Nuclear Localization of Protein Kinase U-α Is Regulated by 14-3-3*

14-3-3 proteins are intracellular, dimeric molecules that bind to and modify the activity of several signaling proteins. We used human 14-3-3ζ as a bait in the yeast two-hybrid system to screen a murine embryonic cDNA library. One interacting clone was found to encode the carboxyl terminus of a putative protein kinase. The coding sequence of the human form (protein kinase Uα, PKUα) of this protein kinase was found in GenBankTMon the basis of sequence homology. The two-hybrid clone was also highly homologous to TOUSLED, an Arabidopsis thaliana protein kinase that is required for normal flower and leaf development. PKUα has been found by coimmunoprecipitation to bind to 14-3-3ζ in vivo. Our confocal laser immunofluorescence microscopic experiments revealed that PKUα colocalizes with the cytoplasmic intermediate filament system of cultured fibroblasts in the G1 phase of the cell cycle. PKUα is found in the perinuclear area of S phase cells and in the nucleus of late G2 cells. Transfection of cells with a dominant negative form of 14-3-3η promotes the nuclear localization of PKUα. These results suggest that the subcellular localization of PKUα is regulated, at least in part, by its association with 14-3-3.

14-3-3 proteins are intracellular, acidic dimeric molecules that play a role in signal transduction pathways (1,2). They have been identified in many eukaryotic organisms, including plants and fungi, and are primarily found in the cytoplasmic compartment of eukaryotic cells. The biological function of 14-3-3 is best modeled in the budding yeast Saccharomyces cerevisiae. Certain yeast strains that lack both 14-3-3 homologues, BMH1 and BMH2, are inviable (3). Furthermore, strains that lack BMH1 and BMH2 can be partially "rescued" by overexpression of the Ras-stimulated kinase TPK1 or by overexpression of clathrin heavy chain. These results suggest that BMH proteins play a role in both the Ras pathway and the membrane sorting pathway. In Drosophila, 14-3-3 proteins positively regulate Ras signaling in R7 photoreceptor development (4,5). Genetic epistasis analyses in Drosophila suggest that 14-3-3 acts between Ras and mitogen-activated protein kinase/extracellular signal-regulated kinase kinase (4).
In vertebrate organisms, 14-3-3 proteins regulate several facets of cell physiology, including binding to and promotion of the activation of tyrosine and tryptophan hydroxylases that are important in neurotransmitter synthetic pathways (6). 14-3-3 proteins bind to the protein kinases Raf-1 (7-10), KSR-1 (11), BCR (12), and protein kinase C (13) and are thought to modulate the activity of these kinases. In the case of protein kinase C, most data demonstrate that 14-3-3 binding inhibits its activity (13). The interaction of 14-3-3 with Raf-1 is required for the Ras-dependent activation of Raf (14 -17). 14-3-3 also interacts with the cell cycle protein phosphatase Cdc25c (18) and the apoptosis-promoting protein BAD (19). These interactions may play an important role in the regulation of apoptosis and the cell cycle.
14-3-3 preferentially binds to serine-phosphorylated proteins (14, 15, 20 -22), but the biochemical significance of this is not clear, and there are several models of 14-3-3 "behavior" that are not mutually exclusive. In one, 14-3-3 binding alters the conformation of a target protein, altering its enzymatic activity. The ability of 14-3-3 to promote the activation of tyrosine and tryptophan hydroxylases in vitro supports this hypothesis (6). In another model, 14-3-3 functions as a "competitive inhibitor" that prevents the binding of other proteins to the target. This model is supported by data demonstrating that 14-3-3 binding to BAD inhibits the ability of BCL-X L to bind to BAD (19). Another possibility is that 14-3-3 is a scaffolding protein that promotes the assembly of oligomeric signaling complexes. Indeed, Raf-1 and BCR can form a complex that is mediated by 14-3-3 protein (23). A fourth possibility is that 14-3-3 is an attachable nuclear export signal that promotes the ability of binding partners to translocate out of the nucleus (24).
