Deficiency of PORCN, a regulator
of Wnt signaling, is associated
with focal dermal hypoplasia
Karl-Heinz Grzeschik1, Dorothea Bornholdt1, Frank Oeffner1,
Arne Ko ¨nig2, Marı ´a del Carmen Boente3, Herbert Enders4,
Barbara Fritz1, Michael Hertl2, Ute Grasshoff4, Katja Ho ¨fling5,
Vinzenz Oji6, Mauro Paradisi7, Christian Schuchardt8,
Zsuzsanna Szalai9, Gianluca Tadini10, Heiko Traupe6&
Focal dermal hypoplasia (FDH) is an X-linked dominant
multisystem birth defect affecting tissues of ectodermal and
mesodermal origin. Using a stepwise approach of (i) genetic
mapping of FDH, (ii) high-resolution comparative genome
hybridization to seek deletions in candidate chromosome
areas and (iii) point mutation analysis in candidate genes,
we identified PORCN, encoding a putative O-acyltransferase
and potentially crucial for cellular export of Wnt signaling
proteins, as the gene mutated in FDH. The findings
implicate FDH as a developmental disorder caused by a
deficiency in PORCN.
FDH (also known as Goltz syndrome; OMIM %305600) is an X-linked
dominant disorder characterized by defective development of ectoder-
mal and mesodermal tissues1. These findings hint at a disturbance of
cellular signals involved in the molecular cross-talk between these
tissues as the underlying cause (Supplementary Fig. 1 online).
Linkage analysis in a family with six affected individuals (Tu1;
Supplementary Note online) suggested an FDH locus at either
Xp11.3–Xq21.32 (flanked by the genetic markers DXS1003 and
DXS990) or Xq24–Xq28 (DXS1001 to DXS1193), encompassing a
total of 74 Mb of DNA (Fig. 1a and Supplementary Table 1 online).
We scrutinized both regions by high-resolution comparative genome
hybridization (CGH)2for chromosome rearrangements in ten un-
related individuals with the clinical diagnosis of FDH (individuals
GG1 through GG10; Table 1 and Supplementary Methods online).
Two of these individuals (GG1 and GG2) showed deletions in Xp11.23
that overlapped by B80 kb. This smallest region of overlap included
four genes, SLC38A5, FTSJ1, PORCN and EBP (Fig. 1b). The deletion
breakpoints of individual GG1 were separated by a 16-bp insertion of
unknown origin (Fig. 1c). The junction fragment was also present in
an affected sister but not in the DNA of unaffected individuals (data
not shown). Multiple amplifiable probe hybridization of coding
PORCN exons confirmed the presence of both deletions (Supplemen-
tary Methods and Supplementary Fig. 2 online). GG11 and her
affected family members likewise showed a deletion of one copy
of the coding exons of PORCN (Table 1 and Supplementary Fig. 2
online). The extent of the deletion beyond this gene has not yet
Individuals GG3–GG10 (who did not have any detectable deletions)
were scrutinized for point mutations in PORCN (Table 1). The wide
range of defects in ectodermal and mesodermal tissues reported in
individuals with FDH1suggested a protein with a key function in
developmental control as the most likely candidate: PORCN, a
member of the porcupine (Porc) gene family encoding endoplasmic
reticulum proteins with multiple transmembrane domains. Functional
studies on the human protein have not been reported3. PORC
proteins are predicted to be putative O-acyltransferases involved in
the palmitoylation and secretion of Wnt (Wingless and INT-1)
signaling proteins, which are key regulators of embryonic develop-
ment4,5. Germline mutations in the Wnt pathway have been associated
with several hereditary diseases, and somatic mutations cause cancer
in various tissues6,7. Based on this predicted role, we chose PORCN
as a suitable candidate. There have been no reports of a human
phenotype associated with mutations of SLC38A5, which encodes a
putative transmembrane amino acid transporter (OMIM *608065).
Mutations in FTSJ1, which encodes an S-adenosylmethionine-binding
protein (OMIM *300499), cause nonsyndromic X-linked mental
retardation8, whereas mutation in EBP, which encodes a 3-beta-
hydroxysteroid isomerase, is associated with X-linked dominant
chondrodysplasia punctata (OMIM #302960)9.
