Altered control of the rat cell cycle induced by adenovirus requires expression of transformation region ElA, but not of E1B, E2A, E2B, or late genes. We show here that neither E3 nor E4 is required, so the effect results directly from an ElA product. Mutants with defects in the 289-amino-acid (aa) ElA product had little or no effect on the rat cell cycle even at 1,000 IU The adenovirus transformation gene ElA has distant sequence homology with the retrovirus oncogenes v-myc and v-myb and can replace myc in transformation of primary cells (24,26). ElA induces expression of other viral early genes (2,12,27), and at least two cell genes, thymidine kinase and the 70,000-molecular-weight (70K) heat shock protein (1,3,8,11,21). The function of ElA in transformation appears to be "immortalization" of the cell and reduction of requirements for serum growth factors. The ability of adenovirus to induce a growth cycle in rodent cells arrested in the Gi phase by deprivation of serum or reduction of Ca2+ concentration, and to alter cycle progression in growing cells (1,5,8,18), may be related to the ElA transformation functions. Experiments with mutant viruses showed that alteration of control of the rat cell growth cycle requires expression of ElA, but not of E1B, E2A, E2B, or late genes (1,5,8,19). However, because ElA induces all of the other early regions, it remained possible that the cell cycle effects were actually due to E3 or E4, under the control of ElA. In this paper, we show that expression of human adenovirus type 5 (AdS) regions E3 and E4 is not necessary for alteration of rat cell cycle progression, which must therefore be a direct effect of ElA. Experiments with six mutants having different defects in the ElA 289-amino-acid (aa) product showed complete or almost complete defects in alteration of cell cycle progression even at 1,000 IU per cell. In this respect, these mutants lack the "multiplicity-dependent leakiness" reported for ElA mutants in human cells (20,27 ElA function in rat cells. The leakiness of some ElA mutants at high multiplicity in rat cells appears to be due to residual function of the mutant proteins. Additional leak in HeLa cells may be due to a cellular ElA-like function.
MATERIALS AND METHODSViruses and cells. Ad2/5 pm975 (17) was a generous gift from A. J. Berk, University of California, Los Angeles; AdS hrA (28) was from D. Solnick, Yale University, New Haven, Conn.; and inSOO (6) was from N. C. Jones, Purdue University, West Lafayette, Ind. The origins of other AdS ElA mutants have been described previously (3). Locations of the ElA mutations are shown in relation to the early El mRNA species in Fig. 1. ElA mutants were grown in 293 cells and titrated by methods described previously (4,5,18,19), and with the exception of pm975 had a ratio of titer on 293 cells to that on HeLa cells of 103 to 104.Mutant d1808, deleted from between 91.4 and 92.0 to between 97.2 and 98.4 map units (m.u.) (7), and thus defective in subregions 1 or 2 through 7 of early region E4, was kindly supplied by G...