a clinical evaluation of each patient and reviewed the patient\'s charts including the original clinical, laboratory, and instrumental evaluation, and the subsequent documentation covering at least 1 year of follow-up and treatment response. On this basis they were requested to judge whether individual patients had SS.
Statistical analysis
--------------------
Quantitative data are shown as means ± standard deviation (SD). Specificity and sensitivity were assessed with their 95% confidence intervals (CI). Differences in frequencies were evaluated by means of chi-square statistics or the Fisher exact test, as appropriate. Given the known limitations of diagnostic accuracy as a parameter for measuring the diagnostic performance of a test, specificity and sensitivity were compared using receiver operating characteristic (ROC) curves \[[@B14]\]. A *P*value of less than 0.05 was considered to indicate statistical significance. All tests were two-sided. Analyses were performed with Statistica for Windows (StatSoft Inc, 2002, Tulsa, OK, USA) and MedCalc software.
Results
=======
Baseline examination
--------------------
The study series comprised 138 patients, 65 of whom had a baseline FS = 0, 14 with 0 \< FS \< 1, 18 with 1 ≤ FS \< 2, and 41 with FS ≥ 2. Eighteen patients had incomplete clinical data that hampered either the AECG classification or the clinical re-evaluation. These patients (8 with FS = 0, 3 with 0 \< FS \< 1, 3 with 1 ≤ FS \< 2, and 4 with FS ≥ 2) were excluded from further analysis. The final series included 120 patients, for whom demographic, biopsy, and clinical data and the result of the clinical re-evaluation are presented in Table [1](#T1){ref-type="table"}.
Histological re-evaluation
--------------------------
In 96 (80%) of the 120 biopsies, the FS group did not change after serial sectioning and calculation of the cFS. In 14 of these biopsies, the FS group changed but this did not affect that patient\'s negative or positive status. In the biopsies for the other 10 patients, 1 (1.7%) of the 57 with a baseline FS = 0 and 1 (9%) of the 11 with a baseline score of 0 \< FS \< 1 switched to a FS consistent with SS according to AECG criteria (FS ≥ 1). At clinical re-evaluation, these two patients were considered not to have SS. Seven (46%) of the 15 patients with a baseline score of 1 ≤ FS \< 2 and one (3%) of 37 with a baseline FS ≥ 2 switched to a grade inconsistent with SS (FS \< 1). On clinical re-evaluation, 7 of these 8 patients were assessed as not having SS.
Patient reclassification according to AECG criteria
---------------------------------------------------
When the cFSs were entered in the AECG criteria set \[[@B5]\], the baseline classifications of the 63 non-SS patients were not changed, while the classifications of 7 of the 57 patients originally classified as having SS were changed to non-SS (Table [2](#T2){ref-type="table"}). The classification was changed in 6% of the 120 patients. Six of these seven patients had a baseline score of 1 ≤ FS \< 2 and one had a baseline FS ≥ 2. On clinical re-evaluation, all these seven patients were judged not to have SS. The clinical re-evaluation also refuted 7 of the 113 (6.2%) classifications that had not been changed at biopsy revision. Considering the clinical re-evaluation as the reference gold standard, the number of false-negative AECG classifications did not change (3 of 63 AECG non-SS cases), while the number of false positives was reduced from 11 to 4 (63.6% reduction).
Comparison of sensitivity and specificity between baseline and multilevel FS evaluation
---------------------------------------------------------------------------------------
In the present series of 120 patients fully evaluated for SS, the sensitivity and specificity of the baseline AECG criteria set were 93.9% and 84.5%, respectively. Reclassification with cFS did not affect sensitivity, whereas specificity changed to 94.4% (*P =*0.056), increasing the accuracy from 88.3% (95% CI 81.2--93.5) to 94.2% (95% CI 88.3--97.6). Pairwise comparison of the ROC curves showed a statistically significant difference between patient classification before and after multilevel FS evaluation (difference between areas: 0.049 \[SE 0.021\]; 95% CI 0.009--0.089; *P =*0.016) (Fig. [1](#F1){ref-type="fig"}). Sensitivity and specificity did not change for biopsies with FS = 0 or FS \< 1 (inconsistent with SS), while specificity increased substantially in biopsies consistent with SS (FS ≥ 1) (Table [2](#T2){ref-type="table"}). Pairwise comparison of the ROC curves showed a statistically significant difference (*P =*0.013) only in biopsies with 1 ≤ FS \< 2 (difference between areas: 0.43 \[SE: 0.17\]; 95% CI 0.09--0.76; *P =*0.013; Fig. [1](#F1){ref-type="fig"}). The diagnostic accuracy of the MSGB histological analysis considered independently of other criteria changed from 85.8% (95% CI 78.3--91.5) to 90.8% (95% CI 84.2--95.3), but the comparison of the ROC curves did not show a statistically significant difference (*P =*0.15).
Discussion
==========
In the present study, we show that the histopathological evaluation of salivary gland biopsies with multilevel sectioning and assessment of a cumulative focus score (cFS) changes the baseline classification in 6% of patients evaluated for SS and increases the diagnostic performance of the criteria recently proposed by the AECG for SS classification \[[@B5]\]. In particular, multilevel evaluation improved the diagnostic accuracy of biopsies with a baseline FS between 1 and 2, which is the most critical cutoff in SS histopathological evaluation.
The present study was prompted by a recent paper documenting that MSGB grading of inflammation was scarcely reproducible at different section depths, and that the difference between grades recorded at baseline and at deeper levels was sufficient to change the biopsy from positive to negative or vice versa in 10% of grade I (FS = 0), 44.4% of grade II (0 \< FS \< 1), 88.8% of grade III (1 ≤ FS \< 2), and 40% of grade IV (FS ≥ 2) biopsies \[[@B13]\]. The authors of that paper recommended that multiple sections of MSGB should be examined to improve the reliability of the histopathological grading. However, they did not suggest how many sections should be examined or how to deal for diagnostic purposes with the different scores obtained at different levels, nor did they give a clinical interpretation of their results by entering them in a criteria set for SS patient classification.
On this basis, we aimed at assessing if the histopathological evaluation of a larger area of MSGB tissue, as obtained by cutting the biopsy sample at additional section levels, could increase the diagnostic performance of the histopathological study and of the AECG criteria set proposed for the classification of SS. We chose a minimum requirement of three different section levels, by analogy with the procedure standardized for the histopathological study of endomyocardial biopsies \[[@B15]\], assuming that a 200-μm distance should ensure the detection of independent foci on each section while reducing the chance of missing the smaller ones, thus allowing estimation of the overall density of inflammatory foci with sufficient precision.
With reference to the diagnostic gold standard, when patients were classified according to the AECG criteria set including the cFS, specificity increased by 9.8%, and the pairwise comparison of the ROC curves showed a statistically significant improvement of the diagnostic performance, mostly due to the increased test specificity in biopsies with 1 ≤ FS \< 2, whereas the increase was minimal in FS ≥ 2 and null in biopsies inconsistent with SS (0 \< FS \< 1). One advantage of the proposed method of MSGB evaluation is that specificity is increased without affecting sensitivity; on the other hand, it was shown that improving sensitivity by means of increasing the cutoff value of positive FS resulted in a substantial reduction of specificity \[[@B16]\].
To explain the increased specificity observed with examination of multilevel salivary gland biopsies, it should be considered that, because of the uneven distribution of inflammatory infiltrates in the gland \[[@B14]\], the examination of a single tissue section might easily either overestimate or underestimate the FS, while the observation of a larger area of biopsy sample would allow a more precise quantification of the focus distribution, provided that the sections are distant enough to avoid recutting and rescoring of the same focus. In accordance with this hypothesis, and confirming previous results \[[@B13]\], after multilevel examination the higher numbers of FS changes proven to be relevant for classification and clinical diagnosis were seen in patients with mild to moderate MSGB inflammatory infiltrates (1 ≤ FS \< 2), while very few relevant changes were recorded in patients with negative or highly positive biopsies (FS \< 1 or FS ≥ 2). We suggest that in mild inflammation, lymphocytic foci are unevenly distributed through the gland, so that positive baseline sections can occasionally be followed by sections with less or no inflammation, whereas negative or highly positive biopsies (FS \< 1 and ≥ 2) are likely to be more homogeneous. Our observations also confirmed the common knowledge that no single test can be reliably applied to the diagnosis of SS \[[@B2]-[@B9]\]. In fact, the performance of the test was significantly improved when the cFS was entered in the criteria set, but not when the histopathological test was considered alone.
One potential limit of the present study is represented by the need to introduce a gold standard reference to assess the diagnostic accuracy of the test, independent of the widely accepted AECG criteria set for SS classification. In fact, after clinical re-evaluation, which we adopted as a gold standard, some patients appeared to have been misclassified according to AECG criteria. This only partial correspondence between the judgement of experienced clinicians and classification criteria is a well-known problem in the diagnosis of rheumatological disorders and justifies the requirement of a wide criteria set for patient classification. In the absence of single, straightforward diagnostic parameters, a thorough patient\'s chart and follow-up revision by experienced rheumatologists was chosen as reference gold standard, by analogy with what has been done in many rheumatological studies, including that of the European Community Study Group on Diagnostic Criteria for SS \[[@B3]-[@B5]\]. Accordingly, a multicenter study would be useful to better standardize the procedure of evaluating FSs by oral pathologists, backed by a larger panel of experienced clinicians, because the clinical performance of SS classification criteria could be improved.
Conclusion
==========
The assessment of a cumulative focus score (cFS) obtained at three different section levels on minor salivary gland biopsies, cut at least 200 μm apart, can improve the diagnostic accuracy of the criteria set used for SS classification, especially in biopsies with a baseline FS between 1 and 2. Since the value of the MSGB biopsy has been confirmed by the recent AECG revision of the SS classification criteria \[[@B5]\], the increase of the diagnostic performance of the histological study will further help to correctly identify SS patients.
Abbreviations
=============
AECG = American-European Consensus Group; cFS = cumulative FS; CI = confidence interval; FS = focus score; MSGB = minor salivary gland biopsy; ROC = receiver operating characteristic; SE = standard error; SS = Sjögren\'s syndrome.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
PM participated in the design of the study, performed the histopathological analysis, coordinated the study, and drafted the manuscript. AM and RC reviewed and discussed patients\' charts for clinical re-evaluation. OE performed all salivary gland biopsies. CV participated in case collection and data analysis. CT participated in the design of the study and performed the statistical analysis. ES and CM conceived the study and participated in its design. CM also participated in the clinical re-evaluation of patients. All authors read and approved the final manuscript.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Statistical comparison of the diagnostic performance of the American-European Consensus Group (AECG) criteria for Sjögren\'s syndrome with baseline and cumulative focus scores (FSs). Receiver operating characteristic (ROC) curves were used to compare the sensitivity and specificity of the AECG criteria with the baseline focus score and with the FS obtained after multilevel histopathological evaluation, with respect to the gold standard of patient re-evaluation by the experienced rheumatologists. The diagnostic performance was significantly improved in the overall series (top left panel; *P=*0.016), mostly because of the improvement in the group of patients with 1 ≤ FS \< 2 (bottom left; *P=*0.013). No difference was observed when FS = 0. No ROC curve could be obtained in the group of patients with 0 \< FS \< 1, because of the absence of cases classified as Sjögren\'s syndrome at clinical re-evaluation (positive gold standard). CI, confidence interval.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Demographic, biopsy, and clinical data for 120 patients given salivary gland biopsies for Sjögren\'s syndrome (SS)
:::
Clinical and laboratory parameters FS^a^= 0 0 \< FS \< 1 1 ≤ FS \< 2 FS ≥ 2
---------------------------------------------------- ------------- -------------- ------------- ------------
No. of patients 57 11 15 37
Sex 9M/48F 1 M/10 F 0 M/15 F 3 M/34 F
Age (years) 46 ± 12 46 ± 11 54 ± 14 56 ± 13
Baseline biopsy area (mm^2^) 6.1 ± 5.6 8.3 ± 2.8 8.6 ± 4.3 4.7 ± 2.6
Cumulative area of three biopsies (mm^2^) 17.1 ± 15.1 21.8 ± 9.3 22.6 ± 12.7 12.9 ± 6.4
Findings \[No. (%)\]
Dry eyes 49 (86) 9 (81) 14 (93) 33 (89)
Xerostomia 45 (79) 7 (63) 14 (93) 35 (94)
Positive Schirmer\'s test 24 (42) 4 (36) 9 (60) 18 (51)
Reduced salivary flow rate 30 (53) 7 (63) 13 (86) 34 (91)
Antinuclear antibodies 29 (51) 8 (73) 8 (53) 31 (83)
Ro/SS-A 18 (31) 3 (27) 6 (40) 25 (67)
La/SS-B 3 (5) 0 3 (20) 10 (27)
Rheumatoid factor 27 (47) 5 (45) 7 (46) 33 (89)
SS according to AECG criteria \[No. (%)\] 7 (12) 1 (9) 15 (100) 34 (92)
SS according to clinical re-evaluation \[No. (%)\] 7 (12) 0 (0) 8 (53) 34 (92)
^a^The focus score (FS) is the number of inflammatory infiltrates of at least 50 cells present in 4 mm^2^of salivary gland area. AECG, American-European Consensus Group; F, female; M, male; SS-A, anti-Ro60 antibodies; SS-B, anti-La antibodies.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Changes in classification determined by multilevel salivary gland biopsies for Sjögren\'s syndrome (SS)
:::
-----------------------------------------------------------------------------------------------------------------
Test results and diagnostic accuracy FS^a^= 0 0 \< FS \< 1 1 ≤ FS \< 2 FS ≥ 2 Total
-------------------------------------- -------------- -------------- -------------- -------------- --------------
No. of patients 57 11 15 37 120
AECG classification changes 0 0 6 (40°%) 1 (3%) 7
Baseline sensitivity (95% CI) 85.7%\ \- 100%\ 94.1%\ 93.9%\
(42.2--97.6) (62.9--100) (80.3--99.1) (83.1--98.6)
Revised sensitivity (95% CI) 85.7%\ \- 100%\ 94.1%\ 93.9%\
(42.2--97.6) (62.9--100) (80.3--99.1) (83.1--98.6)
Baseline specificity (95% CI) 98%\ \- 0%^b^\ 33.3%^b^\ 84.5%\
(89.3--99.7) (0--41.1) (5.5--88.4) (74.0--92.0)
Revised specificity (95% CI) 98%\ \- 85.7%\ 66.7%\ 94.4%\
(89.3--99.7) (42.2--97.6) (11.6--94.5) (86.2--98.4)
-----------------------------------------------------------------------------------------------------------------
^a^The focus score (FS) is the number of inflammatory infiltrates of at least 50 cells present in 4 mm^2^of salivary gland area. ^b^Very low specificity is due to the absence (1 ≤ FS \< 2) or extremely low number (FS ≥ 2) of patients classified as non-SS according to AECG criteria. AECG: American-European Consensus Group; CI: confidence interval; -, could not be evaluated with the available data.
:::
|
PubMed Central
|
2024-06-05T03:55:54.951081
|
2005-1-17
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065324/",
"journal": "Arthritis Res Ther. 2005 Jan 17; 7(2):R343-R348",
"authors": [
{
"first": "Patrizia",
"last": "Morbini"
},
{
"first": "Antonio",
"last": "Manzo"
},
{
"first": "Roberto",
"last": "Caporali"
},
{
"first": "Oscar",
"last": "Epis"
},
{
"first": "Chiara",
"last": "Villa"
},
{
"first": "Carmine",
"last": "Tinelli"
},
{
"first": "Enrico",
"last": "Solcia"
},
{
"first": "Carlomaurizio",
"last": "Montecucco"
}
]
}
|
PMC1065325
|
Introduction
============
Thrombin is a multifunctional protease that can activate hemostasis and coagulation through the cleavage of fibrinogen to form fibrin clots. Increasing fibrin deposition is a predominant feature of rheumatoid arthritis (RA) in synovial tissue, which contributes to chronic inflammation and progressive tissue abnormalities \[[@B1]\]. Thrombin also acts as a mitogen to stimulate the abnormal proliferation of synovial cells during RA pathogenesis. In this regard, thrombin can elevate the expression of nuclear factor-κB, interleukin-6, and granulocyte colony-stimulating factor in fibroblast-like cells of the RA synovium \[[@B2],[@B3]\]. By a similar mechanism, thrombin can upregulate the transcription of vascular endothelial growth factor receptor and thereby induce the permeability, proliferation, and migration of capillary endothelial cells or their progenitors during angiogenesis \[[@B4]-[@B6]\]. Thrombin also plays an important role in the proinflammatory process by stimulating neutrophil adhesion to vessel walls and releasing prostacyclin \[[@B7]\]. Thus, thrombin is essential for enhancing synovial thickness and inflammation during the pathogenesis of RA.
The principal plasma inhibitor of thrombin is antithrombin, a single-chain 51 kDa glycoprotein that is synthesized in liver. The inhibitory activity of antithrombin on thrombin is significantly enhanced by heparin, a type of glycosaminoglycan (GAG) \[[@B8]\]. The GAG family comprises large anionic polysaccharides with similar disaccharide repeats of uronic acid and hexosamine. Physiologically important GAGs include hyaluronic acid (HA), chondroitin sulfates, keratan sulfate (KS), heparin, and heparan, which are the major components of joint cartilage, synovial fluid, and other soft connective tissues \[[@B9],[@B10]\]. Along with the destruction of RA joint tissue, a remarkable quantity of various GAG molecules, especially HA, are released from the extracellular matrix of the synovium \[[@B9],[@B10]\], which is a key feature of RA progression. Because GAGs and heparin share a similar molecular structure, we investigated how HA and other GAGs affect antithrombin activity.
Methods
=======
Highly purified HA, chondroitin sulfate A (CSA), chondroitin sulfate B (CSB), chondroitin sulfate C (CSC), KS, heparin, or heparan (Seikagaku, Tokyo, Japan) were incubated for 24 hours with human antithrombin III at 150 μg/ml (Sigma, St. Louis, MO, USA) at 37°C in working buffer (100 mmol/l Tris-HCl, pH 7.5) containing 5 mmol/l CaCl~2~or FeCl~3~. The concentration of antithrombin was determined according to its physiologic level in synovial fluid \[[@B11],[@B12]\]. The reaction was stopped with EDTA. Residual activity of antithrombin was analyzed using the chromogenic Actichrome AT III (American Diagnostica, Greenwich, CT, USA) kit, which quantifies antithrombin III activity as follows. After exposure to GAGs, antithrombin was incubated with the thrombin reagent provided with the kit and residual thrombin activity was determined by incubation with the thrombin-specific chromogenic substrate in the kit. Absorbance was measured at a wavelength of 405 nm. Hence, the inhibitory ability of antithrombin on thrombin was inversely proportional to the residual thrombin activity. This assay method is usually used in the clinical setting. We prepared a series of control tests in which HA, CSA, CSB, CSC, and KS were digested in 0.1 mol/l phosphate buffer (prepare 100 ml of the buffer with 94 ml of 0.1 M KH~2~PO~4~and 6 ml of 0.1 M K~2~HPO~4~, pH 6.2) at 37°C for 2 hours with 0.1 units/ml hyaluronidase (Seikagaku, Japan) before incubation with antithrombin. Hyaluronidase preferentially digests HA rather than other GAGs.
To determine whether HA can prevent heparin from stimulating antithrombin, we simultaneously incubated heparin (10 μg/ml) and various concentrations of HA with antithrombin (150 μg/ml) at 37°C for 24 hours in the presence of 5 mmol/l CaCl~2~. To investigate the effect of HA on antithrombin in the presence of other metal ions, we incubated HA (1 mg/ml) and human antithrombin III (150 μg/ml) at 37°C for 24 hours in the presence of CaCl~2~, FeCl~3~, KCl, MgCl~2~, and NaCl at various concentrations. Residual antithrombin activity was measured as described above.
Results
=======
In the absence of heparin, antithrombin partly inhibited thrombin activity. Low concentrations of HA did not significantly affect antithrombin activity, regardless of the presence or absence of Ca^2+^or Fe^3+^. However, HA concentrations above 250 μg/ml considerably suppressed the inhibitory ability of antithrombin against thrombin in the presence of Ca^2+^or Fe^3+^, and 1 mg/ml HA completely blocked antithrombin activity under the same conditions. Consequently, thrombin activity was gradually elevated by increasing HA concentrations between 250 and 1000 μg/ml. However, HA at concentrations above 1000 μg/ml progressively lost the ability to prevent inhibition of thrombin activity by antithrombin. Furthermore, HA after digestion with hyaluronidase inhibited antithrombin activity at relatively low concentrations (100 μg/ml) in the presence of Ca^2+^. This observation indicated that the inhibitory effect of HA on antithrombin was not caused by impurities in the reagent. The control without antithrombin indicated that HA does not directly affect thrombin (Fig. [1](#F1){ref-type="fig"}).
CSA, CSB, and CSC also inhibited the antithrombin effect in the presence of Ca^2+^but to a lesser extent than did HA (Fig. [2](#F2){ref-type="fig"}). KS did not significantly affect antithrombin activity. Exposing CSs and KS to hyaluronidase did not clearly change this effect, indicating that CSs themselves inhibit antithrombin (data not shown). In contrast to HA, heparin and heparan clearly stimulated thrombin inhibition by antithrombin (Fig. [2](#F2){ref-type="fig"}). However, the stimulatory effect of heparin was considerably decreased in the presence of HA and Ca^2+^. Moreover, the ability of HA to prevent heparin activity was progressively strengthened with increased concentrations of HA within the range 250--1000 μg/ml (Fig. [3](#F3){ref-type="fig"}). Other metal ions, including K^+^, Mg^2+^, and Na^+^, did alter the effect of HA on antithrombin (Fig. [4](#F4){ref-type="fig"}).
Discussion
==========
The destruction of joint tissue is a primary feature of RA. In the inflamed RA synovium, proliferating macrophages and colonizing lymphocytes, together with persistent angiogenesis, produce large amounts of matrix metalloproteinases that destroy the surrounding cartilage and extracellular matrix of connective tissue \[[@B13]\]. Because GAGs are the basic structural components of joint cartilage, synovial fluid, and soft tissues \[[@B9],[@B10]\], the RA synovium produces an abundance of free GAGs during tissue destruction. Among these, HA is a predominant component of the articular surface and synovial fluid, in which the HA concentration is between 1500 and 2500 μg/ml \[[@B14],[@B15]\]. Pitsillides and coworkers \[[@B14]\] found that the ratio of free HA to bound HA was significantly increased in the RA (4.53 ± 0.40) as compared with the healthy (1.87 ± 0.42) synovium, although the total concentration of hyaluronan was not increased in the rheumatoid synovium. Their histochemical staining also showed that hyaluronan was concentrated in the lining layer of noninflamed synovial membrane but was more uniformly distributed throughout rheumatoid samples. On the other hand, the HA level is very low among various other tissues. For example, the concentration of serum HA from healthy individuals averages 16 ng/ml, which is 1 × 10^5^fold lower than that in synovial fluid \[[@B16],[@B17]\].
The present study found that HA at concentrations between 250 and 1000 μg/ml significantly blocked the ability of antithrombin to inhibit thrombin. This finding helps to explain why RA occurs and develops in joint tissue, because the inflamed RA synovium is uniquely rich in free HA and other GAGs, along with extracellular matrix degeneration. Although the HA levels are higher in RA than in healthy sera \[[@B18]\], we demonstrated that the relatively low levels of HA do not prevent antithrombin activity and thus cannot cause blood clots in the circulation. Hence, only the conditions in the RA synovium can drive the pathogenesis of thrombin-related RA, which includes abnormal angiogenesis, extreme proliferation of fibroblast-like cells, excessive fibrin deposition, and proinflammatory processes. Thus, thrombin-related RA worsens because of the snowball effect of HA release in inflamed joints.
Our notion is supported by many other studies. Jones and coworkers \[[@B11]\] found that antithrombin activity is selectively depressed in RA synovial fluid as compared with that in osteoarthritis, although the concentration of the antithrombin--thrombin complex was significantly increased. Ohba and coworkers \[[@B12]\] also found high levels of thrombin activity in RA synovial fluid. These findings support the notion that inhibiting antithrombin activity plays an essential role in RA pathogenesis. Wang and coworkers \[[@B10]\] recently constructed a model of arthritis by injecting various GAGs into mice. We postulate that the injected GAGs significantly disrupted the inhibition of thrombin by antithrombin, which therefore caused connective tissue disease through abnormally activated angiogenesis, proinflammatory processes, and fibrin deposition. On the other hand, heparan, which has an almost identical structure to that of heparin but contains fewer sulfates, stimulated antithrombin activity in a similar manner to heparin. These observations indicate that the diverse effects of GAGs on antithrombin are due to differences in their molecular configurations. Heparin pentasaccharide can form complexes with antithrombin and expose a reactive proteinase binding loop on the protein surface \[[@B19],[@B20]\]. Because the molecular structure of HA is analogous to that of heparin, HA might exert its effect by binding to the heparin-binding region of antithrombin. However, such binding did not stimulate the activity of antithrombin as did heparin and heparan; in fact, it blocked the ability of antithrombin to inhibit thrombin. In the present study, the stimulatory effect of heparin on antithrombin was considerably decreased in the presence of HA, supporting the notion that HA could compete with heparin for the heparin-binding region of antithrombin.
Remarkably, HA affected the inhibition by antithrombin only within the range 250--1000 μg/ml. At concentrations above 2000 μg/ml, HA either lost its inhibitory effect or elevated the ability of antithrombin to inhibit thrombin. The physiologic level of free HA in the RA synovium is just within the range 500--1000 μg/ml \[[@B14]\]. Some clinical studies have shown that injecting HA into articular rheumatoid joints can ameliorate inflammation \[[@B21],[@B22]\]. Although further investigation is required to elucidate the exact mechanism by which HA inhibits antithrombin, the results of the present study do not refute the notion that optimal proteoglycan uptake can improve overall articular function in patients with arthritis.
Why HA inhibited antithrombin more after than before hyaluronidase digestion remains obscure. Perhaps the small HA molecule can easily bind and thus exert a more inhibitory role on antithrombin. Nagaya and coworkers \[[@B23]\] found high hyaluronidase activity in the synovial fluid and serum of RA patients, implying an abundance of small HA molecules in the RA synovium. Maneirio and coworkers \[[@B24]\] reported that HA at various molecular weights had different effects on the interleukin-1 induced synthesis of both nitric oxide and prostaglandin E~2~in chondrocytes. How Ca^2+^and Fe^3+^are involved in inhibiting antithrombin by HA is also poorly understood. Some investigators found that Ca^2+^dramatically promotes the ability of heparin to drive antithrombin activity \[[@B8],[@B25],[@B26]\]. Thus, both Ca^2+^and Fe^3+^ions might play similar roles in HA-induced changes in the configuration of antithrombin.
Synovial fluid from RA patients contains a far greater abundance of free iron than that from patients with osteoarthritis \[[@B27],[@B28]\]. It was reported that Fe^3+^stored in the RA synovium perpetuates inflammation by supporting the production of oxygen radicals and by promoting hyaluronic acid degradation, as well as the release of lysosomal enzymes \[[@B29]\]. Telfer and coworkers \[[@B30]\] recently found that proinflammatory cytokines produced in the RA synovium increased the accumulation of iron in synovial fluid. On other hand, Davies and coworkers \[[@B31]\] reported that neutrophils from synovial fluid and the circulation of RA patients could increase the release of free Ca^2+^at inflammatory sites. Caruthers and coworkers \[[@B32]\] also showed that calcium signaling is altered in T lymphocytes from RA patients.
Genome-wide single nucleotide polymorphism analysis has shown that peptidylarginine deiminase (PADI4), an enzyme that post-translationally catalyzes peptidyl arginine to citrulline, is closely associated with RA \[[@B33]\]. We recently found that recombinant human PADI4 protein inactivated human antithrombin III via citrullination *in vitro*. We also detected an increased level of citrullinated antithrombin in the plasma of RA patients \[[@B34]\]. PADI4 is extensively expressed in RA synovial tissue \[[@B35],[@B36]\]. Thus, we suggested that the citrullination of antithrombin is one potential pathway through which PADI4 contributes to the pathogenesis of RA \[[@B34]\]. This notion does not contradict the current findings. We postulate that the genetic, single nucleotide polymorphism-associated disorder of PADI4 and its excessive citrullination of antithrombin play important roles in initiating the RA pathogenic process, whereas inhibition of antithrombin by HA contributes to the development of RA rather than its initiation, because free HA in the synovium achieves high concentrations along with RA progression. Because of abundant Fe^3+^and altered Ca^2+^metabolism together with significant hyaluronidase activity in the RA synovium, thrombin-related RA specifically worsens in joint tissue as a result of antithrombin inactivation by local PADI4 and free HA (Fig. [5](#F5){ref-type="fig"}).
HA is an important component of the extracellular matrix. Thrombin and antithrombin play key roles in hemostasis and are involved in the pathogenic processes of many diseases \[[@B6],[@B37],[@B38]\]. The findings presented here should also be useful in investigating the nature of other diseases.
Conclusion
==========
At concentrations of 250--1000 μg/ml *in vitro*, HA blocked the thrombin-inhibitory ability of antithrombin in the presence of Ca^2+^and Fe^3+^. This finding suggested that the high concentration of free HA in diseased RA synovium locally blocks antithrombin under physiologic conditions and thereby deregulates the activity of thrombin. These processes in turn drive the thrombin-related pathogenesis of RA, which includes extensive fibrin deposition, extreme angiogenesis, and abnormal fibroblast-like cell proliferation. Our findings are consistent with those of previous reports regarding increased coagulation activity in the RA synovium.
Abbreviations
=============
CS = chondroitin sulfate; GAG = glycosaminoglycan; HA = hyaluronic acid; KS = keratan sulfate; PADI = peptidylarginine deiminase; RA = rheumatoid arthritis.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
XC designed and executed the study and prepared the manuscript. RY and KY supervised the project, evaluated data, and assisted in preparing the manuscript.
Acknowledgements
================
We thank every member of the Rheumatology Diseases Laboratory of Riken for their general contribution to making this study possible.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Effect of hyaluronic acid (HA) on antithrombin (AT). Various concentrations of HA, digested or not with hyaluronidase, were incubated with antithrombin in the presence of 5 mmol/l CaCl~2~or FeCl~3~. Thrombin activity in the absence of both HA and antithrombin (blank) was considered as 1 and the activities of the other tests were normalized based on comparisons with blank. Values are expressed as mean ± standard deviation of data from triplicate experiments.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Effects of various glycosaminoglycans (GAGs) on antithrombin (AT). Hyaluronic acid (HA), chondroitin sulfate A (CSA), chondroitin sulfate B (CSB), chondroitin sulfate C (CSC), keratan sulfate (KS), heparin, or heparan (500 μg/ml) was incubated with 150 μg/ml antithrombin and 5 mmol/l CaCl~2~. Controls consisted of only GAG or AT and blank (working buffer only). Thrombin activity of blank was considered as 1 and the activities of other tests were normalized based on comparisons with blank. Values are expressed as mean ± standard deviation of data from triplicate experiments.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Heparin stimulates antithrombin (AT) activity in the presence of hyaluronic acid (HA). Heparin (10 μg/ml) and various concentrations of HA were incubated with 150 μg/ml antithrombin in presence of 5 mmol/l CaCl~2~. Thrombin activity of blank (reaction buffer only) was considered as 1 and the activities of other tests were normalized based on comparisons with blank. Values are expressed as mean ± standard deviation of data from triplicate experiments.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Effects of various metal ions on ability of hyaluronic acid (HA) to inhibit the activity of antithrombin (AT). HA (1000 μg/ml) and antithrombin (150 μg/ml) were incubated with various concentrations of CaCl~2~, FeCl~3~, KCl, MgCl~2~, or NaCl. Thrombin activity of blank (reaction buffer only) was considered as 1 and the activities of other tests were normalized based on comparisons with blank. Values are expressed as mean ± standard deviation of data from triplicate experiments.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Proposed mechanism of involvement of hyaluronic acid (HA) and peptidylarginine deiminase (PADI4) in the pathogenesis of rheumatoid arthritis. VEGF, vascular endothelial growth factor.
:::
:::
|
PubMed Central
|
2024-06-05T03:55:54.953576
|
2005-1-11
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065325/",
"journal": "Arthritis Res Ther. 2005 Jan 11; 7(2):R268-R273",
"authors": [
{
"first": "Xiaotian",
"last": "Chang"
},
{
"first": "Ryo",
"last": "Yamada"
},
{
"first": "Kazuhiko",
"last": "Yamamoto"
}
]
}
|
PMC1065326
|
Introduction
============
Although randomized clinical trials have confirmed the advantage of cyclo-oxygenase (COX)-2 inhibitors over traditional nonselective nonsteroidal anti-inflammatory drugs (NSAIDs) with respect to gastrotoxicity \[[@B1]-[@B8]\], a large number of spontaneous reports have incriminated COX-2 inhibitors \[[@B9]\]. Numerous editorials and letters have been published that question the safety of these products \[[@B10]-[@B17]\]. The randomized clinical trial is the design best suited to determine drug efficacy, but it is inadequate for the evaluation of effectiveness, which applies to heterogeneous patient populations and patterns of drug use observed in a real life setting. In addition to pharmacological differences across products, the dosages used for the various indications \[[@B18]\] and past experience with the drug (through the \'depletion of susceptibles\' effect) \[[@B19]\] account for differences in the risk of adverse effects. In an observational setting, such as postmarketing surveillance, the decision to prescribe one product over another is influenced by the characteristics of the patient, the prescriber and the health care system \[[@B20]\]. In the absence of randomization, it is consequently of utmost importance, when comparing the risks associated with individual drug classes, to determine whether the patient populations are indeed comparable.
The present study was conducted to compare the prevalence of selected risk factors for upper gastrointestinal bleeding among patients prescribed COX-2 inhibitors with those among patients prescribed traditional nonselective NSAIDs, and to compare the characteristics of patients prescribed celecoxib and rofecoxib, which are the two COX-2 inhibitors marketed in Canada at the time of the study.
Methods
=======
Design
------
A case--control analysis was conducted in which the prevalence of selected gastrointestinal risk factors and medical histories of patients prescribed COX-2 inhibitors (the cases) were compared with those of users of traditional nonselective NSAIDs (the controls).
Setting
-------
The study involved prescriptions acquired through community pharmacies by members of the Quebec public drug program. Identification of eligible patients and acquisition of study variables were conducted via linkage with four administrative health care databases containing information on beneficiaries, health professionals, pharmaceutical services and medical services.
Study population
----------------
The study targeted all ambulatory adult residents (aged 18 years or older) of the province of Quebec who were members of the public drug coverage program. In Quebec, coverage of prescribed medications was universal for all elderly residents (those aged 65 years or older) regardless of income as well as for all welfare recipients. The program was broadened in 1997 to include patients who do not have access to a private insurance program regardless of age. For everyone, the program now includes a deductible payment and a co-payment, with a monthly premium that depends on the beneficiary\'s income. In practice, the program includes the following segments of the population: the great majority of community-dwelling elderly persons (\>94%), welfare recipients and patients younger than 65 years who do not have access to private insurance (e.g. the self-employed).
A sample of 100,000 drug plan members who received at least one celecoxib or rofecoxib prescription between 1 January and 31 December 2000 was randomly selected. A sample of 60,000 nonselective NSAID users was also randomly selected during the same time period, and patients who used low-dose aspirin (acetylsalicylic acid \[ASA\] ≤325 mg/day) only were excluded from the comparison group. The study population included both new (incident) users and longer time (prevalent) users. The status of patients with respect to being a user of COX-2 inhibitor or nonselective NSAID was determined at the end of the study year. Patients who had received both a COX-2 inhibitor and a nonselective NSAID were considered to be COX-2 inhibitor users. The index date was defined as the date of first dispensation of a COX-2 inhibitor or, for the traditional nonselective NSAID group, the date of the first dispensation of a nonselective NSAID.
The following inclusion criteria were applied: participants were required to have been a resident of Quebec for at least 2 years before the index date; and they were required to have had continuous coverage of medical and pharmaceutical services for at least 2 years before the index date. These criteria were verified through the beneficiary database.
Study variables
---------------
The dependent variable was the prescription of COX-2 inhibitors (celecoxib or rofecoxib) or traditional nonselective NSAIDs. The independent variables were selected risk factors for upper gastrointestinal bleeding: patient demographic characteristics (age, sex); prescribed dosage; concomitant use of corticosteroids or anticoagulants; history of gastropathy (using four indicators: prior diagnosis of gastropathy, history of upper gastrointestinal procedures, prior dispensation of gastroprotective agents, and prior referral to a gastroenterologist); and prior history of NSAID use. Comparisons were controlled for prescriber specialty, patient overall health status (using the Chronic Disease Score \[CDS\]) \[[@B21]\], income level, past use use of low-dose ASA, acetaminophen, antidepressants and benzodiazepines, and time period.
Risk factors for gastrointestinal events
----------------------------------------
Patient demographic characteristics included age, sex and income level, which were sought from the beneficiary database. For reasons of confidentiality, only age on 1 July 2000 was available. Income level was indirectly derived from the type of coverage (amount of deductible payment and co-payment), which was assigned to the patient based on their income.
History of gastropathy was assessed during the year before the index date through the presence of a diagnosis consistent with upper gastrointestinal bleeding in the medical services database. When present, this diagnosis was found to be reliable \[[@B22]\]. However, because it is not mandatory for the physician to be reimbursed, it is often missing. Consequently, three other markers were used: presence of an upper gastrointestinal procedure (e.g. gastroscopy, radiological examination) in the medical database; prior referral to a gastroenterologist, using physician specialty in the medical database; and prior dispensation of gastroprotective agents in the prescription database. Prescribed daily dosage of the COX-2 inhibitors and the traditional nonselective NSAIDs was derived from the dose per unit, quantity dispensed and prescribed duration. Daily dosages were subsequently categorized into low, standard and high, (for each product the dosage thresholds are listed in Table [1](#T1){ref-type="table"}). Standard dosages were the recommended anti-inflammatory dosages. The threshold for low-dose corresponded to the maximum approved over-the-counter dosage, or, for products available on perscription only, dosages below the recommended prescribed anti-inflammatory dosage. High dosages were those above the maximum recommended anti-inflammatory dosage.
Details regarding the dispensation of acetaminophen, low-dose ASA, corticosteroids (excluding asthma-related drugs) and anticoagulants during the year before the index date were obtained from the prescription database. Past use of NSAIDs was assessed through records of the dispensation of these agents during the year before the index date. Patterns of use were defined using three categories of recency (last dispensation ≤1 month, \>1 to 3 months, and \>3 to 12 months before the index date). For recent users, two categories of duration of use were obtained: chronic (defined as at least one dispensation in each quarter of the previous year) and nonchronic (defined as less than one dispensation in each quarter).
Covariables
-----------
Other variables may influence the prescription of NSAIDs and could act as confounders if they are also associated with risk factors for gastrointestinal events. Patient overall health status was assessed through records on medications dispensed during the year before the index date using the CDS \[[@B21]\]. Scores are weighted according to the number of different chronic diseases under treatment and the severity of the diseases. The CDS has been found to predict subsequent mortality and hospitalization rates. Because health status at the index date was the variable most likely to influence the physician\'s prescription, dispensing data for the year before were used for the calculation. Based on the distribution of scores, four categories were defined: 0, 1--4, 5--9 and ≥10. In addition, prescriptions of antidepressants and benzodiazepines were also considered to confirm the findings of a previous unpublished study that demonstrated an association between antidepressant and benzodiazepine use and prescription of COX-2 inhibitors. Prescriber specialty at the index date was determined from the prescription database.
Index dates were categorized into three time periods during the study year in order to account for differences in the date of entry of COX-2 inhibitors into the formulary of reimbursed medications (July 1999 and April 2000 for celecoxib and rofecoxib, respectively). The time periods considered were January--June, July--September and October--December 2000.
Statistical analysis
--------------------
The strength of the association between each patient characteristic and prescribed drug class was measured using odds ratios. The concomitant effect of patient characteristics was examined using multivariate logistic regression. Three models were used: COX-2 inhibitors as a class versus traditional nonselective NSAIDs, celecoxib versus traditional nonselective NSAIDs, and rofecoxib versus traditional nonselective NSAIDs. All data were analyzed using the SAS statistical package (SAS versions 6.12 and 8.0 for Windows; SAS Institute Inc., Cary, NC, USA). The level of statistical significance was set at 0.05 and the statistical uncertainty of the estimates was assessed using 95% confidence intervals.
Ethical considerations
----------------------
No patient or physician identifiers were provided to the researchers; only scrambled identifiers were used throughout the study. The study was approved by the Université de Montréal Health Sciences Ethics Committee.
Results
=======
After applying the selection criteria, 42,422 celecoxib, 25,674 rofecoxib and 12,418 traditional non-selective NSAID users were identified for the study. The characteristics of the study population are presented in Table [2](#T2){ref-type="table"}. Because of the very large sample size, all differences were statistically significant and therefore *P*values are not reported. Patients treated with celecoxib were on average slightly older than those treated with rofecoxib or traditional nonselective NSAIDs, and a larger proportion of women were treated with COX-2 inhibitors as opposed to traditional nonselective NSAIDs. For each of the four indicators of prior history of gastropathy, there was a larger proportion of COX-2 inhibitor users with a positive history as compared to nonselective NSAID users. For all indicators used, the proportion was also greater for celecoxib than for rofecoxib. Very few patients had used anticoagulants during the year before the index date, but again the prevalence of use was greater for COX-2 inhibitors than for traditional nonselective NSAIDs.
Using the data presented in Table [2](#T2){ref-type="table"}, we were able to determine that, overall, very few patients had used a nonselective NSAID for the first time during the month before the index date. The proportion of patients who had received their last NSAID prescription in the distant past (between 3 and 12 months before index date) was greater for celecoxib than for rofecoxib. Of the patients treated with rofecoxib, 72.8% had not received any NSAIDs during the prior year, which means that it was often used as a first treatment obtained under prescription. This proportion was lower for celecoxib (63.7%) and traditional nonselective NSAIDs (55.2%). Only 6.3% of rofecoxib users had received their last NSAID prescription between 1 and 3 months before the index date, as compared with 7.8% among celecoxib users and 15.7% among nonselective NSAID users.
The great majority of NSAIDs were prescribed by general practitioners (85.9% of traditional nonselective NSAIDs, 85.3% of celebrex and 88.3% of rofecoxib prescriptions). Dosage levels were highly heterogeneous across products. A large proportion of traditional nonselective NSAIDs were prescribed at dosages lower than those recommended for anti-inflammatory indications (22.1%) in comparison with celecoxib (3.4%) and rofecoxib (18.2%). Conversely, the majority of COX-2 inhibitors were prescribed at standard anti-inflammatory dosages (65.3% of celecoxib and 73.0% of rofecoxib prescriptions). A relatively high proportion of COX-2 inhibitors, especially celecoxib, were prescribed at dosages in excess of standard recommendations (31.2% of celecoxib and 8.8% of rofecoxib prescriptions). There was a strong correlation between dosage and age. For all products, the proportion of low dosages increased with age, and conversely the proportion of high dosages decreased with age (data not shown). This relationship was also found for overall health status; the higher the CDS, the higher was the proportion of prescriptions for low dosages (30.4% of all prescriptions were of low dosages for patients with a CDS 10+ versus 14.5% for those with a CDS of 0).
Results of the multivariate logistic regression are presented in Table [3](#T3){ref-type="table"}. Increasing age and female sex were both associated with greater likelihood of receiving a COX-2 inhibitor. Compared with patients aged 18--54 years, older patients were more likely to receive a COX-2 inhibitor, but this association was greatly confounded by dosage category. Income level marginally influenced the choice of product; patients with lower income favoured the less costly traditional nonselective NSAIDs. According to crude odds ratio estimates, there was a positive association between each indicator of history of gastropathy and the probability of receiving a COX-2 inhibitor. However, when all the indicators were fitted simultaneously in the multivariate model, a history of gastrointestinal procedures was no longer significant; this finding is probably attributable to correlation between the various indicators. The analyses revealed an association between the CDS scores and the probability of receiving a COX-2 inhibitor, although no trend was observed.
Use of acetaminophen, corticosteroids, anticoagulants, antidepressants and benzodiazepines during the year before the index date were all associated with the prescription of COX-2 inhibitors. On the other hand, patients who had received low-dose ASA during the previous year were more likely to receive a traditional nonselective NSAIDs than a COX-2 inhibitor. Specialists were less likely to prescribe a COX-2 inhibitor than were general practitioners.
Results from the multivariate logistic regression models specific for celecoxib and rofecoxib are presented in Table [4](#T4){ref-type="table"}. As shown, the strength of the association with gastrointestinal risk factors was significantly greater for celecoxib than for rofecoxib for age, past use of NSAIDs between 1 and 3 months before the index date, and recent chronic NSAID use. Point estimates of odds ratio for sex, other patterns of NSAID use, prior dispensation of gastroprotective agents, prior referral to a gastroenterologist, prior gastrointestinal procedures, prior use of antidepressants and benzodiazepines, and anticoagulants were greater for celecoxib than for rofecoxib, but the difference was not significant.
Because rofecoxib was only included in the list of reimbursed medications in April 2000, it was not available for half of the first time period, which explains its lower likelihood of being prescribed than nonselective NSAIDs (odds ratio 0.24, 95% confidence interval 0.22--0.26). However, for the second period (July--Sept) there was no significant difference between rofecoxib and celecoxib.
Discussion
==========
This study provides empirical evidence that channeling exists in the prescription of COX-2 inhibitors. Patients with risk factors for gastropathy were more likely to receive a COX-2 inhibitor than a traditional nonselective NSAID. Age, sex and history of gastropathy are well known independent risk factors for gastrointestinal bleeding, and it is therefore not surprising that they influenced prescribing practices. The effect of sex may be explained by greater use of over-the-counter NSAIDs in the past, not recorded in the databases, for the treatment of dysmenorrhoea. The effect of corticosteroids and anticoagulants is also not surprising, given that these drugs represent contraindications to the prescription of traditional nonselective NSAIDs. These findings are consistent with those obtained in a recent study conducted in a UK primary care setting \[[@B23]\] but they contradict those reported in a elderly Medicare population in the USA \[[@B24]\]. In the latter study it appeared that there was over-treatment with COX-2 inhibitors in patients without risk factors, and under-treatment in patients who had at least one risk factor. The effect of past NSAID use is more difficult to interpret because of the lack of data regarding reasons for discontinuation of NSAIDs. Although past NSAID use has been found to be associated with decreased incidence of gastrointestinal bleeding, the impact that such a \'depletion of susceptibles\' effect may have on prescribing practices remains to be clarified. Regardless of the underlying mechanism, it can be concluded from these results that past NSAID use is likely to confound risk comparisons across drug classes because it is an independent risk factor for gastrointestinal problems as well as influencing prescribing practices.
Patients who had received acetaminophen in the past were more likely to switch to a COX-2 inhibitor than to a traditional nonselective NSAIDs. Patients who had received antidepressants and benzodiazepines were also more likely to receive a COX-2 inhibitor than a traditional nonselective NSAID. This empirical finding is difficult to interpret. It may be hypothesized that physicians may be more likely in general to prescribe newer agents to patients who are anxious. More studies are needed to explore further the interaction between patients and physicians in order to elucidate this issue.
Although there was an association between physician specialty and prescription of COX-2 inhibitors or traditional nonselective NSAIDs, the results did not confound the associations between patient characteristics and prescription practices.
Results for patient overall health status and prescribing practices were highly confounded by dosage. This suggests that, for the sickest patients, prescribing practices are largely determined by dosage rather than by drug class. Patients with a high level of comorbidity still receive traditional nonselective NSAIDs but at lower dosages. Such findings are likely to be time-sensitive because COX-2 inhibitors were just introduced into the Canadian market during the study period, and there might have been reluctance in the medical community to prescribe newer agents to sicker patients.
Comparisons between the two COX-2 inhibitors indicated that for several risk factors under investigation the channeling process is stronger for celecoxib than rofecoxib. However, these findings should be interpreted with caution because for several of the risk factors investigated the differences between products were not statistically significant. On the other hand, celecoxib was not always at a disadvantage; past chronic NSAID use, which, according to the depletion of susceptibles effect, places patients at a lower risk for upper gastrointestinal events \[[@B19]\], was associated with a greater probability of being prescribed celecoxib than rofecoxib.
Many risk factors for gastrointestinal bleeding could not be ascertained in this study, such as smoking status and alcohol use, which are known risk factors for gastrointestinal events and have also been found to influence prescribing practices \[[@B23]\]. Also, there were no data on indications but we controlled for dosage, which, according to Griffin and coworkers \[[@B18]\], is more likely to influence the risk of gastrointestinal bleeding than indication *per se*. Dosage had a very large impact on the results, and its exclusion would have produced spurious differences across products. There were no data on over-the-counter use of NSAIDs such as aspirin and ibuprofen. Therefore, it was not possible to explore the concomitant use of nonprescribed NSAIDs. Finally, data are generalizable only to recent use (in the previous year). We were not able to explore the impact of more distant history. Nevertheless, the use of retrospective data obtained from administrative databases allowed us to examine the various associations in a truly observational setting without influencing prescribing practices in any way. In addition, the large sample size allowed us to conduct comparisons across individual products.
Conclusion
==========
Our results provide empirical evidence that the introduction of a new class of medications into the market results in the channeling of patients at high risk for adverse effects. However, as shown by the present study, differences across individual products cannot be predicted from their order of entry into the formulary. Other factors, such as marketing strategies, play a major role as well. Neverthless, one may conclude that selective prescribing results in a positive association between risk factors and drug use, which could confound risk comparisons across products.
Abbreviations
=============
ASA = acetylsalicylic acid; CDS = chronic disease score; COX = cyclo-oxygenase; NSAIDs = nonsteroidal anti-inflammatory drugs.
Competing interests
===================
This study was funded through an unrestricted grant from Pharmacia Corporation. YM was a paid consultant for this study. JFB, TD, NM and SP declare that they have no competing interests. SZ was an employee of Pharmacia Corporation at the time the study was conducted.
Authors\' contributions
=======================
YM, as principal investigator of the study, designed and coordinated the study, interpreted study results, and wrote the manuscript. TD conducted the statistical analyses. JFB participated in the design of the strategy for the sampling of the study population and helped to draft the manuscript. NM assisted in the conduct of the statistical analyses and contributed to the interpretation of the study results. SP assisted in the review of the literature and determined the relevance of the study. SZ conceived the study and participated in its design. All authors read and approved the final manuscript.
Acknowledgements
================
We are grateful to Mr Jacques Barry and all other members of the Department of Statistical Services at the Régie de l\'assurance-maladie du Québec for providing us with the necessary data for this study. We also wish to thank Drs Rajaa Lagnaoui and Ghada Salamé-Miremont for their methodological contribution.
Figures and Tables
==================
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Dosage categories for each product
:::
Generic name Low dosage (mg/day) Standard dosage (mg/day) High dosage (mg/day)
------------------------------------------------------- --------------------- -------------------------- ----------------------
Celecoxib ≤100^a^ \>100 to 200 \>200
Rofecoxib \<25 25 to \<50 ≥50
Acetylsalicylic acid ≤1300 \>1300 to \<4000 ≥4000
Diclofenac (including Voltaren + Cytotec = Arthrotec) ≤50 \>50 to 100 \>100
Diflunisal ≤500 \>500 to 1000 \>1000
Etodolac ≤300 \>300 to 900 \>900
Fenoprofen \<1800 1800 to 2400 \>2400
Flurbiprofen ≤50 \>50 to 200 \>200
Ibuprofen \<1000 1000 to 1200 \>1200
Indomethacin ≤50 \>50 to 100 \>100
Ketoprofen ≤50 \>50 to 200 \>200
Mefenamic Acid \<750 750 to 1000 \>1000
Naproxen ≤550 \>550 to 1100 \>1100
Piroxicam ≤10 \>10 to 20 \>20
Salsalate ≤500 \>500 to 1000 \>1000
Tiaprofenic Acid ≤200 \>200 to 600 \>600
Tolmetin ≤600 \>600 to 1200 \>1200
^a^According to our references, 100 mg celecoxib would be considered a standard dose. However, because none of the patients were prescribed lower dosages, we included 100 mg as a low dose (in order to avoid a 0 cell).
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Characteristics of the study population
:::
Traditional nonselective NSAIDs (n = 12,418) Celecoxib (*n*= 42,422) Rofecoxib (*n*= 25,674)
----------------------------------------------- ---------------------------------------------- ------------------------- -------------------------
Age (years)
18--34 18.6 3.7 6.7
35--44 15.8 6.5 9.7
45--54 13.2 9.9 12.1
55--64 13.5 15.9 16.4
65--74 18.8 30.5 27.1
75--84 15.1 27.1 22.7
85+ 4.1 6.5 5.1
Sex
Female 55.4 67.4 65.5
Male 44.6 32.6 34.5
Income level
Low 14.8 10.9 11.4
Nonlow 85.2 89.1 88.6
Dosage category
High 11.6 31.2 8.8
Standard 66.3 65.3 73.0
Low 22.1 3.4 18.2
Prior diagnosis of gastropathy 3.6 7.7 5.0
Prior gastrointestinal procedures 2.0 4.5 2.7
Prior dispensation of gastroprotective agents 14.0 29.9 24.3
Prior referral to a gastroenterologist 2.8 6.0 3.9
History of NSAID use
Recent, first time 3.0 2.2 2.8
Recent, chronic 6.6 4.3 4.2
\>1 to 3 months 15.7 7.8 6.3
\>3 to 12 months 19.4 22.0 14.0
No use 55.2 63.7 72.8
Anticoagulants 1.3 3.3 3.0
Corticosteroids 11.7 19.6 16.9
Benzodiazepines 23.3 38.2 33.0
Antidepressants 10.5 17.2 15.7
Chronic Disease Score
≥10 5.2 10.3 8.2
5--9 19.6 28.6 25.4
1--4 27.9 33.6 32.6
0 47.3 27.6 33.8
Prescriber specialty
General practitioner 85.9 85.3 88.3
Cardiology 1.0 0.4 0.3
Internal Medicine 2.3 3.3 2.1
Neurology 0.3 0.3 0.2
General surgery 1.4 0.8 1.0
Obstetrics/gynaecology 1.5 0.2 0.3
Orthopaedic surgery 1.2 3.4 3.2
Rheumatology 2.5 2.5 1.7
Other 2.8 3.8 2.9
Values are expressed as percentages. NSAID, nonsteroidal anti-inflammatory drug.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Multivariate analysis of the factors associated with dispensation of selective COX-2 inhibitors versus traditional nonselective NSAIDs
:::
Crude OR (95% CI) Adjusted OR (95% CI)
----------------------------------------------- ------------------- ----------------------
Age group (years)
75+ 3.17 (3.01--3.34) 4.22 (3.95--4.51)
55 to ≤74 2.86 (2.74--2.99) 3.23 (3.06--3.40)
18--54 Reference Reference
Female sex 1.61 (1.52--1.66) 1.52 (1.45--1.58)
Income level (lower) 0.72 (0.68--0.76) 0.90 (0.85--0.96)
Prior diagnosis of gastropathy 1.93 (1.75--2.11) 1.21 (1.08--1.36)
Prior gastrointestinal procedures 1.94 (1.71--2.22) 1.09 (0.94--1.27)
Prior dispensation of gastroprotective agents 2.37 (2.25--2.50) 1.57 (1.47--1.67)
Prior referral to gastroenterologist 1.89 (1.69--2.12) 1.23 (1.08--1.39)
Prior history of NSAID use
Recent, first time 2.39 (2.14--2.66) 2.17 (1.93--2.43)
Recent, chronic 1.58 (1.46--1.70) 1.21 (1.11--1.32)
\>1 to 3 months 1.12 (1.06--1.19) 0.95 (0.89--1.01)
\>3 to 12 months 0.93 (0.88--0.98) 0.84 (0.80--0.89)
No use in past year Reference Reference
Corticosteroids 1.72 (1.62--1.82) 1.16 (1.07--1.24)
Anticoagulants 2.53 (2.15--2.98) 1.56 (1.32--1.85)
Antidepressants 1.69 (1.59--1.80) 1.37 (1.28--1.46)
Benzodiazepines 1.87 (1.78--1.95) 1.20 (1.14--1.26)
Acetaminophen 1.85 (1.76--1.94) 1.41 (1.34--1.49)
Low dose ASA 0.85 (0.81--0.89) 0.56 (0.52--0.59)
Chronic Disease Score
10+ 2.88 (2.65--3.14) 1.26 (1.13--1.41)
5--9 2.21 (2.10--2.32) 1.28 (1.20--1.37)
1--4 1.88 (1.80--1.97) 1.26 (1.19--1.33)
0 Reference Reference
Specialist (versus GP) 0.99 (0.94--1.04) 0.89 (0.84--0.94)
Dosage
High dose 1.91 (1.80--2.03) 2.19 (2.06--2.33)
Low dose 0.40 (0.38--0.42) 0.29 (0.27--0.30)
Standard dose Reference Reference
Time period
January--June 0.72 (0.68--0.76) 0.54 (0.51--0.57)
July--Sept 1.02 (0.95--1.09) 0.99 (0.92--1.06)
October--December Reference Reference
All covariables included simultaneously in the models are listed in this table; models were not adjusted for any other factors. ASA, acetylsalicylic acid; CI, confidence interval; COX, cyclo-oxygenase; GP, general practitioner; NSAID, nonsteroidal anti-inflammatory drug; OR, odds ratio.
:::
::: {#T4 .table-wrap}
Table 4
::: {.caption}
######
Multivariate analysis of the factors associated with dispensation of celecoxib and rofecoxib versus traditional nonselective NSAIDs
:::
Celecoxib Rofecoxib
----------------------------------------------- ------------------- -------------------
Age group
75+ 5.34 (4.96--5.75) 3.06 (2.83--3.30)
55 to ≤74 3.65 (3.45--3.87) 2.62 (2.46--2.78)
18--54 Reference Reference
Female sex 1.55 (1.47--1.62) 1.45 (1.38--1.52)
Income level (lower) 0.95 (0.88--1.02) 0.83 (0.77--0.89)
Prior diagnosis of gastropathy 1.21 (1.07--1.37) 1.11 (0.97--1.27)
Prior gastrointestinal procedures 1.21 (1.02--1.42) 0.97 (0.81--1.16)
Prior dispensation of gastroprotective agents 1.59 (1.48--1.71) 1.51 (1.41--1.63)
Prior referral to gastroenterologist 1.27 (1.11--1.45) 1.19 (1.03--1.37)
Prior history of NSAID use:
Recent, first time 2.25 (1.99--2.54) 2.02 (1.79--2.29)
Recent, chronic 1.68 (1.52--1.85) 0.75 (0.68--0.84)
\>1 to 3 months 1.36 (1.26--1.45) 0.48 (0.44--0.52)
\>3 to 12 months 0.85 (0.80--0.91) 0.81 (0.76--0.87)
No use in past year Reference Reference
Corticosteroids 1.11 (1.02--1.20) 1.20 (1.11--1.31)
Anticoagulants 1.61 (1.34--1.94) 1.48 (1.23--1.78)
Antidepressants 1.38 (1.28--1.48) 1.37 (1.27--1.48)
Benzodiazepines 1.18 (1.12--1.25) 1.15 (1.09--1.22)
Acetaminophen 1.39 (1.31--1.48) 1.37 (1.28--1.45)
Low-dose ASA 0.67 (0.62--0.71) 0.58 (0.54--0.62)
Chronic Disease Score
10+ 1.26 (1.11--1.43) 1.20 (1.06--1.37)
5--9 1.28 (1.19--1.38) 1.23 (1.14--1.33)
1--4 1.25 (1.17--1.33) 1.26 (1.18--1.34)
0 Reference Reference
Physician specialty 0.96 (0.90--1.03) 0.83 (0.77--0.89)
Dosage
High dose 3.36 (3.15--3.58) 0.76 (0.70--0.82)
Low dose 0.09 (0.09--0.10) 0.73 (0.69--0.78)
Standard dose Reference Reference
Time period
January--June 1.25 (1.16--1.34) 0.24 (0.22--0.26)
July--September 1.09 (1.00--1.19) 0.93 (0.87--1.01)
October--December Reference Reference
ASA, acetylsalicylic acid; NSAID, nonsteroidal anti-inflammatory drug.
:::
|
PubMed Central
|
2024-06-05T03:55:54.955181
|
2005-1-17
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065326/",
"journal": "Arthritis Res Ther. 2005 Jan 17; 7(2):R333-R342",
"authors": [
{
"first": "Yola",
"last": "Moride"
},
{
"first": "Thierry",
"last": "Ducruet"
},
{
"first": "Jean-François",
"last": "Boivin"
},
{
"first": "Nicholas",
"last": "Moore"
},
{
"first": "Sylvie",
"last": "Perreault"
},
{
"first": "Sean",
"last": "Zhao"
}
]
}
|
PMC1065327
|
Introduction
============
Cartilage degradation in osteoarthritis (OA) and rheumatoid arthritis constitutes a major structural change in the joint, which may severely impair its function and cause pain and disability. This degradation is accompanied by the release in the synovial fluid of degraded matrix constituents that primarily result from an increased matrix catabolism \[[@B1]\]. Various factors are directly involved in this process. Endothelin-1 (ET-1), a potent vasoconstrictor and promitogen peptide for many cell types, including chondrocytes, was recently identified as one such factor \[[@B2],[@B3]\].
ET-1 binds to the specific endothelin A or endothelin B receptors expressed on chondrocytes \[[@B4]\] and triggers a cascade of intracellular events, including phospholipase C activation \[[@B5]\], an increase in intracellular calcium \[[@B6],[@B7]\], prostaglandin production \[[@B8]\] and nitric oxide (NO) release \[[@B9]\]. The effect of ET-1 on DNA and protein synthesis in chondrocytes is biphasic. The potent initial stimulatory effect of ET-1 decreases progressively with time and is followed by an inhibition \[[@B3],[@B8]\]. The inhibitory effect seems to be mediated by NO and cGMP, both produced in response to ET-1 stimulation \[[@B8],[@B9]\]. Additionally, we have recently demonstrated that ET-1 is significantly increased locally in OA cartilage and synovial membrane when compared with normal tissues. In OA cartilage, ET-1 is involved in cartilage catabolism through metalloprotease (MMP) regulation and the induction of type II collagen breakdown \[[@B2]\].
MMPs are a family of structurally related zinc-dependent neutral endopeptidases classified into subgroups of collagenases, gelatinases, stromelysins, membrane-type MMPs and other MMPs \[[@B10]\]. When activated, MMPs degrade a broad spectrum of substrates, including collagens and other matrix macromolecules. As a whole, MMPs play an important role in the extracellular matrix remodelling that occurs under physiological and pathological conditions. Among all the MMPs, we have recently demonstrated an induction in the synthesis, secretion and activation of two collagenases (MMP-1 and MMP-13) by ET-1 \[[@B2]\]. These MMPs play an active role in the progression of OA pathology as they are the most effective at initiating collagen destruction during the inflammatory process and the remodelling phase of the disease \[[@B11],[@B12]\].
Another deleterious agent in joint cartilage is the NO radical \[[@B13],[@B14]\], which downregulates DNA \[[@B8]\] and matrix synthesis \[[@B14]\] and upregulates matrix degradation via increased MMP synthesis \[[@B15]\]. Indeed, inhibition of NO production was shown to slow down the progression of OA. It has been demonstrated that, *in vitro*, NO could also upregulate MMP synthesis and activity in joint chondrocytes and cartilage \[[@B15]\]. *In vivo*in an OA animal model, selective inhibition of the inducible nitric oxide synthase (iNOS) provides a protective effect on OA joint tissues more specifically in regard to the degradation of the extracellular matrix and the downregulation of MMP \[[@B16]\].
The aim of the present study was to further investigate the role of ET-1 in human OA chondrocytes, focusing on NO, MMP-1 and MMP-13 production as well as the relevant signalling pathways activated by ET-1 in human OA chondrocytes in regard to these factors.
Materials and methods
=====================
Specimens
---------
Human cartilage was obtained with the consent of 12 OA patients (mean ± standard error of the mean age, 58 ± 6 years) undergoing total knee replacement. The Institutional Ethics Committee Board of Notre Dame Hospital in Montreal, Canada approved the study protocol. Tissue specimens were embedded in paraffin, were sectioned and stained with Safranin O and fast green, and were evaluated using the Mankin histological/histochemical scale \[[@B17]\]. Only tissues corresponding to a moderate degree of OA severity (Mankin 3--7) were included in this study. Cartilage was sectioned from the tibial plateaus, rinsed and finely chopped, and the cells released by enzymatic digestion performed as previously described \[[@B2],[@B11]\]. The cells were seeded in culture flasks at the density of 10^4^cells/cm^2^and were grown to confluence in DMEM (Gibco BRL, Burlington, ON, Canada) containing 10% heat-inactivated FCS (Hyclone, Logan, UT, USA) and 1% penicillin/streptomycin (Gibco BRL). Only first-passage-cultured cells were used.
MMP-1 and MMP-13 quantification
-------------------------------
MMP-1 and MMP-13 protein levels were determined in the culture media using specific ELISA assays. The ELISA assay (Amersham Biosciences Corp., Baie d\'Urfé, QC, Canada) for MMP-1 specifically detected the total human MMP-1 (i.e. active MMP-1, the latent enzyme and the enzyme complexed with inhibitors such as tissue inhibitor of matrix metalloproteinases 1). The sensitivity of this assay is 1.7 ng/ml, and there is no significant cross-reactivity or interference with MMP-3, MMP-2 and MMP-9. The MMP-13 ELISA assay (R&D Systems Inc., Minneapolis, MN, USA) is a monoclonal polyclonal-based assay specific for both the active and latent MMP-13. Its sensitivity is 0.032 ng/ml, and there is no cross-reactivity with MMP-1, MMP-2, MMP-3, MMP-7, MMP-8, MMP-9 and MT1-MMP. Results are expressed as nanograms per 5 × 10^5^cells.
The effect of ET-1, protein kinase inhibitors and a guanylate cyclase inhibitor (LY83583) on MMP-1, MMP-13 and NO production
----------------------------------------------------------------------------------------------------------------------------
MMP-1 production, MMP-13 production and NO production were studied in the absence of and in the presence of ET-1, using various inhibitors: 1 μM SB 202190 (inhibitor of p38 mitogen-activated protein \[MAP\] kinase), 10 μM PD 98059 (a selective mitogen-activated protein kinase kinase 1/2 \[MEK1/2\] inhibitor), 100 nM Wortmannin (a phosphatidyl inositol 3 kinase inhibitor), 4 μM KT5720 (a protein kinase A \[PKA\] inhibitor), or 2 μM LY83583 (an inhibitor of NO-dependent soluble guanylate cyclase inhibitor). All inhibitors were purchased from Calbiochem EDM Biosciences Inc. (San Diego, CA, USA), and the active concentrations chosen are based on the literature or were assayed in preliminary experiments \[[@B18],[@B19]\]. ET-1 was purchased from (Sigma-Aldrich, Oakville, ON, Canada). Confluent OA chondrocytes were preincubated for 30 min with these inhibitors and then 10 nM ET-1 was added for 24 hours. Following incubation, the MMP-13 and MMP-1 protein levels and NO levels were determined in the media of six independent cultures as described in the following.
NO determination
----------------
Nitrite (NO~2~^-^), a stable end product of NO, was measured in the media of cultured cells using a spectrophotometric method based on the Griess reaction \[[@B20]\]. To examine the effects of ET-1 on NO production, a dose--response curve was performed by incubating OA chondrocytes for 24 hours with increased concentrations (0--100 nM) of ET-1, or by pretreating with protein kinase inhibitors or a guanylate cyclase inhibitor and ET-1 as already described. NO production was also evaluated in the presence of the iNOS inhibitor L-NIL (L-N^6^(1-iminoethyl)lysine) (Calbiochem EDM Biosciences Inc.). Chondrocytes were preincubated for 30 min with 0--50 μM L-NIL and were then incubated for 24 hours with 10 nM ET-1. The media were collected and the released NO levels were determined. Results are expressed as nanomoles per 5 × 10^5^cells ± standard error of the mean or as a percentage of the control cultures.
Western blot
------------
Confluent OA chondrocytes were incubated in the presence of or in the absence (control) of 10 nM ET-1, and the cells were lysed in 0.2 ml lysis buffer (25 mM HEPES, 5 mM MgCl~2~, 1 mM EDTA, 1 mM PMSF, 10 μg/ml pepstatin, 10 μg/ml leupeptin, pH 7.5). The protein concentration of the lysate was determined with the Bradford dye assay (Bio-Rad Laboratories, Hercules, CA, USA). For western blot, 10 μg lysate protein was separated by electrophoresis on a 10% SDS discontinuous gradient polyacrylamide gel. Separated proteins were then transferred electrophoretically onto a nitrocellulose membrane (Hybond C extra; Amersham, Pharmacia Biotech, Chalfont St Giles, UK). The membranes were immersed overnight in the Super Block Blocking buffer (Pierce, Rockford, IL, USA), rinsed and incubated for 24 hours at 4°C with one of the mouse monoclonal primary antibodies (New England Biolabs, Mississauga, ON, Canada) specifically recognizing phosphorylated p38 or total p38 (dilution, 1/1000), phosphorylated p44/42 (dilution, 1/5000), phosphorylated Akt (dilution, 1/2000), phosphorylated stress-activated protein kinase/Jun-N-terminal kinase (SAP/JNK) (dilution, 1/1000), or actin C-terminal fragment (dilution, 1/5000). iNOS was detected with a rabbit polyclonal antibody (dilution, 1/1000; Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA).
Following incubation with primary antibody, membranes were carefully washed and reincubated for 1 hour at 4°C with a second antibody (anti-rabbit IgG). Anti-mouse horseradish peroxidase-conjugated IgG (dilution, 1/25,000) was used for the detection of the monoclonal antibody, and sheep anti-rabbit horseradish peroxidase-conjugated IgG (dilution, 1/40,000) was used for the polyclonal antibody. Detection was performed using the Super Signal Ultra Western blot chemiluminescence system (Pierce) \[[@B11]\].
Apoptosis
---------
Apoptosis was investigated in OA chondrocytes cultured on Lab-Tec chamber slides (Nalge Nunc International, Naperville, IL, USA). At confluence, the cells were rinsed and incubated at 37°C for 72 hours in DMEM containing 2.5% heat-inactivated FCS in the absence of or in the presence of 10 nM human recombinant ET-1. Apoptotic cells were detected by *in situ*staining using the TUNEL method (Trevigen Inc., Gaithersburg, MD, USA). Both pro-apoptotic Bad and anti-apoptotic Bcl2 proteins were determined by immunocytochemical detection using specific anti-Bad and anti-Bcl2 antibodies (Upstate Biotechnology, Lake Placid, NY, USA). The results are expressed as the mean percentage of positively stained cells according to a previously published method \[[@B21],[@B22]\].
Statistical analysis
--------------------
Data are expressed as the mean ± standard error of the mean of five or six independent cultures. Statistical significance was assessed by the Mann--Whitney test, and *P*\< 0.05 was considered significant.
Results
=======
ET-1 induces MMP-1 and MMP-13 production
----------------------------------------
The effects of ET-1 and those of various inhibitors on MMP-1 production and MMP-13 production are shown in Fig. [1](#F1){ref-type="fig"}. At 10 nM ET-1 the production of both enzymes was significantly increased (*P*\< 0.005). SB202190, a p38 inhibitor, completely suppressed the ET-1-stimulated production of both enzymes, whereas the phosphatidyl inositol 3 kinase inhibitor Wortmannin and the PKA inhibitor KT5720 partially but significantly (*P*\< 0.01) decreased the level of MMP-13 only. Interestingly, the most potent inhibitor of MMP-1 and MMP-13 production was LY83583, an inhibitor of NO-dependent soluble guanylate cyclase and of cGMP. This agent not only suppressed the ET-1-induced stimulation, but also decreased the level of both enzymes below the basal level: a significant difference was found for both MMP-13 and MMP-1 when compared with the ET-1 stimulation (*P*\< 0.005) and for MMP-13 when compared with the control (*P*\< 0.05). Although a decrease in MMP-13 was noted with the MEK1/2 kinase inhibitor PD98059 at the concentration tested, it did not reach statistical significance. With this inhibitor, no effect was found on MMP-1 production.
ET-1 induces NO production
--------------------------
The effects of ET-1 on NO release and on iNOS expression are shown in Fig. [2](#F2){ref-type="fig"}. Figure [2a](#F2){ref-type="fig"} shows that ET-1 greatly stimulated NO production and was released in a concentration-dependent manner. Incubation with increasing concentrations of ET-1, from 0.1 to 100 nM, augmented almost 12-fold the linear accumulation of NO. To determine the mechanism involved in the ET-1-induced NO production, the effects of the major intracellular signalling pathways were investigated. Figure [2b](#F2){ref-type="fig"} shows that the ET-1-induced NO release was significantly inhibited by p38 inhibition and prevented by KT5720, a PKA inhibitor. No significant effect was noted for MEK1/2 inhibition by PD98059 and by Wortmannin. Moreover, the guanylate cyclase inhibitor LY83583 reduced the NO production as significant differences were found when compared with either the ET-1 stimulation (*P*\< 0.05) or with the control (*P*\< 0.05), and this inhibitor also decreased both the endogenous and ET-1-induced iNOS level (Fig. [2d](#F2){ref-type="fig"}). The ET-1-induced NO release occurs via iNOS as shown in Figure [2c](#F2){ref-type="fig"}: complete inhibition of iNOS by 50 μM allosteric iNOS inhibitor L-NIL, as expected, almost completely inhibited NO release. Figure [2d](#F2){ref-type="fig"} shows the effects of various inhibitors on iNOS expression, as determined by western blot analysis of cell extracts. The 24-hour incubation of cells with ET-1 results in an increase of iNOS protein (Fig. [2d](#F2){ref-type="fig"}, lane 2). The ET-1-induced iNOS protein expression was completely suppressed by SB202190 and LY83583, and was partially suppressed by Wortmannin and KT5720. PD98059 had no effect.
Intracellular protein kinase phosphorylation in the presence of ET-1
--------------------------------------------------------------------
Figure [3a--d](#F3){ref-type="fig"} show the effects of ET-1 on the phosphorylation of p38, Akt, p44/42 and SAP/JNK kinases as detected by western blot of cell extracts. ET-1 at 10 nM induced p38, Akt, p44/42, and SAP/JNK phosphorylation in a time-ordered manner. For p38, the maximal effect following cell exposure to ET-1 was obtained at 10 min. For Akt, the maximal effect was observed at 2 min of cell exposure and this effect persisted during 30 min, followed by a decline at 45 min. At this time (45 min), both p38 kinase and Akt phosphorylated forms were diminished. The maximal effect was obtained at 15 min for p44/42 kinase and at 45 min for SAP/JNK. The SAP/JNK phosphorylated forms were not detected at 60 min, whereas that of p44/42 decreased but was still present even at 60 min.
ET-1 did not affect apoptosis
-----------------------------
As ET-1 induces NO release and because the accumulation of NO causes apoptosis, we explored this potential effect. OA chondrocytes incubated in the absence of (control) or in the presence of ET-1 (10 nM) for 72 hours showed that ET-1 did not affect apoptosis (TUNEL reaction; data not shown) or the production of either anti-apoptotic Bcl2 or pro-apoptotic Bad proteins. A similar percentage of positively stained cells was found for Bcl2 (42.8 ± 5.1% and 43.2 ± 4.3% for the control and for ET-1, respectively) and for Bad (10.1 ± 3.8% and 9.5 ± 2.1%, respectively).
Discussion
==========
This study shows an overproduction of NO, MMP-1 and MMP-13 in human OA chondrocytes stimulated by ET-1. This result goes beyond previous results \[[@B2]\], which showed that human OA synovial tissue and joint cartilage express the ET-1 gene and overproduce ET-1, resulting in an excessive synthesis of MMP-1 and MMP-13 in the same tissues. In addition, the result goes beyond these findings and enlightens on the mechanism by which ET-1 accomplishes this action. Strong evidence was obtained for the key role played by NO, whose production and release were also upregulated by ET-1.
NO induces smooth muscle cell relaxation by activating soluble guanylate cyclase and by increasing the intracellular concentration of cGMP. LY83583 suppresses the effect of NO by inhibiting this NO-dependent production of cGMP \[[@B23]\]. In the present study, LY83583 was also shown to strongly inhibit MMP-1 and MMP-13 production by unstimulated and ET-1-stimulated OA chondrocytes, showing the key role of cGMP for the synthesis of these enzymes. This finding confirms a previous observation that cGMP is necessary for protein synthesis \[[@B9]\], and brings further evidence that an excess of NO is harmful to cells.
It is generally accepted that progressive tissue destruction in rheumatoid arthritis and in OA results from an excessive breakdown mediated by various proteolytic enzymes and other catabolic agents such as free radicals and NO \[[@B1],[@B13],[@B24],[@B25]\]. Our results suggest that ET-1 should also be added to the list of potential deleterious agents that may account for articular cartilage destruction in rheumatic diseases. The action of ET-1 seems to be dual via an increase in MMP and NO production. ET-1-induced stimulation of MMP-1 and MMP-13, as well as the induction of iNOS gene expression with subsequent NO overproduction by OA chondrocytes, may interfere with the proinflammatory cytokine pathways. Indeed, we and other workers have shown that IL-1β upregulates the synthesis of ET-1 \[[@B3]\], which in turn can induce IL-1β gene transcription and consequently the production of the protein \[[@B26]\]. We previously demonstrated \[[@B2]\] that MMP-13 expression was induced similarly by ET-1 and IL-1β; however, although they both enhanced MMP-1 expression, the effect of IL-1β was more potent on this enzyme. Interestingly, using a specific immunoassay measuring the C telopeptide of type II collagen fragments on OA cartilage explants, we also found that the level of the cleaved collagen fragments were significantly increased in the presence of both IL-1β and ET-1 with a more potent effect observed for ET-1. This could be explained by a putative synergy between ET-1 and IL-1β as ET-1 induces IL-1β and as IL-1β has a positive feedback on ET-1 synthesis \[[@B19],[@B27]\].
NO is an important signalling molecule at physiological concentrations \[[@B28]\], but when overproduced via iNOS gene activation it is toxic to cells \[[@B29]\]. NO triggers the transcription of several proinflammatory genes \[[@B28],[@B30]\], interacts with the cysteine residues of many proteins (S nitrosylation) and may alter their structure and function. In the presence of the superoxide anion, NO generates peroxynitrite and hydroxyl radicals that are cytotoxic, inducing peroxidation of lipids and damaging other molecules, such as DNA, and matrix macromolecules. This finally results in the inhibition of many cellular processes that impair the capacity of the cells to synthesize matrix macromolecules and to repair damaged tissue \[[@B8],[@B31]\].
In addition to the findings already discussed, the present study sheds more light on the major signalling pathways involved in the ET-1-induced MMP-1 and MMP-13 production and in NO production. In OA chondrocytes, ET-1 seems to stimulate the production of these enzymes through activation of, at least, two kinases, p38 MAP kinase and PKA. As shown by western blot analysis of the cell extracts, incubation of cells for a short period of time with ET-1 results in the phosphorylation of p38 MAP, p44/42, SAP/JNK and Akt kinases. This effect occurs within minutes following a challenge with ET-1, and disappears after 45 and 60 min for the p-38 and SAP/JNK kinases, respectively. The activation of these kinases is probably necessary for the induction by ET-1 of MMP-1 production and MMP-13 production. The inhibition of p38 kinase is associated with a suppression of the ET-1-induced stimulation of both enzymes, whereas the inhibitions of adenyl cyclase-dependent PKA kinase is associated with a partial suppression of the ET-1-induced stimulation of MMP-13 production only. This suggests that these inhibitors are specific for the ET-1-activated pathways since they do not affect the basal levels of MMP-1 and MMP-13.
Another point also deserves consideration. Tardif and colleagues \[[@B32]\] have described two OA chondrocyte populations distinctive by their MMP-13 content and their response to IL-1β. One population contains small amounts of MMP-13 protein and is highly sensitive to IL-1β stimulation; the other population is enriched in MMP-13 protein but poorly responds to the cytokine. The cell heterogeneity of OA cartilage may explain some variability of the results observed in our study, particularly in the case of using low doses of the MEK1/2 inhibition followed by ET-1 stimulation. In fact, when MAP kinase pathways (extracellular signal-regulated kinase, JNK and p38) are activated in chondrocytes, their inhibition is dependent of the inhibitor concentration used, particularly for SB 203580 and PD 98059 \[[@B18]\]. PD 98059, which had no effect in the present study at the concentration of 10 μM on ET-1-induced iNOS expression and NO production, was demonstrated in other studies to suppress NO induction in human chondrocytes, as shown by Gemba and colleagues \[[@B18]\].
The phosphorylation of p38 MAP kinase by ET-1 was also described in osteoblast-like cells \[[@B33]\] and in cardiac myocytes \[[@B34]\], while in chondrocytes overproducing MMP-1 and MMP-13 this MAP kinase was shown to be phosphorylated principally by IL-1β \[[@B35]\]. Activation of PKA was shown to be required for the upregulation of iNOS, and for the subsequent production and release of NO by several cell types such as vascular smooth muscle cells \[[@B36]\], cardiac myocytes \[[@B37]\] and human macrophages \[[@B38]\]. It is also associated with the cytokine-induced NO production in human OA articular chondrocytes \[[@B39]\]. Our results suggest that the activation of PKA is also required for the ET-1-induced upregulation of iNOS and for subsequent production of NO by human OA chondrocytes. However, PKA activation seems to be less required for the ET-1-induced upregulation of MMP-13 and not at all necessary for the upregulation of MMP-1 since the inhibition of PKA with KT5720 does not affect the ET-1-induced overproduction of this enzyme. In the present study, subtle differences are shown in the pattern of inhibition of the ET-1-induced overproduction of MMP-1 and MMP-13. The effect of ET-1 on MMP-13 production was more sensitive to the inhibitors of protein kinases than on MMP-1 production. As suggested earlier, these variable responses point to possible different cell populations producing these two enzymes or to relevant signalling pathways eliciting the ET-1-induced stimulations \[[@B35]\].
We also tested the hypothesis that ET-1 may act in OA through induction of apoptosis. This was based on the findings that cells of the superficial layer disappear during cartilage degeneration \[[@B40]\], that ET-1 is preferentially produced in this layer \[[@B2]\], and that NO may induce apoptosis and cell death at high concentrations \[[@B29]\]. Indeed, chondrocyte death may represent one of the contributing factors in cartilage destruction. However, as shown in the present study, ET-1 does not appear to induce chondrocyte apoptosis or cell death. Using the TUNEL technique (which was recently shown to detect both apoptosis and cell death \[[@B29]\]), and using Bcl2 and Bad protein determination, no differences were found between ET-1-treated cultures and control cultures.
Conclusion
==========
The present study shows that ET-1 causes an overproduction of NO, MMP-1 and MMP-13 in human OA chondrocytes. The signalling pathway used by ET-1 in these cells was also demonstrated. The fact that ET-1 possesses the biological properties described acknowledges this peptide as an important catabolic factor contributing to the cartilage destruction via induction of the deleterious molecules such as MMPs and NO. NO seems to be a key molecule that is produced in parallel with the ET-1-induced overproduction of the MMPs. Blocking the effects of ET-1 may thus become a useful therapeutic approach aimed at stopping cartilage destruction in rheumatic conditions such as rheumatoid arthritis and OA.
Abbreviations
=============
DMEM = Dulbecco\'s modified Eagle\'s medium; ELISA = enzyme-linked immunosorbent assay; ET-1 = endothelin-1; FCS = foetal calf serum; IL = interleukin; iNOS = inducible nitric oxide synthase; L-NIL = L-N^6^(1-iminoethyl)lysine; MAP = mitogen-activated protein; MEK1/2 = mitogen-activated protein kinase kinase 1/2; MMP = metalloprotease; NO = nitric oxide; OA = osteoarthritis; PKA = protein kinase A; SAP/JNK = stress-activated protein kinase/Jun-N-terminal kinase; TUNEL = terminal deoxynucleotidyl transferase-medulated dUTP nick end labelling.
Competing interests
===================
The author(s) declare there are no competing interests.
Authors\' contributions
=======================
CAM executed the study, contributed to the preparation of the manuscript and performed statistical analysis. MR-B and FSS assisted in the experiments and in the isolation of chondrocytes from human cartilage. JCF, JM-P and J-PP assisted with the design of experiments and obtained human tissues. DRM evaluated and interpreted data and assisted with the preparation of the manuscript. FM designed the study, supervised the project, evaluated and interpreted data, and prepared the manuscript.
Acknowledgements
================
This work was supported by grants from the Canadian Institutes of Health Research (CIHR) (DSH-44200 and MOP-57760) and Dr Moldovan is the recipient of a scholarship from the FRSQ. The authors thank Heather Yampolsky for her excellent assistance in manuscript preparation.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Effect of protein kinase inhibitors and LY83583 on endothelin-1 (ET-1)-induced MMP-13 and MMP-1 production by human osteoarthritis chondrocytes. Confluent monolayer chondrocytes were preincubated 30 min at 37°C with SB 202190 (1 μM), PD98059 (10 μM), Wortmannin (100 nM), KT5720 (4 μM) or LY83583 (2 μM) for 30 min at 37°C, and were then challenged with ET-1 for 24 hours. MMP-13 and MMP-1 proteins were measured in the culture media using specific ELISA assays. *P*values indicate significant differences comparing experimental conditions with ET-1 treatment alone (\*) and versus the control cultures (^\#^). Values are expressed as the mean ± standard error of the mean of five independent experiments performed in duplicate. Significant differences: ^\#,^\* *P*\< 0.05; ^\#\#,^\*\* *P*\< 0.01; ^\#\#\#,^\*\*\* *P*\< 0.005.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Effect of endothelin-1 (ET-1) on nitric oxide (NO) release and inducible nitric oxide synthase (iNOS) expression by human osteoarthritis (OA) chondrocytes. NO was measured in the culture media, and iNOS protein was detected in cell extracts and revealed by western blot using specific antiserum, as described in Materials and methods. **(a)**Concentration-dependent ET-1-induced NO accumulation in the culture media from confluent human OA chondrocytes treated with ET-1 (0--100 nM) at 37°C for 24 hours. **(b)**Effect of protein kinase inhibitors and of guanylate cyclase inhibitor on ET-1-induced NO release in OA chondrocytes. Confluent monolayer chondrocytes were preincubated with SB 202190 (1 μM), PD98059 (10 μM), Wortmannin (100 nM), KT5720 (4 μM) or LY83583 (2 μM) for 30 min at 37°C and then challenged with ET-1 for 24 hours, and NO was determined in the culture media. **(c)**Effect of iNOS inhibition on NO release induced by ET-1 in human OA chondrocytes. The chondrocytes were pretreated with the allosteric inhibitor of iNOS, L-N^6^(1-iminoethyl)lysine (L-NIL) (0--50 μM), for 30 min and were then incubated with ET-1 (10 nM) for an additional 24 hours. The NO level was measured in the culture media. **(d)**Effect of protein kinase inhibitors and LY83583 on ET-1-induced iNOS in human OA chondrocytes. Chondrocytes were preincubated with SB 202190 (1 μM), PD98059 (10 μM), Wortmannin (100 nM), KT5720 (4 μM) or LY83583 (2 μM) for 30 min at 37°C and then challenged with ET-1 for 24 hours, and iNOS was then quantified. M.W., **molecular weight**. (a)--(c) Values are the mean ± standard error of the mean of six independent experiments performed in duplicate. (d) Representative blot of three independent experiments. *P*values indicate the significant difference between ET-1 treated cells and cells treated with indicated inhibitors + ET-1 (\*) and versus control (^\#^). Significant differences: ^\#,^\* *P*\< 0.05, ^\#\#\#,^\*\*\* *P*\< 0.005.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Phosphorylation of p38 mitogen-activated protein (MAP) kinase, Akt, p44/42 and stress-activated protein kinase/Jun-N-terminal kinase (SAP/JNK) by endothelin-1 (ET-1) in human osteoarthritis (OA) chondrocytes. **(a)**Western immunoblot of p38 MAP kinase. Confluent human OA chondrocytes were incubated with ET-1 (10 nM) for 10 or 45 min and the cell extracts were prepared as described in Materials and methods. Western immunoblots used antiserum against activated (phospho-p38) and total p38 MAP kinase (p38 T). Representative result of three different experiments. **(b)**Western immunoblot of Akt. Cells were incubated for 2, 5, 15, 30, 45 or 60 min in the presence of ET-1 (10 nM) and cell extracts were prepared as described in Materials and methods. Western immunoblot was carried out using an antiserum specific for phospho Ser 473 of Akt. Representative result of three different experiments. **(c)**Western immunoblot of p44/42. Confluent human OA chondrocytes were incubated with ET-1 (10 nM) for 0, 5, 15 or 60 min and cell extracts were prepared as described in Materials and methods. **(d)**Western immunoblot of SAP/JNK protein kinase. Confluent human OA chondrocytes were incubated with ET-1 (10 nM) for 0, 5, 30, 45 and 60 min, and cell extracts were prepared as described in Materials and methods. Actin detection was used as a control of the level of proteins loaded. Representative blot of three independent experiments.
:::
:::
|
PubMed Central
|
2024-06-05T03:55:54.959579
|
2005-1-17
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065327/",
"journal": "Arthritis Res Ther. 2005 Jan 17; 7(2):R324-R332",
"authors": [
{
"first": "Christina Alexandra",
"last": "Manacu"
},
{
"first": "Johanne",
"last": "Martel-Pelletier"
},
{
"first": "Marjolaine",
"last": "Roy-Beaudry"
},
{
"first": "Jean-Pierre",
"last": "Pelletier"
},
{
"first": "Julio C",
"last": "Fernandes"
},
{
"first": "Fazool S",
"last": "Shipkolye"
},
{
"first": "Dragoslav R",
"last": "Mitrovic"
},
{
"first": "Florina",
"last": "Moldovan"
}
]
}
|
PMC1065328
|
Introduction
============
Patients suffering from autoimmune diseases are characterized by the presence of autoantibodies directed to a wide range of autoantigens. Mixed connective tissue disease (MCTD) is a relatively rare systemic autoimmune disease and includes a group of patients with overlapping clinical symptoms of systemic lupus erythematosus (SLE), systemic sclerosis (SSc), rheumatoid arthritis and polymyositis/dermatomyositis. Sharp and colleagues were the first to describe MCTD as a distinct rheumatic disease \[[@B1]\], but whether MCTD can be regarded as a distinct disorder has been a subject of discussion \[[@B2]\]. A characteristic serological feature that distinguishes MCTD patients from patients with other connective tissue diseases is high levels of autoantibodies directed against the U1 small nuclear ribonucleoprotein (snRNP) particle \[[@B1],[@B3]\]. The U1 snRNP is a highly conserved RNA--protein complex, located in the nucleus, where it is involved in the processing of pre-mRNA \[[@B4],[@B5]\]. It consists of the U1 snRNA molecule and several proteins: the U1A, U1C and U1-70K (70K) proteins are components specific for the U1 snRNP, whereas the seven Sm proteins (B/B\', D1, D2, D3, E, F and G) are shared with other U snRNPs \[[@B6]\]. Most U1 snRNP components are autoantigenic in MCTD and SLE. Autoantibodies directed against U1A, U1C, 70K and the U1 snRNA molecule are mainly found in MCTD patients, whereas autoantibodies targeting Sm-D, Sm-B/B\' and the E.F.G complex are more specifically associated with SLE \[[@B7],[@B8]\].
The mechanisms through which such autoantigens, generally highly conserved and ubiquitously expressed molecules, escape tolerance and are recognized by the immune system as non-self remain unclear, but it is proposed that cell death is important in the initiation of autoimmune responses \[[@B9],[@B10]\]. Recently, secondary necrosis has also been put forward as a source of proteolytically modified autoantigens \[[@B11]\], but the modifications that occur on autoantigens during apoptosis were studied most extensively. Apoptotic modifications on autoantigens include specific cleavage by caspases or granzyme B, (hyper)phosphorylation, dephosphorylation, citrullination, methylation and transglutaminase cross-linking \[[@B10],[@B12],[@B13]\], and it is thought that these modifications might be seen by the immune system as novel \'cryptic\' epitopes. It is believed that these novel epitopes induce the primary immune response, and that secondary immune responses and epitope spreading result in autoantibodies that are directed against unmodified regions of the autoantigens and antigens that are associated with the initially modified autoantigen \[[@B9]\].
One of the apoptotic modifications occurring on the U1 snRNP is the cleavage of 70K at residue 341 by caspase-3 \[[@B14],[@B15]\]. Antibodies against 70K are in general the first autoantibodies to appear in anti-U1 snRNP (often referred to as anti-RNP) positive patients, indicating that 70K is important as an initial autoantigen \[[@B16]\]. The molecular and immunological characteristics of the major apoptotic isoform of 70K, a 40 kDa cleavage product that remains associated with the U1 snRNP complex \[[@B17]\], and its role in the triggering of the primary and possibly secondary autoimmune response, are therefore intriguing.
Recently it was shown that sera of some anti-U1 snRNP positive patients contain antibodies that specifically bind to the apoptotic form of 70K, which displays an epitope that is not present on the intact form \[[@B18],[@B19]\]. This epitope is dependent on the region between amino acids 180 and 205, partly overlapping with the RNA-binding domain and overlapping with the most common T cell epitope \[[@B20]\]. In this study we analyzed a cohort of MCTD and control patients for the presence of autoantibodies against intact and apoptotic 70K. Moreover, we longitudinally analyzed sera from another group of MCTD patients. Our results show that, early in disease, autoantibodies directed against the apoptotic form of 70K (70K^apop^) are more strongly represented than autoantibodies against the intact form. Longitudinal studies also show that autoantibodies against 70K^apop^are not significantly correlated with disease flares.
Methods
=======
Patient sera
------------
All patients were seen at the Department of Rheumatology of the University Medical Centre Nijmegen or the St Maartenskliniek Nijmegen (The Netherlands), and were classified in accordance with standard criteria for each disease. All MCTD patients (*n*= 53) tested positive for anti-U1 snRNP autoantibodies by counter-immunoelectrophoresis, and for antibodies against one or more components of the U1 snRNP complex by immunoblotting. Most of the sera (91%) were also RNP positive as shown by U1 snRNA co-immunoprecipitation. Longitudinal serum collections were obtained from 12 MCTD patients and have been described previously \[[@B21]\]. From each patient, over a period of 4--15 years (average 10 years), 8 to 33 serum samples (average 18 samples) were available and were analyzed. During the follow-up study, the patients were regularly monitored for clinical and serological parameters. At each visit the disease activity was measured in accordance with a validated SLE disease activity index described by Ter Borg and colleagues \[[@B22]\]. Medication was given as indicated by the clinical status. Additionally, patient sera were collected from SLE (*n*= 48), polymyositis/dermatomyositis (*n*= 26), primary Sjögren\'s syndrome (*n*= 18), SSc (*n*= 10), rheumatoid arthritis (*n*= 3), Raynaud\'s phenomenon (*n*= 3) and undefined connective tissue disease (*n*= 1). Informed consent was obtained from all participants in accordance with the medical ethical regulations of the local ethics committee. Sera were stored at -70°C until use.
Cell lines, induction of cell death and preparation of cell extracts
--------------------------------------------------------------------
Jurkat (human T cell leukemia) suspension cells were grown in RPMI 1640 medium (Gibco-BRL), supplemented with 1 mM sodium pyruvate, 1 mM penicillin, 1 mM streptomycin and 10% heat-inactivated fetal calf serum (Gibco-BRL), in a humidified 37°C incubator containing 5% CO~2~. Cells were maintained at a concentration of 10^6^cells/ml and were induced to undergo apoptosis by the addition of 10 μg/ml anisomycin. Eight hours after induction, apoptotic cells were harvested by centrifugation at 800 *g*for 10 min and washed with PBS. Apoptotic and non-apoptotic Jurkat cells were resuspended in Nonidet P40 (NP40)-containing lysis buffer (25 mM Tris-HCl, pH 7.6, 100 mM KCl, 10 mM MgCl~2~, 0.25 mM dithioerythritol, 1% NP40, Complete™ protease inhibitor cocktail \[Roche\]) at a concentration of 10^8^cells/ml. Cells were lysed on ice for 30 min and subsequently centrifuged for 30 min at 12,000 *g*and 4°C. Supernatants were used immediately or stored at -70°C.
SDS--polyacrylamide gel electrophoresis and western blotting
------------------------------------------------------------
Cell extracts of 1.3 × 10^7^non-apoptotic Jurkat cells and 1.3 × 10^7^apoptotic Jurkat cells, either separately or mixed, were separated by SDS--polyacrylamide gel electrophoresis. Directly after gel electrophoresis, proteins were transferred to a nitrocellulose membrane (Schleicher & Schuell) by semi-dry electroblotting. Staining of the membrane with Ponceau S (Sigma) was used to verify protein transfer.
Probing western blots with patient sera
---------------------------------------
All incubation steps were performed at approximately 20°C on a shaking table. Western blots containing non-apoptotic and apoptotic Jurkat cell extracts were pre-blocked with 5% non-fat dried milk in PBS containing 0.1% NP40 (MPBS/NP40) for 2 hours. Subsequently, membranes were incubated with patient serum, diluted 1000--5000-fold in MPBS/NP40, for 1 hour. After extensive washing with PBS containing 0.1% NP40 (PBS/NP40), membranes were incubated with horseradish peroxidase-labeled rabbit anti-human IgA/IgG/IgM antibody (Dako Immunoglobulins), diluted 1000-fold in MPBS/NP40, for 1 hour. After several washes with PBS/NP40 and PBS, bound antibodies were detected by enhanced chemiluminescence. Antibody reactivities against 70K and 70K^apop^were scored ranging from 0 to 5 by three researchers independently. In each experiment several control antibodies were used.
Results
=======
In this study, patient sera were analyzed for the presence of autoantibodies against 70K and its apoptotic product (70K^apop^), on western blots containing extracts of non-apoptotic and apoptotic Jurkat cells. Two positive controls for the detection of 70K and 70K^apop^were included in each experiment: anti-70K mouse monoclonal antibody 2.73 \[[@B23]\], which displays higher reactivities with 70K than with 70K^apop^, and serum from MCTD patient B16, which reacts with both 70K and 70K^apop^. The position of 70K^apop^on western blots was confirmed by a recombinant monoclonal antibody recognizing both 70K and 70K^apop^(Fig. [1a](#F1){ref-type="fig"}) \[[@B24]\]. The results show that in these apoptotic cells 70K is converted almost completely into 70K^apop^. Besides positive controls for 70K and 70K^apop^, mouse monoclonal antibody ANA125 directed against Sm-B/B\' (Fig [1a](#F1){ref-type="fig"}), and anti-U1-A/U2-B\" mouse monoclonal antibody 9A9 (not shown) were also used. To be able to detect autoantibody reactivities to the intact 70K and its apoptotic 40K fragment simultaneously and to facilitate a direct comparison of these reactivities, a mixture of apoptotic and non-apoptotic cell extracts was used to prepare western blots. An additional advantage of this approach was that differences between blots could be excluded, thereby allowing a more accurate comparison of reactivities with 70K and 70K^apop^in a single patient serum. Serum antibody reactivities against 70K and 70K^apop^were scored ranging from 0 to 5. Figure [1b](#F1){ref-type="fig"} shows a western blot containing such a mixture of non-apoptotic and apoptotic Jurkat cell extracts, probed with a serial dilution of serum from MCTD patient B16. It can be seen that the signals for 70K and 70K^apop^increase when the serum is applied at a lower dilution, indicating that the western blot assay can be used for semi-quantitative interpretation.
Autoantibodies against 70K are more easily detected with 70K^apop^
------------------------------------------------------------------
The presence of high levels of autoantibodies directed against components of the U1 snRNP, such as 70K, is one of the criteria for the diagnosis of MCTD \[[@B2]\]. However, anti-70K antibodies are also found in some SLE and SSc patients \[[@B3]\]. To compare the disease specificity of anti-70K^apop^and anti-70K autoantibodies, sera from a group of MCTD patients and from a group of patients suffering from a variety of autoimmune disorders were analyzed. As shown in Table [1](#T1){ref-type="table"}, most MCTD patients (54%) displayed antibody reactivities that preferentially recognized 70K^apop^over the intact 70K protein. Seven patients (13%) reacted with 70K and 70K^apop^with similar efficiencies, and only 6% of the MCTD patients reacted preferentially with the intact 70K protein. Fourteen sera (27%) did not react detectably with either 70K polypeptide, although the sera were anti-RNP positive by several techniques. These results indicate that 70K^apop^is a better antigen than the intact 70K protein for the detection of anti-70K autoantibodies. Antibody reactivity with 70K^apop^was found in only 2% of sera from control groups, whereas antibody reactivity with 70K was found in 5% of patient sera from control groups.
Autoantibodies against 70K^apop^are not correlated with disease activity
------------------------------------------------------------------------
It has been described that, in some patients with MCTD, antibody titers against the U1 snRNA molecule are correlated with disease activity, and could even possess prognostic value \[[@B21]\]. In contrast, most studies did not find a correlation between disease activity and antibody responses to 70K, either by serum analysis using recombinant protein as antigen in ELISA \[[@B21],[@B25]\] or by analysis on western blots using native protein from cell extracts \[[@B26]\]. Only one study, using ELISA with recombinant 70K as technique, has reported decreasing disease activity concomitant with decreasing anti-70K antibody levels \[[@B27]\]. Because apoptotic modifications on autoantigens, such as the cleavage of 70K, are believed to be involved in the primary autoimmune response, we proposed that immune complexes containing anti-70K^apop^antibodies might also be important for triggering disease flares. Serum samples were collected longitudinally from 12 MCTD patients by a follow-up during variable time intervals (4--15 years; average 10 years). All samples were analyzed for the presence of autoantibodies against 70K and 70K^apop^on western blots containing non-apoptotic and apoptotic Jurkat cell extracts, and the presence of these autoantibodies was compared with the disease activity of each patient. The overall conclusion of this longitudinal study was that no significant correlations between antibody titres against either 70K^apop^or 70K and disease exacerbations could be observed.
Autoantibodies against 70K^apop^are more prevalent early in disease
-------------------------------------------------------------------
As mentioned above, it has been proposed that apoptotic modifications trigger the primary immune response towards self proteins and that, through secondary immune responses and epitope spreading, autoantibodies directed against unmodified regions on the autoantigen appear at later stages of the disease. To investigate this possibility for the 70K autoantigen, the longitudinal serum collection \[[@B21]\] of 12 MCTD patients was re-examined, now for antibodies against 70K and for antibodies against 70K^apop^. Three patients produced antibodies reacting strongly with 70K^apop^, whereas no or only weak reactivity against 70K was observed. In one of these three patients, autoantibodies against 70K^apop^were more prevalent in early serum samples, and the level decreased in time. Eight patients were found to have high titres of antibodies with reactivities to both 70K and 70K^apop^. Interestingly, in three of these patients early serum samples showed a higher reactivity with 70K^apop^than with 70K, whereas later samples showed comparable reactivities with both antigens, or higher reactivities with 70K. An example of this type of reactivity profile is shown in Fig. [2](#F2){ref-type="fig"}. One of the 12 patient sera did not detectably contain antibodies directed against 70K or 70K^apop^. These results thus support the idea that antibodies against 70K^apop^appear earlier in the disease than antibodies against the complete 70K protein.
Discussion
==========
Greidinger and colleagues recently showed that antibodies against the 70K protein in RNP-positive patients are often accompanied by antibodies directed against the apoptotic cleavage product of this autoantigen and that the B cell epitopes recognized on the apoptotic product are antigenically different from those contained in the intact form of the 70K protein \[[@B18],[@B19]\]. This study is the first to confirm and extend these findings and strongly suggests that the reactivity of a patient serum with anti-70K antibodies depends on the presence of antibodies against epitopes shared by 70K and 70K^apop^, and on the presence of antibodies against epitopes exclusively present on 70K^apop^. The major apoptosis-specific epitope on 70K has been shown to be located in the region containing the RNA-binding domain, and its formation depends on amino acids 180--205, overlapping with the most common T cell epitope \[[@B20]\]. Other apoptosis-specific epitopes on 70K have not yet been described. In our recent studies, monoclonal recombinant human antibodies against 70K were isolated from phage display libraries derived from SLE patients, and several of these monoclonal antibodies preferentially recognized 70K^apop^on a western blot and in immunoprecipitation experiments (\[[@B24]\], and D Hof, unpublished results). It is believed that the apoptotic cleavage of 70K leads to the exposure of a neo-epitope that, if presented to the immune system, triggers the autoimmune response. Greidinger and colleagues showed that a mutant consisting of the amino-terminal 205 amino acids, was indeed able to induce an anti-70K^apop^antibody response in mice, with subsequent epitope spreading. Interestingly, some of the immunized mice developed pulmonary lesions comparable to lesions found in lungs of MCTD patients. This finding supports the hypothesis that apoptosis-specific epitopes, and antibodies directed against them, might have a pathological role in the triggering and maintenance of the human autoimmune response to 70K \[[@B19]\].
In our study, a minority of MCTD sera (4%) contained autoantibodies exclusively reacting with intact 70K. We suggest that these sera derive from patients in a relatively late disease phase and primarily contain antibodies resulting from expanded epitope spreading. Most epitopes recognized by these sera might therefore be dependent on the carboxy-terminal part of the protein, which is cleaved off during apoptosis and is not present on 70K^apop^. Patients that tested negative in our western blot experiments might either have low levels of anti-70K antibodies or might not produce such antibodies at all. Instead, other components of the U1 snRNP, such as the U1 RNA molecule, U1A or U1C, might be targeted by these sera and might explain their anti-U1 snRNP reactivity.
We show here that most U1 snRNP-positive patient sera preferentially recognize 70K^apop^, which is most probably explained by the presence of antibodies targeting an apoptotic 70K epitope. These results are in line with reports by Greidinger and colleagues \[[@B18],[@B19]\], who found that about 50% of their RNP-positive sera contained 70K^apop^autoantibodies.
How disease flares are induced is not completely understood. Correlations between serum levels of certain autoantibodies and disease activity have been reported for MCTD and SLE \[[@B21],[@B22]\], but it can be disputed whether these antibodies contribute to the disease flares or are merely epiphenomena. Our data show that antibodies against 70K^apop^are not significantly correlated with disease activity, suggesting that there is no important role for 70K^apop^in the initiation of disease flares. However, it is possible that the variations in antibody levels against the apoptosis-specific epitope are masked by the presence of antibodies against other epitopes on 70K and U1-70K^apop^. Furthermore, a polyspecific secondary antibody was used to detect bound serum antibodies, and as a consequence variations in isotype-specific antibody levels might have remained undetected.
It has been reported that the first autoantibodies to appear in anti-RNP-positive patients are generally antibodies against 70K \[[@B16],[@B27]\]. Our results suggest that 70K^apop^drives the primary autoimmune response to 70K, because in several patients antibodies against an epitope associated with 70K^apop^preceded the appearance of reactivity with intact 70K. The fact that the first serum samples from relatively few patients exclusively contain anti-70K^apop^antibodies might be due to the stage of disease development at which the patient enters the rheumatological clinic. It is likely that the first symptoms, later followed by the diagnosis of the disease, had been established years before the start of the longitudinal study. Moreover, it is possible that autoantibodies, especially those generated by the primary immune response, were already present before the manifestation of clinical symptoms and that subsequent epitope spreading might have occurred before the patient entered the rheumatological clinic. For example, anti-cyclic citrullinated peptide autoantibody is a very specific marker for rheumatoid arthritis, and such antibodies can be detected in patients up to 10 years before the occurrence of the first clinical symptoms \[[@B28],[@B29]\]. In our opinion this might explain why a relative enrichment of anti-70K^apop^antibodies could not be detected in the early sera of all patients.
During apoptosis, the U1 snRNP complex is modified in several ways. In addition to cleavage of 70K, U1 snRNA and the Sm-F protein are cleaved, and phosphorylated serine--arginine proteins associate with the complex \[[@B30]\]. Apoptotic modifications of the U1A and U1C proteins have not yet been described. 70K can be cleaved by caspase-3 and granzyme B, and it can be oxidatively fragmented in the presence of metals, resulting in products of 40, 60 and 55 kDa, respectively. Correlations between the recognition of specific 70K fragments and disease manifestations are interesting. For example, patients suffering from Raynaud\'s phenomenon preferentially recognize the oxidatively modified 55 kDa fragment of 70K \[[@B31]\]. The findings that early MCTD sera are enriched for antibodies against the 40 kDa apoptotic fragment (70K^apop^) and that most sera show a higher reactivity with this fragment suggest that caspase-3 cleaved 70K has a role in breaking tolerance in these patients. Although granzyme B is postulated to have a role in breaking tolerance \[[@B32]\] to 70K, it is unknown whether specific patient groups preferentially recognize the 60 kDa cleavage product generated by granzyme B, which would be interesting to study in more detail.
Conclusions
===========
Analysis of a group of MCTD patient sera by western blotting showed that the majority of patient sera recognized 70K^apop^more efficiently than the intact form of the 70K protein. The fact that the presence of these antibodies in most patients precedes the occurrence of other anti-70K antibodies suggests that 70K^apop^is particularly important for the early detection of this disease in patients.
Abbreviations
=============
70K = U1-70K; 70K^apop^= apoptotic cleavage product of U1-70K protein of about 40 kDa; MCTD = mixed connective tissue disease; NP40 = Nonidet P40; RNP = ribonucleoprotein; SLE = systemic lupus erythematosus; snRNP = small nuclear ribonucleoprotein; SSc = systemic sclerosis; U1-70K = 70 kDa protein component of the U1 snRNP complex.
Competing interests
===================
JMHR works and holds shares in ModiQuest BV, a company producing antibodies for research purposes, but will not gain or lose financially from publication of this paper.
Authors\' contributions
=======================
DH conceived of the study, participated in the design of the study and was involved in the analysis of the immunoassay results. KC performed and analyzed the immunoassays. DR collected the patient sera. FH provided patient data. GP participated in the design of the study and in the analysis of the immunoassay results. WV conceived of the study and participated in the design of the study. JR conceived of the study and was involved in its design and coordination. All authors read and approved the final manuscript.
Acknowledgements
================
We thank Léon Peeters (ModiQuest BV, Nijmegen, The Netherlands) for assisting with the data analysis.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Anti-U1-70K and anti-70K^apop^detection by western blotting. Apoptosis was induced in Jurkat cells by incubation with anisomycin for 8 hours. Western blots were prepared with the resulting cell extracts, and the positions of relevant polypeptides were revealed with patient sera and monoclonal antibodies with the use of a chemiluminescent detection procedure. The positions of the various proteins are indicated on the left, and molecular mass marker positions on the right. **(a)**U1-70K (70K) detected with a serum from MCTD patient B16 (lanes 1 and 5), anti-70K monoclonal antibody 2.73 (lanes 2 and 6) and an anti-70K single-chain recombinant antibody (scFv; lanes 3 and 7) (70K); lanes 4 and 8, Sm-B/B\' detected with a monoclonal anti-Sm-B/B\' antibody (ANA125); the position of U1A, which is also recognized by serum from MCTD patient B16 (lanes 1 and 5), was determined by a U1A-specific monoclonal antibody (not shown). In apoptotic cells (lanes 5--8), 70K is present as a 40 kDa species (70K^apop^). **(b)**A serum sample from MCTD patient B16 was applied at 5000-fold (lane 1), 10,000-fold (lane 2) and 20,000-fold (lane 3) dilution on a western blot containing a mixture of non-apoptotic and apoptotic Jurkat cell extracts. In lane 4 the 70K protein was detected with mouse monoclonal antibody 2.73, which reacts much more efficiently with 70K than with 70K^apop^.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Longitudinal anti-70K analysis of patient T2. Eighteen serum samples taken over a period of 7 years with approximately equal time intervals were analyzed on western blots containing non-apoptotic and apoptotic Jurkat cell extracts. The positions of U1-70K (70K), of which two isoforms are visible, and 70K^apop^are indicated on the left. In lane 19, 70K was detected with mouse monoclonal 2.73, which reacts much more strongly with 70K than with 70K^apop^. MoAb, monoclonal antibody.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Recognition of U1-70K (70K) and its apoptotic cleavage product (70K^apop^) by patients with different autoimmune disorders
:::
No. (%) of patients with disorder
------------ ---- ----------------------------------- --------- -------- ------- -------
MCTD 53 15 (28) 14 (26) 7 (13) 1 (2) 2 (4)
SLE 48 0 1 (2) 1 (2) 0 1 (2)
Non-SLE^a^ 61 0 0 0 0 2 (3)
^a^The control group of non-systemic lupus erythematosus (non-SLE) patients consists of patients suffering from polymyositis/dermatomyositis (*n*= 26), primary Sjögren\'s syndrome (*n*= 18), systemic sclerosis (*n*= 10), rheumatoid arthritis (*n*= 3), Raynaud\'s phenomenon (*n*= 3) and undefined connective tissue disease (*n*= 1). MCTD, mixed connective tissue disease.
:::
|
PubMed Central
|
2024-06-05T03:55:54.961597
|
2005-1-11
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065328/",
"journal": "Arthritis Res Ther. 2005 Jan 11; 7(2):R302-R309",
"authors": [
{
"first": "Daniëlle",
"last": "Hof"
},
{
"first": "Kalok",
"last": "Cheung"
},
{
"first": "Dirk-Jan RAM",
"last": "de Rooij"
},
{
"first": "Frank H",
"last": "van den Hoogen"
},
{
"first": "Ger JM",
"last": "Pruijn"
},
{
"first": "Walther J",
"last": "van Venrooij"
},
{
"first": "Jos MH",
"last": "Raats"
}
]
}
|
PMC1065329
|
Introduction
============
Osteoarthritis (OA) is the leading cause of chronic disability in developed countries, symptomatically affecting about 14% of the adult population in the United States alone. Among the signs of early OA are collagen disorganization, an increase in water content, a decrease in superficial proteoglycan, and alterations in glycosaminoglycans \[[@B1]\]. The later changes include cartilage loss (thinning effect), fibrillation, and surface erosion. Current imaging technologies are limited in their ability to monitor changes in articular cartilage \[[@B2]\]. Furthermore, symptoms are an unreliable indicator of disease progression \[[@B3]\]. Since the cartilage response to intervention cannot be monitored in a noninvasive or minimally invasive manner, assessing the effectiveness of these drugs and following the progression of the disease remain a challenge. This deficiency is the basis of the current US National Institutes of Health OA initiative to find solutions to this major healthcare dilemma \[[@B3]\]. A diagnostic technique capable of high-resolution imaging of articular cartilage *in vivo*could be invaluable to detect the onset of disease, follow its progression, and monitor therapeutic effectiveness.
Other imaging technologies play an important role in managing OA, but they have limitations. While conventional x-rays have an obvious role in managing arthritis, this technology lacks the resolution to monitor changes within the cartilage \[[@B2],[@B4]\]. Magnetic resonance imaging is invaluable for globally evaluating the joint noninvasively, with a typical clinical resolution of 250--300 μm at 10T \[[@B5]\]. However, the resolution of this technique is problematic, since cartilage is typically less than 2--3 mm thick and the evaluation would rely heavily on the interpretation of a few pixels \[[@B6],[@B7]\]. In addition, its high cost, relatively long imaging time, large size of equipment, and limited availability could limit its widespread clinical use. Arthroscopy is also widely used in the diagnosis of joint disorders \[[@B8]\]. While it provides magnified views of the articular surface, it is unable to assess subsurface.
Optical coherence tomography (OCT) is a recently developed imaging technique that can generate cross-sectional images of tissue microstructure \[[@B9],[@B10]\]. OCT is analogous to ultrasound, but measures the intensity of infrared light rather than sound. It is an attractive imaging alternative for OA because it permits imaging in near-real time with unprecedented high resolution (4--15 μm), 10 to 100 times as fine as that of current clinical imaging modalities. Since OCT is based on fiber-optic systems, the apparatus is compact, roughly the size of an ultrasound unit. Imaging catheters can be constructed with diameters less than 0.006 inches (Lightlab Imaging Inc, Westford, MA, USA; <http://www.lightlabimaging.com>). Recently, OCT has been adapted for high-resolution imaging in nontransparent tissue. In addition, a variety of spectroscopic techniques can be incorporated, such as absorption, dispersion, and polarization spectroscopy \[[@B11]-[@B13]\].
Preliminary work demonstrated the feasibility of OCT in assessing joint cartilage pathologies *in vitro*\[[@B11],[@B14]\]. Microstructures such as fibrillations, cartilage thinning, and new bone growth can be identified on OCT images \[[@B14]\]. Comparison with histology reveals strong correlation between OCT images and corresponding histological sections. In addition, OCT has demonstrated superior qualitative and quantitative performance against both 30- and 40-MHz ultrasound, the current clinical technology with the highest resolution \[[@B15],[@B16]\].
Polarization-sensitivity OCT imaging of articular cartilage has also been performed \[[@B11],[@B14]\]. With this technique, the OCT image changes with change in the polarization state of the incident light. In the previous *in vitro*study, polarization-sensitive changes on OCT images of cartilage were directly correlated with collagen organization \[[@B11]\], as assessed by picrosirius staining. Loss of both polarization sensitivity and collagen organization were noted to take place before cartilage thinning and fibrillation, making it a potential additional marker of early OA in addition to structural imaging. These results have been recently confirmed also in tendons and ligaments, and also in studies with theoretical modeling \[[@B17],[@B18]\]. Through this work, quantitative methods have now been developed and are being studied, including the use of the fast Fourier transform or rate of peak change with rotation of the source optical axis.
This study extends our previous *in vitro*work \[[@B11],[@B14]\]. In this study, observations on the ability of OCT to perform *in vitro*imaging of the human knee were confirmed *in vivo*using a novel handheld probe.
Materials and methods
=====================
The principle behind OCT has been described in detail previously \[[@B9],[@B10]\]. A schematic drawing of the OCT system used in this study is shown in Fig. [1a](#F1){ref-type="fig"}. In this study, a novel, compact, handheld OCT imaging probe capable of performing lateral scanning of the articular cartilage subsurface during open knee surgeries was used. The probe had dimensions of \~1.5 cm in diameter and \~15 cm in length (see Fig. [1b](#F1){ref-type="fig"}) and was developed and used to deliver, focus, scan, and detect the returning beam. It consisted of a four-lens relay and a scanning mirror. The measured resolution was approximately 11 μm (axial) and 30 μm (transverse) with a working distance (as defined by the distance between the distal end of the probe and the beam focus) of about 2.5 cm, which provided enough space to perform noncontact imaging. A 532-nm visible beam (green) with a very low power (\<0.2 mW) was coupled into the handheld probe for aiming purposes. OCT images were stored in digital format and also recorded on a super VHS tape for future analysis.
The protocol for OCT imaging during open knee surgery was approved by the investigational review board of the Massachusetts Institute of Technology and West Roxbury Veterans Association Hospital. Six patients 65 to 75 years of age who had been diagnosed with severe OA and were scheduled for treatment through partial or total knee replacement surgery were contacted about 4 weeks before surgery and their informed consent was obtained. Patients underwent routine surgical preparation procedures, and OCT imaging did not commence until the articular surface of the femur/tibia was fully exposed. OCT imaging was performed under sterile conditions. Both visually normal and visually abnormal regions were imaged. Imaging planes were marked with sterile dye (methylene blue) for tissue registration. During imaging, the probe did not contact the cartilage surface and the air distance between the probe and the cartilage surface was maintained at \~2.5 cm to insure that the imaged sites remain in focus. Images of 512 × 256 pixels (transverse × axial) were generated at four frames per second. Each OCT image corresponded to a two-dimensional tissue cross section 5 mm wide by 2.6 mm deep. Multiple sites on the articular surface were imaged within the allotted 10-min imaging period. After OCT imaging, surgery resumed as usual. Upon completion of the surgical procedures, the methylene blue dots were re-marked with India ink to improve visualization during post processing. The cartilage was then immediately fixed in 10% buffered formalin and then decalcified with EDTA followed by routine histological processing and stained with Masson trichrome blue.
Results
=======
A representative OCT image and the corresponding histology of normal knee articular cartilage are shown in Fig. [2](#F2){ref-type="fig"}. The OCT image (Fig. [2a](#F2){ref-type="fig"}) reveals that the cartilage was thick and uniform with a rather smooth surface. The same characteristics can also be seen in the histology as shown in Fig. [2b](#F2){ref-type="fig"}. A banding pattern is seen in the OCT image (Fig. [2a](#F2){ref-type="fig"}, red arrows). Previous work showed that this pattern represents alternating maximum and minimum intensities of back scattering, which results from rotation of the polarization state of back-reflected light as it passes through the organized collagen. During the imaging process, it was noted that the position of the bands moved as the polarization state of the incident light was changed (induced by moving the fiber in the sampling arm).
Fig. [3](#F3){ref-type="fig"} illustrates a representative OCT image (Fig. [3a](#F3){ref-type="fig"}) and the corresponding histology (Fig. [3b](#F3){ref-type="fig"}) of moderately diseased cartilage. Regions of diminished back scattering are noted in the OCT image, which correlate with areas of hypocellularity and diminished matrix in histological preparations. On the OCT image, the banding pattern is disrupted and correlates with histologically abnormal staining and cellularity.
Fig. [4](#F4){ref-type="fig"} shows an OCT image (Fig. [4a](#F4){ref-type="fig"}) and the corresponding histology (Fig. [4b](#F4){ref-type="fig"}) of severely diseased cartilage. Distinctive thinning of the cartilage was observed only on the left portion of both OCT image and histology. In addition, an irregular cartilage surface is seen in the OCT image, with multiple fibrillations evident in the corresponding histology. The OCT image is highly heterogeneous and the cartilage and bone interface are poorly identified. No banding appearance or polarization sensitivity was observed on this image. On the right portion of the OCT image and the histology section, cartilage is absent and the bone is exposed to the surface.
An OCT image of thick cartilage with no evidence of surface erosion and early degenerative changes is shown in Fig. [5](#F5){ref-type="fig"}. The OCT structural image is relatively homogeneous but the banding pattern is lost. The abnormal region seen on histology consists of an area of hypocellularity over a region of hypercellularity.
Fig. [6](#F6){ref-type="fig"} shows normal and diseased cartilage in close approximation in two sections of cartilage. The region on the left of both images is presumed normal cartilage, while on the right, the polarization sensitivity and back-scattering intensity abruptly changes. In addition, since these two samples come from the femur (Fig. [6a](#F6){ref-type="fig"}) and patella (Fig. [6b](#F6){ref-type="fig"}), respectively, the figure confirms that the polarization phenomenon exists in areas other than the tibia.
Discussion
==========
The current study demonstrates that osteoarthritic structural changes in cartilage can be visualized with OCT *in vivo*using a handheld probe. Structural changes including cartilage thinning and fibrillations were observed at a resolution substantially higher than that of any current clinical imaging technology. While normal cartilage demonstrates a banding pattern with a relatively homogeneous intensity (as seen in Fig. [2](#F2){ref-type="fig"}), areas of hypocellularity appear to lose this banding pattern (as seen in Fig. [3](#F3){ref-type="fig"}). These changes are dramatic enough to distinguish between adjacent areas of healthy and diseased tissue (as in Fig. [6](#F6){ref-type="fig"}). These results indicate that OCT may be able to be used by surgeons to aid in the evaluation of the microstructural integrity of articular cartilage during surgical procedures.
It can ultimately be envisioned that OCT imaging will be performed with a surgical arthroscope or a needle arthroscope for assessing the articular cartilage in a minimally invasive fashion. Future efforts will be on the development of a small OCT arthroscope capable of being either used in combination with or integrated into a standard arthroscope. Endoscopic imaging using an OCT probe introduced through the accessory port of an endoscope has been demonstrated in the human gastrointestinal tract \[[@B19],[@B20]\].
The collagen matrix in healthy cartilage is a highly organized structure \[[@B21],[@B22]\]. The banding pattern seen on the OCT images (e.g. Figs [2](#F2){ref-type="fig"}, [3](#F3){ref-type="fig"}, and [6](#F6){ref-type="fig"}) are due to tissue birefringence and are related to collagen organization \[[@B11],[@B14]\]. Changes in collagen organization, although not necessarily in collagen content, are among the earliest changes in OA \[[@B1]\]. It has been shown in animals that a decrease in birefringence, evident on histological evaluation, precedes fibrillations and can even be noted after chronic long-distance running \[[@B23],[@B24]\]. The diminishing and absent banding pattern on the OCT images (e.g. Figs [3](#F3){ref-type="fig"},[4](#F4){ref-type="fig"},[5](#F5){ref-type="fig"},[6](#F6){ref-type="fig"}), an observation supported by *in vitro*work, represents a reduction and loss of the birefringence of the cartilage, which is caused by the reduction or loss of collagen structural organization \[[@B14]\]. This has recently been confirmed in experimental models of OA in the rat \[[@B25],[@B26]\]. That study indicated that changes in the birefringent properties of cartilage (as with OA) are reflected in the polarization sensitivity of OCT images. In the current study, polarization changes were not quantitatively measured. However, as the fiber of the sample arm moved, it would induce a polarization state shift, allowing quick assessment of the polarization sensitivity of the area being imaged. Protocols are now available using fast Fourier transforms to quantitate single-detector OCT.
Conclusion
==========
A true clinical need exists for monitoring therapeutic intervention with regard to osteoarthritic cartilage. This study demonstrates real-time, high-resolution OCT imaging of articular tissues *in vivo*during joint replacement surgery at resolutions on a scale of micrometers. Abnormalities such as cartilage thinning and fibrillations were detected and qualitatively correlated with the corresponding histology. In addition, birefringence changes between osteoarthritic and normal cartilage were noted in this study, indicative of a loss of collagen organization. OCT represents a promising new technology for the evaluation of articular cartilage *in vivo*.
Abbreviations
=============
OA = osteoarthritis; OCT = optical coherence tomography.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
XL designed and constructed the OCT system. SM performed studies in patients, which included gaining their consent and postoperative observation. CP assisted in the construction of the OCT system. RG assisted with the construction of the handheld probe. DS advised on histological preparations. MH processed the tissues. JF consulted on the design of the OCT system. MB was involved with the engineering design, OCT protocol, evaluation of data, and writing of manuscript. All authors read and approved the final manuscript.
