Sensitization of Tumor Cells to Apo2 Ligand/TRAIL-induced Apoptosis by Inhibition of Casein Kinase II¹

Genetic aberrations that render cells incapable of executing apoptosis not only promote tumorigenesis, but also underlie the observed resistance of human cancers to anticancer agents. Unraveling mechanisms to unleash the apoptotic program in tumor cells could aid the design of effective therapeutic interventions against resistant cancers. Tumor-cell death can be triggered by engagement of specific death receptors belonging to the TNF-receptor³ gene superfamily with the “death ligand,” Apo2L/TRAIL (1, 2). TRAIL/Apo2L induces apoptosis of many cancer cell lines in vitro, and its tumoricidal activity and safety in vivo has been confirmed in preclinical animal models of human cancer xenografts (2). Although most human cancer cell lines express death receptors for Apo2L/TRAIL, many remain resistant to TRAIL/Apo2L-induced death (2). The identification of key survival signals responsible for protecting tumor cells from death-receptor-induced apoptosis could aid the design of Apo2L/TRAIL-based combination regimens for treatment of such cancers.

The death receptors for TRAIL/Apo2L, TRAIL-R1 (DR4), and TRAIL-R2 (DR5, KILLER) are type I transmembrane proteins containing cytoplasmic sequences, termed “death domains,” that recruit and cross-activate the initiator procaspase-8 (1). Caspase-8 cleaves and activates BID, a “BH-3 domain only” prodeath member of the Bcl-2 family (3, 4). The active truncated form of BID (tBID) triggers the mitochondrial activation of caspase-9 by inducing the homooligomerization and allosteric activation of BAK or BAX, two multidomain proapoptotic members of the Bcl-2 family (5). However, death-receptor-induced activation of caspase-9 is inhibited by Bcl-x_L, an NF-κB-inducible Bcl-2 family member that sequesters tBID and curtails its ability to activate BAX (6, 7, 8). Therefore, molecules that govern the relative balance between tBID and Bcl-x_L may be key determinant(s) of the sensitivity of tumor cells to Apo2L/TRAIL-induced death.

Protein kinase CK2 is an evolutionarily conserved holoenzyme composed of two catalytic α (and/or a’) subunits and two regulatory β subunits (9). CK2 is increased in response to diverse growth stimuli, and its activity is aberrantly elevated in diverse tumor types including breast carcinomas, colorectal carcinomas, squamous cell carcinomas and adenocarcinomas of the lung, squamous cell carcinomas of the head and neck, prostate carcinomas, ovarian carcinomas, and melanomas (9). Adult transgenic mice expressing CK2α in lymphocytes display a stochastic propensity to develop lymphoma (10), and overexpression of CK2α in the mammary gland results in breast tumors (11). In addition to its demonstrated role in oncogenesis, we now report that CK2 is at the nexus of two survival signaling pathways that protect tumor cells from Apo2L/TRAIL-induced apoptosis. We find that CK2 inhibits Apo2L/TRAIL-induced caspase-8-mediated cleavage of BID. In addition to preventing the formation of tBID, our studies demonstrate that CK2 promotes NF-κB-mediated expression of Bcl-x_L. We show that cancer cells with constitutive activation of CK2 exhibit a high Bcl-x_L/tBID ratio and fail to activate caspase-9 or undergo apoptosis in response to Apo2L/TRAIL. Conversely, reduction of the Bcl-x_L/tBID ratio by inhibition of CK2 with DRB (12), the natural plant flavonoid, apigenin (13), or the anthraquinone derivative, emodin (14), renders such cancer cells sensitive to Apo2L/TRAIL-induced activation of caspase-9 and apoptosis. Using isogenic cancer cell lines that differ only in the presence or absence of either the p53 tumor suppressor gene or the BAX gene (15, 16), we further demonstrate that TRAIL/Apo2L-induced death of p53^+/+- or p53^−/−- BAX-proficient, but not BAX-deficient, cancer cells is augmented by inhibition of CK2. These observations indicate that CK2 inhibitors can augment Apo2L/TRAIL-induced tumor-cell death by elevating the tBID/Bcl-x_L ratio and promoting the activation of BAX.

