In most human cancers the Myc proto-oncogene is highly activated. Dysregulation of Myc oncoprotein contributes to tumorigenesis in numerous tissues and organs. Thus, targeting Myc stability could be a crucial step for cancer therapy. Here we report Smad7 as a key molecule regulating Myc stability and activity by recruiting the F-box protein, Skp2. Ectopic expression of Smad7 downregulated the protein level of Myc without affecting the transcription level, and significantly repressed its transcriptional activity, leading to inhibition of cell proliferation and tumorigenic activity. Furthermore, Smad7 enhanced ubiquitylation of Myc through direct interaction with Myc and recruitment of Skp2. Ablation of Smad7 resulted in less sensitivity to the growth inhibitory effect of TGF-β by inducing stable Myc expression. In conclusion, these findings that Smad7 functions in Myc oncoprotein degradation and enhances the cytostatic effect of TGF-β signaling provide a possible new therapeutic approach for cancer treatment.
The MYC gene encodes a nuclear transcription factor that is involved in a diverse group of cellular processes, including cell proliferation, cell-cycle regulation, apoptosis and development (Grandori et al., 2000). The Myc protein contains two regions important for its function: the N-terminal transactivation domain (TAD) and the C-terminal basic helix-loop-helix leucine zipper (B/HLH/LZ). The B/HLH/LZ domain interacts with the MAX protein and binds to specific E-box elements, whereas the TAD, which contains Myc boxes (MB) I and II, is responsible for regulating the transcription of target genes involved in cell growth, cell cycle regulation and apoptotic cell death (Grandori et al., 2000). Furthermore, Myc is known to be implicated in oncogenesis and its deregulation has been identified in several human cancers of different origins, including colon cancer, glioblastoma, melanoma and diffuse large B-cell lymphoma (Albihn et al., 2010). The expression level of Myc is increased as a result of amplification and mutation of the MYC gene, which affects the stability of Myc. Consequently, the stable and prolonged presence of the Myc protein is a contributor to the induction phase of carcinogenesis (Bahram et al., 2000; Grandori et al., 2000). Therefore, it is important to understand the factors involved in the in vivo molecular stability of the Myc protein which might inform the development of novel targeted molecules for cancer therapy.
Previous reports have implicated ubiquitin-mediated modulation as an important factor in Myc stability and function. In fact, recent studies have identified at least four ubiquitin ligases involved in the regulation of Myc protein turnover (Adhikary et al., 2005; Kim et al., 2003; Popov et al., 2010; Popov et al., 2007; von der Lehr et al., 2003). Among those ubiquitin ligases, SCF (Skp1–Cul1–F-box protein) complexes including F-box proteins, such as S-phase kinase associated with protein 2 (Skp2) and F-box and WD repeat domain containing 7 (Fbw7), have been well characterized. In SCF complexes, F-box proteins act as specific substrate targeting factors and Cul1 ubiquitinase induces ubiquitylation of the substrates (Nakayama and Nakayama, 2005; Zheng et al., 2002). Among F-box proteins, Fbw7 and Skp2 recognize Myc protein and regulate ubiquitylation and degradation of Myc differently by targeting the MBI and MBII domains of Myc, respectively. In particular, Skp2 binds to the MBII and HLH/LZ domains of Myc through leucine-rich repeats. Interaction of Skp2 with the MBII domain of Myc mediates ubiquitylation and proteasomal degradation of Myc. However, Skp2 also increases transcriptional activity of Myc by acting as a co-factor (Kim et al., 2003; von der Lehr et al., 2003). Regulation of Myc stability by Fbw7 is more complicated and requires additional signaling pathways. Fbw7 destabilizes Myc in a phosphorylation-dependent manner by recognizing phosphorylated Myc at threonine 58 (T58) in the MBI domain, with the help of glycogen synthase kinase 3 (Gsk3) (Welcker et al., 2004b; Yada et al., 2004). This interaction facilitates the degradation of Myc and prevents its biological functions (Welcker et al., 2004a).
Previous studies have demonstrated that Myc protein downregulation is one of the key events in the cellular growth inhibitory response to TGF-β signaling. TGF-β-mediated Myc downregulation reduces expression of cell growth-related Myc target genes, including ID2, thereby strongly inhibiting cell growth (Kowanetz et al., 2004; Murphy et al., 2004; Nilsson et al., 2004). TGF-β signaling regulates transcription of the Myc gene through the TGF-β inhibitory element, which enables the binding of Smad3/4, E2F4 and p107, to the promoter of the Myc gene (Chen et al., 2002; Frederick et al., 2004; Yagi et al., 2002). Although it is well characterized that TGF-β signaling regulates Myc at the transcription level, post-transcriptional regulation of Myc protein by TGF-β has not been demonstrated.
