LPS (purified from Escherichia coli 055:B5) was purchased from Sigma-Aldrich (St. Louis, MO) and fetal bovine serum (FBS) was purchased from Invitrogen (Carlsbad, CA). Antibodies against iNOS, COX-2, IKKα/β, NF-κB p65, IκBα and phospho-IκBα (Ser-32/36) were obtained from Cell Signaling Technology (Beverly, MA). RAW 264.7 cells were cultured in a RPMI-1640 medium supplemented with 10% of FBS, 2 mM of L-glutamine, 100 U/ml of penicillin, and 100 μg/ml of streptomycin at 37˚C in a 5% CO₂ humidified incubator. We measured cell viability by counting, with a stain of 0.2% trypan blue solution.

Glyceollins I, II and III (Fig. 1A, B and C) were semi-purified from elicited soybeans, with a slight modification, as previously described (8,11). In brief, we used soybeans (AGA, variant no. 3, a gift from KNU Soyventure, Daegu, Korea) that overproduce isoflavones to the rate of 10 mg/g of soybeans. The soybeans were elicited by a Rhyzopus sp. for de novo synthesis of glyceollins, which produced three kinds of isomers (data not shown). The cultures of Rhyzopus sp. were grown at 25˚C in a culture room on potato dextrose agar media and the inoculums were harvested after 5 days. We first washed the soybeans by dipping in a 65% ethanol solution for 1 min; thereafter, we washed them with deionized water before the soaking process. Soybean seeds were soaked in sterile, deionized water for 6 h, and then placed into a Sanyo MLR-351H growth chamber (Carlsbad, CA). After saturation with distilled water, the seeds were cut into 4 pieces by a knife. A spore suspension (100 μl) of Rhyzopus sp. was dispersed on the cut surfaces of the seeds. The seeds (20 g) were incubated at 26˚C for 4 days, extracted with 50 ml of 80% (v/v) ethanol for 1 h, then cooled and centrifuged at 20,000 × g for 10 min. The extracts were filtered by a membrane filter (0.45 μm) from Sartorius (Aubagne Cedex, France). Each crude extract was freeze-dried and dissolved in dimethyl sulfoxide. The extract was 100 μg/ml, and was used for further purification. For purification, we used a high-performance liquid chromatography system, the HPLC, PerkinElmer Series 200 (Waltham, MA). A Brownlee Choice C18 (150×4.6 mm) reverse-phase column and guard column were used. The injection volume and column temperature were 10 and 30 μl, respectively. Detection was monitored at 260 and 285 nm, and the flow rate was 1.0 ml/min at the following solvent condition (A, acetonitrile, B, 1% acetate in water; 0 to 45% A for 10.2 min; 45 to 90% A for 6 min, holding at 90% A for 3.6 min). The glyceollins were separated by thin-layer chromatography (TLC) for use in further experiments because the concentration was low and the amounts that we needed were small. The glyceollin phase was separated by open-column chromatography (OCC). A methanolic extract was fractionated on a Silica gel-60 (<0.063 mm: 0.063 mm, 8:2) developed in methylene chloride:methanol (97–99:1–3, v/v) solution. A sample of OCC extract was fractionated on aluminum-backed Silica gel-60 TLC plates developed in hexane:acetone (4:1.5, v/v). Glyceollins were visualized with 20% aqueous sulfuric acid spray reagent, and UV light at 254 nm. A single glyceollin band (Rf=0.2) was confirmed by a standard glyceollin, and further purified by preparative TLC. The concentration of glyceollin isomers in the fraction was up to ~90% as analyzed by preparative HPLC. The relative ratio of glyceollin I, II and III in the fraction was 17:2:1. Thereafter, we designated the fraction as glyceollins.

