1201540 A

7/28/2019 1201540 A

1/9

The large E1B protein together with the E4orf6 protein target p53 foractive degradation in adenovirus infected cells

Wilma T Steegenga1

, Nicole Riteco1

, AG Jochemsen2

, Frits J Fallaux1

and Johannes L Bos1

1 Laboratory for Physiological Chemistry, Utrecht University, PO Box 80042, 3508 TA Utrecht and 2 Laboratory of MolecularCarcinogenesis, Sylvius Laboratories, Leiden University, PO Box 9503, 2300 RA Leiden, The Netherlands

It has recently been shown that an adenovirus mutantlacking expression of the large E1B protein ( D E1B)selectively replicates in p53 decient cells. However,apart from the large E1B protein the adenovirus earlyregion encodes the E1A and E4orf6 proteins which alsohave been reported to a ect p53 expression as well as itsfunctioning. After infection with wild-type adenovirus we

observed a dramatic decrease in wild-type p53 expressionwhile no down-regulation of p53 could be detected afterinfection with the D E1B virus. The di erent e ects of thewild-type and D E1B adenovirus on p53 expression werenot only found in cells expressing wild-type p53 but werealso observed when tumor cells expressing highlystabilized mutant p53 were infected with these twoviruses. Infection with di erent adenovirus mutantsindicated the importance of a direct interaction betweenp53 and the large E1B protein for reduced p53expression after infection. Moreover, coexpression of the E4orf6 protein was found to be required for thisphenomenon, while expression of E1A is dispensable. Inaddition, we provide evidence that p53 is activelydegraded in wild-type adenovirus-infected cells but notin D E1B-infected cells.

Keywords: p53; adenovirus; E1A; large E1B; E4orf6

Introduction

DNA tumor viruses have in common that theyinactivate the tumor suppressor protein p53. Thisinactivation is essential for lytic infection and also incell transformation products of these viruses inactivate

p53. For instance, the E6 protein of the humanpapilloma viruses (HPV) 16 and 18 binds to p53 and,via interaction with the cellular E6 associating protein,causes ubiquitin-regulated degradation of p53 (Hui-bregtse et al ., 1993a; Sche ner et al ., 1990). Also thelarge E1B protein of human adenovirus Ad2 and Ad5and the large T antigen of SV40 bind to p53. Althoughin these cases p53 is stabilized rather than degraded,association of p53 with the viral proteins results in aninactivation of the tumor suppressor protein (Linzerand Levine, 1979; Oren et al ., 1981; Reich et al ., 1983;Zantema et al ., 1985).

Detailed analysis of the e ects of the adenoviruslarge E1B protein on p53 showed that the non-

oncogenic adenovirus serotypes (Ad2 and Ad5)express a large E1B protein which directly interacts

with p53 (Sarnow et al ., 1982) and subsequently causemigration of nuclear p53 into cytoplasmic bodies(Zantema et al ., 1985). Many investigators have notbeen able to nd a direct interaction of p53 with thelarge E1B protein of the oncogenic adenovirus type 12,although expression of Ad12 large E1B was found tocause p53 stabilization (Grand et al ., 1993; Mak et al .,

1988; van den Heuvel et al ., 1990, 1993; Zantema et al .,1985). Only recently, direct interaction of Ad12 largeE1B with p53 has been described (Grand et al ., 1994,1996), but this binding seems to be less stable than theone between Ad2 or Ad5 large E1B and, moreover,does not change the nuclear localisation of p53(Zantema et al ., 1985).

The interaction between p53 and the large E1Bprotein of Ad2 has extensively been studied by Berk andcoworkers. They have shown that the area betweenamino acid 224 and 354 of the large E1B protein wasessential for the binding to p53 (Kao et al ., 1990). Thisbinding does not disrupt the DNA binding capacity of p53 but tethers, via p53, the transcription-repressingdomain of E1B to p53-responsive genes and results ininhibition of transcription activation by p53 (Yew andBerk, 1992; Yew et al ., 1994). Bischo and coworkersreported recently that an adenovirus mutant lacking theexpression of the large E1B protein selectively replicatesin p53-decient cells (Bischo et al ., 1996). This resultsuggested that functionally active p53 protects cellsagainst lytic infection by the D E1B adenovirus.However, the large E1B protein is not the onlyadenoviral protein reported to a ect p53. Lowe andRuley have shown that also the E1A proteins can causep53 stabilization (Lowe and Ruley, 1993). Moreover,the E1A proteins have been observed to inhibit

transcription activation by p53 (Steegengaet al

.,1996), as well as p53-mediated repression of transcrip-tion (Horikoshi et al ., 1995). More recently, a thirdcomponent of the adenovirus early region has beenidentied to a ect p53: the E4orf6 gene product. Likethe E1A and E1B proteins, the E4orf6 protein caninhibit both activation and repression of transcriptionby p53 (Dobner et al ., 1996; Nevels et al ., 1997). Incontrast to the e ects of E1A and E1B, however,E4orf6 was observed to decrease the p53 half-life inadenovirus-transformed cells and thereby reducing thep53 expression levels (Moore et al ., 1996).

After infection of cells with the D E1B virus, asdescribed by Bischo and coworkers, the p53

regulators E1A and E4orf6 are still expressed. Theaim of our study was to examine the e ects of theseadenoviral proteins on p53 after infection by the D E1Badenovirus. In agreement with what has been publishedby Grand et al . (1994), we found a signicant reductionin p53 expression after infection of p53 procient cells

Correspondence: WT SteegengaReceived 24 April 1997; revised 1 September 1997; accepted 3September 1997

Oncogene (1998) 16, 349357 1998 Stockton Press All rights reserved 0950 9232/98 $12.00

7/28/2019 1201540 A

2/9

with wild-type adenovirus while the D E1B virus did notcause down-regulation of p53 in these cells. With theuse of di erent E1B adenovirus mutants we found thatbinding of p53 to the large E1B protein is essential forthe observed decrease in p53 expression. We provideevidence that the decreased p53 expression in theinfected cells is caused by a dramatic reduction of thehalf-life of the protein. For the active degradationexpression of both the large E1B protein and theE4orf6 protein is essential while the process isindependent of E1A expression.

