WEE1 Tyrosine Kinase, A Novel Epigenetic Modifier

Chromatin integrity and histone synthesis

Chromatin integrity - the synthesis and packaging of the nascent DNA with histones - is critical for proper chromosome condensation, segregation, epigenetic inheritance, and genome stability. Eukaryotic cells tightly regulate synthesis of core chromatin components, during each cell cycle. Although processes that interfere with DNA replication compromise genetic integrity, alterations in histone stoichiometry or mutations in histone genes are linked to chromosome loss, altered chromatin architecture, and cancer [1–6]. In addition, both histones and DNA are modified epigenetically, and inheritance of these epigenetic marks is essential to maintain cell lineage during development [7, 8]. Understanding how cells coordinate the synthesis of chromatin components and retain a subset of epigenetic marks during each cell cycle as well as identifying the key regulatory factors governing these processesa re active areas of research.

All eukaryotic cells maintain a precise ratio of core histones to the newly synthesized DNA; both higher and lower ratios of histones to the DNA have deleterious effects (Table 1). The importance of regulating this histone/DNA ratio comes from elegant genetic and biochemical studies [1, 2, 9]. In the model eukaryotic organism Saccharomyces cerevisiae, histone imbalance created synthetically by selectively over-expressing either the H2A-H2B or H3-H4 dimers, affected mitotic fidelity leading to the loss of chromosomes [1]. Moreover, histone levels are also critical for proper partitioning of chromosomes to the daughter cells [10]. Another undesirable outcome of overproduction of histones, observed in Schizosaccharomyces pombe, is a marked increase in chromosomal instability due to the inclusion of canonical H3 along with the centromere-specific H3 variant CENP-A at centromeres, which consequently increases nucleosome density and alters the chromatin structure [10]. Conversely, incorporation of CENP-A (CID in Drosophila) at non-centromeric sites leads to the formation of dicentric chromosomes, chromosome breakage during mitosis, and genome instability, which is associated with malignancy [11, 12]. Consequently, cells have developed specific mechanisms to prevent mis-incorporation of CENP-A and maintain chromatin integrity [13]. In addition, transcription of histones is tightly regulated and coordinated with the cell cycle. Genetic studies in yeast revealed presence of a surveillance machinery mediated by the checkpoint kinase, RAD53, which monitors histone dosage during replication [14].

Termination of histone gene expression

Just as histone synthesis is exquisitely regulated during the cell cycle, its termination is also precisely synchronized with cell cycle progression [2, 15–17]. Higher eukaryotes have multiple copies of histone genes for the core histones, each encoding a fraction of the total histone protein to deal with the large-scale histone synthesis required for packaging newly synthesized DNA into chromatin during S-phase [18]. These genes are organized in three major clusters in humans, HIST1 (55 histone genes), HIST2 (6 histone genes), and HIST3 (3 histone genes), with copies being arranged in tandem. At the end of S-phase or upon replication inhibition, histone levels are rapidly lowered [2, 19, 20]. In addition, histones are regulated post-transcriptionally in higher eukaryotes through a reduction in mRNA half life at the end of S phase [21–23]. However, the rapid turnover of histone mRNA has to be coupled with cessation of histone mRNA synthesis to achieve the precipitous drop in histone transcript levels observed at the end of DNA synthesis. The process by which higher eukaryotic cells halt histone transcription has remained elusive for a number of years. Recent work demonstrated that transcription of the histone genes is itself regulated by histone modification [17]. Nucleosomes upstream of the mouse histone gene cluster I (HistI) were found to be decorated by histone H2B tyrosine phosphorylation at 37 residue (pY37-H2B) precisely at the end of S-phase when DNA synthesis is completed. This mechansism is conserved in S. cerevisiae, and mutation of the equivalent site, Y40 to alanine, (H2BY40A) leads to a significant increase in all the core histone mRNA levels [17].

