The enzyme 5,10 methylenetetrahydrofolate reductase (MTHFR) catalyzes the conversion of 5,10 methylenetetrahydrofolate into 5-methyl-tetrahydrofolate, which is the major circulating form of folate. Folate in turn is used in many biochemical pathways, including the methylation of homocysteine and other compounds and the synthesis of nucleotides.

The gene for MTHFR is located on chromosome 1 at 1p36.3. Of the allelic variants described so far, most are rare, induce severe MTHFR deficiency, and are one of the causes of homocystinuria, a rare metabolic disorder. Two alleles, however, the thermolabile or C677T allele and the A1298C allele (also referred to as C1298A allele) are common and are not associated with homocystinuria. The C677T allele causes a mild enzymatic dysfunction that results in mild homocystenemia in persons whose folate status is not optimal. The C677T allele has been linked to an increased risk of spina bifida and anencephaly, two severe neural tube defects (NTDs). Anencephaly is invariably fatal and is a leading cause of infant death due to congenital anomalies. Spina bifida causes neurologic disabilities, including motor paralysis, that can lead to lifelong disability and premature death.

The prevalence of the C677T allele has not been established with precision but appears to differ across populations. Allele frequency ranges from less than 10 percent in some groups of Africans from sub-Saharan Africa and some blacks of African origin living in the Americas, to 40 percent and higher among Hispanics in California and Italians, respectively. Compared to infants without the C677T allele, infants homozygous for the C677T allele are at a moderately increased risk for NTDs, with a pooled odds ratio (OR) from published studies of approximately 1.6 (95 percent confidence interval [95% CI], 1.2 - 2.0). Mothers homozygous for the C677T allele also appear to be at an increased risk of having an NTD-affected pregnancy (pooled OR, 2.0; 95%CI, 1.3 - 3.0), compared to mothers who are not homozygous, although homozygous fathers are not. If the association is causal, the fraction of cases attributable to C677T homozygosity appears to be low (pooled attributable fraction, 6 percent).

More limited data is available on the A1298C allele of the MTHFR gene. These data so far do not suggest that the A1298C variant is a risk factor for NTDs per se, although some studies suggest that compound heterozygosity (C677T/A1298C genotype) may confer an increased disease risk.

Gene-gene and gene-environment interactions are currently being explored. One study suggests that vitamin use reduces the risk of spina bifida in infants homozygous for the C677T variant, but larger studies are needed to establish whether this reduction is substantially different compared to infants with other MTHFR genotypes. An interaction for NTD risk has been suggested to occur between the C677T allele of MTHFR and either an insertion mutation of cystathionine-beta-synthase or a the Gly919 mutation of methionine synthase, though confirmatory data are clearly needed.

The current understanding of MTHFR and its effects on health and disease still presents critical gaps. Large, population-based studies will be instrumental in assessing the prevalence of MTHFR genotypes in well-defined, diverse populations, in expanding the spectrum of health outcomes (including other birth defects) related to MTHFR alleles, and in determining the associated risks (absolute, relative and attributable). In addition, the systematic study of interactions, in particular with micronutrient intake, may suggest opportunities to further improve public health strategies for primary prevention of NTDs.

The MTHFR gene is located on chromosome 1 at 1p36.3. The cDNA sequence is 2.2 kilobases long and appears to consist of 11 exons (1). Alternative splicing of the gene has been observed both in humans and in mice (1). The major product of the MTHFR gene in humans is a catalytically active 77-kDa protein, although a smaller isoform of approximately 70kDa has been observed in some tissues (2). MTHFR (EC 1.5.1.20) catalyzes the conversion of 5,10 methylenetetrahydrofolate into 5-methyltetrahydrofolate (5-MTHF) which is the major circulating form of folate.

