Identification of key biomarkers and immune infiltration in the peripheral blood of osteoarthritis by bioinformatics analysis

doi:10.21203/rs.3.rs-1650299/v1

Download PDF

Research Article

Identification of key biomarkers and immune infiltration in the peripheral blood of osteoarthritis by bioinformatics analysis

https://doi.org/10.21203/rs.3.rs-1650299/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background Osteoarthritis (OA) is a common chronic arthritis disease characterized by cartilage degeneration, sub-cartilage bone hyperplasia, osteophyte formation and joint space stenosis. Traditionally, OA treatment consists of pain management with joint replacement for end-stage disease. Patients diagnosed with OA have developed irreversible organic changes, so the focus should be shifted to prevention, early diagnosis and intervention of OA.

Methods The differentially expressed genes (DEGs) were identified from the gene expression profile GSE48556 from the GEO database, and functional enrichment analyses were performed. Boruta algorithm and the least absolute shrinkage and selection operator (LASSO) analysis were used to select hub genes, then we used the LightGBM to evaluate the accuracy of hub genes prediction for OA. The CIBERSORT algorithm was used to analyze the immune infiltration of peripheral blood mononuclear cells (PBMCs) between OA and normal controls.

Results A total of 35 DEGs including 29 downregulated genes and 6 upregulated genes were mainly involved in response to tumor necrosis factor (TNF) and TNF-mediated signaling pathway through GO analysis and were mainly enriched in cAMP signaling pathway, NF-κB signaling pathway through KEGG. The LightGBM prediction of the hub gene model showed that area under curve (AUC) of ROC curve and PR curve reached 0.988 and 0.996 respectively in verification set, indicating that hub genes have a high diagnostic accuracy for OA. Gene ADRB2 was the only up-regulated gene in OA group. H3F3B showed the largest AUC which means a high diagnostic power of OA. From the immune infiltration analysis, the content of CD8⁺ T cells and resting mast cells showed significant differences between the two groups and hub genes were closely related to immune cells.

Conclusion The hub genes in PBMCs between OA and normal controls may potentially serve as biomarkers in both prevention and diagnosis of OA and indicate for the exploration of the mechanism of OA.

osteoarthritis

peripheral blood mononuclear cells

immune infiltration

bioinformatics analysis

Osteoarthritis (OA) is a chronic degenerative joint disease that affects the quality of life of the elderly, which clinical manifestations are chronic pain, swelling, malformation, and stiffness of the joint^[¹^]. According to the World Health Organization, approximately 9.6% of men and 18% of women over 60 years old suffer from OA, and 25% of OA patients have disabilities^[²^]. OA is a progressive disorder that is characterized by the degradation of hyaline articular cartilage, along with subchondral sclerosis, narrowing of the joint space, osteophyte formation, synovial hyperplasia, and structural alterations of peripheral muscles and ligaments^[³^,⁴^].

In the clinic, symptoms are the prompt that leads patients to seek medical attention, and the diagnosis is made by case history, physical examination, and imaging. Traditionally, OA treatment consists of pain management and joint replacement for end-stage disease^[^5-7^]. When OA patients are diagnosed, they have already developed irreversible organic changes, so the focus should be shifted to prevention, early diagnosis and intervention of OA^[⁸^]. Many proinflammatory cytokines have been identified in the synovial fluid of OA patients, including IL-6, MCP-1, IP-10, and CXCL9^[⁹^], meanwhile, macrophages mainly infiltrate in the synovial tissue of OA patients^[¹⁰^]. Therefore, the occurrence and development of OA is closely related to inflammation and immune infiltration.