Smad7 is well known as an intracellular antagonist of TGF-β signaling through a negative feedback loop (Nakao et al., 1997). Upon activation of TGF-β signaling, Smad7 induces the degradation of the activated TβRI through association with various E3 ubiquitin ligases, including Smurf 1 and 2 (Chen et al., 2000; Ebisawa et al., 2001; Hayashi et al., 1997). In the nucleus, Smad7 also antagonizes TGF-β signaling by interfering with the Smad4–DNA complex formation on the target gene promoter (Zhang et al., 2007). Thus, Smad7 blocks many TGF-β-induced cellular or biological processes, including embryonic development, inflammation and fibrosis (Dooley et al., 2003; Monteleone et al., 2001; Zhao et al., 2000). However, in contrast to its inhibitory function, Smad7 occasionally enhances TGF-β-mediated cellular processes. For example, TGF-β-induced apoptosis is known to be mediated by Smad7 in prostate cancer cells, visceral epithelial cells and mesangial cells (Lallemand et al., 2001; Landström et al., 2000; Okado et al., 2002; Schiffer et al., 2001). Also, Smad7 reinforces the anti-inflammatory effect of TGF-β through inhibition of tumor necrosis factor α (TNFα) inflammatory signaling (Hong et al., 2007).
In this manuscript, we suggest that Smad7 functions as a key mediator in regulating Myc stability. Ectopic expression of Smad7 downregulated Myc protein levels, leading to decreased Id2 expression. Regulation of Myc degradation was mediated through the polyproline-tyrosine (PY) motif of Smad7 by recruitment of Skp2. HaCaT cells lacking Smad7 were less sensitive to TGF-β-induced growth inhibition signaling. These studies provide strong evidence that Smad7 plays a role as an integrated mediator of Myc proteolysis, contributing to amplification of the cytostatic effect of TGF-β.
Smad7 downregulated Myc expression by modulating Myc stability
To investigate the subcellular functions of Smad7, microarray analysis was performed using SNU638, a human gastric cancer cell line, and MCF7, a human breast cancer cell line, stably expressing Smad7. We had identified 20 genes from the expression profile of 10,000 genes by comparing control cells to Smad7-overexpressing cell lines. Among selected genes, members of the Id family, which are direct targets of Myc, were all significantly suppressed by Smad7. Semi-quantitative and quantitative RT-PCR analysis demonstrated that Smad7 reduced the level of ID1, ID2 and ID3 transcripts by 5- to 15-fold (Fig. 1A). Transcription of other Myc-target genes, such as BRCA1, CCND1 and ODC, were also decreased by the presence of Smad7 (supplementary material Fig. S1). Interestingly, the expression of Myc protein was markedly downregulated in both SNU638 and MCF7 cells stably expressing Smad7 without affecting transcription of the MYC gene (Fig. 1B,C, supplementary material Fig. S2A). Smad7 did not influence the expression of Max and Mad4 that associate with Myc complexes (supplementary material Fig. S2A). These results led us to evaluate whether loss of Smad7 increases the expression of Myc and its target molecule, Id2. As anticipated, decreasing the level of Smad7 by treatment with Smad7 siRNA led to an increase of Id2 and Myc protein expression (supplementary material Fig. S2B). The effect of Smad7 on Myc regulation was further extended to in vivo experiments using a Smad7 transgene-inducible mouse system (Han et al., 2006). The Smad7 transgene was induced in K5.Smad7 transgenic mice by treatment with RU486 (10−7 M) for 12 hours. Induction of the Smad7 transgene in primary epidermal keratinocytes inhibited the expression of Myc (Fig. 1D).
To further validate the effect of Smad7 on Myc regulation, we used HaCaT-tetMyc cell lines, which were able to induce the human MYC transgene, without regulation of TGF-β, under the control of the tet regulator (Seoane et al., 2002). Ectopic expression of Smad7 by adenoviral infection resulted in downregulation of Myc in a dose-dependent manner (Fig. 1E). We then examined the effect of Smad7 on the stability of Myc protein using pulse–chase experiments. Indeed, the ectopic expression of Smad7 accelerated Myc turnover (Fig. 1F). Knowing that MYC and ID family genes were involved in regulating cell proliferation (Albihn et al., 2010; Lasorella et al., 2005), we next examined whether Smad7-mediated downregulation of Myc and Id2 could affect cell proliferation and tumorigenesis. Cells stably expressing Smad7, which inhibited MYC and ID family gene expression, exhibited a slower growth rate compared with control cells, and introduction of Myc into cells expressing Smad7 rescued cell proliferation (Fig. 1G; supplementary material Fig. S3A,B). Furthermore, ectopic expression of Smad7 significantly inhibited the tumorigenic activity of SNU638 cells (P = 0.022; Fig. 1H). In addition, transient transfection of siRNAs against Myc or Id2 led to significant inhibition of [3H]thymidine incorporation (P<0.0001 and P = 0.0006; supplementary material Fig. S3C,D). These results confirmed that Smad7 inhibited cell proliferation through regulation of Myc and Id2 expression. Taken together, the results of these experiments suggest that Smad7 regulates Myc protein stability and its physiological functions.
