Importance of Dimer Formation of Myocardin Family Members in the Regulation of Their Nuclear Export

Abstract

Myocardin (Mycd) family members function as a transcriptional cofactor for serum response factor (SRF). Dimer formation is necessary to exhibit their function, and the coiled-coil domain (CC) plays a critical role in their dimerization. We have recently revealed a detailed molecular mechanism for their Crm1 (exportin1)-mediated nuclear export. Here, we found other unique significances of the dimerization of Mycd family members. Introduction of mutations in the CC of myocardin-related transcription factor A (MRTF-A) and truncated Mycd resulted in significant decreases in their cytoplasmic localization and increases in their nuclear localization. In accordance with such subcellular localization changes, their binding to Crm1 were reduced. These results indicate that the dimerization of Mycd family members is necessary for their Crm1-mediated nuclear export. We have recently found that the N-terminal region of Mycd consisting of 128 amino acids (Mycd N128) self-associates to Mycd via the central basic domain (CB), resulting in masking the Crm1-binding site. Such self-association of MRTF-A would be unlikely. In this study, we also revealed that the dimerization of Mycd was also necessary for this self-association. Wild-type Mycd activated SRF-mediated transcription more potently than Mycd lacking the Mycd N128 (Mycd ΔN128) did. These results suggest two possible functions of the Mycd N128: 1) stabilization of Mycd dimer to enhance SRF-mediated transcription and 2) positive regulation of the transactivation ability of Mycd. These findings provide a new insight into the functional regulation of Mycd family members.

Introduction

The myocardin (Mycd) family members, Mycd, myocardin-related transcription factor A (MRTF-A/MAL/MKL1), and MRTF-B (MRTF-B/MAL16/MKL2), are a specific coactivator of serum response factor (SRF), and play a critical role in the activation of SRF-mediated transcription (Wang et al., 2002; Miralles et al., 2003). Although Mycd is specifically expressed in cardiac and smooth muscles (Wang et al., 2001) and plays a critical role in the differentiation of these cell lineages (Wang et al., 2001; Li et al., 2003), Mycd is recognized as a master gene for SMC differentiation. This notion is based on the following facts. Mycd knockout mice are defective in vascular smooth muscle development and have died by embryonic day 10.5, but their heart development is unaffected (Li et al., 2003). The forced expression of Mycd induces the SMC differentiation program in 10T1/2 cells (Wang et al., 2003). Contrary, suppression of Mycd function results in the elimination of SMC gene expression (Wang et al., 2004; Hayashi et al., 2006). In contrast, MRTF-A/B are distributed in a wide variety of cells and tissues (Ma et al., 2001; Mercher et al., 2001; Miralles et al., 2003), and they participate in various biological processes and cell functions (Oh et al., 2005; Li et al., 2006; Medjkane et al., 2009). Notably, MRTF-A/B play a critical role in the transforming growth factor β1-induced epithelial-mesenchymal transition, which arises from the enhanced expression of several cytoskeletal proteins triggered by activation of a Rho signaling (Miralles et al., 2003; Morita et al., 2007). This biological process is also closely related to tumor cell migration and metastasis (Medjkane et al., 2009; Yoshio et al., 2010).

Mycd is constitutively located in the nucleus (Wang et al., 2001), whereas MRTF-A/B mainly reside in the cytoplasm, and transiently translocate to the nucleus in response to Rho signaling and actin polymerization (Miralles et al., 2003: Du et al., 2004; Hinson et al., 2007). The importin α/β1 heterodimer and Crm1 (exportin1) regulate their nuclear import and export, respectively (Nakamura et al., 2010; Pawlowski et al., 2010; Hayashi and Morita, 2013). In the N-termini of Mycd family members, three actin-binding RPEL motifs are present. Guettler et al. have reported that MRTF-A/B’s RPEL motifs have a much higher binding capacity for G-actin than that of Mycd does. And, they also have proposed a model for actin dynamics-dependent regulation of the MRTF-A’s nucleocytoplasmic shuttling; G-actin binding to MRTF-A’s RPEL motifs inhibits its nuclear import, resulting in promotion of its nuclear export (Guettler et al., 2008). Until now, we have revealed the following facts about the nuclear import of Mycd family members. The N-terminal basic domain (NB) of Mycd family members, which is also known as B2 (Miralles et al., 2003), is identified to be a binding site for the importin α/β1 heterodimer and functions as a nuclear localization signal (NLS). Of interest, the central basic domain (CB) is not involved in the nuclear import of wild-type Mycd, whereas it plays a significant role in the nuclear import of Mycd ΔN128 lacking the N-terminal 128 amino acids including the NB (Mycd N128) (Nakamura et al., 2010). Actin dynamics does not affect the interaction between Mycd and the importin α/β1 heterodimer and the nuclear localization of Mycd. On the other hand, G-actin significantly suppresses the interaction between MRTF-A/B and importin α/β1, and affects their nuclear import (Nakamura et al., 2010). The Treisman group has also demonstrated a similar competitive inhibition by G-actin (Pawlowski et al., 2010). Such inhibition by G-actin is a compelling model for understanding the actin dynamics-dependent nuclear import of MRTF-A/B, but dose not completely make clear the distinct subcellular localization of Mycd family members. Even in the absence of G-actin, Mycd shows a higher binding to importin α/β1 than MRTF-A/B do, and the serum-induced nuclear import of MRTF-A/B also correlates with their binding to the importin α/β1 heterodimers (Nakamura et al., 2010). Based on these evidences, we have proposed that the differences in the binding to importin α/β1 among Mycd family members play a vital role in their subcellular localization (Nakamura et al., 2010). Further, we have recently revealed a novel regulation in the nuclear export of Mycd family members. Two leucine-rich sequences in the N-terminal RPEL motif (L1) and in the Q-rich domain (Q) (L2) are present in the molecules of Mycd family members. Both of the L1 and the L2 are functional Crm1-binding sites and act as a nuclear export signal (NES) of MRTF-A. In contrast, in the case of Mycd, although only the L2 is the Crm1-binding site, the Crm1-mediated nuclear export system itself does not affect the subcellular localization of Mycd. This is because multiple inhibitory mechanisms isolate Mycd from such nuclear export system. Therefore, the binding of Mycd to Crm1 is much lower than that of MRTF-A. However, the L2 exceptionally acts as a functional NES of truncated Mycd, Mycd ΔN128. And, the binding of Mycd ΔN128 to Crm1 is about 4-fold higher than that of Mycd (Hayashi and Morita, 2013).

