Abstract

Objective. To investigate the pathophysiological effect of immunoglobulin G (IgG) from systemic lupus erythematosus (SLE) patients on pleural mesothelial cells and related mechanisms.

Methods. Serum IgG from 28 lupus patients and 13 healthy controls was purified by protein-G affinity chromatography. The concentrations of anti-dsDNA-, anti-histone- and/or anti-nucleohistone-containing IgGs were determined by enzyme-linked immunosorbent assay (ELISA). Lupus patients were divided into an active (n = 12) and an inactive group (n = 16) on the basis of the SLE Disease Activity Index (SLEDAI). The binding of IgG to a human pleural mesothelial cell line (MeT-5A) under different conditions, including pretreatment with DNase and preincubation with exogenous histone, DNA or nucleohistone, was examined using flow cytometry and cellular ELISA. The effect of IgG on MeT-5A cell proliferation was studied using an MTT assay. Gene expression and protein synthesis for interleukin 1β (IL-1β), monocyte chemoattractant protein 1 (MCP-1) and transforming growth factor β1 (TGF-β1) in MeT-5A cells were determined using reverse transcription–polymerase chain reaction and ELISA.

Results. The binding of IgG to MeT-5A cells was higher in the active lupus group than the inactive lupus group (P = 0.047) and controls (P = 0.003). The binding decreased in both lupus groups following pretreatment of MeT-5A cells with DNase. The binding of IgG to MeT-5A cells was greater by 112% in the active lupus group after preincubation with histone (P < 0.001), but not with DNA or nucleohistone. Exposure of MeT-5A cells to IgG from either lupus group induced cell proliferation when compared with IgG from healthy controls (P = 0.04). Gene expression and protein synthesis of MCP-1, TGF-β1 and IL-1β in MeT-5A cells were significantly increased after incubation with IgG from patients with active lupus when compared with IgG from the inactive lupus and control groups (P < 0.01). The concentration of anti-dsDNA antibodies correlated with the binding of IgG to MeT-5A cells and the synthesis of cytokines by MeT-5A cells. The serum level of anti-histone antibodies in the active lupus group was higher than that in the inactive group (P = 0.015) and the serum concentration correlated with cell binding and MCP-1 production.

Conclusions. IgG from lupus patients can bind to MeT-5A cells and the binding is modulated by DNA or histone. Binding of anti-dsDNA-containing IgG to MeT-5A cells induces the synthesis of proinflammatory cytokines. Our findings suggest that the binding of anti-dsDNA antibodies, particularly the IgG isotype, to pleural mesothelium plays a direct pathogenetic role in inducing inflammatory injury in the serositis of SLE.

Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by the presence of autoantibodies to various nuclear or cytoplasmic antigens. Antibodies to double-stranded DNA (anti-dsDNA) are found in 40–90% of patients with SLE. Evidence has shown that anti-dsDNA antibodies, particularly of immunoglobulin (Ig) G isotype, play important roles in the pathogenesis of SLE and that a high serum concentration of anti-dsDNA antibodies is closely related to disease activity [1, 2].

Pleuritis occurs frequently in active lupus and an immune-complex-mediated mechanism may be a common way in which pleuritis develops. The anti-dsDNA titre is elevated in both serum and pleural fluid from patients with lupus pleuritis [3, 4]. Immunoglobulin deposits are detected in the pleura of patients with lupus pleuritis [5] and DNA–anti-dsDNA immune complexes are present in pleural effusion [6]. Whether tissue immune deposits are derived from circulating immune complexes or are formed in situ remains uncertain. Little is known of any direct pathogenic effect of anti-dsDNA antibodies on pleural mesothelial cells.

Our previous work has shown that anti-DNA antibodies are capable of binding to human glomerular mesangial cells [7] and endothelial cells [8]. The binding of anti-dsDNA antibodies to the endothelial cell can be mediated by either preformed DNA–anti-dsDNA immune complexes or by Ig directly [8]. Anti-dsDNA antibodies stimulate the release of cytokines, including interleukin (IL)-1, IL-6, IL-8 and transforming growth factor-β (TGF-β) from endothelial cells, which could induce endothelial damage [9, 10]. This binding of anti-dsDNA antibodies is mediated by DNA and histone and correlates closely with disease activities [11]. Furthermore, anti-dsDNA antibodies bind to and stimulate mononuclear cells to produce proinflammatory cytokines such as IL-1β, IL-6, IL-8, IL-10 and tumour necrosis factor α (TNF-α) [12].

Few studies have focused on the effect of anti-dsDNA antibodies on pleural mesothelial cells. The pleural mesothelial cells are vulnerable to various aetiological agents and participate actively in acute pleural inflammation through cell proliferation and the release of cytokines, including intercellular adhesion molecule 1 (ICAM-1), vascular cell adhesion molecule 1 (VCAM-1), IL-8, monocyte chemoattractant protein 1 (MCP-1), IL-1β and IL-6 [13]. Subsequently, the pleural mesothelial cells play a contributory role in fibroproliferative reactions by releasing fibrogenetic cytokines such as TGF-β, platelet-derived growth factor (PDGF) and fibroblast growth factor (FGF) [14, 15].

In this study, we examined the binding of anti-dsDNA-containing IgG from lupus patients to cultured pleural mesothelial cells (MeT-5A). The possible roles of different antigens, including DNA, histone and nucleohistone, in the binding of anti-dsDNA to MeT-5A cells were tested. We also investigated the effects of anti-dsDNA antibodies on the proliferation of MeT-5A cells and cytokine synthesis by these cells.

Materials and methods

Patients and controls

The study was conducted in accordance with the Declaration of Helsinki and was approved by the ethics committee for studies in humans of the University of Hong Kong. All subjects (patients and healthy controls) gave written consent for serum and tissue collection. Serum samples were collected from 28 randomly selected lupus patients (mean age ± s.e.m., 34.8 ± 8.7 yr; 26 females and two males) and 13 healthy controls (30.4 ± 3.9 yr; 12 females and one male). Lupus patients fulfilled four or more diagnostic criteria of the American Rheumatism Association for SLE [16] and they were categorized as ‘active’ (n = 12) if the SLE Disease Activity Index (SLEDAI) was ≥10 and ‘inactive’ (n = 16) if the SLEDAI was ≤4. Patients with SLEDAI between 4 and 10 were excluded from this study. Of the 12 active patients, two had pleural effusion and two had pleuritic rub. One inactive patient had a small pericardial effusion detected by echocardiography and one had pleural effusion. At the time of serum collection, the active patients had not yet been commenced on treatment and the inactive patients were maintained on low-dose immunosuppressive therapy consisting of prednisolone and azathioprine.