In an attempt to identify additional 14-3-3-binding partners, we performed a yeast two-hybrid screen with human 14-3-3 as a bait. One interacting clone was found to encode a serine/ threonine kinase, named protein kinase U-␣ (PKU␣). 1 This protein kinase is homologous to a plant protein, TOUSLED, that is required for normal flower and leaf development (25). TOUSLED is constitutively localized in the nucleus of plant cells and is thought to play a role in cell cycle regulation (26).

EXPERIMENTAL PROCEDURES
Yeast Two-hybrid Screen-Full-length human 14-3-3 was inserted into the vector pAS1 (gift of Stephen Elledge, Baylor University) as an in-frame fusion with the transactivation domain of GAL4, as described previously (7). A mouse embryonic d12.5 cDNA yeast two-hybrid library was screened (gift of Stan Hollenberg, Oregon Health Sciences University), and pAS1/14-3-3 was used as the bait. Yeast strain Y190 was cotransfected with pAS1/14-3-3, and the mouse embryonic cDNA library and yeast were plated on media lacking histidine, tryptophan, and leucine. Colonies that grew in the absence of histidine were assayed for ␤-galactosidase activity by use of 5-bromo-4-chloro-3-indolyl-␤-Dgalactopyranoside (X-gal) as a substrate. Positive colonies were plated onto media containing cycloheximide and tryptophan to expel the bait plasmid. Yeast that contained only the cDNA library plasmid were mated with yeast strain Y189 containing either pAS1/14-3-3 or pAS1/ * This work was supported by a grant from the Missouri Affiliate of the American Heart Association, by the Barnes-Jewish Hospital Foundation, and by Grant GM54670 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  lamin that was grown on plates lacking leucine and tryptophan and reassayed for ␤-galactosidase activity.
Clones that specifically interacted with 14-3-3 were rescued and sequenced. DNA sequencing was performed with a Amersham Pharmacia Biotech 377 automated sequencer. BLAST searches (National Center for Biotechnology Information) were performed by use of the DNA and putative amino acid sequences of the two-hybrid clones.
Monoclonal Antibody Generation-A peptide corresponding to a region near the amino terminus of murine PKU␣ (amino acids 670 -684, sequence AYRKEDRIDVQQLAC) was synthesized by standard Fmoc (N-(9-fluorenyl)methoxycarbonyl) chemistry. The peptide was purified by high pressure liquid chromatography, coupled to keyhole limpet hemocyanin, and injected into mice. After multiple injections, spleens were harvested, and splenic lymphocytes were isolated and fused to myeloma cells. Clonal populations of fusion cells were tested for antibody production with the antigenic peptide in enzyme-linked immunoabsorbent assay reactions. Highly concentrated antibody was obtained from murine ascites after the intraperitoneal injection of hybridoma cells.
Northern Blot Analysis-Murine premade multiple tissue and embryonic poly(A) ϩ Northern blots were obtained from CLONTECH. The murine two-hybrid clone 52b, which corresponds to the carboxyl terminus of PKU␣, and a human skeletal ␣-actin coding region cDNA (amino acids 202-374) were used to generate probes for Northern blot analysis. These probes were labeled with [␣-32 P]dCTP by use of random hexamers and the Klenow fragment of DNA polymerase 1. Blots were prehybridized for 1 h at 42°C in 50% formamide, 5ϫ Denhardt's solution, 4ϫ SSPE (0.6 M sodium chloride, 46 mM sodium phosphate, 5 mM EDTA), and 1% sodium dodecyl sulfate (SDS). Blots were washed under stringent conditions and were then visualized by autoradiography with Kodak XAR5 film. Equal loading of RNA was confirmed by ethidium bromide staining of the blots.
