Abstract

Objective. We investigated SF and serum proteomic fingerprints of patients suffering from RA, OA and other miscellaneous inflammatory arthritides (MIAs) in order to identify RA-specific biomarkers.

Methods. SF profiles of 65 patients and serum profiles of 31 patients were studied by surface-enhanced laser desorption and ionization–time-of-flight-mass spectrometry technology. The most discriminating RA biomarkers were identified by matrix-assisted laser desorption ionization–time of flight and their overexpression was confirmed by western blotting and ELISA.

Results. Three biomarkers of 10 839, 10 445 and 13 338 Da, characterized as S100A8, S100A12 and S100A9 proteins, were the most up-regulated proteins in RA SF. Their expression was about 10-fold higher in RA SF vs OA SF. S100A8 exhibited a sensitivity of 82% and a specificity of 69% in discriminating RA from other MIAs, whereas S100A12 displayed a sensitivity of 79% and a specificity of 64%. Three peptides of 3351, 3423 and 3465 Da, corresponding to the α-defensins-1, -2 and -3, were also shown to differentiate RA from other MIAs with weaker sensitivity and specificity. Levels of S100A12, S100A8 and S100A9 were statistically correlated with the neutrophil count in MIA SF but not in the SF of RA patients. S100A8, S100A9, S100A12 and α-defensin expression in serum was not different in the three populations.

Conclusion. The most enhanced proteins in RA SF, the S100A8, S100A9 and S00A12 proteins, distinguished RA from MIA with high accuracy. Possible implication of resident cells in this increase may play a role in RA physiopathology.

Introduction

RA is the most common inflammatory arthritis, with a prevalence range in European countries of between 0.3 and 0.8% [1]. Mechanisms of synovium inflammation are still unclear and often lead to progressive joint destruction and deformation. Cells of the adaptative immunity pathway were thought to be the principal actors of RA pathogenesis. However, an increasing body of evidence suggests that cells of the innate immunity and resident articular cells are involved in RA pathogenesis [2, 3]. Indeed, the activation of synovial macrophages or fibroblast-like synoviocytes may trigger lymphocytes and subsequently synovium inflammation since inactivating these cells reduces arthritis in experimental models [4]. It has been hypothesized that several cytosolic peptides, when released in the extracellular matrix during cell necrosis, could be recognized by cells of the innate immunity pathway as danger signals [3]. Among them, S100 proteins could be implicated in RA pathogenesis. The S100A8, S100A9 and S100A12 proteins are predominantly expressed by myelomonocytic cells and constitute ∼40% of the polymorphonuclear neutrophil (PMN) cytosolic proteins. They have a critical role in inflammation as they can activate the innate immunity pathway sensed by receptors for advanced glycation end products (RAGEs) or Toll-like receptors [5–7]. S100A8 and S100A9 are also expressed by keratinocytes and epithelial cells under inflammatory conditions [8].

Several reasons prompted us to perform SF proteome profiling using surface-enhanced laser desorption and ionization–time-of-flight mass spectrometry (SELDI–TOF-MS) [9, 10]. First, the currently available biological tests are not precocious and accurate enough, leading to delayed diagnosis of the disease [11]. The discovery of new biomarkers is of particular interest for earlier diagnosis. Secondly, SELDI–TOF-MS technology, whose relevance has been emphasized not only in cancer [12–14] but also in the rheumatology field [15–18], is likely to detect numerous RA-specific biomarkers. Thirdly, since the synovium–cartilage junction is a critical zone of the disease development [19], disease-specific proteins may localize into SF rather than in other biological materials. Thus, we decided to investigate the SF and the serum proteomic fingerprint of patients suffering from RA, OA and miscellaneous inflammatory arthritides (MIAs) in order to identify RA-specific biomarkers.

Patients and methods

Patient populations

Seventy-four patients were recruited in the rheumatology clinic between March 2001 and 2007. The study was approved by the local ethics committee (Comité de Protection des Personnes). Sixty-five SFs and 31 sera were obtained during therapeutic or diagnostic procedures, after patients gave their informed consent to the study. The RA group comprised 30 patients for the SF analysis and 14 patients for the serum analysis, fulfilling the 1987 modified ACR criteria for RA [20]. The 1986 clinical and radiographic criteria were fulfilled by 24 patients of the OA group (17 SF and 7 sera) [21]. The inflammation control group was composed of 18 patients (nine AS, six pseudogout, one SLE and two juvenile idiopathic arthritis) for the SF analysis and 10 patients (eight AS, one pseudogout and one SLE) for the serum study. All patients suffering from AS, juvenile idiopathic arthritis and SLE fulfilled, respectively, the modified New York criteria [22], ILAR criteria [23] and the 1982 ACR criteria [24]. The SF of patients suffering from pseudogout displayed a leucocyte count over 2000/mm3 and CP crystals. Demographic parameters are summarized in Table 1.

Table 1

Demographic and biological characteristics of the population

 RA (n = 44) OA (n = 24) MIA (n = 28) P-value 
Age, years 55.0 (44.0–62.0) 71.0 (62.0–78.0) 52.0 (41.0–67.0) 0.001 
Disease duration, year 8.5 (2.8–12.5) 0.5 (0.1–5.0) 4.0 (0.9–7.0) 0.885 
RF, IU/ml 19.5 (10.5–81.0) NA 6.0 (3.5–31.3) 0.035 
ESR, mm/h 39.0 (26.0–67.0) 20.0 (8.5–46.5) 33.5 (8.3–73.0) 0.018 
CRP, mg/l 30.5 (12.0–49.5) 5.0 (5.0–18.0) 60.0 (24.0–146.0) 0.049 
SF count, cells/mm3     
    Leucocytes 10 000 (8125–18 000) 200 (120–290) 7000 (5000–10 000) <0.001 
    PMN 6885 (4620–13320) 67 (17–80) 5355 (2100–9685) <0.001 
    Lymphocytes 1260 (480–1822) 48 (35–80) 560 (2–845) <0.001 
Treatment, n (%)  NA   
    Biologicals 19 (43.2)  3 (10.7) 0.024 
    DMARDs 21 (47.7)  8 (28.6) 0.243 
    Steroids 8 (18.1)  2 (7.1) 0.012 
 RA (n = 44) OA (n = 24) MIA (n = 28) P-value 
Age, years 55.0 (44.0–62.0) 71.0 (62.0–78.0) 52.0 (41.0–67.0) 0.001 
Disease duration, year 8.5 (2.8–12.5) 0.5 (0.1–5.0) 4.0 (0.9–7.0) 0.885 
RF, IU/ml 19.5 (10.5–81.0) NA 6.0 (3.5–31.3) 0.035 
ESR, mm/h 39.0 (26.0–67.0) 20.0 (8.5–46.5) 33.5 (8.3–73.0) 0.018 
CRP, mg/l 30.5 (12.0–49.5) 5.0 (5.0–18.0) 60.0 (24.0–146.0) 0.049 
SF count, cells/mm3     
    Leucocytes 10 000 (8125–18 000) 200 (120–290) 7000 (5000–10 000) <0.001 
    PMN 6885 (4620–13320) 67 (17–80) 5355 (2100–9685) <0.001 
    Lymphocytes 1260 (480–1822) 48 (35–80) 560 (2–845) <0.001 
Treatment, n (%)  NA   
    Biologicals 19 (43.2)  3 (10.7) 0.024 
    DMARDs 21 (47.7)  8 (28.6) 0.243 
    Steroids 8 (18.1)  2 (7.1) 0.012 

Values are presented as median (interquartile range) except for treatment where values are numerical (%). Statistical significance was calculated using ANOVA or χ2-test where appropriate. RF assessed by Waaler–Rose technique (n < 15 IU/ml); ESR (n < 20 mm); CRP (n < 10 mg/l); NA: not applicable.

