Phosphorylation of the Anaphase-promoting Complex/Cdc27 Is Involved in TGF-β Signaling

Abstract

Loss of TGF-β-induced growth inhibition is a hallmark of many human tumors. Previous studies implied that activation of the anaphase-promoting complex (APC/cyclosome) is involved in the TGF-β signaling pathway, which facilitates the destruction of SnoN, a transcriptional co-suppressor, which leads in turn to the transactivation of TGF-β-responsive genes for cell cycle arrest. The function of APC was demonstrated in TGF-β signal transduction, but the mechanism by which it is activated in response to TGF-β signaling remains unclear. We report here that phosphorylation of Cdc27, a core subunit of APC, in response to TGF-β signaling can facilitate the activation of APC. We have demonstrated that casein kinase II (CKII) is involved in the phosphorylation of Cdc27 in response to TGF-β signaling. Depletion of CKII by shRNA abolishes the TGF-β-induced phosphorylation of Cdc27 and subsequent degradation of SnoN. Disruptive mutation of Cdc27 (S154A) attenuates TGF-β-induced SnoN degradation. In addition, expression of a phosphorylation-resistant Cdc27 mutant significantly attenuates TGF-β-induced growth inhibition. Together, the results suggest that phosphorylation of Cdc27 by CKII is involved in TGF-β-induced activation of APC.

Introduction

TGF-β is an inhibitory growth factor for a variety of epithelial cells. Loss of TGF-β growth inhibition is a hallmark for many types of human tumors (1–4). SnoN and its partner Ski are proto-oncoproteins in that they play critical roles in coordinating the balance between proliferation and differentiation during early development in mice (5). SnoN and Ski, two co-repressors of transcription, are rapidly degraded upon stimulation by TGF-β in an ubiquitin-dependent manner (6, 7). The snoN gene is localized on chromosome 3q26, an oncogene locus that is often amplified in human cancer (8). SnoN, as well as its three alternative splice variants (SnoN2, SnoI, and SnoA), is expressed in humans. The expression levels of SnoN and Ski are important for their biological function. An excess amount of SnoN induces oncogenic transformation of normal cells (9, 10). Overexpression of SnoN in TGF-β-responsive cells blocks TGF-β-induced cell growth arrest (6, 7). In normal cells, the levels of SnoN have to be tightly regulated to maintain a normal sensitivity to TGF-β. Therefore, proteolysis provides a crucial mechanism to modulate its protein level. It is thought that the turnover of specific inhibitors of transcription is required for induction of TGF-β-responsive genes. We (11) and others (12) have shown that the anaphase-promoting complex (APC)³ is the E3 ubiquitin ligase activated by TGF-β signaling that allows for the degradation of SnoN. SnoN and its three splice variants contain a functional destruction motif (RXXLXXXX(N/D)) that ensures the recognition of SnoN by APC. Cdh1 serves as a substrate specificity factor targeting SnoN for E3 ligase APC. However, the mechanism by which E3 ligase APC is activated by TGF-β remains unknown. Elucidation of this mechanism will allow a thorough understanding of the TGF-β signaling pathway.

Previous work demonstrated the TGF-β-induced interaction among Smad2/3, Cdh1, and Cdc27 (11). The formation of a ternary complex of Smad2/3-Cdh1-Cdc27 in response to TGF-β signaling has drawn our attention to Cdh1 and Cdc27 as components necessary to elevate APC activity in response to TGF-β (11, 12). Previous work suggested that regulation of Cdh1 during the cycle or in response to environmental signaling is mediated principally by a kinase/phosphatase circuitry (13, 14), where Cdc14 and some unknown kinases are players that govern Cdh1-APC. On the other hand, phosphorylation of Cdc27 serves as an important mechanism in the activation of APC at a different level, where whether Cdc27 is phosphorylated would determine the affinity between APC and its substrate specificity activator, Cdc20 or Cdh1 (15). Our initial efforts to study the alteration of Cdh1 in response to TGF-β did not provide us any clues as to how APC is activated by TGF-β signaling. However, our efforts spent on the dissection of the TGF-β-induced alteration of Cdc27 have revealed a dramatic change in the phosphorylation of Cdc27 induced by stimulation with TGF-β, which sheds light on the mechanism via which APC is activated by TGF-β signaling.

