The etiology and pathogenesis of Wegener's granulomatosis is still elusive. Autoimmune phenomena, most probably in response to infectious agents, have been discussed as the initiating event1,2. There is ample evidence for activation of the immune system, including activation of T lymphocytes3,4,5,6,7, autoantibody formation, particularly to polymorphonuclear neutrophil (PMN)-derived cytoplasmic antigens (ANCA) [reviewed in 8–10], and for cytokine production11,12. Also, the effector cells are activated, and enhanced cytokine production by monocytes has been reported13, as well as increased adhesion of PMN14. Moreover, studies with isolated PMN showed activation by ANCA15,16,17,18. How these events are related to each other and how they contribute to the pathogenesis of the disease is still under investigation.
We recently described expression of major histocompatibility complex (MHC) class II antigens on PMN of patients with clinically active Wegener's granulomatosis19. This finding was of particular interest, because MHC class II antigens are normally found only on professional antigen presenting cells, such as dendritic cells, B cells, monocytes, or macrophages, but not on PMN of healthy donors. Since the only known function of MHC class II molecules is the presentation to T cells of antigens, our data provided first evidence that under pathological conditions PMN might acquire the capacity to function as accessory cells for T cell activation.
Our current study addressed the question of whether MHC class II expression was characteristic for Wegener's granulomatosis or whether it also occurred in patients with microscopic polyangiitis; moreover, we tested whether PMN acquired CD80 and CD86 (also known as B7.1 and B7.2). As co-stimulatory molecules, both CD80 and CD86 bind to CD28, one of the most relevant signal-transducing events in T cell activation20,21.
The consequence of MHC class II expression on PMN was tested with PMN of healthy donors. We found that PMN were able to present to T-cell superantigens and peptide antigens, and that they induced T cell proliferation in a MHC class II-restricted manner.
METHODS
Patients
After approval by the Ethikkommission (Ethics committee) of the University Heidelberg Hospital and after having obtained informed consent, patients with Wegener's granulomatosis or microscopic polyangiitis (MPA) attending the renal unit of the Heidelberg University Hospital were included in a prospective study. Wegener's granulomatosis was diagnosed according to the definition of the Chapel Hill conference22 and to the American College of Rheumatology (ACR) criteria23. Disease activity was determined by the "Birmingham Vasculitis Activity Score" (BVAS)24. In all patients, ANCA titers (normal value <1:20), proteinase 3 (PR3)-ANCA (<10%) and myeloperoxidase (MPO)-ANCA (<2 U) were measured as reported elsewhere25,26. Laboratory tests were performed by routine methods (Hitachi autoanalyzer; Hitachi, Tokyo, Japan; blood cell counts by Coulter counter; Coulter Electronics, Hialeah, FL, USA). Urine analysis was done by test strips (Combur 8 test kit; Boehringer Mannheim, Mannheim, Germany), by Addis Count and by measurement of 24-hour protein excretion with Biuret; urinary sediment was analyzed by phase contrast microscopy.
Patients with vasculitis and additional complications including acute infections as detected by x-ray, blood cultures, antiviral IgM titers or by measuring procalcitonin levels >1 ng/mL27 were excluded, as were patients with chronic inflammatory disease (cholecystolithiasis, inflammatory bowel disease, bronchitis, osteomyelitis), patients with kidney transplants or patients on dialysis.
The first patients entered the study in September 1996; the study was closed in June 2000. Only patients with systemic disease and more than two organ involvement were included in the study. Many patients were seen repeatedly, usually in 8- to 12-week intervals, or when the disease became active again or other complications arose. Data on patients and clinical findings are summarized in Table 1. In short, the following criteria were used to define organ involvement: (a) skin, such as palpable purpura, ulcers or nodules; (b) musculoskeletal, such as arthralgias, arthritis or polymyalgia; (c) eyes, such as (epi-)scleritis, keratitis, uveitis or retinal vasculitis; (d) ear/nose/throat (ENT), such as nasal mucosal ulceration with or without epistaxis, otitis media, sinusitis or tracheal stenosis; (e) lower respiratory tract (lung), such as hemoptysis, alveolar hemorrhage, respiratory failure, asthma, coin lesions or radiographic proof of pulmonary infiltrates in the absence of infectious etiology; (f) kidney, where renal involvement was defined as elevated serum creatinine, decreased creatinine clearance, microscopic hematuria (>5 red blood cells per high power field) and/or red cell casts and/or proteinuria (>0.3 g/dL); the majority of patients had biopsy-proven pauci-immune crescentic glomerulonephritis and in two patients mesangioproliferative GN was seen; (g) nervous system involvement included seizures, mononeuritis multiplex or peripheral neuropathy; (h) heart, involvement comprised pericarditis, myocardial infarction or cardiomyopathy. Data of patients with Wegener's granulomatosis gathered between September 1996 and April 1998 were already published19. In the present study, only patients attending the renal unit between December 1998 and June 2000 were considered.
