Sialylation levels of antiproteinase 3 antibodies are associated with the activity of granulomatosis with polyangiitis Wegener's.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 63, No. 7, July 2011, pp 2105–2115 DOI 10.1002/art.30362 © 2011, American College of Rheumatology Sialylation Levels of Anti–Proteinase 3 Antibodies Are Associated With the Activity of Granulomatosis With Polyangiitis (Wegener’s) Cécile Espy,1 Willy Morelle,2 Niloufar Kavian,1 Philippe Grange,1 Claire Goulvestre,1 Vivian Viallon,1 Christiane Chéreau,1 Christian Pagnoux,1 Jean-Claude Michalski,2 Loı̈c Guillevin,1 Bernard Weill,1 Frédéric Batteux,1 and Philippe Guilpain1 Birmingham Vasculitis Activity Score (BVAS) (P < 0.0001). Similar results were obtained using the BVAS/ GPA. The area under the receiver operating characteristic curve for the sialylation ratio of anti-PR3 antibodies, as a test to determine the activity of GPA, was 0.82 (P ⴝ 0.0006). The characterization of N-glycans showed a decrease in 2,6-linked sialylated N-glycans and an increase in dHex1Hex3HexNAc4 (mass/charge 1,836) agalactosylated structures in purified IgG from patients with active disease compared with controls. The antiPR3 antibody–induced oxidative burst of neutrophils was inversely correlated with the sialylation levels of anti-PR3 IgG. Conclusion. The sialylation level of anti-PR3 antibodies contributes to the clinical activity of GPA, by modulating the oxidative burst of neutrophils induced by these autoantibodies. Objective. To investigate whether the glycosylation and sialylation levels of anti–proteinase 3 (antiPR3) antibodies could affect their pathogenicity, and whether these levels could be correlated with the activity of granulomatosis with polyangiitis (Wegener’s) (GPA). Methods. Forty-two serum samples positive for anti-PR3 antibodies from 42 patients with active or weakly active/inactive GPA were included. Anti-PR3 antibodies were assayed by enzyme-linked immunosorbent assay, and their levels of glycosylation and sialylation were assessed by enzyme-linked lectin assay. The glycosylation and sialylation levels of IgG purified from the serum of healthy donors and patients with active, remitted, or weakly active disease were assessed by permethylation and mass spectrometry analysis of glycans, following neuraminidase digestion. The neutrophil oxidative burst induced by purified IgG was assayed by spectrofluorimetry. Results. The mean sialylation ratio of anti-PR3 antibodies was significantly lower in patients with active disease than in patients with weakly active or inactive disease, and this was inversely correlated with the Granulomatosis with polyangiitis (Wegener’s) (GPA) is a small-sized vessel necrotizing vasculitis associated with giant cell granulomas and necrosis. GPA usually affects the upper respiratory tracts, lungs, and kidneys, but any organ may be involved, including the skin, peripheral nerve, gastrointestinal tract, heart, and central nervous system. Antineutrophil cytoplasmic autoantibodies (ANCAs) (1–3) exhibiting a classic cytoplasmic (cANCA) pattern of fluorescence and specific for proteinase 3 (PR3) are detected in 85–90% of patients with GPA. Anti-PR3 antibodies are considered to play a pathogenic role through the induction of the oxidative burst of neutrophils (4), but the measurement of anti-PR3 antibody levels as a tool to monitor disease activity remains controversial. Their levels are usually not correlated with disease activity and clinical outcome (5–7), although some authors found a clear association 1 Cécile Espy, PharmD, Niloufar Kavian, PharmD, PhD, Philippe Grange, PhD, Claire Goulvestre, PharmD, PhD, Vivian Viallon, PhD, Christiane Chéreau, PharmD, Christian Pagnoux, MD, PhD, Loı̈c Guillevin, MD, Bernard Weill, MD, PhD, Frédéric Batteux, MD, PhD, PharmD, Philippe Guilpain, MD, PhD: Université Paris Descartes and Hôpital Cochin, Assistance Publique Hôpitaux de Paris, Paris, France; 2Willy Morelle, PhD, Jean-Claude Michalski, PhD: UMR 8576, Université des Sciences et Technologies de Lille, Villeneuve D’Ascq, France. Drs. Batteux and Guilpain contributed equally to this work. Address correspondence to Frédéric Batteux, MD, PhD, PharmD, Laboratoire d’Immunologie, UPRES EA 1833, 24 Rue du Faubourg St. Jacques, 75679 Paris Cedex 14, France. E-mail: firstname.lastname@example.org. Submitted for publication November 30, 2009; accepted in revised form March 15, 2011. 2105 2106 ESPY ET AL between increased anti-PR3 antibody levels and ensuing disease activity (8). Moreover, anti-PR3 antibodies do not constantly disappear during clinical remission, suggesting that persistent anti-PR3 antibodies are devoid of pathogenic effects. Recent advances have been made in the understanding of the mechanisms that determine the biologic activity of IgG antibodies. The interaction between Fc␥ fragments and their receptors, the clearance of IgG molecules, and the deposition of Ig antibodies in tissue (9–11) are dependent on the type and level of glycosylation of the Fc␥ fragments. Two examples of this process can be presented. First, the level of sialylation of Fc␥ fragments was found to modulate the pathogenic effect of autoantibodies in a murine model of immune thrombocytopenia (12). Second, the level of sialylation of the IgG antibodies in intravenous immunoglobulin preparations can modulate their antiinflammatory effects. We therefore hypothesized that the sialylation of anti-PR3 antibodies could contribute to their pathogenic role. We designed an experimental model to determine whether variations in the sialylation of anti-PR3 antibodies are associated with anti-PR3 antibody–mediated neutrophil activation, disease activity, and clinical features of GPA. PATIENTS AND METHODS Patients and serum sampling. Forty-two patients with GPA (42 serum samples from 22 men and 20 women, mean ⫾ SD age 44.7 ⫾ 17.4 years, range 16–73 years) were included in the study. All patients were positive for anti-PR3 antibodies and met the Chapel Hill Consensus Conference Nomenclature/Criteria for Vasculitis (13) and/or the American