Cell viability was measured by the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay (25). Cells were seeded on 96-well plates at a concentration of 5×10⁴ cells/well. After seeding, cells were incubated with MTT solution (1 mg/ml) for 2 h. The medium was discarded and the formazan product was solubilized with dimethyl sulfoxide. Viability was assessed by measuring absorbance at 570 nm with a Bio-Rad microplate reader (Hercules, CA).

RAW 264.7 cells were treated with 1 μg/ml LPS plus sample for 16 h. Nitric oxide synthesis can be determined by assaying nitrite in the culture supernatant because nitrite is a stable reaction product of NO (26). Briefly, cell-free culture media were reacted with a 1:1 mixture of Griess reagent (1% sulfanilamide, 0.1% naphthylethylenediamine dihydrochloride and 2.5% phosphoric acid) at room temperature for 10 min. The OD of the assay sample was measured spectrophotometrically at a wavelength of 570 nm. Nitrite concentration was calculated from a standard curve prepared using NaNO₂ under the same assay conditions.

Proteins were separated by SDS-PAGE and immunoblotted onto a nitrocellulose membrane in a buffer containing 20% methanol, 25 mM of Tris, and 192 mM of glycine, as described elsewhere (27). The membranes were then blocked with 5% non-fat dry milk and incubated with the primary antibody overnight. Subsequently, membranes were washed in Tween-Tris buffer saline (TTBS), incubated for 4 h with horseradish peroxidase-conjugated goat anti-mouse or goat anti-rabbit (1:4,000) antibodies, and finally developed using an enhanced ECL system (KPL Inc., Gaithersburg, MD). The membranes were then reprobed with a β-actin antibody as a control.

Total-RNA was extracted from RAW 264.7 cells using an easy-BLUE total-RNA extraction kit from Intron Biotechnology (Sungnam, Korea) according to the manufacturer's protocols and was quantified by measuring absorbance at 260 nm. RNA was reverse-transcribed using 2.5 μM oligo(dt) primers, 1 mM dNTPs, and Moloney murine leukemia virus (MMLV) reverse transcriptase (Promega, Madison, WI), and the resulting cDNAs were amplified with SuperTherm DNA polymerase SR Product (Kent, UK). β-actin primers were used to standardize the amount of RNA in each sample. PCR products were resolved on 1% agarose gels and visualized by ethidium bromide staining (28).

RAW 264.7 cells were subcultured in 6-well plates treated with the indicated dose of glyceollins for 2 h in the presence or absence of LPS (1 μg/ml) for 16 h (29). The culture media was collected for determination of prostaglandin E2 (PGE₂) concentration by an enzyme immunoassay kit from Cayman (Ann Arbor, MI).

RAW 264.7 cells were incubated with glyceollins and LPS as indicated (30). The cells were harvested in PBS containing 2% serum, washed twice with ice-cold PBS, and resuspended in 500 μl of buffer A (10 mM of HEPES, pH 7.9, 5 mM of MgCl₂, 10 mM of KCl, 1 mM of ZnCl₂, 0.2 mM of EGTA, 1 mM of Na₃VO₄, 10 mM of NaF, 0.5 mM of dithiothreitol, 0.5 mM of PMSF and protease inhibitors). After the cells had been incubated on ice for 10 min and lysed by adding 50 μl of 20% Nonidet P-40, to a final concentration of 2%, their nuclei were harvested by centrifugation. The nuclear pellets were then resuspended in 60 μl of extraction buffer (10 mM HEPES, pH 7.9, 5 mM MgCl₂, 300 mM NaCl, 0.2 mM EGTA, 25% glycerol, 1 mM Na₃VO₄, 10 mM NaF, 0.5 mM dithiothreitol, 0.5 mM PMSF and protease inhibitors), and incubated on ice for 15 min. Nuclear debris was then removed by centrifugation (13,000 rpm, 10 min), and the nuclear extracts were subjected to gel shift analysis. Protein concentrations were determined using a Bradford method (31).

