EXPERIMENTAL and MOLECULAR MEDICINE, Vol. 36, No. 4, 358-366, August 2004

The role of Rho GTPases in the regulation of the rearrangement of actin cytoskeleton and cell movement

Rokeya Begum1,3, M.S.A. Nur-E-Kamal2 Rho, Rac and Cdc42 GTPases in the regulation and M.A. Zaman1 of the rearrangement of actin cytoskeleton. Our results, for the first time, demonstrate that 1Department of Botany RhoGAP domain of RhoGAP could be used to University of Dhaka study the molecular mechanism of Ras-media- Dhaka 1000, Bangladesh ted signalling in growth, differentiation and car- 2Department of Pharmacology cinogenesis. Robert Wood Johnson Medical School University of Medicine Keywords: actin cytoskeleton; cell motility; Cdc42; Dentistry of New Jersey Rac; RHG; Rho 675 Hoes Lane Piscataway, New Jersey 08854, USA 3Corresponding author: Tel, 88-2-8626885; Introduction Fax, 88-2-8615583; E-mail, [email protected] Cell migration is essential for many normal and ab- normal biological processes including embryogenesis, Accepted 24 July 2004 tissue repair and directing the cells to the sites of infections (Martin et al., 1997). The rearrangement of Abbreviations: DMEM, Dulbecco's modified eagle's medium; the actin cytoskeleton appears to play an important DMSO, dimethyl sulfoxide; EDTA, ethylene-diaminetetra-acetic role in this cell motility (Nobes and Hall, 1999). This acid; FBS, fetal bovine serum; GAP, GTPase activating protein; rearrangement has been shown to be regulated by GDI, guanine nucleotide dissociation inhibitor; GDP, guanine cellular proteins (Ayscough, 1998). The Rho family of diphosphate; GEF, guanine nucleotide exchange factor; G protein, small GTP binding proteins (Rho, Rac and Cdc42) GTP binding protein; GTP, guanine triphosphate; PBS, phosphate has been shown to involved in producing signals that buffer saline; RhoGAP, Rho GTPase-activating protein affect the rearrangement of the actin cytoskeleton (Ridley, 1996; Zigmond, 1996). The constitutively acti- vated Rho, Rac and Cdc42 induce distinct morpho- logical changes of the actin cytoskeleton (Ridley and Abstract Hall, 1992; Nobes and Hall, 1995). The activated The rearrangement of the actin cytoskeleton has Cdc42, Rac and Rho induce the formation of filo- been shown to play a critical role in the devel- podia, lemellipodia and membrane ruffles, focal ad- opment of transformation and malignant pheno- hesion complexes and actin stress fibres, respectively type of cancer cells. Rho family GTPases regu- (Ridley and Hall, 1992; Nobes and Hall, 1995). Des- late the arrangement of the actin cytoskeleton. pite the central role postulated for these Rho GTPases, it is not clear how these small GTPases are By wound-healing assay, we have found that activated or how they regulate the actin assembly NIH 3T3 fibroblast cells move towards the wound- (Teramoto et al., 2003). gaps by extending filopodial and lamellipdial These Rho GTPases are the members of Ras structures at the leading edge of the moving superfamily. Ras, a small G-protein, plays the pivotal cells. We have inactivated the function of Rho role in the control of transducing signals of normal GTPases of v-Ras transformed NIH 3T3 cells by and transformed cell growth (Lowy and Willumsen, overexpressing Rho GTPase-activating (RhoGAP) 1993). Rho, Rac and Cdc42 have been shown to be domain of RhoGAP of p190. We have observed involved in transducing Ras signals for various cellular that inactivation of Rho, Rac and Cdc42 GTP- functions including transformation (Khosravi-Far et al., ases by overexpressing RHG causes inhibition 1995; Qiu et al., 1995a, b; 1997; Nur-E-Kamal et al., of: (i) polymerization of actin to form filaments, 1999) but molecular mechanisms by which Rho (ii) formation of lamellipodia, filopodia and stress GTPases mediate these signals remained unknown fibres, (iii) cell motility, (iv) cell spreading and (Ahmed et al., 2004). (v) cell-to-cell adhesions. These results further The Rho, Rac and Cdc42 GTPases are thought to strengthen the current knowledge on the role of be downstream