Sharing is caring? Skin microbiome insights into staphylococci in atopic dermatitis patients and caregivers

Atopic dermatitis (AD) is a common inflammatory skin disease characterized by pruritus and a chronic relapsing course with disease flares. AD patients commonly suffer sleep disruption related to pruritus and may also experience depression and anxiety. In general, the burden of disease in AD is high and corresponds to disease severity. With AD’s prevalence and disease burden, extensive research has focused on understanding etiology and improving therapeutic management.

The pathophysiology of AD is multifactorial and includes genetic susceptibility, skin barrier dysfunction, immunologic dysregulation, and environmental exposures. For example, mutations in the filaggrin gene (FLG), which encodes a protein produced by keratinocytes and contributes to the skin barrier, is associated with skin barrier disturbance and severe atopic dermatitis. In addition, deeper understanding of the immunological abnormalities in AD have led to recent advances in treatment and management. Understanding AD pathophysiology is further complicated by disease heterogeneity in which subtypes continue to be defined and by the inherent complexities of microbiome studies.^1–6

Another key characteristic of AD is microbial imbalance. For instance, AD patients are susceptible to viral infections including molluscum contagiosum and herpes simplex virus (HSV), and HSV infections in AD can lead to severe sequelae including vision-threatening ocular disease and, in 3% of patients, eczema herpeticum.¹ More commonly, S. aureus can be cultured in 80–100% of patients with AD. With the high frequency of S. aureus colonization and infections in AD, treatments often include antimicrobial approaches, e.g., topical and systemic antibiotics and dilute sodium hypochlorite baths. The link with S. aureus and clinical improvement with antimicrobial approaches in some patients has led to interest in studying the potential contributions of S. aureus to disease.

Culturing methods can be challenging for some microbes and potentially bias against slow-growing species. Microbiome amplicon sequencing enables identification of the microbial composition within a clinical sample, e.g., targeting bacterial 16S ribosomal RNA (rRNA) genes. Earlier amplicon sequencing studies in AD showed higher relative abundances of staphylococci during flared disease as compared to after use of topical steroids or at baseline disease state.⁵ Consistent with culture-based studies, most AD patients had higher S. aureus relative abundances during disease flares and a smaller subset of AD patients had higher S. epidermidis relative abundances.⁵

While sequencing specific regions of the 16S rRNA gene can help differentiate amongst staphylococcal species, taxonomic classification beyond the species-level is important due to potential biological differences.⁴ For example, limited S. aureus strains carry the toxin genes responsible for clinical phenotypes of toxic shock syndrome or staphylococcal scalded skin syndrome. Shotgun metagenomics sequencing can permit identification of bacteria, fungi, and DNA viruses within a clinical sample as well as determination of strain-level differences and the microbiome functional potential. Prior shotgun metagenomics studies investigating staphylococcal strains in AD patients showed that the predominant S. aureus strains were often clonal whereas S. epidermidis strains were much more diverse.^{2, 3}

In this issue, Chia et al.⁷ built on their prior AD microbiome work to investigate skin microbiomes in children with AD and their healthy primary caregivers. The authors used shotgun metagenomics and whole genome sequencing of cultivated S. aureus isolates from Singaporean AD patient-caregiver pairs (n=30) and control-caregiver pairs (n=30) to demonstrate that individuals from AD households had different skin microbiomes as compared to control households. They examined a ratio of S. aureus to S. hominis (A/H) which distinguished AD versus control households (AUC >0.80, fair to good). Interestingly, the stronger correlation between SCORAD severity and the A/H ratio in contrast to normalized S. aureus relative abundances raises the concept that looking beyond S. aureus may be important in AD.

Consistent with other human microbiome studies, members of the same household had similar skin microbiota. In comparison to other studies on S. aureus culture-positive rates in AD patients and in general population cohorts, S. aureus culture positivity was relatively low in this cohort of AD patients [55% (n=10/18) lesional and 57% (n=17/30) non-lesional skin] and relatively high for control households (40%). The low S. aureus positivity in AD may be due to a focus on non-lesional skin; higher positivity in controls may suggest that the study’s findings are population-specific. A total of 55 high-quality genomes were obtained after sequencing and paired isolates were limited to 6 AD patient-caregiver pairs and 5 control child-caregiver pairs. S. aureus strains matched in 5 of 6 AD patient-caregiver pairs and 1 of 5 control child-caregiver pairs. In contrast to control household strains, S. aureus isolates from AD households were enriched in genes encoding virulence factors, including enterotoxins Q (seq), K2, and K (sek). This study demonstrated AD children and their primary healthy caregivers shared more skin species and strains in comparison to controls.

The case-control design with child-caregiver pairs was suited to address shared microbiota, but the low AD culture positivity rate limited the numbers of matched pairs for strain analyses and combining swabs of different skin sites limited the ability to assess S. aureus distribution. The skin sampling was well-controlled, requiring ≥12 hours after bathing and additionally restricting use of topical steroids and medications which would otherwise have resulted in lower S. aureus culture rates. A limitation addressed by the authors is the cross-sectional study design. Longitudinal studies following the natural history of AD may have potentially increased the yield of cultured S. aureus isolates as well as provided potential insights into the directionality of shared strains amongst household members.

To address whether members of the same household have shared strains, shotgun metagenomics alone would have been insufficient due to the combined challenges of small amounts of microbial DNA on skin and sequencing resolution limits. The integration of shotgun metagenomics and whole genome sequencing of bacterial isolates cultures from the same individuals was critical for deeper strain-level analyses. Since S. aureus is most frequently associated with AD, this study focused on this particular species. Given the prior culture-dependent and culture-independent studies that demonstrated a relative increase in S. epidermidis in some AD disease exacerbations^{2, 5}, the role of S. epidermidis in AD may also be of interest.⁸

An unresolved debate is whether S. aureus contributes to the pathophysiology of AD or is a bystander that adheres particularly well to a disrupted skin barrier. Chia and colleagues noted that S. aureus on non-lesional skin suggested that microbial communities can precede clinical disease, yet healthy AD caregivers had similar microbiota highlighting that non-microbial factors contribute to disease. Analyses of skin biopsies from non-lesional skin in AD patients have described immunological activation more similar to lesional skin as compared to controls⁹, indicating that normal-appearing skin may have subclinical abnormalities which in turn may result in S. aureus colonization. This query is clinically relevant in that understanding the role of S. aureus in AD directly affects therapeutic approaches and the use of antimicrobial treatments and decolonization strategies. While systematic reviews have not found strong evidence for anti-staphylococcal methods in managing AD¹⁰, the research into subtypes of AD may indicate that, in certain patients, antimicrobial approaches may be effective. The current study highlights that primary caregivers can share S. aureus strains with AD patients⁷, raising the possibility that caregiver sharing could confound effective antimicrobial therapies in AD. Additional investigations are needed to assess more deeply clinical outcomes of carriage and decolonization. As researchers continue to disentangle the role of staphylococci in AD and potential disease subtypes and determine how findings can inform clinical medicine, it is likely that clinicians will continue to incorporate common antimicrobial and/or decolonization approaches when faced with the driving need to alleviate the burden of disease in AD patients.