Isolation of IgG

IgG was isolated from sera by protein-G affinity chromatography (Pharmacia, Uppsala, Sweden). The purified IgG was dialysed after centrifugal concentration (Amicon, Beverley, MA, USA). The purity of the IgG antibodies was ascertained by sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE). The purified IgG was stored at −70°C until use.

Determination of anti-dsDNA, anti-histone and anti-nucleohistone antibodies by ELISA

An enzyme-linked immunosorbent assay (ELISA) was developed for the determination of anti-dsDNA, anti-histone and anti-nucleohistone antibody concentrations in IgG preparations. Calf thymus DNA was obtained from Sigma (St Louis, MO, USA). Histone and nucleohistone (Worthington Biochemical Corporation, Lakewood, NJ, USA) were dissolved in 8 m urea solution and dialysed against phosphate-buffered saline (PBS). Nucleohistone represented nucleosomes containing the nucleoprotein complex of DNA and histones. Briefly, wells of microtitre plates (Immunon 2; Dynatech, Marnes la Coquette, France) were coated with 100 μl of calf thymus DNA (10 μg/ml), histone (40 μg/ml) or nucleohistone (10 μg/ml) in 0.015 m sodium carbonate and 0.03 m sodium bicarbonate (pH 9.8) overnight at room temperature. Two hundred microlitres of 3% bovine serum albumin (BSA; Sigma) in TBS was added to each well, and the plates were incubated for 2 h in order to block non-specific binding. After washing, 100 μl of serum (diluted 1:100) or purified IgG samples (100 μl, 200 μg/ml) in TBS (Tris-buffered saline) containing 10% fetal calf serum (Gibco, Chargrin Falls, NY, USA) and 3% BSA were added to each well and the plates were incubated for a further 2 h. After washing, bound Igs were detected using goat anti-human IgG–alkaline phosphatase (Biosource International, Camarillo, CA, USA) conjugated in TBS (1:1000), followed by p-nitrophenyl phosphate (Sigma) as substrate. The optical density (OD) at 405 nm was read on a spectrophotometer and the concentration of anti-dsDNA antibodies was then determined according to a standard curve constructed using the international standard. The anti-histone and anti-nucleohistone levels were expressed as OD at 450 nm (A450).

Cell culture

The pleural mesothelial cell line (MeT-5A) was obtained from the American Type Culture Collection (Rockville, MD, USA). The cells were maintained with RPMI 1640 culture medium (Gibco) containing 10% fetal calf serum, 100 U/ml penicillin and 100 μg/ml streptomycin in an atmosphere and 5% CO2/95% room air at 37°C. Subculture or preparation of cells for assay was done when the cells reached confluence.

Reverse transcription–polymerase chain reaction

MeT-5A cells were grown to confluence in six-well cell culture plates, growth-arrested, and exposed to IgG for defined time periods at 37°C. Total cellular RNA was then extracted and reverse-transcribed to cDNA. The polymerase chain reaction (PCR) was carried out as described previously [12]. The oligonucleotide sequences of cDNA primers were designed from GenBank sequences. The primer sequences and PCR profiles for amplifying the various target genes and their product sizes are summarized in Table 1. For quantification, human glyceraldehyde 3-phosphate-dehydrogenase (GAPDH) primers were included in every reaction as an internal control. The PCR products were separated by 2% weight/volume agarose gels and stained with ethidium bromide (Sigma), and the gel image was captured and analysed using the Gel Doc 1000 Densitometry System and Quantity One (Bio-Rad Laboratories, Hercules, CA, USA). The PCR result was expressed as an amplicon ratio with respect to GAPDH.

Table 1.

Primer sequence and size of PCR product for reverse transcription–PCR

 5′ to 3′ sequence of Size of PCR 
Primer PCR primers product (bp) 
MCP-1  499 
    Forward primer AGT CTC TGC CGC CCT TCT GT  
    Backward primer CCC CAA GTC TCT GTA TCT AA  
IL-1β  648 
    Forward primer GCC CTA AAC AGA TGA AGT GCT  
    Backward primer CTG CTT GAG AGG TGC TGA TGT  
TGF-β1  191 
    Forward primer GCC CTG GAC ACC AAC TAT TGC T  
    Backward primer AGG CTC CAA ATG TAG GGG CAG G  
GAPDH ACC ACA GTC CAT GCC ATC AC 452 
 TCC ACC ACC CTG TTG CTG TA  
GAPDH TGA AGG TCG GAG TCA ACG GAT TTG GT 983 
 CAT GTG GGC CAT GAG GTC CAC CAC  
 5′ to 3′ sequence of Size of PCR 
Primer PCR primers product (bp) 
MCP-1  499 
    Forward primer AGT CTC TGC CGC CCT TCT GT  
    Backward primer CCC CAA GTC TCT GTA TCT AA  
IL-1β  648 
    Forward primer GCC CTA AAC AGA TGA AGT GCT  
    Backward primer CTG CTT GAG AGG TGC TGA TGT  
TGF-β1  191 
    Forward primer GCC CTG GAC ACC AAC TAT TGC T  
    Backward primer AGG CTC CAA ATG TAG GGG CAG G  
GAPDH ACC ACA GTC CAT GCC ATC AC 452 
 TCC ACC ACC CTG TTG CTG TA  
GAPDH TGA AGG TCG GAG TCA ACG GAT TTG GT 983 
 CAT GTG GGC CAT GAG GTC CAC CAC  

