• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of molcellbPermissionsJournals.ASM.orgJournalMCB ArticleJournal InfoAuthorsReviewers
Mol Cell Biol. Apr 2009; 29(7): 1796–1813.
Published online Jan 21, 2009. doi:  10.1128/MCB.01423-08
PMCID: PMC2655612

A Rac GTPase-Activating Protein, MgcRacGAP, Is a Nuclear Localizing Signal-Containing Nuclear Chaperone in the Activation of STAT Transcription Factors[down-pointing small open triangle]

Abstract

In addition to their pleiotropic functions under physiological conditions, transcription factors STAT3 and STAT5 also have oncogenic activities, but how activated STATs are transported to the nucleus has not been fully understood. Here we show that an MgcRacGAP mutant lacking its nuclear localizing signal (NLS) blocks nuclear translocation of p-STATs both in vitro and in vivo. Unlike wild-type MgcRacGAP, this mutant did not promote complex formation of phosphorylated STATs (p-STATs) with importin α in the presence of GTP-bound Rac1, suggesting that MgcRacGAP functions as an NLS-containing nuclear chaperone. We also demonstrate that mutants of STATs lacking the MgcRacGAP binding site (the strand βb) are hardly tyrosine phosphorylated after cytokine stimulation. Intriguingly, mutants harboring small deletions in the C′-adjacent region (βb-βc loop region) of the strand βb became constitutively active with the enhanced binding to MgcRacGAP. The molecular basis of this phenomenon will be discussed, based on the computer-assisted tertiary structure models of STAT3. Thus, MgcRacGAP functions as both a critical mediator of STAT's tyrosine phosphorylation and an NLS-containing nuclear chaperone of p-STATs.

The STAT (signal transducers and activators of transcription) family proteins (STAT1 to -4, -5A, -5B, and -6) are phosphorylated by cytokine stimulation, form homo- or heterodimers, and enter the nucleus, where they regulate expression of their target genes (6, 13). STATs have a variety of functions, including antiapoptosis, proliferation, differentiation, inflammation, and development under physiological conditions, and of note, the oncogenic activities of STAT3 and STAT5 have also been demonstrated under pathological conditions (5).

A small GTPase Rac1 is implicated in cytoskeletal organization, membrane ruffling, production of superoxide, phagocytosis, and chemotaxis as well as regulation of the cell cycle (15, 39). Recent studies revealed its distinct roles in nuclear translocation of phosphorylated STATs (p-STATs) and β-catenin and also its nuclear accumulation in the G2 phase, promoting cell division (17, 31, 47). MgcRacGAP is an evolutionarily conserved GTPase-activating protein (GAP) for Rho family GTPases. We and others previously showed that MgcRacGAP controls the mitotic spindle through associating α-, β-, and γ-tubulin, Rho family GTPases, and a kinesin protein, MKLP1, and plays essential roles in the completion of cytokinesis, accumulating to the midbody (12, 16, 32). Very recently, Yamada et al. reported that conditional knockout of MgcRacGAP results in acute apoptosis even before the failure of cytokinesis in interleukin-7 (IL-7)-expanded B220+ cells (48), indicating that MgcRacGAP is not simply involved in cell division but also in cell survival, at least in IL-7-expanded B220+ cells.

Molecular trafficking between the nucleus and cytoplasm occurs via nuclear pore complexes. To enter the nucleus, nuclear proteins larger than ~50 kDa usually harbor a functional nuclear localization signal (NLS) or bind NLS-containing chaperones. The best-characterized NLS is the mono- or bipartite polybasic NLS. Polybasic NLS-containing proteins are usually recognized by importin α/β heterodimers, importin β docks the ternary complex to the nuclear pore, and the complex migrates into the nucleus. Then, the GTP-bound form of small GTPase Ran directly binds to importin β in the complex, followed by disassembly of the complex inside the nucleus (11, 29). In addition to the polybasic NLS proline, tyrosine-containing NLSs (PY-NLSs) have been reported to be a different class of NLS; these NLSs mediate direct binding of NLS-containing proteins to karyopherin β2 (24).

How activated STATs are transported to the nucleus has been investigated; activated STAT1 and STAT3 were reported to bind importin α5 and several importin αs, respectively, which mediated the nuclear transport of STATs (25, 26, 30, 42, 44). Molecules other than importins and Ran also participate in the regulation of the nuclear translocation of STATs (27). We recently found that GTP-bound Rac1 and MgcRacGAP form a ternary complex with p-STATs and play critical roles in the nuclear translocation of p-STATs via the importin α/β pathway in an in vitro nuclear transport assay (17). However, it remained elusive how GTP-bound Rac1 and MgcRacGAP mediate the complex formation of p-STATs with importin αs. Moreover, the regulation of nuclear import of activated STATs by MgcRacGAP has not been fully demonstrated in vivo, because MgcRacGAP also plays an essential role in the completion of cytokinesis (12, 16, 33, 45). Therefore, the phenotypes observed in MgcRacGAP-depleted cells may include the effects of cytokinesis failure.

In the present work, we demonstrate that the NLS of MgcRacGAP plays essential roles in the nuclear import of p-STATs not only in vitro but also in vivo, by using MgcRacGAP conditional knockout DT40 cells expressing MgcRacGAP mutants lacking the NLS. We also found that the STAT mutants that did not bind MgcRacGAP were hardly phosphorylated on their tyrosine residues by cytokine stimulation, while the STAT mutants that preferentially bound MgcRacGAP exerted strong transcriptional activities even without cytokine stimulation. Thus, MgcRacGAP accompanied by GTP-bound Rac1 is not only an NLS-containing nuclear chaperone of p-STATs but also a critical activator of STAT proteins.

MATERIALS AND METHODS

Purification of recombinant proteins in Sf-9 cells.

Purification of recombinant proteins using Sf-9 cells was done as described previously (17). To confirm the purity, the eluted proteins were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), followed by Coomassie blue (CBB) staining or by Western blotting. Purified His-tagged L69Ran and His-tagged N24Ran were purchased from Cytoskeleton, Inc.

Import assays with permeabilized cells.

HeLa cells were grown on poly(l-lysine)-coated coverslips and permeabilized with 40 μg/ml digitonin (Roche) in transport buffer (TB; 20 mM HEPES, pH 7.3, 110 mM KO-acetate, 2 mM Mg(O-acetate)2, 1 mM EGTA, 2 mM dithiothreitol, 0.4 mM phenylmethylsulfonyl fluoride, 3 μg/ml of aprotinin, 2 μg/ml of pepstatin A, 1 μg/ml of leupeptin, 20 mg/ml of bovine serum albumin) for 10 min at room temperature. Subsequently, the cells were washed twice in TB to wash out cytoplasmic proteins. Incubation with 50 μl import mix (IM) was performed at 37°C for 30 min. The IM contained TB, an energy-regenerating system (ERS; 0.5 mM ATP, 0.5 mM GTP, 10 mM creatine phosphate, 30 U/ml creatine phosphokinase) and combinations of 1 μM of the purified proteins indicated in Fig. Fig.2B,2B, below (see also Fig. S2 and S3 in the supplemental material). Following the import reaction, the cells were washed with ice-cold TB and immunostained.

FIG. 2.
Nuclear translocation of p-STAT5A requires the NLS of MgcRacGAP. (A) The recombinant proteins used in the experiments are shown, after CBB staining of purified proteins (a) or Western blot analysis of the STAT5A-Flag protein purified from Sf-9 cells with ...

Immunostaining and antibodies.

HeLa cells were immunostained as described previously (12). Rabbit anti-STAT5A, anti-STAT3, antihemagglutinin, anti-NF-κB p65, goat anti-glutathione S-transferase (anti-GST) (Santa Cruz), mouse anti-Flag (M2; Sigma), rabbit anti-Flag (Zymed), and affinity-purified rabbit anti-MgcRasGAP (12) antibodies (Abs) were used for the first Ab. For the secondary Ab and DNA staining, fluorescein isothiocyanate- or rhodamine-conjugated goat anti-rabbit immunoglobulin (Ig; Wako), fluorescein isothiocyanate- or rhodamine-conjugated goat anti-mouse Ig (Sigma), or rhodamine-conjugated anti-goat Ig (Wako) and 4′,6-diamino-2-phenylindole (DAPI) were used.

Microscopy.

