IL-33-activated KU812 cells were treated with various concentrations of PHF, DP and GA, followed by measurement of cell surface expression of adhesion molecule. As shown in Figure 1, PHF, DP and GA could significantly suppress the expression of IL-33-induced intercellular adhesion molecule (ICAM)-1 on KU812 cells in a dose-dependent manner (all p < 0.05). The reduced ICAM-1 expression on IL-33 activated KU812 cells suggests that PHF, DP and GA may inhibit the endothelial transmigration of basophiles to the inflamed sites, and dampen the subsequent allergic responses [30,31,32].

Chemokine CCL2 (monocyte chemotactic protein-1/MCP-1), CCL5 (regulated and normal T cell expressed and secreted/RANTES) and CXCL8 (IL-8) are known to be associated with inflammation [33,34,35,36]. CCL2 is known to be an AD-associated chemokine that recruits dendritic cells, monocytes and memory T cells from the circulation to inflammatory sites [33]. CXCL8 is a potent chemo-attractant for immune cells, especially neutrophils, to the inflamed sites where CCL5 is chemotactic for T helper type 2 (Th2) cells, eosinophils and basophils [33,34,35,36].

Figure 2 shows that the release of CCL2, CCL5 and CXCL8 from IL-33-activated KU812 cells was significantly suppressed by the treatment with PHF, DP and GA. There was a dose-dependent suppression of CCL2 release by PHF and GA (Figure 2A). The suppressed release of these inflammation-related chemokines from IL-33-activated KU812 cells may reduce the infiltration of immune cells such as Th cells, eosinophils, basophils and macrophages to the inflamed sites, thereby lessening the subsequent inflammation and other allergic responses. Th cells including Th1 and Th2 which play important roles for the cell-mediated immunity and humoral immunity, respectively [35]. Since Th2 cells is responsible for the IgE production for the sensitization and activation of mast cells in type I hypersensitivity, the suppression of the release of Th2 chemokine CCL5 by PHF, DP and GA can subsequently reduce the infiltration of Th2 cells into the inflammatory site and hence the allergic inflammation [34,35].

The proinflammatory cytokine IL-6 is produced by T cells and macrophages at the inflammatory site. It plays an important role in the regulation of immune responses, immediate and late-phase allergic inflammation, and Th17 cell activation [37,38,39]. The combined effects of IL-6 with soluble IL-6 receptor-α can lead to the transition of acute to chronic inflammation by means of switching transmigration from neutrophils to monocytes/macrophages into the inflamed site [37,38,39]. As shown in Figure 3, the dose-dependent suppression of IL-6 release by PHF, DP and GA from IL-33-activated KU812 cells may reduce the progression from acute to chronic inflammation in allergic diseases.

p38 MAPK, NF-κB and JNK are involved in the intracellular signaling pathways contributing to the inflammatory responses in granulocytes such as basophils and eosinophils [20,27,40]. Both p38 MAPK and JNK play crucial roles in regulating cell apoptosis, cell proliferation and inflammatory cytokines release [41,42,43,44]. NF-κB, bound by inhibitory protein IκBα, is a key regulator of pro-inflammatory mediator expression and inflammatory cytokines induction in lymphocytes, granulocytes, macrophages and fibroblasts [20,40,41,42,43,44]. The level of the phosphorylation of IκBα is directly related to the release of active NF-κB and the subsequent translocation into the nucleus to activate the genes transcription for cytokines, chemokines and adhesion molecules. Figure 4 shows that the IL-33-upregulated phosphorylation of p38 MAPK, IκBα and JNK, in KU812 cells was significantly inhibited by GA (all p < 0.05). However, GA did not exhibit any effect on the phosphorylation of extracellular signal regulated kinase (ERK) (data not shown). Therefore, the reduction of the phosphorylation of p38 MAPK, IκBα and JNK by GA could be the underlying intracellular mechanisms accounting for the suppression of ICAM-1 expression and release of CCL2, CCL5, CXCL8 and IL-6. In Figure 4A, since GA at dose 10 μM could potently suppress the IL-33-induced phosphorylation of p38MAPK, GA (100 μM) did not show any significant further suppression on the IL-33-induced phosphorylation of p38MAPK (p > 0.05). However, the mean of the phosphorylation of p38MAPK of GA (100 μM) is still lower than that of GA (10 μM). Since PHF and DP are complex heterogeneous mixtures but not single and purified compound, it could be difficult to interpret their effect on the individual intracellular signaling pathways. Therefore, PHF and Danpi were not investigated for their effects on intracellular signaling molecules in the present study.

