Cyclic stretch is an artificial model of mechanical force loading, which induces the reorientation of vascular endothelial cells and their stress fibers in a direction perpendicular to the stretch axis. Rho family GTPases are crucial for cyclic‐stretch‐induced endothelial cell reorientation; however, the mechanism underlying stretch‐induced activation of Rho family GTPases is unknown. A screen of short hairpin RNAs targeting 63 Rho guanine nucleotide exchange factors (Rho‐GEFs) revealed that at least 11 Rho‐GEFs – Abr, alsin, ARHGEF10, Bcr, GEF‐H1 (also known as ARHGEF2), LARG (also known as ARHGEF12), p190RhoGEF (also known as ARHGEF28), PLEKHG1, P‐REX2, Solo (also known as ARHGEF40) and α‐PIX (also known as ARHGEF6) – which specifically or broadly target RhoA, Rac1 and/or Cdc42, are involved in cyclic‐stretch‐induced perpendicular reorientation of endothelial cells. Overexpression of Solo induced RhoA activation and F‐actin accumulation at cell–cell and cell–substrate adhesion sites. Knockdown of Solo suppressed cyclic‐stretch‐ or tensile‐force‐induced RhoA activation. Moreover, knockdown of Solo significantly reduced cyclic‐stretch‐induced perpendicular reorientation of endothelial cells when cells were cultured at high density, but not when they were cultured at low density or pretreated with EGTA or VE‐cadherin‐targeting small interfering RNAs. These results suggest that Solo is involved in cell–cell‐adhesion‐mediated mechanical signal transduction during cyclic‐stretch‐induced endothelial cell reorientation.

INTRODUCTION

Mechanical forces play fundamental roles in various cell functions, including morphogenesis, migration, proliferation, apoptosis and differentiation, and thereby make important contributions to the regulation of many pathophysiological processes such as embryogenesis, organogenesis, angiogenesis, tumorigenesis and tissue remodeling and homeostasis (Wozniak and Chen, 2009; Lecuit et al., 2011). In blood vessels, vascular endothelial cells are constantly exposed to mechanical forces, including periodic stretching/relaxation and shear stress caused by pulsatile blood pressure and flow, respectively (Hahn and Schwartz, 2009). Endothelial cells in the vessel exhibit an elongated spindle‐like morphology, orient their long axis parallel to the blood flow, and form a monolayer sheet that covers the inner layer of the vessel. When endothelial cells are cultured in vitro under static conditions, they have a polygonal shape and are orientated randomly. However, when they are artificially subjected to uniaxial cyclic stretch, they become elongated and align perpendicular to the stretch axis (Dartsch and Betz, 1989; Yano et al., 1996). Concomitantly, actin stress fibers are aligned perpendicular to the stretch direction. These results suggest that the mechanical forces of periodical stretching and relaxation contribute to the morphology and orientation of endothelial cells within vessels and to the rearrangement of their stress fibers. Cells sense mechanical forces through cell–substrate and cell–cell adhesion sites. Integrins and cadherins are involved in mechanical‐stretch‐induced cell responses (Geiger et al., 2009; Huveneers and de Rooij, 2013). Several signaling pathways involving RhoA, focal adhesion kinase (FAK, also known as PTK2), Src and Ca2+ are involved in cyclic‐stretch‐induced reorientation of endothelial cells and their stress fibers (Yano et al., 1996; Naruse et al., 1998; Kaunas et al., 2005). However, it is not well understood how mechanical forces are sensed and transduced to biochemical signals that lead to actin cytoskeletal rearrangement and cell reorientation.

Rho family small GTPases play key roles in actin cytoskeletal reorganization (Etienne‐Manneville and Hall, 2002). RhoA is activated by various mechanical stimuli. Inhibition of RhoA or its downstream effectors, Rho‐associated protein kinase (ROCK) or mammalian Diaphanous (mDia), changes cyclic‐stretch‐induced cell and stress fiber reorientation, indicating that RhoA plays a crucial role in mechanical signal transduction (Kaunas et al., 2005; Goldyn et al., 2009; Lee et al., 2010; Lessey et al., 2012). Rac1 is also activated by mechanical stresses and is involved in the mechanical‐stress‐induced responses of endothelial cells and other cell types (Tzima, 2006; Liu et al., 2007; DiPaolo et al., 2013). However, the mechanisms by which mechanical force induces activation of RhoA and/or Rac1 are unknown. Rho guanine nucleotide exchange factors (Rho‐GEFs) are responsible for the conversion of Rho GTPases from inactive GDP‐bound forms to active GTP‐bound forms, and these activated forms associate with downstream effector proteins to induce diverse cellular responses (Schmidt and Hall, 2002; Rossman et al., 2005; Cook et al., 2014). In the human genome, there are ∼70 distinct members of the Dbl‐related Rho‐GEF gene family, and they are thought to stimulate the GDP‐GTP exchange of Rho family GTPases (Cook et al., 2014). They possess a Dbl homology (DH) domain that is responsible for Rho‐GEF activity and a pleckstrin homology (PH) domain that mediates membrane association or modulates the Rho‐GEF activity and the specificity of the DH domain (Rossman et al., 2005). Given the crucial roles of Rho family GTPases in mechanotransduction, it is plausible that Rho‐GEFs make important contributions to mechanical‐force‐induced cell responses upstream of Rho family GTPases. Two RhoA‐targeting GEFs, LARG (also known as ARHGEF12) and GEF‐H1 (also known as ARHGEF2), have been shown to be involved in the integrin‐mediated response to mechanical force (Guilluy et al., 2011). However, little is known about the role of each Rho‐GEF in mechanical signal transduction.

This study aimed to identify Rho‐GEFs that are involved in cyclic‐stretch‐induced reorientation of human umbilical vein endothelial cells (HUVECs). A screen of a short hairpin RNA (shRNA) library targeting 63 human Rho‐GEFs revealed that at least 11 Rho‐GEFs – Abr, alsin, ARHGEF10, Bcr, GEF‐H1, LARG, p190RhoGEF (also known as ARHGEF28), PLEKHG1, P‐REX2, Solo (also known as ARHGEF40) and α‐PIX (also known as ARHGEF6) – are involved in the cyclic‐stretch‐induced reorientation of HUVECs and their stress fibers to the direction perpendicular to the stretch axis. We also provide evidence that Solo, a RhoA‐targeting GEF, contributes to cyclic‐stretch‐induced cell and stress fiber reorientation by transducing mechanical force signals at cell–cell adhesion sites.

RESULTS

Cyclic stretch induces the reorientation of HUVECs and their stress fibers

Cyclic uniaxial stretch causes endothelial cells and their stress fibers to reorientate in a direction perpendicular to the stretch axis (Yano et al., 1996; Kaunas et al., 2005). Prior to the screening of Rho‐GEFs involved in the cyclic‐stretch‐induced cell and stress fiber reorientation responses, we analyzed the effects of cyclic stretch stimuli on the orientation of HUVECs and their stress fibers. Cells were cultured to almost 100% confluency on a fibronectin‐coated silicone chamber and subjected to uniaxial cyclic stretch (20% stretch) at frequency of 1 Hz for 1 hour. Then, cells were fixed and stained with Alexa‐Fluor‐546‐tagged phalloidin to detect cell morphology and actin cytoskeletal organization. Whereas untreated HUVECs had a polygonal shape and were almost randomly orientated, cells that were treated with uniaxial cyclic stretch adopted an elongated shape and aligned their long axis and stress fibers in a direction almost perpendicular to the stretch axis (Fig. 1B). Cell orientation was quantified by fitting each cell outline to an ellipse and measuring the angle (θ) between the long axis of the ellipse and the stretch direction (Fig. 1A). Histogram analysis showed that HUVECs were almost randomly oriented prior to cyclic stretch, whereas they aligned almost perpendicular to the stretch axis after cyclic stretch treatment (Fig. 1C). The stress fiber orientation before and after cyclic stretching was also analyzed by measuring the angle (θ) between the stress fiber and the stretch direction (Fig. 1D). The percentages of the cells and stress fibers orientated perpendicular to the stretch axis (70°<θ<110°) were 34% and 20% before stretching and 67% and 74% after stretching, respectively. The value should be 22% (40° divided by 180°) if cells are oriented randomly. Thus, as previously reported (Yano et al., 1996; Kaunas et al., 2005), cyclic stretch stimuli induced the reorientation of HUVECs and their stress fibers to the direction perpendicular to the stretch axis.

Fig. 1.

Cyclic stretch induces cell and stress fiber reorientation.(A) Analysis of cell and stress fiber orientation. Cell shape was analyzed by fitting each cell outline to an ellipse. Cell and stress fiber orientation was analyzed by measuring the orientation angle (θ) of the long axis of the ellipse or stress fiber relative to the stretch axis. (B) Effects of cyclic stretch on cell and stress fiber orientation. HUVECs were co‐transfected with CFP and control shRNA plasmids and cultured for 24 hours. Cells were replated onto a fibronectin‐coated silicone stretch chamber, cultured for 24 hours and subjected to 20% cyclic uniaxial stretch for 1 hour at 1 Hz. Cells were fixed and stained with Alexa‐Fluor‐546–phalloidin. Scale bar: 50 µm. (C,D) Histograms of the percentage of cells (C) and stress fibers (SF, D) with the indicated orientation angle θ, measured as in A.

