From top to bottom: Staphylococci in atopic dermatitis

Abstract

In this issue of Cell Host & Microbe, Kashaf et al. and Key et al. examine isolates of Staphylococcus aureus among individuals with atopic dermatitis, revealing insights into evolution, antibiotic resistance, transmission mechanisms, skin colonization, and virulence factors. These findings further our understanding of disease pathogenesis and potential treatments.

Atopic dermatitis (AD) is a multifactorial disease associated with various host and environmental components.¹ Host genetic factors include dysfunction in skin barrier integrity (e.g., filaggrin) or immune responses (e.g., Th2-cell differentiation), which account for only 20% of disease heritability and cannot explain the increased prevalence of AD in industrialized countries.^2-4 However, a role for the skin microbiome as emerged as a key culprit. Previous shotgun metagenomic studies have shown that staphylococcal abundance on AD-lesional skin correlates with worsened disease, caused in part by bacterial virulence factors and immune system modulation.^4,5 Finally, treatments for AD include antimicrobials and immunomodulatory therapies that alleviate staphylococcal abundance and disease severity, further highlighting the microbiome as an active player.^6,7

Considering the impact staphylococci have on AD disease severity, information on the genetic basis of staphylococcal genotypes, phenotypes, and global differences among AD patients is sorely lacking. To address this gap in knowledge, Kashaf et al. and Key et al. went beyond our current understanding in extensive and data-rich studies by pairing whole genome sequencing of staphylococcal species from AD patients with metagenomic (population level) data.^8,9 This uncovered new insights into the role of staphylococci in AD pathogenesis, including global antibiotic resistant patterns, bacterial transmission among household members, sharing of genetic elements between bacteria on an individual, and genetic loci that correlated with AD severity (Figure 1).

Figure 1. Staphylococcal phenotypes in atopic dermatitis uncovered using metagenomics with whole genome sequencing of skin isolates: from a global down to a genetic loci perspective.

The study by Kashaf et al. revealed geographical patterns in staphylococci (predominantly S. aureus and S. epidermidis) from AD patients, ST and antibiotic resistance (e.g., Fusidic acid). Staphylococcal strain diversity on AD patients was also influenced by within-household transmission (e.g., siblings). On an individual level, metagenomics revealed that S. aureus colonization was associated with greater AD severity than S. epidermidis. Moreover, examination of the “mobilome,” which consists of phages, plasmids, and pathogenicity islands, identified transfer and stable acquisition of these mobile genetic elements between S. aureus and S. epidermidis. A mGWAS revealed that S. aureus virulence factor genetic loci correlated with AD, whereas genetic loci for S. epidermidis metabolic and inter-species interaction factors were associated with AD. Lastly, Key et al. identified that loss-of-function mutations in the S. aureus capD gene are associated with strains isolated from AD patients. Abbreviations: AMR, antimicrobial resistance; ST, sequence types; AD, atopic dermatitis. Created with BioRender.com.

Using publicly available metagenomic data from AD patients, Kashaf et al. examined global patterns in staphylococci sequence types (ST). Interestingly, overall disease-associated Staphylococcus aureus strains were not associated with any single ST on a global scale. In the USA, ST5 and ST8 predominated, whereas ST1 was notably absent despite its predominance in Europe and Singapore. Upon further analysis, global patterns emerged regarding antibiotic resistance among AD patients’ staphylococcal isolates. For instance, patients from the U.K., Singapore, and Denmark carried fusidic acid resistance determinants, whereas these were absent in USA isolates. The authors postulated that this is likely due to the commonplace use of fusidic acid for the treatment of skin infections in Europe and East Asia where ST1 predominates, but not in the USA where ST5 and ST8 prevail.¹⁰ Surprisingly, the authors found that S. epidermidis but not S. aureus were largely methicillin-resistant, potentially a result of overuse of treatment agents combating methicillin-resistance in the USA.

