HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 20, 18206-18214, May 17, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/20/18206    most recent M201028200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Itahana, K. Articles by Campisi, J. Search for Related Content PubMed PubMed Citation Articles by Itahana, K. Articles by Campisi, J. Social Bookmarking                      What's this?

A Role for p53 in Maintaining and Establishing the Quiescence Growth Arrest in Human Cells^*

Koji Itahana, Goberdhan P. Dimri^§, Eiji Hara^¶, Yoko Itahana, Ying Zou, Pierre-Yves Desprez, and Judith Campisi^**

From the  Life Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720, the ^§ Department of Radiation Oncology, New England Medical Center, Boston, Massachusetts 02111, ^¶ Paterson Institute for Cancer Research, Manchester, M20 4BX, United Kingdom, and  California Pacific Medical Center, San Francisco, California 94115

Received for publication, January 30, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The p53 tumor suppressor protein induces transient growth arrest or apoptosis in response to genotoxic stress and mediates the irreversible growth arrest of cellular senescence. We present evidence here that p53 also contributes to the reversible, growth factor-dependent arrest of quiescence (G₀). Microinjection of expression vectors encoding either MDM2 or a pRb-binding mutant of SV40 T antigen, both of which abrogate p53 function, stimulated quiescent normal human fibroblasts to initiate DNA synthesis and were 40-70% as effective as wild-type T antigen. Electrophoretic mobility shift and p53 transactivation assays showed that p53 activity was higher in quiescent and senescent cells compared with proliferating cells. As proliferating cells entered G₀ after growth factor withdrawal, the p53 mRNA level increased, followed by transient accumulation of the protein. Shortly thereafter, the expression (mRNA and protein) of p21, a p53 target gene and effector of cell cycle arrest, increased. Finally, stable expression of the HPV16 E6 oncogene or dominant negative p53 peptide, GSE-22, both of which inhibit p53 function, delayed entry into quiescence following growth factor withdrawal. Our data indicate that p53 is activated during both quiescence and senescence. They further suggest that p53 activity contributes, albeit not exclusively, to the quiescent growth arrest.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Complex organisms contain postmitotic cells, which cannot proliferate, and mitotic cells, which proliferate when appropriately stimulated or a need arises for cell replacement or tissue repair. In the absence of signals that favor proliferation, mitotic cells arrest growth in a reversible state termed quiescence or G₀. A fundamental difference between postmitotic and mitotic cells is that only mitotic cells give rise to cancer.

Cell proliferation (used here interchangeably with growth) generally requires release from negative growth signals and stimulation by positive signals. Much of our knowledge about these processes originates from studies of fibroblasts in culture. Subconfluent fibroblasts grow exponentially in culture, and proliferation depends on the presence of growth factors, typically supplied as serum. When growth factors are withdrawn, the cells cannot pass a restriction point in late G₁, and they stably but reversibly arrest growth in G₀. Fibroblasts, and many other cell types, can remain in G₀ for long periods without undergoing apoptosis, providing nutrients and survival factors are present. When growth factors are restored, quiescent cells initiate DNA synthesis after a lag that exceeds the duration of G₁ (1).

Mitotic cells can also enter an essentially irreversible growth-arrested state termed senescence. Normal cells enter senescence when challenged with potentially oncogenic stimuli such as short telomeres, DNA damage, or certain oncogenes (2). A significant difference between quiescent and senescent cells is their response to growth factors. In both cases, growth factors induce the expression of a common set of genes. However, whereas quiescent cells resume proliferation, senescent cells remain arrested with a G₁ DNA content. Most cancer cells lose the ability to become quiescent and/or senescent in response to appropriate signals (1, 2).

Among the most common events during carcinogenesis is a loss of p53 function. p53 is a transcription factor that interacts with multiple proteins. It lies at the heart of a major tumor suppression pathway comprised of upstream regulators and downstream effectors (3). p53 is best known for its ability to transiently arrest the cell cycle or induce apoptosis in response to DNA damage (4, 5). p53 is also a critical regulator of the senescence growth arrest (2, 6-8).

Both quiescent and senescent human fibroblasts initiate DNA synthesis, in the absence of growth factors, in response to SV40¹ T antigen (9-12), a viral oncogene that binds and inactivates both p53 and the retinoblastoma (pRb) tumor suppressor. In contrast to p53, pRb is a cell cycle regulator that controls, among other processes, progression from G₀ and G₁ into S phase (13, 14). Both p53 and pRb are critical for the growth arrest of senescent cells as indicated by their response to wild-type and mutant T antigens, dominant negative p53 mutants, and antisense oligonucleotides (15-18). By contrast, although a role for pRb in the quiescent growth arrest is established, a role for p53 in quiescence has been less apparent.

Several lines of evidence indirectly suggest that p53 may function in quiescence. First, a T antigen mutant defective in binding pRb, but not p53, stimulated quiescent, but not senescent, cells to synthesize DNA (11, 19). Second, proliferating, senescent, and quiescent cells displayed different p53 phosphorylation patterns (20), suggesting that p53 functions differently in these growth states. Third, quiescent and senescent cells express high levels of p21 (21, 22), a cyclin-dependent protein kinase inhibitor (CDKI) that negatively regulates pRb (23) and is a target for p53 transactivation (24). It is not known if p53 is responsible for the expression of p21 by quiescent cells since p21 can be induced by p53-independent mechanisms (25). Nonetheless, loss of p21 function stimulated quiescent cells to initiate DNA synthesis (22). On the other hand, p27, a CDKI not known to be induced by p53, has been implicated in maintaining the quiescent growth arrest (26-29).