Fig. 1.

Cyclic stretch induces cell and stress fiber reorientation.(A) Analysis of cell and stress fiber orientation. Cell shape was analyzed by fitting each cell outline to an ellipse. Cell and stress fiber orientation was analyzed by measuring the orientation angle (θ) of the long axis of the ellipse or stress fiber relative to the stretch axis. (B) Effects of cyclic stretch on cell and stress fiber orientation. HUVECs were co‐transfected with CFP and control shRNA plasmids and cultured for 24 hours. Cells were replated onto a fibronectin‐coated silicone stretch chamber, cultured for 24 hours and subjected to 20% cyclic uniaxial stretch for 1 hour at 1 Hz. Cells were fixed and stained with Alexa‐Fluor‐546–phalloidin. Scale bar: 50 µm. (C,D) Histograms of the percentage of cells (C) and stress fibers (SF, D) with the indicated orientation angle θ, measured as in A.

Construction of a human Rho‐GEF shRNA library

To identify Rho‐GEFs that are involved in the cyclic‐stretch‐induced reorientation of HUVECs, we constructed a shRNA library targeting human Dbl‐like Rho‐GEFs and searched for the Rho‐GEF shRNAs that perturb the cyclic‐stretch‐induced cell reorientation response. To obtain effective shRNA constructs against each human Rho‐GEF, candidate shRNA sequences were designed according to a previously reported method (Ui‐Tei et al., 2004) and inserted into the pSUPER vector (Brummelkamp et al., 2002). To assess the silencing effect, each shRNA plasmid was introduced into Jurkat cells together with the reporter plasmid, which contained a chimeric cDNA composed of luciferase cDNA and a cDNA fragment (360–540 base pairs) of the target gene (supplementary material Fig. S1A). The primer sequences used to amplify the Rho‐GEF cDNA fragments are listed in supplementary material Table S1. If any shRNA construct effectively suppressed expression of its target gene, luciferase expression was expected to be reduced. Thus, the knockdown effect of each shRNA on expression of its target gene could be easily assessed by measuring luciferase activity. Using this reporter assay, one or two effective shRNA constructs targeting each of the 63 human Dbl‐like Rho‐GEFs were obtained (supplementary material Fig. S1B). The shRNA sequences targeting each Rho‐GEF are listed in supplementary material Table S2.

Identification of Rho‐GEFs involved in cyclic‐stretch‐induced reorientation of HUVECs and their stress fibers

We examined the effect of knockdown of each Rho‐GEF on cyclic‐stretch‐induced alignment of HUVECs perpendicular to the stretch axis. Cells were transfected with each Rho‐GEF shRNA plasmid, transferred to a fibronectin‐coated silicone chamber and then subjected to cyclic stretch (20% stretch) at 1 Hz for 1 hour. Cell orientation was analyzed by measuring the orientation angle (θ) of the long axis of the cell relative to the stretch direction (Fig. 1A), and the percentage of cells that were orientated perpendicular to the stretch axis (70°<θ<110°) was calculated (Fig. 2A). Of the shRNAs targeting the 63 Rho‐GEFs, the shRNAs targeting 11 Rho‐GEFs (Abr, alsin, ARHGEF10, Bcr, GEF‐H1, LARG, p190RhoGEF, PLEKHG1, P‐REX2, Solo and α‐PIX) suppressed the cyclic‐stretch‐induced realignment of HUVECs perpendicular to the stretch axis (Fig. 2A). Fig. 2B and supplementary material Fig. S2A show the fluorescence images of phalloidin staining of HUVECs transfected with each Rho‐GEF shRNA after cyclic stretching and the histograms of the percentage of cells orientated at the indicated angles. We also analyzed the effect of knockdown of each of the 11 Rho‐GEFs on cyclic‐stretch‐induced stress fiber reorientation. Knockdown of each Rho‐GEF suppressed the cyclic‐stretch‐induced perpendicular reorientation of stress fibers in HUVECs (supplementary material Fig. S2B,C). These results suggest that at least 11 Rho‐GEFs are involved in the cyclic‐stretch‐induced reorientation of HUVECs and their stress fibers to the direction perpendicular to the stretch axis.

Fig. 2.

Screening of Rho‐GEFs that are involved in cyclic‐stretch‐induced reorientation of HUVECs.(A) Effects of Rho‐GEF‐targeting shRNAs on cyclic‐stretch‐induced cell reorientation perpendicular to the stretch axis. HUVECs were co‐transfected with plasmids encoding CFP and each Rho‐GEF‐targeting shRNA and cultured for 24 hours. Cells were replated onto a fibronectin‐coated silicone stretch chamber, cultured for 24 hours and subjected to 20% cyclic uniaxial stretch for 1 hour at 1 Hz. Cells were fixed and stained with Alexa‐Fluor‐546–phalloidin. The orientation angle of the cell was measured as in Fig. 1 and the percentages of cells with an orientation angle of 70°<θ<110° were calculated. Data represent the means of two independent experiments (at least 50 cells per experiment). (B) Fluorescence images of HUVECs that were transfected with control or Rho‐GEF‐targeting shRNAs, subjected to cyclic stretch and stained with Alexa‐Fluor‐546–phalloidin. The lower panels show the histogram of the percentage of cells with the indicated orientation angle (θ). Scale bar: 50 µm.

Fig. 2.

Screening of Rho‐GEFs that are involved in cyclic‐stretch‐induced reorientation of HUVECs.(A) Effects of Rho‐GEF‐targeting shRNAs on cyclic‐stretch‐induced cell reorientation perpendicular to the stretch axis. HUVECs were co‐transfected with plasmids encoding CFP and each Rho‐GEF‐targeting shRNA and cultured for 24 hours. Cells were replated onto a fibronectin‐coated silicone stretch chamber, cultured for 24 hours and subjected to 20% cyclic uniaxial stretch for 1 hour at 1 Hz. Cells were fixed and stained with Alexa‐Fluor‐546–phalloidin. The orientation angle of the cell was measured as in Fig. 1 and the percentages of cells with an orientation angle of 70°<θ<110° were calculated. Data represent the means of two independent experiments (at least 50 cells per experiment). (B) Fluorescence images of HUVECs that were transfected with control or Rho‐GEF‐targeting shRNAs, subjected to cyclic stretch and stained with Alexa‐Fluor‐546–phalloidin. The lower panels show the histogram of the percentage of cells with the indicated orientation angle (θ). Scale bar: 50 µm.

To examine whether the 11 Rho‐GEFs identified in this study are expressed in HUVECs, we performed reverse transcription (RT)‐PCR analysis using the primers listed in supplementary material Table S1. Transcripts of all these Rho‐GEFs were detected in HUVECs (supplementary material Fig. S3A).

Characteristics of Rho‐GEFs involved in cyclic‐stretch‐induced reorientation of HUVECs

Schematic structures of the 11 Rho‐GEFs identified as being involved in the cyclic‐stretch‐induced reorientation of HUVECs are shown in supplementary material Fig. S3B. As a whole, they are structurally diverse with distinct domain structures (except for the DH and PH domains) and are distributed widely in the phylogenetic tree of Rho‐GEFs; however, two pairs of Rho‐GEFs are classified into the same subgroups (Abr and Bcr; GEF‐H1 and p190RhoGEF) (Cook et al., 2014). As summarized in supplementary material Table S3, the reported targets of these Rho‐GEFs are also diverse, as follows: Abr targets RhoA, Rac1 and Cdc42; alsin targets Rac1; ARHGEF10 targets RhoA, RhoB and RhoC; Bcr targets RhoA, Rac1 and Cdc42; GEF‐H1 targets RhoA; LARG targets RhoA; p190RhoGEF targets RhoA; P‐REX2 targets Rac1, Rac2, Rac3, Cdc42, RhoG and TC10 (also known as RHOQ); Solo targets RhoA and RhoC; and α‐PIX targets Rac1 and Cdc42 (Cook et al., 2014; see also references in supplementary material Table S3). The target of PLEKHG1 is unknown; therefore, we examined which Rho family GTPases are the targets of PLEKHG1. 293T cells were co‐transfected with YFP‐tagged wild‐type PLEKHG1 [PLEKHG1(WT)] or its inactive L258E mutant, in which Leu‐258 was replaced with Glu, and FLAG‐tagged RhoA, Rac1 or Cdc42. Thereafter, cell lysates were subjected to GST pull‐down assays using the GST‐tagged RhoA‐binding domain (RBD) of rhotekin and the p21 (Rac1/Cdc42)‐binding domain (PBD) of PAK1, to which active GTP‐bound forms of RhoA and Rac or Cdc42 specifically bind, respectively. This assay revealed that PLEKHG1(WT) activated Rac1 and Cdc42, but not RhoA, whereas PLEKHG1(L258E) did not activate any Rho family GTPases (supplementary material Fig. S3C). In addition, overexpression of CFP‐tagged PLEKHG1(WT) in HeLa cells induced membrane ruffling and dorsal stress fibers, in place of marginal stress fibers observed in the surrounding PLEKHG1‐non‐expressing cells (supplementary material Fig. S3D). These phenotypes are frequently observed in Rac‐activated cells (Vallenius, 2013), further suggesting that PLEKHG1 activates Rac. These results suggest that a wide spectrum of Rho‐GEFs with distinct domain structures and target specificities are involved in cyclic‐stretch‐induced cell reorientation.