Acknowledgements
================
Dr Xingde Li is now at the Department of Bioengineering, University of Washington, Seattle, WA 98195, USA. The authors would like to thank Tony Ko, Pei-Lin Hsiung, Christine Jesser, Kathleen Saunders, Dr David Golden, Dr Wolfgang Drexler, and Dr Christian Chudoba for their technical and laboratory assistance, and Charlie Pye for his help in coordinating the clinical studies. This research is sponsored by the National Institutes of Health, contracts R01-AR44812 (MEB), R01-EB000419 (MEB), R01 AR46996 (MEB), R01-HL55686 (MEB), R01-EB002638 (MEB), NIH-RO1-HL63953 (MEB), NIH-1-R29-HL55686 (MEB), NIH-9-RO1-EY11289 (JGF), NIH-1-RO1-CA75289 (JGF); by the Medical Free Electron Laser Program, Office of Naval Research Contract Grant N00014-97-1-1066 (JGF and MEB); and by Whitaker Foundation Contract 96-0205 (MEB).
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Schematic drawing of the optical coherence tomography (OCT) system and the imaging probe used. The OCT system **(a)**includes a light source with a broad wavelength distribution (called a low-coherence light source), an interferometer (for dividing/recombining the light), and detection electronics. A compact, pen-sized, handheld probe was used for lateral scanning of the articular cartilage, in conjunction with an aiming beam. The handheld OCT imaging probe **(b)**consists of a four-lens relay and a scanning mirror. The outer shell of the probe can be detached for ease of sterilization. A/D, analog-to-digital converter; VCR, video cassette recorder.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Normal human knee articular cartilage. The optical coherence tomography (OCT) image **(a)**of the cartilage is relatively thick and uniform. The pronounced banding pattern on the OCT image is due to the birefringence of the highly organized structure of the collagen (red arrows). The alternating maximum and minimum intensities are due to changes in back scattering as light travels through the tissue while the plane of light polarization rotates. Previous work has shown that it is due to the presence of organized collagen that alters the polarization state of the light. Note: darker gray scale represents higher-intensity back scattering. The corresponding histology is shown in **(b)**.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Representative optical coherence tomography (OCT) image **(a)**and the corresponding histology **(b)**of mild to moderate osteoarthritic knee cartilage. Regions of lost back scattering are noted in the OCT image. These regions correlate with abnormalities detected on the corresponding histology (b). Areas of hypocellularity are indicated by the red arrows.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
An optical coherence tomography (OCT) image **(a)**and the corresponding histology **(b)**of severely degenerated cartilage. The heterogeneity of the cartilage and loss of the polarization sensitivity are noted. The subchondral bone interface is indicated by either white (a) or red (b) arrows. Black arrows indicate areas in which cartilage is absent with the bone exposed.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Optical coherence tomography (OCT) image **(a)**of cartilage with evidence of early degenerative changes and the corresponding histology **(b)**. Areas of hypocellularity are indicated with red arrows.
:::
:::
::: {#F6 .fig}
Figure 6
::: {.caption}
######
Optical coherence tomography image of cartilage from femur and patella consisting of adjacent areas of normal and diseased tissue. The banding pattern is attenuated and lost in diseased areas (on the right portion of each image). In addition, back-scattering intensity is abruptly reduced.
:::
:::
|
PubMed Central
|
2024-06-05T03:55:54.963930
|
2005-1-17
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065329/",
"journal": "Arthritis Res Ther. 2005 Jan 17; 7(2):R318-R323",
"authors": [
{
"first": "Xingde",
"last": "Li"
},
{
"first": "Scott",
"last": "Martin"
},
{
"first": "Costas",
"last": "Pitris"
},
{
"first": "Ravi",
"last": "Ghanta"
},
{
"first": "Debra L",
"last": "Stamper"
},
{
"first": "Michelle",
"last": "Harman"
},
{
"first": "James G",
"last": "Fujimoto"
},
{
"first": "Mark E",
"last": "Brezinski"
}
]
}
|
PMC1065330
|
Introduction
============
Zymosan, a polysaccharide from the cell wall of *Saccharomyces cerevisiae*, is composed primarily of glucan and mannan residues \[[@B1]\]. *In vitro*, it has served as a model for the study of innate immune responses, because it is capable of stimulating inflammatory cytokine production \[[@B2]\] and can activate complement in the absence of immunoglobulins \[[@B3]\]. Zymosan is recognized and phagocytosed principally by monocytes and macrophages and leads to cellular activation \[[@B4]\]. Zymosan-induced arthritis (ZIA) in mice was first described by Keystone in 1977 \[[@B5]\]. Arthritis was induced by intra-articular injection of zymosan and was thought to be mediated by activation of the alternative pathway of complement and the release of lysosomal hydrolases from activated macrophages \[[@B6]\].
The recent discovery of pattern recognition receptors and their role in innate immunity has led to a re-evaluation of our concepts of zymosan-induced inflammation. Toll-like receptors (TLRs) are a family of type 1 transmembrane proteins that consists of an extracellular leucine-rich repeat domain and a cytoplasmic domain homologous to the cytoplasmic domain of the human interleukin 1 (IL-1) receptor \[[@B7]\]. The ligands of TLR2 include lipopeptides and peptidoglycan \[[@B8],[@B9]\], and TLR2 is a receptor for zymosan, acting in collaboration with CD14 and TLR6 \[[@B2],[@B10]\]. Ligand binding to TLRs induces the activation of NF-κB and the production of the inflammatory cytokines IL-1, IL-6, IL-8, and IL-18 as well as the expression of the co-stimulatory molecule B7.1 \[[@B7]\]. Additionally, zymosan is able to induce maturation of dendritic cells *in vitro*and to stimulate their production of IL-2 \[[@B11],[@B12]\], providing evidence for a link between the innate and the adaptive immune responses.
The inflammatory response triggered by zymosan is linked to its phagocytosis, a process that is mediated by a set of different receptors from the TLRs. The non-opsonic recognition of zymosan by macrophages is mediated by Dectin-1. Dectin-1 is a type 2 membrane receptor with an extracellular C-type lectin-like domain fold and a cytoplasmic immunoreceptor tyrosine-based activation motif \[[@B13]\] and is expressed on macrophages, dendritic cells and neutrophils \[[@B14]-[@B16]\]. Dectin-1 mediates the binding of *Saccharomyces cerevisiae*and *Candida albicans*in a β-glucan-dependent manner and may also have a pro-inflammatory function \[[@B17]\].
In the light of the above findings, we have re-investigated ZIA to elucidate the roles of the innate and adaptive immune responses in this model and to compare the effects of TLR2 deficiency and complement C3 deficiency. The role of Dectin-1 in zymosan-induced inflammation was also investigated. Our results indicate that TLR2 is the major pathway of pro-inflammatory signalling in ZIA and is necessary for the development of specific immune responses to zymosan.
Materials and methods
=====================
Animals
-------
C3-deficient mice (C3^-/-^) on a C57bl/6 background were generated by Professor M Botto \[[@B18]\]. TLR2-deficient mice (TLR2^-/-^) on a C57bl/6 background were provided by Dr Kiyoshi Takeda (Department of Host Defense, Research Institute for Microbial Diseases, Osaka University) \[[@B19]\]. Wild-type (WT) C57bl/6 mice were purchased from Charles River (L\'Arbresle, France). All mice were bred in our animal house facility. Double knockout and double WT mice were generated by mating TLR2^-/-^and C3^-/-^mice. The genotypes of all mice used were confirmed by polymerase chain reaction analysis of genomic DNA extracted from mice tails. The primer sequences used were as follows: TLR2 sense, 5\' -GTTCTCCCAGCATTTAAAATCATT-3\' ; TLR2 antisense, 5\' -GTCTCCAGTTTGGGAAAAGAACC-3\' ; TLR2 NEO antisense, 5\' -CGACACAGCTGCGCAAGCAAC-3\' ; C3 sense, 5\' -CTTCATAGACTGCTGCAACCA-3\' ; C3 antisense, 5\' -AACCAGCTCTGTGGGAAGTG-3\' ; C3 NEO antisense, 5\' -AAGGGACTGGCTGCTATTGG-3\'.
Induction of ZIA
----------------
Zymosan A from *Saccharomyces cerevisiae*(Sigma, St Louis, MO, USA) (300 mg) was resuspended in 10 ml of endotoxin-free saline, boiled and homogenized by sonic emulsification. The suspension was autoclaved and stored in aliquots at -20°C. Arthritis was induced by intra-articular injection of 180 μg (6 μl) of zymosan through the suprapatellar ligament into the joint cavity. In specified experiments, the contralateral knee was injected with an equal amount of sterile saline (6 μl) as control.
Laminarin was co-injected at a dose of either 500 μg or 100 μg together with 180 μg of zymosan into the knee joint.
Approval was obtained from the local animal health committee for these experiments.
Isotopic quantification of joint inflammation *in vivo*
-------------------------------------------------------
Joint inflammation was measured by ^99m^Tc uptake in the knee joint as described \[[@B20]\]. Mice were sedated by the intra-peritoneal administration of sodium pentobarbital (50 mg/kg) and then injected subcutaneously in the neck region with 10 μCi of ^99m^Tc. The accumulation of the isotope in the knee was determined by external gamma-counting after 15 min. The ratio of ^99m^Tc uptake in the inflamed arthritic knee to ^99m^Tc uptake in the contralateral control knee was calculated. A ratio higher than 1.1:1 indicated joint inflammation.
Histological grading of arthritis
---------------------------------
Mice were killed at day 8 and at day 25. Knees were dissected and fixed for 2 weeks in 10% buffered formalin. Fixed tissues were decalcified for 2 weeks in 15% EDTA, dehydrated and embedded in paraffin. Sagittal sections (5 μm) of the whole knee joint were stained with safranin-O and counterstained with fast green/iron hematoxylin. Histological sections were graded by two observers unaware of animal genotype or treatment. Synovial cell infiltrate and exudate were scored from 0 (no cells) to 6 (maximum number of inflammatory cells). Cartilage proteoglycan depletion (damage), reflected by a loss of safranin-O staining intensity, was scored on a scale from 0 (fully stained cartilage) to 6 (totally unstained cartilage) in proportion to severity. For each histopathological measure the score (mean ± SEM) of all slides was calculated.
T cell proliferation assay
--------------------------
Mice were killed in accordance with the experimental protocol. Inguinal lymph nodes were removed and single-cell suspensions were incubated in RPMI supplemented with 2-mercaptoethanol, penicillin, streptomycin and 1% autologous serum. Lymph node cells (LNC; 4 × 10^5^per 200 μl per well) were plated in 96-well flat-bottomed plates and stimulated with zymosan at specified concentrations. Concanavalin A at 4 μg/ml was used as non-specific mitogen.
The cells were incubated for 48 hours at 37°C in 5% CO~2~, then \[^3^H\]thymidine (1 μCi per well) was added to the cultures for 18 hours. The cells were harvested, and \[^3^H\]thymidine uptake was measured with a beta scintillation counter.
Determination of interferon-γ production *in vitro*
---------------------------------------------------
Culture supernatants from LNC cultured with or without 4 μg/ml zymosan were harvested after 72 hours for determination of interferon (IFN)-γ levels. Quantification of cytokine production was performed with an enzyme-linked immunosorbent assay (ELISA) kit specific for murine IFN-γ (Amersham Pharmacia, Dubendorf, Switzerland).
TLR2 immunohistochemistry
-------------------------
Immunohistochemistry was performed with affinity purified anti-mouse TLR2 antibody (clone 6C2; eBioscience, San Diego, CA, USA). Specificity of the antibody was tested on bone marrow cells derived from c57bl/6 TLR2^+/+^and TLR2^-/-^mice.
Dissected knees were embedded in Tissue-Tek OCT, then immediately frozen in precooled hexane and stored at -70°C until use. Sections 7 μm thick were cut on a motor-driven Leica cryostat with a retraction microtome and a tungsten carbide knife at a cabinet temperature of -25°C and mounted on Menzel Super Frost Color glass slides.
Phagocytosis assay
------------------
RAW 264.7 cells (5 × 10^5^to 10^6^per chamber) were plated on a Lab-Tek II Chamber Slide system (Nalge Nunc International). After adherence, cells were either preincubated with 100 or 500 μg/ml laminarin \[[@B21]\] for 20 min followed by the addition of 25 zymosan particles per cell, or laminarin was co-administrated with zymosan. After incubation for 3 hours at 37°C in 5% CO~2~, cells were washed twice with PBS and fixed for 10 min in acetone. Cell-bound and phagocytosed particles were stained by periodic acid Schiff, a stain specific for insoluble glucose polymers, and quantified by light microscopy.
Quantification of IgG levels
----------------------------
Serum levels of total IgG were quantified with ELISA. In brief, rabbit anti-mouse IgG (Dako, Carpinteria, CA, USA) was coated on 96-well plates (Nunc, Roskilde, Denmark). Murine sera from naive and ZIA mice (dilution 1:100,000) were added and incubated for 2 hours. Secondary alkaline-phosphatase-linked anti-mouse IgG (Sigma, Buchs, Switzerland) was added and *p*-nitrophenyl phosphate (Sigma, Buchs, Switzerland) completed the reaction.
Serum levels of specific anti-zymosan IgG were also quantified by ELISA. Zymosan particles at 1 mg/ml were coated on 96-well plates and murine sera (dilution 1:100) were added and incubated for 2 hours. The reaction was developed as previously described.
Statistical analysis
--------------------
The Wilcoxon rank sum test for unpaired variables (two-tailed) was used to compare differences between groups. The unpaired Student *t*-test was used to compare the groups with normally distributed values. A level of *P*\< 0.05 was considered statistically significant.
Results
=======
Zymosan-mediated inflammation in the knee joint is biphasic
-----------------------------------------------------------
In experiments on WT C57bl/6 mice, we observed a biphasic course of inflammation, with an initial peak of ^99m^Tc uptake at day 1 (1.71 ± 0.08), followed by a decrease to a trough value at day 7 (1.29 ± 0.05) and a secondary increase in uptake at day 14. Inflammation measured by ^99m^Tc uptake persisted up to day 25 (1.40 ± 0.06) (Fig. [1a](#F1){ref-type="fig"}).
Histological assessment of the mice at day 8 showed a low score for cellular infiltration (1.00 ± 0.32) and for cartilage destruction (0.7 ± 0.2) (Fig. [1b](#F1){ref-type="fig"}), whereas scoring at day 25 was characterized by an increase in cellular infiltration (2.5 ± 0.37) while cartilage destruction remained low (0.71 ± 0.24) (Fig. [1c](#F1){ref-type="fig"}).
Histology and immune responses at day 25 of ZIA
-----------------------------------------------
To determine whether zymosan particles persisted in the joint at day 25, periodic acid Schiff staining was performed on joint tissues obtained at day 25 and showed persistence of zymosan particles in the synovial membrane of mice injected with zymosan (Fig. [2a](#F2){ref-type="fig"}).
To verify that joint inflammation was associated with the development of specific immune responses to zymosan, we assessed both humoral and cellular responses in WT mice. Proliferation of LNC in response to zymosan was significantly increased in day 25 WT ZIA mice compared with LNC of naive mice (3.5 stimulation index in ZIA WT mice versus 1.5 in naive mice; *P*\< 0.001), whereas mitogenic response to the non-specific mitogen concanavalin A at 4 μg/ml showed no difference between groups (Fig. [2b](#F2){ref-type="fig"}). No difference in proliferation in response to zymosan was observed between ZIA and naive mice at day 8 (data not shown).
The humoral response to zymosan was measured by ELISA. In arthritic mice, the serum levels of anti-zymosan IgG antibodies were significantly increased at day 25 in comparison with those in untreated naive mice (antibody ratio for WT = 0.944 versus naive = 0.677; *P*\< 0.02) (Fig. [2c](#F2){ref-type="fig"}).
In addition, *in vitro*stimulation of WT ZIA LNC with zymosan at 4 μg/ml induced the secretion of IFN-γ at 1200 pg/ml, whereas unstimulated LNC produced undetectable levels of IFN-γ (Fig. [2d](#F2){ref-type="fig"}).
Synovial expression of TLR2 and its role in ZIA
-----------------------------------------------
We wished next to evaluate whether TLR2 might have a role in the recognition of zymosan *in vivo*and in mediating inflammation in ZIA. Specific antibody for TLR2 was used to stain synovium from WT mice that had developed ZIA at day 25. Figure [3a](#F3){ref-type="fig"} shows a representative example of the distribution of TLR2 expression in the synovial cell lining. Control antibody staining was negative (Fig. [3b](#F3){ref-type="fig"}). Antibody specificity was confirmed by a lack of staining in TLR2^-/-^mice (data not shown).
To explore whether the deficiency of TLR2 had an effect on the course of ZIA, we measured knee joint inflammation in TLR2^+/+^and TLR2^-/-^mice by ^99m^Tc uptake at different time points up to day 24 (Fig. [3c](#F3){ref-type="fig"}). In two independent experiments we observed an attenuation of inflammation in TLR2^-/-^mice at days 1, 3, 14, 17 and 24, although only the decrease observed at days 1 and 24 reached statistical significance (*P*\< 0.05).
TLR2 deficiency ameliorates histological features of ZIA
--------------------------------------------------------
We compared the histological features of arthritic knee joints from TLR2^+/+^and TLR2^-/-^mice (Fig. [3d](#F3){ref-type="fig"}). In both groups, arthritis was histologically present in all knees that had been injected with zymosan. In TLR2^+/+^mice, on day 25 of ZIA, the synovial membrane was thickened, mainly as a result of invasion by inflammatory cells (see Fig. [1c](#F1){ref-type="fig"}). In TLR2^-/-^mice, synovial infiltrate was significantly decreased in comparison with TLR2^+/+^mice (4.9 ± 0.33 in TLR2^+/+^mice \[*n*= 15\] versus 3.1 ± 0.67 in TLR2^-/-^mice \[*n*= 12\] on day 25 after arthritis onset; *P*\< 0.045).
TLR2^-/-^mice showed no difference from WT mice in terms of cartilage destruction, as assessed by the loss of safranin-O staining at day 25 (Fig. [3d](#F3){ref-type="fig"}).
Effect of TLR2 deficiency on cellular responses
-----------------------------------------------
The role of TLR2 on the cellular response to zymosan was examined by isolating LNC from ZIA mice. The proliferation of LNC induced by zymosan was significantly lower in cells isolated from TLR2^-/-^mice than in TLR2^+/+^mice. A significant difference was found at both concentrations of zymosan studied (4 and 8 μg/ml; both *P*\< 0.05) (Fig. [3e](#F3){ref-type="fig"}). No differences were observed in proliferation stimulated by the non-specific mitogen concanavalin A (data not shown).
The serum levels of anti-zymosan IgG antibodies, measured by ELISA, were decreased by 50% in TLR2^-/-^mice at day 25 in comparison with the serum levels in controls (antibody ratio for WT = 1.00 versus TLR2^-/-^= 0.51, *P*= 0.047) (Fig. [3f](#F3){ref-type="fig"}).
Lack of effect of C3 on inflammation in ZIA
-------------------------------------------
The availability of C3-deficient mice in a C57bl/6 background allowed us to reassess the role of C3 in ZIA. No effect, either in ^99m^Tc uptake or in histological scoring, was observed in C3-deficient (*n*= 25) mice in comparison with WT mice (*n*= 25). In addition, humoral and cellular responses were similar in C3^-/-^and C3^+/+^mice (data not shown).
Generation of TLR2/C3 double-deficient mice gave similar responses as TLR2^-/-^mice, excluding a synergistic effect of double deficiency and confirming no role for the alternative pathway component of the complement cascade (Fig. [4a](#F4){ref-type="fig"}). Histological scoring showed the presence of arthritis in both groups of animals. In TLR2/C-3 double-deficient mice, synovial infiltrate was significantly decreased in comparison with control (4.0 ± 0.65 in control mice \[*n*= 5\] versus 1.9 ± 0.62 in TLR2/C-3 double-deficient mice \[*n*= 5\] on day 25 after arthritis onset; *P*\< 0.05) (Fig. [4b](#F4){ref-type="fig"}).
TLR2/C-3 double-deficient mice also showed a significantly decreased cartilage destruction in comparison with WT mice at day 25 (1.7 ± 0.12 in control mice \[*n*= 5\] versus 0.9 ± 0.29 in TLR2/C-3 double-deficient mice \[*n*= 5\]; *P*\< 0.05) (Fig. [4b](#F4){ref-type="fig"}).
Stimulation of LNC with zymosan *in vitro*showed a significant decrease of stimulation in double-deficient mice compared with WT littermates, similar to that observed in TLR2^-/-^mice (data not shown).
In addition, a decreased production of zymosan-specific IgGs was observed in the double-deficient mice (ratio for WT = 0.944 versus double knockout = 0.616; *P*\< 0.05) (Fig. [4c](#F4){ref-type="fig"}).
Dectin-1 has a minor role in inflammation in ZIA
------------------------------------------------
The identification of the β-glucan receptor Dectin-1 and its ability to bind zymosan particles *in vitro*stimulated us to study the role of Dectin-1 *in vivo*in ZIA. *In vitro*blockade of the Dectin-1 receptor by laminarin led to a 50% decrease in a phagocytosis assay with RAW 264.7 cells. This decrease was not dependent on the time of administration of laminarin, because it was not modified by preincubation or co-incubation with zymosan particles (Fig. [5a](#F5){ref-type="fig"}).
Co-administration of laminarin and zymosan in the knee joint of C57bl/6 mice showed a trend to a decrease of ^99m^Tc uptake in the early phase of inflammation in a laminarin-treated knee, compared with an untreated knee, at 4 hours and 1 day after administration, but did not reach statistical significance (Fig. [5b](#F5){ref-type="fig"}).
Discussion
==========
For more than 50 years zymosan has been a tool in the study of microbial recognition by the innate immune system. The mechanisms mediating the recognition and phagocytosis of zymosan *in vivo*are complex. Phagocytes, including monocytes, macrophages and dendritic cells, express receptors such as the TLRs, complement receptor 3, scavenger receptors (such as acetylated LDL receptors) and Dectin-1 \[[@B22]-[@B24]\], which have all been implicated in the cellular response to zymosan \[[@B25]\]. In addition, zymosan is capable of activating the alternative pathway of complement through C3 \[[@B3]\], which may serve to amplify the inflammatory response.
To elucidate how zymosan induces inflammation *in vivo*, we re-investigated the ZIA model that was first studied in the 1970s. This model has been often used as a tool to dissect non-immune mechanisms of joint inflammation \[[@B26]-[@B28]\]. In our experiments we observed that ZIA was not as short lived as originally described. Arthritis persisted beyond day 14 and in fact beyond day 25. After an initial peak of inflammation at about day 3, inflammation subsided by day 7. Subsequently, inflammation returned to levels that could be as high (as measured by ^99m^Tc uptake) as the initial peak, suggesting that ZIA has early and late phases. Histologically, the joint inflammation was characterized by mononuclear cell infiltration in the sublining layer and hypertrophy of the lining layer as well as cartilage damage. Histological changes were milder at day 8 than at day 25. Zymosan particles were present in the synovium at day 25.
We then investigated the role of TLR2 in ZIA because the macrophage inflammatory response to zymosan depends largely on its recognition by a heterodimer of TLR2 and TLR6 \[[@B2],[@B10]\]. In TLR2^-/-^mice there was a significant attenuation of the early and late inflammatory phases of ZIA, indicating that a ligand that activates the innate immune response through TLR2 can lead to a chronic local inflammatory reaction.
In the absence of TLR2, joint inflammation was not totally blocked. This would suggest that, *in vivo*, the inflammatory response to zymosan is not dependent on TLR2 signalling alone and that receptors other than TLR2 might have a role. This is supported by the observation that inhibition of TLR2 and MyD88 by dominant-negative mutants blocked pro-inflammatory signalling but not zymosan uptake *in vitro*. Recent data have shown that Dectin-1 and SIGNR1 \[[@B29]\] on macrophages and pentraxin-3, an opsonin for the recognition of zymosan by Dectin-1, are involved in zymosan recognition and internalization \[[@B30]\]. We therefore investigated the role of Dectin-1 by using the β-glucan laminarin as a competitive inhibitor of zymosan \[[@B16]\]. We confirmed that laminarin inhibited zymosan uptake by RAW 264.7 cells and did not observe any difference in the blocking capacity of laminarin, whether administered before or at the same time as zymosan. In both cases and at two different concentrations, we observed a 50% decrease in cell-bound zymosan particles. On the basis of these results, we proceeded to assess the effect of laminarin on ZIA. Although there was a trend towards reduced ^99m^Tc uptake in the treated animals, this was not statistically significant. It is possible that 50% inhibition of zymosan phagocytosis is insufficient to modulate inflammatory signalling through TLR2. Furthermore, a redundancy in the multiple mechanisms that mediate zymosan phagocytosis could also explain the lack of effect of laminarin inhibition *in vivo*\[[@B31]\].
The biphasic course of ZIA and its modulation by TLR2 led us to study the acquired immune response to zymosan and the effects of TLR2 deficiency on it. We compared the cellular proliferative and antibody responses to zymosan in WT and TLR2^-/-^mice at day 25. First, we were able to detect zymosan-induced lymphocyte proliferation and enhanced IFN-γ production in the draining LNC of mice with ZIA, and second, this was accompanied by the formation of a zymosan-specific IgG. In TLR2^-/-^animals, the proliferative response was blunted and only reached 50% of that observed in WT ZIA animals. There was also a significant decrease in the zymosan-specific IgG response, which was about 50% lower than in WT mice. At day 8 we did not observe any difference between ZIA and naive WT mice in their proliferative response to zymosan (data not shown). These findings suggest that inflammation in the later phase of ZIA is paralleled by the development of acquired immune response to zymosan. The finding that zymosan particles persisted in the joint even at day 25 suggests that they could become a target for specific immune responses. The decrease in acquired immune responses in TLR2^-/-^mice might be a result of the decreased antigen presentation efficiency of dendritic cells in the absence of TLR2 \[[@B32]\] or the lack of co-stimulatory signals through TLR2 expressed on activated T cells \[[@B33]\].
A significant role for the alternative pathway of complement in this model of inflammation was excluded by the phenotype observed in C3^-/-^mice. Both phases of ZIA were comparable to that observed in WT controls. Mice with combined deletions of the C3 and the TLR2 genes did not show a significant decrease in ^99m^Tc uptake in comparison with TLR2^-/-^mice. Histologically, we observed a significant decrease in cartilage damage in mice with the combined deficiency of TLR2 and C3, which did not occur in C3^-/-^mice. We interpret this effect to be due in principle to the lack of TLR2, because TLR2^-/-^mice also showed a diminished cartilage score (although it did not reach statistical significance). Combined with recent data showing that phagocytosis of zymosan is not mediated by complement receptor 3 \[[@B16]\], complement activation does not seem to contribute to zymosan-induced joint inflammation *in vivo*.
The expression of TLR2 in arthritic synovium from WT mice in the ZIA model and the modulation of joint inflammation in TLR2^-/-^animals show that TLR2 may have a general role in amplifying local inflammation. TLR2 has been shown to be expressed on neutrophils and lymphocytes as well as macrophages, and they all are participants in the inflammatory process in this model. The data in human arthritis would also support such a role for TLR2. Increased expression of TLR2 in synovial lining layer and by CD16^+^peripheral blood mononuclear cells in RA indicate that its expression is upregulated during chronic inflammation \[[@B34]\]. TLR2 is also expressed on RA synovial fibroblasts, and incubation of cultured RA synovial fibroblasts with pro-inflammatory cytokines increases levels of TLR2 mRNA \[[@B35]\]. Although the precise role of TLR signalling in RA is unclear at present, increased TLR2 expression might modulate synovial inflammation if endogenous or exogenous TLR2 ligands gain access to the joint, thus amplifying specific and innate immune pathways of synovial inflammation. Furthermore, our results provide a model by which stimulation of the innate immune response can lead to chronic inflammation in the joint. These pathways may be of relevance to the development of reactive arthritis in man.
Conclusion
==========
The results of the present study indicate that the biphasic joint inflammation in ZIA is mediated primarily by activation of the innate immune system. In the early phase of arthritis TLR2 plays a vital role, and in the later phase the development of a secondary immune response to zymosan may contribute to joint inflammation. Innate immune responses may be important amplificatory pathways of joint inflammation in man.
Abbreviations
=============
ELISA = enzyme-linked immunosorbent assay; IFN = interferon; IL = interleukin; LNC = lymph node cells; TLR = Toll-like receptor; WT = wild-type; ZIA = zymosan-induced arthritis.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
MF contributed to breeding and genotyping, performed technetium uptake measurements and immunoassays, and participated in coordination of the study. DT participated in technetium uptake measurements. VC performed histological stainings and scoring. NB performed statistical analysis and participated in the design of the study. AS conceived of the study, participated in its design and coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.
Acknowledgements
================
We thank Professor Marina Botto (Division of Medicine, Imperial College London, UK) for the gift of the C3^-/-^mice, and Dr Didier Le Roy (Laboratoire de Maladies Infectieuses, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland) and Professor Kiyoshi Takeda (Department of Host Defense, Research Institute for Microbial Diseases, Osaka University) for providing us with the TLR2^-/-^mice.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Biphasic kinetic of inflammation in zymosan-induced arthritis (ZIA). **(a)**^99m^Tc uptake measurement shows a biphasic course of inflammation in wild-type (WT) mice. Zymosan was injected into the right knee and PBS was injected into the contralateral control knee. The values obtained correspond to the ratio between the right and left knee (*n*= 15). The severity of histological signs of arthritis was assessed by scoring synovial thickness and cartilage destruction on a scale from 0 to 6. Results are expressed as means ± SEM. **(b)**WT mice with ZIA showed mild inflammation at day 8 as judged by cellular infiltration (histological score 1.00 ± 0.32) and cartilage destruction (0.7 ± 0.2; *n*= 10), which became more severe at day 25. **(c)**There was an increase in cellular infiltration (2.5 ± 0.37) but cartilage destruction remained slight (0.71 ± 0.21; *n*= 10). The arrows in (b) and (c) indicate inflammatory infiltrate in the synovial membrane.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Local and immune responses to zymosan at day 25. **(a)**Periodic acid Schiff staining shows the persistence of zymosan particles (arrow) within the synovial membrane in zymosan-induced arthritis (ZIA). Original magnification ×100. **(b)***In vitro*lymph node mononuclear cell proliferation in response to zymosan at day 25 of ZIA and in naive mice. Single-cell suspensions were incubated with 4 μg/ml zymosan. Concanavalin A (ConA) was used as a non-specific mitogen. **(c)**Antibody production against zymosan in zymosan-treated mice compared with naive mice measured by enzyme-linked immunosorbent assay (ELISA). The results are expressed as a ratio of the amount of zymosan-specific IgGs (in arbitrary units) in murine serum to total IgGs (also in arbitrary units) (*P*\< 0.05). **(d)**Interferon-γ (IFN-γ) production by zymosan-stimulated and unstimulated lymph node cells of WT ZIA and naive mice in culture. IFN-γ was measured by specific ELISA. n.d., not detectable.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
TLR2 mediates an inflammatory response in zymosan-induced arthritis (ZIA). **(a)**Immunohistochemistry of TLR2 expression in synovial membrane sections of WT mice (day 25 ZIA) showed staining in the sublining (sl) of synovial membrane, whereas inflammatory lymphocytes were not stained. **(b)**Negative control was performed with 0.5% bovine serum albumin. (100× magnification). **(c)**^99m^Tc uptake measurement showed an attenuation of the inflammatory response at days 1 and 25 (*P*\< 0.05) in TLR2^-/-^ZIA mice (*n*= 5) compared with WT ZIA mice (*n*= 5). The technetium ratio was measured as detailed in the Materials and methods section. **(d)**Histological scoring at day 25 of ZIA showed a significant decrease in cell infiltration in TLR2^-/-^mice (*P*\< 0.02) compared with control mice, whereas cartilage destruction was similar in both groups. **(e)**Lymph node cell proliferation to 4 and 8 μg/ml zymosan (Zym 4 and Zym 8, respectively). **(f)**Antibody production against zymosan in zymosan-treated TLR2^+/+^and TLR2^-/-^mice.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
C3 has no role in mediating inflammatory responses in zymosan-induced arthritis. **(a)**^99m^Tc uptake measurement showed an attenuation of inflammation at early and late time points in TLR2/C3 double-deficient (KO) mice (*n*= 5) compared with double wild-type (WT) mice (*n*= 5) but did not reach statistical significance. **(b)**Histological scoring showed a significant decrease in both cell infiltration (*P*\< 0.05) and cartilage destruction (*P*\< 0.05) in TLR2/C-3 double-deficient mice (*n*= 5) compared with littermate control (*n*= 5). **(c)**Production of specific anti-zymosan IgGs was reduced in double-deficient mice in comparision with double WT littermates. Values correspond to the ratio of zymosan-specific IgG to total IgG mentioned in Fig. 2c.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Dectin-1 inhibition *in vitro*and *in vivo*. **(a)**In a phagocytosis assay, 2.5 × 10^6^particles of zymosan were incubated with 10 × 10^5^RAW 264.7 cells. Laminarin was administered either before (pre adm.) or at the same time as (co adm.) the zymosan particles. Incubation of zymosan particles was performed for 4 hours at 37°C. Laminarin decreased the amount of zymosan bound to and phagocytosed by RAW 264.7 cells by 50% in comparison with a control to which no laminarin had been added (*P*\< 0.05). **(b)**Laminarin (500 μg) and zymosan (180 μg) in a final volume of 15 μl were co-injected into the right knee joint of C57bl/6 mice (*n*= 6). Phosphate-buffered saline (15 μl) was injected into the contralateral knee. ^99m^Tc uptake was measured up to day 7. Comparison of ^99m^Tc uptake with control mice (*n*= 8), which received zymosan alone in the right knee, showed an attenuation of inflammation at 4 and 24 hours in the laminarin-treated mice, but did not reach statistical significance. After 72 hours the values were similar.
:::
:::
|
PubMed Central
|
2024-06-05T03:55:54.965923
|
2005-1-21
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065330/",
"journal": "Arthritis Res Ther. 2005 Jan 21; 7(2):R370-R379",
"authors": [
{
"first": "Matthias E",
"last": "Frasnelli"
},
{
"first": "David",
"last": "Tarussio"
},
{
"first": "Veronique",
"last": "Chobaz-Péclat"
},
{
"first": "Nathalie",
"last": "Busso"
},
{
"first": "Alexander",
"last": "So"
}
]
}
|
PMC1065333
|
Introduction
============
Rheumatoid arthritis (RA) is an inflammatory joint disease characterized by hyperplasia of synovial tissue and pannus formation growing invasively into the cartilage, followed by cartilage and bone destruction. Analyses of hyperplastic synovial tissues of patients with RA reveal features of transformed long-living cells such as the presence of somatic mutations, expression of oncogenes, and resistance to apoptosis \[[@B1]-[@B3]\]. Resistance to apoptosis further contributes to synovial hyperplasia and is closely linked to the invasive phenotype of synovial fibroblasts \[[@B4],[@B5]\].
Apoptosis is a tightly regulated process of elimination of aged cells without disrupting cellular integrity (reviewed in \[[@B6],[@B7]\]). Apoptosis may be initiated by extracellular stimuli through activation of death receptors on the cell surface, and intracellularly by the release of mitochondrial cytochrome c into the cytoplasm. Both pathways induce expression of apoptosis genes and activation of the caspase cascade, resulting in DNA fragmentation. The apoptosis signals are abrogated by the family of apoptosis-inhibiting proteins (IAPs).
A number of disturbances in the apoptosis machinery have been pointed out in RA patients. Fibroblasts from RA synovia are relatively resistant to apoptosis induced by extracellular Fas stimulation. Moreover, co-culture of synovial fibroblasts from RA joints with T cells and B cells induces anergy of lymphocytes. Increased levels of soluble Fas in RA synovial fluid have been suggested as one possible explanation for this fact \[[@B8]\]. Indeed, administration of antagonistic anti-Fas antibodies or of Fas ligand has been shown effective in abrogation of arthritis in animal models \[[@B9],[@B10]\]. Resistance to Fas-induced apoptosis in RA synovium correlates with a markedly increased expression of sentrin-1 \[[@B11]\]. Sentrin-1/SUMO is a molecule whose binding to a protein results in the prevention of ubiquitin-related processing and degradation of that protein. Sentrin-mediated protection has been shown for such proteins as p53 and IkBa. Upregulation of anti-apoptotic molecules belonging to the Bcl family and of the caspase-8 inhibitor FLIP has been repeatedly reported in RA \[[@B12]\]. Inhibited apoptosis has been shown to contribute to the pathogenesis of experimental arthritis \[[@B13],[@B14]\].
Survivin is a 142-amino-acid protein that belongs to the IAP family, and it inhibits the activity of caspase 3, caspase 7, and caspase 9, but not of the upstream initiator protease caspase 8. Survivin can thereby downregulate, directly or indirectly, both death-receptor-mediated and mitochondria-mediated pathways of apoptosis \[[@B15]\]. Survivin has been also suggested to regulate cell division during mitosis. Indeed, survivin is the only one of IAPs that is tightly connected to the cell cycle being upregulated in the G~2~/M phase. Inside the dividing cell, survivin is found incorporated in centrosomes and mitotic spindles, and relocates to midbodies in the late telophase. Disruption of survivin function by negative mutation or by introduction of anti-sense oligonucleotides results in a cell-division defect \[[@B16],[@B17]\]. Survivin is abundantly expressed in all the most common human cancers and in transformed cell lines \[[@B15]\], while most normal differentiated adult tissues do not express this molecule. A few adult tissues reported to express survivin include the spleen, the testes, the thymi, the placentas, and the colonic crypts.
In the present study we demonstrate high levels of the anti-apoptotic protein survivin extracellularly in plasma and synovial fluid of patients with RA. In all the cases but one, high levels of survivin were associated with the erosive type of joint disease. Moreover, it is demonstrated that autoantibody responses to survivin led to a more benign (non-erosive) course of RA. The latter finding may have potential therapeutic consequences.
Methods
=======
Participants
------------
Plasma and synovial fluid samples were collected from 131 RA patients who attended the rheumatology clinics at Sahlgrenska University Hospital, Göteborg for acute joint effusion. RA was diagnosed according to the American College of Rheumatology criteria \[[@B18]\]. At the time of synovial fluid and blood sampling all the patients received non-steroidal anti-inflammatory drugs.
Disease-modifying anti-rheumatic drugs (DMARDs) were used by 96 patients, 67 of which used methotrexate (MTX). Forty-two of these 67 patients combined medication of MTX with the inhibitors of tumour necrosis factor alpha (TNF-α), two other patients combined MTX with sulfasalazine, one patient combined it with cyklosporine A, and the remaining 22 patients were treated with MTX alone. DMARDs other then MTX were used by 14 patients, six patients were treated with sulfasalazine, five patients were treated with cyklosporine A (one patient in combination with azathioprine, one patient with leflunomide, two with sulfasalazine, and the remaining patient with infliximab), four patients used parenteral or oral gold salt compounds, one patient used leflunomide, and one patient used azathioprine. The inhibitors of TNF-α were used in 47 patients (42 patients in combination with MTX, three patients in combination with azathioprine, one patient in combination with cyklosporine, and the remaining patient in combination with cyclophosphamide). The remaining 35 of 131 patients had no DMARD treatment at the time of blood and synovial fluid sampling.
Recent radiographs of the hand and foot skeletons for all patients were studied. The presence of bone erosions, defined as the loss of cortical definition at the joint, was recorded in proximal interphalangeal joints, metacarpophalangeal joints, carpus joints, wrist joints, and metatarsophalangeal joints. The presence of one erosion was sufficient to fulfil the requirement of an erosive disease. We considered the presence of rheumatoid factor (RF) of any of the immunoglobulin isotypes as positive. Informed consent was obtained from the patients and the controls. The study was approved by the Ethics Committee of Sahlgrenska University Hospital.
Analyses of survivin and antibodies to survivin
-----------------------------------------------
Synovial fluid samples were obtained by arthrocentesis of knee joints. Synovial fluid was aspirated aseptically and transferred into tubes containing sodium citrate (0.129 mol/l; pH 7.4). We obtained blood samples simultaneously from the cubital vein and directly transferred them into sodium citrate medium. Blood samples from healthy individuals (*n*= 34; age range, 18--62 years; mean age, 42 ± 7 years) were used as controls. Collected blood and synovial fluid samples were centrifuged at 800 × *g*for 15 min, aliquoted, and stored frozen at -20°C until use.
Survivin levels were determined by a sandwich ELISA using a pair of matched antibodies (rabbit anti-human survivin; R&D Systems, Stockholm, Sweden). Briefly, 96-well polystyrene dishes (Nunc, Roskilde, Denmark) were coated with capture antibodies and were left overnight at room temperature. Following washing, plates were blocked with PBS--BSA containing 5% sucrose. Matched samples of plasma and synovial fluid were introduced into the parallel strips, at a dilution of one in 10 in PBS--BSA. Horseradish peroxidase-labelled detection antibodies and the corresponding substrate were used for colour development. Double-wave reading at 450 and 570 nm was used and the difference of absorbances was calculated. The obtained absorbance values were compared with the serial dilution of recombinant survivin and are presented as picograms per millitre.
Antibodies of IgG and IgM class specific for survivin were measured in blood and synovial fluid samples by an ELISA. Briefly, 96-well polystyrene dishes (Nunc) were coated with human recombinant survivin (R&D Systems). Reconstituted survivin (0.5 μμg/ml) was introduced in each well and left overnight at room temperature. Following washing with PBS containing 0.1% Tween-20, plates were blocked with 1% ovalbumin (Sigma, St Louis, MO, USA) in PBS for 2 hours at room temperature. Matched samples of plasma and synovial fluid were introduced into the parallel strips, in a dilution of one in 100 using PBS--1% ovalbumin. This dilution was established as being on a linear scale in preliminary titration experiments. Horseradish peroxidase-labelled detection antibodies (rabbit F(ab\')2-anti-human IgG and IgM; Dako, Glostrup, Denmark), ExtrAvidin peroxidase conjugate (Sigma), and the corresponding substrate were used for colour development. The absorbance at 405 nm was registered. Absorbances of the patient samples were compared with the mean values obtained in the control group of healthy individuals.
Interaction with survivin transcription
---------------------------------------
Peripheral blood mononuclear cells (PBMC) were prepared from heparinized blood of healthy individuals by separation on a Lymphoprep density gradient. We washed the cells, and resuspended in complete medium (Iscoves medium containing 1% l-glutamine, 5 × 10^-5^M β-mercaptoethanol, 50 μg/ml gentamycin sulphate, and 10% heat-inactivated FCS). We cultured PBMC in 24-well plates in a humidified atmosphere of 5% CO~2~at 37°C. In addition, we cultured the human monocytic cell line THP-1 (American Type Culture Collection, Manassas, VA, USA) in 10-ml culture flasks (Nunc) in RPMI 1649 medium supplemented with 10% FCS, 1% sodium pyruvate, gentamycin, and 2.5% Hepes in a humidified atmosphere of 5% CO~2~at 37°C. For the experiments, 4-day-old THP-1 cells were harvested, washed, and adjusted to 1 × 10^6^cells/ml.
For the transfection experiments, phosphorothioated oligonucleotides containing the anti-sense-targeting human survivin gene \[[@B19]\] were synthesized by MWG Oligo (Ebersberg, Germany). The following anti-sense sequences were used: aSur 1, 5\'-CCCAGCCTTCCAGCTCCTTG-3\' ; and aSur 2, 5\'-GCACCTAGTCTCCCTGCACC-3\'. Irrelevant non-sense sequences were used as controls: non-sense 1, 5\'-GTCCTCCACTGGCCTCACTC-3\' ; and non-sense 2, 5\'-CCCCGATTCACCTCGTCCGT-3\'. Oligonucleotides were delivered to THP-1 cells using oligofectamine reagent (Invitrogen, Carlsbad, CA, USA). Before the transfection procedure we seeded THP-1 cells in 96-well tissue culture plates and cultured them overnight in RPMI medium free of antibiotics and FCS. Transfection was performed in RPMI medium supplemented with 2.5% Hepes and 100 mg/ml CaCl~2~. We mixed 0.6 μl oligofectamine with diluted oligonucletides and added it to the washed THP-1 cells. Following 4 hours of incubation at 37°C in a CO~2~incubator, the transfection procedure was discontinued by adding RPMI medium containing a threefold excess of FCS. At this time point, we also stimulated the cells with phytohaemagglutinine (PHA) (1.5 μg/ml) if required. Following 48 hours of stimulation, THP-1 cultures were aseptically collected, centrifuged at 1000 × *g*for 5 min, and kept frozen at -20°C until analysis. We prepared cell lysates by incubating the cell pellet for 1 hour in 1 mM EDTA buffer containing 6 M urea and proteinase inhibitors (Complete MiniTab; Boehringers, Ingelheim, Germany). These preparations were assessed for proliferation, survivin expression, and IL-6 levels.
Cell survival and apoptosis in the transfected cultures were assessed by surface expression of annexin V and propidium iodide intake. Following transfection and stimulation for 48 hours, THP-1 cells were washed and stained with FITC-marked anti-annexin V antibodies and were subjected to flow cytometry (FACSort; Becton Dickinson, San Jose, CA, USA). The results were analysed using the CELLQuest software (Becton Dickinson).
Proliferation of THP-1 cells was assessed by incubating the cell suspension with the test substance for 48 hours. The cells were then pulsed for 12 hours with 1 μCi \[^3^H\]thymidine (specific activity, 42 Ci/mmol; Amersham, Bucks, UK). Cells were collected onto glass fibre filters. Thymidine incorporation was measured in a beta-counter. We compared the counts obtained in cells transfected with survivin anti-sense oligonucleotides and those incubated with oligofectamine alone. The results were expressed as a percentage.
The level of IL-6 in supernatants was assessed by a bioassay. The effect of test samples on proliferation of the IL-6-dependent cell line B13.29 \[[@B20]\] was assessed following 72 hours of culturing. The results were analysed by incorporation of \[^3^H\]thymidine (Amersham) during the last 4 hours of incubation at 37°C. Cells were collected onto a glass fibre filter. Proliferation in the presence of test samples was compared with that induced by standard dilutions of recombinant IL-6 (Genzyme, Cambridge, MA, USA). The results were further recalculated as in the proliferation assay.
Statistical analysis
--------------------
We expressed the level of survivin and antibodies against survivin in the blood, in synovial fluid samples, as well as in cell lysates as the mean ± standard error of the mean. The survivin levels in the matched blood and synovial fluid samples were analysed by the paired Student *t*test. We further performed a comparison of survivin levels between the patient blood samples and the healthy controls using the Mann--Whitney U test.
We stratified the patient material according to radiological findings (erosive RA versus non-erosive RA) and calculated the difference in survivin levels between the groups employing the Mann--Whitney U test. An arbitrary level of survivin corresponding to three standard deviations of the control group (300 pg/ml) was chosen as a cut-off. The RA patients were further stratified as having \'high\' (\>300 pg/ml) or \'low\' (\<300 pg/ml) levels of survivin. We performed the evaluation of survivin as a prognostic factor for the development of joint destruction, comparing the group having \'high\' and \'low\' survivin levels in a multivariate analysis. In order to control for the role of other prognostic factors (RF, disease duration, age, presence of antibodies against survivin), a multivariate logistic regression was performed. Odds ratios (with 95% confidence interval) are given for descriptive purposes. All tests were two-tailed and conducted at the 5% significance level.
We evaluated a possible influence of the ongoing treatment on the survivin levels, and we stratified patient material according to DMARD treatment (treated versus untreated). For the simultaneous comparison of the survivin levels in more than two groups the equality of variance *F*test was employed. The inter-relation between the survivin levels and duration of the joint disease, age, white blood cell (WBC) count, and C-reactive protein was calculated employing the Spearman correlation coefficient. For all the statistical evaluation of the results, *P*\< 0.05 was considered significant. All statistical evaluations were performed using StatView PowerPC software.
Results
=======
Clinical and demographic data of the patient population and the control group are presented in Table [1](#T1){ref-type="table"}. The patient group showed no difference regarding gender compared with controls, while individuals from the control group were younger (*P*\< 0.05). After stratification of the RA patients with respect to radiological changes, the group with erosive joint disease (ERA) was, as expected, more often positive for RF compared with the group for non-erosive joint disease (NRA) (91% versus 23%, *P*\< 0.0001), and had longer duration of RA (*P*= 0.0002) as compared with NRA patients. With respect to treatment, 68% of ERA patients were treated with MTX, and 48% in combination with TNF-α inhibitors. Among NRA patients, only 28% were treated with MTX (*P*\< 0.025), and 12% with TNF-α inhibitors. NRA patients were significantly more often without DMARDs at the time of blood sampling compared with ERA patients (63% versus 20%, *P*\< 0.0001).
Extracellular survivin determines the erosive course of RA
----------------------------------------------------------
Plasma of the RA patients contained significantly higher levels of survivin as compared with the controls (330 ± 123 pg/ml versus 121 ± 2 pg/ml, *P*= 0.002). Survivin levels in plasma correlated strongly to their levels in synovial fluid (*r*= 0.89). Evaluation of the survivin level was performed in RA patients with respect to the erosivity of joint disease (Fig. [1](#F1){ref-type="fig"}). Patients with ERA had a significantly higher level of survivin compared with NRA patients in plasma (430 ± 108 pg/ml versus 127 ± 5 pg/ml, *P*= 0.0022) and in the synovial fluid (434 ± 181 pg/ml versus 124 ± 2 pg/ml, *P*= 0.0029). The levels of survivin did not differ significantly between the patients positive for RF (*n*= 90) and those who were RF-negative (*n*= 41) (418 ± 107 pg/ml versus 151 ± 20 pg/ml, not significant). Survivin levels showed no significant correlation to the serum levels of C-reactive protein and WBC count, and neither to the synovial fluid leukocyte count and IL-6 levels.
The RA patients were further stratified as having \'high\' (\>300 pg/ml) or \'low\' (\<300 pg/ml) levels of survivin, departing from the level of survivin that corresponded to a mean + three standard deviations of the control group as a cut-off. The difference in the mean survivin level between the \'high\' and the \'low\' groups was about 10-fold (1180 ± 309 pg/ml versus 97 ± 9 pg/ml). High levels of survivin were detected in 28 of 131 patients (21%). All but one (96%) of the patients with a high survivin level displayed erosive RA. A dominance of a high survivin level among the ERA patients was consequently found both in plasma and in synovial fluid samples. Comparison between the ERA patients having high and low levels of survivin (Table [2](#T2){ref-type="table"}) revealed, beside erosivity, an association between high levels of survivin and increased circulating C-reactive protein as well as elevated WBC counts. In contrast, age, gender, RF-positivity, and duration of the disease were similar in the ERA patients with high levels of survivin as compared with those with low levels.
The level of survivin was also studied in RA synovial fluid samples separated with respect to the cell pellet and the supernatant by centrifugation (*n*= 9). Survivin levels found in supernatants and in the lysates of synovial fluid cells obtained from the same sample revealed a strong correlation (*r*= 0.87, *P*\< 0.0001). These data indicate that survivin is produced and secreted locally in the joints of RA patients.
To evaluate the predictive value of high survivin levels for the development of destructive joint disease, a logistic regression model was constructed, taking erosive changes at radiological examination of the hand and foot skeletons as a dependent variable. We found that high levels of survivin were significantly associated with erosive changes (odds ratio, 18.76; 95% confidence interval, 2.45--143.65; *P*= 0.0048). To determine whether survivin was independently associated with erosive RA, we developed a multivariate logistic regression model with radiological changes as the dependent variable and with RF, duration of RA, gender, and the survivin level as independent variables. After adjusting for the presence of RF, gender, and the duration of RA, a high level of survivin was significantly associated with erosive RA (odds ratio, 16.02; 95% confidence interval, 2.02--127.19; *P*= 0.013). Our data thus demonstrate that RA patients having high levels of survivin are 16 times more likely to develop erosive joint disease compared with those with low levels of survivin.
Taking into account the fact that the increased survivin levels were observed predominantly among the ERA patients, we assessed the effect of DMARD treatment on survivin levels in this patient group. To analyse the putative influence of anti-rheumatic treatment on the level of survivin, ERA patients were stratified with respect to their treatment modality at the time of sampling into three groups. Group 1 included patients receiving MTX (*n*= 18), group 2 included patients treated with combination of MTX and TNF-α inhibitors (*n*= 42), group 3 included patients treated with DMARDs other than MTX (*n*= 10), and group 4 included patients having no treatment with DMARD at the time of sampling (*n*= 18) (Fig. [2](#F2){ref-type="fig"}). The highest level of survivin, both in blood and in synovial fluid, was found in the group of patients having no DMARD at the time of sampling (blood, 666 ± 473 pg/ml and synovial fluid, 830 ± 610 pg/ml, respectively). This was significantly higher than in the patients treated with MTX (322 ± 174 pg/ml, *P*= 0.02) and in the patients treated with other DMARDs (280 ± 82 pg/ml, *P*\< 0.001). These three groups of patients were similar with respect to the duration of the disease, age, WBC counts in blood and synovial fluid, and levels of C-reactive protein. Patients treated with combination of MTX and TNF-α inhibitors exhibited no significant difference in survivin plasma levels compared with the patients treated with MTX alone. This was despite the fact that patients obtaining TNF-α inhibitors were younger (*P*\< 0.05) and had lower levels of WBC and C-reactive protein (*P*\< 0.05).
Autoantibodies specific for survivin relate to the non-erosive course of RA
---------------------------------------------------------------------------
An ELISA was used for the evaluation of antibodies against survivin of IgG and IgM isotypes in plasma and in synovial fluid of 129 patients with RA and of 34 healthy controls. The absorbance values revealed a significantly higher antibody reactivity with human recombinant survivin in the case of RA patients compared with the controls (Fig. [3](#F3){ref-type="fig"}). This was true both for IgG (0.19 ± 0.02 versus 0.11 ± 0.012, *P*= 0.022) and for IgM (0.60 ± 0.03 versus 0.28 ± 0.03, *P*\< 0.0001) isotypes of antibodies. There was a weak, although significant, correlation between the antibodies of IgG and IgM isotypes in blood (*r*= 0.389, *P*\< 0.001), but not in synovial fluid (*r*= 0.146, not significant). No significant difference in the IgG antibody levels was found between blood and synovial fluid (0.19 ± 0.02 versus 0.20 ± 0.03, not significant), while the level of IgM antibodies was significantly higher in blood samples than in synovial fluid samples (0.60 ± 0.03 versus 0.43 ± 0.03, *P*= 0.031).
Stratification of the patient material with respect to radiological changes revealed that the level of antibodies against survivin was higher in NRA patients compared with ERA patients (Fig. [4](#F4){ref-type="fig"}). The difference was most pronounced in synovial fluid samples (IgG, 0.18 ± 0.02 versus 0.22 ± 0.02, *P*= 0.038; IgM, 0.31 ± 0.03 versus 0.59 ± 0.03, *P*= 0.0007). Among the ERA patients, a distinct group of patients with high extracellular levels of survivin was outlined. These patients had significantly higher levels of antibodies against survivin both in blood (IgG, 0.25 ± 0.02 versus 0.15 ± 0.02, *P*\< 0.0001; IgM, 0.64 ± 0.03 versus 0.55 ± 0.03, not significant) and in synovial fluid (IgG, 0.21 ± 0.02 versus 0.16 ± 0.02, not significant; IgM, 0.40 ± 0.03 versus 0.27 ± 0.03, *P*= 0.023) as compared with those ERA patients with low survivin levels. However, no significant correlation between the level of extracellular survivin and the level of antibodies against survivin was observed (*r*= 0.05).
Influence of survivin expression on inflammatory responses
----------------------------------------------------------
PBMC from healthy individuals and from RA patients were stimulated with various B-cell and T-cell mitogens, superantigen, and TNF-α (10--100 ng/ml lipopolysaccharide, 0.5--5 μg/ml Concanavalin A, 10--100 ng/ml TNF-α, 10--100 ng/ml TSST-1, 0.5--5 μg/ml PHA) for 6--48 hours. Supernatants and cell lysates were evaluated for survivin expression by an ELISA. Detectable levels of survivin were not found in supernatants. In the cell lysates, levels of survivin varied in response to the aforementioned stimuli (Fig. [4](#F4){ref-type="fig"}). In the tested panel, the T-cell mitogen PHA was found to be a potent inducer of survivin expression both by PBMC originating from RA patients (*n*= 3) and by PBMC from healthy controls (*n*= 6)
Stimulation of THP-1 with PHA was therefore used in the subsequent transfection experiments. To assess the role of survivin in the inflammatory process, the human mononuclear cell line THP-1 was transfected with oligonucleotides targeting different regions of survivin mRNA. Oligonucleotides were delivered in complex with oligofectamine as described in Materials and methods. Successful transfection with the inhibitory sequence was confirmed by a downregulation of survivin expression in THP-1 lysates as assessed by ELISA. THP-1 cells displayed, as expected, high spontaneous intracellular expression of survivin, which correlated well with their proliferative activity.
Following the transfection procedure, cells were stimulated with PHA (1.5 μg/ml) for 48 hours and the cultures were assessed for proliferation and secretion of IL-6. Two different anti-sense sequences were tested, and both anti-sense oligonucleotides downregulated survivin expression (from 100% to 30--44%, *P*\< 0.05). In contrast, non-sense oligonucleotides showed no significant suppression of survivin expression as compared with the THP-1 cultures incubated with oligofectamine alone (Fig. [5a](#F5){ref-type="fig"}). In the THP-1 cultures displaying suppressed survivin expression, a significant downregulation (*P*\< 0.01) of IL-6 production was observed, decreasing from 100% to 21--30% (Fig. [5c](#F5){ref-type="fig"}). To assess whether low survivin expression was related to apoptosis and cell death in the transfected cell cultures, cell proliferation and the expression of annexin V were measured using FACS analysis. THP-1 cells transfected with anti-sense oligonucleotides showed no significant difference regarding annexin V expression (24--37% versus 20--27%, not significant) or proliferation rate (57--68% versus 64--80%, not significant) (Fig. [5b](#F5){ref-type="fig"}) compared with the cells transfected with non-sense oligonucleotides. These data indicate that the production of inflammatory cytokine IL-6 participating in the regulation of inflammatory responses is directly related to survivin expression by monocytes.
Discussion
==========
Suppression of apoptosis has been suggested as a key mechanism supporting selection and accumulation of distinct lymphocyte subsets in chronically inflamed joint tissues \[[@B21]\]. Indeed, synovial T cells in RA are highly differentiated and would not normally be expected to survive for a prolonged time within inflamed joints unless their death was actively inhibited \[[@B22]\]. In the present study we demonstrate that high expression of survivin, a member of the IAP family, is a new and potentially important mechanism of apoptosis suppression in patients with RA. Survivin is known as a multipotent inhibitor of apoptosis, neutralizing several caspases at the final steps of the apoptosis cascade, thus abrogating signals from both the death-receptor-dependent and mitochondrial pathways of apoptosis. Together with previous findings of upregulation of other caspase inhibitors (Bcl and FLIP) \[[@B12],[@B13]\], high levels of survivin give new insights in numerous alterations of the apoptosis machinery during the course of RA.
We observed that survivin levels were clearly increased in synovial fluid and plasma of RA patients compared with the healthy controls. Survivin expression was originally considered a reflection of cell proliferation. Indeed, survivin is continuously overexpressed in cancer cells \[[@B23]\]. Survivin gene transcription is repressed by wild-type p53 \[[@B24]-[@B26]\]. Multiple mutations and functional dysregulation of p53 have been demonstrated in the synovial tissue of RA patients \[[@B3],[@B27]\] and constitute one of the possible reasons for increased survivin production in this non-malignant condition. Notably, high survivin levels (over three standard deviations of the mean of healthy blood donors) were registered exclusively in patients with erosive joint disease and were associated with markers of inflammation such as WBC count and C-reactive protein levels, as well as with the absence of immunosuppressive treatment. This category of RA patients typically displays chronic joint inflammation, progressive joint destruction, and early mortality \[[@B28],[@B29]\].
Altogether these findings place survivin at the centre of attention as a potential prognostic factor for the destructive course of disease in RA. Indeed, using logistic regression analysis, we demonstrated that RA patients having high levels of survivin had a 16 times higher risk to develop destructive joint disease as compared with the patients with low levels of survivin. Moreover, in a multivariant model we showed that the role of survivin is independent of the presence of RF, the duration of the rheumatic disease, and gender. Interestingly, survivin expression has been shown to be an important prognostic factor in acute leukaemia \[[@B30],[@B31]\], and a predictor of recurrence in soft-tissue sarcomas \[[@B32]\] and urinary bladder cancer \[[@B33],[@B34]\]. In the latter case, extracellular urinary survivin levels were used for the evaluation of treatment and recurrence of cell carcinoma.
Survivin expression determined locally in the inflamed joints and also systemically in circulation of patients with RA was measured extracellularly. Whether survivin found extracellularly originates from dead cells or is a subject of active secretion is presently unknown. The number of *in vitro*leukocyte-activating stimuli (e.g. lipopolysaccharide, PHA, TSST-1, Concanavalin A) will not induce secretion of survivin. This observation suggests, but does not prove, that extracellular survivin found in synovial fluid originates from dead cells. Alternatively, some other cells (e.g. fibroblasts) or endogenous stimuli give rise to secretion of this molecule. Little is known about extracellular functions of survivin. Survivin has been suggested to function as a self-antigen in patients with haematologic malignancies and solid tumours. In our patient material we demonstrate the presence of antibodies to survivin in the plasma and synovial fluid of patients with RA. Interestingly, reactivity against survivin was significantly higher in the patient group with non-erosive RA. Notably, patients with non-erosive RA have extracellular survivin levels undistinguishable from these of the healthy controls. The association of a high level of antibodies against survivin with non-erosive joint disease may be a reflection of a protective autoimmune mechanism existing in these patients.
To assess the role of survivin in the inflammatory process, we first studied its inducibility in differentiated mature human PBMC. Most of the pro-inflammatory stimuli including lipopolysaccharide, Concanavalin A, TSST-1, and TNF-α leading to a significant release of inflammatory cytokines and chemokines, failed to induce survivin expression by PBMC. In contrast, downregulation of survivin expression using specific anti-sense oligonucleotides resulted in the decrease of IL-6 production by human monocytes. These two observations suggest that the regulatory role of survivin in inflammation is mediated by an increase of cytokine production. The connection between survivin expression and production of IL-6 deserves special attention in the view of recent success of the neutralization of IL-6 for alleviation of RA \[[@B35]\]. These observations support the regulatory role of survivin in the pathogenesis of arthritis.
Studying the variability of survivin levels in patients with RA, we observed that in most cases survivin levels were inclined to decrease in survivin-positive patients and almost never converted from absent to high in survivin-negative cases (data not shown). We also showed that the decrease of survivin levels could be mediated by treatment with DMARDs. This suggests survivin to be a transient phenomenon in the course of RA and may explain a relatively low frequency of patients having high survivin levels (21%) in the cohort tested. However, the results of our study may be affected by the fact that most of the patients were treated with DMARDs at the time of sampling, and even those without ongoing DMARD therapy might have received immunosuppressive treatment previously.
Conclusions
===========
Our study suggests that survivin regulates the inflammatory and destructive process inside the joints of patients with RA. Indeed, high levels of extracellular survivin are associated with chronic erosive arthritis, indicating poor prognosis. In contrast, antibodies against survivin are characteristic of the patients with the non-erosive, benign course of RA. Our findings on survivin expression and autoimmunity to this molecule provide new insight regarding the role of apoptosis in RA.
Abbreviations
=============
BSA = bovine serum albumin; DMARD = disease-modifying anti-rheumatic drug; ELISA = enzyme-linked immunosorbent assay; ERA = erosive rheumatoid arthritis group; FACS = fluorescence-activated cell sorting; FCS = foetal calf serum; FITC = fluorescein isothiocyanate; IAP = inhibitor of apoptosis proteins; IL = interleukin; MTX = methotrexate; NRA = non-erosive rheumatoid arthritis group; PBMC = peripheral blood mononuclear cells; PBS = phosphate-buffered saline; PHA = phytohaemagglutinine; RA = rheumatoid arthritis; RF = rheumatoid factor; TNF-α = tumour necrosis factor alpha; WBC = white blood cell.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
MB contributed to the study design, to the clinical, laboratory and statistical evaluation of material from RA patients, and to preparation of the manuscript. SL performed some of the cell experiments. DB performed ELISA assays, bioassays, and some of the transfection experiments. AT contributed to the conception of the study and the study design, to statistical evaluation of the results, and to preparation of the manuscript.
Acknowledgements
================
The work was supported by the Göteborg Medical Society, the Swedish Association against Rheumatism, King Gustaf V:s Foundation, the Swedish Medical Research Council, the Nanna Svartz\' Foundation, Börje Dahlin\'s Foundation, the National Inflammation Network, the Lundberg Foundation, Åke Wiber\'s Foundation, and the University of Göteborg.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Survivin levels in plasma and synovial fluid of patients with rheumatoid arthritis (RA) are significantly increased in the case of erosive joint disease. SEM, standard error of the mean.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Influence of disease-modifying anti-rheumatic drugs on survivin levels of rheumatoid arthritis patients with erosive joint disease. DMARDs, disease modifying anti-rheumatic drugs; MTX, methotrexate; TNF-α inh, tumour necrosis factor alpha inhibitors; SEM, standard error of the mean.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Synovial fluid antibodies of both IgG and IgM isotypes specific for survivin are higher in rheumatoid arthritis patients with the non-erosive course compared with the erosive course of the joint disease. SEM, standard error of the mean.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Expression of survivin in lysates from peripheral blood mononuclear cells of rheumatoid arthritis (RA) patients and from healthy controls following stimulation with various mitogens. Survivin expression was measured following 48 hours of stimulation. TNF-α, tumour necrosis factor alpha; PHA, phytohaemagglutinine; ConA, Concanavalin A; LPS, lipopolysaccharide.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Modulation of **(a)**survivin expression, **(b)**proliferation, and **(c)**IL-6 production following transfection of THP-1 cells with anti-sense oligonucleotides specific for survivin mRNA and non-sense sequences. Data are provided as the percentage of phytohaemagglutinine-stimulated THP-1 cells. The concentration of oligonucleotides throughout was 300 nM.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Clinical and demographic characteristics of patients with rheumatoid arthritis (RA) and of healthy controls
:::
Erosive RA (*n*= 88) Non-erosive RA (*n*= 43) Controls (*n*= 34)
--------------------------------------------------- ---------------------- -------------------------- --------------------
Age (years) \[mean ± standard deviation (range)\] 63 ± 2 (28--85) 53 ± 3 (19--83) 42 ± 1.8 (18--67)
Sex (male/female) 26/62 12/31 12/22
Duration of the disease (years) 12.7 ± 1.2 8 ± 1.4 \-
Rheumatoid factor (+/-) 80/8 10/33 n.a.
Treatment with DMARDs
Methotrexate (*n*= 25) 18 7 \-
Other DMARDs (*n*= 13) 9 4
TNF-α blockade (*n*= 47) 42 (37\*) 5 (5\*)
None (*n*= 45) 18 27
n.a., not assesed; DMARD, disease modifying anti-rheumatic drug; TNF-α, tumour necrosis factor alpha. \*In combination with methotrexate.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Clinical comparison of patients with rheumatoid arthritis (RA) expressing high\* and low levels of survivin
:::
Survivin high, erosive RA (*n*= 27) *P* Survivin low, erosive RA (*n*= 61) *P* Survivin low, non-erosive RA (*n*= 42)^†^
------------------------------------- ------------------------------------- ---------- ------------------------------------ ---------- -------------------------------------------
Survivin levels (pg/ml)
Blood 1180 ± 309 \<0.0001 97 ± 9 0.013 127 ± 5
Synovial fluid 1039 ± 523 0.016 132 ± 4 n.s. 124 ± 2
Disease duration (years) 15.5 ± 2.4 n.s. 13.6 ± 1.2 0.0002 8.3 ± 1.4
Age (years) 58 ± 3 n.s. 60 ± 2 0.05 53 ± 3
Rheumatoid factor-positive (*n*) 25 n.s. 53 \<0.0001 10
C-reactive protein (mg/l) 45 ± 9 0.035 29 ± 5 n.s. 39 ± 7
White blood cell count (× 10^9^/ml)
Blood 8.7 ± 0.6 0.038 7.2 ± 0.3 n.s. 7.1 ± 0.3
Synovial fluid 10.8 ± 1.9 n.s. 11.2 ± 2.9 n.s. 13.1 ± 2.8
Continuous parameters are presented as the mean ± standard error of the mean. n.s., not significant. \*Level of survivin above 300 pg/ml was considered \'high\'. ^†^One patient having a high survivin level is excluded.
:::
|
PubMed Central
|
2024-06-05T03:55:54.968450
|
2005-1-21
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065333/",
"journal": "Arthritis Res Ther. 2005 Jan 21; 7(2):R349-R358",
"authors": [
{
"first": "Maria",
"last": "Bokarewa"
},
{
"first": "Sofia",
"last": "Lindblad"
},
{
"first": "Dmitriy",
"last": "Bokarew"
},
{
"first": "Andrej",
"last": "Tarkowski"
}
]
}
|
PMC1065334
|
Introduction
============
Articular cartilage matrix undergoes substantial structural, molecular, and mechanical changes with ageing, including surface fibrillation, alteration in proteoglycan structure and composition, increased collagen cross-linking, and decreased tensile strength and stiffness \[[@B1],[@B2]\]. Deterioration in chondrocyte function accompanies these changes in the extracellular matrix \[[@B3]\]. Recently, attention has been given to the suggestion that cartilage ageing and chondrocyte senescence play an important role in the pathogenesis and development of osteoarthritis (OA) \[[@B4],[@B5]\]. Several reports revealed that chondrocyte senescence contributes to the risk for cartilage degeneration by decreasing the ability of chondrocytes to maintain and repair the articular cartilage tissue \[[@B4]-[@B6]\]. The mitotic and synthetic activity of chondrocytes decline with advancing donor age \[[@B5]\]. In addition, human chondrocytes become less responsive to anabolic mechanical stimuli with ageing and exhibit an age-related decline in response to growth factors such as the anabolic cytokine insulin-like growth factor-I \[[@B6]\]. These findings provide evidence supporting the concept that chondrocyte senescence may be involved in the progression of cartilage degeneration.
Telomeres, the terminal guanine-rich sequences of chromosomes, are structures that function in the stabilization of the chromosome during replication by protecting the chromosome end against exonucleases \[[@B7],[@B8]\]. The telomere DNA may function as a timing mechanism that, when reduced to a critical length, signals a cell to stop dividing and to enter cellular senescence \[[@B7]-[@B9]\]. More recent reports demonstrated that the telomere length of chondrocytes shortened with donor ageing and that decreased mean telomere length was closely related to the increase in senescence-associated β-galactosidase expression in human chondrocytes, suggesting that chondrocyte senescence, at least in part, participates in the age-related loss of chondrocyte function responsible for deterioration in articular cartilage structure and function \[[@B10]\]. An understanding of the mechanisms of chondrocyte senescence would be helpful to our efforts to devise new approaches to the prevention and treatment of OA.
Mechanical and chemical stresses are thought to induce increased free radical production, consequently leading to oxidative damage to the tissue \[[@B11]-[@B14]\]. Oxidative damage not only can initiate apoptosis through caspase activation but also may lead to irreversible growth arrest, similar to replicative senescence \[[@B11],[@B12],[@B15]\]. Furthermore, it has been reported that oxygen free radicals (O~2~^-^and peroxynitrite) directly injure the guanine repeats in the telomere DNA, indicating that oxidative stress directly leads to telomere erosion, regardless of cell active division \[[@B16]\]. Generally, it is now thought that oxidative stress/antioxidative capacity may be prominent among factors that control telomere length \[[@B17]-[@B19]\]. These findings strongly suggest that oxidative stress could induce chondrocyte telomere instability with no requirement for cell division in articular cartilage, leading to chondrocyte senescence.
Numerous reports have demonstrated that oxidative damage due to the over-production of nitric oxide (NO) and other reactive oxygen species (ROS) may be involved in the pathogenesis of OA \[[@B20]-[@B23]\]. However, because of the highly reactive nature of these oxygen reactive species and their short half-lives, it had been difficult to investigate oxidative damage *in vivo*\[[@B24]\]. ROS and NO cannot be directly and accurately measured in a cartilage sample. Recently, a reaction product of ROS and NO, namely nitrotyrosine, was used as evidence of oxidative damage in several ageing tissues \[[@B25],[@B26]\]. Loeser and coworkers \[[@B26]\] demonstrated that nitrotyrosine is over-expressed in normal cartilage from elder donors and in OA cartilage, suggesting the presence of oxidative damage in ageing and degenerative cartilage. These findings provide evidence to support the concept that oxidative stress in articular cartilage affects chondrocyte function, resulting in changes in cartilage homeostasis that are relevant to cartilage ageing, chondrocyte senescence and the development of OA.
Based on the properties of chondrocyte senescence and oxidative stress in OA cartilage, as discussed above, we postulated that oxidative stress induces telomere instability and dysfunction in chondrocytes, subsequently resulting in cartilage ageing and the development of OA through a mechanism involving the acceleration of chondrocyte senescence. It is now thought that oxidative stress/antioxidative capacity is prominent among factors that control telomere length, and hence replicative lifespan \[[@B17],[@B18]\]. To clarify the role of oxidative damage in the pathogenesis of OA, we looked for the presence of oxidative damage in degenerated cartilage from OA patients and examined whether chemical oxidative stress (ROS) affects chondrocyte telomere DNA, replicative lifespan, and function in cultured chondrocytes and in explants of articular cartilage. We also examined the effects of the antioxidative agent ascorbic acid on the oxidative stress induced downregulation of cellular lifespan and function in chondrocytes.
Methods
=======
Articular cartilage tissue and chondrocyte culture
--------------------------------------------------
Articular cartilage samples were obtained from OA patients (*n*= 9) who had undergone arthoplastic knee surgery (all female, age \[mean ± standard deviation\] 61.5 ± 5.4 years). The patients had given informed consent, in accordance with the ethical committee of the university. All samples were obtained in accordance with institutional protocol, with review board approval. Donor articular cartilage samples were evaluated macroscopically using a modified Collins scale from 0 to 5, as described previously \[[@B27]-[@B29]\].
To obtain sufficient numbers of cells for the experiments, cultured chondrocytes were isolated from macroscopically intact zones of cartilage. Cartilage tissue was cut into small pieces, washed in phosphate-buffered saline (PBS), and digested in Dulbecco\'s modified Eagle\'s medium (DMEM; Sigma, St. Louis, MO, USA) containing 1.5 mg/ml collagenase B (Sigma). Digestion was carried out at 37°C overnight on a shaking platform. Cells were centrifuged, washed with PBS, and plated with fresh DMEM.
Basically, chondrocytes were cultured in DMEM supplemented with 10% heat-inactivated foetal calf serum, 2 mmol/l l-glutamine, 25 mmol/l HEPES, and 100 units/ml penicillin and streptomycin at 37°C in a humidified 5% CO~2~atmosphere \[[@B30]\]. To avoid loss of chondrocyte phenotypes during passages, we used cultured chondrocytes only from passages 1--4. In parallel cultures, we checked the cell morphology and potential to produce proteoglycan in order to examine whether chondrocyte phenotype had been maintained during the passage. Data from chondrocyte mass cultures with loss of chondrocyte phenotypes were excluded from the analysis.
Chondrocytes were cultured in the presence of an antioxidant (100 μmol/l ascorbic acid-2-*O*-phosphate \[Asc2P; Wako Junyaku, Tokyo, Japan\]) or a ROS (H~2~O~2~) at a concentration of 0.1 μmol/l, which was not cytotoxic to the cells \[[@B17]\]. We had already investigated the effect of H~2~O~2~(0.1--500.0 μmol/l) on chondrocyte viability *in vitro*. Concentrations of 0.1--200.0 μmol/l of H~2~O~2~exhibited no inhibitory effects on chondrocyte viability (data not shown). In addition, we had also studied the time course of H~2~O~2~treatment (0.1--100.0 μmol/l) *in vitro*. Based on our preliminary experiments, in the present study we conducted the cell culture and the organ culture in the presence or absence of H~2~O~2~(0.1 μmol/l).
In each culture group, the medium including freshly prepared Asc2P or H~2~O~2~was changed every 2 days. Human chondrocytes were subcultured weekly. At each passage, the total number of collected cells in the dish was determined. Then, 2.5--5.0 × 10^5^cells were transferred to a new dish for the next passage, and the number of attached cells was determined 6 hours after seeding. From each passage, the remaining cells after subculture were stored at -180°C until the analysis of cellular activity, telomere length and telomerase activity was conducted.
Oxidative stress in human articular cartilage
---------------------------------------------
We compared the degree of oxidative stress (antioxidative potential) of the intact cartilage with that of degenerative cartilage tissue. Cartilage samples from the same donor joint were cut and divided into two groups (the degenerated region group, which exhibited macroscopic changes of OA; and the intact region group, which was macroscopically normal).
In these donor matched pairs of articular cartilage samples, antioxidative potential of the tissue was measured using an assay that is based on reduction of Cu^2+^to Cu^+^and the measurement was conducted according to the manufacturer\'s instructions (OXIS Health Products, Inc., Portland, OR, USA). This assay measures the total contribution of all antioxidants in the tissue sample. The results of the assay were calculated as mmol/l uric acid equivalents, and expressed as a ratio of antioxidative potential of the degenerating region to that of the corresponding intact region from each donor.
Immunohistochemistry
--------------------
For immunostaining of human articular cartilage, paraffin blocks of articular cartilage tissues were prepared using standard histological procedures. Serial sections of paraffin-embedded bone and cartilage tissues were cut and immunostained using an antibody for nitrotyrosine. The sections were deparaffinized and hydrated. Then, the slides were stained using horseradish peroxidase method \[[@B26]\]. Briefly, the slides were blocked with 3% H~2~O~2~. After blocking nonspecific protein binding with blocking agent (Dako, Carpinteria, CA, USA), the sections were incubated with a monoclonal antibody to nitrotyrosine (1:100 dilution; BIOMOL Research Laboratories Inc., Plymouth Meeting, PA, USA) for 1 hour at room temperature, followed by incubation with biotinylated goat anti-mouse IgG (Dako) for 30 min at room temperature. After washing with PBS, the sections were incubated with streptavidin--horseradish peroxidase complex (LSAB2 kit; Dako) for 30 min at room temperature We used diaminobenzidine (Sigma) as a visible peroxidase reaction product. Sections were counterstained with Mayer\'s haematoxylin (Sigma).
Cells positive and negative for nitrotyrosine were counted in the 20 areas of cartilage at 200× magnification (0.785 mm^2^/field). The level of immunostaining for nitrotyrosine was expressed as a mean number of nitrotyrosine-positive cells per field.
Chondrocyte activity
--------------------
Chondrocyte activity was measured as the production of glycosaminoglycan (GAG) by cultured chondrocytes \[[@B15]\]. After undergoing continuous treatment with ROS or ascorbic acid (initial subculture at the start of the experiment: 1 × 10^5^cells/dish, chondrocytes from passage 2), the cells were collected with trypsin and washed with PBS. Then, chondrocytes (1 × 10^5^cells/dish) were plated in the culture dishes and incubated for 12 hours, and the amount of GAG in the supernatant was measured using a spectrophotometric assay with dimethylmethylene blue (Aldrich Chemical, Milwaukee, WI, USA) \[[@B31]\].
Determination of the lifespan of cultured chondrocytes
------------------------------------------------------
The increase in cumulative population doublings at each subculture was calculated based on the number of cells attached and the cell yield at the time of the next subculture. Population zero was the primary culture of human chondrocytes, and the number of each successive generation was calculated using the following formula \[[@B32],[@B33]\]: generation number at the start of the subculture + log~2~(\[the number of collected cells at the time of the next subculture\]/\[the number of attached cells at the start of the subculture\]). Senescence was defined as less than one population doubling in 4 weeks. The *in vitro*lifespan (remaining replicative capacity) was expressed as population doublings up to cellular senescence \[[@B34]\].
Telomere length of cultured chondrocytes
----------------------------------------
Telomere length was determined using terminal restriction fragment Southern blot analysis, as described previously \[[@B35],[@B36]\]. Genomic DNA from 10^6^chondrocytes from each subculture (initial subculture at the start of the experiment: 1 × 10^6^cells/dish, chondrocytes from passage 3 or 4) was digested with 400 μl DNA extraction buffer (100 mmol/l NaCl, 40 mmol/l Tris \[pH 8.0\], 20 mmol/l EDTA, and 0.5% SDS) and proteinase K (0.1 mg/ml). Extraction was performed using phenol chloroform. Extracted DNA (5--10 μg) was digested with 10 units of *Msp*I and *Rsa*I (Boehringer Mannheim, Indianapolis, IN, USA) for 12--24 hours at 37°C. The integrity of the DNA before digestion and the completeness of digestion were monitored by gel electrophoresis. Electrophoresis of digested genomic DNA was performed in 0.5% agarose gels in 45 mmol/l Tris-borate EDTA buffer (pH 8.0) for a total of 660--700 V-h. After electrophoresis, gels were depurinated in 0.2 N HCl, denatured in 0.5 mol/l NaOH and 1.5 mol/l NaCl, transferred to a nylon membrane using 20× SSC, and dried for 1 hour at 70°C. The telomeric probe (TTAGGG)~3~(Genset, La Jolla, CA, USA) was 5\' end-labelled with \[α-^32^P\]ATP using T4 PNK (Boehringer Mannheim). Prehybridization and hybridization were performed at 50°C using 5× Denhardt\'s, which was composed of 5× SSC, 0.1 mol/l Na~2~HPO~4~, 0.01 mol/l Na~4~P~2~O~7~, 30 μg/ml salmon sperm DNA, and 0.1 mmol/l ATP. The mean terminal restriction fragment length was determined from densitometric analysis of autoradiograms, as described previously \[[@B35]\].
Tissue culture of human articular cartilage
-------------------------------------------
Procedures for preparing articular cartilage were generally the same as mentioned above. Briefly, articular cartilage was excised in small, full-depth slices (typically 1.0 cm square) from patients with OA (*n*= 4) who had undergone arthroplastic knee surgery (all females; ages 61, 65, 67 and 68 years). The cartilage explants were cut, weighed and divided into three groups as follows: control group, antioxidative agent + oxidative stress treated group, and oxidative stress treated group. Control and experimental cartilage explants (site-matched pairs) were placed in individual dishes (diameter 6.0 cm) with 10.0 ml DMEM with 10% foetal bovine serum, 100 units/ml penicillin/streptomycin. The process of harvesting the cartilage tissue resulted in significant catabolic activity that was measurable in the absence of interleukin-1 stimulation, presumably due to secretion of proteases in response to trauma. The contribution of this basal catabolic activity could be minimized by culturing for 24 hours before aspiration of the culture medium, washing with PBS, and adding fresh culture medium \[[@B37],[@B38]\]. For the antioxidative agent + oxidative stress treated group, the cartilage explants were incubated in the culture medium with 100.0 μmol/l Asc2P plus 0.1 μmol/l H~2~O~2~. For the oxidative stress treated group, the explants were incubated in the culture medium in the presence of 0.1 μmol/l H~2~O~2~. For each group, culture medium including freshly prepared Asc2P or H~2~O~2~was changed every day.
At the end of each incubation period (48, 72, 96, 120 and 120 hours), the cartilage samples and the culture media were collected and re-weighed for analyses. The cartilage samples were washed with PBS. Some parts of cartilage samples were fixed with 4% paraformaldehyde at 4°C, and then paraffin blocks were prepared using standard histological procedures. For nitrotyrosine staining, the sections were deparaffinized and hydrated, and then were immunostained using antibody for nitrotyrosine in accordance with the method described above.
Other cartilage samples and supernatants were stored at -80°C for the determination of GAG concentration and isolated chondrocyte telomere length. Catabolic changes to GAG in cartilage were analyzed by determining the GAG content remaining in cartilage tissue relative to the total amount of GAG in the culture (GAG released into the culture media plus GAG in the tissue) in the presence of the antioxidative agent or ROS \[[@B2],[@B39]\]. GAG contents were measured using a spectrophotometric assay mentioned above. Procedures for cultured chondrocyte preparation from tissue cultured explants and telomere length assay were generally the same as those described above.
Statistical analysis
--------------------
Results were expressed as a mean value ± standard deviation. Comparison of the means was performed by analysis of variance. *P*\< 0.05 was considered statistically significant.
Results
=======
Oxidative damage in human articular cartilage tissues
-----------------------------------------------------
To determine whether oxidative damage was present in OA degenerated cartilage, we measured the antioxidative potential of the intact region and degenerated region isolated from the same articular cartilage tissue of patients who had undergone arthroplastic knee surgery. In the donor-matched pair of intact and degenerated regions from same articular cartilage, the antioxidative potential in the intact region was significantly greater than that in the degenerated region of articular cartilage in the OA patient group (*n*= 9; mean percentage antioxidative capacity of degenerative cartilage compared with intact cartilage: 45.5 ± 16.8%), suggesting that degenerated cartilage may exhibit more oxidative damage than an intact region from the same OA cartilage.
Presence of nitrotyrosine in articular cartilage from patients with osteoarthritis
----------------------------------------------------------------------------------
To clarify the relationship between oxidative damage and development of OA, immunostaining for nitrotyrosine was examined in the donor-matched pair of intact and degenerated articular cartilage sections from the same OA sample.
Figure [1](#F1){ref-type="fig"} shows a representative example of immunohistochemical staining for nitrotyrosine in the articular cartilage from an OA patient (female, 67 years old). Immunostaining for nitrotyrosine was most apparent in the degenerated regions of articular cartilage that showed histological changes consistent with OA (nine patients; positive cells/field, intact cartilage versus degenerated cartilage: 0.3 ± 0.1 versus 7.4 ± 2.4; *P*\< 0.01). Nine of 10 donor samples with degenerated regions were highly positive for nitrotyrosine. Nitrotyrosine was present both within chondrocytes and in the cartilage matrix, and was seen mainly in the more superficial regions. The degree of immunostaining for nitrotyrosine (number of positive cells/field) correlated with the level of histological change in donor cartilage tissues (*n*= 9, r^2^= 0.4671; *P*\< 0.01). In contrast to the immunostaining in the degenerated regions, almost all intact regions isolated from the same articular cartilage were negative for nitrotyrosine, even in superficial and deep zones (Fig. [1](#F1){ref-type="fig"}).
*In vitro*chondrocyte activity under the different oxidative conditions
-----------------------------------------------------------------------
Figure [2](#F2){ref-type="fig"} shows that GAG synthesis from cultured chondrocytes decreased gradually in a time dependent manner, regardless of the presence of H~2~O~2~or an antioxidative agent *in vitro*. The H~2~O~2~treated group showed a significant decrease in proteoglycan production by chondrocytes as compared with the control group at any incubation time. In contrast, in the antioxidative agent group the level of proteoglycan production tended to increase as compared with that in control groups, although no significant differences were observed between control groups and antioxidative agent groups at any incubation time (Fig. [2](#F2){ref-type="fig"}).
Chondrocyte replicative potential under the different oxidative conditions
--------------------------------------------------------------------------
To clarify the effect of oxidative stress on the replicative potential of chondrocytes, we analyzed the cellular replicative potential of chondrocytes in the presence of the antioxidative agent or ROS *in vitro*. As shown in Fig. [3](#F3){ref-type="fig"}, the replicative potential of cultured chondrocytes was expressed as the cumulative number of cells dividing at each incubation time. After 20 days of incubation the H~2~O~2~treated group exhibited lesser replicative potential as compared with the control group at any incubation time. In contrast, treatment with the antioxidative agent increased the cellular replicative potential at all incubation times after 20 days (Fig. [3](#F3){ref-type="fig"}).
During the 4 weeks after a 50- to 60-day incubation, the cumulative population doubling levels of all groups reached a plateau, indicating that the cultured chondrocytes in each group reached the limit of their ability to divide, namely cellular senescence, after about 8 weeks of incubation. The mean lifespan to cellular senescence was 23 population doublings in the antioxidative agent treated group, 18 population doublings in the control group, and 14 population doublings in the ROS-treated group (Fig. [3](#F3){ref-type="fig"}).
Chondrocyte telomere length under the different oxidative conditions
--------------------------------------------------------------------
To clarify the effect of oxidative stress on the telomeric instability in chondrocytes, we analyzed the telomere length of chondrocytes in the presence of an antioxidative agent or ROS *in vitro*(Fig. [4a](#F4){ref-type="fig"}). After five to six population doublings, telomere lengths of chondrocytes were shorter in H~2~O~2~treated groups than in control groups at any level of population doubling. Treatment with an antioxidative agent resulted in a tendency of chondrocyte telomere length to elongate (*n*= 9; Fig. [4b](#F4){ref-type="fig"}).
Immunohistochemical staining for nitrotyrosine of human articular cartilage cultured under different oxidative conditions
-------------------------------------------------------------------------------------------------------------------------
To examine the influence of an antioxidative agent or ROS in human articular cartilage, immunohistochemical staining for nitrotyrosine was evaluated in cartilage samples that were treated with an antioxidative agent or ROS (H~2~O~2~) in organ culture. Cartilage from an OA patient was cut and divided into three groups as follows: control group, antioxidative agent (Asc2P) treated group, and H~2~O~2~treated group. After a 48-hour incubation in explant culture, OA articular cartilage in both the control group and the H~2~O~2~treated group exhibited positive immunostaining for nitrotyrosine (Fig. [5a](#F5){ref-type="fig"}). The degree of nitrotyrosine staining was higher in the H~2~O~2~treated group than in the control group (Fig. [5b](#F5){ref-type="fig"}). In contrast to these two groups, articular cartilage treated with the antioxidative agent showed less staining for nitrotyrosine (Fig. [5b](#F5){ref-type="fig"}).
Catabolic changes to articular cartilage matrix under different oxidative conditions in organ culture
-----------------------------------------------------------------------------------------------------
To investigate whether oxidative stress resulted in catabolic changes to the articular cartilage matrix, we examined the amount of GAG remaining in cartilage tissue and that was released into the culture medium in organ culture in the presence of an antioxidative agent or ROS. Catabolic changes to proteoglycan in the tissue were quantified as the percentage of proteoglycan remaining in the cartilage relative to total amount in the culture medium plus cartilage.
During culture, the amount of proteoglycan remaining in the cartilage tissue in the control group and H~2~O~2~-treated group decreased gradually in a timedependent manner. After 72 hours of incubation, the percentage of proteoglycan remaining in the cartilage tissue was significantly lower in the H~2~O~2~treated group than in the control group. In contrast, the antioxidative agent (Asc2P) treated group exhibited a tendency to maintain tissue proteoglycan even in the presence of H~2~O~2~during the incubation period we studied in organ culture (Fig. [6](#F6){ref-type="fig"}).
Telomere length of chondrocytes from human articular cartilage explants cultured under different oxidative conditions
---------------------------------------------------------------------------------------------------------------------
To clarify the effect of oxidative stress on chondrocyte telomeric instability in the cartilage, we analyzed the telomere length of chondrocytes that were isolated from cartilage explants cultured in the presence of an antioxidative agent (Asc2P) or ROS (H~2~O~2~) *in vitro*. After 144 hours of incubation, the telomere length of chondrocytes was significantly shorter in H~2~O~2~treated groups (lane 4 in Fig. [7a,b](#F7){ref-type="fig"}) than in control group (lane 2 in Fig. [7b](#F7){ref-type="fig"}). Treatment with an antioxidative agent showed a tendency to maintain chondrocyte telomere length (lane 3 in Fig. [7](#F7){ref-type="fig"}).
Discussion
==========
The present study clearly demonstrates for the first time that oxidative stress affects chondrocyte telomeric DNA, cellular replicative lifespan, chondrocyte function, and cartilage matrix proteoglycan structure and composition *in vitro*and *in vivo*. These findings are consistent with a large body of data showing that reactive oxidative species, such as NO and ROS, are important in the pathogenesis of OA \[[@B11]-[@B16]\]. More recently, a suggestion that chondrocyte senescence may contribute to the risk for cartilage degeneration by decreasing the ability of the cells to maintain and to repair cartilage tissue has attracted attention \[[@B3]-[@B6]\]. Age-dependent changes in articular cartilage increase the risk for joint deterioration that causes the clinical syndrome of OA. However, the exact mechanism of chondrocyte senescence remains unclear. Our findings, demonstrating the oxidative stress (ROS) induced telomere erosion and replicative senescence in chondrocytes, suggest the involvement of oxidative stress in both the progression of cartilage ageing (chondrocyte senescence) and the development of OA.
Our results also show the presence of oxidative damage in degenerated cartilage from OA patients. Chondrocytes have been shown to be capable of producing ROS and NO \[[@B15],[@B20],[@B40]\]. In the present study, stronger staining for nitrotyrosine, a marker of oxidative stress, was observed in degenerating regions as compared with intact regions from the same articular cartilage samples. In addition, the degree of immunostaining was correlated with the level of histological change in articular cartilage. These findings suggest that local accumulation of proteins altered by the reaction between ROS and NO may be important in the pathogenesis of OA. Oxidative damage in cartilage may affect chondrocyte function, resulting in changes in cartilage homeostasis that are relevant to cartilage ageing and the development of OA.
We also measured the antioxidative potential of articular cartilage tissue using an assay based on reduction in Cu^2+^to Cu^+^by the combined action of all antioxidants present in the cartilage sample. Numerous reports have demonstrated that hypoxia is suitable for chondrocyte proliferation *in vitro*\[[@B41]-[@B43]\]. During chondrocyte differentiation, hypoxia may promote the process, although the exact mechanisms of chondrocyte differentiation have not been investigated to date. In addition, there is a general consensus that tissue oxygen partial pressures within articular cartilage decrease with increasing depth from the cartilage surface to deep layers \[[@B38],[@B44],[@B45]\]. Oxygen gradients do indeed exist in joint articular cartilage. These findings suggest that hypoxia may be required for homeostasis and maintenance of articular cartilage as well as chondrocyte cell growth and differentiation. During the development of OA, mechanical and chemical stresses may affect cellular adaptation to hypoxia, consequently leading to oxidative damage and changes in the microenvironment due to oxidative damage, resulting in the downregulation of chondrocyte synthesis. Indeed, our results revealed that antioxidative potential was significantly lower in degenerating regions than in intact regions from the same articular cartilage sample in OA.
To clarify the involvement of oxidative damage in the development of OA, we focused on chondrocyte telomere instability. Cumulative cell damage from oxidative stress provides an alternative explanation for cellular senescence. Oxygen free radicals directly damage guanine repeats in telomeric DNA, resulting in telomere erosion regardless of cell division \[[@B16]-[@B19]\]. DNA single strand damage by oxygen free radicals results in telomere shortening during DNA replication. Oxidative stress increases the telomere shortening rate by up to one order of magnitude \[[@B46]\]. From these findings, we postulated that oxidative stress directly induces chondrocyte telomere instability in OA cartilage tissue, resulting in chondrocyte senescence with no requirement for cell division. Our results, demonstrating chondrocyte telomere shortening in the presence of H~2~O~2~, at a noncytotoxic concentration, supports this hypothesis.
In addition to oxidative stress-induced telomere shortening, chondrocytes under chemical oxidative stress showed lower replicative lifespan and proteoglycan production as compared with normal chondrocytes *in vitro*. These findings also indicate that oxidative stress affects chondrocyte viability, and replicative potential and function, as well as telomere erosion.
We investigated catabolic changes to articular cartilage matrix under different oxidative conditions in tissue culture. The degree of immunostaining for nitrotyrosine was significantly higher in ROS (H~2~O~2~) treated cartilage tissues than in control cartilage tissues that were derived from the same articular cartilage. In addition, the GAG released to the medium was increased in the presence of ROS, suggesting that oxidative damage induces catabolic changes to cartilage matrix proteoglycan in articular cartilage. These observations led us to the hypothesis that oxidative stress may induce catabolic changes in cartilage matrix, consequently leading to the development of OA. This hypothesis is supported by the results of the present study, demonstrating that treatment of articular cartilage with the antioxidative agent ascorbic acid resulted in less immunopositivity for nitrotyrosine and maintenance of GAG content in articular cartilage in tissue culture.
Interestingly, treatment of cultured cartilage with an antioxidative agent not only inhibited GAG loss but also maintained telomere length of chondrocytes from cultured cartilage in contrast to data obtained from cultured cartilage under normal or ROS-treated conditions. These findings may very well indicate the role played by endogenous oxidative agents in catabolic changes to cartilage matrix proteoglycan and telomere length. This is an important observation and will validate the hypothesis that oxidative agents play a role *in situ*in chondrocytes and in cartilage changes in OA. These results also support the concept that antioxidative agents may prevent oxidative stress-induced chondrocyte dysfunction and degeneration in cartilage.
The findings of the present study suggest that cumulative oxidative stress leads to a decrease in antioxidative capacity in articular cartilage, resulting in chondrocyte telomere shortening, regardless of cell proliferation. Oxidative stress may be closely involved in telomere erosion, cellular senescence in chondrocytes and resultant cartilage ageing.
Conclusion
==========
This study provides insight into the involvement of oxidative stress in the pathogenesis of OA from the viewpoint of oxidative stress induced genomic instability, especially telomere erosion, and chondrocyte senescence. Our findings clearly show the presence of oxidative stress in degenerating cartilage, and the resultant telomere erosion and dysfunction of chondrocytes *in vitro*and *in vivo*, suggesting a role for oxidative stress in the development of OA. Also, our results suggest that antioxidative agents are effective in preventing and overcoming oxidative stress induced cartilage degeneration. New efforts to prevent the development and progression of OA may include strategies and interventions aimed at reducing oxidative damage in articular cartilage.
Abbreviations
=============
Asc2P = ascorbic acid-2-*O*-phosphate; DMEM = Dulbecco\'s modified Eagle\'s medium; GAG = glycosaminoglycan; NO = nitric oxide; OA = osteoarthritis; PBS = phosphate-buffered saline; ROS = reactive oxygen species.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
KY carried out *in vitro*studies (cell culture and organ culture), participated in the design of the study, conducted sequence alignment and drafted the manuscript. NvT carried out the immunoassays. HN, KH-M, TK and KN conceived the study, participated in its design and coordination, and helped to draft the manuscript. All authors read and approved the final manuscript
Acknowledgements
================
This study was supported by grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan, the Ministry of Health, Labour and Welfare of Japan, and the Japan Rheumatism Foundation.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Representative immunohistochemical staining for nitrotyrosine in donor articular cartilage. Cartilage sections were immunostained using an anti-nitrotyrosine antibody. In donor-matched pairs of degenerative and intact regions from same cartilage explants (67-year-old donor), positive immunostaining for nitrotyrosine was observed in chondrocytes and in the cartilage matrix in degenerated regions, whereas the intact region from same cartilage sample showed no positive staining for nitrotyrosine. Original magnification: 40×.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Glycosaminoglycan (GAG) production from cultured chondrocytes under different oxidative conditions. After the incubation times indicated, in the presence of 0.1 μmol/l H~2~O~2~or 100.0 μmol/l ascorbic acid (initial subculture at the start of the experiment: 1 × 10^5^cells/dish, chondrocytes at passage 2), chondrocytes were collected and transferred to a new culture dish (1 × 10^5^cells/dish). Following 12 hours of incubation, the amount of GAG in the supernatant was measured using a spectrophotometric assay with dimethylmethylene blue. Values are expressed as the mean ± standard deviation of nine donors (*n*= 4 culture dishes per treatment group at each incubation period; \**P*\< 0.05, \*\**P*\< 0.01, versus control group at each incubation time). The H~2~O~2~treated group exhibited a significant decrease in GAG production by chondrocytes as compared with the control group at all incubation times. In the antioxidative agent group the level of proteoglycan production tended to increase as compared with the control group, although no significant differences were observed between the control groups and antioxidative agent groups at any incubation time.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Chondrocyte replicative capacity under the various oxidative conditions. At each subculture (initial subculture at the start of the experiment: 5 × 10^4^cells/dish, primary culture), the total number of cells in the dish was determined, and the cells (1 × 10^5^cells/dish) were placed in a new dish. The number of cells that had attached 6 hours after seeding was determined. The increase in cumulative population doublings (number of cell divisions) at each subculture (*n*= 4 per treatment group) was calculated based on the number of cells attached and the cell yield at the time of the next subcultivation. Cell cultures were considered to have achieved their proliferative limit (senescence) when they did not exceed a twofold increase in 4 weeks. Values are expressed as mean ± standard deviation of four donors. \**P*\< 0.05 and \*\**P*\< 0.01, versus control group at each incubation time.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Southern blot analysis of chondrocyte telomere lengths in cultured chondrocytes at each passage under the different oxidative conditions. **(a)**Representative image of Southern blot analysis. Telomere lengths in chondrocytes (1 × 10^6^cells/dish, initial subculture at the start of the experiment: chondrocytes at passage 3 or 4) were determined using the terminal restriction fragment (TRF) assay. **(b)**The mean lengths of the chondrocytes were calculated by densitometric molecular weight analysis and were plotted against the number of cell population doublings. \**P*\< 0.05, versus control group at each incubation time. ROS, reactive oxygen species.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Tissue culture of articular cartilage tissue. **(a)**Representative immunohistochemical staining for nitrotyrosine in cartilage explants treated with reactive oxygen species (ROS) or an antioxidative agent in tissue culture. Osteoarthritis (OA) cartilage explant from a 67-year-old donor was cut and divided into three groups: control group, H~2~O~2~treated group, and antioxidative agent (ascorbic acid-2-*O*-phosphate \[Asc2P\]) treated group. After the end of the incubation period (48 hours of incubation), the cartilage sections were immunostained with anti-nitrotyrosine antibody. Original magnifications are given in parentheses. **(b)**The number of nitrotyrosine positive cells were counted in the 20 areas of tissue-cultured cartilage at 200× magnification (0.785 mm^2^/field). A statistical analysis of immunostaining was performed. \**P*\< 0.05, \*\**P*\< 0.01, versus control group.
:::
:::
::: {#F6 .fig}
Figure 6
::: {.caption}
######
Glycosaminoglycan (GAG) remaining in the cartilage extract treated with reactive oxygen species (ROS) or antioxidative agent in tissue culture. Catabolic change in articular cartilage matrix was analyzed by determining the GAG content remaining in the cartilage extract relative to the total amount of GAG in the supernatant and the cartilage digest. Values are expressed as mean ± standard deviation of nine donors (three cartilage extracts per donor). \**P*\< 0.05, \*\**P*\< 0.01, versus control group at each incubation time.
:::
:::
::: {#F7 .fig}
Figure 7
::: {.caption}
######
Telomere length of cultured chondrocytes from tissue cultured cartilage explants under the different oxidative conditions. After 144 hours\' incubation of tissue culture, chondrocytes were isolated from cartilage explants, which were incubated in the presence or absence of H~2~O~2~(0.1 μmol/l) or ascorbic acid-2-*O*-phosphate (Asc2P; 100.0 μmol/l). Telomere lengths in chondrocytes (1 × 10^6^chondrocytes of passage 3--4 after isolation) were determined using the terminal restriction fragment (TRF) assay. **(a)**Representative image of telomere length assay of chondrocytes after 144 hours of incubation. Lane 1, pretreated group (telomere length of isolated chondrocytes from cartilage explants before tissue culture); lane 2, Asc2P + H~2~O~2~treated group; lane 3, control group; lane 4, H~2~O~2~treated group. **(b)**Treatment with Asc2P (lane 2) showed a tendency to elongate the mean telomere length of chondrocytes in comparison with control. Mean telomere length in H~2~O~2~treated group was significantly shorter than in the control group (*n*= 9; *P*\< 0.05).
:::
:::
|
PubMed Central
|
2024-06-05T03:55:54.971815
|
2005-1-26
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065334/",
"journal": "Arthritis Res Ther. 2005 Jan 26; 7(2):R380-R391",
"authors": [
{
"first": "Kazuo",
"last": "Yudoh"
},
{
"first": "Nguyen",
"last": "van Trieu"
},
{
"first": "Hiroshi",
"last": "Nakamura"
},
{
"first": "Kayo",
"last": "Hongo-Masuko"
},
{
"first": "Tomohiro",
"last": "Kato"
},
{
"first": "Kusuki",
"last": "Nishioka"
}
]
}
|
PMC1065335
|
Introduction
============
The adaptive immune system uses various potent effector mechanisms for the elimination of foreign pathogens. Because these mechanisms are potentially damaging to the host, an essential feature of the immune system is its ability to distinguish self from non-self antigens and to develop tolerance to the former. With regard to T cell tolerance, the immune system has evolved several strategies. Most autoreactive T cells are eliminated during (primary) maturation in the thymus, a process described as negative selection, resulting in central T cell tolerance. Autoreactive T cells that escape negative selection will nevertheless be prevented from being activated as they are confronted with auto-antigen in the periphery. Several mechanisms have been proposed to account for this peripheral tolerance. One of those is suppression by a subset of T cells that express both CD4 and CD25. Evidence for the important role of these cells is overwhelming \[[@B1]\]. For example, when CD4^+^T cells isolated from peripheral lymphoid tissues of normal mice are depleted of CD4^+^CD25^+^T cells and injected into *nu*/*nu*mice, the recipients develop a high incidence of organ-specific autoimmune disease \[[@B2]\]. Co-transfer of the CD4^+^CD25^+^population prevents the induction of disease. CD4^+^CD25^-^and CD4^+^CD25^+^T cells are therefore often designated as, respectively, T~eff~and T~reg~cells. CD4^+^CD25^+^T~reg~cells are generated in the thymus. Their development is directed by relatively high-avidity interactions between the TCR and self-peptide ligands \[[@B3]-[@B5]\]. The CD4^+^CD25^+^T~reg~cell population constitutes 5 to 10% of the mature CD4^+^cell population in the adult thymus and the peripheral lymphoid tissue and blood.
*In vitro*, CD4^+^CD25^+^T~reg~cells inhibit polyclonal T cell activation \[[@B6],[@B7]\]. The suppression is mediated by a cytokine-independent, cell contact-dependent mechanism that requires activation of the CD4^+^CD25^+^cells via the TCR with specific antigen \[[@B8]\]. However, once stimulated, they are competent to suppress in an antigen-independent manner. Although the exact mechanism by which T~reg~cells exert their regulatory function is still unknown, there are indications that interaction of transforming growth factor-β (TGF-β) with its receptor \[[@B9]-[@B11]\], inhibition of IL-2 production \[[@B6]\] or downregulation of co-stimulatory molecules on antigen-presenting cells \[[@B12]\] could be involved.
T~reg~cells have proved to be important in various animal models of autoimmune diseases. Administration of anti-CD25 antibody *in vivo*induces organ-localised autoimmune diseases \[[@B13]\]. Inoculation of CD4^+^T cells depleted of CD25^+^cells in *nu/nu*mice results in autoimmune diseases such as gastritis, thyroiditis and insulitis \[[@B2]\]. Thus, transfer of T~reg~cells prevents autoimmune gastritis after neonatal thymectomy, and inhibits gastritis induced by H/K ATPase-reactive effector T cells \[[@B14]\]. MBP-specific CD25^+^CD4^+^T cells prevent spontaneous autoimmune encephalomyelitis in TCR-transgenic mice deficient in the recombination activating gene RAG-1 \[[@B15]\]. Similarly, CD4^+^CD25^+^T~reg~cells suppress central nervous system inflammation during active experimental autoimmune encephalomyelitis \[[@B16]\].
Collagen-induced arthritis (CIA) is a well-described animal model for rheumatoid arthritis. The disease is induced in genetically susceptible DBA/1 mice by immunisation with collagen type II (CII), and both T cell and B cell autoimmune responses are required for its development \[[@B17]-[@B19]\]. IFN-γ receptor knock-out (IFN-γR KO) mice have been found to suffer an accelerated and more severe form of CIA \[[@B20]-[@B23]\]. Moreover, knocking-out of the IFN-γ gene makes genetically resistant strains of mice susceptible to CIA \[[@B24],[@B25]\]. These data indicate that deletion of the IFN-γ response somehow disrupts an endogenous protective mechanism against CIA.
Morgan and colleagues \[[@B26]\] have recently demonstrated that CD25^+^T~reg~cells are important in the pathogenesis of CIA. In the present study we confirmed the importance of T~reg~cells in the pathogenesis of CIA by rendering wild-type DBA/1 mice deficient in T~reg~cells by depleting anti-CD25 antibodies. Anti-CD25-treated mice developed a significantly more severe arthritis, comparable to the disease course in IFN-γR KO mice. Thus, we proposed that the higher susceptibility of IFN-γR KO DBA/1 mice to CIA might be ascribed to defects in the production (differentiation and homeostasis) or function of these CD4^+^CD25^+^T~reg~cells. We therefore determined the numbers of T~reg~cells in central and peripheral lymphoid organs of IFN-γR KO and wild-type mice. We further investigated whether T~reg~cells of IFN-γR KO mice have defects in the ability to suppress TCR-induced *in vitro*proliferation of CD4^+^CD25^-^T~eff~cells.
Materials and methods
=====================
Mice and experimental conditions
--------------------------------
The generation and the basic characteristics of the mutant mouse strain (129/Sv/Ev) with a disruption in the gene coding for the α-chain of the IFN-γ receptor (IFN-γR KO) have been described \[[@B27]\]. These IFN-γR KO mice were backcrossed with DBA/1 wild-type mice for 10 generations to obtain the DBA/1 IFN-γR KO mice used in the present study. The homozygous IFN-γR KO mice were identified by PCR as described \[[@B23]\]. Wild-type and IFN-γR KO DBA/1 mice were bred in the Experimental Animal Centre of the University of Leuven. The experiments were performed in mice 6 to 10 weeks old, but in each experiment the mutant and wild-type mice were age-matched within 5-day limits. The male : female ratio was kept between 0.8 and 1.3 in each experiment group, unless otherwise mentioned. All animal experiments were approved by the local ethical committee (University of Leuven).
Induction and clinical assessment of arthritis
----------------------------------------------
Native chicken CII (Sigma-Aldrich, St Louis, MO, USA) was dissolved at 2 mg/ml in PBS containing 0.1 M acetic acid by stirring overnight at 6°C and emulsified in an equal volume of complete Freund\'s adjuvant (CFA; Difco Laboratories, Detroit, MI, USA) with added heat-killed *Mycobacterium butyricum*(0.5 mg/ml). IFN-γR KO and wild-type mice were sensitised with a single intradermal injection at the base of the tail with 100 μl of the emulsion on day 0. From day 0 after immunisation, mice were examined for signs of arthritis five times a week. The disease severity was recorded with the following scoring system for each limb: score 0, normal; score 1, redness and/or swelling in one joint; score 2, redness and/or swelling in more than one joint; score 3, redness and/or swelling in the entire paw; score 4, deformity and/or ankylosis.
Media, reagents and antibodies
------------------------------
All cells were grown in RPMI 1640 (Bio Whittaker Europe, Verviers, Belgium), supplemented with 10% heat-inactivated FCS (Gibco, Paisley, UK), penicillin (100 IU/ml; Continental Pharma, Brussel, Belgium), streptomycin (100 μg/ml; Continental Pharma), 2 mM L-glutamine, 10 mM Hepes (Gibco), 0.1 mM nonessential amino acids (ICN, Asse Relegem, Belgium), 1 mM sodium pyruvate (Gibco) and 50 μM 2-mercaptoethanol (Fluka, AG, Switzerland).
Anti-CD25 IL-2Rα monoclonal antibody was produced by hybridoma PC61 in an INTEGRA CELLine CL1000 (Elscolab, Kruibeke, Belgium) and is a rat IgG1 antibody. The hybridoma supernatant was purified by Protein G-Sepharose chromatography (Amersham Biosciences, Roosendaal, The Netherlands) for administration *in vivo*.
The hamster monoclonal antibody, directed against the mouse CD3 complex, was prepared from the culture supernatant of 145-2C11 hybridoma cells \[[@B28]\]. The antibodies were purified by affinity chromatography with Protein A-Sepharose (Amersham Biosciences). Batches of anti-CD3 antibody were tested for endotoxin content with the *Limulus*amebocyte lysate QCL-1000 kit (Bio Whittaker) and were found to contain less than 3 ng/ml endotoxin.
Cell purification
-----------------
Lymph nodes (axillary, inguinal and mesenteric) and spleens were harvested from mice 6 to 8 weeks old. Lymph nodes and spleens were gently cut into small pieces and passed through cell strainers (Becton Dickinson Labware, Franklin Lakes, NJ, USA). Red blood cells were lysed by two consecutive incubations (5 and 3 min at 37°C) of the suspension in NH~4~Cl (0.83% in 0.01 M Tris-HCl, pH 7.2). Remaining cells were washed, resuspended in cold PBS and counted. Lymph node preparations were then enriched for CD4^+^T cells with the Mouse T cell CD4 Subset Column Kit (R&D systems, Abingdon, UK). To purify CD4^+^CD25^+^and CD4^+^CD25^-^cells, the enriched CD4^+^T cells were incubated for 20 min at 4°C with FITC-conjugated anti-CD25 and phycoerythrin (PE)-conjugated anti-CD4 antibodies (10 μg per 10^8^cells) in PBS containing 2% FCS. They were sorted by flow cytometry on a FACS Vantage (Becton Dickinson, San Jose, CA, USA). The resultant purity of the CD4^+^CD25^-^population was 99%, whereas the purity of the CD4^+^CD25^+^population varied from 96% to 99%. Alternatively, CD4^+^T cells were labelled with PE-conjugated anti-CD25 monoclonal antibody, followed by incubation with magnetic-activated cell sorting (MACS) anti-PE beads (CD25 Microbead Kit; Miltenyi Biotec, Bergisch Gladbach, Germany). CD4^+^CD25^+^T cells were selected on an LS column in a magnetic field and the flow-through was collected as CD4^+^CD25^-^T cells. After removal of the column from the magnetic field, CD4^+^CD25^+^T cells were flushed out by a plunger. The purity of the CD4^+^CD25^-^population was 99% and the purity of the CD4^+^CD25^+^population varied from 90% to 95%.
T cell-depleted spleen suspensions were prepared by MACS (Miltenyi Biotec) and used as accessory cells (ACs). For MACS separation, the cell suspension was magnetically labelled with CD90 (Thy1.2) microbeads and passed through a CS separation column, placed in a magnetic field. The unlabelled CD90^-^cells ran through.
Flow cytometry
--------------
Single-cell suspensions (5 × 10^5^cells) were incubated for 15 min with the Fc-receptor-blocking antibodies anti-CD16/anti-CD32 (CD16/CD32; BD Biosciences Pharmingen, San Diego, CA, USA). Cells were washed with PBS containing 2% FCS and stained with the indicated FITC-conjugated antibodies (0.5 μg) for 30 min, washed twice and incubated for 30 min with the indicated PE- or biotin-conjugated antibodies. For the biotin-conjugated antibodies, a third staining step with streptavidin conjugated with peridinin chlorophyll *a*protein (PerCP) was performed. After washing, propidium iodide (Sigma-Aldrich) was added at a final concentration of 4 μg/ml to distinguish dead cells from living cells. Biotin-conjugated anti-CD25 (7D4), FITC-conjugated anti-CD25 (7D4), FITC-conjugated CD69 (H1.2F3), PE-conjugated anti-CD4 (RM4-5) and PerCP-conjugated streptavidin were purchased from BD Biosciences Pharmingen. FITC-conjugated anti-CD62L (MEL-14) and anti-CD44-FITC (IM7.8.1) were from CALTAG Laboratories (Burlingame, CA, USA).
For intracellular staining with anti-CTLA-4-PE (UC10-4F10-11; BD Biosciences Pharmingen), 10^6^cells were first labelled with anti-CD25-FITC as described above. Then, cells were fixed, permeabilised and stained with anti-CTLA-4-PE using the Cytofix/Cytoperm™ Kit (BD Biosciences Pharmingen) according to the recommendations of the manufacturers.
Flow-cytometric analysis was performed on a FACScan flow cytometer with Cell Quest software (Becton Dickinson).
Proliferation assays
--------------------
CD4^+^CD25^-^cells (5 × 10^4^per well) were cultured in U-bottomed 96-well plates (200 μl) with ACs (5 × 10^4^per well, 30 Gy γ-irradiated or treated with mitomycin-C (Sigma-Aldrich)), 3 μg/ml anti-CD3 and the indicated numbers of CD4^+^CD25^+^cells for 48 hours at 37°C in 7% CO~2~. Cultures were pulsed for the last 16 hours with 1 μCi of \[^3^H\]TdR and harvested. The suppressive activity of the T~reg~cells can be presented by plotting the percentage of inhibition (100 × (Radioactivity in condition without T~reg~cells -- Radioactivity in condition with T~reg~cells)/Radioactivity in condition without T~reg~cells) against the number of T~reg~cells.
Antibody administration
-----------------------
DBA/1 mice were immunised with CII in CFA; 13 days after immunisation, the mice were treated every second day with 0.25 mg of anti-CD25 (PC61) or control IgG antibodies, for 4 weeks (injected intraperitoneally).
Histological examination
------------------------
Forelimbs and hindlimbs were fixed in 10% formalin and decalcified with formic acid (31.5% (v/v) formic acid and 13% (w/v) sodium citrate). The paraffin sections were stained with haematoxylin and eosin.
Measurement of serum anti-CII antibodies
----------------------------------------
Blood samples were taken from the orbital sinus and were allowed to clot at room temperature for about 1 hour, and at 4°C overnight. Individual sera were tested by ELISA for antibodies directed against chicken CII. In brief, ELISA plates (Maxisorb; Nunc, Wiesbaden, Germany) were coated overnight at 4°C with native CII (1 μg/ml; 100 μl per well) in coating buffer (50 mM Tris-HCl, pH 8.5, 0.154 mM NaCl), followed by incubation for 2 hours with blocking buffer (50 mM Tris-HCl, pH 7.4, 0.154 mM NaCl and 0.1% caseine) to saturate non-specific binding sites. Serial twofold dilutions of the sera in assay buffer (50 mM Tris-HCl, pH 7.4, 154 mM NaCl and 0.05% Tween 20) were added and incubated for 2 hours at room temperature. The plates were then incubated for 2 hours with peroxidase-conjugated goat anti-mouse IgG (Jackson ImmunoResearch Laboratories, West Grove, PA, USA). Finally, the substrate 3,3\',5,5\'-tetramethylbenzidine (Sigma-Aldrich) in reaction buffer (100 mM sodium acetate/citric acid, pH 4.9) was added for a 10 min incubation and absorbance was determined at 450 nm. Plates were washed five times between each step with PBS containing 0.05% Tween 20. A serial twofold dilution series of a purified standard was included to permit a calculation of the antibody content of each sample. The standard was purified by affinity chromatography from pooled sera obtained from various arthritic wild-type and IFN-γR KO mice.
Quantitative RT-PCR
-------------------
Isolated CD4^+^CD25^+^and CD4^+^CD25^-^cells were pelleted and directly used for total RNA isolation, using the Micro-to-Midi Total RNA Purification System (Invitrogen Life Technologies, Carlsbad, CA, USA). Total RNA (1 μg) was used for random primed cDNA synthesis with RAV-2 reverse transcriptase (Amersham, Aylesbury, Bucks., UK). The reaction mixture was incubated for 80 min at 42°C and the reverse transcriptase was inactivated by incubating the cDNA samples for 5 min at 95°C.
The cDNA samples were then subjected to real-time quantitative PCR, performed in the ABI prism 7700 sequence detector (Applied Biosystems, Foster City, CA) as previously described \[[@B29]\]. The sequences of the forward (-FW) and reverse (-RV) primers and probes (-TP) for β-actin and Foxp3 were as follows: β-actin-FW, AGA GGG AAA TCG TGC GTG AC; β-actin-RV, CAA TAG TGA TGA CCT GGC CG T; β-actin-TP, CAC TGC CGC ATC CTC TTC CTC CC; Foxp3-FW, CCC AGG AAA GAC AGC AAC CTT; Foxp3-RV, TTC TCA CAA CCA GGC CAC TTG; Foxp3-TP, ATC CTA CCC ACT GCT GGC AAA TGG AGT C; TGF-β-FW, TGA CGT CAC TGG AGT TGT ACG G; TGF-β-RV, GGT TCA TGT CAT GGA TGG TGC; TGF-β-TP, TTC AGC GCT CAC TGC TCT TGT GAC AG. Probes were dual-labelled with 5\'-FAM and 3\'-TAMRA.
All primers and probes were designed with the assistance of the computer program Primer Express (AB) and were purchased from Eurogentec (Seraing, Belgium). The 5\'-nuclease activity of the *Taq*polymerase was used to cleave a nonextendable dual-labelled fluorogenic probe. Fluorescent emission was measured continuously during the PCR reaction. PCR amplifications were performed in a total volume of 25 μl containing 5 μl of cDNA, 12.5 μl of Universal PCR Master Mix, no AmpErase UNG (AB), each primer at 100 to 300 nM, and the corresponding detection probe at 200 nM. Each PCR amplification was performed in triplicate wells under the following conditions: 94°C for 10 min, followed by 40 or 45 cycles at 94°C for 15 s and 60°C for 1 min. cDNA plasmid standards, consisting of purified plasmid DNA specific for each individual target, were used to quantify the target gene in the unknown samples, as described \[[@B29]\]. All results were normalised to β-actin and/or hypoxanthine--guanine phosphoribosyltransferase (HPRT) to compensate for differences in the amount of cDNA in all samples. Results were similar whether β-actin or HPRT was used as the housekeeping gene.
Results
=======
Effect of treatment *in vivo*with depleting anti-CD25 antibodies on the development of CIA in wild-type DBA/1 mice
------------------------------------------------------------------------------------------------------------------
In a first set of experiments we tested the importance of T~reg~cells in the pathogenesis of CIA by rendering wild-type mice deficient in T~reg~cells by treating the mice with depleting anti-CD25 antibody. Starting from day 11 or 13 after immunisation with CII in CFA, wild-type DBA/1 mice were treated every second day with anti-CD25 antibodies or control IgG. In a first experiment, female mice were chosen because these are only moderately sensitive to CIA \[[@B30],[@B31]\], so that we would be able to detect both increased and decreased disease severity after CD25^+^cell depletion. Blood samples were taken at intervals to confirm the depletion of the CD25^+^population (Fig. [1a](#F1){ref-type="fig"}). In control-treated mice, the development of arthritis (day of onset, incidence and mean limb score) was reminiscent of our previously reported findings in which mice received a single immunisation with CII in CFA \[[@B20]\]. In contrast, mice treated with the anti-CD25 antibodies developed a significantly more severe arthritis with a higher incidence and earlier onset than those receiving control IgG1 (Fig. [1b](#F1){ref-type="fig"}). In fact, the disease course in antibody-treated mice was very similar to that of IFN-γR KO mice \[[@B20]-[@B22]\]. The results were confirmed in an additional experiment with female mice. A third experiment was also performed on male animals. The data are plotted in Fig. [1c](#F1){ref-type="fig"}. Here again, anti-CD25-treated mice developed a higher incidence and a more severe form of arthritis than control-treated mice, whereas the onset of arthritis was not significantly earlier (Fig. [1d](#F1){ref-type="fig"}). The data from the three experiments were pooled and the percentages of limbs with the different scores from only arthritic mice in the two groups are shown in Fig. [1d](#F1){ref-type="fig"}. It can be seen that, at an early time point (day 27 after immunisation), the highest scores of arthritis (scores 3 and 4) were already present in anti-CD25-injected mice, but not yet in their control counterparts. On day 40 after immunisation, mice treated with anti-CD25 developed more limbs with a maximum score of 4 than control-treated mice. The mean limb score on the two days for the two groups are indicated and are significantly different (*P*\< 0.05, Mann--Whitney *U*-test). The mean number of involved limbs, ± SEM, on day 40 was 2.8 ± 0.2 and 2.2 ± 0.2 for the treated and control mice, respectively (*P*= 0.07; Mann--Whitney *U*-test). Representative pictures of the most severe case of arthritis of anti-CD25-injected and control mice on day 25 after immunisation are shown in Fig. [1e](#F1){ref-type="fig"} and Fig. [1f](#F1){ref-type="fig"}, respectively. To ensure that the more severe form of arthritis in the anti-CD25-treated mice was not merely due to oedema, some mice were killed at day 42 for histological evaluation. The presence of hyperplasia and infiltration of immunocompetent cells in the synovium, pannus formation and osteoclast-like multinucleated giant cells confirmed the authenticity of arthritis (Fig. [1g](#F1){ref-type="fig"}).
On day 35 after immunisation, the titres of collagen-specific antibodies in the sera were determined. No differences in antibody levels in sera of mice treated with anti-CD25 or control IgG could be detected (data not shown).
Number and phenotype of CD4^+^CD25^+^T~reg~cells in IFN-γR KO and wild-type mice
--------------------------------------------------------------------------------
To test whether T~reg~cells might be less numerous in IFN-γR KO than in wild-type mice -- because this might explain the differences in susceptibility to CIA -- we counted CD4^+^CD25^+^cells in thymus, lymph nodes and spleen by flow cytometry. IFN-γR KO and wild-type mice were immunised with CII in CFA on day 0. Thymocytes, splenocytes and lymph node cells were obtained on day 21, a time point at which the difference in severity of arthritis between the two groups of mice is most pronounced \[[@B20]-[@B22]\]. Groups of naive IFN-γR KO and wild-type mice were also included. A typical CD4/CD25 staining pattern of thymocytes and lymph node cells from IFN-γR KO and wild-type mice is shown in Fig. [2](#F2){ref-type="fig"}; percentages of CD4^+^CD25^+^and CD4^+^CD25^-^cells are indicated. It can be seen that IFN-γR KO mice did not have smaller proportions of CD4^+^CD25^+^cells in the thymus and lymph nodes. Immunised mice, whether wild-type or IFN-γR KO, had rather lower proportions of total CD4^+^cells than naive counterparts (for example 31% versus 50% in wild types). However, the real numbers of CD4^+^cells per organ were in fact higher after immunisation and did not differ in IFN-γR KO from those in wild-type mice. In fact, the lower percentages of CD4^+^cells after immunisation were due to a still larger expansion of the myelopoietic population, a well-recognised phenomenon arising from the use of CFA \[[@B22],[@B32]\].
When over a total of six experiments (Table [1](#T1){ref-type="table"}) the numbers of CD4^+^CD25^+^cells were expressed as fractions of total CD4^+^cell numbers, it appeared that spleens and lymph nodes of IFN-γR KO mice, naive as well as immunised ones, contained slightly higher percentages of CD4^+^CD25^+^cells. In spleens and lymph nodes of wild-type mice, 5 to 10% of the CD4^+^T cells were CD25^+^, conforming to previously published figures obtained in other mouse strains. Thymuses contained lower percentages of CD4^+^CD25^+^cells. A possible explanation might be that thymic CD4^+^T cell populations contain not only CD4^+^CD8^-^but also CD4^+^CD8^+^cells, the latter being mostly CD25^-^. In the peripheral lymphoid organs of IFN-γR KO mice, the percentage of CD4^+^CD25^+^cells was higher (7 to 14%) than in the wild-type mice (Table [1](#T1){ref-type="table"}).
