Staphylococcus aureus Virulent PSMα Peptides Induce Keratinocyte Alarmin Release to Orchestrate IL-17-Dependent Skin Inflammation

Abstract

Staphylococcus aureus commonly colonizes the epidermis, but the mechanisms by which the host senses virulent, but not commensal, S. aureus to trigger inflammation remain unclear. Using a murine epicutaneous infection model, we found that S. aureus-expressed phenol-soluble modulin (PSM)α, a group of secreted virulence peptides, is required to trigger cutaneous inflammation. PSMα induces the release of keratinocyte IL-1α and IL-36α, and signaling via IL-1R and IL-36R was required for induction of the pro-inflammatory cytokine IL-17. The levels of released IL-1α and IL-36α, as well as IL-17 production by γδ T cells and ILC3 and neutrophil infiltration to the site of infection, were greatly reduced in mice with total or keratinocyte-specific deletion of the IL-1R and IL-36R signaling adaptor Myd88. Further, Il17a^-/-f^-/- mice showed blunted S. aureus-induced inflammation. Thus, keratinocyte Myd88 signaling in response to S. aureus PSMα drives an IL-17-mediated skin inflammatory response to epicutaneous S. aureus infection.

Keywords: Agr virulence; IL-1; IL-36; Myd88; PSMs; S. aureus; alarmins; pathogen virulence; skin infection.

Figure 1. Keratinocyte Myd88 is important for epicutaneous S. aureus-induced skin inflammation

A, C57BL/6 (WT), K14-CreMyd88^fl/fl (Myd88^Δker) and Myd88^−/− mice were epicutaneously colonized with S. aureus. WT mice treated with PBS are shown for comparison. Representative macroscopic images and hematoxylin and eosin (HE)-stained skin sections 7 days after colonization (n=3 to 7 mice per group). Scale bars, 100 μm. B, Day 7 skin disease scores of WT, Myd88^Δker and Myd88^−/− mice colonized with S. aureus. C–E, The number of S. aureus colony forming unit (CFU) (C), numbers of neutrophils per high power field (D), and epidermal thickness (E) in the skin of WT, Myd88^Δker and Myd88^−/− mice 7 days after colonization with S. aureus. Each dot represents a mouse (B, C). Data are presented as mean ± SD (D, E). Results shown represent combined data of 2 independent experiments. ND; not detected, n.s.; not significant, *P<0.05; **P<0.01, by Kruskal-Wallis test.

Figure 2. Both IL-1R and IL-36R contribute to skin inflammation induced by epicutaneous S. aureus colonization

A, WT, Il1r^−/−, WT mice treated with IL-36R blocking antibody (IL-36RAb), and Il1r^−/− mice treated with IL-36RAb were epicutaneously colonized with S. aureus for 7 days. Representative macroscopic images and HE-stained skin sections of mice colonized with S. aureus or treated with PBS (n=4 to 8 mice per group). Scale bars, 100 μm. B–E, Day 7 skin disease scores (B), S. aureus CFU in the skin (C), the number of neutrophils in the skin (D) and epidermal thickness (E) of WT, Il1r^−/−, WT mice treated with IL-36RAb, and Il1r^−/− mice treated with IL-36RAb (n=4 to 8 mice per group). WT mice treated with PBS are shown for comparison. Each dot represents a mouse (B, C). Data are presented as mean ± SD (D, E). Data represent combined results from 3 independent experiments. ND; not detected, n.s.; not significant, *P<0.05 and **P<0.01, by Kruskal-Wallis test (B, C) or by one-way ANOVA test with Bonferroni’s correction (D, E).

Figure 3. IL-1α and IL-36α are induced by S. aureus via Myd88 signaling in keratinocytes

A, WT mice treated with IL-1α blocking antibody (IL-1αAb) or isotype-matched control Ab were epicutaneously colonized with S. aureus. Representative macroscopic images of mice colonized with S. aureus (n=5 mice per group). B–C, Skin disease scores (B) and S. aureus CFU in the skin (C) of WT mice treated with IL-1α Ab or control Ab. Each dot represents a mouse. D, Skin tissues of WT, and Myd88^Δker and Myd88^−/− mice colonized with S. aureus or treated with PBS were stained with Hoechst stain (blue) and antibody against IL-1α (red) or Hoechst stain (blue) and antibodies against S. aureus (red) and IL-36α (green). Scale bars, 50 μm. E, The numbers of S. aureus per high power field are (HPF) are shown. Each dot represents average results from an individual mouse. Data are representative of 2 independent experiments. F–K, IL-1α (F, H, J) and IL-36α (G, I, K) release of differentiated primary KCs isolated from WT and Myd88^−/− mice (F, G), WT and Ilr1^−/− mice (H, I), or WT mice in the presence of anti-IL-36 neutralizing Mab or isotype-matched control Mab (J, K) and stimulated with culture supernatants of S. aureus for indicated time. IL-1α and IL-36α were detected by ELISA assay and immunoblotting, respectively. β-actin in whole cell lysates is shown as loading control. Data are presented as mean±SD. Data are presented as mean ± SD. Data are representative of at least 2 independent experiments. ND; not detected, n.s.; not significant, *P<0.05 and **P<0.01, by unpaired two-tailed Mann-Whitney U test (B, C, F, H, J) or Kruskal-Wallis test (E).

Figure 4. IL-17 is critical for skin inflammation induced by epicutaneous S. aureus colonization

A, Production of IL-17A, IL-17F, IL-22, IFN-γ and GM-CSF by skin cells isolated from WT mice colonized with S. aureus or treated with PBS. Intracellular cytokine production was assessed in gated CD45⁺CD90⁺ cells on day 7 after pathogen colonization by flow cytometry. Representative flow cytometry profiles (left panels) and the number of cytokine-producing cells (right panel). Results in right panel represent mean ± SD of 2 experiments. B, Production of IL-17A by CD45⁺ cells in the skin of WT and Myd88^−/− mice 7 days after epicutaneous infection. Representative flow cytometry profiles (left panels) and the number of IL-17A-producing cells (right panel). Results in right panel represent mean ± SD of 2 experiments. C, IL-17A and IL-17F production in skin tissues of WT mice, Myd88^Δker mice, Il1r^−/− mice, WT mice treated with IL-36RAb and Il1r^−/− mice treated with IL-36RAb 7 days after pathogen colonization. WT mice treated with PBS are shown for comparison. Each dot represents a mouse. Data represent combined data of 3 independent experiments. D WT and Il17a^−/−f^−/− mice were epicutaneously colonized with S. aureus or treated with PBS. Representative macroscopic images and HE-stained skin sections 7 days after colonization (n=7 mice per group). Scale bars, 100 μm. E–H, Skin disease scores (E), S. aureus CFU in the skin (F), the numbers of neutrophils (G) and epidermal thickness (H) of WT and Il17a^−/−f^−/− mice colonized with S. aureus. WT mice treated with PBS are shown for comparison. Each dot represents a mouse (C, E, F). Data are presented as mean ± SD (A, B, G, H). Results shown represent combined data of 3 independent experiments. ND; not detected, n.s.; not significant, *P<0.05; **P<0.01, by unpaired two-tailed Mann-Whitney U test (A, B, E, F, G, H) or Kruskal-Wallis test (C).

Figure 5. Both γδ T cells and ILC3 contribute to skin inflammation after S. aureus colonization

A, IL-17A-producing γδ T cells, ILC3 and αβ T cells were evaluated by flow cytometric analysis after epicutaneous S. aureus colonization. Flow cytometric analysis of lineage (B220, CD11b, CD11c, Gr-1, NK1.1)-negative cells and γδ T cells was performed on gated CD45⁺CD90⁺IL-17A⁺ skin cells. Data are representative of 3 independent experiments. B, WT and Tcrδ^−/− mice were epicutaneously colonized with S. aureus. WT mice treated with PBS are shown for comparison. Representative macroscopic images and HE-stained sections of mouse skin on day 7 after colonization (n=10 to 15 mice per group). Scale bars, 100 μm. C–E, Day 7 skin disease scores (C), S. aureus CFU in the skin (D), neutrophil numbers in the skin (E) and epidermal thickening (F) of WT and Tcrδ^−/− mice colonized with S. aureus. Each dot represents a mouse (C, D). Data are presented as mean ± SD (E, F). Results represent combined data of 4 independent experiments. G, IL-17A-producing γδ T cells and ILC3 were evaluated by flow cytometric analysis in WT and Tcrδ^−/− mice after S. aureus colonization. Representative flow cytometric profiles of CD3 and lineage labeling on CD45⁺CD90⁺IL-17A⁺ skin cells (left panels). The number of IL-17A⁺ γδ T cells and ILC3 in WT and Tcrδ^−/− mice (right panels). Results on right panels represent mean ± SD of 3 experiments. H, WT mice and Tcrδ^−/− mice treated with control Ab and anti-CD90 Ab were epicutaneously colonized with S. aureus. Representative macroscopic images and HE-stained sections of mouse skin on day 7 after colonization (n=7 mice per group). Scale bars, 100 μm. I–J, Skin disease scores (I) and S. aureus CFU in the skin (J) of WT mice and Tcrδ^−/− mice treated with control Ab and anti-CD90 Ab. Each dot represents a mouse. Results represent combined data of 2 independent experiments. ND; not detected, n.s.; not significant, *P<0.05 and **P<0.01, by using unpaired two-tailed Mann-Whitney U test (C–G) or by one-way ANOVA test with Bonferroni’s correction (I, J).

Figure 6. PSMα peptides induce the release of keratinocyte IL-1α and IL-36α to mediate skin inflammation

A–B, IL-1α release (A) and cytotoxicity (B) of primary KCs from WT mice stimulated with culture supernatant of WT and Δpsma S. aureus (LAC strain) for indicated time. Data are presented as mean ± SD. C, IL-36α release from primary KCs stimulated with culture supernatant of WT or Δpsma S. aureus for indicated time. IL-36α was detected by immunoblotting. β-actin in whole cell lysates is shown as loading control. D, Representative macroscopic images (top panels) and HE-stained sections (middle upper panels), and sections stained with Hoechst stain (blue) and antibody against IL-1α (red) (middle lower panels) and stained with Hoechst stain (blue) and antibodies against S. aureus (red) and IL-36α (green) (bottom panels) of the skin from WT mice colonized with WT and Δpsma S. aureus or treated with PBS (n=5 to 8 per group) 7 days post infection. Epidermis/dermis border is marked by dotted white line. Scale bars, 100 μm (middle upper panels), 50 μm (middle lower panels) and 25 μm (bottom panels). E–H, Day 7 skin disease scores (E), S. aureus CFU in the skin (F), quantification of neutrophils in the lesional skin (G) and epidermal thickening (H) of WT mice colonized with WT and Δpsma S. aureus or treated with PBS (n=5 to 8 per group). Each dot represents a mouse (E, F, I). Data are presented as mean ± SD (G, H). I, The amounts of IL-17A and IL-17F in the lesional skin of WT mice colonized with WT or Δpsma S. aureus or WT mice treated with PBS for 7 days. Each dot represents a mouse. Results represent combined data of 3 independent experiments. Data are representative of at least 2 independent experiments. ND; not detected, n.s.; not significant, *P<0.05 and **P<0.01, by unpaired, two-tailed Mann-Whitney U test.