was first identified as the mutated gene in Goltz-Gorlin syndrome or FDH in 2007 
. Since then, at least 80 different mutations and large genomic deletions of the PORCN
gene have been found in FDH patients 
. The resulting phenotype in human patients varies widely from mild skin and distal skeletal defects to severe forms of FDH, with multiple defects that include severe limb abnormalities involving long bones, aplasia cutis, limb-body wall complex anomaly, pentalogy of Cantrell, and Van Allen-Myhre syndrome 
. Identification and characterization of PORCN
mutations as the cause of FDH was important for genetic counseling and diagnostic testing in suspected cases. However, an animal model for the condition and functional cell-based assays are needed to determine the functional in vivo
consequences of Porcn
mutations on Wnt signaling and to investigate new therapies.
We first attempted to generate mice from ES cells with a genetrap insertion in the Porcn
locus. This produced a single chimeric animal, but not unexpectedly, there was no germline transmission of this null allele. We then generated a conditional allele of Porcn by introducing loxP
sites flanking exons 3 and 7. When targeted mice are bred with early embryonic Cre-expressing mice (EIIa-Cre
), ectoderm-specific Krt14-Cre
, and limb mesenchyme-specific Prx-Cre
mice, a variety of interesting phenotypes are observed that recapitulate various aspects of the human condition and prove that these mice provide a reliable animal model for human FDH. Our findings are in agreement with and significantly expand on those reported by Barrott and colleagues during the course of this work in a mouse model with a deletion of exons 2 and 3 of Porcn
. We also made important novel observations that may be relevant for other human conditions potentially resulting from Porcn
mutations or other defects in the Wnt signaling pathway.
Although lethality of Porcn
mutations in hemizygous males has already been confirmed in studies with embryos resulting from aggregation of CSD256 ES cells with tetraploid blastocysts 
and in vivo
, our data provides the first in vivo
evidence that many heterozygous females with inactivating mutations of PORCN
also die in utero
. We also demonstrate that, even though neural tube defects are not typically present in liveborn humans with FDH, severe cranial neural tube defects are common in the heterozygous embryos. This is consistent with expression of Porcupine and Wnt proteins in the developing central nervous system and suggests that associated defects could be responsible for the lack of surviving humans with FDH that have NTDs or other severe brain abnormalities 
. We also demonstrated in different ways that mosaicism for a Porcn
mutation is important for the phenotype: we observed limb, skin, and internal organ abnormalities in the CSD256 genetrap chimera and in chimeric animals carrying the Porcn-ex3-7Neo-flox
allele, as well as mosaic skin defects in mice with Porcn-ex3-7
deletion driven by Krt14-Cre
and by EIIa-Cre
. Having the current mouse model will also allow us to study the contribution of X-inactivation patterns for the mosaic expression of the phenotype in females.
inactivation also causes complete absence of hair follicle development and differentiation. This is in agreement with the lack of hair follicles described with ubiquitous Cre recombinase-mediated inactivation 
, but our observation that inactivating Porcn
only in ectoderm using Krt14-Cre
is sufficient to cause this phenotype is novel. In addition, we found that the defects in hair follicle development phenocopy those observed after Krt14-Cre
-driven β-catenin inactivation, implying that they are caused by defective canonical Wnt signaling in ectodermal derivatives. The absence of P-cadherin and Sox9 indicates a very early defect in specification of the hair placode. Our findings of abnormal dentition in the animals are also novel and may result from similar disruption in ectodermal Wnt signaling in the tooth placode.
Another novel observation is that the chimeric animals have some of the rarer phenotypic features occasionally described in FDH, such as urogenital anomalies. Considering the known role of Wnt proteins in the development of the urogenital tract, the Porcn-ex3-7flox mice will be valuable to study how Porcupine regulates Wnt signaling during the development of the urogenital tract.
Finally, the cell culture assays provide evidence that PORCN
containing mutations that cause human FDH can influence the secretion and signaling of WNT proteins from the cell. Consistent with prior data 
, we demonstrate that knockdown of PORCN
by siRNA, as well as some of the FDH-causing mutations that we studied, cause retention of WNT3a in the cell. In contrast, similar experiments with WNT1 show a lesser and different effect on WNT protein secretion and indicate that the effects of identical mutations in PORCN
cannot be generalized to all WNT proteins. A recent study also showed that differential palmit(e)oylation of two residues of Wnt1 has both overlapping and distinct consequences, whereby one is Porcn
-dependent and the other is Porcn
independent. It has been proposed that this differential residue modification may play a role in determining participation in the canonical versus non-canonical Wnt-signaling pathway 
. This data provides further evidence that Porcupine has a complex role in the regulation of secretion and signaling of Wnt ligands and that individual Wnt-proteins are differently influenced. This is consistent with data by others that indicate that effects of Porcupine on Wnt signaling can be uncoupled from Wnt ligand secretion 
The mouse Porcn
alleles that we generated, supported by cell-based assays, provide evidence for an important role of Porcupine in ectodermal and mesodermal signaling of Wnt proteins. This and other generated mouse models 
will be important tools for further detailed characterization of the function of PORCN
in development and disease, and have provided a model for investigating potential therapies for some of the features of FDH that progress postnatally, such as skin defects and peri-orificial papillomas.