SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Objective

Systemic sclerosis (SSc) is an autoimmune disease characterized by fibrosis due to excessive and dysregulated collagen production by fibroblasts. Previously, we reported that anti–DNA topoisomerase I (anti–topo I) antibodies bound specifically to fibroblast surfaces; however, we had not identified their antigenic target. We undertook this study to characterize the target of anti–topo I antibodies on fibroblasts and the effects of their binding.

Methods

Purified topo I or topo I released from apoptotic cells was tested for surface binding to a number of human cell types by cell-based enzyme-linked immunosorbent assay, flow cytometry, and indirect immunofluorescence. Antibodies purified from SSc patient and normal control sera were used to detect topo I binding. The consequences of topo I and anti–topo I binding to fibroblasts were assessed by coculture with THP-1 monocytes.

Results

The autoantigen topo I itself was found to bind specifically to fibroblasts in a dose-dependent and saturable manner, where it was recognized by anti–topo I from SSc patients. The binding of anti–topo I subsequently stimulated adhesion and activation of cocultured monocytes. Topo I released from apoptotic endothelial cells was also found to bind specifically to fibroblasts.

Conclusion

The findings of this study thus confirm and extend the findings of our previous study by showing that topo I binding to fibroblast surfaces is both necessary and sufficient for anti–topo I binding. Second, topo I–anti–topo I complex binding can then trigger the adhesion and activation of monocytes, thus providing a plausible model for the amplification of the fibrogenic cascade in anti–topo I–positive SSc patients.

Systemic sclerosis (SSc), or scleroderma, is an autoimmune disease characterized by excessive extracellular matrix (ECM) deposition, fibrosis, and vascular obliteration in connective tissues of the skin, lungs, gastrointestinal tract, heart, and kidneys (1). Following a still-undefined initial event, mononuclear cells such as lymphocytes and monocytes are recruited into connective tissues, where the subsequent release of inflammatory cytokines leads to fibroblast activation, collagen accumulation, and tissue hypoxia (2).

One of the hallmarks of SSc is the presence, in >90% of patient sera, of highly specific autoantibodies against a variety of nuclear proteins (i.e., antinuclear autoantibodies [ANAs]) (3). The most common ANAs associated with SSc include anti–DNA topoisomerase I (anti–topo I), anti–centromeric protein B (anti–CENP-B), anti–RNA polymerase III, and anti-Th/To autoantibodies. ANAs have been associated with different clinical manifestations and various degrees of SSc severity. For example, anti–topo I are usually associated with more diffuse cutaneous SSc (4, 5), whereas anti–CENP-B are detected predominantly in patients with limited cutaneous SSc (4, 6, 7). Also, anti–topo I are strongly associated with pulmonary fibrosis in SSc (6, 8), while pulmonary arterial hypertension is more common in patients with anti–CENP-B (9). In spite of the prevalence of ANAs and their association with different SSc subsets and selective visceral involvement, their direct or indirect role in the pathophysiology of SSc is still unclear.

Recent in vitro studies have demonstrated that some SSc autoantibodies directed against cell surfaces could display direct pathogenic effects. For example, anti–endothelial cell antibodies (AECAs), which are present in 25–85% of SSc patients (10), have been shown to trigger antibody-dependent cellular cytotoxicity in the presence of leukocytes (11, 12), while antifibroblast antibodies (AFAs), which are present in 26–58% of SSc patients (13, 14), have been shown to induce a proadhesive and proinflammatory phenotype in fibroblasts upon binding (13, 15). Endothelial cell apoptosis and dysregulated fibroblast activation are cornerstones of SSc pathophysiology, the former being one of the earliest events observed during the disease (16) and the latter being directly responsible for the increased and continuous deposition of ECM, notably type I collagen, in tissues (17–19). Although there is now evidence that AECAs and AFAs are directly implicated in the pathophysiology of SSc, no such evidence yet exists for ANAs.

In a previous study, we demonstrated that anti–topo I from SSc patients could bind specifically to the cell surface of fibroblasts via an unknown surface target (14). The aims of the present study were therefore 1) to identify the fibroblast surface target of anti–topo I and 2) to characterize the consequences of anti–topo I binding to fibroblasts on the mobilization and activation of monocytes.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Patients and cell lines.

Thirty-six SSc patients (12 anti–topo I negative, 24 anti–topo I positive) were selected randomly from a French Canadian cohort of 309 SSc patients diagnosed between April 1984 and September 1999 at the Connective Tissue Diseases and Vascular Medicine Clinics of Notre-Dame Hospital, Centre Hospitalier de l'Université de Montréal, in Montreal (20, 21). All patients fulfilled the American College of Rheumatology (formerly, the American Rheumatism Association) classification criteria for SSc.

Sera were obtained at the time of SSc diagnosis (i.e., prior to treatment), coded, and stored at −80°C. Control sera were collected as previously described from age- and sex-matched French Canadian volunteers (22). Primary human dermal fibroblasts were isolated from biopsy samples obtained from lesional or nonlesional skin of patients with SSc or from age- and sex-matched normal adult skin as previously described (23). Fibroblasts were cultured in Dulbecco's modified Eagle's medium supplemented with 10% inactivated fetal bovine serum (FBS) and 100 μg/ml of gentamicin sulfate (Wisent, St. Bruno, Quebec, Canada). Primary fibroblasts were used between passages 3 and 7.

The following primary cell populations and their respective media were from Cambrex (Walkersville, MD). Normal human lung fibroblasts (NHLFs) as well as normal human dermal fibroblasts from adults (NHDF-Ad) and neonates (NHDF-Neo) were cultured in fibroblast basal medium with 2% FBS, 5 μg/ml of bovine insulin, 1 ng/ml of human fibroblast growth factor B (hFGF-B), 0.1% gentamicin sulfate, and 0.1% amphotericin B. Human umbilical vein endothelial cells (HUVECs) as well as human microvascular endothelial cells from dermis (HMVEC-D) and lung (HMVEC-L) were grown in endothelial basal medium with basal medium for microvascular endothelial cells (5% FBS, 0.1% insulin-like growth factor 1, 0.1% ascorbic acid, 0.4% hFGF-B, 0.1% human epidermal growth factor [hEGF], 0.04% hydrocortisone, 0.1% human vascular endothelial growth factor, and 0.1% gentamicin). Human pulmonary artery smooth muscle cells from adults (PASMC-Ad) and neonates (PASMC-Neo) were grown in smooth muscle basal medium to which was added the SmGM-2 bullet kit (5% FBS, 0.1% hEGF, 0.1% insulin, 0.2% hFGF-B, and 0.1% gentamicin; Cambrex). All primary cells were used between passages 3 and 5. Subculturing was achieved before confluence using trypsin–EDTA and trypsin neutralizing solution (Cambrex).

