Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Wegener's granulomatosis (WG) is a disseminated, necrotizing granulomatous vasculitis that primarily involves the upper airway, lungs, and kidneys (1) and is classically associated with the presence of the cytoplasmic antineutrophil cytoplasmic autoantibodies (c-ANCA) in the serum. c-ANCA is a sensitive and specific marker for WG and it is used to monitor disease activity (2). Proteinase 3 (PR3), the target antigen for c-ANCA, is a serine proteinase found in the azurophilic granules of neutrophils and monocytes that has substrate specificities similar to neutrophil elastase (NE) (6). PR3 is not normally expressed on the surface of resting neutrophils. However, neutrophils from patients with active WG express PR3 on their cell membranes in high amounts as a result of the ongoing inflammatory process (7, 8). Further, neutrophils from normal healthy volunteers can be induced to express PR3 by incubation with proinflammatory cytokines, e.g., tumor necrosis factor (TNF)-. This causes translocation of PR3 from its intralysosomal site to the cell membrane, where it is made available for c-ANCA binding (7).

The neutrophil plays a central role in the pathogenesis of WG. Renal biopsies from active WG patients demonstrate the presence of increased numbers of activated neutrophils in the affected tissue, and the number of activated neutrophils correlates with the degree of renal impairment (9, 10). Bronchoalveolar lavage fluid (BALF) taken from patients with active WG demonstrates high levels of neutrophils and c-ANCA. This contrasts with several other granulomatous lung diseases where the BALF typically contains high levels of lymphocytes (11). Primed neutrophils expressing high levels of PR3 can be activated in vitro by c-ANCA and monoclonal anti-PR3 antibodies, with resultant generation of reactive oxygen intermediaries, cytokine release, and degranulation with release of lytic enzymes (9, 12, 13). c-ANCA and monoclonal anti-PR3 antibodies activate the neutrophil by binding to PR3 and simultaneously crosslinking FcRIIa with its Fc component (14, 15). There are in vitro studies which suggest that c-ANCA-induced intravascular neutrophil activation may contribute to endothelial injury (16, 17).

Monocytes also express PR3, and the ability of c-ANCA and anti-PR3 antibodies to induce interleukin (IL)-8 production in primed monocytes has been demonstrated (18). Moreover, 1-antitrypsin (A1AT), the physiologic inhibitor of PR3, has been shown to block c-ANCA and anti-PR3 antibody-induced IL-8 production by primed monocytes. There are many clinical studies that associate A1AT deficiency with WG, suggesting a role for protease-antiprotease imbalance in the pathogenesis of this disease (19- 22). However, this has not been evaluated using neutrophils from patients with WG. We investigated the effect of A1AT on anti-PR3 immunoglobulin (Ig) G antibody-mediated activation of TNF--primed neutrophils from healthy volunteers and neutrophils isolated from patients with active WG. We demonstrate here for the first time that A1AT can inhibit anti-PR3 IgG antibody binding and induce activation of oxidative burst in TNF--primed neutrophils from healthy volunteers and neutrophils from patients with active WG.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

WG Patients

Six patients with newly diagnosed, biopsy-proven WG presented to our unit during this study period, four females and two males. Mean age was 58 yr (range: 22 to 72 yr). c-ANCA titers ranged from 1:80 to 1:1,200. Venous blood was drawn from these patients before commencing therapy and their neutrophils were isolated as described later.

Neutrophil Isolation

Neutrophils were isolated from heparinized venous blood (10 U/ml; Sarstedt, Nümbrecht, Germany) obtained from healthy volunteers (n = 15) and patients with newly diagnosed WG. Briefly, density gradient centrifugation was carried out in Ficoll-Paque (Pharmacia Biotech, Uppsala, Sweden) to separate the red cell pellet containing the neutrophil population from the lymphocytes. Neutrophils were separated from erythrocytes by sedimentation in a 3% dextran solution (Sigma-Aldrich, Poole, Dorset, UK). Residual erythrocytes were removed by hypotonic lysis solution and the isolated neutrophils were resuspended in RPMI-1640 (Sigma-Aldrich) at a final concentration of 2 × 10⁶/ml and used immediately. Cell viability was confirmed by trypan blue dye exclusion. All incubations were carried out at 37°C in 5% CO₂ for 30 min, unless otherwise stated.

