- Email: [email protected]
Searching for the Major Histocompatibility Complex Psoriasis Susceptibility Gene
REVIEW ARTICLES
Searching for the Major Histocompatibility Complex Psoriasis Susceptibility Gene Francesca Capon, Margo Munro, Jonathan Barker,* and Richard Trembath
Division of Medical Genetics, University of Leicester, Leicester, U.K.; *St John's Institute of Dermatology, Kings College, London, U.K.
linkage disequilibrium ®ne mapping. Studies of positional candidate genes have also been undertaken, focusing on HLA-C, corneodesmosin, and a-helix coiled-coil rod homolog genes. Methodologic approaches, results, and interpretations of these studies are discussed, as well as future research objectives. In particular, we emphasize the importance of characterizing PSORS1 linkage disequilibrium patterns and developing functional assays for diseaseassociated alleles. Key words: a-helix coiled-coil rod homolog/complex disease/corneodesmosin/HLA/linkage disequilibrium. J Invest Dermatol 118:745±751, 2002
Psoriasis, a common skin disorder, is widely regarded to be multifactorial in origin including gene±gene and gene±environment interactions. Genetic and allelic heterogeneity, multifactorial inheritance, and low penetrance of susceptibility alleles substantially complicate both study design and interpretation of results. Notwithstanding these dif®culties, genome-wide scans for psoriasis susceptibility have generated robust evidence for a major locus lying within the major histocompatibility complex (PSORS1, Psoriasis Susceptibility 1), on the short arm of chromosome 6. Subsequent studies have sought to re®ne the PSORS1 boundaries by means of
P
SOME KEY METHODOLOGIC CONSIDERATIONS
soriasis is a common in¯ammatory skin disorder affecting approximately 2% of Caucasians. Though rarely fatal, the disease is signi®cantly debilitating not least when complicated by the occurrence of a sero-negative, potentially disabling arthritis, as seen in some 5% of psoriasis patients (Christophers and Sterry, 1993). Psoriasis is associated with keratinocyte hyperproliferation and dermal in®ltration of activated T lymphocytes (Barker, 1991), yet the underlying pathogenic mechanisms remain poorly characterized. A signi®cant genetic component to psoriasis susceptibility has long been supported by observations of family disease clustering, elevated concordance rates amongst monozygous twins, and numerous reports of disease association with the HLA-Cw6 antigen (Elder et al, 1994). These data were mostly derived using patients presenting at a young age (< 40 y), leading to the suggestion that HLA-Cw6 predisposes to an earlier onset of psoriasis (Enerback et al, 1997). Over the last few years, a number of research groups have pursued major studies investigating the molecular genetic basis of psoriasis. It therefore seems timely to review the current state of psoriasis genetics research. Particular attention is given to the genetic analysis of the major histocompatibility complex (MHC) on human chromosome 6, a region subject to intense scrutiny, as evidenced by the rush of recent publications (Asumalahti et al, 2000; Nair et al, 2000; Chia et al, 2001; O'Brien et al, 2001).
Psoriasis is etiologically a complex disorder with a widely held view that environmental factors trigger the disease in genetically susceptible individuals. Such multifactorial pathways are considered likely to underlie the majority of common medical disorders including in¯ammatory bowel disease, asthma, or diabetes. Re¯ecting the major public health concerns attributed to these common disease states and the need to better understand their pathogenesis, considerable effort has been exerted to characterize associated molecular genetic determinants. Signi®cant breakthroughs have recently been achieved, e.g., the identi®cation of the NOD2 and CAPN10 genes as contributors to susceptibility for Crohn's disease and type II diabetes, respectively (Horikawa et al, 2000; Hugot et al, 2001; Ogura et al, 2001). It is clear, however, that the identi®cation of common disease susceptibility genes remains a major challenge. Some of the recognized dif®culties are brie¯y discussed below. Detailed reviews providing greater insight into such methodologic issues have recently been published elsewhere (Terwilliger and Weiss, 1999; Risch, 2000). Linkage analysis Linkage analysis of extended pedigrees has proven a most effective tool for the assignment of loci involved in monogenic disorders. The ef®ciency of such mapping strategies is critically dependent upon correct assumptions regarding the disease mode of inheritance. As the transmission of complex traits does not follow simple Mendelian inheritance patterns, novel linkage tools have been devised. These statistics typically implement ``nonparametric'' linkage analysis strategies, which highlight regions of the genome that demonstrate excess of allele sharing between affected family members (Risch, 2000). Further compounding factors need to be considered. Susceptibility to complex disorders is probably determined by the interplay between several independent genes, such that detecting the effect of a single locus requires due care in study design. Usually, large family cohorts have to be recruited, capable of
Manuscript received August 30, 2001; revised January 8, 2002; accepted for publication January 14, 2002. Reprint requests to: Professor Richard Trembath, Division of Medical Genetics, Adrian Building, University Road, Leicester LE1 7RH, U.K. Email: [email protected] Abbreviations: CDSN, corneodesmosin; HCR, a-helix coiled-coil rod homolog; LD, linkage disequilibrium; SNP, single-nucleotide polymorphism; TDT, transmission disequilibrium test. 0022-202X/02/$15.00
´ Copyright # 2002 by The Society for Investigative Dermatology, Inc. 745
746
CAPON ET AL
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
Table I. Published psoriasis susceptibility loci Chromosomal location
Locus name
Linkage reports
1p 1q
PSORS7 PSORS4
Veal et al (2001) Capon et al (1999a) Bhalerao and Bowcock (1998) Veal et al (2001) Bhalerao and Bowcock (1998) Trembath et al (1997) Enlund et al (1999b) Samuelsson et al (1999) Bhalerao and Bowcock (1998) Samuelsson et al (1999) Matthews et al (1996) Nair et al (1997) Trembath et al (1997) Burden et al (1998) Capon et al (1999b) Enlund et al (1999a) Jenisch et al (1998) Samuelsson et al (1999) Veal et al (2001) Veal et al (2001) Trembath et al (1997) Bhalerao and Bowcock (1998) Veal et al (2001) Samuelsson et al (1999) Nair et al (1997) Tomfohrde et al (1994) Nair et al (1997) Samuelsson et al (1999) Enlund et al (1999a) Lee et al (2000) Veal et al (2001) Nair et al (1997) Trembath et al (1997)
2p 2q 3q
PSORS5
4q13 4q34 6p21.3
PSORS3 PSORS1
7 8q 14q 15 16q 17q25
PSORS2
19p13
PSORS6
20p
generating adequate power to declare linkage; in general, the weaker the contribution of a single locus, the larger the sample size needed to detect it. Recruiting large data sets raises issues of phenotypic and genetic heterogeneity, however. Distinct genes may contribute to disease susceptibility in different populations or disease subtypes. Patient clinical phenotypes should therefore be meticulously characterized, in order to minimize the introduction of disease heterogeneity. Furthermore, many populations show a remarkable degree of genetic admixture, re¯ecting a long history of migration. The gene-¯ow underlying such admixed populations can introduce multiple genetic determinants for a given trait, thereby decreasing the power to detect linkage to a single locus (Pfaff et al, 2001). Efforts to ascertain families from a similar ethnic background may reduce the level of genetic admixture. A number of investigators have described the theoretical advantages of this approach (de la Chapelle and Wright, 1998) and several molecular genetic studies have been performed on isolated populations. For instance, an atopy susceptibility locus has recently been assigned through the analysis of a Finnish cohort (Latinen et al, 2001). Even carefully selected data sets may still present signi®cant heterogeneity, however, through undetected population strati®cation or the occurrence of phenocopies (nongenetic examples of the disorder). Failures to replicate reported linkage assignments should be considered in the context of these confounding factors. Lack of replication may re¯ect false positive linkage results. For example, a linkage of schizophrenia to chromosome 5q could not be replicated, despite the assignment being strongly supported by a highly signi®cant LOD score. It was eventually agreed that the original result represented a spurious assignment (McGuf®n et al, 1990). Alternatively, the inability to replicate previous ®ndings may simply re¯ect locus heterogeneity, which substantially reduces the power to detect linkage (Leal and Ott, 2000).
In psoriasis, independent genome-wide scans have suggested the involvement of a large number of chromosomal regions (Table I), but replication has only been achieved for a limited number. Profoundly robust evidence supports the presence of a major susceptibility locus (PSORS1) within the MHC, as linkage to this region has been observed by many authors (Table I). The assignment of other loci, such as PSORS2 and PSORS4, has been replicated in independent studies. In contrast, involvement of other genomic regions (e.g., chromosome 20p) has been suggested in several studies, yet in none has genome-wide signi®cance been achieved. Such regions may represent loci with minor effects potentially requiring very large data sets to reveal their contribution to disease susceptibility. Finally, it is noteworthy that four psoriasis susceptibility loci, namely PSORS2, PSORS4, PSORS5, and the chromosome 20p locus, overlap susceptibility regions for atopic dermatitis (Fig 1) (Cookson et al, 2001). This observation suggests the intriguing possibility that common genetic susceptibility pathways underlie in¯ammatory skin disorders. Association studies In complex diseases, genome-wide scans typically assign susceptibility loci to broad genomic regions, spanning 20 cM or more. Re®nement of these locations cannot be achieved by the traditional crossover family-based analysis. In multifactorial diseases, genetic heterogeneity and the low penetrance of at-risk alleles each contribute to blur the relationship between phenotype and single locus genotype. Alternative approaches to susceptibility locus re®nement have been proposed. These methods typically rely on linkage disequilibrium (LD) ®ne mapping (Jorde, 1995). The rationale for these experiments is the assumption that an ``at-risk allele'' arose on an ancestral haplotype (a segment of DNA inherited from a common ancestor), and then transmitted through a number of generations to be inherited by patients under analysis (Fig 2). As the ancestral haplotype will have been broken down by recombination, a process occurring at each meiosis, only marker alleles lying in close proximity to the ``true'' susceptibility locus will continue to be coinherited and therefore can be found at increased frequency among affected individuals. Based on this rather simple model, the broad susceptibility regions identi®ed by genome-wide scans may be re®ned by increasing the number of genetic markers typed within a region, and allele frequencies compared between patients and controls. LD is typically maintained across much smaller genetic distances (typically less than 1 cM) than would be detected in linkage studies in family-based analysis. The marker alleles in LD with the disease will therefore highlight the most likely location for the susceptibility gene. Although potentially powerful, the LD mapping approach has some recognized drawbacks. The results of case-control studies can easily be biased through inappropriately matched controls and also through undetected population strati®cation. Indeed, these shortcomings have been held responsible for the poor reproducibility plaguing association studies of complex diseases (Altshuler et al, 2000). Moreover, the use of unrelated subjects does not allow the construction of haplotypes, a powerful tool for mapping susceptibility genes in complex traits (Rioux et al, 2001). Hence, in recent years a number of family-based association tests have been developed (Schaid, 1998), the most popular of which is the socalled transmission disequilibrium test (TDT). The TDT requires genotyping of nuclear families including at least an affected offspring together with the parents. In the analysis of such trios, parental alleles transmitted to the affected offspring are identi®ed and scored as cases. Conversely the untransmitted parental alleles are scored as controls. Allele frequencies are determined for the two groups and the signi®cance of the difference is assessed by a c2 test (Spielman et al, 1993). So far, efforts to re®ne psoriasis susceptibility loci using LD mapping strategies have focused on the PSORS1 region. The results of these studies are reviewed below. How large is the minimal PSORS1 region? The initial approach to susceptibility locus ®ne mapping relied on the analysis
VOL. 118, NO. 5 MAY 2002
PSORIASIS SUSCEPTIBILITY AND THE MHC
747
Figure 1. The psoriasis susceptibility loci overlapping with loci for other in¯ammatory disorders. The approximate extent of the mapped region is highlighted with a solid black bar to the right-hand side of the banded chromosome and the associated diseases are de®ned in the text. The approximate locations of the NOD2 (Crohn's disease gene), MHC genes, and the epidermal differentiation complex genes are denoted by a bracket to the left-hand side of the banded chromosome and labeled accordingly.
