The expression pattern of GPX3 in pancreatic adenocarcinoma (PAAD; n=178) and the normal tumor-adjacent tissues (n=4) was downloaded from UALCAN database (https://ualcan.path.uab.edu/index.html) (16). The association between GPX3 expression (cut-off value, 50%; to determine high and low GPX3 expression levels) and prognosis [overall survival (OS), disease-free survival (DFS) and relapse-free survival (RFS)] in PAAD was obtained from the GEPIA (http://gepia.cancer-pku.cn/) (17) and Kaplan-Meier Plotter (http://kmplot.com/analysis/) databases (18).

The human pancreatic ductal epithelial cell line HPDE6c7 (cat. no. BNCC359453) was obtained from BeNa Culture Collection. The PC cell lines BxPC-3 (cat. no. CL-0042), SW1990 (cat. no. CL-0448B) and PANC-1 (cat. no. CL-0184) were purchased from Wuhan Procell Life Science & Technology Co., Ltd. HPDE6c7 and BxPC-3 cells were cultured in RPMI-1640 medium, while SW1990 and PANC-1 cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS; MilliporeSigma) and 1% penicillin/streptomycin solution (Life Technologies; Thermo Fisher Scientific, Inc.). All cells were cultured at 37˚C in a humidified incubator with 5% CO₂.

Total RNA was extracted from cells utilizing a TRIzol reagent (MilliporeSigma), according to the manufacturer's instructions. After purity verification and concentration measurement, 1 µg total RNA was reverse transcribed into complementary DNA using the riboScript Reverse Transcription Kit (Guangzhou RiboBio Co., Ltd.) according to the manufacturer's instructions. Subsequently, qPCR was carried out using the SYBR Premix Ex Taq II kit (Takara Bio Inc.) on a real-time PCR instrument (Bio-Rad Laboratories, Inc.). The reaction protocol was as follows: Initial denaturation at 95˚C for 30 sec; 40 cycles at 95˚C for 5 sec and 60˚C for 20 sec, and a final extension step at 72˚C for 10 min. The primer sequences used in this study were as follows: GPX3 forward, 5'-AGCAGTATGCTGGCAAATATGTCC-3' and reverse, 5'-CAGACCGAATGGTGCAAGCTCTTC-3'; β-actin forward, 5'-AGCGAGCATCCCCCAAAGTT-3' and reverse, 5'-GGGCACGAAGGCTCATCATT-3'. The expression levels of GPX3 were calculated using the 2^-ΔΔCq method (19) and β-actin served as an endogenous control.

Protein samples were prepared following cell lysis with radioimmunoprecipitation assay (RIPA) lysis buffer (Beyotime Institute of Biotechnology) containing protease and phosphatase inhibitors (Roche Diagnostics). Following protein concentration assessment using a bicinchoninic acid protein assay kit (Pierce; Thermo Fisher Scientific, Inc.), equal amounts of protein extracts (30 µg/lane) were separated by 15% SDS-PAGE and were then transferred onto polyvinylidene fluoride membranes (MilliporeSigma). Non-specific binding was blocked following membrane incubation with 5% skimmed milk for 2 h at room temperature. Subsequently, membranes were incubated with primary antibodies, including anti-GPX3 (1:1,000; cat. no. ab275965, Abcam), anti-E-cadherin (1:1,000; cat. no. ab40772, Abcam), anti-N-cadherin (1:5,000; cat. no. ab76011, Abcam), anti-Snail (1:1,000; cat. no. ab216347, Abcam), anti-Bax (1:1,000; cat. no. ab32503, Abcam), anti-Bcl-2 (1:1,000; cat. no. ab32124, Abcam), anti-p-JNK (1:1,000; cat. no. 9251, Cell Signaling Technology, Inc.), anti-JNK (1:1,000; cat. no. 9252, Cell Signaling Technology, Inc.), anti-p-c-Jun (1:1,000; cat. no. ab32385, Abcam), anti-c-Jun (1:1,000; cat. no. ab40766, Abcam), anti-p21 (1:1,000; cat. no. ab109520, Abcam), anti-c-Myc (1:1,000; cat. no. ab32072, Abcam), and anti-GAPDH (1:2,500; cat. no. ab9485, Abcam) at 4˚C overnight followed by incubation with HRP-conjugated secondary antibody (1:2,000; cat. no. ab6721, Abcam) for 2 h at room temperature. Finally, the protein bands were visualized utilizing an Enhanced Chemiluminescence System Reagent (Nanjing KeyGen Biotech, Co., Ltd.) and quantified using ImageJ software version 1.48 (National Institutes of Health).