Table 1 - Laboratory findings and clinical characteristics of all patients participating in the study.
Full table The following groups of patients were formed according to the clinical presentation of patients with Wegener's granulomatosis: patients with newly diagnosed systemic active disease (BVAS >1; N = 10); patients developing a relapse under immunosuppressive therapy (N = 5); patients with inactive disease (BVAS 1 or 0) under immunosuppressive therapy (N = 25); and patients with inactive disease and receiving no immunosuppressive therapy (N = 20; Table 1).
Microscopic polyangiitis (MPA) was defined by the presence of necrotizing crescentic glomerulonephritis associated with extrarenal vasculitis involving small vessels. None of the patients had asthma or granuloma formation. Pulmonary manifestation was dominated by lung hemorrhage with characteristic chest radiographs. There were no nodules or cavitations. Patients with upper respiratory tract involvement were not considered to have Wegener's granulomatosis when there was no evidence for granulomatous infection or invasive bony disease. Most patients had radiographic features of sinusitis. The upper respiratory involvement was generally mild as described also by others22,28. All patients seen between September 1996 and January 2000 were included, and as described for patients with Wegener's granulomatosis, grouped according to the clinical presentation when they were first seen. Nine patients with newly diagnosed MPA are included in the study, as well as two patients developing relapses under immunosuppressive therapy, and 16 patients with inactive disease Table 1.
Immunosuppressive therapy in patients with Wegener's granulomatosis and MPA consisted of a combination of methylprednisolone, cyclophosphamide or azathioprine.
Patients with acute bacterial infections.
For comparison, twenty patients (recruited from January to September 1999) with acute bacterial infections, but no symptoms of autoimmune or chronic inflammatory disease, were included in the study. Diagnosis was based on bacterial culture, high plasma levels of C reactive protein (>8 mg/L), leukocytosis (>10 
103/
L), procalcitonin levels higher 1 ng/mL, and fever.
Healthy individuals.
For comparison, PMN of healthy individuals, comparable with regard to age and sex were analyzed. Twenty-five donors were tested within the same time as the patients with Wegener's granulomatosis, 28 during the time the MPA study was performed. The two control groups were not different from each other.
Determination of surface antigen expression by cytofluorometry
The respective surface molecules were detected by double-labeling with fluorescein (FITC)-conjugated antibody to CD66b as marker for PMN, CD14 as marker for monocytes and phycoerythrin (PE)-conjugated antibodies to MHC class II (anti-DP, -DQ, -DR; Serotec, purchased from Biozol, Eching, Germany), CD80 and CD86 (Serotec). For comparison, isotype-matched mouse IgGs (Serotec) were used in comparable protein concentrations. To avoid changes in surface expression due to handling procedures, antibody-labeling was performed in whole blood within two hours after withdrawal. As described previously no major changes were observed within that time19. Results are expressed as % CD66b positive PMN or as mean fluorescence units (MFI) for monocytes. Normal values were established from analyzing 53 healthy donors. The 90% percentile was set as threshold and values higher than 5.5% for MHC class II; 5% for CD80; 12% for CD86; and 12% for CD64 were considered positive. For cytofluorometry of isolated PMN 5 105 to 1 106 cells were suspended in phosphate-buffered saline (PBS) containing 1% bovine serum albumin (BSA) and 0.1% Na-azide and incubated with the respective antibodies as described above.
Statistical analysis
The results for MHC Class II, CD80 and CD86, and CD64 are presented as box blots, with the box containing 50% of the values. The highest, lowest and the median values are depicted. Differences between the patients' group and the healthy controls were analyzed using the Wilcoxon test. Patients with bacterial infections (N = 20) were considered as a separate group.
Isolation of polymorphonuclear neutrophils and of T cells
From peripheral blood of healthy donors or of selected patients mononuclear cells and PMN were separated by centrifugation on PolymorphPrep™ (Nycomed, Oslo, Norway). For the functional experiments (antigen presentation) and the reverse transcription-polymerase chain reaction (RT-PCR) the PMN fraction was further purified by adsorption to CD15-beads by magnetic cell separation (autoMACS) using the devices supplied by MiltenyBiotech (Bergisch-Gladbach, Germany). As judged by cytofluorometry with antibodies to CD66b/CD18 as a marker for PMN, the purity was greater than 99%. From the mononuclear cell fraction, T cells were further purified by adsorption to anti-CD3, B cells by anti-CD19 and monocytes by absorption to anti-CD14 using autoMACS (MylteniBiotech). The procedure yielded highly homogenous preparations as seen by cytofluorometry.
