June 2006
Volume 47, Issue 6
Free
Cornea  |   June 2006
Cross-talk among Rho GTPases Acting Downstream of PI 3-Kinase Induces Mesenchymal Transformation of Corneal Endothelial Cells Mediated by FGF-2
Author Affiliations
  • Jeong Goo Lee
    From the Doheny Eye Institute and the
  • EunDuck P. Kay
    From the Doheny Eye Institute and the
    Department of Ophthalmology, Keck School of Medicine of the University of Southern California, Los Angeles, California.
Investigative Ophthalmology & Visual Science June 2006, Vol.47, 2358-2368. doi:https://doi.org/10.1167/iovs.05-1490
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Jeong Goo Lee, EunDuck P. Kay; Cross-talk among Rho GTPases Acting Downstream of PI 3-Kinase Induces Mesenchymal Transformation of Corneal Endothelial Cells Mediated by FGF-2. Invest. Ophthalmol. Vis. Sci. 2006;47(6):2358-2368. https://doi.org/10.1167/iovs.05-1490.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

purpose. Endothelial–mesenchymal transformation (EMT), in which the contact-inhibited corneal endothelial cells (CECs) become multilayers of spindle-shaped cells containing protrusive processes, is mediated by fibroblast growth factor (FGF)-2. The involvement in EMT of cross-talk among Rho GTPases mediated by FGF-2 was also investigated.

methods. GTP pull-down assays were performed to identify the activated Rho GTPases. Transfection of CECs with either constitutively active (ca) or dominant negative (dn) Rho GTPase vectors was performed. Protein–protein interaction was investigated by coimmunoprecipitation and a yeast two-hybrid assay.

results. The alteration of morphology and actin cytoskeleton caused by FGF-2 was mediated by active Rac and inactive Rho. Prolonged treatment of CECs with FGF-2 induced formation of protrusive processes through activated Cdc42. All FGF-2 actions were blocked by the phosphatidylinositol (PI) 3-kinase inhibitor LY294002. Cells transfected with caRacG12V acquired elongated morphology; the actin cytoskeleton was reorganized to the cortex. Formation of protrusive processes was observed in the elongated cells expressing caCdc42G12V or dominant negative (dn)RhoT19N, whereas polygonal cells expressing dnRacT17N, caRhoG14V, or dnCdc42T17N had stress fibers. Further analysis demonstrated that Rac was associated with Cdc42 or Rho through a 32- or 30-kDa Dbl homology/pleckstrin homology–containing protein.

conclusions. These findings suggest that alteration of cell shape and actin cytoskeleton are closely linked to the sequential activation of Rho GTPases through PI 3-kinase in response to FGF-2 stimulation. Cortical actin is formed via active Rac and inactive Rho followed by formation of protrusive processes mediated by active Cdc42 and inactive Rho.

