Regulation of p21-activated Kinase-independent Rac1 Signal Transduction by Nischarin*

Nischarin regulates Rac1-dependent cell motility by interaction with and inhibition of the p21-activated kinase (PAK1). In addition to regulating the activation of PAK1, Rac1 controls multiple downstream pathways to regulate cell growth and differentiation, as well as cell motility. Signaling by a constitutively activated Rac1 mutant deficient in PAK binding (Rac1Q61L-40C) was examined to determine whether Nischarin impinges on these other Rac1 effector pathways. Nischarin formed immunoprecipitatable complexes with Rac1Q61L and Rac1Q61L-40C when the proteins were co-expressed. In NIH3T3 cells, Rac1Q61L and Rac1Q61L-40C stimulation of a minimal NF- (cid:1) B response element or the cyclin D1 promoter, a downstream target of NF- (cid:1) B, was inhibited by co-expression of Nischarin. Addition-ally, suppression of endogenous Nischarin protein with small interfering RNA in PC12 cells enhanced Rac1Q61L and Rac1Q61L-40C activation of NF- (cid:1) B. In further support of Nischarin suppressing PAK independent Rac signaling, foci formation in monolayers of NIH3T3 cells by Rac1Q61L-40C in cooperation with c-Raf/C AAX was inhibited by the presence of Nischarin. Nischarin alters the cellular localization of

The Rho GTPases are a family of 20 proteins that act as molecular switches that assist in the transduction of signals through the interior of the cells. Activation of a variety of cell surface receptors shifts the Rho GTPases from the inactive GDP bound state to the active GTP bound state through the activity of numerous guanine nucleotide exchange factors. The active GTPase can then interact with a myriad of effectors and modulate their function. GTPase activating proteins suppress signaling by the Rho GTPases by stimulating the hydrolysis of GTP to GDP (1,2).
Rac1 is an important and well studied member of the Rho GTPase family of proteins. Rac1 signaling regulates numerous and disparate cellular functions including motility, cell cycle tra-verse, cytoskeletal remodeling, endocytosis, and cellular transformation (2)(3)(4)(5)(6)(7). To mediate these effects on cellular phenotype, Rac1 interacts with a variety of effector proteins. For example, Rac1 regulation of cell motility is mediated, in part, by interaction with the family of p21-activated kinases (PAK). 1 Activated PAK stimulates the formation of cortical actin networks, which are important for Rac-induced lamellipodia, by suppressing the actin-depolymerizing activity of cofilin via the activation of LIM kinase. PAK also readies the cell for movement by suppressing stress fiber formation and increasing focal adhesion turnover, in part, by disrupting the actin myosin network through inactivation of myosin light chain kinase (8). Independently from PAK, Rac1 interacts with the dormant, multiprotein WAVE complex through PIR121 and Nap125, to stimulate the dissociation of WAVE and allowing it to activate the actin polymerizing Arp2/3 complex. Rac has also been found to complex with WAVE through IRSp53 to modulate WAVE function (9,10). Similarly, Rac1 can regulate proliferative signals through various means such as activation of PAK and its subsequent phosphorylation of Raf/mitogen-activated protein kinase components or via other mechanisms, such as stimulation of NF-B through the NF-B inducing kinase (11)(12)(13)(14)(15)(16).
Nischarin was cloned from a mouse embryonic cDNA library using the cytoplasmic domain of the integrin ␣5 subunit as "bait" (17). Nischarin interacts directly with ␣5 through membrane proximal residues of the ␣5 cytoplasmic domain (18). Elevated levels of intracellular Nischarin impede ␣5-dependent cell motility in fibroblast and epithelial cells (17). Overexpression of Nischarin inhibits Rac1-stimulated cell motility and this directly correlates with its ability to inhibit signaling via the Rac1 effector PAK1 (18,19). Nischarin preferentially binds activated PAK1 and forms a complex with PAK1 via its N terminus and the kinase domain of PAK. The association of Nischarin with PAK1 inhibits PAK1 activation and, thus, PAK1-stimulated migration (20).
As described above, in addition to PAK, Rac1 has multiple downstream effectors and can act independently of PAK to regulate processes such as cell proliferation and transformation (2,3,6,16,21). Using a PAK binding-deficient mutant of Rac, Rac1Q61L-40C, we have found that Nischarin can interact with activated Rac independently of its interaction with PAK. This interaction is functionally important because Nischarin inhibits PAK-independent Rac1 signal transduction and cell transformation. Thus, in addition to regulating the PAK1 arm of Rac signal transduction, Nischarin may regulate Rac1 signaling via other Rac effector pathways.

MATERIALS AND METHODS
Cell Culture-COS-7 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) with 4,500 mg/liter D-glucose and sodium pyruvate supplemented with L-glutamine and 10% fetal bovine serum (Sigma). NIH3T3 cells were cultured in DMEM supplemented with L-glutamine and 10% calf serum (Colorado Serum Company, Denver, CO). PC12 cells were cultured in DMEM with 10% horse serum (GIBCO®, Invitrogen Cell Culture, Carlsbad, CA) and 5% fetal bovine serum.
