Proanthocyanidins Inhibit In vitro and In vivo Growth of Human Non–Small Cell Lung Cancer Cells by Inhibiting the Prostaglandin E₂ and Prostaglandin E₂ Receptors Free

Lung cancer is the leading cause of cancer-related deaths in men and women in the United States and worldwide (1). One of every three cancer-related deaths is attributable to lung cancer, and the dismal 5-year survival rate of ∼14% has shown no improvement over the past three decades (2, 3). Although a combination of chemotherapy and radiation therapy can improve survival, most patients die of disease progression, which is often associated with acquired or intrinsic resistance to chemotherapeutic drugs (4–6). Cyclooxygenase-2 (COX-2) is frequently constitutively elevated in different human malignancies, including lung cancers (7–10 11) as well as colorectal (12, 13), prostate (14), and breast (15, 16) cancers. Although multiple genetic alterations are necessary for lung cancer development, COX-2 may be a central element in orchestrating this process. Overexpression of COX-2 promotes tumor cell resistance to apoptosis (17, 18) and increases angiogenesis and tumor invasion (19, 20). Cyclooxygenases, which are the rate-limiting enzymes in prostanoid synthesis, convert arachidonic acid into prostaglandin (PG) H₂, a substrate for specific PG synthases (21). Two isoforms of COX, a constitutively expressed form, COX-1, and an inducible form, COX-2, have been identified. COX-2 inhibitors have been shown to inhibit tumor growth and metastasis in several animal models, including models of lung cancer (22). The current treatment strategies for advanced lung cancer include surgical resection, radiation, cytotoxic chemotherapy, and photodynamic therapy (5). In almost two thirds of cases, the cancer has already spread beyond localized disease at the time of diagnosis, limiting therapeutic options (23, 24). Therefore, the exploration and development of more effective chemopreventive/chemotherapeutic agents and therapies that can target the molecules associated with tumor proliferation, angiogenesis, and apoptosis resistance will lead to improved outcomes in patients with lung cancer.

Phytochemicals, particularly those that can be administered as dietary supplements, offer promising new options for the development of more effective chemopreventive and chemotherapeutic strategies. Grape seed proanthocyanidins (GSP) have anticarcinogenic properties (25) and seem to exhibit minimal toxicity (26, 27). GSPs are a mixture of polyphenols/flavanols and primarily contain proanthocyanidins (89%), which constitute dimers, trimers, tetramers, and oligomers/polymers of monomeric catechins and/or (−)-epicatechins (25, 27). GSPs are readily available as an extract of grape seeds and this extract, rather than the individual constituents, is used in determining the chemotherapeutic effects of GSPs as it represents a feasible and affordable dietary phytochemical.

Recently, we found that dietary GSPs resulted in the significant inhibition of the growth of human non–small cell lung cancer (NSCLC) cell xenografts in nude mice and that this was associated with the inhibition of lung tumor cell proliferation and the inhibition of angiogenic factors (28). In the present study, we investigated the mechanism responsible for the inhibition of lung cancer cell proliferation using in vitro and in vivo models. We report that treatment of NSCLC cells with GSPs induces apoptotic cell death, and this cell death is mediated through the inhibition of COX-2 expression and the associated inhibition of PGE₂ and PGE₂ receptors. Our results provide a convincing rationale for the pharmacologic activity of GSPs against NSCLC cells in in vitro and in vivo models.

The human NSCLC lines (A549, H1299, H226, H460, H1975, H1650, HCC827, and H157) were purchased from the American Type Culture Collection. Normal (nonmalignant) human bronchial epithelial cells (BEAS-2B) from the American Type Culture Collection were used as a control. The lung cancer cell lines were cultured as monolayers in Ham's F-12 or RPMI 1640 culture media supplemented with 10% heat-inactivated fetal bovine serum (Hyclone), 100 μg/mL penicillin, and 100 μg/mL streptomycin and maintained in an incubator with a humidified atmosphere of 95% air and 5% CO₂ at 37 °C. The GSPs were dissolved in a small amount of DMSO before addition to the complete cell culture medium [maximum concentration of DMSO, 0.1% (v/v) in media] and addition of the media to subconfluent cells (60–70% confluent). Cells treated only with DMSO served as a control. To study the effect of PGE₂, cells were treated with PGE₂ (10 μg/mL) for 10 min. For studies of the effects of inhibitors on PGE₂ treatment, the cells were pretreated with the respective inhibitors at the indicated concentrations 1 h before the addition of PGE₂.

