Identification of a novel interaction of 14-3-3 with p190RhoGEF.

Activation of Rho GTPases by guanine nucleotide exchange factors (GEFs) mediates a broad range of cytoskeletal alterations that determine cell shape. In the nervous system, Rho GTPases are essential for establishing highly asymmetrical neuronal forms and may fine-tune the shape of dendrites in differentiated neurons. p190RhoGEF is a brain-enriched, RhoA-specific GEF whose highly interactive C-terminal domain provides potential linkage to multiple pathways in the cell. In the present study, a yeast two-hybrid screen was used to identify 14-3-3eta and 14-3-3epsilon as additional binding partners of p190RhoGEF. Interactions between p190RhoGEF and 14-3-3eta were confirmed biochemically and by colocalization of the respective proteins when fused to fluorescent markers and transfected in neuronal cells. We also mapped a unique phosphorylation-independent binding site (I(1370)QAIQNL) in p190RhoGEF. Deletion of the binding site abolished interactions in vitro as well as the ability of 14-3-3eta to alter the cytoplasmic aggregation of p190RhoGEF in cotransfected cells. The findings suggest a potential role for 14-3-3 in modulating p190RhoGEF activity or in linking p190RhoGEF to the activities of other pathways in the neuron.

Activation of Rho GTPases by guanine nucleotide exchange factors (GEFs) mediates a broad range of cytoskeletal alterations that determine cell shape. In the nervous system, Rho GTPases are essential for establishing highly asymmetrical neuronal forms and may fine-tune the shape of dendrites in differentiated neurons. p190RhoGEF is a brain-enriched, RhoAspecific GEF whose highly interactive C-terminal domain provides potential linkage to multiple pathways in the cell. In the present study, a yeast two-hybrid screen was used to identify 14-3-3 and 14-3-3⑀ as additional binding partners of p190RhoGEF. Interactions between p190RhoGEF and 14-3-3 were confirmed biochemically and by colocalization of the respective proteins when fused to fluorescent markers and transfected in neuronal cells. We also mapped a unique phosphorylation-independent binding site (I 1370 QAIQNL) in p190RhoGEF. Deletion of the binding site abolished interactions in vitro as well as the ability of 14-3-3 to alter the cytoplasmic aggregation of p190RhoGEF in cotransfected cells. The findings suggest a potential role for 14-3-3 in modulating p190RhoGEF activity or in linking p190RhoGEF to the activities of other pathways in the neuron.
Rho GTPases are a subgroup of Ras GTPases and include at least 14 different members (1). They are best characterized for their role in regulating the actin cytoskeleton but also partake in other activities, such as cell proliferation, gene expression, and apoptosis (2,3). Like other small GTPases, they exhibit both GDP/GTP binding and GTPase activities, and their active states are mediated by cycling between GTP-bound active and GDP-bound inactive forms. Regulation of the cycle is achieved through the opposing effects of three different classes of protein: 1) GTPase-activating proteins, 2) guanine nucleotide exchange factors (GEFs), 1 and 3) guanine nucleotide dissociation inhibitors. Determination of GDP-bound or GTP-bound forms of Rho proteins is linked to other pathways by interactions of regulatory proteins with other components in the cell.
RhoGEFs activate Rho GTPases by catalyzing the exchange of GDP with GTP at the nucleotide binding site. More than 35 RhoGEFs have been identified (1), all containing the signature tandem array of Dbl homology and pleckstrin homology do-mains responsible for nucleotide exchange activity. RhoGEFs are large proteins with additional functional domains that modulate nucleotide exchange activities or link the RhoGEFs to other pathways in the cell (4 -8). Diverse domains on Rho-GEFs have evolved in support of varying Rho GTPase activities in different tissues and cells.
