A Slew of Science In Seattle

CROI served up a snapshot of HIV’s Envelope trimer and the latest on PrEP and protective immune responses

By Regina McEnery and Richard Jefferys

As the global campaign against AIDS enters its fourth decade, the development of a broadly preventive HIV vaccine remains among its most vexing challenges.

In his opening remarks at the 19th Conference on Retroviruses and Opportunistic Infections (CROI), held March 5-8 in Seattle, Tufts University virologist John Coffin, the scientific organizer of the event, noted that the failure of previous vaccine candidates had convinced many scientists that antibodies capable of preventing HIV infection could not be elicited through vaccination.

But the 31.2% efficacy demonstrated in the RV144 vaccine trial in Thailand—which, though modest, provided the first evidence of vaccine-induced protection from HIV—has helped lift the gloom from such speculations. Volunteers in that trial who received the prime-boost vaccine combination developed low titers of gp120-binding antibodies that subsequent analyses revealed are correlated with the risk of HIV infection. Separately, a dramatic expansion in the number of broadly neutralizing antibodies (bNAbs) against HIV isolated from volunteers, and the data describing their mechanisms of action, have renewed optimism about the prospects of this vaccine strategy.

“We’re now thinking much more seriously about developing vaccines that might be based on eliciting specific antibodies,” Coffin told the international gathering of more than 4,200 HIV researchers and clinicians.

Reflecting this shift in strategy, CROI organizers selected pioneering antibody researcher Dennis Burton, professor of immunology and microbial science and director of IAVI’s Neutralizing Antibody Center (NAC) at The Scripps Research Institute in La Jolla, California, to kick off the conference. Burton observed that the recent discovery of more than two dozen potent bNAbs by his lab and others, and the elucidation of some of their structural targets on HIV’s Envelope glycoprotein, have revealed weaknesses that can be exploited for both drug and vaccine development. “The tools are all there,” said Burton. “It remains to be seen if immunogen design can take advantage of all these tools.”

But antibodies were far from the only item on CROI’s four-day agenda. The conference also highlighted investigations of the structure of HIV’s Envelope trimer and updates on the continuing analysis of samples collected in the RV144 trial. Other talks of particular interest covered new findings on how a subset of T cells influences antibody development, the evolutionary pathways of HIV and SIV, and the results of several recent studies on ARV-based prevention, which dominated the conference and provoked more than a few animated discussions.

Freezing the fidgety trimer

All known bNAbs against HIV target functional Envelope trimers that—rather sparsely—adorn the viral surface. Many researchers believe that obtaining the detailed molecular structure of this protein complex would bring us much closer to the design of vaccines that induce sterilizing immunity to HIV, the ultimate goal of the AIDS vaccine field.

But that objective is easier defined than attained. The spike—composed of three identical pairs of the extracellular gp120 and transmembrane gp41 proteins—is highly unstable and structurally dynamic. While a structure of the HIV Envelope glycoprotein gp120 bound to CD4 was obtained more than a decade ago through X-ray crystallography (1), scientists haven’t yet obtained a high-resolution structure of the unliganded trimer.

In a riveting plenary talk describing his use of cryo-electron microscopy (cryo-EM) to probe the trimer’s structure, Joseph Sodroski, professor of microbiology and immunology at the Dana-Farber Cancer Institute, Harvard Medical School, compared the Envelope protein in its unliganded state to a mousetrap that has been set but not sprung. Once it binds its receptors on the surface of the T cell, the trap is sprung. At that point its gp120 and gp41 components each assumes a “preferred, lower energy” conformation. “The difference in energy between starting and final conformation,” Sodroski explained, “is used to drive the fusion of viral and target cell membranes, and complete the cell entry process.”

But the only low-resolution images available are of the unliganded HIV Envelope trimer and a brief subsequent state, assumed just before viral entry, known as the prehairpin intermediate. Sodroski said the main problem is that researchers haven’t yet been able to generate crystals of the functional Envelope protein that are amenable to X-ray crystallography. The resolutions achieved using electron tomography, meanwhile, are too low to reveal the details essential to protein engineering.

