Angiotensin-Converting Enzyme Gene Polymorphism Is Associated With Myocardial Infarction but Not With Development of Coronary Stenosis
Background Although both genetic and nongenetic factors contribute to the pathogenesis of coronary artery disease, the identification of specific genetic lesions has lagged behind the identification of critical environmental risk factors. A reported association between myocardial infarction (MI) and the insertion/deletion (I/D) polymorphism of the angiotensin-converting enzyme (ACE) gene in European men suggests a critical role for this genomic region. However, the generality of this association remains to be determined. It also is not clear at what stage in disease progression the association with the ACE I/D polymorphism becomes important.
Methods and Results We evaluated the ACE I/D polymorphism in patients who had undergone coronary angiography (402 men and 295 women) and in 203 representative control subjects. After polymerase chain reaction amplification, genotypes were determined by agarose gel sizing and by hybridization with allele-specific oligonucleotides. After patients were categorized by the degree of coronary artery stenosis and the occurrence of an MI, the distribution of ACE I/D genotypes was evaluated by log linear analysis. Patients were genetically representative of the regional population, and patients with >60% stenosis of their coronary arteries had the same distribution of ACE I/D genotypes as did patients with <10% stenosis. However, among patients with stenosis, the occurrence of an MI was significantly associated with the D allele in all patients (odds ratio [OR], 1.59; P=.002) and in men alone (OR, 1.63; P=.006). The lack of significance in women (OR, 1.40; P=.263) is probably due to the fact that only 36 women in the present study had experienced an MI. Furthermore, the association between MI and the ACE I/D polymorphism was independent of blood pressure, smoking habits, and body mass index.
Conclusions Segregation of the ACE I/D polymorphism is a pervasive genetic risk factor for MI in whites but has no evident effect on the events leading to stenosis of the coronary arteries. This suggests that risk of MI is influenced by two independent processes—atherogenesis that leads to coronary stenosis followed by conversion to MI. The renin-angiotensin system appears to confer significant risk of infarction by influencing the conversion to MI but has no apparent effect on the development of atherostenosis.
Heart attacks, which represent a leading cause of death in most developed countries, tend to cluster in families. However, other than the genetically mediated lipid abnormalities that contribute to stenosis of coronary arteries, little is known about the genetic underpinnings of the events leading to myocardial infarction (MI). Genes that influence the renin-angiotensin system are potential etiological candidates for MI because their products exert a profound systemic effect on vasoconstriction and because angiotensin-converting enzyme (ACE) inhibitors reduce the risk of MI.1 2 This supposition is strengthened by reports that the ACE insertion/deletion (I/D) polymorphism is associated with increased risk for MI in a group of retrospectively studied European men3 4 as well as with coronary disease in French diabetics,5 Japanese coronary patients,6 and a Welsh community.7 However, a prospective case-control study of US male physicians failed to demonstrate that this polymorphism is associated with a significantly increased risk of clinical manifestations of coronary heart disease.8 To resolve this issue and to determine at what stage in disease progression the ACE I/D polymorphism might become clinically relevant, we studied a large series of angiographically assessed white patients.
Our study population9 consisted of 402 men and 295 women (<65 and <70 years old, respectively) who presented for angiography either because of symptoms relating to coronary artery disease (CAD) or because of unrelated conditions, such as valvular disease. We also evaluated 203 population control subjects located through a blood bank. Patients and population control subjects were residents of Utah, southwestern Idaho, and southeastern Wyoming, a population that is genetically representative of US whites.10 Blinded assessment of coronary angiograms by the CASS protocol9 11 identified CAD (defined by >60% stenosis in any major coronary vessel) in 264 men and 98 women, whereas 138 men and 197 women were disease free (<10% stenosis). Quantitative angiography of a small series of angiograms (n=27) confirmed the validity of the estimated degree of coronary occlusion. Patients were matched to subjects by age to within 5 years, so that, on average, patients with CAD were no more than 3.2 years older than disease-free individuals.9 Of the patients with CAD, 159 had experienced an MI that was clinically verified by ECG (123 men and 36 women). All subjects enrolled in the study gave informed consent, and all study protocols were approved by the University of Utah Medical Center Institutional Review Board.