Because CD25 is expressed not only by T~reg~cells but also by other recently activated T cells, the slightly higher proportion of CD4^+^CD25^+^cells in IFN-γR KO mice is not synonymous with a higher proportion of T~reg~cells. In fact, even a lower proportion of such cells cannot be excluded. We therefore compared the CD4^+^CD25^+^T cells from IFN-γR KO and wild-type DBA/1 mice for expression of various other activation markers. Figure [3a,b](#F3){ref-type="fig"} shows flow-cytometric expression patterns of CD69, CD62L, CD44 and cytolytic T lymphocyte-associated antigen (CTLA-4) in CD4^+^CD25^+^T cells from naive and immunised IFN-γR KO and wild-type mice. No major differences in expression levels of these activation markers could be detected between CD4^+^CD25^+^T cells from IFN-γR KO mice and those from wild-type mice, whether naive or immunised. Thus, this analysis did not provide evidence for different proportions of any cell type, including T~reg~cells. A specific marker for T~reg~cells is Foxp3. We determined mRNA for this marker by quantitative PCR in CD4^+^CD25^+^and CD4^+^CD25^-^cells, sorted from the lymph node cells of naive or immunised IFN-γR KO and wild-type DBA/1 mice at day 21. In CD4^+^CD25^-^cells Foxp3 mRNA levels were extremely low (less than 6), and not different between one group of mice and the other. CD4^+^CD25^+^cells, in contrast, displayed high expression levels. In cells from naive IFN-γR KO and wild-type mice, levels were comparable. However, CD4^+^CD25^+^T cells of immunised IFN-γR KO mice contained levels of Foxp3 that were one-third of those of wild-type mice (Fig. [3c](#F3){ref-type="fig"}). This lower expression level might be indicative of a smaller proportion of T~reg~cells in the sorted CD4^+^CD25^+^cell population or of a lower expression level per cell. To distinguish between these alternatives, a tagging anti-Foxp3 antibody would be needed.
Thus, after immunisation, IFN-γR KO mice possessed a slightly higher percentage of CD4^+^CD25^+^cells than wild-type mice. However, the actual T~reg~cells present in this population might be considerably less numerous or might be qualitatively different so as to express less Foxp3.
Reduced suppressive activity of CD4^+^CD25^+^T~reg~cells in arthritic IFN-γR KO mice
------------------------------------------------------------------------------------
To characterise the CD4^+^CD25^+^T~reg~cells functionally, we measured their ability to suppress the anti-CD3-induced proliferation of CD4^+^CD25^-^T~eff~cells *in vitro*. The experiments were performed with CD4^+^CD25^+^cells, CD4^+^CD25^-^cells and ACs. T~reg~suppressive activity was presented by plotting the percentage of inhibition against the number of T~reg~cells. As shown in Fig. [4a,c](#F4){ref-type="fig"}, the patterns of inhibition in naive IFN-gR KO and wild-type mice were very similar: in both cases 2 × 10^4^purified CD4+CD25+ cells were able to inhibit more than 90% of the proliferative response of 5 × 104 T~eff~cells. This result indicates that IFN-γ is not required for T~reg~cells to be able to suppress anti-CD3-induced *in vitro*proliferation.
In a separate set of seven experiments we investigated the suppressive effect of CD4^+^CD25^+^cells from mice that had been immunised with CII in CFA. IFN-γR KO and wild-type DBA/1 mice were immunised on day 0, and CD4^+^CD25^+^cells, T~eff~cells and ACs were isolated on day 21 after immunisation. The data of the individual experiments are plotted in Fig. [4b](#F4){ref-type="fig"} and the means of the seven experiments are shown in Fig. [4c](#F4){ref-type="fig"}. It can be seen that the capacity to suppress TCR-triggered proliferation of T~eff~cells was significantly lower in CD4^+^CD25^+^cells isolated from immunised mice than in those of naive animals. Indeed, to obtain 40% inhibition of proliferation, 4.5 × 10^3^CD4^+^CD25^+^cells from immunised wild-type mice were required, in comparison with only 1.5 × 10^3^CD4^+^CD25^+^cells from naive wild-type mice. Moreover, CD4^+^CD25^+^cells from immunised IFN-γR KO mice were significantly less suppressive than those of immunised wild-type mice: 10^4^CD4^+^CD25^+^cells were necessary to decrease T~eff~cell proliferation by 40%. In an additional experiment we verified whether the deficit in inhibition by CD4^+^CD25^+^cells from immunised IFN-γR KO mice could be corrected by adding excess CD4^+^CD25^+^cells. However, with 2 × 10^4^and 4 × 10^4^CD4^+^CD25^+^cells the inhibition on T cell proliferation was 64.6% and 65.8%, respectively, indicating that a plateau level of suppressive activity had been reached.
Normal levels of TGF-β in IFN-γR KO and wild-type mice
------------------------------------------------------
Several studies have shown the critical role of TGF-β in the induction of Foxp3 and the activity of T~reg~cells \[[@B10],[@B33],[@B34]\]. Because IFN-γ and TGF-β act antagonistically with each other (reviewed in \[[@B35]\]), it is possible that TGF-β is upregulated in wild-type mice as a homeostatic response to IFN-γ produced by their activated T cells, and similarly in IFN-γR KO mice the decreased Foxp3 levels and the decreased suppressive activity of T~reg~cells might be due to inadequate amounts of TGF-β produced in the co-cultures or *in vivo*in mice. We therefore analysed the expression of TGF-β by quantitative PCR in T~reg~cells as well as in co-cultures and in spleens of naive and immunised mice. The following results were obtained. First, the levels of TGF-β from the sorted CD4^+^CD25^+^cells from immunised IFN-γR KO mice were not different from those of wild-type mice (normalised TGF-β mRNA levels were 179 ± 16 and 193 ± 22, respectively; mean ± SEM for three measurements). Second, because TGF-β might be produced by ACs (or T~eff~cells), quantitative PCR was performed on cells obtained from co-cultures (T~reg~plus T~eff~plus ACs) from immunised IFN-γR KO and wild-type mice. It was found that the levels of TGF-β were even increased in IFN-γR KO cells in comparison with wild-type cells (2,184 versus 1,574, respectively, in the condition of 2 × 10^4^T~reg~cells, in a pool of eight mice). Third, the TGF-β levels were also analysed *ex vivo*; that is, in spleen tissue from IFN-γR KO and wild-type mice at day 21 after immunisation (thus at a time point at which T~reg~, T~eff~and ACs were isolated). Here again, the TGF-β levels were found to be slightly increased in spleens from IFN-γR KO mice (816 ± 129 and 633 ± 40 for IFN-γR KO and wild-type mice, respectively). If these results are taken together, the defective activity of T~reg~cells from arthritic IFN-γR KO mice (in comparison with those from wild-type animals) seems not to be associated with a defective TGF-β production.
It was notable that the TGF-β levels were higher in immunised mice than in their naive counterparts (for example, 633 ± 40 and 205 ± 19 for immunised and naive wild-type mice, respectively). These data suggest that the differences in suppressive activity of T~reg~cells from immunised versus naive mice cannot be explained by differences in the TGF-β production.
T~reg~cells from immunised IFN-γR KO mice have the capacity to inhibit proliferation responses
----------------------------------------------------------------------------------------------
We next investigated whether the lower capacity of CD4^+^CD25^+^cells from IFN-γR KO mice to downregulate proliferation responses is due to an intrinsic defect or to an altered activity of surrounding ACs and T~eff~cells. We measured the inhibition of anti-CD3-induced proliferation in co-cultures differently reconstituted of CD4^+^CD25^+^, CD4^+^CD25^-^and ACs, derived either from the same or from different immunised wild-type or immunised IFN-γR KO mice. The combinations tested are indicated in Fig. [5](#F5){ref-type="fig"}.
As expected, when all cells in the reconstituted co-cultures were of IFN-γR KO mouse origin, suppressive activity was less than when all cells were of wild-type origin. In co-cultures of mixed composition, suppressive activity of IFN-γR KO-derived CD4^+^CD25^+^cells was less than that of the wild type only when ACs were from IFN-γR KO origin, but not when they were of wild-type origin. However, such ACs of IFN-γR KO mice were unable to reduce the suppressive effect of wild-type T~reg~cells against wild-type or IFN-γR KO (data not shown) T~eff~cells. These data demonstrate that the defect in inhibiting CD4^+^CD25^-^T~eff~cells acquired the presence of T~reg~cells from immunised IFN-γR KO mice in combination with their autologous ACs.
Discussion
==========
We and others have previously demonstrated that IFN-γ(R) KO mice show an accelerated and more severe from of arthritis than their wild-type counterparts, indicating that endogenous IFN-γ acts as a protective factor in CIA \[[@B20],[@B21],[@B24],[@B25]\]. Because CIA has been defined as a Th1-driven disease (reviewed in \[[@B17]\]), the protective effect of IFN-γ in CIA constitutes an enigma that compromises the Th1/Th2 paradigm as a basis for explaining the regulation of autoimmune diseases. A clue to the enigma seemed to be the use of CFA in the induction procedure of CIA. In the absence of IFN-γ, CFA induces an extensive extramedullary myelopoiesis that goes together with an even more pronounced Th1 cytokine profile than in wild-type counterparts \[[@B22],[@B36]\]. The data suggest that IFN-γ can, under certain circumstances, be a strong Th2 inducer, a finding that has recently been confirmed by others \[[@B37]\]. Here, we tested the hypothesis that this protective action of IFN-γ is due to a stimulatory effect on T~reg~cells. Specifically, we addressed the following two questions. Are T~reg~cells important in modulating CIA? And, because we found that depletion of T~reg~cells in wild-type mice increased the severity of CIA, can the higher susceptibility of IFN-γR KO mice to CIA be explained by defects in the number or function of their T~reg~cells?
As to the first question, we found that administration of a T~reg~cell-depleting anti-CD25 antibody to wild-type DBA/1 mice after CFA-assisted immunisation with CII resulted in accelerated and more severe arthritis. In fact, the disease course in these mice was comparable to that in IFN-γR KO mice \[[@B20]-[@B23]\]. The actual depletion of T~reg~cells was monitored by flow cytometry, and the authenticity of arthritis was verified histopathologically. These results are in line with those of Morgan and colleagues \[[@B26]\], who showed that the administration of depleting anti-CD25 antibody before immunisation (days --28, --24, --21 and --14) hastened the onset of severe CIA. Because in our experiments antibodies were administered starting from day 11 or day 13 after immunisation, we can conclude that T~reg~cells are important in the pathogenesis of CIA, not only in the immunisation phase but also in the effector phase. In contrast to the findings of Morgan and colleagues \[[@B26]\], the accelerated and more severe course of arthritis was, in our experiments, not accompanied by a higher concentration of anti-collagen II antibodies, possibly due to the different regimen of anti-CD25 treatment. Indirect evidence for the involvement of T~reg~cells in the pathogenesis of CIA comes from data of Min and colleagues \[[@B38]\]. They found that the immune tolerance induced by oral feeding of CII before induction of CIA was mediated by IL-10-producing CD4^+^CD25^+^T cells. Notably, in proteoglycan-induced arthritis, another model of autoimmune arthritis, it has been shown that CD4^+^CD25^+^T~reg~cells might not have a critical role \[[@B39]\]. This might result from the use of a different auto-antigen.
To address the second question, we compared CD4^+^CD25^+^cell numbers and T~reg~cell function in IFN-γR KO DBA/1 mice with those in wild-type mice. According to our hypothesis we expected numbers of T~reg~cells in IFN-γR KO mice to be lower. Counter to this expectation, in each of the six experiments done, we found a trend for a higher proportion of CD4^+^CD25^+^T cells in the total CD4^+^cell population. This was true for thymic, splenic and lymph node CD4^+^cells, in both naive and immunised mice. Analysis of all data as one set revealed a significant difference of about 30% and 20% in naive and immunised mice, respectively. CD25 is not an exclusive marker of T~reg~cells: especially in immunised mice, part of the CD4^+^CD25^+^population might be effector rather than regulatory T cells \[[@B40],[@B41]\]. Therefore, to exclude the possibility that we were comparing two completely different populations, we performed additional flow-cytometric characterisation studies on pre-sorted CD4^+^CD25^+^cells.
Expression of CD44, CD69, CTLA-4 and CD62L in CD4^+^CD25^+^cells from IFN-γR KO mice did not differ from expression in cells from corresponding wild-type mice, whether naive or immunised. However, because T~reg~cells display an activated phenotype, activation markers might not be adequate to distinguish T~reg~cells from activated T~eff~cells. According to Fontenot and colleagues \[[@B42]\] a specific marker for T~reg~cells is Foxp3, because it is highly expressed in CD4^+^CD25^+^T~reg~cells and is virtually undetectable in both resting and activated T~eff~cells. We examined Foxp3 expression by determining mRNA levels with PCR. After immunisation, CD4^+^CD25^+^cells contained lower levels of Foxp3 mRNA than those of their naive counterparts. Moreover, mRNA levels in immunised IFN-γR KO mice were less than one-third of those in their wild-type counterparts, indicating that IFN-γR KO mice have a smaller number of T~reg~cells or that expression of Foxp3 in each T~reg~cell is lower.
Recently, Bruder and colleagues \[[@B43]\] have shown linked expression of neuropilin-1 and Foxp3, thereby identifying neuropilin-1 as a specific surface marker for CD4^+^CD25^+^T~reg~cells able to distinguish them from both naive and recently activated CD4^+^CD25^+^non-regulatory T cells.
Nishibori and colleagues \[[@B44]\] demonstrated impaired development of T~reg~cells in naive signal transduction and activators of transcription (STAT)-1-deficient mice, associated with an increased susceptibility to autoimmune disease. Because IFN-γ is among the strongest activators of STAT-1, these observations seem to conflict with ours. However, several cytokines, other than IFN-γ, can also activate STAT-1, including IFN-α, IFN-β, IL-6, IL-9, IL-11, oncostatin M, leukaemia inhibitory factor and the chemokines RANTES and macrophage inflammatory protein 1α \[[@B45],[@B46]\].
To determine whether overall T~reg~cell activity would be lower in IFN-γR KO mice, we co-cultured increasing numbers of CD4^+^CD25^+^T cells with fixed numbers of CD4^+^CD25^-^T~eff~cells and ACs in the presence of anti-CD3 antibody. We observed a dose-dependent inhibition of the proliferative responses by CD4^+^CD25^+^T~reg~cells. By estimating numbers of CD4^+^CD25^+^cells required to attain a selected level of suppression, we could compare suppressive activity in the different groups of mice. In naive mice, the inhibition curves were almost identical, whether the T~reg~cells were derived from wild-type or IFN-γR KO mice, indicating that endogenous IFN-γ is not an important regulator of the function of constitutive CD4^+^CD25^+^T~reg~cells. In co-cultures of cells from immunised wild-type mice, the T~reg~suppressive capacity was about one-third of that in those from corresponding naive mice, and a further halving was noted in co-cultures of cells from immunised IFN-γR KO mice.
The observation that immunisation renders T~reg~cells less suppressive is in line with results of Pasare and Medzhitov \[[@B47]\], who found that microbial triggering of the Toll-like receptor (TLR) pathway by lipopolysaccharide or CpG, which are ligands for TLR4 and TLR9, respectively, blocked the suppressive effect of CD4^+^CD25^+^T~reg~cells. Because mycobacteria also contain TLR ligands, immunisation with CFA can be expected to affect T~reg~cell activity similarly. The decrease in suppressive activity that takes place after TLR4 or TLR9 triggering was found to be dependent on IL-6 production \[[@B47]\]. It might therefore be of interest to note that in our experiments, IL-6 production was enhanced after exposure to CFA-assisted immunisation, and this effect was even more pronounced in IFN-γR KO mice (P Matthys, unpublished data). This could provide an explanation for the fact that CD25^+^T~reg~cells are totally functional before immunisation but lose (part of) their function after immunisation. However, the most important observation is the lower T~reg~suppressive capacity in IFN-γR KO than in wild-type mice after CFA-assisted immunisation, because this supports our hypothesis that the protective effect of endogenous IFN-γ against CIA could be mediated in part by its stimulatory effect on T~reg~cells.
Because the disease is barely detectable in wild-type mice on day 21 after immunisation, we investigated whether the decreased suppressive activity in immunised wild-type mice was further downregulated at a later time point (namely, day 35 after immunisation, when most of the animals show symptoms of arthritis). However, suppressive activity was not further downregulated to the level seen in homogeneous IFN-γR KO co-cultures, but was comparable to that seen in co-cultures from immunised wild-type mice on day 21 after immunisation (maximal inhibition 60%; data not shown). This indicates that the low suppressive activity as evident in immunised IFN-γR KO mice is restricted to conditions under which IFN-γ is abrogated.
The implication is that the CIA immunisation schedule induces a decrease in T~reg~activity and that endogenous IFN-γ largely counteracts this decrease. It therefore becomes important to know by what mechanism, direct or indirect, IFN-γ influences T~reg~cell function. Addition of anti-IFN-γ antibody to the co-cultures failed to affect suppressive activity (data not shown), indicating that the relevant IFN-γ effect takes place *in vivo*before sampling of the T cells. To examine the role of the different cell components, we tested suppressive activity in mixed co-cultures. CD4^+^CD25^+^cells from immunised IFN-γR KO mice, confronted with T~eff~cells and ACs from immunised wild-type mice, were not less suppressive than wild-type CD4^+^CD25^+^confronted with wild-type or IFN-γR KO T~eff~cells and ACs. This suggests that lower levels of suppression in homogeneous IFN-γR KO cultures result in part from the presence of IFN-γR KO-derived ACs. And, indeed, when CD4^+^CD25^+^and T~eff~cells from immunised IFN-γR KO mice were co-cultured with ACs from immunised wild-type mice, suppressive activity was not inhibited. Finally, ACs from IFN-γR KO mice by themselves were unable to downregulate the activity of wild-type T~reg~cells acting on wild-type T~eff~cells. We therefore conclude that the *in vivo*effect of endogenous IFN-γ that accounts for the greater suppressive activity in wild-type mice than in IFN-γR KO mice concerns reprogramming of both ACs and T~reg~cells.
Because CD4^+^CD25^+^T cells from immunised IFN-γR KO mice were not less suppressive than those of immunised wild-type mice in co-cultures with T~eff~cells and ACs from immunised wild-type mice, we can refute the proposition that the lower expression of Foxp3 in the CD4^+^CD25^+^population from immunised IFN-γR KO mice is due to a smaller proportion of T~reg~cells and a larger number of activated T~eff~cells. Indeed, if the CD4^+^CD25^+^population from immunised IFN-γR KO mice contained a higher proportion of activated T~eff~cells, suppression by these CD4^+^CD25^+^cells should be lower, irrespective of the origin of the T~eff~cells and ACs. Another argument is that the addition of more CD4^+^CD25^+^cells failed to improve suppression in co-cultures of cells from immunised IFN-γR KO mice. Our data are therefore more in line with the proposition of a lower Foxp3 expression level per cell.
Expression of Foxp3 could be downregulated by the interaction of T~reg~cells with ACs. ACs might be source of TGF-β, which has been described to convert naive T cells into CD25^+^suppressor cells by inducing Foxp3 expression \[[@B48]\]. Because IFN-γ and TGF-β act antagonistically with each other, the low levels of Foxp3 in arthritic IFN-γR KO mice might be due to inadequate amounts of TGF-β produced by ACs or other cells. However, quantitative PCR performed on isolated T~reg~cells, on cells obtained from co-cultures (T~reg~plus T~eff~plus ACs) and on splenocytes from immunised IFN-γR KO and wild-type mice does not support the concept that the defective activity of T~reg~cells *in vitro*or *in vivo*is due to defects in the production of TGF-β. ACs have also been shown to be able to reverse suppression by CD4^+^CD25^+^cells through the GITR/GITR-ligand system \[[@B49]\]. GITR (glucocorticoid-induced tumour necrosis factor receptor) is expressed on CD4^+^CD25^+^T cells; GITR-ligand is initially upregulated on activated APCs. It remains to be determined whether this process involves a downregulation of Foxp3 expression. This or a similar mechanism might take place during the interaction of T~reg~cells and ACs from immunised IFN-γR KO mice. Co-cultures with ACs of immunised wild-type mice might possibly normalise Foxp3 expression in the T~reg~cells of immunised IFN-γR KO mice, together with their T~reg~suppressive activity.
Conclusions
===========
In conclusion, our experiments support a pathogenesis model that ascribes an important role to T~reg~cells as moderators of the disease course in CIA. In particular we show that T~reg~cells fulfil this role not only during the induction phase but also during the effector phase of the autoimmune response. Furthermore, we were able to refine the model by showing that, after the immunisation with CII in CFA, T~reg~cells lose part of their suppressive potential. This effect is more pronounced in IFN-γR KO than in wild-type mice, indicating that, in this system, IFN-γ acts as an upregulator of T~reg~activity, which might be part of the explanation for the well-known protective effect of endogenous IFN-γ. Finally, we present evidence that the mechanism underlying the effect of IFN-γ on T~reg~cell activity is exerted in part via ACs.
Abbreviations
=============
ACs = accessory cells; CFA = complete Freund\'s adjuvant; CIA = collagen-induced arthritis; CII = collagen type II; CTLA = cytolytic T lymphocyte-associated antigen; ELISA = enzyme-linked immunosorbent assay; fetal calf serum = FCS; FITC = fluorescein isothiocyanate; GITR = glucocorticoid-induced tumour necrosis factor receptor; IFN-γ = interferon-γ; IFN-γR KO = interferon-γ receptor knock-out; IL = interleukin; MACS = magnetic-activated cell sorting; PBS = phosphate-buffered saline; RT-PCR = reverse transcriptase polymerase chain reaction; STAT = signal transduction and activators of transcription; T~eff~= effector T; TGF = transforming growth factor; TLR = Toll-like receptor; T~reg~= regulatory T.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
BDK, HK and TM performed the CIA induction and evaluation. HK, MVB and GL performed the cell purification. DB performed the quantitative PCR. TM and HK performed the flow cytometry. BDK and HK did the *in vitro*experiments. HK, GL and PM designed the study. All authors participated in the interpretation of the data. HK, AB, GL and PM prepared the manuscript. All authors read and approved the final manuscript.
Acknowledgements
================
We thank C Dillen, E Dilissen, W Landuyt and O Rutgeerts for excellent assistance and helpful discussions. This work was supported by grants from the Fund of Scientific Research Flanders (FWO Vlaanderen), from the Regional Government of Flanders (GOA Program), and from the Belgian Federal Government (Interuniversity Network for Fundamental Research, IUAP). PM and DB are postdoctoral research fellows of the FWO Vlaanderen, and HK holds a fellowship from the FWO Vlaanderen.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Wild-type mice treated with anti-CD25 antibodies develop a more severe form of arthritis. In three experiments, wild-type DBA/1 mice were immunised on day 0 with collagen type II in complete Freund\'s adjuvant. From day 11 **(c)**or 13 **(b)**after immunisation onwards, mice were treated every second day with 0.25 mg of depleting anti-CD25 monoclonal antibody (*N*= 7) or with 0.25 mg control rat IgG (*N*= 7). **(a)**Depletion of the CD25^+^cell population was checked in the blood twice a week by flow-cytometric analysis with anti-CD4 and anti-CD25 antibodies. A representative staining pattern on day 27 is shown. The percentages of CD4^+^CD25^+^cells in control-treated mice (left plot) and anti-CD25-treated mice (right plot) are shown. **(b, c)**Cumulative incidence of arthritis (and mean day of disease onset) and the mean limb score of the arthritic mice in female **(b)**and male **(c)**wild-type mice treated with anti-CD25 or control IgG are shown (the maximum score per limb is 4). Error bars indicate SEM. The data from the female mice are representative of two independent experiments. The data of the three experiments were pooled and the percentage of limbs with each limb score on days 27 and 40 after immunisation is shown in **(d)**. The mean limb score of the arthritic mice in the two groups is also indicated for the two time points and is significantly higher in the treated mice (*P*\< 0.05; Mann--Whitney *U*-test) than in those receiving control IgG. **(e, f)**Representative pictures of the most severe case of collagen-induced arthritis on day 25 after immunisation of a mouse treated with anti-CD25 **(e)**and a mouse treated with control IgG **(f)**. **(g)**Haematoxylin-stained paraffin section of the joint of an anti-CD25-treated mouse on day 42 after immunisation. Hyperplasia and infiltration of immunocompetent cells in the synovium (s) and pannus formation (p) that penetrates into the bone (b) can be seen. Note the presence of osteoclast-like multinucleated giant cells (arrow). \**P*\< 0.05 for comparison with control IgG1-treated mice (Mann--Whitney *U*-test).
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
IFN-γ is not required to establish normal numbers of CD4^+^CD25^+^T~reg~cells. Thymus cells **(a)**and lymph node cells **(b)**were isolated from IFN-γR KO and wild-type DBA/1 mice, either naive (upper row) or having been immunised 21 days previously with collagen type II in complete Freund\'s adjuvant (collagen-induced arthritis (CIA), lower row). Cells were stained with anti-CD25-FITC, phycoerythrin-conjugated anti-CD4 and propidium iodide. Dead cells were excluded by gating on propidium iodide-negative cells. The percentages of cells in each quadrant are indicated. Each plot represents a staining pattern of cells from a single female mouse. Identical profiles were observed in male mice. The staining pattern is representative of data obtained in three experiments (Table 1).
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Phenotypic characterisation of CD4^+^CD25^+^T cells from immunised IFN-γR KO and wild-type DBA/1 mice. **(a, b)**CD4^+^CD25^+^T cells isolated from IFN-γR KO and wild-type mice show a similar expression pattern of activation markers, in naive **(a)**and immunised **(b)**conditions. CD4^+^T cells were purified from the lymph node cells of eight IFN-γR KO and wild-type DBA/1 mice, either naive or having been immunised 21 days previously with collagen type II in complete Freund\'s adjuvant (purity more than 99%). CD4^+^T cells were stained for CD25 in combination with CD69, CD62L, CD44 or cytolytic T lymphocyte-associated antigen-4 (CTLA-4). Dead cells were excluded by gating on propidium iodide-negative cells. The numbers represent the percentages of CD4^+^CD25^+^cells within the indicated marker. **(c)**Decreased Foxp3 mRNA levels in CD4^+^CD25^+^T~reg~cells from immunised mice. Lymph node cells were isolated from eight naive or immunised IFN-γR KO and wild-type DBA/1 mice. Purified CD4^+^T cells were stained with anti-CD25-FITC and phycoerythrin-conjugated anti-CD4, and sorted. The purity of the sorted CD4^+^CD25^+^population was more than 97%. cDNA samples were prepared from 2 × 10^5^cells of each population and were subjected to real-time quantitative PCR analyses. The relative quantity of Foxp3 in each sample was normalised to the quantity of β-actin. Error bars indicate standard error of the means of two (CD4^+^CD24^+^cells from naive mice) or three (CD4^+^CD25^+^cells from immunised mice) independent experiments. \**P*\< 0.05 for comparison with Foxp3 expression of cells isolated from immunised wild-type mice (Mann--Whitney *U*-test).
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Suppressive capacity of CD4^+^CD25^+^cells is decreased more in immunized IFN-γR KO than in wild-type mice. **(a, b)**T~reg~cells, T~eff~cells and accessory cells (ACs) were isolated from lymph nodes and spleen of naive **(a)**IFN-γR KO and wild-type DBA/1 mice or from IFN-γR KO and wild-type DBA/1 mice 21 days after immunisation with collagen type II in complete Freund\'s adjuvant **(b)**. In each case, a group of seven to nine mice was used. CD4^+^CD25^-^T~eff~cells (5 × 10^4^) were incubated with anti-CD3 antibody in the presence of ACs and the indicated number of CD4^+^CD25^+^T~reg~cells. The percentage inhibition (100 × (Radioactivity in condition without T~reg~cells -- Radioactivity in condition with T~reg~cells)/Radioactivity in condition without T~reg~cells) of the proliferation of T~eff~cells (CD4^+^CD25^-^) by increasing numbers of CD4^+^CD25^+^T~reg~cells is shown. Two and seven independent experiments are shown in **(a)**and **(b)**, respectively. Each result is the mean of two cups. **(c)**The means of the two (naive mice) or seven (immunised mice) independent experiments shown in **(a)**and **(b)**. Error bars indicate SEM.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Accessory cells (ACs) of immunised IFN-γR KO mice are required for their defective T~reg~activity. T~reg~cells, T~eff~cells and ACs were isolated from lymph nodes and spleen of IFN-γR KO and wild-type DBA/1 mice 21 days after immunisation with collagen type II in complete Freund\'s adjuvant. Mixing experiments were performed as indicated. In each set, 5 × 10^4^CD4^+^CD25^-^T~eff~cells were incubated with anti-CD3 antibody in the presence of ACs and the indicated number of CD4^+^CD25^+^T~reg~cells. The percentage inhibition of the proliferation of T~eff~cells (CD4^+^CD25^-^) by increasing numbers of CD4^+^CD25^+^T~reg~cells is shown. The results are representative of two independent experiments.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Proportion of regulatory T cells to the total CD4^+^T cell population in lymphoid organs of naive and immunised IFN-γ receptor knock-out and wild-type (WT) DBA/1 mice
:::
100 × CD4^+^CD25^+^/CD4^+^(*N*)
------- --- ------------- --------------------------------- ----------------
Naive 1 Thymus 3.2 ± 0.6 (5) 2.2 ± 0.7 (5)
Spleen 10.1 ± 0.9 (3) \* 7.4 ± 0.2 (3)
Lymph nodes 6.9 ± 1.1 (5) 5.1 ± 1.1 (5)
2 Spleen 14.4 (1) 9.9 (1)
Lymph nodes 9.1 ± 0.9 (4) 6.6 ± 0.7 (4)
3 Lymph nodes 11.2 (1) 7.0 (1)
CIA 4 Thymus 3.5 ± 0.9 (3) 4.0 ± 1.6 (3)
Spleen 11.0 ± 1.3 (2) 7.9 ± 0.7 (2)
Lymph nodes 10.2 ± 0.8 (6) 7.7 ± 0.6 (6)
5 Spleen 12.1 ± 2.9 (3) 9.2 ± 0.8 (3)
Lymph nodes 12.9 ± 1.4 (4) 10.3 ± 1.1 (4)
6 Lymph nodes 13.4 ± 0.4 (4) \* 9.2 ± 0.7 (4)
Cells were obtained from thymuses, spleens or lymph nodes of IFN-γ receptor knock-out (IFN-γR KO) and wild-type DBA/1 mice. In experiments 4 to 6, mice were immunised with collagen type II in complete Freund\'s adjuvant on day 0, and cells were obtained on day 21 (collagen-induced arthritis; CIA). Cells were stained with anti-CD25-FITC and phycoerythrin-conjugated anti-CD4 antibodies. The proportion of CD4^+^CD25^+^in the total CD4^+^T cell population is shown. In experiments 1, 2, 4 and 5, *N*(number in parentheses) indicates the number of mice in each experiment; in experiments 3 and 6, *N*represents the number of experiments, each consisting of groups of 5 to 10 mice, from which samples were pooled for analysis. \*Significant difference between IFN-γR KO and wild-type mice (*P*\< 0.05; Mann--Whitney *U*-test).
:::
|
PubMed Central
|
2024-06-05T03:55:54.975021
|
2005-1-28
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065335/",
"journal": "Arthritis Res Ther. 2005 Jan 28; 7(2):R402-R415",
"authors": [
{
"first": "Hilde",
"last": "Kelchtermans"
},
{
"first": "Bert",
"last": "De Klerck"
},
{
"first": "Tania",
"last": "Mitera"
},
{
"first": "Maarten",
"last": "Van Balen"
},
{
"first": "Dominique",
"last": "Bullens"
},
{
"first": "Alfons",
"last": "Billiau"
},
{
"first": "Georges",
"last": "Leclercq"
},
{
"first": "Patrick",
"last": "Matthys"
}
]
}
|
PMC1065336
|
Introduction
============
Whereas classical analysis of synovial tissue in chronic inflammatory arthritis suggested that synovitis is a nonspecific phenomenon, a number of studies using new molecular tools and synovial biopsies obtained during active disease indicated clear histopathological differences between spondyloarthropathy (SpA) and rheumatoid arthritis (RA), which are the two most frequent forms of chronic autoimmune arthritis \[[@B1]-[@B4]\]. These differences were explored as a diagnostic tool in undifferentiated arthritis \[[@B5],[@B6]\], and highly disease-specific markers as well as less pronounced synovial features turned out to be useful in multiparameter classification models \[[@B7],[@B8]\]. Since some of these features are pathophysiologically related to specific disease mechanisms \[[@B3],[@B4]\], it is tempting to hypothesize that histopathological characteristics of inflamed synovium directly reflect the disease process.
In RA, it has been shown that synovial features may help to distinguish different subgroups corresponding to distinct pathogenetic mechanisms and clinical phenotypes. Indeed, three main histological types of synovitis can be distinguished in RA: one type characterized by follicular organization with high numbers of B cells and plasma cells, another type with diffuse infiltration by essentially T lymphocytes and macrophages, and a third, granulomatous type, which is less frequent \[[@B9]\]. The different histological types are stable over time within one individual, are linked with different cellular and molecular disease pathways, as evidenced, for example, by abundant IL-10 in the follicular type, and are related to phenotypic differences such as seronegativity in the diffuse type and extra-articular manifestations in the granulomatous type \[[@B10],[@B11]\]. Besides distinguishing subtypes within one disease, some of these features may also reflect global disease activity and thus be valuable candidates for evaluation as surrogate markers in trials of new, targeted therapies in autoimmune arthritis. Synovial macrophages have been shown to be related to scores for local disease activity as well as to articular damage in RA \[[@B12],[@B13]\]. Sublining macrophages, but also T cells and plasma cells, were increased in clinically involved versus clinically uninvolved knee joints \[[@B14]\], while sublining macrophages were also increased in joints in active RA compared with joints in end-stage disease \[[@B15]\]. Taken together, these various findings strongly suggest that the number of sublining macrophages is a good reflection of active disease processes in RA. This interpretation was further validated by demonstrating that synovial sublining macrophages can be used as surrogate marker for global disease activity in clinical trials evaluating antirheumatic therapy in RA \[[@B16]\].
In contrast with the situation with RA, these issues have not yet been fully assessed in SpA. We have previously demonstrated a correlation between synovial histology and local disease activity in SpA \[[@B1]\] and the absence of manifest differences in synovial histopathology between psoriatic arthritis (PsA) and other SpA subtypes. Considering the data in RA synovitis, the increase of specific macrophage subsets and polymorphonuclear leukocytes (PMNs) in SpA synovium compared with RA \[[@B2]\], and the strong and rapid reduction of synovial macrophages, T lymphocytes, and PMNs during treatment with anti-tumor-necrosis-factor (TNF)-α in SpA \[[@B17],[@B18]\], the objective of the present study was to analyze in more detail whether histopathological features of the synovial membrane reflect specific phenotypes and/or global disease activity in SpA.
Materials and methods
=====================
Patients and samples
--------------------
The study included 99 SpA patients fulfilling the criteria of the European Spondyloarthropathy Study Group \[[@B19]\]. One cohort consisted of 82 patients, including 19 with ankylosing spondylitis (AS), 33 with PsA, 24 with undifferentiated SpA (USpA), 4 with SpA associated with inflammatory bowel disease, and 2 with reactive arthritis. Since we had previously found no major differences between these SpA subgroups for the synovial histopathology markers used in the present study, we considered them collectively as having SpA (unpublished data). All patients had active disease at the time of inclusion, as evidenced by a mean swollen joint count (SJC) of 3.5 ± 4.1 (mean ± standard deviation), a mean serum C-reactive protein (CRP) concentration of 33 ± 45 mg/L, and a mean erythrocyte sedimentation rate (ESR) of 28 ± 24 mm/hour. The mean duration of disease was 5.5 ± 5.4 years, and 23 of the 82 patients were being treated with a disease-modifying antirheumatic drug (DMARD); none of the patients were being treated with corticosteroids. All patients had at least one swollen knee joint, from which synovial biopsies were sampled by needle arthroscopy.
As an independent validation group, a second cohort of 17 SpA patients (4 with AS, 5 with PsA, 8 with USpA) fulfilling the same inclusion criteria was included in the study. This group had a mean SJC of 5.5 ± 5.4, a mean serum CRP of 38 ± 48 mg/L, and a mean ESR of 31 ± 27 mm/hour. None of these patients was receiving DMARDs or corticosteroids. The mean duration of disease in this group was 10.8 ± 10.2 years.
For the control group, we included 86 patients fulfilling the American College of Rheumatology criteria for RA \[[@B20]\]. As for the SpA cohort, all RA patients had at least one swollen knee joint and had active disease, with a mean SJC of 9.2 ± 6.6, a mean serum CRP of 58 ± 67 mg/L, and a mean ESR of 41 ± 27 mm/hour. The mean duration of disease was 6.0 ± 7.5 years. Fourteen patients were receiving DMARDs, 5 patients were receiving corticosteroids, and 21 patients were receiving both.
In all the patients, synovial tissue biopsies (16 from each person) were obtained by needle arthroscopy of a clinically involved knee joint, as described previously \[[@B21]\]. All patients gave their written, informed consent before inclusion in the study, which was approved by the Ethics Committee of the Ghent University Hospital.
Synovial histopathology
-----------------------
Synovial biopsies were fixed, stained, and scored as described previously \[[@B1]-[@B4]\]. Briefly, eight paraffin-embedded biopsies were stained with H&E for histological analysis, including mean thickness of the synovial lining-layer, vascularity of the sublining layer, infiltration of the sublining layer, and the presence of lymphoid aggregates, plasma cells, and PMNs. A separate analysis in 93 samples showed that the evaluation of the number of blood vessels and the number of plasma cells on H&E staining correlated well with staining for, respectively, the endothelial marker CD146 (*r*= 0.436; *P*\< 0.0001) and the plasma cell marker CD138 (*r*= 0.621; *P*\< 0.0001). The remaining eight biopsies were embedded in tissue-freezing medium and used for immunohistochemistry with the following antibodies: anti-CD1a (interdigitating dendritic cells, mouse, monoclonal; Dako, Glostrup, Denmark), anti-CD3 (T cells, mouse, monoclonal; Dako), anti-CD20 (B cells, mouse, monoclonal; Dako), anti-CD68 (monocytes and macrophages, mouse, monoclonal, clone EBM11; Dako), anti-CD163 (scavenger receptor expressed on mature tissue macrophages, mouse, monoclonal, clone Ber-MAC3; Dako), anti-L-citrulline (citrullinated peptides, rabbit, polyclonal; Biogenesis, Poole, UK), and mAb 12A (detecting MHC class II--HC gp39 peptide complexes, mouse, monoclonal; NV Organon, Oss, Netherlands). Parallel sections were stained with irrelevant origin-, isotype-, and concentration-matched antibody as negative control. Stained sections were coded and analyzed by two independent observers, who were blinded to the diagnosis and clinical data. Due to the low number of positive cells in each sample for anticitrulline, anti-CD1a, and mAb 12A staining, these parameters as well as lymphoid aggregates were scored as present or absent. For all other parameters, the analysis included all areas of the biopsies, and a global score was given for each parameter, using a semiquantitative 4-point scale: 0 represented the lowest and 3 the highest level of expression \[[@B1]-[@B4]\]. As some histological markers are more abundant than others, the scoring system was calibrated for each marker separately by examining a representative number of samples. In case of discordant scores between the two observers, the mean of the two scores was used. Since anti-CD68 and anti-CD163 staining, which recognizes a particular subset of the CD68^+^macrophages \[[@B2]\], was observed in both the synovial lining layer and the synovial sublining layer, these markers were scored separately in the two compartments. An overview of the 15 synovial parameters is given in Table [1](#T1){ref-type="table"}.
Statistics
----------
The histopathological features of the synovial membrane in SpA and RA were compared using the Mann--Whitney *U*test for semiquantitative parameters and the χ^2^test for dichotomous parameters. Histopathological subgroups were identified by clustering analysis (within-group average linkage with Pearson correlation) using SPSS version 12.0 software (SSPS Inc, Chicago, IL, USA). Correlations between semiquantitative histological parameters and clinical disease activity markers (SJC, CRP, ESR) were calculated using the Spearman ρ test. For dichotomous histological markers, the clincal disease activity parameters of the positive and negative groups were compared using an unpaired Student\'s *t*-test. A *P*value of less than 0.05 was considered statistically significant.
The relation between the 15 histological parameters and the 3 clinical parameters was also analyzed by SAM (significance analysis of microarray) software, a statistical analysis model that was specifically developed for multiparameter datasets \[[@B22]\] (see also <http://www-stat.stanford.edu/~tibs/SAM/index.html>). Measuring the relation between changes in the input parameter (which are here the 15 histological features) and changes in the response variable (SJC, CRP, and ESR), the software assigns a score (expressed as a value d) reflecting the strength of the observed differences. To assess the significance of this relationship, a value q is calculated by permutations of the measurements to estimate the percentage of parameters identified by chance, the false discovery rate (FDR). Using SJC, CRP, or ESR as quantitative response parameters, d\>2 (indicating that the strength of the association between the histological input parameter and the disease activity outcome parameter was at least twice the expected value) and q\<0.10 (corresponding to an α error of less than 10% in classical statistics or, in other words, indicating that the observed associations had a 90% chance of being real) were considered significant.
Finally, histological parameters that not only are correlated with disease activity but also contribute significantly to the prediction of the disease activity in individual samples were identified by PAM (predictive analysis of microarray) software \[[@B23]\]. Whereas SAM is intended to identify significant differences between groups of samples, PAM is intended to identify those parameters that are most useful in predicting the outcome (in this case, disease activity) in individual samples and to combine those parameters in an optimal multiparameter algorithm to classify single samples or patients.
Results
=======
Comparative histopathology of SpA and RA
----------------------------------------
The scores for the histopathological features of the synovial membrane in SpA and RA are given in Table [2](#T2){ref-type="table"}. There was significantly higher vascularity (*P*= 0.013), lining CD163 (*P*\< 0.001), and sublining CD163 (*P*= 0.003) in SpA than in RA. There was also a trend towards an increase of PMNs in SpA (*P*= 0.062). In contrast, there was a significantly lower score in SpA versus RA for lining-layer thickness (*P*= 0.032), CD1a (*P*= 0.009), lymphoid aggregates (*P*= 0.029), anticitrulline staining (*P*\< 0.001), and mAb 12A staining (*P*\< 0.001). In contrast with CD163, no differences were found for the pan-macrophage marker CD68 in the lining or sublining layer. The number of plasma cells, which were found in 32 of the 82 SpA samples and 40 of the 86 RA samples, was also not different between the two diseases. Although the mean age of the patients was higher in the RA cohort (56.2 ± 14.9 years) than in the SpA cohort (42.6 ± 13.3 years), none of the differentiating parameters was related to age, excluding the possibility that the difference in age could have induced a systematic bias in the comparison. These findings, which are illustrated in Fig. [1](#F1){ref-type="fig"}, are in agreement with previous observations \[[@B1]-[@B4]\] and indicate that the patient cohorts used in the present study are representative of the full-blown SpA or RA synovial histopathology.
Histopathological heterogeneity within SpA
------------------------------------------
With the exception of anticitrulline and mAb 12A staining, which were found almost exclusively in RA, all investigated histopathological parameters showed a wide range of scores within the SpA group, reflecting wide interindividual variability. Therefore, we next tried to identify specific SpA subgroups by combining the different histological features in a multiparameter model using clustering analysis. Unsupervised analysis yielded two main clusters within SpA, consisting of 39 and 43 samples (Fig. [2](#F2){ref-type="fig"}). Although there were slightly more AS samples in cluster 2, the different SpA subtypes were found both in cluster 1 (4 AS, 14 USpA, 16 PsA, 4 inflammatory bowel disease, 1 reactive arthritis) and in cluster 2 (15 AS, 10 USpA, 17 PsA, 1 reactive arthritis), confirming that synovial histopathology is not basically different between SpA subtypes. The mean duration of disease (5.4 ± 5.0 versus 5.7 ± 5.9 years, respectively), the use of DMARDs (in 10 of 39 versus 13 of 43), and the mean SJC (3.1 ± 3.3 versus 3.8 ± 4.7, respectively) were not different between cluster 1 and cluster 2. In contrast, both serum CRP concentrations (14 ± 12 mg/L versus 51 ± 56 mg/L; *P*\< 0.001) and ESR (19 ± 17 mm/hour versus 35 ± 28 mm/hour; *P*= 0.003) were significantly lower in cluster 1 than in cluster 2 (Fig. [3](#F3){ref-type="fig"}), indicating that this unsupervised classification based on the synovial histopathology reflects the global disease activity. In contrast, a similar analysis of the RA cohort yielded five separate clusters, without significant differences in SJC, serum CRP concentrations, or ESR between the clusters (data not shown).
When the two clusters of the SpA cohort were compared, the cluster with higher CRP and ESR was found to show a significant increase of the following histological parameters: vascularity (*P*= 0.001), inflammatory infiltration (*P*\< 0.001), lymphoid aggregates (*P*= 0.027), plasma cells (*P*= 0.001), PMNs (*P*\< 0.001), CD3^+^lymphocytes (*P*\< 0.001), and CD20^+^lymphocytes (*P*= 0.007).
Relation between synovial histopathology and disease activity in SpA
--------------------------------------------------------------------
Since these data suggest that synovial histopathology reflects global disease activity in SpA, we further analyzed the correlation of individual histological parameters with SJC, CRP, and ESR in the SpA cohort. In order to minimize the risk of false-positive results, only the parameters identified by both classical statistics and SAM analysis were considered as significant. As shown in more detail in Table [3](#T3){ref-type="table"} (top), lining-layer thickness, sublining CD68, and sublining CD163 were weakly but significantly correlated with the SJC in SpA. CRP concentrations correlated significantly with inflammatory infiltration, PMNs, and sublining CD163. Finally, ESR correlated with lining-layer thickness, inflammatory infiltration, PMNs, and sublining CD163. Thus, it appears that global disease activity in SpA is essentially associated with inflammatory infiltration with macrophages -- especially the CD163^+^subset -- and PMNs as well as with lining-layer hyperplasia. Although some of the correlations were relatively weak, the number of CD163^+^macrophages in the sublining appeared to be consistently correlated with the three different measures of disease activity, whereas inflammatory infiltration and PMNs showed a stronger correlation with the systemic inflammatory parameters.
For comparison, we performed the same analysis in the RA control group. As shown in more detail in Table [3](#T3){ref-type="table"} (bottom), no histological parameters were significantly correlated with the SJC. Serum CRP concentrations were significantly associated with CD3 and mAb 12A staining. ESR was correlated with CD3 as well as with sublining CD68 and anticitrulline staining. Globally, the correlations were weaker and less consistent in RA than in SpA.
Supervised clustering in relation with disease activity in SpA
--------------------------------------------------------------
On the basis of the previous findings, we redefined two separate clusters within the SpA cohort based not on all synovial features, but on the five synovial characteristics that were significantly associated with disease activity: lining-layer hyperplasia, inflammatory infiltration, PMNs, sublining CD68, and sublining CD163. Cluster 1 (*n*= 62) was characterized by significantly lower SJC (3.0 ± 3.2 versus 5.2 ± 5.8; *P*= 0.037), serum CRP concentrations (23 ± 32 mg/L versus 66 ± 63 mg/L; *P*\< 0.001), and ESR (21 ± 19 mm/hour versus 48 ± 28 mm/hour; *P*\< 0.001) than cluster 2 (*n*= 20) (Fig. [4a](#F4){ref-type="fig"}). Again, there were no significant differences between the two clusters with regard to the SpA subtypes (12 AS, 21 USpA, 24 PsA, 3 inflammatory bowel disease, and 2 reactive arthritis in cluster 1 versus 7 AS, 3 USpA, 9 PsA, and 1 inflammatory bowel disease in cluster 2), indicating the absence of a relation between histopathology and disease phenotypes. Moreover, there was no difference in DMARD treatment (5 of 20 versus 18 of 62) or disease duration (4.0 ± 5.3 versus 5.8 ± 5.5 years) between the two clusters. In other words, the five previously defined histological parameters were able to identify a subgroup of SpA patients with high disease activity independently of the SpA subtype, treatment, and disease duration, thereby confirming that synovial histopathology reflects not only local inflammation but also global disease activity in SpA.
In contrast, a similar analysis of the RA cohort using CD3, sublining CD68, anticitrulline, and mAb 12A as input parameters yielded two clusters (with respectively *n*= 41 and *n*= 45 samples) that were not different with regard to SJC (9.1 ± 6.3 versus 9.2 ± 7.0), serum CRP concentrations (52 ± 43 mg/L versus 64 ± 82 mg/L), or ESR (42 ± 27 mm/hour versus 40 ± 28 mm/hour).
Confirmation of the supervised clustering analysis on an independent SpA cohort
-------------------------------------------------------------------------------
To confirm these findings, we applied the same supervised clustering analysis based on the five previously defined histological parameters to an independent cohort of 17 SpA patients, none of whom were being treated with DMARDs. Two clusters were identified: cluster 1, consisting of 7 patients (1 with AS, 2 with PsA, 4 with USpA) and cluster 2, consisting of 10 patients (3 with AS, 3 with PsA, 4 with USpA). The mean disease duration in the two clusters was similar (10.7 ± 6.9 versus 10.8 ± 12.4 years, respectively). However, the two clusters were again significantly different with regard to SJC (9.3 ± 6.6 versus 2.9 ± 2.3; *P*= 0.012), serum CRP concentrations (75 ± 55 versus 13 ± 18 mg/L; *P*= 0.004), and ESR (55 ± 18 versus 15 ± 20 mm/hour; *P*= 0.001) (Fig. [4b](#F4){ref-type="fig"}). Thus, this analysis in an independent validation cohort confirms that well-defined synovial histopathological features in SpA reflect global disease activity independently of SpA subtype, disease duration, and treatment.
Prediction of global disease activity in individual SpA samples
---------------------------------------------------------------
Since the previous data provided evidence that lining-layer hyperplasia and infiltration with PMNs and macrophage subsets are directly related to the global disease activity in SpA, we next investigated whether synovial histopathology could be a valuable surrogate marker for the prediction of disease activity in individual patients rather than in a patient cohort. Using PAM to classify the 82 SpA patients into tertiles (low, middle, and high disease activity) for respectively SJC, CRP, and ESR, the same histological parameters were identified as having the largest contribution to the predictive algorithms: inflammatory infiltration, sublining CD163, PMNs, sublining CD68, and lining CD68. However, the positive predictive value of these models was relatively poor, as shown by the correct classification of only 51%, 55%, and 49% of the samples for SJC, CRP, and ESR, respectively, compared with an a priori chance of 33%. Similarly, PAM analysis using histopathological parameters was not able to make a good prediction of samples belonging to the highest quartile for disease activity (data not shown). These data indicate that although the previously identified histological parameters are related to disease activity, the wide interindividual variability does not allow a robust prediction in single patients.
Discussion
==========
It has previously been shown that the synovial histopathology in inflammatory arthritis is dependent on both the disease background and the local disease activity \[[@B1],[@B2],[@B12],[@B15]\]. When focusing on SpA patients with active synovitis of the investigated joint, however, we still observe a large heterogeneity in synovial features. In an attempt to translate these histological findings into clinically relevant patterns, we assessed whether this heterogeneity was related to the fact that SpA consists of different subtypes with distinct phenotypes. Confirming a recent report in which we found no significant differences between PsA on the one hand and AS and USpA on the other hand (unpublished data), both unsupervised and supervised clustering analysis of the present data indicated that the synovial heterogeneity is not basically associated with specific SpA phenotypes.
In contrast, the present study reveals that this heterogeneity directly reflects the global disease activity in SpA. Considering that there are no validated global disease parameters for SpA as a whole, the present study used SJC, serum CRP concentrations, and ESR as clinical outcomes. Although these parameters are not elevated in all patients with active SpA, several studies have shown that peripheral arthritis as well as systemic inflammatory parameters are characteristics of severe disease \[[@B24],[@B25]\]. In this context, the present finding that local synovial features are correlated with systemic inflammatory parameters such as CRP and ESR strengthens the concept that peripheral synovitis, although not present in all SpA patients, contributes significantly to disease severity.
Further analysis revealed that not all synovial features were correlated with disease activity. Both classical statistics and SAM analysis demonstrated a correlation with lining-layer hyperplasia and, more consistently, inflammatory infiltration by CD163^+^macrophages and PMNs. We previously shown that these CD163^+^macrophages, but not the overall number of CD68^+^macrophages, are increased in SpA synovitis and play a specific role in the disease pathogenesis \[[@B2],[@B26]\]. In contrast, neither synovial hypervascularity, which is clearly increased in SpA versus RA and contributes to diagnostic classification \[[@B1],[@B7]\], nor lymphocyte-related characteristics (CD3, CD20, plasma cells, lymphoid aggregates) were associated with SJC, CRP, or ESR. This was further confirmed by a supervised clustering analysis that could even better identify a high-disease-activity group on the basis of only five synovial features, both in the first SpA cohort that was used to identify these features and, most importantly, in a completely independent SpA cohort. Both the previous reports and the present study pointed towards the increased presence of CD163^+^macrophages and PMNs in SpA versus RA synovitis \[[@B2],[@B27]\]. Moreover, in the RA control group the global disease activity parameters correlated not with these features, but with lymphocyte-related characteristics such as CD3 and putative B- and T-cell autoantigens in RA (intracellular citrullinated proteins and MHC--HC gp39 complexes) \[[@B3],[@B4]\]. Although certainly not excluding a secondary effector function for lymphocytes in SpA or for macrophages in RA synovitis, these findings point towards distinct pathogenetic mechanisms in the two diseases and fit well with the hypothesis that SpA synovitis is primarly driven by innate immune cells with secondary alterations in lymphocyte activation and functions \[[@B28]\].
Independently of these pathogenetic considerations, the present data also raise the question of the value of synovial histopathology as a biomarker in SpA. Despite the fact that lining-layer hyperplasia and inflammatory infiltration with CD163^+^macrophages and PMNs reflected the global disease activity when cohorts of patients were analyzed, multiparameter models based on synovial histopathology turned out to be relatively poor predictors of disease severity in individual patients. This finding is not totally unexpected in view of the wide variability of individual values for both histology and disease activity and the broad overlap between different clusters in the previous analyses.
Several factors could play a role in this wide individual variability. Firstly, treatment with DMARDs might be a confounding factor. However, the facts that most SpA patients of the first cohort had been given no treatment at all or were being treated exclusively with nonsteroidal antirheumatic drugs, that the clustering analysis was not influenced by treatment, and that the clustering was confirmed in an independent cohort without DMARD treatment are in accord with previous data showing that DMARDs did not bias the synovial histopathology in patients with persistent, refractory peripheral synovitis (unpublished data). Moreover, even after exclusion of the DMARD-treated patients, the prediction models performed poorly (data not shown). Secondly, the present study used a semiquantitatve scoring system for the histopathology, whereas the previously mentioned RA studies used digital image analysis \[[@B16]\]. Whereas semiquantitative scoring has been shown in multiple studies to be robust and reproducible, it is less sensitive to change than digital image analysis and might thus underestimate small variations \[[@B29]\]. Thirdly, recent data obtained with microarrays indicated clearly that setting up profiles using multiple parameters can compensate for the relative lack of precision and the variability of individual parameters \[[@B30]\]. As we have already demonstrated the added value of combining different histopathological features in multiparameter models for diagnostic classification of inflammatory arthritis \[[@B7],[@B8]\], the same might apply to the use of synovial histology as a surrogate marker for global disease activity.
In this context, early--phase, randomized clinical trials in SpA might be of particular interest for the use of synovial histopathology as a biomarker. With the availibility of powerful new treatments such as TNF-α blockers it becomes increasingly important to obtain as much paraclinical and biological information as possible in small patient cohorts early in the clinical development of new drugs. Moreover, in such trials, the emphasis is on groups with uniform treatment schedules rather than on individual patients, and different biological measurements (such as histology, mRNA expression levels, serum protein concentrations) can be combined in multiparameter algorithms, thus overcoming the previously mentioned caveats. In RA, it has recently been demonstrated that the number of sublining CD68^+^macrophages is a sensitive surrogate marker for response to therapy \[[@B16]\], even if this feature was only found to correlate with ESR in RA in the present cross-sectional study and was clearly less robust than the previously discussed SpA parameters. Since the correlations between global disease activity and synovial histopathology of a single joint were consistently stronger in SpA than in RA and since previous studies showed a histopathological response to targeted therapies in SpA and more specifically a decrease of macrophage subsets and PMNs \[[@B17],[@B18],[@B31]-[@B33]\], the data presented here warrant further prospective and longitudinal analysis of synovial histopathology as a surrogate marker in the evaluation of new, targeted therapies for SpA.
Conclusion
==========
The data presented indicate that inflammatory infiltration of the synovium with CD163^+^macrophages and PMNs as well as lining-layer hyperplasia reflect global disease activity in SpA, independently of the SpA subtype. These data support a prominent role for innate immune cells in SpA synovitis and warrant further evaluation of synovial histopathology as a surrogate marker in early-phase therapeutic trials in SpA.
Abbreviations
=============
AS = ankylosing spondylitis; CRP = C-reactive protein; DMARD = disease-modifying antirheumatic drug; ESR = erythrocyte sedimentation rate; H & E = hematoxylin and eosin; IL = interleukin; mAb = monoclonal antibody; MHC = major histocompatibility complex; PAM = predictive analysis of microarray; PMN = polymorphonuclear leukocyte; PsA = psoriatic arthritis; RA = rheumatoid arthritis; SAM = significance analysis of microarray; SJC = swollen joint count; SpA = spondyloarthropathy; TNF = tumor necrosis factor; USpA = undifferentiated spondyloarthropathy.
Competing interests
===================
Annemieke M Boots is employed by Organon NV, Oss, The Netherlands.
Authors\' contributions
=======================
DB, EMV, and FDK designed the study. DB, EK, and LDR sample and analyzed the synovial tissues. HM selected the patients. DB collected and analyzed the data. mAb 12A was provided by AMB. DB, LDR, and AMB prepared the manuscript, and EMV, HM, and FDK reviewed it. All authors read and approved the final manuscript.
Acknowledgements
================
The authors wish to thank Jenny Vermeersch and Virgie Baert for technical assistance. Dominique Baeten is a Senior Clinical Investigator of the Fund for Scientific Research-Flanders (FWO-Vlaanderen). Leen De Rycke is supported by a fund of IWT (Vlaams instituut voor de bevordering van het wetenschappelijk-technologisch onderzoek in de industrie; IWT/SB/11127).
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Distinct synovial features in spondyloarthropathy (SpA) and rheumatoid arthritis (RA). Vascularity, CD163^+^macrophages, and polymorphonuclear leukocytes (PMNs) were significantly increased in SpA, whereas lining-layer hyperplasia, lymphoid aggregates, CD1a^+^dendritic cells, intracellular citrullinated proteins (detected by anticitrulline staining), and MHC--HC gp39 complexes (detected by staining with monoclonal antibody (mAb) 12A) were higher in RA.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Unsupervised clustering analysis using synovial histopathological parameters identified two main clusters within the spondyloarthropathy cohort (*n*= 82). The dendrogram represents the 82 spondylarthropathy cases on the *y*-axis, classified according to their similarity for the histological parameters. The degree of similarity is represented as rescaled distance on the *x*-axis: when two samples are closely similar the distance will be small, whereas a rescaled distance of 25 represents a high degree of histological difference.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Unsupervised clustering analysis using synovial histopathological parameters identified two main clusters within the spondyloarthropathy (SpA) cohort (*n*= 82). Whereas there were no differences for the swollen jount count between the two clusters, cluster 2 was characterized by significantly higher serum C-reactive protein concentrations (mg/L) and erythrocyte sedimentation rate (mm/hour). Data are represented as box--whisker plots, with median, 25th to 75th percentile, and 5th to 95th percentile. Comparisons were performed with the Mann--Whitney *U*test.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Supervised clustering analysis using the synovial histopathological parameters that were significantly associated with disease activity in spondyloarthropathy (SpA) (lining-layer hyperplasia, inflammatory infiltration, polymorphonuclear cells, sublining CD68, and sublining CD163). **(a)**Analysis in the cohort of 82 patients originally used to identify the histopathological parameters identified a cluster (cluster 2, *n*= 20 samples) which was characterized by significantly higher swollen joint count, serum C-reactive protein concentrations (mg/L), and erythrocyte sedimentation rate (mm/hour). **(b)**A similar analysis in an independent cohort of 17 patients confirmed these results by identifying a cluster of 7 patients (cluster 2) that was similarly characterized by significantly higher swollen joint count, serum C-reactive protein concentrations (mg/L), and erythrocyte sedimentation rate (mm/hour). Data are represented as box--whisker plots, with median, 25th to 75th percentile, and 5th to 95th percentile. Comparisons were performed with the Mann--Whitney *U*test.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Histopathological features and scoring systems used to evaluate synovial inflammation
:::
Feature Scoring system
--------------------------------------------------------------- -------------------
*Assessed by histology*
Lining-layer hyperplasia Semiquantitative
Degree of vascularity Semiquantitative
Inflammatory infiltration Semiquantitative
Lymphoid aggregates Present or absent
Plasma cells Semiquantitative
Polymorphonuclear leukocytes Semiquantitative
*Assessed by immunohistochemistry*
CD1a Present or absent
CD3 Semiquantitative
CD20 Semiquantitative
CD68 lining layer Semiquantitative
CD68 sublining layer Semiquantitative
CD163 lining layer Semiquantitative
CD163 sublining layer Semiquantitative
Intracellular citrullinated peptides Present or absent
MHC class II--HC gp39 peptide complex, recognized by mAb 12A Present or absent
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Synovial histopathology in spondyloarthropathy and rheumatoid arthritis
:::
Feature Spondyloarthropathy^a^ Rheumatoid arthritis^a^
------------------------------ ------------------------ -------------------------
Lining-layer thickness 1 (1--3)\* 1.5 (1--3)\*
Vascularity 2 (0--3)\* 1.5 (0--3)\*
Inflammatory infiltration 1.5 (0--3) 2 (0--3)
Plasma cells 0 (0--3) 0 (0--3)
Polymorphonuclear leukocytes 0 (0--3) 0 (0--2.5)
CD3 1.5 (0--3) 2 (0--3)
CD20 1 (0--3) 1.5 (0--3)
CD68 in lining layer 1 (0--3) 1 (0--3)
CD68 in sublining layer 0.5 (0--3) 1 (0--3)
CD163 in lining layer 1.5 (0--3)\* 0.5 (0--3)\*
CD163 in sublining layer 1.5 (0--3)\* 0.5 (0--3)\*
Lymphoid aggregates 17/65\* 32/54\*
CD1a 18/64\* 36/50\*
Anticitrulline staining 0/82\* 27/59\*
mAb 12A staining 1/81\* 28/58\*
^a^Results are expressed as median (range) for semiquantitative scores and present/absent for dichotomous parameters. As indictated in Materials and methods, stained sections were scored by two blinded observers using a semiquantitative scale from 0 (lowest expression) to 3 (highest level of expression). This scale was calibrated for each marker separately and the mean of the two scores was used.
\**P*\< 0.05.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Association of single histopathological features with three measures of disease activity in spondyloarthropathy (top) and in rheumatoid arthritis (bottom)
:::
Histological feature SJC CRP ESR
------------------------------- -------- ------ ------------ ------ ------------ ------
***In spondylarthropathy***
Lining-layer thickness 0.220 2.16 0.242 NS 0.240 2.30
Vascularity NS NS NS NS NS NS
Inflammatory infiltration NS NS 0.472 2.60 0.336 2.34
Plasma cells NS NS NS NS NS NS
PMNs NS NS 0.393 2.15 0.280 2.58
CD3 NS NS 0.236 NS NS NS
CD20 NS NS NS NS NS NS
Lining CD68 NS NS -0.510 NS -0.372 NS
Sublining CD68 0.278 3.14 NS NS NS NS
Lining CD163 NS NS NS NS NS NS
Sublining CD163 0.281 2.14 0.249 2.06 0.299 3.18
Lymphoid aggregates NS 2.23 NS NS NS NS
CD1a NS NS NS NS NS NS
Anticitrulline staining NS NS NS NS NS NS
mAb 12A staining NS NS NS NS NS NS
***In rheumatoid arthritis***
Lining-layer thickness 0.222 NS NS NS NS NS
Vascularity NS NS NS NS -0.235 NS
Inflammatory infiltration NS NS NS NS NS NS
Plasma cells NS NS NS NS 0.212 NS
PMNs -0.255 NS NS NS NS NS
CD3 NS NS 0.297 2.02 0.334 2.49
CD20 NS NS NS NS NS NS
Lining CD68 NS NS -0.227 NS NS NS
Sublining CD68 NS NS NS NS 0.304 2.31
Lining CD163 NS NS NS NS NS NS
Sublining CD163 NS NS NS NS NS NS
Lymphoid aggregates NS NS NS NS NS NS
CD1a NS NS NS NS NS NS
Anticitrulline staining NS NS NS NS *P*= 0.004 2.84
mAb 12A NS NS *P*= 0.035 2.00 NS NS
Only significant associations are shown. ^a^Correlations were calculated either by classicial statistics (results are expressed as correlation coefficient; for dichotomous parameters differences between the positive and negative group are expressed as *P*on Student\'s unpaired *t*-test) or by SAM (significance analysis of microarray) (for which results are expressed as d values). CRP, serum C-reactive protein; ESR, erythrocyte sedimentation rate; NS, not significant; PMNs, polymorphonuclear leukocytes; SJC, swollen joint count.
:::
|
PubMed Central
|
2024-06-05T03:55:54.979232
|
2005-1-21
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065336/",
"journal": "Arthritis Res Ther. 2005 Jan 21; 7(2):R359-R369",
"authors": [
{
"first": "Dominique",
"last": "Baeten"
},
{
"first": "Elli",
"last": "Kruithof"
},
{
"first": "Leen",
"last": "De Rycke"
},
{
"first": "Anemieke M",
"last": "Boots"
},
{
"first": "Herman",
"last": "Mielants"
},
{
"first": "Eric M",
"last": "Veys"
},
{
"first": "Filip",
"last": "De Keyser"
}
]
}
|
PMC1065337
|
Introduction
============
One of the main pathological features of rheumatoid arthritis is marked destruction of cartilage \[[@B1]\]. This destruction starts with reversible proteoglycan depletion, which is followed by irreversible cartilage degradation defined as chondrocyte death and breakdown of collagen type II, eventually leading to matrix erosion. The latter is mainly induced by matrix metalloproteinases (MMPs), which generate specific cleavage sites within matrix molecules \[[@B2],[@B3]\]. MMPs are secreted in an inactive form by IL-1-stimulated chondrocytes, synovial macrophages, and fibroblasts \[[@B4]-[@B6]\]. Activation of MMPs is still poorly understood, but MMP activity is primarily found in experimental immune-complex (IC)-dependent arthritis models.
Immunoglobulin G (IgG)-containing ICs can activate macrophages upon recognition by Fcγ receptors (FcγRs). Three classes of murine FcγR can be distinguished: FcγRI, II, and III. Triggering FcγRI and III activates cellular responses, whereas FcγRII is an inhibitory receptor \[[@B7]\]. Previous studies have showed that activating FcγRI and III are crucial in induction of severe cartilage destruction, since chondrocyte death and MMP-mediated cartilage damage were absent in FcγR-deficient mice after induction of immune-complex-mediated arthritis (ICA) \[[@B8]\]. Furthermore, cartilage damage is aggravated by local overexpression of the proinflammatory T helper (Th)1 cytokine IFNγ \[[@B9]\]. This increase in cartilage destruction was observed only in IC-dependent arthritis models \[[@B9]\]. FcγRI was found to be crucial in the induction of chondrocyte death, whereas both FcγRI and III mediated MMP-mediated expression of VDIPEN \[[@B9]\].
Since the Th1 cytokine IFNγ worsens the arthritic response by up-regulation of the activating FcγRs, overexpression of a Th2 cytokine during arthritis might be protective, because of down-regulation of these receptors. In earlier studies, we found that adenoviral overexpression of IL-4 resulted in reduced MMP-mediated cartilage damage and chondrocyte death during ICA and arthritis induced by collagen type II \[[@B10],[@B11]\]. IL-4 is regarded as a potent anti-inflammatory cytokine by direct inhibition of proinflammatory cytokines such as IFNγ, IL-1, and tumor necrosis factor α \[[@B12]\]. However, IL-4 protein and mRNA are hardly detected in synovial fluid and synovium of rheumatoid arthritis patients \[[@B13]\]. In contrast, IL-13 is expressed in rheumatoid arthritis synovial fluid and synovial fluid macrophages and resembles many functions of IL-4 \[[@B14],[@B15]\]. Systemic overexpression of IL-13 in collagen-type-II-induced arthritis and local overexpression of IL-13 in rat adjuvant-induced arthritis reduced joint inflammation and bone destruction \[[@B16],[@B17]\]. However, the effect of IL-13 on cartilage destruction was not investigated in detail in these studies and remains to be elucidated.
In the present study, we investigated whether IL-13 influences the development of chondrocyte death and MMP-mediated VDIPEN expression in ICA. Subsequently, regulation of FcγR, MMP, and IL-1 expression by IL-13 was studied, as these are important mediators in severe cartilage damage.
The present study demonstrates that overexpression of IL-13 in arthritic knee joints reduces chondrocyte death and MMP-mediated VDIPEN expression despite enhanced joint inflammation. Injection of an adenovirus encoding for IL-13 diminished chondrocyte death, which correlated with down-regulation of FcγRI expression in the synovium. Reduction of MMP-mediated VDIPEN expression was not reflected by MMP mRNA and IL-1 concentrations, as these were increased.
Materials and methods
=====================
Animals
-------
C57Bl/6 male mice (10 to 12 weeks old) were purchased from Elevage-Janvier (Le Genest Saint Isle, France). Mice were fed a standard diet and tap water ad libitum. Ethical approval was obtained from the research ethics committee of the Central Animal Facility in Nijmegen.
Local gene transfer of IL-13
----------------------------
The recombinant adenovirus encoding human IL-13 (AxCAhIL-13) was generated as described before \[[@B17]-[@B19]\] and an empty adenoviral construct (AxCANI) was used as control virus. AxCAhIL-13 or AxCANI (1.10^7^plaque-forming units) was injected intra-articularly in naive knee joints. Patellae with adjacent synovium were dissected in a standardized manner \[[@B20]\] and synovial biopsies were taken with a biopsy punch (diameter of 3 mm). Total RNA was extracted in 1 ml TRIzol reagent and used for quantitative PCR as described below. AxCAhIL-13 or AxCANI was injected intra-articularly 1 day before the induction of arthritis.
Induction of immune-complex-mediated arthritis
----------------------------------------------
Rabbit polyclonal antibodies directed against lysozyme were injected intravenously into mice. ICA was then passively induced by injecting 3 μg lysozyme coupled to poly-L-lysine in 6 μl pyrogen-free saline into the knee joints.
Histology of arthritic knee joints
----------------------------------
Total knee joints were dissected at days 3 and 7 after the onset of arthritis. Joints were decalcified, dehydrated, and embedded in paraffin. Tissue sections (7 μm) were stained with hematoxylin and eosin.
Histopathological changes were scored in two ways. Inflammation was graded on a scale from 0 (no inflammation) to 3 (severely inflamed joint) as influx of inflammatory cells in synovium and joint cavity. Chondrocyte death was scored as the amount of empty lacunae expressed as a percentage of the total number of cells within the cartilage layers.
Immunohistochemical detection of macrophages and polymorphonuclear neutrophils (PMNs)
-------------------------------------------------------------------------------------
Macrophages were detected using a specific antibody against F4/80, a murine macrophage membrane antigen \[[@B21]\]. PMNs were visualized using NIMPR14, a specific rat anti-mouse monoclonal antibody \[[@B22]\]. Primary antibodies were detected using rabbit anti-rat IgG and avidin--horseradish peroxidase conjugate. Finally, sections were counterstained with hematoxylin. Macrophage and PMN subsets were quantitatively measured using an image analysis system. The inflammatory cell mass was selected by hand and the amount of positive features present in this area was displayed using a computer imaging system. Three sections of each knee joint were measured and the mean was calculated. We report the amount of positive features per 100,000 μm^2^inflammatory cell mass in the synovium.
Immunohistochemical VDIPEN staining
-----------------------------------
Sections were digested with proteinase-free chondroitinase ABC (0.25 units/ml in 0.1 M Tris/HCl, pH 8.0; Sigma, Zwijndrecht, The Netherlands) to remove the side chains of proteoglycans followed by incubation with affinity-purified rabbit anti-VDIPEN IgG \[[@B23]\]. The primary antibody was detected using biotinylated goat anti-rabbit IgG, and avidin--streptravidin--peroxidase (Elite kit; Vector, Burlingame, CA, USA). Counterstaining was done with orange G (2%). Areas of immunostaining were expressed as a percentage of the total cartilage surface.
Quantitative detection of FcγR and MMP mRNA using RT-PCR
--------------------------------------------------------
Specific mRNA levels for FcγRI, II, and III and MMP-3, -9, -12, -13, and -14 were detected using the ABI/PRISM 7000 Sequence Detection System (ABI/PE; Foster City, CA, USA). Briefly, 1 μg of synovial RNA was used for RT-PCR. mRNA was reverse transcribed to cDNA using oligodT primers. cDNA (1/100) was used in one PCR amplification. PCR was performed in SYBR Green Master Mix using the following amplification protocol: 2 min at 50°C followed by 40 cycles of 15 s at 95°C and 1 min at 60°C with data collection in the last 30 s. Message for murine FcγRI, II, and III and MMP-3, -9, -12, -13, and -14 was amplified using the primers listed in Table [1](#T1){ref-type="table"} (Biolegio, Malden, The Netherlands) at a final concentration of 300 nmol/l. Relative quantification of the PCR signals was performed by comparing the cycle threshold value (Ct) of the FcγR and MMP genes in the different samples after correction of the GAPDH content for each individual sample.
Determination of cytokine and chemokine concentrations
------------------------------------------------------
To determine concentrations of IL-13, IL-1β, KC (a mouse homologue for human growth-related protein), and macrophage inflammatory protein 1α in patella washouts, synovial specimens were isolated in a standard manner \[[@B20]\] and incubated in 200 μl RPMI 1640 medium (GIBCO BRL, Breda, The Netherlands) for 1 hour at room temperature. Cytokine and chemokine concentrations were determined using the BioPlex^®^system from BioRad (Hercules, CA, USA) for the Luminex^®^multi-analyte system and expressed as pg/ml.
Statistical analysis
--------------------
Differences between experimental groups were tested for significance using the Mann--Whitney *U*test. *P*values \<0.05 were considered statistically significant.
Results
=======
Local IL-13 expression in naive knee joints using adenoviral gene transfer
--------------------------------------------------------------------------
The expression of IL-13 was determined in synovial washouts at days 1, 2, 3, and 7 after injection of the AxCAhIL-13 virus. IL-13 reached a concentration of 0.4 ng/ml after 24 hours. Values increased to 2 ng/ml at day 2 and remained high up to 7 days after injection (Fig. [1a](#F1){ref-type="fig"}). IL-13 was not detected after injection of AxCANI.
We next investigated whether injection of the adenoviral IL-13 construct causes joint inflammation by itself. Using histology, we found that IL-13 overexpression in naive knee joints did not recruit inflammatory cells at day 1, 2, 3, or 7 (Fig. [1c](#F1){ref-type="fig"}). Injection of AxCANI resulted in minor cell influx in the synovial lining and joint cavity (Fig. [1b](#F1){ref-type="fig"}), which was not detectable from day 2 onwards.
IL-13 overexpression during ICA enhances joint inflammation and alters the composition of the cell mass
-------------------------------------------------------------------------------------------------------
To investigate whether IL-13 overexpression ameliorated the arthritic response, we injected AxCAhIL-13 1 day before ICA induction. Joint inflammation was studied 3 and 7 days after arthritis onset.
IL-13 overexpression significantly increased the inflammatory cell mass in joint cavity and synovium, by 60% and 30%, respectively, 3 days after arthritis induction (Fig. [2a](#F2){ref-type="fig"}). After 7 days, joint inflammation seemed to normalize in the IL-13 group (Fig. [2b](#F2){ref-type="fig"}).
To further investigate inflammatory cell types attracted by IL-13, PMNs and macrophages were detected using specific NIMPR14 and F4/80 antibodies respectively using immunolocalization.
At day 3, the amount of PMNs and macrophages was not markedly altered by IL-13 (Fig. [3a](#F3){ref-type="fig"} and [3B](#F3){ref-type="fig"}). At day 7, however, the amount of PMNs in the synovial lining was 10 times higher (Fig. [3a](#F3){ref-type="fig"}), whereas the amount of macrophages in the IL-13 group was half that in the mice without IL-13 (Fig. [3b](#F3){ref-type="fig"}).
KC concentration in synovial washouts is augmented by IL-13
-----------------------------------------------------------
A possible mechanism by which IL-13 can increase joint inflammation in the presence of ICs is elevation of chemokine production. To investigate this, synovial washouts were done on days 3 and 7, and the chemokines KC (chemotactic for neutrophils) and macrophage inflammatory protein1α (chemotactic for macrophages) were measured. Local IL-13 overexpression increased KC concentrations 4- and 18-fold, respectively, at days 3 and 7 after arthritis induction, which correlates with the high amount of PMNs (Table [2](#T2){ref-type="table"}). Macrophage inflammatory protein1α concentrations at day 3 were comparable between the control and IL-13 groups. At day 7, macrophage inflammatory protein1α expression was slightly increased by IL-13 (Table [2](#T2){ref-type="table"}).
IL-13 strongly inhibits chondrocyte death during ICA: down-regulation of FcγRI
------------------------------------------------------------------------------
Because IL-13 enhanced the inflammatory response, we next investigated the effect of IL-13 overexpression on cartilage destruction. A characteristic feature of irreversible cartilage damage is chondrocyte death; this was scored as the percentage of empty lacunae relative to the total amount of chondrocytes present in various cartilage layers in the knee joint.
Three days after ICA induction, chondrocyte death, expressed as the mean for six cartilage layers in the knee joint, was very low in the IL-13 group (5%) and significantly less than in the control arthritic knee joints, which showed 25% chondrocyte death (Fig. [4a](#F4){ref-type="fig"}). At day 7, chondrocyte death was even more significantly reduced (65%) in comparison with the control group (Fig. [4a](#F4){ref-type="fig"}).
In a previous study, we found that FcγRI is the dominant receptor mediating chondrocyte death during ICA \[[@B9]\]. We speculated that the decreased chondrocyte death might be caused by down-regulation of FcγRI by IL-13. For that reason, we determined the effect of IL-13 on mRNA levels of all three classes of FcγRs in synovium. Cycle values of FcγRI, II, and III in synovium of arthritic knee joints injected with AxCANI were subtracted from cycle values of FcγRs after AxCAhIL-13 injection. Interestingly, FcγRI mRNA level was decreased by IL-13 at day 3 after ICA induction (ΔCt = 2), and was still slightly down-regulated at day 7 (ΔCt = 0.5). In contrast, FcγRII and FcγRIII were up-regulated by IL-13, at both days 3 and 7 after ICA induction (Fig. [4b](#F4){ref-type="fig"}).
IL-13 increases IL-1 production and MMP mRNA levels in the arthritic knee joint
-------------------------------------------------------------------------------
Cartilage matrix degradation is largely mediated by MMPs. Production of latent MMPs is mainly regulated by IL-1 and this cytokine has been shown to be crucial in the generation of MMP-mediated neoepitopes \[[@B23]\]. The production of IL-1 was determined in synovial washouts of arthritic knee joints at both days 3 and 7. At day 3, IL-1 concentration was between 450 and 500 pg/ml in both the control and the IL-13 group. However, at day 7, the IL-1 concentration was reduced in the control group but remained high in the IL-13 group (control 54 pg/ml vs IL-13 255 pg/ml).
This sustained IL-1 production at day 7 may result in high concentrations of MMPs in synovium. Levels of MMP-3, -9, -12, -13, and -14 mRNA were detected by quantitative PCR. MMP-12 mRNA levels were increased 10-fold and 8-fold by IL-13 at days 3 and 7, respectively, after the onset of ICA. At day 7, mRNA levels of MMP-3, -9, and -13 were also significantly increased in the IL-13 group (Table [3](#T3){ref-type="table"}).
MMP-mediated VDIPEN expression is reduced by IL-13 overexpression
-----------------------------------------------------------------
Increased IL-1 and MMP concentrations may induce enhanced MMP-mediated proteoglycan degradation and this was further investigated by detection of VDIPEN neoepitope expression in the cartilage.
In the control group, 35% of the cartilage surface expressed VDIPEN neoepitopes after 3 days (Fig. [5](#F5){ref-type="fig"}). Injection with AxCAhIL13 reduced VDIPEN expression by 43%, as only 20% VDIPEN expression was found in the IL-13 group. The inhibitory effect of IL-13 was still present at day 7 after arthritis induction, as only 10% VDIPEN expression was found in the IL-13 group compared to 25% in the control group (Fig. [5](#F5){ref-type="fig"}).
Discussion
==========
In the present study, we have shown that local gene transfer of IL-13 reduced severe cartilage destruction defined as chondrocyte death and MMP-mediated aggrecan damage during ICA.
Local IL-13 overexpression during IC-dependent arthritis enhanced joint inflammation. To exclude the possibility that IL-13 itself induces influx of inflammatory cells, as is found when IL-13 is overexpressed in the lung \[[@B24],[@B25]\], AxCAhIL-13 was injected in naive knee joints. We observed that IL-13 overexpression in the knee joint did not recruit inflammatory cells. This observation indicates that overexpression of IL-13 induces elevated joint inflammation in combination with IC triggering. In our IC-dependent arthritis model, we showed that joint inflammation is determined by activating FcγRIII \[[@B26]\]. In the present study, we find that IL-13 increased expression of FcγRIII within the synovium, which is not in line with the study showing that IL-13 decreases FcγRIII expression on human monocytes \[[@B27]\].
However, regulation of FcγR expression on mouse macrophages by IL-13 has not been described. IL-13 has high similarity with IL-4, which can increase FcγRIII expression on murine mast cells \[[@B28]\]. Binding of IC to FcγRIII on macrophage lining cells leads to activation, resulting in elevated influx of inflammatory cells. We further found that overexpression of IL-13 in arthritic knee joints particularly increased the amount of PMNs. This is in line with earlier studies in which it was shown that stimulation of FcγRIII induces release of PMN attracting chemokines as IL-8, resulting in neutrophil accumulation \[[@B29]-[@B31]\].
The proinflammatory action of IL-13 found in the present study seems to be dependent on costimulation with ICs to trigger arthritis onset, since local overexpression of IL-13 during T-cell-mediated rat adjuvant-induced arthritis diminishes joint inflammation \[[@B17]\]. In the latter model, ICs do not play a role. Whether IL-13 decreases or enhances joint inflammation may also be dependent on systemic or local overexpression. Systemic overexpression of IL-13 during collagen-type-II-induced arthritis, in which FcγRIII is also required for arthritis development \[[@B32]\], decreased joint inflammation \[[@B16]\]. An explanation may be that systemic overexpression of IL-13 hampers the development of the immune response by induction of isotype switching to the nonarthritogenic IgG4 and IgE \[[@B33],[@B34]\], thereby ameliorating the arthritic response. Induction of immunity is hardly affected by local overexpression, as was shown when injection of AdIL-4 (adenovirus expressing IL-4) in knee joints during arthritis induced by collagen type II markedly increased the amount of inflammatory cells \[[@B11]\].
Cartilage destruction during ICA is mostly related to joint inflammation. Despite the enhanced influx of inflammatory cells, however, a significant reduction of chondrocyte death was induced by IL-13. Chondrocyte death may be the result of increased production of oxygen radicals, as reactive oxygen species can mediate apoptosis \[[@B35]\]. In a previous study, we showed that there is a prominent role for FcγRI mediating chondrocyte death during ICA. In FcγRI-deficient mice, chondrocyte death was almost absent. When the Th1 cytokine IFNγ was overexpressed, a significant increase in chondrocyte death was observed, which was dependent on FcγRI \[[@B9]\]. Stimulation of FcγRI leads to production of oxygen radicals via NADPH-oxidase \[[@B36]\]. In the present study, we find that in knee joints injected with AxCAhIL-13, FcγRI expression remained low, whereas in knee joints injected with control virus, FcγRI expression level was enhanced in the synovium. The decrease in chondrocyte death might be due to a reduced FcγRI concentration. Moreover, it has been shown that IL-13 itself down-regulates production of oxygen radicals by inflammatory cells, since IL-13 can inhibit protein-kinase-C-triggered respiratory burst in monocytes \[[@B37]\]. The inhibiting effect of IL-13 on oxygen radical production seemed to be monocyte-dependent, as no reduction was found in PMNs \[[@B38]\].
In addition, IL-13 also reduced MMP-mediated VDIPEN neoepitope expression. It has been reported that IL-13 diminishes the breakdown of collagen and proteoglycans from bovine cartilage, by regulation of MMP expression \[[@B39]\]. Several mechanisms may inhibit MMP-mediated cartilage destruction, as regulation of MMPs occurs at three different levels: MMP synthesis, activation of latent enzyme, and MMP inhibition. IL-1 is a prominent cytokine controlling the production of latent MMPs \[[@B40]\], and diminished production of IL-1 might reduce MMP-mediated cartilage damage. We found, however, that IL-13 overexpression in arthritic knee joints strongly increased IL-1β concentrations. IL-13 is described as an anti-inflammatory cytokine, which in general reduces IL-1β production \[[@B14],[@B27],[@B41]\]. However, the effect of IL-13 on IL-1 production by IC-stimulated macrophages has not been described to date. In addition to macrophages, fibroblasts and PMNs are also present in the knee joint at day 7 after the onset of arthritis.
The sustained production of IL-1 by IL-13 may indeed stimulate MMP production, as reflected by enhanced MMP-3, -9, -12, and -13 mRNA levels 7 days after ICA induction in AxCAhIL-13-injected arthritic knee joints. MMP-12 mRNA level was already increased at day 3 after the onset of arthritis. It has been shown that MMP-12 expression is IL-13-dependent and that MMP-12 is a critical downstream mediator and regulator of IL-13-induced responses \[[@B42],[@B43]\]. Furthermore, IL-13 induction of MMP-2, -9, and -13 is at least partly mediated by MMP-12 \[[@B43]\], indicating that MMP-12 may be a crucial enzyme inducing MMP-mediated cartilage damage.
Furthermore, IL-13 might interfere at the level of activation of MMPs. MMPs are secreted in a latent form and activation occurs after cleavage of a propeptide. Factors that activate latent MMPs are still unknown. However, MMP-mediated VDIPEN expression is mainly found in IC-dependent arthritis models, in which FcγRs are of utmost importance. Down-regulation of the activating FcγRs might reduce VDIPEN expression. Indeed, we found that IL-13 strongly diminished FcγRI expression in synovium. Another mechanism involved in activation of MMPs is production of oxygen radicals. As mentioned above, stimulation of FcγRI results in assembly of the NADPH-oxidase complex, which produces oxygen radicals \[[@B36]\]. Additionally, oxygen metabolites can be converted into H~2~O~2~, which can activate latent proMMPs \[[@B44],[@B45]\]. Taken together, decreased FcγRI expression reduces the production of oxygen radicals, which apart from chondrocyte protection may also result in diminished MMP-mediated VDIPEN expression.
Conclusion
==========
The present study shows that IL-13 is a potent cytokine that protects the cartilage matrix against degradation during ICA. In addition, these results indicate that regulation of the expression of FcγR, particularly FcγRI, might be involved in this process. Therefore, modulation of FcγRI by Th2 cytokines seems to be a promising therapeutic tool diminishing cartilage damage in rheumatoid arthritis.
Abbreviations
=============
AxCAhIL-13 = adenovirus encoding interleukin-13; AxCANI = adenovirus encoding no gene; Ct = cycle threshold; FcγR = Fcγ receptor; IC = immune complex; ICA = immune-complex-mediated arthritis; IFNγ = interferon γ; IgG = immunoglobulin G; IL = interleukin; KC = mouse homologue for human IL-8; MMP = matrix metalloproteinase; NADPH = reduced nicotinamide adenine dinucleotide phosphate; PMN = polymorphonuclear neutrophil; RT-PCR = reverse transcriptase polymerase chain reaction; Th, T helper.
Competing interests
===================
The author(s) declare that they have no competing interests.
Authors\' contributions
=======================
KN designed the experimental design of the study, carried out the experiments, and drafted the manuscript. PL participated in the experimental design of the study and preparation of the manuscript. AH participated in the animal studies. AS participated in isolation of mRNA and performing PCRs. AK provided the adenoviruses and participated in the preparation of the manuscript. TR participated in the preparation of the manuscript. WB participated in the design of the study and preparation of the manuscript. All authors read and approved the final manuscript.
Acknowledgements
================
Supported by grants from the Dutch Arthritis Association (99-1-402); US National Institutes of Health (NIH) grants AR48267, AI40987, and HL58695; funds from the Veterans Administration, USA; and The William D Robinson and Frederick Huetwell endowed professorship.
Figures and Tables
==================
::: {#F1 .fig}
Figure 1
::: {.caption}
######
Adenoviral-vector-mediated IL-13 expression in knee joints of C57Bl/6 mice. **(a)**Naive knee joints and **(b)**total knee joint sections 24 hours after injection of AxCANI (adenovirus encoding no gene) or of **(c)**AxCAhIL-13 (adenovirus encoding interleukin-13). Injection of AxCAhIL-13 resulted in 0.4 ng/ml IL-13 at day 1, which increased to 5.5 ng/ml by day 7 (a). Injection of AxCANI resulted in a mild thickening of the synovial lining (S) and some invading inflammatory cells in the joint cavity (JC) (b), whereas no inflammation was observed after AxCAhIL-13 injection (c). Plotted values are means ± SEM of data from 5 mice. \**P*\< 0.05. Original magnification 200×. F, femur; P, patella.
:::
:::
::: {#F2 .fig}
Figure 2
::: {.caption}
######
Joint inflammation in arthritic knee joints of C57Bl/6 mice injected with AxCANI (adenovirus encoding no gene) or AxCAhIL-13 (adenovirus encoding interleukin-13). At **(a)**day 3 and **(b)**day 7 after the onset of immune-complex-mediated arthritis. The inflammatory cell mass was significantly enhanced by IL-13 in both the joint cavity and the synovium 3 days after arthritis induction. Bars show the means ± SEM for 10 mice. Significance was evaluated using the Mann--Whitney *U*test. \**P*\< 0.05.
:::
:::
::: {#F3 .fig}
Figure 3
::: {.caption}
######
Immunohistochemical detection of inflammatory cells in knee joints of mice with immune-complex-mediated arthritis (ICA). **(a)**Polymorphonuclear neutrophils and **(b)**macrophages in synovium 3 and 7 days after injection of AxCANI (adenovirus encoding no gene) or AxCAhIL-13 (adenovirus encoding interleukin-13). Polymorphonuclear neutrophils were detected using the specific rat anti-mouse monoclonal antibody NIMPR14, and macrophages were detected using an antibody against the membrane marker F4/80. At day 7, the amount of NIMPR14-positive features was significantly higher in the synovium of AxCAhIL-13-injected arthritic knee joints, while the amount of F4/80-positive features was significantly lower. The bars represent means ± SEM for 10 mice. Data were evaluated using the Mann--Whitney *U*test. \**P*\< 0.05.
:::
:::
::: {#F4 .fig}
Figure 4
::: {.caption}
######
Chondrocyte death in the knee joints of mice with immune-complex-mediated arthritis (ICA). **(a)**At day 3 and 7 in arthritic knee joints injected with injected with AxCANI (adenovirus encoding no gene) or AxCAhIL-13 (adenovirus encoding interleukin-13) and **(b)**expression profiles of Fcγ receptor I (FcγRI), II, and III mRNA levels induced by IL-13 in synovium. IL-13 significantly decreased chondrocyte death, both at day 3 and at day 7 (a). Cycle threshold (Ct) values of FcγRI, II, and III in arthritic knee joints injected with AxCANI were subtracted from the Ct values for FcγRs after injection of AxCAhIL-13. Ct values were corrected for glyceraldehyde-3-phosphate dehydrogenase content for each individual sample. (b) FcγRI mRNA level was down-regulated by IL-13, whereas an up-regulation was observed for both FcγRII and III. Bars represent means ± SEM for 10 mice. Mann--Whitney *U*test. \**P*\< 0.05. D, Δ.
:::
:::
::: {#F5 .fig}
Figure 5
::: {.caption}
######
Matrix-metalloproteinase-mediated aggrecan damage in knee joints of mice with immune-complex-mediated arthritis. VDIPEN expression at day 3 and 7 after the induction of immune-complex-mediated arthritis in knee joints injected with AxCANI or AxCAhIL-13. Note that VDIPEN expression was reduced by IL-13 both at day 3 and day 7. Values represent the mean ± SEM for 10 mice. \**P*\< 0.05, Mann--Whitney *U*test. AxCAhIL-13 = adenovirus encoding interleukin-13; AxCANI = adenovirus encoding no gene.
:::
:::
::: {#T1 .table-wrap}
Table 1
::: {.caption}
######
Primers for detection of murine FcγRI, II, and III mRNA
:::
Gene Primer Primer sequence
--------- -------- --------------------------------------------
GAPDH Up 5\'-GGC-AAA-TTC-AAC-GGC-ACA-3\'
Low 5\'-GTT-AGT-GGG-GTC-TCG-CTC-CTG-3\'
FcγRI Up 5\'-ACA-CAA-TGG-TTT-ATC-AAC-GGA-ACA-3\'
Low 5\'-TGG-CCT-CTG-GGA-TGC-TAT-AAC-T-3\'
FcγRII Up 5\'-GAC-AGC-CGT-GCT-AAA-TCT-TGC-T-3\'
Low 5\'-GTG-TCA-CCG-TGT-CTT-CCT-TGA-G-3\'
FcγRIII Up 5\'-GAC-AGG-CAG-AGT-GCA-GCT-CTT-3\'
Low 5\'-TGT-CTT-CCT-TGA-GCA-CCT-GGA-T-3\'
MMP-3 Up 5\'-TGG-AGC-TGA-TGC-ATA-AGC-CC-3\'
Low 5\'-TGA-AGC-CAC-CAA-CAT-CAG-GA-3\'
MMP-9 Up 5\'-GGA-ACT-CAC-ACG-ACA-TCT-TCC-A-3\'
Low 5\'-GAA-ACT-CAC-ACG-CCA-GAA-GAA-TTT-3\'
MMP-12 Up 5\'-GGA-CAT-GAA-GCG-TGA-GGA-TGT-3\'
Low 5\'-GAA-GTC-TCC-GTG-AGC-TCC-AAA-T-3\'
MMP-13 Up 5\'-ACC-TTG-TGT-TTG-CAG-AGC-ACT-AAC-TT-3\'
Low 5\'-CTT-CAG-GAT-TCC-CGC-AAG-AGT-3\'
MMP-14 Up 5\'-AAG-GCT-GAT-TTG-GCA-ACC-AT-3\'
Low 5\'-GTC-CCA-AAC-TTA-TCC-GGA-ACA-C-3\'
Primer sequences used for RT-PCR on synovium. FcγR, Fcγ receptor; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; MMP, matrix metalloproteinase.
:::
::: {#T2 .table-wrap}
Table 2
::: {.caption}
######
Effect of IL-13 on chemokine concentrations (pg/ml) in arthritic joints in mice
:::
Group KC MIP1α
------------- ------------ -----------
ICA, day 3
AxCANI 56 ± 8 303 ± 6.8
AxCAhIL-13 196 ± 31\* 344 ± 96
ICA, day 7
AxCANI 10 ± 6 157 ± 25
AxCAhIL-13 184 ± 26\* 268 ± 98
Concentrations of KC and MIP1α were detected in synovial washouts of arthritic knee joints 3 and 7 days after arthritis induction. KC concentrations were significantly higher in arthritic knee joints injected with AxCAhIL-13 both at day 3 and 7. \**P*\< 0.05 in comparison with AxCANI. AxCAhIL-13, adenovirus encoding interleukin-13; AxCANI, adenovirus encoding no gene; ICA, immune-complex-mediated arthritis; KC, mouse homologue for human IL-8; MIP 1α, macrophage inflammatory protein 1α.
:::
::: {#T3 .table-wrap}
Table 3
::: {.caption}
######
Effect of IL-13 on MMP mRNA levels in synovium of mice with ICA
:::
ICA day 3 ICA day 7
-------- ----------- ------------- ----------- -------------
AxCANI AxCAhIL-13 AxCANI AxCAhIL-13
MMP-3 5.7 ± 0.3 7.1 ± 0.8 4.2 ± 0.5 6.1 ± 0.2\*
MMP-9 5.1 ± 0.2 4.8 ± 0.3 0.2 ± 0.6 3.9 ± 0.5\*
MMP-12 0.6 ± 0.4 5.8 ± 1.1\* 0.9 ± 1 8.1 ± 0.9\*
MMP-13 3.2 ± 0.2 2.7 ± 0.3 4.3 ± 0.3 6.4 ± 0.3\*
MMP-14 3.7 ± 0.4 4.9 ± 0.8 3.7 ± 1 3.7 ± 0.6
Expression profile of MMP-3, -9, -12, -13, and -14 mRNA levels after injection of AxCANI or AxCAhIL-13 in synovial biopsies isolated at day 3 and day 7 after arthritis onset. The Ct values for MMP genes in naive knee joints were subtracted from the Ct values for MMPs at day 3 and 7 after arthritis onset. Ct values were corrected for GAPDH content for each individual sample. Note that MMP-3, -9, -12, and -13 mRNA levels were significantly increased at day 7 by IL-13, and the MMP-12 mRNA level was already elevated at day 3. Values represent means ± SEM for 5 mice. \**P*\< 0.05, Mann--Whitney *U*test. AxCAhIL-13 = adenovirus encoding interleukin-13; AxCANI = adenovirus encoding no gene; Ct = cycle threshold; GAPDH = glyceraldehyde-3-phosphate dehydrogenase; ICA = immune-complex-mediated arthritis; MMP = matrix metalloproteinase.
:::
|
PubMed Central
|
2024-06-05T03:55:54.982834
|
2005-1-26
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065337/",
"journal": "Arthritis Res Ther. 2005 Jan 26; 7(2):R392-R401",
"authors": [
{
"first": "Karin CAM",
"last": "Nabbe"
},
{
"first": "Peter LEM",
"last": "van Lent"
},
{
"first": "Astrid EM",
"last": "Holthuysen"
},
{
"first": "Annet W",
"last": "Sloëtjes"
},
{
"first": "Alisa E",
"last": "Koch"
},
{
"first": "Timothy RDJ",
"last": "Radstake"
},
{
"first": "Wim B",
"last": "van den Berg"
}
]
}
|
PMC1065704
|
As a species, we pride ourselves on the uniqueness of our brain. Relative to our body size, the human brain is bigger than that of any other animal. It may also contain unique structures and patterns of organisation that presumably underlie our intelligence and ability to manipulate our environment. But how did our unique brain originate, and under what selective pressures did it evolve? Some of the answers may lie in the genetic differences that researchers are now uncovering between us and our closest relatives.
What Is So Different about the Human Brain? {#s2}
===========================================
When we compare our brain to those of other animals, the first thing that strikes us is its size. Human brains weigh on average 1,300 grams; a squirrel brain weighs six grams. Some of this difference is because, as larger animals, we need more brain to run our bodies. However, the brains of our nearest relatives, the great apes, weigh only 300--500 grams, even though their body size is similar to ours ([Figure 1](#pbio-0030050-g001){ref-type="fig"}). "Humans sit on the top of the pile when it comes to relative brain size", notes geneticist Bruce Lahn (University of Chicago, Illinois, United States) (see [Box 1](#box1){ref-type="boxed-text"}).
Box 1. Nothing like a Whale
---------------------------
Just how unique is human brain evolution? Neuroscientist Lori Marino (Emory University, Atlanta, Georgia, United States) and her colleagues have used computed tomography to estimate the body and brain size of 36 fossil whale species and have compared these data with those for modern toothed whales. Relative to body size, whales and dolphins have the next biggest brains to us, bigger even than chimpanzees, and, says Marino, "there have been dolphins swimming in the oceans with huge brains for more than 15 million years. We are really the new kids on the block."
Like in humans and other primates, the neocortex in whale brains is huge, but its structure is very different to that of our neocortex. Whales have been independent of other lineages for about 60 million years, notes Marino, and haven\'t shared a common ancestor with primates for 94 million years. "Nevertheless, during evolution, whales have converged upon very similar capacities and behaviours to those of primates, including a highly developed social structure, which tells us that there is more than one way to evolve a complex intelligence."
::: {#pbio-0030050-g001 .fig}
Figure 1
::: {.caption}
###### Comparison of a Human and a Chimpanzee Brain Scale bar = 1 cm
(Image: Todd Preuss, Yerkes Primate Research Center)
:::
:::
Throughout mammalian and primate evolution, there has been a gradual increase in brain size, superimposed with "spikes" of fast growth such as the tripling in human brain size that occurred about 1.5 million years ago, 4 million years after the human lineage diverged from that of the great apes. "Even in the ape lineage, the brain has been expanding but along the human lineage it has really taken off", says Lahn.
In addition, over time, different parts of our brain have increased in size at different rates. The cerebral cortex has expanded more than other areas, and within the cortex, some areas have expanded differentially while others have lagged behind.
> "Humans sit on the top of the pile when it comes to relative brain size."
Paleoanthropologist Ralph Holloway (Columbia University, New York, United States) uses endocasts to look for macroscopic differences in the brains of our human ancestors. "We fill human fossil skulls with vulcanised rubber and once it has set, we pull it out of the large hole at the base of the skull and the rubber snaps back into the shape of the skull", Holloway explains. Endocasts are particularly useful for comparing brain sizes, but they also provide information on when the asymmetries that are present in our brain first appeared. These often reflect cerebral specialisation, and Holloway believes that some of the asymmetries he sees in human fossil skulls may indicate when our ancestors acquired language.
More details about how the shape of our brain differs from that of our closest living relatives are emerging from the work of neuroscientist Karl Zilles (Institute of Medicine, Research Center Jülich, Germany). He prepares magnetic resonance images of monkey, ape, and human brains and then uses a nonlinear elastic algorithm to transform one brain into another ([Figure 2](#pbio-0030050-g002){ref-type="fig"}). "We know what forces we have to apply to the images to do this", he explains, "which tells us which areas of the brain have changed most during primate evolution". Zilles and his colleagues also are currently using molecular imaging techniques to update the existing maps of the different areas within our brains. Until we have this information, it is hard to make meaningful comparisons between our brain and that of chimpanzees. Already, Zilles has discovered that there is much more interindividual variation in human brain organisation than anyone suspected. This means, says Zilles, "that a general statement like 'the neocortex is bigger in human brains than in ape brains' actually tells us very little. It gives us the general direction that evolution has taken but not whether an ape brain is different because of its sensory, motor, or association areas."
::: {#pbio-0030050-g002 .fig}
Figure 2
::: {.caption}
###### Magnetic Resonance Imaging of Brains
Three-dimensional reconstruction of a reference bonobo (pygmy chimpanzee) brain (A) and a reference human brain (B) after magnetic resonance imaging and normalisation of absolute brain sizes. The virtual bonobo brain has been transformed into the virtual human brain using an elastic deformation algorithm. The local deformation vectors are colour-coded and projected onto the virtual human brain (C). The most dramatic changes in brain shape occur in (1) the ventro-orbital prefrontal cortex, (2) the ventral stream of the visual cortex, and (3) the hypothalamic neuroendocrine region.
(Image: Karl Zilles, Hartmut Mohlberg, and Peter Pieperhoff, Research Center Jülich)
:::
:::
Scientists are also using other techniques to investigate more subtle changes in the organisation of the human brain compared to the brains of other mammals and primates. Indeed, says Holloway, the reorganisation of the brain during evolution has been at least important as its increase in size. Neurobiologist John Allman (California Institute of Technology, Pasadena, California, United States) and his collaborators, for instance, have discovered that a special type of large spindle-shaped neuron, first described in the early 20th century by Constantin von Economo, is unique to apes and humans and much more numerous in the latter. These neurons are found in brain areas that are implicated in decision making in uncertain situations so, Allman speculates, they may help humans to interact rapidly in complex social situations.
Costs and Benefits {#s3}
==================
A bigger, more complex brain may have advantages over a small brain in terms of computing power, but brain expansion has costs. For one thing, a big brain is a metabolic drain on our bodies. Indeed, some people argue that, because the brain is one of the most metabolically expensive tissues in our body, our brains could only have expanded in response to an improved diet. Another cost that goes along with a big brain is the need to reorganise its wiring. "As brain size increases, several problems are created", explains systems neurobiologist Jon Kaas (Vanderbilt University, Nashville, Tennessee, United States). "The most serious is the increased time it takes to get information from one place to another." One solution is to make the axons of the neurons bigger but this increases brain size again and the problem escalates. Another solution is to do things locally: only connect those parts of the brain that have to be connected, and avoid the need for communication between hemispheres by making different sides of the brain do different things. A big brain can also be made more efficient by organising it into more subdivisions, "rather like splitting a company into departments", says Kaas. Overall, he concludes, because a bigger brain *per se* would not work, brain reorganisation and size increase probably occurred in parallel during human brain evolution. The end result is that the human brain is not just a scaled-up version of a mammal brain or even of an ape brain.
For natural selection to work, the costs of brain evolution must be outweighed by the advantages gained in terms of fitness. For many years, explains ecological psychologist Robin Dunbar (University of Liverpool, United Kingdom), "people thought that the ability to hunt or forage better was what drove the evolution of our brains. But a better diet had to come before we could grow a bigger brain." Dunbar believes instead that brain evolution in primates and more generally in mammals "has been driven by the need to manage social relationships, and in primates, in particular, to coordinate coherence in social groups through time and space". More complex social interactions, he says, mean that individuals are better able to pool resources to solve problems like finding food, and so they survive better.
This theory, says Dunbar, is supported by a correlation between social group size and neocortex size across primates and modern humans. Furthermore, during primate brain evolution, the trend has been to add more material to the front than the back of the brain. The front of the brain is where information from the rest of the brain is interpreted, and the capacity to interpret information underlies social interactions, says Dunbar. The number of problem-solving cognitive tasks you can do may well depend on how much frontal lobe volume you have and how it is organised. Just think of how few moves you can run a chess game into the future with a 1980s personal computer compared to a 21st century mainframe machine, he suggests.
> The human brain is not just a scaled-up version of a mammal brain or even of an ape brain.
The Genetics of Human Brain Evolution {#s4}
=====================================
Selective pressures like those considered by Dunbar and, before him, by scientists like Holloway work on the raw material of random gene mutations, and molecular biologists now have some clues to the gene changes that may underlie brain evolution. Take brain size, for example ([Figure 3](#pbio-0030050-g003){ref-type="fig"}). Research groups, including those led by Lahn, neurologist Christopher Walsh (Harvard Medical School, Boston, Massachusetts, United States), and clinical geneticist Geoffrey Woods (University of Leeds, United Kingdom), wondered whether the genes that cause microcephaly, an inherited human disorder in which brain size is greatly reduced, might include genes involved in human brain evolution. In 2002, mutations in the genes *ASPM (abnormal spindle-like microcephaly associated)* and *microcephalin* were identified as two causes of microcephaly. Three groups have since reported that both these genes have been under selective pressure during primate evolution. *ASPM* encodes a protein involved in spindle formation, so it is tempting to think that changes in its sequence might result in an increased rate of cell division and hence brain size. But, cautions Walsh, "we really have no idea yet how or even if *ASPM* is involved in brain evolution".
::: {#pbio-0030050-g003 .fig}
Figure 3
::: {.caption}
###### Primate Brain Sizes
These skulls are from the Harvard Museum of Comparative Zoology.
(Image: Christopher Walsh, Harvard Medical School)
:::
:::
Both Lahn and Walsh believe that *ASPM* and *microcephalin* may be only the tip of the iceberg when it comes to genes that have helped to shape our brains. For example, Walsh has recently reported that deletion of a gene called *Nde1* produces mice with very small brains. "Our experiments indicate that the loss of *Nde1* causes neurons to mature prematurely. This stops them dividing so the mice end up with small brains", explains Walsh, who is now investigating whether human *NDE1* variants have been positively selected during human evolution.
Lahn is also searching for additional candidate genes that might help to explain how our brains evolved. In a recent *Cell* paper, Lahn and his colleagues identify several hundred genes that are involved in nervous system biology and show that, as a group, there are significantly higher rates of protein evolution in these genes in primates than in rodents. Protein evolution rates are particularly high in the lineage leading from ancestral primates to humans, notes Lahn, "so some of these genes may regulate brain size and behaviour". However, he warns, as with *ASPM* and *microcephalin*, "definitive proof for this will only come from functional studies, which are difficult to do".
Enter Glutamate Dehydrogenase {#s5}
=============================
For one gene, evidence for an effect on brain function may be emerging. Geneticist Henrik Kaessmann (University of Lausanne, Switzerland) studies the origin of new genes in primates, in particular genes that arise when a DNA copy of an mRNA transcribed from an existing gene is integrated back into the genome. Usually this new "retrocopy" is not expressed, but if the DNA inserts near an active promoter, it can become a transcribable "retrogene". This is the origin of *GLUD2*, a retrogene derived from *GLUD1*, which encodes glutamate dehydrogenase. *GLUD2*, which first appeared 18--23 million years ago in hominoids, probably immediately picked up a brain-specific promoter and then over the next few million years acquired two critical amino acid changes, explains Kaessmann. These allow *GLUD2*-encoded glutamate dehydrogenase to work better in the brain than the *GLUD1*-encoded enzyme. Because glutamate dehydrogenase recycles the neurotransmitter glutamate, the presence of *GLUD2* may permit a higher neurotransmitter turnover and greater neuronal activity in hominoid brains than is possible in monkey brains, which lack *GLUD2*, suggests Kaessmann.
Gene Expression {#s6}
===============
Kaessmann plans to search his extensive database of retrocopies in the human genome for other functional genes that could, like *GLUD2*, be implicated in brain evolution. By contrast, evolutionary neurobiologist Todd Preuss (Yerkes Primate Research Center, Emory University, Atlanta, Georgia, United States) hopes to identify genes involved in human brain evolution by comparing gene expression patterns in different primates. Preuss, who began training as a paleoanthropologist before turning to neuroscience, has been comparing post-mortem human and chimpanzee brains since the mid 1990s, believing that "if we want to understand human brain evolution, we really have to compare humans with chimpanzees, our nearest relatives", even though chimp brains have been evolving separately from ours for 5--7 million years. But, warns Preuss, "we have to do these studies now. There are few chimps left and if we lose the opportunity to study them and their brains, we will lose forever a fundamental source of insight into our own species."
To begin with, Preuss used staining techniques that exploit antibodies to examine the neural components of chimpanzee and human brains. Then in 1998, he was asked to collaborate in a microarray project. "My antibody approach was very labour intensive so I jumped at the opportunity to screen 10,000 genes at once", he says.
Preuss and his collaborators now know that more than 100 genes are differentially expressed in chimpanzee and human brains. "Importantly, when we go back into tissue with probes for these gene products, in some cases there are remarkably different spatial patterns of expression in humans, chimps, and macaques", notes Preuss. "We don\'t know yet what these differences mean in terms of functional organisation in these different brains but our results open up new and exciting vistas", particularly since many of the differentially expressed genes have not previously been considered as being potentially involved in brain evolution. The microarray data produced by Preuss and other researchers also indicate that many of the gene expression changes that have occurred during brain evolution involve gene upregulation. For example, there is increased expression of genes involved in metabolism, synaptic organisation, and synaptic function. "All told, it seems that the human brain may be more dynamic than ape or monkey brains", says Preuss. "The human brain seems to be running hot in all sorts of ways."
Scratching at the Surface {#s7}
=========================
As far as understanding how our brains evolved, more questions remain than have been answered. One problem is that we don\'t really know enough about how our brains differ from those of other mammals and primates, although work by Zilles and others is helping here. We also know very little about how the areas of our brain are physically linked up, and we need to understand that before we can see how we differ from our nearest relatives. And as far as identifying the gene changes that were selected during evolution, although we have several candidates, we don\'t know how or if these gene variants affect our cognitive abilities. It is one thing, concludes Dunbar, to identify genetic or anatomic differences between human and ape brains, but quite another to know what they mean in terms of actual cognitive processes.
**Citation:** Bradbury J (2005) Molecular insights into human brain evolution. PLoS Biol 3(3): e50.
Jane Bradbury is a freelance science news writer based in Cambridge, United Kingdom. E-mail: <janeb@sciscribe.u-net.com>
|
PubMed Central
|
2024-06-05T03:55:54.985035
|
2005-3-15
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065704/",
"journal": "PLoS Biol. 2005 Mar 15; 3(3):e50",
"authors": [
{
"first": "Jane",
"last": "Bradbury"
}
]
}
|
PMC1065705
|
Introduction {#s1}
============
Recent concerns about bioterrorism and emerging diseases have led to a new focus on the development of vaccines and drugs targeting infectious pathogens. An important component of vaccine development is the characterization of immune responses (to vaccination, for example, or following infection in experimental settings) by evaluating the epitopes recognized by antigen-specific receptors of the immune system (antibodies and/or T cell receptors (TCRs)) \[[@pbio-0030091-b1]\]. In recent years, different groups have followed different approaches to the discovery of immune epitopes, and various assay types have been used to generate data for the purpose of epitope definition or validation. We believe that research in this area could be greatly facilitated by a comprehensive knowledge center: a repository of immune epitope data with associated analysis tools. Our goal is the creation of the Immune Epitope Database and Analysis Resource (IEDB).
The IEDB is sponsored by the National Institute for Allergy and Infectious Diseases (NIAID). It will host data relating to both B cell and T cell epitopes from infectious pathogens, as well as experimental and self-antigens (RTP-NIH-NIAID-DAIT-03/31; [www.niaid.nih.gov/contract/archive](www.niaid.nih.gov/contract/archive)). Priority will be placed on those epitopes considered to be potential bioterrorism threats, and emerging diseases as defined by NIAID (so-called category A--C pathogens; see: <http://www2.niaid.nih.gov/Biodefense/bandc_priority.htm>). As a corollary to the IEDB effort, NIAID has also launched a large-scale antibody and T cell epitope discovery program aimed at generating epitope data and analysis resources to be included in the IEDB. Other data sources to be integrated into the IEDB are publications in peer-reviewed journals, published patents or patent applications, and direct submissions from institutions or companies. Everyone who contributes data or analysis resources to the database will be cited, either by authorship or by other acknowledgment of their contributions.
The involvement of the scientific community in the design of the scope and capability of the IEDB will be crucial to the success of the project. The IEDB will be produced in a manner that encourages the incorporation of data and analytical tools derived by research labs at-large. With this paper, we hope to inform the scientific community of our effort and to solicit feedback while the project is still in a design stage. We envision that this resource center will be freely available on the Internet, with a prototype operational in the fourth quarter of 2005. Once the project is online, forms for direct feedback and online submission of data and tools will be provided. Yearly conferences to present data relating to epitope identification and the IEDB itself will be organized, and a newsletter will be published quarterly.
Defining the Scope of the IEDB {#s2}
==============================
Each scientific approach generates a set of epitope data, specific to itself, which must be integrated into a general representation of epitope information. In a programmatic sense, we believe that selecting data that fit one particular epitope definition or experimental bias is not our prerogative and would be unwise. Rather, we have opted to define a comprehensive, all-inclusive representation of information that separates epitope features into intrinsic and extrinsic features. Intrinsic features are those determined by the sequence and structure of an epitope, while extrinsic features are context-dependent attributes determined by the experimental or natural environment. This immunological perspective will be an organizing principle behind the IEDB.
Intrinsic Versus Extrinsic Features of an Epitope {#s3}
=================================================
At the level of T cell epitopes, intrinsic features included in the IEDB are: the molecular structure of the epitope, the binding affinity for different MHC receptors, and the affinity of MHC/ epitope complexes for TCRs of defined sequence. Likewise, at the level of B cell epitopes, intrinsic features include the epitope\'s molecular structure and binding affinity for antibody molecules of defined sequence. These features are unequivocally specified and are singularly associated with a given epitope structure or epitope/receptor combination.
Other features---such as immunogenicity, or whether an epitope is naturally processed---are not intrinsically associated with a given molecular structure of an epitope alone, but rather are context-dependent (i.e., extrinsic). Context information includes, for example, the species of the host in which a response was found, the assay utilized to measure responses, and the dose and route of administration. Likewise, the yield of a given epitope following proteasomal cleavage of a complex protein precursor is dictated by the overall sequence of the protein in which the epitope is contained. Also, the T cell and B cell responses to an epitope are heavily influenced by previous exposure of the immune system to the same or a related antigen. Collectively, these examples show that to meaningfully capture the immunogenicity of an epitope, the context in which it occurs must be described as well.
The IEDB Classes {#s4}
================
Formalizing the above considerations, we defined the main classes of the IEDB data as Reference, Epitope, Binding, and Context ([Figure 1](#pbio-0030091-g001){ref-type="fig"}). These classes represent the top level in the data hierarchy used to store epitope information in the IEDB. The class Reference defines one of three possible sources of data, namely literature, patent, and direct submissions. The Epitope class is subdivided into two categories: Epitope Structure, which specifies the molecular structure of an epitope itself, and Epitope Source, which identifies the pathogen/protein in which the Epitope is present. The Binding class captures intrinsic information relating to how the structure specified in the "Epitope" class interacts with well-defined receptors of the immune system such as MHC molecules or antibodies and TCRs of defined sequence. The Context class is organized into three subclasses, including T cell immune responses, naturally processed peptides, and B cell immune responses. [Table 1](#pbio-0030091-t001){ref-type="table"} is an example how the main features of a T cell epitope described in \[[@pbio-0030091-b2]\] would be displayed in the IEDB. Many more fields exist that are left blank because they are not appropriate for this particular epitope (such as antibody binding data) or are unknown (such as MHC binding data).
::: {#pbio-0030091-g001 .fig}
Figure 1
::: {.caption}
###### Main Classes in the IEDB
:::
:::
::: {#pbio-0030091-t001 .table-wrap}
Table 1
::: {.caption}
###### Sample Epitope Entry in the IEDB
:::
:::
A Scientific Approach for the Development of the Analysis Resource {#s5}
==================================================================
Our proposal includes the establishment and maintenance of an Analysis Resource of online tools for the Immune Epitope Database. Because this resource must be useful to the entire community, it is important that the tools provided cover a broad range of research areas relating to epitope discovery and analysis, and that no particular scientific "school" has priority. To identify tool candidates, we have generated a list of existing tools of interest through extensive literature searches and expert input. This will be periodically revised, taking advantage of input from the scientific community and NIAID.
The current list of candidate tools comprises an extensive menu of prediction tools for the identification of novel antibody and T cell epitopes from genome and protein sequences. At the level of antibody epitope predictions, standard methods of predicting which regions in a protein are likely to be on the surface will be provided, such as hydrophilicity analysis. Tools that use various methods for prediction of MHC binding will also be provided, along with tools predicting proteasomal processing and TAP transport of T cell epitopes.
We will also provide analytical tool resources to assist in vaccine discovery and development. These are designed to project population coverage of epitopes in different ethnicities, to project the degree of cross-reactivity within sets of different MHC molecules, and to assess the degree of conservancy of an epitope in various isolates of the same pathogen, both in related pathogens, and in potential hosts. Finally, tools to visualize data will be provided, such as those that display antibody antigen interactions where 3D structural information is available. We also hope that collection of consistently annotated data in the IEDB will allow the development of new, "context-sensitive," tools.
In deciding how many tools should be hosted in the IEDB, a balance has to be achieved between discriminating too much, which may leave user demands unaddressed, and discriminating too little, hosting so many tools that the collection becomes overly redundant and unmanageable. To facilitate an objective and transparent choice of which predictive tools should be hosted, the predictions of all candidate tools will periodically be evaluated. Most importantly, we plan to make all evaluations publicly available through the IEDB website, and we will encourage all different scientific groups to participate by submitting tools and evaluating data. Such prediction "contests" have had a tremendous positive impact in the field of tool evaluation and prediction of protein structure \[[@pbio-0030091-b3],[@pbio-0030091-b4]\]. To the best of our knowledge, this would represent the first attempt at a rigorous and comprehensive evaluation of prediction tools found on immune responses.
Conclusions {#s6}
===========
We envision a future in which the development of the Immune Epitope Database and Analysis Resource will help researchers throughout the world quickly access relevant information for evaluation of immune responses, assisting them in the development of prophylactic/therapeutic approaches against new and old, emerging and reemerging diseases.
This work was supported by the National Institutes of Health contract HHSN26620040006C.
**Citation:** Peters B, Sidney J, Bourne P, Huynh-Hoa B, Buus S, et al. (2005) The immune epitope database and analysis resource: From vision to blueprint. PLoS Biol 3(3): e91.
Bjoern Peters, John Sidney, Huynh-Hoa Bui, Ward Fleri, Mitch Kronenberg, Ralph Kubo, Muthu Sathiamurthy, Stephen Schoenberger, Steve Wilson, and Alessandro Sette are with the La Jolla Institute of Allergy and Immunology, San Diego, California, United States of America. Phil Bourne and Julia V. Ponomarenko are with the San Diego Supercomputer Center, San Diego, California, United States of America. Soeren Buus is with the University of Copenhagen, Copenhagen, Denmark. Grace Doh is with SH Grace Consulting, Seoul, Korea. Ole Lund is with BioCentrum-DTU, Technical University of Denmark, Lyngby, Denmark. David Nemazee is with The Scripps Research Institute, Department of Immunology, La Jolla, California, United States of America. Scott Stewart, Pamela Surko, and Scott Way are with Science Applications International Corporation, San Diego, California, United States of America.
IEDB
: Immune Epitope Database and Analysis Resource
NIAID
: National Institute for Allergy and Infectious Diseases
TCR
: T cell receptor
|
PubMed Central
|
2024-06-05T03:55:54.986657
|
2005-3-15
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065705/",
"journal": "PLoS Biol. 2005 Mar 15; 3(3):e91",
"authors": [
{
"first": "Bjoern",
"last": "Peters"
},
{
"first": "John",
"last": "Sidney"
},
{
"first": "Phil",
"last": "Bourne"
},
{
"first": "Huynh-Hoa",
"last": "Bui"
},
{
"first": "Soeren",
"last": "Buus"
},
{
"first": "Grace",
"last": "Doh"
},
{
"first": "Ward",
"last": "Fleri"
},
{
"first": "Mitch",
"last": "Kronenberg"
},
{
"first": "Ralph",
"last": "Kubo"
},
{
"first": "Ole",
"last": "Lund"
},
{
"first": "David",
"last": "Nemazee"
},
{
"first": "Julia V",
"last": "Ponomarenko"
},
{
"first": "Muthu",
"last": "Sathiamurthy"
},
{
"first": "Stephen",
"last": "Schoenberger"
},
{
"first": "Scott",
"last": "Stewart"
},
{
"first": "Pamela",
"last": "Surko"
},
{
"first": "Scott",
"last": "Way"
},
{
"first": "Steve",
"last": "Wilson"
},
{
"first": "Alessandro",
"last": "Sette"
}
]
}
|
PMC1065706
|
All organisms have mechanisms to ensure that dividing cells produce new cells with the proper number of chromosomes. The dividing cell closely monitors that chromosomes are copied exactly once and then distributed correctly to daughter cells. After replication, the chromosomes (now comprising two chromatids) align at the center of the cell, and are attached to a structure known as the spindle apparatus. A key point of attachment is the centromere, a characteristic constriction carried by each chromosome. The spindle, which is composed of microtubules, pulls the chromatids apart so that two complete sets of chromosomes are gathered together at each pole of the cell, which can then divide. Cohesion between chromosome copies, which keeps the chromatids together until just the right time, therefore plays a critical part in this process.
Chromosome cohesion is established during S phase (when the chromosomes are replicated) and is then dissolved completely in metaphase to allow sister chromatids to come apart. The dissolution of cohesion is highly regulated; human cell lines that have defects in the regulation of cohesion show the hallmarks of cancer cells \[[@pbio-0030094-b1]\]. Furthermore, it has been suggested that the abnormal karyotypes that result in diseases such as Down syndrome are the result of the improper dissolution of chromosome cohesion \[[@pbio-0030094-b2]\]. Finally, mutation of a factor required to load cohesin---the protein complex responsible for chromosome cohesion---onto chromosomes appears to cause Cornelia de Lange syndrome, a clinically heterogeneous developmental disorder that may include facial dysmorphia, upper-extremity malformations, hirsutism, cardiac defects, growth and cognitive retardation, and gastrointestinal disorders \[[@pbio-0030094-b3],[@pbio-0030094-b4],[@pbio-0030094-b5]\].
Cohesion serves at least three roles in the cell with respect to accurate genome transmission. First, cohesion close to the centromere facilitates bi-orientation of chromosomes, such that each chromosome becomes attached to the two poles of the spindle \[[@pbio-0030094-b6]\]. Second, it prevents the splitting of chromosomes until all bipolar attachments are made \[[@pbio-0030094-b6]\]. The function of cohesion at the centromere is presumably to oppose the force of microtubules, which pull the chromosomes to opposite spindle poles; this force is not exerted along the chromosome arms, which means that cohesion at centromeres and along arms is functionally distinct. Third, cohesion along chromosome arms may be essential for proper chromosome condensation \[[@pbio-0030094-b7],[@pbio-0030094-b8]\], although the function of cohesion at chromosome arms is something of a mystery.
Differences between Arms and Centromeres {#s2}
========================================
Cohesion in eukaryotic cells is mediated by a multi-subunit protein complex called cohesin. Cohesin consists of four proteins: Smc1, Smc3, Scc1/Mcd1 (also known as kleisin), and Scc3 (SA2). The Smc (structural maintenance of chromosomes) proteins form intramolecular coiled coils that have been observed in the electron microscope to form a V shape with sides that are 50 nm long \[[@pbio-0030094-b9]\]. The cohesin complex has been proposed to form a ring structure that encircles sister chromatids \[[@pbio-0030094-b10]\]. Alternately, two rings may snap sisters together via interactions between the coiled coils of the Smc proteins \[[@pbio-0030094-b11]\]. All members of the cohesin complex are essential in budding yeast, Saccharomyces cerevisiae, since mutation results in the precocious dissociation of sister chromatids. Functional orthologs of these proteins exist in all eukaryotes.