Sample processing

Sera were treated by a protease inhibitor cocktail (Roche Diagnostics, Meylan, France), and were subsequently aliquoted and stored at −80°C. SFs were centrifuged [500 g, 10 min, room temperature (RT)], supernatants were incubated with hyaluronidase (40 IU/ml of SF) for 30 min at RT, centrifuged again (1800 g, 10 min, RT), treated with the same protease inhibitor cocktail and aliquoted at −80°C.

SELDI–TOF-MS analysis

Serum and SF protein profiles were obtained by the SELDI–TOF-MS method (Ciphergen Biosystems, Fremont, CA, USA), using a cation exchange array (CM10), an anion exchange array (Q10), a nickel affinity array [immobilized metal affinity capture (IMAC-Ni)] and a copper affinity array (IMAC-Cu).

As SF volume varied considerably among patients, a constant amount of protein in each sample was analysed in order to rule out the influence of protein concentration on the results. Protein concentration of SF was measured using the Bradford method: 5 µg were used for the proteome profiling on CM10, IMAC-Cu and IMAC-Ni, whereas 2 µg were loaded on Q10 arrays. Ten microlitres of serum were mixed with 15 µl of a denaturation buffer [phosphate-buffered saline (PBS), 9 M urea, 1% 3[(3-Cholamidopropyl)dimethylammonio]-propanesulfonic acid (CHAPS) pH 7.3] under constant motion for 20 min and 2 µl of this solution were loaded on each array.

Samples were processed robotically on a Biomek 2000 Automation Workstation (Beckman Coulter, Villepinte, France). IMAC-Cu and IMAC-Ni arrays were activated with, respectively, 100 µl of 100 mM CuSO4 or 100 mM NiSO4. Then spots were washed twice with 200 µl H2O, once with 200 µl sodium acetate pH 4 and finally with 200 µl H2O. Then the following procedure was applied to each chip. Arrays were equilibrated with 200 µl of the corresponding loading/washing buffer [20 mM Tris, 0.1% Triton X-100 (v/v) pH 8 for Q10; 20 mM sodium acetate, 0.1% Triton X-100 pH 4 for CM10; PBS, 0.5 M NaCl, 0.1% Triton X-100 pH 7.3 for IMAC-Cu and IMAC-Ni]. Respective amounts of SF or serum were incubated in 50 µl of loading/washing buffer for 60 min. Spots were washed twice with 200 µl of washing buffer, twice with 200 µl of the same buffer without Triton X-100 and finally with 200 µl of 5 mM Tris pH 8 (Q10), 5 mM sodium acetate pH 4 (CM10) or 5 mM HEPES pH 7 (IMAC-Cu and IMAC-Ni). After drying, spots were loaded twice with 0.8 µl of saturated sinapinic acid in 0.5% trifluoroacetic acid, 50% acetonitrile and chips were analysed on the PCS-4000 ProteinChip reader. Two laser intensities (3200 and 4500 nJ) were used to analyse protein profiles and mass data processing was performed with the Ciphergen Express software, following the classical steps of spectra calibration, normalization, peak detection and aligning.

Protein purification and identification

S100A8, S100A9 and S100A12 proteins were purified from 500 µl of RA SF by protein precipitation with 70% ammonium sulphate. After centrifugation (10 000 g, 30 min, 4°C), the supernatant was dialysed in 5 mM Tris pH 7.5 in Spectra/Por dialysis membranes (molecular weight cut off: 5000 Da) and proteins were lyophilized in a freeze–dry system. The powder was suspended in 1 ml H2O; 900 µg of protein evaluated by the Bradford method were loaded on a 16% acrylamide tricine–SDS–PAGE gel, and the gel was further stained with Coomassie blue R-250 [25]. Bands revealed at about 10 800, 10 400 and 13 300 Da were digested by trypsin and identified by MS using a 4800 MALDI–TOF/TOF analyser. Database searches were performed in the UniProtKB database (release 12.0) using the Mascot (Matrix Science) search algorithm. Validity of the identification results was evaluated by the molecular weighted search (MOWSE) score.

Immunoblotting

SF (300 µg) of patients suffering from RA, OA and AS were loaded on a 12.5% acrylamide SDS–PAGE gel. Recombinant S100A8 and S100A9 (1 µg) were used as positive controls [26, 27]. After electrophoresis, proteins were transferred onto a 0.45 mm nitrocellulose membrane. The membrane was further incubated with home-made rabbit anti-calprotectin (S100A8/S100A9 hetero-complex) antibody (1: 1000 dilution). Anti-rabbit antibodies conjugated to horseradish peroxidase (1: 5000 dilution) were added the next day and revelation was performed with ECL Reagent (GE Healthcare, Little Chalfont, UK).

α-Defensin concentration

Alpha-defensin concentration was evaluated in SF by an ELISA kit according to the manufacturer’s instructions (HyCult Biotechnology, Uden, The Netherlands) [28].

Statistical analysis

Differences in demographic and biological parameters between groups were explored using ANOVA or χ2-test, where appropriate. Mann–Whitney U-test or Kruskall–Wallis test was used to compare SELDI–TOF-MS peak intensities and α-defensin concentration assessed by ELISA (Statview 5.0; SAS Institute, Marlow, UK). Receiver operating characteristic curve analyses were performed to measure the diagnostic accuracy of each protein and correlation between biomarker intensity and synovial PMN count was determined using Spearman’s test. Multivariate comparison between two populations was performed by supervised hierarchical clustering analyses of the characterized biomarkers and results were visualized by heat maps which are a graphical representation of overexpressed (red), decreased (green) or unchanged (black) proteins previously identified as differentially expressed in the two studied populations.