Smad7 decreased Myc transcriptional activity
Because Smad7 influenced Myc stability, we then examined the effect of Smad7 on the transcriptional activity of Myc using a Myc-binding site (MBS)–luciferase vector, which contains four Myc-binding repeat sequences. Smad7 significantly inhibited Myc transcriptional activity in a dose-dependent manner (Fig. 2A). In support of this observation, loss of Smad7 markedly increased the transcriptional activity of Myc (Fig. 2B). Because Id2 is known to be transcriptionally regulated by Myc (Lasorella et al., 2000), the effect of Smad7 on Myc-mediated Id2 promoter activation was evaluated using an Id2 promoter-luciferase construct, pGId2EcoRI-Luc, which contains three Myc-binding repeat sequences (Fig. 2C). The presence of the Myc protein increased the Id2 promoter reporter activity by twofold, whereas Smad7 significantly inhibited Myc-induced Id2 promoter activation (P = 0.0002; Fig. 2D). Furthermore, loss of Smad7 enhanced Myc-mediated Id2 promoter activation by around threefold, whereas Myc alone or co-transfection of Myc with scrambled siRNA (siSCR) activated the Id2 promoter by twofold (Fig. 2E). A chromatin immunoprecipitation (ChIP) assay showed that Myc directly bound to the Id2 promoter, whereas Smad7 did not, indicating that Smad7 inhibited Id2 promoter activity through post-translational regulation of the Myc protein rather than direct interaction with the Id2 promoter (supplementary material Fig. S4). Thus, Smad7-mediated Myc instability influenced Myc transcriptional activity, consequently resulting in the reduced expression of the ID2 gene.
Smad7 promoted degradation of Myc through the ubiquitin-proteasomal pathway
The degradation and recycling of a number of cellular proteins occur through the ubiquitin-proteasomal pathway (Glickman and Ciechanover, 2002). Therefore, to determine whether Smad7 induced the degradation of Myc protein through the ubiquitin-proteasomal pathway, we first checked the level of Myc protein in cells co-transfected with Smad7 and Myc in the presence or absence of the proteasomal inhibitor, MG132. Smad7 markedly decreased the steady-state level of Myc, whereas treatment with MG132 prevented Smad7-mediated degradation of Myc protein (Fig. 3A). Furthermore, MG132 treatment also prevented degradation of endogenous Myc protein by Smad7 in HaCaT-tetMyc cells (Fig. 3B), suggesting that Smad7 regulates Myc stability through the proteasomal pathway.
Previous studies have demonstrated that Smad7 is involved in proteasomal degradation pathway by recruiting ubiquitin E3 ligase through its PY motif (Kavsak et al., 2000; Kim et al., 2004). We, therefore, examined whether the Smad7 PY motif was engaged in Myc degradation. Two Smad7 mutants that had a deletion of the PY motif (ΔPY) or a point mutation from tyrosine to alanine at amino acid 211 of the PY motif (Y211A) were used (Fig. 3C). Ectopic expression of Smad7 markedly inhibited expression of Myc, whereas Smad7 mutants failed to decrease the level of Myc expression (Fig. 3D).
To determine whether Smad7 is involved in Myc ubiquitylation, Myc and ubiquitin were co-transfected in the presence of wild-type Smad7 or its two mutant forms. Strong polyubiquitylation of Myc was observed in the presence of Smad7, whereas the two Smad7 PY motif mutants impeded ubiquitylation of Myc (Fig. 3E). We then examined the effect of Smad7 PY mutants on Myc-mediated transcriptional activity. Smad7 mutants (Y211A and ΔPY) failed to suppress the Myc-mediated transcriptional activation of Myc target genes, including activation of the Id2 promoter (supplementary material Fig. S5A,B). These results further support the role of the Smad7 PY motif in the ubiquitylation and degradation of Myc.
Mapping of interacting domains in Myc and Smad7
To elucidate whether Smad7-mediated regulation of Myc stability occurs by physical interaction of Myc with Smad7, an immunoprecipitation assay and an in vitro GST-pull down assay were performed. The immunoprecipitation assay showed the physical interaction between endogenous Myc and Smad7 protein in the HaCaT cell line (Fig. 4A). The in vitro GST-pull down assay using 35S-labeled Smad7 protein also confirmed that Myc directly interacted with Smad7 (Fig. 4B). Furthermore, immunofluorescence data demonstrated that Smad7 colocalized with Myc in the cell nucleus (Fig. 4C). Taken together, these results suggest that Smad7 regulates the stability of Myc by directly binding to Myc in the nucleus.
To determine the specific domain of Smad7 that binds to Myc, full-length Myc was transfected along with three different Smad7 constructs: the N-terminal region (aa 1–203), the C-terminal region (aa 204–427) including the PY motif and MH2 domain, and full-length Smad7. Myc bound primarily to the C-terminal region of Smad7 (Fig. 4D). In order to map the Smad7-binding domain in Myc, various Myc deletion constructs (Fig. 4E) were used in transfection experiments with a full-length Smad7. The full-length Smad7 interacted with the following Myc fragments: amino acids (aa) 1–454 (WT), aa 162–454 (Δ3) and aa 1–236 (Δ1), but not with aa 237–454 (Δ2; Fig. 4E,F). These data indicate that the C-terminal region of Smad7 and the middle domain (aa 162–236) of Myc are crucial for their physical interaction.