A dual role for WEE1 in S phase regulation and histone synthesis

WEE1 is an evolutionarily conserved nuclear tyrosine kinase (Table 2) that is markedly active during the S/G2 phase of the cell cycle [24, 25]. It was first discovered 25 years ago as a cell division cycle (cdc) mutant-wee1- in the fission yeast, Schizosaccharomyces pombe [26]. Fission yeast lacking WEE1 are characterized by a smaller cell size, and this phenotype has been attributed to the ability of WEE1 to negatively regulate the activity of cyclin dependent kinase, Cdc2 (Cdc28 in budding yeast and CDK1 in human), in the Cdc2/CyclinB complex [27]. Recently, WEE1 was shown to directly phosphorylate the mammalian core histone H2B at tyrosine 37 in a cell cycle dependent manner. Inhibition of WEE1 kinase activity either by a specific inhibitor (MK-1775) or suppression of its expression by RNA interference abrogated H2B Y37-phosphorylation with a concurrent increase in histone transcription [17]. Interestingly, S. cerevisiae lacking the WEE1 homolog, Swe1, also exhibited loss of the equivalent H2B Y40-phosphorylation and histone gene dysregulation [17]. Collectively, these data suggest a role for WEE1 as a ‘chromatin synthesis sensor’ by two sequential phosphorylation events: (i) Y15-phosphorylation of CDK1 throughout S phase to prevent exit from S phase until DNA replication is completed [17, 28], (ii) Y37-phosphorylation of H2B at the end of S phase to terminate histone synthesis, thus maintaining the right histone-DNA stoichiometry prior to mitotic entry [17].

With more than 50 genes coding for core histones in the Hist1 cluster in mouse and humans, eukaryotic cells require a strict mechanism to suppress transcription of so many genes. WEE1 deposits pY37-H2B marks at nucleosomes located upstream of HistI cluster to disengage NPAT (Nuclear protein, ataxia-telangiectasia locus) [17], a transcriptional activator of mammalian histone genes and RNA polymerase II [29, 30]. In addition, this epigenetic modification acts as a beacon for the recruitment of a transcriptional repressor, HIRA (histone regulatory homolog A) [17]. HIRA is the mammalian homologue of the yeast HIR proteins and is a component of the histone chaperone complex that spreads across silenced domains to enforce transcriptional repression [31]. HIR proteins bind to the negative regulatory site NEG located at the promoters of seven of the eight yeast histone genes to repress histone transcription [32, 33]. H2B Y37 phosphorylation is enriched in the histone promoters containing the NEG site [17], consistent with a mechanism for HIR recruitment to suppress histone gene transcription.

Cell cycle analysis indicated that cells rapidly exit S-phase once the histone transcription is completed and the pY37-H2B marks are quickly erased [17]. The temporal and transient nature of pY37-H2B suggests that cells may actively recruit a phosphatase to dephosphorylate pY37-H2B before the cells enter mitosis. Furthermore, continuous repression of histone transcription is likely to cripple cells due to a lack of histones to package nascent DNA. Consistent with this, ectopic expression of HIRA which functionally approximates the continuously phosphorylated state of histone H2B, caused arrest in S phase [34]. Although a pY37-H2B-specific phosphatase has not yet been identified, the members of the CDC14 tyrosine phosphatase family are potential candidates. CDC14 is an evolutionarily conserved dual specificity phosphatase [35, 36] that was recently found to interact with Swe1 [37]. Based on the transient nature of H2B Y37-phosphorylation, WEE1 interaction with a tyrosine phosphatase at specific chromatin loci, wherein a kinase recruits the partner phosphatase, seems likely. Other potential candidates include the EYA family of tyrosine phosphatases, which have been shown to dephosphorylate the variant histone H2AX at tyrosine 142 (Box 1).

