BACKGROUND: Some drugs that augment cell-intrinsic defenses or modulate cell death/survival pathways have been reported to selectively kill HIV-infected cells and/or reduce HIV DNA. We hypothesized that these drugs may differ in their ability to kill cells infected with intact and defective proviruses.
METHODS: We tested drugs currently in clinical use or human trials, including interferon alpha2A, interferon gamma, acitretin (RIG-I inducer), GS-9620/vesatolimod (TLR7 agonist), nivolumab (PD-1 blocker), auranofin (p53 modulator), obatoclax (Bcl-2 inhibitor), FX-1 (Bcl-6 inhibitor), bortezomib (proteasome inhibitor), birinapant (IAP inhibitor), and INK128/sapanisertib (mTOR[c]1/2 inhibitor). Drug concentrations were chosen based on levels attainable in plasma (if known) or previously tested in vitro. PBMCs were isolated from eight ART-suppressed PWH, aliquoted into single or duplicate wells (6x106 cells/well), and cultured for six days with ARVs and either DMSO (negative control), anti-CD3/CD28+IL-2, or individual drugs. After six days, we measured cell counts/viabilities and extracted DNA/RNA. Total, intact, and defective HIV DNA were measured by IPDA (4-10 replicates), normalized to copies/million cells (using DNA mass and housekeeping genes), expressed as percent of the DMSO control, and compared to DMSO (Wilcoxon signed rank test).
RESULTS: A trend toward lower cell viability was observed with auranofin (median=86.7% of DMSO; P=0.11), FX-1 (median=84.6%; P=0.063), and obatoclax (median=93.9%; P=0.078), so the doses were subsequently reduced. Obatoclax reduced intact HIV DNA (median=17.0% of DMSO [range 0-82.9%]; P=0.0078) but not defective or total HIV DNA. A trend towards lower intact HIV DNA was also observed with auranofin (median=53.9%[0-156.5%]; P=0.11), bortezomib (median=78.6%[0-121.5%]; P=0.11), and INK128 (median=33.2%[0-138.2%]; P=0.12). IFNalpha2A resulted in the lowest median intact HIV DNA (18.2%), but effects were not consistent (range:0-232%; P=NS). Vesatolimod reduced 3’-defective HIV DNA (median=63.7%[0-87.3%]; P=0.031) and tended to reduce 5’-defective DNA (median=75.0%[0.5-101.4%]; P=0.063) but not intact or total HIV DNA. Other drugs showed no statistically significant effects.

CONCLUSIONS: Several drugs induced selective ex vivo depletion of cells with intact proviruses (obatoclax, possibly auranofin, INK128, and bortezomib) or defective proviruses (vesatolimod). Their distinct modes of action on cell death/survival and innate immune response provide a strong rationale to compare the effects of drug combinations ex vivo and in animal or human trials.