Crizotinib Induces PUMA-Dependent Apoptosis in Colon Cancer Cells

Published OnlineFirst February 20, 2013; DOI: 10.1158/1535-7163.MCT-12-1146

Molecular Cancer Cancer Therapeutics Insights Therapeutics

Xingnan Zheng1, Kan He1, Lin Zhang2, and Jian Yu1

Abstract Oncogenic alterations in MET or anaplastic lymphoma kinase (ALK) have been identified in a variety of human cancers. Crizotinib (PF02341066) is a dual MET and ALK inhibitor and approved for the treatment of a subset of non–small cell lung carcinoma and in clinical development for other malignancies. Crizotinib can induce apoptosis in cancer cells, whereas the underlying mechanisms are not well understood. In this study, we found that crizotinib induces apoptosis in colon cancer cells through the BH3-only protein PUMA. In cells with wild-type p53, crizotinib induces rapid induction of PUMA and Bim accompanied by p53 stabilization and DNA damage response. The induction of PUMA and Bim is mediated largely by p53, and deficiency in PUMA or p53,butnotBim, blocks crizotinib-induced apoptosis. Interestingly, MET knockdown led to selective induction of PUMA, but not Bim or p53. Crizotinib also induced PUMA- dependent apoptosis in p53-deficient colon cancer cells and synergized with gefitinib or sorafenib to induce marked apoptosis via PUMA in colon cancer cells. Furthermore, PUMA deficiency suppressed apoptosis and therapeutic responses to crizotinib in xenograft models. These results establish a critical role of PUMA in mediating apoptotic responses of colon cancer cells to crizotinib and suggest that mechanisms of oncogenic addiction to MET/ALK-mediated survival may be cell type-specific. These findings have important implications for future clinical development of crizotinib. Mol Cancer Ther; 12(5); 777–86. 2013 AACR.

Introduction (2, 3). In colon cancer, MET overexpression and gene Receptor tyrosine kinases (RTK) are cell surface recep- amplification are associated with advanced diseases and tors that act upon a variety of ligands including growth poor prognosis (4, 5). factors, cytokines, and hormones. RTKs are vital regula- Recent efforts in cancer genomics continue to identify tors of normal cell physiology and play critical roles in the aberrantly activated oncogenic kinases and facilitate development and progression of human cancer (1, 2). the development of targeted agents. This approach is MET is an extensively studied RTK and the receptor for expected to ultimately deliver safer and more effective hepatocyte growth factor (HGF; ref. 3). The activation of cancer therapeutics (6). MET-targeting agents currently in MET by HGF ligation initiates various signaling cascades, clinical use include the monoclonal antibody MetAb such as the PI3K/AKT and RAS/RAF/MAPK pathways, (Roche; ref. 7) and small-molecule tyrosine kinase to induce cell survival, proliferation, migration, and tissue inhibitors (TKI), such as ARQ197 (ArQule/Daiichi Sankyo; regeneration (3). Aberrant activation of MET can result ref. 8), INCB28060 (Incyte; ref. 9), and crizotinib from gene amplification, transcriptional upregulation, (PF02341066, Pfizer; ref. 10). Crizotinib was initially missense mutations, or ligand autocrine loops, and is designed as a selective ATP-competitive MET inhibitor implicated in the pathogenesis of many human cancers and later found to inhibit several related kinases, including anaplastic lymphoma kinase (ALK; ref. 10), C-ros onco- gene1, and receptor tyrosine kinase (ROS1; ref. 11). Crizo- Authors' Affiliations: 1Department of Pathology, University of Pittsburgh tinib has received the approval of the U.S. Food and Drug Cancer Institute; and 2Department of Pharmacology and Chemical Biology, Administration for the treatment of ALK-rearranged non– University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania small cell lung carcinoma (NSCLC), and is being evaluated Note: Supplementary data for this article are available at Molecular Cancer in patients with other malignancies. Crizotinib has gar- Therapeutics Online (http://mct.aacrjournals.org/). nered much attention as it inhibits MET- and ALK-depen- X. Zheng and K. He contributed equally to this work. dent tumor cell growth, migration, and invasion via both Corresponding Author: Jian Yu, Hillman Cancer Center Research Pavil- HGF-dependent and -independent mechanisms (12). Cri- ion, Suite 2.26h, 5117 Centre Ave, Pittsburgh, PA 15213. Phone: 412-623- zotinib also induces apoptosis in cancer cells; however, the 7786; Fax: 412-623-7778; E-mail: [email protected] underlying mechanisms are not well understood. Current- doi: 10.1158/1535-7163.MCT-12-1146 ly, there is no reliable biomarker for crizotinib response 2013 American Association for Cancer Research. other than ALK or ROS1 rearrangement (11, 13, 14).

www.aacrjournals.org 777

Zheng et al.

