Genetic factors affecting HIV infection and progression rates: immunological factors relating to entry inhibitors

Simon Collins, HIV i-Base

The opening session at the meeting was one of only two invited lectures and appropriately for a virology meeting, focused on immunological responses. Entry inhibitors that target human receptors will bring a new awareness of the range of immunological responses in patients. As the newest class of drugs, now in Phase 2/3 studies, this will become an increasing focus for this meeting.

This lecture by Sunil Ahuja from the Veterans Administration Research Center in San Antonio, Texas provided more detail relating to the complexity of viral entry, its relationship to viral fitness on an individual and epidemiological level, and the role played by genetic background. [1]

The larger context of the talk was to provide an explanation for the wide inter-patient variability in terms of exposure and risk of infection. This included why some people appear to be protected from infection, and differences in HIV progressesion rates in individuals who become infected. It was based on research published earlier this year in Science. [2]

From a virological perspective, factors affecting the risk for infection include viral fitness and viral load levels of the source partner. Viral fitness (and resistance) therefore become important factors for the effectiveness of drugs that target the virus life-cycle. As CCR5 inhibitors are dependent on host responses, it is important to understand new immunological factors that contribute to their effectiveness.

Immunological factors contributing to risk of infection, and speed of progression include differences in CCR5 co-receptors. The double deletion in delta32 is well known for providing protection against HIV-infection, but the importance of other sites in the cis region of CCR5 and changes at 641 in CCR2 in the level of protection have received less attention.

In a similar way, levels of human leukocyte antigen (HLA) varies between individuals and is one of the factors that determine whether an individual is able to mount strong or weak CD8+ CTL immune responses.

Genetic differences in the CCR5 ligand network were the main focus for the talk: principally CC chemokine ligand 3-like1 gene number (CCL3L1, also called MIP-1-alphaP), which affect levels of RANTES which also vary considerably between individuals. A ligand is the term for a molecule that binds to a specific protein or receptor to form a larger molecule, and RANTES and MIP-1-alphaP directly block HIV from attaching to the receptor. Ahuja’s research focused on whether higher levels of these chemokines correlated with levels of protection against HIV on both an individual and population level.

Their hypothesis is that these differences explain why some multiply-exposed individuals never become infected, and others are infected after one exposure. They also explain the broad range of responses between the two extremes of rapid progessors and long-term slow progressors and why two people infected with the same strain may have widely different rates of disease progression.

This was tested against data from two large cohorts. The first cohort included 1100 adult HIV-positive patients, half of whom were seroconverting, representing different genetic backgrounds (approximately one third African American) treated at the Wilford Hall Medical Centre, Texas (with a similar number of HIV-negative controls). This is the referral centre for USAF military and is 94% male with median ages at diagnosis of 28 (range 18-70 years). The second cohort included around 800 perinatally-exposed children in an Argentinian cohort (genetically classed as European) half of whom were HIV-positive half were HIV-negative.

The distribution of population differences in gene copy was determined from 1046 samples from the Human diversity genome panel, including humans from 57 populations plus 83 chimpanzees. This showed wide population differences. Approximate mean copy number ranged from 5-7 in Africans, 3-5 in Americans, 2-5 in East Asians and 2-3 in Europeans. Interindividual and interpopulation differences resulted in a median phenotype for African Americans of 3-4 copies compared to 1-2 copies in Non-Africans; ie half the dose in Caucasians has a comparable protective effect. They found no relationship between absolute copy number of CCL3L1 and protective benefit. However, when individual copy number was then compared to the genetic population distribution, they found a clear correlation between CCL3L1 copy number and infection risk in both the Texan and Argentinian cohorts.

Compared to a median of 2 copies, children in the Argentinian cohort with either greater or less than two copies had significantly higher or lower risks respectively of acquiring HIV. Each increase in copy number above the median produced a dose-dependent step-wise decrease in the risk for acquiring HIV.

In HIV-positive individuals, they found a dose dependent correlation between copy number and both viral set point and rate of change in CD4 cells.

When they looked at the phenotypic effect of CCL3L1 dose and CCR5 haplotypes when combined in four mutually exclusive genetic risk groups based on CCL3L1 number relative to population (high or low) and whether the CCR5 genotype was detrimental (det) or accelerating (non-det). If the impact on disease progression was related to both factors, then low CCL3L1 plus detrimental CCR5 geneotype should correlate to one extreme of faster disease progression, and high CCL3L1 plus non-detrimental CCR5 genotype would be linked to slow disease progression. Definitions for detrimental CCR5 genotype were derived from previous genetic analysis, that had showed highly statistically significant associations with disease progression, relating to changes in the non-coding region and delta-32 deletion in CCR5, and the CCR2-64I region of CCR2.

Relative to the high+nondet group, low+det genotypes were associated with a >3-fold increased risk of progressing to 8 of 12 AIDS-defining illnesses (ADIs). Relative to the high+nondet group, the two middle risk groups low+nondet and high+det were associated with a less than 3-fold risk of progressing to 3-4 ADIs.

Summaries of three of these analyses are shown in Tables 1, 2 and 3 below. (EA=European American; AA=African American).

Table 1: RH (95% CI) acquiring HIV by genetic response group in adult cohort

CCL3L1 low (<2EA, <3AA) high (=2 EA, =3 AA)
R5 det. 3.44 (2.54-4.67) 1.75 (1.26-2.42)
R5 non-det. 1.52 (1.22-1.90) 1.00

Table 2: RH (95% CI) acquiring HIV by genetic response group in children exposed perinatally

CCL3L1 low (<2EA, <3AA) high (=2 EA, =3 AA)
R5 det. 4.03 (2.39-6.69) 1.48 (0.97-2.27)
R5 non-det. 2.36 (1.40-4.00) 1.00

Table 3: RH (95% CI) progressing to AIDS by genetic response group in adults diagnosed during seroconversion

CCL3L1 low (<2EA, <3AA) high (=2 EA, =3 AA)
R5 det. 4.87 (2.82-8.39) 1.76 (1.02-3.02)
R5 non-det. 1.45 (0.94-2.25) 1.00

The lecture only covered part of more detailed research from the earlier Science publication, and referral to that paper with the accompanying notes available online is recommended. [2] The conclusions have particular significance now that new drugs are in development that are likely to vary in activity based on immunological parameters. Gene dosage and response to CCR5-blockers in an obvious area for future study, as are other host genetic factors.


  1. Ahuja S. Host genetic determinants of HIV transmission and pathogenesis: should we care? 14th International HIV Drug Resistance Workshop (14th IHDRW), 7-11 June 2005, Quebec City, Canada. Abstract P1.
  2. Gonzalez G, Kulkarni H, Bolivar H et al. The influence of CCL3L1 gene¬Ėcontaining segmental duplications on HIV-1/AIDS susceptibility. Science 307: 1434. 4 March 2005. Published online 6 Jan 2005.

Links to other websites are current at date of posting but not maintained.