Transfection of Cultured Cells-NIH/3T3 fibroblasts were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. The KIAA0137 cDNA with a Myc epitope tag was inserted into pTARGET (Promega), a mammalian expression plasmid that contains a cytomegalovirus promoter. The cDNA encoding a dominant negative form of 14-3-3 (R56A and R60A) with an amino-terminal Myc epitope tag in pcDNA3.1 (Invitrogen) was a gift from Andrey Shaw (Washington University, St. Louis) (15). An amino-terminal FLAG epitope tag was added to the cDNA encoding wild type 14-3-3 (7) by PCR, and the product was inserted into pTARGET (Promega). The coding region of the FLAG epitope-tagged 14-3-3 construct was confirmed by DNA sequencing.
NIH/3T3 fibroblasts were transfected with 10 g of plasmid DNA/ 100-mm dish by use of the calcium phosphate method. Cells were maintained in nonselective medium for 2 days after transfection, treated with trypsin, and replated in selective medium containing 0.5 mg/ml geneticin. After 2-3 weeks, distinct colonies were trypsinized and transferred to multiwell plates for further propagation in the presence of selective medium.
For coimmunoprecipitation assays, protein A/G-Plus agarose (Santa Cruz Biotechnology) was used to immobilize antibody-bound proteins. Immunoprecipitates were washed with lysis buffer with added NaCl (final concentration 1 M) and analyzed by SDS-PAGE as above.
Immunofluorescence Microscopy-NIH/3T3 fibroblasts were plated on chamber slides (Nunc, Inc). After 2 days in culture, cells were fixed with 4% formaldehyde in phosphate-buffered saline (PBS), followed by permeabilization with 1% Triton X-100. Primary antibody, diluted in phosphate-buffered saline with 10 mM glycine, was incubated with fixed, permeabilized cells for 1 h at room temperature. Murine monoclonal anti-FLAG epitope antibody (Santa Cruz Biotechnology) was used at a dilution of 1:300. Murine monoclonal anti-PKU␣ antibody, rabbit polyclonal anti-14-3-3␤ (Santa Cruz Biotechnology), and murine IgG (Promega) were used at dilutions of 1:200. Murine monoclonal anti-Myc epitope antibody (Santa Cruz Biotechnology) and rabbit polyclonal anti-cyclin B1 antibody (Santa Cruz Biotechnology) were used at dilutions of 1:100. Goat polyclonal antivimentin antibody (Chemicon International Inc.) was used at a dilution of 1:40. After incubation with primary antibody, slides were washed three times with PBS. Fluorescein isothiocyanate (FITC)-conjugated anti-goat IgG secondary antibody (used at a 1:200 dilution) (Santa Cruz Biotechnology), cyanine Cy3-conjugated anti-mouse IgG secondary antibody (used at a 1:200 dilution) (Jackson Immunoresearch Laboratories), or FITC-conjugated anti-rabbit IgG secondary antibody (used at a 1:200 dilution) (Jackson Immunoresearch Laboratories) was incubated for 1 h with cells at room temperature. Slides were washed three times with PBS. Coverslips were mounted with Vectashield mounting medium (Vector Laboratories Inc.), and cells were viewed in a confocal laser microscope (MRC 1024, Bio-Rad). To check for specificity of antibody binding, cells were treated with primary or secondary antibody alone, or the PKU␣ antigenic peptide (100 M) was added to the anti-PKU␣ antibody (0.1 mg/ml). Control slides did not display significant fluorescence in any case.
Cell Synchronization-Cultured cells were treated with 5 g/ml aphidicolin, a DNA polymerase-␣ inhibitor, for 24 h to synchronize cells at G 1 /S (27,28). Aphidicolin was washed off and replaced with fresh media for 0 h (G 1 /S-enriched cells), 6 h (S phase-enriched cells), 12 h (G 2 /M phases-enriched cells), and 18 h (G 1 phase-enriched cells). Synchronized cells were processed for immunofluorescence microscopy, as described above. The cell cycle distribution of parallel NIH/3T3 cells at these time points was confirmed by propidium iodine staining and fluorescent cell sorting.