Apoptosis plays an important role in the antitumor #R1055, Zymo Research) according to the manufac- activities of conventional chemotherapeutic agents and turer’s protocol. Total RNA (2 mg) was used to generate targeted therapies (1, 15, 16). The Bcl-2 family proteins complementary DNA using SuperScript III reverse tran- are the central regulators of mitochondria-mediated scriptase (Invitrogen). The following primers were used apoptosis. The BH3-only family members are first for PUMA:Forward:50-CGACCTCAACGCACAGTAC- engaged in response to distinct as well as overlapping GA-30,Reverse:50-AGGCACCTAATTGGGCTCCAT-30, signals. Several of them, such as Bim and PUMA, are b-Actin:Forward:50-GACCTCACAGACTACCTCAT- potent inducers of apoptosis by activating Bax/Bak 30, Reverse: 50-AGACAGCACTGTGTTGGCTA-30. following the neutralization of antiapoptotic Bcl-2 family members (15, 17). We and others have shown that Transfection and siRNA PUMA functions as a critical initiator of apoptosis in The gene-specific siRNA, including MET siRNA (30) both p53-dependent and -independent manners in a and PUMA siRNA (31) were synthesized by Dharmacon wide variety of cell types (18). PUMA transcription is (Lafayette) and transfected into cells with Lipofectamine directly activated by p53 in response to DNA damage 2000 (Invitrogen) according to the manufacturer’s instruc- (18), and lack of its induction renders p53-deficient tions. After 24 hours of transfection, cells were treated cancer cells refractory to chemotherapeutic drugs and with PF02341066 for further analysis. More details are radiation. PUMA induction by nongenotoxic stimuli is found in the Supplementary Data. generally p53-independent, and mediated by transcription factors such as p73 (19, 20), forkhead box O3a Analysis of apoptosis, growth, and mitochondria- (FoxO3a; ref. 21, 22), and NF-kB (23, 24). Upon induc- associated events tion, PUMA potently induces apoptosis by antagonizing Apoptosis was analyzed by counting cells with con- antiapoptotic Bcl-2 family members and/or directly densed chromatin and micronucleation following nuclear activating the proapoptotic members Bax and Bak, staining with Hoechst 33258 (Invitrogen; ref. 27). The leading to mitochondrial dysfunction and caspase acti- methods of colony formation, changes in the mitochon- vation cascade (18). drial membrane potential, and cytochrome c release have In this study, we investigated the underlying mechan- been previously described (25, 32). More details are found isms of crizotinib-induced apoptosis in colon cancer cells, in the Supplementary Data. and found that both p53-dependent and -independent induction of PUMA contributes to crizotinib-induced Reporter assays apoptosis. These results provide novel mechanistic PUMA reporters with or without p53-bindings sites insight into the therapeutic responses of crizotinib, a have been described previously (20). Reporter assays rationale for manipulating PUMA and BH3-only proteins were carried out in 12-well plates as described (33). to improve the efficacy of targeted therapies, as well as Normalized relative luciferase units were plotted. All therapy-induced changes in their expression as potential reporter experiments were carried out in triplicate and biomarkers. repeated 3 times.

Materials and Methods Xenograft studies Cell culture and drug treatment All animal experiments were approved by the Univer- Human colorectal cancer cell lines, including HCT116, sity of Pittsburgh Institutional Animal Care and Use RKO, LoVo, DLD1, and HT29 were obtained from Amer- Committee. Female 5- to 6-week-old Nu/Nu mice ican Type Culture Collection. The isogenic cell lines, (Charles River) were housed in a sterile environment with including PUMA-KO (25), p53-KO (26), and p53-binding microisolator cages and allowed access to water and chow site knockout (BS-KO; ref. 27) HCT116 cells, PUMA-KO ad libitum. Mice were injected subcutaneously in both DLD1 cells (27), and p53-KO RKO cells (28) have been flanks with 4 million wild-type (WT) or PUMA-KO described. More details are found in the Supplementary HCT116 cells. Following tumor growth for 7 days, mice Data for drug treatments. We examine loss of expression were treated daily by oral gavage for 9 consecutive days of targeted protein by Western blotting and conduct with 35 mg/kg PF02341066 in 10% ethanol or 10% ethanol mycoplasma testing by PCR during culture routinely; no without PF02341066 (control buffer), the total volume addition authentication was done by the authors. being approximately 100 mL/mouse. Detailed methods on tumor measurements, harvests, and histologic analysis Western blotting are found in the Supplementary Data as previously Western blotting was carried out as previously described (34–36). described (29). More details on antibodies are found in the Supplementary Data. Statistical analysis Statistical analyses were carried out in Microsoft Excel. Real-time reverse transcription PCR P values were calculated by the Student t test, and were Total RNA was isolated from untreated or drug-trea- considered significant if P < 0.05. The means 1 SD are ted cells using the Mini–RNA Isolation II Kit (cat. displayed in the figures.

778 Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics

PUMA and Crizotinib-Induced Apoptosis

Figure 1. PUMA was induced by crizotinib in colon cancer cells. A, top, chemical structure of crizotinib (PF02342066). Bottom, HCT116 cells were treated with 12 mmol/L crizotinib or PF02341066 (PF) for indicated time. The expression levels of PUMA, p-Met (T1234/1235), total Met, p-AKT (S473), total AKT, p-ERK (T202/Y204), and total ERK were analyzed by Western blotting. B, HCT116 cells were treated with 12 mmol/L PF02341066 and total RNA was extracted at the indicated time points. PUMA mRNA expression was analyzed by semiquantitive reverse transcription PCR (RT-PCR). b-Actin was used as a control. C, PUMA protein levels were analyzed by Western blotting in HCT116 cells treated with increasing doses of PF02341066 for 24 hours. D, HCT116 cells were transfected with either a control scrambled siRNA or a Met siRNA for 24 hours. Met and PUMA expression was analyzed by Western blotting (left) and real-time PCR (right). NS, nonspeciﬁc band for Met. , P < 0.01, MET siRNA þ versus . b-Actin was used as a loading control for Western blot analyses.

