Proteomic analysis of synovial tissue in rheumatoid arthritis, OLFM4 promotes fibroblast like synoviocytes proliferation in RA

Background To understand the roles of synovial tissue in rheumatoid arthritis (RA), we comprehensively investigated the protein profiles of synovial tissues in patients with RA and osteoarthritis (OA). Methods Protein were isolated from synovial tissues of 46 patients (22 with RA, 24with OA). Proteins were digested and labeled with TMT kit. All the peptide were send to LC-MS/MS analysis, and processed with Mascot and Proteome Discovery (version 1.4) search engines. All the protein data were analysis with Gene ontology (GO) and KEGG on line database. Four up-regulated protein were examined with western blotting and IHC. The RNAi were used for revealing the OLFM4 function in FLS. Results In total, over 500 proteins were identified as the different expression protein compared RA and OA synovia tissues, including 239 proteins were up-regulated in RA, and 271 proteins were upregulated in OA. GO and KEGG analysis of the different expressed protein mostly identified as developmental processes and protein processing in ER. RT-qPCR, western blotting and IHC confirmed that there was a strong over expressed OLFM4 in RA synovia tissues. OLFM4 could be up-regulated under inflammatory stimulation, and participate in fibroblast like synoviocytes proliferation. Conclusion patients with RA possessed considerably different protein profiles of synovial tissues from with OA. The unique protein profiles of RA synovial tissues, such as the OLFM4 and MZB1 had been up-regulated compared to the OA, may reflect the pathophysiology of RA. tissue of OA patients. We confirmed the up-regulation of OLFM4 in RA synovial fluid by IHC and immunofluorescence as well as western blotting on protein level. The intervention of OLFM4 could inhibit the FLS proliferation. Pathway analysis of differentially expressed proteins revealed a significant enrichment of genes involved in protein processing in endoplasmic reticulum (ER) in rheumatoid arthritis. mRNA

using 28% NCE, ion fragments were detected in the Orbitrap at a resolution of 17,500. A datadependent procedure that alternated between one MS scan followed by 20 MS/MS scans was applied for the top 20 precursor ions above a threshold ion count of 1.5E4 in the MS survey scan with 60.0 ms dynamic exclusion. The electrospray voltage applied was 2.0 kV. Automatic gain control (AGC) was used to prevent overfilling of the ion trap; 5E4 ions were used to generate the MS/MS spectra. For MS scans, the m/z scan range was 300 to 1800 Da.

Peptides data processing
The MS data were processed using Mascot and Proteome Discovery (version 1.4) search engines.
Tandem mass spectra were searched against the UniProt_human database (161584 sequences) concatenated with the reverse Decoy Database. Trypsin/P was specified as a cleavage enzyme allowing up to 2 missed cleavages. Mass error was set to 20 ppm for precursor ions and 0.1 Da for fragment ions. Carbamidomethylation on cysteine was specified as a fixed modification, and oxidation and TMT were specified as variable modifications. A 1.0% false discovery rate (FDR) was used to filter the identified peptides. All the other parameters in Mascot were set to their default values.

GO analysis and KEGG pathway analysis
The online analysis database, Database for Annotation, Visualization and Integrated Discovery (DAVID) (https://david.ncifcrf.gov/, Version 6.8), to analyze the Gene Ontology (GO) annotation and the enrichment of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways for these data [18,19]. The p value < 0.05 and gene counts >2 were criteria for the both.

Western blotting assay
The synovial tissue from OA and RA patients were described in the synovial tissue segment. Such tissue samples were extracted in RIPA buffer with protease inhibitor cocktail. After 20 min incubation in ice, all the samples were centrifuged at 12000 rpm for 30 min at 4 ℃ and then quantified with BCA kit. 40 ug protein of each samples were loaded on 8-10% SDS-PAGE gels. The antibodies used for western blotting were OLFM4 (NOVOS, NBP2-24535), MZB1 (Abcam-ab181205) and DERL3 (Abcam-ab78233), S100P (Abcam-ab225543), and LAP3 ABclonal-A7101 , rabbit anti-ACTB. All the first antibodies were diluted at 1:1000, and the HRP conjugated goat anti-rabbit IgG antibody was diluted at 1:10000. Signal intensity was determined by Supersignal ® West Pico Kit (Thermo Scientific). And the images for western blotting were taken by Syngene G. All the density of the bands was measured and normalized to β-actin. Final data are expressed by showing one representative image and quantitative result of samples from various patients.