RESULTS
Interactive Cloning of PKU␣-To identify additional binding partners of 14-3-3 protein, the yeast two-hybrid system was used with 14-3-3 as bait to screen a mouse embryonic library. Approximately 2 million colonies were screened, and nine clones were found to specifically interact with 14-3-3, with lamin as a negative control. One of the interacting clones, clone 52b, encoded the carboxyl-terminal portion of a putative serine/ threonine kinase. Additional two-hybrid analysis demonstrated that clone 52b did not interact with the protein kinase Raf-1.
Clone 52b encoded a putative protein product of 94 amino acids. This clone was identical at the amino acid level to a portion of a human cDNA named protein kinase U-␣ (PKU␣) (GenBank TM accession AB004884) (Fig. 1). Clone 52b was also identical at the amino acid level to a portion of a murine cDNA named multiple testes transcript 1 (mtt1) (GenBank TM accession AF045252). The two-hybrid clone was slightly less homologous to the related human cDNAs KIAA0137 (GenBank TM accession D50927) and PKU␤ (85/94 amino acid identity) (Gen-Bank TM accession AB004885) (30). In addition, it was found to share a high degree of sequence similarity with an Arabidopsis thaliana gene, tousled (25), and also with an open reading frame in the Caenorhabditis elegans genome (C07A9.3 in chromosome III, GenBank TM accession P34314).
PKU␣ and PKU␤ each contain two potential coiled-coil domains. The kinase domains of these proteins share significant sequence similarity with protein kinase A and phosphorylase kinase. There is one potential 14-3-3-binding site in the carboxyl terminus of PKU␣ (amino acid motif RKSVSTS) and PKU␤ (amino acid motif RRSNSSG) (14,15,22).
To determine the binding site for clone 52b on 14-3-3, truncation mutant forms of 14-3-3 were generated (Fig. 2). Mutant forms of 14-3-3 were tested for their ability to bind to clone 52b by yeast two-hybrid assay (7). Clone 52b bound with greater affinity to the carboxyl-terminal half of 14-3-3 (amino acids 121-245) than to the amino-terminal region (amino acids  or to the middle portion of 14-3-3 (amino acids 78 -121) (Fig.  2). These results are consistent with previous observations of the interaction of 14-3-3 with phosphorylated tryptophan hydroxylase (31) but do not exclude the possibility that additional contact points exist between 14-3-3 and 52b. Indeed, muta-tional analysis of 14-3-3 has revealed that multiple residues in both the carboxyl-and amino-terminal portions of the protein are important for phosphoserine-mediated binding (15).
PKU␣ Expression in Embryonic and Adult Tissues-Northern blot analysis was performed to determine the gene expression pattern of PKU␣. The clone 52b oligonucleotide probe was found to cross-react with an mRNA species of 4.3 kb in murine, rat, and human tissues (data not shown). PKU␣ was found to be highly expressed in whole murine embryos throughout development (Fig. 3A). In adult murine tissues, PKU␣ was widely expressed, with the highest level of expression found in testes (Fig. 3B). Although the clone 52b probe cross-reacted with a single mRNA species of 4.3 kb in most tissues, additional bands were detected in testes, including a major band of 3.7 kb.
Analysis of PKU␣ Protein-A monoclonal antibody was generated by use of a keyhole limpet hemocyanin-coupled peptide corresponding to the carboxyl terminus of murine PKU␣ (Fig.  1). When tested by enzyme-linked immunoabsorbent assay with the immunogenic peptide, the antibody was found to be efficient for both Western blotting and immunoprecipitation. PKU␣ protein levels were examined with protein lysates generated from cultured NIH/3T3 fibroblasts and 293 cells. The anti-PKU␣ monoclonal antibody specifically recognized a single species with a relative molecular mass of approximately 88 kDa that was detected in NIH/3T3 (Fig. 4) cells. This size corresponds to the molecular mass of TOUSLED (26). Antiserum binding to the 88-kDa species was blocked by addition of the antigenic peptide (Fig. 4).