So Sodroski’s lab turned about three years ago to single-particle analysis of the HIV trimer using cryo-EM. This approach, said Sodroski, permits near atomic-level resolution of molecular structures, allowing researchers to generate reconstructions of molecular assemblies that resist crystallization.

Sodroski and colleagues first expressed the Envelope glycoproteins obtained from a primary HIV-1 strain in JR-FL cells, and then fragmented their membranes (which carried the intact trimer) with detergent to release the soluble proteins. They then purified and froze the molecules, spread them out in thin layers and obtained multiple micrographs of the particles from a variety of angles. To improve the resolution and reduce background noise, they averaged images of particles in similar orientations taken from multiple angles. They then aggregated the results to generate a structure of the trimer.

Sodroski said close to a million single-particle images were used to create the 3D image of the Envelope trimer at nearly atomic-level resolution. The resulting single-particle cryo-EM structure agreed with a structure of the native Envelope protein trimer constructed by cryo-electron tomography in the laboratory of Sriram Subramaniam, chief of the Biophysics Section in the Division of Cell Biology at the US National Cancer Institute (see The Beauty Behind the Beasts, IAVI Report, Nov.-Dec. 2009).

A striking architectural feature of the HIV Envelope trimer in its unliganded state, Sodroski said, is the “doughnut hole” at its center. This conformation, he noted, is distinctly different from the densely packed structure that emerges at the end stage of the viral entry process. The images obtained from cryo-EM show that the gp41 protein is kept from its more energetically favorable conformation by the inner domain of gp120, which acts like a clamp holding the membrane-distal domains of the former far apart. Sodroski pointed out that the unliganded gp120 protein also has an open structure. The only parts of that molecule that interact with each other are the trimer association domains—the V1, V2 and V3 loops.

Sodroski said that the architecture of the unliganded HIV Envelope glycoprotein suggests that broadly neutralizing antibodies must not only recognize conserved epitopes, but approach them from appropriate angles as well. He noted that more detailed reconstructions of the trimer structure could aid the design of immunogens that can elicit similar antibodies.

A Boost in their Prime

After decades of stagnation through much of the 20th century, research into adjuvants—substances that boost immune responses to vaccines—is enjoying a boom of sorts, fueled in part by improved understanding of the interplay between the innate and adaptive immune responses. At CROI, a team from the Atlanta biotech GeoVax offered up a plum example of this phenomenon. GeoVax has devised an immunization regimen in which the gene for GM-CSF, a cytokine produced by macrophages, neutrophils, and a variety of other immune cells, is co-expressed as an adjuvant with a SIVmac239 DNA prime. They have previously shown that, following such priming, a boost with SIVmac239 MVA prevents acquisition of simian immunodeficiency virus (SIV) by 70% of macaques subsequently challenged a dozen times with SIVsmE660. In their latest study, the team examined the durability of that effect. Seven of the macaques who did not acquire SIV following rectal challenge in the initial study—two who did not receive the GM-CSF adjuvant and five who did—were boosted with only the MVA vaccine candidate. They were then re-challenged six months later with 12 weekly doses of SIVsmE660. GeoVax’s chief scientific officer Harriet Robinson reported that six macaques remained uninfected after the challenge, while the seventh animal—from the adjuvanted group—acquired SIV only after the 10th challenge. The late MVA boost, she said, increased CD4+ and CD8+ T-cell responses to levels similar to early peak responses, and boosted antibody responses to the SIV envelope protein to four times the levels observed in earlier peak responses. Robinson and her team suggest that if their DNA/MVA prime-boost regimen includes a GM-CSF adjuvant, a yearly MVA boost could provide long-lasting protection against repeated rectal challenge. GeoVax is now testing the safety and immunogenicity of a GM-CSF adjuvanted DNA/MVA HIV vaccine candidate in a Phase I trial (HVTN 094) that opened for enrollment in April. —RM

 

Of sieves and the sifting of samples

The continuing correlates analysis of samples collected in the RV144 trial provided some grist for discussion as well. Paul Edlefsen, a biostatistician at the Public Health Sciences and Vaccine and Infectious Disease Division of the Fred Hutchinson Cancer Research Center in Seattle, added a little texture to the primary analysis of RV144 samples. That analysis suggested that the presence of non-neutralizing IgG antibodies that bind to the V1/V2 loops of HIV Env correlated with a 43% reduction in HIV infection risk in vaccine recipients (see A Bangkok Surprise, IAVI Report, Sep.-Oct. 2011).

