Proteinase 3, the Target Antigen of - Europe PMC

American Journal of Pathology, Vol. 139, No. 4, October 1991 Copyight C American Association of Pathologists

Proteinase 3, the Target Antigen of Anticytoplasmic Antibodies Circulating in Wegener's Granulomatosis Immunolocalization in Normal and Pathologic Tissues

Mafthias G. Braun,* Elena Csernok,t Wolfgang L. Gross,t and Hans-K. Muller-Hermelink* From the Institute of Pathology, University of Wurzburg and the Department of Clinical Rheumatology

of the Medical University of Lubeck, and the Center for Rbeumatology,t Bad Bramstedt, Federal Republic of Germany

Proteinase 3 was identified as the target antigen of anticytoplasmic antibodies (ACPA/c-ANCA) in Wegener's granulomatosis (WG). Using the murine monoclonal antibody (MAb) WGM2, the authors localizedproteinase 3 on normal andpathologic tissue and in various human cell lines. Proteinase 3 was found in all granulocytes, mast cells and a subset of monocytes (2-10% of peripheral blood monocytes). No crossreactivity of the MAb with other cells or tissue components was observed Comparing various vasculitic and granulomatous tissues, WG granulomas contained more granulocytes. In tissue of WG, no crossreactivity of theMAb with tissue components such as blood vessel wall or endothelium was seen The monocytic line THP-1 and the promyelocytic line HL 60 contained proteinase 3. Since proteinase 3 was only found in neutrophil granulocytes, mast cells; and a subset of monocytes, the presented data provide further evidence thatgranulocytes play a central role in the pathogenesis of WG. (Am JPathol 1991,

139:831-838)

During the past few years autoantibodies directed against a cytoplasmic antigen of human neutrophils and monocytes (ACPA; synonyms: ANCA, c-ANCA) have been established as useful tools in the diagnosis and follow-up of Wegener's granulomatosis (WG).1" ACPA are found in approximately 95% of WG patients with ac-

tive disease,3'4 and ACPA titers in patient sera as determined by indirect immunofluorescence5 6 or ELISA7 correlate with disease activity. The target antigen of ACPA was identified as proteinase 3,8 a neutral serine proteinase of human neutrophil granulocytes. A murine MAb reactive with the affinity-purified antigen was used to localize proteinase 3 in normal and pathologic tissue and in various human cell lines. Ultrastructural analysis with the new MAb WGM1 directed against proteinase 3 confirmed that the antigen is localized in the azurophil granules and in smaller amounts on the plasma membrane of resting normal human polymorphonuclear leukocytes and in myeloperoxidase-positive granules of monocytes.9

Material and Methods

Production of MAb WGM2 Hybridoma technology was used to generate the MAb WGM2 directed against proteinase 3, the target antigen of ACPA, as described in detail elsewhere.9 Briefly, the antigen preparation for immunization was derived from human neutrophils by phorbol-myristate-induced degranulation, as described previously.10 Female CD2F1 mice were immunized by subcutaneous injection of 50 ,ug crude antigen in Freund's complete adjuvant (Difco Laboratories, Detroit, Ml). After 1 week, the mice were injected with a further 50 ,ug antigen in phosphatebuffered saline (PBS) daily for 5 days. The spleen cells were fused with the murine myeloma X63-Ag8.653. CulSupported by grants 01 VM 8622 of the Bundesministenum fur Forschung und Technologie and by the Verein zur Forderung der Erforschung rheumatischer Erkrankungen Bad Bramstedt e. v. Accepted for publication June 11, 1991. Address reprint requests to Dr. Matthias G. Braun, Abt. Klin. Rheumatologie, Rheumaklinik Bad Bramstedt, 2357 Bad Bramstedt, FRG.