College of Rheumatology criteria for GPA (14). The patients’ clinical manifestations at the diagnosis of GPA were as follows: ear/nose/throat (ENT) involvement (37 patients), general symptoms (32 patients), lung involvement (32 patients, including 16 with alveolar hemorrhage), arthromyalgias (25 patients), renal involvement (24 patients), ophthalmologic involvement (16 patients), skin lesions (11 patients), peripheral neuropathy (9 patients), and gastrointestinal involvement (5 patients). Histologic confirmation of vasculitis was obtained in 39 of the 42 patients. Vasculitis was confirmed by biopsy of the kidneys in 20 patients, while ENT involvement was confirmed in 10 patients, skin involvement in 6 patients, lung involvement in 4 patients, orbital tumor in 2 patients, and muscle involvement in 2 patients. Twenty-six serum samples were obtained at the time of active GPA and 16 were obtained during the remission stage or low-activity state of GPA, when the disease was weakly active or remittent. The 42 serum samples were studied to compare patients with active disease to those with weakly active/inactive disease. An additional series of 12 serum samples from patients with remitted GPA (Birmingham Vasculitis Activity Score [BVAS] of 0) were studied to compare, in a crosssectional analysis, these sera at the time of remission with the same patients’ sera at the time of flare (mean ⫾ SD BVAS score 19.6 ⫾ 12.9). Ten serum samples from 10 healthy donors served as controls. All patients and controls gave their written informed consent. Disease activity was assessed by the BVAS index (5,15,16). A BVAS score of 0 corresponded to remitted disease. A BVAS score of 3 or higher corresponded to active disease, while a BVAS score lower than 3 corresponded to weakly active disease. The BVAS/GPA, a specific activity index for GPA, was also calculated (17). When disease activity was assessed in the 42 patients at the time of sampling using the BVAS/GPA, 17 patients were determined to have severe flare, while 2 had limited flare, 13 had persistent disease, and 10 were in remission. At the time of sampling, patients with active GPA had received low doses (⬍10 mg/day) of prednisone (12 patients), intermediate doses (10–30 mg/day) of prednisone (2 patients), azathioprine (4 patients), methotrexate (6 patients), and/or cyclophosphamide (1 patient) or had received no treatment (12 patients). At the time of sampling, patients with weakly active/inactive GPA had received low doses (⬍10 mg/day) of prednisone (14 patients), azathioprine (6 patients), and/or methotrexate (5 patients) or had received no treatment (2 patients). At the time of sampling of the 12 additional sera from patients with remitted disease, obtained for the longitudinal study, this group of patients had received low doses (⬍10 mg/day) of prednisone (10 patients), azathioprine (4 patients), and/or methotrexate (2 patients) or had received no treatment (2 patients). IgG purification from the serum of patients with GPA and control subjects. IgG antibodies from the serum of 4 patients with active systemic GPA, 4 patients with localized GPA, 4 patients with anti-SSA–positive Sjögren’s syndrome, 4 patients with anti–Jo-1–positive myositis, and 7 healthy subjects were purified using a Melon Gel IgG purification kit (Pierce). The purity of the resulting IgG preparation was assessed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis. The purified IgG antibodies were then tested for their ability to induce H2O2 production by polymorphonuclear neutrophils (PMNs), and qualitative and quantitative analyses of their glycosylation were performed. Release and clean-up of the N-linked glycans of immunoglobulins. The IgG antibodies purified from GPA serum and control serum were lyophilized and then resuspended in 200 l of 50 mM ammonium hydrogen carbonate. After the addition of 20 l of 200 mM dithiothreitol, the sample was flushed with argon and incubated at 60°C for 1 hour. After the addition of 80 l of 200 mM iodoacetamide, the sample was again flushed with argon and incubated at room temperature for 1 hour in the dark. Twenty microliters of 200 mM dithiothreitol was then added to react with the iodoacetamide in excess, at room temperature for 1 hour in the dark. The sample was diluted with 600 l of 50 mM ammonium hydrogen carbonate, and the reduced carboxyamidomethylated immunoglobulins were digested with TPCK bovine pancreas trypsin (EC 188.8.131.52; Sigma), with an enzyme-to-substrate ratio of 1:25 (by mass), for 24 hours at 37°C. The reaction was terminated by boiling for 5 minutes before lyophilization. The products were then purified on a C18 Sep-Pak SIALYLATION OF ANTI-PR3 ANTIBODIES IN GPA (Waters Ltd.). After conditioning the C18 Sep-Pak by sequential washing with methanol (5 ml) and 5% acetic acid (10 ml), the sample was loaded onto the Sep-Pak and washed with 15 ml of 5% acetic acid. Peptides/glycopeptides were eluted with 3 ml of acetonitrile/water (80:20 volume/volume) containing 5% (v/v) acetic acid. Acetonitrile was evaporated under a stream of nitrogen and the sample was freeze-dried. PNGase F digestion was carried out in 50 mM ammonium hydrogen carbonate for 16 hours at 37°C. After PNGase F digestion, the reaction was terminated by lyophilization and the products were purified on a C18 Sep-Pak, in the same manner as described above, to isolate the released N-glycans, which were then freeze-dried. Permethylation and mass spectrometry analysis of glycans. Permethylation of the freeze-dried N-glycans was performed according to the procedure developed by Ciucanu and Kerek (18). The reaction was terminated by adding 1 ml of cold 10% (v/v) acetic acid, followed by 3 extractions with 500 l of chloroform. The pooled chloroform phases (1.5 ml) were then washed 8 times with water. The chloroform phase containing the methylated derivatives was finally dried under a stream of nitrogen, and the extracted products were further purified on a C18 Sep-Pak (19). The C18 Sep-Pak was sequentially conditioned with methanol (5 ml) and water (10 ml). The derivatized