Recently, a novel intestinal bacterial metabolite of ginseng protopanaxadiol saponins, i.e., 20-O-(β-d-glucopyranosyl)-20(S)-protopanaxadiol (IH-901), has been reported to induce apoptosis in a variety of cancer cells. Here we show a differential effect of IH-901 on several cell types. Exposure to IH-901 for 48 hours at a supposedly subapoptotic concentration of 40 μmol/L led to both apoptotic cell death and G1 arrest in Hep3B cells, but only resulted in G1 arrest in MDA-MB-231, Hs578T, and MKN28 cells. Additionally, the treatment of MDA-MB-231, but not of Hep3B, with IH-901 up-regulated cyclooxygenase-2 (COX-2) mRNA (2 hours) and protein (6 hours), and enhanced the production of prostaglandin E2. In MDA-MB-231 cells, IH-901 induced the sustained activation of extracellular signal-regulated kinase (ERK), whereas inhibition of mitogen-activated protein/ERK kinase blocked IH-901-mediated COX-2 induction and resulted in apoptosis, suggesting the involvement of an ERK-COX-2 pathway. Combined treatment with IH-901 and nonsteroidal anti-inflammatory drugs inhibited COX-2 enzyme and induced apoptosis in MDA-MB-231 and Hs578T cells. Adenovirus-mediated COX-2 small interfering RNAs also effectively inhibited COX-2 protein expression and enhanced IH-901-mediated apoptosis without inhibiting ERK 1/2 phosphorylation, thus providing direct evidence that COX-2 is an antiapoptotic molecule. Moreover, IH-901-mediated G1 arrest resulted from an increase in p27Kip1 mRNA and protein expression followed by a decrease in CDK2 kinase activity that was concurrent with the hypophosphorylation of Rb and p130. In conclusion, IH-901 induced both G1 arrest and apoptosis, and this apoptosis could be inhibited by COX-2 induction.
Panax ginseng, C.A. Meyer, is an herbal root that has been used for more than 2000 years throughout Far Eastern countries including China, Japan, and Korea. Its beneficial effects have been analyzed by extensive preclinical and epidemiological studies (1–3) . Recently, 20-O-(β-d-glucopyranosyl)-20(S)-protopanaxadiol (IH-901), a novel ginseng saponin metabolite, formed from ginsenosides Rb1, Rb2, and Rc was isolated and purified after giving ginseng extract p.o. to humans and rats (4). IH-901 has been shown to enhance the efficacy of anticancer drugs in cancer cell lines previously resistant to several anticancer drugs (5, 6) , to exhibit antigenotoxic and anticlastogenic activity in rats concurrently treated with benzo(a)pyrene (7), and to induce apoptosis (8, 9) . These studies found that the antitumor activity of IH-901 is attributable to the induction of apoptosis.
Recently, the link between cyclooxygenase (COX)-2 and cancer has been studied (10, 11) . COX catalyzes the conversion of arachidonic acid to prostaglandins and related eicosanoids. Two isoforms of COX, which are the products of distinct genes, have been identified. COX-2 is an inducible and immediate-early gene that is up-regulated in many epithelial malignancies (12–15) . Unlike constitutively expressed COX-1, COX-2 overexpression is associated with neoplastic transformation (12), cell growth (16), angiogenesis (17), invasiveness, and metastasis (18). Although in vivo, the exact mechanism has not been clearly elucidated, in vitro studies have shown in a variety of cell types that COX-2 overexpression is antiapoptotic (19). This antiapoptotic effect is associated with inhibition of Fas (20) and up-regulation of Bcl-2 (21). Even though COX-2 is an important target for cancer therapy, COX-2 is induced by several anticancer agents such as paclitaxel (22), mitomycin C (23), and radiation (24), suggesting its association with drug resistance.
Physiologically, p27Kip1 seems to be important for blocking progression of cells from late G1 into S via its interaction with cyclin E-CDK2 complexes (25). p27Kip1 has been also implicated as a mediator of the growth arrest due to its regulation by protein kinase A activating agents (26). IH-901 stimulated p27Kip1 induction and subsequent cell cycle (G1) arrest.
