Cdk1 and PP2A constitute a molecular switch controlling orderly degradation of atypical E2Fs

  1. Sapir Nachum-Raines
  2. Noy Gamliel
  3. Danit Wasserman
  4. Nasrin Qassem
  5. Inbal Sher
  6. Julia Guez-Haddad
  7. Michael J Emanuele
  8. Jordan H Chill
  9. Amit Tzur

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

Dynamic oscillations in the phosphorylation and ubiquitination of key proliferative regulators are defining features of the eukaryotic cell cycle. Resetting the cell cycle at the mitosis-to-G1 transition requires activation of the E3 ubiquitin ligase Anaphase-Promoting Complex/Cyclosome (APC/C), which ensures cell cycle irreversibility by targeting dozens of substrates for degradation, safeguarding genome integrity. However, the overall coupling of substrate phosphorylation with target recognition and degradation by the APC/C remains relatively unexplored. As a paradigm for further defining these rules, we focused on E2F7 and E2F8 – atypical E2F-family proteins that coordinate cell cycle gene expression by restraining the pro-proliferative transcriptional activity of E2F1. Leveraging complementary cell and cell-free systems, we demonstrate that flexible domains in the amino-termini of E2F7 and E2F8 contain APC/C recognition motifs adjacent to critical Thr residues, whose phosphorylation by Cdk1 is rate limiting for degradation. The removal of this phosphorylation by PP2A phosphatase serves as a molecular switch, coupling the degradation of E2F7 and E2F8 to the G1 phase, coinciding with the rise of E2F1. Collectively, these findings highlight a critical role for Cdk1-PP2A signaling in controlling the orderly degradation of APC/C substrates, ensuring precisely timed assembly of the transcriptional infrastructure that coordinates cell cycle commitment and progression.

Article activity feed

  1. Review Commons

    Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    Manuscript number: RC-2025-02932

    Corresponding author(s): Amit Tzur

    [Please use this template only if the submitted manuscript should be considered by the affiliate journal as a full revision in response to the points raised by the reviewers.

    If you wish to submit a preliminary revision with a revision plan, please use our "Revision Plan" template. It is important to use the appropriate template to clearly inform the editors of your intentions.]

    1. General Statements

    We thank all Referees for their insightful comments and thoughtful review of our manuscript.

    2. Point-by-point description of the revisions

    This section is mandatory. *Please insert a point-by-point reply describing the revisions that were already carried out and included in the transferred manuscript. *

    __! Original comments by Reviewers #1-3 are in gray. __

    Reviewer #1 (Evidence, reproducibility and clarity (Required)):

    The study highlights a dephosphorylation switch mediated by PP2A as a critical mechanism for coupling E2F7/8 degradation to mitotic exit and G1 phase. The study is clear and experiments are well conducted with appropriate controls

    I have some concerns highlighted below:

    Point 1. In this sentence: This intricate network of feedback mechanisms ensures the orderly progression of the cell cycle. What feedback mechanism are the authors referring to?

    Thank you for pointing this out. We aimed for a general comment. The original line was replaced with: “The intricate network of (de)phosphorylation and (de)ubiquitination events in cycling cells establishes feedback mechanisms that ensure orderly cell cycle progression.”

    Point 2. Characterization of disorder in the N-terminal segments of E2F7 and E2F8

    What does it mean disorder in this title?

    “Disorder” is a structural biology term for describing an unstructured (floppy) region in a protein. We suggest the following title in hope to improve clarity: “The N-terminal segments of E2F7 and E2F8 are intrinsically unstructured”

    Point 3. In the paragraph on the untimely degradation of E2F8 the authors keep referring to APC/C Cdc20, however the degradation is triggered by the Ken box which is specifically recognised by APC/C Cdh1. Can it be due to another ligase not APC/C?

    In our anaphase-like system, Cdh1 cannot associate with the APC/C due to persistently high Cdk1 activity, maintained by the presence of non-degradable Cyclin B1. While the KEN-box is classically recognized as a Cdh1-specific motif, previous studies have also clearly demonstrated that APC/C-Cdc20 can mediate the degradation of KEN-box substrates. For example, BubR1 interacts with Cdc20 via two KEN-box motifs (PMIDs: 25383541, 27939943 and 17406666). Nek2A is targeted for degradation by the APC/C in mitotic egg extracts lacking Cdh1, in a manner that depends on both D-box and KEN-box motifs (PMID: 11742988). CENP-F degradation in Cdh1-null cells has been shown to be dependent on both Cdc20 and a KEN-motif (PMID: 20053638). Thus, the most simple explanation for our results is that degradation is KEN box dependent and controlled by Cdc20.

    Regarding alternative E3 ligases, KEN-box mutant variants of non-phosphorylatable E2F8 remained stable in APC/CCdc20-active extracts, suggesting that this degradation is indeed APC/C-specific.

    Please also see our response to Reviewer #3, Point 3.

    Point 4. The assays to detect dephosphorylation are rather indirect so it is difficult to establish whether phosphorylation of CDK1 and dephosphorylation by PP2A on the fragments is direct.

    First, the phosphorylation sites analyzed in this study conform to the full and most canonical Cdk1 consensus motif: S/TPxK/R. While recognizing that other kinases are proline directed as well, the cell cycle dependent manner of this control, and presence of a similar CDK-dependent mechanism for Cdc6, points us towards considering the role of CDKs.

    Second, consistent with the direct role of CDK1 in this regulation, NMR experiments demonstrate conformational shifts of recombinant E2F8 following incubation with Cdk1–Cyclin B1 (not included in manuscript, but shown here for reviewer consideration); see Figure below. We have not yet established equivalent biochemical systems for PP2A.

    Figure legend: NMR-based monitoring of E2F7 (a-c) and E2F8 (d-f) phosphorylation by Cdk1.

    a(d). 15N,1H-HSQC spectrum of E2F7(E2F8) prior to addition of Cdk1. Threonine residues of interest, T45 (T20) conforming to the consensus sequence (followed by a proline), and T84 (T60) lacking the signature sequence are annotated. b(e). Strips from the 3D-HNCACB spectrum used for assigning E2F7(E2F8) residues. Black (green) peaks indicate a correlation with the 13Cα (13Cβ) of the same and previous residues. The chemical shifts assigned to T45 (T20) and T84 (T60) match the expected values for K44(K19) and P83(P59), thereby confirming the assignment. c(f). Top, overlay of subspectra before adding Cdk1 (black) and after 16 h of activity (red) at 298 K. Bottom, change in intensities of the T45/T84 in E2F7 and T20/T60 in E2F8 showing how NMR monitors phosphorylation and distinguishes between various threonine residues.

