WEE1 opposes APC/C CDH1 activity to promote S phase entry

To understand the potential roles of WEE1 in regulation of CDK2 activity during G1 and S phase entry, we constructed an influence diagram describing the known interactions of WEE1 with the G1/S regulatory network (Fig. 1a). WEE1 inhibits cyclin E–CDK2 and cyclin A–CDK2 (ref. ¹⁶). We converted this influence diagram to a series of ordinary differential equations and parametrized the model to reproduce the known dynamics of the G1/S regulatory network (Fig. 1b). We simulated the depletion of WEE1 levels to 50% of baseline and observed that CDK2 was activated earlier and S phase entry was faster than in the presence of 100% WEE1 activity.

To experimentally test whether we could observe accelerated S phase entry in cells when WEE1 activity was reduced, we used AZD1775, a WEE1i previously used in clinical trials (Extended Data Fig. 1a–c). To investigate the role of WEE1 in S phase entry while avoiding confounding effects from WEE1 perturbation during the preceding G2 phase, we synchronized hTert-RPE1 mRuby–PCNA cells in early G0/G1 phase with a CDK4–CDK6 inhibitor (CDK4/6i) (Extended Data Fig. 1d). Labelling of PCNA at the endogenous locus with mRuby generates a robust marker to define the G1/S transition (Methods). Subsequently, we washed out the CDK4/6i, releasing cells into G1, in the presence or absence of WEE1i and quantified G1 length by time-lapse imaging of mRuby-labelled PCNA (Extended Data Fig. 1e). Contrary to our model predictions, WEE1 inhibition increased both G1 length and the fraction of cells that did not enter S phase during the 48 h imaging period (Fig. 1c). To confirm that this was not an artefact of CDK4/6i treatment, we also arrested cells in G0 by contact inhibition, followed by serum starvation (Extended Data Fig. 1f), then replated cells at low density, releasing the cells into G1 in the presence or absence of WEE1i, and again performed time-lapse imaging. Again, we observed increased G1 length and an increased fraction of cells not entering S phase after WEE1 inhibition (Extended Data Fig. 1g).

As kinase inhibitors can have off-target effects, we depleted WEE1 from CDK4/6i-arrested G0/G1 cells using short interfering RNA (siRNA) (Extended Data Fig. 1h), then released the cells back into the cell cycle and quantified G1 length by time-lapse imaging. WEE1 depletion had a more severe impact than WEE1 inhibition, with very few cells entering S phase over the 48 h imaging period compared with control-depleted cells and those that did so having an increased G1 length (Fig. 1d). To be certain that the delay in entering S phase was due to loss of WEE1 function, we generated a cell line with an inducible WEE1–GFP degron (Extended Data Fig. 2) to remove endogenous WEE1 in a highly specific manner. We validated WEE1–GFP localization and normal growth of this cell line and showed that WEE1 could be degraded within 1 h of addition of doxycycline and indole-3-acetic acid (IAA) (Extended Data Fig. 2b,d,e). Degradation of endogenous WEE1 upon release from CDK4/6i-mediated G0/G1 arrest again led to most cells not entering S phase; of those that did enter S phase, the length of G1 was significantly longer in the absence of WEE1 (Fig. 1e).

These data show that, contrary to our predictions and previous understanding of the functions of WEE1, WEE1 inhibition does not accelerate S phase entry. Instead, WEE1 is required for timely S phase entry in cells returning to proliferation from quiescence.

Our model (Fig. 1a,b) predicted an earlier and more rapid increase in CDK2 activity upon WEE1 inhibition. To test this, we used an exogenously expressed fluorescent live-cell reporter (CDK2L–GFP) that measures cyclin E–CDK2 activity in real time during G1 (ref. ¹⁷) (Fig. 2a). hTert-RPE1 mRuby–PCNA CDK2L–GFP cells were released from CDK4/6i G0/G1 arrest in the presence or absence of WEE1i, and CDK2 activity was tracked by time-lapse imaging. We used a WEE1i to perturb WEE1, as it was a clinically relevant drug, and because a higher fraction of cells entered S phase with this approach compared with siRNA or degron approaches, allowing us to make statistically meaningful comparisons. When WEE1 was inhibited, CDK2 activity started to increase earlier during G1 and reached a higher activity than in dimethyl sulfoxide (DMSO)-treated cells (Fig. 2b and Supplementary Video 1). Cells treated with WEE1i therefore entered S phase at a higher level of CDK2 activity. The observation that cells could remain in G1 with high CDK2 activity was surprising; we repeated the experiment using a similar, shorter version of the CDK2 sensor (DHB–mCherry) and observed the same result¹⁸ (Extended Data Fig. 3a,b).

Two key events drive the G1/S transition: an increase in CDK2 activity and inhibition of APC/C^CDH1 E3 ubiquitin ligase activity¹⁹ (Fig. 1b). CDK2 can phosphorylate CDH1 and inhibit APC/C^CDH1 activity^20,21,22. Therefore, we tested whether there was any change in the timing and/or rate of APC/C^CDH1 inactivation relative to the G1/S transition in WEE1i-treated cells that had abnormally high CDK2 activity in G1. We used hTert-RPE1 cells expressing mCherry–geminin(1–110), constitutively expressed from an exogenous promoter, as a live-cell reporter of APC/C^CDH1 activity²¹. Geminin is very effectively degraded by APC/C^CDH1 during G1, and geminin protein only starts to accumulate after the G1/S transition (Extended Data Fig. 3c). Therefore, mCherry–geminin, together with coexpressed mTurquoise2–PCNA (under the control of an exogenous promoter), is a useful reporter for determining the timing and rate of APC/C^CDH1 inactivation relative to the G1/S transition. Using this reporter, we identified that in cells released from CDK4/6i G0/G1 arrest, APC/C^CDH1 activity was switched off just after the G1/S transition in WEE1i-treated cells, with timing and rate indistinguishable from those observed in DMSO-treated cells (Fig. 2c and Extended Data Fig. 3c). Our data suggest that upon WEE1 inhibition, CDK2 activity increases to a higher level in G1, but the timing of APC/C^CDH1 inactivation relative to the G1/S transition remains similar to that in DMSO-treated cells (Fig. 2d). Consequently, APC/C^CDH1 inactivation is more important than reaching a threshold of CDK2 activity with respect to establishing the timing of S phase entry.

Low CDK2 activity during G1 provides a period for licensing of DNA replication origins²³. An early increase in CDK2 activity in G1 in WEE1i-treated cells could decrease DNA replication origin licensing. CDC6 is an essential replication origin licensing factor that is required for formation of prereplicative complexes at replication origins²⁴. We used an hTert-RPE1 cell line expressing mTurquoise2–PCNA, DHB–mCherry and CDC6–mVenus²⁵, all under the control of exogenous constitutively active promoters, to track the dynamics of nuclear CDC6 protein relative to the G1/S transition. In cells released from CDK4/6i G0/G1 arrest in the presence of WEE1i, the licensing window (the time from the rise to the peak in nuclear CDC6–mVenus intensity²⁵) was reduced compared with that in DMSO-treated cells (Fig. 2e). We also quantified replication origin licensing during G1 by quantifying chromatin-bound MCM protein in cells released from CDK4/6i G0/G1 arrest. We observed a decrease in chromatin-bound MCM7 protein in WEE1i-treated G1 cells compared with DMSO-treated cells (Fig. 2f).

In summary, these results demonstrate that WEE1 inhibits CDK2 activity during G1, and that an early and rapid increase in CDK2 activity after WEE1 inhibition can lead to reduced DNA replication origin licensing but not inhibition of APC/C^CDH1 activity. Increased CDK2 activity is insufficient to accelerate the transition of WEE1i-treated G1 cells into S phase, suggesting that WEE1 has other roles during G1.

As progression through G1 was abnormal in WEE1i-treated cells, we examined whether such cells entered S phase normally. We observed that mRuby–PCNA localization at the start of S phase was abnormal in WEE1i-treated cells, with low PCNA intensity and less well-defined PCNA foci compared with DMSO-treated cells (Fig. 3a). In WT cells, as S phase progresses, mRuby–PCNA localization changes, with early DNA replication factories (small puncta) coalescing into larger DNA replication factories (large foci). Thus, the standard deviation of mRuby–PCNA intensity normally increases throughout S phase. We quantified the standard deviation of mRuby–PCNA nuclear intensity; this showed a clear reduction at the G1/S transition in WEE1i-treated cells compared with DMSO-treated cells (Fig. 3b). This decrease also remained throughout S phase, indicating altered S phase progression in the presence of WEE1i, consistent with previous reports^26,27,28.

PCNA recruitment to chromatin in S phase brings substrates, including DNA licensing factor CDT1 and CDK inhibitor protein p21, in contact with the CRL4^CDT2 E3 ubiquitin ligase. This occurs through substrate binding to PCNA through a PCNA interaction protein (PIP) motif²⁹, which leads to rapid substrate degradation and S phase progression. We assayed whether the change in PCNA chromatin recruitment in WEE1i-treated cells affected PIP-containing substrate degradation using an hTert-RPE1 cell line expressing the PIP-FUCCI reporter from an exogenous promoter and time-lapse imaging. We tracked the dynamics of the degradation of this construct, which contains the PIP motif of CDT1 fused to an mVenus fluorophore, at the G1/S transition. In WEE1i-treated cells, the degradation of PIP-mVenus was slower compared with that in DMSO-treated cells (Fig. 3c), suggesting that impaired recruitment of PCNA to chromatin at the start of S phase affects the rate of degradation of PIP-containing substrates.

p21 contains a PIP motif; therefore, p21 protein is normally rapidly degraded at S phase entry^30,31,32, remains low throughout S phase and reaccumulates at the S/G2 transition when CRL4^CDT2 is inhibited by CDK1 (ref. ³³). As PIP-mVenus was more slowly degraded at the G1/S transition after WEE1 inhibition, we proposed that normal p21 protein degradation might also be impaired. To test this hypothesis, we used an hTert-RPE1 mRuby–PCNA cell line expressing endogenously labelled p21–GFP³⁴. We observed that in WEE1i-treated cells, p21–GFP protein accumulated abnormally during S phase (Fig. 3d, Extended Data Fig. 4a,b and Supplementary Video 2). This suggested continuing impaired p21 degradation in S phase, not just at the G1/S transition, upon WEE1 inhibition. We further proposed that WEE1 inhibition in S phase would lead to increased CDK1 activity and premature inactivation of CRL4^CDT2 (ref. ³³). We tested whether treating cells with a CDK1 inhibitor would restore CRL4^CDT2 activity during S phase and prevent accumulation of p21. Combining WEE1 and CDK1 inhibition prevented accumulation of p21–GFP in S phase cells treated with WEE1i alone (Fig. 3e and Extended Data Fig. 4c,d).

If p21 is not degraded during S phase, it can cause severe disruption to DNA replication and DNA damage³⁵. Cells treated with WEE1i had increased levels of DNA damage compared with DMSO-treated cells (Fig. 3f and Extended Data Fig. 4e). These results indicate that abnormal p21 protein accumulation may contribute to DNA damage in WEE1i-treated cells.

High levels of p21 protein can cause cell cycle arrest in G2. We tracked the cell cycle fate of hTert-RPE1 mRuby–PCNA cells released from CDK4/6i G0/G1 arrest and treated with WEE1i. We observed an increase in the fraction of cells arresting in G2 in WEE1i-treated cells compared with DMSO-treated cells and determined that this was p21-dependent (Fig. 3g,h), whereas the increase in G1 arrest after WEE1i treatment was not p21-dependent (Fig. 3g,h). WEE1i-treated cells do not completely replicate their DNA during S phase; therefore, more accurately, this was a ‘G2-like’ arrest³⁶. We confirmed that this was the case by examining cellular DNA content: we observed a large accumulation of cells between 2n and 4n specifically in WEE1i-treated cells (Extended Data Fig. 4f,g).

