Preclinical Evaluation of the WEE1 Inhibitor MK-1775 as Single-Agent Anticancer Therapy Free

Many commonly used anticancer drugs target DNA in dividing cells and ultimately cause DNA damage. This, in turn, triggers activation of cell-cycle checkpoints that arrest progression of the cell cycle (at the G₁, S, or G₂–M phases) to allow the DNA to be repaired before the cell undergoes DNA replication and/or division. From a therapeutic standpoint, inhibition of checkpoint kinases that mediate cell-cycle arrest could force tumor cells to continue cell division before chemically induced DNA damage is repaired, eventually causing apoptosis or mitotic catastrophe (1). Cell line studies support this hypothesis and show chemosensitization and radiosensitization by pharmacologic or genetic disruption of checkpoint kinase activity including CHK1, WEE1, ATR, ATM, and MK2. Inhibitors of these kinases are at various stages of preclinical and clinical development.

The checkpoint kinase WEE1 catalyzes an inhibitory phosphorylation of both CDK1 (CDC2) and CDK2 on tyrosine 15 (2, 3). WEE1-dependent inhibition of CDK1 and CDK2 arrests the cell cycle in response to extrinsically induced DNA damage (4). WEE1 activity is also essential for the unperturbed cell cycle (5, 6). Cell synchronization studies in normal human fibroblasts revealed that similar amounts of WEE1 protein were detected in both S- and G₂–M phases but that its greatest activity was in S-phase of the cell cycle (3). Furthermore, conditional knockout of WEE1 in mouse embryonic fibroblasts results in genomic instability, malfunctioning checkpoints, and premature mitosis (6). This phenotype was explained, in part, by recent findings that show a critical role for WEE1 in DNA synthesis. Knockdown of WEE1 led to DNA double-strand breaks specifically in S-phase cells undergoing DNA replication (7, 8). Data support a model of WEE1-dependent genomic stability in which WEE1 knockdown or inhibition leads to aberrantly high activity of CDK1 and/or 2, resulting in inappropriately timed firing of excessive DNA replication origins. This, in turn, quickly depletes nucleotide pools and leads to stalled replication forks that, in the absence of WEE1 activity, are substrates for the DNA exonuclease SLX4-MUS81 and resolve into DNA double-strand breaks (9).

Deregulated WEE1 expression and activity have been associated with several types of cancer. WEE1 is often overexpressed in glioblastomas where elevated levels of WEE1 mRNA are linked to poor prognosis (10). High expression of WEE1 was found in malignant melanoma and correlated with poor disease-free survival in this population (11). Aberrant WEE1 expression has been implicated in additional tumor types such as hepatocellular carcinoma (12), breast cancer (13), colon carcinoma (14), lung carcinoma (15), and head and neck squamous cell carcinoma (16). Advanced tumors with an increased level of genomic instability may require functional checkpoints to allow the repair of DNA perturbations that accompany genomic instability. Therefore, WEE1 might be an attractive target in advanced tumors where its inhibition may lead to irreparable DNA damage (reviewed in ref. 17).

MK-1775 is a potent and selective ATP-competitive small-molecule inhibitor of WEE1 (18) and is currently under clinical development as a chemosensitizer in combination with chemotherapeutics (19, 20). Because of the DNA-damaging effects resulting from the loss of WEE1 activity, we hypothesized that targeted pharmacologic inhibition of WEE1 in the absence of chemotherapy could be a viable anticancer strategy. We show that treatment with MK-1775 gives rise to DNA damage in S-phase cells even in the absence of standard chemotherapeutic DNA-damaging agents and that premature mitosis is not required for its ability to inhibit cancer cell proliferation. At tolerated doses, MK-1775 leads to tumor growth inhibition (TGI) in multiple xenograft models. Like WEE1, PKMYT1 also phosphorylates and inhibits CDK1 and 2 so we questioned whether this kinase affected cell-based responses to MK-1775 treatment. Our data suggest that low PKMYT1 expression could be a determinant of MK-1775 sensitivity. The results presented here support the use of the WEE1 inhibitor MK-1775 as a DNA-damaging anticancer therapy and suggest reduced PKMYT1 expression as a possible feature of the most responsive tumors to this agent.

