Journal of Allergy and Clinical Immunology: Global (Aug 2024)
Combined metagenomic- and culture-based approaches to investigate bacterial strain–level associations with medication-controlled mild-moderate atopic dermatitis
Abstract
Background: The skin microbiome is disrupted in atopic dermatitis (AD). Existing research focuses on moderate to severe, unmedicated disease. Objective: We sought to investigate metagenomic- and culture-based bacterial strain–level differences in mild, medicated AD and the effects these have on human keratinocytes (HKs). Methods: Skin swabs from anterior forearms were collected from 20 pediatric participants (11 participants with AD sampled at lesional and nonlesional sites and 9 age- and sex-matched controls). Participants had primarily mild to moderate AD and maintained medication use. Samples were processed for microbial metagenomic sequencing and bacterial isolation. Isolates identified as Staphylococcus aureus were tested for enterotoxin production. HK cultures were treated with cell-free conditioned media from representative Staphylococcus species to measure barrier effects. Results: Metagenomic sequencing identified significant differences in microbiome composition between AD and control groups. Differences were seen at the species and strain levels for Staphylococci, with S aureus found only in participants with AD and differences in Staphylococcus epidermidis strains between control and AD swabs. These strains showed differences in toxin gene presence, which was confirmed in vitro for S aureus enterotoxins. The strain from the participant with the most severe AD produced enterotoxin B levels more than 100-fold higher than the other strains (P < .001). Strains also displayed differential effects on HK metabolism and barrier function. Conclusions: Strain-level differences in toxin genes from Staphylococcus strains may explain varying effects on HK, with S aureus and non–aureus strains negatively affecting viability and barrier function. These differences are likely important in AD pathogenesis.