henhaobio
Sign in

bioRxivSystems BiologyPreprint

Integrative Multi-omics Analysis of the Human Skeletal Muscle Response to Endurance or Resistance Exercise: Findings from the Molecular Transducers of Physical Activity Consortium (MoTrPAC)

Hasmik KeshishianiD, Gina M. ManyiD, Gregory R. SmithiD, Natalie M. ClarkiD, Gayatri R. IyeriD, Patrick Hart, Maléne E. LindholmiD, Samuel MontalvoiD, Zidong ZhangiD, Christopher A. JiniD, James A. SanfordiD, Steven A. CarriD, Joshua AdkinsiD, D.R. ManiiD, Sue C. BodineiD, Scott W. TrappeiD, Joseph A. HoumardiD, Nicolas MusiiD, Kim M. HuffmaniD, William E. KrausiD, Lauren Marie SparksiD, Anna Elizabeth Thalacker‐MerceriD, Stuart C. SealfoniD, Ashley Y. XiaiD, Daniel H. KatziD, , MoTrPAC Study Group

PDFDOICode
In-depth full-text digest available · Jump to digest ↓

Abstract

The Molecular Transducers of Physical Activity Consortium (MoTrPAC) was established to systematically characterize the molecular basis of the health benefits of exercise. Here, we present the integrative, multi-omics response of human skeletal muscle to acute endurance (EE) and resistance (RE) exercise. Distinct temporal responses were observed, with changes in ATAC-seq, phosphoproteome, and metabolome occurring before changes in the transcriptome and proteome. These distinct temporal multi-omic dynamics were used to identify transcriptional regulatory hubs converging around MEF2A and NFIC regulation of autophagy, angiogenesis and metabolism. Further, early RE-specific phosphoproteome signatures counteracted epigenetic modifications and downregulated transcripts involved in protein turnover. Additional findings include suppression of HIPK2/3 kinase signatures linked to the acute exercise regulation of sarcomeric proteins TTN, NEB, ANKRD2 and LMOD2. Our data demonstrate distinct temporal regulation across the multi-omic landscape of human skeletal muscle, with EE and RE eliciting common and unique molecular signatures.

In-depth summary

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 2 (8 Oct 2026) · CC BY 4.0

In 174 sedentary adults, resistance exercise changed more muscle molecular features than endurance exercise across omes and time points; network analysis nominates MEF2A and NFIC as regulatory hubs.

Earlier acute-exercise studies, as the authors describe them, mostly profiled one ome at a time, enrolled few participants or pooled heterogeneous studies, leaving the coordinated time course across omes unclear. Here, sedentary adults were randomized to endurance, resistance or non-exercise control, and muscle biopsies taken up to 24 h later were profiled for chromatin accessibility, DNA methylation, transcripts, proteins, phosphosites and metabolites against time-matched controls.

Key results

  1. ShownResistance exercise (RE) altered more muscle features than endurance exercise (EE) at every post-exercise time point. The transcriptome and phosphoproteome carried the most changes and the proteome the fewest.

    Difference-in-change models versus time-matched controls, adjusted p < 0.05. Ranking of differentially abundant (DA) features: transcriptome and phosphoproteome, then epigenome, metabolome, proteome. DA proteins: n = 4 in EE vs n = 31 in RE. Features that were DA in both modalities mostly changed in the same direction. · RE elicits a broader and more sustained…; Discussion; Figure legends

  2. SuggestedNetworks linking transcription factor phosphosites and chromatin accessibility to target transcripts nominated MEF2A and NFIC as hubs. NFIC, which the authors say had no known role in exercise, is predicted to regulate mitochondrial and translation genes.

    SC-ION + MAGICAL: ChIP-seq predicted NFIC targets were filtered in the combined network (1580/2326, 67.9%), including MEF2A, CHCHD3 and ESRRA, and enriched for oxidative phosphorylation, splicing and mitochondrial translation. NR4A1, an NFIC target, was strongly induced by both modalities. In International Mouse Phenotyping Consortium data, NFIC knockout mice of both sexes had lower lean-to-total body mass. · Central regulatory nodes of the acute exercise…; Figure legends

  3. SuggestedAt the FBXO32/KLHL38 locus, chromatin opened after both modalities, yet expression rose with endurance and fell with resistance. The authors propose that RE-specific MAPK- and AMPK-linked FOXO3 phosphorylation, plus ZEB1 changes, override the accessibility signal.

    Locus accessibility and gene expression were reported as highly correlated. FOXO3 was phosphorylated at S425, S413 and S294 after RE but only at S294 after EE; ZEB1 pS679 rose with RE while pS315;T324 fell with EE. Only RE showed an early MAPK/ERK signal and protein-synthesis phosphosites (RPS6 S236, EEF2K S470;S474). · MAPK-driven phosphorylation overrides epigenetic…; Figure legends

  4. SuggestedA HIPK3 kinase signature dropped quickly after both modalities. Phosphosites that tracked HIPK3 autophosphorylation sat on sarcomeric proteins, which the authors read as a link between HIPK3 and sarcomere remodeling, stronger after resistance exercise.