Cancer cell lines were obtained from American Type Culture Collection (not authenticated) and grown in medium recommended by the cell line vendor. Tissue culture media, serum, and supplements were purchased from Life Technologies and Sigma. For the proliferation assay screen, cells were plated in 384-well tissue culture plates in the presence of increasing concentrations of MK-1775 or with dimethyl sulfoxide (DMSO) as a vehicle control. After 96 hours, Cell Titer Glo (Promega) was used according to manufacturer's protocol to approximate live cell content. Assays were run in triplicate, and the proliferation index was calculated as the Cell Titer Glo raw value of treated samples relative to vehicle-treated control wells.

For the knockdown studies, NCI-H460 and KNS62 were transfected with siRNA pools (SMARTpool, Dharmacon) of either control nontargeting or PKMYT1 sequences using DharmaFECT formulation 1 according to manufacturer's protocol. Cells were plated in a 6-well plate at 3 × 10⁵ cells per well and transfected the following day with 25 nmol/L final siRNA concentration. Twenty-four hours after transfection, cells were rinsed with PBS, trypsinized, and seeded at 4 × 10³ cells per well in 96-well tissue culture plates. The following day (48 hours posttransfection) cells were treated with MK-1775 or DMSO for 72 hours. To approximate cell content, ViaLight (Lonza) was used according to the manufacturer's protocol. Samples were run in triplicate, and growth was calculated by determining the percentage of the control raw value for each treatment.

Cells were lysed in mammalian protein extraction reagent (MPER; Thermofisher 78505) and protein concentration was determined with the bicinchoninic acid (BCA) assay. Lysates were run on SDS-PAGE and transferred onto nitrocellulose or polyvinylidene difluoride (PVDF) membranes. Antibodies used for Western blotting at indicated working dilutions are from the following sources: total CDK1 (1:1000 dilution; CST #9112), pCDK1^Y15 (1:1000; CST #9111), pCDK1^T14 (1:1000; CST #2543), pCHK1^S345 (1:1000; CST #2348), γH2AX (1:1000; CST #2577), cyclin A (1:2000; CST #4656), and total PKMYT1 (1:1000; CST #4282) from Cell Signaling Technologies; actin-horseradish peroxidase (HRP) from Santa Cruz Biotechnology (1:10,000; sc-1616 HRP); secondary HRP-conjugated anti-mouse and anti-rabbit antibodies are from GE Healthcare (1:5000 each; NA9340 and NA9310). Blots were exposed with SuperSignal West Femto chemiluminescent substrate (Thermofisher Pierce).

Cells analyzed by flow cytometry were fixed overnight in ice-cold 70% ethanol and propidium iodide (PI)/RNase solution (BD Biosciences 550825) was used to determine total DNA content. To detect DNA double-strand breaks, cells were stained with a fluorescein isothiocyanate (FITC)-conjugated anti-γH2AX (S139) antibody (Millipore kit 17-344). To define the mitotic population, cells were stained with an anti-pHH3-Alexa 647 antibody directed against phospho-serine 28 (BD Biosciences 558217).

For synchronization studies, cells were incubated in serum-free medium for 36 hours, followed by replenishment with 20% FBS. One hour before each harvest, cells were pulsed with 10 μmol/L bromodeoxyuridine (BrdUrd). Cells were fixed and stained for BrdUrd and DNA content with an anti-BrdU FITC-conjugated antibody and with a 7-aminoactinomycin D (7-AAD) dye, respectively, according to the instructions in the BD Pharmingen FITC BrdU Flow Kit (BD Biosciences 559619, 557891). All cytometry data were collected on the BD LSR II flow cytometer using Diva software, and results were analyzed in FlowJo version 7.5.

CD-1 nu/nu female mice aged 5 to 6 weeks were obtained from Charles River Laboratories and housed in our animal care facility at standard laboratory conditions and fed 2018S autoclaveable diet and water ad libitum. The protocol was approved by Merck's Institutional Animal Care and Use Committee. Mice were inoculated with 5 × 10⁶ cells (1:1 Matrigel:PBS) for A427 and LoVo models or with 1 mm³ tumor fragments for the SK-MES-1 model, subcutaneously into the right flank. When tumor volume reached 200 mm³ (±50), mice were pair-matched so each group had a similar mean and SD. Mice received anywhere from 13 to 28 days of either vehicle (0.5% methylcellulose) or MK-1775 at 60 mg/kg, both administered twice daily at a dosing volume of 10 mL/kg (0.2 mL per 20 g mouse). Tumor volume and body weights were recorded biweekly. Percent TGI was calculated as 100 − (100 × ΔT/ΔC) if ΔT > 0 where ΔT = final mean volume − initial mean volume of treated group and ΔC = final mean volume − initial mean volume of vehicle control group.