Priming of Neutrophils from Healthy Volunteers to Induce PR3 Expression and In Vitro Neutrophil Activation

Neutrophils isolated from healthy volunteers (n = 12) were incubated with TNF- (2 ng/ml) (R&D Systems Europe Ltd., Abingdon, Oxon, U.K.) for 30 min. Control (unprimed) neutrophils were incubated in RPMI 1640 alone. After incubation, both groups of cells were washed twice with RPMI 1640 and resuspended in fresh medium. The TNF--primed cells were then incubated with a monoclonal anti-PR3 IgG (CLB, Amsterdam, Holland) at 1 µg/ml for 30 min. The production of reactive oxygen species was determined by the addition of the substrate dihydrorhodamine (Orpegen Pharma, Heidelberg, Germany) to the cell medium, followed by immediate quantitative flow cytometry to determine the production of reactive oxygen intermediaries using a FACScan (Becton Dickinson, Mountain View, CA), as previously described (23). At least 5,000 cells per sample were analyzed and expressed in arbitrary units of mean channel fluorescence (MCF). A separate group of primed cells was preincubated with A1AT (1 mg/ml) (Sigma Diagnostics, Sigma-Aldrich) for 30 min before the addition of the anti-PR3 IgG and subsequent analysis of neutrophil production of reactive oxygen intermediaries. Results are expressed as means ± standard error of the mean (SEM) from 12 separate healthy donors.

Control groups included primed cells incubated with A1AT only and unprimed neutrophils incubated either with A1AT (1 mg/ml) alone or with monoclonal anti-PR3 IgG antibody (1 µg/ml) alone before the addition of dihydrorhodamine and quantitative flow cytometry.

To evaluate the optimal dose of A1AT to use in these experiments, a dose-response of A1AT on anti-PR3 IgG activity was carried out. TNF--primed neutrophils were preincubated with varying amounts of A1AT (0.25, 0.5, 1, 2, and 4 mg/ml) before the addition of anti-PR3 IgG and subsequent analysis for reactive oxygen intermediaries. A1AT concentrations of 1 and 4 mg/ml showed similar inhibitory effects, and we used A1AT concentrations of 1 mg/ml for the remainder of this study. Results are expressed as means ± SEM for six healthy donors.

Anti-PR3 IgG Binding to Primed Neutrophils from Healthy Volunteers in the Presence of A1AT

For anti-PR3 IgG to activate primed neutrophils, it must crosslink PR3 and FcR11a on the surface of the neutrophil (14, 15). The observed effect of A1AT on reducing anti-PR3 IgG-mediated activation of primed neutrophils is probably due to A1AT preventing the anti-PR3 IgG binding to either PR3 or FcR11a, thereby preventing neutrophil activation. Therefore, the binding of anti-PR3 IgG to primed neutrophils in the presence of A1AT was examined first. TNF--primed and unprimed neutrophils from healthy volunteers were incubated with anti-PR3 IgG (1 µ/ml) for 30 min at 4°C. A separate group of primed cells was preincubated with A1AT (1 mg/ml) for 30 min before the addition of anti-PR3 IgG. Cells were then washed twice with 2% phosphate-buffered saline (PBS)/bovine serum albumin (BSA) (Sigma Diagnostics, Sigma-Aldrich). Fluorescein isothiocyanate (FITC)- labeled goat antimouse F(ab)₂ secondary antibody (DAKO AS, Glostrup, Denmark) was added to all groups of cells for 30 min at 4°C and followed by a wash in 2% PBS/BSA before quantitative flow cytometry using FACScan (Becton Dickinson). Cells were stimulated with argon laser light (488 nm) and emitted fluorescence at 580 nm was quantified in a minimum of 5,000 cells and expressed in arbitrary units of MCF. Results are expressed as means ± SEM for 10 subjects. As controls, the binding of monoclonal IgG1 isotype control (1 µ/ml) (DAKO) to primed and unprimed neutrophils in the presence of A1AT, as described earlier, was performed.

In a similar fashion, to ensure that the observed effect of A1AT on anti-PR3 IgG binding was not due to an anti-NE action, the effect of secretory leukoprotease inhibitor (SLPI) on anti-PR3 IgG binding to primed neutrophils was also examined. SLPI preincubation was performed in one group of primed neutrophils before the addition of the anti-PR3 IgG, and subsequent evaluation for anti-PR3 IgG binding was performed as described earlier. Results are expressed as means ± SEM for 12 subjects.