Results

Complex formation between the large E1B protein and p53 after adenovirus infection is correlated withdecreased p53 expression levels

To examine the e ect of the large E1B protein on theexpression levels of p53 after adenovirus infection, therhabdoid kidney tumor cell line G401, containingfunctionally wild-type (wt) p53 (Steegenga et al .,1995b), was infected with the wt and D E1B(= dl 1520)adenovirus. Lysates were made 0, 8 and 22 h post-infection and p53 expression levels were analysed byWestern Immunoblotting. As can be seen in Figure 1a,22 h after infection with the wt virus a dramaticdecrease was observed in the expression levels of p53.In contrast, no reduction of p53 protein was foundwhen the cells were infected with the D E1B adenoviruswhile expression of the early adenoviral proteins E1A,E4orf6, E4orf6/7 and E1B/21 kDa was comparable inboth cases (Figure 1a).

We used two previously described adenovirusmutants to analyse the e ect of complex formationbetween p53 and the large E1B protein on p53expression after infection in more detail. The R443adenovirus mutant produces a large E1B proteinwhich still can bind to p53 but has lost its ability toinhibit transcription activation by p53 due to aninsertion of 4 aa on position bp3342 and the H326mutant expresses a large E1B protein not capable inbinding to p53 due to an insertion of 4 aa on positionbp2991 (Yew et al ., 1990). G401 cells were infectedwith either the R443 or the H326 adenovirus mutant

and again p53 expression was examined by WesternImmunoblotting. As can be seen in Figure 1a, afterinfection with the R443 virus, the p53 expression levelis decreased, comparable to the situation observedafter infection with the wt virus. However, afterinfection with the H326 mutant, no decreased p53expression was found. Together these results indicatethat, due to a direct interaction between p53 and thelarge adenovirus protein, p53 protein levels are down-regulated after adenovirus infection.

The e ects of infection with the same set of adenoviruses on p53 expression levels were alsoinvestigated in the osteosarcoma cell line U2OS, tocheck whether the observed e ects could be detected inother wt p53 expressing cells. Like in the G401 cells, asignicant decrease in p53 expression was observed inU2OS cells after infection with the wt adenovirus andR443 mutant, while slightly increased p53 expressionwas observed after infection with the D E1B or H326virus (Figure 1b).

Reduced expression of mutant p53 after adenovirusinfection

p53 is the most frequently mutated gene in humantumors and mutations in the p53 gene lead, in general,to conformational changes, functional loss andincreased protein stability. We were interested whetherinfection with wt adenovirus leads to down-regulation

of highly stabilized mutant p53 in human tumor celllines. As can be seen in Figure 1b, infection of theMD468 mammacarcinoma cells and C33A cervicalcarcinoma cells, both containing mutant p53, with wtand R443 adenovirus resulted in strongly decreasedexpression of p53, while infection with the two other

p53

E1A

E1B/55kDa

E4orf6

E4orf6/7

E1B/21 kDa

0 8 22 0 8 22 0 8 22 0 8 220 8 22

C o n

t r o l

W i l d

- t y p e

E 1 B

R 4 4 3

H 3 2 6

a

b

U2OS

MD468

C33A

0 8 22 0 8 22 0 8 22 0 8 220 8 22

C o n

t r o l

W i l d

- t y p e

E 1 B

R 4 4 3

H 3 2 6

Figure 1 p53 expression is di erently a ected after infection withwt or mutant adenoviruses. ( a) G401 cells were either mockinfected or infected with the wt, D E1B, R443 or H326adenoviruses. At time points 0, 8, and 22 h post-infection totalcell lysates were examined by Western Immunoblotting. Theexpression of wild-type p53 in the G401 cells was signicantlyreduced at 22 h after infection with either the wt or R443 mutantvirus, while infection with the D E1B of H326 virus did not alterthe p53 expression levels. Expression of the adenovirus proteinsE1A, large E1B, E4orf6, E4orf6/7 and E1B/21 kDa was checkedin same lysates. ( b) The same infections were performed on U2OScells, expressing wt p53, and on MD468 and C33A cells bothexpressing mutant p53. Western Immunoblotting shows a

dramatic decrease in all cell types after infection with the wtand R443 adenovirus but not after infection with the D E1B andH326 virus

Degradation of p53 after adenovirus infectionWT Steegenga et al

350

7/28/2019 1201540 A

3/9

mutant viruses did not inuence the p53 levels. Basedon these results we conclude that down-regulation of p53 after adenovirus infection not only a ects wt p53with a short half-life, as is the case in G401 and U2OScells, but alters the highly stabilized mutant p53 inMD468 and C33A cells in the same way.

Adenovirus infection overrules the e ect of adenovirustransformation on p53 expression

The results presented in Figure 1 indicate therequirement for expression of the large E1B proteinfor decreased p53 expression after adenovirus infection.In contrast, E1B causes stabilization of p53 aftertransformation resulting in increased expression levelsof the protein. To analyse these apparently contra-dictory e ects into more detail, we stably transfectedG401 cells with the Ad5 E1B/55 kDa gene. WesternImmunoblotting showed that, in agreement with earlier

studies (van den Heuvel et al ., 1993; Zantema et al .,1985), expression of the non-oncogenic Ad5 E1B/55 kDa resulted in increased p53 expression (Figure 2a).

Subsequently, these stable transfectants were in-fected with the wt and mutant adenoviruses and thee ects on p53 expression were examined. As can beseen in Figure 2b, a dramatic reduction in p53 levelswas observed after infection of two independent stabletransfectants, G55C2 and G55C5, with wt adenovirusat 22 h after infection. This result indicates that thestabilized p53 in these transfectants is still sensitive forthe down-regulatory e ect of adenovirus infection.Moreover, the D E1B and the H326 mutants, whichboth did not cause decreased expression of p53 in allthe cell lines tested so far, showed the same reductionin p53 as the wt and R443 mutant virus in both stabletransfectants. These observations indicate that stableexpression of the large E1B protein in cells cancomplement for the defects of the mutant E1B virusesand targets p53 to be down-regulated after adenovirusinfection.

