At the recent CROI meeting, David Goldstein from Duke University gave a plenary presentation on genetic determinants of viral load set point in HIV-infected individuals. Goldstein’s talk was an update on work published recently with CHAVI (Center for AIDS Vaccine Immunology) collaborator Amalio Telenti from the University of Lausanne. Their approach involved analyzing a staggering 500,000 different single nucleotide polymorphisms (SNPs) present in the human genome to see if they were associated with lower viral load set points (although Goldstein noted that this initial work is restricted to SNPs documented in >1% of the population and thus excludes rare polymorphisms). As reported in the Science paper, three sets of significant associations emerged. The first was a SNP in a gene called HCP5. This is linked to an immune response gene called HLA B*5701 which is well known to be over-represented among HIV-infected long-term non-progressors (HLA B*5701 makes a receptor on CD8 T cells which appears particularly good at recognizing HIV epitopes). However, HCP5 is also an endogenous retroviral element (a part of the human genome derived from an ancient retroviral infection which gained access the human germ line by infecting an egg or sperm cell) and so Goldstein initially speculated that the SNP in HCP5 might somehow have a direct anti-HIV effect. At CROI, he reported that studies in which the SNP-containing version of HCP5 was overexpressed in cells showed no inhibition of HIV replication, suggesting that the SNP is mediating its effect via other means (exactly how is under investigation).
The second association uncovered by Goldstein’s work was with a SNP in the gene for HLA-C. HLA-C molecules are part of a family called class I HLA molecules that are involved in the recognition of pathogens by CD8 T cells. Thousands of HLA molecules are displayed on the surface of every cell in the body (except red blood cells) and their job is to constantly lift cellular debris from inside the cell and display it to passing CD8 T cells. If an HLA molecule displays a protein fragment (epitope) from a pathogen that has infected the cell, this can trigger the CD8 T cell to release toxic enzymes, which cause the cell to die. The majority of HLA molecules on cells belong to classes HLA-A and HLA-B, while HLA-C molecules are less frequent and, as a result, less studied. But because HIV’s Nef protein is known to cause a reduction in HLA-B molecules on infected cells (as a means of escaping the immune response), Goldstein’s hypothesis is that the SNP he has identified causes more HLA-C molecules to be expressed, thereby making it easier for CD8 T cells to identify HIV-infected cells. He is currently collaborating with Andrew McMichael at Oxford University to measure the effect of the SNP on HLA-C expression.
The third association described by Goldstein involves a set of seven SNPs in or near two genes, ZNRD1 and RNF39. In the Science paper, this association was only seen with disease progression (as assessed by time to a CD4 T cell count of less than 350) but at CROI Goldstein reported that in an expanded analysis involving 1,000 more individuals (in addition to the 446 studied initially), an association with viral load set point was also documented. ZNRD1 encodes a protein involved in RNA transcription and, interestingly, was also identified in the recent, widely publicized study of host proteins needed by HIV to replicate. The function of RNF39 remains to be determined. Goldstein stressed that the discovery of all these associations is a first step, and efforts are underway to uncover the mechanisms by which the identified SNPs mediate their effects.
To give a sense of how a combination of genetic factors can have a profound impact on HIV disease progression, Goldstein showed an analysis that included the SNPs in the HCP5, HLA-C, and ZNRD1/RNF39 genes and two other known favorable genetic polymorphisms in CCR5 and CCR2 genes (CCR5Δ32 and CCR2 V64I). HIV-infected individuals with no favorable mutations in any of these genes showed an average time of less than two years from infection to a CD4 T cell count of less than 350. In contrast, people with one or two favorable mutations in at least four of these genes did not experience a CD4 T cell decline to this level for an average period of more than eight years.
Goldstein’s group is now embarking on an effort to uncover genetic associations with the magnitude of antibody responses generated against a vaccine, using data from the North American efficacy trial of AIDSVAX. This is a potentially important area of study because the ability of an individual to generate a high titer antibody response was correlated with reduced susceptibility to HIV infection in the trial cohort. Some researchers have suggested an analogy with the association between the magnitude of anti-Ad5 antibody responses and susceptibility to HIV infection seen in the placebo group of the recent Merck vaccine trial, so Goldstein’s work may have the potential shed light on that mystery also. Another important area of ongoing study mentioned by Goldstein is an analysis of genetic associations with viral load set point restricted to African American individuals.
The webcast of David Goldstein’s talk is available on the CROI website, it is the second presentation on Monday.
Shortly after CROI, a study of untreated HIV-infected individuals with viral loads under 50 copies – a group now dubbed “elite controllers” – highlighted the complexities and limitations of genetic associations. In a research letter published in the journal AIDS, Yefei Han and colleagues from Bob Siliciano’s laboratory at Johns Hopkins report that, of 16 elite controllers analyzed, only four possessed the SNP in HLA-C identified in the CHAVI work and none had the SNP in HCP5 (although two had HLA B*5702 and HLA B*5703 genes, which are closely related to HLA B*5701). The researchers did find the HCP5 SNP in two other infected individuals with HLA B*5701, one of whom had a low but detectable viral load and the other with progressive disease. The authors also note that the frequency of the HLA-C SNP in the elite controllers was not significantly different from what would be expected in a larger population of people from the same ethnic background. Citing these data and an example of an individual in their cohort who developed an increasing viral load after developing a CD8 T cell escape mutation, they argue that adaptive immune responses are likely an important contributor to the control of viral replication in elite controllers.
AIDS. 22(4):541-544, February 19, 2008.
Research Letters
Han, Yefei; Lai, Jun; Barditch-Crovo, Patricia; Gallant, Joel E; Williams, Thomas M; Siliciano, Robert F; Blankson, Joel N
Abstract:
Elite suppressors (ES) are untreated HIV-1-infected patients who maintain undetectable viral loads. A recent whole-genome analysis identified two independent polymorphisms associated with low viral loads in untreated HIV-1 infection. We screened 16 ES; none were positive for the protective HLA complex 5 gene polymorphism, and only four were positive for the protective polymorphism associated with the HLA-C gene. These results suggest that some ES control viremia by mechanisms independent of the newly-identified genetic factors.
Comments