Patients characteristics

Our study consisted of 42 patients with inflammatory myopathies and 27 healthy controls. All patients underwent muscle biopsy that was guided by positive magnetic resonance imaging (MRI) findings from affected muscles [12]. The specific pattern of muscle biopsy, including the immunohistological [13, 14] and clinical investigation, showed that 17 patients suffered from DM and 25 from other types of myositis. Based on the novel clinicoserological criteria [15], four patients with myositis could be classified as pure PM, while 21 had overlap myositis. These patients had at least one clinical overlap feature and/or an overlap antibody [15].

Patients were recruited from the inpatient and outpatient departments of the Institute of Rheumatology in Prague. Disease activity was assessed using myositis disease activity assessment visual analogue scales (MYOACT) that globally score constitutional, articular, cardiac, pulmonary, gastrointestinal, cutaneous, muscle organs or systems [16]. Manual muscle testing of eight muscle groups (MMT8) was performed and included one axial, five proximal (two upper extremity, three lower extremity), and two distal muscles (one upper, one lower extremity) [17]. Written informed consent from each participant was obtained prior to enrolment, and the study was approved by the local ethics committee.

Laboratory measurements

Peripheral blood was obtained from all patients at the time of clinical assessment and from healthy donors. A routine laboratory analysis was performed on fresh serum of patients with inflammatory myopathies. C-reactive protein (CRP) levels in patients and healthy controls were determined by an immuno-turbidimetric technique using an Olympus biochemical analyser (model AU 400, Tokyo, Japan). Creatine kinase and myoglobin levels were measured by routine laboratory methodology using Olympus analyser. All collected serum was stored at -80°C until further analysis. Serum resistin levels were measured with commercially available ELISA according to the manufacturer's protocol (Biovendor, Brno, Czech Republic). The levels of IL-1β, IL-6, TNFα and MCP-1 (RayBiotech, Norcross, GA, USA), along with the levels of perforin (Abcam, Cambridge, UK) and granzyme (Abcam) in cell culture supernatants were measured using commercially available ELISA kits. The absorbance was measured at 450 nm using an ELISA reader (Tecan Sunrise, Salzburg, Austria).

Immunohistochemistry

Samples for immunohistochemistry were obtained from patients with DM (n = 5) and myositis (n = 5) at the time of diagnostic muscle biopsy. In these patients, the muscle biopsy was guided by positive magnetic resonance imaging (MRI) findings from affected muscle as previously described [12]. Muscle tissue samples from patients with myasthenia gravis (MG, n = 5) were used as non-inflammatory controls. Muscle tissues were snap frozen in isopentane (2-ethylbutane, Sigma-Aldrich, St. Louis, MO, USA) cooled in liquid nitrogen. For the purposes of immunohistochemistry, 5-μm frozen sections were fixed in acetone and 4% paraformaldehyde and blocked with 0.3% H₂O₂. The sections were incubated in tris buffered saline (TBS) buffer and incubated with a monoclonal anti-resistin antibody (Phoenix Pharmaceuticals, Inc., Burlingame, CA, USA) at a dilution 1:500 for one hour. Afterwards, the slides were rinsed again in TBS buffer. Antigen-antibody complexes were visualised with a Histofine detection system (Nichirei Biosciences Inc., Tokyo, Japan) using 3, 3'diaminobensidine as a chromogene. Rabbit IgG (Dako Cytomation, Glostrup, Denmark), diluted 1:1,000, was used as a negative control. The sections were slightly counterstained with Harris' hematoxylin. All sections were analysed semiquantitatively by two experienced pathologists who were blinded to the clinical data. The intensity of resistin expression was scored on a four-point scale (0 to 3). A staining intensity of 0 represented a negative result, and scores of 1 to 3 represented weak, moderate and strong positive staining intensity, respectively.