The coding sequences of GPX3 (accession no. NM_002084.5) were cloned into pcDNA3.1 (Shanghai Genechem Co., Ltd.) to construct GPX3-overexpressing vector (Oe-GPX3), and the empty pcDNA3.1 vector served as the negative control (Oe-NC). Upon achieving 60-70% confluency, PANC-1 cells were transfected with 5 µg/ml Oe-NC or Oe-GPX3 using Lipofectamine^® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.) at 37˚C for 48 h, according to the manufacturer's protocol. At 48 h following transfection, the cell transfection efficiency was assessed by RT-qPCR and western blot analysis. For the rescue experiment, the GPX3-overexpressing PANC-1 cells were treated with 0.01 µM anisomycin (Shanghai Yuan Ye Bio-Technology Co., Ltd.), an activator of JNK, for another 24 h at 37˚C.

Cell viability was assessed using a Cell Counting Kit-8 (CCK-8) assay. Briefly, PANC-1 cells were seeded into 96-well plates at a density of 2x10³ cells/well and cultured at 37˚C in an incubator with 5% CO₂. Following incubation for 24, 48 and 72 h, each well was supplemented with 10 µl CCK-8 reagent (Dojindo Laboratories, Inc.) and cells were incubated for an additional 3 h. The absorbance at a wavelength of 450 nm was measured in each well using a microplate reader (BioTek Instruments, Inc.).

A total of 2x10³ PANC-1 cells were inoculated into each well of 6-well plates and cultured at 37˚C in an incubator with 5% CO₂ for 2 weeks. The culture medium was changed with fresh one every 3 days. The formed colonies were fixed with 4% paraformaldehyde for 10 min at room temperature and then stained with 1% crystal violet for 30 min at room temperature. The colonies (>50 cells) were observed under a microscope and counted using ImageJ software version 1.48.

An EdU assay was carried out using the Cell-Light EdU DNA Cell Proliferation Kit (Guangzhou RiboBio Co., Ltd.). Briefly, PANC-1 cells were labeled with EdU for 2 h at 37˚C. Subsequently, cells were stained with DAPI solution (MilliporeSigma) for 5 min. The fluorescence signal was visualized under an inverted fluorescence microscope (Olympus Corporation), and then quantified using ImageJ software version 1.48.

A wound healing assay was conducted to assess the migration ability of PC cells. When PANC-1 cells reached 100% confluency in 6-well plates, a wound was created on the cell monolayer using a 200-µl sterile pipette tip. Following washing with PBS for three times to remove detached cells, cells were incubated in serum-free medium for 24 h. Images of the wound at 0 and 24 h were captured under a light microscope (Olympus Corporation). The cell migration rate was calculated depending on the shortened wound distance of each group: (wound distance at 0 h-wound distance at 24 h)/wound distance at 0 h x100.

To evaluate the invasion ability of PC cells, a Transwell assay was performed in a 24-well plate using a Transwell chamber (Corning; Corning, Inc.) pre-coated with Matrigel (BD Biosciences) at 37˚C for 30 min. Briefly, PANC-1 cells (5x10⁵ cells/ml) were resuspended in serum-free medium and were then added onto the upper chamber of the Transwell insert. The lower chamber was supplemented with complete medium containing 10% FBS. Following incubation at 37˚C for 48 h, the remaining cells on the upper surface of the membrane were removed using cotton swabs. The invasive cells were fixed in 4% paraformaldehyde and were then stained with 1% crystal violet for 30 min at room temperature. Finally, the invasive cells were observed and counted under a light microscope (Olympus Corporation). The cell invasion rate was calculated depending on the number of invasive cells of each group.

PANC-1 cells were treated with increasing concentrations of GEM (0, 1, 2, 5, 10 and 20 µM; APeXBIO Technology LLC) for 48 h. Meanwhile, GEM-resistant PANC-1 cells (PANC-1/GEM; cat. no. IMD-015; Xiamen Immocell Biotechnology Co., Ltd.) were treated with 0, 5, 10, 20, 40, 60, 80 and 100 µM GEM for 48 h. Subsequently, the cell viability of each group was assessed as aforementioned. The half maximal inhibitory concentration (IC₅₀) values were calculated using the concentration-response data in GraphPad Prism 8.0 software (GraphPad Software, Inc.).

Cell apoptosis was assessed using an Annexin V-fluorescein isothiocyanate (FITC)/propidium iodide (PI) kit (BD Biosciences). PANC-1/GEM cells were seeded into 6-well plates (5x10⁵ cells/per well). Following incubation with GEM (80 µM) or PBS at 37˚C for 48 h, PANC-1/GEM cells were washed with ice-cold PBS and resuspended in binding buffer, followed by the addition of Annexin V-FITC and PI, according to the manufacturer's protocol. Following incubation at 37˚C for 15 min in the dark, the apoptotic cells, including early apoptotic cells and late apoptotic cells, were analyzed with the FACScan flow cytometer (BD Biosciences) and CellQuest Pro software version 3.3 (BD Biosciences).