Fixed fluorescence images were analyzed on a confocal microscope (FLUOVIEW FV300 scanning laser biological microscope IX70 system; Olympus). Living cells expressing green fluorescent protein (GFP)-fusion proteins were viewed with a fluorescence microscope IX70 (Olympus) equipped with a SenSys/OL cold charge-coupled-device camera (Olympus).

In vitro binding assay, immunoprecipitation, and Western blotting.

Immunoprecipitation, gel electrophoresis, and immunoblotting were done as described previously (17), with minor modifications. Purified proteins which were preincubated in the TB with 0.5% Triton X in the in vitro binding assay or cell lysates (2 × 107 cells/ml) were incubated at 4°C for 2 h with the indicated antibodies and protein A-Sepharose. The immunoprecipitates were subjected to Western blot analysis with an anti-tyrosine-phosphorylated STAT5 monoclonal Ab (anti-p-STAT5; Upstate), anti-MgcRacGAP (12), anti-importin β1 (Transduction Laboratories), or anti-MKLP-1, anti-STAT5A, anti-Rac1, or anti-importin α1 (Santa Cruz) antibodies. The filter-bound Ab were detected using the enhanced chemiluminescence system (Amersham). Cytosol and nuclear fractions were prepared, as described previously (17). Fractionation was confirmed by Western blotting with the anti-histone deacetylase (HDAC; for the nuclear fraction) or RhoA (for the cytosol fraction) Ab (Santa Cruz).

Generation of MgcRacGAP conditional knockout cells (5C cells) using DT40 cells.

Full-length MgcRacGAP cDNA was cloned into the EcoRI/BamHI site of pUHD10-3 to yield a tetracycline (TET)-sensitive expression plasmid, pUHD-MgcRacGAP. For targeting constructs, a histidinol (hisD) or puromycin (puro) resistance cassette under the control of the β-actin promoter was inserted between the two arms. The targeting constructs and pUHD-MgcRacGAP were transfected into DT40 cells with a GenePulser II electroporator (Bio-Rad). DT40 cells were cultured and transfected as described previously (32). All DT40 cells were cultured at 38°C in Dulbecco's modified medium supplemented with 10% fetal calf serum, 1% chicken serum, penicillin, and streptomycin. To suppress expression of the tetracycline-responsive transgenes, TET (Sigma) was added to the culture medium at a final concentration of 2 μg/ml.

Expression constructs.

Flag-tagged wild-type (WT) and mutant MgcRacGAP or STATs were cloned into EcoRΙ and NotΙ sites of a mammalian expression vector, pME 18S, or a retrovirus expression vector, pMXs-IG and pMXs-neo. The deletion constructs of MgcRacGAP that lack the internal (INT) domain or various regions of the INT domain were generated by overlapping extension using PCR as described in reference 18. Synthesized oligonucleotides encoding the NLS of MgcRacGAP (NLS-MgcRacGAP; amino acids [aa] 182 to 200), mutant NLS of MgcRacGAP (NLS-MgcRacGAP-182AAA/199AA), or the large T antigen-NLS (PKKKRKV) were cloned into EcoRΙ and NotΙ sites of a pMXs-GFP-fusion vector. Site-directed mutagenesis of MgcRacGAP or STATs was done using QuikChange (Stratagene) and the oligonucleotide primers. Several alanine substitution mutants were generated by overlapping extension using PCR. The sequence generated by PCR was confirmed by automated sequencing using an ABI Prism 310 genetic analyzer (Perkin-Elmer).

Production of retroviruses.

High-titer retroviruses harboring Cre-recombinase were produced in transient retrovirus packaging cell line PLAT-E or PLAT-A (36).

Cell culture and transfection.

The 293T cells were grown in Dulbecco's modified Eagle medium supplemented with 5% fetal calf serum and were transiently transfected with plasmids encoding the wild-type or the mutant forms of STATs using Lipofectamine Plus (Gibco-BRL) according to the manufacturer's recommendations.

Luciferase reporter assay.

For the experiments using DT40 cells, pMKIT (mock) or pMKIT/ITD-Flt3 together with 1.0 μg of pME/STAT5A, 1.0 μg of a reporter plasmid carrying a firefly luciferase gene driven by the β-casein promoter, and 0.5 μg of an internal control reporter plasmid with the Rous sarcoma virus long terminal repeat promoter was introduced into 2 × 106 cells of 5C cells with Nucleofector II (Amaxa) set at program B-009 using the Cell Line Nucleofector kit T (Amaxa) according to the manufacturer's instruction. A control vector carrying GFP was introduced to more than 50% of 5C cells under this condition. Sixteen hours after transfection, cells were lysed and were then subjected to a dual luciferase reporter system (Promega). Luciferase activities were also examined in the lysates of 5C transfectants cotransfected with the NF-κB reporter plasmid (k9) carrying a firefly luciferase gene driven by the IL-6 promoter (14) together with an internal control plasmid. After the transfection, cells were incubated with 30 nM phorbol myristate acetate (PMA) and 1 μM ionomycin for 12 h before cell lysates were prepared. For the experiments using 293T cells, cells were transfected with pME (mock) or pME/STAT5As together with 0.5 μg of pME/EPOR, 0.5 μg of a reporter plasmid carrying a firefly luciferase gene driven by the β-casein promoter, and 0.5 μg of an internal control reporter plasmid with the Rous sarcoma virus long terminal repeat promoter, using Lipofectamine Plus (Gibco-BRL). Twenty-four hours after transfection, cells were stimulated with erythropoietin (EPO; 18 ng/ml; 16 h) and subjected to a dual luciferase reporter system (Promega). To examine the transcriptional activities of STAT3 mutants harboring deletions in DB2, a luciferase assay was done in the lysates of unstimulated or IL-6-stimulated (20 ng/ml) and sIL-6R-stimulated (20 ng/ml) 293T cells cotransfected with the STAT3 reporter plasmid carrying a firefly luciferase gene driven by the mouse glial fibrillary acidic protein promoter (35) together with an internal control reporter plasmid and either the mock vector (pME), the expression vector for the Flag-tagged WT STAT3, or a series of STAT3 mutants harboring deletions in DB2.

Semiquantitative RT-PCR.

Gene expression was examined by semiquantitative reverse transcription-PCR (RT-PCR) analysis. One microgram of total RNA was reverse transcribed with random hexamers by using a first-strand cDNA synthesis kit (Pharmacia), and the reaction mixture was subjected to PCR. PCRs were carried out using the following oligonucleotide primers: chicken Bcl-xL sense (5′-CGTACCAGAGCTTTGAGCAGGT-3′) and antisense (5′-GACCAAGCACAAGCACAATCAC-3′) and chicken glyceraldehyde-3-phosphate dehydrogenase (GAPDH) sense (5′-ATGGTGAAAGTCGGAGTCAACGG-3′) and antisense (5′-ACAGTGCCCTTGAAGTGTCC-3′). PCR was performed under the following conditions to amplify the chicken Bcl-xL gene: 96°C for 3 min; 30 cycles of 96°C for 30 s, 63°C for 30 s, and 72°C for 30 s; a final elongation at 72°C for 8 min. To amplify the GAPDH gene, conditions were 96°C for 3 min, 20 cycles of 96°C for 30 s, 55°C for 30 s, and 72°C for 30 s, and a final elongation at 72°C for 8 min. Amplification of the product is not saturated at this number of cycles.

Modeling of mutant structures.

The models of the D1-10, d356P, d357E, d358L, and L358A mutants of STAT3 were built using the 2.25-Å resolution crystal structure of the STAT3β · DNA complex (2) (PDB ID 1BG1) as a template. Initial models were constructed with the MODELER 9v1 program (9, 28, 41). Then, the models were refined by energy minimization with the AMBER99 force field (46) using MOE computer software (Chemical Computing Group Inc.).

RESULTS

The INT domain of MgcRacGAP contains the bipartite NLS.