For a long time it was thought that the large E1Bprotein of the oncogenic adenovirus serotype Ad12did not associate with p53, but Grand and colleaguesshowed recently that the Ad12 E1B/54 kDa proteincan bind directly to p53 (Grand et al ., 1994, 1996),although this interaction seemed to be less stable. To

examine the e ect of the Ad12 large E1B protein onthe expression of p53 after infection, we stablytransfected the G401 cell line with a Ad12 E1B/54 kDa expression vector resulting in increased p53expression levels as expected (Figure 2a). Thesituation observed after infection with the wt andmutant adenovirus di ered from the one observedafter infection of the stable transfectants expressingthe Ad5 E1B/55 kDa protein. Ad12 E1B/54 kDaexpression only partially compensates for the mu-tated E1B proteins of the D E1B and H326 viruses inthe two independent stable transfectants tested.Strongly decreased p53 expression was again foundafter infection with the wt and R443 virus. To exclude

the possibility that the intermediate e ects observedafter infection with the D E1B and H326 virus werecaused by low E1B/54 kDa expression levels, weexamined the function of p53 in these stabletransfectants. The results presented in Figure 2cshow that treatment of the Ad12 E1B/54 kDa

transfectants with camptothecin, an inducer of DNA

damage and thereby an activator of p53 (Nelson andKastan, 1994), did not result in increased p53expression and activation of the p53-responsive genep21 Waf1 could not be detected, in contrast to parentalG401 cells. From these data we conclude that theexpression levels of the Ad12 large E1B protein in

p53

Ad5 E1B/55 kDa

Ad12 E1B/54 kDa

G 4 0 1

G 5 5 C 2

G 5 5 C 5

G 5 4 C 1

G 5 4 C 7

a

G401

G55C2

G55C5

G54C1

G54C7

w t

E 1 B

R 4 4 3

H 3 2 6

w t

E 1 B

R 4 4 3

H 3 2 6

t=8 h t=22 h

p53

p21 Waf1

+ + +

c

b

G 4 0 1

G 5 4 C 1

G 5 4 C 7

Figure 2 Endogenously expressed Ad5 large E1B protein cancompletely overrule the e ects of the D E1B and H326 on p53expression after infection while Ad12 large E1B has only apartial e ect. ( a) Total cell lysates of G401 cells stablytransfected with either the non-oncogenic Ad5 large E1B orthe oncogenic Ad12 large E1B were examined by WesternImmunoblotting on p53, Ad5 E1B/55 kDa and Ad12 E1B/54 kDa expression, respectively. ( b) The same stable transfectantswere mock infected or infected with the wt, D E1B, R443 orH326 adenovirus. Western Immunoblotting of the lysates of theAd5 E1B stable transfectant showed the complete disappearanceof p53 expression at 22 h after infection with all four viruses. Inthe lysates made of the Ad12 E1B stable transfectants, p53 wasalmost absent after infection with the wt or R443 virus but onlypartially reduced after infection with the D E1B and H326adenovirus. ( c) G401 cells and the Ad12 E1B/54 kDa stabletransfectants were incubated with camptothecin (300 n M ) for20 h. The results of Western Immunoblotting show increasedexpression of p53 and p21 Waf after camptothecin treatment of the parental G401 cells but no induction was found in the E1B/54 kDa expressing transfectants with the DNA damaging agent


351

7/28/2019 1201540 A

4/9

these stable transfectants is su cient to inactivate thisp53 function completely, indicating that the partiale ects on p53 expression found after infection withthe D E1B and H326 adenoviruses are probably causedby a di erence in the interaction of the Ad5 andAd12 large E1B proteins with p53.

Active degradation of p53 by adenovirus infection

The almost complete disappearance of wt p53 at 22 hafter infection by the wt and R443 adenovirus mightbe caused by adenovirus-induced inhibition of hostcell protein synthesis. It has been reported previouslythat, due to complex formation between the large E1Band the E4orf6 proteins, the transport of newlysynthesized host cell mRNA from the nucleus to thecytoplasm is blocked after adenovirus infection(Ornelles and Shenk, 1991), thereby inhibiting hostcell protein synthesis. The reduction of highly stable

mutant p53 in the C33A and MD468 cells and thestabilized p53 in the large E1B expressing transfec-tants however, can not solely be explained by thisprocess. We previously examined the stability of p53in the MD468 cells and found the half-life of themutant p53 protein to be approximately 20 h(Steegenga et al ., 1995a). This means that even whenthe adenoviral proteins shut o host cell proteinsynthesis, at 22 h after infection at least half of thetotal amount of p53 should still be present. Based onthese data we expect p53 to be actively degraded afteradenovirus infection. To test this hypothesis MD468cells were mock-infected, or infected with wt or D E1Badenovirus and at 8 h after infection, a time point atwhich we found clear expression of E1B/55 kDa,E4orf6 and E1A (Figure 1a), a pulse chase experi-ment was performed. As can be seen in Figure 3a andb, p53 in the mock infected MD468 cells is highlystable. Treatment of the cells with the wt virus,however, results in a very rapid degradation of themutant p53, an e ect not observed after infection withthe D E1B adenovirus. These results prove that p53 isindeed actively degraded after infection with wild-typeadenovirus.

Apart from the active degradation of p53, inhibitionof host cell protein synthesis might contribute to thedecreased p53 levels observed at 22 h post-infection.

When we metabolically labelled infected G401 cells for2 h prior to isolation we found no changes in the p53synthesis at 8 h post infection but a signicant decreasein p53 synthesis was observed at 22 h post infection.As can be seen in Figure 3c at 22 h post infection inthe wild-type and R443 infected cells the strongestreduction in p53 was found but also the D E1B and theH326 mutant from which the mutated large E1Bprotein can not complex to the E4orf6 proteinanymore (Rubenwolf et al ., 1977) showed a strongdecrease in the synthesis of p53. These results indicatethat the down-regulation of host cell protein synthesisis not solely caused by the large E1B-E4orf6 complexbut that additional factors are involved. Since all

viruses cause signicant down-regulation of the p53synthesis at 22 h post infection while reduction of thep53 expression analysed by Western Immunoblotting(Figure 1) was only observed after infection with thewild-type and R443 adenoviruses we conclude thatthese two viruses make use of a specic mechanism to

cause down-regulation of the cellular p53 expression

levels.

