Elevated matrix metalloproteinase-9 in patients with systemic sclerosis

This study was conducted in accordance with the principles embodied in the Declaration of Helsinki and was approved by the Ethical Committees in the Catholic Research Institutes of Medical Sciences. Before the study, informed consent was obtained from all patients and healthy controls. Forty-two patients (1 man and 41 women), all of whom fulfilled the criteria of the American Rheumatism Association for SSc [24], were studied; their mean age was 43.7 years (range 24–69 years). The mean duration of disease was 80.8 months (range 5–276 months). The comparisons were made with 32 healthy controls (all women) who had no rheumatic disease; their mean age was 44.2 years (range 21–62 years). The ages and sexes of the patient and control groups did not differ significantly.

Clinical and laboratory assessments were done at the time of sampling. The clinical variables were age, sex, disease duration, type of SSc [25], modified Rodan score [26], presence or absence of esophageal involvement on endoscopy and esophageal manometry, interstitial lung disease on chest radioagrapy and/or high-resolution computerized tomography, diffusion capacity (DLCO; diffusion of carbon monoxide in the lung) on the pulmonary function test, arthritis, sicca syndrome, and antibodies to Scl-70 or centromere using ELISA kits (MBL, Nagoya, Japan). Interstitial lung disease was defined as bibasilar interstitial fibrosis on chest radiographs, or, in patients with no abnormalities on chest radiographs, as the presence of alveolitis on high-resolution computerized tomography.

The total MMP-9 and TIMP-1 concentrations were determined in the serum and the culture supernatant using a commercial ELISA kit (R&D Systems Inc, Minneapolis, MN, USA). In accordance with the manufacturer's recommendations, the aliquots of serum were diluted to a ratio of 1:100 in the assay buffer. The detection limits of the MMP-9 and TIMP-1 kits were 0.15 ng/ml and 0.08 ng/ml, respectively. The MMP-9 assay kit detected pro-MMP-9 and complexes of pro-MMP-9 with TIMP-1 and had no significant cross-reactivity with MMP-1, MMP-2, MMP-3, TIMP-1, or TIMP-2. Again, the TIMP-1 detection kit detected TIMP-1 either free or in complex with some MMPs and showed no cross-reactivity or interference with TIMP-2.

Circulating TGFβ was measured in the same samples using ELISA, as described previously [27]. Briefly, 2 μg/ml of monoclonal antibodies to TGFβ1, β2, and β3 (R&D Systems) were added to 96-well plates (Nunc Inc, Roskilde, Denmark). They were incubated overnight at 4°C and blocked with PBS containing 1% bovine serum albumin and 0.05% Tween 20 for 2 hours at room temperature. A sample (50 μl) of each patient's serum was diluted 1:2 with PBS, acidified with 50 μl of 2.5 M acetic acid and 10 M urea for 10 minutes at room temperature and then was neutralized with 50 μl of 2.7 M NaOH and 1 M HEPES. The patient's sera and the standard recombinant TGFβ (R&D Systems) were then put into 96-well plates and incubated at room temperature for 2 hours. Biotinylated polyclonal antibodies (50 ng/ml) to human TGFβ (R&D Systems) were added and the reaction was allowed to proceed for 2 hours at room temperature. Streptavidin–alkaline phosphatase (Sigma Bioscience, St Louis, MO, USA) diluted 1:2000 with PBS was added, and the reaction was again allowed to proceed for 2 hours. p-Nitrophenylphosphate (1 mg/ml) (Sigma Bioscience) dissolved in diethanolamine (Sigma Bioscience) was added to induce a color reaction, and 1 N NaOH (Fisher Scientific, Pittsburgh, PA, USA) was used to stop the reaction. An automated microplate reader (Vmax, Molecular Devices, Palo Alto, CA, USA) was used to measure the optical density at 405 nm. Between each of these steps, the plates were washed four times with PBS containing 0.05% Tween 20. A standard curve was drawn by plotting the optical density versus the log of the recombinant TGFβ concentration. The detection limit for TGFβ was 30 pg/ml.

MMP-9 and MMP-2 activities were also tested by gelatin zymography. A 0.5-μl sample of serum diluted in 30 μl of SDS buffer was separated in 10% SDS–PAGE gel polymerized with 1 mg/ml gelatin (Invitrogen Life Technologies, Carlsbad, CA, USA). Culture supernatants of HT1080 cell lines (malignant human fibroblasts) stimulated with 10 μg/ml of concanavalin A were used as a positive control. Gels were washed once for 3 hours in 2.5% Triton X-100 to remove the SDS and once for 30 minutes in the reaction buffer containing 50 mM Tris/HCl, 200 mM NaCl, 10 mM CaCl₂, and 0.02% (w/v) Brij 35 (pH 7.5). The reaction buffer was changed to a fresh one, and the gels were incubated at 37°C for 24 hours. Gelatinolytic activity was visualized by staining the gels with 0.5% Coomassie brilliant blue and was quantified by densitometry.

