Apoptosis Modulation of Epstein–Barr Virus-Encoded Latent Membrane Protein 1in the Epithelial Cell Line HeLa Is Stimulus-Dependent

Xiangning Zhang,* LiFu Hu,* Bengt Fadeel,† and Ingemar T. Ernberg*,1

*Microbiology and Tumor Biology Center, and †Division of Toxicology, Institute of Environmental Medicine,Karolinska Institutet, S-171 77 Stockholm, Sweden

Received February 25, 2002; returned to author for revision April 5, 2002; accepted July 3, 2002

Epstein–Barr virus (EBV)-encoded latent membrane protein 1 (LMP1) is required for viral transformation and has beenshown to protect lymphocytes from apoptosis. However, the effect of LMP1 on cells of epithelial origin remains poorlyunderstood. Using the epithelial cell line HeLa in which the expression of LMP1 is inducibly regulated by tetracycline, wedemonstrate that apoptosis triggered by ligation of the death receptor, Fas, or by the chemotherapeutic agent, etoposide, ispotentiated by LMP1. Apoptosis was assessed by nuclear condensation and activation of caspase-3-like enzymes withconcomitant proteolysis of the nuclear caspase substrate, poly(ADP-ribose) polymerase. However, the effect of LMP1 inHeLa cells appeared to be stimulus-dependent since apoptosis induced by tumor necrosis factor (TNF) was inhibited.Moreover, we observed an upregulation of the zinc finger protein A20 and a decrease in expression of Bcl-2 upon inductionof LMP1 in HeLa cells. Taken together, these data further our understanding of the function of LMP1 in epithelial cells andsuggest that LMP1, similar to its mammalian homolog CD40, can exert opposing effects on cell survival depending on the

Key Words: apoptosis; caspases; Epstein–Barr virus; Fas

INTRODUCTION

Epstein–Barr virus (EBV) latent membrane protein 1(LMP1) is essential for transformation of primary B lym-phocytes into indefinitely proliferating lymphoblastoidcell lines (Kaye et al., 1993). LMP1 confers a survivaladvantage to EBV-infected B cells by protecting thesecells from apoptosis, and this was shown to depend onLMP1-induced upregulation of the anti-apoptotic proteinBcl-2 (Henderson et al., 1991). LMP1 also inhibits p53-triggered apoptosis of Burkitt lymphoma (BL) cellsthrough a similar upregulation of Bcl-2 (Okan et al., 1995).Inhibition of apoptosis is generally perceived of as beingan important step in the development of cancer (forreview, see Hanahan and Weinberg, 2000). LMP1 there-fore has the potential to contribute to pathogenesis ofEBV-associated malignancy through the promotion ofcell survival as well as cell growth, at least in B cells.However, it has recently been shown that other EBVantigen(s) than LMP1 may play a role in apoptotic inhi-bition of cells harboring EBV (Wade and Allday, 2000).Indeed, EBV-encoded BHRF1, an early lytic protein withsequence homology to Bcl-2, has been shown to sup-press apoptosis induced by diverse stimuli, includingDNA-damaging agents, serum depletion, and calciumionophore stimulation (Henderson et al., 1993; Tarodi et

0042-6822/02 $35.00© 2002 Elsevier Science (USA)All rights reserved.

330

; CD95; latent membrane protein 1; tumor necrosis factor.

al., 1994). Taken together, these observations suggestthat the apoptosis-modulating activity of LMP1 is condi-tional, perhaps depending on the cell type or mode ofapoptosis induction.

In addition to lymphocytes, EBV also infects epithelialcells. A tight association with EBV has been seen inundifferentiated forms of nasopharyngeal carcinoma(NPC), and about two-thirds of the EBV-positive NPC alsoexpress LMP1 (Hu et al., 1995), raising the possibility thatthis viral protein may play a role in the promotion ofmalignant growth in epithelia at certain anatomical sites.However, there are few reports on the role of LMP1 in themodulation of apoptosis in cells of epithelial origin. Pre-vious studies showed that when expressed in the lungcancer cell line H1299, LMP1 blocks p53-mediated apo-ptosis through its induction of the zinc finger protein A20,an anti-apoptotic molecule (Fries et al., 1996). On theother hand, LMP1 has been reported to enhance tumornecrosis factor (TNF)-mediated apoptosis in intestine407 epithelial cells, despite the occurrence of A20 induc-tion (Kawanishi, 2000). Whether LMP1 modulates epithe-lial cell apoptosis instigated by other stimuli has notbeen established.

LMP1 is known to functionally mimic CD40 signalingin B cells (Busch and Bishop, 1999; Uchida et al., 1999).The latter molecule is a member of the TNF receptor(TNFR) family and is expressed on the surface of acti-

nature of the apoptosis trigger. © 2002 Elsevier Science (USA)

1 To whom correspondence and reprint requests should be ad-dressed. Fax: �46 8 31 94 70. E-mail: ingemar.ernberg@mtc.ki.se.

Virology 304, 330–341 (2002)doi:10.1006/viro.2002.1640

vated lymphocytes. It exerts growth-promoting effects inB lymphocytes, but inhibits growth and proliferation in

; APO-1

other cell types (for review, see Young et al., 1998).Interestingly, the apoptosis-modulating properties ofCD40 have been shown to be opposing and to dependon the nature of the apoptotic stimulus. Hence, ligation ofCD40 rescues Ramos-BL cells from calcium ionophore-and antigen receptor-triggered apoptosis (An and Knox,1996), yet potentiates Fas-mediated apoptosis of thesame cells, despite the upregulation of the anti-apoptoticmolecules Bcl-2 and Bcl-XL under these conditions (An etal., 1997). LMP1 and CD40 have been shown to bindseveral of the same TNFR-associated factors (TRAFs),namely TRAF-2, -3, and -5 (Young et al., 1998), and thisinteraction thus provides further evidence that LMP1-mediated signaling can intersect the TNFR pathway. Inaddition, the adaptor protein termed TNFR-associateddeath domain protein (TRADD) has been reported to bindLMP1 (Floettmann et al., 1998), but does not bind CD40directly, indicating the possibility for differential signalingmediated by these two molecules. Importantly, TNFR1and the so-called death receptor known as Fas (APO-1/CD95) have both been implicated in the physiologicalregulation of apoptosis in cells of various origin, includ-ing lymphocytes (for review, see Krammer, 2000). Fasalso recruits an adapter protein, designated Fas-associ-ated death domain protein (FADD), upon binding of itscognate ligand, and this results in the aggregation andactivation of upstream caspases, thus initiating a cas-cade of proteolytic activity with subsequent cleavage ofcellular proteins and dismantling of the cell by apoptosis(Krammer, 2000). In addition to death receptor-mediatedtriggering of apoptosis, a number of death stimuli, includ-ing DNA-damaging agents such as etoposide, elicit mi-tochondrial perturbation with release of apoptogenic fac-tors from these organelles and subsequent activation ofcaspases, including caspase-3, resulting in the rapiddemise of the cell (for review, see Kroemer and Reed,2000).