A wide array of responses was observed when 522 cancer lines representing 16 different tumor types were screened with a selective inhibitor of WEE1, MK-1775 (Fig. 1A; Supplementary Fig. S1). Antiproliferative EC₅₀ values ranged from ≤0.1 μmol/L in 2% (9 of 522) to ≥1 μmol/L in 19% (98 of 522) of the cell lines tested (Supplementary Table S1). Comparing mean EC₅₀ values of the different tumor types revealed that, as a group, colorectal cancer cell lines were less sensitive (mean EC₅₀ = 1.16 μmol/L, n = 66, range, 0.17–>10 μmol/L) and neuroblastoma tumor lines were on average more sensitive to MK-1775 treatment (mean EC₅₀ = 0.28 μmol/L, n = 7, range, 0.12–0.45 μmol/L). The sample size of the latter group is limited, but the notion that neuroblastoma cells tend to be more affected by WEE1 inhibition is consistent with recent findings (21).

Functional genomic screens and validation studies have shown that knockdown of WEE1 leads to DNA double-strand breaks and activation of the DNA damage response (DDR). We selected a 2-hour time point to examine the immediate effects of pharmacologic inhibition of WEE1 in 6 cell lines of varying sensitivity to MK-1775 (Fig. 1B): ES-2 (EC₅₀ = 0.26 μmol/L), A2058 (EC₅₀ = 0.23 μmol/L), A431 (EC₅₀ = 0.17 μmol/L), A427 (EC₅₀ = 0.12 μmol/L), KNS62 (EC₅₀ = 3.41 μmol/L), and NCI-H460 (EC₅₀ = 3.31 μmol/L). Western blot analyses for pCHK1^S345, a surrogate marker for activated DDR (22), showed a dose-dependent activation of the DDR in all 6 cell lines, although the effect was more evident in the most sensitive cell lines, that is, ES-2, A2058, A431, and A427. As expected, a reduction in pCDK1^Y15 was also observed in all 6 cell lines, providing a link between induction of the DDR and elevated CDK activity as a result of WEE1 inhibition. Phosphorylation of CDK1 and CDK2 at T14 by PKMYT1 is also known to impair CDK1 and 2 kinase activity, and MK-1775 inhibits PKMYT1 in vitro at roughly 100-fold higher concentrations than those required to inhibit WEE1 (18). However, we failed to see an MK-1775–dependent effect on pCDK1^T14 at concentrations that induce DNA damage.

To understand when MK-1775–induced DNA damage occurs during the cell cycle, we analyzed TOV21G ovarian cancer cells by flow cytometry. In exponentially growing untreated cells, baseline staining for the DNA double-strand break marker γH2AX was between 1% and 2% (Fig. 1C, left; ref. 23). However, as little as 2 hours after addition of MK-1775 to the culture medium, 22% of cells stained positive for γH2AX (Fig. 1C, middle). Chromosomal content of the γH2AX-positive cells was >2N, suggesting that DNA damage arising from WEE1 inhibition occurs during or after the initiation of DNA synthesis in S-phase. To confirm this, TOV21G cells were treated with MK-1775 and pulse-labeled with BrdUrd. γH2AX was detected almost exclusively in BrdUrd-positive cells (Fig. 1C, right). This finding supports our observation that DNA double-strand breaks due to pharmacologic WEE1 inhibition arise during DNA synthesis and is consistent with similar results using genetic disruption of WEE1 expression (7, 8).

