This study used twenty male mice (6–8 weeks old), comprising ten C57BL/6 mice (National Laboratory Animal Center, Taipei, Taiwan) for bone marrow cells extracted, and five OT-I and five OT-II transgenic mice (The Jackson Laboratory, Bar Harbor, ME, USA) for T-cell isolation. Contact hypersensitivity (CHS) was modeled in mice using 2,4-dinitro-1-fluorobenzene (DNFB). Fifteen males C57BL/6 mice (7 weeks old, 18–22 g) were separated into three groups (n = 5) at random for a single experiment: vehicle, control, and hibifolin. The mice were housed in cages with separate ventilation and under specified pathogen-free (SPF) conditions at National Chung Hsing University. The temperature (22 ± 2°C) and relative humidity (55 ± 10%) were maintained within a standard 12:12 h light/dark cycle. Ad libitum access to autoclaved water and regular laboratory mouse food was given to the animals. A randomized controlled design was applied, with group allocation by random number table; investigators measuring ear thickness and histology were blinded to group assignment. Only healthy mice were included, and no animals or data were excluded. All experimental procedures were conducted strictly in accordance with the protocols approved by the Institutional Animal Care and Use Committee (IACUC) of National Chung Hsing University (IACUC: 112-074) and the university’s ethical guidelines for the use of laboratory animals.

To obtain bone-marrow cells, the mice’s femurs were flushed with RPMI 1640 medium (Gibco™, 31800-022, Grand Island, NY, USA) via syringe fitted with a 25-gauge needle. RBCs were subsequently removed using RBC lysis buffer (BioLegend, 420302, San Diego, CA, USA). Bone marrow cells were cultivated in medium (RPMI-1640) added with (10%) heat inactivated fetal bovine serum (FBS) (Gibco™, A52567-01, Grand Island, NY, USA), 100 units/mL penicillin G, 100 µg/mL streptomycin (Capricorn Scientific GmbH, AAS-B, Ebsdorfergrund, Germany), 20 ng/mL of recombinant mouse granulocyte-macrophage colony-stimulating factor (GM-CSF), and IL-4 (PeproTech, Cat. Nos. 315-03 and 214-14, Cranbury, NJ, USA). Every two days, the cultured media were replaced, and cells were kept at 37°C in an incubator with 5% CO₂. Cells that were weakly adherent and nonadherent were gathered on the seventh day of culture. The CD11c⁺ cells from this population were then magnetically separated using the CD11c MicroBeads UltraPure (Miltenyi Biotec, 130-125-835, Bergisch Gladbach, Germany). Flow cytometry (Accuri™ C5 flow cytometer; Accuri Cytometers, Ann Arbor, MI, USA) verified that the isolated CD11c⁺ cells, which represent immature BMDCs, had a purity of more than 80%.

Hibifolin (ChemFaces, CFN70410, Wuhan, China) was diluted to a 1000-fold stock solution using DMSO (Dimethyl sulfoxide, Alpha Biochemistry, D090405-0500, Taoyuan, Taiwan) as the solvent. The BMDCs (2 × 10⁶ cells/2 mL/well in 6-well) were incubated in medium containing 0.1% DMSO or specific concentrations of hibifolin for 1 h. 100 ng/mL LPS (lipopolysaccharide) (Sigma-Aldrich Chemie GmbH, L2880, Steinheim, Germany) was then used to activate the cells for either 4 or 18 h in order to measure the expression of TNF-α, IL-6, IL-12, and IL-23. Following the collection of the culture supernatant, the quantities of these cytokines were measured by ELISA kits (Catalog 430904, 431304, 433604, and 433704, correspondingly) supplied by BioLegend (San Diego, CA, USA).

The cytotoxicity of hibifolin on BMDCs was assessed using Cell Counting Kit-8 (CCK-8) examination (Sigma-Aldrich, 96992, St. Louis, MO, USA). BMDCs were grown at a concentration of 1 × 10⁵ cells/0.1 mL per well in plates with 96 wells at 37°C in an incubator that was humidified with 5% CO₂. They were then exposed to different concentrations of hibifolin (6.25 to 50 μM) for 24 h under two experimental conditions: (1) hibifolin by itself, or (2) hibifolin in combination with lipopolysaccharide (LPS, 100 ng/mL). Following the manufacturer’s directions, cell viability was evaluated following incubation.

