Identification of immune feature genes and intercellular profiles in diabetic cardiomyopathy

Background: Diabetic cardiomyopathy (DCM) is a multifaceted cardiovascular disorder in which immune dysregulation plays a pivotal role. The immunological molecular mechanisms underlying DCM are poorly understood.

Aim: To examine the immunological molecular mechanisms of DCM and construct diagnostic and prognostic models of DCM based on immune feature genes (IFGs).

Methods: Weighted gene co-expression network analysis along with machine learning methods were employed to pinpoint IFGs within bulk RNA sequencing (RNA-seq) datasets. Single-sample gene set enrichment analysis (ssGSEA) facilitated the analysis of immune cell infiltration. Diagnostic and prognostic models for these IFGs were developed and assessed in a validation cohort. Gene expression in the DCM cell model was confirmed through real time-quantitative polymerase chain reaction and western blotting techniques. Additionally, single-cell RNA-seq data provided deeper insights into cellular profiles and interactions.

Results: The overlap between 69 differentially expressed genes in the DCM-associated module and 2483 immune genes yielded 7 differentially expressed immune-related genes. Four IFGs showed good diagnostic and prognostic values in the validation cohort: Proenkephalin (Penk) and retinol binding protein 7 (Rbp7), which were highly expressed, and glucagon receptor and inhibin subunit alpha, which were expressed at low levels in DCM patients (all area under the curves > 0.9). SsGSEA revealed that IFG-related immune cell infiltration primarily involved type 2 T helper cells. High expression of Penk (P < 0.0001) and Rbp7 (P = 0.001) was detected in cardiomyocytes and interstitial cells and further confirmed in a DCM cell model in vitro. Intercellular events and communication analysis revealed abnormal cellular phenotype transformation and signaling communication in DCM, especially between mesenchymal cells and macrophages.

Conclusion: The present study identified Penk and Rbp7 as potential DCM biomarkers, and aberrant mesenchymal-immune cell phenotype communication may be an important aspect of DCM pathogenesis.