In response to DNA damage, p53 protein accumulates in the cell nucleus causing cells to undergo DNA repair or apoptosis, programmed cell death. Reintroduction of wild-type p53 into tumors with null or mutant p53 offers a novel strategy for controlling tumor growth, by inducing apoptotic death in neoplastic cells. The efficacy of a replication-deficient p53 adenovirus construct was tested against three human breast cancer cell lines expressing mutant p53, MDA-MB-231, -468, and -435. 231 and 468 cells were both highly transduced at a multiplicity of infection of 10. By contrast, 435 cells were rarely transduced. p53 adenovirus-mediated gene therapy was highly effective against 231 and 468 tumor xenografts in nude mice. At a total dose of 2.2 x 10(9) cellular infectious units (CIU), inhibition of 231 tumor growth was 86% (P < or = .01). Thirty-seven percent of that growth inhibition was due to p53, while 49% was adenovirus-specific. Inhibition of 468 tumor growth was 74% (P < or = .001). Forty-five percent of that inhibition was p53-specific, while 28% was adenovirus-specific. The ED50 values for 231 tumors and 468 tumor growth inhibition were 3 x 10(8) CIU and 2 x 10(8) CIU, respectively. Injection of p53 Ad into 231 or 468 tumors induced apoptosis. By contrast, growth inhibition in 435 tumors treated with p53 adenovirus was not significant, probably due to low adenovirus transduction. 231 and 435 cells both expressed high levels of alpha v, beta 1, beta 3, and beta 5 integrin subunits, ruling out lack of the appropriate integrins as the reason for the low infection rate in 435 cells. Our results demonstrate the ability of wild-type p53 to curtail cancerous cell growth in vivo in tumors expressing mutant p53. The ability of beta-gal Ad to infect tumor cells in vitro was generally predictive of in vivo p53 Ad efficacy.