Identification and Clinical Validation of Key Extracellular Proteins as the Potential Biomarkers in Relapsing-Remitting Multiple Sclerosis

Abstract

Background: Multiple sclerosis (MS) is a demyelinating disease of the central nervous system (CNS) mediated by autoimmunity. No objective clinical indicators are available for the diagnosis and prognosis of MS. Extracellular proteins are most glycosylated and likely to enter into the body fluid to serve as potential biomarkers. Our work will contribute to the in-depth study of the functions of extracellular proteins and the discovery of disease biomarkers.

Methods: MS expression profiling data of the human brain was downloaded from the Gene Expression Omnibus (GEO). Extracellular protein-differentially expressed genes (EP-DEGs) were screened by protein annotation databases. GO and KEGG were used to analyze the function and pathway of EP-DEGs. STRING, Cytoscape, MCODE and Cytohubba were used to construct a protein-protein interaction (PPI) network and screen key EP-DEGs. Key EP-DEGs levels were detected in the CSF of MS patients. ROC curve and survival analysis were used to evaluate the diagnostic and prognostic ability of key EP-DEGs.

Results: We screened 133 EP-DEGs from DEGs. EP-DEGs were enriched in the collagen-containing extracellular matrix, signaling receptor activator activity, immune-related pathways, and PI3K-Akt signaling pathway. The PPI network of EP-DEGs had 85 nodes and 185 edges. We identified 4 key extracellular proteins IL17A, IL2, CD44, IGF1, and 16 extracellular proteins that interacted with IL17A. We clinically verified that IL17A levels decreased, but Del-1 and resolvinD1 levels increased. The diagnostic accuracy of Del-1 (AUC: 0.947) was superior to that of IgG (AUC: 0.740) with a sensitivity of 82.4% and a specificity of 100%. High Del-1 levels were significantly associated with better relapse-free and progression-free survival.

Conclusion: IL17A, IL2, CD44, and IGF1 may be key extracellular proteins in the pathogenesis of MS. IL17A, Del-1, and resolvinD1 may co-regulate the development of MS and Del-1 is a potential biomarker of MS. We used bioinformatics methods to explore the biomarkers of MS and validated the results in clinical samples. The study provides a theoretical and experimental basis for revealing the pathogenesis of MS and improving the diagnosis and prognosis of MS.

Keywords: bioinformatics analysis; biomarkers; extracellular protein; protein-protein interactions; relapsing-remitting multiple sclerosis.

Figure 1

Flow chart of the study.

Figure 2

Analysis of gene expression correlation and differential gene expression between MS group and control group in the dataset. (A) Boxplot of gene probe expression levels among samples. There was no significant difference in the median and the upper and lower quartile. (B) Correlation heatmap between samples. Compared with the control group, the intra-group correlation in the MS group was stronger. (C) PCA principal-component analysis. The center points of the MS group and the control group are far apart in space, indicating that the principal components are different. (D) Volcano map of all DEGs in the MS group and the control group analyzed by the limma R package. The top 10 up-regulated and down-regulated genes with the smallest P-value are marked on the map. (E) Heatmap of all DEGs in MS group and control group.

Figure 3

Screening of differentially expressed genes encoding extracellular proteins. (A) The genes encoding extracellular proteins annotated in the HPA database were intersected with DEGs, 69 EP-DEGs were screened out. The genes encoding extracellular proteins annotated in the Uniprot database were intersected with DEGs, 132 EP-DEGs were screened out. The genes screened by the two methods were combined to obtain a total of 133 EP-DEGs. (B) Volcano map of EP-DEGs in MS group and control group. Mark the top 10 up-regulated and down-regulated genes with the smallest P-value. (C) Heatmap of the top 30 up-regulated and down-regulated EP-DEGs.

Figure 4

GO enrichment of EP-DEGs. The dotplots show the Top5 processes enriched by EP-DEGs in BPs, CCs and MFs.

Figure 5

Circle graph in GO enrichment of EP-DEGs. (A–C) The circle graph shows the EP-DEGs enriched in the Top5 GO categories of BPs, CCs, and MFs, respectively. The yellow points represent the GO categories, the color of the line delivered by a point indicates the category of the point in the legend, the size of a point indicates the number of the genes it includes.

Figure 6

KEGG enrichment analysis of EP-DEGs. (A, B) respectively show the pathways to which up-regulated genes and down-regulated genes are enriched.

Figure 7

Construction of PPI network of EP-DEGs and screening of hub genes. (A) The STRING database is used to construct the PPI network of EP-DEGs, with 85 nodes and 185 edges (the legend is in the Supplementary Material ). (B) The node gene cluster with the highest score constructed by the MCODE plug-in in Cytoscape consists of 9 genes. (C) The Cytohubba is used to construct the Top10 hub genes. The figure shows the Top10 hub genes constructed by the MCC method. (D) The Cytohubba was used to predict the first stop node genes that interact with IL17A. A total of 16 genes were predicted, 10 up-regulated and 6 down-regulated.

Figure 8

Levels of IL17A, Del-1, and resolvinD1 in CSF of RRMS patients and their correlation with clinical data. (A–C) Del-1 and resolvinD1 levels were elevated in RRMS patients, and IL17A level were reduced in RRMS patients. (D–F) ResolvinD1 was positively correlated with Del-1, resolvinD1 was negatively correlated with protein and IgA. ***p < 0.001.

Figure 9

ROC, relapse-free survival, progression-free survival curves of Del-1 in RRMS. (A) The area under the ROC curve of Del-1 (AUC=0.947), which is higher than that of IgG (AUC=0.740). (B) The median relapse-free survival time was 30 months in the high Del-1 group, the median relapse-free survival time was 13.5 months in the low Del-1 group, the difference was statistically significant (P=0.044). (C) The probability of progression-free survival in the high Del-1 group was always higher than 50% during the follow-up period, and the median progression-free survival time in the low Del-1 group was 46 months, the difference was statistically significant (P=0.034).