microorganisms Article Changes in Skin and Nasal Microbiome and Staphylococcal Species Following Treatment of Atopic Dermatitis with Dupilumab Caroline Meyer Olesen 1,*, Anna Cäcilia Ingham 2 , Simon Francis Thomsen 1, Maja-Lisa Clausen 1 , Paal Skytt Andersen 2 , Sofie Marie Edslev 2, Yasemin Topal Yüksel 1, Emma Guttman-Yassky 3,4,† and Tove Agner 1,† 1 Department of Dermatology, Bispebjerg Hospital, 2400 Copenhagen, Denmark; [email protected] (S.F.T.); [email protected] (M.-L.C.); [email protected] (Y.T.Y.); [email protected] (T.A.) 2 Department of Bacteria, Parasites and Fungi, Statens Serum Institute, 2400 Copenhagen, Denmark; [email protected] (A.C.I.); [email protected] (P.S.A.); [email protected] (S.M.E.) 3 Laboratory of Inflammatory Skin Diseases, Department of Dermatology, Icahn School of Medicine at Mount Sinai, New York, NY 10029, USA; [email protected] 4 Laboratory for Investigative Dermatology, The Rockefeller University, New York, NY 10065, USA * Correspondence: [email protected]; Tel.: +45-24-49-01-33 † These authors contributed equally to this work. Abstract: Investigation of changes in the skin microbiome following treatment of atopic dermatitis (AD) with dupilumab may provide valuable insights into the skin microbiome as a therapeutic Citation: Olesen, C.M.; Ingham, target. The aim of this study is to assess changes in the AD skin microbiome following treatment A.C.; Thomsen, S.F.; Clausen, M.-L.; of AD with dupilumab (n = 27). E-swabs were collected from nose, lesional, and nonlesional Andersen, P.S.; Edslev, S.M.; Yüksel, skin before and after 16 weeks of dupilumab therapy, and the microbiome was analyzed by 16S Y.T.; Guttman-Yassky, E.; Agner, T. rRNA and tuf gene sequencing. Data for 17 patients with milder disease receiving treatment with Changes in Skin and Nasal non-targeted therapies are also presented. The results show that both groups experienced clinical Microbiome and Staphylococcal improvement (p < 0.001) following dupilumab therapy and that Shannon diversity increased and Species Following Treatment of bacterial community structure changed. The relative abundance of the genus Staphylococcus (S.) and Atopic Dermatitis with Dupilumab. S. aureus decreased, while that of S. epidermidis and S. hominis increased. No significant changes Microorganisms 2021, 9, 1487. were observed for patients receiving non-targeted treatments. The increases in S. epidermidis and https://doi.org/10.3390/ S. hominis and the decrease in S. aureus correlated with clinical improvement. Furthermore, changes microorganisms9071487 in S. hominis and S. epidermidis correlated inversely with S. aureus. In conclusion, treatment with Academic Editor: John E. Gustafson dupilumab significantly changed the skin microbiome and decreased S. aureus. Our results suggest a favorable role of commensal staphylococci in AD. Received: 28 May 2021 Accepted: 7 July 2021 Keywords: atopic dermatitis; skin microbiome; staphylococcus; dupilumab Published: 13 July 2021 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in 1. Introduction published maps and institutional affil- The pathogenesis of atopic dermatitis (AD) is characterized by the interplay between iations. activation of the Th2/Th22-skewed immune response and epidermal barrier impairment [1,2]. This is paralleled by changes in the skin microbiome, with decreased bacterial diversity and increased colonization with Staphylococcus aureus (S. aureus), which correlates with disease severity [3–6]. Conditions in AD skin promote S. aureus colonization [7,8], and Copyright: © 2021 by the authors. S. aureus expresses a broad spectrum of virulence factors that aggravate the disease [9–11]. Licensee MDPI, Basel, Switzerland. Commensal staphylococcal species, such as S. epidermidis and S. hominis, work in concert This article is an open access article with host immunity to limit pathogen colonization [12–15], and were recently shown to be distributed under the terms and reduced in AD [16]. Accordingly, deficiency of commensal staphylococci may be related to conditions of the Creative Commons the increased S. aureus colonization and disease severity in AD. Attribution (CC BY) license (https:// Dupilumab, a fully human monoclonal antibody targeting IL-4 and IL-13 by blocking creativecommons.org/licenses/by/ IL-4Rα signaling [17], has been shown to improve the clinical disease activity, immune 4.0/). Microorganisms 2021, 9, 1487. https://doi.org/10.3390/microorganisms9071487 https://www.mdpi.com/journal/microorganisms Microorganisms 2021, 9, 1487 2 of 11 abnormalities, and epidermal barrier function, thereby highlighting IL-4Rα signaling as a key driver of the AD phenotype [18–20]. One previous study of changes in the skin microbiome and S. aureus colonization