Cranial Nerve Development Requires Co-Ordinated Shh and Canonical Wnt Signaling

Cranial nerves govern sensory and motor information exchange between the brain and tissues of the head and neck. The cranial nerves are derived from two specialized populations of cells, cranial neural crest cells and ectodermal placode cells. Defects in either cell type can result in cranial nerve developmental defects. Although several signaling pathways are known to regulate cranial nerve formation our understanding of how intercellular signaling between neural crest cells and placode cells is coordinated during cranial ganglia morphogenesis is poorly understood. Sonic Hedgehog (Shh) signaling is one key pathway that regulates multiple aspects of craniofacial development, but whether it co-ordinates cranial neural crest cell and placodal cell interactions during cranial ganglia formation remains unclear. In this study we examined a new Patched1 (Ptch1) loss-of-function mouse mutant and characterized the role of Ptch1 in regulating Shh signaling during cranial ganglia development. Ptch1Wig/ Wig mutants exhibit elevated Shh signaling in concert with disorganization of the trigeminal and facial nerves. Importantly, we discovered that enhanced Shh signaling suppressed canonical Wnt signaling in the cranial nerve region. This critically affected the survival and migration of cranial neural crest cells and the development of placodal cells as well as the integration between neural crest and placodes. Collectively, our findings highlight a novel and critical role for Shh signaling in cranial nerve development via the cross regulation of canonical Wnt signaling.


Introduction
The cranial nerves are part of the peripheral nervous system that governs various critical functions such as sensing and controlling movement within the craniofacial region. Previous studies in avian embryos have shown that some of the cranial nerves including the trigeminal (V) and facial nerves (VII) originate from both cranial neural crest cells and ectodermal placode cells [1,2]. Cranial neural crest cells arise in the dorsal neuroepithelium, delaminate via an epithelial to mesenchymal transformation, and migrate sub-ectodermally throughout the head and neck. In the peripheral nervous system, cranial neural crest cells generate neurons and glia. In contrast, ectodermal placodes comprise thickened regions of surface ectoderm cells, which are distinct from the neuroepithlium. Ectodermal placode cells delaminate from the surface ectoderm to establish the neurogenic core of the cranial nerves [3]. Cellular interactions between neural crest cells and placode cells are essential for proper cranial nerve patterning [4][5][6], and many signaling pathways influence cranial nerve formation in vertebrates by regulating cranial neural crest and/or ectodermal placode cell development [7]. However, our knowledge of how, and in what cell type or tissue these signals primarily function, and also how these different signaling pathways interact remains limited. This is due in part to the early embryonic lethality of many mutants in key developmental pathways.
In a previous study, we performed an N-ethyl-N-nitrosourea (ENU) mutagenesis screen in mice and identified multiple recessive alleles important for craniofacial development [8]. Here we characterize one of these ENU induced mutants called Wiggable (Wig) that carries a mutation in the Patched1 (Ptch1) gene. Ptch1 encodes a receptor for the Hedgehog family of morphogens which includes Sonic Hedgehog (Shh). Unlike Ptch1 null mutant mice which are lethal at E9.5 [9], Ptch1 Wig/ Wig mutants survive until E12.0, allowing an analysis of the effects of aberrant Shh signaling on cranial ganglia morphogenesis. In this study, we took advantage of multiple mouse mutants to clarify the role of cross-talk between the Shh and WNT signaling pathways during the formation of the trigeminal and facial nerves. We discovered that elevated Shh signaling restricts canonical Wnt signaling during cranial ganglia development. This affects the survival of migrating neural crest cells, the pattern of placode development and the integration between neural crest cells and placode cells. Our findings describe the importance of cross-talk between Shh and Wnt signaling in regulating tissue interactions during cranial nerve development.

Ethics Statement
This study was carried out in accordance with recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol (2013-0115) was approved by the Institutional Animal Care and Use Committee of The Stowers Institute for Medical Research. Adult mice were euthanized via CO2 and cervical dislocation according to the recommendations of the American Veterinary Medical Association and all efforts were made to minimize any potential suffering.

