Background: Serum stimulation leads to the activation of various signal transduction pathways in cells, and the resultant signals are integrated into the serum response factor (SRF)-dependent transcription of immediate-early genes such as c-fos.
Results: To further characterize this response, we investigated the mechanism which controls serum response transcription in cultured human cells. Frequency of PML (promyelocytic leukaemia)-nuclear bodies (NBs) formation increases shortly after serum stimulation, probably facilitating the interaction of SRF and CBP acetyltransferase at the NBs. PML modulates SRF-mediated c-fos promoter activities upon addition of serum to cells or expression of constitutively active Rho family GTPases. We mapped the region in the SRF that interacts with PML to the C-terminal transactivation domain. An SRF mutant deleted of the transactivation domain neither co-localizes with CBP in NBs nor fulfills its transcriptional role. Under conditions of serum stimulation, the formation of NBs coincides with the immediate-early expression of the endogenous c-fos gene in fibroblasts and in all-trans retinoic acid-treated acute promyelocytic leukaemia NB4 cells.
Conclusion: These data provide an insight into the involvement of NBs in modulating the transcription of serum-induced immediate-early genes.
Growth factor- or mitogen-stimulated transcription of a subset of genes plays a fundamental role in regulating cellular responses such as proliferation, differentiation, growth arrest, senescence and apoptosis. Serum response factor (SRF) is a MADS (MCM1, Agamous, Deficiens and SRF) box transcription factor required for the activation of immediate-early genes including c-fos and Egr-1 (Norman et al. 1988). These gene promoters retain the serum response element (SRE), one of the major destinations for growth signals. SRF and members of the TCF (ternary complex factor) family of Ets domain proteins form transcription factor complexes on the c-fos SRE for its transactivation, together with CREB–binding protein (CBP) as a co-activator (Arias et al. 1994; Ramirez et al. 1997). The activity of the TCF is controlled through phosphorylation of the C-terminal transcriptional activation domain by MAP kinases (Treisman 1996). In contrast, SRF is activated by serum or mitogens, principally through pathways involving Rho-family GTPase Rho A (Hill et al. 1995) and phosphatidylinositol 3-kinase (PI3K) (Wang et al. 1998), and these signal transducers appear to act cooperatively. The activation of c-fos SRE by Rho-family GTPases requires histone H4 hyperacetylation that is induced via the stress-activated protein/Jun kinase (SAPK/JNK)-dependent or -independent pathways (Alberts et al. 1998). However, the mechanisms by which SRF activity is regulated in vivo remain unclear.
PML functions in cell growth and differentiation mechanisms, and it also acts as a target for viral infection or as a mediator of interferon and the immune system (Lin et al. 1999; Zhong et al. 2000; Hatta & Fukamizu 2001). In addition, the t(15;17) translocation in acute promyelocytic leukaemia (APL) yields a fusion of PML with retinoic acid receptor-α, which produces an oncogenic effect. Some parts of PML localize to specified subnuclear domains called nuclear bodies (NBs), also known as Kremer bodies, ND10 or PODs (PML oncogenic domains), which are considered, at least in part, as an important structure for fulfilling PML's multiple roles. Recently, transcription factors and their co-regulators have been identified as associating with PML, emphasizing the transcriptional roles of NBs (Zhong et al. 2000; Doucas et al. 1999). PML promotes transcriptional activation by p53, probably through the acetylation of p53 by CBP at the NBs (Pearson et al. 2000), showing that PML participates in the formation of potential transcription complexes. In addition, CBP has been reported to interact with the hypophosphorylated form of RNA polymerase II (pol II) and PML in a subset of NBs (von Mikecz et al. 2000), suggesting an involvement of NBs in the transcriptional initiation. Nascent pol II transcripts were shown to be within the NBs (LaMorte et al. 1998) or in the periphery of the bodies (Boisvert et al. 2000). These findings support the idea that NBs contribute to the formation of a nuclear environment for the pol II-mediated transcription.
Serum stimulation provides an essential basis for a variety of the cell regulation and is integrated into the SRF-mediated transcriptional activation of many downstream genes (Treisman 1996). It is therefore of considerable interest to study the mechanism and control of serum response transcription via SRF. During an investigation of serum-stimulated human fibroblasts, we discovered that the frequency of NBs formation increases shortly after serum induction, and that PML, CBP and SRF can probably interact in the NBs. We present several lines of evidence indicating the involvement of NBs in the cellular serum response. Because the serum-stimulated expression of immediate-early genes is conserved regardless of the cell types, our observation may offer a general mechanism for regulating serum response in a wide range of cell types.