In the present study, we utilized a tetracycline-regu-lated LMP1 expression system (Floettmann et al., 1996)to obtain inducible LMP1 expression in the epithelial cellline HeLa. Apoptosis of HeLa cells was triggered bythree different stimuli acting via death receptors or at themitochondrial level under conditions which LMP1 wasinduced or uninduced, and assessed by the activation ofcaspase-3-like enzymes and characteristic nuclear apo-ptotic changes. We observed a stimulus dependence ofapoptosis modulation by the inducible LMP1 in HeLacells insofar as TNF-triggered apoptosis was attenuatedby LMP1, whereas Fas- and etoposide-triggered apopto-sis was potentiated in the same cells. The implications ofthese findings for EBV-associated malignancy, and thepossible mechanism(s) underlying the differential out-come of apoptosis triggering in LMP1-expressing cells,are discussed.

RESULTS

Establishment of a tetracycline-regulated LMP1expression system in the epithelial cell line HeLa

HeLa cells were transfected with the tTA-containingplasmid pJEF3 and selected by in vitro culture in thepresence of 500 �g/ml hygromycin B. Outgrowing cellspresenting as visible colonies were transferred to cultureplates, and the clones were tested for tTA activity bytransient transfection with the tTA reporter plasmidpUHC13–3, containing a luciferase gene downstream oftTA responder. Clone Ht27 (Ht � HeLa tTA), selected outof 31 clones, showed the lowest luciferase activity in thepresence of tetracycline (1 �g/ml) and the highest activ-ity in the absence of tetracycline (data not shown). Thisclone was therefore chosen for the next round of trans-fection involving a tTA responder plasmid containingLMP1 cDNA. Outgrowing cells after selection with 2mg/ml geneticin were divided and cultured in the pres-ence or absence of tetracycline. Tetracycline-regulatedLMP1 expression was monitored by Western blottingusing antibodies specific for LMP1. As shown in Fig. 1A,several clones exhibited regulatable LMP1 expression.In addition to stable induction of LMP1 upon multiplepassages, the Ht27/8 clone also displayed marked NF�Bactivation, as determined by the conventional luciferaseassay (Fig. 1B), and this clone was therefore selected foruse in further experiments.

Etoposide-triggered apoptosis of HeLa cells andconcomitant activation of caspase-3-like enzymes isenhanced by inducible LMP1

When HeLa cells were exposed to etoposide, a topo-isomerase II poison known to elicit apoptosis in vitro,blebbing, and later floating cells, indicative of apoptoticcell death, were observed (data not shown). Cells ex-posed to etoposide (20 �g/ml) were also stained with thenuclear dye Hoechst 33258, and the number of apoptoticcells exhibiting fragmented nuclei were counted (Fig. 2).The percentage of apoptotic cells was completely sup-pressed when cells were cultured in the presence of thepan-caspase inhibitor zVAD-fmk2 (10 �M), indicating thatcell death in this model is caspase-dependent. To deter-mine the effect of LMP1 on apoptosis, HeLa cells werecultured in the presence or absence of tetracycline for48 h after plating and then incubated in medium supple-mented with etoposide (20 �g/ml) in the continued pres-ence or absence of tetracycline. The earliest signs ofblebbing were seen at 10 h after the initiation of culture,

2 Abbreviations: BHRF1, BamHI rightward reading frame 1; DEVD-AMC, aspartate-glutamate-valine-aspartate-7-amino-4-methyl-cou-marin; tTA, tetracycline transactivator; zVAD-fmk, benzyloxycarbonyl-valine-alanine-aspartate- fluoromethylketone.

331STIMULUS-DEPENDENT APOPTOSIS MODULATION OF EBV LMP1

and blebbing and floating cells continued to appearduring after 2–3 days of exposure to etoposide (Fig. 3Aand data not shown). Etoposide-triggered cells were alsocollected at different time points and analyzed for acti-vation of caspase-3-like enzymes using the fluorogenicpeptide substrate, DEVD-AMC (50 �M) (Nicholson et al.,1995). As shown in Fig. 3B, etoposide-induced caspaseactivity was enhanced by the expression of LMP1. On theother hand, caspase activation in the absence of etopo-side in HeLa cells expressing or not expressing LMP1was minimal (Fig. 3C). Importantly, the addition of thecompetitive caspase-3 inhibitor DEVD-CHO (50 nM) to

the cell lysates completely abrogated DEVDase activity,thus confirming the specificity of these lysates for thecaspase-3 substrate (Fig. 3D). As expected, etoposide-induced DEVDase activity, and its potentiation by LMP1,was blocked by zVAD-fmk (data not shown). Hence,LMP1 potentiates etoposide-triggered apoptosis up-stream of the activation of caspase-3-like enzymes.

Fas-triggered apoptosis of HeLa cells is enhanced byinducible LMP1

Agonistic anti-Fas antibodies are frequently used tomimic the effect of the endogenous Fas ligand. We foundthat HeLa cells are sensitive to treatment with anti-Fasantibodies (100 ng/ml), albeit to a lesser extent than toetoposide, with pronounced blebbing of cells starting tooccur at 4–5 h after exposure (data not shown). LMP1-mediated modulation of apoptosis in this model wasconfirmed by assessment of Hoechst 33258 staining offragmented nuclei (Fig. 4A). Fas-triggered (100 ng/ml)cells were also collected at the indicated time points andmonitored for AMC release in a continuous fluorometriccaspase assay as described above (Fig. 4B). Importantly,Fas-triggered caspase-3-like activity was consistentlyenhanced by the expression of LMP1. A nonspecific,toxic effect of LMP1 can thus be excluded since caspaseactivation is an indicator of apoptotic—as opposed tonecrotic—cell death (Samali et al., 1999). To control forthe inducibility of LMP1 upon Fas ligation, cells wereanalyzed by Western blot at various time points afterexposure to anti-Fas antibodies. The EBV-positive B cellline Raji was used as a positive control. As shown in Fig.4C, the inducibility of LMP1 was indeed sustained uponFas triggering, although the level of expression of LMP1was decreased at later time points, most likely as aresult of a loss of cells through detachment. Moreover,while LMP1 was shown to induce Fas expression insome cell types (Devergne et al., 1998), our flow cyto-metric analysis of the surface expression of Fas in thepresence or absence of LMP1 appears to rule out sucheffects in the present model (Fig. 4D). Taken together, thecurrent data demonstrate that inducible LMP1 potenti-ates Fas-triggered apoptosis of HeLa cells.

