Monitoring the Cell Cycle of Tumor Cells in Mouse Models of Human Cancer

Abstract

Cell division is obligatory to tumor growth. However, both cancer cells and noncancer cells in tumors can be found in distinct stages of the cell cycle, which may inform the growth potential of these tumors, their propensity to metastasize, and their response to therapy. Hence, it is of utmost importance to monitor the cell cycle of tumor cells. Here we discuss well-established methods and new genetic advances to track the cell cycle of tumor cells in mouse models of human cancer. We also review recent genetic studies investigating the role of the cell-cycle machinery in the growth of tumors in vivo, with a focus on the machinery regulating the G₁/S transition of the cell cycle.

The active cell cycle has four distinct phases—G₁, S, G₂, and M sequentially (Fig. 1A). G₁ and G₂ are gap phases with G₁ accounting for almost half of a complete cycle in normal cells. Replication of DNA occurs during S phase, and division of DNA and cell contents occurs during M phase, which itself consists of prophase, metaphase, anaphase, and telophase defined by stages of chromosome separation. The transition between each phase of the cell cycle is driven by the activity of cyclin-dependent kinases (CDKs), which are each in turn activated by cell phase–specific cyclin activators. Checkpoints ensure that each phase is correctly completed before moving to the next phase. Perturbations in the cell-cycle machinery and these checkpoints are linked to a number of defects and diseases, including cancer (for reviews, see Pack et al. 2019; Lara-Gonzalez et al. 2021; Matthews et al. 2022; da Costa et al. 2023). As an example, the so-called RB pathway controls the G₁/S transition of the cell cycle (Fig. 1B). In this pathway, RB and the two RB-like proteins p107 and p130 are G₁ checkpoint regulators and proliferation repressors that are canonically thought to act through inhibition of the activity of E2F transcription factors. During normal cell cycles, RB is inactivated by phosphorylation by cyclin D-CDK4/6 and cyclin E-CDK2 kinase complexes in response to pro-proliferation factors. These kinases are themselves under the control of small-cell-cycle inhibitors of the p16 and p21 families. Disruption of this pathway is frequently observed in human tumors (for reviews, see Kent and Leone 2019; Rubin et al. 2020; Fassl et al. 2022).

Figure 1.

Cell-cycle progression is driven by cyclin-dependent kinase (CDK) activity and regulated by cell-cycle checkpoints. (A) Schematic of cell-cycle phases with corresponding CDK-cyclin activity. Boxes indicate cell-cycle checkpoints and criteria for progression. (B) The RB1 pathway regulating the G₁/S transition.

Many cells in the adult body exist in a nondividing state and therefore do not undergo this division cycle, either because they exist in terminally differentiated or senescent states, or because they are in a reversible G₀ quiescent state. Under proliferative conditions, these normally quiescent cells can revert to a proliferative state, and if the pro-proliferative signals are sustained, as is the case with certain oncogenic mutations, this can lead to the expansion of precancerous or cancerous cells. Other cells of origin for cancer include stem cells or progenitors that are already cycling and that cycle uncontrollably upon genetic or epigenetic alterations (for reviews, see Sutherland and Visvader 2015; Bajaj et al. 2020). In tumors that have already formed, not all cancer cells divide at the same rate. For example, quiescence in populations of cancer stem cells may be critical for the ability of tumors to sustain long-term growth, evade the immune system, become resistant to therapy, and metastasize (discussed in Batlle and Clevers 2017; Goddard et al. 2018; Smith and Macleod 2019; Massagué and Ganesh 2021).