Antibodies specific for COX-2, an enzyme immunoassay kit for PGE₂ analysis, and an EP4 agonist (PGE1 alcohol) were obtained from Cayman Chemicals. The antibodies specific for EP1, EP2, EP3, and EP4 and their secondary antibodies were obtained from Cell Signaling Technology, Inc. The PCR primers of known sequences of EP1, EP2, EP3, and EP4 were obtained from Invitrogen. The enhanced chemiluminescence detection reagents for Western blotting were purchased from Amersham Pharmacia Biotech.

The effect of GSPs on the proliferative capacity of the cells was determined using the MTT assay as previously described (28). The effect of GSPs on cell viability was assessed as the percent cell viability compared with vehicle-treated control cells, which were arbitrarily assigned 100% viability.

GSP-induced apoptosis of the human NSCLC cells was determined by flow cytometry using the Annexin V–conjugated Alexa Fluor488 (Alexa488) Apoptosis Detection kit following the instructions of the manufacturer, and as described by us (28, 29). Briefly, after overnight serum starvation, cells were treated with varying concentrations of GSPs for 48 h. The cells were then harvested, washed in PBS, and incubated with Alexa488 and propidium iodide in the dark at room temperature. The stained cells were analyzed by fluorescence-activated cell sorting using a FACSCalibur instrument (BD Biosciences) equipped with the CellQuest 3.3 software. The experiments were repeated twice.

We routinely receive GSPs for our research from Kikkoman Corporation, Japan. The GSPs preparation contains ∼89% proanthocyanidins, with dimers (6.6%), trimers (5.0%), tetramers (2.9%), and oligomers (74.8%) as described earlier (26, 27). GSPs are stable for at least 2 y when refrigerated at 4°C. The experimental AIN76A control diet containing GSPs is prepared in pellet form by TestDiet for our research using the GSPs that we provide for this purpose.

Skin or tumor samples were homogenized in 100 mmol/L phosphate buffer (pH 7.4) containing 1 mmol/L ethylenediamine tetraacetic acid and 10 μmol/L indomethacin using a polytron homogenizer (PT3100, Fisher Scientific). The supernatants were collected after centrifugation and the concentration of PGE₂ was determined in supernatants using the Cayman PGE₂ Enzyme Immunoassay kit following the manufacturer's protocol.

Following treatment of the NSCLC cells with or without GSPs, the cells were harvested, washed with cold PBS, and lysed with ice-cold lysis buffer supplemented with protease inhibitors, as detailed previously (28, 29). Lysates of tumor xenografts were prepared similarly. For immunoblotting, the proteins were resolved on 10% Tris-glycine gels and transferred onto a nitrocellulose membrane. After blocking the nonspecific binding sites, the membrane was incubated with the primary antibody at 4°C overnight. The membrane was then incubated with the appropriate peroxidase-conjugated secondary antibody and the immunoreactive bands were visualized using enhanced chemiluminescence reagents. Each membrane was stripped and reprobed with anti–β-actin antibody to verify equal protein loading.

Human-specific COX-2 small interfering RNA (siRNA) was transfected into A549 and H1299 cells using the siRNA Transfection Reagent kit (Santa Cruz Biotechnology, Inc.) according to the manufacturer's protocol. Briefly, 2 × 10⁵ cells per well were seeded in a six-well plate and allowed to grow to 70% confluency. The COX-2 siRNA mix with transfection reagents was overlaid on the cells for ∼6 h at 37°C and transferred into 2× growth medium for about 18 to 20 h. At 24 h posttransfection, fresh medium was added to the cells, and the cells incubated for an additional 48 h. Thereafter, cells were harvested and analyzed for cell death using a trypan blue exclusion assay. The knockdown of COX-2 expression in cells after transfection was confirmed using Western blot analysis.