Cell culture and stimulation assays

Human skeletal muscle cells (Lonza, Basel, Switzerland) were cultured at a density of 100,000 cells in a six-well plate in 2 ml DMEM (Gibco, Carlsbad, CA, USA). PBMC were isolated by standard Ficoll density gradient centrifugation from blood samples donated by healthy donors. Freshly isolated PBMC were resuspended at a density of 10⁶ cells per well in a 12-well plate in 2 ml of RPMI 1640 (Gibco, Carlsbad, CA, USA). Both myocytes and PBMC were stimulated with human recombinant resistin (Biovendor, Modrice, Czech Republic) at a concentration of 10, 100 or 1,000 ng/ml for 6 and 48 hours at 37°C in 5% CO₂ humidified atmosphere. In the initial experiment, PBMC were stimulated with either resistin (1,000 ng/μl) alone or in combination with polymyxin B sulphate (5 ug/ml, Sigma-Aldrich, St. Louis, MO, USA) to exclude possible endotoxin contamination. For the gene expression analysis, cells were lysed in RLT buffer (Qiagen, Hilden, Germany) 6 h after stimulation. The supernatants were collected following 48 h of stimulation. The samples were stored at -80°C until use.

RNA isolation and TaqMan Real-Time PCR

Total RNA was isolated using a MagNA Pure Compact RNA Isolation Kit for the MagNA Pure Compact Instrument (Roche Diagnostics, Mannheim, Germany), and reverse transcription was performed with a High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA). Real-time PCR was performed using gene expression assays (Applied Biosystems), and the reaction was performed using a 7900HT Fast Real-Time PCR System (Applied Biosystems). Data were analysed using the ddCt method for relative quantification, and 18S was used as an endogenous control.

MTT assay

The evaluation of the proliferation of myocytes after 72 h stimulation with resistin (10, 100, 1,000 ng/ml) was carried out with an MTT test using dimethylthiazol diphenyl tetrazolium bromide (Sigma-Aldrich). The absorbance was measured at 570 nm via ELISA reader (Tecan Sunrise, Salzburg, Austria).

Statistical analysis

Data are expressed as the mean ± SD or mean ± SEM where indicated. The Mann-Whitney U test was used for comparisons between two variables, and the Kruskal-Wallis test, along with Dunn's multiple comparison tests, was used for comparisons among more than two variables. Paired T-tests were used for statistical analyses of gene expression. Spearman's and Pearson correlation coefficients were used to correlate any two variables with non-normal and normal distribution, respectively. P-values less than 0.05 were considered statistically significant. The analysis and the graphs were performed using GraphPad Prism 5 (version 5.02; GraphPad Software, La Jolla, CA, USA).

Characteristics of patients

Increased resistin levels and disease activity in patients with inflammatory myopathies

The serum resistin levels were significantly higher in patients with inflammatory myopathies than in healthy controls (8.53 ± 6.84 ng/ml vs. 4.54 ± 1.08 ng/ml, P < 0.0001, Figure 1). Patients with both DM (7.39 ± 3.75 ng/ml, P < 0.01) and overlap myositis (9.65 ± 8.88 ng/ml, P < 0.001) had significantly higher levels of resistin compared to healthy controls. There were no significant differences in the levels of serum resistin between patients with DM and other types of myositis (7.39 ± 3.75 vs. 9.31 ± 8.30 ng/ml, P = 0.626) or between anti-Jo-1 positive and anti-Jo-1 negative patients (9.459 ± 5.948 vs. 8.242 ± 7.153, P = 0.281). The serum levels of resistin were comparable between male and female patients (7.160 ± 3.363 vs. 7.864 ± 4.459, P = 0.761) and were not affected by age (r = -0.154, P = 0.331) or disease duration (r = -0.104, P = 0.513). Furthermore, resistin serum levels did not differ significantly between patients treated with glucocorticoids for less and more than one month, and treatment with other immunosuppressive agents also did not affect the serum resistin levels. The resistin levels were not significantly different between patients with and without arthritis or patients with and without interstitial lung disease although these compartments may be a potential source of resistin.