glycans dissolved in methanol were applied on the cartridge, washed with 3 ⫻ 5 ml water and 2 ml of 10% (v/v) acetonitrile in water, and eluted with 3 ml of 80% (v/v) acetonitrile in water. Acetonitrile was evaporated under a stream of nitrogen and the samples were freeze-dried. Matrixassisted laser desorption ionization–time-of-flight (MALDITOF) mass spectrometry and linkage analyses of permethylated glycans were performed as described elsewhere (19). Neuraminidase digestion. The released N-glycans were digested at 37°C for 48 hours with 50 mU ␣-neuraminidase (from Vibrio cholerae, EC 184.108.40.206; Roche Molecular Biochemicals) in 200 l of 50 mM ammonium formate buffer, pH 5.5. The digestion was terminated by boiling for 5 minutes before lyophilization. An appropriate aliquot was obtained after the digestion, and this was permethylated for MALDI-TOF mass spectrometry analysis after purification on a C18 Sep-Pak. ANCA and anti-PR3 antibody assays. All sera were screened for ANCAs by indirect immunofluorescence using ethanol-fixed fresh normal neutrophils (20). Anti-PR3 antibodies were determined by enzyme-linked immunosorbent assay (ELISA), using a commercially available kit according to the manufacturer’s recommendations (Bio Advance), with results expressed as arbitrary units (AU)/ml. Concentrations of anti-PR3 less than 20 AU/ml were considered negative. Serum levels of sialylated anti-PR3 antibodies, as determined by enzyme-linked lectin assay (ELLA), and sialylation ratio of anti-PR3 antibodies. The levels of sialylated anti-PR3 antibodies were determined by ELLA, adapted from the method of Gornik and Lauc (21). In this assay, wells coated with PR3 (Bio Advance) were first treated with 100 l of 20 mM freshly prepared periodic acid (HIO4; SigmaAldrich) at 4°C for 30 minutes, in order to prevent nonspecific binding of lectins to carbohydrate determinants on PR3. Desialylation of PR3 does not affect the binding of anti-PR3 antibodies, since the levels of anti-PR3 antibodies determined on HIO4-treated PR3 were strongly and significantly correlated with the levels of anti-PR3 antibodies determined on 2107 native PR3. After 2 washes with 300 l of sample buffer (Bio Advance), 100 l of each human serum, diluted 1:50, was added in duplicate and incubated for 60 minutes at 37°C. The wells were then washed 3 times for 5 minutes with 300 l of TBS (50 mmoles/liter Tris HCl, 150 mmoles/liter NaCl, and 0.1 mmoles/liter CaCl2, pH 7.5). In order to detect sialic acids on anti-PR3 antibodies, 100 l of biotinylated Sambucus nigra lectin (Vector), diluted at 2 g/ml in TBS, was added to each well and incubated for 90 minutes at 37°C. The wells were then washed as described above and incubated with 100 l of 1 mg/liter alkaline phosphatase–conjugated streptavidin (Sigma) in TBS for 60 minutes at 37°C. After 3 additional washes, 100 l of 1 gm/liter p-nitrophenyl phosphate substrate (Sigma) in diethanol amine buffer (1 mole/liter diethanol amine, 0.5 mmoles/liter MgCl2, pH 9.8) was added to each well and incubated for 30 minutes at room temperature. The levels of sialylated anti-PR3 antibodies were expressed as the optical density (OD). The OD values were determined at 405 nm, with a reference wavelength of 630 nm, using a Dynatec microplate reader. Each plate comprised wells with positive sera, wells with negative sera, and wells without PR3 (blank controls). For each serum, total anti-PR3 IgG levels were measured using an alkaline phosphatase–conjugated anti-IgG antibody for detection, instead of biotinylated lectin. The sialylation ratio of anti-PR3 antibodies was determined for each sample by calculating the ratio between the levels of sialylated anti-PR3 antibodies and the levels of total anti-PR3 antibodies. Neutrophil oxidative burst induced by IgG purified from the serum of patients with GPA. PMNs from the serum of healthy blood donors were isolated by gradient centrifugation (22). Production of hydrogen peroxide (H2O2) by the neutrophils was measured as described previously (23). Cells (1 ⫻ 105 per well) were seeded in 96-well plates (Costar) and incubated at 37°C in 5% CO2 with 50 l of 200 mM dichlorofluorescein diacetate (Sigma-Aldrich) in phosphate buffered saline. After 30 minutes, 50 l of IgG antibodies purified from GPA serum or control serum were added, and H2O2 production was assayed by spectrofluorimetry after 6 hours (Fusion; PerkinElmer), with results expressed as AU. In another set of experiments, in order to block Fc receptors, PMNs were preincubated for 1 hour at 37°C with 2 g/ml of anti-human CD16 blocking antibody (clone 3G8; BioLegend) or 2 g/ml of anti-CD32 blocking antibody (clone IV.3; Stem Cell Technologies). The assay for H2O2 production was then performed in the same manner as described above. Statistical analysis. All results are expressed as the mean ⫾ SD. Statistical analysis was performed using the nonparametric Mann-Whitney U test for unpaired data, Wilcoxon’s matched pairs test, or regression analysis with the Spearman’s test. A P value less than 0.05 was considered significant. In the stratified analysis, patients with active disease (BVAS score ⱖ3) were compared to patients with weakly active disease (BVAS score ⬍3) or patients with inactive disease (BVAS score 0). A receiver operating characteristic (ROC) curve was constructed to determine whether the levels of anti-PR3 antibodies or their sialylation ratio could be a predictor of the activity of GPA. The area under the ROC curve and P value of the ROC curve were calculated. 