| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print September 30, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Division of Medical Sciences, The School of Medicine, University of Birmingham, Edgbaston, United Kingdom
1 Correspondence: Renal Immunobiology, Division of Medical Sciences, The School of Medicine, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK. E-mail: C.O.S.Savage{at}bham.ac.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(TNF-) and a minimally activated model stimulated with 2 U/ml TNF- and in which ANCA was present as a secondary neutrophil stimulus. CD11a/CD18, CD11b/CD18, and CXCR2 contributed to adhesion and transendothelial migration in both models. However, when the endothelium was minimally activated with TNF-, CD11b/CD18 played an important role in stabilizing adhesion induced by ANCA immunoglobulin G (IgG). Analysis of ß2 integrins and chemokine receptors demonstrated that ANCA IgG had no effect on expression levels at the neutrophil surface but enabled an active conformational change of CD11b/CD18. Similar molecular mechanisms control neutrophil adhesion and migration through highly or minimally TNF--activated endothelium. However, the direct ANCA-mediated neutrophil stimulation is needed to drive migration through minimally activated endothelium, and CD11b/CD18 is recruited for greater stability of adhesion during this process and can undergo an activatory, conformational change in response to ANCA IgG.
Key Words: neutrophil • endothelium • integrins • chemokines
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Previous studies by this group have demonstrated that ANCA can increase the level of adhesion and migration of neutrophils to activated endothelium, and the increase is most prominent in response to minimally activated endothelium [2 U/ml tumor necrosis factor- (TNF-); ref. 7
]. However the molecular mediators of these changes were not examined in detail, although some insight was gained from studying the effects of anti-integrin antibodies on ANCA-neutrophil interactions with platelets under flow. There, CD11b/CD18, but not CD11a/CD18, appeared to play a role in the conversion from neutrophil rolling to stationary adhesion on platelets [8
].
CD11a/CD18 (lymphocyte function-associated antigen-1) and CD11b/CD18 (membrane-activated complex-1) are members of the ß2 integrin family and are most highly expressed on neutrophils. They contribute to the firm adhesion of activated neutrophils to endothelium. Neutrophils are activated by soluble endothelium-derived mediators such as IL-8/CXCL8 and platelet-activating factor, which mediate conformational change of ß2 integrins, and can then interact with their receptors on the endothelium. CD11a/CD18 is thought to be primarily involved in unstimulated neutrophil migration via interactions with endothelial intercellular adhesion molecule-1 (ICAM-1) [9 ]. CD11b/CD18 does not seem to be as critical in this process, as neutrophil migration can still occur in CD11b/CD18-deficient mice but is severely decreased in CD11a/CD18-deficient mice [10 ].
Chemokines are a large group of small chemotactic proteins, which initiate the infiltration of leukocytes to sites of infection and inflammation and may trigger conformational changes in cell surface-expressed integrin molecules through inside-out signaling [11
]. One of the major subgroups of chemokines is termed CXC, and IL-8/CXCL8 is the prototypic member. Chemokines bind to seven transmembrane domain G protein-coupled receptors. Neutrophils express two receptors for IL-8/CXCL8: CXC chemokine receptor 1 [CXCR1; IL-8 receptor type A (IL-8RA)] and CXCR2 (IL-8RB). CXCR1 binds with high-affinity to IL-8/CXCL8 only, but CXCR2 binds to IL-8/CXCL8, growth-related oncogene- (Gro-)/CXCL1, and epithelial-derived neutrophil-activating factor-78 (ENA-78)/CXCL5. CXCR1 is involved in superoxide production, whereas CXCR2 is responsible for initiating neutrophil migration [12
].
