Study objectives: Marked variability exists in coronary artery collaterals in patients with ischemic heart disease. Although multiple factors are thought to play a role in collateral development, the contribution of genetic factors is largely unknown. Hypoxia inducible factor 1 (HIF-1), a transcriptional activator that functions as a master regulator of oxygen homeostasis, is one possible genetic factor that could play an important role in modulating collateral development. Design, setting, and participants: Collateral vessels were determined in 100 patients with [greater than or equal to] 70% narrowing of at least one coronary artery without acute myocardial infarction or prior revascularization. DNA was genotyped for the presence of a single nucleotide (C to T) polymorphism that changes residue 582 of HIF-1[alpha] from proline to serine.
Measurements and results: The frequency of the T allele was significantly higher among patients without collaterals compared to patients with collaterals (0.188 vs 0.037, p < 0.001). In multivariate analyses, two variables affecting collateral formation were detected: two- or three-vessel coronary artery disease was a significant positive predictor (odds ratio [OR], 4.17; 95% confidence interval [CI], 1.61 to 10.8; p = 0.001), whereas the presence of HIF-1[alpha] genotype CT or TT was a negative predictor (OR, 0.19; 95% CI, 0.04 to 0.84; p = 0.03).
Conclusions: These data suggest that variations in HIF-1[alpha] genotype may influence development of coronary artery collaterals in patients with significant coronary artery disease.
Key words: collaterals; genetics; genotype; ischemia
Abbreviations: CI = confidence intereval; HIF-1 = hypoxia-inducible factor 1; OR = odds ratio
Coronary collateral vessels can increase blood flow to regions of the heart supplied by arteries with high-grade stenosis, thus protecting the myocardium from ischemia. (1,2) Collateral formation is highly variable between patients and only partially attributable to differences in the degree of the coronary artery occlusive disease. (3) Several factors associated with decreased collateral formation have been described in animals and humans, including aging, (4,5) hypercholesterolemia, (6-9) hypertension, (10,11) and cigarette smoking. (12) A specific genetic marker for differences in collateral formation remains to be demonstrated. (13)
Hypoxia-inducible factor 1 (HIF-1) is a transcriptional activator that functions as a master regulator of oxygen homeostasis. HIF-1 target genes encode proteins that increase oxygen delivery, such as anglogenic factors, as well as proteins that mediate adaptive responses to oxygen deprivation in ischemic tissue, such as glucose transporters and glycolytic enzymes. (14) HIF-1 consists of a constitutively expressed HIF-1[beta] subunit and an oxygen-regulated HIF-1[alpha] subunit. Two nucleotide sequence variants in exon 12 of the human HIF1A gene that affect the coding sequence of HIF-1[alpha] were recently reported in a study of renal cell carcinoma. (15) Each of these variants alters the amino acid sequence within the carboxyl-terminal domain of HIF-1[alpha] that regulates protein stability and transcriptional activity, (14-17) suggesting that they may have functional consequences. In this study, we investigated whether genetic variation at the locus encoding HIF-1[alpha] influences collateral formation in patients with coronary artery disease.
MATERIALS AND METHODS
The study was approved by the institutional review board of Johns Hopkins University. Written informed consent was obtained from all subjects. Study participants were recruited consecutively from those patients undergoing diagnostic coronary artery catheterization. The indications for catheterization in all patients were the presence of stable or unstable angina pectoris or suspected significant myocardial ischemia. Exclusion criteria were age < 18 years, presence of anemia, acute myocardial infarction, or prior revascularization by percutaneous coronary intervention or coronary artery bypass surgery.
Clinical Data Collection
For each patient a detailed questionnaire was recorded. Angiographic and genotype analyses were performed in an independent, blinded manner by different members of the research team. Angiograms were analyzed by two independent reviewers. Only patients with visual assessment of [greater than or equal to] 70% diameter stenosis of at least one coronary artery of [greater than or equal to] 2.5 mm diameter were included in the study. The Rentrop scoring system (3) was used to grade the extent of collateral circulation: 0 = no visible filling of any collateral channels, 1 = collateral filling of branches of the vessel without dye reaching the epicardial segment of that vessel, 2 = partial collateral filling of the epicardial segment of the vessel, and 3 = complete collateral filling of the vessel. Patients with a collateral score of 0 (n = 32) were compared to patients with a collateral score of [greater than or equal to] 1 (n = 68). Agreement between the two readers was 97%. A third reviewer blinded to the readings of the first two reviewers served as arbitrator of differences.