In this study, we discovered that IH-901, a novel ginseng metabolite, could inhibit human cancer cell proliferation by inducing cell cycle arrest as well as apoptosis. Treatment with IH-901 resulted in cell cycle (G1) arrest by p27Kip1 induction. Additionally, we provide evidence here that COX-2 functions as an antiapoptotic molecule as its inhibition by both nonsteroidal anti-inflammatory drugs (NSAID) and adenoviral-COX-2 small interfering RNAs (siRNA) significantly enhanced IH-901-mediated apoptosis. These findings provide a molecular basis for the antineoplastic effects of IH-901 and use of this ginseng metabolite may aid in enhancing the anticancer drug efficacy by COX-2 inhibition.
Materials and Methods
Reagents. IH-901 was kindly provided by Dr. J-H. Sung (Central Research Institute, Il Hwa Co., Ltd., Korea) and was previously described (9). NS-398 was purchased from Cayman Chemicals (Ann Arbor, MI); actinomycin D from Sigma Chemical Co. (St. Louis, MO); CDK2 (M2), cyclin E (C-19), Bax and COX-2 antibodies from Santa Cruz Biotechnology (Santa Cruz, CA); p21WAF1/CIP1 (C24420) and p27Kip1 (K25020) antibodies from Transduction Laboratories (Lexington, KY); pRB (G3-245) antibodies from PharMingen (San Diego, CA); and Bcl-2 antibodies from DAKO (Denmark).
Cell Cultures. MDA-MB-231, Hep3B, and MKN28 cells were grown in RPMI 1640 supplemented with 10% heat-inactivated fetal bovine serum and gentamicin (10 μg/mL). Hs578T cell line was grown in DMEM supplemented with 10% fetal bovine serum, and gentamicin. Subconfluent monolayers of cells were utilized in all experiments.
3-(4,5-Dimethylthiazol-2-yl)-2,5-Diphenyltetrazoliumbromide Assay. Cell viability was determined by measuring the absorbance of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide (Sigma) dye for living cells, as described earlier (9). IC50's were calculated based on the survival rate of untreated cells being 100%.
Cell Cycle Analysis. Apoptosis was monitored by measuring hypodiploid DNA content (sub-G1). Cells (5 × 105) were seeded in 60 mm dishes and allowed to attach overnight. Cell cultures were then incubated with IH-901 at 40 μmol/L concentration for the indicated times. fluorescence-activated cell sorting analysis using propidium iodide (PI) was done as described previously (27). The relative DNA content per cell was obtained by measuring the fluorescence of DNA bound propidium iodide. Cell cycle analysis was conducted using a FACStar flow cytometer (Becton Dickinson, San Jose, CA) and a ModFit LT V2.0 computer program.
Chromosomal DNA Isolation and Ladder Formation Assay. Cells were harvested and then suspended in a lysis buffer [50 mmol Tris-Cl (pH 8.0), 10 mmol EDTA, and 0.5% laurylsarcosine] containing 50 μg/mL RNase A. Chromosomal DNA was isolated and ladder formation assays were done as described previously (9).
RNA Extraction, Northern Blot Analysis, and mRNA Stability Assay. Total RNA preparation and Northern blotting were done as described previously (28). Blots were hybridized with radiolabeled probes for COX-2 cDNA (kindly provided by Dr. Stephan M. Prescott, University of Utah) and a p27Kip1 cDNA fragment for RNA expression. For mRNA stability assay, the relative band intensities of p27Kip1 and β-actin mRNA at each time point were quantified by determining radioactivity using a Fuji FLA2000 and image analysis software (Image Gauge, version 3.12).