The phenotypic characteristics of BMDCs cultured for seven days were evaluated via fluorescence-activated cell sorting (FACS) to assess surface markers. BMDCs were cultured at 2 × 10⁶ cells/mL and treated according to experimental groups, 100 ng/mL LPS was used to excite the experimental groups when either 0.1% DMSO or 25 µM hibifolin was present. The control group received 0.1% DMSO without LPS. Incubated the cells for 30 min at 4°C before being stained with BB515-conjugated anti-mouse CD11c (1:100, Clone N418, Cat. No. 565586) and one of the PE-conjugated antibodies: CD80 (1:100, Clone 16-10A1, Cat. No. 553769), anti-mouse I-A/I-E (MHC class II, 1:100, Clone M5/114.15.2, Cat. No. 557000), or CD86 (1:100, Clone GL1, Cat. No. 553692). BB515 Hamster IgG2, λ1 (Cat. No. 565404), PE Hamster IgG2, κ (Cat. No. 550085), PE Rat IgG2a, κ (Cat. No. 553930), and PE Rat IgG2b, κ (Cat. No. 555848) were used as corresponding isotype controls. The supplier of all antibodies, including isotype standards, was BD Pharmingen™ (San Diego, CA, USA). Every PE-conjugated marker was assessed using a different double-staining setup with CD11c, and all staining was carried out under identical circumstances. The gating strategy for BMDCs is illustrated in Supplementary Fig. S1. BD Accuri C6 Plus software (version 1.0.23.1, Accuri Cytometers, Ann Arbor, MI, USA) was used to evaluate fluorescence data obtained with an Accuri™ C5 flow cytometer (Accuri Cytometers, Ann Arbor, MI, USA).

2 µg/mL OVAP1 (OVA_257–264)/OVAP2 (OVA_323–339) peptides, provided via Echo Chemical Co., Taiwan, were administered to BMDCs resulting from (OT-I and OT-II) TCR transgenic mice. For one hour, BMDCs (2 × 10⁵ cells/mL) were treated with either 25 µM hibifolin or 0.1% DMSO at 37°C. After that, the cells were stimulated for 18 h using LPS (100 ng/mL), while the control group, given merely 0.1% DMSO. Following the course of treatment, the cells were gathered and given a PBS rinse. OT-I & OT-II animals, used to isolate (OVAP1) specific CD8⁺ T-cells and (OVAP2) specific CD4⁺ T-cells using Miltenyi Biotec’s MACS cell separation technology (South San Francisco, CA). To ensure regulated cell interactions, BMDCs and T cells were cocultured at a 1:5 ratio for 96 h. [³H]-thymidine incorporation was measured to evaluate T-cell proliferation. After 96 h, ELISA was used to check for the generation of IFN-gamma in the culture supernatants (BioLegend, Catalog 430805, San Diego, CA, USA).

Before being exposed to 100 ng/mL LPS, BMDCs (2 × 10⁶ cells/mL) were pretreated with either 0.1% DMSO or 12.5 μM and 25 μM hibifolin for an hour. Following 18 h of LPS induction, BMDCs were harvested, and whole-cell lysates were then made. The control group consisted of cells that had been exposed to 0.1% DMSO. For protein analysis, the Thermo Scientific™ Pierce™ (BCA) Protein Analyses (Thermo Fisher Scientific, Product No. 23227, Waltham, MA, USA) was used to measure the protein content of each lysate. 12% SDS–polyacrylamide gels were used to run 40 µg of protein lysate, which was then transferred onto PVDF membranes (Bio-Rad Laboratories, 1620177, Hercules, CA, USA). After that, 5% skim milk in TBS containing 0.05% Tween 20 was used to block these membranes for an hour. The primary antibodies were incubated at 4°C for one night to detect the presence of phospho-p38 (1:1000, Cat. No. 9211S), total p38 (1:1000, Cat. No. 9212S), phospho-p42/44 ERK1/2 (1:1000, Cat. No. 9101S), total p42/44 ERK1/2 (1:500, Cat. No. 4695S), phospho-JNK (1:1000, Cat. No. 4668S), GAPDH (1:20,000, Cat. No. 5174S), or total JNK (1:500) (provided by SC-571; Santa Cruz Biotechnology, Santa Cruz, CA, USA). Horseradish peroxidase-conjugated secondary antibodies (1:10,000, Jackson Immuno-Research, (111-035-003), West Grove, PA, USA) were added to the membranes after the primary antibody had been incubated. Band intensities were quantified using ImageJ software (version 1.54p, National Institutes of Health, Bethesda, MD, USA), protein bands were analysed using LumiFlash Ultima Chemiluminescent Substrate, HRP System (Visual Protein, LF08-500, Taipei, Taiwan), and the images were visualized with Hansor Luminescence Image System (Taichung, Taiwan).