THP-1 monocytes (American Type Culture Collection, Manassas, VA) were grown in RPMI 1640 supplemented with 10% inactivated FBS and 100 μg/ml of gentamicin sulfate. Cells were grown at 37°C in air with 5% CO2.

Endothelial cell apoptosis.

HUVEC and HMVEC-D apoptosis was induced by growth factor deprivation as previously described (24, 25). Briefly, cells were grown in serum-free medium for 48–72 hours. Apoptosis was confirmed by immunofluorescence with Hoechst/propidium iodide (PI) staining (24, 25) and flow cytometry analysis of phosphatidylserine exposure on cell surfaces using fluorescein isothiocyanate (FITC)–conjugated annexin V (BD Biosciences, Bedford, MA). Supernatants from several apoptotic HUVEC or HMVEC-D cultures were pooled. Dead cells and large apoptotic bodies were sedimented at 20,000g for 15 minutes at 4°C, and pooled supernatants were concentrated 20 times using protein concentrators with a 5-kd molecular weight cutoff (VivaScience, Hannover, Germany). Supernatant proteins were quantitated using the Bradford Protein Assay (Bio-Rad, Hercules, CA), following the manufacturer's instructions.

Topo I.

Purified topo I from calf and/or rabbit thymus was obtained from Immunovision (Springdale, AR) and tested for purity upon arrival by gel electrophoresis and immunoblotting. Topo I major molecular species had a molecular weight of 70 kd; minor species had molecular weights of 80 kd and 100 kd.

Antibody purification.

Human IgG were purified from sera by affinity chromatography using the NAb Protein G Spin Chromatography kit (Pierce, Rockford, IL), following the manufacturer's instructions. The final IgG concentration was determined by the Bradford dye-binding procedure (Bio-Rad) and varied from 2.0 to 11.9 μg/μl (14). Human anti–topo I were purified from SSc sera by affinity chromatography on immobilized topo I using the Vivapure Epoxy Protein Coupling kit (VivaScience), following the manufacturer's instructions, and were then transferred into phosphate buffered saline (PBS) (14). Final anti–topo I IgG concentrations were determined using the Easy-Titer Human IgG Assay kit (Pierce) and varied from 2.1 to 8.3 ng/μl.

Flow cytometry.

Flow cytometry analyses were performed as previously described (14). Briefly, adherent cells were detached with PBS/0.5% EDTA and washed with PBS. Cells were first incubated with 1 μg/ml goat IgG (Sigma, Oakville, Ontario, Canada) in PBS/3% bovine serum albumin (BSA) to block nonspecific binding sites. Cells were then incubated with topo I (0.5 μg/ml) in PBS/3% BSA for 30 minutes, washed with PBS, and incubated with IgG (50 μg/ml) or anti–topo I (400 ng/ml) purified from SSc or normal control sera in PBS/3% BSA for 30 minutes. IgG binding was revealed with phycoerythrin (PE)–conjugated goat anti-human IgG (Jackson ImmunoResearch, West Grove, PA) diluted 1:100. Cell permeability was assessed by addition of 7-aminoactinomycin D (Sigma), and permeable cells were gated out. Fluorescence was detected on a FACScan and analyzed by CellQuest software (BD Biosciences).

Cell-based enzyme-linked immunosorbent assay (ELISA).

Cells were grown in collagen-coated 96-well culture microplates (Corning, Acton, MA) until confluence, washed with PBS, and incubated with 1 μg/ml goat IgG in complete medium. Cells were then incubated for 30 minutes with topo I in complete medium, washed with PBS, and incubated for 1 hour with a mouse anti–topo I monoclonal antibody (Sigma) diluted 1:100, IgG (50 μg/ml) from SSc or normal sera, or anti–topo I (400 ng/ml) purified from SSc sera in complete medium. Antibody binding was revealed with horseradish peroxidase (HRP)–conjugated goat anti-human or anti-mouse IgG (Jackson ImmunoResearch) and o-phenylenediamine/citrate solution. The reaction was stopped by 2M H2SO4, and the optical density (OD) at 490 nm was read in an MRX Revelation microplate reader (Dynex, Chantilly, VA).

Indirect immunofluorescence and confocal microscopy.

Indirect immunofluorescence was performed as previously described (14). Briefly, cells were grown to 70–80% confluence on glass coverslips (Fisher Canada, Ottawa, Ontario, Canada), washed with PBS, and incubated with 1 μg/ml goat IgG in complete medium. Cells were incubated for 30 minutes with topo I (0.5 μg/ml) or for 1 hour with supernatant from apoptotic endothelial cells, washed with PBS, and incubated for 1 hour with a mouse anti–topo I monoclonal antibody diluted 1:50, IgG (50 μg/ml) from SSc or normal sera, or anti–topo I (400 ng/ml) purified from SSc sera in complete medium. Cells were then washed with PBS and fixed with methanol for 5 minutes at −20°C. IgG binding was detected with FITC-conjugated goat anti-human or anti-mouse IgG (Jackson ImmunoResearch). Hoechst 33342 was used to stain nuclei. Cells were examined with an Eclipse E600 fluorescence microscope (Nikon, Melville, NY) using MetaMorph 4.6r9 software (Universal Imaging, Downingtown, PA) or with a 510 confocal laser microscope (Zeiss, Thornwood, NY), as previously described (14).

Electrophoresis and immunoblotting.