Cultivation of PMN
To induce MHC class II surface expression purified PMN (1 106/mL) were cultivated in AIM V™ (Gibco, Eggenstein, Germany) with 2.5% autologous serum (NHS), granulocyte macrophage-colony stimulating factor (GM-CSF) 50 U/mL and interferon gamma (IFN-
) 100 U/mL (Sigma) for the times indicated. Survival of PMN in culture was measured by propidium iodide staining and only experiments with more than 80% PMN surviving were considered reliable29. For cultivation of whole blood, samples were collected in heparin-coated tubes (Sarstedt, Nürmbrecht, Germany); portions of 1 mL were incubated with the cytokines in the concentrations and for the times indicated. Erythrocytes were lysed before analyzing surface expression of the respective antigens as described above.
Determination of MHC class II specific RNA.
RNA was isolated from PMN using a commercially available kit (Roche, Mannheim, Germany). MHC class II-specific RNA and for comparison 
actin-specific RNA were amplified by RT-PCR as described in the study of Radsak et al29.
T cell lines
T cell lines with specificity for elastase were established from a healthy donor. Isolated T cells were cultivated with autologous monocytes and B cells and 5 g heat-inactivated elastase (Calbiochem, Darmstadt, Germany). After five days in culture, T cells were separated and again cultivated with elastase and irradiated autologous monocytes and B cells. The procedure was repeated after 15 days. Then the cells were sub-cloned (100 cells/well) and cells with high proliferation rate for elastase in the presence of autologous mononuclear cells, and no reactivity with other antigens (tetanus toxoid, TT) or mononuclear cells alone were chosen for further propagation. After two more rounds of re-stimulation and subcloning, two lines (ELA5 and ELA10) were used for the experiments. Moreover, one T cell line with reactivity for TT (FRIT-T), established previously29 and two T cell lines with reactivity for Staphylococcus enterotoxin E (SEE); A37/2 and D894, established and given to us by D. Kabelitz) were used30.
T cell proliferation test
Purified PMN (1 105), B cells or monocytes were added to each well of a 96-well flat bottom plate. B cells were irradiated with 80 grays, monocytes with 60 grays, respectively. Then, 1 105 T-cells and elastase (0.2 g/well), TT (0.2 g/well; Calbiochem, La Jolla, CA, USA), or SEE (2 ng/well; Serva), respectively, were added. After co-incubation for four days at 37°C with 5% CO2, cells were pulsed by adding 1 Ci of 3H-thymidine (specific activity 5 Ci/mmol; Amersham Life Science, Braunschweig, Germany) to each well and harvested 14 to 16 hours later. 3H-thymidine incorporation into DNA was measured with a scintillation counter (Packard, Groningen, Netherlands) and expressed as counts per minute (cpm). The values represent the mean 
SD of 6 to 12 parallel wells. In some experiments, nonadherent cells were harvested after four days, and pulsed with 3H-thymidine for 14 to 16 hours. An aliquot of those cells was subjected to cytofluorometry and identified as T cells by the use of an anti-CD3 (<95% CD3 positive cells). For inhibition experiments, monoclonal antibodies (mAbs) against intercellular adhesion molecule-1 (ICAM-1), CD80, CD86, and MHC II were used in concentrations of 1 g/mL (0.2 g/well). IgG1, IgG2a and IgG2b were used as isotype controls in the same final concentration. Antibodies were added together with the antigens and were present throughout the entire incubation time.
RESULTS
Polymorphonuclear neutrophils of patients with active Wegener's granulomatosis are phenotypically altered when compared to the PMN of healthy donors. By cytofluorometry, using forward/side scatter analysis, a subpopulation of PMN (5 to 20%) was seen that appeared larger and more granular. Moreover, the PMN expressed MHC class II antigens, CD80 and CD86, with the altered PMN being more positive. After immunosuppressive therapy, the morphological changes as well as the expression of MHC class II, CD80 and CD86 decreased [an example is shown in Figure 1]. Alteration of PMN morphology and expression of MHC class II, CD80 and CD86 was seen in 9 of 10 patients with active Wegener's granulomatosis Table 2. The patient, who did not express MHC class II nor CD80 or CD86 had pANCA (titer 1:80) and antibodies to MPO (13.2; normal value <2) but not to PR3. By nasal biopsy, however, granuloma was detected, leading to the diagnosis of Wegener's granulomatosis.