In this study, we investigated the role that ß2 integrins (CD18, CD11a, and CD11b) and chemokine receptors (CXCR1 and CXCR2) play in neutrophil adhesion and migration. We used two models. In one, the endothelium was highly activated with 100 U/ml TNF-, which is sufficient to induce secondary neutrophil activation. This could mimic conditions involving a severe inflammatory insult. In the second model, endothelium was weakly activated with 2 U/ml TNF-. This is sufficient to capture rolling neutrophils, but a second stimulus is needed for further neutrophil activation. In these studies, the second stimulus has been provided by ANCA, thereby mimicking inflammation during systemic vasculitis and indicating that ANCA has a strong, proadhesive effect. We found that the molecular mediators supporting the two models were found to differ in their quantitative, but not overall qualitative, contributions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Immunoglobulin G (IgG) isolation
ANCA-positive IgG was isolated from sera taken from patients during the active phase of WG and MPA, and normal (ANCA-negative) IgG was isolated from sera obtained from normal, healthy volunteers, as described previously [2
]. Patients with WG or MPA accorded with the Chapel Hill Consensus Conference definitions for vasculitis [13
]. All patients had clinical evidence of disease activity (Birmingham Vasculitis Activity Score >2), and all had evidence of active renal vasculitis, confirmed histologically by renal biopsy. The presence of PR3-ANCA and MPO-ANCA was confirmed by antigen-specific enzyme-linked immunosorbent assay (Binding Site, Birmingham, UK). ANCA-positive IgG was used at a concentration of 200 µg/ml, as this was shown previously to be the optimum concentration for the conversion of rolling to stationary adhesion [8
]. The normal IgG concentration in serum is 8–18 g/L, and levels are usually within this range in vasculitis patients prior to treatment. The IgG preparations were polyclonal, and specific ANCA IgG was only a fraction of the total IgG.
Human umbilical vein endothelial cell (HUVEC) isolation
Umbilical cords were collected with informed consent. HUVEC were isolated as described previously [14
] and were cultured in medium 199 supplemented with 20% heat-inactivated fetal calf serum, 1 ng/ml endothelial cell growth factor, and 1 µg/ml hydrocortisone. All assays were performed on the first passage endothelial cell cultures, which were grown on the inner surface of glass capillary tubes (microslides) as described previously [14
, 15
]. All reagents were obtained from Sigma-Aldrich (Poole, UK) unless otherwise stated.
Microslide preparation
Glass microslides (rectangular cross-section, 0.3x3 mm; length, 50 mm, Camlab, Cambridge, UK) were coated with 4% 3-aminopropyltriethoxysilane [14
] and coated with 1% gelatin (type B from bovine skin). HUVEC were then seeded onto the slides at a level previously determined to give a confluent monolayer following sedimentation for 2 h at 37°C [15
]. Microslides were attached to a pump system, which enabled a short perfusion of fresh culture medium at 1-h intervals and were cultured for 24 h until confluent [14
].
Confluent HUVEC in microslides were stimulated with 2 U/ml or 100 U/ml TNF- (NIBSC, Herts, UK) for 4 h prior to assays [15
]. Each experiment was carried out using HUVEC from a single primary culture from a single donor. All reagents were obtained from Sigma-Aldrich unless otherwise stated.
Flow assay
The flow-based adhesion assay was similar to that described recently [7
], and all manipulations were carried out in an enclosed perspex chamber at 37°C. The microslides with the confluent HUVEC were glued to a glass microscope slide and were viewed by phase-contrast miscopy. One end of the microslide was attached to a Harvard syringe pump (Harvard Apparatus, South Natic, MA) by silicone rubber tubing, allowing the control of flow rate through the microslide. The other end of the microslide was attached by silicon rubber tubing to an electronic valve (Lee Products, Gerards Cross, UK) to ensure smooth switching between neutrophil suspension and the cell-free buffer.
