1. Cancer Biology
  2. Cell Biology

Cyclin F drives proliferation through SCF-dependent degradation of the retinoblastoma-like tumor suppressor p130/RBL2

  1. Taylor P Enrico
  2. Wayne Stallaert
  3. Elizaveta T Wick
  4. Peter Ngoi
  5. Xianxi Wang
  6. Seth M Rubin
  7. Nicholas G Brown
  8. Jeremy E Purvis
  9. Michael J Emanuele  Is a corresponding author
  1. Department of Pharmacology. The University of North Carolina at Chapel Hill, United States
  2. Lineberger Comprehensive Cancer Center. The University of North Carolina at Chapel Hill, United States
  3. Department of Genetics. The University of North Carolina at Chapel Hill, United States
  4. Department of Chemistry and Biochemistry. University of California at Santa Cruz, United States
https://doi.org/10.7554/eLife.70691
Share this article
Cite this article
  1. Taylor P Enrico
  2. Wayne Stallaert
  3. Elizaveta T Wick
  4. Peter Ngoi
  5. Xianxi Wang
  6. Seth M Rubin
  7. Nicholas G Brown
  8. Jeremy E Purvis
  9. Michael J Emanuele
(2021)
Cyclin F drives proliferation through SCF-dependent degradation of the retinoblastoma-like tumor suppressor p130/RBL2
eLife 10:e70691.
https://doi.org/10.7554/eLife.70691
  2. Download BibTeX
  3. Download .RIS

Abstract

Cell cycle gene expression programs fuel proliferation and are universally dysregulated in cancer. The retinoblastoma (RB)-family of proteins, RB1, RBL1/p107, and RBL2/p130, coordinately represses cell cycle gene expression, inhibiting proliferation, and suppressing tumorigenesis. Phosphorylation of RB-family proteins by cyclin-dependent kinases is firmly established. Like phosphorylation, ubiquitination is essential to cell cycle control, and numerous proliferative regulators, tumor suppressors, and oncoproteins are ubiquitinated. However, little is known about the role of ubiquitin signaling in controlling RB-family proteins. A systems genetics analysis of CRISPR/Cas9 screens suggested the potential regulation of the RB-network by cyclin F, a substrate recognition receptor for the SCF family of E3 ligases. We demonstrate that RBL2/p130 is a direct substrate of SCFcyclin F. We map a cyclin F regulatory site to a flexible linker in the p130 pocket domain, and show that this site mediates binding, stability, and ubiquitination. Expression of a mutant version of p130, which cannot be ubiquitinated, severely impaired proliferative capacity and cell cycle progression. Consistently, we observed reduced expression of cell cycle gene transcripts, as well a reduced abundance of cell cycle proteins, analyzed by quantitative, iterative immunofluorescent imaging. These data suggest a key role for SCFcyclin F in the CDK-RB network and raise the possibility that aberrant p130 degradation could dysregulate the cell cycle in human cancers.

Editor's evaluation

The identification of the tumor suppressor RBL2/p130 as a substrate of cyclin F/SCF adds a new level of understanding about the role of this ubiquitin ligase in cell cycle control and identifies a novel functional interaction that could have implications for cancer. This work will be of interest to researchers in the fields of cell cycle and cancer biology.

https://doi.org/10.7554/eLife.70691.sa0

Introduction

The eukaryotic cell cycle consists of a sequential progression of events that govern cell growth and division. During cell cycle progression, many hundred genes oscillate in expression, contributing to myriad processes, including DNA replication, chromosome segregation, cytoskeletal organization, and so on. The expression of cell cycle genes is repressed during quiescence, a reversible state of growth arrest, and in early G1-phase, presenting a critical barrier to proliferation. The retinoblastoma protein (RB) is a vital regulator of cell cycle gene repression. During quiescence and in early G1-phase, RB binds and inhibits E2F transcription factors, repressing transcription of many cell cycle genes. RB is phosphorylated by Cyclin-Dependent Kinases 4 and 6 (CDK4/6), as well as CDK2 (Narasimha et al., 2014; Rubin et al., 2020). This phosphorylation causes RB to dissociate from E2F, promoting the E2F-dependent expression of cell cycle genes that catalyze S-phase entry and cell cycle progression (Rubin et al., 2020). Due to its critical role in restricting proliferation, RB is a prototypical tumor suppressor (Dyson, 2016).

RB has two closely related family members, RBL1/p107 and RBL2/p130 (Dyson, 1998; Sadasivam and DeCaprio, 2013). Both p107 and p130 are also tumor suppressors that prevent cell cycle gene expression by binding the repressor E2F proteins E2F4/5 (Claudio et al., 1994; Zhu et al., 1993), and both are also regulated by CDK-dependent phosphorylation (Canhoto et al., 2000; Farkas et al., 2002; Hansen et al., 2001). Additionally, p130 functions as part of the mammalian DREAM complex (DP, RB-like, E2F4/5, and MuvB) (Litovchick et al., 2007; Smith et al., 1996). DREAM assembles during quiescence and inhibits cell cycle progression by restricting the transcription of numerous cell cycle genes regulated by E2F, B-MYB, and FoxM1 transcription factors (Fischer et al., 2016; Müller et al., 2012). Accordingly, perturbations to p130 or the DREAM complex allow expression of its cell cycle target genes, shifting the balance from quiescence toward proliferation (Forristal et al., 2014; Iness et al., 2019; Patel et al., 2019).

RB and p130 collaborate to suppress proliferation. In mice, Rb−/−p130−/− mouse embryo fibroblasts (MEFs) grow more rapidly in culture than MEFs deficient in either Rb or p130 alone, and Rb−/−p130−/− mice spontaneously form many more tumors than their respective single-gene knockouts (Dannenberg et al., 2004). In mouse models of small cell lung cancer, p130 knockout increases tumor size and overall tumor burden, even in the background of Rb and p53 loss (Ng et al., 2020; Schaffer et al., 2010). Consistent with its role as a tumor suppressor, p130 cooperates with RB to repress G2-M genes in response to genotoxic stress (Schade et al., 2019). And, p130 loss in primary human fibroblasts leads to increased expression of cell cycle genes compared to loss of Rb alone (Schade et al., 2020). These observations highlight the importance of p130 in cell cycle control, as well as its role in tumor suppression. These results also illustrate the importance of the broader CDK-RB network in normal proliferation, and the consequence of its dysregulation in the aberrant cell cycles observed in cancer. RB mutations, overexpression of cyclin D and cyclin E, loss of p130 protein, and dysregulation of the mammalian DREAM complex have all been implicated in increased cellular proliferation and tumorigenesis (Forristal et al., 2014). Interestingly, p130 mutations are infrequent compared to other tumor suppressors like RB, CDKN2A, and p53, suggesting the possibility that post-translational mechanisms could account for its inactivation.

Cyclin F is a non-canonical cyclin—it neither binds nor activates CDKs (Bai et al., 1994; D’Angiolella et al., 2013). Instead, cyclin F is one of ~70 F-box proteins, a family of substrate recognition receptors that recruit substrates to the Skp1-Cul1-Fbox protein (SCF) E3 ligase (Bai et al., 1996; Cardozo and Pagano, 2004). SCF ligases play an evolutionarily conserved role in promoting cell cycle progression by triggering the destruction of cell cycle inhibitors. For example, yeast SCFCdc4 and human SCFSkp2 trigger the destruction of CDK inhibitors Sic1 and p27, respectively (Carrano et al., 1999; Feldman et al., 1997; Schwob et al., 1994; Skowyra et al., 1997).

Cyclin F mRNA and protein levels oscillate during the cell cycle, giving cyclin F cell cycle-dependent activity (Bai et al., 1994). Cyclin F begins to accumulate at the G1/S transition, peaks in G2, and its protein levels are subsequently downregulated via proteasomal degradation in mitosis and G1 (Choudhury et al., 2016; Mavrommati et al., 2018) where cyclin F is ubiquitinated by SCFβTRCP following cyclin F phosphorylation by casein kinase II (Mavrommati et al., 2018). Cyclin F is also ubiquitinated by the cell cycle E3 Anaphase Promoting Complex/Cyclosome (APC/C) (Choudhury et al., 2016). Further, cyclin F has been implicated in the ubiquitination and degradation of several cell cycle proteins (Emanuele et al., 2020). Taken together, its dynamic regulation and substrate repertoire highlight the importance of cyclin F in cell cycle control.

We demonstrate here that cyclin F binds, ubiquitinates, and regulates the RB-family tumor suppressor p130. We identify the regions in both cyclin F and p130 that are required for their interaction. Interfering with cyclin F-p130 regulation, by expressing a mutant version of p130 which cannot be ubiquitinated, causes a severe defect in proliferation. Taken together, these data implicate cyclin F as a new, key player in CDK-RB network regulation, highlighting a critical ubiquitin-based mechanism that controls cell proliferation.

Results

Cyclin F/CCNF fitness correlates with the CDK-RB network

The Project Achilles Cancer Dependency Map (DepMap) has performed near genome-wide, CRISPR/Cas9 loss-of-function screens in hundreds of cancer cell lines. Fitness scores, corresponding to each gene in each cell line, reflect the relative impact of individual gene knockout on proliferation and survival. Despite differences in protocols and reagents, DepMap scoring is highly concordant with Project Score, another large, pan-cancer, CRISPR/Cas9 screening platform (Dempster et al., 2019). Importantly, systems genetics analyses of these data can be used to identify physically or genetically linked genes/proteins, since their loss often similarly impact overall fitness (Munoz et al., 2016; Boyle et al., 2018; Doench et al., 2016; Kory et al., 2020; McDonald et al., 2017; Pan et al., 2018; Wang et al., 2017). The Pearson’s correlation coefficient was computed for DepMap fitness gene scores across cell lines (17,634 genes in 789 cell lines at the time of analysis). This analysis effectively detects pathways involved in proliferation. For example, the gene most highly correlated with the DNA helicase component MCM2 is its complex member MCM4 (Figure 1—figure supplement 1A). The genes most highly correlated with mitogen-activated protein kinase MEK1/MAP2K1 are its upstream activator BRAF and its downstream effector ERK2/MAPK1 (Figure 1—figure supplement 1A). Genes that encode proteins in the Skp2-Cyclin E/A-CDK2-E2F pathway are highly correlated, as are those involved in autophagy (Figure 1—figure supplement 1A).

Fitness scores for those genes in CDK-RB network which promote proliferation (e.g., CDK4, CCND1/Cyclin D1, etc.) are well correlated between DepMap and Project Score (Figure 1—figure supplement 2; Dempster et al., 2019). The gene most highly correlated with CDK4 is its coactivator CCND1/Cyclin D1, demonstrating that relevant genetic relationships in the CDK-RB network are correlated (Figure 1—figure supplement 3A). Interestingly, the ninth most highly correlated gene with CDK4 was CCNF, which encodes cyclin F (Bai et al., 1994; Bassermann et al., 2014). CCNF is among the top 1% of highly correlated genes when analyzing several other members of the CDK-RB network, including CDK4, CDK6, CCND1, and RBL1 (Figure 1—figure supplement 3A). Accordingly, when we determined those genes most highly correlated with CCNF knockout, CDK4, CDK6, CCND1/Cyclin D1, RBL1, and RBL2 are among the top 85 genes, out of more than 17,630 (~0.5%; Figure 1A). Thus, the impact of CCNF knockout on fitness is highly correlated with genes in the CDK-RB network.

Figure 1 with 3 supplements see all
Download asset Open asset
Analysis of the Cancer Dependency Map reveals that CCNF is highly correlated with the CDK-RB network.

(A) Cancer Dependency Map data from project Achilles were analyzed to identify the impact of gene loss-of-function on cellular fitness, and fitness correlation with that of CCNF, based on pooled CRISPR/Cas9 gene knockout screens performed in 789 cell lines. Pearson’s correlation coefficients are reported for all gene pairs (each dot corresponds to a single CCNF-gene X pair). The Pearson’s correlations for CCNF compared to 17,634 other genes are shown. The CDK-RB network members highlighted in red all score in the top 0.5% of genes whose impact on fitness is most highly correlated with CCNF. (B) Gene ontology (GO) analysis was performed for the top 0.05% of genes whose impact on fitness is most highly correlated with CCNF. The top 10 enriched GO terms and their corresponding p-value is shown. (C) The top 0.05% of genes whose impact on fitness is most highly correlated with CCNF were sorted by the GO term cell division (GO:0051301). A graphical representation of the remaining 21 genes is shown. Genes are grouped by their known associations with specific functional pathways or complex, including CDK-RB, Nde1-kinetochore, or APC/C. (D) MCF-7 and T47D cells were engineered to contain a TET-inducible cyclin F transgene. Cells were treated with 0 (vehicle control), 5, 25, or 100 ng/ml of doxycycline to induce cyclin F expression, and the indicated proteins were analyzed by immunoblot. No band was detected for CDK6 in T47D cells. Representative of n=3 experiments. (E) BJ, IMR-90, and NHF-1 human fibroblast cell lines were synchronized in G0 by 48 hr serum starvation (quiescent) or allowed to proliferate normally (asynchronous). Whole-cell extracts were collected for immunoblot analysis. Representative of n=3 experiments. (F) NHF-1 cells were synchronized in G0 by serum starvation for 48 hr. Cells were then released into the cell cycle upon the addition of serum-containing media supplemented with either MLN4924 or vehicle (DMSO) as a control. Cells were collected at the indicated time points after release. Protein levels were assessed by immunoblot. Data represent n=3 independent experiments.

Gene ontology (GO) analysis on the top 94 genes correlated with CCNF (Pearson correlation>0.15) enriched GO terms related to cell cycle, regulation of ubiquitin protein ligase activity, mitotic G1-phase and G1/S transition, Cyclin D1-CDK4-CDK6 complex, G1/S specific transcription, and Cyclin A-CDK2 associated events at S-phase entry (Figure 1B). When we filtered the top CCNF correlated genes by the GO term ‘cell division’ (Figure 1C), three sub-networks emerged: the Nde1-kinetochore complex, the Anaphase Promoting Complex/Cyclosome (APC/C), and the CDK-RB network. Identification of eight subunits of the APC/C complex is consistent with our previous observation of a reciprocal relationship between SCFcyclin F and APC/CCdh1 (Choudhury et al., 2016). Among the remaining genes/proteins were two known cyclin F substrates: E2F7, which encodes the transcriptional repressor E2F7 (Yuan et al., 2019), and FZR1, which encodes the APC/C coactivator protein, Cdh1 (Choudhury et al., 2016).

We examined endogenous proteins corresponding to CDK-RB network genes that correlate most strongly with CCNF using two breast cancer cell lines, MCF-7 and T47D, engineered for doxycycline-inducible expression of cyclin F (Wasserman et al., 2020). Cells were treated with increasing amounts of doxycycline for 24 hr, and immunoblotted for CDK4, CDK6, cyclin D1, p107, and p130. We observed a specific downregulation of endogenous p130 in both cell lines, whereas all other proteins remained unchanged (Figure 1D). Similarly, p130 levels decrease in a time-dependent manner following doxycycline treatment, and this can be prevented by inhibiting the proteasome with MG132, or cullin RING E3 ligases, using the neddylation inhibitor MLN4924 (Figure 1—figure supplement 3B; Ohh et al., 2002; Soucy et al., 2009). Taken together, these data suggest that cyclin F might regulate p130 degradation.

To further assess the relationship between cyclin F and p130, we compared asynchronous human fibroblasts to those same cells synchronized in quiescence by serum-starvation for 48 hr. In each of cell lines tested (BJ, IMR-90, and NHF-1), cyclin F protein levels significantly decrease in quiescence, whereas p130 protein levels increase, establishing an inverse expression pattern (Figure 1E). Next, NHF-1 cells were synchronized in quiescence by serum deprivation and then released into the cell cycle. Cyclin F protein began to accumulate as cells re-entered the cell cycle at G1/S, similar to what was previously reported (Choudhury et al., 2017; D’Angiolella et al., 2012; Figure 1—figure supplement 3C). Meanwhile, p130 was degraded as cells exited quiescence. The accumulation kinetics of cyclin F is similar to another F-box protein, Skp2, which ubiquitinates p27, marking it for degradation. Skp2 has also been implicated in p130 degradation (Bhattacharya et al., 2003; Tedesco et al., 2002). However, p130 has a different degradation pattern than p27, and in a previous study, p130 was still degraded in Skp2 knockout cells (Tedesco et al., 2002).

To determine whether the SCF and cullin E3 ligase family is required for p130 degradation at quiescence exit, we synchronized NHF-1 and T98G cells in quiescence and then released cells into the cell cycle in the presence of either DMSO (control) or MLN4924. In cells treated with MLN4924, neither p130 nor p27 is degraded, cyclin E accumulates but does not get degraded, and there is a significant delay in cyclin A accumulation (Figure 1F and Figure 1—figure supplement 3D).

