Eur. J . Biochem. 245, 730-737 (1997) 0 FEBS 1997

Redox-mediated regulation of p21waf1'cip1 expression involves a post-transcriptional mechanism and activation of the mitogen-activated protein kinase pathway Franca ESPOSITO', Franca CUCCOVILLO I, Marco VANONl', Filiberto CIMINO I, Carl W. ANDERSON3, Ettore APPELLA4 and Tommaso RUSSO'

' Dipartimento di Biochimica e Biotecnologie Mediche, Universita degli Studi di Napoli Federico TI, Naples, Italy ' Dipartimento di Fisiologia e Biochimica Generali, Universiti degli Studi di Milano, Milano, Italy ' Biology Department, Brookhaven National Laboratory, Upton NY, USA

Laboratory of Cell Biology, NCI, National Institutes of Health, Bethesda MA, USA

(Received 6 January 1997) - EJB 97 0017/1

p21w'<' l 'CL]7l gene expression is induced by DNA damage in cells with wild-type p53 and contributes to the arrest of cell growth. It was demonstrated that under many experimental conditions, including oxida- tive stress, p21"""~"p' expression can be induced through p53-independent pathways. Since most of these experimental conditions induce the phosphorylation of mitogen-activated protein kinase (MAPK) and thus its activation, we evaluated p21 w."''c'pl mRNA levels in cells exposed to an oxidative stress, induced by diethylnialeate (Et,Mal), and in which the MAPK pathway was blocked. The expression of a dominant- negative mutant of MEK, the MAPK kinase that phosphorylates and activates MAPK, and of a dominant- negative [Asn17]Ras mutant prevented the EtZMal-induced accumulation of p21w~f1'c'p' mRNA. Similarly, the expression of MEK- and of [Asn17]Ras mutants decreased the 12-0-tetradecanoyl-phorbol 13-acetate (TPA)-mediated p21 WLlt'"lP' induction. Furthermore, TPA-induced and serum-induced p21 u.afl'c'pI mRNA accumulation was blocked by pretreating the cells with the antioxidant compound N-acetylcysteine, sug- gesting that oxidative stress is involved in these responses.

p21 w.,t Iiclpl mRNA levels reached a maximum within 2 h of adding Et,Mal or TPA; however, the rate of transcription from a p21 W""'c'p'-promoter construct did not increase during this period. In contrast, cells treated with actinomycin D show an increase of p21'"~'"'"~" mRNA stability after Et,Mal treatment.

This result suggests that the increase in p21""11'c'~" mRNA at early times results from post-transcrip- tional regulatory events. Longer exposure to TPA may activate p21 *~"'"p~ gene transcription through an Spl -dependent mechanism, while Et,Mal treatment gradually inhibits p21W"t1'c'P' gene transcription through oxidative changes that affect S p l binding to DNA.

Keywords: p53; p21W~"1"'p' ; mitogen-activated protein kinase; reactive oxygen species.

The inhibition of cell growth after exposure to DNA-damag- ing agents is a consequence of the elevated expression of several genes regulated by the p53 tumor-suppressor protein [ 1, 21. One of these is the p21""''"p1 gene, which encodes a general inhibitor of cyclin-dependent kinases and a modulator of proliferating- cell-nuclear antigen activity [3 -51. Several observations indi- cate that the expression of p21"""'"P' gene is also activated through p53-independent pathways. For example, in p53-null cells derived from knock-out mice, serum stimulation after se- rum starvation, and exposure to various purified growth factors stimulate ~21""' ""p' gene [6-81. TPA and other differentiating agents such, as retinoic acid and vitamin D,, induce p21""'""'p' mRNA expression in HL60 cells, which lack p53 [9-111. In all

Correspondence to T. Russo, Dipartimento di Biochimica e Biotec- nologie Medicbe, Universita degli Studi di Napoli Federico IT, Via S. Pansini 5 , 1-80131 Napoli, Italy

Fax: +39 81 7463650. E-mail: russoto@ds.unina.it Abbreviations. Ac-Cys, N-acetylcysteine; DMEM, Dulbecco's madi-

fied Eagle's medium ; EtZMal, diethylmaleate; MAPK, mitogen-activated protein kinase; MEK, mitogen-activated-protein kinase kinase; MKP, mitogen-activated-protein-kinase phosphatase; TPA, 12-0-tetradeca- noyl-phorbol 13-acetate; PDGF, platelet-derived growth factor.

cases, the increase in p21wnf'/c1p1 mRNA was observed in the ab- sence of protein synthesis [ I 11. A similar p53-independent acti- vation of the p21w'"''c1p1 gene was observed in KG-1 cells treated with ?-rays or tumor-necrosis factor [12], i n human breast carci- noma cells treated with etoposide [S], in HaCaT cells exposed to transforming growth factor [13], and in PC3 M p53-null cells exposed to taxol [14].

