A WEE1 family business: regulation of mitosis, cancer progression, and therapeutic target

The WEE1 kinase family consists of three serine/threonine kinases sharing conserved molecular structures and encoded by the following genes: WEE1 (WEE1 G2 Checkpoint Kinase), PKMYT1 (membrane-associated tyrosine- and threonine-specific cdc2-inhibitory kinase), and WEE1B (WEE2 oocyte meiosis inhibiting kinase). In eukaryotic somatic cells, WEE1 and PKMYT1 play a key role in cell cycle regulation and, in particular, they are involved in the entry into mitosis [1]. Their role as regulators is crucial during normal cell cycle progression and in response to DNA damages, as part of the DNA damage response (DDR) pathways. Similarly, WEE2 regulates cell cycle progression and, in particular, meiosis [2]. Briefly, WEE2 plays a dual regulatory role in oocyte meiosis by preventing premature restart prior to ovulation and permitting metaphase II exit at fertilization [3]. Despite the identification of WEE2 somatic mutations (1.9% of cases) and copy number (CN) alterations (22.5% of patients with CN loss and 22.5% with CN gain) across several cancer types (https://​portal.​gdc.​cancer.​gov), they have not been functionally linked to tumor development so far. Therefore, the following sections will be focused on WEE1 and PKMYT1 kinases that have a well-recognized role in oncology and hemato-oncology.

WEE1 and PKMYT1 act as tumor suppressors in non-malignant eukaryotic somatic cells. Similarly to other DDR-related kinases, their main biological function is to prevent replication of cells with altered DNA. The main downstream target of WEE1 family kinases is the cyclin-dependent kinase 1 (CDK1)-cyclin B1 complex, also known as mitotic-promoting factor (MPF). WEE1 phosphorylates CDK1 on Tyr15 while PKMYT1 has a dual activity on Tyr15 and Thr14 [4] (Fig. 1a). The phosphorylation of those residues keeps the MPF complex inhibited until the cell approaches mitosis. WEE1 is located in the nucleus, while PKMYT1 is associated with the endoplasmic reticulum and Golgi apparatus [5, 6], and regulates Golgi membrane reassembly following mitosis [7]. Together, WEE1 and PKMYT1 ensure that CDK1 remains inactive as it shuttles into and out of the nucleus [8]. Through its extra-nuclear localization, PKMYT1 can also promote CDK1 cytosolic segregation. At the G2/M border, if no DNA damage has been detected, CDK1 phosphorylation on Tyr15 and Thr14 is rapidly removed by CDC25C phosphatase. In the nucleus, the CDK-activating kinase (CAK) complex composed by cyclin-dependent kinase 7 (CDK7), cyclin H1, and MAT1 promotes MPF complex activation through the phosphorylation of CDK1(Thr161) [9, 10]. The active MPF complex is then imported into the nucleus through phosphorylation of cyclin B1 (Ser126, Ser128, Ser133, and Ser147) [11]. This event is required to enter mitosis. The relevance of WEE1 and PKMYT1 regulation of CDK1 has been recently confirmed by in vivo studies. Indeed, the replacement of the CDK1 inhibitory phosphorylation sites with non-phosphorylatable amino acids (CDK1^T14A/Y15F) was embryonic lethal in mice [12]. Once activated, the MPF complex can phosphorylate WEE1 and PKMYT1 to promote their inactivation via different cascades [5, 13, 14]. WEE1 is phosphorylated (Ser123) by CDK1 at the onset of mitosis, thereby generating a binding motif for polo like kinase 1 (PLK1) and casein kinase 2 (CK2), that in turn phosphorylate WEE1 (Ser53 and Ser121, respectively) [14, 15]. Together, the phosphorylation of the three Ser residues serves as a tag for the degradation of WEE1 by the ubiquitin ligase SCFβ-TrCP [13]. PKMYT1 is also phosphorylated by CDK1 and PLK1 and this event promotes its degradation [16]. In addition to the checkpoint function at the G2/M border, recent findings highlighted a role of WEE1 in the regulation of replication dynamics during S phase (intra S phase checkpoint). When cells reach the S phase, replication is initiated from a large number of replication origins triggered through the activation of the pre-replication complex [17] and following the activation of S phase specific CDK, primarily CDK2 [18, 19]. Similarly to CDK1, CDK2 regulation is controlled through Tyr15 phosphorylation status, that is balanced by WEE1 (Fig. 1a) and cell division cycle 25A (CDC25A) activity [20]. Both WEE1 and CDC25A/C have been shown to modulate unperturbed replication through regulating CDK1/CDK2 activity. Monoallelic expression of CDK1^T14A/Y15F induced replication stress and S phase arrest in mouse embryonic fibroblasts (MEFs), with substantial increase of γH2AX levels, chromosomal fragmentation, and DDR activation, as a consequence of intra-S phase DNA damage [12]. Moreover, unscheduled origin firing due to loss of WEE1 leads to exhaustion of the replication protein A1 (RPA1) pool and, as a consequence, to death during DNA replication (replication catastrophe). The intra S phase activity of WEE1 is independent from PKMYT1 that is unable to phosphorylate CDK2 [5]. In addition, WEE1, but not PKMYT1, contributes to the control of mitosis exit. Indeed, Wee1-deficient MEFs showed mitotic defects (e.g., in the number and position of centrosomes) that induce arrest in mitosis or, in the majority of cells, mitotic slippage [21, 22]. At the end of mitosis, WEE1 inhibits CDK1 through phosphorylation of its Tyr15 residue (Fig. 1a). This event is dependent on the activation of the CTD phosphatase subunit 1 (FCP1) that dephosphorylates and activates WEE1 and other crucial component of the spindle assembly checkpoint (SAC) complex [23]. Although the precise mechanisms that regulate FCP1 activity is still unknown, it has been showed that FCP1 promotes the dephosphorylation of crucial SAC components, including cell division cycle 20 (CDC20) and ubiquitin specific peptidase 44 (USP44), thus promoting APC/C^Cdc20 activation and chromosome segregation [24‐26]. Moreover, WEE1 directly interacts with APC/C components, including fizzy and cell division cycle 20 related 1 (CDH1), CDC20, cell division cycle 27 (CDC27), and its deletion enforced APC/C activity, resulting in alterations of the level of APC/C substrates and mitosis progression at the expense of genomic stability [21].