Solo induces F‐actin accumulation at cell–cell and cell–substrate adhesion sites

Solo (also named ARHGEF40 and Scambio) is a GEF that targets RhoA and RhoC (Curtis et al., 2004; Tse et al., 2005). Its zebrafish ortholog, called Quattro (Quo), is involved in convergent extension cell movements in gastrula, a process related to mechanotransduction (Daggett et al., 2004). However, the cellular function of Solo is unknown. Here, we focused on the role of Solo in cyclic‐stretch‐induced cell and stress fiber orientation. To examine the role of Solo in actin cytoskeletal remodeling, we analyzed the effect of Solo overexpression on actin organization. HUVECs were transfected with YFP‐tagged Solo(WT) or its GEF‐inactive L1217E mutant, in which Leu‐1217 was replaced with Glu, and stained with Alexa‐Fluor‐568–phalloidin to detect F‐actin. Overexpression of YFP–Solo(WT) induced the formation of aberrantly accumulated F‐actin and thick stress fibers near to cell–cell contact sites in both confluent and subconfluent cells, whereas this was not observed in surrounding cells that did not express YFP–Solo(WT) (Fig. 3A,B). In some instances, expression of YFP–Solo(WT) induced aberrant accumulation of actin filaments at focal adhesions, the sites of stress fiber termination and cell attachment to the extracellular matrix (ECM) (Fig. 3C). In both cases, YFP–Solo(WT) accumulated at regions where F‐actin accumulated. By contrast, expression of YFP–Solo(L1217E) had no apparent effect on actin organization and it localized diffusely in the cytoplasm (Fig. 3D). These results suggest that Solo has the potential to promote the formation of actin stress fibers and focal adhesions by activating RhoA. Time‐lapse fluorescence imaging of HUVECs co‐expressing CFP–Solo(WT) and YFP–Lifeact revealed that Solo colocalized and moved together with F‐actin‐assembled structures (supplementary material Fig. S4A and Movie 1), suggesting that Solo functions in the F‐actin‐assembled regions.

Fig. 3.

Effects of Solo overexpression on actin cytoskeletal organization.‐tagged Solo (WT or L1217E), cultured for 24 hours, fixed and stained with Alexa‐Fluor‐546–phalloidin. Expression of Solo(WT) induced aberrant F‐actin accumulation and thick stress fibers at cell–cell adhesion sites in confluent (A) and subconfluent cells (B). Expression of Solo(WT) also induced aberrant F‐actin accumulation at cell–substrate adhesion sites (C). Solo localized at sites where F‐actin accumulated. By contrast, Solo(L1217E) was diffusely distributed in the cytoplasm and had no apparent effect on the actin cytoskeleton (D). Scale bars: 10 µm.

Fig. 3.

Effects of Solo overexpression on actin cytoskeletal organization.‐tagged Solo (WT or L1217E), cultured for 24 hours, fixed and stained with Alexa‐Fluor‐546–phalloidin. Expression of Solo(WT) induced aberrant F‐actin accumulation and thick stress fibers at cell–cell adhesion sites in confluent (A) and subconfluent cells (B). Expression of Solo(WT) also induced aberrant F‐actin accumulation at cell–substrate adhesion sites (C). Solo localized at sites where F‐actin accumulated. By contrast, Solo(L1217E) was diffusely distributed in the cytoplasm and had no apparent effect on the actin cytoskeleton (D). Scale bars: 10 µm.

Solo is involved in cyclic‐stretch‐induced RhoA activation

To confirm that Solo activates RhoA, YFP‐tagged Solo(WT) or its C‐terminal fragment (amino acids 1058–1519), which contains the DH and PH domains (Fig. 4A), was expressed in 293T cells, and cell lysates were subjected to GST pull‐down assays using GST–rhotekin (RBD). This assay revealed that expression of Solo(WT), but not its C‐terminal fragment, caused RhoA activation (Fig. 4B), indicating that Solo activates RhoA and that the N‐terminal region of Solo is required for exerting its RhoA‐GEF activity in cells.

Fig. 4.

Solo is involved in cyclic‐stretch‐induced RhoA activation.(A) Structures of Solo(WT) and its C‐terminal fragment (amino acids 1058–1519). (B) Expression of Solo(WT), but not Solo(1058–1519), induces RhoA activation. 293T cells were transfected with YFP, YFP–Solo(WT) or YFP–Solo(1058–1519) and cultured for 48 hours. RhoA activity was measured by using GST–rhotekin(RBD) pull‐down assays. Relative RhoA activity is shown as means±s.d. of three independent experiments. **P<0.01 (one‐way ANOVA followed by Dunnett's test). (C) Cyclic stretch induces perpendicular orientation of MDCKII cells and their stress fibers. MDCKII cells were subjected to cyclic stretch, as in Fig. 1. Cells were fixed and stained with Alexa‐Fluor‐546–phalloidin. Scale bar: 20 µm. (D) Effect of Solo‐targeting siRNAs on expression of Solo in MDCKII cells. Cells were transfected with control or dog Solo‐targeting siRNAs and cultured for 48 hours. Cell lysates were analyzed by immunoblotting with an anti‐Solo antibody. (E) Effect of Solo knockdown on cyclic‐stretch‐induced RhoA activation. MDCKII cells were transfected with Solo‐targeting siRNAs and cultured for 24 hours. Cells were replated onto a fibronectin‐coated silicone stretch chamber, cultured for 24 hours and subjected to cyclic stretch for 1 minute at 1 Hz. Cell lysates were subjected to GST–rhotekin(RBD) pull‐down assays. Relative RhoA activity is shown as means±s.d. of 12 independent experiments. *P<0.05; **P<0.01 (one‐way ANOVA followed by Dunnett's test). (F) Suppression of Solo and LARG expression by siRNAs. HUVECs were transfected with Solo‐ or LARG‐targeting siRNAs and cultured for 48 hours. Cell lysates were analyzed by immunoblotting with the indicated antibodies. (G) Effects of double knockdown of Solo and LARG on cyclic‐stretch‐induced cell and stress fiber reorientation. HUVECs were transfected with Solo‐ or LARG‐targeting siRNAs, or both, and subjected to cyclic stretch, as in Fig. 1. Cells were fixed and stained with Alexa‐Fluor‐546–phalloidin, and the orientation parameters (cos2θ) of the cell body and stress fibers were measured. Data represent means±s.d. of three independent experiments (at least 100 cells or 500 stress fibers per experiment). ***P<0.001; n.s., not significant (one‐way ANOVA followed by Tukey's test).

Fig. 4.

Solo is involved in cyclic‐stretch‐induced RhoA activation.(A) Structures of Solo(WT) and its C‐terminal fragment (amino acids 1058–1519). (B) Expression of Solo(WT), but not Solo(1058–1519), induces RhoA activation. 293T cells were transfected with YFP, YFP–Solo(WT) or YFP–Solo(1058–1519) and cultured for 48 hours. RhoA activity was measured by using GST–rhotekin(RBD) pull‐down assays. Relative RhoA activity is shown as means±s.d. of three independent experiments. **P<0.01 (one‐way ANOVA followed by Dunnett's test). (C) Cyclic stretch induces perpendicular orientation of MDCKII cells and their stress fibers. MDCKII cells were subjected to cyclic stretch, as in Fig. 1. Cells were fixed and stained with Alexa‐Fluor‐546–phalloidin. Scale bar: 20 µm. (D) Effect of Solo‐targeting siRNAs on expression of Solo in MDCKII cells. Cells were transfected with control or dog Solo‐targeting siRNAs and cultured for 48 hours. Cell lysates were analyzed by immunoblotting with an anti‐Solo antibody. (E) Effect of Solo knockdown on cyclic‐stretch‐induced RhoA activation. MDCKII cells were transfected with Solo‐targeting siRNAs and cultured for 24 hours. Cells were replated onto a fibronectin‐coated silicone stretch chamber, cultured for 24 hours and subjected to cyclic stretch for 1 minute at 1 Hz. Cell lysates were subjected to GST–rhotekin(RBD) pull‐down assays. Relative RhoA activity is shown as means±s.d. of 12 independent experiments. *P<0.05; **P<0.01 (one‐way ANOVA followed by Dunnett's test). (F) Suppression of Solo and LARG expression by siRNAs. HUVECs were transfected with Solo‐ or LARG‐targeting siRNAs and cultured for 48 hours. Cell lysates were analyzed by immunoblotting with the indicated antibodies. (G) Effects of double knockdown of Solo and LARG on cyclic‐stretch‐induced cell and stress fiber reorientation. HUVECs were transfected with Solo‐ or LARG‐targeting siRNAs, or both, and subjected to cyclic stretch, as in Fig. 1. Cells were fixed and stained with Alexa‐Fluor‐546–phalloidin, and the orientation parameters (cos2θ) of the cell body and stress fibers were measured. Data represent means±s.d. of three independent experiments (at least 100 cells or 500 stress fibers per experiment). ***P<0.001; n.s., not significant (one‐way ANOVA followed by Tukey's test).