Mouse Lines
Ptch1 Wig , Ptch1 LacZ , Wnt1Cre, R26RYFP, TOPgal and Hhat Creface mice were maintained as described previously [8][9][10][11][12][13][14]. The morning of vaginal plug identification was defined as E0.5 for embryo collection and staging. We designated Ptch1 Wig/Wig as Wig homozygous mutants and Hhat Creface/Creface ;Ptch1 Wig/Wig as double-homozygous mice. Either wild-type or heterozygous littermates were used as control mice described in this study. Unless otherwise indicated, we used a minimum of 4 or 5 embryos from multiple distinct litters for each parameter analyzed in this study.
to FVB females to generate founders. The male founders were subsequently mated to FVB females and the resulting daughters were backcrossed to the founders to identify recessive mutant phenotypes. One phenotype encompassed a kinked and dysmorphogenic neural tube in the craniofacial region and was termed Wiggable to reflect the superficial resemblance to Baroque period English wigs. DNA from mutant and control littermate embryos was collected along with tail biopsies from founder males and subjected to microsatellite and single nucleotide polymorphism (SNP) mapping, which narrowed down the affected region to mouse chromosome 13 (qB2 to qB3). This region is gene dense but contained only a few genes known to regulate embryonic neural development, including Ptch1. We used a candidate approach to sequence the genomic (intron/exon) regions of the genes residing in the minimal interval. Sequencing primers were designed to cover the entire mouse Ptch1 gene (available upon request), which was one of our prioritized gene candidates in the minimal affected region, and we discovered a T to A substitution in the 3'end of intron 15 which segregated with the mutant phenotype (S1 Fig.).
Validation of the Ptch1 Wig mutation by qPCR RNA was extracted using the RNeasy kit (Qiagen, Valencia, CA) and quantified using a Bioanalyzer. Primer Express v3.0 was used to make the primers and probes covering the splice junction between exons 15 and 16 of the Ptch1 gene. Amplicon length was 92 base pairs (bp) for the wild type product and 99bp for the Wig mutant product. The following primers were used to generate the amplicon: CAA GCA ACT TCC CCA AAT GTG and ATC CTC CCA GTT TCC CAG TCA. Actin endogenous control probe and primers were ordered as a pre-designed assay from Applied Biosystems (Mouse ACTB (actin, beta) Endogenous Control # 4352933E). For the mutation site, the following dye-coupled sequences were used: Ptch1 wild type probe: 6FAM-TGG CTT CAA GGA CTT C-MGBNFQ; Ptch1 Wig mutant probe: VIC-ACT GGC TTC AAG TTT CTA GGA-MGBNFQ. The linear dynamic range of the qPCR assay was determined by making serial dilutions of the RNA and performing the reverse transcription (RT) reaction and subsequent amplification using the wild type, mutant and Actin endogenous control probes and primers. For the reverse transcription (RT) reaction, we utilized the high capacity cDNA reverse transcription kit from Applied Biosystems (Life Technologies, Grand Island, NY), using 100ng of RNA per reaction in a total volume of 20μL. Each sample was then split into 3 reactions where the wild type, mutant and endogenous control primers and probes were added. The cycling conditions for the cDNA amplification were: denaturation at 95°C for 2 min, followed by 40 cycles of 95°C for 15 seconds (sec) and 60°C for 1 minute (min). Each sample was genotyped by amplification of genomic DNA with forward and reverse primers detailed above using the following cycling conditions: 95°C for 2 min then 32 cycles of 95°C/30sec, 60°C/30 sec, and 72°C/30 sec.) The resulting amplicon was sequenced from both directions using the same primers and aligned to the expected sequence to identify the mutation polymorphism (S2 Fig.).