Epigenetic marks manifested during the cell cycle

In addition to pY37-H2B, a number of other histone marks are regulated in a cell-cycle dependent manner. For example, phosphorylation of serine 10 in the N-terminal tail of histone H3 was shown to be crucial for chromosome condensation and cell-cycle progression during mitosis and meiosis in Tetrahymena [38]. Other marks, such as phosphorylation of histone H3 at serine 28 and threonine 11, were also identified to be mitosis-specific histone modifications in mammalian cells [39, 40]. Methylation of histone H3 (K4me3, K9me1, K9me2) and H4 (K20me1) and acetylation of H4 (K5, K16 and K56) are also regulated in cell cycle dependent manner [41]. Although H4K5 acetylation marks are critical for deposition of nascent histones during chromatin assembly by histone chaperones during S phase [42]; they are likewise erased after nucleosome assembly to restore the chromatin structure [41]. The histone deacetylases, HDAC1, HDAC2, or HDAC3, remove H4K5 acetyl marks and their recruitment to newly synthesized DNA is a regulated process [43]. Chicken HIRA has been shown to interact with HDAC1 and 2 [44]. Whether recruitment of HIRA by pY37-H2B establishes a binding platform for the further recruitment of HDACs to deacetylate H4K5ac in the chromatin remains to be established.

In human embryonic stem cells and induced pluripotent stem (iPS) cells the transcriptionally active chromatin marks, H3K4me3 and H3K9ac, were found to be enriched upstream and downstream of the H3 and H4 gene loci but not at the transcription start sites (TSS) in cell cycle- and transcription-dependent manner [8]. Whether pY37-H2B marks occur in pluripotent stem cells and have an active role in cell cycle-dependent transcriptional suppression of histones genes remains to be examined. Because pluripotent stem cells can differentiate into many different cell types, these cells in particular need to tightly regulate chromatin status, which can significantly impact gene expression profiles. Thus, a regulatory mechanism that maintains balanced histone pools during the cell cycle will be critical in stem cells, and given the functional and evolutionary conservation of pY37-H2B marks during cell cycle, a role in regulating histone gene expression in stem cells seems likely.

WEE1 acts as a hub for regulating chromatin integrity

During S-phase, WEE1 appears to mark the completion of not only DNA replication and genetic integrity via Y15-phosphorylation of CDK1 but also histone synthesis by marking chromatin with H2B Y37-phosphorylation before cells enter mitosis [17, 45–47]. Because coupling of DNA synthesis with histone transcription is critical for proper chromatin formation, by regulating both of these processes WEE1 acts as a master regulator of chromatin integrity. Agents that compromise WEE1 function cause mitotic infidelity, chromosome loss, and apoptosis, effect commonly referred to as mitotic catastrophe [48]. Interestingly, in contrast to single-celled eukaryotes such as budding or fission yeast that can survive without Wee1, knockout of Wee1 in mice is embryonic lethal [49]. Wee1 null embryos do not proceed past the blastocyst stage due to defects in cell proliferation. Moreover, the abnormalities exhibited by Wee1-deficient cells are more severe than that noted for other checkpoint proteins such as ATM, which regulates G2/M entry, suggesting that the loss of proliferation may not be solely attributed to early entry into mitosis [49]. Consistent with this notion, Wee1-deficient cells exhibit increased DNA damage in S-phase as evidenced by increased H2AX S139-phosphorylation, a hallmark of DNA damage, and chromosome aneuploidy [49, 50].