RNA was isolated from NSCLC cell lines (A549 and H1299) treated with or without GSPs using the RNAEasy kit (Qiagen) and were reverse transcribed using the SuperScript II reverse transcriptase (Invitrogen). PCR was done using the Taq polymerase from New England Biolabs. The following primer pairs (F, forward; R, reverse primers 5′-3′; amplicon sizes are in parentheses) were used for EP receptor and β-actin amplification: EP1F GGTATCATGGTGGTGTCGTG, EP1R GGCCTCTGGTTGTGCTTAGA (324 bp); EP2F GCCACGATGCTCATGCTCTTCGCC, EP2R CTTGTGTTCTTAATGAAATCCGAC (655bp); EP3F CGTGTCGCGCAGCTACCGGCG, EP3R CGGGCCACTGGACGGTGTACT (398 bp); EP4F GGGCTGGCTGTCACCGACCTG, EP4R GGTGCGGCGCATGAACTGGCG (485 bp); β-actinF CCATCGAGCACGGCATCGTC, β-actinR TCCAGACGCAGGATGGCATG (363 bp). PCR conditions were 5 min at 95°C followed by 35 cycles (25 for β-actin): 1 min at 95°C, 1 min at 58°C, and 1 min at 72°C, as reported (30).

Details of this experiment are previously described in ref. (28). Briefly, the female athymic nude mice (6- to 7-wk-old) used in these studies were purchased from the National Cancer Institute. National Cancer Institute provided the sterilized AIN76A diet and water ad libitum. Mice in the experimental groups were given the AIN76A diet supplemented with GSPs (28). To determine the in vivo efficacy of GSPs against human NSCLC tumor xenograft growth, exponentially growing A549 and H1299 cells were mixed at a 1:1 ratio with Matrigel (Becton Dickinson), and a 100 μL suspension containing 2 × 10⁶ cells was injected s.c. in the right flank of each mouse (n = 10). At the termination of the experiment, the entire tumor mass was recovered and used for the analysis of biomarkers of interest.

The results of the cell viability, cell death, and PGE₂ analyses were expressed as the mean ± SD. The statistical significance of difference between the values of control and treatment groups was determined using either two-tailed Student's t test or ANOVA followed by post hoc Tukey's test. A P value of <0.05 was considered statistically significant.

We have reported previously that the treatment of NSCLC cells with GSPs inhibits their proliferation as determined using an MTT assay (28). As shown in Fig. 1, these results were confirmed using human NSCLC cell lines with different genetic characteristics, including p53 wild-type (A549), p53 mutated (H1299), and epidermal growth factor receptor positive (A549, H1299, H226, and H460). Treatment of these NSCLC cells with various concentrations of GSPs (0, 20, 40, 60, or 80 μg/mL) for 48 h resulted in a significant reduction in cell viability (10–68%, P < 0.05–0.001) as assessed using the MTT assay (Fig. 1A, left). It is evident from the data that different NSCLC cell lines differ in their sensitivity to GSPs and that the IC₅₀ for the GSPs is cell line dependent. We also examined the effects of GSPs on normal (nonneoplastic) human bronchial epithelial cells (BEAS-2B) under identical conditions. We did not find significant inhibition of cell proliferation of normal human bronchial epithelial cells after GSPs treatment at the concentrations of 20, 40, 60, or 80 μg/mL for 24, 48, or 72 hours (Fig. 1B). Although an inhibitory effect of GSPs on normal bronchial epithelial cells was noted at the 96-hour time point, the inhibition was significantly lower than the effect of equivalent concentrations of GSPs on the NSCLC cells at the same time point (P < 0.001).

