Genetics of Psoriasis and Psoriatic Arthritis: Update and Future Direction (GRAPPA 2007)

Abstract

Psoriasis and psoriatic arthritis (PsA) both have substantive genetic determinants. Numerous candidate regions and genes have now been replicated in disease susceptibility, and to a lesser extent in disease expression, in both disease entities. Intensive efforts are now underway or are being actively planned to perform genome-wide association scans (GWAS) in psoriasis and PsA. A major determinant of success for GWAS is likely to be accumulation of multiple large well-phenotyped cohorts, sophisticated data management, and verification of the findings.

At the 2007 Annual Meeting of the Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA), members of the GRAPPA genetics committee presented a discussion of the genetics of psoriasis and PsA, including future trends. This paper is a summary of that presentation and a review of the literature.

GENETICS OF PSORIASIS

Genetic factors have long been recognized to play an important role in psoriasis. The heritability of psoriasis was first described 200 years ago, evidenced by familial clustering of disease and later by demonstrating increased concordance in monozygotic twins vs. dizygotic twins.^{1, 2} Psoriasis has a complex, multifactorial genetic basis, and this concept has only been strengthened by the discoveries of over 20 candidate loci, using linkage analysis, and more recently, genome-wide association scans (GWAS).

Major Histocompatibility Complex (MHC) and Psoriasis Susceptibility

PSORS1 and HLA-C

The major genetic determinant of psoriasis is believed to reside in an approximately 300kb-segment in the MHC I region on chromosome 6p21.3 known as PSORS1. Over 30 years ago, this region was found to harbor human leukocyte antigen (HLA) genes that associated with autoimmune diseases. Psoriasis was found to be associated with HLA-C and several HLA-B alleles;³ however, the association with HLA-B was later determined to be due to strong extended haplotypes and linkage disequilibrium with HLA-C.⁴ This region was subsequently identified by linkage analysis in 1997^{5, 6} and replicated in numerous populations. Candidate genes just telomeric to HLA-C were appealing since several (CDSN, HCR, and PSORS1C3) are expressed in skin. However, none of the candidates are convincingly associated with psoriasis independent of HLA-C. Extensive study of this segment has been led by Elder and colleagues using recombinant ancestral haplotypes.⁷ Although a 70-kb risk segment telomeric (and excluding) HLA-C initially was believed to confer the most risk,⁸ an international collaborative study extended the risk segment to 300-kb span from just telomeric to HLA-B to beyond CDSN, thus including HLA-C.⁷ After sequencing this segment in 2 risk and 5 non-risk chromosomes, then examining recombinant haplotypes retaining HLA-Cw6 but lacking risk alleles in CDSN, Nair et al concluded that HLA-Cw6 is the PSORS1 risk variant that confers susceptibility to psoriasis.⁹

HLA-C and Disease Expression

A specific allele of the HLA-C region, HLA-Cw*0602, is also the only genetic variant repeatedly observed to associate with phenotypic features of psoriasis. Patients carrying this allele typically have early onset, higher incidence of guttate or streptococcal-induced flares of disease,¹⁰ Koebnerization, and a more severe course. Homozygosity for HLA-Cw*0602 predisposes to the likelihood of development of psoriasis and to earlier onset, but it otherwise does not impact clinical course.¹¹ Women carrying HLA-Cw*0602 are more likely to experience remission with pregnancy.¹² HLA-Cw*0602 is less frequent in patients with PsA (20%)¹³ and does not appear to be a risk factor for later onset of psoriasis (type II), palmar-plantar pustular disease, nail disease, or scalp disease.^{10, 14, 15}

Functional Role of HLA-C

Despite its repeated genetic association with psoriasis, limited data exist to explain the functional role of HLA-C in psoriasis pathogenesis. In vitro studies have suggested that compared to the CD8+ T cells from HLA-Cw6 negative individuals, CD8+ T cells from HLA-Cw6-positive individuals are more responsive to peptides found in both the hyperproliferative keratin K17 and streptococcal M protein, suggesting that HLA-Cw6 may predispose individuals to recognize keratin self-antigens.¹⁶ Responses were 10-fold higher in T cells expressing cutaneous lymphocyte-associated (CLA-positive) skin-homing receptors than in CLA-negative T cells, demonstrating that these responses are targeted to the skin.