Figure 2. Re®nement of a disease susceptibility locus by ancestral haplotype analysis. Upper panel: the assumption is made that an at-risk allele originally arose on a single chromosome carrying a determined haplotype. In the following generations, the haplotype in question is transmitted to patients, together with the at-risk allele (left). As no single susceptibility allele is suf®cient to trigger a multifactorial disease, the at-risk variant and the surrounding haplotype may also be found on some control chromosomes, even though at a lower frequency (right, black bar with black circle). At the same time, some unaffected individuals will inherit chromosomes that did not undergo the unique mutation event generating the susceptibility allele, but still carry the same haplotype (right, black bar without black circle). Lower panel: as generations succeed, all haplotypes are disrupted by recombination. The ancestral haplotype carrying the susceptibility allele is broken down too, so that only the portion lying in close proximity of the at-risk variant can be maintained. Thus, the extension of minimal conserved ancestral haplotypes can be used to support the re®nement of a disease susceptibility locus.
of HLA haplotypes. A study by Schmitt-Egenolf et al (1996) demonstrated an association between psoriasis and the extended haplotype (EH) 57.1, including Cw6, B57, DRB1*0701, DQA1*0201, and DQB1*0303 alleles. This is consistent with the results of previous serologic surveys reporting associations with class I antigens HLA-B13 and HLA-B57, as well as class II antigens HLA-DR4 and HLA-DR7 (Elder et al, 1994). The SchmittEgenolf study, however, indicated that HLA-Cw6 and HLA-B57 are the alleles showing the most signi®cant association, leading the authors to conclude that the susceptibility gene lies towards the class I side of the EH57.1 haplotype. Similar results were also
obtained by Jenisch et al (1998), who demonstrated that the EH57.1 class I alleles were selectively retained among affected individuals. Three papers concerning the LD analysis of the PSORS1 interval have been published (Balendran et al, 1999; Oka et al, 1999; Nair et al, 2000). All three groups observed extended regions of LD within PSORS1; however, distinct statistical approaches were adopted in the independent studies and somewhat differing conclusions were drawn particularly with regard to the extent and location of the minimal PSORS1 region. Balendran et al (1999) examined a sample of 78 north European families by typing 14 markers spanning the entire MHC.
748
CAPON ET AL
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
Figure 3. Outcome of PSORS1 re®nement studies. The position of known genes is reported in the upper part of the ®gure. Black boxes in the lower part indicate published minimal intervals. Grey boxes have been added to illustrate how much farther these regions may extend (see text for details).
Genotypes were analyzed using a battery of linkage and association based methods. Haplotypes were also derived and ancestral MHC chromosomes were inferred. The results of the various test statistics showed a convergent peak at the boundary between class III and class I loci, and were consistent with a PSORS1 location between tumor necrosis factor a and STG genes (Fig 3). In their ®ne mapping study, Oka et al (1999) adopted a different experimental design. They focused their attention on a much smaller genomic region of just over a megabase (»1000 kb) segment surrounding HLA-C. This interval, selected on the basis of the association between psoriasis and HLA-Cw6, was saturated with 11 markers that were typed in 76 affected and 132 controls. Thus, a case-control rather than a family-based cohort was analyzed in this study. The presence of association was assessed using two different statistics: the c2 test and the analysis of deviation from Hardy±Weinberg equilibrium. The resulting p-values peaked within a 111 kb interval telomeric to HLA-C, whereas less signi®cant association was maintained along a broader interval (Fig 3). Nair et al (2000) carried out a comprehensive analysis of the MHC, by typing 62 markers in 478 psoriasis families. The authors observed two conserved haplotypes (RH1 and RH2) separated by a 50 kb interval de®ned by markers M6S111 and M6S169. As RH1 only was present on all the identi®ed risk haplotypes, the authors concluded that the PSORS1 location could be re®ned to the 60 kb RH1 interval (Fig 3). The results from the three ®ne mapping studies are summarized in Fig 3, clearly showing that there is no overlap between the published minimal intervals. All three re®ned regions may extend farther than assumed by the investigators, however. In the Balendran study, the critical susceptibility interval region was mainly inferred from ancestral haplotype analysis, so that the re®ned intervals reported in Fig 3 correspond to the minimal conserved haplotypes observed by the authors. The PSORS1 susceptibility gene does not necessarily lie within conserved haplotype regions, however. Like all crossover-de®ned loci, the minimal PSORS1 interval is actually delimited by the position of the two nearest recombining markers. If the re®ned interval is extended to include the closest markers that are not conserved along the ancestral haplotype, then the size of minimal regions increases.
The interpretation of the Oka study is more problematic as the analysis of unrelated individuals did not allow haplotypes to be derived. The authors assumed that PSORS1 lies in the region where both statistics yielded signi®cant p-values, but signi®cance at a single test is maintained for some distance, on both sides of such interval. With regard to the Nair study, only two markers separate the RH1 and RH2 haplotypes. Taken together, this region extends over a 170 kb conserved interval, closely corresponding to the peak of LD identi®ed by the TDT. Double recombination events around the M6S111 and M6S169 markers would be required to account for the separation of RH1 and RH2. Given that the authors report ®ve distinct haplotypes carrying both RH1 and RH2, several such double recombinations would have to be assumed. An alternative hypothesis suggests the existence of a single RH haplotype carrying two microsatellite markers that have undergone a CA slippage mutation (Timsit, 1999). This hypothesis is consistent with the observation that the two most common risk haplotypes share the same M6S169 alleles and carry M6S111 alleles differing for a single CA repeat. Overall, this interpretation would support a larger PSORS1 interval, extending to include the 50 kb segment harbouring M6S111 and M6S169 (Fig 3). If the extended regions are aligned, some overlap between the various studies can be detected (Fig 3). Indeed the comparison between the different minimal intervals points to a 100 kb segment telomeric to HLA-C, including OTF3, TCF19, and a-helix coiled-coil rod homolog (HCR) genes. Many issues related to the re®nement methodology, however (e.g., reliability of some test statistics), are still a matter of considerable debate among workers in the ®eld. Thus, the 100 kb interval described above can only be considered a provisional assignment, and candidate genes lying in its proximity cannot be excluded only on the basis of their position. CANDIDATE GENE ANALYSES Investigations of PSORS1 candidate genes have been accumulating at an increasing pace as the evidence for an MHC susceptibility gene grew ever more robust. The ®rst and most extensively investigated candidate was HLA-C. The association between psoriasis and the HLA-Cw6 antigen has been observed in a wide range of ethnic groups, and there is a general correlation between
VOL. 118, NO. 5 MAY 2002
the prevalence of the disease and that of Cw6 (Mallon et al, 1999). On this basis, several groups have carried out HLA-C molecular analyses in psoriasis families. These studies have repeatedly indicated the HLA-Cw*0602 allele as the most signi®cant marker for disease risk prediction (Tazi Ahnini et al, 1999; Asumalahti et al, 2000; Enerback et al, 2000). Three different Cw6-positive haplotypes that are associated with psoriasis (Cw6-B57, Cw6B13, and Cw6-B37), however, have been described (Elder et al, 1994). These three haplotypes confer different disease risks (Jenisch et al, 1998), and HLA-C seems to lie outside the most likely PSORS1 interval (Fig 3). Taken together with the lack of a characterized functional role in psoriasis, HLA-Cw6 is widely assumed be a marker in LD with the PSORS1 gene rather than the susceptibility allele itself (Allen et al, 1999; Asumalahti et al, 2000; Nair et al, 2000). A second, extensively investigated candidate is the corneodesmosin (CDSN) gene (OMIM *602593). CDSN lies 160 kb telomeric to HLA-C (Fig 3) and codes for corneodesmosin, a desmosomal protein involved in keratinocyte cohesion and desquamation (Guerrin et al, 1998). Interestingly, a recent study has demonstrated CDSN overexpression and altered distribution in psoriatic epidermis (Allen et al, 2001). These are disease-speci®c changes, as they have not been observed in other in¯ammatory skin disorders such as atopic dermatitis and lichen planus (Allen et al, 2001). The CDSN gene is remarkably polymorphic and a singlenucleotide polymorphism (SNP) frequency of up to 1 each 100 bp has been observed (Enerback et al, 2000). Most CDSN variants have been tested for association, and several studies have reported signi®cant p-values for a coding SNP that causes a Ser®Leu substitution (Allen et al, 1999; Jenisch et al, 1999; Tazi Ahnini et al, 1999; Enerback et al, 2000). This association was not observed in certain ethnic groups, however, including Japanese, Finnish, and Spanish (Ishihara et al, 1996; Asumalahti et al, 2000; Gonzalez et al, 2000). These discrepancies might still be accounted for by allelic heterogeneity, however. As CDSN polymorphism content is so high, SNP haplotypes rather than single alleles are likely to in¯uence the protein structure. In this context, the effect of a variant such as the Ser®Leu substitution could be reinforced or neutralized, depending on the alleles that are in LD with it, in each population. In conclusion, CDSN