Of specific interest for this study is the induction of p21"""/c!p' mRNA expression in response to oxidative stress. We demonstrated that diethylmaleate (Et,Mal), which promotes an increase of reactive oxygen species by causing glutathione de- pletion [lS], inhibits the specific DNA binding and the transacti- vating efficiency of p53 [16]. However, despite this Et,Mal-in- duced impairment of p53-mediated transcription, the levels of P21 "'It mRNA significantly increase after Et,Mal treatment and the cells accumulate in G1 [16]. The finding that Et,Mal induces the p21 w.Lrl'"pl expression in cells lacking functional p53 indicates that a p53-independent pathway for p21""''"~'' induc- '

tion in response to oxidative stress exists [16]. In this report we show that Et,Mal-induced oxidative stress

induces p21"~'f''cLp' gene expression through regulatory mecha- nisms involving the activation of the Ras/mitogen-activated pro- tein kinase (MAPK) cascade, and acts at a post-transcriptional

Esposito et al. ( E m J. Binchem. 245) 73 1

level. 12-0-tetradecanoyl-phorbol 13-acetate (TPA)-induced and serum-induced increases of p21 mRNA expression are, at least in part, due to the activation of the same pathway, and are prevented by antioxidant treatments. This suggests a key role for intracellular redox conditions in the p53-independent p21""f1/c'p' gene regulation.

EXPERIMENTAL PROCEDURES

Cell cultures and treatments. HeLa cells were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10 % calf serum (ICN), 100 U/ml penicillin and 100 pg/ml streptomy- cin at 37 "C in a 5 % CO, atmosphere. TPA and Et,Mal (Sigma) were added to the cell cultures from concentrated solutions at the concentrations and times indicated in the figure legends. N- acetylcysteine (Ac-Cys; Sigma) was dissolved in 10 mM sodium phosphate, 138 mM NaC1, 2.7 mM KC1, pH 7.4 (NaCUP,), the pH was adjusted to 7.4 with 1 M NaOH, and Ac-Cys was added to the cells to 30 mM 2 h before treatment with either Et,Mal, TPA or serum. Serum starvation was achieved by culturing cells in DMEM containing 0.5% calf serum for 24 h before treat- ments. Actinomycin D was dissolved in the medium at 10 pg/ ml. Viability was assessed by trypan-blue exclusion for each treatment.

Plasmids. The plasmid pBABE LIDA (Mek-), expressing a dominant-negative mutant of MAPK kinase (MEK), was kindly provided by C. J. Marshall [17]. The construct expressing the dominant-negative mutant of Ha-Ras ([Asnl7]Ras) was kindly provided by G. M. Cooper [18]. Plasmids pSG5, pSG5-MKP-1 and pSG5-MKP-1CS were gifts of N. K. Tonks [19]. pDM-Luc containing the p53-site-deleted p21 wntl/clpl promoter linked to the luciferase gene, was kindly provided by Dr B. Vogelstein [3].

Transfections and luciferase assay. Transfections of HeLa cells with mutant plasmids were per€ormed by electroporation (Bio-Rad gene pulser, Bio-Rad Lab.) to reach the highest effi- ciency of transfection. 3 X loh cells/lOO-mrn plastic dish were transfected with 25 pg of each plasmid at 960 pF/200 V. Carrier DNA was used for normalization in cotransfection experiments. 48 h after transfection, cells were exposed to the different agents, as described in the figure legends, and harvested for the purification of RNA and/or for the preparation of cell lysates. For the luciferase assays, cells were transfected with 10 pg/ 0.5X lo6 cells/60 mm plastic dish of the pGL2-promoter vector (Promega Corporation), which has the luciferase gene driven by the SV40 early promoter, or with the pDM-Luc plasmids [3] by the calcium-phosphate procedure [20]. 36 h after transfections, cells were exposed to TPA and Et,Mal for the times indicated in the figure legends and harvested for the preparation of cellular extracts. The determination of luciferase activity was carried out according to the manifacturer's instructions (Promega).

RNA extraction and northern analysis. Total cellular RNA was purified by the guanidinium-thiocyanate method as de- scribed [21]. RNA was separated on a 1.5% agarose/formalde- hyde gel and transferred to a nitrocellulose membrane (Nytran, Schleicher & Schiill). Northern analysis was performed with probes from human p21w"f'/c1p', actin and glyceraldehyde-3-phos- phate dehydrogenase cDNAs labeled with [a-"P]dATP by the random-priming method. Hybridizations and washes were per- formed according to the Church protocol [22].

Western blots. Cell cultures (70- 80% confluent) were treated with the indicated agents, washed twice with NaCI/P,, and lysed in 1.6 mM KH,PO,, 12.5 mM K,HPO,, 1 % Triton, 3.4 mM SDS, 0.1 M NaC1, 11.5 mM sodium deoxycholate, 0.1 % sodium azide, 50 mM NaF, 1 mM sodium orthovanadate, 10 mM sodium pyrophosphate and 1 mM phenylmethylsulfonyl

fluoride. Soluble extracts used for western blots were obtained by centrifugation of the lysate at 4°C at 100OXg for 10 min; 10 pg protein were applied to each lane. Proteins were separated by electrophoresis through a 10% polyacrylamide gel. To re- solve phosphorylated from non-phosphorylated MAPK, a 30/0.3 ratio of acrylamide to bisacrylamide was used. Proteins were transferred to Immobilon-P transfer membranes (Millipore), and reacted with polyclonal anti-MAPK Ig, at 0.1 pg/ml (Santa Cruz) for 1 h at room temperature. After incubation with rabbit horseradish-peroxidase-conjugated secondary antibody, the blots were developed by means of enhanced chemiluminiscence (Amersham).

