Nuclear localization of Tob is important for regulation of its antiproliferative activity Abstracttob is a member of an antiproliferative gene family that includes btg1, pc3/tis21/btg2, pc3b, ana/btg3, and tob2. Exogenous overexpression of the family proteins suppresses cell proliferation. These proteins participate in transcriptional regulation of several genes. Here, we show that Tob is a nuclear protein that is imported into the nucleus through a nuclear localization signal (NLS)-mediated mechanism. Mutation in the NLS sequence of Tob affects its nuclear localization and impairs antiproliferative activity. Additionally, Tob contains a nuclear export signal (NES). In oncogenic ErbB2-transformed cells, nuclear export of Tob is facilitated by NES-mediated mechanism, resulting in decrease of its antiproliferative activity. These results indicate that regulation of nuclear localization of Tob is important for its antiproliferative activity. IntroductionIn higher eukaryotes, cell proliferation is regulated by extra- and intracellular signals. These signals lead to activation of cyclin-dependent protein kinases (CDKs), which are important for cell cycle progression (Morgan, 1995). Growth stimulation signals, such as growth factors, cytokines, and antigens, cause cells to progress from the quiescent state to the G1/S transition. In many cases, the receptors for growth factors are associated with protein-tyrosine kinase (PTK) activity that is intrinsic to the receptors or receptor-associated proteins. Upon receptor stimulation, PTKs are activated, and then signals generated by PTKs at the inner surface of the cell membrane activate downstream molecules such as Ras, MAP kinase, PI3-kinase, and phospholipase C-纬. Many of these signals are then transmitted to the nucleus to regulate gene expression (Marshall and Leevers, 1995).In our previous study on ErbB2 PTK signaling, we identified a growth regulatory protein, Tob, that belongs to the Tob/BTG antiproliferative protein family (Matsuda et al., 1996). This family in human consists of six members: BTG1, BTG2, PC3B, Tob, Tob2, and ANA/BTG3 (Rouault et al., 1992, 1996; Guehenneux et al., 1997; Yoshida et al., 1998; Ikematsu et al., 1999; Buanne et al., 2000). The amino-terminal 120 amino acids of these proteins show significant homology, therefore this region is designated as the Tob/BTG homology domain. Within the domain, there are two highly conserved regions, the A box and B box (Guehenneux et al., 1997; Guardavaccaro et al., 2000; Tirone, 2001), corresponding, respectively, to amino-acid residues 40鈥?8 and 86鈥?05 of Tob. All members of the Tob/BTG family suppress cell growth when overexpressed exogenously in NIH3T3 cells (Matsuda et al., 1996; Montagnoli et al., 1996; Rouault et al., 1996; Yoshida et al., 1998; Ikematsu et al., 1999; Buanne et al., 2000; Suzuki et al., 2002). BTG1 and BTG2 associate with the homeodomain-containing transcription factor Hoxb9 (Prevot et al., 2000). All Tob/BTG family proteins are associated with CCR4-associated factor 1 (Caf1), which is a component of the CCR4-NOT transcription complex (Bogdan et al., 1998; Rouault et al., 1998; Ikematsu et al., 1999; Yoshida et al., 2000). We recently reported that Tob is a negative regulator of BMP/Smad signaling in osteoblasts and that Tob regulates Smad-mediated transcription in a BMP2-dependent manner (Yoshida et al., 2000). Moreover, exogenously expressed PC3, a rat homologue of BTG2, and Tob downregulate cyclin D1 transcription, resulting in suppression of G1 progression (Guardavaccaro et al., 2000; Yoshida et al., 2003). These findings suggest that Tob/BTG family proteins interact with transcription factors directly or indirectly and that they participate in regulation of gene expression.The importance of nuclear localization of Tob/BTG proteins was recently reported. BTG1 is localized in the nucleus of confluent avian myoblast QM7 cells, whereas it is localized in the cytoplasm diffusely in proliferating QM7 cell (Marchal et al., 1995; Rodier et al., 1999). Moreover, triiodothyronine (T3) or 8-Br-cAMP treatment of QM7 cells induces nuclear accumulation of BTG1, which is accompanied by reduction of proliferation and stimulation of differentiation of QM7 cells (Marchal et al., 1993, 1995). The B box and the carboxyl-terminal domain of BTG1 (amino-acid positions 117鈥?71) are responsible for the nuclear localization of BTG1 (Rodier et al., 2001). However, it remains elusive as to how these sequences are involved in regulation of nucleus鈥揷ytoplasm trafficking of BTG1.Here, we show that Tob is a nuclear protein and that nuclear localization of Tob is regulated via the bipartite nuclear localization signal (NLS). Mutations in these sequences result in distribution of Tob in the cytoplasm and impairment of the antiproliferative activity. We also show that Tob contains a nuclear export signal (NES). An NES mutant that is defective in nuclear export retains the antiproliferative activity. Thus, nuclear localization of Tob appears