Last year, his team, in collaboration with teams from the US Military HIV Research Program (MHRP) and the University of Washington, presented an analysis of breakthrough viruses—those that emerge in individuals who are not protected by vaccination— from 110 infected recipients across the vaccine and placebo arms of the RV144 trial. Their analysis compared nearly 1,000 amino acid sequences from the envelope proteins of Clade E HIV-1, the dominant subtype contained in the AIDSVAX B/E gp120 boost, from 50 vaccinees and 71 placebo recipients in the trial. These studies, led by MHRP virologists Morgane Rolland and Sodsai Tovanabutra and University of Washington virologist Jim Mullins, found viral escape (known to vaccinologists as a “sieve effect”) to be associated with the V2 and C1 region of the Envelope protein.

This is of particular interest because C1 corresponds to a common antibody-dependent cellular cytoxicity (ADCC) epitope, and ADCC is thought to have played a role in the protection induced by the RV144 regimen. The sieve effect appeared to be associated with a region between amino acids 69-95 in the C1 region. They identified two amino acid positions that appear to help determine whether a given clade E HIV variant escaped the nominal protection afforded by the RV144 vaccine regimen. The amino acids at positions 169 and 181 that correlated with such escape correspond to the crown of the V2 loop and the third amino acid of the α4ß7 binding motif, respectively (see A Bangkok Surprise, IAVI Report, Sep.-Oct. 2011).

Edlefsen shared this year the results of a linear peptide microarray assay—developed by Duke University scientist David Montefiori—that his laboratory used to conduct its secondary correlates analysis of RV144 samples. The assay was designed to detect binding of antibody in the vaccinees’ sera to linear epitopes representing all clades of HIV-1. This showed that people who have antibodies binding to the crown of the V2 loop were less likely to become infected, confirming results of the primary correlates analysis. The detection of IgG binding to linear peptides, he argued, strengthens the relevance of V2 antibody responses observed in the initial correlates studies.

Other findings have similarly boosted the starring role of the V2 loop in RV144. Edlefsen discussed, for example, two novel human monoclonal antibodies (mAbs), CH58 and CH59, retrieved from blood samples collected from a vaccine recipient after the fourth vaccination in the RV144 regimen. The mAbs—cloned by the lab of Duke University scientist Barton Haynes, who led the scientific steering committee that oversaw the search for RV144 correlates—are only weak neutralizers of HIV, unlike the PG9 bNAb, which they resemble. They do, however, bind the crown of the V2 loop implicated by sieve analysis. Interestingly, a substitution of the amino acid alanine at position 169 of the V2 loop abrogates this binding, but a similar substitution at position 181 does not. Exactly why that may be remains unclear.

 

T Follicular Helper Cells in Action

In the red germinal center (stained for cytidine deaminase, a marker for germinal center B cells), T follicular helper cells (stained green for CD3) can be seen interacting with B cells. An outer ring of re-circulating follicular B cells (stained blue for IgD), known as follicular B-cell mantles, surround the germinal center. Other T-cell zones are stained green as well. Image courtesy of Shimpei Kawamoto and Carola Vinuesa.

HIV and the elusive follicular helper

 

Immunologists have long been aware that CD4+ T cells—which are T-helper cells—play a vital role in the antibody response. One of those roles is the delivery of signals that drive the maturation and selection of B cells that generate increasingly potent antibodies. This process, called affinity maturation, is critical to the evolution of bNAbs to HIV.

But only in the last decade has it become clear that there exists a specialized subset of CD4+ T cells that is dedicated to such maturation. Carola Vinuesa, Humoral Immunity & Autoimmunity Group Leader at the John Curtin School of Medical Research, Australian National University in Canberra, spoke to the CROI audience about this subset—T follicular helper (Tfh) cells. Part of the reason Tfh cells eluded identification for so long is that they hang around in the B-cell follicles of lymph nodes and tend to be poorly represented in the circulating population of immune cells.

