Trizol reagents (Life Technologies) were used to isolate total RNA from H9c2 cells transiently transfected with the recombinant RhoGDIβ plasmids or from cells stably expressing RhoGDIβ. Total RNA (20 μg) was separated on a formaldehyde agarose gel, transferred to a nylon filter, and then hybridized with a probe corresponding to the full-length Rac1 cDNA labeled using the NEBlot random labeling kit (New England BioLabs) in the presence of [α-³²P]dCTP. The blot was washed with SSC/SDS solutions (Sodium Chloride, Sodium Citrate/SDS) before autoradiography. Ethidium bromide staining was used to check the integrity of all samples.

H9c2 cells seeded on 10-cm plates were cultured to confluency. They were then scratched with a 200 μl pipette tip and further incubated in DMEM supplemented with 10% FBS. Images were taken at 24, 48, and 72 h with a Zeiss Axiovert 200 microscope. The Image-Pro image analysis system was used to measure the lesion area. The data were expressed as the percentage of recovery (WC%) using the equation: WC% = [1 - (wounded area at T_t/wounded area at T₀)] × 100%, where T_t is the number of hours post-injury and T₀ is the time of injury.

H9c2 cells were cultured with 1 μg/ml doxycycline for 48 h and then treated with lysis buffer (20 mM Tris-HCl, pH 7.5, 100 mM NaCl, 5 mM EDTA, 2 mM PMSF, 1× protease inhibitor) at 4°C for 30 min. The samples were centrifuged at 500 × g at 4°C for 10 min, and the pellets were dissolved in lysis buffer plus 0.1% (w/v) Triton X-100 for the membrane fractions. The supernatants were re-centrifuged at 15,000 rpm at 4°C for 20 min, and the supernatants were saved as cytosolic fractions.

The RhoGTPases act as molecular switches by cycling between the inactive GDP-bound form located in the cytoplasm and an active membrane-associated GTP-bound form. The activities of Rho family proteins are regulated by various proteins, such as guanine nucleotide exchange factors (GEFs), GTPase activating proteins (GAPs), and GDIs. The functions and binding of RhoGDIα to RhoA, Rac1, and Cdc42 are well studied; however, the functions and targets for RhoGDIβ remain unclear, as it binds poorly to RhoA, Rac1, and Cdc42. We therefore sought to determine whether RhoGDIβ stimulates the expression or activities of these GTPases in cardiac cells by western blotting. The total RhoA and Cdc42 levels remained the same; however, cells overexpressing RhoGDIβ had increased levels of Rac1 (Figure 1A). Moreover, the expression of RhoGDIα in H9c2 cells was not affected by the overespression RhoGDIβ (Figure 1A). The expression levels of Rac1 may be causally linked to RhoGDIβ expression or merely an epiphenomenon of the selection of a stable clone. If the former is the case, then Rac1 might be a functionally important downstream target of RhoGDIβ. Northern blot analysis of total RNA indicated that cardiac cells either stably or transiently overexpressing RhoGDIβ had increased Rac1 mRNA levels (Figure 1B). We therefore concluded that RhoGDIβ plays a role in the transcriptional regulation of Rac1 and that the increased level of Rac1 mRNA was not a secondary effect of the selection of a stable RhoGDIβ-expressing clone. Rac1 association with membranes reflects its biological activity 19. To further address the question of whether induction of Rac1 may also influence Rac1 activity, a detergent-insoluble membrane fraction was prepared from RhoGDIβ-overexpressing cells, and the levels of membrane-associated Rac1 were determined by western blotting. Both the levels of membrane-associated Rac1 and cytosolic Rac1 increased in RhoGDIβ-overexpressing cells (Figure 1C). Furthermore, the membrane-associated and cytosolic forms of Cdc42 remained unchanged in RhoGDIβ-expressing cells when compared to parental cells. We further detect the amount of GTP-bound Rac1 in H9c2 RhoGDIβ-overexpressing cells. We found that RhoGDIβ increased GTP loading in Rac1 (Figure 1D). Therefore, overexpression of RhoGDIβ in H9c2 cells increases the level of membrane-associated Rac1 and GTP loading in RAC1 by upregulating Rac1 transcripts. However, the increase in membrane-associated Rac1 in RhoGDIβ-expressing cells may be a secondary effect of increased expression of Rac1, as we were not able to detect any physical interaction between RhoGDIβ and Rac1 by co-immunoprecipitation.