Western Blot Analysis. Cells were washed with ice-cold PBS and suspended in lysis buffer [20 mmol Tris-Cl (pH 7.4), 100 mmol NaCl, 1% NP40, 0.5% sodium deoxycholate, 5 mmol MgCl2, 0.1 mmol phenylmethylsulfonyl fluoride, 0.1 mmol pepstatin A, 0.1 mmol antipain, 0.1 mmol chymostatin, 0.2 mmol leupeptin, 10 μg/mL aprotinin, 0.5 mg/mL soybean trypsin inhibitor, and 1 mmol benzamidine] on ice for 30 minutes. Lysates were cleared by centrifugation at 13,000 rpm for 20 minutes. Equal amounts of cell extracts (100 μg) were resolved by SDS-PAGE, transferred to nitrocellulose membranes, and probed with appropriate primary and horseradish-conjugated secondary antibodies, as described previously (29). Anti-α-tubulin antibody (Sigma) was used as a loading control. Detection was done using an enhanced chemiluminescence system (Amersham Pharmacia Biotech, Buckinghamshire, United Kingdom).
Prostaglandin E2 Production and Determination of COX-2 Activity. Cells were treated with IH-901 and/or celecoxib, a specific COX-2 inhibitor, for 24 hours in complete growth medium containing 10% fetal bovine serum. Afterwards, culture medium was collected to determine the amount of prostaglandin E2 (PGE2) secreted by the cells. For the measurement of COX-2 enzymatic activity, PGE2 quantification was measured using a PGE2 enzyme immunoassay kit (Cayman) according to the manufacturer's instructions, as described previously (30). Results are expressed as picograms of PGE2 per 106 cell ± SD.
Immunoprecipitation and Kinase Assays. CDK2 and cyclin E-associated H1 histone kinase activity were determined as described previously (27). Two hundred micrograms of cell extracts were used per immunoprecipitation with CDK2 antibody coupled to protein A or G beads. After being washed, CDK2 kinase assays on histone H1 was done by incubating the immune complex beads with 30 μL of kinase reaction [0.25 μL (2.5 μg) of histone H1, 0.5 μL (5 μCi) of γ-[32P]ATP, 0.5 μL of 0.1 mmol ATP and 28.75 μL of kinase buffer] for 30 minutes at 37°C. The reaction was stopped by boiling the samples in 2× SDS sample buffer for 5 minutes. Samples were analyzed by 12% SDS-PAGE, and the gels were then dried and subjected to autoradiography. Preimmune serum was also analyzed for possible nonspecific immunoprecipitation in these studies.
Construction of Recombinant Adenoviral Vector of siRNAs for COX-2 Gene. pSUPER vector (31) enabled us to directly clone synthetic oligonucleotides yielding U1 promoter–based knock-down constructs. Target sequences of the siRNA specific for COX-2 and GFP duplex are as follows: siRNA-COX-2#1 (M90100), 5′-AAC CGA GGT GTA TGT ATG AGT GT-3′; siRNA-COX-2#2, 5′-AAT GCA ATT ATG AGT TAT GTG TT-3′; and siRNA-pEGFP (U57609), 5′-GGC TAC GTC CAG GAG CGC ACC-3′. Both siRNAs against COX-2 were controlled for sequence specificity by BLAST searching and did not show homology to other known human genes, especially COX-1. SpeI-EcoRI digested pSUPER containing U1 promoter–directed knockdown constructs were subcloned into pShuttle vector of Adeno-X Expression System (Clontech, Palo Alto, CA), thereby replacing the cytomegalovirus-promoter sequence. The sequences of the inserts and their orientations were examined by DNA sequencing analyses. siRNA expression cassettes for COX-2 or GFP from recombinant pShuttle plasmid DNA were digested and ligated to Adeno-X viral DNA according to the manufacturer's instructions. These adenoviral vectors were infected and amplified in HEK293 cells. CsCl-purified virus was dialyzed against Ca2+- and Mg2+-free DPBS containing 10% glycerol. The virus titer was determined using a standard plaque assay. Viral stock was aliquoted into appropriate volumes and stored at −70°C until use.
Adenovirus Infection In vitro. Recombinant adenoviruses expressing COX-2 or GFP siRNAs were used on the target cells at a multiplicity of infection of 20, and incubated at 37°C for 2 hours in a CO2 incubator. The medium was removed and cells were then cultured for 24 hours before further experimentation.
