Neural Tube Defects (NTDs) are clinically significant congenital malformations affecting the development of the central nervous system. The prevalence of these defects is approximately one in every 2,000 live births per year in the U.S. and 0.5 to 2 per 10,000 live births worldwide [1]. Neural tube closure (NTC) is the morphological process occurring in early embryogenesis that ultimately creates the adult central nervous system including the brain and spinal cord. Initiation of NTC begins with the elevation of neural folds from the neural plate which ultimately fuse at the dorsal midline along the rostral-caudal embryonic axis [1a]. NTDs arise secondary to a failure of neurulation and are phenotypically classified as anencephaly, myelomeningocele (spina bifida) and craniorachischisis [1b]. Fetuses with anencephaly and craniorachischisis die prior to, or shortly after birth [2].

Low maternal folate plasma concentrations have been identified as an import ant risk factor for NTDs. Studies dating back over 30 years show that a significant portion of NTDs can be prevented by maternal folic acid supplementation [2,3,4]. Wald (2001) reported that folic acid (5 mg per day) administered prior to, and at the early stages of pregnancy, can decrease the risk of NTDs by about 85% in women that have average serum folate levels [5]. Our laboratory demonstrated in a murine model in which the primary folate receptor gene (Folr1) which is crucial for folate transport during central nervous system development was inactivated, hence compromising maternal-to-fetal folate transport during development resulted in nullizygous embryos with NTDs. Furthermore, supplementation with folic acid restored the normal phenotype in a significant percentage of the nullizygous Folr1 embryos [6]. Thus, maintaining sufficient maternal folate stores is essential for normal NTC and prevention of NTDs.

Based on strong human epidemiological data and a significant experimental mouse literature as well, it is now clear that genes involved in the metabolism and transport of folate represent excellent candidate genes for evaluating specific contributions to the genetic risk of NTDs. Two genes that code for enzymes that are involved in cytoplasmic metabolism of folate, Methylenetetrahydrofolate Reductase (MTHFR; MIM# 607093) and C₁-Tetrahydrofolate Synthase (MTHFD1; MIM# 172460), have previously been shown to be associated with an increased NTD risk [7,8,9]. Mutations in genes encoding the glycine cleavage system predispose to neural tube defects, as observed in Amt$^{-/-}$ mice and supported by AMT mutations identified in human NTD patients [10]. In our previous study, we found that Mthfd1l$^{-/-}$ mice presented 100% NTDs [11].

A study performed in an Irish NTD population found that a gene encoding the mitochondrially localized C₁-Tetrahydrofol-ate Synthase (MTHFD1L) enzyme influences NTD disease risk [12]. A polymorphism of the MTHFD1L gene, rs3832406, was found to affect the alternative splicing efficiency of the MTHFD1L mRNA transcripts [12]. Three alleles of rs3832406 are recognized that consist of a varying number of short tandem repeats of the sequence ATT. In the Irish population, the presence of allele 1 (ATT$_{7})$ in children showed an increased NTD risk, allele 2 (ATT$_{8})$ was associated with a decreased NTD risk, and allele 3 (ATT$_{9})$ did not influence NTD risk [12]. In this study, we investigate MTHFD1L as a candidate gene for NTD risk association in a US NTD population consisting primarily of Hispanic and non-Hispanic Caucasians. Unlike the Irish population study, no association was found between MTHFD1L alleles in children with spina bifida. However, when examining the mothers' genotype, alleles 1 and 3 were found to have a significant association with spina bifida risk.

During the last step of the mitochondrial 1C pathway, MTHF-D1L is responsible for encoding the monofunctional 10-formyl-THF synthetase to produce formate. Most cellular formate originates from the mitochondria and is transported into the cytoplasm. Disruption of mitochondrial formate production will affect the cytoplasmic formate supply [13]. The association between maternal folate status and risk for NTDs has led to the development of a number of folate-pathway knockout mice that have been used to study the impact of folate on mammalian embryonic development [14]. A knockout mouse model confirms that MTHFD1L plays a critical role in NTC, and deletion of Mthfd1l in mice produced 100% penetrant NTDs in nullizygous embryos [11]. The strong linkage between MTHFD1L and NTD prevalence prompted us to study variants in this gene in a human NTD cohort.

A previous study has shown that the Mthfd1l rs3832406 polymorphism was a risk factor for NTDs in the Irish population studied, and the presence of an additional ATT repeat in allele 2 affected alternative splicing efficiency [12]. In the Irish population, allele 1 increased infant NTD risk, while allele 2 decreased infant NTD risk, and allele 3 had no associated risks. In the current study, no association between infant NTD risk and MTHFD1L rs3832406 alleles 1, 2 and 3 was observed. Interestingly, we identified a maternal association between rs3832406 and NTD risks in the USA study population. Our data shows that mothers with allele 3 of rs3832406 had significantly increased maternal risk of having an NTD baby, while mothers with allele 1 of rs3832406 had significantly decreased risk of offspring with an NTD. Embryos rely on nutrition from their mothers during early development, so it is likely that the defective folate metabolism of a mother could increase the risk of having an NTD infant. Defective MTHFD1L function may impact maternal formate supply to the embryo. Supplementation of Mthfd1l heterozygous dams with formate rescues the growth and NTD phenotype of Mthfd1l nullizygous embryos, supporting the importance of formate transfer from mother to embryo [11].

The different association results between the Irish population and the American population on rs3832406 and NTD risk could be due to several different reasons. Clearly, the two populations have different genetic backgrounds, therefore it is highly likely that there are different modifying genes and different single nucleotide variant (SNV)-SNV interactions which may contribute to the differences in the US and Irish populations with respect to NTD risks. Apart from the genetic background, the different environmental factors encountered by these populations should be considered as a possible explanation for the different effect of the alleles. For example, the diets of the two populations vary, which can impact how much a polymorphism will affect the expression of various genes and ultimately the phenotype. The U.S. also has a mandatory program in place for food fortification with folic acid while Ireland has not implemented a mandatory folate fortification program. These key nutritional differences could be a potential cause of the discrepancies in the results obtained with the Irish and American populations.

This research was supported in part by NIH research grants R01 HD 083809 (RHF and DA), PO1 HD067244 (RHF), and March of Dimes grant #6-FY16-169 (RHF and DA). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Maternal and Child Health, or the National Institutes of Health.

All the authors declare no conflicts of interest.

Paige McKenzie performed the experiments and conceptualized the study with Drs. Lei and Finnell and assisted in writing the manuscript; Yunping Lei conceptualized the study and performed the experiments with Ms. McKenzie and assisted in writing the manuscript; Dean Appling provided scientific input into the study and assisted in writing the manuscript; Jessica Momb provided scientific input into the study and assisted in writing the manuscript; Richard Finnell conceptualized the study, wrote the manuscript and obtained the samples and funding for the study.