Anti-PR3 IgG Fab Fragment Binding in the Presence of A1AT

It is most likely that A1AT binds to the PR3 site on the neutrophil, thereby preventing anti-PR3 IgG binding to PR3 and subsequent crosslinkage of PR3 and FcRIIa. It is also possible that when A1AT is bound to PR3 the access to the FcRIIa receptor is also impeded. To evaluate whether A1AT interferes with anti-PR3 IgG binding to the surface of primed neutrophils at the level of PR3 or the FcRIIa receptor, we first examined the effect of A1AT on the PR3 binding site.

Anti-PR3 IgG Fab fragments are specific for PR3 only and do not bind to FcRIIa receptors. Anti-PR3 IgG and IgG1 isotype control Fab fragments were prepared and the integrity of Fab fragments was confirmed by electrophoresis in 15% sodium dodecyl sulfate polyacrylamide gel electrophoresis as described previously (18). TNF--primed and unprimed (control) neutrophils were incubated with anti-PR3 IgG or isotype control Fab fragments (0.1 µg/ml) for 30 min at 4°C and washed twice with 2% BSA/PBS before addition of a FITC-labeled goat antimouse F(ab)₂ secondary antibody and subsequent flow cytometric analysis. A minimum of 5,000 cells per sample were examined and expressed in arbitrary units of MCF. One group of primed cells was preincubated with A1AT (1 mg/ml) before the addition of anti-PR3 IgG or isotype control Fab fragments. Results are expressed as means ± SEM for seven subjects.

Next, we examined the effect of A1AT on the binding of a monoclonal antibody (mAb) to FcRIIa. The binding of a mAb for FcRIIa in the presence of A1AT was examined. If A1AT is effecting anti-FcRIIa IgG access to this receptor, a decrease in expression should be seen in the presence of A1AT. TNF-- primed neutrophils and control neutrophils were incubated with a mAb specific for FcRIIa (Medarex Ltd., Princeton, NJ) for 30 min at 4°C, washed twice with 2% BSA/PBS, and then incubated with a FITC-labeled goat antimouse F(ab)₂ secondary antibody, followed by a final wash in 2% BSA/PBS and flow cytometric analysis with a minimum of 5,000 cells per sample examined and expressed in arbitrary units of MCF. A separate group of primed cells was incubated with A1AT (1 mg/ml) for 30 min at 37°C before the addition of the anti-FcRIIa mAb and subsequent flow cytometric analysis. Results are expressed as means ± SEM for 13 subjects.

The Effect of SLPI on Anti-PR3 IgG-Induced Oxidative Burst on Primed Neutrophils from Healthy Volunteers

SLPI is a NE antiprotease with no anti-PR3 activity (24). To ensure that the effect of A1AT observed was a true anti-PR3 action not mediated through its anti-NE activity, the effect of SLPI on anti-PR3 IgG-mediated neutrophil activation was investigated. If the observed inhibitory effect of A1AT on anti-PR3 IgG activity is through an anti-NE action, either wholly or in part, that would suggest that SLPI would have a similar inhibitory action. TNF-- primed and control neutrophils from healthy volunteers were incubated with monoclonal anti-PR3 IgG (1 µg/ml). Separately, another group of primed neutrophils was preincubated with either SLPI (1 mg/ml) (R&D Systems Europe Ltd.) or A1AT (1 mg/ ml) for 30 min before the addition of anti-PR3 IgG and subsequent analysis for the production of reactive oxygen intermediaries as described earlier. Results are expressed as means ± SEM for 11 individuals.

PR3 Enzyme-Linked Immunosorbent Assay

To complement the results observed in the flow cytometric experiments, which suggest that A1AT complexed with PR3 prevented binding of the anti-PR3 IgG to the PR3 molecule, we performed a PR3 enzyme-linked immunosorbent assay (ELISA) for detection of PR3 in the presence of A1AT. We also performed a PR3 ELISA for detection of PR3 in the presence of SLPI to show that SLPI has no effect on anti-PR3 IgG binding to PR3. Samples were prepared in which a 10-fold molar excess of A1AT was reacted with PR3 in 0.05 M Tris/0.05 M NaCl, pH 7.5, for 30 min. Similarly, a 10-fold molar excess of SLPI was reacted with PR3 under the same conditions. The samples were then probed for PR3 by ELISA using the anti-PR3 IgG 12.8 as previously described (25). Briefly, the samples were coated overnight to microtiter plates (Immulon 2; Dynatech Laboratories, Alexandria, VA) in 0.1 M carbonated buffer, pH 9.6. Control samples of PR3, A1AT, and SLPI were also coated to the plate. The plate was washed five times with 0.05% Tween-20 in PBS and then incubated with anti-PR3 IgG 12.8 (diluted 1:1,000 in PBS supplemented with 5% normal goat serum, 0.25% Tween-20, and 0.15 M NaCl) for 1 h at room temperature. Primary antibody binding was detected with horseradish peroxidase (HRP)-labeled antimouse secondary antibody for 1 h, followed by substrate development with 2,2'-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS).

