Induction of endothelial cell apoptosis by the binding of antiendothelial cell antibodies to Hsp60 in vasculitis-associated systemic autoimmune diseases.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 52, No. 12, December 2005, pp 4028–4038 DOI 10.1002/art.21401 © 2005, American College of Rheumatology Induction of Endothelial Cell Apoptosis by the Binding of Anti–Endothelial Cell Antibodies to Hsp60 in Vasculitis-Associated Systemic Autoimmune Diseases Christophe Jamin,1 Christophe Dugué,1 Jean-Éric Alard,1 Sandrine Jousse,1 Alain Saraux,1 Loı̈c Guillevin,2 Jean-Charles Piette,3 and Pierre Youinou1 anti-Hsp60–containing AECA-positive sera and was inhibited by preincubation of the ECs with recombinant Hsp60. Conclusion. Our data support the notion that Hsp60 is an important target for AECAs and that such an interaction contributes to pathogenic effects, especially in vasculitis-associated systemic autoimmune disease. Objective. Anti–endothelial cell antibodies (AECAs), which recognize a number of endothelial antigens, are seen in patients with systemic autoimmune diseases, more often in the presence of vasculitis than in its absence. Some AECAs induce apoptosis of endothelial cells (ECs), but their target antigens remain unknown. The aim of this study was to determine whether Hsp60 is a target antigen and whether AECAs induce apoptosis in ECs. Methods. Two-dimensional electrophoresis and conventional Western blotting techniques were used to characterize AECA targets. Hsp60 reactivity was determined by enzyme-linked immunosorbent assay. Results. Hsp60 was shown to be targeted by a proportion of AECAs. The level of reactivity was higher in patients with systemic autoimmune disease and vasculitis than in those without vasculitis and in patients with systemic lupus erythematosus than in patients with other systemic autoimmune diseases. Hsp60 was expressed on the plasma membrane of heat-stressed ECs, and this followed Hsp60 messenger RNA transcription, confinement of the protein to the cytoplasm, and translocation of the protein to the surface. Shedding of Hsp60 from ECs was induced by stress and resulted in the binding of soluble Hsp60 to the surface of ECs, particularly stressed ECs. Apoptosis of ECs was triggered by The 60-kd heat-shock protein (Hsp60) acts as a chaperone in the folding, translocation, and assembly of protein complexes throughout their posttranslational maturation (1). Although originally restricted to the mitochondrial matrix, Hsp60 is then exported to extramitochondrial sites, most particularly the plasma membrane (2). As a result, it becomes detectable on different cell types, including oligodendrocytes (3), macrophages (4), B cell lymphoma cells (5), pancreatic carcinoma cells (6), epithelial cells (7), and endothelial cells (ECs) (8). It has also been established that a diverse array of metabolic stimuli, such as hypoxia, heat shock, fluid pressure, cytokines, and inflammatory reactions, induce the expression of Hsp60 on ECs (8,9). In these settings, Hsp60 expression acts as a protective mechanism to maintain homeostasis. This seminal finding prompted studies of the immunologic consequences of Hsp60 expression, such as those in which it presents immunodominant antigens (10). Specific T cell activation and anti-Hsp60 antibody production have been characterized in infectious diseases, including leprosy, tuberculosis, and malaria (11). Autoantibodies to Hsp60 have also been identified in systemic autoimmune diseases, including rheumatoid arthritis (RA) (12), systemic lupus erythematosus (SLE) (13), and diabetes mellitus (14). In fact, immune reactivity to Hsp60 has been most extensively analyzed in atherosclerosis, which has thus been considered to be an Supported by a grant from the Brest Métropole Océane and from the Conseil Régional de Bretagne, France. 1 Christophe Jamin, PhD, Christophe Dugué, PhD, Jean-Éric Alard, BSc, Sandrine Jousse, MD, Alain Saraux, MD, PhD, Pierre Youinou, MD, PhD: Brest University Medical School, Brest, France; 2 Loı̈c Guillevin, MD: Cochin Hospital, Paris, France; 3Jean-Charles Piette, MD: La Pitié-Salpêtrière Hospital, Paris, France. Address correspondence and reprint requests to Pierre Youinou, MD, PhD, Immunology, Brest University Medical School Hospital, BP824, Brest F29609, France. E-mail: firstname.lastname@example.org. Submitted for publication April 19, 2005; accepted in revised form August 2, 2005. 4028 INDUCTION OF ENDOTHELIAL CELL APOPTOSIS BY ANTI-Hsp60 ANTIBODIES autoimmune condition (8,15). Although the association of anti-Hsp60 antibody with disease has not been completely characterized (16), the general perception is that the interaction of autoantibodies with Hsp60 proteins might be involved in the pathogenesis of systemic autoimmune diseases. They trigger complement-mediated cytotoxicity and antibody-dependent cell-mediated cytotoxicity (8) at the expense of macrophages and ECs. Several investigators, including our group, have reported the presence of anti-EC antibodies (AECAs) in autoimmune conditions, such as SLE (17), polyarteritis nodosa (PAN) (18), and systemic sclerosis (SSc) (19). Furthermore, AECAs have also been found in patients with infectious diseases, such as leprosy (20), hepatitis virus C infection (21), and cytomegalovirusinfected recipients of cardiac or renal allografts (22), as well as in patients with borderline hypertension (23). Interestingly, the binding of certain AECAs induces apoptosis in ECs, both in vitro (24) and in vivo (25). Given the impressive diversity of diseases associated with AECAs, their target antigens are likely to be heterogeneous. This study was aimed at determining whether Hsp60 is one of the target antigens of AECAs and whether such autoantibodies induce the apoptosis of ECs. Using recombinant Hsp60, we demonstrated that reactivity of a group of AECAs with this antigen is more frequent in patients with systemic autoimmune diseases who also have vasculitis than in those who do not. Furthermore, autoantibodies that