The role of the E4orf6 protein in the degradation process

The results we have outlined above indicate animportant role for E1B in the down-regulation of

Chase time (h) 0 1/2 1 2 3 5

Control

wt

E1B

a

b

c

Figure 3 Rapid degradation of mutant p53 in the MD468 cellsinfected with wt adenovirus. ( a) MD468 cells were mock infectedor infected with wt or D E1B adenovirus and 8 h after infection

the cells were metabolically labelled for 2 h. Subsequently, thecells were incubated with normal tissue culture medium and atdi erent time points after labelling lysates were made. p53 wasimmunoprecipitated from the lysates by a mixture of PAb 122and 1801 and loaded on a SDS-polyacrylamide gel. ( b) The samegels were analysed by phosphorImaging which shows a reductionof the highly stable p53 to approximately 30 min after infectionwith wt adenovirus. ( c) G401 cells were infected and metabolicallylabelled for 2 h prior to isolation. p53 was immunoprecipitatedfrom the lysates and loaded on a SDS-polyacrylamide gel whichwas subsequently analysed by PhosphorImaging. The experimentwas performed in duplicate and the error bars represent thevariation between the two experiments


352

7/28/2019 1201540 A

5/9

p53 after adenovirus infection. Since the large E1Bprotein on its own causes stabilization of p53, therehas to be at least one additional viral protein which,in cooperation with the large E1B, is responsible forthe degradation of p53 after infection. Since it hasrecently been reported that the E4orf6 protein candecrease the p53 half-life in 293 cells expressing bothE1A and E1B (Moore et al ., 1996), the E4orf6 proteinseems to be a likely candidate. To test this hypothesis,G401 cells were infected with the previously describedD E4orf6 adenovirus mutant dl 355 (Halbert et al .,1985) and as can be seen in Figure 4a, no decrease inthe p53 expression was observed at 1 or 2 days postinfection but, in contrast, a slight increase wasdetectable. This increased p53 expression is probablydue to stabilization of p53 by the E1A and large E1Bproteins. Apparently, in the absence of the E4orf6protein the large E1B can not sensitise p53 for activedegradation. As can be seen in Figure 4a after

infection with theD

E4orf6 mutant expression of theE4orf6/7 gene product could still be detectedindicating that this protein can not down-regulatep53 expression together with the large E1B protein.As a control we infected two independent G401transfectants stably expressing the E4orf6 proteinwith the D E4orf6 virus. Figure 4a shows that alreadyat day one after infection, p53 expression is stronglyreduced in the GE4-C3 and GE4-C7 cells indicatingagain that the E4orf6 protein is responsible for the

degradation of p53 together with the large E1Bprotein.

As can be seen in Figure 4b the stably E4orf6expressing transfectants did not show a change in thep53 expression levels indicating that E4orf6 on its owncan not alter the p53 half-life.

The E1A proteins are not required for the activedegradation of p53 after adenovirus infection

Apart from the large E1B and E4orf6 proteins, alsothe E1A proteins have been found to inuence thep53 half-life (Lowe and Ruley, 1993). Furthermore,the previously reported reducing e ect of the E4orf6protein on the p53 half-life has been studied in thepresence of E1A expression (Moore et al ., 1996) andin all our infection experiments presented so far, theE1A proteins were coexpressed. To examine whetherthe E1A proteins are required for the degradation of

p53 by the Ad5 large E1B and E4orf6 proteins, aG401 transfectant stably expressing the large E1B wasinfected with Ad5 CMVLacZ. In this recombinantvirus the E1 region has been replaced by the E. coli LacZ gene and thus can not express the E1A andE1B proteins but is able to express the E4orf6protein (Kolls et al ., 1994). If E1A expression is notrequired for the active degradation of p53, infectionof the G55C5 transfectant should result in decreasedp53 expression. Indeed, when this stable transfectantis infected with the CMVLacZ adenovirus, at day 2post-infection decreased p53 was observed and atthree days after infection p53 has almost completelydisappeared (Figure 5a). Although we detect down-regulation of p53 in the absence of the E1A proteinsthe kinetics of the disappearance is delayed comparedto the situation observed after wt adenovirusinfection (Figure 1a). As can be seen in Figure 5aalso the expression of E4orf6 is retarded which canbe explained by the previously reported stimulation of E4orf6 expression by the E1A proteins (Marcellus etal ., 1996). Infection of the parental G401 cells did notshow decreased p53 expression after infection withthe CMVLacZ adenovirus while the other indepen-dent large E1B transfectant G55C2, showed the samedisappearance of p53 as was found for the G55C5(Figure 5b), E4orf6 protein expression was in all

situations comparable (Figure 5a and data notshown). As expected, we could not detect E1B norE1A expression after infection with the CMVLacZvirus (data not shown). The G401 control transfec-tant expressing the E1A proteins did not showdecreased p53 levels after infection with theCMVLacZ adenovirus (Figure 5c). From these datawe conclude that E1A expression is not required forthe active degradation of p53 after adenovirusinfection.

E1B and E4orf6 decrease p53 expression after transienttransfection

To prove that the large E1B and E4orf6 proteinstogether can target p53 for active degradation not onlyafter adenovirus infection but also in a more directassay we performed transient transfections in p53-negative Hep3B cells. As can be seen in Figure 6increased p53 expression was found after cotransfec-

0 1 0 0

5 0 0

0 1 0 0

5 0 0

0 1 0 0

5 0 0

0 1 0 0

5 0 0

G 4 0 1 / d a y

1

G 4 0 1 / d a y

2

G E 4 C 3 / d a y

1

G E 4 C 7 / d a y

1

p53

E1B/55 kDa

E4orf6

E4orf6/7

p53

E4orf6

G 4 0 1

G E 4 C 3

G E 4 C 7

a

b

Figure 4 Coexpression of E4orf6 is essential for the activedegradation of p53 after adenovirus infection. ( a) The parentalG401 cell line and the GE4 C3 and GE4C7 stable transfectantswere infected with di erent concentrations of the dl 355 adenovirus.The lysates were examined by Western Immunoblotting for p53,large E1B and E4orf6 protein expression. ( b) p53 expression in thetwo independent G401-E4orf6 stable transfectants was examinedby Western Immunoblotting


353

7/28/2019 1201540 A

6/9

tion of the large E1B protein while cotransfection of the E4orf6 protein does not alter the expression of p53.However, when the large E1B and the E4orf6 weretogether cotransfected with p53 a signicant reductionin the p53 expression was observed conrming ourprevious conclusion that together these adenovirusproteins target p53 for active degradation.