Dexamethasone, a potent synthetic member of the glucocorticoid class of steroid drugs, is usually prescribed for patients with AD for resolving the tissue inflammation. It has been reported that dexamethasone can promote eosinophilic apoptosis [45] and inhibit basophilic migration [46].

We investigated the expression of adhesion molecules, chemokines and cytokines in IL-33-activated KU812 cells treated with the combined use of dexamethasone at various concentrations (0.01, 0.1 and 1 μg/mL) and the 3 natural products (PHF, DP and GA). As shown in Figure 5, the combined use of a low concentration of dexamethasone (0.01 μg/mL) with GA (10 μg/mL) could further suppress ICAM-1, CCL5 and IL-6 expression of KU812 cells compared to the use of GA (10 μg/mL) or dexamethasone (0.01 μg/mL) alone.

However, higher concentrations of dexamethasone (0.1 and 1 μg/mL) together with GA (10 μg/mL) did not exhibit any significant suppression of ICAM-1, CCL5 and IL-6 comparing with GA or dexamethasone alone (data not shown). In addition, PHF and DP (100 μg/mL) showed similar enhanced suppressive effect on the IL-33-induced expression of ICAM-1, CCL5 and IL-6 (data not shown). The above results suggest that PHF, DP or GA could be used together with a lower dose of dexamethasone for the resolution of inflammation in patients with allergic diseases such as AD, thereby promoting the therapeutic efficacy and lessening the adverse side effect of steroidal drugs.

RPMI-1640 medium and fetal bovine serum (FBS) were purchased from Life Technology Corp. (Carlsbad, CA, USA), while FITC conjugated goat anti-mouse IgG, dexamethasone, PBS, dimethyl sulfoxide (DMSO) and GA were purchased from Sigma-Aldrich Co. (St. Louis, MO, USA). Recombinant human IL-33, FITC-conjugated mouse anti-human ICAM-1/CD54, anti-human L-selectin/CD62 monoclonal antibody were purchased from R&D Systems Inc. (Minneapolis, MN, USA). Purified mouse IgG1κ isotype control, mouse anti-human IκBa (pS32/pS36), mouse anti-JNK (pT183/pY185), mouse anti-p38MAPK (pT180/pY182) were purchased from BD Pharmingen Corp., (San Diego, CA, USA).

PHF and DP were supplied by the Institute of Chinese Medicine, The Chinese University of Hong Kong, from which water extract was prepared by the Hong Kong Biotechnology Institute, Hong Kong. For PHF extract, Lonicerae Flos (20 g), Menthae Herba (10 g), Moutan Cortex (20 g), Atractylodis Rhizoma (20 g) and Phellodendri Cortex (20 g) were refluxed in distilled water (900 mL) of at 100 °C for 2 h and the whole process was repeated twice. The three batches of water extracts were mixed together and centrifuged to remove the herbal debris. Finally, the combined extract was vacuum dried to form herbal powder which was stored in desiccators until use. These processes fulfilled the Good Manufacturing Practice according to the Australian Therapeutic Goods Administration standard. GA was purchased from Sigma-Aldrich. The dosages of DP used in this study were 100–1,000 µg/mL. Based on our previous high-performance liquid chromatography (HPLC) analysis, the amount of GA present in DP was 3.5% w/w. Hence, the dosages of GA used for the experiments were in the range of 10–100 µg/mL. The PHF, Danpi and GA used in this in vitro study were weighed and dissolved in pyrogen-free phosphate buffered saline (PBS) at 40 °C for 4 h, centrifuged at 2,000 rpm for 5 min. The supernatants were filtered through 0.22 μm polyethersulphone filter. If there was any herbal debris, it was dried at 57 °C overnight for weighing. The solutions were stored at 4 °C. The concentration of the solution = (weighted mass before dissolving in PBS − mass of herbal debris) / volume of PBS used.