PML cooperates with SRF in cellular serum response
To elucidate the involvement of PML in serum response, we first investigated the distribution pattern of endogenous PML-nuclear bodies (NBs) and SRF (Fig. 1A). Cultured human fibroblasts were serum-starved for 24 h and then stimulated by administration of 20% foetal calf serum for the indicated periods. Immunofluorescent analysis showed that serum induction promptly increased NBs in both number and size at 15–30 min after stimulation (upper panel). Quantitatively, ≈ 70% of the stimulated cells showed twice the number of NBs in the starved state. It is likely that pre-existing PML may be recruited into NBs, and the dynamics of the NB is analogous to previous reports of NBs induced by interferon (Lavau et al. 1995) or by oncogenic Ras (Pearson et al. 2000; Ferbeyre et al. 2000). SRF showed diffuse localization in the nuclei of the starved cells and then a portion of the SRF concentrated into the enlarged NBs at 15–60 min after stimulation (middle panel). At 90 min, SRF tended to be delocalized from the NBs. Thus, an increased frequency of NB formation was concomitant with the translocation of SRF to the NBs, suggesting that PML may recruit SRF to NBs. To test this possibility, PML was over-expressed in HeLa cells, and the cells were counterstained for PML and SRF. Endogenous SRF was found to form recognizable foci that were co-localized with the enlarged NBs containing PML (Fig. 1B). In contrast, over-expression of green fluorescent protein did not cause the recruitment of SRF into the NBs, emphasizing a PML-dependent recruitment of SRF to the NBs.
We examined whether endogenous PML and SRF form a complex upon a serum-stimulation (Fig. 1C). The starved fibroblasts were serum-stimulated for 40 min and then cross-linked with dimethyl 3,3′-dithiobispropionimidate-2HCl (DTBP) prior to cell lysis to preserve intracellular protein complexes in the lysate (Pearson et al. 2000). After immunoprecipitation with either anti-PML or anti-SRF antibodies, the immunoprecipitates were probed with anti-PML antibodies. PML was present in the SRF- as well as in the PML-immunoprecipitates (lanes 2 and 3). Immune complexes with control IgG did not associate with PML (lane 1). We were hardly able to detect the presence of a complex containing PML and SRF in the starved condition (data not shown). Thus, SRF interacts with PML in response to serum induction in vivo.
To ensure the effect of the serum stimulation in this study, we examined the expression pattern of the endogenous c-fos gene in serum-stimulated fibroblasts, using a reverse transcription-PCR method (Fig. 1D). The transient induction of c-fos mRNA was detected at 30 min, and almost completely repressed at 60 min after stimulation. In addition, protein analysis showed that c-Fos was induced at 60 min after serum stimulation in both fibroblasts and HeLa cells (data not shown).
PML modulates SRF-dependent transcription
The influence of PML on serum response transcription was examined in HeLa cells which have been proved to maintain serum-induced signalling pathways to target genes (Wang et al. 1998). HeLa cells were transfected with c-fos SRE-containing luciferase reporter construct driven by SRE from the c-fos gene, and increasing amounts of PML-expressing plasmid (Fig. 2A). The transfected cells were serum-starved and stimulated by adding serum for 3 h for luciferase analysis. The addition of serum alone increased the c-fos SRE transcription about threefold, and expression of PML further enhanced the luciferase activities in a dose-dependent manner, in contrast to the results from the basal starved condition.