Vinuesa explained that the discovery of Tfh was facilitated by the identification of the chemokine receptor CXCR5 in the mid-1990s (2). This receptor was shown to be required for lymphocyte migration into lymph node follicles and the formation of germinal centers, where B-cell proliferation and affinity maturation of antibodies occur. Subsequent work by several research groups revealed that high levels of CXCR5 expression defined a subset of CD4+ T cells that home in on lymph node follicles and provide help to B cells (3; 4).

The elevation of these Tfh cells to the rank of an independent subset was initially boosted by data revealing that their gene expression patterns are distinct from that of better-known Th1, Th2 and T-regulatory groups of the CD4+ T-cell clan (5). It was subsequently cemented by the identification of the Bcl-6 gene as the master regulator of their differentiation (6). If any doubt lingered about the legitimacy of the promotion, it was dispelled by the finding that, in addition to CXCR5, Tfh cells are distinguished by the expression of high levels of the PD-1 molecule and that they primarily secrete the cytokine IL-21.

Vinuesa highlighted two types of interactions that occur between Tfh and B cells: a brief dalliance at the border of the lymph node follicle and germinal center that leads to the generation of short-lived antibody-producing plasma cells and promotes class-switching of antibodies—from the IgM class to mature IgG, IgA or IgE classes—and longer, secondary liaisons in the germinal center that facilitate affinity maturation and the development of long-lived memory B cells. The secondary interface is crucial to the induction of lasting memory responses by vaccines, but Vinuesa noted that research into the impact of different vaccine approaches on Tfh cell and Tfh/B cell interactions is still in its infancy. The finding that bNAbs are characterized by extensive somatic hypermutation—the genetic changes in B cells that occur during affinity maturation (see Vaccines to Antibodies: Grow Up!, IAVI Report, July-Aug. 2010)—suggests that a better understanding of Tfh cells could contribute significantly to HIV vaccine development.

There followed other intriguing presentations on Tfh cells. Madelene Lindqvist, a postdoctoral researcher in the laboratory of Ragon Institute immunologist Hendrik Streeck, presented the first data on Tfh cells ever collected from individuals with chronic HIV infection. Her study involved 16 individuals with untreated infections (a median CD4+ T-cell count of 444 and a viral load of 35,000 copies/ml), 10 on antiretroviral therapy (median CD4+ T-cell count of 573 and viral load of 49 copies/ml) and seven HIV-uninfected controls. Lindqvist first demonstrated that Tfh cells with the profile outlined by Vinuesa (displaying CXCR5 and PD-1, expressing Bcl-6 and producing IL-21) were present in lymph node samples from participants at a frequency that was 100-fold higher than that measured in peripheral blood.

Comparing the different study cohorts, Lindqvist found that Tfh cells were significantly expanded in untreated HIV infection compared to controls. Gag-specific Tfh cells were found to be a component of this expansion, as the magnitude of such responses was two-fold higher in untreated vs. treated infection. Tfh responses to gp120 were also measured, but were around five times lower than those to Gag. Additional analyses revealed correlations between Tfh expansion and the B-cell dysregulation that has been described in progressive HIV infection (7). Specifically, increased Tfh numbers in untreated HIV infection correlated with an increase in plasma cells and loss of memory B cells in lymph nodes, along with the hypersecretion of IgG antibodies (hypergammaglobulinemia). Lindqvist noted that B-cell dysregulation in HIV had been thought to be a consequence of compromised CD4+ T-cell help. Her data indicate, however, that it is driven by an abnormal expansion of Tfh cells.