HPLC analyses were performed using Hewlett Packard Agilent 1100 series HPLC System, equipped with G1329A ALS Autosampler and G1315A Diode Array Detector (Agilent Technologies, Santa Clara, CA, USA). Calibration curves for GA was prepared using standard solutions containing 0, 5, 10, 15, and 20 μg/mL GA. Danpi aqueous extract was prepared with double distilled water (800 μg/mL). Sample solution was injected onto an Ultrasphere ODS column (250 × 4.6 mm i.d., particle size 5 μm, Beckman Instrument Inc., Fullerton, CA, USA). All solvents were pre-filtered with 0.45 μm Millipore filter disk (Millipore Corp., Billerica, MA, USA) and de-gassed. A gradient elution was carried out using the following solvent systems: mobile phase A—acetonitrile; mobile phase B—double distilled water/phosphoric acid (99.0/1.0; v/v). The analyses were performed for 55 min. The flow rate used was 1.0 mL/min and detection was performed at 274 nm. Each sample (10 μL) was injected into the column after filtration through a 0.45 μm filter disk. The system was monitored by a computer equipped with the 32 Karat Software (Beckman) for data collection, integration and analysis of GA quantity.

The human basophilic cell line KU812 cells (American Type Culture Collection, Manassas, VA, USA) were maintained in RPMI-1640 medium supplemented with 10% (v/v) FBS (RPMI-1640 complete medium). Cells were incubated at 37 °C in atmospheric air enriched with 5% (v/v) CO₂. At 70–80% cell confluence, cells dispersed in RPMI-1640 complete medium to a final cell concentration of 10 × 10⁵ cells/ml for further treatment.

KU812 cells were seeded in 24-well plate at a density of 5 × 10⁵ cells in 500 μL RPMI-1640 complete medium per well. They were then pre-activated with IL-33 (50 ng) for 1 h followed by treatment with various concentrations of PHF, DP or GA in the presence of IL-33 for further 18 h. Cells were then used for the detection of expression of adhesion molecules, while the supernatants collected was subjected to BD cytometric beads array (CBA) assays for allergic inflammation-related chemokines and cytokine (see below) using flow cytometry.

For the analysis of adhesion molecules, cells were washed and suspended in cold PBS followed by blocking with 2% human pooled serum for 20 min at 4 °C. They were incubated with FITC-conjugated mouse anti-human ICAM-1 or mouse IgG1 isotype (R&D Systems) for 40 min at 4 °C in the dark. After washing, the cells were re-suspended in PBS for flow cytometric analysis.

For the analysis of chemokines and cytokines, supernatants obtained were subjected for the bead-based multiplex CBA immunoassay with Human Inflammatory Kit (CXCL8, IL-1β, IL-6, IL-10, TNF-α and IL-12p70) and Human Chemokine Kit (CXCL8, CCL5, CXCL9, CCL2 and CXCL10). The capture beads contained monoclonal antibodies against individual cytokine and chemokine. Supernatant (50 μL) was incubated with different capture bead mixtures (50 μL) and phycoerythrin-conjugated detection antibodies (50 μL) for three hours at room temperature with constant shaking. After incubation, the beads were washed and re-suspended in wash buffer. The samples were subsequently analyzed by BD FACSCalibur flow cytometer (BD Biosciences, San Jose, CA, USA) with BD CBA analysis software.

For the analysis of intracellular expression of phosphorylated signaling molecules, 5 × 10⁵ cells seeded in 24-well culture plate were pre-activated with IL-33 (50 ng) for 30 min or 1 h followed by 15-min treatment with GA. After cells were fixed with 4% paraformaldehyde for 10 min at 37 °C and centrifugation, cells were permeabilized in ice-cold methanol for 30 min and then stained with FITC-conjugated mouse anti-human phosphorylated ERK1/2, phosphorylated JNK, phosphorylated p38 MAPK, phosphorylated IκBα or mouse IgG1 antibodies (BD Pharmingen) for 30 min at 4 °C in the dark. Cells were then washed, resuspended and subjected to analysis.

Expressions of surface adhesion molecule, intracellular phosphorylated signaling molecules and chemokines and cytokines were analyzed by flow cytometry (FACSCalibur) as mean fluorescence intensity (MFI), which includes both the changes of target molecule expression in individual cells and the percentage of cells expressing the target molecules.

All data are expressed as the means ± SEM. Differences between groups were assessed by Students’s t-test. A probability p < 0.05 was considered significantly different. When ANOVA indicated a significant difference, the Bonferroni post hoc test was then used to assess the difference between groups. All analyses were performed using the Statistical Package for the Social Sciences statistical software for Windows, version 16.0 (SPSS Inc., Chicago, IL, USA).