Proteins were diluted in electrophoresis sample buffer (25 mM Tris [pH 6.8], 10% glycerol, 1% sodium dodecyl sulfate, 2% β-mercaptoethanol, and 0.25% bromphenol blue). Proteins were separated on 8% polyacrylamide gels and transferred onto nitrocellulose membranes (Bio-Rad). Membranes were preincubated in blocking buffer (Tris buffered saline–0.5% Tween 20/5% powdered skim milk) and incubated with a mouse IgM anti–topo I monoclonal antibody (BD Biosciences) diluted 1:100 or with IgG (50 μg/ml) from SSc sera. IgM and IgG binding were detected using an HRP-conjugated goat anti-mouse IgM (Jackson ImmunoResearch) and an HRP-conjugated goat anti-human IgG, respectively, and the Super- Signal West Pico chemiluminescence kit (Pierce).

THP-1 monocyte adhesion assay.

Fibroblasts (NHDFAd or NHLFs) were grown in 96-well culture microplates (Corning) until confluence, incubated with topo I (0.5 μg/ml) in complete medium for 30 minutes, washed with PBS, and incubated for 1 hour with IgG (50 μg/ml) or anti–topo I (400 ng/ml) purified from SSc or normal sera in complete medium. Fibroblasts were then washed with PBS. THP-1 monocytes preincubated with bromodeoxyuridine (BrdU) (1% volume/volume) for 3 hours were added to fibroblasts at a concentration of 2.5 × 105/well in complete RPMI medium. THP-1 monocytes and fibroblasts were cocultured for 30 minutes before removal of nonadhered THP-1 monocytes by thorough washing with PBS. THP-1 monocyte adhesion was assessed by anti-BrdU ELISA (Roche, Laval, Quebec, Canada), following the manufacturer's instructions. The reaction was stopped by 2M H2SO4, and the OD at 450 nm was read in an MRX Revelation microplate reader.

THP-1 monocyte activation assay.

Fibroblasts (NHDFAd or NHLFs) were grown in 6-well culture microplates (Corning) until confluence, incubated with topo I (0.5 μg/ml) in complete medium for 30 minutes, washed with PBS, and incubated with IgG (50 μg/ml) or anti–topo I (400 ng/ml) purified from SSc or normal sera in complete medium for 1 hour. Fibroblasts were then washed with PBS, and THP-1 monocytes were added at a concentration of 5.0 × 105/well in complete RPMI medium. After 3 days of coculture, THP-1 monocytes were detached with 0.02% EDTA in cold PBS, and CD14 and CD11b expression was analyzed by flow cytometry using FITC-conjugated anti-CD14 and PE-conjugated anti-CD11b antibodies (Jackson ImmunoResearch). Permeable cells were gated out, and fluorescence was detected as previously described (14).

Preliminary characterization of topo I ligand.

Fibroblasts (NHLFs) were detached with PBS/0.5% EDTA or trypsin–EDTA and washed with PBS. Trypsin–EDTA–treated cells were allowed to recover in complete medium for 0, 10, 20, or 30 minutes at 37°C on a low-adherence plastic substrate. Brefeldin A effects on topo I binding were assessed by pretreating fibroblasts with brefeldin A (Sigma) for 2 hours prior to trypsin–EDTA detachment. Fibroblasts were then allowed to recuperate in complete medium for 20 minutes as above. Topo I (0.5 μg/ml) binding to these fibroblasts was then analyzed by flow cytometry as described above.

Statistical analysis.

Kruskal-Wallis nonparametric test followed by Dunn's multiple comparison test was used for group comparisons after assays for topo I binding to different cell lines and THP-1 monocyte adhesion. Mann-Whitney nonparametric test was used for group comparisons after the THP-1 monocyte activation assay. Statistical tests were performed with GraphPad Prism 4.0 (GraphPad Software, San Diego, CA). P values less than or equal to 0.05 were considered significant.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Topo I binds to the fibroblast surface and recruits anti–topo I.

In a previous study, we found that anti–topo I from SSc patients bound specifically to the cell surface of fibroblasts (14). Anti–topo I binding had been assayed by flow cytometry and indirect immunofluorescence on cells detached from the substratum, a procedure that inevitably damages some cells and provokes leakage of their contents into the medium. When we assessed anti–topo I binding to adherent fibroblasts (i.e., on cells that had not been detached from the substratum), we did not detect any binding (Figure 1A), thus indicating that the antigenic target of anti–topo I was missing under these conditions. Hence, we hypothesized that endogenous topo I released from damaged cells had bound directly to fibroblasts and had become the target of anti–topo I.

thumbnail image

Figure 1. A, Detection of topoisomerase I (topo I) binding by indirect immunofluorescence. Normal human dermal fibroblasts from adults (NHDF-Ad) and human microvascular endothelial cells from the dermis (HMVEC-D) were incubated with 0.5 μg/ml topo I followed by IgG purified from an anti–topo I–positive systemic sclerosis (SSc) serum and a fluorescein isothiocyanate–conjugated second antibody to human IgG (green). Hoechst 33342 was used to detect nuclei (blue). Topo I binding was detected on fibroblast, but not endothelial, surfaces. No antibody binding was observed in the absence of topo I. Bar = 25 μm. B, Detection of topo I in supernatant of cells treated with phosphate buffered saline (PBS)/0.5% EDTA (Sup. EDTA). HMVEC-D were treated with PBS/0.5% EDTA, and cells were centrifuged at 400g for 5 minutes. Supernatants were recovered, and topo I was detected by immunoblotting with purified IgG from an anti–topo I–positive SSc serum. Lysate from untreated HMVEC-D (Lysate) and affinity-purified topo I (Topo I) were used as controls. Similar results were obtained for PBS/0.5% EDTA–treated NHDF-Ad (data not shown). C, Flow cytometry analysis of topo I binding to NHDF-Ad and HMVEC-D. Topo I binding was detected using anti–topo I antibodies purified from an anti–topo I–positive SSc serum and a second phycoerythrin–conjugated antibody to human IgG γ-chain. Data are presented as intact cell counts versus fluorescence intensity. Permeable cells have been gated out after vital staining with 7-aminoactinomycin D, as described in Materials and Methods. Antibody binding to the surface of cells incubated with 0.5 μg/ml topo I (solid line) or without topo I (dotted line) was measured. Cells incubated with secondary antibody alone (shaded area) were used as controls. Anti–topo I bound only to fibroblasts and not to endothelial cells. The binding intensity was increased after incubation with exogenous topo I. Results are representative of 3 independent experiments.