Binding of IgG to MeT-5A cells by flow cytometry

MeT-5A cells were grown to confluence and were harvested using 0.05% trypsin for 5 min at room temperature and then fixed in 1% paraformaldehyde (PFA) at 4°C for 15 min. One hundred microlitres of MeT-5A cells (5 × 105 cells/ml) in staining buffer (PBS with 1% fetal bovine serum and 0.1% sodium azide) was mixed with purified IgG (200 μg/ml) from the active or inactive SLE group or the healthy controls and incubated for 30 min at 4°C. After washing with ice-cold staining buffer, 50 μl of rabbit F(ab′)2 anti-human IgG–FITC (fluorescein isothiocyanate) conjugate (Dako, Glostrup, Denmark) at 50-fold dilution was added and cells were further incubated for 30 min. Cells were spun down, resuspended with 250 μl of 1% PFA and analysed by flow cytometry on a Coulter EPICS analyser (Coulter Electronic, Miami, FL, USA). A minimum of 5000 cells for each sample were analysed. The result was expressed as mean fluorescence intensity (MFI).

For some experiments, MeT-5A cells were treated with DNase (50 μg/ml) in Tris–MgCl2 (10 mm) at 37°C for 15 min before the binding assay, to examine the effect of cell surface-bound DNA on cell binding of IgG. To study the effects of exogenous DNA, histone, or nucleohistone on IgG binding to MeT-5A cells, the cells were preincubated with different concentrations of DNA, histone or nucleohistone (0–5 mg/ml) at room temperature for 30 min before the binding experiment. In order to study whether the binding of IgG was Fc-dependent, F(ab′)2 fragments of IgG were prepared by pepsin digestion. Briefly, purified IgG was digested using immobilized pepsin (Pierce, Rockford, IL, USA) at 37°C for 4 h in a shaking water bath. The F(ab′)2 fragments were separated from the Fc portion and undigested IgG by protein-G affinity chromatography (Pharmacia, Uppsala, Sweden). The binding of F(ab′)2 fragments of IgG to MeT-5A cells was determined by flow cytometry, as described above.

Inhibition assay of cell binding by cellular ELISA

MeT-5A cells at the log phase were harvested and seeded in 96-well culture plates until confluence. MeT-5A cells were fixed with 1% PFA for 10 min and blocked with 3% BSA (bovine serum albumin) at 4°C overnight. In order to see if DNA, histone or nucleohistone could block the binding of IgG to MeT-5A cells, IgG (100 μl, 200 μg/ml), before incubation with MeT-5A cells, was coincubated with DNA (1 mg/ml), histone (1 mg/ml) or nucleohistone (1 mg/ml) in 100 μl of TBS for 1 h. The plates were then incubated for 2 h at room temperature. After three washes with 200 μl of TBS per well, 100 μl of goat F(ab′)2 anti-human IgG–horseradish peroxidase conjugate was added and the plates were further incubated for 1 h. After washing, o-phenylenediamine (OPD) was added and the OD at 450 nm was determined.

Proliferation studies by MTT assay

Cell proliferation was measured using the MTT [3-(4,5- dimethylthiazole-2-yl)-2,5-diphenyltetrazolium bromide] assay (R&D Systems, Minneapolis, MN, USA). Briefly, 1.25 × 105 MeT-5A cells were seeded in each well of 96-well polystyrene plates. The cells were arrested with serum-free medium 1 day before the experiment. Purified IgG (final concentration 200 μg/ml) was added to the wells and the plate was incubated at 37°C in 5% CO2 for half an hour. Ten microlitres of MTT reagent was added to each well and the plate was incubated at 37°C until purple-blue crystals were clearly visible under an inverted microscope. Detergent reagent (100 μl; R&D Systems, Minneapolis, MN, USA) was then added and the plate was incubated at 37°C for 4 h. The OD of the wells was read at 550 nm with a reference wavelength of 650 nm.

Immunoassay for cytokines in cell culture supernatant

MeT-5A cells were grown to confluence in six-well culture plates. The cultured cells were arrested with serum-free medium for 48 h before the experiment. The cells were incubated with IgG from SLE patients or healthy controls for 16 h at 37°C. Culture supernatant was collected and the concentrations of TGF-β1, MCP-1 or IL-1β were measured using an ELISA kit (TGF-β1, R&D Systems; MCP-1, Chemicon, Temecula, CA, USA; IL-1β, MedSystem Diagnostics, Vienna, Austria) according to the manufacturers’ protocols. The sensitivity of the assay was 7, 13.1 and 1.5 pg/ml for TGF-β1, MCP-1 and IL-1β respectively. The intra-batch coefficient of variation was 5.2% for TGF-β1, 7.9% for MCP-1 and 8% for IL-1β.

Statistical analysis

The results were expressed as mean ±s.d. when the normal distribution could be assumed, otherwise as median and range. Differences in parameters between groups were evaluated either by analysis of variance or the Kruskal–Wallis test, as appropriate. Correlations between different study variables were detected by Spearman's correlation test. P < 0.05 was considered significant. The analysis was performed with the SPSS 10.0 statistical package.

Results

Characterization of purified IgG preparations

The concentration of anti-dsDNA antibodies in purified IgG from the lupus patients (median 30 IU/ml, range 4.0–1690 IU/ml) was higher than that for the healthy controls (median 14.4 IU/ml, range 4.0–41.9 IU/ml, P < 0.001). Amongst the lupus patients, the concentrations of anti-dsDNA antibodies were significantly greater in the active group than in the inactive group (median 56.7 IU/ml, range 44.6–1690 IU/ml vs median 14.3 IU/ml, range 4.4–42.3 IU/ml; P < 0.001) (Fig. 1).

Fig. 1.

Concentration of IgG anti-DNA antibodies in patients with SLE, determined by ELISA. The concentration was significantly higher in the active SLE group than in the inactive SLE group and the healthy controls. *P < 0.001.

Fig. 1.

Concentration of IgG anti-DNA antibodies in patients with SLE, determined by ELISA. The concentration was significantly higher in the active SLE group than in the inactive SLE group and the healthy controls. *P < 0.001.