831

832

Braun et al

AJP October 1991, Vol. 139, No. 4

ture supernatants were screened in an ELISA with the degranulation supernatant on the solid phase. Further selection was done using an ELISA with affinity-purified antigen.10 Two hybridomas, designated WGM1 (IgM-class) and WGM2 (IgG1 -class), showed desired reactivity. The antigen specificity was determined in a complete inhibition test, immunofluorescence, and in immunoblots.9

Tissue Selection and Cell Preparations For staining, 4-,um cryostat sections were used. A small number of paraffin sections were stained after deparaffinization in xylol and alcohol. Sixteen normal human tissues and 42 sections of various disease states (inflammatory, autoimmune, malignant) were stained for reactivity with WGM2. The mononuclear cells (MNC) and monocyte preparations were obtained from healthy donors by density centrifugation, as described elsewhere.15 After washing the MNC in PBS pH 7.4, cytocentrifuge slides were obtained by centrifugation of 100 RI of the cell suspension (1 x 106 cells/ml) at 400x g for 1 minute in a cytocentrifuge. The monocyte preparation was obtained by resuspension of MNC in RPMI 1640 medium (Seromed, Munich, FRG), supplemented with 2 mM glutamine, 10% heat-inactivated fetal calf serum, and 1% penicillin/streptomycin (all reagents GIBCO, Paisley, Scotland). The MNC suspension was cultured in cellculture flasks (A/S Nunc, Roskilde, Denmark) at a concentration of 106 cells/ml for 7 days at 37°C in an 8% C02 atmosphere. This culture period ensured that no granulocytes contaminated the cell culture. The monocytes were harvested with a rubber policeman from the bottom of the culture flask and used for preparation of cytocentrifuge slides. Granulocytes were obtained from patients with an elevated sedimentation rate due to other than haematologic causes. After 20 minutes of sedimentation, the erythrocytefree fraction was washed in PBS pH 7.4 and cytocentrifuge slides of the granulocytes and human cell lines HL 60,11 U937,12 THP-1 13, and HUT 7814 were prepared.

was supplemented with 0.1% H202 for 20 minutes. The stainings were performed with the murine MAb WGM2 at a dilution of 1:200. The slides were incubated with 100 ,ul of the MAb for 30 minutes followed by incubation with 100 ,ul of the second antibody, a peroxidase-conjugated rabbit-anti-mouse antiserum (Dakopatts, Hamburg, FRG) for 30 minutes. Finally, incubation with the third antibody, a

peroxidase-conjugated goat-anti-rabbit antiserum (Medac, Hamburg, FRG) followed for 30 minutes. Thorough rinsing of the slides in PBS pH 7.4 followed each incubation step. The immunocomplexes were visualized by incubation with 0.06% diaminobenzidine (DAB) and 0.01% H202 in PBS pH 7.4 for 10 minutes, and the slides were counterstained with an alum-hematoxylin and mounted with glycerin gelatin. Controls omitting the first antibody were performed for every tissue and cell preparation. For additional control purposes, the monocytes that were cultured for 7 days were stained with an irrelevant mouse mab of IgGl-class to exclude unspecific Fc receptor binding of WGM2. For confirmation of the results obtained by the immunoperoxidase stains, some sections were stained by the indirect alkaline phosphatase method. The incubation with sodium azide was omitted, and 100 p.l of an alkaline phosphatase-conjugated rabbit-anti-mouse antiserum (Dakopatts, Hamburg, FRG) as second antibody and 100 [lI of an alkaline phosphatase-conjugated goat-antirabbit antiserum (Sigma, St. Lois, MO) as third antibody followed the WGM2 incubation. Again, thorough rinsing in PBS pH 7.4 followed each incubation step. NaphtolAS-BI-phosphat (Sigma, Munich, FRG) was used as substrate, Fast-Red TR salt (Sigma, Munich, FRG) as coupler for visualization of the immunocomplexes. Substrate and coupler were disolved in N, N-Dimethylformamide (DMF). 1 mM levamisol (Sigma, Munich, FRG) was added to the coupler and the substrate solution to block endogenous phosphatase activity. These sections were also counterstained with an alum-hematoxylin.