still remains an attractive candidate gene because of its function, its skin-speci®c expression pattern, and its alterations in psoriatic epidermis. At this stage, a functional study of the psoriasis-associated CDSN SNP is clearly needed. In particular, in vivo or in vitro experiments are required in order to investigate the effects of the Ser/Leu variant on protein expression and localization. A third candidate, the HCR gene, has lately been under intense scrutiny. HCR (OMIM *605310) was ®rst identi®ed as a putative structural component of skin-related tissues, as it presented trichohyalin, myosin, and laminin homologies (Guilladeux et al, 1998). Asumalahti et al (2000) demonstrated that the gene is ubiquitously expressed but is upregulated in psoriatic epidermis. They also tested the HCR as a PSORS1 candidate by analysing 17 intragenic SNPs in a sample of Finnish origin. Their study disclosed a signi®cant association between psoriasis and a two-SNP haplotype bearing two Arg®Trp substitutions (HCR*Trp-Trp allele). O'Brien et al (2001) subsequently analyzed HCR in a smaller data set of Swedish origin. They replicated the association with the *Trp-Trp allele and observed signi®cant p-values for six additional SNPs, including a third Arg®Trp substitution. When they strati®ed their results according to Cw*0602 presence, however, association was maintained only for one silent polymorphism. On the other hand, stratifying for HCR SNPs did not affect association with Cw*0602. On this basis, the authors concluded that the association with the *Trp-Trp allele was solely due to LD with Cw*0602 and therefore had no pathologic relevance. It could be argued that Cw*0602 is not an SNP but a haplotype, however, as its sequence is de®ned by several nucleotide changes along the HLA-C gene. So, data strati®ed according to Cw*0602 presence
PSORIASIS SUSCEPTIBILITY AND THE MHC
749
may not be comparable to data strati®ed according to HCR single SNPs. A second report questioning HCR candidacy was published by Chia et al (2001). The paper shows that the HCR *Trp-Trp allele lies on most but not all the at-risk haplotypes identi®ed by Nair et al (2000) in their PSORS1 re®nement study. In particular, a Cw8B65 haplotype is described that does not carry the associated *TrpTrp allele. The p-values that originally demonstrated the preferential transmission of the Cw8-B65 haplotype, however, are on the threshold of statistical signi®cance (Nair et al, 2000), and even if the Cw8-B65 haplotype contributes to psoriasis pathogenesis the absence of the *Trp-Trp allele could still be accounted for by allelic heterogeneity. Chia et al also disputed the HCR pathogenic role based on the observation of a common nonrisk haplotype carrying HLA-Cw7 and the *Trp-Trp allele. This ®nding can be explained by low penetrance of the risk allele, however, a feature already noted in other complex disease susceptibility genes (Reich and Lander, 2001). Indeed, the effects of the *Trp-Trp allele could be modulated by haplotype-speci®c SNPs, as hypothesized for CDSN. In this context, it would be interesting to investigate whether the nonrisk Cw7 haplotype also carries the additional HCR risk SNPs identi®ed by O'Brien et al (2001). In conclusion, the data generated so far do not allow the dismissal of HCR as a PSORS1 candidate: the analysis of additional data sets and the functional characterization of the *Trp-Trp allele are still needed, in order to elucidate HCR contribution to psoriasis susceptibility. Other genes lying in the proximity of HLA-C have been analyzed, albeit less systematically. Gonzalez et al (2000) reported an association with one allele of the OTF3 gene in the Spanish population. The authors themselves acknowledge, however, that OTF3 is a somewhat unlikely candidate for disease susceptibility, as the gene encodes a transcriptional factor playing a major rule in embryonic stem cells lineage commitment (Niwa et al, 2000). Teraoka et al (2000) analyzed the TCF19 gene (Fig 3), coding for a cell growth regulator, ubiquitously expressed in the G1±S phases of the cell cycle. They identi®ed three SNPs and one deletion polymorphism, but failed to observe any association. Our group has recently been investigating SPR1 and SEEK1 genes (Fig 3). Direct sequencing of SPR1 failed to detect any highfrequency SNP within this gene (Munro et al, unpublished observations). WHERE DO WE GO FROM HERE? So far, none of the analyzed candidates seems to match the criteria for a disease susceptibility gene (see below). The PSORS1 region may contain as yet uncharacterized transcripts, e.g., the MHC sequence annotation predicts three open reading frames in the interval between HLA-C and OTF3 (Ensembl database of annotated genomic sequences, http://www.ensembl.org). Interestingly, these open reading frames match expressed sequence tags that map to the same region. No gene has yet been characterized, however. This may in part be due to the complexity of the target genomic region, containing a large number of repeats and retrovirus-like elements. On the other hand, these expressed sequence tags may represent expressed pseudogenes. To date, few susceptibility genes have been characterized; hence criteria for declaring disease causality are mostly based on theoretical considerations (Terwilliger and Weiss, 1999; Risch, 2000). For instance, it is widely assumed that the genetic contribution to complex disorders is due to relatively high frequency SNPs with subtle effects on protein function (Chakravarti, 1999). This did not prove to be the case for Crohn's disease, however, in which a lowfrequency frameshift mutation of the NOD2 gene has recently been implicated in disease susceptibility (Hugot et al, 2001; Ogura et al, 2001). Likewise, replication of association has long been considered an essential requirement for susceptibility gene assignment. This implies that the same alleles determine disease susceptibility in distinct populations. The real picture is likely to be more complex, however. For instance, it has been demonstrated
750
CAPON ET AL
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
that the CAPN10 haplotype1 involved in type II diabetes susceptibility confers very different risks in different ethnic groups (Horikawa et al, 2000). Another implicit assumption underlying most association studies is that disease susceptibility is caused by a single or a few closely linked SNPs. Again, this is not necessarily the case. Variation in adjacent genes may underlie disease susceptibility, as is the case with the IDDM superlocus lying within the MHC (Lie et al, 1999). More generally, the effect of single SNPs may vary according to the haplotype they reside on. Indeed this has proven to be the case even for a ``simple'' monogenic disorder like cystic ®brosis, where CFTR mutation effects can be modulated by intronic variants (Estivill, 1996). If similar mechanisms were active in the PSORS1 region, identifying psoriasis susceptibility alleles by association studies alone could prove a formidable task indeed. Despite these dif®culties, susceptibility gene assignment may still be achieved by assessing the functional relevance of diseaseassociated SNPs. Indeed, demonstrating that a given SNP affects gene expression levels or protein secondary structure is expected to provide robust support for the involvement of a candidate gene. This is likely to be particularly true for the PSORS1 locus. The con¯icting interpretations of recent association ®ndings show how genetic analysis may be close to its limit, suggesting that further resolution cannot be achieved owing to LD conservation along the MHC. Before genetic analysis is set aside, however, it will be important to ensure that a near complete assessment of the ancestral history of the numerous genetic variants within the PSORS1 region has been achieved. The MHC is unusual in the way extended regions of LD have been maintained, probably re¯ecting complex mechanisms that ensure such genomic regions remain intact. The availability of a high-resolution genetic map of this region allows the detailed characterization of highly localized ``hot spots'' of recombination within the MHC with important implications for disease causality (Jeffreys et al, in press). These studies will need to be assessed in the light of such analysis. Together with the rich and ¯exible array of functional analysis techniques, researchers have now at their disposal adequate tools to identify the long-hunted-for PSORS1 gene. Note added in proof: Asumalahti et al have recently reported the analysis of HLA-C/HCR/CDSN risk haplotypes in ethnically diverse data sets, con®rming highly signi®cant association of psoriasis with HLA-CW*0602 and HCR*TRP-TRP alleles (Hum Mol Genet 11:589±597, 2002). The authors wish to thank the many psoriasis family members for support, and the National Psoriasis Foundation and the Wellcome Trust for ®nancial aid. FC is a Wellcome Trust Travelling Fellow.
REFERENCES Allen MH, Veal C, Faassen A, Powis SH, Vaughan RW, Trembath RC, Barker JNWN: A non-HLA gene within the MHC in psoriasis. Lancet 353:1589± 1590, 1999 Allen MH, Ishida-Yamamoto A, McGrath J, et al: Coneodesmosin gene expression in psoriasis vulgaris differs from normal skin and other in¯ammatory skin disorders. Lab Invest 81:969±976, 2001 Altshuler D, Hirschorn JN, Klannemark M, et al: The common PPAR Pro12Ala polymorphism is associated with decreased risk of type 2 diabetes. Nature Genet 26:76±80, 2000 Asumalahti K, Latinen T, Itkonen-Vatjus R, et al: A candidate gene for psoriasis near HLA-C, HCR (Pg8), is highly polymorphic with a disease associated susceptibility allele. Hum Mol Genet 9:1533±1542, 2000 Balendran N, Clough RL, Arguello JR, et al: Characterization of the major susceptibility region for psoriasis at chromosome 6p21.3. J Invest Dermatol 113:322±328, 1999 Barker JN: The pathophysiology of psoriasis. Lancet 338:227±230, 1991 1 Nair RP, Chia N, Shahabi L, et al: Sequence variation in a retroviruslike expressed sequence tag in the PSORS1 critical of the MHC: lack of evidence for a causative psoriasis gene. Society of Investigative Dermatology 62nd Annual Meeting. Washington, May 9±12, 2001.