bind Hsp60 exposed on the surface of heat-stressed ECs or shed Hsp60 that is captured by unknown receptors on the surface of ECs induce apoptosis. These results suggest a role of AECA interaction with Hsp60 in the pathogenesis of vasculitisassociated systemic autoimmune diseases. PATIENTS AND METHODS Patients. Sera were obtained from 2 groups of patients with systemic autoimmune disease. The first group comprised 176 patients with systemic autoimmune disease associated with vasculitis: 51 with SLE, 45 with Wegener’s granulomatosis (WG), 36 with PAN, 16 with microscopic polyangiitis (MPA), 16 with Churg-Strauss syndrome (CSS), and 12 with Behçet’s disease. The second group consisted of 95 patients with systemic autoimmune disease, but without vasculitis: 26 with polymyositis, 15 with RA, 41 with primary Sjögren’s syndrome (SS), and 13 with SSc. All patients fulfilled the criteria for their respective diseases. Because patients with RA (26) and patients with primary SS (27) can potentially present with features of vasculitis, we took care that none of them had signs of vasculitis at the time their serum samples were collected. Sera from 30 healthy laboratory and clinical staff 4029 members were used as controls to set the cutoff level for AECA and anti-Hsp60 antibody positivity in our enzymelinked immunosorbent assays (ELISAs). In the 10 patient subgroups and in the control population, the ratios of women to men were ⬃3:1, and the mean ages of the subjects ranged from 23 to 66 years. Measurement of EC reactivity. EA.hy926 cells were kindly donated by Dr. Cora-Jean S. Edgell (University of North Carolina, Chapel Hill, NC). AECAs were identified by Western blotting on EA.hy926 cell lysates, using a membrane preparation method described in detail elsewhere (28). The pattern of reactivity of each serum was determined using a G-700 imaging densitometer and analyzed with Molecular Analyst software (both from Bio-Rad, Marnes la Coquette, France). A serum was considered positive when enhanced intensity for a given peak (optical density [OD] ⬎0.1) was identified relative to the controls or when an additional peak over an OD threshold arbitrarily set at 0.1 was found. Two-dimensional electrophoresis (2DE) followed by Western blotting were also performed as previously described (20). For isoelectric focusing, protein samples were loaded onto nonlinear, pH 3–10, immobilized pH gradient strips (Pharmacia, Uppsala, Sweden) and focused for 45,000 V/hour. The strips were incubated for 20 minutes at room temperature with 50 ml of an equilibration buffer containing 0.05M Tris HCl, pH 6.8, 6M urea, 1% sodium dodecyl sulfate, 30% glycerol, and 0.5% dithiothreitol (DTT). The strips were then incubated in the equilibration solution, which contained 4.5% iodoacetamide (Sigma-Aldrich, St. Louis, MO) instead of 0.5% DTT. Then, 5–15% acrylamide–bisacrylamide gradient gels prepared in our laboratory were used to separate the proteins. The strips were loaded onto the top of the slab gel, subjected to electrophoresis for 4 hours at 50 mA per gel, and transferred to polyvinylidene difluoride (PVDF) membranes. After 2 hours of blocking with phosphate buffered saline (PBS) containing 2% milk protein and 0.05% Tween 20, the PVDF membranes were probed for 2 hours with sera diluted 1:50 in PBS supplemented with 1% milk protein and 0.05% Tween 20. Bound AECAs were revealed with biotinylated anti-human IgG antibody (Jackson ImmunoResearch, West Grove, PA), and developed with horseradish peroxidase (HRP)– streptavidin (Amersham Biosciences, Orsay, France). Autoantigen spots were identified in reference to the EA.hy926 2DE gel map from the Swiss-Prot database (available at http:// www.harefield.nthames.nhs.uk). Measurement of Hsp60 reactivity. The presence of anti-Hsp60 antibodies was determined by Western blotting using recombinant Hsp60 as the substrate (Sigma-Aldrich). This protein was loaded onto a 10% acrylamide–bisacrylamide gel, subjected to electrophoresis, and transferred to a PVDF membrane. Western blotting with AECA-positive sera was performed as described above. After densitometric analysis, the mean ⫹ 2SD of the OD value in control sera was taken as the threshold for positivity. Hsp60 reactivity was also detected using an ELISA developed in our laboratory (29). Briefly, ELISA microtiter wells were sensitized with 0.5 g of recombinant Hsp60, and blocked with 2% milk protein. After 3 washes with 0.1% Tween 20 in PBS, 100 l of serum diluted 1:100 in PBS containing 0.1% Tween 20 was added in triplicate, and plates were incubated for 2 hours at 37°C. Incubation with alkaline 4030 phosphatase–conjugated anti-human IgG antibody (Zymed, South San Francisco, CA) was then performed for a further 2 hours at 37°C, followed by incubation with p-nitrophenyl phosphate. The absorbance was measured at 405 nm, and the OD of control wells without Hsp60 was subtracted from the OD of Hsp60-coated wells. The mean ⫹ 2SD of the OD value in control sera was taken as the threshold for positivity. Detection of Hsp60 expression. Human umbilical vein ECs (HUVECs) were prepared by digestion in 0.1% collagenase (Sigma-Aldrich) and grown to confluence, as previously described (20). The HUVECs were then passaged twice, harvested, and seeded at 105 cells in 100 l of Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal calf serum (FCS) and 100 IU/ml of polymyxin B. Resting HUVECs or HUVECs subjected to stress by exposure to temperatures of 42.5°C for 1–5 hours were permeabilized with 0.5% saponin (Sigma-Aldrich), then tested for the expression of Hsp60. Two hundred microliters of this cell suspension was incubated with anti-Hsp60 monoclonal antibody (mAb; VWR International, Fontenay-sous-Bois, France) for 1 hour at 4°C. The cells were then incubated with fluorescein isothiocyanate (FITC)–conjugated anti-mouse IgG antibody, washed, and suspended in 400 l of PBS for fluorescence-activated cell sorter (FACS) analysis in an Epics Elite flow cytometer (Coulter, Hialeah, FL). Unstained cells