点击显示 收起
1 From the Departments of Human Genetics, Pediatrics, and Biology, McGill University–Montreal Children's Hospital Research Institute, Montreal, Canada (DL, LP, YL, QW, and RR), and the Institut de Recherches Cliniques de Montreal, Montreal, Canada (JSC)
2 Supported by the Canadian Institutes of Health Research (CIHR), by a CIHR Senior Scientist Salary Award (to RR), by a CIHR/Pfizer R & D Investigator Salary Award (to JSC), and by a fellowship (to DL) and a studentship (to LP)s from the Montreal Children's Hospital Research Institute. 3 Address reprint requests to R Rozen, Montreal Children's Hospital Research Institute, 4060 Ste Catherine Street West, Room 242, Montreal, PQ, Canada H3Z 2Z3. E-mail: rima.rozen{at}mcgill.ca.
ABSTRACT
Background: Genetic or nutritional disturbances in folate metabolism may affect embryonic development because of the critical role of folate in nucleotide synthesis and methylation reactions. The possible role of a mild deficiency in methylenetetrahydrofolate reductase (MTHFR) and low dietary folate in pregnancy outcomes and heart morphogenesis requires further investigation.
Objective: We investigated the effect of mild MTHFR deficiency, low dietary folate, or both on resorption rates, on length and weight, and on the incidence of heart malformations in murine embryos.
Design: Female Mthfr +/+ and +/– mice were fed a control diet (CD) or a folic acid–deficient diet (FADD) before mating with male Mthfr +/– mice. On gestational day 14.5, implantation and resorption sites were recorded and viable embryos were examined for gross malformations, growth delay, and congenital heart defects.
Results: Plasma homocysteine in Mthfr +/– dams and in FADD-treated dams was significantly higher than that in Mthfr +/+ dams and CD-treated dams, respectively. A significantly higher rate of resorption and greater developmental delay were observed in hyperhomocysteinemic mice than in CD-treated +/+ dams. Heart defects were identified in 4 of 11, 5 of 10, and 4 of 10 litters from CD-treated +/–, FADD-treated +/+, and FADD-treated +/– dams, respectively, but not in any of those from CD-treated +/+ dams (0/11 litters).
Conclusion: Our findings suggest that mild MTHFR deficiency, low dietary folate, or both in the dams increase the incidence of fetal loss, intrauterine growth retardation, and heart defects. These data support the benefit of folic acid supplementation in pregnant women, particularly in those with MTHFR deficiency.
Key Words: Folate • homocysteine • methionine • methylenetetrahydrofolate reductase • MTHFR • pregnancy complications • intrauterine growth retardation • fetal loss • congenital heart defects
INTRODUCTION
Folate derivatives are involved in a variety of important cellular reactions, including nucleotide synthesis and the generation of methionine and S-adenosylmethionine (SAM) in the methylation cycle. Methylenetetrahydrofolate reductase (MTHFR), a key enzyme in these pathways, catalyzes the conversion of 5,10-methylenetetrahydrofolate to 5-methyltetrahydrofolate, a major carbon donor for methionine synthesis from homocysteine. Cloning of MTHFR (1) led to the identification of a common variant (677CT; A222V) that results in a thermolabile enzyme with only 35–50% residual activity in homozygous mutant (TT) persons. This prevalent genotype (10–15% of many North American and European populations) is the most common genetic risk factor for hyperhomocysteinemia (2). Hyperhomocysteinemia, or low methionine synthesis, which results from mild MTHFR deficiency or inadequate dietary folate, may contribute to several pathologic states through various mechanisms such as direct toxic effects or through indirect effects such as a disruption in methylation or an increase in oxidative or endoplasmic reticulum stress (3-7). The well-studied pathologic states associated with homocysteine metabolism include vascular diseases and neural tube defects (4-6, 8).
A few clinical reports suggested that the common MTHFR polymorphism or low dietary folate may increase the risk of intrauterine growth retardation (IUGR), fetal loss, and other pregnancy complications (9-12); other studies did not show these associations (13-16). The relation between folate and congenital heart defects (CHDs), one of the most common types of birth defects, is also unclear. Two studies that examined the effect of maternal MTHFR mutant genotype on offspring concluded that it was associated with increased risk of heart defects (17, 18), whereas one study of both maternal and offspring genotypes did not reach this conclusion (19).