Bhalerao J, Bowcock A: The genetics of psoriasis: a complex disorder of the skin and immune system. Hum Mol Genet 7:1537±1545, 1998 Burden AD, Javed S, Bailey M, Hodgins M, Connor M, Tillman D: Genetics of psoriasis: paternal inheritance and a locus on chromosome 6p. J Invest Dermatol 110:958±960, 1998 Capon F, Novelli G, Semprini S, et al: Searching for psoriasis susceptibility genes in Italy: genome scan and evidence for a new locus on chromosome 1. J Invest Dermatol 112:32±35, 1999a Capon F, Semprini S, Dallapiccola B, Novelli G: Evidence for interaction between psoriasis susceptibility loci on chromosomes 6p21 and 1q21. Am J Hum Genet 65:1798±1800, 1999b Chakravarti A: Population genetics making sense out of sequence. Nature Genet 21 (Suppl. 1):56±60, 1999 de la Chapelle A, Wright FA: Linkage disequilibrium mapping in isolated populations: the example of Finland revisited. Proc Natl Acad Sci USA 95:12416±12423, 1998 Chia NV, Stuart P, Nair RP, et al: Variations in HCR (Pg8) gene are unlikely to be causal for familial psoriasis. J Invest Dermatol 116:823±824, 2001 Christophers E, Sterry W: Psoriasis. In: Fitzpatrick TB, Eisen AZ, Wolff K, Austin KF, eds. Dermatology in General Medicine. New York: McGraw-Hill, 1993 Cookson WOCM, Ubhi B, Lawrence R, et al: Genetic linkage of childhood atopic dermatitis to psoriasis susceptibility loci. Nature Genet 27:372±373, 2001 Elder JT, Henseler T, Christophers E, Voorhees JJ, Nair RP: Of genes and antigens: the inheritance of psoriasis. J Invest Dermatol 103:150S±153S, 1994 Enerback C, Martinsson T, Inerot A, et al: Signi®cantly earlier age at onset for the HLA-Cw6-positive than for the Cw6-negative psoriatic sibling. J Invest Dermatol 109:695±696, 1997 Enerback C, Enlund F, Inerot A, Samuelsson L, Wahlstrom J, Swanbeck G, Martinsson T: S gene (corneodesmosin) diversity and its relationship to psoriasis; high content of cSNP in the HLA-linked S gene. J Invest Dermatol 114:1158±1163, 2000 Enlund F, Samuelsson L, Enerback C, et al: Analysis of three suggested psoriasis susceptibility loci in a large Swedish set of families: con®rmation of linkage to chromosome 6p (HLA region), and to 17q, but not to 4q. Hum Hered 49:2±8, 1999a Enlund F, Samuelsson L, Enerback C, et al: Psoriasis susceptibility locus in chromosome region 3q21 identi®ed in patients from southwest Sweden. Eur J Hum Genet 7:783±790, 1999b Estivill X: Complexity in a monogenic disease. Nature Genet 12:348±350, 1996 Gonzalez S, Martinez Borra J, Del Rio JS, Santos-Juanes J, Lopez-Vazquez A, Blanco-Gelaz M, Lopez-Larrea C: The OTF3 gene polymorphism confers susceptibility to psoriasis independent of the association of HLA-Cw*0602. J Invest Dermatol 115:824±828, 2000 Guerrin M, Simon M, Montezin M, Haftek M, Vincent C, Serre G: Expression cloning of human corneodesmosin proves its identity with the product of the S gene and allows improved characterization of its processing during keratinocyte differentiation. J Biol Chem 273:22640±22647, 1998 Guilladeux T, Janer M, Wong GKS, Spies T, Geraghty DE: The complete genomic sequence of 424,015 bp at the centromeric end of the HLA class I region: gene content and ploymorphism. Proc Natl Acad Sci USA 95:9494±9499, 1998 Horikawa Y, Oda N, Cox NJ, et al: Genetic variation in the gene encoding calapain10 is associated with type 2 diabetes mellitus. Nature Genet 26:163±175, 2000 Hugot JP, Chamaillard M, Zouali H, et al: Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn's disease. Nature 411:599±603, 2001 Ishihara M, Yamagata N, Ohno S, et al: Genetic polymorphisms in the keratin-like S gene within the human major histocompatibility complex and association analysis on the susceptibility to psoriasis vulgaris. Tissue Antigens 48:182±186, 1996 Jeffreys AJJ, Kauppi L, Neumann R: Intensely punctate meiotic recombination in the class II region of the major histocompatibility complex. Nature Genet, 29:217± 222, 2001 Jenisch, S, Henseler T, Nair RP, et al: Linkage analysis of human leukocyte antigen (HLA) markers in familial psoriasis: strong disequilibrium effects provide evidence for a major determinant in the HLA-B/-C region. Am J Hum Genet 63:191±199, 1998 Jenisch S, Koch S, Henseler T, et al: Corneodesmosin gene polymorphism demonstrates strong linkage disequilibrium with HLA and association with psoriasis vulgaris. Tissue Antigens 54:439±449, 1999 Jorde LB: Linkage disequilibrium as a gene-mapping tool. Am J Hum Genet 56:11±14, 1995 Latinen T, Daly MJ, Rioux JD, et al: A susceptibility locus for asthma-related traits on chromosome 7 revealed by genome-wide scan in a founder population. Nature Genet 28:87±91, 2001 Leal SM, Ott J: Effects of strati®cation in the analysis of affected-sib-pair data: bene®ts and costs. Am J Hum Genet 66:567±575, 2000 Lee YA, Ruschendorf F, Windemuth C, et al: Genomewide scan in German families reveals evidence for a novel psoriasis-susceptibility locus on chromosome 19p13. Am J Hum Genet 67:1020±1024, 2000 Lie BA, Todd JA, Pociot F, et al: The predisposition to type 1 diabetes linked to the human leukocyte antigen complex includes at least one non-class II gene. Am J Hum Genet 64:793±800, 1999 Mallon E, Newson R, Bunker CB: HLA-Cw6 and the genetic predisposition to psoriasis: a meta analysis of published serologic studies. J Invest Dermatol 113:693±695, 1999 Matthews D, Fry L, Powles A, et al: Evidence that a locus for familial psoriasis maps to chromosome 4q. Nature Genet 14:231±233, 1996 McGuf®n P, Sargeant M, Hetti G, Tidmarsh S, Whatley S, Marchbanks RM:
VOL. 118, NO. 5 MAY 2002
Exclusion of a schizophrenia susceptibility gene from the chromosome 5q11q13 region: new data and a reanalysis of previous reports. Am J Hum Genet 47:524±535, 1990 Nair RP, Henseler T, Jenisch S, et al: Evidence for two psoriasis susceptibility loci (HLA and 17q) and two novel candidate regions (16q and 20p) by genomewide scan. Hum Mol Genet 6:1349±1356, 1997 Nair RP, Stuart P, Heneseler T, et al: Localization of psoriasis susceptibility locus PSORS1 to a 60 kilobase interval telomeric to HLA-C. Am J Hum Genet 66:1833±1844, 2000 Niwa H, Miyazaki J, Smith AG: Quantitative expression of Oct-3/4 de®nes differentiation, de-differentiation or self-renewal of ES cells. Nature Genet 24:372±376, 2000 O'Brien KP, Holm SJ, Nilsson S, et al: The HCR gene on 6p21 is unlikely to be a psoriasis susceptibility gene. J Invest Dermatol 116:750±754, 2001 Ogura Y, Bonen DK, Inohara N, et al: A frameshift mutation in NOD2 associated with susceptibility with Crohn's disease. Nature 411:603±606, 2001 Oka A, Tamiya G, Tomizawa M, et al: Association analysis using re®ned microsatellite markers localizes a susceptibility locus for psoriasis vulgaris within a 111kb segment telomeric to the HLA-C gene. Hum Mol Genet 8:2165±2170, 1999 Pfaff CL, Parra EJ, Bonilla C, et al: Population structure in admixed populations: effect of admixture dynamics on the pattern of linkage disequilibrium. Am J Hum Genet 68:198±207, 2001 Reich DE, Lander ES: On the allelic spectrum of human disease. Trends Genet 17:502±511, 2001 Rioux JD, Daly MJ, Silverberg MS, et al: Genetic variation in the 5q31 cytokine gene cluster confers susceptibility to Crohn disease. Nature Genet 29:223±228, 2001 Risch NJ: Searching for genetic determinants in the new millennium. Nature 405:847±856, 2000 Samuelsson L, Enlund F, Torinsson A, et al: A genome-wide search for genes
PSORIASIS SUSCEPTIBILITY AND THE MHC
751
predisposing to familial psoriasis by using a strati®cation approach. Hum Genet 105:523±529, 1999 Schaid DJ: Transmission disequilibrium test, family controls and great expectations. Am J Hum Genet 63:935±942, 1998 Schmitt-Egenolf M, Eiermann TH, Boehncke WH, Stander M, Sterry W: Familial juvenile onset psoriasis is associated with the human leukocyte antigen (HLA) class I side of the extended haplotype Cw6±B57-DRB1*0701-DQA1*0201DQB1*0303: a population- and family-based study. J Invest Dermatol 106:711± 714, 1996 Spielman RS, McGinnis RE, Ewens WJ: Transmission test for linkage disequilibrium: the insulin gene region and insulin dependent diabetes mellitus (IDDM). Am J Hum Genet 52:506±516, 1993 Tazi Ahnini R, Camp NJ, Cork NJ, Mee JB, Keohane SG, Duff GW, Di Giovine FS: Novel genetic association between the corneodesmosin (MHC S) gene and susceptibility to psoriasis. Hum Mol Genet 8:1135±1140, 1999 Teraoka Y, Naruse TK, Oka A, et al: Genetic polymorphisms in the cell growth regulated gene, SC1 telomeric of the HLA-C gene and lack of association of psoriasis vulgaris. Tissue Antigens 55:206±211, 2000 Terwilliger JD, Weiss KM: Linkage disequilibrium mapping of complex disease: fantasy or reality? Curr Opin Biotechnol 9:578±594, 1999 Timsit Y: DNA structure and polymerase ®delity. J Mol Biol 293:835±853, 1999 Tomfohrde J, Silverman A, Barnes R, et al: Gene for familial psoriasis susceptibility mapped to the distal end of chromosome 17q. Science 264:1141±1145, 1994 Trembath RC, Clough RL, Rosbotham JL, et al: Identi®cation of a major susceptibility locus on chromosome 6p and evidence for further disease loci revealed by a two stage genome-wide search in psoriasis. Hum Mol Genet 6:813±820, 1997 Veal CD, Clough RL, Barber RC, et al: Identi®cation of a novel psoriasis susceptibility locus at 1p and evidence of epistasis between PSORS1 and candidate loci. J Med Genet 38:7±13, 2001
Searching for the Major Histocompatibility Complex Psoriasis Susceptibility Gene Francesca Capon, Margo Munro, Jonathan Barker,* and Richard Trembath
Division of Medical Genetics, University of Leicester, Leicester, U.K.; *St John's Institute of Dermatology, Kings College, London, U.K.