The anti-histone antibody level in the active SLE group was higher than that in the inactive SLE group (A450 0.563 ± 0.142 vs 0.441 ± 0.059; P = 0.015). The anti-histone antibody level in the active SLE group was also higher than that in the control group (A450 0.563 ± 0.142 vs 0.423 ± 0.103; P = 0.011). There was no difference in the anti-histone antibody level between the inactive SLE and control groups (A450 0.441 ± 0.059 vs 0.423 ± 0.103; P > 0.05). For anti-nucleohistone antibodies, there were no significant differences between the groups (A450 0.612 ± 0.139, 0.544 ± 0.106 and 0.492 ± 0.084 in active SLE, inactive SLE and healthy controls respectively).

Binding of IgG to MeT-5A cells and effect of DNase treatment on binding

The binding of IgG to MeT-5A cells (expressed as the MFI) was higher in the active SLE group than in the inactive SLE group (MFI 0.894 ± 0.238 vs 0.717 ± 0.088; P = 0.047). There was also a significant difference in binding between the active SLE group and the healthy controls (MFI 0.894 ± 0.238 vs 0.599 ± 0.039; P = 0.003).

After treating the MeT-5A cells with DNase, the binding was decreased in both the active SLE group (no DNase treatment 0.894 ± 0.238, DNase treatment 0.778 ± 0.152; P = 0.018) and the inactive SLE group (no DNase treatment 0.717 ± 0.088, DNase treatment 0.668 ± 0.079; P = 0.002). Such a difference was not observed with IgG isolated from healthy controls (no DNase treatment 0.599 ± 0.035, DNase treatment 0.574 ± 0.039; P > 0.05) (Fig. 2).

Fig. 2.

Binding of IgG to MeT-5A cells and the effect of DNase pretreatment, expressed as MFI. Without DNase treatment (solid columns), IgG from the active SLE group showed greater binding affinity to MeT-5A cells than IgG from the inactive SLE or control group (▵P = 0.047; ▵▵P = 0.003). After DNase treatment (open columns), binding was significantly decreased in the SLE groups compared with no DNase treatment (*P = 0.018; **P = 0.002).

Fig. 2.

Binding of IgG to MeT-5A cells and the effect of DNase pretreatment, expressed as MFI. Without DNase treatment (solid columns), IgG from the active SLE group showed greater binding affinity to MeT-5A cells than IgG from the inactive SLE or control group (▵P = 0.047; ▵▵P = 0.003). After DNase treatment (open columns), binding was significantly decreased in the SLE groups compared with no DNase treatment (*P = 0.018; **P = 0.002).

Binding of IgG to MeT-5A cells preincubated with exogenous histone, nucleohistone or DNA

The binding of IgG from patients with active lupus to MeT-5A cells increased by 112% after preincubation with histone (histone preincubation 1.555 ± 0.134, no histone preincubation 0.735 ± 0.262; P < 0.001). The increase in binding after histone preincubation was concentration-dependent in the active lupus group (Figs 3 and 4). A similar increase in binding after histone preincubation was not observed in patients with inactive lupus or in healthy controls. In contrast, the binding of IgG to MeT-5A cells was not enhanced by preincubation with exogenous DNA or nucleohistone (Fig. 3).

Fig. 3.

Binding of IgG to MeT-5A cells preincubated with exogenous DNA, histone or nucleohistone, expressed as percentage of control (no preincubation with exogenous compound). Binding to MeT-5A cells was increased by 112% in the active SLE group (solid columns) after preincubation with histone (*P < 0.001). There were no significant changes in the inactive SLE group (shaded columns) or healthy controls (open columns). Binding of IgG to MeT-5A cells was unaffected by exogenous DNA or nucleohistone treatment.

Fig. 3.

Binding of IgG to MeT-5A cells preincubated with exogenous DNA, histone or nucleohistone, expressed as percentage of control (no preincubation with exogenous compound). Binding to MeT-5A cells was increased by 112% in the active SLE group (solid columns) after preincubation with histone (*P < 0.001). There were no significant changes in the inactive SLE group (shaded columns) or healthy controls (open columns). Binding of IgG to MeT-5A cells was unaffected by exogenous DNA or nucleohistone treatment.

Fig. 4.

Binding of IgG from patients with active SLE (*) and inactive SLE group (□) and from the control group (▵) to MeT-5A cells after preincubation with different concentrations of histone.

Fig. 4.

Binding of IgG from patients with active SLE (*) and inactive SLE group (□) and from the control group (▵) to MeT-5A cells after preincubation with different concentrations of histone.

Effects of coincubated exogenous histone, nucleohistone and DNA on binding of IgG to MeT-5A cells

In neither the active nor the inactive SLE group was a significant difference observed in the binding of IgG to MeT-5A cells when IgG was coincubated with exogenous DNA, histone or nucleohistone (Fig. 5).

Fig. 5.

Binding of IgG coincubated with exogenous DNA, histone or nucleohistone, determined by cellular ELISA and expressed as percentage of control (no preincubation with exogenous compound). No significant change in binding was observed in the active SLE group (solid columns), the inactive SLE group (shaded columns) or the control group (open columns).

Fig. 5.

Binding of IgG coincubated with exogenous DNA, histone or nucleohistone, determined by cellular ELISA and expressed as percentage of control (no preincubation with exogenous compound). No significant change in binding was observed in the active SLE group (solid columns), the inactive SLE group (shaded columns) or the control group (open columns).

Binding of pepsin-digested IgG to MeT-5A cells

Following the pepsin digestion, the binding of F(ab′)2 fragment of IgG to MeT-5A cells remained significantly higher in the active SLE group (MFI 0.76 ± 0.21) than in the inactive SLE group (0.54 ± 0.09) and the healthy controls (0.47 ± 0.10) (P = 0.0098 and P = 0.002 respectively). No significant difference was found between the inactive SLE group and the healthy controls (Fig. 6).

Fig. 6.

Binding of pepsin-digested IgG to MeT-5A cells measured by flow cytometry. Like the binding of undigested IgG (solid columns), the binding of IgG F(ab′)2 fragments of antibodies (open columns) from the active SLE group was significantly greater than that of fragments from the inactive SLE or control group. *P = 0.0098; **P = 0.002.

Fig. 6.

