点击显示 收起
1 From the Department of Internal Medicine, University of California, Davis, Davis, CA (AMD and CHH), and the Clinical Trial Service Unit and Epidemiological Studies Unit (RC and JB) and the University Division of Clinical Gerontology (JGE), Nuffield Department of Clinical Medicine, Radcliffe Infirmary, University of Oxford, Oxford, United Kingdom
2 Supported by the European Union Commission Demonstration Project contract no. BMH4-98-3549; by grants from the Medical Research Council, the Clothworkers' Foundation, and the Department of Health, London; and by the US Public Health Service, National Institutes of Health grants DK56085 and DK35747 (to CHH). 3 Reprints not available. Address correspondence to AM Devlin, Nutrition Research Program, Child & Family Health Research Institute, Department of Pediatrics, University of British Columbia, Vancouver, V6H 3N1, Canada. E-mail: angela.devlin{at}ubc.ca.
ABSTRACT
Background: Homocysteine concentrations are influenced by vitamin status and genetics, especially several polymorphisms in folate-metabolizing genes.
Objective: We examined the interactions and associations with serum total homocysteine (tHcy) and folate concentrations of polymorphisms in the following folate-metabolizing genes: methylenetetrahydrofolate reductase (MTHFR), reduced folate carrier 1 (RFC1), and glutamate carboxypeptidase II (GCPII).
Design: Healthy volunteers (436 men and 606 women; mean age: 77.9 y) were randomly selected from among residents of Oxford, United Kingdom. We determined the individual effects and interactions of the MTHFR 677CT, MTHFR 1298AC, RFC1 80GA, and GCPII 1561CT polymorphisms on serum tHcy and folate concentrations.
Results: Subjects with the MTHFR 677TT genotype had higher serum tHcy concentrations than did those with the MTHFR 677CC genotype (P < 0.001), and this effect was greater in subjects with low serum folate status (P for interaction = 0.026). The MTHFR 1298AC, RFC1 80GA, and GCPII 1561CT polymorphisms had no individual effects on serum tHcy or folate concentrations. There was no interactive effect of the MTHFR 677CT and MTHFR 1298AC polymorphisms on tHcy concentrations. An interaction (P = 0.05) was observed between the MTHFR 677TT and RFC1 80GG genotypes, whereby persons with this genotype combination had a mean (±SEM) serum tHcy concentration (18.5 ± 1.2 µmol/L) that was 5.1 µmol/L greater than the mean value of 13.4 ± 0.2 µmol/L for the whole population.
Conclusions: Folate and tHcy concentrations were not affected individually by the MTHFR 1298AC, RFC1 80GA, or GCPII 1561CT polymorphisms or by combinations of the MTHFR 677CT and MTHFR 1298AC genotypes. An interaction between the MTHFR 677TT and RFC1 80GG genotypes was observed whereby persons with this combination had higher serum tHcy.
Key Words: Serum folate • serum total homocysteine • methylenetetrahydrofolate reductase 677CT polymorphism • methylenetetrahydrofolate reductase 1298AC polymorphism • reduced folate carrier 1 80GA polymorphism • glutamate carboxypeptidase II 1561CT polymorphism • elderly
INTRODUCTION
Hyperhomocysteinemia is an independent risk factor for cardiovascular disease (1–3), Alzheimer disease (4, 5), cognitive impairment (6), and osteoporotic fractures (7). Folate status is the most important vitamin determinant of blood total homocysteine (tHcy) concentrations (8), and several polymorphisms in genes for enzymes required for folate metabolism have been shown to be associated with elevations in tHcy.
Homocysteine is metabolized in the methionine cycle. Methylenetetrahydrofolate reductase (MTHFR) plays an integral role in the cycle by supplying 5-methyltetrahydrofolate for the remethylation of homocysteine to methionine by methionine synthase, which requires vitamin B-12 as a cofactor (Figure 1). A common polymorphism in the MTHFR gene, 677CT, produces a thermolabile variant of the enzyme. About 10–12% of whites of northern European decent have the MTHFR 677TT genotype, which is associated with higher blood tHcy concentrations than are found in persons with the MTHFR 677CC genotype. Some population-based studies have shown an increased risk of cardiovascular disease in persons with the MTHFR 677TT genotype, especially in those with low folate status (3, 9), whereas others have not (10). Another common polymorphism in the MTHFR gene, 1298AC, has been described (11), and heterozygosity for both the MTHFR 677CT and 1298AC polymorphisms is associated with elevated blood tHcy in the presence of low folate status, similar to what is observed for the MTHFR 677TT genotype (12). However, this finding has not been confirmed (13).
