Adding Fuel to the Fire? The Skin Microbiome in Atopic Dermatitis

Absolute quantification of the living skin microbiome overcomes relic-DNA bias and reveals specific patterns across volunteers

BACKGROUND

As the first line of defense against external pathogens, the skin and its resident microbiota are responsible for protection and eubiosis. Innovations in DNA sequencing have significantly increased our knowledge of the skin microbiome. However, current characterizations do not discriminate between DNA from live cells and remnant DNA from dead organisms (relic DNA), resulting in a combined readout of all microorganisms that were and are currently present on the skin rather than the actual living population of the microbiome. Additionally, most methods lack the capability for absolute quantification of the microbial load on the skin, complicating the extrapolation of clinically relevant information.

RESULTS

Here, we integrated relic-DNA depletion with shotgun metagenomics and bacterial load determination to quantify live bacterial cell abundances across different skin sites. Though we discovered up to 90% of microbial DNA from the skin to be relic DNA, we saw no significant effect of this on the relative abundances of taxa determined by shotgun sequencing. Relic-DNA depletion prior to sequencing strengthened underlying patterns between microbiomes across volunteers and reduced intraindividual similarity. We determined the absolute abundance and the fraction of population alive for several common skin taxa across body sites and found taxa-specific differential abundance of live bacteria across regions to be different from estimates generated by total DNA (live + dead) sequencing.

CONCLUSIONS

Our results reveal the significant bias relic DNA has on the quantification of low biomass samples like the skin. The reduced intraindividual similarity across samples following relic-DNA depletion highlights the bias introduced by traditional (total DNA) sequencing in diversity comparisons across samples. The divergent levels of cell viability measured across different skin sites, along with the inconsistencies in taxa differential abundance determined by total vs live cell DNA sequencing, suggest an important hypothesis for certain sites being susceptible to pathogen infection. Overall, our study demonstrates a characterization of the skin microbiome that overcomes relic-DNA bias to provide a baseline for live microbiota that will further improve mechanistic studies of infection, disease progression, and the design of therapies for the skin. Video Abstract.

Skin microbiome and dermatologic disorders

Human skin acts as a physical barrier to prevent the entry of pathogenic microbes while simultaneously providing a home for commensal bacteria and fungi. Microbiome sequencing studies have demonstrated the unappreciated diversity and selectivity of these microbes. Functional studies have demonstrated the impact of specific strains to tune the immune system, sculpt the microbial community, provide colonization resistance, and promote epidermal barrier integrity. Recent studies have integrated the microbiome, immunity, and tissue integrity to understand their interplay in common disorders such as atopic dermatitis. In this Review, we explore microbiome shifts associated with cutaneous disorders with an eye toward how the microbiome can be mined to identify new therapeutic opportunities.

Single-cell sequencing of human Langerhans cells identifies altered gene expression profiles in patients with atopic dermatitis

Atopic dermatitis (AD) is characterized by dysregulated T cell immunity and skin microbiome dysbiosis with predominance of Staphylococcus aureus, which is associated with exacerbating AD skin inflammation. Specific glycosylation patterns of S. aureus cell wall structures amplify skin inflammation through interaction with Langerhans cells (LCs). Nevertheless, the role of LCs in AD remains poorly characterized. Here, we performed single cell RNA sequencing of primary epidermal LCs and dermal T cells, isolated from skin biopsies of AD patients and healthy control subjects, alongside specific glycoanalysis of S. aureus strains isolated from the AD lesions. Our findings revealed 4 LC subpopulations ie, 2 steady-state clusters [LC1 and LC1H] and 2 proinflammatory/matured subsets [LC2 and migratory LCs]. The latter 2 subsets were enriched in AD skin. AD LCs showed enhanced expression of C-type lectin receptors, the high-affinity IgE receptor, and activation of prostaglandin and leukotriene biosynthesis pathways, upregulated transcriptional signatures related to T cell activation pathways, and increased expression of CCL17 compared with healthy LCs. Correspondingly, T helper 2 and T regulatory cell populations were increased in AD lesions. Complementary, we performed bulk RNA sequencing of primary LCs stimulated with the S. aureus strains isolated from the AD lesions, which showed upregulation of T helper 2-related pathways. Our study provides proof-of-concept for a role of LCs in connecting the S. aureus-T cell axis in the AD inflammatory cycle.