Identifying a fatty acid metabolic gene signature in diabetic cardiomyopathy through integrated bioinformatics and machine learning

Comprehensive Summary

This study investigated diabetic cardiomyopathy (DCM), a major complication of diabetes affecting 463 million people worldwide (projected 700 million by 2045), by integrating multi-omics data to identify key metabolic drivers. Researchers analyzed 7 GEO gene expression datasets and identified 112 differentially expressed genes (84 upregulated, 28 downregulated) using three approaches: limma, Robust Rank Aggregation (RRA), and Weighted Gene Co-expression Network Analysis (WGCNA). Network and feature selection methods, including machine learning techniques (Random Forest and SVM-RFE), narrowed to 7 hub genes (Hmgcs2, Ech1, Acot2, Acot1, Fbp2, Fkbp5, Decr1), all with strong diagnostic performance (AUC > 0.96). Single-nucleus RNA sequencing showed widespread upregulation of several genes, especially in cardiomyocytes. Experimental validation in a streptozotocin (STZ) induced mouse model confirmed increased blood glucose (>11.1 mmol/L), reduced body weight, and impaired cardiac function, including decreased left ventricular ejection fraction (LVEF) and fractional shortening (LVFS). Gene expression analysis showed significant upregulation of 5 key genes at the mRNA level, while protein levels of ACOT1, FBP2, and ECH1 were also elevated. Proteomics identified 572 differentially expressed proteins, enriched in fatty acid β-oxidation and PPAR signaling pathways. Metabolomics revealed that 67.19% of altered metabolites were lipids, with increased fatty acids and decreased NAD⁺ and cADPR, indicating disrupted energy metabolism. Overall, the study demonstrates that DCM involves a metabolic shift from glucose to fatty acid utilization, leading to lipid accumulation, mitochondrial dysfunction, and reduced energy efficiency, with Ech1 and Fbp2 emerging as key therapeutic targets.

Outcomes and Implications

This study highlights important clinical implications by identifying key metabolic genes (such as Ech1 and Fbp2) that contribute to the shift from glucose to fatty acid metabolism in diabetic cardiomyopathy. The findings describe a self-reinforcing cycle of lipid accumulation, NAD⁺ depletion, and mitochondrial stress that progressively impairs heart function. The strong diagnostic performance of these genes (AUC > 0.96) suggests they could be useful as early biomarkers for DCM. In addition, targeting pathways involved in fatty acid metabolism and energy imbalance may offer new therapeutic strategies to improve cardiac function and slow disease progression in patients with diabetes.