To integrate and analyze biological data, bioinformatics, an interdisciplinary field that develops methods and software tools, has been widely employed in search of the differential expression of disease-related genes and biomarkers. We selected hub genes from differentially expressed genes (DEGs) by Boruta algorithm and he least absolute shrinkage and selection operator (LASSO) regression analysis. Boruta is a feature selection algorithm and feature ranking based on the Random Forest (RF) algorithm. Boruta’s benefits are to decide the significance of a variable and to assist the statistical selection of important variables. LASSO regression analysis is a shrinkage method that can actively select from a large and potentially multicollinear set of variables in the regression, resulting in a more relevant and interpretable set of predictors^[¹¹^]. Then we used the light gradient boosting machine (LightGBM) to evaluate the accuracy of hub genes prediction for OA. The LightGBM is an ensemble model of decision trees for classification and regression prediction. The algorithm has exceptional performance in terms of accuracy, model stability and computational efficiency^[¹²^]. The immune infiltration in the peripheral blood was analyzed by performing the CIBERSORT algorithm.

In conclusion, OA is a systemic disease, so it is possible to detect the gene expression profile and immune infiltration of peripheral blood mononuclear cells (PBMCs) that may potentially serve as biomarkers in both prevention and diagnosis of OA and indicate for the exploration of the mechanism of OA.

Raw data acquisition and preprocessing

Gene expression profiles of GSE48556 were downloaded from the GEO database (GPL6947 Illumina HumanHT-12 V3.0 expression beadchip) contain 139 samples, including 33 PBMCs samples from non-OA patients (normal controls) and 106 PBMCs samples from OA patients (OA groups). R software (version 4.1.3) and “Bioconductor” packages were used in the data analyses. Approval by an ethics committee was not necessary because all data were collected from publicly available databases.

Identiﬁcation of DEGs

The linear models for microarray data “limma” package in Bioconductor was used to identify DEGs by comparing the expression values between PBMCs from OA patients and normal controls. The corresponding p value of the gene symbols after t test were used, and adjusted p < 0.05 and |logFC| > 0.5 were used as the selection criteria. The “pheatmap” package was used to draw a heatmap of the DEGs in R software and the results were also visualized as a volcano plot using the “ggplot2” package. DEGs were then analyzed to obtain correlate transcription factors (TFs) by Transcriptional Regulatory Relationships Unraveled by Sentence-based Text mining (TRRUST, http://www.grnpedia.org/trrust/).

Gene ontology (GO) and Kyoto Encyclopedia of genes and genomes (KEGG) enrichment analyses

GO is a major bioinformatics tool for annotating genes and analyzing their biological processes including biological process (BP), cellular component (CC), and molecular function (MF). KEGG is a database resource for understanding high-level functions and biological systems from large-scale molecular datasets generated by high-throughput experimental technologies. The “clusterProfiler” package in Bioconductor was used to perform the GO and KEGG pathway analysis of the DEGs. P < 0.05 was considered statistically significant.

Boruta algorithm and LASSO regression analysis to select hub genes

Boruta algorithm and LASSO regression analysis were used to obtain hub genes from the above DEGs. In other words, these two algorithms were both used for data dimension reduction. First, we implemented Boruta algorithm which is a wrapper built around RF. The Boruta algorithm was based on the “Boruta” package^[¹³^]. Next, LASSO regression analysis was used. We utilized the “glmnet” package to fit the logistic LASSO regression^[¹⁴^]. We utilized ten-fold cross-validation to select the penalty term, λ. The binomial deviance was computed for the test data as measures of the predictive performance of the fitted models. We opted for the λ that minimizes the binomial deviance.

Diagnostic efficiency validation by the LightGBM algorithm

We used the LightGBM algorithm to evaluate the accuracy of hub genes prediction for OA. Here, the LightGBM was optimized by calculating the optimal values of the hyperparameters using 10-fold cross-validation. We split the “GSE48556” dataset into training set and validation set by random splitting. Receiver operating characteristic (ROC) curves and Precision-Recall (PR) curves were used to evaluate the sensitivity, specificity, recall and precision of the gene prediction model. We utilized the “lightgbm” package to establish the LightGBM model.

Immune infiltration analysis by CIBERSORT

The CIBERSORT algorithm is a widely applied method for calculating the proportions of 22 types of leukocytes in complex tissues profiled by microarray^[¹⁵^]. In this study, all samples were screened with a p value < 0.05, and CIBERSORT analysis was performed to characterize immune cell infiltration in the peripheral blood. The composition of immune cells was visualized using a violin plot, and the infiltration levels of each cell between patients with OA and healthy controls were analyzed using the “vioplot” package. In addition, the intersection between immune infiltration and hub genes was showed by heatmap.

