Experimental agents reverse HIV latency and help immune system fight HIV-infected cells

Researchers at the recent Conference on Retroviruses and Opportunistic Infections (CROI 2015) in Seattle, USA, presented data on several experimental agents that may play a role in achieving a 'functional cure' for HIV, or prolonged remission without disease progression. These include drugs that reactivate the latent HIV reservoir, interfere with expression of viral DNA and help the immune system target HIV-infected cells.

As part of its lifecycle, HIV integrates its genetic material into the chromosomes of host cells and uses the cell's machinery to produce new virus. Soon after infection, HIV establishes a reservoir of latent genetic material, known as proviral DNA, in inactive or resting T-cells. While antiretroviral therapy (ART) can effectively control viral replication long-term, the virus soon comes back if the drugs are stopped.

Kick and kill

One cure strategy is known as 'kick and kill'. The idea is to flush out HIV by activating latent proviral DNA in resting cells, in the hope that once inactive cells 'wake up' and start producing new virus they will be recognised and eliminated by the immune system.

The agents that have been most extensively studied for this approach are histone deacetylase or HDAC inhibitors, some of which are already used as cancer drugs. HDACs are enzymes that keep DNA tightly coiled in a cell's nucleus so it cannot be used to produce of new proteins. HDAC inhibitors reverse this process, allowing proviral DNA gene expression and production of new virus. Prior studies have shown that the HDAC inhibitors vorinostat and romidepsin can activate latent HIV gene expression in resting T-cells.

Sarah Palmer from the University of Sydney presented similar findings at CROI for another of these drugs, panobinostat. Fifteen people on ART received oral panobinostat three times weekly, every other week, for eight weeks. The researchers saw increased levels of genetically diverse cell-associated HIV RNA, indicating non-selective activation of resting memory T-cells. Of note, nearly 40% of this cell-associated HIV RNA was 'hypermutated' and not able to produce replication-competent virus.

While HDAC inhibitors can kick inactive HIV-containing cells into gear, these cells may not necessarily become visible to the immune system, and so far the drugs have not been shown to shrink the size of the viral reservoir. Nonetheless, they may have a place in a combination cure strategy.

The Danish team that presented their work on romidepsin at last summer's International AIDS Conference, reported at CROI that romidepsin and panobinostat do not impair the activity of HIV-specific CD8 T-cells, providing support for combining HDAC inhibitors with therapeutic vaccines or other immune-based therapies in HIV eradication trials.

TLR7 agonist

Given that HDAC inhibitors can potentially cause toxicity and immune suppression, Gilead Sciences is looking at a different type of latency-reversing agent, the toll-like receptor 7 (TLR7) agonist GS-9620.

Toll-like receptors on immune cells play a role in defence against viruses and other pathogens. TLR7 activation leads to increased antigen presentation and enhanced activation of natural killer cells, CD8 killer T-cells and CD4 T-cells.

James Whitney from Harvard Medical School and colleagues evaluated an analogue of GS-9620 in ten rhesus macaques with SIV, an HIV-like virus that infects monkeys. The macaques started combination ART about 9 weeks after infection and achieved stable viral suppression. At 45 weeks post-infection, four of the monkeys received seven oral doses (0.1-0.3 mg/kg) of the TLR7 agonist every other week, while the other six received a placebo.

The monkeys showed a consistent increase in CD8 T-cell activation while taking the TLR7 agonist, as well as a smaller increase in CD4 T-cell activation, which returned to baseline levels after the last dose.

The first three doses had no effect on plasma viremia, but after the fourth or fifth dose all treated monkeys had transient increases in plasma viral load (500-1000 copies/ml), which returned to undetectable within 4-7 days after the last dose. The control animals showed no change in viral load.

Viral DNA levels decreased by 30-90% in peripheral blood mononuclear cells (PBMCs) and in colon and lymph node biopsy samples from three of the four treated monkeys; levels again remained unchanged in the control group.

ART was discontinued two weeks after the last TLR7 agonist dose to see if these transient plasma viral blips and decreases in viral DNA corresponded to a reduction in the latent viral reservoir. The speed of plasma viral load rebound was similar in the treated and control monkeys, but those that received the TLR7 agonist had about a 0.5 log₁₀ lower viral set-point than the control group. 


The researchers concluded that the TLR7 agonist "was safe, induced transient plasma viremia, reduced viral DNA content in PBMCs, colon and lymphoid tissues and established lower viral set-point after ART cessation."

"We get some kick, but also get some kill," Whitney explained at a CROI press conference. "After repeated dosing we can elicit viraemic blips, which tells us we're unleashing the latent reservoir." He added that the TLR agonist was very well tolerated.

In a related poster presentation, Gilead scientists reported that GS-9620 induced HIV RNA expression in cultures of PBMCs from 18 people living with HIV on stable suppressive ART who underwent leukapheresis, a procedure in which blood is removed, white cells are separated out, and the rest is returned to the donor. In a laboratory study, the collected PBMCs were exposed to GS-9620 (100 or 1000 nM) or DMSO as a control for four days, along with antiretrovirals.

GS-9620 activated HIV expression more than DMSO: at either dose GS-9620 induced at least 2-fold greater HIV activation in cells from 13 of the 18 donors (72%), with a mean of 9-fold greater and up to 27-fold greater. The drug's variable effect on cells from different donors may be due to variations in the combination of 'kick' and 'kill' effects, the researchers suggested.

GS-9620 also induced production of a wide variety of cytokines (immune cell chemical messengers) including interferons, interleukins, tumour necrosis factor-alfa, IP-10 (CXCL10) and I-TAC (CXCL11). Interferon alfa/beta (type I) receptor signalling was required for maximal HIV activation, and blocking the receptor reduced HIV expression by up to 85%. Viral activation was also associated with peak production of IP-10 and I-TAC, although the role of these chemokines in HIV activation is not yet known.