Blood samples (5 mL) were drawn from the femoral artery sheath placed for the coronary artery catheterization, collected in ethylenediamine tetra-acetic acid tubes, and stored at 4[degrees]C. Genomic DNA was isolated from buffy coats using an automated workstation (Autopure LS; Gentra Systems; Minneapolis, MN). Exon 12 of the HIF1A gene was amplified by polymerase chain reaction using forward and reverse oligonucleotide primers in flanking introns (5'-CAGAAGCAAAGAACCCAT-3' and 5'-TCAAGAATTTGCGTTAG-3', respectively). Nucleotide sequence was determined (3730xl DNA Analyzer; Applied Biosystems; Foster City, CA) using the same primers, and data analysis was performed using software (Sequencer; GeneCodes; Ann Arbor, MI).
Two exon 12 polymorphisms have been reported: a C/T polymorphism at nucleotide 85 of exon 12 results in a Pro/Ser polymorphism at residue 582 of HIF-1[alpha], whereas a G/A polymorphism at nucleotide 103 results in an Ala/Thr polymorphism at residue 588. (15) We isolated DNA from peripheral blood cells of all patients and determined the nucleotide sequence of exon 12.
Data are reported as mean [+ or -] SD. Analysis between groups for statistically significant differences in categorical data was performed with the use of the [chi square] or Fisher Exact Test. Analysis between groups for continuous variables was performed by the t test. Multivariate analysis was performed using variables that in univariate analysis were found statistically significant, as well as conditional logistic regression analysis adjusting for variables that may influence the amount of collateral formation (Stata 6.0; StataCorp; College Station, TX). All p values are based on a two-tailed comparison; p < 0.05 was considered significant.
Among 100 patients studied, 68 had collaterals (Table 1). Among patients with collaterals, there was a significantly increased frequency of multivessel coronary artery disease (p = 0.002), hypercholesterolemia (p = 0.03), and statin treatment (p = 0.04) as compared to patients without collaterals.
All patients were homozygous for the G allele encoding Ala at codon 588 (Ala/Thr polymorphism). For the polymorphism involving codon 582 (Pro/Ser polymorphism), the overall frequencies of the C and T alleles were 0.915 and 0.085, respectively (Table 2). These allele frequencies are similar to those reported in studies (15,16) of individuals without cardiac disease. The study population is in Hardy-Weinberg equilibrium. The frequency of the T allele was significantly increased among patients without collaterals as compared to patients with collaterals (0.188 vs 0.037, p < 0.001). Among patients without collaterals, the frequency of the TT genotype was 0.125, whereas no patient with collaterals had the TT genotype (Table 2).
Due to the small number of patients with the CT or TT genotype, both genotypes were analyzed together and compared to CC. Patients with the CC genotype had a significantly higher incidence of multivessel disease as compared with patients with the TT or CT genotype (p = 0.04; Table 3). Collaterals were present at a significantly higher frequency among patients with the CC as compared to the TT or CT genotype: 63 of 87 patients (72%) vs 5 of 13 patients (38%); odds ratio (OR), 4.2; 95% confidence interval (CI), 1.25 to 14.1 (p = 0.02). Collaterals were present in 5 of 13 of the patients with the CT or TT genotype (p = not significant as compared to those with collaterals).
Multivariate analysis, using variables that were significant in the univariate analysis (Table 1), identified two significant variables that predicted the presence or absence of coronary collateral formation. The presence of two- or three-vessel coronary artery disease predicted the presence of collaterals (OR, 4.17; 95% CI, 1.61 to 10.8; p = 0.001). In contrast, the presence of HIF1A genotype CT or TT predicted the absence of collateral formation (OR, 0.25; 95% CI, 0.07 to 0.94; p = 0.04).
