Curcumin was obtained from C. longa as described previously (24). Curcumin (1.2 g, 3.24 mmol) was dissolved in tetrahydrofuran (THF; 30 ml) and NaBH₄ (50 mg, 1.32 mmol) was added. The reaction mixture was stirred at room temperature for 4 h. Water (200 ml) was then added and the mixture was extracted with EtOAc. The organic phase was washed with water, dried over anhydrous Na₂SO₄ and the solvent was evaporated until dry. The obtained crude product was dissolved in THF (30 ml) and p-toluenesulfonic acid monohydrate (50 mg) was then added. The reaction mixture was refluxed for 3 h, water was added and the mixture was extracted with EtOAc (2×100 ml). The combined organic phase was washed with water and dried over anhydrous Na₂SO₄; the solvent was evaporated and the residue was chromatographed on a silica column (particle size, <0.063 mm; silica gel 60; EMD Millipore) using hexane-EtOAc (3:2) and further purified on a Sephadex LH-20 column eluted with MeOH to yield HMPH (Fig. 1A). A total of 240 mg (21% overall yield from curcumin) of HMPH was obtained (Fig. S1). The ¹H and ¹³C NMR (Figs. S2 and S3) and mass spectral data (data not shown) were consistent with the reported values (24).

The mouse macrophage cell line, RAW264.7 was purchased from American Type Culture Collection. Cells were cultured in RPMI-1640 medium (Corning, Inc.) supplemented with 10% foetal bovine serum (Biochrom, Ltd.), 1% penicillin/streptomycin and 2 mM stable L-glutamine (Gibco; Thermo Fisher Scientific, Inc.) in an atmosphere containing 95% humidity and 5% CO₂ at 37°C. In all experiments, the cell line was cultured for 24 h and cell passages 21–30 were used for experimentation.

To measure cell viability of HMPH, the MTT assay (Sigma-Aldrich; Merck KGaA) was performed. RAW264.7 cells were plated into 96-well plates at a density of 3×10⁵ cells/cm². After 24 h, cells were pretreated with HMPH or curcumin at different concentrations (0.63–20 µM) for 1 h at 37°C, then incubated with LPS (10 ng/ml) (Escherichia coli 0111:B4; Sigma-Aldrich; Merck KGaA) for 24 h. After a 3 h incubation period with MTT (0.5 mg/ml) solution, the formazan crystals were dissolved in DMSO and absorbance was measured at 560 nm.

RAW264.7 cells (3.0×10⁵ cells/cm²) were cultured in 96-well plates. To avoid the direct interaction of the test compounds with LPS and the presence of a high inflammatory background, cells were pretreated with the test compounds for 1 h prior to LPS stimulation. Briefly, cells were incubated with HMPH or curcumin (0.63–20 µM), an inhibitor of IκBα phosphorylation (BAY 11-7082; 10 µM; Sigma-Aldrich; Merck KGaA) and an anti-inflammatory drug (dexamethasone; 10 µM; Sigma-Aldrich; Merck KGaA) for 1 h prior to LPS (10 ng/ml) stimulation for 24 h. Nitrite concentration in the cultured medium was measured using Griess reagent (Sigma-Aldrich; Merck KGaA) (25). The optical density was measured at 540 nm.

RAW264.7 cells were plated at a density of 3.0×10⁵ cells/cm² in 96-well plates for 24 h. Cells were treated with HMPH and curcumin at different concentrations (5–20 µM) and BAY 11–7082 (10 µM) for 1 h followed by LPS (10 ng/ml) stimulation for 24 h. Subsequently, culture medium was analysed using ELISA kits for TNF-α (cat. no. 430904), IL-6 (cat. no. 431304) and IL-1β (cat. no. 432604), according to the manufacturer's protocols (BioLegend, Inc.).

RAW264.7 cells were pretreated with HMPH and curcumin (5–20 µM) and BAY 11-7082 (10 µM) for 1 h followed by LPS (10 ng/ml) stimulation for 15 min. Nuclear and cytoplasmic extracts were prepared using NE-PER™ Nuclear and Cytoplasmic Extraction Reagents (cat. no. 78835; Thermo Fisher Scientific, Inc.) according to the manufacturer's guidelines. The nuclear and cytoplasmic proteins were then assessed by western blot analysis.