We have combined the approaches of mass spectrometry and mutagenesis to identify the candidate kinase that facilitates TGF-β-induced Cdc27 phosphorylation. We have characterized the function of the candidate kinase in mediating TGF-β signaling as phosphorylating Cdc27, which in turn destroys SnoN, allowing the transcription of genes necessary for growth inhibition. Our results suggest that casein kinase II is a TGF-β-responsive kinase that activates APC via phosphorylation of Cdc27. Enhanced interaction between Cdh1 and Cdc27 through TGF-β-induced phosphorylation of Cdc27 could be a mechanism that activates APC activity.

EXPERIMENTAL PROCEDURES

Antibodies and Chemicals

Anti-SnoN antibody (ABM-3002) was obtained from Cascade BioScience, Inc. (Winchester, MA). Anti-HA antibody (sc-805) and anti-casein kinase IIα antibody (sc-6479) were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Anti-casein kinase IIβ antibody (612625) was from Pharmingen. Anti-casein kinase Iϵ antibody (610445) was from BD Transduction Laboratories. Anti-CaMK2 antibody (3362) was from Cell Signaling (Danvers, MA). Anti-tubulin antibody (T-5168) was from Sigma. Anti-Cdc27 polyclonal antibody was a gift from Dr. Marc W. Kirschner (15). TGF-β was from R&D Systems (Minneapolis, MN). Protease mixture inhibitor was from Roche Applied Science.

Plasmids, Mutagenesis, and Engineering of Stable Cell Lines

The primers used for constructing Cdc27 mutants were as follow: Cdc27F, 5′-aaaaggccggccaaccATGACGGTGCTGCAGGAACCC-3′; Cdc27R, 5′-ttggcgcgccAAATTCATCACTTTCAGCTGCATG-3′; Cdc27(R412), 5′-ttggcgcgccTGTCTGTTTGGGATTTTAGGTGG-3′; Cdc27(F101), 5′-aaaaggccggccaaccATGAGCCATGATGATATTGTT-3′; Cdc27(F276), 5′-aaaaggccggccaaccATGCCATTAACCCCAAGTTTT-3′;Cdc27(F313), 5′-aaaaggccggccaaccATGGGAGCCCCTTCAAAAAAG-3′; Cdc27(F418), 5′-aaaaggccggccaaccATGGGAGGAATAACTCAACC-3′; Cdc27S154(F), 5′-CCTCTGGTCTCCCTTTGAAGCATTATGTGAAATAGGTG-3′; and Cdc27S154(R), 5′-CACCTATTTCACATAATGCTTCAAAGGGAGACCAGAGG-3′. Cdc27 mutants were cloned into the pCS2-HA (C terminus) and pREX-IRES-GFP vectors (a gift from Dr. Xuedong Liu, University of Colorado, Boulder, CO).

SnoNΔdb (D box-deleted SnoN) was generated by deleting amino acids 164–72 by a PCR approach based on the parental plasmid pCS2-HA-SnoN. The primers used for construction of this mutant were as follows: 5′-GTTTCAAGTTGGAGGAGGATTCTCTGTTCTCCGAGAATTTAC-3′ and 5′-GTAAATTCTCGGAGAACAGAGAATTCTCCTCCAACTTGAAAAC-3′. SnoNmdb (D box-mutated SnoN) was generated by replacement of arginine and leucine with alanine at amino acids 164 and 167 by site-directed mutagenesis. HA-WT SnoN and HA-SnoNΔdb were then subcloned from the pCS2 vector into the pREX-IRES-GFP vector.

To establish stable cell lines, including Mv1Lu-HA-SnoN, Mv1Lu-HA-SnoNΔdb, Mv1Lu-HA-Cdc27, and Mv1Lu-HA-Cdc27(S154A), pREX-HA-SnoN-IRES-GFP, pREX-HA-SnoNΔdb-IRES-GFP, pREX-HA-Cdc27-IRES-GFP, and pREX-HA-Cdc27(S154A)-IRES-GFP were transfected into the Phoenix packaging line, respectively. The packaged retroviral particles were collected, mixed with Polybrene, and added to target Mv1Lu cells. The GFP-positive cells were sorted by flow cytometry for further assay (7).

