SEARCH

SEARCH BY CITATION

Keywords:

  • deep vein thrombosis;
  • epidemiology;
  • genetics;
  • genome-wide scan;
  • pulmonary embolism;
  • venous thromboembolism

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References
  10. Supporting Information

Summary.  Objectives:  To identify venous thromboembolism (VTE) disease-susceptibility genes.

Patients and methods:  We performed in silico genome wide association scan (GWAS) analyses using genotype data imputed to approximately 2.5 million single-nucleotide polymorphisms (SNPs) from adults with objectively-diagnosed VTE (n = 1503), and controls frequency matched on age and gender (n = 1459; discovery population). Single-nucleotide polymorphisms exceeding genome-wide significance were replicated in a separate population (VTE cases, n = 1407; controls, n = 1418). Genes associated with VTE were re-sequenced.

Results:  Seven SNPs exceeded genome-wide significance (< 5 × 10−8): four on chromosome 1q24.2 (F5 rs6025 [factor V Leiden], BLZF1 rs7538157, NME7 rs16861990 and SLC19A2 rs2038024) and three on chromosome 9q34.2 (ABO rs2519093 [ABO intron 1], rs495828, rs8176719 [ABO blood type O allele]). The replication study confirmed a significant association of F5, NME7 and ABO with VTE. However, F5 was the main signal on 1q24.2 as only ABO SNPs remained significantly associated with VTE after adjusting for F5 rs6025. This 1q24.2 region was shown to be inherited as a haplotype block. ABO re-sequencing identified 15 novel single nucleotide variations (SNV) in ABO intron 6 and the ABO 3′ UTR that were strongly associated with VTE (< 10−4) and belonged to three distinct linkage disequilibrium (LD) blocks; none were in LD with ABO rs8176719 or rs2519093. Our sample size provided 80% power to detect odds ratios (ORs) = 2.0 and 1.51 for minor allele frequencies = 0.05 and 0.5, respectively (α = 1 × 10−8; 1% VTE prevalence).

Conclusions:  Apart from F5 rs6025, ABO rs8176719, rs2519093 and F2 rs1799963, additional common and high VTE-risk SNPs among whites are unlikely.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References
  10. Supporting Information

Venous thromboembolism (VTE), consisting of deep vein thrombosis (DVT) and its complication, pulmonary embolism (PE), is a common and potentially fatal disease. The overall annual VTE incidence among USA whites is 108 per 100 000 person-years; incidence rates increase exponentially with age for both men and women, and for both DVT and PE [1]. For almost one-quarter of PE patients, the initial clinical presentation is sudden death [2]. Family and twin studies suggest that VTE is highly heritable (h2 = 0.62) and likely results from multigenic action as well as environmental exposures [3–5]. Several genetic disorders or mutations that affect the regulation of hemostasis have been associated with VTE in small case–control studies or observed in high VTE-risk pedigrees [6]. These include mutations that impair the anticoagulant pathway, up-regulate (e.g. prothrombin G20210A) or impair downregulation of the procoagulant pathway (e.g. factor [F]V Leiden), downregulate fibrinolysis and up-regulate innate immunity. In a single-nucleotide polymorphism (SNP)-based candidate gene, case–control study, we tested genetic variation within all known important components of the anticoagulant, procoagulant, fibrinolytic and innate immunity pathways for an independent association with VTE and found one or more SNPs within ABO, F2, F5, F11, KLKB1, SELP and SCUBE1 were significantly associated with VTE, including FV Leiden (F5 rs6025 T [minor] allele), prothrombin G20210A, ABO non-O blood type (ABO rs8176719, odds ratio [OR] = 1.47, = 5.68 × 10−12) and a novel association with ABO rs2519093 (OR=1.68, = 8.08 × 10−16) that was independent of blood type [7]. A genome wide association scan (GWAS) confirmed associations of VTE with FV Leiden and with ABO loci (blood groups non-O and A2) [8], and a second GWAS with a larger sample size and more dense genome-wide scan genotype data found associations of VTE with the F11 and FGG loci [9]. We hypothesized that genes within other pathways are also VTE disease-susceptibility genes. To test this hypothesis, we performed a GWAS study on the same VTE case–control (discovery) population used for our candidate gene study, and replicated all SNPs exceeding genome-wide significance in a separate VTE case–control (replication) population. Finally, we performed gene re-sequencing in order to identify potentially ‘causative’ genetic variation in linkage disequilibrium (LD) with haplotype-tagged SNPs that were associated with VTE.

Patients and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References
  10. Supporting Information

Discovery population

The discovery study population consisted of 1503 VTE cases and 1459 controls (Table S1) from a candidate gene (CG) study, as previously described [7]. A subset of the CG study, 1270 VTE cases and 1302 controls, who consented to sending their leukocyte genomic DNA outside of the Mayo Clinic, and newly recruited cases and controls, have GWS genotype data. The Mayo Clinic Institutional Review Board approved the study protocol as well as the data sharing protocol administered through Database for Genotypes and Phenotypes (dbGAP).

Genotyping

Candidate gene study: leukocyte genomic DNA was extracted, quantified and diluted to the appropriate concentration for Illumina Infinium iSelect genotyping within the Genotyping Core Laboratory of the Mayo Clinic Advanced GenomicsTechnology Center, as previously described [7]. Genotyping controls included 2% sample replicates and a CEPH trio for quality control. In addition, case and control DNA sample addresses were randomly assigned across both the 96-well plate as well as the 12-address iSelect BeadChip (Illumina, Inc., San Diego, CA, USA), ensuring approximately equal numbers of case and control DNA samples by each stratum to avoid potential plate and chip effects, respectively. Genotyping results from high-quality control DNA (SNP call rate ≥ 95%) were used to generate a cluster algorithm. Of the 14 612 SNPs submitted to Illumina from 764 genes within the anticoagulant, procoagulant, fibrinolytic and innate immunity pathways, 1585 SNPs (covering 1100 LD bins) failed manufacture, leaving 13 027 successfully genotyped SNPs. Four additional SNPs were custom genotyped on TaqMan, leaving 13 031 SNPs for analysis.

Genome-wide scan study: the Center for Inherited Disease Research (CIDR), one of two genotyping centers supported by the Gene, Environment Association Studies Consortium (GENEVA) consortium [10], genotyped DNA samples using the Illumina Human660-Quad v.1_A BeadChip (Illumina, Inc.) and DNA from a subset of CG cases and controls that consented to sending their DNA outside Mayo. Similar to the CG study, case and control DNA sample addresses were randomly assigned across 96-well plates provided by CIDR while ensuring roughly equal percentages of cases and controls within each plate. Genotype clusters for each SNP were determined using the Illumina BeadStudio Module (version 3.3.7), and combined intensity data from 96.6% of samples were used to define clusters and call genotypes. Overall, 99.9% of samples attempted (2597 of 2600 total) passed quality-control standards. Genotypes were not called if the quality score from BeadStudio was < 0.15. Both the mean SNP call rates and the mean sample call rates were 99.9%. Genotypes were released for 560 816 SNPs (99.88% of those attempted). Genotypes were not released for autosomal SNPs with call rates < 85%, > 1 HapMap replicate error, > 9% difference in call rate between genders, < 0.2 cluster separation or > 11% difference in heterozygote frequency. Duplicate samples from both HapMap and the study were included on each plate. Four HapMap trios (parents and child) were included to detect Mendelian inheritance errors (i.e. whether the alleles are incorrectly passed from each parent to the child). Reproducibility rates in the raw data were 99.99% among 32 duplicated study subjects. The genotype concordance in the 134 HapMap control samples was 99.8%.