Figure 1.
Analysis of polymorphonuclear neutrophils (PMN) of a patient with active Wegener's granulomatosis by flow cytofluorometry. (A) To 100 L heparinized blood, FITC-labeled antibodies to CD66b (as a marker for PMN) together with PE-labeled antibodies were added to HLA-DP, HLA-DQ, HLA-DR, CD80, or CD86. Then the cells were analyzed by cytofluorometry. At least 12,000 events were counted. The forward-side scatter analysis (left panel), revealed two major PMN populations (identified by anti-CD66b), termed R1 (84% of total PMN) and R2 (16% of total PMN), monocytes (R3) and lymphocytes (R4). The PMN of both R1 and R2 expressed HLA-DP, HLA-DQ, HLA-DR, CD80, and CD86 as measured by double-labeling cytofluorometry (right panels). The population in R2 contained more positive PMN with higher expression of the respective antigens. (B) The same experimental set up with blood of the same patients ten days after starting the immunosuppressive therapy. The cells in R2 had declined to 2% of total PMN count. (C) The analysis of blood of a healthy donor. Only 0.2% of PMN were detected in the area of R2.
Table 2 - MHC class II, CD80, CD86 and CD64 expression on PMN of patients with Wegener's granulomatosis, MPA and healthy donors.
Full table In patients with clinically inactive disease, with or without immunosuppressive therapy, only low levels of MHC class II or CD80 expression were seen in a few patients. Expression was significantly lower than in patients with active disease (largest P value 2.4 10-4 for MHC class II, and 0.002 for CD80) and was not different from healthy controls. PMN of patients with relapse under immunosuppressive therapy did not express MHC class II nor CD80 (Table 2 and Figure 2).
Figure 2.
Expression of major histocompatibility complex (MHC) class II, CD80, CD86 and CD64 on PMN of patients with Wegener's granulomatosis, microscopic polyangiitis (MPA) or bacterial infections. Expression was measured as % positive PMN. Box blots are shown for each surface antigen. The boxes contain 50% of the values. The highest, lowest and median values are shown for: Wegener patients with active disease without therapy (active, N = 10) patients suffering a relapse under immunosuppressive therapy (relapse N = 5), patients with inactive disease and immunosuppressive therapy (N = 25), patients with inactive disease and no therapy (N = 20), healthy donors (N = 25), patients with MPA and active disease (active, N = 9), with inactive disease (N = 16) and healthy donors (N = 28), and patients with bacterial infections (acute, N = 20). For MHC class II and CD80 expression the superscript symbol (*) indicates that the respective group differed from all the others (P < 0.001 for MHC class II and P < 0.01 for CD80); (#) the value for patient 6 with pANCA and MPO antibodies. CD86 expression was significantly enhanced in all patients with Wegener's granulomatosis (*largest P = 0.016), as was CD64 (*largest P = 0.006). In patients with MPA, CD64 expression was higher in patients with active disease than in patients with inactive disease or healthy donors (**P = 0.02 and 1.2 10-6, respectively). PMN of patients with bacterial infections (N = 20) did not express MHC class II, CD80 or CD86. Only expression of CD64 differed form that of the healthy donors (***largest P = 3.3 10-6).
High expression of CD86 was seen on PMN of patients with Wegener's granulomatosis irrespective of disease activity, but was not found on PMN of healthy donors (largest P value 0.016). Though the median value for CD86 expression was higher on PMN of patients with active disease than that for patients with inactive disease, the groups did not differ from each other due to the large variations Figure 2. The follow-up studies of single patients, however, showed that expression of CD86 declined rapidly under immunosuppressive therapy as did expression of MHC class II and of CD80 (Figure 1a, data of five patients are summarized in Table 3). In parallel, the abundance of MHC class II-specific mRNA declined, consistent with the fact that MHC class II antigens are synthesized de novo Figure 1b.
Table 3 - Follow-up study in five patients entering the study with active systemic disease of Wegener's granulomatosis (BVAS >1) and receiving immunosuppressive therapy.
Full table The CD86 expression on PMN correlated to some extent with the MHC class II expression (R = 0.7). Other linear correlation, for example, of MHC class II and CD80 surface expression, or abundance of either MHC II, CD80, CD86 and autoantibody titers were not found.
CD64 was expressed on the PMN of Wegener patients. The expression, however, did not differ between patients with active disease or patients with inactive disease, but was different from that of healthy donors Figure 2.
