Plant secondary metabolites as potential myostatin inhibitors: Computational insights for skeletal muscle atrophy

Myostatin (MSTN), a TGF-β superfamily member, inhibits skeletal muscle (SM) growth and differentiation. Dysregulated MSTN signaling is strongly associated with muscle-wasting conditions such as sarcopenia, cachexia, and muscular dystrophies. Inhibiting MSTN, therefore, represents a promising therapeutic strategy to restore muscle mass and function. In the present study, a library of 2,853 plant secondary metabolites (PSMs) from the PSC-db database was systematically screened using a structure-based virtual screening approach to identify potential MSTN inhibitors. Several PSMs displayed stronger predicted binding than the positive control, curcumin (−6.8 kcal·mol⁻ 1 ), with PSCdb02834 (−7.5 kcal·mol⁻ 1 ), PSCdb02056 (−7.3 kcal·mol⁻ 1 ), and PSCdb02107 (−7.3 kcal·mol⁻ 1 ) emerging as the most promising candidates. To evaluate interaction stability, we performed 200-ns all-atom molecular dynamics simulations (MDS) for each MSTN-ligand complex. The complexes retained bound conformations with low-amplitude backbone (Cα) root mean square deviation (RMSD) and persistent ligand–protein hydrogen bonds, indicating equilibrated binding. Post-MD analyses (solvent-accessible surface area (SASA), water–ligand radial distribution functions (RDF), and water-bridge occupancy) supported a solvent-assisted, stable recognition at the pocket. Consistently, MM/PBSA estimates ranked the leads more favorably than the control, with PSCdb02834 showing the most negative ΔG (relative ranking), reinforcing its prioritization for experimental validation. In silico ADMET (absorption, distribution, metabolism, excretion, toxicity) predictions indicated that all three compounds possessed favorable pharmacokinetic properties, low predicted toxicity, and drug-likeness. These findings highlight the potential of PSMs as MSTN inhibitors and provide a robust computational framework for guiding future experimental validation.