Disruption of G1-phase phospholipid turnover by inhibition of Ca2+-independent phospholipase A2 induces a p53-dependent cell-cycle arrest in G1 phase.

The G1 phase of the cell cycle is characterized by a high rate of membrane phospholipid turnover. Cells regulate this turnover by coordinating the opposing actions of CTP:phosphocholine cytidylyltransferase and the group VI Ca2+-independent phospholipase A2 (iPLA2). However, little is known about how such turnover affects cell-cycle progression. Here, we show that G1-phase phospholipid turnover is essential for cell proliferation. Specific inhibition of iPLA2 arrested cells in the G1 phase of the cell cycle. This G1-phase arrest was associated with marked upregulation of the tumour suppressor p53 and the expression of cyclin-dependent kinase inhibitor p21cip1. Inactivation of iPLA2 failed to arrest p53-deficient HCT cells in the G1 phase and caused massive apoptosis of p21-deficient HCT cells, suggesting that this G1-phase arrest requires activation of p53 and expression of p21cip1. Furthermore, downregulation of p53 by siRNA in p21-deficient HCT cells reduced the cell death, indicating that inhibition of iPLA2 induced p53-dependent apoptosis in the absence of p21cip1. Thus, our study reveals hitherto unrecognized cooperation between p53 and iPLA2 to monitor membrane-phospholipid turnover in G1 phase. Disrupting the G1-phase phospholipid turnover by inhibition of iPLA2 activates the p53-p21cip1 checkpoint mechanism, thereby blocking the entry of G1-phase cells into S phase.


Introduction
Phospholipids are the major building blocks of cell membranes, which are crucial to the life of the cell. To successfully form daughter cells, cells must double their phospholipid mass during cell-cycle progression. Phosphatidylcholine (PtdCho) is a major component of phospholipids in mammalian cells, and regulation of its biosynthesis and turnover is crucial in maintaining membrane structure and function (Lykidis and Jackowski, 2001).
PtdCho metabolism varies throughout the cell cycle (Jackowski, 1996;Lykidis and Jackowski, 2001). Although cells in G1 phase rapidly synthesize and degrade PtdCho, they maintain a constant total membrane phospholipid mass (Jackowski, 1994). By contrast, PtdCho turnover ceases in S phase to allow the cells to double their membrane phospholipid content in preparation for cell division, and the synthesis and degradation of membrane phospholipids components are at their lowest point in G2 and M phases (Jackowski, 1994;Jackowski, 1996). It is obvious that a cell must have stringent control mechanisms to keep the phospholipid content in tune with the cell cycle.
Many signals influence cell division and the deployment of the developmental program of a cell during G1 phase. Diverse metabolic, stress and environmental cues are integrated and interpreted during this period to determine whether the cell enters S phase or pauses in its cell cycle. The G1-phase cells maintain a constant membrane phospholipid content by coordinating the opposing actions of CTP:phosphocholine cytidylyltransferase (CCT) and the group VIA Ca 2+independent-phospholipase A 2 (iPLA 2 ) (Baburina and Jackowski, 1999;Barbour et al., 1999); several lines of evidence indicate that this coordination is crucial to normal cell proliferation (Jackowski, 1996;Lykidis and Jackowski, 2001). First, enforced CCT expression stimulates both the incorporation of choline and glycerol into PtdCho as well as the degradation of PtdCho to glycerophosphocholine (GPC) by upregulating iPLA 2 expression (Baburina and Jackowski, 1999;Barbour et al., 1999). Second, cellular proliferation is inhibited when PtdCho is modified to prevent its degradation to GPC (Baburina and Jackowski, 1999). Third, overexpression of iPLA 2 in cells of the insulinoma (INS-1) cell line increased the rate of cell proliferation .