Data Analysis. Results are representative of at least three independent experiments done in triplicate and are presented as mean ± SD. Comparisons between groups were analyzed using Student's paired t test.
Effect of IH-901 on Cell Proliferation of Cancer Cells. We examined the effect of IH-901 on cellular proliferation using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide assay. Significant dose-dependent growth inhibition was observed in two breast cancer cell lines (MDA-MB-231 and Hs578T), as well as gastric (MKN28), and hepatoma (Hep3B) cell lines following treatment with IH-901 for 3 days ( Fig. 1 ). The concentration required for 50% inhibition of growth (IC50) by IH-901 ranged from 33.1 to 42.5 μmol/L. These results are consistent with earlier reports using myeloma cells (8, 9) . In order to examine the mechanism of growth suppression by IH-901, cancer cells were treated with 40 μmol/L IH-901 for 48 hours and then analyzed for cell cycle progression by flow cytometry. The apoptotic sub-G1 fraction increased to 31.74% in Hep3B cells after treatment with IH-901 for 48 hours, whereas those of MDA-MB-231 and MKN28 cells were 5.23% and 2.76%, respectively, indicating that IH-901 induced a significant apoptotic cell death in Hep3B but not in MDA-MB-231 and MKN28 cells ( Fig. 2A ). The increase in the sub-G1 fraction and endonucleolytic DNA cleavage in Hep3B cells by IH-901 was evident in a time-dependent manner ( Fig. 2B). As shown in Fig. 2C, IH-901-induced cell cycle arrest at the G0-G1 phase and a progressive decline of S phase cells occurred in MDA-MB-231 as well as in Hep3B cells. These results suggest that IH-901 induced apoptotic cell death concurrent with cell cycle arrest in Hep3B cells, whereas it caused G0-G1 arrest only in MDA-MB-231 and MKN28 cells. As IH-901-mediated apoptosis is reported to be associated with the activation of caspase-3 and PARP cleavage (8, 9) , levels of antiapoptotic Bcl-2 and proapoptotic Bax proteins were analyzed in Hep3B cells. Bcl-2 expression decreased in a time-dependent manner but Bax expression did not change ( Fig. 2D). Hs578T breast cancer cell lines also showed only G0-G1 arrest after treatment with 40 μmol/L IH-901 (data not shown).
COX-2 Induction by Treatment with IH-901. Because several recent reports have implicated the role of COX-2 in preventing apoptosis in various cell types (19–23) , we examined the ability of IH-901 to induce COX-2 in MDA-MB-231 cells. As shown in Fig. 3A , 40 μmol/L of IH-901 treatment resulted in the induction of COX-2 protein in MDA-MB-231 cells. COX-2 protein and mRNA were induced in a time-dependent manner by 40 μmol/L IH-901; the ethanol vehicle had no effect on COX-2 induction ( Fig. 3B). PGE2 production was also increased by IH-901 treatment in a dose-dependent manner, consistent with the induction of COX-2 protein ( Fig. 3C). IH-901 also induced COX-2 protein in MKN28 and Hs578T cells but did not lead to COX-2 induction in Hep3B cells ( Fig. 3D). Mitogen-activated protein kinase (MAPK) pathways mediate the regulation of COX-2 expression when induced by a variety of extracellular stimuli (32, 33) . To determine whether IH-901 activates MAPK pathways in MDA-MB-231 cells, the levels of phosphorylated extracellular signal-regulated kinase (ERK), p38MAPK, and c-Jun-NH2-kinase proteins were evaluated by Western analysis. Western blots of MDA-MB-231 cell extracts were probed with antibodies selective for the phosphorylated (activated) forms of ERK-1 and ERK-2. As shown in Fig. 4A , IH-901 induced a time-dependent increase in phosphorylated ERK-1 and ERK-2 with levels elevated up to 24 hours. Marginal ethanol-induced ERK activation was also observed but this was very weak compared with IH-901-induced ERK activation (data not shown). To determine whether the activation of ERK-1 and ERK-2 is required for IH-901-mediated COX-2 induction, MDA-MB-231 cells were pretreated with UO126, a selective inhibitor of MAP/ERK kinase 1/2. UO126 (10 μmol/L) strongly inhibited IH-901-induced COX-2 expression for up to 12 hours (> Fig. 4A). Although blocking ERK1/2 with 10 μmol/L UO126, the cell cycle distribution of MDA-MB-231 cells was analyzed by flow cytometry ( Fig. 4B). Cotreatment with IH-901 and UO126 caused the percentage of apoptotic sub-G1 population to increase from 5% to 18%, compared with IH-901 treatment alone, suggesting that MAP/ERK kinase inhibition can enhance IH-901-mediated apoptosis in MAD-MB-231 cells. No significant change in the phosphorylated forms of c-jun-NH2-kinase and p38MAPK was observed until 4 hours of IH-901 treatment ( Fig. 4C).
