Methylome and proteome integration in human skeletal muscle uncover group and individual responses to high‐intensity interval training

Abstract

Exercise is a major beneficial contributor to muscle metabolism, and health benefits acquired by exercise are a result of molecular shifts occurring across multiple molecular layers (i.e., epigenome, transcriptome, and proteome). Identifying robust, across‐molecular level targets associated with exercise response, at both group and individual levels, is paramount to develop health guidelines and targeted health interventions. Sixteen, apparently healthy, moderately trained (VO2 max = 51.0 ± 10.6 mL min−1 kg−1) males (age range = 18–45 years) from the Gene SMART (Skeletal Muscle Adaptive Responses to Training) study completed a longitudinal study composed of 12‐week high‐intensity interval training (HIIT) intervention. Vastus lateralis muscle biopsies were collected at baseline and after 4, 8, and 12 weeks of HIIT. DNA methylation (~850 CpG sites) and proteomic (~3000 proteins) analyses were conducted at all time points. Mixed models were applied to estimate group and individual changes, and methylome and proteome integration was conducted using a holistic multilevel approach with the mixOmics package. A total of 461 proteins significantly changed over time (at 4, 8, and 12 weeks), whilst methylome overall shifted with training only one differentially methylated position (DMP) was significant (adj.p‐value < .05). K‐means analysis revealed cumulative protein changes by clusters of proteins that presented similar changes over time. Individual responses to training were observed in 101 proteins. Seven proteins had large effect‐sizes >0.5, among them are two novel exercise‐related proteins, LYRM7 and EPN1. Integration analysis showed bidirectional relationships between the methylome and proteome. We showed a significant influence of HIIT on the epigenome and more so on the proteome in human muscle, and uncovered groups of proteins clustering according to similar patterns across the exercise intervention. Individual responses to exercise were observed in the proteome with novel mitochondrial and metabolic proteins consistently changed across individuals. Future work is required to elucidate the role of these proteins in response to exercise.

Metabolomic and proteomic applications to exercise biomedicine

“…While beyond the scope of this review, it is worth noting the relative merits of these approaches [ 2 ], also the advent of spatial transcriptomics [ 3 ], which offers tissue level resolution of transcript distribution (and proteins) in relation to co-localization to structures of scientific interest. More recently in exercise biomedicine, there has also been a focus on how exercise may alter gene expression through post-translational modification of DNA such as methylation and acetylation (silencing/activating) gene expression both acutely [ 4 , 5 ] affecting, and perhaps leaving sustained molecular footprints, termed muscle memory [ 6 ]. Readers are referred to several review articles [ 2 , 7 ], [ 8 ], [ 9 ], [ 10 ] if interested in methodologies to confer study at these biological levels.…”

Molecular Landscape of Modality-Specific Exercise Adaptation in Human Skeletal Muscle through Large-Scale Multi-OMICs Integration

“…Human intervention studies are time consuming and invasive and therefore typically have small group sizes, these individual studies often lack sufficient statistical power due to the large number of CpGs investigated. For example, to achieve 95% power in typical exercise and DNA methylation studies Jacques et al, 2023;Voisin et al, 2024), a study would need approximately 316 tissue samples given effect sizes of 0.01-0.02 beta-value and standard errors average 0.06-0.07 betavalue. Therefore, a meta-analysis approach pooling multiple independent studies is ideal and costeffective for drawing powerful and robust associations between DNA methylation and exercise adaptation.…”