CHKA and PCYT1Agene polymorphisms, choline intake and spina bifida risk in a California population

Neural tube defects (NTDs) are complex congenital malformations of the central nervous system. Anencephaly and spina bifida are the most common and severe forms of NTDs. The birth prevalence of NTDs varies from approximately 0.8/1,000 births in many areas of the US to 3.5/1,000 in Mexico [1, 2]. Epidemiologic studies suggest that both genetic and environmental factors contribute to NTD etiologies. Although most factors appear to explain very little of the population burden of NTDs, maternal nutritional factors do appear to substantially contribute to the complex etiologies of NTDs. Foremost among these factors has been the role of periconceptional intake of folic acid in reducing recurrence and occurrence risks of women for NTD-affected pregnancies [3–11].

Nutrients and nutrition-related factors other than folic acid have been observed to influence NTD risks. For example, increased intakes of methionine, zinc, vitamin C, and choline have been associated with reduced NTD risk [12–15]. With respect to choline, it was recently observed that increased periconceptional intakes of diets with choline were associated with reduced risks of NTD-affected pregnancies that were independent of maternal folate intakes.[15] This observation provided evidence to suggest that deficiencies in methyl donors may be associated with NTD risk, that is to say, a less than optimal methyl-donor supply and DNA methylation status has been a suggested area for research efforts for certain birth defects [16]. Choline, like folate, is a methyl donor in the methylation of homocysteine to methionine [17, 18].

Choline is utilized for the de novo synthesis of phosphatidylcholine (PC) and sphingomyelin through the cytidine diphosphocholine (CDP-choline) pathway. There are three reactions in this pathway. The first reaction is catalyzed by the enzyme choline kinase (CHK; ATP:choline phosphotransferase, EC 2.7.1.32), which phosphorylates choline by donating an ATP [19]. The second reaction involves phosphocholine (P-Cho) cytidylyl transferase (CCT), which catalyzes the formation of CDP-Choline from P-Cho and CTP [20]. The final reaction uses choline phosphotransferase (CPT), which catalyzes the condensation reaction of CDP-Choline with diacylglycerol [21]. Phosphatidylcholine (PC) and sphingomyelin are required for maintaining cell membranes and play important roles in regulation of cell growth, differentiation, and death through the production of diacylglycerol (DAG) and ceramide (CER), which are cell signaling molecules [22, 23].

In gastrulation- and neurulation-stage mouse embryos, choline was elucidated to be used primarily for PC synthesis favoring the CDP-choline pathway, although some betaine and acetylcholine was also generated [24]. Using the choline uptake inhibitor 2-dimethlyaminoethanol (DMAE) and an inhibitor of PC synthesis, 1-O-octadecyl-2-O-methly-rac-glycerol-3-phosphocholine (ET-18-OCH₃), Fisher and co-workers observed an increase in cell death and both craniofacial and NTDs in neurulation stage mouse embryos grown in culture [24].

In humans, choline kinase has two isoforms, CHKα and CHKβ, with the α as a dominant isoform. The CHKA gene encoding choline kinase α is located at chromosome 11q13.2. Our study focused on the CHKA gene. PCYT1A and PCYT1B encode CCTα and CCTβ, respectively. PCYT1A located at chromosome 3q29, while PCYT1B is located at chromosome Xp22.11 [25, 26]. In this study we focused on PCYT1A gene.

Given that periconceptional intake of choline has been associated with decreased risk of NTD-affected pregnancies [15], we investigated CHKA and PCYT1A genotypes on risk of spina bifida. We also investigated these genotypes in combination with lowered maternal intake of choline as risk factors of spina bifida.

Genotyping results of all SNPs were in Hardy-Weinberg Equilibrium (HWE) among controls (χ² test: P > 0.05). Among non-Hispanic white and Hispanic white, the minor allele frequencies (MAF) were 0.39 and 0.27 for hCV1562388, 0.20 and 0.12 for hCV1562393, 0.32 and 0.33 for rs939883, 0.38 and 0.47 for rs3772109, respectively. Linkage disequilibrium (LD) were evaluated by D' and r² using the Haploview program. For CHKA gene, hCV1562388 and hCV1562393 are in complete LD in study population (D' = 0.91, r² = 0.065). For PCYT1A gene, D' for rs939883 and rs3772109 was 0.81 and r² was 0.29.

This study investigated an underlying genetic explanation for a previously identified association between choline intake and spina bifida risk [15]. In the current study, we investigated intronic gene variants of two enzymes involved in the metabolism of dietary choline via the CDP-choline pathway. We believe this is the first study to evaluate DNA sequence variants in the human CHKA and PCYT1A genes for a possible association with NTD risk. Reduced risks of spina bifida were found for CHKA SNP hCV1562388, and increased risks were found for SNP rs939883. These risks, however, were not modified by maternal periconceptional intake levels of dietary choline. Thus, our study showed gene-only effects but did not observe gene-nutrient interaction effects associated with choline intake. The results indicate that dietary choline and choline metabolism genes may affect spina bifida risk independently or through some other unknown mechanisms. This interpretation should be taken cautiously owing to limited statistical power; if gene-only effects are true, a lack of gene-nutrient interaction effects may be due to small sample sizes and limited statistical power.

The functional impacts of the CHKA SNP hCV1562388 (A>C), CHKA SNP hCV1562393(C>G), PCYT1A SNP rs939883 and SNP rs3772109, or other sequence variations associated with these tagging SNPs are unknown with respect to choline, PC, or homocysteine concentrations. Unlike the enzymes cystathionine beta-synthase (CBS), methionine synthase reductase (MTR), and 5,10-methylenetetrahydrofolate reductase (MTHFR), that are directly involved in folate-homocysteine metabolism, the CHK metabolic profile may be modulated, as its effect is transmitted through the choline pathway [24].

This study had limitations that lessen our ability to make solid inferences from the results. As noted, the study had limited sample size; therefore, precision was low for many of the estimated effects, particularly those involving gene-nutrient interactions. Our study, explored only four tagging SNPs in two genes, limiting the ability to detect possible genetic modifiers related to choline metabolism.

Despite its limitations, this study provided initial data indicating a potential association between CHKA and PCYT1A gene variants and spina bifida risk. Future studies of additional SNPs within the CHKA and PCYT1A genes should be investigated as potential predictors of spina bifida risks.