While endogenous MgcRacGAP localizes both in the cytoplasm and nucleus, overexpressed exogeneous MgcRacGAP mainly accumulates to the nucleus in the interphase in HeLa cells (12). Although MgcRacGAP does not harbor a classic (polybasic) NLS, we found that the ΔINT mutant localized in the cytoplasm (Fig. (Fig.1C).1C). These results suggested that the INT domain was critical for the nuclear translocation of MgcRacGAP. To identify a functional NLS of MgcRacGAP, we produced a series of Flag-tagged deletion mutants of the INT domain (Fig. (Fig.1A).1A). Protein expression of the mutants and their interaction with a known binding partner, MKLP-1 (34), were confirmed by immunoprecipitation and Western blotting using anti-Flag Ab and anti-MKLP-1 Ab (Fig. (Fig.1B).1B). After the immunostaining experiments in HeLa cells, we found that aa 165 to 204 of MgcRacGAP were required for its nuclear localization (Fig. (Fig.1C).1C). Since this region contained three polybasic regions (178KKR, 182KRR, and 199KK), we tested whether these polybasic regions could serve as an NLS. We produced alanine substitution mutants (178AAA, 182AAA, 199AA, and 182AAA/199AA) and tested their subcellular localizations. While the 178AAA mutant was located in the nucleus, the 182AAA mutant showed partial cytoplasmic localization, and the 199AA and 182AAA/199AA mutants mostly localized in cytoplasm (Fig. (Fig.1C).1C). Next, aa182-200, aa182-200(182AAA/199AA), and a conventional NLS of large T antigen were fused to the N terminus of GFP and we viewed their subcellular localizations with a fluorescence microscope in living cells. As shown in Fig. Fig.1D,1D, aa182-200-GFP and large T-NLS-GFP but not GFP alone or aa182-200(182AAA/199AA)-GFP preferentially accumulated to the nucleus, indicating that aa 182 to 200 is sufficient for a functional polybasic NLS. We also confirmed that leptomycin B treatment, which inhibits nuclear export signal-dependent nuclear export by specific binding to CRM1, did not affect the nuclear or cytoplasmic localization of the WT or 182AAA/199AA mutant, respectively (data not shown). These results suggest that 182KRR and 199KK serve as a functional bipartite NLS for MgcRacGAP. Next, we asked whether MgcRacGAP was transported into the nucleus by importin family members like other proteins carrying NLS. β1Δ450-876, a mutant of importin β1 which lacks the importin α binding site, thereby competitively blocking importin α/β-mediated nuclear import (21), strongly inhibited the nuclear translocation of MgcRacGAP (data not shown). These results implied that MgcRacGAP was transported to the nucleus in an importin α/β-dependent manner.

FIG. 1.
Identification of the nuclear localization signal of MgcRacGAP. (A) Schematic diagram of various Flag-tagged deletion mutants in the INT domain of MgcRacGAP and summary of the localizations of these mutants. (B) Expression of various deletion mutants ...

Importin αs specifically bind MgcRacGAP in yeast through its bipartite NLS.

To determine which member of the importin α family is involved in the nuclear transport of MgcRacGAP, we used a yeast two-hybrid system. To test another possibility, that importin β1 directly mediates the nuclear import of MgcRacGAP, we also examined whether importin β1 could directly bind MgcRacGAP. It was found that importin αs specifically bound MgcRacGAP in yeast (see Fig. S1A in the supplemental material). Moreover, purified importin α1 but not importin β1 pulled down MgcRacGAP from HeLa cell lysate (see Fig. S1B in the supplemental material), suggesting that importin α1 but not importin β1 directly bound MgcRacGAP. The purity of isolated importin α1 and importin β1 was confirmed by CBB staining (see Fig. S1C in the supplemental material). To determine if the bipartite NLS of MgcRacGAP indeed mediated its association with importin α1, we tested the interaction of importin α1 with the INT domain, aa 165 to 205, or each of the alanine mutants of MgcRacGAP (182AAA, 199AA, and 182AAA/199AA), and confirmed that the interaction between importin α1 and MgcRacGAP was mediated by the bipartite NLS of MgcRacGAP (see Fig. S1D in the supplemental material). These results demonstrated that MgcRacGAP directly bound importin αs through its bipartite NLS (182KRR and 199KK).

NLSs of MgcRacGAP and GTP-bound Rac1 were required for nuclear translocation of p-STATs in the in vitro nuclear transport assay.

GTP-bound Rac1 and MgcRacGAP were reported to promote nuclear translocation of p-STATs, forming a ternary complex via the importin α/β pathway in an in vitro nuclear transport assay (17), but the component of the ternary complex that directly binds importin α remained elusive. Neither STAT3 nor STAT5 harbors an apparent polybasic or PY-type NLS. One possibility is that p-STATs themselves harbor unidentified functional NLSs. However, the requirement of MgcRacGAP/GTP-Rac1 for the nuclear translocation of p-STATs raises another possibility that MgcRacGAP functions as an NLS-containing nuclear chaperone of p-STATs. We first asked if the NLS of MgcRacGAP plays a role in the nuclear translocation of p-STAT5A by using an in vitro nuclear transport assay (1). For this, Sf-9-mediated protein purification was done for WT or the 199AA mutant of MgcRacGAP, a constitutively active mutant Rac1 (V12Rac1), STAT5A, or p-STAT5A and for nuclear transporter proteins, including importin α1, importin β1, Ran, and NTF2, which binds Ran and enhances Ran-dependent nuclear import (Fig. (Fig.2A,2A, panel a). We confirmed the purities of recombinant proteins and the tyrosine phosphorylation of STAT5A induced by coexpression with the kinase domain of JAK2 (JH1) (40) in Sf-9 cells (Fig. (Fig.2A,2A, panels a and b). We also examined the extent of STAT5A phosphorylation by phosphate-affinity SDS-PAGE using acrylamide-pendant Phos-tag and found that about 10% was phosphorylated by coexpression of JH1 (data not shown). Although purified p-STAT5A contained only about 10% phosphorylated forms, as we previously reported (17), efficient accumulation to the nuclear envelope of p-STAT5A but not of unphosphorylated STAT5A was achieved in the presence of both purified MgcRacGAP and V12Rac1 (Fig. (Fig.2B,2B, panels g and b), and further addition of the purified nuclear transporters, including importin α1, importin β1, Ran, and NTF2, induced efficient nuclear translocation of p-STAT5A but not of unphosphorylated STAT5A (Fig. (Fig.2B,2B, panels i and d). Addition of the purified nuclear transporters alone was not sufficient for the nuclear translocation of p-STAT5A, and the addition of purified dominant negative N17Rac1 instead of V12Rac1 blocked the nuclear translocation of p-STAT5A even in the presence of purified MgcRacGAP and other nuclear transporters (data not shown). Notably, 199AA-MgcRacGAP did not support the accumulation of p-STAT5A at the nuclear envelope, even in the presence of V12Rac1, and abolished the nuclear translocation of p-STAT5A even in the presence of V12Rac1, importin α1, importin β1, Ran, and NTF2 (Fig. (Fig.2B,2B, panels h and j).

To determine if both the NLS of MgcRacGAP and Rac1 activation are required for complex formation of p-STAT5A with importin α, an in vitro binding assay was done. As we reported previously (17), p-STAT5A, but not unphosphorylated STAT5A, formed complexes with importin α1 only in the presence of WT MgcRacGAP and V12Rac1 (Fig. (Fig.2C).2C). On the other hand, p-STAT5A did not form complexes with importin α1 in the presence of 199AA-MgcRacGAP and V12Rac1 (Fig. (Fig.2C).2C). Addition of importin β1 increased the amount of importin α1 associating with the complex of p-STAT5A, WT MgcRacGAP, and V12Rac1, and further addition of GTP-bound Ran (L69Ran) but not GDP-bound Ran (N24Ran) dissociated the import complex (Fig. (Fig.2D),2D), suggesting that Ran was indeed involved in the nuclear import of p-STAT5A. We also found that 199AA-MgcRacGAP did not support the nuclear translocation of p-STAT3 in the presence of V12Rac1, importin α1, importin β1, Ran, or NTF2 (see Fig. S2 in the supplemental material). Conversely, addition of purified p-STATs and V12Rac1 was required for the nuclear import of GST-MgcRacGAP in the presence of purified importin α/β pathway proteins (see Fig. S3A and B in the supplemental material). These results suggested that the NLS of MgcRacGAP was somehow activated by the association with p-STATs and GTP-bound Rac1, and bound importin α, thereby facilitating nuclear transport of the MgcRacGAP/p-STATs/Rac1 complex. Altogether, the present results indicate that the bipartite NLS was essential for MgcRacGAP to function as a nuclear chaperone of p-STATs, in accordance with GTP-bound Rac1, at least in the in vitro nuclear transport assay.

