A Flurry of Updates from Keystone

Advances in understanding the arms race between HIV and its host, and efforts to capitalize on recent success in the HIV prevention field were among the highlights of this year’s joint symposia

By Andreas von Bubnoff and Regina McEnery

This year’s joint Keystone Symposia on “HIV Evolution, Genomics and Pathogenesis” and on “Protection from HIV,” which took place March 20-25 in Whistler, British Columbia, came at a time when the HIV prevention field has been transformed by success. Recent clinical trials, including RV144, which tested a prime-boost vaccine regimen; CAPRISA 004, which tested an antiretroviral (ARV)-based microbicide gel; and iPrEx, which tested oral pre-exposure prophylaxis (PrEP), have all yielded positive results.

“These trials reinvigorate our research efforts. They provide a signal for the control of transmission, a signal which we can optimize and hopefully improve,” said Mario Roederer, a T-cell immunologist at the Vaccine Research Center (VRC) at the National Institute of Allergy and Infectious Diseases (NIAID), who helped organize the Protection from HIV track. He said one goal of the track was “to really expose the vaccine community to the microbicides people and vice versa because these approaches will have to be undertaken together in the future.” The HIV pathogenesis-related talks also had an additional focus this year—HIV evolution and genomics. The discussion focused on large-scale genomic analyses of hosts and viruses, and also on what clues evolution provides about virus replication and host responses. Select highlights from both tracks are detailed in this article.

HIV restriction factors

In 2002, the human host cell factor APOBEC3G, and its role in inhibiting HIV replication by introducing hypermutations, was first described. Researchers also discovered that HIV counters APOBEC3G with its Vif protein, which triggers APOBEC3G degradation. Therefore, APOBEC3G can restrict replication of HIV particles that are lacking Vif. Other than APOBEC3G, six additional human APOBEC3 proteins are known. “The critical question is, which of these are relevant to HIV?” asked Reuben Harris, a professor of biochemistry, molecular biology, and biophysics at the University of Minnesota.

The answer, Harris reported, is that three other APOBEC3 proteins, in addition to APOBEC3G, contribute to HIV restriction. To determine this, Harris and colleagues used a CD4+ T-cell line that can be infected by HIV lacking Vif because it expresses only low levels of APOBEC3 proteins. They found that four APOBEC3 proteins, including APOBEC3G, could keep the cells from getting infected with HIV lacking Vif. This was also the case for the rhesus macaque versions of these APOBEC3 proteins, suggesting that the ability of these four proteins to restrict HIV infection is evolutionarily conserved.

In other experiments, Harris and colleagues eliminated each of the APOBEC3 genes from a CD4+ T-cell line and checked which and how many APOBEC3 genes they needed to remove until they didn’t see any hypermutations in the Vif-deficient HIV particles used to infect the cells. They found that the APOBEC3-induced hypermutations disappeared when the cells lacked the same four APOBEC3s that they had previously shown to be sufficient for restriction of HIV infection.

The finding that four APOBEC3 proteins are involved in causing hypermutations in HIV could mean that drugs that inhibit Vif to treat HIV infection might unleash even more APOBEC3-driven hypermutations in HIV than previously thought, effectively killing the virus by mutating it to death, Harris said. “[This] is good news because if the [Vif] inhibitor is properly designed, it will unleash four instead of one APOBEC,” said Harris.

Instead of inhibiting Vif, Harris is working on finding chemicals that can inhibit APOBEC3 proteins, which would result in reducing the HIV mutation rate in infected cells. “The [APOBEC3s] contribute to the virus mutation rate and to immune escape and drug resistance,” Harris said. “If you can decrease how fast HIV changes, you will make it less prone to developing resistance and you will make it more susceptible to normal immune responses.” Ideally, it might turn HIV into something more similar to other viruses that can actually be eliminated by the immune response, such as the flu, he added.

Interactions like the one between APOBEC3 and Vif are the result of a molecular arms race between the virus and the host, over millions of years, said Aris Katzourakis, a Royal Society university research fellow at the University of Oxford. “What we see today has been shaped by conflict between viruses and their host that goes back all the way to the beginnings of mammal evolution,” he said. Katzourakis works in the emerging field of “palaeovirology,” which explores the “fossil record” of ancient viral infections in the genomes of animals, taking advantage of the fact that retroviruses, such as HIV, integrate into the genome of cells. When they integrate into germline cells, the integrated virus is inherited, and can potentially still be found millions of years after the initial integration event.

