Association and regulation of casein kinase 2 activity by adenomatous polyposis coli protein

Abstract

Mutations in the adenomatous polyposis coli (APC) gene are responsible for familial adenomatous polyposis coli and also sporadic colorectal cancer development. By using antibodies raised against the N-terminal region of APC protein, we have detected the variable masses of endogenous APC proteins in individual cell lines established from human colorectal carcinomas caused by nonsense mutations of the gene. Phosphorylation of immunoprecipitates of full-length and truncated APC were observed in in vitro kinase reaction, indicating association of APC with protein kinase activity. The kinase activity complexed with APC was sensitive to heparin and used GTP as phosphoryl donor, suggesting an involvement of casein kinase 2 (CK2). Both CK2alpha- and beta-subunits were found to associate with APC in immunoprecipitates as well as in pull-down assays, with preferential interaction of APC with tetrameric CK2 holoenzyme. In synchronized cell populations, the association of APC with CK2 was cell cycle dependent, with the highest association in G(2)/M. Unexpectedly, APC immunoprecipitates containing full-length APC protein inhibited CK2 in vitro, whereas immunoprecipitates of truncated APC had little effect. This was confirmed by using recombinant APC, and the inhibitory region was localized to the C terminus of APC between residues 2086 and 2394. Overexpression of this fragment in SW480 cells suppressed cell proliferation rates as well as tumorigenesis. These results demonstrate a previously uncharacterized functional interaction between the tumor suppressor protein APC and CK2 and suggest that growth-inhibitory effects of APC may be regulated by inhibition of CK2.

Figure 1

Full-length and truncated forms of APC proteins. (A) Detection of full-length and truncated forms of APC proteins. Cell lysates from normal fibroblasts (6 mg of protein) and cancer cells (2 mg of protein) were incubated with anti-APC polyclonal antibody, and the resultant immunocomplexes were separated on SDS/5% PAGE and probed with monoclonal anti-APC antibody (OP44). Lanes: 1, TIG-3; 2, TIG-7; 3, DLD-1; 4, KMS-4; 5, KMS-8; 6, SW480. Exposure for TIG-3, TIG-7 was about two times longer than that for the other cell lines. (B) Association of protein kinase activity with APC. Cell lysates prepared from TIG-3, SW480, and KMS-8 cells were immunoprecipitated with anti-APC polyclonal antibody and incubated for 15 min in the kinase reaction mixture in the presence (+) or absence (−) of 10 ng/ml heparin. The reaction was stopped by adding sample buffer, and radioactive proteins were separated by SDS/5% PAGE and visualized by autoradiography. (C) The protein kinase utilizes GTP as well as ATP. The phosphorylation activity in the immunoprecipitates from SW480 cells was determined in the presence of either [γ-³²P]ATP (left lane) or [γ-³²P]GTP (right lane) in the reaction mixtures. Proteins were separated on SDS/5% PAGE, and phosphorylated proteins were visualized by autoradiography. Truncated APC is indicated by an arrowhead.

Figure 2

Association of CK2 with APC in vivo and in vitro. (A) Detection of CK2 in APC immunocomplex. Lysates (2 mg each) obtained from TIG-3 (lane 1), KMS-4 (lane 2), SW480 (lane 3), or KMS-8 (lane 4) cells were immunoprecipitated with polyclonal anti-APC antibody and analyzed by SDS/12% PAGE, followed by immunoblotting with anti-CK2α, anti-CK2β, and β-catenin. (B) Recombinant GST-CK2α, GST-CK2α′, and GST-CK2β were separated by SDS/12% PAGE, and proteins were visualized by Coomassie blue staining. (C) CK2 pull-down assay. Cell lysates (≈2 mg) were prepared from TIG-3 (Upper) and SW480 (Lower) and incubated with GST-CK2α, GST-CK2α′, GST-CK2β, a 1:1 mixture of GST-CK2α and GST-CK2β, or GST alone (≈1 μg), each immobilized on glutathione-Sepharose beads. After washing thoroughly, binding proteins were solubilized and separated by SDS/5% PAGE, followed by immunoblotting with anti-APC mAb.

Figure 3

Association with CK2 and phosphorylation of APC during the cell cycle. (A) Cell cycle-dependent association of CK2α with APC in vivo. Cell lysates were prepared from TIG-3 and SW480 cells synchronized in G₀, G₁/S, or G₂/M phase of the cell cycle and immunoprecipitated with anti-APC polyclonal antibodies. Proteins were solubilized and separated by SDS/12% PAGE for α or by 5% gels for APC and blotted with anti-CK2α or anti-APC antibodies. Total lysates were analyzed in parallel for the content of CK2α. (B) Phosphorylation of full-length APC from G₂/M-arrested cells by CK2. Anti-APC immunoprecipitates prepared from synchronized TIG-3 cells were combined with (+) or without (−) recombinant CK2α and CK2β to analyze in vitro phosphorylation activity. Samples were separated by SDS/7.5% PAGE and autoradiographed.

Figure 4

Inhibitory effect of APC on CK2 activity. The kinase activities of GST-CK2α (≈0.5 μg) and a 1:1 mixture of GST-CK2α (≈0.5 μg) and GST-CK2β (≈0.5 μg) were measured against substrate peptide in the presence or absence of anti-APC or human IgG immunoprecipitates from TIG-3 or SW480 cells arrested at G₂/M with nocodazole. Results were expressed as mean ± SD of three independent experiments.

Figure 5

C terminus of APC is responsible for inhibition of CK2 activity. (A) APC fragments used in this study with a schematic structure of APC are shown. APC contains various domains including heptad repeat, armadillo repeat (ARD), 20-aa repeat, SAMP, and basic domain. The portion of APC fragments used in this study was indicated as bars with amino acid number. (B) Inhibition of CK2 activity by C-terminal fragment of APC. The kinase activity of GST-CK2α (≈0.5 μg) and a 1:1 mixture of GST-CK2α and GST-CK2β (≈0.5 μg each) were measured against substrate peptide in the presence or absence of recombinant APC fragments (0.1 or 0.5 μg) as indicated. Inhibitory effect of each fragment was summarized in A. (C) Dose-dependent inhibition of CK2 activity by APC. The kinase activity of a 1:1 mixture of GST-CK2α and GST-CK2β (≈0.5 μg each) was measured against substrate peptide in the presence of various concentrations of recombinant APC fragments as indicated. Results were expressed as means ± SD of three independent experiments.

Figure 6

Effect of APC fragments on growth and transformation properties. (A) 293 cells (2 × 10⁴) constitutively expressing an APC fragment 2086–2394 (APC-C) in pEGFP vector (○) or in pCDNA3-FLAG (♦) vs. wild-type cells expressing empty pEGFP (■) were split to 35-mm plates, and cell counts were monitored at each time point. (B) Triplicate plates of SW480 cells (1 × 10⁶) were transfected with 0.3 μg of either pEGFP-APC-C (column 2), pCDNA3-FLAG-APC-C (column 3), or empty pEGFP (column 1), plated in soft agar, and assayed for colony formation after 3 weeks. Results were expressed as means ± SD of three independent experiments.