Corneal fibrosis represents a significant pathophysiological problem—one that causes blindness by physically blocking light transmittance. One clinical example of corneal fibrosis observed in corneal endothelium is the development of a retrocorneal fibrous membrane (RCFM) in Descemet’s membrane. 1 2 In the RCFM, corneal endothelial cells (CECs) are converted to fibroblast-like cells. The contact-inhibited monolayers of CECs are lost, resulting in the development of multilayers of fibroblast-like cells. 3 4 These morphologically altered cells simultaneously resume their proliferation ability and deposit a fibrillar extracellular matrix (ECM) in the basement membrane environment. An in vitro model to elucidate the molecular mechanism of RCFM formation led us to the finding that fibroblast growth factor (FGF)-2 is the direct mediator of the endothelial–mesenchymal transformation (EMT) observed in CECs. 5 6 7 We have reported that, among the phenotypes altered during EMT, FGF-2 directly regulates cell cycle progression through the action of phosphatidylinositol (PI) 3-kinase, and this phenotype alteration leads to a marked stimulation of cell proliferation, 8 9 as opposed to the normal CECs arrested in the G1-phase of the cell cycle throughout their lifespan. We also reported that FGF-2 induces a change in cell shape from a polygonal to a fibroblastic morphology and that it induces a reorganization of actin cytoskeleton via PI 3-kinase. 10 11 Similarly, PI 3-kinase is known to regulate directly the cell cycle progression and morphogenic pathways in other cell systems. 12 13 14 15 16  
It is now clear that the actin cytoskeleton with its polymerization dynamics is central to many aspects of cellular activities, such as cytokinesis, phagocytosis, cell migration, adhesion, polarity, and morphology. 17 18 19 20 21 Assembly and organization of the actin cytoskeleton is controlled by the Rho family of small GTPases, the best characterized of which are Rho, Rac, and Cdc42. 22 23 24 25 In particular, these GTPases have been shown to induce morphologic changes associated with actin polymerization. In Swiss 3T3 fibroblasts, Rac1 induces membrane ruffling and lamellipodium formation, RhoA induces the formation of stress fibers, and Cdc42 induces the formation of microspikes and filopodia, all of which depend on filamentous actin (F-actin) organization. 26 27 28 29 These earlier studies also demonstrated that Rho GTPases are key signal transducers that mediate growth factor-induced changes to the actin cytoskeleton. 26 27  
Numerous studies have demonstrated that cross-talk among the Rho GTPase family exerts a key role in organizing actin cytoskeleton structures. With regard to the cross-talk among these Rho GTPases, there appears to be a generalized consensus on the antagonizing activity between Rac and Rho 30 31 and coincidental activation of Rac and Cdc42. 32 33 Nonetheless, different types of cross-talk, such as the coactivation of Rac and Rho, the activation of Rac by Rho, and the antagonistic cross-talk between Rac and Cdc42 were observed, respectively, in vertebrate gastrulation, in Swiss 3T3 fibroblasts, and during regulation of reactive oxygen species generation. 34 35 36 It is, therefore, crucial to reveal the cross-talk among Rho GTPases and to determine the hierarchy of these proteins during the organization of actin cytoskeleton structures in the course of morphogenesis. 
Our recent investigation using the in vitro system found that there are two key events in the FGF-2-mediated endothelial cell morphogenesis: FGF-2 acts through PI 3-kinase to induce both reorganization of actin cytoskeleton to the cortex and alteration of cell morphology to an elongated shape. The simultaneous treatment of cells with FGF-2 and inhibitors of Rho or Rho-associated kinase (ROCK) leads to the formation of prominent protrusive processes by the FGF-2-modulated cells, through the PI 3-kinase-dependent pathway. 11 The phenotypes are similar to that observed after EMT in vitro and in vivo during wound healing. 4 11 Based on this observation, it is likely that both activation of Cdc42 and complete inactivation of Rho are necessary to generate spindle-shaped cells with prominent protrusive processes. Cdc42 and its downstream effector proteins are known to be involved in actin reorganization, leading to the formation of protrusive processes, 37 38 and Rho and Cdc42 are known to play a role in the transformation of epithelial cells and fibroblasts. 39 40 We therefore investigated whether there was a hierarchy of actin cytoskeleton reorganization in response to FGF-2 stimulation in CECs. Herein, we demonstrate that activation of Rac mediated by FGF-2 and the subsequent inactivation of Rho by active Rac cause a loss of stress fibers and reorganization of actin in the cortex, and that activation of Cdc42, which further antagonizes Rho activity, is necessary for formation of protrusive processes in the FGF-2-treated, elongated endothelial cells. Such acquisition of a protrusive structure is biologically relevant to the phenotypes observed after EMT, as well as to the in vivo wound-healing process. 
Materials and Methods
LY294002 and monoclonal antibodies against β-actin, hemagglutinin (HA), and Rac1 were obtained from Sigma-Aldrich (St. Louis, MO); rhodamine-phalloidin from Invitrogen (Eugene, OR); monoclonal antibody against vinculin and fluorescein isothiocyanate (FITC)–conjugated secondary antibodies from Chemicon (Temecula, CA); FGF-2 and anti-Vav2 Dbl homology (DH)/pleckstrin homology (PH) antibody from Calbiochem (San Diego, CA); and anti-Rho and anti-Cdc42 antibody from Upstate Biotechnology, Inc. (Lake Placid, NY) and Santa Cruz Biotechnology (Santa Cruz, CA), respectively. Mounting solution and biotinylated secondary antibodies were purchased from Vector Laboratories Inc. (Burlingame, CA). Constitutively active (ca) (RhoG14V, RacG12V and Cdc42G12V) and dominant negative (dn) (RhoT19N, RacT17N, and Cdc42T17N) Rho GTPase plasmids were obtained from the UMR cDNA Resource Center (Rolla, MO). 
Cell Cultures
Rabbit eyes were purchased from Pel-Freez Biologicals (Rogers, AR). Isolation and establishment of rabbit CECs were performed as previously described. 41 Briefly, the corneal endothelium-Descemet’s membrane complex was treated with 0.2% collagenase and 0.05% hyaluronidase (Worthington Biochemical, Lakewood, NJ) for 90 minutes at 37°C. Primary cultured cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 15% fetal bovine serum and 50 μg/mL of gentamicin (DMEM-15) in a 5% CO2 incubator. First-passage CECs maintained in DMEM-15 were used for all experiments. Heparin (10 μg/mL) was added to cell cultures treated with FGF-2 (10 ng/mL), because our previous study had shown that FGF-2 activity in CECs requires supplemental heparin. 5 In some experiments, LY294002 at 20 μM was used to block PI 3-kinase; the optimal concentration of the inhibitor had been determined earlier. 11 We used the following plating cell densities throughout the experiments: 1 × 105 cells per 35-mm culture dish and 8 × 105 cells per 100-mm tissue culture dish. 
Transfection of CECs with Rho GTPase Expression Vector
Rabbit CECs were transfected with Rho GTPase eukaryotic expression plasmid in which human ca or dnRho GTPases were expressed as fusion proteins (HA-tagged Rho GTPases) under the control of a CMV promoter. The eukaryotic empty-expression plasmid vector pcDNA3.1 was used for vector-alone mock transfection. CECs (1 × 104/chamber) were seeded on four-well chamber slides and maintained in culture until they reached 60% to 70% confluence. These cells were transiently transfected with 1 μg of one of the Rho GTPase plasmids or pcDNA3.1 (FuGENE 6 transfection reagent; Roche, Pleasanton, CA), according to the manufacturer’s instructions. After the 8-hour incubation, medium containing transfection reagent was removed, and the cells were grown further in DMEM-15, with or without FGF-2; in some experiments, the transfectants were simultaneously treated with LY294002 at 20 μM in the presence of FGF-2. 
Cytochemical Staining and Confocal Microscopy
Cells plated on four-well chamber slides were washed with phosphate-buffered saline (PBS), fixed with 4% paraformaldehyde for 20 minutes, permeabilized with 0.1% Triton X-100, and incubated with PBS containing 2% BSA for 1 hour at room temperature. Primary antibodies diluted in PBS containing 2% BSA were used in a dilution of 1:200 for anti-HA and 1:200 for anti-vinculin and incubated for 1 hour at room temperature. The cells were washed with PBS and incubated with an FITC-conjugated goat anti-mouse IgG (1:200 dilution). F-actin was stained with rhodamine-conjugated phalloidin (1:300). Double labeling of F-actin and focal adhesion or F-actin and HA were examined with a laser scanning confocal microscope (LSM-510; Carl Zeiss Meditec, Inc., Thornwood, NY) with an inverted microscope (100M Axiovert; Carl Zeiss Meditec, Inc.) operating with a 25-mW argon laser tuned to 488 nm and a 1 mW helium-neon laser tuned to 543 nm. Cells were imaged with a 40 × 1.3 NA oil-immersion objective (Zeiss Plan-Neofluar; Carl Zeiss Meditec, Inc.). Images were collected in a single-track configuration in which the red and green emission was split with an NFT 545 dichroic filter and collected in PMT 1 with an LP 560 filter and in PMT 2 with a BP 505-530 filter, respectively. Each microscopic image is representative of 10 fields over a minimum of three experiments, and all images were taken at the equatorial plane of the cell with a single optical section. Image analysis was performed with the standard system operating software provided with the laser scanning microscope (LSM 510; Carl Zeiss Meditec, Inc.). All illustrations were assembled and processed digitally (Photoshop 7; Adobe, San Jose, CA). 
Cross-linking
The procedures used for protein cross-linking have been reported. 42 Dithiobis succinimidyl propionate (DTSP), a substance that has been shown to cross cell membranes, 43 was selected for the study. Cells treated with 0.05% trypsin containing EDTA (5 mM) were washed with DMEM-15 and PBS, and then harvested by centrifugation. The harvested cells were combined with DTSP, vigorously shaken, and placed on ice for 30 minutes. After the reaction, the cells were rinsed with 2 mM glycine in PBS, to block the DTSP activity, and then were lysed with RIPA buffer (50 mM HEPES [pH 7.5] 150 mM NaCl, 10 mM MgCl2, 1% Triton X-100, 0.5% Nonidet P-40, and protease inhibitors). Concentration of lysates was assessed with the Bradford protein assay system (Bio-Rad Laboratories, Hercules, CA). 
Immunodetection of Proteins
Equal amounts of proteins from different treatments were immunoprecipitated, as described previously, 44 and immune complexes were resolved by SDS-PAGE and analyzed by immunoblot analysis, with a commercial kit (ABC Vectastain; Vector Laboratories Inc.) and a chemiluminescence detection kit (ECL; GE Healthcare, Piscataway, NJ). 7 44 The relative density of the polypeptide bands detected on ECL film was estimated (Gel-Doc system; Bio-Rad Laboratories). 
Partial cDNA Cloning of Rabbit RhoA, Rac1, and Cdc42
Total RNA was extracted from CECs (TRIzol reagent; Invitrogen). One microgram of extracted total RNA treated with DNase I was used for cDNA synthesis. The first-strand cDNA was synthesized with a 12 oligo dT adapter primer in a reverse transcription system (Promega, Madison, WI). The conditions used for cDNA synthesis were 1 cycle of 42°C for 15 minutes and 95°C for 5 minutes, followed by incubation at 60°C for 5 minutes. PCR was performed, using the synthesized cDNA library as a template, to amplify partial RhoA (sense, 5′-GCTGCCATCCGGAAGAAACTGGTG-3′; antisense, 5′-CTGCAGGGAAGCCCGGGTGGCCAT-CTC-3′), Rac1 (sense, 5′-CAGGCCATCAAGTGCGTGGTGGTG-3′; antisense 5′-GGGGCACAGGACGGCCCGGATAGC-3′), and Cdc42 (sense, 5′-ATGCAGACAATCAAGTGTGTTGTT-3′; antisense, 5′-GGGAGGCTCCAGGGCAGCCAG-3′) cDNA fragments, resulting in a 570-bp product for RhoA (190 amino acids; GenBank accession No. AY644387; http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD), 534 bp for Rac1 (178 amino acids, GenBank accession no. AY644386), and 540 bp for Cdc42 (180 amino acids, GenBank accession No. AY780571). The PCR protocol was 95°C for 5 minutes, followed by 33 cycles of 95°C for 1 minute, 52°C for 1 minute, and 72°C for 1 minute, with a final incubation at 72°C for 10 minutes. Rho GTPase cDNA fragments amplified by PCR were cloned into a vector (pGEM-T Easy; Promega). 
Yeast Two-Hybrid Assay