Preparation of TGF-β-stimulated Mink Lung Epithelial Cell Extracts

Mink lung epithelial cells (Mv1Lu) were cultured for 24 h in DMEM supplemented with 10% FBS in 5% CO₂. To stimulate cells with TGF-β, they were treated with 100 pm TGF-β. Cells were washed with PBS and harvested by scraping at designated time points. Approximately 1 × 10⁸ harvested cells were resuspended in 500 μl of hypotonic buffer (20 mm HEPES (pH 7.5), 5 mm KCl, 1.5 mm MgCl₂, 1 mm DTT, 1× protease mixture, and energy regeneration mixture) for 30 min to allow cells to swell. Cells were frozen by liquid nitrogen, thawed in a 37 °C water bath, and homogenized with 10 strokes using a Dounce homogenizer. Cell lysates were spun in an Eppendorf microcentrifuge at 14,000 rpm for 1 h at 4 °C. The clear supernatant was collected using a syringe needle and used directly for protein degradation assays (11).

In Vitro Ubiquitylation Assay

Mv1Lu cells were stimulated with TGF-β and harvested at designated time points. APC was purified from the cell lysates using anti-Cdc27 antibody coupled to protein A beads and subjected to ubiquitylation assay in a ubiquitylation mixture. In vitro translated ³⁵S-labeled cyclin B was used as the substrate for APC. Reaction mixtures were incubated for 45 min at room temperature, resolved by SDS-PAGE, and visualized by radiography (11).

Immunoprecipitation Assays

Cell pellets collected at designated time points were lysed in lysis buffer on ice for 30 min. Then, 27-gauge 0.5-inch syringes were used to shred the DNA. The supernatants were collected after centrifugation at 12,000 × g for 30 min. Primary antibody was added to the lysates. After rotation overnight at 4 °C, immobilized protein A/G beads were added to the tubes. After rotation again at 4 °C for 4 h, the beads were collected by centrifugation at 2500 × g for 3 min. Electrophoresis loading buffer was added to the beads after washing. After denaturing at 95 °C for 5 min, the supernatants were subjected to SDS-PAGE and Western blotting.

Mobility Shift Assay

2 μl of in vitro translated ³⁵S-labeled mutant Cdc27 was added to 50 μl of Mv1Lu extracts prepared from cells exposed to TGF-β at different time points or left untreated. The extracts were incubated at room temperature for 30 min. Phosphorylation of Cdc27 reflected by mobility shift was measured by SDS-PAGE-coupled autoradiography (11).

Kinase Assay

In vitro synthesized ³⁵S-labeled mutant Cdc27 proteins were washed twice with buffer containing 100 mm Tris-HCl (pH 8.0), 5 mm EGTA, and 20 mm MgCl₂. The samples were then incubated with the same buffer supplemented with mock IgG beads or immunopurified casein kinase II (CKII). The reaction mixtures were incubated, subjected to SDS-PAGE, and detected by autoradiography.

For evaluation of Cdc27 phosphorylation directly by CKII, ∼10 ng of in vitro synthesized ³⁵S-labeled Cdc27 was incubated with immunopurified CKII complex (Mv1Lu cell lysates were prepared from cells treated with TGF-β at different time points) and kinase buffer supplemented with 10 mm ATP (16). Kinase reaction samples were resolved by SDS-PAGE. Cdc27 phosphorylation reflected by mobility shift was estimated by autoradiography.

Knockdown of CKII

Cells were transfected with CKIIα-specific siRNAs (AACAUUGAAUUAGAUCCACGUUU) or randomized mixtures of double-stranded RNA as a control. At 24 h post-transfection, cells were transfected again with the same siRNA preparations to ensure efficient depletion. To stimulate cells with TGF-β, they were treated with 100 pm TGF-β. Cells were then washed with PBS and harvested by scraping at designated time points. Cell extracts were resolved on SDS-polyacrylamide gels, followed by immunoblotting.

Growth Inhibition Assay

For growth inhibition assay, 5 × 10³ Mv1Lu cells were incubated with various concentrations of TGF-β1 for 3–4 days. The growth of cells was determined by counting and comparison of stimulated versus unstimulated cells (17).

