Introduction

Medulloblastoma is the most common malignant brain tumor in children. Over 800 cases per year occur worldwide with the vast majority occurring in children less than 12 years of age [1]. The mainstays of medulloblastoma therapy continue to be surgery, radiation and cytotoxic chemotherapy [2]. While therapy for standard risk patients has resulted in improved outcomes, high-risk patients do poorly. In particular those showing relapse and MYC amplification have a 5-year survival rate of less than 40% [3]. In addition, there remains significant therapy-related morbidity, particularly in the very young patients [4–6]. Novel therapeutic approaches based on tumor biology are clearly needed to improve outcomes for these children.

Recent genomic analysis has been successfully used to identify medulloblastoma subtypes [7–10]. International consensus has resulted in four molecular subgroups being defined [11]. These are the Wnt and Shh signaling subgroups as well as Group 3 and 4. Group 3 tumors largely represent the MYC amplified tumors whereas there is not a clear molecular definition of the Group 4 tumors [11]. However, finding therapeutic targets from these categories is still challenging [12]. Patients with the Wnt signaling signature are in a very good risk category and efforts are underway to de-escalate therapy for this cohort of patients [13]. For patients with the Shh signature, there are targeted inhibitors currently in early phase trials [13]. Unfortunately molecular targeting for Group 3 and 4 tumors is less clear. This is particularly problematic since Group 3 and 4 tumors constitute 60% of all medulloblastoma tumors [11].

The advent of RNA interference (RNAi) technologies for targeting large sets of genes in mammalian cells allows us to systematically interrogate gene functions in a high throughput manner [14, 15]. This functional genomic approach has successfully resulted in the discovery of genes that were components of Ras oncogene driven tumors [16, 17], of genes that sensitize cells to chemotherapeutic agents [33]. Together with our data these two studies demonstrate the utility of functional genomics to identify novel therapeutic targets in medulloblastoma.

Cell cycle checkpoint genes have previously been implicated in medulloblastoma [34–36]. However to our knowledge this is the first report in medulloblastoma to document the functional importance of mitotic kinases and in particular WEE1. WEE1 controls the G2-M transition by catalyzing the inhibitory phosphorylation of CDK1 thus preventing the Cyclin B-CDK1 complex from driving cells into mitosis [37]. The G2-M cell cycle checkpoint is critical for eukaryotic organisms ensuring cells do not initiate mitosis before DNA damage is repaired [38]. WEE1 is a key regulator of this process [37, 28]. Thus abrogation of the G2-M checkpoint leading to premature mitotic entry and subsequent cell death by mitotic catastrophe has emerged as a promising new therapeutic strategy [39, 38].

WEE1 over expression is associated with several types of cancer [38]. More recently an in silico analysis of a large data set of glioblastoma identified WEE1 as over expressed and a key regulator of mitotic catastrophe [27]. Similarly a functional genomic approach in acute myeloid leukemia also identified WEE1 as a key regulator of chemotherapy sensitivity [40]. This study shows that WEE1 is over expressed and functionally important in medulloblastoma. Our data show that the small molecule inhibitor of WEE1, MK-1775, is a potent inhibitor of medulloblastoma cell growth and synergizes with cisplatin to decrease cell proliferation in vitro. Cisplatin in combination with MK-1775 treatment induces an increase in the percentage of cells in the S and G2-M phases of the cell cycle (Additional file 6: Figure S4A) and subsequently leads to more DNA damage as measured by γH2AX foci than DNA damage induced by cisplatin alone. Similarly an increase in mitosis was demonstrated in the combination treatment by increased phospho-H3 staining as seen in Additional file 6: Figure S4B and C. Interestingly, MK-1775 as a single agent potently inhibits medulloblastoma tumor growth in vivo. These data could be explained by the fact that MK-1775 induces DNA damage and genomic instability. MK-1775 has recently been shown to inhibit growth of sarcoma cells and glioblastoma cells among many other tumor types [31, 41, 42].

Inhibition of WEE1 has largely been investigated in the context of abnormal p53 function, given that cells with impaired p53 function are highly dependent on the G2-M checkpoint to maintain genomic integrity [43]. For example studies have demonstrated that cells with dysfunctional p53 can be sensitized to DNA damage by impairing the G2-M checkpoint through inhibition of WEE1 [30]. Our data showing synergy of WEE1 inhibition in combination with cisplatin were generated in a medulloblastoma cell line that has nonfunctional p53. A more thorough investigation with additional functional p53, nonfunctional p53 and isogenic cell lines are needed to determine the role of p53 on the effectiveness of WEE1 inhibition to sensitize cells to DNA damaging agents.

We have demonstrated that WEE1 inhibition sensitizes medulloblastoma cells to cisplatin in vitro. However it is unlikely that all patients with medulloblastoma will respond in a similar manner. Thus, it will be important to develop suitable biomarkers to predict which patients may benefit the most from such a therapeutic strategy. Detailed animal modeling of medulloblastoma with assessment of potential biomarkers as well as pharmacokinetics will provide further data in evaluating the effectiveness of WEE1 inhibition in conjunction with cisplatin in eradicating medulloblastoma cells in vivo and prolonging patient survival.

Importantly, MK-1775 is a well-tolerated drug with low dose limiting toxicities [44]. Clinical trials are now underway for MK-1775 including combination with carboplatin for ovarian cancer (NCT01164995) and combination with gemcitabine, cisplatin or carboplatin in advanced solid tumors (NCT00648648). Because MK-1775 crosses the blood brain barrier it is a promising agent for brain tumor therapy. In fact a Phase I trial in glioblastoma multiforme in adults is currently underway (NCT01849146) as well as a newly initiated trial for diffuse intrinsic pontine glioma in children (NCT01922076). For medulloblastoma therapy we envision a strategy of administering MK-1775 thrice weekly for 3 weeks with each cycle of cisplatin. Thus once the maximum tolerated dose is obtained in a Phase I trial, we would propose a Phase II/III trial of cisplatin based therapy (the current standard of care) compared to cisplatin plus MK-1775.

In summary our data supports a role for WEE1 in regulating the G2-M checkpoint in medulloblastoma and validated WEE1 as a therapeutic target. We demonstrate that a clinically relevant small molecule inhibitor, MK-1775, is potent in the inhibition of medulloblastoma tumor growth in vivo. In light of our data and that of others, in combination with the good positive safety profile, we suggest that MK-1775 is an exciting new agent in the treatment of pediatric brain tumors, particularly medulloblastoma.