When DNA replication is disrupted by endogenous or exogenous factors such as alkylating agents, oxidative free radicals, ultraviolet-radiation, or nucleotide depletion, cells rapidly mobilize checkpoint and DNA repair proteins to sites of damage. In response to stalled replication forks, the DNA replication checkpoint kinase, ATR (ATM and RAD3-related) is activated. Mammalian ATR regulates activation of the checkpoint kinase CHK1 [51, 52]. CHK1, a serine/threonine kinase not only negatively regulates the CDC25 phosphatase that dephosphorylates CDK1 but also phosphorylates and stabilizes WEE1 [53–55]. Consequently, WEE1 phosphorylates CDK1 in the CDK1/cyclin B complex at tyrosine 15 to prevent mitotic entry until the damaged DNA is repaired (G2/M checkpoint). In addition, WEE1 function appears to be important for the intra-S phase checkpoint as evident from the studies demonstrating that WEE1 inhibitors synergize with CHK1 inhibitors to induce cytotoxicity [56, 57]. During S-phase WEE1 also modulates the activity of CDK1 and CDK2 present at replication origins to control replication initiation (Figure 2). Untimely CDK activity in WEE1 inhibited cells led to a marked increase in the loading of the replication factor, CDC45, at chromatin [58]. As a consequence of augmented replication origin firing, the replication forks slow down and MUS81, a structure-specific endonuclease that directly interacts with WEE1, is recruited to cleave the DNA at the stalled forks generating double strand breaks (DSBs) [58, 59]. It is plausible that WEE1 negatively regulates MUS81 activity at replication forks to prevent untimely DSB formation during DNA replication. Further, the ability of activated WEE1 to phosphorylate Y37-H2B is likely to halt histone mRNA synthesis concomitant with inhibition of DNA synthesis. This is consistent with the earlier observation that showed co-ordinated repression of these two macromolecular processes in response to genotoxic insults [19].

In addition to histone synthesis, chromosome condensation also appears to be regulated by WEE1 kinase [54]. Drosophila Wee1 and Grp (the Drosophila homolog of CHK1) were found to be crucial for the delay in chromosome condensation when DNA replication was perturbed by treatment with specific inhibitors that interfere with S phase progression or to delay metaphase in cells treated with topoisomerase inhibitors [54]. These observations imply that WEE1 co-ordinates completion of multiple chromosomal processes before cells divide by dynamically surveying events to ensure proper chromatin duplication and segregation (Figure 1). These studies suggest that WEE1 has a critical function in S-phase that is outside its well-established role in regulating mitotic entry. Thus, WEE1 is an important regulator in DNA replication, histone transcription, and chromosome condensation; perturbation of any of these processes profoundly affects chromatin integrity and faithful transmission of genetic and epigenetic information to the progeny (Figure 1).

One significant question is how the WEE1-mediated cytoplasmic signaling cascade (WEE1/CDK1) converges with WEE1-mediated nuclear events (WEE1/H2B) to coordinately regulate cell cycle progression. WEE1 phosphorylates H2B at the end of S phase when the peak phosphorylation of its other substrate, CDK1, is receding [17]. This opens up the possibility that when WEE1 senses completion of DNA synthesis, it switches its substrate preference from CDK1 to H2B. In eukaryotes the origin recognition complex (ORC), Cdc6 protein, and the minichromosome maintenance (MCM) protein complex assemble on chromatin before initiation of DNA replication. CDK1 has been shown to interact specifically with the ORCs in mammalian, Xenopus, and yeast cells [60–62]. CDKs regulate DNA replication positively by inducing the initiation of DNA replication at the G₁/S transition and negatively by preventing further rounds of origin firing. WEE1 phosphorylates CDK1 in S phase thus preventing further rounds of origin firing within the same cell cycle [58]. Thus, it is likely that upon completion of DNA synthesis, when WEE1 is prevented from phosphorylating CDK1, it switches to a new substrate, H2B. Thus, by integrating these two seemingly distinct temporal events, WEE1 promotes progression into mitosis with accurately duplicated chromatin.

WEE1-independent modes of histone regulation

Despite its central role in coordinating DNA synthesis and histone transcription, WEE1 is not the only factor regulating these processes. Genetic studies using budding yeast revealed a critical role for the RAD53 checkpoint kinase in detecting excess histones that are not incorporated into the chromatin. Excess histones bind Rad53, which targets them for degradation by the ubiquitin-proteosome machinery [6, 14]. Furthermore, excess histones in budding yeast sensitized cells to DNA replication inhibitors and lead to loss of chromosomes; conversely, decreasing histone pools by deleting specific histone genes facilitates radioresistance by promoting access of the DNA repair factors to the damaged sites [14, 63, 64]. A recent study identified multiple novel ubiquitin ligases (Hel1, Hel2, Pep5 and Snt2) in S. cerevisiae that ubiquitylate ectopically overexpressed histone H3 and targets it for degradation [65]. Although viable, these E3 ligase mutants displayed differential sensitivity to the replication inhibitor, hydroxy urea (HU). Notably, the rad53 hel1 double mutants as well as some of the other double mutant combinations did not display increased HU sensitivity compared with the single rad53 mutant, suggesting that they are either functional redundant or epistatic. Interestingly, increased association of endogenous histones with the histone chaperone Asf1 was observed even in the absence of DNA damage in the hel1, hel2, pep5, and snt2 deletion strains. These results suggest that defects in histone protein turnover exacerbate cellular sensitivity to genotoxic agents- in part by interfering with the function of histone chaperone/chromatin remodeling enzymes by sequestering them with excess histones.

