Author(s): HANSEN, M.1, GROTHEN, J.E.R.2,3, KARLSEN, A.1, PEDERSEN, T.Å.2, SIDIROPOULOS, N.4, HELGE, J.W.1, DELA, F.1,5, Institution: UNIVERSITY OF COPENHAGEN, Country: DENMARK, Abstract-ID: 2434

INTRODUCTION:
Training improves insulin sensitivity in patients with type 2 diabetes (T2D) (1), but the underlying mechanisms are only partly understood. Single-nuclei RNA-sequencing (snRNA-seq) offers the opportunity to study fibre type-specific gene expression changes. This is the first study to examine nuclei-specific gene expression changes in skeletal muscle in individuals with T2D and healthy controls (CON) in response to a short-term high-intensity training (HIIT) program. We hypothesised that snRNA-seq of muscle biopsies would reveal marked diabetes- and training-induced responses.
METHODS:
Ten males with T2D (57±2 years, BMI 31±1 kg/m2, HbA1c 53±1 mmol/mol) and ten male CON (53±2 years, BMI 31±1 kg/m2, HbA1c 37±1 mmol/mol) completed two weeks of one-legged HIIT on a cycle ergometer. Insulin sensitivity (clamp + leg balance technique) was measured previously (2). We prepared a single nuclei suspension from muscle biopsies from untrained (UT) and trained (T) legs and used the 10X Genomics system for the preparation of snRNA-seq Libraries, which were sequenced using an Illumina NextSeq550snRNA-seq. QC, cluster identification and differential expression testing were analysed using the Seurat (v4), and we identified differentially expressed genes (DEGs) between two groups of nuclei using a Wilcoxon Rank Sum test. Fibre type distribution was confirmed through immunohistochemistry.
RESULTS:
Insulin-stimulated leg glucose clearance was 33±24% higher in T legs compared to UT legs in both groups (2).
Single-nuclei RNA-sequencing in 38 biopsies yielded 135,225 nuclei in total, and we profiled 99,970 myonuclei. In the UT legs, there were 249 DEGs between the groups, with equal distribution between fibre types. When comparing UT and T legs, there were no differences between the number of DEGs in the groups (222 DEGs in CON and 234 DEGs in T2D).
Almost no genes related to the citric acid cycle, glycolysis, glycogenolysis, or beta-oxidation differed between the groups in the UT leg. In T2D there was no effect of training on genes related to glycolysis and glycogenolysis. However, CON had multiple DEGs for glycolytic and glycogenolytic enzymes (i.e. enolase, glycogen debranching enzyme, glycogen phosphorylase) in the type 2A and type 2X fibres in the T leg compared to the UT leg. There were no changes in fibre type distribution after training, and the immunostaining and the snRNA-seq fibre type distributions were in accordance with each other.
CONCLUSION:
HIIT improves skeletal muscle insulin sensitivity, but the modest single-nuclei gene expression changes did not mirror this response. Mostly, genes related to glycolysis and glycogenolysis in type 2 muscle fibres displayed significant differences between the groups in response to training. Overall, the pattern of altered gene expressions does not seem to explain the marked change in muscle insulin sensitivity.

References
1. Kanaley et al., Med Sci Sports Exerc, 2022
2. Dela F et al., Acta Physiol, 2019