Retrospective Analysis of Coagulation Factor II Receptor (F2R) Sequence Variation and Coronary Heart Disease in Hypertensive Patients
Objectives— The purpose of this study was to evaluate the role of genetic variants within the coagulation factor II receptor (F2R) in the occurrence of coronary heart disease (CHD).
Methods and Results— Four SNPs (−1738 G/A, 2860 G/A, 2930 T/C, and 9113 C/A) and an ins/del polymorphism −506-/GGCCGCGGGAAGC (D/I), replicating a consensus sequence for Ets-1 transcription factor, and their related haplotypes were tested for association to CHD in 1600 hypertensive patients divided in 2 groups according to presence (cases, n=559) and absence (controls, n=1041) of CHD. Allele I at −506 locus was associated with increased risk of CHD under additive, dominant, and recessive models of inheritance (all P<0.01). Three haplotypes carrying I allele were consistently associated with an increased risk of CHD (all P<0.05). Patients homozygous for the C allele at the 2930 locus also showed an increased risk of CHD (P<0.05). To test the functionality of −506 locus, nuclear extracts were incubated with −506D and −506I sequences by EMSA and F2R promoter activity (F2R-A) were assessed in HUVECs transfected with vectors carrying −506D and −506I sequences and exposed to hypoxia. Presence of the −506I sequence was associated with a 26% reduction of affinity binding to nuclear proteins and to blunted F2R-A in response to hypoxia as compared with the −506D sequence (all P<0.05).
Conclusions— F2R genetic variants may influence the natural history of CHD in patients at high risk of cardiovascular events.
Coronary Heart Disease (CHD), a complex and heterogeneous disease with a heritability exceeding 50%,1,2 represents the leading cause of death among cardiovascular diseases, and almost half of the patients experiencing a first heart attack have high blood pressure levels.3 Individual susceptibility to CHD depends on multiple genetic and environmental factors that influence coronary artery lesion morphology and metabolism and therefore clinical presentation. Identification of genetic variants predisposing to CHD is essential to improve our knowledge of the mechanisms underlying CHD and, on a broader level, to ameliorate cardiovascular disease prevention and pharmacological treatment.
Recently, the Coagulation Factor II Receptor (F2R), the thrombin receptor expressed on human platelets and endothelial cells, has been identified as a powerful modulator of endothelial lining, and more generally, of vascular wall integrity and stability.4,5 In aortic6 and coronary artery rings7 thrombin-activated F2R promotes endothelial-dependent vasodilatation through a nitric oxide-dependent mechanism and infusion of an F2R agonist in mice results in rapid hypotension followed by sustained hypertension.8 In cultured endothelial cells prolonged F2R stimulation induces an inflammatory and angiogenic response through the enhanced expression of proinflammatory cytokines,9vascular endothelial growth factor (VEGF) and VEGF receptors expression.10 Finally, experiments performed in animal model of vascular injury have suggested that inhibition of this receptor might be of clinical relevance in preventing acute arterial thrombosis11 and vascular wall remodeling.12 Taken together these observations strongly support F2R as an attractive candidate gene in the pathogenesis of CHD. A progressive loss of the endothelial lining anatomic and functional integrity represents, in fact, a hallmark of hypertension and a condition underlying atherosclerosis and strongly promoting CHD.
To date the potential role of F2R sequence variation in the occurrence of cardiovascular diseases has only been investigated in 2 published case-control studies,13,14 both failing to find any association between F2R and the occurrence of myocardial infarction.
In the present study we investigated the potential role of genetic variants within the F2R gene in the occurrence of CHD in a population of patients at high risk of cardiovascular events through an integrated genetic and biological approach. We report the analysis of 5 genetic variants and related haplotypes with the occurrence of CHD and the characterization of the biological effect of −506I, the allelic variant associated to the highest risk of CHD.
Methods are described in details in the supplemental materials, available online at http://atvb.ahajournals.org.