The biochemical pathways involving folic acid and MTHFR are complex and have been reviewed in detail (3). Briefly, 5-MTHF, the methylated form of folate, provides the carbon moiety that is used to convert for example homocysteine into methionine, a reaction catalyzed by methionine synthase. The remethylation of homocysteine to methionine is an important step in the metabolic network that regulates the biosynthesis of nucleosides, the methylation of DNA, proteins and lipids, and the levels of homocysteine and methionine (Figure 1). The metabolic network is complex and relies on multiple activators and inhibitors. For instance, a derivate of methionine, S-adenosyl methionine (SAM) (Figure 1), is an allosteric inhibitor of MTHFR and an activator of cystathionine ß-synthase and regulates two main outflow paths of homocysteine. Although the complete effects of normal and abnormal folate metabolism are still incompletely understood, there is growing evidence that normal MTHFR activity may contribute to maintaining the pool of available circulating folate and methionine and prevent a buildup of homocysteine; conversely, abnormally low MTFHR activity may lead to lower levels of circulating folate, lower availability of methionine, and higher levels of homocysteine.

click to view

Figure 1 Simplified metabolic pathways involving 5,10-methylenetetrahydrofolate reductase (MTHFR). MSR, methionine synthase reductase.

How these biochemical effects translate into derangements of developing embryonic structures is still unknown. On a basic level, methionine is used in the formation of SAM; SAM in turn plays a major role in many methylation processes, including methylation of DNA, proteins, neurotransmitters, and phospholipids. It has also been suggested that homocysteine or its derivatives may be toxic in high doses to developing tissues (4). To add to the system’s complexity, many enzymes in addition to MTHFR are involved in folate-related metabolic pathways (Figure 1)(3).

Of the gene variants reported to date, most were identified in patients with homocystinuria, a rare, severe, autosomal recessive metabolic disorder. These 14 mutations were associated with severe MTHFR deficiency (5-7) and are individually very rare, each having been found in only one or two families. Because these variants have not been linked to common birth defects so far, they will not be discussed further.

Two MTHFR variants, however, the C677T and the A1298C mutations, are common in many populations and have been studied in relation to birth defects, mainly spina bifida and anencephaly. One study reported a third common mutation (T1059C), that was described as a silent polymorphism apparently cotrasmitted within the same gene that harbored the A1298C mutation in that particular study group. The C677T and the A1298C mutations will be the focus of this review. Although these two variants also have been associated with an increased risk for diseases other than congenital malformations, in particular adult cardiovascular disease, stroke, and coagulation abnormalities, these associations will not be discussed.

C677T Allele

The C677T allele is a single base pair mutation in which a cytosine is converted to a thymine at basepair 677, resulting in an amino acid substitution (alanine to valine)in the enzyme (2,8). The molecular genetics of the mutation have been recently reviewed (2). Functionally, the encoded protein has a reduced enzymatic activity at 37 degrees C (9) and higher, so that the C677T mutation is often termed "thermolabile". For instance, compared to similarly treated controls, samples from C677T homozygotes have 50 to 60 percent lower MTHFR activity at 37 degrees C, and about 65 percent lower activity at 46 degrees C (2). Heterozygotes are in the intermediate range.

The variations in MTHFR activity associated with the C677T mutation appear to correlate with biochemical abnormalities. For instance, compared to controls, C677T homozygous persons have higher plasmatic levels of homocysteine (2); the homocysteine levels, however, depend in part on folate levels in that homocysteine is increased among those in whom plasma folate is at the lower end of the normal range, but not when folate levels are normal. Erythrocyte folate, on the other hand, is not decreased, since 5-MTHF is the major circulating form of folate and not a significant storage form (2).

A1298C Allele (C1289A)

The A1298C variant of MTHFR was first identified in 1995 during a study of ovarian cancer (10). In the A1298C allele, a point mutation in exon 7 results in the coding of a glutamate instead of an alanine residue (10-12). This allele has also been termed C1289A mutation by other authors (13).

Functionally, the mutation results in mildly decreased MTHFR activity (11, 12). Unlike the case with the C677T mutation, neither homozygous nor heterozygous individuals had significantly higher plasma homocysteine or lower plasma folate compared to controls (11, 12); however, persons who were compound heterozygotes for the A1298C and C677T mutations presented a biochemical profile that included increased homocysteine and decreased plasma folate levels similar to that seen in persons homozygous for the C677T mutation (11).