It is known that the Rho family of small G-proteins affects c-fos expression and SRF-dependent transcription via LIM kinase-1-regulated actin dynamics (Sotiropoulos et al. 1999). Active forms of Rho A can potentiate SRF independently of TCF without the presence of extracellular stimuli (Hill et al. 1995; Wang et al. 1998). To test the transcriptional role of PML, we examined whether PML modulates c-fos SRE activation by introducing a constitutively active form of Rho A (Rho A.V14). Expression of Rho A.V14 increased the luciferase activities more than sevenfold, and the co-transfection of PML further augmented Rho A-activated transcription (Fig. 2A). In addition, Rac1 and cdc42, other Rho family GTPases, have been reported to activate SRE in NIH3T3 and HeLa cells. Ras is also known to activate c-fos SRE, both via the Raf-ERK pathway and via the pathway involving Rho family members (Wang et al. 1998). PML similarly enhanced SRE activation by either Rac1.V12, cdc42.V12 or Ras.V12 (data not shown). These data suggested that PML can regulate c-fos SRE activation via various pathways. Because p53, one of well-known constituents in NBs, is functionally impaired in HeLa cells, MCF-7 cells expressing wild-type p53 were used in Fig. 2B. Rho A-induced c-fos SRE transactivation was also enhanced by expression of PML in the cell line, indicating that the effect of PML is not specific to the cell type. In addition, PML alone had little transactivation activity, as shown in the basal state. As a negative control, the use of either mutated SRE in the reporter construct or C-terminal deletion mutants of SRF gave less response to PML co-expression (Figs 2C and 4D, respectively).
The SRE consists of two distinct binding sites: the CArG box, which binds SRF, and the Ets box, which binds a TCF family of proteins. To confirm the cooperative role of SRF and PML, we used wild- and mutant-type SRE reporter constructs (thereafter referred as wt SRE and mCArG SRE, respectively). The mCArG SRE, which abolishes SRF binding and therefore also eliminates TCF binding (Wang et al. 1998). PML was expressed in HeLa cells in combination with Rho A.V14, and the luciferase activities were determined in both wild- and mutant-type reporter constructs (Fig. 2C). The alteration of the CArG box markedly diminished both Rho A-dependent c-fos SRE activation and enhancement of the transcription by PML. This implied the primary importance of SRF in SRE transactivation. To examine whether PML directly enhances the transcriptional activity of SRF, we used a pSRF-Luc reporter construct containing the CArG box, but not the Ets box, derived from the c-fos SRE promoter. This reporter enabled us to elucidate the transcriptional activity of SRF without any influence by TCF. The pSRF-Luc was introduced into HeLa cells in combination with the expression of Myc-tagged SRF, PML and Rho A.V14 (Fig. 2D, left panel). SRF strongly increased the luciferase activity in basal as well as in Rho A-stimulated states. PML augmented the basal transcription and further enhanced the Rho A-stimulated transcription in the presence of SRF. In addition, PML enhanced the luciferase activities from the pSRF-Luc in the absence of over-expressed SRF, indicating that PML affects endogenous SRF (Fig. 2D, right panel). Accordingly, these data suggest that PML modulates the SRF transactivation.
Complex formation of PML, SRF and CBP in NBs
CBP functions as a general transcriptional coactivator and has been shown to interact with PML in NBs (Lin et al. 1999; Zhong et al. 2000). It has also been shown that CBP contributes to SRF's roles in F9 cells (Ramirez et al. 1997). To investigate whether PML mediates the interaction of SRF with CBP in NBs, we visualized the localization of SRF and CBP in the nucleus (Fig. 3A). The expression of PML was used in HeLa cells to test the effect of the enlargement of NBs on CBP and SRF (d–f and j–l). In the upper panels, endogenous PML and CBP were co-localized at NBs (a–c), and endogenous CBP was recruited into NBs by over-expressed PML (d–f). In the lower panels, enlarged NBs induced the co-localization of CBP and GFP-SRF at these sites (j–l), and this compared with the lack of specific association between these two proteins in the absence of increased NB (g–i). These data suggest the possibility that PML can recruit both SRF and CBP to enlarged NBs. Next, we tested whether CBP affects SRF-mediated transcription. HeLa cells were transfected with pSRF-Luc together in the presence of increasing amounts of CBP. As expected, CBP efficiently enhanced the luciferase activities with increased serum or Rho A.V14 expression (Fig. 3B), supporting the role of CBP in the SRF-mediated transcription. CBP alone moderately increased transcription in the basal condition.