iPLA 2 hydrolyzes the sn-2 fatty acyl bond of phospholipids to liberate free fatty acids and lysophospholipids . It has also been reported to be involved in cell proliferation Roshak et al., 2000;Sanchez and Moreno, 2001;Sanchez and Moreno, 2002). Since the regulated deacylation of PtdCho to GPC is a key process in The G1 phase of the cell cycle is characterized by a high rate of membrane phospholipid turnover. Cells regulate this turnover by coordinating the opposing actions of CTP:phosphocholine cytidylyltransferase and the group VI Ca 2+ -independent phospholipase A 2 (iPLA 2 ). However, little is known about how such turnover affects cell-cycle progression. Here, we show that G1-phase phospholipid turnover is essential for cell proliferation. Specific inhibition of iPLA 2 arrested cells in the G1 phase of the cell cycle. This G1-phase arrest was associated with marked upregulation of the tumour suppressor p53 and the expression of cyclin-dependent kinase inhibitor p21 cip1 . Inactivation of iPLA 2 failed to arrest p53-deficient HCT cells in the G1 phase and caused massive apoptosis of p21deficient HCT cells, suggesting that this G1-phase arrest requires activation of p53 and expression of p21 cip1 . Furthermore, downregulation of p53 by siRNA in p21deficient HCT cells reduced the cell death, indicating that inhibition of iPLA 2 induced p53-dependent apoptosis in the absence of p21 cip1 . Thus, our study reveals hitherto unrecognized cooperation between p53 and iPLA 2 to monitor membrane-phospholipid turnover in G1 phase. Disrupting the G1-phase phospholipid turnover by inhibition of iPLA 2 activates the p53-p21 cip1 checkpoint mechanism, thereby blocking the entry of G1-phase cells into S phase. membrane phospholipid homeostasis, and the inability to degrade excess PtdCho inhibits cellular proliferation (Baburina and Jackowski, 1999), it is possible that degradation of excess PtdCho controls cell proliferation by tethering phospholipid metabolism to endogenous pathways of cell-cycle control. To investigate this possibility, we studied whether disrupting phospholipid turnover by specifically inhibiting iPLA 2 can induce cell-cycle arrest without affecting cell viability. Here, we demonstrate that inhibition of iPLA 2 directly regulates cell proliferation, arresting cells in the G1 phase of the cell cycle. This G1-phase arrest requires activation of the tumour suppressor p53 and expression of the cyclin-dependent kinase inhibitor p21 cip1 . These findings indicate that iPLA 2 cooperates with p53 to monitor a membrane phospholipid turnover in G1 phase.

Results
G1-phase phospholipid turnover is essential for cell proliferation iPLA 2 plays a key role in the regulation of G1-phase phospholipid turnover by degrading PtdCho (Baburina and Jackowski, 1999;Barbour et al., 1999). We and others previously reported that iPLA 2 is involved in cell proliferation Roshak et al., 2000;Sanchez and Moreno, 2001;Sanchez and Moreno, 2002). To investigate whether iPLA 2 -mediated PtdCho degradation is coupled to cell-cycle progression, we examined the effect of disrupting G1-phase phospholipid turnover by inhibiting iPLA 2 .
We first used bromoenol lactone (BEL), an irreversible and mechanism-based specific inhibitor of iPLA 2 , to inhibit the iPLA 2 activity in INS-1 cells Hazen et al., 1991). We found that BEL inhibited INS-1 proliferation in a concentration-dependent manner (Fig. 1A). Cell proliferation was completely inhibited by 10 M of BEL without causing cell death, consistent with a previous report (Fuentes et al., 2003). Cells treated with up to two days of 15 M BEL can proliferate after removal of the inhibitor, but persistent treatment caused reduced viability (Fig. 1B). BEL has recently been reported to inhibit Mg 2+ -dependent cytosolic phosphatidate phosphohydrolase-1 (PAP-1) (Balsinde and Dennis, 1996). To determine whether PAP-1 is involved in BEL-induced inhibition of cell proliferation, we treated the cells with propranolol, a well-known PAP-1-specific inhibitor (Fuentes et al., 2003;Pappu and Hauser, 1983). In accordance Journal of Cell Science 119 (6) for 2 days and then continuously cultured with (red) or without (green) BEL. Untreated cells were used as controls (blue). Error bars represent standard deviation from the mean of three experiments. (C) Transient expression of ARD-iPLA 2 -GFP reduces INS-1 cell proliferation. INS-1 cells (2ϫ10 5 ) were transfected with ARD-iPLA 2 -GFP (5 g/100 mm plate), then counted daily, starting 24 hours after transfection. Mock-transfected cells (no DNA) were used as controls. The cell lysates were prepared and analyzed for expression of ARD-iPLA 2 -GFP over 6 days, by western blotting with anti-GFP and anti-actin antibodies (top). Graph shows analysed data (bottom). Error bars represent the mean ± s.d. of three experiments. (D) Expression of Mut-iPLA 2 -GFP inhibits the proliferation of INS-1 cells. iPLA 2expressing INS-1 cells were transfected with Mut-iPLA 2 -GFP (labelled Mut or Mut-iPLA 2 at 10 g/100 mm plate) or mock transfected (no DNA) with FuGENE. Counting of cells began on the second day after transfection (bottom). After counting, cells lysates were prepared and analyzed by western blotting for GFP and actin (top). Error bars represent the mean ± s.d. of three experiments. *P<0.05. (E) Expressing siRNA against iPLA 2 decreases cell proliferation. Mock transfected INS-1 cells, and those transfected with psiRNA-iPLA 2 were counted on the fifth day after transfection (bottom), and the cell lysates were prepared for Western blot analysis for iPLA 2 (top). *P<0.05. A scrambled siRNA construct was transfected to serve as a negative control. Error bars represent the mean ± s.d. of three experiments. with a previous study (Fuentes et al., 2003), inhibition of PAP-1 by 200 M propranolol caused a massive cell death rather than cell-proliferation arrest. This result indicates that the suppression of cell proliferation by BEL is not attributable to inhibition of PAP-1.