A fusion protein of GST and 14-3-3␤ was produced in bacteria and used to determine whether PKU␣ interacts with 14-3-3 in vitro. PKU␣ protein derived from NIH/3T3 cells bound to immobilized GST/14-3-3␤ fusion protein but not to GST protein alone (Fig. 5A). PKU␣ protein derived from NIH/3T3 cells also bound to immobilized GST/14-3-3 fusion protein (data not shown).
The ability of PKU␣ to associate with 14-3-3 in vivo was tested in coimmunoprecipitation experiments; protein lysates derived from subconfluent unsynchronized NIH/3T3 cells grown in the presence of 10% fetal calf serum were immunoprecipitated with anti-14-3-3. A Western blot revealed that the immunoprecipitate contained PKU␣, suggesting that PKU␣ and 14-3-3 form a complex in vivo (Fig. 5B). Subcellular Localization of PKU␣-To characterize the subcellular localization of PKU␣, confocal laser immunofluorescence microscopy was performed with the monoclonal anti-PKU␣ antibody. Previous work has established that TOUSLED kinase is constitutively localized in the nucleus of plant cells (26). In addition, other investigators have shown that PKU␤ is localized in the nucleus and, to a lesser extent, in the cytoplasm when overexpressed in COS1 cells (30). Analysis of nontransfected subconfluent NIH/3T3 cells grown in 10% fetal calf serum revealed that PKU␣, in contrast to TOUSLED, was found in the cytoplasm in a wavy network pattern, characteristic of intermediate filaments, that extended throughout the cell (Fig. 6) (32, 33). Dual fluorescence experiments revealed that vimentin, an intermediate filament protein, and PKU␣ colocalized in NIH/3T3 fibroblasts (Fig. 6).
To determine whether the localization of PKU␣ is dependent on the cell cycle state of the cell, confluent NIH/3T3 fibroblasts were synchronized by exposure to aphidicolin, which causes cells to accumulate at the G 1 /S border by inhibiting DNA polymerase-␣ activity (27,28). Subcellular localization of PKU␣ was examined by confocal laser immunofluorescence microscopy with the monoclonal anti-PKU␣ antibody at several time points after aphidicolin exposure (0, 6, 12, 18 h). These experiments demonstrated that PKU␣ was primarily localized in the cytoplasm at the G 1 /S border (0 and 18 h after release), but during S phase (6 h after release) it became perinuclear, and during late G 2 (12 h after release) it became nuclear in distribution (Fig. 7).
The subcellular localization of PKU␣ in NIH/3T3 cells that were synchronized by aphidicolin exposure was next examined in parallel with cyclin B1. Previous work has demonstrated that cyclin B1 is primarily localized in the cytoplasm of cells in the G 1 , S, and early G 2 phases of the cell cycle but that during late G 2 cyclin B1 rapidly translocates into the nucleus (34). PKU␣ and cyclinB1 were localized in the cytoplasm of cells at the G 1 /S border (0 hours after aphidicolin release) (Fig. 8). Both PKU␣ and cyclin B1 became nuclear in distribution during late G 2 (12 h after aphidicolin release) (Fig. 8).
In addition, NIH/3T3 cells were transfected with a mammalian expression plasmid encoding Myc epitope-tagged KIAA0137 that comprises amino acids 239 -787 of PKU␤ and that includes the entire kinase domain. Confocal laser immunofluorescence microscopy with a monoclonal anti-Myc epitope antibody demonstrated the KIAA0137 also translocated into the nucleus of G 2 phase cells after aphidicolin exposure (data not shown).