Immunohistochemistry
The paraffin-embedded RA and OA synovial tissues were sectioned and mounted on positively charged slides. All the slides were incubated at 60°C for overnight, followed by de-paraffinization and rehydration. These slides were transferred to the Leica Bond-IIITM and were treated for antigen retrieval at 100°C for 20 minutes in a retrieval solution. Endogenous peroxidase was blocked using peroxidase-blocking reagent, followed by primary antibody incubation for 60 minutes. The blocker buffer were used for blocking the unspecific reaction for 1 hour at room temperature. Then the first antibody of OLFM4, MZB1 (Marginal Zone B and B1 Cell Specific Protein) and DERL3 (Derlin 3), and S100P (S100 Calcium Binding Protein P) were diluted at 1:1000 and then applied to the tissue at 4 ℃ overnight. The next morning all the slides were taken out and balanced at room temperature for 30 min, then washed with PBS for times. Then the secondary antibody were used for 30min at room temperature, followed by the horseradish peroxidase reaction with DAB reaction.

RT-qPCR
The total RNA were isolated from RA and OA patients' synovial tissues, and then quantified. The cDNA were reversed by using First Strand cDNA Synthesis Kit (Roche, USA). The synovial fibroblast were isolated from RA synovial tissues. The FLS were treated with different pro-inflammatory cytokines, 10 ng/mL TNFα, 2 ng/mL IL-1β or 1 ng/mL LPS for 24 hours. All the RNA were extracted with Trizol, quantified and then the mRNA were reversed to cDNA by using First Strand cDNA Synthesis Kit (Roche, USA). The real time PCR was performed with Fast Start Universal SYBR Green Master (ROX) (Roche, USA) by Agilent Mx3000. PCR efficiency and specificity of each primer pair was examined by a standard curve of serially diluted cDNA and melting curve functionality respectively. Fold change was calculated based on 2 −ΔΔ Ct method after normalization to the transcript level of housekeeping gene ACTB. Primer sequences used in the RT-PCR could be found in Table 3.

RNAi
The siRNA kit for the target protein OLFM4 were purchased from Santa Cruz. The 20uM siRNA (siNC or siOLFM4, Santa Cruz) were transfected Lipofectamine™ 2000 Transfection Reagent (Thermo Scientific, 11668019) to the FLS for 72 hours. Then the cell were collected either for RT-qPCR or for cell proliferation assay.

Cell proliferation assay
The transfected NC and siOLFM FLS cell were incubated with PE-labeled cell trace dye (diluted at 1:1000) for 72hours. After 72 hours cultured, all the cell were collected and detected by flow cytometer. The unstained NC cell were used as a negative control. The MFI were measured to determine the cell proliferation rate after knocking down the OLFM4 gene.

Statistics
Statistics were carried out using SPSS and Graphpad Prism software. The student's t-test of the Mann-Whitney U test and unpaired student' s t-tests were used to determine statistical differences between two groups. The Fisher's tests were used to test for enrichment or one-way ANOVA depletion of specific annotation terms among members of the resulting protein cluster. P<0.05 was considered to indicated a statistically significant difference. Data were presented as mean ± SEM.

Clinical characteristic of patients
Of the 22 patients with RA and 24 patients with OA from whom blood sample and synovial tissue were obtained. The age, sex, prevalence of rheumatoid factor (RF), anti-cyclic citrullinated peptide (CCP), anti-CRP (C reactive protein) and erythrocytesedimentationrate (ESR) were listed in Table 1. There were significant differences in patients' age, RF, anti-CCP, CRP and ESR.