While our understanding of cancerous cell division has grown, research interest remains active in dissecting the role of different states of disrupted division in all stages of tumor development from initiation to metastasis. Additionally, as a process with high stakes for error and a heavy regulation process as a result, cell cycle represents a key pathway for identification of vulnerabilities that can be exploited in the therapeutic context (for reviews, see Suski et al. 2021; Fassl et al. 2022; Liu et al. 2022; da Costa et al. 2023). Therefore, methods for isolating and characterizing different phases and states of cell-cycle dysregulation are vital for the progression of cancer research. Here we will first discuss methods that have been used in the past decades to measure the cell-cycle state of cells in tumors, including human tumors and tumors from genetically engineered mouse models of human cancer. Next, we will summarize recent genetic advances allowing a more precise and/or a more quantitative characterization of cell-cycle states in mouse tumors. Finally, we will use recent work focusing on the analysis of the RB pathway in mutant mice to highlight new insights into the mechanisms of cell-cycle regulation, including how they pertain to the response of tumor cells to small molecule inhibitors of the cell-cycle machinery.

CLASSICAL METHODS TO MEASURE THE CELL CYCLE OF TUMOR CELLS IN HUMANS AND MICE

A number of antibodies against components of the cell-cycle machinery have been used over the years to determine the cell-cycle status of cells in human tumors. Some of these antibodies can also be used for immunostaining of frozen or paraffin-embedded sections of mouse tumors. These antibodies can be used in combination with antibodies marking specific populations of cells in tumors (including subpopulations of cancer cells, immune cells, and others), either on tissue sections to retain spatial information or by flow cytometry for bulk analysis.

Ki-67 (sometimes Ki67, encoded by the MKI67 gene, for marker of proliferation Ki-67) is a protein that normally contributes to chromosome clustering during mitosis (Cuylen et al. 2016; Cuylen-Haering et al. 2020). Ki-67 is detectable during all phases of the active cell cycle (G₁, S, G₂, and M) except for G₀. Ki-67 expression is used to determine the fraction of cycling cells in a tumor or tumor cell population, and the Ki-67 index serves as a clinical marker of tumor progression (Iatropoulos and Williams 1996; Brown and Gatter 2002; Guzinska-Ustymowicz et al. 2009; Miller et al. 2018; Uxa et al. 2021). However, expression of Ki-67 cannot easily distinguish between different stages of cell-cycle progression (especially in G₁, S, and G₂, before chromosomes condensate). In addition, lack of Ki-67 detection cannot distinguish between different forms of cell-cycle arrest as quiescent, senescent, and differentiated cells are all negative for Ki-67 expression. Still, Ki-67 expression remains frequently used in the analysis of cell proliferation in both human and murine tumors.

Markers used to monitor progression through S phase, including expression of the replication machinery factors PCNA and MCM2 (Iatropoulos and Williams 1996; Guzinska-Ustymowicz et al. 2009), have historically not been used as often as Ki-67 despite providing a greater degree of cell phase specificity. In mouse tumor models, these markers can be replaced by incorporation of the thymidine analog BrdU during DNA synthesis; BrdU can be injected intraperitoneally in mice with tumors to label cells replicating their DNA in vivo (Leif et al. 2004; Wojtowicz and Kee 2006). The use of various synthetic thymidine analogs (i.e., iodo-deoxyuridine IdU, chloro-deoxyuridine CldU, and ethinyl-deoxyuridine EdU) can further help perform simple lineage-tracing experiments by monitoring successive rounds of cell division or cell turnover (Fig. 2; see examples in Teta et al. 2007; Harris et al. 2018; Podgorny et al. 2018; Ouadah et al. 2019).

Figure 2.

Lineage tracing using thymidine analogs. A combination of the thymidine analogs EdU and BrdU can be used to identify stem-like populations of cells following two rounds of injury. In this example, cells that replicate after both injuries and stain double positive for both BrdU and EdU following harvest are considered to be stem-like precursor cells with regenerative potential. (Analysis grid based on data in Ouadah et al. 2019.)