to be relevant to its antiproliferative activity.ResultsNuclear localization of Tob proteinWe showed previously that Tob is associated with transcriptional regulators, such as Caf1 and Smads (Ikematsu et al., 1999; Yoshida et al., 2000). Tob regulates Smad-mediated transcription in a BMP2-dependent manner, suggesting that nuclear localization of Tob may be important for its function. To examine nuclear localization of Tob, we performed immunofluorescence staining experiments. As endogenous Tob was poorly detectable by our anti-Tob antibodies in immunostaining experiments, we analysed the subcellular distribution of Tob exogenously expressed in normal rat kidney (NRK) cells. NRK cells were infected with adenovirus containing the tob cDNA (Adex-Tob) or the 尾-galactosidase cDNA (Adex-尾-Gal), and the subcellular distribution of each protein was examined by immunostaining with anti-Tob or anti-尾-galactosidase antibody. As shown in Figure 1A, Tob was concentrated in the nucleus, whereas 尾-galactosidase (LacZ) was localized in the cytoplasm. To confirm the subcellular localization of endogenous Tob, we performed subcellular fractionation of NRK cells and detected the endogenous Tob by immunoblotting with anti-Tob antibody. Endogenous Tob was mostly present in the nuclear fraction, and was detected only at very low levels in the cytosolic fraction (Figure 1B). In monitoring the efficiency of fractionation, we detected CBP only in the nuclear fraction and Raf-1 only in the cytosolic fraction. These results indicated that Tob is localized mainly in the nucleus of NRK cells.Figure 1Tob is localized mainly in nucleus. (A) NRK cells were infected with the Adex-Tob (Tob) or Adex-尾-Gal (LacZ) virus, and the subcellular localization of each protein was examined. (B) Subcellular localization of endogenous Tob protein. Proteins from NRK cells were subfractionated, and each fraction was subjected to immunoblotting with anti-Tob, anti-CBP (marker of the nuclear fraction) and anti-Raf-1 (marker of the cytoplasmic fraction). W, whole cell lysate; N, nuclear fraction; C, cytoplasmic fraction. (C) Subcellular localization of Tob in NIH3T3 cells was examined in the absence (鈭? or presence (+) of WGA. (D) Tob contains a functional NLS. Upper: Tob NLS peptide or the mutant NLS peptide was subcloned into the expression vector between EGFP and 尾-galactosidase cDNAs (EGFP鈥揘LS鈥?i>尾-Gal). Lower: Subcellular localizations of EGFP鈥?i>尾-Gal (a), EGFP鈥揟obWT NLS鈥?i>尾-Gal (b), and EGFP鈥揟obQN NLS鈥?i>尾-Gal (c) fusion protein in NIH3T3 cellsFull size imageTo dissect the biological significance of nuclear localization of Tob, we chose NIH3T3 cells for further study. To confirm that subcellular localization of Tob in NIH3T3 cells is virtually identical to that in NRK cells, we transfected the pME鈥揟ob expression plasmid carrying the full-length tob cDNA into NIH3T3 cells. Immunostaining of the transfectants revealed that Tob was localized in the nucleus of more than 50% of cells (Figure 1C, 鈭扺GA). Thus, we assumed that subcellular localization of Tob is regulated similarly in NIH3T3 cells and in NRK cells. We then examined the effect of a nuclear transport inhibitor, wheat germ agglutinin (WGA), on localization of Tob. As shown in Figure 1C, the nuclear transport of Tob was inhibited by WGA (+WGA), indicating that nuclear localization of Tob is regulated by the classical NLS-mediated mechanism.Tob contains a functional bipartite NLSAmong the Tob/BTG family proteins, only Tob and Tob2 contain the sequence that matches the consensus sequence of bipartite NLS (Jans and Hubner, 1996) adjacent to the A box (Matsuda et al., 1996; Ikematsu et al., 1999). As nuclear localization of Tob is mediated through the classical NLS-mediated mechanism (Figure 1C), we hypothesized that the putative NLS (NLS22-39) regulates Tob nuclear localization. To confirm this, we fused the putative NLS to EGFP鈥?i>尾-Gal and examined the subcellular distribution of the fusion protein in NIH3T3 cells (Figure 1D). As the EGFP鈥?i>尾-Gal tag is too large (about 190鈥塳Da) to diffuse passively into the nucleus, the fusion can reveal whether the NLS is functional (Bear et al., 1999; Luo and Shibuya, 2001). The EGFP鈥?i>尾-Gal tag alone was localized exclusively to the cytoplasm (Figure 1D, a). In contrast, the fusion protein with the putative Tob NLS (EGFP鈥揥TNLS鈥?i>尾-Gal) showed complete nuclear localization (Figure 1D, b). We then introduced Arg to Gln substitution mutations at amino-acid positions 22鈥?4 and Lys to Asn substitution mutations at amino-acid positions 37鈥?9. Upon transfection of the resulted construct (pEGFP鈥換NNLS鈥?i>尾-Gal), the mutant protein, EGFP鈥換NNLS鈥?i>尾-Gal, localized exclusively in the cytoplasm (Figure 1D, c). These results indicate that the putative NLS of Tob is functional.To confirm that the nuclear localization of Tob protein is mediated through NLS22鈥?9, we constructed an expression plasmid carrying the full-length tob cDNA containing the QN mutation (TobQN). We also constructed two additional NLS