The HCT116 human colon adenocarcinoma cell line containing wild-type p53 and one intact BAX allele (p53^+/+BAX^+/−) and isogenic p53-deficient (p53^−/−) or BAX-deficient (BAX^−/−) derivatives of HCT116 cells generated by disruption of either p53 or BAX alleles by gene targeting have been described previously (15, 16). HCT116 cells of each genotype (p53^+/+,BAX^+/−; p53^−/−; BAX^−/−) were provided by Dr. Bert Vogelstein (Johns Hopkins University School of Medicine, Baltimore, MD). Isogenic HCT116 cells were cultured in McCoy’s 5A medium supplemented with 10% FCS, penicillin (100 units/ml), and streptomycin (100 μg/ml). The Hs578 and SKBr-3 human breast cancer cell lines were cultured in DMEM (with 1 μg/ml insulin) and McCoy’s 5A medium, respectively, supplemented with 10% FCS, penicillin (100 units/ml), and streptomycin (100 μg/ml).

Exponentially growing cells (2 × 10⁵/well in 6-well plates) were incubated with soluble recombinant human TRAIL/Apo2L (100 ng ml⁻¹) plus enhancer antibody (2 μg ml⁻¹; Alexis, San Diego, CA) for 48 h at 37°C.

Cell lysates were prepared as described (7), 50–100 μg of protein were resolved by SDS-PAGE, transferred onto Immobilon-P PVDF membrane (Millipore, Bedford, MA), and probed with antibodies against caspase-8 (C-20), BID (C-20), BAX (N-20), caspase-9 (H-170), Bcl-x_L (S-18), FLIP (G-11), CK2α (C-18), and actin (C-11) from Santa Cruz Biotechnology (Santa Cruz, CA). Immunoreactive protein complexes were visualized with enhanced chemiluminescence (Amersham, Arlington Heights, IL).

The phosphotransferase activity of CK2 was measured using a Casein Kinase-2 Assay Kit (Upstate Biotechnology, Lake Placid, NY). This assay is based on phosphorylation of a CK2-specific peptide substrate using the transfer of the γ-phosphate of [γ-³²P]ATP by CK-2 kinase. The phosphorylated substrate was separated from the residual [γ-³²P]ATP using P81 phosphocellulose paper, and [³²P] incorporation into the substrate was measured using a scintillation counter and expressed as the calculated pmol phosphate incorporated into CK2 substrate peptide/min/200 ng of CK-2.

CK2 complexes were immunoprecipitated from whole-cell extracts (500 μg) using an antibody against CK2α (C-18; Santa Cruz Biotechnology) and subjected to a CK2 kinase assay at 30°C for 30 min in kinase buffer {100 mm Tris (pH 8.0), 100 mm NaCl, 50 mm KCl, 10 mm MgCl₂, 10 μCi [γ-³²P]GTP, 100 μm Na₃VO₄, 2 μm phenylmethylsulfonyl fluoride, 10 μg/ml aprotinin, 1 μg/ml leupeptin, 1 μg/ml pepstatin} containing 200 ng of GST-IκBα fusion protein (GST-IκBα; Santa Cruz Biotechnology) as substrate (17). The specificity of the kinase reaction was confirmed by the addition of cold CK2-specific peptide substrate H-Arg-Arg-Ala-Asp-Asp-Ser-Asp-Asp-Asp-Asp-Asp-OH (Calbiochem, San Diego, CA). Recombinant CK2 (New England Biolabs) was used as a positive control. The kinase reaction was terminated by the addition of 2× SDS-PAGE sample buffer and subjected to SDS-PAGE and autoradiography.

Inhibition of CK2 was achieved by incubation of cells in graded concentrations of DRB(10 or 40 μm; Calbiochem; Ref 12), apigenin (10 or 20 μm; Sigma-Aldrich, St. Louis, MO; Ref. 13), or emodin (5, 10, or 20 μg/ml; Sigma-Aldrich; Ref. 14).