Phosphohistone H3 (PH3 or PHH3) is commonly used to mark G₂/M cells. Histone H3 phosphorylation at serine 10 and serine 28 is closely correlated with chromosome condensation initiation during early prophase. Metaphase chromosomes are heavily phosphorylated, and interphase cells do not stain as intensely for PH3 as mitotic cells (Colman et al. 2006; Tapia et al. 2006). A high mitotic index defined by PH3 can be used in the clinic as a prognostic marker for lower survival rates (e.g., Kim et al. 2007). PH3 serves as a more quantitative marker of cell-cycle activity than some of the markers discussed above as fewer cells will stain positive owing to the short duration of mitosis, allowing for easier single-cell distinction in a field of dense tumor cells.

In the past few years, decreased costs of sequencing and novel protocol for single-cell RNA sequencing have enabled an increasing number of analyses of bulk tumors at the single-cell level. From these analyses, it is possible to identify cell-cycle signatures that can readily inform the proliferative status of cells in tumors (e.g., Levitin et al. 2019; Hsiao et al. 2020). Similarly, recent advances in spatial transcriptomics and multiplexed imaging approaches (e.g., Keren et al. 2019; Black et al. 2021; Moses and Pachter 2022) suggest that a number of markers routinely used to measure the cell cycle of tumor cells are likely to be combined for a more precise integration of cell-cycle states in human and mouse tumors.

GENETIC TOOLS TO MONITOR THE CELL CYCLE IN VIVO

In addition to markers discussed above, several groups have developed new alleles for cell-cycle analyses in vivo, including knockin reporter genes into cell-cycle-regulated genes or transgenic mice with the promoter of cell-cycle genes driving the expression of a fluorescent reporter.

For instance, E2F transcription factors are key regulators of the G₁/S transition of the cell cycle (Fig. 1), and E2F-based reporters can be used to track progression from G₁ to S phase. E2F1 is a major transcriptional activator of E2F genes, and its own transcription is regulated in a positive feedback loop (Johnson et al. 1994). An E2F1-luc transgenic mouse has been generated and luciferase activity in this model can be used to monitor cell proliferation in mouse tissue and in tumors in vivo (Uhrbom et al. 2004). The RB family member p107 (encoded by Rbl1 in mice) is also an E2F target. Transgenic mice expressing GFP (green fluorescent protein) inserted into a bacterial artificial chromosome (BAC) containing the mouse Rbl1 gene showed GFP expression paralleling endogenous p107 expression, with increased expression upon loss of RB and consequent E2F activation (Burkhart et al. 2008). A similar transgenic BAC knockin approach with the Rb1 gene itself (coding for RB) and a transgenic model with an Rb1 promoter both showed that Rb1 transcription is regulated by RB/E2F complexes (Agromayor et al. 2006; Burkhart et al. 2010).

At later stages of the cell cycle, a cyclin B1-GFP fusion reporter in a lentiviral vector was used to develop a transgenic mouse in which replicating cells in S/G₂/M are marked, which has been used to assess the proliferation of hepatocytes in vivo (Klochendler et al. 2012).

The scaffolding protein Anillin is located in the nucleus in late G₁, S, and G₂, in the cytoplasm and cell cortex in early M, and in the contractile ring and midbody during and immediately upon cytokinesis. A transgenic mouse expression of fusion between GFP and Anillin allows direct visualization of these cell-cycle stages and cell-cycle structures by high-resolution microscopy in live cells in vivo (Hesse et al. 2012).

A knockin allele with the gene coding for the RFP (red fluorescent protein) reporter into the MKI67 gene (Ki67-RFP allele) has also been generated to investigate cycling and quiescent populations of cells in the intestinal epithelium (Basak et al. 2014). This allele has been used thus far to characterize the proliferation of cells in normal mouse tissues (Morcos et al. 2017; Kretzschmar et al. 2018) but could easily be applied to monitor the proliferation of various cell types in tumors, including cancer stem cells that may exist in quiescent states or various immune cell populations. A limitation of this allele and other simple reporter alleles, however, is that they do not allow for multigenerational tracking of the cell-cycle status of single cells in tumors.

