We have previously demonstrated that an immunotoxin (IT) directed against CD45RO and containing chemically deglycosylated ricin toxin A chain (dgA), kills CD4-postive human T-cells which are both latently and actively infected with HIV-1 in vitro. We now propose to determine whether this IT can kill infected T-cells from HIV positive individuals and, in particular, latent cells from individuals receiving HAART. This study will involve perfecting a highly sensitive limiting dilution assay to determine the frequency of replication-competent cells. Experiments will include both in vitro- and in vivo-infected CD4-postive T cells infected with drug-resistant and drug-sensitive isolates of HIV-1. If the IT effectively kills the target cells, we will then determine what effect it has on immune memory in the CD4-positive and CD8-positive T-cell populations. This will be accomplished by treating peripheral blood cells from both HIV-negative and HIV-positive individuals with the IT and then activating cells with mitogens, superantigens, and antigens. Flow cytometry will then be used to evaluate intracellular cytokines in sub-populations of naive and memory CD4-postive and CD8-postive T-cells to determine whether T cell memory is compromised. This is a preclinical study to evaluate a new therapeutic modality. Our results should determine the feasibility of a future clinical trial.