Research Briefs

By Roberto Fernandez-Larsson

A Therapeutic Dendritic-Cell Vaccine for Chronic HIV Infection

Therapeutic AIDS vaccines attempt to bolster the virus-specific immune response in individuals who are already HIV infected. In a recent report (Nat. Med. 10, 1359, 2004), a group led by Wei Lu and Jean-Marie Andrieu present a preliminary study on the efficacy of a therapeutic AIDS vaccine based on dendritic cells (DCs).

The vaccine comprised autologous activated DCs obtained from the volunteers themselves and incubated in vitro with aldrithiol-2 (AT-2)-inactivated autologous HIV-1 isolates—in essence, each volunteer received a ‘personalized’ vaccine. DCs were cultured ex vivo with granulocyte-macrophage colony-stimulating factor (GM-CSF) and interleukin (IL)-4, AT-2-inactivated virus was added and then cultured with IL-1β, IL-6, and tumor necrosis factor (TNF)-α. The 18 inoculated adult volunteers were HIV-infected for at least a year, with a stable mean plasma viral RNA load of 3 x 105 copies/ml and a mean CD4+ T cell count of 550 cells/μL at the time of injection. All patients were immunized three times at 2-week intervals with the AT-2-inactivated HIVpulsed DCs, and followed for one year thereafter without antiviral therapy.

After vaccination, the plasma viral RNA loads of 8 individuals decreased from around 3 x 105 copies/ml to approximately 1 x 105 copies/ml during the first 112 days and remained stable for the rest of the year. There was a lesser and transient reduction in plasma viral loads in the other 10 subjects, and in some cases they exceeded preimmunization values one year after vaccination. But in the individuals who did show a sustained reduction in plasma viral loads there was a positive correlation with the numbers of HIV-specific IL-2- or interferon-γ-expressing CD4+ T cells, which normally decline in the course of untreated infection, and with Gag-specific perforin-expressing CD8+ (effector) T cells. Neutralizing antibody apparently did not contribute to reduction in viral loads since titers remained unchanged in 16 of the 18 subjects.

This study in humans is a follow-up to the authors’ previous publication describing a similar vaccine modality (autologous DCs pulsed with AT-2-inactivated autologous virus) in Chinese macaques with chronic SIV infection (Nat. Med. 9, 27, 2003). In the human study the authors could not rule out a nonspecific ‘adjuvant effect’ of the DC preparation since a control group inoculated with activated DCs not pulsed with HIV was not included. However, in the macaque study an appropriate control group treated with pulsed DCs alone (no AT-2- inactivated virus) did not present an adjuvant effect. The authors suggest that unloaded mature DCs cannot present the virus already present in these monkeys, because as mature DCs they have almost lost their antigen-uptake capacity. However, it remains to be seen just how robust these latest results are since the characterization of the individual’s immune responses was minimal, with patients only sampled at 3 time points after vaccination.

An expensive, time-consuming autologous vaccine will never be practical, but this study provides the first demonstration in humans that an HIV-specific cellular immune response can have a positive effect on some parameters of immunity in vivo, and is possibly encouraging news that could be applied to develop more effective, conventional preventive vaccines (e.g., DNA or recombinant viral vector vaccines) that induce similar HIV-specific immune responses in vivo.

The Influence of a Chemokine Gene Number on Susceptibility to HIV/AIDS

Analyses of sequences of the human genome have identified a large amount of interspersed as well as tandem low-copy repeats or segmental duplications, and the possibility has been raised that variation in the copy number of specific host defense genes may affect the susceptibility to, or the progression or severity of diseases in which the genes play a role.

The copy number of the gene encoding CC chemokine ligand 3-like 1 (CCL3L1) varies among individuals. Importantly, CCL3L1 (also known as MIP-1αP) is the most effective known ligand for the HIV coreceptor CC chemokine receptor 5 (CCR5), and the most active naturally-occurring inhibitor of HIV entry known. To test the hypothesis that segmental duplications of host defense genes causing dosage effects are associated with phenotypic effects in vivo, Sunil Ahuja and his collaborators (Science 307, 1434, 2005) have determined the distribution of chemokine gene-containing segmental duplications in humans and the relationship to HIV infection/susceptibility.