As discussed above, injections of BrdU or BrdU analogs can help perform simple lineage-tracing experiments related to the cell-cycle status. Similar experiments can be performed using H2B-GFP transgenic mice (Foudi et al. 2009; Greco et al. 2009). In these mice, GFP is fused to Histone 2B and thus localized to the DNA when expressed, allowing for simple visualization and quantification (Fig. 3A). Expression of H2B-GFP under a TetO inducible promoter and a double transgenic system with rtTA expressed from a tissue-specific promoter can rapidly label cells of interest with spatial and temporal specificity following a doxycycline pulse. GFP intensity can then provide a readout of division “counts” as H2B-GFP is diluted upon subsequent divisions—similar to what is done with the intracellular fluorescent label CFSE (carboxyfluorescein diacetate succinimidyl ester) ex vivo or in vivo after reinjection of cells labeled ex vivo (Lyons et al. 2013). Some disadvantages of the H2B-GFP approach include the two-transgene system that requires crossing to an additional cancer model, as well as possible leakiness of the promoters, which may make it more difficult to track specific cell populations (Challen and Goodell 2008). Still, combining this approach with single-cell barcoding systems may allow for cell-cycle tracking of distinct clonal populations in vivo (Oren et al. 2021).

Figure 3.

In vivo genetic approaches for identifying cell phase and proliferation. (A) Simple lineage tracing using H2B-GFP expression with spatial and temporal regulation. In this example, a tissue-specific (Tissue X) promoter is used to express rtTA. Following doxycycline pulse, rtTA is able to drive H2B-GFP expression in the tissue of interest, which could be a tumor. After timepoint collection, relative GFP intensity acts as a proxy for counting replication rounds with high GFP indicating few rounds and low GFP indicating many rounds. (B) Overview of the FUCCI (fluorescence ubiquitin cell-cycle indicator) two-color cell-cycle marker system including transgenes from the original FUCCI and FUCCI2 systems.

One of the most widely used systems to track cell-cycle progression is the FUCCI system. This system allows distinction between live cells in the G₁ and S/G₂/M phases of the cell cycle by dual-color imaging. The original FUCCI system marked cells residing in G₁ phase with red fluorescence (fluorescent protein fused to degron derived from the CDT1 factor), while cells in S/G₂/M were labeled in green (fluorescent protein fused to degron derived from Geminin expression); at the G₁/S transition, cells appeared yellow (both probes being expressed) (Fig. 3B). Since its first implementation in 2008, the FUCCI system has been improved and adapted, including for studies in animals such as mice (for review, see Zielke and Edgar 2015; see also Sakaue-Sawano et al. 2017; Zambon et al. 2020). “Fucci2” transgenic mice expressing FUCCI sensors were generated with mCherry (red fluorescence) and mVenus (green) under the control of the Rosa26 promoter, either for direct expression or for conditional expression upon Cre-mediated recombination (Abe et al. 2013). A similar approach (“Fucci2a”) used the Thosea asigna virus 2A (T2A) self-cleaving peptide between the two reporters (Mort et al. 2014). Another study generated independent transgenes for “FucciG₁” and “FucciS/G₂/M” analyses (Yo et al. 2015). Similar to other systems, these transgenic mice need to be crossed to other alleles to track the cell cycle of tumor cells in vivo, which remains a limitation.

Overall, with the new multiplexed imaging platforms available to investigators interested in the cell cycle of tumor cells, the main current advantage of these mouse alleles is the possibility to track the evolution of cell-cycle states over time. While it requires more time and effort to set up inducible and tractable systems, temporal changes in cell-cycle states may be critically important in our understanding of normal tumor progression or in response to therapy. With the development of barcoding systems that can be linked to multiplexed imaging approaches (e.g., Gutierrez et al. 2021; Lomakin et al. 2022; Rovira-Clavé et al. 2022), it is likely that this advantage of genetic reporters may become more limited in the future, at least in the counting of cell divisions.