IKK complexes were immunoprecipitated from whole-cell extracts (500 μg) using an antibody against IKKβ (M-280; Santa Cruz Biotechnology) and subjected to a kinase assay at 30°C for 30 min in kinase buffer [20 mm HEPES (pH 7.6), 3 mm MgCl₂, 10 μm ATP, 3 μCi [γ-³²P]ATP, 10 mm β-glycerophosphate, 10 mm NaF, 10 mmp-nitrophenyl phosphate, 300 μm Na₃VO₄, 1 mm benzamidine, 2 μm phenylmethylsulfonyl fluoride, 10 μg/ml aprotinin, 1 μg/ml leupeptin, 1 μg/ml pepstatin, and 1 mm DTT] containing 500 ng of glutathione S-transferase (GST)-IκBα fusion protein (GST-IκBα; Santa Cruz Biotechnology) as substrate (18). The kinase reaction was terminated by the addition of 2× SDS-PAGE sample buffer and subjected to SDS-PAGE and autoradiography. Inhibition of IKK was achieved by incubation of cells with either ASA (3 mm) or sulindac sulfide (120 μm) from Biomol Research Laboratories, Inc. (Plymouth Meeting, PA).

Nuclear extracts were prepared as described (7). Double-stranded oligonucleotides containing a consensus binding site for NF-κB (5′-GGGGACTTTCCC-3′; Santa Cruz Biotechnology) were 5′-end labeled using polynucleotide kinase and [γ-³²P] dATP. Nuclear extracts (2.5–5 μg) were incubated with ∼1 μl of labeled oligonucleotide (20,000 cpm) in 20 μl incubation buffer [10 mm Tris, 40 mm NaCl, 1 mm EDTA, 1 mm β-mercaptoethanol, 2% glycerol, and 1–2 μg of poly(deoxyinosinic-deoxycytidylic acid] for 20 min at 25°C. The specificity of NF-κB DNA-binding activity was confirmed by competition with excess cold wild-type or mutant oligonucleotide or supershift with an antibody against p65/RelA (Geneka, Montreal, Canada) as described (7). DNA-protein complexes were resolved by electrophoresis in 5% nondenaturing polyacrylamide gels and analyzed by autoradiography and densitometry (Molecular Dynamics).

The bcl-x-CAT reporter containing the human bcl-x promoter region cloned in a promoterless vector expressing CAT reporter gene (pCAT-basic) and the κB site mutant bcl-x-CAT plasmid with an inactivated NF-κB motif at position −232 in the bcl-x promoter (TTTACTGCCC; −298/+22 mκB) have been described previously (provided by Dr. Céline Gelinas, University of Medicine and Dentistry of New Jersey, NJ; Ref. 6). Hs578 cells were transiently cotransfected with 1 μg of either bcl-x-CAT reporter plasmid (−298/+22) or the κB site mutant (mt) plasmid (−298/+22 mκB) together with a β-galactosidase (CMV-β-gal) reporter using lipofectin (InVitrogen). Transfected cells were incubated at 37°C for 24 h in DMEM medium (supplemented with insulin) in the absence or presence of DRB (10 or 40 μm), apigenin (10 or 20 μm), or emodin (5 or 10 μg/ml). Cells were then harvested and assayed for CAT activity (normalized to β-gal activity) as described (6).

Cells were assessed for morphological features of apoptosis (condensed chromatin and micronucleation) by microscopic visualization. Cell viability was assessed at the indicated intervals by trypan blue dye exclusion of harvested cells (adherent + floating in the medium). Cell survival was measured by scoring at least 500 cells in each group, and the average percentage of viability (mean ± SE) was calculated from three different experiments.

CK2 is constitutively activated in Hs578 breast cancer cells (Fig. 1,A; Ref. 17). Exposure of Hs578 cells to the classic CK2-specific inhibitor, DRB (10 or 40 μm), resulted in dose-dependent inhibition of phosphorylation of a CK2-specific peptide substrate (RRADDSDDDDD; Fig. 1,A). Dose-dependent inhibition of CK2 activity in Hs578 cells was also achieved by treatment with the plant flavonoid, apigenin (10 or 20 μm), or the plant anthraquinone derivative, emodin (10 or 20 μg/ml), but not by the nonsteroidal anti-inflammatory drugs, ASA or sulindac sulfide (Fig. 1 A).