To investigate whether PML is associated with SRF and CBP, an in vitro pull-down analysis was carried out (Fig. 3C). Glutathione S-transferase (GST) and GST-fused PML (amino acids 5–452) were immobilized on glutathione-agarose beads and were incubated with the nuclear extract from HeLa cells. Both SRF and CBP were found to bind GST-PML, indicating that PML forms complexes with SRF and CBP. Increased amounts of GST-PML further augmented the association with SRF. To demonstrate the complex formation of PML, SRF and CBP in cells, Flag-tagged PML and Rho A.V14 were expressed in HeLa cells for an immunoprecipitation analysis (Fig. 3D). Endogenous SRF and CBP were detected in the Flag (PML)-immunoprecipitates. SRF and PML were also present in the CBP-immunoprecipitates. Furthermore, PML and CBP were co-immunoprecipitated with SRF. The amount of SRF in both Flag (PML)- and CBP-immunoprecipitates appeared to be relatively low, probably due to the fact that PML and CBP interact with multiple other proteins. Neither SRF, PML nor CBP was detected in the immunoprecipitates with control IgG. We previously confirmed that the antibodies used here did not cross-react each other by Western blot analysis (data not shown). These data emphasized the existence of a complex containing SRF, PML and CBP in vivo.
Interaction of PML, SRF and CBP is crucial for c-fos SRE activation
To examine the biological significance of the interaction between PML, SRF and CBP in the NBs, we constructed two plasmids to express SRF mutants deleted of the C-terminal transactivation domain (Johansen & Prywes 1993): Myc-tagged SRF lacking amino acids 414–508 (SRF(ΔC413)) and 339–508 (SRF(ΔC338)) (Fig. 4F). Because of the presence of the intact MADS domain, these SRF mutants retain their ability to bind SRE and to form heterodimers with endogenous SRF or homodimers with themselves in cells. Wild- and two mutant-type SRFs were expressed together with PML in HeLa cells for immunostaining with anti-Myc and anti-CBP antibodies (Fig. 4A). Wild-type SRF was mostly co-localized with CBP in NBs. In contrast, both SRF(ΔC413) and SRF(ΔC338) were found to reduce the association with NBs and to be localized along the nuclear periphery in about 80–90% of the transfected cells. Thus, the transactivation domain is crucial for the interaction between SRF and CBP in NBs. To elucidate a role of the transactivation domain of SRF in associating with PML, an immunoprecipitation analysis was performed using cell lysates from HeLa cells expressing either Myc-tagged wild-type SRF, SRF(ΔC413) or SRF(ΔC338) together with Flag-tagged PML (Fig. 4B). The Myc-immunoprecipitates were probed with anti-PML antibodies. PML existed in the immune complexes with wild-type SRF (lane 6). In contrast, both C-terminal deletions remarkably reduced their interaction with PML (lanes 4 and 5), which was well correlated with the impaired localization of these mutants to NBs (Fig. 4A). PML was not found in the control precipitates (lane 7). Wild- and mutant-type SRFs were immunoprecipitated with anti-Myc antibodies to similar extents (lower panel). These data suggest that the C-terminal region of SRF is important for the interaction between SRF and PML.
To determine the domain of SRF that interacts with PML, (His)6-Flag-tagged portions of SRF were bacterially prepared for an in vitro pull down assay; termed SRF(1–338), SRF(301–508), SRF(301–451) and SRF(301–395) containing the region of amino acids 1–338, 301–508, 301–451 and 301–395 of the protein, respectively. The recombinant proteins were incubated with GST-PML immobilized on glutathione–agarose beads, and bound proteins were analysed with anti-Flag antibodies (Fig. 4C). Both SRF(301–508) and SRF(301–451) were efficiently concentrated on PML, while SRF(1–338) and SRF(301–395) showed no binding activity to PML. This data suggested that SRF directly binds PML via the region of amino acids 396–451 in the transactivation domain. The fact that SRF(1–338) did not interact directly with PML in vitro suggests the possible dimerization of a part of SRF(ΔC338) with endogenous SRF in vivo, leading to the weak association of this mutant with NBs or PML in cells (Fig. 4A–B).
We further studied the effect of these SRF mutants on c-fos SRE transcription in HeLa cells. Wild- and mutant-type SRFs were expressed with the pSRF-Luc reporter construct, in combination with PML and/or Rho A (Fig. 4D). Wild-type SRF-mediated transcription was activated by PML and/or Rho A. The transcriptional activities of SRF(ΔC338) itself were at lower level, and PML did not stimulate the activity of SRF(ΔC338), in contrast to the case of SRF(ΔC413). These may be explained by the fact that SRF(ΔC413) retains the N-terminal half of the transactivation domain, and that SRF(ΔC338) lacked the entire transactivation domain (Fig. 4F). In addition, the expression of the GAL4-fused C-terminal region (amino acids 266–508) of SRF, termed SRF (ΔN266), in HeLa cells decreased SRF-dependent activation in the presence of either PML, Rho A, or both (Fig. 4E), suggesting that SRF (ΔN266) can act as a dominant negative form for endogenous SRF. Taken together, the transactivation domain of SRF is important for localization to NBs, interaction with PML, and its transcriptional role.
