1 IRF6 loss‐of‐function causes defects in enamel formation and root patterning Emily Y. Chu, H.Fong, Matthew R. LaCourse, Brian L. Foster, Martha J. Somerman, Timothy C. Cox INTRODUCTION Mutations in interferon regulatory factor‐6 (IRF6) are associated with syndromic and nonsyndromic cleft lip with/without clefting of the palate (CLP), notably Van der Woude and Popliteal Pterygium Syndromes. CLP patients with epithelial abnormalities (e.g. wound healing delays and ectopic adhesions) often feature dental aberrations (Advani et al., 2012; Brookes and Canady, 2007; Kondo et al., 2002; Zucchero et al., 2004). However, IRF6 function remains elusive. We have found that IRF6 physically interacts with the hexameric NME (non‐metastatic) complex, consisting of NME1 and NME2. Although IRF6:NME has not been implicated in CLP, NME participates in establishment of epithelial apical and basolateral domains, which affects cell shape, adhesion, and behavior (Woolworth et al., 2009). NME1 interacts with eth TIAM1 (T‐cell lymphoma invasion and metastasis 1) protein that regulates the PAR epithelial polarity complex, known to regulate adherens junctions (AJs) formation (Tanaka et al., 2012). AJs contain nectins and cadherins, and maintain epithelial polarity by delineating apical and basolateral domains (Miyoshi and Takai, 2008). The surface expression of E‐ cadherin is dynamically regulated to influence adhesion strength and ultimately epithelial polarity. Compared to the general population, CLP patients exhibit higher rates of dental anomalies, including hypodontia and enamel defects, reasons still remaining unknown. Specifically, hypodontia and taurodontism are associated with VWS and isolated forms of CLP. Thus, similar mechanisms may control aspects of tooth development and proper palatal fusion. The tooth serves as an isolated system to examine epithelial‐derived tissues. When Nectin1, Rac1 and p‐120 catenin, molecules critical toward the establishment of epithelial adhesion and polarity, are ablated in teeth, mice exhibit severe enamel structure defects, including disorganized and hypomineralized enamel (Barron et al., 2008; Bartlett et al., 2010; Huang et al., 2011). Ameloblasts, which form enamel and are derived from oral epithelium, exhibit easily visualized polarity. Adherens junctions are also found in cell‐cell contacts between secretory ameloblasts, and disturbances in ameloblast adhesion have been reported in Nectin1, Rac1, and p‐120 catenin mutants (Nishikawa et al., 1990). Thus, ameloblasts offer an alternative model to further validate the impact of IRF6 on epithelial cell adhesion and polarity. Irf6 mutant strains die before completion of tooth development (Ingraham et al., 2006), thus an Irf6 conditional knockout model (cKO) driven by the paired‐like homeodomain2 (Pitx2) promoter was generated to investigate the effects of Irf6 on ameloblast polarity and enamel development (Liu et al., 2002). Pitx2, specific to the oral epithelium is expressed in the dental lamina, enamel knot, undifferentiated cervical loops, and pre‐ameloblasts (Li et al., 2014). During the progression from oral epithelium to differentiated ameloblasts, the epithelial derived tissues undergo morphological changes, likely influenced by epithelial polarity and adhesion. Irf6 is expressed in the oral epithelium; thus, our murine model is expected to elucidate the role of Irf6 in epithelial‐derived tooth structures. Tooth sections from stages of tooth development (bud stage, cap stage, and bell stage) were immunolabeled using primary anti‐IRF6 and NME antibodies. IRF6 and NME were detected in developing ameloblasts, specifically adjacent to the developing enamel matrix, suggesting that IRF6 and NME are involved in amelogenesis (Fig. 1). Our studies here were designed to further elucidate the role of IRF6 on CLP by investigating molecular contributions of IRF6 on epithelial adhesion and polarity as well as developing an Irf6 animal model that survives past birth. We hypothesize that IRF6, via interaction with the NME complex, contributes to the establishment dan maintenance of epithelial polarity in developing facial primordia and tooth buds. Ultimately, these studies are expected to provide novel insight into the molecular and cellular consequences of IRF6 mutations that underlie orofacial clefting disorders. MATERIALS AND METHODS Twelve specific IRF6 missense mutations found in familial cases of VWS/PPS were compared to wild type (WT) IRF6 in ability to interact with NME1/2. Wild type IRF6:NME1/2 interaction was used as a positive control. Using site directed mutagenesis, DNA constructs containing missense mutations were generated. These constructs (featuring WT, mutated, and/or hybrid sequences) were cloned into yeast vectors for the yeast two