Stabilization of C-terminal binding protein 2 by cellular inhibitor of apoptosis protein 1 via BIR domains without E3 ligase activity
Abstract
C-terminal binding protein 2 (CtBP2) is a transcriptional co-repressor that regulates many genes involved in normal cellular events. Because CtBP2 overexpression has been implicated in various human cancers, its protein levels must be precisely regulated. Previously, we reported that CtBP1 and CtBP1-mediated transcriptional repression are regulated by X-linked inhibitor of apoptosis protein (XIAP). In the pre- sent study, we sought to investigate whether CtBP2 is also regulated by XIAP or any other human IAP. We found that cIAP1 interacts with CtBP2 via through BIR domains to regulates the steady-state levels of CtBP2 protein in the nucleus. The levels of CtBP2 were gradually increased upon cIAP1 overexpression and downregulated upon cIAP1 depletion. Interestingly, the RING domain of cIAP1 responsible for E3 ligase activity was not required for this regulation. Finally, the levels of CtBP2 modulated by cIAP1 affected the transcription of CtBP2 target genes and subsequent cell migration. Taken together, our data demonstrate a novel function of cIAP1 which involves protecting CtBP2 from degradation to stabilize its steady-state level. These results suggest that cIAP1 might be a useful target in strategies aiming to downregulate the steady-state level of CtBP2 protein in treating human cancers.
1. Introduction
The C-terminal Binding Proteins (CtBPs) are transcriptional co- repressors involved in a wide range of cellular events including cell death and tumorigenesis [1]. In humans, two CtBP family members, CtBP1 and CtBP2, exhibit highly conserved primary se- quences and shared target genes [2]. Despite these commonalities, their knockout (KO) mouse phenotypes are quite different. ctbp2 KO mice have an embryonic lethal phenotype due to severe defects in early embryonic development, while ctbp1 KO mice are relatively normal but with a shorter lifespan [3]. These characteristics suggest that CtBP1 and CtBP2 possess unique functions and regulatory mechanisms. In recent years, several reports have suggested that CtBP2 functions as an oncogene; CtBP2 proteins are overexpressed in various tissue malignancies, including bone [4], breast [5,6], stomach [7], liver [8], ovary [9], and prostate cancers [10]. Also, CtBP2 represses expression of genes related to cell migration such as E-cadherin [2], as well as pro-apoptotic genes such as bax [2] and bik [11]. Taken together, these data suggest that tumorigenesis might be caused by upregulated CtBP2 and subsequent repression of its target gene expression. Therefore, maintaining proper levels of CtBP2 protein are crucial in preventing tumorigenesis.
Inhibitor of Apoptosis proteins (IAPs) were originally identified as negative regulators of cell death. These proteins are evolution- arily conserved and all members of the IAP family possess Bacu- lovirus Inhibitor of apoptosis Repeat (BIR) domains [12]. BIR domains are responsible for mediating interaction with other proteins. In fact, caspase-dependent cell death is inhibited by direct binding of BIR domains to caspases [13]. The Really Interesting New Gene (RING) domain found in five members of the human IAP family has E3 ligase activity [14]. XIAP and cIAP1/2 possess three BIR domains and one RING domain, respectively. The BIR domains in XIAP reported display high affinity to caspase 3 or 7 and there- fore inhibit apoptosis [15]. Moreover, the BIR domains in cIAP1/2 exhibit poor affinity towards caspases, instead binding to target proteins and presenting them for ubiquitination by the RING domain [16]. Interestingly, recent reports have demonstrated a unique function of the BIR domains of IAPs which is distinct from cell death regulation or supporting RING activity. The BIR domain of XIAP contributes to the regulation of Epidermal Growth Factor Receptor (EGFR) translation by masking the inhibitory phosphor- ylation of protein phosphatase 2 (PP2A) [17]. We also reported that the BIR1 and BIR3 domains of cIAP1 function to mediate the interaction with target proteins and their E3 ligase [18].
Previously, we reported that CtBP1 and CtBP1-mediated tran- scriptional repression are regulated by XIAP [19]. In the present study, we sought to investigate whether CtBP2 is also regulated by XIAP or any of the human IAPs, which led us to discover an inter- esting mechanism of CtBP2 regulation by cIAP1. Here, we demon- strate the regulation of CtBP2 by cIAP1 and its BIR domains, as well as a novel function of cIAP1 in stabilizing the steady-state levels of CtBP2 and its target gene expression.