In individuals GG3–GG10, sequencing uncovered seven heterozy-
gous stop mutations distributed throughout PORCN leading to loss of
function and one mutation (720-2A4T) that affects the receptor
splice site of exon 9 (Table 1, Supplementary Table 2 and Supple-
mentary Fig. 3 online). We confirmed the mutations by independent
methodology and did not detect any of them in 100 unrelated control
individuals (Supplementary Methods online). Multiple alternatively
spliced transcript variants encoding distinct isoforms had been
observed in a previous study. Exon 9, which encodes a highly
conserved transmembrane domain, is included in all the alternatively
Received 8 February; accepted 27 April; published online 3 June 2007; doi:10.1038/ng2052
1Department of Human Genetics, University of Marburg, Bahnhofstr. 7, 35033 Marburg, Germany.2Department of Dermatology, University of Marburg, Deutschhaus-
Str. 9, 35033 Marburg, Germany.3Department of Dermatology, Hospital del Nin ˜o Jesu ´s, Pasaje Bertre ´s 224, 4000 San Miguel de Tucuma ´n, Argentina.4Department of
Human Genetics, University of Tu ¨bingen, Calwerstr. 7, 72076 Tu ¨bingen, Germany.5Institute for Medical Microbiology, Immunology and Parasitology, University of
Bonn Sigmund-Freud-Str. 25, D-53105 Bonn, Germany.6Department of Dermatology, University of Mu ¨nster, von-Esmarch-Str. 56, 48149 Mu ¨nster, Germany.7Division
of Pediatric Dermatology, Istituto Dermopatico dell’Immacolata, Via dei Monti di Creta 104, 00167 Roma, Italy.8Klinik Pieper, Vorderdorfstr. 17, 79837 St. Blasien-
Menzenschwand, Germany.9Department of Pediatric Dermatology, Heim Pa ´l Children’s Hospital, U¨llo ¨i u ´t 86, 1089 Budapest, Hungary.10Center for Hereditary Skin
Diseases, Istituto di Scienze Dermatologiche, University of Milan, Via Pace 9, 20122 Milano, Italy. Correspondence should be addressed to K.-H.G.
NATURE GENETICS VOLUME 39 [ NUMBER 7 [ JULY 2007 833
© 2007 Nature Publishing Group http://www.nature.com/naturegenetics
The deletions in GG1 and GG2, eliminating PORCN and
flanking genes, demonstrate that FDH can be a component
of a contiguous gene syndrome (Fig. 1b). In view of the universal
role of Wnt signaling6, loss of PORCN function is expected
to be epistatic to many of the clinical features that might be attri-
buted to deficiency in neighboring genes. Whereas mutations in
FTSJ1 are associated with nonsyndromic X-linked mental retarda-
tion in males, female carriers are reportedly not affected8.
Thus, deletion of one copy of this gene might not lead to mental
retardation. Moreover, specific cutaneous features of X-linked domi-
nant chondrodysplasia punctata were absent in GG1 and GG2,
although EBP was deleted. These affected individuals had not been
scrutinized for epiphyseal stippling in early childhood. We expect
that increased levels of 8(9)cholesterol and 8-dehydrocholesterol
might be present in their cultured fibroblasts originating from an
affected skin region9.
PORC function in a sufficient number of cells might be essential for
survival of early embryos. Lyonization has been invoked to explain
various degrees of severity ranging from mild involvement to a severe
fetal malformation syndrome1. Notably, in DNA extracted from
venous blood of the familial cases (GG1, GG2 and affected individuals
I.2, II.2, II.4, III.2 and III.4 of family Tu1), we observed extremely
imbalanced X chromosome inactivation (Z95:5). Individual III.1
could not be tested because of insufficient material. In contrast, the
X inactivation ratio was 55:45 in unaffected individual III.3 (Fig. 1 and
Table 1). Marked skewing of X inactivation seems to be a trait
segregating in some families10. We hypothesize that in most familial
FDH cases, severe skewing of X inactivation reduces the number of
embryonal cells in which the mutated X chromosome is active.