Population Frequencies

Data sources

Although many reports are available on MTHFR frequency in several countries worldwide, most are based on groups of individuals that are either convenient samples of relatively undefined populations or groups for which the selection procedure is incompletely documented. Thus, conclusions based on most current data should be attempted with caution, recognizing the potential for bias. This review includes data from surveys of healthy individuals as well as from healthy controls included in case-control studies of various diseases. Data from affected controls were not included. We restricted our review to those studies with at least 100 informative individuals, although we included some smaller studies if they were the only source of data for a particular country or if they were studies of non-white races or ethnic groups. We did not include studies published only as abstracts. When needed, we recalculated the measures of interest (allele frequency, frequency of specific genotypes, confidence intervals) from data in the published reports. We computed confidence intervals using normal approximation and correction for continuity (14).

Results

A. C677T Allele

Prevalence by geographic region and racial/ethnic group

The frequency of the C677T allele shows geographic and ethnic/racial variations (Figure 2). It appears to be high in Italy (pooled estimate, 44 percent) and in white hispanics from California (42 percent), and relatively low in some samples of individuals from sub-Saharan Africa (7 percent) and individuals of African descent living outside of Africa (e.g., 14 percent in US blacks, pooled from 3 studies) (Table 1). As expected, the frequency of C677T homozygotes follows a similar pattern (Figure 3). The reason for the high frequency of the C6T77T variant in many populations is unclear. Table 1 summarizes the data and provides pooled estimates by country. When a study includes more than one sample (e.g., newborns screened at birth, healthy adult volunteers, and blood donors), these are entered separately in Table 1. Following is a summary by major ethnic groups.

click to view

Table 1 International distribution of the C677T allele of the methylenetetrahydrofolate reductase (MTHFR) gene, by geographic area and ethnicity*

click to view

Figure 2 Frequency of the C677T allele of MTHFR, with pooled estimates by country, 1995-1999

click to view

Figure 3 Frequency of homozygous genotype of the C677T allele of MTHFR, with pooled estimates by country, 1995-1999.

Whites. Studies from Europe include data on about 7,000 individuals. The C677T allele frequency among Europeans ranged from 22 percent in Norway to 44 percent in Italy (Table 1)(15-37). In Britain and Ireland, two countries with historically high rates of NTDs, the allele frequency was 37 and 32 percent, respectively. Among whites outside of Europe the allele frequency of C677T was similar, in the aggregate, to the frequencies observed in Britain and Ireland. Specifically, the allele frequency among whites from Australia, Brazil, Canada and the United States ranged from 34 to 37 percent (38-50). As expected, countries in which the frequency of C677T allele was highest also had the highest frequency of C677T homozygotes followed; for example, in Italy the frequence of C677T homozygotes reached 18 percent (95% CI, 17-20).

Blacks. In one study from sub-Saharan Africa that included genotype information, no C677T homozygotes were identified among the 234 individuals tested(51). The C677T allele frequency was 7 percent (51). In another study of 89 sub-Saharan Africans from four tribes, the allele frequency was also 7 percent, but the homozygote frequency was not reported (52). These estimates are relatively low compared to those of other ethnic groups. Among blacks living outside Africa, the allele frequency ranged from 5 percent among Brazilian blacks of African descent to 24 percent among members of one group of African-Americans in the United States (53-58).

Asian. The majority of studies of Asians were done on Japanese people. The C677T allele frequency of 34 percent for the Japanese was derived by pooling data from more than 3,300 persons tested in nine studies (59-67). Limited data is available for other Asian populations (51).

Hispanics. In one study from California that included a sample of 169 whites of Hispanic origin (50), the allele frequency was 42 percent (95% CI, 36-47). The allele frequency was similarly high (42 percent) in a group of Colombians (68).

Amerindians. The C677T allele frequency was high in some Amerindian groups but not others. For example, the allele frequency was high in a group of Cayapa Amerindians from Ecuador (52) and in a group of Brazilian Amerindians (51) (43 and 45 percent, respectively); however, in members of the Tupi Parakana tribe, the allele frequency was 11 percent (69).

Other ethnic groups. There are limited data on the distribution of MTFHR mutation in other populations in the world (70), including some small or isolated ethnic groups (51, 52).