Smad7 recruited ubiquitin ligase, Skp2, into the complex with Myc
The fact that Smad7 was involved in the ubiquitin-proteasomal degradation pathway prompted us to identify ubiquitin ligases recruited by Smad7 into the Smad7–Myc complex. Previous reported studies have indicated that Fbw7, a component of an SCF-class ubiquitin ligase (E3) complex, and Skp2, a substrate-binding subunit of the SCFSkp2 E3 ligase complex, mainly associates with Myc (Kim et al., 2003; von der Lehr et al., 2003; Welcker et al., 2004b). Therefore, we examined whether Skp2 and Fbw7 were involved in Smad7-mediated Myc degradation. Co-transfection of Smad7 with Skp2 or Fbw7, along with Myc, showed that the expression of Myc was downregulated in a manner that was Fbw7 or Skp2 dose dependent. Neither Skp2 nor Fbw7 affected the expression of Smad7. Treatment of the cells with MG132 restored the level of Myc even in the presence of Skp2 or Fbw7 with Smad7 (Fig. 5A; supplementary material Fig. S6A), indicating that Smad7 might associate with Skp2 or Fbw7 to facilitate Myc degradation. We further investigated Smad7-induced turnover of Myc protein in the presence of either Skp2 or Fbw7 through pulse–chase experiments. Smad7 rapidly reduced the half-life of Myc in the presence of Skp2 or Fbw7 (Fig. 5B; supplementary material Fig. S6B). Next, we characterized Smad7-mediated Myc turnover by transfecting Skp2-specific siRNA duplexes. Loss of Skp2 reduced Myc instability even in the presence of Smad7 (Fig. 5C; supplementary material Fig. S7). Therefore, we next examined whether Smad7-mediated Myc degradation could be promoted by recruitment of Skp2 or Fbw7. Smad7 was found to enhance formation of the Skp2–Myc complex, but it did not affect the level of the Fbw7–Myc complex (Fig. 5D; supplementary material Fig. S6C). In addition, PY motif deletion in Smad7 failed to increase the level of the Skp2–Myc complex and also inhibited interaction of Smad7 with Myc (Fig. 5D). These results indicate that recruitment of Skp2 through the Smad7 PY motif results in Smad7-mediated Myc degradation.
In order to map the binding sites for Smad7 and Skp2, we performed an immunoprecipitation assay using Smad7 and Skp2 deletion mutants. Interaction of Smad7 with Skp2 was characterized by the observation that the C-terminal region of Smad7, including the PY motif and MH2 domain, interacted with Skp2 (supplementary material Fig. S8A). We also confirmed the interaction between endogenous Smad7 and Skp2 and SCFSkp2 complexes (supplementary material Fig. S8B). Deletion of the PY motif in Smad7 diminished its interaction with Skp2 (Fig. 5E). Smad7 bound to the leucine-rich repeat (3LRR-7LRR) domain of Skp2 (Fig. 5F; supplementary material Fig. S8C,D). Taken together, these data conclude that Skp2 is recruited to Myc through the PY motif of Smad7.
To examine whether Smad7 phosphorylation affects its interaction with either Skp2 or Fbw7, we performed an immunoprecipitation assay using a Smad7 phosphorylation site mutant (Smad7 S248A) (Pulaski et al., 2001). Mutation of the Smad7 phosphorylation site decreased Smad7–Skp2 interaction, but not Smad7–Fbw7 interaction (supplementary material Fig. S9). This indicates that Smad7 phosphorylation is required for its interaction with Skp2, but not with Fbw7.
Lack of Smad7 released cells from cytostatic effect of TGF-β
The experimental evidence that Myc stability is regulated by the Smad7–Skp2 complex and that TGF-β signaling induces Smad7 expression (Nakao et al., 1997) prompted us to examine the role of Smad7-induced Myc regulation in the cytostatic effect mediated by TGF-β signaling. FACS analysis showed that TGF-β1 treatment of HaCaT cells in vitro resulted in the accumulation of these cells in the G1 phase. However, loss of Smad7 increased the S-phase population by 10–15% even with TGF-β1 stimulation (Fig. 6A,B), suggesting that the lack of Smad7 caused the cells to be less sensitive to TGF-β1-mediated cytostatic effect. Loss of Smad7 inhibited TGF-β1-mediated suppression of Id2 and Myc expression. Doing the reverse to provide an increase of Smad7 in the cells resulted in a decrease of Id2 and Myc expression (Fig. 6C). These data indicate that Smad7 acts as a crucial mediator of the cytostatic effect of TGF-β signaling and that this occurs through post-translational regulation of Myc protein levels.