Katzourakis gave an overview of some of his recent findings on the evolution of lentiviruses, the genus of retroviruses to which HIV belongs. He found that the vif gene HIV uses today as a counter defense against APOBEC3 proteins must be quite old because it is found in pSIVgml, a simian immunodeficiency (SIV)-related lentivirus integrated in the genome of the grey mouse lemur (1). Because grey mouse lemurs live in Madagascar, which has been separated from mainland Africa for millions of years, pSIVgml and vif must also be millions of years old, Katzourakis said. “By finding vif in an endogenized ancient lentivirus we can say that primates have probably been infected by lentiviruses and have been evolving countermeasures against them for millions of years,” said Katzourakis.

He also found evidence that some of the ancient viral structures that are preserved as “fossilized” viruses integrated in the genome of vertebrates can still interact with present day host cell factors related to human Trim5α, a well known restriction factor that keeps HIV from infecting cells from many monkey species and is thought to inhibit HIV infection in humans as well. Last year, Katzourakis and colleagues reconstructed the capsid protein from the ancient pSIV found in the genome of the grey mouse lemur and showed that TRIMCyp, a restriction factor related to human Trim5α that is found in present-day owl monkeys and macaques, can interact with this ancient pSIV capsid and keep it from infecting cells (2).

Trim5 was also the topic of a talk by Jeremy Luban, a professor of microbiology and molecular medicine at the University of Geneva. Luban presented evidence that suggests that Trim5 is the first known pattern recognition receptor (PRR) specific for retroviruses (3). Another well-known type of PRR, toll-like receptors (TLRs), recognizes bacterial structures, but so far retrovirus-specific PRRs were not known.

Luban thought that there must be a receptor that recognizes retroviral structures in part because of the structure of the HIV capsid (see image below). “I thought, my god, this thing looks scary,” Luban said. “If I had this in my cytoplasm I would be upset.” He also realized that Trim5 is similar to other PRRs, such as TLRs, because it only recognizes proteins when they are arranged in a certain pattern, in this case the hexagonally arranged CA proteins in the HIV capsid (4; see image below).


HIV Capsid Structure 

For over a decade, the laboratory of Mark Yeager has examined the structure of retroviral CA proteins, which assemble as hexamers and pentamers to form the capsid shell that encloses the RNA genome. In 2009, they solved the crystal structure of the HIV-1 CA hexamer, and earlier this year, Owen Pornillos and colleagues solved the crystal structure of the CA pentamer (5). This allowed them to build an atomic model of the HIV capsid shell, which follows the geometric pattern of a fullerene cage (yellow) assembled from about 250 CA hexamers (blue) packed in a curved hexagonal lattice. The conical shape is conferred by the insertion of exactly 12 CA pentamers (red), 7 at the wide end and 5 at the narrow end. Image courtesy of Owen Pornillos and Mark Yeager, University of Virginia School of Medicine.


Luban and colleagues found that when HIV interacts with cells that are important for eliciting the innate immune response, such as dendritic cells (DCs), Trim5 can activate inflammatory genes that are part of the antiviral innate immune response, and that this activation is increased by an interaction of Trim5 with the HIV capsid. Luban showed this to be the case for HIV interacting with monkey fibroblasts, and for retroviruses other than HIV interacting with human DCs, but not yet for HIV interacting with human DCs. “Our paper does not directly show that HIV-1 will trigger immune responses in human cells,” Luban said. “This is something we are attempting to do now.” In general, Luban said, his study is “the first specific handle of an innate immune receptor that detects retroviruses.”

The finding could lead to the development of more specific adjuvants specifically for HIV, Luban said. And with 68 Trim genes in the human genome, Trim5 is unlikely to remain the only example of a host cell HIV restriction factor that doubles as a PRR, he said. “This is a Manhattan project waiting to be done,” said Luban, adding that other restriction factors like tetherin or APOBEC3 might also act as PRRs by activating the innate immune response.

Antibody developments

Over the last two years, several research groups have reported the isolation of a slew of new broadly neutralizing antibodies (bNAbs) that are more potent than the handful of such antibodies that had been previously identified. Among them was VRC01, a bNAb isolated by researchers at the VRC, which can neutralize over 90% of currently circulating HIV-1 strains (see Raft of Results Energizes Researchers, IAVI Report, Sep.-Oct. 2009). VRC01 and other HIV-specific bNAbs have an unusually high degree of affinity maturation, which means that they are quite different from the germline version of the antibody that they are derived from (see Vaccines to Antibodies: Grow Up!, IAVI Report, July-Aug. 2010). Researchers are now trying to find VRC01-like antibodies from additional HIV-infected volunteers to see how common they are and how these antibodies develop. This work could help them understand how to elicit such antibodies with a vaccine.

