Clinical and Genetic Heterogeneity of Erythrokeratoderma Variabilis

The skin disease erythrokeratoderma variabilis (EKV) has been shown to be associated with mutations in GJB3 and GJB4 encoding connexin (Cx)31 and Cx30.3, respectively. Gap junctions composed of Cx proteins are intracellular channels providing a mechanism of synchronized cellular response facilitating metabolic and electronic functions of the cell. In the skin, Cx31 and Cx30.3 are expressed in the stratum granulosum of the epidermis with a suggested role in late keratinocyte differentiation. Molecular investigations of GJB3 and GJB4 were performed in five pedigrees and three sporadic cases of EKV. Mutational analyzes revealed disease-associated Cx31 or Cx30.3 mutations in only three probands of which two were novel mutations and one was a recurrent mutation. These genetic studies further demonstrate the heterogeneous nature of the erythrokeratodermas as not all individuals that were clinically diagnosed with EKV harbor Cx31 or Cx30.3 mutations. The skin disease erythrokeratoderma variabilis (EKV) has been shown to be associated with mutations in GJB3 and GJB4 encoding connexin (Cx)31 and Cx30.3, respectively. Gap junctions composed of Cx proteins are intracellular channels providing a mechanism of synchronized cellular response facilitating metabolic and electronic functions of the cell. In the skin, Cx31 and Cx30.3 are expressed in the stratum granulosum of the epidermis with a suggested role in late keratinocyte differentiation. Molecular investigations of GJB3 and GJB4 were performed in five pedigrees and three sporadic cases of EKV. Mutational analyzes revealed disease-associated Cx31 or Cx30.3 mutations in only three probands of which two were novel mutations and one was a recurrent mutation. These genetic studies further demonstrate the heterogeneous nature of the erythrokeratodermas as not all individuals that were clinically diagnosed with EKV harbor Cx31 or Cx30.3 mutations. Connexin denaturing high performance liquid chromatography erythrokeratoderma variabilis palmoplantar keratoderma The erythrokeratodermas are a heterogeneous group of disorders of keratinization. They are usually inherited in an autosomal dominant fashion, but autosomal recessive transmission has also been reported. Distinction between the erythrokeratodermas is usually on clinical grounds and is not always easy. Erythrokeratoderma variabilis (EKV; OMIM 133200) usually presents at birth or during infancy. It is characterized by the coexistence of fixed hyperkeratotic plaques mainly on the extensor surfaces of the limbs, with transient erythematous patches, which may be circinate, geographic, or targetoid in shape. Approximately 50% of EKV patients have palmoplantar keratoderma (PPK) (Richard, 2003Richard G. Connexin gene pathology.Clin Exp Dermatol. 2003; 28: 397-409Crossref PubMed Scopus (57) Google Scholar). The lesions have a predilection for the distal extremities, buttocks, and trunk. There is considerable intra- and interfamilial variability in severity of the skin disease, for example, some red patches may last only a few hours whereas others persist for a few days to a few weeks. The patches may migrate followed by fine white scaling. Because of its variable penetrance, the classification of EKV can often be difficult particularly as it shares clinical features with other erythrokeratodermas. Progressive symmetric erythrokeratoderma (PSEK; OMIM 602036) is one of the less prevalent erythrokeratoderma variants initially described byGottron, 1922Gottron H.A. Congential angelegte symmetrische progressive erythrokeratodermie.Zentbl Haut-Geschl Krankh. 1922; 4: 493-494Google Scholar. It is characterized by the development of symmetrical plaques on the limbs, buttocks, and face during early childhood. The plaques progress during childhood and frequently stabilize during the teenage years (Richard and Ringpfeil, 2003Richard G. Ringpfeil F. Ichthyoses, erythrokeratodermas and associated disorders Vol. 1.in: Bolognia J.L. Jorizzo J.L. Rapini J.L. Dermatology. Mosby, London2003: 799-801Google Scholar). It is inherited in an autosomal dominant manner, often with incomplete penetrance. The main feature distinguishing EKV from PSEK is the variable erythema and possibly lack of facial lesions in most cases. Although these erythrokeratodermas are suggested to be distinct clinical syndromes, in reality there is considerable phenotypic overlap (van Steensel, 2004van Steensel M. Does progressive symmetric erythrokeratoderma exist?.Br J Dermatol. 