Rhizoma Paridis Saponins Suppresses Tumor Growth in a Rat Model of N-Nitrosomethylbenzylamine-Induced Esophageal Cancer by Inhibiting Cyclooxygenases-2 Pathway

Abstract

Rhizoma Paridis Saponins (RPS), a natural compound purified from Rhizoma Paridis, has been found to inhibit cancer growth in vitro and in animal models of cancer. However, its effects on esophageal cancer remain unexplored. The purpose of this study was to investigate the effects of RPS on tumor growth in a rat model of esophageal cancer and the molecular mechanism underlying the effects. A rat model of esophageal cancer was established by subcutaneous injection of N-nitrosomethylbenzylamine (NMBA, 1 mg/kg) for 10 weeks. RPS (350 mg/kg or 100 mg/kg) was administered by oral gavage once daily for 24 weeks starting at the first NMBA injection. RPS significantly reduced the size and number of tumors in the esophagus of rats exposed to NMBA and inhibited the viability, migration, and invasion of esophageal cancer cells EC9706 and KYSE150 in a dose dependent manner (all P < 0.01). Flow cytometry revealed that RPS induced apoptosis and cell cycle G2/M arrest in the esophageal cancer cells. The expression of cyclooxygenases-2 (COX-2) and Cyclin D1 in rat esophageal tissues and the esophageal cancer cells were also significantly reduced by RPS (all P < 0.01). Consistently, RPS also significantly decreased the release of prostaglandin E2, a downstream molecule of COX-2, in a dose-dependent manner (P < 0.01). Our study suggests that RPS inhibit esophageal cancer development by promoting apoptosis and cell cycle arrest and inhibiting the COX-2 pathway. RPS might be a promising therapeutic agent for esophageal cancer.

Fig 1. The schematic diagram of the experimental design for NMBA injection and RPS administration.

A. A schematic representation of experimental protocol. The empty arrows indicate subcutaneous injection of saline. The solid arrows indicate subcutaneous injection of NMBA. The shaded area represents oral RPS administration. B. The chemical formula of RPS. The letter “R” indicates that different functional groups can be at that position, resulting in different types of RPS with various molecular structures.

Fig 2. RPS reduced the size and number of tumors on the esophagus of rats exposed to NMBA.

A. Photographs and H&E staining of esophageal tissues of rats from healthy control, NMBA, or NMBA+RPS group. Male F344 rats (n = 10 per group) were subcutaneously injected with saline containing DMSO (Healthy control group), NMBA at 1mg/kg (NMBA group), or NMBA plus oral administration of RPS at 350 mg/kg (NMBA + RPS group). Esophagus was dissected, photographed, and examined under a light microscope. Part of esophageal tissues was fixed in 10% formalin solution and stained with H&E. B. RPS significantly reduced the number of tumors on esophagus. Tumors larger than 1mm in diameter were counted, n = 10. C. RPS significantly decreased tumor size. The volume of the lesions was calculated using the standard formula: volume = length × width × height × 0.52, n = 10. D. RPS reduced the number of papilloma and carcinoma. Two pathologists, who were blinded for the treatment allocation, examined the type and number of tumors. The average number of papilloma and carcinoma is presented, n = 10. * represents significant difference between NMBA group vs healthy control group, P < 0.01. # represents significant difference between NMBA +RPS group vs NMBA group, P < 0.01.

Fig 3. RPS inhibited the viability, migration, and invasion of esophageal cancer cells.