mediators of Ras functions and act RhoGAP inhibits cell movement 359 as molecular switches between GTP- (active) and and then inserted into the EcoRI/HindIII sites of pMV7 GDP-bound (inactive) forms (Hall, 1990). The state at mammalian expression vector (7 kb) which carried a which G-protein will remain is determined by the neomycin resistant selectable marker. This resultant GTPase activating proteins (GAPs), guanine nucleo- plasmid was named RHG, purified by Cesium Chlo- tide exchange factor (GEF) and guanine nucleotide ride Density Gradients centrifugation and used for dissociation inhibitor (GDI) (Boguski and McCormick, subsequent transfection experiment. 1993). The GAPs stimulate the rate of hydrolysis of the bound GTP to the GDP state. In mammalian cells, Transfection of v-Ha-Ras transformed there are GAPs that are specific for Rho GTPases. NIH 3T3 cells with RHG plasmid One of them is p190. It is a tyrosine phosphorylated G protein of 190,000 dalton molecular weight. The The oncogenic v-Ha-Ras transformed NIH 3T3 cells 5 p190 contains a GAP domain at COOH terminus that (5×10 cells/ml) were cultured under standard condi- stimulates the intrinsic GTPase activity of Cdc42, Rac tion in RPMI and 10% heat-inactivated FBS medium and Rho (Settleman et al., 1992a, b; Foster et al., in six-well-plates. When the cells became 50-60% 1994). In the case of Ras-related GTP binding confluent, the transfection was made by the cons- proteins, the rate of hydrolysis is accelerated by the tructed RHG plasmid using lipofectin reagent (GIBCO GAPs. Therefore, it might be possible to attenuate the BRL kit) following the procedure as per instruction of function of a particular G protein by overexpressing the distributor. These cells were incubated overnight o its specific GAP or GAP domain in the cell. at 37 C in a 5% CO2 incubator. The medium was Keeping all these in view, we have investigated the replaced by the standard cell growth medium and o role of Rho, Rac and Cdc42 in transducing Ras incubated for 2-3 days at 37 C to grow. This growth signals for cell growth and transformation by using the medium was then replaced by RPMI, 10% FBS and p190 gene fragment encoding Rho GAP domain 400 µg/ml G418 (a neomycin analogue medium) and (RHG). We are providing here the evidences to show allowed the cells to grow for three weeks. The resul- that overexpression of RHG inhibits Ras signals requ- tant neomycin resistant colonies (transfectants) were ired for transformation, actin polymerisation, growth- cloned and re-cultured separately in RPMI supple- cone development, formation of filopodia, lamellipodia mented with 10% FBS and G418 (400 µg/ml) medi- and actin stress fibres. Our results demonstrate that um for further investigations. Rho GTPases are involved in transducing Ras signals for different function of the cell. Wound-healing assays and quantitative estimation of cell migration of NIH 3T3 and RHG over- expressing v-Ha-Ras transformed NIH 3T3 cells Materials and Methods For wound-healing assay and immunostaining, 5× 5 Cell culture 10 cells/ml of NIH 3T3 cells were seeded in six-well- culture-plates containing cover slips in standard RPMI The NIH 3T3 and v-Ha-Ras cells were cultured in growth medium. When the cells became 60% conflu- RPMI-1650 (Sigma) growth medium with 10% FBS ent, wounds were made (350-400 µm) across the (GIBCO-BRL) and the RHG 2'C and RHG 2'E trans- monolayer cells using 10 µl micropipette tips carefully fectants in RPMI medium containing 10% FBS and so that the surrounding cells might not be disturbed. G418 (400 µg/ml, Invitrogen). The cells were incuba- The medium was replaced very carefully by fresh one o o ted at 37 C in an environment containing 5% CO2 and and incubated at 37 C in a CO2 incubator and moni- 95% atmospheric air until the cells became confluent. tored to examine the migratory behavior of the cells. Quantitative measurements were made at 0 h, 6 h, 12 h, 24 h, 36 h and 48 h by measuring the dis- Construction of plasmid tances between the wound-edges (by averaging the Using the rat