Binding of pepsin-digested IgG to MeT-5A cells measured by flow cytometry. Like the binding of undigested IgG (solid columns), the binding of IgG F(ab′)2 fragments of antibodies (open columns) from the active SLE group was significantly greater than that of fragments from the inactive SLE or control group. *P = 0.0098; **P = 0.002.

Effect of IgG on MeT-5A cell proliferation

The proliferation index was expressed as MTT conversion, which was determined as OD at 550 nm. The proliferation index of the cells cultured with IgG from the active SLE group was higher than that of cells cultured with control IgG (A550 1.43 ± 0.20 vs 1.26 ± 0.16; P = 0.036). The cell proliferation index was also higher in the inactive SLE group than in the control group (A550 1.39 ± 0.16 vs 1.26 ± 0.16; P = 0.04). No significant difference was found between the active and inactive SLE groups (A550 1.43 ± 0.20 vs 1.39 ± 0.16; P > 0.05) (Fig. 7).

Fig. 7.

Effect of IgG on proliferation of MeT-5A cells. The proliferation index was expressed as MTT conversion, measured as OD at 550 nm. There was a significant increase in cell proliferation stimulated by IgG from the active SLE group (solid columns) or the inactive SLE group (shaded columns) when compared with IgG from healthy controls (open columns). *P = 0.036; **P = 0.04.

Fig. 7.

Effect of IgG on proliferation of MeT-5A cells. The proliferation index was expressed as MTT conversion, measured as OD at 550 nm. There was a significant increase in cell proliferation stimulated by IgG from the active SLE group (solid columns) or the inactive SLE group (shaded columns) when compared with IgG from healthy controls (open columns). *P = 0.036; **P = 0.04.

Effect of IgG on differences in cytokine gene expression and synthesis by MeT-5A cells

Expression of the MCP-1, TGF-β1 and IL-1β genes in cultured MeT-5A cells is shown in Figs 8A, 9A and 10A. The expression of MCP-1 mRNA in MeT-5A cells incubated with IgG from the active SLE group (2.45 ± 1.05) was significantly higher than that in MeT-5A cells incubated with IgG from the inactive SLE group (0.69 ± 0.29, P < 0.001) and the healthy controls (0.92 ± 0.21, P < 0.001). Similarly, expression of the TGF-β1 gene in MeT-5A cells incubated with IgG from the active SLE group was higher than that in cells incubated with IgG from the inactive SLE group (1.18 ± 0.10 vs 0.91 ± 0.12; P < 0.001) or control IgG (1.18 ± 0.10 vs 1.05 ± 0.07; P = 0.007). For IL-1β gene expression, there was a significant increase in MeT-5A cells incubated with IgG from the active SLE group compared with cells incubated with IgG from healthy controls (1.13 ± 0.145 vs 0.94 ± 0.10; P = 0.006). However, no significant difference in IL-1β gene expression was observed between the active and inactive SLE groups (1.13 ± 0.145 vs 1.02 ± 0.24; P > 0.05).

Fig. 8.

(A) MCP-1 mRNA expression in MeT-5A cells induced by IgG autoantibodies. Results are expressed as the MCP-1/GAPDH amplicon ratio. The horizontal bar represents the mean of the measurements. MCP-1 mRNA expression induced by IgG autoantibodies from the active SLE group was greater than that induced by IgG autoantibodies from the inactive SLE and control groups. (B) Synthesis of MCP-1 by MeT-5A cells stimulated by IgG autoantibodies. Results are concentrations of MCP-1 in the cell culture supernatant (pg/ml). The horizontal bar represents the mean of the measurements. Synthesis of MCP-1 was greater after stimulation by IgG autoantibodies from the active SLE group when compared with IgG autoantibodies from the inactive SLE and control groups.

Fig. 8.

(A) MCP-1 mRNA expression in MeT-5A cells induced by IgG autoantibodies. Results are expressed as the MCP-1/GAPDH amplicon ratio. The horizontal bar represents the mean of the measurements. MCP-1 mRNA expression induced by IgG autoantibodies from the active SLE group was greater than that induced by IgG autoantibodies from the inactive SLE and control groups. (B) Synthesis of MCP-1 by MeT-5A cells stimulated by IgG autoantibodies. Results are concentrations of MCP-1 in the cell culture supernatant (pg/ml). The horizontal bar represents the mean of the measurements. Synthesis of MCP-1 was greater after stimulation by IgG autoantibodies from the active SLE group when compared with IgG autoantibodies from the inactive SLE and control groups.

Fig. 9.

(A) TGF-β1 mRNA expression in MeT-5A cells induced by IgG autoantibodies, expressed as the TGF-β1/GAPDH amplicon ratio. The horizontal bar represents the mean of the measurements. TGF-β1 mRNA expression induced by IgG autoantibodies from the active SLE group was greater than that induced by IgG autoantibodies from the inactive SLE or control group. (B) Synthesis of TGF-β1 by MeT-5A cells stimulated by IgG autoantibodies. Results are concentrations of TGF-β1 in the supernatant of cell cultures (pg/ml). The horizontal bar represents the mean of the measurements. The synthesis of TGF-β1 was greater after stimulation by IgG autoantibodies from the active SLE group compared with IgG autoantibodies from the inactive SLE or control group.

Fig. 9.

(A) TGF-β1 mRNA expression in MeT-5A cells induced by IgG autoantibodies, expressed as the TGF-β1/GAPDH amplicon ratio. The horizontal bar represents the mean of the measurements. TGF-β1 mRNA expression induced by IgG autoantibodies from the active SLE group was greater than that induced by IgG autoantibodies from the inactive SLE or control group. (B) Synthesis of TGF-β1 by MeT-5A cells stimulated by IgG autoantibodies. Results are concentrations of TGF-β1 in the supernatant of cell cultures (pg/ml). The horizontal bar represents the mean of the measurements. The synthesis of TGF-β1 was greater after stimulation by IgG autoantibodies from the active SLE group compared with IgG autoantibodies from the inactive SLE or control group.

Fig. 10.

