Tumor necrosis factor mediated cytolysis requires the adenovirus E1A protein but not the transformed phenotype

Vim Research, 15 (1989) 231-236 Elsevier

231

VIRUS 00556

Tumor necrosis factor mediated cytolysis requires the adenovirus ElA protein but not the transformed phenotype M. Rodrigues, Department

of Microbiology,

P. Dion, S. Sircar and J.M. Weber

Faculty of Medicine Centre Hospitalier Canada JIH 5N4

Universitaire,

Sherbrooke,

Quebec,

(Accepted 10 November 1989)

Adenovirus transformed cells are susceptible to lysis by human recombinant tumor necrosis factor (TNF). This susceptibility correlates with the presence of Ela in these cells. A flat revertant cell line which expresses a biologically functional Ela but not the transformed phenotype was nevertheless susceptible to TNF. However, flat revertants retransformed by kuacytidine, without concomitant reactivation of Ela, were resistant to TNF-a. This result suggests TNF susceptibility is not transformation but Ela dependent. To study the mechanism of cytolysis in these cell lines, we examined the possibility that changes in the transcription of Ela were brought about by TNF, as it was reported in the case of a c-myc transformed cell line. The results showed that TNF did not affect either Ela or c-myc transcription in our cells during the development of the cytotoxic response. Tumor necrosis factor; Ela and c-myc transcription; Cytotoxicity

Adenovirus transformed

cell;

Tumor necrosis factor (TNF) is secreted by macrophages in response to inflammatory stimuli. TNF has a wide variety of biological effects including the selective killing of some tumor cells (Beutler and Cerami, 1988). The mechanism by which these cells are killed by TNF is still obscure. TNF has been shown to be Correspondence to: J.M. Weber, Department of Microbiology, Faculty of Medicine, Department Microbiology, Centre Hospitalier Universitaire, Sherbrooke, Quebec, Canada J 1H 5N4.

0168-1702/90/$03.50

0 1990 Elsevier Science Publishers B.V. (Biomedical Division)

of

232

cytotoxic to cells transformed by the adenovirus type 2 Ela proteins (Chenetal., 1987) however, if these cells express the adenovirus E3 region (14.7 kDa protein), they are protected from the TNF cytolysis (Gooding et al., 1988). In addition to its antiviral activity (Mestan et al., 1986; Wong and Goeddel, 1986) TNF can induce changes in gene expression in transformed cells including cellular oncogenes (Schutze et al., 1988). In the present communication we asked if Ela expression or transformation per se was the decisive factor rendering cells susceptible to lysis by TNF. We had isolated flat variants resistant to the antiproliferative effects of methylglyoxal-bis(guanylhydrazone) (MGBG) from an adenovirus type 2 transformed rat embryo cell line (F4) (Flint, 1981; Gallimore et al., 1977). These variants were named G2, G4, G5 and have previously described (Sircar et al., 1987; Sircar and Weber, 1987; Rodrigues et al., 1987; Sircar et al., 1988). They had the following properties in common: (i) flat cell morphology, (ii) lack of tumor formation in nude mice, and (iii) resistance to the drug, MGBG. In the course of these studies, subclones have been isolated from G2 and G4 and were named G2a and G4F. Both these cell lines still maintained the transformation revertant phenotype. The G4F cell line, however, had lost the integrated El genes. The Dl cell line is a subclone of G4 transfected with an Ad2-El bearing plasmid (pLB206) and has similar properties to G2 in that it expresses Ela proteins but is not tumorigenic. On the other hand, G4NF has an FClike transformed phenotype and is an MGBG-resistant spontaneous rerevertant of G4. Some of these flat variants and subclones were retransformed by the drug 5azacytidine. Thus we had a series of cell line with different phenotypes derived from a single adenovirus type 2 transformed parent. These variant cell lines were subjected to treatment with recombinant tumor necrosis factor (rTNF-a; Amersham). All cell lines expressing Ela were sensitive to TNF cytolysis (Table 1). These results confirmed a previous report showing a correlation between Ela expression and susceptibility to TNF-mediated cell lysis (Chen et al., 1987). Furthermore these results establish three points concerning TNF susceptibility: (i) the transformed or tumorigenic phenotype was not required as long as Ela was expressed (note Dl, G2 versus G2a), (ii) a low level of Ela expression may be sufficient for tumorigenicity but may not be sufficient for TNF-susceptibility (see GSaza) and (iii) tumorigenicity in absence of the Ela proteins was insufficient to bring about cytolysis (see G2a aza, G4F aza, FR3T3 aza). Although there appears to be a good correlation between Ela and TNF lysis, in view of the nontumorigenic nature of the G2 cell line it may be argued that the Ela proteins were not biologically functional in this cell line. This would cast doubt on the significance of the correlation. To test the functional integrity of Ela, several of these cell lines were infected with an Ela-deletion mutant (d1312) and tested for the ability to replicate the d1312 viral DNA. Cells were suspended, after 72 h post-infection, and gently lysed in the wells of the agarose gel followed by separation of extrachromosomal DNA from the bulk of the cellular DNA by high-voltage electrophoresis. Fig. 1 shows a wild-type level of complementation for G2 witnessing functional Ela proteins. This result confirms a previous report that showed the

