On the Scientific Trail in Santa Fe

A walk through some of the pioneering research presented at the annual Keystone Symposia on HIV Biology and Pathogenesis

By Regina McEnery

Ever since HIV was discovered, researchers have been probing the retrovirus’ life cycle. Some of their recent progress was highlighted during the annual Keystone Symposia on HIV Biology and Pathogenesis, which was held from January 12-17 in Santa Fe, New Mexico. Just an hour’s drive from the Los Alamos National Laboratory where the so-called Manhattan Project coordinated development of the first nuclear weapons, nearly 300 HIV scientists gathered to discuss research that they hope will one day detonate the virus.

An assortment of findings were presented that are shedding light on the drivers of immune activation, components and mechanisms of innate immunity, as well as the mechanisms of HIV transmission.

Some of these advances have been a long time in coming. A major highlight of the meeting was the unveiling of a crystal structure of an integrase protein from the recombinant prototype foamy virus (PFV), a nonpathogenic retrovirus (see Figure 1). Scientists attending the conference described the research as a tour de force. 


Figure 1: Integrase Finally Crystallizes 

Crystals of the prototype foamy virus (PFV) integrase-DNA complex that were used to determine its three-dimensional structure. Image provided by Peter Cherepanov at Imperial College London.

A shadowy enzyme

Researchers succeeded 21 years ago in building crystal structures of an HIV protease complex. They followed suit nine years later with HIV reverse transcriptase. It is the third canonical retroviral enzyme integrase that has been the proverbial black box scientists have been unable to decode. Although structures of several HIV integrase fragments have been determined, it was far from obvious how they could be assembled together and how the full-length protein engages viral and human DNAs. While crystallization of full-length HIV integrase remains an elusive target, Peter Cherepanov, a professor at Imperial College London, and his team took a giant step forward last month when their needle-in-a-haystack search produced a crystal structure of PFV integrase bound to its DNA ends, a complex known as the pre-integration complex (PIC), or intasome (see Figure 2). Formation of the PIC is the stage immediately before reverse transcriptase (RT) copies viral RNA genomes into double-stranded complementary DNA (cDNA), which then get integrated into the host’s DNA (1). Integration is the point of no return in HIV infection because once the viral genome is inserted into the host’s DNA you can’t get rid of it. 


Figure 2: Crystallized Integrase Protein 

An overview of the full-length structure of retroviral integrase from a prototype foamy virus, shown perpendicular to the crystallographic two-fold axis. Viral DNA strands are shown in brown. Subunits of the integrase tetramer engaged with viral DNA are blue and green. Image provided by Peter Cherepanov at Imperial College London.


Cherepanov says he will never forget the day his team identified the successful crystal structure at a resolution of three Angstrom. The team had to set up more than 40,000 crystallization trials, and most of the crystals they succeeded to grow were not of sufficient quality for structure determination. “The rest was quick,” said Cherepanov. “After years of preparatory work we solved the structure within two weeks.” Cherepanov was actually on the London Tube, heading back to his lab, when he got his first glimpse of the partially solved structure. “It was hard not to scream,” he said.

Cherepanov’s lab used PFV to grow the integrase crystal because it was considered a good proxy for HIV and because HIV integrase had a well-deserved reputation for “misbehaving.” The protein’s inherently poor solubility had made it impossible to concentrate HIV integrase to the degree required for crystal formation.

According to Cherepanov, retroviral integrase proteins are very similar. They share domain organization and even sequence homology. “Here, taking two similar proteins with identical function, we can be quite certain their overall structures will be very similar.” Furthermore, Cherepanov says that the amino acid sequence within the active sites of integrase in HIV and PFV is also similar.

When Cherepanov’s team soaked the PFV integrase crystal in solutions of the HIV integrase-inhibiting drugs raltegravir and elvitegravir, they were able to observe, for the first time, how these ARVs bind to and inactivate integrase. “Using the PVF structure as a template, it is now relatively straightforward to generate reliable models for the HIV intasome, which will help improve the design of drugs that target HIV integrase.”