The Effect of A1AT on Neutrophils Isolated from Patients with Active WG

There is a strong association between A1AT deficiency and WG (19) and it has been proposed that there may be a defective A1AT-PR3 interaction in active WG (26). This may be due to a normal A1AT unable to interact with a dysfunctional PR3 expressed on the WG leukocytes. To evaluate this theory, we wished to evaluate the interaction of exogenous A1AT with neutrophils from patients with active WG. Thus, neutrophils were isolated from patients with newly diagnosed, c-ANCA-positive, biopsy-proven WG. Anti-PR3 IgG binding studies were performed, with one group of cells preincubated with A1AT (1 mg/ml) before the addition of anti-PR3 IgG as described earlier. Flow cytometric analysis was performed with a minimum of 5,000 cells per sample examined and expressed in arbitrary units of MCF. Results are expressed as means ± SEM for six subjects.

Because there was a wide range of PR3 expression noted with neutrophils isolated from this group of WG patients, and considering that previous investigators have demonstrated that PR3 expression is directly related to vasculitic activity (8), we opted to prime these neutrophils with TNF to ensure maximal PR3 expression, thereby maximizing anti-PR3 IgG-induced production of reactive oxygen intermediaries and, in turn, optimizing any effect of A1AT in this population of neutrophils. Thus, these neutrophils were primed with TNF- and the effect of anti-PR3 IgG on neutrophil production of reactive oxygen intermediaries was examined as described earlier. One group of primed WG neutrophils was preincubated with A1AT (1 mg/ml) before the addition of the anti-PR3 IgG and immediate quantitative flow cytometry to determine the production of reactive oxygen intermediaries using a FACScan (Becton Dickinson). At least 5,000 cells per sample were analyzed and expressed in arbitrary units of MCF. Results are expressed as means ± SEM for five subjects.

Statistical Analysis

All data are expressed as means ± SEM. Statisical analysis was performed using analysis of variance and Dunn's post hoc analysis and Wilcoxon's paired test analysis, as appropriate.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

A1AT Inhibits Anti-PR3 IgG-Induced Respiratory Burst in Primed Human Neutrophils from Healthy Volunteers

Incubation of primed neutrophils with anti-PR3 IgG led to a significant increase (P < 0.01, n = 12) in neutrophil oxidative burst activity as compared with TNF--treated cells alone (Figure 1A). This increase in oxidative burst activity could be significantly inhibited by preincubating primed neutrophils with 1 mg/ml of A1AT before the addition of anti-PR3 IgG (a reduction of 47 ± 7.7%; P < 0.001, n = 12). A1AT reduced this anti-PR3 IgG-induced oxidative burst in a dose-dependent fashion (Figure 1B). Previous studies have demonstrated that a concentration of 2 mg/ml A1AT inhibited anti-PR3 IgG-induced IL-8 production of primed monocytes (18). We found that an A1AT concentration of 1 mg/ml caused a reduction in oxidative burst activity similar to that observed with 2 or 4 mg/ml of A1AT. A concentration of 1 mg/ml of A1AT was used throughout this study. However, A1AT (1 mg/ml) alone had no effect on oxidative burst activity when added to primed neutrophils in the absence of anti-PR3 IgG or to unprimed neutrophils (data not shown).

