We have previously reported that interfering with actin

We have previously reported that interfering with pi3k inhibitors assembly dynamics by knocking down cofilin-1 in hMSCs increased polymerized actin that promoted osteoblast cell differentiation through a mechanism of enhancing focal adhesion kinase (FAK), p38 and c-Jun N-terminal kinase (JNK) signaling (Chen et al., 2015). In the present study, we examined the effects of interfering in the actin cytoskeletal dynamics on the regulation of adipocyte differentiation of hMSCs.
Materials and methods
Results
Discussion Actin cytoskeleton undergoes cycles of actin assembly and disassembly in a balanced fashion between two states: a monomeric state (globular, G-actin) and a fibrous state (filamentous, F-actin). We have observed that adipocyte differentiation of hMSCs is associated with increased in G-actin and these changes were the opposite of what we have previously reported during osteoblastic differentiation of hMSC (Chen et al., 2015). Since G-actin/F-actin is important determinant of cell shape, our findings corroborate previous studies showing that restructuring of cellular morphology during differentiation influences lineage-specific commitment (Carvalho et al., 1998; McBeath et al., 2004; Spiegelman and Ginty, 1983). We observed significant changes in actin depolymerizing factors during hMSCs differentiation into adipocytes. CFL1 is ubiquitously expressed, actin depolymerizing factor that limits actin polymerization by binding (the non-phosphorylated form) to F-actin without capping and severe the F-actin into monomers free ends actin (G-actin) (Arber et al., 1998). CFL1 activity depends on serine 3 phosphorylation mediated by LIM kinase (LIMK), which in turn regulated by the Rho small GTPase and Rho-associated protein kinase (ROCK) (Maekawa et al., 1999). We observed that high levels of non-phosphorylated CFL1, and low level of LIMK1 and RhoA enhance adipocyte differentiation and lead to cellular enrichment of depolymerized actin. We have previously reported that stabilizing F-actin microfilaments by Phalloidin or siRNAs-mediated decreased levels of CFL1 and DSTN enhanced osteoblast differentiation of hMSCs and heterotopic bone formation in vivo, while disruption of actin cytoskeleton by CytoD or inhibition of LIMK1 activity decreased osteoblast differentiation. In the current study, we observed opposite molecular changes associated with adipocyte differentiation i.e. enhancing actin F-filament stabilization inhibited adipocyte differentiation and decreased actin assembly enhanced adipocyte differentiation. These studies suggest an inverse relationship between the actin assembly state and the osteoblastic versus adipocytic differentiation potential of hMSC. A recent study reported that reduction of actin depolymerisation by CytoD treatment enhances commitment to both osteoblastic and adipocytic lineage based on increased expression of lineage specific markers and that a single intra-tibial injection of CytoD enhances bone formation and bone marrow adipocyte formation (Sen et al., 2015, 2017). This discrepancy may be caused by differences in cellular composition of MSC cultures and/or culture conditions (short term versus long term treatment). Interestingly, our measurements of G/F actin during lineage differentiation in hMSCs showed similar reduction in actin polymerization (with increased G-actin in cells) during the first 3days of osteoblast and adipocyte differentiation suggesting that initial differentiation phase of hMSCs and prior to specific lineage commitment associated with decreased actin assembly which is consistent with the published work of Sen et al. (2017). These authors reported that short-term treatment with CytoD increased the availability of G-actin and enhanced G-actin translocation to the nucleus that increased gene expression of both osteoblastic and adipogenic differentiation (Sen et al., 2017). However, different changes were observed later during the differentiation with formation of enriched cytoplasmic F-actin (Chen et al., 2015) and branching of nuclear actin that are associated with enhanced osteoblast differentiation. In addition, our findings of the presence of an inverse actin microfilaments dynamics during osteoblast versus adipocyte differentiation is more consistent with the notion of hMSCs lineage allocation to osteoblastic or adipocytic cells occur at the level of stem cells early during differentiation and associated with significant cytoskeletal changes (McBeath et al., 2004).