HCC and adjacent non-tumorous tissue samples were obtained from The First Affiliated Hospital of University of Science and Technology of China (Hefei, China). The experimental protocol for the use of tissues was approved by the Ethics Committee of The First Affiliated Hospital of University of Science and Technology of China, and written informed consent was obtained from all study participants.

HepG2, Bel 7402, Hep3B and Huh7, human HCC cell lines, were purchased from Procell Life Science & Technology Co., Ltd. The normal hepatocyte, LO2, was acquired from The Cell Bank of Type Culture Collection of The Chinese Academy of Sciences. The cells were cultured at 37˚C in an incubator with 5% CO₂ in Dulbecco’s Modified Eagle’s Medium (DMEM, Biological Industries, Israel) supplemented with 10% FBS (Biological Industries) and 1% penicillin-streptomycin.

U50488 hydrochloride (U50488h, cat. no.0495/25), a specific agonist of KOR, was purchased from Tocris Bioscience Co., Ltd (Bristol, England); morphine Hydrochloride, a specific agonist of MOR, was obtained from Northeast Pharmaceutical Group (Shenyang, China); and oxycodone hydrochloride, an agonist for both KOR and MOR, was from Mengdi Pharmaceutical Co., Ltd. (Beijing, China). GSK2656157 (cat. no. T2654 from Topscience, Shanghai, China), a highly specific and ATP-competitive eukaryotic translation initiation factor 2α (PERK) inhibitor, was dissolved in dimethylsulfoxide (DMSO) to a stock concentration of 2 mM and stored at − 20°C, and diluted 1:1000 in cell culture medium to make a working solution. Hep3B and Huh7 were incubated with U50488h (0,0.1,1,10 μM), oxycodone (0,0.1,1,10 μM) and morphine (0,0.1,1,10 μM) for 24h and 48h in viability assay. The concentration of drugs and the time of stimulation in the follow-up experiments were chosen based upon successful inhibition in viability assay: U50488h (10 μM) oxycodone (10 μM) and morphine (10 μM) for 48h.

For immunofluorescence, climbing slides were pre-placed in six-well plate, then LO2, Hep3B, and Huh7 cells (3×10⁵/well) were seeded into the six-well plate, cultured in a 37°C and 5%CO₂ incubator overnight. 4% paraformaldehyde (biosharp, China) was used to fix at room temperature for 20 minutes and 50ul of blocking solution for immunofluorescence (Proteintech,China) was dropped on the climbing slide and incubated at room temperature for 1 hour. The following primary antibodies KOR (1:100; cat. no. ab113533; Abcam, UK) and MOR (1:100, cat. no.DF5045; Affinity, China) were diluted in the climbing slides, and incubated overnight at 4°C in a wet box. On the next day, the climbing slides were washed three times with phosphate-buffered saline tween-20(PBST), each time for 5 minutes, then 50ul Cy3 Goat anti-rabbit secondary antibody (1:100, Proteintech, China) was added to the climbing slides, and incubated in a cassette at room temperature for 1 hour. The slides were taken out from the six well plate, and a drop of a fluorescent antiquenching agent (meilunbio, China) containing DAPI was placed on the glass slide, covered under a coverslip, avoiding air bubbles. The cells were observed with the fluorescence microscope (Nikon type 120c, Japan).

Total RNA was extracted from cells using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, USA). Total RNA was reverse transcribed into cDNA using the HiScript Q RT SuperMix for qPCR (Vazyme Biotech, China) according to the manufacturers’ instructions. qPCR was subsequently performed using specific primers and an AceQ qPCR SYBR Green Master Mix (Vazyme Biotech) with the following cycling parameters: 95°C for 5 min, followed by 40 cycles of 95°C for 10 s and 60°C for 30 s. The specific primer pairs used for the qPCR analyses are indicated in the Table S1 of the Supplementary Material . Relative gene expression was calculated using the 2^-ΔΔCq method. GADPH mRNA expression level was used as the internal normalization control.