Cloning of the Ptch1 Wig mutation and Western blot
The sequencing of Wiggable (Wig) founders and their offspring confirmed the mutation introduced a 7 base pair insertion into the Ptch1 cDNA (S1 Fig.). This in turn was predicted to generate a TGA codon 17 base pairs downstream of Exon 16, producing a premature truncation at position 848 in the mouse PTCH1 protein. To model the Wig mutation in cell culture, we used the QuikChange In Vitro mutagenesis kit (Agilent, Santa Clara, CA) to introduce the 7 bp insertion into the full-length mouse Ptch1 cDNA which was cloned in the expression vector pcDNA3.1 (gift from Dr. Kazushi Aoto). The insertion was confirmed by sequencing. Wild type Ptch1 or Ptch1 Wig was transfected into Hek293T cell lines and after 48 hours culture, protein was extracted using a standard lysis buffer (20nM Tris HCl pH8, 137mM NaCl, 10% glycerol, 1% nonidet P-40, 2mM EDTA, protease inhibitors). Protein was quantified using a BSA curve and 40-60ng of protein was loaded per lane. Blots were transferred onto PVDF membranes (Bio-Rad, Hercules, CA) and blocked with 5% BSA solution. To detect PTCH1 protein, blots were probed with an amino-terminal specific PTCH1 antibody (G-19, Santa Cruz Biotechnology, Santa Cruz, CA; 1:1000).
Complementation of the Ptch1 Wig mutation with the Ptch1 LacZ null allele To test whether the Wig mutation was indeed a loss-of-function mutation in Ptch1, we conducted a functional complementation assay with Ptch1 LacZ/ + mice (Jackson Laboratory strain 003081). The Ptch1 LacZ allele contains a lacZ-neo fusion gene that disrupts endogenous Ptch1 gene function [9]. Ptch1 LacZ/ LacZ embryos exhibit neural tube abnormalities and are typically lethal at E9.5. Ptch1 Wig/+ heterozygotes were interbred with Ptch LacZ/+ heterozygotes to generate embryos which were genotyped for the presence of both Wig and LacZ alleles. β-galactosidase staining (described below) was evaluated to determine the spatiotemporal expression of the Ptch1 locus in the Ptch1 Wig/LacZ mutants.

In situ hybridization
Whole-mount in situ hybridization and sectioning was performed as previously described [10].

Results
Truncated PTCH1 protein leads to excessive Shh signaling during mouse embryonic development Through an ENU mutagenesis screen for recessive embryonic phenotypes, we isolated a mutant we named Wiggable (Wig) due to a kinked and overproliferative neural tube that superficially resembled an English wig ( Fig. 1) [8]. Microsatellite and SNP mapping narrowed the candidate region to mouse Chromosome 13 (qB2-qB3). We then focused on candidate genes within the interval known to have embryonic patterning defects. Among them was Patched1 (Ptch1), which encodes a receptor for Hedgehog ligands [9]. Wig ENU mutants segregated with a T to A substitution in the 3'end of intron 15 of Ptch1, which creates a novel splice acceptor site and produces a 7 base pair insertion in the 5' end of Exon 16 (Fig. 1A, S1 Fig.). RT-PCR sequencing of cDNAs derived from Wig carrier mice embryos confirmed the presence of a 7 base pair insertion in the cDNA of Ptch1 (S1 Fig.). This in turn created a premature stop codon17 base pairs into exon 16 which is predicted to truncate the mouse PTCH1 protein at amino acid position 848. The fulllength mouse PTCH1 protein is 1424 amino acids, and the mutation truncates the protein within the 6th extracellular loop, keeping the sterol-sensing domain intact, but abrogating part of the Hedgehog-interacting extracellular domain. Since most of the C-terminal domain is lacking, the Wig ENU mutation is expected to create a Ptch1 loss-of-function allele and abrogate PTCH1 inhibition of Smoothened signaling. In support of this idea, qPCR analysis of Ptch1 expression revealed an approximately 4-fold increase in Ptch1 locus transcription in Ptch1 Wig/Wig mutants relative to wild-type controls (S2 Fig.). This reflects the fact that the Ptch1 locus is under the regulatory control of GLI proteins, and also that Ptch1 loss of function mutations result in the hyper-activation of Shh target genes [16]. The Wig allele of Ptch1 was cloned to test whether it would create a truncated protein in transfected Hek293T cells. Using amino (N)-terminal-specific Ptch1 antibodies to probe the resulting Western blot, we confirmed that the Wig allele of Ptch1 did indeed produce a stable truncated form of PTCH1 that migrated at *90kDa (Fig. 1B). To further evaluate the effect of the Wig mutation on Ptch1 mRNAs levels, we performed a qPCR experiment using mRNA extracted from E10.5 embryos obtained from Ptch1 Wig/+ intercrosses (S2 Fig.). Ptch1 Wig/Wig mutants (red bars, specimens 216, 218, 219, and 224) expressed only the mutant form of Ptch1, with undetectable levels of wild type Ptch1 mRNA expression (S2 Fig.). Ptch1 Wig/+ heterozygotes (Specimens 220-222) showed approximately equal levels of both the wild type and Wig mutant Ptch1 mRNA. No Ptch1 Wig mutant mRNA was detected in wild type embryos. Interestingly, in the Ptch1 Wig/Wig embryos Ptch1 mRNA levels were highly elevated, consistent with deregulated Shh signaling. This is because the Ptch1 locus is under the direct transcriptional regulation of Shh signaling through the binding of GLI transcription factors to its promoter [16].