Conditional logistic regression analysis was performed adjusting for variables that influence collateral formation (age, gender, smoking, hypertension, hypercholesterolemia, diabetes, [beta]-blockers, angiotensin-converting enzyme inhibitors, and number of diseased vessels). After adjusting for all of these variables, the presence of collaterals was still significantly reduced among patients with the CT or TT as compared to the CC genotype (OR, 0.19; 95% CI, 0.04 to 0.84; p = 0.03). A similar analysis revealed that the presence of multivessel disease predicted collateral formation (OR, 6.45; 95% CI, 2.16 to 19.2; p = 0.001).
A novel single nucleotide (C/T) polymorphism was also identified at nucleotide 99 of intron 12 (intron 12 [+ 99]). The overall frequencies of the C and T alleles were 0.93 and 0.07, respectively (Table 2). In contrast to the exon 12 polymorphism, there were no significant differences in the allele or genotype frequencies for the intron 12 (+ 99) polymorphism in patients with or without collaterals.
In this study, we have presented evidence that the CT or TT genotype for the HIF1A exon 12 polymorphism was associated with absence of coronary collaterals in the setting of coronary artery disease. The CT or TT genotype was also associated with single-vessel as compared to multivessel disease and was observed in 0 of 30 patients with triple-vessel disease. These data provide the first evidence that genetic variation at the HIF1A locus encoding HIF-1[alpha] influences the pathogenesis of ischemic heart disease in humans. The C-to-T polymorphism results in the P582S substitution of serine for proline at residue 582 of HIF-1[alpha]. A search of the HomoloGene database (National Center for Biotechnology Information) revealed that proline is found at this location in HIF-1[alpha] in all mammalian (Homo sapiens, Bos taurus, Mus musculus, Rattus norvegicus) and several nonmammalian (Callus gallus, Oncorhyncus mykiss) vertebrate species. We hypothesized that the P582S substitution may result in reduced HIF-1 activity, which could influence the expression of angiogenic growth factors, leading to reduced collateral formation. Alternatively, the P582S substitution may lead to reduced expression of glucose transporters and glyeolytic enzymes, resulting in reduced metabolic adaptation to ischemia and earlier clinical presentation prior to extensive collateral formation.
Arguing against these models, however, are data indicating that the P582S substitution increases HIF-1 activity in tissue culture cells and is associated with increased tumor microvessel density in patients with head and neck cancer. (16) There are at least three possible explanations for these results. First, HIF-1[alpha] may interact with different proteins in the heart as compared to head and neck cancers, such that the functional effect of the P582S substitution is opposite in the two tissues. Second, experiments (17) indicate that in response to hypoxia, HIF-1 can either activate or repress the expression of genes encoding the angiogenic growth factors angiopoietin 1 and 2, depending on the cell type. Any functional consequence of the P582S substitution would have opposite effects on the level of angiopoietin gene expression, depending on whether HIF-1 was activating or repressing expression. Third, the P582S substitution may be in linkage disequilibrium with other polymorphisms in the HIF1A gene that differ in the US cardiac and Japanese cancer patient populations that were studied. HIF-1[alpha] CC genotype was not found as a predictor for collateral formation. Thus, additional studies to investigate the effect of the Pro582Ser polymorphism on HIF-1 activity in the heart and on the progression of ischemic heart disease are warranted.
The level and extent of myocardial ischemia is a major determinant of collateral vessel growth. There was no difference in clinical presentation, and the percentage of patients with or without collaterals and positive stress test was similar. The exact duration of symptoms in every patient, echocardiographic data, and detailed information on the stress tests are not complete. In a genetic study, an allele association study preferably should show an association not with a single genotype, but with a haplotype, composed of different genotypes of different polymorphisms of the gene. In the study, only one out of the three polymorphisms showed a positive association.
ACKNOWLEDGMENT: We are grateful to the nurses and staff of The Johns Hopkins Hospital Cardiac Catheterization Laboratory for their help. We thank Dr. Aravinda Chakravarti for advice on study design and data analysis.