Cells or tissues were lysed in RIPA lysis buffer (Beyotime Institute of Biotechnology) on ice for 30 min and the lysate was centrifuged at 3000g, 4°C for 15 min. The concentration of the proteins in the supernatants was determined using the BCA protein quantification kit (Beyotime Institute of Biotechnology) according to the manufacturer’s instructions. Proteins (30 µg/lane) were separated in 8, 10 or 12% SDS-PAGE, respectively, according to the molecular weight of the target proteins, then transferred onto a PVDF membrane (Millipore, Sigma, USA) using transfer buffer for 60-90 min. The membranes were subsequently blocked with 5% non-fat milk at room temperature for 2 h, and then incubated with the following primary antibodies overnight at 4°C: anti-β-actin (1:5,000; cat. no. AF7018; Affinity Biosciences, USA), anti-KOR (1:900; cat. no. ab113533; Abcam, UK), anti-GRP78 (1:1,000; cat. no. ab191023; Abcam, UK), anti-CHOP (1:500; cat. no. ab11419; Abcam, UK), anti-GADPH(1:8,000; ProteinTech Group, China), anti-caspase-3 (1:3,000; cat. no. 19677-1-AP; ProteinTech Group, China), anti-E-cadherin (1:5,000; cat. no. 20874-1-AP; ProteinTech Group, China), anti-Bax (1:2,000; cat. no. 50599-2-lg; ProteinTech Group, China), anti-PERK (1:1000; cat. no.5683, Cell Signaling Technology, Boston, USA), anti-N-cadherin (1:1,000; cat. no. 13116; Cell Signaling Technology) and anti-vimentin (1:1,000; cat. no. 5741; Cell Signaling Technology). Following the primary antibody incubation, the membranes were washed with TBS-Tween 20 three times and incubated with anti-rabbit (1:10,000; cat. no. A0208; Beyotime Institute of Biotechnology) IgG HRP-conjugated secondary antibodies at 37°C for 2 h. Protein bands were visualized using an ECL kit (Thermo Fisher Scientific, Inc.) and densitometric analysis was performed using ImageJ software (v1.8.0; National Institutes of Health).

A total of 6×10³ Huh7 and Hep3B cells/well were seeded into 96-well plates and incubated overnight. Following the incubation, the cells were exposed to various concentrations (0, 0.1, 1 and 10 μM) of U50488h, oxycodone or morphine for 24 and 48 h. The cell viability was detected using the cell counting kit-8 (CCK-8, BestBio, Shanghai, China) according to the manufacturer’s protocol. The absorbance was measured using an microplate reader (USCN KIT INC, Wuhan, China). Relative cell viability (%) was calculated using the following equation: (Treated cells – blank well)/(control cells – blank well) x100.

For the colony formation assay, 500 cells/well were plated into 6-well plates and cultured with DMEM supplemented with 10% FBS at 37°C in 5% CO₂ for 2 weeks. The cells were fixed with methanol for 25 min and stained with crystal violet solution for 15 min. Visible colonies were imaged and observed using phase-contrast microscope (Olympus, Japan), and counted using ImageJ software.

Huh7 and Hep3B cells were treated with 10 µM U50488h, 10 µM oxycodone or 10 µM morphine at 37°C for 48 h, and then stained with an Annexin V/FITC Apoptosis Detection kit (BestBio, Shanghai, China) according to the manufacturer’s protocol. The percentage of early and late apoptotic cells was analyzed using a FACSCalibur flow cytometer (Becton–Dickinson, San Jose, CA, USA).

The migratory ability of HCC cells was analyzed using a wound healing assay. Briefly, 4◊105 cells were seeded into 6-well plates, cultured to 70% confluence, treated with DMEM, 10 µM U50488h, 10 µM oxycodone or 10 µM morphine and incubated for 48 h at 37°C. The cell monolayer was scratched with a 200ul pipette tip to create an artificial wound, washed with PBS three times, and then incubated with serum-free DMEM at 37°C with 5% CO2 for another 48 h. The width of the wound area was measured at 0 and 48 h after the wound creation. The migratory ability of the cells was observed and calculated as the wound area at 48 h compared with the wound area at 0 h.

Female 5 to 6-week-old and 20-25g Balb/C nude mice were selected to establish the xenograft model. 5◊10⁶ Hep3B cells were injected into the right axilla of 20 mice. The mice were randomly divided into four groups: PBS (control), U50488h (1.25 mg/kg) (21), oxycodone (0.307mg/kg/d for first 15 days and then 0.715 mg/kg/d, equivalent to the analgesic potency of morphine), and morphine (0.714 mg/kg/d for first 15 days and then 1.43 mg/kg/d) (22). PBS and drug treatments were started on day 5 (the tumor grew to about 30-60mm³), intraperitoneally injected every two days. A vernier caliper was used to measure the width (a) and length (b) of the tumors and the tumor volumes were calculated by the formula (a²×b)/2. All the mice were sacrificed on the 21st day post Hep3B injection and the tumors were weighed. The animal experiment was approved by the Animal Protection and Use Committee of the First Affiliated Hospital of the University of Science and Technology of China (2021-N(A)-56).

For IHC staining for PERK and caspase3 expression in HCC xenograft in mice, consecutive 4μm sections of the paraffin-embedded tumor specimens were prepared for hematoxylin-eosin (HE) staining and IHC. The sections were deparaffinized in xylene and rehydrated with a descending alcohol series. After antigen retrieval using sodium citrate and endogenous peroxidase activity blocking in 3% hydrogen peroxide at 37°C for 10 min, the tumor tissues were incubated with a rabbit anti-PERK antibody (1:2000; cat. no.5683, Cell Signaling Technology, Boston, USA), or rabbit anti-caspase3 antibody (1:500; cat. no. 19677-1-AP; ProteinTech Group, Wuhan, China) overnight at 4°C. Following the primary antibody incubation, the sections were rinsed with PBS three times (5 min each time) and incubated with an HRP-conjugated goat anti-rabbit secondary antibody (1:200; cat. no. S0001, Affinity, USA) at 37°C for 30 min. After washing, the slides were stained with diaminobenzidine and counterstained with hematoxylin solution. Finally, the tissue sections were mounted and observed under a light microscope. Obvious brownish staining in the cytoplasm were identified as positive cells. Assessment of the mean of integrated optical density (IOD) staining was evaluated using Image-Pro Plus 6.0 software.