    Third, PP2A is likely the principal phosphatase counteracting Cdk1-mediated phosphorylation during mitotic exit, targeting numerous APC/C substrates (PMID: 31494926). In light of our findings and the extensive literature, it is therefore reasonable to propose that E2F7 and E2F8 may also be direct PP2A targets.

    Fourth, we cannot fully exclude the possibility that dephosphorylation of E2F7 and E2F8 by PP2A occurs indirectly. Nevertheless, indirect studies of PP2A substrate identification in the literature often rely on similar genetic perturbations, chemical inhibition, cell-free systems (coupled with immunodepletion, inhibitory peptides/proteins, and small-molecule inhibitors), and phosphoproteomics. Moreover, more direct assays are not without caveats, as they lack the cellular stoichiometric context, an important limitation for relatively promiscuous enzymes such as phosphatases.

    Importantly, repeated attempts (conventional [Co-IP] and less conventional [affinity microfluidics]) to detect interactions between PP2A and E2F7 and E2F8 were unsuccessful. This result was unfortunate but not surprising, given that transient substrate–phosphatase interactions are often challenging to capture experimentally.

    Given our evidence showing the regulation of E2F7 and E2F8 degradation in a manner that depends on Cdk1 and PP2A, the title of the manuscript remains appropriate: "Cdk1 and PP2A constitute a molecular switch controlling orderly degradation of atypical E2Fs.”

    Please also see our response to Reviewer #3 Point 1.

    Point 5. Although there seems to be a control by phosphorylation and dephosphorylation (which could be indirect), it is difficult to establish the functional consequences of this observation. The authors propose a feedback mechanism which regulates the temporal activation inactivation of E2F7/8 however, there are no evidence in support of this.

    The components being studied here have been extensively characterized, as have the direct and indirect interactions that connect them and ensure orderly cell cycle progression. For example: i) The E2F1–E2F7/8 transcriptional circuitry functions as a negative feedback loop; ii) Cdk1 and PP2A counteract one another’s activity; iii) E2F1 promotes the disassembly of APC/CCdh1; iv) E2F7 and E2F8 are APC/C substrates with cell cycle-relevant degradation patterns; and v) Loss of Cdh1 leads to premature S-phase entry.

    Our study brings these components together into a coherent regulatory module operating in cycling cells, revealed through cell-free biochemistry and newly developed methodologies with broad applicability to signaling research. We believe that advancing mechanistic understanding at this level of central regulators is impactful. And notably, this is a model, which we expect others in the field to test. We stand behind the result of each individual experiment and based on those findings are proposing a feedback circuit.

    To address your suggestion, we incorporated phenotypic analyses (see Figure on the next page). Although modest and variable due to transient overexpression, these data align with the mechanistic model proposed in our study.

    In Panel a, overexpression of E2F7 or E2F8 reduces E2F1 and its target Plk1, consistent with the established negative feedback within the E2F1–E2F7/8 transcriptional circuitry. A broader impact on cell cycle progression was also evident: G1-phase cells increased and S-phase cells decreased (Panel b), hinting at a delayed G1–S transition when E2F1, an essential driver of S-phase and mitotic entry, is downregulated by excess E2F7 or E2F8.

    We next examined the effects of hyper- vs. hypo-phosphorylation–mimicking mutants of E2F7 and E2F8 on E2F1 and Plk1 (Panels c and d). Both raw data (top) and quantification (bottom) are shown. Despite ectopic overexpression, our experimental conditions highlighted the diffenrential outcome of the two phospho-mutant variants. Speificially, E2F1 and Plk1 levels were consistently higher upon expression of non-phosphorylatable variants of E2F7 (T45A/T68A) and E2F8 (T45D/T68D) relative to their phophomimetic counterparts (T45D/T68D; T20D/T44D). These findings suggest that E2F1 downregulation is more pronounced when E2F7/E2F8 are hyper-phosphorylated at Cdk1-regulated sites that control their half-lives. Furthermore, the proportion of S-phase cells was consistently lower for the phospho-mimicking mutants compared with the non-phosphorylatable variants, with complementary, though less pronounced, shifts in G1-phase cells (Panel e).

    Figure legend: Evidence for cell cycle control linked to Cdk1–PP2A regulation of the E2F1–E2F7/E2F8 axis.

    a) Immunoblot analysis showing reduced E2F1 and its target protein Plk1 upon E2F7/E2F8 overexpression. Antibodies used for immunoblotting (IB) are indicated. b) Cell cycle phase distribution after E2F7/E2F8 overexpression, based on DNA content. Left: representative histograms. Right: quantification of G1- and S-phase cells. Means (x) with individual biological replicates (color-coded; N = 4) are shown. c,d) Top: E2F1 and Plk1 protein levels in cells expressing phosphomimetic (TT-DD) or non-phosphorylatable (TT-AA) E2F7 (c) or E2F8 (d) variants. Antibodies used are indicated (*distorted signal excluded). Bottom: quantification relative to loading controls. Means (x) with individual values (N = 3/4) are shown. e) Cell cycle phase distribution following expression of E2F7/E2F8 phospho-mutant variants. Means (x) with individual values (N = 4) are shown. All experiments were performed in HEK293T cells. Cells were fixed 40–44 h post-transfection. DNA content was assessed using propidium iodide (PI). Mutation sites: T45/T68 (E2F7); T20/T44 (E2F8. Statistical significance was determined by two-tailed Student’s t-test; P-values are indicated.

    Taken together, these results support a model in which Cdk1-site (de)phosphorylation modulates the stability of E2F7 and E2F8, thereby shaping E2F1 output and influencing cell cycle preogresion.

    Point 6. Reviewer #1 (Significance (Required)):

    The study is a good and well conducted work to understand the mechanisms regulating degradation of E2F7/8 by APC/C. This is crucial to establish coordinated cell cycle progression. While the hypothesis that disruption of this mechanism is likely responsible for altered cell cycle progression, there are no evidence this is just a back up pathway, whose functional significance could be limited to lack of APC/C Cdh1 activity. These experiments are rather difficult but the authors could comment on the limitation of the study and emphasise the hypothetical alterations which could result from the alterations of the described feedback loop

    We thank Reviewer #1 for this comment. Accordingly, we have expanded the discussion to further elaborate on the potential molecular outcomes and limitations of our study.

    Reviewer #2 (Evidence, reproducibility, and clarity (Required)):

    Summary: The authors provide strong biochemical evidence that the regulation of E2F7 and E2F8 by APC is affected by CDK1 phosphorylation and potentially by PP2A dependent dephosphorylation. The authors use both full length and N-terminal fragments of E2F8 in cell-free systems to monitor protein stability during mitotic exit. The detailed investigation of the critical residues in the N-terminal domain of E2F8 (T20/T44) is well supported by the combination of biochemical and cell biology approaches.

