LAK cell cytolysis

VIROLOGY

180,

818-821

(1991)

Tumorigenicity of Adenovirus-Transformed Cells and Their Sensitivity to Tumor Necrosis Factor LYand NK/LAK Cell Cytolysis DAVID J. KENYON, JANEEN DOUGHERTY, AND KAREL RASKA, JR.’ Department of Laboratory of New Jersey-New Received

Medicine and Jersey Medical September

20,

Pathology, School, 1990;

University of Medicine and Newark, New Jersey 07103-2714

accepted

October

Dentistry

3 1, 1990

Sensitivity of a library of cloned adenovirus-transformed rat cell lines of varying tumorigenicity to cytotoxic action of tumor necrosis factor IY (TNFa) was studied and correlated with their sensitivity to NK/LAK cell cytolysis. Our data confirm earlier reports that expression of the El A oncogene of Ad2 or Ad5 is associated with sensitivity of transformed cells to TNFa and also NK/LAK cytotoxicity. Adz-transformed cell line which expresses the E3 early region in addition to the El gene block is resistant to TNFa, but remains sensitive to NK/LAK cells. All cell lines which express the ElA oncogene of highly oncogenic Ad12 are resistant to NK but not LAK cells. Their sensitivity to TNFa, however, varies over a broad range and does not correlate with either their susceptibility to NK/LAK cytolysis or their tumorigenic potential. 0 1991 Academic Press, Inc.

Oncogenic potential of adenovirus (Ad)-transformed cells varies from highly tumorigenrc to nontumorigenic, depending on serotype used for transformation. While cells transformed with Ad1 2 induce tumors with high efficiency, those transformed with Ad2 or Ad5 typically fail to do so (I). The reasons for this difference, however, which have not been satisfactorily elucidated, appear to be affected by a multitude of factors of host resistance. The El A oncogene of Ad 12 but not of Ad2 or Ad5 strikingly reduces the expression of class I MHC antigen in rat cells (2, 3). It was postulated that tumorigenicity of Ad1 Z-transformed cells is a result of evasion of T cell immunity (4). This hypothesis is not supported by the results on eliciting of T cell mediated transplantation immunity by the Ad1 2 and Ad5, which is controlled by their respective ElA oncogenes. Induction of highly specific transplantation immunity is comparable with the two Ad serotypes (5). On the other hand, there is a striking difference of sensitivity of Ad cells which express the Ad2 or Ad5 ElA oncogenes and those expressing the Ad12 ElA gene to natural killer (NK) cells. The former are highly sensitive, while Ad 12-transformants are resistant to this cytolytic mechanism (6, 7). Both cell types, however, are effectively killed by IL-2 activated LAK cells (8). The degree of sensitivity to these two cytotoxrc effecters is not affected by the level of expression of class I MHC antigen; its modulation by interferon does not affect cytolytic sensitivity (8).

1 To whom correspondence addressed.

The mouse cells stably transfected with recombinant plasmids containing Ad2, Ad5, and also Ad12 ElA oncogene have been reported to be highly sensitive to recombinant TNFa (9). There have been several reports which link theTNFa to the lytic cycle of NK/LAK cell activity (70-72). In fact, it was suggested that TNFcr sensitivity may partially or fully account for the fact that Ad ElA gene induces sensitivity to NK cellmediated cytotoxicity (73). To investigate this point further we have examined in vitro sensitivity to TNFcv of a library of Ad-transformed rat cells which are well characterized as to their expression of Ad El genes. These cells which express Ad2, Ad5 and Ad1 2 El A, and El B oncogenes, or the hybrid Ad5/Adl2 El regions vary broadly in theirtumorigenicity, from nontumorigenic in newborn rats to highly tumorigenic in mature syngeneic animals. Eleven well-characterized Ad-transformed cell lines were used. The A2T8, A2rT13 and T2C4 cells were transformed with Ad2 virions (14). The RT, cells were transformed with Ad1 2 virus (8). The A, B, and AB cells were transformed with the recombinant adenoviruses sub370-12ElA, s&370-1 2El B, and s&370-12ElAB which contain Ad12 ElA, ElB, or ElAB genes in the Ad5 background (15). The RFC, EcoC2 and EcoC3 cells were transformed with the isolated and cloned EcoRI-C fragment of the Ad12 DNA (O-16.5 map units), respectively (8). The XhoC cells were transformed with recombinant plasmid, containing the Xhol -C (O-l 5 map units) fragment of Ad5 DNA. The LIS-cells are an established fibroblastic cell line (6). Tumorigenicity of these cells has been tested in newborn, weanling (21 days), and mature (12 weeks) syngeneic

and requests for reprints should be

0042-6822/g i $3.00 Copyright Q 1991 by Academic Press, inc. Ail rights of reproduction in any form reseNeU.