RESULTS

Activity of APC Is Enhanced in Response to TGF-β Stimulation

Previous studies indicated that the activity of APC is regulated in response to TGF-β signaling (11, 12). To biochemically estimate the alteration in APC activity upon stimulation with TGF-β, we measured the early response of TGF-β in regulating APC by an in vitro ubiquitylation assay using cyclin B as a putative in vitro substrate (11). As shown in Fig. 1A, APC activity was profoundly activated in response to TGF-β. The elevated APC activity was sustained for 3–4 h and quenched after 5 h (Fig. 1A and supplemental Fig. 1). Consistent with the previous observation, SnoN protein levels dramatically dropped in response to TGF-β signaling (Fig. 1B). The TGF-β-induced decrease in SnoN protein levels was significantly retarded in the presence of MG132, a proteasomal inhibitor (Fig. 1B). Depletion of Cdh1, an activator of APC, by RNA interference significantly abrogated APC-mediated SnoN degradation in response to TGF-β signaling (Fig. 1, B and C). Mutation of the D box in SnoN, a specific degron for substrate recognition by APC, stabilized SnoN in the presence of TGF-β (Fig. 1C). Interference with SnoN degradation by the expression of stabilized SnoN or disruption of Cdh1 function antagonized the TGF-β-induced cytostatic effect (Fig. 1, D and E).

FIGURE 1.

Activity of APC is enhanced in response to TGF-β stimulation. A, the activity of APC is enhanced following stimulation by TGF-β. Mv1Lu cells were stimulated with TGF-β and harvested at different times as indicated. APC was subsequently purified from the cell lysate using anti-Cdc27 antibody coupled to protein A beads and subjected to a ubiquitylation mixture in a ubiquitylation assay. In vitro translated ³⁵S-labeled cyclin B was used as a putative substrate for APC activity analysis. Polyubiquitin conjugates of cyclin B were measured by autoradiography. B, knockdown of Cdh1 attenuates SnoN degradation in response to TGF-β stimulation. C, protein levels of stably expressed wild-type SnoN and non-degradable SnoN in response to TGF-β signaling. Upper panel, construction of retroviral expression vectors for wild-type and mutant SnoN. Lower panel, protein expression levels for wild-type and mutant SnoN in response to TGF-β signaling. IRES, internal ribosome entry site. D, expression of degradation-resistant mutant SnoN antagonizes TGF-β signaling. Mv1Lu cells stably expressing SnoN or SnoNΔdb were incubated for 4 days with various concentrations of TGF-β1 as indicated. The growth of cells was quantified by cell counting and compared with the growth of unstimulated cells. E, depletion of Cdh1 blocks TGF-β-induced growth inhibition.

APC Activity Can Be Elevated via TGF-β-induced Cdc27 Hyperphosphorylation

Our results detail the role of APC in mediating TGF-β-induced growth inhibition via degradation of SnoN, a co-suppressor for TGF-β-regulated transactivation. However, the mechanism by which TGF-β activates the ubiquitin ligase APC remains a mystery. To elucidate this mechanism, we investigated whether phosphorylation of major APC subunits, including Cdc27, Cdc23, and Cdc16, in response to TGF-β signaling could be the mechanism for enhanced activity of APC in the presence of TGF-β.

Previous investigators have shown that for both embryonic and somatic cells, post-translational modification of APC is critical for its activation (15, 18). They further revealed that the biological consequence of phosphorylation of APC is to create a favorable conformation in APC for its mitotic activator Cdc20. Coordinated interaction between APC and Cdc20 results in the full activation of APC for degradation of securin. To test the hypothesis that stimulation by TGF-β induces the post-translational alteration of APC, we labeled Mv1Lu cells with [γ-³²P]ATP (19). Mv1Lu cells metabolically labeled with [γ-³²P]ATP were treated with TGF-β (100 pm) for 30 min and subsequently harvested for extraction. For purification of the APC complex, an anti-Cdc27 polyclonal antibody was cross-linked to protein A-Affi-Gel beads (Bio-Rad) as described previously (11). Using anti-Cdc27 beads, APC metabolically labeled with [γ-³²P]ATP was immunopurified from cells stimulated with TGF-β and incubated with [γ-³²P]ATP. Purified APC complex was intensively washed and resolved by SDS-PAGE. Phosphorylation of APC subunits in response to TGF-β stimulation was detected by autoradiography. As shown in Fig. 2A (panel a), Cdc27 was significantly hyperphosphorylated when it was induced by TGF-β signaling. On the other hand, Cdc23 and Cdc16 were not hyperphosphorylated in response to TGF-β. Phosphorylation of Cdc27 in response to TGF-β signaling was further suggested by immunoprecipitation, where the mobility of Cdc27 from the cells exposed to TGF-β was shifted, as shown in Fig. 2A (panel b). To verify that Cdc27 is phosphorylated in response to stimulation with TGF-β, we attempted to identify the phosphorylation site on Cdc27 by mass spectrometry. Consistent with the autoradiography results, purified Cdc27 from cells stimulated with TGF-β was shifted and smeared when it was silver-stained (Fig. 2A, panel c). The smeared band was confirmed as Cdc27 by mass spectrometry.