RAW264.7 cells were cultured at a concentration of 3.0×10⁵ cells/cm² on 6-well plates. Cells were pretreated with HMPH and curcumin (5–20 µM), BAY 11-7082 (10 µM) and dexamethasone (10 µM) for 1 h, then cells were activated with LPS (10 ng/ml) for 15 min for p65 NF-κB detection or 24 h for iNOS and COX-2 detection. Cells were washed three times with ice-cold PBS and lysed with lysis buffer containing protease inhibitors (Cell Signaling Technology, Inc.). Cell lysates were centrifuged at 12,000 × g for 15 min at 4°C. The supernatant was collected and protein concentrations were determined using the detergent compatible protein assay kit (BioRad Laboratories, Inc.). Briefly, 30–50 µg protein was loaded, separated by SDS-PAGE on 7.5 and 10% gels, and then transferred to polyvinylidene fluoride membranes. The membranes were blocked with 5% non-fat dry milk in Tris-buffered saline and 0.1% Tween-20 (TBST) for 1 h at room temperature, and incubated with primary antibodies against iNOS (1:1,000; cat. no. 13120), COX-2 (1:1,000; cat. no. 4842), p65 NF-κB (1:1,000; cat. no. 8242), β-actin (1:5,000; cat. no. 4967) and lamin B1 (1:1,000; cat. no. 13435) (Cell Signaling Technology, Inc.) at 4°C overnight. Each membrane was then washed with TBST followed by incubation with anti-rabbit HRP-conjugated secondary antibody (1:2,000; cat. no. 7074; Cell Signaling Technology, Inc.) at room temperature for 1 h under agitation. Blots were visualized using the chemiluminescent HRP detection reagent (EMD Millipore). GeneSys software version 1.2.5.0 (Synoptics, Ltd.) was used to acquire iNOS, COX-2 (HMPH treatment group) and associated β-actin images, and Image Lab™ Touch Software (Bio-Rad Laboratories, Inc.) was used to acquire the images of p65 NF-κB, lamin B1, COX-2 (curcumin treatment group) and associated β-actin. Signal intensities were densitometrically semi-quantified using ImageJ V1.8.0 software program (National Institute of Health).

RAW264.7 cells (3.0×10⁵ cells/cm²) were seeded on coverslips on 24-well plates for 24 h. Cells were pretreated with HMPH (20 µM), curcumin (20 µM) or BAY 11-7082 (10 µM) for 1 h, and were then treated with LPS (10 ng/ml) for 15 or 30 min. Cells were washed three times with ice-cold PBS, fixed in pre-cooled methanol (−20°C) for 5 min and immersed in cold acetone. The fixed cells were blocked in 1% bovine serum albumin (Sigma-Aldrich; Merck KGaA) in PBS containing 0.1% Tween-20 (PBST) for 30 min and then incubated with anti-p65 NF-κB antibody (1:200; cat. no. 8242; Cell Signalling Technology, Inc.) at room temperature for 1 h. After washing in PBST, the cells were incubated with anti-rabbit secondary antibody conjugated with Alexa Fluor-488 (1:400; cat. no. 4412S; Cell Signaling Technology, Inc.) alongside propidium iodide (2.5 µg/ml) at room temperature for 1 h. The coverslips were mounted on microscope slides and images were captured using a Leica TCS SP5 II laser scanning confocal microscope (Leica Microsystems GmbH).

RAW264.7 cells (3.0×10⁵ cells/cm²) were seeded on 6-well plates and incubated at 37°C for 24 h in an incubator containing 5% CO₂. Cells were incubated in the presence or absence of various concentrations of HMPH and curcumin (5–20 µM) and BAY 11-7082 (10 µM) for 1 h, and were washed three times with serum-free medium before incubation with LPS (10 ng/ml) for 24 h. The culture supernatant was collected and the nitrite accumulated in the supernatant was determined by the Griess assay.