Download figure to PowerPoint

As a direct test of this hypothesis, the presence of topo I in the supernatant of EDTA-detached cells was assessed by immunoblotting. Topo I was found to be present in the supernatant of EDTA-detached cells (Figure 1B). Moreover, flow cytometry analysis showed that the addition of exogenous topo I (0.5 μg/ml) to EDTA-detached NHDF-Ad significantly increased anti–topo I binding (i.e., over and above the binding of topo I released from damaged cells) (Figure 1C). Anti–topo I did not bind to HMVEC-D, whether or not exogenous topo I was present (Figure 1C). Topo I binding to cells was detected with IgG purified from all SSc anti–topo I–positive sera tested (n = 24) as well as with a mouse anti–topo I monoclonal antibody. None of the IgG from SSc anti–topo I–negative sera (n = 12) or normal sera (n = 4) bound to NHDF-Ad or to NHLFs, whether topo I was added or not (data not shown).

Exogenous topo I was also found to bind to live, adherent, unfixed, and unpermeabilized fibroblasts. Exogenous topo I (0.5 μg/ml) was added to live cells, followed by the addition of IgG or anti–topo I purified from an SSc anti–topo I–positive serum. Topo I binding was detected on NHDF-Ad surfaces but not on HMVEC-D surfaces (Figure 1A). Similar results were obtained when NHLFs and HUVECs were assayed under the same conditions (data not shown). Since fluorescence patterns and intensities on NHDF-Ad were the same with both IgG and anti–topo I purified from an SSc anti–topo I–positive serum, we concluded that anti–topo I binding was due to the initial binding of topo I to the fibroblast surface. These results clearly indicate that the topo I bound to the fibroblast surface was the target detected by anti–topo I in our previous work (14).

Dose-dependent and saturable binding of topo I on fibroblasts.

To characterize topo I binding to fibroblasts, we performed cell-based ELISA experiments in which increasing concentrations of topo I (up to 5 μg/ml) were added to adherent and intact NHDF-Ad and HMVEC-D (Figure 2). We found that topo I binding to cell surfaces was dose dependent and saturable. Topo I binding was significantly higher on NHDF-Ad than on HMVEC-D at all concentrations tested (P ≤ 0.05). Similar results were obtained when topo I binding on NHLFs was compared with that on HUVECs (data not shown). Taken together, these results suggest that topo I bound to cells via a specific surface determinant whose expression and/or topo I affinity was increased on fibroblasts compared with endothelial cells.

thumbnail image

Figure 2. Dose-response curves of topo I binding to fibroblasts and endothelial cells by cell-based enzyme-linked immunosorbent assay. Live NHDF-Ad and HMVEC-D were incubated with increasing concentrations of topo I. Topo I binding was detected with anti–topo I purified from SSc sera and with a horseradish peroxidase–conjugated second antibody to human IgG. Data are presented as the mean ± SEM optical density (OD) from pooled triplicate experiments. Topo I binding to NHDF-Ad was significantly higher than binding to HMVEC-D at all concentrations tested (∗ = P ≤ 0.05). See Figure 1 for other definitions.

Download figure to PowerPoint

Specific binding of topo I to fibroblasts.

To further assess the cellular specificity of topo I binding, we performed cell-based ELISA experiments in which topo I (0.5 μg/ml) was added to adherent and intact fibroblasts (NHLFs, NHDF-Ad, and NHDF-Neo), endothelial cells (HUVECs, HMVEC-D, and HMVEC-L), or smooth muscle cells (PASMC-Ad and PASMC-Neo) (Figure 3A). Topo I binding was detected with anti–topo I purified from SSc sera and an HRP-conjugated anti-human IgG. For convenience, NHLFs were used as an interplate control, and the ratio of the binding intensity on each cell line to that on NHLFs was determined. We found that all fibroblast cell lines tested bound topo I with an intensity similar to that of NHLFs, while binding to smooth muscle cells and endothelial cells was significantly lower (P ≤ 0.01 and P ≤ 0.001, respectively). Similar differences were observed at topo I concentrations of 1.0, 2.5, and 5.0 μg/ml (data not shown). These results confirm that fibroblasts displayed a higher affinity for topo I than for other cell types.

thumbnail image

Figure 3. Cell-based enzyme-linked immunosorbent assay analysis of topo I binding to different human cell types (A) and fibroblasts from SSc patients (B). Live cells were incubated with topo I (0.5 μg/ml), and binding was detected with anti–topo I purified from SSc sera and with a horseradish peroxidase–conjugated second antibody to human IgG. Data are presented as the mean and SD ratio of topo I binding to cell lines:topo I binding to normal human lung fibroblasts (NHLFs) (n = 4). Topo I binding to fibroblast cell lines was significantly higher than to smooth muscle cell lines (∗ = P ≤ 0.01) and endothelial cell lines (∗∗ = P ≤ 0.001). There was no significant difference between topo I binding to patient fibroblasts (SSc1, SSc2) and fibroblasts from normal controls (N1, N2). NHDF-Neo = NHDFs from neonates; PASMC-Neo = human pulmonary artery smooth muscle cells from neonates; PASMC-Ad = PASMCs from adults; HUVECs = human umbilical vein endothelial cells; HMVEC-L = HMVECs from lung; L = lesional skin; N-L = nonlesional skin (see Figure 1 for other definitions).

Download figure to PowerPoint

Binding of topo I to fibroblasts from SSc patients.

Compared with normal fibroblasts, SSc fibroblasts have primary metabolic abnormalities and altered responses to environmental signals (18). Therefore, there could be differences between normal and SSc fibroblasts with respect to affinity for topo I. To address this issue, we quantitated topo I binding to primary fibroblasts from SSc patient (n = 2) and normal control (n = 2) biopsy samples (Figure 3B). For each SSc patient, dermal fibroblasts derived from both lesional and nonlesional skin were tested. NHLFs served as an interplate control, as described above. There was no significant difference between the relative binding of topo I to normal and SSc fibroblasts (Figure 3B). Again, the origin of fibroblasts (lesional or nonlesional skin) did not influence topo I binding.

Anti–topo I binding to fibroblasts stimulates adhesion and induces activation of THP-1 monocytes.