Quality-control steps

For both CG and GWS studies, four main categories of quality-control flags were set for autosomal SNPs: (i) unacceptably high rates of missing genotype calls (> 5%); (ii) low minor allele frequency (MAF) (< 0.005); (iii) unacceptably high rates of Mendelian errors (> 1); and (iv) deviation from Hardy–Weinberg equilibrium (HWE, < E-05). For the CG study, of the SNPs perfoming well (12 477/13 031) 181 SNPs were excluded beacause of MAF < 0.005 (n = 127) or a call rate < 0.95 (n = 54), leaving 12 296 SNPs (covering 10 456 bins) for the association analysis. For the GWS study, we flagged 1965 SNPs with > 5% missing genotypes. A total of 1683 (1115/1135) SNPs were flagged overall (cases/controls) for HWE deviation at < E-05.

Population stratification

To test for ‘cryptic relatedness’ in our population, we conducted a population structure analysis on all participating subjects to document genetic diversity among our population. Using 494 Ancestry Informative Markers (AIMS) available on both our candidate gene and GWA studies [11], we ran STRUCTURE on 2962 participants [12]. The triangle plot provided a graphical representation of genetic structure of our participants plus 209 unrelated individuals from HapMap phase II populations (Yorubans [YRI]; European–Americans from the CEPH collection [CEU]; Chinese from Beijing [CHB]; and Japanese from Tokyo [JPT]), giving a clear sense of how our participants fall among the HapMap reference populations (Fig. S1).

Replication population

Single-nucleotide polymorphisms exceeding genome-wide significance in the discovery population and not in LD within a gene were genotyped in a separate replication population of 2825 individuals (VTE cases, n = 1407; controls, n = 1418; Table S3). Replication study leukocyte genomic DNA samples were largely from Olmsted County, MN, USA residents with an objectively-diagnosed incident VTE over the 45-year period, 1966–2010, and previously-matched Olmsted County resident controls, as previously described [13–15]. The replication case and control sample size was augmented by leukocyte DNA from non-Olmsted County Mayo Clinic patients identified as described for the CG study population [7].

DNA resequencing

We selected a subset of the discovery VTE cases and controls for re-sequencing to take advantage of the joint configuration of two ABO SNPs independently associated with VTE, rs8176719 (ABO exon 6 deletion determining type O blood group) and rs2519093 (ABO intron 1 tag SNP) [7]. We randomly sampled 82 cases and 14 controls within 3 of the 9 potential allele frequency cells (Fig. S8). The rs8176719 alleles are −/− (deletion is the major allele), −/G, and G/G (G is the minor allele). The rs2519093 alleles are G/G (G is the major allele), A/G and A/A (A is the minor allele). For each SNP, the genotypes are represented as 0, 1 or 2 copies of the minor allele. We represented the joint allelic configuration of the two SNPs with the number of copies of the rs8176719 minor allele given first as 0/0 (both with 0 copies of the minor allele), 0/1, 0/2 (0 copies of the rs8176719 minor allele), 1/0 (1 copy of the minor rs8176719 allele), 1/1, and 1/2 and 2/0 (2 copies of the rs8176719 SNP minor allele and 0 copies of the rs2519093 minor allele), 2/1 and 2/2. From Fig. S8, one observes discrepancies between cases and controls at the 0/0, 1/1 and 2/2 combinations. We randomly sampled from these three combinations, taking one-third of the case series. For each SNP, we had 28 cases with 0/0 copies of the minor allele, 27 cases with 1/1 copies of the minor allele; and 27 cases with the combination of 2/2 copies of the minor allele. We compared these 82 cases with 14 controls that do not have any of these combinations.

Capture of the target genomic regions was performed using the Agilent custom eArray. The capturing probes (baits) of 120 bases in length were designed based on the paired-end sequencing protocol with a tiling frequency of 3×. The standard repeat masked regions were avoided based on the definition in the UCSC genome database. The repeat regions are mostly in intronic regions of the genes.

Sequencing was performed using Illumina’s HiSeq 2000 sequencer. Twelve samples were multiplexed in each lane of the eight-lane flow cell, and a total of 96 samples (82 VTE cases and 14 controls) were sequenced. The read qualities were examined by FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) which generates QC matrix from the FASTQ files including per-base sequence qualities, per-sequence quality scores, per-base nucleotide content and sequence duplication levels. The FastQC tool also provides warnings for parameters failing to pass QC thresholds. The paired-end 100-base reads were aligned to human genome build 36 using Burrows Wheeler Aligner (BWA) [16] allowing 4% of editing distance within a seed length of 32. If the sequence duplication levels failed to pass the FastQC threshold, the duplicated reads were removed using the SAMtools’s rmdup method [17]. The BWA alignment was then cleaned up and improved using the Genome Analysis ToolKit (GATK) [18] by recalibration and local re-alignments. Single nucleotide variations (SNVs) were called using SNVMix [19] with a cut-off probability score of 0.8 based on our preliminary testing using a HapMap CEPH subject sequenced by the 1000 genome project (data not shown), and INDELs were called by GATK with default parameters setting.

The read depths of each of the A, C, G, T bases at each variant position, as well as the average mapping quality score were provided by curating the BAM pile-up files using SAMtools [17]. If an identified SNV was a known variant from dbSNP or the 1000 Genomes Project, the allele frequencies of CEU, YRI and CHB/JPT populations from HapMap and 1000 Genome Project were provided. Both SNVs and INDELs were annotated by batch submission to the SeattleSeq server, and for SNVs additional annotations were acquired using a locally or cloud installed SIFT [20]. The SNVs or INDELs within a user-defined distance (default: five bases) to exon–intron boundaries were flagged as potential splice variants and the corresponding transcript IDs are provided.

Information was reported on genes hosting SNVs and INDELs, including (i) the KEGG pathway(s) (http://www.genome.jp/kegg) to which the gene belongs; and (ii) tissue expression specificity of the gene.

Statistical methods

Genome-wide scan (Illumina 660Q; 557 112 SNPs) and CG (n = 764 genes relevant to the anticoagulant, procoagulant, fibrinolytic and innate immunity pathways; n = 12 551 SNPs) genotypes from the discovery population VTE cases (objectively-diagnosed; no cancer, catheter or antiphospholipid antibodies; n = 1503) and controls (frequency-matched on case age, gender, race, MI/stroke status; n = 1459) were merged and imputed to approximately 2.5 million SNPs (i.e. in silico genotypes) with MACH using HapMap Phase II (60 CEU) [21]. All genetic coordinates in tables and figures refer to HapMap release 22 build 36. In regions where no candidate gene genotypes were available, only subjects with GWS data were used for the imputation.