p190 cDNA as a template, we have width of four randomly chosen regions of single amplified an EcoRI/HindIII DNA fragment that contains wound) and the percentage of the closure were calcu- a Kozak consensus sequence (GCCGCCACCATG) at lated as described previously (Ahmed et al., 2004). the 5' end and encodes the COOH-terminus Rho The RHG 2'C and RHG 2'E transfectants were GAP domain (residues 1186 to 1513) of p190 by cultured in RPMI medium containing 10% FBS and polymerase chain reaction (PCR). The pair of primers G418 (400 µg/ml) on sterile coverslips in six-well- was 5' GAATTCGCCGCCACCATGGGGCGG 3' and plates. After they reached 50% confluency, for wound- 3' AAGCTTTTAAGAAGACAA 5'. This amplified DNA healing assay, the same procedure mentioned in NIH product was purified using Sephaglos band prepation 3T3 cells was followed. Then like NIH 3T3 cells, the kit (Pharmacia) as per instruction of the distributor actin cytoskeleton and migratory behaviour of these 360 Exp. Mol. Med. Vol. 36(4), 358-366, 2004 stable cell lines were monitored. The wound closure into the wound by extending finger-like structures measurements were taken at 0 h and 24 h later. The resulting in the closure of the gap (Figure 1B-G). wound-healing ability of each transfectant (RHG 2'C Quantitative analyses showed that the percentage and 2'E) was determined by measuring the distances of wound closure at 6 h, 12 h, 24 h, 36 h and 48 between the wound edges as procedure described h were 22.65, 39.69, 63.36, 80.6 and 94.2 respec- earlier for NIH 3T3 cells. Photographs were taken tively. under phase contrast microscope using 5X objective We have also observed that the leading edge cells and 5X eyepiece. first get attached to the substratum, and then extend finger-like protrusions towards the open space of the wound (Figure 4B, straight arrow), followed by forma- Immunostaining of actin filament of NIH 3T3 and tion of membranous sheet-like structure (Figure 4B, RHG overexpressing v-Ha-Ras transformed NIH bent arrow) and membrane ruffles. Front row cells of 3T3 cells the wound develop a polarised morphology by local- The cells were washed twice, 3 min each in PBS ising the finger-like protrusions and membranous at room temperature and fixed for 20 min in 10% sheet-like structure at the direction of the migration buffered formalin. The cells were washed three times and not at the rear side, where they had contact with with PBS and permeabilised by treating with 0.5% the neighbouring cells (Figure 4B). Triton X-100 (freshly prepared) for 10 min and washed in PBS three times. The cells were treated with rhodamine-conjugated-phalloidin (for NIH 3T3 Rho GTPases regulate the rearrangement cells 1:100 and for RHG transfectants 1:200 dilution of the actin cytoskeleton in NIH 3T3 cells in PBS) for 30 min in a humid chamber. Coverslips The role of Rho, Rac and Cdc42 in the regulation were washed thoroughly in PBS, 3 times, 5 min each of actin cytoskeleton of NIH 3T3 cells was studied and finally mounted in Immunofluoro mounting media. by rhodamine-conjugated-phalloidin staining. It was The cells were visualised under a Carl Zeiss fluor- observed that the new cells were migrating by escent microscope (Zeiss, Axiovert 405 M) equipped extending finger-like projection consisting of thin with a 63X phase 3 objectives. The images were parallel bundles of actin filaments, filopodia (Figure captured with Kodak Pro 400 film using a Nikon 4B, arrow). This was followed by membranous lamell- camera. ipodia (Figure 4B, bent arrow) and membrane ruffles. The lamellipodia consist of linked actin filaments. Afterwards, most conspicuous stress fibres, bundle of Results actin filaments that traversed from one end of migratory cells to another, appeared to insert into the Rho GTPases regulate the migration of focal complexes at the periphery (Figure 4B, arrow NIH 3T3 cells head). These migratory NIH 3T3 cells at the leading- Under the phase contrast microscope, we observed edge display very bright stress fibres than those of that the NIH 3T3 cells formed a thin