There are at least two types of cohesin sites: (1) cohesin associated with the centromere and the nearby pericentric domain, and (2) cohesin associated with chromosome arms \[[@pbio-0030094-b12],[@pbio-0030094-b13],[@pbio-0030094-b14],[@pbio-0030094-b15]\]. In S. cerevisiae, cohesin at centromeric and pericentric domains is spread over a broad region (up to 50 kb), large quantities of the protein complex are bound, and binding is not affected by the natural transcriptional and coding status of the regions that are occupied. By contrast, binding sites in arms tend to be much smaller (about 1 kb)---at least in S. cerevisiae, where they have been most extensively characterized---and of lower intensity, and are spaced at approximately every 11 kb (see [Figure 1](#pbio-0030094-g001){ref-type="fig"}). Cohesin in arms localizes to regions lacking transcription in yeast \[[@pbio-0030094-b12],[@pbio-0030094-b16],[@pbio-0030094-b17]\]. This reinforces the view that there may be functional differences in arm and pericentric cohesion and perhaps different mechanisms to load cohesin, as has been proposed for cohesin on meiotic chromosomes for S. pombe \[[@pbio-0030094-b18]\]. A unifying feature of all cohesin-binding sites in S. cerevisiae is high AT (adenine and thymine) content \[[@pbio-0030094-b12],[@pbio-0030094-b15]\].
::: {#pbio-0030094-g001 .fig}
Figure 1
::: {.caption}
###### Cohesin Sites for Sister Chromatids of Chromosome I in S. cerevisiae
Cohesin sites (red ovals) are concentrated at the centromere/pericentric region (where the two chromatids are "pinched"), but also occur along the arms of the chromatids.
:::
:::
Another important difference between cohesin binding along arms and at centromeres is that the arm sites do not appear to be dependent on a DNA consensus sequence, whereas binding to pericentric regions requires specific centromere sequence \[[@pbio-0030094-b13],[@pbio-0030094-b14]\]. The S. cerevisiae centromere sequence is composed of three DNA elements (CDEI, CDEII, and CDEIII). Studies of cohesion at the centromere reveal that as little as 100 bp (a portion of CDEII and the entire CDEIII) are required to direct cohesion \[[@pbio-0030094-b13],[@pbio-0030094-b14],[@pbio-0030094-b19]\]. Mutations in the protein Ndc10 have also been shown to affect cohesin deposition at centromeres. Ndc10 forms part of a structure known as the kinetochore, which forms around the centromere and is responsible for the attachment to the spindle; establishment and maintenance of cohesion at pericentric regions may therefore rely on both the centromere sequence and kinetochore function \[[@pbio-0030094-b13],[@pbio-0030094-b20]\]. Presumably both arm and pericentric cohesion are important for chromosome dynamics, but the functional differences between the two are not well understood.
Cohesion---It\'s Just a Phase {#s3}
=============================
Cohesion can be divided into four phases that occur during the cell cycle ([Figure 2](#pbio-0030094-g002){ref-type="fig"}): (1) deposition in G1 (the gap in the cell cycle before S phase), (2) establishment in S phase, (3) maintenance in G2 (the gap between S and mitosis), and (4) dissolution in mitosis. During G1, Scc2 and Scc4 are responsible for loading cohesin onto unreplicated double-stranded DNA \[[@pbio-0030094-b21]\]. Then, during S phase, several proteins are involved in establishment of cohesion between replicated chromatids. Eco1 and Chl1 are required for establishing cohesion but not for maintenance \[[@pbio-0030094-b22],[@pbio-0030094-b23],[@pbio-0030094-b24]\]. The associations between cohesion and DNA replication have led to a model whereby cohesion is established coincident with the passage of the replication fork \[[@pbio-0030094-b25]\]. This requires an alternative replication factor C (RF-C) complex \[[@pbio-0030094-b26],[@pbio-0030094-b27],[@pbio-0030094-b28]\] and may require the origin recognition complex (ORC) \[[@pbio-0030094-b29]\]. Cohesion is maintained during G2 by the cohesin complex, and is eventually dissolved in mitosis to allow sister chromatids to separate.
::: {#pbio-0030094-g002 .fig}
Figure 2
::: {.caption}
###### Behavior of Cohesin during the Cell Cycle
One cohesin complex is depicted at each site for the sake of simplicity; at the centromere especially there are likely to be many complexes. Cohesion is represented as occurring via the "encircling" model; other models have been proposed.
:::
:::
The dissolution of cohesion is regulated by at least two mechanisms. First, subunits of the complex may be phosphorylated, which facilitates their removal. In S. cerevisiae and human cells, phosphorylation of Scc1/Mcd1 by Polo kinase makes it a better substrate for proteolysis \[[@pbio-0030094-b30],[@pbio-0030094-b31],[@pbio-0030094-b32]\]. In this issue of *PLoS Biology*, one of two related papers exploring the regulation of cohesin in vertebrates shows that phosphorylation of Scc3 (SA2) by Polo-like kinase is apparently sufficient to allow dissociation from chromosome arms, which occurs during prophase \[[@pbio-0030094-b32]\]. In Xenopus extracts, phosphorylation of cohesin also depends on Polo-like kinase, and this phosphorylation reduces the ability of cohesin to bind to chromatin \[[@pbio-0030094-b8]\].
The second mechanism that can facilitate the dissolution of cohesin is proteolysis; this may be particularly important at centromeres. The Scc1/Mcd1 component of the cohesin complex is cleaved by a separase (Esp1) whose activity is held in check by a securin (Pds1) until separation at the metaphase-to-anaphase transition \[[@pbio-0030094-b33],[@pbio-0030094-b34]\]. Separase is a cysteine protease that cleaves Scc1/Mcd1, presumably resulting in the cohesin complex falling apart and being unable to hold sister chromatids together.
Scc1/Mcd1 at pericentric regions is protected from phosphorylation during prophase---and therefore dissociation from chromosomes is prevented---by proteins known as shugoshins \[[@pbio-0030094-b35],[@pbio-0030094-b36],[@pbio-0030094-b37]\]. In the second paper on cohesin in this issue of *PLoS Biology*, McGuinness et al. show that shugoshin specifically protects Scc3 (SA2) at the centromere, so that centromeric cohesion is preserved until the chromosomes are ready to separate \[[@pbio-0030094-b35]\]. Vertebrate shugoshin has been shown to have a strong microtubule-binding domain \[[@pbio-0030094-b36]\] and is found at the kinetochore \[[@pbio-0030094-b37]\]. Recent evidence suggests that shugoshin in S. cerevisiae may sense tension between sister chromatids, acting as part of a spindle checkpoint that monitors whether chromosomes are properly aligned on the mitotic spindle \[[@pbio-0030094-b38]\]. It is currently unclear why the cell has two mechanisms to dissociate cohesin from chromosomes, although it is interesting to speculate that this could be related to different functions of cohesin at chromosome arms versus pericentric domains. For instance, cohesin in chromosome arms may help to organize or condense chromosomes, whereas cohesin at centromeres may be more directly involved in chromosome bi-orientation at the spindle and segregation. These functions may be important during different phases of the cell cycle.
A Link between Chromatin and Cohesin {#s4}
====================================
Several results suggest that transcription and cohesin binding are incompatible. In Drosophila, one of the components that loads cohesin (Nipped-B or Scc2) has also been shown to be required for long-range promoter--enhancer interactions \[[@pbio-0030094-b39],[@pbio-0030094-b40]\]. One model proposed to explain this result is that cohesin can prevent long-range promoter--enhancer interactions and that removal of cohesin can restore these interactions and allow transcription to occur \[[@pbio-0030094-b41]\]. In this model, Nipped-B or Scc2 can act as both a loading factor and an unloading factor for cohesin. This model further speculates that rather than Cornelia de Lange syndrome stemming from a cohesin loading defect, the failure to unload cohesin from regions that need to be transcribed leads to transcriptional defects that cause the syndrome. In S. cerevisiae it has been shown that driving transcription through a centromere via an inducible promoter prevents cohesin from associating and results in chromosome missegregation and cell death \[[@pbio-0030094-b13]\]. Cohesin is found at the boundaries of the HMR locus, the right telomere of Chromosome III, and the RDN1 array, all regions of silent chromatin \[[@pbio-0030094-b16]\]. Cohesin localizes to intergenic regions where transcription is converging \[[@pbio-0030094-b12],[@pbio-0030094-b17]\].
Since transcription and chromatin configuration are intimately related, it is possible that chromatin may play an important role in the localization of cohesin. Indeed, the chromatin remodeling complex RSC (remodels the structure of chromatin) has been shown to be important for establishment of cohesin binding \[[@pbio-0030094-b42]\], and another study suggests RSC is particularly important for cohesin association with chromosome arms \[[@pbio-0030094-b43]\]. The chromatin remodeling complex ISWI (SNF2h) has been shown to be essential for cohesin to localize to Alu repeats (certain DNA sequences) in human cells \[[@pbio-0030094-b44]\]. The possibility also exists that cohesin itself may influence transcriptional status and act as a transcriptional boundary \[[@pbio-0030094-b39],[@pbio-0030094-b40],[@pbio-0030094-b45]\]. The preferential location of cohesin in heterochromatin in pericentric regions in S. pombe also supports the idea that chromatin modification/structure is a key determinant of cohesin localization \[[@pbio-0030094-b46],[@pbio-0030094-b47]\]. It is interesting to speculate that chromatin differences and transcriptional differences between chromosome arms and centric regions will turn out to be related to different mechanisms for loading and removal of cohesin from these regions.
While one of the primary roles for chromosome cohesion in bi-orientation and mitotic chromosome segregation is well-established, the complexities of the regulation of cohesion are still being discovered. Cohesin may be involved in multiple ways in chromosome dynamics. Future studies focusing on the differences between cohesion at chromosome arms versus pericentric domains and the link between cohesion and transcription will likely yield very interesting insights into the function of the cohesin complex in the maintenance of genome integrity.
**Citation:** Gerton J (2005) Chromosome cohesion: A cycle of holding together and falling apart. PLoS Biol 3(3): e94.
Jennifer Gerton is at the Stowers Institute for Medical Research, Kansas City, Missouri, United States of America. E-mail: <jeg@stowers-institute.org>
|
PubMed Central
|
2024-06-05T03:55:54.987714
|
2005-3-15
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065706/",
"journal": "PLoS Biol. 2005 Mar 15; 3(3):e94",
"authors": [
{
"first": "Jennifer",
"last": "Gerton"
}
]
}
|
PMC1065707
|
Imagine this scenario. You\'re the director of one of the world\'s largest medical research charities, and you receive notification from one of your funded investigators in Africa reporting some exciting progress toward the development of a vaccine for malaria. The work has just been published, so you log onto the Web to do a quick keyword search, and a link to the article is brought up on your screen.
Then imagine the frustration when you click on the link to read the message, "Access Denied---access to this journal is restricted to registered institutional and individual subscribers."
And there\'s the rub: this actually happened to the Director of the Wellcome Trust. Prior to this, the committee that advises the Wellcome Trust Library were already asking whether the Trust should adopt a formal position on the continually increasing prices of journal subscriptions and the problems this trend was causing research libraries.
These events encouraged the Trust to investigate the publication of scientific research, to see if there was anything research-funding organisations could be doing to stimulate change in what appears to be a failing market. As it turns out, there is quite a lot. I now believe it is the funders of research---charities, governments, and other publicly funded bodies such as national research agencies---who hold the purse strings that can untie scientific discoveries from a publishing market that is no longer serving the community as well as it could. That is why today the Trust is a leading advocate for enabling free access to research literature through support for new publishing models, such as that of the Public Library of Science, and the establishment of publicly accessible repositories, working in partnership with the United States National Institutes of Health--funded PubMed Central \[[@pbio-0030097-b1]\].
::: {#pbio-0030097-g001 .fig}
:::
It is worth noting that the Trust is not a novice in seeking better ways to disseminate research findings. The fact that the sequence of the human genome is an openly accessible work is due in large measure to the Trust\'s determination that this information be in the public domain and not hidden behind commercial subscriptions. As a consequence of that insistence, we believe, these data are a more widely used and valuable resource.
> "Trust-funded researchers will have to deposit an electronic version of their manuscripts in PMC to be made available for free via the Internet within 6 months of publication."
The Trust began its investigation of the scientific publishing sector by commissioning two pieces of research: one to inform itself of the economics of the publishing sector, and a second to explore whether there were alternative business models out there that could enable research to have the quality assurance it needs (peer review) whilst being available for free, using the Web as the medium of publication.
The Economics of Publishing {#s2}
===========================
The first Trust-commissioned study described how scientific research publishing has traditionally worked and why it can be described, in economic terms, as a failing market \[[@pbio-0030097-b2]\]. Essentially, the producers (researchers as authors) and the consumers (researchers as readers) are isolated from any of the costs within the system. Researchers give away the copyright to their work, for free, to the publishers, who organise the peer review and copyedit the article. The publishers then sell it to libraries at prices that range from enough to cover their costs through to some pretty high profits---some over 30%. These profits escape from an otherwise self-contained financial cycle to satisfy shareholders or run learned societies; unlike typical publishing relationships, none are returned directly to the author (the researcher who wrote the piece) or even to the consulting experts (the researchers who provided the peer review).
At the same time, researchers as readers access the material, if they are able to do so, through their employing institution, either using the library or---more typically now---via the Internet through the institution\'s subscription. To the researcher this access appears free, effectively creating a market system that has no pressures from the producers or consumers to change. One consequence of this is that publishers have been able to increase subscription prices well above inflation; the United Kingdom has seen subscription rates rise by more than 200% in the last ten years (Blackwell\'s periodical price indexes; \[[@pbio-0030097-b3]\]). The money used to fund UK libraries is all public money, and over 90% of the funds paying for research in the UK university system is either government or charitable \[[@pbio-0030097-b4]\]---so in a sense the people who are paying for the research cannot access its findings without paying an additional fee.[](#pbio-0030097-g002){ref-type="fig"}
::: {#pbio-0030097-g002 .fig}
::: {.caption}
###### Access denied at *The Journal of Infectious Diseases*
:::
:::
This then begs the question of what alternatives there are to this traditional system, now that the Internet has become the researcher\'s tool of choice for searching and accessing the literature. The second piece of research commissioned by the Trust looked at different business models for research publishing, in order to address this question \[[@pbio-0030097-b5]\]. It compared open-access journals, which often levy a charge to publish but provide the journal for free, and the majority of the traditional models, which take the research for free but charge readers to read it.
This study convinced the Trust that the best way forward to improve access to research findings would be through open access to scientific research articles. This essentially means two things: first, that the copyright holder or holders must grant to the public a free, irrevocable, perpetual license to use, copy, distribute, and make derivative works of their research article, in any medium for any purpose (excepting those that constitute plagiarism or other dishonest acts, of course); and second, that a digital copy must be deposited in an open public archival repository (for example, the US National Library of Medicine\'s PubMed Central). Whilst a debate continues as to the most appropriate route to achieve open access to all research literature, it is important to bear in mind that the publication and the archiving of research articles are intrinsically linked. Both aspects of open access need to be explored and experimented with, and the Trust is actively pursuing solutions for the problems of both.
Alternative Business Models {#s3}
===========================
The findings of the second report seem to have caused quite a controversy---particularly in the suggestion that moving wholesale to an open-access publishing model might produce savings of up to 30% \[[@pbio-0030097-b6]\]. One common misinterpretation of this conclusion is that any such savings would be due solely to discontinuing the printed versions of publications that are freely available online. This is incorrect. In fact, if savings are to be made in an open-access model, they will largely be found in the variable costs of journal production---since an open-access journal will not have to cover the costs of subscription management, licence negotiations, or sales, and little is required for marketing and distribution.
In a comparison included in the report, an article in a good- to high-quality journal produced in the subscription model is estimated to cost US\$2,750. The equivalent cost under an author-side payment model is estimated as US\$1,950---a comparable saving of 30% on the costs, and a saving of 90% when the variable costs are compared. It must be remembered that cost does not equate to price, so to these figures, regardless of the mode of publication, must be added overhead expenses and, of course, profit. However, if a truly competitive market is created---where payments are directed to publishers not by third parties but by those directly involved in the scientific enterprise, who could easily compare the varying article processing charges of different open-access publishers---then the actual savings might well be substantially higher.
At its essence though, the open-access debate is not about economics, it is about access. That is why the Trust has been in discussion with the US National Library of Medicine about the possibility of creating a UK PubMed Central (UKPMC) as a publicly accessible repository for Trust-funded research.
UK PubMed Central {#s4}
=================
The proposal is that a UKPMC will be run as a proper electronic library: it will collect, collate, and archive whole journals and be developed to receive single articles as well. Submission will be as straightforward as attaching a document to an email. UKPMC will be able to accept manuscripts in any format, including Microsoft Word, and it will be the responsibility of UKPMC to convert the files it receives into extensible markup language (XML) to enable the appropriate document type definition (DTD) to be assigned. UKPMC will also correct the structural, content, and consistency errors that occur when converting text for digital preservation, and provide the conversion process to print a "clear" PDF version of included articles to those users who download them. This is a process well used by the National Library of Medicine, and the one most suited for the long-term, digital preservation of articles.
And once articles are in a digital format they can be searched and used in different ways. For example, genome sequence data, chemical compounds, or protein structures embedded within an article can be searched for in other articles and linked directly to genome or structural databases uncovering new genetic markers, drug uses, or protein functions. The articles themselves become live research material greatly improving the efficacy of the research itself.
For a funder, having all its research in one format, "under one roof", and searchable will improve the efficiency of strategy setting---for example, setting funding priorities---assessing the outputs of the funded research, and even gaining an insight into the impact of the work. As grants management becomes more electronic, there can be a direct link between original research proposals and the research outputs.
For a medical charity like the Trust, I believe it is our duty to actively encourage the most efficient processes available to maximise the likelihood that the research we fund will have the greatest possible health benefit.
That is why the Trust will be making it a requirement of its grant conditions that Trust-funded researchers deposit an electronic version of their manuscripts in UKPMC to be made available for free via the Internet within 6 months of publication. The delay means that this is not open access in the truest sense. However, the Trust considers that the development of a PubMed Central portal in the UK offers the best next step in the transition towards a situation where all high-quality peer-reviewed research is available for free via the Internet, whilst leaving all publishers room for manoeuvre in this changing market.
**Citation:** Terry R (2005) Funding the way to open access. PLoS Biol 3(3): e97.
Robert Terry is Senior Policy Adviser at the Wellcome Trust, Cambridge, United Kingdom. E-mail: <r.terry@wellcome.ac.uk>
UKPMC
: United Kingdom PubMed Central
|
PubMed Central
|
2024-06-05T03:55:54.989054
|
2005-3-15
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065707/",
"journal": "PLoS Biol. 2005 Mar 15; 3(3):e97",
"authors": [
{
"first": "Robert",
"last": "Terry"
}
]
}
|
PMC1065708
|
Since their discovery in the 1970s and 1980s, giant tubeworms at hydrothermal vents and cold seeps have fascinated biologists and laymen alike---not only for their alien morphology ([Figure 1](#pbio-0030102-g001){ref-type="fig"}), but also for epitomizing the perfect animal--microbe symbiosis. They are among the biggest worms on this planet---some over 3 m long---yet they do not eat other organisms. Tubeworms thrive independently of photosynthetic production \[[@pbio-0030102-b1]\]. They have even lost their entire digestive tract. One of the most exciting findings in early tubeworm research was the discovery that the worm\'s food is delivered by bacterial symbionts \[[@pbio-0030102-b2]\]. The chemoautotrophic symbionts live intracellularly in a specialized worm tissue called the trophosome. They are sulfide oxidizers, using the free energy yield from the oxidation of sulfide with oxygen to fix carbon dioxide with their bacterial Rubisco enzyme. In exchange for providing nutrition for the worm, the symbionts are sheltered from grazing, but most importantly, they receive a steady source of sulfide and oxygen via the highly adapted blood circulation system of the worm. (I will never forget how horrified I was as a young student by the amounts of almost human-like blood flowing into my lab dish while dissecting tubeworms to analyze trophosome enzyme activity.) Tubeworm blood physiology, in particular the hemoglobin molecules, are tailored specifically to the needs of the symbionts. However, the host metabolism in itself is not different from that of many other animals, the main source of energy being aerobic respiration of carbohydrates. In other words, tubeworms and their symbionts need oxygen as an electron acceptor---so, after all, they are dependent on photosynthesis, the main oxygen-producing process on earth.
::: {#pbio-0030102-g001 .fig}
Figure 1
::: {.caption}
###### Vestimentiferan Tubeworms
\(A) Close-up photograph of the symbiotic vestimentiferan tubeworm Lamellibrachia luymesi from a cold seep at 550 m depth in the Gulf of Mexico. The tubes of the worms are stained with a blue chitin stain to determine their growth rates. Approximately 14 mo of growth is shown by the staining here. (Photo: Charles Fisher) (B) Close-up photograph of the base of an aggregation of the symbiotic vestimentiferan tubeworm L. luymesi from a cold seep at 550 m depth in the Gulf of Mexico. Also shown in the sediments around the base are orange bacterial mats of the sulfide-oxidizing bacteria Beggiotoa spp. and empty shells of various clams and snails, which are also common inhabitants of the seeps. (Photo: Ian MacDonald)
:::
:::
Classification of Host and Symbiont {#s2}
===================================
With their strange morphology, vent tubeworms were first classified as a novel phylum, Vestimentifera \[[@pbio-0030102-b3]\]. Recently they have been regrouped together with the pogonophoran tubeworms ([Figure 2](#pbio-0030102-g002){ref-type="fig"}) into a family of annelid polychaetes called the Siboglinidae \[[@pbio-0030102-b4],[@pbio-0030102-b5]\]. Vestimentiferan tubeworms of hydrothermal vents grow on chimneys and other hard substrates in the vicinity of active vents, which emit reduced compounds like hydrogen and sulfide \[[@pbio-0030102-b6]\]. Vestimentiferan tubeworms living at cold hydrocarbon seeps, i.e., the lamellibrachids and escarpids, are adapted to a sedimentary environment, with a substantial part of the body and tube of many species extending into the mud. All vestimentiferan tubeworms found today at vents, seeps, and a few other reduced submarine habitats harbor sulfide-oxidizing endosymbionts in their trophosome. These symbionts belong to bacteria of the gamma-proteobacteria clade and are phylogenetically related to each other \[[@pbio-0030102-b7]\]. (For the only known exception see \[[@pbio-0030102-b8]\].)
::: {#pbio-0030102-g002 .fig}
Figure 2
::: {.caption}
###### Pogonophoran Tubeworms Being Sampled at the Haakon Mosby Mud Volcano
(Source: AWI/IFREMER expedition RV POLARSTERN/VICTOR 6000 in 2003)
:::
:::
Tubeworm Mysteries {#s3}
==================
The study of tubeworms is now in its fourth decade, and there are still many fascinating problems to be solved. One of the most interesting---but also most difficult---questions in tubeworm symbiosis is how this obligate and highly integrated interaction between microbes and animals evolved. How can a worm evolve into a perfect home for chemosynthetic bacteria? What are the main evolutionary steps towards this symbiosis, and in which order did they occur? Another intriguing problem is how the worms acquire their endosymbionts, which appear to be taken up from the environment---but so far have not been detected as free-living forms. How does the host recognize its specific symbiont from the vast diversity of gamma-proteobacteria and sulfide oxidizers in the environment? Furthermore, how do tubeworms populate new vents, seeps, and other reducing environments emerging from the ever-changing ocean floor---how do their larvae migrate and settle, and what determines the distribution and lifetime of tubeworm populations in the different mid-ocean ridge and continental margin habitats? Although these questions are still to be answered, new research and techniques are beginning to provide intriguing clues.
Seep Vestimentifera and Their Energy Source {#s4}
===========================================
At some seeps the vestimentiferan tubeworms are so abundant that they form a special habitat that is attractive for a host of other marine species \[[@pbio-0030102-b9]\]. Seep vestimentiferans are usually thinner, have slower growth rates, and have greater longevity than their vent relatives \[[@pbio-0030102-b10]\]. For example, a 2-m-long Lamellibrachia luymesi individual is estimated to be more than 200 y old and hence represents the longest-lived animal on earth \[[@pbio-0030102-b11],[@pbio-0030102-b12]\]. At seeps, geological processes causing fluid and gas seepage can last hundreds to millions of years, whereas hydrothermal vents often have a lifespan on the order of decades. Vent tubeworm colonies will die when their chimneys stop venting, i.e., delivering sulfide, so they are adapted to a rapidly changing environment, as typified by their fast growth and high reproduction.
Like vent vestimentifera, seep vestmentifera also depend on the availability of sulfide in their direct vicinity, but they are sessile, and anchor on hard substrates such as carbonates. Individual aggregations at seeps can consist of hundreds to thousands of worms, requiring sulfide fluxes of half a mole per day---and this for more than 200 y \[[@pbio-0030102-b12]\]. So an ecological problem that has always intrigued biologists and geochemists alike is how these tubeworms obtain their energy over the long term. Because vent and seep vestimentifera depend on sulfide-oxidizing symbionts, their distribution is limited to habitats with high sulfide fluxes lasting for at least a few reproductive cycles. However, at cold seeps, unlike hydrothermal vents, most of the chemical energy occurs in the form of hydrocarbons. Cold seeps are characterized by high fluxes of methane, higher hydrocarbons (such as ethane, propane, butane), and/or petroleum from deep subsurface reservoirs. Often the source fluids and gases do not contain much sulfide, because there are no high-temperature seawater--rock interactions involved in their formation, as there are at vents. Some pogonophoran tubeworms at seeps have teamed with methane-oxidizing symbionts to profit from the high availability of hydrocarbons, but seep vestimentiferans do not appear to be able to directly tap this resource. However, seep vestimentiferans are still capable of producing enormous biomass over many years with the help of their sulfide-oxidizing symbionts. So where does the supply of sulfide come from at seeps that enables such large aggregations to be maintained for so long?
Only recently was it realized that anaerobic microbial processes, namely, the oxidation of hydrocarbons with sulfate, could produce astonishingly high fluxes of sulfide in cold seep settings \[[@pbio-0030102-b13],[@pbio-0030102-b14]\]. At methane seeps, methanotrophic microbial communities inhabiting the surface sediments oxidize methane with sulfate, which results in very high sulfide fluxes \[[@pbio-0030102-b13]\]. If the seepage consists of other hydrocarbons such as petroleum, their degradation with sulfate supports an even higher production of sulfide \[[@pbio-0030102-b14]\]. In some seep sediments, sulfide concentrations can reach 25 mM in subsurface sediments (5--10 cm below the sediment surface). Such concentrations are not known from tubeworm habitats at hydrothermal vents.
However, the zones of high hydrocarbon turnover and sulfide flux at seeps are often limited to only a few centimeters below the seafloor, depending on hydrocarbon flows and the rate of sulfate transport from the bottom water into the sediments. Sulfate is crucial because the free-living hydrocarbon-degrading microbes in seep sediments depend on this electron acceptor for an energy yield. Without sulfate to fuel the oxidation of hydrocarbons, sulfide production stops, even if there is still an enormous reservoir of hydrocarbon available. How might tubeworms, sulfide-oxidizing symbionts, and benthic hydrocarbon degraders overcome these limitations?
Ménage à Trois---A Model Solution {#s5}
=================================
Cordes et al. \[[@pbio-0030102-b15]\] have now provided an answer to how the stability of sulfide production is maintained over such long periods and how the worms optimize sulfide uptake. Seep vestimentifera have specific adaptations to their habitat. A main adaptation is the subsurface part of the lamellibrachids called a "root." The tubeworm root appears to have a special function in the energy cycle of the organism---as in plant roots. Several authors have proposed that the worm roots are not only important in sulfide uptake, but generally in geochemical engineering of the sediments in the direct environment \[[@pbio-0030102-b16],[@pbio-0030102-b17],[@pbio-0030102-b18]\]. Obviously such hypotheses are very difficult to test---today it is still hardly possible to measure gas, petroleum, and sulfide fluxes in the seafloor in situ at depth, especially below tubeworm aggregations. But it is also not possible to recover whole aggregations of worms and to keep them alive in the lab for biochemical and biogeochemical measurements---this would require simulation of seepage under pressure. Instead, Cordes et al. \[[@pbio-0030102-b12],[@pbio-0030102-b15]\] have used geochemical and biological modeling to solve the intriguing question of seep vestimentiferan longevity and how they might also interact with free-living anaerobic microbes to increase sulfide availability.
To explain the persistence of the large tubeworm colonies in the Gulf of Mexico, Cordes et al. suggest a broader mutualistic interaction between the tubeworm, its endosymbiont, and benthic hydrocarbon-degrading and sulfide-producing microbes. Seep tubeworms take up sulfide from the sulfide-rich subsurface sediment zones through the roots, but, crucially, they may also release sulfate through the roots as a byproduct of sulfide oxidation by the tubeworm\'s endosymbiont. Sulfate may also be ventilated through the tube into the sediments. Since anaerobic microbial communities in subsurface hydrocarbon-rich sediments are limited by sulfate influx, any additional supply of sulfate enhances their production of sulfide. Furthermore, the removal of sulfide by the worm will thermodynamically favor anaerobic hydrocarbon oxidation coupled to sulfate reduction. Hence, the tubeworm roots may provide an excellent habitat for anaerobic hydrocarbon oxidizers. For example, Cordes et al. predict in their model that nearly all of the sulfate released through the root will be utilized by benthic microbes for anaerobic hydrocarbon degradation in the direct vicinity of the worm. This process could provide 60% of the sulfide needed by a tubeworm aggregation to persist for 80 y. Hence, it may even be concluded that tubeworms farm anaerobic hydrocarbon degraders to provide a steady supply of sulfide to their endosymbionts. Especially at petroleum seeps, this would guarantee a lifelong energy source and help explain the extraordinary longevity of the worms. The mutual benefit arising from the association of sulfide oxidizers, sulfate reducers, and a host worm is known to be exploited by the oligochaete Olavius algarvensis \[[@pbio-0030102-b19]\]. In this very effective "ménage à trois" the sulfate reducer has even become an endosymbiont of the worm. Interestingly, some of our recent studies at the methane seeps of Hydrate Ridge (Cascadia margin) also show that certain populations of anaerobic methane oxidizers are specifically associated with seep organisms---such as the symbiotic clam Calyptogena and the giant filamentous sulfide oxidizer Beggiatoa \[[@pbio-0030102-b20]\]. But many more examples may be out there, of bacterial and archaeal populations specifically growing in the "rhizosphere" of benthic organisms, potentially profiting from bioturbation, bioirrigation, fecal deposits, and exudates.
The association and interaction between benthic fauna and sedimentary microorganisms is a very interesting field of study, although inevitably still very speculative. So far it has been limited by a lack of appropriate technologies, not only for in situ biogeochemical and biological measurements, but also for quantitative investigation of specific functional microbial populations. Some insight can be provided by clever environmental modeling approaches---such as the one developed by Cordes et al., but ultimately the models need empirical verification. Only very recently has it become possible to combine visually targeted sampling ([Figure 2](#pbio-0030102-g002){ref-type="fig"}) and high-resolution measurements of geochemical gradients with molecular tools for the identification of microbes, such as 16S rDNA and organic-biomarker-based techniques. For the study of continental margin and deep-sea ecosystems, this requires the availability of underwater vehicles ([Figure 3](#pbio-0030102-g003){ref-type="fig"}) as well as multidisciplinary research platforms and extensive, highly detailed lab work---so this is very expensive research. Yet this is the future, if we want to determine whether such an intriguing ménage à trois as proposed by Cordes et al. accounts for the presence and longevity of these extraordinary tubeworms, and possibly also other chemosynthetic symbioses, forming some of the most fascinating marine ecosystems at continental margins.
::: {#pbio-0030102-g003 .fig}
Figure 3
::: {.caption}
###### Harbor Branch Oceanographic Institution\'s Submersible "Johnson SeaLink"
(Source: Gulf of Mexico Cruise SJ0107)
:::
:::
I thank Erik Cordes and Nicole Dubilier for their comments on the text.
Citation: Boetius A (2005) Microfauna--macrofauna interaction in the seafloor: Lessons from the tubeworm. PLoS Biol 3(3): e102.
Antje Boetius is at the Max Planck Institute for Marine Microbiology, Bremen, Germany. E-mail: <aboetius@mpi-bremen.de>
|
PubMed Central
|
2024-06-05T03:55:54.990731
|
2005-3-15
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065708/",
"journal": "PLoS Biol. 2005 Mar 15; 3(3):e102",
"authors": [
{
"first": "Antje",
"last": "Boetius"
}
]
}
|
PMC1065709
|
The Fourth International Conference on Neuroesthetics was not a large event, but it was an unusual one. Held on a single day in the basement auditorium of the Berkeley Art Museum at the University of California at Berkeley, it brought together a typically motley collection of intellectuals who would willingly give over a sunny Saturday to an opportunity to learn from a panel of distinguished speakers. This was not the unusual part. Nor was it unusual that the meeting was touted as an interdisciplinary event, bringing together the best and brightest of different fields. These days, and perhaps it has always been the case, interdisciplinarity is the rule rather than the exception of innovative science.
What set this meeting apart was the fluid progression from art to science, in content as well as in style. The artists were more or less scientific, the scientists more or less artistic. The topic was empathy ("Empathy in the Brain and in Art")---more particularly, man\'s (and not just man\'s) ability to recognize and respond to the expressions of others. What do we respond to in an expression and what are the mechanisms in the brain that underlie these responses? And as the primatologist Frans de Waal (Emory University) highlighted, how much of our empathic natures do we share with our ape cousins?
In a slide presentation of her work and sources of inspiration, portrait photographer Judy Dater clearly captured with great sensitivity an infinite variety of poignant expressions. However, when asked, she could not clearly articulate the choices she had made in posing and photographing her subjects, could not give dimensions to the criteria she was using. In contrast, the performance artist Leonard Pitt had clearly made a science out of expression. His physical demonstrations with Balinese masks, carved into iconic images of happiness, sadness, or anger, gave the audience insight into the variety of subtle expression that could be attributed to the mask with simple postural adjustments. Happiness melted into melancholy, sadness into ennui. "It\'s not about moving," he observed, "it\'s about not moving."
The psychologist Paul Ekman (University of California at San Francisco) brought the official stamp of academia to his science of expression, documenting in the language of training-dependent effects on recognition the subtle range of expressions and microexpressions we can identify. For a practical example, he showed a clip from testimony in the O. J. Simpson trial of a moment in which the infamous "houseguest" Kato Kaelin was caught out in a lie. A fleeting hostile look crossed his otherwise carefully schooled features: invisible until pointed out, unmistakable after.
Where the artist and psychologist show us the richness of the human behavioral repertoire, the neuroscientist tries to break behaviors down into manageable, testable predictions of the associated brain activity. In contrast to the feasts of expression presented by other speakers, the faces representative of basic emotions used by the cognitive neuroscientist Ray Dolan (University College London) to study the neural activity engendered by expressions seemed almost too caricatured to be meaningful. But Dolan, introducing his subject through the portraiture of American colonial artist Gilbert Stewart, deconstructed the information we derive from the expressions of others into five categories---familiarity, identity, emotion, intentionality, and character---and was able to describe neural activity associated with carefully constructed experiments to probe each of these facets.
The Fourth International Conference on Neuroesthetics, \"Empathy in the Brain and in Art,\" took place on 15 January 2005 at the University of California at Berkeley. Further information can be found at <http://plaisir.berkeley.edu/>.
Physiologist Vittorio Gallese (University of Parma) prompted many nods of satisfaction from the audience with his findings of activity in areas of the brain controlling movement when people simply watched the actions of others (see also the Research Article by Iacoboni et al. in this issue of *PLoS Biology* \[DOI: [10.1371/journal. pbio.0030079](10.1371/journal.pbio.0030079) \]). Susan Langer, in her book *Mind: An Essay on Human Feeling*, has defined empathy as the direct physical reaction inherent in the perception of others, an *involuntary* breach of individual separateness, and to see the neural resonance, to see that the same activity patterns were being recreated in actor and observer, was to give substance to the intuition of empathy.
Themed meetings, particularly when the theme does not conform to one discipline, are hard to pull off. It can be nearly impossible to convince successful professionals on the lecture circuit to modify the presentation of their own work to support such a theme. In that respect, this meeting was no different from many---some speakers were hard-pressed to conform to the theme, and it is not clear that many attendees learned information of practical value to their work from speakers across disciplines.
However, it is not often that scientists have the luxury of stepping back and appreciating the context of their work in quite this way. It is not, for instance, usually appropriate to begin a paper on an apoptotic signaling pathway with a philosophical digression into the nature of Death. The abstract dimensions that the visual neuroscientist Alice O\'Toole (University of Texas at Dallas) gave to facial characteristics are supposed to shed light on how we instantaneously recognize the friend we have not seen in 30 years. The electrophysiological signals in the brain that neurophysiologist Aina Puce (West Virginia University) described when we view simple movements is ultimately meant to explain how we identify with the subtle shrugging of shoulders that can transmute insouciance into insecurity.
By reducing the problem to its simplest, most controlled form, scientists hope to shed light on the complexities of life. Auditory physiologists are supposed to tell us how we hear. And yet it will be a long time before they can explain "music heard so deeply that it is not heard at all, but you are the music while the music lasts" (T. S. Eliot, as quoted by the conference organizer, Semir Zeki \[University College London\]). But the richness of the goal makes the journey all the more rewarding.
Hemai Parthasarathy is a senior editor for *PLoS Biology*. E-mail: <hemai@plos.org>
|
PubMed Central
|
2024-06-05T03:55:54.992131
|
2005-3-15
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065709/",
"journal": "PLoS Biol. 2005 Mar 15; 3(3):e105",
"authors": [
{
"first": "Hemai",
"last": "Parthasarathy"
}
]
}
|
PMC1065718
|
[10.1371/journal.pbio.0020356](10.1371/journal.pbio.0020356)
In *PLoS Biology,* volume 2, issue 11.
Figure 3C should have appeared as shown below. The GFP-AD-Cterm (3x Thr) and GFP-AD-Cterm (3x Asp) constructs are slightly less soluble than their headless counterparts. This may be due to the fact that they are somewhat more prone to aggregation over time than the headless proteins. The change does not affect the conclusions of the paper.
This correction note may be found online at DOI: [10.1371/journal.pbio.0030119](10.1371/journal.pbio.0030119). Published March 15, 2005
Citation: (2005) Correction: Dictyostelium myosin bipolar thick filament formation: Importance of charge and specific domains of the myosin rod. PLoS Biol 3(3): e119.
|
PubMed Central
|
2024-06-05T03:55:54.993066
|
2005-3-15
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1065718/",
"journal": "PLoS Biol. 2005 Mar 15; 3(3):e119",
"authors": [
{
"first": "Daniel",
"last": "Hostetter"
},
{
"first": "Sarah",
"last": "Rice"
},
{
"first": "Sara",
"last": "Dean"
},
{
"first": "David",
"last": "Altman"
},
{
"first": "Peggy M",
"last": "McMahon"
},
{
"first": "Shirley",
"last": "Sutton"
},
{
"first": "Ashutosh",
"last": "Tripathy"
},
{
"first": "James A",
"last": "Spudich"
}
]
}
|
PMC1069645
|
Introduction {#s1}
============
Molecular switches have been implicated in many types of cell-biological processes including the storage of decisions about cell fate \[[@pbio-0030107-b01]\], genetic control \[[@pbio-0030107-b02]\], and memory storage in the brain \[[@pbio-0030107-b03]\]. The mechanisms of such switches generally depend on some kind of autocatalytic process. If a switch is composed of a small number of molecules, stochastic fluctuations are significant and a deterministic description is not sufficient \[[@pbio-0030107-b04]\]. Because of the dynamic interaction of opposing reactions, such fluctuations can spontaneously reset the state of a switch. Reset events of this kind impose a temporal limit on the usefulness of the switch for information storage. It is thus crucial to understand the factors that control switch stability and to develop quantitative insight into how the stability required for a particular biological process could be achieved. The stability problem of switches has so far been studied primarily in relation to genetic switches \[[@pbio-0030107-b05],[@pbio-0030107-b06],[@pbio-0030107-b07],[@pbio-0030107-b08],[@pbio-0030107-b09]\].
The problem of switch stability is of particular relevance to synaptic function \[[@pbio-0030107-b10],[@pbio-0030107-b11]\] since memory is thought to be encoded by changes in synaptic strength \[[@pbio-0030107-b12]\] and because there are indications that synaptic strength is controlled by molecular switches \[[@pbio-0030107-b13],[@pbio-0030107-b14],[@pbio-0030107-b15]\]. By a molecular switch, we mean a molecule or a small group of molecules that can undergo a persistent change in state. In our definition, the change in state occurs in a discrete rather than a smoothly graded way. Clearly, spontaneous reset of a synaptic switch that encodes memory would be problematic because it would lead to loss of the stored memory. The fact that at least some memories persist for a human lifetime indicates that storage processes of extraordinary stability are present.
The mechanisms that underlie synaptic information storage are beginning to be elucidated \[[@pbio-0030107-b16]\]. It has been demonstrated that brief periods of strong stimulation can lead to an increase in the strength of synapses, a process termed long-term potentiation (LTP) \[[@pbio-0030107-b17]\]. In vivo studies show that LTP can persist for at least a year \[[@pbio-0030107-b18]\]. The initiation of LTP is caused by activation of [N]{.smallcaps}-methyl-[D]{.smallcaps}-aspartate (NMDA) channels and elevation of intracellular calcium (Ca^2+^) concentration \[[@pbio-0030107-b19],[@pbio-0030107-b20]\]. There is general agreement that the resulting activation of calcium/calmodulin-dependent protein kinase II (CaMKII) plays a critical role in LTP (reviewed in \[[@pbio-0030107-b03]\]). CaMKII activation is persistent \[[@pbio-0030107-b21]\], is required for LTP \[[@pbio-0030107-b22],[@pbio-0030107-b23],[@pbio-0030107-b24]\], and is sufficient by itself to produce potentiation \[[@pbio-0030107-b25]\]. Genetic modification of CaMKII that prevents its sustained activation prevents long-term memory, as defined in behavioral tests \[[@pbio-0030107-b24]\]. The possibility that CaMKII is a synaptic memory molecule is further strengthened by the finding that it has autocatalytic properties that would allow it to function as a molecular switch \[[@pbio-0030107-b10],[@pbio-0030107-b26],[@pbio-0030107-b27]\]. Although CaMKII is required for long-term synaptic modification and is therefore a strong candidate as a memory molecule, whether evidence that its persistent activation is necessary for the maintenance of LTP remains an open question \[23\].
In a previous analysis of the CaMKII switch, Zhabotinsky and Lisman \[[@pbio-0030107-b28],[@pbio-0030107-b29]\] proposed a model that incorporated many key biochemical properties of CaMKII holoenzymes and the phosphatase-1 (PP1) enzymes that dephosphorylate them \[[@pbio-0030107-b30],[@pbio-0030107-b31]\]. It was shown that an interplay between autophosphorylation of CaMKII holoenzymes and dephosphorylation by PP1 molecules can give rise to two stable states of phosphorylation at basal levels of free Ca^2+^. Therefore, a transient input of high Ca^2+^ (such as during the stimulation protocol used in LTP induction) can switch the system from an unphosphorylated (DOWN) state to a persistent, highly phosphorylated (UP) state. Such a persistent change in activation of CaMKII following LTP induction could underlie the persistent change in synaptic strength. In these previous modeling efforts, chemical reactions were described deterministically by the law of mass action, precluding estimates of the switch stability. The need for considering the limits on stability imposed by stochastic fluctuations is made more urgent by recent measurements showing that the number of CaMKII holoenzymes in the postsynaptic density (PSD) of a single synapse is relatively small \[[@pbio-0030107-b32]\]. For a typical PSD, there are about 30 holoenzymes \[[@pbio-0030107-b32]\]. In the absence of a theory that relates switch stability to the number of switch molecules, the implications of this finding are unclear. The current work addresses this issue using Monte Carlo simulations of the stochastic chemical reactions in the CaMKII/PP1 system. This approach allows us to estimate quantitatively the stability (lifetime) of a CaMKII switch and its dependence on the number of molecules. Our results thus provide new information about the potential for the CaMKII switch within the PSD to store long-term memories.
A second goal of our work is to analyze the impact of molecular turnover on switch stability. Because biological switches are themselves composed of molecules that are unstable, turnover must occur. Such turnover is likely to have a detrimental effect on switch stability \[[@pbio-0030107-b33]\]. However, turnover need not necessarily lead to switch reset since new molecules may adopt a state that is dependent on the state of the other molecules in the switch \[[@pbio-0030107-b33],[@pbio-0030107-b34]\]. Specifically, when the CaMKII switch is in the UP state, the PP1 activity should be saturated. This saturation reduces the effectiveness of the phosphatase so that when a phosphorylated holoenyzme is replaced by a newly synthesized unphosphorylated one, the new holoenzyme will become phosphorylated as a result of the autophosphorylation even at basal Ca^2+^ levels. This can restore the state of the switch that was present before the turnover event. Direct measurements show that the CaMKII at synapses turns over about once per day \[[@pbio-0030107-b35]\], a timescale much shorter than synaptic memory. However, no theory has been developed for any type of molecular switch that allows an estimation of how this turnover quantitatively affects stability. Here we examine this issue with regard to the CaMKII switch. Our findings reveal general principles with implications for other kinds of molecular switches.
Results {#s2}
=======
Autocatalysis Leads to Bistability {#s2a}
----------------------------------
To understand the effect of stochastic fluctuations in molecular switches, we have implemented simulations of the CaMKII/PP1 switch model \[[@pbio-0030107-b28],[@pbio-0030107-b29]\]. In this implementation, reactions are modeled stochastically using Monte Carlo methods and the number of CaMKII and PP1 molecules that are individually considered is comparable to the numbers contained within the PSD at single synapses.
A CaMKII holoenzyme is composed of two rings, each with six kinase subunits. Each subunit has a single phosphorylation site at Thr286/287 that, when phosphorylated, makes the subunit active even when Ca^2+^/calmodulin is no longer bound. Autophosphorylation of the site on a given "substrate" subunit proceeds if two necessary conditions are fulfilled \[[@pbio-0030107-b36]\]. Ca^2+^/calmodulin must bind to the "substrate" subunit in order to reveal its Thr286/287 site. Also, the counterclockwise neighboring "catalyst" subunit must be active. Hence, the initial autophosphorylation necessary to switch a ring "on" requires the binding of two molecules of Ca^2+^/calmodulin. Subsequent phosphorylation of other subunits within a ring is faster (see [Figure 1](#pbio-0030107-g001){ref-type="fig"}A) since the phosphorylated subunit is constitutively active without Ca^2+^/calmodulin. Thus, only a single Ca^2+^/calmodulin is required to phosphorylate a "substrate" subunit if its counterclockwise "catalyst" neighbor is already phosphorylated. (Note that our results are unaffected by the direction of autophosphorylation, but based on geometric considerations, we assume it is asymmetric \[[@pbio-0030107-b27],[@pbio-0030107-b37]\].) At the resting Ca^2+^ concentration, with our standard parameters, the initial autophosphorylation occurs at an average rate of one per 3.5 h per unphosphorylated ring, while the further phosphorylation steps occur at approximately one per 4 min per available "substrate" subunit. We assume that the molecules of PP1 held in the PSD can dephosphorylate any of the sites on any of the holoenzymes in the PSD. Furthermore, PP1 becomes saturated when the kinase becomes hyperphosphorylated \[[@pbio-0030107-b29]\].
::: {#pbio-0030107-g001 .fig}
Figure 1
::: {.caption}
###### CaMKII Autophosphorylation and PP1 Saturation Lead to Bistability in the Phosphorylation States of CaMKII at Resting Ca^2+^
\(A) Phosphorylation of the first subunit is slow at resting Ca^2+^ because of the requirement for two Ca^2+^/calmodulin molecules (see text). Subsequent phosphorylation is faster because only one Ca^2+^/calmodulin is required.
\(B) Schematic figure indicating how bistability arises from the dependence of phosphorylation and dephosphorylation rates on the number of subunits phosphorylated. Stable states are at the left (DOWN) and right (UP) intersections of the two curves. The middle crossing is unstable. The greater the area of the shaded region between a stable steady state and the unstable steady state, the harder it is for fluctuations to destabilize that stable steady state (the larger their basins of attraction).
\(C) A 2-s pulse of high Ca^2+^ switches the system (with 16 holoenzymes) from a low state of phosphorylation to a higher state within the basin of attraction of the UP state (see \[D\]). Phosphorylation fraction is *Sp~tot~*/12*N~CaMK~*.
\(D) After the end of the Ca^2+^ pulse, it can take tens of minutes for the system to reach the dynamic equilibrium in the UP state.
\(E) The UP state is stable for many years in a system with 16 holoenzymes.
:::
:::
[Figure 1](#pbio-0030107-g001){ref-type="fig"}B indicates schematically how the total rates of autophosphorylation and dephosphorylation lead to two stable states at the resting Ca^2+^ level. The curves in the figure show how these reactions vary as a function of the total number of sites phosphorylated. The intersection points (where dephosphorylation balances phosphorylation) define three steady states, of which the left and right ones are stable and the middle one is unstable. Switching can occur if phosphorylation of the system is forced (either by a transient signal or by a spontaneous fluctuation) far enough away from one stable value that it passes the unstable value; the system will then fall into the basin of attraction of the second stable point. Once in this basin, the intrinsic dynamics of the system set the timescale of drift to the second stable state.
The model presented here includes stochastic turnover of CaMKII holoenzymes with an average lifetime of 30 h that is independent of phosphorylation level, as experimentally determined \[[@pbio-0030107-b35]\]. We assume, except where stated otherwise, that once a holoenzyme is removed, it is immediately replaced by an unphosphorylated holoenzyme. If the switch is in the DOWN state, the newly inserted holoenzyme is likely to stay off: any spontaneous phosphorylation will be countered by dephosphorylation, which removes a subunit in approximately 5 min, on average. However, if the switch is in the UP state, the phosphatase is so saturated by other phosphorylated holoenzymes, that the average time for a newly inserted subunit to be dephosphorylated is almost 1 h. Hence, the newly inserted holoenzyme can turn on as a result of the activation of the kinase by basal Ca^2+^ levels \[[@pbio-0030107-b28],[@pbio-0030107-b29]\], since the time for it to be phosphorylated is significantly less than the time for turnover. Thus, the UP state of the switch is stable, in spite of turnover.
Parameters for the model were constrained according to the references cited in [Table 1](#pbio-0030107-t001){ref-type="table"}. Where the constraints allowed a range of variation, we chose values of parameters so that the system would be bistable and have approximately equal lifetimes of the UP and DOWN states (see below). We required that our standard system have equal numbers of CaMKII holoenzymes and PP1 molecules, as their concentrations are known to be similar, but adjusted the less well-determined Hill constants, *Κ~H1~* and *Κ~H2~,* to maximize system lifetime (see below). While our model is more sensitive to the value of *Κ~H1~* than any other parameter, bistability still exists in a significant range (10% around its optimal value) when all other parameters are fixed. Compensatory covariation of other parameters maintains the system\'s bistability at fixed calcium when *Κ~H1~* varies by more than a factor of three. Our study aims to test whether a plausible set of kinetic parameters enables a molecular switch, with a small number of participating molecules, to be stable in spite of fluctuations.
::: {#pbio-0030107-t001 .table-wrap}
Table 1
::: {.caption}
###### Parameters Used in the Model
:::
PKA, cAMP-dependent protein kinase
:::
Two critical requirements exist for the switch to function. First, the initial (P0 to P1) phosphorylation step must be significantly slower than further phosphorylation steps. This is true as the Hill constant for Ca^2+^ activation of CaMKII \[[@pbio-0030107-b38],[@pbio-0030107-b39]\] is significantly greater than the average Ca^2+^ concentration in the physiological resting state (0.7 μM versus 0.1 μM in most of this paper). Second, the phosphatase activity must saturate so the rate of dephosphorylation per phosphorylated subunit is significantly slower in the UP state than in the DOWN state. This is achieved as the Michaelis constant, *Κ~M~,* of PP1 is much lower than the concentration of CaMKII subunits (0.4 μM versus 400 μM in this paper, though we also test the model with a *Κ~M~* of 10 μM \[[@pbio-0030107-b40]\]). Thus, according to the latest experimental data, the two critical requirements for a functioning switch are met.
[Figure 1](#pbio-0030107-g001){ref-type="fig"}C and [1](#pbio-0030107-g001){ref-type="fig"}D (with 16 holoenzymes simulated) show that a large, 2-s-long Ca^2+^ elevation of the kind that may occur during LTP induction \[[@pbio-0030107-b41]\] can switch the system from the unphosphorylated DOWN state to a highly phosphorylated, persistent UP state. The system drifts toward the UP state even after Ca^2+^ falls to its basal level ([Figure 1](#pbio-0030107-g001){ref-type="fig"}D). Such a drift has been observed experimentally \[[@pbio-0030107-b42]\]. During the drift period, which can take tens of minutes, the system would be more vulnerable to depotentiation. In this particular example (with 16 holoenzymes), the stable UP state is reached in under an hour ([Figure 1](#pbio-0030107-g001){ref-type="fig"}D). As seen in [Figure 1](#pbio-0030107-g001){ref-type="fig"}E, the resulting UP state remains stable for at least 10 y.
Spontaneous Transitions between the Baseline and Memory States {#s2b}
--------------------------------------------------------------
We examined the distribution of times between spontaneous switching events as a measure of the stability of memory storage. [Figure 2](#pbio-0030107-g002){ref-type="fig"}A shows that the total phosphorylation level of a small system with eight holoenzymes is only stable on the time scale of months, not years; sporadically, fluctuations cause the system to change from one state to the other, as indicated by the random switching of phosphorylation level. Analysis reveals that the times spent in either the DOWN or UP state (i.e., the lifetimes) are distributed exponentially ([Figure 2](#pbio-0030107-g002){ref-type="fig"}B) (apart from brief transition times). Such an exponential distribution \[[@pbio-0030107-b43]\] indicates that the probability of transition per unit time is constant for a given state. The exponential distribution of lifetimes has a characteristic time constant that equals the average lifetime in the state (and the inverse of the probability of transition per unit time).
::: {#pbio-0030107-g002 .fig}
Figure 2
::: {.caption}
###### Switch Stability Is a Trade-Off between Lifetimes of UP and DOWN States
\(A) Spontaneous switching between UP and DOWN states in a system with eight CaMKII holoenzymes.
\(B) The distribution of lifetimes between switching events is exponential, as demonstrated by the straight line fit for lifetimes of the UP state on a semi-logarithmic scale.
(C and D) Removing one PP1 enzyme (seven instead of eight) (C) leads to a longer lifetime for the UP state but shorter lifetime for the DOWN state; whereas adding one PP1 enzyme (nine instead of eight) (D) yields the opposite effect.
\(E) Dependence of average lifetimes of the UP state (squares) and DOWN state (circles) as a function of the number of PP1 enzymes (with eight CaMKII holoenzymes). Filled blue symbols correspond to data points from (A), (C), and (D). Lines are approximate, analytic results (based on [Materials and Methods](#s4){ref-type="sec"} and \[[@pbio-0030107-b51]\]). The optimal lifetime of the switch is defined by the intersection point of the two curves, at which the lifetimes of the UP and DOWN states are equal. Blue indicates reference parameters. Red indicates *k* ~1~ = 0.75 s^−1^, one-half of the standard value. The lifetime does depend on kinetic parameters, but maximum stability is approximately the same, albeit with a different number of PP1 molecules.
:::
:::
The overall stability of the system is dependent on the lifetimes of both the UP and DOWN states. In general, any change in the system that increases the rate of phosphorylation tends to increase the lifetime of the UP state while reducing the lifetime of the DOWN state. The opposite is true for an increase in dephosphorylation rate. This tradeoff between lifetimes of the two states can be demonstrated by varying the number of phosphatase enzymes in the system while all other parameters are fixed. As seen in [Figure 2](#pbio-0030107-g002){ref-type="fig"}C, a reduction in the amount of phosphatase resulted in an increased lifetime of the UP state, but destabilization of the DOWN state; in contrast, increasing the amount of phosphatase had the opposite effect ([Figure 2](#pbio-0030107-g002){ref-type="fig"}D). [Figure 2](#pbio-0030107-g002){ref-type="fig"}E (blue) shows the lifetimes averaged over 400 transitions, as a function of the number of phosphatase enzymes. Again, the lifetime of the UP state decreases, and the lifetime of the DOWN state increases with the number of PP1 molecules. We define the "system\'s lifetime" as the smaller of the two lifetimes, because we assume that a random potentiation of a synapse is as equally undesirable as a random depotentiation; a spontaneous transition from either state would be detrimental for memory. It follows that the system is optimal at the crossing of the two curves ([Figure 2](#pbio-0030107-g002){ref-type="fig"}E), where the two lifetimes are equal (so that neither lifetime is too small). Such an optimum corresponds to a balance between the processes of phosphorylation and dephosphorylation. We therefore define the phosphatase concentration at which the two curves cross to achieve balance as the optimal concentration.
In the simulations described above, we set the optimal phosphatase concentration equal to the concentration of the CaMKII holoenzymes (R. J. Colbran, personal communication; see [Table 1](#pbio-0030107-t001){ref-type="table"}). We adjusted the less well-constrained parameters to achieve this. To see how sensitive the system\'s lifetime is to these particular choices, we simulated the system with a different value of the phosphorylation rate constant for the kinase, changing *k* ~1~ from 1.5 s^−1^ to 0.75 s^−1^. [Figure 2](#pbio-0030107-g002){ref-type="fig"}E (red) shows that the optimal phosphatase concentration is also reduced, but at this concentration the system lifetime remains as high as in the original system (the intersection of the two curves is shifted but remains at the same lifetime). We also tested the system\'s robustness to a 25-fold larger value, 10 μM, for *Κ~M~*. With an increased optimal phosphatase concentration, the system of 20 holoenzymes was stable for over 10 y. Thus, achieving long lifetimes does not require a specific level of enzyme activity, but does require an appropriate balance between phosphorylation and dephosphorylation rates. See Discussion for how this might be achieved.
Stability Increases Exponentially with the Number of Molecules {#s2c}
--------------------------------------------------------------
We next considered how the system\'s lifetime varies with the number of holoenzymes (while PP1 varies in proportion, as does the system\'s volume). [Figure 3](#pbio-0030107-g003){ref-type="fig"}A and [3](#pbio-0030107-g003){ref-type="fig"}B show the behavior of the model switch with four and 16 holoenzymes, respectively. We found ([Figure 3](#pbio-0030107-g003){ref-type="fig"}C) that the system stability increases exponentially with the number of holoenzymes (i.e., the system size). As a result of this exponential dependence, the lifetime of the system almost doubles for each additional holoenzyme in the PSD. These simulations show that a switch made of only four holoenzymes can only be expected to have stability on the order of days to weeks, whereas increasing the system to 16 holoenzymes could result in a switch that is stable for a human lifetime.
::: {#pbio-0030107-g003 .fig}
Figure 3
::: {.caption}
###### Stability of the Switch Increases Exponentially with System Size (Number of CaMKII Holoenzymes and PP1 Molecules)
\(A) Spontaneous transitions in a system with four CaMKII holoenzymes and four PP1 enzymes. Phosphorylation fraction is *Sp~tot~*/12*N~CaMK~*.
\(B) System with 16 holoenzymes and 16 PP1 enzymes, with four times the volume of (A). Note different timescales between (A) and (B). Phosphorylation fraction is *Sp~tot~*/12*N~CaMK~*.
\(C) The switch\'s lifetime increases exponentially with system size. Numbers of all species scale together with system volume. Circles are data points, line is a linear fit, indicating an exponential dependence, because the ordinate is in logarithmic scale. The red asterisks indicate data points where we included the PP1--I1P fluctuations explicitly.
:::
:::
Protein Turnover Limits Memory Lifetime {#s2d}
---------------------------------------
In order to understand more deeply the cause of spontaneous switching, we examined what was occurring in the switch during the period preceding switching events. The two examples in [Figure 4](#pbio-0030107-g004){ref-type="fig"}A and [4](#pbio-0030107-g004){ref-type="fig"}B show the total instantaneous level of phosphorylation of the system during the time preceding a spontaneous transition to the DOWN state ([Figure 4](#pbio-0030107-g004){ref-type="fig"}A is for a system of four holoenzymes, [Figure 4](#pbio-0030107-g004){ref-type="fig"}B for a system of 16). Holoenzyme turnover events are evident in these traces, because a turnover event causes an abrupt drop in the level of phosphorylation (marked by red arrows). In [Figure 4](#pbio-0030107-g004){ref-type="fig"}A, four turnover events occur in a 3-h period prior to a downward switching event, and in [Figure 4](#pbio-0030107-g004){ref-type="fig"}B, six turnover events occur in the same length of time (*t* = −7 h to −4 h), after which intrinsic dynamics take over to complete the downward transition. Based on the average time for turnover of 30 h per holoenzyme, one would expect a turnover event every 7.5 h in a system with four holoenzymes, and every 1.9 h in a system with 16 holoenzymes. Hence, the figures indicate that high numbers of turnover events occur in the periods before a transition, as expected if turnover initiates switching to the DOWN state (as explained on the next page). Such high amounts of turnover result from the stochastic nature of the turnover process, and occurred prior to such spontaneous transitions in all the traces we examined. Hence, protein turnover, and, in particular, its stochastic nature, strongly affects the system\'s ability to store information.
::: {#pbio-0030107-g004 .fig}
Figure 4
::: {.caption}
###### The Rate of Protein Turnover Limits the Maximal Lifetime of the System and Leads to a Minimal Rate of Energy-Consuming Activity
(A and B) Stochastic changes in total phosphorylation during a transition from the UP to DOWN state, with turnover events marked by arrows, in a system with (A) four CaMKII holoenzymes and (B) 16 CaMKII holoenzymes. Red arrows indicate turnover events, which cause an abrupt drop in the level of phosphorylation.
\(C) Log--log plot of lifetime of the switch as a function of turnover rate for the system with eight CaMKII holoenzymes (the red asterisk marks 30-h turnover, used as standard in this paper).
\(D) The rate for an individual ring of subunits to switch off as a function of the total number of rings that are on (shown here for a system with eight CaMKII holoenzymes). As more rings are turned on, the phosphatase activity saturates and the equilibrium level of phosphorylation per ring increases. As a result, the switching-off rate for rings in the UP state for the system approaches the turnover rate (dashed line), because the probability of total ring dephosphorylation by PP1 becomes small.
\(E) Dynamic equilibrium between turnover (vertical solid black lines) and switching on of rings (colored step-like lines) when the system is in the UP state. At the time of the first turnover, five rings are already unphosphorylated by prior turnover. The system is stable because the rate of rings switching on matches the rate of turnover of phosphorylated rings (each turnover event can result in the loss of zero, one, or two phosphorylated rings). A system with 20 holoenzymes is shown.
:::
:::
We next investigated how the rate of protein turnover,*ν~T~,* affects the switch\'s stability. We found that an increase in the rate of protein turnover has little effect on the lifetime of the DOWN state, but dramatically reduces the lifetime of the UP state. Again, this is to be expected if turnover is responsible for initiating a switch DOWN in the system. Since the system is optimal when the two lifetimes are similar, in a series of simulations where we used different rates for protein turnover, we also adjusted the amount of PP1 to return to an optimal system (where UP and DOWN state lifetimes are similar). Hence, we can plot in [Figure 4](#pbio-0030107-g004){ref-type="fig"}C the optimal lifetime of the switch as a function of average time for protein turnover. For turnover times of less than 1 mo, the optimal system stability is strongly dependent on the rate of turnover, suggesting that protein turnover is a limiting factor in the stability of the switch. Indeed, making the turnover rate very fast (hourly) can cause the system to lose all bistability.
In the UP state of our system, protein turnover replaces phosphorylated holoenzymes with unphosphorylated ones and is thus effectively acting like a phosphatase. It was therefore of interest to compare this effective phosphatase activity to the rate of dephosphorylation produced by the phosphatase itself. We proceeded by calculating the total rate for individual phosphorylated rings to switch off, that is, to become an unphosphorylated ring in state P0. Such a switching-off rate is the sum of the turnover rate and the rate for dephosphorylation by phosphatase. Our approximate calculation is accurate when the switching-on and switching-off rates for a ring are much slower than the rates for individual subunits to be phosphorylated or dephosphorylated, as it assumes a ring has time to reach all configurations of phosphorylation before it switches off (see [Materials and Methods](#s4){ref-type="sec"}). Given that assumption, we obtained the proportion of time a ring that is on spends in each configuration of phosphorylation. The unphosphorylated state, P0, where a ring is off, can be reached either by turnover, or by dephosphorylation from the state with a single phosphorylated site, P1. Hence, the average rate at which a ring switches off ([Figure 4](#pbio-0030107-g004){ref-type="fig"}D) is given by *ν* ~3~ *Ρ~Ρ~* ~1~ *+ ν~T~,* where *ν* ~3~ is the rate per phosphorylated subunit of phosphatase activity (*ν* ~3~ of [equation 8](#pbio-0030107-e008){ref-type="disp-formula"}), *Ρ~Ρ~* ~1~ is the proportion of time a ring that is on spends in the configuration P1 and *ν~T~* is the protein turnover rate. Note that as the number of rings switched on increases, so the total phosphorylation of the system increases, causing both *ν* ~3~ and *Ρ~Ρ~* ~1~ to decrease. As is evident in [Figure 4](#pbio-0030107-g004){ref-type="fig"}D, when more than half the rings in the system are on, the rate of rings switching off becomes identical to the turnover rate, *ν~T~,* itself (horizontal dashed line in [Figure 4](#pbio-0030107-g004){ref-type="fig"}D). Hence, with a 30-h turnover rate and the optimal concentration of phosphatase, the phosphatase is unable to switch a ring off in the UP state; loss of phosphorylated rings is purely due to turnover.
We next sought to visualize how the system remains in the UP state even while turnover is causing the replacement of approximately two-thirds of the phosphorylated holoenzymes with unphosphorylated ones during a 30-h period (on average). In [Figure 4](#pbio-0030107-g004){ref-type="fig"}E, we present a snapshot of a few hours of activity to show the time course for turnover and rephosphorylation of individual rings. The system in [Figure 4](#pbio-0030107-g004){ref-type="fig"}E contains 20 holoenzymes, so the UP state is stable for many decades. Turnover events are marked by vertical lines. Unphosphorylated rings that become phosphorylated are indicated by the colored step-like lines, where each step indicates the phosphorylation of a subunit. It should be noted that at any one time, even though the system is in the UP state, a number of rings are unphosphorylated because of previous turnover. The rate of switching on for rings is proportional to the number of rings off. On average, the total rate of switching on matches the rate of phosphorylated rings lost by turnover. This dynamic equilibrium between the switching on of rings and turnover determines the average number of unphosphorylated rings at any time. In the system of 20 holoenzymes (and therefore 40 rings), the number of unphosphorylated rings typically varies from four to eight. The number is five before the first turnover event on [Figure 4](#pbio-0030107-g004){ref-type="fig"}E and reaches a maximum of nine following the two closely spaced turnover events to the right in [Figure 4](#pbio-0030107-g004){ref-type="fig"}E. The figure also illustrates how once one subunit is phosphorylated on a ring, the others rapidly follow.
Effect of Fluctuations in Reactant Concentrations {#s2e}
-------------------------------------------------
In our simulations thus far, we have not considered noise in the signaling pathways that control the kinase and phosphatase reactions. A careful analysis of this issue requires an understanding of the signals that lead to bidirectional synaptic modification, as well as a consideration of the noise reduction mechanisms; both are beyond the scope of this paper. Nevertheless, we wanted to determine whether the switch we have modeled could tolerate reasonable noise levels in its input. In this class of models, reaction rates are nonlinear in Ca^2+^ concentration, so fluctuations in Ca^2+^ concentration affect the mean reaction rate, as well as providing additional noise about the mean rate. Moreover, the functional dependence on Ca^2+^ is not the same for all reaction steps. In particular, fluctuations in Ca^2+^ concentration increase the average rate of the slow initial (P0 to P1) phosphorylation step, which requires two Ca^2+^/calmodulins, to a greater extent than any other reaction steps. The change in relative reaction rates means that, in principle, large enough fluctuations can destroy the bistability altogether, whatever the system size. This class of switch has no absolute protection against Ca^2+^ fluctuations---indeed, we require in our model that a strong enough change in Ca^2+^, as occurs during LTP, leads to a switch from the DOWN to the UP state (see [Figure 1](#pbio-0030107-g001){ref-type="fig"}C). However, the system should be robust to smaller, realistic fluctuations that occur in the absence of LTP. [Figure 5](#pbio-0030107-g005){ref-type="fig"}A (blue squares) shows that the system with our standard parameters (but with a lower, 0.07 μM, baseline) is able to tolerate the moderate fluctuations that might arise from Ca^2+^ influx through NMDA-receptor-mediated channels (0.1 μM amplitude with 100 ms decay time; \[[@pbio-0030107-b44]\]) due to spontaneous presynaptic action potentials (a 0.5-Hz Poisson train) with only modest reduction in stability. Although the lifetime increases less steeply with system size than in the case without fluctuations ([Figure 5](#pbio-0030107-g005){ref-type="fig"}A, black circles), extrapolation suggests that a system with 20 holoenzymes would be sufficient to have a lifetime of 100 y. If we use larger fluctuations, of amplitude 1.0 μM, it is important to adjust parameters to compensate for the change in average activity produced by fluctuations. With such an adjustment (see [Materials and Methods](#s4){ref-type="sec"}) the system with 1.0-μM fluctuations in free Ca^2+^ concentration (with 100 ms decay time, above a 0.1**-**μM baseline in a 0.5-Hz Poisson train) is slightly more stable than the original system without fluctuating Ca^2+^ ([Figure 5](#pbio-0030107-g005){ref-type="fig"}A, red triangles). We conclude that plausible levels of Ca^2+^ fluctuations in spines do not necessarily compromise switch stability.
::: {#pbio-0030107-g005 .fig}
Figure 5
::: {.caption}
###### Switch Stability in the Presence of Spontaneous Fluctuations in Free Calcium Concentration and the Total Number of Enzymes
\(A) Effect of Ca^2+^ fluctuations on stability (lifetime) as a function of number of CaMKII holoenzymes. Circles indicate the original system without Ca^2+^ fluctuations. Squares indicate the original system with free Ca^2+^ fluctuations of amplitude 0.1 μM and baseline 0.07 μM. Triangles indicate adjusted system with free Ca^2+^ fluctuations of amplitude 1.0 μM and baseline 0.1 μM. The adjusted system has alternative parameters, such that *N~PP~* ~1~ = *N~CaMK~*/2, *k* ~1~ = 6 s^−1^, *k* ~2~ = 7 s^−1^, and *K~H~* ~1~ = 4.0 μM. The ordinate is in logarithmic scale.
\(B) Effect of fluctuations in the number of PP1 molecules on the lifetime of UP states (squares/solid line) and DOWN states (circles/dashed line) (16 holoenzymes). The timescale on the abcissa is the average time for the number of PP1 molecules to increase or decrease by one. Red indicates 12 \< *N~PP~* ~1~ \< 20. Blue indicates 8 \< *N~PP~* ~1~ \< 24. The ordinate is in logarithmic scale.
\(C) Lifetime of UP states (squares/solid line) and DOWN states (circles/dashed line) when the number of holoenzymes and PP1 molecules fluctuate in the respective ranges 14 \< *N~CaMK~* \< 18 and 8 \< *N~PP~* ~1~ \< 24. The timescale for PP1 fluctuations varies along the abcissa. The timescale for CaMKII fluctuations is fixed by the turnover rate (30 h per holoenzyme). The ordinate is in logarithmic scale.
:::
:::
In our simulations so far, we have assumed fixed numbers of enzyme molecules, but in principle these numbers may fluctuate with time, potentially compromising stability. We implemented several sets of simulations to address this issue. In each set, we carried out a number of simulations corresponding to a range of timescales for fluctuations of PP1. Specifically, we varied the average time between random steps of plus one or minus one in the number of PP1 molecules in the PSD, and plotted this time as the x-axis in [Figure 5](#pbio-0030107-g005){ref-type="fig"}B and [5](#pbio-0030107-g005){ref-type="fig"}C. It should be noted that the total time the system spends away from its optimum is usually the sum of many such time steps. In the first set of simulations, with the number of CaMKII holoenzymes held fixed, we assumed the number of PP1 molecules could fluctuate by plus or minus 25%, so in the particular case of 16 holoenzymes, the number of PP1 molecules varies between 12 and 20. The resulting lifetimes for the UP and DOWN states are given in [Figure 5](#pbio-0030107-g005){ref-type="fig"}B in red. The lifetime decreases as the average time between changes in PP1 increases. This is because slow fluctuations lead to long periods of time when the system is far from optimal. In contrast, if the fluctuations are rapid, the system may not have time to make a transition even if the system loses bistability temporarily. Still, even with a change every 6 h, the system with 16 holoenzymes is stable for over about 20 y. Second, we increased the amplitude of fluctuations in PP1 to plus or minus 50% (a variation of a factor of three, from eight to 24 in this case). The simulation results in [Figure 5](#pbio-0030107-g005){ref-type="fig"}B (blue) indicate that such an increase in amplitude of fluctuation causes the average lifetime for the system to decrease. Again, if the fluctuations are relatively rapid, they do not seriously degrade the switch. It is somewhat remarkable that when the number of PP1 molecules varies by a factor of three over a timescale of tens of minutes, the switch lifetime still averages over 10 y. Third, we introduced slow fluctuations in the number of CaMKII holoenzymes, assuming stochastic insertion of holoenzymes as well as stochastic turnover. Since the timescale of removal is set at 30 h per holoenzyme \[[@pbio-0030107-b35]\], the average rate of insertion is fixed (at 16 every 30 h) to ensure the appropriate average of 16 holoenzymes within the PSD. As above, the simulations covered a range of timescales for the fluctuations of PP1, whose number could vary between 8 and 24. We found that variations in the number of holoenzymes are more deleterious than variations in PP1 alone. In particular, loss of holoenzymes from the PSD destabilizes the UP state. This is because the effect seen above, of a nonoptimal CaMKII to PP1 ratio, is exacerbated by a reduction in switch size when holoenzymes are lost (cf. [Figure 3](#pbio-0030107-g003){ref-type="fig"}C). [Figure 5](#pbio-0030107-g005){ref-type="fig"}C indicates the resulting lifetimes when the number of CaMKII holoenzymes varied between 14 and 18. Without PP1 fluctuations (equivalent to a time step of zero) the DOWN state is little affected by these slow fluctuations in the number of holoenzymes (circles/dashed line), but stability of the UP state is greatly reduced (squares/solid line). Including slow PP1 fluctuations of **±**50% reduces the lifetimes of both UP and DOWN states to below 10 y. Although a system averaging 20 holoenzymes would be more stable, we conclude that during turnover, a holoenzyme removed from the PSD needs to be replaced relatively rapidly---on a timescale of minutes, not hours---to avoid degradation of the switch. Moreover, if CaMKII were not anchored, but able to freely diffuse in and out of the PSD, fluctuations in the number of holoenzymes present would be increased, and bistability would not be possible \[[@pbio-0030107-b45],[@pbio-0030107-b46]\].
Discussion {#s3}
==========
In this paper, we have considered the stability against fluctuations of a bistable switch based on the interaction of CaMKII and PP1 in the PSD. Although a deterministic model of such a switch has been presented before \[[@pbio-0030107-b28],[@pbio-0030107-b29]\], it was not previously possible to assess quantitatively the potentiality of the switch for long-term information storage, because the rate of spontaneous reset was not known. Given the small number of CaMKII molecules at synapses \[[@pbio-0030107-b32]\], stochastic fluctuations in the reactions must necessarily lead to switch reset on some timescale. Our results show that that this timescale depends crucially on the number of molecules that make up the switch (see [Figure 3](#pbio-0030107-g003){ref-type="fig"}C). Indeed, this dependence is highly nonlinear, scaling exponentially with the number of molecules involved. We have shown that this timescale can exceed human lifetimes when the number of holoenzymes is greater than 15 (see [Figure 3](#pbio-0030107-g003){ref-type="fig"}B), provided the parameters of the system are in an optimal range. A substantially smaller number of holoenzymes, such as four, would result in spontaneous transitions on a timescale of a week (see [Figure 3](#pbio-0030107-g003){ref-type="fig"}A). One interesting possibility is that initially a small number of CaMKII holoenzymes is sufficient for the immediate information storage and with time, during an initial consolidation period \[[@pbio-0030107-b47],[@pbio-0030107-b48]\], a larger number of holoenzymes accumulate at the PSD, allowing for more permanent memory formation (see [Figure 3](#pbio-0030107-g003){ref-type="fig"}C). Our general conclusion is that relatively small groups of CaMKII molecules, such as are found in the PSD (where the average is 30 holoenzymes) can function as highly stable switches and could potentially subserve information storage for very long periods.
The exponential dependence of lifetime on system size is consistent with general theoretical considerations \[[@pbio-0030107-b11]\]. This is because the switch can be described as a biochemical system with a "double well" effective energy potential in which two minima are separated by a barrier. The fluctuations in the reactions generate noise-driven hopping over the barrier in a manner analogous to thermally driven hopping over a real potential barrier \[[@pbio-0030107-b11],[@pbio-0030107-b49],[@pbio-0030107-b50],[@pbio-0030107-b51]\]. The effective barrier height is proportional to the system size \[[@pbio-0030107-b49]\], and it is well-known that with a constant noise source, the time for transitions across a barrier increases exponentially with barrier height \[[@pbio-0030107-b51],[@pbio-0030107-b52]\]. Intuitively, a transition from the UP to DOWN state is triggered when a critical number of CaMKII rings (proportional to the system\'s size *N*) are dephosphorylated at the same time. In terms of probability, the decrease in likelihood with *N* is the same effect as the increase in expected number of coin tosses necessary to obtain *N* heads in a row, each with a probability *p* = 1/2. The probability for *N* consecutive heads is *p^N^*, and the expected time (number of coin tosses) it takes before this happens is (1/*p*)*^N^* = 2*^N^,* which grows exponentially with *N*. By analogy, the larger the system, the more rings of CaMKII have to turn off randomly without recovery before they are able to cause a switch in the whole system. If the dephosphorylation events do not occur "in a row," switching off a critical number of CaMKII rings within a short time interval, the opposing reactions (autophosphorylation) have time to counteract and turn rings on, allowing the natural dynamics to drive the system back to the UP equilibrium state.
An additional finding is that the molecular turnover of CaMKII strongly limits switch stability (see [Figure 4](#pbio-0030107-g004){ref-type="fig"}). If turnover were absent, switch stability could be an order of magnitude higher (see [Figure 4](#pbio-0030107-g004){ref-type="fig"}C). Our analysis shows an interesting set of relationships between the rates of phosphatase and kinase activities and the rate of protein turnover with regards to their effect on switch stability. In principle, lowering the basal phosphatase and kinase rates in proportion increases switch stability. Such lowering of the dephosphorylation rate has a second desirable feature of lowering energy consumption. This is because the UP state consists of what biochemists call a "futile cycle," in which the rate of ATP-utilizing phosphorylation equals the rate of dephosphorylation. Although minimizing energy utilization dictates that the system be "cooled" (lowering the phosphatase and kinase rates), our results show that there are limits to how much cooling is effective and that this limit is set by the protein turnover rate. Specifically, if cooling sets the phosphatase and kinase rates too low, the system cannot regain steady-state values after a molecular turnover event (newly inserted unphosphorylated kinase molecules will not become fully phosphorylated, as they do in [Figure 4](#pbio-0030107-g004){ref-type="fig"}E). In this case, unphosphorylated kinase molecules will accumulate, leading inexorably toward the threshold for reset to the DOWN state.
Our results suggest that the stability of a switch depends sensitively on a balance of phosphorylation and dephosphorylation rates. Hence we assessed how changes in the ratio of PP1 to CaMKII affect the lifetimes of states of the switch (see [Figure 2](#pbio-0030107-g002){ref-type="fig"}E). We find that short-term fluctuations in the ratio, on a timescale of tens of minutes, do not significantly degrade the switch (see [Figure 5](#pbio-0030107-g005){ref-type="fig"}B). Slower fluctuations are more deleterious, particularly as stability of the UP state is compromised if the number of CaMKII holoenzymes becomes too low (see [Figure 5](#pbio-0030107-g005){ref-type="fig"}C). In contrast to its robustness to short-term fluctuations, our system requires the long-term average ratio of activities is constrained (see [Figure 2](#pbio-0030107-g002){ref-type="fig"}E). Hence, if all other parameters and concentrations are fixed, the ratio of numbers of PP1 molecules to CaMKII molecules should lie in a narrow optimal range (see [Figure 2](#pbio-0030107-g002){ref-type="fig"}E). It will thus be of interest to see whether special mechanisms exist to stabilize the ratio of PP1 to CaMKII in the long term. Promoting the necessary fixed ratio of PP1 to CaMKII may be one of the functions of the scaffolding proteins that hold CaMKII and PP1 within the PSD structure \[[@pbio-0030107-b53],[@pbio-0030107-b54],[@pbio-0030107-b55],[@pbio-0030107-b56],[@pbio-0030107-b57],[@pbio-0030107-b58]\]. Moreover, our results (see [Figure 5](#pbio-0030107-g005){ref-type="fig"}A) indicate that moderate Ca^2+^ fluctuations are tolerable on short timescales, but we find the average level must be tightly regulated over the long term (to within several percent; data not shown). Since the kinase activity depends on the level of free calmodulin, it may further be expected that free calmodulin is regulated over long timescales. This may be an important function of the known calmodulin buffers \[[@pbio-0030107-b59],[@pbio-0030107-b60]\]. In the absence of control mechanisms, the stability of the switch would be greatly reduced.
Several limitations of our study should be noted. While we addressed the effect of Ca^2+^ fluctuations (see [Figure 5](#pbio-0030107-g005){ref-type="fig"}A), we did not include the detailed reaction steps of Ca^2+^ binding to calmodulin in our model, but used the steady-state values for reaction rates based on Ca^2+^/calmodulin. These steady-state rates may not be reached during rapid changes in concentration. Hence, the phosphorylation rates may not vary with Ca^2+^ precisely as we have modeled, in which case other parameters or concentrations would have to be altered to maintain an optimal system. However, including calmodulin-binding steps, and removing the assumption of excess, freely available calmodulin, would reduce the effect of sharp, brief rises in free Ca^2+^. Hence, the influx of Ca^2+^ necessary to cause LTP could be greater than presented here (see [Figure 1](#pbio-0030107-g001){ref-type="fig"}C), and the system may be stable to larger Ca^2+^ fluctuations than those we include. Quantitative measurements of spontaneous Ca^2+^ fluctuations in vivo will be needed to assess whether the switch stability is robust against realistic fluctuations. A second type of simplification that we have made is likely to lead to an underestimate of stability. We have assumed that the CaMKII molecules bound in the PSD are operating with the same kinetic constants measured in free solution. However, some or all of the CaMKII in the PSD may be bound to NMDA receptors \[[@pbio-0030107-b61],[@pbio-0030107-b62]\]. This binding increases the rate of autophosphorylation of the first site, allowing it to occur with only one calmodulin bound rather than two. If such binding to NMDA receptors occurs significantly only after an LTP event, when the system is in the UP state, the effect of protein turnover will be reduced, because unphosphorylated rings would become rapidly phosphorylated before the whole system has time to switch to the DOWN state.
While the switch we have described could be stable for a human lifetime, long-term information storage may not require stability of this magnitude. One possibility is that such long-term stability is not solely a property of the switch, but emerges from interaction between the switch and an attractor network created by memory-specific synaptic connections. According to this idea \[[@pbio-0030107-b47],[@pbio-0030107-b48]\], reactivation of the attractor, perhaps during sleep, may serve to refresh the memory by setting switches back to their correct state. For such a system to work, average switch stability need only be larger than the time between reactivations of the attractor. Such reactivations appear to be important in the early stages of memory, when consolidation is important \[[@pbio-0030107-b47],[@pbio-0030107-b63]\]. However, we know little quantitatively of how frequently such reactivation processes take place. Moreover, the role of reactivation in long-term maintenance of a memory trace, after initial consolidation, remains unclear. Advances in this direction would enhance our understanding of the interplay of molecular and network properties in determining the overall stability of memory in the brain.
Materials and Methods {#s4}
=====================
{#s4a}
### The model {#s4a1}
We treat each ring of six subunits of CaMKII as an independent entity \[[@pbio-0030107-b37],[@pbio-0030107-b64],[@pbio-0030107-b65],[@pbio-0030107-b66]\]. Each subunit can be in one of two states: either phosphorylated or unphosphorylated. The set of possible configurations among the six subunits results in 14 distinguishable states for a ring, labeled here by the number of phosphorylated subunits: P0, P1, P2 (three configurations for P2 because the two phosphorylated subunits can be either neighboring, or separated by either one or two unphosphorylated subunits), P3 (four configurations), P4 (three configurations), P5, and P6. The different configurations have different multiplicities, which are counted when calculating rates of reactions that change configurations. Phosphorylation of the first subunit of a ring (P0 to P1) requires the binding of two activated calmodulins, so is slow at resting Ca^2+^ concentrations (see [Figure 1](#pbio-0030107-g001){ref-type="fig"}A). Once a single subunit is phosphorylated, it can catalyze the phosphorylation of neighboring subunits in a directional manner \[[@pbio-0030107-b27]\], so that further phosphorylation steps are faster at resting Ca^2+^ concentrations (see [Figure 1](#pbio-0030107-g001){ref-type="fig"}A). We refer to a ring in the unphosphorylated state, P0, as off, and a ring with any other level of phosphorylation as on. Once a ring is on, we do not include further slow steps for that ring in the calculations, because the faster, directional steps dominate.
Taking a Hill coefficient of three \[[@pbio-0030107-b40],[@pbio-0030107-b67]\], we have for the initial phosphorylation rate per subunit:
and for the phosphorylation of a clockwise neighboring subunit:
The dephosphorylation occurs through the Michaelis--Menten scheme:
with Michaelis constant, , and where *Sp* denotes a phosphorylated subunit while *S* denotes an unphosphorylated one. In the simulations presented here, we assume *k* ~+~ = *k* ~2~/*K~M~* and set *k* ~−~ to zero, but we find the results are not noticeably influenced by the relative values of *k* ~−~ and *k* ~+~ at fixed *k* ~2~ and *K~M~*.
The phosphatase is deactivated by phosphorylated inhibitor-1 (I1P), a noncompetitive antagonist \[[@pbio-0030107-b68],[@pbio-0030107-b69],[@pbio-0030107-b70]\]. We follow the formulation of Zhabotinsky \[[@pbio-0030107-b28]\], assuming the level of free inhibitor-1 (I1) is constant (at 0.1 μM) in the PSD owing to free exchange of I1 with the larger cell volume. Such free exchange with the larger cell is important, as the number of free I1 molecules in the PSD is less than one on average. Even the spine itself can contain fewer I1 molecules than there are PP1 molecules in the PSD. However, the rapid and strong binding of PP1 to the inhibitor means that the PSD acts as a sink of free I1, and the total concentration of all I1 in the PSD is significantly greater than that of free I1. Importantly, I1 exchanges between the PSD and spine volume vary rapidly, and between the spine and parent dendrite with τ ≤ 1 s, a timescale much faster than that of the phosphatase and kinase reactions.
I1 is phosphorylated by cAMP-dependent protein kinase and dephosphorylated by calcineurin. The rate of dephosphorylation of I1P by calcineurin increases with Ca^2+^, with a Hill coefficient of three \[[@pbio-0030107-b71]\], because calcineurin requires Ca^2+^/calmodulin to activate. Hence I1 is less phosphorylated and the phosphatase is less inhibited at higher Ca^2+^ concentrations. These reaction steps (modeled by Zhabotinsky \[[@pbio-0030107-b28]\]) are assumed to be fast, so we can write down the stationary level of free I1P concentration as
and the fraction, *f~e~,* of phosphatase that is free of inhibitor and hence active as
with *K~I~ = k* ~4~ */k* ~3~.
In the majority of results presented here, we do not simulate the reaction of phosphatase inhibition stochastically. Since the reaction occurs on a timescale faster than other reaction steps, we can use the quasi-steady-state assumption and use only its equilibrium values \[[@pbio-0030107-b72]\]. We did test this assumption by carrying out simulations that included both a stochastic step for phosphorylation of I1 (instead of [equation 4](#pbio-0030107-e004){ref-type="disp-formula"}) and for inhibition of PP1 (instead of [equation 5](#pbio-0030107-e005){ref-type="disp-formula"}). Simulating such fast processes (two to three orders of magnitude faster than other reactions) means a huge increase in the total number of reactions per unit time and hence a corresponding increase in the computer time required. The results for the three systems we tested (red asterisks in [Figure 3](#pbio-0030107-g003){ref-type="fig"}C) showed no significant difference from the results without such fast fluctuations. Hence, we have confidence in our use of the equilibrium values for other data points.
Turnover occurs at a rate, *ν~T~,* and acts equivalently to a non-saturating dephosphorylation process, as we assume all holoenzymes are replaced by unphosphorylated ones, and any attached PP1 is released back to the system.
It is possible to calculate analytically *ν* ~3~, the rate of dephosphorylation by PP1 per phosphorylated subunit. Note that from [Figure 1](#pbio-0030107-g001){ref-type="fig"}B, we require *ν* ~3~ to decrease at high phosphorylation so that the total concentration of subunits dephosphorylated per unit time, *ν* ~3~·\[*Sp~tot~*\], saturates (we have written the total concentration of phosphorylated subunits, \[*Sp~tot~*\] = \[*Sp*\] + \[*PP*1.*Sp*\]). To write the rate of dephosphorylation on the right hand side of [equation 3](#pbio-0030107-e003){ref-type="disp-formula"} as *ν* ~3~·\[*Sp~tot~*\], we assume the intermediate product, \[*PP*1.*Sp*\], is at steady state. We combine [equation 3](#pbio-0030107-e003){ref-type="disp-formula"} with the other two ways that \[*PP*1.*Sp*\] can change. That is, noncompetitive inhibition leads to the reaction
while turnover leads to
So we calculate an effective dephosphorylation rate constant, *ν* ~3~, from *ν* ~3~·\[*Sp~tot~*\] = *k* ~2~\[*PP*1.*Sp*\] assuming *d*\[*PP*1.*Sp*\]/*dt =* 0. The result is, writing \[*E* ~0~\] = *N~PP~* ~1~ *N~A~*/*vol,*
where the concentration of phosphorylated subunits without phosphatase attached, \[*Sp*\], is given by
As expected, the rate, *ν* ~3~, decreases significantly for \[*Sp~tot~*\] \> *K~M~* such that at large total phosphorylation in the UP state, the product is a constant (see [Figure 1](#pbio-0030107-g001){ref-type="fig"}B). In the limit of negligible phosphorylation, , [equation 9](#pbio-0030107-e009){ref-type="disp-formula"} simplifies to give
hence, from [equation 8](#pbio-0030107-e008){ref-type="disp-formula"},
In the limit of high phosphorylation (where ), [equation 9](#pbio-0030107-e009){ref-type="disp-formula"} becomes
leading to
Our derivation differs from the standard Michaelis--Menten approach because we cannot assume that at all times, so \[*Sp*\] ≠ \[*Sp~tot~*\].
### Monte Carlo simulations {#s4a2}
We conducted Monte Carlo simulations of this model, in which all the microscopic configurations of CaMKII holoenzymes were counted and chemical reactions between these states were simulated as stochastic Markov processes. We used the algorithm of Gillespie \[[@pbio-0030107-b51],[@pbio-0030107-b73]\], which can be summarized as follows. We identify all the possible configurations of rings. There are 56 in total, because for a configuration with a given number, *n,* of phosphate groups, the number of PP1 bound can vary from 0 to *n* (assuming *n* \< *N~PP~* ~1~). So for each of the 14 configurations of phosphate groups, the number of configurations including enzymes is multiplied by *n +* 1. For example, a ring with two subunits phosphorylated (P2) can have up to two PP1 enzymes attached. As there are three distinguishable configurations for two phosphorylated subunits on a ring, and each configuration can have zero, one, or two PP1 enzymes attached, we include nine separate configurations for P2. We have the following numbers of configurations: P0 (1), P1 (1 × 2), P2 (3 × 3), P3 (4 × 4), P4 (3 × 5), P5 (1 × 6), and P6 (1 × 7), to obtain 56 in total. We do not explicitly count the different relative positions of PP1 bound to phosphate groups, but just select one of the phosphorylated subunits at random when the dephosphorylation occurs.
At any point in time, the system is in a state that is defined by the number of rings in each configuration, denoted by {*N~i~*}, *i* = 1, 2,..., 56. The numbers of rings in each configuration {*N~i~*} determine the rates of each of the possible reaction steps, {*r~j~*}. A reaction step takes a ring from one of the 56 configurations to another. With the three types of reaction steps (phosphorylation, PP1 binding, or dephosphorylation) able to occur from most of the configurations, in total the system has 144 distinguishable reaction steps. However, most of the rates are zero at any particular instance. Protein turnover is treated stochastically like any other process. Turnover results in two randomly chosen rings (i.e., one holoenzyme) being replaced by two rings in the state P0 (totally unphosphorylated).
The fundamental assumption behind the simulations is that any particular rate, *r~j~*(*t*), depends only on the present configuration, not on the history of the system. This makes the reaction scheme a Markov process. The key step that is necessary when dealing with small numbers of molecules is to recall that the deterministic rate is the macroscopic average of a stochastic process, such that the probability, *P~j~dt* of reaction step *j* in a small time interval, *dt,* is given by
that is, the rate is simply the probability of occurrence per unit time, and, given the Markov assumption, we now have a Poisson process.
Gillespie\'s algorithm proceeds by first summing the rates of all reaction steps to obtain the average rate, *R~T~,* for any change in configuration:
The distribution of times elapsed before a reaction step follows an exponential decay,
which is a standard result, easily verified as it satisfies the necessary requirement:
that is, the probability of reaction in time step from *τ* to *τ* + *dτ* equals *R*(*τ*)*dτ* multiplied by the probability that a reaction did not happen before *τ*.
Once the time for the next reaction step is randomly selected, a second random selection is taken to decide which particular reaction occurs. The relative probabilities, *p~j~*, of each reaction step are simply proportional to their rates, *p~j~* = *r~j~*/*R~T~*.
Given the time and type of the next reaction step, the total time for the system is advanced by *τ*, and the quantities, {*N~i~*}, in relevant configurations are updated, as are the reaction rates for the affected reaction steps. The process now repeats itself with a new set of reaction rates. Total phosphorylation is monitored, and thresholds for determining a switch to the UP state or a switch to the DOWN state are set according to the equilibrium values, but typically if the total falls to below 10% phosphorylation, we record a transition to the DOWN state, and if it spontaneously rises to 70% phosphorylation, we record a transition to the UP state. It should be noted that the time it takes for the system to switch between UP and DOWN states is on the order of hours, which gives an overestimate of the lifetime of a state when in reality the system is "in transition." However, such an error on the order of hours is insignificant (compared to many years for typical average lifetimes) except in the most unstable systems presented here.
Simulations begin with either 0% or 100% phosphorylation, but within a very short time (at least compared to the lifetimes of UP and DOWN states) the system settles near the equilibrium values for UP or DOWN states, respectively.
### Switching-off rate calculations {#s4a3}
We define a ring in the unphosphorylated state, P0, as off, and one with any subunits phosphorylated as on. We calculate the rate for rings to switch off by assuming that all rings in the system that are on reach a dynamic equilibrium in their phosphorylation levels, in the time between switching-on and switching-off events for rings. This approximation, known as separation of timescales, is valid when the individual subunit phosphorylation and dephosphorylation rates are much faster than the rates for rings to switch on and off (as evident in [Figure 4](#pbio-0030107-g004){ref-type="fig"}E).
For a given amount of total phosphorylation, we know exactly the average activity of the phosphatase. Knowing this activity means average rates can be calculated for all reactions. The different average rates determine the rate of change from one configuration of a holoenzyme to another, so knowing the average rates allows us to obtain numerically the relative amounts of time, *Ρ~i~,* spent in each configuration, *i*. Hence we can calculate when the system has a given total level of phosphorylation, *Sp~tot~*, what is the average phosphorylation level per ring, , where *P~i~* is the phosphorylation of configuration *i*. The total level of phosphorylation, *Sp~tot~*, when a given number of rings are on, *N~on~,* is given by
Since *Sp~tot~* is proportional to *P~av~* and *P~av~* depends on *Sp~tot~*, we iterate the equations to find a self-consistent solution for *Sp~tot~* and *P~av~* for each value of *N~on~*.
To find the switching-off rate per ring, plotted in [Figure 4](#pbio-0030107-g004){ref-type="fig"}D, for each number of rings switched on, the total phosphorylation, *Sp~tot~,* is calculated as above. The value of *Sp~tot~* determines the phosphatase activity per subunit, *ν* ~3~, which affects the proportion of time spent by a ring in each configuration. Notably, the lower *ν* ~3~ is, the more phosphorylated are the on rings, and the lower is *Ρ~P~* ~1~. The rate for a ring to switch off is the sum of the turnover rate, *ν~T~*, and rate of dephosphorylation by PP1 of holoenzymes with only a single phosphorylated subunit, *ν~3~Ρ~P~* ~1~.
### Ca^2+^ influx and fluctuations {#s4a4}
For the LTP induction protocol, we use a burst of Ca^2+^ influx constituted by a Poisson train of Ca^2+^ pulses at 100 Hz. Each pulse is a 0.1-μM step increase, followed by a 100 ms exponential decay in free Ca^2+^ concentration.
In the study of the effects of background Ca^2+^ fluctuations, we assume Ca^2+^ entry through NMDA receptors occurs as a random Poisson train with an average rate of 0.5 Hz. We model each Ca^2+^ influx as a step rise of 0.1 or 1.0 μM, followed by exponential decay with a time constant of 100 ms (\[[@pbio-0030107-b44]\], assuming a membrane potential of −50 mV and a spine volume of 0.1 μm^3^; peak \[Ca^2+^\] rise is 0.14 μM per presynaptic spike). We reduce the base level of Ca^2+^ to 0.07 μM in the 0.1**-**μM amplitude simulations, keeping all other parameters the same, to maintain an approximate balance between phosphorylation and dephosphorylation rates. We assume a refractory period of 2 ms for the presynaptic neuron, so that no two Ca^2+^ influxes occur within such a short interval (hence the train of Ca^2+^ inputs is not quite Poisson, but includes a 2-ms negative correlation).
If the fluctuation amplitude is too large (for example, if it is doubled from 0.1 μM to 0.2 μM in this case) and the model parameters are fixed, then bistability is completely lost, because the fluctuations actually change the average effective kinetic rates of reaction steps and bring them outside of the range for bistability of the system. Hence, in the simulations with amplitude 1.0 μM, we maintain a base Ca^2+^ level of 0.1 μM and adjust other parameters (see legend of [Figure 5](#pbio-0030107-g005){ref-type="fig"}) to maintain an appropriate balance between the different reaction rates. For example, we include a higher *Κ~H~* ~1~ to maintain low rates for the initial phosphorylation step. With the alternative parameters, the system is only bistable if the Ca^2+^ fluctuations are present.
Notice that the number of free Ca^2+^ ions in the PSD is on average less than one. However, Ca^2+^ acts through Ca^2+^-bound calmodulin (which can be at a higher concentration), and the exchange of both free Ca^2+^ and calmodulin between postsynaptic density and the spine is much more rapid than both the fluctuations considered here and typical CaMKII reaction steps. Hence, we can neglect such strong, but fast, "shot" noise. Exchange of free Ca^2+^ or calmodulin between the dendritic shaft and spine will cause significant additional fluctuations, but only if it is on a slower timescale than the dissociation steps between Ca^2+^ and calmodulin, or Ca^2+^/calmodulin and CaMKII. Future work, including on these specific binding reaction steps \[[@pbio-0030107-b74]\], is necessary to clarify more precisely the dynamical effects of Ca^2+^ changes on the system.
### Parameters {#s4a5}
Parameters for the standard system are given in [Table 1](#pbio-0030107-t001){ref-type="table"}. In figures where one parameter varies, all others are fixed according to the table, unless otherwise stated. Figures where the number of holoenzymes changes also include a proportionally changed volume and number of phosphatases, to maintain fixed concentrations. The standard concentration of PP1 is equal to that of CaMKII holoenzymes, at 20 molecules per 10^6^ nm^3^ or 33 μM. Our maximal system with 20 holoenzymes is slightly smaller than an average synapse of 30 holoenzymes \[[@pbio-0030107-b32]\]. Hence, the volume is smaller than average, corresponding to a cylindrical PSD of diameter 250 nm (compared to an average diameter of 350 nm in \[[@pbio-0030107-b32]\]), assuming the holoenzymes are predominantly in a single layer of a little over 20 nm thickness, to give a volume of 10^6^ nm^3^ for the domain of reactions.
For those parameters in the table without experimental references, we chose values that were in a reasonable range given the values for similar chemical reactions. We picked simple values that would work well for our system. For example, a low Michaelis constant is beneficial because the range of bistability increases \[[@pbio-0030107-b28]\]. In general, variation of any one parameter, without alteration of the others, leads to an effect like that shown in [Figure 2](#pbio-0030107-g002){ref-type="fig"}, where number of phosphatases is varied. If, for example, the activity of calcineurin is higher (e.g, giving *ν* ~Ca*N*~= 2.0 instead of *ν* ~Ca*N*~=1.0), then the amount of I1P is halved and the amount of uninhibited PP1 increases. Hence, the stability of the UP state decreases while stability of the DOWN state increases. However, a system with a lower overall PP1 concentration would work as well as the original system. So, modification of individual parameters does degrade the system, reducing the lifetime of one state relative to the other. Nevertheless, if the cell is able to maintain concentrations of species in an optimal range determined by the actual value of parameters, then the best results shown here can be achieved (cf. \[[@pbio-0030107-b28]\]).
PM is grateful to the National Institutes of Heath (NIH) National Institute of Mental Health for support under a Mentored Career Development Award, K25-MH064497. JEL, AMZ, and XJW are funded by the Packard Interdisciplinary Science Program. JEL, AMZ, and XJW are also supported by NIH National Institute of Neurological Disorders and Stroke (NINDS) Collaborative Research in Computational Neuroscience grant 1 R01 NS 50944--01, and JEL by NIH NINDS grant 2 R01 NS027337--16. The authors are grateful to Irving Epstein and Eugene Carter for comments on the manuscript.
**Competing interests.** The authors have declared that no competing interests exist.
**Author contributions.** AMZ, JEL, and XJW conceived the problem and designed the model. PM designed and performed the computational experiments. PM, AMZ and XJW analyzed the data. PM, JEL, and XJW wrote the paper.
Citation: Miller P, Zhabotinsky AM, Lisman JE, Wang XJ (2005) The stability of a stochastic CaMKII switch: Dependence on the number of enzyme molecules and protein turnover. PLoS Biol 3(4): e107.
CaMKII
: calcium/calmodulin-dependent protein kinase II
I1
: free inhibitor-1
I1P
: phosphorylated inhibitor-1
LTP
: long-term potentiation
NMDA
: [N]{.smallcaps}-methyl-[D]{.smallcaps}-aspartate
PP1
: protein phosphatase-1
PSD
: postsynaptic density
*Sp~tot~*
: total number of phosphorylated subunits
|
PubMed Central
|
2024-06-05T03:55:54.993754
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069645/",
"journal": "PLoS Biol. 2005 Apr 29; 3(4):e107",
"authors": [
{
"first": "Paul",
"last": "Miller"
},
{
"first": "Anatol M",
"last": "Zhabotinsky"
},
{
"first": "John E",
"last": "Lisman"
},
{
"first": "Xiao-Jing",
"last": "Wang"
}
]
}
|
PMC1069646
|
Introduction {#s1}
============
Over 1 billion people in more than 90 countries are at risk from filarial nematode infections, and 150 million people are infected. The parasitic nematodes are insect-borne and are responsible for lymphatic or cutaneous filariasis, leading to medical conditions including elephantiasis or onchocerciasis (African river blindness). Lymphatic filariasis is caused predominantly by Wuchereria bancrofti and Brugia malayi and affects 120 million individuals, a third of whom show disfigurement, while onchocerciasis, caused by *Onchocerca volvulus,* affects 18 million people of whom 500,000 have visual impairment and 270,000 are blind \[[@pbio-0030121-b01],[@pbio-0030121-b02]\]. Within these filarial parasites are intracellular bacteria that were first observed almost 30 y ago \[[@pbio-0030121-b03],[@pbio-0030121-b04],[@pbio-0030121-b05],[@pbio-0030121-b06]\].
The establishment in 1994 of a Filarial Genome Project funded by the World Health Organization (WHO/Tropical Disease Research/United Nations Development Programme/World Bank) contributed to the rediscovery of these endosymbiotic bacteria. In the analysis of cDNA libraries generated from different life cycle stages of *B. malayi,* the presence of rare non-*Escherichia-coli-*like, alpha-proteobacterial sequences implicated the occurrence of endobacterial DNA \[[@pbio-0030121-b07]\]. Phylogenetic analyses subsequently identified the bacteria as *Wolbachia* \[[@pbio-0030121-b08]\]. These endosymbionts have now been found in the vast majority of filarial nematode species, with notable exceptions \[[@pbio-0030121-b03],[@pbio-0030121-b09],[@pbio-0030121-b10],[@pbio-0030121-b11],[@pbio-0030121-b12],[@pbio-0030121-b13],[@pbio-0030121-b14],[@pbio-0030121-b15],[@pbio-0030121-b16],[@pbio-0030121-b17],[@pbio-0030121-b18],[@pbio-0030121-b19]\]. *Wolbachia* appear to be absent in nonfilarial nematodes \[[@pbio-0030121-b20]\].
In nematodes that contain *Wolbachia* and which have been well examined, the bacteria are located in the lateral chords (invaginations of the body wall hypodermis that project into the body cavity) in both sexes. They are also localized in oocytes but not in the male reproductive tract. The endosymbionts appear to be present in 100% of individuals within a population, when that species contains them, suggesting that they are required for worm fertility and survival \[[@pbio-0030121-b10],[@pbio-0030121-b21],[@pbio-0030121-b22]\]. They are therefore potential therapeutic targets for filariasis control.
Certain antialpha proteobacterial agents, most notably tetracycline and doxycycline, but also rifampicin and azithromycin, show inhibitory effects on parasitic nematode development and fertility \[[@pbio-0030121-b13],[@pbio-0030121-b23],[@pbio-0030121-b24],[@pbio-0030121-b25],[@pbio-0030121-b26],[@pbio-0030121-b27],[@pbio-0030121-b28],[@pbio-0030121-b29],[@pbio-0030121-b30],[@pbio-0030121-b31],[@pbio-0030121-b32],[@pbio-0030121-b33]\]. After antibiotic treatment, immunogold staining, using *Wolbachia-*specific cell-surface probes, shows the absence of *Wolbachia* in the female reproductive tract and the degeneration of embryos, while *Wolbachia* remain in the lateral chords, albeit in reduced numbers \[[@pbio-0030121-b34]\]. Genchi et al. \[[@pbio-0030121-b35]\] have also shown that *Wolbachia* are present at 1000X lower frequencies after antibiotic treatment and can still be detected by PCR from female hypodermis tissues, but not from female reproductive tissue. No antibiotic effects are observed in filarial nematodes that do not harbor *Wolbachia,* nor are they observed with other antibiotics (e.g., penicillin, gentamicin, ciprofloxacin, or erythromycin), suggesting that these effects correlate with *Wolbachia* presence \[[@pbio-0030121-b11],[@pbio-0030121-b12],[@pbio-0030121-b13],[@pbio-0030121-b36],[@pbio-0030121-b37]\]. Human trials using doxycycline, undertaken in Ghana, have shown that this antibiotic interferes with embryogenesis in adult female filariae with a concomitant depletion of *Wolbachia* from both adults and microfilariae (first stage larvae) of O. volvulus and W. bancrofti \[[@pbio-0030121-b38],[@pbio-0030121-b39],[@pbio-0030121-b40],[@pbio-0030121-b41],[@pbio-0030121-b42]\]. Thus, as in animal models, *Wolbachia* appears to be a therapeutic target for human filarial parasitic infections.
The use of anti-*Wolbachia* chemotherapy against filarial parasites has initiated a novel approach for filarial disease control and eradication. Previous strategies for elimination of filariasis have included vector control in the presence or absence of antiparasitic drugs \[[@pbio-0030121-b43],[@pbio-0030121-b44],[@pbio-0030121-b45],[@pbio-0030121-b46],[@pbio-0030121-b47]\]. Diethylcarbamazine, albendazole, and ivermectin have been the most recent drugs of choice for prevention of filarial infections, but they have little effect on adult worms. Thus repeated doses in endemic areas are required to eliminate infections that can arise again within months of treatment \[[@pbio-0030121-b39],[@pbio-0030121-b44],[@pbio-0030121-b48]\]. In addition, the possibility of drug resistance, as observed with intestinal helminths in animals is a concern \[[@pbio-0030121-b49],[@pbio-0030121-b50],[@pbio-0030121-b51]\]. No new therapeutics have been developed in over 20 y, and there is a need for better drugs that permanently sterilize or kill adult worms.
*Wolbachia* play a role in the host immunological response to filarial parasite invasion. Infection by filarial parasites results in B-cell proliferation and the generation of antibodies directed toward parasite- and *Wolbachia*-specific antigens, including those to *Wolbachia* surface protein, heat shock protein, aspartate aminotransferase, and Htr serine protease \[[@pbio-0030121-b11],[@pbio-0030121-b52],[@pbio-0030121-b53],[@pbio-0030121-b54],[@pbio-0030121-b55],[@pbio-0030121-b56],[@pbio-0030121-b57]\]. Other *Wolbachia*-specific molecules also play roles in the immune response to filarial infections including the release of stimulatory and modulatory factors from neutrophils and monocytes, which may be related to *Wolbachia* release upon worm death \[[@pbio-0030121-b58],[@pbio-0030121-b59],[@pbio-0030121-b60],[@pbio-0030121-b61]\]. One component of the host immune response appears to mimic a lipopolysaccharide (LPS)-like response, typically observed as a host immune response to Gram-negative bacteria (such as the alpha-proteobacterial *Wolbachia*) \[[@pbio-0030121-b22],[@pbio-0030121-b58],[@pbio-0030121-b62],[@pbio-0030121-b63],[@pbio-0030121-b64],[@pbio-0030121-b65]\]. Further, LPS-like products of *Wolbachia* appear to be involved in the eye inflammation observed in African river blindness. Leukocytes (neutrophils and eosinophils) infiltrate the cornea as a result of microfilarial invasion and death within the eye, leading to a loss of corneal transparency \[[@pbio-0030121-b66]\]. LPS-like molecules are implicated in this process due to activation of the toll-like receptor 4 (TLR4) pathway by *Wolbachia* \[[@pbio-0030121-b61],[@pbio-0030121-b67]\].
Release of filarial worm-associated molecules, especially after drug treatments that cause worm death in the host, leads to pathogenesis ("Mazzotti Reaction") \[[@pbio-0030121-b68],[@pbio-0030121-b69],[@pbio-0030121-b70],[@pbio-0030121-b71],[@pbio-0030121-b72]\], and *Wolbachia* has been associated with chronic and acute infection states of filariasis (reviewed in \[[@pbio-0030121-b59]\]). Repetitive exposures to LPS-like molecules due to release of *Wolbachia* following death of microfilaria are thought to induce chronic inflammation events giving rise to immune tolerance \[[@pbio-0030121-b65],[@pbio-0030121-b73]\], as hyporesponsiveness occurs with increasing parasite load \[[@pbio-0030121-b74],[@pbio-0030121-b75],[@pbio-0030121-b76]\].
*Wolbachia* endosymbionts can be separated into six supergroups based upon 16S rRNA, *Wolbachia* surface protein, and *fts*Z phylogenetics \[[@pbio-0030121-b08],[@pbio-0030121-b11],[@pbio-0030121-b15],[@pbio-0030121-b77],[@pbio-0030121-b78],[@pbio-0030121-b79],[@pbio-0030121-b80],[@pbio-0030121-b81],[@pbio-0030121-b82]\]. Four supergroups contain *Wolbachia* from arthropods while supergroup C contains *Wolbachia* from the nematodes O. volvulus and *Dirofilaria immitis,* and supergroup D contains *Wolbachia* from *B. malayi, W. bancrofti,* and Litomosoides sigmodontis \[[@pbio-0030121-b11],[@pbio-0030121-b82]\]. In nematodes, the evolution of *Wolbachia* parallels the phylogenetics of their hosts, while in the other supergroups, horizontal transmission appears to have occurred \[[@pbio-0030121-b11],[@pbio-0030121-b14],[@pbio-0030121-b15],[@pbio-0030121-b79],[@pbio-0030121-b82]\]. The closest bacterial relatives to the *Wolbachia* are in the Order Rickettsiales, including *Rickettsia, Ehrlichia, Cowdria,* and *Anaplasma,* all parasites of mammals that require arthropod vectors for transmission \[[@pbio-0030121-b83],[@pbio-0030121-b84]\].
Up to 70% of all insect species appear to harbor *Wolbachia* \[[@pbio-0030121-b85],[@pbio-0030121-b86],[@pbio-0030121-b87]\]. While parasitic and maternally inherited in insects, they appear not to be required for host survival. But when present in appropriate genetic backgrounds, they confer developmental effects leading to sex ratio disturbances, feminization of genetic males, parthenogenesis, cytoplasmic incompatibilities and/or reciprocal-cross sterility \[[@pbio-0030121-b79],[@pbio-0030121-b88],[@pbio-0030121-b89],[@pbio-0030121-b90]\]. It has been suggested that endosymbionts, including *Wolbachia,* might be of medical importance and used for insect vector control to deliver antiparasitic products to recipient hosts \[[@pbio-0030121-b91],[@pbio-0030121-b92],[@pbio-0030121-b93],[@pbio-0030121-b94],[@pbio-0030121-b95],[@pbio-0030121-b96],[@pbio-0030121-b97],[@pbio-0030121-b98],[@pbio-0030121-b99],[@pbio-0030121-b100],[@pbio-0030121-b101], [@pbio-0030121-b102]\]. For these reasons, a genome project was initiated and completed on the *Wolbachia* endosymbiont of Drosophila melanogaster (*w*Mel) \[[@pbio-0030121-b103]\].
Identification of *Wolbachia* in parasitic nematodes, their role in pathogenesis, their potential as a target for development of antifilarial therapeutics, and their widespread occurrence in arthropods triggered a meeting held in 1999 to initiate a consortium of *Wolbachia* researchers \[[@pbio-0030121-b104],[@pbio-0030121-b105]\]. Three additional meetings have been held (see <http://www.wolbachia.sols.uq.edu.au/index.html>, and eight additional *Wolbachia* genomes responsible for diverse phenotypes are being sequenced.
We report the second complete genome sequence of *Wolbachia* and the first from a parasitic nematode, B. malayi (*W. pipientis,* BruMal TRS strain; *Wolbachia* endosymbiont of *B.* malayi \[*w*Bm\]). We also describe a comparative analysis of reductive evolution in different lineages of endosymbiotic bacteria, a major evolutionary trend in all intracellular parasites and symbionts. Features of the *w*Bm genome are presented as a systematic comparison to *w*Mel and *Rickettsia* spp., the closest fully sequenced relatives of *w*Bm and more distant intracellular parasites and symbionts of the gamma-proteobacterial lineage, such as *Buchnera* (aphid endosymbiont), *Blochmannia* (ant endosymbiont), and *Wigglesworthia* (tsetse fly endosymbiont) \[[@pbio-0030121-b106],[@pbio-0030121-b107],[@pbio-0030121-b108],[@pbio-0030121-b109],[@pbio-0030121-b110],[@pbio-0030121-b111],[@pbio-0030121-b112]\]. We also delineate the metabolic pathways that might account for the mutualistic relationship between *Wolbachia* and its nematode host.
Results/Discussion {#s2}
==================
Genome Properties and General Comparison with the Genomes of Other Parasites and Endosymbionts {#s2a}
----------------------------------------------------------------------------------------------
The genome of *w*Bm is represented by a single circular chromosome consisting of 1,080,084 nucleotides and is 34% G+C. The size agrees with the 1.1 Mb length previously determined by both pulsed-field gel electrophoresis and restriction mapping \[[@pbio-0030121-b113],[@pbio-0030121-b114]\]. The origin of replication *(oriC)* was tentatively mapped immediately upstream of the *hem*E gene on the basis of GC- and AT-skew analyses \[[@pbio-0030121-b115]\] ([Figure 1](#pbio-0030121-g001){ref-type="fig"}). The genome of *w*Bm has an extremely low density of predicted functional genes compared to all other bacteria, with the exceptions of R. prowazekii ([Table 1](#pbio-0030121-t001){ref-type="table"}) and *Mycobacterium leprae.* Both *Wolbachia* spp. and *Rickettsia* spp. have undergone considerable gene loss in many metabolic pathways, relative to other alpha-proteobacteria ([Table 2](#pbio-0030121-t002){ref-type="table"}). A comparison of predicted functional genes in *w*Bm and *Rickettsia* spp. reveals a large core set that is conserved among these genomes, as well as smaller sets unique to each genome ([Figure 2](#pbio-0030121-g002){ref-type="fig"}). In contrast, nearly all observed pseudogenes are unique to each genome ([Figure 2](#pbio-0030121-g002){ref-type="fig"}), suggesting substantial independent genome degradation. *Wolbachia* (*w*Bm) and R. conorii contain, in addition to many demonstrable pseudogenes, a considerable number of short open reading frames (ORFs), which have no detectable orthologs in current protein databases but are recognized as probable genes by gene prediction programs. However, most of these sequences, which comprise approximately 5% of the total predicted gene number in *w*Bm, are likely to be fragmented genes as well ([Table 1](#pbio-0030121-t001){ref-type="table"}).
::: {#pbio-0030121-g001 .fig}
Figure 1
::: {.caption}
###### Genogram of the Complete Circular Genome of *w*Bm
The scale indicates coordinates in kilobase pairs (kbp) with the putative origin of replication positioned at 0 kbp. The outermost ring indicates the GC-skew over all bases in the forward strand using a window size of 40 kbp and a step size of 1 kbp. Positive and negative skew are shaded gold and blue, respectively. Features are shown as paired rings separated by a circular baseline. In each pair, the outer and inner rings represent the forward and reverse DNA strands, respectively. Working inward from the scale, the features displayed are as follows: identified genes and their broad functional classification (multihued, as listed); tRNA (blue)/rRNA (red) genes; putative pseudogenes (green); repeated sequences (red) and transposon-related repeats (blue).
:::
:::
::: {#pbio-0030121-g002 .fig}
Figure 2
::: {.caption}
###### Venn Diagram Showing Comparison of Conserved and Unique Genes and Pseudogenes in *w*Bm *(Wolbachia* from *B. malayi), Rickettsia prowazekii, Rickettsia conorii,* and in *w*Bm and *w*Mel (among Those Assigned to COGs)
\(A) Predicted functional protein-coding genes.
\(B) Pseudogenes.
\(C) Combined results for comparison between *wBm* and *w*Mel.
G, intact gene; P, pseudogene.
:::
:::
::: {#pbio-0030121-t001 .table-wrap}
Table 1
::: {.caption}
###### Comparison of Genome Features of Proteobacterial Endosymbionts and Endoparasites
:::
^a^ Independent estimates obtained during this work. *w*Bm, *Wolbachia* from *B. malayi; w*Mel, *Wolbachia* from *Drosophila melanogaster; R. conorii, Rickettsia conorii; R. prowazekii, Rickettsia prowazekii; B. aphidicola, Buchnera aphidicola; B. floridanus, Blochmannia floridanus; W. glossinidia, Wigglesworthia glossinidia.*
IS, insection element sequence
:::
::: {#pbio-0030121-t002 .table-wrap}
Table 2
::: {.caption}
###### Gene Loss and Decay in *Wolbachia* and *Rickettsia*
:::
Gene conservation and loss were determined with respect to the set of 1,177 genes that are represented by confidently identifiable orthologs in all free-living alpha-proteobacteria. For each category, the first number indicates retained genes, the second number indicates lost genes, and the third number indicates pseudogenes. The sum of these numbers equals the total number of genes in this category in the alpha-proteobacterial core set. *w*Bm, *Wolbachia* from *B. malayi; w*Mel, *Wolbachia* from *Drosophila melanogaster; R. conorii, Rickettsia conorii; R. prowazekii, Rickettsia prowazekii.*
:::
The *w*Bm genome contains one copy of each of the ribosomal RNA genes (16S, 23S, and 5S), which do not form an operon, as also observed in *w*Mel and *Rickettsia* but in contrast to most other bacteria, and 34 tRNA genes that include cognates for all amino acids. Probable biological function was assigned to 558 (approximately 70%) of the 806 protein coding genes; a more general prediction of biochemical function was made for an additional 49 ORFs. Most of the predicted genes (617, 76%) could be included in clusters of orthologous groups of proteins (COGs) with orthologs not only in *w*Mel and *Rickettsia* but also in more distant organisms.
A lack of flagellar, fimbrial or pili genes indicates that *w*Bm is probably nonmotile ([Table 2](#pbio-0030121-t002){ref-type="table"}). However, some intracellular pathogens, including spotted fever group *Rickettsia,* exploit a different motility mechanism that makes use of the host cell actin polymerization to promote bacterial locomotion. Actin-based motility of *Rickettsia* depends upon activation of the host Arp2/3 complex by the WASP family protein RickA \[[@pbio-0030121-b116],[@pbio-0030121-b117]\]. A gene coding for WASP family protein (Wbm0076) was identified in *w*Bm suggesting that it might be able to employ actin polymerization for locomotion and cell-to-cell spread.
Informational and Regulatory Systems {#s2b}
------------------------------------
Comparison with an obligatory gene set characteristic for free-living alpha-proteobacteria ([Table 2](#pbio-0030121-t002){ref-type="table"}) shows that both *Wolbachia* spp. and *Rickettsia* have retained an almost intact gene set for translational processes (greater than 84%). Several RNA metabolism genes are among the few shared losses, including tRNA and rRNA modification enzymes (LasT, RsmC, Sun, TrmA, CspR) and even pseudouridine synthase, TruB (pseudogenes in both lineages). TruB is present in all gamma-proteobacterial endosymbionts but absent in other parasites and endosymbionts, including *Mycoplasma, Chlamydia,* and spirochetes. It is likely that the lack of these modifications affects reading frame maintenance and translation efficiency in both *Wolbachia* spp. and in *Rickettsia.* Further reduction of genes involved in RNA modification occurs specifically in *w*Bm and in *w*Mel, which have lost several genes involved in queuosine biosynthesis (COG0809, COG603, COG702, COG0602, COG0780) \[[@pbio-0030121-b118]\] and 16S rRNA uridine-516 pseudouridylate synthase. The absence of RNA methylase (COG1189) highlights the loss of RNA modification systems, which is a general trend in evolution of endosymbionts among various lineages \[[@pbio-0030121-b119]\].
Although *w*Bm retains most of the genes for DNA replication and repair, the loss of several genes present in other alpha-proteobacteria (except *w*Mel) is notable. These include the chi subunit of DNA polymerase III (HolC), chromosome partitioning proteins ParB and ParA, repair ATPase (RecN), exonuclease VII (XseAB), and the RNA processing enzyme RNase PH (Rph).
Both *Wolbachia* spp. and *Rickettsia* have a complete repertoire of UV-excision (UVR-ABCD-mediated), recombinational synaptic (RecA/RecFOR-mediated), and postsynaptic (RuvABC-mediated) DNA repair pathways. In contrast, *Buchnera* and *Blochmannia* are devoid of conventional homologous recombination and uvr pathways, although they encode a putative phrB familyphotolyase \[[@pbio-0030121-b107],[@pbio-0030121-b109],[@pbio-0030121-b110],[@pbio-0030121-b112],[@pbio-0030121-b120],[@pbio-0030121-b121]\]. *Wolbachia, Rickettsia, Buchnera,* and *Wigglesworthia* all encode enzymatic machinery to counter the deleterious effects of various types of base oxidative damage, which could be important for defense against mutagenic metabolic by-products in the intracellular environment \[[@pbio-0030121-b103],[@pbio-0030121-b108],[@pbio-0030121-b109],[@pbio-0030121-b119],[@pbio-0030121-b122]\].
Many proteins categorized as being involved in protein fate in the two *Wolbachia* spp. and *Rickettsia* spp. (CcmF, CcmB, CcmH, CcmE, CcmC, Cox11, CtaA), but which are absent in the genomes of gamma-proteobacterial endosymbionts, are involved in biogenesis of cytochrome c oxidase and c-type cytochromes typical of alpha-proteobacterial aerobic respiratory chains. Respiratory chains of gamma-proteobacterial endosymbionts employ quinol oxidase rather than cytochrome c oxidase.
A major loss of transcriptional regulators likely occurred in the common ancestor of *Wolbachia* and *Rickettsia* spp. ([Table 2](#pbio-0030121-t002){ref-type="table"}). Only a few of these genes have been additionally lost in the *w*Bm lineage, including those from COG1396, COG1959, COG1329, COG1678, and COG1475. This is a general trend in evolution of endosymbionts and parasites \[[@pbio-0030121-b118],[@pbio-0030121-b122],[@pbio-0030121-b123]\], suggesting that most of their genes are likely constitutively expressed. Those few regulators found in *w*Bm that are not present in other alpha-proteobacteria, including two Xre-like regulators (COG5606), may be of interest for future experimental characterization. Similarly, most genes implicated in signal transduction systems are absent in both *Wolbachia* and *Rickettsia* spp. Several regulatory proteins that remain in the genome are involved in various stress responses (Wbm0660, MerR/SoxR family; Wbm0707, cold shock protein; Wbm0494, stress response morphogen; Wbm0061, TypA-like GTPase) or in cell cycle regulation (Wbm0184, PleD-like regulator; Wbm0596, cell cycle transcriptional regulator CtrA).
Metabolic Capabilities of *w*Bm are Key to Understanding its Interaction with the Host {#s2c}
--------------------------------------------------------------------------------------
One of the roles of *w*Bm as an obligate endosymbiont may be to provide its host with essential metabolites. Although *w*Bm has retained more metabolic genes than *Rickettsia* spp., its biosynthetic capabilities appear to be rather limited. Unlike *Buchnera* spp. \[[@pbio-0030121-b107],[@pbio-0030121-b109],[@pbio-0030121-b112],[@pbio-0030121-b122],[@pbio-0030121-b123]\], *w*Bm is able to make only one amino acid---*meso*-diaminopimelate (*meso*-DAP), a major peptidoglycan constituent. In most bacteria, it is produced as an intermediate in the pathway of lysine biosynthesis. Similar to *Rickettsia* spp. \[[@pbio-0030121-b122]\], *w*Bm lacks *meso*-DAP decarboxylase (LysA, COG0019), necessary for lysine biosynthesis, such that the biochemical pathway ends with *meso*-DAP.