HLA-C also serves as a ligand for killer immunoglobulin-like receptors (KIR) on natural killer (NK) and natural killer-T (NK-T) cells, which may also have a role in psoriasis.¹⁷ Inheritance of activating KIRs, encoded on chromosome 19q13.4, particularly KIR2DS1 and KIR2DS2, have been associated with psoriasis,¹⁸ and lack of inhibitory KIRs or their corresponding HLA-C ligand have been associated with the development of PsA.¹⁹ However, this function of HLA-C appears unlikely to account for the strong associations between psoriasis and HLA-Cw6, as several other HLA-C alleles manifest the same binding specificity for KIRs.

Susceptibility Loci Outside the MHC

Although the PSORS1 locus is generally understood to confer the most risk for psoriasis, numerous susceptibility loci also have been identified outside of the MHC region. Linkage scans were used to identify and replicate the intervals designated PSORS2-PSORS9, as previously reviewed.²⁰ Although dense microsatellite markers and sequencing within these loci have identified candidates, lack of replication of the specific risk variants, and lack of a clear role of variants that do not lie within functional genes, has slowed our understanding of the magnitude of the contribution and the relevance of these loci. GWAS using single nucleotide polymorphism (SNP) technology have identified new candidates within and outside of linkage peaks. The PSORS intervals and their candidate genes are summarized in Table 1.

Table 1.

Psoriasis Susceptibility Loci and Gene Candidates

Locus	Region	Gene Candidates/Function	Lead Author and Year of Publication of Psoriasis Susceptibility Studies*
PSORS1	6p21.3	HLA-Cw6 CDSN, HCR, HERV-K, HCG2, 7PSORS1C3, POU5F1, TCF19, CCHCR1, LMP, SEEK1, SPR1.	Samuelsson L, 1999; Lee YA, 2000; Elder JT, 2001; Veal CD, 2001; Zhang XJ, 2002; Foerster J, 2004; Sagoo GS, 2004
PSORS2	17q25	RUNX1 RAPTOR SLC9A3R1 NAT9 TBCD	Tomfohrde J, 1994; Nair RP, 1997; Enlund F, 1999; Samuelsson L, 1999; Helms C, 2003; Zheng Y, 2003; Capon F, 2004; Stuart P, 2006; Capon F, 2007
PSORS3	4q34	IRF-2≠	Matthews D, 1996; Hida S, 2000; Foerster J, 2004
PSORS4	1q21	Loricrin≠ Filaggrin≠ Pglyrp3,4≠ S100 genes within epidermal differentiation complex	Bhalerao J, 1998; Capon F, 1999; Semprini S, 2002; Giaardian E, 2004; Giardina E, 2006; Sun C, 2006; Zhao Y, 2007
PSORS5	3q	SLC12A8 Cystatin A≠ Zn finger protein 148≠	Enlund F, 1999; Samuelsson L, 1999; Hewett, 2002; Samuelsson L, 2004; Huffmeier U, 2005
PSORS6	19p13	JunB	Lee YA, 2000; Zenz R, 2005
PSORS7	1p	PTPN22≠ (1p13) IL-23R (1p32.1–31.2)	Veal CD, 2001; Tsunemi Y, 2002; Nistor I, 2005; Duerr RH, 2006; Huffmeier U, 2006; Capon F, 2007; Cargill M, 2007
PSORS8	16q	CX3CL1, CX3R1 NOD2/CARD15≠	Nair RP, 1997; Karason A, 2003; Young C, 2003; Plant D, 2006
PSORS9	4q31	IL-15	Bhalerao J, 1998; Samuelsson L, 1999; Zhang XJ, 2002; Bowcock AM, 2004; Sagoo GS, 2004; Sun LD, 2007; Zhang XJ, 2007
PSORS10	18p11		Veal CD, 2001; Asumalahti K, 2003
-	5q31.1– 33.1	IL-12B SLC22A4≠ SLC22A5≠ IL-13 IL3, IL4, IL5, CSF2 and IRF1	Tsunemi Y, 2002; Duerr RH, 2006; Friberg C, 2006; Capon F, 2007; Cargill M, 2007, Nair et al, 2008
-	9q33– 34		Zhang XJ, 2002; Yan KL, 2007
-	6p22	CDKALI	Wolf N, 2007
	19q34	KIR2DS1, KIR2DL1, KIR2DL5	Suzuki Y, 2004; Luszczek W, 2004

^*Full citations NOT included in Reference list.

