The active form of MMP-3 is a marker of synovial inflammation and cartilage turnover in inflammatory joint diseases

All the reagents used in this study were standard high quality chemicals from Sigma (St.Louis, MO, USA) and Merck (Whitehouse Station, NJ, USA) unless specifically mentioned. All the peptides for monoclonal antibody development were a) immunogenic peptide: FRTFPGIPKW-GGC b) screening peptide: FRTFPGIPKW-biotin c) standard peptide: FRTFPGIPKW d) elongated peptide: HFRTFPGIPKW. All the peptides were purchased from the Chinese Peptide Company, China.

All the mice were specific pathogen free (SPF) animals and housed in SPF animal facility with 12 h light/dark cycle. The mice had free access to food and water. All the work on mice was approved by Beijing laboratory animal administration office and animal ethics committee of Nordic Bioscience (Beijing).

We used the first 10 amino acids of the N-terminal (¹⁰⁰’FRTFPGIPKW’¹⁰⁹) as the immunogenic peptide to generate specific neo-epitope monoclonal antibodies. The methods used for monoclonal antibody development were as previously described [23]. Briefly, six Balb/c mice (female, 4 to 6 weeks old) were immunized subcutaneously with 200 μl emulsified antigen and 60 μg of KLH conjugated immunogenic peptide. Consecutive immunizations were performed at two-week intervals in Freund's incomplete adjuvant, until stable sera titer levels were reached, and the mice were bled from the 3rd immunization on. At each bleeding, the serum titer was detected and the mouse with highest antiserum titer and the best native reactivity was selected for fusion. The selected mouse was rested for 1 month followed by intravenous boosting with 50 μg of KLH conjugated immunogenic peptide in 100 μl 0.9% sodium chloride solution 3 days before isolation of the spleen for cell fusion.

The fusion procedure has been described [24]. Briefly, the spleen cells from the immunized mouse with best antiserum titer and native reactivity were fused with SP2/0 myeloma fusion partner cells. The fusion cells were raised in 96-well plates and incubated in a 5% CO2 incubator. Here standard limited dilution was used to promote monoclonal growth. After seven to ten days of culture, supernatants were screened in a competitive ELISA setting. Cell lines specific to standard peptide and without cross-reactivity to elongated peptide were selected and sub-cloned. At last the antibodies were purified.

10 μg of Pro-MMP-3 (cat.no PF063, Calbiochem) was dissolved in 100 μL MMP buffer (100 mM Tris-HCl, 100 mM NaCl, 10 mM CaCl₂, 2 mM Zn acetate, pH 8.0). 1 μg pro-MMP-3 was mixed with 1.1 μL 10 mM APMA and incubated at 37°C for 3 hours.

Synovial membrane was obtained from total knee replacements of osteoarthritis patients at Gentofte Hospital, Gentofte, Denmark. The study was approved by the Ethics Committee of the Capital Region of Denmark, DK-3400 (approval no. HD-2007-0084). Patients were informed about the purpose of the study and provided written consent.

Synovial membrane was isolated during surgery and kept in DMEM + 10% FCS at 4°C until the next day, where experiments were initiated. Synovial membrane was washed 5 times in PBS to limit contamination and to remove excess blood. The synovial membrane was divided into equal pieces (explants) of about 30 mg and placed in a 96 well plate. Explants for the metabolic inactive (MI) group were inactivated by three freeze–thaw cycles. They were placed in a 37°C water bath for 5 min, then immediately transferred to liquid nitrogen and stayed for another 5 min, and repeated for 3 times. The explants was cultured in DMEM:F12 for 14 days, where medium was changed every second or third day. Conditioned medium was kept at -20°C until analysis.

The human cartilage explants (HEX) culture was performed as described previously [25]. Briefly, human cartilage was collected from cartilage replacement surgery. The study was approved by the Ethics Committee of the Capital Region of Denmark, DK-3400 (approval no. HD-2007-0084). The cartilage explants (16 ± 4 mg) were placed in 96-well plates and cultured at 37°C, 5% CO2 incubator. Each explant was cultured in 200 μL of DMEM for 21 days and divided into two groups: 1) without catabolic factors (W/O), 2) with the catabolic cytokines oncostatin M (10 ng/mL) and TNF-α (20 ng/mL) (O + T) to stimulate MMP activity. The culture medium was changed every 2-3 days and the supernatant was collected and stored in -20°C for further use.

