Are HIV reservoirs larger than previously suspected?

The perhaps biggest challenge to curing HIV infection is that the virus hides in latently infected, resting memory CD4+ T cells. These cells harbor integrated HIV DNA—the so-called provirus—in their genome, and unless they are activated to produce virus, they are indistinguishable from uninfected resting CD4+ T cells. One strategy to eradicate this reservoir is the so-called “kick and kill” approach: activate the latently infected cells so that they give themselves away by producing virus again, and then kill these virus-producing cells. 

Because in vitro assays that activate these latently infected cells cause less than 1% of them to produce virus, researchers used to think that the rest of them harbor nonfunctional proviruses. But a new study suggests that the fraction of latently infected cells that harbor functional provirus may be 60 times larger than previously thought (Cell 155, 540, 2013). 

A team led by Robert Siliciano of the Johns Hopkins University School of Medicine isolated resting CD4+ T cells from the blood of eight patients on highly active antiretroviral therapy and measured which cells could produce virus when they were activated in vitro. Many could not, and sequencing revealed that the HIV DNA in many of the resting CD4+ T cells that couldn’t produce virus had crippling deletions or mutations. 

The researchers found, however, that about 12% of the cells that failed to produce virus in the in vitroactivation assay harbored complete HIV sequences without any inactivating mutations or deletions. This suggested that, in principle, these cells should be able to produce functional virus. And indeed, when the researchers synthesized DNA molecules that were identical to the functional HIV sequences, they could use the molecules to make completely functional and fully infectious HIV particles in cultured cell lines. 

“This work provides us with a better understanding of the true size of the latent reservoir,” says Siliciano. If all of the intact proviruses observed in the study can also be induced in vivo, then the true size of the reservoir is about 60-fold greater than previously estimated, the researchers say. 

Their work also suggests that in vitro activation assays may underestimate the extent of the reservoir, because activating latently infected cells doesn’t always prompt them to produce virus. The only way to check the extent of the reservoir in infected people may be to sequence the HIV DNA inserts in their latently infected cells. Currently, Siliciano and colleagues are trying to develop simplified assays that are based on the DNA sequence and can be used in the clinic to check reservoir size. 

The researchers are also looking for the cellular or viral factor(s) that keep many latently infected cells from producing virus when they are activated. Finding such a factor might lead to more efficient strategies to induce virus production in latently infected cells than the ones that are currently available.