mutants: TobRQ (Gln substituted for Arg at amino-acid positions 22鈥?4) and TobKN (Asn substituted for Lys at amino-acid positions 37鈥?9) (Figure 2A). Wild-type Tob, Tob mutants, and 尾-Gal expression plasmids were transfected into NIH3T3 cells. The representative immunostaining data are shown in Figure 2B. In 53.5% of cells, exogenously expressed wild-type Tob was concentrated in the nucleus (Figure 2B, b, and graph). TobRQ and TobKN proteins were not concentrated in the nucleus (Figure 2B, d and e); TobRQ was distributed throughout the cell, and TobKN distributed in spots at the outer periphery of nucleus in the majority of cells. TobQN was localized exclusively in cytoplasm (85.9% of TobQN expressed cells, Figure 2B, c and graph). These data indicate that the NLS at amino-acid positions 22鈥?9 of Tob is important for its nuclear localization.Figure 2Localization of Tob is regulated by the bipartite NLS sequence in the N-terminal domain. (A) Mutations in the putative NLS in the Tob expression plasmids. (B) Subcellular localization of Tob mutants in NIH3T3 cells. The cells were immunostained with anti-尾-galactosidase antibody (a) or anti-Tob monoclonal antibody (b鈥揺): (a) 尾-Gal; (b) wild-type Tob; (c) TobQN; (d) TobRQ, and (e) TobKN. Graph: Relative proportion of cells with different subcellular localizations of Tob. Immunofluorescence staining analysis of more than 1000 cells are summarized. Bars represent mean卤s.e.m. (n=3). Percentage of cells with Tob concentrated in nuclei (nucleus (N) cytoplasm (C), black bar), throughout the entire cell (N=C, shadowed bar), and enriched in the cytoplasm (N C, white bar) are shown. *P 0.01 vs WT/N and #P 0.01 vs WT/CFull size imageSubcellular localization of Tob changes depending on cell cycle progressionWe observed that Tob was present in the cytoplasm or distributed throughout the cells in 12.4. or 33.9%, respectively, of wild-type Tob expressing cells (see N C or N=C in Figure 2B graph). The data suggest that subcellular localization of Tob changes depending on cell growth states. Indeed, when cells were cultured in the serum-depleted medium for 48鈥塰 and arrested at G0 phase, Tob was mainly localized in the nucleus (Figure 3A, 0鈥塰). Until 1鈥塰 after serum-stimulation (early G1 phase), most of Tob proteins were localized in the nucleus. When the cells moved to the late S鈥塸hase (Figure 3A, 18鈥塰), more than half of Tob proteins was detected in the cytosolic fraction. These data indicated that the subcellular distribution of Tob changes during the cell cycle. The detailed mechanism that regulates subcellular localization of Tob remains to be investigated. Interestingly, when we transfected Tob expression plasmid into NIH3T3 cells transformed by activated tyrosine kinases, such as ErbB2 and v-Src, cytosolic localization of Tob was greatly increased (N=C+N C is 44.8% in NIH3T3 and 77.9% in ErbB2/NIH3T3). The representative data are shown in Figure 3B.Figure 3Subcellular localization of Tob changes depending on cell growth. (A) Subcellular distributions of endgenous Tob in NIH3T3 cells at different points of the cell cycle. NIH3T3 cells were serum-starved for 48鈥塰, and then stimulated with serum for 0, 0.25, 1 and 18鈥塰. Proteins were subfractionated as described in Figure 1B. Each fraction was subjected to immunoblotting with anti-Tob, anti-CBP and anti-Raf-1. W, whole cell lysates; N, nuclear fraction; C, cytoplasmic fraction. (B) Subcellular localization of Tob in ErbB2-transformed NIH3T3 cells. Wild-type Tob plasmid was transfected into NIH3T3 cells (a鈥揷) or ErbB2-transformed NIH3T3 cells (d鈥揻). The cells were then immunostained with anti-Tob monoclonal antibody (a and d) or anti-ErbB2 antibody (b and e); (c) is a merged image of the data from (a) and (b), and (f) is a merged image of the data from (d) and (e)Full size imageTob contains two consensus sequences of NESTo examine whether Tob is exported from the nucleus, we tested the effect of leptomycin B (LMB) (Wolff et al., 1997), which interferes with binding of the leucine-rich Rev-type NES to the export receptor exportin 1 (Fornerod et al., 1997; Fukuda et al., 1997; Ossareh-Nazari et al., 1997), on subcellular localization of Tob. pME鈥揟ob was transfected into NIH3T3 cells, and cells were treated with LMB. In comparison with untreated cells, the proportion of Tob localized to the nucleus was increased in LMB-treated cells (50.7 vs 81.2%, data not shown). These data suggest the existence of leucine-rich Rev-type NES in the amino-acid sequence of Tob. Two regions, amino-acid positions 82鈥?2 (NES82鈥?2) and 226鈥?34 (NES226鈥?34) in Tob, showed significant similarity to the previously identified NES sequences (NES consensus sequence is Lx(1鈥?)Lx(2鈥?)LxL; L is changeable with hydrophobic residues and x indicates any residues (Wen et al., 1995; Fukuda et al., 1996); Figure 4A). In addition, there is a report that the sequence in BTG1 that corresponds to the amino acids 82鈥?2 of Tob serves for its nuclear export (Rodier et al., 2001). To determine whether