Discussion

Wt p53 has been reported to protect cells against lyticinfection by an adenovirus mutant lacking expressionof the large E1B protein ( D E1B) but not againstinfection by wt adenovirus (Bischo et al ., 1996). Thisresult suggested that the protective function of wt p53against virus replication can be inhibited by the largeE1B protein. Indeed, we observed strongly decreasedexpression of wt p53 after infection with the wtadenovirus while no reduction in p53 was detectableafter infection by the D E1B virus. These di erente ects on p53 expression by the two viruses were notonly detected for wt, instable p53 but also for highlystabilized mutant p53. By performing pulse chaseexperiments we measured a dramatic reduction of thehalf-life of mutant p53 in the MD468 cells afterinfection with the wt adenovirus while the D E1B virusdid not alter the half-life of p53. From these results we

conclude that active degradation of p53 takes placeafter wt adenovirus infection.After infection with the D E1B virus we found a

slight increase in wt p53 expression in the G401 andU2OS cells and no enhancement at all of the mutantp53 levels in the MD468 and C33A cells. Grand andcolleagues also found enhanced p53 expression afterinfection with an adenovirus mutant lacking expressionof the large E1B protein, but the e ect was muchstronger. They showed that this e ect was caused bythe E1A proteins (Grand et al ., 1994, 1995). The factthat we do not observe an increase in the half-life of mutant p53 in MD468 and C33A cells after D E1Binfection is probably due the already high stability of p53 in these cells. Also in G401 cells p53 is alreadypartially stabilized (Steegenga et al ., 1995a) explainingthe relative minor e ect of the E1A proteins on p53 inthese cells compared to the results published by Grandet al . (1994, 1995).

The data presented by Bischo and colleaguesindicate that the p53 binding domain of theadenovirus large E1B protein is essential for lyticinfection in wt p53 expressing cells (Bischo et al .,1996). We found that the previously described H326mutant virus (Yew et al ., 1990), expressing a large E1Bprotein not capable to bind to p53, could not sensitisep53 for active degradation indicating that a direct

interaction between the two proteins is essential. It hasrecently been shown that the mutation in the large E1Bprotein of the H326 virus also causes loss of thebinding of this protein to the E4orf6 product(Rubenwolf et al ., 1997) and we can not exclude thatthis property is also involved in the degradation of p53after viral infection. The results we have obtained afterinfection of the Ad5 and the Ad12 large E1B stabletransfectants show that the stability of the binding of p53 to the di erent E1B proteins is correlated with thedegree in which these large E1B proteins cancompensate for the mutations in the D E1B and H326adenovirus. These data conrm our hypothesis that adirect interaction between the large E1B and p53

proteins is important for decreased p53 levels afterinfection. In cell transformation, adenovirus large E1Bprotein, like the E1A proteins, cause p53 stabilization(Lowe and Ruley, 1993; Zantema et al ., 1985). Thissuggested to us that after adenovirus infection at leastone additional adenoviral protein has to be expressed

a

b

c

G55C5

G401

G55C2

GE1A/21 k

Ad-CMVLacZ

p53

E4orf6

E4orf6/7

+ + +

d1 d2 d3

Ad-CMVLacZ

p53

p53

+ + +

d1 d2 d3

+p53

E4orf6E4orf6/7

Figure 5 E1A expression is not required for the activedegradation of p53 induced by the large E1B and E4orf6proteins. ( a ) The G401 transfectant G55C2, stably expressingthe large E1B protein, was infected with the CMVLacZadenovirus lacking the expression of the E1 proteins. WesternImmunoblotting showed signicantly reduced p53 expression atday 2 post-infection and almost complete disappearance of p53 at3 days after infection. ( b) The same infections were performed onthe parental G401 showing no reduction of p53 expression whilean independent E1B/55 kDa expressing transfectant showed againsignicant degradation of p53. ( c) As a control, the stable G401transfectant expressing E1A together with the small E1B viruswas infected with the CMVLacZ adenovirus. Western Immuno-

blotting showed clear E4orf6 expression but no e ect on p53could be observed

0 0 0 0 0.5 0.5 0.5 0 2.5 2.5 2.50 0.5 2.5 0 0 0.5 2.5 0 0 0.5 2.5 E4orf6

E1B/55 kDa

p53

E1B/55 kDa

E4orf6

Figure 6 The large E1B and E4orf6 proteins together can down-regulate p53 in transient transfections. p53-negative Hep3B cellswere transiently transfected with 0.1 mg pCMVp53 and cotrans-fected with either pCMVneo, pCMV-E1B, pCMV-E4orf6 orcombinations of these plasmids as indicated. Lysates were madeat 24 h after transfection and analysed by Western Immunoblot-ting


354

7/28/2019 1201540 A

7/9

to observe p53 degradation. Active degradation of p53was not observed after infection of G401 cells with anD E4orf6 adenovirus mutant unless the G401 cellsstably expressed the E4orf6 protein. Infection of G401 cells with a CMVLacZ virus, lacking the E1Aand E1B genes but expressing the E4orf6 protein, didnot signicantly alter the p53 expression. However,when the CMVLacZ virus was used to infect G401cells stably expressing the large E1B protein, weobserved down-regulation of p53, indicating that theE4orf6 protein together with the large E1B protein caninduce degradation of p53. Since the E1A proteinswere not expressed under these conditions we concludethat the latter proteins are dispensable for the activedegradation of p53 after adenovirus infection.

In most normal cells the p53 protein is expressed atvery low levels due to the relatively short half-life of the protein. Maki and Howley showed that alterationin p53 ubiquitination might be responsible for the

enhanced p53 protein stability induced by u.v.radiation (Maki and Howley, 1997). However,Kubbutat and Vousden reported that also inhibitionof calpain, a nonlysosomal calcium-activated neutralprotease, is correlated with increased p53 proteinstability (Kubbutat and Vousden, 1997). Moreover,the observation made by Maki and Howley that p53stabilization after X-ray radiation is not correlatedwith changes in the ubiquitination of the tumorsuppressor protein also indicates that apart fromubiquitination at least one other mechanism canregulate p53 stability in vivo (Maki and Howley,1997). The rapid degradation of p53 as a result of HPV 16 or 18 infections has extensively been studiedand it has been shown that the HPV E6 protein can,via interaction with cellular protein E6-AP, enhanceubiquitin-mediated degradation of p53 (Huibregtse etal ., 1993a,b; Sche ner et al ., 1990). Not only wt butalso mutant p53 can be targeted for degradation by theHPV E6 (Crook and Vousden, 1992; Sche ner et al .,1992). Comparable to the situation after HPVinfection, we report here the rapid degradation of both wt and mt p53 after adenovirus infection. Thequestion remains whether the adenovirus proteins alsoactivate ubiquitin-mediated degradation of p53 and, if so, whether the E6-AP is involved in this degradationpathway. On the other hand, since also calpain can