Results crizotinib has a broader effect on the levels of PUMA and PUMA is induced by crizotinib in colon cancer cells other BH3-only proteins. We ﬁrst investigated the effects of crizotinib (PF02341066, PF) on MET signaling and PUMA expres- PUMA mediates crizotinib-induced apoptosis sion in HCT116 colon cancer cells. Crizotinib treatment Next, we determined the role of PUMA in crizotinib- led to rapid dephosphorylation of MET and AKT without induced apoptosis using isogenic PUMA-KO HCT116 affecting their total levels. However, the levels of phos- cells and siRNA. Crizotinib treatment induced approxi- phorylated extracellular signal-regulated kinase (ERK) mately 15% to 65% apoptosis from 24 to 48 hours in WT only decreased transiently, and recovered within 6 hours HCT116 cells, which was associated with the activation of (Fig. 1A). PUMA protein and mRNA were induced within caspase-3 and caspase-9, mitochondrial membrane depo- 6 hours of treatment, suggesting transcriptional regula- larization, and cytochrome c release (Fig. 2A and B). In tion (Fig. 1A and B). Interestingly, the expression of Bcl-2 contrast, apoptosis in PUMA-KO cells was suppressed by family members, such as Bim and Mcl-1, also increased, more than 70% with little or no activation of caspases, whereas that of Bad, Bid, Bcl-2, and Bcl-xL remained mitochondrial membrane depolarization, or cytochrome c unchanged by 48 hours (Supplementary Fig. S1A). PUMA release at 48 hours (Fig. 2A–C). Annexin V/propidium induction was dose-dependent and stimulated by as little iodide staining conﬁrmed the apoptotic resistance of as 1 mmol/L crizotinib (Fig. 1C). To determine whether PUMA-KO cells (Supplementary Fig. S2A). Consistent MET regulates PUMA expression directly, we depleted with blocked apoptosis, PUMA-KO cells showed much MET by siRNA. MET knockdown led to increased PUMA improved clonogenic survival (Fig. 2D). We have shown mRNA and protein (Fig. 1D), but not that of Bim (Sup- previously that PUMA induces Bax-dependent apoptosis plementary Fig. S1B). Taken together, these data suggest (25, 29). As expected, BAX-KO HCT116 cells were also that MET inhibition selectively induces PUMA, whereas resistant to crizotinib-induced apoptosis (Fig. 2E and

www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 779

Zheng et al.

Figure 2. PUMA is required for the apoptotic activity of crizotinib. A, left, WT and PUMA-KO HCT116 cells were treated with 12 mmol/L PF02340166 and harvested at the indicated time points. Apoptosis was analyzed by a nuclear fragmentation assay., P < 0.01; , P < 0.001, WT versus PUMA-KO. Right, activation of caspase-3 and -9 was analyzed by Western blotting in WT and PUMA-KO HCT116 cells treated with 12 mmol/L PF02341066 for 48 hours. b-Actin was used as a loading control. B, WT and PUMA-KO HCT116 cells treated with 12 mmol/L PF02341066 for 24 hours were stained with MitoTracker Red CMXRos, and mitochondrial membrane potential was measured by ﬂow cytometry. C, the cytoplasm and mitochondria were fractionated from WT and PUMA-KO HCT116 cells treated with 12 mmol/L PF02341066 for 36 hours. The distribution of cytochrome c (Cyt c) was analyzed by Western blotting. Tubulin and cytochrome oxidase subunit IV (COX IV) were probed as cytoplamic and mitochondrial fraction controls. D, WT and PUMA-KO HCT116 cells treated with 12 mmol/L PF02340166 for 30 hours were subjected to colony formation assays as described in Materials and Methods. Colony numbers were scored 14 days later after plating. Representative pictures of colonies (left) and quantiﬁcation of colony numbers (right) are shown. E, HCT116 cells and BAX-KO HCT116 cells were treated with 12 mmol/L PF02341066 for 48 hours. Apoptosis was determined by a nuclear fragmentation assay. , P < 0.001, WT versus PUMA-KO.

Supplementary Fig. S2B). Transient PUMA knockdown damage (18). Earlier work suggested that other small- with siRNA also led to the reduced apoptosis in HCT116 molecule MET inhibitors can cause DNA double-strand and LoVo cells following crizotinib treatment (Supple- breaks (37, 38). We found that crizotinib treatment mentary Fig. S2C and S2D). Despite Bim induction, Bim resulted in increased phosphorylation of p53 and knockdown by siRNA did not render HCT116 cells resis- H2AX,aswellasp53stabilizationinHCT116,LoVo, tant to crizotinib (Supplementary Fig. S1B and S2E). and RKO cells, all with WT p53 (Fig. 3A and Supple- Collectively, these results show an essential role of PUMA mentary Fig. S3A). Induction of PUMA,andBIM to a and the mitochondrial pathway in crizotinib-induced lesser extent but not MCL-1, was observed in HCT 116 apoptosis in colon cancer cells. cells and reduced in p53-KO HCT116 cells (Fig. 3B). Furthermore, p53 deﬁciency attenuated crizotinib- p53-Dependent induction of PUMA by crizotinib induced apoptosis and caspase activation in HCT116 Both HCT116 and LoVo cells contain WT p53 gene, cells (Fig. 3C and Supplementary Fig. S3B), as well as in and p53 mediates PUMA induction following DNA RKO cells (Supplementary Fig. S3C and S3D). In time