Synovial membrane culture supernatant, HEX supernatant and act-MMP-3 were separated by SDS-PAGE, and proteins were transferred onto nitrocellulose membranes. The membranes were blocked with 5% skim milk powder in TBS-T buffer, and incubated in room temperature for 2 hours. The membranes were incubated with primary antibody at 4°C overnight, followed by three times wash in TBS-T buffer. Then the membranes were incubated in the secondary peroxidase conjugated antibody, and followed by three times wash. Finally, the results were visualized with ECL detection system (cat# RPN2109, Amersham Pharmacia). The antibody recognizes both pro and active form of MMP-3 was from Abcam (ab38916).

ELISA-plates used for the assay development were Streptavidin-coated from Roche cat.: 11940279. All ELISA plates were analyzed with ELISA reader from Molecular Devices, SpectraMax M, (CA, USA). The selected monoclonal antibody 7B2 was labeled with horseradish peroxidase (HRP) using the Lightning link HRP labeling kit according to the instructions of the manufacturer (Innovabioscience, Babraham, Cambridge, UK). A 96-well streptavidin plate was coated with screening peptide FRTFPGIPKW-biotin dissolved in coating buffer (50 mM Tris, 137 mM NaCl, 1% BSA, 0.05% Tween-20, 0.36% Bronidox L5, pH 8.0) and incubated 30 minutes at 20°C. 30 μL of standard peptide or sample dissolved in incubation buffer (50 mM Tris, 137 mM NaCl, 1% BSA, 0.05% Tween-20, 0.36% Bronidox L5, 5% liquid II, pH 8.0) were added to appropriate wells, and followed by 100 μL of conjugated antibody 7B2-HRP. Synovial membrane culture and HEX culture supernatant, serum samples measured in the ELISA were pre-diluted 1:2; act-MMP-3 and pro-MMP-3 inhibition test was performed in a serial dilution with the concentration from 1 ng/ul. Then the assay was allowed to incubate at 4°C for 20 ± 1 hours. Finally, 100 μL tetramethylbenzinidine (TMB sens) (Kem-En-Tec cat.4850E) was added and the plate was incubated 15 minutes at 20°C in the dark. All the above incubation steps included shaking at 300 rpm. After each incubation step the plate was washed five times in washing buffer (20 mM Tris, 50 mM NaCl, pH 7.2). The TMB reaction was stopped by adding 100 μL of stopping solution (0.1% H₂SO) and measured at 450 nm with 650 nm as the reference. A standard curve was performed by serial dilution of the standard peptide with the concentration of 0, 4.88, 9.77, 19.53, 39.06, 78.13, 156.25, 312.5, 625, 1250, 2500 and 5000 pg/mL. Calibration curve was plotted using a 4-parametric mathematical fit model.

The intra- and inter-assay variation was determined by 10 independent runs of 8 quality control (QC) samples consisting of serum, and each run consisting of double determinations of the samples. The lower limit of detection (LLOD) was determined from 21 zero samples (incubation buffer) and calculated as the mean 3x standard deviation. Dilution recovery was calculated as a percentage of recovery of diluted QC samples from the 100% sample. Spike recovery was calculated by comparing different concentrations of human cartilage explants supernatant in incubation buffer and in human serum. Finally, for each assay, a master calibrator prepared from synthetic peptides accurately quantified by amino acid analysis was used for calibration purposes.

Serum samples from 201 AS patients and 47 control RA patients were analyzed [26]. Of all the patients, 142 AS patients and 47 RA patients received 3 months anti-TNF-α treatment after recruitment. At baseline, the Bath AS Disease Activity Index (BASDAI) and the modified Stoke AS Spine Score (mSASSS) were recorded for all the AS patients, while DAS and HAQ were recorded for RA patients. CRP (C-reactive protein) concentration and ESR (erythrocyte sedimentation rate) were also recorded for both AS and RA patients. For those patients who have received anti-TNF-α drug, the characteristics mentioned above, except for mSASSS, were also recorded after treatment. In 118 AS patients, mSASSS were recorded after 2 years of followup. All the clinical characteristics used have previously been published by Bay-Jensen AC et al[26]. Serum samples were thawed from -80°C storage and act-MMP-3 concentrations were measured both at baseline and after treatment. The study was approved by local ethics committee and all patients provided the written informed consent.