CK2 phosphorylates the PEST domain of IκBα (19). To analyze CK2-dependent phosphorylation of IκBα in Hs578 cells, CK2α was immunoprecipitated from whole-cell lysates and subjected to an in vitro CK2 kinase assay using a GST-IκBα fusion protein as substrate. Kinase assays demonstrated strong constitutive phosphorylation of GST-IκBα that was reduced by competition with a CK2-specific peptide substrate (Fig. 1,B). The specific involvement of CK2 in the high basal IκBα-kinase activity was also confirmed by treatment of cells with the CK2-inhibitors, DRB, apigenin, or emodin. Incubation of Hs578 cells with DRB (40 μm), apigenin (10, 20 μm), or emodin (5, 10 μg/ml) led to a dose-dependent inhibition of GST-IκBα phosphorylation by CK2 immune complexes, but did not influence the phosphorylation of GST-IκBα by IKKβ (Fig. 1,B). Conversely, treatment of Hs578 cells with the nonsteroidal anti-inflammatory drugs, ASA, or sulindac sulfide inhibited IKKβ-mediated GST-IκBα phosphorylation, but did not interfere with CK2-dependent phosphorylation of GST-IκBα (Fig. 1 B).

In addition to promoting phosphorylation-induced degradation of IκBα, CK2 activates NF-κB transcriptional activity by phosphorylating RelA/p65 (20). To evaluate whether the constitutive activation of CK2 promotes NF-κB activity, nuclear extracts of Hs578 cells incubated for 16 h with graded concentrations of DRB, apigenin, or emodin were subjected to electrophoretic mobility shift assays for NF-κB DNA-binding activity. Treatment with DRB, apigenin, or emodin, led to inhibition of NF-κB DNA-binding activity (Fig. 1 C).

The human bcl-x promoter contains a κB DNA site (TTTACTGCCC; 298/+22) responsible for its Rel-dependent induction (6). To determine whether CK2 promotes NF-κB-dependent transcriptional activation of Bcl-x_L, Hs578 cells were transiently cotransfected with either the bcl-x-CAT reporter plasmid (−298/+22) or the κB site mutant (mt) plasmid (−298/+22 mκB) and incubated in insulin-supplemented medium in the absence or presence of DRB, apigenin, or emodin. Inhibition of CK2 with DRB, apigenin, or emodin led to a dose-dependent loss of NF-κB-dependent transcriptional activation of the bcl-x-CAT reporter in Hs578 cells (Fig. 1,D). Consistent with the decline of NF-κB-dependent bcl-x transcription by inhibition of CK2, exposure of Hs578 cells to DRB, apigenin, or emodin led to a dose-dependent reduction of Bcl-x_L protein levels in immunoblot analysis (Fig. 1,E). Akin to Bcl-x_L, CK2 inhibitors also reduced expression of c-FLIP, an NF-κB-inducible protein that prevents death-receptor-induced activation of the initiator pro-caspase-8 (Fig. 1 E; Ref 21). These data indicate that CK2 activation promotes NF-κB-mediated expression of Bcl-x_L and FLIP in tumor cells.

Engagement of death receptors for TRAIL/Apo2L leads to recruitment and cross-activation of the initiator caspase-8, which in turn, cleaves and activates BID (3, 4). The active truncated form of BID (tBID) triggers the mitochondrial activation of caspase-9 via activation of BAK or BAX (5, 18, 22). CK2 promotes NF-κB-mediated expression of the caspase-8-inhibitor, c-FLIP (Fig. 1,E), and phosphorylation of BID by CK2 has been reported to render BID resistant to caspase-8 mediated cleavage (23). In addition, CK2 promotes expression of Bcl-x_L (Fig. 1 E), which in turn, sequesters tBID and prevents tBID-induced activation of caspase-9 (8).

To assess whether the constitutive activation of CK2 in tumor cells plays a role in regulating Apo2L/TRAIL-induced caspase-8-mediated cleavage of BID and activation of caspase-9, Hs578 cells were treated with Apo2L/TRAIL in the absence or presence of DRB, apigenin, or emodin. Immunoblot analyses showed that treatment with any of these CK2 inhibitors facilitated Apo2L/TRAIL-induced cleavage of procaspase-8, caspase-8-mediated truncation of BID, and activation of caspase-9 (Fig, 1F). These data indicate that elevation of the Bcl-x_L/tBID ratio by constitutive activation of CK2 in tumor cells inhibits the activation of caspase-9 in response to Apo2L/TRAIL. Conversely, reduction of the Bcl-x_L/tBID ratio by CK2 inhibitors facilitates Apo2L/TRAIL-induced activation of caspase-9 in tumor cells.