We next performed a genetic complementation assay of the Ptch1 Wig allele with the Ptch1 LacZ mutation [9]. The Ptch1 LacZ mutation creates a null allele that in Ptch1 LacZ/LacZ embryos leads to severe patterning defects and lethality at E9.5. In contrast the Ptch1 Wig mutation is less severe, with Ptch1 Wig/Wig embryos surviving until E11.5-12.5 (Fig. 1C,D), although this could be due in part to strain background differences. The Ptch1 LacZ allele can also be used to monitor Ptch1 locus activity via β-galactosidase staining, and is strongly expressed in ventral neural tissues and the endoderm in E9.5 control embryos (Fig. 1E). While Ptch1 LacZ/+ and Ptch Wig/+ heterozygotes are normal (Fig. 1E), Ptch1 LacZ/Wig mutants exhibited severe neural tube closure and brain patterning defects together with early embryonic lethality (Fig. 1F). The failure to complement confirmed that the Wig mutation is a novel loss-of-function allele of Ptch1. The fact that β-galactosidase staining was more intense in the Ptch1 LacZ/Wig E9.5 embryos than in controls is indicative of elevated Hh signaling and is consistent with the Ptch1 locus being a direct target of activated Shh signaling (Fig. 1F, S2 Fig.).

Ptch1 Wig/Wig embryos exhibit disrupted cranial nerve development
The survival of Ptch1 Wig/Wig mutants until E11.5-12.5 afforded us the opportunity to examine the role of elevated Shh signaling in the formation and patterning of the cranial nerves. Immunohistochemistry on E9.5 control embryos revealed that the trigeminal (V) and facial nerves (VII) develop as clusters of Neuronal Class III β-tubulin (TUJ1) labeled neurons ( Fig. 2A). In contrast, Ptch1 Wig/Wig mutant embryos exhibited severely disorganized trigeminal and facial nerves (Fig. 2D). A similar phenotype could be seen via neurofilament (2H3) staining ( Fig. 2B and E). By E10.5 the axonal branches of the trigeminal and facial nerves should have begun projecting to the distal facial tissues they eventually innervate as can be seen in control embryos (Fig. 2C). However, in Ptch1 Wig/Wig mutant embryos the cranial nerves do not develop properly or project to their appropriate target tissues (Fig. 2F). These results suggest that the Ptch1 Wig/Wig mutation perturbs cranial nerve development presumably in association with elevated Shh signaling.