Doublestaining by Indirect Immunofluorescence

Immunohistologic Stains: Indirect Immunoperoxidase/Alkaline Phosphatase

Staining of two different antigens on one cell or tissue section was achieved by a two-step technique as described elsewhere.20 At first, the slides were incubated

Indirect immunoperoxidase stains of the frozen and deparaffinized tissues and cytocentrifuge slides of the cell preparations were performed as described in detail elsewhere.16 Briefly, the slides were fixed in ethanol at 40C for 10 minutes, followed by airdrying. To block endogenous peroxidase activity, the slides were incubated in a 0.1% M sodium azide solution (Merck, Darmstadt, FRG) that

with 25% human AB serum for 30 minutes and then the mab WGM2 diluted 1:200 was applied for 30 minutes, followed by incubation with a fluorescein-isothiocyanateconjugated (FITC) rabbit-anti-mouse antiserum (Dako, Hamburg, FRG), diluted 1:40 for 30 min. The slides were incubated for 60 minutes with unconjugated rabbit-antimouse antiserum (Dako, Hamburg, FRG) diluted 1:5 to block free binding sites of the first antibody. Thereafter,

Immunolocalization of Proteinase 3

833


Results

ber of monocytes was evaluated with the MAb KiMl. To exclude misinterpretations with granulocytes, MNC cultures from two of the blood donors were cultivated for 7 days, and similar results were obtained. To determine the phenotype of the WGM2-positive monocytes, doublestainings were performed with MNC of two donors. The doublestaining experiments showed that the WGM2-positive subset of peripheral blood monocytes were all KiMl, KiM6, KiM8, and HLA DR positive. However, only between 3% and 6% of the monocytes expressing these markers were WGM2 positive. We found no certain evidence that antigen-presenting cells contained proteinase 3. In the lymphoid organs (lymph node, Peyer's patches, spleen, tonsil) no WGM2-positive cells were seen in the germinal centre, the location of the antigen presenting dendritic reticulum cell. Furthermore, Langerhans' cells of the skin were also negative. We saw WGM2-positive macrophages in the interfollicular zone (Figure 3). It cannot be excluded that some of these proteinase-3-containing cells were antigen-presenting interdigitating-reticulum cells. Using morphologic criteria, we also found WGM2-positive macrophages in WG granulomas. Although difficult to assess, it seems unlikely that their number was increased in these granulomas.

Immunofluorescence and Blocking Tests with WGM2

Staining Results on Normal Tissue

Table 1. Second Monoclonal Antibodies Used for Double Stainings Monoclonal antibody Specificity KiMl (CD1 1 c) Macrophages, IDC, Langerhans' cells, monocytes (17) KiM6 (CD68) Macrophages, monocytes (18) KiM8 Macrophages, monocytes, IDC (19) HLA DR MHC class 11

the incubation with the antibodies listed in Table 1 followed for 30 minutes. Finally, these antibodies were visualized by a 30-minute incubation with tetramethylrhodamin isocyanate-conjugated (TRITC) rabbit-antimouse antiserum (Dako, Copenhagen, Denmark), diluted 1:20. Each incubation step was followed by extensive rinsing in PBS pH 7.4. For confirmation of the results, the MAb WGM2 and the antibodies listed in Table 1 were exchanged for the staining procedure. The controls were performed by omitting the first, the second, or both MAbs.

On normal granulocytes the fluorescence pattern obtained with WGM2 was identical to that of an ACPApositive serum (Figure 1 A). A fine granular fluorescence of the cytoplasm, sometimes accentuated around the nucleus, could be seen. For further investigations of WGM2, blocking experiments were performed. Granulocytes were incubated with WGM2 diluted 1:200, followed by incubation of an ACPA-positive serum (ACPA titer 1:64), diluted 1:32. The fluorescence intensity induced by the serum was not decreased by the MAb. Similar results were obtained using sera from four different ACPApositive WG patients. In contrast, preincubation with an ACPA-positive serum followed by incubation with WGM2 decreased the MAb-induced fluorescence by approximately 50% (Figure 1 B).