These data show that WEE1 inhibition leads to an abnormal G1/S transition that is exacerbated by the detrimental effects of WEE1 on DNA replication.

We next aimed to determine the mechanism preventing WEE1-inhibited G1 cells with high CDK2 activity from entering S phase prematurely. As APC/C^CDH1 seemed to be inactivated with normal rate and timing around the G1/S transition in WEE1i-treated cells, we considered whether WEE1 inhibition could cause increased or prolonged APC/C^CDH1 activity during G1 that could lead to a delay in S phase entry. This hypothesis was consistent with a previous report showing that WEE1 physically interacts with and can inhibit the ubiquitin ligase activity of APC/C^CDH1 (ref. ³⁷).

First, we revised our ordinary differential equation model to test the impact of WEE1 inhibition of APC/C^CDH1 on the G1/S regulatory network (Fig. 4a). We considered three hypotheses for this interaction: first, that WEE1 inactivates CDH1 through a mechanism identical to that of CDK2 (‘CDK2-like model’), that is, WEE1 phosphorylates the same sites in CDH1 as in CDK2 (Extended Data Fig. 5a,b); second, that WEE1 does not inactivate CDH1 directly but rather increases the affinity of cyclin E–CDK2 for APC/C^CDH1 (‘threshold model’), which has otherwise been shown to be negligible in vitro^20,38 (Extended Data Fig. 5e,f); and, third, that WEE1 inactivates CDH1 independently of CDK2, by phosphorylating separate amino acid residues in CDH1 (‘independent model’) (Extended Data Fig. 5i,j).

All three models reproduced G0 arrest upon complete WEE1 depletion (Fig. 1d,e and Extended Data Fig. 5d,h,l). However, they differed markedly with respect to the emergent APC/C^CDH1 activity during G1 and at G1/S upon partial (50%) WEE1 inhibition. Thus, each set of assumptions made distinct, experimentally verifiable predictions. The threshold model predicted a gradual inactivation of APC/C^CDH1 around the G1/S transition (Extended Data Fig. 5f), which was inconsistent with our time-lapse data (Fig. 2c). This left the CDK2-like and independent models of CDH1 inhibition by WEE1. These two models differed principally in terms of the change in APC/C^CDH1 activity during G1 after WEE1 inhibition (Extended Data Fig. 5a,b,i,j). In the CDK2-like model, APC/C^CDH1 activity remained unchanged after WEE1 inhibition, although it was prolonged, whereas in the independent model, APC/C^CDH1 activity increased after WEE1 inhibition. To differentiate between the two possibilities, we quantified levels of APC/C^CDH1 activity during G1 in WEE1i-treated cells.

To estimate APC/C^CDH1 activity in G1 cells, we used time-lapse imaging to quantify the levels of two APC/C^CDH1 substrates that accumulate during G1: CDC6 (ref. ³⁹) and CDC25A (ref. ⁴⁰). We considered using the mCherry–geminin reporter, but as geminin is very efficiently degraded by APC/C^CDH1, the mCherry–geminin signal intensity was negligible in G1 (Extended Data Fig. 3c). Quantification of the peak intensity of nuclear CDC6–mVenus protein from our previous analysis showed that CDC6–mVenus protein levels were decreased in WEE1i-treated G1 cells compared with DMSO-treated cells (Fig. 2e). This was owing to decreased CDC6–mVenus protein stability in WEE1i-treated cells (Extended Data Fig. 3d,e). To analyse CDC25A protein levels in real time, we generated an hTert-RPE1 mRuby–PCNA cell line in which we introduced a GFP fluorophore into the carboxyl terminus of the endogenous CDC25A gene (Extended Data Fig. 6a). We validated these cells (Extended Data Fig. 6b–f). In unperturbed cells released from CDK4/6i G0/G1 arrest, CDC25A–GFP showed accumulation around the G1/S transition (Fig. 4c and Supplementary Video 3). We confirmed that CDC25A–GFP was a reliable readout of APC/C^CDH1 activity in G1 by depleting CDH1 using siRNA, then quantifying CDC25A–GFP levels and showing that they were increased (Extended Data Fig. 6g). After WEE1i treatment in cells released from CDK4/6i G0/G1 arrest, CDC25A–GFP levels were much reduced, compared with those in DMSO-treated cells, around the G1/S transition (Fig. 4c). This reduction was due to a change in CDC25A protein stability, rather than a change in CDC25A transcription (Extended Data Fig. 6h,i). The decreased CDC6 and CDC25A protein levels after WEE1i treatment during G1 were consistent with increased APC/C^CDH1 activity after WEE1 inhibition and thus the independent model of WEE1 inhibition of CDH1 (Fig. 4b and Extended Data Fig. 5i–l).

To test whether increased APC/C^CDH1 activity caused a longer G1 phase in WEE1i-treated cells, we used siRNA to deplete CDH1. This led to a short G1 length (Fig. 4d and Extended Data Fig. 7a). CDH1 depletion and simultaneous inhibition of WEE1 also led to a short G1 length, consistent with our model and supporting the hypothesis that in WEE1i-treated cells, APC/C^CDH1 causes a prolonged G1 phase (Fig. 4d). Using an hTert-RPE1 cell line expressing exogenous, doxycycline-inducible EMI1 to inhibit APC/C^CDH1 (ref. ⁴¹), we observed a robust decrease in CDH1 protein levels and a decrease in the proportion of G0/G1-arrested cells in the presence of WEE1i (Fig. 4e and Extended Data Fig. 7b). Notably, we did not observe any increase in CDH1 protein levels or change in CDH1 stability in G1 cells after WEE1i treatment (Extended Data Fig. 7c–e), suggesting that WEE1 does not regulate CDH1 protein stability.

To test whether WEE1 phosphorylated CDH1 directly, we used in vitro kinase reactions with purified WEE1 and CDH1 proteins. WEE1 phosphorylated tyrosine(s) in CDH1 (Fig. 4f and Extended Data Fig. 7f) in a time-dependent manner, whereas phosphorylation of related APC/C coactivator CDC20 was negligible (Extended Data Fig. 7g). Mass spectrometry analysis showed that Tyr91 and Tyr130 of CDH1 were phosphorylated by WEE1. We re-analysed the samples using a targeted liquid chromatography–mass spectrometry approach, focusing acquisition on the identified phosphorylated peptides. This provided exceptionally high-quality fragmentation data, allowing unambiguous localization of the phosphorylation sites and demonstrating phosphorylation of Tyr91 and Tyr130 of CDH1 by WEE1 (Extended Data Fig. 7h,i). Tyr91Phe–Tyr130Phe CDH1 mutants had substantially reduced WEE1-driven pTyr phosphorylation, indicating that these sites contribute substantially to CDH1 phosphorylation, although WEE1 may phosphorylate other sites (Extended Data Fig. 7j). These sites, particularly Tyr130, are conserved across species (Extended Data Fig. 8), and both phosphorylation events have previously been detected in cells, although the kinase responsible was not identified^42,43.

Scrutinizing the structure of human APC/C^CDH1 (ref. ⁴⁴), we found that the Tyr91 and Tyr130 sites were part of a CDH1 amino-terminal extension that interacted extensively with APC/C. Tyr91 sat at an interface between CDH1 and APC1 and formed a hydrogen bond with His1231 in APC1 (Fig. 4g). Tyr130 formed intramolecular hydrogen bonds with Asp46 and Arg47 and seemed to stabilize the C-box motif in CDH1 (ref. ⁴⁵) (DRFIP), which has a crucial role in maintenance of the interaction with APC8b (ref. ⁴⁶) (Fig. 4h). These observations suggest that phosphorylation of CDH1 at Tyr91 and Tyr130 could weaken the engagement of CDH1 with APC/C and thereby reduce APC/C^CDH1 activity (Fig. 4i). To test this, we examined whether phosphomimetic substitution of the WEE1-targeted residues affected APC/C binding by CDH1. In APC/C binding assays in Xenopus extract, both CDH1 Tyr91Glu and CDH1 Tyr130Glu showed reduced association with APC/C, with Tyr130Glu consistently causing the stronger defect. Notably, the CDH1 Tyr91Glu–Tyr130Glu double mutant was almost completely defective in APC/C binding (Fig. 4j). Consistent with this binding defect, in vitro cyclin destruction assays showed that CDH1 Tyr130Glu was impaired in APC/C activation, whereas CDH1 Tyr91Glu caused only a modest defect. The double phosphomimetic mutant was severely compromised (Fig. 4k). These findings support a model in which WEE1-mediated phosphorylation of CDH1 inhibits APC/C^CDH1 by weakening CDH1 engagement with APC/C.

On the basis of our data and modelling, we propose a mechanism of G1/S control whereby WEE1 phosphorylates and inhibits APC/C^CDH1, independently of CDK2-mediated inhibition of CDH1, to control G1 length and promote robust S phase entry.

Having shown that WEE1 may suppress APC/C^CDH1 activity and maintain low levels of p21 protein during S phase, we next investigated the implications of these findings for WEE1i treatment in the clinic. To determine whether WEE1 inhibition resulted in similar phenotypes in cancer cells to those observed in non-transformed hTert-RPE1 cells, we repeated experiments in TP53-wild-type (WT) KU1919 bladder cancer and MCF7 breast cancer cell lines engineered to express endogenously labelled mRuby–PCNA. These cell lines were chosen as they have a functional p53–p21 pathway and are sensitive to CDK4/6i, which enabled us to robustly arrest cells in G0/G1. G1 length increased in WEE1i-treated KU1919 and MCF7 cells, and cells of both lines underwent significant G1 and G2 arrest after treatment with WEE1i, similar to phenotypes of hTert-RPE1 cells (Fig. 5a,b and Extended Data Fig. 9a–f). G2 arrest was p21-dependent, and we observed increased cell death during G2 in WEE1i-treated cells compared with DMSO-treated cells in the absence of p21 in all cell lines tested (Figs. 3g and 5a,c and Extended Data Fig. 9a). These data indicate that cells lacking p21 may be more sensitive to WEE1 inhibition. CDKN1A (encoding p21) is rarely mutated in cancers, except in bladder cancer, for which between 10% and 15% of patients have predicted loss-of-function mutations of p21 (ref. ⁴⁷). These mutations occur early and can be present in normal bladders⁴⁸. Therefore, we investigated whether WEE1i treatment could be used to eradicate pretumorigenic CDKN1A mutant lesions to prevent a subset of bladder cancers. We generated p21-knockout cell lines in the non-transformed bladder urothelial cell line HBLAK and found that these cells were indeed more sensitive to WEE1i than their WT counterparts (Fig. 5d and Extended Data Fig. 9g).

We reasoned that depletion of CDH1 might also sensitize cells to WEE1i by allowing the cells to enter S phase and incur WEE1i-mediated DNA damage. Indeed, siRNA-mediated depletion of CDH1 increased sensitivity of hTert-RPE1 cells to WEE1i (Fig. 5e).