A. RPS inhibited the viability of esophageal cancer cells. After overnight incubation, cells in 96-well plates were treated with RPS at 0, 2.5, 5, 7.5, 10, 15, 20, 30, 40, or 60 μg/mL for 48 hours and then incubated with 20 μL of MTT (5mg/mL) for 4 hours. The absorbance at 550 nm was measured in a microplate reader. The percentage of viability relative to the controls without RPS was calculated. Each concentration was tested with 6 repeats in each experiment, n = 3. B. RPS reduced the invasion of esophageal cancer cells. Cells were cultured in the transwell chamber (5 x 10⁴/chamber, 8 μm pore size and coated with 1mg/mL matrigel) containing serum-free media with RPS at 0, 5, 10, or 20 μg/mL in the upper chamber and RPMI-1640 medium + 10% FCS in the lower chambers for 48 h. The penetrated cells at the lower surface of the filter were fixed and counted under a microscope. A total of 5 fields of each chamber were randomly selected and the average cell number of the 5 fields was used, n = 3. * and # represents significant difference at 5, 10, and 20 μg/mL of RPS compared to 0 μg/mL of RPS in EC9706 cell and KYSE150 cells, respectively, P < 0.01. C. Images of wound healing of EC9706 cells, 20 x. D. Images of wound healing of KYSE150 cells, 20 x. E. RPS reduced the migration of EC9706 cells. F. RPS reduced the migration of KYSE150 cells. A total of 5 scratches were used for each concentration of RPS. Cells were cultured in serum-free media with RPS at 0, 3.25, 7.5, or 15 μg/mL for EC9706 cells, or at 0, 5, 10, or 20 μg/mL for KYSE150 for 24 h. Wounds were photographed under a phase-contrast inverted microscope, and the percentage gap closure was calculated as (width at 0h –width at 24 n)/width at 0h * 100%, n = 3. * represents significant difference between RPS at specified concentration vs 0 μg/mL, P < 0.01.

Fig 4. RPS induced apoptosis in esophageal cancer cells.

Cells were treated with RPS at 0, 5, 10, or 20 μg/mL for 24 h. Both the attached cells and the floating cells in the media were collected by centrifugation at 1000 g for 5min. Cells were than incubated with 5 μL Annexin V-FITC and 10 μL PI for 15 min in dark, diluted in 200 μL PBS, and analyzed on BD FACS Calibur flow cytometer. Apoptosis was analyzed using FlowJo software, n = 3.

Fig 5. RPS promoted cell cycle G2/M phase arrest.

Cells used for analyzing apoptosis were also analyzed for cell cycle using FlowJo software, n = 3. * and # represents significant difference at 5, 10, and 20 μg/mL of RPS compared to 0 μg/mL of RPS in EC9706 cell and KYSE150 cells, respectively, P < 0.01.

Fig 6. RPS reduced the expression of COX-2 and Cyclin D1 and the release of PGE2.

A. RPS reduced the expression of COX-2 and cyclin D1 in esophageal cancer cells. EC9706 cells were treated with RPS at 0, 7.5, or 15 μg/mL for 24h. KYSE150 cells were treated with RPS at 0, 10, or 20μg/mL for 24h. The cells were then harvest and lysed. Protein extract (30 μg) were separated by SDS-PAGE and transferred to the nitrocellulose membranes. The membranes were probed with rabbit anti rat β-actin, COX-2, or cyclin D1. Representative images were presented. B. RPS reduced the expression of COX-2 and cyclin D1 in esophageal tissues. Esophageal tissues from rats were homogenized and protein extract (30 μg) were separated by SDS-PAGE and transferred to the nitrocellulose membranes. The membrane was probed with rabbit anti rat β-actin, COX-2, or cyclin D1. Representative images are presented. C. Densitometry analysis of the western blot in B. Average values of 5 rats are presented, n = 5. * represents significant difference between NMBA group vs healthy control group, P < 0.01. # represents significant difference between NMBA +RPS group vs NMBA group, P < 0.01. D. RPS decreased the release of PGE2 from esophageal cancer cells. Cells were cultured in serum-free media containing 0, 5, 10, or 20 μg/mL RPS for 24 h. The culture supernatants were collected and centrifuged at 12, 000 rpm for 10 min to remove cell debris. The level of PGE2 in the supernatants was determined by ELISA kit according to the protocol provided by the manufacturer, n = 3. * and # represents significant difference at 5, 10, and 20 μg/mL of RPS compared to 0 μg/mL of RPS in EC9706 cell and KYSE150 cells, respectively, P < 0.01.