Study participants consisted of 1600 patients affected by essential hypertension retrospectively divided in 2 groups according to the presence (cases, n=559) or absence (controls, n=1041) of CHD. Chronic stable angina (SA) was diagnosed in 238 patients and a former acute coronary syndrome event (ACS) in 321 patients.
Single Nucleotide Polymorphisms Genotyping and Haplotype Generation
Six single nucleotide polymorphisms (SNPs) were identified from the HapMap database (public release#11, 09/02/2004; www.hapmap.org) on sequence with GenBank accession no. AF391809: rs2227744 (−1738 G/A, nucleotide [n]: 352), rs27135 (G/A n: 2860), rs27593 (T/C n: 2930), r2227774 (C/A n: 9113), rs153311 (A/C n:17841) and rs2227800 (C/T n:19390). Two additional SNPs reported as polymorphic in the NCBI database (www.ncbi.nlm.nih.gov/), rs11267092 (−506D/I, −/GGCCGCGGGAAGC, n:1631 to 1643) and rs2230849 (T/A n: 18627), were included in the analysis. Validation assays confirmed that −1738 G/A, −506D/I, 2860 G/A, 2930 T/C, and 9113 C/A were polymorphic in this population.
Haplotype counts and frequencies were inferred from original genotype data using the THESIAS program.15
Assessment of −506 Locus Biological Function
Biological properties of the two alleles at the −506 locus were assessed by electrophoretic mobility shift assay (EMSA) and by luciferase activity (LA) assay.
Nuclear extracts from a human cervical cancer cell line (Hela) (Santa Cruz) were incubated with 32P- labeled probes encoding −506D, −506I, and an Ets-1 consensus sequence. Complexes were run on 8% polyacrilamyde gel and visualized by autoradiography.
Pooled human umbilical vein endothelial cells (HUVECs; Cambrex) were transiently transfected with empty, p-506D and p-506I vectors and stimulated for 4, 6, and 8 hours with Angiotensin II (ATII, 10−6 mol/L) isoprotenerol (Is, 10−7 M), or incubated in hypoxic conditions and left to recover overnight in normoxic conditions. LA, normalized by β galactosidase light units in the same sample, was expressed as fold increases in relative LA versus p-506D baseline values.
Differences in genotypic, allelic, and haplotype frequencies among cases and controls were calculated by χ2 test. Haplotype counts and frequencies were inferred from genotypes. Unadjusted odds ratios (ORs) with 95% confidence interval (CI) were estimated for −506 D/I and 2930 T/C variations and for haplotype configurations. Age, male sex, and diabetes were identified as independent predictors of CHD and treatment with angiotensin receptor blockers (ARBs), diuretics, and dihydropyridine Ca++ antagonists (DHP Ca++ ant) as potential confounders and were adjusted for in a multivariable logistic regression analysis.
In vitro assays were performed 3 times in duplicate. Results are expressed as mean±SEM. Differences among groups were evaluated using the Kruskall-Wallis H test.
Demographic, anthropometric, biochemical, and therapeutic data collected in the present study population are outlined in Table 1. Family history of hypertension was more common among controls, whereas history of CHD was more frequent in cases. At diagnosis, diastolic blood pressure (DBP) was higher in controls (P<0.0001). However, when blood pressure values were obtained throughout the course of years, both groups showed comparable systolic blood pressure (SBP) and DBP values. Compared with controls, cases were older (P<0.0001), predominantly of male sex (P<0.0001), and current smoking (P<0.05) and overt diabetes mellitus (P<0.0001) were more frequently observed. No differences were noted in the occurrence of hypercholesterolemia between the 2 groups of patients. However, among cases, hypercholesterolemia was more frequent in SA patients as compared with patients who had suffered ACS (χ22DF:6.53 P=0.03).
The following risk factors were identified as independent predictors of CHD (OR, (95%CI)): age 1.02 (1.01 to 1.03), male sex 1.81 (1.37 to 2.39), and diabetes mellitus 3.98 (2.78 to 5.69) and pharmacological treatment with ARBs 0.54 (0.44 to 0.67), DHP Ca++ ant 0.55 (0.44 to 0.68), and diuretics 0.55 (0.45 to 0.67) were identified as potential confounders in genotype/phenotype association study and adjusted for in a multivariable logistic regression analysis.
