Microglia Exhibit a Unique Intact HIV Reservoir in Human Postmortem Brain Tissue

Abstract

A proviral reservoir persists within the central nervous system (CNS) of people with HIV, but its characteristics remain poorly understood. Research has primarily focused on cerebrospinal fluid (CSF), as acquiring brain tissue is challenging. We examined size, cellular tropism, and infection-dynamics of the viral reservoir in post-mortem brain tissue from five individuals on and off antiretroviral therapy (ART) across three brain regions. Microglia-enriched fractions (CD11b⁺) were isolated and levels of intact proviral DNA were quantified (IPDA). Full-length envelope reporter viruses were generated and characterized in CD4⁺ T cells and monocyte-derived microglia. HIV DNA was observed in microglia-enriched fractions of all individuals, but intact proviruses were identified only in one ART-treated individual, representing 15% of the total proviruses. Phenotypic analyses of clones from this individual showed that 80% replicated efficiently in microglia and CD4⁺ T cells, while the remaining viruses replicated only in CD4⁺ T cells. No region-specific effects were observed. These results indicate a distinct HIV brain reservoir in microglia for all individuals, although intact proviruses were detected in only one. Given the unique immune environment of the CNS, the characteristics of microglia, and the challenges associated with targeting these cells, the CNS reservoir should be considered in cure strategies.

Keywords: CNS; HIV; brain; cellular tropism; microglia; replication-competent.

Figure 1

Quantification of the proviral reservoir. Copy numbers are indicated per million cells for all three brain regions in all five individuals. (A) The total proviral reservoir quantified via LTR copies. (B) The intact, psi, and env proviral copies are quantified with the IPDA. Single psi and env copies represent defective reservoirs. Traces of proviral DNA are defined as follows: for the LTR assay, reservoirs measured with 2–4 positive droplets, and for the IPDA, fewer than 7 detected copies, based on the limit of detection of the assay.

Figure 2

Infection dynamics of viral clones of participant 3. These viral clones of three different brain regions were used to infect CD4⁺ T cells (A,C,E) and MDMi (B,D,F) for 14 days. The Y-axis values are comparable but not identical between model systems and brain regions. All viruses were also monitored in the absence and presence of Maraviroc, a CCR5 inhibitor, prior to infection. Some viruses can replicate in MDMi, whereas all viral replications were inhibited by Maraviroc. MVC = Maraviroc.

Figure 3

Maximum likelihood phylogenetic tree of envelope sequences of participant 3. The bootstrap values are shown next to each branch and a value of >70% support clustering. Created with MEGA X v10.0.1.