Nature Genetics
20, 366 - 369 (1998)
doi:10.1038/3840
Mutations in the human connexin gene GJB3 cause erythrokeratodermia
variabilisGabriele Richard1, Lisa E. Smith1, Regina A. Bailey1, Peter Itin2, Daniel Hohl3, Ervin H. Epstein Jr4, John J. DiGiovanna1, 5, John G. Compton1
& Sherri J. Bale.11 Genetic Studies Section, Laboratory of Skin Biology,
National Institute of Arthritis and Musculoskeletal and Skin Diseases, National
Institutes of Health, Bethesda, Maryland, USA. 2 Department of Dermatology, University of Basel,
Switzerland. 3 Department of Dermatology, Hôspital Beaumont,
Lausanne, Switzerland. 4 Department of Dermatology, San Francisco General Hospital,
University of California, San Francisco, California, USA. 5 Division of Dermatopharmacology, Department of Dermatology,
Brown University, Rhode Island Hospital, Providence,
Rhode Island, USA.
Correspondence should be addressed to Sherri J. Bale. sherrib@box-s.nih.gov or Gabriele Richard Gabriele.Richard@mail.tju.eduErythrokeratodermia variabilis (EKV, OMIM 133200) is an autosomal dominant
genodermatosis with considerable intra- and interfamilial variability1. It has a disfiguring phenotype characterized by the independent
occurrence of two morphologic features: transient figurate red patches and
localized or generalized hyperkeratosis (Fig. 1). Both
features can be triggered by external factors such as trauma to the skin.
After initial linkage to the RH locus on 1p (Refs 2,3), EKV was mapped to an interval of 2.6 cM on 1p34-p35,
and a candidate gene (GJA4) encoding the gap junction protein  -4
(connexin 31, Cx31) was excluded by sequence analysis4. Evidence
in mouse suggesting that the EKV region harbours a cluster of epidermally
expressed connexin genes5,
6 led us to characterize the human
homologues of GJB3 (encoding Cx31) and GJB5 (encoding Cx31.1).
GJB3, GJB5 and GJA4 were localized to a 1.1-Mb YAC in the
candidate interval. We detected heterozygous missense mutations in GJB3
in four EKV families leading to substitution of a conserved glycine by
charged residues (G12R and G12D), or change of a cysteine (C86S). These mutations
are predicted to interfere with normal Cx31 structure and function, possibly
due to a dominant inhibitory effect. Our results implicate Cx31 in the pathogenesis
of EKV, and provide evidence that intercellular communication mediated by
Cx31 is crucial for epidermal differentiation and response to external factors.
Gap junctions permit the rapid exchange of ions, secondary messengers and
small metabolites between neighbouring cells, which is crucial for the coordination
of cellular activities. They are formed by connexins, a multigenic family
of 13 polytopic membrane proteins with a common
sequence of structural motifs7,
8 (Fig. 2),
which are classified into  -and  -subtypes.
Connexins assemble by hexameric oligomerization to hemichannels
(connexons), which dock with identical (homotypic) or compatible (heterotypic)
connexons in adjacent cells to form channels with selective properties for
size, charge, gating sensitivity or post-translational regulation. In rodent
skin, gap junction intercellular communication is mediated by at least eight
different connexins, including Cx31, Cx31.1 and Cx30.3 (Refs 5,6,9,10,11), that are preferentially
expressed in the differentiated layers of the epidermis12. The
overlapping spatial and differential expression patterns of connexins in human
skin appears similar to those of rodent skin13,
14,
15,
16,
but not all human homologues have been cloned and analysed.