Disruption of phospholipid turnover arrests cells in G1 phase
Since phospholipid turnover is a characteristic of G1-phase cells (Jackowski, 1996;Lykidis and Jackowski, 2001), we asked whether inhibition of iPLA 2 -mediated PtdCho degradation affects S-phase entry. We measured 5bromodeoxyuridine (BrdU) incorporation by fluorescenceactivated cell sorting (FACS) after inhibition of iPLA 2 . BrdU incorporation markedly decreased with increasing BEL concentrations; 15 M BEL almost completely blocked the incorporation of BrdU ( Fig. 2A), suggesting that phospholipid turnover in G1 phase is required for cells to enter S phase. Furthermore, staining for the mitosis biomarker phosphohistone H3 (Nowak and Corces, 2004) showed that, in contrast to the G2-M-transition arrest caused by nocodazole treatment, few cells progressed to G2-M transition after 30 hours of BEL treatment (Fig. 2B).
To better characterize this cell-cycle arrest, we treated cells with both BEL and nocodazole and analyzed them by FACS (Agami and Bernards, 2000). Although nocodazole treatment alone showed an accumulation of G2-M-transition cells with time ( Fig. 2C), treatment of cells with both BEL and nocodazole showed a BEL-concentration-dependent increase in the number of cells in G1 phase and a concomitant decrease in S phase and G2-M transition, confirming the G1-phase arrest.
To verify that BEL affects only iPLA 2 , we overexpressed wild-type iPLA 2 -GFP or ARD-iPLA 2 -GFP in INS-1 cells and used FACS to sort cells with varying levels of GFP (Fig. 3A). Consistent with previous data , overexpression of iPLA 2 increased the proliferation rate of INS-1 cells. After FACS sorting, a mixed population (R1 population) of GFPcontaining cells showed a similar cell-cycle distribution for both fusion proteins. When we selected a cell population with higher levels of ARD-iPLA 2 -GFP (R2 or R3), we observed more cells in G1 phase and fewer in S phase and G2-M transition. By contrast, selection for higher iPLA 2 -GFP expression gave a similar cell-cycle distribution, regardless of the gate used (Fig. 3A). These results indicate that cells are arrested in G1 phase by ARD-iPLA 2 in a concentrationdependent manner.
To further confirm that expression of ARD-iPLA 2 prevents cell progression to G2-M transition, we used fluorescence microscopy to analyze transfected cells treated with nocodazole ( Fig. 3B). In the ARD-iPLA 2 -GFP-transfected plate, the cells expressing ARD-iPLA 2 -GFP exhibited interphase characteristics: noncondensed chromosomes and an intact nuclear envelope, visualized by the ring structure of the nuclear pore complex protein mAB414 staining (Fig. 3Bb,c). However, two cells on the same plate that did not express ARD-iPLA 2 -GFP had clearly progressed to G2-M transition, exhibiting condensed chromosomes and disrupted nuclear envelopes ( Fig. 3Bb and c, respectively).
By contrast, the iPLA 2 -GFP transfected cells did not stay in interphase under the same conditions. Several cells progressed to G2-M transition, with obviously condensed chromosomes and broken-down nuclear envelopes. FACS analysis of the cells in the presence of nocodazole showed that 61.5% cells with ARD-iPLA 2 -GFP, but only 37% of the cells with iPLA 2 -GFP, were in G1 phase (Fig. 3C). These data strongly suggest that G1-phase phospholipid turnover is a programmed cellular event required for cell-cycle progression.

Inhibition of iPLA 2 induces upregulation of p53
The above results prompted us to explore the mechanisms by which cells might modulate progression through the cell cycle when G1-phase phospholipid turnover is disrupted. LPA (lysophosphatidic acid), a mitogen that stimulates cell proliferation (Mills and Moolenaar, 2003), is converted from iPLA 2 -generated lysophospholipids by lysophospholipase D. We examined the effect of LPA or Lyso-PC (lysophosphatidylcholine) on cell proliferation when iPLA 2 was inhibited. As shown in Fig. 4, LPA or Lyso-PC did not reverse the inhibitory effect of BEL on cell proliferation under our culture conditions, indicating that iPLA 2 must act directly on phospholipid turnover to control cell proliferation.
