Two SNPs rs1801133 and rs1801394 in folate pathway are associated with the risk of nonsyndromic cleft lip with or without cleft palate CURRENT STATUS:

Background Prenatal intake of folic acid is important for prevention of nonsyndromic cleft lip with or without cleft palate (NSCL/P). Associated genes in folate pathway are major enzymes of folic acid metabolism that is crucial for preventing birth defects. The present study aims to investigate the association between four single nucleotide polymorphisms (SNPs) in folate pathway genes and the risk of NSCL/P. Methods Comprehensive bioinformatics analysis was used to predict the functional pathogenicity of genetic variation. The PubMed, Embase database and Google Scholar were intensively searched by two researchers to ascertain all relevant studies. Stata 11.0 software was used to analyze the results. Subgroup analysis was carried out to assess the influence of genetic background. Sensitivity analysis, regression analysis and publication analysis were also conducted to enhance the strength of our results. Results It is estimated that the probability of two missense mutation rs1801133 in MTHFR and rs1801394 in MTRR are more likely to be damaging by bioinformatics analysis. A total of 34 publications were included in the present study analysis. Our results showed a significant association between rs1801133 and risk of NSCL/P in two genetic models: TT allele vs CC allele (OR=1.333 95%CI=1.062-1.674, P =0.013), and recessive model (OR=1.325 95%CI=1.075-1.634, P =0.008). A significant association between rs1801394 and the risk of NSCL/P in Asian (GG allele vs AA allele, OR=0.520 95%CI=0.321-0.841, P =0.008) was also observed. Meta-regression, sensitivity analysis, and publication bias analysis confirmed that the results of the present study were statistically significant. Conclusions The present study highlights two SNPs rs1801133 in MTHFR and rs1801394 in MTRR, associated with the increasing risk of NSCL/P. Further, larger studies should be performed to confirm these findings.

Here, therefore, we conduct a systematic review and performed an updated meta-analysis of all related studies published before April 2019, to provide more precise statistical results. Our study could provide basic data for exploring effective therapeutic strategies for NSCL/P.

Literature search
The present meta-analysis was conducted according to the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analysis 2009 (PRISMA2009).

Inclusion criteria
All studies were intensively reviewed by two researchers. Studies were included in the meta-analysis if they met the following criteria: (1) original observational study of human participants; (2) an association study between rs1801133 and/or rs1801194 and/or rs1801398 and/or rs3733890 and NSCL/P; (3) case-control study or cohort study; (4) allele data were available; (5) largest sample size or sufficient useful data were included in duplicate publications from the same population; and (6) studies were published in English.

Quality score assessment
Study quality was assessed to guarantee the strength of results and conclusions. The Newcastle-Ottawa scale (NOS) score was calculated to assess the quality of studies. A maximum of nine scores, including selection, comparability and exposure items, could be awarded [37]. Any variances in comparison were decided by a third researcher.

Data extraction
The data were extracted independently by two researchers from all included studies using an integrated and standardized form. The following information was extracted: (1) first author name; (2) publication year; (3) population ethnicity; and (4) genotype distribution.

Computational and Statistical analysis
Polyphen2, SIFT, CADD, phyloP, and LRT were used for bioinformatics prediction. The same genetic models were performed for "rs1801394", "rs1801198", and "rs3733890". A P value of P<0.05 was established as the significant threshold for each genetic model. Two subgroups, including Caucasian and Asian, based on ethnicity were analyzed to reduce the heterogeneity and influences from the genetic background. Meta-regression and one-way sensitivity analysis were also included in the meta-analysis. Egger's regression test was used to evaluate the publication bias with quantitative analysis [38]. The trim and fill method was also used when publication bias exists.

Study characteristics
Based on the search strategy, 926 publications were identified in the initial search. After evaluating the titles and abstracts, 801 publications were excluded, and 125 full-text publications were further reviewed. By applying the inclusion criteria, 34 publications were used for the final meta-analysis ( Figure. 2). A total of 30 publications with 5517 cases and 7770 controls were included in the rs1801133 group; ten publications with 1767 cases and 2029 controls were included in the rs1801394 group; six publications with 1815 cases and 898 controls were included in the rs1801198 and five studies with 1253 cases and 1562 controls were included in the rs3733890 group. A total of seven studies in the control group (not excluded) were found to deviate from HWE. A total NOS assessment score of three or lower, four to six, and seven or greater were defined as low, medium and high quality, respectively. The main characteristics of the included publications are shown in Table 2.

