Modulating influence on HIV/AIDS by interacting RANTES gene variants

Abstract

RANTES (regulated on activation normal T cell expressed and secreted), a ligand for the CC chemokine receptor 5, potently inhibits HIV-1 replication in vitro. We tested the influence of four RANTES single nucleotide polymorphism (SNP) variants and their haplotypes on HIV-1 infection and AIDS progression in five AIDS cohorts. Three SNPs in the RANTES gene region on chromosome 17 (403A in the promoter, In1.1C in the first intron, and 3'222C in the 3' untranslated region) are associated with increased frequency of HIV-1 infection. The common In1.1C SNP allele is nested within an intronic regulatory sequence element that exhibits differential allele binding to nuclear proteins and a down-regulation of gene transcription. The In1.1C allele or haplotypes that include In1.1C display a strong dominant association with rapid progression to AIDS among HIV-1-infected individuals in African-American, European-American, and combined cohorts. The principal RANTES SNP genetic influence on AIDS progression derives from the down-regulating RANTES In1.1C allele, although linkage disequilibrium with adjoining RANTES SNPs including a weaker up-regulating RANTES promoter allele (-28G), can modify the observed epidemiological patterns. The In1.1C-bearing genotypes account for 37% of the attributable risk for rapid progression among African Americans and may also be an important influence on AIDS progression in Africa. The diminished transcription of RANTES afforded by the In1.1C regulatory allele is consistent with increased HIV-1 spread in vivo, leading to accelerated progression to AIDS.

Figure 1

(a) The genomic structure of the RANTES gene on chromosome 17q11.2 and the haplotype structure and frequency of the RANTES variants. Exons are shown as open boxes, and intron sizes are indicated. Locations of nucleotide variants are indicated by arrows. Variants in the promoter region are numbered according to the transcription start site, and the SNP in the 3′ untranslated region is numbered from the first nucleotide of the 3′ untranslated region. Variants In1.1 and In1.2 are located at nucleotide positions 168923 and 170226, respectively, in GenBank sequence AF088219. (b) Competitive electrophoretic mobility-shift assay of DNA–nuclear protein binding at the In1.1T/C site. Nuclear extracts from CD4⁺-enriched lymphocytes bound to the intron 1 fragment containing In1.1T (lanes A–D) or In1.1C (lanes E–H). Lanes A and E, no extract; lanes B and F, In1.1T and In1.1C probe, respectively, without competitors; lanes C and G, In1.1T and In1.1C probe, respectively, with a 100-fold excess of cold In1.1T probe as competitor; lanes D and H, In1.1T and In1.1C probe, respectively, with a 100-fold excess of cold In1.1C probe as a competitor. The arrowheads indicate the specific DNA–protein complexes associated with In1.1T/C. (c) Luciferase activities of pBL3-promoter constructs containing RANTES intron 1 fragments with either In1.1T (pGL3-In1.1T) or In1.1C (pGL3-In1.1C). The pBL3-promoter vector served as a control. Bars indicate mean values with standard deviations; P values were determined by Student's t test. Results are the mean of two experiments performed in triplicate. (d) Luciferase activities of the RANTES haplotype constructs containing the RANTES promoter (−403G/A, −28C/G) and the intron 1 element (In1.1T/C) allele combinations. The pGL3-basic vector constructs contain the promoter fragments only (hatched bar), promoter and intron 1 fragments carrying In1.1T (black bars), and promoter and intron 1 fragments carrying In1.1C (dotted bars). Haplotypes are denoted by the nucleotide at positions −403, −28, and In1.1, respectively. These constructs were also tested in the U937 monocyte cell and CCRF-SB B cell lines. The gene transcription activities were similar in the U937 cells but were much weaker in the CCRF-SB B cells, compared with those in Jurkat T cells (data not shown). The pBL3-basic vector (pB) served as a control. Note: haplotype AGT has not been observed in the population studied.

Figure 2

Comparison of frequency of genotypes carrying RANTES SNPs or the R4 haplotype between high-risk HIV-1-exposed but uninfected and HIV-1 seroconverters for EA (Left) and combined EA and AA (Right). Numbers above bars are the numbers of subjects in each group; * indicates a significant frequency difference (dominant model) by a two-sided Fisher's exact test.

Figure 3

Kaplan–Meier survival curves for progression to AIDS-1987 (a–d), CD4 < 200, AIDS related death comparing the influence of the RANTES In1.1C allele, and In1.1C-carrying haplotypes on progression to AIDS by HIV-1-infected EA (Left) and AA (Right). Cox model RH and Kaplan–Meier Wilcoxon P values comparing each variant genotype or haplotype to the no-In1.1C group are shown for each factor. (a and b) Rate of progression to AIDS-1987 based on In1.1C genotype: black, no In1.1C; blue, one copy; and purple, two copies of In1.1C for EA (a) and AA (b). (c–f) Rate of progression to AIDS partitioned according to In1.1C-containing haplotypes: black, no In1.1C; red, R3; green, R4; and orange, R5. (c and e) Progression of EA to AIDS-1987 (c) and CD4 < 200 (e). (d and f) Progression of AA to AIDS-1987 (d) and AIDS-related death (f). Subjects carrying two different In1.1C haplotypes (two EA and six AA) are omitted.

Figure 4

Defined disease category analysis of RANTES In1.1C and the In1.1C-carrying R3 haplotype comparing slow versus rapid progressors to AIDS-1987 and AIDS-related death. Bars show frequencies of homozygotes (black) and heterozygotes (gray) for In1.1C or the R3 haplotype in the slow and rapid groups for AA (Left), EA (Center), and combined ethnic groups (Right). Number of subjects considered and significance for a codominant model by Fisher's exact test are indicated (*, P < 0.05; **; P < 0.01; ***, P < 0.001; ****, P < 0.0001).