Comparative analysis of adipose-, bone marrow-, and amniotic membrane-derived MSC secretomes and EVs reveals shared and source-specific therapeutic signatures for osteoarthritis

Aim: Mesenchymal stromal cells (MSCs) exert their therapeutic effects in osteoarthritis (OA) primarily through paracrine signaling, including secreted proteins and extracellular vesicle (EV)-associated microRNAs (miRNAs). However, the contribution of tissue origin to the composition and function of these secretomes remains unclear. This study aimed to provide a comprehensive molecular and functional comparison of secretomes from adipose-derived (ASCs), bone marrow-derived MSCs (BMSCs) and human amniotic membrane-derived MSCs, with a specific focus on OA-relevant pathways. Methods: MSCs were immunophenotyped by flow cytometry. Secretomes were profiled for 200 factors and 784 EV-miRNAs. Functional enrichment was performed using Gene Ontology and Reactome databases. In vitro, secretomes were tested on interleukin (IL)-1β-stimulated human chondrocytes to assess modulation of OA-related gene expression. Results: All MSC secretomes shared a core of factors enriched in anti-inflammatory and matrix-regulatory functions. ASCs showed the differential expression of a few modulators, potentially shifting their chondroprotective phenotype. EV-miRNAs further distinguished the MSC types. ASCs and BMSCs clustered closely in both overall miRNA content and functional enrichment, which included pathways for extracellular matrix organization, angiogenesis and IL-6 signaling. BMSC-and ASC-EVs had a higher ratio of OA-protective to destructive miRNAs, including miR-24-3p, miR-125b-5p and miR-222-3p. Functional assays confirmed that all MSC secretomes were effective in suppressing key OA-related genes in inflamed chondrocytes, with ASCs and BMSCs having a stronger activity. Conclusion: These findings support the development of MSC-derived cell-free therapies and emphasize the importance of molecular profiling in MSC source selection. Further studies are warranted to validate these observations and optimize MSC-based interventions for clinical translation in OA.