TNF-induced apoptosis and caspase activation inHeLa cells is inhibited by inducible LMP1

The TNFR triggers a signaling cascade similar but notidentical to the Fas-mediated pathway (see Introduction).To test whether the effect of LMP1 was specific for Fas,the effect of TNF on HeLa cells was examined. After 2 hof exposure to TNF (5 ng/ml) and cycloheximide (CHX; 20�g/ml), a few blebbing cells were evident, and at 10 hafter exposure, a large number of blebbing and floatingcells started to accumulate (data not shown). Apoptosisinduced by TNF was confirmed by staining of cells withHoechst 33258, and typical fragmented nuclei were vis-

FIG. 1. Tetracycline-regulated expression of LMP1 in HeLa cells. (A)Three HeLa clones derived from outgrowing cells after selection inhygromycin and geneticin were cultured in the presence or absence oftetracycline (1 �g/ml) for 24 h. Cell lysates were then assessed byWestern blotting using antibodies specific for LMP1 (clone S-12). TheEBV-positive B cell line Raji is used as a positive control. Ht � HeLa tTA(tetracycline transactivator). (B) HeLa cells (clone Ht27/8) were culturedin the presence or absence of tetracycline for the indicated time points,and NF�B activity was determined using the luciferase assay as de-scribed under Materials and Methods. The activity registered in tetra-cycline (tet) � samples is set at 1 arbitrary unit and the activity in tet�samples is presented as fold numbers of induction. Data shown aremean � SD of three independent experiments.

332 ZHANG ET AL.

ible in cells upon TNFR1 ligation (Fig. 5A). Activation ofcaspase-3-like enzymes was monitored as above usingthe peptide substrate DEVD-AMC (50 �M). Interestingly,a marked inhibition by LMP1 of TNF-triggered caspase-3-like enzymes was observed (Fig. 5B). Similarly, Westernblotting showed that processing of poly(ADP-ribose)polymerase (PARP), a known substrate for caspase-3,and the generation of the characteristic 85 kDa fragment,was attenuated in the LMP1-expressing cells (Fig. 5C).Fas-triggered Raji cells were used as a positive control.Importantly, Western blotting of HeLa cells expressing ornot expressing LMP1 demonstrated that LMP1 did notaffect the level of expression of TNFR1 in these cells (Fig.5D). These data thus show that inducible LMP1 attenu-

ates TNF-triggered apoptosis of HeLa cells, upstream ofcaspase-3 activation.

Inducible LMP1 regulates the expression of A20 andBcl-2 in HeLa cells

The anti-apoptotic effect of LMP1 in certain cell typeshas been demonstrated to correlate with an upregulationof the zinc finger protein A20 (Fries et al., 1996). More-over, an LMP1-induced upregulation of Bcl-2 has beenreported in B cells (Henderson et al., 1991), while LMP1has no effect on Bcl-2 levels in the leukemic T cell lineJurkat (Kawanishi, 1997a). Therefore, we investigated theeffect of LMP1 in HeLa cells on these apoptosis-regulatingmolecules using antibodies specific for A20 and Bcl-2,

FIG. 2. Etoposide-triggered apoptosis of HeLa cells is caspase-dependent. Cells were incubated in the presence or absence of the pan-caspaseinhibitor, zVAD-fmk (10 �M), prior to stimulation with etoposide (20 �g/ml) for the indicated time points. Apoptosis was determined as the percentageof cells exhibiting characteristic fragmented nuclei. Data shown are mean � SD of one out of three independent experiments.

333STIMULUS-DEPENDENT APOPTOSIS MODULATION OF EBV LMP1

respectively. As seen in Fig. 6A, an induction of LMP1 anda concomitant upregulation of A20 can be seen in HeLacells incubated for 48 h in the absence of tetracycline. Onthe other hand, the level of Bcl-2 is markedly decreased

upon induction of LMP1 in these cells (Fig. 6B). Takentogether, these data suggest that LMP1 can exert selective,cell lineage specific effects on the expression of the anti-apoptotic molecules A20 and Bcl-2.

FIG. 3. Inducible LMP1 potentiates etoposide-triggered apoptosis in HeLa cells. (A) HeLa cells expressing or not expressing tetracycline-regulatable LMP1 were cultured in the presence or absence of etoposide (20 �g/ml) for 24 h prior to assessment of Hoechst 33258 staining undera fluorescence microscope. Arrows indicate typical apoptotic cells with fragmented nuclei. Original magnification, �60. (B) HeLa cells incubated asabove were monitored for caspase-3-like activity using the fluorescent peptide substrate, DEVD-AMC (50 �M). The maximum rate of AMC release(pmol/min) was estimated by linear regression (r2 � 0.99). Data shown are mean � SD of three independent experiments. (C) Caspase-3-like activityin Ht27/8 cells cultured in regular medium, i.e., in the absence of etoposide. Cells were cultured with or without tetracycline, and caspase-3-likeenzyme activity was measured as described above. Data shown are mean � SD derived from two independent experiments. (D) Real-time recordingof AMC release in lysates of HeLa cells stimulated for 24 h with 20 �g/ml etoposide in the absence (open circles) or presence (solid circles) of thecompetitive inhibitor of caspase-3-like enzymes, DEVD-CHO (50 nM). Each data point represents the mean value of determinations performed intriplicate.

334 ZHANG ET AL.

FIG. 4. Inducible LMP1 potentiates Fas-triggered apoptosis in HeLa cells. (A) HeLa cells expressing or not expressing tetracycline-regulatableLMP1 were cultured in the presence or absence of agonistic anti-Fas antibodies (100 ng/ml) and then assessed for fragmented nuclei using theHoechst 33258 dye. Typical fragmented nuclei are indicated with arrows. Original magnification, �60. (B) HeLa cells were incubated as above for theindicated time points and caspase-3-like activity was monitored by the in vitro cleavage of the fluorescent substrate DEVD-AMC (50 �M). Data shownare mean � SD of three independent experiments. (C) HeLa cells cultured in the presence or absence of tetracycline (tet; 1 �g/ml) and thencoincubated with agonistic anti-Fas antibodies (100 ng/ml) for the indicated time points were assessed for LMP1 expression by Western blotting. TheEBV-positive B cell line Raji was used as a positive control. (D) Flow cytometric analysis of Fas surface expression in HeLa cells cultured with orwithout tetracycline for 24 h. The leukemic T cell line Jurkat was used as a positive control. The solid line depicts staining with the anti-Fas mAb; thedotted line depicts secondary antibody alone. Data are representative for three independent experiments.

