MK-1775, a Potent Wee1 Inhibitor, Synergizes with Gemcitabine to Achieve Tumor Regressions, Selectively in p53-Deficient Pancreatic Cancer Xenografts

This article is featured in Highlights of This Issue, p. 2603

Pancreatic ductal adenocarcinoma (PDA) remains one of the least curable cancers, ranking fourth in cancer related deaths in the United States (1). Worldwide, PDA contributes to more than 230,000 deaths annually (2). Gemcitabine (GEM), a pyrimidine antimetabolite, is the accepted standard treatment for patients with advanced and metastatic PDA, but the benefit is small (3). The addition of other chemotherapeutics or monoclonal antibodies or radiation to GEM has not resulted any meaningful improvement in the survival of PDA patients (4–6). As the current therapies offer very limited benefit, there is significant unmet medical need to identify novel molecular targets and develop novel therapeutic strategies to treat this devastating disease (7).

Treatment efficacy of DNA damaging agents is determined not only by the amount of therapy-induced DNA damage but also by the capacity of tumor cells to repair the damaged DNA. Many of the conventional anticancer treatments including antimetabolites, ionizing radiation, alkylating agents, platinum compounds, and DNA topoisomerase inhibitors exert their antitumor effects by damaging the DNA in tumor cells which leads to apoptosis (8). However, when cells are treated with DNA damaging agents, multiple checkpoints are activated, including G₁, intra-S, and G₂/M, leading to cell cycle arrest, thus providing time for the cell to repair the damage and to evade apoptosis before resuming the cell cycle (9–11). Tumor cells can exploit these repair mechanisms in response to DNA damaging chemotherapeutics, rendering tumors refractory to current therapeutic interventions. Therefore, abrogation of checkpoints function may drive tumor cells toward apoptosis and enhance the efficacy of oncotherapy.

p53 is a key regulator of the G₁ checkpoint and is one of the most frequently mutated genes in cancer. Since G₁ checkpoint is frequently compromised due to loss-of-function mutation of p53 gene in 50% to 70% of all cancers, the G₂/M checkpoint plays a pivotal role in preventing the programmed cell death in p53-deficient tumors (12, 13). Hence, inhibitors of the G₂ checkpoint can selectively sensitize cancer cells with deficient p53 to killing by DNA-damaging anticancer agents while sparing normal tissue from toxicity.

The G₂ DNA damage checkpoint ensures maintenance of cell viability by delaying progression into mitosis when cells have suffered DNA damage. Wee1 is a cellular protein kinase which inhibits Cdc2 activity, thereby preventing cells from proceeding through mitosis by maintaining G₂ arrest (14). Wee1 reversibly arrests the cell cycle by inactivating Cdc2 through phosphorylation at Tyr-15 (15). Disruption of this phosphorylation site abrogates checkpoint-mediated regulation of Cdc2 (16). Wee1 knockdown with siRNA has been reported to abrogate the G₂ DNA damage checkpoint arrest and to sensitize cancer cells to DNA damaging agents (17). Wee1 kinase inhibition is expected to potentiate the antitumor effect of DNA damaging chemotherapeutics by overcoming the G₂ arrest and thereby promoting checkpoint bypass to facilitate the preferential killing of p53-deficient tumor cells via mitotic catastrophe.

MK-1775 is the first reported Wee1 inhibitor with high potency, selectivity, and oral bioavailability in preclinical animal models and is currently being evaluated in several phase I clinical trials (18–21). Given that p53 mutations are common in pancreatic cancer (22), we sought to investigate efficacy and pharmacodynamic effects of MK-1775 alone and in combination with GEM in pancreatic cancer xenografts with p53 wild-type or p53-deficient status. Our results provide compelling evidence that MK-1775 synergizes with GEM to achieve tumor regressions, selectively in p53-deficient pancreatic cancer xenografts. These results may help to frame clinical investigation of Wee1 inhibitors along with chemotherapy to benefit cancer patients whose tumor cells are devoid of functional p53 gene.