The primary outcome was VTE status, a binary measure. The covariates were age at interview or blood sample collection, gender, stroke/myocardial infarction (MI) status and USA state of residence (Table S1). We tested for an association between each SNP and VTE using unconditional logistic regression, adjusting for age, gender, stroke/MI status and USA state of residence using PLINK v 1.07 [22]. A similar analysis was performed for the replication study and the covariates were age at interview or blood sample collection, sex and USA state of residence (Table S3). Haplotype analyses were performed on the 25 chromosome 1q.24.2 SNPs most significantly associated with VTE, considering 25- and 10-SNP-sliding-windows, as implemented in PLINK v 1.07 [22]. Novel ABO SNVs identified by re-sequencing were tested for an association with VTE using age-, sex-adjusted unconditional logistic regression and Fisher’s exact test. Population attributable risk (PAR) was estimated for F5 rs6025 (FV Leiden), F2 rs1799963 (prothrombin G20210A) and ABO blood type haplotype, as previously performed [7].

Venous thromboembolism prevalence was estimated using the resources of the Rochester Epidemiology Project which has mapped Olmsted County, MN, USA residency for over 500 000 people from 1/1/1966 to 12/31/2010. All Olmsted County residents with objectively-diagnosed, incident VTE over the 40-year period, 1966–2005, have been followed forward in time until death or last follow-up. Venous thromboembolism point prevalence was recorded for 1/1/2006 using these data and adjusted to the USA white population data from the 2000 Census.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References
  10. Supporting Information

Discovery in silico GWAS results

Among the 1503 VTE cases and 1459 controls in the discovery population, the age and gender distributions were similar between cases and controls; however, we found a significant difference between VTE cases and controls for prior stroke/MI; and USA state of residence (Table S1). From the population stratification analysis, 98.64% of samples were classified as European and 1.36% as ‘other’ ancestry (includes individuals of mixed ancestry; Fig. S1). As age, gender, prior stroke or MI have been shown or suggested to be risk factors for VTE [1], and owing to the significant difference between VTE cases and controls regarding USA state of residence, we adjusted for these four characteristics in the logistic regression analyzes.

After adjusting for age, gender, stroke/MI and USA state of residence, 39 SNPs exceeded genome-wide significance ( 5 × 10−8; Fig. 1A and Table S2); F2 rs1799963 (prothrombin G20210A) was borderline significant (OR = 2.46, = 1.7 × 10−6). The quantile–quantile plot of the expected vs. observed genome-wide −log(P-values) (Fig. S2) shows the impact of chromosome regions 1q24.2 and 9q on VTE risk. Table 1 shows the association results for significant SNPs not in LD within a gene (F5 rs6025, BLZF1 rs7538157, NME7 rs16861990, SLC19A2 rs2038024, and ABO rs2519093, rs8176719 [ABO blood type non-O] and rs495828).

image

Figure 1.  (A) Manhattan plot of −log10(P-values) from the case–control association analysis on the merged/imputed venous thromboembolism (VTE) data assuming an additive genetic model. The horizontal line at 1.99 × 10−08 represents Bonferroni’s correction. (B) Location and linkage disequilibrium for single-nucleotide polymorphisms (SNPs) on chromosome 1q24.2 surrounding F5. The black diamond corresponds to F5 6025 (factor V Leiden) the most significant SNP. (C) Location and linkage disequilibrium for SNPs on chromosome 9q surrounding ABO. The black diamond corresponds to the most significant SNP, rs495828. For B,C the linkage disequilibrium measure, r2, between the most significant SNP and any other SNP in the region is symbol coded; the square, star, triangle, circle, and cross symbols correspond to r2 > 0.95, > 0.8 and ≤ 0.95, > 0.5 and ≤ 0.8, > 0.2 and ≤ 0.5, and ≤ 0.2, respectively.

Download figure to PowerPoint

Table 1.   Discovery odds ratio (OR) and 95% confidence intervals (CIs), minor allele, minor allele frequency (MAF) and imputation quality for single-nucleotide polymorphisms (SNPs) showing genome-wide significance, adjusted for age, gender, stroke/myocardial infarction (MI) status and USA state of residence
SNPGeneChromosomePosition*Minor alleleMAFOR (95%CI) P Imputed (r2)/Genotyped
  1. *NCBI build 36 of the human genome. Genotype available in all subjects through the candidate gene data set. Genotype available in only 2570 subjects (87% of our combined samples) through the GWA data set.

SNPs in genes previously reported
 rs6025 F5 1167785673T0.0633.57 (2.76, 4.60)1.68E-22Yes (0.97)/Yes
 rs8176719 ABO 9135122729G0.4191.47 (1.32, 1.64)5.68E-12No/Yes
 rs2519093 ABO 9135131691A0.2431.69 (1.48, 1.91)8.08E-16No/Yes
SNPs in genes not previously reported
 rs7538157 BLZF1 1167618168C0.0622.69 (2.09, 3.45)1.04E-14Yes (0.901)/No
 rs16861990 NME7 1167401751C0.0992.02 (1.66, 2.45)1.69E-12Yes (0.932)/Yes
 rs2038024 SLC19A2 1167722606C0.1771.53 (1.32, 1.78)1.12E-08Yes (0.893)/Yes
 rs495828 ABO 9135144688T0.2721.65 (1.46, 1.86)2.96E-16Yes (0.988)/Yes

Replication and combined results findings

In the separate replication population consisting of 1407 VTE cases and 1418 controls, no significant difference was found between age, gender and USA state of residence (Table S3). Because BLZF1 rs7538157 was in high LD with F5, SLC19A2 and ATP1B1, the BLZF1 SNP (rs7538157) was replaced by ATP1B1 rs12061601 (Fig. 1B). After adjusting for age, gender and USA state of residence, F5 rs6025 (= 1.4 × 10−12), NME7 rs16861990 (= 4.9 × 10−9), and ABO rs2519093 (= 1.2 × 10−17), rs495828 (= 2.4 × 10−17) and rs8176719 (= 5.7 × 10−16), were significantly associated with VTE, whereas ATP1B1 rs12061601 (P = 0.02), F2 rs1799963 (= 0.03) and SLC19A2 rs2038024 (= 0.09) were marginally associated (Table 2). All replicated SNP ORs were in the same direction and of similar magnitude as that of the discovery population.