233 TABLE 1 TNF sensitivity of transformed and revertant cell lines: correlation with ElA expression and tumorigenicity Cell line

F4 G4NF Dl G2 G2a G5 G4F G2aAZAd G5 AZAd G4F AZA * FR3T3 FR3T3 AZA ’

ElA Gene

Protein a

+ + + + + + _ + + -

+ + + +

-

-

+/-

Tumor in nude mice b

Sensitivity ’

+ + -

94 91 96 97 15 17 14 8 14 11 4 3

+ + + +

(W)

a ElA was tested by immunoprecipitation as previously described (Sircar et al., 1987) b Tumorigenicity in nude mice was tested as previously described (Sircar et al., 1987) ’ In vitro cytotoxicity assay was as described by Chen et al. (1987). Cells in 0.1 ml of medium were seeded (20,000 cells/well) 96-well linbro plates. Twenty-four hours later various amounts of rTNF (O-32 units/ml) was added and incubated for another 48 h at 37 o C. After incubation, the remaining, adherent viable cells were stained with 0.5% crystal violet and the absorbance at 540 nm measured with a microelisa autoreader (Bio-TEK microplate reader). The average absorbance of triplicate wells without TNF was taken as 100% viable. Sensitivity indicates the percentage of cells killed by 10 units of TNF (average of several experiments). * Cells treated with 5-azacytidine (Weber et al., unpublished results).

G2 cells to be resistant to retransformation by El bearing plasmids but not by other oncogenes (12). This shows that the introduction of new Ela genes fails to transform the cells and therefore implies that the resident Ela genes encode functionally normal Ela proteins. Our findings suggest that transformation per se does not confer susceptibility to TNF cytolysis; however, the expression of the oncogene Ela is required. The mechanisms by which retransformation by the drug 5azacytidine mediates resistance to cytolysis by rTNF-a are unknown. One possible explanation could be an induced overexpression of as yet unidentified oncogenes which may be potent in their induction of rTNF-cu resistance. In support of this explanation Hudziak et al. (1988) have identified an oncogene that was overexpressed and induced resistance to rTNF-ar in the transformed NIH3T3 cells. In search of a mechanism for the TNF-induced lysis of Ela expressing cells, we considered the following facts: (i) Ela is present on the external surface of the cell membrane (Bellgrau et al., 1988), (ii) Ela is a myc-like oncogene (Zerler et al., 1986; Figge et al., 1988), (iii) studies have shown that growth regulatory genes like myc (Schutze et al., 1988; Kronke et al., 1987; Yarden and Kim&i, 1986; &stein and Baglioni, 1988) are at the receiving end of the post-receptor signaling pathway of TNF and (iv) overexpression of Ela can be toxic to some cells (Jochemsen et al,, 1987). It is plausible therefore that TNF may cause excessive transcription of Ela

234

G4f

I

G5 G2a G2 DI

G4f G5 G2aG2

F4

DI

72h

1hP.i. 1

I

F4 V

P.i. I

Fig. 1. Complementation of d1312 replication by Ela positive cells. Cells were infected with 20 PFU/cell of d1312 (Jones and Shenk, 1979; Shenk et al., 1979) and 1 h or 72 h later 200,000 cells were loaded into each well of a horizontal 0.75% agarose lysing gel and electrophoresed as described by Gardella et al. (1984). The gel was blotted onto a nylon membrane and hybridized with adenovirus type 2 DNA. The arrow indicates the position of viral DNA. Lane V contained 32 ng of viral DNA.

