Distribution in Healthy and Coronary Populations of the Methylenetetrahydrofolate Reductase (MTHFR) C677T Mutation
Modest elevations of circulating homocyst(e)ine are common in patients with vascular disease. We explored in normal and coronary artery disease (CAD) populations the distribution of a mutation in the 5,10-methylenetetrahydrofolate reductase (MTHFR) gene that results in enzyme thermolability and reduced activity and in homocyst(e)ine elevation to assess its relevance to risk. We identified the C to T substitution at the MTHFR locus and compared the distributions of genotypes in 565 patients aged ≤65 years without and with angiographically documented CAD and in 225 healthy subjects. In the patients, we also assessed interrelations between genotypes and CAD occurrence and severity, as well as standard risk factors. The frequency of homozygotes for the mutation was the same in patients with and without CAD and in healthy subjects (11.6%, 11.0%, and 10.7%, respectively; P>.5 for each). There was also no excess among the 419 patients with severe disease (ie, one or more vessels with >50% luminal obstruction) compared with those with no or mild CAD (odds ratio: 1.004; 95% confidence interval: 0.59 to 1.70). Homozygosity for the mutation was also not associated with a history of myocardial infarction or the presence or severity of angina. However, body mass index increased linearly with the presence of the mutant allele (P=.005), and the mutation and hypertension were weakly associated (P=.036). We conclude that the MTHFR genotype is not a risk factor for coronary disease in this Australian population but that the strong association found with body mass index should be explored further.
- methylenetetrahydrofolate reductase gene
- thermolabile MTHFR
- coronary artery disease
- body mass index
- Received November 13, 1995.
- Revision received February 22, 1996.
A modest elevation of plasma homocyst(e)ine has been documented in ≈20% of patients with coronary, cerebral, and peripheral vascular disease;1 2 3 4 5 6 7 8 venous thrombosis;9 and chronic renal failure,10 11 among whom cardiovascular risk is also greatly increased. For CAD, the enhanced risk associated with a 5-μmol/L elevation of total plasma homocyst(e)ine was estimated to be the same as that associated with a 0.5-mmol/L increase in total cholesterol in a recent meta-analysis.12
The essential sulfur-containing amino acid methionine is metabolized in several steps to homocysteine.13 On a normal diet, ≈50% is further catabolized to cystathionine, and then cysteine, and eventually excreted as inorganic sulfate. The rate-limiting step is the activity of the enzyme cystathionine β-synthase, the gene for which is located on chromosome 21.13 14 On a normal diet, the remaining 50% of the homocysteine formed is remethylated back to methionine, thus conserving methyl groups and completing a cycle. The principal remethylating enzyme involved is MTHFR, the gene for which is located on chromosome 1.15 Mutations in the genes coding for both of these enzymes lead to a group of disorders in which marked elevation of circulating homocyst(e)ine and homocystinuria are common features.13 16 17 18 19 20 In untreated individuals, these disorders may result in various severe phenotypic features, including precocious vascular disease.13
A significant proportion of subjects with mild homocyst(e)ine elevation have a thermolabile MTHFR that is defective in its enzymatic activity.21 22 The presence of a thermolabile MTHFR was also found in one study to be predictive of coronary artery stenosis independent of other risk factors.23 Recently, Frosst and colleagues19 identified a C to T substitution at nucleotide 677 of the MTHFR gene that converts an alanine to a valine residue. They showed that this mutation was responsible for the thermolability of MTHFR and that homozygotes for the mutation had about 30%, and heterozygotes about 65%, of the MTHFR activity found in individuals without the mutation. Homozygotes for the mutation also had elevated circulating homocyst(e)ine.19 These findings have been confirmed by van der Put and colleagues,24 who showed that the mutation was not only associated with thermolability of the enzyme and increased plasma homocyst(e)ine but also with increased red blood cell folate, although plasma folate levels tended to be in the low normal range.
We undertook the present study to determine whether this point mutation at the MTHFR gene associated with increased plasma homocyst(e)ine could serve as a genetic predictor for CAD risk in addition to other known risk factors. Since factors initiating coronary atherogenesis may not necessarily be the same as those responsible for its progression, we explored possible associations between the MTHFR mutation and both the occurrence and severity of CAD. We compared the distributions of the mutation in patients with and without CAD and in healthy subjects to assess associations with the occurrence of CAD. To investigate relationships with the severity of CAD, we determined the distribution of the mutation among patients with different numbers of significantly diseased (>50% luminal obstruction) major epicardial coronary arteries documented angiographically and among patients with different coronary severity scores.
