S. aureus USA300 JE2 (Fey et al., 2013) was grown overnight in CnT-Prime medium (CELLnTEC), diluted to an initial optical density value at 600 nm (OD₆₀₀) of 0.05 in fresh CnT-Prime medium and grown with shaking (180 rpm) at 37°C either to the exponential (OD₆₀₀ = 0.5) or post-exponential (OD₆₀₀ = 0.5 + 4 h) growth phase.

Ex vivo skin explants were obtained from skin tumor surgery in which pieces of healthy skin were removed for surgical reasons to compensate for the circular excisions upon surgical closure. This approach was approved by the local ethics committee of the Medical Faculty RWTH University of Aachen, Germany (EK 404/19). In accordance with the Declaration of Helsinki Principles guidelines written consent was obtained from all participants involved in the study.

Immediately after surgery, subcutaneous adipose tissue was removed and skin explants were punched into defined pieces (4 mm in diameter) using surgical biopsy punches (pfm medical AG). Skin samples were incubated for 12 h in DMEM (Pan Biotech) supplemented with 10% fetal calf serum (Pan Biotech) and 1x antibiotic-antimycotic (Gibco^TM 15240062) to remove the remaining skin flora. The skin area was routinely disinfected prior to surgery and subsequent incubation in DMEM supplemented with 1x antibiotic-antimycotic (12 h) ensured a significant reduction of the natural skin flora. In order to enable colonization with S. aureus samples were washed twice with pre-warmed 37°C phosphate-buffered saline (PBS) [Pan Biotech] and placed into Millicell^® 24-well cell culture inserts with a 0.4μm pore size membrane (Merck). Inserts were placed into 24-well plates prefilled with 300 μl of CnT-Prime medium and explants were cultured at the air-liquid interface at 37°C and 5% CO₂. After 4 days, CnT-Prime medium was added to the wells, ensuring the continuous supply of nutrients to the tissue from below. Either exponential or post-exponential grown S. aureus USA300 JE2 bacteria were harvested by centrifugation at 4.000 rpm for 10 min and resuspended in CnT-Prime medium. 9 μl of bacterial suspension corresponding to 1 × 10⁶ colony-forming units (CFUs) were applied step wise in 3–4 μl steps to the surface of the skin explants.

As a control, we randomly applied the concentrated medium to blood agar plates to examine for the presence of S. aureus. No growth of S. aureus was detected in any of the plates (data not shown). For transcript analyses infected skin samples were incubated for 15 min, 1, 2, 4, 6, 18, 28, and 74 h, as well as 5, 6, 7, and 8 days at 37°C and 5% CO₂. At the defined times, skin explants were removed from the inserts, cut into small pieces and dissolved in 1 ml TRIzol^TM LS reagent (Thermo Fisher Scientific) for RNA isolation. The bacterial inoculum that was applied to the skin’s surface was also transferred into 1 ml of TRIzol^TM LS reagent. In addition, wells containing 300 μl of CnT Prime medium only, were also inoculated with 1 × 10⁶ CFUs of the inoculum suspension to analyze the in vitro growth without skin. Bacteria grown in medium only were also transferred into 1 ml of TRIzol^TM LS reagent for RNA isolation at the same time points already defined for the ex vivo skin explants up to 18 h.

Bacteria were lysed and RNA isolation was performed as described previously (Burian et al., 2010a). To eliminate contaminating DNA each RNA sample was digested with 8 U of RNase-free DNase I (Roche), 2 μl 10 × incubation buffer (Roche), and 16 U of RNasin ribonuclease inhibitor (Promega) for 30 min at 25°C. DNase treatment was carried out twice for each sample. DNase I treatment was stopped using DNase inactivation reagent (Thermo Fisher Scientific). Three microliter of total RNA were transcribed into cDNA using Super Script III Reverse Transcriptase (Thermo Fisher Scientific) and 200 ng of random hexamer primers (Thermo Fisher Scientific). Reverse transcription was performed as described in the instructions of the Superscript manufacturer. cDNAs were diluted 1:3 with nuclease-free water (Thermo Fisher Scientific) and frozen at −20°C using Eppendorf LoBind Tubes (Eppendorf, Hamburg, Germany) for prolonged storage.

Relative quantification of S. aureus transcripts by qPCR was carried out using the 7300 Real Time PCR instrument (Applied Biosystems) in combination with the KAPA SYBR^® FAST qPCR Master Mix (2x) ABI Prism (Merck). Master mixes were prepared as following: 8 μl KAPA SYBR^® FAST qPCR Master Mix (2x) ABI Prism, 8 μl nuclease-free water, 1 μl of each primer (see Supplementary Table 1; Burian et al., 2010a,b), and 2 μl cDNA. The following temperature profile was utilized for amplification: denaturation for 1 cycle at 95°C for 15 s and 55 cycles at 95°C for 27s, 55–60°C for 10 s, and 72°C for 27 s with fluorescence acquisition at 72°C. Melting-curve analysis was done at 60–97°C with stepwise fluorescence acquisition. Relative quantities of transcripts were calculated by a standard curve for each gene generated using sixfold serial dilution of S. aureus USA300 wild type RNA.

The membrane-permeable dye thiazolyl blue tetrazolium bromide (MTT) was used to assess skin explant viability. Four millimeter skin biopsies were cultured as described above up to 10 days. From day 5 onward skin explants were washed twice with PBS and further incubated in 2 mg/ml MTT (Sigma-Aldrich) for 2 h at 37°C and 5% CO₂. MTT is reduced from yellow color to purple formazan in living skin biopsies.