 | | |
| | |
We identified two expressed sequence tags, ESTs AA079696 and AA235826,
from the human EST database by their similarity to mouse Gjb3 and
Gjb5, respectively. By radiation hybrid mapping, we placed them in proximity
with STS WI-1721 (AA079696, 9.1 cR; AA235826, 9.9 cR), which is also linked
to GJA4 (14.8 cR). YAC clones (899_e_11, 1,120 kb; 802_d_04, 1,380
kb) known to contain GJA4 and microsatellite marker D1S195 also
harboured both EST sequences, suggesting a tandem array of at least three
connexin genes in the EKV critical region. We determined cDNA sequences of
the genes from which both ESTs were derived by genome walking with gene-specific
primers and 5´/3´-RACE using total placental cDNA. Sequence similarity
to rodent connexin genes established them as human homologues of Gjb3
and Gjb5, encoding Cx31 and Cx31.1, respectively. Cx31 and Cx31.1 belong
to the -connexin family, which also includes Cx26 and Cx32. Both
GJB3 and GJB5 were shown to be expressed in human epidermis by
qualitative RT-PCR from total epidermal RNA using gene-specific, intron-crossing
primers (data not shown). DNA sequence analysis of the coding region of
GJB5 failed to detect pathogenic mutations in a panel of 12 unrelated
EKV patients, and was not evaluated further.
Comparison of genomic and cDNA sequence of GJB3 revealed an exon-intron
organization common to that of genes encoding connexins. The complete coding
sequence was contained in a single, uninterrupted ORF of 813 nt preceded by
a putative splice junction located 25 nt upstream of the ATG initiation site
and followed by the 3' UTR with a polyadenylation signal (AATAAA) at position
1,583. The human gap junction protein Cx31 (of predicted molecular weight
30.8 kD) consists of 270 aa (Fig. 2) and differs from
its rodent homologues at 40 residues (82% similarity) that are mainly confined
to the cytoplasmic loop17,
18. Protein structure analysis confirmed
a structural organization typical for -connexins19, including
a conserved arrangement of three cysteine residues in each extracellular loop
(Fig. 2). The predicted arrangement differs from all
other connexins in having one additional residue in E2 that might contribute
to restricted interactions of Cx31 with other connexins8.
We screened coding, 5' and 3' flanking sequences of GJB3 for nucleotide
variations in one affected individual from each EKV family. Sequence analysis
of genomic DNA revealed three different heterozygous missense mutations in
four EKV families (Figs 2,3).
In a Swiss patient with localized hyperkeratosis, we detected a heterozygous
G C transversion at nt 34 (Fig. 3a) that
results in a nonconservative change (G12R) from glycine (GGT) to
a positively charged arginine (CGT). This heterozygous mutation,
which eliminates a BsrBI site in one GJB3 allele and creates
an altered DNA digestion pattern, completely cosegregated with the disease
in the extended family of the proband (Fig. 3a).
In a parent-offspring pair with generalized disease, a missense mutation of
the neighbouring nucleotide was identified (35 GGTGAT),
changing glycine to aspartic acid (G12D) and altering the same BsrBI
site as above (Fig. 3b). Neither mutation was
detected in 140 European control alleles, excluding the possibility that these
variations represent common polymorphisms. The third mutation was a heterozygous
TA substitution at nt 256, which resulted in replacement of cysteine
with serine (C86S) and eliminated a NlaIII site. This mutation was
found in a three-generation family with localized hyperkeratosis, and also
in a sporadic case with generalized hyperkeratosis. Restriction endonuclease
digestion of GJB3 amplicons with NlaIII revealed the mutation
in each affected individual, whereas no such mutation was detected in unaffected
family members or 104 alleles of the European control cohort (
Fig. 3c). In addition to these disease-associated mutations,
we also detected three single-nucleotide polymorphisms in GJB3, including:
a silent sequence variant (N119N) in the C-terminal domain (frequency .03)
and two polymorphisms in the 3' UTR, C or A at position 856 (frequency 0.76
and 0.24, respectively) and G or A at position 866 (frequency 0.66 and 0.34,
respectively). The remainder of the GJB3 coding sequence, flanking
5' UTR (125 bp; including the splice site) and 3' UTR (213 bp) did not exhibit
other sequence aberrations. We have not identified GJB3 mutations in
eight EKV families. This may be explained by the occurrence of mutations that
affect protein expression in portions of the gene that we have not screened.