2004; 150: 1043-1045Crossref PubMed Scopus (23) Google Scholar). Moreover, individuals with both a PSEK and EKV phenotype have been described in the same family (Macfarlane et al., 1991Macfarlane A.W. Chapman S.J. Verbov J.L. Is erythrokeratoderma one disorder? a clinical and ultrastructural study of two siblings.Br J Dermatol. 1991; 124: 487-491Crossref PubMed Scopus (46) Google Scholar). Other rarer EKV variants described include erythroderma en cocardes, also known as Degos' disease (Rajagopalan et al., 1999Rajagopalan B. Pulimood S. George S. Jacob M. Erythrokeratoderma en cocardes.Clin Exp Dermatol. 1999; 24: 173-174Crossref PubMed Scopus (8) Google Scholar), reticulate erythrokeratoderma (Itin et al., 2003Itin P.H. Moschopulos M. Richard G. Reticular erythrokeratoderma: A new disorder of cornification.Am J Med Genet. 2003; 120: 237-240Crossref Scopus (8) Google Scholar), EKV with erythema gyratum repens-like lesions (Landau et al., 2002Landau M. Cohen-Bar-Dayan M. Hohl D. Ophir J. Wolf C.R. Gat A. Mevorah B. Erythrokeratodermia variabilis with erythema gyratum repens-like lesions.Pediatr Dermatol. 2002; 19: 285-292Crossref PubMed Scopus (23) Google Scholar).Saba et al., 2005Saba T.G. Montpetit A. Verner A. Rioux P. Hudson T.J. Drouin R. Drouin C.A. An atypical form of erythrokeratodermia variabilis maps to chromosome 7q22.Hum Genet. 2005; 116: 167-171Crossref PubMed Scopus (26) Google Scholar recently described EKV-like lesions, ichthyosis, sensorineural hearing loss, peripheral neuropathy, psychomotor retardation, congenital chronic diarrhea, and an elevation of very long chain fatty acids in a French–Canadian pedigree, mapping to chromosome 7q22, a region containing connexin (Cx)31.1. The identification of the gene defects associated with this syndrome and the varer EKV varients would be of interest. Autosomal dominant EKV was initially linked to chromosome 1p34–1p35 where there is a cluster of Cx genes encoding the gap junction proteins Cx30.3 (GJB4), Cx31 (GJB3), Cx31.1 (GJB5), and Cx37 (GJA4) (van der Schroeff et al., 1984van der Schroeff J.G. Nijenhuis L.E. Meera Khan P. et al.Genetic linkage between erythrokeratrodermia variabilis and Rh locus.Hum Genet. 1984; 68: 165-168Crossref PubMed Scopus (40) Google Scholar; Richard et al., 1997Richard G. Lin J.P.J. Smith L. et al.Linkage studies in erythrokeratodermias; fine mapping, genetic heterogeneity and analysis of candidate genes.J Invest Dermatol. 1997; 109: 666-671Crossref PubMed Scopus (58) Google Scholar; Macari et al., 2000Macari F. Landau M. Cousin P. et al.Mutation in the gene for connexin 30.3 in a family with erythrokeratodermia variabilis.Am J Hum Genet. 2000; 67: 1296-1301Abstract Full Text Full Text PDF PubMed Scopus (128) Google Scholar). Subsequently, pathogenic GJB3 mutations in four families with EKV were identified (Richard et al., 1998Richard G. Smith L.E. Bailey R.A. et al.Mutations in the human connexin gene GJB3 cause erythrokeratodermia variabilis.Nat Genet. 1998; 20: 366-369Crossref PubMed Scopus (287) Google Scholar), demonstrating a link between defects in Cx31-associated gap junctions and aberrant epidermal differentiation and function. To date, six dominant GJB3 mutations have been identified in eight unrelated EKV families (Richard et al., 1998Richard G. Smith L.E. Bailey R.A. et al.Mutations in the human connexin gene GJB3 cause erythrokeratodermia variabilis.Nat Genet. 1998; 20: 366-369Crossref PubMed Scopus (287) Google Scholar,Richard et al., 2000Richard G. Brown N. Smith L.E. et al.The spectrum of mutations in erythrokeratodermias- novel and de novo mutations in GJB3.Hum Genet. 2000; 106: 321-329Crossref PubMed Scopus (76) Google Scholar; Wilgoss et al., 1999Wilgoss A. Leigh I.M. Barnes M.R. et al.Identification of a novel mutation R42P in the gap junction protein beta-3 associated with autosomal dominant erythrokeratoderma variabilis.J Invest Dermatol. 