Prevalence by sex

Most published studies either do not specify the gender composition of the samples, do not comment on differences of genotype frequencies by sex, or state that genotype frequency is not significantly different in males and females. One study reported a lower proportion of C677T homozygotes in newborn females compared to newborn males (71).

Prevalence by age

One study from Japan reported a lower C677T allele frequency in older people than in younger people and both in males than in females (72). The frequency of C677T homozygotes was 7 percent among people 80 years of age or older, compared to 14 percent among people 55 to 79 years old, and 19 percent among people 14 to 55 years old (72).

In the Netherlands, the frequency of the C677T mutation was significantly lower in the elderly (85 years or older) than in the young (18 to 40 years) (73); however, the difference in genotype distribution by age was only present in males.

B. A1298C allele

To date, data on the prevalence of the A1298C allele in the population is limited to relatively small groups of controls from case-control studies. The frequency of A1298C homozygotes among controls was approximately 9 percent in two studies, one from Canada (11) and one from the Netherlands (12). Still among controls, the frequency of C677T/A1298C compound heterozygotes was 15 percent in the Canadian study (11) (18 percent in the mothers and 11 percent in the children), 20 percent in the Dutch study (12), and appeared to be 17 percent in a U.S. study (13).

C. T1059C allele

The T1059C allele has been recently described (13) and was found to be present in 85 percent (allele frequency) of a group of control subjects from Iowa (U.S.). The distribution by genotype was not provided. This silent mutation was apparently cotrasmitted with the A1298C mutation (13).

Almost all studies on the relationship between MTHFR variants and birth defects have focused on NTDs. A vast literature is available on NTDs. A set of reviews (74) that address the epidemiology, embryologic mechanisms, animal models, genetics, efficacy of folic acid, fetal therapy, clinical treatment, and ethical issues related to NTDs has recently been published.

Briefly, NTDs comprise a range of congenital malformations associated with the failure of the neural tube to close during embryonic development. NTDs include spina bifida, in which the bony spine fails to close posteriorly, causing the hearniation of meninges (meningocele), neural tissue (myelocele), or both (myelomeningocele). NTDs also include anencephaly (a condition in which the brain is exposed and incompletely developed due to incomplete closure of the cranial portion of the neural tube), and encephalocele (a condition in which meninges and neural tissue herniate through a defect in the skull, posteriorly [occipital encephalocele] or anteriorly [anterior encephalocele]).

NTDs are clinically severe conditions. Anencephaly is fatal. Spina bifida is associated with a broad variety of neurologic deficits that disrupt bowel, bladder, sexual, and motor function. Often joint and bone deformities add to the already considerable disease burden. Learning and developmental disabilities are also frequent. The estimated lifetime cost of a single case of spina bifida has been estimated as 300,000 US dollars in 1992 and is probably even more now (75). The incidence of NTDs is unknown, since many affected pregnancies end in spontaneous abortion. The prevalence among births varies considerably among by country and ethnic group and range from as high as 1 case in 100 births in some regions of China (76) to about 1 case in 5,000 or less in some Scandinavian countries (77). In many countries the prevalence is approximately 1 in 1,000 births (77). Increasingly, prenatal diagnosis and termination of affected pregnancies tends to spuriously lower the prevalence at birth of NTDs, thus concealing in part the impact of NTDs on human pregnancies. It is estimated that folic acid, if taken daily in a sufficient amount, can reduce NTD occurrence as much as 50 to 70 percent (78, 79).

Neural tube defects

The search for genetic determinants of NTD risk has recently focused on genes involved in folate and homocysteine metabolism, because of the association between NTD occurrence and folic acid use, and because some studies suggested that some women (about 20 percent) who had had an NTD-affected pregnancy also had abnormal homocysteine metabolism (80, 81). Data from the studies of the C677T and A1298C mutations are summarized in Table 2 and are discussed below. The estimated odds ratios are also charted in Figures 4 and 5.

click to view

Table 2 Results of case-control studies of the association between the C677T allele of the methylenetetrahydrofolate reductase (MTHFR) gene and risk for spina bifida, by case group (people with a neural tube defect or their parents) and genotype*

click to view

Figure 4 Risk of neural tube defects associated with C677T homozygous mutation in infants, by country and year of study, and pooled estimates for C677T homozygous and heterozygote genotypes, 1995-1999.

click to view

Figure 5 Risk of neural tube defects associated with C677T homozygous mutation in mothers and fathers of infants with a neural tube defect, by country and year of study, and pooled estimates for C677T homozygous and heterozygote genotypes, 1995-1999.

