Mechanically induced osteogenic differentiation--the role of RhoA, ROCKII and cytoskeletal dynamics

Abstract

Many biochemical factors regulating progenitor cell differentiation have been examined in detail; however, the role of the local mechanical environment on stem cell fate has only recently been investigated. In this study, we examined whether oscillatory fluid flow, an exogenous mechanical signal within bone, regulates osteogenic, adipogenic or chondrogenic differentiation of C3H10T1/2 murine mesenchymal stem cells by measuring Runx2, PPARgamma and SOX9 gene expression, respectively. Furthermore, we hypothesized that the small GTPase RhoA and isometric tension within the actin cytoskeleton are essential in flow-induced differentiation. We found that oscillatory fluid flow induces the upregulation of Runx2, Sox9 and PPARgamma, indicating that it has the potential to regulate transcription factors involved in multiple unique lineage pathways. Furthermore, we demonstrate that the small GTPase RhoA and its effector protein ROCKII regulate fluid-flow-induced osteogenic differentiation. Additionally, activated RhoA and fluid flow have an additive effect on Runx2 expression. Finally, we show RhoA activation and actin tension are negative regulators of both adipogenic and chondrogenic differentiation. However, an intact, dynamic actin cytoskeleton under tension is necessary for flow-induced gene expression.

Fig. 1.

Oscillatory fluid flow has the potential to regulate the activation of RhoA and its effector protein, ROCKII. (A) Using agarose beads with GST-tagged fusion protein corresponding to residues of rhotekin RhoA-binding domain, GTP-bound RhoA was isolated using a pull-down assay. (B) Western blot analysis of GTP-bound and total RhoA indicated that a 1 hour exposure to oscillatory fluid flow induces a 2.0±0.2-fold increase (P≤0.01) in active RhoA. (C) Furthermore, ROCKII was isolated by immunoprecipitation and incubated with the ROCKII substrate protein, MYPT1, and kinase activity analyzed by western blot analysis of phosphorylated MYPT1. (D) Here we show that a 1 hour exposure to dynamic fluid flow induced a significant 3.9±0.8-fold increase in kinase activity (P≤0.01). Error bars: s.e.m. (n≥4).

Fig. 2.

Micrographs of each biochemical treatment, with or without flow, were examined to characterize alterations in actin microstructure and cell morphology. Untreated cells in the presence of flow appeared to have a denser actin misconstrue. Similar increases in actin microstructure were observed with a 1 hour incubation with LPA, and this was further enhanced with LPA treatment and dynamic flow. Inhibiting actin tension with Y27632, which inhibits ROCKII, and blebbistatin, which inhibits myosin II function, resulted in a lack of actin fibril organization and ruffled cell edges; however, there were no other gross alterations to cytoskeletal organization or cell morphology with flow. Disruption of actin polymerization with cytochalasin D incubation resulted in punctuate actin fibrils that became more pronounced with flow exposure. Treatment with jasplakinolide, which stabilizes existing filaments and inhibits actin reorganization or dynamics, appeared to have increased actin density that was not affected by flow exposure. In all cases, treatment with each biochemical in the presence of flow did not appear to induce any gross morphological alterations. Scale bar: 50 μm.

Fig. 3.

The RhoA/ROCKII pathway ultimately leading to an intact, dynamic actin cytoskeleton under tension has an important role in mechanically induced Runx2. (A) One hour incubation with LPA leading to increased amounts of active RhoA elicits a 2.6±0.8-fold increase in Runx2 expression (P≤0.01). Additionally, dynamic flow further induced Runx2 expression by 2.4±0.2-fold (P≤0.01) in LPA-treated cells, suggesting that active RhoA and oscillatory fluid flow act synergistically to enhance Runx2 upregulation. On the other hand, inhibiting ROCKII function significantly decreases Runx2 expression by 3.2±0.3-fold (P≤0.01) and abrogates flow induced Runx2 expression. (B) Multiple characteristics of the actin cytoskeleton downstream of RhoA and ROCKII activity influence flow induced Runx2 expression. Specifically, if we remove cytoskeletal tension by inhibiting myosin II activity, the cells lose the ability to upregulate Runx2 with flow. Furthermore, altering cytoskeletal polymerization with cytochalasin D or inhibiting cytoskeletal dynamics with jasplakinolide abrogates flow induced Runx2 expression. Error bars: s.e.m. (n≥6).

Fig. 4.

A dynamic actin cytoskeleton under tension is necessary for flow-induced Sox9 and PPARγ expression. (A) Altering the cytoskeletal dynamics had significant effects on both PPARγ and Sox9 basal expression. RhoA activation, and thus increased actin organization, significantly decreased PPARγ by 2.2-fold (P<0.05), while disrupting the actin cytoskeletal organization with cytochalasin D had an opposing effect and increased expression by 2.3-fold (P<0.01). Cytochalasin D had a similar effect on basal Sox9 expression, resulting in a fivefold upregulation (P<0.01). Furthermore, inhibiting actin cytoskeletal tension using Y27632 and blebbistatin also significantly increased Sox9 basal expression by threefold (P<0.01) and twofold (P<0.01), respectively. (B) Oscillatory flow-induced Sox9 expression is attenuated in the presence of pharmacological agents that inhibit ROCKII activation, myosin II ATPase function and actin polymerization. Inhibition of actin dynamics by jasplakinolide treatment resulted in a 1.85±0.17-fold decrease in Sox9 expression (P≤0.01). LPA incubation resulting in increased levels of active RhoA did not have a significant effect on flow-induced Sox9 expression. (C) Flow-induced PPARγ expression was abrogated under all conditions, indicating that an intact, dynamic actin cytoskeleton under tension is necessary. LPA treatment also abrogated flow-induced PPARγ expression, suggesting that active RhoA has a negative effect on adipogenic differentiation. Error bars: s.e.m. (n≥6).

Fig. 5.

Mechanical and biochemical regulation of gene expression. (A) Oscillatory fluid flow has the potential to regulate multiple transcription factors involved in unique lineage pathways; however, they are not regulated by the same mechanism, and the specific molecular pathways remain unknown. (B) Runx2 expression requires ROCKII activity and a dynamic actin cytoskeleton under tension. Furthermore, RhoA activation is sufficient to induce Runx2 upregulation and has an additive effect on flow-induced osteogenic differentiation. By contrast, RhoA activation downregulates PPARγ and attenuates its flow induced expression. RhoA does not alter Sox9 in either basal or flow-exposed conditions, suggesting that it does not have a significant role in chondrogenic differentiation. Interestingly, any biochemical that disrupted the actin network resulted in a decrease in Runx2 expression, and an increase in PPARγ and/or Sox9 indicating actin microstructure and ultimately, cell shape, may regulate commitment between these three fates; however, in all cases an intact actin network was necessary for flow-induced gene expression. Based on this, the presence of an actin cytoskeleton under `pre-stress', controlled by RhoA and ROCKII, is necessary for the transduction of the physical signal into alterations in gene expression. Nonetheless, other signaling pathways must also be activated by with flow to induce PPARγ and Sox9 upregulation.