The experimental data were statistically analyzed using GraphPad Prism 8.0 software (GraphPad Software, Inc.; Dotmatics). All experiments were repeated at least three times. The experimental data were normally distributed and were expressed as the mean ± standard deviation (SD). The differences among multiple groups were compared using one-way ANOVA, followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

Bioinformatics analysis was performed to assess the effect of GPX3 on PC. As shown in Fig. 1A, the expression levels of GPX3 in tumor tissues were lower compared with those in normal ones, thus suggesting that GPX3 was downregulated in PC. In addition, the increased expression levels of GPX3 were positively associated with enhanced OS, DFS and RFS (Fig. 1B and C). Additionally, the expression levels of GPX3 were also detected in PC cell lines and the results consistently disclosed that both the mRNA and protein expression levels of GPX3 were markedly decreased in the PC cell lines BxPC-3, SW1990 and PANC-1 compared with HPDE6c7 cells (Fig. 1D). Since the lowest expression levels of GPX3 were observed in PANC-1 cells, this cell line was selected for the subsequent in vitro experiments.

To explore the regulatory role of GPX3 in PC, gain-of-function experiments were carried out. PANC-1 cells were first transfected with Oe-GPX3. The markedly increased mRNA and protein expression levels of GPX3 in the Oe-GPX3 group confirmed that the cells were successfully transfected (Fig. 2A). As shown in Fig. 2B and C, GPX3 overexpression significantly inhibited cell viability and reduced the colony formation ability of PC cells, thus supporting the anti-proliferative capacity of GPX3 in PANC-1 cells. The above finding was further verified by the reduced number of EdU-positive cells in the Oe-GPX3 group (Fig. 2D). Furthermore, wound healing and Transwell assays revealed that GPX3 overexpression not only inhibited the healing velocity of PANC-1 cells within 48 h, but also reduced their invasion ability (Fig. 2E), thus suggesting that GPX3 could inhibit the migration and invasion of PANC-1 cells. The results also demonstrated that GPX3 notably increased the protein expression levels of E-cadherin and decreased those of N-cadherin and Snail (Fig. 2F), thus indicating that GPX3 attenuated the epithelial-mesenchymal transition (EMT) of PANC-1 cells.

In addition to the enhanced proliferation and invasion abilities of cancer cells, chemo-resistance is also significantly associated with poor prognosis in PC (20). Therefore, the present study also aimed to investigate the effect of GPX3 on the chemo-resistance of PANC-1 cells to GEM. The results revealed that PANC-1/GEM cells possessed a higher GEM IC₅₀ value compared with PANC-1 cells, while GPX3 overexpression markedly reduced the IC₅₀ value of GEM in PANC-1 and PANC-1/GEM cells (Fig. 3A). The flow cytometric analysis showed that cell treatment with GEM promoted the apoptosis of PANC-1/GEM cells, which was further enhanced by GPX3 overexpression (Fig. 3B). Consistently, the western blot analysis results showed that Bax was upregulated and Bcl-2 was downregulated in the Oe-GPX3 group, thus further verifying the anti-apoptotic activity of GPX3 in PANC-1/GEM cells (Fig. 3C). The aforementioned findings indicated that GPX3 overexpression could robustly improve the chemo-sensitivity of PC and GEM-resistant PC cells to GEM.

Subsequently, the present study aimed to uncover the detailed mechanism underlying the effect of GPX3 on antagonizing the malignant behavior of PC cells. Therefore, western blot analysis showed that compared with the Oe-NC group, the protein expression levels of phosphorylated (p)-JNK, p-c-Jun, c-Jun and c-Myc were markedly decreased, while those of p21 were robustly elevated in the Oe-GPX3 group (Fig. 4). The aforementioned finding indicated that GPX3 significantly inhibited JNK/c-Jun signaling in PANC-1 cells.

Finally, to clarify the significance of JNK/c-Jun signaling in the antitumor activity of GPX3 in PC, GPX3-overexpressing PANC-1 cells were treated with 0.01 µM anisomycin, an activator of JNK, and then a series of in vitro experiments were performed. As shown in Fig. 5A-C, anisomycin significantly weakened the anti-proliferative effect of GPX3 on PANC-1 cells, as evidenced by the enhanced cell viability and cell colony formation ability, and the increased number of EdU-positive cells in the anisomycin + Oe-GPX3 group compared with the Oe-GPX3 group. In addition, compared with the Oe-GPX3 group, cell treatment with anisomycin enhanced the migration and invasion rates of PC cells (Fig. 5D). Furthermore, E-cadherin downregulation and N-cadherin and Snail upregulation following cell treatment with anisomycin indicated that the inhibitory effect of GPX3 overexpression on EMT was partially restored by anisomycin treatment (Fig. 5E). In addition, compared with the Oe-GPX3 group, the reduced cell apoptosis rate and Bax/Bcl-2 ratio in the anisomycin + Oe-GPX3 group suggested that cell treatment with anisomycin abrogated the beneficial effect of GPX3 on the chemo-sensitivity of PC cells to GEM (Fig. 6A and B). These fidings indicated that GPX3 inhibited cell proliferation, migration, invasion and chemo-resistance in PC cells via inhibiting the JNK/c-Jun signaling pathway.