To examine the role of Solo in cyclic‐stretch‐induced RhoA activation, we analyzed the effect of Solo knockdown on cyclic‐stretch‐induced RhoA activation. We used Madin‐Darby canine kidney II (MDCKII) cells in this experiment, because MDCKII cells reorientated perpendicular to the stretch direction (Fig. 4C) and consistently exhibited RhoA activation in response to cyclic stretching. The small interfering RNAs (siRNAs) targeting dog Solo effectively suppressed the expression of endogenous Solo in MDCKII cells (Fig. 4D). Cells were treated with control or dog Solo‐targeting siRNAs and subjected to cyclic stretch for 1 minute at 1 Hz. Measurement of activated RhoA by GST–RBD pull‐down assays revealed that RhoA is activated by cyclic stretch and that knockdown of Solo significantly decreases the level of cyclic‐stretch‐induced RhoA activation (Fig. 4E). This indicates that Solo is involved in the cyclic‐stretch‐induced RhoA activation.

Effect of double knockdown of Solo and LARG on cyclic‐stretch‐induced cell and stress fiber orientation

To examine the relationships between the roles of Solo and other RhoA‐targeting GEFs, we analyzed the effects of double knockdown of Solo and LARG on cyclic‐stretch‐induced cell and stress fiber reorientation and compared them with the effects of Solo or LARG single knockdown. The siRNAs targeting Solo and LARG effectively suppressed the expression of the respective target proteins (Fig. 4F). The cell and stress fiber orientations with respect to the stretch axis were expressed by using the orientation parameter (cos2θ) (Goldyn et al., 2009). The values of cos2θ = 1, 0 and −1 indicate the perfectly parallel, random and perpendicular orientation, respectively. Double knockdown of Solo and LARG suppressed the cyclic‐stretch‐induced perpendicular orientation of cells to a similar extent to that of Solo or LARG single knockdown (Fig. 4G, left panel), suggesting that Solo and LARG function in a similar pathway for stretch‐induced cell alignment response. By contrast, double knockdown of Solo and LARG suppressed the stretch‐induced perpendicular orientation of stress fibers to a significantly greater extent than Solo or LARG single knockdown (Fig. 4G, right panel), which suggests that Solo and LARG contribute to the cyclic‐stretch‐induced stress fiber reorientation in different ways.

Effects of cell density on Solo‐knockdown‐induced suppression of perpendicular orientation of cells and stress fibers

Cells sense environmental mechanical forces through cell–cell and cell–substrate adhesion sites (DuFort et al., 2011). To examine whether Solo is involved in the mechanical force transduction through cell–cell contacts or cell–substrate adhesions, HUVECs were transfected with control or Solo‐targeting shRNAs and cultured at a high or low cell density. Thereafter, the effects of Solo knockdown on the cyclic‐stretch‐induced cell reorientation response were analyzed. When HUVECs were transfected with control shRNA, cultured at a high density (the usual conditions used for Rho‐GEF screening) where most cells contact each other, and subjected to cyclic stretch, cell orientation changed from random (cos2θ = −0.09) to an almost perpendicular direction (cos2θ = −0.70) to the stretch axis (Fig. 5A). By contrast, when control shRNA cells were cultured at a low density, where most cells do not contact each other, and subjected to cyclic stretch, cell orientation changed from random (+0.06) to slightly perpendicular direction (−0.19) (Fig. 5B). These results indicate that cyclic stretching weakly, but significantly, increased the number of cells orientated perpendicular to the stretch axis at low cell density and that cell–substrate adhesion only partially mediates cyclic‐stretch‐induced cell reorientation. Comparison of the results of high and low cell density cultures suggests that at high cell density, cell–cell contacts play a larger role than cell–substrate adhesions in mediating cyclic‐stretch‐induced cell reorientation.

Fig. 5.

Knockdown of Solo suppresses cyclic‐stretch‐induced perpendicular cell and stress fiber orientation when cells are cultured at high density, but not when they are cultured at low density.‐targeting shRNAs and seeded on a fibronectin‐coated silicone stretch chamber at a high (A) or low density (B). After 24 hours, cells were subjected to cyclic stretch as in Fig. 1, fixed and stained with Alexa‐Fluor‐546–phalloidin. Scale bars: 50 µm. The orientation parameters (cos2θ) of cell bodies and stress fibers were measured. Data represent means±s.d. of three independent experiments (at least 60 cells and 300 stress fibers per experiment). *P<0.05; **P<0.01; ***P<0.001; n.s., not significant (one‐way ANOVA followed by Tukey's test).

Fig. 5.

Knockdown of Solo suppresses cyclic‐stretch‐induced perpendicular cell and stress fiber orientation when cells are cultured at high density, but not when they are cultured at low density.‐targeting shRNAs and seeded on a fibronectin‐coated silicone stretch chamber at a high (A) or low density (B). After 24 hours, cells were subjected to cyclic stretch as in Fig. 1, fixed and stained with Alexa‐Fluor‐546–phalloidin. Scale bars: 50 µm. The orientation parameters (cos2θ) of cell bodies and stress fibers were measured. Data represent means±s.d. of three independent experiments (at least 60 cells and 300 stress fibers per experiment). *P<0.05; **P<0.01; ***P<0.001; n.s., not significant (one‐way ANOVA followed by Tukey's test).

To examine the effects of Solo knockdown, HUVECs were transfected with two independent Solo‐targeting shRNAs. When cells were cultured at a high density and subjected to cyclic stretching, the cell orientation parameters (cos2θ) of Solo‐knockdown cells (−0.41, −0.41) were significantly increased, compared with that in control shRNA‐treated cells (−0.70) (Fig. 5A). By contrast, when cells were cultured at a low density and subjected to cyclic stretching, the cell orientation parameters of Solo‐knockdown cells (−0.24, −0.27) and control shRNA‐treated cells (−0.19) did not differ significantly (Fig. 5B). Similar results were obtained when we compared the orientation parameters of stress fibers in control and Solo‐knockdown cells cultured at high and low cell density (Fig. 5A,B, bottom panels). These results suggest that Solo is primarily involved in cell–cell‐contact‐mediated reorientation of HUVECs and their stress fibers in response to cyclic stretch.

Effects of EGTA or VE‐cadherin siRNA treatment on Solo‐knockdown‐induced suppression of perpendicular orientation of cells and stress fibers

To further examine the role of Solo in cell–cell‐contact‐mediated reorientation of HUVECs, we analyzed the effects of Solo knockdown on cyclic‐stretch‐induced cell and stress fiber reorientation in the presence or absence of EGTA, which is expected to inhibit cell–cell adhesion through inhibition of cell adhesion molecules, including vascular endothelial (VE)‐cadherin. Treatment with EGTA reduced VE‐cadherin localization in the cell–cell contact sites in HUVECs (Fig. 6A). When control siRNA‐transfected cells were cultured in the presence of EGTA and subjected to cyclic stretch, the values of the orientation parameters of cell bodies (−0.31) and their stress fibers (−0.45) were increased, compared with those in control EGTA‐non‐treated cells (−0.71 and −0.73 for cell bodies and stress fibers, respectively) (Fig. 6B,C), indicating that EGTA treatment suppresses cyclic‐stretch‐induced perpendicular orientation of cells and stress fibers. The orientation parameters of cells and stress fibers after cyclic stretch were significantly increased by Solo knockdown in EGTA‐non‐treated cells, but they did not significantly differ between control cells and Solo‐knockdown cells after EGTA treatment (Fig. 6B,C), which suggests that Solo is involved in cyclic‐stretch‐induced cell and stress fiber reorientation primarily through cell–cell‐contact‐mediated mechanical force transduction.

Fig. 6.

EGTA treatment abrogates the suppressive effect of Solo knockdown on cyclic‐stretch‐induced perpendicular cell and stress fiber orientation.(A) Treatment with EGTA reduces the localization of VE‐cadherin at the cell adhesion sites. HUVECs were fixed and stained with an anti‐VE‐cadherin antibody. (B) Effects of EGTA treatment on cyclic‐stretch‐induced cell and stress fiber orientation. HUVECs were transfected with control or Solo‐targeting siRNAs, cultured for 24 hours and transferred to a fibronectin‐coated silicone stretch chamber. After culture for a further 24 hours, cells were treated with 3 mM EGTA, cultured for 1 hour and subjected to cyclic stretch as in Fig. 1. Then, cells were fixed and stained with Rhodamine–phalloidin. Scale bars: 20 µm. (C) The orientation parameters (cos2θ) of cell bodies and stress fibers were measured. Data represent means±s.d. of three independent experiments (at least 100 cells and 500 stress fibers per experiment). **P<0.01; ***P<0.001; n.s., not significant (one‐way ANOVA followed by Tukey's test).

