DESCRIPTION (provided by applicant): While chronic HIV infection is associated with a complex, heterogeneous population of HIV variants, a newly infected individual harbors only a single or few variants of virus, almost all of which utilize the CCR5 co-receptor for cell entry. Although this transmission "bottleneck" was recognized over a decade ago, the biologic basis for this phenomenon has yet to be elucidated. Are these viruses selected because of their replicative advantage in the transmitting compartment (genital secretions, breast milk, and blood) and/or their selective amplification within the new host? Viral selection also occurs in mother to child transmission (MTCT) of HIV. Children are sequentially at risk for infection via three distinct routes of transmission; in utero infection across the placenta (blood borne), intrapartum infection by contact with blood and/or genital secretions and finally by exposure to breast milk. Thus, HIV infection of children provides the unique opportunity to assess the properties of HIV transmitted by different mechanisms after exposure to a known source. We propose to study two potential viral bottlenecks in transmission: the first being viral replication in the milieu of the transmitting compartment, and the second being these viruses' ability to infect a potential primary target cell in the new host. Our central hypothesis asserts that transmitted strains possess enhanced abilities to interact with cellular receptors (CD4/CCR5). A corollary to the hypothesis contends that viral transmission routed through the breast milk milieu will select for distinct viral envelopes (Env) possessing enhanced interactions with the CCR5 receptor relative to virus transmitted through in utero transmission. We also predict that strains that traverse the placental barrier will have a higher affinity for CD4 since placental cells express low levels of this receptor. To address these hypotheses, we specifically propose to: 1. Compare the phenotypic properties of HIV Env (gp 160) variants in breast milk and blood. 2. Compare the phenotypic properties of maternal gp160 variants to those in the child infected by different routes (in utero vs. breast milk). 3. Define the genotypic features of non-transmitted and transmitted gp160 by different routes (in utero and breast milk) and to correlate these features to biologic phenotype.