335STIMULUS-DEPENDENT APOPTOSIS MODULATION OF EBV LMP1

DISCUSSIONLMP1 has been reported to modulate lymphocyte ap-

optosis through its induction of the anti-apoptotic mole-cules, Bcl-2 (Henderson et al., 1991) and A20 (Laherty etal., 1992), and more recently through the upregulation ofthe Bcl-2 homologues, Mcl-1 and Bfl-1 (Wang et al., 1996;

D’Souza et al., 2000). In the present study, apoptosismediated by TNFR1, but not by the related death receptorFas or the chemotherapeutic agent etoposide, was in-hibited by tetracycline-regulated LMP1 in the epithelialcell line HeLa. Moreover, the modulation of TNF-trig-gered apoptosis correlated with an induction of A20 in

FIG. 5. Inducible LMP1 attenuates TNF-triggered apoptosis in HeLa cells. (A) Apoptosis was detected by Hoechst 33258 staining of LMP1-expressing or nonexpressing HeLa cells exposed for 4 h to TNF (5 ng/ml) plus CHX (20 �g/ml). A typical apoptotic cell is indicated with an arrow.Original magnification, �60. (B) Caspase-3-like activity was monitored in lysates of cells treated as above for the indicated time points. The maximumrate of DEVD-AMC cleavage (pmol/min) was estimated by linear regression (r2 � 0.99). Data shown are mean � SD of three independent experiments.(C) HeLa cells (LMP1-expressing and nonexpressing) incubated in the presence or absence of TNF (5 ng/ml) for 4 h were assessed by Westernblotting using antibodies specific for PARP. Raji cells exposed to agonistic anti-Fas antibodies (100 ng/ml) for 8 h were used as a positive control.The intact PARP molecule (116 kDa) and the characteristic caspase-mediated cleavage product (85 kDa) are indicated with arrows. (D) TNFR1expression as determined by Western blot. Equal numbers of HeLa cells induced and uninduced for LMP1 expression were probed with anti-TNFR1and anti-LMP1 antibodies, respectively. Raji cells were used as a positive control. Data shown in C and D are representative of at least threeindependent experiments.

336 ZHANG ET AL.

these cells. Previous studies have shown that an en-forced expression of A20 fails to diminish Fas-mediatedapoptosis (Jaattela et al., 1996), and the fact that LMP1yielded a potentiation of Fas-triggered apoptosis in HeLacells despite the upregulation of A20 is therefore notsurprising. Studies on the apoptosis-modulating functionof LMP1 in epithelial cells are scarce; nevertheless,published observations suggest that LMP1, when ex-pressed in epithelial cells, exhibits a uniform pattern ofinhibition or promotion of apoptosis in the same cell line(Fries et al., 1996; Kawanishi, 2000). However, in themodel system used herein, its modulation is clearly stim-ulus-dependent. The current observations are thus rem-iniscent of those obtained in studies of the EBV-encodedprotein BHRF1, which also exerts an apoptosis-modulat-ing effect which is stimulus- and cell type-dependent(Foghsgaard and Jaattela, 1997). Furthermore, whileLMP1 expression in NPC cells has been shown to confergrowth enhancement, expression of LMP1 in gastric car-cinoma cells resulted in enhanced apoptosis and hencein growth suppression (Sheu et al., 1998), thus indicatingthat the context of LMP1 expression is crucial for itseffect. An elevated A20 protein level may explain theinhibition of TNF-induced apoptosis described herein, asit has been reported that A20 selectively blocks thepathways stimulated by certain apoptotic triggering

agents, notably TNF (Opipari et al., 1992; Jaattela et al.,1996). Another possible mechanism of apoptosis inhibi-tion involves the molecular interaction of the intracellularportion of LMP1. Two functional domains in the C-termi-nal cytoplasmic tail of LMP1 associate with TNFR-asso-ciated factors and the TNFR-associated death domainprotein, respectively, and mediate NF�B activation (Izumiet al., 1999 and references therein). The apoptosis-atten-uating effect of LMP1 in cells exposed to TNF could thusbe explained by its possible competitive binding withTRADD and subsequent attenuation of the cytotoxic sig-nal mediated through TNFR1. However, it is also reason-able to consider a role for NF�B in the apoptosis mod-ulation exerted by LMP1. Hence, we have observed apotentiation of etoposide-induced caspase activation inLMP1-expressing HeLa cells incubated in the presenceof a chemical NF�B inhibitor (X. Zhong, L. Hu, B. Fadeel,and I. Ernberg, unpublished data). Clearly, further studiesare needed to delineate the putative role of NF�B, and ofTNFR-associated proteins, in the apoptosis signaling inthe current model.

Remarkably, while TNF-induced apoptosis is attenu-ated by LMP1, our data show that caspase activation andapoptosis of HeLa cells induced via the TNFR-relateddeath receptor, Fas, is enhanced. LMP1 thus differs fromthe other EBV-encoded apoptosis-regulating proteinBHRF1, which was shown previously to protect epithelialcells from apoptosis induced either through TNFR1 orFas stimulation (Kawanishi, 1997b). On the other hand,Lu et al. (1996) reported the induction of apoptosis in theepithelial cell line Rhek-1 by high levels of LMP1, aneffect that was most apparent under conditions of serumstarvation. Upregulation of Fas expression upon ligationof CD40, the mammalian homolog of LMP1, has beendemonstrated (An et al., 1997) and these findings weretaken to account for the apoptosis-enhancing effect ofCD40. However, we failed to detect any consistent alter-ations in surface expression of Fas upon LMP1 inductionin the present model. Apoptosis triggered by TNFR1 andFas shares many common features, such as the activa-tion of “executioner” caspases, yet there are also differ-ences which may account for the differential apoptosis-modulating effect of LMP1. For instance, Fas signaling incertain cell types involves mitochondrial amplification ofthe original stimulus, with caspase-8-mediated cleavageof the Bcl-2 family member Bid acting as a link betweenthe death receptor at the plasma membrane and mito-chondria (Scaffidi et al., 1998; Yin et al., 1999). Impor-tantly, etoposide-triggered apoptosis of HeLa cells isalso enhanced by LMP1, and our data indicate that LMP1elicits its effect upstream of caspase-3 both in the Fasand in the etoposide model. It is well established thatetoposide triggers apoptosis via mitochondrial perturba-tion with subsequent activation of cytoplasmic caspases,including caspase-3 (Sun et al., 1999), and depending onthe dose etoposide was shown to trigger the release of

FIG. 6. LMP1 regulates the expression of A20 and Bcl-2 in HeLacells. (A) HeLa cells (clone Ht27/8) were incubated in the presence orabsence of tetracycline (tet; 1 �g/ml) for 48 h and cell lysates were thenmonitored by Western blotting for the expression of LMP1 and A20using specific antibodies. The EBV-positive B cell line Raji was in-cluded as a positive control. Data are representative of four indepen-dent experiments. (B) HeLa cells were incubated in the presence orabsence of tetracycline (1 �g/ml) for 48 h and lysates were probed forexpression of LMP1 and Bcl-2, respectively. Raji cells were used as apositive control. Data shown are representative of three independentexperiments.