Identiﬁcation of DEGs

After processing the raw microarray results from GSE48556, DEGs were identified. A total of 35 differentially expressed genes including 29 downregulated genes and 6 upregulated genes were detected. The results were validated with a volcano plot of all downregulated genes and upregulated genes (Figure 1A). Fig. 1B shows the DEGs expression heatmap. 8 related TFs including PGR, RELA, NFKB1, ATF2, STAT6, PPARA, HDAC1 and TP53 were obtained from the TRRUST (Table 1).

Functional enrichment analysis of the candidate DEGs

GO analysis showed that in terms of BP, the candidate DEGs were significantly enriched in response to temperature stimulus, response to tumor necrosis factor (TNF) and TNF-mediated signaling pathway. With respect to CC, the candidate DEGs were mainly enriched in nuclear speck, focal adhesion and cell-substrate junction. With respect to MF, the candidate DEGs were mainly enriched in ubiquitin protein ligase binding and ubiquitin-like protein ligase binding (Table 2, Figure 2A). KEGG pathway analysis revealed that the DEGs were mainly enriched in cAMP signaling pathway, IL-17 signaling pathway, NF-κB signaling pathway, C-type lectin receptor signaling pathway and TNF signaling pathway (Table 3, Figure 2B).

Hub genes selection

In order to filtrate hub genes from DEGs, 14 hub genes were screen out from 35 DEGs by Boruta algorithm and LASSO regression analysis (Figure 3A). The most important thing for Lasso regression is to select a suitable λ value, which can be realized by “cv.glmnet” function. The built-in function in R produces two automatic λ’s——one that minimizes the binomial deviance and one representing largest λ that is still within 1 standard error of the minimum binomial deviance. We opted for the former λ to ensure the least binomial deviance. For the present analyses, the λ values ranged from 0.00002878 to 0.26225840 with a minimal binomial deviance achieved at 0.00764497 (Figure 3B). The standard errors of the LASSO coefficients were obtained via bootstrapping within the primary sampling unit and strata^[¹⁶^].

The hub genes model was composed of H3F3B, ADRB2, NFKBIA, SNORD68, SFRS5, NT5C3, SNORD13, NBPF10, LIMS1, EIF1AY, GNL3L, LOC649841, ZNF148, RGS18 (Table 4). We compared the expression levels of these 14 genes in the control group and the OA group. As figure 4 shows, gene ADRB2 was the only up-regulated gene in the OA group, while the remaining 13 genes were down-regulated. Furthermore, ROC curves were used for each gene to show the sensitivity and specificity of prediction of OA genes (Figure 5). Among the hub genes, the largest area under curve (AUC) was 0.879 which was contributed by gene H3F3B.

GO function and KEGG pathway enrichment analysis of hub genes

As shown in Figure 6A and Table 5, the 14 identified hub biomarkers were primarily involved in some critical BPs, such as TNF-mediated signaling pathway, regulation of NIK/NF-κB signaling. The hub genes were also found to be enriched in telomerase holoenzyme complex in the CC category and peptide binding and amide binding in the MF category. Moreover, the results of KEGG enrichment analysis indicated that the selected hub genes played a critical role in cAMP signaling pathway, regulation of lipolysis in adipocytes and cytosolic DNA-sensing pathway (Figure 6B and 6C, Table 6).

The LightGBM’s prediction of the hub gene model

AUC was the area under ROC curve or PR curve. Due to its intuitiveness, we used AUC value as the evaluation index of the model. The AUC value of the training set almost reaches 1. However, the results of the verification set are more important. The results of the verification set showed that AUC of ROC curve and PR curve reached 0.988 and 0.996 respectively, indicating that hub genes have a high diagnostic accuracy for OA (Figure 7).

