Gene expression in canine atopic dermatitis and correlation with clinical severity scores

Abstract

Background

Canine atopic dermatitis (cAD) is a common condition in dogs that may be a naturally occurring model for human atopic dermatitis (hAD). Despite this, comparative research is limited, particularly into the genetic background of cAD.

Objectives

• Measure candidate gene expression in cAD skin using quantitative real time PCR (qPCR)

• Correlate gene expression to clinical cAD scores (Canine Atopic Dermatitis Extent and Severity Index [CADESI]-03 and intradermal allergen test [IDT]).

Methods

mRNA was extracted from biopsies of non-lesional and lesional skin from atopic dogs, and healthy skin from non-atopic dogs. Gene expression was quantified using qPCR, and compared between non-lesional atopic, lesional atopic and healthy skin. Gene expression in atopic skin was correlated with clinical severity (CADESI-03) and the number of positive reactions on an IDT.

Results

Of the 20 quantified genes, 11 demonstrated statistically significant altered mRNA expression between atopic and healthy skin; dipeptidyl-peptidase-4 (DPP4), phosphatidylinositol-3,4,5-trisphosphate-5-phosphatase-2 (INPPL1), serine protease inhibitor kazal type-5 (SPINK5), sphingosine-1-phosphate lyase-1 (SGPL1), peroxisome proliferator-activated receptor gamma (PPARγ), S100 calcium-binding protein A8 (S100A8), Plakophilin-2 (PKP2), Periostin (POSTN), Cullin4A, TNF-α and metalloproteinase inhibitor-1 (TIMP-1). Three genes correlated with CADESI-03: serum amyloid A 1 (SAA-1), S100A8, and PKP2; and four with IDT results: mast cell protease I (CMA1), SAA-1, S100A8 and SPINK5.

Conclusion

Genes with altered expression included those relevant to skin barrier formation and immune function, suggesting both are relevant in the pathogenesis of AD. Many of these genes reflect the proposed pathogenesis in hAD, supporting the use of dogs as a model for hAD. Furthermore, these genes may be considered suitable targets for future genetic and protein function studies in human and canine AD.

Introduction

Canine atopic dermatitis (cAD) is a common condition that can be considered a naturally occurring, spontaneous model of human atopic dermatitis (hAD). There are many animal models which display clinical signs consistent with hAD but dogs are of particular interest as they share the same environment as humans, unlike the commonly used rodent models. The pathogenesis of the disease in both humans and dogs is strongly associated with immunological hyper-reactivity, although skin barrier function, microbial colonisation and infection may also have a role [1], [2]. The prevalence of AD over the last three decades has increased in both dogs [3] and humans [4], suggesting common environmental and/or genetic components.

Unlike most murine models cAD is not caused by a single gene defect, and, like hAD, is a polygenic disorder with complex inheritance and interactions with environmental influences. It is therefore a potential complex model for the human condition. Using dogs as a model to study of the genetic basis of AD is advantageous because dog breeds form genetically isolated populations in strong linkage disequilibrium [5]. Fewer genetic markers are required to identify an association with a given phenotype and smaller sample sizes can be used to find these associations compared to human genetic studies [6]. However, research into cAD has been limited to reporting cytokine and chemokine profile changes [7] until recently, when a message RNA (mRNA) expression microarray study was performed [8]. Similarities between dysregulated genes in hAD and cAD were identified, although the number and relevance of the genes tested were limited. In hAD, microarray and qPCR analysis of gene expression, linkage studies and candidate gene analyses have been used to identify causative or susceptibility alleles. These studies have implicated a large number of genes and regions [9]. Twelve genes with relevant epidermal or immune functions (Cystatin A [10], [11], CARD4 [12], P-selectin [13], [14], PKP2 [15], PPARγ [16], [17], SGPL1 [18], TNF-α [19], [9], Cadherin-13 [20], CMA1 [21], [22], DPP4 [23], SPINK5 [24], [25], SAA-1 [26]), were selected as potential candidate genes for AD. Eight further genes (ARTS-1, Cullin-4A, INPPL1, S100A8, POSTN, SCCA2, STAT2, TIMP-1) were selected using data from a canine mRNA expression microarray [8], showing that they were dysregulated in cAD.

The aim of this study was to quantify the expression of these candidate genes in lesional atopic, non-lesional atopic and healthy canine skin. Gene expression in atopic skin was also correlated with two clinical measures: the Canine Atopic Dermatitis and Severity Index (CADESI-03), which is similar to the human SCORAD index, and the number of positive reactions on intradermal testing (similar to prick tests in human patients) with individual allergens (IDT). Determining whether cAD and hAD share a similar genetic background will provide evidence that cAD is a suitable complex model for the human condition, and help identify novel target genes for further study of pathogenesis and intervention.