To identify changes that could drive acquired resistance to WEE1i, we generated a WEE1i-resistant MCF7 cell line (Fig. 5f). We profiled these cells by RNA sequencing and found that WEE1i-resistant cells showed upregulation of nucleotide biosynthetic pathways, consistent with previous work showing that nucleoside supplementation reduces double-strand breaks caused by increased replication initiation in WEE1i-treated cells²⁶ (Extended Data Fig. 9h–j and Supplementary Data Table 1). WEE1i-resistant cells also showed downregulated TP53 and CDKN1A expression, as well as components of APC/C^CDH1 (ANAPC7 and ANAPC10, encoding APC7 and APC10, respectively) (Fig. 5g). These observations were consistent with our mechanistic findings, as cells would need downregulation of the p53–p21 and APC/C^CDH1 pathways to enable continued proliferation in the presence of WEE1i. However, loss of these proteins would lead to increased sensitivity to acute WEE1i treatment, as we observed for p21 and CDH1 (Fig. 5c,e,h). Analysis of DepMap for genes that synergized with WEE1 loss identified several APC/C^CDH1 components (Extended Data Fig. 9k).

In conclusion, our data suggest that CDKN1A mutation status or APC/C^CDH1 subunit expression could be used to improve patient selection and potentially reduce toxicity in WEE1i clinical trials.

The substantial side effects associated with WEE1i treatment reported in clinical trials, including myelosuppression, gastrointestinal issues and fatigue, have been attributed to potential off-target effects on PLK1. An alternative explanation is that WEE1 has other functions. We report a role for WEE1 in the regulation of re-entry into the cell cycle from quiescence. This could account for some of the effects of WEE1i observed in healthy tissues.

Our findings lead to an apparent conundrum, in that WEE1 inhibits CDK2 to prevent S phase entry yet also inhibits APC/C^CDH1 to promote S phase entry. We propose that WEE1 overall promotes S phase entry. By inhibiting APC/C^CDH1 activity, WEE1 permits CDC25A protein to accumulate; this in turn reverses WEE1 inhibitory phosphorylation of CDK2. WEE1 may be required for initial inhibition of APC/C^CDH1 before CDK2-mediated CDH1 phosphorylation and degradation. Future work involving mutation of WEE1 and CDK phospho-sites in CDH1 and identification of tyrosine phosphatases that act on CDH1 could determine the relative importance of each pathway with respect to S phase entry. mTOR-mediated phosphorylation of CDH1 Thr129 transiently inactivates APC/C^CDH1 (ref. ⁴³). This site is adjacent to CDH1 Tyr130, which is phosphorylated by WEE1, and phosphorylation of Ser131 has also been detected in cells⁴²; these findings indicate a potential key role for this motif in regulation of APC/C^CDH1 activity, probably involving stabilization of the interaction of the C-box in CDH1 with APC8. mTOR inhibition decreased Thr129 and Tyr130 phosphorylation in cells; however, mTOR only phosphorylates Thr129 (ref. ⁴³). Therefore, mTOR phosphorylation may precede WEE1 phosphorylation of CDH1 Tyr130. Together, these data indicate that APC/C^CDH1 may function as a signal integrator in G1 to determine whether cells commit to cell proliferation.

Our experimental data are most consistent with the model in which WEE1 inhibition of APC/C^CDH1 activity is independent of CDK2-mediated APC/C^CDH1 inhibition. However, there remain two discrepancies. First, although both modelling and experimental data showed reduced CDC25A levels during G1 after WEE1 inhibition, the model predicts that CDC25A levels should increase after S phase entry, owing to inactivation of APC/C^CDH1 at the G1/S transition. This post-G1/S increase in CDC25A does not occur in cells. This is likely to be because Cdc25A is degraded by CHK1-mediated phosphorylation in S phase cells⁴⁹ (Extended Data Fig. 10a). Second, whereas the model is consistent with our data in that cells remain arrested in G0/G1 at very low levels of WEE1 activity, the model predicts high CDK2 activity and inhibition of Rb during the arrest, which we did not observe (Extended Data Fig. 10b). The reasons for this are unclear but could involve reduced CAK activity in arrested cells⁵⁰.

WEE1i were designed to be synthetically lethal with TP53 mutations, which would cause DNA-damage-induced p53-mediated p21 expression to be inactive. However, WEE1i show efficacy in TP53-WT tumours; this demonstrates a need to understand the parameters that define WEE1i sensitivity. Here we identified expression of p21 and APC/C^CDH1 complex components as potential indicators of sensitivity to WEE1i. Acute loss of p21 or CDH1 sensitizes cells to WEE1 inhibition, whereas reduced p53–p21 and APC/C^CDH1 expression in resistant cells may reflect adaptation to chronic drug exposure. These findings indicate that CDKN1A status and APC/C^CDH1 pathway activity may merit further investigation as biomarkers for WEE1i-based therapies, including in CDKN1A mutant bladder cancer. Our data also suggest that WEE1i toxicity may arise, at least in part, from increased APC/C^CDH1 activity in quiescent stem cells, which impairs their ability to re-enter proliferation and maintain tissue function.

Cell culture

All cell lines used are listed in Supplementary Table 1.

Growth media

hTert-RPE1 and MCF7 cells were cultured at 37 °C and 5% CO₂ with the corresponding growth conditions: DMEM (Gibco, 11594486) supplemented with 10% fetal bovine serum (FBS; Gibco F7514) and 1% penicillin–streptomycin (Gibco 15140); KU1919 and HBLAK cells were cultured under the same environmental conditions, but the media were RPMI 1640 (Gibco, A1049101) and Cnt Prime (Cellntec, CnT-PR), respectively.

Cell maintenance

All cell lines were routinely tested for mycoplasma using an enzyme-linked immunosorbent assay test kit (MycoAlert, LT07-318). Cells in culture were routinely split 1:10 every 3–4 days. For cell splitting, cells were washed with phosphate-buffered saline (PBS) and detached by incubation for 3 min with Gibco Trypsin-EDTA (0.25%) and phenol red (Gibco; 25200056) before being replated in fresh media. Accutase (Thermo Fisher Scientific, 00-4555-56) was used for HBLAK cells.

Cell synchronization

Contact inhibition

hTert-RPE1 cells were arrested in G0/G1 by contact inhibition so that they reached 100% confluency for 3 days in a 6-well plate. After 6 days, cells were replated at low cell density in serum-free DMEM for 24 h, according to the protocol described in ref. ²⁵. Cells were released into the cell cycle with addition of serum-containing DMEM (as above for growth medium).

CDK4/6i

hTert-RPE1 cells were treated with 0.5–1 µM palbociclib (Selleckchem, S4482) for 24 h to arrest cells in G0/G1 (ref. ⁵¹). Cells were washed three times with PBS to release cells back into the cell cycle with addition of full growth medium. KU1919 cells were treated with 1 µM palbociclib for 24 h, then washed three times with PBS to release cells back into the cell cycle with addition of full growth medium. MCF7 cells were treated with 1 µM palbociclib for 24 h, before being washed three times with PBS and released into the cell cycle with addition of appropriate medium plus serum.

Escherichia coli transformation

Chemically competent E. coli DH5a cells (NEB, C2987H) were used for transformation. Competent cells were thawed on ice for 30 min, after which 0.5 μg of the plasmid DNA of interest was added, followed by incubation on ice for a further 20 min. To facilitate the uptake of plasmid DNA, the cells were subjected to heat shock for 45 s at 42 °C in a water bath and subsequent incubation on ice for 2 min. After this, 250 μl of room-temperature SOC medium was added to the cells, and the mixture was incubated in a horizontal shaker at 37 °C for 1 h. The bacteria were then spread on the appropriate antibiotic-containing LB agar plates and incubated at 37 °C overnight.

Generation of CDH1 mutant plasmids

Generation of CDH1 mutant plasmids carrying Y91E, Y130E and T129D

CDH1 mutant plasmids carrying Y91E, Y130E or T129D or combinations of these mutations are listed in Supplementary Table 2. All point mutations were introduced in the HA-CDH1 plasmid (Addgene, 11596) by Gibson assembly with Gibson Assembly Master Mix (E2611S, NEB), according to the manufacturer’s instructions. The Y91E fragments were PCR-amplified with forward primer 5′-ctggggacgtcggagcaagc-3′ and reverse primer 5′-agggcagaCTCGGCCAGGCCGTCTTTG-3′ for the insert and forward primer 5′-ggccGAGTCTGCCCTGCTCAAGAATG-3′ and reverse primer 5′-gctccgacgtccccaggcagaatgg-3′ for the backbone. The Y130E fragments were PCR-amplified with forward primer 5′-ctggggacgtcggagcaagc-3′ and reverse primer 5′-gggaCTCcgtgaacagacccttcttc-3′ for the insert and forward primer 5′-ggtctgttcacgGAGtcccttagcacca-3′ and reverse primer 5′-gctccgacgtccccaggcagaatgg-3′ for the backbone. The Y91E/Y130E fragments were PCR-amplified with forward primer 5′-ggccGAGTCTGCCCTGCTCAAGAATG-3′ and reverse primer 5′-gggaCTCcgtgaacagacccttcttc-3′ for the insert and forward primer 5′-ggtctgttcacgGAGtcccttagcacca-3′ and reverse primer 5′-agggcagaCTCGGCCAGGCCGTCTTTG-3′ for the backbone. The Y91E/T219D/Y130E fragments were PCR-amplified with forward primer 5′-ggccGAGTCTGCCCTGCTCAAGAATG-3′ and reverse primer 5′-gggaCTCGTCgaacagacccttcttctcaggc-3′ for the insert and forward primer 5′-ggtctgttcGACGAGtcccttagcacca-3′ and reverse primer 5′-agggcagaCTCGGCCAGGCCGTCTTTG-3′ for the backbone. Subsequently, CDH1 WT and mutated sequences were subcloned in the pHY22 (ref. ⁵²) plasmid backbone by Gibson assembly. The pHY22 backbone was PCR-amplified using forward primer 5′-GGATCGATCCTCTAGAGTCGACCT-3′ and reverse primer 5′-GCCATATGCGGTAAAATAACGTTTA-3′. HA-CDH1 WT and mutated sequences were PCR-amplified using forward primer 5′-TTTACCGCATATGGCatggaccaggactatgagcg-3′ and reverse primer 5′-CTAGAGGATCGATCCttaccggatcctggtgaagagg-3′.

Generation of CDH1 N186 mutant plasmids carrying Y91F, Y130F and Y91F/Y130F

Plasmids expressing the N-terminal 186 residues of human CDH1 (WT, Y91F, Y130F and Y91F/Y130F) with a C-terminal His6 tag were generated by PCR amplification of the corresponding double-stranded gBlocks (Integrated DNA Technologies, BVBA) using primers containing 5′ NdeI and 3′ BamHI sites. The PCR products were digested with NdeI and BamHI and subcloned into the pET-NGST vector. All constructs were confirmed by DNA sequencing.

Cell line generation

Generation of endogenous WEE1–GFP–AID-tagged hTert-RPE1 cell line

The GFP–AID–P2A–Neo tag (a gift from Francis Barr’s laboratory⁵³) was inserted at the C terminus of the endogenous WEE1 coding sequence using CRISPR–Cas9-mediated editing.

gRNA cloning. The guide RNA (gRNA) used was forward 5′-ACCTCCCCTGAACACTGTG-3′, reverse 5′-CACAGTGTTCAGGGGGAGGT-3′. This gRNA sequence was cloned into pX459 pSpCas9 using the Bbsl restriction enzyme. pSpCas9(BB)−2A-Puro (PX459) V2.0 was a gift from F. Zhang (Addgene plasmid 62988)⁵⁴.