(1) Koerselman J, van der Graaf Y, de Jaegere PP, et al. Coronary collaterals: an important and underexposed aspect of coronary artery disease. Circulation 2003; 107:2507-2509
(2) Hansen J. Coronary collateral circulation: clinical significance and influence on survival in patients with coronary artery occlusion. Am Heart J 1989; 117:290-295
(3) Rentrop KP, Cohen M, Blanke H, et al. Changes in collateral channel filling immediately after controlled coronary artery occlusion by an angioplasty balloon in human subjects. J Am Coll Cardiol 1985; 5:587-592
(4) Rivard A, Fabre JE, Silver M, et al. Age-dependent impairment of angiogenesis. Circulation 1999; 99:111-120
(5) Arthur WT, Vernon RB, Sage EH, et al. Growth factors reverse the impaired sprouting of microvessels from aged mice. Microvasc Res 1998; 55:260-270
(6) VanBelle E, Rivard A, Chen D, et al. Hypercholesterolemia attenuates angiogenesis but does not preclude augmentation by angiogenic cytokines. Circulation 1997; 96:2667-2674
(7) Chen CH, Henry PD. Atherosclerosis as a microvascular disease: impaired angiogenesis mediated by suppressed basic fibroblast growth factor expression. Proc Natl Acad Set U S A 1997; 109:351-361
(8) Bucay M, Nguy J, Barrios R, et al. Impaired adaptive vascular growth in hypercholesterolemic rabbit. Atherosclerosis 1998; 139:243-251
(9) Chen CH, Cartwright J, Li Z, et al. Inhibitory effects of hypercholesterolemia and ox-LDL on angiogenesis like endothelial growth in rabbit aortic explants: essential role of basic fibroblast growth factor. Arteriosel Thromb Vase Biol 1997; 17:1303-1312
(10) Kyrialddes ZS, Kresmastinos DT, Michelakakis NA, et al. Coronary collateral circulation in coronary artery disease and systemic hypertension. Am J Cardiol 1991; 67:687-690
(11) Karpanou EA, Vyssoulis GP, Skoumas JN, et al. Significance of arterial hypertension on coronary collateral circulation development and left ventricular function in coronary artery disease. J Hypertens Suppl 1988; 6:S151-153
(12) Heinle RA, Levy RI, Gorlin R. Effects of factors predisposing to atherosclerosis on formation of coronary collateral vessels. Am J Cardiol 1974; 33:12-16
(13) Hoehberg I, Roguin A, Nikolsky E, et al. Haptoglobin phenotype and coronary artery collaterals in diabetic patients. Atherosclerosis 2002; 161:441-446
(14) Semenza GL. Surviving ischemia: adaptive responses mediated by hypoxia-inducible factor 1. J Clin Invest 2000; 106:809-812
(15) Clifford SC, Astuti D, Hooper L, et al. The pVHL-associated SCF ubiquitin ligase complex: molecular genetic analysis of elongin Band C, Rbx1, and HIF-1[alpha] in renal cell carcinoma. Oncogene 2001; 20:5067-5074
(16) Tanimoto K, Yoshiga K, Eguchi H, et al. Hypoxia-inducible factor-1[alpha] polymorphisms associated with enhanced transactivation capacity, implying clinical significance. Carcinogenesis 2003; 24:1779-1783
(17) Kelly BD, Hacker SF, Hirota K, et al. Cell type-specific regulation of angiogenic growth factor gene expression and induction of angiogenesis in nonisehemic tissue by a constitutively active form of hypoxia-inducible factor 1. Circ Res 2003; 93:1074-1081
* From the Division of Cardiology (Drs. Resar, Roguin, Nasir, Hennebry, and Miller, and Mr. Voner), Departments of Medicine and Pediatrics, McKusick-Nathans Institute of Genetic Medicine (Ms. Ingersoll and Ms Kasch); and Institute for Cell Engineering (Dr. Semenza), The Johns Hopkins University School of Medicine, Baltimore, MD.
This work was supported in part by grants R01-HL55338 and P01-HL65608 from the National Institutes of Health (Dr. Semenza).
Manuscript received December 2, 2004; revision accepted February 22, 2005.
Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml).
Correspondence to: Jon R. Resar, MD, Johns Hopkins Hospital, Blalock 524, 600 North Wolfe St, Baltimore, MD 21287; e-mail: email@example.com
COPYRIGHT 2005 American College of Chest Physicians
COPYRIGHT 2005 Gale Group