 

A Tip of the Hat to Cure Research

There was a time when the notion that HIV infection might be cured was considered quixotic at best in scientific circles. That is no longer the case. One measure of how much things have changed was the first ever dedicated inclusion in this year’s CROI of a symposium covering advances in that quest. The crowded event featured overviews of the state of the science, and much of the same material covered in the recent HIV Persistence Workshop in St. Maarten (see In Pursuit of a Cure, IAVI Report, Jan.-Feb. 2012), including a study by David Margolis, director of the University of North Carolina School of Medicine, that evaluated the activity of a single dose of SAHA (vorinostat) in humans. Sharon Lewin, Professor and Director of Infectious Diseases, Alfred Hospital and Monash University in Melbourne, offered a glimpse at a similar trial she is currently leading. But unlike Margolis’ study, hers involves 14 days of vorinostat administration. Ten of the desired 20 participants have so far enrolled in her trial. They have a median CD4+ T-cell count of 710, and average around seven years on suppressive ART. Due to the repeat dosing, grade 1 and 2 adverse events have been common, including lethargy, nausea, vomiting, diarrhea, thrombocytopenia (decreased platelet counts) and increased levels of the enzyme alkaline phosphatase. Median onset of side effects was 2-3 days into the dosing period. Those side effects, however, did not persist once the participants completed the necessary 14 days of treatment. No evidence of T-cell activation has been observed. Data measuring the effects of the regimen on latent HIV reservoirs are pending, and will likely be available next year. One of the unanswered questions in cure research has been whether induction of HIV RNA expression in latently infected CD4+ T cells will suffice to induce cell death. At CROI, Liang Shan, a graduate student in the laboratory of Robert Siliciano at Johns Hopkins, offered a sobering answer: latently infected CD4+ T cells do not die after exposure to vorinostat, but require functional HIV-specific CD8+ T cells to deliver the coup de grace. Shan studied whether HIV-specific CD8+ T cells sampled from several different groups—naïve controls, HIV-infected individuals on suppressive ART and elite controllers—could kill latently infected CD4+ T cells treated with vorinostat in the laboratory. CD8+ T cells from three elite controllers went about this task with vigor, but only one out of eight individuals on ART showed similar activity, indicating that HIV-specific CD8+ T-cell dysfunction in chronic infection will need to be addressed if elimination of latently infected CD4+ T cells is to be achieved. Shan reported that stimulation of the CD8+ T cells with HIV antigens prior to mixing with the infected CD4+ T cells restored their lethality, suggesting that therapeutic HIV vaccines may be an important component of anti-latency strategies. Shan’s work was published shortly after his presentation (8). While one goal of cure research is to eradicate HIV, there is also interest in the possibility of a “functional cure,” defined as control of the virus and prevention of disease progression in the absence of ART. In support of that effort, Emmanouil Papasavvas, senior scientist in Luis Montaner’s laboratory at the Wistar Institute in Philadelphia, evaluated the effects of pegylated alpha interferon—a cytokine therapy used to treat hepatitis C—on HIV viral load following interruption of ART. A total of 20 participants on suppressive ART with CD4+ T-cell counts over 450 were recruited and, after eight weeks, randomized to receive one of two pegylated alpha interferon doses given weekly. After an additional five weeks, ART was interrupted for up to 24 weeks before being restarted, while pegylated alpha interferon was continued as monotherapy. The primary endpoint of the study was the proportion of participants with viral loads less than 400 copies/ml at week 12 following interruption of ART. Intriguingly, nine out of 20 participants maintained control of viral load below this level at the week 12 timepoint, a percentage far higher than that obtained in prior ART interruption studies. Papasavvas showed that this salutary outcome was associated with the ability of the participant’s natural killer (NK) cells to respond to alpha interferon signaling. Binding of alpha interferon to its cellular receptor typically leads to a cascade of signals that cause phosphorylation of a protein called STAT1, which in turn translocates to the nucleus and steps up transcription of alpha interferon-stimulated genes, driving antiviral responses. Papasavvas measured the ability of participant NK cells to phosphorylate STAT1 after exposure to alpha interferon in vitro and found that such activity correlated significantly with the control of viral load observed in participants (p=0.005). A similar association was observed for measures of NK cell cytotoxicity. The results suggest that alpha interferon signaling pathways can play a key role in the immunological control of HIV. Although pegylated alpha interferon has an infamous array of severe side effects—including depression, nausea, vomiting and neutropenia—likely to dampen enthusiasm about its use as a monotherapy, Papasavvas’s work points to novel mechanisms that may have the potential to be exploited by researchers pursuing a functional cure. —RJ

Learning from the living

Researchers are learning more about what drives protective immune responses to HIV-like retroviruses by studying, in nonhuman primates, the effects of live attenuated vaccines (LAVs) made from weakened strains of simian immunodeficiency virus (SIV). Such experimental vaccines provide macaques with the most robust protection obtained against SIV. Although preventive LAVs against HIV are generally considered too risky for use in humans, the field has placed a high priority on understanding why precisely they work so well in experimental models and applying that information to improve the design of HIV vaccines.