Dermal fibroblasts were obtained from affected skin of two SSc patients and from two healthy controls, as described previously [28]. Fibroblasts were grown from explants in Dulbecco's modified Eagle's medium (DMEM) at 37°C in 5% CO₂. The cells were then centrifuged at 500 g, resuspended in DMEM supplemented with 10% fetal calf serum (Gibco-BRL, Grand Island, NY, USA), 2 mM glutamine, penicillin (100 U/ml), and streptomycin (100 μg/ml), and plated in 25-cm² flasks. The cultures were kept at 37°C in 5% CO₂ and the culture medium was replaced every 3 days. When cells approached confluence, they were detached with trypsin, passed after dilution 1:3 with fresh medium, and recultured until use. Cells were housed in a 37°C humidified incubator with 5% CO₂. Second- or third-passage cells were used for all experiments. Fibroblasts were seeded in 24-well plates at 5 × 10⁴ cells per well in serum-free DMEM supplemented with insulin–transferrin–selenium A (ITSA; Gibco BRL). After the cells had been grown in selected medium alone for 12 hours, we added cytokines – IL-1β (10 ng/ml), tumor necrosis factor α (TNF-α) (10 ng/ml), and TGFβ (10 ng/ml) – to stimulate the fibroblasts. After 24 hours of incubation, cell-free media were collected and stored at -20°C until assay. All cultures were set up in triplicate or quadruplicate.

The serum concentrations of MMP-9 were significantly higher in patients with SSc (n = 42) than in healthy controls (n = 32) (317.6 ± 33.5 ng/ml versus 81.2 ± 6.8 ng/ml, P < 0.001) (Fig. 1). The serum concentration of TIMP-1, an inhibitor of MMP-9, was also higher in SSc patients than in healthy controls (157.1 ± 13.2 ng/ml versus 77.7 ± 12.5 ng/ml, P < 0.001), but SSc patients had higher MMP-9/TIMP-1 ratios than healthy controls (233.0 ± 27.1 versus 69.5 ± 24.3, P < 0.001). There was no correlation between MMP-9 and TIMP-1 concentrations in SSc patients or in healthy controls. SSc patients with the diffuse type (n = 23) and had higher concentrations of MMP-9 than those with the limited type (n = 19) (364.6 ± 32.4 ng/ml versus 260.0 ± 34.6 ng/ml, P = 0.034) (Fig. 2). No significant differences were found between the two groups of patients with regard to age, sex, disease duration, and prednisolone usage or the kinds of immunosuppressive agents being used (e.g. D-penicillamine and cyclosphosphamide) (Table 1).

We used gel zymography to study sera of 20 SSc patients and 10 healthy controls, all selected unsystematically, to ascertain the serum gelatinase activity of MMP-9. As can be seen in Fig. 3, the 92 kDa band, consistent with the latent form of MMP-9, was detected in the sera of all subjects. The bands in Fig. 3 represent the latent form of MMP-9 (92 kDa, upper band) and the latent form of MMP-2 (72 kDa, lower band). The serum MMP-9 activities of SSc patients were higher than those of healthy controls. Densitometric analysis in sera of 20 SSc patients and 10 healthy controls indicated that the mean MMP-9 activity for SSc patients was 137.2 ± 21.7 densitometry units and for healthy controls, 38.5 ± 4.2 densitometry units (P < 0.001). Furthermore, a good linear correlation was found between the densitometry units measured by zymogram and the respective concentrations of MMP-9 measured by immunoassay in the sera of SSc patients (r = 0.875 and P < 0.001; data not shown). However, the intensity of the 86 kDa band (active MMP-9) was generally weak and was often not measurable.

To determine the association of MMP-9 concentrations with a definite clinical manifestation of SSc, we compared the serum MMP-9 concentrations with clinical and laboratory characteristics in patients (n = 35) with SSc. The patients with severe skin involvement (n = 18), defined by a modified Rodnan score ≥20, had significantly higher concentrations of circulating MMP-9 than those with mild to moderate skin involvement (n = 17) (modified Rodnan score <20) (Table 2). Moreover, the serum MMP-9 concentrations correlated well with the Rodnan scores (n = 35, r = 0.425, P = 0.011) and with the serum TGFβ concentrations (n = 41, r = 0.736, P < 0.001) (Fig. 4a,4b). However, a correlation between MMP-9 and TGFβ was not found in the sera from healthy controls (data not shown). There were no differences in the MMP-9 concentrations with respect to age, the presence of esophageal involvement, interstitial lung disease, decrease of diffusion capacity (DLCO < 70%), digital ulcer, arthritis, sicca syndrome, and antibodies to Scl-70 or centromere-B (Table 2).

The finding that MMP-9 concentrations correlated with skin scores prompted us to investigate the in vitro MMP-9 production by dermal fibroblasts from SSc patients. The spontaneous MMP-9 concentrations in the culture supernatants of dermal fibroblasts were not greatly different between SSc patients and healthy controls (Fig. 5). However, stimulation of SSc fibroblasts with IL-1β, TNF-α, or TGFβ strongly increased MMP-9 production relative to the unstimulated concentration, by factors of 3.5, 3.2, and 2.3, respectively, whereas fibroblasts of healthy controls responded weakly to these cytokines (by factors of 1.6, 1.5, and 1.2, respectively). The increase in MMP-9 production by IL-1β and TNF-α appears to be triggered at least in part by a cyclosporin A (CsA)-sensitive pathway, since 500 ng/ml CsA limited MMP-9 production in SSc fibroblasts stimulated with IL-1β or TNF-α to 63% and 57% of original responses, respectively.