F2R SNPs and Haplotypes and the Risk of CHD
Genotype and allele frequencies at −1738 G/A, −506D/I, 2860 G/A, 2930 T/C, and 9113 C/A loci are reported in Table 2. No significant deviation from Hardy-Weinberg equilibrium was observed in controls or cases at any of the 5 loci. LD data are reported in supplemental Table II.
At −1738 G/A, 2860 G/A, and 9113 C/A, no significant differences were found in genotype and allele frequencies among cases and controls. At −506D/I locus, genotype II was more frequent in cases (0.11 versus 0.05 in controls, P<0.0001) with allele I having a frequency of 0.31 in cases and 0.22 in controls (P<0.0001) and at 2930T/C locus frequency of C allele was higher in cases (0.25 versus 0.22, P=0.02). We tested the hypothesis of allele I and C being associated with an increased OR of CHD. As shown in Table 3, presence of the I allele at −506 locus was associated with an increased risk, OR (95%CI), of CHD assuming an additive (unadjusted 1.24 [1.11 to 1.39]; adjusted 1.25 [1.10 to 1.42]), dominant (unadjusted 1.68 [1.36 to 2.09]; adjusted 1.72 [1.35 to 2.19]), and recessive (unadjusted 2.40 [1.61 to 3.58]; adjusted 2.63 [1.67 to 4.13]) models of inheritance. At the 2930 locus, patients homozygous for the C allele were at increased risk of CHD at both unadjusted 1.62 (1.04 to 2.53) and adjusted 1.65 (1.00 to 2.73) logistic regression analyses, whereas no significant association was found under additive or dominant models of inheritance.
Neither of these two loci showed a significant association with the presence of any of the cardiovascular risk factors identified in this population and no epistatic interactions were found.
Overall, 32 different haplotypes were inferred from the original genotype data. Table 4 lists counts and frequencies of the haplotypes observed in cases and controls with a frequency higher than 0.01 as well as the OR (95%CI) associated to each haplotype. Haplotypes bearing the I allele at −506 locus were more frequent in cases. The effect of the I allele was associated with an increased risk of CHD in patients carrying Hap2 GIGTA, Hap4 AIATC, and Hap8 GIGCC in both un- and adjusted logistic regression analyses (all P<0.05). The I allele, as compared with D allele, was associated with an OR 1.77 (1.19 to 2.62) P=0.004 on the G-GTA background; to an OR of 1.95 (1.28 to 2.97) P=0.001 on A-ATC background; and to an OR of 2.79 (1.4 to 5.58) P=0.003 on G-GCC background. No apparent increase in the risk of CHD was associated with the presence of allele I in any of the other haplotypic configurations in this population. Hap8 represents the haplotype bearing I and C allele with the highest frequency, all the other haplotypes having a frequency ranging from 0.0003 to 0.003.
Consistently, haplotypes carrying the discordant allele at −506 locus Hap0 ADATC, Hap3 GDGTA, and Hap6 GDGTC and Hap9 ADATA were more frequently observed in controls. At multivariable analysis Hap6 was associated with a reduced risk of CHD (OR: 0.41 [0.21 to 0.80]). At the analysis of the haplotypic background, however, the protective effect seemed to be related to the presence of T allele on a GDG-C background with an OR 0.48 (0.30 to 0.76) P=0.005; presence of either D or I allele did not change, in fact, the OR associated to the G-GTC haplotype (OR:1.21 [0.52 to 2.80] P=0.65).
Assessment of −506 Locus Biological Function
Given −506I being the allelic variant showing the strongest effect on CHD at the genotype/phenotype association analyses, we assessed the functionality of −506D and −506I sequences by means of EMSA. Thus, we tested the hypothesis that sequence variation at the −506 locus might influence F2R-A in response to ATII, Is, and hypoxia, 3 stimuli mimicking a microenvironment that favors development and progression of CHD.