The crystal structure of human MD2-lipid IVa complex (PDB code 2E59) was used as the initial coordinates for the molecular docking calculations using AutoDock4.2 (26). The docking input files and analysed docking results were generated from AutoDockTools1.5.6 (26). A grid box dimension of 60×60×60 ų with a spacing of 0.375 Å between the grid points was created and centred on the ligand. The Lamarckian genetic algorithm was employed for the docking calculations to enable modification of the gene population (27). The docking parameters were as follows: A total of 250 independent runs, a population size of 150, and a maximum of 25,000,000 energy evaluations. MD2 protein was set rigid but all torsional bonds of the ligand were freely rotated. Non-polar hydrogens of the molecule were merged and each atom was assigned with Gasteiger partial charges. The lowest docked energy of each conformation in the most populated cluster was selected for the detailed analysis and further study. Visualisation of the docking results was performed using Biovia Discovery Studio Visualizer (Dassault Systèmes).

The energy minimisations of the docked MD2-HMPH and MD2-curcumin complexes were performed using the SANDER module of AMBER16 packages (28). To remove bad contacts, the systems were minimised starting with the steepest descent followed by the conjugate gradient. The minimized complexes were solvated by water molecules. Then, the systems were heated progressively from 0 to 300 K for 100 ps and equilibrated at 300 K for 200 ps in the canonical (NVT) and isothermal-isobaric (NPT) ensembles, respectively. The last snapshot obtained from the equilibrated phase was used as the suitable structure of the complex for calculating the binding free energy value using the MM/PBSA module (∆G_MM/PBSA) in the AMBER16 programme; this methodology provides more reliable binding energy values.

Data are presented as the mean ± SEM of three independent experiments. Statistical significance was evaluated by one-way or two-way ANOVA followed by Bonferroni multiple comparison test. P<0.05 was considered to indicate a statistically significant difference. The statistical analyses were performed using GraphPad Prism 6 (GraphPad Software, Inc.).

The present study investigated the effect of HMPH on the cell viability of RAW264.7 macrophages. Cells were treated with the test compounds in increasing concentrations between 0.63 and 20 µM, followed by treatment with LPS (10 ng/ml) for 24 h. HMPH did not affect the viability of RAW264.7 cells at concentrations up to 20 µM in combination with 10 ng/ml LPS; similar results were observed for curcumin (Fig. 1B).

The present study further explored the inhibitory activity of HMPH on NO in LPS-activated RAW264.7 macrophages. The inhibitory effect of HMPH on NO in LPS-activated RAW264.7 cells was assessed by Griess assay. The results revealed that LPS significantly induced NO production compared with in the control cells (without LPS treatment) (Fig. 2A). Pretreatment with HMPH, at concentrations ranging between 0.63 and 20 µM, significantly inhibited LPS-induced NO secretion in a dose-dependent manner, whereas curcumin suppressed NO production at higher concentrations of 1.25–20 µM (Fig. 2A). These findings indicated that HMPH was more effective than curcumin at inhibiting NO production. The suppression of iNOS protein expression levels by HMPH were further investigated in LPS-activated RAW264.7 cells. Upon LPS stimulation, the protein expression levels of iNOS were significantly increased (Fig. 2B and C). Pretreatment with 5–20 µM HMPH and curcumin significantly decreased the protein expression levels of iNOS in LPS-activated macrophages compared with LPS treatment alone. The anti-inflammatory drugs dexamethasone and BAY 11-7082 also significantly inhibited NO production and iNOS protein expression compared with LPS-treated cells.

As curcumin has been shown to inhibit COX-2 protein expression (20), the present study investigated the inhibition of COX-2 expression by HMPH. HMPH at 20 µM (Fig. 3A and B) significantly inhibited the protein expression levels of COX-2 in LPS-activated RAW264.7 macrophages; curcumin exhibited a similar effect at the same concentration. Compared with HMPH and curcumin, dexamethasone possessed a stronger inhibitory effect on COX-2 protein expression (Fig. 3).