To verify protein–protein interaction among three Rho GTPases, a yeast two-hybrid assay was performed using the Saccharomyces cerevisiae L40 strain harboring reporter genes HIS3 and β-galactosidase under the control of upstream LexA-binding sites. PCR was performed to make a bait vector for RhoA (sense, 5′-GAATTCCTGCCATCC-GTAAGAAA-3′; antisense, 5′-CTGCAGACTGCAGGGCAGCCCGGGT-3′), Rac1 (sense, 5′-GAATTCCAGGCCATCAAGTGCGTG-3′; antisense 5′-CTGCAGTTAGGGGCACAGGACGGCCCG-3′), and Cdc42 (sense, 5′-GAATTCCAGACAATCAAGTGTGTT-3′; antisense, 5′-CTGCAGGGGAGGCTCCAGGGCAGC-3′), using rabbit Rho GTPases cDNA fragments as the templates. The PCR products were subcloned in-frame into the EcoRI– PstI site of pBHA (a bait vector containing LexA DNA-binding domain), resulting in the construction of pBHA-RhoA, pBHA-Rac1, and pBHA-Cdc42. Similar subcloning was conducted to construct prey vector for RhoA (sense, 5′-CTCGAGCTGCCATCCGTAAGAAA-3′; antisense, 5′-GAATTCACTGCAGGGCAGCCCGGGT-3′), Rac1 (sense, 5′-CTCGAGAGGCCATCAAGTGCGTG-3′; antisense 5′-GAATTCTTAGGGGCACAGGACGGC-3′), and Cdc42 (sense, 5′-CTCGAGAGACAATCAAGTGTGTT-3′; antisense, 5′-GAATTCGGGAGGCTCCAGGGCAGC-3′), using PCR. The resultant PCR fragments were inserted in-frame into the XhoI–EcoRI site of pGAD10 (a prey vector with the Gal4 activation domain) and named pGAD10-RhoA, pGAD10-Rac1, and pGAD10-Cdc42, respectively. The yeast L40 strain was cotransformed with these bait and prey plasmids. Transformation of the yeast strain was performed using the lithium acetate method. 45 Cotransformants were plated on solid YNB (0.67% yeast nitrogen base without amino acids, 2% dextrose) + adenine + histidine medium (YNB+AH) and grown at 30°C. The colonies were replica printed to X-Gal plates at 30°C and checked daily for blue color for 3 days. Colonies that turned blue were selected on YNB lacking histidine (YNB+A)–selective media and then retested in X-Gal plates, to confirm that they produced β-galactosidase, which is selective for interaction among Rho, Rac, and Cdc42. 
Rho and Rac GTPase Activation Assay
Rho and Rac activity were quantified by measuring the amounts of Rho and Rac precipitated in a pull-down reaction from cell lysates, with the GTPase-binding domain of Rhotekin-RBD and PAK-CRIB used as bait, using Rho (Cytoskeleton, Denver, CO) and Rac (Pierce, Rockford, IL) activation assay kits, respectively, according to the manufacturers’ instructions. Briefly, CECs were cultured in the designated culture condition and then lysed with the lysis buffer supplied by the manufacturer. The entire 9 mg of cleared cell lysates was split into three equal aliquots. As negative and positive controls for the pull-down assay, two of the aliquots were added to 100 μM GDP or GTPγS, respectively, and incubated for 15 minutes at 30°C, with agitation, to deplete or enrich Rho/Rac-GTP. The third aliquot remained untreated and was kept on ice while the controls were loaded. Glutathione S-transferase (GST)-Rhotekin-RBD beads or GST-PAK-CRIB beads were added to each aliquot, and the reaction mixtures were incubated for 45 minutes at 4°C with gentle agitation. After the pull-down reaction, the supernatants were removed by brief centrifugation, and the precipitated proteins bound to the beads were subjected to immunoblot analysis with monoclonal antibody to Rho or to Rac1. The total amount of each small G protein was also determined by Western blot analysis of total cell lysates and used to normalize the protein concentration of each lane. 
Cdc42 GTPase Pull-down Experiment Using Purified GST-PAK-CRIB Domain
To purify GST-PAK-CRIB domain, Escherichia coli BL21 cells were transformed with expression plasmid for GST fusion proteins with PAK-CRIB and grown at 37°C to an absorbance of 0.3. Expression of GST fusion protein was induced by adding 0.1 mM isopropyl-1-thio-β-d-galactopyranoside (final concentration) at an absorbance of 1.0. After induction, the cells were collected and lysed by sonication in bacterial lysis buffer (50 mM Tris-HCl [pH 8.0], 2 mM MgCl2, 2 mM dithiothreitol, 10% glycerol, 20% sucrose, 0.2 mM Na2S2O, 1 μg/mL leupeptin, 1 μg/mL pepstatin, 1 μg/mL aprotinin, and 1 mM phenylmethylsulfonyl fluoride [PMSF]). The lysate was centrifuged at 4°C for 10 minutes at 10,000g, and the supernatant was used to purify the GST-PAK-CRIB domain by affinity purification, using glutathione-Sepharose beads (Sigma-Aldrich). Beads loaded with the GST fusion proteins were washed twice with PBS and used immediately for Cdc42 GTPase pull-down experiments. 
CECs were lysed with ice-cold GST-Fish lysis buffer (50 mM Tris [pH 7.4], 100 mM NaCl, 2 mM MgCl2, 10% glycerol, 1% Nonidet P-40, 1 μg/mL leupeptin, 1 μg/mL pepstatin, 1 μg/mL aprotinin, and 1 mM PMSF). 31 46 The detergent-soluble supernatant was recovered after centrifugation for 15 minutes at 14,000g, and 3 mg of total cell lysates were used for the Cdc42 pull-down assay experiments. GTP-Cdc42 proteins were precipitated for 1 hour with 50 μL of purified GST-PAK fusion protein at 4°C. The complexes were washed three times with ice-cold PBS, resuspended, and boiled with SDS sample buffer. Bound Cdc42 proteins were detected by Western blot, using anti-Cdc42 antibody. Positive and negative controls were as stated earlier. Total Cdc42 was used to normalize the protein concentration of each lane. 
Results
FGF-2-Induced Morphologic Change and Cortical Actin Formation through PI 3-kinase
Our previous study showed that FGF-2 altered the polygonal endothelial cells into elongated cells through the action of PI 3-kinase and that FGF-2 also altered the actin cytoskeleton structure. 11 47 To confirm whether cell shape change caused by FGF-2 was correlated to the reorganization of actin cytoskeleton and focal adhesion, CECs were treated with FGF-2 for either 1 hour or 16 hours, whereas CECs maintained in serum-containing medium (DMEM-15) served as the control. Cells maintained in DMEM-15 showed well-organized stress fibers, oriented radially across the cell, and abundant focal adhesion complexes (Fig. 1A) . Initial disruption of stress fibers and loss of focal adhesion were observed in cells treated with FGF-2 for 1 hour. Of interest, smaller cells demonstrated actin cytoskeleton at the cortex. The prolonged stimulation of cells with FGF-2 for 16 hours further modulated cell shape to an elongated cell morphology, and the cortical actin was strongly organized with a small degree of lamellipodia formation (Fig. 1B , arrowheads). Simultaneously, focal adhesion complexes were almost lost, except where lamellipodia were present. In these cells, it appears that FGF-2 overcomes the RhoA activation, as evidenced by the loss of stress fibers and focal adhesions. These alterations caused by FGF-2 were completely blocked in cells treated with LY294002 (Fig. 1B) . Elongated cells reverted to polygonal cells, in which stress fiber formation and the focal adhesion complex were reestablished. These findings suggest that the loss of stress fibers and focal adhesion and the reorganization of the actin cytoskeleton at the cortex are mediated by PI 3-kinase and that actin cytoskeleton reorganization at the cortex is responsible for CECs acquiring their elongated cell morphology in response to FGF-2 stimulation. 
To further confirm that cortical actin formation in response to FGF-2 stimulation was indeed mediated by Rac activation and Rho inactivation, Rac and Rho GTPase activities were measured. Figure 2Ashows negligible Rac activity in CECs maintained in DMEM-15. In contrast, there was a time-dependent activation of Rac in response to FGF-2 stimulation: after 1 hour of FGF-2 stimulation, Rac activity was increased nearly sixfold compared with the level in control cells (DMEM-15). Maximum Rac activity was observed in cells treated with FGF-2 for 16 hours, after which activity was reduced. LY294002 treatment completely blocked Rac activation by FGF-2. Rac-GTP levels were similar to the basal levels of CECs maintained in DMEM-15 (Fig. 2A)
In contrast, FGF-2 had a negative effect on Rho activity. Figure 2Bshows that CECs maintained in DMEM-15 demonstrated very high levels of RhoA activity and that FGF-2 rapidly and completely blocked Rho activation: the amount of activated Rho in the lysates treated with FGF-2 for 1 hour was decreased to one fourth the level observed in cells maintained in DMEM-15. This inhibitory activity of FGF-2 on Rho activity was markedly reversed by LY294002. Together, these results suggest that FGF-2 facilitates Rac activation, but completely blocks Rho activation, through the action of PI 3-kinase. Figures 1 and 2demonstrate that the cortical actin is organized through active Rac and inactive Rho pathways. Furthermore, these data clearly demonstrate that PI 3-kinase is the upstream regulator for Rac and Rho in response to FGF-2 stimulation. 
Relationship between Morphology and Actin Cytoskeleton Structure Caused by Rac and Rho in CECs
Previous studies have shown that stress fiber formation mediated by lysophosphatidic acid via RhoA activation is inhibited by constitutively active forms of Rac. 31 48 49 Therefore, we investigated whether there is a hierarchy between Rac and Rho, by using transfected cells with caRacG12V, dnRacT17N, caRhoG14V, or dnRhoT19N (Fig. 3) . CECs transfected with caRacG12V lost their characteristic polygonal cell morphology. Stress fibers were mostly disrupted, and actin cytoskeleton was reorganized at the cortex (Fig. 3A) . When these transfectants were treated with FGF-2, cell proliferation was greatly stimulated and cells became round; however, the actin cytoskeleton organized at the cortex was maintained (Fig. 3B) . In contrast, cells transfected with dnRacT17N demonstrated a polygonal cell shape and well-defined stress fibers, confirming that Rac antagonizes Rho activity. When these cells were stimulated with exogenous FGF-2, which is able to activate Rac while inactivating Rho, the FGF-2 was not able to inactivate Rho in these CECs expressing inactive Rac. Prominent stress fibers were maintained in the characteristic polygon-shaped cells (Fig. 3B) . The null activity of FGF-2 in CECs expressing dnRacT17N suggests that Rac is an upstream molecule to RhoA in CECs. 
CECs expressing caRhoG14V have well-organized stress fibers, similar to those observed in control cells or in CECs expressing inactive Rac (Fig. 3A) . Exogenous FGF-2 caused a slight disruption of stress fibers and slightly facilitated lamellipodia formation (Fig. 3B , arrowheads), suggesting that the exogenous FGF-2 is able to activate Rac further, albeit to a low degree. In contrast, a marked disruption of stress fibers and a reorganization of actin cytoskeleton at the cortex were observed in CECs expressing inactive RhoA (dnRhoT19N; Fig. 3A ), suggesting that Rho inactivation is absolutely essential for cortical actin formation. Cell shape was also altered from the characteristic polygonal morphology to fibroblast-like cells containing protrusive processes. Exogenous FGF-2 further modulated cell shape to the elongated fibroblast-like cells with prominent protrusive processes (Fig. 3B) . Similar to the results of previous studies, 31 48 49 our data also demonstrated a hierarchy between the two Rho GTPases. 
We further searched the biochemical evidence for physical interaction between the endogenous Rac and Rho in CECs using coimmunoprecipitation and immunoblot analysis with either anti-Rac1 or anti-Rho antibody. We used a cross-linking method to remove artificial interaction and to promote the in vivo association of the two proteins. Figure 4Ademonstrates that immune complex precipitated with anti-Rac1 antibody contained Rho and that immune complex precipitated with anti-Rho antibody contained Rac1, suggesting that there is a protein–protein interaction between the two endogenous Rho GTPases. However, the degree of such physical interaction between Rac1 and Rho was not altered by FGF-2. 
To determine further whether Rac directly binds to Rho or vice versa, we used a yeast two-hybrid assay, which has been extensively used to examine protein–protein interactions. 50 51 We subcloned rabbit Rac1 and RhoA cDNA into the yeast two-hybrid expression vectors (pBHA-Rac and pGAD-Rho or pBHA-Rho and pGAD-Rac) to use in our rabbit CECs. Cells transformed with the positive control vector had high β-galactosidase activity and grew in YNB+A medium (Fig. 4B) . However, the yeast cotransformant of pBHA-Rac1 and pGAD-RhoA or pBHA-RhoA and pGAD-Rac1 showed no β-galactosidase activity; and these yeast cotransformants did not grow in YNB+A, suggesting that no direct interaction takes place between Rac and Rho (Fig. 4B) . Because the subcloned rabbit Rho GTPases contained no carboxyl terminal polybasic region, which controls the diverse functions of these small GTPases, 52 53 we also tested a yeast two-hybrid assay with a combination of caRhoG14V and caRacG12V or dnRhoT19N and caRacG12V. The yeast two-hybrid assay with these full-length human Rho GTPase genes demonstrated neither β-galactosidase activity nor growth in YNB+A (data not shown). Thus, altogether, these data support the notion that the physical interaction between Rho and Rac may be mediated by other proteins. 
Activation of Cdc42 by FGF-2 through PI 3-kinase in Inducing the Protrusive Processes in CECs