RAW 264.7 cells, in 10-cm diameter dishes (10⁷ cells/dish), were pretreated with or without glyceollins for 2 h and then incubated with LPS (1 μg/ml) for 16 h. For the gel shift assay, a consensus sequence for the NF-κB DNA binding site, sc-2505 from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), was used. The mutant binding sequence for NF-κB was identical to sc-2505 except for a G→C substitution in the NF-κB-DNA binding motif (sc-2511, Santa Cruz Biotechnology, Inc.). Beforehand, the NF-κB probe was biotin-end labeled by the biotin 3′ end DNA labeling kit from Thermo Scientific Pierce (Rockford, IL). Briefly, EMSA binding reactions were performed by incubating 20 μg of nuclear extract with the annealed oligos with the LightShift EMSA kit from Thermo Scientific Pierce, according to the manufacturer's instructions. The reaction mixture was subjected to electrophoresis on a 4% native gel in a 0.5xTBE buffer. After transfer, the membrane was immediately cross-linked for 15 min on a UV transilluminator equipped with 312 nm bulbs. A chemiluminescence detection method utilizing a luminol/enhancer solution and a stable peroxide solution from Thermo Scientific Pierce was used as described by the manufacturer's manual, and the membranes were exposed to X-ray films for 2–5 min before developing (32).

Statistical differences between mean values ± SD were determined by the Dunnett's multiple range test. The significance was set at P<0.05 (33).

To examine whether glyceollins exhibit cytotoxicity in cells, we first measured whether glyceollins affect cell proliferation in RAW 264.7 cells. The result showed that, at concentrations up to 100 μM, glyceollins had no toxic effect on cell viability (Fig. 1A). Major and/or minor fractions did not have any morphological changes in microscopic observation (data not shown). Therefore, we decided to investigate whether glyceollins have potential in reducing iNOS and COX-2 expressions, which could be a landmark for the assessment of molecular inflammation.

To assess the inhibitory effect of glyceollins on LPS-induced NO production in RAW 264.7 cells, the cells were treated with LPS (1 μg/ml) for 16 h after treatment in the presence or absence of glyceollins (0.1, 1, 10 or 50 μM) for 2 h. The amount of nitrite, a stable metabolite of NO, was used as the indicator of NO production in the medium. During the 16 h of incubation, RAW 264.7 cells produced up to 6.4±0.02 μM of nitrite in the resting state. When LPS (1 μg/ml) was added, NO production was dramatically increased up to 56.2±0.01 μM (Fig. 1B). In this condition, adding glyceollins inhibited LPS-induced NO production in a concentration-dependent manner corresponding to 10.6 and 58.7% inhibition at 0.1 and 50 μM, respectively (Fig. 1C). To further investigate whether the inhibitory effect of glyceollins on NO production was associated with the inhibition of corresponding gene expression, the protein and mRNA expressions of iNOS were determined by semi-quantitative RT-PCR and western blot analysis, respectively. In unstimulated RAW 264.7 cells, the iNOS mRNA and protein expressions were almost undetectable (Fig. 1D, first bands of each set); however, LPS treatment augmented the protein and mRNA expressions of iNOS remarkably; pretreatment of the cells with different concentrations of glyceollins also dramatically reduced LPS-induced iNOS mRNA and protein expressions in a concentration-dependent fashion (Fig. 1D, compare the second and sixth bands of each set at 0 to 50 μM). The intensity was a 5- and 12.5-fold decreased compared with that of LPS-treated cells. The data suggest that glyceollins can downregulate LPS-induced iNOS expression at the transcription level.