Once the microslide was inserted into the flow system, HUVEC were perfused with wash buffer (PBS/BSA) to remove residual TNF-. Neutrophil suspension was flowed over the HUVEC monolayer at a wall shear stress of 0.1 Pa (equivalent to 1 dyn/cm2) for 4 min, followed by a period of washout. The wall shear stress used was adequate to ensure that binding of HUVEC required selectins and could not take place directly from flow-through, integrin-mediated adhesion. Video recordings were made of the cells 4–5 min after starting neutrophil perfusion (after 1 min of wash-out), 10–11 min after starting neutrophil perfusion (after 6 min of wash-out), and 15–16 min after starting neutrophil perfusion (after 11 min of wash-out).
Video analysis
Video recordings were used to analyze the state of perfused neutrophils. Three populations of adherent cells were counted: rolling adherent cells, characterized by cells moving over the surface at a velocity much slower than the free-flowing cells and maintaining their spherical shape; stationary-adhered neutrophils, which were bright-phase and distorted and had stopped and spread over the endothelium; and transmigrated neutrophils, which were phase-dark cells, indicating that they had migrated underneath the endothelial monolayer [15
]. The cells in each of six fields were characterized, and an average was calculated.
Data are displayed in graphs as the mean number of neutrophils per square mm per 106 perfused. In addition, data are expressed in text as the percentageof cells undergoing one type of interaction as a percentage of the total interactions, for example, the percentage of cells rolling in relation to the total number of cells interacting with the endothelium at that time-point. This allows comparison of the relative levels of interactions (rolling, stationary-adhered, or migrated). In the results, the percentage of interacting neutrophils rolling is shown at 5–6 min, the percentage of interacting neutrophils that were stationary-adhered is shown at 5–6 min, and the percentage of interacting neutrophils transmigrated is shown at 15–16 min. In addition, to quantify the stability of these interactions, the change in the number of cells interacting (rolling, stationary-adhered, or migrated) at 5–6 min compared with the number of cells interacting at 15–16 min was calculated (numbers interacting at 15–16 min/numbers interacting at 5–6 min) and will be referred to as the change in the total number of interacting cells over time.
Treatments
Normal IgG or ANCA IgG (200 µg/ml) was added to the neutrophil suspension just before neutrophil perfusion. Following neutrophil perfusion, PBS/BSA containing normal or ANCA IgG was perfused over the HUVEC monolayer. In these experiments, five PR3 and three MPO ANCAs have been used, and each experiment was carried out at least three times. Repeated experiments used different donors of neutrophils, and every experiment used different HUVEC derived from a single umbilical cord.
In monoclonal antibody (mAb)-blocking studies, neutrophils were treated with the antibodies for 20 min at room temperature prior to perfusion over the HUVEC monolayer. Integrin-blocking antibodies against CD18 (R6.5E), CD11a (DA36), and CD11b (KIM 249) were gifts from Dr. Martyn Robinson (Celltech, Slough, UK); they were used at 10 µg/ml. Function-blocking antibodies against CXCR1 and CXCR2 (Biosource, Nivelles, Belgium) were used at 2 µg/ml. Mouse IgG1 mAb was used as isotype control in these experiments.
Surface expression of integrins CXCR1 and CXCR2
Indirect immunofluorescence was used to analyze surface expression of CD18, CD11a, CXCR1, and CXCR2. Primary antibodies (R6.5E at 20 µg/ml; DA36 at 20 µg/ml; mouse anti-human CXCR1 or CXCR2 from R&D Systems, Abingdon, UK, both at 5 µg/ml) were incubated with neutrophils (100 µl at 3x106cells/ml) for 45 min at 4°C. The cells were washed in wash buffer (PBS/BSA) by centrifugation for 5 min at 150 g and then incubated with fluorescein isothiocyanate (FITC)-conjugated anti-mouse (Sigma-Aldrich) for a further 45 min at 4°C. Cells were washed in wash buffer and then fixed (2% paraformaldehyde). Cells were stained for CD11b using a retinal pigment epithelial-conjugated mouse anti-CD11b (Dako, Cambridgeshire, UK, at 1/100) and then were washed and fixed as above. Fluorescence intensity of labeled cells was measured by flow cytometry (Becton Dickinson, Beds, UK), and results were expressed as mean fluorescence intensity (MFI) or percentage positively stained (i.e., with fluorescence intensity, >95% of cells labeled using a control, nonspecific mouse IgG).