Associations between the four SNPs of folate pathway gene and NSCL/P in the overall population
The meta-analysis results showed that there was a significant association between rs1801133 and NSCL/P risk in two genetic models: TT allele vs CC allele (OR=1.333 95%CI=1.062-1.674, P=0.013) and recessive model (OR=1.325 95%CI=1.075-1.634, P=0.008) ( Table 3, Figures 3-4). There was no statistically significant association between rs1801394 of the MTRR, rs1801198 of the TCN2, rs3733890 of the BHMT and NSCL/P risk in the overall population (Tables 4-6).

Subgroup analysis by ethnicity stratification
To assess the role of genetic background in the meta-analysis, we carried out a subgroup analysis, where the overall population was divided into two subgroups, (i) Asian and (ii) Caucasian. The subgroup analysis showed that there was significant association between rs1801394 and NSCL/P risk in Asian (GG allele vs AA allele, OR=0.520 95%CI=0.321-0.841, P=0.008), but no associations in Caucasian (Table 4).

Meta-regression and influence analysis
The following covariates were considered for meta-regression: publication year, sample size and HWE in controls. The results showed that there was no influence for the four SNPs (P>0.05). To identify the degree to which a single study affected the overall OR estimates, one-way sensitivity analysis was performed by repeating the meta-analysis, sequentially excluding one study at a time. The results showed that no study was found to exert an excessive influence on the pooled effect.

Publication bias
There was publication bias for rs1801133 in the Asian population in genotype model CT vs CC (  [10]. Similarly, the present study didn't find the significant association between the C776G and NSCL/P.
BHMT, a zinc dependent cytosolic enzyme, is important for homocysteine metabolism and methionine synthesis. In 2010, Mostowska et al. first found that rs3733890 of the BHMT was associated with NSCL/P, and other studies also indicated its association with coronary artery disease and neural tube defects [23]. In the present study, we found no evidence showing rs3733890 playing any significant role [23]. We inferred several factors may contribute to the result. First, we found a relative high value of heterogeneity among studies, which cause a different distribution of genotype. Second, the number of included studies and sample size are relatively small. So the subgroup analysis was not conducted based on ethnicity.
MTRR plays a vital role in functional regeneration of methionine synthase, and it may be associated with increasing the congenital heart disease risk [18]. But the meta-analysis showed that the change of genetic variant will influence the protein function and predispose to cause the disease (Table 1). However, the association between rs1801133 and the risk of NSCL/P is inconsistent. The allelic frequencies vary in different ethnic groups and the minor allele frequency (MAF) of MTHFR rs1801133 in Asian are lower than that in European and American, so it is very valuable to summarize and analyze by systematic statistical methods. In the present study, we included 30 studies including 5517 cases and 7770 controls and found T allele can increase the risk of NSCL/P.
The strength of this meta-analysis is that it expands to a large number of related studies, and the most updated publications were included. To guarantee the quality of the included studies, a strict procedure for search strategy, literature inclusion, data extraction, and quality assessment was performed, which could avoid potential influences and increase the strength of the results. Meta-regression and sensitivity analysis were also performed to strengthen the conclusions. We confirmed the previous investigation by summarizing a larger number of closely related studies.
There are some limitations in the present meta-analysis. Firstly, studies published only in English were included in the meta-analysis, and studies published in other languages were excluded. Secondly, environmental factors also contribute to NSCL/P, and in the present study, non-genetic factors and other potential interactions, including environment-gene interactions, were limited. Additionally, some potential covariates (e.g., age, sex, folate etc) were not included in the analysis due to insufficient information.

Conclusion
In this study, we successfully identified two SNPs rs1801133 in MTHFR and rs1801394 in          Forest plot for pooled ORs for the associations between TT vs CC model of rs1801133 and NSCL/P risk. Each square is proportional to the study-specific weight.

Figure 4
Forest plot for pooled ORs for the associations between recessive model of rs1801133 and NSCL/P risk. Each square is proportional to the study-specific weight.