Integrative identification and preliminary validation of senescence-associated secretory phenotype-related candidate biomarkers in type 2 diabetes.

Abstract

Senescence-associated secretory phenotype (SASP)-related inflammatory programs have been implicated in type 2 diabetes mellitus (T2DM), yet blood-based candidates have not been systematically screened and validated across independent cohorts. We adopted standardized senescence signature scoring and multi-cohort independent validation to screen SASP-related candidate genes through literature mining, differential expression analysis, and machine learning, followed by functional enrichment analysis, immune cell composition analysis, and immune cell proportion adjustment. We identified CCL8 and MMP9 as key SASP-related candidates; MMP9 was significantly positively correlated with senescence scores, and both genes were implicated in T2DM progression through E2F targets and interferon gamma response pathways. T2DM led to significant immune microenvironment alterations, and the expression changes of CCL8 and MMP9 were independent of peripheral blood mononuclear cell (PBMC) subset frequencies after CIBERSORT adjustment. Expression patterns were consistently validated across all cohorts including the large-scale GSE184050 dataset and single-cell RNA-seq dataset GSE280401, with RT-qPCR providing exploratory support in a small cohort. Collectively, this study provides supportive evidence that CCL8 and MMP9 may represent SASP-related biomarkers in T2DM pathogenesis. Moreover, we present an integrative workflow for systematic screening prioriti zation and validati on of SASP-related candidate genes across multiple cohorts.