To determine p130 levels in the absence of cyclin F, we first utilized CCNF knockout HeLa cells generated using CRISPR/Cas9 gene editing (Choudhury et al., 2016). In HeLa cells, p130 protein levels are higher in CCNF KO cells (sgCCNF) than in control cells (sgCTRL; Figure 2A). Next, CCNF KO and control HeLa cells were synchronized at G1/S by double thymidine block. After release, p130 levels were higher in the CCNF knockouts at every time point analyzed (Figure 2—figure supplement 1A). We then transiently depleted cyclin F from NHF-1 or IMR-90 cells. Cyclin F depletion increased p130 in both cell lines and with two different siRNAs (Figure 2B). Neither CCNF KO HeLa cells, nor the transiently depleted NHF-1 or IMR-90 cells, exhibited significant defects in overall cell cycle compared to their respective controls, judged by propidium iodide staining and analyzed by flow cytometry (Figure 2—figure supplement 1B-D). Consistent with the regulation of p130 by cyclin F resulting from a physical interaction, endogenous cyclin F co-immunoprecipitated endogenous p130 from HEK293T and U2OS cells (Figure 2C).

Figure 2 with 1 supplement see all
Download asset Open asset
Cyclin F regulates and interacts with endogenous p130.

(A) Asynchronously proliferating CCNF CRISPR/Cas9 knockouts (sgCCNF) and control (sgCtrl) HeLa cells were blotted for levels of the indicated proteins. Representative of n=3 experiments. (B) NHF-1 and IMR-90 cells were transfected with two different siRNAs targeting CCNF or a control siRNA targeting Firefly Luciferase (siFF). Whole-cell lysates were immunoblotted for the indicated proteins. Representative of n=3 experiments. (C) Endogenous cyclin F was immunoprecipitated from asynchronously proliferating HEK293T cells (left) or asynchronously proliferating U2OS cells (right). Indicated proteins were immunoblotted. SE=short exposure; LE=long exposure; representative of n=3 experiments.

p130 degradation is proteasome- and neddylation-dependent and requires the canonical cyclin F substrate-binding site

We determined if exogenously expressed cyclin F affects the stability of exogenously expressed p130 protein. In HEK293T cells transfected with HA-tagged p130 (HA-p130), increasing amounts of FLAG-tagged cyclin F (FLAG-cyclin F) reduced p130 levels in a dose-dependent manner (Figure 3A). Four human SCF E3 ligases are involved in marking substrates for degradation to promote entry and progression through S-phase: SCFcyclin F, SCFSkp2, SCFβTRCP, and SCFFbxw7. We therefore transiently expressed HA-p130 alone or together with cyclin F, Skp2, Fbxw7, βTRCP1, or βTRCP2, and in both HEK293T and U2OS cells. p130 levels decreased only upon co-expression with cyclin F and not with any of the other F-box proteins (Figure 3B and Figure 3—figure supplement 1A).

Figure 3 with 1 supplement see all
Download asset Open asset
Cyclin F promotes p130 degradation.

(A) HEK293T cells transiently expressing HA-p130 with an empty FLAG vector control (lane 1) or together with increasing amounts of FLAG-cyclin F (lanes 2–4). Cells were collected and analyzed by immunoblot 24 hr post-transfection. The antigen being immunoblotted for is represented by the underline, here and in all experiments below. Representative of n=3 experiments. (B) HEK293T cells transiently expressing HA-p130 with an empty FLAG vector control (lane 1) or together with FLAG-cyclin F (lanes 2–4). MG132 (proteasome inhibitor) or MLN4924 (neddylation inhibitor) were added for 6 hr prior to harvesting. Cells were collected and analyzed by immunoblot 24 hr post-transfection. Representative of n=3 experiments. (C) HEK293T cells transiently expressing HA-p130 with an empty FLAG vector control (lane 1) or together with FLAG-cyclin F WT (lane 2) or FLAG-cyclin F(M309A L313A) (lane 3), as indicated. Cells were collected and analyzed by immunoblot 24 hr post-transfection. Representative of n=3 experiments. (D) HEK293T cells transiently expressing HA-p130 with an empty FLAG vector control (lane 1) or together with the indicated FLAG-tagged F-box proteins (lanes 2–6). Cells were collected and analyzed by immunoblot 24 hr post-transfection. Representative of n=3 experiments.

In HEK293T cells, the reduction in HA-p130 caused by FLAG-cyclin F expression is reversed by addition of MG132 or MLN4924 (Figure 3C). Cyclin F recognizes its substrates through a hydrophobic patch motif in its cyclin domain (sequence MRYIL at amino acids M309-L313). Previous studies have shown that mutating the methionine and leucine residues to alanine (Cyclin F M309A L313A) can impair cyclin F binding to substrates (Choudhury et al., 2016; D’Angiolella et al., 2012; D’Angiolella et al., 2010). Notably, whereas cyclin F (WT) robustly downregulated p130 protein levels, cyclin F (M309A L313A) was unable to promote p130 degradation (Figure 3D). These data suggest that p130 is a cyclin F substrate.

p130 is a cyclin F substrate

Cyclin F recognizes Cy Motifs in substrates, which corresponds to the amino acid sequence RxL/I (where x=any amino acid)(Choudhury et al., 2016; D’Angiolella et al., 2012; D’Angiolella et al., 2010). There are 13 putative Cy motif sequences in p130 that span the length of the protein. We therefore created six HA-tagged p130 truncations, while considering p130 domain structure (Figure 4A). We transiently expressed each HA-p130 truncation alone, or together with FLAG-cyclin F (WT) and included conditions with and without proteasome and neddylation inhibitors. p130 degradation depended on the presence of a flexible spacer region, located between regions A and B in the p130 pocket domain (Figure 4A, Figure 4—figure supplement 1A). Within the spacer, there are two RxL/I motifs: R658-I660 and R680-L682. We mutated the first and last amino acid in each motif to alanine (AxA) in full-length HA-p130. Only the p130 (R680A L682A) mutant was resistant to cyclin F mediated degradation, demonstrating that these two amino acids alone are required for cyclin F to trigger p130 degradation (Figure 4B). That sequence and many surrounding residues are conserved in p130 proteins found in other species, including clawed frogs (Xenopus tropicalis), chickens (Gallus gallus), and mice (Mus musculus) (Figure 4B).

Figure 4 with 1 supplement see all
Download asset Open asset
Cyclin F binds p130 directly and promotes p130 degradation through a conserved degron motif.

(A) Graphical depiction of p130 domain structure. Indicated p130 truncation mutants were screened for their ability to be degraded following co-overexpression with cyclin F. The data supporting these conclusions are shown in Figure 4—figure supplement 1. (B) The first and last amino acids in the two potential cyclin F binding sites in the p130 spacer domain (R658-I660 and R680-L682) were mutated to alanine (AxA). HEK293T cells transiently expressed HA-p130 alone (WT or AxA mutants, as indicated) or together with FLAG-cyclin F WT. Cells were collected and analyzed by immunoblot 24 hr post-transfection. Representative of n=3 experiments (top) and amino acid sequence alignment for human, mouse, frog, and chicken p130 (bottom). (C) GST-p130593–790 (WT) and GST-p130593–790 (AA) were produced in Escherichia coli and purified. FLAG-cyclin F was transiently expressed in HEK293T cells. GST pulldowns (left) and FLAG pulldowns (right) were used to determine the interaction of p130 and cyclin F. Interaction was assessed by immunoblot after pulldown (representative of n=3 experiments). (D) GST-p130(WT) and FLAG-cyclin F(WT) or FLAG-cyclin F(M309A L313A) were expressed as described in (C). Interaction and binding were assessed as in (C) (representative of n=3 experiments). (E) Fluorescence polarization anisotropy assay to detect direct association of p130 with cyclin F. (Left) 10 nM TAMRA-p130674–692 probe was titrated with increasing concentrations of purified GST-cyclin F25–546-Skp1. (Right) The p130 probe bound with 0.5 μM GST-cyclin F25–546-Skp1 was displaced with increasing concentrations of the indicated p130 protein construct or E2F184–99 peptide. Experiments were performed in triplicate, and the standard deviation is reported as the error.

Since p130(R680A L682A; hereafter referred to as p130(AA)) cannot be degraded by cyclin F, we asked whether it still binds to cyclin F. GST-tagged p130 truncations were purified from Escherichia coli (amino acids 593–790; hereafter referred to as GST-p130), and mixed with total lysates from HEK293T cells expressing FLAG-cyclin F. In a GST-pulldown assay, FLAG-cyclin F bound to GST-p130(WT), but not to GST-p130(AA) or GST alone (Figure 4C, left panel). Conversely, FLAG-cyclin F, purified from HEK29T cells precipitated GST-p130(WT) but not GST-p130(AA) (Figure 4C, right panel). Consistent with its inability to trigger p130 degradation, cyclin F(M309A L313A) could not bind to GST-p130 (Figure 4D). Thus, the p130 Cy motif at R680-L682 and the cyclin F hydrophobic patch (M309-L313) are required for the p130-cyclin F interaction and for p130 degradation.

To assess the direct binding and affinity of p130 for cyclin F, we developed a fluorescence polarization anisotropy assay using a peptide containing the relevant Cy motif in p130. We used a version of cyclin F, produced in Sf9 cells, which lacks the first ~25 amino acids and its C-terminal PEST domain to improve solubility (cyclin F25–546). We titrated recombinant, purified cyclin F25–546-Skp1 dimer into a solution of synthetic, TAMRA-labeled, p130 peptide (residues 674–692) and observed a Kd of 0.21±0.01 μM (Figure 4E, left panel). Binding was efficiently competed away using purified GST-p130(WT) protein but not using GST-p130(AA), which had >50-fold higher Ki in the competition assay (Figure 4E, right panel). We found that an unlabeled E2F1 peptide containing the relevant Cy motif (residues 84–99), which was previously shown to mediate cyclin F-E2F1 association (Clijsters et al., 2019), also competed with the p130 probe. We performed a similar competition assay using a TAMRA labeled E2F184–99 peptide as a probe, and we found that we could efficiently compete the E2F1 peptide with GST-p130(WT) but not GST-p130(AA) (Figure 4—figure supplement 1B). We conclude that p130 directly binds cyclin F-Skp1 at a site that overlaps with the E2F1 binding site and that the association is dependent on critical interactions made by R680 and L682 in p130.

To analyze p130 ubiquitination, we reconstituted the SCFcyclin F ubiquitination reaction in vitro using purified components. Cyclin F25–546 was combined with neddylated Cul1-Roc1(Rbx1), Skp1, substrate, and ubiquitin. The human homolog of ariadne (ARIH1) is a RING-Between-RING (RBR) E3 ligase that has been shown to work with the E2 UBCH7 and other cullin ring E3 ligases, including the SCF, to ubiquitinate substrates (Horn-Ghetko et al., 2021; Scott et al., 2016). Therefore, ARIH1/UBCH7 and the chain-elongating E2 CDC34B were used as the ubiquitin transfer module.

SCFCyclin F robustly ubiquitinated GST-p130(WT) (Figure 5A). This modification was dependent on the inclusion of NEDD8-Cul1-Roc1, Skp1-Cyclin F, and ARIH1/UBCH7, since the exclusion of any of these components completely abrogated ubiquitination (Figure 5A). Significantly, GST-p130(AA) was unable to be ubiquitinated by SCFcyclin F.

Figure 5 with 2 supplements see all
Download asset Open asset
SCFCyclin F regulates the ubiquitination and stability of p130.

(A) Ubiquitination reactions were performed with SCFcyclin F, ARIH1/UBCH7, and CDC34B. GST-tagged p130593–790 WT and GST-p130593–790 were used as substrates and detected by immunoblot against GST. Data are representative of n=3 experiments. (B) FLAG-cyclin F, HA-p130(WT), and/or HA-p130(AA) were transiently expressed in HEK293T cells for 24 hr. The protein synthesis inhibitor cycloheximide (CHX) was added and cells were collected at the indicated time points. Protein levels were determined by immunoblot. Representative of n=3 experiments. (C) Quantification of (B). *=p<0.05 (Student’s t-test). Data are shown as mean ± SEM for n=3 experiments. (D) Inducible p130 NHF-1 cells were grown in media-containing 100 ng/ml doxycycline for 14 days to induce p130 expression, or in media-containing vehicle control. On indicated days, cells were pulsed with EdU for 30 min prior to harvest/fixation, DNA was stained with DAPI, and cells were analyzed by flow cytometry. Data for S- and G2/M phases are in Figure 6—figure supplement 1D. Data represent mean ± SEM for n=3 replicates.

Because cyclin F can ubiquitinate p130(WT), we determined the stability of p130(WT) and p130(AA) in the absence or presence of cyclin F. To determine p130 half-life, we expressed full-length HA-p130(WT) or HA-p130(AA) with or without FLAG-cyclin F in HEK293T or U2OS cells for 24 hr. Cells were treated with the protein synthesis inhibitor cycloheximide (CHX) and then analyzed for HA-p130 half-life. While p130(WT) half-life was significantly reduced by ectopic expression of cyclin F, p130(AA) remained stable for the length of the experiment in the presence or absence of cyclin F (Figure 5B–C and Figure 5—figure supplement 1A), indicating that protein levels of p130(AA) are resistant to cyclin F-mediated degradation.

We asked whether p130(AA) accumulates more than p130(WT) during a specific cell cycle phase. To evaluate p130 levels at different points during the cell cycle, we first engineered NHF-1 cells to express HA-tagged p130(WT) or p130(AA) in response to doxycycline treatment (Figure 5—figure supplement 1B). NHF-1 cells expressing p130 were synchronized with vehicle control, aphidicolin (S-phase synchronization), or nocodazole (Mitosis synchronization) for 16 hr. Interestingly, p130(WT) levels are decreased in nocodazole-synchronized cells compared to p130(AA) levels (Figure 5D). Consistently, nocodazole synchronization significantly decreased endogenous p130 protein levels in NHF-1 cells (Figure 5—figure supplement 1C). Taken Together, these data indicate that p130(WT) levels decrease in nocodazole synchronized NHF-1 cells, whereas levels of p130(AA) remain stable.

The Cy motif that we mapped at amino acids 680–682 was previously implicated in binding cyclins A and E in SF9 cell extracts (Lacy and Whyte, 1997). To evaluate binding in human cells, we utilized the NHF-1 cells engineered to express p130(WT) or p130(AA). We immunoprecipitated (IP) HA and found that similar amounts of cyclin A and E co-purify with HA-p130(WT) and HA-p130(AA) (Figure 5—figure supplement 2A). Reciprocally, following IP of endogenous cyclin A, both HA-p130(WT) and HA-p130(AA) co-immunoprecipitated (Figure 5—figure supplement 2B). Further, we assessed whether mutation of the p130 Cy motif would impair p130 phosphorylation. HA-tagged p130(WT) and p130(AA) immunoprecipitated from NHF-1 cells are similarly phosphorylated on the CDK site at S672 (Figure 5—figure supplement 2C). Additionally, HA-p130(WT) and HA-p130(AA) migrate similarly in SDS-PAGE, their migration was similarly increased by post-lysis phosphatase treatment with calf intestinal phosphatase (CIP), and this could be prevented by the addition of phosphatase inhibitors (Figure 5—figure supplement 2D). These findings are consistent with previous reports mapping additional cyclin E and A binding sites to sequences in the p130 amino-terminus, and cyclin D binding to a carboxy-terminal helix (Castaño et al., 1998; Hansen et al., 2001; Topacio et al., 2019). Further, HA-tagged p130(WT) and p130(AA) can associate with the DREAM complex based on the co-IP of Lin54 following an HA pulldown (Figure 5—figure supplement 2C). Finally, we determined that the Cy motif mutations did not impact p130 binding to large T-antigen in HEK293T cells or the E7 oncoprotein in HeLa cells (Figure 5—figure supplement 2E-F).

Loss of P130 regulation by cyclin F causes G0/G1 arrest and apoptosis

Interfering with RB phosphorylation, by expression of a mutant harboring substitutions at CDK phosphorylation sites, impairs cell cycle progression (Fry et al., 2004; Lukas et al., 1997). To similarly isolate the phenotypic impact of cyclin F on p130, independent of other substrates, we utilized the NHF-1 cells engineered with a doxycycline inducible expression of p130(WT) or p130(AA) (Figure 5—figure supplement 1A). To determine whether p130(AA) accumulates more than p130(WT), we collected cells at various time points after induction. Although protein levels of p130(WT) and p130(AA) are initially equal, p130(AA) accumulated to higher levels compared to p130(WT) over time (Figure 6A), consistent with p130(AA) being resistant to degradation.

Figure 6 with 2 supplements see all
Download asset Open asset
Cells expressing p130(AA) exhibit proliferation defects.

(A) NHF-1 cells were engineered to express a TET-inducible HA-p130(WT) or HA-p130(AA) transgene. Doxycycline (100 ng/ml) was used to induce p130 expression, and water was used as a vehicle control. HA-p130 levels were assessed by immunoblot after cells were grown in doxycycline for up to 14 days. (B) Inducible p130 NHF-1 cells were grown in 100 ng/ml doxycycline-containing media for 14 days to induce p130 expression. Cells were counted on the indicated days. Data represent mean ± SEM for n=3 independent experiments. (C) Doubling time was calculated from counting experiment in (B). Error bars are SEM for n=3 independent experiments. Representative of n=2 independent experiments. (D) NHF-1 cells were grown in media supplemented with nocodazole for 16 hr, then immunoblotted for the indicated proteins. Representative of n=2 independent experiments. (E–G) Inducible p130 NHF-1 cells were grown in 100 ng/ml doxycycline for indicated times. Cells were fixed and analyzed by iterative immunofluorescent staining and imaging. Distributions of single-cell measurements are shown for nuclear phosphorylated versus total RB (E, left) and representative images of phosphorylated RB are shown for indicated days (E, right). Distributions of single-cell measurements are also shown for cytoplasmic p21 (F) and nuclear versus cytoplasmic p130 (G), as indicated.