Animal experiments were conducted following approval and accordance with Animal Care and Use Committee guidelines of Johns Hopkins University. Fresh pancreatic cancer specimens resected from patients at the time of surgery, with informed written patient consent, were implanted subcutaneously into the flanks of 6-week-old female nu/nu athymic mice (Harlan). The patients had not undergone chemotherapy or radiation therapy before surgery. Grafted tumors were subsequently transplanted from mouse to mouse and maintained as a live PancXenoBank according to an Institutional Review Board (IRB) approved protocol (23). Tumor-specific mutations of protein-coding genes (exomic sequencing) in these xenografts have been recently reported (24). Most importantly, these xenografts were not placed in culture and appear to retain most of the genetic features of the original tumor, despite serial passing across several generations of mice (25, 26).

The Wee1 inhibitor, MK-1775, was provided by Merck Research Laboratories. GEM (Eli Lilly) was purchased from pharmacy.

Nine pancreatic cancer xenografts (6 with p53-deficient and 3 with p53 wild-type status) from PancXenoBank were allowed to grow separately on both flanks of athymic mice. When tumors reached a volume of approximately 200 mm³, mice were individually identified and randomly assigned to treatment groups, with 5 to 6 mice (8–10 evaluable tumors) in each group: (1) control; (2) MK-1775 (30 mg/kg. p.o., once daily for 4 weeks; 3) GEM (100 mg/kg, i.p., twice weekly on days 1 and 4) for 4 weeks; (4) GEM followed 24 hour later by MK-1775 in the above mentioned dose. Tumor growth was evaluated twice per week by measurement of 2 perpendicular diameters of tumors with a digital caliper. Individual tumor volumes were calculated as V = ab²/2, where a being the largest diameter, b the smallest. Relative tumor growth index (TGI) on day 28 was calculated using the formula: (mean tumor volume of drug-treated group/mean tumor volume of control group) × 100. Number of tumors that regressed more than 50% of its initial size in each xenograft was noted. Animals were sacrificed 1 hour after the last dose of GEM or MK-1775 and tumors were harvested for analysis except 3 mice each from GEM and combination treatment group, which were kept longer to check tumor regrowth after the treatment. Mice kept for the regrowth study were sacrificed when the tumors reached the size of control tumors in that xenograft.

Protein extracts were prepared from tumors according to previously published method (27). Briefly, tumors (50 mg) were minced on ice in prechilled lysis buffer containing protease cocktail (Roche Molecular Biochemicals). The minced tissue was homogenized and centrifuged at 16,000 × g at 4°C for 10 minutes. Protein content in the supernatants was measured with the Pierce Protein Assay kit using bovine serum albumin (BSA) as a standard. Protein lysates (30 μg) were boiled in Laemmli sample buffer (Bio-Rad Laboratories) resolved by electrophoresis on 4% to 12% Bis-Tris precast gels (Bio-Rad Laboratories), and transferred to Immobilon-P membranes (Millipore). Membranes were blocked at room temperature with 5% nonfat milk (Pierce) for 1 hour. Primary antibodies against Wee1, phospho-Wee1^Ser642, Cdc2 (p-Cdc2), phospho-Cdc2^Tyr15 (p-Cdc2^Tyr15), phospho-histone H3^Ser10 (p-HH3^Ser10), cleaved poly (ADP-ribose) polymerase^Asp214 (C-PARP^Asp214), phospho-H2AX^Ser139 (γ-H2AX), cellular inhibitor of apoptosis protein-2 (cIAP2) and cyclin B1 (Cell Signaling Technology) were diluted (1:1,000) in 5% BSA and incubated overnight at 4°C with mild rocking. Membranes were probed with secondary anti-rabbit IgG horseradish peroxidase (HRP)-conjugated antibody (GE Healthcare) at a final dilution of 1:2,000 for 2 hours. After washing 3 times with TBST (Tris-buffered saline and Tween 20), bound antibodies were detected by enhanced chemiluminescence (GE Healthcare) or Supersignal West Femto (Pierce). Equal loading was verified with β-actin antibody.

