Consistent effects of TSG101 genetic variability on multiple outcomes of exposure to human immunodeficiency virus type 1

Abstract

Tumor susceptibility gene 101 (TSG101) encodes a host cellular protein that is appropriated by human immunodeficiency virus type 1 (HIV-1) in the budding process of viral particles from infected cells. Variation in the coding or noncoding regions of the gene could potentially affect the degree of TSG101-mediated release of viral particles. While the coding regions of the gene were found to lack nonsynonymous variants, two polymorphic sites in the TSG101 5' area were identified that were associated with the rate of AIDS progression among Caucasians. These single-nucleotide polymorphisms (SNPs), located at positions -183 and +181 relative to the translation start, specify three haplotypes termed A, B, and C, which occur at frequencies of 67%, 21%, and 12%, respectively. Haplotype C is associated with relatively rapid AIDS progression, while haplotype B is associated with slower disease progression. Both effects were dominant over the intermediate haplotype A. The haplotypes also demonstrated parallel effects on the rate of CD4 T-cell depletion and viral load increase over time, as well as a possible influence on HIV-1 infection. The data raise the hypothesis that noncoding variation in TSG101 affects the efficiency of TSG101-mediated release of viral particles from infected cells, thereby altering levels of plasma viral load and subsequent disease progression.

FIG. 1.

SNPs and the corresponding haplotypes observed in the 5′ area of TSG101. (A) Schematic map of the nine SNPs identified in the 5′ area of TSG101. SNP positions are determined relative to the “A” nucleotide of the ATG start codon, which is shown within exon 1 (black box). (B) Haplotypes based on the nine SNPs. The haplotype frequencies were estimated based on sequencing data among 79 randomly chosen Caucasian seroconverters. Only haplotypes with frequencies of >0.01 are listed. Alleles with minor frequencies are shown in boldface. F, frequency, CI, confidence interval. (C) Haplotypes based on four SNPs estimated in 1,895 Caucasian individuals. (D) Structure and frequencies of haplotypes A, B, and C estimated in 2,071 Caucasian individuals.

FIG. 2.

Kaplan-Meier plots for the three TSG101 haplotypic groups. Four AIDS outcomes were analyzed using the Cox proportional-hazards model.

FIG. 3.

Decline in CD4 T-cell count among AIDS patients with different TSG101 haplogenotypes. The data were analyzed using a random-effects linear model (see Materials and Methods). (A) Square-root CD4 counts for 380 individuals from the MACS cohort were plotted as a function of time from seroconversion, and fitted lines for each genotype group were generated based on 4,918 measurements. Slopes for protective, neutral, and susceptible genotypic groups are as follows (95% CI in parentheses): −1.11 (−1.22, −1.00), −1.33 (−1.42, −1.25), and −1.39 (−1.50, −1.27), respectively. (B) CD4 data from the SHCS patients were plotted starting from the point at which CD4 cell counts fell in the range of 500 to 600 cells/mm³. By chance, the susceptible haplogenotypic group had a mean CD4 count that was higher in this range than that of the protective or neutral group. The fitted lines are based on 3,551 measurements for 310 people. Slopes (95% CI) for protective, neutral, and susceptible groups are −0.52 (−0.64, −0.39), −0.57 (−0.69, −0.47), and −0.81 (−0.99, −0.64), respectively.

FIG. 4.

Viral-load increase over time in 373 MACS patients. Fitted lines for the three TSG101 genotypic groups were constructed based on 3,569 measurements of log₁₀ HIV RNA over time after seroconversion. Statistical analysis was performed using the multivariate random-effects linear model (see Materials and Methods). The best-fit lines for the protective, neutral, and susceptible haplogenotypic groups have slopes (95% CI) of 0.01 (−0.01, 0.02), 0.04 (0.03, 0.05), and 0.05 (0.04, 0.07), respectively.