Peter Kwong, chief of the structural biology section at the VRC, reported the structure of two additional VRC01-like antibodies, VRC03 (which was isolated from the same HIV-infected donor as VRC01) and VRCPG04 (which was isolated from a different donor from IAVI’s cohort of chronically HIV-infected individuals), bound to gp120. This structural work revealed that both VRC03 and VRCPG04 recognize gp120 in a very similar way to VRC01, even though the amino acid sequence of the variable domains of the three antibodies differs by up to about 50%.

All three VRC01-like antibodies were isolated using a probe that bound to them. Now Kwong and colleagues have developed a less expensive deep sequencing strategy that uses 454 next generation sequencing technology to find VRC01-like antibodies by determining the sequences of the antibody genes found in B cells. “From 10 [milliliters] of blood from any person you have about one million B cells, so you potentially have one million new sequences,” Kwong said.

So far, Kwong, together with the VRC’s John Mascola and their colleagues, have used the deep sequencing approach in the VRC01/03 donor and the VRCPG04 donor. To identify sequences of novel VRC01-like antibodies, they first looked for sequences with a high degree of affinity maturation that were somewhat similar to VRC01, and that were also derived from the same genomic precursor as the VRC01 antibody, an IgG heavy chain allele or gene variant called IGHV1-02*02. However, given that VRC01, VRC03, and VRCPG04 differ by up to about 50% in the sequence of their variable domain, looking for similar sequences alone would likely miss VRC01-like antibodies that differ in sequence, but not in structure.

To overcome this problem, researchers made the assumption that antibodies that have different sequences but similar structures probably share a similar precursor or intermediate along their affinity maturation process. To find such intermediates, they did a phylogenetic analysis of the sequences, using them to construct an evolutionary tree. They found that the different VRC01-like antibodies shared precursors or intermediate sequences along the affinity maturation pathway even if their sequences were not very similar. “The end results are 50% different, [but] the unexpected finding is that [the] maturation pathway for different VRC01[-like] antibodies from different people is similar,” Kwong said.

So far, the VRC researchers have synthesized approximately 80 different antibodies from VRC01-like sequences identified this way from the VRC01/03 and the VRCPG04 donors. Of those, 25 appear to be able to broadly neutralize a handful of HIV strains in preliminary experiments, Kwong said. He and colleagues have also started to use the deep sequencing approach alone to find VRC01-like antibodies in additional donors.

The intermediate stages along the affinity maturation pathway identified by the phylogenetic analysis may also help in developing immunogens that can induce such intermediates and direct the affinity maturation process in vaccinees toward fully affinity matured, VRC01-like antibodies, Kwong said. “With this information, we can now create one set of immunogens or one immunogen [that] should work not just for one unique person but for a population,” he said.

However, because VRC01-like antibodies are all derived from the Immunoglobulin (Ig)G heavy chain allele IGHV1-02*02, a person must be born with this variant to be able to produce VRC01-like antibodies, according to Kwong. The good news is that “it’s one of the most common alleles in the population,” he said.

While bNAbs like VRC01 are very potent, they may not always provide the best protection from HIV if they are used alone. Shannon Allen, a graduate student in Tom Hope’s laboratory at Northwestern University, showed movies from preliminary experiments in which she and her colleages added a mix of VRC01 bNAbs and fluorescently labeled HIV particles to cervical mucus taken from HIV-uninfected women. They found that when antibody was added, the movement of fluorescently labeled HIV particles in the mucus was slower than the HIV particle movement when no antibody was added. However, in cervical mucus taken from an HIV-infected woman, HIV movement was completely blocked, which may be due to a combination of many different antibodies in her mucus that don’t necessarily have to be neutralizing, Hope said. “Antibodies don’t work as monoclonals,” he said. “Antibodies work as swarms.” This could mean that a vaccine that protects by immobilizing HIV, making it harder for the virus to reach its target cells, may only have to induce a combination of many antibodies that aren’t necessarily broadly neutralizing.

Next, Hope wants to understand what it is in the mucus of an HIV-infected woman that is able to block the movement of HIV. “What kind of antibodies are there, is it [a] certain isotype, are there other factors there?” are the questions he hopes to answer.