Dimer formation of SRF and Mycd family members are critically important to exhibit their functions (Norman et al., 1988; Wang et al., 2001; Miralles et al., 2003). Several studies have characterized the relationships between multi-merization of protein factors and their subcellular localization and functions. Most of them have demonstrated that the dimer formation suppresses their Crm1-mediated nuclear export (McBride and Reich, 2003; Wang et al., 2010). In this study, we reported the relationships between the dimer formation of Mycd family members and their nuclear export and a unique role of the N-terminal region of Mycd (Mycd N128) to exhibit its function.

Materials and Methods

Reagents and antibodies

The antibodies used in this study are as follows: anti-Flag M2-agarose, anti-Flag (F7425) and anti-α-tubulin (DM 1A) antibodies (Sigma-Aldrich, St. Louis, MO); anti-HA affinity matrix and anti-HA (3F10) antibody (Roche Applied Science, Mannheim, Germany); anti-DYKDDDDK (anti-Flag) antibody (Trans Genic, Kobe, Japan); anti-glutathione S-transferase (GST) antibody (Wako, Osaka, Japan). Secondary antibodies were conjugated to Alexa 568 (Molecular Probes, Eugene, OR). Leptomycin B (LMB) was purchased from Calbiochem (La Jolla, CA).

Plasmids

Construction of the expression plasmids used in this study has been described elsewhere (Hayashi et al., 2006; Nakamura et al., 2010; Hayashi and Morita, 2013). In brief, the cDNAs of mouse full-length Mycd (accession number AF384055), MRTF-A (accession number AK044188.1), and human Crm1 (accession number BC032847) were amplified by reverse transcription PCR and inserted into a mammalian expression plasmid, pCS2+, with the indicated tags. A series of expression plasmids for truncated and/or mutated derivatives of each of Mycd family members was constructed with PCR-mediated methods and their sequences were confirmed. Introduction of mutations in the coiled-coil domain (CC) of MRTF-A (CCmut1 and CCmut2) and the CC of Mycd (CCmut) (Wang et al., 2003) are shown in Fig. 1A and Fig. 2A, respectively. Mycd derivative with CBmut carries a mutated central basic domain (CB), in which all the lysine residues are changed to alanine. Introduction of mutations in the L2 of Mycd ΔN128 was performed as follow: the L2 sequence (LFLQL) was changed to LFLQA (L295A). Drs. Sekimoto and Yoneda (Osaka University Graduate School of Medicine) kindly provided a bacterial expression plasmid for the GST-fusion protein of the Flag-tagged constitutive active form of human Ran (RanQ69L) (Hieda et al., 1999). We also constructed a bacterial expression plasmid for the GST-fusion green fluorescence protein (GFP) (GST-GFP). Construction of SM22α promoter-luciferase gene (SM22P-Luc) was described in our previous report (Hayashi et al., 2006).

Fig. 1

Effect of the mutations in the CC on the subcellular localization of MRTF-A. (A) Schematic representation of MRTF-A. Two leucine-rich sequences (L1 and L2) are indicated by black vertical lines. Other abbreviations are as follows: NB, N-terminal basic domain; CB, central basic domain; Q, Q-rich domain; CC, coiled-coil domain; TA, transactivation domain. Sequences of the wild-type and the mutant CCs are shown. Red letters indicate possible candidates for the critical amino acids that are necessary for the dimerization of MRTF-A. Blue letters indicate the mutant amino acids in the CCmut1 and the CCmut2. (B) COS-7 cells were transfected with the expression plasmids for the indicated Flag-tagged MRTF-As under serum-stimulated conditions for 4 hours. For a further 20 hours, the cells were cultured under serum-starved conditions (serum-), and then were re-stimulated with 10% serum for 10 minutes (serum+). The cells were stained with anti-DYKDDDDK (Flag) antibody (red) and Hoechst 33258 (blue). Representative images from at least three independent experiments are shown. Bar=20 μm. (C) The images were quantified as described in Materials and Methods: nuclear-specific localization (N), diffuse distribution in the nucleus and the cytoplasm (NC), and cytoplasmic localization (C). Statistical differences were calculated using the Student t test. *, p<0.05 versus wild-type MRTF-A in the respective localization categories.