Complete pathways for de novo biosynthesis of purines and pyrimidines are found in *w*Bm, as opposed to *Rickettsia* and many other endosymbionts and parasites, including *Buchnera, Blochmannia, Mycoplasma,* and *Chlamydia* ([Table 3](#pbio-0030121-t003){ref-type="table"}). The general trend for nucleotide biosynthesis pathways to be lost in these organisms appears to be independent of the presence of ADP/ATP translocase (COG3202) (present only in *Rickettsia* and *Chlamydia*), which facilitates the uptake of nucleotide-triphosphates from the hosts. This observation suggests that *w*Bm produces nucleotides not only for internal consumption but also for supplementation of the nucleotide pool of the host ([Figure 3](#pbio-0030121-g003){ref-type="fig"}) when needed, such as during oogenesis and embryogenesis, where the requirement for DNA synthesis is likely very high \[[@pbio-0030121-b124]\].
::: {#pbio-0030121-g003 .fig}
Figure 3
::: {.caption}
###### Metabolic Pathways Retained in wBm
Pathways shared by *Wolbachia* and *Rickettsia* are shown with black arrows. Pathways present in *Wolbachia* but not in *Rickettsia* are shown with green arrows. Numbering alongside pathway arrows reflects enzyme annotation, a table of which is available at <http://tools.neb.com/wolbachia/>.
:::
:::
::: {#pbio-0030121-t003 .table-wrap}
Table 3
::: {.caption}
###### Differential Loss of Functionality and Differentially Preserved Functionality, if Only a Few Compared Alpha- and Gamma-Proteobacterial Parasite/Symbiont Genomes Have Lost or Preserved This Functionality
:::
*w*Bm, *Wolbachia* from B. malayi; *w*Mel, *Wolbachia* from *Drosophila melanogaster; R. conorii, Rickettsia conorii; R. prowazekii, Rickettsia prowazekii; B. aphidicola, Buchnera aphidicola; B. floridanus, Blochmannia floridanus; W. glossinidia, Wigglesworthia glossinidia.*
:::
All genes required for biosynthesis of fatty acids and all but one gene for biosynthesis of phospholipids (phosphatidylglycerol, phosphatidylserine, and phosphatidylethanolamine) are present in the *w*Bm genome. The absent gene in phospholipid biosynthesis is glycerol-3-phosphate acyltransferase (COG2937), which catalyzes the transfer of the first fatty acid to glycerol-3-phosphate. However, a "fatty acid/phospholipid biosynthesis enzyme" PlsX is present, which can complement the absence of glycerol-3-phosphate acyltransferase in E. coli \[[@pbio-0030121-b125]\]. All but one gene for biosynthesis of isoprenoids has been found in the genome. This absent gene is 1-deoxy-D-xylulose-5-phosphate synthase (COG1154), an essential gene in the nonmevalonate pathway. It is possible that this biochemical function could be complemented by a transketolase or transaldolase, two highly promiscuous enzymes encoded by the *w*Bm genome or, alternatively, 1-deoxy-D-xylulose-5-phosphate must be supplied by the host.
Unlike *Rickettsia, w*Bm contains all the enzymes for the biosynthesis of riboflavin and flavin adenine dinucleotide ([Figure 3](#pbio-0030121-g003){ref-type="fig"}). *w*Bm could be an important source of these essential coenzymes for the host nematode. No genes for riboflavin biosynthesis have been detected in the ongoing B. malayi genome data (9X coverage) \[[@pbio-0030121-b126]\]. Similar to most other endosymbionts, *w*Bm lacks complete pathways for de novo biosynthesis of other vitamins and cofactors such as Coenzyme A, NAD, biotin, lipoic acid, ubiquinone, folate, and pyridoxal phosphate, retaining only a few genes for the finals steps in some of these pathways. These incomplete pathways may make *w*Bm dependent upon the supply of those precursors from the host.
Heme serves as a prosthetic group of cytochromes, catalase and peroxidase, and may be another metabolite provided by *w*Bm to *B. malayi. w*Bm has all but one gene for heme biosynthesis and has maintained all genes for maturation of c-type cytochromes. The absent gene in the heme biosynthesis pathway encodes protoporphyrinogen oxidase, a gene not identified in many alpha-proteobacteria. It is likely that these bacteria contain a functional form of protoporphyrinogen oxidase, which is not yet known, or that the missing function is complemented by another gene function, as in E. coli \[[@pbio-0030121-b127]\].
Heme could play an important role in filarial reproduction and development. It is possible that molting and reproduction are regulated by ecdysteroid-like hormones, since the insect hormones ecdysone and 20-hydroxyecdysone and their inhibitors affect molting and microfilarial release in D. immitis and B. pahangi \[[@pbio-0030121-b128],[@pbio-0030121-b129]\]. In *Drosophila,* five enzymatic reactions in the pathway of ecdysteroid biosynthesis are catalyzed by microsomal and mitochondrial cytochrome P450 mono-oxygenases \[[@pbio-0030121-b130]\]. If similar enzymes participate in the pathway of biosynthesis of filarial steroid hormones, heme depletion caused by elimination of *w*Bm could result in a decreased activity of these enzymes, which might account for the effects on nematode viability, larval development, and reproductive output observed following antibiotic treatment of filarial parasites.
There is currently no evidence of heme biosynthesis enzymes in B. malayi (analysis of the draft genome sequence of B. malayi does not identify any genes for heme biosynthesis \[[@pbio-0030121-b126]\]). These enzymatic activities have been detected in *Setaria digitata,* a cattle filarial parasite, which is devoid of typical cytochrome systems, yet has heme-containing enzymes, such as microsomal cytochrome P450, catalase, and peroxidase \[[@pbio-0030121-b131]\]. It is not known whether S. digitata contains *Wolbachia* and whether heme biosynthesis detected in this worm is due to the presence of endosymbiotic bacteria. However the closely related filarial parasites, *S. equina, S. tundra,* and S. labiatopapillosa are devoid of endosymbiotic *Wolbachia* \[[@pbio-0030121-b15],[@pbio-0030121-b16]\]; perhaps they have retained the genes for heme biosynthesis.
Genes for biosynthesis of glutathione are present in the *w*Bm genome (Wbm0556; Wbm0721). Two physiological roles of glutathione in bacteria are known: one is detoxification of methylglyoxal \[[@pbio-0030121-b132]\], and the other is protection against oxidative stress through activation of the glutathione peroxidase--glutathione reductase system \[[@pbio-0030121-b133],[@pbio-0030121-b134]\]. Methylglyoxal is accumulated in phosphate-limited environments, such as those encountered by *Salmonella* inside macrophages \[[@pbio-0030121-b132]\]. It is possible that *w*Bm encounters phosphate-limited conditions inside the host and therefore needs glutathione as a quencher of methylglyoxal. This view is supported by the presence of the gene encoding the Kef-type potassium efflux system, a participant in methylglyoxal detoxification through acidification of cytosol \[[@pbio-0030121-b132]\]. However, no homologs of *E. coli glo*A--*glo*B genes responsible for glutathione-dependent methylglyoxal detoxification were found in the genome. Glutathione peroxidase is also absent, hence the physiological role of glutathione in *w*Bm is unclear. Although genes for glutathione biosynthesis are present in the B. malayi genome, it is possible that *w*Bm provides glutathione to the host, since the latter needs high levels of this essential metabolite for protection against oxidative stress \[[@pbio-0030121-b135]\] and detoxification \[[@pbio-0030121-b136]\].
Intermediates for these biosynthetic pathways are likely derived from gluconeogenesis, the nonoxidative pentose phosphate shunt, and the tricarboxylic acid (TCA) cycle. Glycolytic enzymes encoded by *w*Bm probably function in a gluconeogenesis pathway ([Figure 3](#pbio-0030121-g003){ref-type="fig"}), since the genes coding for two enzymes catalyzing irreversible glycolytic reactions, 6-phosphofructokinase and pyruvate kinase, are absent. Instead, the gluconeogenic enzyme fructose-1,6-bisphosphatase (Wbm0132) and pyruvate-phosphate dikinase (Wbm0209), which functions predominantly in gluconeogenesis in bacteria, are present suggesting that the pathway functions as gluconeogenesis, albeit ending with fructose-6-phosphate rather than glucose-6-phosphate. While fructose-6-phosphate is necessary for biosynthesis of the peptidoglycan components N-acetylglucosamine and N-acetylmuramate, no enzymes capable of utilizing glucose-6-phosphate as a substrate are encoded in the *w*Bm genome.
It is reasonable to suggest that the most likely growth substrates for *w*Bm would be those compounds that are highly abundant in the worm. In adult *B. malayi, B. pahangi,* and *Dipetalonema viteae (Acanthocheilonema viteae),* these include the excretory metabolites lactate and succinate, which are the principal products of glucose utilization under both aerobic and anaerobic conditions, and a disaccharide trehalose, which is used by the worms as a storage compound \[[@pbio-0030121-b137],[@pbio-0030121-b138]\]. Nuclear magnetic resonance studies of adult B. malayi identified phosphoenolpyruvate as the major energy reservoir \[[@pbio-0030121-b139]\]. However, *w*Bm is not predicted to be able to utilize lactate due to the absence of genes coding for lactate dehydrogenases and is likely unable to grow on sugars, as evidenced by the lack of genes encoding sugar transporters or sugar kinases. Thus, the most likely growth substrates for *w*Bm are pyruvate and TCA cycle intermediates derived from amino acids, with enzymes present for amino acid degradation, a pyruvate dehydrogenase complex, a complete TCA cycle, and a respiratory chain typical of alpha-proteobacteria ([Figure 3](#pbio-0030121-g003){ref-type="fig"}). Amino acids are likely imported from the extracellular environment where they are obtained by proteolysis of host proteins by proteases and peptidases. Indeed, the genome of *w*Bm encodes a variety of proteases, including predicted metallopeptidases (at least seven Zn-dependent proteases of four distinct families compared to only one in *Rickettsia*) (Wbm0055, Wbm0153, Wbm0221, Wbm0311, Wbm0419, Wbm0418, Wbm0742). In addition, two Na^+^/alanine symporters were found (Wbm0197, Wbm0424), which are absent in *Rickettsia.*
Cell Wall Structure {#s2d}
-------------------
A dramatic case of lineage-specific gene loss in both *Wolbachia* spp. includes approximately 20 genes for enzymes of cell-envelope LPS biosynthesis. It has been reported that soluble endotoxin-like products of *Wolbachia* endosymbionts of filarial nematodes, including *B. malayi, B. pahangi, L. sigmodontis, O. volvulus,* and *D. immitis,* contribute to the immunology and pathogenesis of filarial diseases through induction of potent inflammatory responses, including production of tumor necrosis factor alpha, interleukin-1-beta, and nitric oxide by macrophages \[[@pbio-0030121-b22],[@pbio-0030121-b58],[@pbio-0030121-b59],[@pbio-0030121-b60],[@pbio-0030121-b71],[@pbio-0030121-b72],[@pbio-0030121-b140],[@pbio-0030121-b141]\]. Chemokine and cytokine responses to the sterile extracts of *Brugia* and *Onchocerca* were dependent on signaling through TLR4 and could be blocked by neutralizing antibodies to CD14 and by the antagonistic lipid A analogs, indicating that the inflammatory response was induced by an LPS-like molecule. Recently the major surface protein of *Wolbachia* spp. was implicated as the inducer of the immune response acting in a TLR2- and TLR4-dependent manner \[[@pbio-0030121-b141]\]. However, it is not clear whether this protein is the only *Wolbachia*-specific molecule eliciting a TLR4-dependent innate immune response.
Analysis of the *w*Bm genome indicates that, like Ehrlichia chaffeensis and Anaplasma phagocytophilum \[[@pbio-0030121-b142]\], it lacks homologs of the genes responsible for biosynthesis of lipid A. Although lipid A structure can vary in different bacteria, it always consists of a polysaccharide backbone carrying fatty acid residues. The only predicted genes belonging to the glycosyltransferase family were those participating in peptidoglycan biosynthesis, and one glycosyltransferase pseudogene is present. Similarly, the only genes from the acyltransferase family are those participating in fatty acid and phospholipid biosynthesis. Thus, it is unlikely that the cell wall of *w*Bm contains LPS-like molecules. This idea is supported by the absence of the gene products responsible for maintaining the outer membrane structure in Gram-negative bacteria, such as TolQ, TolR, TolA, and TolB.
Several lines of evidence suggest that the structure of the *w*Bm peptidoglycan is very unusual, and peptidoglycan derivatives might be responsible in part for the observed inflammatory responses. First, although all the genes necessary for biosynthesis of lipid II are present in the *w*Bm genome, there are no homologs of alanine and glutamate racemases responsible for synthesis of pentapeptide components D-alanine and D-glutamate. While the genomes of *Rickettsia* spp. contain L-alanine racemase that could catalyze racemization of both alanine and glutamate, the only amino acid racemase present in the genomes of both *Wolbachia* is *meso*-DAP epimerase (Wbm0518), an enzyme catalyzing interconversions of LL- and *meso-*isomers of diaminopimelate. It is possible that *meso*-DAP epimerase is able to catalyze racemization of alanine and glutamate, although this activity has never been experimentally demonstrated. Alternatively, instead of the usual D-isomers, *w*Bm peptidoglycan might contain L-isomers of alanine and glutamate.
Second, Gram-negative bacteria (including *Rickettsia* spp.) usually contain two monofunctional transpeptidases. One of them, FtsI (also known as PBP3), is localized to the septal ring and is required for peptidoglycan biosynthesis in the division septum, while the other, PBP2, is localized preferentially to the lateral cell wall \[[@pbio-0030121-b143]\]. FtsI and PBP2 are recruited to the sites of their action by two membrane proteins, FtsW and RodA, respectively. In the *w*Bm genome, only functional orthologs of E. coli RodA and PBP2 were found; the orthologs of FtsW--FtsI are disrupted by multiple frameshifts.
Third, genomes of bacteria that have peptidoglycan in their cell wall usually contain at least one gene coding for a high molecular weight penicillin-binding protein responsible for cross-linking of the murein sacculus. The transpeptidase and transglycosylase domains of this protein catalyze transpeptidation and transglycosylation of the murein precursors, respectively, to form the carbohydrate backbone of murein and the interstrand peptide linkages. No homologs of bifunctional transpeptidase/transglycosylase or monofunctional biosynthetic transglycosylase were found in the genomes of *Wolbachia* spp., although they are present in the *Rickettsial* genomes. The homolog of lytic transglycosylase, which is responsible for hydrolysis of the carbohydrate backbone during bacterial growth and division, is also absent from the genomes of both *Wolbachia* spp. Thus, their peptidoglycan can be cross-linked by the interstrand peptide linkages, but the carbohydrate backbone is not polymerized. These observations suggest that peptidoglycan of *w*Bm has some features in common with the peptidoglycan-derived cytotoxin produced by Neisseria gonorrhoeae and Bordetella pertussis \[[@pbio-0030121-b144],[@pbio-0030121-b145]\] and that muramyl peptides derived from *w*Bm peptidoglycan could elicit the inflammatory response contributing to the pathogenesis of filarial infection.
Other Host Interaction Systems {#s2e}
------------------------------
As expected, functional Type IV secretion genes were found in the *w*Bm genome, including two operons: Wbm0793--Wbm0798 and Wbm0279--Wbm0283. These systems are indispensable for successful persistence of endosymbionts within their hosts \[[@pbio-0030121-b146]\]. Similar genes have been observed in the sequence of *w*Mel \[[@pbio-0030121-b103]\].
A role in the adaptation to the intracellular existence seems likely for several genes that are present in *w*Bm, *w*Mel, and *Rickettsia.* Thus, *w*Bm encodes five ankyrin-repeat-containing proteins and, in addition, has at least seven related pseudogenes, while *w*Mel contains 23 ankyrin -repeat-containing genes. *Rickettsia* contains two or three functional ankyrin-repeat genes (and probably one pseudogene) \[[@pbio-0030121-b147]\]. In eukaryotes, ankyrins connect cell membranes, including membranes of endosymbionts to the cytoskeleton \[[@pbio-0030121-b148]\], while in bacteria the function of ankyrin-like proteins remains largely unknown. One physiological function of bacterial ankyrin-like proteins was demonstrated in *Pseudomonas aeruginosa,* where ankyrin repeat AnkB is essential for optimal activity of periplasmic catalase, probably serving as a protective scaffold in the periplasm \[[@pbio-0030121-b149]\]. Another ankyrin-repeat protein, AnkA from *E. phagocytophila,* was detected in association with chromatin in infected cells, suggesting its possible role in regulation of host cell gene expression \[[@pbio-0030121-b150]\].
Another interesting protein is a member of the WASP family and is conserved in *Rickettsia* and *w*Bm (Wbm0076). Eukaryotic homologs of these proteins are suppressors of the cAMP receptor and regulate the formation of actin filaments \[[@pbio-0030121-b151]\]. The genes for an ankyrin-repeat protein and a WASP protein might have been acquired from a eukaryotic host by the common ancestor of *Rickettsia* and *Wolbachia* and could have contributed to the evolution of the intracellular lifestyle of these bacteria. *w*Bm also encodes several proteins with large nonglobular or transmembrane regions or internal repeats, orthologs of which are present also in the *w*Mel genome (Wbm0010, Wbm0304, Wbm0362, Wbm0749, and others). These proteins are likely to be surface proteins interacting with host cell structures.
Further Comparisons of *w*Bm and *w*Mel {#s2f}
---------------------------------------
One of the most striking characteristics of the *w*Mel genome is a large amount of repetitive DNA and mobile genetic elements, including three prophages, altogether comprising more that 14% of genomic DNA (and about 134 ORFs). Despite the abundance of repeats in the *w*Bm genome (5.4%) ([Figure 4](#pbio-0030121-g004){ref-type="fig"}), the percentage of repetitive DNA in *w*Bm is considerably less than in *w*Mel. This may reflect a stronger selection in *w*Bm for repeat loss and, as no prophages were identified in the *w*Bm genome, little exposure to foreign DNA. No plasmid maintenance genes were identified in the *w*Bm genome.
::: {#pbio-0030121-g004 .fig}
Figure 4
::: {.caption}
###### Organization of Direct and Palindromic Repeats in *w*Bm
Circles represent the complete genomic sequence of *w*Bm. Repeats were identified using the REPuter program \[[@pbio-0030121-b182]\] and are connected by line segments. Direct repeats are shown in the graphs in the top row, while palindromic repeats are shown in the lower row of graphs. The left column graphs display repeats of 50 to 500 bp in length. The rightmost graphs display repeats of greater than 500 bp in length.
:::
:::
Comparison of the repetitive elements between these two genomes suggests the invasion of mobile genetic elements occurred after the divergence of the two *Wolbachia* along the *w*Mel branch, or that the majority of the transposons and phages were eliminated (degraded) specifically in the *w*Bm lineage. There is a similarly large difference in the amount of repetitive DNA in the two *Rickettsia* species ([Table 1](#pbio-0030121-t001){ref-type="table"}). While an appropriate outgroup would be useful in both comparisons, the apparent degradation of repetitive DNA in *Buchnera* spp. \[[@pbio-0030121-b111],[@pbio-0030121-b112],[@pbio-0030121-b152],[@pbio-0030121-b153],[@pbio-0030121-b154],[@pbio-0030121-b155]\] suggests the specific elimination of nonessential DNA is a result of reduced selection on gene functions no longer necessary in the host cells in *Wolbachia* spp. \[[@pbio-0030121-b156]\]. The large number of repeats and an apparently active system of DNA recombination suggest that extensive genome shuffling within *w*Bm and *w*Mel has eliminated colinearity between their genomes ([Figure 5](#pbio-0030121-g005){ref-type="fig"}). Frequent rearrangements in *Wolbachia* might be expected, given the exceptionally high levels of repeated DNA and mobile elements and the presence of several prophages in *w*Mel. It has been suggested that the surprisingly high percentage of repetitive DNA in *w*Mel might reflect a lack of selection for its elimination \[[@pbio-0030121-b103]\]. An alternative hypothesis might be that in *Wolbachia* there is a selective benefit to systems that maintain genetic diversity and that a high percentage of repeats may contribute to genome plasticity, as has been suggested for *Helicobacter* \[[@pbio-0030121-b157]\]. It has been suggested that the presence of a high level of repetitive DNA in *w*Mel, relative to *w*Bm, might reflect recurrent exposures to mobile elements and bacteriophages, as a result of its parasitic lifestyle \[[@pbio-0030121-b156],[@pbio-0030121-b158]\].
::: {#pbio-0030121-g005 .fig}
Figure 5
::: {.caption}
###### Absence of Gene Order Colinearity between *w*Bm and *Rickettsia* and Disruption of Gene Colinearity between *w*Bm and *w*Mel
Each dot represents a pair of probable orthologs defined as reciprocal BLAST best hits with E-value less than 0.001.
\(A) Genome dot-plot comparison of *w*Bm *(Wolbachia* from *B. malayi)* and *Rpro (R. prowazekii).*
\(B) Genome dot-plot comparison of *w*Bm *(Wolbachia* from *B. malayi)* and *Rcon (R. conorii).*
\(C) Genome dot-plot comparison of *Rpro (R. prowazekii)* and *Rcon (R. conorii).*
\(D) Genome dot-plot comparison of *w*Bm and *w*Mel.
:::
:::
Comparative analysis of the genes assigned to COGs in both *w*Mel and *w*Bm shows that the genome of *w*Bm is more reduced ([Figure 2](#pbio-0030121-g002){ref-type="fig"}; [Table 2](#pbio-0030121-t002){ref-type="table"}). In total, 696 individual proteins from wBm have an ortholog in the *w*Mel genome; 84 such proteins are not assigned to COGs, and a considerable fraction of them are specific for only these two genomes. At least half of these predicted genes are larger than 100 amino acids, and orthologs have a similar length and presumably encode functional proteins. One of the important differences between the two *Wolbachia* for which genomes are available is that *w*Bm is apparently a mutualistic symbiont of its host, while *w*Mel is parasitic. The smaller size of the *w*Bm genome might be related to this difference. *w*Mel likely has to retain genes required for infecting host cells and avoiding host defense systems, whereas wBm may have lost many of these genes, as has been seen in organelles and other mutualistic symbionts such as the *Buchnera* symbionts of aphids.
Despite there being considerably fewer predicted genes in *w*Bm ([Table 1](#pbio-0030121-t001){ref-type="table"}), the metabolic capabilities of *w*Mel and *w*Bm are very similar. Unlike *w*Bm, *w*Mel has retained some enzymes for folate and pyridoxal phosphate biosynthesis, two subunits of cytochrome bd-type quinol oxidase, and a few additional enzymes for amino acid utilization (proline dehydrogenase and threonine aldolase). Among the genes unique to *w*Bm, there are two extracellular metallo-peptidases (Wbm0384, Wbm0742) that are only distantly related to counterparts in the *w*Mel genome. These results suggest a basic common strategy used by *w*Bm and *w*Mel during the evolution of their host symbiosis. In the case of *w*Bm, the basis of the interaction may be to provide essential vitamin cofactors, heme biosynthesis intermediates, and nucleotides while requiring amino acids and perhaps other nutrients supplied by the host.
Both *Wolbachia* have lost a considerable number of membrane biogenesis genes that make them apparently unable to synthesize lipid A, the usual component of proteobacterial membranes. However, a few differences do exist. For example, in *w*Mel there is a predicted gene belonging to the family of GDSL-like lipases (WD1297), similar to the major secreted phospholipase of Legionella pneumophila \[[@pbio-0030121-b159]\], which also has phospholipid-cholesterol acyltransferase activity. Its ortholog in *w*Bm is disrupted by a frameshift (Wbm0354 corresponds to the C-terminal portion of the gene). However, it is still possible that, similar to E. chaffeensis and A. phagocytophilum \[[@pbio-0030121-b160]\], *w*Bm and *w*Mel incorporate cholesterol into their cell walls. Furthermore, *w*Mel retains several genes absent in *w*Bm that might be involved in cell wall biosynthesis. These include a small gene cluster (WD0611--WD0613) and several other enzymes (WD0620, WD0133, WD0431), suggesting that *w*Mel might produce peptidoglycan modified with an oligosaccharide chain, while *w*Bm makes unmodified peptidoglycan. Possible differences in peptidoglycan structure may be additionally predicted by the already mentioned loss of FtsW--FtsI genes in *w*Bm and their presence in *w*Mel. These differences may reflect the occurrence of a mutualistic lifestyle (*w*Bm) in contrast to a parasitic lifestyle (*w*Mel).
Somewhat surprisingly, no recent apparent horizontally transferred genes from hosts were found in either *Wolbachia* genome. Moreover, an aforementioned WASP protein homolog, apparently acquired by a common ancestor of *Wolbachia* and *Rickettsia* from an animal host, is disrupted in the *w*Mel genome (WD0811). However, in *w*Mel there are two proteins encoded in the region of the prophages (WD0443, WD0633) that have "eukaryotic" OTU-like protease domains with their predicted catalytic residues apparently intact \[[@pbio-0030121-b161]\]. Proteases from this family are shown to be involved in ubiquitin pathways \[[@pbio-0030121-b162]\]. To our knowledge, this is a rare appearance of these proteases in prokaryotic genomes, although they are present in the genomes of C. pneumoniae \[[@pbio-0030121-b161]\] and in a closely related genome, Chlamydophila caviae (CCA00261).
Conclusions {#s2g}
-----------
Comparing the genomes of *w*Bm and *Rickettsia* to those of gamma-proteobacterial symbionts points to general similarities and distinctions in the evolution of endosymbionts. The genomes of R. conorii and *Wolbachia* species contain numerous repeats of various classes that are much more abundant than in the gamma-proteobacterial endosymbionts ([Table 1](#pbio-0030121-t001){ref-type="table"}). This correlates with the minimal gene colinearity between the genomes of *Wolbachia* and *Rickettsia* \[[@pbio-0030121-b103],[@pbio-0030121-b114],[@pbio-0030121-b163]\] ([Figure 5](#pbio-0030121-g005){ref-type="fig"}). By contrast, gamma-proteobacterial endosymbionts share a variety of operons with one another, and even with free-living relatives, despite the dramatic gene loss. Furthermore, gamma-proteobacterial endosymbionts (with the exception of *Wigglesworthia*) have lost crucial genes involved in recombinational repair, whereas almost no gene loss in this functional class was observed in *Wolbachia* or *Rickettsia* spp. Active recombination between repeats might have led to both gene loss and genome shuffling in *Wolbachia* and *Rickettsia* spp., whereas other mechanisms of genome reduction were probably involved in the evolution of gamma-proteobacterial endosymbionts \[[@pbio-0030121-b109],[@pbio-0030121-b120],[@pbio-0030121-b121],[@pbio-0030121-b122],[@pbio-0030121-b123],[@pbio-0030121-b164]\]
Comparative genome analysis highlights the different metabolic capabilities that render endosymbionts indispensable to their hosts \[[@pbio-0030121-b108],[@pbio-0030121-b119],[@pbio-0030121-b121]\]. For example, *Buchnera* and *Blochmannia* retain a nearly complete repertoire of amino acid biosynthesis pathways and supply amino acids to their insect hosts \[[@pbio-0030121-b110],[@pbio-0030121-b112]\]. In contrast, *w*Bm, *w*Mel, and *Wigglesworthia* \[[@pbio-0030121-b103],[@pbio-0030121-b108]\] have lost nearly all of these pathways but retain the pathways for the biosynthesis of nucleotides and some coenzymes ([Table 3](#pbio-0030121-t003){ref-type="table"}). Thus, endosymbiotic organisms in different divisions of proteobacteria independently evolved distinct strategies for symbiont--host interactions.
Genomic analysis of the alpha-proteobacterium *w*Bm, the first sequenced endosymbiont from a human parasitic nematode, provides new insights into the evolution of intracellular bacterial symbiosis and clues to the role of *Wolbachia* in the mutualistic relationship with the nematode. It is anticipated that continued genome analysis of nematodes and their endosymbionts will provide novel targets for antimicrobials aimed at the elimination of human filarial parasites.
Materials and Methods {#s3}
=====================
[Materials and Methods](#s3){ref-type="sec"} {#s3a}
--------------------------------------------
B. malayi microfilaria worms were purchased from TRS Labs (Athens, GA, United States) for preparation of DNA. Because of the difficulties in obtaining purified *Wolbachia* DNA from the B. malayi host, bacterial artificial chromosome (BAC) libraries were created \[[@pbio-0030121-b114]\]. From these libraries, a minimum tiling path of 21 *Wolbachia* BACs was created and used for subcloning into plasmid vectors for genomic sequencing. This ordered BAC approach was useful in the assembly phase of the project because of the highly repetitive nature of this genome.
For plasmid library generation, equal amounts of BAC DNAs were pooled and 50 μg of DNA from the pool was sheared into 2.0--3.0 kb fragments (HydroShear device, GeneMachines, Genomic Solutions, Ann Arbor, Michigan, United States). Sheared DNA was purified from a 0.7% agarose gel, blunted, and cloned into cleaved, dephosphorylated plasmid vectors. Libraries were generated containing DNA from 1 to 9 BACs.
Plasmid DNA was isolated by a modified alkaline lysis protocol. Sequencing reactions were performed at Integrated Genomics (Chicago, Illinois, United States) using the DYEnamic ET Dye Terminator Cycle Sequencing Kit (Amersham Biosciences, Little Chalfont, United Kingdom). Unincorporated dye was removed by isopropanol precipitation as recommended by the manufacturer. Samples were run on MegaBace 1000 (Amersham Biosciences) sequencers; 87% of plasmid sequencing reactions were successful. The genome was sequenced to an average coverage of 10.7X and at 2X minimum coverage (at least once in each direction) and assembled.
The sequence was assembled into contigs by using PHRED--PHRAP--CONSED \[[@pbio-0030121-b165],[@pbio-0030121-b166],[@pbio-0030121-b167]\], and gaps were initially closed by primer walking (1,766 reactions). Regions considered to be potential frame shifts or sequencing errors after the first round of annotation were resequenced from direct genomic PCR products. The completed sequence was used to identify homologous sequences in the independent ongoing B. malayi sequence project (TIGR parasites genome database: <http://www.tigr.org/tdb/e2k1/bma1/> \[[@pbio-0030121-b126]\]). The sequence of one BAC had been previously determined \[[@pbio-0030121-b163]\]. The final assembly was in full agreement with the BAC physical map \[[@pbio-0030121-b114]\].
Integrated Genomics ERGO software \[[@pbio-0030121-b168]\] and other software programs \[[@pbio-0030121-b169]\] were used for ORF calling, gene identification, and feature recognition. Computational analysis of the genome sequence was performed as previously described. Briefly, the tRNA genes were identified using the tRNA-SCAN program \[[@pbio-0030121-b170]\], and the rRNA genes were identified using the BLASTN program \[[@pbio-0030121-b171]\]. For the identification of the protein-coding genes, the genome sequence was conceptually translated in six frames to generate potential protein products of ORFs longer than 100 codons. These potential protein sequences were compared to the database of proteins from the COG database using COGNITOR \[[@pbio-0030121-b172]\].
After manual verification of the COG assignments, the validated COG members from *w*Bm were called as protein-coding genes. The COG assignment procedure was repeated with ORFs of greater than 60 codons from the intergenic regions. Additionally, the potential protein sequences were compared to the nonredundant protein sequence database using the BLASTP program \[[@pbio-0030121-b171]\] and to a six-frame translation of unfinished microbial genomes using the TBLASTN program \[[@pbio-0030121-b171]\], and those sequences that produced hits with E (expectation) values less than 0.01 were added to the protein set after an examination of the alignments. Finally, protein-coding regions were predicted using the GeneMarkS program \[[@pbio-0030121-b173]\]. After manual refinement, the genes predicted with these methods in the regions between evolutionarily conserved genes were added to produce the final protein set. Protein function prediction was based primarily on the COG assignments. In addition, searches for conserved domains were performed using the Conserved Domain Database (CDD) search option of BLAST (<http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi>) and the SMART system \[[@pbio-0030121-b174]\], and in-depth, iterative database searches were performed using thePSI-BLAST program \[[@pbio-0030121-b175]\]. The KEGG database \[[@pbio-0030121-b176]\] (<http://www.genome.ad.jp/kegg/metabolism.html>) and the Integrated Genomics ERGO database pathway collection \[[@pbio-0030121-b168]\] were used, in addition to the COGs, for the reconstruction of metabolic pathways. Paralogous protein families were identified by single-linkage clustering after comparing the predicted protein set to itself using the BLASTP program \[[@pbio-0030121-b171]\]. Signal peptides in proteins were predicted using the SignalP program \[[@pbio-0030121-b177]\], and transmembrane helices were predicted using the MEMSAT program \[[@pbio-0030121-b178]\]. Gene orders in bacterial genomes were compared using the Lamarck program \[[@pbio-0030121-b179]\].
Two closely related genome sequences were completed and published since the above comparative analysis was undertaken \[[@pbio-0030121-b180],[@pbio-0030121-b181]\].
Supporting Information {#s4}
======================
Data Access {#s4a1}
-----------
DNA sequence, ORF, as well as annotation and positional information tables, are available at the following Web site: <http://tools.neb.com/wolbachia/>.
Accession Number {#s4a2}
----------------
The genome sequence was deposited in GenBank (<http://www.ncbi.nlm.nih.gov/>) under accession number AE017321.
We gratefully acknowledge Drs. L. McReynolds, L. Raleigh, and R. Roberts for intellectual discussions and encouragement throughout this project. We thank the members of the Filarial Genome Project and *Wolbachia* Consortium communities for their discussions and support, in particular Drs. S. O\'Neill, J. Werren, M. Blaxter, M. Taylor, A. Scott, S. Williams, and C. Bandi. We also thank Drs. J. Eisen, H. Ochman, J. Wernegreen, S. Bordenstein, A. Osterman and R. Overbeek for insightful comments. We gratefully acknowledge helpful comments from three anonymous reviewers. Financial support was provided by internal funding from New England Biolabs, Inc. Dedicated to the memory of Mikhail Mazur.
**Competing interests.** The authors have declared that no competing interests exist.
**Author contributions.** J. Foster, A. Lapidus, E. Ghedin, V. Joukov, K. Tsukerman, and B. Slatko conceived and designed the experiments. J. Foster, M. Ganatra, I. Kamal, J. Ware, A. Bhattacharyya, V. Kapatral, J. Ingram, L. Moran, A. Lapidus, E. Goltsman, V. Joukov, O. Ostrovskaya, K. Tsukerman, M. Mazur, and B. Slatko performed the experiments. J. Foster, M. Ganatra, I. Kamal, J. Ware, K. Makarova, N. Ivanova, A. Bhattacharyya, V. Kapatral, S. Kumar, J. Posfai, T. Vincze, A. Lapidus, M. Omelchenko, N. Kyrpides, E. Ghedin, E. Goltsman, V. Joukov, K. Tsukerman, M. Mazur, E. Koonin, S. Wang, and B. Slatko analyzed the data. J. Foster, M. Ganatra, I. Kamal, J. Ware, K. Makarova, N. Ivanova, A. Bhattacharyya, S. Kumar, J. Posfai, J. Ingram, A. Lapidus, N. Kyrpides, E. Ghedin, E. Goltsman, D. Comb, E. Koonin, and B. Slatko contributed reagents/materials/analysis tools. J. Foster, M. Ganatra, J. Ware, K. Makarova, N. Ivanova, S. Kumar, J. Ingram, L. Moran, D. Comb, E. Koonin, and B. Slatko wrote the paper.
¤a Current address: Biochemistry Department, Faculty of Science Ain Shams University, Abassiah, Cairo, Egypt
¤b Current address: Joint Genome Institute, Walnut Creek, California, United States of America
¤c Current address: Department of Molecular Biology and Microbiology, Case Western Reserve University School of Medicine, Cleveland, Ohio, United States of America
Citation: Foster J, Ganatra M, Kamal I, Ware J, Makarova K, et al. (2005) The *Wolbachia* genome of *Brugia malayi:* Endosymbiont evolution within a human pathogenic nematode. PLoS Biol 3(4): e121.
BAC
: bacterial artificial chromosome
COGs
: clusters of orthologous groups of proteins
kbp
: kilobasepairs
LPS
: lipopolysaccharide
*meso*-DAP
: *meso*-diaminopimelate
ORF
: open reading frame
TCA
: tricarboxylic acid
TLR4
: toll-like receptor 4
*w*Bm
: *Wolbachia* endosymbiont of *Brugia malayi*
*w*Mel
: *Wolbachia* endosymbiont of Drosophila melanogaster
|
PubMed Central
|
2024-06-05T03:55:54.999865
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069646/",
"journal": "PLoS Biol. 2005 Apr 29; 3(4):e121",
"authors": [
{
"first": "Jeremy",
"last": "Foster"
},
{
"first": "Mehul",
"last": "Ganatra"
},
{
"first": "Ibrahim",
"last": "Kamal"
},
{
"first": "Jennifer",
"last": "Ware"
},
{
"first": "Kira",
"last": "Makarova"
},
{
"first": "Natalia",
"last": "Ivanova"
},
{
"first": "Anamitra",
"last": "Bhattacharyya"
},
{
"first": "Vinayak",
"last": "Kapatral"
},
{
"first": "Sanjay",
"last": "Kumar"
},
{
"first": "Janos",
"last": "Posfai"
},
{
"first": "Tamas",
"last": "Vincze"
},
{
"first": "Jessica",
"last": "Ingram"
},
{
"first": "Laurie",
"last": "Moran"
},
{
"first": "Alla",
"last": "Lapidus"
},
{
"first": "Marina",
"last": "Omelchenko"
},
{
"first": "Nikos",
"last": "Kyrpides"
},
{
"first": "Elodie",
"last": "Ghedin"
},
{
"first": "Shiliang",
"last": "Wang"
},
{
"first": "Eugene",
"last": "Goltsman"
},
{
"first": "Victor",
"last": "Joukov"
},
{
"first": "Olga",
"last": "Ostrovskaya"
},
{
"first": "Kiryl",
"last": "Tsukerman"
},
{
"first": "Mikhail",
"last": "Mazur"
},
{
"first": "Donald",
"last": "Comb"
},
{
"first": "Eugene",
"last": "Koonin"
},
{
"first": "Barton",
"last": "Slatko"
}
]
}
|
PMC1069647
|
We all know that our memories are stored somehow in our brains. But exactly how do we remember the way to our office or what our mother looks like or the date we got married? Scientists attribute our ability to store apparently infinite numbers of memories for decades to long-lasting changes in the electrical, structural, and biochemical properties of neurons. One cellular mechanism proposed to be involved in the storage of memories---long-term potentiation---involves alterations in the strength of messages passed from one neuron to another across structures known as synapses.
The initiation of long-term potentiation is caused by activation of [n]{.smallcaps}-methyl-[d]{.smallcaps}-aspartate receptors on the receiving neuron and a subsequent increase in the intracellular calcium concentration in a region of the neuron that is called the postsynaptic density. The increase in calcium, in turn, activates the calcium/calmodulin-dependent protein kinase II (CaMKII). This enzyme seems to play a critical role in long-term potentiation, and has been proposed as one of the leading candidates to act as the molecular switch that maintains stable synapse-specific cellular changes. To fulfill this role, CaMKII would need to have stable UP and DOWN positions, or states, much like a light switch.
Xiao-Jing Wang and colleagues now provide a new analysis that strengthens the argument that CaMKII is a molecular switch involved in the storage of long-term neural changes. The activity of the CaMKII holoenzyme (the complete enzyme consisting of both regulatory and catalytic subunits) is controlled by its autophosphorylation state---the enzyme is able to add phosphate groups to specific amino acids within itself. Previous modeling studies have shown that the interplay between the autocatalytic addition of phosphate groups to CaMKII and the removal of phosphate groups by protein phosphatase-1 (PP1) enzymes produces two stable states of the CaMKII enzyme at basal free calcium levels. The DOWN state is unphosphorylated; the UP state is highly phosphorylated. When there is a transient high input of calcium, as happens when long-term potentiation is induced, the CaMKII enzyme flips from a DOWN state to a persistent UP state.
The questions that Wang and colleagues have now asked are what factors affect the stability of the state of this switch, and how many CaMKII holoenzymes are needed to construct a switch that could last a lifetime. These questions are important because a switch that could be spontaneously reset by small, random fluctuations of the conditions within the postsynaptic density would not be useful in maintaining stable long-term changes. The researchers have used a mathematical probabilistic modeling technique known as Monte Carlo simulation, together with the known biochemical and thermodynamic characteristics of CaMKII and PP1, to test how random fluctuations in the chemical reactions involved in the CaMKII/PP1 system change the state of the switch.
They report that switch state stability requires a balance between the phosphorylation and dephosphorylation rates of CaMKII, and that the turnover rate of the kinase---the replacement of old molecules with new ones---critically affects switch stability. However, their main finding is that the lifetime of states of the switch increases exponentially with the number of CaMKII holoenzymes that are present. This finding is important because experimental work by other researchers has estimated that there are about 30 CaMKII holoenzymes present in a typical postsynaptic density, and until Wang\'s team did their modeling it was unclear whether this number of holoenzymes could build a switch stable enough to last a lifetime. In fact, Wang and co-workers estimate that a switch containing as few as 15 holoenzymes can remain activated for longer than a human lifetime. Thus, the researchers conclude, CaMKII switches may indeed play a critical role in preserving our precious memories throughout our lives.
|
PubMed Central
|
2024-06-05T03:55:55.004749
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069647/",
"journal": "PLoS Biol. 2005 Apr 29; 3(4):e124",
"authors": []
}
|
PMC1069648
|
*"Great fleas have little fleas upon their backs, to bite \'em. And little fleas have lesser fleas and so on,* ad infinitum."---*Augustus DeMorgan, based on a Jonathan Swift poem*
More than a billion people are at risk for infection with filarial nematodes, parasites that cause elephantiasis, African river blindness, and other debilitating diseases in more than 150 million people worldwide. The nematodes themselves play host to bacteria that live within their cells, but in this case, the relationship is classic mutualism, with each benefiting from the other. Indeed, the Wolbachia bacterium is so crucial to its host nematode that apparently eradicating it with antibiotics severely compromises the nematode\'s ability to complete its life cycle within its human host. Thus, understanding the details of this symbiosis may help identify new strategies for controlling diseases caused by filarial nematodes. In a new study, Barton Slatko and colleagues present the complete DNA sequence of the Wolbachia pipientis strain within Brugia malayi, a parasitic nematode responsible for lymphatic filariasis, and analyze its genome for clues to the interdependence of the two species.
This Wolbachia genome is small, only about a million base pairs, and many metabolically critical genes have degraded through mutation to the point of uselessness. This phenomenon, called reductive evolution, is typical of long-term symbioses, as the two partners increasingly complement one another\'s biochemical activities, reducing the selection pressure on otherwise lethal mutations. Wolbachia\'s translational machinery and DNA repair equipment are largely intact. The bacterium appears to supply nucleotides to its host, as it contains complete pathways for biosynthesis of both purine and pyrimidine nucleotides. This is in contrast to Rickettsia, a close relative of Wolbachia and a mammalian parasite. Slatko and colleagues enumerate a variety of other pathways that have either been degraded or preserved, and highlight patterns in the genome structure through comparisons with both Rickettsia and another Wolbachia strain, found in fruit flies. For example, the two Wolbachia strains appear to have different membrane structures, possibly reflecting their different lifestyles (mutualistic versus parasitic).[](#pbio-0030148-g001){ref-type="fig"}
::: {#pbio-0030148-g001 .fig}
::: {.caption}
###### Over a billion people are at risk for infection by filarial nematodes, parasites that cause elephantiasis (Photo: Dr. Steven A. Williams)
:::
:::
Wolbachia can manufacture riboflavin and FAD, which are essential metabolic coenzymes and which do not appear to be made by its host. Conversely, it cannot synthesize amino acids and a variety of other vitamins and cofactors, and probably depends on the nematode to supply them. One discovery of possible significance is the presence in the bacterium of the synthetic pathway for heme---the oxygen-carrying iron component of hemoglobin. The nematode may require heme for synthesis of developmental hormones, so Wolbachia\'s heme pathway may be an inviting target for therapy against nematode infection. Since no new antifilarial has been developed in two decades, these results may quickly lead to new therapeutic strategies against these parasites.
|
PubMed Central
|
2024-06-05T03:55:55.005275
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069648/",
"journal": "PLoS Biol. 2005 Apr 29; 3(4):e148",
"authors": []
}
|
PMC1069649
|
Introduction {#s1}
============
The acquisition of complex motor sequences, such as swinging a golf club or playing the piano, can be thought of as reinforcement learning. This learning process requires the exploration of a range of motor actions and the concomitant evaluation of the resulting performance, reinforcing motor programs that lead to improved outcomes \[[@pbio-0030153-b01]\]. Similarly, juvenile songbirds explore a large range of vocalizations by continuously varying their song \[[@pbio-0030153-b02]\], utilizing auditory feedback to improve their performance \[[@pbio-0030153-b03]\]. Thus, song learning encompasses the two ingredients of reinforcement learning: exploratory motor behavior, and performance evaluation.
In the songbird, two main neural pathways are involved in song production and song learning ([Figure 1](#pbio-0030153-g001){ref-type="fig"}A). The "motor pathway" controls the vocal motor program through the hierarchical organization of several premotor nuclei \[[@pbio-0030153-b04]\]. A key nucleus in the motor pathway is the robust nucleus of the arcopallium (RA), which projects to brainstem nuclei controlling the vocal and respiratory muscles \[[@pbio-0030153-b05]\]. During singing, RA neurons in adult birds generate a highly stereotyped sequence of bursts \[[@pbio-0030153-b06],[@pbio-0030153-b07]\], which appear to be driven by precisely timed inputs from a higher premotor vocal area, nucleus HVC \[[@pbio-0030153-b08]\]. RA also receives input from the "anterior forebrain pathway" (AFP), a circuit homologous to the basal ganglia thalamo-cortical loops \[[@pbio-0030153-b09],[@pbio-0030153-b10]\] that may be involved in controlling motor behavior and stereotypy in mammals \[[@pbio-0030153-b11]\]. Lesions of the AFP in juvenile zebra finches have devastating effects on song development, whereas the same manipulations in adults have few short-term consequences for song production \[[@pbio-0030153-b12],[@pbio-0030153-b13]\].
::: {#pbio-0030153-g001 .fig}
Figure 1
::: {.caption}
###### Inactivation of LMAN Significantly Reduces Vocal Experimentation, Making the Otherwise Variable Song of the Juvenile Zebra Finch Highly Stereotyped
\(A) Two major pathways in the vocal control system of the songbird. The motor pathway (gray) includes motor cortex analogs HVC and RA, while the AFP (white), a basal ganglia thalamo-cortical circuit, consists of Area X, the dorsolateral anterior thalamic nucleus (DLM), and LMAN, which, in turn, projects to RA. To inactivate the output of the AFP, injections of TTX and muscimol (red bolus) were made into LMAN.
\(B) Examples of a juvenile zebra finch song (57 dph) showing large variability in the sequence and acoustic structure of song syllables.
\(C) Inactivating LMAN with TTX produces an immediate reduction of sequence and acoustic variability, revealing a highly stereotyped song produced by the motor pathway.
The song snippets shown in (B) and (C) are from consecutive song bouts, immediately before and 1 h after drug injection. Songs are displayed as spectral derivatives calculated as described \[[@pbio-0030153-b36]\]. The frequency range displayed is 0--8.6 kHz. For audio of song bouts before and during LMAN inactivation in this bird, refer to Audios [S1](#sa001){ref-type="supplementary-material"} and [S2](#sa002){ref-type="supplementary-material"}, and [S3](#sa003){ref-type="supplementary-material"} and [S4](#sa004){ref-type="supplementary-material"}, respectively.
:::
:::
While the critical importance of the AFP for song learning has been established, its specific role remains unknown \[[@pbio-0030153-b14]\]. It has been proposed that the AFP may be involved in comparing the auditory feedback of the bird\'s vocal output with a stored auditory template of the desired song---an evaluation process that could provide a corrective signal to the motor pathway needed for learning \[[@pbio-0030153-b15]\]. However, recent results showing that the firing patterns of neurons in the lateral magnocellular nucleus of the nidopallium (LMAN) of adult birds are insensitive to distorted auditory feedback have called this idea into question \[[@pbio-0030153-b16],[@pbio-0030153-b17]\]. Here we test the alternative hypothesis that, in juvenile songbirds, LMAN is involved in generating vocal variability \[[@pbio-0030153-b18]\]---the other important ingredient of reinforcement learning.
Results {#s2}
=======
Our approach was to transiently inactivate LMAN in juvenile zebra finches (*n* = 7 birds, see [Materials and Methods](#s4){ref-type="sec"}), and observe whether and how their songs were affected. Birds were briefly head-restrained, and injections of a sodium channel blocker, tetrodotoxin (TTX, 30 nl, 50 μM), were made in LMAN in both hemispheres, inactivating the nucleus (see Figures [S1](#sg001){ref-type="supplementary-material"} and [S2](#sg002){ref-type="supplementary-material"}). After injections, birds were returned to a sound-isolated chamber, where they typically began to sing after 0.5--1.5 h. In all birds probed, LMAN inactivation resulted in an immediate loss of acoustic variability across song renditions. The effect was particularly dramatic in birds at an early stage of song development (approximately 55 d post hatch \[dph\]) because these birds normally exhibit greater song variability ([Figures 1](#pbio-0030153-g001){ref-type="fig"}B, [1](#pbio-0030153-g001){ref-type="fig"}C, and [S3](#sg003){ref-type="supplementary-material"}; [Audios S1--S4](#sa001){ref-type="supplementary-material"}).
To quantify song variability, experiments were carried out in slightly older birds with less sequence and acoustic variability (*n* = 6 birds; age range, 59--72 dph) ([Figure 2](#pbio-0030153-g002){ref-type="fig"}). This allowed us to reliably identify song syllables, the basic acoustic units of zebra finch song, across song renditions ([Figure 2](#pbio-0030153-g002){ref-type="fig"}A). The variability score *(V)---*a measure reflecting the acoustic variability of a syllable across renditions (see [Materials and Methods](#s4){ref-type="sec"})---was calculated for all identified syllables before and after TTX injection. Without exception, the syllables showed a highly significant reduction in variability as a consequence of LMAN inactivation ([Figure 2](#pbio-0030153-g002){ref-type="fig"}B; *n* = 25 syllables; 〈*V*〉~before~ = 0.46, 〈Δ*V*〉 = 0.2; *p* ~ave~ \< 0.0001, *t*-test). In fact, the juvenile song after inactivation was significantly less variable than songs of adult zebra finches singing undirected song (i.e., songs not directed to a female; [Figure 2](#pbio-0030153-g002){ref-type="fig"}D; *p* \< 0.001, *t*-test). LMAN inactivation also eliminated 75% of the difference in mean variability between juvenile song and adult directed song---the most highly stereotyped form of song \[[@pbio-0030153-b19]\].
::: {#pbio-0030153-g002 .fig}
Figure 2
::: {.caption}
###### Analysis of the Effect of Bilateral LMAN Inactivation on Song Variability
\(A) Consecutive renditions of a repeating song motif of 0.5 s duration in a juvenile bird (59 dph) arranged vertically. Note the large variations in acoustic structure within individual syllables before LMAN inactivation (left). Following TTX injection into LMAN, the acoustic variability is dramatically reduced (middle), only to return to the original level by the following day (right). Numbers below each column indicate the variability index (See [Materials and Methods](#s4){ref-type="sec"} section) calculated for the four renditions of the syllables shown.
\(B) Scatter plot of variability scores before and during LMAN inactivation with TTX (red) and muscimol (blue). Also shown are results for bilateral TTX injection into MMAN (black; see text), and saline injection into LMAN (green).
\(C) Time course of variability reduction following TTX (red) and muscimol (blue) injections show a time dependence that reflects the known in vivo pharmacology of the respective agents. Data were averaged over four identified syllables and taken from the same bird over consecutive days (dph = 70 and 71; muscimol inactivation followed by TTX inactivation).
\(D) Distribution of variability scores for all syllables analyzed in the TTX and muscimol experiments (25 unique syllables, six birds) before (black) and during (red) LMAN inactivation in juvenile birds. Shown for comparison are the variability scores for adult zebra finch syllables (18 syllables, 4 birds; undirected song, green; directed song, light blue). Dots represent raw data, while the lines are smoothed running averages.
\(E) TTX inactivation of LMAN significantly increased syllable sequence stereotypy. Sequence stereotypy scores (see [Materials and Methods](#s4){ref-type="sec"}) for six birds before (black) and after (red) TTX injections into LMAN. For comparison, the average stereotypy score for adult birds singing directed song was 0.95 (*n* = 4 birds).
:::
:::
To verify that the loss of variability resulted from silencing LMAN neurons, and not from inactivating fibers of passage near LMAN, a GABA~A~ receptor agonist (muscimol, 30 nl, 25 mM) was injected bilaterally into LMAN (*n* = 2 birds; 66 and 70 dph). Again, all syllables showed a dramatic reduction in variability after injection (*n* = 8 syllables; 〈*V*〉~before~ = 0.43, 〈Δ*V*〉 = 0.16; *p* ~ave~ \< 0.0001, *t*-test). While the reduction in acoustic variability was similar to that resulting from TTX injections ([Figure 2](#pbio-0030153-g002){ref-type="fig"}B), the duration of the effect of muscimol was substantially shorter than observed for TTX ([Figure 2](#pbio-0030153-g002){ref-type="fig"}C). This difference in temporal profile was in good agreement with the known in vivo pharmacology of TTX and muscimol \[[@pbio-0030153-b20],[@pbio-0030153-b21]\], suggesting a direct link between suppression of spiking activity in LMAN and loss of song variability.
An additional effect of LMAN inactivation was a significant reduction in sequence variability, a measure of the variability in syllable ordering ([Figure 2](#pbio-0030153-g002){ref-type="fig"}E; *p* \< 0.005, paired *t*-test; see [Materials and Methods](#s4){ref-type="sec"}). In fact, the sequential ordering of syllables after TTX injection was comparable in stereotypy to that of adult song. Thus, LMAN activity may influence sequence generation, possibly through an indirect feedback pathway going from RA to HVC, the putative sequence generator \[[@pbio-0030153-b06],[@pbio-0030153-b08],[@pbio-0030153-b22]\].
We confirmed that the loss of song variability following injections into LMAN did not result from diffusion of the drugs into the medial magnocellular nucleus of the nidopallium (MMAN), a nucleus approximately 1.25 mm medial from LMAN with projections to HVC. Bilateral injections of TTX into MMAN, done in the same birds in which LMAN injections were previously made, had no significant effect on acoustic variability ([Figure 2](#pbio-0030153-g002){ref-type="fig"}B).
We next considered the neural mechanisms by which LMAN affects variability in the motor pathway. One intriguing possibility is that song variability is driven by fast synaptic input from LMAN. If true, then acoustic variability should be accompanied by variability in the firing patterns of RA-projecting LMAN neurons. To test this idea explicitly, we recorded single-unit signals from 29 LMAN neurons in singing juvenile birds (*n* = 3 birds; age range, 62--79 dph) ([Figure 3](#pbio-0030153-g003){ref-type="fig"}). In all, 17 of these were antidromically identified as RA-projecting LMAN neurons (see [Materials and Methods](#s4){ref-type="sec"}). These neurons exhibited song-related changes in firing rate (spontaneous activity, 12 ± 4 Hz; during singing, 39 ± 6 Hz \[mean ± standard deviation\]), and generated significantly more bursts during singing ([Figure 3](#pbio-0030153-g003){ref-type="fig"}C). Raster plots of the spike trains aligned to the song motif showed that the patterns of spikes and bursts generated by individual neurons were different each time the bird sang ([Figure 3](#pbio-0030153-g003){ref-type="fig"}A and [3](#pbio-0030153-g003){ref-type="fig"}B).
::: {#pbio-0030153-g003 .fig}
Figure 3
::: {.caption}
###### Song-Aligned Firing Patterns of RA-Projecting LMAN Neurons in Singing Juvenile Zebra Finches Are Highly Variable
\(A) Three successive renditions of a 67-d-old bird\'s song motif. Displayed under each spectrogram is the simultaneously recorded voltage waveform of an antidromically identified RA-projecting LMAN neuron (verified by collision testing). Average syllable variability for the three motifs is 0.31. Motif alignment was done at the onset (yellow lines) of syllable C.
\(B) Raster plot showing the spike patterns for 50 consecutive motif renditions for the same cell as in (A). The motifs from (A) are indicated in green.
\(C) Relative frequency of inter-spike intervals during singing (black) and non-singing (blue) for all the 17 identified projection neurons (units are intervals per second; bin size is 0.04 log units).
\(D) Distribution of spike-train correlations across all pairs of motifs for the cell in (B) (solid red line). Correlations calculated with random time shifts added to the spike trains have a similar distribution (dashed red line; see [Materials and Methods](#s4){ref-type="sec"}). Also shown is the correlation distribution for the population of identified projection neurons (solid black line; mean correlation indicated by solid arrowhead), and for the population with random time shifts added (dashed black line). In comparison, spike trains of neurons in premotor nucleus RA of the adult bird are highly stereotyped (from \[[@pbio-0030153-b23]\]; mean correlation indicated by open arrowhead).
:::
:::
Correlations in the spike trains across different renditions of the motif were small (0.054 ± 0.34 \[mean ± standard deviation\]) compared to those observed in premotor neurons of adult birds (0.90 ± 0.1) \[[@pbio-0030153-b07]\]. We also compared the correlation distributions to those calculated after random time shifts were added to the spike trains (see [Materials and Methods](#s4){ref-type="sec"}). In general, the correlation distributions of the randomized spike trains were very similar to those calculated for the motif-aligned spike trains ([Figure 3](#pbio-0030153-g003){ref-type="fig"}D), confirming that the firing patterns of LMAN neurons are highly variable. Nevertheless, in 13 out of the 17 identified RA-projecting neurons the correlation distributions were still significantly different from those of the randomly shuffled spike trains (*p* \< 0.01, Kolmogorov-Smirnov test), suggesting that while LMAN activity is highly variable, it is not completely random with respect to the song.
Guided by the neural data, we next tested the hypothesis that LMAN drives song variability by providing excitatory glutamatergic input to RA---which in the zebra finch is mediated almost exclusively by [N]{.smallcaps}-methyl-[D]{.smallcaps}-aspartate (NMDA)--type receptors \[[@pbio-0030153-b24]\]. In contrast, glutamatergic inputs to RA from HVC are mediated by a mixture of NMDA and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)--type receptors ([Figure 4](#pbio-0030153-g004){ref-type="fig"}A) \[[@pbio-0030153-b25]\]. Thus, if LMAN drives song variability through glutamatergic input to RA, then blocking NMDA receptors should reduce this variability, while sparing the AMPA-mediated drive from HVC. In line with our hypothesis, bilateral injections of the NMDA receptor antagonist 2-amino-5-phosphonovalerate (AP5, 50 nl, 30 mM) into RA significantly reduced acoustic variability in all song syllables examined ([Figure 4](#pbio-0030153-g004){ref-type="fig"}B and [4](#pbio-0030153-g004){ref-type="fig"}C; *n* = 4 birds; age range, 57--73 dph; 11 syllables; 〈*V*〉~before~ = 0.47, 〈Δ*V*〉 = 0.16; *p* ~ave~ \< 0.0001, *t*-test). The time course of the variability reduction ([Figure 4](#pbio-0030153-g004){ref-type="fig"}D) was consistent with the temporal profile of AP5 effects seen in other in vivo studies \[[@pbio-0030153-b26]\].
::: {#pbio-0030153-g004 .fig}
Figure 4
::: {.caption}
###### Bilateral Injections of the NMDA Receptor Antagonist AP5 into RA Significantly Reduced Song Variability
\(A) Excitatory synaptic inputs to RA from LMAN and HVC are mediated by a different mix of glutamate receptor types (see text). Using AP5 we could block LMAN input while only partially inactivating HVC input.
\(B) Eight sequential renditions of one song syllable in a juvenile zebra finch (63 dph) before and after AP5 injection into RA. Note the rapid fluctuations in pitch, the appearance of noisy acoustic structure, and variations in syllable duration before injection. The average variability scores *(V)* before and after injections for the eight shown syllable renditions were 0.50 and 0.25, respectively.
\(C) Following injection of AP5 into RA, fluctuations in acoustic structure were substantially reduced. Variability scores of 11 syllables in four birds before and after injection of AP5 into RA.
\(D) Time course of acoustic variability following drug injection averaged over all identifiable syllables for the bird in (B).
:::
:::
Given that AP5 has effects beyond blocking LMAN input to RA, it may influence the song in ways other than reducing variability. To examine whether AP5 injections affected the acoustic structure of syllables, we compared the acoustic features of syllables after AP5 injection to the same syllables before injection (average similarity score 78.0, 11 syllables; see [Materials and Methods](#s4){ref-type="sec"}). In comparison, the average similarity score across renditions of the same syllables prior to injection was 77.7, suggesting that the effect of AP5 injection was largely limited to song variability.
Discussion {#s3}
==========
Previous studies have shown that permanent LMAN lesions in the juvenile bird disrupt song learning and result in an impoverished and prematurely stereotyped song \[[@pbio-0030153-b12],[@pbio-0030153-b13]\]. Such lesions are known to produce synaptic maturation in RA within a few days \[[@pbio-0030153-b27]\], perhaps because of a loss of neurotrophic input from LMAN \[[@pbio-0030153-b12],[@pbio-0030153-b13]\]. Because of the long delay from lesioning to singing (often several days), these studies could not address whether increased stereotypy was caused by synaptic reorganization in RA, or by a more immediate mechanism such as the loss of fast synaptic input from LMAN. In our experiments, we observe singing within an hour after injection, and find that LMAN inactivation reduces song variability reversibly and on a short timescale. This observation implies that, in addition to slow neurotrophic effects, LMAN acts on RA rapidly to drive or control song variability, a necessary ingredient of reinforcement learning. Thus, our results suggest that the loss of vocal plasticity following permanent lesions of LMAN may, in part at least, be due to the immediate loss of exploratory behavior.
What is the mechanism by which neural activity in LMAN controls motif-to-motif variability in the song? Our experiments tested the hypothesis that fluctuations in the song are driven directly by synaptic input from LMAN \[[@pbio-0030153-b25]\]. In this view, the premotor circuit generates a stereotyped song sequence upon which the AFP acts to drive variations. This hypothesis requires that neural activity in LMAN be highly variable across different song motifs, a prediction that was borne out by our recordings in LMAN (see [Figure 3](#pbio-0030153-g003){ref-type="fig"}). In comparison, premotor neurons in adult birds (singing song of comparable stereotypy to our LMAN-inactivated juvenile birds) generate extremely stereotyped, song-locked spike patterns \[[@pbio-0030153-b06],[@pbio-0030153-b07] [@pbio-0030153-b08]\]. In itself, the result that LMAN neurons are only weakly time-locked to the song may not be surprising. The significance of this observation becomes apparent when considering that these neurons send excitatory projections to the motor pathway, and that they are necessary for the expression of song variability as demonstrated by our inactivation results. Together with the finding that electrical stimulation of LMAN in adult birds can drive transient changes in the song \[[@pbio-0030153-b19]\], these observations make LMAN a likely source for the variability in the premotor pathway.
Because LMAN input to RA neurons is mediated almost exclusively by NMDA receptors, another strong prediction of our hypothesis was that blockade of NMDA receptors in RA should reduce song variability. Our results from the injection of AP5 into RA confirmed this. However, given the presence of NMDA receptors in the projection from HVC to RA \[[@pbio-0030153-b24]\], and perhaps in recurrent connections within RA, blockade of NMDA receptors is likely to have effects on RA circuitry beyond the loss of direct synaptic input from LMAN. Thus, this experiment cannot preclude other hypotheses---for example, that LMAN acts to regulate stochastic processes intrinsic to the premotor circuit, through some yet unknown mechanism.
Further support for the idea that LMAN can drive song variations comes from studies in the adult zebra finch. Song-related neural activity in LMAN is variable also in the adult bird, and this variability has been shown to be larger during undirected as compared to directed singing \[[@pbio-0030153-b27],[@pbio-0030153-b28]\]. A recent study \[[@pbio-0030153-b19]\] linked the increased neural variability in LMAN during undirected singing to an increase in motif-to-motif variability in song features (see also [Figure 2](#pbio-0030153-g002){ref-type="fig"}D).
How does the role and function of LMAN change as song variability is reduced during learning and finally during song crystallization? To the extent that the variability of LMAN firing patterns in the adult bird during undirected song \[[@pbio-0030153-b28]\] is similar to that in the juvenile bird, an essential part of song development may be a reduction of the gain by which LMAN drives RA. This could occur as a result of synaptic changes within RA that weaken input from LMAN and/or strengthen the projections from HVC. While there is evidence that this may indeed occur \[[@pbio-0030153-b26],[@pbio-0030153-b29]\], more experiments are needed to establish how the developmental reduction in song variability is related to changes in song circuitry.
Reinforcement learning requires that variability in the motor output be accompanied by a mechanism that evaluates the resulting performance. In the songbird, such an evaluation signal could be sent directly to the motor system (e.g., to RA), perhaps via a neuromodulator \[[@pbio-0030153-b30] [@pbio-0030153-b31]\], to reinforce the states of the motor pathway that lead to a better-than-expected match to the memorized template. A reinforcement signal could also be sent to the AFP to shape or regulate the fluctuations introduced into the motor pathway via LMAN. This would make LMAN more than a simple "noise generator," allowing it to bias vocal fluctuations in the direction of the desired song. Such bias is suggested by the presence of small but significant correlations in the motif-aligned firing pattern of LMAN neurons (see [Figure 3](#pbio-0030153-g003){ref-type="fig"}). This bias could permit a more efficient exploration of motor space, and even allow LMAN activity to drive plastic changes in the motor circuitry.
The exploratory motor behavior exhibited by juvenile songbirds may also provide general insights into how the brain generates fluctuations required for learning. Such fluctuations could be generated within the motor pathway or by brain regions projecting to it, and could result from stochastic processes, such as randomness in synaptic release \[[@pbio-0030153-b32]\], noise propagated by summation of irregular patterns of inhibitory postsynaptic potentials and excitatory postsynaptic potentials \[[@pbio-0030153-b33]\], or complex collective dynamics of the neuronal network \[[@pbio-0030153-b34]\]. Our results strongly suggest that, whatever the detailed biophysical mechanisms, the neural circuits generating these fluctuations are located outside the motor pathway in a specialized pathway involving the basal ganglia. The output of this circuit acts on the motor pathway, allowing the song system to explore the vocal space in a purposeful manner. Whether inducing exploratory motor behavior is a general feature of basal ganglia circuits is an intriguing idea that remains to be explored.
Materials and Methods {#s4}
=====================
{#s4a}
### Subjects {#s4a1}
Subjects were juvenile male zebra finches (54--79 dph). Birds were obtained from the Massachusetts Institute of Technology zebra finch breeding facility (Cambridge, Massachusetts), and from the aviary at the Rockefeller Field Research Station (Millbrook, New York). The care and experimental manipulation of the animals were carried out in accordance with guidelines of the National Institutes of Health and were reviewed and approved by the Massachusetts Institute of Technology Institutional Animal Care and Use Committee.
### Reversible inactivation {#s4a2}
Birds underwent a brief surgery to attach to the skull a means of restraining the head during drug injections. The animals were anesthetized with isoflurane (2%) and placed in a stereotaxic apparatus (MyNeuroLab.com, St. Louis, Missouri, United States). Two stainless-steel screws (\#0--80 6 mm long) were secured to the skull with dental acrylic. Small holes (approximately 300 μm in diameter) were drilled through the cranium bilaterally over LMAN or MMAN, or RA using stereotaxic coordinates. The holes were covered with a thin layer of Kwik-Kast (World Precision Instruments, Sarasota, Florida, United States). The animals were then placed in a custom sound-isolation chamber where they began to sing prolifically after a few days---typically 200--1,000 song motifs per hour.
Inactivation of song control nuclei in the singing bird was carried out by placing the bird, unanesthetized, in a small foam restraint and attaching the head-mounted screws to a metal plate bolted to the stereotaxic apparatus. The Kwik-Kast over the cranial holes was removed, and 30 nl of TTX (50 μM, \#T5651, Sigma, St. Louis, Missouri, United States) or muscimol (25 mM, \#M1523, Sigma) was injected bilaterally into the brain region of interest using a Nanoject II injector (Drummond Scientific, Broomall, Pennsylvania, United States). The procedure of injecting the birds took approximately 10 min. Experimental confirmation of the physiological effects of TTX injections showed that LMAN was likely completely inactivated after our injections (see [Figure S2](#sg002){ref-type="supplementary-material"}). Regions immediately surrounding LMAN were also affected, and we cannot rule out an indirect contribution from the partial inactivation of these regions. For inactivation of NMDA-mediated synapses in RA, AP5 (\#A5282, Sigma) was injected bilaterally into RA (50 nl, 30 mM). The injection site was guided by electrophysiological recordings of spontaneous activity in RA.
Injected solutions also contained dye-conjugated dextrans (\#D22912, Molecular Probes, Eugene, Oregon, United States). All injection sites were verified by histological examination and were found to be within the target nucleus (see [Figure S1](#sg001){ref-type="supplementary-material"}), except for TTX injections in LMAN in two birds: one in which the LMAN injection site in one hemisphere was found to be approximately 100 μm anterior to the edge of LMAN, the other in which the injections were approximately 200 μm posterior to LMAN, but right in the middle of the fiber tract leading from LMAN to RA. The results from these birds were similar to those from other birds, and were included in the analysis.
### Chronic neural recordings in LMAN {#s4a3}
Experiments were timed such that the birds were at an age at which they produced readily identifiable syllable sequences, yet showed variable acoustic syllable structure across song renditions. Recordings were carried out using a motorized microdrive described previously \[[@pbio-0030153-b35]\]. Cells were isolated by searching for spontaneous or antidromically evoked spiking activity; units typically had signal-to-noise ratios greater than 10:1. Antidromic identification of RA-projecting LMAN neurons was carried out with a bipolar stimulating electrode implanted in RA using techniques described previously for antidromic identification of RA-projecting HVC neurons \[[@pbio-0030153-b08]\]. Neurons exhibiting a short-latency antidromic spike (\<5 ms) with a root-mean-squared latency jitter of less than 100 μs (at a stimulation current of approximately 10% above threshold) were counted as identified RA-projecting neurons. Of the 17 antidromically identified neurons in our dataset, ten were further validated with collision tests \[[@pbio-0030153-b08]\]. An additional ten putative projection neurons did not respond to RA stimulation with a short-latency spike, but exhibited spike patterns and correlations similar to the identified projection neurons. For the cells in our dataset, we recorded signals for many song motifs (range, 5--133 motifs; mean, 56).
### Data analysis {#s4a4}
To assess the effects of drug injections on acoustic variability and average acoustic structure, analysis was done on reliably identifiable song syllables (range, 2--5 per bird; see [Figure 2](#pbio-0030153-g002){ref-type="fig"}A for an example). Each data point was derived from 45 pairwise comparisons made across ten consecutive renditions of a given syllable, recorded immediately before and after injection. Acoustic variability was quantified using the Sound Analysis Pro 1.04 software \[[@pbio-0030153-b36]\], and pairwise comparisons of the acoustic features of identified syllables were made using the local similarity measure ("accuracy"). This measure is based on pitch, frequency modulation, amplitude modulation, Wiener entropy, and goodness of pitch, and is calculated in 9-ms intervals and averaged over the duration of the syllable; syllables were aligned in time so as to maximize the similarity, allowing for 5% time warping. For the variability measurements, the resulting similarity score (S, ranging from zero to 100) was converted, through a linear remapping, to a variability score *(V)* by the following formula:
〈*S* ~min~〉 is the average similarity score of randomly chosen pairs of syllables from unrelated birds, which in our finch colony was measured to be 50 ± 12 (mean ± standard deviation, *n* = 200 pairwise comparisons; comparisons were made across syllables of birds from different fathers). The similarity of identical syllables, *S* ~max~, is 100 by definition of the similarity measure. Thus, a variability score of one means that syllables are as different as two unrelated syllables would be on average, while variability score of zero means that the syllables are identical. Error bars for *V* in the figures all denote standard error of the mean. 〈*V*〉 denotes the average variability score across birds and syllables for a given condition.
The variability of syllable ordering in a song was quantified using the stereotypy score of Scharff and Nottebohm \[[@pbio-0030153-b13]\], excluding the variability in the number of introductory notes and in the end syllable of a song bout. The score is a combination of "sequence linearity," which addresses the way in which notes are ordered, and "sequence consistency," a measure of the frequency with which the main motif sequence appears. Complete stereotypy yields a score of one, while a completely random sequencing will have a score close to zero. Stereotypy scores were calculated over ten consecutive song bouts, before and after LMAN injections.
For the analysis of the neural recordings in LMAN, we determined the sequence of song syllables most frequently produced by each bird. Motifs that matched this sequence were identified and time-aligned using the onset of one of the syllables. The alignment syllable was chosen for a sharp onset in acoustic power. The relative jitter in the timing of other syllables in the motif was found to be less than 9 ms (root mean squared). Spike times were extracted, and the instantaneous firing rate during each motif rendition was estimated by smoothing the spike train with a Gaussian of half-width 20 ms (to the 1/e points). Correlations were calculated between the firing rate functions for all pairs of smoothed spike trains. Correlations were also calculated for all pairs of spike trains after a random time shift. The shift was circular, such that spikes wrapped around to the beginning of the motif; time shifts were chosen randomly from a uniform distribution with the width of the motif. For each cell the correlation distribution of the time-shifted firing rates was calculated with 100 different ensembles of random shifts. This random shift ensured zero mean correlation while preserving spike statistics. Thus, the distribution of time-shifted correlations provides a zero-correlation baseline with which to compare our results.
Supporting Information {#s5}
======================
Figure S1
::: {.caption}
###### Histology Confirming the Injection Sites for the LMAN Inactivation Experiments in [Figures 1](#pbio-0030153-g001){ref-type="fig"} and [2](#pbio-0030153-g002){ref-type="fig"}
\(A) A parasaggital Nissl-stained section of a zebra finch brain showing the location of LMAN.
\(B) Inverted darkfield image showing LMAN in one of the juveniles injected (red markers in \[D\] and \[E\]).
\(C) Combined darkfield and fluorescence image showing the spread of the dye that was co-injected with the drug.
(D and E) Estimated injection sites relative to the boundaries of LMAN for all birds in [Figures 1](#pbio-0030153-g001){ref-type="fig"} and [2](#pbio-0030153-g002){ref-type="fig"} in the saggital (D) and coronal (E) planes, respectively (individual birds are color coded).
\(F) Estimated maximum diameter of LMAN in the saggital plane.
\(G) Estimated lateral extent of LMAN in the coronal plane.
The estimates in (F) and (G) are based on the contrast borders seen in the darkfield images (see \[B\]). Note that fibers from LMAN to RA leave the posterior edge of LMAN.
(369 KB PDF).
:::
::: {.caption}
######
Click here for additional data file.
:::
Figure S2
::: {.caption}
###### Dose- and Distance-Dependent Effects of TTX Injections in and around LMAN
\(A) Decrease in acoustic variability (ΔV) approximately 1 h after injection, as a function of location and concentration of TTX injections. Red bars indicate dose response for TTX injections in LMAN (*n* = 2 birds; 8 syllables; injection sites for the two birds correspond to the blue and grey markers in [Figure S1](#sg001){ref-type="supplementary-material"}). Blue bars indicate 30-nl saline injections in LMAN (*n* = 2 birds; 7 syllables). Green bars indicate 30-nl (50 μM) TTX injections 1.25 mm medial (MMAN, *n* = 2 birds; 6 syllables) and dorsal ("above," *n* = 2; 8 syllables) from the center of LMAN.
(B and C) Summary of experiments done to verify the physiological spread of TTX. Experiments were done in anesthetized birds (2% isoflurane). A bipolar stimulating electrode was placed in RA, and a recording electrode in LMAN, producing antidromically evoked activity in LMAN (stimulus pulses, 175 μA, 0.2 ms, 0.5 Hz ). TTX (30 nl, 50 μM) was injected at different distances away from the recording electrode. (B) Examples of recorded signals for TTX injections 400 μm (top) and 1,250 μm (bottom) away from the recording electrode (averaged over 30 stimulus pulses). The baseline stimulus artifact recorded 1 mm above LMAN is shown in the green boxes (left). Signal recorded in LMAN immediately before injection is shown in the black boxes (middle). Signal recorded 1 h after injection is shown in the red boxes (right). (C) Summary of evoked activity 1 h after TTX injections made at different distances away from the recording site. Evoked activity was measured as the root-mean-squared deviation of the signal from the baseline in the interval 1.5--4.5 ms after the stimulation pulse (six birds, two at 400 μm, two at 600 μm, and one each at 800 μm and 1,250 μm).
(1.1 MB PDF).
:::
::: {.caption}
######
Click here for additional data file.
:::
Figure S3
::: {.caption}
###### Example of a Juvenile Zebra Finch Song (54 dph) Showing a Loss of Sequence and Acoustic Variability following LMAN Inactivation by TTX Injection
The song snippets shown are from three consecutive song bouts, immediately before and 1 h after TTX injection. Tutor song is shown for comparison.
(1.8 MB PDF).
:::
::: {.caption}
######
Click here for additional data file.
:::
Audio S1
::: {.caption}
###### Example of a Song from the Bird in [Figure 1](#pbio-0030153-g001){ref-type="fig"} prior to TTX Inactivation of LMAN (Bout 1)
(545 KB WAV).
:::
::: {.caption}
######
Click here for additional data file.
:::
Audio S2
::: {.caption}
###### Example of a Song from the Bird in [Figure 1](#pbio-0030153-g001){ref-type="fig"} prior to TTX Inactivation of LMAN (Bout 2)
(455 KB WAV).
:::
::: {.caption}
######
Click here for additional data file.
:::
Audio S3
::: {.caption}
###### Example of a Song from the Bird in [Figure 1](#pbio-0030153-g001){ref-type="fig"} during TTX Inactivation of LMAN (Bout 1)
(430 KB WAV).
:::
::: {.caption}
######
Click here for additional data file.
:::
Audio S4
::: {.caption}
###### Example of a Song from the Bird in [Figure 1](#pbio-0030153-g001){ref-type="fig"} during TTX Inactivation of LMAN (Bout 2)
(360 KB WAV).
:::
::: {.caption}
######
Click here for additional data file.
:::
We thank Edward Soucy, Stephen Baccus, Isabella Nebel, Carlos Lois, and members of the Fee lab for comments on the manuscript. We also acknowledge Thomas Ramée for assistance with histology and animal care.
**Competing interests.** The authors have declared that no competing interests exist.
**Author contributions.** BPÖ, ASA, and MSF conceived and designed the experiments, performed the experiments, analyzed the data, contributed reagents/materials/analysis tools, and wrote the paper.
Citation: Ölveczky BP, Andalman AS, Fee MS (2005) Vocal experimentation in the juvenile songbird requires a basal ganglia circuit. PLoS Biol 3(5): e153.
AFP
: anterior forebrain pathway
AMPA
: α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid
AP5
: 2-amino-5-phosphonovalerate
dph
: days post hatch
LMAN
: lateral magnocellular nucleus of the nidopallium
MMAN
: medial magnocellular nucleus of the nidopallium
NMDA
: [N]{.smallcaps}-methyl-[D]{.smallcaps}-aspartate
RA
: robust nucleus of the arcopallium
TTX
: tetrodotoxin
|
PubMed Central
|
2024-06-05T03:55:55.005772
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069649/",
"journal": "PLoS Biol. 2005 May 29; 3(5):e153",
"authors": [
{
"first": "Bence P",
"last": "Ölveczky"
},
{
"first": "Aaron S",
"last": "Andalman"
},
{
"first": "Michale S",
"last": "Fee"
}
]
}
|
PMC1069650
|
From the "ecstatic sound" of Thomas Hardy\'s thrush to the "full-throated ease" of Keats\'s nightingale, the dulcet tones of songbirds have long inspired poetic explorations of the human spirit. Scientists have more recently found inspiration in songbirds, but it is their behavior and not their song that tickles the scientific imagination. Just as the vocal explorations of toddlers reflect the (no doubt) consequential conversations of their elders, the highly variable chirps and warbles of juvenile songbirds echo the precise melodies of the adult songbird. Through trial and error and random forays into harmolodic dissonance, the young bird patterns his performance after a tutor song (usually performed by dad) until he produces a workable facsimile. It is this behavior--- known as reinforcement learning---that makes songbirds an ideal model for studying the interplay between experience, brain activity, and learning.
Michale Fee\'s lab studies the neural basis of song learning in the zebra finch, the organism of choice for birdsong researchers. In a new study, Bence Ölveczky, Aaron Andalman, and Fee study just how young songbirds generate the vocal explorations that help the apprentice master its song.
Two major neural pathways control zebra finch song. The motor pathway controls vocal outputs through the RA (for robust nucleus of the arcopallium) neuron cluster, which indirectly stimulates vocal and respiratory muscles. When adult birds sing, RA neurons show a signature sequence of bursts during each syllable. Another pathway, called the anterior forebrain pathway (AFP), appears to be critical for song learning. AFP shares characteristics with the mammalian basal ganglia, which regulates movement and motor learning in mammals.[](#pbio-0030162-g001){ref-type="fig"}
::: {#pbio-0030162-g001 .fig}
::: {.caption}
###### The vocal explorations of young zebra finches shed light on the neural basis of learning motor tasks (Photo: Daniel D. Baleckaitis)
:::
:::
To explore the nature of the AFP\'s contributions to song learning, Fee and colleagues recorded brain activity from young zebra finches (54--79 days old) learning to sing. Then they injected young birds with drugs that temporarily blocked activity in a brain region that is part of the AFP called LMAN (lateral magnocellular nucleus of the nidopallium). Zebra finch songs typically contain three to seven syllables---the basic acoustic units of zebra finch songs---that follow a specific sequence. Thirty to 90 minutes after LMAN inactivation, the birds sang with less syllabic variation. This effect was especially dramatic in the youngest birds, which normally exhibit the greatest acoustic variation. LMAN inactivation, the authors note, "eliminated 75% of the difference in mean variability between juvenile song and adult directed song \[wooing a mate, for example\]---the most stereotyped form of song." LMAN inactivation also reduced the birds\' variation in syllable sequence, which again hewed closer to the orthodoxy of adult song than to the exuberance of youthful experimentation.
The authors go on to show that changes in the firing patterns of LMAN neurons projecting into the motor pathway accompany changes in song. That LMAN inactivation reduces song variability quickly and reversibly, the authors argue, indicates that LMAN supports experimental behavior and controls song variability by providing rapid inputs to the motor pathway. This model requires that LMAN neurons show high variability across different song motifs---which is what Fee and colleagues found. As the bird sings, some as yet unknown brain areas must also evaluate the song against a template, modulating the actions of the motor pathway as a conductor might correct a performer\'s mistakes in note and pitch until she masters the tune.
It\'s thought that birdsong serves multiple purposes---staking a territorial claim, for example, and attracting a mate---though precisely how the song relates to fitness is still an open question. Whether inducing the type of exploratory motor behavior that\'s so critical to motor learning is a fundamental feature of basal ganglia circuits also remains to be determined. But it does seem clear that these circuits play a significant role in generating the variability that songbirds need in order to acquire the communication skills of their parents---a finding that may shed light on how the brain produces the fluctuations required for learning other tasks. For more on song learning, see the primer by Fernando Nottebohm (DOI: [10.1371/journal.pbio.0030164](10.1371/journal.pbio.0030164), available online May 2005).
|
PubMed Central
|
2024-06-05T03:55:55.009030
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069650/",
"journal": "PLoS Biol. 2005 May 29; 3(5):e162",
"authors": []
}
|
PMC1069657
|
Background {#s1}
==========
It is an ongoing challenge to share health information with resource-poor communities that is locally relevant and owned by the communities themselves. Too often, new information brought to these settings is seen as coming from "outside" and therefore as having little local relevance. People may look with suspicion at those who bring such information. Many factors---the background, attitudes, clothing, employers, and the language of those who bring information---may have more impact on the way new ideas are received than the actual relevance of the ideas themselves. We have only to consider our own attitudes to politicians for an example of how such factors influence our receptivity to information.
When health information from outside the community goes against deeply held beliefs and attitudes about personal and sexual matters, this challenge becomes still greater. Providing opportunities for people to discuss the impact of HIV and AIDS on their communities in a relaxed and open manner is key to enabling people to engage in potentially lifesaving discussion and attitude change.
But as positive a step as open discussion is, unless poor people can access and accept the information they need, they will not be able to make informed decisions regarding their lives and future. Ignorance about the impact of HIV and AIDS and how the virus is transmitted is potentially life-threatening; we urgently need to raise awareness in ways that are accepted by local communities.
PILLARS: Partnership in Local Language Resources {#s2}
================================================
Between 1995 and 1999, Tearfund UK, a Christian community development and disaster relief charity, conducted research into the flow of information at grassroots level in Uganda and Ghana, with the support of the education division of the United Kingdom Department for International Development \[[@pmed-0020046-b1]\]. The findings highlighted the importance of small groups in sharing information, the lack of relevant printed materials for the poor, and the need for end users to be involved in the creation of relevant printed information in their own languages. These findings have now been translated into practical action through Partnership in Local Language Resources (PILLARS). PILLARS guides provide small community groups with simple printed information, written in local languages, on community development issues such as nutrition, food security, micro-credit, HIV and AIDS, and community mobilisation (see [www.tilz.info/resources](www.tilz.info/resources)). These guides are not seen by communities as information from an outside entity, but rather as locally consolidated information on relevant issues for groups to discuss in their regular meetings. Rather than acting as passive recipients of information, group members can bring useful experience, knowledge, and insights into the discussion ([Figure 1](#pmed-0020046-g001){ref-type="fig"}).
::: {#pmed-0020046-g001 .fig}
Figure 1
::: {.caption}
###### Participants Discussing a Topic on Leadership Styles during a PILLARS Workshop in Delhi
:::
:::
Each guide contains 20--24 topics, with illustrations, text, and discussion questions. Any group that meets on a regular basis can set aside time to read through and discuss one of the topics. A trained leader is not required, though it helps if someone in the group knows how to facilitate discussion. The guides build on existing knowledge and experience shared among group members.
Empowering the Community {#s3}
========================
The ultimate goal of PILLARS is to empower community groups in developing nations by building their capacity for collective learning, consensus-building, and subsequent action. Use of the guides restores their right to receive and share information in their own tongue and to participate in the development of their communities. The generation, use, and distribution of information in local languages encourages and gives confidence and value to marginalised groups. The discussion process helps groups manage their own change and engage in local decision-making processes.
Guides are now available on nine subjects relating to community development. They are designed for ease of translation into any language using a CD-ROM with design and text files. Pages or illustrations can be contextualised to meet local needs, and participatory bible studies are included at the back of the guides for faith-based groups to use.
A further aspect of PILLARS is that development workers can be equipped to translate and write new guides over the course of three workshops. During the first two workshops, participants learn skills in translation, reviewing, and field testing. In the final workshop, participants write their own guide and plan for future sustainability. Training in facilitation skills and participatory techniques equips participants to use the guides.
This process has been piloted in southern Sudan, Ethiopia, Nigeria, Burkina Faso, Brazil, and Myanmar. In Myanmar, development workers produced guides in Burmese and then replicated the training with a further 13 language groups, generating considerable energy and empowerment. A facilitator commented: "We have so many languages in our country. Through this programme, people are encouraged to value their culture and to share useful information about development with the community." In Ethiopia, participants wrote a guide on "harmful traditional practices". Training has also been conducted there with a refugee community from southern Sudan, helping them plan for repatriation.
A recent evaluation, led by Dr Clinton Robinson of PILLARS in Myanmar, Brazil, and Ethiopia, revealed a dearth of written information in the languages of minority groups. Access to information and to the media was generally low. The evaluation found that improving access to simple, relevant and practical information in local languages increased people\'s self-confidence and their ability to make positive change. It commented on the benefits of the emphasis PILLARS places on collective learning rather than on individual reading.
Discussion-Based Learning on HIV and AIDS {#s4}
=========================================
A new guide, *Responding More Effectively to HIV and AIDS*, is now available ([Figure 2](#pmed-0020046-g002){ref-type="fig"}) \[[@pmed-0020046-b2]\]. With funding from Development Corporation Ireland, this guide is being translated into a further nine languages: French, Spanish, Portuguese, Hindi, KiSwahili, Amharic, Khmer, Kinyarwandan, and Chinese.
::: {#pmed-0020046-g002 .fig}
Figure 2
::: {.caption}
###### A Recent PILLARS Guide
:::
:::
The guide first gently challenges misconceptions to ensure that people have the correct facts about HIV/AIDS and how the virus is passed on. Issues raised include traditional practices that might spread HIV, the need for HIV testing and accompanying counselling, and the needs of children who lose their parents from HIV/AIDS. The guide encourages discussion of how to talk about sexual issues with children, with partners, and within faith-based teaching. The burden that can fall upon carers who respond to the needs of families living with HIV/AIDS can be immense, and several topics encourage people to discuss this. The guide also addresses relevant and challenging questions, including who provides the caring, who else could help, what support systems are available, and how the local community can increase its support. Recent advances in antiretroviral therapy and the latest advice regarding breastfeeding by mothers with HIV are also covered.
At a recent workshop in Nairobi, staff from Sudan gained facilitation skills and learned techniques to share information effectively. They worked in small groups to develop simple role-plays to introduce topics from the HIV and AIDS guides. Though few of the participants had used this method before, they produced some amazingly powerful role-plays that provided a very effective introduction to the group discussion and learning that followed.
The Local Production of Information {#s5}
===================================
A number of organisations have developed tools and training to help with the local production of information ([Box 1](#box1){ref-type="boxed-text"}). PILLARS differs from these approaches, both in its focus on providing information for discussion-based learning targeted at grassroots community groups, and in providing technical support and design files to simplify the translation process.
Box 1. Tools for the Local Production of Information {#sb1}
----------------------------------------------------
Agricultural kits produced by the International Institute for Rural Reconstruction ([www.iirr.org](www.iirr.org)).Shell booklets produced by the Summer Institute of Linguistics ([www.sil.org](www.sil.org)), often used as literacy primers in local languages.The REFLECT approach ([www.reflect-action.org](www.reflect-action.org)), a method of increasing literacy using participatory techniques. Each literacy circle produces its own learning materials, analysing its own local community and its immediate circumstances.The STAR (Stepping Stones and Reflect) Initiative ([www.healthcomms.org/pdf/STARsummary.pdf](www.healthcomms.org/pdf/STARsummary.pdf)), which provides draft guidelines that support communities or organisations to analyse and tackle issues that affect them, from agriculture to war, in the context of HIV and AIDS.The Quest manual, from Healthlink Worldwide ([www.healthlink.org.uk/consult/quest.html](www.healthlink.org.uk/consult/quest.html)), a tool to support organisations to develop their capacity to produce effective communication and information resources.
PILLARS guides have now been translated into more than 30 different languages and are being used with basic literacy programmes and training workshops, and as discussion-based materials for women\'s, farmers\', and credit groups. They offer a simple, yet potentially very effective method of sharing health messages. The use of print means that such messages can be used over the long term and widely distributed.
A free copy of the HIV and AIDS guide for groups in resource-poor nations is available from E-mail: <roots@tearfund.org>.
**Citation:** Carter I (2005) Creating locally relevant health information. PLoS Med 2(3): e46.
PILLARS
: Partnership in Local Language Resources
[^1]: Isabel Carter is the coordinating editor for the Programme Development Team, Tearfund, United Kingdom. E-mail: <isabel.carter@tearfund.org>
[^2]: **Competing Interests:**The author declares that she has no competing interests.
|
PubMed Central
|
2024-06-05T03:55:55.009894
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069657/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e46",
"authors": [
{
"first": "Isabel",
"last": "Carter"
}
]
}
|
PMC1069658
|
Preparing for Death {#s1}
===================
How does one prepare for death? Those who have created a public persona must add to any spiritual ponderings about eternity the mundane chore of organizing their literary archives to protect any of life\'s secrets that seem worth the effort. That task involves choosing what diaries, letters, drafts, and laundry lists to donate to a university or to leave in a closet for legions of biographical ragpickers to quote, misquote, or variously interpret in as yet unimaginable contexts---or to burn.
Many well-known figures contemplating their posthumous selves have been foiled in exercising control over their literary remains. Purposefully confounding future biographers, Sigmund Freud burned his early papers and admonished his wife Martha to destroy their love letters. Instead, she bequeathed us this charming insight into the youthful exuberance of the patriarch of psychoanalysis, written in 1884: "Woe to you, my Princess, when I come. I will kiss you quite red and feed you till you are plump. And if you are forward, you shall see who is stronger, a gentle little girl who doesn\'t eat enough or a big wild man who has cocaine in his body" \[[@pmed-0020060-b1]\].
Anaïs Nin, whose voluminous diaries recorded her daily life in exquisite, compulsively recorded detail, had better luck in choreographing her literary afterlife. While alive, she published volumes of carefully edited literary diaries. When someone at a seminar remarked to her that her life seemed more, well, racy than those diaries revealed, she smiled mysteriously and said that after the death of all concerned, "unexpurgated" editions would be published. Several decades later, companion volumes to the literary diaries revealed passionate incest with her father, Joachim Nin, an affair with her analyst, Otto Rank, and successfully bigamous marriages in New York and California.
When André Gide revealed that Oscar Wilde had had sexual relations with a young Arab boy in Egypt, Wilde\'s friend Robert Sherard lamented: "Heavens! The task of shooing hyenas away from the graves of the illustrious dead." Sherard meant Wilde\'s literary grave---but what about actual graves? What about history\'s corpus delicti?
The Line between Scientist and Grave Robber {#s2}
===========================================
How many giants and tyrants unlucky enough to have left body parts or ashes behind when they shuffled off the mortal coil could have imagined what scientists and medical practitioners of the future would do with their physical remains? Here, the line between the scientist and the grave robber blurs, as corpses are exhumed and cremation urns raided to provide organic remnants for any number of curious purposes.
Ethical debates about the appropriate care and maintenance of biological relics often begin at the autopsy table. Having removed Albert Einstein\'s brain, pathologist Thomas Harvey chopped it into 240 pieces and stored it in a cookie jar in his basement, often shipping slabs (mailed in mayonnaise jars) to brain researchers eager to count glia and neurons. Forty years later, Harvey lugged what remained of the brain cross-country to deliver it to Evelyn Einstein, a woman rumored to be the physicist\'s daughter from an affair with a New York dancer. Dr. Charles Boyd had tried to prove this paternity with his brain-chunk, but Einstein\'s DNA proved "too denatured to decipher."
Harvey\'s volunteer driver, Michael Paterniti, described getting his hands in the cookie jar: "I actually feel as if I might puke. The pieces are sealed in celloidin---the pinkish, liver-colored blobs of brain rimmed by gold wax. I pick some out of the plastic container and hand a few to Evelyn. They feel squishy, weigh about the same as very light beach stones. We hold them up like jewelers, marveling at how they seem less like a brain than---what?---some kind of snack food, some kind of energy chunk for genius triathletes" \[[@pmed-0020060-b2]\].
Pilferers cannot resist snipping body parts. While Einstein was being autopsied, his ophthalmologist, Dr. Henry Abrams, dropped by and filched Einstein\'s brown eyes as a keepsake, storing them in a jar in a Philadelphia bank vault. There were rumors that singer Michael Jackson, a collector of body parts, offered Abrams several million dollars for the eyes.
> Does confidentiality extend beyond the grave?
Beethoven\'s ears were hacked out and soon went missing. René Descartes\'s middle finger was stolen. (His head was also separated from his body for shipping---a philosopher\'s in-joke, since Descartes introduced the mind/ body split into Western philosophy.) Napoleon\'s reputed penis went on a picaresque odyssey of its own, being displayed at the Museum of French Art in New York, auctioned, and finally ending up in the possession of a urologist---or so the story goes. Josef Haydn\'s head was stolen by phrenologists at his burial.
In 2004, Dr. Anunciada Colon presided over the opening of a golden trunk from the 16th century, containing ashes and bone fragments presumed to belong to her ancestor Christopher Columbus, an event chronicled by a television crew. Officials at the Seville Cathedral allowed researchers at the University of Granada to borrow the bones for a DNA study. Being unsuccessful at extracting DNA from pulverized fragments, Professor José A. Lorente loaded the bones in a shoulder bag and flew them to Dallas, Texas, where more sophisticated DNA tests (developed for the victims of the terrorist attack of 9/11) provided a disappointingly short and impure sequence of mitochondrial DNA. Remaining ashes and shards were inelegantly deposited on a metal storage shelf in a lab, in a Styrofoam picnic basket labeled "Colon" in black marker, awaiting better tests \[[@pmed-0020060-b3]\].
Vladimir Ilyich Lenin remains the most visible deceased person. His body, or what remains of it since his brain and other organs were removed, has been viewed by the millions who have passed by his open casket in a mausoleum on Moscow\'s Red Square. A waterproof suit under his uniform holds in the embalming fluid. His hands and head are bathed frequently. His microtomed (31,000 sections) and dyed brain resides down the street from his body at the Moscow Brain Institute, joining the brains of his countrymen Stalin and Tchaikovsky. Many Russians who find Lenin\'s public resting place a macabre embarrassment think his soul will only rest (and theirs with it) once he goes underground. But who can decree his burial?
When I was four, my mother found me exhuming a goldfish we had ceremoniously buried in the garden in a little fish coffin a few days before. How different, I wonder now, was my childish curiosity and wonderment at the mysterious process happening to my no-longer-swimming fish below the earth from that of grown-up exhumers? Consider Gira Fornaciari, who unearthed 49 members of the Medici family to confirm various causes of death, or the committee that had Beethoven and Schubert dug up to transfer them to more secure zinc coffins (borrowing both heads for a bit more measuring, and swiping Schubert\'s luxuriant, larvae-laden hair while they were at it). Archaeologists have braved curses and biohazards to retrieve mummies from pyramids. Doctors from Japan, however, were not allowed to take DNA from King Tut\'s mummy to sort out his genealogy; the Egyptian government\'s supreme council of antiquities, after first agreeing, reversed the decision. A non-invasive x-ray of the mummy suggests a murder plot: King Tut may have been done in by a blow to the back of the skull.
Guidelines for Bioethical Research {#s3}
==================================
When a committee was convened to decide whether specimens of Lincoln\'s blood and bones should be tested for DNA to discover whether he suffered from Marfan syndrome, ethicists voted yes but scientists vetoed the plan, claiming that the precious material should not be destroyed in case future tests would prove more effective \[[@pmed-0020060-b4],[@pmed-0020060-b5]\]. But what if they were even asking the wrong question? Lincoln once told his biographer and friend William Herndon that he had been infected with syphilis by a prostitute in Beardstown around 1835 \[[@pmed-0020060-b6]\]. What if a future test could prove that Lincoln had spoken the truth? Imagine, if you will, a press release from the Armed Forces Institute of Pathology revealing that hot potato about the most beloved of American presidents.
The Lincoln testing question spurred bioethicist Lori Andrews and her colleagues at the Chicago Historical Society to join with the Illinois Institute of Technology to review existing ethical issues of biohistorical research. Their conclusion, after studying professional codes from 23 other organizations: none contained guidelines for conducting biohistorical research and analysis \[[@pmed-0020060-b7]\]. They recommend genetic testing for "historically significant" questions. But who is to define that loaded phrase?
The newly dead are warm, soft, and somehow still human; by contrast, aged corpses and skeletons rising from the cold ground are the stuff of horror films, vampires and ghouls. While fascinating, they also unnerve. Medical examiners in fiction (Kay Scarpetta) and television (Dr. Quincy, Jordan Cavanaugh) capture wide audiences with their gruesome and graphic dissection of putrefied, maggot-ridden corpses, all in the service of solving some medical mystery.
Respect for the Dead {#s4}
====================
Does confidentiality extend beyond the grave? Should doctors publish articles in medical journals about diagnoses that were confidential when the patient was alive? Physicians have often raced to put pen to paper and reveal the signs and symptoms of their more illustrious deceased patients. According to Anne Sexton\'s biographer Diane Wood Middlebrook, who used tapes of hundreds of hours of therapy sessions given to her by Sexton\'s therapist Dr. Martin Orne, the dead have no rights \[[@pmed-0020060-b8]\]. Although Dr. Orne insisted that Sexton had given him permission to do what he thought appropriate with the tapes, his colleagues howled that he had made a travesty of doctor-patient confidentiality, Sexton\'s wishes be damned.
The long-dead are latecomers to the game of lobbying for rights. Who owns their bones? Who is to choose the right test, the right time, the appropriate question to ask? Who gets to decide whether they should be sliced, diced, dyed, pulverized, displayed, x-rayed, photographed, and subjected to the esoteric tests developed for forensic laboratories to reveal secrets they carefully took to their graves or urns? An interdisciplinary committee? The law? The government? Should such decisions be made by bioethicists, scientists, medical examiners, lawyers, archaeologists, descendants of the deceased? Where does simple respect for the dead play into this issue?
The answers change over time and from place to place. The quagmire of ethical, legal, moral, and even aesthetic questions that surround the use (and misuse) of leftover body parts can only become more complex and contentious, not less.
A word of warning, then, to the famous not-yet-deceased: consider the disposition of your physical remains as carefully as you consider the packaging of your archive.
Swear your doctor to posthumous secrecy.
Be cremated.
And have your ashes scattered to the wind.
::: {#pmed-0020060-g001 .fig}
::: {.caption}
###### Is it ethical to remove body parts to make a tissue diagnosis?
(Illustration: Margaret Shear, Public Library of Science)
:::
:::
::: {#pmed-0020060-g002 .fig}
::: {.caption}
###### Victor McKusick of the Johns Hopkins School of Medicine chaired a committee to decide whether specimens of Lincoln\'s blood and bones should be tested for Marfan syndrome
(Photo: Alexander Gardner, Library of Congress)
:::
:::
**Citation:** Hayden D (2005) Alas, poor Yorick: Digging up the dead to make medical diagnoses. PLoS Med 2(3): e60.
[^1]: Deborah Hayden is the author of *POX: Genius, Madness, and the Mysteries of Syphilis* (Basic Books 2004), a biographical study of the effects of syphilis on cultural icons. She has recently published articles in the *New Statesman* and the *The Wildean: A Journal of Oscar Wilde Studies*, and has been interviewed for "High Hitler," a History Channel special pertaining to Adolf Hitler\'s syphilis diagnosis. E-mail: <debhayden@sbcglobal.net>
[^2]: **Competing Interests:** The author declares that she has no competing interests.
|
PubMed Central
|
2024-06-05T03:55:55.011135
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069658/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e60",
"authors": [
{
"first": "Deborah",
"last": "Hayden"
}
]
}
|
PMC1069659
|
Epidemics of overt toxicity following widespread environmental contamination from commercial toxins heralded the discovery of children\'s enhanced vulnerability to lead, methyl mercury, polychlorinated biphenyls (PCBs), and tobacco \[[@pmed-0020061-b1],[@pmed-0020061-b2],[@pmed-0020061-b3],[@pmed-0020061-b4],[@pmed-0020061-b5]\] ([Box 1](#box1){ref-type="boxed-text"}). Over the past three decades, researchers have found that remarkably low-level exposures to these toxins are linked with less overt symptoms of toxicity---intellectual impairments, behavioral problems, spontaneous abortions, or preterm births \[[@pmed-0020061-b6],[@pmed-0020061-b7],[@pmed-0020061-b8],[@pmed-0020061-b9],[@pmed-0020061-b10],[@pmed-0020061-b11],[@pmed-0020061-b12],[@pmed-0020061-b13],[@pmed-0020061-b14], [@pmed-0020061-b15],[@pmed-0020061-b16],[@pmed-0020061-b17],[@pmed-0020061-b18],[@pmed-0020061-b19],[@pmed-0020061-b20],[@pmed-0020061-b21],[@pmed-0020061-b22],[@pmed-0020061-b23],[@pmed-0020061-b24],[@pmed-0020061-b25],[@pmed-0020061-b26],[@pmed-0020061-b27], [@pmed-0020061-b28],[@pmed-0020061-b29],[@pmed-0020061-b30],[@pmed-0020061-b31],[@pmed-0020061-b32],[@pmed-0020061-b33],[@pmed-0020061-b34],[@pmed-0020061-b35],[@pmed-0020061-b36],[@pmed-0020061-b37],[@pmed-0020061-b38],[@pmed-0020061-b39],[@pmed-0020061-b40]\]. Moreover, there is emerging evidence that decrements in intellectual abilities and low birth weight linked with lead or tobacco are, for a given increment of exposure, greater at lower levels than those found at higher levels \[[@pmed-0020061-b10],[@pmed-0020061-b41],[@pmed-0020061-b42],[@pmed-0020061-b43]\].
Box 1. Poisoning following Widespread Environmental Contamination from Commercial Toxins {#sb1}
========================================================================================
**Lead:** One hundred years ago, an epidemic of lead poisoning was described among children who ingested leaded house paint \[[@pmed-0020061-b2],[@pmed-0020061-b3]\]. The children developed anemia, encephalopathy, paralysis, and blindness.**Methyl Mercury:** In the 1950s, in the Japanese fishing village Minamata Bay, which was contaminated with methyl mercury, children developed cerebral palsy, limb defects, and mental retardation \[[@pmed-0020061-b4]\].**PCBs:** In Taiwan and Japan during the 1960s and 1970s, the ingestion of PCB-contaminated rice bran oil by pregnant women led to fetal wasting and cola-colored, dull, apathetic children \[[@pmed-0020061-b5]\].**Tobacco:** During the past century, widespread tobacco use has led to an epidemic of undersized, premature babies and children with repeated bouts of wheezing or asthma \[[@pmed-0020061-b6],[@pmed-0020061-b7],[@pmed-0020061-b8],[@pmed-0020061-b9],[@pmed-0020061-b10]\].
The consequences of exposure to many other chemicals or mixtures of chemicals, such as insecticides---chemicals oftentimes specifically designed to be toxic---are largely unknown \[[@pmed-0020061-b33],[@pmed-0020061-b34],[@pmed-0020061-b35],[@pmed-0020061-b44]\]. Many of these chemicals or their metabolites are routinely found in the blood and body fluids of pregnant women and children \[[@pmed-0020061-b45]\].
Children\'s Vulnerability to Environmental Toxins {#s2}
=================================================
The developing fetus and young child is particularly vulnerable to certain environmental toxins \[[@pmed-0020061-b46],[@pmed-0020061-b47],[@pmed-0020061-b48],[@pmed-0020061-b49],[@pmed-0020061-b50]\]. Critical neurodevelopmental processes occur in the human central nervous system during fetal development and in the first three years of life. These processes include cortical functional differentiation, synaptogenesis, myelination, and programmed apoptosis \[[@pmed-0020061-b46]\].
Children\'s exposure to environmental toxins is insidious. Environmental toxins covertly enter a child\'s body transplacentally during fetal development or by direct ingestion of house dust, soil, and breastmilk and other dietary sources during early childhood \[[@pmed-0020061-b51],[@pmed-0020061-b52],[@pmed-0020061-b53],[@pmed-0020061-b54],[@pmed-0020061-b55],[@pmed-0020061-b56]\]. Our ability to directly measure the actual levels of environmental chemicals in human tissues and body fluids using biologic markers (biomarkers) enables scientists to more effectively link exposures to environmental toxins with disability or disease \[[@pmed-0020061-b57]\].
Despite our increased knowledge of the toxicity of environmental chemicals, testing for developmental neurotoxicity (DNT) and reproductive toxicity is rarely done. DNT testing uses animal experiments to provide information on the potential functional and morphologic toxicity to the fetal nervous system that results from the mother\'s exposure to toxins during pregnancy and lactation. Paradoxically, DNT testing of a chemical is seldom requested, and then typically requested only if there is pre-existing evidence that it is neurotoxic.
The Prevalence of Diseases and Disabilities Linked to Environmental Toxins {#s3}
==========================================================================
Based on parental reports, one in six United States children has one or more developmental disabilities, from a subtle learning disability to overt behavioral or emotional disorders \[[@pmed-0020061-b58]\]. Exposures to environmental toxins have been linked with higher rates of mental retardation, intellectual impairment, and behavioral problems, such as conduct disorder and attention deficit hyperactivity disorder \[[@pmed-0020061-b16],[@pmed-0020061-b17],[@pmed-0020061-b18], [@pmed-0020061-b19],[@pmed-0020061-b20],[@pmed-0020061-b21],[@pmed-0020061-b22],[@pmed-0020061-b23],[@pmed-0020061-b24],[@pmed-0020061-b25],[@pmed-0020061-b26],[@pmed-0020061-b27],[@pmed-0020061-b30],[@pmed-0020061-b31],[@pmed-0020061-b36],[@pmed-0020061-b37], [@pmed-0020061-b38],[@pmed-0020061-b39],[@pmed-0020061-b40],[@pmed-0020061-b41],[@pmed-0020061-b42],[@pmed-0020061-b43],[@pmed-0020061-b59],[@pmed-0020061-b60],[@pmed-0020061-b61]\].
One in ten US babies is born preterm and about 5% have low birth weight \[[@pmed-0020061-b62],[@pmed-0020061-b63]\]. Preterm birth, defined as birth at less than 37 weeks of gestation, is a major determinant of infant mortality and morbidity throughout childhood \[[@pmed-0020061-b62],[@pmed-0020061-b63],[@pmed-0020061-b64]\]. Exposures to environmental toxins such as lead, tobacco smoke, and DDT have been linked with an increased risk for spontaneous abortion, low birth weight, or preterm birth \[[@pmed-0020061-b6],[@pmed-0020061-b9],[@pmed-0020061-b10],[@pmed-0020061-b13],[@pmed-0020061-b14],[@pmed-0020061-b15],[@pmed-0020061-b28],[@pmed-0020061-b32],[@pmed-0020061-b65],[@pmed-0020061-b66]\]. The rate of occurrence for many of these diseases or disabilities has been rising, as has treatment for attention deficit hyperactivity disorder and depression in children \[[@pmed-0020061-b62],[@pmed-0020061-b63],[@pmed-0020061-b67],[@pmed-0020061-b68],[@pmed-0020061-b69],[@pmed-0020061-b70]\].
Multiple risk factors, including both genetic and environmental influences, interact in complex and often unknown ways to cause disease and disability in children. But efforts can be undertaken to prevent or reduce environmental exposures linked to disease without full elucidation of the underlying mechanism \[[@pmed-0020061-b71]\]. Thus, conducting some sort of test to identify pesticides and industrial chemicals that could cause reproductive or neurobehavioral toxicity before the chemical reaches widespread use is essential to protect pregnant women and children.
Origin and Evolution of DNT Tests {#s4}
=================================
The process for testing potential developmental neurotoxins in laboratory animals evolved out of a series of tragic epidemics. Widespread use of the drug thalidomide during the 1950s led to an epidemic of phocomelia, an absence or deformity of limbs and other congenital defects in children exposed in utero to the drug \[[@pmed-0020061-b72]\]. Subsequently, in 1965, the Food and Drug Administration (FDA) developed the Teratology Guidelines. Because thalidomide induced gross defects in rabbits but not in rats, these guidelines called for toxicity tests in two species. Moreover, these guidelines focused on gross abnormalities; they did not require testing for behavioral or DNT.
Following the outbreak of methyl mercury poisoning in Minamata Bay ([Box 1](#box1){ref-type="boxed-text"}), Japan and the United Kingdom added behavioral (DNT) guidelines to their teratology requirements in 1974 and 1975, respectively \[[@pmed-0020061-b73]\]. In 1978, the Collaborative Behavioral Teratology Study (CBTS) was conceived to standardize and evaluate methods for DNT testing in the US \[[@pmed-0020061-b74]\]. The final report was issued in 1985, and shortly thereafter, Dr. Donald Kennedy, who was then Commissioner of the FDA, supported the adoption of the CBTS recommendations. But the FDA failed to implement these recommendations after Kennedy\'s departure.
> Children\'s exposure to environmental toxins is insidious
In 1990, the US Environmental Protection Agency (EPA) identified nine developmental neurobehavioral teratogens for both humans and animals (lead, PCBs, methyl mercury, cocaine, alcohol, phenytoin, heroin, methadone, and ionizing radiation) and developed rules for DNT testing in laboratory animals \[[@pmed-0020061-b49],[@pmed-0020061-b50]\]. By 1991, the Developmental Neurotoxicity Test Guidelines (OPPTS 870.6300) had been established for use when submitting chemical data to the EPA \[[@pmed-0020061-b49]\]. In 1993, the National Research Council recommended that DNT data be included in the EPA\'s evaluations of pesticides, which include classes of chemicals specifically designed to be toxic \[[@pmed-0020061-b44]\].
The Precarious US Framework for Protecting Children {#s5}
===================================================
Despite numerous attempts to upgrade the regulatory system, such as the CBTS, the framework to protect children from environmental toxins is precarious. Under current regulations, manufacturers of commercial chemicals (excluding pesticides) are not required to supply any toxicity data before selling their products. Nor are pesticide manufacturers obligated to supply basic premarket toxicity and exposure data necessary to ensure that children will be protected from exposure and potential harm from use of those pesticides. Indeed, the vast majority of chemicals have not been tested for DNT. The most basic toxicity tests in animals are lacking for 75% of the 3,000 highest production volume chemicals---chemicals for which annual production exceeds 1 million pounds per year \[[@pmed-0020061-b49],[@pmed-0020061-b75],[@pmed-0020061-b76],[@pmed-0020061-b77]\]. The US EPA has entered into an agreement with the American Chemistry Council, the chemical manufacturer\'s trade association, to provide basic toxicity screening tests for the high-production-volume chemicals by 2005 (<http://www.epa.gov/chemrtk/volchall.htm>), but this is voluntary.
For new pesticides intended for use on food crops---one of the areas in which regulations are most stringent---regulations require only that DNT testing be evaluated for substances already known or suspected of being toxins. Further, neurotoxicity testing need be conducted only in adult animals. The EPA acknowledges that over 140 registered pesticides are neurotoxic (i.e., specifically designed to act against pests by interfering with neurotransmitters or other processes shared by mammals and insects), but the EPA has received DNT testing using validated protocols for only nine pesticides \[[@pmed-0020061-b49],[@pmed-0020061-b75],[@pmed-0020061-b76],[@pmed-0020061-b77]\].
There is no general requirement that pesticides or other chemicals be tested for potential DNT prior to their registration and use \[[@pmed-0020061-b49]\]. For pesticides---which undergo more premarket testing than other chemicals---the EPA has relied on a tiered system of toxicity testing. The assumption underlying this system is that positive findings on earlier, more basic tests of neurotoxicity in adult animals will "trigger" the EPA to request more extensive testing by manufacturers, including tests in immature animals. Unfortunately, this tiered process has failed to result in appropriate DNT testing. In 1998, an internal EPA Toxicology Working Group concluded that these triggers may not be sufficient to identify all chemicals that have the potential to produce DNT \[[@pmed-0020061-b75]\]. Moreover, this tiered system discourages industry from conducting testing in immature animals because the findings could necessitate further costly testing and hinder a chemical from reaching the market.
The European Framework: "REACH" {#s6}
===============================
In 2001, the European Commission affirmed that the European Union\'s legislative framework did not provide adequate information about the adverse effects of chemicals on human health, and that when hazards were identified the regulatory agencies were slow to assess risks and to introduce measures to reduce those risks \[[@pmed-0020061-b78]\]. Indeed, chemical manufacturers are not required to "prove" that a chemical is safe before marketing it. The European Commission proposed a new regulatory framework for chemicals, REACH (Registration, Evaluation, and Authorization of Chemicals) \[[@pmed-0020061-b78],[@pmed-0020061-b79]\] ([Figure 1](#pmed-0020061-g001){ref-type="fig"}).
::: {#pmed-0020061-g001 .fig}
Figure 1
::: {.caption}
###### Flow Chart Summarizing REACH (Registration, Evaluation, and Authorization of Chemicals)---the European Commission\'s Regulatory Framework for Chemicals
(Illustration by Sapna Khandwala, Public Library of Science, adapted from \[[@pmed-0020061-b86]\])
:::
:::
Under REACH, chemical manufacturers would have to assume a much greater burden for showing the lack of harm from use of their products. Specifically, REACH would require both European and non-European manufacturers doing business in Europe to submit more extensive toxicity data for about 30,000 chemicals on the market, including reproductive and DNT data for those chemicals produced in highest quantity. Chemicals found to be hazardous would be subject to an authorization procedure to show that they can be used safely or that there are no safer alternatives. This registration process would not guarantee that chemicals are safe, but it is a step in the right direction.
The American Chemistry Council has objections to REACH, stating that "the proposed regulation is burdensome, costly, and impractical" (<http://www.accnewsmedia.com/site/page.asp?TRACKID=&VID=1&CID=359&DID=1256>). The pharmaceutical industry used similar objections to ward off regulations before the thalidomide epidemic ushered in requirements for pharmaceutical agents to undergo extensive premarket testing in clinical trials \[[@pmed-0020061-b80]\].
Limitations of Existing Animal Tests for DNT {#s7}
============================================
The US EPA has been slower than the EU to adapt to the overwhelming evidence that low-level exposure to environmental toxins can be harmful. The EPA continues to rely heavily on data from animal (toxicity) testing conducted on only a single animal species and in adult animals. Furthermore, EPA guidelines for a general developmental toxicity screening test typically examine only crude toxicological endpoints such as death, body weight, or organ dysfunction. In contrast, the DNT includes tests of locomotor activity, acoustic startle, learning, and memory. But, as currently designed, the existing tests may miss important effects such as mood changes, impulsive behaviors, and attentional problems that in humans have been shown to result from exposures to environmental toxins \[[@pmed-0020061-b24],[@pmed-0020061-b27],[@pmed-0020061-b30],[@pmed-0020061-b37],[@pmed-0020061-b40]\]. While these effects might seem subtle, they can seriously interfere with a child\'s social and emotional well-being. It is also uncertain whether tests conducted under current EPA guidelines will detect subtle deficits in key human skills such as reading.
There are other problems with relying principally on adult animals to signal the potential for DNT in humans. The structure and development of the cerebral cortex of animals commonly used in these studies differs markedly from that of humans. A chemical\'s effects on one type of animal may differ from its effects on other animals and on humans. In the case of thalidomide, high-dose fetal exposure had adverse morphologic effects on rabbits, but not rats; functional effects have only recently been described \[[@pmed-0020061-b81]\].
Although there is some concordance of human and animal data for the adverse effects of lead, mercury, and PCBs, intake limits for these compounds established exclusively on the basis of rodent studies have not been sufficiently protective of human health compared with epidemiologic studies \[[@pmed-0020061-b47]\]. Indeed, there is compelling evidence from epidemiologic studies of widespread contaminants such as lead, tobacco, and PCBs that human studies are essential to ensure that children are not harmed by low levels of exposure \[[@pmed-0020061-b11],[@pmed-0020061-b12],[@pmed-0020061-b13],[@pmed-0020061-b14],[@pmed-0020061-b15],[@pmed-0020061-b16],[@pmed-0020061-b17],[@pmed-0020061-b18],[@pmed-0020061-b19],[@pmed-0020061-b20],[@pmed-0020061-b21],[@pmed-0020061-b22], [@pmed-0020061-b23],[@pmed-0020061-b24],[@pmed-0020061-b25],[@pmed-0020061-b26],[@pmed-0020061-b27],[@pmed-0020061-b28],[@pmed-0020061-b29],[@pmed-0020061-b30],[@pmed-0020061-b31],[@pmed-0020061-b32],[@pmed-0020061-b33],[@pmed-0020061-b34],[@pmed-0020061-b35], [@pmed-0020061-b36],[@pmed-0020061-b37],[@pmed-0020061-b38],[@pmed-0020061-b39],[@pmed-0020061-b40]\].
From a scientific standpoint, data from epidemiologic studies represent the "gold standard" for detecting subtle effects of environmental toxins on humans. But epidemiological studies are expensive to mount, difficult to execute, and take years to complete. Using observational studies to disentangle the adverse consequences of a single toxin from other environmental influences and to promulgate regulations is a difficult and painfully slow process. There is also a financial disincentive for chemical registrants to voluntarily fund such studies because a positive epidemiological study could lead to stricter regulations. More importantly, if society continues to rely on epidemiologic studies to evaluate the toxicity of chemicals only after they are marketed, many children will first be harmed.
Steps to Protect Children from Environmental Toxins {#s8}
===================================================
Children must be better protected from both new and existing chemicals that are known or possible toxins \[[@pmed-0020061-b49]\]. To protect children from existing toxins, such as lead, mercury, and tobacco, the US EPA and FDA need more authority and resources to regulate and reduce emissions and exposures. Under our current system, efforts to enhance regulations to protect children from confirmed toxins are costly and protracted. Indeed, countless communities across the globe suffer from widespread environmental contamination. If there is any lesson from our experience with environmental toxins, it is that we need to identify environmental chemicals that are toxic before they are marketed or widely disseminated.
For new commercial chemicals, toxicity testing in animals should be required before they are marketed. For all new chemicals, including pesticides, extensive premarket testing should be required in multiple animal species of both sexes and at different developmental stages. These tests should be designed to have adequate statistical power to detect subtle differences within the ranges of exposure that occur in human populations. If implemented, these testing requirements would represent a dramatic departure from existing regulations, while providing a powerful incentive for industry to develop less toxic chemicals.
Toxicity testing in animals is essential but insufficient to protect pregnant women and children. For one thing, uncertainties about the safety of a chemical for humans will persist even after toxicity testing in animals is successfully completed. One additional safeguard that deserves further debate is whether prevalent environmental chemicals to which children could be exposed should undergo more extensive testing in human trials before they are marketed. If done, these trials should examine exposure, uptake (using biomarkers), and adverse effects among children or other populations only when the product is used as intended. For example, once animal toxicity testing of a residential pesticide is complete (including DNT and reproductive toxicity testing), a pesticide could undergo further testing in the home environment. Using an experimental group and a control group, researchers would compare levels of pesticides found in settled dust, on children\'s hands, and in their blood, urine, or hair. Children would be followed, when indicated, to ensure that an excess of neurobehavioral problems or other relevant outcomes did not develop among those whose homes were assigned to receive the pesticide application.
If such trials were undertaken, families would need to be fully informed about the purpose, potential benefits, and risks of participating. The trials should be conducted by the federal government---or other independent entities that do not have any ties to the chemical industry---and funded by an industry fee or tax. Community representatives would need to be involved in the review and approval of such trials, and ethical standards would need to be established regarding, for example, the role of data safety and monitoring boards. Many families would undoubtedly find it objectionable and would choose not to participate. Indeed, some products might never undergo testing if they failed to offer meaningful benefits to families, in which case the product would either be taken off the market or never reach the market.
This type of trial sounds extreme, but it is quite rational when compared to the existing approach of disseminating a potential toxin into children\'s environments without any human data about exposure, uptake, or toxicity. Furthermore, under our existing system, families are neither informed nor given an option to decline involvement in what ultimately are experiments exposing millions of pregnant women and children to potential toxins. Thus, we need to thoughtfully deliberate about whether these types of trials can be done in an ethical fashion. We also need to have further debate about whether it is ethical to continue to disseminate chemicals of unknown toxicity into children\'s environments or to allow children to continually be exposed to prevalent toxins, like lead, despite considerable evidence that they are toxic \[[@pmed-0020061-b82]\]. Too often, it is left up to a few investigators or community leaders to discover and quantify the adverse effects of toxins, and advocate efforts to reduce children\'s exposure.
Conclusion {#s9}
==========
In contrast with the EU\'s proposed REACH program, which would require industry to conduct more tests or analyses to demonstrate that high-production chemicals will not cause harm to fetuses or children, the Bush administration has argued---in unison with the American Chemistry Council---that such regulations would harm industry \[[@pmed-0020061-b83],[@pmed-0020061-b84]\]. It is time to acknowledge that the existing requirements for toxicity testing and regulations are inadequate to safeguard pregnant women and children. Until a formal regulatory system is developed to effectively screen and identify new and existing chemicals that are toxic to pregnant women and children, we are left to await the next epidemic to warn us about an environmental disaster. Unfortunately, by then we will have once again fouled our nest \[[@pmed-0020061-b85]\].
::: {#pmed-0020061-g002 .fig}
::: {.caption}
###### The US framework to protect children from environmental toxins is precarious
(Photo: Earl Dotter, <http://www.earldotter.com>)
:::
:::
We would like to acknowledge the comments of David Wallinga.
**Citation:** Lanphear BP, Vorhees CV, Bellinger DC (2005) Protecting children from environmental toxins. PLoS Med 2(3): e61.
CBTS
: Collaborative Behavioral Teratology Study
DNT
: developmental neurotoxicity
EPA
: Environmental Protection Agency
FDA
: Food and Drug Administration
PCB
: polychlorinated biphenyl
[^1]: Bruce P. Lanphear is Director of the Environmental Health Center at Cincinnati Children\'s Hospital Medical Center, and Professor of Pediatrics and of Environmental Health in the Departments of Pediatrics and Environmental Health at the University of Cincinnati, Ohio, United States of America. Charles V. Vorhees is Professor of Pediatrics and Environmental Health at Cincinnati Children\'s Hospital Medical Center and the University of Cincinnati. David C. Bellinger is Professor in the Department of Neurology at Children\'s Hospital and Harvard Medical School, and in the Department of Environmental Health at Harvard School of Public Health, Boston, Massachusetts, United States of America.
[^2]: **Competing interests:** BPL has served as an expert witness for the state of Rhode Island and for counties in California and Wisconsin, for which Cincinnati Children\'s Hospital Medical Center was compensated. CVV served as an expert witness for the Federal Daubert hearing in the Fen-Phen multidistrict litigation, consulted on DEET neurotoxicity studies (see Fundam Appl Toxicol 21: 355--365), and advised pharmaceutical companies on methods for conducting rodent neurotoxicity evaluations. DCB declares that he has no competing interests.
|
PubMed Central
|
2024-06-05T03:55:55.012588
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069659/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e61",
"authors": [
{
"first": "Bruce P",
"last": "Lanphear"
},
{
"first": "Charles V",
"last": "Vorhees"
},
{
"first": "David C",
"last": "Bellinger"
}
]
}
|
PMC1069660
|
In Stockton, California, a city of 269,000 people nestled in California\'s largest agricultural valley, residents are reported to speak 100 different languages. Acculturation is difficult in the best of circumstances, but what happens when those people with limited or no proficiency in English have a medical problem? Many United States hospitals are required to provide some manner of interpreter services for people with limited English proficiency---but do those services also bridge the cultural divide?