Immune infiltration analysis by CIBERSORT

Two immune cells including CD8⁺ T cells and resting mast cells were detected which showed statistically significant differences in immune infiltration between OA group and normal group (p = 0.019, p = 0.025, respectively) (Figure 8A). CD8⁺ T cells showed statistically significant negative correlation with hub genes including NFKBIA, LIMS1, EIF1AY, SNORD68 and SFRS5 and positive correlation with ADRB2 which was the only upregulated gene. Resting mast cells showed statistically significant negative correlation with H3F3B, EIF1AY and NT5C3. Although several immune cells showed no statistically significant differences between normal group and OA group, they had significant correlation with 14 hub genes (Figure 8B).

With the aging of the population and chronic sports injuries, osteoarthritis has become a prominent problem affecting the quality of life of the elderly. Currently, most diagnostic OA methods are only based on images and patients’ symptoms; thus, most OA patients are not diagnosed early, and late and terminal stage of patients can only undergo surgery to rehabilitate their daily activity. The early diagnosis role of the peripheral blood in OA is often ignored by doctors. However, the biomarkers of PBMCs not only affects the progression OA, but can also serve as early diagnostic biomarkers. From this study, we can provide some useful novel early biomarkers for diagnosis even as the therapy targets for OA patients.

Some of the previous OA studies emphasized articular cartilage^[¹⁷^] and synovium^[³^,¹⁸^,¹⁹^] ignored the role of PBMCs in the diagnosis of OA. In this study, we tried to establish diagnosis model by identifying the hub genes related to PBMCs in OA. 14 hub genes were screened out by bioinformatics analysis, and the biological functions and signaling pathways involved were analyzed in the present study. We found that these hub genes and the enriched pathways were involved in many immune responses, so we analyzed the immune cell infiltration by the CIBERSORT algorithm method, and the results showed that there was a significant difference between OA PBMCs and normal controls.

As a result, 14 genes, including the human replacement histone H3.3B gene (H3F3B), β₂-adrenergic receptor (ADRB2), nuclear factor-kappa-B-inhibitor alpha (NFKBIA) Small Nucleolar RNA, C/D Box 68 (SNORD68), serine and arginine rich splicing factor 5 (SFRS5), 5'-nucleotidase, cytosolic III (NT5C3), small nucleolar RNA, C/D box 13 (SNORD13), neuroblastoma breakpoint family member 10 (NBPF10), LIM zinc finger domain containing 1 (LIMS1), eukaryotic translation initiation factor 1A Y-linked (EIF1AY), G protein nucleolar 3 like (GNL3L), LOC649841, zinc finger protein 148 (ZNF148), regulator of G protein signaling 18 (RGS18) were identified as differentially expressed hub biomarkers.

Gene ADRB2 was the only up-regulated gene in the OA group and was associated with a variety of immune cells. ADRB2 was positively correlated with CD8⁺ T cells, γδ T cells, activated DC cells, and negatively correlated with monocytes, M₀ macrophages, and memory resting CD4⁺ T cells. On this basis, we wanted to further explore the relationship between this gene and OA. The gene ADRB2 encodes the β₂-adrenergic receptor (β₂-AR) and β-adrenergic receptor subtypes were detected in various tissues related to OA development such as synovium, cartilage, and subchondral bone^[²⁰^,²¹^]. In the temporomandibular joint (TMJ) OA model, activation of the β₂-AR resulted in subchondral bone loss^[²²^]. Furthermore, conditional deletion of ADRB2 in mesenchymal stem cells attenuates osteoarthritis-like defects in TMJ^[²³^]. β₂-AR plays a major role in OA-related cartilage calcification and subchondral bone changes but the contribution of other AR subtypes should be taken into consideration^[²⁴^]. Hence, the underlying mechanism of the relationship between ADRB2 and OA needs to explore in the long run.

