Insights into the aberrant CDK4/6 signaling pathway as a therapeutic target in tumorigenesis

Abstract

The recent findings advance our knowledge for the prevention of the premature activation of the major oncogenic pathways including MYC and the cyclin D-cyclin-dependent kinases 4 and 6 (CDK4/6) axis. D-type cyclins are frequently deregulated in human cancer and promote cell division in part through activation of CDK4/6. Therefore, the activation of the cyclin D-CDK4/6 axis stimulates cell proliferation and cancer progression, which represents a unique therapeutic target. However, we have shown that inhibition of CDK4/6 upregulates protein levels of RB1 and CDK6 for acquisition of drug resistance to CDK4/6 inhibitors. Here, we review new progress in the control of cyclin D-dependent cancer cell cycle and proliferation, along with identification of novel E3 ligase for the stability of cyclin D. Cullin4-RING E3 ligase (CRL4)^AMBRA1 complex plays a critical role in regulating D-type cyclins through their protein destabilization to control S phase entry and maintain genomic integrity. We also summarize the strategy for inhibition of the cyclin D-associated kinases CDK4/6 and other potential cell cycle regulators for targeting cancer with altered cyclin D expression. We also uncover the function of CK1ε as an effective target to potentiate therapeutic efficacy of CDK4/6 inhibitors. Moreover, as the level of PD-L1 is considered in the severe clinical problem in the patients treated with CDK4 inhibitors, we assume that a therapeutic combination using PD-L1 immunotherapy might lower the development of drug resistance and targeting cyclin D will likely inhibit tumor growth and overcome resistance to cyclin D-associated CDK4/6 inhibitors.

1. Introduction

Proliferating cells require high demands of nucleotides, amino acids, and lipids for DNA replication in S phase and subsequent cell division (Bensaad et al., 2006; Gordan, Thompson, & Simon, 2007; Hanahan & Weinberg, 2011; Manning & Cantley, 2007; Wang et al., 2017). This cell division is regulated by key protein components in the cell cycle machinery for the maintenance of the genome without error (Nakayama & Nakayama, 2006). The replication of the genome is regulated by two main signaling pathways including the MYC signaling and the cyclin D/cyclin-dependent kinases (CDKs)/retinoblastoma protein (RB) pathway, which are both deregulated in human cancers with increased genomic instability (Otto & Sicinski, 2017). Therefore, the well-maintained proteins in the cell cycle machinery are essential for proper cellular functions and avoidance of human disorders such as cancer. As such, cyclin D and its associated CDKs have appeared as potential therapeutic targets. To this end, the ubiquitin-proteasome system (UPS) is the main mechanism for the maintenance of intracellular proteins by the serial action of enzymes, linking the ubiquitinated proteins to proteasomal degradation. The UPS system participates in many critical biological functions, such as antigen presentation, transcription and the cell cycle (Bassermann, Eichner, & Pagano, 2014; Ciechanover, 1994; Liu et al., 2015; Muratani & Tansey, 2003). Thus, understanding the physiological functions of UPS-mediated protein maintenance in the machinery of cell cycle would also be beneficial in effectively treating cancer.

Cyclin proteins including D-type cyclins (cyclins D1, D2 and D3) are periodically expressed in a cell cycle-dependent manner. They support cell proliferation through activation of CDK4 and CDK6, which subsequently inactivate the RB family member proteins in a phosphorylation-dependent manner (Malumbres & Barbacid, 2009). The activation of CDK is associated with the sequential completion of DNA replication for DNA damage checkpoint in response to defects in the mitotic spindle to determine cell division (Besson, Dowdy, & Roberts, 2008; Sherr & Roberts, 2004). The activity of CDK is also coordinated with the abundance of the regulatory cyclin subunits, followed by phosphorylation events mediated by the catalytic CDK subunit (Besson et al., 2008; Sherr & Roberts, 2004). Hence, activation of cyclin D-CDK4 or CDK6 facilitates cell cycle progression by the phosphorylation of its substrates, such as RB (Fig. 1). Moreover, D-type cyclins have non-catalytic functions that interact with diverse transcription factors leading to the transcriptional regulation of several genes involved in proliferation and differentiation (Musgrove, Caldon, Barraclough, Stone, & Sutherland, 2011). Cyclin-dependent kinases play critical roles in the stability of proteins involved in the regulation of cell cycle progression (Hydbring, Malumbres, & Sicinski, 2016; Malumbres & Barbacid, 2005). The cyclin D-CDK4/6 complexes also regulate downstream substrates with the function in mitochondrial performance, centrosome duplication, cell growth, adhesion and motility (Musgrove et al., 2011; Nelsen et al., 2005; Wang et al., 2006). These functions, in addition to the role of D-type cyclins in facilitating DNA repair and cytoskeletal modeling (Choi & Anders, 2014; Jirawatnotai, Hu, Livingston, & Sicinski, 2012; Welsh et al., 2001), implicate suitable therapeutic modules for targeting D-type cyclins in cancer therapies.