Twenty-four patients with MPA were analyzed Table 1. In two of the nine patients with active disease, minor MHC class II expression (7.4% and 6.45%, respectively) on PMN was seen. When analyzed as a group, there was no difference between patients with active disease, inactive disease or healthy controls Figure 2. The same observation was made for CD80 and CD86 expression. None of the patients developing a relapse expressed MHC class II, CD80 or CD86 on PMN. In all patients with active disease, however, up-regulation of CD64 was seen (Figure 2 and Table 2).
As an example for acute inflammatory disease, patients with bacterial infections were included in the study. Neither MHC class II antigens nor CD80 were found. CD86 was only mildly enhanced in some patients. The median values for MHC class II, CD80 and CD86 expression did not differ from that of healthy controls. The PMN, however, expressed high levels of CD64, indicating activation Figure 2.
In addition to PMN, monocytes were analyzed with regard to MHC class II expression. Monocytes of healthy donors express MHC class II molecules. From 40 healthy donors a median value for the mean fluorescence of 112 MFI was determined. Many, but not all, patients with Wegener's granulomatosis showed enhanced levels of MHC class II expression on monocytes. The median value for patients with active disease, however, was not statistically different from that of patients with inactive disease Figure 3. Expression of MHC class II on monocytes determined as MFI did not correlate linearly to MHC class II on PMN (R = 0.45), however, monocytes of patients with MHC class II-positive PMN (>5.5%) expressed significantly more MHC class II than patients with MHC class II-negative PMN Figure 3. Follow-up studies with single patients showed that, together with MHC class II expression on PMN, the MFI of monocytes declined Table 3. On monocytes of patients with active MPA no up-regulation of MHC class II was detectable (data not shown).
Figure 3.
Expression of MHC class II antigens on monocytes. (A) Box blots show the expression of MHC class II (measured as MFI) of the patients with Wegener's granulomatosis (c.f. Figure 2) and of healthy donors. The patient groups were not different from each other, but all differed from the healthy donors (largest P value 0.005). (B) MHC class II expression on PMN did not correlate linearly with MHC class II expression on monocytes (insert); patients, however, with PMN expressing MHC class II (>5.5% positive PMN) had significantly higher MHC class expression also on monocytes.
Full figure and legend (30K)Various questions arose from the in vivo observations. The two most pertinent questions were how PMN are activated to acquire MHC class II, CD80 and CD86, and what would be the functional consequences of this activation.
In that respect, induction of MHC class II expression on PMN of healthy donors was tested. A first set of experiments was performed with highly purified PMN derived from healthy donors. When PMN were co-cultivated with autologous T cells and MHC class II expression, the expression of CD80 and CD86 increased within 24 to 72 hours (Figure 4 and Table 4). Essentially similar results were obtained when PMN were cultivated with heterologous T cells or cells of the various T cell lines.
Figure 4.
Induction of MHC class II expression on PMN by co-culture with T cells. Highly purified PMN (106) were co-cultivated with T-cells (106) for 48 hours. By forward-side scatter analysis the two cell population can be easily distinguished and identified as PMN (R1) or T cells (R2) by the use of monoclonal antibodies. When setting the gate around the PMN population, it can be seen that a portion of PMN had acquired MHC class II, and CD86 (filled peaks; the lines show the respective isotype controls).
Full figure and legend (57K)Table 4 - Induction of MHC class II, CD80 and CD86 synthesis on PMN after co-culture with autologous T cells.
Full table As described previously, culture of PMN with either T cell supernatants or INF- alone or in combination with GM-CSF induced synthesis and surface expression of MHC class II29. Following culture, the PMN showed a similar shift in the forward-side scatter analysis as the PMN derived from the patients with active Wegener's granulomatosis, and again, similarly to the in vivo observation, the shifted PMN expressed more of MHC class II per cell than PMN in the original position Figure 5; c.f. Figure 1.
Figure 5.
Induction of MHC class II antigens on PMN. PMN from a healthy donor were cultivated in the presence of IFN- (100 U/mL) and GM-CSF (50 U/mL) for 48 hours. The forward-side scatter analysis revealed several changes of the original (O) PMN population. A portion of the PMN were dead or dying. In R4 mainly cellular debris is seen in R3 apoptotic cells as judged by propidium iodide (PI) exclusion. In addition to the viable cells in R2 (R2 corresponds roughly to the original position) an additional population was seen (R1) comprising 15% of cells in R2. PMN in R1 and R2 expressed MHC class II, in contrast to PMN before culture. The binding of IgG, not shown here, was not altered when the three cell populations were compared.