To assess proliferation, we induced expression of low levels of p130(WT) or p130(AA) and allowed cells to proliferate in culture for 14 days. Cells were split regularly, provided fresh media supplemented with doxycycline every 48 hr, and never reached confluency. Expression of non-degradable p130(AA) led to a profound decrease in proliferation/viability based on a PrestoBlue assay (Figure 6—figure supplement 1A). After 14 days, the PrestoBlue signal was decreased 5.8-fold (~83%) in p130(AA)-expressing cells relative to p130(WT) controls.

Next, we directly quantified proliferation by counting cells every 48 hr for 2 weeks after p130(WT) or p130(AA) induction. The p130(AA)-expressing cells proliferated much more slowly than the p130(WT)-expressing cells (Figure 6B). After 14 days, there were 7.7-fold (~87%) less cells in the p130(AA) population compared to p130(WT)-expressing controls. The average doubling time was 33 hr for p130(WT)-expressing cells and 48 hr for p130(AA)-expressing cells (Figure 6C). As a control, both cell lines were treated with vehicle alone. There was no difference in proliferation across a 2-week experiment (Figure 6—figure supplement 1B-C). Expectedly, ectopic expression of p130(WT) slowed the doubling time compared to cells treated with vehicle, consistent with previous studies which report p130 overexpression slows cell growth (Lacy and Whyte, 1997). However, p130(AA) had a much more severe impact on proliferation/survival.

We investigated whether the slow growth observed following p130(AA)-induction resulted from cell cycle arrest. We performed flow cytometry to analyze cell cycle phase distribution using EdU incorporation and DNA staining. NHF-1 were pulsed with EdU for 30 min prior to fixation at several time points after p130 induction. Total DNA was stained with DAPI, and cells were analyzed by flow cytometry. The percent of p130(AA)-expressing cells in G0/G1 significantly increased at days 5 and 7 compared to p130(WT)-expressing cells, while the percentage of S-phase cells was significantly decreased (Figure 6D and Figure 6—figure supplement 1D). Surprisingly, beyond day 10, there was no difference in G0/G1% between p130(AA) and p130(WT)-expressing cells, suggesting that the cell population may adapt to the increased levels of p130, which remain elevated throughout the entire 2-week experiment (Figure 6D).

We used iterative indirect immunofluorescence imaging (4i) to determine the levels of cell cycle markers in p130(WT)- and p130(AA)-expressing cells. 4i is accomplished by repeated antibody staining, followed by imaging, stripping, and re-staining, that collectively produce a set of targeted protein measurements in single cells. The resulting images are ‘stacked’ to quantify the expression of multiple proteins and antigens across an identical set of individual cells (Gut et al., 2018). Automated and semi-automated methods are used to identify cell boundaries and extract quantitative protein-level information for each cell (Gut et al., 2018).

To determine whether cells were arresting in G1, we stained for total and phospho-RB. RB is phosphorylated in proliferating cells, while unphosphorylated RB is a hallmark of G1/G0. RB phosphorylation was decreased in p130(AA)-expressing cells at days 5, 7, and 10 compared to p130(WT)-expressing cells or vehicle-treated controls (Figure 6E). We also found that p21 was increased in p130(AA)-expressing cells at days 5, 7, and 10, consistent with a G0/G1 arrest (Figure 6F). To assess cell cycle distribution in each population, we quantified total DNA content. DNA content analysis revealed that a larger percentage of p130(AA)-expressing cells had <4C DNA content compared to p130(WT)-expressing cells at days 5, 7, and 10, consistent with cells accumulating in G0/G1-phase (Figure 6—figure supplement 2A). Finally, we assessed the subcellular localization of p130 at each time point by quantifying the ratio of nuclear:cytoplasmic intensity from immunofluorescence images. We found p130(AA) was more localized to the nucleus at days 5 and 7 compared to p130(WT) (Figure 6G and Figure 6—figure supplement 2B), consistent with p130(AA) mediating cell cycle arrest through cell cycle gene repression.

As a component of the DREAM complex, p130 functions to prevent the transcription of myriad early (S-phase) and late (G2/M) cell cycle genes (Engeland, 2018). Because we observe an increase in G0/G1% for populations of cells expressing p130(AA), we sought to determine the expression levels of several DREAM target genes including: CCNF, CCNE1, CDC6, DHFR, E2F1, and CDT1. There was a significant reduction in expression of all of these in p130(AA)-expressing cells compared to those expressing p130(WT) (Figure 7A). Since both p130(WT)-expressing cells and p130(AA)-expressing cells grew more slowly than cells not expressing a p130 transgene, we also performed RT-qPCR on control cells; the expression of CCNF, CCNE1, CDC6, DHFR, E2F1, and CDT1 was lower in both p130(WT)- and p130 (AA)-expressing cells compared to controls (Figure 7—figure supplement 1A). Thus, ectopic expression of p130 reduces transcription of DREAM target genes, which is further decreased by the expression of p130(AA). To confirm that the reduction in gene expression of DREAM targets led to decreased protein expression, we used 4i to quantify two DREAM targets by immunofluorescence: CDC6 and Cdt1. Consistent with DREAM activation, as well as observed G0/G1 arrest, protein levels of both CDC6 and Cdt1 were decreased from days 5 to 10 in p130(AA)-expressing cells compared to p130(WT)-expressing cells (Figure 7B–C).

Figure 7 with 2 supplements see all
Download asset Open asset
DREAM targets are downregulated in cells expressing p130(AA).

(A) Inducible p130(WT) and p130(AA) NHF-1 cells were grown for 8 days in media-containing 100 ng/ml doxycycline to induce p130 expression. RNA was extracted for rt-qPCR analysis. Gene expression is relative to GAPDH expression and normalized to the vehicle control. Data are mean of n=3 experiments, and error bars are SEM. (B, C) Inducible p130 NHF-1 cells were grown in 100 ng/ml doxycycline for indicated times, and protein levels of CDC6 (B) and Cdt1 (C) were analyzed by iterative immunofluorescent staining and fixed cell imaging. Distributions of single-cell measurements of Cytoplasmic CDC6 (B) and nuclear CDT1 (C) were plotted (left) and representative images are shown for the indicated days (right).

Because not all cells arrest in G0/G1, we asked whether another factor may also be contributing to the reduced proliferation of the p130(AA)-expressing population. High p130 levels have previously been linked to apoptosis (Pentimalli et al., 2018; Ventura et al., 2018), whereas low p130 levels have been shown to protect against apoptosis (Bellan et al., 2002). We therefore examined the apoptotic markers cleaved PARP and cleaved caspase 3. As a positive control, cell lines were treated with 100 nM staurosporine to induce apoptosis. We observe cleaved PARP in the p130(AA)-expressing population at days 7, 10, and 14, and cleaved caspase 3 at days 7 and 10, indicating that cells in these populations are undergoing apoptosis. We did not observe cleaved PARP or cleaved caspase in any of the populations for p130(WT)-expressing cells (Figure 7—figure supplement 2A). Additionally, we performed flow cytometry for Annexin V and propidium iodide incorporation in unfixed cells to assess apoptosis/necrosis. We observed a higher percentage of apoptotic/necrotic cells in the p130(AA)-expressing population compared to the p130(WT)-expressing population at day 8 (Figure 7—figure supplement 2B). Since DNA damage can cause apoptosis, and because high p21 levels can be an indicator of DNA damage, we used 4i to determine the expression of proteins in the DNA damage response pathway, including 53BP1, phospho-Chk1, and phospho-H2A.X. We found that in the p130(AA)-expressing cells at days 5, 7, and 10, that levels of 53BP1, phospho-Chk1, and phospho-H2A.X were downregulated (Figure 7—figure supplement 2C-E). A decrease in DNA damage markers is consistent with fewer cells being in S-phase and suggests that DNA damage is not the cause of apoptosis. We also examined senescence but saw no apparent increase in beta-galactosidase staining in p130(AA)-expressing cells.

Discussion

We demonstrate that the RB-like protein p130 is a substrate of the E3 ubiquitin ligase SCFCyclin F. Cyclin F loss allows p130 to accumulate, whereas its overexpression promotes p130 degradation. Cyclin F and p130 bind directly, and SCFcyclin F can ubiquitinate p130 in a fully reconstituted in vitro system. We mapped a critical cyclin F binding motif in p130, and when that sequence is mutated, it prevents p130 ubiquitination and degradation. This mutant version of p130 accumulates in normal human fibroblast cells, causing slowed proliferation, an accumulation in G0/G1-phase, and apoptosis.

Taken together, our results suggest that cyclin F contributes to inactivation of p130 and is therefore a regulator of the CDK-RB network. This finding is consistent with our observation that CCNF KO is correlated with CDK4 and CCND1/Cyclin D1 KO (Figure 1—figure supplement 3). Interestingly, the DepMap has tested hundreds of cell lines for their sensitivity to over 1000 oncology and non-oncology drugs. We queried these data for gene-drug similarities related to the selective CDK4/6 inhibitors Palbociclib and Ribociclib, which are used to treat metastatic, hormone receptor positive breast cancer (Agostinetto et al., 2021; Kay et al., 2021). Comparing Palbociclib to CRISPR knockout screens, revealed CCND1/Cyclin D1 and CDK4 as the most highly correlated genes, validating this approach. Notably, CCNF ranked ninth among over 17,000 genes queried (Figure 8A). For Ribociclib, CCND1 is the most highly correlated, CDK4 is ranked 5th, and CCNF was 125th, falling within the top 1% of most highly correlated genes (Figure 8B). Conversely, we determined which drugs are highly correlated with CDK4, CCND1, and CCNF KO. Palbociclib and Ribociclib are among the top four most highly correlated drugs with CDK4 and CCND1 (Figure 8C–D). Remarkably, when CCNF/Cyclin F knockout is compared to all drugs tested, Palbociclib is ranked 1st, and Ribociclib is ranked 24th (Figure 8E). Thus, CCNF KO is highly correlated with both genetic and pharmacologic inactivation of the CDK-RB network across hundreds of cell lines.

Figure 8
Download asset Open asset
CDK4/6 inhibitors and CCNF knockout correlate highly.

Project Achilles Dependency Map data sets were analyzed to determine: (A) Pearson’s correlation coefficients for gene knockout correlated with Palbociclib treatment. (B) Pearson’s correlation coefficients for gene knockout correlated with Ribociclib treatment. (C) Pearson’s correlation coefficients for the correlation between CDK4 knockout and 1000 drug treatments. (D) Pearson’s correlation coefficients for the correlation between CCND1 knockout and 1000 drug treatments. (E) Pearson’s correlation coefficients for the correlation between CCNF knockout and 1000 drug treatments. For (A–E), only the top scoring, most statistically significant associations are shown.

RB-family protein dynamics throughout the cell cycle

The activity of RB and the RB-like proteins, p107 and p130, oscillates during the cell cycle and quiescence. Cyclin-CDK complexes phosphorylate RB, p107, and p130, which prevents their binding to activator E2Fs, allowing for the transcription of cell cycle target genes. In addition, RB-like proteins are also controlled at the level of their abundance. As cells exit quiescence, p130 protein is degraded by the ubiquitin-proteasome system. Then, when cells transition from proliferation to quiescence, p130 protein levels significantly increase (Smith et al., 1996). The role of phosphorylation in controlling RB and the RB-like proteins has been well-established (Canhoto et al., 2000; Farkas et al., 2002; Hansen et al., 2001); however, much less is known about how changes in the levels of RB-like proteins might contribute to changes in pathway activity or their ability to restrain E2F transcription factors. Notably, as cells grow bigger, RB and its yeast ortholog Whi5, are diluted, contributing to cell cycle progression (Schmoller et al., 2015; Zatulovskiy et al., 2020). Our studies suggest that the targeted degradation of p130 by cyclin F, which rapidly reduces its concentration, also plays a significant role in promoting cell cycle progression.

Despite a concrete understanding of the role of phosphorylation in regulating RB-like proteins, less is known about the role of ubiquitin ligases. The phosphatase subunit NRBE3 was previously linked to RB ubiquitination (Wang et al., 2015). A nucleolar protein, U3 protein 14a (hUTP14a), was also suggested to be a novel type of E3 ubiquitin ligase, capable of promoting both RB and p53 degradation and cancer cell proliferation (Liu et al., 2018). However, neither NRBE3 nor hUTP14a are ubiquitin ligases. Finally, RB degradation by an unknown, Cul2-based E3 was shown in cells expressing HPV E7 oncoprotein (White et al., 2012). Previous studies have reported that SCFSkp2 recognizes p130 and promotes its ubiquitination (Bhattacharya et al., 2003; Tedesco et al., 2002). However, p130 still cycles in Skp2 knockout cells, suggesting additional E3 ligase(s) may be involved in p130 regulation. Surprisingly, for unknown reasons, Skp2 overexpression had no effect on p130 levels in our experiments.

Cyclin F as a regulator of the cell cycle gene transcription

Cyclin F has previously been linked to cell cycle gene transcription via its regulation of several proteins, including B-MYB, SLBP, E2F1, E2F2, E2F3a, E2F7, and E2F8 (Emanuele et al., 2020). In response to DNA damage, cyclin F is able to bind to B-MYB, and rather than ubiquitinate it, inhibits the ability of B-MYB to promote the expression of mitotic genes (Klein et al., 2015). In G2, cyclin F ubiquitinates the stem-loop binding protein (SLBP), which is required during S-phase to mediate histone biogenesis (Dankert et al., 2016). Also, during G2, cyclin F catalyzes the ubiquitination of the activator E2Fs: E2F1, E2F2, and E2F3a (Burdova et al., 2019; Clijsters et al., 2019). The ubiquitination of activator E2Fs allows for their degradation, and expression of mutant E2Fs that can no longer bind to cyclin F results in increased expression of E2F target genes. Further, during G2-phase, cyclin F was shown to regulate the atypical repressor E2Fs, E2F7, and E2F8, promoting their degradation (Wasserman et al., 2020; Yuan et al., 2019). Each of these substrates implicates cyclin F in indirectly controlling cell cycle gene expression, and our study uncovers a new mechanism by which cyclin F regulates transcriptional control at the start of the cell cycle, by ubiquitinating p130.

Adding to the complexity of the role of cyclin F in cell cycle transcription, not only does cyclin F regulate p130 degradation, but p130 regulates the expression of cyclin F. Thus, cyclin F and p130 appear to exist in a double negative feedback loop, where each can downregulate the other. Interestingly, in response to DNA damage, cyclin F is downregulated by degradation (D’Angiolella et al., 2012), and p130 has been suggested to become activated to repress cell cycle gene expression. Thus, the degradation of cyclin F in response to genotoxic stress could preserve cell cycle arrest by allowing for the accumulation of p130, while also contributing to increased nucleotide pools through the accumulation of another cyclin F substrate, RRM2 (D’Angiolella et al., 2012).

Effect of P130 overexpression on proliferation

We have shown that expressing the mutant p130(AA), but not p130(WT), causes growth arrest and apoptosis. These results are consistent with clinical observations of outcomes of p130 expression levels in cancerous tissues. In lung cancers, p130 expression negatively correlates with histological grading and metastasis ( Baldi and Esposito, 1997). For endometrial cancer patients who receive surgery, low p130 levels are significantly associated with increased recurrence and death (Susini et al., 1998). In vulvar carcinomas, loss of p130 and p27 has been shown to contribute to carcinogenesis (Zamparelli et al., 2001). In oral squamous cell carcinoma, p130-negative cases have worse prognoses than p130-positive cases (Tanaka et al., 2001). And, finally, in thyroid neoplasms, reduced p130 expression has been linked to the aggressive characteristics of anaplastic carcinoma, while high p130 expression in micropapillary carcinoma has been linked to the smaller size of those tumors (Ito et al., 2003). Thus, cancers with low p130 expression or p130 loss, seem to grow more quickly and aggressively, whereas cancers with high p130 expression are correlated with better prognoses. Our findings fit within the context of these studies; as p130(AA) accumulates, the cells are less proliferative and some die by apoptosis. Since p130 mRNA levels are not regulated during the cell cycle, it is interesting to speculate that enhancing p130 degradation in some malignancies could contribute to defects in the regulation of cell cycle gene expression and hyper-proliferation. Likewise, blocking p130 degradation could phenocopy CDK4/6 inactivation and could potentially be therapeutically advantageous, and would be consistent with our analysis showing that loss-of-function in cyclin D, CDK4, and cyclin F are highly correlated across nearly 800 cancer cell lines.

Cyclin F is a strongly cell cycle-regulated gene. Cyclin F mRNA levels peak in G2/M, and no studies have identified CCNF as an E2F target at G1/S. It is therefore unclear how its protein levels increase early in the cell cycle, in lockstep with Skp2, a bona fide E2F target gene. We previously showed that AKT can phosphorylate and stabilize cyclin F, and that this could contribute to its accumulation near G1/S (Choudhury et al., 2017). Thus, the activation of cyclin F by AKT could potentially lead to the degradation of p130. Given the recurrent activation of the PI3K-AKT pathway in many cancers, it will be interesting in the future to determine if cyclin F phosphorylation promotes p130 degradation. If true, this would suggest that hyper-activated PI3K, either through mutation of PIK3CA or loss of PTEN, functions through cyclin F to render p130 inactive via the ubiquitin pathway.