HIV in Cervical Mucus

Section of the endocervix that was incubated with HIV-1 and stained with fluorescent markers. It shows a layer of cervical mucus (stained green) that lines the simple columnar epithelium of the endocervix (the nuclei of the columnar epithelium cells of the endocervix are stained in blue). Trapped within this cervical mucus layer is HIV-1 (stained red), although some virions are able to maneuver through the layer of mucus to reach the simple columnar epithelium. Image courtesy of Mike McRaven, Northwestern University.

Combining prevention strategies

The recent CAPRISA 004 trial showed for the first time that a microbicide gel containing 1% of the antiretroviral tenofovir (TDF) was able to reduce HIV incidence by 39% in a group of South African women. Prior to that, the results of the large RV144 trial in Thailand provided the first evidence that a prime-boost vaccine regimen can provide modest protection against HIV. Robin Shattock, a professor of mucosal infection and immunity at Imperial College in London and a co-organizer of the Protection from HIV track of this year’s Keystone meeting, made the point that partially effective prevention strategies such as these could be used in combination. “Rather than waiting for the vaccine that is 100% effective and only needs one shot, which we may never realize, [we need to] start filling that gap with things that have a degree of efficacy and see whether their combination improves that efficacy significantly,” Shattock said.

This could have several advantages. For example, microbicides might be able to turn a high-risk into a lower-risk population by reducing the HIV incidence, Shattock said, referring to concerns that the RV144 vaccine regimen might not work in high-risk populations. Microbicides might also protect vaccinees from a temporarily increased risk of HIV acquisition as was observed in the STEP trial, Shattock said. “[If] there was a window where a vaccine increased susceptibility based on say animal models, you could cover that window before a robust immune response was in place,” Shattock said.

Furthermore, microbicides might actually have effects beyond protecting vaccinees from infection. They might also lead to what Shattock calls “protected challenge” (in animals) or “protected exposure” (in humans)—the induction of virus-specific immune responses following a virus challenge or natural exposure to HIV that does not establish an infection. In 2008, Shattock’s group reported CD4+ and CD8+ T-cell immune responses to a rectally applied challenge virus in rhesus macaques, even though the macaques were completely protected against infection by a microbicide containing TDF (6). Shattock said that researchers are now trying to see if this kind of protected exposure effect can also be observed in humans, by looking for HIV-specific T-cell responses in volunteers who have been protected from HIV infection in the CAPRISA 004 trial.

The T-cell immune responses that Shattock observed in his 2008 microbicide study in macaques are likely the result of exposure to the virus for a very short time, enough to elicit an immune response, but not enough to establish infection, Shattock said. In high-risk individuals who become exposed to HIV but are protected by a microbicide, protected exposure might therefore result in a vaccine-like effect, giving the immune system time to mature its response, added Shattock.

In combination with a vaccine, a microbicide might also be able to boost or broaden the immune responses to the vaccine, a hypothesis Shattock is currently testing in nonhuman primates (NHPs).

Vaccine effects on HIV transmission

Analysis of transmitted founder viruses has become quite commonplace in recent years with the use of methods such as single genome amplification that make it possible to isolate and characterize single clones of viral variants (see Luck Favors the Prepared, IAVI Report, Sep.-Oct. 2010). James Whitney, an instructor in medicine at the Beth Israel Deaconess Medical Center and Harvard Medical School, reported results from what he said was the first study that looked at the effect of a candidate vaccine on the number of transmitted founder viruses in rhesus macaques, and showed that vaccination can reduce the number of transmitted founder viruses.

Whitney, in collaboration with researchers at the VRC and elsewhere, vaccinated rhesus macaques with SIVmac239 Gag, Pol, and Env immunogens delivered as three DNA injections followed by an adenovirus serotype 5 (Ad5) injection. About five months later, the animals received a repeat-low-dose rectal challenge with SIVmac251 or SIVsmE660. All animals were negative for a major histocompatibility (MHC) class I allele called Mamu A*01 that is associated with control of viremia.

All 20 animals challenged with SIVmac251 became infected, but showed a one to two log reduction of peak viremia compared with the unvaccinated control animals. In contrast, only 12 of the 25 vaccinated animals that were challenged with SIVsmE660 became infected after 12 repeat-low-dose rectal challenges, compared with 22 of the 25 unvaccinated control animals. In both groups, the vaccinated animals had fewer transmitted founder viruses. While this was only a trend in the animals challenged with SIVmac251, the reduction was significant in the animals challenged with SIVsmE660—at least eight of the 22 infected control animals had “significantly more” than one transmitted founder virus, whereas 11 of the 12 infected vaccinated animals had just one transmitted founder virus, Whitney said.