Alternatively, EKV may be heterogeneous and mutations in another gene, possibly
a functional partner of Cx31, could give rise to a similar phenotype. Such
a candidate is Cx30.3, which has yet to be cloned.
| | Figure 3. GJB3 mutations in EKV. | | |  | Pedigrees of four EKV families are shown. Open symbols, unaffected individuals;
filled symbols, affected persons. All numbered individuals were examined and
their DNA samples obtained. Mutant (MT) and wild-type (WT) DNA sequence of
GJB3 for each mutation are shown, as well as results of restriction endonuclease
analysis confirming each mutation. The lanes are numbered to correspond with
the pedigree figures above. U, uncut control; C, normal control sample; 100-bp
ladder. a, Swiss family with localized EKV (30). The sequence ladder of the non-coding strand of affected individuals
shows a CG transversion of one allele (arrow) corresponding to alteration
of the normal glycine codon to an arginine codon. Digestion of a 766-bp DNA
fragment with BsrBI resulted in a ladder of 69 bp, 268 bp and 429 bp
of the wild-type alleles, and an additional undigested band of 337 bp in the
mutant alleles. b, Family with generalized EKV. A heterozygous CT
transition detected in affected individuals in the non-coding strand (arrow
sequence panel) results in substitution of aspartic acid for glycine, as shown
on the coding strand. The wild-type sequence of an unaffected family member
(1) is shown on the right. Interpretation of the BsrBI digestion is
similar to (a). c, Sequence analysis in EKV patients of both
families demonstrates a TA transversion in one allele (arrow), which
replaces a serine for cysteine in codon 86. NlaIII digestion of wild-type
alleles produces a ladder of 41 bp, 110 bp, 155 bp, 206 bp and 254 bp; mutant
alleles give rise to an additional band of 460 bp.
 Full Figure and legend (31K) |
| |
Our data constitute the first report linking mutations in a gene encoding
a connexin to a human skin disorder, which is probably due to impaired gap
junction intercellular communication; thus, EKV joins the growing list of
diseases caused by connexin defects, including Charcot-Marie-Tooth disease
(Cx32), non-syndromic deafness (Cx26) and cataract20,
21,
22
(Cx50). We have presented genetic evidence that mutations in GJB3 are
the proximal cause of EKV. Glycine 12 lies in the cytoplasmic N-terminal domain
in a stretch of six highly conserved residues, and is invariably present at
this position in connexins across many species (Fig. 4).
The charge of this motif influences the gating polarity of connexons in
vitro23. As both G12R and G12D introduce charged residues,
changes in the polarity of Cx31 channels can be predicted. Mutant versions
of this residue in other connexins have been implicated in autosomal recessive
deafness and X-linked Charcot-Marie-Tooth disease20,
24,
25.
Moreover, cellular localization of mutant Cx32 in mammalian cells demonstrates
that G12S results in a severe defect of protein trafficking, and in a potentially
toxic cytoplasmic accumulation of the protein26. The C86S mutation
occurs in the conserved M2 domain, which is critical for the regulation of
voltage gating8. The neighbouring P87 residue has been described
as a transduction element between voltage sensing and voltage gating of intercellular
channels in Cx26 (27), and mutations affecting
this proline residue and adjacent residues in different connexin genes are
pathogenic28. Residue 86 is serine in Cx26 and threonine in
Cx32, whereas cysteine is in this position in connexins, whose expression
is largely restricted to the epidermis (Cx30.3, 31, 31.1). It is possible
that the serine substitution causes Cx31 to acquire properties adversely affecting
an epidermal-specific connexin function. GJB3 mutations may cause EKV
through haploinsufficiency of gap junction channels, or by dominant inhibition
of normal connexin channel activity by affecting assembly, transport, docking
of connexons, or channel properties. Further functional in vitro and
in vivo studies are required to test these hypotheses, to understand how
mutant Cx31 alters differentiation of the epidermis (hyperkeratosis) and affects
the cutaneous microcapillary system (transient erythema) and to help explain
the clinical features of the disease.
| | |
Methods
Patients and biologic material.
Twelve unrelated families
with sixty affected individuals were ascertained and examined as described4. Five families were consistent with the localized type of EKV and
seven had generalized hyperkeratosis. DNA was obtained from the blood or buccal
mucosa of each individual using standard procedures29.
Localization of human GJB3 and GJB5.