780 Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics

PUMA and Crizotinib-Induced Apoptosis

Figure 3. PUMA induction by crizotinib is largely dependent on p53. A, HCT116 cells were treated with 12 mmol/L PF02340166 for indicated time. The expression levels of p-H2AX (S139), p-p53 (S15), total p53, and PUMA were analyzed by Western blotting. B, PUMA mRNA levels in WT and p53-KO HCT116 cells treated with 12 mmol/L PF02341066 at indicated times were determined by real-time RT-PCR. b-Actin was used as the normalized control., P < 0.001; WT versus p53- KO. C, WT and p53-KO HCT116 cells were treated with 12 mmol/L PF02340166 and harvested at indicated time points. Apoptosis was analyzed by a nuclear fragmentation assay. , P < 0.001; WT versus p53- KO. D, PUMA and Bim protein levels in WT and p53-KO HCT116 cells treated with 12 mmol/L PF02341066 at the indicated times were analyzed by Western blotting. E, HCT116 cells were transfected with the indicated reporters for 24 hours and subjected to 8 mmol/L PF02341066 treatment for 24 hours. The ratios of normalized relative luciferase activities (to the empty vector pBV-Luc as 1) were plotted. b-Actin was used as a loading control for Western blot analyses.

course experiments, PUMA and Bim were induced in p53-Independent induction of PUMA by crizotinib HCT116 p53-KO or RKO p53-KO cells after 24 hours, To further probe p53-independent induction of though at much lower levels compared with WT cells PUMA, we treated HCT116 cells harboring a deletion (Fig. 3D and Supplementary Fig. S3D). To further inves- of 2 p53-binding sites in the PUMA promoter (BS-KO; tigate p53-dependent activation of PUMA, we used a ref. 27) with crizotinib, and found much reduced PUMA series of PUMA promoter luciferase reporters (within induction compared with WT HCT116 cells (Fig. 4A). 2 kb; ref. 20), and found that the reporters containing the Moreover, PUMA, but not Bim, was induced by crizo- 2 p53-binding sites, such as fragments "A", "abc," and tinib in 2 p53-mutant colon cancer cell lines DLD1 and "E" had high activities after crizotinib treatment (Fig. HT29 in the absence of p53 phosphorylation or stabili- 3E). Notably, the most proximal fragment "de" ( 200 zation (Fig. 4B). Other transcription factors, such as NF- bp), lacking 2 p53-binding sites, still had a moderate kB subunit p65 and FoxO3a can bind to respective sites activity (Fig. 3E). The higher activity of the shorter located promixal to p53-binding sites in the PUMA fragment "abc" may be due to the removal of GC-rich promoter. Despite phosphorylation changes associated sequences and binding sites for transcriptional repres- with activation of p65 and FoxO3a, p65 or FoxO3a sors (18). Together, these data suggest that PUMA knockdown did not affect PUMA induction following inductionbycrizotinibismediatedbybothp53-depen- crizotinib treatment (Supplementary Fig. S4). Crizotinib dent and -independent mechanisms, and p53 is the also induced apoptosis in both DLD1 and HT29 cells, major transcriptional activator of PUMA in WT p53 which was suppressed by PUMA gene ablation or colon cancer cells. siRNA (Fig. 4C and 4D). DLD1 cells showed lower

www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 781

Zheng et al.

Figure 4. Crizotinib induces PUMA- and p53-independent apoptosis in colon cancer cells. A, PUMA protein levels in WT and BS-KO HCT116 cells treated with 12 mmol/L PF02341066 at the indicated times were analyzed by Western blotting. BS-KO HCT116 cells harbor the deletion of 2 p53-binding sites in the PUMA promoter. B, p53-mutant colon cancer cell lines DLD1 and HT29 were treated with 12 mmol/L PF02341066 for 24 hours. The expression levels of PUMA, Bim, p-p53, and total p53 were analyzed by Western blotting. C, top, PUMA expression was analyzed by Western blotting in WT and PUMA-KO DLD1 cells. Bottom, apoptosis was analyzed by nuclear fragmentation in WT and PUMA-KO DLD1 cells treated with the indicated doses of PF02341066 for 48 hours. , P < 0.001; WT versus p53-KO. D, HT29 cells were transfected with either a scrambled siRNA or PUMA siRNA for 24 hours and then treated with 12 mmol/L PF02341066 for 48 hours. Left, Western blotting conﬁrmed PUMA depletion by siRNA in HT29 cells. Right, apoptosis was determined by a nuclear fragmentation assay. , P < 0.001; si-PUMA versus Scrambled. b-Actin was used as a loading control for Western blot analyses.

