Results
Of the 523 WNT signaling genes, 173 contained qualifying RDVs (GATK PASS, AF < 1%, CADD phred ≥ 20, coverage ≥ 90%) in the MA myelomeningocele population (S2 Text). When comparing RDV mutational burden in the 173 genes between the two Hispanic populations, ten genes associated with a risk for myelomeningocele by yielding Fisher’s exact test P values below 0.05 and odds ratios above 1 (Table 1). This included PORCN , CDH2 , PRICKLE2 , CPE ,FUZ , PTPRU , PSMD3 , TNRC6B , PPP2R1A , and FERMT2 . Of these, PORCN , CPE , and TNRC6Bwere detected caudally in the closing neural tube in humans (Krupp et al., 2012). PORCN , CDH2 , and FUZ have been previously associated with NTD phenotypes in mouse models (Gray et al., 2009; Heydeck, Zeng, & Liu, 2009; W. Liu et al., 2012; Radice et al., 1997). This analysis gave a Bonferroni correction value of 2.9 x 10-4, which none of the genes’ P values fell below.
[t][/t] table 1 near here
In the EA myelomeningocele population, 189 genes contained qualifying variants (S3 Text). After comparing those 189 genes between the two European populations, seven genes were associated with a risk for myelomeningocele by giving P values below 0.05 and an odds ratio above 1 (Table 1). These genes included DDB1 , SDC1 ,CSNK1G2 , SOSTDC1 , PLCB1 , DVL2 , andTLE3 . Of these, DDB1 , PLCB1 , and DVL2 were detected in the human neural tube during closure and DVL2 is also associated with NTDs in mouse models (Hamblet et al., 2002; Krupp et al., 2012). This analysis gave a Bonferroni correction value of 2.6 x 10-4, which none of the genes’ P values fell below.
Subjects from both ethnicities tended to have a similar number of genes that contained qualifying variants, but MA subjects tended to have more nominally significant disrupted genes (Fig 2). Nominally significant disrupted genes were identified in 46 (18%) EA and 66 (26%) MA myelomeningocele subjects. Also, the disrupted genes that associated with EA myelomeningocele subjects are different from disrupted genes associated with MA myelomeningocele subjects, i.e. the nominally significant genes from each comparison did not overlap.
Gene Ontology analysis regarding biological processes of the disrupted genes in both populations reveals that both groups have disrupted genes enriched for regulation of WNT signaling and negative regulation of canonical WNT signaling. However, disrupted genes in only the MA subjects were enriched for functions involving PCP and SCF-KIT signaling whereas disrupted genes in only the EA subjects were enriched for genes associated with myoblast differentiation, inositol phosphate metabolism, disassembly of the destruction complex, and recruitment of AXIN to the membrane (Table 2).
[t][/t] table 2 near here