Cardiovascular disease (CVD) occurs 15 years earlier in diabetics compared to those without, but the underlying mechanisms driving diabetes-accelerated CVDs (DACVDs) pathogenesis remains incompletely understood. Endothelial cells (ECs) and macrophages (MΦ) are key players in vascular wall and their crosstalk is crucial in DACVD. In diabetes, ECs activation enables monocytes recruitment, which transmigrate across the intima and differentiate into macrophages (MΦ). Beyond this established diapedesis model, EC-MΦ interplay is highly intricate and heterogenous. Here, we leveraged human mesenteric arteries of varied diabetic state and used single cell (sc)- and spatial transcriptome mapping to interrogate the functional crosstalk between MΦ subtypes and ECs in diabetes. We identified diabetes-induced EC-MΦ interactions and explored their consequences in vitro and in vivo, in peripheral artery disease (PAD), a prominent DACVD. Integrative scRNA-seq and spatial transcriptome (Visium) profiling identified a remarkable increase of MΦ subtype expressing triggering receptor expressed on myeloid cells (TREM2) with concomitant increased EC-derived TREM2 ligands in human diabetic vs non-diabetic vessels. Such induction of TREM2 MΦ was consistently observed in arteries from type I and II diabetes mouse models. Compared to the condition medium from MΦ cultured under normal glucose, that from MΦ cultured under high glucose conditions reduced EC migration and increased inflammatory markers, which were reversed by the medium from MΦ with TREM2-knockdown. We further evaluated the functional importance of EC-TREM2 interaction in vivo in a mouse hindlimb ischemia (HLI) model. In diabetic mice under HLI, a TREM2 neutralizing antibody promoted tissue recovery and perfusion. Finally, we showed that TREM2 was increased, especially in proximity to ECs, in ischemic sites, as compared to non-ischemic regions in limb muscles from human PAD patients.Taken together, our study presents a diabetic artery atlas and identifies EC-TREM2 interaction as a crucial event in DACVDs, such as PAD. Our findings provide novel insights into the EC-MΦ interplay in diabetes and vascular inflammation and how these cellular interactions may be targeted to ameliorate DACVD.