To examine whether Smad7 affects TGF-β-mediated transcriptional repression of Myc, cells expressing either control or Smad7 were treated with TGF-β, and then Myc transcription was compared. Smad7 inhibited TGF-β-mediated transcriptional regulation of the MYC as well as the PAI-I gene (supplementary material Fig. S10). Taken together, this result indicates that Smad7 does not act as an enhancer of Smad3/4–E2F4 complex function, but acts as an inducer of Myc protein degradation.
The oncoprotein Myc is known to be tightly regulated by ubiquitylation-mediated proteolysis. When this process fails, cells can grow abnormally forming a tumor (Bahram et al., 2000; Bhatia et al., 1993). Therefore, understanding the proteolysis mechanism(s) responsible for Myc degradation is important for developing novel approaches to cancer prevention and therapy. Previous studies have demonstrated that TGF-β suppresses the expression of the MYC gene at the transcriptional level through the TGF-β inhibitory element in the promoter of the MYC gene. This transcriptional repression of the MYC gene is dependent on the direct binding of the Smad3 protein to the repressive Smad binding element (RSBE) (Frederick et al., 2004). Although the transcriptional mechanisms of Myc regulation are reasonably well understood, post-translational regulation of Myc by TGF-β signaling has not been previously reported.
Our study demonstrates a molecular mechanism involving a crucial role for Smad7 in ubiquitylation-mediated proteolysis of Myc. Smad7 is shown to be involved as an integrated mediator to regulate the proteolysis of the Myc protein by recruiting the F-box protein, Skp2. In our experiments Smad7 repressed Myc transcriptional activity by imparing Myc stability. This in turn led to the inhibition of Id2 mRNA expression, and consequently inhibition of the proliferative effect of Myc. Therefore, Smad7impaired has a transductory property in TGF-β-mediated Myc downregulation, as in TGF-β-induced apoptosis (Lallemand et al., 2001; Landström et al., 2000; Schiffer et al., 2001). It is clear from our present studies that Smad7-induced proteolysis of Myc is one of the mechanisms that are involved in the observed TGF-β-mediated cytostatic effect (Fig. 6D).
It has been reported that Skp2 participates in the induction of Myc degradation through a ubiquitylation mechanism. Skp2 is also known to be involved in the enhancement of Myc transcriptional activity, working as a co-factor of Myc (Kim et al., 2003). This is consistent with evidence that Myc and Skp2 are upregulated in various cancers and co-operate in the degradation of p27Kip1 (Albihn et al., 2010; Gstaiger et al., 2001; Sicari et al., 2012). However, stable expression of Skp2 can limit activity of Myc by inducing rapid turnover of the Myc protein. The role of Skp2 in the regulation of Myc can be changed by adaptor proteins that interact with Skp2. Kalra and Kumar showed that the X protein of hepatits B virus (HBx) enhanced Myc stability and tumor formation by inhibiting the interaction between Skp2 and Myc, indicating that Skp2 limits Myc activity in carcinogenesis (Kalra and Kumar, 2006). Kimura et al. demonstrated that MM-1, a Myc-binding protein, reduced stability and activity of Myc through recruitment of Skp2, suggesting that Skp2 could be involved in only the degradation of Myc by MM-1 (Kimura et al., 2007). Our studies support the molecular mechanism whereby Skp2 is able to suppress or limit Myc activity when Smad7 is present and co-operates with Skp2. Indeed, our finding that Smad7 suppressed tumorigenic activity and cell proliferation could be shown to be dependent on Myc degradation through recruitment of Skp2 to Myc. Thus, experimental evidence suggests that Smad7 modulates the role of Skp2 in Myc regulation through the proteasomal degradation pathway.
Several studies demonstrated that Smad7 interacted with E3 ubiquitin ligases, such as Smurf1/2, Jab1 and Arkadia (Kavsak et al., 2000; Kim et al., 2004; Koinuma et al., 2003). Interaction of Smurf1/2 with TβRI and Smad7 resulted in degradation of TβRI, leading to the suppression of TGF-β signaling. Our studies demonstrated that Smad7 induces Myc degradation by recruiting Skp2 through the PY motif of Smad7. It has been reported that Smurf2 and Jab1 also bind to the PY motif of Smad7 (Kavsak et al., 2000; Kim et al., 2004). Therefore, the possibility that these Smad7-binding E3 ligases, Smurf1/2 and Jab1, directly or indirectly participate in the regulation of Myc stability remains to be studied.
The PY motif is known to be recognized by proteins containing WW domains (Aragón et al., 2012). Smurf2 also binds to the PY motif in Smad7 through its WW domains and regulates Smad7 stability (Kavsak et al., 2000). The leucine-rich repeat (LRR) motif is known to bind to various proteins (Kobe and Kajava, 2001). Although Skp2 does not have a WW domain, our study showed that the LRR motif of Skp2 interacts with the C-terminal domain of Smad7 including the PY motif (Fig. 5F; supplementary material Fig. S8) and mutation of the PY motif decreased the Smad7–Skp2 interaction (Fig. 5E).