The −506I sequence (Figure 1, L5) showed a 26% reduction in the affinity binding to nuclear proteins as compared with −506D (Figure 1, L3; ADU±SEM; 88804±4354 versus 119717±4975, P<0.05). The Ets-1 consensus sequence was used as positive control (L1). Specificity of the reaction was controlled by adding a molar excess of the unlabeled Ets-1 and −506D probes (L6 and L7).
F2R-A was selectively increased in HUVECs transfected with p-506D (4.44±0.82 LA fold increase versus baseline, P<0.05) after 8 hours of hypoxia, a change that reverted after exposure to normoxic conditions (Figure 2C, gray bars). In contrast, cells transfected with p-506I showed a blunted response of F2R-A (Figure 2C, black bars). No differences in F2R-A were observed among HUVECs transfected with p-506D or p-506I in response to ATII or Is stimulation (Figure 2A and 2B, respectively).
In the present study we have explored the potential role of genetic variants within F2R gene in the occurrence of CHD in a population of hypertensive patients. Our findings demonstrate that 2 SNPs, mapping in the F2R promoter region at locus −506, and in the intron at position 2930, and 3 related haplotypes are associated with an increased risk of CHD. Presence of the I allele at the −506 locus was the strongest genetic predictor of CHD in this cohort. Within the 13 base pairs replicated in the insertion lies a consensus sequence for Ets-1 transcription factor.16 Ets-1 is a strong activator of endothelial cell specific promoters, and it has been recently described as one of the major regulators of hypoxia/ischemia-driven transcription.17 This polymorphism occurs in a conserved region of the promoter, and there is evidence for at least one expressed sequence tag (EST) mapping directly over this position, indicating that it could regulate alternative promoter usage. To explore the biological relevance of this polymorphism we have tested if the alternative alleles at the −506 locus might display different biological properties. Our results show that presence of the I allele at −506 locus reduces binding affinity to nuclear proteins and is associated with a blunted response to hypoxia.
A crucial role of hypertension in the occurrence of CHD has been highlighted in epidemiological as well as clinical studies.18 However, differences in the individual response to blood pressure lowering drugs and copresence of other risk factors determine major differences in the rate of occurrence of CHD among hypertensive patients. Hypertension represents a major confounding factor in observational genetic studies as well as in the analysis of the mechanisms regulating vascular wall homeostasis. The analysis of the potential role of candidate genes in the occurrence of complex disease in patients selected for an intermediate phenotype may represent a suitable approach to reduce the distance between genes and complex phenotypes.19,20
In this population, age, male sex, and diabetes were identified as independent predictors of CHD, whereas hypercholesterolemia did not show a significant association with CHD. This apparent paradox has already been observed in other studies21 and related to the use of drugs that lower serum cholesterol levels making the discrimination between exposed and nonexposed individuals more difficult to establish. In our population, the percentage of patients taking lipid lowering drugs, as well as the serum lipid profiles, were comparable in cases and controls, but hypercholesterolemia was more frequently observed in the subgroup of patients affected by SA. We cannot exclude that, given the complexity and heterogeneity of CHD, serum cholesterol levels may differently influence various forms of CHD. Likewise, serum lipid profiles differentially influence the occurrence of various stroke subtypes.22
The F2R gene spans 22 Kb on the long arm on human chromosome 5 in position 5q1.3. It is composed of an upstream regulatory region and 2 exons separated by a large intron. The analysis of 5 SNPs within F2R gene in our study population has identified two loci, −506 and 2930, whose allelic frequencies were significantly different among cases and controls. The 2930T/C polymorphism influenced the risk of CHD only in patients carrying variant C in homozygosity, whereas allele I was consistently associated with an increased risk of CHD in hypertensive patients both in genotype and haplotype analyses. Patients carrying haplotypes identified as Hap2, Hap4, and Hap8 were exposed to an increased risk of CHD and we cannot exclude that in a longitudinal observational study these patients might exhibit differences with regard to prognosis or to response to pharmacological treatment.
