Irf6 homeostasis is required for neurulation through a direct interaction with Tfap2a. Y. A. Kousa1, H. Zhu2, A. Kinoshita3, W. D. Fakhouri4, M. Dunnwald5, R. R. Roushangar1, T. J. Williams6, B. A. Amendt7, Y. Chai8, R. H. Finnell2, B. C. Schutte9 1) Biochemistry and Molecular Biology, Michigan State University, East Lansing, MI; 2) Dell Pediatric Research Institute, Department of Nutritional Sciences, University of Texas at Austin, 78723 Austin, Texas, USA; 3) Department of Human Genetics, Nagasaki University, Nagasaki, Japan; 4) Department of Diagnostic & Biomedical Sciences, School of Dentistry, University of Texas at Houston, 77054 Houston, Texas, USA; 5) Department of Pediatrics and Interdisciplinary Program in Genetics and Molecular and Cellular Biology, University of Iowa, 52242 Iowa City, Iowa, USA; 6) Department of Craniofacial Biology, University of Colorado Denver at Anschutz Medical Campus, 80045 Aurora, Colorado, USA; 7) Department of Anatomy and Cell Biology, University of Iowa, 52242 Iowa City, Iowa, USA; 8) 8Center for Craniofacial Molecular Biology, Ostrow School of Dentistry, University of Southern California, 90033, Los Angeles, California, USA Irvine, California, USA; 9) Department of Microbiology and Molecular Genetics, Michigan State University, 48824 East Lansing, Michigan, USA.

   Mutations in Interferon Regulatory Factor 6 (IRF6) cause Van der Woude syndrome (VWS), an autosomal dominant form of cleft lip and palate. A DNA variant (rs642691) in the IRF6 enhancer MCS9.7 is found in 30% of the worlds population and contributes 12% of the risk for nonsyndromic orofacial clefting. In vitro, rs642691 abrogates one of four TFAP2 binding sites within MCS9.7 and increases risk for orofacial clefting. Mutations in TFAP2A cause branchio-oculo-facial syndrome, which typically includes an orofacial cleft. In the mouse, Tfap2a is also required for neurulation. In this work, we dissected the nature of Irf6-Tfap2a interaction. We show that AP-2a is necessary for complete MCS9.7 enhancer activity in the mouse, but that AP-2a was ectopically expressed in Irf6-/- embryos. To determine the mechanism of this interaction, we altered the dosage of these two genes in mutant mice. We showed that 10% of Tfap2a+/- embryos have exencephaly. However, consistent with a negative feedback loop, the Irf6 null allele completely rescues Tfap2a+/- embryos. Furthermore, we observed that 19% of transgenic embryos over-expressing Irf6 have neural tube defects (NTD) that phenocopy Tfap2a heterozygous and knockout embryos. At the molecular level, we found that Irf6 and AP-2a co-localized in non-neural ectoderm of wild type embryos. However, in embryos that over-express Irf6, there was a loss of AP-2a expression in early delaminating neural crest cells. These data suggest that Tfap2a and Irf6 interact through a negative feedback loop. In addition, we found that compound heterozygous embryos for the Irf6 null and hypomorphic alleles have a curly tail, a NTD considered to be analogous to spina bifida in humans. Thus, both over-expression and under-expression of Irf6 can lead to NTD in the mouse. Finally, to assess the role of IRF6 in human neurulation, we sequenced 192 spina bifida patients. We found a nonsynonymous substitution at a highly conserved amino acid in exon 9 of IRF6 (D427Y) that is predicted to be damaging by PolyPhen/SIFT and is not found in control databases. Interestingly, D427Y was previously identified in a patient with VWS. While IRF6 haploinsufficiency has not been associated with spina bifida, our murine studies support the possibility that IRF6 mutations can lead to human NTDs. These results establish a negative feedback loop between Irf6 and Tfap2a and identify a role for Irf6 in neurulation.

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