Skin dysbiosis in the microbiome in atopic dermatitis is site-specific and involves bacteria, fungus and virus

Abstract

Abstract Background Microbial dysbiosis with increased Staphylococcus aureus (S. aureus) colonization on the skin is a hallmark of atopic dermatitis (AD), however most microbiome studies focus on bacteria in the flexures and the microbial composition at other body sites have not been studied systematically. Objectives The aim of the study is to characterize the skin microbiome, including bacteria, fungi and virus, at different body sites in relation to AD, lesional state, and S. aureus colonization, and to test whether the nares could be a reservoir for S. aureus strain colonization. Methods Using shotgun metagenomics we characterized microbial compositions from 14 well defined skin sites from 10 patients with AD and 5 healthy controls. Results We found clear differences in microbial composition between AD and controls at multiple skin sites, most pronounced on the flexures and neck. The flexures exhibited lower alpha-diversity and were colonized by S. aureus, accompanied by S. epidermidis in lesions. Malassezia species were absent on the neck in AD. Virus mostly constituted Propionibacterium and Staphylococcusphages, with increased abundance of Propionibacterium phages PHL041 and PHL092 and Staphylococcus epidermidis phages CNPH82 and PH15 in AD. In lesional samples, both the genus Staphylococcus and Staphylococcus phages were more abundant. S. aureus abundance was higher across all skin sites except from the feet. In samples where S. aureus was highly abundant, lower abundances of S. hominis and Cutibacterium acnes were observed. M. osloensis and M. luteus were more abundant in AD. By single nucleotide variant analysis of S. aureus we found strains to be subject specific. On skin sites some S. aureus strains were similar and some dissimilar to the ones in the nares. Conclusions Our data indicate a global and site-specific dysbiosis in AD, involving both bacteria, fungus and virus. When defining targeted treatment clinicians should both consider the individual and skin site and future research into potential crosstalk between microbiota in AD yields high potential.