The reorganization of NBs coincides with immediate-early c-fos induction in vivo
In APL-derived NB4 cells with t(15;17) translocation, NBs structure is altered into numerous small speckles by the presence of a fusion of PML with retinoic acid receptor-α, and treatment with all-trans retinoic acid (ATRA) reorganizes the NBs to a normal structure (Dyck et al. 1994; Weis et al. 1994). We investigated the effect of ATRA treatment and subsequent serum stimulation on the formation of NBs in NB4 cells. The cells were treated with 1 µm ATRA for 72 h, resulting in regaining normal NBs (Fig. 5A). The ATRA-treated cells were then serum-starved for 24 h prior to serum induction. During the serum starvation, ATRA was withdrawn to minimize an influence on unknown signal pathways. After serum stimulation for the indicated periods, the cells were subjected to immunofluorescence analysis (Fig. 5B). Reorganized NBs were maintained during the serum starvation (a). Serum stimulation induced an increased frequency of formation of NBs by 60 min (b and c). Most of the stimulated cells showed threefold the number of NBs in the starved condition (data not shown). A portion of SRF was found to concentrate in some NBs (e and h), similarly to cultured fibroblasts (Fig. 1A). The small speckles in primary NB4 cells appeared to be little changed by serum stimulation (data not shown).
Next, we examined whether endogenous c-fos induction in NB4 cells was affected by the existence of reorganized NBs, using a reverse transcription-PCR (Fig. 5C). Full induction of the c-fos gene was observed in ATRA-treated cells with reorganized NBs as quickly as 30 min after serum stimulation, as were the cases in cultured fibroblasts (Fig. 1D) and other previous report (Hill & Treisman 1995). In contrast, ATRA-untreated control cells showed a delay in activation of the c-fos gene, and c-fos mRNA was detected at 60 min after stimulation. We confirmed that treatment with ATRA did not affect c-fos expression itself in cultured human fibroblasts (Fig. 5D). The fibroblasts were starved for 24 h and then serum-stimulated for the indicated periods in the presence and absence of ATRA. The c-fos mRNA was detected at 30 min and then repressed at 60 min after the stimulation in both ATRA-treated and untreated fibroblasts. Our observation that ATRA did not change c-fos induction pattern by serum in fibroblasts was consistent with previous reports (Cosgaya et al. 1998; Smith et al. 2001; Sejal & Niles 1997). Therefore, the close correlation between formation of NBs and c-fos induction suggests that an increased frequency of NB formation is required for the initiation of a proper timing of serum response transcription.
Cooperation of PML, SRF and CBP in NBs for serum response
We have presented the importance of an increased NB formation as well as the interaction of SRF and CBP within the NBs for serum response. PML exerts biological functions, at least in part, through the formation of NBs (Zhong et al. 2000), but a mechanism for an increase of NBs remains to be elucidated. Müller et al. reported that arsenic trioxide (As2O3) promotes the formation of NBs by facilitating the SUMO-1 modification of PML (Müller et al. 1998). As2O3 is also known to be a strong inducer of MAP kinase signalling cascades (Chen et al. 1998; Liu et al. 1996; Ludwig et al. 1998; Porter et al. 1999). Like As2O3, serum stimulation activates MAP kinase pathways through Ras and Rho-family GTPases, and PI3K. Our data showed that serum stimulation augments the number and size of NBs in cultured fibroblasts (Fig. 1) and ATRA-treated APL cells (Fig. 5). Collectively, increased frequencies of NBs may be a target downstream of the activated MAP kinase cascades. In fact, specific inhibitors of PI3K and MEK disturbed the NB formation and altered c-fos gene expression by serum stimulation (data not shown), supporting the idea of involvement of MAP kinases in the formation of the NBs. Although the phosphorylation of PML was reported to be correlated with the delocalization from NBs (Everett et al. 1999), the role of the modification of PML remains to be further determined.