2. Materials and Methods
2.1. Cell culture, transfection, and RNA interference
HeLa cells and U2OS cells were maintained in DMEM or RPMI 1640 medium (WelGENE) containing 10% fetal bovine serum (FBS) (WelGENE) at 37 ◦C in a humidified 5% CO2 atmosphere. Trans- fection was performed using PEI (Sigma-Aldrich). U2OS cells were transfected with 500 pmol of CtBP2- or cIAP1-targeting siRNA (siCtBP2, sicIAP1, Dhamacon) or scrambled RNA (scRNA, Dhamacon) using Lipofectamine 2000 (Invitrogen). Forty-eight hours (h) after transfection, cells were harvested and lysed with RIPA buffer (150 mM NaCl, 50 mM Tris pH 7.4, 0.1% SDS, 1% Triton X-100, 0.5% sodium-deoxycholate, protease inhibitor cocktail (Roche)) for Western blot (WB) analysis. The siRNA sequences used were: CtBP2-specific, GCGCCUUGGUCAGUAAUAG; cIAP1-specific, UCGCAAUGAUGAUGUCAAA.
2.2. Antibodies and co-immunoprecipitation (co-IP)
Anti-CtBP2 (BD Biosciences), anti-cIAP1 (R&D Systems), anti- Myc (Millipore), anti-HA (Covance), and anti-actin (Bethyl Labora- tories) antibodies were used for WB or co-IP. For co-IPs, whole cell lysates (WCL) were incubated with the indicated antibody at 4 ◦C for 2 h, and immunocomplexes recovered via protein-A sepharose (Sigma-Aldrich) binding were washed three times with IP-Washing buffer (20 mM Tris pH 8.0, 150 mM NaCl, 1 mM EDTA, 0.1% Triton X- 100) and subsequently analyzed by WB with the indicated antibodies.
2.3. In vivo ubiquitination assay
HeLa cells were transfected for in vivo ubiquitination. At 24 h after transfection, cells were treated with 20 mM of MG132 (Enzo Life Science) for 6 h and then lysed by sonication in lysis buffer (50 mM Tris pH 8.0, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 10%
glycerol, protease inhibitor cocktail). Co-IPs were performed with anti-CtBP2 antibody and analyzed by anti-HA antibody.
2.4. Measurement of protein half-life
U2OS cells were transfected with scRNA or sicIAP1. At 48 h after transfection, cells were treated with 200 mg/ml cyclohexamide (CHX, Sigma-Aldrich) and harvested at the indicated time points. Cells were lysed with RIPA buffer and analyzed by WB with the indicated antibodies.
2.5. Bimolecular fluorescence complementation (BiFC) assay
For the BiFC assay, we cloned the CtBP2 gene into the pBiFC-VN 173 vector. Additionally, the cIAP1 and cIAP2 genes were cloned into the pBiFC-VC 155 vector. HeLa cells were co-transfected with the BiFC expression vectors. At 16 h after transfection, the cells were washed with DPBS (WelGENE), fixed in 3.7% paraformaldehyde (Sigma-Aldrich), and washed again three times with DPBS before staining with DAPI and mounted onto glass slides. Confocal mi- croscopy was used to capture the BiFC images (Carl Zeiss, LSM 800).