2108 ESPY ET AL RESULTS Characterization of the N-glycans of IgG. N-glycan IgG antibodies from the serum of 7 healthy subjects, 4 patients with active systemic GPA, and 4 patients with inactive/remitted GPA (BVAS score 0) were released using PNGase F digestion, after reduction, carboxyamidomethylation, and trypsin digestion of IgG. N-glycans were permethylated, purified on a SepPak C18 cartridge, and analyzed by MALDI-TOF mass spectrometry (Figure 1). Pseudomolecular ions corresponding to the molecular weight of known N-glycan structures were observed. Two of the most intense ions of the 3 spectra were at the mass/charge (m/z) positions 1,836 and 2,040, which correspond to glycans with the compositions dHex1Hex3HexNAc4 and dHex1Hex4HexNAc4, respectively. These glycans were of the fucosylated biantennary type, and are partially galactosylated. However, the major ion in the serum of healthy subjects and patients with inactive GPA was at m/z 2,040 (Figures 1A and C), while that in the sera of patients with systemic active GPA was at m/z 1,836 (Figure 1B). In addition, ions at m/z 2,431 (NeuAc1Hex5HexNAc4), 2,605 (NeuAc1dHex1Hex5HexNAc4), and 2,792 (NeuAc2Hex5HexNAc4), which correspond to sialylated biantennary structures, were less intense in the patients with systemic active GPA (Figure 1B). These results suggest that the glycosylation pattern of IgG is modified in patients with systemic active GPA, with a decrease in sialylated N-glycans and an increase in agalactosylated structures (dHex1Hex3HexNAc4; m/z 1,836). Decreased levels of sialylated biantennary structures in serum IgG antibodies. The total sialylation level of IgG antibodies was evaluated in all patients by assessing the signals given by ions at m/z 2,431 (NeuAc 1 Hex 5 HexNAc 4 ), 2,605 (NeuAc 1 dHex 1 Hex5HexNAc4), and 2,792 (NeuAc2Hex5HexNAc4). The sialylation level was decreased in patients with active GPA compared to that in patients with inactive GPA and healthy controls (mean ⫾ SD 36.79 ⫾ 18.12% in active GPA versus 93.27 ⫾ 17.98% in inactive GPA and 64.05 ⫾ 17.51% in healthy controls; P ⬍ 0.05 for each) (Figure 2A). The sialylation of ions at m/z 2,431, 2,605, and 2,792 of IgG in the serum of patients with active or Figure 1. Matrix-assisted laser desorption ionization–time-of-flight (MALDI-TOF) mass spectrometry profiles of permethylated N-glycans released from IgG purified from the serum of 1 healthy subject (A), 1 patient with systemic active granulomatosis with polyangiitis (Wegener’s) (GPA) (B), and 1 patient with remitted GPA (C). After reduction, carboxyamidomethylation, and digestion with trypsin, N-glycans were released using PNGase F, and then permethylated and purified on a Sep-Pak C18 cartridge before MALDI-TOF mass spectrometry analysis. Data were acquired in positive ion mode, or [M⫹Na]⫹. Only the structures of the major N-glycans are given: galactose (open circles), mannose (closed circles), N-acetylglucosamine (squares), fucose (triangles), and N-acetyl-D-neuraminic acid (diamonds). m/z ⫽ mass/charge. SIALYLATION OF ANTI-PR3 ANTIBODIES IN GPA 2109 Figure 2. Levels of sialylated biantennary structures in serum IgG antibodies purified from the serum of 4 patients with active granulomatosis with polyangiitis (Wegener’s) (GPA), 4 patients with inactive GPA, and 7 healthy control subjects. The total sialylation levels (A) and sialylation levels for ions at mass/charge (m/z) 2,431 (B), m/z 2,605 (C), and m/z 2,792 (D) are shown. Bars show the mean ⫾ SD results from experiments performed in triplicate. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01. NS ⫽ not significant. inactive GPA and healthy controls are presented in Figures 2B, C, and D, respectively. The 2,6-linkage of sialic acid on N-glycans. In order to determine the linkage between the sialic acid residue and the penultimate galactose on the complex glycans, the permethylated N-glycans of the IgG antibodies from the serum of 4 healthy subjects and 3 patients with systemic active GPA were then hydrolyzed and analyzed as their partially methylated alditol acetate derivatives using gas chromatography mass spectrometry. The results (Table 1) revealed a preferential 2,6linkage, as indicated by the presence of 6-linked galactose and the absence of 3-linked galactose. In addition, after ␣-neuraminidase treatment, 6-linked galactose disappeared, indicating that sialic acid residues were attached at the 6-position of galactose prior to desialylation (Table 1). Anti-PR3 antibody levels and sialylation ratio in patients with active and those with inactive/weakly active GPA. Mean anti-PR3 antibody levels tended to be higher in patients with active disease (BVAS score ⱖ3) than in patients with weakly active (BVAS score ⬍3) or inactive (BVAS score 0) disease (mean ⫾ SD 169.3 ⫾ 160.9 UI/ml versus 76 ⫾ 32.3 UI/ml), but this difference was not significant (P ⫽ 0.08) (Figure 3A). Similarly, when the BVAS/WG was used to classify disease activity, mean anti-PR3 antibody levels tended to be higher in patients with active GPA (i.e., those with severe or limited flares and those with persistent disease) than in patients with remitted GPA (mean ⫾ SD 153.5 ⫾ 149.2 UI/ml versus 70.1 ⫾ 28.3 UI/ml), but this difference was not significant (P ⫽ 0.11) (results not shown). In the 12 patients with remitted disease whose serum was studied longitudinally, anti-PR3 antibody levels were not statistically significantly different between the samples obtained at the time of flare and those obtained during clinical remission of the disease (Figure 3B). A weak relationship was observed between anti-PR3 antibody levels and the BVAS score (r ⫽ 0.53, P ⫽ 0.003) (Figure 3C) and between anti-PR3 antibody levels and the BVAS/WG score (r ⫽ 0.54, P ⫽ 0.002) (results not shown). The mean sialylation ratio of anti-PR3 antibodies was significantly lower in patients with active disease (systemic or limited; BVAS score ⱖ3) than in patients with weakly active (BVAS score ⬍3) or inactive (BVAS score 0) disease (mean ⫾ SD 0.437 ⫾ 0.323 versus 2110 ESPY ET AL Table 1. Linkage analysis of partially methylated alditol acetate derivatives of the PNGase F–released N-glycans of IgG isolated from the sera of 4 healthy subjects and 3 patients with GPA* Relative abundance Assignment Retention time, minutes Characteristic fragment ions, m/z T1 T2 T3 T4 S1 S2 S3 Terminal Fuc Terminal Gal 2-linked Man 6-linked Gal 3,6-linked Man 3,4,6-linked Man Terminal GlcNAc 4-linked GlcNAc 4,6-linked GlcNAc 15.17 20.50† 23.57 26.49‡ 30.12 31.46 33.06 36.27 41.09 115, 118, 131, 162, 175 102, 118, 129, 145, 161, 162, 205 129, 130, 161, 190 99, 102, 118, 129, 162, 189, 