Table 2.   Replication odds ratio (OR) and 95% confidence intervals (CIs), minor allele, minor allele frequency (MAF; overall and by case/control status) using 2825 subjects (1407 cases and 1418 controls), adjusted for age, gender and USA state of residence
SNPGeneChromosome N MAFMAF-casesMAF-controlsOR (95%CI) P
rs6025 F5 125430.0600.0860.0352.56 (1.97, 3.32)1.40E-12
rs12061601 ATP1B1 125210.1390.1510.1261.21 (1.03, 1.43)0.02
rs16861990 NME7 125590.0970.1220.0711.79 (1.47, 2.18)4.89E-09
rs2038024 SLC19A2 126030.1720.1820.1621.13 (0.98, 1.31)0.09
rs2519093 ABO 925990.2350.2890.1821.80 (1.58, 2.07)1.15E-17
rs495828 ABO 926270.2610.3170.2051.73 (1.52, 1.96)2.38E-17
rs8176719 ABO 925320.4160.4760.3571.63 (1.45, 1.84)5.62E-16
rs1799963 F2 1126200.0190.0240.0151.69 (1.06, 2.41)0.03

These SNPs cluster in two genomic regions located on chromosome 1q24.2 (Fig. 1B) and on chromosome 9q (Fig. 1C). Chromosome 1q24.2 and 9q LD patterns observed in controls are shown in Figs S3,S4, and the ORs and respective 95% confidence intervals for an association of SNPs in these two genomics regions with VTE are depicted in Figs S5,S6. As ATP1B1, NME7, BLZF1 and SLC19A2 are in close proximity to F5, we repeated the association analysis including the F5 rs6025 T (minor) allele (FV Leiden) as a covariate and found that only the ABO SNPs remained significantly associated with VTE at the genome-wide level. Similarly, in the replication study only the ABO SNPs and SLC19A2 rs2038024 remained significantly associated with VTE after adjusting for F5 rs6025 T. Although SLC19A2 rs2038024 was marginally associated with VTE, the OR was in the opposite direction from that of the discovery population, suggesting a spurious association (Table 3).

Table 3.   Replication odds ratio (OR) and 95% confidence intervals (CIs) using 2825 subjects (1407 cases and 1418 controls), adjusted for age, gender, USA state of residence and F5 rs6025 T allele (factor V Leiden)
SNPGene N MAFMAF-casesMAF-controlsOR (95%CI) P
rs12061601 ATP1B1 24770.1370.1500.1250.92 (0.76, 1.10)0.35
rs16861990 NME7 25050.0970.1230.0711.17 (0.89, 1.54)0.25
rs2038024 SLC19A2 24740.1700.1780.1630.77 (0.65, 0.92)0.004
rs2519093 ABO 24740.2350.2890.1811.85 (1.61, 2.13)1.37E-17
rs495828 ABO 25010.2610.3170.2051.76 (1.54, 2.01)3.60E-17
rs8176719 ABO 24770.4160.4740.3591.58 (1.40, 1.78)9.75E-14
rs1799963 F2 24850.0190.0230.0151.71 (1.12, 2.63)0.014

Receiver-operator characteristic (ROC) areas under the curve (AUC), calculated individually and jointly using all significant SNPs, were not different for covariates (age, gender, stroke/MI and USA state of residence) plus F5 rs6025 T (AUC = 0.653) compared with covariates plus F5 rs6025 T, ATP1B1 rs12061601, NME7 rs16861990 and SLC19A2 rs2038024 (AUC = 0.653), whereas the addition of ABO SNPs did increase the AUC (AUC = 0.675; Fig. 2).

image

Figure 2.  Receiver-operator characteristic (ROC) curve analyses of chromosome 1q24.2 and ABO single-nucleotide polymorphisms (SNPs).

Download figure to PowerPoint

Haplotype analysis results

Haplotype analysis of the 25 chromosome 1q.24.2 SNPs (including ATP1B1, NME7, BLZF1, C1orf114, SLC19A2, F5 and SELP; Fig. 1B and Table S4) most significantly associated with VTE showed that only the two haplotypes containing F5 rs6025 T were significantly associated with VTE (Table 4). Together, these data suggest that this chromosome 1q24.2 region is inherited as a haplotype block of approximately 0.52 mb. Among all F5 rs6025 T carriers (N = 369; 280 cases and 89 controls; 186 males and 183 females), 235 (63.7%; 177 cases and 58 controls) carriers inherited at least one of these two 1q24.2 haplotypes. The haplotype containing the F5 rs6025 T and the SELP rs3917862 G (minor) alleles slightly increased the association of the haplotype with VTE (OR = 3.59) compared with the haplotype containing the F5 rs6025 T and the SELP rs3917862 A alleles (OR = 3.16, Table 4). However, a 10-SNP-sliding-window with nine SNPs in F5 (including the F5 rs6025 T allele) and one SNP in SELP (SNP rs3917862 G [minor] allele) considerably increased the association of this haplotype with VTE (OR = 4.55, = 2.07 × 10−13) compared with the association for the haplotype containing the F5 rs6025 T and SELP rs3917862 A alleles (OR = 3.09, = 1.83 × 10−11; Table S5).

Table 4.   Haplotype analysis of the 25 most significant chromosome 1q24.2 single-nucleotide polymorphisms (SNPs) adjusted for age, gender, stroke/myocardial infarction and USA state of residence
Haplotype*†‡FORSTAT P
  1. *The first SNP of the haplotype is ATP1B1 rs12061601, and the last SNP of the haplotype is SELP rs3917862. Allele in bold is F5 rs6025 (T is the minor allele [Factor V Leiden]). Underlined allele is SELP rs3917862 (G is the minor allele).

TATAGGTAGAGCTATGCAGCATGCA0.1360.807.650.006
TATAGGTAGAGCTATGCAGCAGGCA0.4470.922.020.16
TATAAGTAGAGCTATGCAGCATGCA0.07950.823.680.06
TATAAGTAGAGCTATGCAGCAGGCA0.02730.831.040.31
CCCCAACCAGAACTACGGATCGATG0.01853.5928.97.46E-08
TATAAGTCAAAACTACGGATCGACA0.01271.040.03130.86
CCCCAACCAGAACTACGGATCGATA0.02273.1629.26.55E-08
CATAGGTAGAGCTATGCAGCAGGCA0.0290.970.02660.87
TATAGGTAGAGCTATGCAGCAGGCG0.02111.070.1090.74
TATAAGTCAAGCTATGCAGCATGCA0.01960.890.3850.54
TCCAAACAGAGCTATGCAGCAGGCA0.01290.731.510.22
TATAAGTCGAGCTATGCAGCAGGCA0.01621.271.150.28

Re-sequencing results

To further evaluate ABO sequence variation for an association with VTE, we performed ABO re-sequencing, including 10 kb of the flanking regions, using Illumina HiSeq 2000. Excluding the intronic repeat regions, 98% of the targeted area (Figure S7) was sequenced with > 20 × coverage in 96 samples (82 VTE cases and 14 controls). On average, ∼35–40 SNVs and ∼5–10 insertion-deletions (INDELs) were detected in each sample. Fifteen novel SNVs in ABO intron 6 and the ABO 3′ UTR were associated with VTE (all < 10−04) and belonged to three distinctive LD blocks; none were in LD with ABO rs8176719 (blood type O allele) or rs2519093 although ABO rs8176746 (blood type A allele) and rs8176747 (blood type B allele) were in moderate LD (r2 = 0.56–0.65) with novel SNPs in block 3 (Fig. 3; Table S6). Although ABO rs8176746 and rs8176747 did not reach genome wide significance in the primary analysis, a secondary analysis (adjusted for age, gender, stroke/MI and USA state of residence) of the derivation population that considered ABO blood type (either haplotype/genotype or phenotype; VTE cases, n = 1459; controls, n = 1422) found strong associations of ABO blood types A (OR = 1.64; 95%CI: 1.41, 1.91; = 7.83 × 10−11) and O (OR = 0.52; 95%CI: 0.44, 0.60; P = 2.40 × 10−16) with VTE, but no association with blood types B (OR = 1.15; 95%CI: 0.91, 1.44; = 0.25) or AB (OR=1.22; 95%CI: 0.87, 1.70; = 0.25; Table S7). Finally, a significant, protective, frame-shifting single base (G) deletion of one of three guanines at ABO chr9:135120878-80 was found within exon 7. This deletion likely corresponds to ABO rs8176750 (ABO blood-type A2).