2.6. Wound healing assay
For this assay, we seeded CtBP2- or cIAP1-depleted U2OS cells onto 60-mm cell culture dishes. Using sterile tips, triplicate scratches were introduced and then the cells were incubated in RPMI 1640 medium containing 2% FBS. Phase-contrast images were obtained at the indicated time points using a digital camera and phase-contrast microscopy (Olympus). Wound areas were measured and the percentage change was calculated using the formula: Area (%) 100 (wound area at 0, 8, 16, 24 h ÷ wound area at 0 h). The data are represented as the mean ± SEM for the 0, 8, 16, and 24 h time points (n ¼ 6, ***P < 0.0001; one-way ANOVA). 2.7. RNA extraction and quantitative reverse transcription-PCR (qRT-PCR) Total RNA was isolated from U2OS cells using easy-spin™ (iNtRON Biotechnology) according to manufacturer’s instructions. Synthesis of cDNA was performed with M-MLV reverse transcrip- tase (Enzynomics). This cDNA was used for qRT-PCR analysis using the SensiMix™ SYBR No-ROX Kit (Bioline) following the manufac- turer’s instructions. Gene expression was quantified using the comparative Ct (cycle threshold) method and the results were normalized to GAPDH. All qRT-PCR experiments were replicated at least three times. The primer sets used were: Bik (sense: 50- TCCTATGGCTCTGCAATTGTCA-30, antisense: 50-GGCAGGAGT- GAATGGCTCTTC-30), E-cadherin (sense: 50-TGCCCA- GAAAATGAAAAAGG-30, antisense: 50-GGATGACACAGCGTGAGAGA-30), Sirt1 (sense: 50-GCAGATTAGTAGGCGGCTTG-30, antisense: 50- TCTGGCATGTCCCACTATCA-30), and GAPDH (sense: 50-AGTCAACG- GATTTGGTCGT-30, antisense: 50-TTGATTTTGGAGGGATCTCG-30). 3. Results and discussion 3.1. CtBP2 protein interacts with cIAP1 and cIAP2 in the nucleus The primary sequence of human CtBP2 is approximately 80% homologous to human CtBP1 [20], which is regulated by XIAP and the ubiquitin proteasome system [19]. Therefore, we hypothesized that XIAP and members of the IAP family may interact with CtBP2 and regulate its steady-state levels in human cells. To test our hy- pothesis, we expressed 6Myc-XIAP, 6Myc-cIAP1, 6Myc-cIAP2, and Myc/His-ML-IAP into HeLa cells and examined their interaction with endogenous CtBP2 by co-immunoprecipitation with anti- CtBP2 antibody. The results show that endogenous CtBP2 inter- acted with XIAP, cIAP1, and cIAP2, but not with ML-IAP (Fig. 1A). Interestingly, cIAP1 and cIAP2 interacted with CtBP2 more strongly than XIAP did. To determine whether and where CtBP2 interacts with the cIAP1 and cIAP2 proteins in cells, we performed bimolecular fluorescence complementation (BiFC) assays [18,21]. Thus, we generated several fusion proteins, namely CtBP2 fused to the N-terminal half of Venus fluorescent protein (CtBP2-VN173) and cIAP1 fused to the C-ter- minal half of Venus fluorescent protein (cIAP1-VC155) (Fig. 1B), and then co-expressed them into HeLa cells (Fig. 1C). We observed bright green fluorescence almost exclusively in the nucleus, indi- cating that the interaction of CtBP2 and cIAP1 occurs in this compartment. Similar results were obtained with the co- expression of CtBP2-VN173 and cIAP2-VC155 (Fig. 1D). Fig. 1. CtBP2 protein interacts with cIAP1 and cIAP2 in the nucleus. (A) HeLa cells were transfected with the indicated IAP family members. After 24 h, whole cell lysates (WCL) were co-immunoprecipitated (IP) with anti-CtBP2 antibody and analyzed by Western blot (WB) with anti-Myc antibody (top). Expression of IAPs and CtBP2 was determined by WB with anti-Myc and -CtBP2 antibodies (bottom). *IgG heavy chain. (B) Schematic diagram of cIAP1-VC155 (or cIAP2-VC155) and CtBP2-VN173 fusion proteins generated for the BiFC assay. U2OS cells were co-transfected with cIAP1-VC155 (C) or cIAP2-VC155 (D) and CtBP2-VN173. Sixteen hours after transfection, cells were fixed and permeabilized before GFP fluorescence was analyzed by confocal microscopy. DAPI staining (blue) was used to visualize the nucleus. Scale bar, 20 mM. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Collectively, these results suggest that CtBP1 and CtBP2 have differential affinity for interacting with IAP family members. Considering CtBP2 is a member of a transcriptional repressor complex, cIAP1 and cIAP2 might be involved in transcriptional repression events via interaction with CtBP2. 3.2. Levels of CtBP2 protein are modulated by levels of cIAP1 protein To examine whether overexpression of IAPs affects the steady- state level of CtBP2 protein, we expressed 6Myc-cIAP1, 6Myc- cIAP2, or 6Myc-XIAP into U2OS cells. The results demonstrate that the level of endogenous CtBP2 protein gradually increased when cIAP1 or cIAP2 was overexpressed in a dose-dependent manner (Fig. 2A). However, when cIAP1 was depleted with sicIAP1, the steady-state level of CtBP2 protein was decreased (Fig. 2B). These results suggest that IAPs function to stabilize the steady-state level of CtBP2 protein. Fig. 2. Stability of CtBP2 is modulated by cIAP1. (A) U2OS cells were transfected with increasing amounts of 6Myc-cIAP1 (left), 6Myc-cIAP2 (middle), and 6Myc-XIAP (right). After 24 h, whole cell lysates (WCL) were analyzed by Western blot (WB) with anti-Myc, -CtBP2, and -actin antibodies. (B) U2OS cells were transfected with scRNA (scrambled RNA) or cIAP1-targeted siRNA (sicIAP1) for 48 h. Anti-cIAP1, -CtBP2, and -actin antibodies were used for WB analysis. (C) U2OS cells were transfected with scRNA or sicIAP1 and treated with 200 mg/ml CHX at 48 h post-transfection. Cells were harvested at the indicated times and subjected to WB analysis. The numbers below the CtBP2 bands indicate the level of CtBP2 protein relative to actin as determined by densitometry using the NIH Image J program. (D) HeLa cells were co-transfected with vectors expressing 6Myc-cIAP1 and HA-Ub WT or HA-Ub R63K and co-immunoprecipitations were performed with anti-CtBP2 antibody and analyzed by WB with anti-HA antibody. *IgG heavy chain. To further examine this mechanism, we analyzed the half-life of CtBP2 in cells treated with CHX, which blocks de novo protein synthesis. Our data show that the half-life of CtBP2 in cIAP1- depleted U2OS cells was substantially reduced compared to parental control cells (Fig. 2C). These results were unexpected because we had hypothesized that CtBP2 is regulated by cIAP1 via a mechanism similar to CtBP1, which is regulated by XIAP and the ubiquitin proteasome system. However, our results demonstrate that degradation of CtBP2 might be accelerated in the absence of cIAP1, suggesting that cIAP1 may play a protective role in CtBP2 regulation. Ubiquitin (Ub) has seven lysine residues, each of which can form different types of isopeptide Ub linkages [22]. These linkages seem to be complex and may provide a unique signal. In general, while K48-linkages activate degradation via proteasome, K63-linkages are recognized as a signal for further cellular events [23]. For example, cIAP1 has been reported to ubiquitinate RIP1 via K63- isopeptide linkages, forming pro-survival signaling complex [24]. Therefore, we tested whether cIAP1 ubiquitinates CtBP2 via K63-isopeptide linkages. HeLa cells were co-transfected with vectors expressing 6Myc-cIAP1 and either wild-type (HA-Ub WT) or mutant Ub (HA-Ub-R63K) in which all lysine residues except for K63 were mutated to arginine. Our data demonstrate that the level of CtBP2 ubiquitination was similar between the WT and R63K Ub mutant (Fig. 2D), suggesting that the level of CtBP2 protein might be stabilized via K63-linked ubiquitination by cIAP1. 3.3. Stabilization of CtBP2 protein occurs via interaction with BIR domains of cIAP1 without requirement for RING activity To map the region of cIAP1 that interacts with CtBP2, we generated a series of cIAP1 truncation mutants (Fig. 3A) and expressed them in HeLa cells. The co-IP results showed that the mutant which retained the whole BIR domains but lacked the CARD and RING domains (6Myc-cIAP1 BIR1-BIR3) interacted with endogenous CtBP2 stronger than cIAP1 WT (6Myc-cIAP1 FL). cIAP1 mutants lacking BIR1 (6Myc-cIAP1 BIR2-RING) or the BIR1~BIR2 region (6Myc-cIAP1 BIR3-RING) exhibited very weak interactions with