Lack of a skeletal muscle phenotype in adult human bone marrow stromal cells following xenogeneic-free expansion

Abstract

Abstract Background Many studies have elegantly shown that murine and rat bone marrow-derived mesenchymal stromal cells (bmMSCs) contribute to muscle regeneration and improve muscle function. Yet, the ability of transplanted human bmMSCs to manifest myogenic potential shows conflicting results. While human adipose- and umbilical cord-derived MSCs can be differentiated into a skeletal muscle phenotype using horse serum (HS), bmMSCs have only been shown to differentiate towards the skeletal muscle lineage using a complex mixture of cytokines followed by transfection with notch intracellular domain. Methods Since xenogeneic-free growth supplements are increasingly being used in the expansion of bmMSCs in clinical trials, we investigated the effects of human plasma and platelet lysate (P/PL) on the expression of neuromuscular markers and whether P/PL-expanded human bmMSCs could be differentiated towards a skeletal myogenic phenotype. Neuromuscular markers were measured using the highly sensitive droplet digital polymerase chain reaction for measuring the expression of Myf5, MyoD, MyoG, ACTA1, Desmin, GAP-43, and Coronin 1b transcripts, by performing immunofluorescence for the expression of Desmin, GAP-43, and MEF2, and flow cytometry for the expression of CD56/neural cell adhesion molecule (NCAM). Results Despite that bmMSCs expressed the myogenic regulatory factor (MRF) MEF2 after expansion in P/PL, bmMSCs cultured under such conditions did not express other essential MRFs including Myf5, MyoD, MyoG, or ACTA1 needed for myogenesis. Moreover, HS did not induce myogenesis of bmMSCs and hence did not induce the expression of any of these myogenic markers. P/PL, however, did lead to a significant increase in neurogenic GAP-43, as well as Desmin expression, and resulted in a high baseline expression of the neurogenic gene Coronin 1b which was sustained under further P/PL or HS culture conditions. Fetal bovine serum resulted in equally high levels of GAP-43 and Coronin 1b. Moreover, the proportion of CD56/NCAM-positive bmMSCs cultured in P/PL was 5.9 ± 2.1. Conclusions These data suggest that P/PL may prime a small portion of bmMSCs towards an early neural precursor cell type. Collectively, this shows that P/PL partially primes the cells towards a neurogenic phenotype, but does not prime adult human bmMSCs towards the skeletal muscle lineage.