An analysis of the potential role of the −506D/I polymorphism in the occurrence of myocardial infarction did not show any significant association in a French population.13 This discrepancy may be related to the fact that we performed our study in a population selected for hypertension, and we did not perform association analysis in the 2 subgroups of cases, SA or ACS, to avoid small sample size bias. On the contrary, the −506II genotype was associated with a reduced risk of deep vein thrombosis in male patients.23 In our population, incidence of deep vein thrombosis and pulmonary embolism was extremely low (2 patients in controls and 1 patient in cases), and we could not perform any analysis to replicate this finding. It has been suggested that human thrombotic diathesis shows a focal nature and that the same genetic variant may differentially affect predisposition to deep vein and arterial thrombosis.24 With regard to F2R this hypothesis is supported by the observation that F2R elicits vasorelaxation in arteries and vascular wall cell proliferation and vasoconstriction in veins,25 and therefore the same genetic variation may differentially affect the development of thrombosis in different vascular beds.
One major caveat in observational genetic studies is that positive association findings may represent spurious results because of the presence of population stratification.16 According to NCBI database (www.ncbi.nlm.nih.gov/), the I allele frequency in the European panel is 25%; however, in our genotype data 3 of the 5 SNPs examined, −1738 G/A, 2860 G/A, and 9113 C/A, did not show any difference in allelic frequency between cases and controls speaking against the presence of stratification in this population.
To test the biological relevance of sequence variation at the −506 locus we have explored the possibility that duplication of the Ets-1 binding site might change the functionality of this locus. In our experimental conditions, presence of the I sequence was associated with reduced nuclear binding affinity and blunted response to prolonged hypoxia as compared with the D sequence, suggesting that duplication of the Ets-1 consensus sequence might be responsible for a reduced Ets-1 binding and a blunted response to hypoxia. An abnormal response of gene machinery to a hypoxic stimulus might be detrimental for endothelial and vascular wall metabolism and can therefore be a factor promoting coronary artery lesion development and growth. Endothelial responses to the same stimulus, however, are different in each vascular bed and depend on the local microenvironment and further studies in atherosclerotic as well as healthy coronary arteries are needed to conclusively demonstrate this biological phenomenon.
Several limitations of the present study need to be acknowledged. The study design is based on a case-control retrospective approach and is limited to hypertensive patients. To reach more general conclusions with regard to the role of F2R on the occurrence of CHD, a prospective longitudinal study as well as the replication in a population not selected for hypertension are warranted. Genetic variation at a specific locus will vary among different populations26,27 and will change as more extensive knowledge of human genetic variability is gained. In the current (January 2007) HapMap release 3 more tSNPs for this locus, 2 mapping in the intron and 1 in the mRNA untranslated region, have been reported, therefore 63% of the total genetic variation at this locus has been explored in the present study and the haplotypes identified may well be implemented and extended in future studies. Finally, although our results suggest that changes in Ets-1 binding affinity are involved in the selective response of −506D and −506I to hypoxia, new experiments are needed to investigate whether the conformational change within the promoter sequence actually reduces Ets-1 binding or enhances binding of other transcription factors with opposite effects.
In conclusion, our results demonstrate that F2R sequence variation influences the occurrences of CHD. The −506I allele, which throughout the study exhibited the strongest effect, has peculiar biological properties, suggesting that it may play a role in the onset and progression of coronary artery lesions modulating F2R activity.
Sources of Funding
This work was supported by two consecutive Marie Curie Fellowships (European Commission Contract Number QLGA-CT-2000-60005 to B.G.) within the EUROGENDIS program, a grant from the Italian Ministry of Health and from Italian Society of Hypertension Year 2004 (to B.T.), the Swedish Research Council (09533), and the Swedish Heart and Lung Foundation (to U.d.F.).
Original received July 12, 2006; final version accepted February 26, 2007.
Fischer M, Broeckel U, Holmer S, Baessler A, Hengstenberg C, Mayer B, Erdmann J, Klein G, Riegger G, Jacob HJ, Schunkert H. Distinct heritable patterns of angiographic coronary artery disease in families with myocardial infarction. Circulation. 2005; 111: 855–862.