(A) IL-1β mRNA expression in MeT-5A cells induced by IgG autoantibodies, expressed as the IL-1β/GAPDH amplicon ratio. The horizontal bar represents the mean of the measurements. IL-1β mRNA expression was higher after induction by IgG autoantibodies from the active SLE group than after induction by antibodies from the control group. (B) Synthesis of IL-1β by MeT-5A cells stimulated by IgG autoantibodies. Results are concentrations of IL-1β in the cell culture supernatant (pg/ml). The horizontal bar represents the mean of the measurements. The synthesis of IL-1β was greater after stimulation by IgG autoantibodies from the active SLE group compared with stimulation by IgG autoantibodies from the inactive SLE or control group. The synthesis of IL-1β was also greater in cells treated with IgG from the inactive SLE patients than in cells treated with IgG from the healthy controls.

Fig. 10.

(A) IL-1β mRNA expression in MeT-5A cells induced by IgG autoantibodies, expressed as the IL-1β/GAPDH amplicon ratio. The horizontal bar represents the mean of the measurements. IL-1β mRNA expression was higher after induction by IgG autoantibodies from the active SLE group than after induction by antibodies from the control group. (B) Synthesis of IL-1β by MeT-5A cells stimulated by IgG autoantibodies. Results are concentrations of IL-1β in the cell culture supernatant (pg/ml). The horizontal bar represents the mean of the measurements. The synthesis of IL-1β was greater after stimulation by IgG autoantibodies from the active SLE group compared with stimulation by IgG autoantibodies from the inactive SLE or control group. The synthesis of IL-1β was also greater in cells treated with IgG from the inactive SLE patients than in cells treated with IgG from the healthy controls.

The concentrations of MCP-1, TGF-β1 and IL-1β in culture supernatants of MeT-5A cells stimulated with different IgG preparations are shown in Figs 8B, 9B and 10B. IgG from patients with active lupus enhanced the synthesis of MCP-1 in MeT-5A cells (97.60 ± 8.49 pg/ml) compared with IgG from either the inactive lupus group (67.08 ± 16.10 pg/ml) and healthy controls (58.86 ± 19.24 pg/ml) (P < 0.001). In parallel with the gene expression, there was increased synthesis of TGF-β1 in MeT-5A cells incubated with IgG from active lupus subjects (67.31 ± 11.95 pg/ml) compared with values from the inactive lupus group (15.29 ± 13.26 pg/ml, P < 0.001) and the healthy controls (32.49 ± 16.53 pg/ml, P < 0.001). The synthesis of IL-1β by MeT-5A cells was significantly higher after incubation with IgG from the active SLE (682.40 ± 233.59 pg/ml) than after incubation with IgG from the inactive SLE group (385.40 ± 135.67 pg/ml, P = 0.003) or the healthy controls (288.10 ± 54.26 pg/ml, P < 0.001).

Correlation analysis

The binding of IgG from the patients with SLE to MeT-5A cells correlated with the serum concentrations of anti-dsDNA (r = 0.615, P < 0.001) and anti-histone antibodies (r = 0.375, P = 0.034) but not with the concentration of anti-nucleohistone antibodies.

The concentration of anti-dsDNA antibodies correlated with the gene expression of MCP-1 (r = 0.578, P = 0.001), TGF-β1 (r = 0.409, P = 0.029) and IL-1β (r = 0.422, P = 0.028) by MeT-5A cells. There was also a positive correlation between anti-histone antibody titre and MCP-1 gene expression (r = 0.386, P = 0.043), but not with TGF-β1 or IL-1β gene expression

Similar findings were observed with respect to the correlation between anti-dsDNA antibody titre and the MeT-5A cell supernatant concentrations of MCP-1 (r = 0.578, P = 0.001), TGF-β1 (r = 0.637, P < 0.001) and IL-1β (r = 0.616, P < 0.001). Again, there was an association between anti-histone antibody titre and the synthesis of MCP-1 (r = 0.368, P = 0.046) but not with TGF-β1 or IL-1β.

There was a good correlation between gene expression and protein synthesis by MeT-5A cells for MCP-1 (r = 0.68, P < 0.001) and for TGF-β1 (r = 0.565, P = 0.003). Such correlation was not observed with IL-1β.

Discussion

Previous studies have shown that anti-dsDNA antibodies bind to endothelial and glomerular mesangial cells [79]. The binding of anti-dsDNA antibodies to endothelial or mesangial cells is increased in active disease, vasculitis or lupus nephritis, suggesting that the binding of these antibodies to target cells may play a direct pathogenic role [810]. In lupus patients with pleuritis, the anti-dsDNA antibody titre is raised in both serum and pleural fluid [3] and DNA–anti-dsDNA immune complexes are detected in pleural fluid [6]. However, little is known about the interaction of polyclonal IgG anti-dsDNA antibodies with pleural mesothelial cells and their pathogenic roles in lupus pleuritis. This is the first study to explore the direct pathogenic effect of anti-dsDNA-containing IgG isolated from patients with SLE in a pleural mesothelial cell line. An immortalized cultured cell line was used in the present study because the use of primary human pleural mesothelial cells obtained from pleural effusion by thoracentesis has drawbacks as an in vitro model. These include (i) difficulty in obtaining a consistent supply of human pleural effusion with a sufficient number of pleural mesothelial cells; (ii) batch-to-batch inconsistency due to donor variation; (iii) potential activation of pleural mesothelial cell obtained from disease-induced effusion; and (iv) the slow growth rate of human primary pleural mesothelial cells.

