Activation of p53 tumor suppressor induces either cell cycle arrest or apoptosis through transcription-dependent and independent pathways; however, their relative roles in apoptosis induction and how these pathways are regulated remains elusive. Here, we report a unique role for glycogen synthesis kinase-3β (GSK-3β) in regulating p53 functions in human colorectal cancer cells. Pharmacologic modulation of GSK-3β markedly impaired p53-dependent transactivation of targets including p21 and Puma but promoted p53-dependent conformational activation of Bax, resulting in cytochrome c release, loss of mitochondrial membrane potential, and caspase-9 processing. Thus, p53-mediated damage response is converted from cell cycle arrest to apoptosis following exposure to a variety of chemotherapeutic agents. We found that this effect is associated with the modulation of inhibitory Ser9 phosphorylation of GSK-3β but not with the activating tyrosine phosphorylation. We further show that the induction of apoptosis is through a direct mitochondrial pathway that requires Bax but not Puma. Our results underscore the importance of transcription-independent mechanism in p53-induced apoptosis and indicate that GSK-3β plays distinct dual roles in regulating p53 pathways: promoting p53 transcriptional activity in the nucleus but suppressing p53-mediated direct apoptotic function at the mitochondria. Importantly, our data suggest that small-molecule inhibition of GSK-3β might represent a novel approach for modulating chemotherapy.

Activation of the p53 pathway by a diverse range of noxious stimuli results in either cell cycle arrest or apoptosis. The process requires an accumulation of p53 protein, which then mediates its physiologic effects by transcriptional activation of target genes (14). Current evidence suggests that cells undergo cell cycle arrest through transcriptional activation of cyclin-dependent kinase inhibitor p21 (59) or apoptosis through induction of proapoptotic genes such as Puma, Noxa, and Bax (4, 10, 11). The precise mechanisms governing the cell fate to p53 response is poorly understood, although it has been proposed that the p53 response ultimately depends on a balance between transcription of p53 targets that favor growth arrest versus apoptosis. Most chemotherapeutic agents induce cell death through activation of the p53 pathway. Therefore, mutations in the p53 gene result in an evasion of apoptotic pathways, leading to more aggressive phenotypes and resistance to chemotherapy. The aim of p53-based cancer pharmacology is to translate our understanding of p53 biology to viable therapeutic strategies and hence to promote apoptosis in cancer cells. Genetic approaches that manipulate the level of p53 target gene (12) or p53 transcription cofactor (13) have been used in different systems to modulate the fine balance in expression levels of p53 targets, consequently promoting p53-dependent apoptosis. Whereas these experiments have been important in understanding the cellular control of p53-induced cell fate, therapeutic outcomes have yet to arise from these observations.

Recent studies have indicated that p53 can directly induce apoptosis at the mitochondria. p53 has been shown to directly activate proapoptotic Bcl-2 family members Bax or Bak (12, 14, 15) and forms inhibitory complexes with antiapoptotic bcl-XL and Bcl2 proteins (15). These interactions eventually result in mitochondria permeabilization and cytochrome c release. This alternative p53 death pathway correlates with an early phase of apoptosis, which occurs independently of p53 transcriptional activity (16), and can be modulated by direct pharmacologic activator of p53 (16, 17). For the moment, little is known about the regulation of this pathway. Nonetheless, pharmacologic activation of this p53-dependent, transcription-independent apoptotic pathway might provide a feasible strategy for cancer therapeutics.

Glycogen synthase-3 kinase (GSK-3) is a constitutively expressed serine-tyrosine kinase that has two isoforms (α and β) and is involved in diverse signaling pathways (18). Initially described as key enzymes involved in glycogen metabolism, they are now known to regulate a diverse array of cell functions, including a proapoptotic role for GSK-3β in neurons and other tissues (1921). Isolated studies have identified a specific role for GSK-3β in the p53 response: promoting p53-dependent apoptosis in neuroblastoma cells (19) and regulating endoplasmic reticulum stress-mediated p53 activation (22). Currently, the role of specific GSK-3β inhibitors is being explored in the treatment of diabetes and neurodegenerative diseases (23).

Herein, we identify GSK-3β as a unique regulator of p53 activity and show that GSK-3β participates in regulating p53 activity in both the nucleus and mitochondria. In contrast to its proapoptotic role in other tissues, we found that pharmacologic modulation of GSK-3β dramatically promoted p53-induced apoptosis in human colorectal cancer cells. Apoptosis occurred even with concurrent impairment of p53 transcriptional activity and was mediated through direct mitochondrial p53 activity, which requires Bax but not Puma. These findings reveal a novel mechanism used by cells to regulate p53 activity in colorectal cancer cells and provide a novel approach that can be used to modulate chemotherapy.

Cell culture and drug treatments. Human colon cancer cell line HCT116 and its derived isogenic p53−/−, Puma−/−, Bax+/−, and p53-inducible DLD cell lines were kindly provided by Dr. Bert Vogelstein (Johns Hopkins University, MD). RKO and RKO/E6 cells were from American Type Culture Collection (Manassas, VA). Cells were cultured in DMEM containing 10% fetal bovine serum. All culture reagents and media were from Invitrogen (Carlsbad, CA). 5-Fluorouracil (5-FU), Adriamycin, etoposide (VP-16), lithium chloride was purchased from Sigma (St. Louis, MO). GSK-3β-specific small-molecule inhibitor LY2119301 was a gift from Eli Lilly (Indianapolis, IN) and SB-216763 and SB-415286 were from BIOMOL Research Laboratories, Inc.(Plymouth Meeting, PA).

Western blotting, immunoprecipitation, and microarray analysis. Cells were scraped, collected, and lysed as described previously (24). Protein samples (50 μg) were separated by SDS-PAGE and transferred onto Immobilon membranes (Millipore, Bedford, MA). Antibodies against the following proteins are used: p53 (DO-1), p21, poly(ADP-ribose) polymerase (PARP), GSK-α/β, α-tubulin, and β-actin (Santa Cruz Biotechnology, Santa Cruz, CA); Puma (Ab-1, Oncogene, Uniondale, NY); Bax (Sigma), phosphor-(Ser9/21)-GSK3α/β, and caspase-9 (Cell Signaling Technologies, Beverly, MA); phosphor-(Tyr216/279)-GSK3α/β (Upstate Biotechnologies, Lake Placid, NY); and cytochrome c (BD PharMingen, San Diego, CA). To detect Bax conformational change, cells were lysed in 1% CHAPS buffer and the soluble fraction was immunoprecipitated with anti-Bax monoclonal antibody (6A7, BD PharMingen) and followed by immunobloting with anti-Bax polyclonal antibody. Microarray analysis was done as described (11).