Fig. 6.

EGTA treatment abrogates the suppressive effect of Solo knockdown on cyclic‐stretch‐induced perpendicular cell and stress fiber orientation.(A) Treatment with EGTA reduces the localization of VE‐cadherin at the cell adhesion sites. HUVECs were fixed and stained with an anti‐VE‐cadherin antibody. (B) Effects of EGTA treatment on cyclic‐stretch‐induced cell and stress fiber orientation. HUVECs were transfected with control or Solo‐targeting siRNAs, cultured for 24 hours and transferred to a fibronectin‐coated silicone stretch chamber. After culture for a further 24 hours, cells were treated with 3 mM EGTA, cultured for 1 hour and subjected to cyclic stretch as in Fig. 1. Then, cells were fixed and stained with Rhodamine–phalloidin. Scale bars: 20 µm. (C) The orientation parameters (cos2θ) of cell bodies and stress fibers were measured. Data represent means±s.d. of three independent experiments (at least 100 cells and 500 stress fibers per experiment). **P<0.01; ***P<0.001; n.s., not significant (one‐way ANOVA followed by Tukey's test).

VE‐cadherin mediates cell–cell adhesions and mechanical‐stress‐induced endothelial cell responses (Tzima et al., 2005; Liu et al., 2007). To further examine the role of Solo in cell–cell‐adhesion‐mediated endothelial cell reorientation, we analyzed the effects of Solo knockdown on cyclic‐stretch‐induced reorientation under the VE‐cadherin‐depleted conditions. Treatment with VE‐cadherin‐targeting siRNAs reduced the expression of endogenous VE‐cadherin in HUVECs (Fig. 7A). When HUVECs were treated with VE‐cadherin siRNAs and subjected to cyclic stretch, the values of the orientation parameters of cell bodies (−0.50) and their stress fibers (−0.59) were significantly increased, compared with those in control siRNA‐treated cells (−0.75 and −0.86 for cell bodies and stress fibers, respectively) (Fig. 7B,C), indicating that depletion of VE‐cadherin suppresses cyclic‐stretch‐induced perpendicular orientation of cells and stress fibers. The orientation parameters of cells and stress fibers after cyclic stretch were significantly increased by Solo knockdown in control HUVECs (Fig. 5A; Fig. 6C, −EGTA), but they were not significantly changed by Solo knockdown in VE‐cadherin‐depleted cells (Fig. 7B,C). Taken together with the findings in Figs 5 and 6, these results suggest that Solo is primarily involved in cell–cell‐contact‐mediated mechanical force transduction during the cyclic‐stretch‐induced cell and stress fiber reorientation responses.

Fig. 7.

Knockdown of VE‐cadherin abrogates the suppressive effect of Solo knockdown on cyclic‐stretch‐induced perpendicular cell and stress fiber orientation.(A) Effect of VE‐cadherin‐targeting siRNAs on expression of VE‐cadherin in HUVECs. Cells were transfected with control or VE‐cadherin‐targeting siRNAs and cultured for 48 hours. Cell lysates were analyzed by immunoblotting with an anti‐VE‐cadherin antibody. (B) Effects of VE‐cadherin knockdown on cyclic‐stretch‐induced cell and stress fiber orientation. HUVECs were transfected with VE‐cadherin‐targeting siRNA with or without Solo‐targeting siRNAs, cultured for 24 hours and transferred to a fibronectin‐coated silicone stretch chamber. After culture for a further 24 hours, cells were subjected to cyclic stretch as in Fig. 1. Then, cells were fixed and stained with Rhodamine–phalloidin. Scale bar: 20 µm. (C) The orientation parameters (cos2θ) of cell bodies and stress fibers were measured. Data represent means±s.d. of three independent experiments (at least 100 cells and 500 stress fibers per experiment). ***P<0.001; n.s., not significant (one‐way ANOVA followed by Tukey's test).

Fig. 7.

Knockdown of VE‐cadherin abrogates the suppressive effect of Solo knockdown on cyclic‐stretch‐induced perpendicular cell and stress fiber orientation.(A) Effect of VE‐cadherin‐targeting siRNAs on expression of VE‐cadherin in HUVECs. Cells were transfected with control or VE‐cadherin‐targeting siRNAs and cultured for 48 hours. Cell lysates were analyzed by immunoblotting with an anti‐VE‐cadherin antibody. (B) Effects of VE‐cadherin knockdown on cyclic‐stretch‐induced cell and stress fiber orientation. HUVECs were transfected with VE‐cadherin‐targeting siRNA with or without Solo‐targeting siRNAs, cultured for 24 hours and transferred to a fibronectin‐coated silicone stretch chamber. After culture for a further 24 hours, cells were subjected to cyclic stretch as in Fig. 1. Then, cells were fixed and stained with Rhodamine–phalloidin. Scale bar: 20 µm. (C) The orientation parameters (cos2θ) of cell bodies and stress fibers were measured. Data represent means±s.d. of three independent experiments (at least 100 cells and 500 stress fibers per experiment). ***P<0.001; n.s., not significant (one‐way ANOVA followed by Tukey's test).

Knockdown of Solo suppresses tensile‐force‐induced RhoA activation

To further examine the role of Solo in mechanical force transduction, we analyzed whether knockdown of Solo affects the tensile‐force‐induced RhoA activation. E‐cadherin‐coated magnetic beads were put on cultured MDCKII cells stably expressing CFP–RhoA and tensile force was applied by placing a permanent magnet on the top of the culture dish (Fig. 8A). To measure activated RhoA, cell lysates were subjected to GST–rhotekin pull‐down assays. This assay revealed that RhoA is significantly activated 10 minutes after tensile force application and that knockdown of Solo suppresses tensile‐force‐induced RhoA activation (Fig. 8B). This result suggests that Solo is involved in the cadherin‐mediated mechanical‐force‐induced RhoA activation.

Fig. 8.

Effect of Solo knockdown on tensile‐force‐induced RhoA activation.(A) Schematic of the process of tensile force application. The E‐cadherin‐coated magnetic beads were put on cultured MDCKII cells stably expressing CFP–RhoA. Tensile force (F) was applied by placing a permanent magnet on the top of the culture dish. (B) Effect of Solo knockdown on tensile‐force‐induced RhoA activation. Tensile force was applied to MDCKII cells as in A. Cell lysates were subjected to GST–rhotekin(RBD) pull‐down assays. Relative RhoA activity is shown as means±s.d. of six independent experiments. *P<0.05; n.s., not significant (one‐way ANOVA followed by Dunnett's test).

Fig. 8.

Effect of Solo knockdown on tensile‐force‐induced RhoA activation.(A) Schematic of the process of tensile force application. The E‐cadherin‐coated magnetic beads were put on cultured MDCKII cells stably expressing CFP–RhoA. Tensile force (F) was applied by placing a permanent magnet on the top of the culture dish. (B) Effect of Solo knockdown on tensile‐force‐induced RhoA activation. Tensile force was applied to MDCKII cells as in A. Cell lysates were subjected to GST–rhotekin(RBD) pull‐down assays. Relative RhoA activity is shown as means±s.d. of six independent experiments. *P<0.05; n.s., not significant (one‐way ANOVA followed by Dunnett's test).

DISCUSSION

Mechanical forces play essential roles in many cell activities and functions. Rho family GTPases, including RhoA, Rac1 and Cdc42, are crucially involved in mechanical‐force‐induced cell responses, such as cell shape change, migration and proliferation, mostly by regulating actin cytoskeletal reorganization (Tzima, 2006; Birukov, 2009; Lessey et al., 2012). However, little is known about the mechanisms by which mechanical forces induce the activation of Rho family GTPases. Using a shRNA library targeting 63 Rho‐GEFs, we identified 11 Rho‐GEFs that are involved in cyclic‐stretch‐induced alignment of HUVECs and their stress fibers perpendicular to the stretch axis. Consistent with the important roles of RhoA and Rac1 in mechanical force responses, the 11 Rho‐GEFs identified in this study have distinct target specificities towards RhoA, Rac1 and Cdc42 (supplementary material Table S3). Among these 11 Rho‐GEFs, GEF‐H1, LARG and p190RhoGEF are RhoA‐targeting GEFs that are involved in integrin‐mediated mechanotransduction at focal adhesions (Lim et al., 2008; Tomar and Schlaepfer, 2009; Guilluy et al., 2011). In addition, single‐nucleotide polymorphism analysis revealed that the genes encoding ARHGEF10 and PLEKHG1 are associated with atherothrombotic stroke and high blood pressure traits, respectively (Matsushita et al., 2010; Franceschini et al., 2013), suggesting that they have roles in blood vessel homeostasis and functions. However, little is known about the regulation mechanisms and functional roles of these Rho‐GEFs in mechanotransduction in vascular endothelial cells. Our results indicate that multiple Rho‐GEFs with distinct target specificities are involved in the transmission of mechanical signals to activate distinct Rho family GTPases.