Immunoprecipitation-For Nischarin/Rac1 binding, COS-7 cells were transfected with 1 g of Myc-Nischarin plasmid and 0.5 g of HA-Rac1Q61L or HA-Rac1Q61L-40C plasmids using FuGENE 6 (Roche). After 48 h, the cells were lysed in Rac lysis buffer (50 mM Tris-HCl, pH 7.5, 250 mM NaCl, 5 mM EDTA, 1 mM dithiothreitol, 1% Nonidet P-40) with protease and phosphatase inhibitors. The lysates were pre-cleared with Protein G-Sepharose® (Amersham Biosciences) and then incubated with a 1:100 dilution of anti-Myc antibody (9E10, Covance Research Products, Berkeley, CA) or anti-HA antibody (HA.11, Covance) at 4°C for 2 h. The immune complexes were precipitated with Protein G-Sepharose®, and the precipitates were resolved on a 4 -20% gradient gel. The immunoprecipitated proteins were detected with the antibodies used for the various epitope tags with ECL TM (Amersham Biosciences) using standard techniques. For the immunoprecipitation of endogenous Rac1, exponentially growing PC12 cells were lysed in the Rac lysis buffer ϩ 0.1% Triton X-100. A 1:100 dilution of an anti-Rac1 monoclonal antibody was used to immunoprecipitate Rac1 as described above (anti-Rac1, clone 23A8, Upstate Cell Signaling Solutions, Lake Placid, NY). Normal mouse IgG from Santa Cruz Biotechnology was used at an equivalent concentration in parallel as a control. The immunoprecipitated Rac1 was detected with the same antibody, and the bound Nischarin was detected with a rabbit polyclonal antibody generated against a Nischarin peptide (aa 219 -236).
Stimulated Rac/Nischarin Interaction-COS-7 cells were transfected with 2 g of Myc-Nischarin plasmid and 1 g of AU5-Rac1 using FuGENE 6. Twenty-four hours after the beginning of the transfection, the plates were washed twice with phosphate-buffered saline (PBS) and placed in serum-free DMEM for 18 h. The cells were treated with 25 ng/ml recombinant human epidermal growth factor (EGF) for 10 min (Upstate Cell Signaling Solutions). The cells were lysed in Rac lysis buffer plus 0.1% Triton X-100. Myc-Nischarin was immunoprecipitated with a 1:100 dilution of 9E10 monoclonal antibody. The blots were probed with a biotinylated, monoclonal anti-AU5 antibody (Covance) and detected with streptavidin-horseradish peroxidase conjugate and ECL TM (Amersham Biosciences).
Reporter Gene Assays-NIH3T3 cells were plated in 6-well plates at 1 ϫ 10 5 cells per well and grown overnight. The cells were transiently transfected with 3 g of plasmid for Myc-Nischarin, Myc-N terminus, Myc-C terminus, or Myc-␤-galactosidase and 1 g of HA-Rac1Q61L, HA-Rac1Q61L-40C, or pCGN plasmid, and 1 g of 3ϫNF-B-LUC or Ϫ963-CD1LUC plasmid with SuperFect TM (Qiagen, Valencia, CA). Twenty-four hours after transfection the cells were washed with PBS and incubated overnight in DMEM ϩ 0.5% calf serum. After starvation, the cells were washed with PBS and lysed in 150 l of 1ϫ Reporter Lysis Buffer (Promega). The luciferase activity was measured using the Luciferase Assay System from Promega on a Monolight 2010 luminometer (Analytical Luminescence Laboratory, San Diego CA).
For PC12 cells, 12-well plates were coated with 0.5 ml of 20 g/ml collagen in 30% ethanol and dried overnight. The coated plates were seeded with 4 ϫ 10 5 cells and grown overnight. The cells were transfected with Lipofectamine TM 2000 (Invitrogen) with 900 ng of plasmid for Myc-Nischarin or Myc-␤-galactosidase and 300 ng of HA-Rac1Q61L, HA-Rac1Q61L-40C, or pCGN plasmids and 300 ng of 3ϫNF-B-LUC. Twenty-four hours after transfection the cells were washed with PBS, starved in serum-free DMEM for 12-15 h, harvested, and assayed for luciferase activity.
For suppression of endogenous Nischarin protein levels by siRNA, a previously described 21-base pair siRNA oligonucleotide for rat Nischarin, rCrCUrCrGUrGrCrArCrCUUrGrArCrCUrGTT (20), or a nonspecific control duplex siRNA (#D-001206-07-20) were synthesized by Dharmacon, Lafayette, CO. The siRNA oligomers at a concentration of 150 nM were co-transfected with 500 ng of HA-Rac1Q61L, HA-Rac1Q61L-40C, or pCGN plasmid and 500 ng of 3ϫNF-B-LUC into PC12 cells with Lipofectamine TM 2000 as described above. Forty-eight hours after transfection the cells were washed with PBS, starved in serum-free DMEM for 12-15 h, harvested, and assayed for luciferase activity. Nischarin levels were determined with a monoclonal anti-Nischarin antibody (BD Biosciences). ␣-Tubulin was detected with a monoclonal anti-␣-tubulin antibody (Sigma). All of the luciferase assays were performed in triplicate, and the luciferase activity was normalized to protein levels.