American Heart Association Heart Disease and Stroke Statistics—2005 Update. In. Dallas, Tex.; 2005.
Minami T, Sugiyama A, Wu SQ, Abid R, Kodama T, Aird WC. Thrombin and phenotypic modulation of the endothelium. Arterioscler Thromb Vasc Biol. 2004; 24: 41–53.
Kataoka H, Hamilton JR, McKemy DD, Camerer E, Zheng YW, Cheng A, Griffin C, Coughlin SR. Protease-activated receptors 1 and 4 mediate thrombin signaling in endothelial cells. Blood. 2003; 102: 3224–3231.
Kaplanski G, Fabrigoule M, Boulay V, Dinarello CA, Bongrand P, Kaplanski S, Farnarier C. Thrombin induces endothelial type II activation in vitro: IL-1 and TNF-alpha-independent IL-8 secretion and E-selectin expression. J Immunol. 1997; 158: 5435–5441.
Leger AJ, Jacques SL, Badar J, Kaneider NC, Derian CK, Andrade-Gordon P, Covic L, Kuliopulos A. Blocking the protease-activated receptor 1–4 heterodimer in platelet-mediated thrombosis. Circulation. 2006; 113: 1244–1254.
Andrade-Gordon P, Derian CK, Maryanoff BE, Zhang HC, Addo MF, Cheung W, Damiano BP, D’Andrea MR, Darrow AL, de Garavilla L, Eckardt AJ, Giardino EC, Haertlein BJ, McComsey DF. Administration of a potent antagonist of protease-activated receptor-1 (PAR-1) attenuates vascular restenosis following balloon angioplasty in rats. J Pharmacol Exp Ther. 2001; 298: 34–42.
Watanabe D, Takagi H, Suzuma K, Suzuma I, Oh H, Ohashi H, Kemmochi S, Uemura A, Ojima T, Suganami E, Miyamoto N, Sato Y, Honda H. Transcription factor Ets-1 mediates ischemia-and Vascular Endothelial Growth Factor-dependent retinal neovascularization. Am J Pathol. 2004; 164: 1827–1835.
Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL Jr, Jones DW, Materson BJ, Oparil S, Wright JT Jr, Roccella EJ. Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension. 2003; 42: 1206–1252.
Burdon KP, Bento JL, Langefeld CD, Campbell JK, Carr JJ, Wagenknecht LM, Herrington DM, Freedman BI, Rich SS, Bowden DW. Association of protein tyrosine phosphatase-N1 polymorphisms with coronary calcified plaque in the Diabetes Heart Study. Diabetes. 2006; 55: 651–658.
Tirschwell DL, Smith NL, Heckbert SR, Lemaitre RN, Longstreth WT Jr, Psaty BM. Association of cholesterol with stroke risk varies in stroke subtypes and patient subgroups. Neurology. 2004; 63: 1868–1875.
Arnaud E, Nicaud V, Poirier O, Rendu F, Alhenc-Gelas M, Fiessinger JN, Emmerich J, Aiach M. Protective effect of a thrombin receptor (protease-activated receptor 1) gene polymorphism toward venous thromboembolism. Arterioscler Thromb Vasc Biol. 2000; 20: 585–592.
Yang Z, Ruschitzka F, Rabelink TJ, Noll G, Julmy F, Joch H, Gafner V, Aleksic I, Althaus U, Luscher TF. Different effects of thrombin receptor activation on endothelium and smooth muscle cells of human coronary bypass vessels. Implications for venous bypass graft failure. Circulation. 1997; 95: 1870–1876.
Mueller JC, Lohmussaar E, Magi R, Remm M, Bettecken T, Lichtner P, Biskup S, Illig T, Pfeufer A, Luedemann J, Schreiber S, Pramstaller P, Pichler I, Romeo G, Gaddi A, Testa A, Wichmann HE, Metspalu A, Meitinger T. Linkage disequilibrium patterns and tagSNP transferability among European populations. Am J Hum Genet. 2005; 76: 387–398.