    We thank Reviewer #2 for their encouraging feedback.

    Point 1. Major: It is unclear how critical the APC-dependent destruction of E2F7 and E2F8 is for cell cycle progression or other cellular processes. Prior studies have reported that Cyclin F regulation of E2F7 is critical for DNA repair and G2-phase progression. This study would be improved if the authors could provide a cellular phenotype caused by the lack of APC dependent regulation of E2F8 and/or E2F7.

    We thank Reviewers #2 and #1 for this comment, which prompted substantial revisions. Below, we reiterate our response to Reviewer #1.

    The molecular components examined in this study are well established in the literature. Key principles include: (i) the reciprocal regulation between E2F1 and its repressors, E2F7 and E2F8, which forms a transcriptional feedback loop; (ii) the opposing activities of Cdk1 and PP2A; (iii) the capacity of E2F1 to attenuate APC/CCdh1 activity; (iv) the fact that E2F7 and E2F8 are APC/C substrates with defined cell cycle–dependent degradation patterns; and (v) the requirement for Cdh1 to prevent premature S-phase entry.

    Our study integrates these elements into a unified framework operating in proliferating cells. This framework is supported by biochemical reconstitution experiments and newly developed methodological tools, which we anticipate will be broadly applicable for dissecting signaling pathways. We view this type of mechanistic synthesis as valuable for the field. Importantly, we do not present this as a definitive model, but rather as a testable regulatory circuit constructed from robust individual findings.

    In response to your request, we incorporated additional phenotypic analyses (see Figure, next page). Although modest and variable due to transient overexpression, the results are consistent with the regulatory architecture we propose.

    In panel a, elevating E2F7 or E2F8 levels reduces E2F1 and its downstream target Plk1, consistent with the established inhibitory feedback exerted by E2F7 and E2F8 on E2F1. Additionally, we observed an increase in G1-phase cells and a decrease in S-phase cells (Panel b), hinting at a delayed G1–S transition when E2F1, a key transcriptional engine of S- and M-phase entry, is downregulated by excess E2F7 or E2F8.

    Figure legend: Evidence for cell cycle control linked to Cdk1–PP2A regulation of the E2F1–E2F7/E2F8 axis.

    a) Immunoblot analysis showing reduced E2F1 and its target protein Plk1 upon E2F7/E2F8 overexpression. Antibodies used for immunoblotting (IB) are indicated. b) Cell cycle phase distribution after E2F7/E2F8 overexpression, based on DNA content. Left: representative histograms. Right: quantification of G1- and S-phase cells. Means (x) with individual biological replicates (color-coded; N = 4) are shown. c,d) Top: E2F1 and Plk1 protein levels in cells expressing phosphomimetic (TT-DD) or non-phosphorylatable (TT-AA) E2F7 (c) or E2F8 (d) variants. Antibodies used are indicated (*distorted signal excluded). Bottom: quantification relative to loading controls. Means (x) with individual values (N = 3/4) are shown. e) Cell cycle phase distribution following expression of E2F7/E2F8 phospho-mutant variants. Means (x) with individual values (N = 4) are shown. All experiments were performed in HEK293T cells. Cells were fixed 40–44 h post-transfection. DNA content was assessed using propidium iodide (PI). Mutation sites: T45/T68 (E2F7); T20/T44 (E2F8. Statistical significance was determined by two-tailed Student’s t-test; P-values are indicated.

    We next examined how phospho-regulation of E2F7 and E2F8 influences cell cycle control by comparing the effects of phospho-mimetic and non-phosphorylatable variants on E2F1 levels and cell cycle distribution (panels c and d). Both the raw data and the corresponding quantitative analyses are presented. Despite exogenous overexpression, we identified conditions that distinguish the behaviors of the two mutant classes. Cells expressing the phospho-mimetic variants consistently exhibited lower E2F1 and Plk1 levels than those expressing the non-phosphorylatable forms. This pattern supports a model in which phosphorylation of key Cdk1 sites in E2F7 and E2F8 elevates their stability, thereby enhancing their ability to suppress E2F1. Panel e extends these observations to cell cycle behavior: compared with the non-phosphorylatable variants, The phospho-mimetic forms of E2F7 and E2F8 consistently lower the proportion of S-phase cells, accompanied by corresponding shifts in the G1 population.

    The central aim of this manuscript is to define how the Cdk1–PP2A axis is integrated into the APC/C–E2F1 regulatory network controlling cell cycle progression. Collectively, our findings support a model in which Cdk1/PP2A-dependent (de)phosphorylation modulates the stability of E2F7 and E2F8, thereby fine-tuning E2F1 activity and cell cycle progression.

    Point 2. Minor: All optional: It would have been interesting to see the T20A/T44A/KM in the live cell experiment (Figure 3F).

    This is an excellent point. Following Reviewer #2’s request, we generated a stable cell line expressing a KEN-box mutant variant of E2F8-T20A/T44A (N80 fragment). The figure below demonstrates the impact of the KEN-box mutation on the dynamics of N80-E2F8-T20A/T44A in HeLa cells. Together, our data from both cellular and cell-free systems show that the temporal dynamics of both wild-type and non-phosphorylatable variants of E2F8 depends on the KEN degron. Please note that due to differences in the flow cytometer settings used for acquiring the original measurements and those newly generated at the Reviewer’s request, the numeric data for N80-E2F8-T20A/T44A-KEN mutant will not be integrated into the original plots shown in the original Figure 3c–e in the manuscript.

    Figure legend: Dynamics of mutant variants of N80-E2F8-EGFP in HeLa cells.

    Top: Bivariate plots showing DNA content (DAPI) vs. EGFP fluorescence, with G1/G1-S phases and G2/M phases highlighted (black and gray frames, respectively). Bottom: Histograms showing EGFP signal distributions within these cell cycle phases. Blue arrows highlight subpopulations of G2/M cells with relatively low EGFP levels. The data was generated by flow cytometry.

    Point 3. Figure 4C-D - include the corresponding blots for the WT E2F7.

    This is a good point, which we previously overlooked. The requested data will be integrated in the revised manuscript.

    Point 4. It is unclear how selective or potent the PP2A inhibitors are that are used in Figure 5. Is it possible to include known targets of PP2A (positive controls for PP2A inhibition) in the analysis performed in Figure 5?

    Thank you for this helpful suggestion. Following Reviewer #2’s comment, we performed gel-shift assays of Cdc20 and C-terminal fragment of KIF4 (Residues: 732-1232), both known targets of PP2A (PMIDs: 26811472; 27453045). See data below.