Foci Forming Assays-NIH3T3 cells were plated in 60-mm dishes at 4 ϫ 10 5 cells per plate. The cells were transfected with 1 g of Rac1Q61L-40C, 100 ng of c-Raf-CAAX, both, or the pCGN vector control and 2 g of Myc-Nischarin or pcDNA using SuperFect TM . The cells were kept at confluence for 21-24 days with media changes every 3-4 days. The colonies were visualized and counted after staining with 0.1% crystal violet. The experiments were performed with triplicate or quadruplicate plates for each treatment.
Confocal Microscopy-NIH3T3 cells were plated in 6-well plates at 2 ϫ 10 5 cells per well and grown overnight. The cells were transiently transfected with 3 g of Myc-Nischarin, Myc-N terminus, Myc-C terminus, or pcDNA vectors and 1 g of HA-Rac1Q61L, HA-Rac1Q61L-40C, or pCGN with Lipofectamine TM 2000 (Invitrogen). After 24 h of transfection, the cells were trypsinized, and half of the cells were replated on fibronectin-coated coverslips (10 g/ml) in DMEM ϩ 10% calf serum, and grown overnight. The cells were washed with PBS, fixed in 3% paraformaldehyde, permeabilized in 0.5% Triton X-100, and blocked in 2% bovine serum albumin for 1 h at room temperature. The coverslips were stained with FITC-conjugated anti-Myc (9E10) and Alexa 594-conjugated anti-HA (HA.11) antibodies (Covance) for 1 h at room temperature. The coverslips were washed with PBS and rinsed with H 2 O. The stained cells were observed on an Olympus Confocal FV300 fluorescent microscope with a 60ϫ oil immersion objective; images were acquired by using Olympus Fluoview software.

Nischarin Interacts with Rac1
Independently of PAK-Nischarin directly interacts with activated PAK1 and suppresses PAK1 kinase activity. Additionally, co-expression of activated Rac1 with PAK1 and Nischarin stimulates association of PAK1 with Nischarin without preventing association of Rac1 with PAK1. All three proteins can be found in Nischarin immunoprecipitates from cells overexpressing Nischarin, PAK1, and activated Rac1 (20).
The presence of Rac1 in the Nischarin-PAK1 intracellular complex may solely depend on the association of Rac1 with PAK1. Alternatively, Rac1 could associate with Nischarin through PAK1-independent interactions. Overexpression of HA-tagged, constitutively activated Rac1 (HA-Rac1Q61L) with Myc-tagged Nischarin (Myc-Nischarin) in the absence of exogenous PAK1 in COS-7 cells allowed co-precipitation of Myc-Nischarin with the immunoprecipitated HA-Rac1Q61L (Fig. 1A). Conversely, immunoprecipitation of Myc-Nischarin also co-precipitates HA-Rac1Q61L (Fig. 1B). Nischarin does not exhibit detectable interaction with the unrelated protein, HA-ERK1, under these experimental conditions nor does HA-Rac1Q61L interact with Myc-GFP. This indicates that the Nischarin/Rac1 interaction is specific. The retention of the Rac1/Nischarin interaction in the absence of exogenously expressed PAK1 suggests that the binding of Rac1 to Nischarin may have a PAK-independent component because the levels of endogenous PAK are unlikely to be sufficient to mediate this interaction.
To directly test if the interaction of Nischarin with Rac1 required PAK1, a PAK binding-deficient mutant of Rac1, HA-Rac1Q61L-40C, was used. Several laboratories have used this Rac1 effector domain mutant to investigate PAK-independent functions of Rac1 (16,21). Myc-Nischarin and Myc-PAK1 were co-expressed in COS-7 cells with HA-Rac1Q61L or HA-Rac1Q61L-40C. Immunoprecipitation of Nischarin or PAK1 by their Myc epitopes consistently co-precipitated HA-Rac1Q61L.
To determine whether Nischarin and Rac1 can interact at the normal, endogenous levels of the proteins, PC12 cells were used because Nischarin is expressed well in PC12 cells (17,20). Lysates were made from exponentially growing cells, and Rac1 was immunoprecipitated with a Rac1-specific monoclonal antibody. The endogenous Nischarin was co-immunoprecipitated in four separate experiments (Fig. 1D). Thus, the interaction between Rac1 and Nischarin occurs at physiological levels of expression.
Nischarin interacts with ␣5 integrin and PAK1 through overlapping domains in the N terminus (aa 1-802) of the protein (20). To determine which region of Nischarin Rac1 interacts, Myc-tagged forms of the N and C terminus of Nischarin (aa 970 -1354) were co-expressed with HA-Rac1Q61L in COS-7 cells. The N terminus of Nischarin consistently co-immunoprecipitated with HA-Rac1Q61L when expressed in COS-7 cells.