Here, we explore the role of p53 in the quiescent growth arrest of human fibroblasts. We show that p53 binding and transactivation activities are elevated in quiescent cells. Following growth factor withdrawal, there is rapid induction of p53 mRNA and protein, followed by an increase in p21 mRNA and protein. Furthermore, abrogation of p53 function delays entry into G₀ following growth factor withdrawal. Our findings suggest that p53 may play a role in establishing and maintaining the quiescent state of normal cells.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Culture-- WI-38 human fibroblasts were obtained, cultured, and assessed for quiescence or senescence as described (30). Briefly, subconfluent cells (1500-4000/cm²) were made quiescent by washing with serum-free medium and incubating in 0.2% serum for 3-4 days. They were made senescent by serial passage in 10% serum. Cells were given 10 µCi/ml [³H]thymidine for 1 (quiescence) or 3 (senescence) days to determine the percent radiolabeled nuclei (% LN) after autoradiography (30). Cultures with >70% LN were considered presenescent, and <10% LN were considered quiescent or senescent. U2OS, Saos2, and GM05823 cells were from the American Type Culture Collection and Coriell Cell Repositories, and AT2SF cells were from E. Blakely (Lawrence Berkeley National Laboratory, Berkeley, CA).

Vectors and Retroviruses-- CMV1, CMV-T, CMV-T[K1], and CMV-gal have been described (12, 30). To construct CMV-T[p53], the BamHI-BglI fragment of Tdl434-444 (from J. Tevethia, Pennsylvania State College of Medicine, Hershey, PA), which encodes T antigen lacking amino acids 434-444 (p53 binding-defective) (31), was cloned into CMV1. CMV-MDM2 was from B. Vogelstein (Johns Hopkins University, Baltimore, MD). Cells producing LXSN-HPV16-E6 retrovirus were provided by V. Band (New England Medical Center, Boston, MA) and D. Galloway (Fred Hutchinson Cancer Center, Seattle, WA), who also provided the mutant E6 retroviral vector (LXSN-HPV16-E6-8S9A10T). pBabe-GSE22 encodes an interfering p53 fragment and was from A. Gudkov (Lerner Research Institute, Cleveland, OH). CMV-p53 was from J. R. Smith (University of Texas, San Antonio, TX). pBKCMV was from Stratagene. Infectious virus was produced, and cells were infected and selected as described (32).

Microinjection-- Cells were microinjected with plasmid DNA as described (12). Briefly, nuclei of subconfluent cells were injected with 10-20 ng/µl expression vector and/or 5-10 ng/µl CMV-gal. [³H]thymidine was added 30-60 min later. After 3 days, cells were stained for -galactosidase (-gal) activity to identify injected cells and processed for autoradiography to identify cells that synthesized DNA.

Electrophoretic Mobility Shift Assays-- Nuclear extracts were prepared as described (30). p53 binding activity was detected by elec- trophoretic mobility shift assays using oligonucleotides containing the human p21 promoter p53 binding site (24) that was either wild-type (binding site underlined) 5'-GGTCAGGAACATGTCCCAACATGTTGACC-3'), or mutant (mutant bases in lowercase): 5'-GGTCAGGAAtggaTCCCAAacctTTGACC-3'). AP1 and Sp1 binding sites were 5'-GGCAGTCAGTTTTCATTTAAAAATG-3' and 5'-ATTCGATCGGGGCGGGGCGAGC-3', respectively. Oligonucleotides were end-labeled with polynucleotide kinase and gel-purified prior to use. Where indicated, extracts were preincubated with 0.1 µg of p53 (Ab6; DO-1) or Myc (Ab1) antibodies (Calbiochem) for 1 h on ice prior to binding. Extracts (1-2 µg of protein) were incubated in binding buffer (60 mM NaCl, 20 mM HEPES, 2.5 mM MgCl₂, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1 µg of poly(dI·dC) or 3 µg of poly(dA·dT)) for 2 min at 25 °C, labeled oligonucleotide (10⁴ cpm) was added, and incubation was continued for 20 min. Protein-DNA complexes were separated by elec- trophoresis for 2 h at 150 V through 5% nondenaturing polyacrylamide gels in 0.25× Tris borate-EDTA buffer. The gels were dried and exposed for autoradiography.