Nuclear extracts were obtained from DN-14-3-3-transfected NIH/3T3 cells and were analyzed by immunoblotting with the monoclonal anti-PKU␣ antibody. Transfected cells exhibited a marked increase in nuclear PKU␣ protein compared with untransfected controls (Fig. 9, C and D). Confocal laser immunofluorescence microscopy was also performed on DN-14-3-3transfected cells with the monoclonal anti-PKU␣ antibody, and this revealed a significant increase in nuclear PKU␣ protein compared with untransfected controls (Fig. 10). DN-14-3-3 localization was also examined by confocal laser immunofluorescence microscopy experiments with a monoclonal anti-Myc epitope antibody, and these revealed that DN-14-3-3 was located primarily in the cytoplasm of transfected cells (Fig. 10E). DISCUSSION 14-3-3 proteins are ubiquitously expressed intracellular dimeric proteins that regulate several aspects of cellular physi- Cultured NIH/3T3 cells were synchronized at G 1 /S by use of aphidicolin treatment. Cells were examined at various time points after release from aphidicolin exposure, fixed, and analyzed by confocal laser immunofluorescence microscopy to determine the intracellular localization of PKU␣. A, 0 h after release from aphidicolin (G 1 /S border). B, 6 h after release from aphidicolin (S phase). C, 12 h after release from aphidicolin (G 2 /M border). D, 18 h after release from aphidicolin (G 1 phase). The cell cycle distribution of parallel NIH/3T3 cells at these time points after release from aphidicolin was confirmed by propidium iodine staining and fluorescent cell sorting. ology and bind to signaling, cell cycle, cytoskeletal, and apoptotic proteins (1,2). The varied biochemical functions of 14-3-3 are dependent on binding to a partner protein; this binding may alter the enzymatic activity of the partner (e.g. tyrosine hydroxylase, protein kinase C, and Raf-1) (1, 2, 7-10), seques-ter it (e.g. BAD) (19), enhance its solubility (e.g. keratin K8) (35), link it to other signaling proteins (e.g. BCR and Raf-1) (23), or promote its nuclear export (e.g. Cdc25) (24). 14-3-3 preferentially binds to proteins that contain serine-phosphorylated residues (14, 15, 20 -22, 36, 37), a requirement that suggests that serine kinases play a critical role in the regulation of 14-3-3 binding. Indeed, in the case of the apoptosis-promoting protein BAD, the serine kinase Akt or other serine kinases may be required for BAD phosphorylation that leads to 14-3-3 binding (38). Not only do serine kinases regulate 14-3-3 binding, but it also appears that 14-3-3 regulates the activity of a variety of serine kinases, such as Raf-1 and protein kinase C.
In this work, we sought new binding partners of 14-3-3 by performing a yeast two-hybrid screen. The carboxyl-terminal portion of a murine serine/threonine kinase, named PKU␣, was found to interact with 14-3-3. PKU␣ is homologous to an A. thaliana protein, TOUSLED, that is required for normal flower and leaf development (25). The identity of the signal transduction cascade in which TOUSLED participates is unclear; there is a homologue of TOUSLED in C. elegans, but the function of the worm protein is unknown.
In this study, we documented in Northern blot experiments that the PKU␣ gene is highly expressed throughout murine embryonic development and is widely expressed in adult murine tissues. GST/14-3-3␤ and GST/14-3-3 fusion proteins were used to determine that PKU␣ binds to 14-3-3

in vitro.
Coimmunoprecipitation experiments demonstrated that PKU␣ and 14-3-3 form a complex in vivo.
PKU␣ is found in the cytoplasmic phase cells, and in the nucleus of late G 2 cells. This localization differs from that of TOUSLED protein, which is found entirely in the nuclei of plant cells at all phases of the cell cycle (26). TOUSLED lacks a putative 14-3-3-binding site and this may explain the difference in subcellular localization. In transfected COS1 cells, previous work has demonstrated that overexpressed PKU␤ is found in the nucleus with some cytoplasmic localization (30), but the subcellular localization of native PKU␤ and PKU␣ has not been previously determined.