Apoptosis was assessed on skin tissue using an in situ Apoptosis Detection Kit (Sigma-Aldrich) according to the manufacturer’s protocol. After the final washing step, sections were mounted in ibidi Mounting Medium with DAPI (ibidi). Staining was visualized using a Leica DMI4000 B microscope (Leica).

Hematoxylin and eosin (H&E) staining, as well as immunostaining, were carried out as previously described by Huth et al. (2018). Briefly, for immunofluorescence, 4 μm cryosections were fixed for 10 min in acetone at 4°C. First antibodies keratin 10 (Dako) and filaggrin (AKH1) (Santa Cruz) were diluted with Antibody Diluent (Dako) and incubated at room temperature for 1 h. Following washing steps with PBS, the sections were incubated in fluorochrome-conjugated secondary antibody Alexa Fluor 488 IgG H + L (Molecular Probes) for 1 h at room temperature. Cell nuclei were stained with DAPI (Applichem). All image processing and analyses were performed with ImageJ software (National Institutes of Health).

Statistical analysis was performed with the Prism 9.0.2 package (GraphPad Software) using multiple t-tests. P < 0.05 was considered to be statistically significant.

Healthy human skin tissue was cultured at the air-liquid interface for up to 10 days (Figure 1A). To assess the viability of the skin explants we analyzed the ability of the tissue to reduce the MTT. Reduction from yellow color to purple formazan in living skin biopsies was observed for up to 8 days. At later time points, the cell viability decreased in skin explants (Figure 1B). In addition, we performed a validation of the integrity of our skin explants following the validation described by Neil et al. (2020). Using H&E stained slides, neither signs of spongiosis, necrosis or cell death could be observed. The epidermis and dermis are not separated over this period of time and no alterations in the corneal layer were observed. Furthermore, the terminal differentiation marker filaggrin and keratin 10 were unaffected during this period (Supplementary Figure 2). We decided to follow the S. aureus adaptation process for up to day 8. During this period, we confirmed stable colonization of the skin samples with S. aureus by demonstrating persistent expression levels of the house keeping gene gyrB (Figure 1C). gyrB transcripts were stable over time, both from skin samples that were colonized with exponentially grown bacteria as well as from skin samples colonized with post exponentially grown bacteria (mean gyrB exponential bacteria 2.20 × 10⁴ and mean gyrB post exponential bacteria 1.60 × 10⁴). As expected, gyrB values of bacteria grown in vitro (medium only without skin) increased over time, indicating a growth in the cell culture medium (Figure 1C).

To profile the expression pattern of S. aureus during skin colonization we performed transcript analysis by quantitative real-time PCR (qPCR) starting 15 min after initial colonization. To evaluate, whether contact with the epidermis induced a uniform change in the expression pattern independently of the initial transcription level, we colonized skin samples with either exponentially or post-exponentially grown S. aureus (Figure 1A). In addition, wells containing medium only, were also inoculated with the corresponding inoculum to analyze gene expression during in vitro growth in the absence of skin contact. Initially, we analyzed transcription of RNAIII of the quorum sensing system agr, which is usually repressed during exponential growth, and transcription of clfB as marker for an exponentially expressed gene (Figure 2). Strikingly, agr transcription was immediately switched off upon bacterial contact to the skin surface. Notably, the massive agr downregulation also took place when skin samples were colonized with post-exponentially grown bacteria in which agr is highly active (Figure 2A). Thus, contact with skin led to a downregulation of agr in post-exponentially and exponentially grown bacteria and this downregulation lasted for up to 18 h. In contrast to agr suppression, transcription of clfB was induced in post exponentially grown bacteria upon skin contact or remained high when inoculated with expontential grown bacteria (Figure 2B). In summary, S. aureus contact with human skin results in consistent transcriptional changes independent of the inoculum’s growth phase.

Subsequently, we analyzed the expression of 21 S. aureus genes associated with a variety of cellular functions such as virulence regulation, toxin production, adhesion, immune evasion and cell wall dynamics. The expression of these genes during colonization of the human nasal cavity was previously assessed (Burian et al., 2010b) and was used for comparison (Figure 3). Skin explants were colonized with exponentially grown bacteria and gene expression was followed over 8 days during the colonization process. To define early and late colonization, time points up to 74 h were defined as early and 5–8 days as late colonization (Figure 1A). The results are summarized according to the functional categorization of target genes.

The data support that toxicity of S. aureus is dampened through inactivation of virulence regulators such as agr. Moreover, toxins might be inactivated through proteolytic cleavage (Gimza et al., 2021). On the other hand S. aureus secreted proteases might also contribute to tissue degradation and inactivation of host defense systems (Singh and Phukan, 2019). For Staphylococcus epidermidis the protease EcpA can be a deleterious component of the skin microbiome in AD (Cau et al., 2021). Therefore, we subsequently focused on the analysis of genes coding for major proteases: the metalloprotease aureolysin (aur), the serine proteases sspA (V8 protease), and splA (member of the protease-like proteins), and the cysteine proteases staphopain A and B (encoeded by scpA and sspB). Transcription of all proteases was strongly induced during the entire time period of colonization starting already after 15 min with the exception of splA, which was transcribed at low levels and peaked only during late colonization (Figure 4).