Reporter Assays-- The p53 reporter vector used the p53 responsive element in the human ribosomal gene cluster (33). Either 13 copies of the wild-type element (PG13; binding site underlined) 5'CCTGCCTGGACTTGCCTGG-3' or 15 copies of a mutant element (5'-CCTTAATGGACTTtaaTGG-3' (MG15: mutant bases in lowercase) were inserted into the NarI-EcoRI site of pAD (adenoviral minimal promoter upstream of the bacterial lacZ gene; CLONTECH). Cells (5-8 × 10³/cm² on 35-mm dishes) were transfected 24 h after plating with 0.5 µg of PG13, MG15, or control (pAD) plasmids and 0.1 µg of pCMV-Luc (made by replacing the lacZ gene in pCMV--gal (30) with the luciferase gene in pGL2 from Promega) using LipofectAMINE-Plus (Invitrogen). -galactosidase and luciferase activities were measured using Galacto-Star (Tropix) and luciferase assay systems (Promega) kits as directed by the suppliers. -galactosidase activity was normalized to luciferase activity.

Northern Analysis-- Total cellular RNA (15-20 µg), purified by the total RNA isolation system (Promega), was analyzed by Northern blotting as described (30). The blot was hybridized to ³²P-labeled cDNA probes and rehybridized to QM or Gi2 cDNA probes (34) to control for RNA integrity and quantity. Signals were quantified by densitometry.

Western Analysis-- Cells were washed with phosphate buffered saline (PBS), lysed in Laemmli sample buffer (35) lacking -mercaptoethanol, and analyzed immediately or frozen at 80 °C. Prior to analysis, -mercaptoethanol was added, and samples were heated at 95 °C for 10 min. Proteins were separated in 10 or 4-15% denaturing polyacrylamide gels and transferred to polyvinylidene difluoride membranes (Immobilon P, Millipore) by electrophoresis. Membranes were blocked with 10% nonfat milk in Tris borate saline (TBS)-Tween and incubated with primary and secondary antibodies, washing with TBS-Tween after each step. Secondary antibodies were detected using a chemiluminescence kit (ECL Plus, Amersham Biosciences). Antibodies recognizing p53 (Ab6; DO-1), Myc (Ab1), and -tubulin (Ab1) were from Calbiochem, anti-phospho-p53 (Ser-15) (9284) was from Cell Signaling, anti-p21 (6B6) was from PharMingen, and anti-QM (C-17) was from Santa Cruz Biotechnology. Signals were quantified by densitometry.

Immunofluorescence-- Cells were cultured in 4-well glass slide chambers, fixed with 3.7% formaldehyde in PBS, washed with PBS, and permeabilized with 0.5% Triton X-100 in PBS. Slides were blocked with 10% nonfat milk in PBS, incubated with primary and secondary antibodies in blocking solution for 1 h each, and mounted in VectaShield containing DAPI (Vector Laboratories) to visualize nuclear DNA. Cells were viewed by epifluorescence. The p21 (6B6) and Ki67 (NCL-Ki67p) primary antibodies were from Novocastra Laboratories.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

T Antigen Mutants Stimulate Quiescent Cells to Synthesize DNA-- We first determined the extent to which p53 and pRb inactivation contributes to the ability of T antigen to stimulate DNA synthesis in quiescent or senescent cells. We made proliferating normal human fibroblasts quiescent by depriving them of serum (0.2%, 3 days) senescent cels, which do not proliferate, were similarly serum-deprived. We then microinjected control (CMV1) or T antigen (CMV-T, CMV-T[K1] or CMV-T[p53]) expression vectors with a marker vector (CMV-gal), and incubated the cells with [³H]thymidine for 72 h. We identified injected cells by histochemical staining for the marker (-galactosidase, or -gal), and scored -gal-positive cells for radiolabeled nuclei, indicative of DNA synthesis.

As expected, T antigen stimulated 70-75% of senescent and quiescent cells to synthesize DNA compared with <5% stimulation by the control vector (Fig. 1). We used two T antigen mutants to determine the contributions of p53 and pRb inactivation. T[K1] has a point mutation (E107K) that prevents binding to pRb and related proteins, but not p53 (36). T[p53] has a small deletion (residues 434-444) that abolishes p53, but not pRb, binding (31). Either mutation substantially reduced the ability of T to stimulate senescent cells (10-35% of wild-type activity) but only modestly reduced stimulation of quiescent cells (70-75% of wild-type activity) (Fig. 1). The mutants complemented each other (Fig. 1), confirming they were not markedly compromised in functions other than pRb or p53 binding (31, 36). These results verify that both p53 and pRb inactivation are needed to overcome the senescence arrest. Although T antigen has functions other than p53 and pRb binding (37), the results further suggest that inactivation of either p53 or pRb can overcome quiescence.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   DNA synthesis induced by T antigens and/or MDM2 in quiescent or senescent cells. WI-38 normal human fibroblasts were made quiescent or senescent, microinjected with control (CMV1) or expression vector plasmids, labeled with [3H]thymidine for 3 days, and processed to quantify the percentage of cells that synthesized DNA (% LN) as described under "Experimental Procedures." Expression vectors encoded wild-type T antigen (T), T antigen defective in pRb binding T[K1], T antigen defective in p53 binding T[p53], or human MDM2 (MDM2). Given below each bar are the total number of injected cells that were analyzed (two-five experiments) and the expected status (active = +; inactive = ) of p53 and pRb in the injected cells.