    __Figure legend: PP2A inhibitor LB-100 block protein dephosphorylation in G1-like extracts. __

    Time-dependent gel shifts of mitotically phosphorylated Cdc20 and the C-terminal fragment of KIF4 (residues 732–1232) following incubation in G1 extracts supplemented with LB-100 or okadaic acid (OA; positive control). Substrates (IVT, 35S-labeled) were resolved by PhosTag SDS–PAGE and autoradiography.

    Point 5. Is the APC still active in LB-100 or OA treated conditions? Is it possible to demonstrate the APC is active using known substrates in this assay (e.g., Securin (Cdc20) and Geminin (Cdh1) or similar).

    This is an excellent point and we should have clarified this previously. Importantly, treatment with 250 µM LB-100 does not abolish APC/C-mediated degradation (otherwise, the assay would not be viable), but it does attenuate degradation kinetics. This is reflected by the prolonged half-lives of Securin and Geminin relative to mock-treated extracts (see below). Consistently, we noted in the manuscript: “Although APC/C-mediated degradation is also affected, it remains efficient, allowing us to measure relative half-lives of APC/C targets that cannot undergo PP2A-mediated dephosphorylation.” Following this comment, and one by Reviewer #3, these data will be included in the revised manuscript.

    __Figure legend: APC/C-specific activity in cell extracts treated with LB-100. __

    Time-dependent degradation of EGFP–Geminin (N-terminal fragment of 110 amino acids) and Securin in extracts supplemented with LB-100 and/or UbcH10 (recombinant). A control reaction contained dominant-negative (DN) UbcH10. Proteins (IVT, 35S-labeled) were resolved by SDS-PAGE and autoradiography.

    Reviewer #2 (Significance (Required)): Advance: A detailed analysis is provided for the critical N-terminal residues in E2F7 and E2F8 that when phosphorylated are capable of restricting APC destruction. The work builds on prior work that had identified the APC regulation of E2F7 and E2F8.

    Point 6. Audience: The manuscript would certainly appeal to a broad basic research audience that is interested in the regulation of APC substrates and/or E2F axis control via E2F7 & E2F8. The study could have a broader interest if the destruction of E2F7 or E2F8 could be shown to be biologically relevant (e.g., critical for cell fate decision G1 vs G0, G1 length, timely S-phase onset, or expression of E2F1 target genes in the subsequent cell cycle).

    To clarify, we subdivided Reviewers’ comments into separate points. Reviewer #2’s Points 1 and 6 address essentially the same issue; our detailed response is therefore provided under Point 1. We again thank Reviewer #2 for raising this concern, which led to substantial revisions to both the manuscript text and the supporting data.

    We thank Reviewer #2 for their constructive comments and criticism.

    Reviewer #3 (Evidence, reproducibility and clarity (Required)):

    This manuscript presents a well-structured study on the regulatory interplay between Cdk and Phosphatase in controlling the degradation of atypical E2Fs, E2F7 and E2F8. The work is relevant in the field of cell cycle regulation and provides new mechanistic insights into how phosphorylation and dephosphorylation govern APC/C-mediated degradation. The use of complementary cell-based and in vitro approaches strengthens the study, and the findings have significant implications for understanding the timing of transcriptional regulation in cell cycle progression.

    Point 1. However, several points in this paper require further clarification for it to have a meaningful impact on the research community. The characterization of the phosphatase is unclear to me. The use of OA is necessary to guide the research, but it is not precise enough to rule out PP1 and then identify which PP2A is involved - PP2A-B55 or PP2A-B56. To clarify this, the regulatory subunits should either be eliminated or inhibited using the inhibitors developed by Jakob Nilsson's team.

    We are grateful for this comment, which prompted an extensive series of experiments that have undoubtedly strengthened our manuscript.

    First, we wish to clarify that LB-100, unlike okadaic acid (OA), is not considered a PP1 inhibitor.

    Second, we have conducted a large set of experiments to address this important question of the strict identity of the phosphatase involved in the dephosphorylation of atypical E2Fs.

    I. We initially attempted to immunodeplete the catalytic subunit of PP2A (α) from G1 extracts as a means to validate PP2A-dependent dephosphorylation. In retrospect, this was a naïve approach given the protein’s high abundance; although immunoprecipitation was successful, immunodepletion was inefficient, preventing us from using this strategy (see Panel a in the figure below). As an alternative, we incubated immunopurified PP2A-Cα with mitotic phosphorylated E2F7 and E2F8 fragments (illustrated in Panel b). A time-dependent gel-shift assay demonstrated enhanced dephosphorylation in the presence of immunopurified PP2A-Cα (Panel c) compared to immunopurified Plk1 (control reaction), suggesting that mitotically phosphorylated E2F7 and E2F8 are targeted by PP2A.

    Figure legend: Immunopurified PP2A-Cα facilitates dephosphorylation of E2F7 and E2F8 in cell extracts. a) Inefficient immunodepletion (ID) of the catalytic subunit α of PP2A (PP2A-Cα) from cell extracts despite three rounds of immunopurification, as detected by immunoblotting (IB) with anti-PP2A-Cα and anti-BIP (loading control; LC) antibodies (BD bioscience, Cat#: 610555; Cell Signaling Technology, Cat#: 3177). Briefly, G1 cell extracts were diluted to ~10 mg/mL in a final volume of 65 μL. Anti-PP2A-Cα antibodies (3 μg) were coupled to protein G magnetic DynabeadsTM (15 μL; Novex, Cat#: 10004D) for 20 min at 20 °C. For each depletion round, antibody-coupled beads were incubated with cell extracts for 15 min at 20 °C. Cell extracts and beads were sampled after each step to assess immunodepletion and immunopurification (IP) efficiency. Equivalent immunopurification steps are shown for Plk1 (bottom). b) Schematic of the dephosphorylation assay using mitotically phosphorylated in vitro translated (IVT) targets and immuno-purified PP2A-Cα/Plk1. c) Dephosphorylation of mitotically phosphorylated E2F7 and E2F8 fragments, detected by electrophoretic mobility shifts in Phos-Tag SDS-PAGE. Immunopurified Plk1 was used for control reactions (antibodies: Santa Cruz Biotechnology: Cat#: SC-17783). *Image was altered to improve visualization of mobility shifts.

    II. Next, we used pan-B55-specific antibodies for immunodepletion of all B55-type subunits. This approach was unsuccessful despite five rounds of immunopurification (see Panel a in the figure below). Both suboptimal binding and the high abundance of endogenous B55 subunits likely contributed to this outcome. Thus, dephosphorylation in B55-depleted extracts could not be tested.