RESULTS Accumulation of p21"afl'c'p' mRNA in response to EkMal and TPA is correlated with activation of RasMAPK pathway. Since TPA treatment of p53-I- cells has an effect on p21"~'"/"pl mRNA similar to that of Et,Mal, we explored the possibility that oxidative stress acts by modulating the activity of the enzyme systems known to be targets of TPA. We reported previously that Et,Mal induces phosphorylation of MAPK, but, in contrast to other oxidant or prooxidant conditions [23, 241, it did not affect PKC activity 1161. To evaluate whether MAPK activation is involved in the induction of p2lw""/"P' mRNA accumulation by Et,Mal, we explored the effect of the oxidative stress in cells in which the pathway of MAPK activation was blocked. To this end, HeLa cells were transiently transfected with plasmid vec- tors driving the expression of dominant-negative mutants acting on the Ras pathway upstream of MAPK. Over-expression of the transfected dominant-negative mutant of MEK [17], the MAPK kinase that phosphorylates and activates MAPK, prevented the Et,Mal-induced p21 wail/c'pl mRNA increase. Similarly, a cDNA encoding the [Asn17]Ras dominant-negative mutant [IS], tran- siently transfected in HeLa cells before Et,Mal treatment, pre- vented the induction of the p21"'1"/c1p1 mRNA provoked by the Et,Mal treatment. These two treatments block MAPK phosphor- ylation and activation by inhibiting separate steps in the Ras signaling cascade upstream of MAPK. The MAPK assay (Fig. 1 B) demonstrated that the amount of phosphorylated MAPK was high in mock-transfected cells treated with Et,Mal, but this increased phosphorylation was abolished in cells transfected with the dominant-negative MEK mutant.

To evaluate the specificity of this observation, we asked whether induction of the p21w""/c1~' mRNA accumulation in re- sponse to TPA-treatment was modified by the transient expres- sion of dominant negative mutants of the Ras pathway. Transient transfection of dominant-negative MEK or [Asnl7]Ras caused a significant decrease in p21"""'"Pl mRNA accumulation after TPA treatment. Consistent with the involvement of MAPK in mediating p21 ~ ~ ~ f ~ ~ ~ I P 1 mRNA accumulation, the over-expression of the MEK mutant was accompanied by a significant decrease of MAPK phosphorylation (Fig. 2 B).

These results suggest that Et,Mal and TPA could act up- stream of Ras, but do not exclude the possibility that the ob- served MAPK activation is a consequence of the inhibition of MAPK-specific phosphatases (MKP), which would result in an accumulation of phosphorylated, active form of MAPK [ 191. To examine this possibility, we transiently transfected HeLa cells with the cDNA encoding MKP under the control of a constitu- tive promoter. Over-expression of this phosphatase resulted in a strong inhibition of MAPK phosphorylation. Under these condi- tions neither Et,Mal nor TPA were able to activate MAPK (Fig. 3), and the Et,Mal treatment was unable to induce the p21 wsfl/ctpl mRNA accumulation (Fig. 3 B), suggesting that the target of these treatments is not this phosphatase.

732

A

Esposito et al. ( E m J. Biochem. 245)

A mock + + - - - - MEK- - - + + - - A m 1 7 - - - - + +

mock + + - - - - ~ M E K - - - A m 1 7 - - - - + +

1 EtzMal - + - + - + I / T P A - + - + - + I + WAFl

+- Act in

a b c d e f

6 -

5-

4-

a b c d e Lanes

f

- MAPK

a b c d e

Fig. 1. p21"af1"'p' mRNA accumulation induced by Et,Mal is abol- ished in cells in which the MAPK cascade is blocked. (A) HeLa cells were transfected, as described under Experimental Procedures, with plas- mid vectors expressing MEK and [Asnl7]Ras dominant-negative mu- tants, under the control of a constitutive promoter, or with the empty vector (mock). 48 h after transfection, the cells were exposed to 1 mM Et,Mal for 3 h and harvested for RNA preparation. RNA samples were analyzed by northern blotting with human p21"""'"p' cDNA as a probe. The same blot was probed with human actin cDNA as a control for gel loading and transfer. Lanes a and b, mock-transfected cells; lanes c and d, cells transfected with MEK- mutant cDNA; lanes e and f, cells transfected with [Asnl7]Ras mutant cDNA; lanes b, d and f, cells treated with Et,Mal. (B) Quantitation of p21wk'f''c'p' mRNA induction. The autoradiogram shown in (A) was scanned with a densitometer. The amount of ~21"""'"~' (white bars) and actin(hatched bars) mRNA in each of the lanes is reported. (C) HeLa cells, transfected as described above, were treated with 1 mM Et,Mal for 30min or with 50ng/ml TPA for 10 min and harvested for protein-extract preparation. The protein ex- t rx t s were anaiyzed by western blotting with anti-MAPK Ig. Lanes a- c, mock-transfected cells; lanes d and e, cells transfected with MEK- mutant cDNA; lane b, cells treated with TPA; lanes c and e, cells treated with Et,Mal. Phosphorylated (activated) MAPK has a slightly decreased mobility compared with inactive MAPK.