Fig. 2

Effect of LMB on the subcellular localization of MRTF-As. COS-7 cells were transfected with each of the indicated expression plasmids and were cultured under serum-stimulated conditions for 4 hours. Then, the cells were cultured under serum-starved conditions for a further 20 hours. They were incubated with vehicle (LMB-) or LMB (5 ng/ml) (LMB+) for the last 2 hours. The cells were then stained with anti-DYKDDDDK (Flag) antibody and Hoechst 33258. Representative images of the cells expressing wild-type MRTF-A and MRTF-A CCmut1 under LMB-treated conditions are shown (upper panel). Bar=20 μm. Other images without LMB are referred to the images shown in Fig. 1B. The images were quantified as described in the legend for Fig. 1 (lower graph). Statistical differences were calculated using the Student t test. *, p<0.05 versus MRTF-A (LMB-); #, p<0.05 versus MRTF-A CCmut1 (LMB-) in the respective localization categories.

Cell culture and transfection

COS-7 cells were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum. Transfection of the indicated expression plasmids was performed using Trans IT-LT1 (PanVera Corporation, Madison, WI) according to the manufacturer’s instructions. The transfected cells were then cultured under indicated conditions for 24 to 48 hours.

Immunocytochemistry

Cells were fixed with 4% formaldehyde for 30 minutes, and then were permeabilized and blocked with 0.1% Triton X-100, 10% normal goat serum, and 0.2% bovine serum albumin in phosphate-buffered saline for 1 hour at room temperature. Thereafter, the cells were incubated with an anti-DYKDDDDK antibody (Flag-antibody) for 1 hour followed by a secondary antibody conjugated to Alexa 568 with Hoechst 33258 for 1 hour at room temperature. Fluorescent images were collected with the aid of a Biorevo BZ-9000 fluorescence microscope (Keyence, Osaka, Japan). The expression patterns of Mycd family members were categorized into three groups: nuclear-specific localization (N; nuclear stain-ing>cytoplasmic staining), diffuse distribution in the nucleus and the cytoplasm (NC; nuclear staining=cytoplasmic staining), and cytoplasmic localization (C; nuclear staining<cytoplasmic staining). The expression pattern was defined as an NC when the differences in fluorescent intensities between the nucleus and the cytoplasm were within 10%. In each experiment (n=at least 3 independent experiments), 100–200 cells in several random fields were counted. The proportion of Flag-positive cells exhibiting the respective expression patterns defined above (mean±SEM) is calculated.

Protein-protein interaction analyses

A GST-Flag-RanQ69L protein expressed in Escherichia coli was purified with Glutathione Sepharose 4B (GE Healthcare BioSciences) according to the manufacturer’s instructions. The recombinant protein thus obtained was associated with GTP as described elsewhere (Hieda et al., 1999). In brief, purified GST-Flag RanQ69L protein was incubated for 1 hour in phosphate-buffered saline containing 2 mM GTP and 1 mM 2-mercaptoethanol on ice. Other proteins for immunoprecipitation (IP) analysis were prepared by using the TNT SP6 High-Yield Expression System based on an optimized wheat germ extract (Promega, Madison, WI) according to the manufacturer’s instructions. We preliminary confirmed that there was no significant protein crossreaction between each of antibodies against Mycd family members or nuclear import/export proteins and this wheat germ extract itself, and checked the expression levels of in vitro translated proteins by immunoblotting (IB) with the specified antibodies (data not shown). The IP/IB experiments were performed as described elsewhere (Hayashi and Morita, 2013). In brief, the indicated in vitro-translated proteins with or without the recombinant RanQ69L protein were first incubated with a control gel for 1 hour in IP buffer (20 mM Tris-HCl (pH 7.5), 0.5% Nonidet P-40, 150 mM NaCl, 5 mM MgCl₂, 1 mM EDTA, 50 mM NaF, 10 mM β-glycerophosphate, and proteinase inhibitors (c0mplete Mini; Roche Applied Science) to clear non-specific interactions, and then incubated with either a control gel or the specified affinity gel for 6 hours at 4°C. Proteins in the immunoprecipitates were detected by means of IB with the indicated antibodies. Target proteins were detected with a SuperSignal chemiluminescence detection kit (Pierce, Rockford, IL). Quantification of the respective IB signals’ intensities was performed with NIH image software. These interaction analyses were repeated at least three times. Representative data are shown.

Promoter assays

COS-7 cells were transfected with SM22P-Luc, pSVβ-gal (Promega) and the indicated expression plasmids, and then were cultured for 48 hours. The cell extracts prepared by using a passive lysis buffer (Promega) were subjected to luciferase assay with a luciferase assay kit (Promega). The relative promoter activity was expressed in luminescence units normalized to the β-galactosidase activity of pSVβ-gal in the cell extracts and the expression levels of Mycds in respective cell extracts were checked by means of IB. These assays were performed in triplicate and were repeated at least three times.