The activity of WEE1 through the cell cycle can explain its tumor suppressor function, at least in nonmalignant cells. This observation was confirmed and disentangled in preclinical studies. Indeed, conditional Wee1 heterozygous deletion in the murine mammary epithelium caused enhanced proliferation, with cells progressing into mitosis while still undergoing DNA replication, and consequent accumulation of DNA damage, resulting in genomic instability and, ultimately, in tumor development [21]. Biological processes such as DNA replication and homologous recombination involve the formation of branched DNA structures that physically link chromosomes. Such DNA structures needs to be disengaged prior to entry into mitosis, in order to ensure proper chromosome segregation. Eukaryotic cells evolved different mechanisms to identify and process branched DNA structures (Y-shape DNA) and the most important one involves the structure-selective endonuclease MUS81. MUS81 forms heterodimeric complexes with the non-catalytic subunits EME1 or EME2 and recognizes Y-shape DNA structures during DNA replication or during mitosis (homologous recombination). The activity of MUS81-EME1/2 complex is crucial to recover stalled replication forks, during prolonged S phase arrest, and to reset DNA junction between twin chromatids during homologous recombination [27]. In unperturbed cells, WEE1 protects replication forks and prevents the generation DNA damages and chromosome pulverization through an indirect inhibition of MUS81 functionality [28]. Indeed, WEE1 phosphorylates CDK1 and CDK2, thus preventing the CDK-mediated phosphorylation and activation of MUS81-EME1/2 complexes [29]. Lack of WEE1-dependent regulation of MUS81-EME1/2 endonucleases may lead to cleavage of unwanted DNA structure (excessive replication forks), which would slow down replication progression and increase genomic instability [27, 28] (Fig. 1b).