nuclear export of Tob is mediated through these putative NESs, we mutated the leucine residues of Tob (Leu90 and Val 92, or Leu232 and Leu234) that were thought to be critical for the NES function (Fischer et al., 1995; Gorlich and Mattaj, 1996; Nigg, 1997) (Figure 4A). The expression plasmids pME鈥揟obL90A/V92A (TobLAVA) and pME鈥揟obL232A/L234A (TobLALA), which encode the mutated proteins, were then transfected into NIH3T3 cells. Immunostaining of the transfectants with anti-Tob antibody revealed that TobLALA was not exported to the cytoplasm (Figure 4B, b and C), suggesting that NES226鈥?34 functions as leucine-rich Rev type NES. Unexpectedly, TobLAVA, like TobQN, was hardly imported to the nucleus (Figure 4B, d and C). This finding suggested that NES82鈥?2 may virtually function as NLS. Then, to assess whether the NES82鈥?2 functions as NLS, we fused it to EGFP鈥?i>尾-Gal and examined subcellular distribution of the fusion protein in NIH3T3 cells. In contrast to the WT NLS (NLS22鈥?9) fusion protein (Figure 5A, b), the NES82鈥?2 fusion protein was not transported into the nucleus (Figure 5A, c) These results suggest that NES82鈥?2 is important for nuclear localization of Tob but does not function as a bona fide NLS.Figure 4Tob contains two consensus sequences of NES. (A) Schematic representation of Tob and Tob NES mutants. Two consensus NESs are present in Tob (residues 82鈥?2 and 226鈥?34), and mutations were introduced into these sequences (TobL90A/V92A, LAVA and TobL232A/L234A, LALA). (B) Subcellular localization of the NES mutants: TobWT (a), TobLALA (b), TobQN (c), and TobLAVA (d). (C) Relative proportion of cells with specific subcellular localization patterns for Tob in NIH3T3 cells. More than 1000 cells were analysed as described in the legend to Figure 2. Bars represent means卤s.e.m. (n=3). *P 0.01 vs WT/N, **P 0.01 vs WT/N/C and #P 0.01 vs WT/C. (D)Relative proportion of cells with specific subcellular localization patterns for Tob in ErbB2-transformed NIH3T3 cells. More than 1000 cells were analysed as described in the legend to Figure 2. Bars represent means卤s.e.m. (n=3). *P 0.01 vs WT/N, **P 0.01 vs WT/N/C and #P 0.01 vs WT/CFull size imageFigure 5The function of NES82鈥?2 and NES226鈥?34. (A) NES82鈥?2 sequence cannot act as NLS. Tob NES82鈥?2 containing peptide was subcloned into the expression vector, which was described in Figure 1D and subcellular distribution of fusion proteins were examined in NIH3T3 cells: (a) EGFP鈥?i>尾-Gal, (b) EGFP鈥揟obWTNLS鈥?i>尾-Gal, and (c) EGFP鈥揟obNES82鈥?2鈥?i>尾-Gal fusion protein. (B) NES82鈥?2 or NES226鈥?34 containing sequence was subcloned into the NES activity detectable vector (pRev(1.4)鈥揋FP). (C) The subcellular distribution of exogenously expressed NES鈥揋FP fusion proteins in NIH3T3 cells, before (鈭扐ct.D) or after (+Act.D 3鈥塰) blocking nuclear import. Rev(1.4) indicates Rev(1.4)鈥揋FP (no NES) fusion protein and Rev NES, NES82鈥?2 or NES226鈥?34 is Rev(1.4)鈥揘ES鈥揋FP fusion protein. The proportion of cells with nuclear (N), nuclear+cytoplasmic (N+C) or cytoplasmic (C) NES-GFP is shown graphically. Bars represent means卤s.e.m. (n=5)Full size imageNES82鈥?2 and NES226鈥?34 possess weak NES activityWe examined whether NES82鈥?2 and NES226鈥?34, respectively, act as NES by using pRev(1.4)鈥揋FP expression vector (Figure 5B) (Henderson and Percipalle, 1997; Henderson and Eleftheriou, 2000). Rev carries both NLS and NES (Kubota et al., 1989; Meyer and Malim, 1994). Since the NES is mutated in Rev(1.4), Rev(1.4)鈥揋FP fusion protein accumulates in the nucleus (Henderson and Percipalle, 1997). Insertion of a functional NES between Rev(1.4) and GFP enables the fusion protein to be exported to cytoplasm. Moreover, the Rev NLS can be inactivated by treating cells with actinomycin D, which allows detection of the activity of very weak NESs (Henderson and Eleftheriou, 2000). Then, we fused NES82鈥?2 or NES226鈥?34 to Rev(1.4)鈥揋FP and examined the subcellular distribution of the fusion proteins. As controls, we used Rev(1.4)鈥揋FP (no NES) and Rev(1.4)鈥揜evNES鈥揋FP (wild-type Rev NES was inserted) in NIH3T3 cells. Consistent with the previous data (Henderson and Eleftheriou, 2000), we observed the nuclear accumulation of Rev(1.4)鈥揋FP and the cytoplasmic distribution of Rev(1.4)鈥揜evNES鈥揋FP in NIH3T3 cells (Figure 5C and Table 1). After 3鈥塰 treatment with actinomycin D, the subcellular distribution of Rev(1.4)鈥揋FP was not changed, while the proportion of cytoplasmic Rev(1.4)鈥揜evNES鈥揋FP was slightly increased. When Rev(1.4)鈥揘ES226鈥?34鈥揋FP was expressed in NIH3T3 cells, GFP signals were detected throughout the cell (N+C) in 18.3% of the fusion protein-expressing cells. Actinomycin D treatment raised the proportion of cytoplasmic Rev(1.4)鈥揘ES226鈥?34鈥揋FP (Figure 5C and Table 1, N+C: 49.9%). Similar results were obtained in case of the Rev(1.4)鈥揘ES82鈥?2鈥揋FP protein (Figure 5C and Table 1). According to the NES scoring system that was proposed by Henderson and Eleftheriou (2000), NES activities of NES226鈥?34 and NES82鈥?2 both correspond to 1+, whereas