Hs578 cells were markedly resistant to Apo2L/TRAIL, with death of only 9 ± 3% death over 48 h. To determine whether inhibition of CK2 could sensitize cancer cells to Apo2L/TRAIL-induced death, Hs578 cells were treated for 48 h with graded concentrations of DRB, apigenin, or emodin in the presence or absence of Apo2L/TRAIL (100 ng/ml). DRB, apigenin, or emodin alone induced only limited cell death at the maximum concentrations used (Fig. 2, A and B). However, treatment with any of these CK2 inhibitors led to a dose-dependent increase in Apo2L/TRAIL-induced death of Hs578 cells (Fig. 2, A and B). Similarly, inhibition of CK2 with DRB, apigenin, or emodin also led to dose-dependent sensitization of HER2/neu-overexpressing SKBr-3 breast cancer cells to Apo2L/TRAIL-induced apoptosis (Fig. 2, A and C).

The active truncated form of BID (tBID) triggers the mitochondrial activation of caspase-9 in tumor cells by inducing the homooligomerization and allosteric activation of BAX (18, 22). The sequestration of tBID by Bcl-x_L counteracts the mitochondrial activation of caspase-9 (8, 18). By increasing the ratio of tBID to Bcl-x_L, CK2 inhibitors may facilitate Apo2L/TRAIL-induced activation of caspase-9 via BAX. To determine whether CK2 inhibitors augment Apo2L/TRAIL-induced apoptosis of tumor cells via activation of BAX, we studied isogenic derivatives of HCT116 colon cancer cells that differ only in the presence or absence of the BAX gene (16). Ninety-four percent of HCT116 cells have an intact BAX allele (BAX^+/−) and express functional BAX protein. BAX-deficient HCT116 cells (BAX^−/−) were generated by targeted inactivation of the wild-type BAX allele in a BAX heterozygote (16). BAX^+/− or BAX^−/− HCT116 cells were treated for 48 h with graded concentrations of either apigenin (10, 20 μm) or emodin (10, 20 μg/ml) in the presence or absence of Apo2L/TRAIL (100 ng/ml). Inhibition of CK2 with either apigenin or emodin led to a dose-dependent augmentation of TRAIL/Apo2L-induced death of BAX-proficient HCT116 cells (BAX^+/−; Fig. 3, A–C). In contrast, BAX-deficient HCT116 cells (BAX^−/−) remained resistant to TRAIL/Apo2L-induced death even in the presence of the highest tested concentrations of either apigenin or emodin (Fig. 3, A–C).

CK2-mediated phosphorylation of p53 has been reported to modulate p53-dependent transcription (24). CK2α transgenic mice that are deficient in p53 develop thymic lymphomas at a markedly accelerated rate when compared with p53-deficient mice lacking the transgene (25). To determine whether the augmentation of Apo2L/TRAIL-induced apoptosis by CK2 inhibitors requires p53, we studied isogenic derivatives of HCT116 colorectal cancer cells that differ only in the presence or absence of the p53 gene. HCT116 cells have wild-type p53 (p53^+/+) and express intact functional p53 protein (15). Isogenic p53-deficient derivatives of HCT116 cells were generated by targeted inactivation of both p53 alleles (p53^−/−; Ref. 15). p53^+/+ or p53^−/− HCT116 cells were treated for 48 h with graded concentrations of either apigenin (10, 20 μm) or emodin (10, 20 μg/ml) in the presence or absence of Apo2L/TRAIL (100 ng/ml). In contrast to isogenic BAX^−/− HCT116 cells, p53^−/− and p53^+/+ HCT116 cells exhibited equivalent sensitivity to induction of apoptosis by either TRAIL/Apo2L or each of the CK2 inhibitors (Fig. 3, A–C). Exposure to either apigenin or emodin augmented the sensitivity of both p53^−/− and p53^+/+ HCT116 cells to TRAIL/Apo2L in a dose-dependent manner (Fig. 3, A–C). These data indicate that the enhancement of Apo2L/TRAIL-induced tumor-cell death by CK2 inhibitors is independent of p53.

