PTPN22 polymorphisms may indicate a role for this gene in atopic dermatitis in West Highland white terriers

Background Canine atopic dermatitis is an allergic inflammatory skin disease common in West Highland white terriers. A genome-wide association study for atopic dermatitis in a population of West Highland white terriers identified a 1.3 Mb area of association on CFA17 containing canine protein tyrosine phosphatase non-receptor type 22 (lymphoid) PTPN22. This gene is a potential candidate gene for canine atopic dermatitis as it encodes a lymphoid-specific signalling mediator that regulates T-cell and possibly B-cell activity. Findings Sequencing of PTPN22 in three atopic and three non-atopic West Highland white terriers identified 18 polymorphisms, including five genetic variants with a bioinformatically predicted functional effect. An intronic polymorphic repeat sequence variant was excluded as the cause of the genome-wide association study peak signal, by large-scale genotyping in 72 West Highland white terriers (gene-dropping simulation method, P = 0.01). Conclusions This study identified 18 genetic variants in PTPN22 that might be associated with atopic dermatitis in West Highland white terriers. This preliminary data may direct further study on the role of PTPN22 in this disease. Large scale genotyping and complementary genomic and proteomic assays would be required to assess this possibility.


Findings
Canine atopic dermatitis (AD) is an allergic inflammatory skin disease that is common in West Highland white terriers (WHWTs) [1]. Following a genome-wide association (GWAS) in a group of related WHWTs, we found a 1.3 Mb area on CFA 17 which was significantly associated with the disease [2]. Based on its biological functions, expression patterns and proximity to this area of association, PTPN22 was selected as a candidate gene for AD in this population. This gene encodes a lymphoid tyrosine phosphatase (PTPN22), a signalling mediator that regulates generic and specialised immune functions in mammals [3]. Activation of T and B lymphocytes is a key event in the pathogenesis of atopic disease [4], and the disruption of these pathways could cause hyper-reactive pathogenic T-cell responses, as well as affect B-cell selection, maturation and function [5,6]. In humans and dogs, genetic variants in the gene PTPN22 have been associated with auto-immune diseases [7][8][9]. In humans, these include psoriasis, a chronic immune-mediated inflammatory skin disease that shares susceptibility loci with human AD [10,11]. To date, no association has been found between PTPN22 variants and atopic disease in humans [12].
The University of Queensland Animal and Human Ethics Committees, and the University of Sydney Animal Ethics Committee approved this study. Written consent was obtained from all participating dog owners.
Criteria used to classify dogs in the present study are described elsewhere [1]. Fourteen set of primers were designed with primer3 [13], to sequence a total of 12.6 Kb of PTPN22 in 14 PCR products (Table 1). Amplification reactions used the HotStar HiFidelity PCR Kit (QIAGEN Pty Ltd, Doncaster, Vic, Australia) and 0.5 μM (PCR products 5 and 12), 1.5 μM (6 and 14) or 1 μM (remaining PCR products) of primers; at 55°C (PCR
Sequence data were analyzed with ChromasPro v1.5 (Technylisium, Tewantin, Qld, Australia) and compared with the 1.5× poodle (version 1) and the boxer 7.6× whole-genome sequences (CanFam2.0). Among 18 variants identified [14], five variants showed a medium to high disease-associated risk as predicted by FASTSNP [15] and Mutation Taster [16]; three single-nucleotide polymorphisms (SNPs) in a predicted regulatory region of the gene, one synonymous SNP, and a variable sequence repeat in a predicted splice site (Table 2). These variants formed five different haplotypes (Table 3). There were no recombinant events within this 12.6 Kb interval.
Variant sequence repeat c.2137-20 T(17_22) (Figure 1) has not been previously reported in dogs or other species and was bioinformatically predicted to have indirect structural effects on PTPN22. Comparable intronic repeat variations might interfere with normal gene expression [17][18][19] and have been associated with alternative splicing and disease in humans [20][21][22][23]. Thus, fluorescently labelled, amplified-fragment length genotyping of this variant was performed in 72 WHWTs, including 54 dogs from the GWAS. Primers and PCR conditions for amplification of PCR product 11 were used. Genotyping was performed on a 3130xl Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) and analyzed using Genemapper (Applied Biosystems, Foster City, CA, USA). SIB-PAIR [24] showed no significant evidence for allelic association between this variant and the trait (gene-dropping simulation method, P = 0.01). Large scale genotyping and complementary genomic and proteomic assays would be required to assess any potential effect of the remaining genetic variants in PTPN22.

Availability of supporting data
The data set supporting the results of this article is available in the National Center for Biotechnology Information Reference Assembly dbSNP repository, http:// www.ncbi.nlm.nih.gov/SNP/snp_viewTable.cgi? handle=O_LEARY_ATOPY.