On high alert

The detrimental effects of chronic immune activation, which is the result of HIV infection, have become increasingly clear (see Everything from Cause to Cure, IAVI Report, July-Aug. 2009). Less clear are the biological components that drive immune activation in HIV-infected individuals.

Daniel Douek, chief of the Human Immunology Section at the Vaccine Research Center at the National Institute of Allergy and Infectious Diseases (NIAID), has looked at whether elevated levels of biological products associated with microbial translocation—the leakage of endotoxins and other microbial products across the gastrointestinal barrier and into systemic circulation—is a key driver of immune activation, and therefore disease progression. In previous experiments, Douek’s team found elevated levels of bacterial lipopolysaccharide (LPS)—a component of Gram negative bacterial cell walls that stimulates toll-like receptors on macrophages and dendritic cells (DCs)—in HIV- infected individuals compared to HIV-uninfected individuals, as well as elevated levels of other bacterial markers such as 16S ribosomal RNA and soluble CD14 (sCD14; see Down, But Not Out, IAVI Report, March-April 2008). Douek’s team believes microbial translocation occurs because of HIV-inflicted damage to the epithelium and the immune component of the gut.

At Keystone, Douek presented data from 700 HIV-infected individuals that were part of the SMART study. In the SMART study, nearly 5,500 HIV-infected individuals with CD4+ T-cell counts of at least 350 were randomly assigned to either a viral suppression arm that offered continuous use of antiretroviral therapy for the duration of the trial, or a drug conservation arm, in which the administration of ARVs was deferred until the HIV-infected individual’s CD4+ T-cell counts dropped below 250, at which point ARVs were initiated and continued until their CD4+ T-cell counts increased to more than 350. The cycle of intermittent therapy continued in the drug conservation arm for the duration of the trial. The SMART study was halted in 2006 after data showed a significantly increased risk of disease or death from any cause in those undergoing intermittent treatment (2).

In the SMART study sub-analysis, Douek found that HIV-infected individuals were more likely to have elevated levels of LPS, 16S ribosomal RNA, and sCD14 when compared with a control group of unmatched HIV-uninfected individuals.

“We found plasma sCD14 levels were a predictor of death both in treated and untreated [HIV-infected] individuals, independent of other markers of inflammation and independent of viral load,” said Douek. “We believe these microbial products contribute to immune activation.”

His group also found higher levels of a fourth biomarker—intestinal fatty acid binding protein (I-FABP)—in plasma. I-FABP, which is generally confined to the lower intestine, is associated with a loss of epithelial cell integrity, usually in response to blood supply restriction or tissue injury. “We think I-FABP may be a nice way of measuring gut damage rather than through gut biopsies,” said Douek.

Douek said they were unable to compare whether the time of initiation of highly active antiretroviral therapy (HAART) impacted the levels of LPS, sCD14, or I-FABP. Nonetheless, he said the results underscore the fact that immune activation is undesirable. And although ARVs can successfully suppress viral replication, this may not be enough to eliminate immune activation. “When you reduce the virus to very, very low levels, you don’t fix the problem completely. You can have immune activation, all this evil going on, even when there is no [detectable] virus,” said Douek.

But Douek cautioned not to trivialize the role of the virus in this chain reaction of events. “None of this happens without the virus causing damage in the first place,” said Douek. “You don’t get AIDS from immune activation unless you have HIV.” 


New Antiviral Targets  

The keynote address at the Keystone Symposia on HIV Biology and Pathogenesis was given by Columbia University molecular biologist Steve Goff, who noted how astonishingly complex HIV is compared to other microorganisms. While HIV contains just nine genes that code for 15 proteins, Goff said the virus may exploit as many as 200 human proteins, known as HIV dependency factors (HDFs), to survive and replicate (3). “It’s a nightmare,” said Goff, calling attention to a slide that mapped the vast array of HDFs that conspire in a broad array of cellular functions that begin with the virus binding to CD4+ T cells, and end with the budding of a new virus.