Previous studies indicate that the phosphorylation at serine 62 stabilizes Myc, whereas phosphorylation at threonine 58 is responsible for Myc degradation (Sears et al., 1999; Sears et al., 2000; Yeh et al., 2004). These two residues, T58 and S62, in the MBI domain of Myc are frequently mutated in various tumors (Bahram et al., 2000; Bhatia et al., 1993), and all MYC genes recovered in the transformed retroviruses harbor a mutation at T58 (Papas and Lautenberger, 1985). Fbw7-mediated Myc turnover occurs when Myc was phosphorylated on T58 (Welcker et al., 2004b; Yada et al., 2004). In contrast, Skp2 binding to Myc did not require T58 phosphorylation for Myc turnover because it recognized the MBII and B/HLH/LZ regions of Myc (Kim et al., 2003; von der Lehr et al., 2003). Unlike Skp2 and Fbw7, Smad7 interacted with the middle region (aa 162–236) of Myc and recruited Skp2 to Myc, eventually contributing to Myc turnover (Fig. 4E; Fig. 5D). Therefore, Smad7 functions as an intermediary to induce Myc degradation, regardless of the phosphorylation status of Myc. This study provides information that Smad7 could be a candidate target in the treatment of cancers that have a mutation in the T58 site of the Myc protein.
Although we observed the association of Fbw7 with Smad7, Smad7 did not enhance the ability of Fbw7 to bind to Myc (supplementary material Fig. S6). Because Fbw7 facilitates Smad7-mediated Myc degradation, the regulation of Myc stability by the Smad7–Fbw7 complex might be different from that by the Smad7–Skp2 complex. One possible mechanism is that Smad7 could influence Fbw7-mediated Myc degradation by regulating an additional signaling pathway such as Myc phosphorylation (Welcker et al., 2004b; Yada et al., 2004).
In conclusion, our study defines a previously unknown intracellular role of Smad7 in regulating the stability and activity of the Myc protein. Our findings provide an important understanding of the relationship between the cytostatic effect of TGF-β and regulation of the proto-oncogene Myc.
MATERIALS AND METHODS
Cell culture and reagents
The human gastric cancer cell line SNU638 was maintained in RPMI 1640 (Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS), 100 units/ml of penicillin, and 100 µg/ml of streptomycin (Invitrogen, Carlsbad, CA, USA) at 37°C in an atmosphere of 5% CO2, 95% air (Han et al., 2004). MCF7, HaCaT and HEK 293T cells were maintained in Dulbeeco's modified Eagle's medium (DMEM; Gibco) supplemented with 10% FBS, 100 units/ml of penicillin and 100 µg/ml of streptomycin. HaCaT-tetoffMyc cells were maintained as previously described (Warner et al., 1999). To establish HaCaT cell lines expressing Myc in a doxycycline-inducible manner (HaCaT-tetonMyc), the human MYC cDNA was cloned into pRetroX retroviral vector (Clontech, Mountain View, CA, USA). Viruses containing the MYC gene or tetracycline response activator gene were produced according to the manufacturer's protocol (Cell Biolabs, Inc., San Diego, CA, USA). HaCaT-tetonMyc cells were generated by infection with these viruses and selected using the antibiotics, G418 (2 mg/ml) and puromycin (2 µg/ml). SNU638 cell lines stably expressing FLAG-tagged Smad7 (SNU638-Smad7) were generated by reroviral infection using pLPCX retroviral vector, as described previously (Kim et al., 2004) and maintained with 400 ng/ml puromycin. Cells containing empty pLPCX vector (SNU638-LPCX) were used as a control. Anti-Myc (9E10), Smad7 (N19), Id2 (C20), CDC4 (Fbw7, H300), Skp2 (H435), ubiquitin and GST antibodies were purchased from Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA. Flag (M2) and β-actin (AC-15) antibodies were obtained from Sigma, St. Louis, MO, USA.
Total RNA was purified using TRIzol and RNAeasy (Qiagen, Valencia, CA, USA) from five clones of SNU638-LPCX and SNU638 LPCX-Smad7 cells. Purified RNA (10 µg) was reverse-transcribed to label with Cy3 and Cy5 (GE Healthcare, Piscataway, NJ, USA) using the FairPlay Microarray Labeling kit (Stratagene, La Jolla, CA, USA). Microarray analysis was carried out using individually isolated RNA from at least triplicate samples in each group and human array chips (45K) obtained from the NCI Microarray Facility. Experiments and analysis were performed according to the manufacturer's instructions and the Center for Cancer Research, NCI.
GST-Myc expression plasmids were constructed by inserting full length and fragments of human Myc into BamHI–NotI sites of the pEBG vector. Smad7 PY domain mutants (ΔPY and Y211A) were kindly provided by J. L. Wrana (University of Toronto). FLAG- or HA-Skp2 wild-type and deletion mutants were obtained from X. Lin (Liang et al., 2004). FLAG-Smad7 S248A was made by a site-directed mutagenesis.