Detection of the ACE Alu I/D Allele
The presence (allele I) or absence (allele D) of the 287-bp Alu repeat in intron 16 of the ACE gene was determined by evaluating the size of DNA fragments after polymerase chain reaction (PCR) amplification, using the primers and PCR conditions described by Rigat et al.12 Genotypes were scored by visualizing the fragments under UV light after running the reaction products on a 2% FMC Seakem agarose gel in Tris borate EDTA buffer containing ethidium bromide for 2 hours at 100 V. Genotypes were confirmed by allele-specific oligonucleotide (ASO) hybridization. After amplification, samples were denatured and blotted onto nylon membranes, cross-linked with UV light, neutralized, and hybridized overnight at 40°C with 32P-labeled probe D (5′-CACATAAAAGTGACTGTATAGGCAG-3′) for the deletion allele and probe I (5′-AAAAAAAAAAAGTGACTGTA-3′) for the insertion allele. The membranes were washed at 45°C or 65°C with 6× SSC/0.1% SDS for 30 minutes for probe I or probe D, respectively. Autoradiographs of the dot blots were then scored for genotypes.
Allele and genotype frequencies were determined from observed genotypic counts, and departure from Hardy-Weinberg expectations were evaluated by χ2 analysis. Genetic associations were evaluated by likelihood ratio test (LRT) statistics calculated for log-linear analysis of cross-classified data using the glim13 package. Multiway log-linear analyses were performed on the data for patients with CAD with the significance of two- and three-way associations determined by comparing LRT statistics generated by successive hierarchical models. The relation between genotype and concomitant variables (eg, lipids) was evaluated by ANOVA and linear regression, also using glim. Statistical power for detecting odds ratios was calculated according to Schlesselmann.14
The frequency of the D allele in the 697 angiographically assessed patients (0.54) was virtually identical with the frequency observed in the sample of 203 representative controls (0.55), as was the frequency of each of the three genotypes (II, 0.23 versus 0.21; ID, 0.47 versus 0.48; DD, 0.30 versus 0.31). Furthermore, these allele frequencies were the same as those previously reported for a large European control population3 4 as well as a small sample of control subjects from Colorado.15 Therefore, with respect to the ACE locus, our sample of angiographically assessed patients is genetically representative not only of the regional population but also of white populations in general.
Association Between the ACE I/D Polymorphism and CAD
When the distribution of the ACE I/D polymorphism was evaluated within the angiographically assessed patients, we observed no significant difference in the frequencies of alleles or of genotypes between the 362 patients with CAD (>60% stenosis) and the 335 patients without CAD (<10% stenosis) (Table 1⇓). This lack of association persisted when the comparison was restricted to men or women. Overall, there is no evidence that the ACE I/D polymorphism is associated with the development or progression of arterial stenosis in these 697 angiographically assessed patients.
Association Between the ACE I/D Polymorphism and MI
In contrast, when the analysis was restricted to patients with significant stenosis, patients who had had an MI had a significantly different distribution of genotypic frequencies (P=.005) and of allelic frequencies (P=.002) compared with patients with CAD who had not had an MI (Table 1⇑). When stratified by sex, both genotypic associations and allelic associations remained statistically significant in men (P=.018 and P=.006, respectively). Although statistical significance was not obtained for women alone, this was probably due to the small number of women with an MI (n=36), as the odds ratio for women (1.40) approached that for men (1.63).
In this sample, patients with CAD who had had an MI had more extensive stenosis than those who had not: 86.2% of patients with an MI had >90% occlusion compared with only 65.8% of patients without an MI. However, when the patients with CAD were stratified into three groups by degree of coronary occlusion (60% to 75% stenosis, 75% to 90% stenosis, and >90% stenosis), the association between the D allele and MI occurred in each group, with no variability by degree of coronary occlusion. Because smoking and high blood pressure, both risk factors for MI, might be expected to interact with the renin-angiotensin system, we repeated the analyses within strata defined by blood pressure and smoking status. Neither risk factor modified the association between the ACE I/D polymorphism and MI, nor did the distribution of blood pressure vary by ACE genotype (data not shown).