With the indirect immunoperoxidase staining method, the fine granular pattern seen with immunofluorescence could not be observed. The cytoplasm of granulocytes was more homogeneously positive, and the perinuclear accentuation was not as evident. In all normal adult human tissues, granulocytes were positive. Depending on the tissue investigated, either solitary granulocytes (e.g., stomach) or clusters (e.g., lung, spleen, tonsil, nasal mucosa) were seen. In the bone marrow, promyelocytes and myelocytes contained proteinase 3. Erythropoetic cells and megakaryocytes showed no reactivity with WGM2. In epithelial tissues (skin, colon), mast cells reacted positive; their cytoplasmic granules showed a strong reaction with MAb WGM2. No other cell type or tissue components reacted positive. In conclusion, all granulocytes and mast cells as identified by morphologic criteria were positive for proteinase 3.

Monocyte/Macrophage Staining of the MAb WGM2. Staining Results on Pathologic Tissue In normal and pathologic tissue, a subset of cells identified by morphologic criteria to be macrophages or monocytes were positive for the MAb WGM2. In mononuclear cell fractions isolated from 10 healthy blood doners only 2 to 10% of the monocytes as judged by morphologic appearance were WGM2 positive (Figure 2). The total num-

After establishing that proteinase 3 was only localized in neutrophil granulocytes, mast cells and a subset of monocytes on a cross section of normal tissue, we extended our investigation of the reactivity of WGM2 to disease states. Our aim was to determine whether the dis-

834

Braun et al


*.f

bt

.......

ol' ....... .......

..........

00,

..... ...... ..

.i:

...

Figure 1. A: Positive staining of neutrophil granulocytes with MAh WGM 2. B: Preincubation with ACPA positive serum diminishes reactivity of MAb WGM 2. C: Control: Incubation with buffer and FITC-conjugated anti-human-Ig. Figure 2. Peripheral blood monocyte stained with WGM2, alkaline phosphatase, x400. Figure 3. Spleen containing proteinase-3 positive cells in the interfollicular zone, immunoperoxidase, x400.


835


tribution of proteinase 3 changed or whether crossreactivity with other tissue structures existed in 1) inflammation due to infective agents or autoimmunity, 2) malignancies, or 3) diseases of leukocytes. Therefore, our selection of pathologic tissues included a wide range of acute and chronic inflammatory diseases involving different inflammatory cells. Furthermore, malignancies of different origin and differentiation were stained. In all of these tissue sections, the specific reactivity of WGM 2 was limited to granulocytes and a subset of monocytes. On some slides a discrete unspecific background reactivity was visible (skin, tonsil, purulent peritonitis). This was never regarded as a relevant impairment in evaluating the section.

Staining Results on Cell Lines Which cell lines of the granulocyte/monocyte or lymphocyte lineage reacted with WGM2 was of interest. No peripheral blood lymphocyte reacted positively. No reactivity with WGM2 was found on cytocentrifuge preparations of the T-cell lymphoma line HUT 78 and monoblastic U937 cells. The majority of the promyelocytic HL 60 cells showed a faint cytoplasmic reactivity without perinuclear accentuation. The reactivity was markedly less intense than on granulocytes. Approximately 50% of the monocytic THP-1 cells reacted positively with WGM2. The staining pattern differed from the pattern observed on granulocytes and HL 60 cells: THP-1 cells contained up to five granulalike patches that reacted positively for WGM2. Finally, special attention was given to WG and other granulomatous and vasculitic diseases. Sections of liver, lung, and kidney from three different WG patients were investigated. The granulomas contained high numbers of WGM2-positive cells (Figure 4). The vast majority of these cells were granulocytes. Again, no crossreactivity with other tissue structures or with other cells was observed. Figure 5 shows a blood vessel in a lung section of a WG patient in whom no positive reaction could be seen except in the surrounding granulocytes and macrophages; the endothelium of blood vessels always proved negative in WG tissue. Due to the high number of granulocytes in granulomas of WG patients, the staining appearance was characteristic. Around the WGM2-positive granulomas a patchy positive staining was observed. This staining pattern was irregular and its appearance differed from the faint crossreactivity with fibrous tissue. This was probably due to spillage of proteinase 3 from disintegrating and necrotic granulocytes. This appearance was only seen in granulomas of WG patients. In contrast to these findings, the granulomas of other diseases, such as tuberculosis, sarcoidosis, panarteritis