Sections were prepared from formalin-fixed, paraffin-embedded tumor samples. To ensure uniform handling of samples, all sections were processed simultaneously. Sections were deparaffinized by routine techniques and were quenched with 3% H₂O₂ for 10 minutes. The slides were steamed in citrate buffer (pH 6.0) for 20 minutes at 90°ºC+ and were blocked with 10% FBS solution (Invitrogen) for 30 minutes prior to incubation with primary antibodies. For the p-Cdc2 staining, p-Cdc2^Tyr15 antibody (Cell Signaling Technology) was used at a dilution of 1:200 followed by 1 hour incubation. For p-HH3 detection, p-HH3^Ser10 antibody (Cell Signaling Technology) was used at a 1:250 dilution and incubated overnight at 4°ºC. Sections were incubated with EnVision/HRP rabbit antibody for 30 minutes and diaminobenzidine (DAB+) chromogen for 10 minutes. The sections were counterstained with hematoxylin, dehydrated, and mounted. Slides were evaluated by a pathologist under 20X objective and digitized using a color camera mounted to the microscope. p-Cdc2^Tyr15 and p-HH3^Ser10 stained tumor cells were counted over total number of tumor cells and two sided Fisher's exact test (GraphPad Prism 5.03 software) was used for significance calculation.

All error bars are represented as the standard error of the mean. Significance was analyzed using unpaired Student's t-test. The differences were considered significant when P value was less than 0.05.

The principal aim of this study was to investigate the efficacy of MK-1775 as a single agent and in combination with GEM in PDA xenografts and to assess whether the status of the p53 gene had any role in dictating treatment efficacy. We used a xenograft model which is freshly generated from the tumors taken from pancreatic cancer patients and selected nine xenografts (6 xenografts with p53-deficient and 3 xenografts with p53 wild-type status) for this study. As shown in Figure 1A, tumors in vehicle-treated animals grew rapidly. Single agent MK-1775 treatment produced greater than 50% inhibition of tumor growth in 2 xenografts (PANC286 and PANC198). However, 5 of 9 xenografts treated with GEM and 6 of 9 xenografts treated with GEM plus MK-1775 produced complete tumor growth inhibition and in fact resulted tumor shrinkage compared to control and MK-1775 treated animals (Fig. 1A). These data suggest that single agent MK-1775 is unlikely to be effective in patients with PDA but that the combination of this agent with GEM has a substantial level of activity, and should be prioritized for clinical development.

Overall, none of the xenografts with p53-deficient status in the GEM treatment group produced 50% regression of initial tumor volume (Fig. 1A). However, combination of GEM and MK-1775 resulted in greater than 50% regression of initial tumor volume in 4 of 6 xenografts (66.66%) with p53-deficient status (Fig. 1A). The number of tumors that regressed more than 50% of its initial tumor size in each xenograft upon completion of treatment is provided in Table 1. Tumors with wild-type p53 status did not regress with treatments (Table 1). Among the xenografts with p53-deficient status, GEM alone treatment-induced regression in 7 of 55 tumors (12.72%), while MK-1775 in combination with GEM induced regressions in 25 of 49 tumors (51.10%) in the 6 xenografts. Tumor growth regressions in GEM plus MK-1775 treated mice were found to be significant in PANC198 (P < 0.0001), PANC215 (P < 0.005), and PANC185 (P < 0.005) as compared to GEM-treated mice. There was an overall 4.01-fold increase in total number of tumors regressed in the combination treatment group compared to GEM alone treatment (Fig. 1B). K-ras and SMAD4 status do not influence the tumor regression pattern in the xenografts (Table 1). One limitation of preclinical studies is that the threshold of activity that translates into positive clinical outcome is not known. Often, drugs are selected for clinical development based on tumor growth inhibition in preclinical models. As our experience with freshly generated PDA models increases and more comparison data is available, we are observing that indeed only agents that result in marked tumor regressions in this model have the potential to impact patient outcome. This is illustrated by our recent work on AZD0530 and nab-paclitaxel. AZD0530, a Src kinase inhibitor, induced only modest inhibition of tumor growth in PDA xenografts and, as expected, failed in a phase II clinical trial (27, 28). In contrast, nab-paclitaxel, in combination with GEM, resulted in marked tumor regression in this model, which successfully predicted a positive phase II study (29, 30).