1999; 113: 1119-1122Crossref PubMed Scopus (62) Google Scholar). All nucleotide changes lead to heterozygous amino acid substitutions. The mutation L34P in GJB3 was the first homozygous change providing molecular evidence for a recessive variant resulting in EKV (Gottfried et al., 2002Gottfried I. Landau M. Glaser F. et al.A mutation in GJB3 is associated with recessive erythrokeratodermia variabilis (EKV) and leads to defective trafficking of the connexin 31 protein.Hum Mol Genet. 2002; 11: 1311-1316Crossref PubMed Scopus (67) Google Scholar). A second Cx31 mutation (E100K) has also been described associated with recessive EKV (Terrinoni et al., 2004Terrinoni A. Leta A. Pedicelli C. et al.A novel recessive connexin 31 (GJB3) mutation in a case of erythrokeratodermia variabilis.J Invest Dermatol. 2004; 122: 837-839Crossref PubMed Scopus (22) Google Scholar; Terrinoni and Melino, 2005Terrinoni A. Melino G. Recessive EKV.J Invest Dermatol. 2005; 124: 270-271Crossref Google Scholar). GJB3 mutations have been shown to alter gap junction communication and/or induce cell death in vitro (Di et al., 2002Di W.L. Monypenny J. Common J.E. et al.Defective trafficking and cell death is characteristic of skin disease- associated connexin 31 mutations.Hum Mol Genet. 2002; 11: 2005-2014Crossref PubMed Scopus (66) Google Scholar; Diestel et al., 2002Diestel S. Richard G. Doring B. Traub O. Expression of a connexin31 mutation causing erythrokeratodermia variabilis is lethal for HeLa cells.Biochem Biophys Res Commun. 2002; 296: 721-728Crossref PubMed Scopus (38) Google Scholar; Rouan et al., 2003Rouan F. Lo C.W. Fertala A. et al.Divergent effects of two sequence variants of GJB3 (G12D and R32W) on the function of connexin 31 in vitro.Exp Dermatol. 2003; 12: 191-197Crossref PubMed Scopus (26) Google Scholar; Terrinoni and Melino, 2005Terrinoni A. Melino G. Recessive EKV.J Invest Dermatol. 2005; 124: 270-271Crossref Google Scholar). In transfected keratinocytes, EKV-associated Cx31 mutant proteins had a cytoplasmic distribution, suggesting that the mutant forms of this protein will not form normal Cx31 gap junctions between adjacent cells (Di et al., 2002Di W.L. Monypenny J. Common J.E. et al.Defective trafficking and cell death is characteristic of skin disease- associated connexin 31 mutations.Hum Mol Genet. 2002; 11: 2005-2014Crossref PubMed Scopus (66) Google Scholar). The initial molecular studies of EKV suggested that there was genetic heterogeneity, as approximately half of the reported families did not have mutations in GJB3. Genetic studies in a large Israeli–Kurdish family with EKV associated with erythema gyratum repens, revealed a heterozygous missense mutation (F137L) in the gene GJB4 encoding Cx30.3, demonstrating a role for another Cx gene in the pathogenesis of EKV (Macari et al., 2000Macari F. Landau M. Cousin P. et al.Mutation in the gene for connexin 30.3 in a family with erythrokeratodermia variabilis.Am J Hum Genet. 2000; 67: 1296-1301Abstract Full Text Full Text PDF PubMed Scopus (128) Google Scholar). In 13 unrelated EKV families without mutations in GJB3, six distinct missense mutations in GJB4 in five families and a sporadic case of EKV have also been described (Richard et al., 2003Richard G. Brown N. Rouan F. et al.Genetic heterogeneity in erythrokeratodermia variabilis: Novel mutations in the connexin gene GJB4 (Cx30.3) and genotype-phenotype correlations.J Invest Dermatol. 2003; 120: 601-609Crossref PubMed Scopus (102) Google Scholar). In this paper, we describe the clinical phenotypes and molecular findings in five pedigrees and three sporadic cases of EKV. Affected individuals from five families and three sporadic cases displaying clinical features of EKV were assessed for disease-causing germline mutations in the coding sequence of GJB3 and GJB4 encoding Cx31 and Cx30.3, respectively. As with previously described cases of EKV and its variant forms, none of the affected individuals analyzed in this study had hearing, nail or hair anomalies. In addition to the dominant pathogenic mutations discovered in this study, several polymorphic sequence variants were identified and their allele frequencies determined in control