Yoshinori Fukazawa, a postdoctoral researcher at the Oregon Health & Science University’s Vaccine and Gene Therapy Institute, presented new data at CROI that described the strongest correlates of LAV protection obtained so far. Fukazawa’s experiment evaluated five different LAVs with six macaques in each vaccine group and six unimmunized controls. Challenge was performed intravenously with SIVmac239 at week 50 after LAV inoculation. The extent of protection achieved was diverse, with only SIVmac239Δnef and SIVmac239Δ3 offering complete protection, which Fukazawa primarily defined as no—or only transient—replication of the challenge virus, and no depletion of mucosal CD4+ T cells.

Eleven different immunological parameters were assessed as potential correlates of immunity, including ADCC, neutralizing antibodies and SIV-specific T-cell responses (both CD4+ and CD8+) in blood, lymph nodes and lung. Highly significant correlations with protection were only observed for SIV-specific CD4+ and CD8+ T-cell responses measured by intracellular cytokine staining (ICS) in the lymph nodes (p=0.0036 and p=0.0070, respectively), after Bonferroni corrected for multiple comparisons. The T cells displayed an effector memory phenotype. Using an assay designed to measure the ability of SIV-specific CD8+ T cells to suppress viral replication in autologous infected CD4+ T cells, Fukazawa also found differences between protected and non-protected animals. But these were only statistically significant for SIV-specific CD8+ T cells sampled from lymph nodes, not from blood.

The subject of Tfh cells cropped up again in Fukazawa’s talk, when he produced evidence suggesting that LAVs preferentially replicate in that population. Furthermore, the amount of LAV RNA in lymph nodes correlated significantly with the magnitude of the SIV-specific CD4+ and CD8+ T-cell responses at the site. Fukazawa concluded that ongoing LAV replication is probably essential to maintaining protective effector memory T-cell responses in lymph nodes. This suggests that persistent vaccine vectors will in all likelihood be required to induce and maintain this type of immunity in humans.

Molecular warfare

Researchers continue to study HIV’s ancestral tree to better understand the emergence of viral defense mechanisms. The tussle between viruses and hostile host environments leads to an “evolutionary arms race” that has shaped the evolution of HIV.

It has, for one thing, spurred the selection of several HIV-1 proteins to counteract the effects of defensive host cell proteins—restriction factors—that curtail viral replication. The latest restriction factor identified is a cellular protein named SAMHD1, which limits the ability of HIV-1 to replicate in dendritic cells (DCs) and monocytes by depleting the nucleotide components the virus needs to assemble new virions (9). Some members of the lentivirus family to which HIV belongs possess a protein, Vpx, which has the ability to degrade SAMHD1, allowing these viruses—including HIV-2 and its simian virus precursors—to infect dendritic cells and monocytes.

Olivier Fregoso, a postdoctoral fellow in the laboratory of Michael Emerman at the Fred Hutchinson Cancer Research Center in Seattle, described his efforts to ascertain whether the ability to degrade SAMHD1 is a function that HIV-1 has lost, or one that its viral relatives have gained. The question is not merely academic: understanding the emergence of adaptations can shed light on hidden vulnerabilities that emerge at various points in the viral life cycle.

Fregoso showed that the ability to degrade SAMHD1 was acquired initially by Vpr in a lineage of lentiviruses distinct from the one out of which HIV-1 evolved. The Vpr protein in this lineage subsequently evolved into Vpx, a protein specializing in the degradation of SAMHD1. HIV-1, it would appear, has not evolved to avoid infecting DCs and monocytes, as was suggested when SAMHD1 was first identified. Rather, Fregoso’s group believes that HIV-1’s greater pathogenicity compared to HIV-2 could be related to its inability to infect these cell types (10).

