The skin microbiome in atopic dermatitis

The human skin is colonized by a variety of microorganisms, interacting with the host andmodulating immunity. In atopic dermatitis (AD), the pathogenic bacterium Staphylococcusaureus expand, worsening disease. Technological advances now allow characterizingwhole communities of microbiota and enhance our knowledge on imbalanced microbialcompositions, dysbiosis. The overall aim was to gather knowledge on the skin microbiomeand its role in AD, establish methodology and investigate the skin microbiome, includingbacteria, fungi and virus, across multiple skin sites in patients with AD.In an initial systematic review, we identified AD skin to have low bacterial diversity, mostpronounced at lesional sites, with higher relative abundance of S. aureus and S. epidermidisand lower abundances of Cutibacterium and Malassezia. A causal role of dysbiosis indermatitis was indicated. We furthermore identified great variability in methodologicalapproaches, likely affecting the outcomes. In comparing sampling strategies in ourmethodological study, eSwabs were preferable. Reducing human DNA in vitro beforeshotgun metagenomic sequencing did not skew microbial populations, however, may havecaused low success of library preparation in the case-control study. The case-control studycomprised 10 adult AD patients and 5 skin-healthy age- and sex-matched controls sampledon 14 non-overlapping skin areas. The microbial compositions differed clearly between ADand control, most pronounced at the flexures and neck. Besides from the phenotype foundin our systematic review, AD skin was characterized by higher relative abundances ofMoraxella osloensis and Micrococcus luteus and lower relative abundances of S. hominisand C. acnes. The AD-virome had increased abundances of Propionibacterium phages,PHL041 and PHL092, and Staphylococcus epidermidis phages, CNPH82 and PH15. Higherabsolute abundances of Staphylococcus phages, Ipla5 and Ipla7, in lesional AD skin. Wefound great subject specificity in skin microbiomes, including strains of S. aureus.In conclusion, robust and standardized methodologies for investigating the skin microbiomeis still warranted and we would recommend dealing with human DNA computationally, postmetagenomic sequencing. Dysbiosis in AD involve both the bacteriome, mycobiome andvirome. The results imply a key role of the skin microbiome in AD where phages support thebacterial dysbiosis, potentially by lysing commensals and providing S. aureus and S.epidermidis with virulence genes.