Human cytomegalovirus inhibits human immunodeficiency virus replication in cells productively infected by both viruses.

We have been studying the role of human cytomegalovirus (HCMV) as a potential cofactor in human immunodeficiency virus (HIV)-related disease. The clinical relevance of HCMV is highlighted by the fact that it is a principal viral pathogen in patients with AIDS and is known to infect the same cells as HIV. In this study, we focused on the molecular interactions between HIV and HCMV in human fibroblasts and in the human glioblastoma/astrocytoma-derived cell line U373 MG, cells which can be productively infected by both viruses. Because these cells are CD4-, we used HIV pseudotyped with a murine amphotropic retrovirus as described previously (D. H. Spector, E. Wade, D. A. Wright, V. Koval, C. Clark, D. Jaquish, and S. A. Spector, J. Virol. 64:2298-2308, 1990). Initial studies showed that when cells were preinfected with HIV (Ampho-1B) for 5 days and then superinfected with HCMV, HIV antigen production dropped significantly in the coinfected cells but continued to rise in cells infected with HIV (Ampho-1B) alone. HCMV production, however, was unaffected by the presence of HIV. Further analysis showed that HIV steady-state RNA levels and gag and env protein production were also inhibited in the presence of HCMV. The transcriptional inhibition of HIV was particularly surprising in view of the previous results of several other laboratories as well as our own that HCMV infection stimulates HIV long terminal repeat-chloramphenicol acetyltransferase (LTR-CAT) expression in transient expression assays. To investigate this further, we transfected the HIV LTR-CAT construct into either uninfected cells or cells which had been preinfected with HIV. The cells were infected with HCMV 24 h posttransfection and assayed for CAT gene expression at 48 h after HCMV infection. Although there was some stimulation of the LTR-CAT in cells that were dually infected by HIV and HCMV, it was 16-fold less than that in the cells infected only with HCMV. This suggests that in the presence of the HIV infection, the stimulation of the HIV LTR-CAT gene by HCMV is significantly reduced. Experiments with UV-irradiated HCMV and the HCMV DNA polymerase inhibitor ganciclovir showed that HCMV transcription is necessary for the reduction in HIV production to occur; however, replication of the HCMV genome or any events which take place after DNA replication are not necessary. These results, coupled with the observation that inhibition is usually first seen between 8 and 24 h after HCMV infection, suggest that an HCMV early protein is involved in repression of HIV.