View larger version (26K): [in this window] [in a new window]   Figure 1.   (A) A1AT inhibits anti-PR3 IgG-induced respiratory burst in primed human neutrophils from healthy volunteers. Neutrophils from healthy volunteers were incubated with TNF- (2 ng/ml) to induce PR3 expression. A group of these cells was then incubated with anti-PR3 IgG (1 µg/ml) for 30 min at 37°C. A separate group of TNF--treated cells was preincubated with A1AT (1 mg/ml) for 30 min at 37°C before the addition of anti-PR3 IgG. Neutrophil oxidative burst activity was measured by the oxidation of nonfluorescent dihydrorhodamine to fluorescent rhodamine 123 by flow cytometry. Oxidative burst in the group of cells incubated with anti-PR3 IgG (TNF/Anti-PR3 IgG) was significantly higher than those incubated with TNF (*P < 0.01). Oxidative burst in those cells preincubated with A1AT (TNF/ A1AT/Anti-PR3 IgG) was significantly lower (**P < 0.001) than that in the primed cells incubated with anti-PR3 IgG (TNF/ Anti-PR3 IgG). Results are expressed as means ± SEM from 12 separate healthy donors. (B) Dose-response of the effect of A1AT on anti-PR3 IgG-induced activation of primed neutrophils. Primed neutrophils were preincubated with increasing doses of A1AT (0.25, 0.5, 1, 2, and 4 mg/ml) for 30 min at 37°C. The cells were then incubated with anti-PR3 IgG (1 µg/ml) for 30 min at 37°C and oxidative burst was measured. Results are expressed as means ± SEM from six separate healthy donors.

A1AT Inhibits Anti-PR3 IgG Binding to Primed Neutrophils from Healthy Volunteers by Preventing Anti-PR3 IgG Binding to the PR3 Site

For anti-PR3 IgG to induce oxidative burst in the primed neutrophil it must crosslink PR3 and FcRIIa. A1AT may act at one or both of these sites to inhibit anti-PR3 IgG activation of primed cells. To examine this, initially we looked at the effect of A1AT on the binding of anti-PR3 IgG to neutrophils from healthy volunteers. Incubation of healthy human neutrophils with TNF- induced a significant increase in PR3 expression, as measured by anti-PR3 IgG binding (Figure 2A). Preincubation of primed cells with A1AT before the addition of anti-PR3 IgG resulted in a significant fall in anti-PR3 IgG binding (a reduction of 83 ± 1.5%; P < 0.001, n = 10). A1AT preincubation had no significant effect on the binding of the isotype control to primed cells when compared with controls (data not shown).

View larger version (20K): [in this window] [in a new window]   Figure 2.   (A) A1AT inhibits anti-PR3 IgG binding to primed neutrophils from healthy volunteers. Unprimed and TNF-- primed cells were incubated with anti-PR3 IgG (1 µg/ml) for 30 min at 4°C. One group of primed cells was incubated with A1AT (1 mg/ml) for 30 min at 37°C before the addition of anti-PR3 IgG. FITC-labeled secondary antibody was added for 30 min at 4°C before quantitative flow cytometry. Anti-PR3 IgG binding in the group of primed cells preincubated with A1AT (TNF/A1AT/ Anti-PR3 IgG) was significantly lower; **P < 0.001. Results are expressed as means ± SEM from 10 separate healthy donors. (B) A1AT inhibits anti-PR3 Fab fragment binding to primed neutrophils. Unprimed and TNF--primed cells were incubated with anti-PR3 IgG Fab fragments (0.1 mg/ml) for 30 min at 4°C. One group of primed cells was incubated with A1AT (1 mg/ml) for 30 min at 37°C before the addition of anti-PR3 IgG Fab fragments. FITC-labeled secondary antibody was then added for 30 min at 4°C before quantitative flow cytometry. Preincubation of primed neutrophil with A1AT (TNF/A1AT/Anti-PR3 IgG Fab Fragment) results in a significant reduction in anti-PR3 IgG Fab fragment binding; *P < 0.01. Results are expressed as means ± SEM from seven separate healthy donors. (C ) A1AT has no significant effect on monoclonal anti-FcRIIa antibody binding to TNF-- primed neutrophils. Unprimed and TNF--primed neutrophils were incubated with a specific mAb for FcRIIa for 30 min at 4°C. FITC-labeled secondary antibody was added for 30 min at 4°C before quantitative flow cytometry. A separate group of primed cells was preincubated with A1AT (1 mg/ml) before the addition of the anti-FcRIIa mAb and subsequent binding analysis. Incubation with TNF- (TNF/Anti-FcRIIa antibody) caused a significant (**P < 0.01) decrease in anti-FcRIIa antibody binding. Preincubation with A1AT (TNF/A1AT/Anti-FcRIIa antibody) had no significant effect on anti-FcRIIa antibody binding. Results are expressed as means ± SEM from 13 separate healthy donors.