With increased formation of NBs in frequency after serum stimulation, a part of SRF was translocated into the NBs, suggesting that PML can functionally associate with SRF. PML enhanced not only c-fos SRE activation upon serum treatment, Ras and Rho family GTPases, but also SRF-dependent transactivation. In addition, CBP cooperated with SRF in the transactivation, and PML can form a complex with SRF and CBP in enlarged NBs. Therefore, the compartmentalization of SRF and CBP by PML in the NBs is likely to facilitate their cooperative interaction, leading to the enhancement of SRF-mediated transcription. CBP possesses an intrinsic histone acetyltransferase activity, and it associates with transcriptional co-activators and the hypophosphorylated form of pol II in NBs (von Mikecz et al. 2000). There are nascent pol II transcripts within the NBs or in the periphery of the bodies (LaMorte et al. 1998; Boisvert et al. 2000), suggesting the necessity of a transcription-initiating complex assembled near the NBs. The observation that the hyperacetylation of histone H4 is important for SRF-mediated c-fos activation (Alberts et al. 1998) may be explained by the fact that PML-CBP-SRF complexes are formed after serum stimulation. Collectively, we propose that the association of PML, CBP and SRF in NBs is one of major regulatory mechanisms in serum response transcription. By directly interacting with the transactivation domain of SRF, PML may function as a scaffold protein to mediate the interaction of SRF with co-activators such as CBP, or to increase the local concentration of the transcription factor and its coactivator complex. Therefore, NBs may play a role in PML-CBP-SRF connection and SRF-mediated transactivation of target genes during cellular serum response.
Role of NBs in immediate-early gene expression by serum stimulation
Immediate-early genes are normally induced within 30 min after the presence of extracellular stimuli, as shown in cultured fibroblasts. In NB4 cells, however, there was a significant difference in chromosomal c-fos induction between untreated and ATRA-treated conditions under serum stimulation. ATRA-induced reorganization of NBs in NB4 cells is likely to restore the maximal expression of c-fos mRNA at 30 min after serum addition. These results suggested that reorganized NBs are responsible for proper serum response transcription. Furthermore, we found that c-fos expression in fibroblasts delayed to 60 min after adding serum under the inhibition of PI3K or MEK, and that these inhibitions impaired the formation of NBs (data not shown). Thus, the condition of NBs is correlated with SRF-mediated c-fos transactivation in vivo. PML seems to delicately modulate the transcriptional initiation of serum-induced genes through forming NBs.
Co-repressors such as c-Ski, nuclear hormone receptor co-repressor (N-CoR), silencing mediator of retinoic acid and thyroid hormone receptors (SMRT), Sin3A and histone deacetylases have been recently reported to associate with PML-mediated transcriptional repression (Khan et al. 2001; Wu et al. 2001). Among them, SMRT is shown to interact with SRF and inhibit its transactivation following serum induction (Lee et al. 2000). The combination of rapid activation and subsequent repression of the c-fos gene suggests a dynamic association of SRF with co-activator and co-repressor complexes. A portion of SRF was found to be localized in NBs at 15–60 min after serum stimulation, suggesting an involvement of PML in both transcriptional activation and repression by this molecule. Because PML interacts directly with SRF, CBP and several constituents of co-repressor complexes, it is possible that SRF cooperates with co-activator and co-repressor complexes in NBs. In addition, delayed activation of the c-fos gene coincides with the impairment of NB organization in untreated NB4 cells. These lines of observations may be explained by the idea that PML is a positive and negative regulator of SRF activity during cellular serum response. Therefore, the altered cooperation of SRF with NBs may cause the inadequate regulation of SRF-mediated transcription. In conclusion, our data suggest that the recruitment of SRF to NBs by PML is important for the immediate-early expression and possible repression of c-fos gene upon serum stimulation.
Rabbit anti-SRF G-20, anti-CBP A22 polyclonal and mouse anti-PML PG-M3 monoclonal antibodies (Santa Cruz Biotechnology), mouse anti-Flag M5 (Sigma) and anti-Myc 9E-10 monoclonal antibodies (Roche Diagnostics) were used.
Cell lines and cultures
HeLa, MCF-7 cells and human fibroblasts (MRC5) immortalized by telomerase reverse transcriptase were grown in Dulbecco's modified Eagle's minimum essential medium and Ham's F12 medium (Invitrogen) supplemented with 5–10% (v/v) heat-inactivated foetal calf serum (FCS). NB4 cells were cultured in RPMI-1640 medium (Sigma) supplemented with 10% FCS. For serum induction, these cells were serum-starved in the medium containing 0.5% FCS for 24 h and then stimulated by 20% FCS for the indicated periods.