DISSECTING THE ROLE OF THE RB TUMOR SUPPRESSOR IN GENETICALLY ENGINEERED MOUSE MODELS

A number of recent studies have also directly altered the expression or the activity of factors in the cell-cycle machinery in mouse tumors to determine the role of these factors in the growth of tumors. We will give some examples, focusing on the RB tumor suppressor to show what can been learned from this type of study at the interface between the regulation of cell-cycle control and the development of tumors.

The retinoblastoma tumor suppressor RB acts as a negative regulator of cell-cycle progression in G₁ (Fig. 1B). Not surprisingly, loss of RB function in mice results in a number of hyperproliferative phenotypes that lead to the development of tumors, including in the pituitary and the thyroid glands of these Rb1 mutant mice (Jacks et al. 1992; Maandag et al. 1994; Vooijs et al. 1998; Vooijs and Berns 1999). The RB family members p107 and p130 can functionally compensate for loss of RB, as shown by enhanced tumor development in double- or triple-knockout mice, including retinoblastoma development that is not observed in RB-only mutant mice (Robanus-Maandag et al. 1998; Chen et al. 2004; Dannenberg et al. 2004; Zhang et al. 2004; Ajioka et al. 2007; Viatour et al. 2008, 2011; Lázaro et al. 2017). More surprisingly, while loss of the entire RB family leads to an inability for fibroblasts to arrest in G₁ in culture (Dannenberg et al. 2000; Sage et al. 2000), it is still compatible with cell-cycle arrest in G₁ in other cell types, including neuronal cells in culture and hepatocytes in vivo (Wirt et al. 2010; Ehmer et al. 2014). The basis of this RB family-independent arrest in G₁ is not fully understood but, in the case of hepatocytes in the liver, may relate to a dominant role for organ size signals via the Hippo signaling pathway and its downstream effector the YAP transcriptional regulator (Ehmer et al. 2014). In the thymus, regulation of the expression of the differentiation factor FOXN1 by E2F is critical for proliferation and organ growth (Garfin et al. 2013), indicating that even E2F target genes that are not directly related to the regulation of proliferation can be essential for progression from G₁.

Knockin approaches have also been used to investigate how single amino acids or small domains can contribute to the mode of action of RB pathway member function in vivo. For instance, transgenic mice with active forms of RB in which serine/threonine residues normally phosphorylated by cyclin/CDK kinases have been substituted to alanine have been shown to surprisingly develop breast cancer, possibly because expression of active RB may not just slow the cell cycle but also promote the survival of cells via cell-cycle-independent mechanisms, including by regulating differentiation states (Jiang and Zacksenhaus 2002). These mice also display additional body-wide signs of premature aging (Jiang et al. 2022).

As mentioned above, RB and its family members p107 and p130 repress cell-cycle progression via regulation of E2F activity. Knockout experiments in mice have confirmed these functional interactions between RB family members and E2F family members in tumor development (e.g., Yamasaki et al. 1998; Parisi et al. 2007). However, a number of more recent experiments have painted a more complex picture. First, while cells in culture require the E2F1/2/3a transcriptional activators, some cells in vivo (e.g., in the developing lens) can still cycle in the absence of these transcription factors (Wenzel et al. 2011; Liu et al. 2015), further emphasizing the need to conduct cell-cycle studies in vivo to more accurately understand how cells control their cycle. Mice with a mutant form of RB that is severely compromised for its ability to interact with the transactivation domain of E2F factors have been generated (Cecchini et al. 2014). Surprisingly, these mice do not spontaneously develop tumors, suggesting that loss of general E2F repression by RB alone is not sufficient to predispose to cancer development (Cecchini et al. 2014), but this mutation can promote tumor development in the context of other cancer mutations (Thwaites et al. 2017, 2019), indicating that decreased binding to E2F is still an important tumor suppressor function of RB. RB can also form specific complexes with E2F1 (via the carboxyl terminus of RB), and a mutation in RB that specifically disrupts this unique interaction results in the development of lymphoma in mice, suggesting that this RB/E2F1 interaction is important in some contexts to suppress tumor development (Ishak et al. 2016).