To examine whether glyceollins inhibit PGE₂ production, the cells were pre-incubated with glyceollins for 2 h and then activated with 1 μg/ml of LPS for 16 h. As shown in Fig. 2A, unstimulated RAW 264.7 cells mildly decreased PGE₂ production when compared with treatment with glyceollins alone. In Fig. 2B, LPS induced a 5.2-fold increase in the biosynthesis of PGE₂ as compared with untreated cells, but glyceollins strongly inhibited the production of PGE₂ in a dose-dependent manner. Because the biosynthesis of PGE₂ is catalyzed by COX-1 and COX-2 enzymes, we next measured the effect of glyceollins on LPS-induced COX-2 activities. We also detected that glyceollins inhibited the COX-2 mRNA and protein expressions in a dose-dependent manner (Fig. 2C, 5- and 2.5-fold, respectively). The data suggest that glyceollins can downregulate LPS-induced COX-2 expression at the transcription level. Inhibition of COX-2 expression by glyceollins was responsible for the decrease of PGE₂ production.

Glyceollins were found to inhibit the pro-inflammatory mediators, such as NO and PGE₂, most potently. Therefore, to test whether glyceollins could effectively for regulate inflammatory and anti-inflammatory cytokines in RAW 264.7 cells, we examined the mRNA levels of TNF-α, IL-1β, IL-18 and IL-10 by RT-PCR. Though IL-1β mRNA expression was only slightly inhibited in a dose-dependent manner (Fig. 3), TNF-α and IL-18 mRNA expression was significantly reduced by pretreatment of RAW 264.7 cells with glyceollins (0.1, 1, 10 or 50 μM) (Fig. 3). In contrast, in terms of the anti-inflammatory cytokines, the ability of glyceollins to induce IL-10 increased up to 38±0.5% in the LPS-activated RAW 264.7 cells (Fig. 3), while that of β-actin was not altered.

Because the phosphorylation of IKK, and its subsequent phosphorylation of IκBα, are key signals for the activation of NF-κB, we examined the effect of glyceollins on LPS-induced phosphorylation of IKK and the degradation of IκBα protein by western blot analysis. A time-course experiment showed that IκBα in the cytoplasm was almost completely degraded within 10 min, and that it recovered at 30 min, after LPS (1 μg/ml) stimulation (Fig. 4A). Pretreatment with glyceollins prevented the induced degradation of IκBα protein at 5 and 15 min; the recovery of IκBα, which is under the control of NF-κB, was also suppressed (data not shown). IκBα phosphorylation was also examined by western blot analysis. As shown in Fig. 4A, the phosphorylated form of IκBα was hardly detectable in the resting RAW 264.7 cells; however, upon exposure to LPS (1 μg/ml) alone for 10 min, IκBα phosphorylation was initiated. At 30 min after LPS stimulation, pretreatment of glyceollins moderately inhibited LPS-mediated IκBα phosphorylation (Fig. 4A). Because IKK-α and β are the upstream kinases of IκB in the NF-κB signaling pathway and are activated via phosphorylation, we examined the effect of glyceollins on LPS-induced IKKα/β activations by western blot analysis using an a phospho-specific IKKα/β antibody. RAW 264.7 cells were pretreated with glyceollins (50 μM) for 2 h and then stimulated with LPS (1 μg/ml) for various courses of time. As shown in Fig. 4A, LPS was found to strongly induce IKKα/β phosphorylation, whereas glyceollin pretreatment significantly inhibited this phosphorylation. These findings indicate that glyceollins suppressed IKK and IκBα phosphorylation in LPS-induced RAW 264.7 cells.

To investigate whether glyceollins affect the DNA binding ability of the NF-κB complex in the RAW 264.7 cells, we performed an electrophoretic mobility shift assay (EMSA). RAW 264.7 cells were pretreated with 0.1–50 μM of glyceollins for 2 h, and then the cells were stimulated with LPS for 16 h. RAW 264.7 cells with LPS strongly induced the DNA binding activity of NF-κB. In contrast, pretreatment with glyceollins significantly suppressed the induced the DNA-binding activity of NF-κB by LPS in a dose-dependent manner (Fig. 4B). Taken together, the above findings indicate that glyceollins suppress NO production as well as expressions of iNOS, COX-2, TNF-α, IL-1β and IL-18, at least in part via an NF-κB-dependent mechanism.