Bimodal Expression Patterns, and Not Viral Burst Sizes, Predict the Effects of Vpr on HIV-1 Proviral Populations in Jurkat Cells

Integration site landscapes, clonal dynamics, and latency reversal with or without vpr were compared in HIV-1-infected Jurkat cell populations, and the properties of individual clones were defined. Clones differed in fractions of long terminal repeat (LTR)-active daughter cells, with some clones containing few to no LTR-active cells, while almost all cells were LTR active for others. Clones varied over 4 orders of magnitude in virus release per active cell. Proviruses in largely LTR-active clones were closer to preexisting enhancers and promoters than low-LTR-active clones. Unsurprisingly, major vpr(+) clones contained fewer LTR-active cells than vpr(−) clones, and predominant vpr(+) proviruses were farther from enhancers and promoters than those in vpr(−) pools. Distances to these marks among intact proviruses previously reported for antiretroviral therapy (ART)-suppressed patients revealed that patient integration sites were more similar to those in the vpr(+) pool than to vpr(−) integrants. Complementing vpr-defective proviruses with vpr led to the rapid loss of highly LTR-active clones, indicating that the effect of Vpr on proviral populations occurred after integration. However, major clones in the complemented pool and its vpr(−) parent population did not differ in burst sizes. When the latency reactivation agents prostratin and JQ1 were applied separately or in combination, vpr(+) and vpr(−) population-wide trends were similar, with dual-treatment enhancement being due in part to reactivated clones that did not respond to either drug applied separately. However, the expression signatures of individual clones differed between populations. These observations highlight how Vpr, exerting selective pressure on proviral epigenetic variation, can shape integration site landscapes, proviral expression patterns, and reactivation properties.