FIGURE 1.. Metabolic interactions of glutamate carboxypeptidase II (GCPII), reduced folate carrier 1 (RFC1), and methylenetetrahydrofolate reductase (MTHFR) in the small intestine and liver. B12, vitamin B-12; BBM, brush border membrane; BLM, basolateral membrane; MS, methionine synthase; 5-MTHF, 5-methyltetrahydrofolate; 5,10-MTHF, 5,10-methylenetetrahydrofolate; THF, tetrahydrofolate.
Conflicting data exist on the effect of genetic variants of proteins required for the intestinal absorption of dietary folates on blood tHcy concentrations. Intestinal absorption of polyglutamylfolates, the predominant form of naturally occurring dietary folates, first involves cleavage of the glutamate side chain by glutamate carboxypeptidase II (GCPII), followed by transport of the monoglutamyl folate derivative across the intestinal brush border membrane by the reduced folate carrier 1 (RFC1) (Figure 1; 14, 15). Our group (AMD and CHH) were the first to identify a CT polymorphism at nucleotide 1561 in the GCPII gene and found that the GCPII 1561CT genotype was associated with lower serum folate and higher tHcy concentrations in 75 healthy subjects (16); 2 larger studies, however, failed to confirm this association (17, 18). A GA polymorphism at nucleotide 80 in the RFC1 gene has also been identified, but no effect on serum folate and tHcy concentrations was observed in 169 subjects, although higher tHcy concentrations were found in 4 subjects with the combined MTHFR 677TT and RFC 80GG genotypes (19).
The goal of the present study was to examine the potential interactions and associations of the MTHFR 677CT, MTHFR 1298AC, GCPII 1561CT, and RFC1 80GA polymorphisms with serum folate and tHcy concentrations. We analyzed DNA samples from the Oxford Healthy Ageing Project (a contributor to the Medical Research Council Cognitive Function and Ageing Study). The objectives of this project included evaluation of the relations of the values of serum vitamin concentrations, tHcy, and methylmalonic acid to the risk of dementia and cognitive impairment (20, 21).
SUBJECTS AND METHODS
Study population
The Oxford Healthy Ageing Project involves a random sample of 2740 persons aged 65 y who resided in the city of Oxford, United Kingdom, when first examined between 1991 and 1994 (20). The sample was drawn from general practice registers to provide equal numbers of subjects aged 65–74 y and 75 y. All participants took part in a structured interview in their own homes. The collected data included the subjects' medical history, smoking habits, alcohol consumption, and use of multivitamin supplements. Between 1994 and 1996, all surviving participants were invited to provide a blood sample. The mean (±SEM) age at blood collection was 77.9 ± 0.2 years. DNA samples and phenotypic data were obtained from 1041 persons, of whom 436 were men and 606 were women (Table 1). All participants provided informed consent according to procedures required by the Local Research Ethics Committee, which also approved the study protocol. DNA samples and phenotypic data were anonymous.
View this table:
TABLE 1. Selected characteristics of the study population1
Laboratory methods
As previously described (20), serum tHcy concentrations were measured on an Abbott IMx autoanalyzer in Oxford by means of a fluorescence polarization immunoassay [FPIA; AXIS-Shield, Oslo, Norway (22, 23]. The CV for the FPIA tHcy assays was <3.5% in the Oxford laboratory (24). For a subset of 400 samples, repeat tHcy measurements were made by gas chromatography–mass spectrometry (GCMS) at the University of Bergen, and the correlation coefficient between the GCMS and FPIA tHcy assays was 0.99 (25). Serum folate assays were carried out by use of a microbiological method that had an interassay CV of 6% (26). Serum vitamin B-12 concentrations were measured by a competitive protein binding assay at Aarhus University Hospital, Denmark, on an ACS Centaur with an automated chemiluminescence detection system (Bayer A/S, Oslo, Norway) that had an analytic imprecision of <10%. Methylmalonic acid concentrations were measured by GCMS (25) at the University of Bergen, as described previously (20). DNA was extracted from blood samples by using the DNeasy blood kit (Qiagen, West Sussex, United Kingdom) according to the manufacturer's protocol. Genotyping for each polymorphism was accomplished by polymerase chain reactions followed by restriction digests as described previously for the MTHFR 677CT polymorphism (27), the MTHFR 1298AC polymorphism (12), the GCPII 1561CT polymorphism (16), and the RFC1 80GA polymorphism (19). Combinations of MTHFR 677CT and 1298AC genotypes were analyzed by Haploview (Cambridge, MA) (28).
