Crosstalk of the mitotic spindle assembly checkpoint with p53 to prevent polyploidy

Treatment of cells with microtubule inhibitors results in activation of the mitotic spindle assembly checkpoint, leading to mitotic arrest before anaphase. Upon prolonged treatment, however, cells can adapt and exit mitosis aberrantly, resulting in the occurrence of tetraploid cells in G1. Those cells subsequently arrest in postmitotic G1 due to the activation of a p53-dependent G1 checkpoint. Failure of the G1 checkpoint leads to endoreduplication and further polyploidization. Using HCT116 and isogenic p53-deficient or spindle checkpoint compromised derivatives, we show here that not only p53 but also a functional spindle assembly checkpoint is required for postmitotic G1 checkpoint function. During transient mitotic arrest, p53 stabilization and activation is triggered by a pathway independent of ATM/ATR, Chk1 and Chk2. We further show that a prolonged spindle checkpoint-mediated mitotic arrest is required for proper postmitotic G1 checkpoint function. In addition, we demonstrate that polyploid cells are inhibited to re-enter mitosis by an additional checkpoint acting in G2. Thus, during a normal cell cycle, polyploidization and subsequent aneuploidization is prevented by the function of the mitotic spindle checkpoint, a p53-dependent G1 checkpoint and an additional G2 checkpoint.     

Functional cooperation of RKTG with p53 in tumorigenesis and epithelial-mesenchymal transition

Raf kinase trapping to Golgi (RKTG) is a potential tumor suppressor gene due to its negative roles in regulating Ras/Raf/MEK/ERK (extracellular signal-regulated kinase) pathway and GPCR (G protein-coupled receptor) Gβγ subunit signaling. Interestingly, RKTG-deficient mice are free of tumors, although they are prone to form skin cancer on carcinogen administration. On the other hand, p53 is a well-characterized tumor suppressor gene and p53 heterozygous mice develop sarcoma and other tumors starting from 12 months of age. In RKTG-null mouse embryonic fibroblasts, lypophosphatidic acid (LPA), but not EGF (epidermal growth factor), could stimulate hyperphosphorylation of AKT and GSK3β, accompanied by increases in phosphorylation of p53 at Ser15 and accumulation of p53, as well as its target genes p21 and p16. Spontaneous skin cancer-like tumors were detected in about 25% of RKTG nullizygous and p53 heterozygous mice within 7 months of age. Hyperplasia and epithelial-mesenchymal transition (EMT) were observed in the tumor-overlying epidermis, in which LOH of p53 occurred and EMT features emerged. In p53-mutated A431 epithelial carcinoma cells, knockdown of RKTG led to enhancement of LPA-stimulated AKT and GSK3β phosphorylation, together with increased accumulation of β-catenin and appearance of EMT features that were antagonized by p53 overexpression. In HepG2 epithelial cells, LPA-stimulated AKT phosphorylation and EMT features reached maximum when both RKTG and p53 were simultaneously silenced. In summary, these results not only indicate that RKTG has an in vivo tumor suppressor function to cooperate with p53 in tumorigenesis but also suggest that p53 has an EMT checkpoint function and the loss of this function can combine with loss of RKTG to drive EMT and tumor progression.     

Mechanism of hyperploid cell formation induced by microtubule inhibiting drug in glioma cell lines

Checkpoint mechanism plays a crucial role in ensuring genomic integrity during cell cycle. Loss of checkpoint function is known to induce genomic instability and to alter ploidy of dividing cells. In this study, we examined mechanisms of hyperploid formation in glioma cells by treatment with nocodazole, which activates spindle assembly checkpoint by inhibiting microtubule polymerization. By prolonged nocodazole treatment, U251MG human glioma cell, which has a p53 mutation, underwent transient arrest at mitosis, and subsequently exited from mitotic arrest (termed 'mitotic slippage') followed by DNA replication without cytokinesis, resulting in hyperploid formation. Additionally, the heterogeneity in the number of centrosomes per cell increased during the hyperploid formation, suggesting that these hyperploid cells have genomic instability. By employing LN382 glioma cell that has a temperature-sensitive p53 mutation, we found that the activation of p53 prevents hyperploid formation after the prolonged nocodazole treatment. Furthermore, staurosporine, an inhibitor for a broad range of serine/threonine kinases including cdc2, was found to enhance hyperploid formation in U251MG cells by accelerating the induction of mitotic slippage. Interestingly, inhibitors specific for cdc2 kinase prevented the G2 to M transition but did not accelerate mitotic slippage, suggesting that staurosporine-sensitive kinases other than cdc2 are required for maintenance of spindle assembly checkpoint. Moreover, the enhancement of hyperploid formation by staurosporine was also blocked by p53-dependent G1 checkpoint. These results suggest that abrogation of G1 checkpoint is a critical factor for formation of hyperploid cells after the mitotic slippage.     