Because longevity has recently been associated with an elevated frequency of the D allele,16 we also investigated whether age influences the association with MI. An age-stratified analysis identified a consistent elevation of the D allele in patients with MI within age tertiles, with no evidence of heterogeneity of risk by age (men: χ2(2)=1.98, P=.39; women: χ2(2)=1.51, P=.47). Thus, the association between the D allele and MI appears invariant for those between the ages of 50 and 70 years (the age range of our sample). Furthermore, in men the frequency of the D allele actually declined with age in both patients and control subjects, consistent with a higher risk of mortality due to MI. Therefore, it is unlikely that the D allele is associated with longevity in the Utah population.
However, despite this unambiguous association between the ACE D allele and risk of MI, we were unable to confirm the initial report3 of a stronger association in lean individuals with low apolipoprotein B (apoB) levels. Even though our sample size was adequate to detect the reported odds ratio of 3.2 for the “low-risk” group,3 we failed to find an increased association, either in patients with stenosis (Table 1⇑) or in patients without CAD (data not shown). Accordingly, we investigated the interaction with plasma lipid levels. Regression of lipid levels on ACE genotype showed no trend with the D allele in control subjects or in patients with CAD without an MI. However, in the patients with CAD with an MI, the D allele was associated with a significant increase in total cholesterol (II, 187.3; ID, 202.9; DD, 217.3; P<.008) and LDL cholesterol (II, 119.1; ID, 134.0; DD, 145.9; P<.007) and a nonsignificant increase in apoB levels (II, 80.6; ID, 88.9; DD, 91.9, P>.10). These data suggest an interaction between the ACE genotype and lipid levels in patients most at risk for MI.
Collectively, these results suggest that in the Utah population and, by extension, in the US white population, the D allele confers a consistent risk of MI, and this risk is not influenced by age, degree of angiographically defined coronary stenosis, hypertension, or smoking habits.
Associations Among ACE Genotype, Body Mass Index, and MI
Because we failed to confirm an increased association between the ACE I/D polymorphism and MI in “low-risk” patients, we carried out a three-way log-linear analysis to determine whether the association with MI was independent of body mass index (BMI). As indicated in Table 2⇓, when patients with CAD were cross-classified with respect to ACE genotype, BMI, and MI, the D allele was found to be associated with low BMI when both sexes were combined (odds ratio, 1.5; P=.023) and in men alone (odds ratio, 1.7; P=.01). In neither case was there a significant association between MI and BMI (P=.29 and P=.45, respectively). Furthermore, the model incorporating only the independent two-way associations (ACE · MI and ACE · BMI) gave an excellent fit to the data for men (χ2(3)=0.36, P=.95) and an adequate fit to the total data set (χ2(3)=4.18, P=.24), indicating that the association between the ACE I/D polymorphism and MI is independent of BMI (Table 2⇓). When the analysis was constrained to women, none of the three-way log-linear models provided a reasonable fit, in part because of the unusual genotype frequency distribution in women with CAD, characterized by low BMI and absence of MI—a group in whom the genotype frequencies depart significantly from Hardy-Weinberg equilibrium.
Overall, this analysis provides no support for the contention that lean individuals possessing the D allele have a significantly greater risk of MI compared with more obese individuals. In particular, the lack of association between MI and BMI indicates that in men, the excess risk of MI associated with the D allele (odds ratio, 1.6; Table 1⇑) is equal for the low- and high-BMI groups.