Generation of MgcRacGAP conditional knockout DT40 cells, in which expression of exogenous MgcRacGAP is under control of a tetracycline-repressible promoter.

To determine if the 199AA or 182AAA/199AA mutant could alter the transcriptional activation of STATs in vivo, we first performed a luciferase assay using 293T cells. The IL-6-induced activation of STAT3 was clearly enhanced by cotransfection with the wild-type MgcRacGAP (~45-fold, compared with mock transfection [~10-fold]). Unexpectedly, cotransfection of the 199AA or 182AAA/199AA mutant modestly inhibited the IL-6-induced transactivation of STAT3 (~25-fold) compared with WT but, rather, enhanced it when compared with mock treatment (data not shown). It is possible that these MgcRacGAP mutants form a heterodimer with endogenous MgcRacGAP and that the heterodimer but not the homodimer of the 199AA or 182AAA/199AA mutant enhanced IL-6-induced transcriptional activation of STAT3. To exclude the effects of endogenous MgcRacGAP, we attempted to establish an MgcRacGAP conditional knockout using DT40 cells. We generated DT40 mutants in which cells with a disrupted MgcRacGAP gene were sustained by expression of the exogenous MgcRacGAP cDNA under the control of a tetracycline-repressible promoter. As shown in Fig. S4A in the supplemental material, an MgcRacGAP-targeting construct was generated such that the 8.0-kb genomic fragment encoding the open reading frame was replaced with one of the two selection cassettes. We transfected the MgcRacGAP-targeting construct containing the histidinol resistance cassette into DT40 cells and isolated MgcRacGAP+/− clones. One MgcRacGAP+/− clone was cotransfected with a chicken MgcRacGAP transgene under the control of a TET-repressible promoter and a TET-repressible transactivator containing a zeocin (ZEO) resistance cassette. We selected ZEO-resistant colonies and identified several clones carrying these constructs integrated at random sites in the genome (MgcRacGAP+/−/MgcRacGAP transgene). Five clones with the MgcRacGAP+/−/MgcRacGAP transgene genotype were transfected with another MgcRacGAP-targeting construct harboring a puromycin selection marker to disrupt the remaining MgcRacGAP allele. We obtained 24 clones with the MgcRacGAP−/−/MgcRacGAP transgene genotype, and one clone, 5C, was chosen for further analysis (see Fig. S4B, panels a and b, in the supplemental material). Exogenous MgcRacGAP protein under the control of a TET-repressible promoter in 5C cells was not detected by Western blot analysis with anti-chicken MgcRacGAP Ab at 6 to 12 h after addition of TET, indicating that a TET-repressible promoter of this clone worked successfully and that MgcRacGAP was actively turned over (Fig. (Fig.3A).3A). When the expression of the MgcRacGAP transgene was suppressed by adding TET, cell growth of 5C cells was suppressed together with the inhibition of cytokinesis, and the cells formed multinucleated cells, eventually undergoing apoptosis within 48 h (Fig. (Fig.3B3B and data not shown). This phenotype is consistent with the previous result indicating that MgcRacGAP is required for completion of cytokinesis.

FIG. 3.
The NLS of MgcRacGAP is required for the transcriptional activation of p-STAT3 in 5C cells. (A) Suppression of MgcRacGAP by TET in 5C cells. The 5C cells were treated with TET for the time indicated and lysed. Cell lysates were separated on SDS-PAGE and ...

The 199AA and 182AAA/199AA mutants inhibited nuclear translocation and transcriptional activation of p-STAT3 in MgcRacGAP knockout cells.

To determine if MgcRacGAP is required for the nuclear translocation of p-STAT3 in vivo, we investigated whether depletion of MgcRacGAP affected the subcellular distribution of p-STAT3 after granulocyte colony-stimulating factor (G-CSF) stimulation using the 5C cells. The 5C cells, which had been transiently transfected with a vector carrying a G-CSF receptor, were treated with TET for 4 h and were stimulated with G-CSF (15 min), followed by nuclear-cytosol fractionation analysis. We also confirmed that after treatment with TET for 4 h only a portion of the 5C cells formed multinucleated cells (less than 10%) (data not shown). The nuclear-cytosol fractionation analysis revealed that depletion of MgcRacGAP resulted in a decreased amount of the G-CSF-induced p-STAT3 as well as inhibition of the nuclear accumulation of p-STAT3 (Fig. (Fig.3C).3C). These results implied that MgcRacGAP mediates the G-CSF-induced phosphorylation and nuclear translocation of p-STAT3. Alternatively, it was possible that depletion of MgcRacGAP indirectly affected activation of STAT3 by disturbing cell cycle machineries. Next, to avoid this possibility, 5C cells were infected with mock or the retrovirus expression vector pMXs-IG carrying WT or the 182AAA, 199AA, or 182AAA/199AA mutant of MgcRacGAP using amphotropic packaging PLAT-A cells (36). The infection efficiencies of these cells were around 10 to 30%, as assessed from the coexpression of GFP using an internal ribosome entry site sequence. After addition of TET, GFP-positive cells grew from the 5C cells transduced with WT or the mutants, while all of the mock-transduced cells became multinucleated, indicating cytokinesis failure, and eventually underwent apoptosis (Fig. (Fig.3B3B and data not shown). These results indicated that the cytokinesis failure of 5C cells after adding TET was prevented by either expression of the WT MgcRacGAP or NLS mutants of MgcRacGAP. All of the 182AAA-, 199AA-, or 182AAA/199AA-expressing cells grew slower than the WT-expressing cells in the presence of TET (Fig. (Fig.3D),3D), suggesting that the NLS of MgcRacGAP plays some role in enhancing cell growth but is dispensable for completion of cytokinesis in 5C cells. Expression levels of WT MgcRacGAP or 182AAA, 199AA, or 182AAA/199AA mutant were comparable as assessed in Western blot analyses (Fig. (Fig.3E).3E). We next investigated whether disruption of the NLS of MgcRacGAP affected the subcellular distribution of endogenous p-STAT3 after G-CSF stimulation by using the 5C transfectants in the presence of TET. The 5C cells expressing WT MgcRacGAP or the 199AA or 182AAA/199AA mutant, which had been transiently transfected with a vector carrying the G-CSF receptor, were stimulated with G-CSF (15 min). The amounts of G-CSF-induced p-STAT3 in 5C transfectants expressing the WT and those expressing the NLS-lacking mutants were found to be comparable (Fig. (Fig.3F).3F). Interestingly, the nuclear-cytosol fractionation analysis revealed that the 199AA- or 182AAA/199AA-MgcRacGAP hardly entered the nucleus, and the G-CSF-induced nuclear accumulation of p-STAT3 was strongly inhibited in 199AA- or 182AAA/199AA-expressing cells compared with those in the 182AAA- and WT-expressing cells (Fig. (Fig.3G).3G). We also performed a semiquantitative RT-PCR analysis to test if induction of Bcl-xL mRNA (one of the target genes of STAT3) was affected in the transfectants expressing the NLS-lacking mutants of MgcRacGAP after the G-CSF stimulation, and we found that induction of Bcl-xL mRNA in response to G-CSF stimulation was severely impaired in the transfectants expressing 199AA or 182AAA/199AA (Fig. (Fig.3H3H).

The 199AA and 182AAA/199AA mutants specifically blocked transcriptional activation of p-STATs in MgcRacGAP knockout cells.

Next, we performed a luciferase assay using the 5C transfectants cultured in the presence of TET. Transcriptional activities of STAT3 in response to G-CSF stimulation were strongly inhibited in the 199AA- or 182AAA/199AA-expressing 5C cells compared to those in the WT-expressing cells (Fig. (Fig.4A).4A). We obtained similar results for STAT5; transcriptional activation of STAT5 induced by ITD-Flt3 was profoundly inhibited in the 199AA- or 182AAA/199AA-expressing 5C cells compared to those in the WT-expressing cells (Fig. (Fig.4B).4B). However, NF-κB p65, whose NLS is unmasked by IκBα degradation and binds importins α3 and α4 (8, 23), entered the nucleus after stimulation even in the 199AA- or 182AAA/199AA-expressing cells (Fig. (Fig.4C).4C). In addition, transcriptional activities of NF-κB in response to stimulation with PMA/ionomycin were not affected in the 199AA- or 182AAA/199AA-expressing 5C cells compared to those in the WT-expressing cells (Fig. (Fig.4D),4D), indicating that MgcRacGAP does not work as a general nuclear chaperone.