Our present findings show that IgG from patients with SLE is, indeed, able to bind to pleural mesothelial cells. The binding of IgG from the active lupus group was significantly stronger than that of IgG from either the inactive lupus or the control group, suggesting that this binding is closely related to disease activity. After pepsin digestion, the binding of IgG F(ab′)2 fragments of antibodies from the active lupus group was still higher than that of IgG from the inactive lupus or control group; this is similar to what occurred with undigested IgG. Our result indicates that, unlike the binding of antibodies to glomerular basement membrane (GBM) through the Fcγ fragment [17], the binding of IgG to MeT-5A cells is dependent on the F(ab′)2 fragment. Reduction of binding by IgG isolated from lupus patients after DNase treatment of MeT-5A cells indicates that DNA on the cell surface is essential in the binding of anti-dsDNA-containing IgG to various cell types, including pleural mesothelial cells. Our previous study showed that the binding of the anti-dsDNA antibodies to endothelial cells was influenced by histone and DNA [11]. Results from the present study demonstrate that IgG from active lupus patients contains a raised level of anti-histone antibodies with high binding activity to MeT-5A cells. This is consistent with other work suggesting that anti-histone antibodies are associated with disease activity in lupus nephritis [18]. We further showed that preincubation of histone with MeT-5A cells increased the binding of IgG from active lupus patients to MeT-5A cells in a concentration-dependent fashion. It has been demonstrated that cell-surface proteoglycans, such as perlecan, can interact with histone molecules [19]. We speculate that the enhancement of binding after preincubation of MeT-5A cells with histone could be due to the immobilization of histone with DNA or other proteins present on the cell membrane of MeT-5A cells. Anti-histone-containing IgG will then bind to the immobilized histone. The binding of IgG to MeT-5A cells remained unchanged when IgG was coincubated with histone. This could have been due to the steric hindrance of the binding of the exogenous histone with proteoglycans on the cell membrane after the formation complexes between exogenous histone and anti-histone antibodies. Further investigation is warranted to confirm this hypothesis.

Several studies have shown that nucleosome-specific antibodies can be detected before the occurrence of the anti-dsDNA or anti-histone antibodies in the lupus-prone mouse, and that the antibodies continue to exist during the course of the disease along with the anti-dsDNA and anti-histone antibodies [20, 21]. It has also been suggested that nucleosomes play an essential role in the development of lupus nephritis by targeting antinuclear antibodies to mesangial cells [22]. The monoclonal nucleosome-specific antibodies bind to the rat GBM only when complexed to nucleosomes, but not to antibodies alone [23]. Despite the importance of anti-nucleohistone antibodies in SLE suggested by a previous study [24], we failed to demonstrate increased cell proliferation or cytokine release by MeT-5A cells after binding to anti-nucleohistone antibodies. We also found that preincubating MeT-5A cells with nucleohistones did not result in any enhancement of binding of IgG from lupus patients. We speculate that the ineffectiveness of nucleohistone in the binding of IgG from SLE patients to MeT-5A was due to (i) lack of significant differences in the anti-nucleohistone levels between active SLE, inactive SLE and healthy controls groups, and (ii) the neutral pI of 6.9 in nucleosomes formed by combining histone (with the cationic N-terminal portion) and DNA of an anionic nature [25]. The cationic region of the nucleosome can be further neutralized after binding with anti-histone antibodies, which affect binding to negatively charged proteins on the cell surface.

Pleural mesothelial cells participate in acute pleural inflammation and the specific immunity is regulated through cell proliferation and the release of cytokines, including IL-1β, TGF-β1 and MCP-1 [13, 14]. We demonstrated that IgG from active lupus patients significantly induced the gene expression and protein synthesis of IL-1β, TGF-β1 and MCP-1 when compared with IgG from the inactive lupus or healthy control group. Moreover, the synthesis of these cytokines correlated well with the anti-dsDNA antibody titre. The purified IgG preparations used in our in vitro studies were free of complement or cytokines that could activate MeT-5A cells [26, 27]. Our findings suggest that a direct up-regulatory effect of anti-dsDNA antibodies on cytokine production is a possible pathogenetic mechanism of pleural mesothelium injury in SLE. The direct binding of anti-DNA antibodies to proteins on the cell surface membrane or nuclei leads to altered cell functions, including proliferation, apoptosis and protein synthesis [2830]. It is likely that the proliferation of MeT-5A cells is due to the direct binding of anti-DNA-containing IgG to DNA or negatively charged protein on the cell membrane. It has been shown that extracellular histone bound to perlecan stimulates myoblast proliferation [19]. Proteoglycans on the cell surface may mediate the internalization of histone and induce cell proliferation [19]. Nonetheless, whether similar mechanism(s) operate after binding of anti-DNA to MeT-5A cells remains to be explored.

In conclusion, we have provided in vitro evidence that the binding of autoantibodies from active lupus patients to MeT-5A cells is closely associated with disease activity. The binding of these antibodies to MeT-5A cells is mediated by DNA and histone. Anti-dsDNA antibodies exert a direct pathogenic effect in lupus pleuritis by stimulating the production of proinflammatory cytokines after binding to the pleural mesothelial cells.

This study was supported by the Mr and Mrs Liu Lit Ching Research Fund and H. Guo was supported by the Mrs Ivy Wu Fellowship.

References

1
Ebling FM, Hahn BH. Restricted subpopulations of DNA antibodies in kidneys of mice with systemic lupus. Comparison of antibodies in serum and renal eluates.
Arthritis Rheum
 
1980
;
23
:
392
–403.
2
Emlen W, Pisetsky DS, Taylor RP. Antibodies to DNA. A perspective.
Arthritis Rheum
 
1986
;
29
:
1417
–26.
3
Naylor B. Cytological aspects of pleural, peritoneal and pericardial fluid from patients with systemic lupus erythematosus.
Cytopathology
 
1992
;
3
:
1
–8.
4
Wang DY, Yang PC, Yu WL, Shiah DC, Kuo HW, Hsu NY. Comparison of different diagnostic methods for lupus pleuritis and pericarditis: a prospective three-year study.
J Formos Med Assoc
 
2000
;
99
:
375
–80.
5
Andrews BS, Arora NS, Shadforth MF, Goldberg SK, Davis JS 4th. The role of immune complex in the pathogenesis of pleura effusions.
Am Rev Respir Dis
 
1981
;
124
:
152
–61.
6
Riska H, Fyhrquist F, Selander RK, Hellstrom PE. Systemic erythematosus and DNA antibodies in pleural effusions.
Scand J Rheumatol
 
1978
;
7
:
159
–60.
7
Chan TM, Leung JK, Ho KN, Yung S. Mesangial cell-binding of anti-DNA antibodies in patients with systemic lupus erythematosus.
J Am Soc Nephrol
 