Integrin-conformational studies
To examine the effects of the treatments on conformational changes of CD11b, we examined the detection of an activation-specific epitope on CD11b. Neutrophils were suspended at 1 x 106/ml in PBS/BSA. Neutrophil suspension (100 µl) was added to 5 µl FITC-conjugated mouse IgG1 isotype control antibody (Dakocytomation, Denmark) or FITC-conjugated anti-CD11b activation epitope antibody CBRM1/5 (eBioscience, San Diego, CA) with the relevant treatment, which included buffer alone, manganese chloride 1 mM as a positive control, normal human IgG (250 µg/ml), or ANCA IgG (250 µg/ml). Cells were incubated at 37°C for 10 min and then were cooled on ice. Neutrophils were then washed in cold Hanks’ balanced salt solution containing HEPES, and cells were fixed in 2% paraformaldehyde. Fluorescence intensity of the neutrophils was measured in a flow cytometer (Becton Dickinson, Oxford, UK). All treatment/antibody combinations were done in duplicate with each donor’s neutrophils; neutrophils from two donors were used. Data acquisition and analysis were performed on Cellquest v3.3. Comparison of average geometric MFI was done by t-test.
Statistical analysis
The effects of different treatments were tested by ANOVA, using the SPSS package (SPSS, Chicago, IL) or Minitab computer package (Minitab, State College, PA). Individual comparisons were made by paired t-test where appropriate. All data are expressed as the mean ± SEM or % mean ± % SEM unless stated otherwise.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2 U/ml TNF-, or TNF- and ANCA IgG supported mainly stationary adherence (27.4±18.9%) and transmigration (64.7±14.8%) of neutrophils at 15–16 min. The mean number of neutrophils in contact with the endothelium stayed relatively constant over the study period, 689 ± 42 cells/mm2/106 perfused at 5–6 min–665 ± 56 cells/mm2/106 perfused during 15–16 min (Fig. 1A
).
|
(minimally activated model) supported mainly rolling of neutrophils (61.6±8.7% by 15–16 min), and only a small percentage showed stationary adherence (32.2±7.3% by 15–16 min). Over the 16-min analysis time, 
55 ± 5.9% of the neutrophils lost contact with the endothelium and rolled off during washing (Fig. 1B)
. The relative proportions of cells engaged in rolling, stationary adherence, or migration, i.e., the "pattern" of neutrophil interaction in the highly activated model, was significantly different from that observed in the minimally activated model (P<0.001). Neutrophils, perfused over endothelial cells that were not pretreated with TNF-, displayed negligible interaction with the endothelium (results not shown).
When neutrophils were treated with ANCA IgG prior to perfusion over endothelium (minimally activated model with ANCA), there was a dramatic change in their pattern of interaction (Fig. 1C) . The majority of the neutrophils were stationary-adhered (23.7±3.34% by 15–16 min) or migrating (59.1±17.1% by 15–16 min), and only a small proportion rolled over the endothelium (15.0±7.6% by 15–16 min). The mean cell number in contact with the endothelium over the 16-min period remained relatively constant (485±169 cells/mm2/106 perfused by 5–6 min–504±234 cells/mm2/106 perfused by 15–16 min) with a similar pattern of interaction to that observed when neutrophils were perfused over 100 U/ml TNF--treated endothelium, although the absolute number of cells in contact with the endothelium was lower. Treatment of cells with normal IgG (Fig. 1D)
was similar to untreated neutrophils (Fig. 1B)
. These findings are thus similar to those published previously [7
].
Using these two models of neutrophil activation on endothelium, triggered in one instance by factors secreted by highly activated endothelium and in the other, by autoantibodies binding to neutrophils rolling on minimally activated endothelium, we went on to dissect the molecular mechanisms involved using blocking antibodies to integrin proteins and chemokine receptors.