resulting in toxicity. To test this hypothesis, we measured the effect of rTNF on the transcription of Ela and c-myc in F4 cells which were highly sensitive to its toxic action. Cells (2OO,OOO/well)were seeded in 24-well linbro plates and 24 h later rTNF (1 pg/ml) was added. After different lengths of tune the cells were trypsinized, counted and resuspended in phosphate buffer and spotted onto nitrocellulose filter using a dot-blot filter apparatus (Schleicher and Schuell, Keene, N.H.) as described (Paeratakul et al., 1988). The Ela mRNA was detected by hybridization with a cDNA probe representing the 13s Ela message. The results showed a slight increase in RNA signal with increased incubation tune (compare 0 h with 16 h in Fig. 2). This, however, was merely due to cell division as shown by the equivalent RNA signal at 16 h in the absence of rTNF. Similar results were obtained with c-myc (data not shown) and a control gene, beta-actin (Fig. 2, right panel). We conclude that rTNF does not alter the transcription of Ela or c-myc in these sensitive F4 cells and this therefore is not the cause for the cytotoxicity. Recently, in TNF-sensitized cells such as murine L929 cells (S&mid et al., 1987), human breast carcinoma cells (Dealtry et al., 1987) and adenovirus-transformed rat fibroblasts (Laster et al., 1988), TNF induced an apoptic form of cell death, i.e. nuclear disintegration and DNA fragmentation. These results suggest that the

235

Actin

ElA 1

I

F4

FR3T3

F4(RNasd

0

0,s TNF+

2 4 8 16

TNF-

16

Fig. 2. Effect of rTNF on the transcription of Ela; 50,000 cells, treated with 1 pg/ml rTNF for the indicated times, were dot-blotted as described (Paeratakul et al., 1988) and hybridized with an Ela probe (left panel) or a /3-a&n probe (right panel). Controls consisted of Ela-negative FR3T3 cells and F4 cells digested with 1 mg/ml RNase (DNase free, Sigma).

mechanism of TNF-induced cytolysis of these sensitive cells is via activation of their endogenous self-destruct programs. It is possible that Ela plays a role in this putative program. Resistant cells can be made sensitive to TNF by treatment with inhibitors of transcription or translation. It has been suggested that these cells have an endogenous protective mechanism that is dependent upon de novo protein synthesis (Ostrove et al., 1979; Kull et al., 1981; Hahn et al., 1985). Thus the synthesis of these essential proteins could also be repressed in Ela-transformed cells, Ela having been demonstrated to be a transcription repressor as well as a transcription activator (Berk, 1988; Nevins et al., 1988). Furthermore, Fletcher et al. (1987) have demonstrated that another oncogene, pp60src, can overcome the resistance of a cell to the cytolytic action of TNF by inhibiting the formation of gap junctions. In summary, we find that Ela plays a critical role in inducing susceptibility to TNF cytolysis, even in the absence of the transformation phenotype. Furthermore, the transcription of Ela, c-myc and actin were not affected by exposure to TNF. Acknowledgements This work was supported by a grant from the National Cancer Institute of Canada. J.M.W. was a Senior Research Scientist of the N.C.I.C., P.D. was the recipient of a studentship via our M.R.C. Biotechnology Training Center. We thank Joseph Horvath for helpful discussions and Jocelyn Raymond for assistance.