Materials and methods

Acquisition and analysis of DEPMAP data

Request a detailed protocol

Gene co-dependencies were determined from the Achilles data set from depmap.org (Achilles_gene_effect.csv, downloaded 7/19/19). The Achilles data set contains dependency scores from genome-scale essentiality screens scores of 789 cell lines. As a measure of co-dependency, the Pearson’s correlation coefficient of essentiality scores was computed for all gene pairs. GO analysis for the top 100 genes co-dependent with CCNF was performed using MetaScape. Potential cyclin F substrates were identified by proteins encoded by genes that have a Pearson’s correlation coefficient of >0.15 when compared to CCNF and are classified by the GO term: cell division (0051301).

Cell lines

Request a detailed protocol

Cell lines used include: HEK293T (ATCC), U2OS (ATCC), HeLa (ATCC), HeLa sgCTRL, and HeLa sgCCNF (Choudhury et al., 2016), MCF-7 (ATCC), T47D (ATCC), NHF-1 William Kaufman Lab (UNC; retired), IMR-90 (Yue Xiong Lab, UNC; retired), and T98G (Tissue Culture Facility, UNC). Additional cell line details are listed in the table in Appendix 1. Cell line authenticity was verified by SPR. All cell lines were checked for mycoplasma contamination periodically throughout the experimental process.

Cell culture

Request a detailed protocol

All cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM; GIBCO) supplemented with 10% fetal bovine serum (FBS; VWR) and 1% Pen/Strep (GIBCO). Cells were incubated at 37°C, 5% CO2.

DNA transfection experiments were performed in HEK293T cells for 24 hr using either Lipofectamine 2000 (Thermo Fisher Scientific) or PolyJet (SignaGen) transfection reagents according to the manufacturer’s protocol. RNA transfection experiments were performed in NHF-1 and IMR-90 cells for 48 hr using RNAiMAX (Thermo Fisher Scientific) transfection reagent according to the manufacturer’s protocol. Plasmid and siRNA information is the table in Appendix 1.

To produce lentivirus, HEK293T cells were transfected with pLV[Exp]-CMV> Tet3G/Hygro, pLV[TetOn]-Neo-TRE3G > HA/{p130 WT}, or pLV[TetOn]-Neo-TRE3G > HA/{p130 AA}, and the lentivirus packaging plasmids VSV-G, Gag-pol, Tat, and Rev. Media-containing virus particles were collected after 48 hr, filtered with a 0.45-µM filter, and frozen at –80°C. NHF-1 cells were transduced with a 1:1 ratio of fresh media plus lentivirus-containing media. NHF-1 cells were first transduced with the Tet3G plasmid and selected in media-containing 50 µg/ml hygromycin. Then, NHF-1 cells stably expressing Tet3G were infected with either p130 WT or p130 AA and selected in media-containing 375 µg/ml G418.

Cloning and Directed Mutagenesis pDEST-HA3-p130, pDEST-HA3-p1301–417, pDEST-HA3-p130418–1139, pDEST-HA3-p130418–616, pDEST-HA3-p130418–827, pDEST-HA3-p130418–1024, pDEST-HA3-p130828–1139, and pINDUCER20 CCNF were produced using Gateway Cloning Technology. pLV[Exp]-CMV> Tet3G/Hygro, pLV[TetOn]-Neo-TRE3G > HA/{p130 WT}, or pLV[TetOn]-Neo-TRE3G > HA/{p130 AA} were all ordered from VectorBuilder. pGEX-GST-p130593–790 was a gift from Peter Whyte (McMaster University) and pINDUCER20 was a gift from Stephen Elledge (Addgene #44012). pDEST-HA3-p130 R658A I660A, pDEST-HA3-p130 R680A L682A, and pGEX-GST-p130593–790 R680A L682A were generated by site-directed mutagenesis using the Q5 Site-Directed Mutagenesis Kit (NEB) according to the manufacturer’s instructions. For recombinant expression, Cyclin F25–546 was subcloned into a pFastBac vector with an N-terminal GST tag and TEV protease site. The Skp1 gene (a gift from Dr. Bing Hao) was subcloned into a PGEX-4T-1 vector previously engineered to contain a TEV protease site. Sequences for all primers are included in tables provided in Appendix 1.

Cell lysis and immunoblotting

Request a detailed protocol

Cells were lysed on ice for 10 min in NETN (20 mM Tris pH 8.0, 100 mM NaCl, 0.5 mM EDTA, 0.5% NP40) supplemented with 10 µg/ml aprotinin, 10 µg/ml leupeptin, 10 µg/ml pepstatin A, 1 mM sodium orthovanadate, and 1 mM AEBSF(4-[two aminoethyl] benzenesulfonyl fluoride). Following incubation, cells were spun at 20,000×g in a benchtop microcentrifuge at 4°C for 10 min. Protein concentration was determined by Bradford assay (Bio-Rad) and samples were prepared by boiling in Laemmli buffer (LB). Protein was separated by electrophoresis on either homemade or TGX (Bio-Rad) stain-free gels and then was transferred to nitrocellulose membranes. Blocking was performed in 5% non-fat dry milk (Blotting Grade blocker; Bio-Rad) diluted in TBS-T (137 mM NaCl, 2.7 mM KCl, 25 mM Tris pH 7.4, 1 % Tween-20). All primary antibody incubations were carried out overnight, rocking, at 4°C, and all HRP-conjugated antibody incubations were carried out for 1 hr, rocking, at room temperature. TBS-T was used for all wash steps. Protein abundance was visualized by chemiluminescence using Pierce ECL (Thermo Fisher Scientific). A detailed list of antibodies and dilutions is provided in table in Appendix 1.

Immunoprecipitation

Endogenous protein pulldowns:

Request a detailed protocol

Asynchronous HEK293T, U2OS, HeLa, or inducible p130 NHF-1 cells were lysed and total protein was quantified as described above. 10% of the protein sample was retained as the input, and the remaining protein was divided in half to be incubated with either rabbit IgG control or Cyclin F antibodies for IP. Cell lysates were incubated for 6 hr, rotating, at 4°C with 1 µg antibody per mg protein. During the incubation, Pierce Protein A/G agarose beads (Thermo Fisher Scientific) were washed 3× for 20 min in NETN, then resuspended at a 50% slurry in NETN. 40 µl slurry per 1 mg protein was added to the protein/antibody IP mix and incubated for 45 min, rotating, at 4°C. Beads were then washed 3× for 5 min in NETN, and finally re-suspended in 2× LB and boiled. Co-IP was assessed by immunoblotting as described.

Exogenous pulldowns:

Request a detailed protocol

Asynchronously proliferating inducible p130 NHF-1 cells were treated with 100 ng/ml doxycycline for 24 hr. Cells were lysed and total protein was quantified as above. 10% of the protein sample was retained as the input, and the remaining protein was divided in half to be incubated with either rabbit IgG control or rabbit anti-HA antibodies for IP. IP and washes were carried out as described above. Alternatively, inducible p130 NHF-1 cells were treated with 100 ng/ml doxycycline or water as a vehicle control for 24 hr. Cells were lysed and total protein was quantified as described above. 50 µl of EZView Red anti-HA affinity gel slurry was used to isolate HA-tagged and co-immunoprecipitating proteins. Washes were performed as described above. Co-IP was assessed by immunoblotting.

Protein purification

Request a detailed protocol

In order to purify GST and GST-p130 for the in vitro binding assay, BL21 E. coli were transformed with GST, GST-p130WT or GST-p130AA and a 0.5-L culture was grown to an O.D. of 0.6 then induced with 1 mM IPTG (Isopropyl β-D-1-thiogalactopyranoside) for 4 hr. Cells were harvested by centrifugation and incubated for 20 min in NTEN E. coli Lysis buffer (100 mM NaCl2, 20 mM Tris pH 8.0, 1 mM DTT, 1 mM EDTA, and 1% NP40) supplemented with 10 µg/ml aprotinin, 10 µg/ml leupeptin, 10 µg/ml pepstatin A, and 10 mg/ml lysozyme. Then, cells were lysed by sonication at 50% power for 3× for 30 s pulses, with 1 min on ice between pulses. Lysates were then spun at 30,000×g for 30 min at 4°C to clarify. During the clarification step, glutathione agarose resin (GoldBio) was equilibrated 3× for 10 min in NTEN E. coli Lysis buffer. 250 µl of equilibrated glutathione agarose resin was then added to the clarified lysate and incubated for 2 hr, rotating, at 4°C. Beads were washed 3× for 5 min in NTEN E. coli Lysis buffer. Bound GST-p130 was eluted by three sequential elution steps: 2× for 30 min and one overnight incubation, rotating, at 4°C in elution buffer (50 mM Tris pH 8.0, 5 mM reduced glutathione). Elution fractions were pooled, and buffer exchange into 50 mM Tris was performed by 5× buffer exchange in 30,000 MWCO spin columns.

For in vitro ubiquitination assays, UBA1, CDC34B, UBCH7, ARIH1, Cul1-Roc1, and ubiquitin were expressed and purified as described previously (Kamadurai et al., 2013; Scott et al., 2016). The neddylation of Cul1-Roc1 was also performed as described previously (Duda et al., 2008). GST-p130wt (residues 593–790) and GST-p130AA (residues 593–790 R680A L682A) fusions were expressed in E. coli BL21-CodonPlus (DE3)-RIL, and then purified by glutathione-affinity and size-exclusion chromatography.

For Cyclin F25-546, 1.2 L of Sf9 cells were infected with baculovirus at a density of 2×106 cells/ml of culture and harvested after 3 days. Pelleted cells were resuspended in lysis buffer (25 mM Tris, 200 mM NaCl, 1 mM DTT, 1 mM PMSF, 1× protease inhibitor cocktail, pH 8.0) and passed through a C3 emulsiflex homogenizer (AVESTIN, Inc). After lysate clarification, the supernatant was loaded onto a GS4B glutathione agarose (Cytvia) column. The column was washed in lysate buffer lacking protease inhibitors and eluted in the same buffer with 10 mM glutathione. Purified TEV protease (1% by mass) was added overnight while protein was dialyzed in buffer containing 25 mM Tris, 200 mM NaCl, and 1 mM DTT (pH 8.0). Protein was then passed back through a GS4B column to remove free GST, concentrated, and loaded in 1 mL onto a Superdex 200 column (Cytvia) equilibrated in similar buffer. Peak fractions from the column elution were pooled and protein concentrated to 8 μM.

Full-length Skp1 was expressed in E. coli as a GST fusion protein. 6 L of BL21(DE3) cells were grown to OD of 0.6 then induced with 1 mM IPTG (Isopropyl β-D-1-thiogalactopyranoside) for 16 hr at room temperature. Protein was purified similar to as described above for cyclin F25–546, except that following the initial GS4B elution, protein was further purified by Source Q Sepharose ion-exchange chromatography prior to TEV cleavage. For the fluorescence polarization anisotropy assays, the GST tag was left on the cyclin F for stability, and the fusion protein was mixed with Skp1 prior to the experiment.

P130-cyclin F in vitro binding assays

Request a detailed protocol

GST pulldowns were performed by loading 1 µg of purified GST or GST-p130 onto 15 µl of glutathione beads (GoldBio) in NETN (20 mM Tris pH8.0, 100 mM NaCl, 0.5 mM EDTA, 0.5% NP40) for 1 hr, rotating at 4°C. Loaded beads were washed 3× for 5 min in NETN, then incubated with 0.5 mg of whole-cell extract of HEK293T cells transiently transfected with a FLAG-Cyclin F plasmid for 2 hr, rotating, at 4°C. Beads were again washed 3× for 5 min in NETN and p130-cyclin F interaction was assessed by immunoblot.

FLAG pulldowns were performed by immunoprecipitating FLAG-Cyclin F from whole-cell extracts of HEK293T cells transiently transfected with FLAG-Cyclin F. IP was performed by incubating 1 mg whole-cell extract with 50 µl EZView Red anti-FLAG M2 affinity gel (MilliporeSigma) for 2 hr, rotating, at 4°C. Loaded affinity gel was washed 3× for 5 min with NETN, then incubated with 1 µg purified GST or GST-p130 for 2 hr, rotating, at 4°C. Loaded affinity gel was again washed 3× for 5 min in NETN and p130-cyclin F interaction was assessed by immunoblot.

In vitro ubiquitination assay

Request a detailed protocol

Multiple turnover reactions were set up by combining 0.1 µM UBA1, 5 mM MgATP, 0.8 µM CDC34B, 0.3 µM ARIH1, 0.8 µM UBCH7, 0.4 µM NEDD8 to Cul1-Roc1,0.4 µM SKP1, 0.4 µM Cyclin F25–546, and 0.6 µM of either GST-p130 or GST-p130AA in the assay buffer (20 mM HEPES pH = 8, 200 mM NaCl) while kept on ice. The mixtures were then equilibrated to room temperature and the reaction was started by the addition of 100 µM of ubiquitin. Reactions were quenched by adding SDS loading buffer at specified time points. After SDS-PAGE, the ubiquitination of GST-p130 was monitored by western blot analysis using α-GST antibody (SC-138, mouse) and a fluorescent secondary antibody (goat anti-mouse IgG (H + L), Alexa Fluor 633, A-21052). Fluorescence was observed using the Amersham Typhoon 5.

De-phosphorylation assay with calf intestinal phosphatase

Request a detailed protocol

Cells were lysed as described above in NETN lacking EDTA. 20 µg total protein lysates were incubated at 37°C for 1 hr with 1× NEB 2.1 buffer (NEB) with 1 unit of calf intestinal phosphatase (NEB) or buffer as a negative control. As a positive control, 1 mM sodium orthovanadate was added to a reaction containing calf intestinal phosphatase for the duration of the 1-hr incubation. Phosphorylation was assessed by immunoblotting.

Proliferation assays

Cell counting:

Request a detailed protocol

NHF-1 inducible p130 cells were seeded in 10 cm plates, and p130 WT or AA was induced by the addition of 100 ng/ml doxycycline (or DMSO as a control) to media. Cells were trypsinized (Gibco), counted with an automatic cell counter (Bio-Rad), and replated in fresh doxycycline-containing media every 48 hr for 2 weeks. Cell count was plotted versus time, differences were assessed using a Student’s t-test, and cell count was used to calculate doubling times. Experiments were carried out in triplicate and repeated three times.

PrestoBlue:

Request a detailed protocol

NHF-1 inducible p130 cells were seeded in 96-well plates, and p130 WT or AA was induced by the addition of 100 ng/ml doxycycline to media. Media were replaced with fresh doxycycline-containing media every 48 hr for 2 weeks, and growth was assessed with the PrestoBlue cell viability reagent (Thermo Fisher Scientific) on days 1, 7, and 14. Experiments were carried out with six technical replicates, and repeated three times.

Crystal violet assay:

Request a detailed protocol

NHF-1 inducible p130 cells were seeded in 60 mm plates, and p130 WT or AA was induced by the addition of 100 ng/ml doxycycline (or DMSO as a control). Media were replaced with fresh doxycycline-containing media (or DMSO-containing media as a control) every 48 hr for 10 days. On day 10, confluence was visualized by staining with Crystal Violet Staining Buffer (0.5% crystal violet, 20% methanol) for 15 min at room temperature, and de-staining with water. Experiments were repeated three times.

RT-qPCR

Request a detailed protocol

In NHF-1 inducible p130 cells, p130 WT or AA was induced by the addition of 100 ng/ml doxycycline (or DMSO as a control) for 14 days. Cells were harvested and RNA was extracted using the RNeasy Plus Mini Kit (QIAGEN). 1 µg of extracted RNA was used to generate cDNA libraries using the SuperScript III First-Strand synthesis system (Thermo Fisher Scientific) following the manufacturer’s instructions. Samples were diluted 1:10, except for GAPDH analysis that were diluted 1:1000. Transcript abundance was quantified using the SSO Advanced Universal SYBR Green Supermix (Bio-Rad) and measured with a QuantStudio 6 Flex Real-Time PCR System (Thermo Fisher Scientific), and transcript levels were normalized to GAPDH. Relative quantity of transcripts was quantified using the 2-∆∆CT method. Each sample was run in triplicate. Primers are listed in tables provided in Appendix 1.

Flow cytometry

Request a detailed protocol

In NHF-1 inducible p130 cells, p130 WT or AA was induced by the addition of 100 ng/ml doxycycline (or DMSO as a control) for the indicated number of days. For cell cycle analysis: 30 min prior to fixation, cells were pulsed with 10 µM EdU and were fixed in 4% formaldehyde/PBS for 15 min at room temperature. Cells were pelleted and re-suspended in 1% BSA/PBS. EdU was labeled with Alexa Fluor 488 using click chemistry as previously described (Franks et al., 2020). For Annexin V apoptosis/necrosis: the dead cell apoptosis kit was used following manufacturer instructions (Thermo Fisher Scientific). Flow was carried out on an Attune NxT Flow Cytometer (Thermo Fisher Scientific), and data were analyzed using FlowJo software.