+ WAFl

+ GraP-DH

a b c d e f

B

a b - c d e f

Lanes

C

I TPA - + + - 1

C- ERK2

a b c d Fig. 2. The increase in p21waf1'cip1 mRNA induced by TPA is reduced in cells in which the MAPK cascade is blocked. (A) HeLa cells were transfected as described in the legend of Fig. 1. 48 h later they were treated with 50 ng/ml TPA for 2 h and harvested for RNA preparation. These RNA samples were analyzed by northern blotting with human p21w""'L'p' cDNA as a probe. The same blot was probed with the human glyceraldehyde dehydrogenase (GraP-DH) cDNA as a control of gel loading and transfer. Lanes a and b, mock-transfected cells; lanes c and d, cells transfected with MEK- mutant cDNA; lanes e and f, cells transfected with [Asnl7]Ras mutant cDNA; lanes b, d and f, cells treated with TPA. (B) Quantitation of northern blot shown in (A) of p21 W ' i i l / C l P l (white bars) and actin (hatched bars). (C) HeLa cells, transfected as described above, were treated with 50 ng/ml TPA for 10 min and harvested for protein-extract preparation. The western blot was carried out as described in the legend of Fig. 1. Lanes a and b, mock-transfected cells; lanes c and d, cells transfected with MEK- mu- tant cDNA; lanes b and c, cells treated with TPA.

A change in intracellular redox conditions is necessary for the MAPK-dependent induction of p21""""'P'. It was impor- tant to evaluate whether Et,Mal activates MAPK by mechanisms similar to or different from those responsible for TPA-mediated and serum-mediated MAPK activation. We demonstrated pre-

Esposito et al. ( E m J. Biockem. 245) 733

-Fig.3. The MAPK activation induced by Et,Mal or TPA is not due to an inhibition of MKP. (A) HeLa cells were transfected, as described under Experimental Procedures, with a plasmid vector encoding MKP, under the control of a constitutive promoter, or with the empty vector (mock). 48 h after transfection the cells were exposed to 1 mM Et,Mal for 30 rnin or SO ng/ml TPA for 10 rnin and harvested to prepare protein extracts. The activation state of MAPK was analyzed by western blotting with an anti-MAPK Ig. Lanes a-c, mock-transfected HeLa cells; lanes d-f, cells transfected with MKP expression vector; lanes b and e, cells treated with TPA; lanes c and f, cells treated with Et,Mal. (B) HeLa cells were transfected as described under Experimental Procedures with a plasmid vector encoding MKP, under the control of a constitutive promoter, or with the empty vector (mock). 48 h later they were treated with Et,Mal (1 n7M) for 3 h and harvested for RNA preparation. These RNA samples were analyzed by northern blotting with human p21"""/ 'v' cDNA as a probe. The same blot was probed with human actin cDNA as a control of gel loading and transfer. Lanes a and b, mock-transfected cells; lanes c and d, cells transfected with MKP-encoding plasmid vec- tor; lanes b and d, cells treated with Et,Mal.

viously that the effect of Et,Mal is dependent upon the oxidative stress induced by this compound. Treatment of the cells with Ac-Cys, a reducing agent which increases the amount of gluta- thione, prevents the p21waf'"'p' mRNA induction provoked by Et,Mal treatment [16]. To determine if the mechanism of TPA action is similar to that of Et,Mal, we exposed the cells to 30 mM Ac-Cys before TPA treatment. As observed in the cells treated with Et,Mal, Ac-Cys treatment decreased the p2lWdf~"'p' mRNA accumulation induced by TPA (Fig. 4A). Consistent with this result, the MAPK activation induced by TPA was decreased by the Ac-Cys treatment (data not shown). Serum stimulation induces an increase in p21"af"'1p1 mRNA accumulation [6-81. As for TPA, treatment of the cells with Ac-Cys decreased the response of p21 w"flic'pl to serum. These results suggest that the TPA-induced and serum-induced accumulation of p21 wa"i'lpl

mRNA is in part due to a cellular redox-mediated mechanism that is affected by these treatments.

Fig. 4. The antioxidant Ac-Cys prevents the p21""f'"'p' mRNA increase induced by TPA and by serum. (A) HeLa cells were treated with 30 mM Ac-Cys for 2 h and with SO ng/ml of TPA for 2 h or with 1 mM Et,Mal for 3 h. The RNAs extracted from these cells were ana- lyzed by northern blotting with human p21""'"''p' cDNA as a probe. The same blot was probed with the human glyceraldehyde dehydrogenase (GraP-DH) cDNA as a control for gel loading and transfer. Lane a, un- treated HeLa cells; lanes b, e and f, cells treated with Ac-Cys; lanes c and e, cells treated with Et,Mal; lanes d and f, cells treated with TPA. (B) HeLa cells were incubated in O S % calf serum/DMEM for 24 h and released from serum deprivation by the addition of IS% calf serum/ DMEM for 2 h. The RNAs extracted from these cells were analyzed by northern blotting with human p21""f1/"p' cDNA as a probe. The same blot was probed with human actin cDNA as a control for gel loading and transfer. Lane a, untreated HeLa cells; lanes b and c, cells stimulated with IS% calf serum/DMEM for 2 h; lanes c and d, cells treated with 30 mM Ac-Cys for 2 h.