Fig. 1.

A schematic diagram of cell cycle and structure of the cyclin D proteins. (A) The progression of cell cycle is dependent on cyclin-dependent kinases (CDKs), which are activated and form holoenzymes with their regulatory subunits, the indicated cyclins. RB protein plays a key role in regulating G1 checkpoint and inhibiting S phase entry through repressing E2F transcriptional activity that is required for the transition from G1 to Sphase. The CDK4/cyclin D complex phosphorylates RB in mid-G1, while CDK2/cyclin A or CDK2/cyclin E phosphorylates RB at the G1 to S transition. CDK2/cyclin A complex is catalytically active during S phase, while CDK2/cyclin A and CDK2/cyclin B complexes are active during G2 phase. (B) The cyclin D proteins including cyclin D1, cyclin D2 and cyclin D3 are identical in humans for the first 240 amino acids but diverge at the carboxy-terminal that contains protein-protein interaction domains. All D-type cyclins include RB-binding LxCxE motif and C-terminal PEST domain. The PEST domain contains conserved threonine residue subjected to phosphorylation to promote cyclin D nuclear export and/or triggers proteasomal degradation. The LLxxxL is leucine-rich motif of the cyclin D1, which binds an LxxLL motif in the steroid receptor coactivator SRC1 (Zwijsen, Buckle, Hijmans, Loomans, & Bernards, 1998).

This review emphasizes the function of cell cycle regulators with a special focus on the role of D-type cyclins due to the weight of their function evidenced in human cancer with drug resistance. We also provide insights into our recently identified cell cycle regulators involved in the PI3K/AKT signaling (Liu et al., 2014), RAF/ERK signaling (Wan et al., 2017), Hippo/YAP signaling (Kim et al., 2019), and cell metabolism pathway (Liu et al., 2021) in the control of cell cycle progression and their relevance in human cancer. We also show that treatment with CDK4/6 inhibitor can change the protein level of RB1, CDK6, and PD-L1, leading to cancer cell resistance to CDK4/6 inhibitor after escaping from cell cycle checkpoints and immune surveillance. Finally, we summarize the recent progress in developing new inhibitors for targeting cyclin D-CDK4/6 as potential cancer therapeutics.

2. Aberrant cell cycle checkpoints cooperate with cancer progression

The SKP2/IDH1/2 checkpoint.

Differentiation of normal cells depends on supplying energy from the glucose metabolism or oxidative phosphorylation in the tricarboxylic acid (TCA) cycle. However, the proliferation of cells enforces aerobic glycolysis or the “Warburg effect” to supply more energy even in the presence of oxygen (Hanahan & Weinberg, 2011; Vander Heiden, Cantley, & Thompson, 2009; Warburg, 1956). Although quiescent cells have been shown to metabolize glucose differently than proliferating cells in cancer, the utilization of glucose metabolism either through oxidative phosphorylation or glycolysis determined in each cell cycle process had not been investigated. Ubiquitination-dependent degradation of cyclins regulates transition of the cell cycle, where SCF (SKP1-Cullin1-F-box protein) E3 ligases complex play an important role in this transition (Suzuki, Ohashi, & Kitagawa, 2013). For example, the S phase kinase-associated protein 2 (SKP2) is an F-box protein component of the SCF complex that targets p27^KIP1, a cyclin-dependent kinase inhibitor, for proteasomal degradation in the ubiquitin-dependent pathway. We recently found that mammalian cells execute the TCA cycle in G1 phase, while predominantly utilize glycolysis in S phase. Since IDH1/2 enzymes play a critical role in the TCA cycle, we hypothesized that their protein stability might be regulated upon the cell cycle process. Notably, we found that depletion of Cullin1, but not Cullin3, Cullin4A, or Cullin4B, led to accumulation of IDH1 and IDH2. We also found that SKP2 could bind to and regulate IDH1 protein stability through its ubiquitination and degradation in the UPS system, in which the elevated level of SKP2 in cancer could drive IDH1 degradation leading to activation of glycolysis and tumorigenesis (Liu et al., 2021). Conversely, the inhibition of SKP2 could accumulate IDH1/2 protein and prevent metabolic shift between TCA cycle and glycolysis. The IDH1 and IDH2 proteins were also accumulated in Ccne1 and Ccna2 knockout MEFs, and CDK2 inhibition led to a reduction of IDH1 ubiquitination. More interestingly, during the G1/S transition, accumulation of cyclin E could activate CDK2 to phosphorylate IDH1 at Thr157. This phosphorylation is an essential step in the SCF^SKP2-dependent degradation of IDH1 protein upon cell cycle process. These findings revealed the mechanistic link between cancer cell metabolism and cell cycle, which was previously unknown.

The CDH1/cyclin A2/AKT axis.