Meeting the challenge of providing health care for a multicultural population is now a major movement that is affecting health care in developed countries, principally the US but also in European countries and Australia. Although the bulk of studies and commentaries on the subject began to appear in the 1990s, the literature dates back much further, to articles written in the 1960s and 1970s by medical anthropologists, sociologists, nurses, mental health professionals, and others.
Wake-Up Calls {#s2}
=============
In the US, the first major alert on this problem came in 1985, when the *Report of the Secretary\'s Task Force on Black and Minority Health* was issued \[[@pmed-0020062-b1]\]. (The "Secretary" was the head of the Department of Health and Human Services (DHHS).) The report painted a bleak picture of the quality of health care afforded to African-Americans and other racial and ethnic minorities.
A decade later, reports from the US Institute of Medicine began to appear. Three of the ten reports, which spanned a ten-year period, dealt with the need to greatly diversify the health professions work force---still a somewhat unachieved goal. The most recent, considered a new wakeup call, was the 2003 report *Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care* \[[@pmed-0020062-b2]\]. It minced few words in describing the problems faced by racial and ethnic minorities who sought health care: "The conditions in which many clinical encounters take place---characterized by high time pressure, cognitive complexity, and pressures for cost-containment---may enhance the likelihood that these processes will result in care poorly matched to minority patients\' needs. Minorities may experience a range of other barriers to accessing care, even when insured at the same level as whites, including barriers of language, geography and cultural familiarity" ([Figure 1](#pmed-0020062-g001){ref-type="fig"}).
::: {#pmed-0020062-g001 .fig}
Figure 1
::: {.caption}
###### Minority Groups Face a Disparity in the Quality of Health Care
(Illustration: Giovanni Maki, adapted from \[[@pmed-0020062-b2]\]).
The Institute of Medicine defined "disparities" in health care as racial or ethnic differences in the quality of health care that are not due to access-related factors, clinical needs, patient preferences, or appropriateness of intervention \[[@pmed-0020062-b2]\].
:::
:::
Soon afterward, another US government arm, the Agency for Healthcare Research and Quality of the DHHS, issued two other reports: the *National Healthcare Disparities Report* \[[@pmed-0020062-b3]\] and the *National Healthcare Quality Report* \[[@pmed-0020062-b4]\], with annual updates promised. The reports focused on seven clinical conditions, including cancer, diabetes, and mental health, and discussed the quality of care and differences in access to such care for special population groups, including minorities and the disabled.
All of these reports make it clear that health care professionals and health systems need to change. In recent years, in order to improve their lives economically or avoid war and/or famine, many people have migrated from less to more developed areas of the world, changing the demographics of the US and a number of other societies. Evidence that they and nonmigrant minorities experience inequities in attaining quality health care is abundant \[[@pmed-0020062-b5]\].
Studies also indicate that although genetics is involved in some health-related differences between racial and ethnic groups, such as in the incidence of certain diseases and responses to pharmaceuticals, it is probably not a major factor in explaining health disparities \[[@pmed-0020062-b6]\].
The Era of Action {#s3}
=================
A primary result of these reports on health disparities? A truly dizzying array of offices, centers, programs, and initiatives within the main DHHS as well as in some of its major branches such as the National Institutes of Health and the Centers for Disease Control and Prevention, all designed to improve health care for racial and ethnic minorities in one way or another. Some of these programs also fund grants to outside organizations, public and private, and coordinate with state offices of minority health.
And there are more activities devoted to reducing health disparities: (1) university-level institutes, offices, and programs, such as those at the UMDNJ-Robert Wood Johnson Medical School and Georgetown University, (2) private foundations, such as the California Endowment, (3) agencies and programs within the various states, such as the very active Ohio Commission on Minority Health, and (4) combinations of groups, such as DiversityRx ([www.diversityrx.org](www.diversityrx.org)), an informational organization sponsored by the National Conference of State Legislatures, Resources for Cross-Cultural Health Care, and the Henry J. Kaiser Family Foundation.
All these efforts might suggest that there are no problems left to be solved, but this is hardly the case. Providing quality health care to those who differ from a country\'s majority population in terms of language and culture (and often race) is a mammoth task that does not yield to easy or quick fixes, but rather to consistent and determined efforts at improvement.
Cultural Competence {#s4}
===================
The most common term used in this effort is "cultural competence," essentially defined as a respectful knowledge of and attitude toward people from different cultures that enables health professionals who work with people from another culture to develop and use standard policies and practices that will increase the quality and outcome of their health care.
With cultural competence as the centerpiece, social and behavioral scientists have started consulting companies to (1) train health care professionals working in private and public health care settings (hospitals, community clinics, managed health care plans) in cultural competence, and (2) propose as well as study the effects of such changes in these settings. Some hospitals and managed health care plans have developed their own programs; examples that stand out are the M.D. Anderson Hospital in Texas and Kaiser Permanente health plans.
In 2000, the M.D. Anderson Hospital established an Office of Institutional Diversity, which emphasizes the use of employees with a variety of backgrounds and experiences to examine cancer and its impact on all kinds of people. Educational forums, employee network groups, and the use of evidence-based hypotheses to design and implement pilot interventions are all part of the effort to improve care of culturally diverse patients.
::: {#pmed-0020062-g002 .fig}
:::
Kaiser Permanente\'s Institute for Culturally Competent Care selects and coordinates Kaiser Permanente\'s several Centers of Excellence, which each serve specific populations. For example, a West Los Angeles center focuses on the diagnosis, treatment, and management of conditions prevalent among African-Americans, such as sickle cell disease and prostate cancer. The National Diversity Department emphasizes a diverse workforce and has published a number of providers\' handbooks on culturally competent care for specific racial or cultural patient groups, such as Latino patients.
Not to Be Left Out {#s5}
==================
Pharmaceutical companies have also discovered multicultural medicine. Many that offer continuing medical education courses to help publicize their new drugs now also offer courses on diseases more prevalent in certain racial and ethnic groups than others (such as diabetes in the Hispanic/ Latino population). These courses include information on how to treat such groups with the company\'s drugs.
Interestingly, in 2004 a clinical trial proved the effectiveness of the first drug specifically designed for the treatment of congestive heart failure in African-Americans \[[@pmed-0020062-b7]\]. The drug, a combination of fixed doses of isosorbide and hydralazine, may now be nearing the market. Despite the fact that the Association of Black Cardiologists was a cosponsor of the trial, the trial drew criticism on the basis that it allowed race to interfere with treatment decisions \[[@pmed-0020062-b6]\].
A Global Issue {#s6}
==============
The increased diversity of European populations, with the expected stress on entrenched health care systems and on the migrants themselves, has led to Migrant-Friendly Hospitals (<http://www.mfh-eu.net>), a "European initiative to promote health and health literacy of migrants and ethnic minorities" begun in October 2002.
With funding from the European Commission and the Austrian Federal Ministry for Education, Science and Culture, a network of 12 pilot hospitals from European Union member states has been implementing and evaluating the effectiveness of three health care models for migrants and minorities. The models are: the improvement of interpreting in clinical communication, the creation and distribution of migrant-friendly information and training in mother and child care, and staff training in cultural competence. Results of the pilot experiences were reported at a final conference in December 2004 and will form the basis of European recommendations on migrant-friendliness as a quality criterion for hospital development and on the role of hospitals in promoting health and health literacy for migrants and ethnic minorities.
One of the 12 pilot hospitals mentioned above is the Bradford Hospitals NHS Trust, long active among a number of other hospitals and health projects in the UK that strive to improve services for racial and ethnic minorities in their areas.
Australia also has a multicultural society, and The Centre for Culture and Health of the University of New South Wales in Sydney has an active program aimed at increasing cultural competency, both among medical students at the University and in the country\'s medical community at large (<http://cch.med.unsw.edu.au/>). The Centre offers graduate certificates and diplomas in public health (culture and health), as well as a Masters in Public Health with a concentration in multicultural health, and a postgraduate research degree. It emphasizes the establishment of partnerships with Area Health Services around New South Wales, grassroots organizations, and governmental organizations. A number of research projects also are underway. There are, for example, intervention strategies designed to reduce risk for cardiovascular disease in various cultural groups, such as the Arabic and Farsi-speaking communities, and studies of cancer among Chinese families in Australia.
Conclusion {#s7}
==========
People\'s basic medical needs do not vary greatly; they can be accommodated with appropriate understanding, awareness, and education. In the end, medicine and health care can only be enhanced and informed by the broadening of cultural awareness.
Further Reading on Multicultural Medicine
-----------------------------------------
Here are three captivating books that yield knowledge through narrative. *The Spirit Catches You and You Fall Down: A Hmong Child, Her American Doctors, and the Collision of Two Cultures* by Anne Fadiman (Farrar, Straus and Giroux, 1997).*Healing by Heart: Clinical and Ethical Stories of Hmong Families and Western Providers* by Kathleen Culhane-Pera and coauthors (Vanderbilt University Press, 2003).*Healing Latinos: Realidad y Fantasia*, a collection of physician-patient vignettes edited by David E. Hayes-Bautista and the late Roberto Chiprut (Cedars Sinai Health Systems and the UCLA Center for Latino Health, 1998).
**Citation:** McBride G (2005) The coming of age of multicultural medicine. PLoS Med 2(3): e62.
DHHS
: Department of Health and Human Services
[^1]: Gail McBride is a medical journalist and editor based in Sutter Creek, California, United States of America. She is the former managing editor of *Medicine of the Americas*, a journal devoted to multicultural medicine that was published in 2000--2001. E-mail: <gmcbride@twinwolf.net>
[^2]: **Competing Interests:** The author declares that she has no competing interests.
|
PubMed Central
|
2024-06-05T03:55:55.014708
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069660/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e62",
"authors": [
{
"first": "Gail",
"last": "McBride"
}
]
}
|
PMC1069661
|
PRESENTATION of CASE {#s1}
====================
A 54-y-old white male presented to the hospital during the winter of 2002 with complaints of fever, shortness of breath, and left-sided pleuritic chest pain of 2 d duration. He was a nonsmoker without any significant family history of pulmonary disease. He was retired, and denied any past exposure to chemicals including beryllium. On physical examination, the patient was febrile with tachycardia and normal blood pressure. Air entry at both lung bases was diminished. The remainder of his history and physical examination was unremarkable.
He had been hospitalized twice before for similar complaints during the prior 6 wk. Investigations during the first admission showed mild leukocytosis with normal electrolytes. Chest radiograph at that time showed bilateral pleural effusions and cardiomegaly, while echocardiography showed cardiac tamponade for which he underwent emergency pericardiocentesis. A computed tomography (CT) scan of the chest after the pericardiocentesis showed bilateral pleural effusions without hilar or mediastinal lymphadenopathy. A CT scan of the abdomen and pelvis, to look for any occult malignancy, was normal. Therapeutic thoracentesis was performed to relieve his symptoms. Pleural and pericardial fluid analyses were exudative without malignant cells, and were negative on culture for bacterial, mycobacterial, and fungal organisms. Investigations for HIV, syphilis, rheumatological diseases, hepatitis, and occult malignancies were all negative. The patient improved symptomatically with pericardiocentesis and thoracentesis, and he was discharged with the differential diagnosis of an atypical vasculitic syndrome or a paraneoplastic syndrome.
On the second admission, he presented with increasing shortness of breath, and a CT scan of the chest showed recurrent bilateral pleural effusions without pulmonary disease. Therapeutic thoracentesis was performed, relieving his symptoms, and he was discharged.
During his third (current) admission, laboratory investigations were unremarkable. Repeat CT scan of the chest showed a left-sided pleural effusion, normal lung parenchyma, a small pericardial effusion, and mediastinal lymph nodes ([Figure 1](#pmed-0020063-g001){ref-type="fig"}). Video-assisted thoracoscopic pleural biopsy showed nonspecific chronic inflammation. Mediastinal lymph node biopsy showed benign reactive lymph nodes with focal epithelioid, non-caseating granulomas consistent with sarcoidosis. The patient was treated with oral prednisone, 20 mg daily. He remained asymptomatic when seen at his 1-y follow-up, when he was taking 7.5 mg of prednisone daily.
::: {#pmed-0020063-g001 .fig}
Figure 1
::: {.caption}
###### CT of the Chest during the Third Hospitalization
The CT shows mediastinal lymphadenopathy (pink arrows) and left pleural effusion.
:::
:::
DISCUSSION {#s2}
==========
Sarcoidosis is characterized by non-caseating granulomas in affected organs. It primarily affects the lungs. Pleural effusion occurs in 5% of patients and may be the presenting feature of the disease \[[@pmed-0020063-b01]\]. Cardiac manifestations include bundle branch block, arrhythmia, congestive heart failure, pericarditis, and cardiomyopathy. Asymptomatic minimal pericardial effusion has been shown to occur in 20% of cases \[[@pmed-0020063-b02],[@pmed-0020063-b03]\]. Sarcoidosis presenting with pleural and pericardial effusion is extremely rare, and only one previous case has been reported \[[@pmed-0020063-b04]\].
Patients presenting with coincident pleural and pericardial effusions need to be investigated for rheumatological diseases, occult malignancies, and chronic infections such as HIV, tuberculosis, hepatitis, and syphilis. The diagnosis of sarcoidosis requires the presence of clinical and radiographic findings suggestive of sarcoidosis, non-caseating granulomas found on biopsies obtained from one or more sites, and the exclusion of other granuloma-forming diseases \[[@pmed-0020063-b05]\]. The differential diagnosis of non-caseating granulomas includes mycobacterial infections, berylliosis, and sarcoidosis. Our patient had clinical and radiological features suggestive of sarcoidosis and had non-caseating granulomas on his mediastinal lymph node biopsy, and we excluded other possible causes of recurrent pleural and pericardial effusions.
The optimal management strategy for sarcoidosis still remains uncertain. Asymptomatic pulmonary sarcoidosis is best treated with a wait-and-watch approach \[[@pmed-0020063-b06]\]. But steroids remain the mainstay of treatment for systemic sarcoidosis involving the cardiovascular system, nervous system, or eyes, and for cases with progressive pulmonary involvement \[[@pmed-0020063-b07],[@pmed-0020063-b08]\]. Earlier institution of steroids in cardiac sarcoidosis may prevent progressive disease and improve outcomes \[[@pmed-0020063-b09],[@pmed-0020063-b10]\]. Cytotoxic agents are used as steroid-sparing agents in patients requiring large doses of steroids and who experience serious side effects \[[@pmed-0020063-b11]\].
Learning Points {#sb1}
---------------
• Patients with recurrent pleural and pericardial effusion should be investigated for chronic infections, rheumatological illnesses, and malignancies.
• Sarcoidosis should be included in the differential diagnosis of patients presenting with bilateral pleural and pericardial effusion, as early treatment may improve the outcome in cardiac sarcoidosis.
• Steroids remain the mainstay of treatment for systemic sarcoidosis involving the cardiovascular system, nervous system, or eyes, and for cases with progressive pulmonary involvement.
**Citation:** Navaneethan SD, Venkatesh S, Shrivastava R, Mehta J, Israel R (2005) Recurrent pleural and pericardial effusions due to sarcoidosis. PLoS Med 2(3): e63.
CT
: computed tomography
[^1]: **Competing Interests:** The authors have declared that no competing interests exist.
[^2]: **Author Contributions:** SV, RS, and JM cared for the patient. SDN, SV, and RI wrote the manuscript.
|
PubMed Central
|
2024-06-05T03:55:55.015943
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069661/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e63",
"authors": [
{
"first": "Sankar D",
"last": "Navaneethan"
},
{
"first": "Sundar",
"last": "Venkatesh"
},
{
"first": "Rakesh",
"last": "Shrivastava"
},
{
"first": "Jagat",
"last": "Mehta"
},
{
"first": "Robert",
"last": "Israel"
}
]
}
|
PMC1069662
|
Introduction {#s1}
============
This article is an invitation to doctors around the world to participate in the CRASH-2 trial (Clinical Randomisation of an Antifibrinolytic in Significant Haemorrhage), a large, multi-centre, randomised controlled trial of a simple and widely practicable treatment for traumatic hemorrhage. The rationale for the trial and contact details for those who would like to take part are given below.
Evidence from randomised controlled trials is essential for improving health care. In the case of widely practicable treatments for common health problems, even modest treatment effects can result in substantial health gains. However, to detect such modest effects requires large multi-centre randomised trials involving hundreds of collaborating health professionals internationally. Many health professionals would be pleased to collaborate in such trials if they knew that they were underway, but at present there is no easy way to bring these trials to their attention.
Three years ago, in the context of the CRASH-1 trial (Corticosteroid Randomisation After Significant Head Injury), the trial investigators sent a message to the electronic mailing list of the World Association of Medical Editors, asking them to consider publishing an editorial about the CRASH-1 trial that invited doctors around the world to participate. In response to this request, many medical journals around the world published the CRASH-1 trial editorial in various different languages, and as a result, many more doctors joined the CRASH-1 trial. The trial was completed in May 2004 and involved around 400 hospitals in almost 50 countries, and because of its size provided a reliable answer to an important question (see [www.crash.lshtm.ac.uk](www.crash.lshtm.ac.uk)).
This current article is being published as the result of a similar such request to medical editors in the context of the CRASH-2 trial.
A Possible Role for Antifibrinolytics {#s2}
=====================================
For people at ages five to 45 years, trauma is second only to HIV/AIDS as a cause of death. Each year, worldwide, over three million people die as a result of trauma, many after reaching hospital \[[@pmed-0020064-b1]\]. Among trauma patients who do survive to reach hospital, exsanguination is a common cause of death, accounting for nearly half of in-hospital trauma deaths \[[@pmed-0020064-b2]\]. Central nervous system injury and multi-organ failure account for most of the remainder, both of which can be exacerbated by severe bleeding \[[@pmed-0020064-b3]\].
The haemostatic system helps to maintain the integrity of the circulatory system after severe vascular injury, whether traumatic or surgical in origin \[[@pmed-0020064-b4]\]. Major surgery and trauma trigger similar haemostatic responses, and any consequent massive blood loss presents an extreme challenge to the coagulation system. Part of the response to surgery and trauma, in any patient, is stimulation of clot breakdown (fibrinolysis) which may become pathological (hyper-fibrinolysis) in some \[[@pmed-0020064-b4]\]. Antifibrinolytic agents have been shown to reduce blood loss in patients with both normal and exaggerated fibrinolytic responses to surgery, and do so without apparently increasing the risk of post-operative complications; most notably there is no increased risk of venous thromboembolism \[[@pmed-0020064-b5]\].
Systemic antifibrinolytic agents are widely used in major surgery to prevent fibrinolysis and thus reduce surgical blood loss. A recent systematic review \[[@pmed-0020064-b6]\] of randomised controlled trials of antifibrinolytic agents (mainly aprotinin or tranexamic acid) in elective surgical patients identified 89 trials including 8,580 randomised patients (74 trials in cardiac, eight in orthopaedic, four in liver, and three in vascular surgery). The results showed that these treatments reduced the numbers needing transfusion by one third, reduced the volume needed per transfusion by one unit, and halved the need for further surgery to control bleeding. These differences were all highly statistically significant. There was also a statistically non-significant reduction in the risk of death (relative risk = 0.85: 95% confidence interval, 0.63--1.14) in the antifibrinolytic-treated group.
Why a Large Trial Is Needed {#s3}
===========================
Because the haemostatic abnormalities that occur after injury are similar to those after surgery, it is possible that antifibrinolytic agents might also reduce blood loss, the need for transfusion and mortality following trauma. However, to date there has been only one small randomised controlled trial (70 randomised patients, drug versus placebo: zero versus three deaths) of the effect of antifibrinolytic agents in major trauma \[[@pmed-0020064-b7]\]. As a result, there is insufficient evidence to either support or refute a clinically important treatment effect. Systemic antifibrinolytic agents have been used in the management of eye injuries where there is some evidence that they reduce the rate of secondary haemorrhage \[[@pmed-0020064-b8]\].
A simple and widely practicable treatment that reduces blood loss following trauma might prevent thousands of premature trauma deaths each year, and secondly, could reduce exposure to the risks of blood transfusion. Blood is a scarce and expensive resource, and major concerns remain about the risk of transfusion-transmitted infection. Trauma is common in parts of the world where the safety of blood transfusion is not assured. A recent study in Uganda estimated the population-attributable fraction of HIV acquisition as a result of blood transfusion to be around 2%, although some estimates are much higher \[[@pmed-0020064-b9],[@pmed-0020064-b10]\]. Only 43% of the 191 WHO member states test blood for HIV, Hepatitis C, and Hepatitis B viruses. Every year, unsafe transfusion and injection practices are estimated to account for 8--16 million Hepatitis B infections, 2.3--4.7 million Hepatitis C infections, and 80,000--160,000 HIV infections \[[@pmed-0020064-b11]\]. A large randomised trial is therefore needed of the use of a simple, inexpensive, widely practicable antifibrinolytic treatment such as tranexamic acid (aprotinin is considerably more expensive and is a bovine product with consequent risk of allergic reaction and hypothetically transmission of disease), in a wide range of trauma patients, who when they reach hospital are thought to be at risk of major haemorrhage that could significantly affect their chances of survival.
A Call to Health Professionals {#s4}
==============================
The CRASH-2 trial will be a large, international, placebo-controlled trial of the effects of the early administration of the antifibrinolytic agent tranexamic acid on death, vascular events and transfusion requirements (<http://www.crash2.lshtm.ac.uk>). The trial aims to recruit some 20,000 patients with trauma and will be one of the largest trauma trials ever conducted. However, it will only be possible to conduct such a trial if hundreds of health care professionals worldwide work together to recruit patients to the trial in order to make it a success. If you are interested in recruiting patients, please contact Ian Roberts at the CRASH-2 trial coordinating centre ([Box 1](#box1){ref-type="boxed-text"}).
Box 1. Contact Information for the CRASH-2 Trial
------------------------------------------------
Ian RobertsCRASH-2 Trial Co-Ordinating CentreLondon School of Hygiene and Tropical MedicineKeppel Street, London WC1E 7HTPhone: 0207 958 8128Fax: 0207 299 4663Web site: [www.crash2.lshtm.ac.uk](www.crash2.lshtm.ac.uk)E-mail: <Ian.roberts@lshtm.ac.uk>
*A similar version of this article is being published in several medical journals worldwide.*
::: {#pmed-0020064-g001 .fig}
:::
**Citation:** Coats T, Hunt B, Roberts I, Shakur H (2005) Antifibrinolytic agents in traumatic haemorrhage. PLoS Med 2(3): e64.
CRASH-1
: Corticosteroid Randomisation After Significant Head Injury
CRASH-2
: Clinical Randomisation of an Antifibrinolytic in Significant Haemorrhage
[^1]: Tim Coats is at the University of Leicester, Leicester, United Kingdom. Beverley Hunt is at the Guy\'s and St Thomas\' Trust, London, United Kingdom. Ian Roberts and Haleema Shakur are at the London School of Hygiene and Tropical Medicine, London, United Kingdom.
[^2]: **Competing Interests:** The authors are the investigators of the CRASH-2 trial but have no other competing interests.
|
PubMed Central
|
2024-06-05T03:55:55.017039
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069662/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e64",
"authors": [
{
"first": "Tim",
"last": "Coats"
},
{
"first": "Beverly",
"last": "Hunt"
},
{
"first": "Ian",
"last": "Roberts"
},
{
"first": "Haleema",
"last": "Shakur"
}
]
}
|
PMC1069663
|
Introduction {#s1}
============
In 1962, Norman Stoll, the distinguished Rockefeller Institute scientist who helped to establish human parasitology research in North America, described the unique health impact of hookworm as follows \[[@pmed-0020067-b1]\]: "As it was when I first saw it, so it is now, one of the most evil of infections. Not with dramatic pathology as are filariasis, or schistosomiasis, but with damage silent and insidious. Now that malaria is being pushed back hookworm remains the great infection of mankind. In my view it outranks all other worm infections of man combined...in its production, frequently unrealized, of human misery, debility, and inefficiency in the tropics."
Like many other global disease experts who witnessed dramatic reductions in malaria prevalence as a result of DDT spraying during the late 1950s \[[@pmed-0020067-b2]\], Stoll did not anticipate malaria\'s imminent re-emergence in India. However, he articulated with eloquence the magnitude of the disease burden resulting from hookworm infection. He further offered the silent and insidious character of hookworm as a partial explanation for its neglect by the global medical community.
This neglect subsequently intensified during the 1970s, 1980s, and 1990s with the omission of hookworm from the list of diseases covered by the World Health Organization\'s Special Programme for Research and Training in Tropical
> Hookworm has proven to be extremely difficult to eliminate or eradicate in areas of poverty and poor sanitation.
Diseases, as well as from other global health initiatives. Over the last ten years, however, there has been increasing recognition of the global health importance of hookworm. Today, new international efforts to control the morbidity of hookworm and other soil-transmitted helminth infections are in progress ([www.who.int/wormcontrol](www.who.int/wormcontrol)).
Etiology and Global Distribution {#s2}
================================
Human hookworm infection is caused by blood-feeding nematode parasites of the genus Ancylostoma and the species Necator americanus. Worldwide, N. americanus is the predominant etiology of human hookworm infection, whereas A. duodenale occurs in more scattered focal environments \[[@pmed-0020067-b3]\]. These two hookworms, together with the roundworm, Ascaris lumbricoides, and the whipworm, Trichuris trichiura, are often referred to collectively as soil-transmitted helminths (STHs).
No international surveillance mechanisms are in place to determine the prevalence and global distribution of hookworm infection. However, based on an extensive search of the literature since 1990, the worldwide number of cases of hookworm was recently estimated to be 740 million people \[[@pmed-0020067-b4]\]. The highest prevalence of hookworm occurs in sub-Saharan Africa and eastern Asia ([Figure 1](#pmed-0020067-g001){ref-type="fig"}). High transmission (defined below) also occurs in other areas of rural poverty in the tropics, including southern China \[[@pmed-0020067-b5]\], the Indian subcontinent \[[@pmed-0020067-b6]\], and the Americas \[[@pmed-0020067-b7]\]. In all regions, there is a striking relationship between hookworm prevalence and low socioeconomic status ([Figure 2](#pmed-0020067-g002){ref-type="fig"}) \[[@pmed-0020067-b4]\]. Hookworm\'s neglected status partly reflects its concentration among the world\'s poorest 2.7 billion people who live on less than \$2 a day.
::: {#pmed-0020067-g001 .fig}
Figure 1
::: {.caption}
###### Global Distribution of Human Hookworm Infection
(Illustration: Margaret Shear, Public Library of Science, adapted from \[[@pmed-0020067-b4]\])
:::
:::
::: {#pmed-0020067-g002 .fig}
Figure 2
::: {.caption}
###### The Relationship between Poverty and Hookworm Prevalence
(Illustration: Margaret Shear, Public Library of Science, adapted from \[[@pmed-0020067-b4]\])
:::
:::
Clinical Features, Epidemiology, and Disease Burden {#s3}
===================================================
Hookworm infection is acquired by invasion of the infective larval stages through the skin (A. duodenale larvae are also orally infective). Following host entry, the larvae undergo a journey through the vasculature, then the lungs and other tissues, before they enter the gastrointestinal tract and molt twice to become one-centimeter-long adult male and female worms \[[@pmed-0020067-b3]\]. The worms mate and the female hookworms produce up to 30,000 eggs per day, which exit the host\'s body in the feces ([Figure 3](#pmed-0020067-g003){ref-type="fig"}).
::: {#pmed-0020067-g003 .fig}
Figure 3
::: {.caption}
###### Life Cycle of the Human Hookworm N. americanus
The BZA anthelminthics albendazole and mebendazole remove adult hookworms from the gastrointestinal tract. In contrast, the *Na*-ASP-2 Hoookworm Vaccine is designed to target third-stage infective larvae (filariform larvae). Humoral immunity to the vaccine inhibits the entry of larvae into the gastrointestinal tract and thereby prevents their development into blood-feeding adult parasites. (Illustration: Sapna Khandwala, Public Library of Science, adapted from \[[@pmed-0020067-b3]\] and \[[@pmed-0020067-b33]\])
:::
:::
Because hookworms do not replicate in humans, the morbidity of hookworm is highest among patients that harbor large numbers of adult parasites. Estimates of the intensity of hookworm infection are typically obtained by using quantitative fecal egg counts as a surrogate marker for worm burden. The World Health Organization defines moderate-intensity infections as those with 2,000--3,999 eggs per gram of feces, and heavy-intensity infections as those with 4,000 or more eggs per gram (p. 26 in \[[@pmed-0020067-b8]\]). Compared to other STH infections and schistosomiasis, hookworm infection exhibits a unique age-intensity profile---whereas the intensities for the former peak in childhood and adolescence, hookworm intensity usually either steadily rises in intensity with age or plateaus in adulthood \[[@pmed-0020067-b3],[@pmed-0020067-b9]\]. The biological basis for this observation is unknown \[[@pmed-0020067-b10]\].
Adult hookworms cause morbidity in the host by producing intestinal hemorrhage \[[@pmed-0020067-b3]\]. The adult hookworms then ingest the blood, rupture the erythrocytes, and degrade the hemoglobin \[[@pmed-0020067-b11]\]. Therefore, the disease attributed to hookworm is silent blood loss leading to iron deficiency anemia and protein malnutrition. There is a correlation between parasite intensity and host intestinal blood loss \[[@pmed-0020067-b12]\]; in children, women of reproductive age, and other populations with low iron stores, there is often a correlation between parasite intensity and reductions in host hemoglobin \[[@pmed-0020067-b3],[@pmed-0020067-b12],[@pmed-0020067-b13],[@pmed-0020067-b14],[@pmed-0020067-b15],[@pmed-0020067-b16]\]. In children, chronic heavy-intensity infections are associated with growth retardation, as well as intellectual and cognitive impairments; in pregnant women, they are associated with adverse maternal--fetal outcomes \[[@pmed-0020067-b3],[@pmed-0020067-b12],[@pmed-0020067-b13],[@pmed-0020067-b14],[@pmed-0020067-b15],[@pmed-0020067-b16]\].
When measured in disability-adjusted life years, the global disease burden from hookworm exceeds all other major tropical infectious diseases with the exception of malaria, leishmaniasis, and lymphatic filariasis (pp. 192--193 in \[[@pmed-0020067-b17]\]). In addition, hookworm has been associated with impaired learning, increased absences from school, and decreased future economic productivity \[[@pmed-0020067-b18]\]. Therefore, like other neglected diseases, chronic infection with hookworm promotes long-term disability and increases the likelihood that an afflicted population will remain mired in poverty.
Hookworm Control Strategies {#s4}
===========================
Because of its high transmission potential, hookworm has proven to be extremely difficult to eliminate or eradicate in areas of poverty and poor sanitation \[[@pmed-0020067-b19]\]. Indeed, in the absence of comprehensive economic development, the impact of sanitation, footwear, and health education has been minimal \[[@pmed-0020067-b19]\]. Control efforts have therefore shifted to reducing morbidity through mass treatment (also known as "deworming") of affected populations with anthelminthic drugs \[[@pmed-0020067-b19]\].
Although benzimidazoles (BZAs) are the most commonly used agents for treating STH infections, levamisole and pyrantel may also be used in some circumstances. Periodic and repeated deworming with BZAs and praziquantel, complemented by basic sanitation and adequate safe water, is considered the most cost-effective means to control the morbidity caused by STH and schistosome infections \[[@pmed-0020067-b19],[@pmed-0020067-b20],[@pmed-0020067-b21],[@pmed-0020067-b22]\]. Efforts led by the World Health Organization have focused on annual, twice-yearly, or thrice-yearly doses in schools because the heaviest intensities of STH infections are most commonly encountered in school-age children \[[@pmed-0020067-b23]\].
Among the health benefits of periodic deworming of schoolchildren are improvements in iron and hemoglobin status, physical growth and fitness, and cognition \[[@pmed-0020067-b20],[@pmed-0020067-b21],[@pmed-0020067-b22],[@pmed-0020067-b23]\]. In addition, there are important externalities, including improvements in education and reduced community-based transmission of ascaris and trichuris infections \[[@pmed-0020067-b23]\]. Accordingly, at the 54th World Health Assembly in 2001, a resolution was passed urging member states to attain a minimum target of regular deworming of at least 75% and up to 100% of all at-risk school-age children by 2010 \[[@pmed-0020067-b20],[@pmed-0020067-b23]\].
Developing a New Control Tool: The *Na*-ASP-2 Hookworm Vaccine {#s5}
==============================================================
Deworming satisfies a number of United Nations Millennium Development Goals including those related to poverty reduction, child health, and education. However, there are also several reasons to believe that the school-based deworming programs could have less of an impact on the control of morbidity from hookworm than from other STH and schistosome infections \[[@pmed-0020067-b3]\]. As noted above, heavy-intensity hookworm infections are common among both adults and children, so school-based programs would not be expected to have an impact on hookworm transmission in the community \[[@pmed-0020067-b24]\]. School-based programs are also not likely to affect either preschool children or pregnant women, despite evidence for the health benefits from BZAs in both populations \[[@pmed-0020067-b16],[@pmed-0020067-b25]\]. Finally, a single dose of mebendazole (one of the two major BZAs) has variable efficacy against hookworm \[[@pmed-0020067-b26]\], and following treatment, hookworm reinfection to pre-treatment levels can occur within 4--12 months \[[@pmed-0020067-b27]\]. This, and the observation that the efficacy of mebendazole against hookworm can diminish with frequent and repeated use, has prompted concerns about the possible emergence of BZA resistance \[[@pmed-0020067-b28]\].
As a complementary strategy, the Human Hookworm Vaccine Initiative (HHVI) is developing a safe, efficacious, and cost-effective vaccine, the *Na*-ASP-2 Hookworm Vaccine, that would provide an additional tool for the control of hookworm \[[@pmed-0020067-b29],[@pmed-0020067-b30]\]. The HHVI is a non-profit partnership comprising research, process development, vaccine manufacturing and control, and pre-clinical and clinical testing units at the George Washington University, London School of Hygiene and Tropical Medicine, and Oswaldo Cruz Foundation (FIOCRUZ), and sponsored by the Sabin Vaccine Institute ([www.sabin.org](www.sabin.org)).
The HHVI selected the hookworm larval antigen ASP-2 (ancylostoma secreted protein-2) based on studies that (1) identified the molecule as a protective antigen linked to earlier-generation irradiated infective larval vaccines \[[@pmed-0020067-b29]\], (2) determined a relationship between human anti-ASP-2 antibodies and reduced risk of heavy hookworm infection in populations living in hookworm-endemic regions of Brazil and China (\[[@pmed-0020067-b30]\]; J. Bethony, A. Loukas, M. J. Smout, S. Brooker, S. Mendez, et al., unpublished data), and (3) confirmed the ability of recombinant ASP-2 to partially protect laboratory animals against larval hookworm challenges \[[@pmed-0020067-b30],[@pmed-0020067-b31],[@pmed-0020067-b32]\].
Process development, cGMP manufacture and control, and pre-clinical testing of *Na*-ASP-2 from N. americanus were completed in 2004 ([Figure 4](#pmed-0020067-g004){ref-type="fig"}). Pending United States Food and Drug Administration approval, clinical testing of the vaccine will take place in 2005. The *Na*-ASP-2 Hookworm Vaccine will be developed almost entirely in the non-profit sector. Ultimately, the vaccine will be indicated for the active immunization of susceptible individuals against moderate and heavy necator infection. Vaccination would reduce the number of hookworm infective larvae entering the gastrointestinal tract, thereby reducing the number of adult worms and the fecal egg counts in individuals exposed to the larvae.
::: {#pmed-0020067-g004 .fig}
Figure 4
::: {.caption}
###### Scheme for the Development and Quality-Control Testing of the *Na*-ASP-2 Hookworm Vaccine, and Its Transition from the Laboratory into the Clinic
After the selection of ASP-2 from N. americanus (*Na*-ASP-2) as the lead candidate antigen based on a series of research and development (R&D) tests---which included immunoepidemiology studies identifying human correlates of immunity to hookworm and confirmatory laboratory animal vaccine trials---the recombinant antigen was expressed in yeast and then developed as a biologic through a well-defined product development strategy (PDS). By following the product development strategy, process development (PD) and manufacturing led to the generation of pilot batches at different scales prior to technology transfer to a cGMP manufacturing facility. Both process development and manufacturing rely on developing assays for the product\'s identity, color and appearance, purity, immunological recognition, and potency, as well as qualification of the assays for sensitivity, specificity, accuracy, and reproducibility. Each of these processes must maintain a high level of quality control by following a set of policies, protocols, and standard operating procedures. After the manufacturing of a cGMP product and the required pre-clinical animal testing, a clinical development plan (CDP) was generated. Because the *Na*-ASP-2 Hookworm Vaccine is a product destined for the world\'s poorest, it is being developed almost exclusively in the non-profit sector, along with government manufacturers in middle-income countries.
:::
:::
Hookworm as a Model {#s6}
===================
Because immunization would only affect hookworm larvae and not adult hookworms already residing in the gastrointestinal tract of infected individuals, the first dose of the vaccine would be administered following deworming. Therefore, use of the vaccine could build on the infrastructures developed as part of school-based programs. Given that hookworm afflicts only the world\'s most impoverished, a major hurdle for the development of the *Na*-ASP-2 Hookworm Vaccine is its small commercial market. Innovative financing mechanisms must be considered to produce this orphan biologic. Towards that end, the HHVI has partnered with manufacturers in hookworm-endemic middle-income countries that would commit to industrial scale-up of the *Na*-ASP-2 Hookworm Vaccine pending proof-of-principle for its efficacy. This approach might help to inform the development of business models for the production and distribution of orphan biologics for other neglected diseases.
The work discussed here was supported by the Human Hookworm Vaccine Initiative (HHVI) of the Sabin Vaccine Institute, a March of Dimes Clinical Research Grant (6FY-00-791), and the China Medical Board of New York. SB is supported by a Wellcome Trust Advanced Training Fellowship (073656). JB is supported by an International Research Scientist Development Award (IRSDA K01 TW00009).
**Citation:** Hotez PJ, Bethony J, Bottazzi ME, Brooker S, Buss P (2005) Hookworm: "The great infection of mankind." PLoS Med 2(3): e67.
BZA
: benzimidazole
HHVI
: Human Hookworm Vaccine Initiative
STH
: soil-transmitted helminth
[^1]: Peter Hotez, Jeff Bethony, and Maria Elena Bottazzi are Professor and Chair, Assistant Professor, and Associate Research Professor, respectively, in the Department of Microbiology and Tropical Medicine, The George Washington University, Washington, D. C., United States of America. Simon Brooker is Lecturer at the London School of Hygiene and Tropical Medicine, London, United Kingdom. Paulo Buss is President of the Oswaldo Cruz Foundation (FIOCRUZ), Rio de Janeiro, Brazil. The authors are also principal scientists of the Human Hookworm Vaccine Initiative of the Sabin Vaccine Institute, Bethesda, Maryland, United States of America.
[^2]: **Competing Interests:** PJH and MEB are inventors on an international patent application (PCT/US02/33106, filed November 11, 2002) entitled "Hookworm Vaccine." PJH is also Co-Chair of the Scientific Advisory Council of the Sabin Vaccine Institute and a member of the Academic Advisory Board for the Pfizer Fellowships in Infectious Diseases. The other authors declare that they have no competing interests.
|
PubMed Central
|
2024-06-05T03:55:55.018059
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069663/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e67",
"authors": [
{
"first": "Peter J",
"last": "Hotez"
},
{
"first": "Jeff",
"last": "Bethony"
},
{
"first": "Maria Elena",
"last": "Bottazzi"
},
{
"first": "Simon",
"last": "Brooker"
},
{
"first": "Paulo",
"last": "Buss"
}
]
}
|
PMC1069664
|
Health Benefits of Physical Activity {#s1}
====================================
Regular physical activity has been recognized to confer health benefits since antiquity \[[@pmed-0020068-b1]\]. However, for most of humankind, voluntary discretion over whether or not to exercise is a recent phenomenon limited to advanced industrialized societies.
A large body of epidemiological literature consistently documents greater longevity in persons who are physically active on a near-daily basis, and reveals inverse relationships between levels of daily exercise and incidence of major chronic disorders such as obesity \[[@pmed-0020068-b2]\], hypertension \[[@pmed-0020068-b3]\], diabetes \[[@pmed-0020068-b4]\], ischemic heart disease, and all causes of mortality \[[@pmed-0020068-b5],[@pmed-0020068-b6],[@pmed-0020068-b7],[@pmed-0020068-b8],[@pmed-0020068-b9],[@pmed-0020068-b10],[@pmed-0020068-b11],[@pmed-0020068-b12]\]. From a public health perspective, there is little question that even modest increases in daily activities such as walking or stair climbing would have important positive consequences in reducing the burden of illness.
However, knowledge of the likely health benefits accruing to the physically active so far has not been a sufficient stimulus to promote sustained changes in behavior for most of the American population. If education and public policies are insufficient to promote behavioral changes to increase physical activity among most people, can advances in biotechnology confer such benefits to individuals unable or unwilling to perform the necessary physical effort?
Translating Knowledge of Exercise Biology to Novel Therapeutics {#s2}
===============================================================
Greater knowledge of how cells and tissues are modified in response to recurring bouts of exercise provides a basis for more precise recommendations as to the mode, intensity, and amount of exercise required to produce specific health benefits (e.g., treatment of dyslipidemia \[[@pmed-0020068-b13]\], control of body weight \[[@pmed-0020068-b14]\], or prevention of diabetes \[[@pmed-0020068-b15]\]). In addition, an understanding of the molecular signaling events that drive the beneficial effects of exercise on human physiology could foster the development of novel drugs, devices, or biological agents designed to substitute for exercise.
Many individuals who otherwise would develop diabetes or cardiovascular disease would benefit if advances in exercise biology revealed novel measures to promote the favorable effects on insulin sensitivity, lipoprotein metabolism, and blood pressure that are known to accrue through regular physical activity.
Physiological Properties of Skeletal Muscle {#s3}
===========================================
What do we know about basic muscle and exercise biology? The cells that constitute our skeletal muscles are called myofibers---large multinucleated cells that may extend for the full length of individual muscles. There are different types of myofibers, which vary in size and with respect to metabolic and contractile capability \[[@pmed-0020068-b16]\] ([Figure 1](#pmed-0020068-g001){ref-type="fig"}). Skeletal myofibers are innervated by motor neurons that contact each myofiber, and the intensity, duration, and timing of each muscle contraction are determined by the pattern of motor neuron firing. A pattern of occasional intense contractions separated by longer periods of rest is called "phasic," while a pattern characterized by brief contractions occurring repeatedly over an extended period is called "tonic." Endurance training regimens like running or cycling employ tonic patterns of contractile work, and it is this form of habitual activity that serves best to reduce risk for obesity, diabetes, hypertension, and heart disease.
::: {#pmed-0020068-g001 .fig}
Figure 1
::: {.caption}
###### Specialized Myofibers in a Mammalian Skeletal Muscle
A cross-section of the gastrocnemius muscle of a mouse has been stained to detect myoglobin, which is found selectively in slow oxidative and fast oxidative myofibers (stained brown), but not in fast glycolytic myofibers (unstained). Human muscles exhibit a similar mosaic pattern. In response to sustained periods of motor nerve stimulation repeated daily for several weeks, the percentage of myofibers that contain myoglobin is increased, in synchrony with an increased abundance of mitochondria and a shift of myosin subtypes from fast glycolytic to slow or fast oxidative.
:::
:::
Dynamics of Muscle Mass {#s4}
=======================
Maintenance of normal muscle mass requires some minimal level of ongoing work activity, and building and maintaining muscle mass is most effectively done through phasic contractions. A slow but inexorable loss of muscle mass is a feature of advancing age in human populations \[[@pmed-0020068-b17]\]. Loss of muscle mass and strength is an important determinant of injury and disability in the elderly, but even rigorous weight training programs cannot completely counteract this age-related decline that becomes particularly troublesome in the eighth and ninth decades of life. Efforts to develop effective countermeasures to maintain muscle mass in the elderly constitute an active and important area of current research \[[@pmed-0020068-b18],[@pmed-0020068-b19],[@pmed-0020068-b20]\].
Although the molecular signaling mechanisms that transduce the effects of phasic patterns of work activity to modify muscle mass are incompletely understood, recent evidence implicates pathways that include the signaling molecules PI3 kinase, Akt, mTOR, S6K, and ERK, the ubiquitin ligases MAFbx and MuRF1, and transcription factors of the FOX superfamily in the control of both catabolic and anabolic processes \[[@pmed-0020068-b21],[@pmed-0020068-b22],[@pmed-0020068-b23],[@pmed-0020068-b24]\].
Contractile and Metabolic Properties {#s5}
====================================
With respect to variations in contractile and metabolic properties, myofibers are classified on a spectrum between two extremes on the basis of contractile (fast versus slow) and metabolic (glycolytic versus oxidative) properties. At one extreme, the fastest glycolytic fibers have high levels of enzymes that generate ATP via glycolysis but few mitochondria (approximately 1% of cell volume). At the other end of the spectrum, slow oxidative fibers generate force with slower kinetics but are capable of long periods of repeated contraction without fatigue. They are rich in mitochondria (3%--10% of cell volume). Other myofibers, called fast oxidative, are both relatively fast and resistant to fatigue, and are rich in mitochondria (like the slow oxidative fibers). Muscles composed primarily of fast glycolytic fibers are needed for rapid movements (e.g., escape from predators) but fatigue when sustained periods of activity are required (e.g., migration).
Most human muscles exhibit a mosaic pattern of different fiber types ([Figure 1](#pmed-0020068-g001){ref-type="fig"}), with a great deal of variation among individuals, which is influenced at least in part by patterns of use. When we exercise daily, or at least several times weekly, we deliver a stimulus to the specific muscle groups involved in these activities that is sufficient to alter specialized properties of myofibers within these muscles. While habitual physical activity promotes a great variety of physiological adaptations that alter vascular reactivity, cardiac function, adipocyte function, and neurophysiology, adaptive responses of skeletal myofibers confer at least some of the health benefits.
Patterning of skeletal muscle fiber composition is initially determined during embryonic development, but can be partially or completely overturned by stimuli applied to fully mature adult myofibers: by hormonal influences (e.g., thyroid hormone), but most importantly by different patterns of motor nerve activity and contractile work. Myofibers that experience phasic patterns of contractile work---brief bursts of activity interspersed within long periods of inactivity---will assume the fast glycolytic phenotype. Myofibers subjected to tonic patterns of work activity---sustained periods of repetitive contraction on a habitual basis---will take on fast oxidative or slow oxidative properties. Under experimental conditions in laboratory animals, it is possible to transform muscles completely from one myofiber phenotype to another in a reversible manner, solely by altering the pattern of neural stimulation. We know that having a high proportion of oxidative muscle fibers conveys health benefits, and the possibility to control fiber composition through therapeutic intervention is promising.
Molecular Signaling Pathways {#s6}
============================
At a cellular and molecular level, how does a fast glycolytic myofiber sense a tonic pattern of contractile activity and transduce that information to transform itself into a cell with fast oxidative or slow oxidative properties? We know that such signals must be transduced to the nucleus, activating certain genes and suppressing others, for myofiber plasticity to occur. We know the identities of some of the nuclear transcription factors that carry these signals, and of other proteins that regulate the function of these transcription factors ([Figure 2](#pmed-0020068-g002){ref-type="fig"}).
::: {#pmed-0020068-g002 .fig}
Figure 2
::: {.caption}
###### Molecular Signaling Pathways Link Changes in Contractile Activity to Changes in Gene Expression That Establish Myofiber Diversity
A tonic pattern of motor nerve activity promotes changes in intracellular calcium that trigger a variety of intracellular events that modify the function of nuclear transcription factors. The pathway transduced by calcineurin and NFAT is highlighted in larger type. Other signals are received by cell surface receptors to activate similar or parallel signaling events. Signaling proteins that participate in transducing effects of contractile activity to specific genes include ion channels (TRP), scaffolding proteins (Homer), protein phosphatases and protein kinases (calcineurin, CAMK, p38MAPK), DNA-binding transcription factors (shown in red; NFAT, MEF2, PGC-1, ATF2), and endogenous inhibitors (shown in blue; GSK3, HDAC, and MCIP) (inhibitors antagonize gene activation via the pathways indicated, in some cases acting as negative feedback regulators).
:::
:::
Quite a variety of intracellular messengers have been proposed to provide the proximate signals in exercising muscles to stimulate activity-dependent gene regulation. This discussion will focus on a signaling cascade mediated by calcineurin, a calcium-regulated protein phosphatase that signals to the nucleus via transcription factors of the nuclear factor of activated T cells (NFAT) family. Upon receipt of the appropriate calcium signal, calcineurin is activated and removes phosphate groups from NFAT, thereby permitting translocation of NFAT to the nucleus. Within the nucleus, NFAT binds DNA and activates transcription (in concert with other transcription factors) of relevant downstream target genes that encode proteins necessary for fast oxidative or slow oxidative myofiber phenotypes.
Calcineurin and NFAT proteins are abundant in skeletal myofibers, and several lines of evidence support the viewpoint that the calcineurin--NFAT pathway plays a role in mediating activity-dependent gene regulation in muscle \[[@pmed-0020068-b25],[@pmed-0020068-b26],[@pmed-0020068-b27],[@pmed-0020068-b28],[@pmed-0020068-b29],[@pmed-0020068-b30],[@pmed-0020068-b31],[@pmed-0020068-b32],[@pmed-0020068-b33],[@pmed-0020068-b34],[@pmed-0020068-b35],[@pmed-0020068-b36],[@pmed-0020068-b37],[@pmed-0020068-b38],[@pmed-0020068-b39]\]. For example, in mice genetically engineered to distinguish the inactive (cytoplasmic) from active (nuclear) forms of NFAT by means of a sensor, it is evident that NFAT is inactive in resting muscles, but activated by tonic patterns of muscle contraction (running or electrical stimulation of the motor nerve) \[[@pmed-0020068-b40]\]. Using other genetic manipulations in mice to produce in muscle a form of calcineurin that remains active even in the absence of calcium signals, myofibers are converted from fast glycolytic to fast oxidative or slow oxidative forms \[[@pmed-0020068-b41]\]. And in muscles of mice genetically engineered to lack calcineurin, fiber type switching is impaired \[[@pmed-0020068-b42]\].
Cellular Memory {#s7}
===============
Muscle contractions are initiated under the influence of the motor nerve by release of calcium from the sarcoplasmic reticulum, which triggers actin--myosin crossbridge cycling ([Figure 3](#pmed-0020068-g003){ref-type="fig"}). Calcium released via ryanodine receptors is completely sufficient to activate muscle contractions, and the effects are immediate (within milliseconds). It is also sufficient to initiate calcineurin--NFAT signaling to the nucleus, but cannot by itself sustain the signal in a manner necessary to promote myofiber remodeling \[[@pmed-0020068-b40]\]. Changes in gene expression evoked by neuromuscular activity are not immediate but require that the stimulus be sustained for an extended period (minutes to hours). Moreover, tonic stimulation of the motor nerve must be repeated daily, or nearly so, over several weeks for the changes in myofiber properties to become fully manifest. We have characterized this requirement for repetition of the activity stimulus over days as a form of "cellular memory." The effects of the tenth or 20th day of exercise are not the same as the effects of the first day. The myofiber somehow "remembers" not only the pattern of activity it has experienced today, but what has gone on over the preceding days or weeks, such that the changes in abundance of proteins that control contractile function and metabolism accrue over time.
::: {#pmed-0020068-g003 .fig}
Figure 3
::: {.caption}
###### Proposed Model for Cellular Memory, Based on Activity-Induced Changes in TRPC3---A Putative Store-Operated Calcium Channel
Neural activation triggers muscle contraction by releasing calcium stored within the sarcoplasmic reticulum (SR) through mechanisms that involve channel proteins called dihydropyridine receptors (DHPR) and ryanodine receptors (RYR). Inactive myofibers have a low abundance of TRPC3 channels, and calcium released from SR is not sufficient to maintain the calcium-regulated transcription factor NFAT in the nucleus. Under conditions of tonic activity (training stimulus), TRPC3 channels become more abundant, and are regulated by the scaffold protein Homer, which binds RYR. Under these conditions, the combined effect of calcium entering the cell via TRPC3 channels and exiting the SR via RYR channels maintains NFAT in the nucleus, where it promotes transcription of genes that establish the slow oxidative phenotype in myofibers. Once the slow oxidative phenotype is established (trained myofiber), the continued expression of TRPC3 allows this state to be maintained even with a less intensive tonic activity pattern of neural stimulation. (Figure adapted from \[[@pmed-0020068-b40]\].)
:::
:::
To explain this cellular memory, we propose that, as the bursts of contractile activity are sustained over time (through a tonic pattern of neural stimulation), a second source of calcium is mobilized from outside of the cell and enters via a class of calcium channels that are called "store-operated" or "non-voltage-dependent." This second source of calcium is not required for muscle contractions, but is required to sustain calcium-dependent signaling to the nucleus. Phasic patterns of contractile activity do not promote calcium entry via store-operated channels. Tonic patterns of activity, in contrast, would not only promote the mobilization of extracellular calcium but also increase the number of store-operated calcium channels with each bout of exercise. Myofibers would thereby grow progressively more responsive to tonic activity. Consistent with this model, we know that daily running increases the expression of a putative store-operated calcium channel called TRPC3. Moreover, increasing the abundance of TRPC3 in cultured myotubes prolongs the period in which intracellular calcium is elevated following a depolarizing stimulus, sustains the transcription factor NFAT within the nucleus, and augments expression of NFAT-dependent target genes \[[@pmed-0020068-b40]\].
A great deal of additional research remains to be done before we have a comprehensive understanding of how habitual physical activity promotes changes in gene expression in skeletal muscles, and in turn improves fitness and reduces risk for diabetes, hypertension, dyslipidemia, and coronary artery disease. However, studies of the relationships between the proteins of calcium metabolism and calcium-regulated signaling pathways---as described here in a simplified manner with respect to TRPC3, calcineurin, and NFAT proteins---are illustrative of progress in this field. Other notable findings point to additional signaling proteins (CAMK, p38MAPK, and AMPK) and transcription factors (PGC-1, MEF2, ATF2, PPARs) active in pathways that intersect with calcineurin--NFAT signaling \[[@pmed-0020068-b31],[@pmed-0020068-b43],[@pmed-0020068-b44],[@pmed-0020068-b45],[@pmed-0020068-b46],[@pmed-0020068-b47],[@pmed-0020068-b48]\] (see [Figure 2](#pmed-0020068-g002){ref-type="fig"}). It is encouraging that some of these proteins are attractive targets for drug discovery.
Summary and Conclusions {#s8}
=======================
Long the province of physiologists who have contributed valuable insights in past decades, exercise science more recently has attracted the attention of molecular biologists, who have recognized the biological interest and medical importance of this field. Biotechnology and pharmaceutical companies also are beginning to take interest.
This review has focused on adaptive responses of skeletal muscle to changing patterns of physical activity, and on the role of the calcium--calcineurin--NFAT signaling cascade in controlling gene expression in skeletal myofibers. Further advances in our understanding of signaling mechanisms that govern activity-dependent gene regulation in skeletal muscle could lead to drugs, gene therapy, or devices that can, at least in part, substitute for daily exercise. Although it is unlikely that such technologies would fully recapitulate exercise-induced adaptations that affect other tissues of the body, beneficial effects on work performance and whole-body metabolism have been demonstrated using gene transfer techniques to alter skeletal muscles in animal models. If it proves possible to drive similar effects in skeletal muscles in humans, the interventions capable of providing such effects would almost certainly find broad clinical application.
**Citation:** Williams RS, Kraus WE (2005) Exercise and health: Can biotechnology confer similar benefits? PLoS Med 2(3): e68.
NFAT
: nuclear factor of activated T cells
[^1]: R. Sanders Williams is Dean of the School of Medicine at Duke University, Durham, North Carolina, United States of America. His research interests are centered on the molecular biology of skeletal and cardiac muscle. Williams E. Kraus is Associate Professor of Medicine at Duke University, Durham, North Carolina, United States of America. His research interests are centered on the genetic determinants of cardiovascular disease and studies of exercise as a preventive and therapeutic modality in humans.
[^2]: **Competing Interests:** The authors declare that they have no competing interests.
|
PubMed Central
|
2024-06-05T03:55:55.019676
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069664/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e68",
"authors": [
{
"first": "R. Sanders",
"last": "Williams"
},
{
"first": "William E",
"last": "Kraus"
}
]
}
|
PMC1069665
|
Introduction {#s1}
============
Cell--cell and cell--matrix interactions play essential roles in tumor metastasis and angiogenesis. Integrins are one of the main classes of receptors involved in these processes. In addition to having adhesive functions, integrins transduce messages via various signaling pathways and influence proliferation and apoptosis of tumor cells, as well as of activated endothelial cells. One prominent member of this receptor class is the integrin αvβ3. It has been demonstrated that αvβ3 is an important receptor affecting tumor growth, local invasiveness, and metastatic potential \[[@pmed-0020070-b01],[@pmed-0020070-b02]\]. This integrin is expressed on various malignant tumors and mediates adhesion of tumor cells on a variety of extracellular matrix proteins, allowing these cells to migrate during invasion and extravasation \[[@pmed-0020070-b03]\].
The integrin αvβ3 is also highly expressed on activated endothelial cells during angiogenesis \[[@pmed-0020070-b04]\]. In contrast, expression of αvβ3 is weak in resting endothelial cells and most normal organ systems \[[@pmed-0020070-b05]\]. On activated endothelial cells, the receptor mediates migration through the basement membrane during formation of the new vessel, which is essential for sufficient nutrient supply of the growing tumor. Inhibition of the αvβ3-mediated cell--matrix interaction has been found to induce apoptosis of activated endothelial cells. Thus, the use of αvβ3 antagonists is currently being evaluated as a strategy for tumor-specific anti-cancer therapies \[[@pmed-0020070-b06],[@pmed-0020070-b07],[@pmed-0020070-b08]\]. Owing to the weak expression on non-activated endothelial cells, treatment with αvβ3 antagonists does not affect preexisting blood vessels. Inhibition of blood vessel formation in tumor models using αvβ3 antagonists not only blocks tumor-associated angiogenesis, but in some cases results in tumor regression \[[@pmed-0020070-b09]\].
However, αvβ3 antagonists can induce apoptosis not only of activated endothelial cells but also of αvβ3-positive tumor cells \[[@pmed-0020070-b10]\], resulting in a direct cytotoxic effect on tumor cells. Moreover, blocking of the receptor expressed on tumor cells can reduce invasiveness and spread of metastases \[[@pmed-0020070-b11]\]. Furthermore, αvβ3-binding molecules have been successfully used to "target" a variety of therapeutic agents to the tumor tissue. These include chemotherapeutic agents \[[@pmed-0020070-b12]\], cDNA-encoding anti-angiogenic genes \[[@pmed-0020070-b13]\], and T lymphocytes \[[@pmed-0020070-b14]\].
These encouraging experimental studies have already led to initial clinical trials evaluating the use of αvβ3 antagonists (e.g., vitaxin \[[@pmed-0020070-b15]\] and cilengitide \[[@pmed-0020070-b16]\]) in patients with various malignant tumors \[[@pmed-0020070-b17],[@pmed-0020070-b18],[@pmed-0020070-b19],[@pmed-0020070-b20]\]. Currently available imaging techniques are limited in monitoring treatment with this class of drugs. Anti-tumor activity is generally assessed by determining the percentage of patients in whom a significant reduction in tumor size is achieved during a relatively short period of therapy ("response rate"). Thus, this method may not be applicable for a form of therapy that is aimed at disease stabilization and prevention of metastases. New methods are urgently needed for planning and monitoring treatments targeting the αvβ3 integrin.
Based on cyclo(-Arg-Gly-Asp-[D]{.smallcaps}Phe-Val-) \[[@pmed-0020070-b21]\], a variety of radiolabeled αvβ3 antagonists for single photon emission tomography and positron emission tomography (PET) have been developed (for review see \[[@pmed-0020070-b22],[@pmed-0020070-b23]\]). \[^18^F\]Galacto-RGD (arginine--glycine--aspartic acid), a glycosylated cyclic pentapeptide, resulted from a consequent tracer optimization \[[@pmed-0020070-b24]\] based on the first-generation peptide \[^125^I\]-3-iodo-[D]{.smallcaps}Tyr^4^-cyclo(-Arg-Gly-Asp-[D]{.smallcaps}Tyr-Val-) \[[@pmed-0020070-b25]\] and showed high affinity and selectivity for the αvβ3 integrin in vitro, receptor-specific accumulation in αvβ3-positive tumors, and high metabolic stability in a murine tumor model, as well as rapid, predominantly renal elimination \[[@pmed-0020070-b26],[@pmed-0020070-b27]\]. Here we describe how \[^18^F\]Galacto-RGD allows quantification of αvβ3 expression in vivo, show that tumor-induced angiogenesis can be monitored in a murine tumor model, and for the first time, to our knowledge, demonstrate that this class of tracers can be used in patients for noninvasive determination of αvβ3 expression.
Methods {#s2}
=======
Tracer Synthesis {#s2a}
----------------
Synthesis of the labeling precursor and subsequent ^18^F-labeling was carried out as described \[[@pmed-0020070-b27]\]. For application in patients, after high-performance liquid chromatography the collected fraction was evaporated to dryness; 0.5 ml of absolute ethanol and 10 ml of phosphate-buffered saline (pH 7.4) were added; and the product was passed through a Millex GV filter (Millipore, Eschborn, Germany). Quality control of the product was carried out according to the demands of the local regulatory authorities.
Murine Tumor Models {#s2b}
-------------------
For in vivo evaluation, xenotransplanted human melanoma models (M21 and M21-L) and a human squamous cell carcinoma model (A431) were used. The M21 cell line expressing αvβ3 \[[@pmed-0020070-b25],[@pmed-0020070-b28]\] acted as a positive control and the M21-L cell line, a stable variant cell line of M21 failing to transcribe the αv gene, as a negative control \[[@pmed-0020070-b29]\]. Cell culture conditions for M21 and M21-L cells are described elsewhere \[[@pmed-0020070-b26]\]. Similar protocols were used for A431 cells.
The experimental protocol involving animals was approved by the Committee of Veterinarian Medicine of the State of Bavaria; handling of animals was performed according to the standards set by the Committee of Veterinarian Medicine.
In order to study the correlation between αvβ3 expression and tumor uptake of \[^18^F\]Galacto-RGD, we injected mice subcutaneously with mixtures of M21 and M21-L cells. Pilot experiments had indicated that injection of 1.5 × 10^6^ M21 cells leads within 4 wk to the formation of tumors with a diameter of approximately 8 mm. To obtain similarly sized M21-L tumors, it was necessary to inject 6 × 10^6^ cells. In order to study tumors with approximately 10%, 25%, 50%, and 75% M21 cells, we injected mice with the following mixtures of M21 and M21-L cells: 1.5 × 10^5^ / 5.4 × 10^5^, 3.8 × 10^5^ / 4.6 × 10^6^, 7.5 × 10^5^ / 3 × 10^6^, and 1.1 × 10^6^ / 1.5 × 10^6^. Four weeks after inoculation, nude mice were injected with 7.4 MBq of \[^18^F\]Galacto-RGD and scanned at the small-animal PET. Subsequently, tumors and other organs of interest were dissected, immediately counted, cut in two pieces, and frozen for further workup.
For experiments with the squamous cell carcinoma model, approximately 10^6^ A431 cells were injected subcutaneously in nude mice. Two weeks after inoculation, 7.4 MBq of \[^18^F\]Galacto-RGD was injected, and mice were scanned in the animal PET. Animals were sacrificed, and organs of interest were dissected and subsequently weighed and counted or used for immunohistochemical analysis.
Immunohistochemistry {#s2c}
--------------------
For immunohistochemical investigation, frozen tumor tissues from mice, as well as from patients, were sectioned (6 μm) and stained using the biotinylated monoclonal anti-αvβ3 antibody LM609 (1:100; Chemicon Europe, Hofheim, Germany). For staining the murine β subunit, a monoclonal hamster anti-mouse antibody (1:10; Chemicon Europe) and a biotinylated mouse anti-hamster IgG secondary antibody (1:200; Chemicon Europe) were used. Sections were processed by peroxidase staining (peroxidase substrate KIT AEC, Vector Laboratories, Burlingame, California, United States).
Western Blotting {#s2d}
----------------
The frozen tumor tissue was homogenized and extracted with lysis buffer (50 mM Hepes (pH 7.5), 150 mM NaCl, 10% Glycerol, 1% Triton X-100, 1 mM EDTA, 1 mM EGTA, 10 mM Na~4~P~2~O~7~, 1 mM MSF, 10 μg/ml Aprotinin, 10 μg/ml Leupeptin). Protein concentration was determined according to Bradford \[[@pmed-0020070-b30]\] and adjusted to equivalent values using lysis buffer. After SDS-PAGE and transfer, immunoblotting was carried out using a polyclonal rabbit anti-αv antibody (1:500; Chemicon Europe) and a ^125^I-labeled polyclonal donkey anti-rabbit IgG antibody (1:400; 477 kBq/μg, Amersham Buchler, Braunschweig, Germany). For analysis, blots were placed on a phosphor screen for 2 d. For readout out, a Molecular Dynamics PhosphorImager 445 SI (Sunnyvale, California, United States) was used.
PET Studies with a Small-Animal Scanner {#s2e}
---------------------------------------
PET imaging of tumor-bearing mice was performed using a prototype small-animal positron tomograph (Munich Avalanche Photodiode PET; \[[@pmed-0020070-b31]\]). The animal scanner consists of two sectors, comprising three detector modules each, which rotate around the animal for acquisition of complete projections in one transaxial slice (30 angular steps). Each module consists of eight small (3.7 × 3.7 × 12 mm3) lutetium-oxy-orthosilicate crystals read out by arrays of avalanche photodiodes. List mode data are reconstructed using statistical, iterative methods including the spatially dependent line spread function. Reconstructed image resolution is 2.5 mm (full width at half maximum) in a transaxial field of view of 7.5 cm, and the slice thickness is 2 mm. Ninety minutes after the injection of approximately 7.4 MBq of \[^18^F\]Galacto-RGD, animals were positioned prone inside the tomograph, and a transaxial slice through the tumor region was measured for 480 s
Patient Study {#s2f}
-------------
The study protocol was approved by the ethics committee of the Klinikum Rechts der Isar ([Protocol S1](#sd001){ref-type="supplementary-material"}), and each patient gave written and informed consent prior to the study ([Protocol S2](#sd002){ref-type="supplementary-material"}). Nine patients were scanned (five female and four male; age, 26--75 y), who suffered from either malignant melanoma with lymph node metastasis (stage IIIb; patients 1--3), malignant melanoma with distant metastasis (stage IV; patients 4 and 5), chondrosarcoma (patient 6), soft tissue sarcoma (patient 7), osseous metastasis of renal cell carcinoma (patient 8), or villonodular synovitis of the knee (patient 9). Patient selection was focused on melanoma and sarcoma because there is considerable evidence that these tumor types express αvβ3.
Diagnosis prior to scanning was made by biopsy (patients 6--8), by CT (patients 1, 2, and 4--8), by MRI (patient 9), and/or by \[^18^F\]fluorodeoxyglucose (\[^18^F\]FDG)--PET (patients 1--4 and 8). After scanning, the diagnosis was confirmed by surgery and histopathological examination of the resectioned specimen (patients 1, 2, 5, 6, 8, and 9) or by combined analysis of morphological imaging, \[^18^F\]FDG PET, and the patient\'s clinical data and history (patients 3, 4, and 7).
For immunohistochemistry, sampled specimens (patients 1, 2, 5, 6, 8, and 9) were snap frozen in liquid nitrogen and stored at −70 °C until staining was performed. Tissue samples were taken within 1 wk after scanning from the tumor regions with the maximum tracer uptake. Light microscopic evaluation of the density of αvβ3-positve microvessels was performed as described previously \[[@pmed-0020070-b32]\]. Briefly, areas with the highest density of αvβ3-positve microvessels were identified using scanning magnification. Subsequently, αvβ3-positve microvessels were counted in three adjacent microscopic fields using a 40× magnifying lens and 10× ocular, corresponding to an area of 0.588 mm^2^. Determination of microvessel density was performed by one senior pathologist (M. S.), who was blinded for the results of the corresponding standardized uptake value (SUV) analysis of tracer accumulation. Then the correlation between the mean values of the vessel counts and the corresponding SUVs was analyzed.
PET scanning was performed using an ECAT Exact PET scanner (Siemens-CTI, Knoxville, Tennessee, United States). After injection of 144--200 MBq of \[^18^F\]Galacto-RGD, three consecutive emission scans (starting at 7 ± 2.7 min, 37 ± 10.5 min, and 79 ± 18.4 min post injection \[p. i.\]) from the body stem and, if necessary, from tumor regions outside the body stem were obtained. For one patient, only one scan starting 120 min. p. i. was carried out. Attenuation- and decay-corrected images were reconstructed by using an ordered subsets expectation maximization algorithm. The accumulation of \[^18^F\]Galacto-RGD was evaluated by calculating the mean SUV normalized to the patient\'s body weight according to the following formula \[[@pmed-0020070-b33]\]: (measured activity concentration \[Bq/ml\] × body weight \[kg\]) / injected activity \[Bq\]. The axial slice of the lesion with the maximum activity accumulation was chosen by visual estimation, a region of interest with a diameter of 15 mm was selected on the lesion, and the resulting mean SUV was used for further analysis. For lesions smaller than 2 cm in diameter, a region of interest with a diameter of 10 mm was used and the analysis was based on maximum SUV rather than mean SUV, in order to minimize partial volume effects, which could lead to an underestimation of the SUV.
Dosimetry calculations are based on the MIRDOSE 3.0 program. Data from six patients were analyzed by selecting regions of interest with a diameter of 1.5 cm on the source organs (lung, liver, spleen, kidneys, muscle, bladder, intestine, and heart \[left ventricle\]). Activity measurements (in Becquerels per cubic centimeter) were performed for all three consecutive scans (mean time p.i. ± standard deviation, 7 ± 2.7 min, 37 ± 10.5 min, and 79 ± 18.4 min, respectively), using a monoexponential fit for calculation of residence times. The volume of the source organs lung, liver, spleen, and kidneys was measured by CT volumetry (Siemens, Forchheim, Germany) in four patients. For the other source organs in these four patients and all organs in the remaining two patients, standardized volume values of the source organs adapted to the patient\'s body weight were used.
Statistical Methods {#s2g}
-------------------
All quantitative data are expressed as mean +/− one standard deviation. The correlation between quantitative parameters was evaluated by linear regression analysis and calculation of Pearson\'s correlation coefficient. Statistical significance was tested by using analysis of variance (ANOVA).
Results {#s3}
=======
Correlation of Tracer Uptake with αvβ3 Expression {#s3a}
-------------------------------------------------
We have previously demonstrated, using a murine tumor model in which the tumor cells are either αvβ3-positive (human melanoma M21) or αv-negative (human melanoma M21-L), that \[^18^F\]Galacto-RGD shows receptor-specific accumulation in the αvβ3-positive tumor \[[@pmed-0020070-b26]\]. Here we studied the correlation of \[^18^F\]Galacto-RGD uptake with the level of αvβ3 expression. We injected tumor cell mixtures containing increasing amounts of αvβ3-positive M21 cells subcutaneously into nude mice. Transaxial images of mice 4 wk after cell inoculation and 90 min after tracer injection using a prototype small-animal PET scanner \[[@pmed-0020070-b31]\] showed increasing tracer uptake in the tumor corresponding with the percentage of receptor-positive cells ([Figure 1](#pmed-0020070-g001){ref-type="fig"}A and [1](#pmed-0020070-g001){ref-type="fig"}B).
::: {#pmed-0020070-g001 .fig}
Figure 1
::: {.caption}
###### Preclinical Evaluation of \[^18^F\]Galacto-RGD
\(A) Transaxial images of nude mice bearing tumors with increasing amounts of αvβ3-positive M21 cells (0% \[M21-L\], 25%, 75%, and 100% \[M21\]) 90 min p. i. provided by a prototype small-animal PET scanner show an increasing tracer uptake in the tumor and low background activity.
\(B) Immunohistochemical staining of tumor tissue sections prepared after PET imaging with an anti-human αvβ3 monoclonal antibody (LM 609) indicate that there is a correlation between tracer uptake and αvβ3 expression.
\(C) Western blots of the dissected tumors show a band at 25 kDa that corresponds with the mass of the αv subunit under reductive conditions, and indicate the increasing αvβ3 density in the murine tumor model used.
\(D) The correlation between receptor expression and \[^18^F\]Galacto-RGD accumulation is confirmed by quantitative analysis based on the tumor/background ratios and tumor/muscle ratios calculated from the PET images and from the biodistribution studies, respectively, and by the relative αv expression in Western blot analyses.
:::
:::
We validated these qualitative results by determining the relative amount of the αv subunit in the dissected tumors through Western blot analysis ([Figure 1](#pmed-0020070-g001){ref-type="fig"}C). These data were correlated with the tumor/background ratios resulting from the quantitative analysis of the PET images ([Figure 1](#pmed-0020070-g001){ref-type="fig"}D), as well as with the tumor/muscle ratios resulting from the biodistribution analysis carried out after the PET study ([Figure 1](#pmed-0020070-g001){ref-type="fig"}D). Both analyses showed a significant correlation between \[^18^F\]Galacto-RGD and relative αv expression, thus confirming the qualitative analysis by immunohistochemistry.
The systematic difference between tumor/background and tumor/muscle ratios is due to the fact that the region of interest used to define the tumor region in the PET images will always contain not only tumor, but also normal tissues with low \[^18^F\]Galacto-RGD uptake, such as muscle and lung. This is due to the limited spatial resolution of the PET scanner, which does not allow a sharp distinction between tumor and normal tissue. Accordingly \[^18^F\]Galacto-RGD uptake by the tumor tissue will be underestimated, and the tumor/background ratio will be lower than the tumor/muscle ratio. Furthermore, tissue sampling was performed 30 min after the start of the PET scan. Clearance of radioactivity from the muscle tissue during this time period will also systematically increase the tumor/muscle ratio compared to the tumor/background ratio calculated from the PET images.
When correlating the weight of the tumor with the relative αv expression, we found a nonsignificant trend for lower αv expression in larger tumors (*r* = 0.34, *p* = 0.09). This is probably related to the presence of necrotic regions in larger tumors, which do not demonstrate αv expression. Thus, it can be excluded that the positive correlation between αv expression and \[^18^F\]Galacto-RGD uptake is due to systematic differences in the size of tumors.
Noninvasive Determination of αvβ3 Expression on Endothelial Cells {#s3b}
-----------------------------------------------------------------
To determine whether PET with \[^18^F\]Galacto-RGD allows noninvasive determination of αvβ3 expression on activated endothelial cells, we used A431 tumor xenografts. A431 cells do not express αvβ3, but induce extensive angiogenesis when subcutaneously transplanted into nude mice. \[[@pmed-0020070-b34]\]. Immunohistochemical staining of tumor sections using a monoclonal anti-human αvβ3 antibody confirmed that the tumor cells do not express this integrin ([Figure 2](#pmed-0020070-g002){ref-type="fig"}A). In contrast, staining with a polyclonal antibody against the murine β3 subunit demonstrated expression of β3 on endothelial cells of the tumor vessels. Since αIIbβ3, the only further integrin containing a β3 subunit, is mainly expressed on platelets, it can be excluded that staining depends on this receptor. Thus, in this case, staining for the β3 subunit correlates with αvβ3 expression.
::: {#pmed-0020070-g002 .fig}
Figure 2
::: {.caption}
###### Noninvasive Monitoring of αvβ3 Expression on the Tumor Vasculature
\(A) Immunohistochemical staining of tumor section using the anti-αvβ3 monoclonal antibody LM609 demonstrates that squamous cell carcinoma cells of human origin do not express the αvβ3 integrin. In contrast, staining of section with an antibody against the murine β3 subunit indicates that the tumor vasculature is αvβ3-positive.
\(B) Transaxial images of a nude mouse bearing a human squamous cell carcinoma at the right shoulder (left) acquired at the small-animal PET 90 min after tracer injection show a clearly contrasting tumor. Tracer accumulation in the tumor (right, top image) can be blocked by injecting 18 mg of cyclo(-Arg-Gly-Asp-[D]{.smallcaps}Phe-Val-) per kilogram of mouse 10 min prior to tracer injection (right, bottom image), indicating receptor-specific accumulation.
:::
:::
Transaxial images of tumor-bearing mice 90 min after injection of \[^18^F\]Galacto-RGD showed a contrasting tumor on the right flank of the mouse, reflecting αvβ3-targeted tracer accumulation on endothelial cells of the tumor vasculature ([Figure 2](#pmed-0020070-g002){ref-type="fig"}B). Moreover, we demonstrated receptor-specific tracer accumulation at the tumor site by injecting 18 mg of the pentapeptide cyclo(-Arg-Gly-Asp-[D]{.smallcaps}Phe-Val-) per kilogram of mouse 10 min prior to tracer injection. After blocking tracer accumulation, we found approximately 25% of the initial activity in the tumor (0.28 ± 0.05% injected dose per gram versus 1.07 ± 0.33% injected dose per gram).
Studies in Humans {#s3c}
-----------------
For the initial evaluation in humans, we imaged nine patients (five with malignant melanomas, two with sarcomas, one with osseous metastasis from renal cell carcinoma, and one with villonodular synovitis) with approximately 185 MBq of \[^18^F\]Galacto-RGD. For all patients, rapid, predominantly renal excretion was observed, resulting in fast tracer elimination from blood and low tracer concentration in most of the organs. Besides the kidneys (SUV = 5.5 ± 3.7; 79 min), the highest activity concentration was found in spleen (SUV = 2.5 ± 0.5; 79 min p.i.), liver (SUV = 2.4 ± 0.5; 79 min p.i.), and intestine (SUV = 2.1 ± 0.8; 79 min p.i.). In tumor lesions, tracer accumulation showed great heterogeneity, with SUVs ranging from 1.2 to 10.0. The SUV in the villonodular synovitis was 3.2. The radioactivity was retained in the tumor tissue for more than 60 min ([Table 1](#pmed-0020070-t001){ref-type="table"}), whereas in all other organs a decrease of activity concentration was observed over time. Tumor/blood and tumor/muscle ratios 79 min p. i. were 3.8 ± 2.6 and 8.8 ± 6.0, respectively. Although for one melanoma patient multiple lesions were detected by the \[^18^F\]FDG scan, which indicates viable tumor cells, no activity accumulation was found using \[^18^F\]Galacto-RGD ([Figure 3](#pmed-0020070-g003){ref-type="fig"}A). For other patients, however, similar uptake patterns were observed for \[^18^F\]FDG and \[^18^F\]Galacto-RGD ([Figure 3](#pmed-0020070-g003){ref-type="fig"}B). The metabolite analysis of blood samples 10, 30, and 120 min p. i. showed in the soluble fractions more than 96% intact tracer (*n* = 4) over the whole observation period and confirmed our preclinical data \[[@pmed-0020070-b27]\]. An effective radiation dose of 18.0 ± 3.2 μSv/MBq was calculated on the basis of the patient data (*n* = 5). The highest absorbed dose was found in the urinary bladder wall (0.20 ± 0.04 mGy/MBq).
::: {#pmed-0020070-g003 .fig}
Figure 3
::: {.caption}
###### Comparison of \[^18^F\]FDG and \[^18^F\]Galacto-RGD Scans
Coronal image sections, acquired 60 min p. i.
\(A) Patient with malignant melanoma stage IV and multiple metastases in liver, skin, and lower abdomen (arrows): marked uptake of \[^18^F\]FDG in the lesions (left), but no uptake of \[^18^F\]Galacto-RGD (right).
\(B) Patient with malignant melanoma stage IIIb and a solitary lymph node metastasis in the right axilla (arrow): intense uptake of both \[^18^F\]FDG (left) and \[^18^F\]Galacto-RGD (right).
:::
:::
::: {#pmed-0020070-t001 .table-wrap}
Table 1
::: {.caption}
###### SUVs Determined Approximately 5 min, 35 min, and 75 min Post Injection
:::
Values are given as mean ± standard deviation (*n* = 8, unless otherwise indicated)
^a^ *n* = 9
^b^ *n* = 7
^c^ *n* = 10
:::
Immunohistochemical staining of sections obtained from tumor tissue after surgery using an anti-αvβ3 antibody showed αvβ3 expression on the endothelial cells of the tumor vasculature (6/6), and for two patients expression on the tumor cells as well (2/6) ([Figure 4](#pmed-0020070-g004){ref-type="fig"}). The density of αvβ3-positive vessels showed wide variation intraindividually and between individual cases. Light microscopic quantification revealed between one (inflammation of the knee due to previous operation) and 35 (soft tissue sarcoma of the knee, same patient) αvβ3-positive vessels per microscopic field. Moreover, in the six cases under analysis, density of immunohistochemically determined αvβ3-positive vessels was significantly associated with tracer accumulation as determined by SUV analysis (*r* = 0.94, *p* = 0.005).
::: {#pmed-0020070-g004 .fig}
Figure 4
::: {.caption}
###### Correlation of Tracer Accumulation and αvβ3 Expression
(A--C) patient with a soft tissue sarcoma dorsal of the right knee joint. (A) The sagittal section of a \[^18^F\]Galacto-RGD PET acquired 170 min p. i. shows circular peripheral tracer uptake in the tumor with variable intensity and a maximum SUV of 10.0 at the apical-dorsal aspect of the tumor (arrow). (B) The image fusion of the \[^18^F\]Galacto-RGD PET and the corresponding computed tomography scan after intravenous injection of contrast medium shows that the regions of intense tracer uptake correspond with the enhancing tumor wall, whereas the non-enhancing hypodense center of the tumor shows no tracer uptake. (C) Immunohistochemistry of a peripheral tumor section using the anti-αvβ3 monoclonal antibody LM609 demonstrates intense staining predominantly of tumor vasculature.
(D--F) patient with malignant melanoma and a lymph node metastasis in the right axilla. (D) The axial section of a \[^18^F\]Galacto-RGD PET acquired 140 min p. i. shows intense focal uptake in the lymph node (arrow). (E) Image fusion of the \[^18^F\]Galacto-RGD PET and the corresponding computed tomography scan after intravenous injection of contrast medium. (F) Immunohistochemistry of the lymph node using the anti-αvβ3 monoclonal antibody LM609 demonstrates intense staining predominantly of tumor cells and also blood vessels.
:::
:::
Discussion {#s4}
==========
Recently, we demonstrated that radiolabeled RGD peptides allow receptor-specific monitoring of αvβ3 expression in murine tumor models \[[@pmed-0020070-b24],[@pmed-0020070-b25],[@pmed-0020070-b26],[@pmed-0020070-b27],[@pmed-0020070-b35]\]. Here we have translated these findings to the clinical setting and for the first time, to our knowledge, demonstrated noninvasive imaging of αvβ3 expression in patients with malignant tumors. Furthermore, we have shown that the activity accumulation in the tumor correlates with the receptor density, determined by immunohistochemistry and Western blotting. This indicates that a noninvasive quantitative determination of αvβ3 expression is feasible. Furthermore, we have demonstrated in a squamous cell carcinoma model that the sensitivity of PET is adequate to image expression of αvβ3 in the tumor vasculature. This indicates that PET with \[^18^F\]Galacto-RGD can be applied to study αvβ3 expression during angiogenesis.
The correlation between \[^18^F\]Galacto-RGD uptake in the tumor and αv expression shows considerable scattering. This is probably due to several factors. As for any imaging probe, tumor uptake of \[^18^F\]Galacto-RGD is not only influenced by the expression of the αvβ3 integrin, but also by unspecific factors such as perfusion and vascular permeability. Furthermore, heterogeneous tracer uptake within a tumor, e.g., due to the presence of necrotic areas, will influence the correlation between \[^18^F\]Galacto-RGD uptake and αv expression, since separate samples were used for measurements of tracer uptake and quantitative assessment of αv expression. Finally, the present study evalutated \[^18^F\]Galacto-RGD uptake only at a fixed time, 90 min p. i. Imaging of the dynamics of \[^18^F\]Galacto-RGD accumulation in the tumor tissue and tracer kinetic modeling may allow a better quantitative assessment of αvβ3 expression by PET imaging, and this approach should be evaluated in animal models as well as in patients. Nevertheless, the significant correlation between the uptake of \[^18^F\]Galacto-RGD at a fixed time after injection and αvβ3 expression is very important for clinical studies, since it suggests that estimates of αvβ3 expression levels may be obtained from simple whole-body PET scans.
It has been shown that the highly bent integrin conformation is physiological and has low affinity for biological ligands, such as fibrinogen and vitronectin. Inside-out and outside-in signaling involve a switchblade-like opening to an extended structure with high affinity for endogenous ligands, as well as integrin antagonists (for overview see \[[@pmed-0020070-b36]\]). The inside-out activation is induced by conformational changes in the membrane-proximal regions of the α and β subunit (e.g., by intracellular proteins like talin). Outside-in signaling is triggered by Mn^2+^, which defines by quaternary rearrangements a pathway for communication from the ligand-binding site to the cytoplasmatic proximal segments. However, it is also reported that cyclo(-Arg-Gly-Asp-[D]{.smallcaps}Phe-Val-), in addition to binding to the high-affinity conformer, can bind to the low-affinity conformation when used at concentrations far above its dissociation constant, resulting in a similar activation as found for Mn^2+^. The nanomolar concentration used in our radiotracer approach is approximately 10,000-fold lower than that reported for the activation of the low-affinity conformation. Thus, PET with \[^18^F\]Galacto-RGD is expected to provide information not only about the expression of αvβ3 but also about the functional status of this integrin.
The glycopeptide \[^18^F\]Galacto-RGD showed high metabolic stability in patients and rapid, predominantly renal elimination, resulting in good tumor/background ratios and, thus, in high-quality images. Moreover, this finding confirms the general advantage of the glycosylation approach \[[@pmed-0020070-b24],[@pmed-0020070-b26],[@pmed-0020070-b37],[@pmed-0020070-b38],[@pmed-0020070-b39]\] in designing peptide-based tracers with favorable imaging properties for clinical applications. Another approach to optimize pharmacokinetics is based on the conjugation of polyethyleneglycol \[[@pmed-0020070-b40],[@pmed-0020070-b41],[@pmed-0020070-b42],[@pmed-0020070-b43],[@pmed-0020070-b44],[@pmed-0020070-b45]\]. It has been demonstrated that such polyethyleneglycolated peptides also improve pharmacokinetics and tumor retention. However, a direct comparison of tracers resulting from the different strategies has not yet been carried out.
The correlation between regional tracer uptake in the lesion and density of αvβ3-positive vessels confirms that this technique allows not only visualization but also noninvasive quantitative assessment of the integrin expression. Interestingly, our study demonstrated high both inter*-* and intraindividual variances in tracer accumulation in the different lesions, indicating a great diversity in receptor expression. This finding demonstrates the value of noninvasive techniques for appropriate selection of patients entering clinical trials of αvβ3-targeting therapies. This is further emphasized by the fact that in some cases no \[^18^F\]Galacto-RGD uptake was found, despite viable tumor cells being detected via a \[^18^F\]FDG scan.
Furthermore, PET imaging with \[^18^F\]Galacto-RGD can be applied to assess successful blocking of αvβ3 by therapeutic agents, thereby providing essential information for the dose and dose scheduling of αvβ3 antagonists. Further studies are needed to demonstrate the impact of this new technique as a novel prognostic indicator in cancer. However, the first evidence of the prognostic value is given by Gasparini et al. \[[@pmed-0020070-b46]\], who found αvβ3 expression in tumor vasculature "hot spots" to be a significant prognostic factor predictive of relapse-free survival in both node-negative and node-positive patients.
αvβ3 is also found on endothelial cells during wound healing, in restenosis, in rheumatoid arthritis, and in psoriatic plaques. Thus, radiolabeled RGD peptides may be used to characterize not only malignant tumors but also inflammatory diseases. Most recently, we demonstrated in a murine model for cutaneous delayed-type hypersensitivity reaction that \[^18^F\]Galacto-RGD allows noninvasive assessment of αvβ3 expression in inflammatory processes \[[@pmed-0020070-b47]\]. Our preliminary data from a villonodular synovitis show that αvβ3 expression on endothelial cells in this lesion can be monitored in patients. Altogether, these findings indicate that \[^18^F\]Galacto-RGD might also become a new biomarker for disease activity in inflammatory processes.
The primary advantage of PET in imaging molecular processes is its high sensitivity combined with high penetration of the gamma radiation resulting from positron decay. Thus, PET imaging allows quantification of regional radioactivity concentrations in human studies. The optical imaging approach has an even higher sensitivity, but suffers from the low penetration of light in most tissues. This results in a very limited ability to carry out tomographic imaging and to quantify the optical signal. Thus, optical imaging is currently limited to preclinical studies in mice, whereas PET can be performed in preclinical as well as in clinical studies. Magnetic resonance imaging provides high spatial resolution and can combine morphological and functional imaging, but has approximately 1,000-fold lower sensitivity compared with PET. Thus, PET is the most appropriate technique for noninvasive determination of molecular processes in patients at the current time. Obviously, the patient is exposed to high-energy γ-rays during this procedure. However, based on our radiation dose estimates, the effective radiation dose for a \[^18^F\]Galacto-RGD scan is in the same range as for a \[^18^F\]FDG scan, an approved routine examination in the clinic in many countries \[[@pmed-0020070-b48]\]. In preclinical studies, different targeted magnetic resonance contrast agents have been evaluated, using either anti-αvβ3 antibody-conjugated polymerized liposomes \[[@pmed-0020070-b49]\] or nanoparticles \[[@pmed-0020070-b50]\], or nanoparticles linked with an αvβ3 peptidomimetic antagonist \[[@pmed-0020070-b51]\]. In those studies, depending on the contrast agent and animal model used, an average magnetic resonance signal intensity enhancement between approximately 20% and 120% was found, a finding which has not yet been confirmed in clinical studies. In our patient study using \[^18^F\]Galacto-RGD and PET, a 9-fold higher activity accumulation, on average, was found in the tumor than in muscle, further indicating the currently superior properties of this radiotracer for molecular imaging. Moreover, recent developments in combining PET with computed tomography or future possibilities to combine PET with magnetic resonance imaging will allow correlation of these processes with the corresponding morphology.
To further improve tumor retention of αvβ3 radioligands, multimeric RGD peptides were recently introduced. Our group developed different series of multimeric structures with up to eight RGD units linked via different spacers \[[@pmed-0020070-b40],[@pmed-0020070-b41],[@pmed-0020070-b42]\]. These multimeric RGD peptides showed increased binding affinities in vitro and improved tumor accumulation and tumor/background ratios in rodents compared with the monomeric compounds. These data and data from other groups \[[@pmed-0020070-b52],[@pmed-0020070-b53],[@pmed-0020070-b54]\] indicate that the multimeric ligand approach may be used for optimization of the performance of peptide-based tracers. However, studies in patients will be necessary to demonstrate the potential of this approach in clinical settings.
In summary, this new class of PET tracer may offer insights into molecular processes during tumor development and dissemination in preclinical as well as clinical settings, and will be a helpful tool in planning and controlling novel αvβ3-directed therapies.
Supporting Information {#s5}
======================
Protocol S1
::: {.caption}
###### Approval of Ethics Committee
(1.4 MB PPT).
:::
::: {.caption}
######
Click here for additional data file.
:::
Protocol S2
::: {.caption}
###### Patient Consent Form
(4.2 MB PPT).
:::
::: {.caption}
######
Click here for additional data file.
:::
Patient Summary {#sb1}
---------------
### {#sb1a}
#### Background {#sb1a1}
Tumor cells express many different molecules on their surface. These cell membrane molecules are involved in a variety of different processes, such as those that hold cells together, trigger cell death, or determine whether the tumor spreads. Some of these molecules can be tagged with radiolabeled compounds, called tracers. These tracers can show where these molecules are found and how many there are by methods such as PET and SPECT scans that don\'t require a biopsy, i.e., are not invasive. These methods can then be used for planning treatment with anti-cancer drugs that bind these molecules
#### What Did the Investigators Do? {#sb1a2}
They induced tumors in mice and injected them with a tracer for one cell surface molecule---an integrin. They showed that the amount of the molecule on the tumor could be measured by the intensity of tracer seen on a PET scan. They also showed that the same molecule was present on the new blood vessels that tumors produce. In a small study of patients with various tumors, including melanomas, the researchers found that the same tracer could be used to measure the expression of the integrin on tumor cells as well as on endothelial cells, such as those found in blood vessels, and hence measure the amount of new vessels in the tumors.
#### What Does This Mean for Patients? {#sb1a3}
This tracer could be useful to determine integrin expression noninvasively, to determine how many new vessels tumors have, to get information for planning anti-cancer therapies targeting integrin, and to study response to anti-cancer drugs. However, this study involved only nine patients, so much more work will need to be done before such a technique is shown to be generally reliable.
#### Where Can I Get More Information? {#sb1a4}
The National Cancer Institute has information on melanomas for patients: <http://www.nci.nih.gov/cancertopics/pdq/treatment/melanoma/patient>
Radiology Info explains PET scanning: <http://www.radiologyinfo.org/content/petomography.htm>
We thank W. Linke, C. Bodenstein, J. Carlsen, B. Blechert, and C. Schott for their excellent technical assistance. The RDS-cyclotron and PET team, especially M. Herz, G. Dzewas, and C. Kruschke, are gratefully acknowledged. We thank D. A. Cheresh of the Scripps Institute, La Jolla, California, for providing the melanoma cell lines M21 and M21-L. This work was financially supported by the Sander Foundation (grant number 96.017.3) and by a grant from the Münchner Medizinische Wochenschrift. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
**Citation:** Haubner R, Weber WA, Beer AJ, Vabuliene E, Reim D, et al. (2005) Noninvasive visualization of the activated αvβ3 integrin in cancer patients by positron emission tomography and \[^18^F\]Galacto-RGD. PLoS Med 2(3): e70.
\[^18^F\]FDG
: \[^18^F\]fluorodeoxyglucose
PET
: positron emission tomography
p. i.
: post injection
SUV
: standardized uptake value
[^1]: **Competing Interests:** The authors have declared that no competing interests exist.
[^2]: **Author Contributions:** R. Haubner, W.A. Weber, H.J. Wester, H. Kessler, and M. Schwaiger conceived and designed the experiments. R. Haubner, A.J. Beer, E. Vabuliene, D. Reim, K.F. Becker, M. Goebel, and R. Hein performed the experiments. R. Haubner, W.A. Weber, M. Sarbia and A.J. Beer analyzed the data. R. Haubner, W.A. Weber, A.J. Beer, H.J. Wester and M. Schwaiger contributed to the writing of the paper.
[^3]: ¤a Current address: Universitätsklinik für Nuklearmedizin, Medizinische Universität Innsbruck, Austria
[^4]: ¤b Current address: Department of Molecular and Medical Pharmacology, David Geffen School of Medicine, University of California, Los Angeles, California, United States of America
|
PubMed Central
|
2024-06-05T03:55:55.021299
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069665/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e70",
"authors": [
{
"first": "Roland",
"last": "Haubner"
},
{
"first": "Wolfgang A",
"last": "Weber"
},
{
"first": "Ambros J",
"last": "Beer"
},
{
"first": "Eugenija",
"last": "Vabuliene"
},
{
"first": "Daniel",
"last": "Reim"
},
{
"first": "Mario",
"last": "Sarbia"
},
{
"first": "Karl-Friedrich",
"last": "Becker"
},
{
"first": "Michael",
"last": "Goebel"
},
{
"first": "Rüdiger",
"last": "Hein"
},
{
"first": "Hans-Jürgen",
"last": "Wester"
},
{
"first": "Horst",
"last": "Kessler"
},
{
"first": "Markus",
"last": "Schwaiger"
}
]
}
|
PMC1069666
|
Andreas Stang, Hans-Werner Hense, and Karl-Heinz Jöckel\'s Viewpoint: It Is Unethical When a Beneficial Standard Treatment Exists {#s1}
=================================================================================================================================
A better understanding of the aetiology and pathological mechanisms of diseases often results in new ideas for their treatment. It is then necessary to put these ideas to a formal empirical test in a trial setting. The randomized controlled trial (RCT) is the closest that clinical research can get to the experimental situation. In the RCT, patients are assigned at random to an intervention of putative effectiveness, with the aim of minimizing the potential for bias inherent in nonrandomized clinical research settings. The triumphal advance of RCTs is reflected in their prominent role as one of the pillars of evidence-based medicine.
Initially, when there is uncertainty about the efficacy of a new treatment, clinical researchers are advised to compare the experimental intervention with a placebo. Placebo-controlled trials serve to show that a specific treatment has a beneficial effect on defined clinical endpoints beyond that attributable to mere administration of the intervention by medical professionals. Thus, the early trials of antihypertensive medications and statins were placebo-controlled and were considered to be proof of their beneficial effects.
But what about the next phase? What happens when a treatment for a certain condition, such as hypertension, has been shown to be effective in placebo-controlled RCTs but a newer intervention has been developed for that condition? Let us assume that there is evidence from basic and early clinical trials that the new intervention has a biological effect and has no major side effects in appropriate doses. Should the researchers test it against placebo to prove the superiority of the new treatment?
It is arguably unethical to withhold a therapy of proven efficacy from any patient in a research trial just for the purpose of increasing scientific knowledge. Paragraph 29 of the Declaration of Helsinki states: "The benefits, risks, burdens and effectiveness of a new method should be tested against those of the best current prophylactic, diagnostic, and therapeutic methods" \[[@pmed-0020072-b1]\]. A note of clarification for paragraph 29 states: "The World Medical Association hereby reaffirms its position that extreme care must be taken in making use of a placebo-controlled trial and that in general this methodology should only be used in the absence of existing proven therapy" \[[@pmed-0020072-b1]\].
Rothman and Michels have argued that the declaration should include specific examples showing how placebo trials are unethical: "It might suggest as one such example that even in studies of new analgesics to study relief from pain such as headache, the new remedies should be compared only with existing analgesics, and never with placebo. The example will reinforce the point that this principle is not a blurry boundary" \[[@pmed-0020072-b2]\].
Critics of the declaration argue that forbidding placebo trials puts the manufacturers of a new treatment at a scientific and commercial disadvantage. The manufacturers of a new treatment, say the critics, have to prove that their treatment is as good as an existing one, whereas the manufacturers of the existing treatment had to pass a "lesser test" (superiority over placebo) to get their drug on the market.
For practitioners, though, the crucial question in evaluating a new treatment is how it compares with the standard available treatment, and not whether it is better than placebo. So the important issue is to decide when it is that we can call a therapy "standard"---that is, when can we speak of an indisputable benefit that would make a currently available treatment\'s use in a trial control group ethically imperative?
Clinical guidelines or recommendations based on high-quality evidence sometimes exist to support use of such a therapy. In situations where no such guidance exists, it is important to assess both the benefits of the therapy (for example, in terms of survival, and relative and absolute risk reduction) and possible harms (including side effects, impaired quality of life, and economic costs). There may be therapies that prolong survival (there is a "gross benefit") but that cannot be considered to be beneficial because the adverse effects cancel out any survival benefit (there is no "net benefit"). Such therapies cannot be considered "standard" treatment.
One framework for grading the quality of evidence and strength of recommendations on any treatment was published last year by the GRADE working group \[[@pmed-0020072-b3]\]. The framework stresses the need for judgments based on a formally structured consideration of the balance between benefits and harm, the quality of the evidence, translation of the evidence in specific clinical situations or settings, and the certainty of baseline risks, including resource utilisation.
So when is use of a placebo trial unethical? It is unethical if, in accordance with an assessment similar to the one suggested by the GRADE working group (that is, balancing gross and net benefits in a given trial through a transparent and formalised process), therapies other than the experimental one are judged to be beneficial and are available. In the many situations where such a decision is not clear-cut, the use of placebo may be considered ethically appropriate.
Erick Turner and Martin Tramèr\'s Viewpoint: It Can Be Unethical *Not* to Use Placebo {#s2}
=====================================================================================
It is generally agreed that placebo is unethical when its use is likely to result in irreversible harm, death, or other serious morbidity. A common argument against placebo is that its use is unnecessary, and therefore unethical, when "proven effective therapy" exists, in which case any new treatment should be tested against this existing treatment. The argument is that if a study drug appears to perform at least as well as a drug that has already been "proven effective", then the study drug must be effective as well.
The problem with this reasoning is that drug efficacy is not a simple all-or-none matter. If a drug with historical evidence of efficacy could be relied upon to be unfailingly effective---and placebo unfailingly ineffective---in all future clinical trials, we would readily admit that placebo is unnecessary and therefore unethical.
The reality is that "proven effective therapy"---better called "assumed effective therapy" (AET)---often fails to show superiority to placebo. This is not because these drugs are in fact ineffective, but because the trials in question lack assay sensitivity \[[@pmed-0020072-b4],[@pmed-0020072-b5]\]. Assay sensitivity is defined as the ability of a trial to distinguish an effective from an ineffective therapy.
Unfortunately, the extent of this problem is poorly appreciated because of publication bias: the tendency for studies that are positive to be published and the tendency of negative and indeterminate studies never to see the light of day \[[@pmed-0020072-b6]\]. Thus, the myth of infallible "proven therapy" is sustained. But, like a mirage, it vanishes on closer examination.
Khan et al. gained access to unpublished, as well as published, clinical trials data on antidepressants from the Food and Drug Administration (FDA) via the United States Freedom of Information Act \[[@pmed-0020072-b7]\]. They obtained the FDA review documents on 51 clinical trials on nine antidepressants approved between 1985 and 2000. Of 92 active treatment arms (all involving doses that were eventually approved), 47 (51%) failed to demonstrate statistical superiority to placebo. Of these, there were seven cases (15%) in which the placebo arm was actually superior to AET. Thus, it can be seen that the phrase "proven effective therapy" should be taken with a grain of salt.
Now, what if the FDA had not had the benefit of looking at the placebo arms and relied on an equivalence or noninferiority design \[[@pmed-0020072-b8]\] comparing study drug with AET? Khan et al. list 12 flexible-dose studies in which (now-approved) study drug outperformed AET (previously approved antidepressants) \[[@pmed-0020072-b7]\]. Many opponents of placebo would argue that each of these 12 trials provides ample evidence for efficacy of the study drug. However, because these trials did include placebo arms, we discover that in four of them (33%), neither AET nor study drug beat placebo. (In fact, in two of these four trials, AET was numerically inferior to placebo.) Therefore, in these four antidepressant trials, the two "active" drugs were not equally effective, but rather equally ineffective. This critical distinction would have been lost without placebo, and it would have been impossible to ascertain that these seemingly positive trials were in fact false positive trials.
The problem of assay sensitivity is not confined to antidepressants or even to psychotropic drugs in general. A meta-analysis by Tramèr et al found that, among 52 possible comparisons between the "proven" antiemetic ondansetron and placebo, 19 (37%) failed to show a difference \[[@pmed-0020072-b5]\]. Additionally, many drug classes have shown problems with assay sensitivity ([Box 1](#box1){ref-type="boxed-text"}). The potential for reaching erroneous conclusions by omitting placebo also exists outside of drug studies, as in the example regarding the usefulness of prophylactic respiratory physical therapy on pulmonary function after cardiac surgery \[[@pmed-0020072-b9]\]; in this specific case, the placebo would be a no intervention control.
Box 1. Drug Classes That Have Shown Problems with Assay Sensitivity {#sb1}
-------------------------------------------------------------------
AnalgesicsAntiemeticsAnxiolyticsAntihypertensivesHypnoticsAntianginal agentsAngiotensin-converting enzyme inhibitors for heart failureBeta-blockers given after myocardial infarctionAntihistaminesNonsteroidal asthma prophylaxisMotility-modifying drugs for gastroesophageal reflux disease
If we were to rely on equivalence or noninferiority designs in studying drugs for indications for which assay sensitivity cannot be assumed, we would risk approving ineffective drugs. It is conceivable that even placebo itself could be approved under such conditions.
According to the Declaration of Helsinki, "Medical research is only justified if there is a reasonable likelihood that the populations in which the research is carried out stand to benefit from the results of the research" \[[@pmed-0020072-b1]\]. To approve ineffective drugs based on flawed science and to let them loose on an unsuspecting public would be unethical. This is akin to the phenomenon of hypercorrection, in which, in trying very hard to be grammatically correct, the person ends up being grammatically incorrect \[[@pmed-0020072-b10]\]. In this case, by trying very hard to be ethical and adhering too rigidly to the anti-placebo dogma, one can end up being unethical.
In order to best serve the public health, we must ensure that our clinical drug trials yield scientifically valid results. Where assay sensitivity can be guaranteed, equivalence or noninferiority trials omitting placebo may be ethically preferable. However, where assay sensitivity cannot be guaranteed---and this problem is probably more widespread than we yet realize---difference-showing superiority studies, usually involving placebo, either as monotherapy or add-on therapy, are ethically preferable.
Stang, Hense, and Jöckel\'s Response to Turner and Tramèr\'s Viewpoint {#s3}
======================================================================
There is some common agreement between Turner and Tramèr\'s viewpoint and ours---we agree that it is unethical to use placebo when a proven effective therapy exists. However, we question their very narrow definition of "proven effective therapy". In their definition, a placebo is unethical when the proven effective therapy "could be relied upon to be unfailingly effective---and placebo unfailingly ineffective---in all future clinical trials". It is possible to argue that all empirical evidence or knowledge is temporary and uncertain. Replication carries no implication for validity \[[@pmed-0020072-b11]\]. Corroborated hypotheses---in this case, about the effectiveness of a drug---merely "survive" and the degree of corroboration depends on the number and "severity" of tests the hypothesis has survived \[[@pmed-0020072-b12]\].
We are therefore left with the difficult task of having to evaluate the current evidence of the effectiveness of available treatments in order to decide whether placebo is ethical or not. In other words, we have to make some evaluation on what constitutes "proven effective therapy" based on our current knowledge. The evaluation of current evidence cannot protect us against misinterpretations in the light of future evidence.
It appears to us that Turner and Tramèr think that superiority of a new drug to a control drug could only be established if *all* trials consistently show a statistically significant superiority of the new drug over the control. But studies that evaluate the effectiveness of a new drug may not show identical results for several reasons. Features of the study design, including sample size, dosage, patients\' inclusion and exclusion criteria, choice of active control treatment, quality of study conduct, patients\' compliance, and other factors can all have an influence on the trial results. Therefore, the proportion of trials showing statistically significant superiority (bullet counting) is an inappropriate indicator of drug superiority and a proportion less than 100% is no indicator of lack of superiority. Several analytical techniques, including meta-analysis and meta-regression, that account for design features are available and provide better insights into the superiority of drugs than bullet counting.
Turner and Tramèr\'s Response to Stang, Hense, and Jöckel\'s Viewpoint {#s4}
======================================================================
Stang and colleagues quote from part of a clarification to the Declaration of Helsinki. But the clarification continues: "a placebo-controlled trial may be ethically acceptable, even if proven therapy is available...where for compelling and scientifically sound methodological reasons its use is necessary to determine the efficacy or safety of a prophylactic, diagnostic or therapeutic method" \[[@pmed-0020072-b1]\]. Our viewpoint was essentially an evidence-based discussion of this clarification and its ethical implications. Assuming that medical research successfully rids itself of publication bias \[[@pmed-0020072-b13],[@pmed-0020072-b14]\], it should become increasingly obvious that, for many drug classes ([Box 1](#box1){ref-type="boxed-text"}), the emperor of "proven therapy" is wearing no clothes \[[@pmed-0020072-b15]\].
But this debate is not only about efficacy; it is also about harm. In the absence of a placebo group, it may be impossible to interpret a drug\'s potential for harm. Let us look at Stang and colleagues\' example of analgesics. The Vioxx Gastrointestinal Outcomes Research (VIGOR) trial showed a five-fold difference in the incidence of myocardial infarction in the rofecoxib (Vioxx) group compared with the naproxen group \[[@pmed-0020072-b16]\].
Nonsteroidal anti-inflammatory drugs such as naproxen, however, inhibit platelet function and therefore might have a myocardial protective effect \[[@pmed-0020072-b17]\]. Since the VIGOR trial did not include a placebo group, it remained unclear whether there was an increased risk of myocardial infarction with rofecoxib or a decreased risk with naproxen. Four years later, and after tens of millions of patients had received rofecoxib \[[@pmed-0020072-b18]\], Merck announced they were withdrawing the drug because of an increased cardiovascular risk \[[@pmed-0020072-b19]\]. The decision was based on the unpublished Adenomatous Polyp Prevention on Vioxx (APPROVe) study, a placebo-controlled three-year trial of rofecoxib.
In his November 2004 testimony before the United States Senate, David Graham of the FDA provided an estimate of the rate of excess cases of Vioxx-related myocardial infarction and sudden cardiac death. He testified that it was as if, for the five years that Vioxx was on the United States market, "2 to 4 jumbo jetliners were dropping from the sky every week" \[[@pmed-0020072-b20]\]. Of those cases, he added, 30% to 40% probably died.
If those who believe that "proven therapy" trials are ethically preferable to placebo-controlled trials had had their way, the APPROVe study would have been blocked, and Vioxx would still be on the market today. It seems ironic that such a stance could be taken in the name of ethics.
::: {#pmed-0020072-g001 .fig}
::: {.caption}
###### (Illustration: Margaret Shear, Public Library of Science)
:::
:::
::: {#pmed-0020072-g002 .fig}
::: {.caption}
###### (Illustration: Margaret Shear, Public Library of Science)
:::
:::
**Citation:** Stang A, Hense HW, Jöckel KH, Turner EH, Tramèr MR (2005) Is it always unethical to use a placebo in a clinical trial? PLoS Med 2(3): e72.
AET
: assumed effective therapy
APPROVe
: Adenomatous Polyp Prevention on Vioxx
FDA
: Food and Drug Administration
RCT
: randomized controlled trial
VIGOR
: Vioxx Gastrointestinal Outcomes Research
[^1]: Andreas Stang is Professor of Clinical Epidemiology at the Institute of Medical Epidemiology, Biometry, and Informatics at the Medical Faculty of the Martin-Luther-University of Halle-Wittenberg, Germany. Hans-Werner Hense is Professor of Clinical Epidemiology at the Institute of Epidemiology and Social Medicine at the Medical Faculty of the University of Münster, Germany. Karl-Heinz Jöckel is Director of the Institute of Medical Informatics, Biometry, and Epidemiology at the University Hospital of Essen, Germany. Erick H. Turner is a former clinical reviewer of psychotropic drugs at the United States Food and Drug Administration. He is currently Medical Director of the Mood Disorders Program at the Portland Veteran Affairs Medical Center, Assistant Professor of Psychiatry, and Assistant Professor of Pharmacology and Physiology at Oregon Health and Science University, Portland, Oregon, United States of America. Martin R. Tramèr is a consultant anesthetist at Geneva University Hospitals, Geneva, Switzerland.
[^2]: **Competing Interests:** AS declares that he has no competing interests. HWH received lecture fees on several occasions from various pharmaceutical companies. His research activities are marginally funded (less than 5% of total) by industrial sponsors. KHJ is involved in many clinical and epidemiological projects, several of them sponsored by the pharmaceutical industry. EHT is on the speaker\'s bureaus of Eli Lilly, AstraZeneca, and Bristol-Myers Squibb. He has provided outside consulting to Bristol-Myers Squibb, Eli Lilly, GlaxoSmithKline, and Sepracor. He has also received funding for clinical drug trials, which can be spent only for research purposes and which has no effect on his income, from Abbott Laboratories, AstraZeneca, Bristol-Myers Squibb, and DOV Pharmaceuticals. MRT has been a scientific consultant to Pfizer, Merck, Janssen-Cilag, and Sintetica. He has also received lecture fees from various pharmaceutical companies.
|
PubMed Central
|
2024-06-05T03:55:55.024490
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069666/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e72",
"authors": [
{
"first": "Andreas",
"last": "Stang"
},
{
"first": "Hans-Werner",
"last": "Hense"
},
{
"first": "Karl-Heinz",
"last": "Jöckel"
},
{
"first": "Erick H",
"last": "Turner"
},
{
"first": "Martin R",
"last": "Tramèr"
}
]
}
|
PMC1069667
|
Introduction {#s1}
============
Numerous genetic and environmental factors have been associated with uric acid \[[@pmed-0020076-b01]\], and serum uric acid values are markedly elevated in patients with gout ([Table 1](#pmed-0020076-t001){ref-type="table"}). Since at least fifty years ago, modestly higher serum uric acid concentrations have been reported in patients with coronary heart disease (CHD) than in controls \[[@pmed-0020076-b02]\], and there have been suggestions that measurement of serum uric acid can enhance the prediction of CHD \[[@pmed-0020076-b03]\]. Prospective epidemiological studies have, however, reported apparently conflicting findings, with several studies reporting positive associations only among women \[[@pmed-0020076-b04],[@pmed-0020076-b05]\], The interpretation of the data has been further complicated by the correlation of serum uric acid concentrations with several established coronary risk factors (such as blood pressure), with the use of cardiovascular medications (such as diuretics), and with clinical conditions associated with CHD (such as chronic renal disease \[[@pmed-0020076-b06]\]). It has been difficult, therefore, to determine whether serum uric acid values are predictive of CHD, and, if so, whether any such associations are independent from established risk factors or from the effects of disease or both.
::: {#pmed-0020076-t001 .table-wrap}
Table 1
::: {.caption}
###### Characteristics of Uric Acid
:::
^a^ Values are mean (SD)
^b^ Conversion to SI units: 1 μmol/l = 59.48 mg/dl
^c^ Approximate correlation between two measurements taken some years apart in the same individuals
:::
To help address these uncertainties, we report a prospective study with more CHD cases than any previous report on serum uric acid, involving 2,459 incident cases of nonfatal myocardial infarction (MI) and CHD death, and 3,969 controls from within a prospective observational study of about 19,000 middle-aged Icelanders without a previous history of MI. To help put these results in context, we also report a meta-analysis of 15 previously published prospective studies of serum uric acid, involving a total of an additional 7,002 incident CHD cases and an additional 151,122 controls, including supplementary information obtained by correspondence from investigators to help assess in more detail the impact of possible confounders. The present analyses have been restricted to prospective cohorts sampled from essentially general populations (i.e., excluding cohorts selected on the basis of existing vascular or other diseases, or on the basis of having risk factors for vascular disease, such as high blood pressure) to reduce any distorting effects of preexisting disease on serum uric acid levels.
Methods {#s2}
=======
The Reykjavik Study {#s2a}
-------------------
The Reykjavik Study, initiated in 1967 as a prospective study of cardiovascular disease, has been described in detail previously \[[@pmed-0020076-b07]\]. All men born between 1907 and 1934 and all women born between 1908 and 1935 who were resident in Reykjavik, Iceland, and its adjacent communities on 01 December 1966 were identified in the national population register and then invited to participate in the Reykjavik Study during five stages of recruitment between 1967 and 1991, yielding 8,888 male and 9,681 female participants without a history of MI (72% response rate). Nurses administered questionnaires, made physical measurements, recorded an electrocardiogram, performed spirometry, and collected fasting venous blood samples, which were stored at −20 °C for subsequent analysis. All participants have been monitored subsequently for all-cause mortality and for cardiovascular morbidity, with a loss to follow-up of less than 1% to date. A total of 2,459 men and women with available serum samples had major coronary events between the beginning of follow-up and 31 December 1995, yielding mean durations of follow-up among CHD cases of 17.5 (standard deviation \[SD\] 8.7) years and, among controls, of 20.6 (SD, 8.2) years. In total, 1,073 CHD deaths and 701 nonfatal MIs were recorded among men (including 564 confirmed MIs and 137 possible MIs), and 385 CHD deaths and 300 nonfatal MIs among women (including 237 confirmed MIs and 63 possible MIs). Deaths from coronary heart disease were ascertained from central registers on the basis of a death certificate with International Classification of Diseases codes 410--414, and the diagnosis of nonfatal MI was based on MONICA criteria. We selected 3,969 controls that were "frequency-matched" to cases on calendar year of recruitment, sex, and age in 5-y bands from among participants who had survived to the end of the study period without a MI. The National Bioethics Committee and the Data Protection Authority of Iceland approved the study protocol, and participants provided informed consent.
Laboratory Methods {#s2b}
------------------
Serum uric acid levels were measured with a Technicon autoanalyzer \[[@pmed-0020076-b08]\]. The measurement of other biochemical analytes has been described previously \[[@pmed-0020076-b07]\]. Baseline measurements of serum uric acid were available on 2,456 out of 2,459 CHD cases and 3,962 out of 3,969 controls. To assess the within-person consistency of serum uric acid levels over time, measurements were made in pairs of samples collected at an interval of about 12 y apart in 379 individuals in the present study.
Statistical Methods and Meta-Analysis {#s2c}
-------------------------------------
Case-control comparisons were made by unmatched stratified logistic regression fitted by unconditional maximum likelihood. Analysis of serum uric acid values was previously specified to be by sex-specific thirds of values in the controls (with subsidiary analyses involving other cut-off values). Adjustment was made for age, sex, smoking status (never, former, current), daily cigarette consumption, blood pressure, body mass index, fasting concentrations of total cholesterol and triglycerides, and various markers of socioeconomic status related to occupation, education, home ownership, and type of accommodation. We assessed variation in the strength of association according to pre-specified sub-groups, using likelihood ratio tests for interaction after adjusting for these factors, with 99% confidence intervals (CIs) used in these exploratory analyses.
For the meta-analysis, studies of serum uric acid and CHD published before May 2003 with greater than a year\'s follow-up conducted in essentially general populations (i.e., excluding cohorts defined on the basis of preexisting cardiovascular or other diseases) were sought by computer-based searches, scanning the reference lists of all relevant studies and review articles, hand-searching of relevant journals, and correspondence with authors of studies. Computer searches using Medline, PubMed, Web of Science, and Embase databases used keywords relating to uric acid in combination with CHD (e.g., coronary heart disease, ischemic heart disease, vascular disease, MI, and atherosclerosis). Relevant endpoints included nonfatal MI (generally using World Health Organization criteria) and CHD death (generally using International Classification of Disease criteria).
The following factors were abstracted from each study: numbers of cases and controls, mean age of cases and percentage of males, mean duration of follow-up, assay type, and those used for adjustment in multivariable assessments. Five studies were excluded because they reported insufficient data or only unadjusted risk ratios \[[@pmed-0020076-b09],[@pmed-0020076-b10],[@pmed-0020076-b11],[@pmed-0020076-b12],[@pmed-0020076-b13]\], but these involved only a total of about 590 CHD cases (or \< 6% of the total number of cases in the present report). Of 16 studies (including four studies that had not previously reported in relation to CHD \[[@pmed-0020076-b14],[@pmed-0020076-b15],[@pmed-0020076-b16],[@pmed-0020076-b17]\]), 11 provided supplementary tabular data on sex-specific "relative risks" (i.e., incidence rate ratios according to sex-specific thirds of serum uric acid distribution in controls) and details of factors adjusted for in multivariable analyses. We excluded female-specific estimates based on fewer than 30 CHD cases, owing to very small sample sizes from two studies \[[@pmed-0020076-b16],[@pmed-0020076-b18]\]. Where data were not available by thirds of serum uric acid levels, the log-relative risk (and its standard error) was estimated from the reported relative risks using log-linear scaling and assuming normality of the uric acid distribution, as described previously \[[@pmed-0020076-b19]\]. Where data were available only in separate age strata, a single pooled estimate was used. Fixed-effect summary measures were calculated as the inverse-variance weighted average of the log-relative risks. Heterogeneity was assessed by the heterogeneity Q statistic and by random effect regression models with restricted maximum likelihood estimation. Subsidiary analyses (conducted only on studies known to exclude individuals with existing CHD) grouped studies by sex, study size, geographical location, sampling framework (population- or workforce-based), degree of adjustment for other cardiovascular risk factors, type of assay, and duration of follow-up. Statistical analyses were conducted using Stata version 7.0. To make some allowance for multiple comparisons, 99% CI were used for individual studies, and 95% CI were reserved for the combined estimates.
Results {#s3}
=======
The Reykjavik Study {#s3a}
-------------------
The mean age at CHD event among cases was 70.2 (SD, 9.7) y. There were highly significant differences between cases and controls with respect to established vascular risk factors such as smoking, body mass index, blood pressure, and serum lipid concentrations ([Table 2](#pmed-0020076-t002){ref-type="table"}). Serum uric acid values were highly significantly associated with male sex, nonmanual occupation, body mass index, diastolic blood pressure, triglycerides, and serum creatinine (*p* \< 0.0001 for each), although most of these associations weakened after adjustment for other vascular risk factors ([Table S1](#st001){ref-type="supplementary-material"}). In 379 participants who provided paired blood samples, on average about 12 y apart, the within-individual correlation coefficient among serum uric acid values was 0.60 (CI, 0.54--0.66), similar to the decade-to-decade consistency observed in values of systolic blood pressure \[0.66 (CI, 0.60--0.72)\] and total serum cholesterol \[0.60 (CI, 0.54--0.66)\] in these participants.
::: {#pmed-0020076-t002 .table-wrap}
Table 2
::: {.caption}
###### Baseline Characteristics of Cases and Controls in the Reykjavik Study
:::
Values are mean (SD) unless indicated otherwise
^a^ Information on occupation was available for only 1,742 cases and 2,888 controls, respectively
^b^ Information on education was available for only 1,292 cases and 2,157 controls, respectively
^c^ Information on home ownership was available for 2,323 cases and 3,754 controls, respectively
^d^ Information on type of residence was available for 2,258 cases and 3,646 controls, respectively. Other categories included "duplex" and "villa."
^e^ Information on serum uric acid was available for 2,456 cases and 3,962 controls, respectively
^f^ Value log transformed for analysis and presented as geometric mean (SD)
:::
The odds ratio for CHD was 1.39 (CI, 1.20--1.61; Wald test statistic, χ^2^ ~1~ = 18.4) in males in the top third compared with those in the bottom third of baseline serum uric acid levels (tertile cut-offs, \> 339 versus \< 286 μmol/l \[[Table 3](#pmed-0020076-t003){ref-type="table"}\]), and this fell to 1.12 (CI, 0.94--1.33; χ^2^ ~1~ = 1.5) after adjustment for smoking, other established risk factors, and indicators of socioeconomic status ([Table 3](#pmed-0020076-t003){ref-type="table"}). The odds ratio for CHD was 1.42 (CI, 1.13--1.79; χ^2^ ~1~ = 9.1) in females in the top third compared with those in the bottom third of baseline serum uric acid levels (tertile cut-offs, \> 280 v \< 232 μmol/l), and this fell to 1.12 (CI, 0.85--1.46; χ^2^ ~1~ = 0.6) after adjustment for smoking, other established risk factors, and indicators of socioeconomic status. In a combined analysis of males and females, the odds ratio for CHD was 1.39 (CI, 1.23--1.58; χ^2^ ~1~ = 27.2) and this fell to 1.12 (CI, 0.97--1.30; χ^2^ ~1~ = 2.4) after adjustment. In analyses restricted to the 2,083 cases without evidence of CHD at baseline, the adjusted odds ratios fell further to 1.08 (CI, 0.90--1.31) in males and 1.00 (CI, 0.75--1.33) in females ([Table 3](#pmed-0020076-t003){ref-type="table"}), but the findings were materially unchanged in analyses excluding the 200 CHD cases with "possible" MI or in analyses varying cut-off levels (e.g., by quarters, fifths, or increases of 1 SD; see [Table 3](#pmed-0020076-t003){ref-type="table"} legend). [Figure 1](#pmed-0020076-g001){ref-type="fig"} indicates that there was no substantial variation in the strength of association between serum uric acid and CHD at different levels of established risk factors, and, in particular, there was no good evidence of interactions with sex or systolic blood pressure (sex, χ^2^ ~1~ = 0.03, *p* = 0.86; smoking, χ^2^ ~1~ = 0.28, *p* = 0.60; body mass index, χ^2^ ~2~ = 1.13, *p* = 0.57; total cholesterol, χ^2^ ~2~ = 2.42, *p* = 0.30; systolic blood pressure, χ^2^ ~2~ = 4.63, *p* = 0.10).
::: {#pmed-0020076-g001 .fig}
Figure 1
::: {.caption}
###### Associations between Serum Uric Acid and CHD in 2,456 cases and 3,962 Controls in the Reykjavik Study at Different Levels of Established Risk Factors
Squares indicate odds ratios, with the size of the square proportional to the effective sample size.