CtBP2. However, no CtBP2 binding was observed when all BIR domains were deleted from cIAP1 (6Myc-cIAP1 CARD-RING) (Fig. 3B). These results indicate that the BIR domains, especially BIR1, are responsible for mediating interactions with CtBP2. Following discovery of E3 ligase activity in the RING domain [12], most studies investigating IAPs have focused on their role as E3 ligases through the RING domain. However, we previously demonstrated that cIAP1 has a novel function as a mediator, in which the BIR domains of cIAP1 mediate the interaction between the E3 ligase, C-terminus of Hsc70-interacting protein (CHIP), and its target protein, eIF4E, for proteasomal targeting. The RING ac- tivity of cIAP1 is not absolutely necessary for this mediator function [18]. Therefore, we investigated whether presence of the cIAP1 mutants possessing only the BIR domains (6Myc-cIAP1 BIR1-BIR3) or lacking RING activity (6Myc-cIAP1 RM, H588A/C592A) (Fig. 3A) affect the steady-state level of CtBP2 protein. Surprisingly, CtBP2 was significantly increased in a dose-dependent manner when either mutant was expressed in U2OS cells (Fig. 3C). The level of increase observed was similar to that seen in cIAP1 FL-expressing cells (Fig. 2A). These results indicate that cIAP1 is able to stabilize CtBP2 protein through BIR domains without need for E3 ligase activity. Fig. 3. Stabilization of CtBP2 proteins occurs via interaction with BIR domains of cIAP1 without requirement for RING activity. (A) Schematic representation of 6Myc-cIAP1 FL (WT), cIAP1 RING mutant (RM) (H588A/C592A), and various cIAP truncation constructs used in this study. (B) HeLa cells were co-transfected with the various 6Myc-cIAP1 con- structs shown in (A). After 24 h, co-immunoprecipitations were performed with anti-CtBP2 antibody and analyzed by Western blot (WB) with anti-Myc antibody. *IgG heavy chain (C) CtBP2 proteins were upregulated in 6Myc-cIAP1 RM (top) or BIR1-BIR3 (bottom)-expressed U2OS. (D) U2OS cells were transfected with 6Myc-cIAP1 FL, 6Myc-cIAP1 RM, or 6Myc-cIAP1 BIR1-BIR3. After 24 h, the cells were irradiated with 20 J/m2 UV/C and incubated for an additional 24 h. Whole cell lysates were analyzed by WB with the indicated antibodies. The numbers below the CtBP2 bands indicate the relative amount of CtBP2 to actin (C, D) as determined by densitometry and the NIH Image J program. Fig. 4. Modulation of CtBP2 level by cIAP1 affects the transcription of CtBP2 target genes and subsequent cell migration. (A) Expression levels of CtBP2 target genes were quantified in scRNA-, siCtBP2-, or sicIAP1-transfected U2OS cells by qRT-PCR. GAPDH was used as an internal control for qRT-PCR. Results are presented as the mean ± SD from triplicate samples (*P < 0.01, **P < 0.001, ***P < 0.0001; unpaired Student’s t-test). or N.S. ¼ not significant. U2OS cells expressing scRNA, siCtBP2 (B), or sicIAP1 (C) were seeded for analysis by the wound healing assay. Wound areas were measured at 0, 8, 16, and 24 h (left), and the change was calculated according to the formula described in Materials and Methods (middle). Whole cell lysates were analyzed by Western blot (WB) with the indicated antibodies (right). (D) Expression levels of Bik, Sirt1, and E-cadherin mRNA were quantified in empty vector (e.v.), 6Myc-cIAP1 FL-, 6Myc-cIAP1 RM-, or 6Myc-cIAP1 BIR1-BIR3-overexpressing U2OS cells by qRT-PCR. Results are presented as the mean ± SD from triplicate samples. (*P < 0.01, **P < 0.001, ***P < 0.0001; unpaired Student’s t-test) (left). Protein levels were assessed by WB using the indicated antibodies (right). Ultraviolet (UV) irradiation downregulates CtBP protein levels in a proteasome-dependent manner [25]. Thus, we hypothesized that cIAP1 might inhibit CtBP2 degradation induced by UV irradiation. We examined the effect of UV irradiation on CtBP2 level in USO2 cells expressing cIAP1 WT (6Myc-cIAP1 FL), a RING-mutant (6Myc- cIAP1 RM), and a BIR only mutant (6Myc-cIAP1 BIR1-BIR3). The results show that UV irradiation-induced downregulation of CtBP2 was reversed in the presence of cIAP1 WT and both mutants, with greater inhibition observed with cIAP1 RM and BIR1-BIR3 expres- sion (Fig. 3D). Thus, CtBP2 might be protected from UV-induced proteasomal degradation when cIAP1 WT or the mutants are expressed. More importantly, cIAP1 RING activity might not be necessary for this protection from degradation. 