and the FITCconjugated developing reagent alone were the controls used to set the level of positivity. For indirect immunofluorescence (IIF) staining, subconfluent HUVECs were permeabilized and fixed in 20% ethanol plus 80% acetone for 5 minutes following application of stress. Cells were then incubated with anti-Hsp60 mAb and revealed using FITC-conjugated anti-mouse IgG antibody. Slides were mounted in glycerol and examined with an Axioplan fluorescence microscope (Zeiss, Le Pecq, France). Soluble Hsp60 released into the supernatant of cultured cells was detected by dot-blot analysis. Untreated or heat-stressed HUVECs were incubated for 18 hours at 37°C. Soluble proteins were precipitated with acetone and loaded onto a nitrocellulose membrane. The membrane was blocked with 2% milk protein, incubated with anti-Hsp60 mAb for 1 hour, revealed with biotinylated anti-mouse IgG antibodies (Jackson ImmunoResearch), and visualized with HRP– streptavidin. Recombinant Hsp60 was used as a positive control. Expression of Hsp60 messenger RNA (mRNA) was evaluated by quantitative reverse transcriptase–polymerase chain reaction (RT-PCR). HUVECs were subjected to heat stress or were untreated, and mRNA was purified. For synthesis of complementary DNA (cDNA), 1 g of mRNA was incubated for 5 minutes at 68°C and for 1 minute on ice with an oligo(dT) and deoxynucleotide mixture in a final volume of 11 l. Seven microliters of PCR buffer was added and incubated for 2 minutes at 42°C. One microliter of Superscript II RT (Invitrogen, Carlsbad, CA) was added, and incubation continued for 50 minutes at 40°C and then for 15 minutes at 70°C. Finally, 1 l of RNase H was added and incubated for 10 minutes at 37°C. For PCR amplification, the reactions were performed in an ABI Prism 7000 system (Applied Biosystems, Warrington, UK) with 1 l of cDNA, 0.5 l of Taq DNA polymerase (5 units/l; Invitrogen), and 2 l of the primer pairs 5⬘-GATGCTGCTGGTGTGGCCTCTCTG-3⬘ and 5⬘-GT- JAMIN ET AL Table 1. Frequencies of AECAs and of anti-Hsp60 antibodies in AECA-positive sera from patients with vasculitic and nonvasculitic systemic autoimmune diseases* Frequency (no.) of patients Anti-Hsp60 Disease AECA By ELISA By Western blotting SLE (n ⫽ 51) WG (n ⫽ 45) PAN (n ⫽ 36) MPA (n ⫽ 16) CSS (n ⫽ 16) BD (n ⫽ 12) PM (n ⫽ 26) RA (n ⫽ 15) Primary SS (n ⫽ 41) SSc (n ⫽ 13) 74 (38) 80 (36) 53 (19) 50 (8) 62 (10) 75 (9) 38 (10) 67 (10) 24 (10) 77 (10) 39 (15) 44 (16) 42 (8) 25 (2) 0 22 (2) 30 (3) 10 (1) 10 (1) 10 (1) 76 (29) 56 (20) 79 (15) 62 (5) 20 (2) 44 (4) 30 (3) 20 (2) 20 (2) 10 (1) * Anti–endothelial cell antibody (AECA) reactivity was determined by Western blotting. Anti-Hsp60 reactivity was determined by enzymelinked immunosorbent assay (ELISA) and by Western blotting. SLE ⫽ systemic lupus erythematosus; WG ⫽ Wegener’s granulomatosis; PAN ⫽ polyarteritis nodosa; MPA ⫽ microscopic polyangiitis; CSS ⫽ Churg-Strauss syndrome; BD ⫽ Behçet’s disease; PM ⫽ polymyositis; RA ⫽ rheumatoid arthritis; SS ⫽ Sjögren’s syndrome; SSc ⫽ systemic sclerosis. GAGGAACACTGCCTTGGGCTTC-3⬘ for Hsp60, and 5⬘GGCTACCACATCCAAGGAAGGCAG-3⬘ and 5⬘-TTGACACTGGCAAAACAATGCAGAC-3⬘ for 18S as follows. After an initial denaturation step, 35 cycles of PCR, consisting of 1 minute at 58°C and 1 minute at 72°C, were performed, with a final 10-minute extension at 72°C. The relative induction or downregulation of gene expression was obtained using the comparative threshold cycle (Ct) method and the 2–⌬Ct formula, as follows: ⌬Ct ⫽ Ct Hsp60 – Ct 18S. The relative fold induction of Hsp60 mRNA was estimated, with normalization of the values found after 1–5 hours of heat-induced stress to the value recorded before the application of stress. RT-PCR products were visualized on 2% agarose gels using ethidium bromide. Assessment of apoptosis. IgG fractions of control sera and patient sera positive for anti-Hsp60 were purified over a protein G–Sepharose column (Sigma-Aldrich). Their concentrations were determined by an ELISA developed in our laboratory. HUVECs were harvested from the second passage of the cultured cells, suspended at a density of 104 cells in 100 l of DMEM supplemented with 10% FCS and 100 IU/ml of polymyxin B, and distributed into 3 sets of duplicate wells. Pilot experiments were conducted to determine how much Hsp60 was required to obtain maximal inhibition of the binding of IgG to Hsp60 coated onto an ELISA plate. Then, 100 l of purified IgG at a concentration of 320 g/ml in the culture medium was preincubated or was not preincubated with 2.5 g/ml of Hsp60 and added to a final concentration of 160 g/ml of IgG. Following a 24-hour incubation, apoptosis was evaluated by measuring FITC-conjugated annexin V staining in association with propidium iodide (PI) to exclude dead cells (kit obtained from Beckman-Coulter, Villepinte, France). Per- INDUCTION OF ENDOTHELIAL CELL APOPTOSIS BY ANTI-Hsp60 ANTIBODIES 4031 Figure 1. Pattern of reactivity of sera from patients with different systemic autoimmune diseases in 1-dimensional and 2-dimensional Western blot analyses. A, EA.hy926 cell lysates were subjected to electrophoresis and then blotted with sera from patients and controls. Representative results for sera from patients with systemic lupus erythematosus (SLE), Wegener’s granulomatosis (WG), polyarteritis nodosa (PAN), microscopic polyangiitis (MPA), and Churg-Strauss syndrome (CSS), as well as serum from a healthy control subject are shown. Molecular weight markers are shown at the left. B, Densitometric analysis of SLE reactivity (thin line) compared with control reactivity (thick line). Antigens strongly recognized by SLE sera are indicated. A.U. ⫽ arbitrary units. C, Electrophoresis was performed on EA.hy926 cell lysates and proteins blotted with sera from patients with SLE, WG, PAN, MPA, and CSS, as well as serum from a healthy control subject. Representative patterns for 2 sera for each disease are shown. Large arrowheads indicate spots for which the molecular weight and the isoelectric point match those of Hsp60. Molecular weight markers are shown at the left. D, On another membrane, the spot corresponding to the area indicated by the large arrowheads in C was cut out, the amino acids were sequenced, and the peptide was aligned with the NH2-terminal end of the Hsp60 protein (Swiss-Prot accession no. P10809), revealing sequence identity with a 19–amino acid segment. 