Results

Subcellular localization of MRTF-A and Mycd carrying the mutant CC

To assess a role of the CC in the subcellular localization of MRTF-A and Mycd, we transfected the expression plasmids for Flag-tagged mutant MRTF-As and Mycds in cultured cells, and analyzed their expressions by immunocytochemistry (Fig. 1, Fig. 2, and Fig. 3). In previous study (Miralles et al., 2003), the CC is revealed to be necessary for the dimerization of MRTF-A. However, little is known about which leucine residues in the CC are critical for its dimerization, because Miralles et al. have analyzed the MRTF-A dimerization using a mutant MRTF-A protein lacking the CC. In contrast, the critical amino acid residues in the CC for the dimerization of Mycd have been identified; I541, L574, and L538 play a vital role (Wang et al., 2001). Sequence comparison of the CC of MRTF-A and the CC of Mycd indicates that a sequence of MRTF-A, IEELTRML, is partly like a sequence of Mycd, INQLTWKL, and the CC of MRTF-A contains a distinct leucine-rich motif, LLRLQL, at its C-terminal portion (Fig. 1A and Fig. 3A). We, therefore, introduced two-types of mutations in the CC of MRTF-A (CCmut1 and CCmut2) to identify which leucine residues are critical for the dimer formation of MRTF-A (Fig. 1A). Although, in COS-7 cells, there were no significant differences among the proportions of cells showing a nuclear-specific localization of wild-type and mutant MRTF-As in both serum-starved and serum-stimulated conditions, obvious increases were observed in the proportions of cells showing a diffused distribution of the mutant MRTF-As in the nucleus and the cytoplasm. Accordingly, the proportions of cells showing their primarily cytoplasmic localization were decreased (Fig. 1B and C). Thus, the introduction of respective mutations in the CC similarly affected the subcellular localization of MRTF-A. However, compared with the effect of a specific inhibitor for Crm1, LMB, the effect by the CC mutations was moderate; treatment with LMB more markedly increased the nuclear accumulation of MRTF-A. The subcellular localization of MRTF-A and MRTF-A CCmut1 in COS-7 cells treated with LMB was comparable (Fig. 2). These results suggest that the mutations in the CC partially inhibit the cytoplasmic localization of MRTF-A.

Fig. 3

Effect of the mutations in the CC on the subcellular localization of Mycds. (A) Schematic representation of Mycd and Mycd ΔN128. Abbreviations indicating several domains on Mycd are described in the legend for Fig. 1. Sequences of the wild-type and the mutant CCs are shown. Red letters indicate the critical amino acids for Mycd dimerization. Blue letters indicate the mutant amino acids in the CCmut. (B) COS-7 cells were transfected with the expression plasmids for the indicated Flag-tagged Mycds under serum-stimulated conditions for 4 hours and were cultured for a further 20 hours. The cells were then stained with anti-DYKDDDDK (Flag) antibody (red) and Hoechst 33258 (blue). Representative images from at least three independent experiments are shown. Bar=20 μm. The images were quantified as described in the legend for Fig. 1. Statistical differences were calculated using the Student t test. *, p<0.05 versus wild-type Mycd; #, p<0.05 versus Mycd ΔN128 in the respective localization categories.

We also similarly analyzed a role of the CC in the sub-cellular localization of Mycd. In the vast majority of COS-7 cells expressing wild-type Mycd (Fig. 3) or mutant Mycd with mutations in the CC (Mycd CCmut) (data not shown), these proteins were primarily observed in the nucleus. In order to assess a role of the CC, we analyzed the subcellular localization of truncated Mycd and its mutant (Mycd ΔN128 and Mycd ΔN128/CCmut) (Fig. 3). This is because the expression of Mycd ΔN128 resulted in a slight but significant decrease and increase in the proportions of cells showing its primarily nuclear localization and its diffused distribution in the nucleus and the cytoplasm, respectively. The expression of Mycd ΔN128/CCmut in COS-7 cells resulted in an obvious increase in the proportion of cells showing its primarily nuclear localization and a decrease in the proportion of cells showing its diffused distribution in the nucleus and the cytoplasm. The change brought by mutations in the CC was equivalent to that by mutations in the L2 (L295A), which is the binding site for Crm1 (Fig. 3B and C). Compared with Mycd ΔN128, the Crm1-binding of Mycd ΔN128/L295A was markedly reduced (data not shown).