Results

SF proteome profiling

RA vs OA. Distinct protein profiles of SF were observed on the four arrays. Among 194 proteins detected in OA and RA SF, 74 were differentially expressed in both populations (P < 0.05). The most differentially expressed biomarkers were three proteins isolated on the four arrays at 10 445 (3.8), 10 839 (4.5) and 13 338 (17.3) Da, which were all overexpressed in RA SF (Fig. 1A). Their mean intensities were four to twelve times higher in RA than in OA, with significant statistical differences between the two populations (P < 0.0001, Table 2). Another cluster of three peptides of ∼3000 Da was differentially expressed in OA and RA SF (Fig. 1B). These biomarkers of 3351 (2.9), 3423 (3.1) and 3465 (2.5) Da, characterized on the IMAC-Cu, IMAC-Ni and CM10 arrays, were statistically enhanced in RA SF (Table 2). They were slightly less accurate than the previous biomarkers for the discrimination of the RA population from the OA group.

Table 2

Characteristics and identification of biomarkers

 Biomarker characteristics
 
Biomarker MALDI–TOF-MS/MS identification
 
Biomarker, Da P-value Sensitivity, % Specificity, % AUC Cut-off Accession number MOWSE score P-value Peptide number Peptide sequence 
RA vs OA           
10 445 <0.0001 88 76 0.89 0.94 P80511 365 <0.01 ELANTIK, GHFDTLSK, KGHFDTLSK, AAHYHTHKE, GHFDTLSKGELK, TKLEEHLEGIVNIFHQYSVR 
10 839 <0.0001 93 83 0.88 0.86 P05109 334 <0.01 10 GADVWFK, MGVAAHKK, MLTELEK, KGADVWFK, GNFHAVYR, SHEESHKE, ALNSIIDVYHK, LLETECPQYIR, LLETECPQYIRK, YSLIKGNFHAVYR 
13 338 <0.0001 91 75 0.80 0.38 P06702 643 <0.01 10 MSQLER, DLQNFLK, LTWASHEK, DLQNFLKK, MTCKMSQLER, LGHPDTLNQGEFK, VIEHIMEDLDTNAD, NIETIINTFHQYSVK, LGHPDTLNQGEFKELVR, MHEGDEGPGHHHKPGLGEGTP 
3351 <0.0001 81 65 0.84 2.32 Not applicable  
3423 0.0002 78 59 0.78 10.11  
3465 <0.0001 81 65 0.87 4.11  
RA vs MIA           
10 445 0.0023 79 64 0.77 1.35 P80511 365 <0.01 ELANTIK, GHFDTLSK, KGHFDTLSK, AAHYHTHKE, GHFDTLSKGELK, TKLEEHLEGIVNIFHQYSVR 
10 839 0.0015 82 69 0.80 2.07 P05109 334 <0.01 10 GADVWFK, MGVAAHKK, MLTELEK, KGADVWFK, GNFHAVYR, SHEESHKE, ALNSIIDVYHK, LLETECPQYIR, LLETECPQYIRK, YSLIKGNFHAVYR 
13 338 0.0429 61 80 0.71 0.92 P06702 643 <0.01 10 MSQLER, DLQNFLK, LTWASHEK, DLQNFLKK, MTCKMSQLER, LGHPDTLNQGEFK, VIEHIMEDLDTNAD, NIETIINTFHQYSVK, LGHPDTLNQGEFKELVR, MHEGDEGPGHHHKPGLGEGTP 
3351 0.0429 71 56 0.67 3.40 Not applicable  
3423 0.0710 75 56 0.66 6.12  
3465 0.0454 75 56 0.70 5.94  
 Biomarker characteristics
 
Biomarker MALDI–TOF-MS/MS identification
 
Biomarker, Da P-value Sensitivity, % Specificity, % AUC Cut-off Accession number MOWSE score P-value Peptide number Peptide sequence 
RA vs OA           
10 445 <0.0001 88 76 0.89 0.94 P80511 365 <0.01 ELANTIK, GHFDTLSK, KGHFDTLSK, AAHYHTHKE, GHFDTLSKGELK, TKLEEHLEGIVNIFHQYSVR 
10 839 <0.0001 93 83 0.88 0.86 P05109 334 <0.01 10 GADVWFK, MGVAAHKK, MLTELEK, KGADVWFK, GNFHAVYR, SHEESHKE, ALNSIIDVYHK, LLETECPQYIR, LLETECPQYIRK, YSLIKGNFHAVYR 
13 338 <0.0001 91 75 0.80 0.38 P06702 643 <0.01 10 MSQLER, DLQNFLK, LTWASHEK, DLQNFLKK, MTCKMSQLER, LGHPDTLNQGEFK, VIEHIMEDLDTNAD, NIETIINTFHQYSVK, LGHPDTLNQGEFKELVR, MHEGDEGPGHHHKPGLGEGTP 
3351 <0.0001 81 65 0.84 2.32 Not applicable  
3423 0.0002 78 59 0.78 10.11  
3465 <0.0001 81 65 0.87 4.11  
RA vs MIA           
10 445 0.0023 79 64 0.77 1.35 P80511 365 <0.01 ELANTIK, GHFDTLSK, KGHFDTLSK, AAHYHTHKE, GHFDTLSKGELK, TKLEEHLEGIVNIFHQYSVR 
10 839 0.0015 82 69 0.80 2.07 P05109 334 <0.01 10 GADVWFK, MGVAAHKK, MLTELEK, KGADVWFK, GNFHAVYR, SHEESHKE, ALNSIIDVYHK, LLETECPQYIR, LLETECPQYIRK, YSLIKGNFHAVYR 
13 338 0.0429 61 80 0.71 0.92 P06702 643 <0.01 10 MSQLER, DLQNFLK, LTWASHEK, DLQNFLKK, MTCKMSQLER, LGHPDTLNQGEFK, VIEHIMEDLDTNAD, NIETIINTFHQYSVK, LGHPDTLNQGEFKELVR, MHEGDEGPGHHHKPGLGEGTP 
3351 0.0429 71 56 0.67 3.40 Not applicable  
3423 0.0710 75 56 0.66 6.12  
3465 0.0454 75 56 0.70 5.94  

Statistical significance, cut-off value, AUC, specificity and sensitivity were obtained from Q10 (10 445, 10 839 and 13 338 Da biomarkers) or CM10 (3351, 3423 and 3465 Da biomarkers) SF profiling. Biomarker identification was performed by MALDI–TOF. AUC: area under the curve.

Fig. 1

Characterization by SF proteome profiling of biomarkers discriminating a RA population from OA patients. Relative SELDI–TOF-MS peak intensity of the 10 445, 10 839 and 13 338 Da proteins on a Q10 array (A) and of 3351, 3423 and 3465 Da peptides on a CM10 array (B). Boxes represent the interquartile range, the line across the box is the median and the whiskers represent the 5th and 95th percentiles. P < 0.0001, P < 0.005. Heat map visualization of the 74 biomarkers significantly differentially expressed in RA vs OA SF (C). Enhancement of the 10 445 (←), 10 839 (forumla), 13 338 (forumla), 3351 (*), 3423 (**) and 3465 Da (***) biomarkers in patients suffering from RA (blue numbers) compared with OA patients (red numbers).