mediate p53 protein degradation it remains possiblethat adenoviruses activate an alternative degradationpathway. The exact role of the E4orf6 protein togetherwith the large E1B in the enhanced degradation of p53is also not clear yet. Alterations in the cellularenvironment rather than resistance to normal degrada-tion were reported to be responsible for the enhancedstability of mutant p53 (Vojtesek and Lane, 1993). Ithas been shown that after transfection of 293 cells withE4orf6 p53 moved from the cytoplasmic bodies to thenucleus (Moore et al ., 1996). This translocation of p53might be important for the enhanced protein degrada-tion. Another option is that, bound to both E1B andE4orf6 the p53 conformation is changed in such a way

that the protein is sensitised for proteolytic cleavage.Since the MDM2 protein binds to the same region of p53 as the large E1B it would be interesting to examinewhether overexpression of this cellular protein can takeover the e ect of the viral protein. The molecularmechanism responsible for the decreased p53 half-life

after infection with wt adenovirus and the role of thelarge E1B and E4orf6 proteins in this process are undercurrent investigation.

Our results show that after infection with the D E1Bvirus both wt and mutant p53 remain present. Ingeneral, mutant p53 proteins have lost the normalfunctions of the tumor suppressor protein (reviewed in:Zambetti and Levine, 1993) and, in addition, havegained new properties (Dittmer et al ., 1993). Althoughmutant p53 expression is not altered after infectionwith the D E1B adenovirus, since mutant p53 has lostits wt activity it can not protect cells against lyticinfection by the adenovirus mutant. The e ects of theD E1B virus on wt p53 on the other hand is less clear.Although the E1A and E4orf6 proteins did not inducesignicant changes in p53 expression in the infectedcells these proteins have also been shown to a ecttranscription regulation by p53 (Dobner et al ., 1996;Horikoshi et al ., 1995; Nevels et al ., 1997; Steegenga et

al ., 1996). There is no evidence however, that p53protects cells against lytic infection by the D E1B virusvia regulating the expression of p53-target genes. P53has been reported to be involved in of a number of di erent cellular processes such as replication, differ-entiation of specic cell types, G1 and G2/M cell cycleregulation, apoptosis, id. (reviewed in: Gottlieb andOren, 1996; Ko and Prives, 1996). These functions of p53 might be di erently a ected by the independentviral proteins. It is possible that one or more of theseadditional features of p53 have to be inhibited by thelarge E1B protein before lytic infection can take place.Further analysis is necessary to unravel the interferenceby all the adenovirus proteins in the di erent processesregulated by p53 during infection.

Materials and methods

Tissue culture and cell lines

The rhabdoid kidney tumor cell line G401 subcloneG401.6TG.C6 (Weissman et al ., 1987) and the cell linesMD468 (Nigro et al ., 1989), C33A (Sche ner et al ., 1991),U2OS (Diller et al ., 1990), 911 (Fallaux et al ., 1996),Hep3B (Puisieux et al ., 1993) and W162 (Weinberg andKetner, 1983) were grown in Dulbecco's modied Eagle'smedium (DMEM) p lus 10% FBS. The large E1Bexpressing stable transfectants were obtained by transfect-ing G401 cells with respectively the pCMV-55K (van denHeuvel et al ., 1993) or the pCMV-54K (van den Heuvel etal ., 1993). The G401-E4 stable cell lines were made bytransfection of the pCMV34K plasmid (Dobner et al .,1996) into these cells. The G401 cell lines expressing Ad5E1A and E1B/21kDa were obtained by cotransfection of pRSV-5E1A (Jochemsen et al ., 1987)+pCMV21K (Stee-genga et al ., 1995a). All stable transfectants were culturedin the same medium as the parental G401 cell line with theaddition of G418 (300 mg/ml).

Adenoviruses and virus techniques

Besides the wt adenovirus type 5 the following mutant

viruses were used: theD

E1B (= dl 1520 or ONYX-0.015)(Bischo et al . , 1996), R443 and H326 adenovirusescontain mutations in the large E1B gene as describedearlier (Yew et al ., 1990). In the CMVLacZ adenovirus theE1 region has been replaced by the E. coli lacZ gene underthe control of the heterologous CMV promoter (Kolls etal ., 1994). In the D E4orf6 virus ( dl 355) the region between


355

7/28/2019 1201540 A

8/9

bp 2331 and 2346 of the orf6 gene has been deleted(Halbert et al ., 1985).

Small scale productions of adenovirus lots were performedessentially according to Stratford-Perricaudet and Perricaudet(Stratford-Perricaudet, 1991). Briey, near-conuent 911monolayers in 600 ml asks were infected with approxi-

mately 5 PFUs per cell in 2 ml PBS containing 2% FBS.After 1 h at 37 8C, the inoculum was replaced by fresh normaltissue culture medium. After 48 h the nearly completelydetached cells were harvested and collected in 1 ml PBS+1%FBS. Virus was isolated from the producer cells by threecycles of ash-freeze/thawing. The lysates were cleared bycentrifugation at 3000 r.p.m. for 10 min. The crude virussupernatants were layered onto CsCl cushions, 2 ml of heavy(1.45 gram/cm 3 CsCl, 10 m M Tris pH 8.1, 1 m M EDTA) and4 ml of light (1.20 gram/cm 3 CsCl, 10 m M Tris pH 8.1, 1 m MEDTA) CsCl and centrifuged in an SW41 rotor at32 000 r.p.m. for 1 h at 17 8C. The collected virus bandswere mixed with 1.33 gram/cm 3 CsCl, 10 m M Tris pH 8.1,1 mM EDTA and centrifuged in a 70Ti rotor at 55 000overnight at 17 8C. To remove the CsCl, the collected virus

bands were mixed with one volume of dialysis bu er (25 m MTris, 137 m M NaCl, 5 m M KCl, 0.73 m M Na 2 HPO 4 , 0.9 m MCaCl 2 , 0.5 m M MgCl 2 , pH 7.45) and dialysed against 2 litresof the dialysis bu er which was refreshed four times, at 4 8C.The nal dialysis was performed in the same bu ercontaining 17% glycerol. Virus stocks were stored at7 808C until further use.