Effect of ß2 integrin-blocking on neutrophil interactions
In the presence of a highly activated endothelium
After blocking with anti-CD18 alone or anti-CD11a and anti-CD11b together, there was a dramatic increase in the proportion of rolling neutrophils and a decrease in stationary adhesion within 5–6 min of perfusion (Fig. 2A
and 2B
), and this was associated with a fall in the total number of neutrophils in contact with the endothelium over time (Fig. 2C)
, particularly after blocking CD18, giving a value of 0.45 ± 0.18 for the change in the number of total interacting cells over time. This corresponded with little migration apparent by 15–16 min with either type of treatment (Fig. 2D)
.
|
Isotype control antibodies showed no significant differences on neutrophil behavior compared with untreated cells (Fig. 2) .
In the presence of minimally activated endothelium, after stimulation of neutrophils with ANCA IgG
In this model, blocking CD18 alone or CD11a and CD11b on neutrophils pretreated with ANCA was associated with a marked increase in rolling and a decrease in the level of stationary adhesion at 5–6 min (Fig. 2E
and 2F)
, and as before, this was accompanied by a fall in the total number of neutrophils interacting with the endothelium over time (Fig. 2G)
. Use of anti-CD11a or anti-CD11b alone was accompanied by an increase in rolling at 5–6 min (Fig. 2E)
, although neither antibody alone significantly affected the level of interacting cells that were stationary-adhered (Fig. 2F)
.
With all treatments, migration was reduced significantly compared with control (Fig. 2H)
. It is interesting that loss of cell contact over time occurred with anti-CD11b but was less marked with anti-CD11a (Fig. 2G)
; loss of cell contact had not been observed when either of these antibodies was applied alone with the highly activated endothelium (100 U/ml TNF-) model (Fig. 2C) . These observations suggest that CD11b may be more important for stabilizing adhesion in the presence of minimal endothelial cell activation and an exogenous neutrophil stimulant. Isotype control antibodies showed no significant difference on neutrophil behavior compared with untreated cells (Fig. 2E
2F
2G
2H)
.
Overall, these data demonstrated that ß2 integrins operate in a similar way in the highly and minimally endothelium-activated models of inflammation.
Effect of CXCR1 and -2 blocking on neutrophil interactions
In the presence of a highly activated endothelium
After blocking both receptors, most of the neutrophils remained rolling (Fig. 3A
), although the level of stationary adhesion (Fig. 3B)
or the number of neutrophils interacting with endothelium was not significantly affected (Fig. 3C)
. In addition, blockade of CXCR1 and -2 was associated with a marked reduction in migration (Fig. 3D)
at all time-points.
|
Isotype control antibodies showed no significant difference in neutrophil behavior compared with untreated cells (Fig. 3A 3B 3C 3D) .
In the presence of a minimally activated endothelium after stimulation of neutrophils with ANCA IgG
Inhibition of CXCR1 and CXCR2 receptors led to a significant increase in rolling by 5–6 min (Fig. 3E)
; however, no significant change was seen in the number of interacting cells over time (Fig. 3G)
or the level of interacting cells stationary adhered at 5–6 min (Fig. 3F)
. This was associated with a significant decrease in migration by 15–16 min (Fig. 3H) . The effects on migration were mimicked by anti-CXCR2-blocking antibody used alone, and blocking CXCR1 did not lead to significant changes in the proportions of cells undergoing rolling, stationary adhesion, or migration (Fig. 3E
3F
3G
3H)
, suggesting that the effects of CXCR1 on induction of firm adhesion or migration are minimal in this model. Isotype control antibodies showed no significant difference on neutrophil behavior compared with untreated cells (Fig. 3E
3F
3G
3H)
.
Overall, in both systems, inhibition of the CXCR1 receptor did not have a major effect on neutrophil interactions with the endothelium. In contrast, blocking CXCR2 receptor led to an increase in unstable rolling and a marked reduction in neutrophil migration.
