According to the currently accepted model, HIV sequentially interacts with two cellular receptors, CD4 and a chemokine receptor. Although several chemokine receptors have been identified, only CCR5 and CXCR4 play a major role as HIV-1 coreceptors. In particular CCR5 is the most widely used coreceptor among HIV-1 isolates. RANTES, MIP-lalpha and MIP-lbeta selectively bind to CCR5 and this binding makes RANTES a potent natural inhibitor of HIV-1. However, the potential therapeutic use of this chemokine has been limited due to the risk of inducing inflammatory side effects or of interfering with the natural physiology of the chemokine system. Recently, evidence has been provided that the functional domain(s) of the chemokines that block HIV infection could be uncoupled from the signaling activity, making this domain(s) a potential template for the design of an effective HIV-coreceptor inhibitor devoid of unwanted inflammatory side-effects. The goal of this project is to combine the potentially powerful anti-HIV activity of RANTES analogs with Osel's Lactobacillus expression technology, to provide a continuous supply of HIV inhibitors at the vaginal mucosa. Specifically, we propose to employ a naturally-occurring vaginal strain of Lactobacillus to express secreted RANTES analogs. We will identify the most potent and selective Lactobacillus-expressed RANTES-based HIV inhibitor, stably integrate its expression cassette into the bacterial chromosome, conduct an in vivo safety and colonization study in the pig-tailed macaque with the selected recombinant Lactobacillus strain, and determine the in situ expression levels of the secreted RANTES analog. Finally, we propose to conduct an efficacy study in the pig-tailed macaque model to determine whether the lactobacillus-expressed RANTES analog can prevent viral transmission following intravaginal infection with SHIV.