Recent studies have reported essential functions for the Anaphase-Promoting Complex/Cyclosome (APC/C) in several cellular processes, including genome stability and tumorigenesis (Kozorovitskiy et al., 2006). The APC/C-mediated degradation of cell cycle components is regulated by two adaptor proteins, CDH1 or CDC20 (Nakayama & Nakayama, 2006; Peters, 2006; Pines, 2011) and is critical for the proper maintenance of mitotic cyclins and DNA replication factors in the M/G1 phases during cell cycle progression. On the other hand, AKT plays a critical function in cell survival, proliferation and metabolism (Hoxhaj & Manning, 2020; Song, Bode, Dong, & Lee, 2019). Hyperactivation of AKT involves human cancer progression (Manning & Cantley, 2007; Toker, 2008; Zoncu, Efeyan, & Sabatini, 2011) and is associated with chemo- and/or radio-therapeutic resistance (Luo, Manning, & Cantley, 2003). We found that AKT activation is altered during the cell cycle correlated with the protein abundance of CDH1 in several cancer cell lines (Liu et al., 2014). We determined four “RXL” cyclin A binding motifs in AKT1(Adams et al., 1996) and revealed that AKT is a CDK2/cyclin A substrate. Notably, AKT phosphorylation status reflects the expression pattern of cyclin A2, while acute depletion of cyclin A2 or CDK2 decreases AKT phosphorylation without changes on upstream kinases PDK1 and mTORC2 (Fig. 2). Interestingly, depletion of CDH1 increases cyclin A2 protein abundance and AKT phosphorylation, followed by cancer cell growth in vitro and tumor formation in vivo. These findings indicate that CDK2/cyclin A directly phosphorylates AKT1 at both S477 and T479 for activation during cell cycle progression for its oncogenic function (Liu et al., 2014). The results further suggest that AKT1-pS477/pT479 may be also regulated by CDK2/cyclin A, DNA-PK, or mTORC2 upon cell cycle progression, DNA damage stress, or stimulation with growth factors. This newly identified phosphorylation of AKT1 is required for AKT1 activation through interaction with mTORC2 for AKT1-pS473 promotion and or couple with cyclin A2 for overactivation of the cell cycle progression and maintaining pro-survival and oncogenic function (Liu et al., 2014).

Fig. 2.

An illustration of the molecular mechanism for regulation of cyclin D-CDK4/6 signaling. Growth factors and receptor tyrosine kinases (RTKs) trigger the Ras/BRAF/MAP kinase and PI3K/AKT pathways, which leads to accumulated D-type cyclins in the early G1 phase through regulation of transcriptional and posttranslational levels. PI3K-dependent AKT activation inhibits glycogen synthase kinase 3β (GSK3β), which plays a critical role in the phosphorylation and inactivation of cyclin D1 through its nuclear export and proteasomal degradation. Subsequently, SCF and APC/C complexes promotes cyclin D degradation in a phosphorylation-dependent manner (Lin et al., 2006; Pawar et al., 2010; Santra, Wajapeyee, & Green, 2009). On the other hand, APC/C^CDH1 ubiquitinates and degrades LATZ1/2 and BRAF proteins for deregulation of Hippo/YAP and MAP kinase signalings for transcriptional repression of E2F and cyclin D, respectively. The full activation of cyclin D-CDK4/6 holoenzymes and cyclin E–CDK2 complexes initiates the phosphorylation of RB and provokes its hyperphosphorylation, leading to release of E2F transcription factors from RB to stimulate the expression of target genes involved in enter S-phase entry.

The APC/C^CDH1/BRAF axis.

The RAF family members of protein kinases, including ARAF, BRAF and CRAF, play critical roles during tumorigenesis (Wellbrock, Karasarides, & Marais, 2004). BRAF mutants are shown in over 60% of melanoma and thyroid cancers (Davies et al., 2002), driving the activation of RAS/RAF/MEK/ERK signaling cascade for tumorigenesis (Bollag et al., 2012; Holderfield, Deuker, McCormick, & McMahon, 2014). Since the upstream signaling(s) for regulating BRAF kinase activity was largely unclear, valuable evidence for targeting BRAF could provide an efficient therapeutic strategy for treating cancers, including melanoma with drug-resistant properties due to BRAF^V600E mutation (Poulikakos & Rosen, 2011). Notably, we found that CDH1 inhibits the activation of ERK through BRAF degradation since APC/C^CDH1 is required for BRAF protein stability in melanoma cells. The BRAF sequence has four putative destruction (D)-box motif (RxxLx(2–5)N/D/E), which can be recognized by CDH1 for its proteasomal degradation, while the R671Q-BRAF mutant fails to interact with CDH1, leading to its stabilization (Fig. 2). On the other hand, in primary melanocytes, CDH1 suppresses BRAF through disrupting BRAF dimerization in APC/C-independent manner (Wan et al., 2017). In this study, we further demonstrated that depletion of Cdh1 and hyperactivation of Akt coactivates Braf/Erk signaling upon Pten loss in the mice. Of note, CDK-mediated phosphorylation of CDH1 reduces CDH1 interaction with the APC/C core complex for the inhibition of APC/C^CDH1 E3 ligase function (Fukushima et al., 2013; Lukas et al., 1999; Luo et al., 2003). Hence, we identified that CDH1 is a direct target of several kinases, including CDK4/cyclin D1 and ERK, in which inhibition of CDK4 and/or BRAF/MEK signaling could restore APC/C^CDH1 ligase activity for suppression of melanomagenesis.