In addition to IFN- and GM-CSF, a number of other cytokines including interleukin (IL)-2, IL-4, IL-6, and IL-8, G-CSF and tumor necrosis factor-
(TNF-) were tested for their ability to induce MHC class II expression. None of the cytokines either alone or in combination had effects on MHC class II expression.
To mimic the in vivo situation closer, a second set of experiments tested the induction of MHC class II synthesis on PMN in whole blood. Again samples of different healthy donors blood were used and it was seen that IFN- and GM-CSF increased the levels of MHC class II expression from 2.8 1.4% in freshly drawn blood to 25.5 15.5% after 24 hours and 33.9 8.7% positive PMN after 48 hours (N = 7; Figure 6).
Figure 6.
Induction of MHC class II antigens on PMN of donors or patients in whole blood. In blood samples of patients with Wegener's granulomatosis under immunosuppressive therapy (N = 11), without therapy (N = 5) or of healthy donors (N = 7) MHC class II expression was measured by cytofluorometry immediately after withdrawal of blood and again 48 hours after incubation with IFN- (100 U/mL) and GM-CSF (50 U/mL). Before culture (
) MHC class II antigen was not expressed; following culture (
) the PMN acquired MHC class II (largest P value 0.001); PMN of patients without therapy or of healthy controls acquired significantly more MHC class II antigens than PMN of patients under immunosuppressive therapy.
Data with cultivated PMN of healthy donors had provided evidence that acquisition of MHC class II, CD80 and CD86 depended on de novo protein synthesis, as it could be inhibited by cycloheximide29. Apparently, also in vivo, those antigens are produced by de novo synthesis, since specific RNA was found in patients with active disease only, but not in patients with immunosuppressive therapy or in healthy donors Figure 7.
Figure 7.
Detection of MHC class II by RT-PCR in PMN of a patient with active Wegener's granulomatosis. In PMN of a patient with active disease, with MHC II specific primer pairs a PCR products was found (lane 1), which could not be detected 10 days after onset of immunosuppressive therapy (lane 2). Lane 3 shows the MHC class II PCR product of IFN- (100 U/mL) treated PMN. M is the size marker.
The observation that MHC class II expression decreased rapidly in patients undergoing immunosuppressive therapy suggests that steroids might prevent up-regulation of MHC class II in vivo. To test that assumption, whole blood of patients under immunosuppressive therapy with steroids was cultivated for 24 to 72 hours in the presence of IFN- and GM-CSF. At all times the MHC class II surface expression was considerably less when compared to the PMN of patients with Wegener's disease in remission without therapy or to the PMN of healthy donors Figure 6.
While the data support the hypothesis that T cells or T cell-derived cytokines such as IFN- or GM-CSF induce MHC class II expression on PMN, the presence of those cytokines in patients' sera could not be demonstrated unequivocally. IFN- was only found in the plasma of two of 12 patients with active Wegener's granulomatosis and GM-CSF in four other patients.
Since the only known function of MHC class II antigens is the presentation of peptide antigens to T cells, we tested whether MHC class II-positive PMN were able to function as accessory cells for T cell activation in an antigen-dependent, MHC class II-restricted manner. In a first set of experiments, MHC class II-positive PMN and T cells derived from patients with active Wegener's disease (N = 3) were co-cultivated with or without antigen. A high proliferation rate of T cells was seen, regardless of the presence of antigen or of PMN of monocytes, respectively. Since the T cells of healthy donors either do not or only to a minor degree do proliferate without antigens or mitogens, the data are indicative of a profound T cell activation in vivo (data not shown). Since these findings precluded the use of patient-derived cells for the in vitro studies, the following experiments were carried out with PMN of healthy donors, induced to express MHC class II, CD80 and CD86 by INF- and GM-CSF for 24 hours. The proliferative response of autologous, peripheral T cells was tested with two different peptide antigens, tetanus toxoid (TT) and elastase. With either elastase or TT only a minor, but statistically significant T cell proliferation was seen Table 4. To enhance the frequency of antigen-specific T cells, lines with reactivity for either elastase or TT were established from a healthy donor. With the respective antigens and autologous PMN considerable T cell proliferation could be induced, while heterologous PMN failed to induce T cell proliferation (data for 2 elastase-specific lines, ELA5 and ELA10, and for one TT-specific line, FRIT-T are summarized in Table 5). Of note was that proliferation of the elastase-specific T cells also occurred without the addition of elastase, whereas proliferation of the TT-specific T cell line required TT. An essentially similar pattern of T cell activation was observed when monocytes were used in place of PMN. On the other hand, with B cells as with antigen-presenting cells, proliferation depended on the addition of elastase Table 5. Though PMN are considered to be non-dividing cells, an experiment was performed where the non-adherent cells were harvested after four days of culture. Then, 3H-thymidine was added to the remaining cell layer and as well as to the harvested cells. Incorporation of 3H-thymidine was only seen in the non-adherent cells that were identified as T cells by cytofluorometry.