Sequential activation and inactivation of G2 checkpoints for selective killing of p53-deficient cells by microtubule-active drugs

By inducing p53-dependent G2 arrest, the pretreatment with low concentrations of DNA damaging drugs (e.g., doxorubicin, DOX) can prevent cell death caused by microtubule-active drugs (e.g., paclitaxel, PTX), thus potentially permitting selective killing of p53-deficient cancer cells. However, DOX still protects a subset of tumor cell lines lacking wt p53 (HL60 and Jurkat leukemia cells), thus limiting the utility of protection of cells with wt p53 (e.g., normal cells). The present work overcomes this obstacle by adding an abrogator of p53-independent checkpoint (e.g., UCN-01) to the DOX-PTX sequence. By inhibiting a p53-independent pathway, UCN-01 overrode DOX-induced G2 arrest and instead induced G1 arrest in HL60 and Jurkat, thus propelling these p53-deficient cells from G2 to G1. Once they entered mitosis, cells were killed by PTX. Induction of G2 arrest with sequential abrogation of a p53-independent checkpoint allows pharmacological manipulation of Raf-1/Bcl-2 hyperphosphorylation, PARP and Rb cleavage and cell death caused by PTX in p53-deficient cells. Unlike previous approaches, this strategy is intended to increase selectivity, not the cytotoxicity of PTX. This rational sequence of agents that induces p53-dependent and abrogates p53-independent arrest represents a cancer-selective strategy for treatment of p53-deficient tumors.     

Protein kinase CK2 is involved in G2 arrest and apoptosis following spindle damage in epithelial cells.

p53 undergoes phosphorylation on several residues in response to cellular stresses that include UV and ionizing radiation, however the influence of spindle damage on this parameter is relatively unclear. Consequently, the effect of nocodazole on serine 392 phosphorylation was examined in two epithelial cell lines. We show that this process is dependent upon the stepwise activation of p38 mitogen-activated protein kinase (p38 MAPK) and protein kinase casein kinase 2 (CK2). Furthermore, this activation correlated with the biochemical regulation of the maturation-promoting factor (MPF, cdc2/cyclin B), as both DRB and antisense depletion of CK2, as well as SB203580 were associated with an inhibition of its activation in response to nocodazole. Strikingly, when the cell cycle characteristics of nocodazole treated cells were examined, we observed that depletion or inhibition of the catalytic subunit of CK2, in the presence of microtubule inhibitors, resulted in a compromise of the G2 arrest (spindle checkpoint). Furthermore, CK2-depleted, nocodazole treated cells demonstrated a dramatic reduction in the apoptotic cell fraction, confirming that these cells had been endowed with oncogenic properties. These changes were observed in both HeLa cells and HCT116 cells. We also show that this effect is dependent on the presence of functional wild-type p53, as this phenomenon is not apparent in HCT116 p53(-/-) cells. Collectively, our results indicate two novel roles for CK2 in the spindle checkpoint arrest, in concert with p53. Firstly, to maintain increased cyclinB/cdc2 kinase activity, as a component of G2 arrest, and secondly, a role in p53-mediated apoptosis. These findings may have implications for an improved understanding of abnormalities of the spindle checkpoint in human cancers, which is a prerequisite for defining future therapies.     