p190RhoGEF is a neuron-enriched Rho-specific GDP/GTP exchange factor that activates RhoA in vitro as well as in neuronal cells (9,10). Upstream of its Dbl homology/pleckstrin homology domain, p190RhoGEF contains a leucine-rich domain and a cysteine-rich zinc finger-like motif, which are dispensable for the catalytic activity of RhoGEF in vivo (9). In addition, p190RhoGEF contains a predicted ␣-helical coiledcoil motif in the C-terminal region of the molecule. This Cterminal region was recently shown to interact directly with microtubules (10) and c-Jun N-terminal kinase-interacting protein-1 (11) as well as with a destabilizing element in the 3Јuntranslated region of the light neurofilament (NF-L) mRNA (12). The highly interactive properties of the C-terminal region of p190RhoGEF are of particular interest in view of their potential involvement in different regulatory processes in the cell.
In this study, we have further explored the interactive properties of C-terminal domain of p190RhoGEF using a two-hybrid screen. We now report the specific binding of 14-3-3 to the C-terminal region of p190RhoGEF. We further show that binding of 14-3-3 is isoform-specific and that the binding site in p190RhoGEF does not conform to previously reported 14-3-3 binding motifs. We also demonstrate interaction between p190RhoGEF and 14-3-3 when fused to fluorescent proteins and transfected in neuronal cells. Interactions with 14-3-3 link p190RhoGEF with multiple pathways in the neuron and may be instrumental in modulating signal transduction phenomena in differentiated neurons.
Yeast Two-hybrid Analysis-Yeast two-hybrid screens were performed according to the manufacture's instructions (Stratagene). YRG-2 yeast strain expressing pBD-GAL4Cam/p190RhoGEF-C was transformed with pAD-GAL4Cam containing a murine embryonic day 12.5 spinal cord cDNA library (Stratagene). Transformants (1.4 ϫ 10 7 ) were selected on synthetic dropout media minus leucine, tryptophan, and histidine and analyzed for ␤-galactosidase activity. cDNAs were rescued from positive yeast clones using XL1-Blue MRFЈ strain of Escherichia coli. Specificity of protein-protein interactions was assessed by co-transforming selected clones with a pBD-GAL4Cam vector lacking the cDNA insert or pBD-GAL4Cam containing a lamin C cDNA (pBD-GAL4Cam/laminC) or p190RhoGEF-C cDNA (pBD-GAL4Cam/ p190RhoGEF-C). Sequences of positive cDNAs in the prey vector were compared with the BLAST data base (www.ncbi.nlm.nih.gov/blast). Protein-protein interactions were also evaluated by quantifying ␤-galactosidase activity in liquid cultures using o-nitrophenyl-␤-galactopyranoside (Sigma).
Cell Culture and Transfection-Neuro 2a and PC12 cells were obtained from ATCC and maintained in Dulbecco's modified Eagle's medium containing 10% fetal calf serum. Cell transfections were performed using the FuGene6 reagent (Roche Molecular Biochemicals).
Immunoprecipitation Assays-Immunoprecipitation was conducted on lysates of Neuro 2a cells transfected with expression vector containing HA-tagged p190RhoGEF-C. Cells were lysed with cell lysis buffer containing 1% Triton X-100, and lysates were cleared by centrifugation, normalized for protein content, mixed with 4 g of anti-14-3-3␤ (an antibody that reacts to all 14-3-3 isoforms) (Santa Cruz) and protein-G-Sepharose (Amersham Pharmacia Biotech), then rocked overnight at 4°C. Immunoprecipitates were washed with cell lysis buffer and boiled in 30 l of 2ϫ SDS loading buffer. Samples were subjected to SDSpolyacrylamide gel electrophoresis analysis and electrotransferred onto polyvinylidene difluoride membranes (Millipore). Membranes were immuno-probed with 1:1000 dilution of anti-HA peroxidase-labeled antibody (Roche Molecular Biochemicals), and immuno-reactivity visualized by Lumi-Light PLUS Western blot substrate (Roche Molecular Biochemicals).
