Identification of Four Biomarkers of Human Skin Aging by Comprehensive Single Cell Transcriptome, Transcriptome, and Proteomics

Abstract

Background: Aging is characterized by the gradual loss of physiological integrity, resulting in impaired function and easier death. This deterioration is a major risk factor for major human pathological diseases, including cancer, diabetes, cardiovascular disease and neurodegenerative diseases. It is very important to find biomarkers that can prevent aging. Methods: Q-Exactive-MS was used for proteomic detection of young and senescence fibroblast. The key senescence-related molecules (SRMs) were identified by integrating transcriptome and proteomics from aging tissue/cells, and the correlation between these differentially expressed genes and well-known aging-related pathways. Next, we validated the expression of these molecules using qPCR, and explored the correlation between them and immune infiltrating cells. Finally, the enriched pathways of the genes significantly related to the four differential genes were identified using the single cell transcriptome. Results: we first combined proteomics and transcriptome to identified four SRMs. Data sets including GSE63577, GSE64553, GSE18876, GSE85358, and qPCR confirmed that ETF1, PLBD2, ASAH1, and MOXD1 were identified as SRMs. Then the correlation between SRMs and aging-related pathways was excavated and verified. Next, we verified the expression of SRMs at the tissue level and qPCR, and explored the correlation between them and immune infiltrating cells. Finally, at the single-cell transcriptome level, we verified their expression and explored the possible pathway by which they lead to aging. Briefly, ETF1 may affect the changes of inflammatory factors such as IL-17, IL-6, and NFKB1 by indirectly regulating the enrichment and differentiation of immune cells. MOXD1 may regulate senescence by affecting the WNT pathway and changing the cell cycle. ASAH1 may affect development and regulate the phenotype of aging by affecting cell cycle-related genes. Conclusion: In general, based on the analysis of proteomics and transcriptome, we identified four SRMs that may affect aging and speculated their possible mechanisms, which provides a new target for preventing aging, especially skin aging.

Keywords: fibroblast; proteomics; senescence; single cell transcriptome analysis; skin aging; transcriptome.

FIGURE 1

The analysis process of this study. We first combined proteomics and transcriptome to identified four key SRMs. Then the correlation between SRMs and aging-related pathways was excavated and validated. Next, we validated the expression of SRMs at the tissue level and qPCR, and explored the correlation between them and immune infiltrating cells. Finally, at the single-cell transcriptome level, we validated their expression and explored the possible pathway by which they lead to aging.

FIGURE 2

Identification of differentially expressed genes in transcriptome and proteome at the same time. (A). PCA analysis showed that there was no deviation from the 6 samples in the protein group; Identification of co-existing molecules in different data sets by VENN (B) and UPSET (C) analysis; (D). Volcanic map of difference analysis; (E). Identification of differentially expressed genes in transcriptome and proteome at the same time.

FIGURE 3

Explore the relationship between four SRMs and aging. (A). GSVA confirmed that there were significant differences in the pathways of senescence and cell senescence between the young group and the old group; (B). Analysis of the relationship between four SRMs and aging genes. The yellow solid line represents positive correlation, the gray dotted line represents negative correlation, and the thicker the line, the stronger the correlation.

FIGURE 4

Explore the relationship between four SRMs and immune cell. (A). The landscape of immune cell infiltration in the tissue was calculated by CIBERSORT. (B). The correlation between the expression of four SRMs and the abundance of immune cell infiltration in tissue, which was calculated by CIBERSORT. (C). The correlation between the expression of four SRMs and the abundance of immune cell infiltration in tissue, which was calculated by Xcell. (D). The correlation between the expression of four SRMs and the abundance of immune cell infiltration in tissue, which was calculated by MCPcounter. The yellow solid line represents positive correlation, the gray dotted line represents negative correlation, and the thicker the line, the stronger the correlation.

FIGURE 5

Single cell data analysis. (A). Cell specific biomarkers in skin tissue can well distinguish cell subsets. (B). The umap map shows that the integrated data containing five normal skin tissue samples are divided into 14 cell clusters by specific biomarkers of skin tissue cells. Examined the expression of four SRMs in the dermis single cell dataset GSE151177 and found that four SRMs were highly expressed in fibroblasts (C–F).

FIGURE 6

GO and KEGG enrichment analysis of genes significantly related to ETF1. (A). KEGG enrichment analysis of genes significantly related to ETF1; (B). The proportion of each pathway in the results of KEGG enrichment analysis; (C). GO-BP enrichment analysis of genes significantly related to ETF1; (D). The proportion of each pathway in the results of GO-BP enrichment analysis.

FIGURE 7

Immune pathway analysis of genes significantly related to ETF1. In order to explore the relationship between ETF1 and immune system, I conducted immune pathway enrichment analysis of genes significantly related to ETF1. The results suggest that a large number of immune-related pathways, including T cell differentiation, B cell differentiation, and immune system processed, are enriched.

FIGURE 8

qPCR technique was used to detect the expression of ETF1 (A), ASAH1 (B), MOXD1 (C), and PLBD2 (D) in fibroblasts of young and old generations, and the differences were compared. * means the p value is less than 0.05.