This shows that in principle, a vaccine can reduce the number of transmitted founder viruses, Whitney said. Given that most people become infected with a small number of HIV variants, he added, a vaccine might theoretically be able “to push that down to essentially zero.” He recommended that the number of transmitted founder viruses should be considered as an endpoint when evaluating candidate vaccines in NHPs in the future. “We may be missing an important correlate where a vaccine that may not appear to be overtly effective may significantly reduce the initial barrage of virus at the portal of entry,” Whitney said.

A different look at elite control

Studies of elite controllers (ECs), HIV-infected individuals whose viral loads remain at undetectable levels without ARV treatment, were another topic covered at the conference. To a large degree elite control is attributed to a difference in the CD8+ T cells of ECs (see Research Briefs, IAVI Report, May-June 2010). “The somewhat conventional and traditional view is that these patients are able to control HIV because they have strong [CD8+] T-cell mediated immunity against HIV,” said Mathias Lichterfeld, an instructor in medicine at Massachusetts General Hospital.

But Lichterfeld reported results from a study that showed for the first time that differences in CD4+ T cells, the major target cells of HIV infection, may also in part explain how ECs can control HIV replication so well (7).

Lichterfeld and colleagues isolated CD4+ T cells from several dozen ECs that are participants in the international HIV controllers study led by Bruce Walker and Florencia Pereyra at the Ragon Institute. They found that HIV replication in cultured CD4+ T cells from these ECs, in the absence of CD8+ T cells, was a lot less effective than HIV replication in CD4+ T cells from HIV-uninfected individuals, or from HIV-infected individuals with normal disease progression, so-called progressors. When the researchers investigated different stages of the HIV replication cycle, they found that all of them were inhibited in CD4+ T cells of ECs compared with CD4+ T cells from HIV-uninfected individuals or from progressors. For example, one week after infection, the CD4+ T cells from ECs produced 10-fold fewer HIV proteins. Transcription from HIV DNA integrated into the host cell’s genome was also inhibited, and as early as one to two days after infection, a time when the host cell reverse-transcribes the RNA genome of HIV into DNA that is then integrated into the genome, the CD4+ T cells from ECs had two- to five-fold fewer reverse transcripts than regular CD4+ T cells. This suggests that HIV replication could already be inhibited at the time of reverse transcription, Lichterfeld said.

None of the known HIV host cell restriction factors such as APOBEC3G or Trim5α seemed to be involved in this inhibition of HIV replication, but a protein called p21 that had been reported to inhibit retroviral gene transfer in gene therapy experiments did seem to play a role. Lichterfeld and colleagues found that p21 protein was expressed at about five-fold higher levels in CD4+ T cells taken from ECs than in cells from progressors or from uninfected people, and that eliminating p21 expression in CD4+ T cells led to increased HIV replication. Finally, they found that p21 inhibits a host cell protein called CDK9, which is required for transcription of HIV DNA that’s integrated into the host cell genome.

Next, Lichterfeld wants to find out why so much more p21 is expressed in CD4+ T cells from ECs. “If we understand the biological mechanisms that are contributing to this, we might be able to design some intervention that would enhance p21 and this could then translate into increased resistance to HIV,” he said.

Monkey STEPs

In 2007, vaccinations in the Phase IIb-test-of-concept STEP trial were halted after Merck’s Ad5-based candidate MRKAd5 failed to prevent transmission of HIV or slow disease progression in volunteers (see A Step Back?, IAVI Report, Sep.-Dec. 2007). Since then, researchers have mined the data collected from the 3,000 volunteers, but questions still remain as to why the vaccine failed to protect. One nagging question is whether the NHP model, which was used to test STEP-like vaccine regimens, is a reliable gatekeeper for the testing of future vaccine candidates.

To try and answer this question, Matthew Reynolds, a scientist in David Watkins’s lab at the University of Wisconsin-Madison, tested the efficacy of the MRKAd5 regimen in NHPs, which he noted during his presentation had never been done before. “There is a common misperception that the true regimen [used in STEP] had been tested extensively prior to human trials,” said Reynolds. “That’s simply not true.”

Reynolds said an SIV version of the precise vaccine used in the STEP trial—a replication-deficient Ad5 vector containing an equal ratio of the gagpol, and nef genes of SIVmac239—was administered intramuscularly three times over a 26-week period to eight Indian rhesus macaques. Upon completion of the vaccination regimen, the vaccinated monkeys and eight controls were then challenged with low-doses of SIVsmE660, a swarm virus that is heterologous to the vaccine immunogens.