monolayer in old ones (Figure 4B). This finding indicates that the the liquid culture and did not form foci by piling up Rho GTPase was activated in response to wounding. cells on top of each other. Therefore, these cells are These actin rich filopodial and lamellipodial structures well-adherent in nature. Morphologically these cells were fewer in older cells compared to the migrating were flat, larger in size, having non-fractile appear- cells (Figure 4B). These data indicate that wound- ance and cell-to-cell attachment (Figure 1A and G). healing process causes activation of Cdc42, Rho and As the migratory cells show clear morphological Rac GTPases resulting in the polymerization of actin structure, the wound-healing assay was performed. structures. This cause induction of more filopodia, Photograph was taken immediately after wounding lamellipodia and stress fibres in these migratory cells (Figure 1A). In this assay, it was observed that the compared to older ones (Figure 4B). These data also new cells started to migrate towards the wound within indicate that in response to wound, the NIH 3T3 cells 6 h (Figure 1B). This migration of cells became clear used the focal complexes, stress fibres, filopodia, after 12 h of wounding (Figure 1C). The migration of lamellipodia and membrane ruffles at the leading new cells became more clear gradually after 24 h edge of the wound and thereby healed it. (Figure 1D) and 36 h (Figure 1E). After 48 h (Figure 1F) nearly entire open space of the wound was filled Effect of RHG on the migration of up with new cells. The new cells were more flat than v-Ha-Ras transformed NIH 3T3 cells the old ones and retained close contact with each p190 has been demonstrated to exhibit GTPase acti- other (Figure 1D-G). Thus, the NIH 3T3 cells re- vating protein (GAP) for Rho family GTPases. Bio- sponded to wounding inducing migration of new cells chemical properties of this protein have recently been RhoGAP inhibits cell movement 361

16.7 µM

Figure 1. The NIH 3T3 cell behaviour at the edge of the wound during movement. NIH 3T3 cells were cultured till 60% confluency. The wound was made across the monolayer cells (dashed line across wound space is 393 mm) and photographed under phase contrast microscope (A). The cells were then incubated and the new migrating cell behaviour was monitored and images taken after 6 h (B), after 12 h (C), after 24 h (D), after 36 h (E) and after 48 h (F). Bar, 133 mm. The distances between the wound-edges were measured and percentages of closures were calculated. High power image of leading edge migrating cells toward the wound extending filopodia protrusions (Arrow) (G). Photographs provided are typical of eight such experiments. 362 Exp. Mol. Med. Vol. 36(4), 358-366, 2004

18.5 µM

Figure 2. Cell behaviour during movement of RHG overexpressing (2'C) v-Ha-Ras transformed NIH 3T3 cells. The RHG 2'C overexpressor was cultured till 50% confluency. A wound was made across the monolayer cells with the help of a micropipette tip and photographed under phase contrast microscope, then incubated and the behaviour of new migrating cells monitored and the images photographed. The RHG 2'C cells before wounding (A) and after wounding (B). Migrating cells (arrow) are crawling towards the wound (C). New cells (arrow) are respreading in the open space (D). The wound-healing ability was determined by measuring the distances between wound-edges. Photographs provided are typical of eight such experiments. characterized and GAP domain (RHG) was found to the cells at the wound-edge were attached to the be localized in the NH2 terminal end (Settleman et substratum and then they started respreading out al., 1992a, b). We have previously demonstrated that (Figure 2D, arrow). But the degree of spreading, overexpression of RHG domain inhibits v-Ras- however, was less than that of NIH 3T3 cells (Figure induced transformation (Wang et al., 1997). To study 4A and B), but more than that of v-Ha-Ras trans- the molecular mechanisms of Rho, Rac and Cdc42 formed NIH 3T3 cells (cf. Nur-e-Kamal et al., 1999). mediated Ras signaling, we have inactivated the Rho The migration of new cells became clearer after 12 GTPases activity by transfecting RHG in v-Ha-Ras h. In wound-healing assay, we have observed that the transformed NIH 3T3 cells. As a result, morpho- migratory transfectant cells were round in shape logically two different stably expressing cell lines were (Figure 2C, arrow and D), their cell-to-cell attachment formed. These were named transfectant RHG 2'C was disrupted and moved towards the wound indiv- (Figure 2A) and RHG 2'E (Figure 3A). These trans- idually (Figure 2B-D). fectants formed a thin monolayer in liquid culture. The monolayer RHG 2'E cells were mostly spher- There was no foci formation or piling up of cells on ical (Figure 3A). In response to wound, cells also top of each other (Figure 2A and 3A). moved bodily towards the empty space (Figure 3B-D, Morphologically the monolayer RHG 2'C stably arrow), as the 2'C transfectants did (Figure 2B, C). expressing cells were somewhat flat, more or less Little spreading was seen in a few 2'E cells (Figure oval shaped (Figure 2A) To see the migratory 3D arrow and 4F). The new migrating cells attached behaviour of transfectant 2'C, a wound was made and to substratum with a few morphologically different photographed immediately under the phase contrast focal complex-like structures (Figure 4E and F). The microscope (Figure 2B). We have observed that in 2 E transfectants started dividing without spreading response to wound-healing, the retracted and rounded (Figure 4E) and filled up empty space of the wound cells of the leading edge moved bodily towards the (Figure 3D). These cells also lost the ability of cell-to- wound (Figure 2C, arrow) within 2 h and started to cell attachment (Figure 3C and D, 4E and F). Each fill up the space without dividing (Figure 2C). Then of them had a rim of cytoplasm around the nucleus RhoGAP inhibits cell movement 363

18.5 µM

Figure 3. Cell behaviour of motile RHG overexpressing 2'E derived from v-Ha-Ras transformant. The RHG 2'E cells were cultured in standard RPMI medium supplemented with G418 (400 µg /ml) till 50% confluency. Wound was made and then incubated and the cell behaviour of new migrating cells monitored. Phase contrast image of the RHG 2'E derived from v-Ha-Ras transformed NIH 3T3 cells (A), cells after wounding (B), cells (arrow) are migrating towards the wound bodily and individually (C) and more closing of the wound by migrating cells (D). The percentage of wound closure was calculated. Photographs provided are typical of eight such experiments.

(Figure 4E and F). For that they reduced in size. refractile appearance (Figure 4C and D) probably due From all these observations it appears that in response to cellular contraction. These cells increased in height, to wound, the migratory cells, at first, moved bodily, may have been the consequence of the lesser then attached themselves to the substratum, moved spreading (Figure 4C and D). The contraction of cell towards the wound-gap without extending any filopod- size was more in RHG 2'E (Figure 3A and 4E-F) than ial or lamellipodial structures and finally spread to that in RHG 2'C (Figure 2A, 4C and D). somewhat. Quantitative analyses of v-Ha-Ras transformed NIH 3T3 cells transfected with RHG (RHG 2'C and 2'E Discussion stable cell lines) showed that the percentage of wound-closure at 24 h was 90. We have studied here the role of Rho, Rac and Cdc42 in the regulation of the rearrangement of actin cytoskeleton in NIH 3T3 cells and the requirements Effect of RHG on the rearrangement of the of these GTPases in transducing Ras signals. In actin cytoskeleton in NIH 3T3 cells order to understand the requirements of Rho, Rac Rhodamine-conjugated-phalloidin staining of the RHG and Cdc42 in transducing Ras signals, we have transfectants derived from v-Ha-Ras transformed NIH cloned the p190 gene fragment encoding Rho GAP 3T3 cells revealed that in contrast to NIH 3T3 cells domain (RHG) in a mammalian expression vector and (Figure 