The drug's effect on the latent HIV reservoir was assessed by using a protein kinase C (PKC) agonist to activate CD4 T-cells. GS-9620 led to decreased HIV expression after activation, but the researchers said more studies are needed to determine if this indicates reduction of the viral reservoir.

A clinical trial of GS-9620 is now underway in ART-treated people with HIV. The drug is also being evaluated as a treatment for hepatitis B, and recently entered a phase 2 trial for that indication.

Targeting HIV gene expression

Several researchers presented findings on a variety of other agents that might one day contribute to functional cure.

Two teams looked at ways to interfere with production of HIV as it emerges from the latent reservoir by blocking viral regulatory proteins needed for HIV gene expression.

Susana Valente from the Scripps Research Institute in Florida, working with Nicholas Chomont's group at the Vaccine and Gene Therapy Institute, evaluated didehydro-Cortistatin A (dCA), a steroid alkaloid that inhibits activity of HIV's Tat protein, which activates viral gene transcription. They noted that Tat is an attractive target because it is active at an early stage of viral replication, sets off a feedback loop that drives exponential virus production and does not affect human cell processes, which can cause toxicity.

In a laboratory study, the researchers showed that dCA interfered with HIV reactivation from latently infected CD4 T-cells obtained from people with HIV on suppressive ART and reduced production of cell-associated HIV RNA.

They concluded that "dCA treatment combined with ART may inhibit and persistently abrogate residual HIV production from cellular reservoirs in blood and tissues from virally suppressed subjects, block viral reactivation, reduce reservoir replenishment and may ultimately decrease the size of the latent reservoir."

In the second study, Jamal Tazi from the French company ABIVAX and colleagues looked at another regulatory protein, Rev. Rev is needed to transport unspliced bits of messenger RNA from the cell nucleus to the cytoplasm, where they direct production of HIV structural proteins used to construct new virus particles.

The researchers screened a library of compounds and found one, dubbed ABX464, that interferes with Rev activity and causes splicing of HIV RNA. Using humanised mouse models, they showed that ABX464 inhibits replication of various HIV subtypes in PBMCs and macrophages. They noted that a phase 1 study of healthy HIV-negative volunteers found that a single dose of ABX464 was safe and well tolerated.

Aiding immune response

Finally, instead of focusing on releasing HIV from the reservoir or interfering with its replication once this happens, other researchers are exploring ways to help the immune system recognise and eliminate HIV-infected cells. The hope is that once ART reduces HIV to a very low level in the body, a stepped up immune response will be able to control the remaining virus without antiretrovirals.

Grant Campbell and colleagues from the University of California at San Diego aimed to determined whether XIAP (X-linked inhibitor of apoptosis protein), which is up-regulated in infected CD4 memory T-cells, might serve as a 'molecular signature' enabling the immune system to selectively target and kill latent cells harbouring HIV.

In a laboratory study, the researchers infected CD4 T-cells from HIV-negative donors, isolated the resting memory T-cells and treated them with GDC-0152 (a XIAP antagonist) or embelin (a XIAP inhibitor).

After 32 days of infection, CD4 T-cells displayed a resting central memory cell phenotype and contained the equivalent of 1 copy of integrated HIV DNA. XIAP expression was significantly increased in these cells compared to uninfected CD4 cells. Adding a XIAP inhibitor or antagonist caused a greater number of latently infected resting memory cells to undergo apoptosis, or programmed cell death.

The researchers concluded that XIAP acts as a molecular signature of latently infected, long-lived resting central memory CD4 T-cells and that targeting XIAP with selective inhibitors might be a way to eliminate latently infected cells without reactivating HIV gene expression.

Another cell signalling molecule, PD-1 (programmed death protein 1), helps regulate immune response by reining in excessive immune activation. PD-1 has been in the news recently as a novel approach to cancer treatment.

PD-1 accumulates over time in activated CD8 and CD4 T-cells, eventually signalling cell exhaustion and marking the cell for death. HIV-specific CD8 T-cells often show up-regulation or increased expression of PD-1 during chronic infection. Blocking PD-1, or its ligand (binding partner) PDL-1, may reinvigorate these cells and restore their ability to fight the virus.

At a community cure workshop prior to CROI, Stephen Mason from Bristol-Myers Squibb reported that the monoclonal antibody PDL-1 inhibitor BMS-936559 reduced viral load and improved T-cell function in SIV-infected macaques. A clinical trial of the drug is now underway in people with HIV.

In a study of eight macaques on ART who were treated with BMS-936559 and then stopped antiretrovirals, responses were highly variable. Some monkeys showed no response and experienced normal viral rebound, some maintained a low viral load throughout the study, some had their viral levels go up and down, and a couple had prolonged undetectable viral load followed by attenuated rebound. But overall, most had a lower viral load than their pre-ART set-point and most did not show an increase in proviral DNA in PBMCs or gut T-cells, suggesting that the PDL-1 inhibitor might prevent expansion of the viral reservoir.

Mason speculated that the PDL1 inhibitor restores the function of HIV-specific T-cells and enables them to attack latently infected reservoir cells, but this response is incomplete since those cells only sporadically and randomly express HIV. Like most cure researchers, Mason expects that long-term HIV remission will require a combination of approaches. For example, a strong latency activator may make more reservoir cells express virus and thus render them more vulnerable to T-cells reinvigorated by a PD1 or PDL1 blocker.

"The cure field started with enthusiasm, and enthusiasm is still there, but progress will be slow," Mellors concluded at the CROI press conference. "It will wind out over years or decades before we have functional cure applicable to many people with HIV."