Among all the hub genes, H3F3B showed the largest AUC with 0.879, which means the gene has a high diagnostic power of OA. Resting mast cells and monocytes showed statistically significant positive correlation with H3F3B, while naive CD4⁺ T cells were negatively correlated. The first comprehensive exploration of gene expression in peripheral blood linked osteoarthritis with apoptotic pathways^[²⁵^] and H3F3B was among the top genes identified. miR-10a-5p/H3F3B axis is regulated to suppress chondrocytes apoptosis in OA, and H3F3B were downregulated in OA cartilage tissues^[²⁶^]. Behjati et al.^[²⁷^] found H3F3B mutated in 73/77 (95%) cases of chondroblastoma, which is one of two genes encoding histone 3.3. Nevertheless, the association between H3F3B and OA pathogenesis remains unknown and needs to further study.

Through GO analysis, we found that the hub genes were mainly involved in response to TNF and TNF-mediated signaling pathway. Elevated level of TNF-α has been found in the synovial fluid, synovial membrane, subchondral bone and cartilage of OA patients^[²⁸^,²⁹^], confirming their important roles in OA pathogenesis^[³⁰^]. It plays a pivotal role in cartilage matrix degradation and bone resorption in OA^[³¹^]. Obese people tended to suffer from OA and showed a higher level of serum TNF-α produced by macrophages derived from adipose tissue^[³²^]. TNF-α was the factor released by fat to negatively regulate cartilage directly. Besides, inflammatory mediators, such as TNF-α was highly upregulated in OA joints and induced ROS production and expression of matrix degrading proteases leading to matrix degradation and joint dysfunction^[³³^].

Based on the KEGG database, 14 hub genes were mainly enriched in cAMP signaling pathway, NF-κB signaling pathway and other signaling pathways. cAMP signaling pathway is an anti-aging target which is a multi-unit protein kinase that mediates signal transduction of G-protein-coupled receptors through its activation by adenyl cyclase-mediated cAMP^[³⁴^]. A systematic evaluation based on Han Chinese population identified significant association signals from NFKBIA with hip OA for the first time^[³⁵^] and NF-κB signaling pathway could be regulated to prevent cartilage degeneration in OA^[³⁶^]. Furthermore, aging contributed to the initiation and progression of osteoarthritis and age was statistically negatively correlated with the expression of NFKBIA^[³⁷^]. Therefore, more studies are needed to define the function of NFKBIA in the pathomechanism of OA. In addition, it has been demonstrated that NF-κB plays an important role in mediating the effects of TNF-α and IL-1β on the expression of type II collagen^[³⁸^].

From the immune infiltration analysis, we found that there was a significant difference in the relative cell content of CD8⁺ T cells and resting mast cells between the two groups. In peripheral blood immune cell composition in OA, CD8⁺ T cells frequency were higher compared to healthy controls, CD8⁺ memory-like cells were more likely to be found in OA and increased frequency of CD8⁺ T cell related to inflammation were also present in OA^[³⁹^]. In accordance with the previous studies, CD8⁺ T cells infiltrated the synovium after OA induction^[⁴⁰^]. CD8⁺ T cells may act as a primary activator to exacerbate synovitis and/or a secondary regulator to destruct the cartilage^[⁴¹^]. The effects of NE on memory CD8⁺ T cells were primarily mediated by ADRB2^[⁴²^]. Memory CD8⁺ T cells expressed a significantly higher level of ADRB2 compared to naive cells, which had a faster and stronger immune response. The content of mast cells in OA patients was less than that in normal people. Due to little literature reported the association between OA and resting mast cells in peripheral blood, the results in synovial fluids (SF) might have something in common. Despite the low number of MC in SF (range 0-1.2%), there is a growing body of evidence for their crucial role in OA pathogenesis^[⁴³^].

Nonetheless, despite these findings, the study still has some limitations. First of all, this study was a retrospective analysis based on a public database, so more prospective approaches are required to substantiate the results. Secondly, the results we discovered through bioinformatics are lacking experimental validation. Finally, Mechanistic insight into the hub genes and their connections with related pathways and immune cells should be investigated.

