We developed an adenovirus vector for transduction of the human CD21 gene (Adv-CD21), the Epstein-Barr virus (EBV)-specific receptor on human B lymphocytes, to overcome the initial barrier of EBV infection in nonprimate mammalian cells. Inoculation of Adv-CDZ1 followed by exposure to recombinant EBV carrying a selectable marker resulted in the successful entry of EBV into three of seven nonprimate mammalian cell lines as evidenced by expression of EBV-determined nuclear antigen (EBNA), The EBV-susceptible cell lines included rat glioma-derived 9L, rat mammary carcinoma derived c-SST-2, and canine kidney derived MDCK. Subsequent selection culture with G418 yielded drug-resistant cell clones, In these cell clones, EBV existed as an episomal form, as evidenced through the Gardella gel technique. Among the known EBV latency-associated gene products, EBV encoded small RNAs, EBNA1 and transcripts from the BamHI-A rightward reading frame (BARF0), and latent membrane protein 2A were expressed in all EBV-infected cell clones, The viral lytic events could be induced in these cell clones by simultaneous treatment with 12-O-tetradecanoylphorbol-13-acetate and n-butyric acid, but they were abortive, and infectious virus was not produced. These results indicate that once the initial barrier for attachment is overcome artificially, EBV can establish a stable infection in some nonprimate mammalian cells, and they raise the possibility that transgenic animals with the human CD21 gene could provide an animal model for EBV infection.