To determine whether the GSP-induced inhibition of cell proliferation or cell viability of the human lung cancer cells was associated with the induction of apoptosis, two representative cell lines, A549 and H1299, were treated with various concentrations of GSPs, and apoptosis was assessed using the Annexin V–conjugated Alexa Fluor 488 (Alexa488) Apoptotic Detection kit, as previously described (29). As shown in Fig. 1C, apoptotic cells were counted as “late” or “early” apoptotic cells, which are shown, respectively, in the upper-right (UR) and lower-right (LR) quadrants of the histograms presented in Fig. 1C. GSP treatment of A549 cells for 48 hours resulted in a highly significant dose-dependent enhancement in the numbers of cells in the early and late stages of apoptosis (Fig. 1C, left): 0 μg/mL (vehicle control, 6.3%), 20 μg/mL (18.6%, P < 0.05), 40 μg/mL (32%, P < 0.01) and 80 μg/mL (49.3%, P < 0.001; summarized in Fig. 1D). Similar results were obtained on GSP treatment of A549 cells for 72 hours except that the percentages of apoptotic cells were slightly higher than 48 hours (data not shown). Treatment of H1299 cells with GSPs at the various concentrations of GSPs also resulted in a significant dose-dependent induction of apoptosis at 48 hours: 0 μg/mL (vehicle control, 3.7 ± 1%), 20 μg/mL (17 ± 2%, P < 0.05), 40 μg/mL (33.2 ± 2%, P < 0.01), and 80 μg/mL (51 ± 3%, P < 0.001; Fig. 1C and D, right). These results provide evidence that human NSCLC cells are sensitive to GSP-induced apoptosis. Treatment of normal human bronchial epithelial cells with GSPs for 48 hours did not result in the significant enhancement of apoptosis of these cells (data not shown). Although a slightly higher number of apoptotic cells was observed on treatment of normal bronchial epithelial cells with 80 μg/mL GSPs for 48 hours, this did not reach significance and was significantly less (P < 0.001) than the levels of apoptosis of A549 or H1299 cells induced by treatment with the same concentration of GSPs at the same time point.

We were interested in determining whether higher levels of COX-2 contribute to the proliferation of lung cancer cells. The levels of COX-2 in lung cancer cells were compared with the levels in normal human bronchial cells. For this purpose, the levels of expression of COX-2 were assessed in cell lysates of various NSCLC cell lines, A549, H1299, H460, H226, H1975, H1650 and HCC827, and BEAS-2B cells, by Western blot analysis. As shown in Fig. 2A, the levels of COX-2 expression were higher in the NSCLC cells than in the normal human bronchial epithelial cells. Among the lung cancer cell lines tested, the A549, H1299, and H460 cell lines expressed higher levels of COX-2. We also determined the levels of PGE₂ production in the same cell lines cultured using an identical protocol. Homogenates of equal numbers of cells were analyzed for PGE₂ production. As shown in Fig. 2B, the levels of PGE₂ were higher in the NSCLC cell lines than the normal human bronchial epithelial cells. As had been observed for the levels of COX-2, the concentrations of PGE₂ were higher in the A549, H1299, and H460 cell lines than the other human lung cancer cell lines tested.

To examine whether GSPs have any effect on the constitutive level of COX-2 in human lung cancer cell lines, we treated the A549, H1299, and 460 cell lines, which have higher levels of COX-2 expression than the other cell lines, with various concentrations of GSPs (0, 20, 40, 60, or 80 μg/mL) for 24 hours. The levels of COX-2 expression in the lysates of the cells was then determined using Western blot analysis. The data revealed that the treatment of lung cancer cells with GSPs resulted in a dose-dependent inhibition of COX-2 expression in all the human NSCLC cell lines tested (Fig. 2C).

As increased expression of COX-2 results in the formation of greater amounts of PG metabolites, we also determined the levels of PG metabolites in these cells with a particular emphasis on PGE₂ because PGE₂ plays a pivotal role in inflammation-associated diseases, including lung cancer. We found that the treatment of A549, H1299, and H460 cell lines with GSPs for 24 hours resulted in a dose-dependent inhibition of PGE₂ production (P < 0.05–0.001) compared with cells which were not treated with GSPs (Fig. 2D).