To obtain evidence that T antigen stimulates quiescent cells in part by inactivating p53 we co-injected an expression vector (CMV-MDM2) encoding MDM2, a cellular protein that enhances p53 degradation (38), together with CMV-T[p53]. MDM2 overexpression restored the ability of T[p53] to stimulate DNA synthesis in senescent and quiescent cells (from 35 to 85% and 70 to 100% of wild-type activity, respectively) (Fig. 1). Moreover, MDM2 overexpression alone stimulated ~25% of quiescent cells to synthesize DNA, supporting the idea that p53 participates, albeit not exclusively, in maintaining quiescence.

p53 Activity in Quiescent Cells-- p53 DNA binding and transactivation activities were shown to increase in senescent cells (6, 39, 40), but little is known about p53 activity in quiescent cells. We therefore measured p53 DNA binding activity in nuclear extracts from quiescent human fibroblasts using electrophoretic mobility shift assays and the p53 recognition sequence in the human p21 promoter. Quiescent cells had ~4-fold higher p53 DNA binding activity than proliferating cells, similar to the level present in senescent cells (Fig. 2). However, although the activity was serum-dependent in presenescent cells, it was elevated regardless of the presence of serum in senescent cells (Fig. 2). We verified that binding to the probe was p53-specific by comparing extracts from cells carrying either wild-type (U2OS) or deleted (Saos2) p53 genes, demonstrating competition by wild-type, but not mutant, oligonucleotides, and disrupting the complex with an anti-p53, but not an irrelevant, antibody (Fig. 2). SP1 DNA binding served to verify the concentration and integrity of the extracts (Fig. 2).

View larger version (95K): [in this window] [in a new window]   Fig. 2.   p53 binding activity in quiescent and senescent cells. Nuclear extracts were prepared from presenescent (Presen) or senescent (Sen) WI-38 cells or proliferating U2OS or Saos2 cells and analyzed as described under "Experimental Procedures." Cells were incubated in 0.2% serum (serum) for 3 days to render presenescent cells quiescent and senescent cells were equivalently serum-deprived. Serum-deprived cells were also stimulated with fresh 10% serum for 20 h (+serum). Extracts were incubated with radiolabeled probes containing a p53 (p53 Binding) or SP1 (Sp1 Binding) site. Where indicated, extracts from quiescent cells were preincubated with 100-fold excess unlabeled probes (C) containing wild-type (wt) p53, mutant (mt) p53, or irrelevant (AP1) binding sites, or antibodies (Ab) reactive against p53 (p53 Ab) or Myc (Myc Ab). p53-specific binding is indicated by the arrow. Free, uncomplexed probe.

We also measured p53 transactivation activity in quiescent human fibroblasts (Fig. 3). We constructed a -gal reporter vector containing wild-type or mutant p53 binding sites (33) upstream of the adenovirus late promoter. We introduced reporter, expression, and normalization vectors into cells using a protocol that achieved 10-15% transient transfection efficiency. We first verified the responsiveness of p53RE by transfecting it into proliferating cells together with control (pBKCMV) or p53 (CMV-p53) expression vectors. Relative to reporter activity driven by endogenous p53 (pBKCMV), cells transfected with CMV-p53 expressed ~20-fold higher activity (Fig. 3A). CMV-p53 also stimulated reporter activity from p53RE containing wild-type, but not mutant, p53 binding sites in Saos2 (p53 null) cells (not shown). We then introduced control (pAD) or p53 (p53RE) reporter vectors into presenescent cells, either proliferating (+) or made quiescent by serum-deprivation (), or senescent cells cultured with serum (+) or serum-deprived (). Relative to proliferating cells, p53-specific transactivation activity was modestly but significantly higher (average 2.7-fold, five experiments) in quiescent cells (Fig. 3B). As expected, activity was also higher in senescent cells (3.6- and 2.8-fold, + and serum, respectively) (Fig. 3B). To verify the specificity of the activity, we expressed the HPV-16 E6 oncoprotein, which accelerates p53 degradation (41), in WI-38 cells and then transfected the cells with the p53 RE vector. p53RE reporter activity was minimal in these cells, consistent with their minimal p53 activity (Fig. 3B). In addition, the p53RE vector containing mutant p53 binding showed little or no activity in normal cells, whether proliferating or quiescent (Fig. 3C). Taken together, these data indicate that p53 DNA binding and transactivation activities are elevated in quiescent cells.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   p53 transactivation activity in quiescent and senescent cells. A, p53 responsiveness of the p53RE reporter vector. Proliferating WI-38 cells were transiently co-transfected with a normalization vector, p53 reporter vector (p53RE), and either a control (pBKCMV) or p53 expression vector (CMVp53). Cell lysates were assayed for the p53 (-galactosidase) and normalization (luciferase) reporters as described under "Experimental Procedures." Normalized p53 Reporter Activity is -galactosidase/luciferase activity, with the activity of the pBKCMV control given an arbitrary value of 1. Transfections were done in triplicate. B, p53RE reporter activity in proliferating, quiescent, and senescent cells. Presenescent (Presen) or senescent (Sen) cells were cultured in 10% serum (+), in which case presenescent cells were exponentially growing, or incubated in 0.2% serum for 3 days (), in which case presenescent cells were quiescent. E6-expressing cells were maintained in 10% serum. Triplicate cultures were co-transfected with the normalization, p53RE, or control (pAD) vectors and assayed as described in A. Normalized control activity was subtracted from normalized activity of p53RE, and the value in proliferating cells was arbitrarily set at 1. Shown is the average of two (E6-expressing cells) or five (other cells) independent experiments. (C) Wild-type and mutant p53RE reporter activity in proliferating and quiescent cells. p53RE containing wild-type (p53RE) or mutant (p53 RE mt) binding sites were assayed for reporter activity in proliferating and quiescent cells as described in B. Shown is the average of three independent experiments. In all cases, error bars show the standard errors of the means.