Plaque assays were performed as described by Grahamand Prevec (1991). Briey, adenovirus stocks were seriallydiluted in 2 ml of DMEM without serum and added to nearconuent 911 cells in six-well plates. After 2 h of incubationat 37 8C the medium was replaced by DMEM containing0.85% agarose (Sigma), 12.3 m M CaCl 2 , 0.0025% L-glutamine and 2% FBS.

Small scale productions and plaque assays for the dl 355

adenovirus were performed according to the above describedprotocols but in W162 cells instead of in 911 cells. Toexamine the e ect of adenovirus infection on p53 expressionlevels exponentially growing cells were infected on 50 mmpetridishes with a MOI of 100 PFUs per cell unlessmentioned otherwise. The infections were performed byincubating the cells for 30 min at 37 8C with the virusdiluted in 500 ml PBS+2% FBS. Afterwards, the virus wasremoved and normal tissue culture medium was added to thecells. At the indicated time points the cells were washed twicewith ice-cold PBS and lysates were made in Giordano/E1Abu er (50 m M Tris-HCL pH 7.4, 0.25 M NaCl, 0.1% Triton,5 mM EDTA, 1 m M PMSF, 1 m M orthovandadate, 1 m Mleupeptin, 0.5 m M trypsin inhibitor and 0.1 mM aprotinin).Lysates were cleared by centrifugation at 14 000 r.p.m. for

10 min. The protein concentrations of the lysates weremeasured by the Bradford Assay (Biorad). All infectionswere performed at least two times.

Western Immunoblotting

For the experiments as depicted in Figures 1, 2, 4, 5 and 6,10 mg of the total cell lysates were boiled for 5 min inLaemmli sample bu er and subsequently separated on12.5% SDS-polyacrylamide gels. Subsequently, the pro-teins were blotted onto Protran nitrocellulose membrane(Scheicher and Schuell). The membranes were incubatedovernight with the rst antibodies: DO-1 (SanverTECH,p53-specic monoclonal antibody), M73 (E1A-specicmonoclonal antibody), 1G11 (E1B21 kDa-specic mono-

clonal antibody). RSA3 (E4orf6-specic monoclonal anti-body), A1C6 or 2A6 (both E1B/55 kDa-specic mono-clonal antibodies), 8A9 (E1B/54 kDa-specic monoclonal

antibody) or SC397 (SanverTECH, p21 Waf1 -specic poly-clonal antibody). Immune complexes were detected byenhanced chemiluminescence (Amersham).

Pulse chase experiment and metabolic labeling

Exponentially growing G401 cells were infected under thesame conditions as described in the section: `Adenovirusesand virus techniques'. At 8 h after infection the cells werelabelled with 80 mCi express protein labeling mix (NEN LifeScience) for 2 h. After labeling, the cells were washed oncewith PBS and subsequently maintained for di erent timeintervals in the normal tissue culture medium. Subse-quently, the cells were washed twice with ice-cold PBS andlysates were prepared in Giordano/E1A bu er (50 m M Tris-HCL pH 7.4, 0.25 M NaCl, 0.1% Triton, 5 m M EDTA,1 m M PMSF, 1 m M orthovandadate, 1 m M leupeptin,0.5 m M trypsin inhibitor, 0.1 mM aprotinin). Lysates werecleared by centrifugation at 14 000 r.p.m. for 10 min andsubsequently, immunoprecipitations were performed with amixture of PAb 122 (p53-specic monoclonal antibody) and

PAb 1801 (SanverTECH, p53-specic monoclonal anti-body). Immunoprecipitates were washed three times withGiordano/E1A bu er, resuspended in Laemmli samplebu er, boiled for 5 min and separated on 10% SDS-polyacrylamide gels. The gels were xated for 1 h inmethanol (25%)/acetic acid (7%) and incubated for30 min in 1 M Sodiumsaliculate. The dried gels weresubsequently exposed to Kadac XAR-5 lms at 7 80 8Cand to a PhosphorImager screen which was analysed byB&L systems Molecular Dynamics software. The experi-ment was performed twice under these conditions showingreproducibly the results as presented in Figure 4a. For themetabolic labeling cells were infected as described in thesection: `Adenoviruses and virus techniques'. At 6 and 20 hpost-infection the infected cells were labeled for 2 h with

80 mCi express protein labeling mix (NEN Life Science).Afterwards lysates were made in Giordano/E1A bu er andthe rest of the procedure was executed following theprotocol of the pulse chase experiment.

Transient transfections

Hep3B cells were transiently transfected by way of thecalcium-phosphate precipitation method on 60-mm petridishes (Steegenga et al ., 1995a). 0.1 mg pCMV-p53 wascotransfected with pCMV-55K and/or pCMV34K asindicated in Figure 6. With the addition of pCMV-neo-Bam the total amount of CMV-containing plasmid wasadjusted to 5.1 mg in all precipitates and in addition 4.9 mgof salm sperm DNA was used as carrier DNA. Lysates

were made in Giordano/E1A bu er and protein concentra-tions were measured by the Bradford (Biorad) assay. Fromeach sample 10 mg protein was loaded on a SDS page,blotted onto nitrocellulose and analysed with the 1801(SanverTECH), A1C6 and RSA3 antibodies.

AcknowledgementsWe would like to thank Dr A Berk for the generous gift of the R443 and the H326 mutant adenoviruses and the 2A6monoclonal antibody and Dr T Shenk for the kind gift of the pCMV-E4orf6 expression vector, the dl 355 adenovirusmutant and the RSA3 monoclonal antibody. Furthermorewe thank Dr G Ketner for the W162 cell line and Dr Kollsfor the CMVLacZ adenovirus and Dr A Balmain and Dr I

Ganly for the D E1B adenovirus. We thank Dr A Shvartsfor carefully reading the manuscript. This project wassupported by a grant from the Dutch Cancer Society.


356

7/28/2019 1201540 A

9/9

References

Bischo JR, Kirn DH, Williams A, Heise C, Horn S, MunaM, Ng L, Nye JA, Sampson-Johannes A, Fattaey A andMcCormick F. (1996). Science , 274, 373 376.

Crook T and Vousden KH. (1992). EMBO J. , 11, 3935

3940.Diller L, Kassel J, Nelson CE, Gryka MA, Litwak G,Gebhardt M, Bressac B, Ozturk M, Baker SJ, Vogelstein Bet al . (1990). Mol. Cell Biol. , 10, 5772 5781.

Dittmer D, Pati S, Zambetti G, Chu S, Teresky AK, MooreM, Finlay C and Levine AJ. (1993). Nat. Genet. , 4, 4246.

