Peripheral arterial disease (PAD) is the most common complication of type 2 diabetes mellitus (T2DM), occurring in approximately 16.2% of patients [1]. A clinical study suggests that approximately 72.4% of hospitalized patients with diabetic foot disease (DF) have complications of PAD [2]. Atherosclerotic vascular disease (ASVD) is the main pathophysiological change underlying PAD. However, in clinical practice, the characteristics of PAD in patients with diabetes and in nondiabetic patients are not exactly the same; the former presents as follows: a) an incidence of 2–4 times that of PAD in nondiabetic patients; b) extensive calcification of distal subpatellar vessels [3,4]. In recent years, more studies have found that vascular endothelial calcification is the main pathological feature of diabetic PAD compared with PAD occurring in nondiabetic patients [5,6]. Since the clinical indicators that are currently in common use do not predict the occurrence of diabetic ASVD in advance, there is an urgent need for new techniques to explore novel markers. Metabolomics is a valid tool used to discover potential biomarkers in chronic diseases and has wide application value in precision medicine [7]. Therefore, specific and sensitive blood biomarkers are needed for PAD diagnosis and risk assessment.

Serum extracellular heat shock protein 90α (eHsp90α) level is an independent risk factor for the progression of diabetic vascular disease [8]. Shen et al. [9] report that inhibition of Hsp90 reduce the anti-calcification effect of aspirin on vascular smooth muscle cells (VSMCs). eHsp90α can be found at multiple extracellular locations and anchors to the plasma membrane by binding to the ectodomains of LRP1 receptors on the cell surface [10]. Moreover, eHsp90α plays a role in promoting survival, motility and invasion in vitro and in vivo through eHsp90α-LRP1 signaling pathway [11]. However, LRP1 has dual roles in the development of atherosclerosis. LRP1 not only participates in the removal of lipoproteins but also mediates the uptake of LDL to promote the formation of foam cells [12]. Runx2 is an essential regulator of osteogenic differentiation and calcification in vascular cells. However, whether eHsp90α-LRP1 signaling can promote endothelial cell calcification, i.e., regulate Runx2 expression, has not been explored.

The physiological concentration of thyroid hormone may have a direct protective effect on vascular smooth muscle calcification in vivo [13]. Free thyroxine (FT4) levels are inversely associated with coronary artery calcification in euthyroid healthy subjects [14]. Further studies are needed to validate whether subjects with decreased FT4 levels within the normal reference range are at a higher cardiovascular disease risk. Based on previous research, we hypothesized that eHsp90α and FT4 play different roles in vascular endothelial calcification, but the specific relationship between them remains unclear.

In our study, we aimed to compare the differential serum metabolite profiles of diabetic ApoE−/− mice and control ApoE−/− mice to help define a diabetic ASVD metabolic signature for comparison with that of patients in the corresponding groups. In vivo serological studies and in vitro cell models were used to explore the direct relationship between metabolites and clinical parameters in ASVD patients and the role of eHsp90α and differentially abundant metabolites in diabetic vascular endothelial calcification injury. The findings will provide a new perspective for investigating early therapeutic targets for diabetic vascular disease.