p53 Protein Increases Transiently As Cells Enter Quiescence-- p53 protein levels are similar in proliferating and replicatively senescent cells, despite elevated DNA binding and transactivation activity in senescent cells (6, 40, 42). However, replicative senescence occurs gradually, and p53 rises transiently when cells undergo relatively rapid senescence in response to agents such as DNA damage (reviewed in Ref. 8). Quiescence is also relatively rapid, and thus p53 may transiently rise as cells enter G₀. To test this possibility, we measured p53 protein after withdrawing serum from proliferating cells. p53 protein levels increased 12-24 h after serum withdrawal, peaking 5-fold over -tubulin and QM levels, which do not change with quiescence (32, 34) (Fig. 4A). p53 remained elevated for ~2 days, before declining over the next 2 days (Fig. 4A). Thus, p53 protein increased transiently as cells entered quiescence.

View larger version (46K): [in this window] [in a new window]   Fig. 4.   p53, phospho-Ser-15 p53, and p21 protein levels as cells enter quiescence. A, p53, phospho-Ser-15 p53, and p21 levels after serum withdrawal. Proliferating cells (+serum, asyn) were washed and incubated in 0.2% serum (serum) for the indicated intervals. Alternatively, quiescent cells were stimulate with 10% serum for 3 h (+serum, 3 h). Lysates were analyzed by Western blotting for p53, phospho-Ser-15 p53, p21, QM (control), and -tubulin (control) proteins. Three independent experiments gave similar results. Signal intensities were quantified by densitometry using several exposures to obtain signals in the linear range of the film and are plotted on the graph below the autoradiograms. The number beneath each band is the signal intensity normalized to the average intensity of the QM and -tubulin signals, with the value for asynchronized cells set at 1. QM and -tubulin signals were both relatively invariant. B, p53, p21, and MDM2 mRNA levels as cells enter quiescence. Proliferating cells (+serum, asyn) were incubated in 0.2% serum (serum). At the indicated intervals thereafter, RNA was isolated and analyzed for p53, p21, MDM2, and QM (control) mRNA. Signal intensities were determined, plotted, and represented as described above (A). Two to three independent experiments showed similar results.

p53 Ser-15 Phosphorylation in Quiescent Cells-- p53 arrests cell growth in response to DNA damage owing in part to N-terminal phosphorylation at Ser-15, which reduces the p53-MDM2 interaction causing p53 accumulation and activation (43, 44). Western analysis using a p53 phospho-Ser-15 antibody showed that 12-24 h after withdrawing serum from proliferating cells Ser-15 phosphorylation increased, peaking 5-fold over the level in proliferating cells (Fig. 4A). This increase equaled the rise in total p53 protein, suggesting that, although the absolute amount of phospho-Ser-15 p53 increased, the fraction of phosphorylated Ser-15 p53 does not change as cells become quiescent.

p53 mRNA Increases As Cells Enter Quiescence-- p53 accumulation is generally mediated by posttranslational modification (45). Consequently, little is known about p53 mRNA regulation. We therefore quantified p53 mRNA levels by Northern analysis as cells entered G₀. Relative to the constitutively expressed QM mRNA (34), p53 mRNA rose ~4-fold within 12 h after serum withdrawal. The mRNA decreased slightly after 2 days but remained high relative to proliferating cells for at least 4 days, and did not change 3-4 h after quiescent cells were stimulated by serum (Fig. 4B). This rise in p53 mRNA preceded the rise in p53 protein (Fig. 4A), suggesting that it contributes to the increase in protein levels. These data suggest that the signal transduction pathway that induces p53 after serum withdrawal act at least in part by inducing p53 mRNA and thus differs from the posttranslational pathway that induces p53 after genotoxic stress.

p21 Increases As Cells Enter Quiescence-- The best-characterized p53 target gene is that encoding the CDKI p21 (24). As cells entered G₀, p21 mRNA and protein increased following the rise in p53 protein (Fig. 4, A and B). p21 mRNA rose ~3-fold 24 h after serum withdrawal, while p21 protein rose 7- to 10-fold after 24-48 h. Thereafter, p21 mRNA and protein levels declined but remained higher than the levels in proliferating cells (Fig. 4, A and B), consistent with the elevated p21 expression reported in quiescent cells (21). p21 mRNA and protein were also transiently induced further by serum stimulation as reported (22)(Fig. 4, A and B). Because p21 can be induced by p53-independent mechanisms (25), we analyzed the expression of other p53-responsive genes. MDM2 (Fig. 4B) and Gadd45 (not shown) mRNA increased in parallel with p21 mRNA, consistent with the rise in p53 protein and activity as cells entered G₀.