To investigate the potential influence of MTHFR deficiency and dietary folate on reproductive outcomes and CHDs, we chose a different approach—the investigation of an animal model of mild MTHFR deficiency that had been generated in our laboratory (20). Only a few animal studies have been conducted on the effects of dietary folate deficiency on reproductive outcomes (21, 22), and none have thoroughly investigated CHD or the role of Mthfr. Mthfr –/– mice have a high mortality in the first few weeks of life, whereas mice with a single knockout allele (ie, Mthfr +/–) are phenotypically normal. The latter mice represent a good animal model for mild MTHFR deficiency in humans because they have 60% residual MTHFR activity and plasma homocysteine concentrations that are 1.6 times higher than those of their wild-type (Mthfr +/+) littermates (20). Mthfr +/– mice also have low methylation potential (20) and were recently shown to have impaired acetylcholine-induced arterial relaxation (23). These findings have all been reported in clinical studies of hyperhomocysteinemia or mild MTHFR deficiency (24, 25), which supports the validity of this animal model for studies of the clinical disorders associated with disturbed folate or homocysteine metabolism. In the current study, we used these Mthfr-deficient mice to examine the effect of MTHFR deficiency, in the presence and absence of low dietary folate, on pregnancy complications and heart defects.
MATERIALS AND METHODS
Mice and diets
Animal experimentation was approved by the Montreal Children's Hospital Animal Care Committee, according to the guidelines of the Canadian Council on Animal Care. Mthfr knockout mice were generated as reported (20) and backcrossed for 10 generations onto a BALB/c background. Mice were maintained in plastic-bottomed cages embedded with shavings and were exposed to a 12-h light-dark cycle in a temperature-controlled room. After weaning, female mice were placed on amino acid–defined diets (Harlan Teklad, Indianapolis, IN) with all the necessary components recommended by the American Institute of Nutrition (26). A control diet (CD) contained the recommended amount of folic acid for rodents (2 mg/kg diet), and the folic acid–deficient diet (FADD) contained 0.3 mg folic acid/kg diet, which is 14.3% of the amount of folic acid in the CD. Both diets contained 1% succinyl sulfathiazole, an antibiotic that is used to prevent the generation of folate by intestinal bacteria (27). After 6 wk of their allotted diets, female mice were housed overnight with Mthfr +/– males (aged 80–100 d). Males were fed standard rodent chow (Agribrands Purina, St-Hubert, Canada) while they were in their own cages. The next morning, females were checked for the presence of vaginal plugs. If plugs were present, that morning was designated as gestational day (GD) 0.5. Females were fed their designated diets during the breeding period and throughout pregnancy until killed. On GD 14.5, pregnant mice were asphyxiated with carbon dioxide, and blood was collected by cardiac puncture. Resorption sites were identified by the smaller size and more necrotic hemorrhagic appearance than were seen in normal embryos and placentas. Resorption rate was calculated as the ratio of resorption sites to the total number of implantation sites (28). All embryos present were dissected and examined for gross malformations, and crown-rump lengths and individual weights were measured. Developmental delay was assessed by examining the morphology of individual viable embryos and staging them according to their resemblance to normal embryos on a given GD (29). The morphologic examination of embryos was completed by a single person, who was blinded to the litters. Embryos were then fixed in 4% paraformaldehyde overnight and transferred to 70% ethanol for storage. Their yolk sacs were dissected and washed thoroughly in phosphate-buffered saline for subsequent embryonic Mthfr genotyping.
Plasma amino acid analysis
Blood was collected into tubes containing EDTA. Plasma was immediately separated from blood by centrifugation at 4000 x g for 7 min at 4 °C. Plasma was then frozen on dry ice and stored at –70 °C. Plasma total homocysteine (tHcy) and methionine were assayed with the use of HPLC as described previously (30, 31).
Mthfr genotyping
Genomic DNA from maternal toes or livers and from embryonic yolk sacs was isolated by using phenol-chloroform extraction. Polymerase chain reaction was performed according to previously described protocols (20).
Histologic tests
Embryos were processed through a series of ethanol, xylene, and paraffin incubations, which took 45 min per step. They were then embedded and sectioned serially in 10-µm transverse sections. All sections were examined under bright-field illumination with the use of an inverted microscope. Selected sections were stained with hematoxylin and eosin and then photographed.