Roles of the CC in the interaction between Crm1 and MRTF-A or Mycd

In our recent study, we have revealed that MRTF-A or Mycd forms a complex containing Crm1 and GTP-bound constitutively active Ran (RanQ69L). In these assays, we used in vitro translated MRTF-A or Mycd proteins, and found no interaction between MRTF-A or Mycd and Crm1 in the absence of GTP-bound RanQ69L (Hayashi and Morita, 2013). We also confirmed that the artifactual effect of GST protein on the MRTF-A-Crm1 (Fig. 4) or the Mycd-Crm1 interaction (data not shown) was negligible level. The intensity of Flag-tagged MRTF-A protein associated with HA-tagged Crm1 protein in the presence of GST-GFP was less than 6% (5.8±1.6%) of that in the presence of GTP-bound RanQ69L. Based on these evidences, we performed the following analyses using in vitro translated proteins (Fig. 5). Both of the mutant MRTF-As (MRTF-A CCmu1 and CCmut2) exhibited a markedly reduced binding to Crm1: their bindings were decreased to 25.0±8.6% and 25.1±10.0% of the wild-type MRTF-A’s binding, respectively (Fig. 5A, left column, lanes IP2 and IP3). When GTP-bound RanQ69L was absent, no such binding occurred (data not shown). To confirm whether these mutant MRTF-As lose the dimerization ability, we examined the dimer formation between Flag-tagged MRTF-A CCmut1 or CCmut2 and HA-tagged wild-type MRTF-A (Fig. 5A, right column). As a result, the mutant MRTF-As exhibited severely reduced dimerization abilities (lanes IP2 and IP3). These results also suggest that either of the leucine-rich motifs in the CC of MRTF-A plays a vital role in its dimerization. Similarly, the binding of Mycd to Crm1 and its dimerization ability were nearly completely lost by the introduction of mutations in the CC (Fig. 5B, left and right columns). In accordance with lost of the dimerization abilities of MRTF-A and Mycd, their enhancement effect on the transcriptional activity of the SM22α promoter was markedly reduced (Fig. 6). However, significant change in the expression levels of endogenous SRF protein was not observed by forced expression of either wild-type or mutant MRTF-A and Mycd. Further, the expression of endogenous Mycd protein was scarcely detectable, but significant level of endogenous MRTF-A protein was expressed (data not shown). Since the transcriptional activity of the SM22α promoter was extremely low in the absence of forced expression of wild-type MRTF-A (Fig. 6), such endogenous MRTF-A hardly contributes the SRF-mediated transcription. And, the forced expression of mutant MRTF-A would never affect the function of endogenous MRTF-A because the mutant MRTF-A is lacking the dimerization ability (biologically inactive form).

Fig. 4

In vitro interaction between Crm1 and MRTF-A. Mixtures of in vitro translated HA-tagged Crm1, Flag-tagged MRTF-A, and purified GTP-bound Flag-tagged GST-RanQ69L or GST-GFP were immunoprecipitated with a control gel or anti-HA-affinity gel, and the resulting immunoprecipitates were analyzed by IB with the indicated antibodies. Positions of molecular weight markers are indicated on the side of IB panels in kilodaltons. Control experiments with the control gel did not show any significant signals on IB (data not shown). The respective IP/IB signals’ intensities were quantified as described in Materials and Methods.

Fig. 5

Effects of the mutations in the CC on the bindings of MRTF-A (A) and Mycd (B) to Crm1 and their dimerization. Mixtures of in vitro translated HA-tagged Crm1, purified GTP-bound Flag-tagged GST-RanQ69L, and each of the indicated in vitro translated Flag-tagged MRTF-As (A) or Flag-tagged Mycds (B) were subjected to IP/IB analyses as described in the legend for Fig. 4 (left columns). Mixtures of HA-tagged wild-type MRTF-A and Flag-tagged wild-type MRTF-A or each of Flag-tagged mutant MRTF-As (A) or mixtures of HA-tagged wild-type Mycd and Flag-tagged wild-type or mutant Mycd (B) were subjected to IP/IB analyses as described earlier (right columns). Positions of molecular weight markers are indicated on the side of IB panels in kilodaltons. Control experiments with the control gel did not show any significant signals on IB (data not shown). The respective IP/IB signals’ intensities were quantified as described in Materials and Methods. The percentage values on the tops of IP columns indicate their relative levels of binding to Crm1 normalized by the binding of wild-type MRTF-A, which was set at 100% (mean±SEM of the results from three independent experiments).

Fig. 6

Effect of the mutations in the CC of MRTF-A (A) or Mycd (B) on the activation of SRF-mediated transcription. COS-7 cells were transfected with 300 ng of SM22P-luc, 100 ng of pSV β-gal (marked with a plus (+)), and the indicated amounts (ng) of the respective expression plasmids for MRTF-As (A) or Mycds (B) and/or empty plasmid (pCS2+) (total: 1.0 μg plasmids/well of a 12-well culture plate). The luciferase activity without MRTF-A or Mycd expression was set at 100. Each value represents the mean±SEM of results from three independent experiments. The expression levels of wild-type and mutant MRTF-A proteins or those of wild-type Mycd and mutant Mycd were comparable (data not shown).

Novel roles of the N-terminal region of Mycd on its functional regulation

In our recent study, we have revealed that the N-terminal region of Mycd (Mycd N128) is intramolecularly interacted with Mycd, and this interaction inhibits the Mycd-Crm1 interaction by masking the L2. However, this type of intramolecular interaction of MRTF-A would be unlikely (Hayashi and Morita, 2013). Here, we suggest that this interaction also requires the dimer formation of Mycd. The basis of this conclusion is as follows. All of the deletion mutants derived from Mycd ΔN128, which are schematically presented in Fig. 7D, lost the binding to Mycd N128 (Fig. 7A, lanes IP2-IP4), suggesting that the C-terminal region containing the CC is necessary for the interaction with Mycd N128. Mycd ΔN128 CCmut hardly interacted with Mycd N128 (Fig. 7B, lanes IP3), indicating that the CC plays a critical role in this interaction. The CB also would play a role in this interaction, because the binding of Mycd ΔN128/ΔCB lacking the CB to Mycd N128 was markedly reduced (34.3±10.5% of the Mycd ΔN128’s binding) (Fig. 7B, IP2). Further, we found that the basic amino acid residues in the CB were essential for this interaction (Fig. 7C), suggesting that the acidic amino acids in Mycd N128 would be the main binding partners. The results of these protein-protein interaction analyses are summarized in Fig. 7D.