Fig. 1

Characterization by SF proteome profiling of biomarkers discriminating a RA population from OA patients. Relative SELDI–TOF-MS peak intensity of the 10 445, 10 839 and 13 338 Da proteins on a Q10 array (A) and of 3351, 3423 and 3465 Da peptides on a CM10 array (B). Boxes represent the interquartile range, the line across the box is the median and the whiskers represent the 5th and 95th percentiles. P < 0.0001, P < 0.005. Heat map visualization of the 74 biomarkers significantly differentially expressed in RA vs OA SF (C). Enhancement of the 10 445 (←), 10 839 (forumla), 13 338 (forumla), 3351 (*), 3423 (**) and 3465 Da (***) biomarkers in patients suffering from RA (blue numbers) compared with OA patients (red numbers).

The 74 biomarkers whose expression was significantly different in OA and RA groups were visualized in a heat map (Fig. 1C). Most of them were increased in RA and decreased in OA SF. Hierarchical clustering classification, according to these proteins, created a first group exclusively constituted of RA samples. The sensitivity of this analysis in classifying RA patients in this group was 87%, with only four misclassified samples. No OA sample was classified in the RA-pattern group, corresponding to a specificity of 100%.

RA vs MIA

Although almost 200 peaks were detected in RA and MIA spectra, only 27 proteins were statistically differentially expressed between the two populations. Among them, the previously characterized 10 445 and 10 839 Da proteins were the most relevant to differentiation of the two populations, whereas the 13 338 Da biomarker was a little less discriminating (Table 2). The average intensity of the 10 839 Da peak was 5.05 (4.40) in the RA set vs 1.88 (1.25) in the MIA group (Fig. 2A), providing the best sensitivity and specificity for the diagnosis of RA in these two populations (Table 2). Concerning the three peptides of ∼3000 Da, mean intensities of each peak were significantly higher in RA SF than in MIA SF (Fig. 2B). However, their specificity and sensitivity in discriminating RA from MIA patients were lower than those of the aforementioned biomarkers (Table 2). Classification of samples according to the 27 proteins whose expression was significantly different in MIA and RA groups, did not improve the breakdown of the population in each group compared with the discriminating potential of individual markers: this hierarchical clustering map displayed a sensibility of 75% and a specificity of 56% in differentiating RA from MIA samples (Fig. 2C).

Fig. 2

Characterization by SF proteome profiling of biomarkers discriminating a RA population from MIA patients. Relative SELDI–TOF-MS peak intensity of the 10 445, 10 839 and 13 338 Da proteins on Q10 array (A) and of 3351, 3423 and 3465 Da peptides on CM10 array (B). Boxes represent the interquartile range, the line across the box is the median and the whiskers represent the 5th and 95th percentile. P < 0.005, §P < 0.05. Heat map visualization of the 27 biomarkers significantly differentially expressed in MIA vs RA SF (C). Enhancement of the 10 445 (←), 10 839 (⇐), 13 338 (forumla), 3351 (*) and 3465 Da (***) biomarkers in patients suffering from RA (red numbers) compared with MIA patients (blue numbers).

Fig. 2

Characterization by SF proteome profiling of biomarkers discriminating a RA population from MIA patients. Relative SELDI–TOF-MS peak intensity of the 10 445, 10 839 and 13 338 Da proteins on Q10 array (A) and of 3351, 3423 and 3465 Da peptides on CM10 array (B). Boxes represent the interquartile range, the line across the box is the median and the whiskers represent the 5th and 95th percentile. P < 0.005, §P < 0.05. Heat map visualization of the 27 biomarkers significantly differentially expressed in MIA vs RA SF (C). Enhancement of the 10 445 (←), 10 839 (⇐), 13 338 (forumla), 3351 (*) and 3465 Da (***) biomarkers in patients suffering from RA (red numbers) compared with MIA patients (blue numbers).

Characterization of the detected biomarkers

The group of the 3000 Da peptides was previously identified as α-defensin-1, -2 and -3 in several SELDI–TOF-MS studies [14, 29]. Evaluation of the whole concentration of the three α-defensins by ELISA confirmed significant overexpression of these peptides in RA SF vs MIA SF [8.81 (6.83) vs 4.46 (7.48) ng/ml, P = 0.006]. Moreover, these values were greatly correlated with SELDI–TOF-MS peak intensity (ρ = 0.722, P < 0.0001).

Then we focused on the 10 445, 10 839 and 13 338 Da biomarkers, which are the most relevant biomarkers for the RA diagnosis. Three proteins at about 10 400, 10 800 and 13 300 Da were purified from a RA SF and were subjected to MS/MS analysis (Fig. 3). They were, respectively, identified as S100A12 (calgranulin C), S100A8 (also called MRP-8 or calgranulin A) and S100A9 (MRP-14 or calgranulin B) proteins (Table 2). The molecular masses of S100A12 (10 445 Da) and S100A8 (10 839 Da) were in agreement with their published masses of 10 444 and 10 835 Da [30]. The difference between the 13 338 Da peak and the S100A9 expected mass (13 243 Da) may be explained by phosphorylation at Thr113 as previously described [31]. Indeed, we confirmed by immunoblotting that S100A8 and S100A9 were increased in RA SF in comparison with OA or MIA SF (Fig. 3A). The relative intensities of bands correlated with SELDI–TOF-MS peak intensities (Fig. 3B and C). We also checked that several parameters did not influence the biomarker expression as evaluated by SELDI–TOF-MS (Table 3).

Fig. 3

Identification of S100A12, S100A8 and S100A9 as biomarkers. S100A8 and S100A9 immunoblotting (A), matching SELDI–TOF-MS peak intensity (B) and SELDI–TOF-MS fingerprint in RA, OA and AS SF (C). Three Coomassie blue stained gel bands migrating at about 10 400, 10 800 and 13 300 Da were cut off (broken lines) and subjected to MS/MS analysis (D).

Fig. 3

Identification of S100A12, S100A8 and S100A9 as biomarkers. S100A8 and S100A9 immunoblotting (A), matching SELDI–TOF-MS peak intensity (B) and SELDI–TOF-MS fingerprint in RA, OA and AS SF (C). Three Coomassie blue stained gel bands migrating at about 10 400, 10 800 and 13 300 Da were cut off (broken lines) and subjected to MS/MS analysis (D).