Primed neutrophils from healthy volunteers preincubated with A1AT gave a significant drop of 70 ± 2.6% in anti-PR3 IgG Fab fragment binding, which is specific for the PR3 binding site only (P < 0.01, n = 7; Figure 2B). The binding of FcRIIa mAb to TNF--primed neutrophils from healthy volunteers exhibited a significant decrease in FcRIIa receptor antibody binding in comparison with the unprimed/control group. Values for the control group were 90.3 ± 12.6 MCF as compared with 48.9 ± 5.5 MCF in the primed group (P < 0.01, n = 7). Preincubation with A1AT had no significant effect on FcRIIa mAb binding as compared with the primed group (Figure 2C). These results suggest that at a cellular level A1AT blocks anti-PR3 IgG binding by blocking access to PR3 and not FcRIIa. This blocking of PR3 prevents the PR3-FcRIIa crosslinkage by anti-PR3 IgG required for neutrophil activation.

SLPI Has No Effect on Anti-PR3 IgG Binding and Induced Oxidative Burst in TNF--Primed Neutrophils

To ensure that the A1AT effect observed was not due in part to an anti-NE effect on anti-PR3 IgG binding or induced oxidative burst, the effect of SLPI on anti-PR3 IgG binding and activation of primed neutrophils from healthy volunteers was examined. Primed neutrophils preincubated with SLPI did not demonstrate a decrease in anti-PR3 IgG binding (Figure 3A). Similarly, SLPI had no effect on anti-PR3 IgG-induced activation of primed neutrophils (Figure 3B). This further suggests that the effect A1AT is exerting is due to anti-PR3 activity rather than anti-NE activity.

View larger version (23K): [in this window] [in a new window]   Figure 3.   (A) SLPI has no effect on anti-PR3 IgG binding to primed human neutrophils. Unprimed and TNF--primed neutrophils were incubated with anti-PR3 IgG (1 µg/ ml) for 30 min at 4°C. FITC-labeled secondary antibody was then added for 30 min at 4°C before quantitative flow cytometry. Separately, a group of primed cells was incubated with either SLPI (1 mg/ml) or A1AT (1 mg/ml) for 30 min at 37°C before the addition of anti-PR3 IgG. This was followed by addition of FITC-labeled secondary antibody and quantitative flow cytometry. In contrast to preincubating primed cells with A1AT (TNF/A1AT/Anti-PR3 IgG; *P < 0.01), SLPI preincubation (TNF/ SLPI/Anti-PR3 IgG) had no effect. Results are expressed as means ± SEM from 12 separate healthy donors. (B) SLPI has no effect on anti-PR3 IgG-induced burst activity on primed human neutrophils. Primed cells were incubated with anti-PR3 IgG (1 µg/ml) for 30 min at 37°C. A separate group of primed cells was incubated with either SLPI (1 mg/ml) or A1AT (1 mg/ml) for 30 min at 37°C before the addition of the anti-PR3 IgG and subsequent assessment of neutrophil oxidative burst activity. SLPI preincubation (TNF/SLPI/Anti-PR3 IgG) had no effect on anti-PR3 IgG-induced burst activity, in contrast to A1AT (**P < 0.01). Results are expressed as means ± SEM from 11 separate healthy donors.

PR3 ELISA

The results in Figure 4 show that the anti-PR3 IgG antibody detected PR3 by ELISA, as previously demonstrated (25). In addition, PR3 in the presence of SLPI was also detected, indicating that SLPI does not inhibit binding of the anti-PR3 IgG to PR3. However, PR3 complexed to A1AT was not detected using the anti-PR3 IgG. This result indicates that A1AT complexation to PR3 obscures the binding site for the anti-PR3 IgG antibody and therefore decreases the interaction of the antibody with PR3. This mechanism may explain the results of the earlier experiments in which oxidant burst and anti-PR3 binding to primed neutrophils were reduced in the presence of A1AT.

View larger version (22K): [in this window] [in a new window]   Figure 4.   PR3 IgG ELISA detection of PR3 is reduced in the presence of A1AT. PR3 was reacted with a 10-fold excess of either A1AT or SLPI and then probed for PR3 using the anti-PR3 IgG 12.8. Primary antibody binding was detected with HRP- labeled anti-mouse secondary antibody for 1 h, followed by substrate development with ABTS. Results are expressed as means ± SEM from three separate experiments.