Table 5 - Proliferation of peripheral T cells or antigen-specific T cell lines in response to antigen presented by PMN, monocytes or B cells.
Full table With either T cell line, PMN-induced proliferation could be inhibited by antibodies to either MHC class II (anti-DR; 40%) or to ICAM-1 (65%).
Since previous data by others described the relevance of Staphylococcus aureus-derived superantigens (SE) in relapses of Wegener's granulomatosis, we tested whether MHC class II-positive PMN of either patients with Wegener's granulomatosis (in relapse and without immunosuppressive therapy) or of healthy donors would present SE and induce T cell proliferation. In both cases proliferation of T cells was found, dependent on MHC class II and ICAM-1 Table 6.
Table 6 - Proliferation of T cells and T cell lines in response to Staphylococcus enterotoxin E (SEE) presented by PMN or monocytes.
Full table DISCUSSION
In the course of ANCA-positive vasculitis the immune system is activated as seen by the up-regulation of receptors on immunocompetent cells, enhanced cytokine levels and production of autoantibodies3,4,5,6,7,8,9,10,11,12,13,14,31. The contributions of the respective activities to the pathogenesis of vasculitis are under intense investigation, and while a correlation to disease activity has been found particularly for autoantibody generation, a causal relationship has not yet been established8,9,10,32.
There is ample evidence for PMN activation during active disease: massive infiltration of PMN into glomeruli and small vessels of the interstitium was described; generation of PMN-derived oxygen radicals and lysosomal enzyme release were detected in correlation to disease activity16. Activation might be due to cytokines, especially to IL-8, which is found in inflammatory lesions33. Moreover, in vitro data show that PMN can be activated by ANCA9,15,18,34. Our data imply that T cell-derived mediators, such as interferon-, are decisive for PMN activation in vasculitis. We found recently that PMN of patients with Wegener's granulomatosis acquired MHC class II antigens when the disease was active19. MHC class II antigens are synthesized de novo upon stimulation of PMN with interferon- or GM-CSF29,35 and, as shown in the present study, also by direct contact with T cells.
In the presence of interferon- PMN escape—at least temporarily—their constitutive apoptosis36,37. In addition to MHC class II they acquire CD80 and CD8629, and they assume new functions, particularly superantigen and antigen presentation29,37,38, the latter being critically dependent on the presence of the co-stimulatory properties of CD80 and CD86.
Owing to the significance of CD80 and CD86, our previous study19 was extended to test for the presence of the co-stimulatory antigens CD80 and CD86 on PMN. Moreover, patients with MPA were included. The patients with Wegener's granulomatosis presented more severe impairment of renal function than the patients with MPA. The patients with MPA had less extrarenal involvement when compared to patients with Wegener's granulomatosis, an observation made before by others15,39. The activity score (BVAS), however, did not differ significantly between the two groups of patients Table 1. PMN of patients with active Wegener's granulomatosis displayed an impressive morphological change and expressed MHC class II, CD80 and CD86 in close correlation to disease activity Figure 1. There was only one exception: PMN of patient #6 did not acquire MHC class II, CD80 or CD86 Table 1. That patient had pANCA and antibodies to MPO, but not to PR3. Nevertheless, the detection of localized granuloma lead to the diagnosis of Wegener's granulomatosis.
Under immunosuppressive therapy, PMN resumed their normal morphological appearance, and expression of MHC class II, CD80 and CD86 declined, being in line with the concept that T cell-derived mediators are involved. In some patients with clinically inactive disease and without immunosuppressive therapy, some MHC class II, CD80 or CD86 expression was seen, suggesting low-level T cell activity. The assumption is supported by the observation that CD25-positive T cells are found also in patients with clinically inactive disease3. The fact that MHC class II was not elevated in patients with a relapse under immunosuppressive therapy is in line with the in vitro findings that PMN derived from patients under immunosuppressive therapy expressed considerably less MHC class II antigens than PMN of healthy donors or of patients without therapy. That a relapse of disease occurs without up-regulation of MHC class II does not rule out a role for class II-positive PMN in the pathogenesis of the disease, but makes participation in the early events rather unlikely.