The APC/C^CDH1/Hippo axis.

Hippo pathway plays a critical role in the regulation of organ size, tissue growth, and regeneration (Dong et al., 2007; Varelas, 2014; Yu, Zhao, & Guan, 2015), and it is dysregulated largely in various human cancers (Harvey, Zhang, & Thomas, 2013; Moroishi, Hansen, & Guan, 2015). The mammalian Hippo pathway includes MST1/2 kinases, WW45 scaffolding protein, LATS1/2 kinases, transcription coactivator YAP and TAZ. The MST1/2 activates LATS1/2 for the phosphorylation of YAP/TAZ for interaction with 14-3-3 in the cytoplasm for subsequent degradation through β-TRCP-dependent ubiquitination (Dong et al., 2007; Liu et al., 2010; Zhao, Li, Tumaneng, Wang, & Guan, 2010; Zhao et al., 2007).

In collaboration with the Yang group, we identified that phosphorylation of YAP (pS127 by LATS kinases)(Liu et al., 2010; Zhao et al., 2010, 2007) is reduced in G1 and S phases with increased YAP localization in the nucleus. Consistently, in CDH1-deficient cells, the YAP protein level is reduced while phosphorylation of YAP is significantly upregulated. Therefore, we have characterized APC/C^CDH1 as a regulator of tumor suppressor LATS kinases that inhibit YAP/TAZ transcription factors. APC/C^CDH1 could destabilize and degrade LATS1/2 kinases in the G1 phase of the cell cycle, leading to promotion in the G1/S transition by the increased YAP/TAZ activities for transcriptional activity of E2F1 (Fig. 2). Furthermore, CDH1 reduction did not increase the YAP phosphorylation in the Lats1/2^−/− cells, and the inhibition of CDK4/6 using palbociclib induces G1 phase arrest led to decrease in YAP expression and localization in the nucleus. Since the Hippo/Yap pathway controls organ size, we also revealed that loss of Cdh1 leads to increased Wts (the Drosophila ortholog of LATS1/2) protein levels, followed by a slight reduction of eye size. Taken together, these results suggest that APC/C^CDH1 promotes YAP activity by regulating LAST1/2 degradation, thereby controlling the Hippo pathway in organ size through promoting Wts degradation in Drosophila. Therefore, we linked the activity of the LATS-regulated YAP/TAZ axis to the cell proliferation in which APC/C^CDH1 promotes LATS1/2 degradation through the ubiquitin-proteasome pathway.

3. AMBRA1 destabilizes cyclin D protein

D-type cyclins are among the most frequently deregulated therapeutic targets in human cancer (Malumbres & Barbacid, 2009). Increased cyclin D promotes cell division by activating CDK4/6 for the subsequent phosphorylation and inactivation of the RB tumor suppressor that stimulates cell proliferation and cancer progression (Malumbres & Barbacid, 2009; Sherr, 1996). Previous studies indicated that SCF and APC/C complexes promote cyclin D degradation in a phosphorylation dependent manner (Lin et al., 2006; Pawar et al., 2010; Qie & Diehl, 2020; Santra et al., 2009), and recent important findings have further contributed to the underlying mechanism of cyclin D stability, through AMBRA1 (activating molecule in beclin-1-regulated autophagy) (Chaikovsky et al., 2021; Maiani et al., 2021; Simoneschi et al., 2021).

AMBRA1 is an essential factor in several cellular signaling through binding to partner proteins (Malumbres & Barbacid, 2009). AMBRA1 regulates cell proliferation and tumorigenesis (Cianfanelli, De Zio, et al., 2015), while lacking Ambra1 leads to embryonic lethality (Fimia et al., 2007). AMBRA1 interacts with the Cullin 4-RING E3 ligase (CRL4) complex to stabilize the autophagy proteins (Antonioli et al., 2014; Nazio et al., 2013), which in turn dephosphorylates and degrades the protooncogene c-MYC in a PP2A-dependent manner (Cianfanelli, Fuoco, et al., 2015). Although AMBRA1 had been known as a critical regulator for controlling cell proliferation, the underlying mechanism of AMBRA1 in the link with D-type cyclin protein stability had not been well addressed.