In the present study, a total of 35 DEGs and 14 hub genes were identified. The LightGBM prediction of the hub gene model showed that AUC of ROC curve and PR curve reached 0.988 and 0.996 respectively in verification set, indicating that hub genes have a high diagnostic accuracy for OA. The biological functions and pathways of the identified genes provide a more detailed molecular mechanism for diagnosis of OA. Gene ADRB2 was the only up-regulated gene in the OA group and gene H3F3B has the highest diagnostic power of all hub genes. Hub genes were mainly involved in response to TNF and TNF-mediated signaling pathway through GO analysis. Based on the KEGG database, hub genes were mainly enriched in cAMP signaling pathway, NF-κB signaling pathway and other signaling pathways. From the immune infiltration analysis, we found that the content of CD8⁺ T cells and resting mast cells was significantly different between the two groups. However, further studies are needed to confirm the relation between the key genes and immune infiltration.

osteoarthritis (OA), differentially expressed genes (DEGs), the least absolute shrinkage and selection operator (LASSO), Random Forest (RF), the light gradient boosting machine (LightGBM), peripheral blood mononuclear cell (PBMC), transcription factor (TF), Transcriptional Regulatory Relationships Unraveled by Sentence-based Text mining (TRRUST), Gene ontology (GO), Kyoto Encyclopedia of genes and genomes (KEGG), biological process (BP), cellular component (CC), molecular function (MF), eeceiver operating characteristic (ROC), Precision−Recall (PR), tumor necrosis factor (TNF), area under curve (AUC), β₂-adrenergic receptor (β₂-AR), temporomandibular joint (TMJ), synovial fluids (SF)

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Author contributions

Tiancong Ma and Guanglei Zhao analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the paper.

Xin Huang and Jinyang Lyu analyzed the data, prepared figures and/or tables, and approved the final draft.

Siqun Wang, Yibing Wei and Gangyong Huang authored or reviewed drafts of the paper.

Jun Xia conceived the study, authored or reviewed drafts of the paper, and approved the final draft.

All authors read and approved the final manuscript.

Competing Interests

The authors declare that they have no competing interests.

Funding

Not applicable.

Availability of data and materials

The data is available at NCBI GEO (https://www.ncbi.nlm.nih.gov/geo): GSE48556.

Author details

¹Department of Orthopaedic Surgery, Huashan Hospital Fudan University, 12th Wulumuqi Middle Road, Jing’an District 200040, Shanghai, China

²Department of Orthopaedic Surgery, North Branch of Huashan Hospital Fudan University, 518th Jingpohu Road, Bao’shan District 200030, Shanghai, China