Northern blot analysis and RT-PCR
Total RNA was prepared from SNU638 and MCF7 cells stably expressed Smad using a TRIzol kit (Life Technologies, Carlsbad, CA, USA). 20 µg of RNA were subjected to northern blot analysis as previously described (Han et al., 2008). Briefly, blots were hybridized to 32P-labeled DNA probes corresponding to Smad7, Myc and Id2 mRNA and then incubated in northern hybridization buffer (Amersham Pharmacia Biotech Inc., Piscataway, NJ, USA) at 65°C overnight. The blots were washed in high (2× SSC, 1% SDS) and low (0.2× SSC, 0.1% SDS) stringency washing buffer at 65°C and then visualized by autoradiography. RT-PCR was performed using AmpliTaq Gold DNA polymerase (Applied Biosystems, Carlsbad, CA, USA) with iCycler (Bio-Rad, Hercules, CA, USA). The PCR condition used was 50°C for 2 minutes and 95°C for 10 minutes followed by 95°C for 15 seconds and 60°C for 40 seconds for 40 cycles. Primers used for qRT-PCR and RT-PCR analysis are given in supplementary material Table S1.
Immunoblotting and immunoprecipitation
The immunoprecipitated protein or total cell lysate were subjected to subsequent immunoblotting analysis as previously described (Kang et al., 2012). Briefly, cells were lysed in a buffer [25 mM HEPES (pH 7.5), 150 mM NaCl, 1% Triton X-100, 10% glycerol, 5 mM EDTA and aprotease inhibitor cocktail (Roche)]. After immunoprecipitation using the appropriate antibodies, the total cell lysate and the immunoprecipitant were separated by SDS-PAGE, followed by transfer to polyvinylidene difluoride (PVDF) membranes. After incubating with the appropriate primary and secondary antibodies, proteins were visualized by chemiluminescence, according to the manufacturer's instructions (Pierce, Rockford, IL, USA).
In vitro labeling of cells with [S35]methionine was performed as described previously (Kim et al., 2003). Cells were infected with adenovirus containing the lacZ or SMAD7 gene for 18 hours. The cells were then starved of methionine and cysteine by replacing the culture medium with DMEM lacking L-methionine and L-cysteine for 30 minutes. Cells were labeled with [S35]Methionine/Cysteine EXPRESS Protein Labeling Mix (Perkin Elmer, Boston, MA, USA) for 1 hour. After labeling, cells were immediately washed once with DMEM containing L-methionine, L-cysteine and then incubated in the presence of cyclohexamide for the indicated chase times. Cells were harvested, and labeled Myc proteins were immunoprecipitated with anti-Myc antibody at 4°C for overnight, followed by immunoprecipitation with GammaBind agarose beads (Invitrogen) and washing with lysis buffer containing protease and phosphatase inhibitors. Labeled Myc was visualized by autoradiography.
2×103 cells per well were seeded in 96-well plates with the appropriate culture medium. Cell proliferation was checked for 3 days after seeding using CellTiter-Glo Luminescent Cell Viability Assay Reagent (Promega, Madison, WI, USA) or [3H]thymidine incorporation. For CellTiter-Glo Luminescent Cell Viability Assay, the absorbance at 470 nm was measured using a SpectraMax L luminescence microplate reader (Molecular Devices).
Colony forming assay
1×103 cells per well were seeded in six-well plates. Colony formation was measured at 14 days after seeding. Cells were fixed in 50% ethanol and stained with Methylen Blue. Colony number was quantified using ImageJ software.
Design and transfection of siRNAs
Specific siRNAs for the ID2 and MYC genes were purchased from Dharmacon Inc., Lafayette, CO, USA. The small interfering RNA design tool (Dharmacon Inc.) was used to synthesize Smad7 and Skp2 siRNAs. Two human Smad7-specific siRNA target sequences were used as follows: 5′-GAAACTGAAGGAGCGGCAG-3′ and 5′-ATTCGGACAACAAGAGTCA-3′ (Smad7 nucleotides 303–321; 905–923, GenBank accession number AF015261). Skp2 siRNA target sequences were as follows: 5′-GCAGACCTTAGACCTCACAGGTAAA-3′ (Skp2 nucleotides 463–487, GenBank accession number AY029177). Pre-synthesized Smad7, Skp2 and non-specific control siRNA duplexes were purchased from Dharmacon Inc. For the rescue experiment, Smad7 siRNA (nucleotide 303–321) was co-transfected with FLAG-tagged Smad7 construct that had a single nucleotide change at position 309 (G to A), but maintained a functional leucine residue. HaCaT cells were seeded at 60% density the day before transfection. Transfections were performed using TransIT-TKO reagent (Mirus, Madison, WC, USA), according to the manufacturer's instructions.