linkage disequilibrium ®ne mapping. Studies of positional candidate genes have also been undertaken, focusing on HLA-C, corneodesmosin, and a-helix coiled-coil rod homolog genes. Methodologic approaches, results, and interpretations of these studies are discussed, as well as future research objectives. In particular, we emphasize the importance of characterizing PSORS1 linkage disequilibrium patterns and developing functional assays for diseaseassociated alleles. Key words: a-helix coiled-coil rod homolog/complex disease/corneodesmosin/HLA/linkage disequilibrium. J Invest Dermatol 118:745±751, 2002
Psoriasis, a common skin disorder, is widely regarded to be multifactorial in origin including gene±gene and gene±environment interactions. Genetic and allelic heterogeneity, multifactorial inheritance, and low penetrance of susceptibility alleles substantially complicate both study design and interpretation of results. Notwithstanding these dif®culties, genome-wide scans for psoriasis susceptibility have generated robust evidence for a major locus lying within the major histocompatibility complex (PSORS1, Psoriasis Susceptibility 1), on the short arm of chromosome 6. Subsequent studies have sought to re®ne the PSORS1 boundaries by means of
P
SOME KEY METHODOLOGIC CONSIDERATIONS
soriasis is a common in¯ammatory skin disorder affecting approximately 2% of Caucasians. Though rarely fatal, the disease is signi®cantly debilitating not least when complicated by the occurrence of a sero-negative, potentially disabling arthritis, as seen in some 5% of psoriasis patients (Christophers and Sterry, 1993). Psoriasis is associated with keratinocyte hyperproliferation and dermal in®ltration of activated T lymphocytes (Barker, 1991), yet the underlying pathogenic mechanisms remain poorly characterized. A signi®cant genetic component to psoriasis susceptibility has long been supported by observations of family disease clustering, elevated concordance rates amongst monozygous twins, and numerous reports of disease association with the HLA-Cw6 antigen (Elder et al, 1994). These data were mostly derived using patients presenting at a young age (< 40 y), leading to the suggestion that HLA-Cw6 predisposes to an earlier onset of psoriasis (Enerback et al, 1997). Over the last few years, a number of research groups have pursued major studies investigating the molecular genetic basis of psoriasis. It therefore seems timely to review the current state of psoriasis genetics research. Particular attention is given to the genetic analysis of the major histocompatibility complex (MHC) on human chromosome 6, a region subject to intense scrutiny, as evidenced by the rush of recent publications (Asumalahti et al, 2000; Nair et al, 2000; Chia et al, 2001; O'Brien et al, 2001).
Psoriasis is etiologically a complex disorder with a widely held view that environmental factors trigger the disease in genetically susceptible individuals. Such multifactorial pathways are considered likely to underlie the majority of common medical disorders including in¯ammatory bowel disease, asthma, or diabetes. Re¯ecting the major public health concerns attributed to these common disease states and the need to better understand their pathogenesis, considerable effort has been exerted to characterize associated molecular genetic determinants. Signi®cant breakthroughs have recently been achieved, e.g., the identi®cation of the NOD2 and CAPN10 genes as contributors to susceptibility for Crohn's disease and type II diabetes, respectively (Horikawa et al, 2000; Hugot et al, 2001; Ogura et al, 2001). It is clear, however, that the identi®cation of common disease susceptibility genes remains a major challenge. Some of the recognized dif®culties are brie¯y discussed below. Detailed reviews providing greater insight into such methodologic issues have recently been published elsewhere (Terwilliger and Weiss, 1999; Risch, 2000). Linkage analysis Linkage analysis of extended pedigrees has proven a most effective tool for the assignment of loci involved in monogenic disorders. The ef®ciency of such mapping strategies is critically dependent upon correct assumptions regarding the disease mode of inheritance. As the transmission of complex traits does not follow simple Mendelian inheritance patterns, novel linkage tools have been devised. These statistics typically implement ``nonparametric'' linkage analysis strategies, which highlight regions of the genome that demonstrate excess of allele sharing between affected family members (Risch, 2000). Further compounding factors need to be considered. Susceptibility to complex disorders is probably determined by the interplay between several independent genes, such that detecting the effect of a single locus requires due care in study design. Usually, large family cohorts have to be recruited, capable of
Manuscript received August 30, 2001; revised January 8, 2002; accepted for publication January 14, 2002. Reprint requests to: Professor Richard Trembath, Division of Medical Genetics, Adrian Building, University Road, Leicester LE1 7RH, U.K. Email: [email protected] Abbreviations: CDSN, corneodesmosin; HCR, a-helix coiled-coil rod homolog; LD, linkage disequilibrium; SNP, single-nucleotide polymorphism; TDT, transmission disequilibrium test. 0022-202X/02/$15.00
´ Copyright # 2002 by The Society for Investigative Dermatology, Inc. 745
746
CAPON ET AL
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
Table I. Published psoriasis susceptibility loci Chromosomal location
Locus name
Linkage reports
1p 1q
PSORS7 PSORS4
Veal et al (2001) Capon et al (1999a) Bhalerao and Bowcock (1998) Veal et al (2001) Bhalerao and Bowcock (1998) Trembath et al (1997) Enlund et al (1999b) Samuelsson et al (1999) Bhalerao and Bowcock (1998) Samuelsson et al (1999) Matthews et al (1996) Nair et al (1997) Trembath et al (1997) Burden et al (1998) Capon et al (1999b) Enlund et al (1999a) Jenisch et al (1998) Samuelsson et al (1999) Veal et al (2001) Veal et al (2001) Trembath et al (1997) Bhalerao and Bowcock (1998) Veal et al (2001) Samuelsson et al (1999) Nair et al (1997) Tomfohrde et al (1994) Nair et al (1997) Samuelsson et al (1999) Enlund et al (1999a) Lee et al (2000) Veal et al (2001) Nair et al (1997) Trembath et al (1997)
2p 2q 3q
PSORS5
4q13 4q34 6p21.3
PSORS3 PSORS1
7 8q 14q 15 16q 17q25
PSORS2
19p13
PSORS6
20p
generating adequate power to declare linkage; in general, the weaker the contribution of a single locus, the larger the sample size needed to detect it. Recruiting large data sets raises issues of phenotypic and genetic heterogeneity, however. Distinct genes may contribute to disease susceptibility in different populations or disease subtypes. Patient clinical phenotypes should therefore be meticulously characterized, in order to minimize the introduction of disease heterogeneity. Furthermore, many populations show a remarkable degree of genetic admixture, re¯ecting a long history of migration. The gene-¯ow underlying such admixed populations can introduce multiple genetic determinants for a given trait, thereby decreasing the power to detect linkage to a single locus (Pfaff et al, 2001). Efforts to ascertain families from a similar ethnic background may reduce the level of genetic admixture. A number of investigators have described the theoretical advantages of this approach (de la Chapelle and Wright, 1998) and several molecular genetic studies have been performed on isolated populations. For instance, an atopy susceptibility locus has recently been assigned through the analysis of a Finnish cohort (Latinen et al, 2001). Even carefully selected data sets may still present signi®cant heterogeneity, however, through undetected population strati®cation or the occurrence of phenocopies (nongenetic examples of the disorder). Failures to replicate reported linkage assignments should be considered in the context of these confounding factors. Lack of replication may re¯ect false positive linkage results. For example, a linkage of schizophrenia to chromosome 5q could not be replicated, despite the assignment being strongly supported by a highly signi®cant LOD score. It was eventually agreed that the original result represented a spurious assignment (McGuf®n et al, 1990). Alternatively, the inability to replicate previous ®ndings may simply re¯ect locus heterogeneity, which substantially reduces the power to detect linkage (Leal and Ott, 2000).
In psoriasis, independent genome-wide scans have suggested the involvement of a large number of chromosomal regions (Table I), but replication has only been achieved for a limited number. Profoundly robust evidence supports the presence of a major susceptibility locus (PSORS1) within the MHC, as linkage to this region has been observed by many authors (Table I). The assignment of other loci, such as PSORS2 and PSORS4, has been replicated in independent studies. In contrast, involvement of other genomic regions (e.g., chromosome 20p) has been suggested in several studies, yet in none has genome-wide signi®cance been achieved. Such regions may represent loci with minor effects potentially requiring very large data sets to reveal their contribution to disease susceptibility. Finally, it is noteworthy that four psoriasis susceptibility loci, namely PSORS2, PSORS4, PSORS5, and the chromosome 20p locus, overlap susceptibility regions for atopic dermatitis (Fig 1) (Cookson et al, 2001). This observation suggests the intriguing possibility that common genetic susceptibility pathways underlie in¯ammatory skin disorders. Association studies In complex diseases, genome-wide scans typically assign susceptibility loci to broad genomic regions, spanning 20 cM or more. Re®nement of these locations cannot be achieved by the traditional crossover family-based analysis. In multifactorial diseases, genetic heterogeneity and the low penetrance of at-risk alleles each contribute to blur the relationship between phenotype and single locus genotype. Alternative approaches to susceptibility locus re®nement have been proposed. These methods typically rely on linkage disequilibrium (LD) ®ne mapping (Jorde, 1995). The rationale for these experiments is the assumption that an ``at-risk allele'' arose on an ancestral haplotype (a segment of DNA inherited from a common ancestor), and then transmitted through a number of generations to be inherited by patients under analysis (Fig 2). As the ancestral haplotype will have been broken down by recombination, a process occurring at each meiosis, only marker alleles lying in close proximity to the ``true'' susceptibility locus will continue to be coinherited and therefore can be found at increased frequency among affected individuals. Based on this rather simple model, the broad susceptibility regions identi®ed by genome-wide scans may be re®ned by increasing the number of genetic markers typed within a region, and allele frequencies compared between patients and controls. LD is typically maintained across much smaller genetic distances (typically less than 1 cM) than would be detected in linkage studies in family-based analysis. The marker alleles in LD with the disease will therefore highlight the most likely location for the susceptibility gene. Although potentially powerful, the LD mapping approach has some recognized drawbacks. The results of case-control studies can easily be biased through inappropriately matched controls and also through undetected population strati®cation. Indeed, these shortcomings have been held responsible for the poor reproducibility plaguing association studies of complex diseases (Altshuler et al, 2000). Moreover, the use of unrelated subjects does not allow the construction of haplotypes, a powerful tool for mapping susceptibility genes in complex traits (Rioux et al, 2001). Hence, in recent years a number of family-based association tests have been developed (Schaid, 1998), the most popular of which is the socalled transmission disequilibrium test (TDT). The TDT requires genotyping of nuclear families including at least an affected offspring together with the parents. In the analysis of such trios, parental alleles transmitted to the affected offspring are identi®ed and scored as cases. Conversely the untransmitted parental alleles are scored as controls. Allele frequencies are determined for the two groups and the signi®cance of the difference is assessed by a c2 test (Spielman et al, 1993). So far, efforts to re®ne psoriasis susceptibility loci using LD mapping strategies have focused on the PSORS1 region. The results of these studies are reviewed below. How large is the minimal PSORS1 region? The initial approach to susceptibility locus ®ne mapping relied on the analysis
VOL. 118, NO. 5 MAY 2002
PSORIASIS SUSCEPTIBILITY AND THE MHC
747
Figure 1. The psoriasis susceptibility loci overlapping with loci for other in¯ammatory disorders. The approximate extent of the mapped region is highlighted with a solid black bar to the right-hand side of the banded chromosome and the associated diseases are de®ned in the text. The approximate locations of the NOD2 (Crohn's disease gene), MHC genes, and the epidermal differentiation complex genes are denoted by a bracket to the left-hand side of the banded chromosome and labeled accordingly.