Proteome profiles of RA and OA synovial tissues.
All the peptide collected in this experiment were mainly distributed in 10 ppm ( Fig.1 A), the MASCOT score was over 20 ( Fig.1 B), and the protein ratio from AR and OA groups were merely to 1 ( Fig.1 C), which suggested the identification results were reliable could use for further analysis. Approximately 4822 proteins were found in synovial tissues via LC/MS, 239 proteins with a quantitative ration of more than 1.2 fold change were considered up-regulated, 271 proteins were considered downregulated if the quantitative ratio was less than 1.2 fold change ( Fig.1 D).

Differential protein expression, GO term and KEGG analysis in the synovial tissues
We summarized the up-regulated ( Fig Table 2. To fully understand the function of each protein as well as the cellular and biological processes involved, we performed GO term analysis. The GO term analysis ( Fig.2C) showed that most significantly changed proteins were divided in 3 groups. Some were associated with biological process (BP), which were mainly in developmental processes, anatomical structural development, extracellular structure organization, extracellular matrix organization, skeletal system development etc. Molecular function study showed calcium ion binding, glycosaminoglycan binding, RAGE receptor binding, metal ion binding, cation binding were changed significantly. As well as the extracellular matrix, proteinaceous extracellular matrix, extracellular region, collagen trimmer and fibrillar collagen trimmer were changed significantly in RA synovial tissues in cellular component analysis.
For further understanding the interaction and coordination of different protein function in the synovial tissues, the KEGG pathway analysis were carried out (Fig.3A). The KEGG pathway enrichment results illustrated that the significantly changed pathways were chemical carcinogenesis, drug metabolismcytochrome P450, protein processing in ER, metabolism of xenobiotic by cytochrome P450, and ECMreceptor interaction (Fig.3B).

Verification the candidate genes expression in RA and OA synovial tissues.
RT-qPCR were used to select the mRNA expression of different expressed genes in RA and OA synovial tissues. The results showed that, comparing with OA synovial tissues, only 4 genes, MZB1, OLFM4, DERL3 and S100P (Fig.4 A-D) were significantly up-regulated in RA synovial tissues. On the contrary, some candidates, like SEC61B (Fig.4 E) and RAC2 (Fig.4 K), only showed the increasing expression trend.

Verification of protein expression in RA and OA synovial tissues.
Western blotting and IHC were employed to further verify differential protein expression in order to ensure the reliability of proteomic results from iTraq. The protein expression of MZB1 and OLFM4 from RA and OA synovial tissues also showed significantly up-regulated (Fig.5 A-C), while the DERL3 and LAP3 were not that significantly up-regulated in RA synovial tissues than OA, when the western blotting was applied (Fig.5 D-E). What's more, the IHC results indicated that the expression levels of OLFM4, DERL3 and MZB1 were significantly higher in the RA synovial tissues compared with OA synovial tissues (Fig.5 F-J), which was consistent with proteomic analysis.

RA FLS cell proliferation via OLFM4
The expression of OLFM4 were determined in RA synovial tissue via IF, as the result showed in Fig.6A, the OLFM4 was mainly expressed in Vimentin positive cells. The FLS were isolated from synovial tissues of RA patients. The protein expression of OLFM4 could be significantly up-regulated after inflammation cytokine TNF-α, IL-1β or LPS treatment (Fig.6B). Meanwhile the mRNA expression of OLFM4 and IL-6 were checked at different time points in TNF-α treated RA FLS. The results showed that the mRNA expression of OLFM4 was persistently up-regulated after 2 hours TNF-α priming until 48 hours, while the IL-6 mRNA expression was up-regulated by TNF-α stimulation from 1 hour to 4 hours, and then up-regulated again on 12 hours to 48 hours (Fig.6C). The function of OLFM4 were reported as an anti-apoptosis protein. So the function of OLFM4 in RA FLS were studied by RNAi. The OLFM4 could be down-regulated via siRNA successfully (Fig.6D). The proliferation of FLS was tested by CFSE method. As the results showed in Fig.6E-F, after knocking down the expression of OLFM4, the PE labeled cells were significantly more than the NC group.