The early increase in p21"af1"iP' mRNA does not require transcription activation of the gene. The accumulation of P21 wn'licip' mRNA in response to Et,Mal, TPA or serum is an early phenomenon, which can be detected within 30 rnin of ini- tiating treatment and reaches a maximum within a few hours [7, 11, 161. Similarly, the p21 w.lfl'c'pl protein increases several fold within 3-4 h of initiating treatment [ l l ] . Recently, it was re- ported that TPA activates the transcription of the p2lWdr""p' gene in U937 cells and that this activation is mediated by the S p l factor [25]. However, in these experiments promoter activity was tested only 24 h after TPA treatment. Because p21 wotl'c'pl

mRNA and protein accumulation begins at a much early time, we asked whether there was an increase in the efficiency of the

gene promoter at earlier times after exposure to TPA or Et,Mal. HeLa cells were transfected with the pGL2-promoter control vector and pDM-luc, a vector in which luciferase gene expression is driven by the p21""f""'P' gene promoter; the cul- tures were then treated with Et,Mal or TPA. Because the half- life of luciferase in mammalian cells is relatively short (about 3 h) [26], luciferase activity reflects promoter activity during the

p21 warlicipl

734 Esposito et al. ( E m 1. Biochem. 245)

0 2 12 24 0 2 12 24 Time(h)

PDM pGI,2

TPA

C

Et zMal

0 2 12 2 12

PDM pGL2

DEM (0.5 mM 0 , 1 mM

h

- Actin a b c d e

Fig. 5. Effects of TPA and Et,Mal treatments on p21""""ip1 gene promoter efficiency. HeLa cells were transfected by the calcium-phosphate procedure, as described in Experimental Procedures, with the plasmids pGL2 or pDM. 36 h after transfection, cells were exposed to 200 nM TPA (A) or to 0.5 mM and 1 mM Et,Mal (8) for the indicated times. The luciferase activities reported represent the means of at least three experiments. Standard errors are indicated on the top of each bar. (C) The RNAs extracted from these cells were analyzed by northern blotting with human p21".~tl/c,p' cDNA as a probe. The same blot was probed with human actin cDNA as a control for gel loading and transfer. Lane a, untreated HeLa cells; lane b, cells treated with 200 nM TPA for 2 11; lane c, cells treated with 200 nM TPA for 12 h ; lane d, cells treated with 0.5 mM Et2Mal for 2 h ; lane e, cells treated with 0.5 mM Et2Mal Fur 12 h.

preceding few hours. No significant induction of the transcrip- tion activity from the p21"""""~" gene promoter was observed 2 h after adding Et,Mal or TPA. However, northern blot analysis (Fig. 5 C ) showed that 2 h after adding either agent, p21"""""p1 mRNA had accumulated to its maximum level.

Only in the case of TPA treatment was an increase in tran- scription from the p21"""'"1" gene promoter observed, and only after 12 h treatment. This late activation of the p21"""~c'p' gene promoter by TPA is consistent with the report that TPA activates p21""' I / C i P l mRNA transcription through a mechanism involving the transcription factor Spl [25]. In contrast, luciferase expres- sion from the pGL-2 control vector and pDM-luc decreased by 12 h after Et,Mal treatment. These results are consistent with a previous finding that the DNA-binding activity of Spl is sensi- tive to oxidation [27].

As far as the mRNA turnover is concerned, we tested the effect of EtzMal in cells treated with actinomycin D to block RNA biosynthesis. The half-life of p21 u*iI /c'pt mRNA was about 1 h, whereas after exposure to Et,Mal this half-life increased to 3 h (Fig. 6). The amount of p21""f""'~'i mRNA continued to increase during the first 30 min after exposure to actinoinycin

D, probably because of a short lag before the complete abolish- ment of RNA biosynthesis.

DISCUSSION The experiments reported in this paper demonstrate that the

p53-independent induction of p21 wr"i''pl mRNA accumulation in response to Et,Mal or TPA treatment involves the Ras-MAPK pathway. The observation that treatment of cultures with Ac-Cys decreased the ability of Et,Mal, TPA and serum to induce the accumulation of p21 wn'~"clpi mRNA suggests that the effects of these agents are mediated, at least in part, through changes in the intracellular oxidative environment.

Several observations strongly suggest a regulatory role for reactive oxygen species, which can act as second messengers for membrane-initiated signal transduction [28, 291. The p53- independent pathway through which Et,Mal, TPA and seium mediated the induction of p2lW"*'/"P' expression thus represents a potentially important example of this regulatory mechanism.

may be modu- Other p53-independent inducers of p21"~'i''c1p1 lated by the cell's redox state. For example, it was demonstrated

Esposito et al. ( E m J . Bimchem. 245) 735

f g h i j k

B

a b c d e f g h i j k

Lanes Fig. 6. Post-transcriptional stabilization of p21"""'P'mRNA after Et,Mal treatment. HeLa cells were treated with actinomycin D (Act D; 10 pg/ml) and with actinomycin D and Et,Mal (1 mM) for the indicated times, then harvested for RNA preparation. These RNA samples were analyzed by northern blotting with human p21""t""P' cDNA as a probe. The lower panels of each experiment shows a photograph of the ethid- ium-bromide-stained gel to confirm equivalent RNA loading. (A) Lane a, untreated HeLa cells; lanes b-e, HeLa cells exposed to actinomycin D (Act D) for the indicated times; lanes f-k, HeLa cells treated with Et,Mal for the indicated times and, after 3 h, also with Act D for the indicated times. The levels of mRNA from HeLa cells treated (hatched bars) or not (withe bars) with Et,Mal are shown.

