Perivascular collections of macrophages and endogenous microglial cells appear to be major reservoirs of HIV- l within the central nervous system (CNS) and a source of viral and host cell products which have been implicated in the pathophysiology of dementia and other neurologic manifestations of HIV-l infection. Mechanisms responsible for the accumulation of these mononuclear leukocytes (MNL) in the CNS and, specifically, the contribution of peripheral blood leukocytes to these interstitial infiltrates have not been determined, nor is it clear whether the virus is carried to the CNS within infected PBL. Preliminary studies demonstrate that HIV- l infection stimulates cell surface expression of molecules on MNL which mediate adhesive interactions with endothelium and neural cells. Moreover we have found that peripheral blood monocytes from HIV- l infected subjects are frequently activated. They display increased quantities of these same adhesins and are more likely to adhere to cultured neural cells in vitro. The experiments proposed in this application will evaluate the ability of MNL from HIV-I infected subjects to migrate across endothelial cell monolayers. Using in vitro methods which allow us to harvest, enumerate and compare the phenotypes of leukocytes which have the capacity to adhere to, and migrate across endothelium, we propose to investigate whether HIV- l infected cells are particularly abundant in this population and evaluate whether infection with putatively "neurotropic" strains of HIV- l enhances their ability to migrate across endothelium. Factors which may modulate endothelial transmigration of MNL will also be investigated, including the effects of astrocytes growing in contiguity with endothelial monolayers, infection of endothelial cells with cytomegalovirus and stimulation with cytokines, such as IL- l, that are likely to be released by activated MNL or endothelial cells. Adhesive interactions between MNL, which have demonstrated their ability to migrate through endothelial barriers, and neural cells will be investigated in an attempt to identify the molecular basis for these interactions. Monoclonal antibodies (mAbs) specific for beta integrins, selectins and other molecules known to mediate heterotypic adhesive cellular interactions will be used in an attempt to block the binding of MNL to neural cells. We also plan to raise antibodies to ligands that resist the ability of presently available mAbs to block adhesive interactions in order to fully characterize the array of molecules used by HIV-l infected and/or activated MNL to attach to neural cells. Upon completion, we postulate that the proposed investigations will provide a better understanding of the mechanisms in HIV- l infected individuals which regulate trafficking of MNL across blood vessels and cause them to be retained within the CNS.