Chromosomal breaks during DNA synthesis would be expected to activate the DNA replication checkpoint and slow progression through S-phase. To address this, we analyzed the effects of MK-1775 treatment on cell populations synchronized by serum depletion. We avoided cell synchronization approaches targeting DNA synthesis (e.g., double-thymidine block, aphidicolin, hydroxyurea, actinomycin D, etc.) because these methods are disruptive to DNA replication (causing stalled forks) and may confound analyses by inadvertently sensitizing cells to MK-1775 treatment. Instead, we opted to induce G₀ synchronization through serum withdrawal. We selected the ES-2 line for these studies because it was amenable to synchronization by serum depletion. Complete serum withdrawal for 36 hours in ES-2 cells did not reduce viability but shifted the G₀–G₁ fraction to 75% to 80% (data not shown). Addition of 20% FBS caused vehicle-treated ES-2 cells to almost double their S-phase population by 8 hours and peak near 50% by 12 to 14 hours (Fig. 2A, top panel). However, when 500 nmol/L MK-1775 was included with the addition of 20% FBS to G₀ synchronized ES-2 cells, there was no detectable change in the S-phase population by 12 hours and peak levels (∼50%) were delayed until 24 hours post-FBS addition. Western blot analysis presented in Fig. 2A (bottom) confirmed the delayed accumulation of cyclin A (indicative of S-phase), a more rapid and robust induction of pCHK1^S345, and inhibition of pCDK1^Y15 in MK-1775–treated relative to vehicle-treated cells. Phosphorylation of pCDK1^Y15 was initially reduced by MK-1775 (Fig. 2A, compare lanes 1 and 7) but increased between 12 and 24 hours after addition (Fig. 2A, compare lanes 7 and 11), although the reason for this is unknown.

Although the percentage of cells in S-phase peaked near 50% for both vehicle- and MK-1775–treated samples by 24 hours post-FBS addition, the mean fluorescent intensity of incorporated BrdUrd was far lower in MK-1775 than in vehicle-treated cells, suggesting slowed DNA replication in the MK-1775–treated population (y-axis, Fig. 2B). We also analyzed γH2AX induced by MK-1775 in serum-starved versus serum-stimulated cell populations. Following 36-hour serum deprivation, ES-2 cells were treated with 500 nmol/L MK-1775 alone or in the presence of 20% FBS. A 24-hour treatment with MK-1775 in the presence of 20% FBS resulted in an increased S-phase population compared with the serum-starved control group (48% compared with 18%, Fig. 2C). γH2AX was detectable in approximately 3 times as many cells (63% compared with 23% γH2AX-positive) following MK-1775 treatment under conditions of serum stimulation (Fig. 2C and Supplementary Fig. S2). These data support our own and others' observations that disruption of WEE1 kinase activity results in DNA double-strand breaks as a result of deregulated DNA replication.

WEE1 is required for the temporal regulation of both CDK2 and 1 in S- and G₂ phases of the cell cycle, respectively. Inhibition of WEE1, therefore, is expected to lead to both S-phase defects (DNA double-strand breaks during DNA replication) and G₂–M defects (premature mitosis). To question whether either or both of these events is necessary or sufficient for MK-1775–driven cytotoxicity, we examined γH2AX and phosphorylated histone H3 (pHH3), a marker of mitosis, in 3 MK-1775–sensitive cell lines, A2058, HT-29, and LoVo (24, 25). For each line, an approximate EC₉₀ concentration of MK-1775 was selected on the basis of cell proliferation assays (Supplementary Fig. S3). After 24 hours of treatment with MK-1775, the percentage of pHH3-positive cells increased in 2 of the 3 cell lines, from 2% to 26% in A2058 cells and from 5% to 77% in HT-29 cells (Fig. 3). Importantly, only the HT-29 cell line contained a substantial mitotic population with <4N DNA, which indicates premature mitosis from S-phase cells that have not completed DNA replication (46%, Fig. 3). Therefore, premature mitosis could be an underlying driver of cytotoxicity in some cellular contexts, such as HT-29, but not all, such as LoVo and possibly A2058. On the contrary, substantial increases in γH2AX-positive cell populations were observed in all 3 cell lines following MK-1775 treatment (28% in A2058, 77% in HT-29, 53% in LoVo; Fig. 3, bottom). These data suggest that induction of DNA damage rather than premature mitosis is the primary cytotoxic consequence of WEE1 inhibition by MK-1775 in sensitive cell lines.

To determine the effect that MK-1775 monotherapy treatment has on tumor growth in vivo, a maximum tolerated dose (MTD) was established at 60 mg/kg for twice daily dosing. Mean body weight loss over the course of a 28-day study at this dose and schedule did not exceed 5% in the treated group (data not shown). MK-1775 inhibits proliferation of the A427 non–small cell lung cancer (NSCLC) cell line at low concentrations (EC₅₀ = 116 nmol/L) and readily induces the DDR (Fig. 1). In the A427 xenograft model, MK-1775 treatment caused regression to approximately 50% of the initial mean tumor volume (Fig. 4A). Individual tumor analysis shows that 9 of the 10 vehicle-treated A427 tumors grew between 2- and 6-fold over their starting volume (Fig. 4B). In contrast, the final volumes of all 10 MK-1775–treated tumors were smaller than their initial volumes (Fig. 4B). Tumor growth-inhibitory effects of MK-1775 were observed in additional xenograft models chosen for their in vitro sensitivity: 92% TGI in the SK-MES-1 NSCLC model (Fig. 4C), 13% tumor regression at day 13 (dosing was abbreviated to 13 days) in a LoVo colorectal cancer xenograft model (Fig. 4D), 88% TGI in A431 epidermoid tumor model (data not shown), and 64% TGI in NCI-H2122 NSCLC model (data not shown). Collectively, these results show the anticancer therapeutic potential of MK-1775 in the absence of any additional targeted or DNA damaging agents.

