The identification of chemokine receptors as coreceptors for HIV entry, not only has contributed to the understanding on viral tropism but has provided an additional target for therapeutic intervention for HIV disease. Several chemokine receptors have been shown to function as coreceptors for HIV-1 entry. The main ones are CXCR4 (for T-cell line tropic viruses) and CCR5 (for macrophage-tropic viruses). Because of the capacity of HIV to adapt when selective pressures are imposed, it is likely that any drug designed to block the interaction of HIV with one coreceptor will force the virus to use additional coreceptors. Thus, the determination of the complete coreceptor repertoire is necessary. Because CXCR6 (STRL33) is expressed in all lymphoid tissues, in collaboration with Dr J. Farber, NIAID, we tested it for coreceptor activity with HIV. CXCR6 expression in Jurkat cells conferred increased permissivity to infection by the ELI1 isolate of HIV-1. Thus, CXCR6 can act as an HIV-1 coreceptor in vitro. As well as testing the coreceptor activity of CXCR6 with a number of HIV-1 strains of different phenotypes, we have begun studies with HIV-2 and SIV. We have shown, in an infectivity assay, that the MAL strain of HIV-1 and the mac239 isolate of SIV use CXCR6 but not as well as they use CCR5. The appearance of virus only after about 30 days in culture was indicative of adaptation. To confirm this, virus emerging after about 35 days was used to infect fresh Jurkat-CXCR6 cells as well as the parent Jurkat cells. In this second passage, virus production was seen after about 12 days, thus demonstrating that both SIVmac239 and HIV-1 MAL had adapted to use CXCR6 more efficiently. Importantly, these passaged viruses were still unable to infect Jurkat cells. That the passaged virus had adapted to use CXCR6 was demonstrated by the fact that an antibody raised to CXCR6 inhibited virus infection. We have cloned 12 envelope genes from the CXCR6-adapted MAL virus and 6 env genes from the CXCR6-adapted SIVmac239. We have characterized 7 functional env genes from MAL and four from SIVmac239. All of the adapted clones had enhanced capacity to use CXCR6 over CCR5 in both a single-cycle assay and a productive infection assay. Sequncing of the MAL envs demonstrated that, while changes were found in several regions of gp120 and gp41, changes in the V3 region of gp120 were sufficient to confer the adapted phenotype. The activity of individual mutations in MAL has been assessed and changes in V3 and gp41 can confer the ability to use CXCR6, while changes in V2 and V5 cannot. Continued passage of the CXCR6-adapted MAL has resulted in virus that can infect the parent Jurkat cells, strongly suggesting that adaptation has been to CXCR4 use. We have demonstrated that infection of monocyte-derived macrophages (MDM) and monocyte-derived dendritic cells (MDDC) by R5 strains, but not X4 strains, of HIV-1 results in the induction of IP-10 and I-TAC, chemokines for CXCR3, but no induction was seen for CCR5, CXCR4, or CCR5 ligands. This is of interest for HIV disease, since all CCR5-expressing memory CD4 cells also express CXCR3, and thus infected tissue macrophages or DC would secrete IP-10 and I-TAC, which would then recruit memory CD4 T cells, which could be infected and thus disseminate infection. Recruited CD8 T cells could also play a role but in this case that in host defence. Induction of IP-10 and I-TAC depends on viral replication, since the effect is blocked by the RT inhibitor ZDV, and association of virus with the cellular receptors is not sufficient. Current work is directed towards understanding the mechanism of the induction.