Genetic variation in ABC transporter A1 contributes to HDL cholesterol in the general population

Abstract

Homozygosity for mutations in ABC transporter A1 (ABCA1) causes Tangier disease, a rare HDL-deficiency syndrome. Whether heterozygosity for genetic variation in ABCA1 also contributes to HDL cholesterol (HDL-C) levels in the general population is presently unclear. We determined whether mutations or single-nucleotide polymorphisms (SNPs) in ABCA1 were overrepresented in individuals with the lowest 1% (n=95) or highest 1% (n=95) HDL-C levels in the general population by screening the core promoter and coding region of ABCA1. For all nonsynonymous SNPs identified, we determined the effect of genotype on lipid traits in 9,259 individuals from the general population. Heterozygosity for ABCA1 mutations was identified in 10% of individuals with low HDL-C only. Three of 6 nonsynonymous SNPs (V771M, V825I, and R1587K) were associated with increases or decreases in HDL-C in women in the general population and some with consistent trends in men, determined as isolated single-site effects varying only at the relevant SNP. Finally, these results were consistent over time. In conclusion, we show that at least 10% of individuals with low HDL-C in the general population are heterozygous for mutations in ABCA1 and that both mutations and SNPs in ABCA1 contribute to HDL-C levels in the general population.

Figure 1

Topological model of ABCA1 according to Fitzgerald et al. (18). The amino acid substitutions identified in the present study are superimposed (SNPs are underlined, mutations are in bold). Location of the variants is deduced from ref. . ABC transporters are composed of 4 parts: 2 membrane-integral domains (H1–H6 and H8–H12 + H14), each of which spans the membrane 6 times, and 2 ATP-hydrolyzing domains (ABCs), which contain the highly conserved Walker A (GXXGXGKS/T) and B motifs (hhhhD; h, hydrophobic amino acid) connected by an ABC family–specific signature motif (LSGGQQ/R/KQR) (38). H1–H13, hydrophobic segment 1 to 13; R1 and R2, regulatory segments 1 and 2. Asterisks indicate new mutations identified in the present study.

Figure 2

Alignment of ABCA1 between species. Human ABCA1 (NP_005493.2) and the orthologous murine (NP_038482.1), rat (NP_835196.1), and chicken (AAL56247.1) protein sequences were aligned with the ClustalW program (http://www.ebi.ac.uk/clustalw/). Selected parts of the protein with the identified amino acid substitutions are marked in underscored (SNPs) and bold (mutations), respectively. Underscored amino acids (VVILD) at residues 1053–1057 mark the Walker B domain. At residue 1587, K (A allele) is reported to be the WT in humans (GenBank protein accession number NP_005493.2). In this study, R (G allele) is the most common amino acid, and similar results have been reported by others (29). Asterisks indicate identical amino acid residues in all sequences aligned.

Figure 3

Alignment of human ABCAs. ABCAs are all full transporters and are involved in sterol transport, either tissue specific or more generalized. The ABCA11 gene sequence/protein sequence is not available in Homo sapiens. ABCA1 to ABCA10 and ABCA12 to ABCA13 Homo sapiens protein sequences, NP_005493.2, NP_001597.1, NP_001080.1, NP_000341.1, NP_061142.2, NP_525023.2, NP_061985.1, NP_009099.1, NP_525022.2, NP_525021.2, NP_775099.2, NP_689914.2, were aligned with the ClustalW program (http://ebi.ac.uk/clustalw/). Selected parts of the proteins with the identified amino acid substitutions are marked in bold (mutations) and underscored (SNPs), respectively. Asterisks indicate identical amino acid residues in all sequences aligned.

Figure 4

In left panels, plasma HDL-C (mmol/l) as a function of 6 nonsynonymous SNPs in ABCA1 regardless of variation at the other 5 sites (overall effects) in women and men from the general population. In right panels plasma HDL-C (mmol/l) as a function of 6 SNP genotypes differing only at the SNP of interest (isolated single site effects) and in the order (left to right) R219K, V771M, V825I, I883M, E1172D, R1587K in women and men from the general population. The WT (the most common genotype), heterozygote (He), and homozygote (Ho) 6 SNP genotypes are given to the right of the bar graphs with the relevant SNP underlined. To increase power for V825I, the combined genotypes also varied at the R219K site, since R219K did not affect HDL-C levels. Values are mean ± SEM. *P < 0.0083 for all ANOVAs when corrected for multiple comparisons using the Bonferroni method. Post hoc tests: **P < 0.05; ^#P < 0.01.

Figure 5

Plasma HDL-C (mmol/l) measured at the second examination (1981–1984) and HDL-C and apoAI measured at the third examination (1991–1994) of the Copenhagen City Heart Study as a function of 6 SNP genotypes differing only at the relevant SNP (isolated single site effects) as detailed in the legend to Figure 4, right panels. Values are mean ± SEM. *P < 0.0083 for all ANOVAs when corrected for multiple comparisons using the Bonferroni method. Post hoc tests: **P < 0.05; ^#P < 0.01.

Figure 6

Plasma HDL-C (mmol/l) as a function of 6 SNP genotypes differing only at the relevant SNP (isolated single-site effects) as detailed in the legend to Figure 4, right panels. Values are estimated marginal mean ± SE for 2 independent measurements of plasma HDL-C obtained from the same individuals at 2 different time points corresponding to the HDL-C levels shown individually in Figure 5, respectively, for the second (1981–1984) and third examinations (1991–1994) of the Copenhagen City Heart Study. Analyses by repeated measures ANOVA. *P < 0.0083 for all ANOVAs when corrected for multiple comparisons using the Bonferroni method. Post hoc tests: **P < 0.05; ^#P < 0.01.