Roles of genetic polymorphisms in the folate pathway in childhood acute lymphoblastic leukemia evaluated by Bayesian relevance and effect size analysis

Abstract

In this study we investigated whether polymorphisms in the folate pathway influenced the risk of childhood acute lymphoblastic leukemia (ALL) or the survival rate of the patients. For this we selected and genotyped 67 SNPs in 15 genes in the folate pathway in 543 children with ALL and 529 controls. The results were evaluated by gender adjusted logistic regression and by the Bayesian network based Bayesian multilevel analysis of relevance (BN-BMLA) methods. Bayesian structure based odds ratios for the relevant variables and interactions were also calculated. Altogether 9 SNPs in 8 genes were associated with altered susceptibility to ALL. After correction for multiple testing, two associations remained significant. The genotype distribution of the MTHFD1 rs1076991 differed significantly between the ALL and control population. Analyzing the subtypes of the disease the GG genotype increased only the risk of B-cell ALL (p = 3.52×10(-4); OR = 2.00). The GG genotype of the rs3776455 SNP in the MTRR gene was associated with a significantly reduced risk to ALL (p = 1.21×10(-3); OR = 0.55), which resulted mainly from the reduced risk to B-cell and hyperdiploid-ALL. The TC genotype of the rs9909104 SNP in the SHMT1 gene was associated with a lower survival rate comparing it to the TT genotype (80.2% vs. 88.8%; p = 0.01). The BN-BMLA confirmed the main findings of the frequentist-based analysis and showed structural interactional maps and the probabilities of the different structural association types of the relevant SNPs especially in the hyperdiploid-ALL, involving additional SNPs in genes like TYMS, DHFR and GGH. We also investigated the statistical interactions and redundancies using structural model properties. These results gave further evidence that polymorphisms in the folate pathway could influence the ALL risk and the effectiveness of the therapy. It was also shown that in gene association studies the BN-BMLA could be a useful supplementary to the traditional frequentist-based statistical method.

Figure 1. Schematic overview of the intracellular folate pathway.

Key molecules and derivatives of the pathway are denoted as ovals, regulatory enzymes as rectangles, affected DNA mechanisms as double rectangles. Genetic polymorphisms are showed in rectangles only where significant results were found. dUMP = Deoxyuridine monophosphate, dTMP = Deoxythimidine monophosphate, Vit.B12 = Vitamin B12, SAM = S-adenosylmethionine, SAH = S-adenosylhomocysteine, TIMP = Thioinosine monophosphate, MetTIMP = Methyl-thioinosine monophosphate, AICART = AICAR (5-Aminoimidazole-4-carboxamide ribonucleotide) formyltransferase, GART = GAR (glycinamide ribonucleotide) formyltransferase.

Figure 2. Illustration of different dependency relations between relevant SNPs and hyperdiploid ALL susceptibility.

Top panel: An interactional map including hyperdiploid ALL susceptibility (red node) and strongly relevant SNPs with respect to it (other nodes). The width of the edges is proportional to their a posteriori probability. The directed edges represent only probabilistic relationships between the variables which are not necessary causal. Bottom panel: The posterior probability of strong relevance (blue columns), edge (direct strong relevance, red columns), pure interaction (green columns) and association (purple columns) of the variables to hyperdiploid ALL susceptibility according to the BN-BMLA method.

Figure 3. Redundancies and interactions according to the BN-BMLA method.

The figure shows the magnitude of redundancies (blue curved lines) and interactions (red curved lines) between the variables in the overall (i.e. ALL susceptibility, A panel), in the B-cell lineage (B panel), in the T-cell lineage (C panel) and in the hyperdiploid patient group (D panel) according to the BN-BMLA method. The width of the curved lines is proportional to the strength of the effect. The a posteriori probability of the strong relevance of the variables is proportional to the length of the dark red columns next to the variable in the inner gray colored ring. The corresponding genes and chromosomes of the SNPs are shown on the outer ring.