Table 3

Biomarker expression according to population characteristics and sample preparation

Biomarkers Age
 
Gender
 
RF
 
Treatment
 
Frozen duration
 
 < 55 years ⩾ 55 years P-value Male Female P-value < 15 IU/ml ⩾ 15 IU/ml P-value No biological Biologicals P-value < 2 years > 2 years P-value 
S100A12                
    Median 2.3 2.3 0.950 2.4 2.3 0.924 2.3 2.3 0.535 2.6 2.1 0.198 5.6 2.3 0.794 
    (IQR) (1.9–3.7) (1.7–3.1)  (1.5–3.8) (1.7–3.3)  (1.4–5.5) (1.9–3.9)  (1.93.7) (1.3–2.3)  (1.6–11.3) (1.6–3.1)  
S100A8                
    Median 3.9 3.9 0.901 3.6 4.3 0.680 3.6 5.7 0.214 5.3 3.3 0.292 2.5 4.3 0.884 
    (IQR) (2.6–6.8) (2.5–5.6)  (1.6–6.0) (3.0–5.8)  (2.1–5.5) (3.2–7.0)  (3.1–6.3) (1.5–5.4)  (2.3–3.9) (2.3–5.8)  
S100A9                
    Median 1.3 1.2 0.950 1.2 1.3 0.504 1.2 1.5 0.457 1.3 1.1 0.412 4.5 1.2 0.999 
    (IQR) (0.3–1.9) (0.6–2.7)  (0.2–1.8) (0.6–2.7)  (0.5–1.9) (0.6–2.8)  (0.5–2.5) (0.3–1.9)  (3.0–6.5) (0.5–2.4)  
α-Defensin-2                
    Median 43.3 39.2 0.158 14.0 49.4 0.273 16.8 54.1 0.051 45.5 43.3 0.594 28.6 49.4 0.999 
    (IQR) (4.9–60.8) (2.6–69.1)  (2.1–52.1) (23.4–73.5)  (2.1–53.4) (40.0–74.0)  (9.4–86.9) (2.7–54.1)  (3.2–52.0) (4.0–54.1)  
α-Defensin-1                
    Median 33.2 35.1 0.205 13.9 35.8 0.247 15.3 37.3 0.072 36.9 15.9 0.332 16.0 37.1 0.528 
    (IQR) (7.1–40.7) (3.6–67.7)  (1.2–48.2) (20.1–54.7)  (2.3–50.0) (30.6–74.2)  (11.0–69.6) (2.3–35.6)  (5.1–26.6) (6.1–48.2)  
α-Defensin-3                
    Median 16.0 12.4 0.371 6.8 18.1 0.319 6.6 20.8 0.111 15.8 17.5 0.734 84.2 54.1 0.698 
    (IQR) (6.5–22.1) (5.1–27.4)  (4.3–17.6) (8.4–28.7)  (4.1–18.7) (15.5–28.1)  (7.3–29.8) (4.3–21.2)  (6.5–140.5) (6.5–120.8)  
Biomarkers Age
 
Gender
 
RF
 
Treatment
 
Frozen duration
 
 < 55 years ⩾ 55 years P-value Male Female P-value < 15 IU/ml ⩾ 15 IU/ml P-value No biological Biologicals P-value < 2 years > 2 years P-value 
S100A12                
    Median 2.3 2.3 0.950 2.4 2.3 0.924 2.3 2.3 0.535 2.6 2.1 0.198 5.6 2.3 0.794 
    (IQR) (1.9–3.7) (1.7–3.1)  (1.5–3.8) (1.7–3.3)  (1.4–5.5) (1.9–3.9)  (1.93.7) (1.3–2.3)  (1.6–11.3) (1.6–3.1)  
S100A8                
    Median 3.9 3.9 0.901 3.6 4.3 0.680 3.6 5.7 0.214 5.3 3.3 0.292 2.5 4.3 0.884 
    (IQR) (2.6–6.8) (2.5–5.6)  (1.6–6.0) (3.0–5.8)  (2.1–5.5) (3.2–7.0)  (3.1–6.3) (1.5–5.4)  (2.3–3.9) (2.3–5.8)  
S100A9                
    Median 1.3 1.2 0.950 1.2 1.3 0.504 1.2 1.5 0.457 1.3 1.1 0.412 4.5 1.2 0.999 
    (IQR) (0.3–1.9) (0.6–2.7)  (0.2–1.8) (0.6–2.7)  (0.5–1.9) (0.6–2.8)  (0.5–2.5) (0.3–1.9)  (3.0–6.5) (0.5–2.4)  
α-Defensin-2                
    Median 43.3 39.2 0.158 14.0 49.4 0.273 16.8 54.1 0.051 45.5 43.3 0.594 28.6 49.4 0.999 
    (IQR) (4.9–60.8) (2.6–69.1)  (2.1–52.1) (23.4–73.5)  (2.1–53.4) (40.0–74.0)  (9.4–86.9) (2.7–54.1)  (3.2–52.0) (4.0–54.1)  
α-Defensin-1                
    Median 33.2 35.1 0.205 13.9 35.8 0.247 15.3 37.3 0.072 36.9 15.9 0.332 16.0 37.1 0.528 
    (IQR) (7.1–40.7) (3.6–67.7)  (1.2–48.2) (20.1–54.7)  (2.3–50.0) (30.6–74.2)  (11.0–69.6) (2.3–35.6)  (5.1–26.6) (6.1–48.2)  
α-Defensin-3                
    Median 16.0 12.4 0.371 6.8 18.1 0.319 6.6 20.8 0.111 15.8 17.5 0.734 84.2 54.1 0.698 
    (IQR) (6.5–22.1) (5.1–27.4)  (4.3–17.6) (8.4–28.7)  (4.1–18.7) (15.5–28.1)  (7.3–29.8) (4.3–21.2)  (6.5–140.5) (6.5–120.8)  

Statistical significance was calculated from Q10 (S100 proteins) or CM10 (α-defensins) SELDI–TOF-MS protein biomarker intensities (i.e. peak height from baseline) using Mann–Whitney U-test. RF assessed by Waaler–Rose technique (n < 15 IU/ml).

Origin of the S100A12, S100A8 and S100A9 proteins according to the disease

S100A12, S100A8 and S100A9 are mainly produced by PMN in inflammatory diseases. Although synovial levels of these proteins were significantly different between RA and MIA populations, no statistical difference in the synovial neutrophil count was observed (Table 1). Moreover, we found a significant correlation between peak intensity of S100A12, S100A8 and S100A9 and neutrophil count in the SF of MIA patients (ρ = 0.965, P = 0.0005; ρ = 0.833, P = 0.0027; and ρ = 0.832, P = 0.0058, respectively) but not in the RA population (ρ = 0.234, P = 0.25; ρ = 0.321, P = 0.12; ρ = 0.187, P = 0.39, respectively) (Fig. 4A–C). Similarly, we demonstrated a statistical correlation between α-defensin concentration and neutrophil count in MIA SF (ρ = 0.857, P = 0.003) but not in RA SF (ρ = 0.316, P = 0.18) (Fig. 4D). Additionally, the α-defensin concentration was correlated with S100A12 (ρ = 0.820, P = 0.003), S100A8 (ρ = 0.903, P = 0.001) and S100A9 levels (ρ = 0.934, P < 0.001) in the MIA group, whereas in the RA population, no correlation could be found between α-defensin concentration and S100A12 (ρ = 0.277, P = 0.16), S100A8 (ρ = 0.292, P = 0.14) or S100A9 intensity (ρ = 0.289, P = 0.19).