The majority of cancer cell lines that we treated with MK-1775 show at least some degree of sensitivity to MK-1775 treatment (Supplementary Fig. S1). However, not all cell lines are equally susceptible to WEE1 inhibition and antiproliferative EC₅₀s vary by as much as 10-fold (Supplementary Fig. S1). One potential determinant of sensitivity to WEE1 inhibition is activity of a functionally related CDK-inhibitory kinase, PKMYT1. Phosphorylation of CDK1 or CDK2 at either of 2 N-terminal sites, T14 or Y15, causes inactivation of the kinase despite the presence of an otherwise activating cyclin-binding partner. WEE1 is known to phosphorylate Y15 of CDK1 and 2, and PKMYT1 has been shown to similarly inhibit CDK1 and 2 through phosphorylation at T14 and/or Y15 (26).

We used siRNA knockdown to determine whether PKMYT1 expression can specifically alter the response to WEE1 inhibition in 2 cell lines, NCI-H460 and KNS62. These lines were selected because they show both relative insensitivity to MK-1775 treatment and relatively high expression of PKMYT1 (data not shown). Cells were transfected with a pool of 4 distinct siRNAs, all targeting PKMYT1, and analyzed in proliferation assays for sensitivity to different cytotoxic agents (Fig. 5A). In a representative experiment shown in Fig. 5A, MK-1775 antiproliferative EC₅₀s for NCI-H460 (n = 3) and KNS62 (n = 2) shifted from 677 to 104 nmol/L and from 487 to 93 nmol/L, respectively, when PKMYT1 was knocked down. Reduction of PKMYT1 potentiated MK-1775 an average of 4.7-fold in NCI-H460 cells (n = 3) and 4.9-fold in KNS62 cells (n = 2). The specificity of PKMYT1-dependent potentiation of MK-1775 is confirmed by identical dose response curves for carboplatin, the MEK inhibitor PD-0325901, or doxorubicin in both the control and PKMYT1 siRNA–transfected cells (Fig. 5A).

Western blot analysis of KNS62 cells (Fig. 5B) revealed that PKMYT1 knockdown results in slightly lower basal phosphorylation of CDK1 and 2 on Y15 and markedly reduced basal phosphorylation of T14 (lane 9 vs. lanes 1 and 5). Knockdown of PKMYT1 alone did not induce γH2AX (Fig. 5B, lane 9), but in the presence of MK-1775, knockdown of PKMYT1 led to a larger increase in both pCHK1^S345 and γH2AX. This is consistent with the observations that MK-1775–mediated cytotoxicity arises from DNA damage (Fig. 3) and that PKMYT1 knockdown further sensitizes cells to MK-1775–dependent antiproliferative effects (Fig. 5A).

Because PKMYT1 knockdown leads to increased sensitivity to MK-1775, we reasoned that low PKMYT1 expression might be common among MK-1775-responsive cell lines. To address this, we used the Broad-Novartis Cancer Cell Line Encyclopedia (CCLE), a publicly available cell line database (27), to find PKMYT1 mRNA expression levels in the 522 cell lines that we had treated with MK-1775 (Supplementary Fig. S1). PKMYT1 mRNA expression data were available for 305 of the 522 cancer cell lines assayed for sensitivity to MK-1775. Plotting relative PKMYT1 expression from the CCLE database against our cell line response data at 450 nmol/L of MK-1775 failed to show a correlation between PKMYT1 mRNA and MK-1775 sensitivity (Fig. 6A). However, 24 of the 33 cell lines (73%) that were killed by 450 nmol/L MK-1775 (response < 0.25 on an adjusted scale, indicated by a dashed vertical line in Fig. 6A) had less than mean expression of PKMYT1 mRNA (i.e., 413 ± 154).