Inhibition of COX-2 Enhances IH-901-Mediated Apoptosis. Celecoxib inhibited IH-901-induced production of PGE2 in a dose-dependent way in MDA-MB-231 cells ( Fig. 3C). To address whether COX-2 protein induced by IH-901 plays an important role in protecting cells from IH-901-mediated apoptosis, we examined the effects of COX-2 inhibition on IH-901-mediated apoptosis using the selective COX-2 inhibitors, celecoxib, or NS-398 ( Fig. 5A ). Briefly, MDA-MB-231 cells were treated with celecoxib (25 or 100 μmol/L) or NS-398 (100 or 500 μmol/L) for 30 minutes before adding 40 μmol/L IH-901. Cell cycle analysis showed that although celecoxib and NS-398 alone had no significant effect on the apoptotic fraction, pretreatment with celecoxib (25 μmol/L) followed by IH-901 enhanced the apoptotic sub-G1 fraction from 5% to 40%. Similarly, NS-398 (500 μmol/L) pretreatment followed by addition of IH-901 increased the apoptotic fraction up to 30%. The combined treatment of a COX-2 inhibitor (25 μmol/L celecoxib or 200 μmol/L NS398) with 40 μmol/L IH-901 increased the apoptotic sub-G1 fraction in Hs578T cells. These results strongly suggest that COX-2 enzyme activity protects MDA-MB-231 and Hs578T cells from IH-901-induced apoptosis. We analyzed the expression of Bcl-2 and Bax in the absence or presence of NS-398 (250 μmol/L) on 40 μmol/L IH-901 in MDA-MB-231 cells. Bcl-2 expression decreased in the presence of NS398 up to 24 hours, whereas Bax expression did not change ( Fig. 5B). Therefore, the inhibition of COX-2 in IH-901-treated MDA-MB-231 cells changes the effect of IH-901 on the level of Bcl-2. To more specifically test the requirement for COX-2 induction in the antiapoptotic mechanism, we down-regulated the expression level of COX-2 protein by using the siRNA technique in MDA-MB-231 cells. The target sequences of siRNA used were specific for the COX-2 gene but not the COX-1 gene. We applied adenovirus-mediated delivery of siRNAs strategy to specifically reduce the expression of the COX-2 gene. The adenoviral vectors expressing siRNAs effectively reduced COX-2 gene expression, compared with parental and GFP-siRNA expressing cells, without interrupting ERK 1/2 phosphorylation following treatment with IH-901 ( Fig. 5C). COX-2 knockdown resulted in a dramatic increase in the apoptotic sub-G1 fraction after IH-901 treatment, compared with mock-infected and GFP knockdown ( Fig. 5D). These results show that the COX-2 signaling pathway is critical in blocking IH-901-mediated apoptosis and that COX-2 plays an important role as an antiapoptotic molecule.