FIG. 4.
The NLS of MgcRacGAP is not required for activation of NF-κB p65 in 5C cells. (A) The NLS of MgcRacGAP was required for transcriptional activities of STAT3. Luciferase activities were examined in the lysates of 5C transfectants cotransfected with ...

The series of STAT3 mutants harboring deletions in the two strands (βa′ and βb) of the β-barrel lost transcriptional activities, while the mutants harboring deletions in the region following the strand βb (βb-βc loop) showed constitutively active phenotypes.

We previously found that STAT3 and STAT5 directly bound MgcRacGAP through aa 338 to 362 and aa 341 to 365 in their DNA binding domain, respectively (termed DB2-STAT3 and DB2-STAT5) and that the STAT3 and STAT5A mutants lacking DB2 (STAT3-dDB2 and STAT5A-dDB2) lost not only the capability for binding to MgcRacGAP but also their transcriptional activities (17). The DB2 region is well-conserved among STAT family proteins. In this study, we produced a series of deletion mutants lacking a three-amino-acid stretch in the DB2-STAT3 region (STAT3-dD1 to -8) and in the next six amino acids (STAT3-dD9 and -10) (Fig. (Fig.5A).5A). Tyrosine phosphorylation of STAT3-dDB2, -dD1, -dD3, -dD4, or -dD5 in response to IL-6 stimulation was diminished, whereas tyrosine phosphorylation of STAT3-dD2 was prominent even in the absence of IL-6 (Fig. (Fig.5C,5C, middle panel). In addition, association of STAT3-dDB2, -dD1, -dD3, -dD4, or -dD5 with MgcRacGAP was not detected, while binding of STAT3-dD2 with MgcRacGAP increased compared with that of WT in the absence or presence of IL-6 stimulation (Fig. (Fig.5C,5C, upper panels). Nonetheless, the mutants lacking D1 to -5 (STAT3-dD1 to -5), including STAT3-dD2, did not show detectable transcriptional activities in response to IL-6 stimulation (Fig. (Fig.5B).5B). Surprisingly, STAT3-dD6 to -9 mutants exerted considerable transcriptional activities even without cytokine stimulation, and this was further enhanced by IL-6 stimulation (Fig. (Fig.5B).5B). These mutants were constitutively tyrosine phosphorylated, and their tyrosine phosphorylation was augmented after IL-6 stimulation (Fig. (Fig.5C).5C). STAT3-dD10 was constitutively tyrosine phosphorylated but did not show detectable transcriptional activities, as was the case for STAT3-dD2. We next examined whether these STAT3 mutants harbored the DNA binding activities in an electrophoretic mobility shift assay using unstimulated 293T cells and found that unlike the other constitutively tyrosine-phosphorylated STAT3 mutants, STAT3-dD2 and STAT3-dD10 lost their DNA binding affinities (data not shown). Notably, STAT3-dD7 showed the strongest transcriptional activities in the absence of cytokine stimulation among the STAT3-dD1 to -10 mutants, and its transcriptional activities in the absence of cytokine stimulation were much stronger than that of the WT after IL-6 stimulation (Fig. (Fig.5B).5B). We next produced a series of mutants lacking each single amino acid of the three amino acids in the region of D7 (STAT3-d356P, -d357E, and -d358L). Interestingly, STAT3-d356P, -d357E, and -d358L, which strongly bound MgcRacGAP, displayed the constitutive activities in the absence of IL-6 stimulation (Fig. 5B and C). STAT3-d356P and STAT3-d358L exerted the strongest transcriptional activity among these mutants and a reported constitutively active mutant of STAT3C (3). These results suggest that the two strands (βa′ and βb) in DB2 are required for the IL-6-induced tyrosine phosphorylation of STAT3 that mediates the interaction with MgcRacGAP, whereas the deletion mutants in the C terminus of DB2 following the strand βb (βb-βc loop) tend to become constitutively active with enhanced binding to MgcRacGAP.

FIG. 5.
Correlation between binding abilities of STAT3 to MgcRacGAP and activities of STAT3. (A) Schematic diagrams showing a series of the deletion sites of STAT3 mutants. (B) Transcriptional activities of STAT3 mutants harboring deletions in DB2. Luciferase ...

We also produced a series of STAT5A mutants lacking a three-amino-acid stretch in the region corresponding to DB2-STAT3 (STAT5A-dD1 to -8) and in the next six amino acids (STAT5A-dD9 and -10) (Fig. (Fig.6A).6A). We found that the series of deletion mutants of STAT5A in DB2 showed phenotypes similar to those of STAT3 mutants (Fig. 6B and C); STAT5A mutants harboring deletions in the two strands (βa′ and βb) of the β-barrel (STAT5A-dD1 to -5) were not tyrosine phosphorylated by EPO stimulation and lost transcriptional activity, while the mutants harboring deletions in the region following the strand βb (STAT5A-dD7 to -9) showed gain-of-function phenotypes. STAT5A-dD8, -d363H, and -d364M also showed constitutively active phenotypes (Fig. 6B and C), although the transcriptional activities of these mutants without stimulation were weaker than that of another constitutively active mutant, STAT5A1*6 (37). Constitutive activities of STAT5A mutants were well-correlated with constitutive binding to MgcRacGAP. Association between the constitutively active STAT5A mutants and MgcRacGAP was stronger than that of the WT and MgcRacGAP in the absence of EPO stimulation (Fig. (Fig.6C).6C). Thus, the molecular basis of the STAT-MgcRacGAP interaction is well-conserved between STAT3 and STAT5A.

FIG. 6.
The series of deletion mutants of STAT5A in DB2 showed similar phenotypes to those of STAT3. (A) Schematic diagrams showing a series of the deletion mutants of STAT5A. (B) The mutants lacking D1 to -5 and D10 (STAT5A-dD1 to -5 and STAT5A-dD10) as well ...

The constitutively active mutant STAT3-d358L promoted cell proliferation of a mutant cell line derived from BaF-BO3 cells.

We next examined whether STAT3-d358L was biologically functional, showing physiological roles of STAT3 activation. It has been reported that BaF-BO3 cells expressing a G-CSF receptor mutant (G133 cells) are able to proliferate in response to G-CSF stimulation via activating SHP-2/mitogen-activated protein kinase and JAK/STAT3 pathways (10). This mutant receptor, G133, is a chimeric receptor composed of the extracellular domain of the G-CSF receptor and the transmembrane and cytoplasmic domains of gp130 truncated 133 amino acids from the transmembrane domain. Mutation of the tyrosine residue in the YXXQ motif within G133 (G133F3) abolished STAT3 activation and, thus, cell proliferation driven by G-CSF (10). G133F3 cells were transduced with the pMXpuro vector carrying the WT STAT3, STAT3C, STAT3-d358L, a constitutively active mutant H-Ras (H-RasV12), or the expression vector alone (mock). Expression levels of WT STAT3, STAT3C, and STAT3-d358L were similar as judged by Western blotting (Fig. (Fig.7B,7B, upper panel). The expression and activation of H-RasV12 were confirmed by Western blotting (Fig. (Fig.7B,7B, lower panel, input lane) and by a pull-down assay using GST-Raf-RBD (Fig. (Fig.7B,7B, lower panel, other lanes), respectively. As shown in Fig. Fig.7A,7A, STAT3-d358L but neither WT, STAT3C, H-RasV12, nor mock treatment promoted cell proliferation of G133F3 cells under G-CSF stimulation, suggesting that the STAT3-d358L mutant is biologically functional and could be a useful tool to study the physiological roles of STAT3 activation. This result also indicates that STAT3-d358L is stronger than STAT3C in inducing STAT3-dependent cell growth, which is consistent with their transcriptional activities (Fig. (Fig.5B5B).

FIG. 7.
A constitutively active STAT3 mutant, STAT3-d358L, preferentially bound MgcRacGAP and Rac1 and accumulated to the nucleus. (A) STAT3-d358L supports proliferation of BaF-BO3-G133F3 cells in the presence of G-CSF. BaF-BO3-G133F3 cells expressing mock vector ...