hybrid assay or into a vector containing a myc tag for co‐immunoprecipitation (Co‐IP) studies. In the yeast two hybrid, co‐transformants were assessed for activity of reporter genes, HIS3 and lacZ. Results were validated using Co‐IP followed by SDS‐PAGE and Western blotting. Rac1 and RhoA activity were measured to assess the contribution of IRF6:NME complex to maintenance of epithelial adhesion and polarity. HEK293T cells were transfected with either WT and mutant IRF6 and NME constructs. After twenty four hours, cells were serum starved for sixteen hours to reduce GTPase levels to a basal state, therefore 2 enhancing detection of effects on their activity. Following serum starvation, cell lysate was harvested, and all Rac1/RhoA activity measurements were performed using standard glutathione S‐transferase (GST) pull down kits followed by SDS‐ PAGE and Western blotting using anti‐Rac1/anti‐RhoA antibodies. A conditional knockout mouse driven by the Pitx2 promoter was generated. Mice possessing Irf6 exons 3 and 4 flanked by LoxP sites were crossed with mice possessing the Pitx2 promoter linked to Cre recombinase. Floxed Irf6 mice were kindly provided by Dr. Brian Schutte (Michigan State University), and Pitx2‐Cre mice were kindly provided by Dr. James Martin (Baylor College of Medicine). Mice were genotyped, and mandibles were collected at 4‐84 days postnatal (P4 to P84). Controls had either Pitx2+/+; Irf6fl/+, , Pitx2Cre/+; Irf6fl/+, or a Pitx2+/+; Irf6fl/fl genotype, and Irf6‐cKO refers to a Pitx2Cre/+; Irf6fl/fl genotype. Control and Irf6‐cKO samples were analyzed by radiography, microCT, scanning electron microscopy (SEM), histology, and immunohistochemistry for amelogenin (AMELX), IRF6, and NME. RESULTS IRF6 and NME point mutations disrupt IRF6:NME complex In the yeast two‐hybrid assay, co‐transformants with IRF6 and NME plasmids (WT, mutated, and/or hybrid sequences) were assessed for HIS3 and lacZ activity. Several missense mutations resulted in disruptions to the IRF6:NME interaction as demonstrated on SD‐Leu‐Trp‐His+3AT plates and X‐Gal Assay (Fig. 2 and Table 1). Similar results were obtained when cells were co‐transformed with NME1 (data not shown). Co‐immunoprecipitation studies further validated the IRF6:NME interaction and demonstrated that certain point mutations identified in VWS patients disrupted the IRF6:NME interaction. Selected mutations caused Rac1 and RhoA up‐regulation (Fig. 2). Loss of Irf6 causes defects in tooth patterning and initiation Irf6‐cKO mice survived past birth and were indistinguishable from control littermates in regards physical and behavioral appearance. On closer examination, Irf6‐cKO mice exhibited defects in tooth patterning and initiation, resulting in an array of crown abnormalities (Fig. 3). Overall, molars were smaller in the Irf6‐cKO samples (microdontia) compared to controls. In 25% of samples examined (5 out of 20), abnormal cusp patterns were noted including significantly smaller crown shapes with single peg‐shaped roots (Fig. 3). In 10% of samples (2 out of 20), 3rd molars were missing. Alterations in Irf6‐cKO molar cusp shape were observed in P7 and P14 samples, ages prior to tooth eruption (Fig. 4). Loss of Irf6 causes alterations in root formation Irf6‐cKO mice presented root patterning defects, including severely taurodontic mandibular second molars in 100% of Irf6‐cKO samples (N=20) examined (Fig. 3). Teeth were characterized as taurodonts with the following criteria: enlarged pulp chamber, apical displacement of the pulpal floor, reduction of separation between or fusion of mesial and distal roots, and reduced constriction at the cementoenamel junction. Enamel maturation defect observed in Irf6‐cKO samples Irf6‐cKO mice displayed rapid enamel attrition with wear reaching dentin 7 days post‐eruption (Fig. 3). Micro‐ct analysis revealed that Irf6‐cKO enamel was hypomineralized compared to controls in incisors and molars. Cross sections of the incisors were examined below four landmarks: alveolar crest, mental foramen, mesial root of the mandibular first molar, and mesial root of the mandibular second molar. Incisor enamel at the alveolar crest and the mental foramen landmarks was comparable between Irf6‐cKO and control samples, whereas micro‐ct and histological analysis revealed that incisor enamel at the mandibular molar landmarks was less mineralized in Irf6‐cKO samples (Fig. 5). Scanning electron microscopy (SEM) analysis revealed comparable prismatic structure between Irf6‐cKO and control samples at mature enamel sites (toward the tip), although the transition zone between immature and mature enamel appears to be located more apically in control samples (Fig. 5). Immature enamel sites (toward the apex) were also comparable between Irf6‐cKO and control samples