In order to test the ability of 14-3-3 to regulate the subcellular localization of PKU␣, NIH/3T3 cells were transfected with a dominant negative form of 14-3-3 that is mutated at two arginine residues (R56A and R60A). Dominant negative forms of 14-3-3 were first identified by a genetic screen in Drosophila melanogaster, where Chang and Rubin (4) demonstrated that three missense mutant forms of Dm14-3-3⑀ inhibited wild type 14-3-3. Subsequent mutagenesis studies with human 14-3-3 and 14-3-3 established that additional mutant forms of 14-3-3, including the R56A and R60A double mutant form of 14-3-3, were potent at inhibiting the activity of wild type 14-3-3 (15,38). Previous work has demonstrated that arginine 56 and arginine 60 are located in the phosphoserine binding pocket of 14-3-3 and that mutating these residues does not inhibit the ability of 14-3-3 monomers to dimerize nor does it result in the production of an unstable protein (39). The presumed mechanism of dominant negative forms of 14-3-3 is that they form inactive heterodimers with wild type 14-3-3 proteins (40), although this remains to be proved. Our findings indicate that transfection of cultured cells with a dominant negative form of 14-3-3 promotes the nuclear localization of PKU␣, and this is consistent with the attachable nuclear export signal model of 14-3-3 action (24). However, these results do not exclude the possibility that dominant negative 14-3-3 indirectly causes PKU␣ to accumulate in the nucleus.
Recently, a leucine-rich nuclear export signal (NES)-like sequence in the fission yeast 14-3-3 protein Rad24 was described that regulates the subcellular localization of Cdc25 (24). The nuclear export factor Crm1 binds to NES-like sequences, but it has not been established whether Crm1 binds to 14-3-3 (24). The NES-like sequence in Rad24 is conserved in mammalian forms of 14-3-3, and crystallographic analysis suggests that several key residues, including leucine-216 and leucine-220 of 14-3-3, are located on one side of the amphipathic groove that binds to phosphoserine-containing peptide motifs (39). Mutation of leucine 220 of 14-3-3 to aspartic acid abrogates binding to Raf-1 kinase, and this demonstrates that residues in the NES-like sequence are important for phosphoserine motif binding (39). One hypothetical model that explains our results is that wild type 14-3-3 forms a oligomeric complex with PKU␣ and Crm1 in the nucleus of cultured cells that mediates PKU␣ export into the cytoplasm and that DN-14-3-3 forms inactive heterodimers that are unable to bind simultaneously to both PKU␣ and Crm1. Experiments are ongoing to test this model of 14-3-3-mediated nuclear export.
The intranuclear substrates of PKU␣ and TOUSLED, if any, have not been identified. The intranuclear biochemical function of TOUSLED is obscure, although its role in proliferative events in plant development suggests that it may have a cell cycle-related activity (25,26). Further studies are needed to investigate this possibility. FIG. 10. Laser confocal immunofluorescence images of NIH/ 3T3 cells transfected with dominant negative 14-3-3. A, laser confocal immunofluorescence image of untransfected NIH/3T3 cells by use of a murine monoclonal anti-PKU␣ primary antibody and a Cy3conjugated anti-murine IgG secondary antibody. No cellular staining was observed when mouse IgG was used as a primary antibody or when antigenic peptide (100 M) was added to the anti-PKU␣ monoclonal antibody (data not shown). B, laser confocal immunofluorescence image of NIH/3T3 cells transfected with wild type 14-3-3 by use of a murine monoclonal anti-PKU␣ primary antibody and a Cy3-conjugated antimurine IgG secondary antibody. C, laser confocal immunofluorescence image of NIH/3T3 cells transfected with dominant negative 14-3-3 by use of a murine monoclonal anti-PKU␣ primary antibody and a Cy3conjugated anti-murine IgG secondary antibody. D, laser confocal immunofluorescence image of untransfected NIH/3T3 cells by use of a murine monoclonal anti-Myc epitope primary antibody and a Cy3conjugated anti-murine IgG secondary antibody. E, laser confocal immunofluorescence image of NIH/3T3 cells transfected with wild type 14-3-3 by use of a murine monoclonal anti-FLAG epitope primary antibody and a Cy3-conjugated anti-murine IgG secondary antibody. F, laser confocal immunofluorescence image of NIH/3T3 cells transfected with dominant negative 14-3-3 by use of a murine monoclonal anti-Myc epitope primary antibody and a Cy3-conjugated anti-murine IgG secondary antibody.