Tumor suppressor WARTS ensures genomic integrity by regulating both mitotic progression and G1 tetraploidy checkpoint function

Defects in chromosomes or mitotic spindles activate the spindle checkpoint, resulting in cell cycle arrest at prometaphase. The prolonged activation of spindle checkpoint generally leads to mitotic exit without segregation after a transient mitotic arrest and the consequent formation of tetraploid G(1) cells. These tetraploid cells are usually blocked to enter the subsequent S phase by the activation of p53/pRb pathway, which is referred to as the G(1) tetraploidy checkpoint. A human homologue of the Drosophila warts tumor suppressor, WARTS, is an evolutionarily conserved serine-threonine kinase and implicated in development of human tumors. We previously showed that WARTS plays a crucial role in controlling mitotic progression by forming a regulatory complex with zyxin, a regulator of actin filament assembly, on mitotic apparatus. However, when WARTS is activated during cell cycle and how the loss of WARTS function leads to tumorigenesis have not been elucidated. Here we show that WARTS is activated during mitosis in mammalian cells, and that overexpression of a kinase-inactive WARTS in Rat1 fibroblasts significantly induced mitotic delay. This delay resulted from prolonged activation of the spindle assembly checkpoint and was frequently followed by mitotic slippage and the development of tetraploidy. The resulting tetraploid cells then abrogated the G(1) tetraploidy checkpoint and entered S phase to achieve a DNA content of 8N. This impairment of G(1) tetraploidy checkpoint was caused as a consequence of failure to induce p53 expression by expressing a kinase-inactive WARTS. WARTS thus plays a critical role in maintenance of ploidy through its actions in both mitotic progression and the G(1) tetraploidy checkpoint.     

p53 activation in response to mitotic spindle damage requires signaling via BubR1-mediated phosphorylation

The mitotic spindle checkpoint plays a crucial role in regulating accurate chromosome segregation and preventing the adaptation of multiploid progeny cells. Recent reports have indicated that the induction of p53 by mitotic checkpoint activation is essential for protecting cells from abnormal chromosome ploidization caused by mitotic failure. However, although studies have shown that p53 deficiencies arrest mitosis, compromise apoptosis, and may cause profound aneuploidy, the molecular mechanisms leading to p53 induction following mitotic checkpoint activation remain unknown. Here, we show that the BubR1 mitotic checkpoint kinase interacts with p53 both in vitro and in vivo, with higher levels of interaction in mitotic cells. This interaction contributes to p53 phosphorylation. Silencing of BubR1 expression reduces the phosphorylation and stability of p53, whereas exogenous introduction of BubR1 proteins into BubR1-depleted cells recovers p53 stability. In addition, inhibition of BubR1 expression in the presence of a microtubule inhibitor accelerates chromosomal instability and polyploidy in p53-null cells. These results collectively suggest that p53 activation in response to mitotic spindle damage requires signaling via BubR1-mediated phosphorylation.     

The Aurora kinase inhibitor VX-680 induces endoreduplication and apoptosis preferentially in cells with compromised p53-dependent postmitotic checkpoint function

VX-680 is a potent inhibitor of Aurora kinases that induces the accumulation of cells with > or =4N DNA content, followed by cell death. Here, we define the role of p53 and p21(Waf1/Cip1) in cell cycle perturbations following exposure to VX-680. Endoreduplication and apoptosis in response to VX-680 are limited in A549 and MCF-7 cells expressing wild-type p53, and markedly enhanced in cells lacking p53, including those engineered to express the HPV16-E6 oncoprotein or short interfering RNA pools targeting p53. In contrast, endoreduplication and apoptosis occur in the p53 wild-type cell lines, RKO and U2OS. The difference in response to VX-680 among these cell lines correlates with the timing of induction of p21(Waf1/Cip1) and its ability to inhibit cyclin E-cdk2 activity. In A549 cells, VX-680 induces the expression of p53 and p21(Waf1/Cip1) within 24 hours, with consequent inhibition of cyclin E-cdk2, and reduction of retinoblastoma protein phosphorylation, limiting endoreduplication. In RKO and U2OS cells, the induction of p21(Waf1/Cip1) is delayed and associated with higher residual cyclin E-cdk2 kinase activity and retinoblastoma protein phosphorylation, followed by progressive endoreduplication and apoptosis. Abrogation of p21(Waf1/Cip1) expression by short interfering RNA targeting in A549 cells results in a substantial increase in the degree of endoreduplication, whereas inducible expression of p21(Waf1/Cip1) in p53-negative NCI-H1299 cells inhibits VX-680-induced endoreduplication and cell death. These data suggest that the integrity of the p53-p21(Waf1/Cip1)-dependent postmitotic checkpoint governs the response to Aurora kinase inhibition. Although cells with intact checkpoint function arrest with 4N DNA content, those with compromised checkpoint function are more likely to undergo endoreduplication followed by eventual apoptosis.     