The formation of protrusive processes (Fig. 3)observed in CECs expressing inactive Rho, enhanced by the exogenous FGF-2, raised the possibility that Cdc42 was also involved in morphologic alteration and actin cytoskeleton reorganization in CECs. We first determined Cdc42 activity in CECs. Cells maintained in DMEM-15 demonstrated the absence of Cdc42 activity, whereas a time-dependent activation of Cdc42 was observed in CECs stimulated by FGF-2 (Fig. 5A) . The maximum activity was observed in cells treated with FGF-2 for 8 hours, after which the activity was maintained at near maximum. LY294002 completely blocked Cdc42 activation, suggesting that Cdc42 was also activated by FGF-2 through the action of PI 3-kinase. To test further whether Cdc42 activation induces filopodia and microspikes, the genuine phenotypes mediated by active Cdc42, actin cytoskeleton of CECs expressing caCdc42G12V or dnCdc42T17N were examined 24 hours or 48 hours after transfection. Disruption of stress fibers was strongly observed in caCdc42G12V-expressing cells after a 24-hour transfection, whereas characteristic filopodia were barely detectable in these cells (Fig. 5B) . Of great interest, protrusive processes were observed 48 hours after transfection in CECs expressing active Cdc42 (Fig. 5B , arrowhead); in these cells, no filopodia were observed and cell shape was altered to an elongated morphology. In contrast, CECs expressing dnCdc42T17N demonstrated stress fiber formation in polygonal cells in a time-dependent manner, suggesting the presence of active Rho. When CECs expressing active caCdc42 were treated with exogenous FGF-2, reorganized actin at the cortex was strongly observed in elongated cells containing protrusive processes (Fig. 5B , arrowhead), suggesting that formation of cortical actin and protrusive processes uses separate pathways: cortical actin formation is mediated by active Rac, whereas protrusive processes are facilitated by active Cdc42. When CECs expressing inactive Cdc42 were treated with FGF-2, these cells completely lost their acquired phenotypes (stress fibers and polygonal cell shape) and developed a strong cortical actin ring structure mediated by FGF-2 (inactive Rho and active Rac). 
To examine whether Cdc42 physically associates with Rac and Rho, a reciprocal coimmunoprecipitation of the cross-linked cellular fractions was performed using anti-Rac1, -Rho, and -Cdc42 antibody. The immune complex precipitated with anti-Cdc42 antibody did not contain RhoA, and anti-Rho antibody failed to bring down Cdc42 (Fig. 6) . FGF-2 did not affect the protein–protein interaction between Cdc42 and RhoA. In contrast, Cdc42 was coimmunoprecipitated with anti-Rac1 antibody, and Rac1 was detected in the immune complex precipitated with anti-Cdc42 antibody (Fig. 6) . FGF-2 did not quantitatively affect the degree of interaction between Cdc42 and Rac. 
A yeast two-hybrid assay was used to identify further whether Cdc42 directly binds Rac. Yeast two-hybrid expression vectors, pBHA-Cdc42 and pGAD-Cdc42, containing rabbit Cdc42 cDNA (minus polybasic region), were constructed and then transformed with pGAD-Rac1 and pBHA-Rac1, individually or as pair-wise combinations, into the yeast L40 strain. There was no appreciable β-galactosidase activity beyond the level of the negative control in yeast cotransformants containing pBHA-Cdc42 and pGAD-Rac1 or vice versa (pBHA-Rac1 and pGAD-Cdc42). These transformants did not grow on YNB+A selection media (data not shown), suggesting that Cdc42 does not interact directly with Rac and that other proteins need a protein–protein interaction between Cdc42 and Rac. We also tested a yeast two-hybrid assay with a combination of caRacG12V and caCdc42G12V, using full-length human sequences. However, there was no positive protein–protein interaction between caRac and caCdc42 in β-galactosidase activity and growth in selection media (data not shown). 
To search for proteins that mediate protein–protein interaction among Rho GTPases, we tested whether the physical interaction between Rac and Rho or Rac and Cdc42 was mediated by Dbl family guanine nucleotide exchange factors (GEFs). Cell lysates obtained from CECs with or without FGF-2 were subjected to an immunoprecipitation assay, with anti-Vav2 DH/PH antibody, followed by an immunoblot analysis using anti-Rho, -Rac1, or -Cdc42 antibody. Figure 7Ademonstrates that the complex precipitated with anti-Vav2 DH/PH antibody contained all three Rho GTPases: Rho, Rac1, and Cdc42. This finding was further confirmed by reversing the experimental procedures: Immune complexes were respectively brought down with anti-Cdc42, Rac1, or Rho antibody and subjected to immunoblot analysis with anti-Vav2 DH/PH antibody (Fig. 7B) . All three immune complexes were shown to contain two polypeptide bands of 32 and 30 kDa. Immune complex by Rac antibody contained an equal ratio of the two polypeptide bands. Anti-Cdc42 antibody brought down more of a 32-kDa band than a 30-kDa band, whereas immune complex precipitated by anti-RhoA antibody contained a heavy 30-kDa band and a very faint 32-kDa band. Because, as Figure 7Ashows, the anti-Vav2 antibody equally precipitated both active and inactive Rho GTPases, it is not known whether the interaction of Rho GTPases with these Vav2-like peptides is mediated by both active and inactive forms of these proteins. The identities of the two polypeptide bands that appear to have homology to Vav2 DH/PH domain are yet to be determined. 
Finally, we confirmed that cortical actin formation precedes the formation of a protrusive structure. Actin cytoskeleton organization of CECs treated with FGF-2 was compared in a time-dependent manner (Fig. 8) . Cells treated with the growth factor for 24 hours demonstrated a simultaneous loss of stress fibers and reorganization of actin at the cortex. The presence of the cortical actin was predominantly observed in cells treated with FGF-2 for 48 hours. Further treatment of cells with FGF-2 (72 hours) resulted in elongated cells with prominent protrusive processes. These observations clearly indicate that loss of stress fibers precedes cortical actin ring formation, which is followed by protrusive structure formation. Sustained activity of Cdc42 in response to FGF-2 stimulation (Fig. 5A)also confirms this sequence of actin cytoskeleton reorganization. 
Discussion
Corneal fibrosis observed in Descemet’s membrane in vivo causes loss of vision by physically blocking light transmittance. Our previous studies demonstrated that such fibrosis is a consequence of EMT of corneal endothelial cells. 3 4 In this nonregenerative wound-healing pathway, endothelial cells resume their cell proliferative activity, which is otherwise arrested in the G1-phase of the cell cycle throughout the cells’ lifespan. These cells lose their contact inhibition, leading to the formation of multilayers of fibroblasts that in turn produce and deposit fibrillar ECM molecules. We also reported that the FGF-2 exerts a key role in such EMT: FGF-2 directly regulates cell cycle progression by degrading p27Kip1 through the PI 3-kinase pathways, 9 54 and FGF-2, via PI 3-kinase, upregulates the steady state levels of α1(I) collagen RNA by stabilizing the message 55 and subsequently facilitates synthesis and secretion of type I collagen into the extracellular space. 5 In contrast to these well-characterized phenotypes observed in our in vitro EMT model, it is not known how FGF-2 causes an alteration of the polygonal monolayer of CECs to multilayers of elongated fibroblastic cells. To understand the whole spectrum of EMT and corneal fibrosis, it is essential to elucidate the morphogenic pathways mediated by FGF-2 in CECs. 
It is now clear that cytoskeletal elements with their polymerization dynamics are central to many cellular activities, including morphogenesis and wound healing. 56 57 58 Assembly and organization of the actin cytoskeleton is regulated by Rho small GTPases, Rho, Rac, and Cdc42. 22 23 24 25 The conclusion that Rho, Rac, and Cdc42 regulate three separate signal-transduction pathways (stress fiber, lamellipodia, and filopodia) linking plasma membrane receptor to the assembly of distinct F-actin structures has been confirmed in a wide variety of cell systems. 22 23 24 25 26 27 28 29 We reported that, unlike these other cell systems, CECs demonstrate unique phenotypes in response to FGF-2 stimulation: the formation of lamellipodia and filopodia is not readily observed in CECs. FGF-2 alone facilitates cortical actin formation, whereas simultaneous treatment of cells with FGF-2 and Rho/ROCK inhibitor further induces protrusive processes in the elongated cells containing actin cytoskeleton at the cortex. 11 These unusual findings led us to investigate further the morphologic change and actin cytoskeletal reorganization of CECs mediated by FGF-2. 
The present study confirms the antagonizing activities between Rho and Rac and between Rho and Cdc42. It also demonstrates that Rac is upstream to Rho and that PI 3-kinase regulates all three of these Rho kinases (Fig. 9) . However, it is not known whether there is a hierarchy between Rac and Cdc42. We also showed that FGF-2 is able to disrupt stress fibers and that FGF-2 reorganizes actin cytoskeleton at the cortex by inhibiting Rho activity and activating Rac. Thus, the reorganized actin at the cortex is formed when Rho is inactivated and stress fibers are disrupted. Such inactivation of Rho is mediated by activated Rac through the action of PI 3-kinase. Simultaneously, the polygonal cell morphology achieved by prominent stress fibers and focal adhesion complexes is altered to an elongated cell shape through PI 3-kinase. This study further demonstrated that formation of protrusive processes is a late event (Fig. 8)and is mediated by activated Cdc42 through PI 3-kinase (Fig. 5) . Thus, reorganization of the actin cytoskeleton in response to FGF-2 stimulation is an orderly event: Disruption of stress fibers is a prerequisite to cortical actin formation, which is followed by the formation of protrusive processes. 
It is likely that these elongated cells with prominent protrusive processes represent the wound phenotypes that are actively involved in migration into the injury sites and that are capable of producing fibrillar ECM proteins to repair the injury sites, rapidly and nonqualitatively. Thus, it is likely that Cdc42 is involved in the final stage of actin cytoskeleton reorganization. We recently reported that interleukin-1β released by the infiltrating leukocytes during injury-related inflammation induces de novo synthesis of FGF-2 in CECs. 7 The present study further confirms that FGF-2 is sufficient to cause mesenchymal transformation of CECs by reorganizing the actin cytoskeleton in favor of de-adhesion and migratory phenotypes. It has been reported that cells transformed by Cdc42 grow in an adhesion-independent manner, exhibit altered morphologies, and display increased motility and invasiveness, all of which reflect alterations in the actin cytoskeleton. 59 60  
Our data also showed that protein–protein interaction between Rac and Cdc42 may be mediated by a DH/PH-containing a 32-kDa protein, whereas the interaction between Rac and Rho may be mediated by another DH/PH containing a 30-kDa protein. Identification of these proteins that share identity with the DH/PH domain of Vav2 is in progress. These proteins may be members of GEFs that are closely related to Vav2, even though their molecular weights are far lower than that of Vav2. Although it is not known how the Vav2-like peptides facilitate the protein–protein interaction between Rho GTPases in CECs, GEFs are likely to have a role in this EMT model, because PI 3-kinase is the major signaling molecule in cell shape change and actin cytoskeleton organization of CECs. The involvement of PI 3-kinase in Rho GTPases is known to be mediated through GEFs, which contain a PH domain to which phosphatidylinositol-(3,4,5)-triphosphate binds. 15 16 Such interaction leads to activation of Rac and/or Cdc42 15 61 and a subsequent reorganization of actin cytoskeleton. In addition, it has been reported that the activated forms of Rac and Cdc42 directly bind p85, a regulatory subunit of PI 3-kinase, resulting in the activation of PI 3-kinase 62 63 and, consequently, in an increase in the level of phosphatidylinositol-(3,4,5)-triphosphate. Thus, these direct and/or indirect activation pathways may provide a link between the PI 3-kinase and the Rac1 and/or Cdc42 signaling pathways. Together, these studies demonstrate that two stages of activation of Rho GTPases (Rac and Cdc42) are necessary to inhibit Rho activity completely and to cause the loss of stress fibers and the subsequent loss of the adhesion phenotype of CECs. Subsequent formation of protrusive processes mediated by the sustained activation of Cdc42 finalizes the EMT of CECs and completes the nonregenerative wound-healing process (corneal fibrosis). It is likely that the FGF-2 present in Descemet’s membrane induces CECs in vivo to organize the actin cytoskeleton at the cortex. 64 Under physiological conditions, CECs in vivo maintain their polygonal cell shape with cortical actin under the minimal influence of FGF-2. 11 When injury-mediated inflammation modulates the local concentration of FGF-2, the FGF-2 is able to exert its activities further on Rho GTPases, as observed in the present study, causing the polygonal cells to convert to the spindle-shaped cells with prominent protrusive processes that are responsible for nonregenerative wound healing. 
 