Simoneschi et al. identified that CRL4^AMBRA1 acts as an E3 ubiquitin ligase that targets all D-type cyclins, leading to defects in neural tubes in the nervous system (Simoneschi et al., 2021). Depletion of AMBRA1 diminished degradation of D-type cyclins, especially nuclear cyclin D1 in the S phase with RB hyperphosphorylation (Ser807/Ser811) and induction of E2F upon nutrient deprivation and genotoxic stress (Fig. 2). They stabilized and accumulated D-type cyclins through treatment of human cell lines with MG132 or MLN4924, an inhibitor of all CRLs. More importantly, they rescued the hyperproliferative and neuronal defect with abemaciclib, an FDA-approved CDK4/6 inhibitor (Simoneschi et al., 2021). Furthermore, they reported that cancer mutations in the D-type cyclins retract their binding to AMBRA1, which could degrade and destabilize D-type cyclin proteins, but a loss of AMBRA1 could eliminate sensitivity to CDK4/6 inhibitors. Along with these findings, Maiani et al. also reported that AMBRA1 regulates cell cycle transition from G1 to S phase as an upstream master regulator for the prevention of replication stress, mainly in S and G2 phases (Maiani et al., 2021). Indeed, knockdown of AMBRA1 could prolong mitosis and the formation of anaphase bridges, followed by remarkable cell death. They found that AMBRA1 could directly bind and stabilize N-MYC through the phosphatase PP2A, suggesting transcriptional and posttranslational regulation of cyclins D1 and D2. Moreover, they identified the role of DDB1 (in the CRL4 complex) as a specific E3 ligase for the stability of cyclin D1 in AMBRA1 dependent manner. Conversely, a phosphorylation-deficient mutant of cyclin D1 (T286A) was resistant to proteasome degradation. Downregulation of AMBRA1 function could increase the levels of CHK1 expression and hyperphosphorylation (Ser345) of CHK1, resulting in induction of cyclin E2 expression. Therefore, CHK1 could be a therapeutic target in tumors lacking AMBRA1, and inhibition of CHK1 could exhibit remarkable replication stress and sensitivity in several tumors (Murga et al., 2011; Sen et al., 2017; Syljuasen et al., 2005). Meantime, Chaikovsky et al. observed the same discovery revealing that loss of AMBRA1 significantly increased the levels of cyclin D in vitro and in mice, which in turn stimulated cell proliferation and decreased sensitivity to the CDK4/6 inhibitors (Chaikovsky et al., 2021). They found that low levels of AMBRA1 were associated with poor survival in patients with lung adenocarcinoma, and loss of AMBRA1 promoted the growth of lung adenocarcinoma in an animal model.

Taken together, the AMBRA1/cyclin D axis has a potential cell-cycle regulatory function linked with genome stability and controls replication-phase entry during embryonic development and tumorigenesis. Therefore, they identified the mechanism of D-type cyclins integrated in the RB pathway during cell cycle progression in human cancer. The AMBRA1 E3 ligase adaptor could regulate cellular level of cyclin D through its destabilization for the sensitization of cancer cells in response to CDK4/6 inhibitors.

4. Immunomodulatory function of cell cycle regulators

The Cyclin D-CDK4/CRL3^SPOP/PD-L1 axis.

Programmed death-ligand 1 (PD-L1) is a critical factor for suppressing the adaptive immune system via binding to the inhibitory checkpoint molecule PD-1(Chemnitz, Parry, Nichols, June, & Riley, 2004). This inhibition reduces the proliferation of antigen-specific T cells while lowers apoptosis in regulatory T cells (suppressive T cells). Therefore, PD-L1 might be employed as a target for anti-cancer immunity (Razaghi, Mansouri, Brodin, Bjornstedt, & Lundahl, 2022). It is shown that nuclear PD-L1 (nPD-L1) in the binding with p-ERK translocates into the nucleus. The nPD-L1 interacts with thyroid hormone receptor-associated protein 3 (THRAP3) to upregulate the expression of cell cycle regulator BUB1 (budding uninhibited by benzimidazoles 1) for an acceleration of cell cycle progression and cell proliferation (Ma et al., 2022). Interestingly, PD-1 inhibits upregulation of the SCF^SKP2 ubiquitin ligase activity required for ubiquitin-dependent degradation of p27^KIP1 CDK inhibitor, thereby accumulating p27^KIP1 and impairing cyclin A, cyclin E and enzymatic activation of CDK2 (Patsoukis et al., 2012). This indicates that PD-1 inhibits cell cycle progression through upregulation of the CDK inhibitor, KIP family.