To disrupt these cellular processes, Goff’s laboratory combs through gene fragments that can potentially interfere with HIV. But while the list of antiviral genes continues to grow, the big question is whether scientists will be able to harness their potential to either prevent or treat disease. —RM 


Aging on HAART

Another one of the perils of chronic immune activation in HIV-infected individuals, it seems, may be accelerated aging. Previous data indicate that HIV-infected individuals on long-term HAART are at greater risk of non-HIV related conditions than age-matched HIV-uninfected individuals. Many of these complications are similar to those observed in the aging population, and include cardiovascular, liver, and kidney disease, as well as osteoporosis and non-HIV related cancers. While antiretroviral drug toxicity is considered to be one reason for the early manifestation of some of these premature age-related conditions, researchers believe there are other factors, including immunologic dysfunction and inflammation that persists even during suppressive antiretroviral therapy, that are driving onset of age-related diseases in HIV-infected individuals.

Steven Deeks, a professor of medicine at the University of California in San Francisco, analyzed an array of immunological markers associated with aging or with cardiovascular disease, which occurs more frequently as people age, in 100 females from the Women’s Interagency HIV Study (WIHS) and 300 HIV-infected men and women from the SCOPE study, and compared them to age-matched uninfected individuals.

The WIHS analysis showed that levels of immunosenescent T cells, as measured by their lack of CD28 expression, was higher in untreated and treated HIV-infected volunteers, compared to uninfected controls. Activation and immunosenescence of T cells—which occurs naturally during aging—were both associated with vascular dysfunction.

The individuals analyzed from the SCOPE study confirmed many of these observations. Although the treated HIV-infected volunteers had higher levels of proliferating CD4+ T cells (as measured by Ki67 expression), the frequency of these cells able to divide ex vivo was low, as compared to the frequency of these cells in uninfected controls. These proliferative defects are also considered a common trait of an aging immune system.

The SCOPE data also showed that HIV-infected individuals on HAART had fewer naive CD8+ T cells, and more CD8+ CD28- T cells than age-matched HIV uninfected individuals, as well as elevated serum levels of C-reactive protein, a biomarker for aging that is usually triggered by infection or tissue injury.

Deeks also presented data from a recently published SCOPE-related analysis showing that among long-term ARV-treated individuals, the levels of cytomegalovirus (CMV)-specific T cells are dramatically expanded, with levels at least twice as high as that seen in age-matched uninfected individuals (4). CMV seropositivity is strongly associated with accelerated aging of the immune system. These cells have also been associated with heart disease in the HIV-infected population.

Natural born killers

A number of presentations at Keystone also explored the role of innate immunity in preventing and controlling HIV infection. Galit Alter, assistant professor of medicine at the Ragon Institute, presented data from a study showing the first example of an innate immune-driven escape mutation in HIV-infected individuals, which could help lead researchers to a better understanding of how NK cells—thought to be non-specific—are able to recognize HIV. This data could also help researchers ascertain how HIV can evade the early expansion of certain NK cells.

Certain polymorphisms of the killer cell immunoglobulin-like receptors (KIRs) present on the surface of NK cells (including KIR3DS1 and some alleles of KIR3DL1), in combination with specific HLA-B alleles such as HLA-B57, are known to delay progression to AIDS in HIV-infected people (see Perspective: Natural Killer Cells: Bridging Innate and Adaptive Immunity?, IAVI Report, May-June 2006). However, the mechanism for this is unclear.

Earlier experiments in acutely infected HIV-infected individuals have shown a remarkable expansion of NK cells following acute infection, with a preferential increase in the frequency of KIR-3DS1 in the presence of its putative ligand HLA-B Bw4801. Researchers have also found that certain subtypes of KIR-3DL1, a highly polymorphic allele, lead to different expression levels of this receptor on the surface of NK cells, and that the alleles that are expressed at high levels are also able to provide protection from disease progression in the presence of their HLA-B ligands.