Since G1-phase phospholipid turnover is required for cells to enter S phase, we asked whether preventing this turnover would halt cell-cycle progression by activating a protection mechanism similar to the DNA damage checkpoint that induces p53 stabilization and G1-phase arrest. To answer this question, we transfected INS-1 cells with ARD-iPLA 2 -GFP to inhibit cell proliferation and then analyzed p53 expression by western blot. Remarkably, we found that increasing expression of ARD-iPLA 2 -GFP caused a dramatic accumulation of p53 (Fig. 5A). To verify this finding, we treated INS-1 cells with BEL for 30 hours followed by immunoblotting and fluorescence microscopy analyses. Similar to ARDoverexpression, BEL induced considerable increases in p53 levels (Fig. 5B).
In situ immunofluorescent staining showed that p53 colocalized with DAPI, indicating that the protein was in the nucleus (Fig. 5C). p53 acts primarily as a transcription factor and an essential transcriptional target of p53 in the induction of G1-phase arrest is p21 cip1 , a cyclin-dependent kinase inhibitor (CKI). Accumulation of p21 cip1 inhibits cyclin-E/CDK2 activity and the G1-S transition (Kastan and Bartek, 2004;Vogelstein et al., 2000). We found that, although p21 cip1 was barely detected in control cells, it was dramatically elevated in ARD-iPLA 2 -GFP-transfected (Fig. 5A) and BELtreated INS-1 cells corresponding to the accumulation of p53 (Fig. 5B). Our finding reveals first time that the inhibition of iPLA 2 leads to G1-phase arrest by activating the p53/p21 cip1 checkpoint pathway.
Activated cyclin-E/CDK2 phosphorylates p27 kip1 , a CKI that inhibits cyclin-E/cdk2 activity, targeting it for polyubiquitination and degradation (Massague, 2004). This suggests that inhibition of cyclin-E/CDK2 activity by p21 cip1 might in turn lead to stabilization of p27 kip1 . Indeed, immunoblotting showed that p27 kip1 levels were increased in BEL-treated INS-1 cells (Fig. 5B). Moreover, inhibition of the cyclin E/cdk2 kinase results in persistence of the RB-E2F interaction and blocks the transcription of S-phase genes such as cyclin A, which is required for formation of the functional cyclin A/cdk2 complex needed for DNA synthesis (DeGregori et al., 1995). Not surprisingly, levels of cyclin A in BEL-treated INS-1 cells were, therefore, barely detectable (Fig. 5B). Taken together, these results demonstrated that inhibition of iPLA 2 results in a G1-phase arrest through activation of p53, and this, in turn, leads to accumulation of p21 cip1 and the subsequent inhibition of cyclin-E/CDK2 activity.

G1-phase arrest induced by inactivation of iPLA 2 requires p53
To substantiate the role of p53 in G1-phase arrest induced by Journal of Cell Science 119 (6) inhibition of iPLA 2 , we compared the effect of BEL on the cellcycle distribution in isogenic human HCT116 colon cancer (HCT) cells that either expressed wild-type p53 (HCTp53 +/+ ) or not (HCTp53 -/-). Both HCTp53 +/+ and HCTp53 -/cells were treated for 10 hours with increasing concentrations of BEL before FACS analyses. The number of HCTp53 +/+ cells in G1 phase was decreased 50% in the presence of BEL, whereas the number of cells in S phase was decreased threefold compared with cells that had not been treated with BEL (Fig.  6A). Consistent with a role for p53 in this cell-cycle arrest, BEL treatment for 10 hours induced the accumulation of p53 and p21 cip1 in a concentration-dependent manner (Fig. 6B). Interestingly, p27 kip1 levels showed little variation after 10 hours of treatment with increasing concentrations of BEL, confirming that p27 kip1 does not play a major role in this event.
By contrast, HCTp53 -/cells showed no significant change in the number of cells in G1 phase and only a moderate decrease in the number of cells in S phase and G2-M transition (Fig. 6A). This suggests that the effect of BEL on HCTp53 -/cells is not phase-specific and that p53 is required for G1-phase arrest induced by a disruption of the phospholipid turnover.
To further demonstrate the role of p53 in this G1-phase arrest, we analyzed the proliferation of both HCTp53 +/+ and HCTp53 -/cells after 28 hours of BEL treatment. Similar to INS-1 cells, this extended BEL treatment inhibited the proliferation of HCTp53 +/+ cells without resulting in cell death (Fig. 6C), and caused the accumulation of p53, p21 cip1 , and p27 kip1 in a concentration-dependent manner (Fig. 6D). By contrast, HCTp53 -/proliferation was only partially inhibited by treatment with 15 M BEL (Fig. 6C).