Fig. 7

Interaction between Mycd N128 and Mycd ΔN128. (A) Differences in the bindings of Mycd ΔN128 and its truncated derivatives to Mycd N128. (B and C) Effects of the deletion of the CB, the mutations in the CC (B), and the mutations in the CB (C) on the interaction between Mycd ΔN128 and Mycd N128. Mixtures of HA-tagged Mycd N128 and each of the indicated Flag-tagged Mycd derivatives were subjected to IP/IB analyses as described in the legend for Fig. 4. Positions of molecular weight markers are indicated on the side of IB panels in kilodaltons. Control experiments with the control gel did not show any significant signals on immunoblots (data not shown). The respective IP/IB signals’ intensities were quantified as described in Materials and Methods. The percentage values on the top of IP column indicate their relative levels of binding to Mycd N128 normalized by the binding of Mycd ΔN128, which was set at 100% (mean±SEM of the results from three independent experiments). (D) Schematic presentation of the structures of Mycd derivatives and summary of these protein-protein interactions. The relative bindings of Mycd derivatives to Mycd N128 are: –, no binding; +, weak binding; ++, strong binding.

We compared the transactivation abilities of Mycd and Mycd ΔN128 in COS-7 cells (Fig. 8). Even though the expression levels of Mycd were much lower than those of Mycd ΔN128 (more than 3-fold), the transactivation ability of Mycd was comparable with that of Mycd ΔN128, suggesting a significant role of the N-terminal region of Mycd to enhance its activity as a transcriptional cofactor.

Fig. 8

Effect of the Mycd N128 on the activation of Mycd/SRF-mediated transcription. COS-7 cells were transfected with 300 ng of SM22P-luc, 100 ng of pSV β-gal (marked with a plus (+)), and the indicated amounts (ng) of the respective expression plasmids for Mycds and/or empty plasmid (pCS2+) (total: 1.0 μg plasmids/well of a 12-well culture plate). The upper panel displays relative luciferase activities normalized to the β-galactosidase activity. The luciferase activity without Mycd expression was set at 100. Each value represents the mean±SEM of results from three independent experiments. Statistical differences were calculated using the Student t test. There were no statistical differences between Mycd and Mycd ΔN128 when 200 or 400 ng of respective expression plasmids were transfected, but Mycd only a little potently activated the luciferase activity than Mycd ΔN128 did when 600 ng of respective expression plasmids were transfected (*, p<0.05 versus lane 4). The expression levels of Mycds’ proteins in corresponding cell extracts were analyzed by IB with indicated antibodies (lower panel). The level of α-tubulin served as a loading control.

Discussion

Our present study has demonstrated that the dimerization of MRTF-A and Mycd is necessary for their interaction with Crm1 followed by their nuclear export. These are unique features of Mycd family members, because although several studies have characterized the relationships between multimerization of protein factors and their subcellular localization, most of them have reported that the dimer formation suppresses the Crm1-mediated nuclear export. For example, one study has found that an NES of STAT1 appears to be masked when its dimer is bound to DNA, but it becomes accessible to the Crm1 after dissociation from DNA (McBride and Reich, 2003). In another study, a nuclear accumulation of survivin, an oncofetal protein, is seen to be promoted by cyclic AMP response element-binding protein-dependent acetylation on lysine 129 (Wang et al., 2010). This molecular mechanism is regulated as follows: survivin acetylation at this position results in its homodimerization, while deacetylation promotes the formation of survivin monomers that are likely to interact with Crm1 and facilitate its nuclear export (Wang et al., 2010). Unlike in the case of Mycd family members, the NESs of these proteins are inaccessible to Crm1 only when they form a dimer.

Since the mutations in the CC never suppressed the nuclear import of MRTF-A and Mycd (Fig. 1, Fig. 2, and Fig. 3), the NB of MRTF-A and Mycd would be accessible to the importin α/β1 heterodimer when they are either a monomer or a dimer. However, their dimerization would induce a structural change in the L1 and/or the L2 to interact with Crm1. It is still obscure which conformation of the L1 and/or the L2 is necessary for the interaction with Crm1, but some changes in the structures of Mycd family members would be caused by their dimerization. The results shown in Fig. 6 indicate the basis of this speculation; the enhancement of SRF-mediated transcription by mutant MRTF-A or Mycd with mutations in the CC is markedly lower than that by wild-type MRTF-A or Mycd. Tertiary structure analysis is necessary to reveal these points. Although the nuclear export system does not affect the subcellular localization of Mycd, this system plays a critical role for the functional regulation of MRTF-A. Although our present data have clearly demonstrated that the dimerization of MRTF-A is critically necessary for its Crn1-mediated nuclear export, we have not yet got an answer to the biological significance of the dimerization-dependent nuclear export of MRTF-A. Specific binding of Crm1 to MRTF-A dimer might be a selection system to export the transcriptionally active MRTF-A molecules to the cytoplasm for suppression of SRF-mediated transcription. Moreover, it remains to be elucidated how the disruption of the CC of endogenous MRTF-A or Mycd affect their subcellular localization and the cell functions. Further study is also necessary to reveal these points.