Figure 2. Re®nement of a disease susceptibility locus by ancestral haplotype analysis. Upper panel: the assumption is made that an at-risk allele originally arose on a single chromosome carrying a determined haplotype. In the following generations, the haplotype in question is transmitted to patients, together with the at-risk allele (left). As no single susceptibility allele is suf®cient to trigger a multifactorial disease, the at-risk variant and the surrounding haplotype may also be found on some control chromosomes, even though at a lower frequency (right, black bar with black circle). At the same time, some unaffected individuals will inherit chromosomes that did not undergo the unique mutation event generating the susceptibility allele, but still carry the same haplotype (right, black bar without black circle). Lower panel: as generations succeed, all haplotypes are disrupted by recombination. The ancestral haplotype carrying the susceptibility allele is broken down too, so that only the portion lying in close proximity of the at-risk variant can be maintained. Thus, the extension of minimal conserved ancestral haplotypes can be used to support the re®nement of a disease susceptibility locus.
of HLA haplotypes. A study by Schmitt-Egenolf et al (1996) demonstrated an association between psoriasis and the extended haplotype (EH) 57.1, including Cw6, B57, DRB1*0701, DQA1*0201, and DQB1*0303 alleles. This is consistent with the results of previous serologic surveys reporting associations with class I antigens HLA-B13 and HLA-B57, as well as class II antigens HLA-DR4 and HLA-DR7 (Elder et al, 1994). The SchmittEgenolf study, however, indicated that HLA-Cw6 and HLA-B57 are the alleles showing the most signi®cant association, leading the authors to conclude that the susceptibility gene lies towards the class I side of the EH57.1 haplotype. Similar results were also
obtained by Jenisch et al (1998), who demonstrated that the EH57.1 class I alleles were selectively retained among affected individuals. Three papers concerning the LD analysis of the PSORS1 interval have been published (Balendran et al, 1999; Oka et al, 1999; Nair et al, 2000). All three groups observed extended regions of LD within PSORS1; however, distinct statistical approaches were adopted in the independent studies and somewhat differing conclusions were drawn particularly with regard to the extent and location of the minimal PSORS1 region. Balendran et al (1999) examined a sample of 78 north European families by typing 14 markers spanning the entire MHC.
748
CAPON ET AL
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
Figure 3. Outcome of PSORS1 re®nement studies. The position of known genes is reported in the upper part of the ®gure. Black boxes in the lower part indicate published minimal intervals. Grey boxes have been added to illustrate how much farther these regions may extend (see text for details).
Genotypes were analyzed using a battery of linkage and association based methods. Haplotypes were also derived and ancestral MHC chromosomes were inferred. The results of the various test statistics showed a convergent peak at the boundary between class III and class I loci, and were consistent with a PSORS1 location between tumor necrosis factor a and STG genes (Fig 3). In their ®ne mapping study, Oka et al (1999) adopted a different experimental design. They focused their attention on a much smaller genomic region of just over a megabase (»1000 kb) segment surrounding HLA-C. This interval, selected on the basis of the association between psoriasis and HLA-Cw6, was saturated with 11 markers that were typed in 76 affected and 132 controls. Thus, a case-control rather than a family-based cohort was analyzed in this study. The presence of association was assessed using two different statistics: the c2 test and the analysis of deviation from Hardy±Weinberg equilibrium. The resulting p-values peaked within a 111 kb interval telomeric to HLA-C, whereas less signi®cant association was maintained along a broader interval (Fig 3). Nair et al (2000) carried out a comprehensive analysis of the MHC, by typing 62 markers in 478 psoriasis families. The authors observed two conserved haplotypes (RH1 and RH2) separated by a 50 kb interval de®ned by markers M6S111 and M6S169. As RH1 only was present on all the identi®ed risk haplotypes, the authors concluded that the PSORS1 location could be re®ned to the 60 kb RH1 interval (Fig 3). The results from the three ®ne mapping studies are summarized in Fig 3, clearly showing that there is no overlap between the published minimal intervals. All three re®ned regions may extend farther than assumed by the investigators, however. In the Balendran study, the critical susceptibility interval region was mainly inferred from ancestral haplotype analysis, so that the re®ned intervals reported in Fig 3 correspond to the minimal conserved haplotypes observed by the authors. The PSORS1 susceptibility gene does not necessarily lie within conserved haplotype regions, however. Like all crossover-de®ned loci, the minimal PSORS1 interval is actually delimited by the position of the two nearest recombining markers. If the re®ned interval is extended to include the closest markers that are not conserved along the ancestral haplotype, then the size of minimal regions increases.
The interpretation of the Oka study is more problematic as the analysis of unrelated individuals did not allow haplotypes to be derived. The authors assumed that PSORS1 lies in the region where both statistics yielded signi®cant p-values, but signi®cance at a single test is maintained for some distance, on both sides of such interval. With regard to the Nair study, only two markers separate the RH1 and RH2 haplotypes. Taken together, this region extends over a 170 kb conserved interval, closely corresponding to the peak of LD identi®ed by the TDT. Double recombination events around the M6S111 and M6S169 markers would be required to account for the separation of RH1 and RH2. Given that the authors report ®ve distinct haplotypes carrying both RH1 and RH2, several such double recombinations would have to be assumed. An alternative hypothesis suggests the existence of a single RH haplotype carrying two microsatellite markers that have undergone a CA slippage mutation (Timsit, 1999). This hypothesis is consistent with the observation that the two most common risk haplotypes share the same M6S169 alleles and carry M6S111 alleles differing for a single CA repeat. Overall, this interpretation would support a larger PSORS1 interval, extending to include the 50 kb segment harbouring M6S111 and M6S169 (Fig 3). If the extended regions are aligned, some overlap between the various studies can be detected (Fig 3). Indeed the comparison between the different minimal intervals points to a 100 kb segment telomeric to HLA-C, including OTF3, TCF19, and a-helix coiled-coil rod homolog (HCR) genes. Many issues related to the re®nement methodology, however (e.g., reliability of some test statistics), are still a matter of considerable debate among workers in the ®eld. Thus, the 100 kb interval described above can only be considered a provisional assignment, and candidate genes lying in its proximity cannot be excluded only on the basis of their position. CANDIDATE GENE ANALYSES Investigations of PSORS1 candidate genes have been accumulating at an increasing pace as the evidence for an MHC susceptibility gene grew ever more robust. The ®rst and most extensively investigated candidate was HLA-C. The association between psoriasis and the HLA-Cw6 antigen has been observed in a wide range of ethnic groups, and there is a general correlation between
VOL. 118, NO. 5 MAY 2002
the prevalence of the disease and that of Cw6 (Mallon et al, 1999). On this basis, several groups have carried out HLA-C molecular analyses in psoriasis families. These studies have repeatedly indicated the HLA-Cw*0602 allele as the most signi®cant marker for disease risk prediction (Tazi Ahnini et al, 1999; Asumalahti et al, 2000; Enerback et al, 2000). Three different Cw6-positive haplotypes that are associated with psoriasis (Cw6-B57, Cw6B13, and Cw6-B37), however, have been described (Elder et al, 1994). These three haplotypes confer different disease risks (Jenisch et al, 1998), and HLA-C seems to lie outside the most likely PSORS1 interval (Fig 3). Taken together with the lack of a characterized functional role in psoriasis, HLA-Cw6 is widely assumed be a marker in LD with the PSORS1 gene rather than the susceptibility allele itself (Allen et al, 1999; Asumalahti et al, 2000; Nair et al, 2000). A second, extensively investigated candidate is the corneodesmosin (CDSN) gene (OMIM *602593). CDSN lies 160 kb telomeric to HLA-C (Fig 3) and codes for corneodesmosin, a desmosomal protein involved in keratinocyte cohesion and desquamation (Guerrin et al, 1998). Interestingly, a recent study has demonstrated CDSN overexpression and altered distribution in psoriatic epidermis (Allen et al, 2001). These are disease-speci®c changes, as they have not been observed in other in¯ammatory skin disorders such as atopic dermatitis and lichen planus (Allen et al, 2001). The CDSN gene is remarkably polymorphic and a singlenucleotide polymorphism (SNP) frequency of up to 1 each 100 bp has been observed (Enerback et al, 2000). Most CDSN variants have been tested for association, and several studies have reported signi®cant p-values for a coding SNP that causes a Ser®Leu substitution (Allen et al, 1999; Jenisch et al, 1999; Tazi Ahnini et al, 1999; Enerback et al, 2000). This association was not observed in certain ethnic groups, however, including Japanese, Finnish, and Spanish (Ishihara et al, 1996; Asumalahti et al, 2000; Gonzalez et al, 2000). These discrepancies might still be accounted for by allelic heterogeneity, however. As CDSN polymorphism content is so high, SNP haplotypes rather than single alleles are likely to in¯uence the protein structure. In this context, the effect of a variant such as the Ser®Leu substitution could be reinforced or neutralized, depending on the alleles that are in LD with it, in each population. In conclusion, CDSN