337STIMULUS-DEPENDENT APOPTOSIS MODULATION OF EBV LMP1

cytochrome c from the mitochondrial intermembranespace through either a direct or an indirect mechanism(Robertson et al., 2000). The potentiating effect of LMP1on etoposide- and Fas-triggered apoptosis may thusinvolve its interactions with factors upstream or at thelevel of the mitochondrion, including members of theBcl-2 family, with subsequent potentiation of downstreameffector molecules such as caspase-3. Our observationthat Bcl-2 is downregulated by LMP1 in HeLa cells, andthe fact that overexpression of Bcl-2 in HeLa cells pro-tects against both Fas- and etoposide-mediated apopto-sis (Mandal et al., 1996; Asoh and Ohta, 1997; Yin andSchimke, 1995; Lock and Stribinskiene, 1996), concurswith this notion. Furthermore, Liu and associates (2002)recently reported that Bcl-2 is significantly downregu-lated by LMP1 expression in NPC cell lines, thus provid-ing further evidence that the effect of LMP1 on Bcl-2expression may be cell lineage specific.

Subversion of apoptotic control is a common phenom-enon in the development of cancer. Viral proteins that arerequired for survival in the host cell during the stages oflatency and lytic replication frequently target the apopto-tic machinery of the cell and may thus contribute totumorigenesis (for review, see Hardwick, 1998; Thomson,2001). LMP1 is such a viral factor which most likelyevolved to protect against untimely apoptosis during thecourse of EBV infection of B cells, an important if notexclusive reservoir of the latent virus. Based on thewell-established anti-apoptotic function of LMP1 in Bcells, it is reasonable to assume a similar function inepithelial cells. However, as discussed above, the cur-rent data show that the effect of LMP1 on epithelial cellapoptosis depends critically on the nature of the stimu-lus. LMP1 is consistently expressed in cancer in situ andin premalignant lesions of the nasopharynx and is there-fore a natural candidate protein to contribute to NPCdevelopment (Pathmanathan et al., 1995). Nevertheless,it is conceivable that a viral protein with multiple pheno-typic properties may contribute differently to the initialprocess of tumor development and the subsequent pro-cess of tumor growth and progression. In this context,the fact that the EBV genome frequently is lost afterlong-term culture of established NPC cell lines indicatesthat EBV infection may be more important in the earlydevelopment of NPC than for the maintenance of contin-uous tumor growth. In addition, in trying to understandthe processes underlying tumor growth, one should con-sider the extensive lymphoid infiltration evidenced inNPCs, since such inflammatory cells may also contributedeath and survival signals, including CD40 ligand (Sbih-Lammali et al., 1999), resulting in a complex in vivosituation with multiple cell–cell and receptor–ligand in-teractions.

Tumor sensitivity to radiation depends, at least in part,on the proper functioning of the intrinsic apoptosis sig-naling pathways of cancer cells (for review, see Zhivo-

tovsky et al., 1999). One may thus speculate that the highsensitivity of NPCs to radiation, the first choice of treat-ment of these tumors, may relate to the enhanced sen-sitivity of LMP1-expressing HeLa cells to etoposide andto Fas ligation. Indeed, the observation that LMP1 poten-tiates apoptosis instigated by these stimuli is in line withour previous observation that LMP1-positive NPCs havea better prognosis than LMP1-negative tumors (Hu et al.,1995). Interestingly, previous studies have demonstratedthat EBV-positive NPC specimens constitutively expressFas (Sbih-Lammali et al., 1999). Moreover, Fas ligandwas recently shown to be upregulated in an EBV-positiveNPC cell line upon irradiation, and these data were takento suggest that Fas–Fas ligand interactions are involvedin the antitumor effect of ionizing radiation in these cells(Abdulkarim et al., 2000). Furthermore, we recently ob-served a negative correlation between Fas ligand andLMP1 expression upon screening of a set of 50 NPCbiopsies (J-Y. Shao, I. Ernberg, P. Biberfeld, T. Heiden,Q-L. Wu, Y-X. Zeng, and L-F. Hu, unpublished observa-tions), which supports the notion of enhanced elimina-tion of Fas ligand-expressing, LMP1-positive cells inNPC. Clearly, further studies are needed to address theputative role of LMP1 in radiation-induced apoptosis andsensitivity of NPC to treatment.

In sum, the present studies demonstrate the effect ofLMP1 on apoptosis in cells of epithelial origin and serveto underscore the marked stimulus-dependent effect ofthis EBV-encoded apoptosis modulator. Elucidating thefunction of LMP1 is important for the understanding ofthe tumorigenic potential of EBV, and further studies incells of epithelial origin are warranted to unravel themolecular mechanism(s) underlying the differential out-come of LMP1 on epithelial cell apoptosis. It is hopedthat a better understanding of the signaling pathwaysthat control survival and death of epithelial cells willallow treatment regimens to be optimized for EBV-asso-ciated malignancies such as nasopharyngeal carci-noma.

MATERIALS AND METHODS

Cell lines and plasmids

HeLa is a cell line derived from a human cervicalcarcinoma. The cells were maintained in Iscove’s modi-fied Dulbecco’s medium (Life Technologies), supple-mented with 10% fetal calf serum (FCS), penicillin, andstreptomycin (Sigma). EBV-positive Raji B cells and leu-kemic Jurkat T cells were maintained in RPMI 1640 me-dium (Life Technologies) supplemented with 10% FCSand antibiotics (Sigma). PJEF3 is a tTA responder con-structed by inserting a hygromycin-resistance gene intothe plasmid pUHD15–1. pJEF6 is a plasmid containingthe LMP1 cDNA (Liebowitz et al., 1992) downstream of atTA responder. The two plasmids are described in Floett-mann et al. (1996) and were the kind gift of Dr. Martin

338 ZHANG ET AL.

Rowe, Cardiff, U.K. pCon A �B luc contains three �B-binding sites upstream of the luciferase gene.