the Rev NES activity corresponds to 9+ in NIH3T3 cells (Figure 5C and Table 1). These results suggest that NES226鈥?34 and NES82鈥?2 contain NES activity that is as low as that of the p53 protein (Henderson and Eleftheriou, 2000), and confirm that NES82鈥?2 and NES226鈥?34 of Tob are bona fide NESs. Although NES82鈥?2 possesses weak NES activity, point mutations in this region impaired nuclear localization (Figures 4C and 5C). We assume that the Tob protein was structurally altered by the point mutations so that it simply became difficult to translocate into the nucleus. (See Discussion in detail.)Table 1 Percentage of transfected NIH3T3 cells displaying nuclear, nuclear and cytoplasmic, or exclusively cytoplasmic GFP fusion protein shown from the average of five experimentsFull size tableNuclear export of Tob is facilitated in ErbB2-transformed cellsIn ErbB2-transformed NIH3T3 cells, the proportion of Tob distributed to the cytoplasm was increased (Figure 3A). To examine whether this phenomenon is due to inhibition of NLS-mediated nuclear import or due to facilitation of NES-mediated nuclear export, we analysed the localization of TobLALA (NES226鈥?34 mutant) in ErbB2-transformed cells. Unlike WT-Tob, the proportion of TobLALA localized mostly in the nucleus in the ErbB2-transformed cells as in untransformed cells (Figure 4C and D). The data suggest that Tob is exported to the cytoplasm through NES-mediated mechanism when cell growth is extensively stimulated.Correlation of Tob nuclear localization with antiproliferative activityWe previously showed that exogenously overexpressed Tob suppresses growth of NIH3T3 cells (Matsuda et al., 1996; Suzuki et al., 2002). Here, we examined whether disruption of nuclear localization of Tob affects its antiproliferative activity. pME鈥揟ob or pME鈥揟obQN was microinjected into serum-starved NIH3T3 cells, and the number of cells that entered S鈥塸hase were counted after serum stimulation. Whereas 71.2% cells microinjected with pME鈥?i>尾-Gal incorporated BrdU, only 18.6% cells microinjected with pME鈥揟ob showed incorporation of BrdU (P 0.05; Figure 6). In contrast, the ability of the TobQN mutant to suppress cell growth was greatly impaired; 55.7% cells microinjected with the pME鈥揟obQN plasmid incorporated BrdU. The data indicated that cytoplasmic Tob is poorly antiproliferative. In addition, the cells microinjected with the pME鈥揟obLALA (NES226鈥?34 mutant) showed suppression of cell growth to an extent similar to cells expressing wild-type Tob. Importantly, the cell growth suppression by TobLAVA (NES82鈥?2 mutant), which is not imported into the nucleus, was reduced. The degree of growth suppression by TobLAVA and TobQN may be reflecting their subcellular localization; the proportion of TobLAVA distributed throughout the cell (23.8%) was larger than that of TobQN (7.5%). These data indicate that nuclear localization of Tob is important for antiproliferative activity.Figure 6Comparison of the antiproliferative activities of wild-type Tob and Tob mutants. Serum-starved NIH3T3 cells were microinjected with the indicated expression plasmids, and BrdU incorporation of each microinjected cell was examined as described in Materials and methods. Bars represent means卤s.e.m. (n=3). **P 0.01 vs 尾-Gal and ##P 0.01 vs WTFull size imageDiscussionAmong the members of the Tob/BTG family of antiproliferative protein, only Tob and Tob2 contain a putative NLS at their N-termini (Matsuda et al., 1996; Ikematsu et al., 1999). The putative NLS sequence of Tob, NLS22-39, is bipartite and similar to that of p53 (Liang and Clarke, 1999). In the present study, we show that NLS22鈥?9 is functional. NLS22鈥?9 fused 尾-Gal protein is imported into the nucleus, whereas 尾-Gal alone is not, and Tob with mutation in NLS22鈥?9 showed cytoplasmic localization (Figures 1D and 2B). Our data strongly suggest that the antiproliferative activity of Tob correlates with its nuclear localization (Figure 6). BTG1 is also imported into the nucleus, and its nuclear localization is correlated with myogenic activity in myoblasts (Rodier et al., 2001). These findings suggest that other Tob/BTG family members also function in the nucleus. Consistent with this idea, recent reports show that Tob and BTG1 associate with various transcription factors and transcriptional regulators such as Hoxb9 (Prevot et al., 2000) and Caf1 (Bogdan et al., 1998; Rouault et al., 1998; Ikematsu et al., 1999; Yoshida et al., 2000; Nakamura et al., 2004). PSORT II (http://psort.nibb.ac.jp/cgi-bin/runpsort.pl), a program that predicts subcellular localization sites of proteins on the basis of their amino-acid sequences, indicates that all the other family members show nuclear localization. Recent report suggested that the B box of BTG1 is relevant to its nuclear translocation (Rodier et al., 2001). The B box is conserved among all the Tob/BTG family proteins, suggesting the presence of common regulatory mechanism in the B box-mediated nuclear translocation. It is possible that only Tob and Tob2 