To prospectively test the hypothesis that low PKMYT1 expression might predict MK-1775 sensitivity, we selected 13 additional cell lines from the CCLE database with varying expression levels of PKMYT1 mRNA that had not previously been tested with MK-1775. The antiproliferative EC₅₀ values of MK-1775 across the 13 cell lines are related to PKMYT1 mRNA expression with R² = 0.496 (Fig. 6B, left) and PKMYT1 protein expression with R² = 0.310 (Fig. 6B, right). Accordingly, PKMYT1 mRNA and protein levels also show good correlation (R² = 0.51), although PKMYT1 protein levels do not necessarily predict PKMYT1 kinase activity in these cells. Although lower PKMYT1 expression did not invariably result in greater sensitivity to MK-1775 in our cell panel of 305 lines, our data do support the hypothesis that low PKMYT1 expression is a common feature among the most MK-1775–responsive cell lines.

MK-1775 is a potent and selective inhibitor of the WEE1 kinase. As of this publication, it is the only WEE1 inhibitor that the authors are aware of currently undergoing evaluation as an anticancer agent in combination with chemotherapy in early-stage clinical trials (19, 20, 28). Previous studies using MK-1775 have shown its potentiation of DNA damage–based therapeutics by forcing unscheduled mitosis and ultimately resulting in apoptosis or mitotic catastrophe (4, 18, 29–32). However, the potential therapeutic effects of WEE1 inhibition in the absence of chemotherapies have not been widely explored. RNA interference knockdown of WEE1 is known to inhibit proliferation of cancer cell lines (13, 33), and more recently, it was shown that MK-1775 alone can induce apoptosis in sarcoma cell lines treated in vitro (34). Our results similarly highlight a requirement for WEE1 activity to maintain cellular viability and genomic stability. Furthermore, we provide the first demonstration of TGI with MK-1775 monotherapy and suggest that low levels of PKMYT1 mRNA or protein could be a possible indicator of sensitivity to WEE1 inhibition by MK-1775.

In the only other instance where MK-1775 effects were studied independently of a DNA damaging partner, Kreahling and colleagues attribute the cell-based cytotoxicity of MK-1775 to induction of premature mitosis (34). This alone, however, is unlikely to account for the antiproliferative activity of MK-1775 in their study. For example, Kreahling and colleagues reported a proliferative EC₅₀ of 169 nmol/L for MK-1775 in the more sensitive HT1080 cell line, yet only 5% of the population was pHH3-positive after 500 nmol/L MK-1775 treatment (34). Forced mitosis, as evidenced by increased pHH3 staining, is not consistent with proliferation-based sensitivities to MK-1775. Recent findings underscore a critical role for WEE1 in regulating appropriate initiation and progression of DNA replication forks and thereby maintaining genomic integrity by preventing DNA double-strand breaks during DNA replication (7–9, 35). Kreahling and colleagues did not examine markers of DNA damage following MK-1775 treatment so the relative contribution of premature mitosis versus DNA damage cannot be appreciated. We also found that some cell lines display a large increase in pHH3 staining in S-phase cells, indicative of premature mitosis (e.g., HT-29, Fig. 3), but we also found that premature mitosis was not a requirement of sensitivity to WEE1 inhibition (e.g., LoVo, Fig. 3). Because a strong induction of DNA damage accompanied MK-1775–driven cytotoxicity, regardless of the effect on mitotic indices, our results suggest that DNA damage rather than premature mitotic entry is the dominant, although not exclusive, mechanism underlying effectiveness of WEE1 inhibition.

PKMYT1 and WEE1 both catalyze inhibitory phosphorylations on CDK1 and 2. Our observations that low PKMYT1 mRNA expression is common among the most sensitive cell lines to MK-1775 and that knockdown of PKMYT1 can sensitize less responsive cell lines to MK-1775 together suggest functional redundancy between PKMYT1 and WEE1. In support of this, siRNA studies have shown that knockdown of PKMYT1 leads to similar, although less pronounced, abrogation of G₂ cell-cycle arrest and sensitization to DNA-damaging agents (31, 36, 37). Furthermore, and similar to WEE1, overexpression of PKMYT1 is sufficient to induce a G₂ cell-cycle delay in HeLa cells (38). Interestingly, this study found that the interaction of PKMYT1 with the CDK1–cyclin B1 complex, rather than PKMYT1 phosphorylation of CDK1–cyclin B1, was responsible for the cell-cycle delay. This argues in favor of PKMYT1 expression rather than PKMYT1 activity as a potential indicator of MK-1775 sensitivity. In our evaluation of sensitivity and PKMYT1 mRNA expression among 305 cell lines, we found that many lines with relatively low levels of PKMYT1 did not respond to MK-1775 treatment (Fig. 6A). Despite the caveats inherent in comparing two independent data sets (internal response data and CCLE expression data), this suggests that PKMYT1 expression could be one of multiple prognostic factors when trying to predict the outcome of WEE1 inhibition.

