Temporal and Spatial Variation of the Skin-Associated Bacteria from Healthy Participants and Atopic Dermatitis Patients

Several factors have been shown to influence the composition of the bacterial communities inhabiting healthy skin, with variation between different individuals, differing skin depths, and body locations (spatial-temporal variation). Atopic dermatitis (AD) is a chronic skin disease also affecting the skin-associated bacterial communities. While the effects of AD have been studied on these processes individually, few have considered how AD disrupts the spatial-temporal variation of the skin bacteria as a whole (i.e., considered these processes simultaneously). Here, we characterized the skin-associated bacterial communities of healthy volunteers and lesional and nonlesional skin of AD patients by metabarcoding the universal V3-V4 16S rRNA region from tape strip skin samples. We quantified the spatial-temporal variation (interindividual variation, differing skin depths, multiple time points) of the skin-associated bacteria within healthy controls and AD patients, including the relative change induced by AD in each. Interindividual variation correlated with the bacterial community far more strongly than any other factors followed by skin depth and then AD status. There was no significant temporal variation found within either AD patients or healthy controls. The bacterial community was found to vary markedly according to AD severity, and between patients without and with filaggrin mutations. Therefore, future studies may benefit from sampling subsurface epidermal communities and considering AD severity and the host genome in understanding the role of the skin bacterial community within AD pathogenesis rather than considering AD as a presence-absence disorder. IMPORTANCE The bacteria associated with human skin may influence skin barrier function and the immune response. Previous studies have attempted to understand the factors that regulate the skin bacteria, characterizing the spatial-temporal variation of the skin bacteria within unaffected skin. Here, we quantified the effect of AD on the skin bacteria on multiple spatial-temporal factors simultaneously. Although significant community variation between healthy controls and AD patients was observed, the effects of AD on the overall bacterial community were relatively low compared to other measured factors. Results here suggest that changes in specific taxa rather than wholesale changes in the skin bacteria are associated with mild to moderate AD. Further studies would benefit from incorporating the complexity of AD into models to better understand the condition, including AD severity and the host genome, alongside microbial composition.