Figure 1.
 
FGF-2 mediated cell shape change in CECs. First-passage cells were plated in a four-well chamber at a cell density of 1 × 104 and maintained for 1 day in DMEM-15. On day 2, cells were further maintained in DMEM-15 with or without FGF-2 or were treated simultaneously with FGF-2 and LY294002 (LY) for 1 hour (A) or 16 hours (B). FGF-2 treatment of CECs led to significant stimulation of cell morphologic change from polygonal cobble-like to fibroblast-like cell shape. Arrowheads: region of lamellipodia formation induced by FGF-2. The data are representative of results in four experiments.
Figure 1.
 
FGF-2 mediated cell shape change in CECs. First-passage cells were plated in a four-well chamber at a cell density of 1 × 104 and maintained for 1 day in DMEM-15. On day 2, cells were further maintained in DMEM-15 with or without FGF-2 or were treated simultaneously with FGF-2 and LY294002 (LY) for 1 hour (A) or 16 hours (B). FGF-2 treatment of CECs led to significant stimulation of cell morphologic change from polygonal cobble-like to fibroblast-like cell shape. Arrowheads: region of lamellipodia formation induced by FGF-2. The data are representative of results in four experiments.
Figure 2.
 
Effect of FGF-2 on Rac and Rho activity through PI 3-kinase. For Rac and Rho pull-down assay, CECs were cultured until they reached 70% confluence in DMEM-15 and then maintained with the indicated media condition and time. Cells were lysed with lysis buffer and active GTPase were affinity-precipitated with PAK-CRIB for Rac (A) or Rhotekin-RBD for Rho (B), eluted from the bead and analyzed by Western blot with the relevant antibody. For each time point, a fraction of total lysates was immunoblotted to monitor the amount of GTPase before precipitation, and the amount of activated Rac and Rho was normalized to the total amount of Rac and Rho in the cell lysates to compare Rac and Rho activity (level of GTP bound Rac and Rho) in different cell lysates. GTPγS and GDP treatment served as positive and negative controls, respectively, for the pull-down assay. As expected, Rac and Rho GTPase did not pull down in the presence of GDP, confirming the specificity for the pull-down of Rac and Rho GTPases. D-15 and LY designate DMEM-15 and LY294002, respectively; these abbreviations are used in all figure legends. The data represent the average of the results of two independent experiments.
Figure 2.
 
Effect of FGF-2 on Rac and Rho activity through PI 3-kinase. For Rac and Rho pull-down assay, CECs were cultured until they reached 70% confluence in DMEM-15 and then maintained with the indicated media condition and time. Cells were lysed with lysis buffer and active GTPase were affinity-precipitated with PAK-CRIB for Rac (A) or Rhotekin-RBD for Rho (B), eluted from the bead and analyzed by Western blot with the relevant antibody. For each time point, a fraction of total lysates was immunoblotted to monitor the amount of GTPase before precipitation, and the amount of activated Rac and Rho was normalized to the total amount of Rac and Rho in the cell lysates to compare Rac and Rho activity (level of GTP bound Rac and Rho) in different cell lysates. GTPγS and GDP treatment served as positive and negative controls, respectively, for the pull-down assay. As expected, Rac and Rho GTPase did not pull down in the presence of GDP, confirming the specificity for the pull-down of Rac and Rho GTPases. D-15 and LY designate DMEM-15 and LY294002, respectively; these abbreviations are used in all figure legends. The data represent the average of the results of two independent experiments.
Figure 3.
 
Requirement of Rac activation and Rho inactivation for the induction of cortical actin and fibroblast-like cell shape containing protrusive processes. CECs were transiently transfected with caRacG12V, dnRacT17N, caRhoG14V, and dnRhoT19N. Transfectants were maintained in DMEM-15 (A) or DMEM-15 with FGF-2 (B) for 48 hours and then fixed. F-actin and HA were labeled with rhodamine-phalloidin and FITC-conjugated anti-HA antibody, respectively. Negative control was stained in the absence of primary antibody. Arrowheads: the region of lamellipodia. The data are representative of results in four experiments.
Figure 3.
 
Requirement of Rac activation and Rho inactivation for the induction of cortical actin and fibroblast-like cell shape containing protrusive processes. CECs were transiently transfected with caRacG12V, dnRacT17N, caRhoG14V, and dnRhoT19N. Transfectants were maintained in DMEM-15 (A) or DMEM-15 with FGF-2 (B) for 48 hours and then fixed. F-actin and HA were labeled with rhodamine-phalloidin and FITC-conjugated anti-HA antibody, respectively. Negative control was stained in the absence of primary antibody. Arrowheads: the region of lamellipodia. The data are representative of results in four experiments.
Figure 4.
 
Physical association of Rac with Rho. (A) CECs were cultured in DMEM-15 with or without FGF-2 for 24 hours and then lysed. Cell lysates were immunoprecipitated with the designated antibody, and the immunoprecipitated complexes were immunoblotted with either anti-Rac or anti-Rho antibody. Antibody to Rho reacts with all three forms of Rho (RhoA, RhoB, and RhoC). (B) Yeast two-hybrid assay was performed using the yeast L40 strain harboring reporter gene HIS3 and β-galactosidase. Partial rabbit Rac and Rho coding regions were subcloned into the pBHA and pGAD10 plasmids of the yeast two-hybrid system and transformed reciprocally into the yeast cells. Cotransformants were plated on X-Gal plates and selective media (YNB+A) to ascertain the interaction between Rac and Rho. Plates were photographed after a 2-day incubation at 30°C. The data are representative of results in four experiments.
Figure 4.
 
Physical association of Rac with Rho. (A) CECs were cultured in DMEM-15 with or without FGF-2 for 24 hours and then lysed. Cell lysates were immunoprecipitated with the designated antibody, and the immunoprecipitated complexes were immunoblotted with either anti-Rac or anti-Rho antibody. Antibody to Rho reacts with all three forms of Rho (RhoA, RhoB, and RhoC). (B) Yeast two-hybrid assay was performed using the yeast L40 strain harboring reporter gene HIS3 and β-galactosidase. Partial rabbit Rac and Rho coding regions were subcloned into the pBHA and pGAD10 plasmids of the yeast two-hybrid system and transformed reciprocally into the yeast cells. Cotransformants were plated on X-Gal plates and selective media (YNB+A) to ascertain the interaction between Rac and Rho. Plates were photographed after a 2-day incubation at 30°C. The data are representative of results in four experiments.
Figure 5.
 
Effect of Cdc42 activation caused by FGF-2 through PI 3-kinase on morphologic alteration in CECs. (A) Culture and lysis of CECs was performed as described in Figure 2 . Active Cdc42 was affinity precipitated with GST-CRIB-PAK and analyzed by immunoblot with the anti-Cdc42 antibody. A fraction of total lysates for each culture time point was immunoblotted to monitor the total amount of GTPase before precipitation, and the amount of activated Cdc42 was normalized to the total amount of Cdc42, for the comparison of Cdc42 activity (level of GTP-bound Cdc42) in different cell lysates. GTPγS and GDP treatment served as positive and negative controls, respectively, in the pull-down assay. As expected, Cdc42 did not pull down in the presence of GDP, confirming the specificity of the pull-down of Cdc42. The data represent the average of the results in two independent experiments. (B) CECs were transiently transfected with caCdc42G12V or dnCdc42T17N. Transfectants were cultured in DMEM-15, with or without FGF-2 for the indicated time and then fixed with paraformaldehyde. Cells were labeled with F-actin (rhodamine-phalloidin) and HA (FITC) antibody, respectively. Arrowheads: the region of protrusive structures. The data are representative of results in four experiments.
Figure 5.
 
Effect of Cdc42 activation caused by FGF-2 through PI 3-kinase on morphologic alteration in CECs. (A) Culture and lysis of CECs was performed as described in Figure 2 . Active Cdc42 was affinity precipitated with GST-CRIB-PAK and analyzed by immunoblot with the anti-Cdc42 antibody. A fraction of total lysates for each culture time point was immunoblotted to monitor the total amount of GTPase before precipitation, and the amount of activated Cdc42 was normalized to the total amount of Cdc42, for the comparison of Cdc42 activity (level of GTP-bound Cdc42) in different cell lysates. GTPγS and GDP treatment served as positive and negative controls, respectively, in the pull-down assay. As expected, Cdc42 did not pull down in the presence of GDP, confirming the specificity of the pull-down of Cdc42. The data represent the average of the results in two independent experiments. (B) CECs were transiently transfected with caCdc42G12V or dnCdc42T17N. Transfectants were cultured in DMEM-15, with or without FGF-2 for the indicated time and then fixed with paraformaldehyde. Cells were labeled with F-actin (rhodamine-phalloidin) and HA (FITC) antibody, respectively. Arrowheads: the region of protrusive structures. The data are representative of results in four experiments.
Figure 6.
 
Physical association of Cdc42 with Rac, but not with Rho. CECs were cultured in DMEM-15, with or without FGF-2, for 24 hours and then lysed. Immunoprecipitation and immunoblot analysis were performed with the designated antibody, as described in Figure 4 . The data are representative of results in four independent experiments.
Figure 6.
 
Physical association of Cdc42 with Rac, but not with Rho. CECs were cultured in DMEM-15, with or without FGF-2, for 24 hours and then lysed. Immunoprecipitation and immunoblot analysis were performed with the designated antibody, as described in Figure 4 . The data are representative of results in four independent experiments.
Figure 7.
 