4032 JAMIN ET AL centages of annexin V–positive cells within the PI-negative population were calculated. Cells in the early stage of apoptosis were positive for annexin V and negative for PI, because PI-positive cells were recorded as dead cells. As in previous studies (24,30), these annexin V data were confirmed by findings of nuclear condensation, hypoploidy of the cells, and DNA fragmentation, and by TUNEL staining. RESULTS Identification of Hsp60 as a target antigen for AECAs. Using 1-dimensional Western blotting with EA.hy926 cell lysates as the substrate, EC reactivity was identified in 176 sera from patients with various systemic autoimmune diseases associated with vasculitis and in 95 sera from patients with nonvasculitic systemic autoimmune diseases. Among the subgroups of patients with associated vasculitis, 50–80% were positive for AECAs, and among the subgroups of patients without associated vasculitis, 24–77% were positive for AECAs, depending on the diagnosis (Table 1). Western blot analyses revealed a range of bands in the sera of patients with systemic autoimmune diseases (Figure 1A). Densitometric analysis indicated that, compared with the control sera, there was greater reactivity with several antigens in the SLE sera (Figure 1B) and the WG sera (results not shown). Among these reactivities was a 60-kd band. To identify the nature of the bands, a 2DE was performed on EA.hy926 cell lysates blotted with randomly selected sera from patients in each disease subgroup. Noteworthy was the restriction of 1 spot to SLE and WG sera (Figure 1C). The pI of the spot was 5.7, and its molecular weight was 60-kd. These stereotypical characteristics fit the Hsp60 protein in the databases. To more accurately confirm its nature, a PVDF membrane was stained with Ponceau S solution, the corresponding spot was cut out, and the protein was sequenced by the Edman degradation method. Given that the identified peptide (Figure 1D) matched exactly with the NH2-terminal end of Hsp60 (Swiss-Prot accession no. P10809), this was confirmed to be the target for AECAs in the positive SLE and WG sera. Although disease activity and drug regimen are most likely to be influential in Hsp60 reactivity, such associations could not be established, particularly in SLE, WG, and PAN. Screening of AECA-positive sera. AECA-positive sera were further screened by Western blotting using recombinant Hsp60. Sera were considered positive for Hsp60 when the OD was ⬎0.170 (mean ⫹ 2SD of the control) (Figure 2A). Overall, more sera from patients with vasculitic systemic autoimmune diseases were positive for Hsp60 (Figure 2B), except for sera from the Figure 2. Reactivity of anti–endothelial cell antibody (anti-AECA)– positive sera with Hsp60. A, Densitometric analysis of the reactivity of sera from patients with systemic autoimmune diseases, with and without vasculitis, against Hsp60 recombinant protein by Western blotting. Broken line shows the cutoff level for positivity (optical density 0.170, representing the mean ⫹ 2SD of control values). B, Percentage of sera positive for AECAs. SLE ⫽ systemic lupus erythematosus; PM ⫽ polymyositis; SS ⫽ Sjögren’s syndrome; RA ⫽ rheumatoid arthritis; SSc ⫽ systemic sclerosis; PAN ⫽ polyarteritis nodosa; MPA ⫽ microscopic polyangiitis; WG ⫽ Wegener’s granulomatosis; BD ⫽ Behçet’s syndrome; CSS ⫽ Churg-Strauss syndrome. CSS patients, than sera from patients with nonvasculitic systemic autoimmune diseases. The percentage of positivity ranged from 44% to 79% and from 10% to 30% in the respective groups (P ⬍ 0.01). This Hsp60 reactivity was confirmed in the ELISA, where sera were considered positive when the OD was ⬎0.150. Although the sensitivity of this ELISA was not as high as that of the Western blotting (Table 1), again, there were more positive sera among the vasculitic than among the nonvasculitic systemic autoimmune disease groups (22–44% versus 10–30%; P ⬍ 0.01). Membrane expression of Hsp60. Next, experiments were performed to determine how Hsp60 is expressed on EC membranes and thereby becomes accessible to autoantibodies. It was impossible to detect Hsp60 by FACS analysis in the cytosol or on the cell INDUCTION OF ENDOTHELIAL CELL APOPTOSIS BY ANTI-Hsp60 ANTIBODIES 4033 Figure 3. Hsp60 expression on endothelial cells (ECs). Permeabilized (⫹ saponin) or nonpermeabilized (– saponin) human umbilical vein ECs were analyzed by fluorescence-activated cell sorting before (A) and after exposure to heat-induced stress for 4 hours (B) and 5 hours (C) (see Patients and Methods for details). The expression of Hsp60 was evaluated using IgG anti-Hsp60 monoclonal antibody (mAb), as revealed by fluorescein isothiocyanate (FITC)–conjugated antibody. Thin lines indicate fluorescence obtained with a control antibody. The percentage of positive cells as well as the mean fluorescence intensity (in parentheses) are indicated. ECs were also analyzed by indirect immunofluorescence before (D) and after exposure to heat-induced stress for 3 hours (E), 4 hours (F), and 5 hours (G). The presence of Hsp60 was evaluated using IgG anti-Hsp60 mAb, as revealed by FITC-conjugated antibody. Bar ⫽ 30 m. Expression of Hsp60 mRNA was then determined after different durations of heat–induced stress (H). Total mRNA was extracted from ECs, the quantity of Hsp60 mRNA was normalized against that of 18S, and the numbers of copies were compared with those in freshly isolated ECs. membrane of freshly isolated cells (Figure 3A). This lack of Hsp60 detection was verified by IIF analysis of permeabilized cells (Figure 3D). Given that Hsp60 may be induced on the surface of arterial (31) and venous (32) ECs by