GST Pull-down Assay-In vitro association experiments were carried out with GST fusion proteins containing full-length 14-3-3 proteins. The fusion proteins were expressed in E. coli strain BL21 and purified according to the manufacturer's instruction (Amersham Pharmacia Biotech). Neuro 2a cells were transfected with wild-type or mutant forms of pHM6/p190RhoGEF-C, and cell lysates were prepared as described above. Cell lysates were incubated with 10 g of GST alone or GST fusion proteins immobilized on glutathione-Sepharose beads. Protein complexes were recovered by centrifugation, washed four times with cell lysis buffer, and analyzed by standard SDS-polyacrylamide gel electrophoresis and immunoblotting techniques as described above. For alkaline phosphatase treatment, HA-tagged p190RhoGEF-C was immunoprecipitated with anti-HA and incubated for 60 min at 37°C in 200 l of phosphatase reaction buffer containing 100 units of alkaline phosphatase (Biolabs) before GST pull-down assays. For negative control, phosphatase inhibitor (50 mM NaF/5 mM Na 3 VO 4 ) was added before phosphatase treatment.
Confocal Microscopy-Neuro 2a cells were cotransfected with pD-sRed1/14-3-3 and pEGFP/p190RhoGEF-C or pEGFP/p190RhoGEF. Transfected cells were fixed with 4% paraformaldehyde for 30 min on ice, washed, and stained with 4,6-diamidino-2-phenylindole for 20 min at room temperature. After mounting, cells were visualized using a Bio-Rad 1024 UV laser-scanning confocal microscope. For quantitation of cytoplasmic aggregation of EGFP fusion proteins, at least 500 transfected cells with and without aggregates were counted at each time point.

Binding of 14-3-3 to the C-terminal Domain of
p190RhoGEF-The C-terminal region (aa 1276 -1582) of p190RhoGEF was fused in-frame to pBD-GAL4Cam (pBD-GAL4Cam/p190RhoGEF-C) and used as bait to screen a mouse embryo spinal cord cDNA library (Fig. 1A). Transformants were selected for growth on media lacking leucine, tryptophan, and histidine. Specific protein-protein interactions were confirmed by co-transforming positive clones with empty two-hybrid bait vector or bait vector containing lamin C or p190RhoGEF-C. Six 14-3-3 clones and one 14-3-3⑀ clone were identified by screening 1.4 ϫ 10 7 clones. Yeast co-expressing 14-3-3 and p190RhoGEF-C grew on media minus leucine, tryptophan, and histidine and activated the lacZ reporter ( Biochemical interactions between p190RhoGEF and 14-3-3 were then tested by GST pull-down assay. Lysates of Neuro 2a cells transfected with HA-tagged p190RhoGEF-C were incubated with GST or GST/14-3-3 fusion proteins. p190RhoGEF-C bound to immobilized GST/14-3-3 and GST/14-3-3⑀ fusion proteins but not to GST protein when lysate proteins were eluted from immobilized GST proteins and immunoblotted with anti-HA ( Fig. 2A). The same GST pull-down assay showed that p190RhoGEF-C also bound to immobilized ␤ and ␥ but not to or isoforms of 14-3-3 (Fig. 2B).
To assess the interaction of p190RhoGEF protein with endogenous 14-3-3, lysates were obtained from cells transfected with HA-tagged p190RhoGEF-C and immunoprecipitated with anti-14-3-3. The immunoprecipitates were washed extensively and immunoblotted with anti-HA. HA-tagged p190RhoGEF-C was readily detected in immunoprecipitates from Neuro 2a and P19 cells (Fig. 2C), indicating the presence of sufficient amount and form of 14-3-3 in both neuronal cell lines for interaction with p190RhoGEF-C.
