Identification of SOCS3 and PTGS2 as new biomarkers for the diagnosis of gout by cross-species comprehensive analysis

Abstract

Background: Gout is the most common inflammatory arthritis in adults. Gout is an arthritic disease caused by the deposition of monosodium urate crystal (MSU) in the joints, which can lead to acute inflammation and damage adjacent tissue. Hyperuricemia is the main risk factor for MSU crystal deposition and gout. With the increasing burden of gout disease, the identification of potential biomarkers and novel targets for diagnosis is urgently needed.

Methods: For the analysis of this subject paper, we downloaded the human gout data set GSE160170 and the gout mouse model data set GSE190138 from the GEO database. To obtain the differentially expressed genes (DEGs), we intersected the two data sets. Using the cytohubba algorithm, we identified the key genes and enriched them through GO and KEGG. The gene expression trends of three subgroups (normal control group, intermittent gout group and acute gout attack group) were analyzed by Series Test of Cluster (STC) analysis, and the key genes were screened out, and the diagnostic effect was verified by ROC curve. The expression of key genes in dorsal root nerve and spinal cord of gout mice was analyzed. Finally, the clinical samples of normal control group, hyperuricemia group, intermittent gout group and acute gout attack group were collected, and the expression of key genes at protein level was verified by ELISA.

Result: We obtained 59 co-upregulated and 28 co-downregulated genes by comparing the DEGs between gout mouse model data set and human gout data set. 7 hub DEGs(IL1B, IL10, NLRP3, SOCS3, PTGS2) were screened out via Cytohubba algorithm. The results of both GO and KEGG enrichment analyses indicate that 7 hub genes play a significant role in regulating the inflammatory response, cytokine production in immune response, and the TNF signaling pathway. The most representative hub genes SOCS3 and PTGS2 were screened out by Series Test of Cluster, and ROC analysis results showed the AUC values were both up to 1.000. In addition, we found that PTGS2 expression was significantly elevated in the dorsal root ganglia and spinal cord in monosodium urate(MSU)-induced gout mouse model. The ELISA results revealed that the expression of SOCS3 and PTGS2 was notably higher in the acute gout attack and intermittent gout groups compared to the normal control group. This difference was statistically significant, indicating a clear distinction between the groups.

Conclusion: Through cross-species comprehensive analysis and experimental verification, SOCS3 and PTGS2 were proved to be new biomarkers for diagnosing gout and predicting disease progression.

Keywords: Biomarkers; Cross-species; Diagnostic; Gout; PTGS2; SOCS3.

Fig. 1

Flow chart.

Fig. 2

Identification of DEGs. (A, B): Volcano plots of upregulated DEGs in both human gout samples (GSE160170) and mice gout samples (GSE190138). The green and red colors respectively represent that compared to control groups, the gene expression state in gout group was down and up. (C): Venn diagram of co-DEGs in both human and mice gout samples. (D): Heatmap of co-DEGs and correlated pathways by GO analysis. (E): Heatmap of co-DEGs and correlated pathways by KEGG analysis. P value of <0.05 and a FDR of <0.25 were considered statistically significant. The top 10 KEGG pathways were shown based on enrichment score (−log10 [p-value]). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

PPI (Protein-protein interaction) network and identification of key genes. (A): PPI network of co-DEGs in both human and mice gout samples. (B) The details of genes clustered in one module was exhibited by MCODE. (C) An Upset plot displaying the intersections of key genes searched for by seven algorithms in the CytoHubba plugin. (D): The co-expression analysis of the seven hub genes by GENEMANIA plugin. (E): Expressions of seven hub genes in human gout samples (GSE160170).

Fig. 4

Further exploration of key genes. (A) Gene Ontology (GO) enrichment results of seven key genes. P value of <0.05 and a FDR of <0.25 were considered statistically significant. (B) Bar chart of seven key genes by KEGG analysis. The top 10 KEGG pathways were shown based on enrichment score (−log10 [p-value]). (C) ROC analysis of seven key genes. (D) Heatmap for expression correlation analysis of seven key genes. (E) Selection of the optimal l value. (F) Least absolute shrinkage and selection operator (LASSO) regression of seven key genes.

Fig. 5

Further exploration of subgroups of acute and chronic gout in human gout samples (GSE160170). (A) Sample box plots after standardization and normalization. (B) PCA analysis based on gene expression. (C) Correlation heatmap based on gene expression level. (D) Gene clustering trend chart for all genes. All genes are clustered into 12 categories based on gene expression. (E) Analysis of SOCS3 gene expression subgroups. (F) Analysis of PTGS2 gene expression subgroups.

Fig. 6

Analysis of the expression of SOCS2 and PTGS2 genes in mice gout samples (GSE190138). (A) ROC analysis of SOCS2 in ankle joint samples. (B) ROC analysis of PTGS2 in ankle joint samples. (C,D) Expression of (C) SOCS3 and PTGS2 (D) in the dorsal root ganglion samples. (E,F) Expression of (E) SOCS3 and PTGS2 (F) in the spinal cord samples.

Fig. 7

Correlation between seven key genes and 28 immune cells. (A) Differential analysis of 28 types of immune cell infiltration between normal and disease samples. (B) Correlation heatmaps of 28 immune cells. (C) Correlation heatmap between key genes and 29 types of immune cells.

Fig. 8

Verification of the expression of SOCS3 and PTCS2 genes. (A,B) ROC analysis shows the diagnostic value of SOCS3 and PTCS2. (C,D) Comparison of ROCS2 and PTGS2 expression between hyperuricemia, acute gout, intermittent gout and normal samples.NC, normal control;HCG, hyperuricemia group;IGG, intermittent gout group;AGG, acute gout attack group.