Ptch1 Wig/Wig mutation results in cranial neural crest cell death
Cranial nerves are derived from both cranial neural crest cells and ectodermal placodes. The defects in cranial nerve development in Ptch1 Wig/Wig mutant embryos could therefore reflect abnormal neural crest formation and migration, and/or aberrant placode cell development. To address the role of the cranial neural crest in the pathogenesis of the Wig phenotype, we crossed Ptch1 Wig/+ mice into Wnt1Cre;R26RYFP to genetically label cranial neural crest cells and their derivatives [13,14]. We did not observe any major defects in the neural crest cells emerging from the anterior hindbrain and migrating into the first pharyngeal arch in Ptch1 Wig/Wig embryos ( Fig. 3A vs Fig. 3E; red arrowhead). However, there were clearly far fewer migrating neural crest cells within the second pharyngeal arch and facial nerve regions in Ptch1 Wig/Wig mutants compared to controls (Fig. 3A vs Fig. 3E; white arrowhead). We used Sox10 staining to further evaluate the effect of the Wig mutation on the migratory cranial neural crest cells. Interestingly, Ptch1 Wig/Wig mutant embryos exhibited fewer SOX10 protein (Fig. 3B vs Fig. 3F; white arrow) and Sox10 mRNA (Fig. 3D vs. Fig. 3H; red arrowhead) positive neural crest cells specifically within the opV region relative to controls. However in contrast to the situation in the opV region, and taking into account the reduced number of GFP-positive neural crest cells in the facial nerve region, the relative proportion of SOX10/Sox10 positive neural crest cells in Ptch1 Wig/Wig mutants embryos was similar to controls (Fig. 3C vs Fig. 3G, Fig. 3D vs Fig. 3H; white arrowhead). Thus there is a consistent diminishment of neural crest cells in both the opV and facial nerve regions in Ptch1 Wig/Wig compared to controls.
Ptch1 together with Hh signaling are collectively known to play critical roles in cell proliferation and survival [17][18][19]. Therefore we hypothesized that the diminished neural crest cell population in the opV and facial nerve regions of Ptch1 Wig/Wig embryos may be due to apoptosis of neural crest cells. Indeed, TUNEL staining in combination with GFP labeling of neural crest cells revealed a substantial elevation of cell death specifically in neural crest cells in the opV region of Ptch1 Wig/Wig mutants compared to control littermates (Fig. 3I vs Fig. 3M). A similar increase of cell death in GFP-positive neural crest cells was also observed in the facial nerve region of Ptch1 Wig/Wig mutants compared to controls (Fig. 3J vs Fig. 3N). Consistent with these results, TUNEL staining together with SOX10 immunohistochemistry also revealed a significant elevation of cell death in cranial neural crest cells in Ptch1 Wig/Wig mutants (S3 Fig.). In  contrast, we did not observe any difference in apoptosis of TUJ1 positive neuronal cells either in the opV or facial nerve of control versus Ptch1 Wig/Wig embryos (S3 Fig.). These findings demonstrate that aberrant cranial nerve development in Ptch1 Wig/Wig embryos is associated with elevated apoptosis of migratory cranial neural crest cells but not of post-mitotic neuronal cells. Since the Wig mutation in Ptch1 results in elevated Shh signaling, our results suggest that the  spatiotemporal activity of this important mitogen has important consequences for the survival of migratory cranial neural crest cells during cranial nerve development.
We recently discovered that Shh signaling restricts the extent of canonical Wnt signaling in the developing craniofacial prominences [10]. Furthermore it has been well established that Wnt signaling is important for neural crest cell survival, as well as development and patterning of the cranial nerves [20][21][22][23]. Therefore we investigated the possibility that elevated Shh signaling in the Ptch1 Wig/Wig mutants regulates canonical Wnt signaling in the migratory crest and/or placodes, leading to the pathogenesis of cranial nerve defects. Firstly, we performed in situ hybridization for Wnt1, which demarcates the dorsal neuroepithelial territory from which neural crest cells are derived and plays an important role in neural crest development and survival [23]. We found a marked down-regulation of Wnt1 expression in the neuroepithelium of Ptch1 Wig/Wig mutants (Fig. 3O vs Fig. 3K, arrowheads). Secondly, to further characterize the possibility of an association of perturbed canonical Wnt signaling with aberrant cranial nerve development in Ptch1 Wig/Wig mutants, we crossed Ptch1 Wig/+ mice with the TOPgal reporter [11]. We observed a considerable reduction of TOPgal activity in Ptch1 Wig/Wig mutant embryos compared to controls (Fig. 3P, vs. Fig. 3L, arrowheads). These observations are consistent with previous work that has shown that Ctnnb1 (β-catenin) loss-of-function in migrating neural crest cells leads to neural crest cell death [22]. Collectively, these results illustrate that elevated Shh signaling restricts canonical Wnt signaling in cranial neural crest cells and influences their survival.