among Tob/BTG family members contain a classical NLS because their molecular masses are larger than those of other family members. Adam et al (1989) reported that nuclear proteins with relative molecular masses larger 20鈥?00鈥?0鈥?00 are probably transported actively into the nucleus through the nuclear pore complex due to a specific NLS. Since the molecular masses of Tob and Tob2 are 45 and 43鈥塳Da, respectively, and those of the other family members are less than 30鈥塳Da, specific sequences for nuclear localization may be necessary for Tob and Tob2 to be efficiently transported into the nucleus.We found that Tob contains two putative NES consensus sequences (Figure 4A: NES82鈥?2 and NES226鈥?34). Mutation experiments and nuclear export assays revealed that NES226鈥?34 functions as NES in Tob protein (Figures 4C and 5C, and Table 1). Nuclear export assay suggested that NES82鈥?2 also functions as NES (Figure 5C). This result is consistent with the finding that BTG1 NES is present in the region corresponding to Tob NES82鈥?2 (Rodier et al., 2001). Unexpectedly, TobLAVA, and another NES82鈥?2 mutant, TobV82A showed subcellular localization pattern similar to the NLS mutant, TobQN (Figure 4B, C and data not shown). A possible reason for these inconsistent results might be due to the overlap of NES82鈥?2 with B box. As mentioned above, Rodier et al. (2001) showed that B box is relevant to BTG1 nuclear translocation by deletion analysis. However, direct assay of the NLS activity of NES82鈥?2 revealed that it does not function as a bona fide NLS (Figure 5A). Importantly, our unpublished results on the X-ray crystallography of the Tob protein suggest that the B box, which overlaps with residues 82鈥?2, and the surrounding sequence form a unique three-dimensional structure. It is likely that the LAVA mutation easily disrupt the structure, resulting in mislocalization of the mutant protein. Although both Tob and BTG1 contain NES, the biological significance of nuclear export of the molecules remains unclear. One possibility is that the nuclear export may be important for rapid degradation of Tob/BTG1 family proteins, whose half-life is very short (Sasajima et al., 2002).Our data suggest that subcellular localization of Tob changes depending on cell growth status because a proportion of wild-type Tob is localized in the cytoplasm in asynchronously growing cells (Figure 2B) and the proportion of cytoplasmic Tob is increased in oncogenic ErbB2-transformed NIH3T3 cells (Figure 3B). Indeed, we found that the distribution of endogenous Tob changed depending on the cell cycle (Figure 3A). Moreover, BTG1 is localized in the nuclei of confluent myoblasts and cytoplasmic BTG1 gradually increases during myotube formation and maturation (Marchal et al., 1993, 1995; Rodier et al., 1999). Taken together, these findings suggest that trafficking of Tob family proteins between the nucleus and cytoplasm may be important for regulation of cell cycle progression and differentiation.In ErbB2-transformed cells, cytoplasmic Tob is remarkably increased (Figures 3B and 4D). Despite the abundance of Tob protein in these cells, proliferation was not suppressed (Matsuda et al., 1996), probably due to both cytoplasmic localization of Tob and facilitation of Tob phosphorylation, possibly in the nucleus, by Erk1/2 that is activated downstream of the ErbB2 receptor. Therefore, the antiproliferative activity of Tob appears to be regulated by two means: phosphorylation by Erk1/2 (Suzuki et al., 2002) and nuclear export. Similar results were obtained in oncogenic-Ras- and oncogenic-Src-tranformed cells (data not shown). These data implicate that impaired nuclear localization as well as facilitated phosphorylation of Tob is involved in oncogene-induced cellular transformation and tumorigenesis.In conclusion, trafficking of Tob between the nucleus and the cytoplasm is mediated by NLS and NES, and nuclear localization of Tob is essential for its antiproliferative activity.Materials and methodsCellsNIH3T3 cells were grown in complete Dulbecco\'s modified Eagle medium (DMEM, Nissui) with 10% calf serum (CS). Rat normal kidney (NRK) cells were grown in complete DMEM with 10% fetal bovine serum (FBS). For inhibition of nuclear import by WGA (SIGMA L0636), cells were preincubated with 400鈥?i>渭g/ml WGA in the culture medium for 30鈥塵in at 37掳C before fixation. For inhibition of nuclear export by leptomycin B (LMB, kindly gift from Dr M Yoshida), cells were preincubated with 10鈥塶g/ml LMB in the culture medium for 1鈥塰 at 37掳C before fixation.Construction of plasmids and adenoviral vectorspME鈥揟ob expression plasmid was previously described (Matsuda et al., 1996). pME鈥揟obRQ encoding QQQ substituted for RRR at amino-acid position 22鈥?4 was generated by the method of Kunkel (Kunkel and Soni, 1988). Similarly, pME鈥揟obKN having NNN for KKK mutation at position 37鈥?9, pME鈥揟obQN having both RRR to QQQ mutation and KKK to NNN mutation, pME鈥揟obL90A/V92A, and pME鈥揟obL232A/L234A were generated by the same