FIGURE 2.

Elevation of APC activity could be via TGF-β-induced Cdc27 hyperphosphorylation. A, panel a, incorporation of [γ-³²P]phosphate into Cdc27 (P-Cdc27) is increased in response to TGF-β. Panel b, Cdc27 is hyperphosphorylated in response to TGF-β signaling. The mobility of Cdc27 protein shifted and smeared in response to TGF-β. Panel c, the mobility shift of Cdc27 protein in response to TGF-β is confirmed by mass spectrometry. B, Cdc27 protein mobility is shifted in response to TGF-β signaling in a cell-free system. Cdc27 mobility was shifted in TGF-β-stimulated Mv1Lu (ML) cell extracts. The Cdc27 mobility shift was blocked by λ-phosphatase (λ-PPase). IP, immunoprecipitate; IB, immunoblot.

To assay the phosphorylation of Cdc27 in response to TGF-β, we developed a somatic cell extract system for studying phosphorylation of Cdc27 by TGF-β signaling. This system is able to detect the modification (phosphorylation and degradation) of various components of the signal transduction pathway in response to ligand stimulation as described previously (20). As shown in Fig. 2B, the mobility of the Cdc27 (but not Cdc23 or Cdc16) protein shifted significantly in response to TGF-β stimulation (data not shown). The protein mobility shift of Cdc27 in response to TGF-β was attenuated by λ-phosphatase, which further suggests that Cdc27 is phosphorylated in response to TGF-β.

Identification of a Candidate Kinase Involved in the Phosphorylation of Cdc27 by TGF-β

Cdc27 is a tetratricopeptide repeat (TPR) protein that contains eight TPR domains. One TPR domain is localized near its N terminus, and the other seven TPR domains are distributed near the C terminus, as indicated in Fig. 3A (panel a). The TPR domain is thought to mediate protein-protein interactions (21). The TPR region on the C terminus of Cdc27 has been shown to facilitate binding to the substrate specificity activator Cdh1 (21). The phosphorylation of Cdc27 during mitosis has been described previously (18). Multiple phosphorylation sites on Cdc27 have been suggested to be important in mitosis for its regulation, although phosphorylation of Cdc27 in G₁ is unknown. Using the PHD motif search program combined with Scansite (set at medium stringency), 18 putative phosphorylation sites on Cdc27 were predicated, as shown in Fig. 3A (panel b).

FIGURE 3.

Putative phosphorylation sites and motifs in Cdc27. A, panel a, diagram of the distribution of phosphorylation sites in Cdc27. The shaded regions represent the TPR domains. Panel b, putative phosphorylation sites and motifs that potentially regulate Cdc27 in response to TGF-β signaling. B, construction of a series of Cdc27 mutants for mapping TGF-β-induced phosphorylation sites in Cdc27. Panel a, diagram of the strategy for mutagenesis of Cdc27. Panel b, in vitro synthesized ³⁵S-labeled mutant Cdc27 proteins. C, mapping the phosphorylation region of Cdc27 by a protein mobility shift assay. Panel a, evaluation of the mobility shift for a set of Cdc27 mutants in TGF-β-stimulated extracts. Panel b, summary of mobility shift analysis.

On the basis of the distribution of the predicted phosphorylation sites, we designed five truncation mutants, Cdc27(1–412), Cdc27(413–824), Cdc27(101–824), Cdc27(276–824), and Cdc27(313–824), as indicated in Fig. 3B (panels a and b). The truncated mutant fragments were amplified using PCR and subsequently cloned into the pCS2(+) vector. All Cdc27 mutants were translated in vitro as shown in Fig. 3B (panel b). To identify the region on Cdc27 that is phosphorylated in response to TGF-β signaling, we detected the protein mobility shift for the five Cdc27 mutants using the cell extract prepared from the cells exposed to TGF-β ligand (20). As shown in Fig. 3C (panels a and b), the protein mobility of full-length Cdc27, Cdc27(1–412), and Cdc27(101–824) was significantly shifted in response to stimulation with TGF-β signaling, but no significant protein shift was observed for Cdc27(413–824), Cdc27(276–824), and Cdc27(313–824). These results suggest that the region from amino acids 101–276 is probably the region that is phosphorylated in response to TGF-β, which provides us with a critical insight into the identity of the kinase that regulates Cdc27 in the TGF-β signaling pathway.