Florescence polarization anisotropy assay

Request a detailed protocol

For the GST-cyclin F25–546-Skp1 titration, TAMRA-labeled synthetic peptide (E2F184–99 or p130674–692) at 10 nM was mixed with varying concentrations of protein complex in a buffer containing 25 mM Tris, 150 mM NaCl, 1 mM DTT, 0.1% (v/v) Tween-20, pH 8.0. For the Ki measurements, varying concentrations of GST-p130593–790 WT, GST-p130593–790 R680A/L682A mutant, or synthetic unlabeled E2F184–99 or p130674–692 peptide were mixed with 10 nM TAMRA-labeled E2F184–99 and 0.5 μM GST-cyclin F-Skp1 in a buffer containing 25 mM Tris, 150 mM NaCl, 1 mM DTT, 0.1% (v/v) Tween-20, pH 8.0. 40 µl of the reaction were used for the measurement in a 384-well plate. Fluorescence anisotropy (FA) measurements were made in triplicate, using a PerkinElmer EnVision plate reader. The KD and Ki values were calculated using Prism 8 (Version 8.4.3).

Iterative indirect immunofluorescence imaging (4i)

Request a detailed protocol

Cells were plated in glass‐bottom plates (Cellvis) treated as required and prepared as follows. In between each step, samples were rinsed 3× with phosphate-buffered saline (PBS) and incubations were at room temperature, unless otherwise stated. Cells were fixed with 4% paraformaldehyde (Thermo Fisher Scientific, 28908) for 30 min, permeabilized with 0.1% Triton X-100 in PBS for 15 min and inspected for sample quality control following Hoechst staining (1:1000, MilliporeSigma, 99403) in imaging buffer (IB: 700 mM N-acetyl-cysteine (Sigma-Aldrich, A7250) in ddH2O. Adjust to pH 7.4). Cells were rinsed 3× with ddH2O and incubated with elution buffer (EB: 0.5 M L-Glycine (Sigma-Aldrich, 50046)), 3 M Urea (Sigma-Aldrich, U4883), 3 M Guanidine chloride (Thermo Fisher Scientific, 15502-016), and 70 mM TCEP-HCl (Sigma-Aldrich, 646547) in ddH2O. Adjusted to (pH 2.5) 3× for 10 min on shaker to remove Hoechst stain. Sample was incubated with 4i blocking solution (sBS: 100 mM maleimide [Sigma-Aldrich, 129585], 100 mM NH4Cl [Sigma-Aldrich, A9434], and 1% bovine serum albumin in PBS) for 1 hr and incubated with primary antibodies diluted as required (anti-phospho-RB [S807/S811] [1:1000, Cell Signaling Technology, 8516]; anti-RB [1:500, Cell Signaling Technology, 9309]; anti-p21 [1:200, R&D Systems, AF1047]; anti-p130 [1:100, Cell Signaling Technology, 13610]; anti-phospho-H2A.X [Ser139] [1:200, Cell Signaling Technology, 80312], anti-phospho-CHK1 [S317] [1:800, Cell Signaling Technology, 12302], anti-53BP1 [1:250, Abcam, ab36823], CDT1 [1:200, Cell Signaling Technology, 8064], CDC6 [1:100, Santa Cruz, sc-9964], HA [1:200, BioLegend, 901501]) in conventional blocking solution (cBS: 1% bovine serum albumin in PBS) overnight at 4°C. Samples were rinsed 3× with PBS and then incubated in secondary antibodies (1:500, donkey anti-rabbit AlexaFluor Plus 488 [Thermo Fisher Scientific, A32790], donkey anti-mouse AlexaFluor Plus 555 [Thermo Fisher Scientific, A32773], and donkey anti-goat AlexaFluor Plus 647 [Thermo Fisher Scientific, A32758]) and Hoechst for 1 hr on shaker, then rinsed 5× with PBS and imaged in IB. Samples were imaged using the Nikon Ti Eclipse inverted microscope with a Nikon Plan Apochromat Lambda 40× objective with a numerical aperture of 0.95 and an Andor Zyla 4.2 sCMOS detector. Stitched 8×8 images were acquired for each condition using the following filter cubes (Chroma): DAPI(383-408/425/435-485nm), GFP(450-490/495/500-550nm), Cy3(530-560/570/573-648nm), and Cy5(590-650/660/663-738nm). After imaging, samples were rinsed 3× with ddH2O, antibodies were eluted, and re-stained iteratively as described above.

4i image analysis

Request a detailed protocol

Nuclear segmentation and quantification were performed using standard modules in CellProfiler (v3.1.8) as described below. For each round of immunofluorescence images obtained by 4i, individual nuclei were automatically detected and segmented using the IdentifyPrimaryObjects module. Cytoplasmic segmentation was performed by making a 5-pixel ring outside the nucleus using the ExpandOrShrinkObjects and IdentifyTertiaryObjects modules. Nuclear and cytoplasmic intensities were quantified using the MeasureObjectIntensity module. Distributions of single-cell intensities were visualized using the seaborn library (v0.11.0) in Python (v3.7.1).

Apoptosis assays

Request a detailed protocol

Cells were grown for 0–14 days in 100 ng/ml doxycycline or water as a vehicle control. As a positive control for apoptosis. As a positive control, cells were treated with 100 nM staurosporine for 6 hr to induce apoptosis. Presence of cleaved PARP and cleaved Caspase three was assessed by immunoblot. And, apoptosis/necrosis were assessed using an Annexin V/propidium iodide staining kit (Thermo Fisher Scientific) coupled with flow cytometry.

Quantification and statistical analysis

Request a detailed protocol

Western blot quantification was carried out using the Fiji software. All statistical analyses were carried out using GraphPad Prism v9.

Appendix 1

Key reagents table

Lists key reagents used in this study.

Appendix 1—key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Gene (Homo sapiens)CCNFGenBankGene ID: 899
Gene (H. sapiens)RBL2GenBankGene ID: 5934
AntibodyCDK4 (rabbit monoclonal) antibodyCSTCat. #: D9GE3; RRID:AB_2799229IB (1:1000)
AntibodyCDK6 (rabbit monoclonal) antibodyCSTCat. #:D4S8S; RRID:AB_2721897IB (1:1000)
AntibodyCyclin A1 (rabbit monoclonal) antibodyAbcamCat. #: ab53699; RRID:AB_879763IB (1:1000)
AntibodyCyclin A1 (mouse monoclonal) antibodySanta CruzCat. #: SC-751; RRID:AB_631329IB (1:5000); IP (1 µg)
AntibodyCyclin D1 (rabbit monoclonal) antibodyCSTCat. #: 2978; RRID:AB_2259616IB (1:1000)
AntibodyCyclin E1 (mouse monoclonal) antibodyCSTCat. #: 4129; RRID:AB_2071200IB (1:5000); IP (1 µg)
AntibodyCyclin E1 (rabbit monoclonal) antibodyCSTCat. #: 20808; RRID:AB_2783554IB (1:1000)
AntibodyCyclin F (rabbit polyclonal) antibodySanta CruzCat. #: SC-952; RRID:AB_2071212IB (1:5000)
AntibodyFLAG (HRP-conjugated mouse monoclonal) antibodySigma-AldrichCat. #: A8592; RRID:AB_439702IB (1:10,000)
AntibodyGAPDH (mouse monoclonal) antibodySanta CruzCat. #: sc-47724; RRID:AB_627678IB (1:5000)
AntibodyGST (HRP-conjugated mouse monoclonal) antibodyGeneTexCat. #: GTX114099; RRID:AB_1949436IB (1:5000)
AntibodyHA (mouse monoclonal) antibodyCovance/BioLegendCovance Cat. #: MMS-101P; RRID:AB_2314672IB (1:2000)
AntibodyLin54 (rabbit polyclonal) antibodyBethylCat. #: A303-799A; RRID:AB_11218173IB (1:1000)
AntibodyNormal rabbit IgG (polyclonal) antibodyProteinTechCat. #: 30000-0-AP; RRID:AB_2819035IP (1 µg)
Antibodyp107 (rabbit monoclonal) antibodyCSTCat. #: D3P3C; RRID:AB_2800144IB (1:1000)
Antibodyp130 (rabbit monoclonal) antibodyCSTCat. #: D9T7M; RRID:AB_2798274IB (1:1000)
Antibodyp130 pS672 (rabbit monoclonal) antibodyAbcamCat. #: ab76255; RRID:AB_2284799IB (1:5000)
Antibodyp27 (rabbit monoclonal) antibodyCSTCat. #: 2552; RRID:AB_10693314IB (1:1000)
AntibodySkp2 (rabbit monoclonal) antibodyCSTCat. #: 2652; RRID:AB_11178941IB (1:5000)
AntibodyTubulin (mouse monoclonal) antibodySanta CruzCat. #: 32293; RRID:AB_628412IB (1:5000)
AntibodyGoat anti-mouse IgG HRP-conjugated (goat polyclonal) antibodyJackson ImmunoResearchCat. #: 115-035-003; RRID:AB_10015289IB (1:5000)
AntibodyGoat anti-rabbit IgG HRP-conjugated (goat polyclonal) antibodyJackson ImmunoResearchCat. #: 111-035-003; RRID:AB_2313567IB (1:5000)
AntibodyPhospho-RB(S807/S811) (rabbit monoclonal) antibodyCSTCat. #: 8516; RRID:AB_111786584i (1:1000)
AntibodyRB (mouse monoclonal) antibodyCSTCat. #: 9309; RRID:AB_8236294i (1:500)
Antibodyp21 (goat polyclonal) antibodyR&D SystemsCat. #: AF1047; RRID:AB_22447044i (1:200)
Antibodyp130 (rabbit monoclonal) antibodyCSTCat. #: 13610; RRID:AB_27982744i (1:100)
AntibodyPhospho-H2A.X(ser139) (mouse monoclonal) antibodyCSTCat. #: 80312; RRID:AB_27999494i (1:200)
AntibodyAnti-phospho-Chk1 (rabbit monoclonal) antibodyCSTCat. #: 12302; RRID:AB_27838654i (1:800)
Antibody53 BP1 (rabbit polyclonal) antibodyAbcamCat. #: ab36823; RRID:AB_7224974i (1:250)
AntibodyCDT1 (rabbit monoclonal) antibodyCSTCat. #: 8064; RRID:AB_108968514i (1:200)
AntibodyCDC6 (mouse monoclonal) antibodySanta CruzCat. #: sc-9964; RRID:AB_6272364i (1:100)
AntibodyDonkey anti-rabbit AlexaFluor Plus 488 (Donkey polyclonal) antibodyThermo Fisher ScientificCat. #: A32790; RRID:AB_27628334i (1:500)
AntibodyDonkey anti-goat AlexaFluor plus 647 (Donkey polyclonal) antibodyThermo Fisher ScientificCat. #: A32758; RRID:AB_27628284i (1:500)
AntibodyDonkey anti-mouse AlexaFluor Plus 555 (Donkey polyclonal) antibodyThermo Fisher ScientificCat. #: A32773; RRID:AB_27628484i (1:500)
Strain, strain background (Escherichia coli)BL21 Competent Escherishia coliNEBCat. #: C2530H
Strain, strain background (E. coli)DH5-ɑ Competent Escherishia coliNEBCat. #: C2988J
Commercial Assay or KitGeneJET Plasmid Miniprep KitThermo Fisher ScientificCat. #: K0503
Commercial Assay or KitQ5 Site-Directed Mutagenesis Kit (without competent cells)NEBCat. #: E0552S
Commercial Assay or KitRneasy Plus Mini KitQIAGENCat. #: 74134
Commercial Assay or KitSuperScript III First-Strand Synthesis SystemThermo Fisher ScientificCat. #: 18080051
Commercial Assay or KitDead Cell Apoptosis Kit with Annexin V FITC and PI for flow cytometryThermo Fisher ScientificCat. #: V13242
Chemical compound, drugSSO Advanced Universal SYBR Green SupermixBio-RadCat. #: 1725271
Chemical compound, drugBio-Rad Protein Assay Dye Reagent ConcentrateBio-RadCat. #: 5000006
Chemical compound, drugPrestoBlue Cell Viability ReagentThermo Fisher ScientificCat. #: A13261
Chemical compound, drugCalf Intestinal Phosphatase (CIP)NEBCat. #: M0290
Chemical compound, drugClarity ECL Western Blot SubstrateBio-RadCat. #: 1705060
Chemical compound, drugCyclohexamideMilliporeSigmaCat. #: 1810
Chemical compound, drugMG132Selleck ChemicalsCat. #: S2619
Chemical compound, drugMLN4924Active BiochemCat. #: A-1139
Chemical compound, drugGlutathione Agarose ResinGoldBioCat. #: G-250-5
Chemical compound, drugEZView Red Anti-FLAG M2 Affinity GelMilliporeSigmaCat. #: F2426
Chemical compound, drugEZView Red Anti-HA Affinity GelMilliporeSigmaCat. #: E6779
Chemical compound, drugPierceTM Protein A/G agarose BeadsThermo Fisher ScientificCat. #: 20421
Chemical compound, drugRnase AMilliporeSigmaCat. #: R6513
Chemical compound, drugLipofectamine RNAiMAXThermo Fisher ScientificCat. #: 13778150
Chemical compound, drugPolyJet Transfection ReagentSignaGenCat. #: SL100688
Chemical compound, drugLipofectamine 2000Thermo Fisher ScientificCat. #: 11668019
Cell line (H. sapiens)293TATCCCat. #: CRL-3216, RRID:CVCL_0063
Cell line (H. sapiens)U2OSATCCCat. #: HTB-96; RRID:CVCL_0042
Cell line (H. sapiens)HeLaATCCCat. #: CCL-2; RRID:CVCL_0030
Cell line (H. sapiens)HeLa sgCTRLPMID: 27653696
Cell line (H. sapiens)HeLa sgCCNFPMID: 27653696
Cell line (H. sapiens)MCF7 pIND CCNFThis paperSee Figure 1, Figure 1—figure supplement 3
Cell line (H. sapiens)T47D pIND CCNFThis paperSee Figure 1, Figure 1—figure supplement 3
Cell line (H. sapiens)NHF-1William Kaufman Lab (UNC; retired)
Cell line (H. sapiens)IMR-90Yue Xiong Lab (UNC; retired)
Cell line (H. sapiens)T98GTissue Culture Facility, UNC
Cell line (H. sapiens)NHF-1 doxy-inducible p130 WTThis paperSee Figure 6, Figure 7, Figure 5—figure supplements 1 and 26,7
Cell line (H. sapiens)NHF-1 doxy-inducible p130 AAThis paperSee Figure 6, Figure 5—figure supplements 1 and 2
Transfected Construct (H. sapiens)pBABE-p130Gift from Larisa Litovchick lab (VCU)For lentiviral transfection
Transfected Construct (H. sapiens)pDEST-HA3-p130This paperTransfected construct (human); See Figures 3—5, Figure 4—figure supplement 1
Transfected Construct (H. sapiens)pDEST-HA3-p130 1–417This paperTransfected construct (human); See Figure 4—figure supplement 1
Transfected Construct (H. sapiens)pDEST-HA3-p130 418–1139This paperTransfected construct (human); See Figure 4—figure supplement 1
Transfected Construct (H. sapiens)pDEST-HA3-p130 418-616This paperTransfected construct (human); See Figure 4—figure supplement 1
Transfected Construct (H. sapiens)pDEST-HA3-p130 418-827This paperTransfected construct (human); See Figure 4—figure supplement 1
Transfected Construct (H. sapiens)pDEST-HA3-p130 418-1024This paperTransfected construct (human); See Figure 4—figure supplement 1
Transfected Construct (H. sapiens)pDEST-HA3-p130 828-1139This paperTransfected construct (human); See Figure 4—figure supplement 1
Transfected Construct (H. sapiens)pDEST-HA3-p130 R658A I660AThis paperTransfected construct (human); See Figure 4—figure supplement 1
Transfected Construct (H. sapiens)pDEST-HA3-p130 R680A L682AThis paperTransfected construct (human); See Figure 4—figure supplement 1
Transfected Construct (H. sapiens)pDEST-FLAG-Cyclin FPMID: 27653696Transfected construct (human)
Transfected Construct (H. sapiens)Cyclin F M309A L131APMID: 20596027Transfected construct (human)
Transfected Construct (H. sapiens)pDEST-FLAG-Cyclin F M309A L313AThis paperTransfected construct (human)
Transfected Construct (H. sapiens)pGEX-GST-p130 593-790PMID: 9188854For protein expression in Escherichia coli
Transfected Construct (H. sapiens)pGEX-GST-p130 593-790 R680A L682AThis paperFor protein expression in Escherichia coli
Transfected Construct (H. sapiens)pINDUCER20PMID: 21307310Addgene #44012; RRID:Addgene_44012For lentiviral transfection
Transfected Construct (H. sapiens)pINDUCER20 CCNFThis paperFor lentiviral transfection
Transfected Construct (H. sapiens)pLV[Exp]-CMV> Tet3G/HygroVectorBuilderID:VB180123-1018bxqFor lentiviral transfection
Transfected Construct (H. sapiens)pLV[TetOn]-Neo-TRE3G > HA/{p130 AA}this paper, VectorBuilderID:VB200319-6469hpyFor lentiviral transfection
Transfected Construct (H. sapiens)pLV[TetOn]-Neo-TRE3G > HA/{p130 WT}this paper, VectorBuilderID:VB200319-6451nqdFor lentiviral transfection
Sequence-based reagentp130 from AA1 w/attb site ForwardPCR primers5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTAAT
GCCGTCGGGAGGTGACCAG
Sequence-based reagentp130 from AA417 w/attb site ReversePCR primers5′-GGGGACCACTTTGTACAAGAAAGCTGGGTA
CTACACACAAGGGCTATTCTCCTT
Sequence-based reagentp130 from AA418 w/attb site ForwardPCR primers5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTT
AACTCCAGTTTCTACAGCTACG
Sequence-based reagentp130 from AA 1139 w/attb site ReversePCR primers5′-GGGGACCACTTTGTACAAGAAAGCTGGGTA TCAGTGGGAACCACGGTCATT
Sequence-based reagentp130 from AA 616 w/attb site ReversePCR primers5′-GGGGACCACTTTGTACAAGAAAGCTGGG
TACTAAACTCTGTTTTCATTGTCTCT
Sequence-based reagentp130 from AA 827 w/attb site ReversePCR primers5′-GGGGACCACTTTGTACAAGAAAGCTGGG
TACTAACTACTGCTGGTTACAGACTG
Sequence-based reagentp130 from AA 1024 w/attb site ReversePCR primers5′-GGGGACCACTTTGTACAAGAAAGCTGGG
TACTAGTACTTCATGGCAAATGTCTT
Sequence-based reagentp130 from AA 828 w/attb site ForwardPCR primers5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTT
AAATAGACCCAGGAAGACCAGC
Sequence-based reagentp130 R658A I660A ForwardPCR primers5′-CGCCACATCTCCAACCACATTATAC
Sequence-based reagentp130 R658A I660A ReversePCR primers5′-CTGGCTCCAAGTCCTCCAGTATC
Sequence-based reagentp130 R680A L682A ForwardPCR primers5′-GGCCTTTGTTGAGAATGATAGCCCCTC
Sequence-based reagentp130 R680A L682A ReversePCR primers5′-CGGGCTCTGGTAGTGCTGGCTGG
Sequence-based reagentCyclin F from AA 1 w/attb site ForwardPCR primers5′-GGGGACAAGTTTGTACAAAAAAGCAGGC
TTAATGGGGAGCGGCGGCGTGGTCC
Sequence-based reagentCyclin F from end w/attb site ReversePCR primers5′-GGGGACCACTTTGTACAAGAAAGCTGG
GTATTACAGCCTCACAAGGCCCAGG
Sequence-based reagentCCNE1 RT-qPCR ForwardPCR primers5′-AGACATACTTAAGGGATCAGC
Sequence-based reagentCCNE1 RT-qPCR ReversePCR primers5′-CACACCTCCATTAACCAATC
Sequence-based reagentCDC6 RT-qPCR ForwardPCR primers5′-ATGTAAATCACCTTCTGAGC
Sequence-based reagentCDC6 RT-qPCR ReversePCR primers5′-GTCATCCTGTTACCATCAAC
Sequence-based reagentDHFR RT-qPCR ForwardPCR primers5′-TTCCAGAAGTCTAGATGATGC
Sequence-based reagentDHFR RT-qPCR ReversePCR primers5′-CTTCCTTATAAACAGAACTGCC
Sequence-based reagentE2F1 RT-qPCR ForwardPCR primers5′-CTGATGAATATCTGTACTACGC
Sequence-based reagentE2F1 RT-qPCR ReversePCR primers5′-CTTTGATCACCATAACCATCTG
Sequence-based reagentGAPDH RT-qPCR ForwardPCR primers5′-GGCCTCCAAGGAGTAAGACC
Sequence-based reagentGAPDH RT-qPCR ReversePCR primers5′-AGGGGTCTACATGGCAACTG
Sequence-based reagentCyclin F RT-qPCR ForwardPCR primers5′-AGGACAAGCGCTATGGAGAA
Sequence-based reagentCyclin F RT-qPCR ReversePCR primers5′-TCTGTCTTCCTGGAGGCTGT
Sequence-based reagentCDT1 RT-qPCR ForwardPCR primers5′-CCTGGGGAAATGGAGAAG
Sequence-based reagentCDT1 RT-qPCR ReversePCR primers5′-TTGTCCAGCTTGACGTAG
Sequence-based reagentCyclin F subcloning into pfastbac ForwardPCR primers5′-GCTAGGGTCGGATCCAGGAGGCCCCGAAACCTGACC
Sequence-based reagentCyclin F subcloning into pfastbac ReversePCR primers3′-GCTAGGCATAGCGGCCGCACCTTAGCTGT
CTTGTGTCACTCCTAATGCAGC
Sequence-based reagentSkp1 subcloning into PGEX-4T-1 ForwardPCR primers5′-GCTAGGGTCGGATCCATGCCTT
CAATTAAGTTGCAGAGTTCTGATGG
Sequence-based reagentSkp1 subcloning into PGEX-4T-1 ReversePCR primers3′-GCTAGGCATAGCCTCGAGTTACTT
CTCTTCACACCACTGGTTCTC
Sequence-based reagentCCNF #1siRNA5′-UAGCCUACCUCUACAAUGAUU
Sequence-based reagentCCNF #2siRNA5′-GCACCCGGUUUAUCAGUAAUU
Sequence-based reagentsiFFsiRNA5′-CGUACGCGGAAUACUUCGAUU
OtherEdUSigma-AldrichCat. #: T511285For flow cytometry-–10 µM to
cell media for 30 min prior to fixation
OtherDAPIThermo Fisher ScientificCat. #: D1306For flow cytometry (1 µg/ml)
OtherAlexa-Fluor 488 AzideThermo Fisher ScientificCat. #: A10266For flow cytometry (0.2 µM)

