SUMMARY It has yet to be determined whether human brain myeloid cells (BMCs) serve as an important reservoir for persistent HIV-1 (HIV) infection in the central nervous system (CNS). This is hampered by the challenges faced in accessing fresh human brains. Therefore, our current knowledge of HIV CNS reservoirs in humans is largely limited to in vitro microglia/macrophage models of HIV latent infections and animal studies from SIV-infected NHP or HIV-infected humanized mouse models. There is an urgent need for innovative HIV latency models of BMCs that are physiologically close to the cells in the human brain. By working with the “Last Gift” Program, we have established a platform to isolate, expand, and characterize BMCs derived from people with HIV (PWH) on antiretroviral therapy (ART). We showed that a near full-length HIV can be recovered from BMCs, which was replication-competent in both T cells and BMCs. Bioinformatics analyses predicted the R5 tropism of this BMC HIV strain, which was confirmed by in vitro infection in BMCs, during which BMC infection by this strain of HIV was effectively dampened by a CCR5 inhibitor, Maraviroc. Pseudoviruses of Env cloned from this strain of BMC HIV efficiently infected CD4low but CCR5high affinofile cells, supporting its M-tropism nature. Interestingly, human BMCs expressed the newly identified HIV latency gene Nurr1 as well as anti-apoptosis genes, including Bcl-2 and MCL1, which could be essential for the stable HIV reservoir in BMCs. With the needed expertise to successfully isolate CNS cells (CNS T cells and BMCs) from postmortem fresh brains, technical innovations of single-cell technique pipelines, and the unique and rare resource of the “Last Gift” cohorts, this proposal will define the mechanism underlying the stable replication-competent HIV reservoir in BMCs. We will isolate BMCs from the brains of ART-suppressed PWH. We will define replication- competent HIV through de novo infections in T cells and BMCs as well as BMC-QVOA and T cell-QVOA analyses. We will genotype BMC HIV by single genome amplification sequencing and phenotype it by Env- pseudoviral infections in cells expressing low levels of CD4. We will characterize the transcriptionally active HIV as a possible viral reservoir in BMCs (Aim 1). In Aim 2, we will define the survival signaling pathways that may be acquired in the latently infected BMCs. We will study latency reversal using epigenetic modulators in BMCs. We will define the mechanisms of the epigenetic regulation of BMC HIV. We will also test whether previously identified BMC transcription factors play essential roles in HIV latency in postmortem BMCs. In Aim 3, we will examine new tools to directly target the transcription machinery of HIV to control viral reservoirs and delay or prevent viral rebound, which may also limit immune activation. This proposal will address critical knowledge gaps in replication-competent HIV reservoirs in BMCs and underlying molecular mechanism...