Identication of Genes and Pathways of Nonsteroidal Anti-inammatory Drugs acting on Synovia from Women with Knee Osteoarthritis by Bioinformatics Analysis

Objective: Through the bioinformatics analysis, to identify the genes and pathways of nonsteroidal anti-inammatory drugs(NSARDs) acting on synovia from women with knee osteoarthritis (KOA), and to provide reference for clinical application. Methods: We downloaded the gene microarray datasets with the accession number of GSE55457 and GSE55584 from the Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/) database, including 5 untreated KOA patients, 9 NSARDs treated KOA patients and 2 patients without KOA The samples in the untreated KOA group and the NSARDs treated KOA group were used for main analysis. The samples in the untreated KOA group and the normal control group were used for cooperative analysis. Then we performed robust multi-array (RMA) normalization with affy R programming package. After that, differential expression genes (DEGs) in main analysis and cooperative analysis were identied based on limma package separately. Screening the common DEGs from main analysis and cooperative analysis. Enriched gene ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways of DEGs were obtained through the Database for Annotation, Visualization and Integrated Discovery (DAVID). What's more, protein-protein interaction (PPI) network was constructed, and we identied modules of PPI network through Cytoscape to screen valuable targets. The value of gene expression fold change (FC) ≥ 1.4 or ≤ 1/1.4, and P <0.05 were used as the screening conditions. P <0.05 and Associated genes count>5 were used as the screening conditions. Results: There were 338 DEGs in main analysis. Among them, 211 genes were up-regulated and 127 genes were down-regulated. There were 7005 DEGs in cooperative analysis. Among them, 6952 genes were up-regulated and 53 genes were down-regulated. A total of 129 common DEGs were identied between main analysis and cooperative analysis. There are 2 biological processes, 3 cell components and 2 molecular functions for the enrichment of differentially expressed genes. Conclusion: NSARDs may play a certain role in synovia from women with KOA by regulating the mRNA expressions of il-6, TNFRSF11A and CSF1R, which may become one of the indicators for monitoring the ecacy of NSAIDs.


Introduction
Knee osteoarthritis (KOA) is a degenerative disease characterized by degeneration of articular cartilage (1), which currently affects about 21.51% of the middle-aged and elderly (2), and is considered as one of the largest cause of disability (3). The genetic contribution of knee osteoarthritis surgery is higher in women than in men (4). KOA is often accompanied by synovitis, osteophyte formation and subchondral bone sclerosis (5). Clearly, all articular tissues, including the synovium, are involved in the overall pathologic process. In particular, synovitis, which plays a key role in the initiation and development of degenerative changes in cartilage, is associated with more severe pain and joint dysfunction (6).
Synoviocytes produce pro-in ammatory mediators, which in turn attract immune cells, increase angiogenesis and induce a phenotypic shift in chondrocytes. Chondrocytes produce additional cytokines and proteolytic enzymes, which eventually increase cartilage degradation and further induce synovial in ammation (7). Therefore, the study on synovial tissue of KOA patients will gradually attract extensive attention.
Nonsteroidal anti-in ammatory drugs (NSARDs) are a class of widely used drugs with antipyretic, analgesic and anti-in ammatory effects and it is often used in the clinical treatment of KOA, rheumatoid arthritis and other in ammatory diseases (8). It mainly inhibits the release and production of prostaglandin synthetase (PGs) by inhibiting the activity of cox-oxidase (COX) and other pathways, thus exerting the anti-in ammatory effect (9). In addition, NSARDs can also adjust the function of osteoblasts and osteoclasts by inhibiting PGs through the above pathways (10). However, there is little evidence that they can alter or stop the progression of KOA (11). Due to its general need for long-term oral, the resulting toxic side effects such as gastrointestinal reactions, cardiovascular side effects can not be ignored (12,13). Therefore, the indications for its application are open to question, and many mechanisms are unknown. NSARDs last long in synovial uid, and considerable concentrations of drugs can be achieved in synovial uid (14). However, little research has been done on whether NSARDs play an intervention role in the expression of abnormal genes in synovial tissues of KOA patients and the details of its mechanism. Through the bioinformatics analysis, we tried to identify the genes and pathways of NSARDs acting on synovia from women with KOA. We downloaded the gene microarray datasets with the accession number of GSE55457 and GSE55584 from the Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/) database, performed robust multi-array (RMA) normalization with affy R programming package, and differential expression genes (DEGs) in two groups were identi ed based on limma package separately for main analysis and cooperative analysis (15). Through our further analysis, to provide reference for clinical application.