^≠Candidate genes investigated and not believed to confer risk of psoriasis

Results of Genome-wide Association Studies

IL-12 and IL-23

Perhaps the most compelling new gene candidates for psoriasis to date, IL12B and IL23R, have been identified using GWAS rather than linkage analysis.²¹ Using a 25,215 gene-centric SNP platform for discovery and follow up tag SNP and sequencing, this study confirmed a previously reported psoriasis-associated SNP in the IL12B 3’ untranslated-region (rs3212227)²² and found a second SNP (rs6887695) located 60 kb upstream.²¹ This study also identified 2 missense SNPs in IL23R that associated with psoriasis, one of which (rs11209026, Arg381Gln) is also associated with Crohn’s disease.²³ Both the IL12B and IL23R SNPs have since been replicated in 2 United Kingdom psoriasis populations and in a study of United States and German families and cases and controls (see Table 1). The functional relevance of these SNPs remains unclear, but interleukin-12 (IL-12) and interleukin-23 (IL-23), a complex of the p19 and p40 subunits, have a key role in the pathogenesis of psoriasis: IL-12 stimulates interferon (IFN)-γ in naïve Th cells, and IL-23 stimulates IFN-γ production, proliferation of memory Th1 cells, and has a role in the recently described Th17 pathway. The p40 subunit is increased in psoriatic lesions,²⁴ and neutralization of p40 with a human monoclonal antibody causes marked improvement of psoriasis.²⁵

GENETICS OF PSORIATIC ARTHRITIS

Epidemiological evidence implicates a strong genetic basis to PsA. Moll and Wright were the first to demonstrate familial aggregation of PsA, and estimated the recurrence risk ratio in first degree relatives (λ₁) to be 55,²⁶ compared with estimates ranging from 5–10 in cutaneous psoriasis. More recent studies have estimated the λ₁ to be 47 in a British population²⁷ and 30.4 in a Canadian population.²⁸

PsA and Genes within the MHC Region

Polymorphisms in the genes coded in the HLA region on chromosome 6p have been shown to be associated with PsA. Class I antigens (HLA-B13, HLA-B57, HLA-B39, HLA-Cw6, HLA-Cw7) have consistently shown a positive association with psoriasis and PsA in population studies, with the strongest association being with HLA-Cw6.²⁹ HLA antigens may also identify patients with a particular pattern of PsA: HLA-B27 with spinal involvement, B38, and B39 with peripheral polyarthritis.

HLA antigens were identified as prognostic factors in patients with PsA.²⁹ HLA-B39 alone, HLA-B27 in the presence of HLA-DR7, and HLA-DQw3 in the absence of HLA-DR7, each conferred an increased risk for disease progression. HLA-B22 was found to be protective for disease progression.²⁹ The “rheumatoid arthritis (RA) shared epitope” was found to be associated with radiological erosions among patients with PsA.³⁰ Recently, patients with PsA carrying both HLA-Cw6 and HLA-DRB1*07 alleles were found to have a less severe course of arthritis.³¹

There are conflicting reports on the associations of TNFα polymorphisms located on chromosome 6p with PsA.³² A meta-analysis confirmed an association between TNFα -238G/A polymorphism and PsA with an odds ratio of 2.29 (95% confidence interval, 1.48–3.55).³² A recent study reported that TNFα-857C/T may represent a risk factor for PsA (but not for psoriasis) that is independent of the PSORS1 allele.³³

Class I MHC chain-related gene A (MICA) located 47 kb upstream of HLA-B also has been shown to be associated with PsA.^{34, 35} In a Spanish population, MICA-A9 polymorphism corresponding to the MICA-002 allele was associated with PsA (but not psoriasis), independent of HLA Cw*0602 (p<0.00035, relative risk 3.2).³⁴ Similar associations have been shown with Jewish,³⁵ Croatian,³⁶ and British patients³⁷ with PsA.

