Differences in Epigenetic, Transcriptional and Protein Adaptations to Exercise in Skeletal Muscle of Responders and Low-Responders in Insulin Sensitivity

Abstract

Die Dissertation ist gesperrt bis zum 28. Februar 2025 !

Type 2 diabetes mellitus (T2DM) is one of the most challenging diseases of our current century due to its high prevalence and economic burden. It is, therefore, of high interest to reduce the prevalence by improving prevention therapies. Regular physical activity has a high potential to prevent or delay the onset of T2DM. However, it is known that people respond very differently to exercise, even when the relative exercise intensity is comparable. Participants of exercise interventions who improve their insulin sensitivity or other metabolic parameters are called responders (RES), and participants not improving or worsening are called low-responders (LRE). Although multiple tissue and molecular adaptations to exercise have been elucidated, which can lead to the improvement of insulin sensitivity and thus prevention of T2DM, the driving forces for the distinction between RES and LRE are unclear. We, therefore, performed proteomics, transcriptomics, and epigenomics analyses of human skeletal muscle biopsies of RES and LRE in insulin sensitivity (assessed using a standard oral glucose tolerance test) and analyzed serum cytokine abundance after acute exercise and prolonged endurance training. The study was conducted in subjects with a high risk for T2DM due to overweight or obesity who performed a supervised 8-week training intervention consisting of 30 min bike and 30 min treadmill training at 70% VO2peak. Muscle biopsies and blood samples for the multi-omics analyses were collected before and after the intervention in the resting state and after acute exercise. In the first approach, the transcriptome and epigenome datasets were analyzed separately and in combination to select transcripts with robust differential regulation in RES and LRE on transcript and DNA methylation levels. Among them are IL34 and AKR1C3 and its two family members, AKR1C1 and AKR1C2, which showed a negative correlation to the change in insulin sensitivity. The different regulation of AKR1C3 was also found on protein level. The candidates were further analyzed in cell culture experiments by treating human myotubes with soluble IL34 or inducible overexpression of the AKR1C-proteins in the mouse muscle cell line C2C12. The cells were analyzed for insulin-induced AKT phosphorylation and mitochondrial respiration. Neither IL34 nor AKR1C3 altered the analyzed parameters in insulin signaling or respiration. Additional experiments revealed that while human myotubes expressed and released IL34, they did not respond to IL34 by activation of receptor-mediated signaling, which suggests that myofibers are not the target of IL34 within skeletal muscle tissue. AKR1C3 was previously described as an insulin-regulated transcript in other tissues, which might explain the upregulation in skeletal muscle of LRE but not RES. The proteome analysis revealed that over all subjects several mitochondrial proteins were increased after training, many of those correlated with an increased mitochondrial respiration measured in isolated muscle fibers. The increased abundance of PERM1 after training was paralleled by transcriptional upregulation of PERM1 and altered methylation pattern on 4 methylation sites (CpG sites) over all subjects. The increase in PERM1 correlated not only with increased respiration but also with the increase in insulin sensitivity. The increase in PPME1 showed a positive correlation with the change in insulin sensitivity after the training, but also the abundance of PPME1 in the resting biopsies was positively correlated with the pre- and post-intervention insulin sensitivity. The potential function of PERM1 and PPME1 as regulators of insulin sensitivity in skeletal muscle needs to be validated. The fold change of other proteins increased after training did not correlate with the change in insulin sensitivity. As a last approach, changes in serum cytokine concentrations were studied using a proximity extension assay (PEA) and a panel of 92 inflammatory cytokines. This analysis revealed multiple cytokines highly increased in serum after an acute exercise bout that have not been studied in the context of exercise yet. We correlated the acute changes of the cytokines with the clinical parameters and found significant correlations with pre-intervention parameters of obesity that diminished after the training intervention. In contrast, the training intervention did not change resting cytokine levels. In summary, this thesis shows the potential of multi-omics analysis in studying the differential training response of individuals and gives insight into potential molecular mechanisms influencing the impact of regular exercise on insulin sensitivity. Furthermore, the obtained datasets provide a comprehensive pattern of exercise-induced molecular adaptations in skeletal muscle and can be utilized for selecting additional candidates, potentially explaining the different responses of RES and LRE to finally improve diabetes prevention by customizing the training sessions more individually.