Transient transfection and luciferase assay
Transfection and the reporter assay were performed as described previously (Kim et al., 2004). Cells were transiently transfected with pGL3B control, MBS-Luc or pGId2EcoRI-Luc. A β-galactosidase reporter was co-transfected as an internal control. TransIT-TKO reagents (Mirus) were used for HaCaT cells as transfection agents according to the manufacturer's instructions. Luciferase activities were measured using the Luciferase Assay System (Promega) according to the manufacturer's protocol. The absorbance at 470 nm was measured using a SpectraMax L luminescence microplate reader (Molecular Devices, Sunnyvale, CA, USA). β-galactosidase activities were assessed using o-nitrophenyl-β-D-galactoside (ONPG) and 1 M Na2CO3. All values were normalized with β-galactosidase activities. All assays were performed in triplicate, and data presented as the means ± standard deviation (s.d.) of three independent experiments.
Chromatin immunoprecipitation assay
ChIP assays were carried out according to the manufacturer's instructions (Upstate Biotech., Lake Placid, NY, USA). Briefly, HaCaT cells were cross-linked with 1% formaldehyde, resuspended in ChIP lysis buffer (1% SDS, 10 mM EDTA and 50 mM Tris-HCl, at pH 8.0) and sonicated on ice. After immunoprecipitation, the protein–DNA complex was incubated with Dynabeads Protein G (Invitrogen). After incubation, the immunocomplexes were washed to remove nonspecific binding and then eluted. The eluted immunocomplexes were reverse crosslinked and the recovered DNA was resuspended in distilled water and used for amplification by PCR, using the primers: 5′-TCTGTTCCACTGTGGCACGTAT-3′ (sense) and 5′-CTCGATAATGGGGAAACAG TGT-3′ (antisense).
In vitro binding assay
GST and GST–Myc fusion proteins were induced in DH5α or BL21 (DE3) with 0.1 mM isopropyl-1-thio-D-galactopyranoside for 3 hours and subsequently purified by binding to glutathione-Sepharose (Amersham Pharmacia Biotech Inc.). Smad7 proteins were labeled with [35S]methionine using the T7 quick-coupled transcription/translation system (Promega). The GST fusion proteins were incubated with Smad7 proteins for 1 hour at 4°C in binding buffer [25 mM HEPES (pH 7.4), 150 mM NaCl, 1 mM MgCl2 and 0.1% Triton X-100] containing protease and phosphatase inhibitors. After extensive washing in the same buffer, the Smad7 bound proteins were visualized by autoradiography.
Immunofluorescence and confocal analysis
Immunofluorescence assays were performed as described previously (Kang et al., 2012). Cells seeded on chamber slides were fixed in 4% paraformaldehyde for 10 minutes at room temperature. Cells were blocked with 5% BSA for an hour and incubated with anti-Myc and anti-Smad7 antibodies overnight at 4°C. Alex-Fluor-488- or -594-conjugated secondary antibodies (Invitrogen) were used. Immunofluorescence was examined with a Zeiss LSM META 510 confocal microscope (Carl Zeiss, Sachsen, Thüringen, Germany).
Flow cytometry analysis
siRNA transiently transfected HaCaT cells were grown in the presence or absence of TGF-β1 (5 ng/ml) for 24 hours. Cells were trypsinized and washed twice with PBS buffer (pH 7.4). After fixing in 70% methanol for 30 minutes on ice, cells were washed twice, and resuspended in PBS containing 0.1% Triton X-100, 5 µg/ml propidium iodide (PI) and 50 µg/ml ribonuclease A for DNA staining. Cells were then analyzed by flow cytometry using the CELLQUEST program (Becton Dickinson, Franklin Lakes, NJ, USA). A total of 20,000 cells were counted.
All quantitative data were expressed as the means ± standard deviation. Statistical analysis (Kruskal–Wallis test or Dunn's comparison test) was performed with GraphPad Prism (GraphPad Software Inc., La Jolla, CA, USA). A value of P<0.05 was considered statistically significant.
We thank J. Park for critical reading of the manuscript, J. L. Wrana for Smad7-expressing plasmids, J. Massague for the HaCaT-tetMyc cell line, A. Iavarone for the pGId2 promoter plasmids, X. H. Feng for Myc constructs, K. Miyazono for Smad7 adenovirus, D. Wolf for Skp2 adenovirus, X. J. Wang for K5.Smad7 transgenic mice, S. J. Park for technical help and R. Davis for the Myc reporter plasmid.
T.-A.K., J.M.K. and S.-J.K. conceived and designed the experiments; T.-A.K., J.M.K., J.-S.H., B.L., S.J.K., E.-S.Y., S.H. and H.-J.L. performed the experiments; T.-A.K., J.M.K. and S.-J.K. analyzed the data; M.F., J.E.N. and S.-J.K. contributed reagents, materials and analysis tools; and T.-A.K. and J.M.K. wrote the article.
This research was supported by the National Research foundation of Korea [grant number 2009-0081756 to S.-J.K. and T.-A.K.] and the Intramural Research Program of the National Cancer Institute, NIH, Bethesda, MD. The NCI had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Deposited in PMC for release after 12 months.
The authors declare no competing interests.