Homology donor cloning. The genomic region surrounding WEE1 exon 11 (C terminus) was PCR-amplified from hTert-RPE1 genomic DNA using forward primer 5′-agtatttggcctattaaatttgtgttg-3′ and reverse primer 5′-tgcaaattgaagtgataatgttatttc-3′. WEE1 exon 11 was cloned into the pJet vector (Thermo Fisher), and the sequence was verified by Sanger sequencing. For construction of the homology donor plasmid, WEE1 homology arms were PCR-amplified from pJet–WEE1 exon 11 using high-fidelity Q5 DNA polymerase (NEB), according to the manufacturer’s instructions. For PCR amplification of the left homology arm, the following primers were used (the 5′ overhangs for Gibson assembly are underlined): forward primer 5′-acggtatcgataagcttgatattcttttgatttattgcag-3′ and reverse primer 5′-gccccaccaccgctccaccgtatatatagtaaggctgacag-3′. For PCR amplification of the right homology arm, forward primer 5′-cttcttgacgagttcttctgagactactcctttcacacctcc-3′ (the base mutated to change the PAM site in the recombined gene is shown in bold) and reverse primer 5′-ccgggctgcaggaattcgatgataaaaatatattttgtg-3′ were used. GFP–AID–GSG–P2A–Neo was PCR amplified from plasmid pFB8580 (provided by Francis Barr’s laboratory⁵³) using forward primer 5′-ggtgggagcggtggtggcg-3′ and reverse primer 5′-tcagaagaactcgtcaagaaggcgat-3′. The pBluescript (pSKII) vector was digested with EcoRV.

All PCR products were run on 0.8% TBE agarose gels, and bands were gel extracted and assembled into EcoRV-digested pSKII(-) using Gibson assembly with a NEB HiFi Assembly kit, according to the manufacturer’s instructions for four-way assembly. The final homology donor plasmid and the pX459 plasmid containing the gRNA were sequenced using Sanger sequencing to confirm the correct sequence.

Generation of WEE1–GFP–AID-tagged cell line. For generation of the tagged cell lines, hTert-RPE1 MYC-OsTIR1 cells (expressing a doxycycline-inducible MYC–OsTIR1 construct; a gift from H. Hochegger⁵⁵) were seeded in a 6-well plate at 5 × 10⁴ cells per well in 1.5 ml of growth medium, 1 day before transfection. The next day, cells were forward transfected with Lipofectamine 2000 and 500 ng of pX459-gRNA and 500 ng of pSK-WEE1–GFP–AID–P2A–Neo. Cells were allowed to recover for 24 h before trypsinization and seeding in a 15-cm plate with fresh growth media. The next day, 1 mg ml⁻¹ of G418 (Thermo Fisher) was added to the plate; then, fresh medium plus G418 was added every 3–4 days until single colonies became visible. Single-cell colonies were picked using cloning cylinders (Sigma). Positive clones were identified by nuclear GFP fluorescence and analysed by western blot. We confirmed that this cell line had the expected nuclear localization of WEE1–GFP (Extended Data Fig. 2b) and that it grew with normal kinetics (Extended Data Fig. 2d).

Generation of endogenous CDC25A–GFP-tagged hTert-RPE1 cell line

gRNA cloning. For generation of the gRNA targeting the C terminus of human CDC25A for CRISPR–Cas9-mediated gene-tagging, gRNA forward primer 5′-caccgaagaagctctgagggcggc-3′ and gRNA reverse primer 5′-aaacgccgccctcagagcttcttc-3′ were annealed and ligated into BbsI-digested pSpCas9(BB)−2A-Puro (PX459) V2.0.

Homology donor cloning. For PCR amplification of the CDC25A left and right homology arms from hTert-RPE1 genomic DNA, we used NEB Q5 High Fidelity DNA polymerase and the following primers: for the left homology arm (the 5′ overhangs for Gibson assembly are underlined), forward 5′-acggtatcgataagcttgatgtgaaggtgctgggcaggactc-3′ and reverse 5′-cgccaccaccgctcccaccgagcttcttcagacgactgtac-3′; for the right homology arm (the base mutated to remove the PAM sequence in recombined DNA is shown in bold), forward 5′-atcgccttcttgacgagttcttctgagggcggcatgaccagccagcag-3′ and reverse 5′-ccgggctgcaggaattcgatcagagcacctagctttctgtcc-3′. GFP–GSG–P2A–Neo was PCR amplified from plasmid pFB8606 (provided by Francis Barr’s laboratory⁵³), using the following primers: P2A-CT linker forward 5′-ggtgggagcggtggtggcg-3′ and P2A-CT Neo reverse 5′-tcagaagaactcgtcaagaaggcgat-3′.

All PCR products were run on 0.8% TBE agarose gels, and bands were gel extracted and assembled into EcoRV-digested pSKII(-) using Gibson assembly with a NEB HiFi Assembly kit, according to the manufacturer’s instructions for four-way assembly. The final homology donor plasmid and the pX459 plasmid containing the gRNA were sequenced using Sanger sequencing to confirm the correct sequence.

Generation of CDC25A–GFP-tagged cell line. hTert-RPE1 mRuby–PCNA cells⁵⁶ in a six-well plate (Sarstedt, 83.3920) were forward transfected with Lipofectamine 2000 and 500 ng of pX459-gRNA and 500 ng of pSK-CDC25A-CT-GFP, according to the manufacturer’s protocol, and left to recover overnight. The next day, cells were split into 15-cm tissue culture dishes (Corning, 420599) and, 24 h later, clones were selected with 1 mg ml⁻¹ G418. Media and antibiotic were refreshed every 3–4 days. After 2 weeks, single colonies were selected using cloning cylinders (Sigma) and transferred to 24-well plates. Cells were expanded and initially screened for GFP positivity in the nucleus. Genomic DNA was extracted from cells with GFP-positive nuclei, using a Qiagen DNeasy blood and tissue kit, according to the manufacturer’s instructions. To check for targeted integration of the C-terminal tagging construct, the following primers were used: 5′-atgtacagtcgtctgaag and 5′-taccacagcgcctcctgg. All PCR products were sequenced by Sanger sequencing to check for in-frame tagging. Clones were then further validated by western blotting, growth curves and single-cell imaging to check cell cycle phase lengths. We obtained several heterozygously tagged clones that had a slightly decreased proliferation rate, mainly owing to an increase in G1 length (Extended Data Fig. 6b–e). CDC25A–GFP localized correctly to the nucleus (Extended Data Fig. 6f). In unperturbed cells released from CDK4/6i G0/G1 arrest, CDC25A–GFP showed accumulation around the G1/S transition, as previously described⁵⁷.

PCNA tagging in hTert-RPE1 MYC-OsTir1 WEE1–GFP–AID–Neo cells

For endogenous expression of the PCNA reporter, the sequence encoding fluorescent protein mRuby was inserted in the first exon of one allele of the PCNA locus by recombinant adeno-associated virus (rAAV)-mediated homologous recombination⁵⁶ in the hTert-RPE1 MYC-OsTir1 WEE1–GFP–AID–Neo cell line. Then, 293T cells were seeded in a T-75 flask so that they reached 50–70% confluence at the time of transfection. pAAV–mRuby–PCNA (3 μg; gift from Joerg Mansfeld, ICR⁵⁶) and pRC and pHelper plasmids (9 μg total) were mixed with 3.75 ml of Opti-MEM (Invitrogen), after which 19 μl of PLUS Reagent (Invitrogen) was added, and the mixture was incubated for 5–10 min at room temperature. Next, 47 μl of Lipofectamine LTX (Invitrogen) was added to the tube, followed by incubation for 30 min. 15 minutes before the end of the incubation, 293T cells were washed twice with PBS; then, 7.5 ml of Opti-MEM (Invitrogen) was added, and the cells were incubated at 37 °C for 15 min. After this, 1.5 ml of DNA/lipid mixture was added to the cells, followed by incubation at 37 °C for 4–6 h, and then the medium was replaced with complete DMEM. 2–3 days postinfection, cells in medium (and any floating cells) were transferred to a sterile 50-ml Falcon tube. Then, 5 ml of fresh medium was added to the flask, and adherent cells were scraped off and transferred to the same 50-ml Falcon tube. Cells were lysed with three rounds of freezing and thawing (10 min in dry ice and ethanol + 10 min in a water bath at 37 °C). After each cycle, cells were vortexed at full speed for 1 min. Cells were then centrifuged at 10,000g for 30 min at 4 °C. Tubes were wiped with ethanol and placed in the tissue culture hood. Cell-free supernatants were collected, and rAAV preparations were dispensed in small aliquots (3–5 ml). hTert-RPE1 MYC-OsTir1 WEE1–GFP–AID–Neo cells were seeded in a T-75 flask so that they reached 30–40% confluence at the time of infection. The T-75 flask was washed twice with PBS; then, 5 ml of complete medium and 5 ml of rAAV stock were added, and the flask was left to incubate at 37 °C for 4 h. Fresh media were added to bring the volume up to 15 ml (or more), and infection was allowed to continue for 48 h. After 3 days, mRuby–PCNA-positive cells were sorted to single-cell density in 50:50 conditioned/growth media in 96-well plates by fluorescence-activated cell sorting (FACS). Cells were left to expand for 2 weeks, and mRuby–PCNA-positive clones were identified by microscopy.

Generation of p21-knockout cell lines

p21-knockout MCF7, KU1919 and HBLAK cell lines were generated as previously described³⁴. The only modification was that HBLAK cells were transfected by nucleofection on the Amaxa nucleofector, using the MCF7 protocol. Knockouts were validated by immunostaining and western blotting for p21 after Nutlin-3 treatment, to boost the p21 signal.

siRNA transfection

All siRNAs used are listed in Supplementary Table 4. For live imaging and western blotting experiments, cells were reverse transfected with siRNA in 384-well PhenoPlates (Revvity) and 24-well plates (Greiner Bio-One, 662165), respectively. For imaging experiments, transfection was carried out 24 h after CDK4/6 inhibition in a total volume of 30 µl of medium and used at a final concentration of 20 nM. Transfection was performed using Lipofectamine RNAiMAX (Invitrogen 13778150), according to the manufacturer’s instructions. In brief, 40 nl of Lipofectamine RNAiMAX was added to 40 nl of 20 µM siRNA in 10 µl Opti-MEM (Gibco 31985062), and 10 µl of the mixture was subsequently added to cells. Following transfection, cells were incubated for 6–24 h, depending on the experiment. For western blotting, this protocol was carried out with the following differences: cells were grown in 500 µl of medium; and 1 µl of Lipofectamine RNAiMAX was added to 1 µl of 20 µM siRNA in 100 µl Opti-MEM, and the mixture was subsequently added to cells. Following transfection, cells were incubated for 6–24 h, depending on the experiment.

Immunostaining

All antibodies used for immunostaining are listed in Supplementary Table 5. Cells were grown in 96- or 384-well PhenoPlates (Revvity, 6055302 and 6057302, respectively). EdU was added to a final concentration of 10 μM for either 12–24h at G0 release or with a 30-min pulse before fixation for identification of G0 and S phase cells, respectively. Cells were fixed in 4% formaldehyde–PBS for 15 min at room temperature, permeabilized with PBS–0.5% Triton X-100 for 15 min and blocked for 1 h in blocking buffer (PBS with 2% bovine serum albumin (BSA), filtered). Primary antibodies were diluted in blocking buffer and incubated overnight at 4 °C or for 2 h at room temperature. After three washes in PBS, cells were incubated for 1 h at room temperature in the dark with conjugated secondary antibodies (1:1,000; Invitrogen). For EdU incorporation, cells were incubated with 100 mM Tris-HCl pH 7.5, 4 mM CuSO₄, 100 mM ascorbic acid and 5 μM sulfo-cyanine-5 azide for 30 min at room temperature in the dark, washed three times in PBS and counterstained with 1 μg ml⁻¹ Hoechst for 10 min, followed by three washes in PBS. All plate-washing steps were performed on an automated 50TS microplate washer (BioTek). Plates were imaged using an Operetta CLS (Revvity) with a ×20 (numerical aperture (NA) 0.8) objective.