By contrast, skewing of X inactivation in sporadic cases was less
pronounced (Table 1). Here, our data suggest that a second mechanism
minimizes the deleterious effects of PORCN deprivation: postzygotic
22.214.171.124 126.96.36.199188.8.131.52 184.108.40.206
220.127.116.11 18.104.22.168 23 2425 2821.3221.33
Figure 1 Zooming in on mutations associated with FDH. (a) Pedigree and genotypes of X-linked markers of family Tu1 (Supplementary Note). ‘-’ ¼ not
determined. Alleles segregating with FDH are outlined. Electropherograms of an AR sequence polymorphism generated from enzymatically digested
(HpaII+) or undigested (HpaII–) lymphocyte DNA demonstrate extreme skewing of X inactivation in affected individuals (Supplementary Methods). Allele
sizes are below electropherograms; ratios of inactivation are below pedigree symbols (italics). (b) PORCN emerges as a candidate. Genetic mapping
(Supplementary Table 1) suggested X chromosomal regions 1 and 2 (gray bars) as candidates for the FDH locus. Within Xp11.23, high-resolution CGH
detected deletions 1 and 2, overlapping in an interval containing four genes, shown schematically with their intron-exon structures. Stop mutations
(numbered diamonds) detected in individuals GG3–GG10 (Table 1 and Supplementary Table 2) are indicated on a schematic of PORCN-D cDNA (gray bar
with subdivisions depicting splice junctions; 1,942 bp, 15 exons, 461 amino acids (aa)). The black bar above the cDNA schematic is a map of the intron-
exon structure of PORCN-D (vertical black bars indicate exons). (c) Analysis of deletion 1. High-resolution CGH of DNA from individual GG1 uncovered a
heterozygous deletion affecting five genes (graphical representation of log2ratio of hybridization signals of test versus reference sequence shown below a
segment of physical map (in bp) of Xp11.23 (Chr X), with predicted genes indicated by horizontal blue bars). The DNA sequence of the junction fragment
from a cloned allele of GG1 (electropherogram) and comparison of this sequence (Pat, green) with human reference sequence (NCBI build 36.1) (Chr X,
blue) identified a 137,411-bp deletion (g.48191095_48328506del) and an insertion of 16 bp of unknown origin (red) between the deletion breakpoints.
834VOLUME 39 [ NUMBER 7 [ JULY 2007 NATURE GENETICS
© 2007 Nature Publishing Group http://www.nature.com/naturegenetics
mutation, resulting in mosaicism with few
cells carrying the mutation and the majority
harboring the wild-type allele (Supplemen-
tary Fig. 3 and Table 1). In fact, 10% of FDH
cases are male1and almost certainly represent
postzygotic mosaicism. None of the indivi-
duals with apparent allele skewing had
affected family members. We predict that in
a considerable number of female sporadic
cases, FDH might likewise originate from
postzygotic PORCN mutations. The cells
with one mutated PORCN allele that were
present in early embryos might not be repre-
sented in blood cells in all cases. However,
evidence for somatic mosaicism of PORCN in
white blood cell DNA from a woman with
FDH argues against germ cell mosaicism in
unaffected parents and could be an important
clue for genetic counseling.
Wnt signals are involved in virtually
every aspect of embryonic development and
also control homeostatic self-renewal in a
number of adult tissues. Attachment of pal-
mitate by PORC prepares Wnt molecules for
and thus represents a pivotal entry point in
the signaling cascade5. Focal deficiency of
Wnt signaling in FDH could affect cell fate
or result in failure of progenitor cells to
expand. Focal lesions in diverse tissues, such
as hypoplastic skin lesions, defects in limb
and eye development (Supplementary Fig. 1)
and longitudinal striation of the long bones
due to decreased bone density11, seem to
represent an aggregation of various deficien-
cies that individually have been reported
as systemic defects in model organisms and in humans with non-
functional components of the Wnt signaling cascade6,12.
Here we associate a human developmental disorder with the
absence of functional PORCN for the first time. The PORCN isoforms
generated through alternative splicing might affect the processing and
secretion of different Wnt factors3,13. In a given cell or tissue, only a
subset of Wnts stimulates signal response. PORCN mutations affecting
only some of the isoforms might generate phenotypic effects restricted
to tissues in which the associated Wnt proteins are essential. Accord-
ingly, traits expressing part of the FDH phenotypic spectrum, such as
Van Allen-Myhre syndrome14and osteopathia striata with cranial
sclerosis15are strong candidates for PORCN mutations.
Note: Supplementary information is available on the Nature Genetics website.
We are grateful to the affected individuals and their families for their
cooperation, and we thank M. Schad (Deutsches Ressourcenzentrum fu ¨r
Genomforschung, Berlin) for help with inverted graphical presentation of
CGH data. This research was supported by BMBF (Netzwerk fu ¨r Ichthyosen
und verwandte Verhornungssto ¨rungen (NIRK)), the European Union
(Coordination Action GeneSkin) and Forschungspool des Fachbereichs Medizin
der Philipps-Universita ¨t Marburg.