still remains an attractive candidate gene because of its function, its skin-speci®c expression pattern, and its alterations in psoriatic epidermis. At this stage, a functional study of the psoriasis-associated CDSN SNP is clearly needed. In particular, in vivo or in vitro experiments are required in order to investigate the effects of the Ser/Leu variant on protein expression and localization. A third candidate, the HCR gene, has lately been under intense scrutiny. HCR (OMIM *605310) was ®rst identi®ed as a putative structural component of skin-related tissues, as it presented trichohyalin, myosin, and laminin homologies (Guilladeux et al, 1998). Asumalahti et al (2000) demonstrated that the gene is ubiquitously expressed but is upregulated in psoriatic epidermis. They also tested the HCR as a PSORS1 candidate by analysing 17 intragenic SNPs in a sample of Finnish origin. Their study disclosed a signi®cant association between psoriasis and a two-SNP haplotype bearing two Arg®Trp substitutions (HCR*Trp-Trp allele). O'Brien et al (2001) subsequently analyzed HCR in a smaller data set of Swedish origin. They replicated the association with the *Trp-Trp allele and observed signi®cant p-values for six additional SNPs, including a third Arg®Trp substitution. When they strati®ed their results according to Cw*0602 presence, however, association was maintained only for one silent polymorphism. On the other hand, stratifying for HCR SNPs did not affect association with Cw*0602. On this basis, the authors concluded that the association with the *Trp-Trp allele was solely due to LD with Cw*0602 and therefore had no pathologic relevance. It could be argued that Cw*0602 is not an SNP but a haplotype, however, as its sequence is de®ned by several nucleotide changes along the HLA-C gene. So, data strati®ed according to Cw*0602 presence
PSORIASIS SUSCEPTIBILITY AND THE MHC
749
may not be comparable to data strati®ed according to HCR single SNPs. A second report questioning HCR candidacy was published by Chia et al (2001). The paper shows that the HCR *Trp-Trp allele lies on most but not all the at-risk haplotypes identi®ed by Nair et al (2000) in their PSORS1 re®nement study. In particular, a Cw8B65 haplotype is described that does not carry the associated *TrpTrp allele. The p-values that originally demonstrated the preferential transmission of the Cw8-B65 haplotype, however, are on the threshold of statistical signi®cance (Nair et al, 2000), and even if the Cw8-B65 haplotype contributes to psoriasis pathogenesis the absence of the *Trp-Trp allele could still be accounted for by allelic heterogeneity. Chia et al also disputed the HCR pathogenic role based on the observation of a common nonrisk haplotype carrying HLA-Cw7 and the *Trp-Trp allele. This ®nding can be explained by low penetrance of the risk allele, however, a feature already noted in other complex disease susceptibility genes (Reich and Lander, 2001). Indeed, the effects of the *Trp-Trp allele could be modulated by haplotype-speci®c SNPs, as hypothesized for CDSN. In this context, it would be interesting to investigate whether the nonrisk Cw7 haplotype also carries the additional HCR risk SNPs identi®ed by O'Brien et al (2001). In conclusion, the data generated so far do not allow the dismissal of HCR as a PSORS1 candidate: the analysis of additional data sets and the functional characterization of the *Trp-Trp allele are still needed, in order to elucidate HCR contribution to psoriasis susceptibility. Other genes lying in the proximity of HLA-C have been analyzed, albeit less systematically. Gonzalez et al (2000) reported an association with one allele of the OTF3 gene in the Spanish population. The authors themselves acknowledge, however, that OTF3 is a somewhat unlikely candidate for disease susceptibility, as the gene encodes a transcriptional factor playing a major rule in embryonic stem cells lineage commitment (Niwa et al, 2000). Teraoka et al (2000) analyzed the TCF19 gene (Fig 3), coding for a cell growth regulator, ubiquitously expressed in the G1±S phases of the cell cycle. They identi®ed three SNPs and one deletion polymorphism, but failed to observe any association. Our group has recently been investigating SPR1 and SEEK1 genes (Fig 3). Direct sequencing of SPR1 failed to detect any highfrequency SNP within this gene (Munro et al, unpublished observations). WHERE DO WE GO FROM HERE? So far, none of the analyzed candidates seems to match the criteria for a disease susceptibility gene (see below). The PSORS1 region may contain as yet uncharacterized transcripts, e.g., the MHC sequence annotation predicts three open reading frames in the interval between HLA-C and OTF3 (Ensembl database of annotated genomic sequences, http://www.ensembl.org). Interestingly, these open reading frames match expressed sequence tags that map to the same region. No gene has yet been characterized, however. This may in part be due to the complexity of the target genomic region, containing a large number of repeats and retrovirus-like elements. On the other hand, these expressed sequence tags may represent expressed pseudogenes. To date, few susceptibility genes have been characterized; hence criteria for declaring disease causality are mostly based on theoretical considerations (Terwilliger and Weiss, 1999; Risch, 2000). For instance, it is widely assumed that the genetic contribution to complex disorders is due to relatively high frequency SNPs with subtle effects on protein function (Chakravarti, 1999). This did not prove to be the case for Crohn's disease, however, in which a lowfrequency frameshift mutation of the NOD2 gene has recently been implicated in disease susceptibility (Hugot et al, 2001; Ogura et al, 2001). Likewise, replication of association has long been considered an essential requirement for susceptibility gene assignment. This implies that the same alleles determine disease susceptibility in distinct populations. The real picture is likely to be more complex, however. For instance, it has been demonstrated
750
CAPON ET AL
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
that the CAPN10 haplotype1 involved in type II diabetes susceptibility confers very different risks in different ethnic groups (Horikawa et al, 2000). Another implicit assumption underlying most association studies is that disease susceptibility is caused by a single or a few closely linked SNPs. Again, this is not necessarily the case. Variation in adjacent genes may underlie disease susceptibility, as is the case with the IDDM superlocus lying within the MHC (Lie et al, 1999). More generally, the effect of single SNPs may vary according to the haplotype they reside on. Indeed this has proven to be the case even for a ``simple'' monogenic disorder like cystic ®brosis, where CFTR mutation effects can be modulated by intronic variants (Estivill, 1996). If similar mechanisms were active in the PSORS1 region, identifying psoriasis susceptibility alleles by association studies alone could prove a formidable task indeed. Despite these dif®culties, susceptibility gene assignment may still be achieved by assessing the functional relevance of diseaseassociated SNPs. Indeed, demonstrating that a given SNP affects gene expression levels or protein secondary structure is expected to provide robust support for the involvement of a candidate gene. This is likely to be particularly true for the PSORS1 locus. The con¯icting interpretations of recent association ®ndings show how genetic analysis may be close to its limit, suggesting that further resolution cannot be achieved owing to LD conservation along the MHC. Before genetic analysis is set aside, however, it will be important to ensure that a near complete assessment of the ancestral history of the numerous genetic variants within the PSORS1 region has been achieved. The MHC is unusual in the way extended regions of LD have been maintained, probably re¯ecting complex mechanisms that ensure such genomic regions remain intact. The availability of a high-resolution genetic map of this region allows the detailed characterization of highly localized ``hot spots'' of recombination within the MHC with important implications for disease causality (Jeffreys et al, in press). These studies will need to be assessed in the light of such analysis. Together with the rich and ¯exible array of functional analysis techniques, researchers have now at their disposal adequate tools to identify the long-hunted-for PSORS1 gene. Note added in proof: Asumalahti et al have recently reported the analysis of HLA-C/HCR/CDSN risk haplotypes in ethnically diverse data sets, con®rming highly signi®cant association of psoriasis with HLA-CW*0602 and HCR*TRP-TRP alleles (Hum Mol Genet 11:589±597, 2002). The authors wish to thank the many psoriasis family members for support, and the National Psoriasis Foundation and the Wellcome Trust for ®nancial aid. FC is a Wellcome Trust Travelling Fellow.
REFERENCES Allen MH, Veal C, Faassen A, Powis SH, Vaughan RW, Trembath RC, Barker JNWN: A non-HLA gene within the MHC in psoriasis. Lancet 353:1589± 1590, 1999 Allen MH, Ishida-Yamamoto A, McGrath J, et al: Coneodesmosin gene expression in psoriasis vulgaris differs from normal skin and other in¯ammatory skin disorders. Lab Invest 81:969±976, 2001 Altshuler D, Hirschorn JN, Klannemark M, et al: The common PPAR Pro12Ala polymorphism is associated with decreased risk of type 2 diabetes. Nature Genet 26:76±80, 2000 Asumalahti K, Latinen T, Itkonen-Vatjus R, et al: A candidate gene for psoriasis near HLA-C, HCR (Pg8), is highly polymorphic with a disease associated susceptibility allele. Hum Mol Genet 9:1533±1542, 2000 Balendran N, Clough RL, Arguello JR, et al: Characterization of the major susceptibility region for psoriasis at chromosome 6p21.3. J Invest Dermatol 113:322±328, 1999 Barker JN: The pathophysiology of psoriasis. Lancet 338:227±230, 1991 1 Nair RP, Chia N, Shahabi L, et al: Sequence variation in a retroviruslike expressed sequence tag in the PSORS1 critical of the MHC: lack of evidence for a causative psoriasis gene. Society of Investigative Dermatology 62nd Annual Meeting. Washington, May 9±12, 2001.