Fig. 4

Origin of the S100A12, S100A8 and S100A9 proteins according to the disease. Correlation between PMN count in SF and SELDI–TOF-MS intensity of S100A8, S100A9 and S100A12 proteins on Q10 array or concentration of α-defensins assessed by ELISA. S100A12 (A), S100A8 (B), S100A9 (C) and α-defensin (D) concentration correlates with neutrophil count in MIAs, (bold line), but not in RA (broken line).

Fig. 4

Origin of the S100A12, S100A8 and S100A9 proteins according to the disease. Correlation between PMN count in SF and SELDI–TOF-MS intensity of S100A8, S100A9 and S100A12 proteins on Q10 array or concentration of α-defensins assessed by ELISA. S100A12 (A), S100A8 (B), S100A9 (C) and α-defensin (D) concentration correlates with neutrophil count in MIAs, (bold line), but not in RA (broken line).

Serum proteome profiling

We also investigated the serum protein profiles of the three populations in order to determine whether serum could contain biomarkers as discriminating as those characterized in SF. Among 361 peaks detected in spectra of RA and OA samples, 21 were differentially expressed between both groups whereas comparison of RA and MIA populations showed only seven statistically differentially expressed proteins among 393 identified peaks. Contrary to the SF analysis, we did not detect any 10 444 and 3465 Da peaks in the serum of any population. Moreover, no statistical difference in the intensity of the other SF biomarkers, such as the 10 845 (P = 0.76), 13 338 (P = 0.35), 3351 (P = 0.96) and 3423 Da (P = 0.83) proteins was characterized when comparing RA spectra with the profile of other groups.

Discussion

The diagnosis of RA is often delayed because of the poor sensitivity of biological and radiological tests. The discovery of new biomarkers could lead to earlier diagnosis and treatment, a better understanding of the disease and to the development of new therapies. As arthritis is the hallmark of RA, we chose to evaluate in this study the proteome of the SF by high-throughput SELDI–TOF-MS technology. The proteomic profile of the RA group was compared with both an OA population, displaying little joint inflammation, and a MIA group, which exhibited a similar joint inflammation pattern. Among several proteins characteristic of RA, three peptides, the α-defensins-1, -2 and -3 were significantly increased in RA fluids compared with OA samples and to a lesser extent compared with MIA SFs. Though the best RA biomarkers were three proteins of 10 445, 10 839 and 13 338 Da (identified, respectively, as S100A12, S100A8 and S100A9 proteins), which discriminated not only RA from OA but also RA from other inflammatory arthritides with high sensitivity and specificity. Moreover, a statistical correlation was demonstrated between synovial PMN concentration and levels of S100A8, S100A9, S100A12 and α-defensins in inflammatory joint diseases but not in RA, suggesting that these biomarkers were produced not only by neutrophil effusion but rather by RA synovial membrane.

Proteomic tests have been developed in the rheumatology field for several years. First 2D gel electrophoresis [32–35] and more recently, SELDI–TOF-MS technology [16–18] have been used to analyse the proteome of SF [32–34], blood [16, 17, 32, 33] or synovium [35] and to detect RA biomarkers compared with OA. All these surveys reported, among several markers, the enhancement of S100A8 in RA and/or S100A9 in SF [18, 32, 33], serum [16, 17] or tissue [35], sometimes associated with S100A12 [17, 32]. Levels of these inflammation markers correlated with the disease activity [34, 36, 37], radiological damage [37] and response to treatment [33, 38, 39]. In this study, we demonstrated that the synovial concentration of S100A8, S100A9 and S100A12 can also discriminate RA patients from other inflammatory arthritic patients. However, our results of serum proteome profiling showed few biomarkers compared with the SF proteome profiling.

S100A8, S100A9, S100A12 and α-defensins are mainly expressed by PMNs. Although it was not surprising that levels of these proteins were associated with neutrophil count in non-RA inflammatory arthritides, the absence of such a correlation in RA was noteworthy. Indeed α-defensins, which were detected in both synovial membrane [40] and SF [28], are also produced by other cells infiltrating the synovial tissue during RA such as B cells and NK lymphocytes. Moreover, an increase in S100A8 and S100A9 mRNA expression by fibroblast-like synoviocytes in a murine model of RA [41], has been reported and both S100A8 and S100A9 proteins were detected in synovial resident cells in the cartilage–pannus junction of RA synovial membranes [42]. Similarly, S100A12 cannot be found in healthy synovial tissue [38], but is strongly produced by RA synovial membrane. These data suggest that another cell type of the synovium may produce S100A8, S100A9 and S100A12 proteins in RA, potentially taking part in the disease pathogenesis.

Indeed, an increased body of evidence suggests that S100A8, S100A9 and S100A12 may be implicated in joint inflammation. First, S100A8 and S100A9 interfere with control mechanisms of inflammatory pain [43, 44]. Secondly, these proteins are involved in the activation of NADPH oxidase [26], leading to the production of reactive oxygen species and to cartilage destruction. Thirdly, S100A12 binds to RAGE [5, 7] of RA fibroblast-like synoviocytes, which induces in vitro a pseudo-tumoral phenotype characterized by a higher proliferation rate, invasiveness [45] and production of MMPs [46]. Moreover, S100A8, S100A9 and S100A12 are essential neutrophil chemokines [47] and stimulate the secretion of pro-inflammatory cytokines [6]. Finally, the enhanced cardiovascular morbidity in RA may be related to their ability to promote endothelial stress, atherosclerosis [48] and possibly acute coronary syndrome [49].

Our study underlines the potential use of the synovial expression of S100A8, S100A9 and S100A12 for RA diagnosis. As these proteins were detected in the synovium early during disease onset [42], their evaluation may be very helpful in undetermined early arthritides. Further studies are required to explore the possible link between increased S100A8 and S100A9 level and atherosclerosis in RA. Moreover, the control of their expression may represent a therapeutic target in RA, considering their proinflammatory properties.

graphic

Acknowledgements

We would like to thank Pr François Moutet and Pr Dominique Saragaglia for management and technical help.

Funding: This work was supported by unrestricted grants from Abbott France, Wyeth Pharmaceuticals France, the Région Rhône-Alpes (MIRA 2007–2008, ARCUS 2007–2008 and Emergence 2003–2006), the UFR Médecine from Joseph Fourier University (Vivier de la Recherche), the Groupement des Entreprises Françaises et Monégasques dans la Lutte contre le Cancer (Department of Grenoble), the Direction Régionale de la Recherche Clinique (CHU Grenoble), the Ligue Nationale contre le Cancer, the Association Nationale de Défense contre L’Arthrite Rhumatoïde, the CGD Research Trust 2006–2007, the Association Nationale de Défense contre L’Arthrite Rhumatoïde, the CNRS Institute and the Société Française de Rhumatologie.

Disclosure statement: The authors have declared no conflicts of interest.

