Epidermolysis bullosa (EB) constitutes a group of genodermatoses manifesting with fragility of the skin and mucous membranes and presenting with blisters and erosions at birth or shortly thereafter.1,2
The spectrum of phenotypic manifestations is broad; in the milder forms there is a life long blistering tendency with no impact on the overall longevity of the affected individual, while in the most severe forms children die during the early postnatal period from metabolic perturbations, dehydration, and sepsis. Some forms are characterised by debilitating scarring with the propensity to early death from aggressive squamous cell carcinomas of the skin. In addition to skin and mucous membrane involvement, there are a number of extracutaneous manifestations. In some cases, abnormalities in the hair, nails, and teeth can be described. In other forms of the disease, the gastrointestinal tract is affected either in the form of oesophageal strictures or congenital pyloric atresia. A subset of patients manifests with late onset muscular dystrophy. Adding to the heterogeneity of EB is the fact that inheritance can be either autosomal dominant or autosomal recessive. This heterogeneity, coupled with historical classifications riddled with eponyms, has led to suggestions that there may be as many as 30 different subtypes of EB.3
EB is now known to result from mutations in 10 different genes expressed within the cutaneous basement membrane zone (BMZ) at the dermoepidermal junction (fig 1).2
The stratified expression pattern of these genes within the BMZ and in extracutaneous tissues explains the broad phenotypic spectrum of EB. Furthermore, the types and combinations of the mutations and their consequences at the mRNA and protein levels, when superimposed on the individual's genetic background and combined with environmental factors, contribute to the wide range of phenotypes.
Figure 1Complexity of the cutaneous basement membrane zone (BMZ), and classification of epidermolysis bullosa. The figure schematically depicts basal keratinocytes at the lower part of the epidermis, separated from the papillary dermis by a dermoepidermal (more ...)
Traditionally, EB has been divided into three broad categories based on the level of tissue separation, determined by diagnostic electron microscopy and/or immunoepitope mapping (fig 1):4
) the simplex forms of EB (EBS) demonstrate tissue separation within the basal keratinocytes at the bottom layer of epidermis; (b
) the junctional forms of EB (JEB) display tissue separation within the dermoepidermal basement membrane, primarily within the lamina lucida; and (c
) in the dystrophic forms (DEB), tissue separation occurs below the lamina densa within the upper papillary dermis (fig 1). In addition to this traditional classification, a fourth subtype, the hemidesmosomal variants of EB (HEB), has been proposed (table 1).5
Table 1Subclassification of EB with associated mutant genes*
Patients with HEB manifest tissue separation at the basal cell/lamina lucida interface at the level of the hemidesmosomes. While there is overlap between the hemidesmosomal variants and the traditional subtypes, particularly the simplex and junctional forms of EB, the incorporation of HEB into the classification scheme has been extremely helpful in guiding initial mutation detection strategies in this subgroup of patients with EB; it should be noted that this addition to the EB classification has not been recognised by the latest International Consensus Meeting.4
However, the general classification, consisting of EBS, HEB, JEB, and DEB, combined with information derived from diagnostic immunoepitope mapping of the skin in affected individuals, has provided the platform for successful and streamlined mutation detection in the candidate genes in these families.
Certain subtypes of EB are associated with mutations in specific genes, and examination of the mutation database has revealed general genotype‐phenotype correlations. For example, the classic, dominantly inherited, simplex forms are primarily due to mutations in the intermediate keratin filament genes, KRT5 and KRT14, expressed in the basal cells of the epidermis. The dystrophic forms are exclusively due to mutations in the type VII collagen gene, COL7A1. The classic junctional forms are associated with mutations in the laminin 5 genes, LAMA3, LAMB3, and LAMC2; in particular, the Herlitz (lethal) JEB variant frequently displays premature termination codon (PTC) mutations in these genes. The non‐Herlitz (non‐lethal) variant harbours mutations in the same genes, but with milder consequences at the mRNA and protein levels, such as missense or splice junction mutations.
The hemidesmosomal clinical variants have been proposed to include generalised atrophic benign EB (GABEB), which was previously included in the non‐Herlitz category of JEB. Most cases of GABEB have been reported to be caused by mutations in the type XVII collagen/180 kDa bullous pemphigoid antigen gene, COL17A1
although this study provides compelling evidence for mutations in the laminin 5 genes in many of these cases. Another hemidesmosomal variant is EB with pyloric atresia (EB‐PA). While most of these cases are associated with mutations in the α6β4 integrin genes, ITGA6
mutations in the plectin gene (PLEC1
), which encodes a large, ~500 kDa adhesion protein, have also been identified in some patients with EB‐PA.6,7
Finally, mutations in the plectin gene have been identified in patients with EB and late onset muscular dystrophy.8
As the tissue separation in the latter cases is intraepidermal within the basal keratinocytes, patients with EB‐MD have also been included in the simplex category, although tissue separation occurs at a different level in these patients than in those with classic forms of EBS.
The DebRA Molecular Diagnostics Laboratory, established in 1996 at the Department of Dermatology and Cutaneous Biology at Jefferson Medical College, has collected over 1000 families with different forms of EB for analysis for mutations in the candidate genes. Among these families, 422 distinct mutations in the 10 genes underlying the major forms of EB have been found. This number does not count homozygous mutations twice or recurring mutations multiple times. Thus, there is a large number of private, family specific mutations in EB with general genotype‐phenotype correlations and translational implications for prognostics, genetic counselling, and prenatal/preimplantation genetic testing. In this article, we summarise results of mutation analysis in a cohort of 265 patients with the junctional and hemidesmosomal forms of EB.