Ptch1 Wig mutants display defects in cranial placodal formation
In addition to cranial neural crest cells, ectodermal placode cells provide major contributions to the development of the cranial ganglia [24]. We therefore explored whether cranial placode development was altered in association with elevated Shh signaling in Ptch1 Wig/Wig mutant embryos. We used PAX3 and SOX10 co-immunostaining to define the placode cells of the trigeminal region [25], and Ngn1, Ngn2, and NeuroD expression via in situ hybridization as indicators of neuronal maturation in the placode [26]. We observed a marked diminution of PAX3-positive cells in Ptch1 Wig/Wig embryos (Fig. 4A vs. Fig. 4D; green arrowhead). This reduction was further confirmed by Pax3 mRNA in situ hybridization (

Ptch1 Wig/Wig embryos exhibit less integration between cranial neural crest cells and ectodermal placode cells
Cellular and molecular interactions between cranial neural crest cells and ectodermal placodes are essential for proper cranial nerve development [4]. Given that we observed defects in the survival of migrating cranial neural crest cells together with defects in placode development, we carefully examined the interactions between these two cell types during cranial nerve development via TUJ1 immunostaining combined with Wnt1Cre;R26RYFP in control and Ptch1 Wig/Wig embryos. Clear cellular integration between neural crest cells and placode cells was observed in the developing trigeminal (Fig. 5A) and facial nerve of control E9.5 control embryos (Fig. 5B). In contrast, Ptch1 Wig/Wig mutants exhibited considerably reduced mixing of migrating neural crest cells and placodal cells (Fig. 5C-D). Thus, increased Shh signaling in Ptch1 Wig/Wig mutants,  Reduced Shh signaling restores canonical Wnt signaling and early cranial nerve development The observation that elevated Shh signaling results in diminished canonical Wnt signaling in association with cranial nerve defects ( Fig. 3K and L vs Fig. 3O and P) raised the idea that genetically reducing Shh signaling in Ptch1 Wig/Wig embryos might restore Wnt signaling to normal and rescue cranial nerve development. We therefore used the Hedgehog palmitoylase (Hhat) loss-of function mutant (Hhat Creface/Creface ) to lower the levels of Shh signaling in Ptch1 Wig/Wig mutants by producing compound mutants. The disruption of Hhat diminishes palmitoylation of SHH which perturbs its secretion and long range activity [12,27]. Furthermore, we also bred the TOPgal reporter into Hhat Creface/Creface ;Ptch1 Wig/Wig compound mutant mice to visualize the spatiotemporal activity of canonical Wnt signaling. In E9.5 Ptch1 Wig/Wig embryos, Wnt signaling as measured by TOPgal activity was diminished in the trigeminal (red arrowhead) and facial (white arrowhead) nerve regions (Fig. 6A-B). Strikingly, TOPgal activity was restored to wild-type levels in Ptch1 Wig/Wig ;Hhat creface/creface compound mutant embryos (Fig. 6C). Furthermore, Sox10 in situ hybridization confirmed that while Ptch1 Wig/Wig embryos exhibited reduced migrating neural crest cells and hypoplasia of the opV branch of the trigeminal nerve as well as the facial nerve (Fig. 6E), cranial nerve patterning was largely restored in Ptch1 Wig/Wig ;Hhat creface/creface double mutants (Fig. 6F, compare with Fig. 6D). Collectively, these findings suggest that an appropriate balance between Shh and Wnt signaling is required to ensure the survival of cranial neural crest cells and proper development of the cranial neurogenic placodes which are essential for the orchestrated integration of these two cell populations during cranial nerve development.

Discussion
Proper formation of the cranial nerves relies on dynamic tissue interactions mediated by multiple signaling pathways. In this study, we discovered that the interplay between neural crest cells and placodes cells depends on feedback regulation of Shh and Wnt signaling. We focused on patterning of the trigeminal and facial nerves because of their stereotypical morphology and uncovered a causal link between migratory neural crest cell survival and a balance between Shh signaling and the canonical Wnt pathway.