The selective augmentation of antitumor effects in tumors with deficient-p53 was anticipated based on the mechanism of action of the agent. Mammalian cells undergo cell cycle arrest in response to DNA damage due to the existence of multiple checkpoint response mechanisms. In response to DNA damage, the cell cycle halts, preventing the propagation of cells with damaged DNA. DNA damage culminates in the enforcement of cell cycle arrest, mainly at G₁ and G₂ phases. Checkpoint pathways operating at the G1 phase are frequently lost in cancer cells due to mutation of the p53 tumor suppressor gene. Cells lacking functional p53 would not be anticipated to arrest at the G₁ checkpoint and would depend on the G₂ checkpoint to permit DNA repair prior to undergoing mitosis. Thus, G₂ checkpoint abrogation should preferentially kill p53-deficient cancer cells by removing the only checkpoint that protects these cells from premature entry into mitosis in response to DNA damage. Our data strongly suggest that the clinical development of MK-1775 with GEM should be restricted to patients with p53-deficient PDA.

Cdc2 initiates mitosis, which is the ultimate target of DNA replication and repair checkpoints. Chk1, Chk2, Wee1, and Myt1 are key regulators of G₂ checkpoint, which act directly or indirectly to inhibit Cdc2 activity (17, 31). Chk1 and Chk2 are downstream effectors of ataxia telangiectasia-mutated kinase (ATM) and ataxia telangiectasia and Rad3-related kinase (ATR), which induce G₂/M cell cycle arrest by inactivating Cdc25 tyrosine phosphatases through phosphorylation (32). Both Chk1 and Chk2 are known to phosphorylate Cdc25 on Ser216 and this phosphorylation makes Cdc25 functionally inactive (33). Cdc25 is required for removal of inhibitory phosphotyrosines on Cdc2/cyclin B1 kinase complexes that mediate entry into mitosis. On the other hand, the inhibitory phosphorylations at Thr-14 and Tyr-15 sites of Cdc2 are mediated by Myt1 and Wee1 kinases (34, 35). Wee1 is the major kinase phosphorylating the Tyr-15 site and Wee1-dependent phosphorylation of Cdc2 maintains the Cdc2/cyclin B1 complex in an inert form. While Myt1 preferentially phosphorylates the Thr-14 site, it can also phosphorylate the Tyr-15 site. Thus either Cdc25 inactivation and/or Wee1/Myt1 activation could contribute to G₂ cell cycle arrest in response to DNA damage.

Chk1/2 inhibitors are in clinical development (36–38). A recent report indicated that Chk1 is required to maintain genome integrity and cell viability, and that p53 wild-type cells are no less sensitive than p53-deficient cells to Chk1 inhibition in the presence of DNA damage. Thus, combining Chk1 inhibition with DNA damaging agents does not lead to preferential killing of p53-deficient over p53 wild-type cells, and inhibiting Chk1 does not appear to be a promising approach for potentiation of cancer chemotherapy (39). Here we showed that Wee1 inhibition by MK-1775 could potentiate GEM sensitivity and tumor regressions, selectively in p53-deficient pancreatic cancer xenografts.