Reynolds said all 16 animals became infected following challenge and there was no statistically significant difference in the rate of infection between vaccinated and control animals. “When we challenged the animals we failed to see a vaccine protective effect just like the STEP study,” said Reynolds. “This suggests the macaque model can be informative for human trials.”

The vaccinated animals required an average of 3.4 challenges before becoming infected and the controls required 3.8. There was also no significant difference in the viral loads between the two groups. “I think the real message here is that when you do the STEP trial in monkeys exactly as it was done in humans, you come to the same conclusion,” said Nicole Frahm, associate laboratory director at the HIV Vaccine Trials Network, which conducted the STEP trial.

Other viral vectors

Without a doubt, the RV144 trial brought increased attention to the idea of using canarypox viral vectors (seeSeeking Correlates of Protection for RV144, below). Follow-up trials to RV144 will include either ALVAC (Sanofi-Pasteur’s canarypox vector-based candidate) or NYVAC, another canarypox vector that has been tested as part of a prime-boost regimen in Phase II trials by EuroVACC.

A menu of other viral vector vaccine candidates is also moving through the pipeline, including several Ad vectors. One of these is an Ad26 vaccine candidate being tested in a prime-boost combination with a modified vaccinia Ankara (MVA) viral vector-based vaccine candidate. Last year, researchers presented results of a study in 40 rhesus macaques, conducted by the US Military HIV Research Program (MHRP) and the Integrated Preclinical/Clinical AIDS Vaccine Development program, which showed that the Ad26/MVA vaccine candidate provided better protection against SIV and better virologic control compared to three other viral vector regimens or placebo (See A Change of Tune, IAVI Report, Sep.-Oct. 2010).

At Keystone, Nelson Michael, director of the US Military HIV Research Program, presented unpublished results from a recently completed correlates analysis of that NHP study, which showed that the immune correlates associated with protection were different from those associated with virologic control. The primary analysis found that strong antibody binding activity to Env, as measured by ELISA, and the presence of Tier-1 neutralizing antibodies correlated with the number of challenges required to infect the vaccinated animals. The level of antibody-dependent cellular cytotoxicity responses induced in the vaccinated animals, while statistically significant, was not as closely correlated with protection, Michael said. In comparison, researchers found that the magnitude and breadth of Gag-specific responses by ELISPOT was significantly correlated with setpoint viral load in the vaccinated animals that became infected.

An exploratory analysis also showed a trend toward increased Gag-, Pol-, and Env-specific CD8+ effector memory T-cell responses in the vaccinated animals that were protected against infection, suggesting cellular immune responses might have also had a hand. Michael said they also identified nine additional humoral and cellular immune responses that significantly correlated with virologic control. “It ain’t as easy as antibodies control acquisition and [T cells] control viral load,” said Michael. “Probably both are important.”

A clinical trial of the Ad26/MVA prime-boost regimen is slated to begin within a year. The candidates are being formulated to encode mosaic antigens, which are computationally designed to achieve optimal coverage of the many different versions of HIV circulating globally. Michael said the trial will enroll 90 volunteers and be conducted in the US, Uganda, Tanzania, Kenya, and Thailand.


Seeking Correlates of Protection for RV144

Even when biomedical interventions against HIV work in clinical trials, the reasons behind their success are not always clear. Scientists still do not know why the vaccine candidate tested in the RV144 trial in Thailand—Sanofi-Pasteur’s canarypox vector-based candidate ALVAC-HIV (vCP1521) and AIDSVAX B/E, the genetically engineered version of HIV gp120 developed by VAXGEN—resulted in a modest but statistically significant efficacy of 31% (see Raft of Results Energizes Researchers, IAVI Report, Sep.-Oct. 2009). And efforts to define correlates of protection in RV144 could be stymied by the relatively sparse supply of samples that were collected over the course of the six-year, 16,000-person trial, as well as the limitations of the assays that are used to measure immune responses.

Researchers are hoping to gain some clarity when a case-controlled study of 41 HIV-infected vaccinees and a control group of 240 uninfected vaccinated individuals from the RV144 trial gets underway soon. The correlates analysis will be measuring the quantity and breadth of humoral and cellular immune responses and is expected to be done by early August, said Nelson Michael, director of the US Military HIV Research Program, a key collaborator on the RV144 trial. He said researchers will present results from this case-controlled study at the AIDS Vaccine Conference in Thailand in September. —R.M.