Bhalerao J, Bowcock A: The genetics of psoriasis: a complex disorder of the skin and immune system. Hum Mol Genet 7:1537±1545, 1998 Burden AD, Javed S, Bailey M, Hodgins M, Connor M, Tillman D: Genetics of psoriasis: paternal inheritance and a locus on chromosome 6p. J Invest Dermatol 110:958±960, 1998 Capon F, Novelli G, Semprini S, et al: Searching for psoriasis susceptibility genes in Italy: genome scan and evidence for a new locus on chromosome 1. J Invest Dermatol 112:32±35, 1999a Capon F, Semprini S, Dallapiccola B, Novelli G: Evidence for interaction between psoriasis susceptibility loci on chromosomes 6p21 and 1q21. Am J Hum Genet 65:1798±1800, 1999b Chakravarti A: Population genetics making sense out of sequence. Nature Genet 21 (Suppl. 1):56±60, 1999 de la Chapelle A, Wright FA: Linkage disequilibrium mapping in isolated populations: the example of Finland revisited. Proc Natl Acad Sci USA 95:12416±12423, 1998 Chia NV, Stuart P, Nair RP, et al: Variations in HCR (Pg8) gene are unlikely to be causal for familial psoriasis. J Invest Dermatol 116:823±824, 2001 Christophers E, Sterry W: Psoriasis. In: Fitzpatrick TB, Eisen AZ, Wolff K, Austin KF, eds. Dermatology in General Medicine. New York: McGraw-Hill, 1993 Cookson WOCM, Ubhi B, Lawrence R, et al: Genetic linkage of childhood atopic dermatitis to psoriasis susceptibility loci. Nature Genet 27:372±373, 2001 Elder JT, Henseler T, Christophers E, Voorhees JJ, Nair RP: Of genes and antigens: the inheritance of psoriasis. J Invest Dermatol 103:150S±153S, 1994 Enerback C, Martinsson T, Inerot A, et al: Signi®cantly earlier age at onset for the HLA-Cw6-positive than for the Cw6-negative psoriatic sibling. J Invest Dermatol 109:695±696, 1997 Enerback C, Enlund F, Inerot A, Samuelsson L, Wahlstrom J, Swanbeck G, Martinsson T: S gene (corneodesmosin) diversity and its relationship to psoriasis; high content of cSNP in the HLA-linked S gene. J Invest Dermatol 114:1158±1163, 2000 Enlund F, Samuelsson L, Enerback C, et al: Analysis of three suggested psoriasis susceptibility loci in a large Swedish set of families: con®rmation of linkage to chromosome 6p (HLA region), and to 17q, but not to 4q. Hum Hered 49:2±8, 1999a Enlund F, Samuelsson L, Enerback C, et al: Psoriasis susceptibility locus in chromosome region 3q21 identi®ed in patients from southwest Sweden. Eur J Hum Genet 7:783±790, 1999b Estivill X: Complexity in a monogenic disease. Nature Genet 12:348±350, 1996 Gonzalez S, Martinez Borra J, Del Rio JS, Santos-Juanes J, Lopez-Vazquez A, Blanco-Gelaz M, Lopez-Larrea C: The OTF3 gene polymorphism confers susceptibility to psoriasis independent of the association of HLA-Cw*0602. J Invest Dermatol 115:824±828, 2000 Guerrin M, Simon M, Montezin M, Haftek M, Vincent C, Serre G: Expression cloning of human corneodesmosin proves its identity with the product of the S gene and allows improved characterization of its processing during keratinocyte differentiation. J Biol Chem 273:22640±22647, 1998 Guilladeux T, Janer M, Wong GKS, Spies T, Geraghty DE: The complete genomic sequence of 424,015 bp at the centromeric end of the HLA class I region: gene content and ploymorphism. Proc Natl Acad Sci USA 95:9494±9499, 1998 Horikawa Y, Oda N, Cox NJ, et al: Genetic variation in the gene encoding calapain10 is associated with type 2 diabetes mellitus. Nature Genet 26:163±175, 2000 Hugot JP, Chamaillard M, Zouali H, et al: Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn's disease. Nature 411:599±603, 2001 Ishihara M, Yamagata N, Ohno S, et al: Genetic polymorphisms in the keratin-like S gene within the human major histocompatibility complex and association analysis on the susceptibility to psoriasis vulgaris. Tissue Antigens 48:182±186, 1996 Jeffreys AJJ, Kauppi L, Neumann R: Intensely punctate meiotic recombination in the class II region of the major histocompatibility complex. Nature Genet, 29:217± 222, 2001 Jenisch, S, Henseler T, Nair RP, et al: Linkage analysis of human leukocyte antigen (HLA) markers in familial psoriasis: strong disequilibrium effects provide evidence for a major determinant in the HLA-B/-C region. Am J Hum Genet 63:191±199, 1998 Jenisch S, Koch S, Henseler T, et al: Corneodesmosin gene polymorphism demonstrates strong linkage disequilibrium with HLA and association with psoriasis vulgaris. Tissue Antigens 54:439±449, 1999 Jorde LB: Linkage disequilibrium as a gene-mapping tool. Am J Hum Genet 56:11±14, 1995 Latinen T, Daly MJ, Rioux JD, et al: A susceptibility locus for asthma-related traits on chromosome 7 revealed by genome-wide scan in a founder population. Nature Genet 28:87±91, 2001 Leal SM, Ott J: Effects of strati®cation in the analysis of affected-sib-pair data: bene®ts and costs. Am J Hum Genet 66:567±575, 2000 Lee YA, Ruschendorf F, Windemuth C, et al: Genomewide scan in German families reveals evidence for a novel psoriasis-susceptibility locus on chromosome 19p13. Am J Hum Genet 67:1020±1024, 2000 Lie BA, Todd JA, Pociot F, et al: The predisposition to type 1 diabetes linked to the human leukocyte antigen complex includes at least one non-class II gene. Am J Hum Genet 64:793±800, 1999 Mallon E, Newson R, Bunker CB: HLA-Cw6 and the genetic predisposition to psoriasis: a meta analysis of published serologic studies. J Invest Dermatol 113:693±695, 1999 Matthews D, Fry L, Powles A, et al: Evidence that a locus for familial psoriasis maps to chromosome 4q. Nature Genet 14:231±233, 1996 McGuf®n P, Sargeant M, Hetti G, Tidmarsh S, Whatley S, Marchbanks RM:
VOL. 118, NO. 5 MAY 2002
Exclusion of a schizophrenia susceptibility gene from the chromosome 5q11q13 region: new data and a reanalysis of previous reports. Am J Hum Genet 47:524±535, 1990 Nair RP, Henseler T, Jenisch S, et al: Evidence for two psoriasis susceptibility loci (HLA and 17q) and two novel candidate regions (16q and 20p) by genomewide scan. Hum Mol Genet 6:1349±1356, 1997 Nair RP, Stuart P, Heneseler T, et al: Localization of psoriasis susceptibility locus PSORS1 to a 60 kilobase interval telomeric to HLA-C. Am J Hum Genet 66:1833±1844, 2000 Niwa H, Miyazaki J, Smith AG: Quantitative expression of Oct-3/4 de®nes differentiation, de-differentiation or self-renewal of ES cells. Nature Genet 24:372±376, 2000 O'Brien KP, Holm SJ, Nilsson S, et al: The HCR gene on 6p21 is unlikely to be a psoriasis susceptibility gene. J Invest Dermatol 116:750±754, 2001 Ogura Y, Bonen DK, Inohara N, et al: A frameshift mutation in NOD2 associated with susceptibility with Crohn's disease. Nature 411:603±606, 2001 Oka A, Tamiya G, Tomizawa M, et al: Association analysis using re®ned microsatellite markers localizes a susceptibility locus for psoriasis vulgaris within a 111kb segment telomeric to the HLA-C gene. Hum Mol Genet 8:2165±2170, 1999 Pfaff CL, Parra EJ, Bonilla C, et al: Population structure in admixed populations: effect of admixture dynamics on the pattern of linkage disequilibrium. Am J Hum Genet 68:198±207, 2001 Reich DE, Lander ES: On the allelic spectrum of human disease. Trends Genet 17:502±511, 2001 Rioux JD, Daly MJ, Silverberg MS, et al: Genetic variation in the 5q31 cytokine gene cluster confers susceptibility to Crohn disease. Nature Genet 29:223±228, 2001 Risch NJ: Searching for genetic determinants in the new millennium. Nature 405:847±856, 2000 Samuelsson L, Enlund F, Torinsson A, et al: A genome-wide search for genes
PSORIASIS SUSCEPTIBILITY AND THE MHC
751
predisposing to familial psoriasis by using a strati®cation approach. Hum Genet 105:523±529, 1999 Schaid DJ: Transmission disequilibrium test, family controls and great expectations. Am J Hum Genet 63:935±942, 1998 Schmitt-Egenolf M, Eiermann TH, Boehncke WH, Stander M, Sterry W: Familial juvenile onset psoriasis is associated with the human leukocyte antigen (HLA) class I side of the extended haplotype Cw6±B57-DRB1*0701-DQA1*0201DQB1*0303: a population- and family-based study. J Invest Dermatol 106:711± 714, 1996 Spielman RS, McGinnis RE, Ewens WJ: Transmission test for linkage disequilibrium: the insulin gene region and insulin dependent diabetes mellitus (IDDM). Am J Hum Genet 52:506±516, 1993 Tazi Ahnini R, Camp NJ, Cork NJ, Mee JB, Keohane SG, Duff GW, Di Giovine FS: Novel genetic association between the corneodesmosin (MHC S) gene and susceptibility to psoriasis. Hum Mol Genet 8:1135±1140, 1999 Teraoka Y, Naruse TK, Oka A, et al: Genetic polymorphisms in the cell growth regulated gene, SC1 telomeric of the HLA-C gene and lack of association of psoriasis vulgaris. Tissue Antigens 55:206±211, 2000 Terwilliger JD, Weiss KM: Linkage disequilibrium mapping of complex disease: fantasy or reality? Curr Opin Biotechnol 9:578±594, 1999 Timsit Y: DNA structure and polymerase ®delity. J Mol Biol 293:835±853, 1999 Tomfohrde J, Silverman A, Barnes R, et al: Gene for familial psoriasis susceptibility mapped to the distal end of chromosome 17q. Science 264:1141±1145, 1994 Trembath RC, Clough RL, Rosbotham JL, et al: Identi®cation of a major susceptibility locus on chromosome 6p and evidence for further disease loci revealed by a two stage genome-wide search in psoriasis. Hum Mol Genet 6:813±820, 1997 Veal CD, Clough RL, Barber RC, et al: Identi®cation of a novel psoriasis susceptibility locus at 1p and evidence of epistasis between PSORS1 and candidate loci. J Med Genet 38:7±13, 2001