nodosa, M. Crohn or Hodgkin's disease, contained far fewer WGM2-positive granulocytes and monocytes. Also, the other vasculitic diseases investigated, i.e., Purpura Henoch-Sch6nlein and giant-cell arteritis, contained fewer granulocytes. In summary, proteinase 3 is only localized in granulocytes, mast cells, and a subset of monocytes; it cannot be found in cells of other origin or differentiation, and no crossreactivity to tissue structures was observed.

Discussion We used the MAb WGM2 directed against proteinase 3 for immunohistologic localization of this enzyme in various human tissues and cell lines. It was shown that this antigen is not only localized in the azurophil granules of normal granulocytes but also in a subset of monocytes, as demonstrated by electron microscopic analysis.9 The reactivity of this MAb is specific, no crossreactivity of the antibody with other cells or tissue structures could be observed apart from a faint reaction with keratin and connective tissue, which is also seen occasionally with other MAb. WGM2 is unable to block binding of ACPA-positive sera to proteinase 3. This can be explained by the monovalence of the MAb and the polyvalence of sera: the MAb can only bind to one epitope on the proteinase 3 molecule, the remaining epitopes are still accessible for the proteinase 3 antibodies in ACPA-positive sera. This result was confirmed by ELISA using purified proteinase 3 as antigen (unpublished data). However, if ACPA sera bind to granulocytes before incubation with WGM2, the WGM2 specific epitope is no longer freely accessible. This may be caused by conformational alterations on the proteinase 3 molecule after binding of ACPA or overlapping epitopes of ACPA and WGM2. Attempts were made to characterize the subset of WGM2 positive monocytes. In normal humans, we found the number of blood monocytes that contained proteinase 3 to range between 2% and 10%. These monocytes coexpressed the antigens found on the majority of mature blood monocytes and tissue macrophages, i.e., KiMl, KiM6, KiM8, and HLA DR. We found no evidence that antigen-presenting cells contained proteinase 3, although this cannot be ruled out. At present, it is unclear whether the WGM2-positive cells from the monocyte/ macrophage lineage form a homogeneous subset or have functionally distinct features. Of the cell lines investigated, only THP-1 and HL 60 reacted with WGM2. The proteinase 3 content of the HL 60 cell was much lower than that of neutrophil granulocytes and the cytoplasmic distribution pattern was more

Braun et al

836


*44P

4.,4.4 44

44'4. .44.4

.4

. 7,,4, .

I

4

.4

4

.44

,

'.4'

I

!. ..h4 4.

. .. W.,W.

A~~~~N

t:*

4;

a_

lt

Jl. :..

4j. s.

t:.

., :I E

:A';;

.

.4

Figure 4. Wegener's granulomatosis, lung. High number of MAb WGM-2 positive granulocytes in granulomas are shoun, alkaline phosphatase, x 160. Figure 5. Wegener'sgranulomatosis, lung. Proteinase3 is notfoundon blood vessel constituents but only in macrophages and granulocytes, alkaline phosphatase, x250.