References

1
Roux
CH
Saraux
A
Le Bihan
E
, et al.  . 
Rheumatoid arthritis and spondyloarthropathies: geographical variations in prevalence in France
J Rheumatol
 , 
2007
, vol. 
34
 (pg. 
117
-
22
)
2
Leadbetter
EA
Rifkin
IR
Hohlbaum
AM
Beaudette
BC
Shlomchik
MJ
Marshak-Rothstein
A
Chromatin–IgG complexes activate B cells by dual engagement of IgM and Toll-like receptors
Nature
 , 
2002
, vol. 
416
 (pg. 
603
-
7
)
3
Matzinger
P
The danger model: a renewed sense of self
Science
 , 
2002
, vol. 
296
 (pg. 
301
-
5
)
4
McInnes
IB
Schett
G
Cytokines in the pathogenesis of rheumatoid arthritis
Nat Rev Immunol
 , 
2007
, vol. 
7
 (pg. 
429
-
42
)
5
Leclerc
E
Fritz
G
Vetter
SW
Heizmann
CW
 
Binding of S100 proteins to RAGE: an update. Biochim Biophys Acta 2008;1793:993–1007
6
Vogl
T
Tenbrock
K
Ludwig
S
, et al.  . 
Mrp8 and Mrp14 are endogenous activators of Toll-like receptor 4, promoting lethal, endotoxin-induced shock
Nat Med
 , 
2007
, vol. 
13
 (pg. 
1042
-
9
)
7
Ghavami
S
Rashedi
I
Dattilo
BM
, et al.  . 
S100A8/A9 at low concentration promotes tumor cell growth via RAGE ligation and MAP kinase-dependent pathway
J Leukoc Biol
 , 
2008
, vol. 
83
 (pg. 
1484
-
92
)
8
Broome
AM
Ryan
D
Eckert
RL
S100 protein subcellular localization during epidermal differentiation and psoriasis
J Histochem Cytochem
 , 
2003
, vol. 
51
 (pg. 
675
-
85
)
9
Seibert
V
Wiesner
A
Buschmann
T
Meuer
J
Surface-enhanced laser desorption ionization time-of-flight mass spectrometry (SELDI TOF-MS) and ProteinChip technology in proteomics research
Pathol Res Pract
 , 
2004
, vol. 
200
 (pg. 
83
-
94
)
10
Vorderwulbecke
S
Kramer
G
Merz
F
, et al.  . 
Low temperature of GroEL/ES overproduction permits growth of Escherichia coli cells lacking trigger factor DnaK
FEBS Lett
 , 
2005
, vol. 
579
 (pg. 
181
-
7
)
11
Schellekens
GA
Visser
H
de Jong
BA
, et al.  . 
The diagnostic properties of rheumatoid arthritis antibodies recognizing a cyclic citrullinated peptide
Arthritis Rheum
 , 
2000
, vol. 
43
 (pg. 
155
-
63
)
12
Clarke
CH
Buckley
JA
Fung
ET
SELDI–TOF-MS proteomics of breast cancer
Clin Chem Lab Med
 , 
2005
, vol. 
43
 (pg. 
1314
-
20
)
13
Zhang
Z
Bast
RC
Jr
Yu
Y
, et al.  . 
Three biomarkers identified from serum proteomic analysis for the detection of early stage ovarian cancer
Cancer Res
 , 
2004
, vol. 
64
 (pg. 
5882
-
90
)
14
Albrethsen
J
Bogebo
R
Gammeltoft
S
Olsen
J
Winther
B
Raskov
H
Upregulated expression of human neutrophil peptides 1, 2 and 3 (HNP 1–3) in colon cancer serum and tumours: a biomarker study
BMC Cancer
 , 
2005
, vol. 
5
 pg. 
8
 