    Figure legend: PP2A-B55 facilitates dephosphorylation of E2F7 and E2F8 fragments.

    __a) Immunodepletion (ID) of B55 subunits in G1 extracts is inefficient despite five rounds of immunopurification; assessed by immunoblotting (IB) using anti-pan-B55 and anti-Cdk1 (loading control; LC) antibodies (see previous figure for more details). Cell extracts and beads were sampled after each round to monitor immunodepletion and immunopurification efficiency. b) __Schematic of a dephospho-rylation assay using immuno-purified B55 subunits. __c) Dephosphorylation of mitotically phosphorylated E2F7 and E2F8 fragments by immuno-purified B55. Control reactions performed with immuno-purified Plk1. d) __Schematic of a dephosphorylation assay performed in G1 cell extracts supplemented with B55-interacting (B55i) or control peptides (see peptide sequence on next page). RO-3306 was added to limit Cdk1 activity. __e) __Dephosphorylation of E2F7 and E2F8 fragments (mitotically phosphorylated) in G1 extracts supplemented with B55-interacting/control peptides. __f) __Schematic of the dephosphorylation assay using in vitro–translated B55/B56 subunits (unlabeled). __g) __Dephosphorylation of mitotically phosphorylated E2F7 (top) and E2F8 (bottom) fragments in reticulocyte lysate containing B55/B56 subunits. Dephosphorylation was assessed by electrophoretic mobility shifts in Phos-Tag SDS-PAGE. Panels marked with an asterisk were adjusted to improve visualization of gel-shifts. Arrowheads denote distinct, time-dependent mobility-shifted forms of E2F7 and E2F8 fragments. Antibodies used: anti-pan-B55 (ProteinTech, Cat#: 13123-1-AP); anti-Plk1 (Santa Cruz Biotechnology, Cat#: SC-17783); anti-Cdk1 (Santa Cruz Biotechnology, Cat#: SC-53217). Dynabeads™ (Novex, Cat#: 10004D) were used for immunopurification.

    As with PP2A-Cα, we incubated immunoprecipitated B55 subunits with mitotically phosphorylated E2F7 and E2F8 fragments (illustrated in Panel b). The results were less definitive compared to PP2A-Cα; nevertheless, they demonstrated accelerated dephosphorylation in the presence of immunopurified B55 subunits (Panel c) relative to Plk1 (control). These results hint at B55-mediated dephosphorylation of E2F7 and E2F8.

    III. Given that PP2A-B55 could be immunodepleted satisfactorily, despite successful immunoprecipitation, we ordered the B55-specific peptide and corresponding control peptide reported recently by Jakob Nilsson’s team as PP2A-B55 inhibitors (see below).

    Figure legend: Adapted from Kruse, T., et al., 2024; ____Science Advances. Figure 3, Panel B. ____PMID: 39356758.

    Despite our long-anticipated wait for these peptides to arrive, this line of experimentation proved disappointing. We wish to elaborate:

    The study by Kruse et al. (PMID: 39356758) is an elegant integration of classical enzymology, performed at the highest level, with structural insight into the conserved PP2A-B55 binding pocket that governs substrate specificity. Their work identified a consensus peptide that binds PP2A-B55 specifically with nanomolar affinity.

    Kruse et al. provide compelling evidence for a direct and specific interaction between their reported B55 inhibitor (B55i) and PP2A-B55. Their data show that the engineered inhibitor disrupts the binding of helical elements that underlie substrate recognition by PP2A-B55.

    However, we could not find direct evidence of PP2A-B55 enzymatic inhibition by the B55i peptide; for example, a B55-specific in vitro dephosphorylation assay demonstrating sensitivity to B55i in a dose-dependent manner. To the best of our understanding, the sole functional consequence described by Kruse et al. was the delay in mitotic exit observed upon expression of YFP-tagged B55i peptides in cells. However, this approach is indirect, given the long interval between cell manipulation and analysis and the complexity of mitotic exit. Furthermore, we assumed that the requested reagents had been validated in cell-free extracts; however, Kruse et al. do not report any experiments performed in these systems. We, in fact, became uncertain whether we had correctly understood Reviewer #3’s request to use these reagents and therefore sought clarification from the Editor.

    In vitro, Kruse et al. reported nanomolar binding affinities for B55i (Figure S14). In our cell extracts, however, we required concentrations of approximately 250 μM to detect an effect on dephosphorylation, evident as altered electrophoretic mobility of both E2F7 and E2F8 (Panel e). At this concentration, the peptide also caused nonspecific effects, rendering the extracts highly viscous (‘gooey’), at times preventing part of the reaction mixture from passing through a 10 μL pipette tip.

    The gel-shift assays shown in Panel e (Page 16) do demonstrate delayed dephosphorylation in extracts treated with the B55i peptide relative to the control peptide. Nevertheless, we prefer to exclude these data because the peptide concentrations required for the assay compromised extract integrity. Moreover, we believe that the PP2A-B55–specific peptide described by Nilsson et al. requires additional validation before it can be considered a reliable functional inhibitor in cell-free systems or in vivo. Accordingly, we are unable to directly address the experiments as suggested.

    IV. In the final set of experiments (Page 16, Panels f and g), we supplemented dephosphorylation reactions with in vitro–translated B55/B56 subunits (illustrated in Panel f). Although the expected concentration of in vitro–translated proteins in reticulocyte lysate is relatively low (100–400 nM), we reasoned that supplementing the reactions with excess of regulatory B subunits (non-radioactive) could still promote dephosphorylation in a differential manner that reflects the B55/B56 preference of E2F7 and E2F8.

    We cloned and in vitro expressed all nine B55/B56 regulatory subunits. While the exact amount of each subunit introduced into the reaction cannot be precisely determined, their expression levels were reasonably uniform (see figure below).

    __Figure legend: Expression of B55/B56 subunits in reticulocyte lysate. __B55/B56 subunits were cloned into the pCS2 vector and expressed in reticulocyte lysate supplemented with ³⁵S-Methionin. Proteins were resolved by SDS–PAGE and autoradiography.

    Returning to Panel g (Page 16), B55 subunits facilitated the accumulation of lower–electrophoretic mobility forms of both E2F7 and E2F8 fragments to the greatest extent. This is evident from the distinct lower–mobility species that emerge over time (marked by arrowheads) and the smear intensity corresponding to the buildup of dephosphorylated forms. Among the tested subunits, B55β exerted the strongest effect on both substrates, suggesting that mitotically phosphorylated E2F7 and E2F8 display a heightened preference for the PP2A-B55β holoenzyme. Control reactions with reticulocyte lysate are also shown.