:::
:::
::: {#pmed-0020076-t003 .table-wrap}
Table 3
::: {.caption}
###### Relative Odds of Coronary Heart Disease in Individuals Who Had Serum Uric Acid in the Top Third of the Sex-Specific Distribution of Controls Relative to Those Who Had Values in the Bottom Third of This Distribution in the Reykjavik Study
:::
^a^ Systolic blood pressure, total cholesterol, triglycerides, body mass index, smoking (former or current, including number cigs per day), FEV1, history of diabetes
^b^ Odds ratios (males and females combined) using alternative comparisons were: 1.24 (0.99--1.55) top fifth vs. bottom fifth; 1.22 (1.03--1.45) top quarter vs. bottom quarter; 1.08 (1.02--1.15) per standard deviation increase. Sex-specific odds ratios using thirds of the overall (not sex-specific) distribution of serum uric acid were: in males, 1.19 (0.98--1.43); in females 1.34 (0.98--1.82)
:::
Meta-Analysis {#s3b}
-------------
In aggregate, 16 prospective reports \[[@pmed-0020076-b06],[@pmed-0020076-b14],[@pmed-0020076-b15],[@pmed-0020076-b16],[@pmed-0020076-b17],[@pmed-0020076-b18],[@pmed-0020076-b20],[@pmed-0020076-b21],[@pmed-0020076-b22],[@pmed-0020076-b23],[@pmed-0020076-b24],[@pmed-0020076-b25],[@pmed-0020076-b26],[@pmed-0020076-b27],[@pmed-0020076-b28]\] on serum uric acid (including the present study) have involved a total of 9,458 CHD cases and 155,084 controls, with a weighted mean age at entry of 50 y and weighted mean follow-up of 10.5 y ([Table 4](#pmed-0020076-t004){ref-type="table"}). Studies were conducted in the USA \[[@pmed-0020076-b22],[@pmed-0020076-b23],[@pmed-0020076-b24],[@pmed-0020076-b25],[@pmed-0020076-b28]\], Western Europe \[[@pmed-0020076-b06],[@pmed-0020076-b14],[@pmed-0020076-b15],[@pmed-0020076-b17],[@pmed-0020076-b18],[@pmed-0020076-b20],[@pmed-0020076-b27]\], Israel \[[@pmed-0020076-b21]\], and Japan \[[@pmed-0020076-b16],[@pmed-0020076-b26]\], and all reported adjustment for at least smoking and some other established risk factors. Overall, in a comparison of individuals with serum uric acid values in the top third with those in the bottom third of the population, the relative risk for CHD was 1.13 (CI, 1.07--1.20: [Figure 2](#pmed-0020076-g002){ref-type="fig"}), with statistically compatible results in male and females (χ^2^ ~1~ = 1.1; *p* = 0.3). In a subsidiary analysis of seven studies \[[@pmed-0020076-b20],[@pmed-0020076-b22],[@pmed-0020076-b23],[@pmed-0020076-b24],[@pmed-0020076-b25],[@pmed-0020076-b27]\], involving 6,357 CHD cases and 65,978 controls, all of which excluded individuals with known cardiovascular disease at the baseline examination, the relative risk for CHD was 1.10 (CI, 1.03--1.18). There was significant heterogeneity among the 23 sex-specific study estimates (χ^2^ ~2~ = 38.1, *p* = 0.02), but only some of this was explained by study characteristics such as sample size (χ^2^ ~2~ = 11.1), geographical location (χ^2^ ~2~ = 1.0), sampling framework (χ^2^ ~1~ = 0.5), degree of adjustment for possible confounders (χ^2^ ~2~ = 10.0), duration of follow-up (χ^2^ ~1~ = 0.1), and assay type (χ^2^ ~3~ = 4.2) ([Figure 3](#pmed-0020076-g003){ref-type="fig"}). In a random-effects model, that takes additional account of study variation and the joint impact of these characteristics, only degree of adjustment for possible confounders remained a significant source of heterogeneity at the 1% level of significance (sex, *p* = 0.36; sample size, *p* = 0.41; geographical location, *p* = 0.71; sampling framework, *p* = 0.02; degree of adjustment for possible confounders, *p* = 0.001; duration of follow-up, *p* = 0.67; and assay type, *p* = 0.06).
::: {#pmed-0020076-g002 .fig}
Figure 2
::: {.caption}
###### Meta-Analysis of Prospective Observational Studies of Serum Uric Acid and CHD in Essentially General Populations, Subdivided by Sex
Conventions are the same as in [Figure 1](#pmed-0020076-g001){ref-type="fig"}. Combined odds ratios and their CIs are indicated by unshaded diamonds for subtotals and shaded diamonds for grand totals. +, adjustment reported only for age and sex; ++, adjustment for these plus smoking; +++, adjustment for these plus some additional established risk factors; ++++, adjustment for these plus existing cardiovascular disease. Study abbreviations: ARIC, Atherosclerosis Risk in Communities; BIRNH, Belgium Interuniversity Research on Nutrition and Health; BRHS, British Regional Heart Study; CHA, Chicago Heart Association Detection Project in Industry; GRIPS, Göttingen Risk Incidence and Prevalence Study; IIHDS, Israeli Ischemic Heart Disease Study; MONICA, World Health Organization Monitoring Trends and Determinants in Cardiovascular Disease; NHANES, National Health and Nutrition Examination Survey; NHEFS, NHANES I Epidemiologic Follow-Up Study; PROCAM, Prospective Cardiovascular Munster Study.
:::
:::
::: {#pmed-0020076-g003 .fig}
Figure 3
::: {.caption}
###### Prospective Studies of the Association of Serum Uric Acid and CHD, Grouped by Various Characteristics
Conventions are the same as in [Figure 1](#pmed-0020076-g001){ref-type="fig"}. \*, each sex-specific estimate was treated as a "study"; †, two studies (6 and 13) were drawn from general practice registers; §, risk factors adjusted for included: smoking, blood pressure, total cholesterol, triglycerides, alcohol consumption, obesity, use of cardiovascular medication, history of hypertension, and history of diabetes. PTA, phosphotungstic acid.
:::
:::
::: {#pmed-0020076-t004 .table-wrap}
Table 4
::: {.caption}
###### Prospective Studies of Serum Uric Acid and Coronary Heart Disease in Essentially General Populations: Study Characteristics
:::
^a^ Sampling method: Random, a randomly selected subset of eligible persons was invited to participate; complete, all eligible persons in the study population were invited to participate
^b^ Only men were included in analyses, due to a small numbers of female cases: BIRNH, 26 women; Osaka, 4 women; MONICA Augsberg, number of female cases not stated
^c^ Duration of follow-up was "at least 10 years."
Table abbreviations: NS, not specified; Mn, mean; GP, general practice; SD, standard deviation. Study abbreviations: IIHDS, Israeli Ischemic Heart Disease Study; BRHS, British Regional Heart Study; NHANES, National Health and Nutrition Examination Survey; NHEFS, NHANES I Epidemiological Follow-up Study; PROCAM, Prospective Cardiovascular Munster Study; ARIC, Atherosclerosis Risk in Communities; GRIPS, Göttingen Risk Incidence and Prevalence Study; CHA, Chicago Heart Association Detection Project in Industry; BIRNH, Belgium Interuniversity Research on Nutrition and Health; MONICA, World Health Organization Monitoring Trends and Determinants in Cardiovascular Disease
:::
Discussion {#s4}
==========
The present report provides prospective evidence from the largest study so far of serum uric acid and CHD---plus a meta-analysis of 15 previous relevant studies---involving a total of more than 9,000 incident cases and more than 150,000 controls. The overall findings suggest that individuals with baseline serum uric acid values in the top third of the population have about a 10% greater risk of CHD over the subsequent decade than those in the bottom third (with the likelihood that this association would be about twice as strong if based on long-term usual levels of serum uric acid). It is likely, however, that this modest association has been exaggerated by the preferential publication of striking findings in smaller studies ("publication bias"), or by residual confounding by established risk factors, or both. For example, the observation of weaker associations in studies with more comprehensive adjustment for possible confounders lessens the likelihood that any association between serum uric acid and CHD is independent from possible confounders; the odds ratio was only 1.02 (CI, 0.91--1.14), which is not significant, in the eight studies with the most complete reported adjustment for possible confounders ([Figure 3](#pmed-0020076-g003){ref-type="fig"}). The present data also provide no good evidence to support previous claims that the association between serum uric acid and CHD is stronger in females than in males \[[@pmed-0020076-b05]\], or stronger at higher levels of established risk factors, such as in individuals with higher blood pressure recordings \[[@pmed-0020076-b29]\].
The main implication of these data is to refute suggestions made throughout the past half-century that measurement of serum uric acid can importantly enhance the prediction of CHD in general populations. These data do not directly address the question of whether or not serum uric acid may be involved in the causation of CHD through a number of potentially relevant vascular effects (such as through the formation of free radicals or through the oxidation of low-density-lipoprotein cholesterol \[[@pmed-0020076-b01],[@pmed-0020076-b30]\]), but they suggest that serum uric acid levels are unlikely to be a major determinant of CHD.
Supporting Information {#s5}
======================
Table S1
::: {.caption}
###### Comparison of Baseline Values of Risk Factors and Other Characteristics in Controls in the Reykjavik Study by Thirds of Serum Uric Acid Concentration
(67 KB DOC).
:::
::: {.caption}
######
Click here for additional data file.
:::
Patient Summary {#sb1}
---------------
### {#sb1a}
#### Background {#sb1a1}
Defining which risk factors are important for disease is useful for clinicians and patients not only because the presence of risk factors allows the prediction of who is more likely to get a disease, but also because they provide some insight into the underlying causes of disease. One such suspected risk factor for coronary heart disease is the level of uric acid in the blood. The debate over whether uric acid is useful for predicting heart disease has been going on for over fifty years. Most evidence for risk factors comes from studies of populations, in which it can be hard to tease out the effects of many different factors; often studies come to different conclusions. One way of finding out which results are reliable is to pool the results of many studies.
#### What Did the Researchers Find? {#sb1a2}
They looked at the uric acid levels of around 2,500 people with coronary heart disease and almost 4,000 controls measured at the start of a large study in Iceland, and then investigated whether there was a relation between levels of uric acid and development of heart disease. After adjusting for all the other factors that could affect the chance of heart disease, they found that uric acid did not predict heart disease. They then combined these results with those from 15 other studies, and confirmed the findings.
#### What Do These Findings Mean? {#sb1a3}
After fifty years, it now seems clear that measurement of uric acid does not help to predict heart disease. It may still be involved in triggering heart disease, but any effect must be subtle.
#### Where Can I Get More Information? {#sb1a4}
The National Heart Lung and Blood Institute has many pages of information on heart disease: <http://www.nhlbi.nih.gov/health/public/heart/index.htm#ami>
This work was supported by a program grant from the British Heart Foundation and by the Raymond and Beverly Sackler Award in the Medical Sciences. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. We are grateful to Adam Dickinson for assistance with correspondence to investigators. The following investigators kindly provided additional data from their studies: Riitta Antikainen, Dirk de Bacquer, Calle Bengtsson, Lori Boland, Christopher Bulpitt, Gerry Fowkes, Uri Goldbourt, Bo Hedblad, Sandy Irving, Ulrich Keil, Amanda Lee, Angela Liese, Jack Medalie, Dorothea Nagal, Shinichi Sato, Helmet Schulte, and Masako Tomita.
**Citation:** Wheeler JG, Juzwishin KDM, Eiriksdottir G, Gudnason V, Danesh J (2005) Serum uric acid and coronary heart disease in 9,458 incident cases and 155,084 controls: Prospective study and meta-analysis. PLoS Med 2(3): e76.
CHD
: coronary heart disease
CI
: confidence interval
MI
: myocardial infarction
SD
: standard deviation
[^1]: **Competing Interests:** The authors have declared that no competing interests exist.
JD is a member of the editorial board of *PLOS Medicine*.
[^2]: **Author Contributions:** JGW, KDMJ, GE, VG, and JD designed the study. JGW, KDMJ, GE, VG, and JD analyzed the data. JGW conducted statistical analyses. JGW, KDMJ, GE, VG, and JD contributed to writing the paper.
|
PubMed Central
|
2024-06-05T03:55:55.026705
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069667/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e76",
"authors": [
{
"first": "Jeremy G",
"last": "Wheeler"
},
{
"first": "Kelsey D. M",
"last": "Juzwishin"
},
{
"first": "Gudny",
"last": "Eiriksdottir"
},
{
"first": "Vilmundur",
"last": "Gudnason"
},
{
"first": "John",
"last": "Danesh"
}
]
}
|
PMC1069668
|
Stuge et al. report a detailed analysis of the fine specificity of CD8+ T cells against tumor-associated antigen in melanoma patients \[[@pmed-0020077-b1]\]. They compared peptide-vaccination-driven with naturally arising T cell responses against the HLA-A\*0201 restricted melanoma peptide antigens M26 (derived from Melan-A/MART-1) and G209-2M (derived from gp100 protein). A major endpoint of this study was in vitro tumor cell recognition by T cells. Fortunately, this is increasingly used as a "golden" standard in the assessment of tumor-specific T cells. The authors suggest that spontaneously arising antigen-specific T cell populations are qualitatively different from those induced by vaccination with heteroclitic peptides (which are altered for increased HLA binding): tumor cell recognition was found in nearly all T cells from the former, but only in a minority from the latter. As reported previously, these results correlated with recognition efficiency of antigenic peptides. We agree that this has considerable implications for immunotherapy and congratulate the authors for analyzing T cell recognition in great detail. However, in one point our own studies lead to different results: we repetitively found that the majority of T cells generated with the heteroclitic Melan-A M26 peptide were tumor reactive. This was the case for Melan-A-specific T cell populations generated in HLA-A\*0201 transgenic mice \[[@pmed-0020077-b2]\], in vitro \[[@pmed-0020077-b3]\], and in melanoma patients \[[@pmed-0020077-b4]\]. The latter studies also assessed T cells from vaccination-site sentinel lymph nodes, containing T cells that are very likely selected and activated by vaccination and not by the tumor.
The authors point out correctly that tetramer+ T cells comprise many cells unable to recognize and kill tumor cells in an antigen-specific manner, presumably owing to low T cell receptor avidity to cognate antigen. An extreme case is naïve T cell populations, of which the majority are unable to recognize tumor cells, despite their specific binding to MHC/ peptide tetramers \[[@pmed-0020077-b5]\]. Therefore, it is crucial to exclude naïve T cells from studies analyzing tumor recognition. HLA-A\*0201+ humans (healthy individuals and melanoma patients) have 0.07% ± 0.05% naïve Melan-A tetramer+ cells within peripheral blood CD8+ T cells \[[@pmed-0020077-b5],[@pmed-0020077-b6]\]. The three patients studied by Stuge et al. had 0.23%, 0.12%, and 0.50% Melan-A tetramer+ cells. Thus, one can estimate that the studied populations from the three patients contained approximately 30%, 60%, and 15% naïve Melan-A-specific T cells, respectively. This is only a rough estimate---tetramer analysis before vaccination and assessment of CD45RA/ CCR7 expression would give more insight. Nevertheless, it remains likely that the first two patients had considerably more naïve cells than the third patient (i.e., the one without immunotherapy). In addition, naïve-derived CD8+ T cells have a higher clonogenic potential than activated Melan-A-specific T cells from melanoma patients (unpublished data). This means that overrepresentation of clones derived from naïve CD8+ T cells is likely to occur when both naïve and activated antigen-specific CD8+ T cells co-exist in a given lymphocyte population. As mentioned, Stuge et al. found unexpected high frequencies of T cell clones not recognizing tumor cells in the two vaccinated patients. It is conceivable that this was due to the presumably high percentages of naïve Melan-A-specific cells present in the populations used for generating the clones, which would provide an explanation for the discrepancy with the results of our studies \[[@pmed-0020077-b2],[@pmed-0020077-b3],[@pmed-0020077-b4]\].
Ethical considerations limit vaccination studies in healthy humans. In patients, candidate antigens should therefore be tested with strong adjuvants \[[@pmed-0020077-b7]\], to increase the likelihood that the studied responses are predominantly vaccination-driven, with only minor contribution of spontaneous T cell activation \[[@pmed-0020077-b8]\]. It would be desirable to directly compare vaccination with heteroclitic peptide versus vaccination with natural peptide. However, this is hampered by the lack of ex vivo detectable responses to native peptides owing to their low immunogenicity. Another option is to analyze clonal distributions (T cell receptors) of responding T cells extensively: Further support for the notion that spontaneous (tumor driven) responses have increased potential for tumor recognition would be obtained if mono/oligoclonal T cell repertoires are indeed significantly more often found in spontaneous than vaccination-induced responses.
We certainly agree that vaccines must be optimized. Thus, more such studies are desirable, since they have high potential to lead to better understanding of the differences between clinically irrelevant and relevant T cell responses, and to rapidly identify the most promising vaccine formulations that can subsequently be tested in large-scale clinical trials.
**Citation:** Speiser DE, Cerottini JC, Romero P (2005) Tumor cell recognition efficiency by T cells. PLoS Med 2(3): e77.
[^1]: **Competing Interests:** The authors have declared that no competing interests exist.
|
PubMed Central
|
2024-06-05T03:55:55.028823
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069668/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e77",
"authors": [
{
"first": "Daniel E",
"last": "Speiser"
},
{
"first": "Jean-Charles",
"last": "Cerottini"
},
{
"first": "Pedro",
"last": "Romero"
}
]
}
|
PMC1069669
|
Introduction {#s1}
============
The central role of T cells in the pathogenesis of allergic disease is well established \[[@pmed-0020078-b01]\]. Through production of interleukin (IL)-4, IL-5, and IL-13, allergen-specific T helper (Th) 2 cells direct IgE synthesis, eosinophil growth/differentiation, and induction of airway hyperreactivity \[[@pmed-0020078-b02],[@pmed-0020078-b03]\].Until recently, it was assumed that the basis for allergic disease was an imbalanced Th cell response to certain allergens, manifest as a predominance of Th2 cytokines over Th1 cytokines. However, immune suppression may also be a normal consequence of a protective immune response, serving to limit excessive responses that lead to immunopathology \[[@pmed-0020078-b04]\]. The role of regulatory T cell (T~reg~) populations in maintaining homeostasis is increasingly well understood. The term T~reg~ is used to describe a variety of T cell functional phenotypes that display common features. Several studies have described the dependence of T~reg~ function on cell--cell contact. In certain cases regulation was demonstrated to be dependent on IL-10 and/or transforming growth factor β secretion \[[@pmed-0020078-b05],[@pmed-0020078-b06],[@pmed-0020078-b07],[@pmed-0020078-b08],[@pmed-0020078-b09]\].
Regulation of immune responses may be attributable to both naturally occurring (thymus-derived, or "natural") regulatory cells and also naïve or effector T cells that have acquired suppressive activity (adaptive regulatory cells) \[[@pmed-0020078-b10],[@pmed-0020078-b11]\]. Therapeutic administration of short, soluble peptide sequences, in the absence of inflammatory signals, may result in presentation by immature or quiescent antigen-presenting cells (APCs). Immature allogeneic human dendritic cells (DCs) induced non-proliferating, IL-10-producing CD4^+^ T cells with regulatory properties \[[@pmed-0020078-b12]\], while peptide-specific human T~reg~ were induced following administration of antigen-pulsed immature DCs in vivo \[[@pmed-0020078-b13],[@pmed-0020078-b14]\]. DCs producing IL-10 were able to suppress airway inflammation in a murine model of asthma \[[@pmed-0020078-b15]\]. Thus, partially immature or "steady state" DCs, circulating in the lymphatics, may interact with T cells in a tolerogenic milieu, in the absence of concomitant pro-inflammatory stimuli such as pattern recognition receptor triggering \[[@pmed-0020078-b16]\].
An additional mechanism for limiting immune responses may be reducing sensitivity to cognate signals. Up-regulation of CD5, a suppressor of T cell signalling \[[@pmed-0020078-b17]\], has been associated with regulatory cells arising as a consequence of competition for space and resources \[[@pmed-0020078-b18]\]. Under such conditions, suppression was shown to lack antigen specificity and to be mediated by cells that did not exhibit any of the hallmarks of "professional" T~reg~. Recently, Hawiger and colleagues delivered antigen to steady-state DCs via the DEC-205 molecule. Following cognate interaction with these cells, antigen-specific T cells were unresponsive and expressed enhanced levels of CD5 \[[@pmed-0020078-b19]\]. Chronic low-level antigen exposure in the periphery has also been shown to result in anergy in CD8^+^ cells that was associated with increased expression of CD5, further illustrating a role for CD5 in regulation of T cell function \[[@pmed-0020078-b20]\].
In animal models, the administration of low-dose peptide is a well-established mechanism for the induction of T~reg~ that may arise as a result of presentation by steady-state DCs and "non-professional" APCs \[[@pmed-0020078-b21],[@pmed-0020078-b22],[@pmed-0020078-b23],[@pmed-0020078-b24]\]. Similarly, administration of soluble peptides to allergic asthmatic individuals has been shown to result in markedly reduced cutaneous reactions to allergen injection \[[@pmed-0020078-b25],[@pmed-0020078-b26],[@pmed-0020078-b27]\], reduced airway hyperreactivity \[[@pmed-0020078-b27]\], and improvements in symptom scores after nasal allergen challenge \[[@pmed-0020078-b28]\]. Changes in clinical reactivity were associated with decreased Th1 and Th2 cytokines and increased IL-10 production \[[@pmed-0020078-b25],[@pmed-0020078-b26]\]. In the current study, we address the hypothesis that low-dose peptide therapy in allergic individuals results in antigen-specific hyporesponsiveness associated with the induction of a suppressive population of CD4^+^ T cells, together with up-regulation of surface CD5 levels on antigen-specific T cells.
Methods {#s2}
=======
Patients and Study Design {#s2a}
-------------------------
Individuals who were cat-allergic and asthmatic were recruited, diagnosed, and assessed as described in detail elsewhere \[[@pmed-0020078-b29]\]. The study received prior approval from the Ethics Committee of the Royal Brompton and Harefield Hospitals National Health Service Trust (London, United Kingdom). Written, witnessed informed consent was obtained from all patients. Peripheral blood mononuclear cells (PBMCs) were obtained from patients enrolled in two consecutive studies (open study design) of immunotherapy employing short synthetic peptides derived from the sequence of the major cat allergen Felis domesticus allergen 1 (Fel d 1). The studies employed different dosing regimes in order to evaluate dose effects on clinical and mechanistic outcomes. The first study included eight patients (referred to hereafter as Group 1) who received incremental doses of Fel d 1 peptides (0.1, 1, 1, 5, 10, and 25 μg) totalling 42.1 μg of each peptide, while the second study comprised 12 patients (referred to hereafter as Group 2) who received a total of 291 μg (1, 5, 10, 25, 50, 100, and 100 μg) of each peptide. Peptides were synthesised by Fmoc chemistry, purified by HPLC, and presented as lyophilised solids (Advanced Biotechnology Centre, Imperial College London, United Kingdom). Peptides were reconstituted with sterile physiological saline and dispensed into sterile vials for single patient use (Nova Laboratories, Leicestershire, United Kingdom). Peptide sequences were as follows: EICPAVKRDVDLFLTGT, LFLTGTPDEYVEQVAQY, EQVAQYKALPVVLENA, KALPVVLENARILKNCV, RILKNCVDAKMTEEDKE, KMTEEDKENALSLLDK, KENALSLLDKIYTSPL, LTKVNATEPERTAMKK, TAMKKIQDCYVENGLI, SRVLDGLVMTTISSSK, ISSSKDCMGEAVQNTV, and AVQNTVEDLKLNTLGR.
Clinical parameters and outcome measures associated with peptide intervention in donors from whom PBMC samples were obtained are described in detail elsewhere \[[@pmed-0020078-b27],[@pmed-0020078-b28]\]. Briefly, in the first study peptide immunotherapy (PIT) resulted in improved non-specific bronchial hyperreactivity, since a significantly (*p =* 0.02) greater concentration of histamine was required to induce a 20% reduction in forced expiratory volume measured in 1 s. Additionally, a significant reduction (*p =* 0.03) in the magnitude (area in square millimeters) of the late-phase skin reaction was observed post-treatment. In the second study, treatment was associated with a reduction in the magnitude of the late asthmatic reaction induced by inhaled allergen challenge, together with a significant decrease in nasal outcome measurements (number of sneezes, nasal blockage, and weight of nasal secretion; all measurements at 15 min post-challenge, *p =* 0.02).
PBMC Cultures {#s2b}
-------------
PBMCs were isolated from venous blood by density gradient centrifugation (Histopaque-1077; Sigma Chemicals, Poole, United Kingdom) and cryopreserved. All experiments were performed on pre- and post-PIT PBMCs of the same patient in single experiments, to reduce inter-experiment variation within single patients. Prior to in vitro culture, PBMCs were thawed, washed, and labelled with carboxyfluorescein diacetate succinimidyl ester (CFSE) (Molecular Probes, Eugene, Oregon, United States), as follows: 2.5 × 10^6^ each of pre-PIT and post-PIT PBMCs were resuspended in 0.5 ml of RPMI-1640 (Invitrogen, Paisley, United Kingdom), and 0.5 ml of 1 μM CFSE added under constant, gentle agitation, to achieve a final CFSE concentration of 0.5 μM. After 10 min, 1 ml of human AB serum (Sigma, Poole, United Kingdom) was added to terminate labelling, and cells were washed twice. Cells were resuspended at 2.5 × 10^6^ cells/ml of complete medium (RPMI-1640 supplemented with [L]{.smallcaps}-glutamine and 5% human AB serum) and plated at 5 × 10^5^ cells/well in 96-well flat-bottom culture plates (Nunc, Merck Eurolab, Lutterworth, United Kingdom) in 200 μl of final volume, under the following culture conditions: unstimulated, stimulated with 30 μg/ml whole cat allergen (generous gift of Leti Laboratories, Madrid, Spain) and stimulated with plate-bound α-CD3/α-CD28 (10/1 μg/ml; BD Pharmingen, Cowley, United Kingdom).
For suppression experiments, PBMCs were separated into CD4^+^ and CD4^neg^ populations. Limited quantities of peripheral blood were available from study patients. Therefore, for reasons of economy, CD4-depleted PBMCs (CD4^neg^) remaining after selection of CD4^+^ cells were used as target cells in all suppression assays. For each patient, 20 × 10^6^ each of pre- and post-PIT PBMCs were labelled with αCD4 magnetic beads (MACS; Miltenyi, Bisley, United Kingdom) and positively sorted to a mean purity of 94%. CD4^neg^ pre- and post-PIT cells were labelled with CFSE as described above, while CD4^+^ pre- and post-PIT T cells were labelled with PKH-26 (Sigma) as follows: cells were resuspended in diluent C (Sigma) at no more than 10^7^ cells/ml, and an equal volume of a PKH-26 dilution (1 μM) was added to reach a final concentration of 0.5 μM. After 2 min, the reaction was stopped with the addition of 1 ml of human serum, and cells were washed twice. Cells were cultured in the following combinations: pre- and post-PIT CD4^neg^ cells alone, pre-PIT CD4^neg^ plus pre- or post-PIT CD4^+^, and post-PIT CD4^neg^ plus pre- or post-PIT CD4^+^ (CD4^neg^ cells at 0.5 × 10^6^ cells/well and CD4^+^ cells at 0.125 × 10^6^ cells/well in 96-well flat-bottom tissue culture plates, to achieve a ratio of 4:1). All were cultured in the absence or presence of cat allergen (30 μg/ml) for 1 wk in a humidified incubator at 37 °C gassed with 5% CO~2~ in air.
Flow Cytometry {#s2c}
--------------
To determine changes in the proliferation of T cell subpopulations associated with PIT, cells were recovered after 1 wk of culture, washed twice, and stained for 30 min at 4 °C with a combination of αCD4-PE + αCD8-Cy, or αCD45-PE. Isotype controls used were mouse IgG~2a~-PE, mouse IgG~1~-PE, and mouse IgG~1~-Cy. Mean fluorescence intensity (MFI) was determined by FACS (FACScan, BD Pharmingen) of at least 2 × 10^4^ events within the lymphocyte gate. In suppression experiments, the extent of proliferation was measured as above on the CFSE-labelled read-out population without additional antibody staining. Percentages of CD4^+^CD25^+^ T cells, CD4^+^CD5^+^ cells, or CD8^+^CD5^+^ cells were measured for unstimulated cells by double staining with αCD4-Cy + αCD25-FITC, αCD4-Cy + αCD5-PE, or αCD8-Cy + αCD5-PE. Isotype controls used were mouse IgG~1~-Cy and mouse IgG~1~-FITC (all antibodies were from BD Pharmingen).
Cytokine Measurements {#s2d}
---------------------
Culture supernatants of 100 μl were collected from wells 48 h after the start of culture. Cytokines were measured by cytometric bead array Th1/Th2 kit (BD Pharmingen) according to the manufacturer\'s instructions. A total of six cytokines were measured simultaneously. Data for IL-5, IL-10, and interferon (IFN)-γ are shown. Cytokine concentrations were determined using cytometric bead array analysis software (BD Pharmingen). The sensitivity of the assays was 2.4 pg/ml for IL-5, 2.8 pg/ml for IL-10, and 7.1 pg/ml for IFN-γ.
Data Analysis {#s2e}
-------------
FACS cell surface data and CFSE--PKH-26 mixing experiment proliferation data were acquired with Cellquest (BD Pharmingen), and events within the live lymphocyte gate were interpreted using Winmdi 2.8 software (Scripps Research Institute, <http://facs.scripps.edu/software.html>). CFSE proliferation data of T cell subsets were acquired with Cellquest, and events within the CD4^+^, CD8^+^, or CD45RO^+^ gate analysed with the Proliferation Wizard module in ModFit LT software (Verity Software House, Topsham, Massachusetts, United States). Percentage proliferation is defined as the fraction of the starting population that has proliferated during the course of the experiment.
Statistical Analysis {#s2f}
--------------------
For statistical analysis data were analysed for normality using the Shapiro-Wilks test. Normally distributed data were analysed using the paired t-test (parametric). Non-normal data were analysed using the Wilcoxon signed rank test (non-parametric). Analysis was performed by an independent statistician (Turnstat, Reading, United Kingdom).
Results {#s3}
=======
PIT Results in the Inhibition of Cat-Allergen-Induced Proliferation of CD45RO^+^, CD4^+^, and CD8^+^ T cell Subsets {#s3a}
-------------------------------------------------------------------------------------------------------------------
The effect of PIT on cellular proliferation of T cell subsets was evaluated by combining CFSE labelling with cell surface staining. As shown in [Figure 1](#pmed-0020078-g001){ref-type="fig"}A, the majority of cat-allergen-specific T cells resided within the CD45RO^+^ (memory) T cell population. The proliferation of this population was markedly inhibited following PIT ([Figure 1](#pmed-0020078-g001){ref-type="fig"}A--[1](#pmed-0020078-g001){ref-type="fig"}D). Limited allergen-specific proliferation was detected in the CD45RO^−^ (naïve) population, but this appeared less sensitive to the effects of PIT. Data from all nine individuals tested showed a similar reduction in the post-PIT proliferative response ([Figure 1](#pmed-0020078-g001){ref-type="fig"}E; mean proliferation pre-PIT \[20.3%\] was decreased post-PIT \[5.8%\], *p =* 0.004; pre-PIT range 7.9%--41.8%; post-PIT range 0%--16.7%). Proliferation in the absence of a stimulus was less than 2% in all cases and was subtracted. Proliferation to plate-bound α-CD3/α-CD28 (10 μg/ml and 1 μg/ml, respectively, as a mixture) resulted in mean pre-PIT proliferation of CD45RO^+^ T cells of 64.2% and post-PIT proliferation of 60.1% (data not shown).
::: {#pmed-0020078-g001 .fig}
Figure 1
::: {.caption}
###### PIT Reduces Antigen-Specific Proliferation of Memory T Cells
(A--D) PBMCs taken before and after PIT were labelled with CFSE to track cell division after antigen stimulaton. Proliferation of cat-allergen-specific CD45RO^+^ lymphocytes was reduced following PIT (A) and (B). (C) and (D) represent CD45RO^+^ T cells as shown in panels (A) and (B), respectively, analysed with ModFit software. The right-hand peaks represent the parental population, and generations of dividing cells are depicted leftwards along the x-axis.
\(E) Summary of the percentage of proliferating CD45RO^+^ T cells pre- and post-PIT (percent proliferating cells is defined as the fraction of the starting population that has proliferated during the course of the experiment, determined with Modfit) for all nine patients tested. Open symbols represent patients enrolled in treatment Group 1, while solid symbols depict patients from treatment Group 2. Horizontal solid bars show mean levels of proliferation. Background proliferation (in the absence of a stimulus) was less than 2% and was subtracted. The Wilcoxon signed rank test was used for statistical analysis.
:::
:::
The effect of PIT on CD4^+^ and CD8^+^ populations was also addressed. Both CD4^+^ and CD8^+^ subsets proliferated to whole cat allergen. CD4^+^ post-PIT T cell proliferation was significantly reduced (*p =* 0.016; [Figure 2](#pmed-0020078-g002){ref-type="fig"}A), despite a slight increase in proliferation for one patient (mean pre-PIT to post-PIT CD4 proliferation was reduced from 5.4% to 2.1% \[pre-PIT range 0%--12.7%; post-PIT range 0%--7.3%\]). A similar reduction was observed for CD8^+^ T cells (*p =* 0.031; [Figure 2](#pmed-0020078-g002){ref-type="fig"}B data from seven patients available for analysis). Post-PIT CD8^+^ proliferation showed a greater reduction (mean pre-PIT to post-PIT CD8^+^ proliferation was reduced from 8.0% to 2.5% \[pre-PIT range 1.8%--20.8%; post-PIT range 0.3%--4.7%\]).
::: {#pmed-0020078-g002 .fig}
Figure 2
::: {.caption}
###### PIT Reduces Antigen-Specific Proliferation of CD4^+^ and CD8^+^ T Cells
CD4^+^ and CD8^+^ proliferation data were obtained and interpreted as for [Figure 1](#pmed-0020078-g001){ref-type="fig"}. (A) and (B) represent percentage proliferation of PBMCs to cat allergen for each patient as determined with ModFit. Open symbols represent patients from treatment Group 1, while solid symbols depict patients from treatment Group 2. Horizontal solid bars indicate means. Background proliferation has been subtracted. The Wilcoxon signed rank test was used for statistical analysis.
:::
:::
Modulation of Cytokine Secretion following Peptide Immunotherapy {#s3b}
----------------------------------------------------------------
In order to characterise modulation of cytokine responses following PIT, culture supernatants were collected after 48 h. Cytokines were measured simultaneously by flow cytometry. The majority of patients displayed increased IL-10 secretion although this change did not achieve statistical significance. IL-5 secretion was significantly reduced post-PIT (*p =* 0.02; [Table 1](#pmed-0020078-t001){ref-type="table"}). IFN-γ secretion tended to be reduced following PIT, but heterogeneity was observed.
::: {#pmed-0020078-t001 .table-wrap}
Table 1
::: {.caption}
###### Modulation of Cytokine Secretion Profiles in Allergen-Stimulated PBMCs following PIT
:::
^a^ Cytokine concentration in picograms per millilitre with background (cells cultured in medium alone) subtracted
^b^ Patients from Group 1
^c^ND indicates not detected, and assigned a value of zero for statistical analysis. The sensitivity of the assays was 2.4 pg/ml for IL-5, 2.8 pg/ml for IL-10, and 7.1 pg/ml for IFN-γ
^d^ IFN-γ and IL-10 analysed with paired t-test (normal distribution), IL-5 analysed by Wilcoxon (non-normal distribution). Normality determined by Shapiro-Wilks test
:::
PIT Leads to the Induction of a CD4^+^ T Cell Population with Suppressor Activity {#s3c}
---------------------------------------------------------------------------------
To identify populations of T cells with suppressive activity and to attempt to distinguish between active suppression and clonal deletion as potential mechanisms following PIT, pre- and post-PIT CD4^+^ T cells were isolated and their effect on proliferation of the CD4^neg^ fraction measured ([Figure 3](#pmed-0020078-g003){ref-type="fig"}). Two distinct cell-cycle tracking dyes, CFSE and PKH-26, were employed to visually separate the target (CD4^neg^; CFSE) from the effector (CD4^+^; PKH-26) populations, by flow cytometry. PIT resulted in a 69% reduction (14.7% proliferation pre-PIT versus 4.6% proliferation post-PIT) in proliferation of the CD4^neg^ T cell population for the one representative patient shown in detail ([Figure 3](#pmed-0020078-g003){ref-type="fig"}A and [3](#pmed-0020078-g003){ref-type="fig"}B). When pre- or post-PIT CD4^+^ T cells were added to CD4^neg^ PBMCs (at a ratio of 1:4), a marked reduction in proliferation of cat-allergen-specific CD4^neg^ pre-PIT T cells was observed when co-cultured with post-PIT ([Figure 3](#pmed-0020078-g003){ref-type="fig"}E; 7.9% proliferation) but not with pre-PIT CD4^+^ T cells ([Figure 3](#pmed-0020078-g003){ref-type="fig"}C; 17.7% proliferation), indicating that the post-PIT CD4^+^ T cells harboured a suppressor population. As post-PIT CD4^neg^ T cell proliferation was minimal, addition of post-PIT CD4^+^ T cells did not have a further suppressive effect on this cell population ([Figure 3](#pmed-0020078-g003){ref-type="fig"}F). Additionally, removal of the CD4^+^ T cells from post-PIT PBMCs did not cause the depleted PBMC population to proliferate ([Figure 3](#pmed-0020078-g003){ref-type="fig"}B), suggesting that antigen-specific cells in the post-treatment population had already been rendered anergic in vivo as a result of PIT, or possessed the ability to actively suppress responses themselves. Similar experiments were performed in a further four patients. A summary of the results for all five patients is shown in [Figure 3](#pmed-0020078-g003){ref-type="fig"}G. Inhibition of proliferation by CD4^+^ post-PIT T cells ranged from 64.0% to 19.6%, with a mean of 47.5%.
::: {#pmed-0020078-g003 .fig}
Figure 3
::: {.caption}
###### CD4^+^ Cells Isolated after PIT Suppress the Proliferative Response of Baseline CD4^neg^ Cells
PBMCs taken before and after PIT were separated into CD4^+^ and CD4^neg^ populations by immunomagnetic separation. CD4^neg^ cells were labelled with CFSE and served as target cells. CD4^+^ cells were labelled with PKH-26 and were evaluated for suppressor/regulator function by co-culture with CD4^neg^ cells. (A) and (B) show antigen-stimulated proliferation of CD4^neg^ target cells before and after PIT. Proliferation of CD4^neg^ target cells was reduced after PIT (B). In (C) and (E), pre-PIT CD4^neg^ cells were employed as target cells. The addition of post-PIT (E), but not pre-PIT (C) CD4^+^ cells inhibited proliferation. In (D) and (F), post-PIT CD4^neg^ cells were employed as target cells. Addition of either pre-PIT (D) or post-PIT (F) CD4^+^ cells had no further effect on proliferation. Proliferation in the absence of a stimulus was less than 2% in all experiments. Representative data for one patient are shown. Data for an additional four patients were obtained using the same protocol. A data summary of percentage proliferation of pre-PIT CD4^neg^ PBMCs in the presence of pre-PIT or post-PIT CD4^+^ T cells for five patients from treatment Group 2 is shown in (G). The paired t-test was used for statistical analysis.
:::
:::
Phenotypic Characterisation of Candidate Regulatory T Cells Induced Post-PIT {#s3d}
----------------------------------------------------------------------------
T cell surface markers known to be associated with tolerance induction, such as CD25 and CD5, were compared on pre- and post-PIT resting PBMC in an attempt to provide further insight into the nature of the suppressor population. No significant variation was found in CD4^+^CD25^+^ cell numbers (mean pre-PIT to post-PIT proliferation 20.5%--17.9%; data not shown). However, when CD5 expression was determined on both CD4^+^ and CD8^+^ cells, a significant increase in MFI in both populations was observed (*p =* 0.016 and 0.047, respectively). [Figure 4](#pmed-0020078-g004){ref-type="fig"}A and [4](#pmed-0020078-g004){ref-type="fig"}B show increases in CD5 expression on CD4^+^ and CD8^+^ cells (MFI of CD5 expression on CD4^+^ cells: pre-PIT mean, 465.5 \[range, 290.3--908.5\]; post-PIT mean, 559.7 \[range, 302.5--1241.8\]; range of post-PIT percentage change in MFI, 4%--37%; MFI of CD5 expression on CD8^+^ cells: pre-PIT mean, 110.9 \[range, 55.3--345.8\]; post-PIT mean, 149.2 \[range, 60.6--352.2\]; range of post-PIT percentage change in MFI, 7%--118%). Increased MFI resulted not only from a decrease in the numbers of CD5^low^ cells and an associated increase in CD5^+^ cells in both populations, but from an increase in CD5 expression on the CD5^+^ cells as well, as is shown in [Figure 4](#pmed-0020078-g004){ref-type="fig"}C and [4](#pmed-0020078-g004){ref-type="fig"}D for one representative patient.
::: {#pmed-0020078-g004 .fig}
Figure 4
::: {.caption}
###### PIT Enhances CD5 Expression on Resting CD4^+^ and CD8^+^ PBMCs
(A and B) Box-and-whiskers plots representing changes in MFI of CD5 expression levels on unstimulated CD4^+^ (A) and CD8^+^ (B) pre- and post-PIT PBMCs from seven patients in treatment Group 2. Isotype control MFI values have been subtracted.
(C and D) CD5 levels on pre-PIT (heavy black line) and post-PIT (grey, filled) CD4^+^ and CD8^+^ PBMCs of one representative patient. M1 marks the CD5^low^ population, with a pre- to post-PIT decrease in CD5^low^CD4^+^ PBMCs from 6.6% (in black) to 1.5% (in grey), and a decrease in CD5^low^CD8^+^ PBMCs from 32.3% to 13.9%. M2 indicates the concomitant increases in CD5^+^CD4^+^ and CD5^+^CD8^+^ cells post-PIT. Changes in MFI values for the total pre- and post-PIT CD4^+^ or CD8^+^ populations of the single representative patient are shown in the upper right-hand corner of each histogram. The Wilcoxon signed rank test was used for statistical analysis.
:::
:::
The effect of PIT on proliferation, cytokine secretion patterns, phenotype of T cell subsets, and suppressive capacity did not appear to be dependent on the total dose of peptide administered in the two treatment groups.
Discussion {#s4}
==========
Following PIT, proliferation of CD4^+^, CD8^+^, and CD45RO^+^ memory T cells was reduced following culture with whole cat dander allergen extract. Non-specific T cell receptor (TCR) ligation with anti-CD3/CD28 was unaffected, implying that only cat-allergen-specific T cells had been targeted by PIT. The reduction in proliferation was primarily observed within the differentiated memory (CD45RO^+^) rather than the naïve T cell population, the latter displaying minimal cell division.
While the role of CD4^+^ T cells in the pathogenesis of allergic disease is well established, that of CD8^+^ T cells is less well defined. A number of reports suggest that CD8^+^ T cells may be activated in the asthma process. CD8^+^ cells from both bronchoalveolar lavage fluid and peripheral blood from atopic donors were found to produce IL-4 and IL-5 in lavage samples and bronchial biopsies \[[@pmed-0020078-b30],[@pmed-0020078-b31]\]. Moreover, individuals with severe atopic disease have high frequencies of Dermatophagoides pteronyssinus 1--specific CD8^+^ T cells that secrete significantly more IL-4, IL-5, and IL-13 than non-atopic individuals \[[@pmed-0020078-b32]\]. Here we have shown that CD8^+^ T cells proliferate markedly to cat allergen in vitro even in the absence of CD4^+^ T cells. In the context of previous studies, it appears likely that these cells may contribute to disease pathogenesis. Thus, induction of non-responsiveness in CD8^+^ T cells should have a positive therapeutic outcome in allergic disease.
Cytokine profiles of cat-allergen-stimulated PBMCs were established following peptide therapy. Levels of IL-2 and IL-4 were generally below the limit of detection of the assays employed. PIT had no effect on secretion of tumour necrosis factor α (data not shown). Production of IL-5, a cytokine considered particularly relevant in asthma, was significantly reduced following PIT. In approximately half of the patients there were reductions in both Th1 and Th2 cytokines, as previously described \[[@pmed-0020078-b26]\]. We have observed similar results in an unpublished study of PIT for bee venom hypersensitivity. The majority of patients showed increased IL-10 production after PIT, in agreement with our earlier observations. However, in the present study this did not achieve statistical significance, in contrast to a previous report. Enhanced production of IL-10 has been associated with protection from allergic symptoms in both naturally exposed individuals such as beekeepers and in individuals receiving bee venom immunotherapy \[[@pmed-0020078-b33]\]. In contrast, IL-10 production in relation to cat allergen exposure and protection is less well established. A protective effect of high-dose natural exposure to cat allergens (resulting in a "modified Th2 response") has been reported \[[@pmed-0020078-b34],[@pmed-0020078-b35]\]. Woodfolk and colleagues demonstrated elevated IL-10 production in individuals displaying a "modified Th2" profile in which cat-allergen-specific IgG4 appeared to protect from disease \[[@pmed-0020078-b36]\]. In their study, particular regions of the Fel d 1 molecule (carboxy terminus of chain 2) appeared to be associated with presentation by HLA-DRB1\*0701 and were associated with preferential IL-10 induction. For technical reasons, peptides from this region were not included in the preparation used in the present study. Inclusion of such peptides in future studies may enhance vaccine efficacy.
In the present study, cytokine production was evaluated in peripheral blood cells. Cytokine production at local tissues targeted by allergens may provide a more accurate picture of the effects of immunotherapy with peptides or native allergens/allergen extracts. For example, in a related study a significant increase in the number of cutaneous CD4^+^IFNγ ^+^ cells (*p =* 0.03), but not in CD4^+^IL-10^+^ cells, was observed in allergen-challenged skin biopsies \[[@pmed-0020078-b27]\]. Similarly, a significant increase in the number of IFN-γ mRNA(+) cells (*p* = 0.03) was found in nasal biopsies of patients enrolled in a whole-grass-pollen immunotherapy trial, in the absence of significant modifications in IFN-γ secretion by corresponding in vitro stimulated PBMCs \[[@pmed-0020078-b37]\]. Thus, in vivo localization of cells by allergen challenge may reveal patterns of immunomodulation that differ from changes in the blood of the same individual. For this reason, caution should be exercised when interpreting alterations in cytokine profiles in different tissues following immunotherapy.
We addressed the possibility that changes in T cell proliferation and cytokine secretion may be related to the induction of a population of T~reg~ or suppressor T cells, similar to that observed following peptide intervention in murine models \[[@pmed-0020078-b38]\]. PBMCs were separated into CD4^+^ and CD4^neg^ populations. CD4^+^ cells isolated from post-PIT blood were able to actively suppress the proliferation of pre-treatment CD4^neg^ cells. The selection of CD4^neg^ cells, rather than CD4^+^ cells, as targets was due to limitations in the number of cells available. Nevertheless, the results obtained indicate the induction of regulatory and suppressor CD4^+^ T cells following PIT. Interestingly, removal of CD4^+^ cells from the post-PIT PBMC population did not lead to a reversal of the allergen-specific hyporesponsiveness in the pre-PIT CD4^neg^ population. This observation suggests that in addition to active suppression by CD4^+^ cells, enduring effects of therapy can also be detected. Explanations for such observations may include the following: (i) clonal deletion of some antigen-specific CD4^neg^ cells, (ii) the induction of anergy in these cells during the treatment phase, or (iii) the presence of a CD4^neg^ suppressor population. In support of the last possibility, regulatory CD8^+^ T cells have recently been described \[[@pmed-0020078-b39]\]. Studies identifying allergen-specific CD4 and CD8 T cells will be required to address such issues. In future studies it will be of interest to identify the subpopulation or subpopulations of CD4 and CD8 cells responsible for the suppressive effect, by removing candidate T cells from the pre- and post-treatment CD4^+^ T cell populations prior to co-culture.
No increase in numbers of CD4^+^CD25^+^ cells was observed in PBMCs following PIT, in contrast to studies of whole allergen immunotherapy \[[@pmed-0020078-b40],[@pmed-0020078-b41]\]. In fact, numbers of CD25^bright^ cells significantly decreased following peptide therapy (data not shown). Furthermore, CD4^+^CD25^+^ T cells obtained before and after treatment in a related PIT study did not differ in their ability to suppress allergen-specific effector T cell proliferation and IL-13 production, arguing against a major role for this type of regulatory cell in peptide therapy \[[@pmed-0020078-b42]\]. We speculate that PIT results in T cell activation in the absence of inflammatory signals, possibly via presentation by immature APCs, or even by neighbouring T cells. Well-characterised in vitro human models have demonstrated that it is indeed possible to induce T cell anergy following incubation with cognate peptide in the absence of professional APCs \[[@pmed-0020078-b43],[@pmed-0020078-b44]\]. Recently, Apostolou and von Boehmer reported induction of antigen-specific hyporesponsiveness, mediated by regulatory cells, following continuous, low-dose peptide administration in mice \[[@pmed-0020078-b21]\], an observation that supports our current and previous clinical findings. Additionally, Prakken and colleagues have demonstrated induction of IL-10-secreting T~reg~ following oral peptide therapy in patients with rheumatoid arthritis. These cells may similarly represent an induced population of adaptive T~reg~ \[[@pmed-0020078-b45]\].
While CD25 expression is considered to be a marker of a functionally distinct population of T~reg~ (provided the cells have not been recently activated), CD5 expression levels on T cells may be an indicator of a regulatory function \[[@pmed-0020078-b18]\]. CD5 has been shown to be a negative regulator of TCR signalling, influencing the fate of developing thymocytes \[[@pmed-0020078-b17]\]. In the periphery, CD5^neg^ T cells show enhanced proliferation to TCR triggering \[[@pmed-0020078-b46]\]. Conversely, increased membrane levels of CD5 correlate with a lowering of the T cell response to antigen by targeting downstream signalling events \[[@pmed-0020078-b47]\].
In the current study, CD5 levels were significantly elevated on directly ex vivo, unstimulated CD4^+^ and CD8^+^ T cells, following peptide therapy. The increases were slight, which likely relates to the low precursor frequency of the cells targeted. Interestingly, the increases observed on CD8^+^ T cells were partly due to a reduction in the size of the CD8^+^CD5^neg^ T cell population. A distinct CD8^+^CD5^neg^ T cell population that accounts for 3%--10% of the total CD8^+^ T cell population in healthy donors has previously been described \[[@pmed-0020078-b48]\] and appears to be the main producer of lymphotactin (XCL-1) \[[@pmed-0020078-b49]\]. As the average size of CD8^+^CD5^neg^ T cell populations in the allergic asthmatic patients in our study is substantially larger (23.2% of the total CD8^+^ T cells), it is tempting to speculate that this is further evidence for a dysregulated immune response associated with allergic disease. This observation is in agreement with data from lymphopenic mice that developed wasting disease with accelerated kinetics following adoptive transfer of T cells expressing low levels of CD5, whilst CD5^hi^ cells were protective \[[@pmed-0020078-b18]\]. Consistent with these findings, surface levels of CD5 on human T cells also appear to correlate with immune function, as the CD5^neg^ population was increased in bone marrow transplant recipients as well as in patients with advanced AIDS \[[@pmed-0020078-b50],[@pmed-0020078-b51]\]. However, as relatively little is known about the role of CD5 in human T cell tolerance, further investigations are required to establish the relevance of our finding in allergic disease. Isolating Fel d 1--specific T cells should yield valuable information on the functional relevance of increased CD5 expression on allergen-specific cells.
To our knowledge, this is the first demonstration that PIT induces a CD4^+^ T cell population that actively suppresses antigen-induced proliferation of effector T cells. The use of dual labelling with distinctly coloured dyes allowed evaluation of the effect of the CD4^+^ T cell subset on the proliferation of CD4^neg^ (including CD8^+^ cells, natural killer cells, B cells, monocytes, and basophils) using flow cytometry. While dual labelling has been widely used to track cell migration in animal models \[[@pmed-0020078-b52]\], its application in in vitro human T cell proliferation experiments has, to our knowledge, not previously been reported. Single-colour labelling of distinct human PBMC populations has been used to characterise, isolate, and clone peanut-allergen-specific T cells \[[@pmed-0020078-b53]\] and to determine precursor frequencies of recall-antigen-specific T cells \[[@pmed-0020078-b54]\]. Measurement of proliferation by means of CFSE has the additional advantages of requiring relatively low numbers of cells and allowing additional phenotypic (cell surface markers) or functional parameters (intracellular cytokine secretion) to be studied in parallel, in distinct subpopulations \[[@pmed-0020078-b54]\].
In conclusion, our data indicate that low-dose PIT targets both CD4^+^ and CD8^+^ memory T cells and induces a population of active suppressor/regulatory T cells within the CD4^+^ compartment. Suppressor activity may also reside within the CD4^neg^ compartment. Peptide therapy resulted in a heterogeneous modulation of allergen-specific PBMC cytokine responses in vitro, generally characterised by IL-10 induction and IL-5 suppression. Finally, modest but consistent increases were observed in surface CD5 expression on both CD4^+^ and CD8^+^ T cells, an observation that may be linked to the induction of antigen-specific hyporesponsiveness. The ability to modulate antigen-specific T cell function in vivo has important implications for the treatment and prevention of allergic, autoimmune, and allograft-related diseases.
Supporting Information {#s5}
======================
**Accession Numbers**
The SwissProt (<http://www.ebi.ac.uk/swissprot/>) accession numbers for the gene products discussed in this paper are CD5 (P06127), DEC-205 (Q60767), Fel d 1 chain 1 (P30438), Fel d 1 chain 2 (P30440), and lymphotactin (P47992).
Patient Summary {#sb1}
---------------
### {#sb1a}
#### Background {#sb1a1}
Increasing numbers of people are developing allergies to pets and becoming asthmatic. It is not clear what combination of events triggers allergy---for example, whether keeping pets as a child is protective---nor what can be done to treat the allergy once it develops.
#### What Did the Authors Do? {#sb1a2}
They looked at a small group of people who were allergic to cats and asthmatic. They measured the levels of different kinds of T cells in their blood---cells that are associated with the allergy. They then treated the people with small proteins that are very similar to the triggers for the allergy and looked to see how the levels of various T cells changed. They found that the protein treatment triggered a particular type of cell, which seemed able to repress the reactive cells that had triggered the immune reaction previously.
#### What Do These Results Mean for Patients? {#sb1a3}
There are many things that interact to produce allergy, and this study does not help in understanding exactly how this happens. It does suggest a way that treatment with specific small proteins might work in reducing the allergy; however, the results will need to be confirmed in much larger studies.
#### Where Can I Get More Information? {#sb1a4}
Both the American Academy of Allergy Asthma and Immunology, and Asthma UK have large sections of patient information: <http://www.aaaai.org/patients.stm>; <http://www.asthma.org.uk/>
These studies were funded by Asthma UK, the Medical Research Council (United Kingdom), the British Medical Association, and the Royal Brompton and Harefield Hospitals Trust Clinical Research Committee. ML is an Asthma UK Senior Research Fellow. The authors would like to thank Prof. N. A. Mitchison, Prof. B. Askonas, and Dr. D. S. Robinson for constructive review of the manuscript. The authors are grateful to Dr. E. Fernandez-Caldas and Leti Laboratories for the generous gift of cat dander allergen extract and to Mrs J. Turner for statistical analysis. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
**Citation:** Verhoef A, Alexander C, Kay AB, Larché M (2005) T cell epitope immunotherapy induces a CD4^+^ T cell population with regulatory activity. PLoS Med 2(3): e78.
APC
: antigen-presenting cell
CFSE
: carboxyfluorescein diacetate succinimidyl ester
DC
: dendritic cell
Fel d 1
: Felis domesticus allergen 1
IFN
: interferon
IL
: interleukin
MFI
: mean fluorescence intensity
PBMC
: peripheral blood mononuclear cell
PIT
: peptide immunotherapy
TCR
: T cell receptor
Th
: T helper
T~reg~
: regulatory T cell(s)
[^1]: **Competing Interests:** ML and ABK have current research funding from Powderject Pharmaceuticals (currently owned by Chiron Vaccines). Powderject Pharmaceuticals was, until its acquisition by Chiron, developing peptide vaccines, based on those described herein, for commercial purposes. ML is a named inventor on three families of patent applications relating to peptide immunotherapy and peptide immunotherapeutics; ABK is a named inventor on two of these families of patent applications. ML and ABK were formerly paid consultants to Powderject Pharmaceuticals and former stockholders. ML and ABK were formerly stockholders in Circassia, a company that they founded.
[^2]: **Author Contributions:** AV, ABK, and ML designed the study. AV, CA, ABK, and ML performed the experiments. AV, ABK, and ML contributed to writing the paper.
|
PubMed Central
|
2024-06-05T03:55:55.029643
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069669/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e78",
"authors": [
{
"first": "Adrienne",
"last": "Verhoef"
},
{
"first": "Clare",
"last": "Alexander"
},
{
"first": "A. Barry",
"last": "Kay"
},
{
"first": "Mark",
"last": "Larché"
}
]
}
|
PMC1069670
|
After reading the excellent research article by Pedrini et al. \[[@pmed-0020081-b1]\] and the associated synopsis \[[@pmed-0020081-b2]\], one may conclude that the only pathway of statins\' effect on Alzheimer disease (AD) is the regulation of amyloid precursor protein (APP) processing and amyloid-ß protein (Aß) generation. The moderation is provided in the research article\'s patient summary, reminding that "statins are likely to influence the risk for Alzheimer disease by several different pathways." What are these other pathways? It is essential to note that in addition to APP processing and Aß chemistry being modulated by statins, fine tuning of cholesterol homeostasis also affects cholinergic function, ionotropic and metabotropic receptors, tau phosphorylation, neural oxidative stress reactions, and other features of neurodegeneration (reviewed in \[[@pmed-0020081-b3]\]). Moreover, precise regulation of neural cholesterol dynamics and supply is itself essential for synapse function, plasticity, and behaviour \[[@pmed-0020081-b3]\]. These data suggest that in addition to its role in sporadic AD, cholesterol homeostasis break is the unifying primary cause of neuromuscular diseases, Niemann-Pick type C disease, and Down syndrome, and explains why rare cases of familial AD (associated with mutations in APP and presenilin genes) are translated into Alzheimer\'s via membrane cholesterol sensitivity of APP processing by secretases and Aß generation. Also important, is the synopsis\'s \[[@pmed-0020081-b2]\] apparently outdated dividing of APP processing into "harmful" (Aß-generating) and "healthy" (non-amyloidogenic). One should be cautious in calling Aß a harmful molecule. This is because several recent studies have illuminated an essential function for amyloidogenic processing of APP and Aß in neurons \[[@pmed-0020081-b4]\] and synapses \[[@pmed-0020081-b5]\]. In this context, the reciprocal effect of Aß on cholesterol synthesis, cellular uptake, efflux, and esterification, and its relation to the experimental restoration of long-term potentiation (LTP, a synaptic plasticity measure) may represent one of the poorly comprehended physiological functions of Aß \[[@pmed-0020081-b6],[@pmed-0020081-b7]\].
**Citation:** Koudinov AR, Berezov TT (2005) Cholesterol, statins, and Alzheimer disease. PLoS Med 2(3): e81.
[^1]: **Competing Interests:** ARK serves as founding and managing editor of *Neurobiology of Lipids* (ISSN 1683-5506), an unpaid position. ARK and TTB declare that they have no competing financial interests.
|
PubMed Central
|
2024-06-05T03:55:55.032790
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069670/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e81",
"authors": [
{
"first": "Alexei R",
"last": "Koudinov"
},
{
"first": "Temirbolat T",
"last": "Berezov"
}
]
}
|
PMC1069671
|
Gatz\'s statement, "At least half of the explanation for individual differences in susceptibility to Alzheimer disease is genetic" \[[@pmed-0020082-b1]\], is, in my opinion, incorrect. As the one who led the team debating Ashford and Mortimer, whose 2002 article \[[@pmed-0020082-b2]\] supports this statement, at the 2001 conference on Alzheimer disease (AD) in Cincinnati ("Challenging Views of Alzheimer\'s Disease") \[[@pmed-0020082-b3]\], I think that the evidence that dietary and lifestyle factors explain the majority of the individual risk for AD in the US is very strong. My original paper in 1997 \[[@pmed-0020082-b4]\] found that total dietary fat and energy intake were the most important dietary risk factors, while fish and cereal intake were the most important risk reduction factors. These findings have been generally confirmed by Drs. Luchsinger and Morris and others. The reason I did my study was that the Honolulu Heart Study reported that Japanese American men in Hawaii had 2.5 times the risk of AD of native Japanese. African-Americans have about four times the risk of AD of native Nigerians. If genetics were the primary risk factor, those living in the US would have a risk of developing AD very similar to that of individuals living in their ancestral home. The reason this is not the case is that the American diet provides too much food, which is a particular problem for those genetically predisposed to AD.
**Citation:** Grant W (2005) Author\'s reply. PLoS Med 2(3): e82.
[^1]: **Competing Interests:** The author has declared that no competing interests exist.
|
PubMed Central
|
2024-06-05T03:55:55.033772
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069671/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e82",
"authors": [
{
"first": "William",
"last": "Grant"
}
]
}
|
PMC1069672
|
Erick H. Turner \[[@pmed-0020083-b1]\] notes that "ClinicalTrials.gov, a registry authorized by the Food and Drug Modernization Act of 1997, appears not to be comprehensive." While we await the creation of clinical trials registry and results databases that are truly comprehensive, innovative efforts to provide convenient access to credible information about known existing clinical trials need to continue. A Canadian example is provided by OntarioCancerTrials.ca, a consumer-oriented site developed by the Ontario Cancer Research Network (OCRN), with funding from the Ontario government.
**Citation:** Till JE (2005) A Canadian perspective. PLoS Med 2(3): e83.
[^1]: **Competing Interests:** The author was a member of an advisory committee during the initial development of OntarioCancerTrials.ca.
|
PubMed Central
|
2024-06-05T03:55:55.034121
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069672/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e83",
"authors": [
{
"first": "James E",
"last": "Till"
}
]
}
|
PMC1069673
|
In "The Commercialisation of Medical and Scientific Reporting" \[[@pmed-0020084-b1]\], Caulfield calls on journalists to ask researchers about the nature of their funding and the financial relationship of the researchers to the sponsor. This is just one principle addressed in a much broader "Statement of Principles" I wrote this past year for the Association of Health Care Journalists (<http://www.ahcj.umn.edu>). The statement is available online \[[@pmed-0020084-b2]\].
**Citation:** Schwitzer G (2005) Statement of principles for health care journalists. PLoS Med 2(3): e84.
[^1]: **Competing Interests:** The author has declared that no competing interests exist.
|
PubMed Central
|
2024-06-05T03:55:55.034375
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069673/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e84",
"authors": [
{
"first": "Gary",
"last": "Schwitzer"
}
]
}
|
PMC1069674
|
We appreciate the note from Drs. Koudinov and Berezov \[[@pmed-0020085-b1]\]. In our opinion, no model has yet been presented that plausibly accounts for all the data on statins, cholesterol, amyloid-ß protein (Aß), and Alzheimer disease. In our paper \[[@pmed-0020085-b2]\], we present evidence that the isoprenoid pathway contributes to statin-activated shedding of the APP ectodomain in cultured cells. We do not yet know which (if any) other "cholesterol-related" Alzheimer phenomena are also attributable to modulation of isoprenoids, Rho, or ROCK.
Previously, conventional wisdom held that Aß load and hypercholesterolemia were directly related, based on observations that high-fat diet aggravated amyloid pathology in plaque-forming mice \[[@pmed-0020085-b3],[@pmed-0020085-b4],[@pmed-0020085-b5]\]. More recently, however, the formulation that statins act simply via cholesterol-lowering fails to account for several observations that cannot immediately be reconciled, either with the original "dogma" or with each other.
First, Fagan et al. \[[@pmed-0020085-b6]\] questioned the role of cholesterol as the final common pathway in Aß load specification, since, in their experiments, low cholesterol per se apparently had no impact on brain Aß load in plaque-forming transgenic mice. Then, equally puzzling pharmacological data emerged. Atorvastatin was shown to lower brain amyloid load and Aß levels, but brain cholesterol levels were unaffected by the drug \[[@pmed-0020085-b7]\]. In an apparent complete contradiction with the original observations, now, some investigators have been able to devise circumstances under which there is an inverse relationship between cholesterol and Aß, with low neuronal cholesterol increasing Aß generation \[[@pmed-0020085-b8]\], and vice versa \[[@pmed-0020085-b9]\]. These newer observations are unexpected and extremely puzzling, and no comprehensive explanation has yet emerged.
For those readers seeking an update on this challenging area, we would direct your attention to the Alzheimer Research Forum Web page (<http://www.alzforum.org/new/detailprint.asp?id=1135>), where you will find an excellent review of the literature as well as a series of evaluations of how our data fit into existing scenarios and models regarding cholesterol, statins, cerebral amyloidosis, and the cognitive failure of Alzheimer disease.
**Citation:** Gandy S, Petanceska S (2005) Authors\' reply. PLoS Med 2(3): e85.
[^1]: **Competing Interests:** The authors haves declared that no competing interests exist.
|
PubMed Central
|
2024-06-05T03:55:55.034606
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069674/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e85",
"authors": [
{
"first": "Sam",
"last": "Gandy"
},
{
"first": "Suzana",
"last": "Petanceska"
}
]
}
|
PMC1069675
|
Grant \[[@pmed-0020086-b1]\] describes as incorrect the statement that at least half of the explanation for individual differences in risk for Alzheimer disease is genetic. He suggests instead that dietary and lifestyle factors explain the majority of individual susceptibility to Alzheimer disease.
The basis for asserting a 50% or greater role for genetics in Alzheimer disease risk comes from family studies and from twin studies. In family studies, first-degree relatives of individuals with Alzheimer disease are at more than double the risk of Alzheimer disease compared to those with no affected relatives \[[@pmed-0020086-b2],[@pmed-0020086-b3]\]. In twin studies, across different Scandinavian twin registries, estimates of heritability of Alzheimer disease range from 55% to over 70% \[[@pmed-0020086-b4]\].
Genetic risk undoubtedly represents the cumulative influence of many genes, including apolipoprotein E (APOE) and other genes not yet identified. In particular, it appears that the magnitude of the genetic component of Alzheimer disease risk is similar across ethnic communities, but that different genetic factors may contribute differently to that risk in white, Latino, and African American families \[[@pmed-0020086-b5]\].
Further, there are interactions between genetic and environmental risks, for example, between the APOE e4 allele and high cholesterol \[[@pmed-0020086-b6]\] or head injury \[[@pmed-0020086-b7]\].
Clearly Alzheimer disease is the outcome of multiple genetic and multiple environmental influences, operating additively and interactively. If genetic effects account for half of individual differences in liability, then environmental influences also account for half of the variation in susceptibility. From a public health viewpoint, it is vital to identify those influences that are modifiable. Controlling blood pressure and avoiding head trauma are examples. However, it is also important to appreciate that individuals bring differences in genetic risk to the table.
**Citation:** Gatz M (2005) Accounting for individual differences in risk of Alzheimer disease. PLoS Med 2(3): e86.
[^1]: **Competing Interests:** The author has declared that no competing interests exist.
|
PubMed Central
|
2024-06-05T03:55:55.035146
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069675/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e86",
"authors": [
{
"first": "Margaret",
"last": "Gatz"
}
]
}
|
PMC1069676
|
P*LoS Medicine* was launched at a time of unprecedented concern about the influence of hidden competing interests on the medical literature. As a new journal, we had the opportunity to help create "the fully transparent world that is desirable" \[[@pmed-0020088-b1]\]. What are we doing to promote transparency, and is our strategy going far enough?
We ask all authors and reviewers to declare any competing interests---financial, personal, and professional \[[@pmed-0020088-b2]\]---and authors\' declarations are included with all published articles. We reject articles when we believe that authors\' competing interests have compromised their work; for research articles, this means compromised in either the conduct of a study or its interpretation. When we are concerned that reviewers\' competing interests may prevent them from giving an unbiased assessment, we find alternative reviewers. And, as recommended by the International Committee of Medical Journal Editors \[[@pmed-0020088-b3]\], we decline to publish studies when the sponsor controls the decision on publication.
Our main strategy for managing competing interests is disclosure. Financial relationships between industry, researchers, and academic institutions are widespread \[[@pmed-0020088-b4]\], and disclosing these competing interests is a crucial step in helping to protect the public and the reputation of authors and of *PLoS Medicine*. Disclosure also matters because there is increasing evidence that authors\' competing interests have a strong influence on their conclusions. For example, review articles looking at the scientific evidence on the health effects of passive smoking have reached different conclusions; a study of these review articles found that the only factor associated with the conclusion that passive smoking was harmless was whether an author was affiliated with the tobacco industry \[[@pmed-0020088-b5]\]. A study of 159 randomized clinical trials found a significant association between authors\' financial competing interests and their favorable conclusions about an experimental intervention \[[@pmed-0020088-b6]\].
We cannot rely entirely on peer review to detect bias in the conclusions of an article, because although peer review may help to uncover some types of bias, such as bias in the study design, it cannot detect other types, such as bias in the study\'s conduct. Because a competing interests statement accompanies every article published in *PLoS Medicine*, readers can take these interests into account when they assess a paper themselves. The reality is that readers are wary of competing interests. In one study, readers were randomly sent the same paper with or without
> Financial relationships between industry, researchers, and academic institutions are widespread.
the authors having disclosed a financial competing interest \[[@pmed-0020088-b7]\]. Readers scored the paper with the competing interest significantly lower on all five measures: interest, importance, relevance, validity, and believability of the study.
Our policy of asking authors to disclose their competing interests is not, of course, foolproof. The Center for Science in the Public Interest recently examined 163 articles in four scientific journals and identified at least 13 articles for which authors did not disclose relevant conflicts of interest that should have been disclosed according to the journals\' policies \[[@pmed-0020088-b8]\]. Bero and colleagues compared the statement of competing interests of the authors of a 2003 *BMJ* paper on the health effects of secondhand smoke with internal tobacco industry documents describing financial ties between the industry and the authors \[[@pmed-0020088-b9]\]. Although the authors met the *BMJ*\'s requirements for financial disclosure, the disclosure did not provide readers with a full picture of the industry\'s long-standing involvement with the authors.
Should we as journal editors be investigating authors\' ties for ourselves? We do not have the resources to do so, and agree with Bero and colleagues that "an elaborate policing operation is not feasible or necessarily desirable" \[[@pmed-0020088-b9]\]. But in this particular case, say the authors, a quick search of the tobacco documents that are freely available online (e.g., at <http://www.legacy.library.ucsf.edu> or <http://www.bat.library.ucsf.edu>) would have been revealing. Maybe we are heading towards an era when such searches become more common---perhaps initially by randomly selecting papers for investigation.
At last year\'s Council of Science Editors retreat on competing interests, editors came up with a list of questions to think about when formulating a journal\'s competing interests policy ([Table S1](#st001){ref-type="supplementary-material"}). Some are straightforward. Should editors declare their own competing interests? We think so (and have declared ours at <http://medicine.plosjournals.org/perlserv/?request=get-static&name=editors_interests>). Others are more complex. When editors discover that a published author failed to declare a significant competing interest, what should they do? Should they impose sanctions on the author? Should they publish a correction or even retract the paper?
To help us answer such questions, and to advise us on individual cases for which we are concerned about competing interests or broader ethical questions, we have appointed an external advisory group. The group ([Table S2](#st002){ref-type="supplementary-material"}) has expertise in clinical medicine, medical editing, research, health policy, law, and bioethics, and includes a lay member. In addition, an internal committee at PLoS meets monthly to consider competing interests across the organization. We are taking this issue seriously, because we recognize that journals are seen as the gatekeepers of published research. We welcome your feedback on how we are doing at protecting the probity of our content.
Supporting Information {#s2}
======================
Table S1
::: {.caption}
###### Competing Interests Policies: Questions for Editors to Consider
(33 KB DOC).
:::
::: {.caption}
######
Click here for additional data file.
:::
Table S2
::: {.caption}
###### Members of the *PLoS Medicine* Advisory Group on Competing Interests and Publication Ethics
(28 KB DOC).
:::
::: {.caption}
######
Click here for additional data file.
:::
|
PubMed Central
|
2024-06-05T03:55:55.035511
|
2005-3-29
|
{
"license": "Creative Commons Zero - Public Domain - https://creativecommons.org/publicdomain/zero/1.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069676/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e88",
"authors": []
}
|
PMC1069677
|
What is the level of medical evidence that should be used to inform medical practice? At the bottom of the hierarchy of evidence are anecdotes, expert opinion, case reports, and case series, and at the top is the systematic review of published (and sometimes unpublished) evidence. By necessity, systematic reviews come many years after hypotheses are first raised, and in the interim recommendations for practice may sway back and forth. One example of this is the debate over the role of uric acid in heart disease, which has been going on for more than 50 years. It started with a paper published in 1951 in the *Annals of Internal Medicine* that found higher serum uric acid concentrations in patients with coronary heart disease (CHD) compared with controls. Since then, measurement of serum uric acid has been suggested as a predictor of CHD. But many of the studies on serum uric acid are epidemiologic studies---somewhere in the middle of the hierarchy of evidence---and have come to different conclusions about how useful measurement of uric acid is.
Dissecting out the role of uric acid is further complicated by three things: high levels of uric acid are associated with hypertension and being overweight (other risk factors for CHD); levels of uric acid can be altered by drugs such as diuretics that people with CHD often take; and finally, alteration of renal function can affect uric acid levels. Another problem is the type of studies that have been used to address the question of uric acid\'s role in CHD. Retrospective studies may be unable to control adequately for other risk factors---hence prospective, ideally population-based, studies would be the best to answer the question of whether there really is an association between high uric acid and CHD.
In this month\'s *PLoS Medicine*, John Danesh and colleagues from the University of Cambridge, along with investigators from the Icelandic Heart Association, report the single largest prospective study addressing the role of uric acid in heart disease. Further, their systematic review combines their findings with those of 15 previously published prospective studies of serum uric acid---9,458 cases of CHD and 155,084 controls in all.
The paper answers the question of the role of uric acid in prediction of CHD clearly: the risk ratio for prediction of disease was 1.13 (1.07--1.20), but it was only 1.02 (0.91--1.14) in the eight studies that had the most complete adjustment for possible confounders. What this paper does not do is directly address the question of whether or not serum uric acid is involved in causing CHD through intermediates; however, it does suggest that serum uric acid levels are unlikely to be a major determinant of CHD.
Where does such a result leave patients? Well, it is likely that improving diet, losing weight, and controlling blood pressure may all contribute to reducing both one\'s risk of CHD and one\'s serum levels of uric acid. The role of uric acid in CHD is now likely to be of interest only to those studying basic science; for now, the clinical question seems closed.
|
PubMed Central
|
2024-06-05T03:55:55.036190
|
2005-3-29
|
{
"license": "Creative Commons Zero - Public Domain - https://creativecommons.org/publicdomain/zero/1.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069677/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e90",
"authors": []
}
|
PMC1069678
|
What holds cells together or connects them with the extracellular matrix---and what happens when these interactions break down---is one of the keys to determining how tumors metastasize. One group of compounds---integrins---are a central part of these interactions. Not only do integrins play a part in cell--cell and cell--matrix adhesion, but they also are involved in signal transduction (the method by which a cell relays information from receptor binding to cellular response) and in triggering cell death by linking to other molecules. One such member of this receptor family is the avß3 integrin, which is expressed on both the tumor cells and the new vasculature of various tumors, including melanomas. avß3 integrin has a role in cell migration and extravasation, which occurs during metastasis, and also in angiogenesis---the development of new blood vessels that are essential for the growth of tumors. These blood vessels are the target for one class of anti-cancer drugs---angiogenesis inhibitors. Molecules that bind to avß3 integrin have also been used to target therapeutic compounds to tumors: compounds that antagonize this integrin can lead to apoptosis (programmed cell death) of cells that express it.
Haubner and colleagues, the authors of a paper in this month\'s *PLoS Medicine*, have previously developed a fluorine-labeled peptide, \[^18^F\]Galacto-RGD, that has a high affinity for avß3 integrin. \[^18^F\]Galacto-RGD has many of the features essential for a tracer: it is specifically accumulated by tumors that express avß3 integrin, it is efficiently eliminated by the kidneys, and it is stable in vitro and in vivo.
In the research paper in *PLoS Medicine*, Haubner and colleagues take the development of the compound further towards clinical application. First, in a mouse with human melanoma they used highly sensitive positron emission tomography (PET) scanning to show not only that the level of uptake of integrin was specific for the tumor, but also that the uptake was in direct proportion to the amount of avß3 expressed, thus potentially allowing quantification of receptor expression; however, larger tumors showed a poorer correlation, possibly because of the presence of necrotic areas that do not express the integrin.
In humans, this picture was a little less clear; in a small study of patients with tumors including melanoma, the authors found a good deal of difference between patients in the uptake of the marker by tumor cells and the corresponding tumor vasculature. However, there was good correlation between the tracer uptake and conventional staining for the integrin by immunohistochemistry---again suggesting that the marker is truly reflecting the in vivo level of the integrin.
What do these results mean for clinical applications? As well as identifying tumors that express this marker, this approach might also offer a noninvasive way to assess the degree of new vessel formation in tumors. The approach could provide important information for planning and monitoring anti-angiogenic therapies targeting this integrin and could reveal the involvement and role of this integrin in metastatic and angiogenic processes in various diseases.
::: {#pmed-0020091-g001 .fig}
::: {.caption}
###### avß3 integrin expression in tumors
:::
:::
|
PubMed Central
|
2024-06-05T03:55:55.036764
|
2005-3-29
|
{
"license": "Creative Commons Zero - Public Domain - https://creativecommons.org/publicdomain/zero/1.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069678/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e91",
"authors": []
}
|
PMC1069679
|
The prevalence of asthma and allergy has risen in all industrialized countries during recent decades, and there is much debate about exposure to pets in early life and later development of asthma and allergy. Some studies have suggested that keeping pets actually protects against later allergy---i.e., that early exposure may somehow modify an individual\'s immune system to tolerate specific antigens. What might be the mechanism for such protection against allergy? One theory of how allergies arise is that an imbalance in T helper cell subtypes tips the body\'s immune response towards overreacting to a particular antigen. There is some evidence that early exposure to high natural levels of cat allergens can prevent such an inappropriate immune response. Other researchers have suggested that normally immune responses are kept under control by another group of T cells---regulatory T cells. The two mechanisms may be linked, since exposure to high levels of cat allergens may induce regulatory T cells.
Various attempts to modify aberrant immune responses to specific allergens, such as those to cat dander, have been made. Investigators have treated patients with related molecules, either peptides derived from the allergen itself, or much smaller peptides produced synthetically. Although therapy with peptides seems to reduce allergic responses, the mechanism of the response to treatment has not been clear, in particular, exactly which cells, cell surface markers, and cytokines are involved in modifying the immune response.
::: {#pmed-0020094-g001 .fig}
::: {.caption}
###### Identifying the T cells that suppress proliferation to allergens after immunotherapy
:::
:::
In a paper in this month\'s *PLoS Medicine*, Mark Larché and colleagues have attempted to dissect out this pathway in a group of individuals with asthma and allergy to cats. They treated the individuals with short synthetic peptides derived from the sequence of the major cat allergen, *Felis domesticus* allergen 1, and then measured the clinical and immunological response to allergen. They found that treatment with the peptides led to the induction of a population of T cells that were capable of suppressing the proliferation of allergen-reactive T cells in vitro. Peptide treatment also resulted in increased levels of a molecule called CD5 on the surface of blood T cells---CD5 has recently been associated with suppressing T cell sensitivity to stimulation. Finally, the authors found that the degree of suppression was not related to the amount of peptide given to the patients.
Where does this finding leave patients who might wonder about exposure to cats and the development of allergy? The simple answer is that we do not know exactly how exposure to antigen triggers either an immune reaction or tolerance. Once triggered, an immune reaction to a cat may be hard---though not impossible---to reverse, but how or why a specific individual becomes sensitized is as yet far from clear.
|
PubMed Central
|
2024-06-05T03:55:55.037255
|
2005-3-29
|
{
"license": "Creative Commons Zero - Public Domain - https://creativecommons.org/publicdomain/zero/1.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069679/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e94",
"authors": []
}
|
PMC1069680
|
Drs. Speiser, Cerottini, and Romero \[[@pmed-0020095-b1]\] correctly point out that CD8+ T cells from HLA-A\*0201 melanoma patients and healthy donors may contain small populations (on average, 0.07% ± 0.05% in their publications \[[@pmed-0020095-b2],[@pmed-0020095-b3]\]) that bind tetramers made with the heteroclitic Melan-A M26 peptide, and that such cells express a naïve phenotype (CD45RA+). We too observe this phenomenon in a portion of HLA-A\*0201 healthy donors and patients with melanoma that we analyze with M26 tetramers. Importantly, this is not seen in all subjects. These cells do not recognize the native M27 peptide, and represent cross-reactive subsets of naïve CD8+ T cells of multiple specificities \[[@pmed-0020095-b4]\]. We routinely analyze all subjects pre-vaccination, and the four post-vaccination responses analyzed in our report \[[@pmed-0020095-b5]\] did not contain M26 or gp100 tetramer-binding cells pre-vaccination (data not shown). Thus, it was unlikely that M26-cross-reactive cells spontaneously developed post-vaccination (not due to peptide vaccination) and were the basis of some of the low-recognition-efficiency MART-specific clones we analyzed. Furthermore, such a phenomenon has been seen only with M26 and not with the heteroclitic gp100 (G209-2M) peptide, so would not be a factor in the gp100-specific responses we analyzed.
The authors also point out that in their experience, they found that the majority of T cells generated with the heteroclitic Melan-A M26 peptide were tumor reactive, citing their studies in vitro \[[@pmed-0020095-b6]\], in mice \[[@pmed-0020095-b7]\], and in patients with melanoma \[[@pmed-0020095-b8]\]. We focus on their publication on patients with melanoma, as this is most directly relevant to our study. This report \[[@pmed-0020095-b8]\] focused on three patients with melanoma immunized with M26, and analyzed T cells from lymph nodes draining the vaccination site (vaccine-site sential nodes \[VSSNs\]). VSSNs from these three patients contained 0.11% (0.08%--0.15%) MART-specific T cells by tetramer staining. Importantly, contralateral lymph nodes in these subjects (distant from the vaccination site) also contained 0.06% (0.05%--0.09%) MART-specific T cells. With their reported background of less than 0.01%, this suggests the possibility of endogenous MART-specific T cells within lymph nodes in these subjects. These authors have shown in previous studies that endogenous MART-specific T cell responses frequently exist within lymph nodes, even in the absence of such cells in peripheral blood mononuclear cells \[[@pmed-0020095-b9]\]. Furthermore, these VSSN responses were analyzed only after two vaccinations, while the authors could not detect circulating MART-specific T cells in any of these three patients even after six vaccinations. MART-specific T cell lines were generated via tetramer-guided sorting of VSSN cells from patients 2 and 3, then individual clones generated via limiting dilution. They reported 16 of 17 clones killed A2+ MART+ melanoma targets. Without knowing the Vbeta usage of these clones and the Vbeta diversity of the parental MART-specific populations, it is difficult to know what fraction of each response these clones accounted for in the two subjects, as we have done in our study. More importantly, these tumor-reactive clones analyzed may be derived from endogenous T cell responses, possibly amplified by vaccination, rather than from de novo vaccine-elicited T cell responses. If so, these data would in fact fit well with our findings that endogenous responses consist mainly of cells with tumor-cytolytic potential that recognize the native peptide with high recognition efficiency.
In our study \[[@pmed-0020095-b5]\], we analyzed in detail four vaccine-elicited T cell responses (two to M26 and two to G209-2M) via the generation of more than 200 cytotoxic T lymphocyte clones, and assessed the fraction of each response that these clones accounted for collectively by analyzing the Vbeta usage of each clone and the parental peptide-specific populations. From this, we showed that the vaccine-elicited T cells were diverse in their tumor-cytolytic potential, which correlated with their recognition efficiency for the native peptides. It is important to point out that tumor-cytolytic T cells were present in these four subjects, but represented a significantly lower fraction than those derived from endogenous responses. These data are consistent with those we recently reported using a different experimental approach---assessing the fraction of T cells in a tetramer+ population that degranulate (via CD107 mobilization) to tumor stimulation \[[@pmed-0020095-b10]\]. While generating individual cytotoxic T lymphocyte clones and analyzing each for tumor killing and recognition efficiency, as we have done in this study, is the most definitive approach to analyze individual cells within an antigen-specific T cell response, this approach is extremely labor-intensive, and thus not feasible for large numbers of patients. As such, more rapid flow-cytometry-based methods, such as the CD107 mobilization assay and a new method to rapidly assess recognition efficiency of a T cell population via differential TCR downregulation (H. E. Kohrt, C. T. Shu, S. P. Holmes, J. S. Weber, P. P. L., et al., unpublished data), will allow analysis of many more patients to various vaccine formulations and strategies. This knowledge will be vital to the improvement of future cancer immunotherapies.
**Citation:** Stuge TB, Lee PP (2005) Authors\' reply. PLoS Med 2(3): e95.
[^1]: **Competing Interests:** The authors have declared that no competing interests exist.
|
PubMed Central
|
2024-06-05T03:55:55.037754
|
2005-3-29
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1069680/",
"journal": "PLoS Med. 2005 Mar 29; 2(3):e95",
"authors": [
{
"first": "Tor B",
"last": "Stuge"
},
{
"first": "Peter P",
"last": "Lee"
}
]
}
|
PMC1073691
|
Introduction {#s1}
============
In both rodents and humans, the liver is the largest solid organ, and performs critical immunological and metabolic functions. The liver receives nutrient-rich blood from the gut via the portal vein and oxygenated blood from the hepatic artery. It processes blood to remove toxins, synthesizes the majority of serum proteins and lipids, stores glycogen, and performs extensive lipid, cholesterol, and vitamin chemistry and storage. The liver is thought to provide a unique environment for lymphocytes, favoring tolerogenic immune system responses, possibly to prevent reactivity to harmless food antigens \[[@pbio-0030113-b01]\]. However, in response to certain stimuli, acute inflammatory reactions can occur, and result in hepatocyte death (hepatitis) and subsequent regeneration, with progressive fibrosis when stimuli are sustained. Several progressive liver diseases that can lead to liver failure have an autoimmune component \[[@pbio-0030113-b02]\].
The liver is an important site of visceral infection. The low-pressure circulation and high surface area of contact between blood and parenchymal cells and the high phagocytic capacity of multiple cell types in liver provide pathogens with an easy route of access. The tolerizing environment may additionally contribute to immune avoidance. The World Health Organization estimates that approximately 5% and 3% of the world\'s population carry hepatitis B and hepatitis C virus, respectively \[[@pbio-0030113-b03]\]. Many of these cases result in chronic infections that can lead to fatal complications, including hepatocellular carcinoma, cirrhosis, or hemorrhage. Malaria and leishmania also display important liver tropisms \[[@pbio-0030113-b04],[@pbio-0030113-b05]\]. Thus, immune surveillance of the liver for pathogens is an important, but poorly understood, process.
The profile of steady-state hepatic immune cells differs markedly from that in secondary lymphoid organs and in other non-lymphoid tissues, with abundant Kupffer cells and natural killer T cells (NKT cells) supplemented with αβ T cells, γδ T cells, natural killer (NK) cells, dendritic cells, and few, if any, B cells. NKT cells, present at trace levels (\<1%) in many organs, are highly enriched in the liver, where they represent up to 30% of lymphocytes \[[@pbio-0030113-b06]\]. These lymphocytes co-express NK receptors (e.g., NK1.1) and T cell antigen receptors (TCRs). The largest subset of NKT cells includes thymus-derived CD1d-reactive cells, most of which express a semi-invariant TCR containing Vα14--Jα18 and Vα24--Jα15 rearrangements in mouse and human, respectively, and a restricted Vβ repertoire \[[@pbio-0030113-b07]\]. Nearly all of these lymphocytes react with the marine-sponge-derived glycolipid α-galactosyl ceramide (αGalCer) presented by CD1d. The CD1d-specific NKT cells are required for Concanavalin A (ConA)--induced hepatitis \[[@pbio-0030113-b08],[@pbio-0030113-b09],[@pbio-0030113-b10]\], a model of autoimmune hepatitis \[[@pbio-0030113-b11]\], and are implicated in a variety of other hepatic immunological reactions, including tumor rejection, inhibition of hepatitis B replication, and anti-malarial responses \[[@pbio-0030113-b07]\].
Lymphocyte trafficking to and within lymphoid organs and also to extra-lymphoid organs such as inflamed gut and skin is governed in large part by the interactions of selectively expressed adhesion receptors and chemokine receptors on the surface of diverse lymphocytes with ligands on the endothelium \[[@pbio-0030113-b12]\]. This is followed by extravasation of the adherent lymphocytes across the endothelium. This paradigm of lymphocyte homing has emerged mainly from studies of high-velocity vascular beds with continuous endothelium and post-capillary venules in lymphoid tissues and some inflamed non-lymphoid tissues \[[@pbio-0030113-b13]\]. However, the dynamic behavior of lymphocytes in peripheral non-lymphoid tissue such as the liver is largely unknown.
The accumulation of lymphocytes in liver sinusoids (devoid of E/P-selectins, PECAM, CD34, and VE-cadherin, and low in VCAM-1) appears to result from mechanisms distinct from those involved in multistep extravasation of intravascular lymphocytes \[[@pbio-0030113-b14]\]. Accumulation of CD1d-reactive T cells in the liver has been shown to require LFA-1 expression on liver cells other than NKT cells \[[@pbio-0030113-b15]\], which implicates hematopoietic cells such as Kupffer cells in capture of NKT cells since endothelial cells do not express LFA-1. However, all activated lymphocytes upregulate ICAM-1, so other mechanisms are needed to account for the large relative enrichment of NKT cells in liver compared to NKT cells in other sites and compared to other types of lymphocytes in the liver.
CXCR6/Bonzo/STRL33 is a chemokine receptor that can serve in conjunction with CD4 as a co-receptor for infection with some human and most simian immunodeficiency viruses (HIV-1, HIV-2, and SIV) \[[@pbio-0030113-b16]\] and, similarly to CCR5 and CXCR3, has an expression pattern restricted to memory/effector T cells such as NKT cells \[[@pbio-0030113-b17],[@pbio-0030113-b18],[@pbio-0030113-b19],[@pbio-0030113-b20],[@pbio-0030113-b21]\]. CXCR6 has one known ligand, CXCL16, a transmembrane chemokine expressed in spleen, lung, and liver, and on macrophages, dendritic cells, and some spleen sinus-lining cells \[[@pbio-0030113-b19],[@pbio-0030113-b21],[@pbio-0030113-b22]\].
In this investigation, we studied liver CD1d-specific NKT cells using mice in which the *cxcr6* coding region was replaced with a green fluorescent protein (GFP) cDNA. We demonstrate that CXCR6^+^ CD1d-reactive T cells crawl within liver sinusoids at speeds of 10--20 μm/min, and stop crawling upon TCR activation, revealing a novel form of intravascular immune surveillance. Interestingly, a complete deficiency in CXCR6 results in a selective and severe reduction of hepatic CD1d-reactive NKT cells, and is correlated with a decreased susceptibility to T-cell-dependent hepatitis. We observe that CXCR6 deficiency does not detectably alter the behavior of individual hepatic CD1d-specific NKT cells as they crawl along the CXCL16-expressing sinusoidal endothelial cells, but by regulating the number of hepatic NKT cells, at least in part by enhancing their survival, CXCR6 regulates liver-based immune responses.
Results {#s2}
=======
CXCR6^+^ T Cells Are Enriched in the Liver {#s2a}
------------------------------------------
GFP expression in *cxcr6^gfp/+^* mice is restricted to subsets of activated/memory T cells \[[@pbio-0030113-b20]\], including NKT cells \[[@pbio-0030113-b17],[@pbio-0030113-b23]\]. CD1d-reactive NKT cells accounted for 75%--80% of GFP^hi^ cells in the liver ([Figure 1](#pbio-0030113-g001){ref-type="fig"}A), and approximately half of GFP^hi^ cells in the spleen and other organs ([Figure 1](#pbio-0030113-g001){ref-type="fig"}A; data not shown).
::: {#pbio-0030113-g001 .fig}
Figure 1
::: {.caption}
###### CXCR6^+^ Liver CD1d-Reactive T Cells Bind CXCL16 and Are Localized within CXCL16^+^ Liver Sinusoids
\(A) Binding of CXCL16 by splenocytes and liver leukocytes from *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice. Splenocytes and liver leukocytes from *cxcr6^gfp/+^* (+/−) and *cxcr6^gfp/gfp^* (−/−) mice were stained with PE-conjugated CD1d tetramer loaded with α-GalCer (αGC), PerCp-conjugated anti-CD3 antibodies and CXCL16-Fc fusion protein, and Cy5-conjugated goat anti-human Fc. The two black curves represent duplicate samples from two different *cxcr6^gfp/+^* mice.
\(B) Thymocytes, splenocytes, and liver leukocytes from *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice were stained with PE-conjugated CD1d tetramer loaded with α-GalCer and biotinylated anti-NK1.1 followed by PerCp-conjugated streptavidin and APC-conjugated anti-CD3 antibodies. Left panels are gated on CD3^+^ T cells and represent GFP fluorescence intensity on CD1d/α-GalCer-reactive T cells and on conventional (tetramer-negative) T cells. The right panels are gated on CD3^−^ cells and represent GFP fluorescence intensity on NK1.1^+^CD3^−^ NK cells. Numbers are percent obtained from a representative experiment among six performed.
\(C) Liver sinusoidal endothelial cells express CXCL16. Sections (10 μm thick) of PFA-fixed liver from wild-type mice (top panel, left), *cxcr6^gfp/+^* mice (top panel, right), and Tie2-GFP mice (bottom panel) were stained with rat monoclonal antibody against mouse CXCL16 (Clone 10H7, IgG2a), or with an isotype control, followed by goat anti-rat antibody (F\[ab′\]2) conjugated to Cy3.
\(D) GFP^hi^ CD1d-restricted T cells within liver sinusoids. Sections (7 μm thick) of PFA-fixed liver from *cxcr6^gfp/+^* mice were stained with rabbit polyclonal serum against caveolin-1 followed by goat anti-rabbit antibody conjugated to Cy3.
\(E) Flow cytometry analysis of cells harvested by perfusion of *cxcr6^gfp/+^* liver with cold PBS--ethylenediamine tetra-acetic acid.
:::
:::
To confirm the correlation of GFP expression with that of CXCR6 at the plasma membrane on a per-cell basis, *cxcr6^gfp/+^* lymphocytes were analyzed for binding of CXCL16, the only known ligand of CXCR6, using a CXCL16-Fc fusion protein \[[@pbio-0030113-b19]\]. We detected CXCL16 binding in GFP^hi^ cells in *cxcr6^gfp/+^*, but not *cxcr6^gfp/gfp^* mice, in both spleen and liver, confirming that GFP^hi^ cells in heterozygous mice express the receptor and that its expression is bi-allelic ([Figure 1](#pbio-0030113-g001){ref-type="fig"}A, R1- and R2-gated cells). We also observed that while 50% of thymic and 80% of splenic CD1d-tetramer-positive NKT cells expressed GFP, 99% of such cells expressed GFP/CXCR6 in liver ([Figure 1](#pbio-0030113-g001){ref-type="fig"}B). Conventional T cells expressing CXCR6 were also found to be more frequent in liver than in other organs, and a subset of CD3^−^ NK cells expressing CXCR6 represented 20% of NK cells in the liver, but was not observed in other organs, including spleen, thymus, lung, and intestine ([Figure 1](#pbio-0030113-g001){ref-type="fig"}B; data not shown).
CXCL16, the CXCR6 Ligand, Is Expressed in Liver Sinusoids {#s2b}
---------------------------------------------------------
We next examined the expression of the CXCR6 ligand, the transmembrane chemokine CXCL16, by immunofluorescence confocal microscopy of fixed tissue sections. CXCL16 has been shown by both RNA and protein analysis to be expressed in the spleen, lung, and liver \[[@pbio-0030113-b19],[@pbio-0030113-b21],[@pbio-0030113-b22]\]. Studies in mice have suggested that CXCL16 is expressed on dendritic cells and macrophages \[[@pbio-0030113-b19],[@pbio-0030113-b22]\]. Staining of liver sections with a monoclonal anti-CXCL16 antibody showed a pattern consistent with CXCL16 expression on liver sinusoids and the contiguous centro-lobular venules, but not in portal tracts ([Figure 1](#pbio-0030113-g001){ref-type="fig"}C, upper panels; data not shown). Anti-CXCL16 staining of liver from Tie2-GFP transgenic mice, which express GFP in endothelial cells, showed membrane localization of the CXCL16 signal in tight association with the cytoplasmic GFP signal ([Figure 1](#pbio-0030113-g001){ref-type="fig"}C, lower panels), confirming that the chemokine is expressed in sinusoidal endothelial cells. A similar pattern of expression has been observed in human liver (M. Briskin, unpublished data).
CXCR6^+^ NKT Cells Patrol Liver Sinusoids {#s2c}
-----------------------------------------
We initially examined the anatomical localization of CXCR6^+^ T cells in the liver by immunofluorescence confocal microscopy of fixed tissue sections. Surprisingly, most GFP^+^ cells in the liver were localized in hepatic sinusoids, separated from hepatocytes by endothelial cells that stained with antibodies against caveolin-1 and CXCL16 ([Figure 1](#pbio-0030113-g001){ref-type="fig"}C and [1](#pbio-0030113-g001){ref-type="fig"}D). Relatively few GFP^+^ cells were found in the portal tracts. To confirm that the GFP^+^ cells were in fact CD1d-reactive NKT cells, livers were perfused with cold phosphate-buffered saline (PBS)/2 mM ethylenediamine tetra-acetic acid to release cells in the vascular lumen. The effluent contained a number of GFP^hi^ cells similar to those obtained from the entire liver (data not shown), among which 75% were CD1d-α-GalCer-specific T cells ([Figure 1](#pbio-0030113-g001){ref-type="fig"}E). It is also notable that within liver sinusoids there is a distinct population that expresses a very high level of GFP and binds CD1d-α-GalCer tetramer with moderate to low efficiency despite having normal expression of TCRβ ([Figure 1](#pbio-0030113-g001){ref-type="fig"}A). This sub-population is unresolvable in *cxcr6^gfp/gfp^* mice, possibly as a consequence of bi-allelic GFP expression. We have not characterized this population further, but believe that it is a subset of NKT cells with reduced reactivity for CD1d-α-GalCer, possibly the non-Vα14 NKT cells that have been described elsewhere \[[@pbio-0030113-b07]\].
To further understand the behavior of these sinusoidal NKT cells, we used fluorescence confocal imaging to observe the livers of live mice in real time (intravital microscopy). These experiments revealed that more than 90% of GFP^+^ cells in *cxcr6^gfp/+^* mice were stably attached to the sinusoidal wall. Time-lapse imaging revealed that these cells were actively crawling within the sinusoids ([Figure 2](#pbio-0030113-g002){ref-type="fig"}A; Videos [S1](#sv001){ref-type="supplementary-material"} and [S2](#sv002){ref-type="supplementary-material"}) with a mean speed of 16.5 ± 8.3 μm/min ([Figure 2](#pbio-0030113-g002){ref-type="fig"}B; [Video S3](#sv003){ref-type="supplementary-material"}), more than 100-fold slower than typical rolling in post-capillary venules. Various examples of crawling patterns of hepatic NKT cells are shown in [Video S4](#sv004){ref-type="supplementary-material"}. Further analysis showed that crawling NKT cells choose their direction of motion largely randomly, and independently of the location of central veins or the direction of blood flow. This analysis required imaging at relatively high speeds (approximately 1 frame/s) after injecting fluorescent dextran intravenously, allowing us to clearly identify inlets and outlets corresponding to regions proximal to portal triads and central veins ([Video S5](#sv005){ref-type="supplementary-material"}). Moreover, cells were observed to pass each other in opposite directions in the same sinusoid and to reverse direction within a single sinusoid (see Videos [S1](#sv001){ref-type="supplementary-material"} and [S4](#sv004){ref-type="supplementary-material"}). The mean ratio of cellular displacement over actual path length was 0.44 + 0.31 ([Figure 2](#pbio-0030113-g002){ref-type="fig"}C).
::: {#pbio-0030113-g002 .fig}
Figure 2
::: {.caption}
###### CXCR6^+^ Lymphocytes Patrol Liver Sinusoids
\(A) Select confocal microscopic images from intravital videos of liver of an anesthetized *cxcr6^gfp/+^* mouse (40× magnification). CD1d-reactive cells (bright green) can be seen migrating along hepatic sinusoids at an average speed of 16 μm/min. Scale bar is 25 μm. Cell tracks were traced and quantified using Volocity cell-imaging software. Note tracks of cells traveling in opposite directions in the same sinusoid (left side of image, at 0--5 versus 7--12 min).
\(B) Velocity quantification of GFP^+^ lymphocytes. Videos of liver in *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice (10× magnification) were analyzed for the velocity of GFP^+^ cells in 640 cell migration tracks for *cxcr6^gfp/+^* mice and 574 cell migration tracks for *cxcr6^gfp/gfp^* mice. Results demonstrate similar velocities of cells with the *cxcr6^gfp/+^* (filled bars, average velocity 16.5 ± 8.3 μm/min) and *cxcr6^gfp/gfp^* genotypes (unfilled bars, average velocity 18.4 ± 9.5 μm/min).
\(C) Analysis of directedness of cell migration. The same cell tracks as in (B) were analyzed for ratio of overall cell displacement to stepwise-summed path length. Results demonstrate similar degrees of directed crawling by cells from *cxcr6^gfp/+^* (filled bars, average 0.44 ± 0.31) and *cxcr6^gfp/gfp^* (unfilled bars, average 0.42 ± 0.30) mice.
\(D) Analysis of crawling relative to blood flow in peri-central vein areas of *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice. Histograms represent the frequency of cell movements made towards or away from nearby draining areas in likely close proximity to central veins (solid blue curve). The radial step of a cell is the cellular displacement along the axis defined by the cell\'s initial starting point to the central vein. The same step distances were assigned random orientations and the resultant data were plotted (red dashed curves).
:::
:::
The wide range of values reflects the branching nature of liver sinusoids and the chaotic nature of NKT cell migration, with some cells persisting in one direction for tens of microns, resulting in high values, and others frequently changing direction, resulting in low values. We determined whether GFP^+^ cells moved towards or away from central veins by plotting the radial step (the cellular displacement along the axis from the cell\'s initial point to the central vein) for each cell at each consecutive time point. In *cxcr6^gfp/+^* mice, steps towards and away from the central vein were equally frequent ([Figure 2](#pbio-0030113-g002){ref-type="fig"}D, blue solid line), and the average radial step was nearly zero (0.02 μm away from the central vein). We also plotted a distribution attained when the angles for the same steps were randomized, and this curve ([Figure 2](#pbio-0030113-g002){ref-type="fig"}D, red dashed curve) displayed a nearly perfect overlay with the experimental data (blue solid curve). A similar analysis for portal triads showed the same result (not shown). Thus, as a population, NKT cells choose their direction of motion largely randomly, and independently of the location of central veins or the direction of blood flow. We additionally conclude that there is no evidence for a haptotactic gradient of CXCL16 or any other ligand directing cells either towards or away from the central veins.
NKT Cells Stop Patrolling upon TCR Activation {#s2d}
---------------------------------------------
By analogy to conventional T cells, which stop migrating in response to antigen in vitro \[[@pbio-0030113-b24]\], we hypothesized that TCR activation would modify the patrolling behavior of hepatic NKT cells. Indeed, following injection of anti-CD3ɛ antibody or ConA, 70%--90% of patrolling GFP^+^ cells arrested their movement within minutes ([Figure 3](#pbio-0030113-g003){ref-type="fig"}A; [Video S6](#sv006){ref-type="supplementary-material"}). Injection of a control anti-CD1d antibody (clone 1B1) or of several T-cell-binding antibodies, including some matched to isotype of the anti-CD3 antibody, had no effect on patrolling ([Figure 3](#pbio-0030113-g003){ref-type="fig"}A; data not shown). Intravenous delivery of the surrogate NKT cell antigen α-GalCer also triggered the arrest of most patrolling GFP^+^ cells, and was blocked by pre-administration of anti-CD1d antibody ([Figure 3](#pbio-0030113-g003){ref-type="fig"}A; [Video S7](#sv007){ref-type="supplementary-material"}). These results further confirm that most patrolling GFP^+^ cells are CD1d-reactive T cells and indicate that these patrolling liver NKT cells undergo rapid arrest upon activation via TCRs.
::: {#pbio-0030113-g003 .fig}
Figure 3
::: {.caption}
###### CXCR6^+^ Lymphocytes Stop Patrolling upon TCR Activation
\(A) Activation of NKT cells delivers stop signal. *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice were imaged before and after intravenous injection of antibody, lectin, or glycolipid antigen. Percentage of immobile cells was determined in 6-min videos either before any injection or 40 min after antigen delivery.
\(B) Histological examination of hepatic CD1d expression. Liver tissue from a Tie2-GFP transgenic was stained with anti-CD1d antibody (left panel, red) or an isotype control (inset). Green fluorescence indicates sinusoid-lining endothelial cells (expressing GFP) as well as the highly autofluorescent hepatocytes.
\(C) Flow-cytometric examination of hepatic CD1d expression. Cells isolated from liver tissue were stained with antibodies to endothelial cell markers (CD31 and Tie2), a Kupffer cell marker (F4/80), or a dendritic cell marker (CD11c) while co-stained with anti-murine CD1d (filled-in) or an isotype-matched control (dashed line).
:::
:::
CD1d Is Expressed on Hepatic Sinusoid-Lining Cells {#s2e}
--------------------------------------------------
We sought to determine which hepatic cell types might be responsible for presenting α-GalCer to the intravascular NKT cells. Fluorescence microscopy of fixed tissue sections showed bright staining for CD1d in a pattern that was difficult to interpret ([Figure 3](#pbio-0030113-g003){ref-type="fig"}B). The pattern suggests high expression on hepatocytes, including an intracellular peri-nuclear compartment, but whether other cell types also express CD1d was difficult to assess by this method. Using flow cytometry, we observed that sinusoid-lining endothelial cells (CD31^+^ cells and Tie2^+^ cells) express high levels of CD1d on their surface, and Kupffer cells (F4/80^+^) and dendritic cells (CD11c-high cells) express lower, but clearly detectable, levels of surface CD1d ([Figure 3](#pbio-0030113-g003){ref-type="fig"}C). Thus, intravascular hepatic NKT cells have easy access to CD1d-expressing cells at all times.
CD1d-Reactive T Cells Are Selectively Reduced in the Liver of CXCR6-Deficient Mice {#s2f}
----------------------------------------------------------------------------------
To investigate a potential role for CXCR6 in patrolling, we also compared heterozygous *cxcr6^gfp/+^* mice to *cxcr6^gfp/gfp^* (i.e., CXCR6-deficient) littermates. We reasoned that because CXCL16 is a *trans*-membrane chemokine, CXCR6 could be involved in the intravascular movement of the CD1d-reactive T cells by affecting their adhesion or polarization during their migration within sinusoids. However, sinusoidal patrolling by GFP^+^ cells was similar in *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice; the speed (16.5 ± 8.3 μm/min versus 18.4 ± 9.5 μm/min; see [Figure 2](#pbio-0030113-g002){ref-type="fig"}B and Videos [S2](#sv002){ref-type="supplementary-material"} and [S3](#sv003){ref-type="supplementary-material"}), directedness (see [Figure 2](#pbio-0030113-g002){ref-type="fig"}C), direction (see [Figure 2](#pbio-0030113-g002){ref-type="fig"}D), and stopping (see [Figure 3](#pbio-0030113-g003){ref-type="fig"}A) were unaltered. Our data thus do not support a critical role of CXCR6 in general crawling behavior of liver NKT cells.
In comparing the heterozygous to the homozygous mutant mice, however, we noticed a significant reduction in the number of GFP^+^ NKT cells in livers from the CXCR6-deficient animals. Flow cytometry revealed 3- to 5-fold fewer GFP^hi^ CD3^+^ TCRβ^+^ CD1d-reactive NKT cells in livers from homozygous null versus heterozygous mice ([Figure 4](#pbio-0030113-g004){ref-type="fig"}A--[4](#pbio-0030113-g004){ref-type="fig"}C). Notably, GFP^+^ CD1d-reactive tetramer-binding T cells in the thymus, peripheral blood, spleen, bone marrow, and lung were found in similar numbers in *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* littermates ([Figure 4](#pbio-0030113-g004){ref-type="fig"}C; data not shown). GFP^+^ and GFP^−^ CD44^hi^ conventional effector/memory T cells and NK1.1^+^ CD3^−^ NK cells were also found in similar numbers in the liver, thymus, peripheral blood, spleen bone, marrow, lung, and small intestine of *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* littermates (data not shown). No significant difference in number of hepatic CD1d-reactive NKT cells was observed between wild-type and *cxcr6^gfp/+^* mice (data not shown).
::: {#pbio-0030113-g004 .fig}
Figure 4
::: {.caption}
###### CD1d-Reactive NKT Cells Express CXCR6, Are Enriched in Liver, and Are Selectively Reduced in CXCR6-Deficient Mice
Splenocytes, thymocytes, and lung, bone marrow, blood, and liver leukocytes from *cxcr6^gfp/+^* (+/−) and *cxcr6^gfp/gfp^* (−/−) mice were stained with one of the following combinations: K1.1 PE-conjugated antibodies and APC-conjugated anti-TCRβ; PE-conjugated CD1d tetramer loaded with α-GalCer (αGC) and APC-conjugated anti-TCRβ; or control unloaded tetramer and APC-conjugated anti-TCRβ.
\(A) Expression of CD3 and GFP by cells from different organs.
\(B) Reduction in CD1d -α-GalCer-tetramer-positive liver leukocytes in *cxcr6^gfp/gfp^* mice.
\(C) Selective reduction of CD1d-reactive T cells in the liver of CXCR6-deficient mice. Results are mean ± standard deviation from four experiments.
\(D) Reduced ability of CXCR6-deficient NKT cells to accumulate in the liver of recipient mice. Thymocytes from *cxcr6^+/+^* mice were co-transferred with thymocytes from either *cxcr6^gfp/+^* or *cxcr6^gfp/gfp^* littermates into TCRα-deficient mice. The phenotype of hepatic leukocytes 2 d after transfer is shown. Similar results were obtained at 3 days. The results shown are representative of three experiments.
:::
:::
To prove that this defect was T cell intrinsic, 1.5 × 10^7^ thymocytes from 4-wk-old *cxcr6^+/+^* mice were mixed with equal numbers of thymocytes from *cxcr6^gfp/+^* or *cxcr6^gfp/gfp^* littermates and co-transferred intravenously into *tcrα^−/−^* recipients devoid of TCR*α*β lymphocytes. After 2--3 d, donor-derived *cxcr6^gfp/gfp^* NKT cells were underrepresented in the recipient liver in comparison to cells from heterozygous or wild-type control mice ([Figure 4](#pbio-0030113-g004){ref-type="fig"}D). Together, these findings demonstrate that CXCR6 has a role in mediating the accumulation in liver of CD1d-reactive NKT cells.
CXCR6 Expression Prolongs Survival of CD1d-Reactive T Cells {#s2g}
-----------------------------------------------------------
Because CXCR6 deficiency appeared to not alter the steady-state behavior of crawling liver NKT cells, we hypothesized that their reduced number may be caused by their impaired survival or decreased proliferation. Staining of liver sections with antibody specific for activated caspase-3 showed only 1%--2% apoptotic cells among GFP^hi^ cells in *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice, and Annexin V staining of CD1d-reactive GFP^hi^ cells by flow cytometry gave similar results (data not shown). This may reflect a slow rate of apoptosis, rapid scavenging of apoptotic cells by Kupffer cells \[[@pbio-0030113-b25],[@pbio-0030113-b26]\], or detachment of apoptotic NKT cells from the liver sinusoids in vivo. We therefore examined apoptosis ex vivo, by culturing Percoll gradient-purified liver mononuclear cells from *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice for various time periods in 96-well plates. The number of GFP^hi^-tetramer-positive T cells from *cxcr6^gfp/gfp^* mice, recovered after 18 h in culture, was dramatically reduced in comparison with that from *cxcr6^gfp/+^* mice ([Figure 5](#pbio-0030113-g005){ref-type="fig"}A and [5](#pbio-0030113-g005){ref-type="fig"}B). Consistent with this observation, Annexin V staining of CD1d-reactive GFP^hi^ cells isolated from liver indicated that *cxcr6^gfp/gfp^* cells underwent apoptosis more rapidly than *cxcr6^gfp/+^* cells ([Figure 5](#pbio-0030113-g005){ref-type="fig"}C). CD1d-reactive T cells did not proliferate ex vivo, as evaluated by staining with Cell Tracker 633 ([Figure 5](#pbio-0030113-g005){ref-type="fig"}A), suggesting that impaired survival rather than reduced proliferation contributes to the reduction in CD1d-reactive NKT cells recovered from the CXCR6-deficient mice. Addition of exogenous CXCL16 to the ex vivo cultures resulted in a significant enhancement of survival of *cxcr6^gfp/+^* cells, but it is worth noting that even without exogenous CXCL16, *cxcr6^gfp/+^* cells survived longer ex vivo than *cxcr6^gfp/gfp^* cells ([Figure 5](#pbio-0030113-g005){ref-type="fig"}B), suggesting that the pro-survival consequences of CXCR6 signaling can persist for long periods of time (days). However, it is possible that contaminating endothelial cells or soluble CXCL16 in the fetal bovine serum provided additional CXCR6 stimulation during the ex vivo incubation.
::: {#pbio-0030113-g005 .fig}
Figure 5
::: {.caption}
###### Requirement for CXCR6 in Survival of CD1d-Reactive T Cells
\(A) Flow cytometry analysis of liver leukocytes from GFP knock-in mice after overnight culture. Dot plots represent GFP signal versus CD1d-α-GalCer-tetramer staining. Histogram plots represent Cell Tracker 633 labeling gated on tetramer-positive cells, indicating no cell division during the course of the culture.
\(B) Survival of liver NKT cells from *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice. Duplicate samples of 5 × 10^4^ CD1d-reactive T cells/well were incubated for each condition and time point. Histograms represent the percentage of viable, CD1d-α-GalCer-tetramer-positive GFP^hi^ cells, as determined by flow cytometry at the indicated time points. The proportion of CXCL16-cultured cells from control *cxcr6^gfp/+^* mice was set at 100%. Values are mean ± standard deviation from three independent experiments.
\(C) Flow cytometry analysis of liver leukocyte apoptosis in culture. The percentage of cells binding Annexin V was determined by flow cytometry at the indicated time points. Results are representative of two independent experiments. A representative analysis at 10 h is shown on the right, with the percentage of CD1-αGC-tetramer-positive and Annexin V--positive cells indicated.
:::
:::
Decreased ConA-Induced Hepatitis and Reduced Sinusoid Patrolling by NKT Cells in CXCR6-Deficient Mice {#s2h}
-----------------------------------------------------------------------------------------------------
The reduction in the number of patrolling CD1d-reactive T cells in the CXCR6-deficient mice may be expected to decrease the efficacy of their effector functions. ConA-induced hepatitis, considered to be an experimental model for human autoimmune hepatitis \[[@pbio-0030113-b11]\], is strictly dependent on CD1d-reactive NKT cells \[[@pbio-0030113-b09],[@pbio-0030113-b10]\]. We injected ConA into *cxcr6^gfp/gfp^*, *cxcr6^gfp/+^*, and *cxcr6^+/+^* littermates and measured serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels at several time points. Elevated serum levels of AST and ALT were first observed 6--8 h after injection of 20 mg/kg of ConA and peaked at 12 h after injection (data not shown). As shown in [Figure 6](#pbio-0030113-g006){ref-type="fig"}A, serum AST and ALT levels were markedly reduced at 12 h in *cxcr6^gfp/gfp^* mice as compared with *cxcr6^gfp/+^* and *cxcr6^+/+^* littermates. Histological examination of livers from ConA-treated *cxcr6^gfp/+^* and *cxcr6^+/+^* mice showed severe bridging necrosis 12 h after injection in the area between the central veins and the portal tracts ([Figure 6](#pbio-0030113-g006){ref-type="fig"}B). Numerous red blood cells were also observed in the sinusoidal area. In *cxcr6^gfp/gfp^* mice, the histological features of hepatitis were markedly less severe than in the *cxcr6^gfp/+^* littermates ([Figure 6](#pbio-0030113-g006){ref-type="fig"}B). Thus, the reduced presence of CD1d-specific NKT cells in CXCR6-deficient mice was correlated with a decreased severity of acute hepatitis. Interestingly, the decrease in severity of hepatitis (an approximately 8-fold decrease in AST and ALT levels) appeared greater than the decrease in NKT density or visitation rate (3-fold). This is possibly due to cumulative effects of hepatic damage that tend to exacerbate each other, such as local damage leading to local ischemia, causing further damage.
::: {#pbio-0030113-g006 .fig}
Figure 6
::: {.caption}
###### Decreased Patrolling Efficiency and Decreased ConA Hepatitis Severity in CXCR6-Deficient Mice
\(A) Serum transaminase levels in *cxcr6^gfp/gfp^*, *cxcr6^gfp/+^*, and *cxcr6^+/+^* littermates 12 h after a challenge with 20 mg/kg ConA IP. Results are pooled from two independent experiments. The horizontal line represents the upper limit for normal serum transaminase levels, as measured in unchallenged mice. Asterisk indicates a two-tailed student\'s *t*-test with a *p* \< 0.05 in comparison with littermates.
\(B) Hematoxylin and eosin staining of paraffin-embedded liver sections from *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice sacrificed 12 h after a challenge with 20 mg/kg ConA IP.
\(C) Reduced sinusoid patrolling by NKT cells in *cxcr6^gfp/gfp^* mice. By measuring the average inter-nuclei distance of hepatocytes in high-magnification images, the average sinusoidal length of a hepatocyte was determined to be 28.6 μm (data not shown). Utilizing the lymphocyte velocity data (left panel) and assuming that a CD1d-reactive T cell can contact only one hepatocyte at a time, we calculated that each CD1d-reactive T cell can visit 0.58 hepatocytes/min in the *cxcr6^gfp/+^* mice, and 0.64 hepatocytes/min in the CXCR6-deficient mice. The density of GFP^+^ cells in *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice (middle panel) was calculated from the intravital videos used in [Figure 3](#pbio-0030113-g003){ref-type="fig"} and the flow cytometry data in [Figure 2](#pbio-0030113-g002){ref-type="fig"}. Combining the similar crawling velocities with the different steady-state densities, we show decreased rate of patrolling by GFP^+^ lymphocytes, expressed as the average time between "visitation" of any single hepatocyte (right panel).
:::
:::
The reduction in the number of CD1d-reactive T cells in the CXCR6-deficient mice might also be expected to decrease the efficiency of patrolling. Based on the results shown in [Figure 2](#pbio-0030113-g002){ref-type="fig"} and in the intravital videos (e.g., Videos [S2](#sv002){ref-type="supplementary-material"} and [S3](#sv003){ref-type="supplementary-material"}), we found little difference in the rate of surveillance of sinusoids by individual CD1d-reactive T cells in *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice. The cells in *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice "visited" 0.58 and 0.64 hepatocytes/min, respectively ([Figure 6](#pbio-0030113-g006){ref-type="fig"}C). However, because of the reduced density of CD1d-reactive T cells in *cxcr6^gfp/gfp^* mice, we estimate that each hepatocyte is visited only once every 47 min, as compared to every 15 min in *cxcr6^gfp/+^* mice. Therefore, the low number of CD1d-reactive T cells in the CXCR6-deficient mice clearly decreases the efficiency of patrolling.
Discussion {#s3}
==========
This study illuminates several important aspects of the biology of CD1d-reactive NKT cells in vivo and provides a new paradigm for effector lymphocyte action in the liver. We have shown that CD1d-reactive NKT cells patrol liver sinusoids, the vascular space of the liver, rather than extravasate like effector cells in other tissues. The recruitment of NKT cells to the vascular endothelium in the liver thus results in intravascular patrolling rather than cellular extravasation. Patrolling appears to be a search for antigen, since it ceases following antigen receptor signaling. The chemokine receptor CXCR6 regulates the number of hepatic NKT cells, at least in part by transmitting survival signals.
CD1d-Reactive NKT Cells Patrol Hepatic Sinusoids {#s3a}
------------------------------------------------
By replacing the murine *cxcr6* gene with the coding region for GFP, we were able to visualize CXCR6-expressing cells throughout the body. GFP^hi^ cells were mainly located intravascularly in mouse liver sinusoids, and by intravital fluorescence microscopy they were found to be crawling within the sinusoids at high speeds. This patrolling behavior appears to be independent of the direction of blood flow. Activation through the TCRs delivered a rapid stop signal to these cells, indicating that the patrolling is a form of immune surveillance.
The only prior indication for patrolling in the liver was the observation of cells crawling within the sinusoids when viewed by dark-field oblique transillumination microscopy \[[@pbio-0030113-b27]\]. These cells were thought at the time to be Kupffer cells (liver macrophages). However, closer examination suggests that two populations were observed: a phagocytic non-motile population, probably Kupffer cells, and a non-phagocytic, rapidly migrating cell type that probably corresponds to the CD1d-reactive NKT cells examined here, though other effector populations such as NK cells and cytotoxic T lymphocytes may also be capable of patrolling the liver from within the sinusoids.
The finding of TCR^+^ cells that survey tissue for antigen while located intravascularly is unprecedented. Immune surveillance within lymph nodes and the spleen occurs in specialized compartments densely packed with lymphocytes, which are shielded from the high flow rates and shear forces of blood. In contrast, within the liver, hepatocytes, sinusoid-lining endothelial cells, hepatic stellate cells (Ito cells), and Kuppfer cells are all in direct contact with blood. Hence, intravascular patrolling by T cells is likely to be the most efficient means for detection of hepatic antigens.
We estimate that, at steady state, the area around each hepatocyte is visited by approximately four CD1d-reactive T cells/h. In stark contrast, dendritic cells in lymph nodes were recently estimated to be scanned by 5,000 T cells/h \[[@pbio-0030113-b28]\]. However, NKT cells have a highly restricted TCR repertoire, whereas naïve T cells are extraordinarily diverse. Thus, discovery of antigen may in fact be more rapid in the liver than in the lymph nodes, despite a 1,000-fold difference in the visitation rate of antigen-presenting cells.
NKT cells may be particularly well suited for surveillance of the liver because of the extensive lipid metabolism in this organ by hepatocytes, Kupffer cells, hepatic stellate cells, and the sinusoid-lining endothelial cells. CD1d is expressed on hepatocytes, Kupffer cells, hepatic dendritic cells, and sinusoid-lining endothelial cells (see [Figure 3](#pbio-0030113-g003){ref-type="fig"}C), all of which have been demonstrated to have significant antigen-uptake capabilities \[[@pbio-0030113-b01]\]. It is unclear whether intravascular NKT cells could contact CD1d on hepatocytes, which are separated from the sinusoid by the sinusoid-lining endothelial cells. Direct contact might be achieved through the approximately 100-nm fenestrations typical of these endothelial cells. Alternatively, CD1d-reactive T cells may survey only the endothelium or Kupffer cells for inflammatory signals and/or for CD1d-bound endogenous ligands.
We have previously shown that CD4^+^ effector T cells stop migrating in response to antigen in vitro \[[@pbio-0030113-b24]\]. Here we have shown that NKT cells, essentially a population of effector T cells, display rapid and sustained stopping following TCR activation. In contrast, recent studies have shown that naïve T cells in lymph nodes continue to move at high speed for several hours after initial encounter with antigen-presenting dendritic cells, forming only short-lived conjugates, while only at later times after initial antigen encounter (8--20 h) did the T cells form long-term conjugates (\>1 h) with dendritic cells \[[@pbio-0030113-b29],[@pbio-0030113-b30]\].
The cytotoxic function of NKT cells has been implicated as necessary for ConA-induced hepatitis \[[@pbio-0030113-b09]\], although the observation that ConA triggers stopping of patrolling appears to argue against direct NKT-cell-mediated killing. NKT cell activation and subsequent cytokine production are known to recruit and activate large numbers of NK cells \[[@pbio-0030113-b31]\], which may instead be responsible for the direct killing. It is also possible that NKT cells resume patrolling after several hours of immotility, and in this way regain access to all the hepatocytes of the liver for direct killing. Further experiments will be required to resolve these questions.
CXCR6 Controls the Accumulation of CD1d-Reactive NKT Cells in the Liver {#s3b}
-----------------------------------------------------------------------
Our study shows that CXCR6, which is expressed on CD1d-reactive NKT cells, controls the selective accumulation of these cells in the liver. Accordingly, CXCR6 also influences the ability of CD1d-reactive NKT cells to induce hepatitis caused by ConA. According to the multistep paradigm of lymphocyte extravasation, chemokines play a key role in activating adhesion molecules that transform rolling cells to firmly adhered ones. Because we observed that CXCL16, the chemokine ligand for CXCR6, is expressed on sinusoidal endothelial cells, we initially hypothesized that CXCR6 was critical for the recruitment of blood-borne NKT cells to the hepatic sinusoids, either by initial tethering interactions or by signaling to initiate crawling and patrolling. Utilizing high-frame-rate imaging, such recruitment events were occasionally observed (see [Video S8](#sv008){ref-type="supplementary-material"}). Many hours of observation showed no significant difference in the frequency of these recruitment events between *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice (data not shown), but the limited number of events and mouse-to-mouse variability preclude definitive conclusions regarding the role of CXCR6 in this process. In addition, we did not observe any significant deficit in hepatic recruitment of CXCR6-deficient cells after short-term transfer of T cell blasts from *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice and after in vivo expansion of NKT cells with α-GalCer treatment (data not shown). These results suggest the existence of redundant mechanisms for NKT cell recruitment to liver sinusoids. We have therefore been unable to support the hypothesis that a defect in the NKT cell recruitment process contributes to the phenotype observed at steady state in CXCR6-deficient animals.
Our results indicate that the CXCR6-deficient hepatic NKT cells patrol in the same manner as their CXCR6-expressing counterparts: they both crawl rapidly, lack directional bias, and rapidly stop in response to TCR stimulation. Thus, there is no defect in the patrolling behavior of CXCR6-deficient NKT cells that can explain their reduced numbers in the liver.
The reduced accumulation of CXCR6-deficient CD1d-reactive NKT cells in the liver is likely to result at least partially from increased cell death due to the lack of a CXCR6-mediated survival signal. Because CXCL16 is the only known ligand for CXCR6 and is expressed on liver sinusoids, it is reasonable to hypothesize that CXCR6/CXCL16 interactions result in enhanced survival of CD1d-reactive liver NKT cells. This would be a novel role for chemokines in the homeostasis of effector T cells in peripheral tissues. As CXCR6 is expressed not only on NKT cells, but also on activated and memory T cells, it is possible that this is a general mechanism by which chemokines mediate the survival of effector/memory T cells that patrol at sites of potential toxic damage and antigen entry, thus facilitating rapid and efficient memory T cell responses.
The nature of the survival signal provided by CXCR6 remains to be studied. Although unexpected, this behavior is not unprecedented, as CXCR4 and CX~3~CR1 have been reported to mediate survival in certain conditions \[[@pbio-0030113-b32],[@pbio-0030113-b33],[@pbio-0030113-b34]\]. Reports have suggested that Akt activation following ligand binding to the chemokine receptors CXCR4 and CX~3~CR1 results in enhanced survival of cells \[[@pbio-0030113-b32],[@pbio-0030113-b33]\], and it has recently been shown that CXCR6 can similarly activate Akt \[[@pbio-0030113-b35]\]. It is also noteworthy that the defect in CXCR6 expression only affects hepatic NKT cells, and not NKT cells in other locations, suggesting that either the liver is an especially inhospitable environment, or that other survival signals are provided in other NKT-rich organs such as the spleen. The former seems a particularly attractive hypothesis since there is evidence for the liver being a pro-apoptotic environment for lymphocytes \[[@pbio-0030113-b01]\].