PUMA induction and were more resistant to crizotinib- nation resulted in a strong PUMA-dependent synergy in induced apoptosis, compared with HT29 cells (Fig. 4B– cell killing (data not shown). D). These results indicate that PUMA plays a critical Crizotinib treatment decreased phosphorylation of role in the apoptotic responses to crizotinib in both p53 both AKT and ERK in HCT116 cells. However, ERK WT and mutant colon cancer cells. dephosphorylation was only transient and restored after 6 hours, long before apoptosis was detected (Fig. 1A). We Crizotinib synergizes with gefitinib or sorafenib to reasoned that a more durable inhibition of ERK phos- induce apoptosis via PUMA phorylation may potentiate crizotinib in cell killing. The Cooperative signaling of MET and the EGF receptor combination of crizotinib and sorafenib, a Raf inhibitor, (EGFR) can contribute to EGFR-TKI resistance (39). markedly induced apoptosis, expression of PUMA and HCT116 cells are highly resistant to gefitinib-induced Bim, and long-term growth suppression, compared with apoptosis or PUMA expression (Fig. 5A and B). We the single agent (Supplementary Fig. S5). Apoptosis and therefore hypothesized that the combination of gefitinib long-term growth suppression were attenuated in PUMA- and crizotinib may enhance apoptosis. Indeed, this com- KO cells (Supplementary Fig. S5A, S5B, and S5D). Of note, bination induced a much stronger induction of apoptosis, AKT phosphorylation was completely inhibited by all 3 caspase activation, and PUMA and Bim in HCT116 cells, agents, but not ERK phosphorylation, and reduction of compared with either agent alone (Fig. 5A–C). PUMA-KO either p-AKT or p-ERK is not sufficient for the enhanced cells were highly resistant to apoptosis and long-term induction of PUMA and Bim, or apoptosis in the combi- growth suppression induced by this combination (Fig. nation treatments (Fig. 5C and Supplementary S5C). In 5A, B, and D). Similarly, crizotinib and erlotinib combi- addition, Mcl-1 levels did not change or decreased

782 Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics

PUMA and Crizotinib-Induced Apoptosis

Figure 5. Crizotinib synergizes with gefitinib to induce PUMA-dependent apoptosis in colon cancer cells. A, WT and PUMA -KO HCT116 cells were treated with 6 mmol/L PF, 20 mmol/L gefitinib, or their combination for 48 hours. Apoptosis was determined by a nuclear fragmentation assay. , P < 0.001; combination versus single agent in WT cells and KO versus WT in combination. Right, chemical structure of gefitinib. B, caspase-3 activation was analyzed by Western blotting in WT and PUMA-KO HCT116 cells treated as in A. C, HCT116 cells were treated with 6 mmol/L PF02341066, 20 mmol/L gefitinib, or their combination for 24 hours. The expression levels of p-AKT, total AKT, p-ERK, total ERK, PUMA, and Bim were analyzed by Western blotting. D, WT and PUMA- KO HCT116 cells were treated with 6 mmol/L PF02341066, 20 mmol/L gefitinib, or their combination for 30 hours and were then subjected to a colony formation assay as described in Materials and Methods. Colony numbers were scored 14 days after plating. Representative pictures of colonies (top) and relative colonogenic survival (bottom) of WT and PUMA-KO HCT116 cells compared with untreated cells are shown. , P < 0.001; combination versus single agent in WT cells and KO versus WT in combination. b-Actin was used as a loading control for Western blot analyses.

following crizotinib combination with gefitinib or sorafi- Bim expression was evident in WT tumors at day 5 of nib (Fig. 5C and Supplementary Fig. S5C). These results treatment (Fig. 6C). Terminal deoxynucleotidyl transfer- suggest that the combinations of crizotinib with addition- ase–mediated dUTP nick end labeling (TUNEL) and al TKIs are required to effectively target nonoverlapping active caspase-3 staining revealed marked apoptosis in survival pathways in cancer cells to induce apoptosis. WT tumor tissues from crizotinib-treated mice, which decreased by more than 60% in the PUMA-KO tumors PUMA contributes to the antitumor activity of (Fig. 6D and E). These results show that the antitumor crizotinib in a mouse xenograft model activity of crizotinib in vivo is also dependent on the p53/ To determine whether PUMA-mediated apoptosis PUMA axis. plays a critical role in the antitumor activity of crizotinib in vivo, we established WT and PUMA-KO HCT116 xeno- Discussion graft tumors in the flanks of BALB/c (nu/nu) nude mice. Aberrantly activated oncogenic kinases are promising Tumors of different genotypes were established on the drug targets for small molecules; however, biomarkers opposite flanks of the same mice to minimize intermouse and resistance mechanisms of most clinically useful variations in tumor uptake or drug delivery. Crizotinib kinase inhibitors remain largely unknown. Crizotinib has was administered orally to tumor-bearing mice daily for been approved in ALK-rearranged NSCLC (3, 13), and 10 days, and tumor volumes were monitored every 2 days clinical interest is expanding to other solid tumors with for 3 weeks. Compared with the buffer, crizotinib reduced genetic alterations in c-MET, ALK, and ROS-1 (14, 40). growth of WT tumors by 81%, and PUMA-KO tumors by The antitumor activities of crizotinib include the induc- 41% (Fig. 6A and B). Increased phospho-p53, PUMA, and tion of cell-cycle arrest, apoptosis, and inhibition of cell

www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 783

Zheng et al.

