Anne M. Bowcock and William O.C.M. Cookson
Psoriasis and atopic dermatitis are chronic and relapsing inflammatory diseases of the skin associated with various immunologic abnormalities. Approximately 30% of psoriasis patients also have joint involvement, indicative of psoriatic arthritis. Genes and environment play a key role in the pathogenesis of these diseases. Genome-wide linkage scans have identified multiple loci linked to each disease and revealed overlap with psoriasis and atopic dermatitis susceptibility loci on chromosomes 1q21, 3q21, 17q25 and 20p12. The genes from these loci have not yet all been identified, or systematically tested for a role in psoriasis and atopic dermatitis; however, these locations suggest that some susceptibility factors lie within genes or gene families with common effects upon epithelial immunity. A strong HLA association is described for psoriasis, but not for atopic dermatitis. Knowledge of the genetic factors leading to these diseases will lead to an understanding of their variable age at onset, their waxing and waning and the variability of body surface environment. The effect of environmental triggers may also be understood once the altered pathways are elucidated. Genes implicated so far in atopic dermatitis are SPINK5, FcεRI-β and PHF11. Genes implicated in psoriasis so far are HLA-C, SLC9A3R1, NAT9, RAPTOR and SLC12A8. Genetic modifiers such as CARD15 may predispose to psoriatic arthritis.
The term psoriasis is derived from the Greek (psora: scurf, itch, rash). The first clear description was by Willan in 1808 (1). The eponymic designation of this dermatologic condition was proposed by Russel in 1950 (2). Types of psoriasis include plaque, guttate, erythrodermic and pustular (3).
Psoriasis has a worldwide distribution with prevalence varying according to race and geographical location. It is commonest in Scandinavia and Northern Europe where it approaches 3%. In North America and the UK its prevalence is ∼2%. In Japan, prevalence is ∼0.2% of the population, and in Native American Indians it is rare (4). In the USA 3 million office visits for psoriasis are made each year, costing over $3 billion. The concordance of psoriasis in monozygotic twins is 65–72%, versus 15–30% in dizygotic twins. Determination of concordance in older twin pairs from a national twin registry in Denmark revealed nearly 90–100% heritability (5). In an Australian study the monozygotic twin concordance rate is lower (35% for monozygotic twins and 12% for dizygotic twins), giving an estimated heritability of 80% (6).
Males and females are equally affected and 75% of patients develop the disease before the age of 40. First manifestations of the disease are most common in the third decade. Two peaks of age of onset have been described: one at 20–30 years and a smaller peak at 50–60 years. This has given rise to the hypothesis that two forms of the disease exist (7). However, there are exceptions to this rule since some families with early-onset and severe disease appear to be segregating a highly penetrant autosomal dominant susceptibility gene that is distinct from HLA (8,9). Psoriasis can also occur with other inflammatory diseases such as (psoriatic) arthritis in 10–30% (recent NPF survey). Psoriasis also occurs in association with human immunodeficiency virus (HIV) infection (10). It is hypothesized that psoriasis is due to a combination of genetic predisposition and environmental assaults. These can include injury, infection, stress or certain medications. One intriguing characteristic of psoriasis is the ‘Koebner phenomenon’, first reported by Heinrich Koebner in 1872. It refers to the appearance of isomorphic pathological lesions following skin trauma patients with pre-existing cutaneous diseases and is most frequent in patients of psoriasis.
Understanding the pathogenesis of inflammatory diseases such as psoriasis has not been straightforward. An examination of transcripts and peptides with altered expression levels, including a global genome-wide expression study (11), has highlighted a large number of dysregulated genes and gene clusters, particularly those involved in epithelial proliferation and in the immune system. However, these studies have not provided sufficient insights to lead to an identification of the molecular defects underlying the disease.
The normal cycle of maturation of keratinocytes is 28–30 days. In psoriasis this is accelerated to 3–4 days. The immune system has been strongly implicated in the pathogenesis of psoriasis since it resembles a T cell-mediated autoimmune disease (12). During lesion formation, inflammation precedes epidermal hyperproliferation and increased numbers of T cells have been demonstrated in the uninvolved skin of psoriatics (13). T cells isolated from involved psoriatic skin may also enhance keratinocyte proliferation. Both CD4+ and CD8+ T cells in active skin lesions are strongly polarized as Th 1 cells (Th 1 and Tc1, respectively) and there is also a significant increase in circulating type 1 T cells in most patients. Psoriasis serves as the clearest (polar) example of a type 1-deviated skin disease, while atopic dermatitis is the clearest example of a skin disease with opposite (type 2) immune deviation. This is based on frequencies of IFN-γ-producing (type 1) T-cells versus IL-4 producing (type 2) T cells in the circulation.
Indirect evidence for the role of the immune system in psoriasis has come from clinical studies. Drugs that act by suppressing the activity of T cells such as cyclosporin, FK506, and the recently developed biologics are effective in treating psoriasis (14). Yet other evidence for a T cell basis for psoriasis susceptibility has come from bone marrow transplantations, where the psoriatic status of the donor has been transmitted to the recipient (15). Other evidence has come from animal models. For example, injection of activated blood-derived T lymphocytes into SCID mice with autologous human-grafted skin has resulted in psoriatic plaques and the presence of T cells with NK cell receptors (16) that accumulate immediately before the onset of acute lesions.
These type of observations reinforce the conjecture that psoriasis is an autoimmune disease with defects in self-tolerance, although a triggering antigen has not been identified.