Hypoxia attenuates the p53 response to cellular damage

The tumour suppressor activity of p53 in vivo can be subject to pressure from the physiological stress of hypoxia and we report on the development of a cell system to define the p53-dependent stages in the adaptation of cells to hypoxia. p53(+/+) cells exposed to hypoxia exhibited a transient arrest in G2/M, but escaped from this checkpoint and entered a long-term G(0)/G(1) arrest. By contrast, isogenic p53-null cells exposed to hypoxic conditions exhibited a 6-10-fold higher level of apoptosis, suggesting that p53 acts as a survival factor under limiting oxygen concentrations. Surprisingly, hypoxia-dependent growth arrest in p53(+/+) cells did not result in either p21(WAF1) or HIF-1 protein stabilization, but rather promoted a significant decrease in Ser(392)-site phosphorylation at the CK2/FACT site. However, chemically induced anoxia induced Ser(392)-site phosphorylation as well as stabilization of both p53 and HIF-1 proteins. In contrast to hypoxia, 5-flourouracil (5-FU)-induced p53-dependent cell death correlated with enhanced Ser(392) phosphorylation of p53 and elevated p21(WAF1) protein levels. Hypoxia inhibited 5-FU-induced p53-dependent cell death and attenuated p53 phosphorylation at the ATM and CK2/FACT phosphorylation sites. Although anoxia activates the p53 response, hypoxia silences the p53 transactivation pathway and identifies a physiological signalling model to study mechanisms of p53 inactivation under hypoxic conditions.     

DNA damage-induced cell-cycle phase regulation of p53 and p21waf1 in normal and ATM-defective cells

The ATM-dependent accumulation of p53 and induction of p21waf1 are key events for G1 cell-cycle checkpoint arrest following DNA damage. In ATM-null AT cells, even though the p53 and p21waf1 responses are kinetically delayed and quantitatively reduced, the G1 checkpoint is virtually disrupted, suggesting that these proteins arrive too late in G1 to enforce the arrest. As the precise mechanism remains unclear, we examined the response to DNA double-strand breaks generated by gamma-radiation (IR), to determine if ATM deficiency affects the cell-cycle phase regulation of these molecules. We find that, after irradiation, whereas normal LCL-N cells markedly increase their levels of p53 in all phases of the cell cycle, AT cells fail to show any p53 increase in the G1 phase. In addition, whereas in LCL-N p21waf1 is induced in G1 and G2-M, in AT cells this induction is partly seen in G2-M, but not in G1, indicating a different cell-cycle phase regulation of p53 and p21waf1 as a result of ATM deficiency. The levels and catalytic activity of the p53-targeting kinases ATR and DNA-PK in LCL-N and AT cells are very similar throughout the cell cycle, both before and after IR, thus excluding a phase-specific activity for these kinases. Collectively, our findings demonstrate that, in ATM-deficient cells, the p53-dependent p21waf1 response to DNA damage is not only quantitatively reduced, but also specifically suppressed in the G1 phase, thus providing a mechanistic explanation for the severe disruption of the G1 checkpoint in AT cells.     

Radiation-induced phosphorylation of Chk1 at S345 is associated with p53-dependent cell cycle arrest pathways

Because DNA damage-inducible cell cycle checkpoints are thought to protect cells from the lethal effects of ionizing radiation, a better understanding of the mechanistic functions of cell cycle regulatory proteins may reveal new molecular targets for cancer therapy. The two major regulatory proteins of G2 arrest are Chk1 and p53. Yet, it is unclear how these two proteins interact and coordinate their functional roles during radiation-induced G2 arrest. To determine Chk1's role in p53-dependent G2 arrest, we used p53 proficient cells and examined expression of G2 arrest proteins under conditions in which G2 arrest was inhibited by the staurosporine analog, UCN-01. We found that UCN-01 inhibited both G1 and G2 arrest in irradiated p53 proficient cells. The arrest inhibition was associated with suppression of radiation-induced expression of both p21 and 14-3-3 sigma -- two known p53-dependent G2 arrest proteins. The suppression occurred despite normal induction of p53 and normal phosphorylation of p53 at S20 and Cdc25C at S216 -- the two known substrates of Chk1 kinase activity. In contrast, we showed that radiation-induced phosphorylation of Chk1 at S345 was associated with binding of Chk1 to p53, p21, and 14-3-3 sigma, and that UCN-01 inhibited S345 phosphorylation. We suggest that DNA damage-induced phosphorylation of Chk1 at S345, and subsequent p53 binding, links Chk1 with p53 downstream responses and may provide a coordinated interaction between DNA damage responses and cell cycle arrest functions.     