Localization of the Binding Site of 14-3-3 in the C-terminal Domain of p190RhoGEF-To map the 14-3-3 binding site in p190RhoGEF, a series of N-terminal and C-terminal truncated mutations of p190RhoGEF-C were created (Fig. 3A) and cotransformed with 14-3-3 in yeast. Colonies grown on ϪLeu/ ϪTrp plates were streaked onto ϪLeu/ϪTrp/ϪHis media. As shown in Fig. 3B, loss of growth in selection media occurred when N-terminal deletions were extended from aa 1353 (p190RhoGEF-N1) to aa 1376 (p190RhoGEF-N2), indicating that sequence between aa 1353 and 1376 contains the binding site for 14-3-3. C-terminal deletions of p190RhoGEF-C did not diminish growth on ϪLeu/ϪTrp/ϪHis media.
To localize the 14-3-3 binding sites of p190RhoGEF-C, we conducted a series of GST pull-down assays using immobilized GST/14-3-3 fusion protein to react with mutant HA-tagged p190RhoGEF-C proteins. Serine-to-alanine point mutations were created at Ser-1360 and Ser-1362 (S1360A/S1362A) or at Ser-1362 and Ser-1365 (S1362A/S1365A) to disrupt potential phosphoserine binding sites in p190RhoGEF-C. These residues were selected for mutation because the sequence between Arg-1359 to Ser-1365 (RSFSGSS) is very similar to the consensus binding motifs (RSXpSXP (pS is phosphorylated serine) and RX 1-2 SX 2-3 S) of 14-3-3 (13,14). To our surprise, we found that neither S1360A/S1362A nor S1362A/S1365A mutations altered the binding to 14-3-3 in the GST pull-down assay (Fig. 3C). We also found that extensive phosphatase treatment did not alter the binding of 14-3-3 to p190RhoGEF-C (Fig. 3D). We therefore concluded that binding of 14-3-3 to p190RhoGEF, localized to the sequence between aa 1353 and 1376, was not a phosphorylation-dependent interaction.
To further localize the 14-3-3 binding site, a series of internal deletions between aa 1353 and 1376 of p190RhoGEF-C were constructed (Fig. 3A). The GST pulldown assay showed that deleting aa 1366 -1372 diminished the binding of p190RhoGEF-C to 14-3-3, whereas deleting aa 1370 -1376 completely abolished the ability of p190RhoGEF-C to interact with 14-3-3 (Fig. 3E). Point mutations in the binding site showed that N1375A and L1376A replacements decreased the binding of 14-3-3 (Fig. 3F); however, no single amino acid mutation within this region eliminated binding of p190RhoGEF-C to 14-3-3. Our findings establish a novel 14-3-3 binding site comprised of uncharged polar and nonpolar residues (I 1370 QAIQNL) between aa 1370 -1376 of p190RhoGEF.
Co-localization of p190RhoGEF and 14-3-3 proteins in transfected neuronal cells was investigated. In vivo interactions between p190RhoGEF and 14-3-3 were tested by fusing the respective proteins to green (EGFP) and red (DsRed1) fluorescent protein markers and coexpressing the fusion proteins in transfected cells. Confocal microscopy showed that DsRed1/14-3-3 has a fine punctate distribution in cytoplasm and nucleus when transfected alone (data not shown) but becomes redistributed when cotransfected with EGFP/p190RhoGEF-C (Fig. 4, A-C) or with EGFP/p190RhoGEF (Fig. 4, D-F). Coexpression of green EGFP/p190RhoGEF-C (Fig. 4A) and red DsRed1/14-3-3 (Fig. 4B)  show sequence deleted in p190RhoGEF-C/N2, compared with p190RhoGEF-C/N1. Shown in relationship to this sequence are the positions of serine-to-alanine mutations (S1360A/S1362A and S1362A/S1365A) and the internal deletions (⌬1353-1361, ⌬1359 -1365, ⌬1366 -1372, and ⌬1370 -1376). All mutations were made in the context of p190RhoGEF-C. B, growth on ϪLeu/ ϪTrp/ϪHis media of yeast cotransformed with pAD-GAL4Cam/14-3-3 and empty pBD-GAL4Cam (Vector) (1) or p190RhoGEF-C (2), p190RhoGEF-C/N1 (3), p190RhoGEF-C/N2 (4), p190RhoGEF-C/N3 (5), p190RhoGEF-C/C1 (6), p190RhoGEF-C/C2 (7), or p190RhoGEF-C/C3 (8) in pBD-GAL4Cam. Truncation of the N terminus of 190RhoGEF-C from aa 1353 (p190RhoGEF-C/N1) to aa 1376 (p190RhoGEF-C/N2) caused a loss of growth of cotransformants. C, GST pull-down assay of lysates from Neuro 2a cells transfected with HA-tagged p190RhoGEF-C (wild type), p190RhoGEF-C (S1360A/S1362A), or p190RhoGEF-C (S1362A/S1365A). HA-tagged p190RhoGEF-C was readily detected in lysates of cells transfected with intact and mutant p190RhoGEF-C when GST/14-3-3-bound proteins in lysates were immunoblotting with anti-HA. D, treatment with calf alkaline phosphatase (CIP treatment) did not prevent the association of HA-tagged p190RhoGEF-C with 14-3-3 when cells were transfected with HA-tagged p190RhoGEF-C and anti-HA immunoprecipitates were treated with calf alkaline phosphatase and analyzed by GST pull-down assay. nuclei of cotransfected cells as well as from cells transfected with EGFP/p190RhoGEF-C or EGFP/p190RhoGEF alone (data not shown). Coexpression of green EGFP/p190RhoGEF (Fig.  4D) and red DsRed1/14-3-3 (Fig. 4E) fusion proteins led to a more diffuse cytoplasmic distribution of the fusion proteins as well as their exclusion from the nucleus. Colocalization of the fusion proteins was indicated by the alterations of their respective fluorescence signals when compared with the merged image (Fig. 4F). Differences in cytoplasmic distributions of fusion proteins reflected the respective properties of EGFP/ p190RhoGEF-C and EGFP/p190RhoGEF, namely, the self-aggregative properties of the C-terminal domain of p190RhoGEF and reduced aggregations of the C-terminal domain in the context of the full-length protein. 2 The colocalization of DsRed1/14-3-3 with both EGFP/p190RhoGEF-C and EGFP/ p190RhoGEF indicates strong interactive properties of the respective fusion proteins. Further information on the nature of the interactions in vivo were probed by examining the effects on aggregation of fusion proteins by mutating the 14-3-3 binding site in the C-terminal domain of p190RhoGEF (see below).

Coexpression of 14-3-3 and p190RhoGEF Reduces but Does Not Eliminate the Cytoplasmic Aggregation of p190RhoGEF-
The cytoplasmic aggregation of EGFP/p190RhoGEF-C is due to the highly interactive properties of the C-terminal domain of p190RhoGEF, and the reduced aggregations of EGFP/ p190RhoGEF reflect the modulating effects of the full-length sequence. 2 We used these attributes of the fusion proteins to assess the interactions between 14-3-3 and p190RhoGEF in vivo and to determine whether the interactions could be altered by mutating the 14-3-3 binding site of p190RhoGEF. Parallel transfections of Neuro 2a cells were conducted in which EGFP/ p190RhoGEF-C fusion protein with and without deletion of aa 1370 -1376 were transfected alone or cotransfected with DsRed1/14-3-3. The extent to which cotransfection of DsRed1/ 14-3-3 reduced the aggregation of EGFP/p190RhoGEF-C fu-sion proteins was quantitated over time. When transfected alone, both wild-type (EGFP/p190RhoGEF-C) and mutant (EGFP/p190RhoGEF-C(⌬1370 -1376)) fusion proteins were localized in cytoplasmic aggregates (Fig. 5, A and C). The extent of cytoplasmic aggregation was more frequently observed with the mutant fusion protein. Cotransfection with DsRed1/14-3-3 reduced the aggregation of wild-type EGFP/p190RhoGEF-C and increased the numbers of cells with diffuse cytoplasmic localization of fusion protein (compare Figs. 5, B and A). Cotransfection with DsRed1/14-3-3 did not reduce the aggregation of the mutant EGFP/p190RhoGEF-C(⌬1370 -1376) lacking the 14-3-3 binding site or increase the numbers of cells with diffusely localized fusion protein (compare Figs. 5, C and D). The ability of DsRed1/14-3-3 to reduce the aggregation of wild-type EGFP/p190RhoGEF-C fusion protein, but not mutant EGFP/p190RhoGEF-C(⌬1370 -1376) fusion protein, was a time-dependent phenomenon and became more apparent at later time points after cotransfection (Fig. 5, E and F).