Monocytic cells are among the first cells to infiltrate the tissues during the onset of SSc (26). The capacity of topo I to recruit SSc anti–topo I to fibroblast cell surfaces prompted us to examine whether this was sufficient to induce subsequent monocyte adhesion. THP-1 monocytes were cocultured with fibroblasts that had been preincubated with topo I (0.5 μg/ml) and normal or SSc IgG (50 μg/ml). THP-1 monocyte adhesion to NHDF-Ad was significantly increased when cells had been preincubated with topo I and IgG or anti–topo I purified from SSc anti–topo I–positive sera (P < 0.001 versus normal or SSc anti–topo I–negative sera) (Figure 4A).

thumbnail image

Figure 4. Analysis of monocyte adhesion to fibroblasts incubated with topo I and SSc patient antibodies. A, NHDF-Ad were preincubated with topo I (0.5 μg/ml) followed by IgG from normal sera (n = 3), IgG from SSc anti–topo I–negative sera (n = 4), or IgG from SSc anti–topo I–positive sera (n = 4), or with anti–topo I purified from SSc sera (Pur. anti–topo I) (n = 2). Bromodeoxyuridine (BrdU)–labeled THP-1 cells were then allowed to adhere to the pretreated normal human lung fibroblasts. THP-1 cell adhesion to NHDF-Ad incubated with topo I alone (no antibody [Ab]) served as the background value. Data are presented as the mean ratio of BrdU intensity in the presence of antibodies:BrdU intensity in the presence of the no-antibody control, from pooled triplicate experiments. THP-1 monocyte adhesion to NHDF-Ad was significantly increased in the presence of IgG from SSc anti–topo I–positive sera (∗ = P < 0.001 versus normal and versus SSc anti–topo I–negative sera) and in the presence of anti–topo I purified from SSc sera (∗ = P < 0.001 versus normal and versus SSc anti–topo I–negative sera). B and C, THP-1 monocyte activation in the presence of fibroblasts preincubated with topo I and SSc patient antibodies. THP-1 cells were cocultured for 3 days with NHDF-Ad preincubated with topo I (0.5 μg/ml) and IgG from normal sera (n = 2), from SSc anti–topo I–negative sera (n = 3), or from SSc anti–topo I–positive sera (n = 3), or with anti–topo I purified from SSc sera (n = 2). Expression levels of CD14 (B) and CD11b (C) on THP-1 cells were assessed by flow cytometry. Data are presented as the mean percentage of positive cells from pooled triplicate experiments. CD14 and CD11b expression on THP-1 cells was significantly increased in the presence of IgG from SSc anti–topo I–positive sera (∗ = P < 0.0004 versus normal and versus SSc anti–topo I–negative sera) and anti–topo I purified from SSc sera (∗ = P < 0.002 versus normal and versus SSc anti–topo I–negative sera). See Figure 1 for other definitions.

Download figure to PowerPoint

We next determined whether THP-1 monocyte adhesion was followed by monocyte activation. Indeed, 2 markers of monocyte activation, CD14 and CD11b, were significantly up-regulated on THP-1 monocytes cocultured with NHDF-Ad that had been preincubated with topo I and IgG from SSc anti–topo I–positive sera (P < 0.0004 versus normal and versus SSc anti–topo I–negative sera) or anti–topo I purified from SSc sera (P < 0.002 versus normal and versus SSc anti–topo I–negative sera) (Figures 4B and C). Similar results for monocyte adhesion and activation were observed when using NHLFs (data not shown). Taken together, these data indicate that the binding of topo I–anti–topo I complexes to fibroblast surfaces is sufficient for recruitment and subsequent activation of monocytes.

Topo I from apoptotic endothelial cells binds fibroblasts specifically.

Topo I expression is normally restricted to the nucleus. However, nuclear autoantigens are redistributed in apoptotic bodies during apoptosis (27, 28), and some even become available for autoantibody recognition (29). Since endothelial cell apoptosis is a hallmark of SSc (16), we examined whether topo I released from apoptotic HMVEC-D or HUVECs was also capable of specifically binding to fibroblasts. Endothelial cell apoptosis was induced by growth factor deprivation for 72 hours. Under these conditions, apoptosis is induced in >90% of the cells. Apoptosis was confirmed by immunofluorescence analysis with Hoechst/PI staining and flow cytometry analysis of phosphatidylserine exposure on the cell surface using FITC-conjugated annexin V (data not shown). Topo I present in the concentrated culture supernatant of apoptotic endothelial cells displayed the typical molecular species (76 and 80 kd) associated with caspase-cleaved topo I (30) (Figure 5A). A lysate from nonapoptotic endothelial cells and purified topo I were used as controls, the former containing the intact form of topo I (100 kd) and the latter displaying the typical molecular products of topo I proteolysis (major, 70 kd; minor, 80 kd) (31). Topo I is very sensitive to proteolysis and its degradation during purification is common (31).

thumbnail image

Figure 5. Binding of topo I from apoptotic endothelial cells to fibroblasts. Apoptosis of HMVEC-D was induced by serum starvation for 72 hours. A, Topo I in supernatants from apoptotic HMVEC-D was detected by immunoblotting with purified IgG from anti–topo I–positive SSc serum. Nonapoptotic HMVEC-D (Lysate) and affinity-purified topo I (Topo I) were used as controls. B, Binding of topo I from the supernatant of apoptotic HMVEC-D to NHDF-Ad or HMVEC-D was detected by indirect immunofluorescence using IgG purified from SSc anti–topo I–positive serum and a fluorescein isothiocyanate–conjugated second antibody to human IgG (green). Hoechst 33342 was used to detect nuclei (blue). Staining was observed only on NHDF-Ad. Results are representative of 3 independent experiments. See Figure 1 for definitions. (Original magnification × 200.)

Download figure to PowerPoint

The culture supernatant from apoptotic HMVEC-D was added to live, unfixed, and unpermeabilized NHDF-Ad and HMVEC-D for 1 hour, and topo I binding was detected with IgG purified from SSc anti–topo I–positive sera. Consistent with our previous findings (Figure 1C), topo I from apoptotic endothelial cells bound to fibroblasts but not to endothelial cell surfaces (Figure 5B). Similar results were obtained using NHLFs and HUVECs (data not shown). To confirm that topo I was detected, and not other proteins present in the supernatant from apoptotic endothelial cells, the experiments were repeated with anti–topo I purified from SSc sera and a mouse anti–topo I monoclonal antibody. Similar results were obtained (data not shown), confirming that topo I was the polypeptide in the apoptotic supernatant that specifically bound to fibroblast surfaces. No staining was observed when IgG from normal and SSc anti–topo I–negative sera were used (data not shown).

