Zvi Grossman^1,2 and William E. Paul²

1: Department of Physiology and Pharmacology, Sackler Faculty of Medicine, Tel-Aviv University, Tel Aviv, Israel, and 2: Laboratory of Immunology, NIAID, National Institutes of Health, Bethesda, MD 20892, USA

We have proposed that HIV replication occurs in multiple local bursts, associated with chronic immune activation in response to antigens and inflammatory signals, and that antiretroviral treatment substantially reduces the size and frequency of such bursts but does not generally prevent the steady-state regeneration of infected memory cells. In a system in which HIV transmission is predominantly local, there are opportunities for HIV to move among T cell clones of different specificity due to patterns of cross-reactivity among co-stimulated T cells. We envision disease progression during the asymptomatic phase as a progressive penetration of HIV into the memory component of an increasing number of clones. This is further enhanced by the secondary activation of infected, HIV-specific memory cells in any response in which HIV is produced. Under treatment, in particular, such HIV-specific cells may facilitate the expansion of drug resistant variants with relative rapidity. The local nature and clonal segregation of virus transmission limit competition for predominance among viral mutants and facilitate diversity. The more restricted diversity during the acute phase and late in progression can be correlated to different predominant modes of virus transmission. The possibility that HIV-specific CD4 and CD8 T cells may exert opposite effects on HIV variant selection should be considered, as CD4 cells play a dual role in HIV pathogenesis.