Data and biological sample collection

Children (3 months – 16 years old) and adults diagnosed with AD using Hannifin and Rajka clinical criteria⁴³ who attended dermatology, pediatric dermatology, and pediatric allergy clinics at the King Chulalongkorn Memorial Hospital were recruited in this study during the COVID-19 pandemic. The demographic data, including clinical presentation, treatments, and severity scores using EASI and SCORAD^44,45 were recorded. The exclusion criteria were patients with current bacterial skin infection, systemic infection, contact dermatitis, and recent (within the past 4 weeks) use of topical or systemic antibiotics, or systemic corticosteroids, or recent hospital admission. This study was approved by the Institutional Review Board (IRB No. 640/60), Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand. Written informed consent and/or assent forms were obtained from all participants. For human participants that are minors, the informed consent and/or assent forms were obtained from parents and/or legal guardians for both study participation and publication of identifying information in an online open-access publication. We confirm that all methods and experiments were performed in accordance with relevant guidelines and regulations. Biological samples from 3 different sites of the skin (lesion, non-lesion, and nasal mucosa) were swabbed using separate sterile cotton sticks, and samples were inoculated onto 5% sheep blood agar and mannitol salt agar (MSA; a selective media for S. aureus).

Bacterial identification and antibiotic susceptibility testing

The inoculated plates were incubated at 35 °C for 18–24 h in ambient air. The bacterial colonies from MSA and 5% sheep blood agar plates were identified by Gram-staining, catalase, coagulase, biochemical tests, API Staph kit (bioMérieux, France), and confirmed by polymerase chain reaction (PCR) using a specific gene (nuc) as shown in Supplementary Table S1-S2 online.

Antimicrobial resistance in S. aureus isolated was initially detected using the disk diffusion method (CLSI). Briefly, pure colonies of isolated S. aureus were suspended in normal saline solution and adjusted to be equivalent to 0.5 McFarland standard before spreading onto Muller-Hinton agar (BBL, USA). Disks of cefoxitin (30 µg), clindamycin (2 µg), erythromycin (15 µg), trimethoprim/sulfamethoxazole (5 µg), ciprofloxacin (5 µg), tetracycline (30 µg), fusidic acid (10 µg), and mupirocin (200 µg) were mounted on the agar surface and incubated at 35 °C for 18–24 h. The result (susceptible; S, intermediate; I, and resistant; R) was interpreted using the Clinical and Laboratory Standards Institute (CLSI)⁴⁶ and European Committee on Antimicrobial Susceptibility Testing (EUCAST) criteria³¹. S. aureus ATCC 25923 was used as a control strain (zone diameter ≥ 22 mm). Minimum inhibitory concentrations (MICs) of cefoxitin (0.03 to 32 µg/mL), mupirocin (> 256 µg/mL), and fusidic acid (0.03 to 32 µg/mL) were determined by broth microdilution assay for phenotypic detection of methicillin, mupirocin, and fusidic acid resistances.

Detection of antimicrobial resistance genes by PCR

Genomic DNA of the S. aureus was extracted as a template of PCR. mecA, mupA, fusA genes were amplified by PCR using specific primers (Supplementary Fig.S2 online)^9,10,11,47. The PCR conditions were 30 cycles of 94 °C (mecA) or 95 °C (mupA and fusA) for 30 s; 55 °C (mecA) or 57 °C (mupA and fusA) for 30 s and 72 °C for 4 min. Genotypes of methicillin resistance and mupirocin resistance were detected by the presence of mecA and mupA amplicons using agarose gel electrophoresis. The PCR product of the fusA gene was purified using the QIAquick PCR purification kit for nucleotide sequencing at the 1 st BASE Inc, Malaysia.

Analysis of nucleotide sequences

Nucleotide sequences and deduced amino acid sequences were analyzed using Online Software available at the National Center for Biotechnology Information (NCBI). (https://blast.ncbi.nlm.nih.gov), SnapGene (version 5.0.7) and ExPASy (www.expasy.org). Multiple sequence alignment was performed using Multilin (http://multalin.toulouse.inra.fr/multalin). The fusA sequences of fusidic acid-resistant S. aureus isolates were compared with the fusA sequence from S. aureus NCTC 8325 (GenBank accession no. NC_007795), using Mega 4 software (Biodesign Institute, Tempe, AZ, USA).

Molecular docking

The AutoDock v4.2 tool was used to perform a molecular docking simulation of fusidic acid with both wild-type EF-G and compared with mutated EF-G (mEF-G). The structure of wild type EF-G obtained from rcsb.org/(PDB: 2XEX), and mutations were modeled by PyMOL [DeLano, 2002] (The PyMOL Molecular Graphics System, Version 1.2r3pre, Schrödinger, LLC. https://www.pymol.org). Both molecules were analysed in the same manner by deleting the co-crystalized waters, adding polar hydrogen, and computing gasteiger charge. The ligand was modified in silico by adding polar hydrogen and computing gasteiger charge, before initiating the docking process. The grid file was generated with 60 × 60 × 60 as the number of points in x, y, and z directions, and the center spacing was 0.375Å. A docking output file was prepared with Lamarckian Genetic Algorithm using default settings. Finally, both grid and docking output files were performed using a script of autogrid4 and autodock4 (The Scripps Research Institute, https://www.scripps.edu). The interactions were analyzed using MGL tools.