hSMG-1 and ATM sequentially and independently regulate the G1 checkpoint during oxidative stress

Genotoxic stress activates the phosphatidylinositol 3-kinase-like kinases (PIKKs) that phosphorylate proteins involved in cell cycle arrest, DNA repair and apoptosis. Previous work showed that the PIKK ataxia telangiectasia mutated (ATM) but not ATM and Rad3 related phosphorylates p53 (Ser15) during hyperoxia, a model of prolonged oxidative stress and DNA damage. Here, we show hSMG-1 is responsible for the rapid and early phosphorylation of p53 (Ser15) and that ATM helps maintain phosphorylation after 24 h. Despite reduced p53 phosphorylation and abundance in cells depleted of hSMG-1 or ATM, levels of the p53 target p21 were still elevated and the G(1) checkpoint remained intact. Conditional overexpression of p21 in p53-deficient cells revealed that hyperoxia also stimulates wortmannin-sensitive degradation of p21. siRNA depletion of hSMG-1 or ATM restored p21 stability and the G(1) checkpoint during hyperoxia. These findings establish hSMG-1 as a proximal regulator of DNA damage signaling and reveal that the G(1) checkpoint is tightly regulated during prolonged oxidative stress by both PIKK-dependent synthesis and proteolysis of p21.     

UCN-01 inhibits p53 up-regulation and abrogates gamma-radiation-induced G(2)-M checkpoint independently of p53 by targeting both of the checkpoint kinases,Chk2 and Chk1

UCN-01 (7-hydroxystaurosporine) is a cell-cycle checkpoint abrogator that sensitizes cells to ionizing radiation (IR) and chemotherapeutic agents. It has been shown previously that UCN-01 abrogates DNA-damage-induced G(2) checkpoint most selectively in p53-defective cells, by primarily targeting Chk1. Here we show that UCN-01 prevented IR-induced p53 up-regulation and p53 phosphorylation on serine 20, a site previously identified for Chk2 (or/and Chk1) kinase. We found that in human colon carcinoma HCT116 cells, IR treatment enhanced Chk2 kinase activity, whereas Chk1 activity remained unchanged, which suggested that UCN-01 may interrupt IR-induced p53 response by inhibiting Chk2 kinase. This conclusion is supported by in vitro kinase assays, showing that UCN-01 inhibits Chk2 immunoprecipitated from HCT116 cells (IC(50), approximately 10 nM). In addition, UCN-01 efficiently abrogated both the initiation and maintenance of IR-induced G(2) arrest in HCT116 cells and their isogenic p53 (-/-) derivative, indicating that G(2) checkpoint abrogation by UCN-01 is p53 independent. In the p53 (-/-) cells, there was no p21(Waf1/Cip1) induction nor UCN-01-induced apoptosis. Taken together, these observations indicate that UCN-01 can modulate both Chk1 and Chk2 in intact cells and enhance IR-induced apoptosis in p53-deficient, and consequently p21-deficient, cells.     

Effects of the protein kinase inhibitors wortmannin and KN62 on cellular radiosensitivity and radiation-activated S phase and G1/S checkpoints in normal human fibroblasts

Wortmannin is a potent inhibitor of phosphatidylinositol (PI) 3-kinase and PI 3-kinase-related proteins (e.g. ATM), but it does not inhibit the activity of purified calmodulin-dependent protein kinase II (CaMKII). In the present study, we compared the effects of wortmannin and the CaMKII inhibitor KN62 on the response of normal human dermal fibroblast cultures to gamma radiation. We demonstrate that wortmannin confers a phenotype on normal fibroblasts remarkably similar to that characteristic of cells homozygous for the ATM mutation. Thus wortmannin-treated normal fibroblasts exhibit increased sensitivity to radiation-induced cell killing, lack of temporary block in transition from G1 to S phase following irradiation (i.e. impaired G1/S checkpoint), and radioresistant DNA synthesis (i.e. impaired S phase checkpoint). Wortmannin-treated cultures display a diminished capacity for radiation-induced up-regulation of p53 protein and expression of p21WAF1, a p53-regulated gene involved in cell cycle arrest at the G1/S border; the treated cultures also exhibit decreased capacity for enhancement of CaMKII activity post-irradiation, known to be necessary for triggering the S phase checkpoint. We further demonstrate that KN62 confers a radioresistant DNA synthesis phenotype on normal fibroblasts and moderately potentiates their sensitivity to killing by gamma rays, without modulating G1/S checkpoint, p53 up-regulation and p21WAF1 expression following radiation exposure. We conclude that CaMKII is involved in the radiation responsive signalling pathway mediating S phase checkpoint but not in the p53-dependent pathway controlling G1/S checkpoint, and that a wortmannin-sensitive kinase functions upstream in both pathways.     

