Clinical and molecular significance of splice site mutations in the plakophilin 1 gene in patients with ectodermal dysplasia-skin fragility syndrome.

In 1997, McGrath et al. described the first case of a rare autosomal recessive inherited skin disease resulting from loss-of-function mutations in plakophilin 1 (PKP1), a component of desmosomes, cell–cell junctions found primarily in epithelial tissues (1). Plakophilin 1 is a member of the armadillo family of structural and signalling proteins and it contributes to the mechanical integrity and calcium stability of desmosomes (2). Clinically, loss of plakophilin 1 leads to trauma-induced skin fragility as well as developmental anomalies of hair and nails. These previously undocumented observations led to the introduction of a new term, ectodermal dysplasia-skin fragility syndrome (OMIM 604536), to describe such patients, and following the initial publication, six additional patients with PKP1 gene mutations (from the UK, France, Japan, Israel and The Netherlands) were reported (3–7). In this issue of Acta Dermato-Venereologica, Zheng et al. present the clinical and molecular abnormalities of a further example of this genodermatosis, the first Chinese case to be described (p. 394, ref. 8). This new case of ectodermal dysplasia-skin fragility syndrome is a compound heterozygote for two splice site mutations in PKP1, c.[1053TwA + 1054+1GwT] affecting the intron 5 donor splice site and an intron 10 acceptor splice site IVS10-2AwG. In general, splice site mutations account for about 15% of all diseaseassociated gene mutations, but in this syndrome they appear to be a much more common phenomenon. Specifically, thus far 13 of a possible 16 mutated alleles have involved splice site mutations of the PKP1 gene, but just why the molecular pathology of ectodermal dysplasia-skin fragility syndrome should involve so many splice site mutations is difficult to fathom. Moreover, predicting the consequences of splice site mutations on RNA processing is also an uncertain matter and this has implications for establishing a paradigm for genotype-phenotype correlation and for providing accurate genetic counselling for patients and their families. Of the reported patients with this desmosomal genodermatosis, keratinocyte cDNA has been examined in five cases (1, 3, 6–8). These analyses have demonstrated the protean effects of the splice site mutations with evidence of cryptic splice sites, exon skipping and retention of introns. Furthermore, some of these studies have implications for genotype-phenotype correlation. For example, RT-PCR across the PKP1 splice site mutation IVS9+1GwA was shown to result in at least two different transcripts, one of which involved activation of a cryptic donor splice site within exon 9 leading to an in-frame deletion of 15 amino acids (7). This finding provided an explanation for the observation of some detectable plakophilin 1 protein in the skin by immunohistochemistry and for the somewhat milder clinical phenotype (e.g. relative lack of alopecia) compared with several of the other published cases. In the new report of ectodermal dysplasia-skin fragility syndrome in this issue, Zheng et al. also attempted to assess the effects of compound heterozygosity for their two new splice site mutations in PKP1. Unfortunately, they were unsuccessful in attempting to identify any PKP1 transcripts using RNA recovered from formalinfixed paraffin-embedded skin. This probably reflects very low levels of PKP1 mRNA, as b-actin could still be amplified from their patient’s skin. Nevertheless, despite the lack of information from this mRNA analysis, it is now possible to make in silico predictions about the possible consequences of the splice site mutations c.[1053TwA + 1054+1GwT] and IVS102AwG. Specifically, over recent years, a number of computational tools have become available to assess mechanisms of normal and aberrant splicing. These include Automated Splice Site Analyses (https://splice. cmh.edu/index.html) or Splice Site Prediction by Neural Network (http://www.fruitfly.org/seq_tools/splice.html). Using the Delila software package (9), which scans genomic DNA sequences with weight matrices for sites with positive Ri (individual information content) values, as well as the Sequence Walker programme to display the results (10), it is possible to predict the consequences of the PKP1 splice site mutations detected by Zheng et al. These analyses are illustrated in Fig. 1. For the mutation c.[1053TwA + IVS5+1GwT], software analysis predicts that the TwA nucleotide substitution in the penultimate nucleotide of exon 5 actually strengthens the pre-existing donor splice site (Ri increased from 6.2 to 8.4 bits). In contrast, the IVS5+1GwT change reduces the information content of the natural donor splice site from 6.2 to 21.6 bits. The combination of both substitutions, therefore, abolishes the natural donor splice site (Ri decreased from 6.2 to Acta Derm Venereol 2005; 85: 386–388