Discussion
Here, we compared the protein expression levels of synovial tissue from RA and OA patients by using iTRAQ and LC/MS based proteomics approach. As a result, over 500 protein were found different expression pattern in synovial tissues between RA and OA patients. The GO analysis and KEGG pathway analysis results also showed that, in RA synovial tissue, the protein processing in ER associated proteins was significantly up-regulated, as well as the extracellular matrix associated with the upregulated proteins in cell process. Some of the gene which related with these process and signal were also be checked. The candidates were confirmed by RT-qPCR and western blotting.
Compared with OA synovial tissues, the MZB1, OLFM4, DERL3 and S100P genes showed highly expressed in RA synovial tissues. However, the protein expression only showed MZB1 and OLFM4 were over expressed in RA patient's synovial tissues. Furthermore, the OLFM4 expression were evaluated via IHC. Since this gene was first reported in RA, a preliminary study was applied. The OLFM4 expression could be up-regulated after LPS, TNFα or IL1β stimulation. After knocking the OLFM4 expression in FLS, the proliferation of FLS were inhibited, suggesting this gene might promote the cell proliferation.
The GO analysis of process networks suggested that the extracellular matrix associated with the upregulated proteins in RA synovial tissue compared to the OA synovial tissue. Moreover, the KEGG pathway analysis also revealed that the protein processing in ER associated with the upregulated proteins in RA. The ER is the place for folding the secreted and membrane proteins properly in the biosynthesis process [30]. It was also reported that in RA FLS, the ER stress associated genes, like DDIT3 and HSP5, were increased and involved in regulating the inflammatory genes expression in activated FLS from RA patients [31, 32].
To further understand the up-regulated proteins in RA, we focused on the 2 genes. Human OLFM4 (olfactomedin 4, also known as hGC-1, GW112), originally termed human cloned from myeloid precursor cells after granulocyte colony-stimulating factor stimulation [16], is a secreted glycoprotein more commonly known as the anti-apoptotic molecule GW112 [17,18]. It was first cloned in human hematopoietic myeloid cells and is also abundantly expressed in intestinal crypts. [16,19] Recently it also was reported expressed in bone marrow, prostate, small intestine, stomach, colon and pancreas [16,20]. More recently, up-regulated OLFM4 expression has been described in lung and breast [ [23][24][25][26], suggesting that this gene might be involved in the cell proliferating process. And the FLS also showed a rapid growth characteristic in RA [27]. In our latest results showed that OLFM4 were upregulated in FLS from RA synovial tissues. Similarly, the increased OLFM4 levels were also found in the crypt epithelium of inflamed colonic mucosa of inflammatory bowel diseases [28] and in gastric biopsies infected with Helicobacter pylori [29,30]. It has been suggested that OLFM4 is involved in cellular process such as apoptosis and tumor growth [17]. We also found that the OLFM4 gene expression could be up-regulated after proinflammatory agent treatment, such as TNFa, IL1b or LPS, which means OLFM4 could be up-regulated in RA proinflammatory environment. It also reported that

Conclusion
In this paper, we used iTraq-labeled methods and found numerous changed proteins that were mainly in the pathways involved in phagosome, and protein processing in ER, and in the extracellular region and development process and indicated differentially expressed genes, including the first reported OLFM4 function in RA FLS. And the OLFM4 was upregulated in RA FLS, participated the cell proliferation progress. Taken together, we believe that our study could provide the opportunity for developing the new biomarker for diagnosis and treatment RA.

Availability of data and materials
The datasets generated and/or analyzed during the current study are available in the supplementary data. Ethics approval and consent to participate. Written informed consent was obtained from each patient.

Consent for publication
Not applicable.