The C terminus also co-immunoprecipitated with HA-Rac1Q61L, but with reduced efficiency (Fig. 2A). Immunoprecipitation of the Nischarin fragments via the Myc epitope produced a similar binding pattern (data not shown).
The HA-Rac1Q61L-40C mutant exhibited a similar co-immunoprecipitation pattern with Nischarin and its fragments as HA-Rac1Q61L (Fig. 2B). Thus, the formation of Nischarin intracellular complexes with activated Rac1 by either domain of Nischarin does not require the interaction of Rac1 with PAK and the 40C substitution does not alter the interaction of Rac1 with either domain. Additionally, the interaction of Rac1 with the C terminus of Nischarin supports the PAK-independent nature of this interaction because PAK1 exhibited no interaction with the C terminus of Nischarin (20).
Growth Factor-stimulated Interaction of Rac and Nischarin-Rac switches to its activated state in response to signals from various cell surface receptors (2,3). To test if a natural cell stimulus could stimulate the Rac1/Nischarin interaction, COS-7 cells co-expressing wild type AU5-Rac1 and Nischarin were treated with EGF. The cells were harvested after 10 min of treatment with EGF. Myc-Nischarin was immunoprecipitated and the precipitates were examined for the presence of AU5-Rac1. Rac1 was detected in the Myc-Nischarin immunoprecipitate after EGF stimulation (Fig. 3). This mimicked the stimulated binding of PAK1 to Rac1 in parallel immunoprecipitations. This EGF-stimulated interaction between Rac1 and Nischarin points to an important physiological role for this interaction.
Nischarin Regulation of PAK-independent Rac Signaling-The observation of a PAK-independent interaction between Nischarin and Rac1 suggested that Nischarin may be able to regulate Rac1 functions that are not dependent on PAK for their execution. NIH3T3 cells have been used in conjunction FIG. 1. Nischarin co-precipitates with activated Rac1. A, COS-7 cells were co-transfected with Myc-tagged Nischarin, PAK1, or GFP and HA-Rac1Q61L or HA-ERK1 expression vectors. Fortyeight hours after transfection cells were lysed in Rac lysis buffer and immunoprecipitated (IP) with a 1:50 dilution of the monoclonal HA antibody. The blot was probed with anti-Myc or anti-HA antibodies as indicated. B, activated Rac1 co-precipitates with Nischarin. COS-7 cells were co-transfected with Myc-Nischarin, Myc-PAK1, or Myc-GFP and HA-Rac1Q61L or HA-ERK1 vectors. Fortyeight hours after transfection cells were lysed in Rac lysis buffer and immunoprecipitated with a 1:100 dilution of the anti-Myc antibody. The blot was probed with anti-HA or anti-Myc antibodies as indicated. C, Nischarin co-precipitates with PAK binding-deficient HA-Rac1Q61L-40C. Myc-Nischarin or Myc-PAK1 and HA-Rac1Q61L or HA-Rac1Q61L-40C vectors were co-transfected into COS-7 cells. The cells were treated and harvested as in A, and Nischarin or PAK1 were immunoprecipitated in parallel with a 1:100 dilution of the anti-Myc antibody. The blots were probed for the Myc and HA epitopes as indicated. D, endogenous interaction of Nischarin and Rac1. Exponentially growing PC12 cells were harvested in Rac lysis buffer ϩ 0.1% Triton X-100. Rac1 was immunoprecipitated with a monoclonal Rac1 antibody from the cleared lysates. An equivalent amount of lysate was incubated with an equal concentration of normal mouse IgG in parallel. Nischarin was detected by immunoblotting with a polyclonal anti-Nischarin antibody.