Together, these experiments in mouse models of human cancers have largely validated biochemical studies in cells in culture but have also uncovered new and surprising phenotypes indicative of “noncanonical” functions of the RB pathway in the regulation of the cell cycle and tumorigenesis, including, as mentioned above, through regulation of differentiation, but also by regulation of chromatin states and chromosome structure (e.g., for reviews, see Viatour and Sage 2011; Dick et al. 2018).

The role of RB as a tumor suppressor at the time of cancer initiation has been extensively documented in mouse models (e.g., Meuwissen et al. 2003; Berman et al. 2008; Walkley et al. 2008). However, except for retinoblastoma, osteosarcoma, and small-cell lung cancer (SCLC), RB is most often lost at later stages of cancer development in humans, and an important aspect of future studies will include modeling its role later during tumorigenesis. This can be achieved by developing alleles of RB that allow its inactivation in a temporal manner that is compatible with other genetic alterations that initiate cancer development. For example, with a large number of mouse models of cancer being initiated with the Cre/Lox system, a Flp/Frt-based conditional allele of Rb (Vooijs et al. 1998) or a doxycycline-inducible shRNA allele to knock down RB (Doan et al. 2021) could prove useful to investigate how loss of RB function contributes to cancer progression and metastasis. Combining these different genetic strategies with different cancer genes of interest may help deconstruct key mechanisms of cancer evolution.

INVESTIGATING THE MODE OF ACTION OF CYCLIN D-CDK4/6 IN CANCER GENETIC MODELS IN THE CONTEXT OF THE DEVELOPMENT OF CLINICALLY RELEVANT CDK4/6 INHIBITORS

The approval of CDK4/6 inhibitors such as palbociclib for the treatment of breast cancer has led to a renewed interest in the biology of cyclin D-CDK4/6 complexes in normal and cancer cells (Sherr et al. 2016). A number of genetically engineered alleles have been developed to investigate the biology of cyclins and CDKs in the past 20 years (for reviews, see Barbacid et al. 2005; Hydbring et al. 2016; Fassl et al. 2022), and these mouse models can be used to probe how cancer cells respond to CDK4/6 inhibitors in vivo.

It has been known for 20 years that cyclin D1 knockout mice are resistant to the development of breast cancer initiated by oncogenes such as RAS (Yu et al. 2001). Knockin experiments next demonstrated that this requirement for cyclin D1 likely stems from its associated CDK4/6 kinase activity (Landis et al. 2006; Yu et al. 2006). These experiments and similar experiments in mice with lung cancer (Puyol et al. 2010) have provided strong support for the clinical trials with CDK4/6 inhibitors and for the clinical use of these molecules in patients with breast cancer.

These cyclin D1 knockout mice have been used more recently to explore the response of cancer cells to CDK4/6 inhibition in various contexts. For instance, in a mouse model of breast cancer driven by active Wnt signaling, knockout of cyclin D1 leads to increased levels of the immune checkpoint molecule PD-L1, validating experiments in culture showing that cyclin D-CDK4/6 kinase activity controls the protein stability of PD-L1 (Zhang et al. 2018), but also suggesting that cancer cells with high levels of cyclin D may be less responsive to cytotoxic T cells unless they are treated with CDK4/6 inhibitors. In a model of liver cancer, knockout of cyclin D1 in hepatocytes protected obese/diabetic mice from tumor development but the presence or the absence of cyclin D1 had no consequence for tumor development in lean/nondiabetic mice. Accordingly, this cyclin D1 dependency correlated positively with tumor inhibition following palbociclib treatment, suggesting that inhibition of cyclin D1-CDK4/6 may more selectively benefit obese/diabetic patients with liver cancer (Luo et al. 2020). The use of mouse genetics in this type of approach is powerful as it allows investigators to tease out effects of small molecule inhibitors on cancer cells and/or other cells in the tumor microenvironment, including T cells, which may aid in the development of combination therapies.