Atopic dermatitis (AD, eczema) is typified by itchy, inflamed skin. The disease usually begins in infancy and early childhood, and infants with AD are prone to weeping inflammatory patches and crusted areas on the face, neck, extensor surfaces and groin. Children and young adults tend to have dermatitis of flexural skin, particularly in the antecubital and popliteal fossae.
AD is increasingly common in the developed world, affecting up to 15% of children in some countries (17). The cost of treating AD is substantial (18,19), and much of this cost is born by the families of patients with the disease (18). A significant proportion of children with the disease continue with problems into adult life.
The word ‘atopy’, meaning ‘strange disease’ (20) was coined to describe the familial syndrome of asthma and hay fever. AD subsequently came to be considered to be part of the same syndrome. The atopic state is recognized by skin prick tests to common allergens, by the presence of allergen-specific IgE in their serum, and by elevations of the total serum IgE (21).
Approximately 80% of cases of childhood eczema are atopic by these criteria (22,23). Atopic mechanisms consequently dominate current understanding of the pathogenesis of the disease. However, eczema in the 20% of children without atopic manifestations is clinically indistinguishable from disease in the 80% who are atopic (23,24), and it is not clear whether disease in non-atopics is the result of different processes.
Twin studies of eczema show concordance rates of 0.72–0.86 in monozygotic and 0.21–0.23 in dizygotic twin pairs (25,26). Physician-diagnosed asthma exhibits a similar pattern, with concordance of 0.65 in monozygotic twins and 0.25 in dizygotic twins (27). Total serum IgE levels show a heritability of ∼50% (21,28). These studies indicate the presence of strong genetic factors underlying the development of atopy and atopic disease.
Although AD is clinically and pathologically quite distinct from psoriasis, some features are shared by both diseases, including dry, scaly skin and disturbed epidermal differentiation (Fig. 1). Psoriasis is characterized by infiltration of inflammatory cells into the dermis and epidermis is accompanied by hyper-proliferation of keratinocytes. The latter is not seen in AD. However, a recent gene-expression study of 12 000 transcripts indicate that most of them were similarly expressed in both diseases (29). However, inflammatory cells invading the skin in psoriasis are TH1 cells (indicated by the overexpression of IFN-γ), macrophages, dendritic cells and neutrophils while infiltrating inflammatory cells in AD are TH2 cells, eosinophils and mast cells. These cells produce IL-4, IL5, IL-10 and IL-13. It is proposed that these may be attracted to the different chemokines in the skin of patients with each disease. For example, chemokines increased two-fold in AD versus psoriatic skin are CCL13/MCP-4, CCL-18/PARC and CCL-27/CTACK. It has also been proposed that keratinocytes of AD patients have high RANTES expression in lesions (30). Chemokines increased 2-fold in psoriatic versus AD skin are reported to be CCL-4/MIP-1β, CCL20/MIP-3α, CXCL-2/GRO-β, CXCL-8/IL8 and CXCR2/IL8R, as well as MCP-1 and IP-10 (29). Our recent gene expression profiling of psoriatic skin revealed the up-regulation of 19 chemokines in psoriatic skin (11) including most of those described above. Several of these are involved in the formation of secondary lymphoid tissue and we have proposed that the combination of many CCR7+ T cells, dendritic cells and regulating chemokines in psoriatic lesions, together with the detection of dendritic cell activation markers in uninvolved skin, could sustain chronic T cell activation and persistence within focal skin regions.
In 1980 association of psoriasis with HLA class I alleles was demonstrated, with the most highly associated allele being HLA-Cw6 (32). However, the identity of the HLA class I allele driving the association (PSORS1) is currently controversial. This is likely to be due to the extensive linkage disequilibrium that exists in patients within an interval of ∼275 kb between HLA-B and a cluster of genes including HCR (alpha-helix coiled-coil rod homolog) and CDSN (corneodesmosin). Nair et al. (33) localizedPSORS1 to a 60 kb region telomeric to HLA-C (33). Others have proposedCDSN and HCR (34–36) lying in a non-overlapping region ∼150 kb away from HLA-C. Association studies with a dense set of single-nucleotide polymorphisms (SNPS) from throughout this region (37) refined PSORS1 to a 10 kb interval very close to HLA-C.
Despite the association with HLA, however, not all affected members harbor HLA-Cw6 (known as HLA-Cw*0602 when identified with DNA typing). In independent sets of affected individuals/families this allele is only found in ∼40–80% of cases. Moreover the penetrance of HLA-Cw*0602 is ∼10%, implicating environmental effects or additional genetic susceptibility factors.
Patients with psoriasis also have different clinical features depending on whether they are HLA-Cw6 positive or negative. Besides having a lower age of onset, HLA-Cw*0602 positive patients have more extensive plaques on their arms, legs and trunk, more severe disease, higher incidence of Koebner’s phenomenon, reported more often that their psoriasis got worse during or after throat infections (see Environmental Triggers) and more often had a favorable response to sunlight. In contrast, dystrophic nail changes and psoriatic arthritis are more common in the Cw6-negative patients (38).
Linkage to HLA in large multiplex families has not been as convincing. The first demonstration of this was by Lin et al. (39). In large multiplex families when most affecteds harbor HLA-Cw*0602, linkage to HLA is not always demonstrated (8). This is due to the existence of multiple HLA haplotypes segregating in single families. In the remaining families that do not show linkage to HLA-Cw*0602 and where affecteds are not HLA-Cw*0602, linkage to other loci has been demonstrated (see below) in some cases.