with the Rac1Q61L-40C effector domain mutant to identify PAK-independent pathways downstream of Rac1 (16,21,22). One Rac1 pathway that appears to be independent of PAK in NIH3T3 cells is the NF-B pathway (12,23,24). We examined the effect of Nischarin on Rac1 stimulation of NF-B using a 3ϫNF-B response element linked to a luciferase reporter construct, 3ϫNF-B-LUC. Expression of activated Rac1Q61L or Rac1Q61L-40C in NIH3T3 cells stimulated the NF-B re-sponse element 5-and 6-fold, respectively (Fig. 4A). An activated mutant of PAK1, PAK1-T423E, which mimics the kinase domain activation loop phosphorylation, was unable to stimulate 3ϫNF-B-LUC. This underscores the PAK independence of the NF-B signaling in NIH3T3 cells. Co-expression of Nischarin with Rac1Q61L or Rac1Q61L-40C suppressed the Rac1stimulated NF-B response by 59 and 73%, respectively. Expression of the N terminus also significantly suppressed Rac1Q61L activation of NF-B by 72% and Rac1Q61L-40C stimulation by 80%. However, expression of the C terminus did not suppress the activation of NF-B by either form of activated Rac (Fig. 4A). Thus, the N terminus of Nischarin inhibits Rac1-mediated stimulation of the NF-B pathway independently of PAK function. Activated Rac1 stimulates cyclin D1 gene expression. This stimulation can occur independently of PAK activation and is mediated in part by NF-B activation (21,23,25). Co-transfection of Ϫ963-CD1LUC with Rac1Q61L or Rac1Q61L-40C, respectively, stimulated the cyclin D1 promoter by 10-and 7-fold (Fig. 4B). Expression of PAK1-T423E had little effect on the level of cyclin D1 promoter activity. Co-expression of Nischarin with Rac1Q61L suppressed cyclin D1 promoter activation by 46%. Rac1Q61L-40C-stimulated cyclin D1 promoter activation was inhibited 58% by co-transfected Nischarin. The N terminus of Nischarin suppressed Rac1Q61L stimulation of the cyclin D1 promoter by 73% and it inhibited the Rac1Q61L-40C stimulation by 78%. Co-expression of the C terminus of Nischarin with either activated Rac1 did not suppress the activation of the cyclin D1 promoter (Fig. 4B). Thus, the ability of Nischarin to inhibit the Rac1-stimulated cyclin D1 maps to the N terminus and does not require Rac1 to PAK signaling.
To determine whether the physiological role of Nischarin in Rac signal transduction is to act as a suppressor of Rac1 signaling, the effect of siRNA suppression of endogenous Nischarin on NF-B activation by Rac1 was examined in PC12 cells. Similar to the observations in NIH3T3 cells, the overexpression of Nischarin in PC12 cells with Rac1Q61L or Rac1Q61L-40C suppressed Rac1 stimulation of the NF-B pathway by 50 and 37%, respectively (Fig. 4C). Co-transfection of siRNA for Nischarin with either Rac1Q61L or Rac1Q61L-40C in PC12 cells elevated the level of Rac1 stimulation of the NF-B pathway by 2.2-fold for Rac1Q61L and 1.6-fold for Rac1Q61L-40C compared with the level of NF-B activation in cells transfected with control siRNA (Fig. 4D). This elevation of the Rac-stimu-

FIG. 4. Nischarin inhibition of Rac1-stimulated NF-B response element and cyclin D1 promoter activation. A, inhibition of
Rac1-stimulated NF-B activity by Nischarin. Nischarin, the N terminus, C terminus, or ␤-galactosidase were co-transfected with Rac1Q61L, Rac1Q61-40C, pCGN, or Pak1-T423E (Nis:Rac ϭ 3:1) and the 3ϫNF-B-LUC reporter vector into NIH3T3 cells. The cells were allowed to recover overnight in complete media and then were starved for 20 h in DMEM ϩ 0.5% calf serum. The luciferase activity was measured and normalized to protein levels. The data are from three to four experiments performed in triplicate. The normalized luciferase activities in the cells transfected with pCGN and the ␤-galactosidase plasmid were used as the control and set equal to 1. B, inhibition of Rac1-stimulated cyclin D1 promoter activity by Nischarin. Nischarin, the N terminus, C terminus, or ␤-galactosidase were expressed with Rac1Q61L, Rac1Q61-40C, pCGN, or Pak1-T423E (Nis:Rac ϭ 3:1) and a Ϫ963-CD1LUC reporter in NIH3T3 cells. The cells were allowed to recover overnight. The cells were starved for 20 h in 0.5% calf serum. The luciferase activity was determined and normalized as in A. The data are from three to five experiments performed in triplicate. C, inhibition of Rac1-stimulated NF-B activity by Nischarin in PC12 cells. Nischarin or ␤-galactosidase were co-transfected with Rac1Q61L, Rac1Q61-40C, or pCGN (Nis:Rac ϭ 3:1) and the 3ϫNF-B-LUC reporter vector into PC12 cells. The cells were allowed to recover overnight and then were starved for 12-15 h in serum-free DMEM. The luciferase activity was determined and normalized as in A. The data are from four experiments performed in triplicate. In A-C, p Ͻ 0.05, for the difference in -fold activation of Rac1Q61L or Rac1Q61-40C in the presence of ␤-galactosidase compared with the -fold activation in the presence of Nischarin or the N terminus; unpaired Student's t test. D, stimulation of Rac1-mediated NF-B activation by suppression of endogenous Nischarin levels. PC12 cells were co-transfected with Nischarin or nonspecific control siRNA and with Rac1Q61L, Rac1Q61-40C, or pCGN (Nis:Rac ϭ 3:1) and the 3ϫNF-B-LUC reporter vector. The cells were grown for 48 h post-transfection and then starved for 12-15 h in serum-free DMEM. The luciferase activity was determined as in A. The normalized luciferase activity in the presence of siRNA alone was used as the control and set equal to 1. The data are from six to seven experiments performed in triplicate (*, p Ͻ 0.05, for the -fold activation for HA-Rac1Q61L or HA-Rac1Q61L-40C in the presence of control siRNA compared with Nischarin siRNA; unpaired Student's t test). The data in A-D are presented as mean Ϯ S.E. E, suppression of Nischarin by siRNA transfection into PC12 cells. Lysates from two independent Nischarin siRNA-treated samples and two independent control siRNA-treated samples from the reporter assays in D were subjected SDS-PAGE and immunoblotted. The blots were probed for the presence of Nischarin and ␣-tubulin as indicated. lated NF-B response inversely correlated with the siRNA suppression of endogenous Nischarin (Fig. 4E). This data underscores the role of Nischarin as a physiological repressor of Rac1 signaling.