populations (Table I).Table IGJB3 and GJB4 sequence variants of unknown biological relevance discovered in this studyNucleotideAmino acidTotal allele frequency (%)Afro CaribbeanAsianUnknown ethnicityGJB3C–T 94R32W6/878 (0.68)2/3041/803/494G–C 109V37L1/878 (0.11)1/3040/800/494C–T 357N119N51/878 (5.81)19/30410/8022/494GJB4C–G 204F68L1/746 (0.13)0/3040/801/362C–T 207F69L2/746 (0.26)1/3040/801/362C–T 307R103C0/746 (–)0/3040/800/362C–G 507C169W6/362 (1.66)——6/362C–A 451R151S3/362 (0.83)——3/362A–C 611E204A3/362 (0.83)——3/362C–T 693P231P2/362 (0.55)——2/362G–A 699T233T0/362 (–)——0/362C–T 292R98C0/362 (–)——0/362C–G 507/Del GTCTC169W/154del40/362 (–)——0/362Total number of alleles, allele frequencies, and a breakdown of control population ethnicity are given where known. Open table in a new tab Total number of alleles, allele frequencies, and a breakdown of control population ethnicity are given where known. Three disease-associated mutations were detected in our EKV cohort. In the family termed EKV1, both affected individuals (mother and son) had a heterozygous G-to-A transition at nucleotide 35 in codon 12 of GJB3 (G12D), changing glycine to aspartic acid (Figure 1a). This mutation has been previously associated with EKV (Richard et al., 1998Richard G. Smith L.E. Bailey R.A. et al.Mutations in the human connexin gene GJB3 cause erythrokeratodermia variabilis.Nat Genet. 1998; 20: 366-369Crossref PubMed Scopus (287) Google Scholar). Both affected family members displayed similar clinical phenotypes with the classic features of EKV, including symmetrical, rugose, hyperkeratotic plaques on the limbs and trunk with serpiginous borders and more transient, figurate, erythematous patches and plaques on the chest and neck (Figure 1b and c). In both cases, dry, scaly skin was observed in the neonatal period, followed by the development of migrating, erythematous patches or blotches on the trunk and limbs in the subsequent 4 mo. "Mauserung-like" lesions were also noted on the extensor aspect of the joints. In the sporadic case termed EKV2, a novel heterozygous A-to-T transition at nucleotide 568 was identified in GJB4, resulting in the amino acid change of methionine to leucine at codon 190 (M190L) of Cx30.3. M190L was not detected in 878 control chromosomes. On examination, fixed plaques with elevated edges were observed in the flexures (Figure 2a) and there was also evidence of mild palmoplantar desquamation. She also had facial erythema with fine scaling (Figure 2a). Transient, migratory patches were seen on the lower trunk and arms. Erythematous plaques in the flexures were reported from the age of 3–4 y. In the family termed EKV 3, DNA was available from four family members and was screened for mutations in GJB4 and GJB3. Affected family members 1, 2, and 4 were heterozygous for a C-to-T transition at nucleotide 625 in GJB3 (Figure 3) resulting in the substitution of a leucine to phenylalanine at codon 209 of Cx31 (L209F). This change was not found in 146 control samples from mixed ethnic backgrounds. In addition, the unaffected mother and one affected family member also harbor a C/T transition at nucleotide 798 in GJB3 that does not result in an amino acid substitution or segregate with the disease phenotype. C/T 798 in GJB3 is likely to be a polymorphism with no disease association seen in this and other studies (Richard et al., 2000Richard G. Brown N. Smith L.E. et al.The spectrum of mutations in erythrokeratodermias- novel and de novo mutations in GJB3.Hum Genet. 2000; 106: 321-329Crossref PubMed Scopus (76) Google Scholar). The clinical features of this family have already been described (Periš et al., 1993Periš Z. Gruber F. Palle M. Treatment of erythrokeratodermia variabilis with etretinate and acidretin.Acta Dermatovenerol Croatica. 