By evaluating the distribution of the ACE I/D polymorphism in a large series of angiographically defined patients, the present study has shown that segregation of the D allele is not associated with the events leading to stenosis of the coronary arteries. Our failure to demonstrate an association with angiographically defined CAD in patients who had not experienced an MI conflicts with reports of an association with coronary heart disease.5 6 7 However, in the French diabetes study, only 46 patients had not experienced an MI,5 so it is unclear whether the elevated frequency of the D allele in these patients was due to small sample size or to a different disease pathogenesis in diabetics. In the other two studies, coronary angiography was not performed, making comparison difficult. Although the result from the Japanese study can be attributed to the fact that 61% of their coronary patients had experienced an MI,6 it is unclear whether a similar enrichment of patients with MI could account for the findings from the Welsh study. On balance, failing other angiographically defined studies, the consistently lower frequency of the D allele in patients from Utah with CAD who have not experienced an MI appears conclusive.
However, once substantial atherosclerosis (>60%) has occurred, the presence of the D allele is then significantly associated with an increased risk of MI. This increased risk of MI is independent of age and degree of coronary stenosis, as well as such risk factors as high blood pressure, smoking, and BMI. Because the ACE D allele is associated with elevated ACE activity in whites,17 it is tempting to speculate that the ACE I/D polymorphism identifies a genetic variant that contributes to the risk of MI through increased vasopression, vascular cell hypertrophy, and thrombosis after an increase in angiotensin II (Ang II) levels and the inactivation of bradykinin.18 Although the I/D of the Alu element in intron 16 of the ACE gene may directly influence ACE activity, the ACE D allele could also be in linkage disequilibrium with a genetic variant of ACE that is associated with greater vasopressor activity. This, in turn, may initiate a series of events leading to infarction.
These results not only implicate the ACE I/D polymorphism as a pervasive and independent risk factor for the occurrence of MI in white populations but also suggest that MI is due to the interaction of two independent pathogenetic processes. The initiation and development of significant atherosclerosis appear to be independent of the ACE gene and, by implication, of ACE levels. Experimental studies are somewhat contradictory in terms of whether ACE or its product, Ang II, is likely to influence atherogenesis. Although Ang II is known to modulate vascular smooth muscle growth in cell culture,19 20 21 22 23 hyperplasia occurs only in the presence of serum20 ; otherwise, hypertrophy alone is observed.21 These bifunctional effects on growth appear to relate to the induction of transforming growth factor–β1,22 which exerts a key antiproliferative effect modulating the mitogenic properties of induced basic fibroblast growth factor,23 with the net effect being hypertrophy. Although an imbalance in these Ang II–activating signals could promote vascular disease through abnormal vascular smooth muscle growth,23 our data suggest this mechanism is unlikely to occur in vivo.
The results of the present study suggest that elevated ACE activity has a primary impact on the transition of atherosclerosis to MI rather than exerting a differential effect on atherosclerotic development. Once significant stenosis has occurred, the influence of a genetic variant of ACE could lead to greater risk of infarction by a number of mechanisms. From a clinical perspective, this interpretation implies that a therapeutic approach with ACE inhibitors may diminish the risk of MI in patients previously identified with significant stenosis (>60%) of the coronary arteries. Given the potential importance of initiating this therapeutic strategy, additional angiographic data are urgently needed to define a more precise estimate of the relation among ACE I/D genotypes, ACE levels, and risk of MI in both men and women. There also is a need to determine whether the magnitude of this association is comparable in other ethnic groups.
This study was supported by grant HL-38840 from the National Institutes of Health. Dr Lalouel is an investigator of the Howard Hughes Medical Institute. We thank S. Prescott, J. Metherall, and R. Lifton for their critical review of the manuscript.
- Received February 15, 1994.
- Revision received February 13, 1995.
- Accepted February 14, 1995.
- Copyright © 1995 by American Heart Association
Pfeffer MA, Braunwald E, Moyé LA, Basta L, Brown EJ, Cuddy TE, Davis BR, Geltman EM, Goldman S, Flaker GC, Klein M, Lamas GA, Packer M, Rouleau J, Rouleau JL, Rutherford J, Werthheimer JH, Hawkins CM, and the SAVE Investigators. Effect of captopril on mortality and morbidity in patients with left ventricular dysfunction after myocardial infarction. N Engl J Med. 1992;327:669-677.