Cyclic‐stretch‐induced cell and stress fiber reorientation is a multistep process that involves disruption and/or movement of focal adhesions and stress fibers along the stretch direction, lamellipodium formation and cell elongation at the cell ends perpendicular to the stretch direction, and reassembly and alignment of focal adhesions and stress fibers perpendicular to the stretch direction. The 11 Rho‐GEFs identified in this study appear to play crucial roles in the cyclic‐stretch‐induced cell and stress fiber reorientation responses by coordinately regulating the activities of their specific target Rho family GTPases. Thus, it is plausible that RhoA‐targeting GEFs, such as ARHGEF10, GEF‐H1, LARG, p190RhoGEF and Solo, are involved in the reassembly of focal adhesions and stress fibers perpendicular to the stretch axis, whereas Rac1/Cdc42‐targeting GEFs, such as alsin, PLEKHG1, P‐REX2 and α‐PIX, are involved in lamellipodium formation and cell elongation at the cell ends perpendicular to the stretch axis. Inhibition of RhoA or its effectors, ROCK or mDia, causes cyclic‐stretch‐induced stress fiber orientation in the direction parallel to the stretch axis (Kaunas et al., 2005). Knockdown of each of the 11 Rho‐GEFs or double knockdown of Solo and LARG was insufficient to completely block perpendicular alignment of cells or induce parallel orientation of stress fibers, indicating that multiple Rho‐GEFs are required for the cyclic‐stretch‐induced RhoA activation and reorientation of cells and their stress fibers perpendicular to the stretch axis. Further studies on the mechanisms that regulate the activation and localization of each Rho‐GEF will provide better understanding of the functional roles of individual Rho‐GEFs and Rho family GTPases and the mechanism underlying their spatiotemporal coordination during cyclic‐stretch‐induced actin cytoskeletal remodeling and cell reorientation.

Cells sense external mechanical forces primarily through cell attachment sites, including cell–substrate and cell–cell adhesions. Integrin‐mediated linkages between the actin cytoskeleton and the ECM at focal adhesions are involved in the sensing of mechanical signals, such as the spatial distribution of adhesive ligands and matrix rigidity (Roca‐Cusachs et al., 2012; Schwarz and Gardel, 2012). Cell–cell adhesions mediated by cadherins or other transmembrane proteins are also involved in mechanosensing, such as the shear‐stress‐induced response and stretch‐induced proliferation of endothelial cells (Tzima et al., 2005; Liu et al., 2007; Huveneers and de Rooij, 2013). Overexpression of Solo caused marked effects on the actin cytoskeletal organization in HUVECs, including the formation of thick stress fibers and aberrant F‐actin accumulation at cell–cell contact sites and focal adhesions, and Solo localized at sites of F‐actin accumulation. These results suggest that Solo is involved in mechanotransduction at cell–cell and/or cell–ECM adhesion sites. Knockdown of Solo suppressed cyclic‐stretch‐induced reorientation of HUVECs perpendicular to the stretch axis when cells were cultured at high density, but it had no apparent effect when cells were cultured at low density. Furthermore, knockdown of Solo had no apparent effect on cyclic‐stretch‐induced cell reorientation when cells were treated with EGTA or VE‐cadherin siRNA. Therefore, Solo appears to predominantly contribute to cell–cell contact‐ and VE‐cadherin‐mediated mechanotransduction during cyclic‐stretch‐induced cell and stress fiber reorientation of HUVECs. Quo, a zebrafish ortholog of Solo, is involved in convergent extension cell movement during gastrulation of the zebrafish embryo (Daggett et al., 2004). Cell–cell‐contact‐mediated mechanotransduction plays an essential role in convergent extension (Lecuit et al., 2011). Therefore, Quo likely mediates mechanical force signal transduction in zebrafish, in a manner similar to how Solo functions in the stretch‐induced reorientation of HUVECs.

It is unknown how mechanical stimuli activate the 11 Rho‐GEFs identified in this study. Recent studies identified several candidate mechanosensor molecules, such as p130Cas (also known as BCAR1), talin, vinculin, α‐catenin and filamin, which directly sense and transduce mechanical force signals into chemical signals (Sawada et al., 2006; del Rio et al., 2009; Grashoff et al., 2010; Yonemura et al., 2010; Ehrlicher et al., 2011). Rho‐GEFs are activated by their association with other proteins or by modifications, such as phosphorylation. For example, GEF‐H1 is activated by its release from microtubules, whereas p190RhoGEF is activated by FAK association and FAK‐mediated tyrosine phosphorylation (Lim et al., 2008; Tomar and Schlaepfer, 2009). To examine the possibility that Solo is phosphorylated after cyclic stretch, we analyzed the gel mobility shift and tyrosine phosphorylation of Solo before and after cyclic stretch of HUVECs and MDCKII cells by immunoblot analysis with anti‐Solo and anti‐phospho‐tyrosine antibodies; however, neither the gel mobility shift nor the phospho‐tyrosine signal of Solo was detected (supplementary material Fig. S4B). Although Dbl‐like Rho‐GEFs have a common catalytic DH domain, many Rho‐GEFs have additional sequences. We showed that expression of full‐length Solo induces RhoA activation in cells, but its C‐terminal DH‐containing fragment does not, indicating that the N‐terminal region of Solo is required for its activation in cells. One possible explanation for this result is that the Rho‐GEF activity of Solo is enhanced by the tensional‐force‐induced conformational change in the DH domain and that the N‐terminal region functions as one of the supporting points for this process. Thus, Solo might function by itself as a mechanosensor that is activated by mechanical‐force‐induced conformational changes. Identification of the Rho‐GEFs that are involved in cyclic‐stretch‐induced cell reorientation in this study will help future studies to elucidate the mechanisms by which mechanical forces activate Rho‐GEFs and their downstream Rho family GTPases.

MATERIALS AND METHODS

Antibodies

A rabbit polyclonal antibody recognizing human and dog Solo was raised against the C‐terminal peptide (LSRQSHARALSDPTTPL) of human Solo. Other antibodies were purchased as follows: rabbit polyclonal antibodies against GFP (A‐6455, Life Technologies, Grand Island, NY), LARG (sc‐25638, Santa Cruz Biotechnology, Dallas, TX) and VE‐cadherin (ALX‐210‐232‐C100, Enzo Life Sciences, Farmingdale, NY), mouse monoclonal antibodies against FLAG (M2, Sigma‐Aldrich, St Louis, MO), RhoA (sc‐418, Santa Cruz Biotechnology) and β‐actin (AC‐15, Sigma‐Aldrich), and Alexa‐Fluor‐488‐labeled goat anti‐rabbit‐IgG antibody (A11034, Life Technologies).

Cell culture and transfection

HUVECs were prepared from human umbilical cord veins as described previously (Ueki et al., 2010) or purchased from Life Technologies. Cells were cultured at 37°C under a 5% CO2 atmosphere in Medium 199 (Life Technologies) supplemented with 20% fetal calf serum (Life Technologies) and 10 ng/ml human basic fibroblast growth factor (Austral Biologicals, San Ramon, CA). The cells were used within seven passages. Jurkat T cells were obtained from the Cell Resource Center, Tohoku University (Sendai, Japan) and maintained in RPMI 1640 medium supplemented with 10% fetal calf serum. HeLa and MDCKII cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. Cells were transfected with plasmids and siRNAs using Lipofectamine LTX and RNAiMax (Life Technologies), respectively.

Plasmid construction and siRNAs

The cDNAs of human Solo and PLEKHG1 were amplified by RT‐PCR from total mRNA isolated from MCF10A cells. Expression plasmids encoding YFP‐tagged Solo and PLEKHG1 were constructed by inserting the PCR‐amplified cDNAs into the pEYFP‐C1 vector (Clontech, Mountain View, CA). The plasmids for GEF‐inactive Solo(L1217E) and PLEKHG1(L258E) were constructed using a site‐directed mutagenesis kit (Agilent Technologies, Santa Clara, CA). Expression plasmids encoding FLAG‐tagged RhoA, Rac1 and Cdc42 were constructed by inserting the cDNAs for RhoA, Rac1 and Cdc42 into the pFLAG‐C1 vector, which was constructed by replacing the YFP cDNA in the pEYFP‐C1 vector with the FLAG epitope sequence. The siRNAs targeting human Solo, LARG and VE‐cadherin were purchased from Sigma‐Aldrich. The sequences of siRNAs targeting dog Solo are listed in supplementary material Table S4.