4A and B), these migratory transfectant cells transfected in v-Ha-Ras, an oncogenic mutant of Ras were completely devoid of lamellipodia, filopodia and (V12), with this constructed RHG plasmid. stress fibres (Figure 4C-F). In a few cells, however, By rhodamine-conjugated-phalloidin staining, it was sporadic membrane protrusions were observed towa- found that in response to wound-healing, the NIH 3T3 rds the direction of migration (Figure 4C and D). These cells extended filopodia (Figure 4B straight arrow) cells possessed a smaller number of morphologically followed by lamellipodial protrusions (Figure 4B bent different thick blunt focal complex-like structures arrow). Thus these cells spread out by extending (Figure 4C and D). The migrating RHG 2'C cells were finger-like protrusions and membranous sheet-like reduced in size, displayed brighter stain and had structure and acquired flattened morphology. These 364 Exp. Mol. Med. Vol. 36(4), 358-366, 2004

16.7 µM

Figure 4. Rhodomine-conjugated phalloidin staining showing cytoskeletal rearrangement of motile NIH 3T3 and RHG overexpressing v-Ha-Ras transformed NIH 3T3 cells. Normal NIH 3T3 and RHG overexpressing v-Ha-Ras transformed NIH 3T3 cells were cultured on coverslips till 50% confluency. The wound was made with the help of a sterile micropipette tip, then incubated and monitored under fluorescent microscope. To examine the actin cytoskeleton of migrating new cells, the cells were fixed and stained with Rhodamine-conjugated phalloidin following the procedure as mentioned in 'Materials and Methods', analysed under fluorescent microscope and images were recorded. The monolayer normal NIH 3T3 cells (A), migrating NIH 3T3 cells (B) showing stress fibre (arrow-head), filopodia (arrow), lamellipodial protrusion (bent arrow). RHG 2'C (C and D) and RHG 2'E (E and F) migrating cells almost failed to respread and are devoid of stress fibres, filopodial and lamellipodial protrusions. Photographs provided are typical of eight such experiments. data indicate that Cdc42 and Rac are involved in that in response to wound-healing, Rho, Cdc42 and wound-healing, as well as in cell spreading. Rac showed a high level of activity and were able to In NIH 3T3 cells, the filopodial and lamellipodial polymerise the actin cytoskeleton in NIH 3T3 cells. protrusions were completely absent at the rear side, In contrast to NIH 3T3 cells (Figure 4A and B), where the cell had contact with neighbouring cells rhodamine-conjugated-phalloidin staining of RHG (Figure 4B). It seems that at the leading edge, overexpressing v-Ha-ras transformed NIH 3T3 cells lamellipodial and filopodial protrusions generated (2'C and 2'E), did not show any filopodia, lamellipodia membrane tension that pull the cells forward. Pre- or stress fibres (Figure 4C and F). These indicate that vious works suggested that the activation of Rho, Rac inactivation of Rho, Rac and Cdc42 activity by and Cdc42 induce stress fibres and focal adhesions, overexpressing RHG interfered in the formation of the lamellipodia and membrane ruffles and filopodia actin cytoskeleton. These observations suggest that respectively (Ridley and Hall, 1992; Kozma et al., RHG alone can block the signals of Rho, Rac and 1995; Nobes and Hall, 1995). We have observed here Cdc42 leading to inhibition of signals for polymer- that the filopodial (straight arrow) and the lamellipodial isation of the actin filaments. As a consequence, cells (bent arrow) protrusions and stress fibres (arrow fail to produce filopodia, lamellipodia and stress fibres. head) were more in the migrating cells (Figure 4B) Therefore, our findings strongly support the hypoth- than in the older ones. Therefore, our findings indicate esis that actin cytoskeleton is regulated by Rho, Rac RhoGAP inhibits cell movement 365 and Cdc42 GTPases (Hall, 1998). Our observations in many cellular processes, including the cell migra- also support the hypothesis that RhoA, Rac1 and tion, growth-cone development and actin polymeriza- Cdc42 are involved in transducing Ras signals which