Other nuclear processes that alter histone stoichiometry can also hamper genome stability [14, 64, 66]. This is evident from several different yeast strains carrying mutations that affect histone levels. For example, excessive core histone mRNA expression has been reported in S. cerevisiae lacking Trf4, which encodes a poly(A) polymerase, and these mutants are characterized by acute sensitivity to DNA replication inhibitors and defects in chromosome condensation [66]. Although, trf4 mutants contain wild-type Rad53 that can detect excess histone proteins and target them for degradation by the ubiquitin proteosome machinery, histone mRNA overexpression is detrimental to trf4 mutants as evidenced by their sensitivity to genotoxic agents and a longer than average S phase. This finding is echoed by the observation that S. cerevisiae H2B Y40A mutants accumulate excess core histone mRNAs, again suggesting that both transcriptional and post-transcriptional mechanisms are required for maintaining balanced histone pools [17]. Collectively these studies illustrate that eukaryotic cells utilize both genetic and epigenetic controls to tightly regulate histone mRNA levels to ensure genome stability.

WEE1 and cancer epigenetics

The central role of WEE1 in integrating various aspects of cell cycle progression, histone synthesis, and genomic stability makes it an important target for cancer treatment. Gene expression profiling of various tumors revealed that the Wee1 kinase is overexpressed in Glioblastoma multiforme (GBM) [48, 67], luminal, and triple negative breast cancers (TNBC) [68, 69] as well as malignant melanomas [70]. WEE1 overexpression and the resultant decrease in histone levels could lead to inefficient chromatin packaging, making the DNA more accessible to the DNA damage repair machinery and promoting radioresistance [64]. The ability of WEE1 to downregulate histone levels could explain why cancer cells become dependent on its epigenetic activity. In addition to acquiring radioresistance, decreased nucleosomal packaging and consequently local alterations in chromatin architecture may activate transcription of pro-proliferative genes or even oncogenes that are otherwise kept in check in normal cells.

Overexpression of WEE1 in certain cancer types could be exploited by using WEE1-specific small molecule inhibitors such as MK-1775 that is in clinical trials [71]. Such inhibition of WEE1 in actively replicating cancer cells would increase histone dosage and, when combined with DNA damaging agents, would interfere with the DNA repair machinery and compromise genome integrity. Consistent with this hypothesis, the WEE1 inhibitor MK-1775 was able to radiosensitize human lung, breast, skin, brain, and prostate cancer cells to DNA damaging agents, and WEE1 inhibitors were shown synergize with CHK1 inhibitors to induce cytotoxicity [56, 72–76]. Moreover, breast cancer cells treated with hydroxyl urea (which inhibits DNA synthesis) and WEE1 inhibitors were found to contain a marked increase in disorganized mitotic spindles and abnormal mitoses [56].