method. pEGFP鈥揥TNLS鈥?i>尾-Gal, pEGFP鈥換NNLS鈥?i>尾-Gal, and pEGFP鈥揘ES82鈥?2鈥?i>尾-Gal were constructed by inserting the synthesized oligonucleotides 5鈥?span >IndexTermAGG AGA CGT GTC AAC ATT TTT GGT GAA GAA CTT GAA AGA CTT CTT AAG AAG AAA, 5鈥?span >IndexTermCAG CAG CAG GTC AAC ATT TTT GGT GAA GAA CTT GAA AGA CTT CTT AAT AAT AAC and 5鈥?span >IndexTermGAT CCC ATT GAT GAT GTT CGT GGC AAT CTG CCA CAG GAT CTT AGT GTT TGG ATC GAC CCA into pEGFP鈥?i>尾-Gal vector (kindly gift from Dr Luo (Luo and Shibuya, 2001)). pRev(1.4)鈥揘ES226鈥?34鈥揋FP was constructed by inserting the synthesized oligonucleotide 5鈥?span >IndexTermGAT CCA GCC ATC TCT TCC TCA ATG CAC TCT CTG TAT GGG CTT GGC TTG GGT AGC CAG CAA into pRev(1.4)鈥揋FP vector (kindly gift from Dr Henderson (Henderson and Eleftheriou, 2000)). pRev(1.4)鈥揘ES82鈥?234鈥揋FP was constructed by inserting the same synthesized oligonucleotide as pEGFP鈥揘ES82鈥?2鈥?i>尾-Gal into pRev(1.4)鈥揋FP vector. To prepare an adenoviral vector for Tob expression, pCMVXwt1鈥揟ob was constructed by cloning the EcoRI fragment of Tob cDNA and blunt ended into the SwaI site of cassette cosmid pAdex1pCAw. Recombinant adenovirus containing the tob cDNA (Adex-Tob) was obtained by cotransfection pCMVXwt1-Tob and EcoT221-digested Ad5-dlX into HEK293 cells (Miyake et al., 1996). Adenovirus containing the 尾-galactosidase gene (Adex-尾-Gal), cording for the bacterial enzyme 尾-galactosidase, was prepared as a control.Transfection, adeno virus infection and microinjectionCells were transfected using the Gene POTER transfection kit (Gene Therapy Systems) with 10鈥?i>渭g of plasmid DNA per dish and cultured for 2 days. NRK cells were infected with Adex-Tob or Adex-LacZ at m.o.i. 20. Cells were lysed at 48鈥塰 after infection and the expression of each protein was detected by Western blotting as previously described (Suzuki et al., 2002). Microinjection was carried out as previously described (Ikematsu et al., 1999).AntibodiesAnti-Tob polyclonal antibody was previously described (Matsuda et al., 1996), and anti-Tob monoclonal antibody 4B1 was from IBL (Japan). Anti-尾-galactosidase polyclonal antibody was from 5 Prime 鈫?3 Prime, Inc. Anti-BrdU monoclonal antibody BU-4 was purchased from Takara (Japan), FITC-conjugated antibody raised against mouse IgG was from Sigma, FITC-conjugated antibody raised against rabbit IgG was from Southern Biotechnology Associates, Inc. and rhodamine-conjugated antibody raised against mouse IgG was from CHEMICON INTERNATIONAL INC. Anti-CBP (C-20) and anti-Raf-1 polyclonal antibodies were from Santa Cruz Biotechnology, Inc.Immunofluorescence staining and cell fractionationThe cells were seeded on coverslips 24鈥塰 before transfection, microinjection or virus infection and were immunofluorescence stained as described (Ikematsu et al., 1999). For cell fractionation, 6 脳 106 NRK cells were used. Cell fractionation was carried out as described previously (Dignam et al., 1983). ReferencesAdam SA, Lobl TJ, Mitchell MA and Gerace L . (1989). Nature, 337, 276鈥?79.Bear J, Tan W, Zolotukhin AS, Tabernero C, Hudson EA and Felber BK . (1999). Mol. Cell. Biol., 19, 6306鈥?317.Bogdan JA, Adams-Burton C, Pedicord DL, Sukovich DA, Benfield PA, Corjay MH, Stoltenborg JK and Dicker IB . (1998). Biochem. J., 336, 471鈥?81.Buanne P, Corrente G, Micheli L, Palena A, Lavia P, Spadafora C, Lakshmana MK, Rinaldi A, Banfi S, Quarto M, Bulfone A and Tirone F . (2000). Genomics, 68, 253鈥?63.Dignam JD, Lebovitz RM and Roeder RG . (1983). Nucleic Acids Res., 11, 1475鈥?489.Fischer U, Huber J, Boelens WC, Mattaj IW and Luhrmann R . (1995). Cell, 82, 475鈥?83.Fornerod M, Ohno M, Yoshida M and Mattaj IW . (1997). Cell, 90, 1051鈥?060.Fukuda M, Gotoh I, Adachi M, Gotoh Y and Nishida E . (1997). J. Biol. Chem., 272, 32642鈥?2648.Fukuda M, Gotoh I, Gotoh Y and Nishida E . (1996). J. Biol. Chem., 271, 20024鈥?0028.Gorlich D and Mattaj IW . (1996). Science, 271, 1513鈥?518.Guardavaccaro D, Corrente G, Covone F, Micheli L, D\'Agnano I, Starace G, Caruso M and Tirone F . (2000). Mol. Cell. Biol., 20, 1797鈥?815.Guehenneux F, Duret L, Callanan MB, Bouhas R, Hayette S, Berthet C, Samarut C, Rimokh R, Birot AM, Wang Q, Magaud JP and Rouault JP . (1997). Leukemia, 11, 370鈥?75.Henderson B and Eleftheriou A . (2000). Exp. Cell Res., 256, 213鈥?24.Henderson B and Percipalle P . (1997). J. Mol. Biol., 274, 693鈥?07.Ikematsu N, Yoshida Y, Kawamura-Tsuzuku J, Ohsugi M, Onda M, Hirai M, Fujimoto J and Yamamoto T . (1999). Oncogene, 18, 7432鈥?441.Jans DA and Hubner S . (1996). Physiol. Rev., 76, 651鈥?85.Kubota S, Siomi H, Satoh T, Endo S, Maki M and Hatanaka M . (1989). Biochem. Biophys. Res. Commun., 162, 963鈥?70.Kunkel TA and Soni A . (1988). J. Biol. Chem., 263, 14784鈥?4789.Liang SH and Clarke MF . (1999). J. Biol. Chem., 274, 32699鈥?2703.Luo JC and Shibuya M . (2001). Oncogene, 20, 1435鈥?444.Marchal S, Cassar-Malek I, Magaud JP, Rouault JP, Wrutniak C and Cabello G . (1995). Exp. Cell Res., 220, 1鈥?0.Marchal S, Cassar-Malek I, Pons F, Wrutniak C and Cabello G . (1993). Biol. Cell., 78, 191鈥?97.Marshall CJ and Leevers SJ . (1995). Methods Enzymol., 255, 273鈥?79.Matsuda S, Kawamura-Tsuzuku J, Ohsugi M, Yoshida M, Emi M, Nakamura Y, Onda M, Yoshida Y, Nishiyama A and Yamamoto T . (1996). Oncogene, 12, 705鈥?13.Meyer BE and Malim MH . (1994). Genes Dev., 8, 1538鈥?547.Miyake S, Makimura M, Kanegae Y, Harada S, Sato Y, Takamori K, Tokuda C and Saito I . (1996). Proc. Natl. Acad. Sci. USA, 93, 1320鈥?324.Montagnoli A, Guardavaccaro D, Starace G and Tirone F . (1996). Cell Growth Differ., 7, 1327鈥?336.Morgan DO . (1995). Nature, 374, 131鈥?34.Nakamura T, Yao R, Ogawa T, Suzuki T, Ito C, Tsunekawa N, Inoue K, Ajima R, Miyasaka T, Yoshida Y, Ogura A, Toshimori K, Noce T, Yamamoto T and Noda T . (2004). Nat. Genet., 36, 528鈥?33.Nigg EA . (1997). Nature, 386, 779鈥?87.Ossareh-Nazari B, Bachelerie F and Dargemont C . (1997). Science, 278, 141鈥?44.Prevot D, Voeltzel T, Birot AM, Morel AP, Rostan MC, Magaud JP and Corbo L . (2000). J. Biol. Chem., 275, 147鈥?53.Rodier A, Marchal-Victorion S, Rochard P, Casas F, Cassar-Malek I, Rouault JP, Magaud JP, Mason DY, Wrutniak C and Cabello G . (1999). Exp. Cell Res., 249, 337鈥?48.Rodier A, Rochard P, Berthet C, Rouault JP, Casas F, Daury L, Busson M, Magaud JP, Wrutniak-Cabello C and Cabello G . (2001). Oncogene, 20, 2691鈥?703.Rouault JP, Falette N, Guehenneux F, Guillot C, Rimokh R, Wang Q, Berthet C, Moyret-Lalle C, Savatier P, Pain B, Shaw P, Berger R, Samarut J, Magaud JP, Ozturk M, Samarut C and Puisieux A . (1996). Nat. Genet., 14, 482鈥?86.Rouault JP, Prevot D, Berthet C, Birot AM, Billaud M, Magaud JP and Corbo L . (1998). J. Biol. Chem., 273, 22563鈥?2569.Rouault JP, Rimokh R, Tessa C, Paranhos G, Ffrench M, Duret L, Garoccio M, Germain D, Samarut J and Magaud JP . (1992). EMBO J., 11, 1663鈥?670.Sasajima H, Nakagawa K and Yokosawa H . (2002). Eur. J. Biochem., 269, 3596鈥?604.Suzuki T, Kawamura-Tsuzuku J, Ajima R, Nakamura T, Yoshida Y and Yamamoto T . (2002). Genes Dev., 16, 1356鈥?370.Tirone F . (2001). J. Cell. Physiol., 187, 155鈥?65.Wen W, Meinkoth JL, Tsien RY and Taylor SS . (1995). Cell, 82, 463鈥?73.Wolff B, Sanglier JJ and Wang Y . (1997). Chem. Biol., 4, 139鈥?47.Yoshida Y, Matsuda S, Ikematsu N, Kawamura-Tsuzuku J, Inazawa J, Umemori H and Yamamoto T . (1998). Oncogene, 16, 2687鈥?693.Yoshida Y, Nakamura T, Komoda M, Satoh H, Suzuki T, Kawamura-Tsuzuku J, Miyasaka T, Hosoda-Yoshida E, Umemori H, Kunisaki RK, Tani K, Ishii S, Mori S, Suganuma M, Noda T and Yamamoto T . (2003). Genes Dev., 17, 1201鈥?206.Yoshida Y, Tanaka S, Umemori H, Minowa O, Usui M, Ikematsu N, Hosoda E, Imamura T, Kuno J, Yamashita T, Miyazono K, Noda M, Noda T and Yamamoto T . (2000). Cell, 103, 1085鈥?097.Download referencesAcknowledgementsWe thank J Inoue, Y Yamanashi, N Ikematsu, R Ajima, N Nakamura, M Watanabe, T Miyasaka and L Amy for valuable discussions. We thank Dr T Seito at IBL for his kind supply of the monoclonal anti-Tob antibody, and Dr M Yoshida for generously supplying leptomycin B. We also thank Dr JC Luo for kind gift of pEGFP鈥?i>尾-Gal vector and thank Dr BR Henderson for pRev(1.4)鈥揋FP vector.Author informationAffiliationsDivision of Oncology, Institute of Medical Science, University of Tokyo, 4-6-1 Shirokanedai, Minato-Ku, Tokyo, 108-8639, JapanJunko Kawamura-Tsuzuku,聽Toru Suzuki,聽Yutaka Yoshida聽 聽Tadashi YamamotoAuthorsJunko Kawamura-TsuzukuView author publicationsYou can also search for this author in PubMed聽Google ScholarToru SuzukiView author publicationsYou can also search for this author in PubMed聽Google ScholarYutaka YoshidaView author publicationsYou can also search for this author in PubMed聽Google ScholarTadashi YamamotoView author publicationsYou can also search for this author in PubMed聽Google ScholarCorresponding authorCorrespondence to Tadashi Yamamoto.Rights and permissionsReprints and PermissionsAbout this articleCite this articleKawamura-Tsuzuku, J., Suzuki, T., Yoshida, Y. et al. Nuclear localization of Tob is important for regulation of its antiproliferative activity. Oncogene 23, 6630鈥?638 (2004). https://doi.org/10.1038/sj.onc.1207890Download citationReceived: 03 October 2003Revised: 17 May 2004Accepted: 17 May 2004Published: 05 July 2004Issue Date: 26 August 2004DOI: https://doi.org/10.1038/sj.onc.1207890KeywordsTobNLSNESantiproliferative activitysubcellular localization Dandan Li, Li Xiao, Yuetan Ge, Yu Fu, Wenqing Zhang, Hanwei Cao, Binbin Chen, Haibin Wang, Yan-yan Zhan Tianhui Hu Molecular Cancer (2018) Alessandro Didonna, Egle Cekanaviciute, Jorge R. Oksenberg Sergio E. Baranzini Scientific Reports (2016) Mitsuro Kanda, Hisaharu Oya, Shuji Nomoto, Hideki Takami, Dai Shimizu, Ryoji Hashimoto, Satoshi Sueoka, Daisuke Kobayashi, Chie Tanaka, Suguru Yamada, Tsutomu Fujii, Goro Nakayama, Hiroyuki Sugimoto, Masahiko Koike, Michitaka Fujiwara Yasuhiro Kodera Digestive Diseases and Sciences (2015) V Coppola, M Musumeci, M Patrizii, A Cannistraci, A Addario, M Maugeri-Sacc脿, M Biffoni, F Francescangeli, M Cordenonsi, S Piccolo, L Memeo, A Pagliuca, G Muto, A Zeuner, R De Maria D Bonci Oncogene (2013)