Preliminary characterization of topo I ligand on fibroblast surface.

One possible ligand for topo I on the fibroblast surface was surface-bound DNA. Indeed, cell surface DNA has been reported in previous studies (32, 33). However, prior incubation of intact cells with DNase did not affect topo I binding to the fibroblast surface (data not shown). On the other hand, prior trypsin–EDTA (0.125 mg/ml) treatment was found to completely inhibit topo I binding to the fibroblast surface (Figure 6A), suggesting that the topo I ligand was a protein. Interestingly, when trypsin–EDTA–treated cells were allowed to recuperate for 10, 20, or 30 minutes in complete medium at 37°C, gradual recovery of topo I binding to the cell surface was observed (Figure 6A). The speed of this recovery suggested that the topo I ligand was readily available in the cytoplasm for transport to the cell surface. To test this hypothesis, we pretreated fibroblasts with brefeldin A, a well-known inhibitor of protein transport to the cell surface (34). Brefeldin A–treated fibroblasts were then detached with trypsin–EDTA (0.125 mg/ml) and allowed to recuperate in complete medium for 20 minutes, as described above. Under these conditions, no topo I ligand reexpression at the fibroblast surface was observed (Figure 6B). Taken together, these preliminary data suggest that the topo I ligand is a surface protein, or protein complex, that is capable of rapid surface trafficking.

thumbnail image

Figure 6. Effects of trypsin–EDTA and brefeldin A (BrefA) treatment on topo I binding to the fibroblast surface. A, Normal human lung fibroblasts (NHLFs) were detached using PBS/0.5% EDTA or trypsin–EDTA (0.125 mg/ml), and trypsin–EDTA–detached cells were allowed to recuperate in complete medium for 0, 10, 20, or 30 minutes. Topo I (0.5 μg/ml) binding to the cell surface was measured by flow cytometry using IgG from an anti–topo I–positive SSc patient (50 μg/ml) and phycoerythrin–conjugated anti-human IgG antibody. B, NHLFs were pretreated with 1.0 or 5.0 μg/ml brefeldin A or vehicle only for 2 hours in complete medium before being detached using trypsin–EDTA (0.125 mg/ml). Trypsin–EDTA–detached cells were allowed to recuperate in complete medium for 20 minutes. Topo I binding was assessed and measured as above. Data are presented as the mean and SD mean fluorescence intensity (MFI). Results are representative of at least 2 independent experiments. See Figure 1 for other definitions.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

ANAs are well-established markers for SSc. They are present in up to 90% of SSc patient sera and are generally associated with specific SSc subtypes (35–37). However, their role in the pathogenesis of SSc is still unclear, and few studies have directly addressed this subject (for review, see ref. 37). One of the reasons for this situation is that ANAs in SSc are traditionally considered bystander products of an immune system that has gone awry. Sequestered in cell nuclei, ANA targets are presumed to be inaccessible to circulating SSc autoantibodies, therefore leading to the notion that ANAs cannot play an active part in the disease etiology. Our recent findings shed new light on the possible role of some ANAs in SSc pathogenesis. Indeed, we demonstrate for the first time that topo I binds to fibroblast surfaces, recruits SSc anti–topo I, and subsequently induces monocyte adhesion and activation.

Topo I is a 100-kd nuclear enzyme responsible for DNA relaxation during transcription and replication (38). During apoptosis, topo I is cleaved into major fragments of 70 and 80 kd (30, 39) and is redistributed into blebs and apoptotic bodies (30). These apoptotic cell remnants should normally be rapidly cleared by phagocytes. However, in the event of a sudden increase in apoptotic cell numbers, the capacity of normal clearance mechanisms may be exceeded, resulting in the accumulation and progress of apoptotic cells toward a late apoptotic state (i.e., secondary necrosis), thereby allowing the release of apoptotic body contents to the medium. This may be the case in SSc, in which endothelial cell apoptosis is detected simultaneously in several tissues, notably, around small blood vessels (16, 40).

As we have shown above, topo I can be released from apoptotic endothelial cells, probably due to the increased permeability of secondary necrotic bodies. Hence, this released topo I could bind to nearby cells in SSc tissues. Since topo I has a specific affinity for fibroblast surfaces, it could bind to bystander fibroblasts and recruit circulating anti–topo I, present only in SSc patients, which would then induce adhesion and activation of circulating monocytes. Ultimately, this chain of events could lead 1) to amplification of the immune responses due to proinflammatory cytokines released by activated monocytes and 2) to fibrosis due to local secretion of profibrotic cytokines by activated fibroblasts. Thus, the presence of anti–topo I in SSc patients would be directly associated with increased immune responses and fibrosis.

The hypothesis that anti–topo I could directly contribute to the pathogenesis of fibrosis in SSc is supported by several indirect findings. First, anti–topo I autoantibodies are associated with a more severe SSc phenotype, notably, pulmonary fibrosis (8, 41–44). Second, in patients with anti–topo I, these antibodies are almost invariably present very early in the disease course (43). Third, anti–topo I titers correlate with disease severity (42). Moreover, anti–topo I titers are higher in patients with active disease than in those with inactive disease (42). Some SSc patients even display anti–topo I titers that fluctuate in parallel with the total skin score (42). Strikingly, SSc patients who have lost anti–topo I experience significant improvements in pulmonary function and survival compared with those in whom anti–topo I persist (44). Although convincing, all of these findings remain indirect, and an in situ localization of anti–topo I in SSc tissues will be important to further establish their pathogenic role.

Anti–topo I are not the only autoantibodies that bind to the cell surface via a nuclear autoantigen. For example, nucleosomes, or DNA–histone complexes, have been shown to recruit anti–double-stranded DNA autoantibodies to the surface of various cell types (45). In that particular case, an unidentified 94-kd protein has been pinpointed as the target of nucleosomes on cell surfaces. It is therefore likely that topo I binds to cells via a surface protein or protein complex that remains to be identified. Interestingly, this ligand seems to be more abundant on fibroblasts, independently of their origin, than on other cell types. Experiments to identify the cell surface ligand of topo I on fibroblasts are currently in progress in our laboratory.