Materials And Methods
Microarray data.
We downloaded the gene microarray datasets with the accession number of GSE55457 and GSE55584 from the GEO database, including 14 samples of synovia from women with KOA and 2 samples of synovia from women without KOA. Among them, there were 5 KOA patients in the untreated KOA group (untreated KOA) (GSM1337327, GSM1337330 from GSE55457 dataset, and GSM1339628, GSM1339629, GSM1339632 from GSE55584 dataset), 9 KOA patients in the NSARDs treated KOA group (NSARDs treated KOA) (GSM1337328, GSM1337329, GSM1337331, GSM1337334, GSM1337335, GSM1337336 from GSE55457 dataset, and GSM1339630, GSM1339631, GSM1339633 from GSE55584 dataset), and 2 patients without KOA in the normal control group (normal control) (GSM1337306 and GSM1337310 from GSE55457 dataset).The samples in the untreated KOA group and the NSARDs treated KOA group were used for main analysis. To repeat the grouping of BOBIN MI,the samples in the untreated KOA group and the normal control group were used for cooperative analysis.
Identi cation of DEGs.
To identify DEGs for main analysis and cooperative analysis, we did the following several times. We performed RMA normalization with affy R programming package. After that, DEGs in two groups were identi ed based on limma package separately for main analysis and cooperative analysis. The screening criteria for DEGs were: the value of gene expression fold change (FC) ≥ 1.4 or ≤ 1/1.4, and P < 0.05.
Screening for common DEGs.
Screening the common differentially expressed genes (DEGs) from main analysis and cooperative analysis.
Gene ontology(GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis for common DEGs.
Enriched GO terms and KEGG pathways of common DEGs were obtained through the Database for Annotation, Visualization and Integrated Discovery (DAVID). For screening results, P < 0.05 and Associated genes count > 5 were used as the screening conditions. PPI network analysis for common DEGs. PPI network was constructed by using the web-based tool STRING (http://www.string-db.org). Subsequently, the PPI network was visualized by using Cytoscape software (http://www.cytoscape.org/).

Identi cation of DEGs for main analysis
The untreated KOA group and the NSARDs treated KOA group had good clustering and no outlier samples, and all the 14 samples could be used for further analysis. FC ≥ 1.4 or ≤ 1/1.4,and P < 0.05 were used as the screening conditions. There were 338 DEGs in total between the untreated KOA group and the NSARDs treated KOA group. Among them, 211 genes were up-regulated and 127 genes were downregulated. (Table 1 and Table 2).  Identi cation of DEGs for cooperative analysis.
The untreated KOA group and the normal control group had good clustering and no outlier samples, and all the 7 samples could be used for further analysis. FC ≥ 1.4 or ≤ 1/1.4,and P < 0.05 were used as the screening conditions. There were 7005 DEGs in total between the normal control group and untreated KOA. Among them, 6952 genes were up-regulated and 53 genes were down-regulated. (Table 3 and   Table 4).  Screening for common DEGs.
Screening the common DEGs from main analysis and cooperative analysis. There were 129 common DEGs in total between main analysis and cooperative analysis. (Fig. 1).
GO and pathway enrichment analysis for common DEGs.
For screening results, P < 0.05 and Associated genes count > 5 were used as the screening conditions. There were 7 results that have been ltered out. Among them, there are 2 biological processes(BP), 3 cell components(CC) and 2 molecular functions(MF) for the enrichment of differentially expressed genes. (Table 5). There was 0 KEGG pathway enriched by differentially expressed genes. PPI network analysis for common DEGs.

Discussion
In this study, the discussion of DEGs, related pathways and PPIs in synovial tissue of female KOA patients after NSARDs treatment is helpful to strengthen the understanding of researchers on KOA and NSARDs. In this study, CD24, CD70, KCNC1, BIRC3, TNFRSF11A, CSF1R, HP, CD69, CD40LG, and IL6 may play an important role in the synovial tissue of female KOA patients after NSARDs treatment. In addition, NSARDs treatment will also affect a series of pathways: "positive regulation of cell proliferation", "negative regulation of apoptotic process", "cell surface", "integral component of plasma membrane", "plasma membrane", and "protein kinase Binding", "ATP binding". IL-6 is a soluble mediator that has multipotency effects on in ammation, immune response and hematopoiesis (16). It can not only stimulate the aggregation and activation of in ammatory cells, but also promote the formation of pain of the knee joint after trauma (17). IL-6 can mediate cartilage destruction (18). By stimulating synovial cells, IL-6 can produce PGs, which further aggravates joint in ammation, activates immature osteoclasts, and makes them participate in bone resorptive, causing progressive destruction of articular cartilage (18, 19).
TNFRSF11A is a member of the tumor necrosis factor receptor superfamily, which interacts with a variety of TRAF family proteins to induce the activation of NF-Kappa B and MAPK signaling pathway, and is also an important mediator of osteoclast development (20)(21)(22)(23). CSF1 can stimulate CSF1R to promote the proliferation and differentiation of macrophages and the formation of osteoclasts, so in patients with KOA, inhibition of CSF1R is bene cial (24)(25)(26).
The limitation of this study is that only a small number of synovium samples were included, and the inclusion of other samples may change the current results. Therefore, it is necessary to collect more synovial samples from female patients with KOA to detect the expression levels of signi cant DEGs. In addition, the signi cance of PTGS1 and PTGS2 changes was relatively low, which was surprising.
In conclusion, NSAIDs may play a certain role in synovium of female KOA by regulating the mRNA expression of IL-6, TNFRSF11A and CSF1R, which may become one of the indicators for monitoring the e cacy of NSAIDs. Given their reported side effects, they should be appropriately recommended.

Figure 1
Screening for common DEGs from main analysis and cooperative analysis DEGs 1 for main analysis; DEGs 2 for cooperative analysis.