Automated image analysis of fixed cells

All automated image analysis of images acquired using Operetta CLS was performed using Harmony software (Revvity). Nuclei were segmented on the basis of Hoechst intensity, with nuclei at the edge of the field of view excluded from the analysis. The nuclear intensity of proteins was calculated as the mean intensity in the segmented region.

G0 calculation

G0 cells were identified by the absence of EdU staining. Briefly, the nuclear-to-cytoplasmic ratio was calculated by creation of a 4-pixel-width ring around the segmented nucleus, and the ring intensity was calculated as a proxy for the cytoplasmic portion. Nuclear-to-cytoplasmic ratios for individual cells were calculated and plotted in Prism (GraphPad). Cells with EdU nuclear-to-cytoplasmic ratios less than 1.2 were called as G0.

Western blotting

All antibodies used for western blotting are listed in Supplementary Table 4.

Whole-cell lysates were prepared following aspiration of medium from culture plates, followed by washing with PBS on ice. Lysates were collected in 1× Novex Tris-Glycine SDS sample buffer (Novex, LC2676) supplemented with 1× phosphatase inhibitors (Thermo Scientific, 1862495), 1× protease inhibitors (Thermo Scientific, 78429) and 1 mM dithiothreitol (DTT; BioUltra, 43816-10ML). Samples were incubated at 95 °C for 10 min, then centrifuged at 14,000g for 1 min before being loaded on a Novex 4–20% Tris-Glycine gel (Invitrogen, XP04205). Proteins were transferred to PVDF-FL membranes (Merck Life Sciences, IPFL00010), which were then blocked in blocking buffer (Tris-buffered saline (TBS), 5% milk, 10% glycerol, Tween 0.1%) for 1 h at room temperature and incubated with primary antibodies diluted in blocking buffer, overnight at 4 °C. Membranes were washed three times in TBS with 0.05% Triton X-100, then incubated for 1 h at room temperature with HRP-conjugated secondary antibodies diluted in blocking buffer. Membranes were then washed three times in TBS with 0.05% Triton X-100 and developed using Clarity Western ECL Substrate (Bio-Rad, 1705061) unless secondary antibodies were conjugated to Alexa Fluor. Blots were imaged on an Amersham Imager 680.

Doxycycline-inducible WEE1 degradation

For live cell imaging involving WEE1 degradation with doxycycline and IAA, RPE1 WEE1–GFP–AID mRuby–PCNA cells were plated in 384-well PhenoPlates (Revvity) at a density of 500 cells per well in 20 µl medium. Cells were synchronized with 1 µM CDK4/6i for 24 h, and doxycycline and IAA (doxycycline: 1 µg ml⁻¹; IAA: 500 µM) was added to the growth medium when cells were released. For western blots, doxycycline or IAA was also added at release.

In vitro kinase assays

All purified proteins used for in vitro kinase assays are listed in Supplementary Table 6.

To confirm the activity of the WEE1 kinase, we carried out a positive control experiment with cyclin B1–CDK1. The in vitro kinase protocol was provided by J. Pines (ICR). For this reaction, 100 ng of recombinant cyclin B1–CDK1 (Abcam) was diluted in 5 µl of kinase buffer (200 mM Tris pH 7.5, 150 mM NaCl, 10 mM MgCl₂, 1 mM DTT) with 0.5 mM ATP and 18 µM RO-3306 (CDK1 inhibitor, Selleckchem). This mixture was added to 50 ng of recombinant WEE1 (Thermo Fisher Scientific) diluted in 5 µl of kinase buffer with 0.5 µM MK-1775 (WEE1i, Selleckchem) for the negative control. A negative control without addition of WEE1 kinase was used. The reactions were incubated for 30 min at 30 °C, stopped by addition of 5 µl of 3× SDS buffer with 100 mM DTT and boiled at 95 °C for 5 min. Samples were then used for western blotting as described, with the exception that milk was replaced with BSA at the blocking step. pY15-CDK1/2/3/5 and CDK1 antibodies were used to quantify pY15-CDK1 phosphorylation. For monitoring of phosphorylation of CDH1 on tyrosine residues by WEE1, 80 ng of recombinant Strep–CDH1 was diluted in 5 µl of kinase buffer with 0.5 mM ATP. This mixture was added to 50 ng of recombinant WEE1 diluted in 5 µl of kinase buffer with 0.5 µM MK-1775 for the negative control. A negative control without kinase addition was also carried out. The reactions were incubated for 30 min at 30 °C, stopped by addition of 5 µl of 3× SDS buffer with 100 mM DTT and boiled at 95 °C for 5 min. As for the cyclin B1–CDK1 reaction, samples were analysed by western blotting. The phospho-tyrosine antibody pTyr-100 and CDH1 antibody were used to quantify phosphorylation of tyrosine residues on CDH1. For the results shown in Extended Data Fig. 7g, 1.5 µg of bacterially expressed and purified Xenopus GST-FZR N186 (Cdh1) or GST-FZY N159 (Cdc20) protein⁵⁸ was incubated with 160 ng recombinant WEE1 kinase (Thermo Fisher Scientific) in 10 µl kinase buffer. Samples were collected at the indicated time points and analysed by SDS–PAGE and immunoblotting using an anti-phospho-tyrosine antibody (pTyr-100). For the results shown in Extended Data Fig. 7j, bacterially expressed and purified human GST–CDH1 N186 or its tyrosine variants were incubated with recombinant WEE1 kinase in the same kinase buffer for 10 min and analysed by SDS–PAGE and immunoblotting using the same anti-phospho-tyrosine antibody.

CDC6–mVenus protein stability assay

For monitoring of CDC6–mVenus stability upon WEE1 inhibition, hTert-RPE1 CDC6–mVenus DHB–mCherry mTurq2–PCNA cells were plated in a 24-well plate at a density of 50,000 cells per well. Cells were arrested with CDK4/6i for 24 h and treated with WEE1i (1 µM) or DMSO upon release. 2 h after release, cells were treated with cycloheximide (100 µg ml⁻¹) every 30 min for 2 h. Subsequently, cells were lysed in 1× Novex Tris-Glycine SDS sample buffer (Novex, LC2676) supplemented with 1× phosphatase inhibitors (Thermo Scientific, 1862495), 1× protease inhibitors (Thermo Scientific, 78429) and 1 mM DTT (BioUltra, 43816-10ML).

CDC6–mVenus and CDC25A protein levels upon proteasome inhibition

For monitoring of CDC6–mVenus and CDC25A protein levels upon WEE1 and proteasome inhibition, hTert-RPE1 CDC6–mVenus DHB–mCherry mTurq2–PCNA cells were plated in a 24-well plate at a density of 50,000 cells per well, or hTert-RPE1 cells were plated in a 6-well plate at a density of 300,000 cells per well. Cells were arrested with CDK4/6i for 24 h and treated with WEE1i (1 µM) or DMSO upon release. 2 h after release, cells were treated with proteasome inhibitor MG132 (10 µM) for 2 h. Subsequently, cells were lysed in 1× Novex Tris-Glycine SDS sample buffer (Novex, LC2676) supplemented with 1× phosphatase inhibitors (Thermo Scientific, 1862495), 1× protease inhibitors (Thermo Scientific, 78429) and 1 mM DTT (BioUltra, 43816-10ML).

CDC25A mRNA fluorescence in situ hybridization

For monitoring of CDC25A transcription levels, hTert-RPE1 cells were plated in a 96-well plate at a density of 5,000 cells per well. Cells were arrested with CDK4/6i for 24 h and treated with WEE1i (1 µM) or DMSO upon release. 4 h after release, cells were fixed in 4% formaldehyde in PBS for 15 min at room temperature. Cells were then processed with a ViewRNA-ISH cell assay kit (Thermo Scientific, QVC0001). In brief, cells were permeabilized with proteases; then, the CDC25A probe (VA6-3168938 ViewRNA Type 6 probe set CDC25A, Thermo Scientific) was hybridized for 3 h at 40 °C. Subsequently, the preamplifier was hybridized for 30 min at 40 °C, followed by hybridization with the amplifier for 30 min at 40 °C and finally hybridization with the label probe for 30 min at 40 °C. Nuclei were stained with DAPI, and the plate was imaged using an Operetta CLS (Revvity) with a ×40 water (NA 1.1) objective. The cytoplasm was imaged with the digital phase contrast channel. Cytoplasmic segmentation was performed with Harmony software (Revvity) using Find Cytoplasm method D, and cellular CDC25A mRNA foci were detected with Find Spots method A.

Expression and purification of Strep–CDH1

Expression and purification of Strep–CDH1 was performed using a method similar to that described in ref. ⁵⁹, with some modifications. Specifically, codon-optimized complementary DNAs of Strep–CDH1 in pFastBac1 vectors were ordered from GeneArt (Thermo Fisher Scientific) and transformed into MultiBac cells for expression in insect cells. The baculovirus was created in Sf9 cells; this virus was then used to infect Hi5 cells at a cell density of 1 × 10⁶ cells ml⁻¹, followed by incubation for 72 h at 27 °C, 130 rpm. All purification steps were carried out at 4 °C. First, the cell pellet was thawed on ice in wash buffer (50 mM Tris-HCl pH 7.3, 500 mM NaCl, 5% glycerol, 0.5 mM TCE)] supplemented with 5 units ml⁻¹ benzonase and an EDTA-free protease inhibitor (Roche). The cells were sonicated and then centrifuged for 1 h at 48,000g. The supernatant was then bound to a 5-ml StrepTactin Superflow Plus cartridge (Qiagen) using a flow rate of 1 ml min⁻¹. The column was washed extensively with wash buffer and then eluted using elution buffer (50 mM Tris-HCl pH 7.0, 200 mM NaCl, 5% glycerol, 0.5 mM TCEP and 2.5 mM desthiobiotin).

CDH1–APC/C binding assays

Hs CDH1 WT or tyrosine mutant proteins (Y91E, Y130E, Y91E/Y130E, or Y91E/T129D/Y130E), prepared using a coupled in vitro transcription–translation system (TnT, Promega), were mixed with Xenopus egg interphase extracts for 30 min at 23 °C. Endogenous APC/C was then immunoprecipitated using Dynabeads crosslinked to the AF3.1 anti-APC3 antibody. The immobilized APC/C was washed twice with 1× XBcsf-MS (10 mM HEPES-KOH pH 7.8, 50 mM sucrose, 300 mM KCl, 2 mM MgCl₂ and 5 mM EGTA) containing 0.01% NP-40, eluted in 1× SDS sample buffer and analysed by SDS–PAGE and immunoblotting.

Cell-free APC/C-dependent destruction assay

Destruction assays were performed essentially as described previously⁶⁰. APC/C substrates were labelled with [³⁵S]methionine/cysteine (Hartmann Analytic) using TnT (Promega, UK). CDH1-mediated destruction assays were carried out in Xenopus egg cell-free interphase extracts supplemented with Hs CDH1 WT or tyrosine mutant proteins prepared by TnT. Full-length cyclin B and cyclin B Δ67, which lacks the D-box, served as positive and negative controls, respectively. Samples were taken at the indicated time points and analysed by SDS–PAGE and autoradiography.

