TAG HIV Basic Science, Vaccines, and Cure Project Blog

HIV Latency: Trapped in Transition, Lurking in Treg

The question of how HIV latency is typically established in long-lived memory CD4 T cells has yet to be fully resolved. One theory is that latency occurs when HIV infects a CD4 T cell that is in the process of transitioning from an activated state to a resting state, but direct evidence has been lacking. A new paper from the research group of Robert Siliciano, published in the journal Immunity, provides experimental support for this model of HIV latency initiation.

For background, HIV preferentially targets activated CD4 T cells because their genetic code (the genome) is very busy transcribing the proteins required to for the cell to go about its work, and the virus is able to hijack this process by integrating its own DNA into the cell’s DNA in order to generate new virions (which can then go on to infect other cells). Resting CD4 T cells are less amenable to HIV replication because their genome is largely shut down, and the virus has difficulty coopting it for the production of new viruses.

Siliciano’s research offers evidence that HIV latency occurs when the virus’s DNA essentially becomes entrapped in the genome of an activated CD4 T cell as it is in the process of shutting down, due to the transition of the cell into a resting state. An example of when these transitions normally occur is after a CD4 T cell has responded to an infection—when the responsible pathogen is cleared or controlled, a proportion of the activated responding CD4 T cells return to a resting memory state, ready to react again should the same pathogen cause further trouble.

The experiments described in the new paper mimicked the transition process in the laboratory by activating CD4 T cells (modified with a gene, Bcl-2, that enhances their survival in the lab dish) and then allowing them to return to rest. A key finding was that latent infection occurred most frequently when activated CD4 T cells were infected with HIV at a time when analyses showed the transcriptional activity of their genome was shutting down—on days 6, 9 or 15 after receiving an activating stimulus. Studies of CCR5 co-receptor expression found that, at these timepoints, the CD4 T cells were displaying enough CCR5 on their surface to facilitate infection by R5-tropic HIV.

The results contrasted with those obtained when CD4 T cells were infected with R5-tropic HIV immediately after activation, or when they were already in a resting state. In these experiments, essentially no establishment of viral latency could be detected.

The researchers found that X4-tropic HIV was more promiscuous in its ability to cause latent infection in CD4 T cells—whether activated, resting, or in transition—which they note is “probably due to the universal expression pattern of CXCR4 on CD4 T cells.” However, analyses of latently infected CD4 T cells isolated directly from HIV-positive individuals on ART demonstrated that the majority contained R5-tropic virus, implying that the mechanism involving infection of activated CD4 T cells transitioning to a resting state is predominant in creating the latent reservoir.

Interestingly, the laboratory studies revealed that while most of the activated CD4 T cells downregulated CCR5 expression as they transitioned into a resting state, a subset continued to display the molecule. Subsequent assessments of samples from the individuals on ART found that levels of HIV DNA were 10- to 100-fold higher in these CCR5-expressing resting memory CD4 T cells compared to CCR5-negative subsets. The CCR5-expressing resting memory CD4 T cells also harbored a greater proportion of replication-competent HIV, suggesting that CCR5 should be included among the candidate biomarkers for identifying latently infected cells.

In a final set of experiments, the researchers showed that functional HIV-specific CD8 T cells have the potential to recognize and destroy virus-infected activated CD4 T cells that are transitioning to a resting state, thereby reducing the frequency of latent infection. These results indicate that effective HIV-specific CD8 T cell responses have the potential to inhibit the formation of the latent reservoir.

The model of HIV latency described by Siliciano and colleagues differs from some other proposed scenarios. For example, a recent paper identifying CD32a as a biomarker of latently infected CD4 T cells reported that the finding was consistent with direct HIV infection of resting memory CD4 T cells as the initiator of latency. The research group of Sharon Lewin in Australia has also demonstrated that certain chemokines and myeloid dendritic cells can facilitate HIV infection of resting memory CD4 T cells and generate latently infected cells.

Evidence in support of the in vivo relevance of Siciliano’s results may come from studies of the specific locations in the genome of CD4 T cells into which HIV commonly integrates – at the IAS 2017 meeting earlier this year, Jori Symons from Sharon Lewin’s laboratory noted that HIV DNA is more commonly found integrated into cellular genes that would be predicted to be active in activated CD4 T cells (a video of the presentation is available, and the slides can be downloaded).

It may not be an either-or proposition, as both mechanisms might conceivably contribute to the formation of the latent HIV reservoir.