The present study also investigated the suppression of cytokine production in LPS-activated macrophages using ELISA. Upon stimulation with LPS for 24 h, the production of TNF-α (Fig. 4A), IL-6 (Fig. 4B) and IL-1β (Fig. 4C) was significantly increased. Pretreatment of cells with HMPH at 10–20 µM markedly inhibited LPS-induced TNF-α, IL-6 and IL-1β production (Fig. 4A-C). Curcumin exhibited a similar trend with regards to suppression of these inflammatory cytokines. BAY 11-7082 (10 µM) also significantly inhibited the production of all inflammatory cytokines measured.

NF-κB is an important signalling protein involved in the activation of inflammation (7,8). The present study further investigated the effect of HMPH on the suppression of p65 NF-κB nuclear translocation. Upon LPS stimulation, p65 NF-κB translocated into the nucleus (Fig. 5). HMPH (10–20 µM) significantly suppressed p65 NF-κB translocation into the nucleus; however, HMPH had no effect on p65 NF-κB expression in the cytoplasm (Fig. 5A). The mechanism of action of HMPH on the suppression of p65 NF-κB translocation was greater than that of curcumin (Fig. 5B). BAY 11-7082 (10 µM) also significantly suppressed the translocation of p65 NF-κB into the nucleus. These findings are consistent with the results obtained from immunofluorescence staining of p65 NF-κB in RAW264.7 cells treated with LPS for 15 min (Fig. S4) and 30 min (Fig. S5). After treatment with HMPH, p65 NF-κB protein expression, as indicated by green fluorescence, was retained mainly in the cytoplasm of LPS-activated macrophages, similar to that observed in cells treated with BAY 11-7082 (Figs. S4 and S5). Cells pretreated with curcumin exhibited both nuclear and cytoplasmic localisation of NF-κB following LPS stimulation, indicating a less pronounced effect than that observed in response to HMPH.

To investigate whether the anti-inflammatory effect of HMPH on the LPS-induced inflammatory response was due to its effect on the cells, RAW264.7 cells were pretreated with HMPH for 1 h and were then thoroughly washed to remove HMPH before LPS stimulation for 24 h. Curcumin was used for comparison with HMPH treatment. Notably, inhibition of NO production was still observed after the removal of HMPH (5–20 µM) and curcumin (10–20 µM) prior to LPS stimulation (Fig. 6). BAY 11–7082 (10 µM) completely inhibited NO production after the washout, this finding was similar to that observed in cells treated with 20 µM HMPH. These data suggested that HMPH may act on cells and exert anti-inflammatory action prior to LPS stimulation.

Previous results demonstrated that HMPH could inhibit inflammatory responses in activated macrophages and the inhibition of LPS-induced inflammation may be through the complex formation of HMPH with MD2 (10,21). The binding of HMPH to the MD2 protein pocket was subsequently assessed and compared with the binding of curcumin. Molecular docking was applied to predict the possible interaction between HMPH and the MD2 protein using AutoDock4.2. The lowest binding energy represents the best binding conformation between the protein and ligand. Table I summarises the results of the docking study based on binding interactions and energies. The minimised structures of the best conformation of MD2-HMPH and MD2-curcumin docked complexes are shown in Fig. 7. The results revealed that HMPH can form a stronger interaction with MD2 protein than curcumin. There are two hydrogen bond interactions involving amino acids R90 and C133, as well as the hydrophobic interactions with the residues F151, I80, L87 and V135 that were observed in the MD2-HMPH complex (Fig. 7A). Conversely, curcumin interacted with MD2 by forming a hydrogen bond with R90, as well as forming hydrophobic interactions with Y131, F151, I124 and I80 residues of the MD2 protein (Fig. 7B).

The present study further calculated the binding free energy of MD2-HMPH compared with MD2-curcumin complexes using a more reliable method, MM/PBSA. As shown in Table I, HMPH (∆G_MM/PBSA, −29.56 kcal/mol) had a more energetically favourable value in the MD2 binding pocket than curcumin (∆G_MM/PBSA, −25.63 kcal/mol), which indicated strong binding between MD2 and HMPH compared with the MD2-curcumin complex. These results indicated that the effect of HMPH on the inflammatory response in macrophages may be achieved through interfering with MD2 protein.