Proteomics

To determine the tyrosine residues phosphorylated by WEE1 on CDH1, the reactions (5 replicates per condition) were carried out as described in the ‘In vitro kinase assays’ section, with the exception that the reactions were stopped with addition of 1 M guanidine hydrochloride. We diluted 5 μl of the kinase reaction with 45 μl water containing TCEP (final concentration 10 mM) and chloracetamide (20 mM). The reaction was incubated at 37 °C to reduce and alkylate Cys, and digestion was initiated by addition of trypsin and LysC at final concentrations of 5 ng µl⁻¹ and 10 ng µl⁻¹, respectively. After overnight digestion the samples were desalted on 3 mg of oasis HLB buffer. The samples were evaporated in a SpeedVac and resuspended in 30 µl of 0.1% trifluoroacetic acid, and 5 µl was analysed using liquid chromatography–mass spectrometry.

Chromatographic separation was performed using an Ultimate 3000 RSLC nano liquid chromatography system coupled to a Exploris 240 mass spectrometer with an EASY-Spray source. Electrospray ionization was achieved with Bruker PepSep emitters (PN: PSFSELJ10, 10 µm). Peptide solutions were injected directly on to the analytical column (self-packed column, CSH C18 1.7-µm beads, 150 μm × 30 cm) at a working flow rate of 1.25 μl min⁻¹ for 4 min. Peptides were then separated using a 60-min stepped gradient: 0–45% of buffer B for 70 min (composition of buffer A: 95%:5% H₂O/DMSO + 0.1% formic acid; composition of buffer B: 75%:20%:5% MeCN/H₂O/DMSO + 0.1% formic acid), followed by column conditioning and equilibration. Eluted peptides were analysed in data-independent acquisition (DIA) mode: an initial MS1 scan was carried out at 140,000 resolution with an AGC target of 3e6 for a maximum IT of 200 ms, m/z range: 410–1,600. This was followed by 30 MS2 scans at 30 K resolution covering 410–1,600 with variable window sizes predicted by encyclopeDIA. The AGC target was set to 3e6 with maximum IT on auto. Normalized collision energy was set to 27%.

Data were processed using the Spectronaut software platform (Biognosys, 19.9.250512.62635) in direct DIA mode. Searches were carried out against Spodoptera frugiperda concatenated with the human WEE1 sequence, the Trichoplusia ni protein sequence database concatenated with the human CDH1 sequence and a universal protein contaminants database⁶¹. For the DIA analysis, higher stringency settings were used than in the default workflow as described previously⁶². Default settings were used for posttranslational modification analysis, with the probability cutoff for posttranslational modification localization set to 0, before table export. To account for CDH1 intensity variance at the protein level, phospho-site intensities were also normalized to the total protein level. Specific sites of interest were validated using the Spectronaut feature viewer.

The two identified phospho-sites were located on the DGLAY[phospho (Y)]SALLK–Y91 and GLFTY[phospho (Y)]SLSTK–Y130 peptides, both with 2+ charge. We reran the samples targeting the peptides in parallel reaction monitoring and single ion monitoring modes at 240 K resolution and 500 ms maximum injection time. The mass spectrometry proteomics data have been deposited at the ProteomeXchange Consortium through the PRIDE⁶³ partner repository with dataset identifier PXD068577.

Fluorescence-activated cell sorting

FACS of cells was performed at the Flow Cytometry Facility, MRC Laboratory of Medical Sciences (MRC LMS). Cells were washed with PBS and resuspended in FACS sorting buffer (1 ml of PBS, 2% FBS) to a concentration of 8–12 million cells ml⁻¹. Before sorting, cells were filtered through a 30- to 40-μM filter.

Growth curves

Cells were plated at a density of 1,000 cells per well in triplicate in 96-well tissue culture plates. Bright-field images were taken every 4 h for 7 days using a ×10 (NA 0.3) objective, and percentage confluency was calculated on a CellCyteX (Echo) live imaging system. For calculation of confluency, a mask was created using the default settings on the CellCyteX (brightness 0%, contrast 0%, contrast sensitivity 50 a.u., smoothing 2 a.u. and filled hole size 100 µm²). Results were plotted with Python (Seaborn package).

Drug dose curves

Cells were plated at a density of 500 cells per well in 20 µl of medium. For the drug dose curve assays with siRNA transfection, cells were transfected (10 µl) for 24 h before drug addition as described previously. MK-1775 was prepared by 7 × 1:3 serial dilutions (7 × 1:4 in the case of transfection) in growth medium with final concentrations for all assays of 100 µM, 10 µM, 1 µM, 0.1 µM, 0.01 µM and 0.001 µM. Ten microlitres were added per well. After 72 h of incubation, cells were fixed and stained for Hoechst. Plates were imaged using an Operetta CLS (Revvity) with a ×20 (NA 0.8) objective. Cell viability was calculated as a percentage, and normalized half-maximal inhibitory concentration (IC₅₀) values were computed using GraphPad Prism.

EMI1 overexpression

hTert-RPE1 CCNB1–eYFP mTurq2–EMI1 53BP1-mCherry cells (a gift from R. Medema⁶⁴) were seeded in a 384-well PhenoPlate (Revvity) at a density of 500 cells per well in 20 µl of medium. Cells were arrested for 24 h in CDK4/6i and then released with doxycycline (1 µg ml⁻¹) to induce EMI1 expression, EdU, and DMSO or WEE1i for 12 h before fixation for identification of G0 cells. Cells were fixed in 4% formaldehyde–PBS for 15 min at room temperature, permeabilized with PBS–0.5% Triton X-100 for 15 min and blocked for 1 h in blocking buffer (PBS–2% BSA, filtered). For EdU incorporation, cells were incubated with 100 mM Tris-HCl pH 7.5, 4 mM CuSO₄, 100 mM ascorbic acid and 5 μM sulfo-cyanine-5 azide for 30 min at room temperature in the dark, washed three times in PBS and counterstained with 1 μg ml⁻¹ Hoechst for 10 min, followed by three washes in PBS. Inhibition of CDH1 was confirmed by western blot. Cells were plated at a density of 40,000 cells per well in a 24-well plate, then arrested for 24 h with CDK4/6i and released with addition of doxycycline to induce EMI1 expression (1 µg ml⁻¹ and 10 µg ml⁻¹) for 4 h before cell lysis in 1× Novex Tris-Glycine SDS sample buffer (Novex, LC2676) supplemented with 1× phosphatase inhibitors (Thermo Scientific, 1862495), 1× protease inhibitors (Thermo Scientific, 78429) and 1 mM DTT (BioUltra, 43816-10ML).

Time-lapse imaging

Live cell imaging was performed on an Operetta CLS High-Content Analysis microscope (Revvity), with atmospheric control to maintain cells at 37 °C and 5% CO₂ and a breathable membrane (Thermo Fisher Scientific) to stop evaporation. Cells were plated in a 384-well PhenoPlate (Revvity) at a density of 500 cells per well in 20 μl of medium for 24 h (unless otherwise stated) before imaging. Cells were imaged using a ×20 (NA 0.8) objective at 10-min intervals for 48 h (unless otherwise stated) in 50 μl phenol red-free media. For hTert-RPE1 WEE1–GFP–AID mRuby–PCNA and hTert-RPE1 mRuby–PCNA CDC25A–GFP live imaging, cells were imaged on an automated inverted spinning disc confocal microscope system IX83 (Olympus) combined with CellVivo (Olympus). Cells were plated on IbiTreat μ-slide 8-well (ibidi) plates at a density of 5000 cells per well in 200 μl of medium or in a 384-well PhenoPlate (Revvity) at a density of 500 cells per well in 20 μl of medium, 24 h before imaging (unless otherwise stated). Images were typically acquired with a ×20 plan (UCPLFLN) fluorescence objective (NA 0.7).

Automated image analysis of live imaging experiments

Automated cell tracking (Nuclitrack)

Images were exported as tiff files and analysed using NucliTrack⁶⁵ and/or Fiji and Napari⁶⁶. All cells analysed were present throughout the entire movie unless they died. Spontaneous death was quantified manually. mRuby–PCNA was used to define cell cycle phases^34,56. We defined S phase entry as the time point at which the mRuby–PCNA mean intensity started to increase and mRuby–PCNA switched from being homogenously localized throughout the nucleus to having a more textured appearance; PCNA binds to the chromatin at the start of S phase, which can be measured by an increase in the standard deviation of mRuby–PCNA intensity. Example images around the G1/S transition and the output of mRuby–PCNA features around the G1/S transition from Nuclitrack are shown in Extended Data Fig. 1e.

As shown in Fig. 3b, the standard deviation of mRuby–PCNA was used to quantify progression through S phase. Cells in G1 had a low standard deviation, as the mRuby–PCNA intensity was homogenous throughout the nucleus. As cells progressed through S phase, the standard deviation of mRuby–PCNA intensity increased as many small foci coalesced into larger, brighter foci, leading to very dark and very bright areas of the nucleus.

Automated cell tracking (Cellpose + Ultrack)

The code used for single-cell time-lapse image segmentation and tracking was implemented in Python (Jupyter notebook) and is available at GitHub (https://github.com/Alexis-Barr-Lab). This analysis pipeline combined Cellpose⁶⁷ segmentation and tracking using Ultrack⁶⁸ and was run using the Jex HPC cluster at the MRC LMS⁶⁹. In brief, raw TIFF stacks (shape: T × Y × X) were background-subtracted using a rolling ball-style method with a Gaussian filter (σ = 15 px), followed by morphological reconstruction to separate foreground from low-frequency background. Frames were normalized (γ = 0.5) and stored in Zarr format for efficient chunked input–output. Normalized frames were processed with a Cellpose model (cpsam). Masks for each time point were generated, yielding per-frame two-dimensional label arrays written into a Zarr volume. A watershed was run on each frame to recover fine cell boundaries and split merged objects. Thresholding and binary morphology mask detections, followed by distance transform watershed with minima suppression, were performed to generate second label sets, which were also saved as Zarr. Graphics processing unit-accelerated boundary detection was implemented in CuPy; boundaries from both Cellpose and watershed masks were combined and smoothed (σ = 5) to create high-resolution contour maps for downstream linking. Centroids of watershed segments between frames were calculated. The Hungarian algorithm (linear_sum_assignment) was used to link cells, providing centroid-distance statistics (95th percentile, maximum, average and 25th percentile displacements) and informing Ultrack’s max_distance parameter. Ultrack was then configured and executed. For quality checks, tracks and labels were exported and visualized. Background subtraction was then applied to the channel of interest, and properties were exported.

Statistical analysis

Statistical analyses in this work were performed in GraphPad Prism unless otherwise stated.

Generation of MCF7 WEE1i-resistant cells

For generation of MCF7 WEE1i-resistant cells, MCF7 WT cells were plated on a 15-cm tissue culture dish (Corning, 420599) at a density of 5 × 10⁶ cells ml⁻¹. Cells were treated with 0.075 µM MK-1775 24 h later and incubated until they reached confluence. Cells were then split and treated with a double amount of MK-1775 24 h later and incubated again until confluence. This step was repeated until 0.6 µM MK-1775 was reached. The resistance of these cells was measured against the parental cells with a 72-h drug dose curve (IC₅₀).

RNA sequencing

MCF7-resistant and WT cells were plated at a density of 100,000 cells per well in a 6-well plate (Sarstedt, 83.3920). Twenty-four hours later, total RNA was extracted using an RNeasy MinElute Cleanup Kit (Qiagen 74204) according to the manufacturer’s instructions. Quality and concentration were assessed using the Agilent 2100 Bioanalyzer RNA 6000 Nano assay.