The researchers determined the average number of CCL3L1 gene copies in more than 5,300 HIV-infected and -noninfected individuals of different ancestral origins to see if copy number is associated with either the risk of acquiring HIV or the rate at which HIV disease progresses.

They found that increasing CCL3L1 copy number was positively associated with ligand secretion, which is not necessarily the case with all genes. People with different geographical ancestries possessed a significantly different number of CCL3L1 gene copies. This does not mean that any one group is more susceptible to HIV/AIDS than other populations. Rather, using the average CCL3L1 copy number as a reference point for each group, the researchers found that individuals with a CCL3L1 copy number lower than their population specific median were at a higher risk of acquiring HIV infection.

Depending on the study population, each CCL3L1 copy lowered the risk of acquiring HIV by 4.5-10.5%. CCL3L1copy number was also associated with variable rates of disease progression. In the adult cohort of HIV-infected individuals, a gene dose lower than the overall cohort median or population-specific median was associated with a dose-dependent increased risk of progressing rapidly to AIDS. These authors and others had previously shown that small sequence variations within the CCR5 gene influence the risk of acquiring HIV and disease progression. They found here that individuals who possessed a low CCL3L1 copy number along with disease-acceleratingCCR5 variants had an even higher risk of HIV acquisition and rate of progression to AIDS. They found thatCCL3L1 copy number correlated with ligand secretion, as well as a dose-dependent association with the viral set point and rate of change in CD4+ T cell counts, which are predictors of disease progression.

They speculate that these chemokines may exert their HIV-suppressive activity by steric hindrance of the attachment of gp120 to CCR5, or perhaps by inducing the internalization of CCR5 molecules that would then be unavailable for gp120 attachment.

The authors suggest that CCL3L1 gene dose is a novel means of “buffering” against the risk of HIV infection and/or disease progression in the populations examined. These findings could have implications for AIDS vaccine researchers since CCL3L1 gene dose may be an important genetic correlate of vaccine responsiveness. This contention is supported by studies in monkeys that have shown that CCR5 ligand production predicts protection and only protected animals had markedly increased concentrations of chemokines. In human vaccine trials, especially those involving small numbers of individuals, it may be important to determine the genetic profile of volunteers. Testing a vaccine’s efficacy in a group of individuals with higher than average CCL3L1 copy number may falsely indicate that the vaccine does not work or works poorly.

Study Identifies Genes That Play Key Role in the Immune Response Against HIV

In humans, human leukocyte antigen (HLA) class 1 molecules are expressed on the surface of cells where they help the immune system recognize virus-infected cells. When new virus particles are produced within an infected cell, class 1 molecules capture fragments of viral proteins and expose them at the cell surface, alerting CD8+ T cells that the cell is infected and should be lysed. Three genes (HLA-A, HLA-B, and HLA-C) encode class 1 molecules. HLA-B genes are extremely diverse, with 563 different alleles identified as opposed to 309 for HLA-Aand 167 for HLA-C.

An international research team led by Philip Goulder has identified immune system genes that appear to play a key role in the immune response against HIV (Nature 432, 769, 2004).

The study was designed to test the hypothesis that the diversity of HLA class 1 molecules could reflect functional differences in the CD8+ cytotoxic T lymphocyte responses controlled by those molecules. The authors studied 375 HIV-infected, treatment-naïve individuals in southern Africa to determine if a particular type of class 1 molecule controls the CD8+ T cell response against the virus. To study the contributions of individual HLA class I molecules they used a panel of 410 overlapping synthetic peptides, spanning the entire expressed HIV genome, and characterized the T cell responses to these peptides in interferon-γ ELISPOT assays. They found that the association of HLAB alleles with peptide-specific responses far exceeded that of HLA-A and of HLA-C alleles, soHLA-B alleles contribute significantly more to the total HIV-specific CD8+ T cell response in the studied population.

Next they studied the influence of class 1 molecules on plasma viremia in 706 chronically HIV-infected treatment-naïve persons, and found that viral load varied significantly according to the particular HLA-B allele expressed. To further test whether it is a particular HLAB allele that principally influences disease outcome, they looked at levels of the CD4+ T cells that are destroyed by HIV in relation to HLA type. The tests all supported the conclusion that the form of the HLAB molecule inherited by patients makes a significant difference in how well their immune systems cope with the infection.