BMDCs (2 × 10⁶ cells/mL), treated with either (0.1% DMSO) or 12.5 and 25 μM hibifolin for an hour in order to measure NF-κB transcriptional activity. After that, the cells were stimulated for 30 min in 6-well plates using 100 ng/mL LPS. Following stimulation, the BMDCs were gathered, and using the supplied technique, nuclear proteins were extracted using the Cayman Chemical NE-PER Nuclear and Cytoplasmic Extraction equipment (Item 10009277, Ann Arbor, MI, USA). Pierce provided a (BCA) protein analyse kit (Cat. No. 23225, Rock-ford, IL, USA), used to measure protein quantities in the extracts. As directed by the manufacturer, the Active Motif TransAM NF-κB p65 ELISA Kit (Cayman Chemical, Item No. 10007889, Ann Arbor, MI, USA) was used in each experiment to quantify NF-κB p65 activity using nuclear protein.

DNFB (Thermo Scientific, AAA1187122, Waltham, MA, USA) was dissolved in a mixture of acetone (TEDIA, AS1112-001, Fairfield, OH, USA) and olive oil (Sigma-Aldrich, 01514, Louis, MO, USA) (AOO, 4:1, v/v). For sensitization, 50 μL of 0.5% DNFB was topically applied once daily from day 1 to day 5 to a localized shaved area (approximately 1 cm × 3 cm) on the central abdomen of each mouse. The inner and outer surfaces of each mouse ear, challenged with (20 μL) of 0.2% DNFB on day 6. Supplementary Fig. S2 shows the experimental groups and treatment regimens. To summarize, three groups of mice were created (n = 5) at random: (1) control group, (2) DNFB group, and (3) DNFB plus hibifolin group (50 mg/kg). From day 1 to day 5, hibifolin was given orally by gavage once a day after being dissolved in a vehicle solution made up of 10% Cremophor EL (Merck Millipore, 238470, Darmstadt, Germany) in 0.9% NaCl (Honeywell Fluka, 31434-5KG, Charlotte, NC, USA). Control and DNFB groups received the same volume of vehicle solution following the same schedule. The dosage of hibifolin (50 mg/kg) was selected based on earlier in vivo studies showing therapeutic efficacy in models of infectious pneumonia and LPS-induced acute lung injury without signs of toxicity [19]. Experimental grouping and treatment timeline are summarized in Supplementary Fig. S2.

Following CHS induction, mouse ears were removed after euthanasia by CO₂, paraffin-embedded, fixed in neutral-buffered formalin (10%), and sliced at a thickness of 4 μm. Sections were photographed via light microscopy (Olympus, Tokyo, Japan), stained with H&E (hematoxylin and eosin) (Leica Biosystems, 3801522 & 3801600, Richmond, IL, USA). ImageJ software (version 1.54p, Wayne Rasband and collaborators, National Institutes of Health, Bethesda, MD, USA), used to measure epidermal thickness. According to the previously established semi-quantitative histological scoring, inflammatory cell infiltration and epidermal thickening were assessed on a measurement of 0–3 (0 = none, 1 = mild, 2 = moderate, and 3 = severe) [23]. Three sections per mouse were analyzed by two blinded investigators, with scores summed for final histological severity.

We conducted data analysis via version 10.0 of the soft-ware Graph-Pad Prism suite (GraphPad Soft-ware, Boston, MA, USA). Homogeneity of variances was assessed using the Brown-Forsythe test, and normality of residuals was assessed via the Shapiro-Wilk test. Dunnett’s multiple comparison assessment was used after one-way ANOVA to assess datasets that satisfied both assumptions (Figs. 1–4). Non-normal data were investigated using the Kruskal-Wallis test (Fig. 5) and then Dunn’s post hoc analysis. The mean ± standard deviation (SD) is used to display all data. p-values < 0.05 were deemed statistically significant.

AI-based tools were applied during manuscript preparation. ChatGPT (OpenAI, San Francisco, CA, USA; version GPT-4) was used for grammar editing and rephrasing to reduce textual similarity. All AI-generated content was critically reviewed and revised by the authors to ensure accuracy, originality, and compliance with academic standards. BioRender (BioRender.com) was used to create the graphical abstract.