837


homogeneous. Charles et al. reported that HL 60 cells are equally suitable for testing ACPA positive sera and that discrimination between perinuclear (p-ANCA) and cytoplasmic reactivity (c-ANCA) is also possible.21 Unfortunately, this report does not state whether the titer levels obtained from ACPA positive sera were identical in HL 60 cells and granulocytes. Because titer levels in WG correlate with disease activity3 7 this is an important criterion. Our results suggest that due to the lower content of proteinase 3 in HL 60 cells, neutrophil granulocytes are a more suitable substrate for screening ACPA. Special interest was focused on tissue of WG patients. Histologically, these granulomas contained large numbers of granulocytes. The MAb WGM2 reacted with the granulocytes and a subset of monocytes in these granulomas, whereas no reactivity with endothelium or other structures of the blood vessel wall was observed. This contradicts the report by Abbott et al., who concluded that ACPA are directed against glomerular cells.22 The reactivity with ACPA was shown with cultured renal cell endothelium and epithelium cells. Whether the antigenic determinants on these cells cultured in vitro with various supplements are identical to those in vivo remains open. Furthermore, if in WG the endothelium has antigenic determinants recognized by circulating ACPA, the question arises why the pathologic lesions are limited to certain organs. We found no evidence that renal cells contained proteinase 3. We cannot exclude that an ACPA-positive serum also contained autoantibodies recognizing epitopes other than proteinase 3. Furthermore, we cannot exclude that epitope homologies exist between tissue structures and proteinase 3 epitopes not recognized by

WGM2. The target antigen of ACPA, proteinase 3, is localized in cytoplasmic granules. In a viable cell these granules are not accessible for the autoantibodies. It could be demonstrated that proteinase 3 is also expressed on the cell membrane of granulocytes and a subset of monocytes where the epitope is accessible for circulating autoantibodies.9 Recent reports provide evidence for various pathologic aspects of the neutrophil granulocyte in WG. In this disease, an impaired granulocyte chemot-

protein which produced the same immunofluorescence staining pattern as ACPA. This 29-kD protein proved to be proteinase 3.31 Recently a Cathepsin G-like proteinase, 30 kD, was described in human mast cells,32 which may be identical to proteinase 3 observed in mast cells in our study. Jenette and Falk report that c-ANCA and p-ANCA are able to activate neutrophils in vitro, as evidenced by induction of a respiratory burst and degranulation.33 It seems feasible that in vivo the autoantibody binding to the granulocytes may lead to the degradation and necrosis of these cells, resulting in a higher turn-over. Degranulation of granulocytes in blood vessels may induce the vasculitic alterations seen in WG. Proteinase 3 has elastinolytic properties and causes emphysema in vitro.34 The relation of the clinical symptoms of WG to the putative pathogenesis is of interest. The organs affected by the granulomatous and vasculitic alterations are the upper respiratory tract, the lung, and the kidney. In healthy humans, the lung contains about 40% of the entire granulocyte population, and the granulocyte content of the upper respiratory tract mucosa is also high. Intravascular granulocyte lysis may initiate the typical vascular lesions in these organs, either directly by elastinolytic enzymes or via complement activation by immunocomplexes. These immunocomplexes consisting of proteinase 3 and autoantibodies may be deposited in the kidney, causing glomerulonephritis. This hypothesis is consistent with the observation that usually the pathologic lesions are primarily limited to the upper respiratory tract before the active generalized phase.13 Important questions arise regarding which immunomodulation initiates the autoantibody formation in WG and what are the pathologic mechanisms that cause the chronicity of this disease. Nevertheless, increasing evidence exists that ACPA are not just an epiphenomenon but are responsible for the course of the disease and the pathologic alterations. Further investigations are required to determine the physiologic role of proteinase 3 and its functional role in WG resulting from interaction with ACPA.

axis23 and a higher granulocyte turnover24 have been described. Electron microscopically, signs of granulocyte degradation 25,26 and an Ig-coating-like rim around cell organelles were shown.23 Furthermore, ACPA stimulate neutrophil activation in vitro27,28 and intravascular granulocyte lysis has been described.-9 These findings are compatible with the present data. The substrate for the described Ig-coating of cell organelles in the cytoplasm of granulocytes proved to be the anti-proteinase 3 antibodies, as shown with the MAbs WGM1 and WGM2.9 Niles et al.30 generated a MAb directed against a 29-kD

Acknowledgments The authors thank Christa Knaus and Christl Kohaut for their excellent technical assistance, and Jens Ludemann and Martin Pfaff for the critical reading of the manuscript.