Fluorescence-activated cell sorting analysis of DNA content and caspase-3 activity. Cells were harvested and fixed in 70% ethanol. Fixed cells were stained with propidium iodide (50 μg/mL) after treatment with RNase (100 μg/mL). The stained cells were analyzed for DNA content by fluorescence-activated cell sorting (FACS) in a FACSCalibur (Becton Dickinson Instrument, San Jose, CA). Cell cycle fractions were quantified using the CellQuest software (Becton Dickinson). To measure caspase-3 activity, cells were fixed with Cytofix/Cytoperm solution (BD PharMingen) according to the manufacturer's instructions and then stained with FITC-conjugated rabbit anti-active caspase-3 monoclonal antibody (BD PharMingen). Quantification of cells positive for the caspase-3 detection was done by flow cytometry.

RNA interference.SMARTpool GSK-3β and GSK-3α small interfering RNAs (siRNA) and negative control siRNA were purchased from Dharmacon, Inc. (Lafayette, CO); HCT116 cells were transfected with LipofectAMINE 2000 reagent (Invitrogen) according to manufacturer's instructions in the presence of siRNAs.

Adenoviral infections. Adenovirus Ad-LacZ and Ad-p53 were obtained from Dr. Shibin Zhou (Johns Hopkins University). Cells were grown to 50% confluence and infected with recombinant adenovirus as described (25). Twenty-four hours after the infection, cells were treated with drugs for indicated times.

Subcellular fractionation. Cells were lysed in mitochondria lysis buffer [210 mmol/L mannitol, 70 mmol/L sucrose, 10 mmol/L HEPES (pH 7.4), 1 mmol/L EDTA, and protease inhibitor cocktail] with a Dounce homogenizer and subjected to centrifugation at 1,000 × g to pellet nuclei. Post-nuclear supernatant was centrifuged at 10,000 × g to pellet mitochondria-enriched heavy membrane fraction, and the resulting supernatant was further centrifuged to obtain cytosolic fraction. Total proteins (30 μg) from cytosolic or mitochondrial fractions were subjected to immunoblot analysis.

Measurement of mitochondrial permeability transition. Cells were harvested by centrifugation at 300 × g at 4°C for 5 minutes. Mitochondrial permeability transition was determined by staining the cells with JC-1 (BD PharMingen) according to the manufacturer's instructions. Mitochondrial permeability transition was quantified by flow cytometric determination of cells with decreased red fluorescence (i.e., with mitochondria displaying a lower membrane potential, ΔΨm). Data are given in percentage of cells with low ΔΨm.

Small-molecule GSK-3β inhibitor promotes genotoxic agent-induced apoptosis in human colorectal cancer cells in a p53-dependent manner. In human colorectal cancer HCT116 cells, p53 activation by DNA-damaging agent Adriamycin and DNA analogue 5-FU results in cell cycle arrest and apoptosis, respectively (26). To investigate the cellular factors that modulate p53 response, we screened a series of protein kinase inhibitors and discovered that a specific GSK-3β small-molecule inhibitor (LY2119301) efficiently converted Adriamycin-induced p53 response from growth arrest to cell death. Figure 1A shows that HCT116 cells treated with Adriamycin for 24 hours underwent cell cycle arrest without obvious induction of cell death. Cotreatment of cells with Adriamycin and LY2119301 resulted in a marked increase in the sub-G1 population (∼59%) compared with cells treated with Adriamycin or LY2119301 alone (∼5-8%). In contrast, the synergistic induction of cell death was not seen in p53 knockout HCT116 cells (Fig. 1A) under the same condition, suggesting that the cell death is p53 dependent. The cell death was attributable to apoptosis, because similar results were obtained by using caspase-3 activity assay (Fig. 1B). LY2119301 also triggered a p53-dependent apoptosis in response to other DNA-damaging agents including VP-16 and camptothecin (Fig. 1C). Moreover, LY2119301 greatly sensitized 5-FU-indcued apoptosis in a strictly p53-dependent manner (Fig. 1D). Remarkably, dose-effect experiments showed that LY2119301 induced a 10-fold increase in apoptosis induction by 5-FU, which is one of the most commonly used chemotherapeutic agents in the treatment of colorectal cancer. Thus, pharmacologic inhibition of GSK-3β not only converted p53-mediated damage response from growth arrest to apoptosis but also further sensitized p53-dependent apoptotic response. The apoptotic effect of LY2119301 on p53 response was also observed in another p53 wild-type colorectal cancer cell line RKO but not in its p53-inactive counterpart RKO/E6 cells (Fig. 1E). Similarly, apoptosis was not induced in other p53 mutant colon cancer cell lines SW480 and HT29 (data not shown). These findings suggest that GSK-3β inhibitor LY2119301 promotes chemotherapeutic agent–induced apoptosis in a manner that strictly requires intact p53.

Figure 1.

GSK-3β inhibitor LY2119301 promotes p53-dependent apoptosis in response to a variety of genotoxic agents. A, flow cytometry profiles showing cell cycle distribution of HCT116 and HCT116 p53−/− cells treated with LY2119301 (5 μmol/L), Adriamycin (ADR, 1 μmol/L), or both for 24 hours. The percentage of cells in sub-G1 is given. FLA-2, DNA content. B, flow cytometry profiles of HCT116 cells treated as in (A) and stained for active anti-caspase-3. Percentages of cells positive for active caspase-3 are indicated. C, flow cytometry profiles showing cell cycle distribution of HCT116 and HCT116 p53−/− cells treated with etoposide (10 μmol/L) or camptothecin (CPT, 100 nmol/L) for 24 hours, with or without LY2119301. The percentage of apoptotic cells (sub-G1) is given. D, dose response of HCT116 cells to varying 5-FU concentrations, in the presence or absence of LY2119301. Apoptotic fractions determined by flow cytometry. E, flow cytometry profiles of RKO and RKO/E6 cells treated as in (A). The percentage of apoptotic cells (sub-G1) is given.

Figure 1.

GSK-3β inhibitor LY2119301 promotes p53-dependent apoptosis in response to a variety of genotoxic agents. A, flow cytometry profiles showing cell cycle distribution of HCT116 and HCT116 p53−/− cells treated with LY2119301 (5 μmol/L), Adriamycin (ADR, 1 μmol/L), or both for 24 hours. The percentage of cells in sub-G1 is given. FLA-2, DNA content. B, flow cytometry profiles of HCT116 cells treated as in (A) and stained for active anti-caspase-3. Percentages of cells positive for active caspase-3 are indicated. C, flow cytometry profiles showing cell cycle distribution of HCT116 and HCT116 p53−/− cells treated with etoposide (10 μmol/L) or camptothecin (CPT, 100 nmol/L) for 24 hours, with or without LY2119301. The percentage of apoptotic cells (sub-G1) is given. D, dose response of HCT116 cells to varying 5-FU concentrations, in the presence or absence of LY2119301. Apoptotic fractions determined by flow cytometry. E, flow cytometry profiles of RKO and RKO/E6 cells treated as in (A). The percentage of apoptotic cells (sub-G1) is given.