In the above experiments (Figs. 4 and 5), we withdrew serum from sparse cultures (~1500 cells/cm²), which remained subconfluent throughout the experiment. In denser cultures (~8000 cells/cm²), which became confluent 24 h after serum withdrawal (because cells beyond the restriction point divide), p53 rose to the level observed in sparse cultures, but p21 rose to a lesser extent (not shown). This finding suggests that p21 induction is density-dependent, consistent with a role for the CDKI p27 in regulating the quiescence response to cell density (26-29).

We determined the relationship between the rise in p21 and the quiescent growth arrest by immunostaining cells for p21 and Ki67. Ki67 is not expressed by cells in G₀ (46). Coincident with the rise in p53 and p21 1 day after serum withdrawal, many cells adopted a spindle-like morphology (Fig. 5A), which persisted for at least 4 days (not shown). Proliferating human fibroblast cultures typically contain 10-20% senescent cells, identified by their large flat morphology, which did not change after serum withdrawal (not shown), and high levels of p21. The proliferating cultures used here had ~12% p21-positive cells, which were likely senescent based on their morphology. One day after serum withdrawal, p21-positive cells increased to 46%, while Ki67-positive cells declined from 65 to 28% (Fig. 5B and Table I). Significantly, ~95% of the Ki67-positive cells were negative for p21 (Fig. 5B and Table I). By 3 days after serum withdrawal, >90% of cells were p21-positive and Ki67-negative (not shown). Taken together, these data suggest that as cells enter quiescence p53 rises followed by an increase in p21 and that the induction of p21 correlates with the quiescent growth arrest.

View larger version (61K): [in this window] [in a new window]   Fig. 5.   Morphology, p21, and Ki67 expression after serum withdrawal. A, morphology. Presenescent WI-38 cells, proliferating (+serum, asyn), or incubated in 0.2% serum for 1 day (serum, 1 d) were photographed under phase contrast microscopy. Magnifications for upper panel were 40×, for the lower panel, 200×. B, p21-positive cells are growth-arrested. Proliferating cells or cells in 0.2% serum for 1 day were immunostained for p21 and Ki67 as described under "Experimental Procedures." Most p21-positive cells were negative for Ki67. Examples are shown by the arrows.

Quiescence Is Delayed in Cells Lacking p53 Function-- To test the idea that p53 contributes to establishing quiescence, we used normal human fibroblasts that express HPV-16 E6, which accelerates p53 degradation (41). As expected (41, 47), E6 markedly reduced the levels of p53 and p21 in proliferating cells (Fig. 6A). When E6-expressing cells were deprived of serum, p53 and p21 levels rose but remained substantially lower than the level in proliferating control cells (Fig. 6A). A similar trend was seen for p53 Ser-15 phosphorylation (Fig. 6B). Thus, although E6 markedly reduced p53 levels, it did not completely abolish the p53 and p21 response to serum withdrawal. We also used human fibroblasts that express the dominant interfering p53 fragment GSE-22, which was shown to specifically inactivate p53 function (48). GSE-22 stabilized p53 protein levels as reported (48) but strongly suppressed p21 expression (Fig. 6C), consistent with its dominant p53 inhibitory activity (48). Although GSE-22 substantially reduces p53 activity, like E6, it did not completely abolish the p21 response to serum withdrawal (Fig. 6C).

View larger version (37K): [in this window] [in a new window]   Fig. 6.   Quiescence and p53 and p21 expression in E6-expressing and ATM-deficient cells. A, E6 reduces p53 and p21 protein levels. WI-38 cells were infected with an E6-expressing retrovirus as described (32). Proliferating (+serum) control (WI-38) and E6-expressing (E6-WI-38) cells were washed and incubated in 0.2% serum (serum) for 2 or 4 days as indicated. Lysates were analyzed for p53, p21, and QM (control) proteins by Western blotting. Signal intensities were determined and normalized to QM as described in the legend to Fig. 4A. The normalized signal intensity in proliferating control cells was arbitrarily set at 1. Beneath each band, the normalized fold change relative to control cells is given. B, p53, phospho-Ser-15 p53 and p21 levels in serum-deprived E6-expressing cells. Proliferating E6-expressing cells (+serum, asyn) were incubated in 0.2% serum (serum) for the indicated intervals. Alternatively, quiescent E6-expressing cells were stimulated with 10% serum for 3 h. Cell lysates were analyzed by Western blotting for p53, phospho-Ser-15 p53, p21, and -tubulin proteins. The p53, phospho-Ser-15 p53, and p21 signal intensities were very low as expected from Fig. 6A (compare WI-38 with E6-WI-38). C, p53 and p21 levels in serum-deprived GSE-22-expressing cells. Proliferating GSE-22-expressing cells were treated and analyzed similarly to E6-expressing cells as described in B. D, loss of p53 function delays quiescence. WI-38 cells were infected with retroviruses carrying no insert (control), wild-type E6 (E6), an E6 mutant defective in binding p53 (E6-mt), or GSE-22. [3H]thymidine was added to proliferating cells for 24 h (0 days after serum withdrawal). Alternatively, proliferating cells were incubated in 0.2% serum for 1-4 days, and [3H]thymidine was added for 24-h intervals immediately after shifting to low serum. Cells were scored for the fraction that synthesized DNA during the [3H]thymidine pulse (%LN). The %LN of proliferating cultures (0 day) was set at 100%, but the absolute %LN for each culture is shown beneath each bar. At least three independent cultures and 500 cells were counted for each determination. Error bars indicate standard errors of the means. E, ATM deficiency does not delay quiescence. Normal (WI-38) and ATM-deficient (GM05823 and AT2SF) human fibroblasts were labeled with [3H]thymidine and incubated in 0.2% serum. At least two independent cultures and 500 cells were counted for each determination. F, p53, phospho-Ser-15 p53 and p21 levels in serum-deprived ATM-deficient cells. Proliferating cells ATM-deficient (GM05823 and AT2SF) cells (+serum, asyn) were incubated in 0.2% serum (serum) for the indicated intervals. Alternatively, quiescent cells were stimulated with 10% serum for 3 h. Cell lysates were analyzed by Western blotting for p53, phospho-Ser-15 p53, p21, and -tubulin proteins. In the lower panel, the phospho-Ser-15 p53 signal intensities in ATM-deficient cells were compared with those in WI-38 and E6-WI-38 cells. Basal Ser-15 phosphorylation was markedly reduced in ATM cells.