    Taken together, our original and newly added data indicate that PP2A, specifically PP2A-B55, counteracts Cdk1-dependent phosphorylation during mitotic exit. Importantly, cell cycle regulators such as Cdc20 can be targeted by both PP2A-B55 and PP2A-B56 holoenzymes. Thus, while we are confident in concluding that mitotically phosphorylated E2F7 and E2F8 are targeted by PP2A-B55, we cannot rule out the possibility of functional interactions between E2F7/E2F8 and PP2A-B56.

    V. Last, but certainly not least, we used AlphaFold 3 to model interactions between the N-terminal fragments of E2F7 and E2F8 and the PP2A regulatory subunits. To clarify: for us, AlphaFold 3 remains very much a computational “black box,” and although this may sound like an overstatement, we did not anticipate obtaining meaningful or interpretable output.

    According to the AlphaFold 3 developer guidelines, the Interface Predicted Template Modeling (IPTM) score is the primary confidence metric for protein–protein interaction predictions. IPTM values above 0.8 indicate high-confidence predictions, whereas values below 0.6 likely reflect failed interaction predictions. In our models, none of the predicted interactions exceeded 0.6 (see figure below). Nevertheless, for both E2F7 and E2F8 fragments, IPTM scores were consistently higher for B55 subunits than for B56 subunits, with B55β yielding the highest scores (each interaction was modeled five times).

    __Figure legend: AlphaFold 3 predicts preferential interactions between E2F7 and E2F8 and PP2A-B55β. __Protein–protein interaction predictions between N-terminal fragments of E2F7 and E2F8 and B55/B56 regulatory subunits of PP2A were generated using AlphaFold 3 (AF3). The plot shows IPTM scores from five models per protein pair.

    Even if one assumes a scenario in which AlphaFold 3 scores are inaccurate or effectively random, such non-specific behavior would not be expected to produce: (i) a reproducible preference of two distinct substrates for B55β and B55γ, in that order (the modeled fragments of E2F7 and E2F8 share The ability of AlphaFold 3, and specifically the IPTM metric, to predict bona fide PP2A B55/B56–substrate interactions remains unvalidated. Accordingly, we do not rely on these predictions as experimental evidence. Nonetheless, in retrospect, the IPTM scores for the E2F7 and E2F8 fragments proved, unexpectedly, to be highly informative. While we are not the first to explore AlphaFold in the context of PP2A phosphatases (e.g., Kruse et al.), at this early stage of AlphaFold 3 these observations are compelling and may ultimately have implications for PP2A-mediated signaling that extend well beyond the cell-cycle field.

    Point 2. It would also be valuable for this study to investigate the mechanisms underlying this regulation. In particular, is it exclusive to E2F7-8 or could other substrates contribute to the generalisation of this regulatory process?

    Assuming Reviewer #3 is referring to the cell cycle mechanism regulating E2F7 and E2F8 half-life via conditional degrons, we wish to clarify that the temporal dynamics of APC/C targets regulated by dephosphorylation has been demonstrated previously. Examples include KIFC1, CDC6, and Aurora A (PMIDs: 24510915; 16153703; 12208850, respectively).

    Point 3. The observation that Cdc20 may target E2F8 is interesting but needs to be further clarified to ensure that weak Cdh1 activity does not contribute to this degradation. Elimination of Cdc20 would be necessary to support the authors' conclusion.

    We gratefully acknowledge this input. The newly implemented experiment and corresponding findings are presented on the next page. The immunodepletion (ID) procedure (Panel a) achieved >60% reduction of Cdc20 and Plk1 in mitotic extracts (Panel b), as confirmed by immunoblotting (IB). Plk1-depleted extracts were used to validate extract-specific activity after successive rounds of immunodepletion at 20°C. Bead-bound Cdc20 and Plk1 were also analyzed by IB for validation (Panel b, right).

    As expected, the phospho-mimetic E2F8 fragment (T20D/T44D) remained stable in Plk1- and Cdc20-depleted mitotic extracts, serving as negative control (Panel c). In contrast, degradation of the non-phosphorylatable variant (T20A/T44A), as well as the APC/CCdc20 substrate Securin (positive control), was strongly hampered in Cdc20-depleted extracts compared to Plk1-depleted extracts. These results confirm that the untimely degradation of the non-phosphorylatable E2F8 in mitotic extracts is Cdc20-dependent.

    Figure legend: Untimely degradation of the non-phosphorylatable E2F8 in mitotic extracts is Cdc20-dependent.____a) Schematic of the immunodepletion (ID) protocol; additional technical details are provided below. b) Plk1 (top) and Cdc20 (bottom) levels in NDB mitotic extracts before and after three rounds of immunodepletion, as detected by immunoblotting (IB). Plk1 and Cdc20 levels were normalized to Tubulin and Cdk1, respectively. Both normalized and raw values are presented as percentages. Immunoprecipitation (IP) efficiency is shown on the right. c) Degradation profiles of phospho-mutant E2F8 variants and Securin (positive control) in NDB mitotic extracts depleted of Plk1 (control) or Cdc20.

    __ ---__

    Point 4. This study focuses on two proteins of the E2F family. These two proteins share similar domains, phosphorylation sites and a KEN box. However, their sensitivity to APC is different. What might explain this difference? Are there any inhibitory sequences for E2F7? Or why is the KEN box functional in E2F8 but not in E2F7?

    This is an excellent question. Here are our thoughts: The processivity of polyubiquitination by the APC/C varies between substrates in ways that influence degradation rate and timing (PMID: 16413484). Although E2F7 and E2F8 are related, their sequence identity is below

    50%, and their C-terminal domains differ substantially (see below) [FIGURE]. These structural differences likely contribute to differences in APC/C-mediated processivity and, consequently, to variations in protein half-lives. Additionally, E2F8 contains two functional KEN-boxes involved in its degradation, whereas E2F7 has only one. This may increase the kon rate of E2F8 for the APC/C, further enhancing its recognition and ubiquitination. Furthermore, re-examining the study by de Bruin and Westendorp (PMID: 26882548, Figure 2f; copied below), we note that the dynamic of inducibly expressed EGFP-tagged E2F7 in cells exiting mitosis is milder compared to E2F8 (see the black lines in both charts). This, as well as the oversensitivity of E2F7 degradation to Cdh1 downregulation accord with E2F7 being less potent substrate of APC/CCdh1.

    Figure legend: Adapted from Boekhout et al., 2016; ____EMBO Reports. Figure 2, Panel F. ____PMID: 26882548.

