AKR1C3 and CLIC3 mediate the comorbidity of diabetic foot ulcers and hypercholesterolemia via the TGF-β/lipid metabolism axis.

Hypercholesterolemia (HCh) represents a prevalent comorbidity in patients with diabetic foot ulcers (DFUs), frequently exacerbating wound recalcitrance, yet the mechanisms linking systemic lipid perturbation to local immune dysregulation remain poorly defined. Here, we applied an integrated transcriptomic and machine learning approach to explore the molecular basis of this comorbidity. Initial analyses identified 70 shared DEGs between DFUs and HCh, primarily enriched in pathways related to lipid homeostasis disruption and inflammatory cascade activation. Subsequently, a four-algorithm machine learning pipeline highlighted AKR1C3 and CLIC3 as candidate diagnostic biomarkers, both consistently downregulated across disease states and supported by independent external validation cohorts (AUC = 0.779 and 0.920). Immune deconvolution further indicated a dysregulated microenvironment featuring sustained innate immune activation alongside adaptive immune impairment, which correlated with reduced AKR1C3 and CLIC3 expression. Drug repurposing analyses proposed that pirfenidone and statins might attenuate this pathological transcriptomic signature, while molecular docking simulations suggested stable binding of exemestane and indomethacin to AKR1C3 (binding energies: -8.0 and -7.3 kcal/mol, respectively). Collectively, these findings suggest that AKR1C3 and CLIC3 act as molecular bridges linking the TGF-β/lipid metabolism axis to local immune dysfunction, offering a rationale for early risk stratification and targeted drug repurposing in DFUs-HCh comorbidity.