We previously suggest that the intramolecular association of Mycd mediated by the Mycd N128 and the CB inhibits the Mycd-Crm1 interaction (Hayashi and Morita, 2013). In this study, we have proved that the dimer formation of Mycd would be also necessary for this self-association (Fig. 7). This finding leads to a new hypothesis that each of the N-terminal regions of dimerized Mycd would bind to the CB of neighboring Mycd in an intermolecular manner. If this hypothesis is correct, the N-terminal region of Mycd would play a role for stabilization of Mycd dimer. The promoter assay shown in Fig. 8 has revealed that forced expression of Mycd more potently enhances the prompter activity of the SM22α gene than that of Mycd ΔN128 does. Although the nuclear accumulation of Mycd was more likely than that of Mycd ΔN128 (Fig. 3), we speculate that such potent transcriptional activation by Mycd would be due to the ability of Mycd itself rather than its enhanced nuclear accumulation. This is because their expression levels are extremely different; the expression level of Mycd ΔN128 is about 3-fold higher than that of Mycd. The dimer formation of Mycd is critically important for activation of SRF-mediated transcription (Fig. 6 and Wang et al., 2001). Thus, the enhanced transcriptional activation by Mycd would be attributed to the stabilization of Mycd dimer brought by such self-association. Another possibility is that the N-terminal region of Mycd itself may positively affect the transactivation ability. Regarding the N-terminal region of Mycd, Ranson et al. have reported an inhibitory effect of this region on SRF-mediated transcription; the N-terminal region of a mutant Mycd acts in an autoinhibitory fashion to bind Mycd, resulting in severe suppression of the SRF-binding to Mycd (Ransom et al., 2008). However, such inhibitory effect is limited to a rare naturally occurring Mycd mutant (a missense mutation at codon 259 resulting in a lysine to arginine substitution at codon 259 (K259R)). In the case of wild-type Mycd, its self-association would never disrupt SRF-mediated transcription (Fig. 8). Although it remains to be elucidated how the N-terminal region of Mycd actually regulates the functions of endogenous Mycd, taken together with our previous findings (Nakamura et al., 2010; Hayashi and Morita, 2013), the N-terminal region of Mycd plays multiple roles: the regulation of the importin α/β1-mediated nuclear import, the inhibitory regulation of the Crm1-mediated nuclear export, and the enhancement of its transactivation ability.

In conclusion, our novel findings demonstrated in this study are as follows. 1) Dimer formation of MRTF-A and Mycd is critically important for their interactions with Crm1 followed by their nuclear export. 2) Dimerization of Mycd ΔN128 is also necessary for the interaction with Mycd N128. The latter finding leads to the following possibilities. The Mycd N128 plays a role for stabilization of Mycd dimer to enhance its transactivation ability or the Mycd N128 itself positively regulates the ability. Our present study provides new insight for the functional regulation of Mycd family members.

Acknowledgments

This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (23590332 to K. H.).