CKII Is the Putative Kinase That Phosphorylates Cdc27 in Response to TGF-β

Our results from mapping the general phosphorylation region of Cdc27 allowed us to focus on the region between amino acids 101 and 276 (Fig. 3C). On the basis of the peptide sequence analysis, we predicted several putative phosphorylation sites (Fig. 3A). There are seven phosphorylation motifs localized in this region, including sites for Ser¹⁰² (CDK2), Ser¹⁵⁴ (CKII), Thr¹⁶⁷ (Polo-like kinase), Ser¹⁸³ (casein kinase I (CKI)), Ser²³⁰ (CKI), Ser²⁶⁷(CDK1), and Ser²⁷⁶ (Erk1). To test whether the activities of these kinases are altered in response to TGF-β signaling, we performed kinase activity assay by incubating the immunopurified kinase from the Mv1Lu cells exposed to TGF-β at various time points with its corresponding in vitro substrate and [³²P]ATP (22–25). As shown in Fig. 4A, the activities of Erk1, CKI, and CKII (but not CDK1) were significantly increased in response to stimulation with TGF-β signaling. We also did not observe alteration of the kinase activity of CDK2 and Polo-like kinase when stimulated with TGF-β (data not shown).

FIGURE 4.

Identification of CKII as a candidate kinase that phosphorylates Cdc27 in response to TGF-β signaling. A, activation of CKII as well as CKI and MAPK is responsible for TGF-β signaling. B, Cdc27 partially interacts with CKIIα and CKIIβ in response to stimulation with TGF-β, as demonstrated by immunoprecipitation (IP). C, summary of immunoprecipitation of Cdc27 with several candidate kinases. D, the mobility of Cdc27 is shifted only when catalyzed by immunopurified CKII, but not by other candidate kinases. IB, immunoblot; Plk, Polo-like kinase; P-Cdc27, phospho-Cdc27; MBP, myelin basic protein.

Based on the above kinase activity assay, the potential kinases that could regulate Cdc27 in the TGF-β signaling pathway were narrowed down to three kinases, Erk1, CKI, and CKII. To test whether these kinases can phosphorylate Cdc27 in response to TGF-β, we investigated the possible interaction between Cdc27 and these candidates in the absence and presence of TGF-β signaling, although such interaction could be very transient. As shown in Fig. 4 (B and C), Cdc27 was co-immunoprecipitated with CKIIα and CKIIβ, whereas no obvious interaction between Cdc5, Polo-like kinase, CaMK2, and CKI with Cdc27 was observed. We further carried out Cdc27 mobility shift analysis by incubating ³⁵S-labeled Cdc27 with immunopurified endogenous Erk1, CKI, and CKII from cells exposed to TGF-β for 30 min. As indicated in Fig. 4D, only CKII (but neither Erk1 nor CKI) was able to catalyze a shift in protein mobility for Cdc27, suggesting that CKII could be the kinase that phosphorylates Cdc27 and regulates APC-mediated destruction of SnoN in response to TGF-β signaling (16).

To determine whether CKII could be the kinase that mediates TGF-β-induced Cdc27 phosphorylation, we depleted CKII by RNA interference in Mv1Lu cells and then examined the effect of CKII depletion on TGF-β-induced Cdc27 phosphorylation (Fig. 5A). As shown in Fig. 5B, whereas TGF-β-induced Cdc27 phosphorylation was detected in control cells, TGF-β-induced Cdc27 phosphorylation was attenuated in CKII-depleted cells. In addition, TGF-β-induced SnoN degradation was also significantly blocked in response to CKII depletion (Fig. 5, C and D). Taken together, the above results suggest that CKII could be the kinase that mediates TGF-β-induced Cdc27 phosphorylation and SnoN degradation.

FIGURE 5.

Depletion of CKII attenuates TGF-β-induced Cdc27 phosphorylation, which in turn abolishes TGF-β-induced SnoN degradation. A, depletion of CKIIα in Mv1Lu cells. B, depletion of CKIIα attenuates TGF-β-induced Cdc27 hyperphosphorylation. C, knockdown of CKIIα abolishes TGF-β-induced SnoN degradation. D, summary of TGF-β-induced SnoN degradation in wild-type and CKIIα-depleted Mv1Lu cells.