We used the Stanford G3 Radiation Hybrid Panel of 83 clones (Research Genetics)
for radiation hybrid mapping. Analysis was performed by scoring for the presence
or absence of DNA fragments amplified by PCR with human-specific primer pairs,
as visualized by agarose gel electrophoresis and ethidium bromide staining.
Results for GJB3 (Cx31), GJB5 (Cx31.1) and GJA4 (Cx37)
were submitted to the Stanford RH mapping server
(http://shgc-www.stanford.edu/Mapping/rh/search.html), which located
these marker segments relative to each other and the
G3 RH map, and reported lod score analysis of the linkage results.
We obtained two YAC clones (899_e_11 and 802_d_4) identified as carrying
human GJA4 by the Whitehead Institute Center for Genome Research (Research
Genetics). Yeast clones were grown overnight in YPD media and streaked on
AHC media plates (30 °C). Single colonies were harvested, cultured on
AHC plates (48 h) and individual colonies were diluted in water (500  l).
Whole-cell PCR was performed using this yeast template (10 l) with gene-specific
primers (50 l reaction total volume). The products were analysed by electrophoresis
on ethidium bromide-stained 1.5% ME agarose.
Identification and characterization of GJB3 and GJB5.
Human expressed sequences homologous to mouse Gjb3 and
Gjb5 were identified by BLAST analysis of the dbEST database
(http://www.ncbi.nlm.nih.gov/dbEST/index.html). We used partial
sequences contained in the entries AA079696 and AA235826
to design pairs of nested primer sequences (Tm>70 °C) for amplifying
genomic DNA fragments extending 5' and 3' using the Genome Walker kit (Clontech).
We also used nested pairs of gene-specific primers to prepare GJB3
cDNA from placental RNA by RT-PCR, and for 5'- and 3'-RACE (Marathon cDNA
Amplification system, Clontech).
All DNA sequencing was performed on gel-purified, PCR-amplified, double-stranded
templates (Wizard DNA clean-up system, Promega) by PCR cycle-sequencing (Prism
ready reaction system or dRhodamine terminator system, PE Applied Biosytems)
and analysed on an ABI 377 automated sequencer.
We determined alignments, composition and the evolutionary relationships
between the deduced amino acid sequences of human Cx31 and Cx31.1 and those
of mouse and rat, and other human connexin sequences using GeneWorks 2.5.1
(Oxford Molecular).
DNA amplification and mutation analysis.
Coding regions
of GJB3 and GJB5 were PCR amplified from genomic DNA and sequenced
directly. A 1,279-bp fragment of human GJB3 was amplified with primers
derived from the cDNA sequence: 5' UTR, 5'-CATATTTGCTTCCTTCCAGC-3' (corresponding
to bases -253 to -234); 3' UTR, 5'-GTGCCAGCCCTCAAAGGACA-3' (bases 1007-1026).
DNA templates were amplified using Platinum Taq DNA polymerase (Gibco
BRL) and standard PCR conditions (95 °C for 2 min; 35 cycles of 95 °C
for 1 min, 60 °C for 1 min, 72 °C for 1.5 min; and 72 °C for 10
min). Similarly, primers for GJB5 from the 5´ UTR (5´-GACAGTAGTAGAACCCAGCTCCT-3´;
bases -146 to -124) and 3´ UTR (5'-GAATCCTCTCTCTTGCCTGCC-3´; bases
923-943) yielded a PCR product of 1,089 bp. Efficient PCR amplification conditions
were: 95 °C for 2 min; 35 cycles of 95 °C for 1 min, 1.5 min transition
to 60 °C, 60 °C for 1 min, 72 °C for 1.5 min; followed by 72 °C
for 10 min. DNA sequence was obtained by PCR cycle-sequencing using the primers
listed above, and additional nested primers (available upon request).
For restriction endonuclease digestion with BsrBI and NlaIII,
we generated 766-bp amplicons with Cx31 (5'-TAATTCTCTCAGGTAGGCAC-3'; bases
-36 to -17) and Cx31 (5'-GTGGCAGCGGCAGGTGGAAG-3'; bases 707-726) using the
PCR conditions: 95 °C for 2 min; 37 cycles of 95 °C for 1 min, 70
°C for 2 min. The fragments were purified with ChromaSpin-100 columns
(Clontech), digested overnight according to the manufacture s instructions
(New England Biolabs) and analysed on 6% non-denaturing TBE gels (Novex).