Data availability

Unprocessed, uncropped, immunoblots are made available in the supplemental source data. All raw, unprocessed imaging data is available at Dryad. Raw data related to cell proliferation assays (cell counting and Presto-blue analysis), RT-qPCR, immunoblot quantification for cycloheximide chase experiments and flow cytometry is available in the supplemental source data. All reagents related to this work will be made fully available upon request.

The following data sets were generated
    1. Enrico T
    2. Stallaert W
    3. Wick E
    4. Ngoi P
    5. Emanuele M
    6. Rubin S
    7. Brown N
    8. Purvis J
    (2021) Dryad Digital Repository
    Data from: Cyclin F drives proliferation through SCF-dependent degradation of the retinoblastoma-like tumor suppressor p130/RBL2.
    https://doi.org/10.5061/dryad.69p8cz93d
The following previously published data sets were used

References

    1. Baldi A
    2. Esposito V
    (1997)
    Differential Tissues and Expression in Primary Lung and in Normal Human Tissues and in Primary Lung Cancer
    Clinical Cancer Research 3:1691–1697.
    1. Bellan C
    2. De Falco G
    3. Tosi GM
    4. Lazzi S
    5. Ferrari F
    6. Morbini G
    7. Bartolomei S
    8. Toti P
    9. Mangiavacchi P
    10. Cevenini G
    11. Trimarchi C
    12. Cinti C
    13. Giordano A
    14. Leoncini L
    15. Tosi P
    16. Cottier H
    (2002)
    Missing expression of pRb2/p130 in human retinoblastomas is associated with reduced apoptosis and lesser differentiation
    Investig Ophthalmol Vis Sci 43:3602–3608.
    1. Claudio PP
    2. Howard CM
    3. Baldi A
    4. De Luca A
    5. Fu Y
    6. Condorelli G
    7. Sun Y
    8. Colburn N
    9. Calabretta B
    10. Giordano A
    (1994)
    p130/pRb2 has growth suppressive properties similar to yet distinctive from those of retinoblastoma family members pRb and p107
    Cancer Research 54:5556–5560.
    1. Fry DW
    2. Harvey PJ
    3. Keller PR
    4. Elliott WL
    5. Meade M
    6. Trachet E
    7. Albassam M
    8. Zheng X
    9. Leopold WR
    10. Pryer NK
    11. Toogood PL
    (2004)
    Specific inhibition of cyclin-dependent kinase 4/6 by PD 0332991 and associated antitumor activity in human tumor xenografts
    Molecular Cancer Therapeutics 3:1427–1438.

Decision letter

  1. Silke Hauf
    Reviewing Editor; Virginia Tech, United States
  2. Detlef Weigel
    Senior Editor; Max Planck Institute for Developmental Biology, Germany
  3. Michele Pagano
    Reviewer; New York Univ. School of Medicine, United States

Our editorial process produces two outputs: (i) public reviews designed to be posted alongside the preprint for the benefit of readers; (ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Decision letter after peer review:

Thank you for submitting your article "Cyclin F drives proliferation through SCF-dependent degradation of the retinoblastoma-like tumor suppressor p130/RBL2" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Maureen Murphy as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Michele Pagano (Reviewer #3).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions

1) The authors provided evidence that an R680xL682/AxA mutant is unable to interact with cyclin F and that this mutant blocks cell cycle gene expression, cell cycle progression and proliferation. The authors do not acknowledge or account for prior reports that this conserved RxL motif has been implicated in cyclin A and cyclin E binding in both p130 and p107 (Lacy and Whyte, 1997) as well as cyclin D binding in p107 (Leng et al., 2002). Thus, the p130 interaction data and functional data for the AxA mutant are likely to be true and unrelated as the functional data for the AxA likely results from a deficiency in CDK phosphorylation rather than increased stability. The authors need to show that the p130 AxA mutant can still interact with cyclins D, A, and E and be phosphorylated at CDK sites such as S672 for the alternative deficiency in CDK phosphorylation hypothesis to be ruled out for the AxA mutant.

2) If the p130 AxA mutant is competent for CDK phosphorylation, this raises questions regarding the disassembly of the DREAM complex. Previously, CDK4/6 phosphorylation of p130 has implicated in DREAM disassembly but, if the repression of cell cycle gene expression observed in the AxA mutant is due to increased protein stability rather than CDK phosphorylation, this would indicate that p130 degradation is the rate limiting step of DREAM disassembly or that p130 has a transcriptional repression function independent of the DREAM complex. The authors should address p130 degradation within DREAM complex function to resolve this conflict as to reduced p130 degradation leads to increase repression of cell cycle gene expression.

3) Many of the cellular experiments establishing that p130 is cyclin F substrate (Figure 2E, 3, 4B, 5B, S3A) were conducted in 293T cells that express Adenovirus 5 E1A and SV40 Large T antigen which bind and inactivate Rb family members and disrupt the RB-CDK network. Furthermore, SV40 large T promotes dephosphorylation of p130 (Lin and DeCaprio, 2003) explaining why SKP2 overexpression did not alter p130 levels as the SKP2 interacts with phosphorylated p130. SV40 LT also binds directly to FBXW7. Thus, the generalizability of these experiments is limited. Ideally, all the experiments in 293T cells would be repeated in cells where the RB-CDK network was intact but the strong in vitro data for the mapping the cyclin F-p130 interaction on p130 indicates that this may be unnecessary for Figure 2E, 4B, and S3A. At a minimum, the experiments in Figure 3, especially 3B, and 5B should be repeated in cells where the RB-CDK network is intact.

4) The transition from Figure 1 to Figure 2 is a difficult to follow in the text as Figure 2A and 2B are correlative observations. Figure 2A and 2B feel more like a follow up to 1D than addressing the question of whether and SCF cyclin F complex regulates p130 levels. Consequently, please move Figures 2A and 2B to figure 1 and Figure S2 should be brought into the main figure.

5) Figures 6 and 7 are poorly organized and the narrative is a bit difficult to follow and appreciate as the authors bounce between cell cycle, proliferation, and apoptotic readouts. It may strengthen and focus the story to focus on cell cycle gene expression and cell cycle progression in Figure 6 and cellular proliferation and apoptosis in Figure 7. This would also help the reader appreciate the CDK4/6i inhibitor data more as CDK4/6i inhibition is generally considered cytostatic rather than cytotoxic.

6) Since CDT1 protein species are degraded by multiple complexes, please include RT-qPCR data for CDT1 so the reader may appreciate that the decrease in CDT1 levels results from reduced transcription.

7) On lines 293-295, the authors compare p130 levels in S4A and 6A to point out the induced p130 AxA accumulates to higher levels at later time points compared to p130 WT. It is difficult to appreciate this comparison as these are different blots in different figures. Consequently, to make this point, the authors should include an IB as in S4A to show p130 WT and AxA levels over the 14 day time course. Ideally, time points would be collected.

8) The authors should include representative images for key time points and comparison for the multiple figures of quantitative immunofluorescence experiments, particularly for the nuclear to cytoplasmic ratio. Also the authors should more fully explain their analysis workflow as data can be processed through the "standard modules" cell profiler to differing results. If a previously established workflow was used that has been previously described in the literature, please provide a citation for the workflow.

9) Inactivation of p130 by CDK phosphorylation is likely to dominate p130 degradation in functional assay under normal growth conditions. The authors should consider examining the p130 degradation after DNA damage, in presence of reduced serum levels, or during quiescence (by expression a cyclin F KEN box mutant) to discern the functional impact of altered p130 degradation by p130.

10) It would help the reader if the results of the iterative immunofluorescence staining were supplemented by additional data such as representative cell images and/or immunoblots. It is difficult to appreciate the extent and the significance of the observed effects in the format shown in Figures 6F-H, 7B-C and S5C-E.

11) Standard methods such as immunoblotting or protein expression in bacteria, are described in excessive detail, whereas the 4i image analysis that could be helpful to a reader unfamiliar with this technique, is described rather briefly. It took a while and reading some additional papers for this reviewer to understand what in fact was shown in the graphs obtained by this technique.

12) Figure 1. From the description, it appears that results of the DepMap correlation analysis form the premise for this work by showing that depletion of cyclin F has similar impact on the cancer cell fitness as depletion of several components of the Rb-CDK pathway, including p130. The authors then hypothesize that cyclin F may regulate the expression of these factors, which leads to identification of p130 as a cyclin F substrate. Significance of this analysis should be better explained. Indeed, the authors first make the introduction that closely correlated genes in DepMap are often involved in the same protein complex or a functional pathway, and then go on to show that cyclin F and p130 not only have the opposite patterns if expression, but also negatively regulate each other, which would be contradictory to a direct correlation found in DepMap analysis. It would help the reader if the authors explained this apparent contradiction, and ideally validated the conclusions of the DepMap analysis by comparing the effects of cyclin F and p130 knockdown in a panel of select cancer cell lines.

13) Figure 2C: In HeLa cells, all Rb family members are targeted to degradation through the interaction with HPV E7 oncoprotein. Although loss of CCNF appears to increase the levels of p130 in HeLa cells, this result does not strongly support the direct role of CCNF in p130 degradation without ruling out an indirect effect on E7 function. Additionally, loss of CCNF could influence the cell cycle synchronization therefore the cell cycle analysis data should be included with this experiment. In general, since cyclin F is involved in regulation of many cell-cycle related factors, the cell cycle data should be included in the experiments where the levels of CCNF are manipulated such as Figure 2D.

14) Figure 2E: Again, the choice of 293T cell line to demonstrate the binding between the endogenous p130 and cyclin F is surprising. The levels of p130 in these cells are very low because of the presence of SV40 Large T antigen that targets Rb family members for degradation. Furthermore, SV40 LT binds to Rb family members and can indirectly recruit other factors into the complexes with these proteins. Therefore, the endogenous interaction between p130 and cyclin F should be confirmed in another cell line. Figure 3: Same concerns – the presence of SV40 LT could complicate the interpretation of these results. The authors need to confirm these effects in another cell line, in the absence of SV40 LT.

15) Figure 4: Panel B: It would be interesting to know whether R680A/L682A p130 mutant can still bind SV40 Large T or HPV E7 proteins. Panel E: The rationale for using E2F1 displacement for measuring the binding affinity between cyclin F-Skp1 complex and p130 fragments is unclear. Is it possible to show a direct binding between these factors?

16) Figure 5B: It is apparent that p130-AA is still being degraded during the CHX chase, but is unaffected by cyclin F overexpression. Could the proteasome or neddylation inhibitors block this degradation? Is it possible that both Skp2 and cyclin F can degrade p130 through different mechanisms? Panel C: Significance of the differences is not shown.

17) Figure 6: It would be very interesting to see how the p130-AA levels are regulated during the G0/G1 arrest and progression through the cell cycle. This could be a key experiment to demonstrate the role of the 680-RRL site for cell cycle dependent degradation of p130. Panel E: Description of this figure indicates that cells were pulse-labeled with EdU to detect S-phase, but the figure only shows G0/G1 fraction. Changes in S-phase and G2/M fractions could be informative and should be shown. Also, this experiment needs to be better explained. Were the cells allowed to grow to confluency over 14 days? Was the growth media changed or supplemented throughout this experiment? Panels F-H: The legend needs to be revised to explain what is shown in the graphs. Was the statistical significance calculated, and how? Which data points are significantly different?