Using Ptch1 Wig/Wig mice to study the role of Shh signaling in early embryogenesis Ptch1 LacZ is a null allele of Ptch1 that is commonly used to study the distribution, level and effects of perturbed Shh signaling during embryogenesis [9]. However, Ptch LacZ/LacZ homozygous mice are lethal at E9.5 and thus do not survive long enough to examine the effects of elevated Shh signaling on cranial nerve development. In contrast, Ptch1 Wig/Wig homozygous mutants  survive until around E12.5 (Fig. 1), which makes them amendable to study the effect of elevated Shh signaling on cranial nerve development.

Elevated Shh signaling disrupts cranial nerve development via increased apoptosis of cranial neural crest cells
We observed that Ptch1 Wig/Wig embryos exhibited disorganized trigeminal and facial nerve development (Fig. 2). These results are consistent with observations that ectopic SHH administration in chick embryos leads to abnormal trigeminal nerve development [28]. In order to a investigate the mechanisms underlying this phenotype, we first explored a role for apoptosis using TUNEL staining together with tissue specific markers and Wnt1Cre;R26RYFP reporter mice. Surprisingly, we detected no differences in cell death between control and Ptch1 Wig/Wig embryos in the trigeminal or facial nerves (S3 Fig.). However, Ptch1 Wig/Wig embryos exhibited substantially elevated levels of cranial neural crest cell apoptosis ( Fig. 3 and S3 Fig.). These results are consistent with other observations that apoptosis of neural crest cells disrupts cranial nerve development in mice [29] and also that ablation of premigratory neural crest cells in chicken embryos results in disorganized trigeminal nerve formation [2,4,30,31]. Collectively, these findings argue that aberrant cranial nerve formation in Ptch1 Wig/Wig mutant embryos arises primarily as a consequence of a diminished population of cranial neural crest cells, which are lost via apoptosis. Increased cell death in cranial neural crest cells in Ptch1 Wig embryos associated with reduction of survival factors Using the Wnt1Cre;R26RYFP reporter line [11,13,14] in combination with Sox10 staining, we confirmed elevated apoptosis of cranial neural crest cells in Ptch1 Wig/Wig embryos (Fig. 3). We noted the greatest reduction of SOX10 positive cells in Ptch1 Wig/Wig embryos occurred in the opV region of the trigeminal nerve and was associated with the apoptotic loss of neural crest cells that typically invade that territory of the neurogenic placode. This prompted us to look for the source of survival signals that act downstream of Shh signaling to promote the survival of neural crest and placodal cells as they interact to form mature cranial nerves. Elevated Shh signaling had been shown to restrict canonical Wnt signaling in various developmental contexts [10,32,33]. Moreover, Ctnnb1 is thought to function as a survival factor for migrating neural crest cells, and is necessary for activating canonical Wnt signaling in these cells [22,34]. Consistent with these observations, Ptch1 Wig/Wig mutants display a reduction in canonical Wnt signaling in migrating cranial neural crest cells together with elevated cell death (Fig. 3). Furthermore, we succeeded in partially restoring cranial nerve development by reducing Shh signaling in Hhat Creface/Creface ; Ptch1 Wig/Wig double mutants. In these double mutants, Hh ligands are not palmitoylated and secreted properly to form a long-range signaling gradient. In addition, because the Wig lesion results in a C-terminal truncation of Ptch1, Smoothened is presumably no longer inhibited and is therefore able to activate Shh signaling effectors. The compound Hhat Creface/Creface ;Ptch1 Wig/Wig mutants exhibit a restoration of Shh signaling levels to that of normal embryos [10]. Consistent with this result, we observed a restoration of normal cranial nerve development in the Hhat Creface/Creface ;Ptch1 Wig/Wig compound mutants, which was accompanied by a restoration of canonical Wnt signaling (Fig. 6). Collectively, these findings indicate that elevated Shh signaling in Ptch1 Wig/Wig embryos leads to a down-regulation of canonical Wnt-dependent survival signaling, resulting in elevated neural crest cell death and consequently cranial nerve developmental defects.