We next examined whether Rac1 mediates RhoGDIβ-induced hypertrophic growth. We found that H9c2 cells stably expressing a dominant-negative form of Rac1 (Rac1N17) and RhoGDIβ significantly reduced the twofold increase in cell size (Figure 2A) and actin organization induced by RhoGDIβ. Moreover, overexpression of wild-type or a constitutively active (V12) Rac1 in H9c2 cells was sufficient to induce hypertrophic growth (Figure 2A) and actin organization. These findings indicate that the RhoGDIβ-induced hypertrophic growth in H9c2 cells is mediated through increased expression of Rac1, and probably through increased levels of membrane-associated Rac1.

Rac1 is able to regulate cell migration 20. To test whether the effect of RhoGDIβ on cell migration is Rac1 dependent, confluent monolayers of cells stably expressing RhoGDIβ, RhoGDIβ and Rac1, or RhoGDIβ and Rac1N17 were scrape-wounded with a sterile plastic pipette, and the migration of cells into the wound was monitored. The RhoGDIβ-expressing cells closed the wound area faster than control cells (Figure 2B). RhoGDIβ-expressing cells that also expressed a dominant-negative form of Rac1N17 migrated slower than RhoGDIβ-expressing or RhoGDIβ- and Rac1-expressing cells (Figure 2B), suggesting that Rac1 plays a key role in mediating cell migration in RhoGDIβ-expressing cells.

To determine whether Rac1 might play a role in the RhoGDIβ-mediated cell cycle arrest, we examined the growth rate of cells expressing both RhoGDIβ and Rac1N17 and found that it was substantially slower than the growth rate of RhoGDIβ-expressing cells or control cells (Additional file 1). Therefore, the RhoGDIβ-regulated cell arrest was not mediated through increased levels of membrane-bound active Rac1.

Since the cyclin-dependent kinase inhibitors p21^Waf1/Cip1 and p27^Kip1 were expressed in RhoGDIβ-expressing cells at higher levels than they were in control cells, we examined p21^Waf1/Cip1 and p27^Kip1 levels in RhoGDIβ- and Rac1N17-expressing cells to study the effects of this dominant-negative form of Rac1 on the expression of p21 and p27. RacN17 was unable to block the expression of p21^Waf1/Cip1 and p27^Kip1 induced by RhoGDIβ in H9c2 cardiac cells (Additional file 2), suggesting that the RhoGDIβ-induced cell cycle arrest was not mediated through Rac1.

To this point, our results indicated that RhoGDIβ stimulates cell migration through the induction of Rac1 expression due to a proportional increase in the activity of Rac1 in H9c2 cells. To confirm the role of RhoGDIβ in cell migration, two RhoGDIβ knockdown cell lines were used to assess whether RhoGDIβ directly stimulates Rac1 expression to induce cell migration. The specific knockdown of RhoGDIβ using siRNA in H9c2 cells was confirmed by immunoblotting (Fig. 2C). We found that targeted disruption of RhoGDIβ by siRNA effectively blocked expression of Rac1 (Fig. 2C); therefore, RhoGDIβ depletion is associated with Rac1 downregulation. We used migration assays to confirm the role of RhoGDIβ in cell migration. Cells were seeded in an upper chamber of a Transwell on a porous filter, and the migration through the filter pores of H9c2 cells expressing RhoGDIβ and cells expressing both RhoGDIβ and RhoGDIβ- specific siRNA was compared. RhoGDIβ-expressing cells showed increased migration compared to parental cells, whereas migration was inhibited in the siRNA RhoGDIβ knockdown cells relative to the RhoGDIβ-expressing cells (Additional file 3). These results suggest that RhoGDIβ may play a critical role in the regulation of Rac1 expression and H9c2 cell migration.