The Effect of A1AT on Anti-PR3 IgG Binding and Activation of Neutrophils Isolated from Patients with Active WG

Unprimed neutrophils from patients with active WG demonstrated a high level of PR3 cell-surface expression reflected by a high level of anti-PR3 IgG binding compared with controls/unprimed neutrophils from healthy volunteers (Figure 5A). Mean PR3 expression was 327.1 ± 112.0 MCF in the active WG group, as compared with 25.19 ± 5.64 MCF in the control/unprimed neutrophils. PR3 expression had a wide range of expression in this group of patients; minimal value was 121.0 MCF, ranging to 877.7 MCF. Preincubation of these WG neutrophils with A1AT led to a significant decrease in anti-PR3 IgG binding. Binding was reduced from 327.1 ± 112.0 to 40.33 ± 18.2 MCF (P < 0.05, n = 6).

View larger version (19K): [in this window] [in a new window]   Figure 5.   (A) A1AT reduces binding of anti-PR3 IgG binding to neutrophils isolated from patients with active WG. Neutrophils from patients with biopsy-proven, active WG and from healthy control subjects were examined for PR3 cell-surface expression. Cells were incubated with anti-PR3 IgG (1 µg/ml) for 30 min at 4°C. One group of WG neutrophils was incubated with A1AT (1 mg/ml) for 30 min at 37°C before the addition of anti-PR3 IgG. FITC-labeled secondary antibody was added for 30 min at 4°C before quantitative flow cytometry. PR3 expression in the active WG patients (Anti-PR3 IgG WG Patient) was increased in comparison with unprimed normal donors. Preincubation of WG neutrophils with A1AT (A1AT/Anti-PR3 IgG WG Patient) significantly reduces anti-PR3 IgG binding; *P < 0.05. Results are expressed as means ± SEM from six separate WG patients. (B) A1AT reduces anti-PR3 IgG-induced activation on neutrophils from patients with active WG. Neutrophils were isolated from patients with biopsy-proven, active WG. When these cells were examined for PR3 expression, a wide range of PR3 expression was noted. The neutrophils were primed with TNF- (2 ng/ml) to maximize PR3 expression and to optimize any A1AT effect that may be seen. These primed cells were then incubated with anti-PR3 IgG (1 µg/ml) for 30 min at 37°C. One group of primed neutrophils was incubated with A1AT (1 mg/ml) before the addition of anti-PR3 IgG. Oxidative burst was measured as described previously. Preincubation of the primed WG cells with A1AT (TNF/A1AT/Anti-PR3 IgG WG Patient) caused a significant reduction in anti-PR3 IgG-induced activation; *P < 0.05. Results are expressed as means ± SEM from five separate WG patients.

In the experiments evaluating WG neutrophil oxidant burst activity, TNF--primed neutrophils from patients with active WG in the absence of anti-PR3 IgG did not demonstrate a significant difference in neutrophil oxidant burst activity when compared with unprimed neutrophils from healthy volunteers (Figure 5B). Addition of anti-PR3 IgG to primed WG neutrophils led to an increase in neutrophil oxidant burst activity. Preincubation with A1AT before the addition of anti-PR3 IgG demonstrated a significant drop in anti-PR3 IgG-induced burst. Reductions of 56 ± 7.2% in the A1AT-pretreated group were noted (P < 0.05, n = 5; Figure 5B).

These experiments demonstrate that exogenous A1AT can inhibit anti-PR3 binding and activation in neutrophils from patients with active WG in a fashion similar to that seen in neutrophils from primed healthy volunteers.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The neutrophil has long been suspected of playing a central role in the pathogenesis of WG. The azurophilic granules of neutrophils contain abundant amounts of PR3, the primary antigen for c-ANCA antibodies found in the circulation of individuals with WG. A1AT is the physiologic inhibitor of PR3 and the related serine protease NE; and in the WG population as a whole there is an increased incidence of A1AT deficiency, raising the possibility of A1AT playing a role in the pathogenesis of WG.

In this study we have demonstrated that neutrophils isolated from patients with active WG express increased cell-surface levels of PR3 compared with healthy controls. A1AT can block binding and activation of neutrophils by anti-PR3 IgG from both healthy controls and individuals with WG. A1AT appears to mediate its effect by binding to the PR3 site on the neutrophil, as seen by the inhibition of binding of PR3 site-specific anti-PR3 IgG Fab fragments to primed neutrophils. This is further confirmed by PR3 ELISA. Previous studies have shown that crosslinking of both PR3 and FcRIIa is required for neutrophil activation (14, 15), raising the possibility that A1AT prevents this crosslinking by binding to PR3. If A1AT were mediating its effect in part by acting on proteases other than PR3, such as NE, then it would be anticipated that SLPI, the specific NE inhibitor, would have a similar inhibitory action on anti-PR3 IgG. However, in contrast to A1AT, SLPI had no effect on anti-PR3 IgG binding or cell activation, nor did it interfere with anti-PR3 IgG binding to PR3 in an anti-PR3 IgG ELISA. These results strongly suggest that A1AT mediates its effects through an anti-PR3 action with no effect on the FcRIIa site.