stress, we analyzed this phenomenon. Hsp60 protein remained undetectable after 3 hours of exposure to heat (Figure 3E). After 4 hours of heat stress, Hsp60 was demonstrated by FACS (Figure 3B) and by IIF 4034 JAMIN ET AL from ECs and subsequently captured on their surface. Dot-blot analysis revealed soluble Hsp60 only in culture supernatants of stressed ECs (Figure 4A). This established their capacity to release newly synthesized Hsp60 protein. Once released, Hsp60 bound to a small proportion of ECs when they were fresh, but to most of them after exposure to stress (Figure 4B). This suggests that in Figure 4. Release of Hsp60 from endothelial cells (ECs) and trapping of soluble Hsp60 on the surface of ECs. ECs were stressed by exposure to heat or were not stressed. A, Culture supernatants were evaluated for the presence of soluble Hsp60 by dot-blotting. The protein was revealed using anti-Hsp60 antibody, followed by horseradish peroxidase (HRP)–streptavidin antibody. Culture medium and recombinant Hsp60 were used as the negative and positive controls, respectively. B, ECs were incubated with (⫹) or without (–) recombinant Hsp60, and protein binding was evaluated by fluorescence-activated cell sorting using anti-Hsp60 monoclonal antibody, followed by fluorescein isothiocyanate (FITC)–conjugated antibody. Thin lines indicate fluorescence obtained with a control antibody. (Figure 3F) in the cytosol of most of saponin-treated ECs. Hsp60 was seen with the highest fluorescence intensity on the membrane after 5 hours of heat stress (Figures 3C and G). Quantitative RT-PCR was then used to quantify Hsp60 transcripts in freshly isolated and heat-stressed ECs. The number of Hsp60 mRNA copies was markedly increased after a 3-hour period of heat stress (Figure 3H). This rise in the number of transcripts preceded the intracytoplasmic expression of the protein. We further assessed whether Hsp60 was shed Figure 5. Induction of endothelial cell (EC) apoptosis by reactivity with anti-Hsp60. ECs were stressed by exposure to heat. A, ECs were incubated with various concentrations of anti-Hsp60 monoclonal antibody (mAb). After 24 hours of incubation, apoptosis was evaluated by fluorescence-activated cell sorting after staining with fluorescein isothiocyanate–conjugated annexin V and propidium iodide (PI). B, ECs were incubated with 8 g/ml of anti-Hsp60 mAb or with 320 g/ml of IgG purified from anti-Hsp60–containing anti-EC antibody– positive sera (samples 1–6 from patients with systemic lupus erythematosus), with (open bars) or without (solid bars) recombinant Hsp60. Apoptotic cells were the annexin V–positive cells among the PInegative EC population. The percentage of apoptosis was determined by subtracting the number of cells undergoing spontaneous apoptosis. C, The percentage inhibition of the apoptotic response was calculated as described in Patients and Methods. INDUCTION OF ENDOTHELIAL CELL APOPTOSIS BY ANTI-Hsp60 ANTIBODIES addition to proteins translocated to the membrane from the cytosol, other extracellular proteins bind to the membrane of ECs, possibly in an autocrine manner. Proapoptotic effect of anti-Hsp60 antibodies. We next examined the issue of whether the complex of AECAs and Hsp60 delivers antiapoptotic signals. ECs were heat stressed to induce Hsp60 expression and then incubated with an anti-Hsp60 mAb for 24 hours. ECs displayed a dose-dependent apoptotic response (Figure 5A). The number of ECs undergoing apoptosis increased above that obtained spontaneously in the culture medium, reaching a plateau at 8 g/ml of antibody. These data suggested that the binding of anti-Hsp60 to Hsp60 does indeed induce apoptosis. To confirm that apoptosis occurs in disease states, ECs were incubated with IgG from anti-Hsp60 antibody–containing AECA-positive sera from 6 patients with SLE. Such treatment triggered apoptosis in ECs, but the number of apoptotic ECs varied from patient to patient (Figure 5B). Interestingly, when purified IgG was preincubated with saturating concentrations of recombinant Hsp60, as determined in pilot experiments, the apoptotic effect was inhibited to various degrees in sera from the different patients studied (Figure 5B). More than 50% inhibition was observed in sera from patients 1–3, while there was as little as 20% inhibition in sera from patient 4 and negligible effects in sera from patients 5 and 6 (Figure 5C). Overall, our results confirm that in some patients with systemic autoimmune diseases, Hsp60 constitutes a target antigen for AECAs and that the binding of these autoantibodies can induce apoptosis in stressed ECs. DISCUSSION In this study, we sought to determine the level of Hsp60 reactivity of AECA-positive sera from patients with vasculitic systemic autoimmune diseases, compared with sera from patients with nonvasculitic systemic autoimmune diseases. Because AECAs are thought to be involved in the pathogenesis of such conditions, we also assessed the contribution of anti-Hsp60 antibody to the proapoptotic effects of AECAs. Furthermore, experiments were performed to gain insight into the mechanisms used by Hsp60 to become accessible to AECAs. Although the prevalence of AECAs may vary among rheumatic diseases (33), our findings (Table 1) are consistent with those of previous studies of the overall distribution of these autoantibodies. Irrespective of AECA activities, there is still controversy regarding the presence of anti-Hsp60 anti- 4035 bodies in SLE (13–16). Thus, a consensus on the prevalence of AECAs and Hsp60 antibodies in SLE and other systemic autoimmune diseases has not yet been reached. As usual, these discrepancies may be ascribed to variations in patient selection and differences in the tests used (33,34). Each of the 3 different approaches used in the current investigation, however, verified the findings in the other approaches and showed that recombinant Hsp60 reactivity exists in vasculitic and in nonvasculitic systemic autoimmune diseases. Using Western blotting with Hsp60, there were more positive