Enzymes that promote or reverse histone modifications have emerged as major targets for the development of small molecule inhibitors, commonly referred as epigenetic inhibitors. Many of these are already in different stages of clinical trials with significant success in hematologic cancers [77]. The precision with which histone modifying enzymes modify a specific amino acid residue in a histone makes them ideal candidates for drug discovery efforts, however, because of a highly heterogeneous nature of disease, the efficacy of the epigenetic inhibitor is likely to be seen in only a subset of cancer patients that are addicted to the activity of the histone modifying enzyme. One well studied example is histone methyltransferase EZH2 that deposits the repressive mark H3K27me3. Two somatic heterozygous mutations in EZH2, Y641 and A677, have been identified in about 22% of diffuse large B-cell lymphoma (DLBCL) and follicular lymphoma, which caused increased H3K27me3 levels [78–82]. Recently, two small-molecule inhibitors of EZH2 were identified, EPZ005687 and GSK126, which suppressed global H3K27me3 levels and reactivated silenced EZH2 target genes leading to inhibition of EZH2 mutant DLBCL cell line proliferation and the growth of xenograft tumors [83, 84]. This and other early successes, suggest that WEE1 inhibition may be a useful strategy for treatment of a subset of cancers.

Cancer is a complex disease, differences in biology and outcomes exist not only among various clinical states but also within each patient. Personalized medicine offers the potential to optimize treatment for a given patient, based on molecular biomarkers that drive individual variability or drug responses. Thus, elucidation of abundant H2B Y37-phosphorylation in a subset of GBM, TNBC or malignant melanoma tumor biopsies could be a ‘companion diagnostic’ test for administration of WEE1 inhibitor such as MK-1775. This ‘personalized therapy’ for a subset of WEE1-positive GBM, TNBC or malignant melanoma patients could be a significant development as limited therapeutic options and no targeted therapeutic modalities are currently available for these cancers with low survival rates.

Defects in chromatin architecture were recently discovered to play a critical role in pediatric and young adult GBM pathogenesis, a lethal brain tumor in both adults and children [3–5, 85]. Exon sequencing revealed two somatic mutations, K27M and G34R/G34V in the histone H3 variants, H3.3 and H3.1. Interestingly, mutations in ATRX (α-thalassaemia/mental retardation syndrome X-linked) and DAXX (death-domain associated protein) were identified in every tumor harboring a G34R/G34V mutation. ATRX and DAXX are subunits of a chromatin remodeling complex that is essential for H3.3 incorporation at pericentric heterochromatin and telomeres. It has been suggested that the alterations in heterochromatinization is the major consequence of the mutations in H3.1, H3.3, and the ATRX-DAXX chromatin remodeling pathway, leading to alternative lengthening of telomeres and expression of specific gene profiles [3–5]. Whether, GBM exhibits nucleosomal packaging defects selectively in children by acquiring histone mutations, and in adults by WEE1 overexpression would require more studies. However, the alteration in heterochromatic states appears to be the major mechanism of pathogenesis in GBM.

Concluding remarks

The discovery of H2B Y37-phosphorylation by WEE1 provides a direct link between the cell cycle kinase and an epigenetic mark that helps to maintain chromatin homeostasis. This finding raises an important question of whether WEE1 utilizes its ability to phosphorylate H2B to accomplish its other well known function as a mitotic gatekeeper. This expands our perspective of how we define the role of WEE1 in cell cycle regulation and its role in various malignancies, especially those which over express WEE1 kinase; these cancer cells could utilize alterations in histone levels to confer selective proliferative advantage and radioresistance.

Identification of WEE1 epigenetic activity raises several important questions in normal physiology as well as in disease pathology. Are there auxiliary loci epigenetically marked by pY37-H2B modification in cancers with aberrant WEE1 expression? If so, do cancer cells employ a different set of epigenetic readers to read H2B Y37-phosphorylation and are there common motifs that distinguish the promoters marked with this specific modification? It is still unclear at this point if pY37-H2B modulates chromatin architecture by recruiting chromatin-remodeling proteins and thereby impacting patterns of local gene expression. Another area to explore is whether this histone Tyr-modification potentially cross talks with other histone modifications such as the H2B ubiquitylation and H4K5 and H3K56 acetylation, which have known roles in chromatin replication or repair [8, 41, 86–88]. Further research aimed at answering these questions is critical for a clear understanding of the global regulatory role of histone Tyr-phosphorylation and the development of WEE1 inhibitors as potential cancer therapeutics.