We observed that p21 cip1 was slightly upregulated in HCTp53 -/cells (Fig. 6D). Since p21 cip1 has also been implicated in G2phase arrest (Bunz et al., 1998), this upregulation might have contributed to the accumulation of G2-M-transition cells (Fig.  6A) and the partial inhibition of cell proliferation in a p53-independent mechanism. Together, our data indicate that -although we cannot rule out a function in additional pathways -p53 plays a crucial role in the link between the inhibition of iPLA 2 and G1-phase arrest. Disruption of this link through loss of p53 permits unchecked proliferation despite the inhibition of G1-phase phospholipid turnover. p21 cip1 is a key effector of the p53-dependent G1-phase arrest induced by iPLA 2 inhibition To determine the role of p21 cip1 in the p53-dependent G1-phase arrest induced by iPLA 2 inhibition, we examined the effect of BEL on HCTp21 -/cells, one of several isogenic colon carcinoma cell lines. Surprisingly, we observed increasing rates of cell death when BEL concentrations increased in these p21deficient cells (Fig. 7A). Cell death was not observed in HCTp53 -/cells after BEL treatment (Fig. 6C). A hallmark of apoptosis is phosphatidylserine (PS) externalization (Fadok and Henson, 1998). Further analysis of PS externalization by staining with Annexin-V-Fluos showed that BEL treatment of HCTp21 -/cells induced early apoptosis and secondary necrosis (Fig. 7B). By contrast, BEL treatment did not induce apoptosis in HCT wild-type cells (Fig. 7B). These results suggest that inhibition of iPLA 2 in the absence of p21 cip1 activates the p53-dependent apoptotic pathway (Vousden and Lu, 2002). To further determine whether the apoptosis in the p21-deficient cells was p53-dependent, we treated p21deficient cells with small interference RNA (siRNA) against p53 (Fig. 7C). Indeed, downregulation of p53 in p21-deficient cells overcame the BEL-induced apoptosis (Fig. 7C). Taken together, these results demonstrate that p21 cip1 is a key effector of the p53-dependent G1-phase arrest in response to inactivation of iPLA 2 , and depletion of p21 cip1 leads to p53dependent apoptosis.

Discussion
It has been long recognized that membrane phospholipid metabolism is regulated in tune with the cell cycle. However, little is known about the key regulatory enzymes or the biochemical mechanisms that link phospholipid levels to cellcycle control. PtdCho is the major membrane phospholipid in mammalian cells, and regulation of its biosynthesis and turnover is crucial in maintaining membrane structure and function (Kent, 1995).
It has been suggested that mammalian cells have developed a system to maintain a constant membrane phospholipid content during G1 phase by coordinating PtdCho biosynthesis and degradation through the opposing actions of CCT and iPLA 2 (Jackowski, 1996;Baburina and Jackowski, 1999;Barbour et al., 1999). Whereas the bulk of phospholipid production is needed to construct cell membranes for cell Journal of Cell Science 119 (6)  division, iPLA 2 -mediated deacylation of PdtCho might also act as a feedback control to ensure membrane in optimal condition. Indeed, during mitosis, the cellular membranes undergo dramatic structural and functional changes, such as the complete breakdown and reformation of the nuclear envelope, and the fragmentation and stack reformation of the Golgi membrane (Burke and Ellenberg, 2002;Colanzi et al., 2003). After mitosis, the cell must restructure or remodel its membrane system to ensure proper DNA synthesis and cellcycle progression. This suggests that the feedback mechanism is capable of restraining cell-cycle progression until particular conditions are satisfied.
Here, we report that inhibition of iPLA 2 arrests cells in the G1 phase of the cell cycle by inducing p53 accumulation and expression of p21 cip1 . We show that DNA synthesis is blocked and cells remain in interphase after inhibition of iPLA 2 , as evidenced by lack of biomarkers of mitosis, DNA condensation and nuclear envelope breakdown. Moreover, even in the presence of nocodazole, the number of cells in G1 phase increases dramatically with increasing BEL concentrations, suggesting a rapid G1-phase arrest.