Expression of Phosphorylation-resistant Mutant Cdc27 Alters TGF-β-induced Growth Inhibition

Interruption of TGF-β-induced Cdc27 phosphorylation could abolish TGF-β-induced APC activation and therefore could antagonize the TGF-β response. To test this hypothesis, we initially engineered a phosphorylation-resistant Cdc27 mutant by mutating the CKII phosphorylation site at Ser¹⁵⁴ to Arg. HA-tagged wild-type or mutant Cdc27 was cloned into a retroviral vector (Fig. 6A). Stable Mv1Lu cell lines expressing either HA-Cdc27 or HA-Cdc27(S154A) were then established (Fig. 6A). As indicated in Fig. 6B, expression of HA-Cdc27(S154A) significantly attenuated the TGF-β-induced phosphorylation of endogenous Cdc27. In contrast, expression of HA-Cdc27 did not alter Cdc27 phosphorylation in response to TGF-β signaling (Fig. 6B). In addition, TGF-β-induced SnoN degradation was significantly blocked by the expression of phosphorylation-resistant mutant Cdc27, whereas no obvious alteration of TGF-β-induced SnoN degradation was measured in Mv1Lu cells stably expressing wild-type Cdc27 (Fig. 6, B and C). To further determine the impact of Cdc27 phosphorylation status on alteration of TGF-β/APC-mediated SnoN degradation and TGF-β-induced growth inhibition, we examined the effect of TGF-β on cell growth in Mv1Lu cells expressing either wild-type or mutant Cdc27. As shown in Fig. 6D, TGF-β induced significant growth inhibition in the Mv1Lu-HA-Cdc27 cells. In contrast, alteration of Cdc27 phosphorylation status significantly antagonized the TGF-β-induced growth inhibition in Mv1Lu-HA-Cdc27(S154A) cells. Therefore, disruption of TGF-β-induced Cdc27 phosphorylation impairs the TGF-β-induced growth inhibitory effect. These results support the hypothesis that TGF-β-induced activation of APC is possibly mediated by CKII (Fig. 7). APC activated by TGF-β signaling targets SnoN for ubiquitylation and subsequent degradation, resulting in the activation of TGF-β-mediated transcription for growth inhibition (12).

FIGURE 6.

Disruption of the CKII-mediated phosphorylation site in Cdc27 blocks TGF-β-induced SnoN degradation, which in turn attenuates TGF-β-mediated growth inhibition. A, mutation of Ser¹⁵⁴ in Cdc27 abolishes TGF-β-induced hyperphosphorylation of Cdc27. Upper panel, retroviral expression vector constructed for wild-type and mutant Cdc27. Lower panel, suppressed Cdc27 hyperphosphorylation in Mv1Lu cells with expression of mutant Cdc27. B, expression of the phosphorylation-resistant Cdc27 mutant attenuates TGF-β-induced SnoN destruction. C, summary of SnoN degradation in wild-type and phosphorylation-resistant mutant Cdc27-expressing Mv1Lu cells. D, expression of the phosphorylation-resistant Cdc27 mutant antagonizes TGF-β-induced growth inhibition. IRES, internal ribosome entry site; IP, immunoprecipitate; IB, immunoblot.

FIGURE 7.

Proposed model for activation of APC by CKII in response to TGF-β signaling.

DISCUSSION

TGF-β Modulates APC Activity via Regulating Cdc27 Phosphorylation

We (11) have previously shown that TGF-β-induced activation of APC targets SnoN, a transcriptional co-suppressor, for destruction, thereby resulting in the induction of TGF-β-responsive genes such as p21 for growth inhibition. We also demonstrated that TGF-β-activated APC targets Skp2 for degradation, leading to the stabilization of p27, which is necessary for the down-regulation of cyclin E/CDK2 for tumor suppression (17). We have observed by [γ-³²P]ATP incorporation, protein mobility shift assay, and mass spectrometry analysis that Cdc27 is hyperphosphorylated in response to TGF-β signaling. Our data suggest that the consequence of TGF-β-induced Cdc27 phosphorylation is to enhance interaction between Cdc27 and Cdh1, an upstream substrate specificity factor for APC (15). Recapitulation of SnoN ubiquitylation by APC with purified and phosphorylated Cdc27 suggested that the TGF-β-induced Cdc27 phosphorylation is a critical event to achieve elevated APC activity for SnoN ubiquitylation and Skp2 degradation. The identification of CKII as a putative kinase that mediates TGF-β signaling and phosphorylates Cdc27 filled a gap in our understanding of the mechanism by which APC is activated in response to TGF-β signaling.