Construction of expression plasmids
The full-length cDNA for human SRF were PCR-amplified from HeLa cDNAs with a cloned Pfu DNA polymerase (Stratagene) using specific primers: forward (5′-CGCCCTCTCGAGTCGAGCGCCATGTTACCGACC-3′) and reverse (5′-GCCACCCTCGAGAAGCTTCATCCCTTGGGCCATCTGTCCA-3′), each containing the XhoI site (underlined). To construct plasmids encoding SRF(ΔC338) and SRF(ΔC413), PCR-amplifications were done using the following oligonucleotides: 5′-CGCCCTCTCGAGTCGAGC GCCATGTTACCGACC-3′ and 5′-GCTGGTCTCGAGTCACTGCATAAGGCCCCCAGAGGA-3′ for SRFΔC338, and 5′-CGCCCTCTCGAGTCGAGC GCCATGTTACCGACC-3′ and 5′-ATACATCTCGAGTCAATGCGGGCTAGGGTACATCAT-3′ for SRFΔC413. The PCR fragments were digested with XhoI and cloned into pcDNA3 Myc. To generate plasmids expressing (His)6-Flag-fused SRF(301–395), SRF(301–451) and SRF(301–508), PCR was performed using following sets of primers: 5′-TTTCCCGAATTCAACTACCTGGCACCAGTGTCT-3′ and 5′-GGGCTCAAGCTT CTCGAGCTACATGATGGTGGCGGGCAGCGT-3′ for SRF(301–395), 5′-TTTCCCGAATTCAACTACCTGGCACCAGTGTCT-3′ and 5′-GAACACAAGCTT CTCGAGCTATGGCTCCTGGACCTGGCTGTG-3′ for SRF(301–451) and 5′-TTTCCCGAATTCAACTACCTGGCACCAGTGTCT-3′ and 5′-GCCACCAAGCTT CTCGAGCTAACAATAAATAAGTGGTGCCATCCC-3′ for SRF(301–508). PCR fragments were digested with EcoRI and HindIII and ligated into pRSET. To express (His)6-Flag-fused SRF1–338, the cDNA encoding SRF(1–338) was ligated into pRSET predigested with XhoI. The cDNAs for wild-type-SRF was inserted into the XhoI site of pEGFP-C1 (Clontech) (termed pEGFP-SRF). To generate pcDNA3 Flag-PML, cDNA for human PML was PCR-amplified by specific primers: forward (5′-AAGCTGGAATTCATGGAGCCTGCACCCGCCCGA-3′), containing EcoRI site (underlined) and reverse (5′-CTTTTTCTCGAG AAGCTTCTAAATTAGAAAGGGGTGGGGGTA-3′) containing XhoI and HindIII sites (underlined).
HeLa and MCF-7 cells were transfected with 1 µg of reporter plasmids and 0.2 µg of pRL-TK, which was used for monitoring the transfection efficiency, together with expression plasmids for the indicated proteins by Lipofectamine (Invitrogen). At 7 h after transfection, the transfected cells were serum-starved for 24 h, then stimulated with 20% FCS-containing medium for 3 h. The cells were harvested and lysed in a lysis buffer provided by Promega. The insertless pcDNA, pEF-BOS and pCMX were used as mock vectors. Luciferase activities were measured with a luminometer using the dual-luciferase reporter assay system. Values are the means and standard deviations of results from three independent experiments (Fujita et al. 2000).
Confocal laser scanning microscopic analysis
HeLa and MCF7 cells were subjected to immunofluorescence analysis 24 h after the vector transfection using Fugene (Roche). The MRC5 cells were serum-induced by 20% FCS for the indicated periods. NB4 cells were washed twice with PBS and then resuspended in 300 µL PBS (1–2 × 103 cells/vol). The cell suspension was spread over polylysine-coated dish (Sigma). These cells were fixed with 4% paraformaldehyde in PBS for 10 min and permeabilized with 0.2% Triton-X 100 in PBS for 5 min. After washing cells with PBS, the cells were incubated with the primary antibodies at room temperature for 60 min. The samples were further incubated with FITC (Biosource)- or Cy3 (Amersham)-conjugated secondary antibodies for 60 min and visualized with a confocal laser scanning microscope (Olympus). To avoid bleed-through effects in double staining experiments, each dye was independently excited and the images were electronically merged.
