FC22 Assessment of pustular psoriasis genes in generalized and palmoplantar pustular psoriasis suggest transethnic differences in disease susceptibility, an oligogenic inheritance in both psoriatic subtypes and an involvement of purinergic receptors

Abstract Generalized (GPP) and palmoplantar pustular psoriasis (PPP) are psoriatic subtypes affecting fewer psoriatic patients but often more severely than common psoriasis vulgaris. While several susceptibility genes have been identified in GPP mainly in Africa and Europe, there is currently a lack of established risk factors for PPP. Recently, P2RX7 variants were suggested to contribute to chronic non-bacterial osteomyelitis (CNO), a clinically overlapping disease. In order to estimate the effect sizes of susceptibility variants in GPP and CNO genes, we assessed 195 exome/ genome sequences of GPP and PPP systematically and performed a literature search for published datasets of European pustular psoriasis patients. Depending on the reported state of disease-variants, we compared allele or genotype frequencies to those of controls. In the case of bi-allelic variants, we considered ∼4900 individuals in whom allele dosage could be determined. Meta-analysis in GPP revealed significant effect sizes for IL36RN and MPO variants in European carriers of mono-, but especially of bi-allelic variants [IL36RN: odds ratio (95% confidence interval) = 334.2 (106.5–1599.2); MPO: 73.7 (16.5–328.5)], as expected for a monogenic/ oligogenic disease like GPP. Nine PPP patients (8%) carried a coding variant in either IL36RN or MPO, one additional individual with disease-contributing variants in both IL36RN and MPO suggests that oligogenic inheritance might also be relevant in PPP. Not a single PPP patient, but five GPP patients were identified to carry P2RX7 missense variants predicted/shown to have an impact on the signalling pathway; notably, one of the latter patients carried two variants, suggesting autosomal recessive inheritance. Putative disease variants in AP1S3 were not associated, while SERPINA3 variants were too rare to perform reliable comparisons. We did not analyse CARD14, as there are numerous missense variants and the classification of many missense variants into different ACMG categories is rather arbitrary. Our study indicates a continuum of GPP-PPP-CNO and a significant contribution of IL36RN and MPO variants on disease susceptibility in bi-allelic states and in European GPP patients, less commonly in PPP. The lack of association of pustular psoriasis with BTN3A3 in European patients and with MPO variants in Asian patients suggests transethnic differences in disease susceptibility. The overall allele frequencies of up to 4.5% of variants in AP1S3 and BTN3A3 in certain populations and a considerable frequency of homozygous, healthy carriers of risk alleles in BTN3A3 and AP1S3 render an impact on disease susceptibility less likely. As variants in known disease genes affect <40% of GPP patients and a low percentage in PPP, there is a high need to elucidate the molecular basis of pustular psoriasis, in order to offer more targeted, individualized therapy.