Close modal

Pharmacologic modulation of GSK-3β but not GSK-3α promotes p53-dependent apoptosis. To verify that GSK-3β modulation accounts for the above observation, we examined the effects of three other structurally unrelated small-molecule GSK-3 inhibitors: SB-216763, SB-415286, and lithium chloride (LiCl). The former two compounds are competitive ATP inhibitors that are selective GSK-3 inhibitors (27), whereas lithium inhibits GSK-3 by acting as a competitive inhibitor of Mg2+ while having no inhibitory effect on other protein kinases (19, 28, 29). Similar to LY2119301, lithium treatment induced a strong apoptosis in HCT116 cells exposed to chemotherapeutic agents in a p53-dependent manner (Fig. 2A). Surprisingly, SB-216763 and SB-415286 had no effect on Adriamycin-treated cells, even at high concentrations up to 20 μmol/L (Fig. 2B). This is in contrast to LY2119301 that was able to induce apoptosis in Adriamycin-treated cells at a concentration as low as 250 nmol/L (Fig. 2B).

Figure 2.

Modulation of GSK-3β but not GSK-3α accounts for the induction of p53-dependent apoptosis. A, flow cytometry profiles of HCT116 cells treated with lithium and DNA damage as before. The percentage of apoptotic cells (sub-G1) is given. B, flow cytometry profiles showing cell cycle distribution of HCT116 cells treated with Adriamycin (ADR) and SB-216763 or SB-415286. The percentage of apoptotic cells (sub-G1) is given. C, immunoblot analysis for phospho-Ser21/9-GSK3α/β, phosphor-Tyr279/216-GSK3α/β, and total GSK-3α/β in HCT116 cells treated with lithium, LY2119301, SB-216763, or SB-415286. D, HCT116 cells were transfected with GSK-3β- or GSK-3α-specific siRNA and negative control siRNA (NC siRNA) before treated with 10 μmol/L etoposide for 24 hours. Levels of GSK-3α/β and PARP cleavage were analyzed by immunoblotting.

Figure 2.

Modulation of GSK-3β but not GSK-3α accounts for the induction of p53-dependent apoptosis. A, flow cytometry profiles of HCT116 cells treated with lithium and DNA damage as before. The percentage of apoptotic cells (sub-G1) is given. B, flow cytometry profiles showing cell cycle distribution of HCT116 cells treated with Adriamycin (ADR) and SB-216763 or SB-415286. The percentage of apoptotic cells (sub-G1) is given. C, immunoblot analysis for phospho-Ser21/9-GSK3α/β, phosphor-Tyr279/216-GSK3α/β, and total GSK-3α/β in HCT116 cells treated with lithium, LY2119301, SB-216763, or SB-415286. D, HCT116 cells were transfected with GSK-3β- or GSK-3α-specific siRNA and negative control siRNA (NC siRNA) before treated with 10 μmol/L etoposide for 24 hours. Levels of GSK-3α/β and PARP cleavage were analyzed by immunoblotting.

Close modal

To investigate whether these compounds affect GSK-3 differentially, we did immunoblot analysis. GSK-3 contains two isoforms (α and β) and their activities can be regulated by activating phosphorylation on Tyr279/Tyr216, or by inhibitory phosphorylation on Ser21/Ser9 (18). Specific GSK-3 inhibitors can either inhibit activating tyrosine phosphorylation or induce inhibitory serine autophosphorylation (18, 23, 30, 31). As such, phosphorylation status at these specific residues was used as surrogate marker for GSK-3α/β activity. Treatment of HCT116 cells with LY2119301 and lithium resulted in an identical increase in inhibitory serine phosphorylation on GSK-3β and to a much lesser extent on GSK-3α (Fig. 2C,, lanes 2 and 5). In contrast, SB-216763 and SB-415286 only induced Ser21 phosphorylation on GSK-3α and even slightly reduced the Ser9 phosphorylation on GSK-3β (Fig. 2C,, lanes 3-4). Furthermore, LY2119301 almost completely blocked the active tyrosine phosphorylation on both GSK-3α and GSK-3β, whereas SB-216763 and SB-415286 reduced tyrosine phosphorylation on GSK-3α only (Fig. 2C,, lanes 3-4). These findings suggest that LY2119301 and lithium inhibit both GSK-3α and GSK-3β but predominately on GSK-3β, whereas SB-216763 and SB-415286 were only inhibitory to GSK-3α in HCT116 cells, although both of them were previously shown to be equally effective at inhibiting GSK-3α and GSK-3β by using peptide-based in vitro protein kinase assays (27). Therefore, the apoptotic effect of GSK-3 inhibitors seems associated with the specific inactivation of GSK-3β but not GSK-3α. Importantly, lithium, in contrast to LY2119301, did not block the activating tyrosine phosphorylation of GSK-3α/β (Fig. 2C , lane 5) but was still able to induce the apoptotic phenotype. Thus, inhibition of activating tyrosine phosphorylation is not required for the apoptotic effect of GSK-3β inhibitor. Instead, the induction of inhibitory Ser9 phosphorylation of GSK-3β is the common feature of both LY2119301 and lithium and thus seems associated with their apoptotic effects. The induction of Ser9 phosphorylation of GSK-3β by LY2119301 and lithium is not likely the result of activation of other upstream kinases, such as AKT, because lithium is known to increase serine autophosphorylation through direct inhibition of GSK-3 itself, probably through inhibition of GSK-3-dependent phosphatase 1 (30). Furthermore, treatment of cells with LY294002 that inhibits AKT did not block the Ser9 phosphorylation triggered by LY2119301 and lithium (data not shown).

To test if GSK-3β knockdown by RNA interference (RNAi) would also sensitize DNA damage–induced apoptosis, HCT116 cells were transfected with GSK-3β, GSK-3α or negative control siRNA for 48 hours and followed by VP-16 treatment. Figure 2D shows that RNAis efficiently and specifically knocked down the expression of GSK-3α and GSK-3β, respectively. However, knockdown of GSK-3β or GSK-3α did not result in more cell death in response to VP-16, as no obvious PARP cleavage was observed. Thus, merely reduction of GSK-3β level was not sufficient to trigger cell death in response to DNA damage. This observation is consistent with the notion that the induction of inhibitory Ser9 phosphorylation of GSK-3β might be important for the efficient apoptosis induction in colorectal cancer cells in response to DNA damage. GSK-3β siRNA, which could not induce Ser9 phosphorylation, thus failed to mimic the apoptotic effect of GSK-3β inhibitors.

