The variable number of tandem repeat polymorphism of platelet glycoprotein Ibα and risk of coronary heart disease

The glycoprotein (GP) Ib-IX-V complex contains 4 polypeptides, GP Ibα, GP Ibβ, GP IX, and GP V.¹  The largest is GP Ibα, which contains the binding site for von Willebrand factor (VWF) and for several other proteins important in the genesis of vascular disease: thrombin, factor XI, P-selectin, and leukocyte Mac-1.^2-5 

GP Ibα is highly polymorphic, one polymorphism being due to variable numbers of tandem repeats (VNTRs) in a region encoding the macroglycopeptide, a mucin-like stalk separating the ligand-binding region from the plasma membrane. This region may have from one to four tandem 39-base repeats, each encoding a 13-amino acid sequence. The 4 variants are named A, B, C, and D, with 4, 3, 2, and 1 repeat, respectively. It has been hypothesized that because added repeats increase the length of the macroglycopeptide region of GP Ibα, they might also increase the sensitivity of the longer molecule to shear stress, a known modulator of the GP Ibα-VWF interaction.¹ 

Here, we analyzed the association of the GP Ibα VNTR alleles with risk of incident coronary heart disease (CHD) in the Atherosclerosis Risk in Communities (ARIC) case-cohort study.

We analyzed the ARIC study population, consisting of a population-based cohort totaling 15 792 men and women between 45 and 64 years of age from 4 US communities.⁶  We defined CHD incidence as (1) a definite or probable myocardial infarction (MI), (2) a silent MI detected by new electrocardiographic (ECG) changes between examinations, (3) a definite CHD death, or (4) a coronary revascularization procedure.

This was a case-cohort study, with frequencies of GP Ibα alleles determined for CHD cases and a stratified random sample of the ARIC cohort. We identified 349 individuals as cases with CHD. For the reference cohort, we oversampled participants with low average carotid intima-media thickness measurements at baseline (< 30th percentile) and also stratified the sampling by age and sex. We included 383 individuals as a reference cohort (of whom 22 were also CHD cases). Before selecting the case-cohort sample, we excluded participants identified as having prevalent CHD, stroke, or transient ischemic neurologic attack (TIA). The study included participants from 2 ethnic groups: whites and African Americans.

We genotyped each individual for the GP Ibα VNTR polymorphism by polymerase chain reaction as described previously.⁷ 

We used analysis of covariance to compute age-, race-, and sex-adjusted frequencies of different genotypes, and means or percentage values of study variables (including lipid profile, plasma VWF and fibrinogen concentrations, history of hypertension, diabetes mellitus, tobacco or alcohol abuse) for CHD cases versus noncases after appropriate weighting for the stratified case-cohort sampling design. We computed the risk ratios and 95% confidence intervals for the time to development of CHD using a weighted proportional hazard regression, accounting for the stratified random sampling and the case-cohort design by the Barlow method.⁸  Because the prevalence of different GP Ibα genotypes was significantly different between whites and African Americans, we performed separate weighted proportional hazard regression analyses for whites and African Americans. In the weighted proportional hazard regression model, we adjusted for sex, age, and the other factors related to CHD. The number of subjects with BB (5 cases and 5 noncases) and DD (1 case and 3 noncases) genotypes were very small and were combined with BC and CD, respectively, in analysis. Only 4 individuals with the A allele were identified, and these were omitted from the analysis.

The association of 3 genotype groups (BC+BB, CC, and CD+DD) with age, sex, ethnicity, and conventional risk factors is shown in Table 1. The CC and CD+DD groups were smaller by percentage in African Americans than whites, with a corresponding increase in the BC+BB group. Aramaki and Reiner have previously shown the higher frequency of the B allele in African Americans compared with whites.⁹ 

In the entire population, we found no significant differences in genotypes between incident CHD cases and noncases (Table 2). However, when we examined the 2 ethnic groups separately, we found an association in African Americans between the non-CC genotypes and incident CHD (Table 2). Put another way, African Americans with the CC genotype had a significantly lower incidence of CHD than those with either the BC+BB or CD+DD genotypes. The weighted proportional hazard regression analysis showed that GP Ibα genotypes correlate with the development of CHD in African Americans (Table 2) In contrast, whites in either the BC+BB or CD+DD groups did not have a significantly higher risk of CHD (risk ratios, 1.1 and 1.4, respectively).

We also analyzed the association according to the presence of the C allele (on the Blood website; see the Supplemental Table link at the top of the online article). The C allele was associated with lower incident CHD in African Americans (P = .003), whereas no such association was noted in the whole group or in whites (P = .091 and .521, respectively). African Americans homozygous for the C allele had significantly lower incident CHD compared to those with one C allele.

We have found a protective effect of the CC genotype in African Americans participating in the ARIC study, but not in the group overall or in whites. This may indicate a interaction between GP Ibα genotype and another environmental or genetic variable that is more prevalent in African Americans than whites. Alternatively, it is possible that the increased prevalence of the B allele in African Americans unmasks an increased risk for cardiovascular disease associated with this variant.

Of interest, 2 GP Ibα haplotypes for GP Ibα have been reported with much higher frequency in African Americans than in whites, the C/Met145 variant and the B/Thr145 variant, present in this population with frequencies of 2.2% and 6.5%, respectively.⁹  Because Met145 has been associated with an increased prevalence and severity of coronary artery disease,¹⁰  the protective effect of the CC genotype in the African-American population is even more striking.

The hypothesis that the length of GP Ibα affects its function, and as a result the propensity to develop CHD, is not a new one.¹¹  Other groups have also studied the role of the GP Ibα VNTR polymorphism in case-control studies, with contradictory results.^10,12-14  All of these studies were cross-sectional case-control studies. The design of these studies makes it less likely that an ideal control group will be chosen, with the case and control subjects being fully matched for all other risk factors. In the current study, 15 792 individuals were followed prospectively over a period of 10 years.

Another mechanism by which the VNTR polymorphism may influence CHD susceptibility is suggested by the results of our study. Most of the individuals in the non-CC group were heterozygous for 2 length variants, either carrying the BC or the CD genotypes. Thus, it is possible that it is the disparity in length between the 2 allele products in a heterozygous individual that influences the interaction of GP Ibα with the receptor (it has been postulated that GP Ibα may be able to associate into dimers or tetramers to form the functional VWF-binding unit¹ ). Nevertheless, because of the relative shortage of BB homozygotes, a larger study is needed to determine whether it is the length of the macroglycopeptide stalk or a length disparity in heterozygotes that determines the increased risk in the non-CC African-American individuals.

The authors thank the staff and participants in the ARIC study for their important contributions.