Expression of Th1-cytokine mRNA in canine atopic dermatitis correlates with severity of clinical lesions

Abstract

<jats:p>Previous studies indicate that canine atopic dermatitis (AD) is associated with expression of the Th2 cytokine IL‐4, but that mRNA for the Th1 cytokines IFNγ, TNFα and IL‐2 is expressed in lesional atopic skin. This suggests that Th1 cytokines are involved in the pathogenesis of chronic lesional AD. The aim of this study was to determine whether the expression of Th1 cytokine mRNA correlates with the clinical severity of canine AD. Samples were taken from 23 atopic and 12 healthy dogs. Atopic dermatitis was diagnosed on accepted clinical criteria and exclusion of other pruritic dermatoses. The healthy dogs had no history or clinical signs of any skin disease or condition likely to affect immune function. The clinical severity was scored using a modified canine atopic dermatitis extent and severity index (CADESI). Six‐millimetre punch biopsies of nonlesional skin were collected from both healthy and atopic dogs. Lesional skin was taken from areas of erythematous macular papular dermatitis avoiding sites of secondary infection. Gene transcripts for the housekeeping gene gluteraldehyde‐3‐phosphate dehydrogenase (GAPDH), IFNγ, TNFα, IL‐2 and IL‐4 were amplified by RT‐PCR. Forward and reverse primers were designed from published canine sequences. Specificity was confirmed by southern blotting and hybridization to internal sequence probes. The PCR products were run on 1.2% polyacrylamide‐ethidium bromide gels and imaged under 590 nm ultraviolet light. Levels of mRNA in each sample were expressed as (cytokine net band intensity)/(GAPDH net band intensity) using Kodak 1D software. Duplicate biopsies for histopathology were fixed in 10% neutral buffered formalin, paraffin embedded, sectioned and stained with H&amp;E. Five ×400 images of the superficial interfollicular dermis were captured for quantitative analysis. The degree of cell infiltration was assessed by calculating the proportion of the image area that consisted of cell nuclei using Object‐Image software. Spearman's nonparametric correlation was used to test the data. There were significant correlations between cellular infiltration and mRNA levels for IFNγ (<jats:italic>P</jats:italic> &lt; 0.0001, <jats:italic>n</jats:italic> = 13), TNFα (<jats:italic>P</jats:italic> = 0.046, <jats:italic>n</jats:italic> = 8) and IL‐2 (<jats:italic>P</jats:italic> = 0.023, <jats:italic>n</jats:italic> = 12) in lesional skin. No such correlation was seen in either nonlesional (<jats:italic>n</jats:italic> = 23) or healthy skin (<jats:italic>n</jats:italic> = 12). There were no significant correlations between cellular infiltration and IL‐4 mRNA levels in lesional (<jats:italic>n</jats:italic> = 8), nonlesional (<jats:italic>n</jats:italic> = 23) and healthy skin (<jats:italic>n</jats:italic> = 10). There were significant correlations between the CADESI scores and mRNA levels for IFNγ (<jats:italic>P</jats:italic> = 0.02, <jats:italic>n</jats:italic> = 13), TNFα (<jats:italic>P</jats:italic> = 0.04, <jats:italic>n </jats:italic>= 8) and IL‐2 (<jats:italic>P</jats:italic> = 0.008, <jats:italic>n</jats:italic> = 12) in lesional skin, but not in nonlesional skin (n = 23). There was no correlation between CADESI scores and IL‐4 mRNA levels in either lesional (<jats:italic>n</jats:italic> = 8) or nonlesional skin (<jats:italic>n</jats:italic> = 23). This suggests that Th1 cytokines are pivotal in the pathogenesis of chronic lesional AD. In contrast, IL‐4 does not appear to be linked to the development of chronic lesions. These findings support the hypothesis of a switch from Th2 to Th1 polarisation in the pathogenesis of canine AD.</jats:p><jats:p> <jats:italic>Funding: The Wellcome Trust.</jats:italic> </jats:p>