Pharmacologic modulation of GSK-3β impairs p53 accumulation and transactivation of p53 target genes in response to genotoxic stress. To determine the effects of GSK-3β inhibitors on p53 response, we did immunoblot analyses of p53 and p53 targets. HCT116 (Fig. 3A) and RKO (Fig. 3B) cells were treated with Adriamycin, VP-16, and 5-FU in the presence or absence of LY2119301. Cells treated with chemotherapeutic agents exhibited progressive accumulation of p53 and activation of its target genes, such as p21 and Puma, mediating growth arrest and apoptosis, respectively. In the presence of LY2119301, p53 accumulation induced by chemotherapy was markedly and consistently impaired, as was induction of p21 and Puma. Thus, LY2119301 impaired p53 accumulation in response to genotoxic stress and reduced its transcriptional activity. Importantly, expression of Bax, a p53 target known to be essential for mitochondrial-mediated apoptosis in colon cancer (32), was not induced following p53 activation. This is in agreement with previous reports that Bax is not a robust target of p53 in colon cancer cells (25, 33). Accordingly, LY2119301 treatment had no effect on Bax levels (Fig. 3A and B). Similarly, treatment with lithium that inhibits GSK-3β as well also impaired p53 accumulation in response to Adriamycin, VP-16, and 5-FU and reduced the induction of p53 target gene p21 (Fig. 3C,, left). Consistent with the apoptotic phenotype, cleavage of PARP was remarkably increased in cells treated with genotoxic agents in the presence of LY2119301 or lithium. In contrast, SB-216763 and SB-415286 that are only inhibitory to GSK-3α in these cells had no effect on p53 accumulation and p21 induction and did not induce PARP cleavage (Fig. 3C , right).

Figure 3.

Effects of GSK-3 inhibitors on p53 accumulation and transactivation of target genes in response to genotoxic stress. A, immunoblot analysis for p53, p21, puma, Bax, and PARP in HCT116 cells treated with Adriamycin (ADR, right), etoposide (middle), 5-FU (left), in the presence or absence of LY2119301 for the indicated times and doses. β- Actin is shown as a loading control. B, immunoblot analysis for p53, p21, puma, Bax, and PARP in RKO cells treated as in (A). C, immunoblot analysis for p53, p21, puma, Bax, and PARP in HCT116 cells, treated with the indicated drugs, with or without lithium (left) and treated with Adriamycin in the presence of LY119301, SB-216763, or SB-415286 (right). D, microarray analysis showing p53-responsive genes in HCT116 treated with Adriamycin alone or with Adriamycin and LY2119301. Cluster and Tree Viewer (34) were used to cluster 12 selected genes that have been previously identified as p53 targets in HCT116 cells (11). Red represents up-regulation relative to the untreated control (black).

Figure 3.

Effects of GSK-3 inhibitors on p53 accumulation and transactivation of target genes in response to genotoxic stress. A, immunoblot analysis for p53, p21, puma, Bax, and PARP in HCT116 cells treated with Adriamycin (ADR, right), etoposide (middle), 5-FU (left), in the presence or absence of LY2119301 for the indicated times and doses. β- Actin is shown as a loading control. B, immunoblot analysis for p53, p21, puma, Bax, and PARP in RKO cells treated as in (A). C, immunoblot analysis for p53, p21, puma, Bax, and PARP in HCT116 cells, treated with the indicated drugs, with or without lithium (left) and treated with Adriamycin in the presence of LY119301, SB-216763, or SB-415286 (right). D, microarray analysis showing p53-responsive genes in HCT116 treated with Adriamycin alone or with Adriamycin and LY2119301. Cluster and Tree Viewer (34) were used to cluster 12 selected genes that have been previously identified as p53 targets in HCT116 cells (11). Red represents up-regulation relative to the untreated control (black).

Close modal

To extend these observations to other p53 targets, we did global expression profiling of p53-induced genes using microarray analysis. We had previously identified a subset of p53-activated genes in HCT116 cells (11). As shown in Gene Cluster analysis (ref. 34; Fig. 3D), Adriamycin treatment of HCT116 cells resulted in substantial induction of these genes. Concurrent treatment with Adriamycin and LY2119301 uniformly attenuated their inductions, suggesting that GSK-3β plays a critical and global role in p53-dependent transactivation. The paradoxical observation that the GSK-3β inactivation promotes p53-mediated apoptosis, whereas abrogating transactivation of p53 targets, including Puma, raises an intriguing possibility that GSK-3β inhibitors promote p53-dependent apoptosis through a mechanism independent of p53-mediated transcription.

GSK-3β inhibitor LY2119301 promotes apoptosis induced by exogenous p53. We next investigated whether GSK-3β inactivation also promotes apoptosis induced by exogenous overexpression of p53. We used two p53 expressing systems: a tetracycline-inducible p53 expressing colon cancer DLD1 cell line (25) and HCT116 cells infected with an adenovirus expression p53. In both cases, apoptosis induced by enforced p53 expression was significantly increased in the presence of LY2119301 (Fig. 4A-B). Under these circumstances, LY2119301 had no effect on ectopically expressed p53 protein levels, or p53-dependent transcription of p21 (Fig. 4A -B, right), probably due to nonphysiologic regulation of p53 in this context. Therefore, this induction of apoptosis was not likely due to the reduced expression of p21, although it has been reported that altered balance between p21 and p53 proapoptotic targets such as Puma can switch the p53 response from growth arrest to apoptosis (12, 13).

Figure 4.

LY2119301 promotes apoptosis induced by exogenous p53 overexpression. A, p53-inducible DLD1 cells (DLD1 p53 A2) were cultured in medium with (tet off) or without (tet on) 20 ng/mL doxycyclin for 24 hours. Left, flow cytometry profiles of DLD1 p53 A2 cells in the presence or absence of LY2119301. Right, Western blot showing levels of p53, p21, and PARP. B, HCT116 cells infected with Ad-p53 or Ad-lacZ as indicated were subjected to LY2119301 or DMSO treatment for 24 hours. Cells were subjected to flow cytometric analysis (left) and immunoblot analysis for p53, p21, and PARP (right). Abbreviation: NS, nonspecific loading control.

Figure 4.

LY2119301 promotes apoptosis induced by exogenous p53 overexpression. A, p53-inducible DLD1 cells (DLD1 p53 A2) were cultured in medium with (tet off) or without (tet on) 20 ng/mL doxycyclin for 24 hours. Left, flow cytometry profiles of DLD1 p53 A2 cells in the presence or absence of LY2119301. Right, Western blot showing levels of p53, p21, and PARP. B, HCT116 cells infected with Ad-p53 or Ad-lacZ as indicated were subjected to LY2119301 or DMSO treatment for 24 hours. Cells were subjected to flow cytometric analysis (left) and immunoblot analysis for p53, p21, and PARP (right). Abbreviation: NS, nonspecific loading control.