Dobner T, Horikoshi N, Rubenwolf S and Shenk T. (1996).Science , 272, 1470 1473.

Fallaux FJ, Kranenburg O, Cramer SJ, Houweling A, VanOrmondt H, Hoeben RC and Van Der Eb AJ. (1996).Hum. Gene Ther. , 7, 215 222.

Gottlieb TM and Oren M. (1996). Biochim. Biophys. Acta. ,1287, 77 102.

Graham FL and Prevec L. (1991). Gene transfer and expression protocols. The humana press, Clifton, NJ.

Grand RJ, Grant ML and Gallimore PH. (1994). Virology ,203, 229 240.Grand RJ, Lecane PS, Owen D, Grant ML, Roberts S,

Levine AJ and Gallimore PH. (1995). Virology , 210, 323 334.

Grand RJ, Lecane PS, Roberts S, Grant ML, Lane DP,Young LS, Dawson CW and Gallimore PH. (1993).Virology , 193, 579 591.

Grand RJ, Owen D, Rookes SM & Gallimore PH. (1996).Virology , 218, 2334.

Halbert DN, Cutt JR and Shenk T. (1985). J. Virol. , 56,250 257.

Horikoshi N, Usheva A, Chen J, Levine AJ, Weinmann Rand Shenk T. (1995). Mol. Cell. Biol. , 15, 227 234.

Huibregtse JM, Sche ner M & Howley PM. (1993a). Mol.

Cell. Biol. , 13, 775 784.Huibregtse JM, Sche ner M & Howley PM. (1993b). Mol.Cell. Biol. , 13, 4918 4927.

Jochemsen AG, Peltenburg LT, te Pas MF, de Wit CM, BosJL and van der Eb AJ. (1987). EMBO J. , 6, 3399 405.

Kao CC, Yew PR and Berk AJ. (1990). Virology , 179, 806 814.

Ko LJ and Prives C. (1996). Genes Dev. , 10, 1054 1072.Kolls J, Peppel K, Silva M and Beutler B. (1994). Proc. Natl

Acad. Sci. USA , 91, 215 219.Kubbutat MH and Vousden KH. (1997). Mol. Cell. Biol. , 17,

460 468.Linzer DI and Levine AJ. (1979). Cell , 17, 4352.Lowe SW and Ruley HE. (1993). Genes Dev. , 7, 535 545.Mak I, Mak S and Benchimol S. (1988). Virology , 163, 201

204.Maki CG and Howley PM. (1997). Mol. Cell. Biol. , 17, 355 363.

Marcellus RC, Teodoro JG, Wu T, Brough DE, Ketner G,Shore GC and Branton PE. (1996). J. Virol. , 70, 6207 6215.

Moore M, Horikoshi N and Shenk T. (1996). Proc. Natl Acad. Sci. USA , 93, 11295 11301.

Nelson WG and Kastan MB. (1994). Mol. Cell. Biol. , 14,1815 1823.

Nevels M, Rubenwolf S, Spruss T, Wolf H and Dobner T.(1997). Proc. Natl Acad. Sci. USA , 94, 1206 1211.Nigro JM, Baker SJ, Preisinger AC, Jessup JM, Hostetter R,

Cleary K, Bigner SH, Davidson N, Baylin S, Devilee P. etal . (1989). Nature , 342, 705 708.

Oren M, Maltzman W and Levine AJ. (1981). Mol. Cell.Biol. , 1, 101 110.

Ornelles DA and Shenk T. (1991). J. Virol. , 65, 424 429.Puisieux A, Galvin K, Troalen F, Bressac B, Marcais C,

Galun E, Ponchel F, Yakicier C, Ji J and Ozturk M.(1993). FASEB J. , 7, 1407 1413.

Reich NC, Sarnow P, Duprey E and Levine AJ. (1983).Virology , 128, 480 484.

Rubenwolf S, Schutt H, Nevels M, Wolf H and Dobner T.(1997). J. Virol. , 71, 1115 1123.

Sarnow P, Ho YS, Williams J and Levine AJ. (1982). Cell , 28,387 394.Sche ner M, Munger K, Byrne JC and Howley PM. (1991).

Proc. Natl Acad. Sci. USA , 88, 5523 5527.Sche ner M, Takahashi T, Huibregtse JM, Minna JD and

Howley PM. (1992). J. Virol. , 66, 5100 5105.Sche ner M, Werness BA, Huibregtse JM, Levine AJ and

Howley PM. (1990). Cell , 63, 1129 1136.Steegenga WT, Shvarts A, van Laar T, van der Eb AJ and

Jochemsen AG. (1995a). Oncogene , 11, 4957.Steegenga WT, van Laar T, Riteco N, Mandarino A, Shvarts

A, van der Eb AJ and Jochemsen AG. (1996). Mol. Cell.Biol. , 16, 2101 2109.

Steegenga WT, van Laar T, Shvarts A, Terleth C, van der EbAJ and Jochemsen AG. (1995b). Virology , 212, 543 554.

Stratford-Perricaudet LD and Perricaudet M. (1991). HumanGene Transfer , Paris, Inserm John Libbey Euro-text.Van den Heuvel SJ, van Laar T, Kast WM, Melief CJ,

Zantema A and van der Eb AJ. (1990). EMBO J. , 9, 2621 2629.

Van den Heuvel SJ, van Laar T, The I and van der Eb AJ.(1993). J. Virol. , 67, 5226 5234.

Vojtesek B and Lane DP. (1993). J. Cell. Sci. , 105, 607 612.Weinberg DH and Ketner G. (1983). Proc. Natl Acad. Sci.

USA , 80, 5383 5386.Weissman BE, Saxon PJ, Pasquale SR, Jones GR, Geiser AG

and Stanbridge EJ. (1987). Science , 236, 175 180.Yew PR and Berk AJ. (1992). Nature , 357, 8285.Yew PR, Kao CC and Berk AJ. (1990). Virology , 179, 795

805.

Yew PR, Liu X and Berk AJ. (1994). Genes Dev. , 8, 190 202.Zambetti GP and Levine AJ. (1993). FASEB J. , 7, 855 865.Zantema A, Schrier PI, Davis-Olivier A, van Laar T, Vaessen

RT and van der EA. (1985). Mol. Cell. Biol. , 5, 3084 3091.


357

Documents

1201540 A