Statistical analysis
The dam or the litter was considered as the unit for statistical analysis. For parametric data, results were expressed as mean ± SEM and examined by two-factor analysis of variance (ANOVA). For nonparametric data, the Kruskal-Wallis test and two-factor ANOVA of the ranks were used to determine significance for multiple comparisons of resorption rate, percentage of delayed embryos, and incidence of CHD in each litter. Analyses of the contribution of the embryonic genotype to embryonic delay and CHD were conducted by using the two-tailed Fisher's exact test. All statistical analyses were performed with SPSS for WINDOWS software (version 11.0; SPSS Inc, Chicago, IL). P values < 0.05 were considered significant.
RESULTS
Plasma homocysteine and methionine concentrations
On GD 14.5, mean plasma tHcy in CD-treated +/– and FADD-treated dams (both +/+ and +/–) was significantly higher than that in CD-treated +/+ dams (control group), which was 13.81 ± 1.55 µmol/L. The highest concentrations of homocysteine (50.66 ± 1.81 µmol/L) were seen in the FADD-treated +/– dams (Figure 1). The overall effects of genotype and diet were highly significant, and there was a marked interaction (P < 0.01, two-factor ANOVA). Plasma tHcy was significantly higher in Mthfr +/– dams than in Mthfr +/+ dams, and there were significant genotype effects in both dietary groups. FADD-treated dams had significantly higher tHcy than did CD-treated dams. These results indicate that the folate-deficient diet had produced the desired increase in tHcy, and they confirm our previous finding that the Mthfr +/– genotype is associated with mild or moderate hyperhomocysteinemia (20).
FIGURE 1.. Mean (±SEM) plasma total homocysteine (tHcy) concentrations in 80–90-d-old pregnant mice fed a control diet (CD) or a folic acid–deficient diet (FADD). Numbers above bars indicate the number of animals in each group. *Significantly different from Mthfr +/+ mice fed the same diet, P < 0.01. #Significantly different from mice of the same genotype fed the CD, P < 0.01 (two-factor ANOVA). The genotype x diet interaction was also significant, P < 0.05 (two-factor ANOVA).
Plasma methionine concentrations in CD-treated +/– dams and FADD-treated dams (both +/+ and +/–) were lower than those in CD-treated +/+ dams (control group), although the differences due to genotype or diet were not significant (two-factor ANOVA) (Figure 2). There was no significant genotype x diet interaction.
FIGURE 2.. Mean (±SEM) plasma methionine (Met) concentrations in 80–90-d-old pregnant mice fed a control diet (CD) or a folic acid–deficient diet (FADD). Numbers above bars indicate the number of animals in each group. There were no significant differences between genotypes or diets and no significant genotype x diet interaction (P > 0.05, two-factor ANOVA).
Reproductive outcomes
No significant differences were found in mean weights of female mice before mating (data not shown). CD-treated +/– dams and FADD-treated dams (+/+ and +/–) had significantly higher resorption rates on the Kruskal-Wallis test than did the control group. The resorption rate in the control group (CD +/+ dams) was 13.37 ± 3.02%, a value that is similar to that previously reported in BALB/c mice (32). Mthfr genotype had a significant effect on embryonic resorption, with a resorption rate in +/– dams that was twice that in +/+ dams on the control diet. The low-folate diet had a particularly marked effect on resorption: 50% of embryos from FADD-treated dams (both +/+ and +/–) died in utero by GD 14.5. This dietary effect was significant in both genotype groups. Two-factor ANOVA of the ranks found significant overall effects of genotype (P < 0.05) and of diet (P < 0.05) on resorption. However, the effect of genotype in the FADD group was minimal; Mthfr +/– dams showed a rate of resorption only 15% higher than that in Mthfr +/+ dams. The effect of the Mthfr genotype may be relatively minor under circumstances of strong dietary stress.
Examination of the gross morphology of viable embryos did not find obvious anomalies in any group. However, an overall developmental delay (from a half-day up to 2 d) was observed in some embryos, which suggests IUGR. When comparing the percentage of delayed embryos between groups (Kruskal-Wallis test), we found that significantly more embryos with development delays were derived from CD-treated +/– and FADD-treated dams (+/+ and +/–) than from the control group (Table 1). Mthfr genotype had a significant effect on delay in dams on the CD, but not in dams on the FADD. This pattern is similar to that seen for the rate of resorption, which again suggests that the dietary influence may minimize or completely mitigate the effect of genotype. The FADD was associated with a proportion of delayed embryos of almost 20%, but the dietary effect was significant only in the +/+ group.
