The methanol extract of ELL was obtained from the National Institute of Horticultural and Herbal Science. Briefly, the aerial portion of ELL was extracted in 99.99% methyl alcohol at 50℃ using an accelerated solvent extractor (ASE). After filtration, the extract was concentrated using a rotary evaporator (JP-SD1000; Eyela, Tokyo Rikikatai, Japan). The final extract was dissolved in dimethyl sulfoxide (Sigma Aldrich, St. Louis, MO, USA) and then diluted in phosphate buffered saline (PBS). Antibodies against ERK, phospho-ERK, phospho-p38, p38, inhibitor of kappa B (IκB), β-actin, and c-Fos were purchased from Cell Signaling Technology (Beverly, MA, USA). The antibody against NFATc1 was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). All other reagents were obtained from Sigma-Aldrich (St. Louis, MO, USA).

Primary calvarial osteoblasts were obtained from the calvariae of neonatal ICR mice (Samtako Inc., Osan, Korea) as previously described.[15] Bone marrow cells were obtained from the long bones of 4- to 6-week-old ICR male mice. To examine osteoclast formation, mouse bone marrow cells (1×10⁵ cells) were co-cultured with calvarial osteoblasts (5×10³ cells) with 1,25-dihydroxy-vitamin D₃ (1,25-[OH]₂D₃, 10 nM) in the presence or absence of the methanol extract in 96-well culture plates (Corning Life Sciences, Acton, MA, USA). After six days of culture, the cells were fixed, then permeabilized with a mixture of acetone and ethanol (1:1 volume ratio) for 30 sec, and treated with TRAP staining solution (0.01% naphthol AS-MX phosphate [Sigma-Aldrich] and 0.06% Fast Red Violet LB Salt [Sigma-Aldrich] in 50 mM sodium tartrate dehydrate and 45 mM sodium acetate, at pH 5.0). TRAP-positive (TRAP⁺) multinucleated cells (>5 nuclei/ cell) were considered mature osteoclasts.

Bone marrow cells were obtained from the long bones of 8- to 10-week-old ICR mice (Samtako Inc.). Bone marrow cells were cultured in the presence of M-CSF (30 ng/mL; PeproTech Inc., Rocky Hill, NJ, USA) for three days to generate the bone marrow-derived macrophages (BMMs). To examine osteoclast formation, BMMs were treated with the methanol extract of ELL in the presence of M-CSF (30 ng/mL) and RANKL (100 ng/mL; PeproTech Inc.) in 96-well culture plates (Corning Life Sciences). After four days, the cells were fixed and stained for TRAP.

Cell cytotoxicity was determined by the microtitration (MTT) assay. BMMs (1×10⁴ cells/well) were placed in a 96-well plate and cultured with M-CSF (30 ng/mL, R & D) and the methanol extract of ELL in alpha- minimum essential medium (α-MEM) for 48 hr. The MTT solution was added and incubated in the dark. After 5 hr, solubilization buffer (10% sodium dodecyl sulfate [SDS] in 0.01 M hydrochloric acid [HCl]) was added and the cells were incubated overnight. The proliferation of the BMMs was determined by measuring the optical density (OD) of the wells at 570 nm.