Construction of a shRNA library targeting Rho‐GEFs

To construct a shRNA library targeting each of the human Rho‐GEFs, candidate 19‐nucleotide shRNA sequences were designed according to the method of Ui‐Tei et al. (Ui‐Tei et al., 2004) and subcloned into the pSUPER vector (OligoEngine, Madison, WI) (Brummelkamp et al., 2002). To assess the knockdown efficiency, each shRNA plasmid was transfected into Jurkat cells together with the pNUL reporter plasmid, which encodes a chimeric cDNA composed of luciferase cDNA and the cDNA fragment (360–540 bp) of the target Rho‐GEF gene, as reported previously (Tsuji et al., 2010; Hayashi et al., 2013). The cDNA fragments (360–540 bp) of Rho‐GEFs were obtained by RT‐PCR amplification from total poly(A)+ RNA of human placenta (Clontech) using the primers listed in supplementary material Table S1. Two days after transfection, Jurkat cells were lysed with 90 µl of Bright‐Glo Luciferase Assay System reagent (Promega, Madison, WI). The levels of luciferase expression were estimated by measuring the luminescence intensity of lysates of Rho‐GEF‐targeting shRNA‐ and control shRNA‐transfected cells using an LMax II luminometer (Molecular Devices, Sunnyvale, CA). The intensity of luminescence in Rho‐GEF shRNA‐transfected cells relative to that in control shRNA‐transfected cells was used as a measure of silencing efficiency. Using this reporter assay, one or two effective shRNA constructs targeting each of the 63 human Rho‐GEFs were obtained. The target sequences are listed in supplementary material Table S2. GL2 (5′‐CCGTACGCGGAATACTTCGA‐3′) was used as the control shRNA.

Cyclic stretch experiments

HUVECs were removed from the dish using 0.05% trypsin and 0.05% EDTA and transferred to a silicone chamber coated with 1 µg/ml fibronectin. For MDCKII cells, 0.25% trypsin and 30 µg/ml fibronectin were used. Cells were subjected to cyclic stretch (20% uniaxial stretch) for 1 hour at a frequency of 1 Hz with an automated stretch system (STB‐140, Strex, Osaka, Japan). Cells were fixed and stained with Alexa‐Fluor‐568‐ or Rhodamine‐labeled phalloidin (Life Technologies) to detect F‐actin. Cell orientation was analyzed by fitting each cell outline to an ellipse and measuring the angle (θ) of the long axis of the ellipse with respect to the stretch axis. Stress fiber orientation was analyzed by measuring the angle (θ) between stress fibers and the stretch direction. The angles of five brighter stress fibers per cell (more than 60 cells) were measured. The levels of cell and stress fiber reorientating were estimated by constructing a histogram in which cell frequency was plotted against cell orientation angle or calculating the orientation parameter (cos2θ) of cell bodies and stress fibers, as reported previously (Goldyn et al., 2009).

Immunostaining and time‐lapse analysis

Cells were fixed with phosphate‐buffered saline (PBS) containing 4% formaldehyde for 30 minutes and permeabilized with PBS containing 0.05% Triton X‐100 for 5 minutes. After washing with PBS, cells were stained with Alexa‐Fluor‐568‐labeled phalloidin or an anti‐VE‐cadherin antibody. Alexa‐Fluor‐488‐labeled anti‐rabbit IgG antibody was used as the secondary antibody. Fluorescence images were obtained using a fluorescence microscope (DMI 6000B, Leica Microsystems, Wetzlar, Germany) equipped with a PL Apo 63× oil‐immersion objective lens (NA 1.4) and ImageJ image analysis software. Time‐lapse fluorescence images were obtained using a LSM 710 confocal microscope (Carl Zeiss, Jena, Germany) equipped with a PL Apo 63× oil objective lens (NA 1.4) and ImageJ software.

Immunoblotting

Cells were lysed with ice‐cold lysis buffer (50 mM Tris‐HCl pH 7.4, 100 mM NaCl, 1% Nonidet P‐40, 10% glycerol, 1 mM MgCl2, 1 mM Na3VO4, 1 mM NaF, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride and 10 µg/ml leupeptin). To detect the silencing effect of siRNAs, cells were lysed with 1× SDS sample buffer containing 1% SDS. Cell lysates were subjected to immunoblot analysis as described previously (Ohashi et al., 2000).

Pull‐down assays of GEF activity

Active GTP‐bound forms of RhoA, Rac1 and Cdc42 were analyzed by pull‐down assays using GST fusion proteins of the RBD of rhotekin and the PBD of PAK1, as described previously (Nishita et al., 2002). 293T cells were transfected with YFP–Solo (WT or 1058–1519) or co‐transfected with FLAG‐tagged RhoA, Rac1 or Cdc42 and YFP–PLEKHG1 (WT or L258E) and cultured for 24 hours. Cell lysates were then incubated with GST–rhotekin(RBD) or GST–PAK1(PBD) bound to glutathione–Sepharose. The bead pellets were washed and analyzed by immunoblotting using anti‐RhoA or anti‐FLAG antibodies. To analyze the effect of Solo knockdown on stretch‐induced RhoA activation, MDCKII cells were transfected with siRNAs targeting dog Solo, cultured for 24 hours and transferred to a silicone chamber coated with 30 µg/ml fibronectin. After incubation for 24 hours, cells were subjected to cyclic stretch for 1 minute at 1 Hz, and cell lysates were subjected to pull‐down assays using GST–rhotekin(RBD) for measuring the RhoA‐GEF activity.

Force application

Recombinant human E‐cadherin–Fc chimera protein (R&D systems, Minneapolis, MN) was bound to the Protein‐G‐coated magnetic beads (280 nm, Dynabeads Protein G, Life Technologies) by incubating at a ratio of 1 µg of protein per 1 mg of beads in 10 mM HEPES (pH 7.2), 50 mM NaCl, 1 mM CaCl2, 0.1% BSA for 4 hours at 4°C, then washed four times to remove unbound proteins and resuspended in 10 mM HEPES (pH 7.2), 50 mM NaCl, 0.1% BSA. MDCKII cells stably expressing CFP–RhoA were cultured subconfluently in 35‐mm dishes and the E‐cadherin‐coated magnetic beads were put on the cells and incubated for 40 minutes. Force application was conducted as reported previously (Guilluy et al., 2011). A ceramic permanent magnet (DynaMag‐2, Veritas, Tokyo, Japan) was used to generate perpendicular tensile forces on beads attached to the dorsal surface of cells. The manufacturer's value for the magnetization is an average of 4000 gauss. For all experiments, the pole face was parallel with and 1 cm from the culture dish surface.

Statistical analysis

Statistical data are expressed as means±s.d. of more than three independent experiments. Statistical analyses were carried out by using Prism 4 (GraphPad Software, La Jolla, CA). The P‐values were calculated using one‐way ANOVA followed by Tukey's test or Dunnett's test for multiple data set comparisons. In all cases, P<0.05 was considered statistically significant.

Acknowledgements

We thank T. Tsuji, M. Naotsuka and K. Kitatani for their help in constructing a Rho‐GEF shRNA library, and H. Kondo for technical assistance.

Author contributions

H.A., S.F., R.H. and T.M. performed the experiments and analyzed the data. N.S. and M.S. provided expertise. K.O. designed the study, performed the experiments and analyzed the data. K.M. designed the study, discussed the data and wrote the manuscript.

Funding

This work was supported by Grants‐in‐Aid for Scientific Research from the Ministry of Education, Culture, Science, Sports and Technology of Japan [grant numbers 23112005 and 25440076 to K.O. and 24370051 and 24121702 to K.M.]; and the Mitsubishi Foundation (to K.M.).