15
Trocme
C
Marotte
H
Baillet
A
, et al.  . 
Apolipoprotein A-I and platelet factor 4 are biomarkers for infliximab response in rheumatoid arthritis
Ann Rheum Dis
 , 
2008
, vol. 
68
 (pg. 
1328
-
33
)
16
de Seny
D
Fillet
M
Meuwis
MA
, et al.  . 
Discovery of new rheumatoid arthritis biomarkers using the surface-enhanced laser desorption/ionization time-of-flight mass spectrometry ProteinChip approach
Arthritis Rheum
 , 
2005
, vol. 
52
 (pg. 
3801
-
12
)
17
de Seny
D
Fillet
M
Ribbens
C
, et al.  . 
Monomeric calgranulins measured by SELDI–TOF mass spectrometry and calprotectin measured by ELISA as biomarkers in arthritis
Clin Chem
 , 
2008
, vol. 
54
 (pg. 
1066
-
75
)
18
Uchida
T
Fukawa
A
Uchida
M
Fujita
K
Saito
K
Application of a novel protein biochip technology for detection and identification of rheumatoid arthritis biomarkers in synovial fluid
J Proteome Res
 , 
2002
, vol. 
1
 (pg. 
495
-
9
)
19
Pap
T
Muller-Ladner
U
Gay
RE
Gay
S
Fibroblast biology. Role of synovial fibroblasts in the pathogenesis of rheumatoid arthritis
Arthritis Res
 , 
2000
, vol. 
2
 (pg. 
361
-
7
)
20
Arnett
FC
Edworthy
SM
Bloch
DA
, et al.  . 
The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis
Arthritis Rheum
 , 
1988
, vol. 
31
 (pg. 
315
-
24
)
21
Altman
R
Asch
E
Bloch
D
, et al.  . 
Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Association
Arthritis Rheum
 , 
1986
, vol. 
29
 (pg. 
1039
-
49
)
22
van der Linden
S
Valkenburg
HA
Cats
A
Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria
Arthritis Rheum
 , 
1984
, vol. 
27
 (pg. 
361
-
8
)
23
Petty
RE
Southwood
TR
Manners
P
, et al.  . 
International League of Associations for Rheumatology classification of juvenile idiopathic arthritis: second revision, Edmonton, 2001
J Rheumatol
 , 
2004
, vol. 
31
 (pg. 
390
-
2
)
24
Tan
EM
Cohen
AS
Fries
JF
, et al.  . 
The 1982 revised criteria for the classification of systemic lupus erythematosus
Arthritis Rheum
 , 
1982
, vol. 
25
 (pg. 
1271
-
7
)
25
Schagger
H
Tricine-SDS-PAGE
Nat Protoc
 , 
2006
, vol. 
1
 (pg. 
16
-
22
)
26
Berthier
S
Paclet
MH
Lerouge
S
, et al.  . 
Changing the conformation state of cytochrome b558 initiates NADPH oxidase activation: MRP8/MRP14 regulation
J Biol Chem
 , 
2003
, vol. 
278
 (pg. 
25499
-
508
)
27
Paclet
MH
Berthier
S
Kuhn
L
Garin
J
Morel
F
Regulation of phagocyte NADPH oxidase activity: identification of two cytochrome b558 activation states
FASEB J
 , 
2007
, vol. 
21
 (pg. 
1244
-
55
)
28
Bokarewa
MI
Jin
T
Tarkowski
A
Intraarticular release and accumulation of defensins and bactericidal/permeability-increasing protein in patients with rheumatoid arthritis
J Rheumatol
 , 
2003
, vol. 
30
 (pg. 
1719
-
24
)
29
Buhimschi
IA
Buhimschi
CS
Weiner
CP
, et al.  . 
Proteomic but not enzyme-linked immunosorbent assay technology detects amniotic fluid monomeric calgranulins from their complexed calprotectin form
Clin Diagn Lab Immunol
 , 
2005
, vol. 
12
 (pg. 
837
-
44
)
30
McMorran
BJ
Patat
SA
Carlin
JB
, et al.  . 
Novel neutrophil-derived proteins in bronchoalveolar lavage fluid indicate an exaggerated inflammatory response in pediatric cystic fibrosis patients
Clin Chem
 , 
2007
, vol. 
53
 (pg. 
1782
-
91
)
31
Lominadze
G
Rane
MJ
Merchant
M
Cai
J
Ward
RA
McLeish
KR
Myeloid-related protein-14 is a p38 MAPK substrate in human neutrophils
J Immunol
 , 
2005
, vol. 
174
 (pg. 
7257
-
67
)
32
Sinz
A
Bantscheff
M
Mikkat
S
, et al.  . 
Mass spectrometric proteome analyses of synovial fluids and plasmas from patients suffering from rheumatoid arthritis and comparison to reactive arthritis or osteoarthritis
Electrophoresis
 , 
2002
, vol. 
23
 (pg. 
3445
-
56
)
33
Drynda
S
Ringel
B
Kekow
M
, et al.  . 
Proteome analysis reveals disease-associated marker proteins to differentiate RA patients from other inflammatory joint diseases with the potential to monitor anti-TNFalpha therapy
Pathol Res Pract
 , 
2004
, vol. 
200
 (pg. 
165
-
71
)
34
Liao
H
Wu
J
Kuhn
E
, et al.  . 
Use of mass spectrometry to identify protein biomarkers of disease severity in the synovial fluid and serum of patients with rheumatoid arthritis
Arthritis Rheum
 , 
2004
, vol. 
50
 (pg. 
3792
-
803
)
35
Tilleman
K
Van Beneden
K
Dhondt
A
, et al.  . 
Chronically inflamed synovium from spondyloarthropathy and rheumatoid arthritis investigated by protein expression profiling followed by tandem mass spectrometry
Proteomics
 , 
2005
, vol. 
5
 (pg. 
2247
-
57
)
36
Wulffraat
NM
Haas
PJ
Frosch
M
, et al.  . 
Myeloid related protein 8 and 14 secretion reflects phagocyte activation and correlates with disease activity in juvenile idiopathic arthritis treated with autologous stem cell transplantation
Ann Rheum Dis
 , 
2003
, vol. 
62
 (pg. 
236
-
41
)
37
Hammer
HB
Odegard
S
Fagerhol
MK
, et al.  . 
Calprotectin (a major leucocyte protein) is strongly and independently correlated with joint inflammation and damage in rheumatoid arthritis
Ann Rheum Dis
 , 
2007
, vol. 
66
 (pg. 
1093
-
7
)
38
Foell
D
Kane
D
Bresnihan
B
, et al.  . 
Expression of the pro-inflammatory protein S100A12 (EN-RAGE) in rheumatoid and psoriatic arthritis
Rheumatology
 , 
2003
, vol. 
42
 (pg. 
1383
-
9
)
39
Foell
D
Wittkowski
H
Hammerschmidt
I
, et al.  . 
Monitoring neutrophil activation in juvenile rheumatoid arthritis by S100A12 serum concentrations
Arthritis Rheum
 , 
2004
, vol. 
50
 (pg. 
1286
-
95
)
40
Paulsen
F
Pufe
T
Conradi
L
, et al.  . 
Antimicrobial peptides are expressed and produced in healthy and inflamed human synovial membranes
J Pathol
 , 
2002
, vol. 
198
 (pg. 
369
-
77
)
41
van Lent
PL
Grevers
L
Blom
AB
, et al.  . 
Myeloid-related proteins S100A8/S100A9 regulate joint inflammation and cartilage destruction during antigen-induced arthritis
Ann Rheum Dis
 , 
2008
, vol. 
67
 (pg. 
1750
-
8
)
42
Youssef
P
Roth
J
Frosch
M
, et al.  . 
Expression of myeloid related proteins (MRP) 8 and 14 and the MRP8/14 heterodimer in rheumatoid arthritis synovial membrane
J Rheumatol
 , 
1999
, vol. 
26
 (pg. 
2523
-
8
)
43
Dale
CS
Pagano Rde
L
Paccola
CC
, et al.  . 
Effect of the C-terminus of murine S100A9 protein on experimental nociception
Peptides
 , 
2006
, vol. 
27
 (pg. 
2794
-
802
)
44
Mitchell
K
Yang
HY
Tessier
PA
, et al.  . 
Localization of S100A8 and S100A9 expressing neutrophils to spinal cord during peripheral tissue inflammation
Pain
 , 
2008
, vol. 
134
 (pg. 
216
-
31
)
45
Steenvoorden
MM
Toes
RE
Ronday
HK
Huizinga
TW
Degroot
J
RAGE activation induces invasiveness of RA fibroblast-like synoviocytes in vitro
Clin Exp Rheumatol
 , 
2007
, vol. 
25
 (pg. 
740
-
2
)
46
Tolboom
TC
Pieterman
E
van der Laan
WH
, et al.  . 
Invasive properties of fibroblast-like synoviocytes: correlation with growth characteristics and expression of MMP-1, MMP-3, and MMP-10
Ann Rheum Dis
 , 
2002
, vol. 
61
 (pg. 
975
-
80
)
47
Ryckman
C
McColl
SR
Vandal
K
, et al.  . 
Role of S100A8 and S100A9 in neutrophil recruitment in response to monosodium urate monohydrate crystals in the air-pouch model of acute gouty arthritis
Arthritis Rheum
 , 
2003
, vol. 
48
 (pg. 
2310
-
20
)
48
McCormick
MM
Rahimi
F
Bobryshev
YV
, et al.  . 
S100A8 and S100A9 in human arterial wall. Implications for atherogenesis
J Biol Chem
 , 
2005
, vol. 
280
 (pg. 
41521
-
9
)
49
Altwegg
LA
Neidhart
M
Hersberger
M
, et al.  . 
Myeloid-related protein 8/14 complex is released by monocytes and granulocytes at the site of coronary occlusion: a novel, early, and sensitive marker of acute coronary syndromes
Eur Heart J
 , 
2007
, vol. 
28
 (pg. 
941
-
8
)

Author notes

*Athan Baillet and Candice Trocmé contributed equally to this work.

Comments

0 Comments