LMP1 transfection

Two rounds of transfection were performed by coincu-bating the cells with plasmid DNA mixed with 4 �g/mllipofectin (Gibco). The transfected cells after the firstround were selected by 500 �g/ml hygromycin B (Gibco)as determined by cytotoxic titration. The tTA activity wasassessed by transiently transfecting the cells with theplasmid pUHC13–3, containing the luciferase genedownstream of a tTA responder. The clone with thehighest luciferase activity in the absence of tetracyclineand the lowest luciferase activity in the presence oftetracycline was chosen for the second round of trans-fection. After transfection with the pJEF6 plasmid andselection with 2 mg/ml geneticin (Gibco), outgrowingcells were tested for tetracycline-regulated LMP1 ex-pression by Western blotting, using specific anti-LMP1monoclonal antibodies (clone S-12). For determination ofNF�B-driving luciferase activity, the pCon A kB luc plas-mid was transfected as above. The cells were culturedwith or without tetracycline (1 �g/ml) (Stratagene) or itsderivative doxycycline (1 �g/ml) (Stratagene) before har-vesting at the indicated time points. Ten microliters ofcell lysate was mixed with 50 �l of the substrate luci-ferrin (Promega) and immediately monitored with aTurner design luminometer. The transfection efficiencywas normalized by assessment of �-Gal activity ex-pressed by the cotransfected plasmid. The relative lightunits are calculated as the ratio of luciferase units andOD420 values obtained in the �-Gal assay.

Apoptosis triggering

The cells were plated in petri dishes and cultured incomplete medium supplemented with 1 �g/ml tetracy-cline for 48 h to remove basal LMP1 expression. Cellswere then cultured for another 24 h with or withouttetracycline. For apoptosis triggering, the medium wassupplemented with 5 ng/ml TNF (Pharmingen) plus 20�g/ml CHX (Sigma), 100 ng/ml anti-Fas antibody (Ig G1isotype; Boehringer), or 20 �g/ml etoposide (Sigma), andincubated for the indicated time points. For some exper-iments, the general caspase inhibitor, zVAD-fmk (10 �M)(Enzyme Systems Products), was added to cells 30 minprior to coincubation with etoposide. Cells were subse-quently harvested and processed for Western blotting,assessment of caspase-3-like activity, or evaluation un-der the fluorescence microscope, as described below.

Assessment of nuclear morphology

Cells were collected and stained with Hoechst 33258(Sigma) to confirm the induction of apoptosis by classicalnuclear alterations (i.e., chromatin condensation). Forthis purpose, cells were mounted on clean glass slides

by cytospin technique, fixed in acetone:methanol (1:1v/v), and stained with Hoechst 33258 (10 �g/ml) in thedark for 10 min. The fluorescent nuclei were viewedunder a Nikon fluorescent microscope, and the percent-age of fragmented nuclei present in at least 500 cellswas determined.

Caspase-3-like enzyme assay

Caspase-3-like activity was estimated in a continuousfluorometric assay as previously described (Fadeel et al.,1998). Briefly, cell lysates and substrate were combinedin a reaction buffer (100 mM HEPES, 10% sucrose, 5 mMdithiothreitol (DTT), 10�6% NP-40, and 0.1% 3-[(3-cholami-dopropyl) dimethylammonio] propane-1-sulphonic acid(CHAPS), pH 7.25) and added in duplicate to a 96-wellplate. The cleavage of the fluorogenic peptide substrateDEVD-AMC (50 �M) (Peptide Institute, Inc., Osaka, Ja-pan) was monitored by AMC liberation in a Fluoroskan IIplate reader (Labsystems, Stockholm, Sweden) using355 nm excitation and 460 nm emission wavelengths.Fluorescence was measured every 70 s during a 30 minperiod and fluorescence units were converted to pico-mole of AMC using a standard curve generated with freeAMC. Data from duplicate samples were then analyzedby linear regression and are displayed as picomole AMCrelease per min. For experiments conducted in the pres-ence or absence of the specific inhibitor of caspase-3-like enzymes, DEVD-CHO (50 nm), real-time recordingsof DEVDase activity (picomole AMC release), are shown.

Western blotting

Total proteins from cell lysates were separated bySDS–polyacrylamide gel electrophoresis and electro-blotted to nitrocellulose membranes. The membraneswere probed with the following primary antibodies: S-12(murine anti-EBV LMP1 mAb); C4 (murine anti-humanactin mAb; Boehringer); C2–10 (murine anti-human PARP;Pharmingen); 8E8 (murine anti-human A20 mAb; a kindgift from Dr. Claudius Vincenz, Ann Arbor, MI); 124 (mu-rine anti-human Bcl-2; DAKO); C20 (goat anti-human 55kDa TNFR1 antibody; Santa Cruz Biotechnology). Theantibodies were diluted in 0.5% nonfat milk in Tris-buff-ered saline (TBS). After washing with 0.1% Tween 20 inTBS, the filter was incubated with horseradish peroxi-dase (HPR)-linked anti-murine immunoglobulin (Ig), ex-cept for the anti-TNFR1 immunoblotting in which a rabbitanti-goat conjugated (DAKO) was used, and the film wasdeveloped with enhanced chemiluminescence (ECL), ac-cording to the manufacturer’s instructions (Amersham).

Flow cytometry

Flow cytometric analysis of the surface expression ofFas was carried out as previously described (Fadeel etal., 1997). Briefly, the monolayer cells were detachedfrom the flasks by trypsinization. Cells were pooled in

339STIMULUS-DEPENDENT APOPTOSIS MODULATION OF EBV LMP1

complete medium, fixed with 1% paraformaldehyde, andwashed in sterile PBS prior to incubation with anti-FasmAb for 1 h at 4°C. Cells then were washed twice withPBS and incubated with FITC-conjugated rabbit anti-murine immunoglobulin (DAKO). Following incubation for30 min at 37°C, cells were washed twice and resus-pended in PBS. Indirect immunofluorescence was mea-sured using a FACScan flow cytometer (Becton–Dickin-son). Fluorescence data were processed using theCellQuest software and are presented as logarithmichistograms. Jurkat cells were employed as a positivecontrol.

ACKNOWLEDGMENTS

Drs. Martin Rowe, University of Cardiff, U.K. and Claudius Vincenz,University of Michigan, MI are acknowledged for the generous dona-tion of plasmids and anti-A20 antibodies, respectively. X.Z. is supportedby the Cancer Research Institute/Concern Foundation, and B.F. is therecipient of a Swedish Society for Medical Research scholarship.