Interestingly, the tyrosine-phosphorylated form of STAT3-d358L accumulated to the nucleus after IL-6 stimulation more evidently than that of the WT STAT3 (Fig. (Fig.7C).7C). Importantly, the STAT3-d358L bound MgcRacGAP and Rac1 more strongly than the WT in the absence or presence of IL-6 stimulation (Fig. (Fig.7D).7D). Although most of the phosphorylated form of overexpressed WT STAT3 remained in the cytoplasm (Fig. (Fig.7C),7C), this was probably because p-STAT3 requires MgcRacGAP/Rac1 as cofactors to enter the nucleus and these cofactors are limiting the nuclear translocation of p-STAT3. Taken together, our results strongly indicate that the interaction of STATs with MgcRacGAP accompanied by GTP-bound Rac1 plays critical roles in regulating STAT functions through facilitating both tyrosine phosphorylation of STATs and nuclear translocation of p-STATs.

DISCUSSION

We originally identified MgcRacGAP in a search for key molecules that are involved in the IL-6-induced macrophage differentiation of M1 cells (18) and found that MgcRacGAP and Rac1 form a ternary complex with STAT3 and are required for STAT3 activation (43). We also reported that MgcRacGAP localizes to the midbody of dividing cells and plays a crucial role in the completion of cytokinesis, thus playing a distinct role in the mitotic phase (12, 33). We recently found that GTP-bound Rac1 and MgcRacGAP are required for nuclear translocation of p-STATs via the importin pathway in an in vitro nuclear transport assay (17). In this paper, to identify the molecular mechanisms of how GTP-bound Rac1 and MgcRacGAP facilitate complex formation of p-STATs with importin αs, we used MgcRacGAP conditional knockout chicken DT40 cells (5C cells) as well as a nuclear transport assay and demonstrated that the NLS of MgcRacGAP plays a critical role in the nuclear translocation of p-STAT3/5. Although the biological functions of STAT3 and STAT5 are not identical, we demonstrated that nuclear import of p-STAT3 and p-STAT5 was mediated by MgcRaGAP and its NLS, and the molecular mechanisms are common.

Liu et al. (25) reported that constitutive nuclear import of STAT3 monomer is independent of tyrosine phosphorylation and is mediated by importin α3. They found that a deletion mutant of STAT3 (d150-163) (aa 150 to 162; DVRKRVQDLEQKM) did not enter the nucleus. However, the substitution mutant of the basic amino acid cluster in this sequence did not hamper nuclear accumulation. Based on these results, they reasoned that aa 150 to 162 play a role in a conformational structure that is required for nuclear import (25). Similarly, Zeng et al. reported that aa 138 to 165 of STAT5B are required for constitutive nuclear import of STAT5B monomer but that this region does not harbor polybasic amino acids (49). Thus, it was not clear whether STAT3 and STAT5 harbor a functional NLS or whether dimer formation creates a polybasic NLS of STAT3 and STAT5. We here propose that MgcRacGAP accompanied by GTP-bound Rac1 functions as an NLS-containing nuclear chaperone toward p-STATs (Fig. 2B and C). Interestingly, Rac1 was reported to play a role in the nuclear import of SmgGDS and p120 catenin (22), members of the importin α-like armadillo family of proteins (4, 38). The C-terminal region of Rac1, but not Rac2 or Rac3, contains a polybasic region, which may function as an NLS. However, Lanning et al. (22) also suggested that the interaction of Rac1 with its GTP exchange factor SmgGDS was qualitatively different from that of NLS-containing molecules with importin αs. Consistent with this, we failed to detect direct interactions of GTP-bound Rac1 with importin αs in either the in vitro binding assay or yeast two-hybrid assay (data not shown). Therefore, it is unlikely that GTP-bound Rac1 serves directly as an NLS-containing chaperone of p-STATs. The results shown in Fig. Fig.2C2C strongly indicate that GTP-bound Rac1 activates the NLS MgcRacGAP associating with p-STATs; however, the precise molecular mechanism for the requirement of Rac1 remains to be clarified by structural analysis.

The present results demonstrate that the bipartite NLS (182KRR/199KK) of MgcRacGAP is essential for the nuclear transport and the transcriptional activation of p-STATs in living cells (Fig. 3G and H and and4A).4A). The results of the nuclear-cytosol fractionation analysis using the 5C cells expressing 199AA and 182AAA/199AA (Fig. (Fig.3G)3G) also suggested that the preferential nuclear localization of MgcRacGAP is mediated by the importin pathway in living cells. However, in the nuclear transport assay using semi-intact cells and purified proteins, nuclear translocation of MgcRacGAP was not achieved by the addition of importin α/β pathway proteins alone and, interestingly, further addition of p-STATs and GTP-bound Rac1 was required for nuclear translocation of MgcRacGAP (see Fig. S3A and B in the supplemental material). This raised a question of why overexpressed MgcRacGAP predominantly accumulated to the nucleus in HeLa cells, where STATs were not extensively activated. In addition, purified MgcRacGAP pulled down importin αs from the HeLa cell lysate (see Fig. S1B in the supplemental material), while it did not bind purified importin αs in the in vitro binding assay (Fig. (Fig.2C2C and data not shown). These results suggest that the NLS of MgcRacGAP can be activated by other cargo proteins as well as by p-STAT and GTP-bound Rac1. Thus, MgcRacGAP may function as a nuclear chaperone for not only p-STATs but also another nuclear protein(s).

The conditional knockout of MgcRacGAP in 5C cells decreased the G-CSF-induced tyrosine phosphorylation of STAT3 (Fig. (Fig.3C),3C), and small interfering RNA-mediated MgcRacGAP knockdown in Ba/F3 cells also reduced tyrosine phosphorylation of STAT5 (17). These results implied that MgcRacGAP functions as an upstream regulator of STAT activation as well. In relation to this, STAT mutants harboring deletions in the βb-βc loop showed enhanced interaction with MgcRacGAP and became constitutively active (Fig. (Fig.55 and and6).6). It should be noted that the extent of tyrosine phosphorylation of STAT3-d358L without stimulation was weaker than that of WT 16 h after IL-6 stimulation (Fig. (Fig.7D),7D), although the association of STAT3-d358L with MgcRacGAP/Rac1 was stronger than that of WT under the same conditions (Fig. (Fig.5C5C and and7D).7D). These results indicate that the stronger association of the constitutively active STAT mutants with MgcRacGAP was not a secondary result following the enhanced tyrosine phosphorylation of these mutants, and they imply a positive role for MgcRacGAP in facilitating STAT activation. In this context, it is interesting to note that V12Rac1 induced translocation of MgcRacGAP to the plasma membrane (see Fig. S5 in the supplemental material) and that MgcRacGAP bound JAK2 (17). In addition, the interaction of the STAT3-Y704F mutant, which does not undergo tyrosine phosphorylation, with MgcRacGAP was enhanced by IL-6 stimulation, similar to that of the WT STAT3 (data not shown). This indicates that the IL-6-induced MgcRacGAP interaction with STATs does not require tyrosine phosphorylation of STATs and occurs before their tyrosine phosphorylation. Based on these observations, we propose a model of STAT nucleo-cytoplasmic shuttling regulated by Rac1/MgcRacGAP (Fig. (Fig.88).

FIG. 8.
A current model of nuclear import of p-STATs and a working hypothesis for membrane targeting and phosphorylation of STATs. In the present work, we demonstrated that the NLS of MgcRacGAP accompanied by GTP-bound Rac1 is essential for nuclear translocation ...

One important finding of the present paper is that the abilities of STAT mutants to bind MgcRacGAP correlated well with the activation of STATs (Fig. (Fig.55 and and6).6). To reveal the molecular mechanisms of MgcRacGAP-mediated regulation of STAT phosphorylation, we conducted an in vitro kinase reaction of purified STAT5 using purified JAK2 in the presence or absence of Rac1 and MgcRacGAP. However, STAT phosphorylation was not enhanced by the addition of Rac1/MgcRacGAP in this mixture (data not shown). Based on the result that MgcRacGAP binds JAK2, we speculate that MgcRacGAP regulates STAT phosphorylation by conveying STAT proteins to JAK2 or by serving as a scaffold for the interaction of JAK2 and STAT.