We studied 225 genetically unrelated subjects (122 females and 103 males) living in the eastern area of Sydney, Australia, of mean age 42±10.8 years (range, 18 to 68 years) recruited from among healthy volunteers participating in a heart health education program.25 None of these control individuals had reported overt CAD in a questionnaire they had completed. They were selected randomly for the current study after appropriate permission only on the basis that all were white, unrelated, and from the same residential area as the patients. Because lipids were also being measured, venous blood samples were collected after an overnight fast. Written consent was obtained from all subjects.
We studied 565 white patients aged ≤65 years, both men (407) and women (158), consecutively referred to the Eastern Heart Clinic at Prince Henry Hospital for coronary angiography over a 16-month period in 1994 and 1995. We excluded only patients shown to have significant left main disease (>50% luminal obstruction), because it was difficult to categorize this small proportion of the total (5%) within the classification system used (see below). A written consent was obtained from every patient after a full explanation of the study, which was approved by the Ethics Committee of the University of New South Wales.
A 4-mL venous blood sample was drawn into an EDTA sample tube before the angiogram after at least a 6-hour fast. The blood sample was centrifuged within 2 hours, and plasma and cellular components were stored separately at −70°C in aliquots until analysis.
Detection of the C to T Substitution at the MTHFR Locus
DNA was extracted from the frozen cellular blood component by a salting-out method adapted from that described by Miller et al26 for whole frozen blood. The extracted DNA was stored at 4°C until analysis. The DNA samples were subjected to amplification by the polymerase chain reaction (see the Figure⇓), and the restriction enzyme HinfI was used to identify those with the mutation, as described by Frosst and colleagues.19 The mutant allele was designated as “+” and the wild-type as “−.”
TC, HDL-C, and triglyceride levels were measured by the hospital's clinical chemistry department by standard enzymatic methods. The LDL-C levels were calculated using the Friedewald formula.
Documentation of CAD Severity and Other Medical Conditions
The severity of CAD was determined by the number of significantly stenosed coronary arteries as follows. The angiograms were assessed by two cardiologists who were unaware that the patients were to be included in the study. Each angiogram was classified as revealing either no coronary lesion with >50% luminal stenosis or as having one, two, or three major epicardial coronary arteries with >50% luminal obstruction. We also used the Green Lane coronary scoring system, which provides a numerical value for lesion severity and takes account of the amount of myocardium supplied by an affected vessel; the maximal score is 15.27
We obtained each patient's medical history by using a questionnaire with standardized choices of answers to be ticked during the interview as described previously.28
Hardy-Weinberg equilibrium was assessed by χ2 analysis as described by Emery.29 The frequencies of the alleles and genotypes among different subgroups were compared by the χ2 test.
We also used logistic linear regression analysis for associations between CAD severity and the C to T mutation at the MTHFR gene. CAD severity was regarded as the dependent variable, and the MTHFR genotype, sex, hypertension, diabetes, age, lifetime smoking dose, and lipoprotein levels were entered as independent variables.
Characteristics of the Patient Population in Relation to MTHFR Genotype
The demographic information in relation to MTHFR genotypes for all 565 angiographically studied patients (407 males and 158 females) is shown in Table 1⇓. The values for different genotypes were compared by ANOVA. There was no significant difference among patients with different genotypes for any of the listed variables except BMI. There was a linear dose-dependent relationship between the presence of the + mutation and BMI. MTHFR +/+ homozygous patients had the highest BMI, +/− heterozygous patients had an intermediate value, and the patients without the mutation had the lowest BMI. This relationship between BMI and MTHFR was not affected by age or sex. The age- and sex-adjusted BMIs for the +/+, +/−, and −/− genotypes were 29.4±0.2, 28.6±0.2, and 27.8±0.1 kg/m2, respectively (F=3.6, df=2, P=.028). Moreover, neither age (P=.132) nor sex (P=.982) was predictive of BMI in the patient population. This association was also significant among those patients (n=456) who had angiographically proven CAD (F=4.07, df=2, P=.018) with BMI adjusted by age and sex. When the analysis was conducted in the 407 male and 158 female patients separately, we found that this linear relation was significant among males (30.0±0.7, 28.1±0.3, and 27.8±0.3 kg/m2 for +/+, +/−, and −/− genotypes, respectively; F=5.41, P=.0048) but not in the smaller number of female patients. However, female patients without the mutation had the lowest BMI (28.9±1.0, 29.1±0.7, and 27.6±0.5 kg/m2 for +/+, +/−, and −/− genotypes, respectively; F=1.54, P=.22).