sera in the former group of patients than the latter group. Consistent with our previous study (28), the prevalence was lower in the ELISA with Hsp60 analyses than in the Western blotting analyses, but still it predominated in patients with vasculitis. These observations reinforce our 2DE preliminary data indicating that Hsp60 recognition by AECAs was more often associated with vasculitic than with nonvasculitic systemic autoimmune diseases. This finding is consistent with the frequency of antiHsp60 reactivity in SLE, the prototypical vasculitic systemic autoimmune disease, compared with other systemic autoimmune diseases not associated with vasculitis. There is every likelihood that patients with higher reactivity for Hsp60 were those who had active disease at the time their blood was collected, but we could not establish any influence of disease activity in patients with SLE, WG, or PAN. We cannot exclude the possibility that different epitopes of Hsp60 would be targeted, depending on the disease severity, as has been elegantly demonstrated in atherosclerosis (35). However, serial sampling is warranted in order to address this specific question. While the patients’ drug regimens could also have an influence, Hsp60 reactivity did not appear to be affected by corticosteroids or other immunosuppressants. This remains to be confirmed, however. Due to its location within the cytoplasm, Hsp60 is not accessible to AECAs unless the cells are specifically activated. This contention was confirmed by FACS and IIF analyses showing that Hsp60 was not present on the membrane of freshly isolated nonpermeabilized ECs or in the cytosol of freshly isolated permeabilized ECs. However, when ECs were exposed to heat stress, they expressed Hsp60. Initially, the level of Hsp60 mRNA was markedly raised after 3 hours of heat-induced stress, which was followed 4 hours later by cytosolic detection of the protein and another 1 hour later by the translocation of the Hsp60 protein to the cell membrane. Inasmuch as this protein lacks the peptide signal sequence required for cell surface translocation, this transfer must use unconventional (and thus far unknown) 4036 pathways. Our observations are also consistent with those of previous experiments showing that ECs express Hsp60 on their surface under stress, both in vitro (8,36) and in vivo (32). Our findings support the notion that translocation of Hsp60 to the cell surface plays a role in the pathogenesis of vasculitis-associated systemic autoimmune diseases (8). Once expressed, Hsp60 can also be shed from stressed ECs. The increased level of soluble Hsp60 in the sera of patients with early cardiovascular disease (37) might thus result from its accelerated release from ECs. Also supporting our interpretation is the recent observation that levels of soluble Hsp60 correlate with markers of inflammation in atherosclerosis (36). It is highly relevant that SLE patients have been shown to be at risk of developing atherosclerosis (37), such that there has been a recent increase in their proportionate death from vascular disease, particularly from accelerated atherosclerosis (38). Interestingly, human and bacterial heatshock proteins have a 55% homology, and so, the immune response to a bacterial challenge may result in cross-reactivity by targeting Hsp60-expressing host ECs. This is particularly true in Mycobacterium infection (20). Moreover, we found that soluble Hsp60 could be taken up on the surface of ECs. The capacity of ECs to bind soluble Hsp60 was enhanced by stress. This newly identified property may affect pathophysiologic responses. In this setting, 3 important questions arise. First, what are the pathogenic consequences of the presence of Hsp60 on the surface of ECs? Second, which receptor(s) is triggered by soluble Hsp60 that would initiate its pathogenic effects? Third, do “autochthonous Hsp60” and passively bound Hsp60 differ in terms of their immunogenicity, and do these 2 kinds of Hsp60 trigger different signal transduction cascades? To gain further insight into the pathogenicity of AECAs directed toward Hsp60, the apoptotic fate of ECs was evaluated (24,25). All anti-Hsp60–positive sera tested induced apoptosis. Those experiments suggest that binding of anti-Hsp60 to its antigen on the EC surface is sufficient in itself to initiate programmed cell death. However, inhibition experiments in which purified IgG was preincubated with recombinant Hsp60 only partly prevented apoptosis, demonstrating that there is more than 1 candidate autoantibody among apoptosisinducing AECAs. This finding is at variance with that by Dieudé et al (13), who reported that AECAs from SLE sera specifically and exclusively recognized Hsp60. In our experience, this is really exceptional. The apoptotic pathways activated by the binding of AECAs to Hsp60 are unknown. Several mechanisms JAMIN ET AL may be involved in the process. When complexed with Bax in the cytosol, Hsp60 participates in an antiapoptotic complex. Stress factors dissociate this complex by translocating Hsp60 to the plasma membrane and directing Bax to the mitochondria (39). It follows that once apoptosis has been triggered, mitochondrial Hsp60 is released and that this, in turn, accelerates the activation of procaspase 3 (40). A role of surface expression of Hsp60 was also suspected based on the fact that levels of Hsp60 on the surface of apoptotic cells are increased (41). All of these observations converge to suggest that the abnormal expression of Hsp60 on the plasma membrane sensitizes ECs to apoptosis. Furthermore, our results support the view that direct recognition of Hsp60 on stressed ECs further promotes the process. A positive feedback regulation may thus be operant, in that stress factors favor the appearance of Hsp60 on the membrane, and Hsp60-associated AECAs trigger apoptosis, which results in the release of mitochondrial matrix Hsp60 into the cytosol (42). Owing to the stress, Hsp60 then moves to the membrane, where it becomes accessible