We found that abundance ofPD-L1 protein is increased during cell cycle progression in M/ early G1 phases, followed by a reduction in late G1/S phases in multiple cancer cell lines. The abundance of PD-L1 is increased after genetic depletion of all three D-type cyclins (D1, D2 and D3) in mouse MEFs and also in tumor-bearing mice, suggesting that cyclin D1 and lesser extent, cyclin D2 or D3 negatively regulate PD-L1 protein level in vitro and in mice tumor (Zhang et al., 2018). Moreover, the findings showed that depletion of CDK4, but not CDK6 or CDK2, increases the abundance of PD-L1 protein (Bates et al., 1994; Lees, Faha, Dulic, Reed, & Harlow, 1992). In agreement with the notion that RB is commonly inactivated in human cancer (Fry et al., 2004; Takaki et al., 2005), we could successfully reduce PD-L1 protein level after reduction of cyclin D-CDK4/6 inhibitor, or depletion of p16^INK4 in RB-deficient/p16-high cell lines, suggesting the opposite effect between PD-L1 expression and the CDK4 activity. In addition, to demonstrate the UPS-dependent PD-L1 regulation, we screened the potential partner of PD-L1 by investigating its binding with each Cullin family member, and excitingly found that Cullin3, through adaptor protein SPOP, specifically interacts and destabilizes PD-L1. Therefore, we concluded that CDK4/cyclin D along with Cullin3-Ring E3 ligase (CRL3)^SPOP complex negatively regulates protein stability of PD-L1 (Zhang et al., 2018) (Fig. 4). In this study, we further noticed that depletion of the adaptor protein CDH1, but not CDC20, in the composition of the APC/C E3 ligase, increases the SPOP protein level during cell cycle progression. We identified an evolutionarily conserved D-box motif (RxxLxxxxN) (da Fonseca et al., 2011) in SPOP, which could be recognized by CDH1. These findings suggest that SPOP is an APC/C^CDH1 substrate. CDK4/cyclin D-mediated phosphorylation and stabilization of SPOP mostly occurred via recruiting 14-3-3γ for disrupting the binding between SPOP and CDH1 (Zhang et al., 2018). These results indicate that E3 ligase APC/C^CDH1 negatively regulates the stability of SPOP protein in the cell cycle progression. Furthermore, we found that implanted tumors overexpressing cancer-derived SPOP-F102C mutant, compared to wild type, reveals high expression of PD-L1 protein level and markedly reduces the number of CD3⁺ tumor-infiltrating lymphocytes (TIL). Interestingly, these significant differences were lightened after treatment with anti-PD-L1 antibody (Zhang et al., 2018), suggesting that tumorigenicity of SPOP-mutant cells is significantly elevated by PD-L1 expression resulting from highly immune evasion. The number of CD8⁺ TIL was diminished in the samples with SPOP mutations (15 SPOP mutants), indicating that SPOP deficiency is associated with protein abundance of PD-L1 in primary human prostate cancers (Zhang et al., 2018).

Fig. 4.

Molecular regulation of CDK4/6-mediated PD-L1 instability and activation of T cells. (A) The cyclin D-CDK4/6 inactivates CDH1-mediated degradation of SPOP and/or phosphorylates the E3 ligase SPOP to facilitate its binding with 14-3-3γ to regulate the stability of PD-L1 protein. (B) The engagement of PD-L1⁺ tumor cells with PD-1⁺ T cells leads to T cell dysfunction for anti-tumor activity. The increased level of PD-L1 in patients treated with CDK4 inhibitors leads to the engagement of PD-L1⁺ tumor cells with PD-1⁺ T cells for T cell dysfunction, followed by a reduction in anti-tumor activity, tumor immune evasion, and drug resistance. The combination therapy using PD-1/PD-L1 blocking antibodies along with the CDK4/6 inhibitors facilitates the binding of TCR with MHC to activate adaptive immune response as an efficient anti-cancer therapeutic option. TCR, T cell receptor; MHC, major histocompatibility complex.

5. Cell cycle-based targeted cancer therapies

The aberrant progression of cell cycle is a hallmark of cancer. Therefore, targeting CDKs would be an effective anti-cancer therapy through blocking cell proliferation (Otto & Sicinski, 2017). Moreover, D-type cyclins are promising therapeutic targets in human cancer, and relevant mutations in their degrons have been found in several malignancies (Behan et al., 2019).

We found that acute inhibition of the CDK2/cyclin A2 using specific inhibitors decreased AKT phosphorylation, suggesting that CDK2/cyclin A2 is an upstream positive regulator of AKT1 kinase activity (Liu et al., 2014). We determined that the CDK2/cyclin A2-activated AKT1 subsequently phosphorylates SKP2 and FOXO, leading to cell cycle progression and confirmed resistance to the chemotherapeutic agents such as etoposide or camptothecin (Liu et al., 2014). Interestingly, the genetic depletion of Cdh1 could synergize with Pten loss for aberrant co-activation of BRAF/ERK and AKT signaling to drive melanomagenesis (Wan et al., 2017). In fact, loss of CDH1 provides vemurafenib (PLX4032) resistance in BRAF^V600E melanoma cells (Wan et al., 2017), indicating that activation of CDH1 might sensitize vemurafenib-resistant melanoma. Moreover, we found that CDK4/cyclin D1 and ERK-mediated phosphorylation of CDH1 could inhibit APC/C^CDH1 E3 ligase activity (Wan et al., 2017). Since CDK kinase activities reduce the interaction between CDH1 and the APC/C core complex that is required for its appropriate tumor suppressor function, targeting CDK4/6 kinases would be an attractive approach for anti-cancer therapy. Recent clinical trials have approved combinational strategies using BRAF^V600E inhibitor (LGX818 or encorafenib) along with another compound targeting MEK, CDK4/6 or PI3K to efficiently improve survival (Catalanotti & Solit, 2012; Dummer et al., 2018a, 2018b; van Geel et al., 2017; Yoshida, Kunisada, Kusakabe, Nishikawa, & Nishikawa, 1996) (Figs. 2 and 5).