Physical interaction between Rho GTPases and guanine nucleotide exchange factor. (A) First-passage cells were cultured in DMEM-15, with or without FGF-2 for 24 hours and then lysed. Cell lysates were immunoprecipitated with anti-Vav2 DH/PH antibody and then equally loaded on SDS-PAGE. Precipitated endogenous DH/PH domain-containing proteins were probed with anti-Rho (top), Rac1 (middle), and Cdc42 (bottom) antibodies. (B) CECs were maintained in DMEM-15 with FGF-2 for 24 hours and then lysed. Cell lysates were divided into three equal parts and immunoprecipitated with anti-Cdc42, Rac1 and Rho antibody, respectively. Immunoprecipitated complexes were then immunoblotted with anti-Vav2 DH/PH antibody. The data are representative of results in three independent experiments.
Figure 7.
 
Physical interaction between Rho GTPases and guanine nucleotide exchange factor. (A) First-passage cells were cultured in DMEM-15, with or without FGF-2 for 24 hours and then lysed. Cell lysates were immunoprecipitated with anti-Vav2 DH/PH antibody and then equally loaded on SDS-PAGE. Precipitated endogenous DH/PH domain-containing proteins were probed with anti-Rho (top), Rac1 (middle), and Cdc42 (bottom) antibodies. (B) CECs were maintained in DMEM-15 with FGF-2 for 24 hours and then lysed. Cell lysates were divided into three equal parts and immunoprecipitated with anti-Cdc42, Rac1 and Rho antibody, respectively. Immunoprecipitated complexes were then immunoblotted with anti-Vav2 DH/PH antibody. The data are representative of results in three independent experiments.
Figure 8.
 
Induction of sequential actin reorganization by FGF-2. First-passage cells were plated in a four-well chamber and maintained in DMEM-15 with or without FGF-2 for the indicated time. Cells were then fixed and stained for vinculin (FITC) and F-actin (rhodamine-phalloidin). The data are representative results in three experiments.
Figure 8.
 
Induction of sequential actin reorganization by FGF-2. First-passage cells were plated in a four-well chamber and maintained in DMEM-15 with or without FGF-2 for the indicated time. Cells were then fixed and stained for vinculin (FITC) and F-actin (rhodamine-phalloidin). The data are representative results in three experiments.
Figure 9.
 
The FGF-2-derived pathway affects cytoskeletal reorganization and EMT. The diagram illustrates the signal-transduction pathways via PI 3-kinase activated by FGF-2 that affect organization of actin cytoskeleton. PI 3-kinase activates both Rac and Cdc42. Activated Rac further antagonizes Rho, which promotes cortical actin formation, whereas activated Cdc42 also antagonizes Rho, which promotes protrusive structure formation. Both reorganization of actin cytoskeleton at the cortex and acquisition of protrusive processes are nedessary for the EMT process. The nature of the cross-reactivity between Rac and Cdc42 is unknown.
Figure 9.
 