Flavopiridol inversely affects p21(WAF1/CIP1) and p53 and protects p21-sensitive cells from paclitaxel

Resting cells are relatively resistant to microtubule-active drugs including paclitaxel (PTX). By causing p53-mediated arrest, pretreatment with low concentrations of doxorubicin (DOX) protected HCT116 cells from the cytotoxicity caused by PTX. Unlike DOX, flavopiridol (FL) did not protect HCT116 cells. Low concentrations of FL (50 nM) induced p21 but not p53. High concentrations of FL (500 nM) decreased levels of p21 and Mdm-2 but dramatically induced p53. Thus, FL reciprocally affects p21 and p53. In LNCaP, a prostate cancer cell line which is highly sensitive to p21-induced growth arrest (p21-sensitive), low concentrations of FL (50 nM) induced p21 (without induction of p53) and caused G1 and G2 arrest. This precluded mitotic arrest, Bcl-2 and Raf-1 phosphorylation, and diminished cell death caused by PTX. In contrast, FL did not protect PC3M, arrest-resitant and highly aggressive prostate cancer cells. Like LNCaP, HL60 and SKBr3 cells are known to be p21-sensitive. As predicted, low concentrations of FL antagonized PTX-mediated cytotoxicity in HL60 and SKBr3 cell lines. In summary, only low concentrations of FL can induce p21, and, in turn, only p21-sensitive cells are protected from PTX.     

High and low fluences of alpha-particles induce a G1 checkpoint in human diploid fibroblasts

The effects of exposure to high and very low fluence alpha-particles on the G1 checkpoint were investigated in human diploid fibroblasts irradiated and released from density-inhibited confluent cultures by the use of the cumulative labeling index method. Transient and permanent arrests in G1 occurred in fibroblast populations exposed to mean doses as low as 1 cGy, suggesting that nontraversed bystander cells may contribute to the low dose response. In cells exposed to high fluences, the G1 checkpoint is at least as extensive as in gamma-irradiated cells. In contrast to gamma-irradiated cells, neither repair of potentially lethal damage nor a reduction in the fraction of cells transiently or permanently arrested in G1 were observed in cells held in confluence for 6 h after alpha-particle irradiation. Studies with isogenic wild-type, p53-/-, and p21Waf1-/- mouse embryo fibroblasts exposed to either gamma or alpha-particle radiation revealed a total lack of G1 arrest in either p53-/- or p21waf1-/- cells, indicating that the G1 checkpoint in wild-type cells is p53-dependent and that p21Wf1 fully mediates the role of p53 in its induction. In contrast to human cells, mouse embryo fibroblasts do not undergo a permanent G1 arrest. Except under conditions favoring potentially lethal damage repair, a comparable expression pattern of p53, p21Waf1, and other cell cycle-regulated proteins (pRb, p34cdc2, and cyclin B1) was observed in alpha-particle or gamma-irradiated human fibroblasts.     

Growth inhibition of A549 human lung adenocarcinoma cells by L-canavanine is associated with p21/WAF1 induction.

L-Canavanine (CAV) is a higher plant nonprotein amino acid and a potent L-arginine antimetabolite. CAV can inhibit the proliferation of tumor cells in vitro and in vivo, but little is known regarding the molecular mechanisms mediating these effects. We demonstrated that the treatment of human lung adenocarcinoma A549 cells with CAV caused growth inhibition; G1 phase arrest is accompanied by accumulation of an incompletely phosphorylated form of the retinoblastoma protein, whose phosphorylation is necessary for cell cycle progression from G1 to S phase. In addition, CAV induces the expression of p53 and subsequent expression of a cyclin-dependent kinase inhibitor, p21/WAF1. The p53-dependent induction of p21/WAF1 and the following dephosphorylation of the retinoblastoma protein by CAV could account for the observed CAV-mediated G1 phase arrest.