As we found that the treatment of the various NSCLC cell lines used in this study with GSPs resulted in the inhibition of cell growth/proliferation, and inhibition of cell proliferation is associated with a reduction in the levels of COX-2 expression and PGE₂ production, we further sought to determine the effects of indomethacin, a well-known COX inhibitor, on these NSCLC cells. This experiment was done to determine whether the inhibitory effect of GSPs on NSCLC cell growth/proliferation is mediated through its inhibitory effect on COX-2 expression. For this purpose, equal numbers of A549, H1299, H460, and H226 cells were plated in tissue culture plates and treated with various concentrations of indomethacin (0, 20, 40, or 60 μmol/L) for 48 hours. Cell growth and morphology was checked microscopically, and cell viability and cell death were determined using MTT and trypan blue exclusion assays, respectively. As shown in Fig. 3A and B, treatment of the cells with indomethacin resulted in a dose-dependent reduction in the growth of the cells and cell viability compared with nonindomethacin-treated controls (P < 0.01–0.001). The results of the trypan blue exclusion assay revealed that the percentage of dead cells was significantly increased (P < 0.05–0.001) with the increasing doses of indomethacin (Fig. 3C). As overexpression of COX-2 is primarily responsible for rapid cancer cell proliferation and growth, these data suggested that the inhibition of constitutive levels of COX including COX-2 expression in the presence of indomethacin resulted in the induction of NSCLC cell death and inhibition of NSCLC cell proliferation.

It has been shown that the overexpression of COX-2 contributes to cell survival and antiapoptotic effects in cancer cells. We therefore examined whether siRNA knockdown of COX-2 in the lung cancer cells would lead to the inhibition of the growth and induction of cell death of NSCLC cells. The transfection of A549 and H1299 cells with COX-2 siRNA resulted in marked reduction of cell growth and induction of cell death (64–68%, P < 0.001) after 48 hours of transfection compared with control siRNA–transfected A549 and H1299 cells (Fig. 3D).

Next, we determined the effect of PGE₂ on NSCLC cells and examined whether GSPs inhibit PGE₂-induced cell proliferation in human lung cancer cells. For this purpose, A549 and H1299 cells were treated with PGE₂ (10 μmol/L) with and without the treatment of GSPs for 48 hours. The cells were then harvested and cell proliferation was estimated using the MTT assay. We found that the treatment of lung cancer cells with PGE₂ for 48 hours resulted in a significant increase in proliferation as indicated by the enhanced absorbance at 540 nm compared with the cells that were not treated with PGE₂ (Fig. 4A). Pretreatment of A549 and H1299 cells with various concentrations of GSPs (20, 40, or 60 μg/mL) for 48 hours resulted in a dose-dependent inhibition of this PGE₂-induced cell proliferation (Fig. 4A). We also examined the effect of GSPs on the proliferation of NSCLC cells that were treated with PGE₂ (10 μmol/L) or indomethacin (40 μmol/L) alone or in combination. As shown in Fig. 4B, PGE₂ treatment enhanced the proliferation of both A549 and H1299 cells compared with non–PGE₂-treated cells. In contrast, indomethacin treatment significantly inhibited (P < 0.01) the proliferation of both cell lines compared with nonindomethacin-treated control cells, as well as PGE₂-stimulated cell proliferation (P < 0.05). Under similar conditions, the combined treatment of cells with GSPs + indomethacin synergistically decreased (P < 0.01–0.001) PGE₂-stimulated cellular proliferation of both A549 and H1299 cells.

It is known that PGE₂ manifests its biological activity via four known G protein–coupled receptors (i.e., EP1-EP4; ref. 31). Therefore, we determined the effect of GSPs on PGE₂ receptors. Analysis of A549 and H1299 cells treated with concentrations of GSPs (0, 10, 20, 40, or 60 μg/mL) for 48 hours by reverse transcription-PCR indicated a dose-dependent decrease in the levels of EP1 and EP4 transcripts (Fig. 5A). The inhibitory effect of GSPs on EP1 was less prominent than EP4. We did not detect any significant changes in the levels of EP2 or EP3 transcripts after GSP treatment of the cells. These results were further verified by Western blot analysis. As shown in Fig. 5B, the levels of EP1 and EP4 were reduced in a dose-dependent manner on treatment with GSPs.