image

Figure 3.  Linkage disequilibrium (LD) between novel ABO single nucleotide variants (SNVs) identified by re-sequencing, and ABO rs8176474 (blood type A allele), rs8176746 (blood type B allele), rs8176719 (blood type O allele) and rs2519093 (ABO intron 1), among 82 VTE cases and 14 controls selected based on ABO rs8176719 and rs2519093 genotype (see Patients and methods). The numbers inside each square and the square shading indicate the degree of LD between SNPs and/or SNVs (i.e. black squares with no number indicate complete LD (r2 = 1.0); intensity of the gray shading indicates intermediate values between complete LD (r2 = 1.0) and no LD (r2 = 0.0), and white squares with 0 number indicate no LD).

Download figure to PowerPoint

The individual, joint and group population attributable risk (PAR) were calculated for the risk genotypes for FV Leiden (F5 rs6025), prothrombin G20210A (F2 rs1799963) and ABO blood type haplotype or genotype (Table S8). The highest PAR value was from the ABO blood type non-O, followed by ABO blood type A, FV Leiden, and prothrombin G20210A. The PAR values were very similar between the joint and group estimation methods, with highest joint or group PAR (PAR = 0.47) for ABO blood type non-O, F5 rs6025 and F2 rs1799963.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References
  10. Supporting Information

Our most notable finding was the strong association of VTE with three novel and distinctive ABO LD blocks that were independent of ABO type O blood type. ABO encodes for glycosyltransferases responsible for post translational glycosylation of procoagulant FVIII and von Willebrand factor (VWF) [23]. ABO O blood group phenotype and genotype (i.e. ABO rs8176719) is associated with a decreased risk for VTE [7,8,24], presumably because of decreased plasma levels of FVIII and VWF [25,26]. ABO-mediated glycosylation protects VWF from proteolysis by ADAMTS13 [27] such that individuals with type O blood group (homozygous exon 6 ABO rs8176719 deletion) and hypoglycosylated VWF have ∼25% lower plasma FVIII and VWF levels [24,28].

We previously identified an independent association of an ABO intron 1 SNP (rs2519093) with VTE (OR = 1.68, = 8.08 × 10−16) [7]. This intronic region does not affect RNA splicing or harbor any suppressor RNA elements, nor is it close to any other ABO variants with known function, including variants in ABO exons 6 and 7 that have been variably associated with VTE risk [29–33]. Of the 15 novel ABO SNVs found on resequencing, none were in LD with this ABO intron 1 SNP (Fig. 3; Table S6). Consequently, the findings of this study expand on our previous observation and suggest that this ABO intron 1 region along with an LD block in ABO intron 6 and two LD blocks in the ABO 3′ UTR play important roles in ABO expression and associated VTE risk.

In a secondary analysis, the ABO blood type A haplotype was strongly associated with VTE (Table S7) as previously described [31–34]. In the limited and select sample (n = 96) of our discovery population who underwent ABO resequencing, the ABO A allele was in moderate LD (r2 = 0.56–0.65) with SNVs in ABO LD block 3 (Fig. 3) which potentially could account for the association of these novel block 3 SNVs with VTE. The association of VTE with novel SNVs in ABO LD block 2 was independent of blood type A (rs8176747) and O (rs8176719) alleles, and ABO rs2519093 (Fig. 3). ABO LD block 2 may be important for gene regulation because of a potential open chromatin region as reflected by enrichment for multiple epi-genetic markers in the region (ENCODE: http://www.genome.gov/10005107) a transcription factor binding site (ChIP-seq, ENCODE: http://www.genome.gov/10005107) and evolutionary conservation (UCSC phatCons [35] and phylo [36] algorithms; Fig. 4). In addition, strong evidence from dbEST supports the middle LD block as also lying within a novel transcript, probably an extension of the 3′ of ABO (dbEST: http://www.ncbi.nlm.nih.gov/dbEST/; Fig. 4). ABO SNV_135116452 was not part of any ABO LD block but it was in high LD with SNVs from LD block 3, and in moderate (r2 = 0.54) LD with the ABO blood type A allele (rs8176747; Fig. 3). Thus, the association of SNV_135116452 with VTE may also be because of blood type A. Lastly, the SNVs in ABO LD block 1 are in low LD with all the other SNVs and SNPs (Fig. 3). The associations of SNVs in LD block 1 with VTE are independent of blood type A (r2 ∼ 0.27–0.44) and O (r2 ∼ 0–0.02) alleles, and ABO rs2519093 (r2 ∼ 0–0.04; Fig. 3).

image

Figure 4.  Potential functional roles of ABO SNVs identified on re-sequencing, by linkage disequilibrium (LD) block or deletion.

Download figure to PowerPoint

Our ABO re-sequencing data also support the previously reported association of VTE with ABO blood type A2 (ABO rs8176750) [8,31]. Although we did not directly genotype ABO rs8176750 in either our CG or GWAS studies, one directly-genotyped ABO SNP (rs8176704) in strong LD with rs8176750 (r2 ≥ 0.8) was associated with a significantly reduced risk of VTE (OR = 0.78; 95%CI: 0.63, 0.97, = 0.02) that was of similar magnitude to that reported for ABO blood type A2 [8,31]. The ABO rs8176750 deletion results in a frame-shift that abrogates the wild-type stop codon within exon 7 and likely leads to an elongated transcript (Fig. 4) and an absent or dysfunctional protein similar to blood type O.

Our second notable finding was the discovery that FV Leiden is inherited as a haplotype block that spans seven genes on chromosome 1q24.2, including ATP1B1, NME7, BLZF1, C1orf114, SLC19A2, F5 and SELP. F5 encodes for plasma procoagulant Factor(F) V; FV accelerates activation of prothrombin to thrombin by catalyzing assembly of the prothrombinase complex. FV inactivation is key for down-regulating procoagulant activity. Activated protein C (APC), a serine protease, inactivates FV via cleavage at arginine [R]306, R506 and R679; initial APC cleavage at R506 is required for rapid FV inactivation by APC cleavage at R306 and R679 [37]. F5 rs6025 encodes for substitution of glutamine for R506 [38,39]. Factor V Leiden carriers have normal FV procoagulant activity but impaired FV down-regulation as a result of ‘resistance’ to cleavage/inactivation by APC [40,41]. Factor V Leiden is a founder mutation occurring about 30 000 years ago [42], and is associated with approximately 3.5-fold increased risk for VTE (lifetime VTE penetrance ∼5%) [7,38,43].