Two restriction factors were discovered prior to SAMHD1. One was APOBEC3G, a cellular protein with antiretroviral activity that is degraded by the HIV-1 protein Vif, the other, tetherin, which limits viral replication by tethering virions to the cell and preventing their dissemination. Tetherin is counteracted by the HIV-1 protein Vpu. APOBEC3G, meanwhile, belongs to a family of APOBEC3 proteins (designated A, B C, DE, F, G and H). Emerman showed evidence that the selection of a version of APOBEC3DE with increased antiviral activity occurred in response to a virus that infected chimpanzees around 2.5 million years ago (11). The human version lacks similar activity because the virus-driven change occurred after the evolutionary branches of humans and chimpanzees diverged.

Tetherin, too, provides a striking example of the impact of viral adaptation. In chimpanzee SIV (SIVcpz), the precursor to HIV-1, the antiviral effect of tetherin is blocked by the Nef protein. But the specific site on tetherin that SIVcpz Nef binds doesn’t exist in the human version of the protein. HIV-1 group M, the strain primarily responsible for the worldwide pandemic, overcame this obstacle by adapting and deploying the Vpu protein to subvert tetherin. The far rarer HIV-1 group O did not; and in group N viruses the anti-tetherin activity of Vpu is far less consistent. HIV-1 group M’s arms race against tetherin has, in a sense, helped shape human history (12).

Shades of HLA

Not all people respond to HIV infection in the same way. For one thing, some control the amount of virus in their bodies far better than others. Many genetic idiosyncrasies—in both the virus and the host—account for such differences. Srinika Ranasinghe, a postdoc in Hendrik Streeck’s lab at The Ragon Institute, reported one such factor: variations in a key set of host genes encoding the human leukocyte antigen (HLA). The Streeck lab has recently illuminated associations between the breadth of HIV-specific CD4+ T-cell responses and the control of viral load in chronic infection (13). It also has exposed the impact of cytotoxic activity of HIV-specific CD4+ T cells in acute infection on both set-point viral load and clinical outcomes (seeResearch Briefs, this issue, and 14). At CROI, Ranasinghe unveiled new data describing how various HLA alleles might influence CD4+ T-cell responses and, as a consequence, viral load. HIV peptides are presented to CD4+ T cells by class II HLA proteins, which are encoded by highly variable genes. Samples from 1,085 treatment-naive HIV-infected individuals with a mean viral load of 40,472 copies/ml were used for the study. All were of European ancestry and results were adjusted for the presence of class I HLA alleles known to influence viral load, such as B*57, B*27 and B*35px. Examples of alleles associated with lower viral load included HLA DRB1 1502 and HLA DRB1 1302, while HLA DRB1 1301 was linked to higher viral load. Further analysis revealed that HLA alleles associated with lower viral load were more versatile, presenting approximately 30 peptides from the Gag, Nef and Pol proteins. In contrast, the alleles possessed by people with higher viral loads could only present around 10 such antigens, a highly significant difference. Ranasinghe noted that this work represents the first evidence that HLA DRB1 alleles influence viral load at the population level. —RJ

Sticking to PrEP

As it did last year, and the year before, ARV-based prevention dominated discussion at CROI. In particular, attendees sought to mine the implications of fresh data from a number of pre-exposure prophylaxis (PrEP) studies that suggest PrEP may be significantly less feasible in some high-risk populations than had been hoped.

“The reality doesn’t necessarily match the vision,” said Jared Baeten, a University of Washington associate professor of global health and a co-investigator in the Partners PrEP study, which found high efficacy in serodiscordant couples. “We have four completed PrEP trials that demonstrate efficacy, but we have two trials in women with high incidence where the entire study or individual arms have demonstrated futility.”