Fig. 5.

Potential therapeutic strategies to target the cyclin D-CDK4/6 axis in cancer. Inhibition using drugs against the activity of cyclin D-CDK4/6 facilitates restoration of tumor suppressor activity of RB, but causes PD-L1 stabilization and upregulation. The effective CDK4/6 targeted therapy could be achieved in combination with immunotherapy (using PD-1/PD-L1 blocking antibodies), radiotherapy, and targeted therapies using Hippo/YAP, MYC, and PI3K/AKT inhibitors.

Depletion of AMBRA1 decreases sensitivity to CDK4/6 inhibitors (Simoneschi et al., 2021). In AMBRA1 depleted cells, treatment with CDK4/6 inhibitor abemaciclib enhances phosphorylation of RB. Furthermore, AMBRA1 knockout promotes the complex formation of cyclin D1 with CDK2, instead of CDK4, resulting in conferring resistance to palbociclib. Interestingly, inhibition of CDK4/6 function rescues DNA damage phenotype regulated by AMBRA1 knockdown, in which CDK2 depletion in AMBRA1-lacking cells resensitizes cells to abemaciclib and palbociclib in vitro and in tumor-bearing mice (Maiani et al., 2021; Simoneschi et al., 2021). In fact, synthetic lethality from AMBRA1 defects along with inhibition of CHK1 in several tumors exhibited remarkable replication stress and sensitivity to preclinically examined CHK1 inhibitors (Murga et al., 2011; Sen et al., 2017; Syljuasen et al., 2005), suggesting CHK1 as a therapeutic target in tumors lacking AMBRA1 (Figs. 3 and 5).

Fig. 3.

AMBRA1 regulates cyclin D-CDK4/6 signaling. (A) AMBRA1 in the Cullin4-RING E3 ligase (CRL4) complex regulates the G1–S-phase transition through reduction of CDK4/6 activity by ubiquitination and degradation of D-type cyclins, leading to tumor suppression. (B) Impairment of AMBRA1 function (alterations, downregulation, or mutations in the TP motif of D-type cyclins) results in the accumulation of cyclin D protein and also maintenance of RB phosphorylation, leading to a premature entry into S phase, persistent cell cycle progression, faster proliferation, replication stress, genome instability, and tumorigenesis. Moreover, the accumulation of cyclin D protein decreases sensitivity to CDK4/6 inhibitors.

Our previous study showed that treatment of multiple cancer cells with palbociclib or ribociclib could upregulate PD-L1 protein level or stability in vitro and subsequently reduces the number of infiltrating CD3⁺ TILs in vivo (Zhang et al., 2018). Treatment of cells with palbociclib impairs the phosphorylation and protein abundance of SPOP, followed by elevation in the PD-L1 levels. We further demonstrated that treatment of tumor-bearing mice with palbociclib reduces the numbers of TILs, including Granzyme B⁺, IFNg⁺, CD3⁺, CD4⁺, and CD8⁺ cells. Since the success of PD1/PD-L1 blockade is correlated with the expression of PD-L1 level (Herbst et al., 2014; Iwai et al., 2002), inhibition of CDK4/6 along with anti-PD-1/PD-L1 therapy might synergize and enhance the therapeutic effect, as revealed in our findings (Zhang et al., 2018). These findings, together with a study showing that a CDK4/6 inhibitor abemaciclib increases immunogenicity of cancer cells through antigen presentation and stimulating production of type III interferons via an RB-dependent mechanism (Goel et al., 2017) can provide a comprehensive molecular rationale for the combination of CDK4/6 inhibitor along with anti-PD1/PD-L1 immunotherapy to eradicate tumor cells (Fig. 4 and 5).