The pathogenesis of SSc, like that of all systemic autoimmune diseases, is extremely complex (2). The presence of specific autoantibodies is only one immune abnormality among several others in this disease. However, the association of specific ANAs with different SSc subsets and with different patterns of organ involvement suggests that these autoantibodies are directly implicated in the pathogenesis of the disease. Anti–topo I could therefore drive the disease toward a more severe phenotype. A better understanding of the pathogenic role of ANAs in SSc could lead to new therapeutic strategies to improve the survival of patients.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

We wish to thank Ms Mélanie Tremblay for performing the initial experiment showing topo I binding to fibroblasts, Ms Isabelle Clément for expert technical assistance, and Mr. Jacques Moisan for insightful suggestions on statistical considerations.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  • 1
    LeRoy EC, Black C, Fleischmajer R, Jablonska S, Krieg T, Medsger TA Jr, et al. Scleroderma (systemic sclerosis): classification, subsets and pathogenesis. J Rheumatol 1988; 15: 2025.
  • 2
    Jimenez SA, Derk CT. Following the molecular pathways toward an understanding of the pathogenesis of systemic sclerosis. Ann Intern Med 2004; 140: 3750.
  • 3
    Reveille JD, Solomon DH, and the American College of Rheumatology Ad Hoc Committee on Immunologic Testing Guidelines. Evidence-based guidelines for the use of immunologic tests: anticentromere, Scl-70, and nucleolar antibodies. Arthritis Rheum 2003; 49: 399412.
  • 4
    Giordano M, Valentini G, Migliaresi S, Picillo U, Vatti M. Different antibody patterns and different prognoses in patients with scleroderma with various extent of skin sclerosis. J Rheumatol 1986; 13: 9116.
  • 5
    Steen VD, Powell DL, Medsger TA Jr. Clinical correlations and prognosis based on serum autoantibodies in patients with systemic sclerosis. Arthritis Rheum 1988; 31: 196203.
  • 6
    Scussel-Lonzetti L, Joyal F, Raynauld JP, Roussin A, Rich E, Goulet JR, et al. Predicting mortality in systemic sclerosis: analysis of a cohort of 309 French Canadian patients with emphasis on features at diagnosis as predictive factors for survival. Medicine (Baltimore) 2002; 81: 15467.
  • 7
    Weiner ES, Earnshaw WC, Senecal JL, Bordwell B, Johnson P, Rothfield NF. Clinical associations of anticentromere antibodies and antibodies to topoisomerase I: a study of 355 patients. Arthritis Rheum 1988; 31: 37885.
  • 8
    Diot E, Giraudeau B, Diot P, Degenne D, Ritz L, Guilmot JL, et al. Is anti-topoisomerase I a serum marker of pulmonary involvement in systemic sclerosis? Chest 1999; 116: 71520.
  • 9
    Mitri GM, Lucas M, Fertig N, Steen VD, Medsger TA Jr. A comparison between anti-Th/To– and anticentromere antibody–positive systemic sclerosis patients with limited cutaneous involvement. Arthritis Rheum 2003; 48: 2039.
  • 10
    Renaudineau Y, Revelen R, Levy Y, Salojin K, Gilburg B, Shoenfeld Y, et al. Anti-endothelial cell antibodies in systemic sclerosis. Clin Diagn Lab Immunol 1999; 6: 15660.
  • 11
    Marks RM, Czerniecki M, Andrews BS, Penny R. The effects of scleroderma serum on human microvascular endothelial cells: induction of antibody-dependent cellular cytotoxicity. Arthritis Rheum 1988; 31: 152434.
  • 12
    Sgonc R, Gruschwitz MS, Boeck G, Sepp N, Gruber J, Wick G. Endothelial cell apoptosis in systemic sclerosis is induced by antibody-dependent cell-mediated cytotoxicity via CD95. Arthritis Rheum 2000; 43: 255062.
  • 13
    Chizzolini C, Raschi E, Rezzonico R, Testoni C, Mallone R, Gabrielli A, et al. Autoantibodies to fibroblasts induce a proadhesive and proinflammatory fibroblast phenotype in patients with systemic sclerosis. Arthritis Rheum 2002; 46: 160213.
  • 14
    Henault J, Tremblay M, Clement I, Raymond Y, Senecal JL. Direct binding of anti–DNA topoisomerase I autoantibodies to the cell surface of fibroblasts in patients with systemic sclerosis. Arthritis Rheum 2004; 50: 326574.
  • 15
    Ronda N, Gatti R, Giacosa R, Raschi E, Testoni C, Meroni PL, et al. Antifibroblast antibodies from systemic sclerosis patients are internalized by fibroblasts via a caveolin-linked pathway. Arthritis Rheum 2002; 46: 1595601.
  • 16
    Sgonc R, Gruschwitz MS, Dietrich H, Recheis H, Gershwin ME, Wick G. Endothelial cell apoptosis is a primary pathogenetic event underlying skin lesions in avian and human scleroderma. J Clin Invest 1996; 98: 78592.
  • 17
    LeRoy EC. Increased collagen synthesis by scleroderma skin fibroblasts in vitro: a possible defect in the regulation or activation of the scleroderma fibroblast. J Clin Invest 1974; 54: 8809.
  • 18
    Strehlow D, Korn JH. Biology of the scleroderma fibroblast. Curr Opin Rheumatol 1998; 10: 5728.
  • 19
    Trojanowska M, LeRoy EC, Eckes B, Krieg T. Pathogenesis of fibrosis: type 1 collagen and the skin. J Mol Med 1998; 76: 26674.
  • 20
    Lonzetti LS, Joyal F, Raynauld JP, Roussin A, Goulet JR, Rich E, et al. Updating the American College of Rheumatology preliminary classification criteria for systemic sclerosis: addition of severe nailfold capillaroscopy abnormalities markedly increases the sensitivity for limited scleroderma [letter]. Arthritis Rheum 2001; 44: 7356.
  • 21