Neural tube defects and the C677T allele

The first report on the C677T allele, from the Netherlands, compared 55 patients with spina bifida and 130 parents of patients with spina bifida (70 mothers and 60 fathers) to 207 unrelated control subjects (82). Homozygosity for the mutation was associated with a three-fold increased risk of being a spina bifida patient or a mother of a spina bifida patient (Table 2). Fathers were not more likely to carry the mutation compared to controls. A second study from the same group, however, showed somewhat different findings (12). It showed that homozygosity for the C677T allele was associated with a lower and non-significant increased risk for NTDs (OR, 1.7; 95%CI, 0.7-3.9) (recalculated from from Table 1 of ref /(12)). Because of the unspecified overlap with the study population from their first report, the researchers’ earlier findings, based on a smaller sample of spina bifida patients, may be superseded by the later report.

Five subsequent association studies conducted in Ireland and the United States provided further evidence for an association between the C677T allele and an increased risk for NTDs and found that infants homozygous for the C677T allele had a two- to seven-fold increased risk of having an NTD (50, 83-86). Most of these studies were conducted in white populations. One study, however, also included non-white subjects (50) and reported odds ratios for C677T homozygosity of 1.4 (95% CI, 0.6-3.2), 1.9 (95% CI, 0.9-3.9) and 1.4 (95% CI, 0.04-3.82), among non-Hispanic whites, Hispanic whites, and blacks, respectively. Although these odds ratios were not significantly heterogeneous, the sample size was small for some ethnic groups. For instance, only eight spina bifida patients were black.

Other studies conducted in Britain, France, Germany, Ireland, Norway, Turkey and United States failed to find a significant association between C677T mutation and risk for NTDs (87-93). In these studies, the odds ratios associated with homozygous C677T genotype in infants or mothers ranged from 0.6 (95% CI, 0.1-2.7) in France (93) to 2.8 (95% CI, 0.7-10.6) in Turkey (90)(Table 2).

Neural tube defects and the C677T allele: pooled estimates

For those studies that provided sufficient information to do so, we calculated Mantel-Haenszel pooled estimates of odds ratios for patients with an NTD and their parents (Table 2). We excluded one study (82) because of probable overlap with a later, larger study (12).

The pooled odds ratios for NTD among infants with the C677T/C677T genotype was 1.7 (95% CI 1.4-2.2). For the heterozygous genotype, the pooled odds ratio was 1.2 (95% CI 1.0-1.3) (Figure 4). The Mantel-Haenszel ?² test for trend was significant (?²= 13.6, p=0.0002), suggesting a relationship between number of C677T alleles and risk for NTDs. The pooled attributable fraction for C677T homozygosity in infants was approximately 6 percent (Table 2).

The odds ratios for mothers of NTD-affected pregnancies were 2.1 (95% CI 1.5-2.9) and 1.2 (95% CI 0.9-1.5) for the homozygous and heterozygous C677T genotype, respectively, and the ?² for trend also was statistically significant (?²= 7.8, p=0.005). The father’s genotype did not appear to be a significant risk factor (Figure 5).

Multiallelic effects of C677T

The increased risk associated with the mother’s but not the father’s genotype suggests that the maternal effects may play a role in modulating NTD risk in the fetus, perhaps by modifying the biochemical environment to which the fetus is exposed. By extension, a fetus homozygous for the C677T mutation might be at increased risk for NTD if also the mother is homozygous for the mutation. So far the data provide mixed results. One group from Ireland reported that if a child was a C677T homozygote the risk for NTD was not elevated if the mother was also a C677T homozygote compared to if she were a heterozygote or homozygote for the normal allele (94). A group from the Netherlands, however, stated that risk of NTD in infants was increased seven-fold if both the mother and the infant were C677T homozygotes (95).

Neural tube defects and the A1298C allele

Three groups (11-13) studied the association between the A1298C allele and NTD risk. Two of the groups (11, 12) also studied the mutation’s effect on enzyme activity.