Finally, they analyzed HLA-B alleles in HIV-infected mothers and their infants. They found that HIV-infected women who have a particular HLA-B allele are more likely to survive, and also less likely to transmit the virus to their infants, suggesting that HIV may be exerting selective pressure on certain protective HLA-B alleles.

Although these data indicate a dominant role for HLA-B alleles in HIV infection the authors are quick to point out that the underlying mechanism remains to be elucidated. They speculate that a possible mechanism is the greater diversity of peptides bound by HLA-B alleles, which can accommodate various positively and negatively charged residues, while HLA-A binds only hydrophobic residues. The study has implications for AIDS vaccine research because it may lead to ways of circumventing the virus’ ability to avoid vaccine-induced immunity by rapid mutation. It also describes how HIV infection may be driving human evolution, since individuals with protective versions of those genes are more likely to survive and pass the genes to children.

Two Crystal Structures May Help Guide Vaccine and Treatment Approaches

The envelope glycoprotein of HIV, gp120, attaches to specific receptors and co-receptors on the surface of cells, enabling virus entry. Several years ago scientists uncovered the structure of gp120 bound to the CD4 cell receptor. The structure of fragments of gp41 in its post-fusion state is also known. Now, Stephen Harrison and colleagues (Nature 433, 834, 2005) have determined the unbound monomeric structure of the gp120 molecule, allowing them to build a model predicting that gp120 significantly changes its shape after binding CD4, enabling the molecule to present different antigenic sites in its two states. In its natural state, gp120 is assembled as trimers in the virus spikes together with gp41, also in a trimer.

Researchers have sought the structure of unliganded gp120 for almost two decades and, in a Nature commentary, Peter Kwong describes this work as a “technical tour de force.” To crystallize the gp120 molecule they used gp120 ‘cores’ of the closely related simian immunodeficiency virus (SIV). These cores had the variable loops V1, V2 and V3 deleted, as well as parts of the molecule’s carboxy and amino terminals. Unresolved is the question whether the unbound gp120 structure described here is different from the unbound structure of gp120 in the viral spike. But they were able to crystallize the SIV gp120 cores with their full complement of surface glycosylation, revealing the extent to which these sugar moieties coat the molecule and possibly protect antigenic sites from antibody attachment. They found, for example, that amino acids that compose the chemokine-receptor site on gp120, which forms after CD4 binding, are not contiguous in the unliganded structure, but amino acids involved in mutations that generate resistance to entry-inhibiting compounds are all arranged facing a wellconfigured pocket.

Vaccine designers will want to know if the bound and unbound partial structures of gp120 now solved provide information on their antigenic properties that may be helpful to understand the ability of HIV to evade the host’s neutralizing antibody. Arguably, gp120 is more vulnerable in its unbound state before cell attachment, but even in the unbound state alone there may be more than one conformation, when the virus mutates Env to modulate its antigenic structure and escape neutralization. A recent paper by Ronald Montelaro and colleagues (J. Virol. 79, 2097, 2005) emphasizes this ability to evade neutralization by showing for the first time that even point mutations in the intracytoplasmic tail (ICT) of the gp41 can render the virus more resistant to neutralization, seemingly through allosteric effects.

In a second X-ray crystallography paper (Immunity 22, 163, 2005), a group from the Scripps Research Institute led by Dennis Burton and Ian Wilson examined the structure of the broadly neutralizing monoclonal antibody (MAb) 4E10 bound to a peptide fragment identical to the gp41 domain recognized by the antibody. The study identified the amino acids of the epitope recognized by MAb 4E10 as well as its three dimensional structure. Burton and Wilson are members of a consortium of laboratories, the International AIDS Vaccine Initiative’s Neutralizing Antibody Consortium, which is focused on understanding broadly neutralizing antibodies at the molecular level. They hope that the information obtained from the crystal structure will aid scientists in designing immunogens that when used as vaccines could elicit antibodies with properties similar to 4E10.