to AECAs. Despite the results generated in this study, it is not yet possible to categorically assign Hsp60-induced apoptosis to antibodies specific for the translocated form of the protein or to antibodies specific for the soluble form passively bound to ECs. There are a number of reports suggesting that certain anti-Hsp60 antibodies display cross-reactivity with CD51, CD61, and connexin 45 on EC membranes (43). One might thus hypothesize that AECAs bridge plasma membrane Hsp60 with any of these molecules on ECs exposed to stress. Such complexes could be required to trigger the apoptotic signaling cascade. It is intriguing in this respect that Toll-like receptor 4 (TLR-4) and TLR-2 mRNA were shown to be induced by heat stress (Jamin C, et al: unpublished observations, and Wick G: personal communication). TLR-4 is involved in the signaling of soluble Hsp60 (44), and TLR-2 is involved in the activation of the apoptotic pathway (45). After exposure to stress, TLR-4 is likely to be up-regulated, leading to its passive binding to soluble Hsp60. This would promote the expression of TLR-2 (46). The Hsp70 ligand for TLR-2 (47) would then stimulate TLR-2 and enhance apoptosis (48). Alternatively, a direct recognition of soluble Hsp60 by AECAs, when bound to EC membranes, might activate the multimolecular complex and trigger apoptosis. Taken together, the results of our experiments establish that the complex mixture of AECAs contain anti-Hsp60 antibodies. The higher frequency of anti- INDUCTION OF ENDOTHELIAL CELL APOPTOSIS BY ANTI-Hsp60 ANTIBODIES Hsp60 antibodies in vasculitic than in nonvasculitic systemic autoimmune diseases indicates that they play a deleterious role in the pathophysiology of vasculitis. The capacity of AECA-positive sera from SLE patients to induceapoptosisofECsstrengthensthisproposedhypothesis. Overall, Hsp60 seems to be a key antigen involved in vessel damage during inflammatory responses in the presence of AECAs. ACKNOWLEDGMENTS We thank Dr. Cora-Jean S. Edgell (University of North Carolina, Chapel Hill, NC) for providing the EA.hy926 cells, Professors Rizgar Mageed (St. Bartholomew’s and the London Queen Mary School of Medicine and Dentistry, London, UK) and Georg Wick (Innsbruck Medical University, Innsbruck, Austria) for insightful review of the manuscript, and Simone Forest and Cindy Séné for secretarial assistance. REFERENCES 1. Zugel U, Kaufmann SH. Role of heat shock proteins in protection from and pathogenesis of infectious diseases. Clin Microbiol Rev 1999;12:19–39. 2. Soltys BJ, Gupta RS. Immunoelectron microscope localization of the 60-kDa heat chaperonin protein in mammalian cells. Exp Cell Res 1996;222:16–27. 3. Freedman MS, Buu NN, Ruijs TC, Williams K, Antel JP. Differential expression of heat shock proteins by human glial cells. J Neuroimmunol 1992;41:231–8. 4. Wand-Wurttenberger A, Schoel B, Ivanyi J, Kaufmann SH. Surface expression by mononuclear phagocytes of an epitope shared with mycobacterial heat shock protein 60. Eur J Immunol 1991; 21:1089–92. 5. Kaur I, Voss SD, Gupta RS, Schell K, Fisch P, Sondel PM. Human peripheral ␥␦ T cells recognize hsp60 molecules on Daudi Burkitt’s lymphoma cells. J Immunol 1993;150:2046–55. 6. Vendetti S, Cicconi R, Piselli P, Vismara D, Cassol M, Delpino A. Induction and membrane expression of heat shock proteins in heat-treated HPC-4 cells is correlated with increased resistance to LAK-mediated lysis. J Exp Clin Cancer Res 2000;19:329–34. 7. Dziewanowska K, Carson AR, Patti JM, Deobald CF, Bayles KW, Bohach GA. Staphylococcal fibronectin binding protein interacts with heat shock protein 60 and integrins: role in internalization by epithelial cells. Infect Immun 2000;68:6321–8. 8. Wick G, Knoflach M, Xu Q. Autoimmune and inflammatory mechanisms in atherosclerosis. Annu Rev Immunol 2004;22: 361–403. 9. Seitz CS, Kleindienst R, Xu Q, Wick G. Coexpression of heatshock protein 60 and intercellular-adhesion molecule-1 is related to increased adhesion of monocytes and T cells to aortic endothelium of rats in response to endotoxin. Lab Invest 1996;74:241–52. 10. Soltys BJ, Gupta RS. Cell surface localization of the 60-kDa heat shock chaperonin protein on mammalian cells. Cell Biol Int 1997;21:315–20. 11. Young D, Lathigra R, Hendrix R, Sweetser D, Young RA. Stress proteins are immune targets in leprosy and tuberculosis. Proc Natl Acad Sci U S A 1988;85:4267–70. 12. Rudolphi U, Rzepka R, Batsford S, Kaufmann SH, von der Mark K, Peter HH, et al. The B cell repertoire of patients with rheumatoid arthritis. II. Increased frequencies of IgG⫹ and IgA⫹ B cells specific for mycobacterial heat-shock protein 60 or human 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 4037 type II collagen in synovial fluid and tissue. Arthritis Rheum 1997;40:1409–19. Dieude M, Senecal JL, Raymond Y. Induction of endothelial cell apoptosis by heat-shock protein 60–reactive antibodies from anti–endothelial cell autoantibody–positive systemic lupus erythematosus patients. Arthritis Rheum 2004;50:3221–31. De Graeff-Meeder ER, Rijkers GT, Voorhorst-Ogink MM, Kuis W, van der Zee R, van Eden W, et al. Antibodies to human HSP60 in patients with juvenile chronic arthritis, diabetes mellitus, and cystic fibrosis. Pediatr Res 1993;34:424–8. Schett G, Xu Q, Amberger A, van der Zee R, Recheis H, Willeit J, et al. Autoantibodies against heat shock protein 60 mediate endothelial cytotoxicity. J Clin Invest 1995;96:2569–77. Jarjour WN, Jeffries BD, Davis JS IV, Welch WJ, Mimura T, Winfield JB. Autoantibodies to human stress proteins: a survey of various rheumatic and other inflammatory diseases. Arthritis Rheum 1991;34:1133–8. D’Cruz DP, Houssiau FA, Ramirez G, Berguley E, McCutcheon J, Vianna J, et al. Antibodies to endothelial cells in systemic lupus erythematosus: a potential marker for nephritis and vasculitis. Clin Exp Immunol 1991;85:254–61. Salojin KV, le Tonqueze M, Nassonov EL, Blouch MT, Baranov AA, Saraux A, et al. Anti-endothelial cell antibodies in patients with various forms of vasculitis. Clin