To further examine the role of PGE₂ receptor (e.g., EP4) on the proliferation of NSCLC cells and the therapeutic effect of GSPs, A549 and H1299 cells were treated with an EP4 agonist (PGE1 alcohol) for 24 hours with or without the addition of GSPs. As shown in Fig. 5C, treatment of A549 and H1299 cells with the EP4 agonist resulted in the significant enhancement of cellular proliferation (P < 0.01). Treatment of cells with various concentrations of GSPs significantly inhibited (P < 0.01–0.001) EP4 agonist–induced proliferation in a dose-dependent manner. These data suggest that the stimulation of PGE₂ receptor in lung cancer cells has a role in cell proliferation, and that GSPs inhibit the NSCLC cell proliferation, at least in part, by inhibiting the levels of PGE₂ receptor.

We have shown earlier that dietary GSPs inhibit the growth of A549 and H1299 NSCLC cells grown as xenografts in athymic nude mice (28). The inhibitory effect of dietary GSPs on the growth of tumor xenografts are shown in Fig. 6A (28). To examine the effect of dietary GSPs in vivo on the levels of COX-2, PGE₂, and the receptors of PGE₂, we used tumor xenograft samples from our previous experiment (28). As the dietary GSPs at the concentration of 0.5% (w/w) supplemented with AIN76A control diet resulted in a significant inhibitory effect on tumor xenograft growth, we compared the data obtained from the analysis of lysates of A549 and H1299 tumor xenografts obtained from mice fed the control diet without GSPs (group 1) and mice fed the control diet supplemented with GSPs (0.5%; group 2). Western blot analysis revealed that the levels of COX-2 were lower in the tumor xenograft samples of A549 and H1299 cells in mice that were fed the diet supplemented with GSPs than in tumor xenografts from those mice that were not given GSPs in the diet (Fig. 6B). The levels of PGE₂ were also significantly lowered (P < 0.001) in the tumor xenograft samples of A549 (62%) and H1299 (50%) cells in mice that were fed the diet supplemented with GSPs. Similarly, the levels of the PGE₂ receptors EP1, EP3, and EP4 were lower in the tumor xenografts from mice that were fed the GSP-supplemented diet than in the tumor xenografts from control mice that were not given dietary GSPs (Fig. 6B).

COX-2 is frequently overexpressed in a variety of human malignancies and is linked to all stages of tumorigenesis. Elevated tumor COX-2 expression is associated with increased angiogenesis, tumor invasion, and suppression of host immunity and promotes tumor cell resistance to apoptosis (31). Because of its important role in tumor progression, COX-2 is a promising target for cancer therapy (31). Although multiple genetic alterations are necessary for lung cancer development, COX-2 enzymatic products may be central to orchestrating this process. Most of the protumorigenic effects of COX-2 have been attributed to its metabolic product, PGE₂, an important mediator of tumor growth. A high concentration of PGE₂ due to COX-2 overexpression in neoplastic cells shifts the balance of these tumorigenic processes, creating a permissive microenvironment for tumor growth. Therefore, the search of potential COX-2 inhibitors for the prevention or treatment of lung cancer may prove to be an important strategy.

Phytochemicals offer promising options for more effective treatment strategies for lung cancer. GSPs represent one such phytochemical agent that has been shown to have anticarcinogenic activity (25), including activity against lung cancer cells (28). As we had found that dietary GSPs do not induce apparent toxicity in experimental animals, and inhibit cell proliferation of NSCLC cells in vitro and tumor xenograft growth in vivo in athymic nude mice, we conducted further studies to understand the molecular mechanisms involved in the GSP-mediated inhibition of NSCLC cell growth in vitro and in vivo. The significant findings in the present study are that the treatment of NSCLC cells with GSPs induces apoptotic cell death, and that is associated with the inhibition of COX-2 expression and PGE₂ production. The NSCLC cells overexpress COX-2, and the inhibition of COX-2 by GSPs may have contributed to the apoptosis of these cells. This concept is supported by the evidence that the treatment of the NSCLC cells with indomethacin, a potent pan-COX inhibitor, resulted in a reduction in cell proliferation/cell viability and induction of cell death. Similar effects were also noted when lung cancer cells, A549 and H1299, were transfected with COX-2 siRNA. It has been reported that COX-2 inhibitors can induce apoptosis of NSCLC cells (13, 31); however, although certain COX-2 inhibitors primarily induce apoptosis, others may predominantly induce growth arrest (32).

