The E1B-attenuated adenovirus, dl1520, causes cyto-lysis of cisplatin-resistant ovarian tumour cell lines both in vitro and in vivo, suggesting that dl1520 may be a useful agent in the treatment of ovarian cancer in which cisplatin resistance has developed. 1 However, just as acquired resistance to chemotherapy can occur, we report another potential limitation of oncolytic adenoviruses, the development of acquired cellular resistance to cytolysis. To establish viral-resistant cell lines, human ovarian A2780/cp70 cells, known to be permissive for viral replication and cytolysis, 2 were infected with dl1520 at a multiplicity of infection (MOI) of 50 plaque forming units (PFU)/cell. After three successive infections, a surviving colony of cells was isolated and expanded as the line CpVR. To detect the presence of intracellular virus particles, CpVR cells were assayed for expression of the late phase hexon protein by immunofluorescence and FACS analysis and compared to the parental cell line A2780/cp70. Table 1 shows that CpVR expresses adenoviral hexon protein, whereas the parental line A2780/cp70 does not. Thus, CpVR showed hexon expression after growth for more than 10 weeks following virus infection. This could either indicate that the gene for hexon protein expression has been integrated into the host cell genome or that it is present in an episomal form. Alternatively, it could indicate that live virus is present in these cells and that resistance to lysis has developed. To determine if live virus was present, CpVR cells were freeze–thawed and the resulting cell extract assayed for PFU on HEK293 cells. This demonstrated that resistant cells can maintain 60PFU/cell of virus and still replicate normally with an unchanged doubling time compared to the parental cell. As well as infecting HEK293 cells, virus harvested from CpVR was able to infect, replicate and induce cytolysis in the permissive parental line A2780/cp70. To determine if CpVR cells are resistant to the cytolytic effect of wild-type Ad2, we further infected CpVR cells and the parental line A2780/cp70 with wild-type Ad2 adenovirus at an MOI of 10 PFU/cell. In the parental line A2780/cp70, 100% CPE had occurred at 72 hours, whereas no CPE had occurred in CpVR, indicating that CpVR was resistant to the cytolytic effect of Ad2 virus, as well as dl1520 (Table 1). To check that resistance to Ad2-induced cytolysis was not due to inability of the CpVR cells to be infected with adenovirus, infectivity assays were done using a nonreplicating E1-deleted adenovirus with a lacZ reporter construct under control of the CMV promoter Ad5-CMVlacZ. 3 Infectivity of the parental cell line A2780/cp70 was 22% and CpVR 14%(Table 1). These results indicate that the acquired resistance in CpVR is not due to resistance to adenoviral cell entry by loss of adenoviral receptors. Resistance is also not due to loss of expression of the early or late adenoviral genes, since expression of the early gene E1A was detected by Western blotting as well as expression of the late adenoviral hexon protein. Together, these data suggest that there is a block at the lytic stage in these cells leading to the acquisition of resistance to oncolysis by adenovirus in these initially sensitive cells. We have already reported that dl1520 is able to replicate and lyse A2780cp70 that have nonfunctional p53, 4 but not the parental line A2780, which express wild-type p53. Therefore, one possible cellular change that could lead to inhibition of replication of an E1B-attenuated virus would be restoration of p53 function in the CpVR cell line. However, CpVR cells do not cell cycle arrest at G1 phase following exposure to ionizing radiation and cannot activate …