Several DDR inhibitors have proved their efficacy against different cancer types in the preclinical and clinical settings [161‐165]. Among them, WEE1 inhibitor seems to be the most effective ones, also favored by a relative low off-target toxicity. However, despite the number of studies and the promising results, few predictive markers of response have been identified. Recently, cyclin E level has been linked to the efficacy of adavosertib in breast cancer models [89], with cyclin E-high cells, that generally show elevated chromosome instability, being more sensitive compared with cyclin E-low ones. Despite the reported low levels of WEE1 expression in breast cancer, chromosome instability, that has also prognostic potential mainly in grade 2 tumors [89], may explain the effectiveness of WEE1 inhibitors, as supported by the predictive role of cyclin E. Our group and others showed that high PKMYT1 expression associates with reduced sensitivity to adavosertib, indicating a potential compensatory effect [35, 166]. Moreover, high-throughput proteomic profiling demonstrated that small cell lung cancer and ovarian cancer models with primary resistance to adavosertib express high levels of AKT/mTOR pathway molecules and phosphorylated S6 ribosomal protein [137, 138]. In acute leukemia models, the sensitivity to adavosertib has been recently linked to HDAC and MYC regulation. Indeed, by generating adavosertib-resistant models, the researchers found that resistant acute leukemia cell lines are dependent on increased HDAC activity for their survival, partly due to increased KDM5A function. In addition, gene expression analyses demonstrated a HDAC-dependent expression of MYC in the adavosertib-resistant cell lines [167]. These observations support the success of preclinical studies combining WEE1 and HDAC [41, 130, 131] or bromodomain inhibitors [150].

Thanks to a constantly growing amount of preclinical and clinical data, our knowledge on cancer biology is increasing and, consequently, the list of cancer hallmarks has been progressively expanding. Recent findings demonstrated that cancer cells are characterized by functional and molecular alterations in crucial genes involved in the DDR pathway, which is fundamental for cell cycle regulation, DNA damages recognition, and repair. Functional alterations of DDR-gene have a deep impact on tumor progression and on the clinical outcome of cancer patients. Indeed, the efficacy of standard of care chemo/radiotherapy regimens depends on the generation of DNA damages in proliferating malignant cells. In this scenario, the overexpression or uncontrolled activation of DDR pathways has been showed to protect cancer cells from the therapeutic effect of DNA damaging agents. Moreover, a large number of preclinical studies highlighted that cancer cells depend on the functionality of DDR pathways in order to survive, to tolerate the replicative stress induced by the high proliferative rate and to sustain the intrinsic genetic instability. For these reasons, selective inhibitors have been developed in order to exploit cancer cells’ dependency on DDR-gene functionality. Pre-clinical data has proven the efficacy of DDR inhibition in different kinds of hematological and solid tumors, both as monotherapy and in combination with a wide number of DNA damaging agents. Among DDR inhibitors, the most effective once are those targeting PARP1 and WEE1 family kinases. The effectiveness of PARP1 inhibitors is however dependent on homologous recombination (HR) repair deficiency while WEE1 family kinases inhibitors seems to have a widespread efficacy independently from a specific the genetic background. Indeed, cancer cells seem to be strictly dependent on the functionality of WEE1/PKMYT1 kinases to survive, especially those with alterations targeting the G1 checkpoint. WEE1/PKMYT1 kinases are involved in different biological processes and they seem to play diverse roles in nonmalignant and in cancer cells. Indeed, they control cell cycle regulation and genetic stability in nonmalignant cells and for these reasons act as tumor suppressor genes. Conversely, their ability of promote DNA damages repair and cell cycle control makes them act as pseudo-oncogenes in cancer cells. Several molecular studies showed that malignant cells have high expression level of WEE1 and PKMYT1, which has become a good prognostic biomarker for chemo/radiotherapy regimens. However, we currently lack information regarding predictive markers of response to WEE1/PKMYT1 inhibitors. Large preclinical and clinical studies should be conducted in order to identify specific molecular backgrounds in which the use of WEE1/PKYMT1 inhibitors may be recommended. The identification of molecular vulnerabilities in cancer patients will be fundamental to design novel therapeutic regimens using WEE1/PKMYT1 inhibitors in a chemo/radiotherapy-free, synthetic lethality-based approach.