Integrative multi-omics analysis identifies an extracellular vesicle-associated miRNA-RAS-MST1 regulatory network in MASH-associated sarcopenia
Yu Xu, Fan Gong, Tingting Cao, Ruogu Chen, Nuotong Zhuo, Yanhong Wei, Yi YangBackground
Sarcopenia often occurs alongside metabolic dysfunction-associated steatohepatitis (MASH). The molecular basis of liver-muscle metabolic crosstalk is unclear. Extracellular vesicle (EV)-associated factors and renin-angiotensin system (RAS) signaling modulate systemic metabolic and inflammatory remodeling, yet their involvement in MASH-linked sarcopenia remains uncharacterized.
Methods
We integrated EV-associated microRNA (miRNA) profiling, bulk and single-cell transcriptomics, weighted gene co-expression network analysis, Mendelian randomization, immune deconvolution, interpretable machine learning, and cellular functional assays to investigate conserved molecular signatures associated with MASH and sarcopenia across liver and skeletal muscle datasets.
Results
Cross-compartment analysis identified conserved EV-associated miRNA signatures across datasets from liver EVs, circulating EVs and skeletal muscle. Network analysis revealed a RAS regulatory module containing ACE, ACE2 and Hippo kinase MST1. Genetic data associated metabolic liver disease and sarcopenia with RAS-MST1 signatures, while direct causality remained unconfirmed. Single-cell deconvolution showed cell-type-specific enrichment of RAS/MST1 pathways mainly in macrophages, Kupffer cells and regenerative populations. Cellular assays showed that MST1 modulates muscle remodeling-related phenotypes and macrophage polarization in vitro. Interpretable machine learning validated MST1-centric signatures in independent MASH and sarcopenia transcriptomic cohorts.
Conclusions
This study identifies a candidate EV-associated miRNA-RAS-MST1 network potentially involved in MASH-associated sarcopenia, supports a liver-muscle metabolic crosstalk model, and highlights MST1 as a candidate molecular target for further research.