Our experiments demonstrated that RhoGDIβ increased the rate of wound closure. However, our data also suggested that the physical association of ZAK with RhoGDIβ and phosphorylation of RhoGDIβ by ZAK could abolish RhoGDIβ function. To identify the regulatory role of ZAK on cell migration responses in RhoGDIβ-expressing cells, a wound-healing assay was performed using cells expressing RhoGDIβ, ZAK and RhoGDIβ, or ZAKdn and RhoGDIβ. After wounding, control, ZAK-expressing, and ZAKdn-expressing cells closed the gap at a similar rate (Figure 3A). As demonstrated above, RhoGDIβ-expressing cells were highly migratory, but ZAK- and RhoGDIβ-expressing cells were substantially slower to close the wound than ZAKdn- and RhoGDIβ-expressing cells. At 36 h post-wounding, control, ZAK-expressing, ZAKdn-expressing, and ZAK- and RhoGDIβ-expressing cells closed the gap by about 20%, 24%, 25%, and 22%, respectively (Figure 3B). RhoGDIβ-expressing cells and ZAKdn- and RhoGDIβ-expressing cells closed 50% and 39% of the gap. These data suggested a negative regulatory role for ZAK in RhoGDIβ cell migratory function.

To elucidate the role of RhoGDIβ in controlling the localization of Rac1, we performed confocal microscopy. The majority of Rac1 was present in the perinuclear region with some present at the plasma membrane of the control, ZAK-expressing, and ZAKdn-expressing cells (Additional file 4). More Rac1 was present at both the plasma membrane and the perinuclear region in RhoGDIβ-expressing cells. However, co-expression of ZAK, but not ZAKdn, with RhoGDIβ in H9c2 cardiac cells decreased the amount of Rac1 at the plasma membrane (Additional file 4). These data also suggested that ZAK might decrease the total amount of cellular Rac1, probably due to ZAK binding and phosphorylation of RhoGDIβ. To determine whether ZAK negatively regulates RhoGDIβ through expression of Rac1, especially membrane-associated Rac1, we performed western blotting. Consistent with the above experiment, membrane-associated Rac1 increased in RhoGDIβ-expressing cells, whereas, in ZAK- and RhoGDIβ-expressing cells, the levels of membrane-associated Rac1 decreased. However, co-expression of ZAKdn and RhoGDIβ had no such effect (Figure 4A). In this regard, it was of interest whether RhoGDIβ increases the amount of Rac1 at the plasma membrane as a consequence of increasing the total cellular levels of Rac1 or whether RhoGDIβ facilitates the translocation of Rac1 to the plasma membrane. We attempted to co-immunoprecipitate RhoGDIβ and Rac1to investigate this possibility; however, the antibody used was insufficient for this purpose (data not shown). This suggests that it is unlikely that RhoGDIβ facilitates translocation of Rac1 to the plasma membrane.

To test the importance of ZAK in regulating RhoGDIβ-mediated membrane-associated Rac1 and hypertrophic growth, we reduced the levels of ZAK using siRNA. We were able to reduce the levels of ZAK by expressing two different human ZAK siRNAs: ZAKGRi1 (ZAK-RhoGDIβ U6-460i) and ZAKGRi2 (ZAK-RhoGDIβ U6-1712i) (data not shown). The reduced levels of ZAK in these two individual clones were able to restore the levels of membrane-associated Rac1 to levels similar to those of RhoGDIβ-expressing cells, whereas the introduction of a scrambled ZAK siRNA (GRCi) into ZAK- and RhoGDIβ-expressing cells had no effect (Figure 4B). These results confirmed the importance of ZAK as a negative regulator of the effect of RhoGDIβ on the expression of Rac1 and membrane-associated Rac1. We found that ZAK was able to regulate ANF expression. We then studied the effects of RhoGDIβ on the regulation of ANF expression by ZAK by examining the levels of ANF mRNA in ZAK- and RhoGDIβ-expressing cells when compared with ZAK-expressing cells. The levels of ANF mRNA induced by ZAK were decreased in cells expressing both ZAK and RhoGDIβ (Fig. 4C). We also found that ZAK-RhoGDIβ cells has less total amount of Rac1 than RhoGDIβ cells. (Figure 4D). This result suggested that RhoGDIβ negatively regulates the functions of ZAK. Moreover, the data presented here (and consistent with our previous studies) indicate that both ZAK and RhoGDIβ may be hypertrophic growth inducers; however, ZAK physically interacts with RhoGDIβ and phosphorylates RhoGDIβ, thus inhibiting the ability of RhoGDIβ to induce Rac1 expression. The levels of Rac1 induced by RhoGDIβ are associated with the closure rate of wound healing (Figure 2B) and hypertrophic growth (Figure 2A), but they are not associated with cell cycle inhibition (Additional file 1). Thus, RhoGDIβ appears to play a role in signaling pathways regulating Rac1 expression that govern wound healing and hypertrophic growth in cardiac cells.