¹Epithelial Genetics Group, Human Genetics Unit, Division of Pathology and Neuroscience, University of Dundee, Ninewells Hospital and Medical School, Dundee, United Kingdom

Correspondence: Prof. W.H. Irwin McLean, Human Genetics Unit, Ninewells Hospital and Medical School, Dundee DD1 9SY, United Kingdom. E-mail w.h.i.mclean@dundee.ac.uk

Abstract

Evidence is now emerging for enhanced penetration of chemical solutes into uninvolved skin of atopic dermatitis patients. Along with the recent discovery of prevalent null mutations in the gene encoding filaggrin, a protein essential for stratum corneum formation, these data point to an innate epidermal-barrier defect in atopy.

Atopic dermatitis (AD, "eczema") has increased in prevalence in recent decades; it now affects 15%–20% of children in the developed world alone and affects all races to differing degrees. The disease is a classical complex trait — it is clearly highly heritable but with reduced penetrance, and there is strong evidence for environmental influences (Morar et al., 2006). Furthermore, the disease impacts on other areas of medicine, as a high percentage of eczema patients develop a range of additional allergic conditions, including food allergies, asthma, and rhinitis, often occurring in a temporal "program" known as the atopic march (Spergel and Paller, 2003). AD causes considerable morbidity and significantly affects quality of life. This is particularly the case for severe AD persisting into adulthood. Establishing the contributing etiological factors for this complex disease — both genetic predisposing factors and environmental trigger factors — is of paramount importance in the development of new, possibly more effective treatments and/or preventative regimens. Given that the skin is inflamed within an AD lesion, there is an obvious immunological component in the disease; however, recent evidence coming from both molecular genetics and functional analysis strongly suggests that a skin barrier defect is the necessary cause in a considerable proportion of atopy cases.

Although there is considerable functional evidence for impaired skin barrier function within active lesional skin of AD patients, as seen, for example, by increased trans-epidermal water loss, the situation has been less clear in studies of uninvolved skin, where the results have been somewhat contradictory. Ivone Jakasa and colleagues in Amsterdam and Rome (2007) present conclusive, statistically significant evidence for enhanced uptake of an entire series of polyethylene glycols, covering molecular weights in the range 150–590daltons, in non-lesional skin of AD patients. This result is further supported by recently published evidence from the same research group for enhanced uptake of sodium lauryl sulfate in non-lesional AD skin, based on the use of similar methods (Jakasa et al., 2006). Genji Imokawa's group and collaborators in Japan, using a photoacoustic spectography system, showed enhanced penetration of both a lipophilic and a hydrophilic dye through clinically normal skin of AD patients as compared with control subjects. In addition, they showed a significant correlation between penetration rates for the hydrophilic dye and elevated IgE levels for patients with severe AD (Hata et al., 2002). Together, these reports strongly support the hypothesis that patients with AD have an inherent skin barrier defect, probably due to one or a combination of genetic mutations, whether or not a particular area of their skin has active eczema at any given time, implying that this is the main predisposing factor for the development of AD.

Recently, it has been shown that null mutations in the filaggrin gene are the cause of ichthyosis vulgaris (Smith et al., 2006) and that these same mutations are highly prevalent in some European populations where up to 9%–10% of individuals are heterozygous carriers, that is, they are haploinsufficient for filaggrin. It was subsequently shown that these mutations are also an important genetic predisposing factor for AD (Palmer et al., 2006). Further studies have now confirmed the strong association of filaggrin-null alleles and AD, as well as the other allergic conditions that make up the atopic march. At the time of writing, the papers published so far report 15 forms of replication for this association, and so the AD–filaggrin story is already unusually robust in the field of complex-trait genetics. Profilaggrin is the major protein of which keratohyalin granules are composed, and when this highly abundant polyprotein is proteolytically cleaved in the uppermost granular cells, the liberated filaggrin peptides aggregate the keratin cytoskeleton and bring about cell compaction to make the squames. Therefore, this gene–protein system is critical in the formation of the stratum corneum, the cell compartment within which the epidermal-barrier function resides. Haploinsufficiency for filaggrin in mutation carriers, or complete loss in homozygotes or compound heterozygotes, leads to impaired barrier formation, which manifests as varying degrees of dry skin, ichthyosis, and/or eczema. A further function of filaggrin has also been suggested, whereby it may be broken down in the stratum corneum into hygroscopic amino acids and derivatives thereof that act as a "natural moisturizing substance" (Rawlings and Harding, 2004). Thus, filaggrin deficiency may spell double trouble — impaired formation of the stratum corneum as well as a lack of the skin's built-in humectant.

The extracellular lipid lamellae also play an important role in skin barrier function. Ceramides are the most abundant of the lipids in the stratum corneum. In AD, ceramide levels are reported to be decreased and lipid packaging altered (Imokawa et al., 1991; Pilgram et al., 2001). Sphingomyelin deacylase activity has been found to be increased in AD (Hara et al., 2000). This enzyme hydrolyzes sphingomyelin to sphingosylphosphorylcholine and free fatty acid instead of ceramide. This is suggested to explain the decreased ceramide in AD. Is there possibly a link between filaggrin and lipid metabolism? The analysis of epidermal lipids in subjects of known filaggrin genotype may shed light on this important aspect of the skin barrier in eczema.

Here, Jakasa and co-workers (2007) did not genotype their small AD cohort for filaggrin-null variants, as their current study predates the publication of the role of this gene in AD. Interestingly, they excluded AD patients who had concomitant signs of ichthyosis vulgaris and so may have eliminated some or all filaggrin-null carriers from their cohort. It will be fascinating to combine the two functional assays they have developed for measuring epidermal solute transfer with filaggrin genotyping in future AD patient cohorts. Because the experiments reported here showed a breach in skin barrier function in all the AD patients examined as compared with control subjects, one would predict that the AD patients who do not carry filaggrin mutations would still have impaired barrier function. It will also be interesting to see whether combining genetic testing with solute uptake measurement methods can help stratify AD. The strongest filaggrin associations are currently being observed in patients with hospital- or dermatologist-diagnosed eczema, which represents the moderate-to-severe end of the disease spectrum. Do these patients with more severe AD have consistently worse barrier impairment, or does all AD behave similarly in terms of barrier function regardless of filaggrin status?

Another interesting question concerns looking at bigger AD cohorts with the use of Jakasa and colleagues' (2007) methodology. Are there AD patients out there who do not have an obvious barrier impairment? The study by Jakasa et al. (2007) suggests that this may not be the case, but a larger study is needed to address this, especially one in which patients with mild, moderate, and severe eczema are included in reasonable numbers. This might tell us whether or not there is a proportion of AD in which barrier function is not so important and in which perhaps a primarily immunological defect predominates. The current feeling in the AD field is that a barrier defect is a sufficient cause of AD, but one should not dismiss a primary immunological/allergy hypothesis out of hand. Genome-wide linkage studies have shown that AD susceptibility is heterogeneous, and although filaggrin may be important in a good proportion of more severe AD, it is certainly not the only gene that will be linked to this complex trait. With the recent completion of the human genome sequence, the haplotype map, and advances in high-throughput sequencing and typing of polymorphisms, we will hopefully see the emergence of other consistently replicable atopy gene associations in the near future. Once the genetic picture becomes more complete, the environmental trigger factors that have led to the obviously increased incidence of AD in recent decades need to be addressed. Then, the next barrier to breach will be that of translating all this knowledge into more rational, and hopefully more effective, treatments for eczema and related atopic conditions.