RNA sequencing libraries were constructed from 200 ng of total RNA using the Watchmaker Genomics mRNA enrichment and RNA Library Prep kits according to the manufacturer’s instructions. Libraries were dual indexed using IDT xGen Unique Dual Index Primers, and library quality was assessed by Agilent 2100 Bioanalyzer High-Sensitivity DNA assay. Libraries were pooled and run on an Illumina NextSeq 2000, generating a minimum of 60 million paired-end 60-bp reads per sample.

RNA sequencing reads were processed using cutadapt v.4.7 (ref. ⁷⁰) to remove Illumina adapters, to quality trim at Q20, and to filter read pairs that contained N bases or in which either read was less than 31 bp. Processed reads were quantified with Salmon v.1.9.0 (ref. ⁷¹) using transcripts from the GRCh38 Ensembl v.107 annotations. Salmon’s expectation maximization procedures were set to enable modelling of sequencing and GC biases.

In R, transcript level counts from Salmon were imported and aggregated to the gene level using tximport v.1.32.0 (ref. ⁷²). Sample quality assessments were performed in relation to collected RNA integrity values, sample concentrations used in library preparation (ng μl⁻¹), read GC content and percentage read mapping. Normalization, further quality control based on principal component analysis and clustering using normalized values, and differential expression analyses were performed with DESeq2 v.1.44.0 (ref. ⁷⁰). We conducted independent hypothesis weighting to optimize the power of P value filtering⁷³ using IHW v.1.32.0 and log₂ fold change shrinkage using ashr v.2.2-63 (ref. ⁷⁴) to reduce the impact of low expression on estimation of fold change values. Multiple correction testing adjustments were optimized using thresholds for false discovery rate and log₂ fold change at the point of generation of results in DESeq2. Significance was assessed using a q value of 0.01.

Cycloheximide CDH1 protein stability assay

To monitor CDH1 stability upon WEE1 inhibition, we plated RPE1 mRuby–PCNA cells in a 24-well plate at a density of 50,000 cells per well. Cells were arrested with CDK4/6i for 24 h and treated with WEE1i (1 µM) or DMSO and cycloheximide (100 µg ml⁻¹) upon release. Subsequently, cells were lysed every hour for 4 h in 1× Novex Tris-Glycine SDS sample buffer (Novex, LC2676) supplemented with 1× phosphatase inhibitors (Thermo Scientific, 1862495), 1× protease inhibitors (Thermo Scientific, 78429) and 1 mM DTT (BioUltra, 43816-10ML).

APC/C activity calculation from time-lapse data

APC/C activity (k_APC) calculations based on mCherry–geminin levels (APC degron), expressed in the PIP-FUCCI hTert-RPE1 cell line⁷⁵ (a gift from J. Cook, University of North Carolina at Chapel Hill), were calculated as previously described^21,76. Specifically, we assumed degron dynamics were described by differential equation (1):

where k_synth is the constitutive rate of APC degron synthesis and k_APC is the APC activity. The equation can be rearranged for k_APC:

Thus, we were able to use the fluorescence signal of our degron cell lines to estimate each term in equation (2).

First, we aligned our fluorescence intensity time series data from each cell to the G1/S transition point. Subsequently, we averaged the signal across cells at each time point and smoothed the data by calculating the three-point running mean. We then subtracted the background signal from the resulting value at each time point. The resulting series was a measure of degron concentration, corresponding to [APC degron], in arbitrary units. The derivative term was calculated by applying the finite difference method to the consecutive points of the [APC degron] time series. k_synth was estimated from the initial (net) rate of degron accumulation in S phase, when APC/C was assumed to be inactive.

Protein alignment

Clustal Omega⁷⁷ was used to align CDH1 protein sequences, and outputs (consensus frequency of residues, identity matrix and phylogenetic tree) were plotted in Python.

Modelling

We built our cell cycle model using a system of coupled nonlinear differential equations describing the rates of change of biochemical species responsible for regulation of the G1 and S states. The regulation of subsequent cell cycle phases was omitted for simplicity, as our experimental approach was focused on quiescence release rather than cycling cells. A description of the modelling and the parameters used are available in the Supplementary Information.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

1. Parker, L. L. & Piwnica-Worms, H. Inactivation of the p34^cdc2-cyclin B complex by the human WEE1 tyrosine kinase. Science 257, 1955–1957 (1992).

   Article  ADS  CAS  PubMed  Google Scholar 

2. Heald, R., McLoughlin, M. & McKeon, F. Human wee1 maintains mitotic timing by protecting the nucleus from cytoplasmically activated Cdc2 kinase. Cell 74, 463–474 (1993).

   Article  CAS  PubMed  Google Scholar 

3. McGowan, C. H. & Russell, P. Human Wee1 kinase inhibits cell division by phosphorylating p34cdc2 exclusively on Tyr15. EMBO J. 12, 75–85 (1993).

   Article  CAS  PubMed  PubMed Central  Google Scholar 

4. Russell, P. & Nurse, P. Negative regulation of mitosis by wee1⁺, a gene encoding a protein kinase homolog. Cell 49, 559–567 (1987).

5. Gautier, J., Solomon, M. J., Booher, R. N., Bazan, J. F. & Kirschner, M. W. cdc25 is a specific tyrosine phosphatase that directly activates p34^cdc2. Cell 67, 197–211 (1991).

   Article  CAS  PubMed  Google Scholar 

6. Kumagai, A. & Dunphy, W. G. The cdc25 protein controls tyrosine dephosphorylation of the cdc2 protein in a cell-free system. Cell 64, 903–914 (1991).

   Article  ADS  CAS  PubMed  Google Scholar 

7. Strausfeld, U. et al. Dephosphorylation and activation of a p34^cdc2/cyclin B complex in vitro by human CDC25 protein. Nature 351, 242–245 (1991).

   Article  ADS  CAS  PubMed  Google Scholar 

8. Millar, J. B. A. et al. p55CDC25 is a nuclear protein required for the initiation of mitosis in human cells. Proc. Natl Acad. Sci. USA 88, 10500–10504 (1991).

   Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

9. Matheson, C. J., Backos, D. S. & Reigan, P. Targeting WEE1 kinase in cancer. Trends Pharmacol. Sci. 37, 872–881 (2016).

   Article  CAS  PubMed  Google Scholar 

10. Otto, T. & Sicinski, P. Cell cycle proteins as promising targets in cancer therapy. Nat. Rev. Cancer 17, 93–115 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

11. Zhang, C., Peng, K., Liu, Q., Huang, Q. & Liu, T. Adavosertib and beyond: biomarkers, drug combination and toxicity of WEE1 inhibitors. Crit. Rev. Oncol. Hematol. 193, 104233 (2024).

    Article  PubMed  Google Scholar 

12. Hughes, B. T., Sidorova, J., Swanger, J., Monnat, R. J. & Clurman, B. E. Essential role for Cdk2 inhibitory phosphorylation during replication stress revealed by a human Cdk2 knockin mutation. Proc. Natl Acad. Sci. USA 110, 8954–8959 (2013).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

13. Zhao, H., Chen, X., Gurian-West, M. & Roberts, J. M. Phosphorylation in a CDK2AF knock-in mouse causes misregulation of DNA replication and centrosome duplication. Mol. Cell. Biol. 2, 1421–1432 (2012).

14. Wu, C. L. et al. Cables enhances Cdk2 tyrosine 15 phosphorylation by Wee1, inhibits cell growth, and is lost in many human colon and squamous cancers. Cancer Res. 61, 7325–7332 (2001).

    CAS  PubMed  Google Scholar 

15. Baldin, V., Cans, C., Watanabe, N. & Ducommun, B. Evidence for a mammalian Nim1-like kinase pathway acting at the G0-1/S transition. Biochem. Biophys. Res. Commun. 236, 130–134 (1997).

    Article  ADS  CAS  PubMed  Google Scholar 

16. Watanabe, N., Broome, M. & Hunter, T. Regulation of the human WEE1Hu CDK tyrosine 15-kinase during the cell cycle. EMBO J. 14, 1878–1891 (1995).

17. Barr, A. R., Heldt, F. S., Zhang, T., Bakal, C. & Novák, B. A dynamical framework for the all-or-none G1/S transition. Cell Syst. 2, 27–37 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

18. Spencer, S. L. et al. The proliferation-quiescence decision is controlled by a bifurcation in CDK2 activity at mitotic exit. Cell 155, 369–383 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

19. Pennycook, B. R. & Barr, A. R. Restriction point regulation at the crossroads between quiescence and cell proliferation. FEBS Lett. 594, 2046–2060 (2020).

20. Sørensen, C. S. et al. A conserved cyclin-binding domain determines functional interplay between anaphase-promoting complex–Cdh1 and cyclin A-Cdk2 during cell cycle progression. Mol. Cell. Biol. 21, 3692 (2001).

    Article  PubMed  PubMed Central  Google Scholar 

21. Cappell, S. D., Chung, M., Jaimovich, A., Spencer, S. L. & Meyer, T. Irreversible APC^Cdh1 inactivation underlies the point of no return for cell-cycle entry. Cell 166, 167–180 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

22. Lukas, C. et al. Accumulation of cyclin B1 requires E2F and cyclin-A-dependent rearrangement of the anaphase-promoting complex. Nature 401, 815–818 (1999).

    Article  ADS  CAS  PubMed  Google Scholar 

23. Blow, J. J. & Hodgson, B. Replication licensing — origin licensing: defining the proliferative state? Trends Cell Biol. 12, 72–78 (2002).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

24. Yan, Z. et al. Cdc6 is regulated by E2F and is essential for DNA replication in mammalian cells. Proc. Natl Acad. Sci. USA 95, 3603–3608 (1998).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

25. Matson, J. P. et al. Intrinsic checkpoint deficiency during cell cycle re-entry from quiescence. J. Cell Biol. 218, 2169–2184 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

26. Beck, H. et al. Cyclin-dependent kinase suppression by WEE1 kinase protects the genome through control of replication initiation and nucleotide consumption. Mol. Cell. Biol. 32, 4226 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

27. Beck, H. et al. Regulators of cyclin-dependent kinases are crucial for maintaining genome integrity in S phase. J. Cell Biol. 188, 629–638 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

28. Elbæk, C. R. et al. WEE1 kinase protects the stability of stalled DNA replication forks by limiting CDK2 activity. Cell Rep. 38, 110261 (2022).

    Article  PubMed  Google Scholar 

29. Moldovan, G. L., Pfander, B. & Jentsch, S. PCNA, the maestro of the replication fork. Cell 129, 665–679 (2007).

    Article  CAS  PubMed  Google Scholar 

30. Nishitani, H. et al. CDK inhibitor p21 is degraded by a proliferating cell nuclear antigen-coupled Cul4-DDB1^Cdt2 pathway during S phase and after UV irradiation. J. Biol. Chem. 283, 29045–29052 (2008).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

31. Kim, Y., Starostina, N. G. & Kipreos, E. T. The CRL4^Cdt2 ubiquitin ligase targets the degradation of p21^Cip1 to control replication licensing. Genes Dev. 22, 2507–2519 (2008).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

32. Abbas, T. et al. PCNA-dependent regulation of p21 ubiquitylation and degradation via the CRL4^Cdt2 ubiquitin ligase complex. Genes Dev. 22, 2496–2506 (2008).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

33. Rizzardi, L. F. et al. CDK1-dependent inhibition of the E3 ubiquitin ligase CRL4^CDT2 ensures robust transition from S phase to mitosis. J. Biol. Chem. 290, 556–567 (2015).