Receptor expression on neutrophils
To further investigate the role of ß2 integrins and chemokine receptors CXCR1 and -2 in neutrophil migration, we investigated changes in their surface expression, following TNF- priming and ANCA IgG treatment. In the adhesion assays undertaken under flow conditions, the neutrophils are not deliberately exposed to priming agents such as TNF-, only to the normal or ANCA IgG. However, in vitro studies of ANCA-induced neutrophil activation, such as induction of a respiratory burst, require priming, allegedly to increase expression of the target antigens (PR3 and MPO) on the neutrophil surface [4
, 16
]. We have previously suggested that deliberate priming is not required in the flow model, as it occurs during neutrophil recruitment to the endothelium [8
]. Thus, we examined receptor expression in response to ANCA in the absence and presence of TNF-.
Treatment with TNF-, ANCA IgG, or a combination of TNF- and ANCA did not significantly change the level of integrins expressed on the neutrophil surface. There was no change in the percentage receptor expression (data not shown) nor in the MFI. Although, the MFI for integrin CD11b was much higher than that for CD11a, this higher expression for CD11b has been reported previously [17
] (Fig. 4
).
|
(Fig. 5
). This was also observed after TNF- priming and ANCA IgG together (not shown).
|
or ANCA IgG stimulation, although a small decrease was seen in the MFI (P=0.014). This was observed in response to TNF- priming alone or TNF- and ANCA IgG treatment but was not evident when cells were treated with ANCA IgG alone, thereby suggesting this is an effect of priming rather than an ANCA IgG effect. Also, there was no difference between treatments with TNF- compared with TNF- and ANCA or TNF- and ANCA compared with TNF- and normal IgG (Fig. 6A
and 6B
). No significant change in MFI for CXCR2 was observed in response to treatment (P=0.18); however, the percentage expression of CXCR2 was decreased significantly in response to treatment with TNF- or TNF- and ANCA IgG (P=0.034). However, treatment with ANCA IgG alone did not down-regulate CXCR2 levels. As no decrease was seen following ANCA IgG treatment alone, CXCR2 down-regulation was probably a result of a TNF--priming effect (Fig. 6A
and 6B)
.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, mimicking a highly inflamed state in which there is likely to be plentiful chemokine release. In the second, ANCA-treated neutrophils were rolled over minimally activated endothelium in the continuous presence of ANCA IgG in an effort to model the inflammation occurring in vivo in vasculitis.
When endothelium was highly activated (100 U/ml TNF-), perfusion of untreated neutrophils produced mainly stationary-adhered and transmigrated neutrophils, as previously reported [7
]. The data obtained following blockade of CD11a and CD11b suggested that both of these molecules play a role in supporting neutrophil-endothelium interactions, particularly migration. Blockade of CD18 or CD11a and CD11b was associated with a dramatic increase in rolling by 5–6 min, a decrease in stationary adhesion, and a decrease in the overall level of neutrophil-endothelial cell interactions over time. Blocking of chemokine receptor CXCR1 did not significantly alter the pattern of neutrophil adherence or migration. Inhibition of CXCR2, however, reduced the level of migration. This is supported by already published reports that established a role for CXCR2 in cell migration through TNF-treated endothelial monolayers [12
]. The highly activated model is consistent with release of chemokines by the activated endothelium, which initiates signaling through CXCR2 and in turn, activates CD11a and CD11b by inside-out signaling to induce adhesion leading to migration.