Our third finding was that the tag SNP SELP rs3917862 compounds the risk of VTE among FV Leiden carriers. SELP encodes for P-Selectin, a membrane glycoprotein with pro-inflammatory activity that mediates the interaction between activated platelets, leukocytes, immune cells and endothelial cells [44,45]. Cell-surface P-selectin expression and soluble P-selectin concentration are partly controlled by SELP genetic variation [46,47]. Increased soluble P-selectin plasma levels are associated with incident and recurrent VTE [48,49]. We previously showed that SELP rs3917862 was associated with a 1.6-fold increased risk of VTE at a false discovery rate q = 0.006 [7]. SELP is located immediately upstream from F5 (Fig. 1B), and associations between SELP haplotypes and VTE have been attributed to high LD with F5 rs6025 T [50]. In our previous report, while SELP rs3917862 was not in high LD with F5 rs6025 T (r2 = 0.12, D’ = 0.37), the association of rs3917862 with VTE was no longer significant (= 0.3) after controlling for F5 rs6025 T [7]. Our new analyses indicate that the haplotype containing the F5 rs6025 T and the SELP rs3917862 G (minor) alleles increases the association of the haplotype with VTE (OR = 3.59) compared with the haplotype containing the F5 rs6025 T and the SELP rs3917862 A alleles (OR = 3.16, Table 4). Furthermore, the association of the haplotype with VTE increases when considering different numbers of SNPs in the sliding window approach (OR = 4.55 for 10-SNP-sliding-window with nine SNPs in F5 and one in SELP; Table S5). Based on these statistical analyzes, we conclude that the SELP rs3917862 T allele compounds the risk of VTE associated with F5 rs6025 T.

Although we found associations of VTE with ATP1B1, NME7, BLZF1 and SLC19A2, these associations disappeared in both the discovery and replication populations after controlling for F5 rs6025 T (FV Leiden). ATP1B1 encodes for a Na+/K+ ATPase important in endothelial cell and platelet function, NME7 encodes for a nucleoside-diphosphate kinase important in erythropoiesis and CD4 T-cell activation, BLZF1 (JEM-1) encodes for basic leucine zipper nuclear factor 1 which is important for the response of acute promyelocytic leukemia to retinoic acid and SLC19A2 encodes for a thiamine transporter protein (THTR-1) implicated in thiamine-responsive megaloblastic anemia. While the products of these genes are potentially biologically plausible in the causal pathway to VTE, none have been previously associated with VTE. Moreover, ROC curve analyses showed that SNPs within these four genes added no additional information over that provided by patient characteristics and F5 rs6025 T alone. Thus, we believe it is unlikely that these genes are VTE disease-susceptibility genes.

Our sample size provided 80% power to detect ORs of 2.0 and 1.51 for minor allele frequencies (MAF) = 0.05 and 0.5, respectively (α = 5 × 10−8) based on a 1% prevalence of VTE among USA whites for the year 2000 [51]. While single studies or individual-level data meta-analyzes with a larger sample size and/or more dense genome-wide scan genotype data may identify additional SNPs that associate with VTE, we believe that aside from F5 rs6025, ABO rs8176719 and rs2519093, and F2 rs1799963, additional common (i.e. MAF > 0.05) and high VTE-risk (i.e. OR > 2.0) SNPs among persons of non-Hispanic European ancestry are unlikely.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References
  10. Supporting Information

The authors would like to thank C.E. Regnier, J.L. Alkhamis, L.M. Heimer, R.M. Weatherly, R.A. Mueller, D. Tines, A. Xue, R.A. Miller, J.J. Larson, E.N. Jeavons and A.F. Beauseigneur for their excellent technical assistance. All were compensated as part of their regular duties.

We sincerely thank all participants in this study. This project was part of the Gene, Environment Association Studies Consortium (GENEVA) funded by the USA National Human Genome Research Institute (NHGRI) to enhance communication and collaboration among investigators conducting genome-wide studies for a variety of complex diseases. Our group benefited greatly from the work and efforts of the entire consortium, especially the work by the Coordinating Center (directed by B. Weir and C. Laurie of the University of Washington) in data cleaning (D. R. Crosslin) and preparation of these cases and controls for submission to the Database for Genotypes and Phenotypes (dbGAP; S. Bennett). We also acknowledge the leadership of T. Manolio of NHGRI. Genotyping services were provided by the Center for Inherited Disease Research (CIDR), funded through a federal contract from the USA National Institutes of Health (NIH) to Johns Hopkins University.