18) Figure 7A. This is an interesting result because it is hard to explain such robust effect on the DREAM target gene expression by a pretty small difference in the p130 expression levels. This difference is not even apparent in the panel D in this Figure. What is the possible explanation of this result? Also, the endogenous p130 immunoblot should be included in this figure to appreciate the extent of the overexpression (also in figure S4A). Panel D: It appears that only a small fraction of the p130-AA cells undergoes apoptosis. Additional evidence is needed to support the conclusion that this mutant promotes apoptosis.

19) Using DepMap, the authors find that CCNF KO is associated with Palbociclib inactivation of the CDK-RB network, but they fail to show experimentally a correlation between the cyclin F-mediated degradation of p130 and sensitivity to CDK4/6 inhibitors. This should be addressed.

https://doi.org/10.7554/eLife.70691.sa1

Author response

Essential revisions

1) The authors provided evidence that an R680xL682/AxA mutant is unable to interact with cyclin F and that this mutant blocks cell cycle gene expression, cell cycle progression and proliferation. The authors do not acknowledge or account for prior reports that this conserved RxL motif has been implicated in cyclin A and cyclin E binding in both p130 and p107 (Lacy and Whyte, 1997) as well as cyclin D binding in p107 (Leng et al., 2002). Thus, the p130 interaction data and functional data for the AxA mutant are likely to be true and unrelated as the functional data for the AxA likely results from a deficiency in CDK phosphorylation rather than increased stability. The authors need to show that the p130 AxA mutant can still interact with cyclins D, A, and E and be phosphorylated at CDK sites such as S672 for the alternative deficiency in CDK phosphorylation hypothesis to be ruled out for the AxA mutant.

This is a great point. As you point out, an important paper from Lacy and Whyte previously showed that the same sequence motif in p130, which we mapped as being critical for cyclin F binding, stability, ubiquitination, and degradation (amino acids 680-682), could also potentially mediate cyclin E/A binding.

We have addressed this important point with a series of new experiments. First, we showed that the p130(R680A, L682A) mutant (hereafter, p130(AA)) is phosphorylated in vivo at an established CDK site. We performed pulldown experiments with HA-tagged p130(WT) and p130(AA) and blotted them using the published, commercially available, p130 phospho-S672 antibody (new Figure 5 Supplement 2C). This reagent has been well validated by experts in this field (e.g., DeCaprio lab, Litovchick lab, etc.). Both p130(WT) and p130(AA) were similarly phosphorylated on S672. Next, since that antibody only detects p130 phosphorylation at a single site, we also examined in-gel migration of p130. Phosphorylated p130 migrates more slowly in SDS-PAGE. Consistently, HA-tagged versions of p130(WT) and p130(AA) showed similar migration patterns and were similarly impacted by phosphatase treatment post-lysis (new Figure 5 Supplement 2D). Together, these data strongly suggest that p130(AA) is phosphorylated in vivo. Finally, we examined p130 binding to cyclin E and cyclin A by immunoprecipitation (IP). We show that IPs of HA-tagged p130(WT) and p130(AA) copurify cyclin A and cyclin E similarly (See new Figure 5 Supplement 2A). Likewise, IP of endogenous cyclin A co-precipitated indistinguishable levels of HA-tagged p130(WT) and p130(AA) (new Figure 5 Supplement 2B). Altogether, these experiments argue that the impact of the p130(AA) mutant is not the result of a defect in phosphorylation or cyclin E/A binding.

We revisited Lacy and Whyte’s paper, which we should have cited and discussed more clearly. It is notable that they did not analyze phosphorylation of full length p130 in human cells. Rather, they demonstrated that deleting amino acids 680-682 from a fragment of GST-p130 (amino acids 595-675) impaired binding to cyclins in SF9 cell extracts derived from cells overexpressing cyclin A/CDK2 or cyclin E/CDK2 complexes. They showed that this deletion reduced phosphorylation of the GST-p130 fragment, also using insect cell extracts overexpressing cyclin/CDK complexes. They did not test binding of this mutant to cyclins, or its phosphorylation, in cultured human cells.

Also notable are two additional studies, from the Dynlacht and Lukas labs, reporting that cyclin E/A binding to p130 instead relied on a different site in the p130 N-terminus (PMIDs 9710622 and 11157749). And a more recent report showing that p130 phosphorylation by cyclin D-CDK4 depends on a c-terminal helical domain (PMID 30982746). These studies and the Lacy and Whyte paper, have all been cited.

2) If the p130 AxA mutant is competent for CDK phosphorylation, this raises questions regarding the disassembly of the DREAM complex. Previously, CDK4/6 phosphorylation of p130 has implicated in DREAM disassembly but, if the repression of cell cycle gene expression observed in the AxA mutant is due to increased protein stability rather than CDK phosphorylation, this would indicate that p130 degradation is the rate limiting step of DREAM disassembly or that p130 has a transcriptional repression function independent of the DREAM complex. The authors should address p130 degradation within DREAM complex function to resolve this conflict as to reduced p130 degradation leads to increase repression of cell cycle gene expression.

This is an interesting question with respect to whether p130 degradation is limiting, particularly with respect to phosphorylation. Importantly, we found that forced expression of low levels of p130(WT) is sufficient to slow proliferation in NHF cells (Figure 6 Supplement 1C and Figure 6C), consistent with overexpression results from other groups (including the Lacy and Whyte study). Thus, as is the case with RB1, simply increasing the levels of p130 is sufficient to repress proliferation, and this is presumably through its established function in cell cycle gene repression through DREAM.

In our assays, this proliferation defect is profoundly exacerbated by mutation of the cyclin F binding site in p130 (Figure 6), which allows p130 protein to accumulate due to it not being degraded. Our data are consistent with the prior observations where increasing p130 levels in cells is sufficient to slow proliferation. Since the p130(AA) mutant can bind other cyclins and is phosphorylated (see point 1 above), this suggests that degradation is pivotal for p130 inactivation. Further, we determined p130(AA) binding to DREAM, and now show that p130(AA) can assemble into the DREAM complex akin to p130(WT), based on its ability to co-immuno-precipitate with LIN54 (new Figure 5 Supplement 2C). Altogether, this suggests that increasing the level of a phosphorylation proficient p130, by interfering with its destruction, represses cell cycle gene expression and proliferation.

3) Many of the cellular experiments establishing that p130 is cyclin F substrate (Figure 2E, 3, 4B, 5B, S3A) were conducted in 293T cells that express Adenovirus 5 E1A and SV40 Large T antigen which bind and inactivate Rb family members and disrupt the RB-CDK network. Furthermore, SV40 large T promotes dephosphorylation of p130 (Lin and DeCaprio, 2003) explaining why SKP2 overexpression did not alter p130 levels as the SKP2 interacts with phosphorylated p130. SV40 LT also binds directly to FBXW7. Thus, the generalizability of these experiments is limited. Ideally, all the experiments in 293T cells would be repeated in cells where the RB-CDK network was intact but the strong in vitro data for the mapping the cyclin F-p130 interaction on p130 indicates that this may be unnecessary for Figure 2E, 4B, and S3A. At a minimum, the experiments in Figure 3, especially 3B, and 5B should be repeated in cells where the RB-CDK network is intact.

To address this potential concern, we now show that ectopic expression of cyclin F, but not SKP2, bTRCP or Fbxw7, drives p130 degradation in U2OS cells (new Figure 3 Supplement 1A). Further, we now show that cyclin F also regulates p130 half-life in U2OS cells (new Figure 5 Supplement 1A). We have also shown that cyclin F overexpression reduces endogenous p130 protein levels in MCF7 and T47D cells, luminal breast cancer cell lines that do not express large T antigen or HPV oncoproteins (Figure 1D). And, cyclin F depletion increases p130 protein levels in IMR-90 and NHF-1 cell lines, both of which are negative for large T antigen and HPV oncoproteins (Figure 2B).

4) The transition from Figure 1 to Figure 2 is a difficult to follow in the text as Figure 2A and 2B are correlative observations. Figure 2A and 2B feel more like a follow up to 1D than addressing the question of whether and SCF cyclin F complex regulates p130 levels. Consequently, please move Figures 2A and 2B to figure 1 and Figure S2 should be brought into the main figure.

Thank you for these thoughtful suggestions. To accommodate these recommendations, and improve the flow of the manuscript, we have re-organized Figures 1 and 2. Specifically, part of the previous Figure S2, which showed data in NHF-1 cells released from arrest into MLN4924, was moved to Figure 1F. Similarly, the previous Figure 2A, which showed cyclin F and p130 levels in proliferating and quiescent cells across different lines, was moved Figure 1E. The previous Figure 2B, which is similar to the new Figure 1F, was moved to Figure 1 Supplement 3, which also still includes data from T98G cells released from arrest into MLN4924. We have updated the manuscript text accordingly.

5) Figures 6 and 7 are poorly organized and the narrative is a bit difficult to follow and appreciate as the authors bounce between cell cycle, proliferation, and apoptotic readouts. It may strengthen and focus the story to focus on cell cycle gene expression and cell cycle progression in Figure 6 and cellular proliferation and apoptosis in Figure 7. This would also help the reader appreciate the CDK4/6i inhibitor data more as CDK4/6i inhibition is generally considered cytostatic rather than cytotoxic.

We have tried to best accommodate this by making the following changes. Since the apoptotic results are significant, but less well understood, these have been moved to supplemental figures. The phenotype data in Figure 6 begins with showing the protein dynamics of p130(WT) vs p130(AA) and then begins with the broadest of phenotypes, being proliferation. This is followed by cell cycle and 4i imaging of cell cycle related proteins. Then, in Figure 7, we drill down into effects on cell cycle gene expression. We have updated the writing to hopefully make this clearer.

6) Since CDT1 protein species are degraded by multiple complexes, please include RT-qPCR data for CDT1 so the reader may appreciate that the decrease in CDT1 levels results from reduced transcription.

We have included CDT1 RT-qPCR data, demonstrating that the reduction in CDT1 protein levels is due to reduced transcription (see Figures 7A and Figure 7 Supplement 1A).

7) On lines 293-295, the authors compare p130 levels in S4A and 6A to point out the induced p130 AxA accumulates to higher levels at later time points compared to p130 WT. It is difficult to appreciate this comparison as these are different blots in different figures. Consequently, to make this point, the authors should include an IB as in S4A to show p130 WT and AxA levels over the 14 day time course. Ideally, time points would be collected.

This is a great point. We have now included a time course of p130(WT) and p130(AA) levels from zero to 14 days. Levels of p130(AA) accumulate over time (see new Figure 6A).

8) The authors should include representative images for key time points and comparison for the multiple figures of quantitative immunofluorescence experiments, particularly for the nuclear to cytoplasmic ratio. Also the authors should more fully explain their analysis workflow as data can be processed through the "standard modules" cell profiler to differing results. If a previously established workflow was used that has been previously described in the literature, please provide a citation for the workflow.

We have addressed this suggestion by including representative images for phosphorylated RB, CDT1, and CDC6 (Figures 6 and 7). We further elaborated when describing the 4i image acquisition and data processing to clarify methods used.

9) Inactivation of p130 by CDK phosphorylation is likely to dominate p130 degradation in functional assay under normal growth conditions. The authors should consider examining the p130 degradation after DNA damage, in presence of reduced serum levels, or during quiescence (by expression a cyclin F KEN box mutant) to discern the functional impact of altered p130 degradation by p130.

This is a very interesting point. We observed that in several cell lines made quiescent by serum starvation, p130 is upregulated and cyclin F downregulated (Figure 1E). We anticipate that p130 increases in quiescence, at least in part, because of reduced cyclin F. We examined p130 protein levels in response to DNA damage, which has been reported to cause an increase p130 abundance, and during when cyclin F is degraded. While we could recapitulate a p130 increase using MMC at specific doses, as described by others, we see inconsistent results depending on the type of damaging agent applied and with dose of damaging agent. Since these results are inconsistent and inconclusive, we have not included them. We are currently unable to overexpress cyclin F in quiescent cells, precluding that analysis.

Related to this point, and one below, we also now provide new experimental evidence that in NHF-1 cells, p130 is degraded in mitotic NHF cells arrested by nocodazole. Significantly, the p130(AA) mutant which cannot bind or be ubiquitinated by cyclin F, is resistant to degradation in nocodazole arrested cells (see new Figure 5D and Figure 5 Supplement 5C). Examining this further represents an important area of future work but is beyond the scope of the current study.

10) It would help the reader if the results of the iterative immunofluorescence staining were supplemented by additional data such as representative cell images and/or immunoblots. It is difficult to appreciate the extent and the significance of the observed effects in the format shown in Figures 6F-H, 7B-C and S5C-E.

We have addressed this suggestion by including representative images for phosphorylated RB, CDT1, and CDC6 (Figures 6 and 7).

11) Standard methods such as immunoblotting or protein expression in bacteria, are described in excessive detail, whereas the 4i image analysis that could be helpful to a reader unfamiliar with this technique, is described rather briefly. It took a while and reading some additional papers for this reviewer to understand what in fact was shown in the graphs obtained by this technique.

Thank you for this very helpful suggestion. We have addressed this by explaining in better detail the 4i technique and providing representative images of cells from these experiments. We have also reduced other explanations in the manuscript that were described with unnecessary detail.

12) Figure 1. From the description, it appears that results of the DepMap correlation analysis form the premise for this work by showing that depletion of cyclin F has similar impact on the cancer cell fitness as depletion of several components of the Rb-CDK pathway, including p130. The authors then hypothesize that cyclin F may regulate the expression of these factors, which leads to identification of p130 as a cyclin F substrate. Significance of this analysis should be better explained. Indeed, the authors first make the introduction that closely correlated genes in DepMap are often involved in the same protein complex or a functional pathway, and then go on to show that cyclin F and p130 not only have the opposite patterns if expression, but also negatively regulate each other, which would be contradictory to a direct correlation found in DepMap analysis. It would help the reader if the authors explained this apparent contradiction, and ideally validated the conclusions of the DepMap analysis by comparing the effects of cyclin F and p130 knockdown in a panel of select cancer cell lines.

This is an extremely interesting point and we have given this considerable thought. Cyclin D1, CDK4, RBL1 and RBL2, are all highly, positively correlated with cyclin F. We agree that at first glance this is confusing and appears contradictory. However, the more we survey the DepMap dataset, the more it becomes clear that positive and negative correlations cannot always be interpreted as indicating positive and negative relationships between proteins in a pathway.

To be clear, positive and negative correlations often do indicate positive and negative relationships, respectively. For example, the E3 ligase AMBRA1 is negatively correlated with the D-type cyclins, which it targets for degradation. Likewise, CDK4 is positively correlated with Cyclin D1.

However, there are many examples that defy this simple logic. CDK4 is negatively correlated with CDK6, with which it shares a common function. Likewise, D-type cyclins are negatively correlated with each other. These examples are likely explained by the fact that cell lines predominantly use CDK4 or CDK6, or only one D-type cyclin, but not all equally. Nevertheless, these examples demonstrate that caution is needed when interpreting positive and negative DepMap correlations. Furthermore, although Cyclin D1 is indeed positively correlated with CDK4, it is also positively correlated with RBL1 and RBL2, both of which it negatively regulates. Altogether, we think this implies a role for a specific circuitry in a given cell line, while not necessarily implying the direction of the relationship. Simply stated, positive DepMap correlations do not necessarily imply activating or positive enzymatic relationships.

To address your point, we compared fitness scores from the Project Achilles Cancer Dependency map dataset to fitness scores from the Sanger Project Score dataset in ~185 overlapping cell lines utilized by both programs. These represent the two largest and most comprehensive, large scale, CRISPR/Cas9, cancer screening programs. Notably, they are performed independently at different sites, by different individuals, and using different protocols and reagents. Prior analysis showed that these datasets are remarkably well correlated with each other, supporting the reliability of data in both (PMID: 31862961). We also now show correlations between the two datasets for genes in the CDK-RB network in new Figure 1 Supplement 2.

13) Figure 2C: In HeLa cells, all Rb family members are targeted to degradation through the interaction with HPV E7 oncoprotein. Although loss of CCNF appears to increase the levels of p130 in HeLa cells, this result does not strongly support the direct role of CCNF in p130 degradation without ruling out an indirect effect on E7 function. Additionally, loss of CCNF could influence the cell cycle synchronization therefore the cell cycle analysis data should be included with this experiment. In general, since cyclin F is involved in regulation of many cell-cycle related factors, the cell cycle data should be included in the experiments where the levels of CCNF are manipulated such as Figure 2D.

We have addressed this potential concern in several ways. First, we provide cell cycle flow cytometry data from cells in cyclin F siRNA-mediated depletion experiments (Figure 2 Supplement 1B-C) as well as for asynchronously proliferating cyclin F KO and control HeLa cells (Figure 2 supplement 1D). Consistently, we and others previously showed that the cyclin F KO HeLa cells are well adapted and grow quite normally in culture (similarly to knockout of most canonical cyclins). In addition, the HeLa KO synchronization experiment was blotted for cyclin E, which shows very similar kinetics between control KO and cyclin F KO cell lines (see Figure 2 Supplement 1A). Since cyclin F is the only thing different between cell lines, this strongly argues that even if E7 were involved in this cell line, that this still is a cyclin F dependent effect.