    Subcommittee for Scleroderma Criteria of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee. Preliminary criteria for the classification of systemic sclerosis (scleroderma). Arthritis Rheum 1980; 23: 58190.
  • 22
    Senecal JL, Rauch J, Grodzicky T, Raynauld JP, Uthman I, Nava A, et al. Strong association of autoantibodies to human nuclear lamin B1 with lupus anticoagulant antibodies in systemic lupus erythematosus. Arthritis Rheum 1999; 42: 134753.
  • 23
    Auger FA, Lopez Valle CA, Guignard R, Tremblay N, Noel B, Goulet F, et al. Skin equivalent produced with human collagen. In Vitro Cell Dev Biol Anim 1995; 31: 4329.
  • 24
    Hebert MJ, Gullans SR, Mackenzie HS, Brady HR. Apoptosis of endothelial cells is associated with paracrine induction of adhesion molecules: evidence for an interleukin-1β-dependent paracrine loop. Am J Pathol 1998; 152: 52332.
  • 25
    Raymond MA, Vigneault N, Luyckx V, Hebert MJ. Paracrine repercussions of preconditioning on angiogenesis and apoptosis of endothelial cells. Biochem Biophys Res Commun 2002; 291: 2619.
  • 26
    Kraling BM, Maul GG, Jimenez SA. Mononuclear cellular infiltrates in clinically involved skin from patients with systemic sclerosis of recent onset predominantly consist of monocytes/macrophages. Pathobiology 1995; 63: 4856.
  • 27
    Dieude M, Senecal JL, Rauch J, Hanly JG, Fortin P, Brassard N, et al. Association of autoantibodies to nuclear lamin B1 with thromboprotection in systemic lupus erythematosus: lack of evidence for a direct role of lamin B1 in apoptotic blebs. Arthritis Rheum 2002; 46: 2695707.
  • 28
    Casciola-Rosen LA, Anhalt G, Rosen A. Autoantigens targeted in systemic lupus erythematosus are clustered in two populations of surface structures on apoptotic keratinocytes. J Exp Med 1994; 179: 131730.
  • 29
    Miranda-Carus ME, Askanase AD, Clancy RM, Di DF, Chou TM, Libera MR, et al. Anti-SSA/Ro and anti-SSB/La autoantibodies bind the surface of apoptotic fetal cardiocytes and promote secretion of TNF-α by macrophages. J Immunol 2000; 165: 534551.
  • 30
    Samejima K, Svingen PA, Basi GS, Kottke T, Mesner PW Jr, Stewart L, et al. Caspase-mediated cleavage of DNA topoisomerase I at unconventional sites during apoptosis. J Biol Chem 1999; 274: 433540.
  • 31
    Douvas AS, Achten M, Tan EM. Identification of a nuclear protein (Scl-70) as a unique target of human antinuclear antibodies in scleroderma. J Biol Chem 1979; 254: 1051422.
  • 32
    Bennett RM, Davis J, Merritt M. Anti-DNA antibodies react with DNA expressed on the surface of monocytes and B lymphocytes. J Rheumatol 1986; 13: 67985.
  • 33
    Bennett RM, Cornell KA, Merritt MJ, Bakke AC, Hsu PH, Hefeneider SH. Autoimmunity to a 28-30 kD cell membrane DNA binding protein: occurrence in selected sera from patients with SLE and mixed connective tissue disease (MCTD). Clin Exp Immunol 1991; 86: 3749.
  • 34
    Misumi Y, Misumi Y, Miki K, Takatsuki A, Tamura G, Ikehara Y. Novel blockade by brefeldin A of intracellular transport of secretory proteins in cultured rat hepatocytes. J Biol Chem 1986; 261: 11398403.
  • 35
    Harris ML, Rosen A. Autoimmunity in scleroderma: the origin, pathogenetic role, and clinical significance of autoantibodies. Curr Opin Rheumatol 2003; 15: 77884.
  • 36
    Hesselstrand R, Scheja A, Shen GQ, Wiik A, Akesson A. The association of antinuclear antibodies with organ involvement and survival in systemic sclerosis. Rheumatology (Oxford) 2003; 42: 53440.
  • 37
    Senecal JL, Henault J, Raymond Y. The pathogenic role of autoantibodies to nuclear autoantigens in systemic sclerosis (scleroderma). J Rheumatol 2005; 32: 16439.
  • 38
    Champoux JJ. DNA topoisomerases: structure, function, and mechanism. Annu Rev Biochem 2001; 70: 369413.
  • 39
    Casiano CA, Martin SJ, Green DR, Tan EM. Selective cleavage of nuclear autoantigens during CD95 (Fas/APO-1)-mediated T cell apoptosis. J Exp Med 1996; 184: 76570.
  • 40
    Nguyen VA, Sgonc R, Dietrich H, Wick G. Endothelial injury in internal organs of University of California at Davis line 200 (UCD 200) chickens, an animal model for systemic sclerosis (scleroderma). J Autoimmun 2000; 14: 1439.
  • 41
    Hatakeyama A, Kasturi KN, Wolf I, Phelps RG, Bona CA. Correlation between the concentration of serum anti-topoisomerase I autoantibodies and histological and biochemical alterations in the skin of tight skin mice. Cell Immunol 1996; 167: 13540.
  • 42
    Hu PQ, Fertig N, Medsger TA Jr, Wright TM. Correlation of serum anti-DNA topoisomerase I antibody levels with disease severity and activity in systemic sclerosis. Arthritis Rheum 2003; 48: 136373.
  • 43
    Weiner ES, Hildebrandt S, Senecal JL, Daniels L, Noell S, Joyal F, et al. Prognostic significance of anticentromere antibodies and anti-topoisomerase I antibodies in Raynaud's disease: a prospective study. Arthritis Rheum 1991; 34: 6877.
  • 44
    Kuwana M, Kaburaki J, Mimori T, Kawakami Y, Tojo T. Longitudinal analysis of autoantibody response to topoisomerase I in systemic sclerosis. Arthritis Rheum 2000; 43: 107484.
  • 45
    Jacob L, Viard JP, Allenet B, Anin MF, Slama FB, Vandekerckhove J, et al. A monoclonal anti-double-stranded DNA autoantibody binds to a 94-kDa cell-surface protein on various cell types via nucleosomes or a DNA-histone complex. Proc Natl Acad Sci U S A 1989; 86: 466973.