Nischarin Inhibition of Rac1-stimulated Transformation-Activated Rac can transform NIH3T3 cells in cooperation with another activated oncogene, such as c-Raf (21). This transformation of NIH3T3 cells by Rac1 can occur independently of PAK activation. Furthermore, PAK1 does not cooperate with activated Raf to transform NIH3T3 cells (21,22). To determine whether Nischarin could suppress this PAK independent function of Rac1, the formation of transformed foci in the absence and presence of Nischarin was assayed. NIH3T3 cells were co-transfected with Rac1Q61L and c-Raf-CAAX at a ratio of 10:1 and Nischarin or ␤-galactosidase expression vectors. The focus forming activity of the various combinations were assayed 22-24 days after transfection. The combination of Rac1Q61L-40C and c-Raf-CAAX increased the level of focus formation by 3.6-fold over background. The presence of Nischarin with Rac1Q61L-40C and c-Raf-CAAX suppressed their focus forming activity by 55% (Fig. 5). Thus, Nischarin can suppress Rac1 transformation in NIH3T3 cells independently from PAK function.
Co-localization of Rac1 and Nischarin in NIH3T3 Cells-The proper cellular localization of Rac1 is important for efficient transmission of its signal (26 -29). Integrin-directed translocation of activated Rac to lipid-rich rafts in the plasma membrane appears to be critical for Rac1 signaling (26,27,29). Additionally, Rac1 translocation to the endosomal compartment and the nucleus may be important in the regulation of Rac1 signaling (29 -31).
NIH3T3 cells were transfected with Myc-Nischarin and HA-Rac1Q61L or HA-Rac1Q61L-40C. Following a time course similar to the NF-B and cyclin D1 regulation experiments, the localization of Nischarin and Rac1Q61L were examined 2 days post-transfection. Cells expressing Nischarin alone exhibited two distinct patterns of staining. In the majority of the cells Nischarin displayed a diffuse cytoplasmic appearance with minor nuclear staining. In a subset of cells Nischarin displayed a vesicular staining pattern in addition to the general diffuse cytoplasmic staining. The cells expressing Nischarin generally retained their fibroblast phenotype (Fig. 6A). Expression of Rac1Q61L in NIH3T3 stimulated the classic Rac phenotype of well spread cells with numerous membrane ruffles. Rac1Q61L in NIH3T3 cells was found throughout the cytoplasm, the nucleus, some vesicles, and in membrane ruffles (Fig. 6A). Expression of both Nischarin and Rac1Q61L led to strong colocalization of the two proteins. They co-localized to numerous vesicular structures and in the cytoplasm. The co-localization did not normally extend to the plasma membrane. The Nischarin staining extended from the perinuclear region and stopped prior to membrane protrusions, whereas the Rac1Q61L staining could be found at the plasma membrane. Nischarin did not localize to the nucleus with Rac1Q61L, which displayed variable levels of nuclear staining (Fig. 6B). A subset of cells did not exhibit the vesicular structures, but still displayed the strong cytoplasmic co-localization. Cells expressing both Nischarin and Rac1Q61L displayed variable phenotypes, but they were generally intermediate between the Nischarin and Rac1Q61L alone phenotypes. The expression of Rac1Q61L-40C in NIH3T3 cells induced morphological changes similar to that of Rac1Q61L with the addition of more filopodial protrusions (Fig.  6A). Nischarin exhibited a similar pattern of co-localization with Rac1Q61L-40C as that seen with Rac1Q61L. The proteins were found to co-localize in vesicles and in the cytoplasm with the co-localization extending from the perinuclear region to just short of the plasma membrane. As with Rac1Q61L, a portion of the cells did not exhibit the vesicular structures but still displayed the strong cytoplasmic co-localization. The morphological changes induced by Rac1Q61L-40C were muted, but not eliminated, by the co-expression of Nischarin (Fig. 6C).
The localization pattern of the N and C terminus in NIH3T3 cells was examined to determine whether this varied from Nischarin itself. When expressed alone, the N terminus exhibited a staining pattern similar to that of Nischarin. All of the cells exhibited strong cytoplasmic and weak nuclear staining with a subset of the cells exhibiting the vesicular structures similar to full-length Nischarin (Fig. 6A). The C terminus showed a widespread, diffuse staining pattern with the nucleus and cytoplasm containing protein. The C terminus never exhibited the vesicular structures seen with Nischarin or its N terminus (Fig. 6A).