tion. It further indicates that in response to the wound, regulate the rearrangement of actin filaments. Cdc42, Rho and Rac showed a high level of activity We have observed that the NIH 3T3 cells move to heal the wound. All our observations strongly sup- as a sheet-like structure (Figure 1D-G) by protruding port the hypothesis that Cdc42, RhoA and Rac1 are filopodial (straight arrow) and lamellipodial (bent arrow) involved in transducing signals for the regulation of protrusions at the leading edge of the migratory cells the rearrangement of actin cytoskeleton which plays towards the direction of the wound-gap (Figure 1G a significant role in cell motility (Nobes and Hall, and 4B). From this observation, it appears that these 1999). Our findings are consistent with the previous protrusions are essential for cell motility, the rearra- observation that Rho GTPases are involved in trans- ngement of actin cytoskeleton plays a significant role ducing Ras signals for transformation (Johnson, 1999). in this motility and this rearrangement is regulated by Over-expressing RHG domain alone is responsible to Rho, Rac and Cdc42 in NIH 3T3 cells. attenuate the RhoA, Rac1 and Cdc42 GTPases signal Quantitative measurement of v-Ha-Ras transformed transducing activity by stimulating the intrinsic GTPase NIH 3T3 cells transfected with RHG (RHG 2'C and activity and thereby inhibits Rho, Rac and Cdc42. 2'E stable cell lines) showed that the percentage of These results, for the first time, demonstrate that wound-closure at 24 h was 90. This result shows that RhoGAP domain of RhoGAP could be used to study RHG is responsible for the partial blocking of the the molecular mechanism of Ras- mediated signaling wound-closure. It also appears that although Rac and in differentiation, growth and carcinogenesis. Cdc42 are required for efficient cell movement (we observed in NIH 3T3 cells), in this assay it was not Acknowledgement essential. p190 has been demonstrated as a tumor suppressor. Like p190, RHG has been demonstrated This work is partly supported by UAE Terry Fox Foun- to possess anti-cancer property most probably by dation and Research Committee, Faculty of Medicine inhibiting the function of Rho family GTPases (Wang and Health Sciences, UAE University. The exper- et al., 1997; Vincet and Settleman, 1999). Recently, iments were carried out in Biochemistry laboratory, p190 has been shown to be associated with adipo- Faculty of Medicine and Health Sciences, UAE genesis and myogenesis by affecting Rho GTPases University. function (Sordella et al., 2003). Consistent with these observations, our results suggest a critical role for p190 in the regulation of Rho GTPases function in References controlling cell migration. Both the transfectants (RHG 2'C and 2'E) formed Ahmed I, Calle Y, Sayed MA, Kamal JM, Rengaswamy P, Manser E, Meiners S, Nur-E-Kamal A. Cdc42-dependent monolayer in liquid culture and no longer formed foci nuclear translocation of non-receptor tyrosine kinase, ACK. (Figure 2A and 3A), which is the characteristics Biochem Biophys Res Commun 2004;314:571-9 features of oncogenecity found in v-Ha-Ras trans- formed NIH 3T3 cells (Nur-e-Kamal et al., 1999). This Ayscough KR. In vitro functions of actin finding protein. Curr Opin Cell Biol 1998;10:102-11 result indicates that contact inhibition of growth is restored in these transfectants. Boguski MS, McCormick F. Proteins regulating Ras and its The migratory NIH 3T3 cells moved as a sheet-like relatives. Nature 1993;366:643-54 structure and had cell-to-cell attachment (Figure Braga VMM, Machesky LM, Hall A, Hotchin NA. The small 1B-G). In contrast, the v-Ha-Ras transformed NIH 3T3 GTPases Rho and Rac are required for the establishment cells transfected with RHG plasmid, moved towards of cadherin-dependent cell-cell contacts. J Cell Biol 1997; the wound singly as an isolated individual (Figure 137:1421-31 2A-D RHG 2'C, Figure 3A-D RHG 2'E). This result Foster R, Hu K-Q, Shaywitz DA, Settleman J. p190 Rho indicates that RHG inhibits the cell-to-cell adhesions. GAP, the