Accession numbers.
Swiss-Protein database: human Cx26,
P29033; human Cx32, P08034; rat Cx31, P25305; mouse Cx31, P28231; rat Cx31.1,
P28232; mouse Cx31.1, Q02739; rat Cx30.3, P36380; mouse Cx30.3, Q02738. GenBank:
GJB3, AF099730; GJB5, AF099731.
Received 27 July 1998; Accepted 20 October 1998
REFERENCES
- Mendes da Costa, S. Erythro- et keratodermia variabilis in a mother and a daughter. Acta Derm. Venereol. 6, 255-261 (1925).
- van der Schroeff, J.G. et al. Genetic linkage between erythrokeratodermia variabilis and Rh locus. Hum. Genet. 68, 165-168 (1984). | PubMed | ChemPort |
- van der Schroeff, J.G., van Leeuwen-Cornelisse, I., van Haeringen, A. & Went, L.N. Further evidence for localization of the gene of erythrokeratodermia variabilis. Hum. Genet. 80, 97-98 (1988). | PubMed | ChemPort |
- Richard, G. et al. Linkage studies in erythrokeratodermias: fine mapping, genetic heterogeneity and analysis of candidate genes. J. Invest. Dermatol. 109, 666-671 (1997). | Article | PubMed | ISI | ChemPort |
- Haefliger, J.A. et al. Four novel members of the connexin family of gap junction proteins. Molecular cloning, expression, and chromosome mapping. J. Biol. Chem. 267, 2057-2064 (1992). | PubMed | ISI | ChemPort |
- Hennemann, H. et al. Two gap junction genes, connexin 31.1 and 30.3, are closely linked on mouse chromosome 4 and preferentially expressed in skin. J. Biol. Chem. 267, 17225-17233 (1992). | PubMed | ISI | ChemPort |
- Bruzzone, R., White, T.W. & Paul, D.L. Connections with connexins: the molecular basis of direct intercellular signaling. Eur. J. Biochem. 238, 1-27 (1996). | Article | PubMed | ISI | ChemPort |
- Goodenough, D.A., Goliger, J.A. & Paul, D.L. Connexins, connexons, and intercellular communication. Annu. Rev. Biochem. 65, 475-502 (1996). | Article | PubMed | ISI | ChemPort |
- Goliger, J.A. & Paul, D.L. Expression of gap junction proteins cx26, cx31.1, cx37, and cx43 in developing and mature rat epidermis. Dev. Dyn. 200, 1-13 (1994). | PubMed | ISI | ChemPort |
- Butterweck, A., Elfgang, K., Willeke, K. & Traub, O. Differential expression of the gap junction proteins connexin45, -43, -40, -31, and -26 in mouse skin. Eur. J. Cell. Biol. 65, 152-163 (1994). | PubMed | ISI | ChemPort |
- Risek, B., Klier, F.G. & Gilula, N.B. Multiple gap junction are utilized during rat skin and hair development. Development 116, 639-651 (1992). | PubMed | ISI | ChemPort |
- Brissette, J.L., Kumar, N.M., Gilula, N.B., Hall, J.E. & Dotto, G.P. Switch in gap junction protein expression is associated with selective changes in junctional permeability during keratinocyte differentiation. Proc. Natl Acad. Sci. USA 91, 6453-6457 (1994). | PubMed | ChemPort |
- Guo, H., Acevedo, P., Parsa, F.D. & Bertram, J.S. Gap-junctional protein connexin 43 is expressed in dermis and epidermis of human skin: differential modulation by retinoids. J. Invest. Dermatol. 99, 460-467 (1992). | Article | PubMed | ISI | ChemPort |
- Masgrau-Peya, E., Salomon, D., Saurat, J.H. & Meda, P. In vivo modulation of connexins 43 and 26 of human epidermis by topical retinoic acid treatment. J. Histochem. Cytochem. 45, 1207-1215 (1997). | PubMed | ChemPort |
- Larson, D.M., Wrobleski, M.J., Sagar, G.D., Westphale, E.M. & Beyer, E.C. Differential regulation of connexin 43 and connexin 37 in endothelial cells by cell density, growth, and TGF-1. Am. J. Physiol. 272, C405-415 (1997). | PubMed | ISI | ChemPort |