C to T Mutation at the MTHFR Locus in Healthy Subjects and in Patients With and Without CAD
The frequencies for MTHFR +/+, +/−, and −/− genotypes and allele frequencies in the patients with angiographically defined CAD and those with angiographically defined normal coronary arteries, together with those in the healthy subjects, are shown in Table 2⇓. The distributions of the genotypes were in Hardy-Weinberg equilibrium for both patients (χ2=0.387, df=2, P=.824) and healthy subjects (χ2=1.813, df=2, P=.404), and the distribution of the genotypes between male and female patients was the same. When we compared the patients in the angiographic population with CAD (n=456), those with normal coronary arteries (n=109), and the healthy subjects, we found no differences in the frequency of the mutant + allele. There was also no excess of the +/+ homozygotes among patients with CAD compared with those without CAD or the healthy subjects (11.6% in the CAD patients versus 11.0% in those without and 10.7% in the healthy subjects; P>.05 for each). The results were the same when only the 419 patients with one or more significantly diseased vessels (>50% luminal obstruction) were compared with the other two populations (P>.05); the odds ratio was 1.004 and the 95% confidence interval 0.59 to 1.70 when CAD patients were compared with healthy subjects. The statistical power of the analysis is 99.1% as calculated from the size of the patient and healthy groups and their difference in frequency distribution of the mutation.
This lack of a significant association between the MTHFR mutation and the presence of CAD was further confirmed for the occurrence of myocardial infarction in the patients with CAD established by angiography. Among the +/+ homozygotes, the frequency of patients who had a past history of myocardial infarction (0.136, n=32) was not different (χ2=2.87, df=2, P=.22) from those who did not (0.101, n=33).
C to T Mutation at the MTHFR Gene and Severity of CAD
As shown in Table 3⇓, there was no consistent relationship of either the frequency of +/+ homozygotes or the + allele frequency with the number of significantly diseased vessels. This was also true for the coronary scores (5.16±0.6, 5.19±0.3, and 5.15±0.3 for +/+, +/−, and −/− genotypes, respectively; F=0.008, P=.99).
Since the MTHFR +/+ genotype was only regarded as a possible additional risk factor for CAD, we further analyzed the relationship between the mutation and the severity of CAD by controlling other measured risk factors and confounding factors in a logistic linear regression analysis; these included sex, age, BMI, lifetime smoking dose, past history of hypertension, diabetes, family history of premature CAD, lipid levels [TC, HDL-C, LDL-C, triglyceride, Lp(a)], and usage of lipid-lowering and β-adrenergic drugs. The MTHFR genotype remained unrelated to the number of significantly diseased vessels (P=.61).
As discussed above, we found that the MTHFR mutation was associated with increased BMI. When we used cutoff points of 25 kg/m2 (25th percentile of the patient population) and 28 kg/m2 (50th percentile of the patient population) to subgroup the “low”-risk patients according to their BMIs, the MTHFR was still not associated with the number of significantly diseased vessels (χ2=5.15, df=6, P=.52). The relationship also remained nonsignificant when we conducted analyses in the “low”-risk groups, as defined by Lp(a)<300 mg/L or TC/HDL-C<5.0 or by being a nonsmoker (χ2=8.24, df=6, P=.24).