Exp Rheumatol 1996;14: 163–9. Salojin KV, le Tonqueze M, Saraux A, Dueymes M, Nassonov EL, Piette JC, et al. Anti-endothelial cell antibodies and markers of endothelial cell damage: clinical associations in systemic sclerosis and Raynaud’s phenomenon. Am J Med 1997;102:178–85. Dugue C, Perraut R, Youinou P, Renaudineau Y. Effects of anti-endothelial cell antibodies in leprosy and malaria. Infect Immun 2004;72:301–9. Cacoub P, Ghillani P, Revelen R, Thibault V, Calvez V, Charlotte F, et al. Anti-endothelial cell autoantibodies in hepatitis C virus mixed cryoglobulinemia. J Hepatol 1999;31:598–603. Toyoda M, Galfayan K, Galera OA, Petrosian A, Czer LS, Jordan SC. Cytomegalovirus infection induces anti-endothelial cell antibodies in cardiac and renal allograft recipients. Transpl Immunol 1997;5:104–11. Frostegard J, Wu R, Gillis-Haegerstrand C, Lemne C, de Faire U. Antibodies to endothelial cells in borderline hypertension. Circulation 1998;98:1092–8. Bordron A, Dueymes M, Levy Y, Jamin C, Leroy JP, Piette JC, et al. The binding of some human antiendothelial cell antibodies induces endothelial cell apoptosis. J Clin Invest 1998;101:2029–35. Worda M, Sgonc R, Dietrich H, Niederegger H, Sundick RS, Gershwin ME, et al. In vivo analysis of the apoptosis-inducing effect of anti–endothelial cell antibodies in systemic sclerosis by the chorionallantoic membrane assay. Arthritis Rheum 2003;48: 2605–14. Turesson C, McClelland RL, Christianson TJ, Matteson EL. No decrease over time in the incidence of vasculitis or other extraarticular manifestations in rheumatoid arthritis: results from a community-based study. Arthritis Rheum 2004;50:3729–31. Youinou P, Pennec YL, Katsikis P, Jouquan J, Fauquert P, le Goff P. Raynaud’s phenomenon in primary Sjögren’s syndrome. Br J Rheumatol 1990;29:205–7. Revelen R, d’Arbonneau F, Guillevin L, Bordron A, Youinou P, Dueymes M. Comparison of cell-ELISA, flow cytometry and Western blotting for the detection of antiendothelial cell antibodies. Clin Exp Rheumatol 2002;20:19–26. Revelen R, Bordron A, Dueymes M, Youinou P, Arvieux J. False positivity in a cyto-ELISA for anti-endothelial cell antibodies caused by heterophile antibodies to bovine serum proteins. Clin Chem 2000;46:273–8. Sgonc R, Gruschwitz MS, Boeck G, Sepp N, Gruber J, Wick G. Endothelial cell apoptosis in systemic sclerosis induced by anti- 4038 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. body-dependent cell-mediated cytotoxicity via CD95. Arthritis Rheum 2000;43:2550–62. Amberger A, Maczek C, Jurgens G, Michaelis D, Schett G, Trieb K, et al. Co-expression of ICAM-1, VCAM-1, ELAM-1 and Hsp60 in human arterial and venous endothelial cells in response to cytokines and oxidized low-density proteins. Cell Stress Chaperones 1997;2:94–103. Hochleitner BW, Hochleitner EO, Obrist P, Eberl T, Amberger A, Xu Q, et al. Fluid shear stress induces heat shock protein 60 expression in endothelial cells in vitro and in vivo. Arterioscler Thromb Vasc Biol 2000;20:617–23. Meroni PL, Youinou P. Endothelial cell antibodies. In: Peter JB, Shoenfeld Y, editors. Autoantibodies. Amsterdam: Elsevier; 1996. p. 245–52. Youinou P, Meroni PL, Khamashta MA, Shoenfeld Y. A need for standardization of the antiendothelial cell antibody test. Immunol Today 1995;91:363–4. Perschinka H, Mayr M, Millonig G, Mayerl C, van der Zee R, Morrison SG, et al. Cross-reactive B-cell epitopes of microbial and human heat-shock protein 60/65 in atherosclerosis. Arterioscler Thromb Vasc Biol 2003;23:1060–5. Xu Q, Schett G, Perschinka H, Mayr M, Egger G, Oberhollenzer F, et al. Serum soluble heat shock protein 60 is elevated in subjects with atherosclerosis in a general population. Circulation 2000;102:14–20. Pockley AG, Wu R, Lemne C, Kiessling R, de Faire U, Frostegard J. Circulating heat shock protein 60 is associated with early cardiovascular disease. Hypertension 2000;36:303–7. Borchers AT, Keen CL, Shoenfeld Y, Gershwin ME. Surviving the butterfly and the wolf: mortality trends in systemic lupus erythematosus. Autoimmun Rev 2004;3:423–53. Gupta S, Knowlton AA. Cytosolic heat shock protein 60, hypoxia, and apoptosis. Circulation 2002;106:2727–33. Samali A, Cai J, Zhivotovsky B, Jones DP, Orrenius S. Presence of JAMIN ET AL 41. 42. 43. 44. 45. 46. 47. 48. a pre-apoptotic complex of pro-caspase-3, Hsp60 and Hsp10 in the mitochondrial fraction of Jurkat cells. EMBO J 1999;18:2040–8. Sapozhnikov AM, Ponomarev ED, Tarasenko TN, Telford WG. Spontaneous apoptosis and expression of cell surface heat-shock proteins in cultured EL-4 lymphoma cells. Cell Prolif 1999;32: 363–78. Chandra D, Tang DG. Mitochondrially localized active caspase-9 and caspase-3 result mostly from translocation from the cytosol and partly from caspase-mediated activation in the organelle: lack of evidence for Apaf-1-mediated procaspase-9 activation in the mitochondria. J Biol Chem 2003;278:17408–20. Bason C, Corrocher R, Lunardi C, Puccetti P, Olivieri O, Girelli D, et al. Interaction of antibodies against cytomegalovirus with heat-shock protein 60 in pathogenesis of atherosclerosis. Lancet 2003;362:1971–7. Bulut Y, Faure E, Thomas L, Karahashi H, Michelsen KS, Equils O, et al. Chlamydial heat shock protein 60 activates macrophages and endothelial cells through Toll-like receptor 4 and MD2 in a MyD88-dependent pathway. J Immunol 2002;168:1435–40. Aliprantis AO, Yang RB, Weiss DS, Godowski P, Zychlinsky A. The apoptotic signaling pathway activated by Toll-like receptor-2. EMBO J 2000;19:3325–36. Fan J, Frey RS, Malik AB. TLR4 signaling induces TLR2 expression in endothelial cells via neutrophil NADPH oxidase. J Clin Invest 2003;112:1234–43. Wagner M, Hermanns I, Bittinger F, Kirkpatrick CJ. Induction of stress proteins in human endothelial cells by heavy metal ions and heat shock. Am J Physiol 1999;277:1026–33. Asea A, Rehli M, Kabingu E, Boch JA, Bare O, Auron PE, et al. Novel signal transduction pathway utilized by extracellular HSP70: role of toll-like receptor (TLR) 2 and TLR4. J Biol Chem 2002;277:15028–34.