1993; 1: 169-172Google Scholar). Previous work has shown that Cx mutations associated with both recessive and dominant EKV displayed a characteristic non-membranous localization when transfected as an EGFP (enhanced green florescent protein) fusion protein into cell lines. The subcellular localization of (L209F) Cx31-EGFP was analyzed in the immortalized keratinocyte cell line, NEB1. Like other Cx31 EKV-associated mutant alleles L209F had a primarily cytoplasmic localization, characteristic of an in vitro trafficking defect (Figure 4). The sporadic case termed EKV4 had three genetic changes in the coding region of GJB4 (Figure 5) of unclear disease association. A single heterozygous C-to-T substitution at nucleotide 292 in codon 98 (R98C) was identified changing a conserved arginine to a cysteine in Cx30.3. R98C was not detected in the control population. In addition, 154del4 and the more prevalent polymorphism C169W were also detected (allele frequencies shown in Table I). Following subcloning and sequencing, R98C and 154del4 were shown to be together on the same GJB4 allele with C169W on the other. As 154del4 is upstream of R98C, but before the premature stop codon 55 amino acids downstream, this change is unlikely to be pathogenic. It is possible that this genotype may underlie EKV in this patient, as C169W/154del4 in GJB4 was not detected in our control population. We are unable to get wild-type Cx30.3 to traffick (Di et al., 2005Di W.L. Gu Y. Common J.E.A. Aasen T. O'Toole E.A. Kelsell D.P. Zicha D. Patterns of connexin interactions in keratinocytes revealed morphologically and by FRET analysis.J Cell Science. 2005; 118: 1505-1514Crossref PubMed Scopus (40) Google Scholar) and functional studies were therefore not performed with these mutants or M190L (EKV2). The patient (EKV4) developed symmetrical cutaneous lesions in both axillae at an age of 2 y. Transient patches of erythema on the neck and trunk lasting 2–3 wk were described. On examination, she had symmetrical, lichenified, erythematous plaques in both axillae and on the neck with well-demarcated borders (Figure 5). She had a mild diffuse keratoderma affecting her palms and a focal keratoderma of the soles. Although all had clinical features of erythrokeratoderma with a similar phenotype to EKV1 with or without PPK (see Table II), three other families and one sporadic case with classical features of EKV did not harbor a mutation (or rare polymorphism(s)) in the GBJ3 or GBJ4 genes. EKV, erythrokeratoderma variabilis. The histological features of EKV are non-specific. Biopsies were available from six of the eight probands with EKV investigated in this study. Almost all biopsies showed acanthosis of the epidermis with hyperkeratosis and variable parakeratosis. Hypergranulosis of the epidermis was noted in both sporadic cases of EKV with a predominant flexural phenotype (EKV2 and EKV4). As retinoic acid has been shown to downregulate the expression of certain Cx including Cx31 (Grummer et al., 1996Grummer R. Hellmann P. Traub O. Soares M.J. el-Sabban M.E. Winterhager E. Regulation of connexin31 gene expression upon retinoic acid treatment in rat choriocarcinoma cells.Exp Cell Res. 1996; 227: 23-32Crossref PubMed Scopus (22) Google Scholar), the response of this patient cohort to retinoids was also assessed and findings summarized in Table II. Four of six patients treated had a good to excellent response to oral acitretin. Two of the patients had complete clearance of their skin disease (indicated by ++), for example, the skin disease in EKV8 resolved completely on acitretin 20 mg daily. The others had some residual skin disease on treatment, particularly on the legs, for example, EKV1. We report in this study four distinct point mutations in GJB3 and GJB4 encoding Cx31 and Cx30.3, respectively, three of which are novel, in two families and two sporadic cases of EKV. In our patient cohort, one sporadic case and three families did not have detectable mutations in GJB3 or GJB4 confirming the heterogeneity of this disorder. In addition, no mutations in these EKV patients were detected in the two other epidermal disease-associated Cx genes (GJB2 and GJB6: data not shown). The differential diagnosis of EKV includes Greither's disease, keratolytic winter erythema, non-bullous ichthyosiform erythroderma, and cyclic ichthyosis with epidermolytic hyperkeratosis (Richard and Ringpfeil, 2003Richard G. Ringpfeil F. Ichthyoses, erythrokeratodermas and associated disorders Vol. 1.in: Bolognia J.L. Jorizzo J.L. Rapini J.L. Dermatology. Mosby, London2003: 799-801Google Scholar). None of our cases fitted into these clinical categories. Our two patients with mutations in GJB3 (EKV1 and 3) had similar clinical phenotypes with hyperkeratosis which was most marked on the limbs, transient erythema and sparing of the central back. A very similar phenotype was seen in patients/families EKV5-8 although no mutation was found. LikeRichard et al., 2003Richard G. Brown N. Rouan F. et al.Genetic heterogeneity in erythrokeratodermia variabilis: Novel mutations in the connexin gene GJB4 (Cx30.3) and genotype-phenotype correlations.J Invest Dermatol. 