Induction of p27Kip1 by IH-901 Leads to the Inhibition of CDK2 Kinase Activity. Because IH-901 arrested the cell cycle at the G0-G1 phase in MDA-MB-231 and Hep3B cells ( Fig. 2C), we investigated whether IH-901 affects the expressions of G1-specific cell cycle regulatory molecules in MDA-MB-231 cells expressing mutant p53 protein (ref. 34; Fig. 6A ). We examined the protein expression profile of cyclin-dependent kinase inhibitors associated with G1 arrest. Western blot analysis showed that IH-901 treatment of MDA-MB-231 cells strongly induced the protein levels of p27Kip1 in a time-dependent manner, whereas the level of p21WAF1/CIP1 did not change for up to 24 hours. p16INK4a was not detected in MDA-MB-231 or Hs578T cell lines, both of which are homozygously deleted for the INK4 gene (35). The p27Kip1 expression on IH-901 treatment of two cell lines, Hs578T and MKN28, showed similar pattern as in MDA-MB-231 (data not shown). Because the p27Kip1 level was up-regulated in MDA-MB-231 cells by 40 μmol/L IH-901 within 24 hours, we investigated the effect of IH-901 on the kinase activity of CDK2 associated with cyclin E. As shown in Fig. 6B, treatment of MDA-MB-231 cells with IH-901 strongly decreased the histone H1-associated kinase activities of CDK2. The protein levels of cyclin D1 and cyclin E did not change for up to 48 hours after treatment (data not shown). The complexes immunoprecipitated with anti-CDK2 antibody exhibited higher amounts of p27Kip1 from IH-901-treated MDA-MB-231 cells than from mock-treated cells. Next, we investigated the phosphorylation status of the pRb-related proteins pRb/p107 and pRb2/p130. The treatment of MDA-MB-231 cells with 40 μmol/L IH-901 increased the levels of hypophosphorylated forms of pRb and p130 in a time-dependent manner ( Fig. 6C). SNU-16 is a control cell line used as a control for the Rb phosphorylation (27). The increase of p27Kip1 protein levels by IH-901 resulted from the elevation of p27Kip1 mRNA expression ( Fig. 6D, top). We analyzed p27Kip1 mRNA degradation in the absence and presence of IH-901 by Northern analysis, in actinomycin D-treated MDA-MB-231 cells ( Fig. 6D, bottom). The kinetics of p27Kip1 mRNA degradation was similar in both cases, indicating that IH-901 could induce p27Kip1 mRNA at the transcriptional level. The levels of p27Kip1 were also increased in Hep3B following treatment with IH-901 ( Fig. 6A), as observed in MDA-MB-231 cells. These results indicate that in Hep3B cells IH-901 can simultaneously induce p27Kip1-mediated G1 arrest as well as apoptotic cell death without COX-2 induction.
Ginseng saponins exert various important pharmacologic effects with regard to the control of many diseases including cancer. The present study focuses on identifying the mechanism that underlies the antitumor activity of ginseng saponins. The study shows, for the first time, that IH-901, a novel ginseng saponin metabolite, can cause simultaneous growth arrest and apoptosis. Additionally, adenoviral infection of COX-2 siRNAs provides direct evidence that COX-2 protein can function as an antiapoptotic molecule.
IH-901 could induce the expression of COX-2 protein via ERK signaling. The suppression of IH-901-induced COX-2 expression by the MAP/ERK kinase inhibitor UO126 suggests that the ERK signaling pathway is important in the regulation of COX-2 by IH-901. Knockdown of COX-2 expression by siRNAs did not block ERK activation in MDA-MB-231 cells. IH-901 treatment led to the activation of ERK (data not shown) but failed to induce COX-2 in Hep3B cells. Therefore, IH-901-induced ERK activation is required but not sufficient for inhibition of apoptosis.
Several anticancer drugs have been shown to induce COX-2 expression (22, 23) . For example, microtubule- or actin-interfering agents stimulate MAPK signaling and activator protein activity, enhance PGE2 synthesis, and increase levels of COX-2 mRNA and protein (22). Epigallocatechin-3-gallate was found to up-regulate COX-2 expression and PGE2 production in association with the activation of both the ERK and protein tyrosine phosphatase signaling pathways (36). The suppression of apoptosis by COX-2 is thought to favor carcinogenesis by permitting the survival of cells that have acquired mutations, and is viewed as one of the main mechanisms of tumorigenesis. As COX-2 induction by anticancer drugs may result in unwanted effects such as inflammation, the combined use of selective COX-2 inhibitors and conventional anticancer drugs has been suggested (23).