References 1. Parlevliet KJ, Henzen-Logmanns SC, Oe PL, Bronsveld W, Balm AJ, Donker AJ: Antibodies to components of neutrophil cytoplasm; a new diagnostic tool in patients with We-

838

Braun et al


gener's granulomatosis and systemic vasculitis. Q J Med 1988, 249:55-63 2. Venning MC, Arfeen S, Bird AG: Antibodies to neutral cytoplasmic antigen in systemic vasculitis. (Letter to the editor) Lancet 1987, 2:850 3. Gross WL: Wegener's granulomatosis: new aspects of the disease course, immunodiagnostic procedures and stageadapted treatment. Sarcoidosis 1989, 6:15-29 4. Van der Woude FJ, Rasmussen N, Lobatto S, WilkA, Permin H, van Es LA, van der Giessen M, van der Heen GK, The TH: Autoantibodies against neutrophils and monocytes: tool for diagnosis and marker for disease activity in Wegener's granulomatosis. Lancet 1985, 1:425-429 5. Ludemann G, Gross WL: Autoantibodies against cytoplasmic structures of neutrophil granulocytes in Wegener's granulomatosis. Clin Exp Immunol 1987, 69:350-357 6. Savage CO, Winearls CG, Jones S, Marshall PD, Lockwood CM: Prospective study of radioimmunoassay for antibodies against neutrophil cytoplasm in diagnosis of systemic vasculitis. Lancet 1987,1:1389-1393 7. Nolle B, Specks U, Ludemann J, Rohrbach MS, DeRemee RA, Gross WL: Anticytoplasmic autoantibodies: their immunodiagnostic value in Wegener's granulomatosis. Ann Intern Med 1989,111:28-40 8. LOdemann J, Utecht B, Gross WL: Anti-neutrophil cytoplasm antibodies in Wegener's granulomatosis recognize an elastionolytic enzyme. J Exp Med 1990,171:357-362 9. Csemok E, Ludemann J, Gross WL, Bainton DF: Ultrastructural localisation of proteinase 3, the target antigen of anticytoplasmic antibodies circulating in Wegener's granulomatosis. Am J Pathol 1990,137:1113-1119 10. LOdemann J, Utecht B, Gross WL: Detection and quantification of anti-neutrophil cytoplasm antibodies in Wegener's granulomatosis by ELISA using affinity-purified antigen. J Immunol Method 1988,114:167-174 11. Gallagher R, Collins S, Trujillo J: Characterization of the continous, differentiating myeloid cell line (HL 60) from a patient with acute promyelocytic leukaemia. Blood 1979, 54:713733 12. Sundstrom C, Nilsson K: Establishment and characterization of a human histiocytic lymphoma cell line (U937). Intl J Cancer 1976,17:565-577 13. Tsuchiya S, Yamabe M, Yamaguchi Y, Kobayashi Y, Konno T, Tada K: Establishment and characterization of a human acute monocytic leukaemia cell line (THP-1). Intl J Cancer 1980, 26:171-176 14. Gazdar AF, Camey DN, Bunn PA, Russell EK, Jeffe ES, Schechter GP, Guccion JG: Mitogen requirements for the in vitro propagation of cutanous T-cell lymphomas. Blood 1980, 55:409-417 15. Boyum A: Isolation of mononuclear cells and granulocytes from human blood. J Clin Lab Invest 1968, 21:77-89 16. Stein H, Bonk A, Tolksdorf G, Lennert K, Rodt H, Gerdes J: immunohistologic analysis of the organization of the normal lymphoid tissue and non-Hodgkin's lymphoma. J Histochem Cytochem 1980, 28:746-752 17. Radzun HJ, Pawaresch MR, Feller AC, Hansmann ML: Monocyte/macrophage-specific monoclonal antibody Ki

