HIV-1 infection of microglial cells in a reconstituted humanized mouse model and identification of compounds that selectively reverse HIV latency.

Most studies of HIV latency focus on the peripheral population of resting memory T cells, but the brain also contains a distinct reservoir of HIV-infected cells in microglia, perivascular macrophages, and astrocytes. Studying HIV in the brain has been challenging, since live cells are difficult to recover from autopsy samples and primate models of SIV infection utilize viruses that are more myeloid-tropic than HIV due to the expression of Vpx. Development of a realistic small animal model would greatly advance studies of this important reservoir and permit definitive studies of HIV latency. When radiation or busulfan-conditioned, immune-deficient NSG mice are transplanted with human hematopoietic stem cells, human cells from the bone marrow enter the brain and differentiate to express microglia-specific markers. After infection with replication competent HIV, virus was detected in these bone marrow-derived human microglia. Studies of HIV latency in this model would be greatly enhanced by the development of compounds that can selectively reverse HIV latency in microglial cells. Our studies have identified members of the CoREST repression complex as key regulators of HIV latency in microglia in both rat and human microglial cell lines. The monoamine oxidase (MAO) and potential CoREST inhibitor, phenelzine, which is brain penetrant, was able to stimulate HIV production in human microglial cell lines and human glial cells recovered from the brains of HIV-infected humanized mice. The humanized mice we have developed therefore show great promise as a model system for the development of strategies aimed at defining and reducing the CNS reservoir.