Susceptibility Loci for PsA Outside the MHC Region

Only one genome-wide linkage study in PsA has been published.³⁸ With respect to PsA-association studies outside the MHC region, a large number of candidate genes have been tested.^{39, 40} However, only a few genes have been independently replicated and will be reviewed below.

Chromosome 16q (via genome-wide linkage study)

The study was conducted in Iceland, where 178 patients with PsA were identified from 906 patients included in a genetic study of psoriasis.³⁸ A linkage with a LOD score of 2.17 was observed on 16q. When the linkage analysis was conditioned on paternal transmission to affected individuals, a LOD score of 4.19 was obtained, whereas a LOD score of only 1.03 was obtained when conditioned on maternal transmission. This locus is close to the PSORS8 locus identified for psoriasis.⁴¹

Chromosome 2q (IL-1 gene cluster)

The interleukin-1 gene cluster on chromosome 2q also has been investigated for association with PsA. An association has been reported with the IL-1α-889 SNP variant.⁴² A recent study of 29 SNPs at the IL-1 cluster also revealed 2 regions contributing independently to risk of PsA: a region spanned by markers rs3783547, rs3783543, and rs17561 in IL1A, and a region near the end of IL1B, through IL1F7, IL1F8, and into IL1F10.⁴³

Chromosome 19q13.4 (KIR genes)

The activating KIRs, KIR2DS1 and KIR2DS2, have been associated with PsA, particularly in the absence of the HLA ligands for the corresponding inhibitory KIRs (KIR2DL1 and KIR2DL2/3).^{19, 44} Furthermore, it was shown that the susceptibility to PsA may be determined by the overall balance of activating and inhibitory composite KIR-HLA genotypes.⁴⁵

PRESENT DIRECTION OF GENETIC STUDIES IN PSORIASIS AND PsA

At present, genetic association studies are at the forefront of genetic analysis. This is a result of the high density SNP arrays, markedly enhanced sample sizes, and a more affordable cost of high-throughput genotyping. The international HapMap project has also been instrumental in limiting the number of markers to be typed as a result of well-characterized linkage disequilibrium between the markers. Furthermore, as evidenced by Cargill et al, genome-wide pooling studies have been developed that decrease the cost of these investigations.²¹ GWAS appear to be bearing fruit as novel SNPs have been identified in multiple common diseases including Crohn’s disease, obesity, and prostate cancer.

Despite the recent success, many limitations still exist with genetic association studies. As evidenced by recent SNP associations with the interleukin-12 p40 subunit (IL12B) and the interleukin-23 receptor (IL23R), the genotype relative risk for these high-priority genes is quite modest, and these variants account for only a small proportion of the genetic risk.²¹ Much larger sample sizes are required for discovery of novel variants, and new findings should be replicated in numerous large independent cohorts such as the Genetics Association Information Network (GAIN), a public-private partnership created to facilitate GWAS of common human disease. The first phase of GAIN includes genotyping of 1500 psoriasis cases and 1500 controls for 600,000 SNPs. De-identified phenotype information from this study has been deposited in a database managed by the National Center for Biotechnology Information (NCBI) for access by the general research community, with access to genotypes restricted to authorized users who have applied for access and agreed to GAIN guidelines regarding confidentiality, intellectual property, and publication (http://www.fnih.org/GAIN2/home_new.shtml).

Once a genetic variant has been identified and replicated, however, the pathogenesis of the respective disease is not necessarily illuminated. In fact, most the variants being identified are in non-coding regions or belong to genes with unknown function. Functional verification of these results is likely to have the most meaningful impact and is of central importance. Other complexities that require further investigations are genotype/phenotype correlations and gene/environment interactions. For these studies, detailed clinical characterization is required along with sophisticated genetic analysis, due to the extensive data likely to be generated from testing of numerous clinical and environmental variables.

CONCLUSION

As in other multifactorial genetic disorders, the genetics of psoriasis and PsA are now coming into focus, powered by the collection of large case-control samples, advances in genotyping technology, and advanced statistical analysis. The emerging results are complemented by recent advances in immunology and therapeutics. While much remains to be done, the integration of genetics and immunology is becoming a reality for both psoriasis and PsA.