Progress Report

Mutation identification in EBS patients within the UK

Identification of keratin mutations in EBS patients has been extended. Most of the referrals from within the UK are now dealt with routinely within the clinical genetics unit at Ninewells Hospital, by analysis of the hotspots of the genes for keratin proteins K5 and K14, where about 70% of the EBS mutations are found. However we have continued to extend this analysis to full sequencing of K5/K14 in any cases that show up as negative in the service lab. In this way, together with the unit at Ninewells we have raised the success rate of finding the mutation in EBS patients to 78% or more.

Extending the mutation analysis to the USA

In collaboration with Dr Jo-David Fine, one of the world’s clinical experts on EBS addition, we have additionally begun to extend mutation analysis to EBS patients from the USA. A cohort of patients diagnosed with different forms of epidermolysis bullosa simplex, from 39 unrelated families on the US National EB Registry, has been analysed in our laboratory for mutations in the coding regions of the K5 and K14 genes. Mutation analysis has so far identified 29 mutations. We then started to compare the range of mutations found with those identified to date in European patient groups, for which a much larger body of data exists, as can be seen on the Dundee website of intermediate filament mutations (www.interfil.org). Amongst the 29 USA mutations in K5/K14, we detected 13 novel mutations that had not been reported so far in the European populations. This included some unusual double heterozygote mutations in which the patient carries two separate mutations in the same keratin, one on each of the two alleles of one gene. This information is adding to the global database on EBS mutations and will help us understand the variety of phenotypes seen in these diseases.

Functional analysis of new mutations

The effects on keratin function of the new mutations observed in the USA group, plus some of the new mutations recently found in the U.K. patient group, is now being examined by recreating the mutations in cultured cell lines whose properties we can then test. In particular we are looking for information about the relative ability of different mutations to affect keratin filament formation and stability.

The cDNA for K5 and K14 has been cloned, and the new mutations have been introduced into these clones by site-directed mutagenesis. These mutant keratin cDNAs have then been introduced into immortalized epithelial cell culture lines. Each mutant keratin sequence tested so far has been co-transfected together with its natural partner keratin into cultured cell lines, using a cell type which is lacking intrinsic keratin filaments. Keratin sequences need to be expressed in the presence of a partner keratin as they will only make keratin filaments from dimeric K5-K14 subunits: one keratin protein alone cannot polymerise. We chose cells lacking their own keratin filaments to be the recipients of these experimental mutant keratins, because if there had been any endogenous keratin protein, even a severely mutant keratin could assemble onto such a filament and we would learn nothing about the effect of the mutation. Thus we have set up the experiment in such as way that if any filaments are formed, both of the transfected keratins must be contributing to those filaments, and in equal amounts.

The transfected cells have then been analysed morphologically to assess the ability of the mutant protein to form keratin filament networks. Without subjecting the cells to any additional stress assay, a very clear distinction has been seen already between the mutations identified in patients with "severe" EBS (in which the mutant keratin is unable to support filament network formation) versus those with "mild" disease (in which mutant keratin-containing filaments are indistinguishable from wild type ones). This suggests that the "mild" mutations exert a fundamentally different effect on the keratin filament structure from the effects of "severe" mutations. We now intend to take these experiments further and try to understand what the difference is in these effects.

Clinical implications of these findings

We have then gone back to the clinical notes on the "severe" and "mild" mutations. It seems clear that the "polymerisation-incompetent" mutants were all of a type which would probably have been diagnosed as Dowling-Meara type EBS in the UK, but some of these had been classified as Koebner EBS according to current USA practice. This highlights an issue that requires discussion between USA and UK clinicians to reach a consensus and we are currently trying to organise a meeting to discuss this. The results so far do suggest that the boundary between EBS subtypes could now be usefully re-aligned to conform to molecular criteria that are now easier to identify routinely and objectively and for which there is now a molecular-based hypothesis for the phenotypic difference.