Mean values and standard error of the mean (SEM) were calculated using GraphPad Prism, version 6 (GraphPad Software, San Diego, CA, USA). Act-MMP-3 differences between female and male were analyzed using non-parametric Mann–Whitney test. Correlations between act-MMP-3 and patient clinical characteristics (age, disease duration, CRP, ESR, BASDAI, mSASSS) were determined by Spearman’s test. Significant differences of clinical samples between pre-treatment and post-treatment were determined using the Wilcoxon matched-pairs signed rank test. Significant differences were considered if P < 0.05.

When measured in the competitive act-MMP-3 assay applying the 7B2 antibody, the antibody specifically recognized the N-terminal peptide ¹⁰⁰’FRTFPGIPKW’¹⁰⁹ and activated MMP-3 but did not recognize the elongated peptide ⁹⁹’HFRTFPGIPKW’¹⁰⁹ and pro-MMP-3 (Figure 1A,B). The results were confirmed by western blot of in vitro activated MMP-3. Bands were detected at 45 kD and 28 kD, but not at 52 kD, which is the size of the pro-MMP-3. All three bands were detected by the antibody recognizing the total MMP-3 antibody (Figure 1C). To ensure the specificity of the blot, primary antibody incubation was done including either the elongated peptide or the standard peptide. The signal was completely blocked only in the presence of the standard peptide.

The LLOD of the ELISA was 33.7 pg/mL. The average intra- and inter-assay variations were 3.1% and 13.5%, respectively. The dilution and spiking recovery of human serum were within 100 ± 20% (Table 1). The data confirms technically robustness of the assay.

We investigated the act-MMP-3 expression and release in human synovial membrane and HEX cultures, and as expected act-MMP-3 was expressed by human synovial membrane, compared to the negative control MI group (Figure 2A). The data was confirmed by western blotting, which again confirmed that the antibody only recognized the active form of MMP-3 (Figure 2B).

In HEX culture, stimulation with the pro-inflammatory cytokines TNF-α and oncostatin M resulted in the release of act-MMP-3 into culture medium starting from the early culture stage (see Additional file 1: Figure S1), supporting the pathological relevance of act-MMP-3 in inflammatory joint diseases.

We investigated the association between baseline act-MMP-3 levels and age, gender, disease duration in AS patients. The data showed that age and disease duration were not correlated to act-MMP-3 levels (R = -0.046 and R = -0.032, respectively). Act-MMP-3 levels in male and female patients were not significantly different. We also assessed the correlation between baseline act-MMP-3 levels, baseline mSASSS, baseline BASDAI, baseline CRP and baseline ESR in AS patients (Table 2). Act-MMP-3 levels were not correlated with the disease burden markers, mSASSS and BASDAI (R = 0.10 and R = 0.06, respectively), but were significantly correlated to CRP and ESR (R = 0.42 and R = 0.38, respectively, P < 0.001), which are markers of inflammation. Similar data were obtained in samples from RA patients, with baseline serum act-MMP-3 levels significantly correlated to CRP, but with no correlation with disease activity score (DAS) and health assessment questionnaire score (HAQ) (see Additional file 1: Table S1).

To evaluate if baseline act-MMP-3 level had any predictive value response to anti-TNF-α treatment, correlations between baseline act-MMP-3 level, BASDAI reduction (change < 0), CRP reduction (change < 0) and ESR reduction (change < 0) in AS patients were analyzed (Table 3). The data showed that baseline act-MMP-3 levels were significantly and positively correlated to the reduction in CRP and ESR (R = 0.44 and R = 0.39, respectively, P < 0.001). There were no significant correlations between baseline act-MMP-3 level and BASDAI reduction (R = 0.10). Comparable results were observed for the RA patients. Baseline act-MMP-3 level did not predict DAS and HAQ change, but were correlated to CRP and ESR changes (see Additional file 1: Table S2).

To investigate if act-MMP-3 level can predict 2-year radiographic progression, correlation between baseline act-MMP-3 level and 2-year mSASSS increase (change > 0) were analyzed. There was no significant correlation between baseline act-MMP-3 level and 2-year mSASSS change (R = 0.07).

To investigate if act-MMP-3 levels responded to anti-TNF-α treatment, we measured serum levels of act-MMP-3 in AS and RA patients at baseline and after anti-TNF-α treatment. A significant decrease in serum act-MMP-3 after treatment (Figure 3) was observed for both AS and RA, which showed the anti-TNF-α treatment, could effectively reduce act-MMP-3 levels, which correlates well with the previous findings that act-MMP-3 reflects the inflammatory status of the patients.