2003; 120: 601-609Crossref PubMed Scopus (102) Google Scholar, there was no specific clinical characteristics that differentiated these patients from patients with mutations in GJB3 or GJB4. It has been postulated that the GJB4 mutations (e.g., T85P and F137L) associate with the phenotype of EKV with circinate or gyrate lesions (Hohl, 2000Hohl D. Towards a better classification of erythrokeratodermias.Br J Dermatol. 2000; 143: 1133-1137Crossref PubMed Scopus (22) Google Scholar; Richard et al., 2003Richard G. Brown N. Rouan F. et al.Genetic heterogeneity in erythrokeratodermia variabilis: Novel mutations in the connexin gene GJB4 (Cx30.3) and genotype-phenotype correlations.J Invest Dermatol. 2003; 120: 601-609Crossref PubMed Scopus (102) Google Scholar). In our study, the one patient (EKV2) with a bonafide GJB4 mutation (M190L) showed no clinical signs of this variant form. This patient had an unusual clinical phenotype with persistent plaques largely confined to the flexures and transient targetoid erythema on the trunk. Interestingly, a further patient (EKV4) with rare sequence variants in GJB4 had a similar clinical phenotype with persistent plaques confined to the flexures. EKV generally responds well to retinoids. We have noted that when retinoids are stopped, the initial location for recurrence of lesions is usually the flexures. These two patients may therefore represent a more localized, flexural variant of EKV. The G12D mutation is in the cytoplasmic NT domain of Cx31, a highly conserved region that can be expected to alter the function of Cx31, as the NT domain is thought to be important in both trafficking of the protein from the Golgi apparatus to the plasma membrane and voltage gating. In vitro studies have shown that (G12D)Cx31 protein has a trafficking defect after transfection into keratinocytes (Di et al., 2002Di W.L. Monypenny J. Common J.E. et al.Defective trafficking and cell death is characteristic of skin disease- associated connexin 31 mutations.Hum Mol Genet. 2002; 11: 2005-2014Crossref PubMed Scopus (66) Google Scholar). The mutation M190L in Cx30.3 occurs within the fourth transmembrane (TM) domain of GJB4 and is conserved across at least 16 human Cx. The hydrophobic methionine residue within the TM domain is likely to play a crucial role in maintaining the open state lining of the channel pore (Richard et al., 1998Richard G. Smith L.E. Bailey R.A. et al.Mutations in the human connexin gene GJB3 cause erythrokeratodermia variabilis.Nat Genet. 1998; 20: 366-369Crossref PubMed Scopus (287) Google Scholar). Unfortunately, we are unable to get wild-type Cx 30.3 to traffick (Di et al., 2005Di W.L. Gu Y. Common J.E.A. Aasen T. O'Toole E.A. Kelsell D.P. Zicha D. Patterns of connexin interactions in keratinocytes revealed morphologically and by FRET analysis.J Cell Science. 2005; 118: 1505-1514Crossref PubMed Scopus (40) Google Scholar), so transfection studies were not performed using this mutant or the rare sequence variants seen in EKV4. Interestingly,Plantard et al., 2003Plantard L. Huber M. Macari F. Meda P. Hohl D. Molecular interaction of connexin 30.3 and connexin 31 suggests a dominant-negative mechanism associated with erythrokeratodermia variabilis.Hum Mol Genet. 