following dupilumab therapy showed that bacterial diversity increased and the absolute S. aureus abundance decreased [21]. However, the study did not assess treatment-related changes in other staphylococcal species and nasal microbiome. Further understanding of the complex interactions within the bacterial communities and their cross-talk with host immunity is necessary to explore the potential of manipulat- ing the skin microbiome in the management of AD. Accordingly, investigation of changes in the skin microbiome and staphylococcal species that are associated with successful responses to targeted and non-targeted treatments may lead to identification of genera and species that need to be further evaluated with regard to their link to AD. Therefore, the aim of the study was to investigate changes in the skin and nasal microbiome associated with successful response to dupilumab. By applying tuf gene sequencing, changes in the relative abundance of all staphylococcal species were assessed, revealing interesting new data on the potential interplay between different staphylococcal species following treatment. 2. Materials and Methods 2.1. Study Design Patients were recruited from the Department of Dermatology, Bispebjerg Hospital, Copenhagen, from March 2018 to September 2019. Inclusion criteria comprised a diagnosis of AD according to UK criteria, age ≥ 18 years, and no pregnant or lactating women. Patients were scheduled to be evaluated at treatment initiation, after 16 weeks of sys- temic treatments (including dupilumab), and after four weeks of topical treatment. Data on the three most common loss-of-function mutations in the filaggrin gene (FLG) were collected [22]. Patients were instructed not to apply emollients within 24 h of each visit. 2.2. Sample Collection and Disease Severity Assessment Swabs (e-Swab, Copan, Brescia, Italy) were collected from anterior nares and lesional and nonlesional skin from representative areas on the upper or lower extremities. Disease severity was assessed with the Eczema Area and Severity Index (EASI). Blood was collected for total serum IgE quantification. 2.3. DNA Extraction and Microbiome Characterization DNA was extracted from swabs using an enzymatic prelysis step (30 min incubation at 37 ◦C with an enzyme solution containing 4 U lysostaphine (SAE0091), 25 U mutanolysin (sae0092), and 3 mg lysozyme (L4919) (Sigma-Aldrich, St. Louis, MO, USA); then 30 min incubation at 56 ◦C with 20 µL protein kinase K (RPROTKSOL-RO, Sigma-Aldrich, St. Louis, MO, USA), followed by DNA extraction on a MagNa-Pure 96 robot using a DNA and Viral NA Small Volume Kit (Roche, Mannheim, Germany). The V3-V4 region of the 16S rRNA gene and that of the tuf gene were amplified in two separate PCRs (95 ◦C for 3 min; 25 cycles of 98 ◦C for 20 s, 60 ◦C for 15 s, 72 ◦C for 45 s; 72 ◦C for 5 min), using primers (16SrRNA: 341F: 50- CCTACGGGNGGCWGCAG -30; 805R: 50- GACTACHVGGGTATCTAATC-30; tuf : F: 50- CAGAAGAAAAAGAACGTGG-30; R: 50- GTCCTCAACWGGCATCA-30) with preceding heterogeneity spacers [23,24]. Amplicon libraries were constructed using nextera indexing primers (Illumina Inc., San Diego, CA, USA) (PCR program used: 95 ◦C for 3 min; 20 cycles of 98 ◦C for 20 s, 55 ◦C for 15 s, and 72 ◦C for 45 s; 72 ◦C for 5 min) and sequenced on a MiSeq instrument using a 600 cycle V3 kit (Illumina Inc., San Diego, CA, USA). 2.4. Sequence Pre-Processing Demultiplexing of raw reads was performed by using the bcl2fastq Conversion Soft- ware (Illumina Inc., San Diego, CA, USA). Subsequently, cutadapt (version 2.3) [25] was used on 16S rRNA and tuf gene reads for heterogeneity spacer and primer trimming at an Microorganisms 2021, 9, 1487 3 of 11 8% error rate (corresponding to one mismatch per primer) in paired-end mode. Trimmed reads were subjected to quality filtering and amplicon sequence variant (ASV) inference with dada2 (version 1.12.1) [26]. The dada2 pipeline was utilized run-wise with default settings, except for truncation length. 16S rRNA gene reads were truncated at 270 (forward reads) and 210 bp (reverse reads), while tuf gene reads were truncated at 270 (forward reads) and 241 bp (reverse reads). We performed consensus chimera removal. In cases when a sample had a read count <5000 after quality filtering, it was re-sequenced. Taxonomic assignment of 16S rRNA gene sequence-derived ASVs was performed with dada20s ‘assign- Taxonomy’ and ‘addSpecies’ functions, using the Silva reference database and species-level training set (version 132) formatted for dada2, respectively. A staphylococcal-specific taxonomic database was used to classify tuf gene sequence-derived ASVs with the ‘assign- Taxonomy’ function [24]. Using the R package phyloseq v, ASV count tables, taxonomic tables, and