To determine whether diminished p53 affected the ability of cells to enter quiescence, we monitored DNA synthesis after withdrawing serum from control cells or from cells that express E6- or GSE-22. DNA synthesis was measured using a 24-h pulse with [³H]thymidine to determine the percentage of cells with radiolabeled nuclei (%LN). At each 24-h interval after serum withdrawal (0-4 days), the %LN was significantly higher in E6- and GSE-22-expressing cells compared with control cells (Fig. 6D). One day after serum withdrawal (interval of 1-2 days), the %LN of control cells fell to 18% that of proliferating cells. By contrast, the %LN of E6- and GSE-22-expressing cells fell to only 43 and 56% of proliferating cells. Similarly, 48 h after serum withdrawal (interval of 2-3 days), control cells had <3% the LN of proliferating cells, whereas E6- and GSE-22-expressing cells had 32 and 40% the LN of proliferating cells. To confirm that the delayed quiescence of E6-expressing cells was due to diminished p53 levels, we used an E6 mutant (E6mt) that is defective in targeting p53 for degradation but proficient in other E6 functions (e.g. activation of telomerase) (49). When deprived of serum, cells expressing E6mt had %LN values that were very similar to those of control cells (Fig. 6D).

Our finding that the ratio of phospho-Ser-15 p53 to total p53 did not change after serum withdrawal suggested that the mechanism of p53 activation during quiescence might differ from that during genotoxic stress. To explore this idea, we monitored entry into quiescence by two human fibroblast strains (GM05823 and AT2SF) deficient in the ATM kinase, which phosphorylates p53 at Ser-15 after DNA damage (44). ATM-deficient cells entered quiescence with kinetics very similar to those of normal cells (Fig. 6E). Moreover, p53 protein levels rose transiently after withdrawing serum from ATM-deficient cells, although Ser-15 phosphorylation did not rise appreciably (Fig. 6F). These results suggest that least one of the major kinases (ATM) involved in the p53 DNA damage response and Ser-15 phosphorylation does not play an important role in establishing quiescence.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In vivo, many cells exist in a quiescent growth state, from which they can be stimulated to proliferate by appropriate signals including growth factors. Tumor cells often have a reduced requirement for growth factors, which contributes to their unregulated proliferation (1). One critical regulator of the transition from G₀ and G₁ into the S phase of the cell cycle is the pRb tumor suppressor. Cell cycle progression entails the sequential phosphorylation of pRb by cyclin-dependent protein kinases (CDKs), and unphosphorylated pRb halts cell cycle progression (13, 14). pRb phosphorylation is regulated by CDK inhibitors. p53 halts cell cycle progression in response to DNA damage at least in part by inducing the CDK inhibitor p21 (3, 5, 24). Given this and other interactions between the pRb and p53 pathways (4) and the important role that p53 plays in tumor suppression, it seemed likely that p53 might contribute to cellular responses to growth factor withdrawal. In some cells, growth factor deprivation causes p53-dependent apoptosis, for example nerve growth factor withdrawal from cultured neurons (50, 51). In other cells, growth factor deprivation causes a quiescent growth arrest. In contrast to the apoptotic response, little is known about whether or to what extent p53 is important for quiescence.

Abrogation of p53 by wild-type, but not mutant, E6 significantly delayed human fibroblasts from entering quiescence after serum withdrawal. Furthermore, GSE-22, a specific inhibitor of p53 activity, also delayed entry into quiescence. These findings, the elevated p53 activity in quiescent cells and the transient induction of p53 after serum withdrawal, support the idea that p53 contributes to the ability of normal cells to become quiescent in response to growth factor insufficiency.

p21, a well characterized p53 target gene (24), may be an important mediator of this effect of p53. p53 binding to the p21 promoter element was elevated in quiescent cells, and 24 h after serum withdrawal the p21 mRNA level rose following the rise in p53 protein. Single cell analyses showed that 1 day after serum withdrawal many cells expressed high levels of p21, the majority of which were in G₀ as judged by the absence of Ki67 staining. A sequential rise in p53 and p21 was also seen in E6-expressing cells, although it was much attenuated suggesting that the residual p53 in E6-expressing cells retained the ability to respond to serum withdrawal. Finally, E6-expressing cells entered quiescence with delayed kinetics and were markedly deficient in p21. Our findings are consistent with reports that p21 antisense RNA stimulates quiescent human fibroblasts to enter S phase (22) and that disruption of the p21 gene delays human fibroblasts from entering G₀ (52).