Polymorphonuclear neutrophils of patients with MPA were not morphologically altered. Only 2 of the 11 patients with active disease showed minor MHC class II expression on PMN and no enhanced expression on monocytes. However, CD64, the high affinity receptor for IgG, was up-regulated in patients with active disease, indicating PMN activation. A similar observation was made with PMN of patients with acute bacterial infection: up-regulation of CD64 was observed, but not MHC class II Figure 2. The data indicate that induction of MHC class II expression and CD64 up-regulation are regulated independently of each other.
The activation pattern of PMN occurring in active Wegener's granulomatosis resembled that of PMN derived from healthy donors cultivated with T cell-derived mediators or T cells. Considering the presence of activated T cell in patients with active disease, T cells are the most likely source of those cytokines in vivo. The assumption is supported by the observation that monocytes are activated to express more MHC class II antigens. T cell-derived mediators, particularly IFN-, have long been known to affect monocyte MHC class II expression40. The failure to detect the cytokines in the plasma of patients with active disease does not rule out that possibility, since cytokines are rapidly bound to their respective target cells or to extracellular matrix proteins, and soluble circulating mediators usually reflect only the surplus. Thus, it is reasonable to assume that in vivo PMN are activated as consequence of a T cell response.
The most obvious alterations of PMN, extended lifespan and MHC class II, CD80 and CD86 expression, could contribute to the inflammatory process, because (1) escape from apoptosis is thought to be a decisive factor of the inflammatory response41, and (2) MHC class II bearing PMN could present antigen to T cells, and initiate or sustain an antigen-specific T cell response.
In support of a role of PMN in the induction of a specific immune response, we found that PMN were able to present antigens in a MHC class II-restricted manner. Since the high proliferation rate of T-lymphocytes derived from patients with active Wegener's granulomatosis precluded the use of those PMN for in vitro experiments, cells of healthy donors, which are known to acquire MHC class II upon stimulation, were used. Peripheral T cells proliferated only to a minor degree when co-cultivated with PMN or monocytes and TT or elastase as antigens, being in line with a low frequency of antigen-specific precursor cells. With established, antigen-specific T cell lines, in contrast, considerable T cell proliferation could be induced provided autologous PMN were used. Of special note was that the elastase-specific clones also were activated by PMN and monocytes in the absence of exogenously added elastase. One interpretation is that endogenous elastase, which is released from PMN or monocytes upon activation, is sufficiently abundant to serve as the antigen. The presumption is supported by the fact that B cells, which do not express elastase, induced T cell proliferation only when elastase was added.
Because of the notion that the presence of Staphylococcus enterotoxin coincides with relapses of Wegener's granulomatosis2, we tested whether PMN was able to also present Staph enterotoxin. Staph enterotoxin is a superantigen, so-called because it binds outside of the peptide-binding groove of the MHC class II and the antigen-specific domain of the T cell receptor, and consequently activates a large portion of T-cells, preferentially those with a V beta 2 domain42. In accordance with previous studies by us and others29,35, co-culture of PMN with T-cells and SE resulted in T cell proliferation. As expected, proliferation was dependent on MHC class II expression, but was not MHC class II restricted, that is, it occurred with autologous as well as heterologous, MHC class II non-identical T cells.
Taken together, our data demonstrate that by synthesizing and expressing MHC class II antigens as well as co-stimulatory molecules, PMN acquire the capacity to present superantigens and peptide antigens to T cells. Whether PMN by this newly acquired function participate in the pathogenesis of Wegener's disease or whether MHC class II expression merely reflects T cell activation is still a matter of speculation. Accordingly, the question remains open as to whether the differently activated PMN contribute to the differences seen between Wegener's granulomatosis and microscopic polyangiitis, or rather reflect activation of distinct cellular compartments. That MHC class II, CD80 and CD86 expression concurs with PR3-ANCA and the absence of those receptors with MPO-ANCA suggests a connection between PMN activation and the specificity of the autoantibody. A recent review by Franssen et al emphasizes the characteristic differences between anti-PR3- and anti-MPO-associated vasculitis by correlating the specificity of the autoantibody to clinical and histopathological findings39. In that context, an attractive hypothesis would be that MHC class II-positive PMN present endogenous PR3 to T-cells, thereby contributing to the generation of autoantibodies.
References
Acknowledgments
This work was supported by a grant from the Faculty for Medicine, University of Heidelberg. A portion of the data were presented at the 33rd Annual Meeting of the American Society of Nephrology in Toronto, Canada (Oct. 2000).