The results of this study may go some way to explain the more severe clinical manifestations observed in WG patients who are A1AT-deficient. Those WG patients who have the PiZZ phenotype, with A1AT levels of 10 to 15% (27), have a much more aggressive vasculitis than do WG patients who have the MZ or MM phenotype (21). This may be due, in part, to a decreased ability of a reduced A1AT level to inhibit c-ANCA activation of neutrophils, with resultant tissue damage.

However, the majority of WG patients who are A1AT-deficient are of the PiMZ phenotype, and studies of these patients report normal A1AT levels when these patients have an active vasculitis (19, 22). Despite this "normal" A1AT level, those WG patients who have the PiMZ phenotype have a much worse clinical outcome (21, 22). This raises the questions: What is the defective link in the A1AT interaction with PR3 in active WG? Is the PR3 in those patients with active disease abnormal, thereby preventing normal A1AT binding to the dysfunctional PR3 expressed by WG cells? Or is the A1AT in the acute WG patient dysfunctional and thus unable to complex with normal PR3? In this study, we have demonstrated that exogenous A1AT can inhibit anti-PR3 IgG binding and activation of neutrophils from patients with active WG, thereby demonstrating that functional exogenous A1AT can have an anti-PR3 effect on the neutrophils from these patients. This suggests that the PR3 expressed on these WG neutrophils is available for complexing with A1AT.

Esnault (28) has proposed that generation of reactive oxygen species by leukocytes secondary to an ongoing infection may result in oxidative inactivation of A1AT with the generation of a localized relative deficiency of A1AT. There are in vitro studies which demonstrate that activated neutrophils and neutrophil products, such as hydrogen peroxide, can inactivate A1AT (29). Clearly, oxidative inactivation of local A1AT by c-ANCA-activated neutrophils would be more detrimental to the anti-PR3 activity of A1AT in those WG patients who are PiMZ, and especially in those who are PiZZ, which may explain the more aggressive vasculitis observed in these individuals (21). It may also explain why a serum level of A1AT, not normally recognized as severely deficient, may be inadequate to protect the body against the products of inflammation.

The present study complements the work by Ralston and colleagues (18), who demonstrated that A1AT can inhibit anti-PR3 IgG-mediated IL-8 production on primed monocytes, and does suggest an anti-inflammatory role for A1AT. However, the exact role of the monocyte in the vasculitic process of WG is unclear. It has been demonstrated that monocytes from patients with active WG express PR3 at a level similar to that of monocytes from healthy control subjects, which is in sharp contrast to the expression of PR3 on neutrophils from the patients described in this study. This supports the hypothesis that the neutrophil is the primary leukocyte "target" for anti-PR3 antibodies in WG (8).

The demonstration of the ability of exogenous A1AT to block the activation of WG neutrophils as seen in this study has clear implications suggesting potential beneficial effects of A1AT augmentation therapy in this group of patients. Supplemental A1AT could serve to inhibit c-ANCA activation on neutrophils and monocytes, as well as serving to "mop up" other products of inflammation, such as local release of NE and PR3 by leukocytes.

In summary, we have demonstrated that A1AT can inhibit anti-PR3 IgG binding and induction of oxidative burst in primed neutrophils from healthy volunteers. This effect of A1AT appears to be mediated by A1AT preventing anti-PR3 IgG binding to the PR3 on the surface of the neutrophil, which in turn prevents PR3-FcRIIa crosslinkage and cell activation. Moreover, this inhibitory effect of A1AT on anti-PR3 IgG binding and activation was also seen in neutrophils isolated from patients with active WG. We have also demonstrated that exogenous A1AT can inhibit anti-PR3 IgG neutrophil activation in the WG patients, suggesting that the primary defect is not in PR3. This study goes some way to explain the more aggressive vasculitis seen in WG patients who are A1AT-deficient and suggests a possible therapeutic role for A1AT augmentation therapy in active WG.