This study was designed by K.-H.G. and R.H. Clinical phenotype assessment and
proband recruitment were performed by A.K, M.d.C.B., H.E., U.G., M.H., K.H.,
V.O., M.P., C.S., Z.S., G.T., H.T. and R.H.; genetic mapping, mutation analyses
and X inactivation analysis were performed by D.B., B.F. and F.O.; data analysis
was performed by K.-H.G., D.B and F.O. and K.-H.G., F.O., A.K. and R.H.
contributed to the writing of the paper.
COMPETING INTERESTS STATEMENT
The authors declare no competing financial interests.
Published online at http://www.nature.com/naturegenetics
Reprints and permissions information is available online at http://npg.nature.com/
1. Gorlin, R.J., Cohen, M.M.J. & Hennekam, R.C.M. in Oxford Monographs on Medical
Genetics Vol. 42 571–576 (Oxford Univ. Press, Oxford, 2001).
2. Selzer, R.R. et al. Genes Chromosom. Cancer 44, 305–319 (2005).
3. Caricasole, A., Ferraro, T., Rimland, J.M. & Terstappen, G.C. Gene 288, 147–157
4. Willert, K. et al. Nature 423, 448–452 (2003).
5. He, X. & Axelrod, J.D.A. Development 133, 2597–2603 (2006).
6. Clevers, H. Cell 127, 469–480 (2006).
7. Moon, R.T., Kohn, A.D., De Ferrari, G.V. & Kaykas, A. Nat. Rev. Genet. 5, 691–701
8. Freude, K. et al. Am. J. Hum. Genet. 75, 305–309 (2004).
9. Derry, J.M. et al. Nat. Genet. 22, 286–290 (1999).
10. Puck, J.M. & Willard, H.F. N. Engl. J. Med. 338, 325–328 (1998).
11. Happle, R. & Lenz, W. Br. J. Dermatol. 96, 133–135 (1977).
12. Gordon, M.D. & Nusse, R. J. Biol. Chem. 281, 22429–22433 (2006).
13. Tanaka, K., Okabayashi, K., Asashima, M., Perrimon, N. & Kadowaki, T. Eur.
J. Biochem. 267, 4300–4311 (2000).
14. Hancock, S. et al. Am. J. Med. Genet. 110, 370–379 (2002).
15. Behninger, C. & Rott, H.D. Genet. Couns. 11, 157–167 (2000).
Table 1 PORCN mutations associated with FDH
Characteristic clinical features
(age at diagnosis, in years)
Mutation (and amino acid
change, where applicable)
ratio in blood
LSL, ND, BH (36)
LSL, ND, ABD (32)
LSL, ND, SY, ABD, STR,
DD, PMU (10)
Microdeletion 1 (Fig. 1)
Microdeletion 2 (Fig. 1)
370C4TGG4LSL, SY, ABD, DD,
PAP, PMU (8)
GG5LSL, ABD, STR, MR (12)50/50SPM
GG6LSL, ABD, C, MIC, O,
HK, MVP (11)
(Supplementary Fig. 2)
GG7LSL, SY, ABD, DD (16)65/35S
GG8 LSL, PH, SY, ABD, C (61)60/40SPM
GG9LSL,PH,ABD,STR (50)80/20S PM
GG10 LSL, ND (6) N.i.SPM
GG11LSL, ND, ABD, TF (25)97/3F
GG1–GG11 are unrelated individuals with FDH; GG11 is individual III.2 from family Tu1 (Supplementary Note). The
extent of microdeletions 1 and 2 is shown in Figure 1b; the extent of microdeletion 3 beyond PORCN has not been
determined. The point mutations in the DNA of GG3–GG10 were determined experimentally, and the predicted amino
acid change is indicated, as well. The proportion of DNA methylation at the AR locus of each pair of X chromosomes
indicates skewing of X inactivation. N.i., not informative; F, familial; S, sporadic; PM, postzygotic mosaic. ABD,
asymmetric bone defects; BH, breast hypoplasia; C, coloboma; DD, dental defects; HK, horseshoe kidney; LSL, linear
skin lesions; MIC, microphthalmia; MR, mental retardation; MVP, mitral valve prolapse; ND, nail dysplasia; PAP,
perianal papillomas; PH, patchy hairlessness; PMU, papillomas of the mucosa; O, omphalocele; SY, syndactyly; STR,
striation of bones; TF, tetrad of Fallot. Written consent was obtained from participants of this study. The procedures
were approved by the Ethics Committee of the Medical Faculty of Philipps University Marburg.
NATURE GENETICS VOLUME 39 [ NUMBER 7 [ JULY 2007 835
© 2007 Nature Publishing Group http://www.nature.com/naturegenetics