Figure 6. PUMA mediates the antitumor effects of crizotinib in a xenograft model. A, nude mice with established WT or PUMA-KO HCT116 xenografts were treated with 35 mg/kg PF02341066 or buffer for 10 consecutive days. Tumor volume at the indicated time-points was calculated and plotted. Arrows indicate PF02341066 administration. Statistical signiﬁcance is indicated for the comparison of PF02341066-treated WT and PUMA-KO tumors. B, representative tumors at the end of the experiment in A. C, WT and PUMA-KO HCT116 xenograft tumors were harvested from 2 mice as in A after 5 treatments. p-p53, PUMA, and Bim expression in representative WT HCT116 tumors was analyzed by Western blotting. b-Actin was used as a loading control. Representative data are shown. D, TUNEL and active caspase-3 staining of tumor sections as harvested in C. E, quantitation of TUNEL or active caspase-3 positive cells in D. Results of D and E were expressed as means SD (N ¼ 3) tumors of each genotype.

proliferation and invasion (3, 13). Our results indicated crizotinib effectively inhibits MET signaling, and induces that activation of mitochondrial pathway and PUMA PUMA-dependent apoptosis in colon cancer cells. Sur- plays a key role in crizotinib-induced cancer cell death prisingly, MET siRNA did not induce obvious apoptosis in vitro and in vivo. In addition, the combinations of or p53 stabilization in HCT116 cells but only a modest crizotinib with EGFR or Raf inhibitors resulted in the PUMA induction. One possible explanation is that crizo- potent induction of apoptosis in colon cancer cells via tinib treatment is more effective in blocking MET signal- PUMA. ing than MET siRNA. Another more likely explanation is Crizotinib-induced apoptosis can be attributed to Bim that crizotinib inhibits other RTKs and activates p53- in lung cancer cells with MET amplification but not with dependent DNA damage responses in some cells. In MET mutations (41, 42). In colon cancer cells, Bim induc- addition, PUMA induction may further engage cyto- tion, though not required, likely potentiates PUMA- plasmic function of p53 to trigger apoptosis (43, 44). These dependent apoptosis by antagonizing antiapoptotic Bcl- issues might be relevant as the clinically efficacious con- 2 family proteins including Mcl-1, whose levels remain centration of crizotinib used in vitro for cell killing (1–10 high after treatment. It is likely that multiple BH3-only mmol/L) are much higher than those required to selec- proteins are used in the apoptotic response to crizotinib, tively inhibit MET or ALK. whereas a distinct member may serve as cell or tissue- Crizotinib and other TKI combinations enhance apo- specific initiator. The coregulation of Bim and PUMA by ptosis and the induction of PUMA and Bim. Interestingly, crizotinib is interesting and somewhat unexpected, and the inhibition of AKT phosphorylation is not sufficient for requires further investigation. Bim and PUMA are located cell killing or strong PUMA induction. Our data are on different chromosomes and their basal expression consistent with the emerging concept that cross-talk levels are quite different. It is possible that higher order between RTKs is a major mechanism for cancer progres- chromatin changes may be involved in addition to loading sion and therapeutic resistance, and successful therapy of stress-induced transcription factors, such as p53 onto will likely require targeting multiple survival pathways their promoters. The mechanisms of MET inhibitor- (1, 6, 45). However, several challenges are facing the induced DNA damage response (37, 38), and p53-inde- clinical applications of kinase inhibitors: (i) genetic altera- pendent induction of PUMA and Bim remain unclear. tions in EGFR, MET,orALK, and possibly ROS1 are Despite a plethora of oncogenic activities ascribed to infrequent, (ii) not all tumors with same alterations MET and its related kinases (3, 13), our data suggest that respond, (iii) tumor heterogeneity plus preexisting or de

784 Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics

PUMA and Crizotinib-Induced Apoptosis

novo mutations in targets can lead to rapid development of Disclosure of Potential Conflicts of Interest resistance. Therefore, it is important to identify effective No potential conflicts of interest were disclosed. combination therapies that target several survival Authors' Contributions mechanisms in cancer cells to prevent the development Conception and design: L. Zhang, J. Yu of resistance. Our data suggest that the combination of Development of methodology: X. Zheng, K. He, J. Yu crizotinib with other TKIs effectively block MET signaling Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): L. Zhang, J. Yu and alternative survival pathways in colon cancer cells. Analysis and interpretation of data (e.g., statistical analysis, biostatis- The cell lines used in this study are not expected to contain tics, computational analysis): X. Zheng, K. He, L. Zhang, J. Yu ALK or ROS1 rearrangements, whereas the mechanisms Writing, review, and/or revision of the manuscript: X. Zheng, K. He, J. Yu Administrative, technical, or material support (i.e., reporting or orga- described warrant further investigation in this context in nizing data, constructing databases): X. Zheng, K. He, L. Zhang, J. Yu relevant cancers. Study supervision: J. Yu In conclusion, our study provides a novel antitumor mechanism of crizotinib via PUMA-mediated apoptosis Acknowledgments The authors thank Bert Vogelstein (Howard Hughes Medical Institute, through both p53-dependent and -independent means. Johns Hopkins University) for p53-KO HCT116 and p53-KO RKO cells, and These findings are in line with other recent findings in other members of Yu and Zhang laboratories for helpful discussions. which PUMA or Bim mediates the apoptotic response to various kinase inhibitors. Therefore, induction of PUMA Grant Support This work is supported by NIH grant CA129829, American Cancer and Bim may be a predictor for a favorable response to Society grant RGS-10-124-01-CCE, FAMRI (J. Yu), and by NIH grants crizotinib, and possibly other targeted agents or their CA106348, CA121105, and American Cancer Society grant RSG-07-156- combinations. This concept is different from using genetic 01-CNE (L. Zhang). This project used the UPCI shared facilities that were supported in part by award P30CA047904. alterations or steady-state mRNA or protein levels in the The costs of publication of this article were defrayed in part by the tumors and measures a dynamic response that may well payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate be cell-type or patient-specific (36). Because induction of this fact. PUMA and apoptosis is significantly more robust in p53 WT cells, it would be useful to determine whether crizo- Received November 28, 2012; revised January 24, 2013; accepted tinib shows selectivity against p53 WT cancers. February 11, 2013; published OnlineFirst February 20, 2013.