Cambien P, Poirier O, Lecerf L, Evans A, Cambou J-P, Arveiler D, Luc G, Bard J-M, Bara L, Ricard S, Tiret L, Amouyel P, Alhenc-Gelas F, Soubrier F. Deletion polymorphism in the gene for angiotensin-converting enzyme is a potent risk factor for myocardial infarction. Nature. 1992;359:641-644.
Tiret L, Kee F, Poirier O, Nicaud V, Lecerf L, Evans A, Cambou J-P, Arveiler D, Luc G, Amouyel P. Deletion polymorphism in angiotensin-converting enzyme gene associated with parental history of myocardial infarction. Lancet. 1993;341:1991-1992.
Ruiz J, Blanché H, Cohen N, Velho G, Cambien F, Cohen D, Passa P, Froguel P. Insertion/deletion polymorphism of the angiotensin-converting enzyme gene is strongly associated with coronary heart disease in non-insulin-dependent diabetes mellitus. Proc Natl Acad Sci U S A. 1994;91:3662-3665.
Nakai K, Itoh C, Miura Y, Hotta K, Musha T, Itoh T, Miyakawa T, Iwasaki R, Hiramori K. Deletion polymorphism of the angiotensin I–converting enzyme gene is associated with serum ACE concentration and increased risk for CAD in the Japanese. Circulation. 1994;90:2199-2202.
Mattu RK, Needham EWA, Galton DJ, Frangos E, Clark AJL, Caulfield M. A DNA variant at the angiotensin-converting enzyme gene locus associates with coronary artery disease in the Caerphilly Heart Study. Circulation. 1995;91:270-274.
Kreutz R, Lindpaintner K, Pfeffer MA, Stampfer MJ, Gaziano JM, Grodstein F, Boring J, LaMotte F, Hennekens CH, Nadal-Ginard B, Braunwald E. Angiotensin-converting enzyme genotype and risk for coronary heart disease. Circulation. 1993;66(suppl I):I-510. Abstract.
Marshall HW, Morrison L, Wu LL, Anderson JL, Corneli PS, Stauffer DM, Allen A, Karagounis LA, Ward RH. Apolipoprotein polymorphisms fail to define risk of coronary artery disease: results of a prospective angiographically controlled study. Circulation. 1994;89:567-577.
The Principal Investigators of CASS and their Associates. The National Heart, Lung, and Blood Institute Coronary Artery Surgery Study (CASS). Circulation. 1981;63(suppl I):I-81. Abstract.
Rigat B, Hubert C, Corvol P, Soubrier F. PCR detection of the insertion/deletion polymorphism of the human angiotensin converting enzyme gene (DCP1 dipeptidyl carboxypeptidase 1). Nucl Acids Res. 1992;20:1433.
glim 3.77. London, UK: Numerical Algorithms Group, Inc; 1985.
Schlesselmann JJ. Case Control Studies: Design, Conduct and Analysis. New York, NY: Oxford University Press; 1982.
Gavras H, Gavras IJ. Cardioprotective potential of angiotensin converting enzyme inhibitors. Hypertension. 1991;9:385-392.
Berk BC, Vekshtein V, Gordon GM, Tsuda T. Angiotensin II-stimulated protein synthesis in cultured vascular smooth muscle cells. Hypertension. 1989;13:305-314.
Gesiterfer AAT, Peach MJ, Owens GK. Angiotensin II induces hypertrophy not hyperplasia of cultured rat aortic smooth muscle cells. Circ Res. 1988;62:749-756.
Gibbons GH, Pratt RE, Dzau VJ. Vascular smooth muscle cell hypertrophy vs hyperplasia: autocrine transforming factor-β1 expression determines growth response to angiotensin II. J Clin Invest. 1992;90:456-461.
Itoh H, Mukoyama M, Pratt RE, Gibbons GH, Dzau VJ. Multiple autocrine growth factors modulate vascular smooth muscle cell growth response to angiotensin II. J Clin Invest. 1993;91:2268-2274.