REFERENCES

Abdulkarim, B., Sabri, S., Deutsch, E., Vaganay, S., Marangoni, E.,Vainchenker, W., Bongrand, P., Busson, P., and Bourhis, J. (2000).Radiation-induced expression of functional Fas ligand in EBV-posi-tive human nasopharyngeal carcinoma cells. Int. J. Cancer 86, 229–237.

An, S., and Knox, K. A. (1996). Ligation of CD40 rescues Ramos-Burkittlymphoma B cells from calcium inophore- and antigen receptor-triggered apoptosis by inhibiting activation of the cysteine proteaseCPP32/Yama and cleavage of its substrate PARP. FEBS Lett. 386,115–122.

An, S., Yap, D., and Knox, K. A. (1997). Ligation of CD40 potentiatesFas-mediated activation of the cysteine protease CPP32, cleavage ofits death substrate PARP, and apoptosis in Ramos-Burkitt lymphomaB cells. Cell Immunol. 181, 139–152.

Asoh, S., and Ohta, S. (1997). Bcl-2 completely blocks Fas-mediatedapoptosis in mtDNA depleted HeLa cells. Biochem. Biophys. Res.Commun. 237, 659–662.

Busch, L. K., and Bishop G. A. (1999). The EBV transforming protein,latent membrane protein 1, mimics and cooperates with CD40 sig-naling in B lymphocytes. J. Immunol. 162, 2555–2561.

Devergne, O., McFarland, E. C., Mosialos, G., Izumi, K. M., Ware, C. F.,and Kieff, E. (1998). Role of the TRAF binding site and NF-�B activa-tion in Epstein-Barr virus latent membrane protein 1-induced cellgene expression. J. Virol. 72, 7900–7908.

D’Souza, B., Rowe, M., and Walls, D. (2000). The bfl-1 gene is transcrip-tionally upregulated by the Epstein-Barr virus LMP1, and its expres-sion promotes the survival of a Burkitt’s lymphoma cell line. J. Virol.74, 6652–6658.

Fadeel, B., Åhlin, A., Henter, J.-I., Orrenius, S., and Hampton, M. B.(1998). Involvement of caspases in neutrophil apoptosis: Regulationby reactive oxygen species. Blood 92, 4808–4818.

Fadeel, B., Thorpe, C. J., Yonehara, S., and Chiodi, F. (1997). Anti-IgG1antibodies recognizing the same epitope of Fas/APO-1 mediatedifferent biological effects in vitro. Int. Immunol. 9, 201–209.

Floettmann, J. E., Eliopoulos, A. G., Jones, M., Young, L. S., and Rowe, M.(1998). Epstein-Barr virus latent membrane protein-1 (LMP1) signal-ling is distinct from CD40 and involves physical cooperation of itstwo C-terminus functional regions. Oncogene 17, 2383–2392.

Floettmann, J. E., Ward, K., Rickinson, A. B., and Rowe, M. (1996).Cytostatic effect of Epstein-Barr virus latent membrane protein 1

analysed using tetracycline-regulated expression in B cells line.Virology 223, 29–40.

Foghsgaard, L., and Jaattela, M. (1997). The ability of BHRF1 to inhibitapoptosis is dependent on stimulus and cell type. J. Virol. 71, 7509–7517.

Fries, K. L., Miller, W. E., and Raab-Traub, N. (1996). Epstein-Barr viruslatent membrane protein 1 blocks p53-mediated apoptosis throughthe induction of the A20. J. Virol. 70, 8653–8659.

Hanahan D., and Weinberg, R. A. (2000). The hallmarks of cancer. Cell100, 57–70.

Hardwick, J. M. (1998). Viral interference with apoptosis. Semin. CellDev. Biol. 9, 339–349

Henderson, S., Rowe, M., Gregory, C., Croom-Carter, D., Wang, F.,Longnecker, R., Kieff, E., and Rickinson, A. B. (1991). Induction of bcl-2expression by Epstein-Barr virus latent membrane protein 1 protectsinfected B cells from programmed cell death. Cell 65, 1107–1115.

Henderson, S., Huen, D., Rowe, M., Dawson C., Johnson, G., andRickinson, A. (1993). Epstein-Barr virus-coded BHRF1 protein, a viralhomologue of Bcl-2, protects human B cells from programmed celldeath. Proc. Natl. Acad. Sci. USA 90, 8479–8483.

Hu, L. F., Chen, F., Zhen, Q. F., Zhang, Y. W., Luo, Y., Zheng, X., Winberg,G., Ernberg, I., and Klein, G. (1995). Differences in the growth patternand clinical course of EBV-LMP1 expressing and non-expressingnasopharyngeal carcinomas. Eur. J. Cancer 31, 658–660.

Izumi, K. M., Cahir-McFarland, E., Ting, A. T., Riley, E. A., Seed, B., andKieff, E. D. (1999). The Epstein-Barr virus oncoprotein latent mem-brane protein 1 enganges the tumor necrosis factor receptor-asso-ciated proteins TRADD and receptor-interacting protein (RIP) butdoes not induce apoptosis or require RIP for NF-�B activation. Mol.Cell. Biol. 19, 5759–5767.

Jaattela, M., Mouritzen, H., Ellin, F., and Bastholm, L. (1996). A20 zincfinger protein inhibits TNF and IL-1 signaling. J. Immunol. 156, 1166–1173.

Kawanishi, M. (1997a). Expression of Epstein-Barr virus latent mem-brane protein 1 protects Jurkat T cells from apoptosis induced byserum deprivation. Virology 228, 244–250.

Kawanishi, M. (1997b). Epstein-Barr virus BHRF1 protein protects in-testine 407 epithelial cells from apoptosis induced by tumor necrosisfactor alpha and anti-Fas antibody. J. Virol. 71, 3319–3322.

Kawanishi, M. (2000). The Epstein-Barr virus latent membrane protein1(LMP1) enhances TNF alpha-induced apoptosis of intestine 407epithelial cells: The role of LMP1 C-terminal activation regions 1 and2. Virology 270, 258–266.

Kaye, K. M., Izumi, K. M., and Kieff, E. (1993). Epstein-Barr virus latentmembrane protein 1 is essential for B-lymphocyte growth transfor-mation. Proc. Natl. Acad. Sci. USA 90, 9150–9154.

Krammer, P. H. (2000). CD95’s deadly mission in the immune system.Nature 407, 789–795.

Kroemer, G., and Reed, J. C. (2000). Mitochondrial control of cell death.Nat. Med. 6, 513–519.

Laherty, C. D., Hu, H. M., Opipari, A. W., Wang, F., and Dixit, V. M. (1992).The Epstein-Barr virus LMP1 gene product induces A20 zinc fingerprotein by activating nuclear factor kappa B. J. Biol. Chem. 267,24257–24260.

