Neuroinflammatory crosstalk between microglia and astrocytes increases viral replication in an iPSC-derived model of CNS HIV infection.

People living with HIV suffer multiple comorbid conditions related to chronic inflammation at increased rates compared to the general population, even when on effective antiretroviral therapy. In particular, current data indicate that the increased incidence and severity of neurocognitive impairment (NCI) are associated with unresolved neuroinflammation. Attempts to treat NCI in people living with HIV by reducing inflammation have thus far been unsuccessful, suggesting that a more mechanistic understanding of inflammatory processes in the CNS during HIV is necessary. Here, we use iPSC-derived microglia (iMg) and astrocytes (iAst) to model HIV infection in the CNS. We show that our iMg robustly express markers associated with microglial identity and are susceptible to HIV infection, but exhibit lower HIV replication rates and weaker immune response to HIV challenge compared to monocyte-derived macrophages. Coculture of iAst with iMg leads to a much stronger pro-inflammatory immune response, and, surprisingly, a robust increase in rates of HIV replication. Increased replication in iMg/iAst cocultures is associated with higher levels of multiple pro-inflammatory cytokines, including TNFα, which is produced by iAst upon exposure to HIV-infected iMg. Addition of exogenous TNFα to iMg during HIV infection is also sufficient to increase rates of replication, and neutralization of TNFα via adalimumab/Humira treatment in iMg/iAst cocultures reduces replication. Blocking NF-kB signaling with iKK inhibitor Bay-11-7082 (Bay-11) demonstrates that increased HIV replication in iMg/iAst cocultures is due to increased NF-kB activity. Finally, we show that in HIV-infected iMg there is movement of lysosomes to the periphery of the cell membrane and release of lysosomal content into the extracellular space, suggesting that this dysregulated lysosomal flux could further contribute to the pro-inflammatory microenvironment. We propose that this altered lysosomal trafficking and increased cytokine production drives a pro-inflammatory phenotype in glia and represents a potential source of unresolved neuroinflammation in people living with HIV.