In this context, it is surprising that other mouse alleles have not been used yet to explore how other members of the G₁/S regulatory networks may contribute to the response of cells to CDK4/6 inhibitors. For example, the E3 ligase adaptor AMBRA1 was recently identified as a negative regulator of the stability of D-type cyclins and a regulator of the response of cells to CDK4/6 inhibition in culture (Chaikovsky et al. 2021; Maiani et al. 2021; Simoneschi et al. 2021), but these studies did not test AMBRA1's role in the response of cancer cells to CDK4/6 inhibition in vivo in genetically engineered mice with an active immune system. Similarly, a number of studies in culture, with xenografts, and in patients with cancer have identified molecules implicated in the resistance of cancer cells to CDK4/6 inhibition (Li et al. 2018, 2022; Walter et al. 2019; Álvarez-Fernández and Malumbres 2020; Costa et al. 2020; Wander et al. 2020; Freeman-Cook et al. 2021; Asghar et al. 2022), including MYC and cyclin E-CDK2. However, little has been done thus far using mouse alleles for these genes to model or recapitulate resistance, and this should be an area of future investigation as these mice provide a defined system to investigate molecular and cellular mechanisms of drug response and resistance. A number of other inhibitors of the cell cycle, including WEE1, AURK (Aurora kinase), or CHK1 (checkpoint kinase 1) inhibitors could be similarly studied in genetically engineered mouse models. For example, the idea that cancer cells deficient for G₁/S checkpoint proteins may be more sensitive to inhibition of checkpoint proteins in S and G₂/M such as AURK and CHK1 has been tested in a mouse model of SCLC (cancer cells mutant for RB, p53, and with hyperactive MYC) (Dammert et al. 2019).

A great example of the idea that cell-cycle therapeutics may be used differently based on our understanding of mutations in cell-cycle genes comes from SCLC: SCLC cells are very often mutant for RB and thus unresponsive to CDK4/6 inhibitors. However, treatment with the CDK4/6 inhibitor trilaciclib can protect against myelosuppression induced by the chemotherapy used to treat SCLC tumors (Hart et al. 2021). It is likely that the availability of loss-of-function and gain-of-function alleles for cell-cycle genes will be key to our ability to comprehend the role of cell-cycle states in specific populations of tumor cells during tumor progression as well as to identify key targets of molecules targeting cell-cycle regulators in complex tumor microenvironments.

CONCLUDING REMARKS

Mouse models of human cancer remain an important tool to dissect fundamental mechanisms of tumorigenesis and serve as preclinical models to investigate how cancer cells may respond to therapy. Cell-cycle states are a critical aspect of tumor development and must be carefully assessed if we want to understand how tumors grow. While a number of analyses can be performed on human samples, these samples are not always available and offer only a snapshot of tumor biology. Experiments in mice provide unlimited samples and the possibility for longitudinal studies. As discussed in this article, it is likely that future studies with genetically engineered mouse models of human cancer will include multiomics platforms and multiplexed imaging systems, which will allow for more markers to be analyzed at the same time in more samples. New markers of cell-cycle progression such as cell size will also likely be incorporated in future analyses (e.g., Xie and Skotheim 2020; Rovira-Clavé et al. 2022). Senescence is also a distinct cell-cycle state with specific markers and for which therapies are being developed (e.g., Prasanna et al. 2021; Schmitt et al. 2022; Wang et al. 2022). These new approaches and new markers can be combined with new mouse alleles, especially alleles with controllable expression of key cell-cycle regulators (e.g., Suski et al. 2022). The next decade of cell-cycle analyses using mouse models of human cancer promises to uncover critical new aspects of tumor development.

COMPETING INTEREST STATEMENT

The authors declare no competing interests.