Abstract Funding Acknowledgements Type of funding sources: None. Background Patients with type 2 diabetes mellitus (DM) that have coexistent heart failure (HF) have exacerbated symptoms and prognosis, however beside cardiac dysfunction the mechanisms governing these features are incompletely understood. Evidence indicates abnormalities in the periphery could contribute to this worse clinical phenotype, including a role for skeletal muscle whereby disturbances in the transcriptome could disrupt muscle homeostasis/repair to offer a novel therapeutic approach. Purpose Is the skeletal muscle transcriptome distinguishable between DM patients with and without HF? Methods DM patients without (n = 11) or with HF with reduced left ventricular ejection fraction (LVEF) (n = 16) were included. Muscle biopsies were collected from the pectoralis major during pacemaker implantation. Following RNA extraction and cDNA synthesis, non-bias RNA sequencing (RNAseq) was performed (Cambridge Genomic Services, UK) followed by targeted RT-PCR gene expression of relevant targets. DESeq2 identified differentially expressed genes (DEGs) with a false discovery rate (p < 0.05). Gene enrichment analysis was performed with clusterProfiler v3.16.0 to interrogate the gene ontology database, while pathway analysis was conducted using ReactomePA v1.32.0 to interrogate the Reactome database, using an adjusted p value. Values of p < 0.05 were accepted as significant. Results Groups were not different (p > 0.05) for age (74 ± 11 vs. 66 ± 10 years), BMI (31 ± 7 vs 29 ± 6), sex (n = 2 females per group), or HbA1c (56 ± 10 vs. 57 ± 8 mmol/mol), although LVEF was lower in the group with HF (27 ± 8 vs. 54 ± 2%; p < 0.05). Of the 19,544 genes analysed, RNAseq identified 53 DEGs between DM patients with and without HF, with several relevant targets related to myofiber homeostasis such as autophagy (RUBCN), protein synthesis (DGKζ), and inflammation/apoptosis (TLE1). Follow-up RT-PCR analysis confirmed a trend towards upregulation of the autophagy-related machinery p62 (p = 0.043) and BNIP3 (p = 0.085) in the HF group, but not ubiquitin-proteasome (MuRF1, MAFbx; p > 0.05). Gene-enrichment analysis of DEGs identified 7 overrepresented terms (P < 0.05), including lipid metabolism/signalling alongside epigenetic modifications related to histone deacetylases (HDAC6/10). Furthermore, pathway analysis identified 4 terms (p < 0.05) related to NOTCH signalling and phosphatidyl inositol-bisphosphate (PIP2) hydrolysis thus indicating alterations to muscle repair and lipid signalling respectively. Conclusion(s): This study confirms that DM patients with and without HF demonstrate distinct skeletal muscle transcriptome profiles. Key differences related to skeletal muscle myogenesis, autophagy, epigenetic regulation, and lipid signalling were identified that could form part of important therapeutic targets. Whether these underlying muscle transcriptome differences contribute to poorer clinical outcomes in DM patients with HF remains to be determined.
Read full abstract