Our results demonstrate that, G1-phase phospholipid turnover is essential for cell progression and iPLA 2 is a key component for maintaining G1-phase phospholipid homeostasis by mediating PdtCho degradation. iPLA 2 hydrolyzes the sn-2 fatty acyl bond of phospholipids to liberate free fatty acids and lysophospholipids . It has been proposed to play a role in signal transduction through the actions of the hydrolytic products of iPLA 2 , which vary according to the particular cell type and The DNA contents of each sample was analyzed by FACS. Data represent the mean of three experiments. (B) Accumulation of p53 and expression of p21 increases with increasing concentrations of BEL. HCTp53 +/+ and HCTp53 -/cells were treated with increasing concentrations of BEL for 10 hours. Cell lysates were prepared and analyzed by western blot for p53, p21, p27 and actin. (C) Inhibition of iPLA 2 with increasing concentrations of BEL dramatically inhibits the proliferation of HCTp53 +/+ but only mildly inhibits the proliferation of HCTp53 -/cells. HCTp53 +/+ cells (blue bars) and HCTp53 -/cells (red bars) were cultured with increasing concentrations of BEL in 24-well microplates for 28 hours, followed by 6 hours of BrdU labelling. Cell proliferation was then determined on a Quant microplate reader. Error bars represent the mean ± s.d. of three experiments; *P<0.05, **P<0.005. (D) Accumulation of p53 and expression of p21 increases with increasing concentrations of BEL in HCTp53 +/+ (left). In HCTp53 -/cells, by contrast, p21 expression increased only slightly (right). Both HCTp53 +/+ and HCTp53 -/cells were treated with increasing concentrations of BEL for 28 hours in 24-well microplates. Cell lysates were prepared and analyzed by western blot for p53, p21, p27 and actin. situation. For example, it plays a role in insulin secretion in pancreatic islet ␤-cells Song et al., 2005) and in the generation of lysophospholipids that activate storeoperated Ca 2+ (SOC) channels and capacitative Ca 2+ influx in smooth-muscle cells (Smani et al., 2004). It also releases LPC to recruit phagocytes to clear apoptotic debris (Lauber et al., 2003).
It has also been proposed that iPLA 2 plays the housekeeping role in phospholipid metabolism, which includes participation in the remodelling of membrane phospholipids (Balsinde et al., 1997;Balsinde et al., 1995) and the regulation of membrane phospholipid turnover (Baburina and Jackowski, 1999;Barbour et al., 1999). This represents a feedback mechanism to fine-tune the phospholipid composition of membranes and to monitor changes in phospholipid mass during cell-cycle progression. In addition, iPLA 2 -like several proteins involved in cell-cycle control -contains an ankyrin-repeat domain (Ma et al., 1997;Tang et al., 1997). Cell-cycle progression is governed by cyclin-dependent kinases (CDKs) that are activated by cyclinbinding (Morgan, 1995) and inhibited by the CDK inhibitors (Sherr and Roberts, 1999). Inhibitors of the CDK4 (INK4) family include four proteins with four or five tandem ankyrin repeats: p16 INK4a , p15 INK4b , p18 INK4c and p16 INK4d . These proteins bind to and inhibit CDK4 and CDK6 through their ankyrin repeat domains (Baumgartner et al., 1998;Serrano, 1997). In addition, the p53-binding protein 2 (53BP2) and the oncoprotein gankyrin use ankyrin-repeat domains to interact with their partners (Gorina and Pavletich, 1996;Higashitsuji et al., 2000;Krzywda et al., 2004). Our data, suggest that iPLA 2 -with its structural similarities to several cell-cycle control proteins and its housekeeping role in phospholipid metabolism -couples membrane phospholipid turnover to cell-cycle control.
Many previous studies have shown that p53 mediates growth arrest or apoptosis in response to many types of stress: DNA damage, telomere attrition, oncogene activation, hypoxia as well as loss of normal growth and survival signals (Ryan et al., 2001). Recently, it has also been shown that glucose restriction induces activation of a reversible cell-cycle arrest through AMP-activated protein kinase (AMPK)-mediated activation of p53 (Jones et al., 2005). Cell membranes are essential components of all cells; without them the cell can not assert its identity or subdivide its functions into specialized compartments. In addition to this structural role, membrane phospholipids are a reservoir from which cells generate intracellular and intercellular messengers to regulate cell function. The present data indicate that p53, in addition to responding to the types of cellular stress mentioned above, is also essential for cell-cycle arrest induced by phospholipid metabolic stress caused by the inhibition of iPLA 2 .