Statistical methods
Because the distributions of serum tHcy and folate values had a positive skew, all analyses were carried out after logarithmic transformation. Results with approximate SEs are presented after transformation back to the original scale. Differences between mean values of serum tHcy and folate by sex were compared by analysis of variance (ANOVA). ANOVA was also used to compare mean serum concentrations of tHcy between different polymorphism groups. When comparing serum tHcy concentrations between polymorphism groups, the ANOVA models were extended to include age, sex, vitamin B-12, and methylmalonic acid as covariates. Similar ANOVA models were used in the analyses for serum folate concentrations. The ANOVA models were also extended to include first-order interaction terms for serum folate with MTHFR 677CT genotype and to explore differences by sex, polymorphisms, and genotype-genotype interactions when classified by each other. In the event of any significant interactions between genotypes and quintiles of serum folate in the ANOVA model, comparisons of the mean serum tHcy concentrations in the top and bottom quintiles of serum folate for each genotype were made by using a t test with Bonferroni correction for multiple comparisons. All analyses used SAS for WINDOWS, version 8.1 (SAS Institute Inc, Cary, NC).
RESULTS
Selected characteristics of the study population are shown in Table 1. Among the 1041 subjects on whom we obtained data, 42% were men and 58% women; 16% were current cigarette smokers. There were no significant differences between the men and the women in the median and ranges of serum folate concentrations. However, compared with women, men had lower median serum vitamin B-12 concentrations (P = 0.005) and higher serum tHcy concentrations (P = 0.008), which is consistent with findings from the entire Oxford Health Ageing Project population, although methylmalonic acid concentrations were similar in both sexes (20).
Shown in Table 2 are mean serum tHcy and folate concentrations for the combined population of men and women according to genotype of each of the 4 polymorphisms analyzed. The observed frequencies of each of the 4 polymorphisms were compared with the expected frequencies and did not deviate from Hardy-Weinberg equilibrium. ANOVA with adjustment for sex, age, and vitamin B-12 concentrations showed that mean serum tHcy concentrations differed significantly among the MTHFR 677CT genotype groups (P < 0.0001) but not among the MTHFR 1298AC genotype groups. Subjects with the MTHFR 677TT genotype had higher serum tHcy concentrations than did those with the MTHFR 677CC (P = 0.001) or MTHFR 677CT genotypes (P < 0.001), and subjects with the MTHFR 677CT genotype also had significantly higher serum tHcy concentrations than did those with the MTHFR 677CC genotype (P < 0.05). Paradoxically, the mean serum folate concentration in subjects with the MTHFR 677CT genotype was higher than that found in subjects with the MTHFR 677CC genotype (P < 0.05).
View this table:
TABLE 2. Distribution of serum total homocysteine and folate concentrations according to genotype1
Contrary to our previous finding (16), we found no significant effect of the GCPII 1561CT genotype on serum tHcy or folate concentrations. The MTHFR 1298AC and RFC1 80GA polymorphisms also had no individual effects on serum tHcy or folate. There were no significant differences between the sexes in the effects of any of these polymorphisms on serum tHcy or folate.
The distribution of the genotype frequencies and their potential interactive effects on serum tHcy concentrations are shown in Table 3. Six of 9 possible genotype combinations were found for the 2 MTHFR 677CT and 1298AC polymorphisms, whereas no subjects had the MTHFR 677CT/1298CC, 677TT/1298CC, or 677TT/1298AC genotype. Haplotype analysis showed linkage disequilibrium between these 2 genotypes: D1 = 1 (28). There was no significant interaction of the 2 MTHFR genotypes on tHcy concentrations, nor of the GCPII CT genotype with any other genotype. A borderline interaction was observed between the RFC1 80GG genotype and the MTHFR 677TT genotype for serum tHcy concentrations (P = 0.05). The mean serum tHcy concentration for the 31 subjects (3% of the study population) who had the combined RFC1 80GG and MTHFR 677TT genotype was 18.5 µmol/L, which was 5.1 µmol/L higher than the mean (±SEM) serum tHcy concentration of 13.4 ± 0.2 µmol/L for the entire population. No other interactive effects between the other genotypes were observed for serum tHcy concentrations.
