Integrated Multi-Tissue Omics Identifies Acylcarnitine Accumulation as Shared Metabolic Marker of Diabetic Microangiopathy With Cross-Organ Validation

Abstract

Purpose Diabetic retinopathy (DR) is often recognized as a marker of systemic microvascular disease, but the metabolic links to other complications, such as diabetic nephropathy (DN), remain unclear. We aimed to identify systemic metabolic signatures shared by DR and DN and investigate their potential causal mechanisms. Methods Multi-tissue metabolomic profiling of the retina, plasma, and kidney was performed in streptozotocin-induced diabetic mice. Clinical relevance was supported by public human DR and DN transcriptomic datasets. Causal relationships were assessed by two-sample Mendelian randomization (MR) using eQTLGen and genome-wide association study (GWAS) summary statistics. Single-cell in silico perturbation analysis was performed to predict organ-specific functional consequences. Results Cross-organ metabolomic profiling identified a conserved systemic lipotoxic signature, yielding a predictive plasma panel comprised of free carnitine and two long-chain acylcarnitines. Clinical transcriptomics and MR analyses pinpointed the synchronous downregulation of the SLC22A5 and CPT2 axis as a causal genetic signature of this lipid imbalance. Furthermore, in silico single-cell analyses revealed that this shared metabolic disturbance induced distinct transcriptional responses across tissues, suggesting tissue-specific molecular responses that may contribute to organ-specific microvascular dysfunction. Conclusions Both DR and DN are associated with systemic disruption of acylcarnitine metabolism. A circulating carnitine/acylcarnitine signature may serve as a non-invasive indicator of microvascular risk, and the SLC22A5–CPT2 axis represents a potential therapeutic target.