View this table:
TABLE 1. Effects of Mthfr deficiency and folate deficiency on reproductive outcomes per litter1
We also examined the overall genotype and dietary effects for the percentage of delayed embryos by using two-factor ANOVA of the ranks (Table 1). The overall dietary effect was significant, but genotype effect was not; there were no significant interactions between genotype and diet. This observation is consistent with our hypothesis that the genotype may have little or no effect in conditions of severe dietary stress.
Average embryonic crown-rump length and weight were smaller in dams of the Mthfr +/– genotype and in dams fed FADD (Table 1), findings that are consistent with the results for delay. The dietary effect was significant, but the effects of genotype on length and weight had only borderline significance (P = 0.09 and 0.06, respectively). There was no significant diet x genotype interaction for length or weight (two-factor ANOVA).
Cardiac malformations in embryos
Viable embryos were sectioned for anatomical and histologic analyses. As shown in Table 2, cardiac malformations were not detected in any of the 78 embryos from 11 litters of CD-treated +/+ dams. On the other hand, 4 of 11 litters (4/62 embryos) from CD-treated +/– dams, 5 of 10 litters (9/39 embryos) from FADD-treated +/+ dams, and 4 of 10 litters (6/37 embryos) from FADD-treated +/– dams showed heart defects. The Mthfr mutant genotype was associated with a significant increase in heart defects in dams on the CD, whereas genotype did not influence rate of CHD in mice on the FADD. There was a significant dietary effect on embryos from Mthfr +/+ dams but not on embryos from Mthfr +/– dams. These results were obtained by using the Kruskal-Wallis test.
View this table:
TABLE 2. Effects of Mthfr deficiency and folate deficiency on congenital heart defects in viable embryos1
To examine overall effects of genotype and diet, we also performed a two-factor ANOVA (Table 2). The dietary effect was significant for incidence of CHD, whereas there was no overall difference for genotype or for the diet x genotype interaction. As mentioned earlier, the genotype effect may be swamped by the severe dietary stress.
Most of the embryos with heart defects (14 of 19) had isolated ventricular septal defects (VSDs) (Figure 3 e), whereas 2 had double-outlet right ventricle (DORV) (Figures 3f and 3g), and 3 had endocardial cushion defects (ECDs) in conjunction with DORV (Figure 3h). These embryos also had variably thinner myocardia than did embryos from CD-treated +/+ dams (Figure 3h). The 3 embryos with ECD and DORV also had congestive heart failure, as shown by the dilation of vena cava and atria, and liver congestion (see Figure 4 c and d, respectively).
FIGURE 3.. Cardiac malformations in embryos on gestational day (GD) 14.5 from dams with Mthfr deficiency and folic acid deficiency (transverse sections, hematoxylin-eosin staining). AO, aorta; AR, aortic root; CD, control diet; EC, endocardial cushion; FADD, folic acid–deficient diet; IVS, interventricular septum; LA, left atrium; LV, left ventricle; MV, mitral valve; PA, pulmonary artery; RA, right atrium; RV, right ventricle; TV, tricuspid valve. (a–d) Nonconsecutive sections of an embryo from a CD-treated +/+ dam. (a) On GD 14.5, the membranous part of the interventricular septum is closed. (b) The aorta and the PA are perpendicular to each other. (c) The aorta arises from the LV. (d) Normal differentiation of atrioventricular valves (MV and TV) and normal connections of atria to ventricles. (e) Embryo from a CD-treated +/– dam with a membranous ventricular septal defect (). (f) Embryo from a FADD-treated +/+ dam in which the aorta and PA are relatively parallel to each other. (g) A more caudal section of (f) showing the aorta arising from the RV. (f) and (g) are the main features of double-outlet right ventricle (DORV). (h) Embryo from a FADD-treated +/+ dam shows a common atrioventricular valve with a ventricular septal defect, RA committing to LV, and RV hypoplasia. Arrowhead shows a cushion-like anterior leaflet of the atrioventricular valve. The compact layer of ventricular myocardium is much thinner (arrow). (Original magnification: x40.)
FIGURE 4.. Congestive heart failure in embryos with congenital heart defects (CHD) (transverse sections, hematoxylin-eosin staining). CD, control diet; DORV, double-outlet right ventricle; ECD, endocardial cushion defect; FADD, folic acid–deficient diet; LA, left atrium; LSVC, left superior vena cava; RSVC, right superior vena cava. (a) and (b) Embryo from a CD-treated +/+ dam. (c) and (d) Embryo from a FADD-treated +/+ dam, with DORV and ECD. (c) Dilation of the LSVC, RSVC, and LA, as compared with that in the normal embryo in (a). (Original magnification: x40). (d) Dilation of sinusoids in liver (arrow), as compared with the normal appearance in (b). (Original magnification: x100).