It is well known that PGE₂ exerts its multiple actions through four G protein–coupled receptors, EP1, EP2, EP3, and EP4 (31) that can stimulate epithelial cell growth and invasion and promote cellular survival (17, 19). Because PGE₂ is the major PG found in lung cancer cells (31), we focused on the possible involvement of the PGE₂ receptors in GSP-induced inhibition of NSCLC cell proliferation. We observed that A549 and H1299 cells express the PGE₂ receptors, EP1 and EP4, and that the expression of EP1 and EP4 was reduced when cells were treated with GSPs in vitro. These data suggest that the inhibition of the EP1 and EP4 levels by GSPs may contribute to the inhibition of tumor cell growth and induction of apoptosis of lung cancer cells. This assumption is based on the findings that PGE₂ receptors coupled to the GαS, and ligand binding has been reported to increase cyclic AMP levels leading to the activation of PKA and Akt (33). Akt and PKA activation can mediate prosurvival pathways through the inactivation of proapoptotic proteins (34, 35). Our results are consistent with the report that PGE₂ protected gastric mucosal cells in vitro from ethanol-induced apoptosis via EP1 and EP4 activation (36). The inhibitory effect of GSPs on lung cancer cell proliferation through the inhibitory effect on EP1 or EP4 was further confirmed by treating the cells with EP4 agonist. We found that the treatment of A549 and H1299 cells with EP4 agonist (PGE1 alcohol) resulted in enhanced cell proliferation, and that EP4 agonist–induced cell proliferation was inhibited by the treatment of cells with GSPs. This observation further supports the concept that the inhibition of PGE₂ receptors by GSPs may have contributed to the inhibition of proliferation and induction of apoptosis in lung cancer cells.

To verify that GSPs can inhibit the growth of lung cancer cells in vivo through these mechanisms, lung tumor xenografts grown in athymic mice were analyzed for the expression of COX-2, PGE₂, and PGE₂ receptors. A significant finding of the present study was the inhibition of tumor xenograft growth in athymic nude mice fed a diet supplemented with GSPs (0.5%, w/w) was associated with the inhibition of COX-2 and PGE₂ expression and a reduction in the levels of PGE₂ receptors, EP1, EP3, and EP4. These data suggest that the protective effects of the dietary GSPs on the growth of NSCLC cells in vivo are also mediated through the inhibition of PGE₂ and PGE₂ receptors.

It is also important to consider whether the effect of any chemopreventive agent in an animal model can be translated in human system. For appropriate conversion of chemopreventive/chemotherapeutic agent doses from animal studies to humans, the body surface area normalization method has been recommended (37). In this study, we measured that each mouse (mean weight, 20 g) consumed ∼13.5 mg GSPs per day. Based on this information, the human equivalent dose of GSPs was calculated using the following formula:

(K_m factor for mouse = 3; K_m factor for adult human = 37).

If, the normal body weight of a person is considered to be 70 kg, then 3.8 g GSPs will be required for a person per day to produce same level of antilung carcinogenic effects as observed in mice, which seems reasonable, affordable, and attainable.

In summary, the results from this study show for the first time the chemotherapeutic efficacy of GSPs in controlling the proliferation and induction of apoptosis of human NSCLC cells in vitro and tumor xenograft growth in vivo through the inhibition of COX-2 and PGE₂ expression, and the role of PGE₂ receptors in this process. As the overexpression of COX-2 and subsequently overproduction of PGE₂ metabolite play a prominent role in lung cancer risk, the novelty of this study lies in the exploration of a new and more effective chemotherapeutic agent, and that is GSPs. More mechanism-based studies are therefore needed to develop GSPs as a pharmacologic safe agent for the prevention or treatment of lung cancer in humans.

No potential conflicts of interest were disclosed.

Grant Support: Veterans Administration Merit Review Award (S.K. Katiyar).

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.