Close modal

Apoptosis promoted by GSK-3β inhibitor requires Bax but not Puma. Proapoptotic Bcl2 family members Bax and Puma have been implicated in p53-mediated apoptosis (35, 36). To further asses the roles of Bax and Puma in p53-dependent apoptosis after GSK-3β inactivation, we used HCT116 cells with either of these genes deleted by homologous recombination (32, 36). When cotreated with LY2119301 and Adriamycin or VP-16, Bax+/− cells were completely refractory to apoptosis (Fig. 5A). In contrast, Puma−/− cells were still able to undergo marked apoptosis, identical to parental HCT116 cells (Fig. 5B). Consistently, PARP cleavage was not observed in Bax+/− cells (Fig. 5C) but evident in Puma−/− cells (Fig. 5D), although in both cell lines p53 accumulation and induction of p21 after Adriamycin or VP-16 treatment was similarly reduced by LY2119301 (Fig. 5C-D). These results indicate that p53-dependent but transcription-independent apoptosis triggered by GSK-3β inhibitor proceeds through a pathway that requires Bax but not Puma. Supporting this notion, colon cancer LoVo cells that are p53 wild type but do not express Bax protein due to a frame-shift mutation in the Bax gene (37) were resistant to the apoptosis inducted by LY2119301 upon DNA damage (data not shown).

Figure 5.

Apoptosis induced by LY2119301 in response to DNA damage requires Bax but not Puma. A, flow cytometry profiles showing cell cycle distribution of HCT116 Bax+/− cells treated with Adriamycin (ADR) or etoposide in the presence or absence of LY2119301 for 24 hours. Apoptotic (sub-G1) fractions as indicated. B, flow cytometry profile of HCT116 Puma−/− cells as in (A). C, immunoblot analysis for p53, p21, and PARP in HCT116 Bax+/− cells, treated as indicated. D, immunoblot analysis for p53, p21, and PARP in HCT116 Puma−/− cells, treated as indicated.

Figure 5.

Apoptosis induced by LY2119301 in response to DNA damage requires Bax but not Puma. A, flow cytometry profiles showing cell cycle distribution of HCT116 Bax+/− cells treated with Adriamycin (ADR) or etoposide in the presence or absence of LY2119301 for 24 hours. Apoptotic (sub-G1) fractions as indicated. B, flow cytometry profile of HCT116 Puma−/− cells as in (A). C, immunoblot analysis for p53, p21, and PARP in HCT116 Bax+/− cells, treated as indicated. D, immunoblot analysis for p53, p21, and PARP in HCT116 Puma−/− cells, treated as indicated.

Close modal

GSK-3β inhibition promotes p53-dependent conformational activation of Bax and induces mitochondrial cell death. Bax is known to undergo a conformational change during apoptosis (38). We next examined whether GSK-3β inhibitor induces a conformational activation of Bax after DNA damage. Cotreatment of HCT116 cells with LY2119301 and genotoxic agents Adriamycin, VP-16, and 5-FU dramatically increased levels of conformationally active Bax (Fig. 6A), which can be detected by a specific anti-Bax monoclonal antibody 6A7, as previously described (39, 40). Bax activation was p53 dependent, as there was no detectable change in Bax conformation in HCT116 p53−/− cells under the same conditions (Fig. 6A). Consistently, increased cytochrome c release and caspase-9 processing were evident after DNA damage in the presence of LY2119301 (Fig. 6A,, bottom). Bax activation was similarly seen in DLD1 cells and HCT116 cells with concurrent p53 overexpression in the presence of LY2119301 (Fig. 6B). Moreover, a significant reduction of mitochondrial membrane potential, as determined by JC-1 staining and FACS, was detected in Bax+/+ cells cotreated with LY2119301 and other genotoxic treatments but not in Bax+/− cells under the same conditions (Fig. 6C). Taken together, these findings show that GSK-3β inhibitor promotes p53-dependent apoptosis through Bax-mediated mitochondrial pathway.

Figure 6.

LY2119301 promotes p53-dependent Bax activation and mitochondrial cell death. A, p53 wild type and p53 null HCT116 cells were treated with the indicated drugs for 24 hours, after which cells were lysed in buffer containing 1% CHAPS; conformationally changed Bax protein was immunoprecipitated using anti-Bax 6A7 antibody and subjected to Western blot analysis using polyclonal Bax antibody (top). Western blots of cytosolic cell fractions showing cytochrome c release and the proteolysis of caspase-9 in HCT116 cells cotreated with LY2119301 and Adriamycin (ADR) or etoposide (bottom). B, Bax activation in DLD1 p53 A2 cells with p53 induction (left) and HCT116 cells with p53 overexpression (right) and treated with LY2119301. C, flow cytometry profile showing JC-1 staining as an indicator of mitochondrial membrane potential in HCT116 and HCT116 Bax+/− cells. Percentage of cells with ÄΨm loss is indicated. D, time course analysis of p53 accumulation in the cytosol and mitochondria following etoposide and 5-FU treatments (top). p53 accumulation in the presence or absence of LY2119301 following etoposide treatment for 24 hours. Levels of p53 and mitochondria marker CoxIV were determined by Western blot analysis.

Figure 6.

LY2119301 promotes p53-dependent Bax activation and mitochondrial cell death. A, p53 wild type and p53 null HCT116 cells were treated with the indicated drugs for 24 hours, after which cells were lysed in buffer containing 1% CHAPS; conformationally changed Bax protein was immunoprecipitated using anti-Bax 6A7 antibody and subjected to Western blot analysis using polyclonal Bax antibody (top). Western blots of cytosolic cell fractions showing cytochrome c release and the proteolysis of caspase-9 in HCT116 cells cotreated with LY2119301 and Adriamycin (ADR) or etoposide (bottom). B, Bax activation in DLD1 p53 A2 cells with p53 induction (left) and HCT116 cells with p53 overexpression (right) and treated with LY2119301. C, flow cytometry profile showing JC-1 staining as an indicator of mitochondrial membrane potential in HCT116 and HCT116 Bax+/− cells. Percentage of cells with ÄΨm loss is indicated. D, time course analysis of p53 accumulation in the cytosol and mitochondria following etoposide and 5-FU treatments (top). p53 accumulation in the presence or absence of LY2119301 following etoposide treatment for 24 hours. Levels of p53 and mitochondria marker CoxIV were determined by Western blot analysis.