This work was supported by the National Institutes of Health, National Heart, Lung and Blood Institute (grant numbers HL66216, HL83141), National Human Genome Research Institute (grant number HG04735), National Institute on Aging (grant number AG0346760) and National Cancer Institute (grant number CA92153), U.S. Public Health Service; the Centers for Disease Control and Prevention (grant number DD000235); the Mayo Clinic Center for Individualized Medicine; and by Mayo Foundation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References
  10. Supporting Information
  • 1
    Heit JA. The epidemiology of venous thromboembolism in the community. Arterioscler Thromb Vasc Biol 2008; 28: 3702.
  • 2
    Heit JA, Silverstein MD, Mohr DN, Petterson TM, O’Fallon WM, Melton LJI. Predictors of survival after deep vein thrombosis and pulmonary embolism: a population-based, cohort study. Arch Intern Med 1999; 159: 44553.
  • 3
    Souto JC, Almasy L, Borrell M, Blanco-Vaca F, Mateo J, Soria JM, Coll I, Felices R, Stone W, Fontcuberta J, Blangero J. Genetic susceptibility to thrombosis and its relationship to physiological risk factors: the GAIT study. Genetic analysis of idiopathic thrombophilia. Am J Hum Genet 2000; 67: 14529.
  • 4
    Heit JA, Phelps MA, Ward SA, Slusser JP, Petterson TM, De Andrade M. Familial segregation of venous thromboembolism. J Thromb Haemost 2004; 2: 7316.
  • 5
    Larsen TB, Sorensen HT, Skytthe A, Johnsen SP, Vaupel JW, Christensen K. Major genetic susceptibility for venous thromboembolism in men: a study of Danish twins. Epidemiology 2003; 14: 32832.
  • 6
    Heit JA. Thrombophilia: clinical and laboratory assessment and managment. In: Kitchens C, Alving B, Kessler C, eds. Consultative Hemostasis and Thrombosis. Philadelphia, PA: W. B. Saunders, 2007: 21344.
  • 7
    Heit JA, Cunningham JM, Petterson TM, Armasu SM, Rider DN, DE Andrade M. Genetic variation within the anticoagulant, procoagulant, fibrinolytic and innate immunity pathways as risk factors for venous thromboembolism. J Thromb Haemost 2011; 9: 113342.
  • 8
    Tregouet DA, Heath S, Saut N, Biron-Andreani C, Schved JF, Pernod G, Galan P, Drouet L, Zelenika D, Juhan-Vague I, Alessi MC, Tiret L, Lathrop M, Emmerich J, Morange PE. Common susceptibility alleles are unlikely to contribute as strongly as the FV and ABO loci to VTE risk: results from a GWAS approach. Blood 2009; 113: 5298303.
  • 9
    Germain M, Saut N, Greliche N, Dina C, Lambert JC, Perret C, Cohen W, Oudot-Mellakh T, Antoni G, Alessi MC, Zelenika D, Cambien F, Tiret L, Bertrand M, Dupuy AM, Letenneur L, Lathrop M, Emmerich J, Amouyel P, Tregouet DA, et al. Genetics of venous thrombosis: insights from a new genome wide association study. PLoS ONE 2011; 6: e25581.
  • 10
    Cornelis MC, Agrawal A, Cole JW, Hansel NN, Barnes KC, Beaty TH, Bennett SN, Bierut LJ, Boerwinkle E, Doheny KF, Feenstra B, Feingold E, Fornage M, Haiman CA, Harris EL, Hayes MG, Heit JA, Hu FB, Kang JH, Laurie CC, et al. The Gene, Environment Association Studies consortium (GENEVA): maximizing the knowledge obtained from GWAS by collaboration across studies of multiple conditions. Genet Epidemiol 2010; 34: 36472.
  • 11
    Seldin MF, Shigeta R, Villoslada P, Selmi C, Tuomilehto J, Silva G, Belmont JW, Klareskog L, Gregersen PK. European population substructure: clustering of northern and southern populations. PLoS Genet 2006; 2: e143.
  • 12
    Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics 2000; 155: 94559.
  • 13
    Heit JA, Silverstein MD, Mohr DN, Petterson TM, O’Fallon WM, Melton LJI. Risk factors for deep vein thrombosis and pulmonary embolism: a population-based case-control study. Arch Intern Med 2000; 160: 80915.
  • 14
    Heit JA, Petterson TM, Owen WG, Burke JP, de Andrade M, Melton LI. Thrombomodulin gene polymorphisms or haplotypes as potential risk factors for venous thromboembolism: a population-based case-control study. J Thromb Haemost 2005; 3: 7107.
  • 15
    Heit JA, Leibson CL, Ashrani AA, Petterson TM, Bailey KR, Melton LJ III. Is diabetes mellitus an independent risk factor for venous thromboembolism?: a population-based case-control study. Arterioscler Thromb Vasc Biol 2009; 29: 1399405.
  • 16
    Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 2009; 25: 175460.
  • 17
    Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G, Durbin R. The Sequence Alignment/Map format and SAMtools. Bioinformatics 2009; 25: 20789.
  • 18
    McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, Garimella K, Altshuler D, Gabriel S, Daly M, DePristo MA. The genome analysis toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res 2010; 20: 1297303.
  • 19
    Goya R, Sun MG, Morin RD, Leung G, Ha G, Wiegand KC, Senz J, Crisan A, Marra MA, Hirst M, Huntsman D, Murphy KP, Aparicio S, Shah SP. SNVMix: predicting single nucleotide variants from next-generation sequencing of tumors. Bioinformatics 2010; 26: 7306.
  • 20
    Kumar P, Henikoff S, Ng PC. Predicting the effects of coding non-synonymous variants on protein function using the SIFT algorithm. Nat Protoc 2009; 4: 107381.
  • 21
    Li Y, Willer C, Sanna S, Abecasis G. Genotype imputation. Annu Rev Genomics Hum Genet 2009; 10: 387406.
  • 22
    Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, Maller J, Sklar P, de Bakker PI, Daly MJ, Sham PC. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007; 81: 55975.
  • 23
    Matsui T, Titani K, Mizuochi T. Structures of the asparagine-linked oligosaccharide chains of human von Willebrand factor. Occurrence of blood group A, B, and H(O) structures. J Biol Chem 1992; 267: 872331.
  • 24
    Ohira T, Cushman M, Tsai MY, Zhang Y, Heckbert SR, Zakai NA, Rosamond WD, Folsom AR. ABO blood group, other risk factors and incidence of venous thromboembolism: the Longitudinal Investigation of Thromboembolism Etiology (LITE). J Thromb Haemost 2007; 5: 145561.
  • 25
    Kamphuisen PW, Eikenboom JC, Rosendaal FR, Koster T, Blann AD, Vos HL, Bertina RM. High factor VIII antigen levels increase the risk of venous thrombosis but are not associated with polymorphisms in the von Willebrand factor and factor VIII gene. Br J Haematol 2001; 115: 1568.
  • 26
    Tsai AW, Cushman M, Rosamond WD, Heckbert SR, Tracy RP, Aleksic N, Folsom AR. Coagulation factors, inflammation markers, and venous thromboembolism: the longitudinal investigation of thromboembolism etiology (LITE). Am J Med 2002; 113: 63642.
  • 27
    McGrath RT, McKinnon TA, Byrne B, O’Kennedy R, Terraube V, McRae E, Preston RJ, Laffan MA, O’Donnell JS. Expression of terminal alpha2-6-linked sialic acid on von Willebrand factor specifically enhances proteolysis by ADAMTS13. Blood 2010; 115: 266673.
  • 28
    Smith NL, Chen MH, Dehghan A, Strachan DP, Basu S, Soranzo N, Hayward C, Rudan I, Sabater-Lleal M, Bis JC, de Maat MP, Rumley A, Kong X, Yang Q, Williams FM, Vitart V, Campbell H, Malarstig A, Wiggins KL, van Duijn CM et al. Novel associations of multiple genetic loci with plasma levels of factor VII, factor VIII, and von Willebrand factor: the CHARGE (Cohorts for Heart and Aging Research in Genome Epidemiology) Consortium. Circulation 2010; 121: 138292.
  • 29
    Wu O, Bayoumi N, Vickers MA, Clark P. ABO(H) blood groups and vascular disease: a systematic review and meta-analysis. J Thromb Haemost 2008; 6: 629.
  • 30
    Mercier B, Oger E, Le Gal G, Mottier D, Ferec C. Phenotypic but not allelic ABO blood group association with risk of venous thrombosis. Thromb Haemost 2005; 93: 3889.
  • 31
    Morelli VM, De Visser MC, Vos HL, Bertina RM, Rosendaal FR. ABO blood group genotypes and the risk of venous thrombosis: effect of factor V Leiden. J Thromb Haemost 2005; 3: 1835.
  • 32
    Tirado I, Mateo J, Soria JM, Oliver A, Martinez-Sanchez E, Vallve C, Borrell M, Urrutia T, Fontcuberta J. The ABO blood group genotype and factor VIII levels as independent risk factors for venous thromboembolism. Thromb Haemost 2005; 93: 46874.
  • 33
    Paiva SG, Sabino AP, Carvalho MG, Ribeiro DD, Gomes KB, Santos MS, Oliveira MS, Lages GG, Dusse LM, Fernandes AP. Polymorphisms in exons 6 and 7 of the ABO locus and their association with venous thrombosis in young Brazilian patients. Blood Coagul Fibrinolysis 2009; 20: 1228.
  • 34
    Wiggins KL, Smith NL, Glazer NL, Rosendaal FR, Heckbert SR, Psaty BM, Rice KM, Lumley T. ABO genotype and risk of thrombotic events and hemorrhagic stroke. J Thromb Haemost 2009; 7: 2639.
  • 35
    Felsenstein J, Churchill GA. A Hidden Markov Model approach to variation among sites in rate of evolution. Mol Biol Evol 1996; 13: 93104.
  • 36
    Cooper GM, Stone EA, Asimenos G, Green ED, Batzoglou S, Sidow A. Distribution and intensity of constraint in mammalian genomic sequence. Genome Res 2005; 15: 90113.
  • 37
    Kalafatis M, Bertina RM, Rand MD, Mann KG. Characterization of the molecular defect in factor VR506Q. J Biol Chem 1995; 270: 40537.
  • 38
    Juul K, Tybjaerg-Hansen A, Schnohr P, Nordestgaard BG. Factor V Leiden and the risk for venous thromboembolism in the adult Danish population. Ann Intern Med 2004; 140: 3307.
  • 39
    Bertina RM, Koeleman BP, Koster T, Rosendaal FR, Dirven RJ, de Ronde H, van der Velden PA, Reitsma PH. Mutation in blood coagulation factor V associated with resistance to activated protein C. Nature 1994; 369: 647.
  • 40
    Dahlback B, Carlsson M, Svensson PJ. Familial thrombophilia due to a previously unrecognized mechanism characterized by poor anticoagulant response to activated protein C: prediction of a cofactor to activated protein C. Proc Natl Acad Sci U S A 1993; 90: 10048.
  • 41
    Martinelli I, Bottasso B, Duca F, Faioni E, Mannucci PM. Heightened thrombin generation in individuals with resistance to activated protein C. Thromb Haemost 1996; 75: 7035.
  • 42
    Zivelin A, Griffin JH, Xu X, Pabinger I, Samama M, Conard J, Brenner B, Eldor A, Seligsohn U. A single genetic origin for a common Caucasian risk factor for venous thrombosis. Blood 1997; 89: 397402.
  • 43
    Heit JA, Sobell JL, Li H, Sommer SS. The incidence of venous thromboembolism among Factor V Leiden carriers: a community-based cohort study. J Thromb Haemost 2005; 3: 30511.
  • 44
    Andre P. P-selectin in haemostasis. Br J Haematol 2004; 126: 298306.
  • 45
    Nurden AT. Platelets, inflammation and tissue regeneration. Thromb Haemost 2011; 105(Suppl 1): S1333.
  • 46
    Volcik KA, Catellier D, Folsom AR, Matijevic N, Wasserman B, Boerwinkle E. SELP and SELPLG genetic variation is associated with cell surface measures of SELP and SELPLG: the Atherosclerosis Risk in Communities Carotid MRI Study. Clin Chem 2009; 55: 107682.
  • 47
    Marteau JB, Lambert D, Herbeth B, Marie B, Droesch S, Tregouet DA, Visvikis-Siest S. P-selectin polymorphisms’ influences on P-selectin serum concentrations and on their familial correlation: the STANISLAS family study. J Thromb Haemost 2008; 6: 9207.
  • 48
    Blann AD, Noteboom WM, Rosendaal FR. Increased soluble P-selectin levels following deep venous thrombosis: cause or effect? Br J Haematol 2000; 108: 1913.
  • 49
    Kyrle PA, Hron G, Eichinger S, Wagner O. Circulating P-selectin and the risk of recurrent venous thromboembolism. Thromb Haemost 2007; 97: 8803.
  • 50
    Uitte de Willige S, De Visser MC, Vos HL, Houwing-Duistermaat JJ, Rosendaal FR, Bertina RM. Selectin haplotypes and the risk of venous thrombosis: influence of linkage disequilibrium with the factor V Leiden mutation. J Thromb Haemost 2008; 6: 47885.
  • 51
    Skol A, Scott LJ, Abecasis G, Boehnke M. Joint analysis is more efficient than replication-based analysis for two-stage genome-wide association studies. Nat Genet 2006; 38: 20913.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References
  10. Supporting Information