Activation of APC/Cdc27 by CKII Involves TGF-β Signaling

CKII, a ubiquitous serine/threonine protein kinase, regulates a variety of cellular processes, including cell growth, differentiation, apoptosis, and DNA repair (26–29). It is composed of two catalytic subunits (∼42-kDa α and 38-kDa α′) and two regulatory subunits (∼28-kDa β) in α₂β₂, αα′β₂, or α′₂β₂ configuration (26, 27, 30, 31). Previous genetic studies have shown that CKII is essential for cell cycle progression through the G₁/S and G₂/M transitions, where several important cell cycle regulators such as CDK-activating kinase and Cdc34 are tightly regulated by CKII (32). It has been demonstrated that both CKI and CKII are involved in TGF-β signal transduction (25, 33). Studies on ALK-1 (a member of the TGF-β superfamily of receptors) in endothelial cells implied a critical role for CKIIβ in signal transduction by phosphorylating Smad1/5/8 in response to TGF-β1 and BMP-9 (34). Yeast two-hybrid screening for Smad3 interaction led to the identification of an interaction between Smad3 and CKI, where phosphorylation of Smad3 by CKI could cause the subsequent ubiquitylation and degradation of Smad3 (33). Singh and Ramji (35) reported that TGF-β-induced expression of the apolipoprotein E gene requires CKII function. The role of CKII was also linked to TGF-β-induced apoptosis and expression of the type IV collagen gene (36). Consistent with the above observations, our work showed increased CKII activity in response to stimulation with TGF-β. We demonstrated that the TGF-β-responsive activation of CKII is necessary for TGF-β-induced Cdc27 phosphorylation. Enhanced Cdc27 phosphorylation by CKII potentially generates a favorable conformation for APC to communicate with its activator Cdh1 and to interact with its substrate SnoN (15).

Phosphorylation of APC/Cdc27 Orchestrates APC Activity during the Cell Cycle and Signal Transduction

APC has been shown to be a critical regulator bridging various environment signals, including genotoxic stress, mitogenic signaling, and TGF-β, to induce downstream cellular responses (20, 37, 38). Regulation of APC can be achieved at several levels, including phosphorylation of APC, interaction with the Cdc20/Cdh1 activators of APC, and interaction with the Emi1/Cdc14 regulators of Cdc20 and Cdh1 (39, 40). Phosphorylation of Cdc27 was thought to be crucial for activation of APC in mitosis (15, 18). A previous biochemical study using Xenopus egg extract showed that the Cdc27 phosphorylation status determines the affinity between APC and its substrate specificity activator Cdc20, thereby ensuring the activation of APC to destroy anaphase inhibitors, thus permitting chromatid separation (15). Our data suggest that CKII could mediate Smad3-transmitted TGF-β signaling by phosphorylating Cdc27. The consequence of TGF-β-induced phosphorylation of Cdc27 is a greater affinity between the activator Cdh1 and Cdc27, thereby achieving activation of APC. Although our work at a different level has sketched a mechanistic model of how CKII phosphorylates Cdc27, there is still gap that needs to be filled, especially with regard to whether the TGF-β receptor or Smad3 directly or indirectly regulates CKII.

Future Directions

Our biochemical dissection has filled a gap in our knowledge of the mechanism by which APC is activated in response to TGF-β signaling, specifically implying the role of CKII. The combinatorial approach of kinase assay, protein mobility shift assay, and mass spectrometry effectively provided us the tools to explore the mystery of TGF-β-induced activation via Cdc27 phosphorylation by CKII. Results based on the immunoprecipitation suggested that elevated Cdc27 phosphorylation by CKII contributes to a greater affinity between APC and its substrate specificity factor Cdh1, which ensures successful targeting of the SnoN or Skp2 substrate for ubiquitylation. Alteration of SnoN and Skp2 levels by TGF-β-activated APC plays an important role in the TGF-β cytostatic effect, tumor suppression, and developmental regulation (11, 17, 41, 42). At this point, development of a chemical inhibitor that could specifically intercept communication between CKII and Cdc27 could provide a rational strategy to sensitize TGF-β-based therapeutic treatment. Thorough validation of the role of CKII in mediating the TGF-β-induced APC-SnoN-p21 cascade by a xenograft tumor model or a knock-in mouse model could greatly advance our understanding of the molecular basis of TGF-β signal transduction and provide additional valuable insight into novel treatment of cancer.