Atherosclerosis and Diabetic Foot Share a Common Molecular Signature of Cellular Senescence: A Multi-Omics and Machine Learning Study

Abstract Atherosclerosis (AS) and diabetic foot ulcer (DFU) are two major vascular complications of diabetes that frequently co-exist, yet whether they share a common molecular basis remains unclear. Here, we tested the hypothesis that AS and DFU share a common senescence-related molecular signature. We integrated transcriptomic datasets GSE134431 (DFU) and GSE163154 (AS), identified differentially expressed genes common to both diseases, and intersected them with the CellAge senescence-associated gene set to obtain senescence-related DEGs (SR-DEGs). Three machine learning algorithms—LASSO, random forest, and SVM-RFE—were applied to prioritize core feature genes, which were subsequently validated at the single-cell (GSE245703), tissue, and in vitro levels. We identified 72 SR-DEGs shared by AS and DFU, from which three core feature genes—CALCOCO1, CCND2, and FUCA1—were selected by cross-algorithm consensus. A multi-gene nomogram incorporating these genes achieved excellent discriminative performance in both diseases (AUC: 0.96 for DFU, 0.90 for AS). Single-cell analysis revealed cell type–restricted expression patterns of these genes, with fibroblasts emerging as a key senescent cell type in the DFU microenvironment. Clinical validation in an independent cohort of 22 patients (11 DFU vs. 11 non-diabetic foot ulcer) confirmed significant accumulation of senescent cells and dysregulated feature gene expression in DFU tissues. Moreover, metformin treatment effectively attenuated hydrogen peroxide- and high-glucose/high-lipid-induced fibroblast senescence in vitro, partially reversing the expression changes of the three feature genes. These findings demonstrate that AS and DFU share a common senescence-associated molecular signature, identify CALCOCO1, CCND2, and FUCA1 as potential diagnostic and therapeutic targets, and suggest that metformin may exert protective effects through modulation of this senescence-related gene network.