MSCs adipogenesis versus osteogenesis depends on FA recruitment of NMIIB. (A) Non-silencing, NMIIB-silencing, GEF-H1 and NMIIB-silencing, and GEF-H1-silencing MSCs expressing pLKO-vector (mock), pLKO-GEFH1 or pLKO-GEF-H1_DH(m) were treated with mixed differentiation medium (OIM ∶ AIM = 1∶1) for 14 days, and stained for the presence of lipid (Oil Red O; red), the activity of ALP (purple) and Hoechst 33342 (blue). Scale bar: 50 µm. (B) Fold change in cells showing osteogenesis (ALP activity) and (C) adipogenesis (Oil red O level) in GEF-H1-silencing (n = 348 cells), NMIIB-silencing (n = 408 cells) or GEF-H1- and NMIIB-silencing (n = 370 cells), normalized to non-silencing MSCs (n = 571 cells), and GEF-H1-silencing MSCs expressing pLKO-GEFH1 (n = 596 cells) or pLKO-GEF-H1_DH(m) (n = 212 cells), normalized to pLKO-vector (mock) (n = 348 cells). Data are mean±s.e.m. **P<0.005, ***P<0.0001. (D) Model of FA-mediated signaling in the control of MSC commitment to osteogenesis. The FA abundance of GEF-H1 acts as a scaffold protein to recruit NMIIB, driving anisotropic orientation of stress fibers, FA organization and MSC osteoblast differentiation. Interference with the abundance of NMIIB in FAs alters actin cytoskeletal organization, cell stiffness, and MSC commitment to adipogenesis. The GEF-H1- and NMIIB-mediated FA signaling appears necessary to reorganize cytoskeletal mechanics, leading to an osteogenic–adipogenic switch in MSC lineage commitment.