The origin and extent of viral replication within the CNS during early HIV-1 infection remains unclear. We aimed to assess cerebral spinal fluid (CSF) for host cells actively transcribing virus in acutely SHIV infected macaques and compare infected cell frequency and cellular activation status to that observed in peripheral blood and lymph nodes.
Rhesus macaques (n=18) were infected intrarectally with a subtype C, HIV-1 env SHIV (1157ipd3N4). CSF, peripheral blood, and lymph node mononuclear cells (PBMC and LNMC) were analyzed at weeks 2-12 post-infection (PI). CD4 and CD8 T cells and macrophages were sorted, using flow cytometry, from each specimen directly ex vivo. SHIV RNA+ cells were identified by RT-qPCR assays specific for unspliced (gag) and spliced (env) viral RNA. Infected cell frequency was estimated by Poisson distribution statistics for PBMC and LNMC or assigning one infected cell to positive CSF replicates. Markers of cellular activation were measured by flow cytometry (surface staining) and gene expression (RT-qPCR).
CSF specimens yielded an average 620 (120-2,840) CD4 T cells and 130 (2-400) macrophages after sorting. Infected, transcriptionally active (env+gag+) CD4 T cells were detected within the CSF in 25% of animals 4 weeks PI and 12% 12 weeks PI. In animals with SHIV RNA+ CSF CD4 T cells, infected (gag or env RNA+, respectively) CD4 T cell frequency was similar across CSF (0.05-2%, 0.3-1%), PBMC (0.02-7%, 0.02-2%), and LNMC (0.03-2%, 0.06-0.09%), indicating comparable T cell infection rates in these compartments in early infection (Figure 1). CSF blood contamination was minimal by ELISA and distinct cell composition. While macrophage infection was less frequently observed in CSF, the limited number of these cells constrained sampling depth. Surface expression of CD38 was elevated on CD4 and CD8 T cells in both PBMC and CSF during acute SHIV compared to uninfected controls (p<0.05). In contrast, the monocyte activation marker CD169, as well as CD38, was elevated on monocytes in PBMC (p<0.05) but not CSF, indicating T cell but not monocyte activation in CSF during acute infection. CSF CD4 T cells and macrophages both upregulated CXCL10 compared to uninfected controls and therefore might contribute to early CSF inflammation.
Our data supports a model of productive CD4 T cell infection within the CNS during acute HIV/SHIV infection, distinct from the role of macrophages in end-stage neuroencephalitis.