Despite the delay, E6- and GSE-22-expressing cells nonetheless eventually became quiescent. Moreover, a T antigen mutant that could not bind p53 nonetheless stimulated quiescent cells to synthesize DNA, and MDM2 overexpression was not as effective as wild-type T antigen in stimulating quiescent cells. These results suggest that other regulators cooperate with p53 or can compensate for loss of p53 in establishing and maintaining quiescence. The most likely candidates are pRb and the p27 CDKI that regulates its phosphorylation. p27 antisense RNA was shown to stimulate DNA synthesis in serum-deprived rodent fibroblasts (26, 27), although in the presence of serum p27 it did not arrest growth unless cells were at high density (29). By contrast, p21 was shown to arrest cell proliferation in the presence of serum, even if cells lacked functional pRb (53). p21 and p27 may function cooperatively or interchangeably, depending on the cell type and cellular microenvironment. This idea is consistent with the finding that p27 levels increase when E6-expressing human fibroblasts become quiescent (47) and that p53/, p21/ and p27/ mice develop normally. Moreover, fibroblasts from p21/ or p27/ mice eventually become quiescent when deprived of serum (54, 55).

The p53 response to genotoxic stress is mediated primarily by posttranslational modifications (56, 57). A prominent posttranslational modification is Ser-15 phosphorylation carried out by the ATM protein kinase (43, 44). We showed that the absolute, but not relative, amount of phosphorylated Ser-15 p53 increased after serum withdrawal. However, cells deficient in ATM entered G₀ with kinetics similar to that of wild-type cells, despite low levels of Ser-15 phosphorylation. These data suggest that Ser-15 phosphorylation does not contribute significantly to the rise in p53 levels and activity in quiescent cells and that kinases that act on Ser-15 may be constitutively active during entry into quiescence. It is possible that other posttranslational modifications, such as acetylation (44), contribute to the increased p53 activity of quiescent cells. Consistent with this idea, p53 acetylation was shown to decrease when quiescent cells were stimulated with serum (58). p53 can, of course, undergo myriad posttranslational modifications, and determining which of these is important for quiescence will require further investigation.

Serum withdrawal caused a relatively rapid rise in p53 mRNA, which preceded the rise in p53 protein, p21 mRNA, and MDM2 mRNA. This finding suggests that one or more components of the p53 pathway respond rapidly to growth factor insufficiency. The induction of p53 mRNA by serum withdrawal distinguishes the quiescence response from the p53 responses to DNA damage and cellular stress. We do not yet know whether the accumulation of p53 mRNA is caused by an increase in mRNA stability or transcriptional induction. Several transcription factors have been shown to bind the p53 promoter and induce transcription, including AP-1, NF-B, and HOXA5 (59, 60). The p53 promoter has also potential binding sites for Ets1 and 2, which are involved in Ras-induced and replicative senescence (45, 61). Inactivation or deficiency in these transcription factors may contribute to the reduced ability of cancer cells to enter G₀. Indeed, p53 mRNA levels were reported low in several breast cancer cell lines because of loss of HOXA5 expression (60).

In summary, p53 is widely recognized as a tumor suppressor that controls or participates in a remarkable number of basic cellular processes including cell cycle progression, apoptosis, and cellular senescence. Here, we present evidence that p53 also participates in the reversible growth-arrested state of quiescence.

	ACKNOWLEDGEMENTS

We thank P. Yaswen for the mutant E6 retroviral vector, B. Vogelstein for CMV-MDM2, J. Tevethia for the T[p53] cDNA, J. R. Smith for CMV-p53, V. Band for the E6 retrovirus-producing cells, and E. Blakely for the AT2SF cells.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants AG09909 (to J. C.) and CA82548 (to P. Y. D.) and a fellowship from the California Breast Cancer Research Program (to K. I.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^** To whom correspondence should be addressed: Life Sciences Div., Lawrence Berkeley National Laboratory, 1 Cyclotron Rd., MS 84-171, Berkeley, CA 94720. Tel.: 510-486-4416; Fax: 510-486-4545; E-mail: JCAMPISI@LBL.GOV.

Published, JBC Papers in Press, March 5, 2002, DOI 10.1074/jbc.M201028200

	ABBREVIATIONS

The abbreviations used are: SV40, simian virus 40; pRb, retinoblastoma; CDKI, cyclin-dependent protein kinase inhibitor; LN, radiolabeled nuclei; CMV, cytomegalovirus; -gal, -galactosidase; Ab, antibody; DAPI, 4',6-diamidino-2-phenylindole; CDK, cyclin-dependent protein kinase; HPV, human papilloma virus; GSE, genetic suppressor element; ATM, ataxia telangiectasia-mutated protein kinase.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES