DNA repair proteins as the targets for paroxetine to induce cytotoxicity in gastric cancer cell AGS

Abstract

To evaluate the potential anticancer effects of 1175 FDA-approved drugs, cell viability screening was performed using 25 human cancer cell lines covering 14 human cancer types. Here, we focus on the action of paroxetine, which demonstrated greater toxicity toward human gastric adenocarcinoma cell-line AGS cells compared with the other FDA-approved drugs, exhibiting an IC50 value lower than 10 μM. Evaluation of the underlying novel mechanisms revealed that paroxetine can enhance DNA damage in gastric cancer cells and involves downregulation of Rad51, HR23B and ERCC1 expression and function, as well as nucleotide shortage. Enhancement of autophagy counteracted paroxetine-induced apoptosis but did not affect paroxetine-induced DNA damage. Paroxetine also enhanced ROS generation in AGS cells, but a ROS scavenger did not improve paroxetine-mediated DNA damage, apoptosis, or autophagy, suggesting ROS might play a minor role in paroxetine-induced cell toxicity. In contrast, paroxetine did not enhance DNA damage, apoptosis, or autophagy in another insensitive gastric adenocarcinoma cell-line MKN-45 cells. Interestingly, co-administration of paroxetine with conventional anticancer agents sensitized MKN-45 cells to these agents: co-treated cells showed increased apoptosis relative to MKN-45 cells treated with the anticancer agent alone. Unequivocally, these data suggest that for the first time that paroxetine triggers cytotoxicity and DNA damage in AGS cells at least partly by reducing the gene expression of Rad51, HR23B, and ERCC1. Our findings also suggest that paroxetine is a promising candidate anticancer agent and/or chemosensitizing agent for use in combination with other anticancer drugs in cancer therapy. The molecular mechanisms underlying the anticancer activity of co-treatment with paroxetine and chemotherapy appear to be complex and are worthy of further investigation.

Keywords: DNA damage; Gastric cancer; apoptosis; chemosensitizer; drug repurposing; paroxetine.

Figure 1

The flow diagram of drug screening and validation for potential anti-cancer drugs. (A, B) 14 human cancer cell lines were treated with 10 µM FDA-approved 1175 drugs for 72 hours. The cell viability was determined by CCK-8. All experiments were performed in triplicate and in four batches. Cell viability of each cell line under paroxetine treatment is expressed as a percentage compared to 0.1% DMSO-treated control. (C) AGS and (D) MKN-45 were treated with different doses of paroxetine for 48 hours and 72 hours, respectively, and the cell viability and IC50 were determined by CCK-8. Data are representative of four to five independent experiments.

Figure 2

The induction of apoptosis by paroxetine in AGS. A-C. After 48-hour treatment, the levels of sub-G1 in AGS were determined by flow cytometry analysis. D. The apoptotic markers, cleaved PARP, active caspase 3, and Bax were measured by Western blot assays following the treatment of paroxetine. E. The Western blot analysis was performed to assay the activation of AKT, ERK, JNK, and p38 signaling in AGS and MKN-45 with different doses of paroxetine treatment for 48 hours. F. The cell division rate was determined by CMFDA staining followed by flow cytometry analysis after 24-hour treatment of paroxetine. Data are representative of four to six independent experiments.

Figure 3

Paroxetine induces DNA damage in AGS. A. Expression of DNA damage markers and DNA repair proteins in cells treated with different doses of paroxetine for 48 hours were detected by immunoblotting assays. B. AGS was exposed to 10 µM paroxetine and the protein levels of DNA repair proteins and apoptosis markers were measured at indicated time points. C. After 10 µM paroxetine treatment for 24 and 36 hours, the expression of γ-H2AX in AGS cells was detected by immunofluorescence staining, and DMSO treatment was used as a control. Scale bar = 50 μm. D. AGS was exposed to 10 µM paroxetine and the protein levels of phospho-JNK and phospho-ERK were measured at indicated time points. Data are representative of three to five independent experiments.

Figure 4

Rad51, HR23B, and ERCC1 gene expression were down-regulated by paroxetine. A. AGS was exposed to 10 µM paroxetine for 24 hours and the mRNA levels of Rad51, HR23B, and ERCC1 were measured by qRT-PCR. B. HR23B and Rad51 were overexpressed in AGS for 24 hours followed by treatment of paroxetine for 48 hours, and the levels of DNA damage markers and apoptotic proteins were determined by Western blot assays. C. pEGFP plasmids were incubated with 1 μg/ml cisplatin at 37°C for 2 hours. The cisplatin-treated plasmids were purified and transfected into AGS cells. After 12-hour transfection, cells were treated with 10 µM paroxetine for 6, 12, or 24 hours. The untreated pEGFP plasmid was as the control. D. AGS was treated with different concentrations of 4dNTP in the presence or absence of 10 µM paroxetine for 48 hours before measurement of cell viability. E. AGS was treated with 10 µM paroxetine in the presence or absence of 200 µM 4dNTP for 48 hours, and the cleaved caspase 3 (c-caspase 3) and PARP were measured by Western blot assays. Data are representative of three to five independent experiments.

Figure 5

Autophagy counteracts paroxetine-induced apoptosis in AGS. (A, B) AGS cells were treated with 10 µM paroxetine for 48 hours, and then stained with acridine orange to determine the levels of autolysosomes by flow cytometry. (C) Cells were exposed to different doses of paroxetine for 48 hours, and then the levels of autophagy markers were determined by Western blot assays. AGS cells were treated with paroxetine for 42 hours and then co-treated with chloroquine (CQ) for another 6 hours. The levels of apoptosis and autophagy markers (D), DNA damage markers and DNA repair proteins (E) and signaling (F) were determined by Western blot assays. All experiments were performed in four to seven replicates.

Figure 6

ROS generation was induced by paroxetine. (A) Cells were treated with 10 µM paroxetine for 24 hours and the levels of ROS were measured by DCFDA staining followed by flow cytometry. (B, C) AGS cells were exposed to different doses of paroxetine for 24 hours and the levels of ROS were measured. (D-G) Cells were co-treated with 10 µM paroxetine and 10 µM JC-10 reagent for 24 hours, and then subjected to flow cytometry analysis. Changes in the MMP of the cells by paroxetine were analyzed using flow cytometry (D) and quantified (E-G). JC-10 monomer (lower potential, FL530) and JC-10 aggregate (higher potential, FL590) are represented in green and red, respectively. Data are representative of three to five independent experiments.

Figure 7

ROS does not contribute to paroxetine-induced DNA damage, apoptosis, or autophagy. AGS was treated with 10 µM paroxetine for 42 hours and then co-treated with NAC for another 6 hours. The protein levels of apoptosis (A), DNA damage and repair (B), autophagy (C), and signaling molecules (D) were measured by Western blot assays. (E) Cells were pre-exposed to U0126 and SP600125 for 1 hour and then co-treated with 10 µM paroxetine for another 24 hours. The protein levels of γ-H2AX, apoptosis, and autophagy markers were then determined by Western blot assays. Data are representative of three to five independent experiments.

Figure 8

The synergistic effects of Paroxetine combination on cell viability in MKN-45 cells. A-D. MKN45 cells were treated with 10 µM paroxetine in combination with each of the listed anticancer drugs for 72 hours, and cell viability was assessed using the CCK-8. The combined effect of drug interactions was evaluated using a combination index (CI)-isobologram equation method, as described in the Materials and Methods. Data are representative of three to four independent experiments.

Figure 9

Paroxetine enhances anticancer agents-induced cytotoxicity in MKN-45. A, B. MKN-45 cells were co-treated with 10 µM paroxetine and 10 µM 5-FU, 2.5 μg/ml cisplatin, 2 nM docetaxel, and 100 nM doxorubicin for 72 hours, and the levels of sub-G1 were assayed by flow cytometry. C. MKN-45 cells were co-treated with 10 µM paroxetine and each of the listed anticancer drugs for 72 hours, and the corresponding levels of specific proteins associated with (related to) apoptosis, autophagy, and DNA damage responses were analyzed by Western blotting analysis. Data are representative of three to four independent experiments.

Figure 10

Schematic representation of paroxetine inducing DNA damage and cytotoxicity might be via shortage of nucleotides and DNA repair proteins in AGS. Our present results also suggest that paroxetine-enhanced ROS might contribute a miner effect on paroxetine-mediated DNA damage and cytotoxicity.