We also observed that STAT3-d358L bound MgcRacGAP more strongly than WT STAT3 did in yeast (data not shown). In the crystal structure of tyrosine-phosphorylated STAT3β (2) (PDB ID 1BG1), the MgcRacGAP binding DB2 region includes the C terminus of the βa′ strand, the DNA-bound βa′-βb loop, the βb strand, the βb-βc loop, and the N terminus of the βc strand in the β-barrel domain, as shown in Fig. Fig.5A5A and and9A.9A. The tertiary structure of this domain appears to be rigid, as its B-factors (ca. 51 Å2) are considerably lower than those of the other domains of the protein (ca. 66 Å2). Deletion mutants of the βb strand (dD3 to dD5) resulted in loss of the MgcRacGAP binding ability of STAT3, suggesting that the βb strand is the MgcRacGAP binding site. Interestingly, the MgcRacGAP binding ability and the transcriptional activity of STAT3 were enhanced by deletion of the D6 to D9 region (Fig. (Fig.5C).5C). These phenotypes may be explained by the flexibility around the βb strand. Within the β-barrel domain, the DNA-bound βa′-βb loop and the βb-βc loop region are as rigid as the β strands (the B-factors, ca. 40 Å2), while the βc-βx, βc′-βe, and βe-βf loops are much more flexible (Fig. (Fig.9A).9A). The rigidity of the βb-βc loop is probably because Leu358 in this loop is involved in the hydrophobic core formation. When a deletion mutation is introduced in the βb-βc loop (dD6-dD9, d356P, d357E, and d358L), the tertiary structure of the domain may be retained but become more flexible, as the hydrophobic core is packed less tightly than the wild type (see Fig. 9B, C, and D for representative structural models). These STAT3 mutants with more flexibility around the βb strand (binding site of MgcRacGAP) could bind more efficiently to MgcRacGAP, leading to their enhanced activation. In the L358A mutant, which behaves similarly to the d358L mutant (data not shown), the hydrophobic core of the domain is also loosened (Fig. (Fig.9E).9E). Thus, all of these deletion and point mutations may destabilize the hydrophobic core of the domain around the βb strand (binding site of MgcRacGAP) and therefore seem to increase the binding ability of the βb strand to MgcRacGAP. The dD2 and dD10 mutants also strongly bind to MgcRacGAP, possibly because of the distortion of the domain structure, while hyperactive transcription was not observed, as they lacked the DNA binding activities. The loss of the MgcRacGAP binding ability in the dD1 mutant lacking the C terminus of the βa′ strand may be due to a secondary result of structural distortion of the βa′ strand and the DNA-bound βa′-βb loop. Our current hypothesis is that upon binding with MgcRacGAP, the β-barrel domain of STAT undergoes some conformational change so that the MgcRacGAP binding region (the βb strand) becomes more flexible and more exposed.

FIG. 9.
Structure of the wild-type STAT3 and homology models of the hydrophobic core surrounding Leu358. The crystal structure of the β-barrel domain of the wild-type STAT3β (A) and its hydrophobic core region (B). In panel A, the B-factors are ...

MgcRacGAP accompanied by GTP-bound Rac1 functions both as a mediator of the tyrosine phosphorylation of STATs and as an NLS-containing nuclear chaperone of p-STATs during interphase, while MgcRacGAP plays a critical role in cell division. We believe that MgcRacGAP is a molecule which functions in the nucleocytoplasmic transporting system during interphase and in the mitotic apparatus from metaphase to cytokinesis, as is the case with nucleocytoplasmic transporters, including importins and Ran, which are also involved in the formation of the mitotic spindle after the disassembly of the nuclear envelope (7). Although MgcRacGAP is not involved in nuclear translocation of NF-κB, whether it is involved in nuclear transport of other proteins is still open to question.

Supplementary Material

[Supplemental material]

Acknowledgments

We thank Toshio Hirano for kindly providing BaF-BO3 mutants and Dovie Wylie for language assistance.

This work was supported by the RIKEN Structural Genomics/Proteomics Initiative, the National Project on Protein Structural and Functional Analyses, the Sumitomo Foundation, and the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Footnotes

[down-pointing small open triangle]Published ahead of print on 21 January 2009.

Supplemental material for this article may be found at http://mcb.asm.org/.