Total RNA was purified with Easy-Blue (iNtRON Biotechnology Inc., Seongnam, Korea). The cDNA was synthesized from 5 µg of RNA by using the Revert Aid™ first-strand cDNA synthesis kit (iNtRON Biotechnology Inc.) and amplified using real-time PCR or reverse transcription (RT)-PCR. The following primers of osteoclastogenic genes were used in this study: RANKL: 5′-CCA AGA TCT CTA ACA TGA CG-3′ (forward), 5′-CAC CAT CAG CTG AAG ATA GT-3′ (reverse); RANK: 5′-CGA GGA AGA TTC CCA CAG AG-3′ (forward), 5′-CAG TGA AGT CAC AGC CCT CA-3′ (reverse); osteoprotegerin (OPG): 5′-ACG GAC AGC TGG CAC ACC AG-3′ (forward), 5′-CTC ACA CAC TCG GTT GTG GG-3′ (reverse); calcitonin receptor (CTR): 5′-TTT CAA GAA CCT TAG CTG CCA GAG-3′ (forward), 5′-CAA GGC ACG GAC AAT GTT GAG AAG-3′ (reverse); CTK: 5′-CTT CCA ATA CGT GCA GCA GA-3′ (forward), 5′-ACG CAC CAA TAT CTT GCA CC-3′ (reverse); GAPDH: 5′-AAC GGA TTT GGT CGT ATT GGG-3′(forward), 5′-CAG GGG TGC TAA GCA GTT GG-3′ (reverse); β-actin, 5′-TTT GAT GTC ACG CAC GAT TTC C-3′ (forward), 5′-TGT GAT GGT GGG AAT GGG TCA G-3′ (reverse). Real-time PCR reactions for RANK, RANKL and OPG were performed in a total volume of 20 µL using SYBR® Green CTR Master Mix (Applied Biosystems, Foster City, CA, USA) according to the manufacturer's instructions. Thermocycling was performed using a 7500 Real-time PCR System (Applied Biosystems) with the following conditions: initial hold, 95℃ for 10 min; followed by 40 cycles of denaturation at 95℃ for 15 sec, annealing at 58℃, and extension at 60℃ for 1 min. An index mRNA level was assessed using a threshold cycle value and normalized against glyceraldehyde 3-phosphate dehydrogenase (GAPDH) expression. The RT-PCR conditions for CTR, CTK, and β-actin were as follows: an initial denaturation step at 94℃ for 3 min, followed by 28 cycles (CTR) or 22 cycles (CTK, β-actin) of denaturation at 94℃ for 30 sec, annealing at 58℃ for 45 sec, and extension at 72℃ for 60 sec; and a final extension at 72℃ for 10 min. After amplification, the RT-PCR reaction mixtures were electrophoresed on 1% agarose gel and visualized by ethidium bromide staining and ultraviolet (UV) irradiation.

Total cell lysates were separated by SDS-polyacrylamide gel electrophoresis (PAGE) and transferred onto Immobilon-P membranes (Millipore, Bedford, MA, USA). The membranes were blocked with 5% non-fat-milk in PBS-Tween (PBS-T), immunostained with anti-phospho ERK (1:1,000), anti-phospho p38 (1:1,000), anti-ERK (1:1,000), anti-p38 (1:1,000), anti-IκB (1:1,000), anti-NFATc1 (1:200), anti-c-Fos (1:1,000), and anti-β-actin (1:4,000), and incubated with a horseradish peroxidase-conjugated secondary antibody (1:5,000). The membranes were developed using an advanced chemiluminescence detection kit (Amersham Pharmacia Biotech, Buckinghamshire, UK).

BMMs were differentiated on dentine slices with M-CSF (30 ng/mL) and RANKL (100 ng/mL) for four days, and then treated with the methanol extract of ELL for two days. The surfaces of the dentine slices were wiped to remove the cells, and the slices were then stained with toluidine blue (1 µg/mL) (J.T. Baker, Phillipsburg, NJ, USA). The number of pits formed by bone resorption on the dentine slices was counted.

Data were presented as the mean±standard deviation (SD) from at least three independent experiments. The Student's t-test was used for the comparison of means. A value of ^*P<0.05 was considered statistically significant.

First, we examined the effect of ELL on RANKL-induced osteoclast differentiation using mouse BMMs. BMMs were cultured with RANKL and M-CSF in the presence or absence of the methanol extract of ELL for four days. BMMs differentiated into mature TRAP-positive (TRAP⁺) multinucleated osteoclasts (MNCs) during the culture period, while treatment with ELL inhibited osteoclast differentiation in a dose-dependent manner (Fig. 1A). The MTT assay showed the anti-osteoclastogenic effect of ELL was not attributable to cellular toxicity (Fig. 1B). Osteoclast differentiation is associated with the upregulation of specific genes in response to RANKL.[8] We further assessed the inhibitory effect of ELL on osteoclastogenesis by evaluating the RANKL-induced mRNA expression levels of osteoclast-related genes such as CTR and CTK. In the RT-PCR analysis, the expression levels of those genes were dramatically increased in BMMs challenged by RANKL and M-CSF for four days. However, treatment with ELL significantly inhibited RANKL-induced osteoclast-related gene expression (Fig. 1C). We further demonstrated that the mRNA expression level of RANK was significantly decreased in BMMs by the treatment of ELL (Fig. 1D). Bone resorption activity occurs after the terminal maturation and activation of osteoclasts. To examine whether the effect of ELL on osteoclast formation could be reflected in osteoclastic activity, we performed an in vitro resorption pit assay using a dentine slice. Many resorption pits were generated with RANKL-treated cells (Fig. 1E). In contrast, the treatment of ELL strongly inhibited the formation of resorption pits by the RANKL-treated cells. Therefore, these results suggested that the methanol extract of ELL exerted inhibitory effects on osteoclast formation which led to reduced bone resorption.

It is known that osteoblasts express RANKL in response to stimulation by bone-resorbing factors such as 1α,25-(OH)₂D₃. [1] Thus in co-cultures of mouse primary osteoblasts and bone marrow cells, osteoclasts are formed in the presence of 1α,25-(OH)₂D₃.[15] To investigate whether ELL effects the ability of osteoblasts to support osteoclast differentiation, we added ELL to co-cultures of mouse primary osteoblasts and bone marrow cells. ELL inhibited the differentiation of bone marrow cells into osteoclasts induced by 1α, 25-(OH)₂D₃ at 50 µM (Fig. 2A). To clarify whether the expression level of RANKL was involved in the inhibitory effect of ELL on osteoclast formation, we examined the effects of ELL on the mRNA expression level of RANKL. Consistent with a previous report,[16] 1α,25-(OH)₂D₃ induced the mRNA expression of RANKL in primary osteoblasts, but ELL showed little effect on it (Fig. 2B). Osteoblasts are also known to produce a soluble decoy receptor for RANKL, called OPG, which inhibits osteoclast formation by interrupting the interaction between RANKL and RANK.[17] When we further investigated the effect of ELL on the mRNA expression of OPG, ELL had no effect on the level of OPG mRNA suppressed by 1α,25-(OH)₂D₃ (Fig. 2C). These results suggest that ELL inhibits osteoclast formation by direct action on osteoclast precursors, and not via indirect action on osteoblasts.

As it has been suggested that osteoclast precursors may respond to ELL during osteoclastogenesis, we examined the molecular mechanism of the inhibitory effect of ELL using BMMs. We first investigated the effects of ELL on NFATc1 and c-Fos expression levels. It has been reported that NFATc1 is upregulated by RANKL and is important for osteoclast differentiation.[13] Furthermore, c-Fos is known to bind to and regulate the expression of NFATc1.[8,10] As reported previously, RANKL stimulation increased the expression of c-Fos and NFATc1 in BMMs. ELL abolished the RANKL-induced protein expression of c-Fos and NFATc1 (Fig. 3A, B).

RANKL stimulates several signaling pathways that involve MAPKs (ERK, JNK, and p38), and NF-κB.[8,18,19] When we examined the effects of ELL on the early signaling pathways induced by RANKL in BMMs, we found that ELL pretreatment suppressed the phosphorylation by RANKL of p38, but not ERK (Fig. 3C, D). We also examined effects on the NF-κB signaling pathway. Upon RANKL stimulation, NF-κB was activated through IκB-kinase (IKK) activation and the subsequent phosphorylation and degradation of IκB.[20] RANKL stimulation led to the degradation of IκB within 15 min, although ELL failed to interfere with this process (Fig. 3E). Altogether, the results suggested that ELL downregulated RANKL-induced expression of NFATc1 and c-Fos via the p38 MAPK pathway.