A similar comparative cotransfection study was conducted using the EGFP marker protein fused to the full-length p190RhoGEF with and without deletion of aa 1370 -1376. Low levels of cytoplasmic aggregation of wild-type EGFP/ p190RhoGEF (Fig. 6A) was further reduced by cotransfecting DsRed1/14-3-3 (Fig. 6B). Also, the number of cells with diffusely distributed fusion protein was slightly increased (compare Figs. 6, A and B). The extent of cytoplasmic aggregation of mutant EGFP/p190RhoGEF lacking the 14-3-3 binding site (EGFP/p190RhoGEF(⌬1370 -1376)) was not diminished by cotransfecting DsRed1/14-3-3 (compare Figs. 6, C and D). Moreover, the cytoplasmic aggregates of full-length fusion protein were more apparent in cells transfected with the mutant fusion protein (compare Figs. 6, A/B and C/D). The effects of coexpression of DsRed1/14-3-3 on aggregation of wild-type and mutant EGFP/p190RhoGEF fusion protein have been quantitated in Figs. 6, E and F.
We interpret the results of cotransfection experiments as indicating direct interactions between 14-3-3 and the C terminus of p190RhoGEF as well as some overlap in the 14-3-3 binding site with the self-interactive site(s) in p190RhoGEF. The ability of cotransfections to alter the aggregation of wildtype but not the mutant p190RhoGEF fusion proteins indicates a direct interaction mediated by the 14-3-3 binding site. Furthermore, the enhanced aggregation of mutant fusion protein is consistent the view that these changes may reflect interactions with endogenous 14-3-3 with the same binding site. Evidence of indirect interactions of 14-3-3 and p190RhoGEF was also observed, as suggested by the persistence of DsRed1/14-3-3 fluorescence in cytoplasmic aggregates of mutant p190RhoGEF fusion proteins (data not shown). The nature of the indirect interaction was not further pursued. DISCUSSION This study has identified novel interactions between the 14-3-3 adapter protein and the C-terminal domain of p190RhoGEF. The interactions were initially observed using the yeast two-hybrid system and subsequently confirmed by in vitro studies using biochemical methods and by in vivo studies using fluorescent protein-tagged fusion proteins. Biochemical studies also characterized the specificity of p190RhoGEF-C binding to different 14-3-3 isoforms, including , ⑀, ␤, and ␥ but not or isoforms. Isoform binding specificities could modify binding affinities to a particular substrate upon dimerization of isoforms or facilitate interactions between diverse substrates upon heterodimerization of isoforms with differing binding properties (15). Such interactions may be particularly relevant in the nervous system where almost all 14-3-3 isoforms are expressed. In most cases, isoform specificities of the different 2 R. Cañ ete-Soler, unpublished data. ligands have not been examined. In the few reported instances, the patterns of binding and non-binding isoforms differ from those reported here (16 -18).