The FGF-2-derived pathway affects cytoskeletal reorganization and EMT. The diagram illustrates the signal-transduction pathways via PI 3-kinase activated by FGF-2 that affect organization of actin cytoskeleton. PI 3-kinase activates both Rac and Cdc42. Activated Rac further antagonizes Rho, which promotes cortical actin formation, whereas activated Cdc42 also antagonizes Rho, which promotes protrusive structure formation. Both reorganization of actin cytoskeleton at the cortex and acquisition of protrusive processes are nedessary for the EMT process. The nature of the cross-reactivity between Rac and Cdc42 is unknown.
The authors thank Eunjoon Kim for advice regarding the yeast two-hybrid assay system and John G. Collard for generously providing the GST-PAK-CRIB expression plasmid. 
BrownSI, KitanoS. Pathogenesis of the retrocorneal membrane. Arch Ophthalmol. 1966;75:518–525. [CrossRef] [PubMed]
MichelsRG, KenyonKR, MaumenceAE. Retrocorneal fibrous membrane. Invest Ophthalmol. 1972;11:822–831. [PubMed]
KayEP, CheungCC, JesterJV, NimniME, SmithRE. Type I collagen and fibronectin synthesis by retrocorneal fibrous membrane. Invest Ophthalmol Vis Sci. 1982;22:200–212. [PubMed]
LeungEW, RifeL, SmithRE, KayEP. Extracellular matrix components in retrocorneal fibrous membrane in comparison to corneal endothelium and Descemet’s membrane. Mol Vis. 2000;6:15–23. [PubMed]
KayEP, GuX, NinomiyaY, SmithRE. Corneal endothelial modulation: a factor released by leukocytes induces basic fibroblast growth factor that modulates cell shape and collagen. Invest Ophthalmol Vis Sci. 1993;34:663–672. [PubMed]
KayEP, GuX, SmithRE. Corneal endothelial modulation: bFGF as direct mediator and corneal endothelium modulation factor as inducer. Invest Ophthalmol Vis Sci. 1994;35:2427–2435. [PubMed]
LeeHT, LeeJG, NaM, KayEP. FGF-2 induced by interleukin-1 beta through the action of phosphatidylinositol 3-kinase mediates endothelial mesenchymal transformation in corneal endothelial cells. J Biol Chem. 2004;279:32325–32332. [CrossRef] [PubMed]
GuX, SeongGJ, LeeYG, KayEP. Fibroblast growth factor 2 uses distinct signaling pathways for cell proliferation and cell shape changes in corneal endothelial cells. Invest Ophthalmol Vis Sci. 1996;37:2326–2334. [PubMed]
LeeHT, KayEP. Regulatory role of PI 3-kinase on expression of Cdk4 and p27, nuclear localization of Cdk4, and phosphorylation of p27 in corneal endothelial cells. Invest Ophthalmol Vis Sci. 2003;44:1521–1528. [CrossRef] [PubMed]
KayEP, ParkSY, KoMK, LeeSC. Fibroblast growth factor 2 uses PLC- 1 for cell proliferation and PI 3-kinase for alteration of cell shape and cell proliferation in corneal endothelial cells. Mol Vis. 1998;4:22. [PubMed]
LeeHT, KayEP. FGF-2 induced reorganization and disruption of actin cytoskeleton through PI 3-kinase, Rho, and Cdc42 in corneal endothelial cells. Mol Vis. 2003;9:624–634. [PubMed]
ColladoM, MedemaRH, Garcia-CaoI, et al. Inhibition of the phosphoinositide 3-kinase pathway induces a senescence-like arrest mediated by p27Kip1. J Biol Chem. 2000;275:21960–21968. [CrossRef] [PubMed]
Phillips-MasonPJ, RabenDM, BaldassareJJ. Phosphatidylinositol 3-kinase activity regulates alpha-thrombin-stimulated G1 progression by its effect on cyclin D1 expression and cyclin-dependent kinase 4 activity. J Biol Chem. 2000;275:18046–18053. [CrossRef] [PubMed]
NaritaY, NaganeM, MishimaK, HuangHJ, FurnariFB, CaveneeWK. Mutant epidermal growth factor receptor signaling down-regulates p27 through activation of the phosphatidylinositol 3-kinase/Akt pathway in glioblastomas. Cancer Res. 2002;62:6764–6769. [PubMed]
WelchHCE, CoadwellWJ, StephensLR, HawkinsPT. Phosphoinositide 3-kinase-dependent activation of Rac. FEBS Lett. 2003;546:93–97. [CrossRef] [PubMed]
JimenezC, PortelaRA, MelladoM, et al. Role of the PI3K regulatory subunit in the control of actin organization and cell migration. J Cell Biol. 2000;151:249–261. [CrossRef] [PubMed]
SchmitzAA, GovekEE, BottnerB, Van AelstL. Rho GTPases: signaling, migration, and invasion. Exp Cell Res. 2000;261:1–12. [CrossRef] [PubMed]
CaronE, HallA. Identification of two distinct mechanisms of phagocytosis controlled by different Rho GTPases. Science. 1998;282:1717–1721. [CrossRef] [PubMed]
NobesCD, HallA. Rho GTPases control polarity, protrusion, and adhesion during cell movement. J Cell Biol. 1999;144:1235–1244. [CrossRef] [PubMed]
AllenWE, ZichaD, RidleyAJ, JonesGE. A role for Cdc42 in macrophage chemotaxis. J Cell Biol. 1998;141:1147–1157. [CrossRef] [PubMed]
SettlemanJ. Rho GTPases in development. Prog Mol Subcell Biol. 1999;22:201–229. [PubMed]
Van AelstL, D’Souza-SchoreyC. Rho GTPases and signaling networks. Genes Dev. 1997;11:2295–2322. [CrossRef] [PubMed]
HallA. Rho GTPases and the actin cytoskeleton. Science. 1998;279:509–514. [CrossRef] [PubMed]
RidleyAJ. Rho family proteins: coordinating cell responses. Trends Cell Biol. 2001;11:471–477. [CrossRef] [PubMed]
Etienne-MannevilleS, HallA. Rho GTPases in cell biology. Nature. 2002;420:629–635. [CrossRef] [PubMed]
RidleyAJ, PatersonHF, JohnstonCL, DiekmannD, HallA. The small GTP-binding protein rac regulates growth factor-induced membrane ruffling. Cell. 1992;70:401–410. [CrossRef] [PubMed]
RidleyAJ, HallA. The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell. 1992;70:389–399. [CrossRef] [PubMed]
NobesCD, HallA. Rho, Rac and Cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia. Cell. 1995;81:53–62. [CrossRef] [PubMed]
JohnsonDI, PringleJR. Molecular characterization of CDC42, a Saccharomyces cerevisiae gene involved in the development of cell polarity. J Cell Biol. 1990;111:143–152. [CrossRef] [PubMed]
ZondagGCM, EversEE, ten KloosterJP, JanssenL, van der KammenRA, CollardJG. Oncogenic Ras downregulates Rac activity, which leads to increased Rho activity and epithelial-mesenchymal transition. J Cell Biol. 2000;149:775–781. [CrossRef] [PubMed]
SanderEE, ten KloosterJP, van DelftS, van der KammenRA, CollardJG. Rac downregulates Rho activity: reciprocal balance between both GTPases determines cellular morphology and migratory behavior. J Cell Biol. 1999;147:1009–1021. [CrossRef] [PubMed]
KurokawaK, ItohRE, YoshizakiH, NakamuraYO, MatsudaM. Coactivation of Rac1 and Cdc42 at lamellipodia and membrane ruffles induced by epidermal growth factor. Mol Biol Cell. 2004;15:1003–1010. [PubMed]
VidalC, GenyB, MelleJ, Jandrot-PerrusM, Fontenay-RoupieM. Cdc42/Rac1-dependent activation of the p21-activated kinase (PAK) regulates human platelet lamellipodia spreading: implication of the cortical-actin binding protein cortactin. Blood. 2002;100:4462–4469. [CrossRef] [PubMed]
TsujiT, IshizakiT, OkamotoM, et al. ROCK and mDia1 antagonize in Rho-dependent Rac activation in Swiss 3T3 fibroblasts. J Cell Biol. 2002;157:819–830. [CrossRef] [PubMed]
HabasR, DawidIB, HeX. Coactivation of Rac and Rho by Wnt/Frizzed signaling is required for vertebrate gastrulation. Genes Dev. 2003;17:295–309. [CrossRef] [PubMed]
DieboldBA, FowlerB, LuJ, DinauerMC, BokochGM. Antagonistic cross-talk between Rac and Cdc42 GTPases regulates generation of reactive oxygen species. J Biol Chem. 2004;279:28136–28142. [CrossRef] [PubMed]
BurbeloPD, SnowDM, BahouW, SpiegelS. MSE55, a Cdc42 effector protein, induces long cellular extensions in fibroblasts. Proc Natl Acad Sci USA. 1999;96:9083–9088. [CrossRef] [PubMed]
HirschDS, PironeDM, BurbeloPD. A new family of Cdc42 effector proteins, CEPs, function in fibroblast and epithelial cell shape changes. J Biol Chem. 2001;276:875–883. [CrossRef] [PubMed]
ZhongC, KinchMS, BurridgeK. Rho-stimulated contractility contributes to the fibroblastic phenotype of Ras-transformed epithelial cells. Mol Biol Cell. 1997;8:2329–2344. [CrossRef] [PubMed]
ZongH, KaibuchiK, QuilliamLA. The insert region of RhoA is essential for Rho kinase activation and cellular transformation. Mol Cell Biol. 2001;21:5287–5298. [CrossRef] [PubMed]
KayEP, SmithRE, NimniME. Basement membrane collagen synthesis by rabbit corneal endothelial cells in culture: evidence for an alpha chain derived from a larger biosynthetic precursor. J Biol Chem. 1982;257:7116–7121. [PubMed]
GuX, KoMK, KayEP. Intracellular interaction of Hsp47 and type I collagen in corneal endothelial cells. Invest Ophthalmol Vis Sci. 1999;40:289–295. [PubMed]
NakaiA, SatohM, HirayoshiK, NagataK. Involvement of the stress protein HSP47 in procollagen processing in the endoplasmic reticulum. J Cell Biol. 1992;117:903–914. [CrossRef] [PubMed]
KoMK, KayEP. Subcellular localization of procollagen I and prolyl 4-hydroxylase in corneal endothelial cells. Exp Cell Res. 2001;264:363–371. [CrossRef] [PubMed]
ItoH, FukudaY, MurataK, KimuraA. Transformation of intact yeast cells treated with alkali cations. J Bacteriol. 1983;153:163–168. [PubMed]
SanderEE, van DelftS, ten KloosterJP, et al. Matrix-dependent Tiam1/Rac signaling in epithelial cells promotes either cell-cell adhesion or cell migration and is regulated by phosphatidylinositol 3-kinase. J Cell Biol. 1998;143:1385–1398. [CrossRef] [PubMed]
GuX, KayEP. Distribution and putative roles of fibroblast growth factor-2 isoforms in corneal endothelial modulation. Invest Ophthalmol Vis Sci. 1998;39:2252–2258. [PubMed]
ManserE, HuangHY, LooTH, et al. Expression of constitutively active alpha-PAK reveals effects of the kinase on actin and focal complexes. Mol Cell Biol. 1997;17:1129–1143. [PubMed]
RidleyAJ, HallA. Signal transduction pathways regulating Rho-mediated stress fibre formation: requirement for a tyrosine kinase. EMBO J. 1994;13:2600–2610. [PubMed]
KoJ, NaM, KimS, LeeJR, KimE. Interaction of the ERC family of RIM-binding proteins with the liprin-alpha family of multidomain proteins. J Biol Chem. 2003;278:42377–42385. [CrossRef] [PubMed]
ParkerAE, Van de WeyerI, LausMC, VerhasseltP, LuytenWH. Identification of a human homologue of the Schizosaccharomyces pombe rad17+ checkpoint gene. J Biol Chem. 1998;273:18340–18346. [CrossRef] [PubMed]
WilliamsCL. The polybasic region of Ras and Rho family small GTPases: a regulator of protein interactions and membrane association and a site of nuclear localization signal sequences. Cell Signal. 2003;15:1071–1080. [CrossRef] [PubMed]
LanningCC, DaddonaJL, Ruiz-VelascoR, ShaferSH, WilliamsCL. The Rac1 C-terminal polybasic region regulates the nuclear localization and protein degradation of Rac1. J Biol Chem. 2004;279:44197–44210. [CrossRef] [PubMed]
LeeHT, KayEP. Regulatory role of cAMP on expression of Cdk4 and p27(Kip1) by inhibiting phosphatidylinositol 3-kinase in corneal endothelial cells. Invest Ophthalmol Vis Sci. 2003;44:3816–3825. [CrossRef] [PubMed]
KoMK, KayEP. Regulatory role of FGF-2 on type I collagen expression during endothelial mesenchymal transformation. Invest Ophthalmol Vis Sci. 2005;46:4495–4503. [CrossRef] [PubMed]
ConnollyJO, SimpsonN, HewlettL, HallA. Rac regulates endothelial morphogenesis and capillary assembly. Mol Biol Cell. 2002;13:2474–2485. [CrossRef] [PubMed]
YuanXB, JinM, XuX, et al. Signalling and crosstalk of Rho GTPases in mediating axon guidance. Nat Cell Biol. 2003;5:38–45. [PubMed]
SachdevP, ZengL, WangLH. Distinct role of phosphatidylinositol 3-kinase and Rho family GTPases in Vav3-induced cell transformation, cell motility, and morphological changes. J Biol Chem. 2002;277:17638–17648. [CrossRef] [PubMed]
LinR, BagrodiaS, CerioneR, ManorD. A novel Cdc42Hs mutant induces cellular transformation. Curr Biol. 1997;7:794–797. [CrossRef] [PubMed]
QiuRG, AboA, McCormickF, SymonsM. Cdc42 regulates anchorage-independent growth and is necessary for Ras transformation. Mol Cell Biol. 1997;17:3449–3458. [PubMed]
InnocentiM, FrittoliE, PonzanelliI, et al. Phosphoinositide 3-kinase activates Rac by entering in a complex with Eps8, Abi1, and Sos-1. J Cell Biol. 2003;160:17–23. [CrossRef] [PubMed]
ToliasKF, CantleyLC, CarpenterCL. Rho family GTPases bind to phosphoinositide kinases. J Biol Chem. 1995;270:17656–17659. [CrossRef] [PubMed]
BagrodiaS, DerijardB, DavisRJ, CerioneRA. Cdc42 and PAK-mediated signaling leads to Jun kinase and p38 mitogen-activated protein kinase activation. J Biol Chem. 1995;270:27995–27998. [CrossRef] [PubMed]
VlodavskyI, FolkmanJ, SullivanR, et al. Endothelial cell-derived basic fibroblast growth factor: synthesis and deposition into subendothelial extracellular matrix. Proc Natl Acad Sci USA. 1987;84:2292–2296. [CrossRef] [PubMed]
Figure 1.
 
FGF-2 mediated cell shape change in CECs. First-passage cells were plated in a four-well chamber at a cell density of 1 × 104 and maintained for 1 day in DMEM-15. On day 2, cells were further maintained in DMEM-15 with or without FGF-2 or were treated simultaneously with FGF-2 and LY294002 (LY) for 1 hour (A) or 16 hours (B). FGF-2 treatment of CECs led to significant stimulation of cell morphologic change from polygonal cobble-like to fibroblast-like cell shape. Arrowheads: region of lamellipodia formation induced by FGF-2. The data are representative of results in four experiments.
Figure 1.
 
FGF-2 mediated cell shape change in CECs. First-passage cells were plated in a four-well chamber at a cell density of 1 × 104 and maintained for 1 day in DMEM-15. On day 2, cells were further maintained in DMEM-15 with or without FGF-2 or were treated simultaneously with FGF-2 and LY294002 (LY) for 1 hour (A) or 16 hours (B). FGF-2 treatment of CECs led to significant stimulation of cell morphologic change from polygonal cobble-like to fibroblast-like cell shape. Arrowheads: region of lamellipodia formation induced by FGF-2. The data are representative of results in four experiments.
Figure 2.
 
Effect of FGF-2 on Rac and Rho activity through PI 3-kinase. For Rac and Rho pull-down assay, CECs were cultured until they reached 70% confluence in DMEM-15 and then maintained with the indicated media condition and time. Cells were lysed with lysis buffer and active GTPase were affinity-precipitated with PAK-CRIB for Rac (A) or Rhotekin-RBD for Rho (B), eluted from the bead and analyzed by Western blot with the relevant antibody. For each time point, a fraction of total lysates was immunoblotted to monitor the amount of GTPase before precipitation, and the amount of activated Rac and Rho was normalized to the total amount of Rac and Rho in the cell lysates to compare Rac and Rho activity (level of GTP bound Rac and Rho) in different cell lysates. GTPγS and GDP treatment served as positive and negative controls, respectively, for the pull-down assay. As expected, Rac and Rho GTPase did not pull down in the presence of GDP, confirming the specificity for the pull-down of Rac and Rho GTPases. D-15 and LY designate DMEM-15 and LY294002, respectively; these abbreviations are used in all figure legends. The data represent the average of the results of two independent experiments.
Figure 2.
 