In the two studies on enzyme activity (11, 12), the A1298C allele was found to decrease MTHFR enzyme activity though to a lesser degree compared to the C677T allele. In addition, the presence of one A1298C allele further reduced MTHFR activity in those who already had one C677T allele, suggesting that compound heterozygosity for the two alleles may carry some risk of disease.

Of the three association studies, two (11, 12) did not show an significant association between the A1298C allele and NTD risk while the third (13) did (odds ratio, 2.4; 95%CI, 1.4, 4.1). The C677T/A1298C genotype appeared to be associated with an increased NTD risk in two (12, 13) of the three studies; in the first study (12) the increase did not reach statistical significance (odds ratio, 2.0; 95% CI 0.9-4.7), while in the other (13) the odds ratio (comparing the double heterozygotes with homozygous wild type) was 2.8 (95% CI, 1.1-7.6) (recalculated from Table 13 of original paper).

Neural tube defects and the T1059C allele

In the one study that looked at the T1059C allele, this apparently silent polymorphism was associated with an increased risk for NTDs (13). However, because this allele was apparently cotrasmitted with the A1298C allele, the authors caution that this association may not be causal (13). Moreover, the authors note that the association was found in only one of the three case-control sets in the study and that the finding could not be replicated in the other two sets nor in the aggregate data set (13).

Validity and potential sources of biases

In the studies of MTHFR polymorphisms and risk for NTDs, researchers used different approaches to enroll case and control subjects. Overall, most studies used what would usually be considered convenient samples, such as subjects identified in selected hospitals, pediatric clinics (88) or with the assistance of birth defects research centers or patient/parent organizations (12, 82-86, 92, 96). Another approach is exemplified by a study in which researchers recruited prospectively subjects with an NTDs who were born to a cohort of 56,049 women attending their first antenatal visit in one of three hospitals within a well-defined geographic area in Ireland and then selected the control subjects among the same cohort of women (89). In many, the selection process was not clearly defined (87, 90, 91, 93). A minority of studies were population-based, to varying degrees (13, 50).

In addition to the traditional sources of bias common to all case-control studies, case-control studies of genetic factors may generate spurious associations if case and control sets are selected from different underlying populations, so that case-control differences in allele frequencies may reflect differences between the underlying population rather in disease status. A few studies (13, 91, 97, 98), acknowledging this potential problem, used family-based studies such as the transmission disequilibrium test to assess the role of different alleles in NTD risk.

MTHFR and other birth defects

Few studies have investigated the association between MTHFR genotypes and risk for birth defects other than NTDs. Homozygosity for C677T was not a risk factor for cleft lip in a population-based study in California (99). In a another study published as an abstract, the authors stated that homozygosity for C677T was three times higher in individuals with cleft lip and palate compared to controls (18.3 percent vs. 5.7 percent, p<0.01) (100); however, the frequency of homozygotes in the control group appears to be lower than expected on the basis of available population studies (Table 1). The expansion of MTHFR studies to other birth defects could be fruitful in light of the reports that the use of folic acid or multivitamins can be associated with a reduced occurrence of several types of birth defects, including some heart defects, oral clefts, urogenital anomalies, and limb deficiencies.

Because of their role in folate metabolism, MTHFR mutations may possibly interact with other genes involved in folate and homocysteine metabolism and with folate consumption (Figure 1).

Interaction with folic acid and vitamin use

One study so far reported data in which NTD risk was studied in relation to both vitamin use and MTHFR genotype (50). Maternal vitamin use was associated with a reduced risk for spina bifida in both infants with normal alleles and those with the C677T homozygous genotype (OR, 0.30 and 0.20, respectively). When reanalyzed from a different perspective (101), these data suggest that the effect of the C677T mutation depends of folate status; C677T homozygosity was associated with a five-fold increased risk of NTDs if mothers did not use vitamins (odds ratio, 5.2, 95% CI, 2.2-12.6), whereas the risk was not increased if they did use vitamins (odds ratio, 1.2; 95% CI, 0.3-4.3).