3.4. Modulation of CtBP2 level by cIAP1 affects the transcription of CtBP2 target genes and subsequent cell migration Next, we hypothesized that the level of CtBP2 protein modu- lated by cIAP1 might affect the transcription of CtBP2 target genes. To test our hypothesis, we assessed the expression level of CtBP2 target genes, namely Bik [11], Sirt1 [26], and E-cadherin [2], in U2OS cells depleted of cIAP1 or CtBP2. Our results show that the mRNA level of all tested genes were increased similarly in cIAP1-depleted and CtBP2-depleted cells compared to that in control cells (Fig. 4A). E-cadherin is the main effector of cancer cell migration, and loss of CtBP2 expression was found to upregulate this gene, thereby inhibiting cell migration in breast cancer [6]. Thus, we examined the effect of cIAP1 and CtBP2 knockdown on cell migration using the wound healing assay. Cell migration was reduced similarly in cIAP1- and CtBP2-depleted cells compared to control cells (Fig. 4B and C, left and center). As expected based on results shown in Fig. 2B and D, levels of CtBP2 protein were reduced in cIAP1- depleted cells (Fig. 4C, right). These results indicate that, in cIAP1-depleted cells, CtBP2 was degraded, which led to upregula- tion of E-cadherin expression and inhibition of cell migration. Since the steady-state level of CtBP2 protein was upregulated in the presence of the cIAP1 FL, RM, and BIR1-BIR3 mutants (Fig. 3C and D), we examined whether overexpression of these cIAP1s also affect transcription of CtBP2 target genes. Our qRT-PCR results show that the transcription of CtBP2 target genes was repressed in cIAP1 FL-, RM-, and BIR1-BIR3-expressing cells (Fig. 4D, left). Re- sults from Western blot analysis demonstrate that the level of CtBP2 protein was increased following overexpression of cIAP1 FL and both cIAP1 mutants (Fig. 4D, right). Altogether, these results suggest that the stabilization of CtBP2 by cIAP1 influences the transcription of CtBP2 target genes, thereby affecting cell migration. Under normal conditions, CtBP2 also acts as a transcriptional repressor. However, when overexpressed, CtBP2 can function as an oncogene by repressing the expression of tumor suppressor genes [2,11,26]. Therefore, CtBP2 protein levels must be precisely regu- lated in many ways for normal cellular function to occur. We pre- viously reported that CHIP, a cochaperone E3 ligase, ubiquitinates CtBP2 to trigger its proteasomal degradation [27]. The present study provides another line of CtBP2 protein regulation in which cIAP1 stabilizes the steady-state level of CtBP2. Whether CtBP2 K63-linked ubiquitination truly protects CtBP2 from degradation remains unknown (Fig. 2D). However, the BIR domains of cIAP1, especially BIR1, are capable of binding to and protecting CtBP2 without the RING domain (Fig. 2). Expression of the cIAP1 BIR domain was observed exclusively in nucleus [28], [Fig. 1C]. In cells, cIAP1 is cleaved by caspase 3 into two fragments, an N-terminal one containing the BIR domains and a C-terminal one with the RING domain [29], suggesting that the BIR domain may exist apart from the RING domain to stabilize CtBP2 in the nucleus for its normal co- repressor function. The physiological relevance of cIAP1 and CtBP2 to the pathogenesis of human cancer has never been addressed. However, besides CtBP2, cIAP1 amplification or overexpression is also implicated in various cancers [30]. Thus, it is plausible that cIAP1 might protect and upregulate the steady-state level of CtBP2 in transformed cells. Our results propose that cIAP1 might be a useful target in strategies aiming to downregulate SPOP-i-6lc the steady-state level of CtBP2 protein in treating human cancers.