Close modal

The above results raise the possibility that GSK-3β inhibitors might enable the p53 accumulation in the mitochondria thus accounting for the observed apoptosis following DNA damage. To address this issue, we first monitored the p53 accumulation in both cytosol and mitochondria at different time points following VP-16 and 5-FU treatments. We found that p53 was accumulated in the mitochondria as early as 2 hours after VP-16 or 5-FU treatment, whereas the accumulation of p53 in the cytosol was not seen until 6 hours after the drug treatments (Fig. 6D). This is consistent with a recent report showing that DNA damage triggers a rapid p53 mitochondrial accumulation in mouse thymus, which led to the first wave of cell death (16) However, VP-16-induced p53 accumulation in the mitochondria in HCT116 cells failed to induce apoptosis. This shows that p53 accumulation in mitochondria per se is not sufficient to activate apoptosis in HCT116 cells. Furthermore, in the presence of LY2119301, VP-16-induced p53 accumulations were similarly decreased in both cytosol and mitochondria (Fig. 6D), albeit a robust apoptosis was induced under these conditions. Thus, we conclude that the mitochondrial p53 levels in colorectal cancer cells are not correlated with the levels of apoptosis induction in colorectal cancer cells, and that the apoptotic effects of GSK-3β inhibitors on p53 response is not due to increased translocation of p53 to the mitochondria.

Current understanding of p53 biology suggests that the molecular mechanisms regulating apoptosis are orchestrated through transcription-dependent and -independent mechanisms. Nevertheless, regulation of these p53-dependent proapoptotic pathways remains enigmatic. In this study, we have shown that constitutively expressed cellular kinase GSK-3β seems to have dual but distinct roles in regulating p53 pathways: promoting its transcriptional activity in the nucleus but suppressing its direct apoptotic activity at the mitochondria. Using small-molecular inhibitors, we interrogated the p53 response to DNA damage, and based on the observations reported, provide novel insights into the regulation of p53 function.

First, the p53-mediated apoptosis, induced by GSK-3β inhibitors LY2119301 and lithium, was robust even when p53-mediated transactivation was remarkably impaired. This challenges the long-held view that p53 induces apoptosis largely through transcriptional activation of proapoptotic genes. This finding is in agreement with a growing body of evidence that support the premise that p53-dependent transcription is not always required for its apoptotic function (4143) and further reinforce a direct role of p53 in inducing apoptosis (14, 15). Consistent with this notion is our finding that the proapoptotic Bcl2 family member Puma, the most inducible p53 proapoptotic target (11, 35), is not required for GSK-3β inhibitor–induced but p53-dependent apoptosis. This is in sharp contrast to the previous studies where Puma was shown to be essential for p53-mediated apoptosis in mouse primary thymocytes (36, 44, 45) or p21-depleted HCT116 cells (36). However, as recently proposed (16), p53-dependent transactivation of proapoptotic target genes, such as Puma, only contributes to a late stage of apoptosis. This is preceded by an earlier wave of robust apoptosis, which is clearly transcription independent and is mediated by direct p53 mitochondrial function. In support of this model, experiments on thymocytes from Puma−/− mice showed protection from apoptosis only after prolonged treatment with γ-IR after 16 to 72 hours. Remarkably, there was little difference between wild-type and Puma−/− cells within the first 8 hours after damage when apoptosis is believed to be primarily mediated by the direct function of p53 at the mitochondria (16, 44, 45).

Second, although the direct role of p53 in inducing apoptosis has been described by several groups (14, 46, 47), little is known about how this arm of p53 function is regulated. Our study identified a novel regulator of p53-dependent direct apoptotic pathway. Modulation of GSK-3β, by two structurally unrelated inhibitors LY2119301 and lithium, was able to consistently and robustly convert the p53 response from growth arrest to apoptosis. Although GSK-3β inhibition by LY2119301 and lithium can diminish p53-dependent p21 induction, which might contribute to the increased apoptosis, we propose that this is not the primary mechanism through which GSK-3β inactivation triggers p53-dependent apoptosis. Indeed, LY2119301 had no effect on p21 expression induced by transfected p53 but could still dramatically induce p53-dependent apoptosis through direct conformational activation of Bax. The precise mechanism through which LY2119301 and lithium regulate p53 pathway remains unknown. However, the inhibitory Ser9 phosphorylation of GSK-3β seems important during this process. A recent study showed that Ser9 phosphorylation is required for GSK-3β binding to 14-3-3 (48, 49). One of the mechanisms for 14-3-3 to regulate apoptosis is through binding and sequestration of Bax (50) or Bad (51, 52) and thus prevents Bax activation. On this basis, we speculate that increased Ser9 phosphorylation of GSK-3β by LY2119301 or lithium could displace Bax or Bad by competitively binding to 14-3-3 thus promoting Bax activation and cell death. Thus, the modulation of the p53 pathway by GSK-3β inhibitors may be caused by molecular events subsequent to the Ser9 phosphorylation and protein-protein interaction rather than the reduced activity of GSK-3β itself. Consistent with this notion, we found that simply knockdown of GSK-3β expression by siRNA, which could not induce the Ser9 phosphorylation, failed to induce the same phenotype.

Third, p53 translocation to the mitochondria can directly trigger apoptosis through Bax activation (14). In this study, we found that p53 accumulation occurred quickly in the mitochondria under both apoptotic (5-FU-treated cells) and nonapoptotic conditions (VP-16-treated cells). Moreover, LY2119301 treatment induced massive apoptosis in DNA-damaged p53 wild-type HCT116 cells, albeit p53 level was reduced in the mitochondria. These results indicate that merely p53 accumulation in the mitochondria is not sufficient to activate apoptosis and that the level of p53 in the mitochondria is not necessarily correlated to the level of apoptosis in colorectal cancer cells. A recent study shows that siRNA knockdown of Bcl-2 expression in HCT116 cells can trigger a strong p53-dependent apoptosis without requiring the genotoxic stress and p53 accumulation (53). Thus, the apoptosis induction might depend on the fine balance between p53 interaction with proapoptotic and antiapoptotic proteins in Bcl-2 family. Further studies will investigate how the GSK-3β inhibitors interfere with this balance and trigger apoptosis in colon cancer cells.

Fourth, the regulation of GSK-3β on p53 apoptotic response seems context dependent. In human neuroblastoma SH-SY5Y cells, inhibition of GSK-3β by lithium was also found to inhibit p53-dependent transcription but impaired the p53 apoptotic response upon DNA damage (19, 54). Thus, the outcome of the GSK-3β inhibition on p53 response seems tissue specific. The reason for this tissue specificity is unknown. It has been suggested that the control of apoptosis may rest on tissue- and pathway-specific interactions of multiple Bcl-2-related proteins (55). Another possibility is that the mechanism of p53-mediated apoptosis may itself be tissue dependent. It is well documented that p53-dependent transcriptional activity is indispensable for apoptosis in mouse embryonic stem cells, primary thymocytes (56, 57), and normal rat kidney cells (58). However, this has not been consistently shown in cancer cells, and in many cases, p53-dependent transcription has been found not to be essential and even inhibitory for its apoptotic activity (33). Theoretically, this could account for the differences in apoptosis seen after GSK-3β inactivation in colorectal cancer lines, where p53-mediated direct mitochondrial activation may predominate, versus other cancer cell lines, which require p53-transcriptional activity for apoptosis.