Li, Z.C., et al., Functional annotation of rheumatoid arthritis and osteoarthritis associated genes by integrative genome-wide gene expression profiling analysis. PLoS One, 2014. 9(2): p. e85784.
Li, M.H., et al., Regenerative approaches for cartilage repair in the treatment of osteoarthritis. Osteoarthritis Cartilage, 2017. 25(10): p. 1577-1587.
Mathiessen, A. and P.G. Conaghan, Synovitis in osteoarthritis: current understanding with therapeutic implications. Arthritis Res Ther, 2017. 19(1): p. 18.
Rousseau, J. and P. Garnero, Biological markers in osteoarthritis. Bone, 2012. 51(2): p. 265-77.
Carr, A.J., et al., Knee replacement. Lancet, 2012. 379(9823): p. 1331-40.
Pivec, R., et al., Hip arthroplasty. Lancet, 2012. 380(9855): p. 1768-77.
Bijlsma, J.W., F. Berenbaum, and F.P. Lafeber, Osteoarthritis: an update with relevance for clinical practice. Lancet, 2011. 377(9783): p. 2115-26.
Glyn-Jones, S., et al., Osteoarthritis. Lancet, 2015. 386(9991): p. 376-87.
Katz, J.N., K.R. Arant, and R.F. Loeser, Diagnosis and Treatment of Hip and Knee Osteoarthritis: A Review. Jama, 2021. 325(6): p. 568-578.
Felson, D.T., et al., Synovitis and the risk of knee osteoarthritis: the MOST Study. Osteoarthritis Cartilage, 2016. 24(3): p. 458-64.
Tibshirani, R., Regression Shrinkage and Selection Via the Lasso. Journal of the Royal Statistical Society: Series B (Methodological), 1996. 58(1): p. 267-288.
Yan, J., et al., LightGBM: accelerated genomically designed crop breeding through ensemble learning. Genome Biol, 2021. 22(1): p. 271.
Degenhardt, F., S. Seifert, and S. Szymczak, Evaluation of variable selection methods for random forests and omics data sets. Brief Bioinform, 2019. 20(2): p. 492-503.
Friedman, J., T. Hastie, and R. Tibshirani, Regularization Paths for Generalized Linear Models via Coordinate Descent. J Stat Softw, 2010. 33(1): p. 1-22.
Newman, A.M., et al., Robust enumeration of cell subsets from tissue expression profiles. Nat Methods, 2015. 12(5): p. 453-7.
Wolf, L.J., Computer age statistical inference: Algorithms, evidence, and data science. Environment and Planning B-Urban Analytics and City Science, 2017. 44(5): p. 986-987.
Trachana, V., et al., Understanding the role of chondrocytes in osteoarthritis: utilizing proteomics. Expert Rev Proteomics, 2019. 16(3): p. 201-213.
Wang, H., et al., Histomorphology and innate immunity during the progression of osteoarthritis: Does synovitis affect cartilage degradation? J Cell Physiol, 2018. 233(2): p. 1342-1358.
Scanzello, C.R. and S.R. Goldring, The role of synovitis in osteoarthritis pathogenesis. Bone, 2012. 51(2): p. 249-57.
Rösch, G., et al., Sympathectomy aggravates subchondral bone changes during osteoarthritis progression in mice without affecting cartilage degeneration or synovial inflammation. Osteoarthritis Cartilage, 2022. 30(3): p. 461-474.
Speichert, S., et al., Role of Norepinephrine in IL-1β-Induced Chondrocyte Dedifferentiation under Physioxia. Int J Mol Sci, 2019. 20(5).
Jiao, K., et al., β2-Adrenergic signal transduction plays a detrimental role in subchondral bone loss of temporomandibular joint in osteoarthritis. Sci Rep, 2015. 5: p. 12593.
Sun, J.L., et al., Conditional deletion of Adrb2 in mesenchymal stem cells attenuates osteoarthritis-like defects in temporomandibular joint. Bone, 2020. 133: p. 115229.
Rösch, G., et al., β2-Adrenoceptor Deficiency Results in Increased Calcified Cartilage Thickness and Subchondral Bone Remodeling in Murine Experimental Osteoarthritis. Front Immunol, 2021. 12: p. 801505.
Ramos, Y.F., et al., Genes expressed in blood link osteoarthritis with apoptotic pathways. Ann Rheum Dis, 2014. 73(10): p. 1844-53.
Jiang, H., et al., LncRNA SNHG5 promotes chondrocyte proliferation and inhibits apoptosis in osteoarthritis by regulating miR-10a-5p/H3F3B axis. Connect Tissue Res, 2021. 62(6): p. 605-614.
Behjati, S., et al., Distinct H3F3A and H3F3B driver mutations define chondroblastoma and giant cell tumor of bone. Nat Genet, 2013. 45(12): p. 1479-82.
Xue, J., et al., Tumor necrosis factor-α induces ADAMTS-4 expression in human osteoarthritis chondrocytes. Mol Med Rep, 2013. 8(6): p. 1755-60.
Stannus, O., et al., Circulating levels of IL-6 and TNF-α are associated with knee radiographic osteoarthritis and knee cartilage loss in older adults. Osteoarthritis Cartilage, 2010. 18(11): p. 1441-7.
Wang, T. and C. He, Pro-inflammatory cytokines: The link between obesity and osteoarthritis. Cytokine Growth Factor Rev, 2018. 44: p. 38-50.
Klooster, A.R. and S.M. Bernier, Tumor necrosis factor alpha and epidermal growth factor act additively to inhibit matrix gene expression by chondrocyte. Arthritis Res Ther, 2005. 7(1): p. R127-38.
Park, H.S., J.Y. Park, and R. Yu, Relationship of obesity and visceral adiposity with serum concentrations of CRP, TNF-alpha and IL-6. Diabetes Res Clin Pract, 2005. 69(1): p. 29-35.
Ansari, M.Y., N. Ahmad, and T.M. Haqqi, Oxidative stress and inflammation in osteoarthritis pathogenesis: Role of polyphenols. Biomed Pharmacother, 2020. 129: p. 110452.
Enns, L.C. and W. Ladiges, Protein kinase A signaling as an anti-aging target. Ageing Res Rev, 2010. 9(3): p. 269-72.
Tang, H., et al., TLR10 and NFKBIA contributed to the risk of hip osteoarthritis: systematic evaluation based on Han Chinese population. Sci Rep, 2018. 8(1): p. 10243.
Wang, C., et al., Safflower yellow alleviates osteoarthritis and prevents inflammation by inhibiting PGE2 release and regulating NF-κB/SIRT1/AMPK signaling pathways. Phytomedicine, 2020. 78: p. 153305.
Rai, M.F., et al., Relationship of age and body mass index to the expression of obesity and osteoarthritis-related genes in human meniscus. Int J Obes (Lond), 2013. 37(9): p. 1238-46.
Séguin, C.A. and S.M. Bernier, TNFalpha suppresses link protein and type II collagen expression in chondrocytes: Role of MEK1/2 and NF-kappaB signaling pathways. J Cell Physiol, 2003. 197(3): p. 356-69.
Ponchel, F., et al., Changes in peripheral blood immune cell composition in osteoarthritis. Osteoarthritis Cartilage, 2015. 23(11): p. 1870-8.
Scanzello, C.R., et al., Local cytokine profiles in knee osteoarthritis: elevated synovial fluid interleukin-15 differentiates early from end-stage disease. Osteoarthritis Cartilage, 2009. 17(8): p. 1040-8.
Hsieh, J.L., et al., CD8+ T cell-induced expression of tissue inhibitor of metalloproteinses-1 exacerbated osteoarthritis. Int J Mol Sci, 2013. 14(10): p. 19951-70.
Slota, C., et al., Norepinephrine preferentially modulates memory CD8 T cell function inducing inflammatory cytokine production and reducing proliferation in response to activation. Brain Behav Immun, 2015. 46: p. 168-79.
Kriegova, E., et al., Gender-related differences observed among immune cells in synovial fluid in knee osteoarthritis. Osteoarthritis Cartilage, 2018. 26(9): p. 1247-1256.