Another recent theory to explain HIV latency involves the virus having specifically evolved the strategy in order to facilitate crossing the mucosal barrier at the time of initial infection. In the discussion section of their paper, Siliciano and colleagues argue that their data supports a simpler explanation: "latency results from infection of cells in a narrow time window after activation when cells are permissive for infection but not for prolonged HIV-1 gene expression."

Treg and Latent HIV

In the same issue of Immunity, Colleen S. McGary and colleagues report that a previously obscure subset of memory CD4 T cells that express the immune checkpoint receptor CTLA-4, but not PD-1, contribute significantly to the latent virus reservoir. The findings are largely derived from SIV-infected macaques but the researchers also documented the presence of these cells in lymphoid tissue samples from HIV-positive individuals on ART. The majority of the CTLA-4+PD-1- CD4 T cells were found to be of a regulatory (Treg) phenotype, a subset typically involved in dampening down immune activation. Several previous studies have reported that Treg can harbor latent HIV (e.g. Jiao YM et al and Tran TA et al), and a cellular gene that has been identified as a common HIV integration site—BACH2—is known to be involved in Treg differentiation.

The researchers suggest that immune checkpoint inhibitors that target CTLA-4, such as the licensed anticancer antibody ipilimumab, may be able to reverse HIV latency in the CTLA-4+PD-1- CD4 T cell population; in support of this suggestion, they cite a published case report regarding an HIV-positive individual receiving ipilimumab for the treatment of cancer, which documented an increase in cell-associated HIV-1 RNA. Further exploration of this possibility is likely to focus on the SIV/macaque model, at least initially, because ipilimumab has a number of potentially serious side effects.

Locking Down Latent HIV

Strategies for dealing with the reservoir of latent HIV that persists despite antiretroviral therapy (ART) have primarily focused on awakening the virus from its dormant state. But Susana Valente and colleagues from the Scripps Research Institute in Florida are taking a different tack, flipping the idea of latency reversal on its head. Their approach—which they have dubbed block and lock—involves trying to imprison latent HIV in a way that prevents it from ever reactivating. The latest work by Valente’s group was published recently in the open access journal Cell Reports.

The intervention Valente and colleagues are testing is Didehydro-Cortistatin A (dCA), an analogue of a natural substance isolated from the marine sponge Corticium simplex. Previous studies have shown that dCA inhibits the HIV Tat protein, and can have the effect of pushing the virus into a deep state of latency that is not reversed by T cell activation or latency-reversing agents in laboratory tests.

The new paper extends these results by testing the longer-term effects of dCA on latently infected CD4 T cells isolated from five HIV-positive individuals on suppressive ART. An initial evaluation found that the combination of dCA+ART led to greater suppression of HIV RNA production by the cells during prolonged culture when compared to just ART.

Another experiment maintained the cells for a 35-day period in culture in the presence of either ART or dCA+ART, then withdrew the treatments and stimulated the cells with the latency-reversing agent prostratin. HIV rebound occurred immediately in cultures that had been treated with ART, but viral reactivation was inhibited by an average of 99% in the cultures exposed to dCA+ART, with two participant samples showing 100% inhibition. The difference was statistically significant. Several other similar experiments found HIV rebound was significantly delayed and reduced as a result of dCA, including after activating cells with the mitogen PHA.

To gain insight into the potential in vivo relevance of the findings, studies were also performed in humanized mice infected with HIV. Treatment with dCA+ART was associated with 10.5-fold lower levels of HIV RNA in lymph nodes and seven-fold lower levels in brain tissue compared to ART, but differences in the spleen, bone marrow and peripheral blood mononuclear cells (PBMC) were not statistically significant.  

A subsequent test also demonstrated that dCA+ART delayed HIV viral load rebound in HIV-infected humanized mice after treatment was interrupted, compared to ART. The latter group displayed evidence of recrudescing viral load at day three after interruption, extending to all eight animals by day 10. In contrast, among the dCA recipients, six out of 10 still had undetectable viral load at day 10. Viremia eventually became detectable in all the animals in the dCA group at day 19.

The researchers propose that dCA or similar Tat inhibitor compounds could have a role in providing additional suppression of HIV and limiting the formation of the viral reservoir if given in combination with early ART. They also suggest that, while dCA delayed rather then prevented HIV rebound after ART withdrawal in the current study, a greater duration of treatment might lead to more prolonged repression of viral reactivation from latency. The hope is that: “over time (in combination or not with other inhibitors), transcriptional repression could be pushed past a certain threshold where viral reactivation from latency is extremely difficult to overcome, blocking and locking HIV into sustained latency.” Additional studies are planned in macaque models, with the goal of advancing the approach into clinical trials.