2003; 12: 3287-3294Crossref PubMed Scopus (41) Google Scholar recently showed that the Cx30.3 mutation F137L prevents the trafficking of Cx31 and Cx30.3 to the membrane of HeLa cells and leads to a decrease in intercellular communication indicating an important interaction between Cx31 and Cx30.3 in the formation of gap junctions. These findings provide a novel explanation for the similarity in clinical phenotypes between patients with Cx30.3 and Cx31 mutations, reported in the literature. The substitution of a leucine at codon 209 of Cx31 (L209F) with the aromatic, highly hydrophobic phenylalanine is likely to alter the carboxyterminus of Cx31. Functional analysis of this mutation has revealed that like other Cx31 EKV mutations, expression results in a primarily cytoplasmic localization with no gap junction plaques observed between two adjacent cells. We have previously shown that the keratinocyte cell line used in this study, NEB1, expresses low levels of endogneous Cx31 with poor basal dye transfer. Dye is transferred, however, at a steady rate in wild-type Cx31-expressing cells suggesting endogenous protein levels are sufficiently low not to influence the effects of the exogenous protein (Di et al., 2002Di W.L. Monypenny J. Common J.E. et al.Defective trafficking and cell death is characteristic of skin disease- associated connexin 31 mutations.Hum Mol Genet. 2002; 11: 2005-2014Crossref PubMed Scopus (66) Google Scholar). This trafficking defect of the mutant protein is likely to be a contributing factor in the disease phenotype of the patients in these families. Impaired gap junction formation and communication may have detrimental consequences on epithelial differentiation and keratinocyte survival in EKV patients (Di et al., 2002Di W.L. Monypenny J. Common J.E. et al.Defective trafficking and cell death is characteristic of skin disease- associated connexin 31 mutations.Hum Mol Genet. 2002; 11: 2005-2014Crossref PubMed Scopus (66) Google Scholar; Diestel et al., 2002Diestel S. Richard G. Doring B. Traub O. Expression of a connexin31 mutation causing erythrokeratodermia variabilis is lethal for HeLa cells.Biochem Biophys Res Commun. 2002; 296: 721-728Crossref PubMed Scopus (38) Google Scholar; Rouan et al., 2003Rouan F. Lo C.W. Fertala A. et al.Divergent effects of two sequence variants of GJB3 (G12D and R32W) on the function of connexin 31 in vitro.Exp Dermatol. 2003; 12: 191-197Crossref PubMed Scopus (26) Google Scholar). Other GJB4 rare polymorphisms have been detected in patients with skin disease. The heterozygous change R103C in GJB4 was found in a patient with focal PPK and was not detected in our control population. This genotype is unlikely to be the sole cause of the disease phenotype as it has been reported before in an individual without a dermatological condition (Lopez-Bigas et al., 2002Lopez-Bigas N. Melchionda S. Gasparini P. Borragan A. Arbones M.L. Estivill X. A common frameshift mutation and other variants in GJB4 (connexin 30.3): Analysis of hearing impairment families.Hum Mutat. 2002; 19: 458Crossref PubMed Scopus (19) Google Scholar). The homozygous change E204A was observed in a patient with a PPK and a conductive deafness. This form of deafness is not likely to be associated with Cx proteins that cause sensorineural hearing impairment. The PPK, however, could be associated with homozygosity for E204A in GJB4 as the E204A allele was only seen in the control population (3/362) in heterozygosity. In conclusion, this study describes a genetic and clinical investigation of the erythrokeratodermas that highlights the importance of Cx mutations in a subset of EKV cases. A number of EKV cases, however, are clearly because of other gene mutations. Further genetic studies in large families with EKV with or without GJB3 or GJB4 mutations are required to delineate further the molecular and clinical classification of the erythrokeratodermas. We studied 12 affected individuals from five pedigrees and three sporadic cases of erythrokeratoderma. Genomic DNA was extracted from the venous blood of the affected and available unaffected family members (Nucleon, Manchester, UK). The clinical status of each individual was established after examination by at least two dermatologists. Seven of the eight index cases were seen in a single center. Genomic DNA collected from 362 unrelated individuals was used as a control population. For those controls with known ethnicity, the frequency of polymorphisms in different ethnic groups was also determined. Ethical permission for this project was obtained from the local ethics committee of the East London Clinical Health Authority. Patient consent was given for