233 118, 129, 189, 234 118, 333 117, 159, 203, 205 117, 159, 233 117, 159, 261 0.31 0.60 0.64 0.37 0.31 0.11 0.47 1.00 0.21 0.35 0.54 0.78 0.65 0.27 0.17 0.39 1.00 0.33 0.21 0.36 0.41 0.33 0.24 0.13 0.38 1.00 0.18 0.17 0.76 0.75 0.43 0.39 0.11 0.39 1.00 0.21 0.14 0.40 0.36 0.27 0.36 0.16 0.22 1.00 0.16 0.15 0.41 0.33 0.28 0.28 0.14 0.37 1.00 0.23 0.21 0.44 0.47 0.24 0.15 0.07 0.30 1.00 0.18 * The 80% (volume/volume) acetonitrile fractions from Sep-Pak purifications of permethylated N-glycans in the serum of 4 healthy subjects (T1–T4) and 3 patients with systemic active granulomatosis with polyangiitis (Wegener’s) (GPA) (S1–S3) were hydrolyzed, reduced, acetylated, and analyzed by matrix-assisted laser desorption ionization–time-of-flight mass spectrometry. m/z ⫽ mass/charge; Fuc ⫽ fucose; Gal ⫽ galactose; Man ⫽ mannose; GlcNAc ⫽ N-acetylglucosamine. † Signals more intense after treatment of N-glycans with ␣-neuraminidase. ‡ Signals not observed after treatment of N-glycans with ␣-neuraminidase. 1.022 ⫾ 0.6014; P ⫽ 0.0002) (Figure 3D). Similarly, using the BVAS/WG, the mean sialylation ratio of anti-PR3 antibodies was significantly lower in patients with active GPA (i.e., those with severe or limited flares and those with persistent disease) than in patients with remitted GPA (mean ⫾ SD 0.56 ⫾ 0.51 versus 0.97 ⫾ 0.46; P ⫽ 0.0098) (results not shown). In the same 12 patients with remitted disease longitudinally studied, the sialylation ratio of anti-PR3 antibodies was lower during flares than during remissions (mean ⫾ SD 0.320 ⫾ 0.230 versus 1.044 ⫾ 0.62; P ⫽ 0.0005) (Figure 3E). A significant reverse correlation was observed between the sialylation ratio of anti-PR3 antibodies and the BVAS score (r ⫽ ⫺0.68, P ⬍ 0.0001) (Figure 3F) or the Figure 3. Anti–proteinase 3 (anti-PR3) antibody (Ab) levels and the sialylation ratio of anti-PR3 antibodies in patients with active and those with inactive/weakly active granulomatosis with polyangiitis (Wegener’s) (GPA). Serum levels of anti-PR3 antibodies (A and B) and the sialylation ratio (D and E) were compared between 26 patients with active GPA and 16 patients with weakly active or inactive GPA (A and D), and between 12 patients, studied longitudinally, at the time of flare (active) and time of remission (inactive) (B and E). C and F, The relationship between the Birmingham Vasculitis Activity Score (BVAS) and the anti-PR3 antibody levels (C) or between the BVAS and the sialylation ratio (F) was determined in the 42 patients with GPA included in the study. NS ⫽ not significant; AU ⫽ arbitrary units. SIALYLATION OF ANTI-PR3 ANTIBODIES IN GPA 2111 Figure 4. Receiver operating characteristic (ROC) curve for anti–proteinase 3 (anti-PR3) antibody levels (A) and the sialylation ratio of anti-PR3 antibodies (B) in patients with granulomatosis with polyangiitis (Wegener’s). The statistical characteristics for each ROC curve are indicated below the graphs. BVAS/WG score (r ⫽ ⫺0.70, P ⬍ 0.0001) (results not shown). Notably, there was a strong and significant positive correlation between the anti-PR3 antibody levels determined on HIO4-treated PR3 (used to calculate the sialylation ratio of anti-PR3 antibodies) and the anti-PR3 antibody levels determined on native PR3 (r ⫽ 0.88, P ⬍ 0.0001). The area under the ROC curve for anti-PR3 levels, used as a test to determine GPA activity, was 0.66 (95% confidence interval [95% CI] 0.50–0.83; P ⫽ 0.07609) (Figure 4A). The area under the ROC curve for the sialylation ratio of anti-PR3 antibodies was 0.82 (95% CI 0.68–0.96; P ⫽ 0.0006) (Figure 4B). Production of hydrogen peroxide (H2O2) by neutrophils incubated with IgG antibodies purified from the serum of patients with active systemic or inactive GPA. We then investigated the functional consequences of hyposialylation in terms of its effects on the neutrophil oxidative burst. The production of hydrogen peroxide (H2O2) by neutrophils was higher when cells were incubated with IgG antibodies purified from the serum of patients with systemic active GPA (BVAS score ⱖ3) than when cells were incubated with IgG antibodies purified from the serum of patients with inactive GPA (BVAS score 0) or the serum of healthy subjects (mean ⫾ SD levels of H2O2 8,349 ⫾ 545 AU in active disease versus 6,705 ⫾537 AU in inactive disease and 6,544 ⫾ 759 AU in healthy subjects; P ⬍ 0.05 for each). In addition, the hydrogen peroxide levels were higher in patients with active GPA than in patients with anti–Jo1–positive myositis (6,657 ⫾ 379.9 AU; P ⬍ 0.01) or patients with anti-SSA–positive Sjögren’s syndrome (6,599 ⫾ 844.0 AU; P ⬍ 0.05) (Figure 5A). Furthermore, the desialylation of IgG antibodies by sialydase treatment in the serum of patients with inactive GPA significantly increased the production of H2O2 by neutrophils (7,519 ⫾ 567 AU; P ⫽ 0.035 versus untreated cells) (Figure 5A). Finally, in patients with active GPA, the rate of H2O2 production by neutrophils was inversely correlated with the sialylation ratio of anti-PR3 antibodies (r ⫽ ⫺0.76, P ⬍ 0.05) (Figure 5B). The increased production of H2O2, which was observed in cultures with IgG antibodies purified from the serum of patients with systemic active GPA, was significantly reduced when IgG antibodies were coincubated with blocking antibodies to the Fc receptors CD16 and CD32 (P ⫽ 0.0161 and P ⫽ 0.0073, respectively) (Figure 5C). DISCUSSION In the present study, we have shown that a low level of galactosylation and of sialylation of IgG antibodies in the serum of patients with GPA is associated with active disease. The analysis of the glycosylation of serum IgG antibodies from patients with GPA was carried out with MALDI-TOF mass spectrometry of permethylated oligosaccharides. This method can be 2112 Figure 5. Intracellular production of hydrogen peroxide (H2O2) by neutrophils incubated with IgG. A, Neutrophils were incubated with IgG antibodies purified from the serum of 4 patients with active granulomatosis with polyangiitis (Wegener’s) (GPA), 4 patients with remitted GPA, and 7 healthy subjects; for comparisons, phosphate buffered saline (PBS) alone or serum from 4 patients with anti–Jo-1– positive myositis or 4 patients with anti-SSA–positive Sjögren’s syndrome were used. In addition, the effects of desialylation of IgG antibodies on the intracellular production of H2O2 by neutrophils were assessed, using treatment of neutrophils with sialydase. B, The relationship between the sialylation ratio of anti-PR3 antibodies (Abs) and the intracellular production of H2O2 was assessed in neutrophils incubated with IgG antibodies purified from the serum of 4 patients with active GPA and 4 patients with remitted disease. C, The effects of anti-CD16 and anti-CD32 blocking antibodies on the intracellular production of H2O2 were assessed in neutrophils incubated with IgG antibodies purified from the same sera as shown in A. Bars in A and C show the mean ⫾ SD results from experiments performed in triplicate. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01. NS ⫽ not significant; AU ⫽ arbitrary units. ESPY ET AL used for relative quantitation of oligosaccharides and is as reliable as chromatographic methods for elucidating glycan profiles (24). The major N-glycans found on IgG molecules are core-fucosylated, partially truncated biantennary species, which may carry a sialic acid species and/or a bisecting N-acetylglucosamine (25,26). Defective glycosylation, especially hypogalactosylation of serum IgG antibodies, has been previously reported in ANCA-positive vasculitides, including GPA (27), but to date, the association with clinical activity has been unknown. In addition, no specific study had been conducted on the sialylation of IgG antibodies in GPA. The decreased sialylation prompted us to investigate whether this phenomenon also affects anti-PR3 antibodies, which are hallmarks of GPA. The sialylation of anti-PR3 antibodies was analyzed using the ELLA method. ELLA is a wellcharacterized, ELISA-derived method that is used to detect and quantify glycoproteins (21). The level of reactivity of the enzyme–lectin conjugate is proportional to the number of sialic residues on the antibody present. Commercially available PR3-coated plates were used in our study. False-positive results caused by the glycosylation of the PR3-coated antigen were prevented by desialylating PR3 with HIO4, as has also been previously described (21). Low sialylation levels of anti-PR3 antibodies are preferentially associated with the active stages of GPA, while higher levels are associated with inactive or weakly active disease. In our experiments, the measurement of the sialylation level of anti-PR3 antibodies, but not the levels of the anti-PR3 antibodies themselves, appeared to be useful in determining the activity of GPA. However, the published data on the significance of serum anti-PR3 levels and their variations have been so contradictory that no recommendations have been established as yet for their interpretation. Several authors have suggested that anti-PR3 antibody levels are not reliable for determining the activity and the prognosis of GPA (5–7). A recent review of 22 studies similarly concluded that serial ANCA testing should not be performed routinely in clinical practice (28), since anti-PR3 antibody levels remain elevated during clinical remission in some patients. However, some reports have suggested that persistent or reappearing cANCA within the first year are associated with a risk of relapse that is higher than that in patients with persistently undetectable cANCA (29). In addition, some studies have suggested that ANCA levels are correlated with clinical status (8,29,30–32). Thus, the significance of the levels of SIALYLATION OF ANTI-PR3 ANTIBODIES IN GPA anti-PR3 antibodies for determining the activity of GPA remains controversial. We showed herein that, whereas anti-PR3 antibody levels were of no use for assessing the activity of GPA, the sialylation levels of these antibodies could discriminate between active and inactive or weakly active GPA, when ANCAs persisted during remission. Therefore, the determination of the levels of anti-PR3 antibody sialylation could represent a biomarker that might provide strong information on the clinical activity of GPA, which could thus help significantly in the management of patients with GPA. These results suggest that the development of vasculitis could be linked with a low sialylation level of IgG antibodies and, in particular, anti-PR3 antibodies. IgG antibodies are heterogeneous with respect to the composition of the N297-attached sugar moiety. While the heptameric core, consisting of N-acetylglucosamine and mannose, is constant, a high level of heterogeneity is introduced through various additions of terminal sugar residues, such as sialic acid and galactose, and branching residues, including N-acetylglucosamine and fucose (25). More than 30 different IgG glycosylation variants exist, which remain at a remarkably constant level in the serum of healthy individuals. During autoimmune diseases, such as rheumatoid arthritis and systemic lupus erythematosus, however, this glycosylation pattern changes in mice and humans, and IgG glycovariants (IgG-G0) lacking terminal sialic acid and galactose can be found more frequently (33,34). There is great interest in understanding whether this altered glycosylation pattern influences antibodymediated effector functions. In vitro studies have suggested that IgG-G0 antibodies gain the capacity to activate the complement pathway via mannose-binding lectin (MBL), which could contribute to antibodymediated inflammation. Nimmerjahn et al (36) have analyzed the activity of IgG-G0 antibodies in mice with a genetic deletion of MBL (MBL-null mice), with the results demonstrating that IgG-G0 antibodies are unimpaired in MBL-null mice. In contrast, those authors have shown that the activity of these antibody glycovariants is fully dependent on the presence of activating Fc receptors, and that asialylated antibodies result in enhanced Fc receptor binding and, thus, enhanced pathogenicity in vivo (35,36). In contrast, high levels of terminal sialic acid residues of the