Liebowitz, D., Mannick, J., Takada, K., and Kieff, E. (1992). Phenotypes ofEpstein-Barr virus LMP1 deletion mutants indicate transmembraneand amino-terminal cytoplasmic domains necessary for effects inB-lymphoma cells. J. Virol. 66, 4612–4616.

Liu, Y., Wang, X., Lo, A. K. F., Wong, Y. C., Cheung, A. L. M., and Tsao,S. W. (2002). Latent membrane protein-1 of Epstein-Barr virus inhibitscell growth and induces sensitivity to cisplatin in nasopharyngealcarcinoma cells. J. Med. Virol. 66, 63–69.

Lock, R. B., and Stribinskiene, L. (1996). Dual modes of death byetoposide in human epithelial tumor cells allows Bcl-2 to inhibitapoptosis without affecting clonogenic survival. Cancer Res. 56,4006–4012.

Lu, J. J., Chen, J. Y., Hsu, T. Y., Yu, W. C., Su, I. J., and Yang, C. S. (1996).

340 ZHANG ET AL.

Induction of apoptosis in epithelial cells by Epstein-Barr virus latentmembrane protein 1. J. Gen. Virol. 77, 1883–1892.

Mandal, M., Maggirwar, S. B., Sharma, N., Kaufmann, S. H., Sun, S. C.,and Kumar, R. (1996). Bcl-2 prevents CD95 (Fas/APO-1)-induceddegradation of lamin B and poly(ADP-ribose) polymerase and re-stores the NF-kappaB signaling pathway. J. Biol. Chem. 271, 30354–30359.

Nicholson, D. W., Ali, A., Thornberry, N. A., Vaillancourt, J. P., Ding, C. K.,Gallant, M., Gareau, Y., Griffin, P. R., Labelle, M., Lazebnik, Y. A.,Munday, N. A., Raju, S. M., Smulson, M. E., Yamin, T. T., Yu, V. L., andMiller, D. K. (1995). Identification and inhibition of the ICE/CED-3protease necessary for mammalian apoptosis. Nature 376, 37–43.

Okan, I., Wang, Y., Chen, F., Hu, L. F., Imreh, S., Klein, G., and Wiman,K. G. (1995). The EBV-encoded LMP1 protein inhibits p53-triggeredapoptosis but not growth arrest. Oncogene 11, 1027–1031.

Opipari, A. W., Hu, H. M., Yabkowitz, R., and Dixit, V. M. (1992). The A20zinc finger protein protects cells from tumor necrosis factor cytotox-icity. J. Biol. Chem. 267, 12424–12427.

Pathmanathan, R., Prasad, U., Sadler, R., Flynn, K., and Raab-Traub, N.(1995). Clonal proliferations of cells infected with Epstein-Barr virusin preinvasive lesions related to nasopharyngeal carcinoma. N. Engl.J. Med. 333, 693–698

Robertson, J. D., Gogvadze, V., Zhivotovsky, B., and Orrenius, S. (2000).Distinct pathways for stimulation of cytochrome c release by etopo-side. J. Biol. Chem. 275, 32438–32443.

Samali, A., Zhivotovsky, B., Jones, D., Nagata, S., and Orrenius, S. (1999).Apoptosis: Cell death defined by caspase activation. Cell DeathDiffer. 6, 459–496.

Sbih-Lammali, F., Clausse, B., Ardila-Osorio, H., Guerry, R., Talbot, M.,Havouis, S., Ferradini, L., Bosq, J., Tursz, T., and Busson, P. (1999).Control of apoptosis in Epstein Barr virus-positive nasopharyngealcarcinoma cells: Opposite effects of CD95 and CD40 stimulation.Cancer Res. 59, 924–930

Scaffidi, C., Fulda, S., Srinivasan, A., Friesen, C., Li, F., Tomaselli, K. J.,Debatin, K. M., Krammer, P. H., and Peter, M. E. (1998). Two CD95(APO-1/Fas) signaling pathways. EMBO J. 17, 1675–1687.

Sheu, L. F., Chen, A., Wei, Y. H., Ho, K. C., Cheng, J. Y., Meng, C. L., andLee, W. H. (1998). Epstein-Barr virus LMP1 modulates the malignantpotential of gastric carcinoma cells involving apoptosis. Am. J. Pathol.152, 63–74.

Sun, X. M., MacFarlane, M., Zhuang, J., Wolf, B. B., Green, D. R., andCohen, G. M. (1999). Distinct caspase cascades are initiated inreceptor-mediated and chemical-induced apoptosis. J. Biol. Chem.274, 5053–5060.

Tarodi, B., Subramanian, T., and Chinnadurai, G. (1994). Epstein-Barrvirus BHRF1 protein protects against cell death induced by DNA-damaging agents and heterologous viral infection. Virology 201,404–407.

Thomson, B. J. (2001). Viruses and apoptosis. Int. J. Exp. Pathol. 82,65–76

Uchida, J., Yasui, T., Takaoka-Shichijo, Y., Muraoka, M., Kulwichit, W.,Raab-Traub, N., and Kikutani, H. (1999). Mimicry of CD40 signals byEpstein-Barr virus LMP1 in B lymphocyte responses. Science 286,300–303.

Wade, M., and Allday, M. J. (2000). Epstein-Barr virus suppresses aG(2)/M checkpoint activated by genotoxins. Mol. Cell. Biol. 20, 1344–1360.

Wang, S., Rowe, M., and Lundgren, E. (1996). Expression of the Epstein-Barr virus transforming protein LMP1 causes a rapid and transientstimulation of the Bcl-2 homolog Mcl-1 levels in B-cell lines. CancerRes. 56, 4610–4613.

Yin, D. X., and Schimke, R. T. (1995). BCL-2 expression delays drug-induced apoptosis but does not increase clonogenic survival afterdrug treatment in HeLa cells. Cancer Res. 55, 4922–4928.

Yin, X. M., Wang, K., Gross, A., Zhao, Y., Zinkel, S., Klocke, B., Roth, K. A.,and Korsmeyer, S. J. (1999). Bid-deficient mice are resistant to Fas-induced hepatocellular apoptosis. Nature 400, 886–891.

Young, L. S., Eliopoulos, A. G., Gallagher, N. J., and Dawson, C. W.(1998). CD40 and epithelial cells: Across the great divide. Immunol.Today 19, 502–506.

Zhivotovsky, B., Joseph, B., and Orrenius, S. (1999). Tumor radiosensi-tivity and apoptosis. Exp. Cell Res. 248, 10–17.

341STIMULUS-DEPENDENT APOPTOSIS MODULATION OF EBV LMP1