recently that exposure to growth factors, such as platelet-derived growth factor (PDGF) and epidermal growth factor, causes an increase of the intracellular levels of H,O, and that the exposure to H,O, is accompanied by an activation of the tyrosine-kinase activities of several growth-factor receptors [30]. Similarly to growth factors, there are several observations indicating that tu-

mor necrosi: factor and y-rays provoke an increase of reactive oxygen species, in a dose-dependent manner [31-321. There- fore, the p21"""/"p1 mRNA increase observed as a consequence of treatment with serum or with individual growth factors [6] could be a consequence of the H,O, accumulation. The H,O, per se is not the oxidizing agent but, in the presence of electron donors such as Fez+, it spontaneously transforms into HO-, the strongest oxygen reactive radical [33].

The physiological target(s) of the oxidizing radicals are un- known. The kinase cascade leading to the activation of MAPK, which includes Raf and MEK, also known as MAPK-kinase ki- nase and MAPK kinase, respectively, does not seem to be a direct target of the oxidative stress since transient transfection of dominant-negative [Asnl7]Ras inhibited the activation of MAPK and the induction of p21""f'~c'p1. Accordingly, the p53- independent induction of p21"""~c'p1 provoked by tax01 treatment requires c-Raf-I [14]. Thus, the steps upstream of the kinase cascade could be activated by the oxidizing radicals. Ras is acti- vated in cells treated with various redox-modulating molecules, including HzOz, and this activation, at least in vitro, seems to be involved in the promotion of the guanine-nucleotide exchange on Ras [34]. Furthermore, PDGF treatment of smooth-muscle cells increases the H,O, concentration, and H,O, activates the tyrosine-kinase activities of growth-factor receptors. These ef- fects do not seem to be secondary events in growth factor signal transduction, since abolishment of this increase of H,O, concen- tration, by exposing the cells to catalase, completely prevents the tyrosine phosphorylation induced by activation of the PDGF receptor [30].

Our findings that the half-life of p21waf1'c1p1 mRNA was increased after Et,Mal treatment and that the promoter was not rapidly activated in response to Et,Mal indicate that the redox- mediated increase in p21 waf'/L1pl mRNA is controlled through a post-transcriptional mechanism that probably results from an increase in p21waf'/c'p1 mRNA stability. This increase in stability could be a consequence of the presence in the 3' untranslated region of the p21"""/"p1 mRNA of several (A+U)-rich sequences including three AUUUA elements, which have been indicated as possible cis-acting regulatory signals of rapid mRNA turnover (reviewed in [35]). The half-life of p21W"f1/c'p1 mRNA in un- treated human cells is short (about 0.5 h) (Fig. 6 ; [Il l) . Further- more, while the basal level of p21""f'/"'p' mRNA is very low in HeLa cells, promoter activity is quite high. Thus, even a modest increase in mRNA stability would result in the rapid accumula- tion of p21waf1/c1~t mRNA. Our findings are consistent with recent reports of an increase in the half-life of p2Iwt1/"p' mRNA in human KG-1 myeloblastic cells after exposure to tumor necrosis factor-a [36], in normal human fibroblasts (Wl38) after exposure to interleukin-I [37] and in HL-60 cells after treatment with TPA or dihydroxyvitamin D, [ 111. Nevertheless, transcriptional activation may contribute to the p53-independent induction of

The physiological role of the MAPK-dependent, p53-inde- pendent pathway of p2lWaf1/"p1 mRNA induction is unknown. A possibility is that, in the absence of a p53 activation, p21""'1'"p' functions as an early response to mitogenic stimuli: its transient induction would result in inhibition of cyclin-dependent kinases that could provide a protective role against an inappropriate and premature transition from GI to S phase. This hypothesis is sup- ported by various observations. For example, it was demon- strated that growth-factor withdrawal results in decreased p21 wafl/clp I mRNA expression and, at the same time, in increased sensitivity to radiation-induced cell death [38]. Similarly, growth arrest provoked by prostaglandin A, through the induction of p21"af1/c'p', is absent in RKO cells; however, upon prostaglandin A, treatment, these cells show significant cell death, and this is

mRNA expression at later times [7, 251. p2 1 wsfl/c,pl

736 Esposito et al. (Eur. J. Biochem. 245)

correlated to the absence of p21"""'c'~1 induction and to an increased activity of cyclin-dependent kinase activity [39]. The same group demonstrated that MAPK activation plays an impor- tant role in cell survival upon oxidative stress [40]. These obser- vations and our results strongly suggest that MAPK promotes cell survival after the increase of reactive oxygen species through the induction of p21""L''"p'.