MTHFR Mutation and Other Medical Conditions
Although the frequency of +/+ homozygotes tended to be higher in patients with diabetes (16.6%, n=10) than in those without diabetes (10.8%, n=54), this difference was not statistically significant (χ2=2.34, df=2, P=.31). The same trend was also found for hypertension (χ2=2.12, df=2, P=.35), in which +/+ homozygotes were more prevalent in patients with hypertension (13.3%, n=34) than in those without (9.8%, n=30). However, in a log-linear analysis that included all categorical variables in the model, there was a significant three-way interaction among sex, hypertension, and MTHFR genotypes (partial χ2=9.797, df=2, P=.0075). We therefore analyzed the relationship between the genotype and hypertension among male and female patients separately. Although we found no significant association between MTHFR genotype and hypertension in female patients (χ2=3.58, df=2, P=.17), the association was significant (χ2=6.62, df=2, P=.036) among male patients (15.8%, n=27 for those with and 7.7%, n=18 for those without hypertension). The + allele frequency was also significantly higher (P<.05) among hypertensive patients (0.386) than nonhypertensive patients (0.320). There was no interaction between MTHFR genotype and diabetes with other variables.
We found no association between family history of premature CAD and MTHFR genotype in this patient population (χ2=0.529, df=2, P=.767). The presence of the MTHFR mutation was also not associated with the severity of angina (χ2=2.21, df=4, P=.69).
In their important study, Frosst and colleagues19 not only identified the MTHFR mutation responsible for thermolability of the enzyme but also established that subjects homozygous for this mutation had elevated fasting and postmethionine plasma homocyst(e)ine concentrations. The association between the mutation and elevated fasting homocyst(e)ine was confirmed by van der Put and colleagues,24 who also showed that the mutation was associated with decreased MTHFR activity. Frosst et al19 did not explore relationships between the mutation and vascular disease but suggested that studies should be undertaken to determine whether the mutation is a genetic risk factor. The present investigation sought to assess in an Australian white population the relevance of this genetic marker to both the occurrence and severity of CAD.
The results were clear cut in what, as far as we are aware, is the first study to determine the relevance of the MTHFR mutation to angiographically documented CAD. We could not identify any relationship between the mutation and either the occurrence or severity of CAD. Nor was there any relationship with a history of myocardial infarction or the presence or severity of angina. Although we feel confident about these results in the population we studied, as Frosst and colleagues19 point out, the findings could be population-specific and could be related to nutritional status. The subjects of our study, consecutively referred white patients aged ≤65 years coming to coronary angiography and somewhat younger healthy subjects, were representative of residents living in the same area of the city of Sydney. While we did not measure either homocysteine or folate levels, nutritional deficiency is likely to be uncommon in these patients and control subjects, considering their age and background. In fact, the coronary patients as a group were overweight. Thus, our study does not preclude the possibility that the mutation could be a genetic marker for vascular disease in subsets of the population with a reduced folate intake for nutritional reasons and who are therefore likely to have a greater homocyst(e)ine increase. For example, genetically determined predisposition could have been a factor contributing to elevated homocysteine and the presence of internal carotid artery lesions documented by Selhub and colleagues,30 as their study was in an elderly population in which folate deficiency was common.
The evidence available so far indicates that the frequency of the mutation differs with the population studied. For example, van der Put and colleagues24 reported a homozygote frequency of 5% in Dutch patients recruited from a general practice in the Hague. On the other hand, Frosst and colleagues19 found a homozygote frequency of 12% in 114 unselected French Canadians, a frequency very similar to the 11.6% in the 456 patients with CAD, the 11.0% in the 109 without CAD, and the 10.7% in the 225 healthy subjects we assessed. The 11.6% is much less than the 17% frequency of the thermolabile MTHFR variant reported in North American CAD patients by Kang and colleagues.23 However, some of those patients could have been heterozygotes, as they were not genotyped. In support of the conclusions we have drawn from our data, the homozygote frequencies in healthy subjects, in patients with documented mild (≤50% luminal stenosis) and no CAD, and in patients with severe disease (>50% stenosis in one, two, or three major vessels) were not different; the statistical power of the study was 99.1% for this absence of difference.
When we initially explored the hypothesis that mild elevations of circulating homocyst(e)ine could predispose to cardiovascular risk, we chose to study patients with documented early-onset coronary disease (aged ≤50 years) and a paucity of established risk factors, and the results were consistent with our hypothesis.1 The patients in our more recent study were also a low-risk group,3 and this may have been a feature of some other published positive studies. The study of American physicians, who are also likely to be at lower cardiovascular risk than the general population, showed prospectively an association between total plasma homocyst(e)ine levels and the subsequent development of myocardial infarction with only modest elevations in the affected subjects.31 For these reasons, we analyzed separately those among our patients considered to be at below-average risk as assessed by age, lipid profile, and absence of smoking and diabetes. But again, the genotype distribution remained in Hardy-Weinberg equilibrium, and the homozygote frequency was not different from that of the normal subjects.