The results of this paper lead us to a model in which the number of intravascular patrolling NKT cells, which is regulated by CXCR6--CXCL16 interactions, influences the immune response in liver in two important ways: (1) it determines the frequency of new antigen detection by affecting the visitation rate of parenchymal liver cells, and (2) it governs the total cytokine "power" (pooled secretion capacity) of the hepatic NKT cell population. In the case of the ConA hepatitis model, the first factor is non-applicable because the "antigen" is in excess. However, in infectious pathogen models in which rare hepatocytes may be infected, there are likely to be agonist glycolipids that react with only a small subset of NKT cells and thus may escape detection for hours, or even days, if NKT cell density is too low. Potential examples include the recent finding that small subsets of NKT cells recognize the mycobacterial cell-wall antigen PIM~4~ \[[@pbio-0030113-b36]\] and lipophosphoglycan of leishmania \[[@pbio-0030113-b37]\].
In conclusion, we have described a model that may explain how the immune system monitors the status of the liver. Future experiments will be required to identify the molecules involved in both NKT cell crawling and stopping processes and the cells that are surveyed by NKT cells, and to determine the relevance of this system for the pathogenesis of hepatitis and for various immune responses in the liver.
Materials and Methods {#s4}
=====================
{#s4a}
### Animals {#s4a1}
*cxcr6^gfp/+^* knock-in mice were generated as described \[[@pbio-0030113-b20]\] and backcrossed three to eight times onto C57BL/6. Tie2-GFP mice on the FVB/NJ background and *tcrα^−/−^* mice on the C57BL/6 background were purchased from Jackson Laboratory (Bar Harbor, Maine, United States) and Taconic Farms (Germantown, New York, United States), respectively. All mice were maintained in our specific-pathogen-free animal facility according to institutional guidelines, and experiments were done with mice from 4 to 12 wk of age.
### Reagents {#s4a2}
CD1d tetramer loaded with α-GalCer was prepared as described previously \[[@pbio-0030113-b06]\]. The α-GalCer compound DB01--1 was kindly provided by Steven Porcelli. The CXCL16-Fc fusion protein was a kind gift of J. Cyster. Recombinant murine CXCL16, SDF-1, and fractalkine were purchased from R & D Systems (Minneapolis, Minnesota, United States). ConA was purchased from Sigma-Aldrich (St. Louis, Missouri, United States). The following monoclonal antibodies were purchased from PharMingen (San Diego, California, United States): H57--597-PE and -APC (anti-TCRβ); PK136-PE, -biotin, and -PE (anti-NK1.1); IM7-biotin (anti-CD44); and 2.4G2 (anti-FcγRII/III). Rabbit polyclonal serum against caveolin-1 was purchased from Transduction Laboratories (Lexington, Kentucky, United States). Goat F(ab′)2 anti-human IgG Fc conjugated to Cy5, goat F(ab′)2 anti-rat IgG Fc conjugated to Cy3, and goat anti-rabbit Ig conjugated to Cy3 were purchased from Jackson ImmunoResearch (West Grove, Pennsylvania, United States). Anti-CD1d antibody (clone 1B1) was purchased from PharMingen and conjugated to Cy5 succinimide ester (Amersham Biosciences, Little Chalfont, United Kingdom) at a final ratio of 1.7 fluorophores/antibody. Cell Tracker 633 (Bodipy 630/650 MeBr) was purchased from Molecular Probes (Eugene, Oregon, United States).
### Production of anti-murine CXCL16 antibodies {#s4a3}
Female Wistar-Kyoto rats, 6--8 wk old, were immunized intraperitoneally with 100 μg of murine CXCL16. Immunizations were performed with 100 μg of protein emulsified in incomplete Freund\'s adjuvant at 2-wk intervals. After a minimum of three immunizations, the rats were boosted with 100 μg of soluble CXCL16 protein in PBS. Three days post-boost, the spleens were harvested and fused with SP2/0 myeloma cells. The fusion was screened by ELISA using plates coated with CXCL16 protein. Specificity of the hybridomas was determined by ELISA with a panel of murine chemokines including CXCL11, CCL22, CCL5, CCL27, CCL28, CCL17, CXCL10, CCL25, and CXCL9. Hybridomas producing CXCL16-specific antibodies were then subjected to three rounds of subcloning by limiting dilution. Biological activity was confirmed by FACS and inhibition of chemotaxis.
### Isolation and staining of lymphocytes from mouse tissues {#s4a4}
Spleen tissue was minced and mashed through a 70-μm strainer in PBS with 0.5% BSA. The resulting suspension was pelleted by centrifugation, re-suspended in PBS with 0.5% BSA, layered on Ficoll-paque (Pharmacia LKB Technology, Uppsala, Sweden), and centrifuged at 400 *g* for 20 min. Blood was collected in heparinized tubes by cardiac puncture of anesthetized animals. Bone marrow was obtained by flushing femurs with PBS containing 0.5% BSA using a 26 G needle. Cells in the resulting pellets were treated with tris-ammonium chloride to remove red blood cells and then washed extensively before use. Lung or liver tissue was minced and mashed through a 70-μm strainer in PBS containing 0.5% BSA. The resulting suspension was pelleted by centrifugation, re-suspended in 40% Percoll (Pharmacia LKB Technology), layered on 80% Percoll, and centrifuged at 600 *g.* Cells at the gradient interface were harvested and washed extensively before use. Alternatively*,* liver mononuclear cells were recovered by perfusion of liver of anesthetized mice as follows: the diaphragm and the sus-hepatic vein were cut and the liver reclined in a Petri dish; perfusion with cold PBS and 2 mM EDTA was through the aorta, and the effluent was collected through the severed sus-hepatic vein (approximately 5 ml).
### Staining of CD1d on parenchymal hepatic cells {#s4a5}
Fluorescence microscopy of fixed liver sections was performed as described above for CXCL16 analysis. For flow cytometry analysis, liver tissue was mashed with the back of a 5-ml syringe plunger, digested in PBS containing Mg and Ca by 0.2 mg/ml Collagenase D (Roche, Basel, Switzerland) and 0.02 mg/ml DNase I (Roche) for 30 min at 37 °C while agitating, filtered through a 70-μm strainer, pelleted in 40% Percoll (Pharmacia LKB Technology), cleared of red blood cells by tris-ammonium chloride solution, and washed with cold PBS. This suspension of hepatic parenchymal cells was blocked with goat and bovine serum (2% each) and anti-CD16 antibody (PharMingen) for 20 min on ice and then stained with CD31-PE or Tie2-PE (eBioscience, San Diego, California, United States), F4/80-PE (Serotec, Oxford, United Kingdom), or CD11c-PE (PharMingen) along with anti-murine CD1d-Cy5 or rat IgG2b-Cy5. After 30 min on ice, cells were washed twice and analyzed by flow cytometry.
### Thymocyte transfer experiment {#s4a6}
Thymocytes were isolated from 4-wk-old *cxcr6^+/+^*, *cxcr6^gfp/+^*, and *cxcr6^gfp/gfp^* littermates. *tcrα^−/−^* recipient mice (devoid of TCRαβ lymphocytes) were grafted IV with 30 × 10^6^ thymocytes from a 50:50 mix of cells (*cxcr6^+/+^*:*cxcr6^gfp/+^* and *cxcr6^+/+^*:*cxcr6^gfp/gfp^*). After 2 and 3 d, recipient mice were sacrificed, and liver cell suspensions were prepared and stained with antibodies against TCRβ-APC, and NK1.1-PE or CD1d-*α*-GalCer-tetramer-PE. After exclusion of dead cells with PI, cells present in the TCRβ^+^ gate were analyzed for staining with NK1.1 or CD1d tetramer and GFP.
### ConA hepatitis {#s4a7}
During preliminary time-course and dose-escalation experiments using wild-type and TCRα KO 6-wk-old C57BL6 mice, a dose of 20 mg/kg ConA IP was determined to induce acute liver disease with a peak in serum transaminases AST and ALT at 12 h. Groups of 6-wk-old *cxcr6^+/+^*, *cxcr6^gfp/+^*, and *cxcr6^gfp/gfp^* mice were therefore treated with 20 mg/kg ConA diluted in PBS, blood was sampled by tail bleeding after 12 h, serum was aliquoted and stored at −20 °C, and transaminases were measured using a Vitros 950 (Ortho-Clinical Diagnosis, Mississauga, Ontario, Canada).
### In vitro survival/proliferation assay {#s4a8}
Liver leukocytes were isolated from 8- to 12-week-old *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* littermates. Equal numbers of CD1d-α-GalCer-tetramer-PE-positive T cells, as measured by flow cytometry, were cultured in RPMI 1640 medium with 10% FCS, 1% penicillin/streptomycin, and beta-mercapto-ethanol on 96-well plates coated sequentially with 0.5 μg/well anti-CD3 antibody (2C11, PharMingen) and/or 0.5 μg/well recombinant CXCL16 diluted in PBS or with PBS alone. The number of GFP^hi^ CD1d-α-GalCer-tetramer-PE-positive T cells, surface-stained with Annexin V (PharMingen), were determined in duplicate, at the indicated time points, by flow cytometry. The number of GFP^hi^ CD1d-α-GalCer-tetramer-PE-positive T cells in the sample from *gfp/+* mice cultivated in wells coated with CXCL16 was considered to be 100%.
### Confocal microscopy {#s4a9}
Livers were dissected and removed, washed in PBS, sliced, and fixed for 45min at 4 °C in 4% paraformaldehyde. Liver slices were washed with PBS and incubated overnight at 4 °C in 30% sucrose, and then washed again in PBS, placed in OCT medium, and frozen. Sections with a thickness of 7 μm were stained with rabbit polyclonal anti-caveolin-1 serum followed by goat anti-rabbit antibody conjugated to Cy3, with rat monoclonal IgG2b anti-CXCL16 followed by goat anti-rat antibody conjugated to Cy3, and with control serum and isotype control antibodies followed by the corresponding second-step reagents. Slides were mounted with Fluoprep (Biomerieux SA, Marcy l\'Etoile, France) and analyzed with a confocal laser system (LSM 510, Zeiss, Jena, Germany).
### Intravital confocal microscopy {#s4a10}
Surgical preparation for liver imaging was based on methods described previously \[[@pbio-0030113-b38]\]. Mice were anesthetized using a cocktail of ketamine (50 mg/kg), xylazine (10 mg/kg), and acepromazine (1.7 mg/kg) injected intraperitoneally. Beginning 45 min later, anesthesia was maintained by half-dose boosts subcutaneously every 30 min. Hair from the left subcostal region was trimmed, and the liver was exposed through a 1.5-cm horizontal incision. The hepatoform ligament was cut and the tip of the left lobe of the liver gently extruded. The mouse was inverted onto a pre-prepared plastic or aluminum tray in which a coverslip was mounted near the center and narrow strips of paper (1.5 mm × 1.5 cm) were glued. These strips of paper provided friction that helped to immobilize the tissue being imaged. Images were acquired using an inverted epifluorescence Zeiss LSM 510 confocal laser system equipped with a 10×/0.3 Plan Neofluar objective, or Fluar 40×/1.3 objective. A warming fan blowing into an enclosure around the whole microscope was used to keep the area warm, and the microscope objective was thermostatically controlled to maintain 37 °C. Videos were acquired by consecutive frames using appropriate combinations of 488-nm, 546-nm, and 633-nm laser lines and GFP, Cy3, and Cy5 filter sets. Imaging speed varied between videos with a range from 1 to 30 s per time point. During all of our intravital imaging experiments, we were keenly aware of the possibility of local photo-toxicity causing time-dependent artifacts in our data. However, we were unable to detect, either by eye or by quantitative analysis, any indication of such a phenomenon.
### In vivo activation and imaging of NKT cells {#s4a11}
Animals were imaged as described above. Antibodies and antigens were injected in 50 μl of PBS solution containing 100 μg of 70-kDa dextran conjugated to tetramethylrhodamine (Molecular Probes) to confirm intravenous delivery and healthy blood flow in the region being imaged. We injected 5 μg of anti-CD3ɛ (2C11 clone, PharMingen), 250 μg of Alexa-633-conjugated ConA (Molecular Probes), or 0.05--5 μg of α-GalCer (DB01--1 compound, kind gift of S. Porcelli). Initial experiments utilized 5 μg of α-GalCer, but as little as 50 ng was sufficient to trigger identical amounts of NKT stopping. Doses of either 200 or 50 ng of α-GalCer were used in the GK1.5/CD1d blocking experiments shown here. The percentages of immobile cells before and 40 min after injection of antigen were assessed by manual observation: 6 min of video was analyzed, and cells that moved less than one cell-body-length (10 μm) were considered immobile.
### Analysis of intravital imaging videos {#s4a12}
Quantitative analysis was performed only on videos in which there was no detectable whole-liver movement. Cells were tracked using Volocity software version 2.0 (Improvision, Lexington, Massachusetts, United States). Because the software lost some cells due to cells moving out of focus or coming close to one another, the resulting data were an array of cell paths ranging from single frames to the full length of the video (typically 10--15 min). No effort was made to reassemble these partial tracks, but there appeared to be no significant difference in the distribution of track lengths between *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice. In total, 640 tracks were analyzed from four videos of *cxcr6^gfp/+^* mice, and 574 tracks were analyzed from four videos of *cxcr6^gfp/gfp^* mice. Values such as cell velocity, overall displacement, and displacement-to-path-length ratio were calculated for each track by manipulation of Volocity output in spreadsheets. To safeguard against possible flaws or biases in the Volocity software, data were analyzed according to various subsets to look for unexpected trends (such as shorter average path lengths in particular videos) that might cause the data to be biased. Furthermore, all Volocity-based conclusions, such as velocity and directedness, we confirmed by manual observation of a limited number of cells.
To calculate the visitation rate, we calculated the density of GFP^+^ cells in *cxcr6^gfp/+^* from the intravital videos used in [Figure 3](#pbio-0030113-g003){ref-type="fig"}B and [3](#pbio-0030113-g003){ref-type="fig"}C (see Videos [S2](#sv002){ref-type="supplementary-material"} and [S3](#sv003){ref-type="supplementary-material"}). Using intravital images to compare the density of GFP^+^ cells was deemed to be inaccurate because of the relatively low number of events counted and was likely to cause a bias in the choice of fields for collecting videos. Thus, the density of cells in *cxcr6^gfp/gfp^* mice was calculated from the density for *cxcr6^gfp/+^* mice and the relative frequency of cells in *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* mice as quantified by flow cytometry (see [Figure 2](#pbio-0030113-g002){ref-type="fig"}). Using the hepatocyte nuclear exclusion of mitochondrial autofluorescence clearly visible in the higher magnification images of [Figure 3](#pbio-0030113-g003){ref-type="fig"}A, we calculated hepatocytes to be 28 μm long along the sinusoids, at a density of 1,142/mm^2^. Utilizing the velocity data and assuming that a CD1d-reactive T cell can contact only one hepatocyte at a time, we estimated the number of hepatocyte areas visited by CD1d-reactive T cells per minute in the *cxcr6^gfp/+^* mice and in the *cxcr*6-null mice.
Supporting Information {#s5}
======================
Video S1
::: {.caption}
###### NKT Cells Patrol Hepatic Sinusoids
High magnification (40×) intravital video of GFP^+^ cells crawling along the hepatic sinusoids of a *cxcr6^gfp/+^* animal. Video is 300-fold compressed in time.
(6.1 MB AVI).
:::
::: {.caption}
######
Click here for additional data file.
:::
Video S2
::: {.caption}
###### NKT Cells Patrol Similarly in *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* Animals I
Low magnification (10×) intravital video of *cxcr6^gfp/+^* liver showing the typical pattern of crawling observed in each genotype. Video is 300-fold compressed in time.
(9.2 MB AVI).
:::
::: {.caption}
######
Click here for additional data file.
:::
Video S3
::: {.caption}
###### NKT Cells Patrol Similarly in *cxcr6^gfp/+^* and *cxcr6^gfp/gfp^* Animals II
Low magnification (10×) intravital video of *cxcr6^gfp/gfp^* liver showing the typical pattern of crawling observed in each genotype. Video is 300-fold compressed in time.
(2.3 MB AVI).
:::
::: {.caption}
######
Click here for additional data file.
:::
Video S4
::: {.caption}
###### Various Examples of Crawling Patterns of Hepatic NKT Cells
Assortment of clips from various videos displaying examples of NKT cells passing each other in single sinusoids, or turning around within a single sinusoid. These examples suggest that NKT cells are not responding to spatial molecular gradients and are capable of crawling both with and against the direction of blood flow. All clips were 300-fold compressed in time.
(7.9 MB AVI).
:::
::: {.caption}
######
Click here for additional data file.
:::
Video S5
::: {.caption}
###### Real-Time Imaging of Blood Flow
Low-magnification (10×) intravital video of the liver of a wild-type C57BL/6 animal during which fluorescent dextran was injected intravenously (red). Video is shown with no compression in time (i.e., at real time).
(779 KB AVI).
:::
::: {.caption}
######
Click here for additional data file.
:::
Video S6
::: {.caption}
###### NKT Cells Stop Patrolling upon Stimulation by ConA
Low-magnification (10×) intravital video of the liver of a *cxcr6^gfp/+^* animal during which 250 μg of ConA was injected intravenously. Video is 300-fold compressed in time.
(4.5 MB AVI).
:::
::: {.caption}
######
Click here for additional data file.
:::
Video S7
::: {.caption}
###### NKT Cells Stop Patrolling upon Stimulation by α-GalCer
Low-magnification (10×) intravital video of the liver of a *cxcr6^gfp/gfp^* animal during which 5 μg of α-GalCer was injected intravenously. Video is 300-fold compressed in time.
(7.5 MB AVI).
:::
::: {.caption}
######
Click here for additional data file.
:::
Video S8
::: {.caption}
###### Recruitment of Blood-Borne NKT Cell to Hepatic Sinusoidal Endothelium
Low-magnification (10×) intravital video of the liver of a *cxcr6^gfp/gfp^* animal taken at a high frame rate (1 frame/s). The arrival, initial rolling, and subsequent attachment and crawling of a single NKT is shown.
(5.4 MB AVI).
:::
::: {.caption}
######
Click here for additional data file.
:::
Accession Numbers {#s5a9}
-----------------
The LocusLink (<http://www.ncbi.nlm.nih.gov/LocusLink>) accession number for the CXCR6/Bonzo/STRL33 chemokine receptor is 80901and for its CXCL16 ligand is 66102.
We thank S. Jung, G. Eberl, U. von Andrian, U. Frevert, P. Kubes, and P. Askenase for stimulating discussions, M. J. Sunshine for assistance with the mouse colony, Steve Porcelli for α-GalCer, and Jason Cyster for the CXCL16-Fc fusion protein. Research support includes grants from the National Institutes of Health (RO1 AI33856 to DRL, RO1 AI55037 to MLD, and RO1 CA52511 to MK), the Irene Diamond Professorship in Immunology (MLD), and the Howard Hughes Medical Institute (DRL), a Human Frontier Science Program long-term fellowship (FG), and a Cancer Research Institute fellowship (TC).
**Competing interests.** The authors have declared that no competing interests exist.
**Author contributions.** FG, TOC, MLD, and DRL conceived and designed the experiments. FG and TOC performed the experiments. FG, TOC, MLD, and DRL analyzed the data. SS, NM, MK, and MJB contributed reagents/materials/analysis tools. FG, TOC, MK, MLD, and DRL wrote the paper.
Citation: Geissmann F, Cameron TO, Sidobre S, Manlongat N, Kronenberg M, et al. (2005) Intravascular immune surveillance by CXCR6^+^ NKT cells patrolling liver sinusoids. PLoS Biol 3(4): e113.
α-GalCer
: α-galactosyl ceramide
ALT
: alanine aminotransferase
AST
: aspartate aminotransferase
ConA
: Concanavalin A
GFP
: green fluorescent protein
NK
: natural killer
NKT cell
: natural killer T cell
PBS
: phosphate-buffered saline
TCR
: T-cell antigen receptor
|
PubMed Central
|
2024-06-05T03:55:55.038543
|
2005-4-5
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1073691/",
"journal": "PLoS Biol. 2005 Apr 5; 3(4):e113",
"authors": [
{
"first": "Frederic",
"last": "Geissmann"
},
{
"first": "Thomas O",
"last": "Cameron"
},
{
"first": "Stephane",
"last": "Sidobre"
},
{
"first": "Natasha",
"last": "Manlongat"
},
{
"first": "Mitchell",
"last": "Kronenberg"
},
{
"first": "Michael J",
"last": "Briskin"
},
{
"first": "Michael L",
"last": "Dustin"
},
{
"first": "Dan R",
"last": "Littman"
}
]
}
|
PMC1073692
|
Introduction {#s1}
============
Many current drugs were originally discovered through observation of unexpected biological activities (e.g., penicillin, benzodiazepines, sildenafil \[Viagra\]). Broad screens for biological function have the advantage of identifying the best "lock" for each new "key" produced by chemical variation. In contrast, the search for drug-like hits by high-throughput approaches is dominated by in vitro single-enzyme activity--based screens and single-readout cell-based assays. These approaches measure very limited regions of biological space and do not reveal potent effects on pathways not being measured directly. In order to systematize the understanding of the full activity of new small molecules, we quantified dose-dependent morphological changes induced in five cell types, thereby identifying "hit" compounds with unique activities.
The assay is based on the principle that many cellular targets are involved in the control of cellular morphology, DNA content and location, and morphology of the Golgi apparatus (\[[@pbio-0030128-b01],[@pbio-0030128-b02],[@pbio-0030128-b03]\]; C. L. Adams, D. A. Coleman, G. Cong, A. M. Crompton, K. A. Elias, et al., unpublished data). Cell-type-specific components are known to utilize distinct pathways and cellular programs to control fundamental processes affecting the features of the organelles and the overall cellular morphology. Five cell types (lung adenocarcinoma, ovarian cancer, a neuronal glioma, a prostate cancer, and endothelial cells) were included in the morphological screen. The approach has been validated by analysis of known pharmacologically active compounds from ten different mechanism of action classes (actin inhibitors, calmodulin antagonists, endoplasmic reticulum Ca^2+^ ATPase inhibitors, geranylgeranyl transferase-1 inhibitors, G-protein-coupled receptor activators, protein kinase C activators, topoisomerase II inhibitors, tubulin destabilizers, tubulin stabilizers, and kinase inhibitors). In every case, a high percentage of the compounds were accurately classified into the ten different mechanism of action groups using the Cytometrix (TM) system (C. L. Adams, D. A. Coleman, G. Cong, A. M. Crompton, K. A. Elias, et al., unpublished data).
A screen of 107 small molecules comprising four different chemical scaffolds known to inhibit protein kinases with varying selectivity and potency were selected for the Cytometrix screen. In this report, we focus on a hydroxyl-substituted analog, 3-(1-*tert*-butyl-4-amino-1H-pyrazolo\[3,4-d\]pyrimidin-3-yl)phenol (hydroxy-PP), of the known Src-family protein kinase inhibitor 1-*tert*-butyl-3-(4-chlorophenyl)-1H-pyrazolo\[3,4-d\]pyrimidin-4-amine (PP2), because it exhibited a cell response profile distinct from the known kinase inhibitors including the closely related compound PP2. Although the compound collection was dominated by kinase-inhibitor scaffolds, we identified a nonkinase target of hydroxy-PP, carbonyl reductase 1 (CBR1), an NADPH-dependent reductase. Hydroxy-PP and analogs chosen by structure-based design were used to search both for signaling pathways in which CBR1 may be involved and for potential therapeutic uses of CBR1 inhibitors.
Results {#s2}
=======
Selection of Chemical Library {#s2a}
-----------------------------
A collection of 107 compounds containing known protein kinase inhibitors and close structural analogs were screened in the Cytometrix assay for unique phenotypic profiles suggestive of potent inhibition of cellular targets not affected by known protein kinase inhibitors ([Figure 1](#pbio-0030128-g001){ref-type="fig"}A). The well-characterized protein kinase inhibitors (K252a \[[@pbio-0030128-b04]\], SKB203580 \[[@pbio-0030128-b05]\], VK-1911 \[[@pbio-0030128-b06]\], and PP2 \[[@pbio-0030128-b07]\]) served as positive controls and "landmarks" for the phenotypes likely to be induced by the less-characterized compounds in the collection ([Figure S1](#sg001){ref-type="supplementary-material"} contains a complete list of all compounds tested). An advantage of screening compounds closely related to each other is the availability of a wealth of structure--activity relationships in the initial screen that provide a guide to follow-up studies aimed at improving affinity and target selectivity in a second round of chemical synthesis.
::: {#pbio-0030128-g001 .fig}
Figure 1
::: {.caption}
###### Cell Morphology--Based Screen for Biologically Active Small Molecules
\(A) Steps in the drug-screening process. Five human cell types, including one primary and four cancer cell lines, were treated for 24 h with the screening library that included compounds of known function and related analogs. The Cytometrix (TM) data analysis package was used to analyze microscopy data for each treatment condition.
\(B) PCA plot of the phenotypic attributes. Colored spheres represent a single compound at one concentration (ranging from 6 nM to 40 μM by 3-fold increases); lines connecting the spheres indicate a single compound\'s effects over a range of concentrations. Spheres are colored as follows: known protein kinase inhibitors (blue), paclitaxil (green), and novel compounds structurally related to the protein kinase inhibitors (red). The PCA provides aggregate variables termed "components" made up of multiple independent variables, each with a "loading factor." These values are provided in [Table S2](#st002){ref-type="supplementary-material"}. Structures of each compound are given in [Figure S1](#sg001){ref-type="supplementary-material"}.
\(C) Structures of the known kinase inhibitors PP and PP2 (blue), as well as the novel "hit compound" hydroxy-PP (red), are shown. Linker analogs of PP (PP-L) and hydroxy-PP (hydroxy-PP-L) that were used to ascertain the functional tolerance of replacing the *t*-butyl substituent at N-1 of the "hit compound" hydroxy-PP are shown.
\(D) Morphological attribute tabulation for cells treated with 129.4 μM PP2 (blue lines) or 0.4 μM hydroxy-PP (red lines) in each of five cell types. Data for the x-axis is grouped by the probe used (α-tubulin antibody, Hoechst dye, and lectin stain). Each of 14 attributes contributing to the magnitude of the response (y-axis) is shown as a red-filled square.
\(E) Visual morphology of A549 cells treated with hydroxy-PP or PP2. Hoechst dye or α-tubulin antibody was used to stain cells. The PP2-treated cells are more elongated and have more a condensed nuclear structure as compared with hydroxy-PP-treated cells.
:::
:::
Phenotypic Profiling Using Microscopy and Automated Image Analysis {#s2b}
------------------------------------------------------------------
Cellular and organelle morphology changes were measured from segmented images of cells stained with DNA and microtubule markers using algorithms that identify cell and nuclear boundaries (C. L. Adams, D. A. Coleman, G. Cong, A. M. Crompton, K. A. Elias, et al., unpublished data). Combining segmentation and intensity distribution algorithms allows acquisition of multiple shape-, texture-, and intensity-related features for each image collected. For each object identified by the segmentation algorithms, collected attributes include object location, area, perimeter, and axis ratio, as well as pixel-intensity sum, mean, variance, and kurtosis (the degree of peakedness of a distribution).
Cells undergo major morphological changes in the course of cell-cycle progression \[[@pbio-0030128-b08]\]. To separate these changes from ones induced by compound treatment, algorithms were used to classify cells by their cell-cycle status based on the DNA content, morphology, and condensation status. The multiple attributes of individual cells are summarized by a set of statistics that describe distributions of these attributes in a population of cells. These statistics are termed "phenotypic attributes." The attributes used to characterize the screening compounds are listed in [Table S1](#st001){ref-type="supplementary-material"}. These attributes were chosen for their biological information content and their low correlation with each other (C. L. Adams, D. A. Coleman, G. Cong, A. M. Crompton, K. A. Elias, et al., unpublished data).
Phenotypic Landmarks: Compounds with Known Mechanisms of Action {#s2c}
---------------------------------------------------------------
Principle component analysis (PCA) was used to reduce the dimensionality in the dataset to allow visual investigation of patterns in the multivariate signature. This transformation converts a number of correlated variables into a smaller number of uncorrelated variables, or principal components, in such a way that the first few components account for as much of the variability in the dataset as possible \[[@pbio-0030128-b09]\]. Importantly, component values are not physical constants but are dependent on the relative "spread" of the attributes derived from all the images in the dataset.
The plot of principle components 1, 2, and 3 derived from image analysis of the effects of 107 compounds (at eight different concentrations ranging from 6 nM to 40 μM) on five cell types is shown in a scatter plot ([Figure 1](#pbio-0030128-g001){ref-type="fig"}B). In this analysis the morphological features making up the individual components are given in [Table S2](#st001){ref-type="supplementary-material"}. We observed four distinct "phenotypes" induced by various members of the compound collection. At the lowest concentrations tested, compounds cluster with the dimethyl sulfoxide (DMSO) and untreated controls. This region, approximately in the center of the PCA plot, constitutes the attribute profile characteristic of no effect on cellular morphology.
The remaining three phenotypic categories are characterized by compounds that exhibit distinct "trajectories" such that at increasing concentrations different attributes are enhanced, indicative of a measurable dose-dependent phenotype. The cells treated with the microtubule-polymerizing natural product paclitaxel exhibit a pronounced and highly reproducible trajectory and constitute the second phenotype observed. Each of the five cell types exhibited reproducible changes in attributes associated with tubulin staining and cell-cycle status, as expected for a microtubule-polymerizing agent \[[@pbio-0030128-b10]\].
A third characteristic phenotype is exemplified by the potent general kinase inhibitor K252a. The bisindolocarbazole K252a is a potent inhibitor of over 50 known kinases from diverse families \[[@pbio-0030128-b04]\]. This compound induced dose-dependent morphological changes in A549, DU145, HUVEC, and SF268 cells but not SKOV3 cells (data not shown). We attribute the majority of the observed changes in morphological attributes to the kinase-inhibitory activity of K252a, rather than to any off-target effects, because K252a analogs (compounds 103--107 in [Figure S1](#sg001){ref-type="supplementary-material"}) that were lacking the ability to inhibit protein kinases \[[@pbio-0030128-b11]\] clustered with the "no phenotype" controls. That the K252a-induced phenotype was caused by inhibition of cellular kinases is consistent with the clustering of the K252a trajectory with trajectories of other known kinase inhibitors (PP2 and SKB203580). Although the kinase targets of each of these compounds are not fully characterized, the fact that the compounds share a similar profile in the Cytometrix assay suggests that they have overlapping targets (i.e., PP2 and SKB203580 inhibit p38), as has been reported elsewhere \[[@pbio-0030128-b12]\].
The fourth phenotype was produced by a structurally related pyrazolopyrimidine in the collection, hydroxy-PP (compound 87 in [Figure S1](#sg001){ref-type="supplementary-material"}) ([Figure 1](#pbio-0030128-g001){ref-type="fig"}C). The cellular attributes characteristic of this compound are distinct from the phenotypes induced by kinase inhibitors and the microtubule depolymerization inhibitors ([Figure 1](#pbio-0030128-g001){ref-type="fig"}B). In order to best distinguish the cellular attributes unique to hydroxy-PP, a close structural analog, PP2, was used as a reference. Hydroxy-PP bears a *meta*-OH substituent, whereas PP2 has a *para*-Cl substituent on the C-3 phenyl ring \[[@pbio-0030128-b07]\]. Other close analogs of PP2 ([Figure 1](#pbio-0030128-g001){ref-type="fig"}C), such as 1-*tert*-butyl-3-phenyl-1H-pyrazolo\[3,4-d\]pyrimidin-4-amine (PP; compound 1 in [Figure S1](#sg001){ref-type="supplementary-material"}), that lack any substituent on the C-3 phenyl ring showed a phenotypic profile comparable to PP2. (For this reason we used PP2 and PP interchangeably.) These structure--activity relationships suggested that the *meta*-OH substituent was critical for interaction with a protein target that is not affected by PP2.
The cellular effects induced by hydroxy-PP and PP2 were further examined by analysis of the quantitative attribute changes at approximately the EC~50~ value for each compound (as judged by distance in PCA space from the DMSO controls; [Figure 1](#pbio-0030128-g001){ref-type="fig"}B). [Figure 1](#pbio-0030128-g001){ref-type="fig"}D shows that the two compounds hydroxy-PP and PP2 exhibit indistinguishable effects on DU145, SF268, and SKOV3 cells. However, the two compounds exhibit cellular activities that are distinct from each other in A549 lung adenocarcinoma and HUVEC cells. Examples of the cell images that were analyzed by the Cytometrix algorithms and are directly relevant to the morphological differences induced by PP2 or hydroxy-PP are shown in [Figure 1](#pbio-0030128-g001){ref-type="fig"}E. The PP2-treated A549 cells appear more elongated and slightly more condensed than the hydroxy-PP2-treated cells, leading to the quantitative differences plotted in [Figure 1](#pbio-0030128-g001){ref-type="fig"}D. That only two of five cell types exhibited differential responses to two closely related molecules highlights the importance of including cells representing a diversity of tissue sources and genetic makeup in order to explore a wide range of possible small-molecule targets (C. L. Adams, D. A. Coleman, G. Cong, A. M. Crompton, K. A. Elias, et al., unpublished data). Attempts to directly assign the molecular target or targets responsible for the hydroxy-PP-induced changes to morphological features using compounds previously profiled using the Cytometrix™ system were not possible because the subtle differences observed were not strongly characteristic of any compounds previously profiled (unpublished data).
Hydroxy-PP Molecular Target Identification {#s2d}
------------------------------------------
With no cell pathway--specific information about the target or targets of hydroxy-PP, we relied on the differential sensitivity of A549 cells to hydroxy-PP and the kinase inhibitors PP2 and PP. Our hypothesis was that hydroxy-PP and PP share targets in common based on the similarity of their Cytometrix profiles in DU145, SF268, and SKOV3 cells ([Figure 1](#pbio-0030128-g001){ref-type="fig"}D). A corollary of this hypothesis is that hydroxy-PP targets one or more proteins in A549 and HUVEC cells that are not targeted by PP, thus leading to differential morphological attributes in A549 and HUVEC cells ([Figure 1](#pbio-0030128-g001){ref-type="fig"}D).
As a test of the first hypothesis, we focused on identification of enzymes that are potently inhibited (IC~50~ \< 1 μM) by PP and hydroxy-PP. Based on the known protein kinase--inhibitory properties of PP2 and its structural homologs, including PP \[[@pbio-0030128-b07],[@pbio-0030128-b13]\], we predicted that hydroxy-PP was also a protein kinase inhibitor. We tested hydroxy-PP against four divergent protein kinase targets: the tyrosine kinase Fyn, \[[@pbio-0030128-b07],[@pbio-0030128-b13]\], p38-α, protein kinase A, and protein kinase B ([Table S3](#st001){ref-type="supplementary-material"}). The additional -OH moiety of hydroxy-PP, did not diminish kinase-inhibitory activity toward Fyn, the best known target of PP2; instead, it enhanced it such that hydroxy-PP exhibited a 5-nM IC~50~ for Fyn inhibition. Similar to what occurred with PP, hydroxy-PP did not exhibit potent inhibition of any of the three other kinases tested ([Table S3](#st003){ref-type="supplementary-material"}).
To test the corollary hypothesis, we opted for a direct approach utilizing the differential sensitivity of A549 cells, rather than attempting to identify the unique targets of hydroxy-PP among the protein kinases, for which the differential effects are likely to be small due to the highly conserved ATP-binding pockets of protein kinases. Very few biochemical or genetic methods are available for identification of the molecular targets of small molecules in cells \[[@pbio-0030128-b14]\]. The most commonly used method is affinity purification by immobilization of the small molecule on a solid phase matrix. This technique requires both a high-affinity small molecule that allows stringent washing of weakly bound targets and a relatively abundant target to allow for mass spectrometry (MS)--based sequence identification. These two properties are not often found in early "hits" from random screening efforts like ours. Nonetheless, we decided to attempt affinity purification of the targets of hydroxy-PP using hydroxy-PP beads with PP beads as a negative control.
To ensure that attachment of a linker to hydroxy-PP did not destroy the target-binding properties, we synthesized a linker-containing analog of hydroxy-PP, hydroxy-PP-L, and a similar linker analog of PP, PP-L ([Figure 1](#pbio-0030128-g001){ref-type="fig"}C). These N-1 analogs of hydroxy-PP and PP were profiled using the Cytometrix system and were found to have the same trajectories as their parent compounds, albeit with approximately 5-fold lower potencies (data not shown). This modest loss in potency is not uncommon for linker-containing analogs \[[@pbio-0030128-b16]\], and importantly demonstrates that the *tert*-butyl substituent at N-1 is not required for target binding. Reactigel beads presenting either hydroxy-PP or PP2 were subsequently synthesized ([Figure 2](#pbio-0030128-g002){ref-type="fig"}A).
::: {#pbio-0030128-g002 .fig}
Figure 2
::: {.caption}
###### Affinity Purification and Identification of Human CBR1
\(A) Reactigel beads appended with hydroxy-PP or PP (control) were used for affinity purification of hydroxy-PP protein targets.
\(B) Hydroxy-PP-binding proteins in A549 cell lysates. Cytosolic fractions of A549 cell lysate (1.7 mg protein each) were incubated with the indicated affinity resin, and bound proteins were resolved by SDS-PAGE (12% acrylamide gel) followed by silver staining. Untreated beads and PP-control resin samples (lanes 1 and 2) indicate little nonspecific binding. Lanes 3, 4, 5, and 6 were loaded using the hydroxy-PP resin incubated with cell lysate and the indicated competitor. Vehicle or competitor compounds (200 μM) were added to the lysate 30 min before incubation with beads (lanes 4--6). Protein of bands B1--B3 did not bind hydroxy-PP beads when pretreated with hydroxy-PP (lane 5).
\(C) MS/MS peptide sequencing. Two tryptic peptides from bands B1--B3 were used to identify human CBR1.
:::
:::
A wide range of buffer conditions were explored to identify conditions under which proteins specifically bound to hydroxy-PP beads but not PP2 beads or underivatized beads ([Figure 2](#pbio-0030128-g002){ref-type="fig"}B, lane 3 versus lanes 1 and 2). Eight silver-stained protein bands at molecular weights of 15--38 kDa were retained on hydroxy-PP beads under these buffer conditions. To further distinguish those proteins that were specifically targeted by hydroxy-PP, and not by features of the linker or beads, the lysate was pretreated with DMSO, PP, or hydroxy-PP, the latter two compounds at 200 μM. Three bands (B1--B3) were not capable of binding to the hydroxy-PP beads with hydroxy-PP treatment, suggesting that these proteins were the targets of hydroxy-PP. Using matrix-assisted laser desorption/ionization (MALDI) and electrospray ionization (ESI) MS/MS sequencing, B1--B3 were each identified as human CBR1, a member of the short-chain dehydrogenase/reductase family of NAD(P)(H) oxidoreductases ([Figure 2](#pbio-0030128-g002){ref-type="fig"}C). The presence of three forms of CBR1 with differing electrophoretic mobilities has been previously observed and is believed to result from autocatalytic modification of a lysine residue \[[@pbio-0030128-b16]\]. Four of the other five protein bands that were not competed off by hydroxy-PP, but that also bound to the hydroxy-PP affinity matrix, were identified as nucleoside diphosphate kinase, nucleoside diphosphate kinase 2 (nm23) and pyridoxal kinase ([Table S4](#st004){ref-type="supplementary-material"}). Because these proteins were not eluted from the affinity beads following hydroxy-PP treatment ([Figure 2](#pbio-0030128-g002){ref-type="fig"}B, lane 5), we concluded they were recognizing a feature of the linker used to attach hydroxy-PP to the beads and so were not pursued.
To initially validate that the oxidoreductase CBR1 was indeed inhibited by hydroxy-PP, we measured CBR1 catalytic activity in vitro in the presence of hydroxy-PP or PP ([Figure 3](#pbio-0030128-g003){ref-type="fig"}). Hydroxy-PP exhibited potent (IC~50~ = 788 nM) inhibition of CBR1-catalyzed NADPH-dependent reduction of menadione to menadiol \[[@pbio-0030128-b17]\]. In contrast, PP exhibited no inhibition of CBR1 activity up to its solubility limit of 200 μM. This differential inhibition of CBR1 by hydroxy-PP but not PP validated our initial hypothesis that these two compounds possess different targets.
::: {#pbio-0030128-g003 .fig}
Figure 3
::: {.caption}
###### IC~50~ Values against CBR1 and Fyn Kinase Are Tabulated for PP Derivatives
:::
:::
To determine how the inhibition of CBR1 by hydroxy-PP was related to the original Cytometrix screen, the entire screening collection was re-screened in vitro for inhibition of CBR1. The only member of the collection that showed inhibition of CBR1 below an IC~50~ value of 1 μM was hydroxy-PP (data not shown). These screening data suggest that the morphology-based screen provided an efficient measure of the inhibitory potential of CBR1 inhibition, even though the screen included no direct measurements designed to read out CBR1 function. To determine whether the absence of an observable differential effect of hydroxy-PP, as compared with PP2, on SKVO3, DU145, and SF268 cells was due to an absence of CBR1 expression in these cells, we carried out a protein immunoblot analysis of CBR1 expression levels in each of the cell types analyzed by Cytometrix. Each of the cell types expressed CBR1 at approximately equal levels ([Figure S2](#sg002){ref-type="supplementary-material"}), suggesting that multiple factors other than expression level regulate CBR1 activity.
Structural Characterization of Hydroxy-PP--CBR1 Complex {#s2e}
-------------------------------------------------------
In order to develop a pharmacological agent that specifically inhibits CBR1, we addressed the target specificity of hydroxy-PP. In particular, hydroxy-PP\'s ability to potently inhibit the cytoplasmic tyrosine kinase Fyn (IC~50~ = 5 nM) in addition to CBR1 (IC~50~ = 788 nM) makes it a poor tool for probing CBR1 functions exclusively. We overexpressed human CBR1 in Escherichia coli and attempted crystallization of the protein in the presence of hydroxy-PP in an effort to enhance design of a selective CBR1 inhibitor. Within 2 d at room temperature, good diffracting crystals of the orthorhombic space group P2~1~2~1~2~1~ were obtained by vapor diffusion from 100 mM sodium-2-(N-ethylmorpholino)ethanesulfonate (pH 6.5), 2.0 M ammonium sulfate, and 5% PEG 400. Orthorhombic crystals of CBR1--hydroxy-PP diffracted to 1.1 Å. The structure was solved by molecular replacement with the AMoRe program \[[@pbio-0030128-b18]\] using a modified porcine carbonyl reductase \[[@pbio-0030128-b19]\] model and refined with SHELXL \[[@pbio-0030128-b20]\] to 1.24 Å with a crystallographic R-factor of 10.3% and a free R-factor of 13.4%.
Human CBR1 shows very high structural similarity to porcine carbonyl reductase, whose sequence is 85% identical to human CBR1 \[[@pbio-0030128-b21]\]. Although NADP(H) was not present during purification of the enzyme from E. coli nor added to the crystallization experiments, one molecule of NADP was found to be bound in the CBR1--hydroxy-PP structure. The same occurrence has been reported for the structure of porcine carbonyl reductase \[[@pbio-0030128-b19]\]. Hydroxy-PP binds to the substrate-binding site of CBR1, with the pyrazolopyrimidine core of hydroxy-PP mainly surrounded by hydrophobic residues (Trp229, Met141, and Ile140). The phenolic hydroxyl group of hydroxy-PP, however, points deep into the substrate-binding pocket and interacts with Tyr139 and Ser193 of the catalytic triad. The phenolic oxygen is positioned 2.5 Å from O^η^ of Tyr139 and 2.5 Å from O^γ^ of Ser193, thus indicating strong hydrogen bonding. The C4 carbon of the NADP(H) nicotinamide ring is positioned 3.2 Å from the *meta*-hydroxy carbon. This four-point geometry is iso-structural to the structures of other short-chain dehydrogenase substrate complexes (e.g., PDB 2AE2, 1FDS, and 1HZJ) and suggests a substrate-like binding mode of hydroxy-PP. Importantly, the structure of CBR1--hydroxy-PP provides a molecular understanding of the basis for the strong dependence of CBR1 inhibition on the presence of a hydroxyl moiety, as this functional group serves as a key interaction determinant with the catalytic machinery of CBR1. The binding mode of hydroxy-PP in CBR1 also explains the tolerance of the pyrazolopyrimidine core to derivatization at N-1, which allowed attachment of hydroxy-PP to solid support and affinity purification of CBR1.
Design and Synthesis of a CBR1-Selective Inhibitor {#s2f}
--------------------------------------------------
The availability of the X-ray structures of hydroxy-PP bound to CBR1 and of 1-*tert*-butyl-3-p-tolyl-1H-pyrazolo\[3,4-d\]pyrimidin-4-amine (PP1) bound to the tyrosine kinase Hck \[[@pbio-0030128-b22]\] afforded us the opportunity to compare how two virtually identical small molecules (PP1 and hydroxy-PP) are able to bind to two completely structurally and functionally unrelated enzyme targets ([Figure 4](#pbio-0030128-g004){ref-type="fig"}A and [4](#pbio-0030128-g004){ref-type="fig"}C versus 4B and 4D). Strikingly, the two co-crystal structures show grossly isosteric active site surfaces that are complementary to the two pyrazolopyrimidines. The geometries of both the pyrazolopyrimidine C-3 phenyl bond and orientation within the binding clefts are conserved within the two complexes. This finding complicates the design of a hydroxy-PP analog that cannot bind to protein kinases because the two binding clefts are highly similar in geometry. To design an analog of hydroxy-PP containing substituents that would disrupt protein kinase binding, we focused on electronic rather than steric aspects of the complexes. The exocyclic amine of PP2 is known to make key H-bond interactions with O^γ^ of Thr338 and O of Glu339 in the linker region of the protein kinase active site pocket ([Figure 4](#pbio-0030128-g004){ref-type="fig"}F) \[[@pbio-0030128-b22]\]. We designed and synthesized a mono-methyl-substituted version of hydroxy-PP, 3-(7-isopropyl-4-(methylamino)-7H-pyrrolo\[2,3-d\]pyrimidin-5yl)phenol (hydroxy-PP-Me), predicted to disrupt this key H-bonding interaction in kinases. Importantly, structural analysis of the hydroxy-PP--CBR1 co-crystal structure revealed a small space in the active site capable of tolerating a methyl substituent at this position ([Figure 4](#pbio-0030128-g004){ref-type="fig"}E).
::: {#pbio-0030128-g004 .fig}
Figure 4
::: {.caption}
###### Co-Crystal Structures of CBR1--Hydroxy-PP and Hck--PP1
(A, C, and E) show the binding mode of hydroxy-PP in co-crystals with CBR1. The inhibitor is oriented with its *t*-butyl group partially exposed to solvent and points toward the surface of the protein. The phenolic moiety of the inhibitor binds deeply within the substrate-binding pocket and makes close contacts to Ser193 and Tyr139 of the catalytic triad and the bound cofactor NADP. (B, D, and F) show the binding mode of the kinase inhibitor PP1 in complex with Hck. PP1 occupies the ATP-binding pocket as an adenosine analog. Although both protein structures show different folds (A and B), the morphology of CBR1- and Hck-binding sites are similar, and inhibitors hydroxy-PP and PP1 bind to these sites with similar shape complementarity (C and D). Key H-bond interactions between hydroxy-PP and the Ser193 and Tyr139 of CBR1 are indicated (E). The exocyclic amine of PP1 in complex with Hck makes essential H-bonds with the main-chain carbonyl oxygen of Glu339 and the side-chain oxygen of Thr338 (F). Disruption of this key H-bonding interaction by derivatization of the exocyclic amine destroys kinase affinity. The figure was prepared using the PyMol 2002 graphics system (DeLano Scientific, San Carlos, California, United States).
:::
:::
Indeed, hydroxy-PP-Me maintained potent inhibition activity against CBR1 (IC~50~ = 759 nM) but was an extremely poor inhibitor of the cytoplasmic tyrosine kinase Fyn (IC~50~ \> 70 μM; see [Figure 3](#pbio-0030128-g003){ref-type="fig"}). Because hydroxy-PP-Me lacks an H-bond donor group known to be key for potent protein kinase inhibition, we anticipate that it is an extremely poor inhibitor of all cellular protein kinases. A small screen against four protein kinases (Fyn, p38-α, protein kinase A, and protein kinase B) (see [Table S3](#st003){ref-type="supplementary-material"}) has shown that none are inhibited by hydroxy-PP-Me. Further in vitro screens against a large panel of protein kinases will be required to experimentally confirm this assertion.
CBR1 and Cancer Therapy {#s2g}
-----------------------
The carbonyl reductase CBR1 was first isolated from brain \[[@pbio-0030128-b17]\] and has been associated with two cellular functions: (1) detoxification of xenobiotics, such as the anthracycline daunorubicin and (2) metabolism of ketone-containing cellular messengers, such as prostaglandin E (reviewed in \[[@pbio-0030128-b23]\]). Genetic studies that have intended to uncover the in vivo function of this enzyme have focused on the xenobiotic detoxification activity of the enzyme. CBR1 converts daunorubicin into daunorubicinol, a compound that lacks the anti-proliferative activity of the parent daunorubicin and is cardiotoxic. Thus, metabolism of daunorubicin by CBR1 is thought to be responsible for the severe cardiotoxicity associated with daunorubicin treatment. In support of this function, mice heterozygous for a null allele of CBR1 show reduced sensitivity to anthracycline-induced cardiotoxicity because reduced CBR1 expression produces lower levels of doxorubicinol (CBR1 homozygous null mice are nonviable) \[[@pbio-0030128-b24]\]. Further support for this model comes from transgenic mice overexpressing CBR1, which exhibit increased cardiotoxicity associated with doxorubicin treatment \[[@pbio-0030128-b23]\]. It has been suggested \[[@pbio-0030128-b23]\] that because CBR1-dependent metabolism of the anthracyclines doxorubicin and daunorubicin reduces their efficacy in tumor-cell killing, a pharmacological inhibitor of CBR1 should potentiate anthracycline-induced cancer-cell killing.
To test this hypothesis, we measured the ability of hydroxy-PP-Me and PP-L to block CBR1-mediated metabolism of daunorubicin in A549 cells using cell killing as a measure of the cellular status of daunorubicin metabolism. The potent kinase inhibitor PP-L, which does not inhibit CBR1, was used as a negative control to measure general toxicity of combining daunorubicin with a structurally related but inactive molecule. A549 cells were treated at a daunorubicin concentration (440 nM) corresponding to an approximate IC~50~ for cell killing as a single agent such that enhanced cell killing could be scored. Concentrations of PP-L and hydroxy-PP-Me that exhibited minimal cytotoxicity on A549 cells (8 μM) when used alone were selected for combination with daunorubicin. [Figure 5](#pbio-0030128-g005){ref-type="fig"}A shows cell viability results measured by an alamarBlue reduction assay for A549 cells following drug treatments. Hydroxy-PP-Me induced a 25% enhancement of daunorubicin-mediated A549-cell killing consistent with its ability to inhibit CBR1-mediated daunorubicin metabolism. In contrast, PP-L exhibited no enhancement of cell killing, further suggesting the need for CBR1 inhibition to enhance daunorubicin-mediated cell killing. Although the observed 25% enhancement of daunorubicin-mediated cell killing is modest, the hydroxy-PP-Me dose dependence ([Figure 5](#pbio-0030128-g005){ref-type="fig"}B) of this effect is further evidence that it is CBR1 mediated rather than a general toxic response.
::: {#pbio-0030128-g005 .fig}
Figure 5
::: {.caption}
###### CBR1 Inhibitors Enhance Daunorubicin-Mediated A549-Cell Killing, yet Prevent Apoptosis in Serum-Starved Cells
\(A) Cell viability as a function of drug treatment. DMSO, PP-L (8 μM), and hydroxy-PP-Me (8 μM) do not have a pronounced effect on cell viability when used alone. Daunorubicin (DR) alone induces a moderate decrease in cell viability that is accentuated by concomitant treatment with hydroxy-PP-Me.
\(B) Cell viability decreases dose dependently with concomitant daunorubicin (DR) treatment. Daunorubicin treatment induces a moderate decrease in cell viability when used alone. Hydroxy-PP-Me (1--8 μM) induces a dose-dependent decrease in cell viability with concomitant daunorubicin treatment.
\(C) PI staining of dead cells is appreciably decreased in serum-starved cells treated with CBR1 inhibitors. A high number of cells were PI stained 65 h following serum starvation in both control and PP-L treated conditions (top). Cells treated with the CBR1 inhibitors hydroxy-PP-L or hydroxy-PP-Me during serum starvation show appreciably less staining (bottom).
\(D) Quantification of PI-stained cells by fluorescence measurement 65 h following serum starvation. Hydroxy-PP-L and hydroxy-PP-Me induce a dose-dependent decrease in PI staining; whereas PP-L does not.
:::
:::
Anti-Apoptotic Effect of Hydroxy-PP-Me Links CBR1 Activity to Serum-Withdrawal-Induced Cell Stress {#s2h}
--------------------------------------------------------------------------------------------------
The embryonic lethal phenotype of CBR1^--/--^ mice suggests that the enzyme plays a nonredundant role in cell signaling during embryogenesis and development. In order to search for unknown biological roles for CBR1, we utilized the two structurally related pyrimidine-based inhibitors of CBR1 hydroxy-PP-L and hydroxy-PP-Me. The former inhibits CBR1 and protein kinases, whereas the latter lacks the kinase-inhibitory action while maintaining CBR1-inhibitory activity. PP-L was included as a negative control. Thus, the three compounds together constitute a probe set for assessment of CBR1 involvement in a variety of cellular processes.
We chose to focus on signals that induce apoptosis in A549 adenocarcinoma cells because this endpoint is important for many cell-fate decisions, including those relevant to cancer and inflammation \[[@pbio-0030128-b25]\]. CBR1 has been directly implicated in redox reactions leading to H~2~O~2~ generation, a known stimulus for apoptosis \[[@pbio-0030128-b26]\]. A549 cells were subjected to a wide range of cell stresses previously shown to induce apoptosis: (1) interferon-γ + Fas ligand, (2) H~2~O~2~, (3) interleukin-1β, (4) serum withdrawal, or (5) interleukin-1β + serum withdrawal. The CBR1 inhibitors hydroxy-PP and hydroxy-PP-Me showed no enhancement of induction of apoptosis by any of these conditions (data not shown). However, both inhibitors were able to block A549-cell apoptosis induced by serum withdrawal \[[@pbio-0030128-b27]\]. Propidium iodide (PI) staining and phase contrast images show that A549 cells 65 h following serum withdrawal undergo virtually 100% apoptosis whereas cells treated with hydroxy-PP-Me are almost completely protected against apoptosis, as judged by the number of PI-negative cells ([Figure 5](#pbio-0030128-g005){ref-type="fig"}C). The dual CBR1--protein kinase inhibitor hydroxy-PP-L also protects A549 cells against serum-withdrawal-induced apoptosis, although to a lesser extent than hydroxy-PP-Me.
To confirm that inhibition of CBR1 by hydroxy-PP-Me was responsible for the anti-apoptotic effects in serum-starved A549 cells, we turned to RNA interference (RNAi) as a means to validate the role of CBR1 in the process. Three types of 21-nt RNAi that target human CBR1 were tested for their ability to knock down CBR1 expression, and their effectiveness was confirmed with an anti-CBR1 antiserum Western blot ([Figure S3](#sg003){ref-type="supplementary-material"}A). The anti-apoptotic effects of these RNAi elements that target CBR1, as opposed to control RNAi elements that target an irrelevant target found in A549 cells, further validate that hydroxy-PP-Me is a potent CBR1 inhibitor in cells and that CBR1 is involved in serum-withdrawal-induced apoptosis ([Figure S3](#sg003){ref-type="supplementary-material"}B).
To begin to determine whether the observed serum-withdrawal-induced apoptosis in A549 cells is mediated by the well-characterized p53 pathway, we used RNAi elements targeting p53 and showed that loss of p53 also protects A549 cells from serum-starvation-induced apoptosis ([Figure S3](#sg003){ref-type="supplementary-material"}B). A connection between p53-mediated cell death and another NADH-dependent reductase (NQO1) has been discovered recently \[[@pbio-0030128-b28]\]. This reductase is a p53-binding partner, and, upon reactive oxygen species generation in cells, p53 is released from NQO1, which then induces apoptosis. We sought to determine whether CBR1, another nicotinamide-dependent reductase, could act similarly as part of a p53 complex by measuring p53 levels following serum-withdrawal-induced apoptosis. These studies revealed no ubiquitination nor a decrease in p53 concentration following CBR1 inhibition as when NQO1 inhibitors are added to cells, suggesting that CBR1\'s involvement with the apoptotic machinery does not follow the pattern established for other oxidoreductases (data not shown).
Discussion {#s3}
==========
We have explored a conceptually new approach for the discovery of novel potent and selective inhibitors of cellular proteins. Rather than attempt to search extensively through chemical space using large chemical libraries, we greatly expanded the amount of biological target space sampled in a single screen with only a limited collection (107) of small drug-like molecules of limited chemical diversity. The morphology-based screen led to the identification of hydroxy-PP, which exhibited a multi-dimensional morphological signature distinct from a known kinase inhibitor of related structure (PP2). Not surprisingly, hydroxy-PP inhibited protein kinases based on its similarity in structure to the known Src-family kinase inhibitor PP2. In order to discover the new target of hydroxy-PP, an affinity-based screen was carried out. This approach revealed a new protein target for hydroxy-PP that was not inhibited by PP2. The target identified for hydroxy-PP and validated by X-ray crystallography was not a protein kinase, but the NADPH-dependent oxidoreductase, CBR1.
The surprising ability of a protein kinase inhibitor to cross-inhibit a member of a completely distinct protein family by simple hydroxyl-group substitution was rationalized based on X-ray structure analysis of the hydroxy-PP-binding site of CBR1. The protein kinase pocket occupied by PP in Hck is closely related, in terms of overall shape, to the hydroxy-PP-binding site of CBR1. Moreover, the hydroxyl group on hydroxy-PP makes contact with two catalytically essential residues in CBR1, thus providing direct structural insight into the original structure--activity relationships identified by morphology-based screening.
Genetic models have confirmed the importance of CBR1 in producing cardiac myocyte--toxic metabolites of important anthracyclines, such as daunorubicinol \[[@pbio-0030128-b29]\]. The discovery of a small-molecule inhibitor of CBR1 and the evidence of improved cell killing by daunorubicin in conjunction with hydroxy-PP-Me of lung adenocarcinoma cells now allows for assessment of this potential improvement to current adjuvant therapy for treating breast cancer and childhood leukemias by reducing or preventing the cardiotoxicity currently associated with anthracycline therapy. Further studies in mice treated with daunorubicin and hydroxy-PP-Me are being initiated to investigate possible cardiotoxicity reduction once the in vivo organ distribution and pharmacokinetics of hydroxy-PP-Me are determined. In particular, distribution of hydroxy-PP-Me in the heart may be essential to block daunorubicin-induced cardiotoxicity because it is thought that CBR1 activity in the heart is responsible for producing the high local concentration of daunorubicinol that is toxic to cardiomyocytes.
The discovery of a potent and selective CBR1 inhibitor also provides a powerful tool for discovery of pathways in which CBR1 plays a role. The CBR1 knockout is lethal, preventing such studies by genetic means. In a limited screen conducted so far, one process found to be most sensitive to CBR1 inhibition was serum-withdrawal-induced apoptosis. The selective CBR1 inhibitor hydroxy-PP-Me was able to block more than 90% of apoptosis induced by this stimulus. This pathway was not known to be dependent on the oxidoreductase CBR1, thus validating that the compounds discovered in such screens can lead to chemical tools capable of uncovering novel functions of key signaling regulators.
The apparent paradox that CBR1 serves a pro-apoptotic function during serum-withdrawal-induced apoptosis but has a protective function when cells are challenged with anthracyclines is a consequence of the different roles of the natural and unnatural (anthracycline) substrates of CBR1. The means by which inhibition of CBR1 causes increased cell death in anthracycline-challenged cells is through blocking the anthracycline-detoxification activity of CBR1. However, the anti-apoptotic effect of blocking CBR1 activity during serum-withdrawal-induced apoptosis is unclear. Inhibition of CBR1 may prevent generation of H~2~O~2~ via cellular quinone reduction by CBR1 and the subsequent comproportionation reaction of a cellular hydroquinone species and O~2~ to form H~2~O~2~ \[[@pbio-0030128-b30]\]. Further experiments to determine the cellular substrates of CBR1 that are involved in execution of serum-withdrawal-induced apoptosis using methods for directly radiolabeling reduction products are being pursued \[[@pbio-0030128-b31]\].
The present study suggests that current focused collections of small molecules for cell-based screens contain potent ligands for cellular targets that might be missed when screens based on single readouts or single-cell processes are used. Efforts to include more specific readouts in the microscopy-based screens, such as use of phospho-specific antibodies for reading out specific kinase activation, have been recently described \[[@pbio-0030128-b01]\]. Other fluorescent readouts, including tagged fusion proteins for visualization of individual protein trafficking events and fluorescent sensors of metals such as Ca^2+^and Zn^2+^, and other visualization methods may increase the information content available for predicting targets of novel small molecules.
Materials and Methods {#s4}
=====================
{#s4a}
### Cell culture and cell plating {#s4a1}
SKOV3 (ovarian epithelial cancer), A549 (lung epithelial cancer), and SF268 (central nervous system epithelial cancer) were chosen for their broad genetic diversity and obtained from the National Institutes of Health. These human cells were grown and maintained in RMPI medium (Mediatech, Herndon, Virginia, United States) with 5% fetal calf serum (FBS, HyClone, Logan, Utah, United States). DU145 (prostate epithelial cancer) cells ( ATTC, Manassas, Virginia, United States) were maintained in MEM with 5% FBS. HUVEC cells (VEC Technologies, Rensselaer, New York, United States) were maintained in MCDB131 medium (VEC Technologies) with 10% FBS. For the assays 1,000--1,800 trypsinized (Mediatech) cells per well (depending on cell type) were plated in 384-well plates (Corning Costar, Acton, Massachusetts, United States) using a Multidrop (Thermo Labsystems, Beverly, Massachusetts, United States) and incubated for 24 h, the approximate doubling time. Six cell plates were made for each cell line.
### Compounds {#s4a2}
Compound stocks were maintained in DMSO. Compound information was stored in ActivityBase (ID Business Solutions, Guildford, United Kingdom). Compounds were serially diluted in 384-well drug plates to achieve eight concentrations diluting 3× in DMSO between wells using a Multimek (Beckman Coulter, Fullerton, California, United States). Wells were reserved on every drug plate for negative and positive controls. Negative-control wells contained DMSO only and the positive-control wells received a titration of paclitaxel, a tubulin stabilizer (Sigma-Aldrich-Fluka, St. Louis, Missouri, United States). For known kinase inhibitors for which we possessed information about the IC~50~ against known targets, the starting concentration was customized such that the IC~50~ would be in the middle of the dose-response (eight 3-fold dilutions) curve. Small molecules were tested at 40 μM on cells based on the logic that cellular effects on cells at greater than 100 μM concentration are most likely nonspecific and that we would most likely get a full-dose response after eight 3-fold dilutions (6 nM to 40 μM final concentrations).
### Compound addition {#s4a3}
Compounds were added to cell plates from drug plates. Compound mixed with medium was added to the cells to achieve a 0.4% DMSO concentration on prepared cell plates using a PlateTrak (CCS Packard, Torrance, California, United States). Treated cells were incubated for another 24 h, the compound exposure time.
### Staining {#s4a4}
Cells were stained for visualization of the nuclei, Golgi apparatus, and microtubules. Cells were fixed with 4% formaldehyde (Polysciences, Warrington, Pennsylvania, United States) for 1 h; washed in TBS (Teknova, Half Moon Bay, California, United States); blocked in 0.01% Triton X-100 (ICN Biomedicals, Irvine, California, United States) and 1% BSA in TBS (Teknova) and then left to incubate for 1 h; stained with 0.01% Triton X-100, 0.1% BSA, 5 μg/ml Hoechst 33342 (Molecular Probes, Eugene, Oregon, United States), 5 μg/ml FITC-Lectin Lens Culinaris (Sigma-Aldrich-Fluka), and 3 μg/ml Rhodamine Red--labeled monoclonal antibody DM1α (courtesy of Tim Stearns) for 1 h; and then washed in TBS. All of these steps were performed with a PlateTrak using an ELx405 microplate washer (Bio-Tek Instruments, Winooski, Vermont, United States). This was done in triplicate for each cell line.
### Imaging {#s4a5}
Cells were imaged on an inverted Axiovert 100M epifluorescent microscope (Carl Zeiss, Oberkochen, Germany) with a 5× objective and a Xenon lamp (Sutter Instruments, Novato, California, United States). A 5× magnification was chosen (rather than 40×) primarily as a function of the need to acquire a large number of individual cells in order to gather statistically relevant data on relatively rare subpopulations. The number of cells in a randomly acquired 40× image is not sufficient to make meaningful statistical measurements across many cells. Metamorph (Universal Imaging, Downingtown, Pennsylvania, United States) was used to control the motorized x, y, z stage (Prior Scientific, Rockland, Massachusetts, United States) that moved the plate to each well, autofocused, and took three successive fluorescent images with an Orca 100 camera (Hamamatsu, Hamamatsu City, Japan). Exposure times were set to minimize the number of saturated pixels in the image. Images of the negative-control wells typically contained 800 to 1,200 cells, depending on the cell line.
### Image analysis {#s4a6}
Custom software was used to segment objects and extract attributes. Segmentation of nuclei used a gradient method for edge detection, the microtubules were segmented using watershed, and the Golgi apparatus was segmented by expanding the mask from the nuclei. Object attributes are listed in [Table S1](#st001){ref-type="supplementary-material"}.
### Object classification {#s4a7}
Two algorithms using object attributes from the nuclei classify objects into different phases of the cell cycle. The cell-cycle algorithm classifies each cell by total intensity of its nuclei (DNA content) as Gap 1, Synthesis, or Gap 2. The condensation algorithm classifies each cell as condensed or not condensed using the mean intensity and area of the nuclei. Nuclei condensation is a surrogate marker for mitosis; meaning that condensed cells are typically mitotic cells but may be rounded-up cells.
### Data storage and analysis {#s4a8}
Custom software was written to automatically identify, register, organize, and analyze the images. Results from image analysis were automatically stored in an Oracle (Redwood Shores, California, United States) database. Information about the experimental conditions was manually entered into the database. Finally, software was written to compile compound information from ActivityBase, experiment design, image analysis, and data analysis into one report. SpotFire DecisionSite (Somerville, Massachusetts, United States) was used to visualize the attribute and PCA data.
### Quality control {#s4a9}
Quality metrics computed on the negative- and positive-control wells were used to determine the overall quality of the plate. Poor-quality plates were discarded and new plates were made. Image statistics---background, contrast, and saturation--were measured, and poor quality images were discarded. Results were aggregated over good images from replicated cell plates.
### Chemicals and chemical synthesis {#s4a10}
Reactigel 6X was purchased from Pierce Biotechnology (Rockford, Illinois, United States). NADPH, glutathione, menadione, and daunorubicin were purchased from Sigma. AlamarBlue was purchased from Biosource (Camarillo, California, United States) International. Boronic acids were purchased from Combi-Blocks (San Diego, California, United States), and palladium complexes were purchased from Strem Chemical (Newburyport, Massachusetts, United States). Other starting materials and synthetic reagents were purchased from Aldrich unless otherwise noted.
4-Amino-1-*tert*-butyl-3-phenylpyrazolo\[3,4-*d*\] pyrimidines (PP2 and hydroxy-PP) were synthesized according to Hanefeld et al.\[[@pbio-0030128-b32]\]. Benzyl bromide was used to protect the hydroxyl group for the hydroxy-PP synthesis. Hydrogenation (10% palladium on carbon) resulted in benzyl deprotection to yield hydroxy-PP.
Hydroxy-PP: Colorless powder; ^1^H NMR (400 MHz, DMSO-*d*6) δ 9.69 (brs), 8.22 (1H, s), 7.33 (1H, dd, *J* = 7.5, 7.5 Hz), 7.05 (2H, m), 6.86 (1H, dd, *J* = 7.5, 2 Hz), 5.74 (s), 1.73 (9H, s). ^13^C NMR (100 MHz, DMSO-*d*6) δ 158.1 (s), 157.8 (s), 154.6 (d), 153.8 (s), 141.7 (s), 134.5 (s), 130.2 (d), 118.9 (d), 115.6 (d), 115.0 (d), 98.5 (s), 59.6 (s), 28.7 (q). HR-EIMS calculated for C~15~H~17~N~5~O 283.1433 found 283.1434.
Chemical probes with a linker (PP-L and hydroxy-PP-L) were synthesized by the coupling reaction of the corresponding pyrazolo\[3,4-*d*\] pyrimidines with 1-bromo-11-*tert*-butoxycarbamoyl-3,6,9-trioxaundecane by treatment with sodium hydride.
Hydroxy-PP-Me was synthesized in three steps from 4-chloro-5-iodo-7H-pyrrolo\[2,3-*d*\]pyrimidine (compound 1 in [Figure S4](#sg004){ref-type="supplementary-material"}), with a 50% overall yield. Compound 1 was synthesized as previously described \[[@pbio-0030128-b33]\]. Mitsunobu alkylation of compound 1 proceeded efficiently to yield the N-alkylated pyrrolopyrimidine (compound 2 in [Figure S4](#sg004){ref-type="supplementary-material"}). S~N~aryl reaction of this product with methylamine in THF provided compound 3 ([Figure S4](#sg004){ref-type="supplementary-material"}), which was subsequently coupled to *meta*-hydroxyphenylboronic acid under Suzuki conditions to afford hydroxy-PP-Me.
To produce 4-chloro-7,7a-dihydro-5-iodo-7-isopropyl-4aH-pyrrolo\[2,3-*d*\]pyrimidine (compound 2 in [Figure S4](#sg004){ref-type="supplementary-material"}), the following steps were taken. To a dry 50-ml round-bottom flask was added 4-chloro-5-iodo-7H-pyrrolo\[2,3-*d*\]pyrimidine prepared as previously described \[[@pbio-0030128-b33]\] (0.5 g, 1.78 mmol) and PPh~3~ (0.84 g, 3.2 mmol). The materials were dried under high vacuum for 20 min, and the flask was purged with argon. THF (30 ml) and isopropanol (0.3 ml, 3.9 mmol) were added and the flask was cooled in an ethylene glycol/dry-ice bath, and diisopropyl azodicarboxylate (0.47 g, 2.3 mmol) was added drop by drop to the stirred solution. After 18 h, the volatiles were evaporated in vacuo and the resultant oil was dissolved in ethyl acetate (50 ml) and 50% saturated sodium bicarbonate (50 ml). The organics were extracted with ethyl acetate (3 × 50 ml), dried with sodium sulfate, and evaporated in vacuo to yield an orange oil. Silica gel chromatography (ethyl acetate:hexanes) afforded the desired product as a yellow solid (480 mg, 84% yield). ^1^H NMR (399.6 MHz, CDCl~3~) δ 1.5 (6H, d, J = 6.4 Hz), 5.1 (1H, sp, J = 6.8 Hz), 7.4 (1H, s), 8.6 (1H, s).
To produce 7,7a-Dihydro-5-iodo-7-isopropyl-N-methyl-4aH-pyrrolo\[2,3-*d*\]pyrimidin-4-amine (compound 3 in [Figure S4](#sg004){ref-type="supplementary-material"}), 4-Chloro-7,7a-dihydro-5-iodo-7-isopropyl-4aH-pyrrolo\[2,3-d\]pyrimidine (0.3 g, 0.93 mmol) from above was placed within a 15-ml pressure tube. Methylamine (2 M) in THF (15 ml) was added, and the reaction was left to stir overnight. The volatiles were removed in vacuo, and the residue was dissolved in methanol; then 5 ml of silica gel was added, and the volatiles were removed in vacuo. The adhered product was purified by silica gel chromatography (ethyl acetate:hexanes), and the requisite fractions were pooled and evaporated in vacuo to yield the desired product (0.25 g, 85% yield). ^1^H NMR (399.6 MHz, CDCl~3~) δ 1.43 (6H, d, J = 6.8 Hz), 3.13 (3H, d, J = 4.8 Hz), 5.0 (1H, sp, J = 6.8 Hz), 7.02 (1H, s), 8.35 (1H, s).
To produce hydroxy-PP-Me, 7,7a-Dihydro-5-iodo-7-isopropyl-N-methyl-4aH-pyrrolo\[2,3-*d*\]pyrimidin-4-amine (0.15 g, 0.475 mmol) from above was placed in a 50-ml round-bottom flask in which 12 ml of dimethoxy ethylene glycol was added. 3-Hydroxyphenylboronic acid (0.262 g, 1.9 mmol, predissolved in 3.3 ml of ethanol) was added at once and was followed by 1.9 ml of saturated aqueous sodium carbonate. Pd^0^(PPh~3~)~4~ (55 mg, 47 μmol) was added to the reaction; the vessel was purged with argon and set to stir at 80 °C for 48 h. The reaction was subsequently cooled and filtered through a bed of diatomatious earth (Celite, Sigma-Aldrich). The filtrate was evaporated in vacuo, and the residual material was adhered to silica gel using ethyl acetate as a solvent. Silica gel chromatography (ethyl acetate:hexanes) and evaporation in vacuo of the requisite fractions yielded the desired product (94.8 mg, 70.7% yield). ^1^H NMR (399.6 MHz, d^6^-DMSO) δ 1.76 (6H, d, J = 6.8 Hz), 5.03 (3H, d, J = 4.8 Hz), 5.34 (1H, sp, J = 6.4 Hz), 5.53 (1H, q, J = 4.8 Hz), 6.73 (1H, m), 6.85 (1H, m), 7.25 (1H, app t, J = 7.6 Hz), 7.37 (1H, s), 7.59 (1H, s).
### Cell lines {#s4a11}
A549 cells were purchased from ATCC and cultured in the F-12K medium with 10% of FBS. Cells were maintained at 37 °C in an atmosphere of 5% CO~2~. Adherant cells were released for passaging using an isotonic Trypsin solution (0.25% Trypsin, 0.02% Versene). Rabbit anti-human CBR1 antibody was obtained from Dr. Tsuchida \[[@pbio-0030128-b34]\]. The CBR expression vector, pET-11aCR, was obtained from Dr. Wermuth \[[@pbio-0030128-b35]\].
### Preparation of cell extracts {#s4a12}
Cells were collected and washed with PBS buffer once then sonicated in buffer A containing 50 mM Tris-HCl (pH 7.4), 2 mM dithiothreitol, 5 mM EDTA, 5 mM EGTA, 20 mM MgCl~2~, and 1 μl/ml protease inhibitor cocktail Set III (Calbiochem, San Diego, California, United States). The solution was centrifuged for 15 min at 10,000 *g* and 4 °C. The supernatant was recovered and loaded on the affinity matrices or stored at --80 °C.
### Preparation and use of affinity reagents {#s4a13}
Pyrazolopyrimidines were coupled to Reactigel 6X beads at a calculated final concentration of 10--50 μmol/ml of resin. They were stored at 4 °C as a 50% (v/v) slurry in ethanol. 20 μl of the suspension was washed with 1 ml of buffer A with 200 mM NaCl and 0.1% IGEPAL CA-630 (Sigma), then the cell supernatant (1--2 mg protein) was added. After 1 h incubation at 4 °C, the beads were washed with the same buffer followed by addition of 40 μl of 1× Laemmli sample buffer.
### Electrophoresis {#s4a14}
Proteins bound to the affinity matrix were recovered with 1× Laemmli sample buffer. Following heat denaturation for 3 min, the bound proteins were separated by 12% SDS-PAGE followed by immunoblotting analysis or silver staining. Silver staining was performed with the following parameters: fixative, 250 ml of 50% methanol; rinse, MilliQ purified water (Millipore, Billerica, Massachusetts, United States) containing 10 μM DTT, followed by 0.1% AgNO~3~ in MilliQ water (w/v); developer, 15 g of Na~2~CO~3~ in 500 ml of MilliQ water containing 250 μl of 37% formaldehyde.
### Protein identification: Materials {#s4a15}
Siliconized 0.65-ml tubes from PGC Scientifics (Frederick, Maryland, United States) were washed with methanol and water prior to use. Reverse-phase packing material was from Phenomenex (Torrance, California, United States); fused-silica capillary tubing was purchased from Dionex (Sunnyvale, California, United States). The matrix solution used for MALDI experiments containing αCHCA (α-cyano-4-hydroxycinnamic acid) was from Agilent Technologies (Palo Alto, California, United States). Solvents were purchased from Fisher Chemicals (Tustin, California, United States); all other reagents were obtained from Sigma-Aldrich-Fluka.
### Enzymatic digestions {#s4a16}
The gel bands were cut in a laminar flow hood under conditions to minimize contamination. Sample reduction, alkylation, and digestion was carried out a using Genomic Solutions Proprep digestion robot (Genomic Solutions, Ann Arbor, Michigan, United States). The in-gel digestions were performed according to a modified in-house protocol, under laminar flow. Reduction with 10 mM DTT was allowed to proceed for 20 min at 50 °C. Iodoacetamide was dissolved in 20 mM NH~4~HCO~3~ (pH 8.2) with 10% ACN, and alkylation of cysteine residues was carried out for 1 h at room temperature. Tryptic digestion was initiated by the addition of 1% (w/w) of side-chain-modified, TPCK-treated porcine trypsin and was allowed to proceed at 37 °C for 4 h. The digests were extracted manually with 40 μl of ammonium bicarbonate buffer solution followed by two 30-μl extractions with 60% acetonitrile. The digest extracts were pooled and concentrated to approximately 30 μl, and 10 μl of each mixture was desalted using C18 Zip Tips (Millipore). The samples were eluted into 10 μl (60% ACN, 0.2% TFA), and the volume was reduced to 3 μl using vacuum centrifugation. Targets for MALDI were spotted using the dried-droplet method by adding 0.7 μl of the sample and 1.0 μl of α,CHCA matrix solution. For ESI experiments, the remaining sample (approximately 2 μl) was injected onto a nano-capillary C18 column for HPLC separation.
### MALDI-TOF/TOF-MS analysis {#s4a17}
MALDI-MS data were acquired in an automated mode using a 4700 Proteomics Analyzer (Applied Biosystems, Foster City, California, United States). This instrument employed a neodymium:yttrium aluminum garnet frequency-tripled laser operating at a wavelength of 354 nm and a laser repetition rate of 200 Hz. Initially, a MALDI-MS spectrum was acquired from each spot (1,000 shots/spectrum). Then peaks with a signal to noise ratio greater than 15 in each spectrum were automatically selected for MALDI-CID-MS analysis (7500 shots/spectrum). A collision energy of 1 keV was used with air as the collision gas for collision-induced dissociation (CID) accumulation. After acquisition, the data were subjected to automatic baseline correction, mathematically smoothed, and stored in an Oracle database. Assuming that all ions were singly charged, peak lists from all MS/MS spectra were automatically extracted from the Oracle database and submitted for batch-analysis database searching using an in-house copy of Protein Prospector (version 4.3) with the new program LCBatch-Tag or an in-house copy of Mascot, version 1.8 (Matrix Science, Boston, Massachusetts, United States). The latter was managed using the Mascot Daemon (Matrix Science, Boston, Massachusetts, United States) program running on the same computer. The MS/MS mass values submitted to both search engines were limited using the following criteria: minimum S/N threshold was 8--10, masses of 0--60 Da and within 20 Da of the precursor were excluded, and a maximum of 60 peaks per spectrum were submitted.
Protein Prospector searches were performed by specifying the inclusion of high-energy fragment ions characteristic of the TOF/TOF instrument, whereas Mascot searches included only the low-energy fragment ions and internal ions. For externally calibrated spectra, the allowed mass tolerance that was specified between expected and observed masses for searches was ±75 ppm for MS data, ±200 ppm for MS/MS parent ions, and ±250 ppm for MS/MS fragment ions. All samples were searched against the nonredundant National Center for Biotechnology Information database (NCBInr.02.25.2002).
### nLC-ESI-Qq-TOF MS analysis {#s4a18}
Tryptic peptides were subject to LC-MS/MS analysis on a QSTAR Pulsar mass spectrometer (MDS Sciex, Concord, Ontario, Canada) operating in positive-ion mode. Chromatographic separation of peptides was performed as described earlier except that formic acid was used as the ion-pairing agent. The LC eluent was directed to a micro-ionspray source. Throughout the running of the LC gradient, MS and MS/MS data were recorded continuously based on a 5-s cycle time. Within each cycle, MS data were accumulated for 1 s, followed by CID acquisitions of 4 s on ions selected by preset selection parameters of the information-dependent acquisition (IDA) method. In general the ions selected for CID were the most abundant in the MS spectrum, except that singly charged ions were excluded and dynamic exclusion was employed to prevent repetitive selection of the same ions within a preset time. Collision energies were programmed to be adjusted automatically according to the charge state and mass value of the precursor ions. Peak lists for database searching were created using a script from within the Analyst software. Searches were performed using the two search engines meantioned earlier except that only the low-energy CID fragments characteristic of the ESI-Qq-TOF instrument were considered. The allowed mass-tolerance range between expected and observed masses for searches was ±100 ppm for MS peaks and ±0.1 Da for MS/MS fragment ions.
### Expression and purification of recombinant CBR1 {#s4a19}
Isopropyl-β-D-thiogalactoside (final: 1 mM) was added to cultures of E. coli BL21(DE3) harboring pET-11aCR vectors when absorption at 600 nm of the culture became 0.6--0.7 AU. The cells were centrifuged after induction for 4--6 h, then the pellet was suspended in the buffer A and sonicated. The solution was centrifuged for 15 min at 10,000 *g* and 4 °C. The supernatant was loaded on Glutathione Sepharose 4B beads (Amersham Pharmacia Biotech, Little Chalfont, United Kingdom). Beads were washed four times with buffer A. Bead-binding proteins were eluted with 50 mM sodium phosphate (pH 6.1) containing 500 mM NaCl and 20 mM glutathione. The eluate was loaded on PD-10 columns (Amersham Pharmacia Biotech) to exchange the buffer to 10 mM Tris-HCl (pH 7.4).
### CBR1 assay {#s4a20}
CBR1 activity was determined spectrophotometrically at 25 °C. The standard assay mixture consisted of 50 mM sodium phosphate (pH 6.8), 200 μM NADPH, and 200 μM menadione in a total volume of 1 ml. Compounds were dissolved in DMSO as 100× stock solutions. Reactions were initiated by the addition of enzyme, and initial rates were determined by monitoring the disappearance of NADPH at 340 nM. Controls without substrates or enzyme were routinely included.
### Kinase assays {#s4a21}
Fyn and p38 kinases were expressed in bacteria and purified as previously described \[[@pbio-0030128-b13],[@pbio-0030128-b37]\]. Protein kinase A and Protein kinase B were obtained commercially (Upstate Cell Signaling Specialties Charlottesville, Virginia, United States). For the inhibition assay, various concentrations of inhibitor were incubated with 50 mM Tris (pH 8.0), 10 mM MgCl~2~, 1.6 mM glutathione, 1 mg/ml BSA, 0.1 mg/ml of the requisite substrate peptide (see [Figure S1](#sg001){ref-type="supplementary-material"}), 3.3% DMSO, 11 nM (2 μCi) \[γ-^32^P\]ATP (6,000 Ci/mmol, NEN), and kinase in a total volume of 30 μl for 30 min. Reaction mixtures (27 μl) were spotted onto a phosphocellulose disk and washed with 0.5% H~3~PO~4~. The transfer of ^32^P was measured by standard scintillation counting. The IC~50~ values were defined to be the concentration of inhibitor at which the counts per minute was 50% of the control disk. When the IC~50~ value fell between two measured concentrations, it was calculated based on the assumption of an inversely proportional relationship between inhibitor concentration and counts per minute between the two data points.
### Cell viability assay and PI-stained cell assay {#s4a22}
Cell viability was determined by the alamarBlue reagent reduction assay in a 96-well culture plate, measuring the absorbance at 570 and 600 nm spectrophotometrically. The amount of PI-stained cells was estimated by measuring fluorescence following incubation with PI (5 μg/ml). For both experiments, data are shown as the percentages of nontreated control cells.
### p53 and ubiquitination detection in A549 cells {#s4a23}
A549 cells maintained as described earlier were seeded in 6-cm dishes at a density of 1.8 × 10^4^ cells/cm^2^ and incubated in the presence of 3 ml of F-12K and 10% FBS. After 48 h, the medium was replaced with 3 ml of serum-free F-12K containing either PP2, hydroxy-PP, or hydroxy-PP-Me (prepared from 5 mM DMSO stock solutions), each at concentrations of 2, 5, or 10 μM. Experiments were routinely performed in duplicate, and controls with representative DMSO concentrations were included. After 48 h incubation, the medium was removed, and the cells were lysed with 0.5 ml of modified RIPA buffer (1% NP-40, 50 mM Tris, 150 mM NaCl, 2 mM EDTA, 2 mM Na~3~VO~4~, 0.1% SDS, 0.1 mM DTT, and one Complete Mini, EDTA-free protease inhibitor tablet \[Roche, Basel, Switzerland\] per 10 ml) for 10 min. The contents of each plate were transferred to 1.5-ml micro-centrifuge tubes, the tubes were sonicated in a bath sonicator for 5 min, and the lysates were cleared by centrifugation (14,000 *g,* 10 min). The supernatant was collected, and the protein concentration was assayed using the Bio-Rad DC protein assay (Bio-Rad, Hercules, California, United States). Equal quantities of protein boiled for 1 min in 1× Lamelli buffer were subjected to electrophoresis using 7.5% acrylamide Tris-Cl Criterion gels (Bio-Rad). Proteins were electrophoretically transferred to nitrocellulose and immunoblotted with DO-1 anti-p53 HRP conjugated antibody or the anti-ubiquitin antibody Ub followed by anti-mouse IgG conjugated HRP (Santa Cruz Biotechnology, Santa Cruz, California, United States). Immuno-reactive proteins were analyzed by enhanced chemiluminescence (Pierce Biotechnology).
### Protein crystallization {#s4a24}
The CBR1 was expressed in E. coli and purified as described earlier. CBR1 (18 mg ml^−1^) in 30 mM sodium phosphate (pH 6.5), 100 mM KCl, and 20 μM DTT was incubated for 30 min with hydroxy-PP (1 mM final concentration) prior to crystallization. Within 2 d at room temperature, good diffracting crystals of the orthorhombic space group P2~1~2~1~2~1~ were obtained by vapor diffusion from 100 mM 2-(N-ethylmorpholino)ethanesulfonate (pH 6.5), 2.0 M ammonium sulfate, and 5% PEG 400.
### Data collection and structure determination {#s4a25}
Orthorhombic crystals of CBR1--hydroxy-PP diffracted to 1.1 Å. A full dataset was collected at the Advanced Light Source (Berkeley, California, United States) beamline 8.3.1 with an ADSC Quantum 4 CCD detector. The dataset was integrated and merged using the HKL2000 and SCALEPACK programs (HKL Research, Charlottesville, Virginia, United States) \[[@pbio-0030128-b37]\]. The structure was solved by molecular replacement with AMoRe \[[@pbio-0030128-b18]\] using a modified porcine carbonyl reductase model (1N5D). Crystallographic refinement to 1.2 Å was carried out and electron density maps were produced using SHELXL \[[@pbio-0030128-b20]\]. Model building was done using O \[[@pbio-0030128-b38]\] and Quanta 2000 (Molecular Simulations, San Diego, California, United States). Detailed data and refinement statistics are listed in [Table S5](#st005){ref-type="supplementary-material"}.
Supporting Information {#s5}
======================
Figure S1
::: {.caption}
###### Structures of Compounds Included in the Screening Library
(2.6 MB PDF).
:::
::: {.caption}
######
Click here for additional data file.
:::
Figure S2
::: {.caption}
###### Immunoblot for CBR1 Expression in the Six Cell Lines Analyzed
(554 KB AI).
:::
::: {.caption}
######
Click here for additional data file.
:::
Figure S3
::: {.caption}
###### siRNA Knockdown of p53 and CBR1 during Serum-Withdrawal-Induced Apoptosis
(2.3 MB AI).
:::
::: {.caption}
######
Click here for additional data file.
:::
Figure S4
::: {.caption}
###### Synthesis of Hydroxy-PP-Me
(518 KB AI).
:::
::: {.caption}
######
Click here for additional data file.
:::
Table S1
::: {.caption}
###### Attributes Used to Characterize Screening Compounds
(30 KB DOC).
:::
::: {.caption}
######
Click here for additional data file.
:::
Table S2
::: {.caption}
###### Definition of Components in [Figure 1](#pbio-0030128-g001){ref-type="fig"}B
(19 KB DOC).
:::
::: {.caption}
######
Click here for additional data file.
:::
Table S3
::: {.caption}
###### Inhibition Values of PP, Hydroxy-PP, PP2, and Hydroxy-PP-Me against Four Protein Kinases
(30 KB DOC).
:::
::: {.caption}
######
Click here for additional data file.
:::
Table S4
::: {.caption}
###### MS/MS Identification of Affinity-Purified Peptides
(31 KB DOC).
:::
::: {.caption}
######
Click here for additional data file.
:::
Table S5
::: {.caption}
###### Data Collection and Refinement
(49 KB DOC).
:::
::: {.caption}
######
Click here for additional data file.
:::
Accession Numbers {#s5a10}
-----------------
The SwissProt (<http://www.ebi.ac.uk/swissprot/>) accession numbers for CBR1 and NQO1 are P16152 and P15559, respectively. The ProSite (<http://au.expasy.org/prosite/>) accession number for human CBR1 is PS00061. The Protein Data Bank (<http://www.rcsb.org/pdb/>) accession numbers for the other gene products discussed in this paper are porcine carbonyl reductase (PDB 1N5D), Hck (PDB 1QCF), and short-chain dehydrogenase substrate complexes (PDB 2AE2, 1FDS, and 1HZJ). The accession number for the crystal structure of the modified human carbonyl reductase model refined with SHELXL to 1.24 Å with a crystallographic R-factor of 10.3% and a free R-factor of 13.4% is PDB 1WMA.
We would like to thank David Savage and Robert Stroud for invaluable assistance in collecting X-ray diffraction data on the in-house Rigaku generator and beamline 8.3.1 at the Advanced Light Source at the Lawrence Berkeley National Laboratory, Gerald Forrest for kindly providing anti-human CBR1 polyclonal sera, Bendicht Wermuth for plasmid DNA encoding human CBR1, and Dejah Petsch and John Wood for K252a analogs. We thank Pam England, David Julius, Tom Scanlan, Jack Taunton, and members of the Shokat laboratory for helpful comments on the manuscript. This work was supported by an award to Kevan Shokat from the Sandler Program for Asthma Research. MS studies were carried out at the UCSF Mass Spectrometry Facility supported by National Institutes of Health grant number NCRR RR01614. Daniel Rauh was supported by funds from the Deutsche Forschungsgemeinschaft. Thanks to Reginald de la Rosa, Corey Nislow, Vadim Kutsyy, and members of the Screening Operations, Software Development, and Data Analysis groups at Cytokinetics for technical contributions during this project. We also thank Donald Oestreicher for making the collaboration possible.
**Competing interests.** EV, SR, JKT, and CLA are affiliated with/employed by Cytokinetics.
**Author contributions.** MT, RB, DR, CLA, and KS conceived and designed the experiments. MT, RB, DR, SR, KCH, and CLA performed the experiments. MT, RB, DR, EV, and CLA analyzed the data. CZ, ALB, and JKT contributed reagents/materials/analysis tools. KS wrote the paper.
Citation: Tanaka M, Bateman R, Rauh D, Vaisberg E, Ramachandani S, et al. (2005) An unbiased cell morphology--based screen for new, biologically active small molecules. PLoS Biol 3(5): e128.
CBR1
: carbonyl reductase 1
CID
: collision-induced dissociation
DMSO
: dimethyl sulfoxide
ESI
: electrospray ionization
FBS
: fetal bovine serum
hydroxy-PP
: 3-(1-*tert*-butyl-4-amino-1H-pyrazolo\[3,4-d\]pyrimidin-3-yl)phenol
hydroxy-PP-Me
: 3-(7-isopropyl-4-(methylamino)-7H-pyrrolo\[2,3-d\]pyrimidin-5yl)phenol
MALDI
: matrix-assisted laser desorption/ionization
MS
: mass spectrometry
PCA
: principal component analysis
PI
: propidium iodide
PP
: 1-*tert*-butyl-3-phenyl-1H-pyrazolo\[3,4-d\]pyrimidin-4-amine
PP1
: 1-*tert*-butyl-3-p-tolyl-1H-pyrazolo\[3,4-d\]pyrimidin-4-amine
PP2
: 1-*tert*-butyl-3-(4-chlorophenyl)-1H-pyrazolo\[3,4-d\]pyrimidin-4-amine
RNAi
: RNA interference;
|
PubMed Central
|
2024-06-05T03:55:55.043926
|
2005-4-5
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1073692/",
"journal": "PLoS Biol. 2005 May 5; 3(5):e128",
"authors": [
{
"first": "Masahiro",
"last": "Tanaka"
},
{
"first": "Raynard",
"last": "Bateman"
},
{
"first": "Daniel",
"last": "Rauh"
},
{
"first": "Eugeni",
"last": "Vaisberg"
},
{
"first": "Shyam",
"last": "Ramachandani"
},
{
"first": "Chao",
"last": "Zhang"
},
{
"first": "Kirk C",
"last": "Hansen"
},
{
"first": "Alma L",
"last": "Burlingame"
},
{
"first": "Jay K",
"last": "Trautman"
},
{
"first": "Kevan M",
"last": "Shokat"
},
{
"first": "Cynthia L",
"last": "Adams"
}
]
}
|
PMC1073693
|
Introduction {#s1}
============
The complexity and coordination of cellular functions are remarkable in view of the disparate histories of the genes that make up contemporary genomes. In eukaryotes, new genes arise primarily through the duplication of existing genes \[[@pbio-0030130-b01],[@pbio-0030130-b02],[@pbio-0030130-b03],[@pbio-0030130-b04]\], while some ancestral genes are inactivated or eliminated over time. In contrast, prokaryotic genomes undergo substantial rates of gene acquisition from foreign sources \[[@pbio-0030130-b05]\], as well as duplication and loss of existing genes. Thus, if we consider the gene repertoire of a particular bacterial cell, some genes have been transmitted vertically for very long periods of time, perhaps from the time of the common ancestor of all cellular life-forms, whereas other genes were acquired or generated at various points in the history of the lineage, including some very recently.
Although the role of vertical transmission and horizontal transfer are both well documented, as yet, we have no comprehensive, quantitative picture of the genome-wide history of gene gain and loss over time for any particular prokaryotic group. The availability of many complete genome sequences of bacteria presents the possibility of tracing the history of individual genes within evolving lineages by identifying the points at which genes originate through acquisition or duplication, and the points at which genes are lost. The resulting picture would address several outstanding questions and paradoxes concerning bacterial genomes. For example, if there is a robust estimate of the cell (or organismal) phylogeny for a set of lineages, can we identify the events of gene acquisition, duplication, and loss that lead to the current gene repertoires of individual cells? Is the incidence of gene acquisition ongoing or episodic, and do acquired genes come from very close relatives or from distant sources? Of acquired genes, what types and what proportion become permanently installed within descendant genomes, and which are lost?
It is clear that gene duplication, gene loss, and gene transfer all impact bacterial genomes; but the relative contributions of each remain controversial \[[@pbio-0030130-b06],[@pbio-0030130-b07],[@pbio-0030130-b08],[@pbio-0030130-b09],[@pbio-0030130-b10],[@pbio-0030130-b11],[@pbio-0030130-b12],[@pbio-0030130-b13]\]. The situation is confounded by the fact that, in bacteria, the presence of two or more homologous sequences within a single genome might reflect the acquisition of a gene copy from a foreign source rather than the duplication of a resident gene. In the absence of further analysis, such homologs cannot be confidently described as paralogs (or duplicates) \[[@pbio-0030130-b14]\] or as xenologs (acquired via horizontal transfer) \[[@pbio-0030130-b15]\], and we propose the term "synologs" as an agnostic name for homologs within a genome arising from either process. Distinguishing the origins of synologs within genomes allows both the accurate dissection of gene families and the full reconstruction of events responsible for the contents of cellular genomes.
Here we investigate the full protein-coding gene repertoires within the γ-Proteobacteria, a group chosen because the large number of fully sequenced genomes, combined with their well-supported phylogenetic relationships, allows us to trace the origins of new genes in organisms that differ widely in their gene inventories (ranging from 564 protein-coding genes in Buchnera aphidicola to 5,540 in Pseudomonas aeruginosa) \[[@pbio-0030130-b16]\]. This group is an ancient bacterial phylum, at least several hundreds of million years old, based on the sequence divergence within the group \[[@pbio-0030130-b17]\] and on its containing at least one ancient subclade (*Buchnera*) that has cospeciated with hosts for over 100 million years. Available genome sequences include species displaying diversified lifestyles and subject to varying degrees of gene acquisition \[[@pbio-0030130-b05],[@pbio-0030130-b06],[@pbio-0030130-b18],[@pbio-0030130-b19],[@pbio-0030130-b20]\].
By assessing the history of every gene family, we find that gene acquisition is a major factor contributing to genomic diversity of these bacteria, but that, paradoxically, they rarely exchange genes. In addition, duplication appears to have played a secondary role in the evolution of gene repertoires, as multigene families are scarce and a substantial fraction of the genetic redundancy observed in genomes is better explained by gene acquisition from a distant source. These results support the view that bacterial genomes evolve mainly by incorporation of completely novel genes rather than by intragenomic duplication or by replacement of resident genes with distant homologs.
Results {#s2}
=======
Having defined all gene families of homologs present in 13 sequenced γ-proteobacterial genomes \[[@pbio-0030130-b16]\], we partitioned them according to their distribution among species and their incidence of synology. We examined the congruence of each of these families with the organismal phylogeny using maximum-likelihood (ML) tests, and we provide two estimates (one stringent and one permissive) of the number of lateral gene transfers (LGTs) found in these families ([Figure 1](#pbio-0030130-g001){ref-type="fig"}).
::: {#pbio-0030130-g001 .fig}
Figure 1
::: {.caption}
###### Distribution of Gene Families and Occurrence of LGT in γ-Proteobacteria
\(A) Numbers of species and numbers of synologs corresponding to the 14,158 gene families. Single-copy gene families (red bars) comprise the large majority of the families. The numbers of families in categories exceeding 300 members are displayed on top.
\(B) Losses required to reconcile gene distribution with organismal phylogeny \[[@pbio-0030130-b16]\] for gene families represented in fewer than six species. For each family, we inferred an initial acquisition event in the most recent ancestor of the species containing a gene from the family and tallied the minimum number of independent events of loss required to explain the phylogenetic distribution. Most distributions can be explained without invoking multiple gene losses, supporting the hypothesis of a single acquisition.
\(C) Percentage of families containing fewer than three synologs (\# gene copies − \# genomes = 0, 1, or 2) showing evidence of LGT by the method described in Lerat et al. \[[@pbio-0030130-b16]\] and [Figure 2](#pbio-0030130-g002){ref-type="fig"}. Boxes represent the conservative estimate of LGT and dashed bars represent the corresponding permissive estimate (see text). For families containing additional synologs (white bars), it was not practical to apply the same method; instead, we built neighbor-joining trees (see [Materials and Methods](#s4){ref-type="sec"}).
:::
:::
Single-Copy Genes {#s2a}
-----------------
Among single-copy genes present in six to 12 genomes ([Figure 1](#pbio-0030130-g001){ref-type="fig"}A: red bars in categories 6--12), 1%--5% display statistically supported incongruence with the organismal phylogeny (red bars, [Figure 1](#pbio-0030130-g001){ref-type="fig"}C), a low incidence in view of the high frequency of acquired genes in some of these genomes. The more permissive estimates (dashed lines in [Figure 1](#pbio-0030130-g001){ref-type="fig"}C) imply that up to 15% of these families with no synologs may have experienced LGT. This low rate of LGT in gene families not universally distributed was statistically indistinguishable from that of genes present in all 13 genomes (χ^2^ test, *p* \> 0.1). The absence of these gene families from one or more genomes could result either from presence in the ancestor followed by loss in some lineages, or from absence in the ancestor followed by acquisition from a distant source in a descendant lineage. Although this implies ongoing loss and acquisition of genes, our results indicate that even genes initially acquired from distant sources are rarely transferred subsequently among lineages of γ-Proteobacteria.
Occurrence and Source of Synology in Bacterial Genomes {#s2b}
------------------------------------------------------
The sizes of gene and protein families in bacterial genomes have previously been shown to follow a power law distribution \[[@pbio-0030130-b21],[@pbio-0030130-b22],[@pbio-0030130-b23],[@pbio-0030130-b24]\]. Within the γ-Proteobacteria, we find that, overall, very few gene families contain synologs, as evident in the low frequencies of families in which members outnumber genomes ([Figure 1](#pbio-0030130-g001){ref-type="fig"}A). When present, synology might be expected to be associated with LGT, given that gene acquisition is itself a possible source of synologs within a genome. Applying tests of phylogenetic congruence to distinguish between intragenomic duplication and LGT ([Figure 2](#pbio-0030130-g002){ref-type="fig"}), our stringent estimate is that LGT can be implicated as the cause of synology in 22% (51) of the 231 families in which gene copies outnumber genomes by one or two. In families with one or two synologs, LGT occurs at significantly higher frequencies than in families of similar species distribution but lacking synologs in both the stringent and permissive tests (*p* \< 0.0001; see [Figure 1](#pbio-0030130-g001){ref-type="fig"}C). This difference between the amount of LGT in families with and without synologs was also found to be significant using the more permissive estimates of LGT. In only 3% of the 231 families was the conflict with the organismal phylogeny not confined to synologs (see last column in [Figure 2](#pbio-0030130-g002){ref-type="fig"}) and more readily explained as due to a gene transfer elsewhere in the tree. Hence, most of the incongruence observed in families containing synologs can be explained by a single transfer event of a gene for which a homolog was already present in the recipient genome. In [Table 1](#pbio-0030130-t001){ref-type="table"}, the proportion of gene families showing evidence for LGT is classified according to their functional categories. Most families are of unknown function (and are listed as "hypothetical proteins"); however, several have been assigned to broad functional categories, such as "energy metabolism," "cellular processes," "transport and binding proteins," and "amino acid biosynthesis."
::: {#pbio-0030130-g002 .fig}
Figure 2
::: {.caption}
###### Testing for LGT and Duplication as Sources of Synology
The illustrated case is for families with a single synolog, that is, in which one genome contains two gene copies. We tested two alignments, each retaining one of the two copies (red or blue), against the reference organismal phylogeny \[[@pbio-0030130-b16]\]. When both alignments agreed with the reference tree (+/+), the synology could be attributed to recent intragenomic duplication, whereas in cases of phylogenetic incongruence of one of the alignments (+/− or −/+), LGT of one synolog was invoked. If both alignments rejected the reference tree (−/−), the family was considered as containing one or several LGT events. Tests of LGT were conducted similarly for families with an additional synolog (one genome with three synologs or two genomes each with two synologs). In such cases, each possible alignment containing a single copy per genome was tested. In addition to these tests, all family trees were inspected to confirm diagnoses of LGT.
:::
:::
::: {#pbio-0030130-t001 .table-wrap}
Table 1
::: {.caption}
###### Numbers of Gene Families Showing Evidence for LGT
:::
Numbers represent the conservative (and permissive) estimates of gene families showing evidence for LGT
:::
For families with larger numbers of synologs, the direct comparison of gene trees with the reference topology also indicates high levels of LGT (up to 60%), although its inference is less definite due to uncertainties in reconstructing complex histories of multiple gene gains and losses. Families with three or more synologs are few in number (\<2% of the 14,158 families) but include some instances of ancient gene duplication preceding the diversification of lineages.
Phylogenetic Signal and the Evidence for Vertical Inheritance {#s2c}
-------------------------------------------------------------
In tests for phylogenetic congruence, alignments that do not reject the reference organismal phylogeny are usually interpreted as reflecting vertical inheritance. However, such results, that is, the absence of a significant difference from the reference topology, can also be caused by phylogenetically uninformative alignments. Such problems are most likely for extremely divergent sequences, for which alignment and phylogenetic inference procedures are prone to failure, or for very short sequences, which may lack sufficient numbers of informative sites. Our gene families were constructed so as to exclude extremely divergent sequences, leaving the possibility that short genes are the most problematic ones. But, in our tests, there was no significant difference (*p* \> 0.2) in the incidence of LGT among genes of different size categories, implying that lack of sufficient information was not a primary reason for failing to reject the reference topology. Furthermore, to explain the result whereby families with synologs display more LGT than those without, one would need to hypothesize that the lack of phylogenetic signal is restricted to families without synologs. However, the difference in the frequency of LGT between the families with and without synologs remains evident in each of the size categories (see [Figure 1](#pbio-0030130-g001){ref-type="fig"}C). Cumulatively, these analyses indicate that our tests of phylogenetic incongruence have sufficient signal to infer vertical inheritance and are not affected by gene size.
Genes with Very Limited Phylogenetic Distributions {#s2d}
--------------------------------------------------
About half of the gene families (7,655 of 14,158; [Figure 1](#pbio-0030130-g001){ref-type="fig"}A) contain a single member confined to a single genome. The fraction of these genes in a genome varies as a function of the local phylogenetic sampling (from \<5% in Yersinia pestis CO92 to 40% in *Pseudomonas*) and of the evolutionary constraints on a genome (with few such genes in the highly reduced genomes of the endosymbionts *Buchnera* and *Wigglesworthia*). Two reasons may account for the exclusion of these genes from other families: first is the possibility that the threshold for delineating families was too restrictive and did not allow inclusion of distant homologs. In this case, there is a chance that a very quickly evolving gene might be assigned to its own, single-member family. Alternatively, genes that are unique to genomes may represent recent acquisitions from distant sources outside of the γ-Proteobacteria. To discriminate between these situations, we conducted a blastp search on each of the unassigned proteins that were confined to one γ-proteobacterial genome on the database containing all proteins present in sequenced bacterial genomes (EMGLib release 5 \[[@pbio-0030130-b25]\]).
This analysis, in which the cutoff for protein matches is based on e-values rather than on an empirically determined percentage of the maximal bit score, provided evidence that the majority of the single-member gene families within γ-proteobacterial genomes could be attributed to LGT. Only 17.5% of the proteins unique to a single genome had matches in other γ-proteobacterial genomes. These potentially represent quickly evolving genes that were originally excluded from protein families because of insufficient similarity. In contrast, 40% of the unique proteins gave hits in organisms outside of the γ-Proteobacteria, a distribution that will most likely arise by LGT between distantly related lineages. The remaining 42.5% of the single-member gene families correspond to orphan open reading frames (ORFans), that is, genes that have no homologs in the current databases.
Alternatively, this last category could result from the misannotation of genome sequences. However, a recent study of ORFans in Escherichia coli demonstrated that most encode functional proteins \[[@pbio-0030130-b26]\]. ORFan genes tend to be short and enriched in A/T nucleotides when compared to the rest of the genome, features that suggest that they originated in parasitic elements, such as bacteriophages \[[@pbio-0030130-b26]\]. An analysis of the base composition of the sets of unique genes in the γ-Proteobacteria demonstrate that in all genomes (with the exceptions of *Buchnera, Wigglesworthia,* and *Haemophilus,* each possessing few, if any, unique genes), ORFans are significantly biased toward A+T at the third codon positions when compared with other genes in the genome (averaging a 5% difference in A+T contents; *p* \< 0.05). This result is consistent with the hypothesis that these genes, which have no matches in current databases, have been recently acquired from bacteriophages, whose diversity is largely unsampled and unknown \[[@pbio-0030130-b27]\]. Therefore, the prevalence of gene families restricted to one or a few genomes ([Figure 1](#pbio-0030130-g001){ref-type="fig"}A) supports gene acquisition as a principal source of new genes in this group of bacteria.
For families containing single members in four or five genomes, ML tests supported phylogenetic congruence for nearly 100% of cases (results not shown). However, this high degree of congruence could reflect, in part, the large number of gene families shared by closely related genomes (e.g., the two *Yersinia* or the two xanthomonads). To further evaluate those families (with and without synologs) present in two to five genomes, we enumerated the gene losses required to explain the phylogenetic distribution of the family under the assumption of no LGT following a single initial appearance in a lineage ([Figure 1](#pbio-0030130-g001){ref-type="fig"}B). For 74% of these families, the occurrence of homologs among the taxa can be explained as a single acquisition by their common ancestor, followed by vertical inheritance with inference of, at most, a single subsequent loss.
Cumulatively, the phylogenetic evidence (for gene families present in six or more genomes) and the distributional evidence (for gene families present in fewer than six genomes) indicate that high levels of foreign gene acquisition have introduced the majority of genes of γ-proteobacterial genomes, but that this gene acquisition has little impact on gene phylogenies within this group. Massive gene uptake does not cause phylogenetic inconsistencies because (i) acquired genes come from sources outside of this group, (ii) they rarely have homologs within the recipient genome, and (iii) subsequent to their initial acquisition, genes tend to be vertically transmitted.
Extent of Gene Origination and Acquisition among Taxa {#s2e}
-----------------------------------------------------
The incidence of LGT varies enormously among the lineages included in our tree. For instance, in addition to possessing a very large number of unique genes, the genome of P. aeruginosa contains numerous genes from families whose phylogenetic distribution can only be explained by a very large number of gene losses in other lineages or by LGT ([Figure 1](#pbio-0030130-g001){ref-type="fig"}B). This species shares numerous genes with one other distantly related species (e.g., 28 with E. coli and 28 with *Salmonella enterica,* each of which would require the inference of five independent gene losses under the assumption of a single initial acquisition) or with a distant sister pair (43 with *Escherichia* + *Salmonella,* and 50 with the two *Yersinia,* corresponding each to a scenario invoking four gene losses). Additionally, P. aeruginosa is the only species in this group for which instances of LGT for single-copy, broadly distributed genes have been detected \[[@pbio-0030130-b16]\]. At the other extreme, the endosymbiotic species (*Buchnera* and *Wigglesworthia*) show virtually no evidence of gene acquisition.
Discussion {#s3}
==========
Previous attempts to reconstruct the history of gene repertoires in bacteria have examined gene distributions on a species phylogeny \[[@pbio-0030130-b09],[@pbio-0030130-b13],[@pbio-0030130-b28]\]. But ignoring the relationships among homologs will lead to incorrect assessments of the relative contributions of gene gain, loss, and duplication to genome inventories. For example, if a gene has spread widely through LGT, an analysis based on gene occurrence would conclude that this ubiquitous distribution resulted solely from vertical inheritance. Moreover, such methods cannot distinguish between LGT and duplication as the origin of synology and therefore provide a distorted view of the extent of duplication in bacterial genomes. Only by evaluating the evidence for concordance between the gene phylogenies and the organismal phylogeny is it possible to trace the history of gain, loss, and duplication affecting each gene family.
It was previously shown only about 200 single-copy genes are shared by these genomes \[[@pbio-0030130-b16]\] and that only 1% of these broadly distributed genes display statistically supported evidence of LGT. However, most of these genomes contain several thousand genes, indicating that the majority of genes in the genome were not present in the ancestor to all γ-Proteobacteria and that they originated either through LGT or by duplications as lineages diversified. By conducting an exhaustive phylogenetic analysis of all genes present in completely sequenced γ-proteobacterial genomes, we have evaluated the factors responsible for altering gene inventories and contributing to genomic innovation. It has long been recognized that duplication and LGT contribute to the genome composition of evolving bacterial lineages and, in particular, of lineages in the γ-Proteobacteria \[[@pbio-0030130-b05],[@pbio-0030130-b29],[@pbio-0030130-b30],[@pbio-0030130-b31]\], and we provide a quantitative assessment of the roles of these processes on a genome-wide scale.
An enormous incidence of gene acquisition is suggested by the large number of genome- or clade-restricted gene families, but beyond their initial acquisitions, few gene histories conflict with the organismal tree. Our results show that most acquired genes lack homologs in the recipient genome and in other γ-Proteobacteria. Therefore, most of the genes present in contemporary genomes have arisen from distant sources. Although these genes may have been transmitted from unrelated cellular organisms, recent work revealing the previously overlooked diversity of bacteriophages \[[@pbio-0030130-b27],[@pbio-0030130-b32]\] and their probable role in bacterial evolution \[[@pbio-0030130-b26],[@pbio-0030130-b33]\] suggest that they have contributed significantly to the evolution of bacterial gene repertoires.
Traditionally, high levels of LGT have been considered to be incompatible with a tree-like representation of bacterial evolution. However, the diversity of gene families unique to single genomes indicates that the pool of available genes is very large, allowing the rate of gene acquisition to be both high for a genome and very low for a particular gene. Interestingly, there is no evidence that genes with narrower phylogenetic distributions were more likely to undergo LGT, suggesting that the essentiality of a gene, as denoted by its universal presence among species, is not a predictor of its propensity for LGT. Hence, once acquired, most genes appear to strictly follow the organismal phylogeny.
Whereas in eukaryotes, most multicopy genes arise from duplications, we find that LGT underlies a substantial proportion of the cases of synology in bacterial genomes. But, overall, synology is rare among gene families. Because duplicates are only rarely retained in bacterial genomes for long periods of time, hidden paralogy, that is, the differential loss of paralogs in independent lineages, is an unlikely explanation for phylogenetic incongruence. The overall paucity of families with synologs and their association with high rates of LGT indicate that duplications are not a major mechanism for diversifying functions in these bacteria. Although duplications play an important role in the short-term adaptation of bacteria \[[@pbio-0030130-b29],[@pbio-0030130-b30]\], only a few duplicated genes are retained and subject to selection for diversifying functions. The fixation of duplicates requires the gradual evolution of sequence changes conferring differences in expression or function, whereas genes arriving through LGT are likely to be operationally distinct from those already present in a genome and, thus, immediately able to contribute unique functions and to be maintained in the genome by selection.
The large number of genes that are confined to a single genome indicates frequent gene acquisition in this group of bacteria. In contrast, substantially fewer genes are distributed in families present in more than one proteobacterial genome. Therefore, based on the distributions of gene families and on the abundance of genes confined to a single genome, recently acquired genes are lost most readily. This implies that genes are continuously integrated into the genomes but rarely persist long enough for hosts to diversify \[[@pbio-0030130-b31],[@pbio-0030130-b34]\]. Although a few such genes could be present in multiple species, but quickly evolving and unrecognizable due to loss of sequence similarity, this situation cannot apply widely given the close relationships of some of the genomes \[[@pbio-0030130-b26]\]. Rather, most genes confined to a single genome reflect recent acquisition from a source outside of the sampled γ-Proteobacteria.
Cumulatively, the picture emerging from these studies is that bacterial lineages are constantly subjected to the input of new genes from a large available pool. Conversely, resident genes are continually lost. As a result, genomes contain sequences that have been resident in a particular lineage for very different durations ([Figure 3](#pbio-0030130-g003){ref-type="fig"}). The extent of gain and loss can vary widely among lineages: among the γ-Proteobacteria, P. aeruginosa is at one extreme and contains a very heterogeneous assemblage of genes with distinct histories and varying widely in persistence, whereas *Buchnera* comprises genes with very long evolutionary histories within the cell lineage and essentially no recently acquired sequences. The coordination of complex networks of cellular functions is all the more remarkable given that the genes within a genome lack a cohesive history together.
::: {#pbio-0030130-g003 .fig}
Figure 3
::: {.caption}
###### LGT and Genome Evolution in γ-Proteobacteria
Only a small proportion of genes have been retained since the common ancestor of γ-proteobacteria (in red). Under the assumption that ancestral and contemporary genome sizes are similar, most of the genes present in this ancestral genome (in white) have been replaced by nonhomologous genes (yellow to green), usually via LGT from organisms outside of this clade. Once a new gene is acquired, its transmission follows vertical inheritance. The abundance of genes unique to a species (in blue) indicates that these bacteria (with the exception of the endosymbionts) constantly acquire new genes, most of which do not persist long-term within lineages. (Numbers of protein-coding genes, excluding those corresponding to known IS elements and phages, are in parentheses for each genome).
:::
:::
Our results, based on the distributions and phylogenies of all genes of a set of related genomes, provide a context for understanding several findings that previously seemed contradictory: extremely high levels of LGT \[[@pbio-0030130-b05]\], congruence among gene trees at various depths within bacteria \[[@pbio-0030130-b06],[@pbio-0030130-b16],[@pbio-0030130-b35],[@pbio-0030130-b36]\], and general agreement of sequence-based gene trees with phylogenies based on genome contents \[[@pbio-0030130-b37],[@pbio-0030130-b38]\]. We focused on the most intensively sequenced bacterial clade: as more genomic sequence data become available, similar approaches can be applied to determine if genome contents evolve in the same manner in other groups.
Materials and Methods {#s4}
=====================
{#s4a}
### Defining gene families {#s4a1}
To investigate the history of all protein-coding genes, we defined all gene families present in the following γ-Proteobacteria: E. coli K12 \[[@pbio-0030130-b39]\], B. aphidicola APS \[[@pbio-0030130-b40]\], H. influenzae Rd \[[@pbio-0030130-b41]\], Pasteurella multocida Pm70 \[[@pbio-0030130-b17]\], S. enterica serovar Typhimurium LT2 \[[@pbio-0030130-b42]\], Y. pestis CO-92 \[[@pbio-0030130-b19]\], Y. pestis KIM5 P12 \[[@pbio-0030130-b43]\], Vibrio cholerae (chromosomes I and II \[[@pbio-0030130-b44]\]), Xanthomonas axonopodis pv. *citri* 306 \[[@pbio-0030130-b45]\], X. campestris \[[@pbio-0030130-b45]\], Xylella fastidiosa 9a5c \[[@pbio-0030130-b46]\], P. aeruginosa PA01 \[[@pbio-0030130-b47]\], and *W. glossinidia brevipalpis* \[[@pbio-0030130-b48]\]. Protein sequences from complete genomes were retrieved from GenBank \[[@pbio-0030130-b49]\] and filtered to remove proteins annotated as insertion sequences or as bacteriophage sequences. Accession numbers for these genomes can be found in the Accession Numbers section of this paper.
Homologous genes (and resulting gene families) were defined using a cutoff for the degree of similarity among proteins reflected in the blastp bit scores \[[@pbio-0030130-b50]\]. The procedure for defining gene families was described in Lerat et al. \[[@pbio-0030130-b16]\] and is briefly summarized as follows: first, a bank containing all annotated protein sequences from all included species was queried with all the proteins in each of the genomes via blastp, such that all proteins were searched against both their resident genome proteins and those from the other species. To establish the threshold for grouping genes into a family, we examined the distribution of the ratio of the bit score to the maximal bit score (i.e., protein match against itself) based on that observed for the proteins of E. coli compared against proteins of the other genomes. In each case, there is a bimodal distribution, with a first peak of low similarity values, which is constant among comparisons and represents random matches, and a second peak of higher values, which varies from one comparison to another and therefore probably represents true homologs. The height of the second peak varies according to the number of gene family constituents and can range from one, for single member families, to hundreds. The two phases of the distribution are partitioned at approximately 30% of the maximal bit score, and thus proteins having bit score values ≥ 30% of the maximal bit score were considered homologous and members of the same gene family.
Genes were assigned to families by a simple link rule such that if gene A matches gene B, and gene B matches genes C, then all three are grouped into the same family. Comparisons among the families resolved after applying different thresholds (10%, 20%, 30%, or 40% of the maximal bit score) revealed that the 30% cutoff maximized the number of families containing genes from all 13 species, indicating that this criterion is optimal for the interspecific identification of homologous sequences. (Information about the distribution and constituents of gene families is available upon request from the authors.)
### Gene origins and ancestries {#s4a2}
Of the 14,158 gene families, 205 families are present as exactly one copy in each of the 13 genomes, and previous work has established that 99% (203) of these single-copy, widely distributed gene families are consistent with a single phylogeny, as expected if they share a history of vertical transmission through the replicating cell lineages \[[@pbio-0030130-b16]\]. This reference phylogeny provides a scaffold upon which the ancestry of every member of every gene family could be examined. To investigate how each gene originates within a genome and how gene families are generated, all protein-coding genes within each family were subjected to phylogenetic analysis.
Although strong evidence of LGT can be gained by a phylogenetic approach, several factors, including the sensitivity of the tests employed and the varied causes of phylogenetic incongruence (such as hidden paralogy or long branch attraction, besides LGT), can confound the interpretation of such analyses. Therefore, we estimated the frequency of LGT in gene families by both stringent and permissive approaches. The conservative estimates rely upon the analysis of four different ML tests of phylogenetic congruence and the visual inspection of the trees and alignments for each family. In this case, we require that at least three tests support phylogenetic incongruence and that this incongruence not be explicable parsimoniously by hidden paralogy or ambiguous alignments. In the permissive estimates, LGT is inferred when at least one out of the four tests supported phylogenetic incongruence and when the tree needed more than two independent gene losses to be explained by hidden paralogy.
Gene families differ in their distribution among the sampled genomes and in the numbers of members per genome, and we considered the following cases:
### Families without synologs {#s4a3}
We first focused on gene families that contained no synology (i.e., the number of genes equals the number of genomes in which family members are found) and whose members are present in at least six of the genomes considered. Sequences were aligned using ClustalW version 1.83 \[[@pbio-0030130-b51]\], and the best ML tree was inferred using proml from the PHYLIP package version 3.6 \[[@pbio-0030130-b52]\] with the JTT model of amino acid change \[[@pbio-0030130-b53]\] and a model of heterogeneity of evolutionary rates among sites (α parameter estimated from the dataset on the best tree, using Tree-Puzzle 5.1 \[[@pbio-0030130-b54]\]). The likelihood of this tree was then compared to the reference species phylogeny \[[@pbio-0030130-b16]\], using the different ML tests (Shimodaira-Hasegawa test \[[@pbio-0030130-b55]\], the one- and two-sided Kishino-Hasegawa tests \[[@pbio-0030130-b56],[@pbio-0030130-b57]\], and the expected likelihood weights \[[@pbio-0030130-b58]\]) implemented in Tree-Puzzle 5.1 \[[@pbio-0030130-b54]\] with a confidence interval of 5%. LGT was inferred from the results of these different tests and by visual inspection of the tree and alignment for each family.
### Families with synologs {#s4a4}
In cases where a gene family contained one or two synologs (i.e., \# species \< \# genes ≤ \# species + 2), we addressed whether synology arose from LGT or from intragenomic duplication by analyzing all possible combinations of genes from an alignment but including only one gene per species via individual ML tests (see [Figure 2](#pbio-0030130-g002){ref-type="fig"} for an explanation of the case with one synolog). When the tree including a particular synolog was incongruent with the reference species tree, we considered that synolog as potentially arising from LGT. This diagnostic was subsequently confirmed by analyzing the tree based on all gene family members applying the procedures (described below) used for families containing more than two synologs. In such cases, LGT was inferred when the synology could otherwise be explained only by a scenario invoking at least three independent gene losses.
Because procedures that reconstruct all possible phylogenies using individual synologs are difficult to interpret when numerous synologs are present, the ML tests were not applied to families with more than two synologs. The number of such families was small, which enabled us to infer cases of LGT by inspection of tree topologies. For each family containing multiple synologs, a tree based on the whole family was built with "Neighbor" using a distance matrix obtained from protdist (JTT model of amino acid change \[[@pbio-0030130-b53]\]) from the PHYLIP package version 3.6 \[[@pbio-0030130-b52]\]. Distances were computed under the γ-based method for correcting the heterogeneity of rates among sites with the α parameter obtained from the dataset on the best tree, using Tree Puzzle 5.1 \[[@pbio-0030130-b54]\].
### Families present in few species {#s4a5}
For gene families present in fewer than six genomes, ML analyses either are not possible (when family members are present in fewer than four species) or might overestimate congruence (when pairs of very closely related genomes are included, such as the two *Yersinia* or the two xanthomonads). To further evaluate the incidence of LGT in gene families distributed in two to five genomes, we inferred an initial acquisition event in the most recent ancestor of the species containing a homolog and tallied the minimum number of independent events of loss required to explain the phylogenetic distribution. Families requiring the inference of zero, one, or two losses can most readily be interpreted as vertically transmitted following their origin in the shared ancestor. In contrast, families requiring inference of many losses would be most reasonably interpreted as having undergone multiple acquisition events from outside sources or transfer between lineages of γ-Proteobacteria.
Supporting Information {#s5}
======================
Accession Numbers {#s5a1}
-----------------
The GenBank (<http://www.ncbi.nlm.nih.gov/Genbank/index.html>) accession numbers for genomes discussed in this paper are Escherichia coli K12 (NC 000913), Buchnera aphidicola APS (NC 002528), Haemophilus influenzae Rd (NC 000907), Pasteurella multocida Pm70 (NC 002663), Salmonella enterica serovar Typhimurium LT2 (NC 003197), Yersinia pestis CO-92 (NC 003143), Yersinia pestis KIM5 P12 (NC 004088), Vibrio cholerae (NC 002505 \[chromosome I\] and NC 002506 \[chromosome II\]), Xanthomonas axonopodis pv. *citri* 306 (NC 003919), Xanthomonas campestris (NC 003902), Xylella fastidiosa 9a5c (NC 002488), Pseudomonas aeruginosa PA01 (NC 002516 \[[@pbio-0030130-b47]\]), and *Wigglesworthia glossinidia brevipalpis* (NC 004344).
Financial support was provided by Department of Energy grant DEFG0301ER63147 to HO and National Science Foundation grant 0313737 to NAM.
**Competing interests.** The authors have declared that no competing interests exist.
**Author contributions.** EL, VD, and NAM conceived and designed the experiments. EL performed the experiments. EL, VD, HO, and NAM analyzed the data. VD contributed reagents/materials/analysis tools. EL, VD, HO, and NAM wrote the paper.
Citation: Lerat E, Daubin V, Ochman H, Moran NA (2005) Evolutionary origins of genomic repertoires in bacteria. PLoS Biol 3(5): e130.
LGT
: lateral gene transfer
ML test
: maximum-likelihood test
ORFans
: orphan open reading frames
|
PubMed Central
|
2024-06-05T03:55:55.049652
|
2005-4-5
|
{
"license": "Creative Commons - Attribution - https://creativecommons.org/licenses/by/4.0/",
"url": "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1073693/",
"journal": "PLoS Biol. 2005 May 5; 3(5):e130",
"authors": [
{
"first": "Emmanuelle",
"last": "Lerat"
},
{
"first": "Vincent",
"last": "Daubin"
},
{
"first": "Howard",
"last": "Ochman"
},
{
"first": "Nancy A",
"last": "Moran"
}
]
}
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.