- Reed, K.E. et al. Molecular cloning and functional expression of human connexin 37, an endothelial cell gap junction protein. J. Clin. Invest. 91, 997-1004 (1993). | PubMed | ISI | ChemPort |
- Hoh, J.H., John, S.A. & Revel, J.P. Molecular cloning and characterization of a new member of the gap junction gene family, connexin-31. J. Biol. Chem. 266, 6524-6531 (1991). | PubMed | ISI | ChemPort |
- Hennemann, H., Schwarz, H.J. & Willecke, K. Characterization of gap junction genes expressed in F9 embryonic carcinoma cells: molecular cloning of mouse connexin31 and -45 cDNAs. Eur. J. Cell Biol. 57, 51-58 (1992). | PubMed | ISI | ChemPort |
- Yeager, M. & Nicholson, B.J. Structure of gap junction intercellular channels. Curr. Opin. Struct. Biol. 6, 183-192 (1996). | Article | PubMed | ISI | ChemPort |
- Bergoffen, J. et al. Connexin mutations in X-linked Charcot-Marie-Tooth disease. Science 262, 2039-2042 (1993). | PubMed | ISI | ChemPort |
- Kelsell, D.P. et al. Connexin 26 mutations in hereditary non-syndromic sensorineural deafness. Nature 387, 80-83 (1997). | Article | PubMed | ISI | ChemPort |
- Shiels, A., Mackay, D., Ionides, A., Berry, V. & Moore, A. A missense mutation in the human connexin 50 gene (GJA8) underlies autosomal dominant "zonular pulverulent" cataract, on chromosome 1q. Am. J. Hum. Genet. 62, 526-532 (1998). | Article | PubMed | ISI | ChemPort |
- Verselis, V.K., Ginter, C.S. & Bargiello, T.A. Opposite voltage gating polarities of two closely related connexins. Nature 368, 348-351 (1994). | Article | PubMed | ISI | ChemPort |
- Denoyelle, F. et al. Prelingual deafness: high prevalence of a 30delG mutation in the connexin 26 gene. Hum. Mol. Genet. 6, 2173-2177 (1997). | Article | PubMed | ISI | ChemPort |
- Zelante, L. et al. Connexin 26 mutations associated with the most common form of non-syndromic neurosensory autosomal recessive deafness (DFNB1) in Mediterraneans. Hum. Mol. Genet. 6, 1605-1609 (1997). | Article | PubMed | ISI | ChemPort |
- Deschênes, M. et al. Altered trafficking of mutant connexin 32. J. Neuroscience 17, 9077-9084 (1997). | ISI |
- Suchyna, T.M., Xu, L.X., Gao, F., Fourtner, C.R. & Nicholson, B.J. Identification of a proline residue as a transduction element involved in voltage gating of gap junctions. Nature 365, 847-849 (1993). | Article | PubMed | ISI | ChemPort |
- Krawczak M. & Cooper D.N. The human gene mutation database. Trends Genet. 13, 121-122 (1997). | Article | PubMed | ISI | ChemPort |
- Richards, B. et al. Multiplex PCR amplification from the CFTR gene using DNA prepared from buccal brushes/swabs. Hum. Mol. Genet. 2, 159-163 (1993). | PubMed | ISI | ChemPort |
- Itin, P., Levy, C.A., Sommacal-Schopf, D. & Schnyder, U.W. Family study of erythrokeratodermia figurata variabilis. Hautarzt 43, 500-504 (1992). | PubMed | ISI | ChemPort |
Acknowledgments
The authors are grateful to the families for their generous participation
in our studies, and to the Foundation for Ichthyosis and Related Skin Types
and the National Registry for Ichthyosis for patient referral. We also appreciate
the expert research nursing assistance by M. Miller and M. Anderson, and outstanding
technical services of G. Poy in oligonucleotide synthesis and DNA sequencing.
D.H. and P.I. were supported by the Swiss National Science Foundation.
|