Intermittent PrEP

While most of the clinical and pre-clinical research of oral PrEP has been focused on once-daily regimens, data surrounding intermittent use of PrEP (iPrEP), is less abundant, and scientists are still not sure which iPrEP regimens or dosing schedules are effective in preventing HIV transmission.

Researchers from the US Centers for Disease Control and Prevention (CDC) reported last year that a single dose of Truvada—a combination of TDF and emtricitabine (FTC)—before or after challenge worked as well as daily dosing at protecting rhesus macaques against a low-dose, repeat simian/human immunodeficiency virus (SHIV) challenge (see Prevent and Conquer, IAVI Report, Jan.-Feb. 2010). They also reported that a single dose of GS-7340, a next-generation, pro-drug of tenofovir, given three days before rectal SHIV challenge failed to protect rhesus macaques.

Researchers were surprised by this result because GS-7340 was designed to deliver TFV-diphosphate (TFV-DP)—the active form of TDF—more efficiently to lymphoid cells and tissues than TDF. CDC researcher Walid Heneine suggested elevated levels of deoxyadenosine triphosphate (dATP) may be the reason why a single dose of GS-7340 did not work. dATP is a naturally occurring purine compound that competes with TFV-DP.

Heneine, and CDC colleague Gerardo Garc’a-Lerma, measured rectal and systemic TFV-DP exposure in lymphocytes from blood, lymphoid, and rectal tissues in four macaques that received a single dose of oral GS-7340. TFV-DP was also measured in peripheral blood mononuclear cells (PBMCs) of four SHIV-infected macaques and two uninfected macaques. Levels of dATP were also measured in PBMCs, lymphoid tissue, and rectal mononuclear cells.

Three days after dosing with GS-7340, the TFV-DP concentrations in PBMCs were 50-fold higher than those achieved with oral TDF, and they remained high for up to seven days. TFV-DP also accumulated in lymphoid and rectal tissues. But concentrations of dATP and dATP/TFV-DP ratios were 100-fold higher in rectal lymphocytes than in PBMC or lymphoid tissue, suggesting that TFV-DP may have less ability to block replication and infection at the rectal mucosa.

Additionally, dATP levels and dATP/TFV-DP ratios in PBMCs were higher in animals that became infected during the study than in those that remained uninfected. Heneine said their findings suggest that higher TFV-DP concentrations may be required to protect against rectal HIV exposure, leading some researchers to question whether it might make sense to look at dATP levels in future trials to see if its expression correlates with infection despite PrEP use.

The genital tract

HIV is primarily a mucosal infection and identifying factors that increase susceptibility to sexual transmission is important. Jo-Ann Passmore, an associate member of the Institute of Infectious Disease and Molecular Medicine at the University of Cape Town, presented results of an analysis that looked at cytokine concentrations in the genital tract before and after HIV infection in 22 women from the CAPRISA 002 acute infection study, a natural history study of clade C infection in heterosexual women, and 27 HIV-infected women from other CAPRISA cohorts in South Africa.

Passmore said previous studies have shown that inflammation in the genital tract during early HIV infection is associated with lower CD4+ T-cell counts. “We wanted to see if genital tract inflammation during acute infection played a role in disease progression and HIV shedding from the genital tract,” she said.

To do this, Passmore and colleagues measured concentrations of inflammatory and regulatory cytokines in samples that were collected by cervicovaginal lavage (CVL) from women with acute and chronic HIV infection using Luminex and ELISA. They then compared the cytokine concentrations to those from a matched set of pre-infection samples.

They also looked at T-cell activation as measured by expression of CD38, CCR5, HLA-DR (which is expressed primarily on B cells and presents proteins for recognition by CD4+ T-cell receptors), and T-cell proliferation, as measured by Ki67 expression, in both cervical fluids and blood samples.

The analysis showed that cytokine concentrations from women with acute HIV-infection did not differ significantly from their pre-infection levels. However, concentrations of a dozen cytokines that were elevated prior to HIV infection were significantly associated with lower CD4+ T-cell counts a year following infection, and a subset of the 12 cytokines were associated with higher setpoint viral load. During the first year of infection, intermittent shedding of HIV RNA into genital secretions was significantly associated with elevated genital tract inflammation.

Passmore also touched on results of a separate analysis that looked at whether the 1% tenofovir gel tested in the recent CAPRISA 004 trial was associated with any changes in genital inflammation that might have increased susceptibility to HIV in a subset of 885 HIV-uninfected women from the trial.