However, the interpretation of previous NSAID studies was complicated by possible COX-independent mechanisms. In many cases, NSAIDs alone induced apoptosis at much higher concentrations than their IC50's. COX-2 mutant proteins lacking COX activity have been reported to inhibit cell cycle progression in a variety of cell types (37). In order to confirm that COX-2 protein is an antiapoptotic molecule, we constructed adenoviral COX-2 siRNAs for selective targeting of the COX-2 protein itself rather than opting to use NSAIDs to inhibit COX-2 enzymatic activity. Our data shows that both NSAID treatment and knock-down of the COX-2 gene by siRNAs could cause apoptosis in MDA-MB-231 cells, providing direct evidence that COX-2 protein plays an important role in suppressing IH-901-mediated apoptosis.
Thus far, the antitumor activity of IH-901 has been attributed solely to apoptosis (8, 9) . However, in this study, we show a second mechanism: cell cycle arrest. IH-901 strongly induced p27Kip1 and inhibited CDK2 activities followed by an increase in hypophosphorylated levels of pRb and p130. These molecular effects of IH-901, especially the up-regulation of p27Kip1, may play important roles in the G1 arrest seen in MDA-MB-231, MKN28, Hs578T, and Hep3B cells. The observed up-regulation of p27Kip1 by IH-901 and resulting decrease of CDK2 kinase activity is important and consistent with an earlier report that decreased p27Kip1 expression in breast carcinomas are associated with aggressive phenotypes and that adenoviral infection with human p27Kip1 resulted in cell cycle arrest (38). The pRb-related proteins, p107 and p130, cooperate to regulate cell cycle progression at G1-S checkpoint. It has been reported that p130 and p27Kip1 are mutually involved in the negative regulation of cellular proliferation (39). IH-901 reduced CDK2-cyclin E kinase activity, and increased the hypophosphorylated levels of Rb-related proteins. Hep3B is known to be Rb-negative (40) and has mutant p53 (41). Even though Hep3B cells showed marginal G1 arrest, the progressive decline of the S phase was indicative of G1 arrest ( Fig. 2C). However, it is difficult to exclude the possibility of the involvement of p21WAF1/CIP1 in IH-901-mediated growth arrest because all four cell lines used in this study express mutant p53. We found the induction of p21WAF1/CIP1 mRNA and protein in a carcinoma cell line containing wild-type p53 after the treatment with IH-901 (data not shown). Collectively, the growth inhibitory effects of IH-901 in cells could be caused by the following progression: induction of p27Kip1 leads to a decrease in kinase activity of CDKs followed by modulation of further downstream targets such as pRb-related proteins and the E2F family of transcription factors family.
In summary, the findings of this study show that IH-901, a novel ginseng saponin metabolite, induces apoptosis and inhibits cell proliferation by G1 arrest via increase of p27Kip1 expression. COX-2 induction caused by IH-901 could diminish the antitumor effects, and strongly suggests the therapeutic use of COX-2 siRNAs to inhibit COX-2. Further study will contribute to the development of IH-901 or ginseng-related drugs as potential chemotherapeutic or chemopreventive agents.
Grant support: Supported in part by the National Cancer Control R&D Program 2003, Ministry of Health and Welfare, Republic of Korea; by Seoul National University Cancer Research Institute; and by the 2002 BK21 Project for Medicine, Dentistry, and Pharmacy.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank the Il Hwa Company for providing IH-901, Drs. S.M. Prescott for cDNA construct of COX-2 and J-W. Soh for his helpful discussion, and to J.S. Koh for his critical editing of this manuscript.
- Received May 18, 2004.
- Revision received November 16, 2004.
- Accepted December 29, 2004.
- ©2005 American Association for Cancer Research.