HeLa cells were transfected with the indicated plasmids by Lipofectamine (Invitrogen), and were lysed at 36 h after transfection with a RIPA buffer (150 mm NaCl, 50 mm Tris-HCl [pH 8.0], 5 mm EDTA, 0.1% sodium dodecyl sulphate (SDS), 1% DOC, 1% NP40, 5% glycerol, 50 µg of aprotinin/mL, 100 µg of phenylmethanesulphonyl fluoride/mL, 10 µg of leupeptin/mL and 180 µg of sodium orthovanadate/mL) on ice for 30 min. After centrifuging (13 000 r.p.m.) for 20 min, the supernatants were incubated with the appropriate antibodies for 90 min and then with protein A/G beads (Calbiochem) at 4 °C for 2 h. The beads were washed five times with a buffer containing 150 mm NaCl, 50 mm Tris-HCl [pH 8.0] and 0.5% NP40. Immunoprecipitates were suspended in a Laemmli sample buffer (2% SDS, 100 mm dithiothreitol (DTT), 60 mm Tris-HCl [pH 6.8] and 0.001% bromophenol blue) and separated on SDS-8% polyacrylamide gel electrophoresis. For the immunoblot analyses, gels were transferred to nitrocellulose membranes, and specific proteins were visualized using an enhanced chemiluminescence detection system (Amersham).
For immunoprecipitations of endogenous proteins, cells were treated with 5 mm DTBP (dimethyl 3,3′-dithiobispropionimidate-2HCl) (Pierce) at 4 °C for 30 min. The cross-linking reaction was terminated by washing the cells with a buffer containing 150 mm NaCl and 100 mm Tris-HCl [pH 8.0]. The treated cells were lysed in a lysis buffer containing 150 mm NaCl, 50 mm Tris-HCl [pH 8.0], 5 mm EDTA, 0.5% SDS, 5% DOC, and 5% NP40. The lysates were diluted to 1 : 5 by adding a buffer (150 mm NaCl and 50 mm Tris-HCl [pH 8.0]) and then mildly sonicated. Subsequent procedures were the same as those above.
GST pull-down assay
GST-fused PML and GST on glutathione-conjugated beads were incubated with HeLa nuclear extract (Promega) or bacterially expressed SRF proteins at 4 °C for 60 min in a binding buffer containing 10 mm HEPES [pH 7.5], 50 mm KCl, 50 µm ZnCl2, 2.5 mm DTT and 0.025% NP40. The beads bound by protein complexes were washed five times with the binding buffer and then boiled in a Laemmli sample buffer.
Poly(A)+RNA was isolated using Micro-FastTrack (Invitrogen) and then reverse-transcribed with oligo-dT primer using Superscript II (Invitrogen). The cDNA product was amplified with specific primers:
for c-fos, forward (5′-GGTGGAACAGTTATCTCCAG-3′) and reverse (5′-ATGCTCTTGACAGGTTCCAC-3′); for glyceraldehyde-3-phosphate dehydrogenase (GAPDH), forward (5′-AAGGCTGGGGCTCATTT-3′) and reverse (5′-CCGTATTCATTGTCATACCA-3′). PCR condition was 95 °C for 5 min, and then 30 cycles of 95 °C for 1 min, 51 °C for 30 s and 72 °C for 30 s.
We thank R.M. Evans for pCMX-PML, S. Ishii for GST-PML vector, F. Ishikawa for immortalized human fibroblasts, Y. Honma for NB4 cells, K. Kaibuchi for pEF-BOS-HA-cdc42.V12, pEF-BOS-HA-Rac1.V12 and pEF-BOS-HA-RhoA.V14, I. Okamoto for pCMV-HA-Ras.V12, Y. Kawano and D. Murakami for mCArG SRE-luciferase vector, pFA2-CBP and SRFΔ266 vector, and T. Ichimura for making graphic files. This work was supported by a Grant-in-Aid for Scientific Research on Priority Areas, and by Special Coordination Funds for Promoting Science and Technology from the Ministry of Education, Culture, Sports, Science and Technology of the Japanese Government (to M.N.).