Tables 1 to 6 are available in the Supplementary Files section

No competing interests reported.

Table1.RelatedTFsobtainedbyTRRUST..xlsx
Table 1. Related TFs obtained by TRRUST.
Table2.GOenrichmentanalysisofDEGstop10termsofeachcategorywerelisted..xlsx
Table 2. GO enrichment analysis of DEGs (top 10 terms of each category were listed).
Table3.KEGGpathwayanalysesofDEGstop10termsofeachcategorywerelisted..xlsx
Table 3. KEGG pathway analyses of DEGs (top 10 terms of each category were listed).
Table4.Coefficientof14hubgenes..xlsx
Table 4. Coefficient of 14 hub genes.
Table5.GOenrichmentanalysesresultsofhubgenestop10accordingtopvalue..xlsx
Table 5. GO enrichment analyses results of hub genes (top 10 according to p value).
Table6.KEGGpathwayanalysesresultsofhubgenestop10accordingtopvalue..xlsx
Table 6. KEGG pathway analyses results of hub genes (top 10 according to p value).

Download PDF

Version 1

posted

You are reading this latest preprint version

Identification of key biomarkers and immune infiltration in the peripheral blood of osteoarthritis by bioinformatics analysis

Status:

Version 1

Abstract

Figures

Introduction

Methods

Results

Discussion

Conclusion

Abbreviations

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1