The unique pattern of 14-3-3 isoform binding to p190RhoGEF may relate to the novel 14-3-3 binding site in the C-terminal domain of p190RhoGEF. Whereas most 14-3-3 binding sites are dependent upon a specific phosphoserine residue within a serine-rich consensus binding motif (13,14), the binding site in p190RhoGEF (I 1370 QAIQNL) does not contain a target residue for phosphorylation, and its binding properties are not affected by treatment with nonspecific phosphatase. Both phosphorylation-dependent and -independent binding of ligands to 14-3-3 are believed to occur at a conserved amphipathic groove, with apposing hydrophobic and charged surfaces in all 14-3-3 isoforms (19,20). It seems likely that the polar and nonpolar residues in the p190RhoGEF binding site may be instrumental in binding to the charged and uncharged surfaces in the amphipathic groove, as disposed in the different isoforms.
Whereas deletion of the binding site in p190RhoGEF was able to abolish interaction with 14-3-3 in a GST pull-down assay and abrogated 14-3-3 effects on self-aggregation of EGFP/p190RhoGEF fusion proteins, deletion of the binding site did not prevent colocalization of DsRed1/14-3-3 fusion protein in the cytoplasmic aggregates. The latter findings could be interpreted as indicating additional interactive sites on p190RhoGEF in vivo or the presence of wild-type endogenous p190RhoGEF in the cytoplasmic aggregates. Alternatively, the colocalization of DsRed1/14-3-3 with EGFP/p190RhoGEF-C(⌬1370 -1376) could reflect the highly interactive properties of the respective fusion protein and their binding to common components in the cell. The latter interpretation would favor the presence of 14-3-3 and p190RhoGEF in a multimolecular complex and could potentiate their respective interactions with the multiple binding partners already defined.
The ability of DsRed1/14-3-3 to reduce the aggregation of EGFP/p190RhoGEF-C and EGFP/p190RhoGEF indicates that binding of 14-3-3 is able to interfere with self-aggregation sites in the C-terminal domain of p190RhoGEF. The presence of endogenous 14-3-3 may therefore serve to maintain the exchange factor in a non-aggregative state. Indeed, the self-aggregative properties of C-terminal domain of p190RhoGEF are reduced in the context of the full-length protein. In cotransfected cells, the full-length EGFP/p190RhoGEF has a diffuse cytoplasmic distribution with focal submembranous accumulations, not unlike the immunofluorescence localization of endogenous p190RhoGEF (10).
p190RhoGEF caused a redistribution of fusion proteins in the cytoplasm but did not alter the exclusion of EGFP/ p190RhoGEF-C and EGFP/p190RhoGEF from the nucleus. The findings indicate that binding to 14-3-3 does not affect sites in p190RhoGEF governing its lack of nuclear transport, as noted with other 14-3-3 ligands (21)(22)(23). This may be relevant in view of the likely considerable excess of endogenous 14-3-3, compared with p190RhoGEF, in neuronal cytoplasm (24).
Recent studies have uncovered some novel and unique features of p190RhoGEF which suggest that interactions between 14-3-3 and p190RhoGEF may have a potentially important role in modulating p190RhoGEF function in diverse neuronal pathways. First, p190RhoGEF has been identified as an RNA-binding protein that binds to a destabilizing element and regulates the stability of light neurofilament (NF-L) mRNA (12). Second, p190RhoGEF is interconnected to the c-Jun N-terminal kinase pathway by virtue of its binding to c-Jun N-terminal kinaseinteracting protein-1 (11) and by the anti-apoptotic activity conferred by transfecting EGFP/p190RhoGEF-C in neuronal cells. 2 Third, the activity of p190RhoGEF as an exchange factor is reported to require the binding of an unknown factor in vivo (10). Finally, the pattern of p190RhoGEF expression, 2 like that of c-Jun N-terminal kinase-interacting protein-1 (25) and 14-3-3 (26 -29), suggests a functional activity in differentiated neurons. Interactions between 14-3-3 and p190RhoGEF could serve to interface the varying activities of the exchange factor with other pathways in the cell. Integration of signals from diverse pathways may be important in maintaining the homeostasis of differentiated neurons. The extensive binding of 14-3-3 to multiple substrates could be viewed as a buffer-like activity for transitioning neurons to a differentiated state and limiting interactions that might be disruptive in a fully differentiated state.