References 1. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. 11. Bergethon K, Shaw AT, Ignatius Ou S-H, Katayama R, Lovly CM, Cell 2011;144:646–74. McDonald NT, et al. ROS1 rearrangements define a unique molecular 2. Gschwind A, Fischer OM, Ullrich A. The discovery of receptor class of lung cancers. J Clin Oncol 2012;30:863–70. tyrosine kinases: targets for cancer therapy. Nat Rev Cancer 12. ChristensenJG,ZouHY,ArangoME,LiQ,LeeJH,McDonnellSR, 2004;4:361–70. et al. Cytoreductive antitumor activity of PF-2341066, a novel 3. Comoglio PM, Giordano S, Trusolino L. Drug development of MET inhibitor of anaplastic lymphoma kinase and c-Met, in experimental inhibitors: targeting oncogene addiction and expedience. Nat Rev models of anaplastic large-cell lymphoma. Mol Cancer Ther 2007;6: Drug Discov 2008;7:504–16. 3314–22. 4. Di Renzo MF, Olivero M, Giacomini A, Porte H, Chastre E, Mirossay L, 13. Ou SH. Crizotinib: a novel and first-in-class multitargeted tyrosine et al. Overexpression and amplification of the met/HGF receptor gene kinase inhibitor for the treatment of anaplastic lymphoma kinase during the progression of colorectal cancer. Clin Cancer Res rearranged non-small cell lung cancer and beyond. Drug Des Devel 1995;1:147–54. Ther 2011;5:471–85. 5. Takeuchi H, Bilchik A, Saha S, Turner R, Wiese D, Tanaka M, et al. c- 14. Stumpfova M, Janne€ PA. Zeroing in on ROS1 rearrangements in non- MET expression level in primary colon cancer. Clin Cancer Res small cell lung cancer. Clin Cancer Res 2012;18:4222–4. 2003;9:1480–8. 15. Adams JM, Cory S. The Bcl-2 apoptotic switch in cancer development 6. Hait WN, Hambley TW. Targeted cancer therapeutics. Cancer Res and therapy. Oncogene 2007;26:1324–37. 2009;69:1263–7. 16. Yu J, Zhang L. Apoptosis in human cancer cells. Curr Opin Oncol 7. Jin H, Yang R, Zheng Z, Romero M, Ross J, Bou-Reslan H, et al. 2004;16:19–24. MetMAb, the one-armed 5D5 anti-c-Met antibody, inhibits orthotopic 17. Leibowitz B, Yu J. Mitochondrial signaling in cell death via the Bcl-2 pancreatic tumor growth and improves survival. Cancer Res family. Cancer Biol Ther 2010;9:417–22. 2008;68:4360–8. 18. Yu J, Zhang L. PUMA, a potent killer with or without p53. Oncogene 8. Munshi N, Jeay Sb, Li Y, Chen C-R, France DS, Ashwell MA, et al. ARQ 2008;27:S71–S83. 197, a novel and selective inhibitor of the human c-Met receptor 19. Melino G, Bernassola F, Ranalli M, Yee K, Zong WX, Corazzari M, et al. tyrosine kinase with antitumor activity. Mol Cancer Ther 2010;9: p73 Induces apoptosis via PUMA transactivation and Bax mitochon- 1544–53. drial translocation. J Biol Chem 2004;279:8076–83. 9. Liu X, Wang Q, Yang G, Marando C, Koblish HK, Hall LM, et al. A novel 20. Ming L, Sakaida T, Yue W, Jha A, Zhang L, Yu J. Sp1 and p73 kinase inhibitor, INCB28060, blocks c-MET-dependent signaling, neo- activate PUMA following serum starvation. Carcinogenesis 2008; plastic activities, and cross-talk with EGFR and HER-3. Clin Cancer 29:1878–84. Res 2011;17:7127–38. 21. You H, Pellegrini M, Tsuchihara K, Yamamoto K, Hacker G, Erlacher M, 10. Zou HY, Li Q, Lee JH, Arango ME, McDonnell SR, Yamazaki S, et al. FOXO3a-dependent regulation of Puma in response to cytokine/ et al. An orally available small-molecule inhibitor of c-Met, PF- growth factor withdrawal. J Exp Med 2006;203:1657–63. 2341066, exhibits cytoreductive antitumor efficacy through anti- 22. Dudgeon C, Wang P, Sun X, Peng R, Sun Q, Yu J, et al. PUMA induction proliferative and antiangiogenic mechanisms. Cancer Res 2007;67: by FoxO3a mediates the anticancer activities of the broad-range 4408–17. kinase inhibitor UCN-01. Mol Cancer Ther 2010;9:2893–902.

www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 785

Zheng et al.

