Allelic heterogeneity of and interactions between polymorphic RET enhancers affecting Hirschsprung disease risk. S. Chatterjee, A. Kapoor, A. Chakravarti Institute of Genetic Medicine, Johns Hopkins University School of Medicine, Baltimore, MD.

   Hirschsprung disease (HSCR, congenital aganglionosis) is the most common form of functional intestinal obstruction in neonates (1 in 5,000 live births). We, and others, have identified >15 genes with rare, deleterious mutations in HSCR, the most common being those at the receptor tyrosine kinase gene RET. We have previously identified a polymorphism (rs2435357 (C/T); risk allele T) within a intron 1 RET enhancer with marked effect on HSCR risk and associated with the more common male, short segment, isolated probands with HSCR. By genome-wide association studies, we have now identified a second non-coding polymorphism (P=2.7E-16; rs2506030 (A/G); risk allele G) located ~125 kb upstream of the RET transcription start site. This second variant also disrupts an enhancer since: (a) a 1kb human DNA fragment containing either the G or A allele shows significant loss-of-expression from the risk allele in Neuro2A cells; (b) transient transgenic zebrafish enhancer assays show that the 1 kb element can drive expression in vivo, overlapping ret, in the migratory neural crest cells at 48hpf and dorsal root ganglion at 3dpf in the developing embryo; (c) deletion of a 10nt fragment containing rs2506030 leads to loss of activity in in vitro and in vivo enhancer assays. Furthermore, enhancer assays demonstrate synergistic effects in Neuro2A cells when the 1 kb element contains both rs2435357 and rs2506030. For establishing genetic-phenotypic interactions, we studied the joint genotypes of 337 HSCR patients and 379 controls of European ancestry (1000 Genomes). We find the expected enrichment of both risk alleles in cases, as expected (rs2435357 T: 57%/27% in cases/controls, P=5.46E-31; rs2506030 G: 56%/40% in cases/controls, P=1.24E-9) but a very wide variation in risk by genotype (odds ratio: 0.31 - 9.24, P<0.0015). Moreover, the haplotype distribution of both risk alleles shows greater association in cases (r2=0.25, P=9.8E-18) than in controls (r2=0.10, P=9.2E-8), a difference that is statistically significant (P=0.0023) and signifies interaction of their effects. Analysis of the joint effects at both loci suggests that the primary RET susceptibility to HSCR is recessive and from the enhancer at rs2435357 with the new enhancer at rs2506030 modulating that effect. In summary, this is one of the first descriptions of the molecular basis and genetic effects of allelic heterogeneity and interactions of enhancer polymorphisms in human disease.

You may contact the first author (during and after the meeting) at