REFERENCES

1. Adam, S. A., R. S. Marr, and L. Gerace. 1990. Nuclear protein import in permeabilized mammalian cells requires soluble cytoplasmic factors. J. Cell Biol. 111807-816. [PMC free article] [PubMed]
2. Becker, S., B. Groner, and C. W. Müller. 1998. Three-dimensional structure of the Stat3β homodimer bound to DNA. Nature 394145-151. [PubMed]
3. Bromberg, J. F., M. H. Wrzeszczynska, G. Devgan, Y. Zhao, R.G. Pestell, C. Albanese, and J. E. Darnell, Jr. 1999. The JAK-STAT pathway: summary of initial studies and recent advances. Stat3 as an oncogene. Cell 98295-303. [PubMed]
4. Chook, Y. M., and G. Blobel. 2001. Karyopherins and nuclear import. Curr. Opin. Struct. Biol. 11703-715. [PubMed]
5. Darnell, J. E., Jr. 2002. Transcription factors as targets for cancer therapy. Nat. Rev. Cancer 2740-749. [PubMed]
6. Darnell, J. E., Jr. 1996. The JAK-STAT pathway: summary of initial studies and recent advances. Recent Prog. Horm. Res. 51391-403. [PubMed]
7. Dasso, M. 2001. Running on Ran nuclear transport and the mitotic spindle. Cell 104321-324. [PubMed]
8. Fagerlund, R., L. Kinnunen, M. Köhler, I. Julkunen, and K. Melén. 2005. NF-κB is transported into the nucleus by importin α3 and importin α4. J. Biol. Chem. 28015942-15951. [PubMed]
9. Fiser, A., R. K. Do, and A. Sali. 2000. Modeling of loops in protein structures. Protein Sci. 91753-1773. [PMC free article] [PubMed]
10. Fukada, T., M. Hibi, Y. Yamanaka, M. Takahashi-Tezuka, Y. Fujitani, T. Yamaguchi, K. Nakajima, and T. Hirano. 1996. Two signals are necessary for cell proliferation induced by a cytokine receptor gp130: involvement of STAT3 in anti-apoptosis. Immunity 5449-460. [PubMed]
11. Gorlich, D., and U. Kutay. 1999. Transport between the cell nucleus and the cytoplasm. Annu. Rev. Cell Dev. Biol. 15607-660. [PubMed]
12. Hirose, K., T. Kawashima, I. Iwamoto, T. Nosaka, and T. Kitamura. 2001. MgcRacGAP is involved in cytokinesis through associating with mitotic spindle and midbody. J. Biol. Chem. 2765821-5828. [PubMed]
13. Ihle, J. N. 1996. Janus kinases in cytokine signalling. Philos. Trans. R. Soc. Lond. B 351159-166. [PubMed]
14. Isshiki, H., S. Akira, O. Tanabe, T. Nakajima, T. Shimamoto, T. Hirano, and T. Kishimoto. 1990. Constitutive and interleukin-1 (IL-1)-inducible factors interact with the IL-1-responsive element in the IL-6 gene. Mol. Cell. Biol. 102757-2764. [PMC free article] [PubMed]
15. Jaffe, A. B., and A. Hall. 2005. Rho GTPases: biochemistry and biology. Annu. Rev. Cell Dev. Biol. 21247-269. [PubMed]
16. Jantsch-Plunger, V., P. Gonczy, A. Romano, H. Schnabel, D. Hamill, R. Schnabel, A.A. Hyman, and M. Glotzer. 2000. CYK-4: a Rho family gtpase activating protein (GAP) required for central spindle formation and cytokinesis. J. Cell Biol. 1491391-1404. [PMC free article] [PubMed]
17. Kawashima, T., Y. C. Bao, Y. Nomura, Y. Moon, Y. Tonozuka, Y. Minoshima, T. Hatori, A. Tsuchiya, M. Kiyono, T. Nosaka, H. Nakajima, D.A. Williams, and T. Kitamura. 2006. Rac1 and a GTPase activating protein MgcRacGAP are required for nuclear translocation of STAT transcription factors. J. Cell Biol. 175937-946. [PMC free article] [PubMed]
18. Kawashima, T., K. Hirose, T. Satoh, A. Kaneko, Y. Ikeda, Y. Kaziro, T. Nosaka, and T. Kitamura. 2000. MgcRacGAP is involved in the control of growth and differentiation of hematopoietic cells. Blood 962116-2124. [PubMed]
19. Kinoshita, E., E. Kinoshita-Kikuta, K. Takiyama, and T. Koike. 2006. Phosphate-binding tag, a new tool to visualize phosphorylated proteins. Mol. Cell. Proteomics 5749-757. [PubMed]
20. Koradi, R., M. Billeter, and K. Wüthrich. 1996. MOLMOL: a program for display and analysis of macromolecular structures. J. Mol. Graph. 1451-55. [PubMed]
21. Kutay, U., E. I. Izaurralde, F.R. Bischoff, I.W. Mattaj, and D. Görlich. 1997. Dominant-negative mutants of importin-β block multiple pathways of import and export through the nuclear pore complex. EMBO J. 161153-1163. [PMC free article] [PubMed]
22. Lanning, C. C., R. Ruiz-Velasco, and C. L. Williams. 2003. Novel mechanism of the co-regulation of nuclear transport of SmgGDS and Rac1. J. Biol. Chem. 27812495-12506. [PubMed]
23. Latimer, M., M. K. Ernst, L. L. Dunn, M. Drutskaya, and N. R. Rice. 1998. The N-terminal domain of IκBα masks the nuclear localization signal(s) of p50 and c-Rel homodimers. Mol. Cell. Biol. 182640-2649. [PMC free article] [PubMed]
24. Lee, B. J., A. E. Cansizoglu, K. E. Süel, T. H. Louis, Z. Zhang, and Y. M. Chook. 2006. Rules for nuclear localization sequence recognition by karyopherin beta 2. Cell 126543-558. [PMC free article] [PubMed]
25. Liu, L., K. M. McBride, and N. C. Reich. 2005. STAT3 nuclear import is independent of tyrosine phosphorylation and mediated by importin-alpha3. Proc. Natl. Acad. Sci. USA 1028150-8155. [PMC free article] [PubMed]
26. Ma, J., and X. Cao. 2006. Regulation of Stat3 nuclear import by importin α5 and importin α7 via two different functional sequence elements. Cell Signal. 181117-1126. [PubMed]
27. Marg, A., Y. Shan, T. Meyer, T. Meissner, M. Brandenburg, and U. Vinkemeier. 2004. Nucleocytoplasmic shuttling by nucleoporins Nup153 and Nup214 and CRM1-dependent nuclear export control the subcellular distribution of latent Stat1. J. Cell Biol. 165823-833. [PMC free article] [PubMed]
28. Marti-Renom, M. A., A. Stuart, A. Fiser, R. Sanchez, F. Melo, and A. Sali. 2000. Comparative protein structure modeling of genes and genomes. Annu. Rev. Biophys. Biomol. Struct. 29291-325. [PubMed]
29. Mattaj, I., and L. Englmeier. 1998. Nucleocytoplasmic transport: the soluble phase. Annu. Rev. Biochem. 67265-306. [PubMed]
30. McBride, K. M., G. Banninger, C. McDonald, and N. C. Reich. 2002. Regulated nuclear import of the STAT1 transcription factor by direct binding of importin-α. EMBO J. 211754-1763. [PMC free article] [PubMed]
31. Michaelson, D., W. Abidi, D. Guardavaccaro, M. Zhou, I. Ahearn, M. Pagano, and M. R. Philips. 2008. Rac1 accumulates in the nucleus during the G2 phase of the cell cycle and promotes cell division. J. Cell Biol. 181485-496. [PMC free article] [PubMed]
32. Minoshima, Y., T. Hori, M. Okada, H. Kimura, T. Haraguchi, Y. Hiraoka, Y. C. Bao, T. Kawashima, T. Kitamura, and T. Fukagawa. 2005. The constitutive centromere component CENP-50 is required for recovery from spindle damage. Mol. Cell. Biol. 2510315-10328. [PMC free article] [PubMed]
33. Minoshima, Y., T. Kawashima, K. Hirose, Y. Tonozuka, A. Kawajiri, Y.C. Bao, X. Deng, M. Tatsuka, S. Narumiya, W. S. May, Jr., T. Nosaka, K. Semba, T. Inoue, T. Satoh, M. Inagaki, and T. Kitamura. 2003. Phosphorylation by Aurora B converts MgcRacGAP to a RhoGAP during cytokinesis. Dev. Cell 4549-560. [PubMed]
34. Mishima, M., S. Kaitna, and M. Glotzer. 2002. Central spindle assembly and cytokinesis require a kinesin-like protein/RhoGAP complex with microtubule bundling activity. Dev. Cell 241-54. [PubMed]
35. Miura, M., T. Tamura, and K. Mikoshiba. 1990. Cell-specific expression of the mouse glial fibrillary acidic protein gene: identification of the cis- and trans-acting promoter elements for astrocyte-specific expression. J. Neurochem. 551180-1188. [PubMed]
36. Morita, S., T. Kojima, and T. Kitamura. 2000. Plat-E: an efficient and stable system for transient packaging of retroviruses. Gene Ther. 121063-1066. [PubMed]
37. Onishi, M., T. Nosaka, K. Misawa, A. L. Mui, D. Gorman, M. McMahon, A. Miyajima, and T. Kitamura. 1998. Identification and characterization of a constitutively active STAT5 mutant that promotes cell proliferation. Mol. Cell. Biol. 183871-3879. [PMC free article] [PubMed]
38. Peifer, M., S. Berg, and A. B. Reynolds. 1994. A repeating amino acid motif shared by proteins with diverse cellular roles. Cell 76789-791. [PubMed]
39. Ridley, A. J. 2006. Rho GTPases and actin dynamics in membrane protrusions and vesicle trafficking. Trends Cell Biol. 16522-529. [PubMed]
40. Saharinen, P., K. Takaluoma, and O. Silvennoinen. 2000. Regulation of the Jak2 tyrosine kinase by its pseudokinase domain. Mol. Cell. Biol. 203387-3395. [PMC free article] [PubMed]
41. Sali, A., and T. L. Blundell. 1993. Comparative protein modelling by satisfaction of spatial restraints. J. Mol. Biol. 234779-815. [PubMed]
42. Sekimoto, T., N. Imamoto, K. Nakajima, T. Hirano, and Y. Yoneda. 1997. Extracellular signal-dependent nuclear import of Stat1 is mediated by nuclear pore-targeting complex formation with NPI-1, but not Rch1. EMBO J. 167067-7077. [PMC free article] [PubMed]
43. Tonozuka, Y., Y. Minoshima, Y. C. Bao, Y. Moon, Y. Tsubono, T. Hatori, H. Nakajima, T. Nosaka, T. Kawashima, and T. Kitamura. 2004. A GTPase activating protein binds STAT3 and is required for IL-6-induced STAT3 activation and for differentiation of a leukemic cell line. Blood 1043550-3557. [PubMed]
44. Ushijima, R., N. Sakaguchi, A. Kano, A. Maruyama, Y. Miyamoto, T. Sekimoto, Y. Yoneda, K. Ogino, and T. Tachibana. 2005. Extracellular signal-dependent nuclear import of STAT3 is mediated by various importin alphas. Biochem. Biophys. Res. Commun. 330880-886. [PubMed]
45. Van de Putte, T., A. Zwijsen, O. Lonnoy, V. Rybin, M. Cozijnsen, A. Francis, V. Baekelandt, C. A. Kozak, M. Zerial, and D. Huylebroeck. 2001. Mice with a homozygous gene trap vector insertion in mgcRacGAP die during pre-implantation development. Mech. Dev. 10233-44. [PubMed]
46. Wang, J. M., P. Cieplak, and P. A. Kollman. 2000. How well does a restrained electrostatic potential (RESP) model perform in calculating conformational energies of organic and biological molecules? J. Comput. Chem. 211049-1074.
47. Wu, X. M., X. L. Tu, K. S. Joeng, M. J. Hilton, D. A. Williams, and F. X. Long. 2008. Rac1 activation controls nuclear localization of beta-catenin during canonical Wnt signaling. Cell 133340-353. [PMC free article] [PubMed]
48. Yamada, T., T. Kurosaki, and M. Hikida. 2008. Essential roles of MgcRacGAP in multilineage differentiation and survival of murine hematopoietic cells. Biochem. Biophys. Res. Commun. 372941-946. [PubMed]
49. Zeng, R., Y. Aoki, M. Yoshida, K. Arai, and S. Watanabe. 2002. Stat5B shuttles between cytoplasm and nucleus in a cytokine-dependent and -independent manner. J. Immunol. 1684567-4575. [PubMed]

Articles from Molecular and Cellular Biology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...