Distinct Myogenic Stages Recapitulate Transcriptomic Networks in COPD Cachexia
BackgroundCachexia is an extrapulmonary manifestation of chronic obstructive pulmonary disease (COPD) characterized by weight loss and muscle wasting. Transcriptomic analysis of skeletal muscle may provide insight to COPD-cachexia-relevant dysregulation, including at the mitochondrial level, where dysfunction is a hallmark of muscle wasting. As muscle biopsies are invasive and yield finite tissue, human muscle derived cultures (HMDCs) can expand limited biopsy material for studying skeletal muscle dysregulation. However, utility of such models depends on their recapitulation of bulk skeletal muscle transcriptomic signatures. To address this, we tested whether COPD and COPD-cachexia dysregulated transcriptional signatures in bulk skeletal muscle are preserved across stages of early differentiation. MethodsVastus lateralis biopsies were collected from 15 (7M/8F, 64{+/-}9 years) participants; COPD n=6, COPD cachexia n=4, and n=5 age-matched controls. Cachexia was defined as weight loss coupled with reduced muscle strength, fatigue, anorexia, low muscle mass and/or systemic inflammation. Satellite cells were isolated and differentiated into HDMCs (myoblasts, myocytes, and myotubes). Transcriptomics data was generated from bulk skeletal muscle and HDMCs. Differential expression analysis identified transcripts significantly dysregulated (p<0.05) in COPD and/or COPD-cachexia. Weighted gene co-expression network analysis (WGCNA) preformed at the whole transcriptome and mitochondrial transcriptome levels, identified modules of co-expressed genes. Modules were tested for correlation with clinical traits and preservation between bulk tissue and HDMCs (Z-summary >2). Gene set enrichment analysis was performed for all modules. Results660 genes were significantly differentially expressed between COPD and control bulk skeletal muscle. The top upregulated and downregulated genes were IL21R-AS1 (Log2-Fold-Change [L2FC]=5.3, p=4.2x10-5) and top downregulated POU5F1B (L2FC=-5.8, p=3.1x10-2), respectively. Among the 492 genes significantly differentially expressed between COPD and COPD-cachexia bulk skeletal muscle. The top upregulated and downregulated genes were LINC02274 (L2FC=5.6, p=3.6x10-5) and ZFY (L2FC=-6.1, p=9.7x10-3) respectively. Modules 1, 9, A, B, D and H correlated with cachexia-relevant traits, and Modules 3, 7, G, and I correlated most strongly with COPD severity. Most modules (1, 2, 4, 7, 8, 9, A, B, H) were preserved across all states; Modules D and G were preserved in myoblasts only, and Module I in myotubes only. Preserved modules were enriched for contractile, inflammatory, oxidative phosphorylation, and fatty acid metabolism pathways. ConclusionsCo-expression modules linked to COPD and COPD-cachexia in bulk skeletal muscle were broadly preserved across all differentiation stages, with myoblast and myotubes most completely recapitulating disease-relevant transcriptional signatures. These findings support HMDCs, as tractable in vitro models preserving contractile, inflammatory, and mitochondrial signatures of COPD-associated muscle dysfunction.