Hypoxia has been shown to select for p53-deficient cells in human neoplasia [41]. This observation suggests that environ- mental conditions that prevent the formation of reactive oxygen species (e.g. hypoxia), block a p53-independent pathway for growth arrest that normally favours cell survival. Such condi- tions would allow the induction of p53-dependent apoptosis and in turn would provide a positive selection for p53-deficient cells. The use of p21"'af'"'P'-null cell lines and mice may permit these ideas to be directly tested.

This work was supported by grants from Associazione ltaliana per la Ricerca sut Cancro, from Consiglio Nazionale delte Richerche special project 'Invecchiamento', and from Minister0 Universiti Ricerca Scien- tifica e Tecnologica 40% and 60%.

REFERENCES 1. Bates, S. & Vousden, K. H. (1996) p53 in signaling checkpoint ar-

rest or apoptosis, Cum Opin. Genet. Dev. 6, 12-19. 2. Gottlieb, T. M. & Oren, M. (1996) p53 in growth control and neo-

plasia, Biochim. Biophys. Acta 1287, 77- 102. 3. El-Deiry, W. S., Tokino, T., Velculesku, V. E., Levy, D. B., Parsons,

R., Trent, J. M., Lin, D., Mercer, W. E., Kinzler, K. W. & Vogel- stein, B. (1993) WAF1, a potential mediator of p53 tumor sup- pression, Cell 75, 817-825.

4. Harper, J. W., Adami, G. R., Wei, N., Keyomarsi, K. & Elledge, S. J. (1993) The p21 Cdk-interacting protein Cipl is a potent inhibi- tor of G1 cyclin-dependent kinases, Cell 75, 805-816.

5. Hunter, T. & Pines, J. (1994) Cyclins and cancer. 11: cyclin D and CDK inhibitors come of age, Cell 79, 573-582.

6. Michieli, P., Chelid, M., Lin, D., Pierce, J. H., Mercer, W. E. & Givol, D. (1994) Induction of WAFKIP1 by a p53-independent pathway, Cancer Rex 54, 3391-3395.

7. Macleod, K. F., Sherry, N., Hannon, G., Beach, D., Tokino, T., Kinzler, K., Vogelstein, B. &Jacks, T. (1995) p53-dependent and independent expression of p21 during cell growth, differentiation, and DNA damage, Genes & Dev. 9, 935-944.

8 . Saeed Sheikh, M., Li, X.-S., Chen, J.-C., Shao, Z.-M., Ordonez, J. V. & Fontana, J. A. (1994) Mechanisms of regulation of WAFll CIPl gene expression in human breast carcinoma: role of p53- dependent and independent signal transduction pathways, Onco- gene 9, 3407-3415.

9 . Steinman, R. A,, Hoffman, B., Iro, A., Guillouf, C., Liebermann, D. A. & El-Houseini, M. E. (1994) Induction of p21 (WAFlKIP1) during differentiation, Oncogene 9, 3389-3396.

10. Jaing, H., Lin, J., Su, Z., Collart, F. R., Huberman, E. & Fisher, P. B. (1994) Induction of differentiation in human promyelocytic HL-60 leukemia cells activates p21, WAFl/CIPl, expression in the absence of p53, Oncogene 9, 3397-3406.

11. Schwaller, J., Koeffler, H. P., Niklaus, G., Loetscher, P., Nagel, S., Fey, M. F. & Tobler, A. (1995) Posttranscriptional stabilization underlies p53-independent induction of p21 WAF1'C'P"SD'l in dif- ferentiating human leukemic cells, J. Clin. Invest. 95, 973-979.

12. Akashi, M., Hachiya, M., Osawa, Y., Spirin, K., Suzuki, G. & Koeffler, H. P. (1995) hadiation induces WAFl expression through a p53-independent pathway in KG-1 cells, J. Biol. Chem. 270, 19181-19187.

13. Datto, M. B., Li, Y., Panus, J. F., Howe, D. J., Xiong, Y. & Wang, X.-F. (1995) Transforming growth factor b induces the cyclin- dependent kinase inhibitor p21 through a p53-independent mecha- nism, Proc. Nut1 Acad. Sci. USA 92, 5545-5549.

14. Blagosklonny, M. V., Schulte, T. W., Nguyen, P., Mimnaugh, E. G., Trepel, J. & Neckers, L. (1995) Taxol induction of p2IwAF' and p53 requires c-raf-I, Cancer Res. 55, 4623-4626.

15. Plummer, J. L., Smith, B. R., Sies, H. & Bend, J. R. (1981) Chemi- cal depletion of glutathione in vivo, Methods Enzymol. 77, 50- 59.

16. Russo, T., Zambrano, N., Esposito, F., Ammendola, R., Cimino, F., Fiscella, M., Jackman, J., O'Connor, P. M., Anderson, C. W. & Appella, E. (1995) A p53-independent pathway for activation of WAFl/CIPI expression following oxidative stress, J. Biol. Chem.

17. Cowley, S., Paterson, H., Kemp, P. & Marshall, C. J. (1994) Activa- tion of MAP kinase kinase is necessary and sufficient for PC12 differentiation and for transformation of NIH 3T3 cells, Cell 77,

18. Cai, H., Szeberenyi, J. & Cooper, G. M. (1990) Effect of a dominant inhibitory Ha-ras mutation on mitogenic signal transduction of NIH 3T3, Mol. Cell. Biol. 10, 5314-5323.