We were also interested in long-term tumor growth control and followed 3 xenografts after treatment for an extended period of time. Tumor regrowth data, as shown in Figure 2 suggest that not only does the combination of GEM with MK-1775 lead to synergistic tumor growth inhibition, but the effect of the combination therapy is also longer-lasting than that seen with GEM alone (Fig. 2A, B, and C). It was noteworthy, however, that tumors eventually recur, albeit at a slower pace.

In order to determine the target modulation by MK-1775, we examined Wee1, Cdc2, and their phosphorylated forms in posttreatment tumor specimens. MK-1775 treatment strongly inhibited phosphorylation of Tyr-15 of Cdc2, the primary substrate of Wee1 (Figs. 3 and 4A and C), suggesting increased Cdc2 kinase activity. In addition, the Wee1 protein was consistently reduced by MK-1775 treatment as shown by Western blotting (Fig. 3), likely due to degradation of Wee1 as MK-1775 treatment activates Cdc2 which in turn phosphorylates Wee1, ultimately leading to its ubiquitin-proteasome–dependent destruction (40).

To determine whether combination therapy promotes mitotic entry, we measured the expression of p-HH3 by Western blotting as well as by immunohistochemistry. When administered in combination with GEM, MK-1775 promoted mitotic entry as measured by enhanced p-HH3^Ser10 expression (Fig. 3, 4B and D). In addition, the combined treatment resulted in the upregulation of C-PARP as well as downregulation of cIAP2, suggesting that combination therapy facilitates apoptotic death of tumor cells (Fig. 3). GEM, as a chain terminator, requires an active cell cycle to be effective for inhibiting tumor growth, and might induce cell cycle halt and enforce cell cycle checkpoints, which may play an important role in escalating the resistance to therapy. Thus, there is a strong rationale in combining checkpoint inhibitors with GEM as a means to enhance tumor response (41, 42). Here we showed that GEM induces G₂ arrest, which correlates with an increased Cdc2 inhibitory phosphorylation at Tyr-15 and prevents mitotic entry as evidenced by decreased p-HH3^Ser10 (Fig. 4B and D). However, the decreased Cdc2 inhibitory phosphorylation at Tyr-15 caused by MK-1775 treatment indicates that MK-1775 has the ability to abrogate the G₂ arrest induced by GEM and promote mitotic entry as demonstrated by enhanced p-HH3^Ser10 (Fig. 4B and D). Cyclin B1 was examined as a marker of G2 phase (43). Expression of Cdc2 was not altered by treatments, while the expression of cyclin B1 was strongly inhibited by MK-1775 as well as combination of MK-1775 and GEM treatment compared to control and GEM-treated tumors of PANC198 (Fig. 3). Loss of cyclin B1 accumulation in the MK-1775 as well as combination of MK-1775 and GEM-treated tumors indicate the exit from G2 phase arrest (Fig. 3). The levels of γ-H2AX were used as a surrogate for unrepaired DNA damage (44). γ-H2AX expression was clearly elevated in the combination of MK-1775 and GEM treatment group compared to GEM-treated tumors of PANC198, indicating the persistence of unrepaired DNA damage in the tumors (Fig. 3). Overall, in addition to providing mechanistic support to the observations made above, the data provides important clues for potential biomarkers for clinical development of this drug combination.

In conclusion, our results provide compelling evidence that MK-1775 treatment leads to the inhibition and subsequent loss of Wee1 and activation of its substrate, Cdc2. The MK-1775 and GEM combination promoted the mitotic entry of tumor cells and eventually led to apoptotic death, and delayed the tumor progression compared to the GEM treatment. These findings have important clinical implications and raise the hope for potential therapeutic benefit to many PDA patients whose cancer cells are deficient for p53 function.

M. Hidalgo received funding from Merck Research Laboratories. J. Watters, D. Brooks, T. Demuth, S.D. Shumway, S. Mizuarai, and H. Hirai are employees of Merck. The other authors declare no potential conflicts of interest.

The study was supported by Merck Research Laboratories as well as NIH grants CA116554 and CA129963 to M. Hidalgo.

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.