View this table:
TABLE 3. Frequency of genotype combinations and interactions of genotypes on serum total homocysteine (tHcy) concentrations1
The higher serum tHcy concentrations in persons with the combined RFC1 80GG and MTHFR 677TT genotype was not explained by marginal vitamin B-12 status because the mean (±SEM) concentration in this group was 333 ± 58.3 pmol/L compared with 252 ± 3.1 pmol/L for the entire population. There were no significant trends in mean vitamin B-12 concentrations among the different genotype groups. Moreover, the effect of this genotype combination on serum tHcy concentrations remained significant (P = 0.026) after further adjustments for age, sex, methylmalonic acid, and vitamin B-12 concentrations. The higher serum tHcy concentrations in this genotype combination cannot be explained by marginal folate status because the mean (±SEM) serum folate concentration in this group was 13.6 ± 1.5 nmol/L.
We also investigated the interactive effects of the MTHFR 677CT, MTHFR 1298AC, GCPII 1561CT, and RFC1 80GA polymorphisms with folate status on serum tHcy concentrations (Table 4). A significant interaction (P = 0.026) between folate status and the MTHFR 677CT polymorphism was observed. Subjects with the MTHFR 677TT genotype who were in the lowest folate quintile had significantly higher mean serum tHcy concentrations than did those of the same genotype in the highest serum folate quintile (P = 0.0002) when the 2 levels were compared by using a t test with a Bonferroni correction for multiple comparisons. No significant differences were observed for the other 2 MTHFR 677 genotypes across folate quintiles.
View this table:
TABLE 4. Unadjusted serum total homocysteine concentrations according to genotype and quintile of serum folate1
DISCUSSION
The present study sought to determine the interactive effects of the folate-metabolizing gene polymorphisms MTHFR 677CT, MTHFR 1298AC, GCPII 1561CT, and RFC1 80GA on serum folate and tHcy concentrations in a large population sample of elderly subjects. At the same time, we reassessed the association of the GCPII 1561CT polymorphism with serum folate and tHcy concentrations and sought to confirm the interaction of the MTHFR 677CT/1298AC genotype and the MTHFR 677TT/RFC1 80GG genotype on serum tHcy concentrations.
Our study had several main findings. First, although we confirmed the effects of the MTHFR 677CT and 677TT polymorphisms on serum tHcy concentrations (27), we found no individual effects of the MTHFR 1298AC, GCPII 1561CT, or RFC1 80GA polymorphisms or interactions between the GCPII 1561CT and RFC1 80GA polymorphisms on serum folate or tHcy concentrations. Second, we confirmed the influence of folate status on the effect of the MTHFR 677TT genotype on tHcy concentrations (29). Third, we found no interaction between the MTHFR 677CT and 1298AC genotypes on tHcy concentrations. Fourth, we found a genotype-genotype interaction in which subjects with the MTHFR 677TT/RFC1 80GG genotype had significantly higher serum tHcy concentrations than the mean serum tHcy concentration observed for the whole population, which supports the findings of a smaller previous study (19). Last, we found no effect of the GCPII 1561CT genotype, which therefore did not confirm our previous finding in a smaller population of higher serum tHcy and lower serum folate concentrations in subjects with this genotype (16).
The MTHFR 677CT polymorphism is the most important genetic factor that influences blood tHcy concentrations in the general population (9, 27). In the present study, 11.6% of the study population had the MTHFR 677TT genotype and had 2.6-µmol/L higher serum tHcy concentrations than did those with the MTHFR 677CC genotype and 2.1-µmol/L higher serum tHcy concentrations than did the overall study population (Table 2). Furthermore, subjects with the MTHFR 677CT genotype (41% of the study population) had serum tHcy concentrations that were 0.6-µmol/L higher than those of subjects with the MTHFR 677CC genotype, which is consistent with findings from other studies (9, 27). In the present study, the effect of the MTHFR 677TT genotype on serum tHcy concentrations was most significant among those with low folate status (Table 4), which confirms an earlier report on the interactive effect of folate status and MTHFR 677TT genotype on blood tHcy concentrations (29).
We also determined the frequency of the 9 possible genotypes that could result from various combinations of the MTHFR 677CT and MTHFR 1298AC polymorphisms (Table 3). We found no subjects with the MTHFR 677CT/1298CC, MTHFR 677TT/1298CC, or MTHFR 677TT/1298AC genotype, as was reported for other populations (12, 30–32). Another study found the MTHFR 677CT/1298CC and MTHFR 677TT/1298CC genotypes only in fetal tissue derived from spontaneous and therapeutic abortions, which suggests that these genotype combinations may be detrimental to fetal viability (31). Although we confirmed that the MTHFR 677CT and MTHFR 1298AC polymorphisms exist in linkage disequilibrium (32), we did not confirm an interactive effect on serum tHcy concentrations that was reported by others (12, 31, 32).
We found that subjects with the MTHFR 677TT/RFC1 80GG genotype (3% of our study population of 1041 subjects) had a mean serum tHcy concentration that was 1.5 times higher than the tHcy concentration of those with the MTHFR 677CC/RFC1 80AA genotype and 1.4 times greater than the mean serum tHcy concentration of 13.4 ± 0.2 µmol/L for the entire study population (Table 3). The present data confirm findings from a previous report in which this genotype-genotype interaction on plasma tHcy was found in a much smaller population (19). The higher serum tHcy concentrations observed in subjects with the MTHFR 677TT/RFC1 80GG genotype combination could result from the combined biological effects of decreased production of 5- methyltetrahydrofolate by impaired MTHFR activity observed with the MTHFR 677TT variant and diminished folate transport across cell membranes by impaired functioning of the RFC1 80GG variant.
Circulating concentrations of 5-methyltetrahydrofolate are regulated in part by the intestinal absorption of dietary folates and are therefore dependent on the ability of GCPII to hydrolyze the glutamate side chain of folylpoly--glutamates and of RFC1 to transport monoglutamylfolates across the intestinal brush border membrane (14, 15). In the presnet study, we found no effect of the GCPII 1561CT polymorphism on serum tHcy or folate concentrations, in contrast with our previous finding that 6 (12%) of 75 healthy English subjects who had the GCPII 1561CT genotype had lower serum folate concentrations and higher serum tHcy concentrations than did subjects with the GCPII 1561CC genotype (16). This discrepancy could be a result of the small sample size in our previous report compared with the 1041 subjects analyzed in the current study. A previous study in a larger population of cardiovascular disease patients (190 subjects) and controls (601 subjects) found that the GCPII 1561CT and GCPII 1561TT genotypes had no effect on pre- or post-methionine-load plasma tHcy concentrations and higher plasma and erythrocyte folate concentrations (18). Another study of 1913 healthy subjects found that the GCPII 1561CT and GCPII 1561TT genotypes in men were associated with higher plasma folate concentrations than in men with the GCPII 1561CC genotype but found no effect of the GCPII 1561CT polymorphism on plasma folate concentrations in women or on plasma tHcy concentrations in men and women (17). A more recent study also showed that subjects with the GCPII 1561CT genotype had higher serum and erythrocyte folate concentrations but found no effect of the GCPII 1561T allele on the absorption of a polyglutamyl folate supplement (33).
In summary, the findings of the present study illustrate that the effects of polymorphisms in folate-metabolizing genes on serum tHcy may be confounded by interactions with other polymorphisms or with environmental determinants that influence folate status. Although we failed to confirm our original finding on the functional significance of the GCPII 1561CT genotype (16), additional studies in larger populations are required to determine the potential effects of the described genotype interactions on clinical phenomena related to elevated serum tHcy concentrations.
ACKNOWLEDGMENTS
AMD performed all the genotyping and most of the writing. RC provided all the phenotypic data and shared in data analysis and writing. JB performed all the statistical analyses. JGE and CHH provided oversight to the project. None of the authors had any conflicts of interest with this work.
REFERENCES