To determine whether KIR on NK cells could place direct pressure on the virus in vivo, Alter, along with colleagues at the Ragon Institute and Microsoft Research, sequenced the viral DNA of 91 chronically infected individuals who were not on ARV therapy and looked for potential escape mutations associated with all known KIR receptors. They found 22 KIR-driven mutations and determined that at least two of these mutations directly provide a means by which the virus is able to avert detection by KIR-expressing NK cells.

Todd Allen at the Ragon Institute is now working with colleagues at the Broad Institute, a joint project between Harvard University and the Massachusetts Institute of Technology, to use ultra-deep sequencing to get a more in-depth look at viral evolution in larger sets of samples, and to expand this dataset to define all KIR-associated escape mutations that may occur in the virus to escape recognition by NK cells beginning in acute HIV infection. These studies will allow researchers to nail down exactly which variants of the virus select for the escape mutations and when the mutations occurred. In doing so, researchers hope to have a better idea of the role that innate immunity plays on viral containment and diversification.

Alter’s group believes HIV may be evading recognition/detection by NK cells by interfering with the activating receptor NKG2D, which modulates NK-cell function, and the protein ligands MICA and MICA/B, which NKG2D expresses during periods of cell stress. Chronically infected, untreated individuals appear to secrete higher levels of MICA, which appears to dampen the expression of NKG2D on the surface of NK cells. Similar mechanisms of NK cell evasion have been reported in certain cancer models. Alter postulates that HIV may be inducing these ligands, setting off a chain reaction that cleaves NKG2D and impairs NKG2D-dependent NK-cell mediated cytoxicity. “Bottom line is, I think NK cells are playing an important role, early on, in HIV infection,” said Alter.

Visualizing transinfection

NK cells are not the sole component of innate immunity that has been occupying the minds of HIV researchers lately. Researchers are also focusing on dendritic cells (DCs).

DCs engulf and degrade HIV to present viral antigens to CD4+ T cells. But in doing so, a portion of the virus can remain intact inside DCs. Two years ago, David McDonald, an assistant professor of cell and molecular virology at Case Western Reserve University, used antibody staining to show that these intact particles left inside DCs reside in compartments contiguous to the cell membrane, making it possible for the viral particles to be passed along to CD4+ T cells in a process known as transinfection (5; see Figure 3). This finding helped spur further study of how innate immune cells can both help and hinder the spread of HIV. 


Figure 3: Viral Crossing  

Image shows an infectious synapse where the larger human antigen-presenting cell is passing HIV virions (green) to the two smaller target CD4+ cells. Image courtesy of David McDonald, Case Western Reserve University.


At Keystone, McDonald used imaging technology to show that cell-surface markers on the plasma membranes of the T cell entered the HIV pocket of the DC at the infectious synapse, a region on the surface of T cells where CD4 and its co-receptors are recruited.

McDonald said there are ligands between T cells and DCs that enable the cell membranes to stick together. And when the T cell starts sniffing around looking for peptides presented by the DCs, the T cell also runs the risk of sticking its membrane into the intact virus. If there happens to be CD4 present on the surface of the T cell, HIV will slide in and infect it.

Although immune activation drives mucosal DCs into lymphoid tissues, where the vast majority of HIV replication occurs, it is not clear how important or extensive these interactions are. McDonald’s group has hypothesized that DCs amplify HIV infection within lymphoid tissues by continually picking up and passing on infectious HIV during interactions with CD4+ T cells. “Our hypothesis, through indirect evidence, is that DCs help to drive the disease,” said McDonald. “But we don’t have any in vivo evidence to support that. Our next push is to extend our live-cell imaging capability to study these cellular interactions within lymphoid tissues.”

1. Nature 464, 232, 2010
2. N. Engl. J. Med. 355, 2283, 2006
3. Science 319, 921, 2008
4. PLoS One 5, e8886, 2010
5. PLoS Pathog. 4, e1000134, 2008