Influence of embryonic genotypes
The Mthfr genotype distribution in viable embryos from each group is shown in Table 3. The ratio of embryonic Mthfr genotypes (+/+, +/–, and –/–) in each group of dams was not significantly different from the expected Mendelian ratios. These data suggest that there is no significant loss of homozygous mutant mice during gestation. We did not observe significant changes in the incidence of embryonic delay or CHDs between embryos of different genotypes within each group of dams. However, this could be due to the small numbers of embryos in each genotype group.
View this table:
TABLE 3. Effect of embryonic Mthfr genotype on reproductive outcome and congenital heart defects1
DISCUSSION
Our results suggest that Mthfr deficiency in mice increases the incidence of pregnancy loss, fetal growth retardation, and CHDs. The conclusions from clinical studies have not been consistent with respect to the role of MTHFR in these types of complications (9-19). Some of the limitations in the clinical studies were the small sample sizes, the lack of information on dietary folate, and the mixed ethnic background of the subjects. The availability of an animal model with MTHFR deficiency has allowed us to overcome most of these potential confounders.
A common link between MTHFR deficiency and folic acid deficiency is an elevation in plasma homocysteine. The influence of MTHFR deficiency on plasma homocysteine was shown in numerous clinical studies and in our original report on Mthfr-deficient mice (2, 20). This study confirms that Mthfr-deficient mice have elevated plasma homocysteine and that dietary folate deficiency raises plasma homocysteine to an even greater extent. The genetic and dietary folate deficiencies appear to have additive effects on plasma homocysteine, and each component contributes to an increase of 15 µmol homocysteine/L. The amount of folic acid in our well-defined CD was the required amount for rodents. However, the plasma homocysteine in mice fed our CD was significantly higher than that previously reported for these mice fed laboratory chow (20), because chow contains folate concentrations that are 3-fold the recommended amount.
Our findings support an association between hyperhomocysteinemia and adverse pregnancy outcomes, as reported previously (11, 33). Whether homocysteine is an independent teratogen or a biomarker for disturbed folate metabolism or methylation remains a controversial topic. Some in vitro evidence supports the former argument—that high homocysteine alone is embryotoxic—in a dose-dependent manner (34, 35). The precise pathogenic mechanism is unknown, but hyperhomocysteinemia might lead to placental vasculopathy, because hyperhomocysteinemia is known to be toxic to the vasculature. This vasculopathy could result in placental infarcts, which would compromise the uteroplacental circulation and reduce fetal blood supply and thus lead to IUGR and spontaneous abortion (33). Examination of the placenta in hyperhomocysteinemic states might prove useful in addressing this question. An alternative mechanism could be that homocysteine results in an increase in S-adenosylhomocysteine (SAH), which can inhibit methylation reactions that are dependent on SAM. Altered DNA methylation was shown in hyperhomocysteinemia and in MTHFR deficiency (20, 24, 36). Disturbances in DNA methylation are associated with changes in gene expression, which could affect the growth and development of embryos. Folate deficiency with hyperhomocysteinemia may also result in imbalances in nucleotide pools and may affect DNA synthesis or DNA repair (3, 33).
The 3 groups of mice with hyperhomocysteinemia had significantly higher rates of adverse pregnancy outcomes than did the control group (CD-treated +/+ dams). However, we observed a resorption rate in FADD-treated +/+ dams (48.9%) that was almost twice that in CD-treated +/–-dams (27.9%), despite relatively similar homocysteine values (25.5 µmol/L and 28.9 µmol/L, respectively). Furthermore, FADD-treated Mthfr +/+ and +/– dams had similar resorption rates, whereas the +/– dams had homocysteine concentrations twice those of the +/+ dams (50.6 and 25.5 µmol/L, respectively). These observations suggest that maternal plasma homocysteine may not be the only etiologic factor. Other compounds in folate metabolism—eg, methionine or SAM concentrations, nucleotide pools, or methylation intermediates—might be involved and could provide a better correlation with outcomes. Plasma methionine concentrations were lower in Mthfr +/– dams and in FADD-treated dams than in CD-treated Mthfr +/+ dams, although the differences were not significant. Plasma homocysteine and methionine concentrations, although useful, may not accurately reflect the concentrations of these compounds in tissues. Changes in intracellular metabolites or alterations in methylation reactions could account for the delay in embryonic development as well as for the fetal loss in maternal Mthfr or folate deficiency, particularly because extreme IUGR may lead to fetal death (33).