Figure S1. STRUCTURE triangle plot of 2962 discovery study individuals and 209 unrelated individuals from HapMap Phase II populations (YRI – cluster 1, CEU – cluster 2 and CHB/JPT – cluster 3) using 494 ancestry informative markers.

Figure S2. Q-Q plot for –log10 (P-values) of all autosomal SNPs not flagged for QC metrics among 2962 VTE cases and controls.

Figure S3. Linkage disequilibrium (r2) patterns among controls subjects for markers on chromosome 1q24.2 showing evidence of an association with venous thromboembolism at the genome-wide level of significance.

Figure S4. Linkage disequilibrium (r2) patterns among controls subjects for SNPs in ABO showing evidence of association with venous thromboembolism at the genome-wide level of significance.

Figure S5. The estimated odds ratios (case) and corresponding 95% confidence intervals under an additive model for an association of SNPs in chromosome 1.q24.2 with venous thromboembolism by SNP physical position.

Figure S6. The estimated odds ratios (case) and corresponding 95% confidence intervals under an additive model for an association of ABO SNPs with venous thromboembolism by SNP physical position.

Figure S7. ABO resequencing coverage of the top ABO SNPs from the genome-wide association results.

Figure S8. Number of cases (total n = 1488) and controls (total n = 1439) with the rare alleles of ABO rs2519093 and ABO rs8176719, represented as 0, 1, and 2 versions of the rare allele for each SNP.

Table S1. Demographic and clinical characteristics for venous thromboembolism cases and controls in the discovery population.

Table S2. In silico analyses of merged genome-wide scan (GWS) and candidate gene (CG) SNPs with < 10−5 for an association with venous thromboembolism, adjusted for age, sex, stroke or myocardial Infarction, and USA state of residence.

Table S3. Demographic and clinical characteristics for venous thromboembolism cases and controls in the replication population.

Table S4. Description of the 25 most significant chromosome 1q24.2 SNPs associated with venous thromboembolism, including minor allele, minor allele frequency (MAF), odds ratio (OR) and P-value, adjusted for age, sex, stroke or myocardial infarction, and USA state of residence.

Table S5. 10-SNP-sliding-window haplotype analysis of the 25 most significant chromosome 1q24.2 SNPs associated with venous thromboembolism, adjusted for age, sex, stroke or myocardial infarction, and USA state of residence.

Table S6. ABO single nucleotide variants (SNVs) significantly associated with venous thromboembolism by linkage disequilibrium blocks displayed in Figure 3.

Table S7. Association analysis results between the four ABO blood types (i.e. ABO blood type haplotype/genotype or phenotype) and venous thromboembolism, adjusted for age, sex, state of residence, and prior stroke or myocardial infarction.

Table S8. Attributable risk (AR) and associated 95% confidence intervals (95%CI) for F5 rs6025 (Factor V Leiden), F2 rs1799943 (prothrombin G20210A) and ABO blood type haplotype or genotype assuming an additive model for each SNP or haplotype.

FilenameFormatSizeDescription
JTH_4810_sm_TableS1-S8-FigureS1-S8.doc2440KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.