Importantly, inactivation of iPLA 2 fails to arrest p53deficient cells in G1 phase, indicating that this cell-cycle arrest is p53-dependent. p53 exerts its cell-cycle arrest through its transcriptional target, the cyclin-dependent kinase inhibitor Journal of Cell Science 119 (6) Fig. 7. BEL treatment induces p53-dependent apoptosis in p21deficient HCT cells. (A) HCTp21 -/cells have a high death rate after inactivation of iPLA 2 by BEL. HCTp21 -/cells (5ϫ10 5 ) were treated with varying concentrations of BEL for 28 hours. Viable and nonviable cells were determined by Trypan Blue staining and counted. Error bars represent the mean ± s.d. of six experiments. (B) Inactivation of iPLA 2 induces apoptosis of p21-deficient cells. Both wild-type HCT and HCTp21 -/cells were incubated with or without BEL for 10 hours and then analyzed for apoptosis by Annexin-V-Fluos staining and FACS. R1 represents living cells. R2 represents cells undergoing early apoptosis, and R3 representes secondary necrotic cells. Each panel shows a typical flow-cytometric histogram of 10,000 cells/sample from one of three experiments. (C) BEL-associated death of HCTp21 -/cells can be prevented by p53 depletion. HCTp21 -/cells were cultured overnight in the presence of BEL with 1 nmol of SMARTpool-p53. dsRNA was used as a negative control. The cells were harvested for both western blot analysis for p53 (top) and cell death analysis (bottom). Recombinant TNT p53 was loaded as indicated. *P<0.05.
(CKI) p21 cip1 . Accumulation of p21 cip1 inhibits cyclin-E/CDK2 activity and, therefore, the G1-S transition (Kastan and Bartek, 2004;Vogelstein et al., 2000). Interestingly, we also observed that p21 cip1 is slightly upregulated in HCTp53 -/cells after BEL treatment, which is correlated to an increase in the number of G2-M-transition cells, suggesting the existence of a p53-independent G2-M-transition arrest through p21 expression (Niculescu et al., 1998). This might explain the apparently unchanged phasic distribution in some cell lines after BEL treatment alone (Sanchez and Moreno, 2001).
Furthermore, we showed that HCTp21 -/cells undergo extensive apoptosis when treated with the iPLA 2 inhibitor BEL, indicating that p21 cip1 is required for p53-dependent G1phase arrest. The fact that this apoptosis can be overcome by siRNA downregulation of p53 further indicates that p53 is essential for the cell-cycle arrest and might, in some cases, commit cells to apoptosis. Disruption of this checkpoint through loss of p53 permits uncontrolled proliferation in response to the disturbance of phospholipid homeostasis, which may explain the viability of iPLA 2 knockout mice (Bao et al., 2004).
In normal cells, p53 activity is maintained at low levels by the MDM2 protein, which inhibits the transcriptional activity of p53 and targets it for degradation (Haupt et al., 1997;Kubbutat et al., 1997). Cellular insults trigger posttranslational modification of p53, including phosphorylation of Ser15, which stabilizes the protein in part by disrupting its interaction with MDM2 (Shieh et al., 1997). This process is mediated by the ataxia telangiectasia-mutated (ATM) kinase (Banin et al., 1998;Canman et al., 1998) and ataxia telangiectasia and Rad-3-related (ATR) kinase (Tibbetts et al., 1999) or by AMPK (Jones et al., 2005). Whether accumulation of p53 by inhibition of iPLA 2 uses the similar pathways remains to be determined.
iPLA 2 contains an ankyrin-repeat domain (Ma et al., 1997;Tang et al., 1997), and it is interesting that 53BP2 and Ankrd2 proteins bind to p53 through their ankyrin-repeat domains (Gorina and Pavletich, 1996;Kojic et al., 2004). BEL irreversibly inhibits iPLA 2 through a covalent modification (Hazen et al., 1991) that may cause conformational changes, which expose the ankyrin repeat domain of iPLA 2 to p53. Moreover, we showed that forced expression of ARD-iPLA 2 enhances p53 levels and G1-phase arrest in a concentration-dependent manner. Interestingly, the endogenous expression of an ankyrin-repeat domain of iPLA 2 has recently been described by another group of researchers, who postulate that this expression serves as a way to regulate iPLA 2 activity throughout the cell cycle (Manguikian and Barbour, 2004). It remains to be seen whether alternative splicing of iPLA 2 represents a way for the protein to be involved in cell-cycle control by interaction with p53.
In summary, our study reveals hitherto unrecognized cooperation between p53 and iPLA 2 to monitor membrane phospholipid turnover in G1 phase. iPLA 2 has been shown to mediate proliferation of human peripheral blood B and T lymphocytes (Roshak et al., 2000), murine 3T6 fibroblasts (Sanchez and Moreno, 2001) and human colorectal carcinoma (Caco-2) cells (Sanchez and Moreno, 2002), suggesting that this effect is not cell-type specific. Our data suggest that G1phase phospholipid turnover is systematically monitored in G1-phase cells and that the cells respond to disruption of this turnover by turning on protective mechanisms.