Interaction with other genes

Genes whose mutations have been studied in conjunction with MTHFR alleles include cystathione beta synthase (CBS), methionine synthase, and methionine synthase reductase. CBS, located at 21q22.2, catalyzes the conversion of homocysteine to cystathionine, is the first enzyme in the transulfuration path of eukaryotes, and provides an outflow route for homocysteine (Figure 1). One study (86) reported data compatible with the presence of interaction between mutations in MTHFR and CBS, specifically the C677T allele of MTHFR and the insertion-mutation of CBS. In that study (see reference (102) for a reanalysis of reference (86)), the presence of mutations in both genes was associated with a five-fold increased risk spina bifida, which appeared to be higher than expected based on the estimated effect of CBS and MTHFR mutations alone. A second study reported in abstract (103) showed similar findings. Because these studies were small, further data are needed to confirm the potential interaction and to assess its magnitude in different populations. Another study (98) reported that a variant of methionine synthase (the Gly919 variant) in combination with the C677T allele of MTHFR occured more frequently than expected in mothers of NTD offspring (OR 3.9).

Finally, homozygosity for a common mutation of methionine synthase reductase, the A66G allele, when combined with C677T homozygosity, was found in one study (104) to increase an infant’s NTD risk four-fold (odds ratio, 4.1; 95% CI, 1.0-16.4)

MTHFR mutations can be identified by direct sequencing; however, the detection of MTHFR’s common polymorphisms is simplified by the fact that these mutations create (C677T) or abolish (A1298C) specific restriction sites. Thus, the C677T variant can be detected by PCR amplification of genomic DNA, followed by digestion with the restriction enzyme HinfI, and by gel electrophoresis (8). This procedure has been performed on DNA extracted from many sources, including blood, dried blood spots, and amniotic fluid. The A1298C mutation abolishes an MboII restriction site and can be detected by analogous procedures (10).

We have been unable to find data on the analytic parameters of these tests (i.e., the specificity, sensitivity, and predictive value of the tests to classify the underlying genotype).

The population prevalence of hMLH1/hMSH2 mutation carriers in the Scottish population aged 15–74 years has been estimated at 1 in 3,139 (78). A recent UK National Screening Committee workshop concluded that there is currently no case to offer population screening in an attempt to identify mutation carriers (Rose et al., UK National Screening Committee, unpublished data). Authors in the United States have reached similar conclusions, agreeing that more information regarding the prevalence and penetrance of mismatch repair gene mutations and more evidence of effective intervention strategies are essential prerequisites for implementing screening outside of the research context (79–84).

There are essentially two strategies that could be employed to search for mutations in the context of population screening: searching the entire gene(s) for mutations using the techniques considered above or looking for specific mutations. The latter option is far less expensive and labor-intensive and could be of particular benefit in countries where specific "founder" mutations are prevalent. It may also be possible to apply DNA pooling strategies in this context to enhance efficiency (85). However, this approach is not currently feasible because of the extreme heterogeneity of mismatch repair gene variants and the low allele frequency of individual mutations. The ethical issues inherent in genetic screening, coupled with the poor efficiency and high cost of detecting mutations using current technology, mean that population testing in any form is unlikely to be recommended in the near future.

Another approach to identifying mutation carriers is performing mutation analysis in colorectal cancer patients deemed to be at high risk of harboring mutations, and subsequently performing "cascade screening" of their relatives. The major issue in the context of a cascade screening program is that of how resources can be efficiently targeted towards the identification of kindreds with hMLH1/hMSH2 mutations. This issue is considered in detail in an overview of findings from one research group (52), and the sensitivity and specificity of various clinical criteria are considered by Syngal et al. (86).

Currently, most mutation carriers are identified by referral of patients with a family history of colorectal cancer to cancer genetics services. Another option, under investigation in an ongoing program in Scotland, is to search for mismatch repair gene mutations among persons with early-onset colorectal cancer and subsequently perform cascade screening in the relatives of mutation carriers. The phenotypic features of age at onset, family history, and MSI are commonly used selection criteria in mutation analysis studies, as summarized for reference in the second supplementary tables and appendix table .

At the present time, there is no consensus regarding the most efficient approach to identifying mutation carriers. It is clear, however, that further understanding of the role of mismatch repair genes in colorectal cancer has important scientific and clinical implications.