    Phosphosite signature enrichment (PTM-SEA): 52 (EE) and 77 (RE) signatures at adj. p < 0.05, mostly at 15 min. HIPK3 pY359 fell after RE; the leading edge was BAG3 pS173/pT285, dephosphorylated in both. Sites correlated with pY359 (Pearson r ≥ 0.8) were enriched for sarcomere organization (TTN, ACTN2, MYPN, CSRP3, LMOD2); LMOD2 dephosphorylation at 15 min preceded its RE-specific protein increase. · Exercise-induced HIPK3 dephosphorylation is…; Figure legends

  5. ShownEarly metabolite responses differed by modality. Endurance raised medium-to-long-chain acylcarnitines and lowered ceramides; resistance lowered acylcarnitines, depleted ATP and CTP more, and reduced leucine and isoleucine.

    Acylcarnitines with >6 carbons rose at 15 min after EE and fell after RE. RE also raised 3-hydroxybutyric acid at 15 min, plus hypoxanthine, inosine and xanthosine; EE raised 2-hydroxy-3-methylbutyric acid at early and middle time points. · RE elicits a broader and more sustained…; Modality-specific responses to acute exercise…

Methods, data and reproducibility

Vastus lateralis biopsies taken before and 15 min, 3.5 h and/or 24 h after cycling, a resistance circuit or supine rest were profiled by ATAC-seq, methylation capture, RNA-seq, proteomics, phosphoproteomics and metabolomics. Difference-in-change models tested exercise against time-matched controls; integration used non-negative matrix factorization, PLIER, PTM-SEA signatures and SC-ION/MAGICAL networks.

  • Vastus lateralis biopsies, MoTrPAC pre-COVID-19 cohort of healthy sedentary adults randomized to control, endurance or resistance exercise; pre-exercise plus 15 min, 3.5 h and/or 24 h post-exercise · human · n = 174 participants (CON 37, EE 64, RE 73) · NCT03960827 (trial registration)
  • TMT 16-plex LC-MS/MS proteomics and IMAC phosphoproteomics of muscle, split between the Broad Institute and PNNL · human · n = 379 muscle samples; 14 analyzed at both sites
  • Pre-exercise RNA, proteome and phosphoproteome data used for baseline non-negative matrix factorization clustering · human · n = 159 samples
  • Muscle ATAC-seq, EPIC methylation capture sequencing, poly(A) RNA-seq, and targeted plus untargeted metabolomics and lipidomics (per-assay counts in Extended Data Fig. 1a) · human
  • NFIC knockout versus wild-type phenotypes from the International Mouse Phenotyping Consortium repository · mouse (C57/Bl6 background)
  • Published transcriptomic exercise results of Reitzner et al., compared by log2 fold-change correlation
Methods & tools
ENCODE ATAC-seq pipeline, STAR, RSEM, Bismark, Malax, variancePartition (dream), MSGF+, CAMERA-PR, PLIER, PTM-SEA, SC-ION, MAGICAL
Software & environment
Versions stated: ENCODE ATAC-seq pipeline v1.7.0, cutadapt v2.5/v1.18, Bowtie 2 v2.3.4.3, MACS2 v2.2.4, bedtools v2.29.0, Bismark v0.20.0, STAR v2.7.0d, RSEM v1.3.1, hg38 with GENCODE v29; RNA-seq pipeline in openWDL on Google Cloud Platform. No container or requirements file named.
Assay conditions & compute
Methylation testing with Malax was split into 100 subsets run in parallel on cluster computers; no hardware, runtime or memory figures are given.
Reproducibility gaps
The text gives no data accession, portal URL or data-availability statement; sharing is described only as following NIH Genomic Data Sharing policy and participant consent levels. Proteomics code is referenced without a link; details are deferred to companion papers.
Limitations (6)
  • Temporal and omic downsampling: each participant gave a limited number of biopsies, and not every sample was run on every assay because of assay limits and study design. · stated by the authors
  • Sample size was too small to rigorously test effects of sex, age or ethnicity; a baseline cluster enriched for Black women (Fisher p < 0.01) could not be explored or adjusted for. · stated by the authors
  • The authors caution that the larger RE response may reflect protocol-specific demands, such as heavier mechanical loading, intermittent high-force contractions and probably more local metabolic stress in sedentary participants. · stated by the authors
  • Limited sample sizes meant methylation models could include only sex and age as covariates, and metabolomics was left out of baseline clustering because 20 participants lacked full platform coverage. · stated by the authors
  • The MEF2A and NFIC hubs, the HIPK3-BAG3 link and the FOXO3/ZEB1 model rest on correlation and network inference without perturbation in human muscle; the NFIC support comes from public knockout mouse phenotypes that do not involve exercise. · noted in this summary
  • Proteomic cell-type terms (mesenchymal cells, monocytes, erythrocytes) come from bulk biopsies, so they may reflect shifts in tissue composition that the design cannot separate from signaling within fibers. · noted in this summary

Read this if you study muscle signaling or exercise biology and want a time-resolved human multi-omic reference contrasting endurance and resistance bouts against controls, or an example of tying phosphosites and chromatin to transcripts in networks.

Citations
2

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISGroup '26Ahn '26Group '26Reitzner '23Robbins '26Group '22
74% semantic matchSystems Biology
MoTrPAC Study Group, Daniel H. Katz, Christopher A. Jin+43 more·bioRxiv·Sep 30, 2026