ala

SHORT TABLE TUMORIGENICIW

Tumor-positive Cell line

Haplotype

Newborn

RT-1’ RT-1’ RT-1’ RT-lc RT-1’ RT-1” RT-1” RT-1” RT-lC RT-lC RT-1’ RT-lC

z/5 515 o/5 515 o/5 515 515 515 515 o/5 o/5 o/5

Al AB2 A2T8 AZiT13 Bl EcoC.2 EcoC3 RFCl RT2 T2C4 XhoC LIE-Y

CELLS

RATS

Serotype origin of Ad El genes

recipIentsa Weanling O/5 515 O/5 N.D. N.D. 515 N.D. 515 515 o/5 o/5 o/5

Mature

ElA

ElB

o/5 o/5 N.D. 515 o/5 515 515 o/5 o/5 N.D. o/5 N.D.

12 12 2 2 5 12 12 12 12 2 5 -

5 12 2 2 12 12 12 12 12 2 5 -

a Randomized newborn, weanling (21 days old) and mature (12 weeks old) LIS (RT-1”) or Fischer (RT-1’) rats were injected subcutaneously with 2 X 1 O6 syngeneic cells and monitored for tumor development for 1 year.

rats. Results of tumorigenicity assays are summarized in Table 1. A2T8, T2C4, B4, and XhoC cells are nontumorigenic. The Al, AB2, RFCl,, and RT, cells induce tumors in immature, but not mature syngeneic animals The EcoC2, EcoC3, and A2/T13 cells induce tumors in mature rats. The sensitivity of these cell lines to highly enriched syngeneic NKcell killing was determined next (Fig. 1). It

0

84

XhoC

A2T8

,2C4

ApIT,

819

1

OF ADENOVIRUS-TRANSFORMED IN SYNGENEIC

COMMUNICATIONS

Al

A62

EcoC3

WC,

RT2

FIG. 1. NK cell sensitivity of adenovirus-transformed cells. The cytolytic assays were performed under conditions described earlier (8). The effector cells were prepared from rat spleen cell suspensions by fractionation on mylon wool and discontinuous Percoll gradient. The results show cytolytic effects at the E/T ratio of 200 after 5-hr incubation. The cells which express the Ad2 or Ad5 ElA oncogene are shown in (A); the cells expressing the Ad12 ElA oncogene are shown in (B).

0

84

XhoC

APT3

T2C4Ap/T,3

At

AB2

EcaC3

RFCl

RT2

FIG. 2. Sensrtivity of adenovirus-transformed cells to cytolytic activity of IL-2 stimulated LAK cells. The LAK cells were generated by 4-day incubation of NK cell-enriched fraction of spleen cells with 20 U/ml of recombinant IL-2 and used as effecters against Ad-transformed cells as described by Kenyon and Raska, at effecter/target ratio of 100 in 4-hr cytolytic assays (8). The ETT cells expressing the Ad2 or Ad5 El A genes are shown in (A); the cells whrch express the Ad12 ElA gene are shown in (B).

is apparent that all cells which express the Ad2 or Ad5 El A oncogenes are sensitive to NK cytolytic mechanisms. The T2C4 and A2TT13 cells are least sensitive in this group. All cells which express the Ad12 ElA gene, however, are much more resistant. All of the cells studied are effectively killed by IL-2-stimulated LAK cells (Fig. 2). The relevance of the NK cytolytic mechanism in the lack of tumorigenicity of the cells bearing the Ad2 or Ad5 El early genes was documented by tumorigenicity assays in syngeneic animals with NK cell pool depleted by three injections of anti-asialo GM1 antibody (WACO Chemicals, Waco, TX) at a dose (65 lug of protein) which effectively eradicated NK cytolytic activity without significantly affecting theT cell proliferative response after both PHAstimulation and in the mixed lymphocyte culture. While no tumors developed in unmanipulated animals with 2 X 1 O6 transformed cells, A2T8 cells induced tumors in 5/5 rats, T2C4 in 414 rats, and XhoC cells in 3/5 syngeneic rats treated with the anti-asialo GM1 serum. The sensitivity to TNFa was tested next. Results of typical experiments are shown in Figs. 3 and 4. The general patterns of sensitivity shown were confirmed in a minimum of four additional independent experiments. The results with cell lines which express ElA oncogene of Ad2 or Ad5 (Fig. 3) show that all these lines are relatively sensitive to cytotoxic action of TNFar except the T2C4 cells, which are strikingly resistant, as is the fibroblastic cell line LIS-Y. Results show that TNFcr at a concentration of 10 U/ml causes 40-60% killing of these cells except for T2C4 cell line. This result is in harmony with earlier reports that cells which express the Ad2 or Ad5 genes are sensitive to TNF (9,