Co-expression of Rac1Q61L or Rac1Q61L-40C with the N terminus led to a pattern of co-localization similar to that of full-length Nischarin. Activated Rac1Q61L-40C and the N terminus displayed cytoplasmic and vesicular co-localization. The cytoplasmic co-localization did not extend to the plasma membrane; Rac1 was detected at the membrane in ruffles, but the N terminus staining stopped prior to reaching the membrane. The N terminus also did not co-localize with nuclear Rac1 (Fig.  6D, data not shown). Thus, Nischarin and its N terminus, which suppress Rac1 signaling, exhibit the same pattern of vesicular co-localization with Rac1.
Expression of the C terminus and Rac1Q61L-40C together produced a broad pattern of staining. Rac1Q61L-40C and the C terminus were seen in the nucleus and the cytoplasm, more FIG. 5. Inhibition of Rac1Q61L-40Cstimulated focus formation by Nischarin. NIH3T3 cells were transfected with 1 g of HA-Rac1Q61L-40C vector, 100 ng of c-Raf-CAAX vector, both, or vector controls (pCGN) and 2 g of Myc-␤galactosidase or Myc-Nischarin vector. The cells were kept at confluence for 22-24 days with media changes every 3-4 days. The transformed foci were visualized after staining with 0.1% crystal violet. The data are from four experiments performed with triplicate or quadruplicate plates. The data are presented as mean Ϯ S.E. The level of foci formation for the vector controls, Myc-␤-galactosidase/pCGN, was set equal to 1 (*, p Ͻ 0.05, for Nischarin/HA-Rac1Q61L-40C/c-Raf-CAAX compared with ␤-galactosidase/ HA-Rac1Q61L-40C/c-Raf-CAAX; unpaired Student's t test). strongly in the perinuclear region. Additionally, the presence of the C terminus was detected in membrane ruffles with Rac1Q61L-40C. Unlike Nischarin or the N terminus, the noninhibitory C terminus exhibited no detectable vesicular costaining with Rac1Q61L-40C in cells co-expressing these proteins. Rac1Q61L exhibited a similar pattern of staining with the C terminus when they were expressed together in NIH3T3 cells (Fig. 6E, data not shown). DISCUSSION Nischarin appears to be an important modulator of Rac1regulated signal transduction pathways. Nischarin suppresses Rac1-stimulated migration by directly interacting with the kinase domain of PAK1 and inhibiting the kinase activity of PAK1 (20). In this work, we demonstrate that Nischarin also regulates Rac1 signal transduction pathways that are PAK independent.
Nischarin interacts with Rac1 in its activated state. Both the N and C terminus of Nischarin exhibit the ability to form a complex with active Rac1. The intracellular interaction between Nischarin and Rac1 does not require the interaction of PAK with Rac1 because the PAK binding-deficient mutant, Rac1Q61L-40C, also binds to Nischarin through interactions with both the N and C termini. Activated PAK interacts with Nischarin via multiple contacts in the N terminus, but PAK makes no detectable contacts with the C terminus (20). Thus, the formation of intracellular Rac1-Nischarin complexes, although not absolutely requiring PAK, would likely be facilitated by both PAK-dependent and PAK-independent interactions with Nischarin. Furthermore, the interactions between these three proteins are not mutually exclusive because Nischarin, PAK1, and Rac1 can be co-immunoprecipitated from cells in the presence of activated Rac1, but the interaction of Nischarin with PAK1 or Rac1 does not require the binding of the other protein to Nischarin (20).
In NIH3T3 cells Rac1 has the ability to signal to multiple effector pathways that are not dependent on activation of PAK (16,21). The activation of NF-B function and the cyclin D1 promoter positively correlate with cellular transformation by Rac1 (23,32). Importantly, the activation of these pathways also appears to be important in human cancer (33). Rac1 activation of these pathways in NIH3T3 cells does not require PAK activation (21,23,24). The interaction of Nischarin with Rac1Q61L-40C suppresses the ability of Rac1 to stimulate the function of NF-B, the cyclin D1 promoter, and the formation of foci. Thus, the ability of Nischarin to suppress these pathways does not require Nischarin repression of PAK function. The enhanced Rac1 to NF-B signal after repression of endogenous Nischarin levels in PC12 cells supports the physiological importance of the inhibitory role of Nischarin in Rac signal transduction.