Antibody or ligand-mediated engagement of death receptors for Apo2L/TRAIL offers an attractive strategy for inducing apoptosis of tumor cells. However, cancer cell lines exhibit a wide heterogeneity in their sensitivity to Apo2L/TRAIL-induced apoptosis, and several tumor cell lines remain resistant to Apo2L/TRAIL, even though they express death receptors, TRAIL-R1/DR4, and TRAIL-R2/DR5 (2). The efficacy of Apo2L/TRAIL against such cancers may be improved by inhibition of the critical molecule(s) that interfere with Apo2L/TRAIL-induced death signaling. Apo2L/TRAIL-induced tumor-cell death involves caspase-8-mediated cleavage of BID, tBID-mediated activation of BAX, and BAX-mediated mitochondrial activation of caspase-9 (18, 22). The results of our study indicate that CK2 is at the nexus of two survival signaling pathways that play an instrumental role in protecting tumor cells from Apo2L/TRAIL-induced apoptosis. We find that constitutive activation of CK2 in tumor cells inhibits caspase-8-mediated formation of tBID in response to Apo2L/TRAIL. In addition, CK2 promotes NF-κB-mediated expression of Bcl-x_L, which in turn, sequesters tBID and curtails its ability to activate BAX. Consistent with the high expression of Bcl-x_L and reduced formation of tBID, we find that tumor cells with constitutively high CK2 activity fail to activate caspase-9 and remain resistant to apoptosis in response to Apo2L/TRAIL. Conversely, reduction of the Bcl-x_L/tBID ratio by CK2 inhibitors sensitizes such cancers to Apo2L/TRAIL-induced activation of caspase-9 via activation of BAX. We further demonstrate that the enhancement of Apo2L/TRAIL-induced tumor-cell death by CK2 inhibitors requires BAX, but not p53.

CK2 is active in rapidly proliferating tissues, and its activity is increased in response to diverse growth stimuli including insulin, insulin-like growth factor-1, epidermal growth factor, and androgens (9). A variety of human cancers and transformed cell lines exhibit constitutively high CK2 activity (9). Overexpression of the catalytic α subunit of CK2 in transgenic mice leads to T-cell lymphoma (10), and CK2α overexpression accelerates lymphomagenesis caused by loss of p53 (25). Although these observations have identified a role of CK2 in cell growth and neoplastic transformation, our findings suggest that CK2 also plays an instrumental role in protecting cancer cells from Apo2L/TRAIL-induced apoptosis. This mechanism of resistance may frequently operate in diverse tumor types with constitutive activation of CK2 via genetic aberrations, such as overexpression of receptor tyrosine kinases (HER2/neu, epidermal growth factor receptor, insulin-like growth factor-1 receptor). Our results indicate that such resistant cancers can be rendered sensitive to Apo2L/TRAIL-induced death by CK2 inhibitors. DRB, apigenin, and emodin competitively inhibit CK2 catalytic activity by directly interacting with the nucleotide-binding sites of subunits of CK2 (12, 13, 14). Apigenin is a plant flavonoid that is a major constituent of herbal chamomile and is found naturally in many fruits, vegetables, and plant-derived beverages (chamomile tea and wine). Emodin is a plant anthraquinone derivative isolated from Rheum Palmatum. Although our results provide a biological rationale for combining TRAIL/Apo2L with CK2 inhibitors such as apigenin or emodin for treatment of cancers, further studies are required to evaluate and optimize the therapeutic ratio of such regimens. In addition to defining CK2 as a key molecular determinant of the resistance of cancers to TRAIL/Apo2L-induced apoptosis, our findings may aid the development and genotype-specific application of TRAIL/Apo2L-based combinatorial regimens for the treatment of diverse human cancers.

We thank Dr. Bert Vogelstein for his kind gift of isogenic HCT116 cells and Dr. Céline Gelinas for kindly providing the bcl-x-CAT reporter plasmids.