In patients with inflammatory myopathies, but not in healthy controls, the serum resistin levels correlated with the CRP levels (r = 0.328, P = 0.040) and, interestingly, positively correlated with the global MYOACT score (r = 0.382, P = 0.026). In contrast, resistin levels correlated neither with the creatine kinase and myoglobin levels nor with the MMT8 score in patients with inflammatory myopathies. Furthermore, we found that resistin serum levels significantly correlated with CRP levels (r = 0.717, P = 0.037) and MYOACT (r = 0.798, P = 0.007) in anti-Jo-1 positive patients, but not in anti-Jo-1 negative patients (CRP r = 0.140, P = 0.469; MYOACT r = 0.283, P = 0.181). In addition, there was a trend towards correlation between serum resistin levels and myoglobin (r = 0.650, P = 0.067) in anti-Jo-1 positive patients.

Expression of resistin in muscle tissue

The assessment of resistin staining intensity in inflammatory infiltrates of the sections from muscle biopsies revealed a mean intensity of 1.8 in myositis and 1.6 in DM patients on the 0 to 3 scale. As expected, no lymphocytes were present in control individuals with MG. Therefore, muscle tissue samples from MG patients were used as non-inflammatory controls. We found that resistin expression in muscle tissues from patients with inflammatory myopathies was significantly higher when compared with control muscle tissues from patients with MG (Figure 2). The increased expression of resistin was particularly localised in scattered mononuclear cells and in mononuclear cells in inflammatory infiltrates surrounding large vessels and muscle fibres in patients with inflammatory myopathies. Mild resistin expression was observed in the cytoplasm of some muscle fibres. Resistin staining was observed in regenerating muscle fibres in four-fifths of myositis and three-fifths of DM patients, but not in control patients with MG (Figure 2). Identification of regenerating muscle fibers was based upon hematoxylin and eosin staining. Regenerating muscle fibers were smaller and splitting, with centrally located vesicular myonuclei. The cytoplasm of regenerating muscle fibers appeared basophilic.

Effects of resistin on mononuclear cells

PBMC (n = 8) were stimulated with increasing concentrations of resistin (0, 10, 100, 1,000 ng/ml) for 6 or 48 h. As shown in Figure 3, treatment of PBMC with resistin resulted in a significant and dose-dependent induction of IL-1β (up to 6.5-fold), IL-6 (up to 203-fold) and MCP-1 (up to 10.8-fold). The induction of TNFα at the mRNA level was only mild (up to two-fold). This expression pattern was followed by the increased cytokine release into the cell culture supernatants (Figure 4a-d). Cytokine levels after stimulation with the lowest doses of resistin were comparable with those of unstimulated cells but these values were significantly different after stimulation with higher doses of resistin. To study the effect of resistin on the cytotoxic activity of PBMC, the perforin and granzyme levels were analysed in cell culture supernatants. Only the highest concentrations of resistin resulted in a reduction in the release of both perforin and granzyme (Figure 4e, f).

To exclude endotoxin contamination of human recombinant resistin, PBMC (n = 3) were stimulated with resistin (1,000 ng/ml) alone or in combination with polymyxin B sulphate (5 μg/ml) for 6 h. There were no differences in the expression of the abovementioned cytokines, indicating that there was no nonspecific effect mediated by recombinant resistin (data not shown).

Effects of resistin on myocytes

To assess the ability of resistin to induce a possible inflammatory response in skeletal muscle cells, as it does in mononuclear cells, myocytes (n = 9) were stimulated with increasing concentrations of resistin (0, 10, 100, 1,000 ng/ml). Although the treatment of myocytes with resistin for 6 hours contributed to the slightly increased gene expression of some pro-inflammatory cytokines (Figure 3), the release of these cytokines into the cell culture supernatants after 48 hours from stimulated and unstimulated cells did not significantly differ. There was also no resistin-associated effect of on the expression of several type I interferon induced genes, such as CXCL10, IFI27, IFI44, IFI44L, RSAD2, OAS1, ISG15, IFIT1, MX1 (data not shown). Furthermore, as assessed by MTT assay, resistin did not affect proliferation of skeletal muscle cells in vitro.