Possible pathway to inhibit trigeminal placode development from elevated Shh signaling Ectodermal placodes are one of the two principal sources of cells that contribute to the cranial nerves with the other being the neural crest [1,2]. Pax3 is widely used as a marker of trigeminal placodal cells [25] and Pax3 loss-of-function results in defects in trigeminal nerve development [35]. Ptch1 Wig/Wig mutant embryos exhibited a significant reduction in Pax3 expressing cells, consistent with a specific effect on trigeminal nerve development (Fig. 4). Interestingly, Wnt signaling [36,37] is required for Pax3 expression in the trigeminal placode, and a recent study demonstrated direct regulation of Pax3 by CTNNB1 [38]. Thus it is possible that the reduction of Pax3 expression in Ptch1 Wig/Wig embryos is due to the down-regulation of canonical Wnt signaling, which affects the regulatory events required for activation of the Pax3 gene and subsequent placode development. It is important to note that other placodal markers such as Ngn1, Ngn2 and NeuroD1 [26] were also down-regulated in the trigeminal placode in Ptch1 Wig/Wig embryos, but not in other placodes. This suggests that the development of individual cranial placodes may have different sensitivities to Shh signaling levels, with the trigeminal being particularly sensitive. It is also important to note altering Shh signaling affects the dorso-ventral patterning of neural tube [39], and this could potentially affect the patterning of neural crest cells emigrating out the neural tube boundary region as was observed in the case of Zic5 loss-offunction [40]. This in turn can then have cascading effects on the proper development and patterning of the cranial nerves.

Possible contribution of poor intermixed cranial neural crest and placodal cells to cranial nerve defects
In this study, we observed relatively poor integration between neural crest and neurogenic cells in the trigeminal and facial nerves of Ptch1 Wig/Wig embryos compared to controls (Figs. 5 and 7). Recent studies have shown that neural crest cells play an important role during cranial nerve development by forming "corridors" or "conduits" which guide and surround neuronal cells [5,15]. Thus, any loss of migratory neural crest populations can impact cranial nerve formation by severing the cellular conduit for nascent axons growing out of the placodal cores of the cranial ganglia. In terms of molecular signals that can orchestrate cranial neural crest cell and placodal interactions, several studies have shown that cell adhesion molecules such as N-cadherin (Cdh2) or αN-catenin (Ctnna2) are important for cranial nerve integration [6,41]. Excessive N-cadherin (Cdh2) expression in Xenopus results in abnormal aggregation and migration of neural crest cells [42]. This suggests that inadequate cell adhesion in neural crest cells could affect the integration of neural crest and placodal cells. Furthermore, CTNNB1 can regulate cell adhesion by interacting with Cadherins [43]. Collectively, these studies suggest that the disorganized cranial nerves in Ptch1 Wig/Wig embryos may also be due in part to disrupted Wnt signaling-dependent cell adhesion in cranial neural crest or/and placode cells, in addition to the reduced numbers of neural crest cells that is caused by apoptosis.
Elevated Shh signaling results in various well characterized craniofacial anomalies including anencephaly, hyperterolism, and cleft lip and palate [10,44,45] as well as a disease spectrum in humans called congenital cranial dysinnervation disorders, represented by Duane retraction syndrome (DRS) or Moebius syndrome (MBS) [46,47]. MBS is characterized by specific cranial nerve defects including trigeminal, abducens (VI) and facial nerves together with hyperterolism, which is a phenotype known to correlate with elevated Shh signaling [48,49]. Our results indicate that the co-ordinated integration of Shh and Wnt signaling plays a key role in promoting the survival and interaction of neural crest cells and ectodermal placodes during cranial nerve development. Moreover, perturbation of these molecular and cellular interactions may underpin the etiology of some congenital cranial dysinnervation disorders.
In conclusion, we have demonstrated a role for cross-talk between Shh and Wnt signaling in neural crest cells and placodes cells during cranial nerve development. Elevated Shh signaling did not affect the formation and migration of cranial neural crest cells, but did result in selective death of migratory neural crest cells. Elevated Shh signaling also perturbed the development of the neurogenic placodes, particularly the trigeminal placode. Together these events disrupted neural crest and placode cell integration and led to aberrant cranial nerve development. Shh signaling restricts canonical Wnt signaling, and reduced Wnt signaling is associated with neural crest cell death and abnormal cranial nerve development. Using genetic approaches to reduce the levels of elevated Shh signaling, we were able to restore the canonical Wnt signaling in the neural crest and placodal territories of the developing head and rescue cranial ganglia morphogenesis.