Ml recognizes interdigitating reticulum cells. Am J Pathol 1984,117:441-450 18. Pawaresch MR, Radzun HJ, Kreipe H, Hansmann ML, Barth J: Monocyte/macrophage reactive monoclonal antibody Ki M6 recognizes an intracytoplasmatic antigen. Am J Pathol 1986,124:141-150 19. Radzun HJ, Kreipe H, B6dewadt S, Hansmann ML, Barth J, Pawaresch MR: Ki M8 monoclonal antibody reactive with an intracytoplasmatic antigen of monocyte/macrophage lineage. Blood 1986, 69:1320-1327 20. Peters U, Papadopoulos T, Muller-Hermelink HK: MHC class 11 antigens on lung epithelia of human fetuses and neonates. Ontogeny and expression in lungs with histologic evidence of infection. Lab Invest 1990 in press 21. Charles LA, Falk RJ, Jennette JC: Reactivity of antineutrophil cytoplasmic antibodies with HL 60 cells. Clin Immunol Immunopathol 1989, 53:243-253 22. Abbott F, Jones S, Lockwood CM, Rees AJ: Autoantibodies to glomerula antigens in patients with Wegener's granulomatosis. Nephrol Dial Transplant 1989, 4:1-8 23. Niinaka T, Okochi T, Watanabe Y, Yamamura T, Yamamura Y: Chemotactic defect in Wegener's granulomatosis. J Med 1977, 8:161-169 24. Dale DC, Fauci AS, Wolff SM: The effect of cyclophosphamide on leucocyte kinetics and susceptibility to infection in patients with Wegener's granulomatosis. Arthrit Rheum 1973, 16:657-663 25. Gross WL: Wegener'sche Granulomatose: immunologische Aspekte zur Diagnostik, Genese und Therapie. Immun Infekt 1987, 15:15-21 26. Schubert C, Gross WL, Schroder J: Functional and ultrastuctural studies on PWN in Wegener's granulomatosis. J Invest Dermatol 1990, in press 27. Jennette JC, Falk RJ: Anti-neutrophil cytoplasmic antibodies. N Engl J Med 1988, 319:1417 28. Falk RJ, Terrell RJ, Honeycutt-Calder L, Jennette JC: Antineutrophil cytoplasmic autoantibody (ANCA) stimulate neutrophil activation in vitro. Kidney Intl 1989, 35:347-353 29. Donald KJ, Edwards RC, McEvoy JDS: An ultrastructural study of the pathogenesis of tissue injury in limited Wegener's granulomatosis. Pathology 1976, 8:161:172 30. Niles JL, McCluskey RT, Ahmad MF, Amaout MA: Wegener's granulomatosis autoantigen is a novel neutrophil serine proteinase. Blood 1989, 74:1888-1893 31. Jenne DE, Tschopp J, Ludemann J, Utecht B, Gross WL: Wegener's antigen decoded. Nature 1990,146:320 32. Schechter NM, Irani AA, Sprows JL, Abernethy J, Wintroub B, Schwartz LB: Identification of a Cathepsin G-like Proteinase in the MC type of human mast cell. J Immunol 1990, 145:2652-2661 33. Jennette JC, Falk RJ: Anti-neutrophil cytoplasmic autoantibodies: new insight into crescentic glomerulonephritis, pulmonary-renal syndrome and systemic vasculitis. AKF Nephr Letter 1989, 6:11 34. Kao RC, Wehner NG, Skubitz KM, Gray BH, Hoidal JR: Proteinase 3: a distinct human polymorphonuclear leucocyte proteinase that produces emphysema in hamsters. J Clin Invest 1988, 82:1963-1973