    Article  CAS  PubMed  Google Scholar 

34. Barr, A. R. et al. DNA damage during S-phase mediates the proliferation-quiescence decision in the subsequent G1 via p21 expression. Nat. Commun. 8, 14728 (2017).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

35. Galanos, P. et al. Chronic p53-independent p21 expression causes genomic instability by deregulating replication licensing. Nat. Cell Biol. 18, 777–789 https://doi.org/10.1038/ncb3378 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

36. Domínguez-Kelly, R. et al. Wee1 controls genomic stability during replication by regulating the Mus81-Eme1 endonuclease. J. Cell Biol. 194, 567–579 (2011).

    Article  PubMed  PubMed Central  Google Scholar 

37. Vassilopoulos, A. et al. WEE1 murine deficiency induces hyper-activation of APC/C and results in genomic instability and carcinogenesis. Oncogene 34, 3023–3035 (2015).

    Article  CAS  PubMed  Google Scholar 

38. Brandeis, M. & Hunt, T. The proteolysis of mitotic cyclins in mammalian cells persists from the end of mitosis until the onset of S phase. EMBO J. 15, 5280–5289 (1996).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

39. Mailand, N. & Diffley, J. F. X. CDKs promote DNA replication origin licensing in human cells by protecting Cdc6 from APC/C-dependent proteolysis. Cell 122, 915–926 (2005).

    Article  CAS  PubMed  Google Scholar 

40. Donzelli, M. et al. Dual mode of degradation of Cdc25 A phosphatase. EMBO J. 21, 4875–4884 (2002).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

41. Feringa, F. M. et al. Hypersensitivity to DNA damage in antephase as a safeguard for genome stability. Nat. Commun. 7, 12618 (2016).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

42. Hornbeck, P. V. et al. PhosphoSitePlus, 2014: mutations, PTMs and recalibrations. Nucleic Acids Res. 43, D512–D520 (2015).

    Article  CAS  PubMed  Google Scholar 

43. Paul, D. et al. Transient APC/C inactivation by mTOR boosts glycolysis during cell cycle entry. Nature 646, 198–207 https://doi.org/10.1038/s41586-025-09328-w (2025).

44. Höfler, A. et al. Cryo-EM structures of apo-APC/C and APC/CCDH1:EMI1 complexes provide insights into APC/C regulation. Nat. Commun. 15, 10074 (2024).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

45. Schwab, M., Neutzner, M., Möcker, D. & Seufert, W. Yeast Hct1 recognizes the mitotic cyclin Clb2 and other substrates of the ubiquitin ligase APC. EMBO J. 20, 5165 (2001).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

46. Vodermaier, H. C., Gieffers, C., Maurer-Stroh, S., Eisenhaber, F. & Peters, J. M. TPR subunits of the anaphase-promoting complex mediate binding to the activator protein CDH1. Curr. Biol. 13, 1459–1468 (2003).

    Article  CAS  PubMed  Google Scholar 

47. Abbas, T. & Dutta, A. p21 in cancer: intricate networks and multiple activities. Nat. Rev. Cancer 9, 400–414 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

48. Lawson, A. R. J. et al. Extensive heterogeneity in somatic mutation and selection in the human bladder. Science 370, 75–82 (2020).

    Article  ADS  CAS  PubMed  Google Scholar 

49. Goto, H. et al. Chk1-mediated Cdc25A degradation as a critical mechanism for normal cell cycle progression. J. Cell Sci. 132, 223123 https://doi.org/10.1242/jcs.223123 (2019).

    Article  CAS  Google Scholar 

50. Hirschi, A. et al. A Cdk7-Cdk4 T-loop phosphorylation cascade promotes G1 progression. Mol. Cell 50, 250–260 https://doi.org/10.1016/j.molcel.2013.04.003 (2013).

51. Trotter, E. W. & Hagan, I. M. Release from cell cycle arrest with Cdk4/6 inhibitors generates highly synchronized cell cycle progression in human cell culture: Cdk4/6 induction synchronisation. Open Biol. 10, 200200 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

52. Yamano, H., Gannon, J. & Hunt, T. The role of proteolysis in cell cycle progression in Schizosaccharomyces pombe. EMBO J. 15, 5268 (1996).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

53. Alfonso-Pérez, T., Hayward, D., Holder, J., Gruneberg, U. & Barr, F. A. MAD1-dependent recruitment of CDK1-CCNB1 to kinetochores promotes spindle checkpoint signaling. J. Cell Biol. 218, 1108–1117 (2019).

    Article  PubMed  PubMed Central  Google Scholar 

54. Ran, F. A. et al. Genome engineering using the CRISPR-Cas9 system. Nat. Protoc. 8, 2281 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

55. Hégarat, N. et al. Cyclin A triggers mitosis either via the Greatwall kinase pathway or cyclin B. EMBO J. 39, e104419 (2020).

    Article  PubMed  PubMed Central  Google Scholar 

56. Zerjatke, T. et al. Quantitative cell cycle analysis based on an endogenous all-in-one reporter for cell tracking and classification. Cell Rep. 19, 1953–1966 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

57. Hoffmann, I., Draetta, G. & Karsenti, E. Activation of the phosphatase activity of human cdc25A by a cdk2-cyclin E dependent phosphorylation at the G1/S transition. EMBO J. 13, 4302–4310 (1994).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

58. Kimata, Y., Baxter, J. E., Fry, A. M. & Yamano, H. A role for the Fizzy/Cdc20 family of proteins in activation of the APC/C distinct from substrate recruitment. Mol. Cell 32, 576–583 (2008).

    Article  CAS  PubMed  Google Scholar 

59. Chang, L., Zhang, Z., Yang, J., McLaughlin, S. H. & Barford, D. Atomic structure of the APC/C and its mechanism of protein ubiquitination. Nature 522, 450–454 (2015).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

60. Yamano, H., Trickey, M., Grimaldi, M. & Kimata, Y. In vitro assays for the anaphase-promoting complex/cyclosome (APC/C) in Xenopus egg extracts. Methods Mol. Biol. 545, 287–300 (2009).

    Article  CAS  PubMed  Google Scholar 

61. Frankenfield, A. M., Ni, J., Ahmed, M. & Hao, L. Protein contaminants matter: building universal protein contaminant libraries for DDA and DIA proteomics. J. Proteome Res. 21, 2104–2113 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

62. Baker, C. P., Bruderer, R., Abbott, J., Arthur, J. S. C. & Brenes, A. J. Optimizing Spectronaut search parameters to improve data quality with minimal proteome coverage reductions in DIA analyses of heterogeneous samples. J. Proteome Res. 23, 1926–1936 (2024).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

63. Perez-Riverol, Y. et al. The PRIDE database at 20 years: 2025 update. Nucleic Acids Res. 53, D543–D553 (2025).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

64. Hornsveld, M. et al. A FOXO-dependent replication checkpoint restricts proliferation of damaged cells. Cell Rep. 34, 108675 (2021).

    Article  CAS  PubMed  Google Scholar 

65. Cooper, S., Barr, A. R., Glen, R. & Bakal, C. NucliTrack: an integrated nuclei tracking application. Bioinformatics 33, 3320–3332 https://doi.org/10.1093/bioinformatics/btx404 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

66. Sofroniew, N. et al. napari: a multi-dimensional image viewer for Python. Zenodo https://doi.org/10.5281/ZENODO.17367124 (2025).

67. Pachitariu, M., Rariden, M. & Stringer, C. Cellpose-SAM: superhuman generalization for cellular segmentation. Preprint at bioRxiv https://doi.org/10.1101/2025.04.28.651001 (2025).

68. Bragantini, J. et al. Ultrack: pushing the limits of cell tracking across biological scales. Nat. Methods 22, 2423–2436 https://doi.org/10.1038/s41592-025-02778-0 (2025).

69. Young, G. & Coktas, A. Jex: high performance computing at the MRC Laboratory of Medical Sciences. Zenodo https://doi.org/10.5281/ZENODO.10245365 (2023).

70. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

71. Patro, R., Duggal, G., Love, M. I., Irizarry, R. A. & Kingsford, C. Salmon provides fast and bias-aware quantification of transcript expression. Nat. Methods 14, 417–419 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

72. Soneson, C., Love, M. I. & Robinson, M. D. Differential analyses for RNA-seq: transcript-level estimates improve gene-level inferences. F1000Research 4, 1521 (2016).

    Article  Google Scholar 

73. Ignatiadis, N., Klaus, B., Zaugg, J. B. & Huber, W. Data-driven hypothesis weighting increases detection power in genome-scale multiple testing. Nat. Methods 13, 577–580 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

74. Stephens, M. False discovery rates: a new deal. Biostatistics 18, 275–294 (2017).

    MathSciNet  PubMed  PubMed Central  Google Scholar 

75. Grant, G. D., Kedziora, K. M., Limas, J. C., Cook, J. G. & Purvis, J. E. Accurate delineation of cell cycle phase transitions in living cells with PIP-FUCCI. Cell Cycle 17, 2496–2516 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

76. McGuinness, B. E. et al. Regulation of APC/C activity in oocytes by a Bub1-dependent spindle assembly checkpoint. Curr. Biol. 19, 369–380 (2009).

    Article  CAS  PubMed  Google Scholar 

77. Madeira, F. et al. The EMBL-EBI Job Dispatcher sequence analysis tools framework in 2024. Nucleic Acids Res. 52, W521–W525 (2024).

    Article  PubMed  PubMed Central  Google Scholar 

78. Baldin, V. & Ducommun, B. Subcellular localisation of human wee1 kinase is regulated during the cell cycle. J. Cell Sci. 108, 2425–2432 (1995).

    Article  CAS  PubMed  Google Scholar 

Download references

Author notes

1. These authors contributed equally: Luc-Alban Vuillemenot, Lucas G. Morales, Calin-Mihai Dragoi

2. Deceased: Béla Novák

Authors and Affiliations

1. Institute of Clinical Sciences, Imperial College London, London, UK

   Luc-Alban Vuillemenot, Betheney R. Pennycook & Alexis R. Barr

2. MRC Laboratory of Medical Sciences, London, UK

   Lucas G. Morales, Edward Harris, Liuyang C. Lin, William A. Weston, Jade-Ellen Brown, Emma Graham, Leelou Moulin, Ivan Andrew, Georgia Roumelioti, Alex Montoya, Laurence Game, George R. Young, Keshava K. Datta, Pavel Shliaha & Alexis R. Barr

3. Department of Biochemistry, University of Oxford, Oxford, UK

   Calin-Mihai Dragoi & Béla Novák

4. University College London Cancer Institute, London, UK

   Hiroyuki Yamano

Contributions

A.R.B. conceived the project, supervised experiments, helped to analyse data, secured funding for the work and wrote the manuscript. L.-A.V., L.G.M., B.R.P., E.H., L.C.L., W.A.W., J.-E.B., E.G., L.M. and H.Y. performed experiments and analysed data. L.-A.V., L.G.M., B.R.P., E.H. and L.M. performed all human cell line experiments. L.C.L., W.A.W., J.-E.B. and E.G. performed cancer cell line imaging experiments. H.Y. performed in vitro APC/C biochemical experiments. C.-M.D. and B.N. developed mathematical models. L.-A.V., L.G.M., C.-M.D., B.R.P and H.Y. prepared figures and helped to write the manuscript. I.A. and L.G. prepared samples for RNA sequencing analysis. G.R., P.S. and K.K.D. prepared samples for proteomics. A.M., P.S. and K.K.D. analysed mass spectrometry data. G.R.Y. analysed RNA sequencing data.

Corresponding author

Correspondence to Alexis R. Barr.

Competing interests

The authors declare no competing interests.

Peer review information

Nature thanks Mardo Koivomagi, Devin Bradburn, Steven Cappell and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions