Paeonol reduced lesions in LL-37-induced rosacea-like mice, correlating with suppressed S100A9-NF-κB signaling
Synopsis
Using network pharmacology to predict shared targets of paeonol and rosacea, then validating in an LL-37-induced rosacea-like BALB/c mouse model, the study found that 150 mg/kg paeonol markedly lowered redness area and score, reduced inflammatory cell and CD4+ T-cell infiltration and angiogenesis, and downregulated S100A9 and p65 phosphorylation, indicating its effects correlate with suppressed S100A9-NF-κB signaling.
FIGURE 1 The interaction of mice and the network pharmacology of Pae and rosacea. (a) All groups were shaved 24 h before treatment. The model groups were injected intradermally on the back skin with LL-37 (40 mL, 320 mM) four times at intervals of 12 h. The control groups were also administered intradermal injections of PBS (40 mL) at the same frequency as the model groups. The DH and Pae groups received the first intradermal injection of an equivalent volume of LL-37 (320 mM). After the first interval of 12 h, the first injection was repeated in the DH and Pae groups, and the first intraperitoneal injection was performed at 30 mg/kg for DH and 50, 100, and 150 mg/kg for Pae. The DH and Pae groups were administered a third intradermal injection of LL-37 after the second interval of 12 h. At 12 h, the DH and Pae groups received the final intradermal and intraperitoneal injections at the same dose as the first intradermal and intraperitoneal injections. Then, 12 h after the last injection, all mice were sacrificed. (b) The heatmap shows the differentially expressed genes in rosacea. Red indicates upregulated differentially expressed genes, and green indicates downregulated differentially expressed genes. (c) The volcano plot illustrates differentially expressed genes. (d) The Venn diagram illustrates Pae- related targets and rosacea-related targets. (e) PPI network for overlapping proteins between Pae-related targets and rosacea-related targets. (f) Simplified PPI network for the overlapping proteins (larger circles and brighter colors indicate closer interactions).
· Page 4Interpretation
Network pharmacology identified 11 intersecting targets between 96 paeonol-related targets and 913 rosacea-related differentially expressed genes, with CXCL8, IL-1β, and CCL2 ranking as the top three hub targets in a PPI network of 11 nodes and 396 edges. No prior report on paeonol in rosacea existed; this work first intersects paeonol's pharmacological target profile with the rosacea differential expression profile to produce a candidate molecule list. Database integration across PubChem, TCMSP, HIT, BATMAN 2.0, Swiss Target Prediction, and GEO dataset GSE65914 (10 healthy volunteers and 19 rosacea patients) with STRING/Cytoscape network analysis; this is computational prediction, and the authors state animal experiments were needed for confirmation.
In LL-37-induced rosacea-like mice, 150 mg/kg paeonol reduced mean redness area from 11.97 ± 3.73 mm² to 0.35 ± 0.23 mm² and redness score from 2.63 ± 0.59 to 1.00 ± 0.38 (both p < 0.05), whereas the 50 and 100 mg/kg groups showed no significant improvement in redness area or score (p > 0.05). This provides in vivo phenotypic evidence for a dose-related effect of paeonol on rosacea-like lesions, with the high-dose group histologically not significantly different from controls (p > 0.05). 42 female BALB/c mice randomized to control, model, doxycycline, and three paeonol dose groups (n = 7 per group), dosed once daily for two consecutive days, scored and photographed 12 h after the final injection, with sample size based on power analysis.
Paeonol treatment was accompanied by reduced cutaneous CD31, CD4+ T cells, and mast cell infiltration, with CD31 significantly decreased at 150 mg/kg (p < 0.01) and CD4+ T cells significantly decreased at 100 and 150 mg/kg (both p < 0.01). Extends paeonol's action beyond pure anti-inflammation to vascular and adaptive immune dimensions, matching the vascular and immune abnormalities characteristic of rosacea. Histological quantification by immunofluorescence (CD31, CD4) and toluidine blue staining (mast cells counted in five random 200× fields), which are correlative observations.
Paeonol (100 and 150 mg/kg) significantly reduced S100A9 expression and inhibited p65 phosphorylation, with 150 mg/kg stronger than 100 mg/kg; the authors therefore propose the therapeutic effects correlate with suppressed S100A9-NF-κB signaling. Links the targets suggested by network pharmacology to the S100A9-NF-κB axis in vivo, offering direction for follow-up mechanistic work. Consistent S100A9 and p-P65 changes across immunofluorescence, RT-qPCR, ELISA, and Western blot, though the authors explicitly note the mechanism currently rests on correlative evidence without in vitro functional validation.
Perspective
The work sits at the early drug-discovery stage for rosacea inflammation intervention: network pharmacology and molecular docking generate candidate target hypotheses, and the LL-37-induced BALB/c mouse model provides in vivo phenotypic and signaling validation. For readers, its value lies in preliminary clues about paeonol's dose-response and the S100A9-NF-κB axis, useful for follow-up mechanistic experiments, formulation development, or combination-therapy design; the authors also note paeonol may be a candidate for alternative or combinatorial intervention rather than a substitute for doxycycline.
Boundaries noted by the authors include: the LL-37 model is a 48-hour acute inflammation model, so long-term efficacy, safety, and toxicity remain unproven; the S100A9-NF-κB mechanism rests on correlative evidence without in vitro functional validation; PPAR-γ-related mechanisms were not characterized within this study's scope; and whether S100A9-IL-17 or TLR4/MyD88-dependent cascades mediate paeonol's improvement of rosacea-like lesions was not experimentally validated. In addition, the loaded text reflects an incomplete reading scope, and some figure content (such as the specific images and full statistical details of Figures 1–5) is not included, so complete interpretation of individual values and between-group comparisons should refer to the original figures.