2002
;
13
:
1219
–29.
8
Chan TM, Yu PM, Tsang KLC, Cheng KP. Endothelial cell binding by human polyclonal anti-DNA antibodies: relationship to disease activity and endothelial functional alterations.
Clin Exp Immunol
 
1995
;
100
:
506
–13.
9
Lai KN, Leung JCK, Lai KB, Li PKT, Lai CKW. Anti-DNA autoantibodies stimulate the release of interleukin-1 and interleukin-6 from endothelial cells.
J Pathol
 
1996
;
178
:
451
–7.
10
Lai KN, Leung JC, Lai KB, Lai CK. Effect of anti-DNA autoantibodies on the gene expression of interleukin 8, transforming growth factor-beta, and nitric oxide synthase in cultured endothelial cells.
Scand J Rheumatol
 
1997
;
26
:
461
–7.
11
Chan TM, Frampton G, Staines NA, Hobby P, Perry GJ, Cameron JS. Different mechanisms by which anti-DNA MoAbs bind to human endothelial cells and glomerular mesangial cells.
Clin Exp Immunol
 
1992
;
88
:
68
–74.
12
Sun KH, Yu SJ, Tang SJ, Sun GH. Monoclonal anti-double-stranded DNA autoantibody stimulates the expression and release of IL-1β, IL-6, IL-8, IL-10 and TNF-α from normal human mononuclear cells involving in the lupus pathogenesis.
Immunology
 
2000
;
99
:
352
–60.
13
Antony VB, Mohammed KA. Cytokines and pleural disease. In: Nelson S, Martin TR, editors. Cytokines in pulmonary disease—infection and inflammation. New York: Marcel Dekker,
2000
:
117
–30.
14
Antony VB, Godbey SW, Sparks JA, Hott JW. Pleural mesothelial cells release a growth factor for fibroblasts.
Eur Respir Rev
 
1993
;
3
:
156
–8.
15
Godbey SW, Holm KA, Yu L, Hott JW, Panadero FR, Antony VB. Role of mesothelial cell in pleural fibrosis following successful talc poudrage: identification of basic fibroblast growth factor (FGF-2) in pleural fluids.
Am J Respir Crit Care Med
 
1995
;
151
:
A353
.
16
Tan EM, Cohen AS, Fries JF et al. The
1982
revised criteria for the classification of systemic lupus erythematosus.
Arthritis Rheum
 
1982
;
25
:
1271
–7.
17
Gussin HAE, Tselentis HN, Teodorescu M. Noncognate binding to histones of IgG from patients with idiopathic systemic lupus erythematosus.
Clin Immunol
 
2000
;
96
:
150
–61.
18
Minota S, Yoshio T, Iwamoto M, et al. Selective accumulation of anti-histone antibodies in glomeruli of lupus-prone lpr mice.
Clin Immunol Immunopathol
 
1996
;
80
:
82
–7.
19
Henriquez JP, Casar JC, Fuentealba L, Carey DJ, Brandan E. Extracellular matrix histone H1 binds to perlecan, is present in regenerating skeletal muscle and stimulates myoblast proliferation.
J Cell Sci
 
2002
;
115
:
2041
–51.
20
Burlingame RW, Rubin RL, Balderas RS. Genesis and evolution of antichromatin autoantibodies in murine lupus implicates T-dependent immunization with self antigen.
J Clin Invest
 
1993
;
91
:
1687
–96.
21
Amoura Z, Chabre H, Koutouzov S, et al. Nucleosome-restricted antibodies are detected before anti-dsDNA and/or anti-histone antibodies in serum of MRL-mp lpr/lpr and +/+ mice and are present in kidney eluate of lupus mice with proteinuria.
Arthritis Rheum
 
1994
;
37
:
1684
–8.
22
Coritsidis GN, Beers PC, Rumore P. Glomerular uptake of nucleosomes: evidence for receptor-mediated mesangial cell binding.
Kidney Int
 
1995
;
47
:
1258
–65.
23
Kramers C, Hylkema MN, van Bruggen MCJ, van de Lagemaat R, Dijkman HBPM, Assmann KJM. Anti-nucleosome antibodies complexed to nucleosomal antigens show anti-DNA reactivity and bind to rat glomerular basement membrane in vivo.
J Clin Invest
 
1994
;
94
:
568
–77.
24
Bruns A, Bläss S, Hausdorf G, Burmester GR, Hiepe F. Nucleosomes are major T and B cell autoantigens in systemic lupus erythematosus.
Arthritis Rheum
 
2000
;
43
:
2307
–15.
25
Berden JH, Licht R, van Bruggen MC, Tax WJ. Role of nucleosomes for induction and glomerular binding of autoantibodies in lupus nephritis.
Curr Opin Nephrol Hypertens
 
1999
;
8
:
299
–306.
26
Daniele RP. The pleura in local and systemic immune disorders. In: Chretien J, Bignon J, Hirsch A, eds. The pleura in health and disease. New York: Marcel Dekker,
1985
:
369
–83.
27
Adamson IY, Bakowska J, Prieditis H. Proliferation of rat pleural mesothelial cells in response to hepatocyte and keratinocyte growth factors.
Am J Respir Cell Mol Biol
 
2000
;
23
:
345
–9.
28
Raz E, Ben-Bassat H, Davidi T, Shlomai Z, Eilat D. Cross-reactions of anti-DNA autoantibodies with cell surface proteins.
Eur J Immunol
 
1993
;
23
:
383
–90.
29
Hsieh SC, Sun KH, Tsai CY, et al. Monoclonal anti-double stranded DNA antibody is a leucocyte-binding protein to up-regulate interleukin-8 gene expression and elicits apoptosis of normal human polymorphonuclear neutrophils.
Rheumatology
 
2001
;
40
:
851
–8.
30
Vlahakos D, Foster MH, Ucci AA, Barrett KJ, Datta SK, Madaio MP. Murine monoclonal anti-DNA antibodies penetrate cells, binding to nuclei, and induce glomerular proliferation and proteinuria in vivo.
J Am Soc Nephrol
 
1992
;
2
:
1345
–54.

Comments

0 Comments