Effect of FGF-2 on Rac and Rho activity through PI 3-kinase. For Rac and Rho pull-down assay, CECs were cultured until they reached 70% confluence in DMEM-15 and then maintained with the indicated media condition and time. Cells were lysed with lysis buffer and active GTPase were affinity-precipitated with PAK-CRIB for Rac (A) or Rhotekin-RBD for Rho (B), eluted from the bead and analyzed by Western blot with the relevant antibody. For each time point, a fraction of total lysates was immunoblotted to monitor the amount of GTPase before precipitation, and the amount of activated Rac and Rho was normalized to the total amount of Rac and Rho in the cell lysates to compare Rac and Rho activity (level of GTP bound Rac and Rho) in different cell lysates. GTPγS and GDP treatment served as positive and negative controls, respectively, for the pull-down assay. As expected, Rac and Rho GTPase did not pull down in the presence of GDP, confirming the specificity for the pull-down of Rac and Rho GTPases. D-15 and LY designate DMEM-15 and LY294002, respectively; these abbreviations are used in all figure legends. The data represent the average of the results of two independent experiments.
Figure 3.
 
Requirement of Rac activation and Rho inactivation for the induction of cortical actin and fibroblast-like cell shape containing protrusive processes. CECs were transiently transfected with caRacG12V, dnRacT17N, caRhoG14V, and dnRhoT19N. Transfectants were maintained in DMEM-15 (A) or DMEM-15 with FGF-2 (B) for 48 hours and then fixed. F-actin and HA were labeled with rhodamine-phalloidin and FITC-conjugated anti-HA antibody, respectively. Negative control was stained in the absence of primary antibody. Arrowheads: the region of lamellipodia. The data are representative of results in four experiments.
Figure 3.
 
Requirement of Rac activation and Rho inactivation for the induction of cortical actin and fibroblast-like cell shape containing protrusive processes. CECs were transiently transfected with caRacG12V, dnRacT17N, caRhoG14V, and dnRhoT19N. Transfectants were maintained in DMEM-15 (A) or DMEM-15 with FGF-2 (B) for 48 hours and then fixed. F-actin and HA were labeled with rhodamine-phalloidin and FITC-conjugated anti-HA antibody, respectively. Negative control was stained in the absence of primary antibody. Arrowheads: the region of lamellipodia. The data are representative of results in four experiments.
Figure 4.
 
Physical association of Rac with Rho. (A) CECs were cultured in DMEM-15 with or without FGF-2 for 24 hours and then lysed. Cell lysates were immunoprecipitated with the designated antibody, and the immunoprecipitated complexes were immunoblotted with either anti-Rac or anti-Rho antibody. Antibody to Rho reacts with all three forms of Rho (RhoA, RhoB, and RhoC). (B) Yeast two-hybrid assay was performed using the yeast L40 strain harboring reporter gene HIS3 and β-galactosidase. Partial rabbit Rac and Rho coding regions were subcloned into the pBHA and pGAD10 plasmids of the yeast two-hybrid system and transformed reciprocally into the yeast cells. Cotransformants were plated on X-Gal plates and selective media (YNB+A) to ascertain the interaction between Rac and Rho. Plates were photographed after a 2-day incubation at 30°C. The data are representative of results in four experiments.
Figure 4.
 
Physical association of Rac with Rho. (A) CECs were cultured in DMEM-15 with or without FGF-2 for 24 hours and then lysed. Cell lysates were immunoprecipitated with the designated antibody, and the immunoprecipitated complexes were immunoblotted with either anti-Rac or anti-Rho antibody. Antibody to Rho reacts with all three forms of Rho (RhoA, RhoB, and RhoC). (B) Yeast two-hybrid assay was performed using the yeast L40 strain harboring reporter gene HIS3 and β-galactosidase. Partial rabbit Rac and Rho coding regions were subcloned into the pBHA and pGAD10 plasmids of the yeast two-hybrid system and transformed reciprocally into the yeast cells. Cotransformants were plated on X-Gal plates and selective media (YNB+A) to ascertain the interaction between Rac and Rho. Plates were photographed after a 2-day incubation at 30°C. The data are representative of results in four experiments.
Figure 5.
 
Effect of Cdc42 activation caused by FGF-2 through PI 3-kinase on morphologic alteration in CECs. (A) Culture and lysis of CECs was performed as described in Figure 2 . Active Cdc42 was affinity precipitated with GST-CRIB-PAK and analyzed by immunoblot with the anti-Cdc42 antibody. A fraction of total lysates for each culture time point was immunoblotted to monitor the total amount of GTPase before precipitation, and the amount of activated Cdc42 was normalized to the total amount of Cdc42, for the comparison of Cdc42 activity (level of GTP-bound Cdc42) in different cell lysates. GTPγS and GDP treatment served as positive and negative controls, respectively, in the pull-down assay. As expected, Cdc42 did not pull down in the presence of GDP, confirming the specificity of the pull-down of Cdc42. The data represent the average of the results in two independent experiments. (B) CECs were transiently transfected with caCdc42G12V or dnCdc42T17N. Transfectants were cultured in DMEM-15, with or without FGF-2 for the indicated time and then fixed with paraformaldehyde. Cells were labeled with F-actin (rhodamine-phalloidin) and HA (FITC) antibody, respectively. Arrowheads: the region of protrusive structures. The data are representative of results in four experiments.
Figure 5.
 
Effect of Cdc42 activation caused by FGF-2 through PI 3-kinase on morphologic alteration in CECs. (A) Culture and lysis of CECs was performed as described in Figure 2 . Active Cdc42 was affinity precipitated with GST-CRIB-PAK and analyzed by immunoblot with the anti-Cdc42 antibody. A fraction of total lysates for each culture time point was immunoblotted to monitor the total amount of GTPase before precipitation, and the amount of activated Cdc42 was normalized to the total amount of Cdc42, for the comparison of Cdc42 activity (level of GTP-bound Cdc42) in different cell lysates. GTPγS and GDP treatment served as positive and negative controls, respectively, in the pull-down assay. As expected, Cdc42 did not pull down in the presence of GDP, confirming the specificity of the pull-down of Cdc42. The data represent the average of the results in two independent experiments. (B) CECs were transiently transfected with caCdc42G12V or dnCdc42T17N. Transfectants were cultured in DMEM-15, with or without FGF-2 for the indicated time and then fixed with paraformaldehyde. Cells were labeled with F-actin (rhodamine-phalloidin) and HA (FITC) antibody, respectively. Arrowheads: the region of protrusive structures. The data are representative of results in four experiments.
Figure 6.
 
Physical association of Cdc42 with Rac, but not with Rho. CECs were cultured in DMEM-15, with or without FGF-2, for 24 hours and then lysed. Immunoprecipitation and immunoblot analysis were performed with the designated antibody, as described in Figure 4 . The data are representative of results in four independent experiments.
Figure 6.
 
Physical association of Cdc42 with Rac, but not with Rho. CECs were cultured in DMEM-15, with or without FGF-2, for 24 hours and then lysed. Immunoprecipitation and immunoblot analysis were performed with the designated antibody, as described in Figure 4 . The data are representative of results in four independent experiments.
Figure 7.
 
Physical interaction between Rho GTPases and guanine nucleotide exchange factor. (A) First-passage cells were cultured in DMEM-15, with or without FGF-2 for 24 hours and then lysed. Cell lysates were immunoprecipitated with anti-Vav2 DH/PH antibody and then equally loaded on SDS-PAGE. Precipitated endogenous DH/PH domain-containing proteins were probed with anti-Rho (top), Rac1 (middle), and Cdc42 (bottom) antibodies. (B) CECs were maintained in DMEM-15 with FGF-2 for 24 hours and then lysed. Cell lysates were divided into three equal parts and immunoprecipitated with anti-Cdc42, Rac1 and Rho antibody, respectively. Immunoprecipitated complexes were then immunoblotted with anti-Vav2 DH/PH antibody. The data are representative of results in three independent experiments.
Figure 7.
 
Physical interaction between Rho GTPases and guanine nucleotide exchange factor. (A) First-passage cells were cultured in DMEM-15, with or without FGF-2 for 24 hours and then lysed. Cell lysates were immunoprecipitated with anti-Vav2 DH/PH antibody and then equally loaded on SDS-PAGE. Precipitated endogenous DH/PH domain-containing proteins were probed with anti-Rho (top), Rac1 (middle), and Cdc42 (bottom) antibodies. (B) CECs were maintained in DMEM-15 with FGF-2 for 24 hours and then lysed. Cell lysates were divided into three equal parts and immunoprecipitated with anti-Cdc42, Rac1 and Rho antibody, respectively. Immunoprecipitated complexes were then immunoblotted with anti-Vav2 DH/PH antibody. The data are representative of results in three independent experiments.
Figure 8.
 
Induction of sequential actin reorganization by FGF-2. First-passage cells were plated in a four-well chamber and maintained in DMEM-15 with or without FGF-2 for the indicated time. Cells were then fixed and stained for vinculin (FITC) and F-actin (rhodamine-phalloidin). The data are representative results in three experiments.
Figure 8.
 
Induction of sequential actin reorganization by FGF-2. First-passage cells were plated in a four-well chamber and maintained in DMEM-15 with or without FGF-2 for the indicated time. Cells were then fixed and stained for vinculin (FITC) and F-actin (rhodamine-phalloidin). The data are representative results in three experiments.
Figure 9.
 
The FGF-2-derived pathway affects cytoskeletal reorganization and EMT. The diagram illustrates the signal-transduction pathways via PI 3-kinase activated by FGF-2 that affect organization of actin cytoskeleton. PI 3-kinase activates both Rac and Cdc42. Activated Rac further antagonizes Rho, which promotes cortical actin formation, whereas activated Cdc42 also antagonizes Rho, which promotes protrusive structure formation. Both reorganization of actin cytoskeleton at the cortex and acquisition of protrusive processes are nedessary for the EMT process. The nature of the cross-reactivity between Rac and Cdc42 is unknown.
Figure 9.
 
The FGF-2-derived pathway affects cytoskeletal reorganization and EMT. The diagram illustrates the signal-transduction pathways via PI 3-kinase activated by FGF-2 that affect organization of actin cytoskeleton. PI 3-kinase activates both Rac and Cdc42. Activated Rac further antagonizes Rho, which promotes cortical actin formation, whereas activated Cdc42 also antagonizes Rho, which promotes protrusive structure formation. Both reorganization of actin cytoskeleton at the cortex and acquisition of protrusive processes are nedessary for the EMT process. The nature of the cross-reactivity between Rac and Cdc42 is unknown.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×