Finally, these findings raise an intriguing therapeutic potential that may be exploited clinically. GSK-3β inhibitors are already in use for cancer treatment. It is therefore conceivable that specific GSK-3β inhibitors can be used in combination with chemotherapeutic agents to augment cancer cell death. In vitro assays have shown a profound synergistic effect on 5-FU mediated apoptosis, which is the first line of chemotherapy used in colorectal cancer. Importantly, the tissue specificity of GSK-3β function and its dual regulations on p53 response suggest a further level of therapeutic benefit: GSK-3β inactivation could selectively promote p53-mediated apoptosis through a direct mitochondrial pathway in cancer cells, whereas having no or protective effect on normal tissues where transcription-dependent death pathway of p53 is impaired.

Grant support: Agency for Science, Technology, and Research of Singapore.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Dr. Bert Vogelstein for p53, Bax, or Puma-depleted HCT116 cells and p53-inducible DLD1 cells; Dr. Shibin Zhou for Ad-p53; and Samuel Ong for technical assistance.

1
Burns TF, El-Deiry WS. The p53 pathway and apoptosis.
J Cell Physiol
1999
;
181
:
231
–9.
2
Gottlieb TM, Oren M. p53 and apoptosis.
Semin Cancer Biol
1998
;
8
:
359
–68.
3
Levine AJ. p53, the cellular gatekeeper for growth and division.
Cell
1997
;
88
:
323
–31.
4
Vousden KH, Lu X. Live or let die: the cell's response to p53.
Nat Rev Cancer
2002
;
2
:
594
–604.
5
Deng C, Zhang P, Harper JW, Elledge SJ, Leder P. Mice lacking p21CIP1/WAF1 undergo normal development, but are defective in G1 checkpoint control.
Cell
1995
;
82
:
675
–84.
6
Chan TA, Hwang PM, Hermeking H, Kinzler KW, Vogelstein B. Cooperative effects of genes controlling the G(2)/M checkpoint.
Genes Dev
2000
;
14
:
1584
–8.
7
Brugarolas J, Chandrasekaran C, Gordon JI, et al. Radiation-induced cell cycle arrest compromised by p21 deficiency.
Nature
1995
;
377
:
552
–7.
8
el-Deiry WS, Tokino T, Velculescu VE, et al. WAF1, a potential mediator of p53 tumor suppression.
Cell
1993
;
75
:
817
–25.
9
Bates S, Ryan KM, Phillips AC, Vousden KH. Cell cycle arrest and DNA endoreduplication following p21Waf1/Cip1 expression.
Oncogene
1998
;
17
:
1691
–703.
10
Fridman JS, Lowe SW. Control of apoptosis by p53.
Oncogene
2003
;
22
:
9030
–40.
11
Kho PS, Wang Z, Zhuang L, et al. p53-regulated transcriptional program associated with genotoxic stress-induced apoptosis.
J Biol Chem
2004
;
279
:
21183
–92.
12
Seoane J, Le HV, Massague J. Myc suppression of the p21(Cip1) Cdk inhibitor influences the outcome of the p53 response to DNA damage.
Nature
2002
;
419
:
729
–34.
13
Iyer NG, Chin SF, Ozdag H, et al. p300 regulates p53-dependent apoptosis after DNA damage in colorectal cancer cells by modulation of PUMA/p21 levels.
Proc Natl Acad Sci U S A
2004
;
101
:
7386
–91.
14
Chipuk JE, Kuwana T, Bouchier-Hayes L, et al. Direct activation of Bax by p53 mediates mitochondrial membrane permeabilization and apoptosis.
Science
2004
;
303
:
1010
–4.
15
Mihara M, Erster S, Zaika A, et al. p53 has a direct apoptogenic role at the mitochondria.
Mol Cell
2003
;
11
:
577
–90.
16
Erster S, Mihara M, Kim RH, Petrenko O, Moll UM. In vivo mitochondrial p53 translocation triggers a rapid first wave of cell death in response to DNA damage that can precede p53 target gene activation.
Mol Cell Biol
2004
;
24
:
6728
–41.
17
Chipuk JE, Maurer U, Green DR, Schuler M. Pharmacologic activation of p53 elicits Bax-dependent apoptosis in the absence of transcription.
Cancer Cell
2003
;
4
:
371
–81.
18
Cohen P, Frame S. The renaissance of GSK3.
Nat Rev Mol Cell Biol
2001
;
2
:
769
–76.
19
Watcharasit P, Bijur GN, Song L, et al. Glycogen synthase kinase-3β (GSK3β) binds to and promotes the actions of p53.
J Biol Chem
2003
;
278
:
48872
–9.
20
Hetman M, Cavanaugh JE, Kimelman D, Xia Z. Role of glycogen synthase kinase-3β in neuronal apoptosis induced by trophic withdrawal.
J Neurosci
2000
;
20
:
2567
–74.
21
Bijur GN, De Sarno P, Jope RS. Glycogen synthase kinase-3β facilitates staurosporine- and heat shock-induced apoptosis. Protection by lithium.
J Biol Chem
2000
;
275
:
7583
–90.
22
Qu L, Huang S, Baltzis D, et al. Endoplasmic reticulum stress induces p53 cytoplasmic localization and prevents p53-dependent apoptosis by a pathway involving glycogen synthase kinase-3β.
Genes Dev
2004
;
18
:
261
–77.
23
Cohen P, Goedert M. GSK3 inhibitors: development and therapeutic potential.
Nat Rev Drug Discov
2004
;
3
:
479
–87.
24
Yu Q, La Rose J, Zhang H, et al. UCN-01 Inhibits p53 up-regulation and abrogates γ-radiation-induced G(2)-M checkpoint independently of p53 by targeting both of the checkpoint kinases, Chk2 and Chk1.
Cancer Res
2002
;
62
:
5743
–8.
25
Yu J, Zhang L, Hwang PM, et al. Identification and classification of p53-regulated genes.
Proc Natl Acad Sci U S A
1999
;
96
:
14517
–22.
26
Bunz F, Hwang PM, Torrance C, et al. Disruption of p53 in human cancer cells alters the responses to therapeutic agents.
J Clin Invest
1999
;
104
:
263
–9.
27
Coghlan MP, Culbert AA, Cross DA, et al. Selective small molecule inhibitors of glycogen synthase kinase-3 modulate glycogen metabolism and gene transcription.