14) Figure 2E: Again, the choice of 293T cell line to demonstrate the binding between the endogenous p130 and cyclin F is surprising. The levels of p130 in these cells are very low because of the presence of SV40 Large T antigen that targets Rb family members for degradation. Furthermore, SV40 LT binds to Rb family members and can indirectly recruit other factors into the complexes with these proteins. Therefore, the endogenous interaction between p130 and cyclin F should be confirmed in another cell line. Figure 3: Same concerns – the presence of SV40 LT could complicate the interpretation of these results. The authors need to confirm these effects in another cell line, in the absence of SV40 LT.

We addressed this potential concern by performing an endogenous IP for cyclin F in U2OS cells, which showed that these endogenous proteins interact in this cell line as well (see new Figure 2C). In addition, we also now show that cyclin F expression can cause the degradation of p130 in U2OS cells, but that SKP2, bTRCP and Fbxw7 cannot (Figure 3 Supplement 1A).

15) Figure 4: Panel B: It would be interesting to know whether R680A/L682A p130 mutant can still bind SV40 Large T or HPV E7 proteins. Panel E: The rationale for using E2F1 displacement for measuring the binding affinity between cyclin F-Skp1 complex and p130 fragments is unclear. Is it possible to show a direct binding between these factors?

We have addressed this suggestion by analyzing binding of HA-tagged versions of p130(WT) and p130(AA) to E7 in HeLa cells and large T antigen in HEK293T cells. We see no difference in binding in either of these situations (see new Figure 5 Supplement 2E-F).

We also re-analyzed cyclin F IP-MS data from 293T cells for the presence of large T antigen and found that we did not enrich for large T antigen in those experiments. These data are not yet published.

16) Figure 5B: It is apparent that p130-AA is still being degraded during the CHX chase, but is unaffected by cyclin F overexpression. Could the proteasome or neddylation inhibitors block this degradation? Is it possible that both Skp2 and cyclin F can degrade p130 through different mechanisms? Panel C: Significance of the differences is not shown.

It is possible that Skp2 contributes to p130 degradation, and it would be difficult to rule this out completely. The focus of our paper is not to discern the potential role of Skp2 in p130 degradation. We tested Skp2 by overexpression in two cell lines, one of which was used in the original studies reporting a role for Skp2 (HEK293T cells) and obtained two negative results. We are concerned about overinterpretation of the CHX data, as this results in complete inactivation of all translation, and causes significant cellular stress. If MLN4924 completely blocked degradation in that assay, it still would not imply a role for Skp2 versus the hundreds of other cullin E3s.

17) Figure 6: It would be very interesting to see how the p130-AA levels are regulated during the G0/G1 arrest and progression through the cell cycle. This could be a key experiment to demonstrate the role of the 680-RRL site for cell cycle dependent degradation of p130. Panel E: Description of this figure indicates that cells were pulse-labeled with EdU to detect S-phase, but the figure only shows G0/G1 fraction. Changes in S-phase and G2/M fractions could be informative and should be shown. Also, this experiment needs to be better explained. Were the cells allowed to grow to confluency over 14 days? Was the growth media changed or supplemented throughout this experiment? Panels F-H: The legend needs to be revised to explain what is shown in the graphs. Was the statistical significance calculated, and how? Which data points are significantly different?

These are helpful suggestions. We had originally shown only G0/G1 cells in Figure 6 for simplicity but have now added the remaining data to the Figure 6 Supplement 1D.

The cells were not grown to confluency at any point in our proliferation assays, but rather were split/fed regularly, and this is now noted in the text. The data in panels is more clearly explained in the legend, and they have been updated to reflect statistical significance.

To address your first point, we synchronized cells expressing HA-tagged versions of p130(WT) and p130(AA) in different phases of the cell cycle. While we observed little or no difference in most conditions, to our surprise, we found that p130(WT) is degraded in nocodazole arrested NHF cells (see new Figure 5D). Likewise, endogenous p130 protein is degraded in nocodazole arrested NHF cells (see new Figure 5 Supplement 1C). Significantly, we did not observe p130(AA) degradation in nocodazole-synchronized cells (Figure 6E).

18) Figure 7A. This is an interesting result because it is hard to explain such robust effect on the DREAM target gene expression by a pretty small difference in the p130 expression levels. This difference is not even apparent in the panel D in this Figure. What is the possible explanation of this result? Also, the endogenous p130 immunoblot should be included in this figure to appreciate the extent of the overexpression (also in figure S4A). Panel D: It appears that only a small fraction of the p130-AA cells undergoes apoptosis. Additional evidence is needed to support the conclusion that this mutant promotes apoptosis.

To address the point regarding apoptosis we have now carried out annexin V staining and flow cytometry, and we observe the same effect. The increase in apoptosis is clear, reproducible, and statistically significant. We had originally analyzed apoptosis after seeing the dramatic impact on cell growth/fitness, and to rule out a role for apoptosis. Defying our hypothesis, our data indicated that a small amount of apoptosis is indeed occurring. These new flow cytometry data are included in Figure 7 Supplement 2B.

Since p130(AA) can assemble into the DREAM complex, and the protein can still be phosphorylated, the data in Figure 7A argues strongly that even a small increase in p130 expression is sufficient to impair cell cycle progression. Since there is likely little p130 in cycling cells, blocking its ubiquitination is sufficient to cause a strong defect in cell cycle gene expression programs and proliferation.

19) Using DepMap, the authors find that CCNF KO is associated with Palbociclib inactivation of the CDK-RB network, but they fail to show experimentally a correlation between the cyclin F-mediated degradation of p130 and sensitivity to CDK4/6 inhibitors. This should be addressed.

The DepMap consortium utilized barcoding strategies in hundreds of cells lines to determine sensitivities to myriad drugs, in an effort at drug repurposing. Our results show that cyclin F regulates p130. The DepMap analysis is consistent with our results, in that cyclin F has an impact on cells that is highly similar to CDK4, cyclin D, and also Palbociclib, all of which are negative regulators of the RB-family of proteins, and whose inactivation can enhance resistance to Palbociclib. Determining chemical-genetic interactions between CDK4/6 inhibitors and cyclin F is an important area of future study, but at this time, beyond the scope of this current study. We have therefore moved this observation to the Discussion, following consultation with the editor, and discussed it there appropriately.

https://doi.org/10.7554/eLife.70691.sa2

Article and author information

Author details

  1. Taylor P Enrico

    1. Department of Pharmacology. The University of North Carolina at Chapel Hill, Chapel Hill, United States
    2. Lineberger Comprehensive Cancer Center. The University of North Carolina at Chapel Hill, Chapel Hill, United States
    Contribution
    Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing - original draft, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-5453-2868

  2. Wayne Stallaert

    Department of Genetics. The University of North Carolina at Chapel Hill, Chapel Hill, United States
    Contribution
    Formal analysis, Investigation, Methodology, Visualization, Writing – review and editing
    Competing interests
    No competing interests declared

  3. Elizaveta T Wick

    1. Department of Pharmacology. The University of North Carolina at Chapel Hill, Chapel Hill, United States
    2. Lineberger Comprehensive Cancer Center. The University of North Carolina at Chapel Hill, Chapel Hill, United States
    Contribution
    Investigation, Methodology, Writing - original draft, Writing – review and editing
    Competing interests
    No competing interests declared

  4. Peter Ngoi

    Department of Chemistry and Biochemistry. University of California at Santa Cruz, Santa Cruz, United States
    Contribution
    Formal analysis, Investigation, Methodology, Visualization, Writing – review and editing
    Competing interests
    No competing interests declared

  5. Xianxi Wang

    Lineberger Comprehensive Cancer Center. The University of North Carolina at Chapel Hill, Chapel Hill, United States
    Contribution
    Investigation, Methodology
    Competing interests
    No competing interests declared

  6. Seth M Rubin

    Department of Chemistry and Biochemistry. University of California at Santa Cruz, Santa Cruz, United States
    Contribution
    Conceptualization, Formal analysis, Funding acquisition, Methodology, Project administration, Supervision, Visualization, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1670-4147

  7. Nicholas G Brown

    1. Department of Pharmacology. The University of North Carolina at Chapel Hill, Chapel Hill, United States
    2. Lineberger Comprehensive Cancer Center. The University of North Carolina at Chapel Hill, Chapel Hill, United States
    Contribution
    Conceptualization, Formal analysis, Funding acquisition, Methodology, Project administration, Supervision, Visualization, Writing – review and editing
    Competing interests
    No competing interests declared

  8. Jeremy E Purvis

    1. Lineberger Comprehensive Cancer Center. The University of North Carolina at Chapel Hill, Chapel Hill, United States
    2. Department of Genetics. The University of North Carolina at Chapel Hill, Chapel Hill, United States
    Contribution
    Conceptualization, Formal analysis, Funding acquisition, Methodology, Project administration, Supervision, Visualization, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-6963-0524

  9. Michael J Emanuele

    1. Department of Pharmacology. The University of North Carolina at Chapel Hill, Chapel Hill, United States
    2. Lineberger Comprehensive Cancer Center. The University of North Carolina at Chapel Hill, Chapel Hill, United States
    Contribution
    Conceptualization, Funding acquisition, Project administration, Supervision, Visualization, Writing - original draft, Writing – review and editing
    For correspondence
    emanuele@email.unc.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-4104-7449

Funding

National Institute of General Medical Sciences (R01GM120309)

  • Taylor P Enrico
  • Xianxi Wang
  • Michael J Emanuele

National Institute of General Medical Sciences (R01GM134231)

  • Taylor P Enrico
  • Xianxi Wang
  • Michael J Emanuele

National Institute of General Medical Sciences (R35GM128855)

  • Elizaveta T Wick
  • Nicholas G Brown

American Cancer Society (RSG-18-220-01-TBG)

  • Taylor P Enrico
  • Xianxi Wang
  • Michael J Emanuele

National Cancer Institute (R01CA163834)

  • Elizaveta T Wick

National Institute of General Medical Sciences (R01GM127707)

  • Peter Ngoi
  • Seth M Rubin

National Institute of General Medical Sciences (GM138834)

  • Wayne Stallaert
  • Jeremy E Purvis

National Institute of General Medical Sciences (DP2-HD091800)

  • Wayne Stallaert
  • Jeremy E Purvis

National Science Foundation (1845796)

  • Jeremy E Purvis

National Institute of General Medical Sciences (T32 GM007040)

  • Taylor P Enrico

National Institute of General Medical Sciences (T32 GM007040)

  • Taylor P Enrico

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

The authors thank lab members for helpful discussions throughout this project. The authors thank Brenda Schulman (Max Planck Institute of Biochemistry) for generously providing SCF complex reagents used for in vitro ubiquitination assays. The authors thank Larisa Litovchick (Virginia Commonwealth University) for providing the pBABE-p130 vector. The authors thank Dennis Goldfarb (Washington University) for help with DepMap data downloads and analysis. The Emanuele lab (TPE and MJE) is supported by the UNC University Cancer Research Fund (UCRF), the National Institutes of Health (R01GM120309, R01GM134231), and the America Cancer Society (Research Scholar Grant; RSG-18-220-01-TBG). The Brown lab (ETW and NGB) is supported by UCRF and National Institutes of Health (R35GM128855) and ETW is partially supported by R01CA163834. The Rubin lab (PN and SMR) is supported by the National Institutes of Health (R01GM127707). The Purvis lab (WMS and JEP) is supported by by NIH grants R01-GM138834 (JEP), DP2-HD091800 (JEP), and NSF CAREER Award 1845796 (JEP).

Senior Editor

  1. Detlef Weigel, Max Planck Institute for Developmental Biology, Germany

Reviewing Editor

  1. Silke Hauf, Virginia Tech, United States

Reviewer

  1. Michele Pagano, New York Univ. School of Medicine, United States

Version history

  1. Preprint posted: April 24, 2021 (view preprint)
  2. Received: May 26, 2021
  3. Accepted: November 19, 2021
  4. Accepted Manuscript published: December 1, 2021 (version 1)
  5. Version of Record published: December 14, 2021 (version 2)

Copyright

© 2021, Enrico et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 2,173
    Page views
  • 239
    Downloads
  • 7
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Taylor P Enrico
  2. Wayne Stallaert
  3. Elizaveta T Wick
  4. Peter Ngoi
  5. Xianxi Wang
  6. Seth M Rubin
  7. Nicholas G Brown
  8. Jeremy E Purvis
  9. Michael J Emanuele
(2021)
Cyclin F drives proliferation through SCF-dependent degradation of the retinoblastoma-like tumor suppressor p130/RBL2
eLife 10:e70691.
https://doi.org/10.7554/eLife.70691

Categories and tags

Research organism

Further reading

    1. Cancer Biology
    2. Computational and Systems Biology

    Pancreatic cancer symptom trajectories from Danish registry data and free text in electronic health records

    Jessica Xin Hjaltelin, Sif Ingibergsdóttir Novitski ... Søren Brunak
    Research Article

    Pancreatic cancer is one of the deadliest cancer types with poor treatment options. Better detection of early symptoms and relevant disease correlations could improve pancreatic cancer prognosis. In this retrospective study, we used symptom and disease codes (ICD-10) from the Danish National Patient Registry (NPR) encompassing 6.9 million patients from 1994 to 2018,, of whom 23,592 were diagnosed with pancreatic cancer. The Danish cancer registry included 18,523 of these patients. To complement and compare the registry diagnosis codes with deeper clinical data, we used a text mining approach to extract symptoms from free text clinical notes in electronic health records (3078 pancreatic cancer patients and 30,780 controls). We used both data sources to generate and compare symptom disease trajectories to uncover temporal patterns of symptoms prior to pancreatic cancer diagnosis for the same patients. We show that the text mining of the clinical notes was able to complement the registry-based symptoms by capturing more symptoms prior to pancreatic cancer diagnosis. For example, ‘Blood pressure reading without diagnosis’, ‘Abnormalities of heartbeat’, and ‘Intestinal obstruction’ were not found for the registry-based analysis. Chaining symptoms together in trajectories identified two groups of patients with lower median survival (<90 days) following the trajectories ‘Cough→Jaundice→Intestinal obstruction’ and ‘Pain→Jaundice→Abnormal results of function studies’. These results provide a comprehensive comparison of the two types of pancreatic cancer symptom trajectories, which in combination can leverage the full potential of the health data and ultimately provide a fuller picture for detection of early risk factors for pancreatic cancer.

    1. Cancer Biology

    Disseminating cells in human oral tumours possess an EMT cancer stem cell marker profile that is predictive of metastasis in image-based machine learning

    Gehad Youssef, Luke Gammon ... Adrian Biddle
    Research Article

    Cancer stem cells (CSCs) undergo epithelial-mesenchymal transition (EMT) to drive metastatic dissemination in experimental cancer models. However, tumour cells undergoing EMT have not been observed disseminating into the tissue surrounding human tumour specimens, leaving the relevance to human cancer uncertain. We have previously identified both EpCAM and CD24 as CSC markers that, alongside the mesenchymal marker Vimentin, identify EMT CSCs in human oral cancer cell lines. This afforded the opportunity to investigate whether the combination of these three markers can identify disseminating EMT CSCs in actual human tumours. Examining disseminating tumour cells in over 12,000 imaging fields from 74 human oral tumours, we see a significant enrichment of EpCAM, CD24 and Vimentin co-stained cells disseminating beyond the tumour body in metastatic specimens. Through training an artificial neural network, these predict metastasis with high accuracy (cross-validated accuracy of 87-89%). In this study, we have observed single disseminating EMT CSCs in human oral cancer specimens, and these are highly predictive of metastatic disease.

    1. Cancer Biology
    2. Medicine

    Tissue-derived exosome proteomics identifies promising diagnostic biomarkers for esophageal cancer

    Dingyu Rao, Hua Lu ... Defa Huang
    Research Article

    Esophageal cancer (EC) is a fatal digestive disease with a poor prognosis and frequent lymphatic metastases. Nevertheless, reliable biomarkers for EC diagnosis are currently unavailable. Accordingly, we have performed a comparative proteomics analysis on cancer and paracancer tissue-derived exosomes from eight pairs of EC patients using label-free quantification proteomics profiling and have analyzed the differentially expressed proteins through bioinformatics. Furthermore, nano-flow cytometry (NanoFCM) was used to validate the candidate proteins from plasma-derived exosomes in 122 EC patients. Of the 803 differentially expressed proteins discovered in cancer and paracancer tissue-derived exosomes, 686 were up-regulated and 117 were down-regulated. Intercellular adhesion molecule-1 (CD54) was identified as an up-regulated candidate for further investigation, and its high expression in cancer tissues of EC patients was validated using immunohistochemistry, real-time quantitative PCR (RT-qPCR), and western blot analyses. In addition, plasma-derived exosome NanoFCM data from 122 EC patients concurred with our proteomic analysis. The receiver operating characteristic (ROC) analysis demonstrated that the AUC, sensitivity, and specificity values for CD54 were 0.702, 66.13%, and 71.31%, respectively, for EC diagnosis. Small interference (si)RNA was employed to silence the CD54 gene in EC cells. A series of assays, including cell counting kit-8, adhesion, wound healing, and Matrigel invasion, were performed to investigate EC viability, adhesive, migratory, and invasive abilities, respectively. The results showed that CD54 promoted EC proliferation, migration, and invasion. Collectively, tissue-derived exosomal proteomics strongly demonstrates that CD54 is a promising biomarker for EC diagnosis and a key molecule for EC development.