this study. The study was conducted according to Declaration of Helsinki Principles. We amplified two overlapping fragments from genomic DNA that covered the entire coding sequence of human Cx31 (GJB3) using two primer pairs (Table III). These PCR products were analyzed by denaturing high-performance liquid chromatography (DHPLC; Transgenomic, Crewe, UK) and sequenced. A similar strategy was employed for the analysis of GJB4 with two overlapping fragments covering the entire coding regions using the overlapping oligonucleotide primers pairs (Table III). The DNA fragments were amplified under the following PCR conditions: 35 cycles of 30 s at 94°C, 30 s at 62°C and 30 s at 72°C. The PCR product was purified using a QIAquick PCR purification kit (QIAgen Ltd, Crawley, UK) and directly sequenced with v3.1 Big-Dye terminator system and analyzed on a ABI 377 automated sequencer (PE Applied Biosytems, Warrington, UK). Sequence traces were analyzed by eye and with Sequence Navigator v1.01 software (Applied Biosystems, Foster City, California). For heteroduplex formation and DHPLC analysis, all PCR products were heated to 95°C for 5 min followed by cooling at -0.03°C per s to a final temperature of 40°C. DHPLC was performed using a WAVE DNA Fragment analysis system (Transgenomic). Subcloning of the PCR product for the GJB4 fragment amplified from the genomic DNA from EKV9 was performed using the Topo cloning kit (Invitrogen, Paisley, UK) after gel extraction using the QIAquick gel extraction kit (QIAgen Ltd). After selection of positive colonies and mini-preparations (QIAgen Ltd), cloned DNA was sequenced with T7 and SP6 oligonucleotide primers.Table IIIOligonucleotide primer sequencesPrimer nameNucleotide sequenceGJB3 F5-GCCTGGTACATAGTAAATGC-3GJB3 R25-CTGAACAGGTAGGTCCACCA-3GJB3 F25-GCCGGCACCGCCAGAAACAC-3GJB3 R5-CAGCCCCTGTAGGACCTCTCC-3GJB4 F5-GTCAATCGCACCAGCATTAAG-3GJB4 R25-AGCAAGTACGTCCACCACAGTC-3GJB4 F25-GGTGGACGTACTTGCTGAGCC-3GJB4 R5-TACCCACCTGCATCCACTGG-3(L209F)Cx31 FGTGAGCTCTGCTACTTCATCTGCCACAGGG(L209F)Cx31 RCCCTGTGGCAGATGAAGTAGCAGAGCTCACCx, connexin. Open table in a new tab Cx, connexin. Forward and reverse oligonucleotides were generated with Sal1 and HindIII restriction sites, respectively, for inframe cloning into pEGFP-N3. These primers allowed the amplification of the whole-length gene of interest from normal genomic DNA as checked against public published sequence data (NM_024009). Primers used and characterization of this clone has already been published (Di et al., 2002Di W.L. Monypenny J. Common J.E. et al.Defective trafficking and cell death is characteristic of skin disease- associated connexin 31 mutations.Hum Mol Genet. 2002; 11: 2005-2014Crossref PubMed Scopus (66) Google Scholar). Site-directed mutagenesis was carried out according to manufacturers instructions and as previously described for other Cx31 mutations (Stratagene, La Jolla, California) (Di et al., 2002Di W.L. Monypenny J. Common J.E. et al.Defective trafficking and cell death is characteristic of skin disease- associated connexin 31 mutations.Hum Mol Genet. 2002; 11: 2005-2014Crossref PubMed Scopus (66) Google Scholar)). Briefly, an EGFP plasmid containing h(WT)Cx31 was manipulated with oligonucleotide primers containing the desired mutation (L209F)Cx31-EGFP in both the forward and reverse primer (Table III). Transfection was performed with the lipofectin-based Transfast reagent according to the manufacturers instructions (Promega, Southampton, UK). Plasmid DNA at a 1:2 ratio with Transfast was added to NEB1 keratinocytes (Morley et al., 1995Morley S.M. Dundas S. James J. et al.Temperature sensitivity of the keratin cytoskeleton and delayed spreading of keratinocyte lines derived from EBS patients.J Cell Sci. 1995; 108: 3463-3471Crossref PubMed Google Scholar) seeded on 13 mM coverslips for an incubation time of 2 h at 37°C. Cells were fixed in 4% paraformaldyhyde 48 h post-transfection. Coverslips were mounted in antifade containing DAPI (Vector Labs, Peterborough, UK) to counterstain cell nuclei before confocal imaging (Zeiss, Orberkochen, Germany). This work was supported by grants from the Research Advisory Board of Barts and the Royal London and from the BBSRC (D. P. K).