major Ras GAP-associated protein binds GTP Recent works suggested that Rho, Rac and Cdc42 directly. Mol Cell Biol 1994;14:7173-81 are required for maintaining cell-to-cell adhesions Hall A. The cellular functions of small GTP-binding proteins. (Braga et al., 1997; Kuroda et al., 1997; Takaishi et Science 1990;249:635-40 al., 1997). Therefore, RHG is responsible for the Hall A. Rho GTPases and the actin cytoskeleton. Science reduction of Rho, Rac and Cdc42 mediated cell-to-cell 1998;279:509-14 adhesion activity. But the molecular mechanism is still unknown and needs to be investigated. Johnson DI. Cdc42: An Essential Rho-Type GTPase con- In the light of all our findings, we conclude that trolling eukaryotic cell polarity. Microbiol Mol Biol Rev 1999;63:54-105 Rho, Rac and Cdc42 GTPases play important roles 366 Exp. Mol. Med. Vol. 36(4), 358-366, 2004

Khosravi-Far R, Solski PA, Clark GJ, Kinch MS, Der CJ. anchorage-independent growth and is necessary for Ras Activation of Rac1, RhoA and mitogen-activated protein transformation. Mol Cell Biol 1997;17:3449-58 kinases is required for Ras transformation. Mol Cell Biol 1995;15:6443-53 Ridley AJ, Hall A. The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress Kozma R, Ahmed S, Best A, Lim L. The Ras-related protein fibers in response to growth factors. Cell 1992;70:389-99 Cdc42Hs and bradykinin promote formation of peripheral actin microspikes and filopodia in Swiss 3T3 fibroblasts. Mol Ridley AJ. Rho: theme and variations. Curr Biol 1996;6: Cell Biol 1995;15:1942-52 1256-64 Kuroda S, Fukata M, Fujii K, Nakamura T, Izawa I, Kaibuchi Settleman J, Albright CF, Foster LC, Weinberg RA. Assoc- K. Regulation of cell-cell adhesion of MDCK cells by Cdc42 iation between GTPase activators for Rho and Ras families. and Rac1 small GTPases. Biochem Biophys Res Commun Nature 1992a;359:153-4 1997;240:430-5. Settleman J, Narasimhan V, Foster LC, Weinberg RA. Lowy DR, Willumsen BM. Function and regulation of Ras. Molecular cloning of cDNAs encoding the GAP-associated Ann Rev Biochem 1993;62:851-91 protein p190: implications for a signaling pathway from ras to the nucleus. Cell 1992b;69:539-49 Martin H, Mendoza A, Rodriguez-Pachon JM, Molina M, Nombela C. Characterization of SKM1, a Saccharomyces Sordella R, Jiang W, Chen GC, Curto M, Settleman J. cerevisiae gene encoding a novel Ste20/PAK-like protein Modulation of Rho GTPase signaling regulates a switch kinase. Mol Microbiol 1997;23:431-44 between adipogenesis and myogenesis. Cell 2003;113: 147-58 Nobes CD, Hall A. Rho, Rac, and Cdc42 GTPases: regu- lators of actin structures cell adhesion and motility. Biochem Takaishi K, Sasaki T, Kotani H, Nishioka H, Takai Y. Soc Trans 1995;23:456-9 Regulation of cell-cell adhesion by Rac and Rho small G Proteins in MDCK cells. J Cell Biol 1997;139:1047-59 Nobes CD, Hall A. Rho GTPases control polarity, protrusion and adhesion during cell movement. J Cell Biol 1999;144: Teramoto H, Malek RL, Behbahani B, Castellone MD, 235-44 Leeand NH, Gutlind S. Identification of H-Ras, RhoA, Rac1 and Cdc42 responsive genes. Oncogene 2003;22:2689-97 Nur-E-Kamal MSA, Kamal JM, Qureshi MM, Maruta H. The Cdc42-specific inhibitor derived from ACK-1 blocks v-Ha- Vincent S, Settleman J. Inhibition of RhoGAP activity is Ras-induce transformation. Oncogene 1999;18:7787-93 sufficient for the induction of Rho-mediated actin reorgani- zation. Eur J Cell Biol 1999;78:539-48 Qiu RG, Chen J, Kirn D, McCormick F, Symons M. An essential role for Rac in Ras transformation. Nature 1995a; Wang DZ, Nur-E-Kamal MS, Tikoo A, Montague W, Maruta 374:457-9 H. The GTPase and Rho GAP domains of p190, a tumor suppressor protein that binds the M(r) 120,000 Ras GAP, Qiu RG, Chen J, McCormick F, Symons M. A role for Rho independently function as anti-Ras tumor suppressors. Can- in Ras transformation. Proc Natl Acad Sci USA 1995b;92: cer Res 1997;57:2478-84 1781-5 Zigmond SH. Signal transduction and actin filament organ- Qiu RG, Abo A, McCormick F, Symons M. Cdc42 regulates ization. Curr Opin Cell Biol 1996;8:66-73