23. Wang P, Qiu W, Dudgeon C, Liu H, Huang C, Zambetti GP, et al. PUMA 35. Leibowitz BJ, Qiu W, Liu H, Cheng T, Zhang L, Yu J. Uncoupling p53 is directly activated by NF-[kappa]B and contributes to TNF-[alpha]- functions in radiation-induced intestinal damage via PUMA and p21. induced apoptosis. Cell Death Differ 2009;16:1192–202. Mol Cancer Res 2011;9:616–25. 24. Dudgeon C, Peng R, Wang P, Sebastiani A, Yu J, Zhang L. Inhibiting 36. Sun J, Sun Q, Brown MF, Dudgeon C, Chandler J, Xu X, et al. The multi- oncogenic signaling by sorafenib activates PUMA via GSK3b and NF- targeted kinase inhibitor sunitinib induces apoptosis in colon cancer kB to suppress tumor cell growth. Oncogene 2012;31:4848–58. cells via PUMA. PLoS ONE 2012;7:e43158. 25. Yu J, Wang Z, Kinzler KW, Vogelstein B, Zhang L. PUMA mediates the 37. Medova M, Aebersold DM, Blank-Liss W, Streit B, Medo M, Aebi S, apoptotic response to p53 in colorectal cancer cells. Proc Natl Acad et al. MET inhibition results in DNA breaks and synergistically sensi- Sci U S A 2003;100:1931–6. tizes tumor cells to DNA-damaging agents potentially by breaching a 26. Bunz F, Hwang PM, Torrance C, Waldman T, Zhang Y, Dillehay L, et al. damage-induced checkpoint arrest. Genes Cancer 2010;1:1053–62. Disruption of p53 in human cancer cells alters the responses to 38. Ganapathipillai SS, Medova M, Aebersold DM, Manley PW, Berthou S, therapeutic agents. J Clin Invest 1999;104:263–9. Streit B, et al. Coupling of mutated Met variants to DNA repair via Abl 27. Wang P, Yu J, Zhang L. The nuclear function of p53 is required for and Rad51. Cancer Res 2008;68:5769–77. PUMA-mediated apoptosis induced by DNA damage. Proc Natl Acad 39. Engelman JA, Zejnullahu K, Mitsudomi T, Song Y, Hyland C, Park JO, Sci U S A 2007;104:4054–9. et al. MET amplification leads to gefitinib resistance in lung cancer by 28. Bunz F. Human cell knockouts. Curr Opin Oncol 2002;14:73–8. activating ERBB3 signaling. Science 2007;316:1039–43. 29. Ming L, Wang P, Bank A, Yu J, Zhang L. PUMA dissociates Bax and 40. Yauch RL, Settleman J. Recent advances in pathway-targeted cancer Bcl-XL to induce apoptosis in colon cancer cells. J Biol Chem 2006; drug therapies emerging from cancer genome analysis. Curr Opin 281:16034–42. Genet Dev 2012;22:45–9. 30. Fornari F, Milazzo M, Chieco P, Negrini M, Calin GA, Grazi GL, et al. 41. Okamoto W, Okamoto I, Arao T, Kuwata K, Hatashita E, Yamaguchi H, MiR-199a-3p regulates mTOR and c-Met to influence the doxorubicin et al. Antitumor action of the MET tyrosine kinase inhibitor crizotinib sensitivity of human hepatocarcinoma cells. Cancer Res 2010;70: (PF-02341066) in gastric cancer positive for MET amplification. Mol 5184–93. Cancer Ther 2012;11:1557–64. 31. Han J, Goldstein LA, Gastman BR, Rabinowich H. Interrelated roles for 42. Tanizaki J, Okamoto I, Okamoto K, Takezawa K, Kuwata K, Yamaguchi Mcl-1 and BIM in regulation of TRAIL-mediated mitochondrial apo- H, et al. MET tyrosine kinase inhibitor crizotinib (PF-02341066) shows ptosis. J Biol Chem 2006;281:10153–63. differential antitumor effects in non-small cell lung cancer according to 32. Yu J, Wang P, Ming L, Wood MA, Zhang L. SMAC/Diablo mediates the MET alterations. J Thorac Oncol 2011;6:1624–31. proapoptotic function of PUMA by regulating PUMA-induced mito- 43. Chipuk JE, Bouchier-Hayes L, Kuwana T, Newmeyer DD, Green DR. chondrial events. Oncogene 2007;26:4189–98. PUMA couples the nuclear and cytoplasmic proapoptotic function of 33. Yu J, Zhang L, Hwang PM, Kinzler KW, Vogelstein B. PUMA induces p53. Science 2005;309:1732–5. the rapid apoptosis of colorectal cancer cells. Mol Cell 2001;7: 44. Chipuk JE, Green DR. How do BCL-2 proteins induce mitochondrial 673–82. outer membrane permeabilization? Trends Cell Biol 2008;18:157–64. 34. Qiu W, Carson-Walter EB, Liu H, Epperly M, Greenberger JS, Zambetti 45. Dent P, Curiel DT, Fisher PB, Grant S. Synergistic combinations of GP, et al. PUMA regulates intestinal progenitor cell radiosensitivity and signaling pathway inhibitors: mechanisms for improved cancer gastrointestinal syndrome. Cell Stem Cell 2008;2:576–83. therapy. Drug Resist Updat 2009;12:65–73.

786 Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics

Crizotinib Induces PUMA-Dependent Apoptosis in Colon Cancer Cells

Xingnan Zheng, Kan He, Lin Zhang, et al.

Mol Cancer Ther 2013;12:777-786. Published OnlineFirst February 20, 2013.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-12-1146

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2013/02/26/1535-7163.MCT-12-1146.DC1

Cited articles This article cites 45 articles, 23 of which you can access for free at: http://mct.aacrjournals.org/content/12/5/777.full#ref-list-1

Citing articles This article has been cited by 4 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/12/5/777.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/12/5/777. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.