    The stability of the E2F7 fragment in cells and extracts was unexpected. We initially hypothesized that the unique N-terminal tail of E2F7 masks the KEN-box, functioning as an inhibitory sequence. However, removal of this region did not restore degradation (original manuscript; Figure 1e). Furthermore, extending the fragment by 20 additional residues failed to confer degradation (original manuscript; Figure S2). These observations suggest that E2F7 may require a distal or modular docking site for APC/C recognition. We did not pursue this question further.

    Point 5. An additional element that could strengthen this work would be referencing the study by Catherine Lindon: J Cell Biol, 2004 Jan 19;164(2):233-241. doi: 10.1083/jcb.200309035. In Figure 1 of this article, there is a degradation kinetics analysis of APC/C complex substrates such as Aurora-A/B, Plk1, cyclin B1, and Cdc20. This could help position the degradation of E2F7/8 relative to known APC/C targets. This can be achieved by synchronizing cells with nocodazole and then removing the drug to allow cells to progress and complete mitosis.

    This is an interesting point and one we should have clarified better previously. The temporal dynamics of E2F8 in synchronized HeLa S3 cells, relative to three known APC/C substrates, were reported in our previous study (PMID: 31995441; Figure 1a, copied on the right). Specifically, protein levels were measured for Cyclin B1, Securin, and Kifc1. Unlike Cyclin B1 and Securin, which are targeted by both APC/CCdc20 and APC/CCdh1, Kifc1 is degraded exclusively by APC/CCdh1. Cells were released from a thymidine–nocodazole block.

    Following Reviewer #3’s comment, we re-blotted the original HeLa S3 synchronous extracts. The new data [FIGURE] can be incorporated into the revised manuscript if requested.

    Point 6. Minor points: Does phosphorylation of E2F7-8 proteins alter their NMR profile? This could help understand how phosphorylation/dephosphorylation affects their sensitivity to the APC/C complex.

    Excellent suggestion. Indeed, we had originally aimed to include a more extensive set of NMR data in this manuscript. Our goal was to monitor E2F7 and E2F8 fragments in cell extracts and assess structural changes induced by phosphorylation and dephosphorylation during mitosis and mitotic exit. However, purifying the E2F7 fragment proved more challenging than anticipated. In addition, the extract-to-substrate ratio requires further optimization: Substrate concentrations must be high enough for reliable NMR detection, but below levels that would saturate the enzymatic activity in the extracts.

    That said, the short answer to the reviewer’s question is Yes: NMR profiles of E2F7 and E2F8 fragment do change following incubation with recombinant Cdk1–Cyclin B1 (see next page). If possible, we wish to exclude these NMR data from the manuscript.

    Point 7. Do these substrates bind to the APC/C complex before degradation? Does E2F7 bind better than E2F8?

    We were unable to detect interactions between endogenous E2F7 and E2F8 and the APC/C complex. In general, detecting endogenous E2F8, and especially E2F7, by immunoblotting proved challenging, making co-immunoprecipitation (Co-IP) even more difficult.

    Figure legend: NMR-based monitoring of E2F7 (a-c) and E2F8 (d-f) phosphorylation by Cdk1.

    a(d). 15N,1H-HSQC spectrum of E2F7(E2F8) prior to addition of Cdk1. Threonine residues of interest, T45 (T20) conforming to the consensus sequence (followed by a proline), and T84 (T60) lacking the signature sequence are annotated. b(e). Strips from the 3D-HNCACB spectrum used for assigning E2F7(E2F8) residues. Black (green) peaks indicate a correlation with the 13Cα (13Cβ) of the same and previous residues. The chemical shifts assigned to T45 (T20) and T84 (T60) match the expected values for K44(K19) and P83(P59), thereby confirming the assignment. c(f). Top, overlay of subspectra before adding Cdk1 (black) and after 16 h of activity (red) at 298 K. Bottom, change in intensities of the T45/T84 in E2F7 and T20/T60 in E2F8 showing how NMR monitors phosphorylation and distinguishes between various threonine residues.

    However, interactions between EGFP-tagged E2F7 snd E2F8 and Cdh1 have been demonstrated previously (PMID: 26882548, Figure 2e). In contrast, only the N-terminal fragment of E2F8, but not the corresponding fragment of E2F7, was found to bind Cdh1 (see figure on the right). This observation is consistent with the stability of the E2F7 fragment in APC/C-active extracts.

    __Figure legend: N-terminal fragment of E2F8 but not E2F7 binds Cdh1. __

    Co-Immunoprecipitation (IP) was performed in HEK293 cells transfected with EGFP-tagged E2F7/E2F8 fragments, using GFP-Trap® (Chromotek, Cat#: GTMA-20). Antibodies used for immunoblotting: ant-GFP (Santa Cruz Biotechnology: Cat#: SC-9996); anti-Cdh1 (Sigma-Aldrich, Cat#: MABT1323).

    Point 8. Why do the authors state that 250 µM of LB-100 has little effect on APC/C activity?

    We thank Reviewers #2 and 3 for raising this point. As shown in the manuscript, treatment with 250 µM LB-100 does not abolish APC/C-mediated degradation (otherwise, the assay would not be viable). However, it does attenuate degradation kinetics, as reflected by the prolonged half-lives of Securin and Geminin (see figure below).

    __Figure legend: APC/C-specific activity in cell extracts treated with LB-100. __

    Time-dependent degradation of EGFP–Geminin (N-terminal fragment of 110 amino acids) and Securin in extracts supplemented with LB-100 and/or UbcH10 (recombinant). A control reaction contained dominant-negative (DN) UbcH10. Proteins (IVT, 35S-labeled) were resolved by SDS-PAGE and autoradiography.

    Point 9. How can E2F8 be a substrate for both the SCF and APC/C complexes? (If I understood correctly.)

    This can happen because they are degraded by different E3 at different times during the cell cycle. To clarify further, certain proteins can be targeted by both the APC/C and SCF complexes, reflecting distinct regulatory needs. A classic example is CDC25A, as shown by M. Pagano and A. Hershko in 2002 (PMID: 12234927). Additional examples include the APC/C inhibitor EMI1 (PMIDs: 12791267 [SCF] and 29875408 [APC/C]).

    Reviewer #3 (Significance (Required)): This manuscript presents a well-structured study on the regulatory interplay between Cdk and Phosphatase in controlling the degradation of atypical E2Fs, E2F7 and E2F8. The work is relevant in the field of cell cycle regulation and provides new mechanistic insights into how phosphorylation and dephosphorylation govern APC/C-mediated degradation. The use of complementary cell-based and in vitro approaches strengthens the study, and the findings have significant implications for understanding the timing of transcriptional regulation in cell cycle progression.

    We wish to thank Reviewer #3 for their positive and encouraging view of our work.