IgG molecules resulted in a reduced affinity of IgG for all mouse Fc␥ receptors and impaired the activity of human IgG in a similar manner (37,38). Despite this reduction in affinity, IgG rich in sialic acid had an enhanced antiinflammatory activity. Interestingly, the linkage of sialic acid to the galactose residue 2113 was of great importance. Whereas both the 2,3- and 2,6-linked sialic acid residues resulted in a reduced affinity for cellular Fc␥ receptors, only the IgG molecules with 2,6-linked sialic acid had the capacity to suppress autoimmune diseases (12). Our results are consistent with those results, since we observed not only a hypogalactosylation, but also a 2,6-hyposialylation of the IgG molecules in patients with active GPA, but not in patients with remitted disease. Finally, a low level of 2,6-sialylation of anti-PR3 IgG antibodies could contribute to stimulate cell activation through the Fc receptors and augment their capacity to activate neutrophils, which are key cells in the pathogenesis of GPA. Our in vitro data confirm this hypothesis, since the oxidative burst of PMNs induced by the hyposialylated anti-PR3 antibodies from the serum of patients with active GPA was stronger than that induced by the other IgG antibodies tested. The observation that the desialylation of IgG antibodies in patients with remitted GPA induced a stronger oxidative burst than did nondesialylated IgG antibodies also supports the notion of a pivotal role of glycosylation of anti-PR3 IgG antibodies in the pathogenesis of GPA. The reason sialylation levels are decreased in systemic autoimmune disease could be related to the recent demonstration that ligand binding to Toll-like receptors 2, 3, and 4 can induce Neu1 sialydase activity within minutes in myeloid cells (39). In those cells, lipopolysaccharide-induced sialydase activity is dependent on p42/p44 MAPK-mediated tumor necrosis factor ␣ (TNF␣) production. Subsequently, the TNF␣ that is produced mediates p38 MAPK activation, which then induces sialydase activity (40). TNF␣ plays a key role in GPA, and its increased level is associated with disease activity (41). Therefore, we can hypothesize that, during active GPA, TNF␣ induces sialydase activity, which favors the production of desialylated anti-PR3 antibodies that strongly activate neutrophils and macrophages through their Fc␥ receptors to produce reactive oxygen species and proinflammatory cytokines, such as TNF␣, thus inducing a pathogenic loop (42). Our study has therefore shown that the sialylation level of anti-PR3 antibodies is associated with the clinical activity of GPA; of note, the sialylation level of anti-PR3 antibodies was found to be lower in patients with active GPA, suggesting that it could also be considered a more accurate biomarker of clinical activity than is the level of anti-PR3 antibodies. Additional prospective studies are required to confirm whether the sialylation level of anti-PR3 antibodies determines the clinical 2114 ESPY ET AL activity of GPA through the modulation of the intensity of the oxidative burst of neutrophils in vivo. AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Batteux had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Espy, Grange, Pagnoux, Guillevin, Weill, Batteux, Guilpain. Acquisition of data. Espy, Morelle, Kavian, Goulvestre, Chéreau, Pagnoux, Guilpain. Analysis and interpretation of data. Espy, Morelle, Kavian, Viallon, Chéreau, Michalski, Batteux, Guilpain. REFERENCES 1. Davies DJ, Moran JE, Niall JF, Ryan GB. Segmental necrotising glomerulonephritis with antineutrophil antibody: possible arbovirus aetiology? Br Med J (Clin Res Ed) 1982;285:606. 2. Falk RJ, Jennette JC. Anti-neutrophil cytoplasmic autoantibodies with specificity for myeloperoxidase in patients with systemic vasculitis and idiopathic necrotizing and crescentic glomerulonephritis. N Engl J Med 1988;318:1651–7. 3. Van der Woude FJ, Rasmussen N, Lobatto S, Wiik A, Permin H, van Es LA, et al. Autoantibodies against neutrophils and monocytes: tool for diagnosis and marker of disease activity in Wegener’s granulomatosis. Lancet 1985;1:425–9. 4. Falk RJ, Terrell RS, Charles LA, Jennette JC. Anti-neutrophil cytoplasmic autoantibodies induce neutrophils to degranulate and produce oxygen radicals in vitro. Proc Natl Acad Sci U S A 1990;87:4115–9. 5. Hellmich B, Flossmann O, Gross WL, Bacon P, Cohen-Tervaert JW, Guillevin L, et al. EULAR recommendations for conducting clinical studies and/or clinical trials in systemic vasculitis: focus on anti-neutrophil cytoplasm antibody-associated vasculitis. Ann Rheum Dis 2007;66:605–17. 6. Finkielman JD, Merkel PA, Schroeder D, Hoffman GS, Spiera R, St.Clair EW, et al, for the WGET Research Group. Antiproteinase 3 antineutrophil cytoplasmic antibodies and disease activity in Wegener granulomatosis. Ann Intern Med 2007;147:611–9. 7. Langford CA. Antineutrophil cytoplasmic antibodies should not be used to guide treatment in Wegener’s granulomatosis. Clin Exp Rheumatol 2004;22:S3–6. 8. Boomsma MM, Stegeman CA, van der Leij MJ, Oost W, Hermans J, Kallenberg CG, et al. Prediction of relapses in Wegener’s granulomatosis by measurement of antineutrophil cytoplasmic antibody levels: a prospective study. Arthritis Rheum 2000;43: 2025–33. 9. Kaveri S, Prasad N, Vassilev T, Hurez V, Pashov A, LacroixDesmazes S, et al. Modulation of autoimmune responses by intravenous immunoglobulin (IVIg). Mult Scler 1997;3:121–8. 10. Siberil S, de Romeuf C, Bihoreau N, Fernandez N, Meterreau JL, Regenman A, et al. Selection of a human anti-RhD monoclonal antibody for therapeutic use: impact of IgG glycosylation on activating and inhibitory Fc␥R functions. Clin Immunol 2006;118: 170–9. 11. Siberil S, Dutertre CA, Boix C, Bonnin E, Menez R, Stura E, et al. Molecular aspects of human Fc␥R interactions with IgG: functional and therapeutic consequences. Immunol Lett 2006;106: 111–8. 12. Anthony RM, Nimmerjahn F, Ashline DJ, Reinhold VN, Paulson JC, Ravetch JV. Recapitulation of IVIG anti-inflammatory activity with a recombinant IgG Fc. Science 2008;320:373–6. 