References

Birukov
KG
2009
Small GTPases in mechanosensitive regulation of endothelial barrier.
Microvasc. Res.
77
46
52
Brummelkamp
TR
,
Bernards
R
,
Agami
R
2002
A system for stable expression of short interfering RNAs in mammalian cells.
Science
296
550
553
Cook
DR
,
Rossman
KL
,
Der
CJ
2014
Rho guanine nucleotide exchange factors: regulators of Rho GTPase activity in development and disease.
Oncogene
33
4021
4035
Curtis
C
,
Hemmeryckx
B
,
Haataja
L
,
Senadheera
D
,
Groffen
J
,
Heisterkamp
N
2004
Scambio, a novel guanine nucleotide exchange factor for Rho.
Mol. Cancer
3
10
Daggett
DF
,
Boyd
CA
,
Gautier
P
,
Bryson-Richardson
RJ
,
Thisse
C
,
Thisse
B
,
Amacher
SL
,
Currie
PD
2004
Developmentally restricted actin-regulatory molecules control morphogenetic cell movements in the zebrafish gastrula.
Curr. Biol.
14
1632
1638
Dartsch
PC
,
Betz
E
1989
Response of cultured endothelial cells to mechanical stimulation.
Basic Res. Cardiol.
84
268
281
del Rio
A
,
Perez-Jimenez
R
,
Liu
R
,
Roca-Cusachs
P
,
Fernandez
JM
,
Sheetz
MP
2009
Stretching single talin rod molecules activates vinculin binding.
Science
323
638
641
DiPaolo
BC
,
Davidovich
N
,
Kazanietz
MG
,
Margulies
SS
2013
Rac1 pathway mediates stretch response in pulmonary alveolar epithelial cells.
Am. J. Physiol.
305
L141
L153
DuFort
CC
,
Paszek
MJ
,
Weaver
VM
2011
Balancing forces: architectural control of mechanotransduction.
Nat. Rev. Mol. Cell Biol.
12
308
319
Ehrlicher
AJ
,
Nakamura
F
,
Hartwig
JH
,
Weitz
DA
,
Stossel
TP
2011
Mechanical strain in actin networks regulates FilGAP and integrin binding to filamin A. Nature
478
260
263
Etienne-Manneville
S
,
Hall
A
2002
Rho GTPases in cell biology.
Nature
420
629
635
Franceschini
N
,
Fox
E
,
Zhang
Z
,
Edwards
TL
,
Nalls
MA
,
Sung
YJ
,
Tayo
BO
,
Sun
YV
,
Gottesman
O
,
Adeyemo
A
et al. 
Asian Genetic Epidemiology Network Consortium
2013
Genome-wide association analysis of blood-pressure traits in African-ancestry individuals reveals common associated genes in African and non-African populations.
Am. J. Hum. Genet.
93
545
554
Geiger
B
,
Spatz
JP
,
Bershadsky
AD
2009
Environmental sensing through focal adhesions.
Nat. Rev. Mol. Cell Biol.
10
21
33
Goldyn
AM
,
Rioja
BA
,
Spatz
JP
,
Ballestrem
C
,
Kemkemer
R
2009
Force-induced cell polarisation is linked to RhoA-driven microtubule-independent focal-adhesion sliding.
J. Cell Sci.
122
3644
3651
Grashoff
C
,
Hoffman
BD
,
Brenner
MD
,
Zhou
R
,
Parsons
M
,
Yang
MT
,
McLean
MA
,
Sligar
SG
,
Chen
CS
,
Ha
T
et al. 
2010
Measuring mechanical tension across vinculin reveals regulation of focal adhesion dynamics.
Nature
466
263
266
Guilluy
C
,
Swaminathan
V
,
Garcia-Mata
R
,
O'Brien
ET
,
Superfine
R
,
Burridge
K
2011
The Rho GEFs LARG and GEF-H1 regulate the mechanical response to force on integrins.
Nat. Cell Biol.
13
722
727
Hahn
C
,
Schwartz
MA
2009
Mechanotransduction in vascular physiology and atherogenesis.
Nat. Rev. Mol. Cell Biol.
10
53
62
Hayashi
A
,
Hiatari
R
,
Tsuji
T
,
Ohashi
K
,
Mizuno
K
2013
p63RhoGEF-mediated formation of a single polarized lamellipodium is required for chemotactic migration in breast carcinoma cells.
FEBS Lett.
587
698
705
Huveneers
S
,
de Rooij
J
2013
Mechanosensitive systems at the cadherin-F-actin interface.
J. Cell Sci.
126
403
413
Kaunas
R
,
Nguyen
P
,
Usami
S
,
Chien
S
2005
Cooperative effects of Rho and mechanical stretch on stress fiber organization.
Proc. Natl. Acad. Sci. USA
102
15895
15900
Lecuit
T
,
Lenne
P-F
,
Munro
E
2011
Force generation, transmission, and integration during cell and tissue morphogenesis.
Annu. Rev. Cell Dev. Biol.
27
157
184
Lee
CF
,
Haase
C
,
Deguchi
S
,
Kaunas
R
2010
Cyclic stretch-induced stress fiber dynamics – dependence on strain rate, Rho-kinase and MLCK.
Biochem. Biophys. Res. Commun.
401
344
349
Lessey
EC
,
Guilluy
C
,
Burridge
K
2012
From mechanical force to RhoA activation.
Biochemistry
51
7420
7432
Lim
Y
,
Lim
ST
,
Tomar
A
,
Gardel
M
,
Bernard-Trifilo
JA
,
Chen
XL
,
Uryu
SA
,
Canete-Soler
R
,
Zhai
J
,
Lin
H
et al. 
2008
PyK2 and FAK connections to p190Rho guanine nucleotide exchange factor regulate RhoA activity, focal adhesion formation, and cell motility.
J. Cell Biol.
180
187
203
Liu
WF
,
Nelson
CM
,
Tan
JL
,
Chen
CS
2007
Cadherins, RhoA, and Rac1 are differentially required for stretch-mediated proliferation in endothelial versus smooth muscle cells.
Circ. Res.
101
e44
e52
Matsushita
T
,
Ashikawa
K
,
Yonemoto
K
,
Hirakawa
Y
,
Hata
J
,
Amitani
H
,
Doi
Y
,
Ninomiya
T
,
Kitazono
T
,
Ibayashi
S
et al. 
2010
Functional SNP of ARHGEF10 confers risk of atherothrombotic stroke.
Hum. Mol. Genet.
19
1137
1146
Naruse
K
,
Yamada
T
,
Sai
XR
,
Hamaguchi
M
,
Sokabe
M
1998
Pp125FAK is required for stretch dependent morphological response of endothelial cells.
Oncogene
17
455
463
Nishita
M
,
Aizawa
H
,
Mizuno
K
2002
Stromal cell-derived factor 1α activates LIM kinase 1 and induces cofilin phosphorylation for T-cell chemotaxis.
Mol. Cell. Biol.
22
774
783
Ohashi
K
,
Nagata
K
,
Maekawa
M
,
Ishizaki
T
,
Narumiya
S
,
Mizuno
K
2000
Rho-associated kinase ROCK activates LIM-kinase 1 by phosphorylation at threonine 508 within the activation loop.
J. Biol. Chem.
275
3577
3582
Roca-Cusachs
P
,
Iskratsch
T
,
Sheetz
MP
2012
Finding the weakest link: exploring integrin-mediated mechanical molecular pathways.
J. Cell Sci.
125
3025
3038
Rossman
KL
,
Der
CJ
,
Sondek
J
2005
GEF means go: turning on RHO GTPases with guanine nucleotide-exchange factors.
Nat. Rev. Mol. Cell Biol.
6
167
180
Sawada
Y
,
Tamada
M
,
Dubin-Thaler
BJ
,
Cherniavskaya
O
,
Sakai
R
,
Tanaka
S
,
Sheetz
MP
2006
Force sensing by mechanical extension of the Src family kinase substrate p130Cas.
Cell
127
1015
1026
Schmidt
A
,
Hall
A
2002
Guanine nucleotide exchange factors for Rho GTPases: turning on the switch.
Genes Dev.
16
1587
1609
Schwarz
US
,
Gardel
ML
2012
United we stand: integrating the actin cytoskeleton and cell-matrix adhesions in cellular mechanotransduction.
J. Cell Sci.
125
3051
3060
Tomar
A
,
Schlaepfer
DD
2009
Focal adhesion kinase: switching between GAPs and GEFs in the regulation of cell motility.
Curr. Opin. Cell Biol.
21
676
683
Tse
S-W
,
Broderick
JA
,
Wei
M-L
,
Luo
M-H
,
Smith
D
,
McCaffery
P
,
Stamm
S
,
Andreadis
A
2005
Identification, expression analysis, genomic organization and cellular location of a novel protein with a RhoGEF domain.
Gene
359
63
72
Tsuji
T
,
Ohta
Y
,
Kanno
Y
,
Hirose
K
,
Ohashi
K
,
Mizuno
K
2010
Involvement of p114-RhoGEF and Lfc in Wnt-3a- and dishevelled-induced RhoA activation and neurite retraction in N1E-115 mouse neuroblastoma cells.
Mol. Biol. Cell
21
3590
3600
Tzima
E
2006
Role of small GTPases in endothelial cytoskeletal dynamics and the shear stress response.
Circ. Res.
98
176
185
Tzima
E
,
Irani-Tehrani
M
,
Kiosses
WB
,
Dejana
E
,
Schultz
DA
,
Engelhardt
B
,
Cao
G
,
DeLisser
H
,
Schwartz
MA
2005
A mechanosensory complex that mediates the endothelial cell response to fluid shear stress.
Nature
437
426
431
Ueki
Y
,
Sakamoto
N
,
Sato
M
2010
Cyclic force applied to FAs induces actin recruitment depending on the dynamic loading pattern.
Open Biomed. Eng. J.
4
129
134
Ui-Tei
K
,
Naito
Y
,
Takahashi
F
,
Haraguchi
T
,
Ohki-Hamazaki
H
,
Juni
A
,
Ueda
R
,
Saigo
K
2004
Guidelines for the selection of highly effective siRNA sequences for mammalian and chick RNA interference.
Nucleic Acids Res.
32
936
948
Vallenius
T
2013
Actin stress fibre subtypes in mesenchymal-migrating cells.
Open Biol.
3
130001
Wozniak
MA
,
Chen
CS
2009
Mechanotransduction in development: a growing role for contractility.
Nat. Rev. Mol. Cell Biol.
10
34
43
Yano
Y
,
Saito
Y
,
Narumiya
S
,
Sumpio
BE
1996
Involvement of rho p21 in cyclic strain-induced tyrosine phosphorylation of focal adhesion kinase (pp125FAK), morphological changes and migration of endothelial cells.
Biochem. Biophys. Res. Commun.
224
508
515
Yonemura
S
,
Wada
Y
,
Watanabe
T
,
Nagafuchi
A
,
Shibata
M
2010
alpha-Catenin as a tension transducer that induces adherens junction development.
Nat. Cell Biol.
12
533
542

Competing interests

The authors declare no competing or financial interests.

Supplementary information