  2. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    This manuscript presents a well-structured study on the regulatory interplay between Cdk and Phosphatase in controlling the degradation of atypical E2Fs, E2F7 and E2F8. The work is relevant in the field of cell cycle regulation and provides new mechanistic insights into how phosphorylation and dephosphorylation govern APC/C-mediated degradation. The use of complementary cell-based and in vitro approaches strengthens the study, and the findings have significant implications for understanding the timing of transcriptional regulation in cell cycle progression.

    • However, several points in this paper require further clarification for it to have a meaningful impact on the research community. The characterization of the phosphatase is unclear to me. The use of OA is necessary to guide the research, but it is not precise enough to rule out PP1 and then identify which PP2A is involved - PP2A-B55 or PP2A-B56. To clarify this, the regulatory subunits should either be eliminated or inhibited using the inhibitors developed by Jakob Nilsson's team. It would also be valuable for this study to investigate the mechanisms underlying this regulation. In particular, is it exclusive to E2F7-8 or could other substrates contribute to the generalisation of this regulatory process?

    • The observation that Cdc20 may target E2F8 is interesting, but needs to be further clarified to ensure that weak Cdh1 activity does not contribute to this degradation. Elimination of Cdc20 would be necessary to support the authors' conclusion.

    • This study focuses on two proteins of the E2F family. These two proteins share similar domains, phosphorylation sites and a KEN box. However, their sensitivity to APC is different. What might explain this difference? Are there any inhibitory sequences for E2F7? Or why is the KEN box functional in E2F8 but not in E2F7?

    • An additional element that could strengthen this work would be referencing the study by Catherine Lindon: J Cell Biol, 2004 Jan 19;164(2):233-241. doi: 10.1083/jcb.200309035. In Figure 1 of this article, there is a degradation kinetics analysis of APC/C complex substrates such as Aurora-A/B, Plk1, cyclin B1, and Cdc20. This could help position the degradation of E2F7/8 relative to known APC/C targets. This can be achieved by synchronizing cells with nocodazole and then removing the drug to allow cells to progress and complete mitosis.

    Minor points:

    • Does phosphorylation of E2F7-8 proteins alter their NMR profile? This could help understand how phosphorylation/dephosphorylation affects their sensitivity to the APC/C complex.

    • Do these substrates bind to the APC/C complex before degradation? Does E2F7 bind better than E2F8?

    • Why do the authors state that 250 µM of LB-100 has little effect on APC/C activity?

    • How can E2F8 be a substrate for both the SCF and APC/C complexes? (If I understood correctly.)

    Significance

    This manuscript presents a well-structured study on the regulatory interplay between Cdk and Phosphatase in controlling the degradation of atypical E2Fs, E2F7 and E2F8. The work is relevant in the field of cell cycle regulation and provides new mechanistic insights into how phosphorylation and dephosphorylation govern APC/C-mediated degradation. The use of complementary cell-based and in vitro approaches strengthens the study, and the findings have significant implications for understanding the timing of transcriptional regulation in cell cycle progression.

  3. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    Summary:

    The authors provide strong biochemical evidence that the regulation of E2F7 and E2F8 by APC is affected by CDK1 phosphorylation and potentially by PP2A dependent dephosphorylation. The authors use both full length and N-terminal fragments of E2F8 in cell-free systems to monitor protein stability during mitotic exit. The detailed investigation of the critical residues in the N-terminal domain of E2F8 (T20/T44) is well supported by the combination of biochemical and cell biology approaches.

    Major:

    It is unclear how critical the APC-dependent destruction of E2F7 and E2F8 is for cell cycle progression or other cellular processes. Prior studies have reported that Cyclin F regulation of E2F7 is critical for DNA repair and G2-phase progression. This study would be improved if the authors could provide a cellular phenotype caused by the lack of APC dependent regulation of E2F8 and/or E2F7.

    Minor:

    All optional: It would have been interesting to see the T20A/T44A/KM in the live cell experiment (Figure 3F). Figure 4C-D - include the corresponding blots for the WT E2F7. It is unclear how selective or potent the PP2A inhibitors are that are used in Figure 5. Is it possible to include known targets of PP2A (positive controls for PP2A inhibition) in the analysis performed in Figure 5? Is the APC still active in LB-100 or OA treated conditions? Is it possible to demonstrate the APC is active using known substrates in this assay (e.g., Securin (Cdc20) and Geminin (Cdh1) or similar).

    Significance

    Advance: A detailed analysis is provided for the critical N-terminal residues in E2F7 and E2F8 that when phosphorylated are capable of restricting APC destruction. The work builds on prior work that had identified the APC regulation of E2F7 and E2F8.

    Audience: The manuscript would certainly appeal to a broad basic research audience that is interested in the regulation of APC substrates and/or E2F axis control via E2F7 & E2F8. The study could have a broader interest if the destruction of E2F7 or E2F8 could be shown to be biologically relevant (e.g., critical for cell fate decision G1 vs G0, G1 length, timely S-phase onset, or expression of E2F1 target genes in the subsequent cell cycle).

  4. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    The study highlights a dephosphorylation switch mediated by PP2A as a critical mechanism for coupling E2F7/8 degradation to mitotic exit and G1 phase. The study is clear and experiments are well conducted with appropriate controls

    I have some concerns highlighted below :

    1. In this sentence : This intricate network of feedback mechanisms ensures the orderly progression of the cell cycle. What feedback mechanism are the authors referring to?
    1. Characterization of disorder in the N-terminal segments of E2F7 and E2F8

    What does it mean disorder in this title?

    1. In the paragraph on the untimely degradation of E2F8 the authors keep referring to APC/C Cdc20, however the degradation is triggered by the Ken box which is specifically recognised by APC/C Cdh1. Can it be due to another ligase not APC/C?

    1. The assays to detect dephosphorylation are rather indirect so it is difficult to establish whether phosphorylation of CDK1 and dephosphorylation by PP2A on the fragments is direct.

    2. Although there seems to be a control by phosphorylation and dephosphorylation (which could be indirect), it is difficult to establish the functional consequences of this observation. The authors propose a feedback mechanism which regulates the temporal activation inactivation of E2F7/8 however, there are no evidence in support of this.

    Significance

    The study is a good and well conducted work to understand the mechanisms regulating degradation of E2F7/8 by APC/C. This is crucial to establish coordinated celll cycle progression. While the hypothesis that disruption of this mechanism is likely responsible for altered cell cycle progression, there are no evidence this is just a back up pathway, whose functional significance could be limited to lack of APC/C Cdh1 activity. These experiments are rather difficult but the authors could comment on the limitation of the study and emphasise the hypothetical alterations which could result from the alterations of the described feedback loop

  5. Version published to 10.1101/2025.02.23.639703 on bioRxiv