6. Targeting CK1ε for overcoming resistance to CDK4/6 inhibitors

RB1 pathway plays a critical role in the regulation of cell cycle progression (Giacinti & Giordano, 2006). However, acquired drug resistance is frequently observed, and underlying mechanisms remain to be elusive. To this end, we tested CDK4/6 inhibitors, abemaciclib, palbociclib, and ribociclib, to investigate the molecular mechanism of resistance in breast cancer. We found that silencing either RB1 or CDK6 resulted in a significant change in drug sensitivity of cancer cells to CDK4/6 inhibitors (Dang et al., 2021). The inhibitor treatments results in significant reduction in RB1 protein levels and increase in protein abundance of CDK6, cyclin E1, and cyclin D1. Interestingly, cyclin D1 or CDK4 depletion dramatically reduces the RB protein level, suggesting that cyclin D1-CDK4/6 inhibition may promote RB1 proteasomal degradation. In fact, we revealed that the cyclin D1-CDK4/6 suppression in G1 phases triggers β-TRCP-mediated RB1 degradation upon treatment with CDK4/6 inhibitors (Dang et al., 2021). However, depletion of Cullin1 enhances the suppressed effect of CDK4/6 inhibitor on cancer cell proliferation. Since the β-TRCP-mediated substrate recognition and subsequent degradation require prior phosphorylation of target proteins (Skaar, Pagan, & Pagano, 2013; Wang, Liu, Inuzuka, & Wei, 2014), we found that ectopic expression of CK1ε, but not other kinases, led to a reduction in the protein abundance of RB1. Inversely, knockdown or inhibition of CK1ε results in RB1 accumulation. Interestingly, the inhibition of CDK4/6 activates the SP1/p300/CBP complex for transcriptional activation of CDK6 (Dang et al., 2021). Furthermore, depletion or inhibition of CK1ε, but not CK1α, accelerates SP1 degradation to reduce its accumulation on the CDK6 promoter region, leading to a reduction in the CDK6 mRNA level. Consistently, treatment with CK1ε inhibitor sensitizes the resistant cancer cells to ribociclib treatment, leading to suppressing cancer cell proliferation in vitro and in vivo. Therefore, CK1ε appears to be an effective target in preventing RB1 degradation and CDK6 accumulation after treatment with the CDK4/6 inhibitors, thereby overcoming resistance to CDK4/6 inhibitors in breast cancer cells.

7. Discussion

Cyclin D and its associated CDKs are potential therapeutic targets but have poor efficacy in clinical trials. It has been identified that amplification, mutation, and overexpression of cyclin D has contributed to cancer progression. Since catalytic and non-catalytic functions of cyclin D are both critical in the physiology of normal and neoplastic cell function, targeting cyclin D should be effectively designed alone or in combination with inhibition of CDK4/6 activity in cyclin D-dependent cancers. This will rely on better patient selection through biomarkers in response to therapy and understanding of the signature of cyclin D-genomics and proteomics.

Recent biochemical and structural studies have shed some light on the molecular signature by which E3 ligases regulate oncoprotein stability. The APC/C^CDH1 has been known as a tumor suppressor due to restraining proliferation and maintaining quiescent/G1 cells. Our findings also confirmed that CDH1 exerts a key tumor suppressor function in cell cycle progression, inhibiting proliferation via degradation of LATS1/2, AKT1, BRAF and SPOP in different contexts. CDH1 could regulate destabilization of LATS1/2 kinases in the G1 phase of cell cycle that increases YAP/TAZ activities by promoting E2F1 expression for G1/S transition (Kim et al., 2019). This links the function of CDH1 to the cell cycle regulation. Moreover, deregulation of the RAS/RAF/MEK/ERK molecular cascade is associated with tumor progression in human cancer (Roberts & Der, 2007). We found that CDH1 could negatively regulate BRAF protein abundance via APC/C-dependent BRAF degradation and or through APC/C-independent disruption of BRAF dimerization. This finding suggests that CDH1 might exert its novel function upstream of the BRAF by inhibiting its oncogenic signaling pathway. These findings introduce the multifunctional role of CDH1 in cell cycle regulation and cell fate determination through the regulation of several unique substrates. Moreover, we summarized the function of CRL4^AMBRA1 as a novel ubiquitin E3 ligase that degrades all three D-type cyclins during normal cell cycle progression; therefore, AMBRA1 secures as a tumor suppressor function in the normal cell cycling (Simoneschi et al., 2021). There is the redundancy of DNA damage in S phase, mitotic abnormalities and premature S-phase entry, all of which indicate the critical function of lacking AMBRA1 in the replication stress (Burrell et al., 2013; Wilhelm et al., 2014). The reduction in AMBRA1 might be one of the mechanisms for desensitization of CDK4/6 inhibitors. Therefore, the regulation of cyclin D protein turnover along with the use of combination therapies might be an alternative strategy to target the function of cyclin D in cancer.

Since the increased level of cyclin D expression and activation of CDK4/CDK6 are critical in therapeutic resistance, the important question would be on discovery of secondary partners in the regulation of cyclin D overexpression. Therefore, understanding of mechanisms involved in the resistance to inhibition of cyclin D and or CDK4/6 would be a paradigm shift in cell cycle therapy in cancer. We found that level of PD-L1 is considered a severe clinical problem in patients treated with CDK4 inhibitors. This might be underlying in the resistance to CDK4 inhibitors by avoiding immune surveillance checkpoint. Moreover, targeting the activation of CK1ε is critical for overcoming resistance to CDK4/6 inhibitors. Based on the observation, the requirement for inhibitors to target CDK2, CDK4, CDK6 and CK1ε simultaneously might be a promising approach. Furthermore, a therapeutic combination using PD-L1 immunotherapy might also lower the development of resistance to manage PD-L1 blockade. This strategy provides a unique molecular mechanism for targeting CDK4/6 along with the introduction of combination therapy as an efficient clinical option for cancer treatment.