References

• Du, K.L., Chen, M., Li, J., Lepore, J.J., Mericko, P., and Parmacek, M.S. 2004. Megakaryoblastic leukemia factor-1 transduces cytoskeletal signals and induces smooth muscle cell differentiation from undifferentiated embryonic stem cells. J. Biol. Chem., 279: 17578–17586.
• Guettler, S., Vartiainen, M.K., Miralles, F., Larijani, B., and Treisman, R. 2008. RPEL motifs link the serum response factor cofactor MAL but not myocardin to Rho signaling via actin binding. Mol. Cell. Biol., 28: 732–742.
• Hayashi, K., Nakamura, S., Nishida, W., and Sobue, K. 2006. Bone morphogenetic protein-induced MSX1 and MSX2 inhibit myocardin-dependent smooth muscle gene transcription. Mol. Cell. Biol., 26: 9456–9470.
• Hayashi, K. and Morita, T. 2013. Differences in the nuclear export mechanism between myocardin and myocardin-related transcription factors A. J. Biol. Chem., 288: 5743–5755.
• Hieda, M., Tachibana, T., Yokoya, F., Kose, S., Imamoto, N., and Yoneda, Y. 1999. A monoclonal antibody to the COOH-terminal acidic portion of Ran inhibits both the recycling of Ran and nuclear protein import in living cells. J. Cell Biol., 144: 645–655.
• Hinson, J.S., Medlin, M.D., Lockman, K., Taylor, J.M., and Mack, C.P. 2007. Smooth muscle cell-specific transcription is regulated by nuclear localization of the myocardin-related transcription factors. Am. J. Physiol. Heart Circ. Physiol., 292: H1170–H1180.
• Li, S., Wang, D.Z., Wang, Z., Richardson, J.A., and Olson, E.N. 2003. The serum response factor coactivator myocardin is required for vascular smooth muscle development. Proc. Natl. Acad. Sci. USA, 100: 9366–93670.
• Li, S., Chang, S., Qi, X., Richardson, J.A., and Olson, E.N. 2006. Requirement of a myocardin-related transcription factor for development of mammary myoepithelial cells. Mol. Cell. Biol., 26: 5797–5808.
• Ma, Z., Morris, S.W., Valentine, V., Li, M., Herbrick, J.A., Cui, X., Bouman, D., Li, Y., Mehta, P.K., Nizetic, D., Kaneko, Y., Chan, G.C., Chan, L.C., Squire, J., Scherer, S.W., and Hitzler, J.K. 2001. Fusion of two novel genes, RBM15 and MKL1, in the t(1;22)(p13;q13) of acute megakaryoblastic leukemia. Nat. Genet., 28: 220–221.
• McBride, K.M. and Reich, N.C. 2003. The ins and outs of STAT1 nuclear transport. Sci. STKE., 2003: RE13.
• Medjkane, S., Perez-Sanchez, C., Gaggioli, C., Sahai, E., and Treisman, R. 2009. Myocardin-related transcription factors and SRF are required for cytoskeletal dynamics and experimental metastasis. Nat. Cell Biol., 11: 257–268.
• Mercher, T., Coniat, M.B., Monni, R., Mauchauffe, M., Nguyen, Khac, F., Gressin, L., Mugneret, F., Leblanc, T., Dastugue, N., Berger, R., and Bernard, O.A. 2001. Involvement of a human gene related to the Drosophila spen gene in the recurrent t(1;22) translocation of acute megakaryocytic leukemia. Proc. Natl. Acad. Sci. USA, 98: 5776–5779.
• Miralles, F., Posern, G., Zaromytidou, A.I., and Treisman, R. 2003. Actin dynamics control SRF activity by regulation of its coactivator MAL. Cell, 113: 329–342.
• Morita, T., Mayanagi, T., and Sobue, K. 2007. Dual roles of myocardin-related transcription factors in epithelial mesenchymal transition via slug induction and actin remodeling. J. Cell Biol., 179: 1027–1042.
• Nakamura, S., Hayashi, K., Iwasaki, K., Fujioka, T., Egusa, H., Yatani, H., and Sobue, K. 2010. Nuclear import mechanism for myocardin family members and their correlation with vascular smooth muscle cell phenotype. J. Biol. Chem., 285: 37314–37323.
• Norman, C., Runswick, M., Pollock, R., and Treisman, R. 1988. Isolation and properties of cDNA clones encoding SRF, a transcription factor that binds to the c-fos serum response element. Cell, 55: 989–1003.
• Oh, J., Richardson, J.A., and Olson, E.N. 2005. Requirement of myocardin-related transcription factor-B for remodeling of branchial arch arteries and smooth muscle differentiation. Proc. Natl. Acad. Sci. USA, 102: 15122–15127.
• Pawlowski, R., Rajakylä, E.K., Vartiainen, M.K., and Treisman, R. 2010. An actin-regulated importin α/β-dependent extended bipartite NLS directs nuclear import of MRTF-A. EMBO J., 29: 3448–3458.
• Ransom, J.F., King, I.N., Garg, V., and Srivastava, D. 2008. A rare human sequence variant reveals myocardin autoinhibition. J. Biol. Chem., 283: 35845–35852.
• Wang, D., Chang, P.S., Wang, Z., Sutherland, L., Richardson, J.A., Small, E., Krieg, P.A., and Olson, E.N. 2001. Activation of cardiac gene expression by myocardin, a transcriptional cofactor for serum response factor. Cell, 105: 851–862.
• Wang, D.Z.Li, S., Hockemeyer, D., Sutherland, L., Wang, Z., Schratt, G., Richardson, J.A., Nordheim, A., and Olson, E.N. 2002. Potentiation of serum response factor activity by a family of myocardin-related transcription factors. Proc. Natl. Acad. Sci. USA, 99: 14855–14860.
• Wang, H., Holloway, M.P., Ma, L., Cooper, Z.A., Riolo, M., Samkari, A., Elenitoba-Johnson, K.S., Chin, Y.E., and Altura, R.A. 2010. Acetylation directs survivin nuclear localization to repress STAT3 oncogenic activity. J. Biol. Chem., 285: 36129–36137.
• Wang, Z., Wang, D.Z., Pipes, G.C., and Olson, E.N. 2003. Myocardin is a master regulator of smooth muscle gene expression. Proc. Natl. Acad. Sci. USA, 100: 7129–7134.
• Wang, Z., Wang, D.Z., Hockemeyer, D., McAnally, J., Nordheim, A., and Olson, E.N. 2004. Myocardin and ternary complex factors compete for SRF to control smooth muscle gene expression. Nature, 428: 185–189.
• Yoshio, T., Morita, T., Tsujii, M., Hayashi, N, and Sobue, K. 2010. MRTF-A/B suppress the oncogenic properties of v-ras- and v-src-mediated transformants. Carcinogenesis, 31: 1185–1193.