Integrins signal to NF-B through phosphatidylinositol 3-kinase and Rac1 (34). Rac1, as well as Cdc42 and RhoA, can stimulate the phosphorylation of IB, an inhibitor of NF-B function, leading to its degradation and NF-B activation (35,36). The phosphorylation of IB downstream of active Rac1 is mediated by the activation of the IB kinase ␤ through Rac1 activation of the NF-B inducing kinase, NIK. In NIH3T3 cells, the Rac-NIK-IB kinase ␤ pathway appears to be the primary pathway for regulating NF-B function (12). In other cell types, Rac1 can stimulate NF-B in a mitogen-activated protein kinase/extracellular signal-regulated kinase kinase kinase 1-dependent manner or by increasing the level of intracellular reactive oxygen species (37)(38)(39)(40)(41). Reactive oxygen species can stimulate IB kinase ␤ via Src activation and subsequent stimulation of protein kinase D activity via protein kinase C␦ and the Abl tyrosine kinase (42,43). Activated PAK1 can stimulate NF-B function, but it alone does not stimulate the activity of IB kinase ␣ or ␤ (12,44,45). However, activated Rac1 with a mutation of tyrosine at residue 40 was able to stimulate NF-B in NIH3T3 cells (23). In NIH3T3 cells, we observed no stimulation of NF-B by activated PAK1, whereas Rac1Q61L-40C efficiently stimulated NF-B function. Thus, disruption of the Rac1 to NF-B signal by Nischarin does not require PAK, but may involve the pathways described above.
Rac1 plays a central role in mediating adhesion-dependent signaling to cyclin D1 (25). In NIH3T3 cells, activated Rac1 stimulates cyclin D1 transcription via activation of NF-B through cooperative interactions at a NF-B and an ATF-2 binding site in the cyclin D1 promoter. An activated Rac1 with a mutated tyrosine 40 residue can also elicit this NF-B-dependent activation of the cyclin D1 promoter (23). Similarly, we observed that the Rac1Q61L-40C mutant retained the ability to stimulate the cyclin D1 promoter. Furthermore, an activated PAK1 was unable to stimulate the cyclin D1 promoter. Nischarin repressed cyclin D1 promoter stimulation by Rac1Q61L-40C, indicating that this repression is also independent of PAK. Because Nischarin represses the ability of Rac to activate NF-B function, the repression of cyclin D1 likely results from reduced NF-B activation.
The inhibition of Rac1 function maps to the N terminus of Nischarin. However, both Rac1Q61L and Rac1Q61L-40C form complexes with both the N and C terminus of Nischarin. The difference between these interactions with Rac1 may be their differential cellular localization. Nischarin and the N terminus exhibited similar patterns of co-localization with Rac1. Both were found with Rac1 in the cytoplasm and vesicles. They exhibited no co-localization with Rac1 in the nucleus. This localization was not dependent on PAK because both Rac1Q61L and RacQ61L-40C displayed a similar co-localization. The C terminus exhibited a diffuse cytoplasmic and nuclear localization when present with both forms of activated Rac1. There was not any strong vesicular co-localization with the C terminus.
The cellular localization of Rac is critical for its function. The translocation of active Rac1 to the plasma membrane facilitates the transduction of Rac1 signals to its effectors. Rho-GDI bound to Rac1 in the cytoplasm inhibits its signaling. Activated integrins direct translocation of Rac1 to cholesterol-rich, lipid raft signaling complexes at the plasma membrane where Rac1 can signal. Internalization of these Rac1 containing lipid raft signaling complexes dampen the Rac1 signal (26,27). Disruption of macropinocytotic internalization of Rac1 with a dominant negative dynamin-2 leads to aberrant localization of active Rac1 and inhibition of lamellipodia formation (29).
Additionally, movement of Rac1 to the endocytic compartment in an Arf6-dependent manner can repress the induction of membrane ruffles of Rac1 (30). Interestingly, recent investigations indicate that active Rac1 stimulates the translocation of the IB NF-B complex to the membrane. This membrane localization of IB/NF-B facilitates activation of NF-B through IB degradation (46).
The human Nischarin homolog, IRAS, has been recently demonstrated to co-localize to early/sorting and recycling endosomes and was postulated to be a sorting nexin. This localization was dependent on a Phox domain and a coiled-coil domain in the N terminus of IRAS. Elevated levels of IRAS shifted ␣5 integrin from the membrane to the endosomal compartment (47). Nischarin lacks the Phox domain, but retains a coiled-coil domain at aa 380 -405 within the N terminus (48,49). In NIH3T3 cells, this coiled-coil domain of Nischarin, and the N terminus, may be sufficient for the formation of vesicles. The co-localization of active Rac1 to these vesicles did not require Rac1 interaction with PAK because the Rac1Q61L-40C was also found to co-localize here. The vesicular co-localization may contribute to the suppression of the Rac1 signal in the presence of Nischarin or its N terminus by altering Rac1 cellular localization as described above. The absence of vesicular co-localization of the C terminus and Rac1 and the lack of an effect of the C terminus on Rac1 signaling further suggests that co-localization in these vesicular structures are important for the negative regulation of Rac1 signaling by Nischarin.
Nischarin represses Rac1 signal transduction pathways by PAK-dependent and -independent mechanisms. The ability to repress the PAK arm of the Rac1 signal transduction cascade results from the direct association of Nischarin with PAK. The repression of the PAK-independent signals may result from Nischarin altering the cellular localization of activated Rac1. Additionally, suppression by Nischarin of other Rac1 effector pathways may be mediated by the disruption of Rac1 interactions with other effector molecules or the interaction of Nischarin with and modulation of the function of other Rac1 effectors.