The associations between the mutation and BMI and hypertension were confined to the male patients. The association was strong for BMI (P<.005), and there was also a trend in the smaller number of female patients with the −/− genotype to have lower BMIs. The association for hypertension was relatively weak (P<.05) and should be interpreted with caution. We have no explanation for the strong association with BMI and recognize that it needs to be confirmed. There could, however, be nutritional implications, including the possibility of an additional obesity-related gene on chromosome 1,32 and the observation should be explored further.
In conclusion, we could find no association between the MTHFR genotype and either the occurrence or severity of CAD in male and female white patients with CAD documented angiographically, nor was there an association with a history of myocardial infarction or the occurrence and severity of angina pectoris. Thus, the MTHFR genotype does not appear to be a risk factor for coronary disease in this population. There was, however, a highly significant association with BMI, which requires further evaluation.
Selected Abbreviations and Acronyms
|BMI||=||body mass index|
|CAD||=||coronary artery disease|
This work was supported by grants from the National Health and Medical Research Council of Australia. We wish to thank Dr Bridget Wilcken for reviewing the manuscript; Lily Fenech, Shelly Brown, Steven Brouwer, Dr C.J. Ma, Dr Greg Cranny, and nurses in the Eastern Heart Clinic for their assistance in clinical data collection; Dr Jun Wang for his laboratory assistance; and J. Kessey for the data entry. We are also very grateful to the cardiologists in the Department of Cardiovascular Medicine for allowing us to study their patients.
Wilcken DEL, Wilcken B. The pathogenesis of coronary artery disease: a possible role for methionine metabolism. J Clin Invest. 1976;57:1079-1082.
Dudman NPB, Wilcken DEL, Wang J, Lynch JF, Macey D, Lundberg P. Disordered methionine/homocysteine metabolism in premature vascular disease: its occurrence, cofactor therapy, and enzymology. Arterioscler Thromb. 1993;13:1253-1260.
Wilcken DEL, Dudman NPB. Homocystinuria and atherosclerosis. In: Lusis AJ, Rother JI, Sparkes RS, eds. Molecular Genetics of Coronary Artery Disease: Candidate Genes and Processes in Atherosclerosis. Basel, Switzerland: Karger; 1992;14:311-324.
Mudd SH, Levy HL, Skovby F. Disorders of transsulfuration. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The Metabolic Basis of Inherited Disease. 6th ed. New York, NY: McGraw-Hill International Book Co; 1995:1279-1326.
Hu FL, Gu Z, Kozich V, Kraus JP, Ramesh V, Shih VE. Molecular basis of cystathionine beta-synthase deficiency in pyridoxine responsive and nonresponsive homocystinuria. Hum Mol Genet. 1993;2:1857-1860.
Kang SS, Wong PWK, Bock HG, Horwitz A, Grix A. Intermediate hyperhomocysteinemia resulting from compound heterozygosity of methylenetetrahydrofolate reductase mutations. Am J Hum Genet. 1991;38:546-551.
Kang SS, Passen EL, Ruggie N, Wong PW, Sora H. Thermolabile defect of methylenetetrahydrofolate reductase in coronary artery disease. Circulation. 1993;88:1463-1469.
Miller SA, Dykes DD, Polesky HF. A simple salting-out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. 1988;16:1215.
Wang XL, Tam C, McCredie RM, Wilcken DEL. Determinants of severity of coronary artery disease in Australian men and women. Circulation. 1994;89:1974-1981.
Emery AEH. Hardy-Weinberg equilibrium: the estimation of gene frequencies. In: Emery AEH, ed. Methodology in Medical Genetics: An Introduction to Statistical Methods. Edinburgh, Scotland: Churchill Livingstone; 1976:3-9.
Hegele RA, Brunt JH, Connelly PW. Genetic variation on chromosome 1 associated with variation in body fat distribution in men. Circulation. 1995;92:1089-1093.