820

SHORT

COMMUNICATIONS

BO60. 40. 20-

0.1 TNF@

I

10

DOSE

(U/ml)

100

1000

FIG. 3. ln vitro cytolytic action of recombinant TNFo( on the Ad2 or Ad5 transformed cells. The cytotoxicity assays were performed in 96.well flat-bottom tissue culture plates under conditions described by Chen et a/. (9); after 48.hr incubation with TNFa (Cetus, Emeryville, CA) the cells were washed with PBS, and stained with crystal vrolet and their absorbance quantitated in an ELISA plate reader of 570 nm. The absorbance of wells with cells not treated with TNFor was taken as 100% vrability. The maximum lysis was determined in the wells lysed with 4 M quanidine hydrochloride (0% viability). The viability of the experimental samples was calculated from the formula viability (%) = 100 x (absorbance of sample - absorbance at 0% viabilrty)/(absorbance of control - absorbance of 0% viability). The results show means of a minimum of quadruplicate parallel wells. All cells were free to mycoplasma contamination as determined by fluorescence assays. n , LIS-Y; 0, T2C4; 0, A2T8; a. A21 T13; A, B4; l , XhoC.

13). Their sensitivity, however, does not reach that of prototypic TNF sensitive targets. Under the conditions of our cytolytic assay, 40-600/o killing of L929 cells is seen already with 1 U/ml of TNF. The fact that the T2C4 cells are resistant to TNF under these conditions is not surprising, since among the cell lines studied the T2C4 line is the only one which actively expresses the Ad2 E3 region (16). It has been established earlier that the Ad2 E3 14.7-kDa protein inhibits cellular destruction by TNF (17). The sensitivity to cytotoxic action of TNFa was then tested with the cells which express the ElA oncogene of Ad1 2 and are uniformly resistant to syngeneic NK cells (Fig. 4). The cytolytic sensitivity of these cells varies over a broad range. The Al, EcoC3, and RFCl cells are resistant, with minimum cytotoxicity detected with even 500 U/ml of TNFa. The ABl, EcoC2, and RT2 cells, on the other hand, are more sensitive, with 50% cytotoxicity seen between 10 and 100 U/ml concentrations of TNFcz. This result shows that sensitivity to TNFa among the cells which express the Ad12 ElA oncogene does not correlate with their NWLAK cell cytolytic sensitivity, since all these cell lines are resistant to NK cells, but sensitve to IL+stimulated LAK

cells. The level of sensitivity of these cells to TNFa also does not correlate with their tumorigenicity. All cells which express the entire Ad1 2 El region are highly tumorigenic. The Ad1 2 El A expression in all these cell lines also is high and comparable (not shown). Yet, there are striking differences in sensitivity to TNFa. Interestingly, the least tumorigenic Al cells, which express the El A of Ad1 2 and El B of Ad5 are highly resistant Many cultured cells, both normal and transformed are resistant to TNF (18, 19), although some such cells express receptors for TNF (20). Sensitivity to TNFa also does not correlate with the number or affinity of TNF receptors (19). Our studies confirm that expression of Ad2 or Ad5 El A oncogene induces sensitivity of transformed cells to cytotoxic action of TNFa, which can be suppressed by simultaneous expression of the Ad E3 region in the transformed cells (17). Earlier studies have established that both 13 and 12 S ElA mRNA are each alone sufficient for induction of sensitivity of murine transformed cells to both NK and TNF cytolysis. A threshold level of their expression which exceeds that for transformation is clearly required (13). The very recent analysis (21) suggested that expression of the highly conserved region 1 (CRl) within the El A sequence was required for the induction of TNFa sensitivity. It did not reflect, however, requirement for either transforming or transcriptional repression activity. The transformation defective CR2 deletion mutants also induced TNFcv sensitivity. Our studies show that while expression of the Ad2 E3 region in transformed cells is associated with high degree of resistance to TNFa, these cells remain relatively sensitive to NK cytolysis. The expression of Ad1 2 ElA oncogene, which fails to induce sensitivity to NK

0’

0.1 TNFa

1

IO

DOSE

(U/ml)

100

1000

FIG. 4. Sensitivity of cells expressing Ad1 2 ElA oncogene to TNFa. The cytotoxicity assays were performed as described in Fig. 3. n , LIS-Y; 0, EcoC2; 0, EcoC3; A, RFCl; A, Al; 0, AB2; 0, RT,.