13. Jennette JC, Falk RJ, Andrassy K, Bacon PA, Churg J, Gross WL, et al. Nomenclature of systemic vasculitides: proposal of an international consensus conference. Arthritis Rheum 1994;37: 187–92. 14. Leavitt RY, Fauci AS, Bloch DA, Michel BA, Hunder GG, Arend WP, et al. The American College of Rheumatology 1990 criteria for the classification of Wegener’s granulomatosis. Arthritis Rheum 1990;33:1101–7. 15. Flossmann O, Bacon P, de Groot K, Jayne D, Rasmussen N, Seo P, et al. Development of comprehensive disease assessment in systemic vasculitis. Ann Rheum Dis 2007;66:283–92. 16. Luqmani RA, Bacon PA, Moots RJ, Janssen BA, Pall A, Emery P, et al. Birmingham Vasculitis Activity Score (BVAS) in systemic necrotizing vasculitis. QJM 1994;87:671–8. 17. Stone JH, Hoffman GS, Merkel PA, Min YI, Uhlfelder ML, Hellmann DB, et al, for the International Network for the Study of the Systemic Vasculitides (INSSYS). A disease-specific activity index for Wegener’s granulomatosis: modification of the Birmingham Vasculitis Activity Score. Arthritis Rheum 2001;44:912–20. 18. Ciucanu I, Kerek F. A simple and rapid method for the permethylation of carbohydrates. Carbohydr Res 1984;131:209–17. 19. Morelle W, Michalski JC. Analysis of protein glycosylation by mass spectrometry. Nat Protoc 2007;2:1585–602. 20. Wiik A. Delineation of a standard procedure for indirect immunofluorescence detection of ANCA. APMIS Suppl 1989;6:12–3. 21. Gornik O, Lauc G. Enzyme linked lectin assay (ELLA) for direct analysis of transferrin sialylation in serum samples. Clin Biochem 2007;40:718–23. 22. Dooley DC, Simpson JF, Meryman HT. Isolation of large numbers of fully viable human neutrophils: a preparative technique using percoll density gradient centrifugation. Exp Hematol 1982;10: 591–9. 23. Servettaz A, Guilpain P, Goulvestre C, Chereau C, Hercend C, Nicco C, et al. Radical oxygen species production induced by advanced oxidation protein products predicts clinical evolution and response to treatment in systemic sclerosis. Ann Rheum Dis 2007;66:1202–9. 24. Wada Y, Azadi P, Costello CE, Dell A, Dwek RA, Geyer H, et al. Comparison of the methods for profiling glycoprotein glycans— HUPO Human Disease Glycomics/Proteome Initiative multi-institutional study [published erratum appears in Glycobiology 2007; 17:10G]. Glycobiology 2007;17:411–22. 25. Arnold JN, Wormald MR, Sim RB, Rudd PM, Dwek RA. The impact of glycosylation on the biological function and structure of human immunoglobulins. Annu Rev Immunol 2007;25:21–50. 26. Endo T, Kochibe N, Kobata A. Structural study of the carbohydrate moieties of two human immunoglobulin subclasses (IgG2 and IgG4). Glycoconj J 1989;6:57–66. 27. Holland M, Takada K, Okumoto T, Takahashi N, Kato K, Adu D, et al. Hypogalactosylation of serum IgG in patients with ANCAassociated systemic vasculitis. Clin Exp Immunol 2002;129:183–90. 28. Birck R, Schmitt WH, Kaelsch IA, van der Woude FJ. Serial ANCA determinations for monitoring disease activity in patients with ANCA-associated vasculitis: systematic review. Am J Kidney Dis 2006;47:15–23. 29. Stegeman CA. Anti-neutrophil cytoplasmic antibody (ANCA) levels directed against proteinase-3 and myeloperoxidase are helpful in predicting disease relapse in ANCA-associated smallvessel vasculitis. Nephrol Dial Transplant 2002;17:2077–80. 30. Han WK, Choi HK, Roth RM, McCluskey RT, Niles JL. Serial ANCA titers: useful tool for prevention of relapses in ANCAassociated vasculitis. Kidney Int 2003;63:1079–85. SIALYLATION OF ANTI-PR3 ANTIBODIES IN GPA 31. Gaskin G, Savage CO, Ryan JJ, Jones S, Rees AJ, Lockwood CM, et al. Anti-neutrophil cytoplasmic antibodies and disease activity during long-term follow-up of 70 patients with systemic vasculitis. Nephrol Dial Transplant 1991;6:689–94. 32. Tervaert JW, Huitema MG, Hene RJ, Sluiter WJ, The TH, van der Hem GK, et al. Prevention of relapses in Wegener’s granulomatosis by treatment based on antineutrophil cytoplasmic antibody titre. Lancet 1990;336:709–11. 33. Mizuochi T, Hamako J, Nose M, Titani K. Structural changes in the oligosaccharide chains of IgG in autoimmune MRL/Mp-lpr/lpr mice. J Immunol 1990;145:1794–8. 34. Parekh RB, Dwek RA, Sutton BJ, Fernandes DL, Leung A, Stanworth D, et al. Association of rheumatoid arthritis and primary osteoarthritis with changes in the glycosylation pattern of total serum IgG. Nature 1985;316:452–7. 35. Malhotra R, Wormald MR, Rudd PM, Fischer PB, Dwek RA, Sim RB. Glycosylation changes of IgG associated with rheumatoid arthritis can activate complement via the mannose-binding protein. Nat Med 1995;1:237–43. 36. Nimmerjahn F, Anthony RM, Ravetch JV. Agalactosylated IgG antibodies depend on cellular Fc receptors for in vivo activity. Proc Natl Acad Sci U S A 2007;104:8433–7. 37. Kaneko Y, Nimmerjahn F, Ravetch JV. Anti-inflammatory activ- 2115 38. 39. 40. 41. 42. ity of immunoglobulin G resulting from Fc sialylation. Science 2006;313:670–3. Scallon BJ, Tam SH, McCarthy SG, Cai AN, Raju TS. Higher levels of sialylated Fc glycans in immunoglobulin G molecules can adversely impact functionality. Mol Immunol 2007;44:1524–34. Amith SR, Jayanth P, Franchuk S, Siddiqui S, Seyrantepe V, Gee K, et al. Dependence of pathogen molecule-induced Toll-like receptor activation and cell function on Neu1 sialidase. Glycoconj J 2009;26:1197–212. Gee K, Kozlowski M, Kumar A. Tumor necrosis factor-␣ induces functionally active hyaluronan-adhesive CD44 by activating sialidase through p38 mitogen-activated protein kinase in lipopolysaccharide-stimulated human monocytic cells. J Biol Chem 2003;278: 37275–87. Nassonov EL, Samsonov MY, Tilz GP, Beketova TV, Semenkova EN, Baranov A, et al. Serum concentrations of neopterin, soluble interleukin 2 receptor, and soluble tumor necrosis factor receptor in Wegener’s granulomatosis. J Rheumatol 1997;24:666–70. Reumaux D, Vossebeld PJ, Roos D, Verhoeven AJ. Effect of tumor necrosis factor-induced integrin activation on Fc␥ receptor II-mediated signal transduction: relevance for activation of neutrophils by anti-proteinase 3 or anti-myeloperoxidase antibodies. Blood 1995;86:3189–95.