One of them was FEM-PrEP, which enrolled nearly 2,000 high-risk heterosexual women from Kenya, South Africa, and Tanzania. The study was discontinued in March, 2011, after a data safety monitoring board determined it was unlikely to establish whether or not daily oral administration of Truvada—a combination of the ARVs tenofovir (TDF) and emtricitabine (FTC)—is effective in reducing HIV acquisition (see Vaccine Briefs, IAVI Report, Mar.-Apr. 2011).

Lut Van Damme, the trial’s principal investigator, presented a final analysis of the trial data that suggests inadequate adherence to the prescribed drug regimen may have undermined the trial, which tallied 33 infections in its Truvada arm and 35 in its placebo arm. Analyses of the blood plasma collected during the trial from the 33 women in the Truvada arm who acquired HIV and about 99 matched uninfected controls found detectable levels of TDF in fewer than half of the samples. Notably, this contrasted with the 86% adherence rate suggested by the weekly pill counts—the number of pills dispensed minus those returned—as well as the claims of 95% of the volunteers who said they had taken the drug diligently. Van Damme noted that these data raise questions about the value of pill counts in measuring adherence, not to mention what became of the pills that were neither taken nor returned.

Van Damme also reported that there were 74 pregnancies (11%) in the Truvada group compared to 51 (7.5%) in the placebo arm of the study, which means that women in the Truvada group spent more time off PrEP because it had to be stopped due to safety concerns. Van Damme noted that pregnancy rates in the trial were highest among oral contraceptive users, which, she suggested, reflected a “problem with daily pill taking.”

Follow-up studies of the Partners PrEP trial underscore the importance of adherence as well. That study, which enrolled 4,758 heterosexual serodiscordant couples in Kenya and Uganda, revealed in 2011 that a daily dose of TDF reduced the risk of HIV infection by 62%, and a similar regimen of Truvada cut that risk by as much as 73% in the study population. Recent findings presented by Deborah Donnell, principal investigator of the HIV Prevention Trials Network and a statistician in the Partners PrEP study, indicate that individuals who remained HIV uninfected in the TDF and Truvada arms had detectable levels of the drug on 83% and 81% of their study visits, respectively. Drug levels were much lower in the 29 participants in both arms of the trial who acquired HIV. Only about a third of them had detectable levels of TDF when investigators first identified HIV antibodies in their blood plasma. “Even in visits prior to seroconversion, these people were not taking their drug as often as those who did not seroconvert,” Donnell said.

Yet adherence does not always predict PrEP efficacy. In her sub-analysis, Donnell noted that nine of the HIV-infected participants who had been assigned to either the TDF or Truvada arms did have detectable levels of TDF. Eight of them had drug levels that were consistently high or detectable throughout the follow-up—an indication the drug was being taken faithfully. “Of course, we don’t know what the level of the drug was at the time they got HIV infected,” said Donnell.

Baeten said that while adherence is arguably the primary determinant of PrEP efficacy, there are important hypotheses to consider about how biologic factors may diminish protection. He said pharmacokinetic studies have found that oral TDF achieves 10-fold or higher concentrations in rectal tissue compared to vaginal tissue, suggesting that PrEP may be more sensitive to non-adherence in women whose primary risk exposure is through vaginal sex, compared with men who have sex with men. He said sexually transmitted infections, genital inflammation and other factors that might increase the risk of HIV acquisition among high-risk heterosexual women could be interacting with PrEP to decrease its efficacy. “Research needs to assess whether this hypothesis can be supported,” he said.

Richard Jefferys is Coordinator, Michael Palm Basic Science, Vaccines & Prevention Project at the Treatment Action Group.

1. Nature 393, 648, 1998
2. Cell 87, 1037, 1996
3. J. Exp. Med. 192, 1545, 2000
4. J. Exp. Med. 192, 1553, 2000
5. J. Immunol. 173, 68, 2004
6. Immunity 31, 457, 2009
7. Nat. Rev. Immunol. 9, 235, 2009
8. Immunity 36, 491, 2012
9. Nat. Immunol. 13, 223, 2012
10. Cell Host Microbe 11, 194, 2012
11. J. Virol. 85, 11361, 2011
12. J. Virol. 84, 7124, 2010
13. J. Virol. 86, 277, 2012
14. Sci. Transl. Med. 4, 123ra25, 2012