On a household level, Kashaf and colleagues studied the intra-host SNP-level variation of S. aureus and S. epidermidis isolates from members of households with AD patients and observed staphylococcal transmissions between siblings that were stably retained.⁸ On an individual level, the authors investigated the potential for inter-staphylococcal transmission of mobile genetic elements from the same skin lesion. Remarkably, on one patient, an identical mecA gene, which encodes for resistance to methicillin and other beta-lactams, was found in both S. aureus and S. epidermidis. Given these findings, Kashaf and colleagues next investigated the potential for inter-staphylococcal transmission of virulence factors or plasmids via phages.⁸ Phages are viruses that infect bacteria with specificity, traditionally grouped phylogenetically based on their host bacterial species. This study challenges the specificity notion as the same phage was identified in multiple staphylococcal species (e.g., S. aureus, S. epidermidis, S. hominis, and S. argenteus). Phage analysis also uncovered the presence of the Sa3int phage, which harbors virulence genes (e.g., Staphylokinase [sak] and Staphylococcal complement inhibitor [scn]), suggesting a role for the microbiome as a melting pot of gene transfer of elements important to AD pathogenesis.

Using a microbial genome-wide association study (mGWAS) of isolates cultured during AD flares of children from a global collection, Kashaf et al. identified genomic loci in both S. aureus and S. epidermidis associated with higher AD disease severity.⁸ In S. aureus, these loci included genes associated with adherence, virulence, growth, and adaptation (e.g., entA, speA, sarU, clfA, and copA), whereas S. epidermidis had enrichment in loci for inter-species and host-microbial interactions to metabolism (e.g., lysP, gseA, galK, and perR), which the authors suggest may enhance resilience or inter-species competition in AD skin. Clues as to how S. aureus and S. epidermidis have adapted to AD skin may be found in the mGWAS hits common between both (e.g., lysS, gabR, perR, and opuD).

In a related longitudinal study in this issue of Cell Host Microbe, Key et al. found that S. aureus colonizes and adapts on AD skin by accumulating de novo mutations.⁹ Similar to the Kashaf et al. study, Key et al. detected clonal expansions, convergent mutations, and bacterial dynamics on the same host, which they did by performing repeated isolations from 23 pediatric patients at 7 various skin sites.^8,9 Interestingly, a single S. aureus lineage dominated each patient’s microbiome in a dramatic sweep over time, and with signs of adaptive evolution, some variants spread across the body within months. Specifically, Key et al. found that adaptive mutations in the S. aureus polysaccharide capsule synthesis gene, capD, which were associated with loss of S. aureus capsule production, were more common in AD globally than other contexts and even occurred in parallel in S. aureus isolated from multiple sites on the same patient.⁹ These remarkable findings suggested that loss-of-function mutations in capD or in genes relating to capsule production have a selective advantage on AD skin.⁹ A potential explanation for this genetic adaptation includes that acapsular S. aureus strains have improved adherence to the AD patient skin.

These incredibly in-depth studies led by the Kong and Lieberman laboratories used extensive and well-designed analyses with their own and publicly available genomic datasets to investigate strain-level differences of staphylococcal species in AD.^8,9 They found that the skin microbiome and staphylococcal strain specificity differed based on geography, disease status, disease subtype, and household composition (e.g., sibling transmission). The precise mechanisms by which AD is initiated is unknown, but is believed to possess a genetic component and, therefore, heritability.^3,4 These studies entertain this notion with a new perspective that heritability stems not only from the patient’s DNA, but also that of the skin microbiome being passed within households, between the staphylococci on the skin, or within S. aureus clones acquiring adaptive genetic mutations and monopolizing across the body.

These insights raise new questions to be answered in future studies. For instance, since the whole genome sequencing data are centered around the U.S. pediatric population, do the findings translate to the broader patient populations and geographic regions? How do these discoveries apply to other diseases with staphylococcal involvement? Moreover, nearly half of all patients did not have staphylococci as the predominant species, which begs the question—what other species affect AD development and severity? Notably, the authors determined that mobile genetic elements were shared among staphylococci species in AD patients, including those harboring antibiotic resistance and virulence genes. Thus, how do these genetic elements influence AD pathogenesis and how do we prevent their transmission?

These robust and exciting studies by Kashaf et al. and Key et al. provide new insights into the role of the microbiome, specifically staphylococcal species, in AD pathogenesis and have the potential to improve the translation of bacterio-therapies (e.g., prebiotics, probiotics, and microbiome transplantation) and other aspects of patient care (e.g., antibiotic use) for AD patients.⁸ The authors demonstrate the importance of studying a disease at various patient population levels (e.g., global, household, and individual) to extend our understanding of cross-species interactions in the microbiome and the relevance to human disease and health.