MK-1775 has been widely studied in preclinical xenograft models as a chemotherapy or radiation sensitizer. These studies generally show that MK-1775 monotherapy is not an effective anticancer treatment. It should be noted, however, that MK-1775 is dosed below its monotherapy MTD of 60 mg/kg twice daily in these studies, arguably accounting for the differing single-agent anticancer activity observed between previous studies and our work presented here. Importantly, in a study of patient-derived pancreatic carcinoma xenograft models, control groups receiving MK-1775 single-agent treatment at doses considerably below MTD showed surprising TGI (39). Unlike responses in the gemcitabine and MK-1775 combination arm, the anticancer activity of MK-1775 was not dependent on p53 mutational status (39), consistent with our own work (both A427 and LoVo xenograft models are wild-type for TP53) and that of Kreahling and colleagues (34).

Recent studies have described synergy between inhibitors of WEE1 and CHK1 kinases (40, 41). Genomic damage resulting from deregulated DNA replication, determined by γH2AX staining in S-phase cells, is not only a hallmark of the response to WEE1 monotherapy (described here) but also both the combination of the WEE1 and CHK1 inhibitors and CHK1 inhibitor monotherapy (42). Our own work supports the in vivo combination benefit from combined WEE1 and CHK1 inhibition (data not shown; ref. 43). Notably, however, when MK-1775 and MK-8776 (formerly SCH-900776) are co-administered, the combination MTD requires both a dose reduction (60 mg/kg each drug alone to 40 mg/kg in combination) and a schedule reduction (twice daily dosing each drug alone to twice weekly dosing in combination), reflecting increased toxicity of the combination (data not shown and (43)). A comparison of the TGI of the MK-1775 and MK-8776 regimen at combination MTD versus MK-1775 alone at monotherapy MTD suggests that despite the strong in vitro synergy of the WEE1 and CHK1 inhibitor combination, the 2 treatments have similar therapeutic indices (91% TGI for combination vs. 13% regression for MK-1775 alone in LoVo colorectal xenograft model; 70% TGI for combination vs. 89% TGI for MK-1775 alone in SK-MES-1 NSCLC xenograft model; 59% TGI for combination vs. 90% TGI for MK-1775 alone in A-431 epidermoid xenograft model). Future studies will be required to determine whether a specific cellular context or altered dosing approach for combined WEE1 and CHK1 inhibitors provides an advantage over either single-agent treatment. Regardless, our work corroborates findings that WEE1 activity is essential to genomic stability and that WEE1 inhibition constitutes a viable therapeutic consideration based on anticancer efficacy of MK-1775 monotherapy.

A.D. Guertin is employed (other than primary affiliation; e.g., consulting) as a scientist, J. Li is employed as a director of Biology Discovery, and G. Gilliland as senior vice president with Merck & Co., Inc. No potential conflicts of interest were disclosed by the other authors.

Conception and design: A.D. Guertin, B. Long, C. Toniatti, L. Zawel, S.E. Fawell, S.D. Shumway

Development of methodology: J. Li, Y. Liu, Y. Benita, L. Zawel

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A.D. Guertin, J. Li, Y. Liu, M.S. Hurd, A.G. Schuller, B. Long, L. Zawel, S.D. Shumway

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A.D. Guertin, J. Li, Y. Liu, A.G. Schuller, H.A. Hirsch, I. Feldman, Y. Benita, C. Toniatti, L. Zawel, S.E. Fawell, S.D. Shumway

Writing, review, and/or revision of the manuscript: A.D. Guertin, A.G. Schuller, H.A. Hirsch, Y. Benita, C. Toniatti, L. Zawel, S.E. Fawell, G. Gilliland, S.D. Shumway

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A.D. Guertin, L. Zawel

Study supervision: J. Li, B. Long, C. Toniatti, L. Zawel, S.E. Fawell, S.D. Shumway

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.