Chem Biol
2000
;
7
:
793
–803.
28
Ryves WJ, Harwood AJ. Lithium inhibits glycogen synthase kinase-3 by competition for magnesium.
Biochem Biophys Res Commun
2001
;
280
:
720
–5.
29
Klein PS, Melton DA. A molecular mechanism for the effect of lithium on development.
Proc Natl Acad Sci U S A
1996
;
93
:
8455
–9.
30
Zhang F, Phiel CJ, Spece L, Gurvich N, Klein PS. Inhibitory phosphorylation of glycogen synthase kinase-3 (GSK-3) in response to lithium. Evidence for autoregulation of GSK-3.
J Biol Chem
2003
;
278
:
33067
–77.
31
Cole A, Frame S, Cohen P. Further evidence that the tyrosine phosphorylation of glycogen synthase kinase-3 (GSK3) in mammalian cells is an autophosphorylation event.
Biochem J
2004
;
377
:
249
–55.
32
Zhang L, Yu J, Park BH, Kinzler KW, Vogelstein B. Role of BAX in the apoptotic response to anticancer agents.
Science
2000
;
290
:
989
–92.
33
Kokontis JM, Wagner AJ, O'Leary M, Liao S, Hay N. A transcriptional activation function of p53 is dispensable for and inhibitory of its apoptotic function.
Oncogene
2001
;
20
:
659
–68.
34
Eisen MB, Spellman PT, Brown PO, Botstein D. Cluster analysis and display of genome-wide expression patterns.
Proc Natl Acad Sci U S A
1998
;
95
:
14863
–8.
35
Yu J, Zhang L, Hwang PM, Kinzler KW, Vogelstein B. PUMA induces the rapid apoptosis of colorectal cancer cells.
Mol Cell
2001
;
7
:
673
–82.
36
Yu J, Wang Z, Kinzler KW, Vogelstein B, Zhang L. PUMA mediates the apoptotic response to p53 in colorectal cancer cells.
Proc Natl Acad Sci U S A
2003
;
100
:
1931
–6.
37
Ionov Y, Yamamoto H, Krajewski S, Reed JC, Perucho M. Mutational inactivation of the proapoptotic gene BAX confers selective advantage during tumor clonal evolution.
Proc Natl Acad Sci U S A
2000
;
97
:
10872
–7.
38
Wolter KG, Hsu YT, Smith CL, et al. Movement of Bax from the cytosol to mitochondria during apoptosis.
J Cell Biol
1997
;
139
:
1281
–92.
39
Suzuki M, Youle RJ, Tjandra N. Structure of Bax: coregulation of dimer formation and intracellular localization.
Cell
2000
;
103
:
645
–54.
40
Nechushtan A, Smith CL, Hsu YT, Youle RJ. Conformation of the Bax C-terminus regulates subcellular location and cell death.
EMBO J
1999
;
18
:
2330
–41.
41
Caelles C, Helmberg A, Karin M. p53-dependent apoptosis in the absence of transcriptional activation of p53-target genes.
Nature
1994
;
370
:
220
–3.
42
Wagner AJ, Kokontis JM, Hay N. Myc-mediated apoptosis requires wild-type p53 in a manner independent of cell cycle arrest and the ability of p53 to induce p21waf1/cip1.
Genes Dev
1994
;
8
:
2817
–30.
43
Gao C, Tsuchida N. Activation of caspases in p53-induced transactivation-independent apoptosis.
Jpn J Cancer Res
1999
;
90
:
180
–7.
44
Jeffers JR, Parganas E, Lee Y, et al. Puma is an essential mediator of p53-dependent and -independent apoptotic pathways.
Cancer Cell
2003
;
4
:
321
–8.
45
Villunger A, Michalak EM, Coultas L, et al. p53- and drug-induced apoptotic responses mediated by BH3-only proteins puma and noxa.
Science
2003
;
302
:
1036
–8.
46
Moll UM, Zaika A. Nuclear and mitochondrial apoptotic pathways of p53.
FEBS Lett
2001
;
493
:
65
–9.
47
Leu JI, Dumont P, Hafey M, Murphy ME, George DL. Mitochondrial p53 activates Bak and causes disruption of a Bak-Mcl1 complex.
Nat Cell Biol
2004
;
6
:
443
–50.
48
Yuan Z, Agarwal-Mawal A, Paudel HK. 14-3-3 binds to and mediates phosphorylation of microtubule-associated tau protein by Ser9-phosphorylated glycogen synthase kinase 3β in the brain.
J Biol Chem
2004
;
279
:
26105
–14.
49
Frame S, Cohen P, Biondi RM. A common phosphate binding site explains the unique substrate specificity of GSK3 and its inactivation by phosphorylation.
Mol Cell
2001
;
7
:
1321
–7.
50
Nomura M, Shimizu S, Sugiyama T, et al. 14-3-3 Interacts directly with and negatively regulates pro-apoptotic Bax.
J Biol Chem
2003
;
278
:
2058
–65.
51
Zha J, Harada H, Yang E, Jockel J, Korsmeyer SJ. Serine phosphorylation of death agonist BAD in response to survival factor results in binding to 14-3-3 not BCL-X(L).
Cell
1996
;
87
:
619
–28.
52
Chittenden T. BH3 domains: intracellular death-ligands critical for initiating apoptosis.
Cancer Cell
2002
;
2
:
165
–6.
53
Jiang M, Milner J. Bcl-2 constitutively suppresses p53-dependent apoptosis in colorectal cancer cells.
Genes Dev
2003
;
17
:
832
–7.
54
Watcharasit P, Bijur GN, Zmijewski JW, et al. Direct, activating interaction between glycogen synthase kinase-3β and p53 after DNA damage.
Proc Natl Acad Sci U S A
2002
;
99
:
7951
–5.
55
Hsu SY, Hsueh AJ. Tissue-specific Bcl-2 protein partners in apoptosis: an ovarian paradigm.
Physiol Rev
2000
;
80
:
593
–614.
56
Attardi LD, Lowe SW, Brugarolas J, Jacks T. Transcriptional activation by p53, but not induction of the p21 gene, is essential for oncogene-mediated apoptosis.
EMBO J
1996
;
15
:
3693
–701.
57
Chao C, Saito S, Kang J, et al. p53 transcriptional activity is essential for p53-dependent apoptosis following DNA damage.
EMBO J
2000
;
19
:
4967
–75.
58
Sabbatini P, Lin J, Levine AJ, White E. Essential role for p53-mediated transcription in E1A-induced apoptosis.
Genes Dev
1995
;
9
:
2184
–92.