19. Sun, H., Charles, C. H., Lau, L. E & Tonks, N. K. (1993) MKP-1 (3CH134), an immediate early gene product, is a dual specificity phosphatase that dephosphorylates MAP hnase in vivo, Cell 75,

20. Graham, F. L. & van der Eb, A. (1993) A new technique for the assay of infectivity of human adenovirus 5, J. Virol. 52, 456- 467.

21. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular clon- ing: u laboruior)? manual, 2nd edn, Cold Spring Harbor Labora- tory, Cold Spring Harbor NY.

22. Church, G. M. & Gilbert, W. (1984) Genomic sequencing, Proc. Nut1 Acad. Sci. USA 81, 1991-1995.

23. Gopalakrishna, R. & Anderson, W. B. (1989) Cat+ and phospho- lipid-independent activation of protein kinase C by selective oxi- dative modification of the regulatory domain, Proc. Nut1 Acad. Sci. USA, 86, 6758-6762.

24. Whisler, R. L., Goyette, M. A., Grants, I. S. & Newhouse, Y. G. (1995) Sublethal levels of oxidant stress stimulate multiple serine/ threonine kinases and suppress protein phosphatases in Jurkat T cells, Arch. Biochem. Biophys. 319, 23-35.

25. Biggs, J. R., Kudlow, J. E. & Kraft, A. S. (1996) The role of the transcription factor Spl in regulating the expression of the WAF1/ CIPl gene in U937 leukemic cells, J. Bid. Chem. 271,901-906.

26. Thompson, J. F., Hayes, L. S. & Lloyd, D. B. (1991) Modulation of firefly luciferase stability and impact on studies of gene regula- tion, Gene (Amst.) 103, 171 - 177.

27. Ammendola, R., Mesuraca, M., Russo, T. & Cimino, F. (1994) The DNA-binding efficiency of Spl is affected by redox changes, Eur. J. Biochem. 225, 483-489.

28. Schreck, R. & Baeuerle, P. A. (1991) A role for oxygen radicals as second messengers, Trends Cell Biol. 1, 39 -42.

29. Remade, J., Raes, M., Toussaint, O., Renard, P. & Rao, G. (1995) Low levels of reactive oxygen species as modulators of cell func- tion, Mut. Res. 316, 103-122.

30. Sundaresan, M., Yu, Z.-X., Ferrans, V. J., Irani, K. & Finkel, T. (1995) Requirement for generation of H,O, for platelet-derived growth factor signal transduction, Science 270, 296-299.

31. Meier, B., Radeke, H. H., Selle, S., Younes, M., Sies, H., Resch, K. & Hebermehl, G. G. (1989) Human fibroblasts release reactive oxygen species in response to interleukin-I or tumor necrosis factor-a, Biochem. .I. 263, 539-545.

32. Scott, M. D., Meshnick, S. R. & Eaton, J. W. (1989) Superoxide dismutase amplifies organismal sensitivity to ionizing radiation, J. Bid. Chem. 264, 2498-2501.

33. Halliwell, B. & Gutteridge, J. M. C. (1986) Oxygen free radicals and iron in relation to Biology and Medicine: some problems and concepts, Arch. Riochern. Biophys. 246, 501 -514.

34. Lander, H. M., Ogiste, J. S., Teng, K. K. & Novogrodsky, A. (1995) p21"' as a common signaling target of reactive free radicals and cellular redox stress, J. B i d . Chem. 270, 21 195-21 198.

3.5. Chen, C.-Y. A. & Shyn, A.-B. (1995) AU-rich elements: character- ization and importance in mRNA degradation, Trends Biochem. Sci. 20, 465-470.

36. Shiohara, M., Akashi, M., Gombart, A. F., Yang, R. & Koeffler, H. P. (1996) Tumor necrosis factor-a: posttranscriptional stabiliza- tion of WAF1 mRNA in p53-deficient human leukemic cells, J. Cell. Physiol. 166, 568-576.

37. Osawa, Y., Hachiya, M., Koeffler, P., Suzuki, G. & Akashi, M. (1995) IL-1 induces expression of WAFl mRNA in human fibro-

270, 29386-29391.

841 -852.

487-493.

Esposito et al. (Eul:

blasts : mechanisms of accumulation, Biochem. Biophys. Rex Commun. 216,429-431.

38. Canman, C. E., Gilmer, T. M., Coutts, S. B. & Kastan, M. B. (1995) Growth factor modulation of p53-mediated growth arrest versus apoptosis, Genes & Dev. 9, 600-611.

39. Gorospe, M. & Holbrook, N. J. (1996) Role of p21 in prostaglandin A2-mediated cellular arrest and death, Cancer Res. 56,475-479.

J. Biochem. 245) 737

40. Guyton, K. Z., Liu, Y., Gorospe, M., Xu, Q. & Holbrook, N. J. (1996) Activation of mitogen-activated protein kinase by H,O,, J. Biol. Chem. 271, 4138-4142.

41. Graeber, T. G., Osmanian, C., Jacks, T., Housman, D. E., Koch, C. J., Lowe, S. W. & Giaccia, A. J. (1996) Hypoxia-mediated selec- tion of cells with diminished apoptotic potential in solid tumors, Nature 379, 88-91.