Even after dozens of studies, many basic questions on the role of MTHFR in health and disease remain unanswered. Research priorities should include the following:

1) Improve the representativeness and precision of the estimates of MTHFR genotype frequencies.
To date, most studies have been based on small groups, convenient samples, or both. A comprehensive evaluation of MTHFR genotype frequencies should include all relevant allelic variants and evaluate different populations and ethnic groups. It is crucial to assemble sufficiently large and representative samples, such as could be achieved by using a sufficiently large set of blood spots selected at random from a universal metabolic screening program.

2) Search for additional disease alleles and genes.
The fraction of NTD cases attributable to the known MTHFR alleles is low and does not account for the 50 to 70% reduction in NTD occurrence resulting from the use of folic acid or multivitamins as observed in clinical trials. This suggests that additional genetic or environmental risk factors for the folate-preventable NTDs remain to be identified. That at least some of these unknown factors also involve folate and homocysteine metabolism is suggested by the fact that in some families with NTD-affected members, some people without the C677T mutation still had, in the aggregate, increased homocysteine levels (105). Other genes, such as those coding for folate receptors or enzymes such as methylenetetrahydrofolate dehydrogenase (MTHFD) or serine hydroxymethyltransferase, are potential candidates for further investigation (106).

3) Expand the spectrum of outcomes associated with MTHFR variants.
Other congenital anomalies, including oral clefts and some heart defects whose risk may be modified by multivitamin or folic acid use, may be candidates for studies of folate- and homocysteine-related genes. For each condition and allelic variant, the disease risks (absolute, relative, and attributable) should be assessed in different populations. To date, even for NTDs data on absolute risk is lacking, and estimates of relative risk are for the most part imprecise.

4) Explore Interactions.
Because of the basic nature and the complexity of the metabolic processes in which MTHFR is involved, it would not be unexpected that MTHFR variants interact with other disease risk factors, including other genes and environmental factors. Obvious candidates include other genes related to folate and homocysteine metabolism, as well as levels of intake of vitamins and other micronutrients

• Table 1 - International distribution of the C677T allele of the methylenetetrahydrofolate reductase (MTHFR) gene, by geographic area and ethnicity*
• Table 2 - Results of case-control studies of the association between the C677T allele of the methylenetetrahydrofolate reductase (MTHFR) gene and risk for spina bifida, by case group (people with a neural tube defect or their parents) and genotype*
• Table 3 - Results of case-control studies of the association between the C677T allele of the methylenetetrahydrofolate reductase (MTHFR) gene and risk for congenital anomalies other than neural tube defects, by case group (people with an anomaly or their parents) and genotype*
• Supplementary Tables and Appendix Table - Notes on populations listed in Table 1, worldwide Distribution of MTHFR C677T Allele.

Figures

• Figure 1 - Simplified metabolic pathways involving 5, 10 methylenetetrahydrofolate reductase (MTHFR).
• Figure 2 - Frequency of the C677T allele of MTHFR, with pooled estimates by country, 1995-1999.
• Figure 3 - Frequency of homozygous genotype of the C677T allele of MTHFR, with pooled estimates by country, 1995-1999.
• Figure 4 - Risk of neural tube defects associated with C677T homozygous mutation in infants, by country and year of study, and pooled estimates for C677T homozygous and heterozygote genotypes, 1995-1999.
• Figure 5 - Risk of neural tube defects associated with C677T homozygous mutation in mothers and fathers of infants with a neural tube defect, by country and year of study, and pooled estimates for C677T homozygous and heterozygote genotypes, 1995-1999.

List of References

Data on Disease Frequency

• EUROCAT (last accessed 5/2007)

General Information on Spina Bifida

• March of Dimes (last accessed 5/2007)
• Spina Bifida Association of America (last accessed 5/2007)

Genetic Information

• Gene Mutation Database (last accessed 5/2007)
• OMIM (last accessed 5/2007)
• GeneAtlas (last accessed 5/2007)
• Entrez (last accessed 5/2007)
• Gene Cards (last accessed 5/2007)

Support Groups

• Spina Bifida Association of America (last accessed 5/2007)