The ability of mature dendritic cells (DCs) to generate and release cytokines that control inflammation and T-cell differentiation—two processes essential to adaptive immunity—is one of their main traits. Thus, we used a CCK-8 assay to first assess the possible cytotoxicity of hibifolin. The findings demonstrated that hibifolin (Fig. 1A) had no discernible effect on BMDC viability at concentrations lower than 25 μM (Fig. 1B). Similarly, hibifolin at similar concentrations did not show appreciable cytotoxicity in the existence of LPS (Fig. 1C). We then evaluated how hibifolin affected the release of pro-inflammatory cytokines brought on by LPS. LPS-induced pro-inflammatory cytokine protein expression, such as TNF-α (Fig. 1D), IL-6 (Fig. 1E), IL-12 (Fig. 1F), and IL-23 (Fig. 1G), was considerably and dose-dependently decreased by hibifolin, as illustrated in (Fig. 1D–G).

DC activation results in increased production of surface molecules, particularly costimulatory markers involved in antigen presentation and major histocompatibility complex (MHC). In order to engage with naive T-cells and promote their activation, these chemicals are essential [24,25]. We employed flow cytometry to evaluate the impact of hibifolin on the levels of these surface molecules in mouse DCs. As demonstrated in Fig. 2A,B, our findings demonstrate that hibifolin strongly suppresses the effect of co-stimulatory molecules like (CD80 and CD86) and MHC class II that are activated by LPS.

We investigated, effects of hibifolin on DCs’ capacity to trigger T-cell responses that are specific to the antigen ovalbumin (OVA). Immature DCs from mice were loaded with either OVA323–339 (OVAP2) or OVA257–264 (OVAP1) peptides. Prior to being activated with LPS, these cells were either pretreated with hibifolin or a control solution. Using [³H]-thymidine incorporation assays, we assessed their ability to stimulate the growth of allo-geneic OVA-specific CD4+ OT-II and CD8+ OT-I T-cells. The findings, which are shown in Fig. 3A,B, showed that OT-II and OT-II T-cell growth was strongly increased by LPS-activated BMDCs. Hibifolin pretreatment of BMDCs, however, considerably lessened this effect by reduction of proliferating both OT-II and OT-I cells. This implies that hibifolin hinders LPS-activated BMDCs’ ability to stimulate CD4⁺ and CD8⁺ T-cell proliferation. Additionally, we evaluated how hibifolin affected T cell production of IFN-γ (interferon-gamma). As seen in Fig. 3C,D, BMDCs treated with hibifolin and subsequently cocultured with OT-II CD4⁺ & OT-I CD8⁺ T-cells had a significant decrease in IFN-γ levels in the medium. The reaction of T-cells cocultured with BMDCs treated with DMSO in the existence of LPS, where production of IFN-γ was higher, was in contrast to this observation.

Prior studies have demonstrated the important functions of NF-κB and MAPKs in DC cytokine generation during LPS stimulation [26,27]. We examined how hibifolin impacts the MAPK & NF-κB pathways in mouse DCs stimulated by LPS in order to look into its suppressive effects. As illustrated in Fig. 5A,B, our findings demonstrated that hibifolin pretreatment reduced p38-MAPK phosphorylation upon LPS induction but had no discernible effect on JNK or ERK phosphorylation. Additionally, using a (Trans^AM) NF-κB transcription factor test kit to analyze nuclear extracts, the effect of hibifolin on NF-κB was assessed. LPS dramatically raised p65 (nuclear translocation) and NF-κB binding activity within 30 min, as shown in Fig. 5C. However, these increases were greatly diminished in BMDCs that had been pretreated with 25 μM hibifolin. These results imply that hibifolin’s capacity to modify these important signaling pathways may be the reason for its inhibitory influence on DC maturation in response to LPS.

When skin is exposed to allergens like DNFB, a T-cell-mediated immune response is triggered, resulting in CHS, a common antigen-specific inflammatory response mediated by DC [28]. We then used this model to look at how hibifolin affected the immune responses in vivo. Oral hibifolin administration (50 mg/kg) substantially decreased ear edema when compared to vehicle control (Fig. 4A). H&E staining (Hematoxylin and eosin staining) further exposed reduced epidermal thickness (Fig. 4B,C) and lower histological scores (Fig. 4D). Although its exact cellular targets in vivo are yet unknown, these results imply that hibifolin may function as a molecule with immunomodulatory qualities, perhaps providing therapeutic benefits for allergic contact dermatitis.