236

References BelIgrau, D., Walker, T.A. and Cook, J.L. (1988) J. Virol. 62, 1513-1519. Berk, A.A. (1988) Annu. Rev. Genet. 20, 45-79. Beutler, B. and Cerami, A. (1988) Ann. Rev. B&hem. 57, 505-518. Chen, M-J., Holskin, B., StrickIer, J., Gorniak, J., Clark, M.A., Johnson, P.J., Mitcho, M. and Shalloway, D. (1987) Nature (London) 330, 581-583. Deahry, G.B., Naylor, M.S., Fiers, W. and Balkwill, F.R. (1987) Eur. J. Immunol. 17, 689-693. Figge, J., Webster, T., Smith, T.F. and Paucha, E. (1988) J. Virol. 62, 1814-1818. Fletcher, W.H., Shiu, W.W., Ishida, T.A., Haviland, D.L. and Ware, C.F. (1987) J. Immunol. 139, 956-962. Flint, S.J. (1981) In: J. Tooze (Ed.), DNA Tumor Viruses, pp. 383-442. Cold Spring Harbor Laboratory; Cold Spring Harbor, New York. GaBimore, P.H., McDougaI, J.K. and Chen, L.B. (1977) Cell 10, 669-678. GardeIIa, T., Medveczky, P., Sairenji, T. and Mulder, C. (1984) J. Virol. 50, 248-254. Gooding, L.R., Ehnore, L.W., ToIIefson, A.E., Brady, H.A. and Wold, W.S.M. (1988) Cell 53, 341-346. Hahn, T., Toker, L., BudiIovsky, S., Aderka, D., Eshhar, Z. and WaIlach, D. (1985) Proc. Natl. Acad. Sci. USA 82, 3814-3818. Hudziak, R.M., Lewis, G.D., Shalaby, M.R., Eessalu, T.E., Aggarwal, B.B., Ulhich, A. and Shepard, H.M. (1988) Proc. Natl. Acad. Sci. USA 85, 5102-5106. Jochemsen, A.G., Peltenburg, L.T.C., Te Pas, M.F.W., De Wit, C.M., Bos, J.L. and van der Eb, A.J. (1987) Eur. Mol. Biol. Organ. J. 6, 3399-3405. Jones, N. and Shenk, T. (1979) Cell 17, 683-689. Kirstein, M. and Baghoni, C. (1988) J. Cell. Physiol. 134, 479-484. Kronke, M., Schluter, C. and Pfizenmaier, K. (1987) Proc. Natl. Acad. Sci. USA 84, 469-473. Kull, F.C. Jr. and Cuatrecasas, P. (1981) Cancer Res. 41, 4885-4890. Laster, S.M., Wood, J.G. and Gooding, L.R. (1988) J. Immunol. 141, 2629-2634. Mestan, J., Digel, W., Mittnaacht, S., Hillen, H., Blohm, D., Moller, A., Jacobsen, H. and Kirchner, H. (1986) Nature (London) 323, 816-819. Nevins, J.R., Raychaudhuri, P., Yee, A.S., Rooney, R.J., Kovesdi, I. and Reichel, R. (1988) B&hem. Cell Biol. 66, 578-583. Ostrove, J.M. and Gifford, G.E. (1979) Proc. Sot. Exp. Biol. Med. 160, 354-358. Paeratakul, U., De Stasio, P.R. and Taylor, M.W. (1988) J. Virol. 62, 1132-1135. Rodrigues, M.T., PaIkonyay, L., Sircar, S., FIeurent, J., Dessureault, J. and Weber, J.M. (1987) B&hem. Biophys. Res. Commun. 147, 675-681. S&mid, D.S., Homung, R., McGrath, K.M., Paul, N. and Ruddle, N.H. (1987) Lymphokine Res. 6, 195-202. Schutze, S., Scheurich, P., Schluter, C., Ucer, U., Pfizenmaier, K. and Kronke, M. (1988) J. Immunol. 140,3000-3005. Shenk, T., Jones, N., Colby, W. and Fowlkes, D. (1979) Cold Spring Harbor Symp. Quant. Biol. 44, 367-375. Sircar, S. and Weber, J. (1988) J. Cell Physiol. 134, 467-472. Sircar, S., Palkonyay, L., Rodrigues, M., Allaire, S., Horvath, J., Thirion, J-P. and Weber, J.M. (1987) Cancer Res. 46,1339-1343. Sircar, S., Rodrigues, M. and Weber, J.M. (1988) Oncogene 3, 725-728. Wong, G.H.W. and Goeddel, D.V. (1986) Nature (London) 323, 819-822. Yarden: E. and Kim&i, A. (1986) Science 234,1419-1421. Zerler, B., Moran, B., Maruyama, K., Mooman, J., Grodzicker, T. and Ruley, H.E. (1986) Mol. Cell. Biol. 6, 887-899. (Received

15 August

1989; revision

received

6 October

1989)