Diabetes. 2026 Jan 21. pii: db250625. [Epub ahead of print]
German Diabetes Study Group*
Recent advances in RNA sequencing (RNA-seq) techniques allow the identification of tissue-specific alternative splicing and can thereby provide new insights into molecular mechanisms of energy metabolism. Full-length transcriptomics based on single-molecule real-time sequencing (SMRT-seq) enable precise detection of isoforms with 99% accuracy in an unbiased manner. In this proof-of-concept study, we integrated SMRT-seq, bulk RNA-seq, and comprehensive metabolic phenotyping to investigate reduced mitochondrial function in the skeletal muscle of individuals with type 2 diabetes. Muscle biopsies were taken from nine individuals with type 2 diabetes and nine age- and BMI-matched glucose-tolerant men. Whole-body insulin sensitivity (WBIS) was assessed by hyperinsulinemic-euglycemic clamps, and muscle mitochondrial respiration was assessed by high-resolution respirometry. In muscle samples, SMRT-seq was used to create full-length reads and isoforms, which were mapped to the genome. Short-read sequencing was used to compare isoform expression between the groups. Participants with diabetes exhibited lower WBIS and fatty acid-driven and complex I-linked respiration compared with control participants. SMRT-seq revealed ∼67,000 isoforms originating from ∼14,000 unique genes. Although isoform numbers per gene did not differ, SMRT-seq-based mapping enabled refined data set clustering compared with conventional short-read sequencing and identified four splicing variants of the ATP5F1A gene encoding a subunit for ATP synthase. Among these, two novel transcripts were expressed exclusively in control participants. This study identified splicing variants of ATP synthase that were differentially expressed between participants with type 2 diabetes and those with normal glucose tolerance, which may contribute to the reduced fatty acid oxidation in diabetes.
ARTICLE HIGHLIGHTS: In our study, we developed a pipeline to integrate single-molecule real-time sequencing (SMRT-seq) with comprehensive metabolic phenotyping to examine reduced mitochondrial respiration in the skeletal muscle of individuals with type 2 diabetes. SMRT-seq revealed ∼67,000 isoforms originating from ∼14,000 unique genes; the isoform numbers per gene did not differ between participants with diabetes and matched control participants. Our data identified novel alternative splicing events, including two variants of the ATP5F1A gene encoding a subunit for ATP synthase. Among these, two novel transcripts were expressed exclusively in control participants. Our findings link transcriptomic changes to impaired mitochondrial respiration in type 2 diabetes, with the potential of providing novel therapeutic targets to improve metabolic health.