Passmore said the sub-analysis did not find any association between tenofovir gel use and changes in genital tract cytokine concentrations that might have increased susceptibility to HIV. The study did show, however, that women who became HIV-infected during the course of the trial had higher levels of genital tract inflammation prior to HIV infection than women who did not become HIV-infected, suggesting that high levels of inflammation in the genital tract may increase susceptibility to HIV infection.


Research Update on Exposed Seronegatives  

One of the longstanding unsolved mysteries baffling AIDS scientists is why some individuals who are exposed repeatedly to HIV remain uninfected (see Individual Armor Against HIV, IAVI Report, July-Aug. 2008). Data presented at Keystone from two recent studies funded by the Center for HIV/AIDS Immunology (CHAVI) and conducted in two high-risk populations offer fresh insights into both the immunological and genetic factors that might be contributing to this phenomenon. Initial findings from the first study, known as CHAVI 002, looked at T-cell responses in HIV-exposed seronegatives (HESNs). The study, which was recently published, also tried to address some of the overriding concerns regarding how to define and recruit HESNs for studies (8).

Adam Ritchie, a scientist at the University of Oxford’s Weatherall Institute of Molecular Medicine, said the laboratory team performed a blinded comparison of T-cell response rates to HIV among 24 HESNs from an HIV-serodiscordant couples cohort in London and compared the response rates to those of a matched group of 28 HIV-unexposed seronegative (HUSN) controls. Exposure to HIV and risk were quantified through detailed sexual behavior questionnaires.

To meet the study definition of HESNs, participants must have reported unprotected anal, vaginal, or oral sex with a known HIV-infected partner on at least 25 occasions in the previous year. Couples who reported monogamous behavior a year prior to the start of the trial and tested negative for both HIV and sexually transmitted diseases were considered to be HUSNs. Blood samples were collected every three months over the course of the year. Twenty-four of the 28 HUSNs and 17 of the 24 HESNs completed all four visits.

No T-cell responses were regularly detected in blood samples from either the HESNs or the HIV-unexposed seronegatives (HUSNs) by IFN-γ ELISPOT assay. But a more sensitive ELISPOT assay where researchers culture the PBMCs with pools of HIV peptides for 10 days before washing and resting the cells for 10 hours, turned up T-cell responses in both HESNs and HUSNs, with a much higher frequency and magnitude of T-cell responses in HESNs over time. Ritchie said it may seem surprising to see HIV-specific T-cell responses in HUSNs, but that the responses seen in HUSNs and some HESNs most likely reflect low frequency T cells that cross-react with HIV and were previously induced by exposure to environmental antigens.

“The significant differences between HESNs and HUSNs confirms that HIV-specific T-cell responses occur in exposed individuals, but whether this is part of the protective phenotype or a marker of exposure remains to be elucidated,” he said.

Findings were also presented from a CHAVI 016 study of hemophiliacs who were exposed to HIV but remained uninfected despite receiving Factor VIII concentrates derived from large pools of blood plasma collected from donors, some of whom were infected with HIV. The study, which began three years ago, set out to identify any key genetic determinants that might explain the apparent resistance of these ESNs to HIV. David Goldstein, a Duke University immunologist heading up CHAVI 016, said whole genome sequencing failed to find any common genetic variants associated with HIV resistance among the 393 HIV-uninfected cases. “Whatever is protective here, it’s pretty unlikely to be a common one,” said Goldstein.

Goldstein said his lab is now in the hunt for much more rare genetic variants that may explain HIV resistance among this cohort of hemophiliacs. Capitalizing on evolving sequencing technology that makes the process faster and more precise, Goldstein’s lab has identified close to 1,000 rare variants of interest, whittled down from a list of 42,000 that might have contributed to protection against HIV acquisition in certain hemophiliacs.

Goldstein acknowledges that this approach is a “needle-in-a-haystack” exploration, but said he does not believe the fact that some hemophiliacs did not become HIV-infected after being exposed to contaminated blood products was merely a matter of luck. —RM

1. Proc. Natl. Acad. Sci. 105, 20362, 2008 
2. Cell Host Microbe 8, 248, 2010
3. Nature 472, 361, 2011
4. Proc. Natl. Acad. Sci. 108, 534, 2011
5. Nature 469, 424, 2011
6. PLoS Med. 5, e157, 2008
7. J. Clin. Invest. 121, 1549, 2011
8. J. Virol. 85, 3507, 2011