DNA constructs
Wild-type iPLA 2 -GFP and ARD-iPLA 2 -GFP expression vectors were constructed by ligation of cDNA of either the full-length or the 1375 bp ankyrin-repeat-domain of iPLA 2 , in-frame upstream of GFP in the vector pEGFP-N2 (Clontech), respectively. The Mut-iPLA 2 -GFP expression vector was constructed by mutating Ser to Ala in the catalytic motif GXSXG in wild-type iPLA 2 -GFP using the QuickChange site-direct mutagenesis kit (Stratagene) according to the manufacturer's protocol. The primers used for to generate the mutation by PCR are 5Ј-GGGTGGCAGGAACCGCCACAGGCGGCATCCTGG-3Ј and 5Ј-CCAGGA-TGCCGCCTGTGGCGGTTCCTGCCACCC-3Ј. The psiRNA-iPLA 2 construct has been described previously (Song et al., 2005).

Cell culture, transfection and treatments
Cells of the insulinoma (INS-1) cell line and human HCT116 colon cancer cells [HCT cells expressing wild-type p53 +/+ (HCTp53 +/+ , and its isogenic p53-deficient (HCTp53 -/-) and p21 cip1 -deficient (HCTp21 -/-) cells] were cultured as described elsewhere (Bunz et al., 1998;Song et al., 2005). FuGENE 6 Transfection Reagent (Roche) was used for transient transfection of cells according to the manufacturer's instructions. For BEL treatment, the desired numbers of cells were seeded on plates to which variable concentrations of BEL (freshly prepared from stocks kept at -80°C, Cayman) were added simultaneously. Because BEL is stable in aqueous solutions for about 12 hrs, BEL was added every 12 hrs.

Depletion of p53 by siRNA
We used SMARTpool p53 (Upstate/Dharmacon) to deplete p53 in a p21-deficient cell line, HCTp21 -/-. HCTp21 -/cells were freshly seeded and cultured overnight. SMARTpool p53 (1 nmol ) was transfected into p21-deficient cells (3ϫ10 6 ) in a 75-cm flask with FuGENE (Roche). Negative control scramble siRNA (Upstate/Dharmacon) was applied at the same time. The transfected cells were split into 12-well plates with fresh BEL on the second day. The rate of cell death was calculated 20 hours after BEL treatment with Trypan Blue staining. Cell lysates were prepared and analyzed for p53 expression by western blot after 10 hours of BEL treatment.

Measurement of cell proliferation, and apoptosis viability
Cell proliferation was determined by cell counting and BrdU incorporation ELISA (colorimetric) kit (Roche). For cell counting, desired numbers of cells were plated on 6-well plates or 60-mm Petri-dishes and cultured for various periods of time. Then, the cells were harvested, stained by Trypan Blue and counted. The number of dead cells was estimated by counting the Trypan Blue-stained cells. For BrdU incorporation analysis, cells were cultured in 24-well microplates in the presence or absence of BEL for 24 hours and then incubated with BrdU in the presence of BEL for an additional 4-10 hours. BrdU incorporation ELISA was performed according to the manufacturer's protocol and absorbance was determined by a Quant microplate reader (BIO-TEK Instruments). Cell apoptosis was analyzed with an Annexin V-FLUOS Staining Kit (Roche Molecular Biochemicals) according to the manufacturer's protocol. The labeled cells were immediately analyzed by flow cytometry on a Becton Dickinson FACSCalibur (Heidelberg, Germany). Living cells were negative in both annexin V and PI staining. Early apoptotic cells were annexin V-positive and PI-negative. Secondary necrotic cells were positive in both annexin V and PI staining.

Detection of BrdU incorporation by FACS
Detection of BrdU-labelled DNA in proliferating cells was accomplished with an in-situ cell proliferation kit FLUOS (Roche). The cells were plated in the presence or absence of the desired concentrations of BEL and incubated for 24 hours. They were then labelled with BrdU or not, and subsequently stained according to the manufacturer's flow-cytometry protocol. Cells were counterstained with 1 g/ml propidium iodide (PI, Sigma) in phosphate-buffered saline (PBS), and analyzed with FACS.

Cell-cycle analysis by FACS
Around 10 5 cells were grown and treated with the respective compounds for various periods of time. After harvesting by centrifugation, the cells were fixed with 70% ethanol and treated with 5 g/ml RNaseA followed by PI staining (1 g/ml in PBS). Cells were analyzed using a FACScan (Becton Dickinson). Doublet exclusion was performed by gating for singlet cells on a height vs area plot. The data was processed with the CellQuest TM Pro (BD Biosciences) for determination of cell numbers in G1 and S phases, and G2-M transition. GFP fluorescence was excited at 488 nm and the emission was measured with a 510 nm filter. PI staining was detected with a 580 nm filter.

Statistical analysis
Data were expressed as mean ± s.d. The statistical significance of differences was analyzed using Student's t-test, where P<0.05 was considered significant.