When endothelium was minimally activated with 2 U/ml TNF-, it did not support stable adherence and migration, and the addition of a second neutrophil-activating stimulus was required. Thus, neutrophils pretreated with ANCA resulted in a high level of stationary adherence and migration, similar to that seen when neutrophils were perfused over highly activated endothelium. In this model, blockade of CD18, CD11a, CD11b, or CD11a and CD11b combined was associated with increased, unstable rolling and reduced migration. Further, blocking CD11b was more effective at preventing development of stationary adhesion, and over time, neutrophil interactions were less stable, suggesting that in this system, CD11b may be particularly necessary as the stabilizing integrin in the sequence. We previously reported a role for CD11b in studies investigating ANCA-activated neutrophil interactions with platelets [8
]; contrary to the present report, preliminary studies with ANCA-activated neutrophil interactions with endothelium did not appear to confirm a role for CD11b [7
]. The role of CD11b/CD18 in conferring stability has been demonstrated in studies of monocytes and eosinophils [18
, 19
]. Studies investigating adhesion efficiency of neutrophils to cells expressing only ICAM-1 demonstrated that although CD11a/CD18 bound ICAM-1 with greater efficiency, CD11b/CD18 maintained stability over time under shear stress. It is hypothesized that initial capture is mediated by CD11a/CD18 (high-avidity receptor), which may allow a shear-dependent increase in the avidity of CD11b, leading to stabilization over time [20
, 21
]. Animal studies have also demonstrated the sequential role of CD11a/CD18 and CD11b/CD18. Although it was CD11a/CD18 that induced more efficient adhesion in inflamed venules in vivo, CD11a/CD18 and CD11b/CD18 had a role to play in the adhesive process [22
].
In the minimally activated model, blocking CXCR1 had little effect on the overall level or pattern of neutrophil-endothelial cell interactions. In contrast, inhibition of CXCR2 was associated with an increase in rolling and significantly reduced the level of migration. Collectively, these results suggest that in this model, neutrophil stimulation by ANCA IgG, which induces a conformational change in CD11b, followed by ligation of CXCR2 jointly signal the initiation of stationary adherence and migration. Further, as bacterial infections are common prior to the onset of vasculitis and subsequent relapse and are accompanied by an increased level of TNF-, the combination of an endothelium activated by TNF- (systemic levels that are in the low/normal range) [23
, 24
] and the presence of circulating ANCA could lead to the induction of neutrophil adherence and transmigration. In the minimally activated model, signaling through CXCR2 was not sufficient to induce integrin activation and stationary adhesion by inside-out signaling. Nevertheless, although a reduced level of chemokines may be produced in this model, chemokines must still make a contribution, as CXCR2 blockade reduced migration. We would hypothesize that chemokines synergize with the ANCA-induced neutrophil signaling to promote stationary adhesion by enhancing ß2 integrin affinity [25
]. There are similarities between our system and that reported by Gopalan et al. [26
], where subthreshold levels of L-selectin cross-linking or IL-8 stimulation alone did not increase the percentage of rolling neutrophils that became adherent, but the combination of the two stimuli resulted in a significant increase in neutrophil adhesion.
The chemokine interacting with CXCR2 is likely to be IL-8/CXCL8, which has been shown to be released from endothelial cells in the flow system [27
]. IL-8/CXCL8 has been proposed to play a role in ANCA-associated vasculitis also [3
]. However, we cannot exclude a role for ENA-78/CXCL5 or GRO-/CXCL1. Given that neutrophil-endothelial cell interactions are considered to play an important role in development of vasculitic lesions [28
], the observations in this study that these interactions use the normal molecular processes of adhesion and migration suggest that it is the inappropriate induction of these processes that causes disease. A further factor may be ANCA-induced neutrophil activation with the untimely release of superoxide and proteolytic enzymes, as neutrophils are adherent to endothelial cells [4
].
The ability of ANCA to convert low-level, unstable interactions to high-level, stable interactions between neutrophils and endothelial cells prompted us to question whether ANCA might enhance neutrophil expression of the integrin or chemokine receptors under study. However, flow cytometry studies revealed that the level of surface expression of ß2 integrins on neutrophils was unaltered in response to treatment with ANCA, irrespective of whether neutrophils were primed with TNF- to enhance expression of the ANCA target antigens. Flow cytometric analysis did not reveal any alteration in the level of CXCR1 or -2 in response to ANCA. TNF- reduced the expression of CXCR2 on the neutrophil surface as previously reported [29
]. However, TNF- pretreatment of endothelial cells in these experiments is not thought to cause a down-regulation of CXCR2, as the TNF- would have been washed off the endothelium prior to neutrophil perfusion.