Synthesis of monolaurin

Figure 1 shows the chemical structure of monolaurin. 0.3 g (2.27 mmol) of 1,2-isopropylideneglycerol in 10 mL of CH₂Cl₂ under N₂ gas conditions was mixed with lauroyl chloride (2.25 mmol) and triethylamine (3.4 mmol)⁴⁸. The reaction mixture was stirred for 30 min at room temperature and washed with water before evaporation using a rotatory evaporator⁴⁹. The concentrated compound (0.5 mmol) was dissolved in 50 mL of acetone, then mixed with 10 mL of aqueous 3 N HCl and stirred for 1.5 h at room temperature. The solution was re-evaporated, neutralized with NaHCO₃ and extracted with CH₂Cl₂. The residue was purified by silica gel column chromatography and the final product (monolaurin) was collected as a powder. The spectral signals of monolaurin as determined by Nuclear Magnetic Resonance (NMR) (MestReNova, Version 15.1.0) of proton¹H NMR) were¹:H-NMR (500 MHz, CDCl₃): δ 4.16 (dd, J = 13.6, 5.3 Hz, 2 H), 3.91 (s, 1 H), 3.67 (d, J = 4.0 Hz, 1 H), 3.59 (d, J = 5.8 Hz, 1 H), 2.42 (s, 2 H), 2.33 (s, 2 H), 1.61 (s, 2 H), 1.24 (s, 16 H), 0.86 (s, 3 H); and carbon¹³C-NMR)¹³:C-NMR (126 MHz, CDCl₃): δ 174.51, 70.35, 65.23, 63.43, 34.25, 31.99, 29.69, 29.54, 29.42, 29.34, 29.21, 24.99, 22.77, 14.21 (as shown in Fig. 1).

Monolaurin susceptibility testing

The MIC and minimum bactericidal concentration (MBC) of monolaurin were determined using the broth microdilution method. Monolaurin powder was dissolved in 30% ethanol (w/v), and concentrations of 0.0625 to 32 µg/mL were adjusted using two-fold dilution with Muller Hinton II broth (cation-adjusted) (BBL, USA). Antimicrobial-resistant S. aureus (1.5 × 10⁸ CFU/mL) clinical isolates were incubated with different concentrations of monolaurin at 35 °C for 24 h. The MIC was determined at the lowest concentration of monolaurin that resulted in no visible growth of isolates. To determine the MBC, each isolate (0.01 mL of bacterial suspension) of antimicrobial-resistant S. aureus at the same concentrations determined in the MIC was subcultured and inoculated onto blood agar and further incubated at 35 °C for 24 h. The concentration of monolaurin with no growth of isolates was identified.

Cytotoxicity assay

Primary epidermal keratinocytes (HEKn) and dermal fibroblasts were seeded (1 × 10⁵ cells/mL) into 96-well plates and incubated with Dermal Cell Basal Medium and Dulbecco’s Modified Eagle Medium (DMEM), respectively (Cytiva, USA) at 37 °C in 5% CO₂ for 24 h. Different concentrations of monolaurin (0.0625–32 µg/mL) were added into each well and the cells were incubated at 37 °C in 5% CO₂ for 24 h. The cell culture media was removed, and fresh medium with 10 µL of thiazolyl blue tetrazolium bromide method (MTT) (Sigma, USA) was added into each well, followed by incubation at 37 °C in 5% CO₂ for 4 h. The solution was discarded and DMSO was added to dissolve the formazan crystals. The optical density (OD) was measured at 570 nm using a microplate reader.

Measurement of inflammatory cytokines

HEKn and dermal fibroblasts (N = 3) were seeded (1 × 10⁵ cells/mL) into 96-well plates and treated with and without peptidoglycan (5 µg/mL, induction of inflammatory cytokine production), or monolaurin (2 µg/mL) for 24 h. The supernatants were collected to measure inflammatory cytokine production using ELISA kits (Thermo, USA); interleukin-6 (IL-6), interleukin-8 (IL-8), and tumor necrosis factor-alpha (TNF-α). The optical density (OD) was measured at 450 and 570 nm by a microplate reader.

Statistical analyses

The statistical analyses were performed using GraphPad Prism version 9.2.0 (GraphPad Software, USA) (https://www.graphpad.com), Pearson’s chi-squared test and Fisher exact test (for categorical data) and Paired t-test (for continuous data). The results are presented as the mean ± standard deviation (SD) and differences with a p-value < 0.05 were considered statistically-significant.