Folate deficiency and MTHFR deficiency are well-characterized risk factors for neural tube defects, and studies of other types of birth defects are now emerging (37). The clinical literature on CHDs is consistent with a preventive role of folate supplementation (38–40), but studies of MTHFR in heart defects are sparse and inconclusive (17-19). A low-folate diet administered to mice resulted in a delay of conotruncal septation, a finding that confirmed the potential effect of folate on heart morphogenesis; however, only 4 embryonic hearts were examined in that report (22). In the current study, we examined >200 embryos and concluded that both folic acid deficiency and Mthfr deficiency independently induced heart defects. Most defects were isolated membranous VSDs, although 2 rare types of CHD—DORV and ECDs—were also found.
Heart morphogenesis is a spatially and temporally dynamic process. Multiple cell types are involved, each of which needs to undergo proliferation, apoptosis, migration, or differentiation at a critical time point. Not surprisingly, many genes and a relatively stable microenvironment are required for normal morphogenesis; significant changes in genetic or environmental modifiers can lead to cardiac malformations. It is interesting that heart defects in our study also were frequently accompanied by myocardial hypoplasia, which could result from inadequate proliferation or increased apoptosis. Previously, Rosenquist et al (34) proposed that neural crest cells might be particularly susceptible to the teratogenic effects of homocysteine. Other recent studies suggested that homocysteine can affect neural crest cell proliferation and migration (41, 42). However, the problem may not be restricted to a single cell type, because we also observed cardiomyopathy and a range of CHDs, including ECDs, that involve mesenchymal cells.
Although it is not clear whether hyperhomocysteinemia or its consequences affect heart development, we observed that the incidence of CHDs in FADD-treated +/+ dams was higher than that in CD-treated +/– dams, despite relatively similar homocysteine concentrations. This observation provides another suggestion that homocysteine may not be the only teratogen or that plasma homocysteine does not accurately reflect intracellular metabolism or toxicity.
Some embryos with severe CHDs also had heart failure, as manifested by venous congestion and liver congestion. These embryos were likely to die in utero. It is also possible that some resorbed embryos had CHD alone or with other birth defects, and that they died as a result. Because of the potential for contamination with maternal tissue, we could not examine the genotypes of resorbed embryos. However, among viable embryos, the distribution of Mthfr genotypes was close to the expected Mendelian ratio in each group, which suggested that there was no increased loss of Mthfr-deficient embryos. Moreover, there appeared to be no consistent correlation between embryonic Mthfr genotype and the incidence of developmental delay or CHDs in each group, although our numbers were too small for definitive conclusions. Our findings therefore suggest that the embryonic Mthfr–/– genotype may have less of a role than does maternal genotype in the determination of pregnancy outcomes and some birth defects. On GD 14.5, most embryonic organs are well developed, but they still rely on the maternal circulation for nutritional components, including folate. Thus, it is not surprising that the disruption of the maternal folate metabolism by genetic or nutritional means would affect embryonic folate supply. In addition, maternal hyperhomocysteinemia could, because of genetic or nutritional folate deficiency, result in embryonic hyperhomocysteinemia through the uteroplacental circulation (33).
Clinical studies have suggested that MTHFR genotype and folate status may have an interactive effect, at least with respect to plasma homocysteine (43). In the current study, we observed this interaction with respect to homocysteine but not with respect to outcomes, because we did not observe significant differences in resorptions, delays, or heart defects between FADD-treated +/– and +/+ dams. The FADD we used is already very severe and may mitigate the effects contributed by Mthfr deficiency. A diet with a smaller decrease in folate concentrations might have allowed us to observe an interactive effect.
In conclusion, the current study found that genetic and dietary disruptions in folate metabolism led to adverse pregnancy outcomes and CHDs. These findings, which were consistent with the multifactorial inheritance model for these disorders, suggested that folate supplementation for pregnant women may be important in the prevention of these common conditions, particularly in women with mild MTHFR deficiency.
ACKNOWLEDGMENTS
We thank Charles Rohlicek for his valuable input on cardiac pathology and Michel Tremblay for technical assistance.
DL, LP, and RR contributed to the design of the experiment, analysis of data, and writing of the manuscript. DL, LP, YL, QW, and JSC contributed to the collection of data. The authors had no financial or personal conflicts of interest.
REFERENCES