SHORT

COMMUNICATIONS

cells (6, 7), is also not always associated with induction of TNFa sensitivity. Sensitivity to TNFa of rat Ad-transformed cells thus does not correlate with either their in vilro sensitivity to syngeneic NK cells, their sensitivity to IL-2-stimulated LAK cells, or their in viva tumorigenic potential. ACKNOWLEDGMENT This work has been tional Cancer Institute.

supported

by Grant

CA-21

196 from

the Na-

REFERENCES 1. HUEBNER, R. J.. ln “Perspectives in Virology” (M. Pollard, Ed.), Vol. 5, pp. 147-166. Academic Press, New York, 1967. 2. SCHRIER, P. I., BERNARDS, R., VAESSEN, T. I., HOUWELING, A., and VAN DER EB, A. J., Nature (London) 305, 771-775 (1983). 3. MELLOW, G. H., FOHRING, B., DOUGHERP/, J., GALLIMORE, P. H., and RASKA, K., JR., Vkology 134, 460-465 (1985). 4. BERNARD% R., SCHRIER, P. I., HOUWELING. A., Bos, J. L., VAN DER EB, A. J., ZIJLSTRA, M., and MELIEF, C. J. M., Nature (London) 305, 776-779 (1983). 5. SAWADA, Y., URBANELLI, D., RASKOVA, J., SHENK, T. E., and RASKA, K., JR., J. fxp. Med. 163, 563-572 (1986). 6. RASKA, K., JR., and GALLIMORE, P. H., virology 123, 8-l 8 (1982). 7. SAWADA, Y., FOHRING, B., SHENK, T., and RASKA, J., JR., Virology 147, 413-421 (1985). 8. KENYON, D. J., and RASKA, K., JR., Virology 155, 644-654 (1986).

821

9. CHEN, M. J., HOLSKIN, B., STICKLER, J., GORNIAK, J., CLARK, M. A., JOHNSON, P. J., MITCHO, M., and SHALLOWAY, D., Nature (London) 330, 58 l-583 (1983). 10. OLD, L. J., Science 230, 630-632 (1985). 11. ORTALDO, J. R.. MASON, L. H., MATHIESON, B. J., LIANG, S. M., FLICK, D. A., and HERBERMAN, R. B., Nature (London) 321, 700-702 (1986). 12. CHANG, A. S. F., SCUDERI, P., GRIMES, N. J., and HERSH, E. M., /. Immunol. 142, 2133-2139 (1989). 73. COOK, J. L., MAY, D. L., WILSON, B. A., HOLSKIN, B., CHEN, M. Y., SHALLOWAY, D., and WALKER, T. A., J. Immunol. 142, 452?4534 (1989). 14. GALLIMORE, P. H., BYRD, J., and GRAND, J. A., In “Viruses and Cancer” (W. J. Rigby and N. M. Wilkie, Eds.), pp. 1255172. Cambridge Univ. Press, Cambridge, 1984. 15. SAWADA, Y., RASKA, K., JR., and SHENK, T., lhroiogy 166, 281284 (1988). 16. KVIST, S., OSTBERG, L., PERSSON, H., PHILIPSON, L., and PETERSON, P. A., Proc. Natl. Acad. Sci. USA 75, 5674-5678 (1978). 17. GOODING, L. R., ELMORE, L. W., TOFFELSEN, A. E., BRADY, H. A., AND WOLD, W. S. M., Cell53, 341-346 (1988). 18. SUGARMAN, B. J., AGGARWAL, B. B., HASS, P. E., FIGARI. I. S., PALLADINO, M. A., and SHEPARD, H. M., Science 230, 943-945 (1985). 19. WILLIAMSON, B. D., CARSWELL, E. A., RUBIN, B. Y., PRENDERGAST, J. S.. and OLD, L. J., Proc. Nat. Acad. Sci USA 80, 5397-6402 (1983). 20. TSUJIMOTO, M., YIN, Y. K., and VILCHEK, J., Proc. Nat/. Acad. SC;. USA 182, 7626-7631 (1985). 21. AMES, R. S., HOLSKIN, B., MITCHO, M., SHALLOWAY, D., and CHEN, M.-J., 1. I/ire/. 64, 41 15-4122 (1990).