In summary, comparing two models of neutrophil-endothelial cell interactions under flow conditions revealed that the neutrophil receptors CD11a/CD18, CD11b/CD18, and CXCR2 play broadly similar roles in both models. Where levels of endothelial activation are limiting, neutrophil stimulation by ANCA can trigger stationary adhesion and migration and reveals subtle roles, particularly for CD11b, whose active conformation is increased following neutrophil exposure to ANCA.
|
ACKNOWLEDGEMENTS |
|---|
Received January 30, 2004; revised July 29, 2004; accepted August 13, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cell Adhes. Commun. 6,491-501[Medline]
subunit cytoplasmic domains Mol. Biol. Cell 10,861-873-induced inflammation by LFA-1 and Mac-1 Blood 99,336-341: a basis for local variations in vascular sensitivity to inflammation Blood 102,2828-2834 J. Immunol. 160,4518-4525This article has been cited by other articles:
|
|
 |
|
  S. L. Nolan, N. Kalia, G. B. Nash, D. Kamel, P. Heeringa, and C. O.S. Savage Mechanisms of ANCA-Mediated Leukocyte-Endothelial Cell Interactions In Vivo J. Am. Soc. Nephrol., May 1, 2008; 19(5): 973 - 984. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Colman, A. Hussain, M. Goodall, S. P Young, T. Pankhurst, X. Lu, R. Jefferis, C. O S Savage, and J. M Williams Chimeric antibodies to proteinase 3 of IgG1 and IgG3 subclasses induce different magnitudes of functional responses in neutrophils Ann Rheum Dis, May 1, 2007; 66(5): 676 - 682. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Williams, T. R. Pettitt, W. Powell, J. Grove, C. O.S. Savage, and M. J.O. Wakelam Antineutrophil Cytoplasm Antibody-Stimulated Neutrophil Adhesion Depends on Diacylglycerol Kinase-Catalyzed Phosphatidic Acid Formation J. Am. Soc. Nephrol., April 1, 2007; 18(4): 1112 - 1120. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A.-J. Ruth, A. R. Kitching, R. Y.Q. Kwan, D. Odobasic, J. D.K. Ooi, J. R. Timoshanko, M. J. Hickey, and S. R. Holdsworth Anti-Neutrophil Cytoplasmic Antibodies and Effector CD4+ Cells Play Nonredundant Roles in Anti-Myeloperoxidase Crescentic Glomerulonephritis J. Am. Soc. Nephrol., July 1, 2006; 17(7): 1940 - 1949. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D. Morgan, L. Harper, J. Williams, and C. Savage Anti-Neutrophil Cytoplasm-Associated Glomerulonephritis J. Am. Soc. Nephrol., May 1, 2006; 17(5): 1224 - 1234. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. C. Jennette, H. Xiao, and R. J. Falk Pathogenesis of Vascular Inflammation by Anti-Neutrophil Cytoplasmic Antibodies J. Am. Soc. Nephrol., May 1, 2006; 17(5): 1235 - 1242. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. M. McGettrick, J. M. Lord, K.-Q. Wang, G. E. Rainger, C. D. Buckley, and G. B. Nash Chemokine- and adhesion-dependent survival of neutrophils after transmigration through cytokine-stimulated endothelium J. Leukoc. Biol., April 1, 2006; 79(4): 779 - 788. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Watts, L. Harper, D. Jayne, J. Levy, C. Pusey, C. Savage, D. G. I. Scott, and J. Williams Translational research in autoimmunity: aims of therapy in vasculitis Rheumatology, May 1, 2005; 44(5): 573 - 576. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
0.1; **, P 