Data are presented as the mean ± SD of at least three independent experiments. Statistical analyses were performed using GraphPad Prism 8.0 software (GraphPad Software, Inc.). A one-way ANOVA was used for multigroup comparisons and Student’s t-test was used to compare the statistical differences between two groups. P<0.05 was considered to be a statistically significant.

For comparison of the expression of KOR and MOR, the normal hepatocyte LO2 and the human HCC cell lines HepG2, Bel-7402, Hep3B and Huh-7 were used. Immunofluorescent staining revealed that KOR and MOR proteins were expressed on the plasma membrane of the HCC cells, the expression level of KOR and MOR were higher in Hep3B and Huh7 cells compared with LO2( Figure 1A ). RT-qPCR analyses indicated that mRNAs of KOR and MOR were highly expressed in Hep3B and Huh7 ( Figure 1B ). KOR and MOR proteins expression level by western blotting was similar to the RT-qPCR results ( Figure 1C ). So Hep3B and Huh7 were chosen for subsequent experiments. Furthermore, western blotting analysis demonstrated that the expression levels of the KOR and MOR were upregulated in the HCC tissues compared with those in the adjacent non-tumorous tissues ( Figure 1D ).

Since expression of KOR or MOR was much higher than that of DOR in the HCC tissues and cell lines as described above, agonists specific for KOR and MOR were tested to observe their effects on malignant phenotypes of HCC cells. As shown in the Figure 2A , treatment of Hep3B and Huh7 cells with increased concentrations of U50488h (a KOR-specific agonist), oxycodone (an agonist for both KOR and MOR), or morphine (an MOR-specific agonist) for 24 and 48 h reduced cell proliferation in a dose- and time-dependent manner. Furthermore, U50488h and oxycodone significantly suppressed the colony formation of both Hep3B and Huh7 cells ( Figure 2B ). However, the MOR-specific agonist morphine did not exert this effect ( Figures 2B, C ). Similarly, wound healing assay revealed that the migration of HCC cells was inhibited by U50488h or oxycodone to varying degrees, but not by morphine, as compared with cell migration in the control group ( Figures 3A, B ). Consistent with the results of wound healing assay, western blotting analyses revealed upregulated expression of E-cadherin and downregulated expression of N-cadherin and vimentin ( Figures 3C–F ) by treatment with U50488h or oxycodone, indicating that U50488h or oxycodone suppressed the epithelial-mesenchymal transition (EMT) of the HCC cells. Morphine also increased the expression of E-cadherin as compared to the control, but to a less extent than U50488h or oxycodone ( Figures 3C, D ). Since the expression of N-cadherin and vimentin in the morphine-treated cells was comparable to that of the control ( Figures 3C, E, F ), morphine did not seem to have an obvious inhibition of EMT. These results demonstrated that the inhibitory effect on HCC cells migration by U50488h or oxycodone was via suppression of EMT. Taken together, malignant phenotypes of HCC cells including cell proliferation, colony formation and migration can be suppressed by KOR activation either via the KOR-specific agonist U50488h or via the mixed agonist oxycodone, but not by MOR activation via the MOR-specific agonist morphine.

For comparison of apoptosis-inducing abilities of the KOR and MOR agonists, Hep3B and Huh7 cells were treated with 10μm of U50488h, oxycodone or morphine for 48h, and the apoptotic cells were then analyzed by flow cytometry. As shown in Figures 4A, B , a significant increase of apoptosis rate was found in the U50488h- or oxycodone-treated cells compared to the control treatment, but not in the morphine-treated cells. In Hep3 and Huh7 cells, pro-apoptotic proteins Bax and cleaved caspase 3 were increased by U50488h or oxycodone treatment, but not by treatment with morphine ( Figures 4C–E ). These results indicate that HCC apoptosis can be induced by KOR activation, but not by MOR activation.

For understanding the underlying mechanism of apoptosis induced by the KOR agonists, RT-qPCR was first performed to analyze the relative mRNA expression of endoplasmic reticulum (ER) stress-associated molecules including GRP78, PERK and CHOP. The results revealed that the expression of these molecules was significantly increased by the KOR agonist U50488h compared with those in the control group ( Figure S1 ). Similarly, increased protein expression of GRP78, PERK and CHOP in the U50488h-treated cells was detected by western blotting ( Figures 5A–D ). However, the MOR agonist morphine did not induce an increase of GRP78, PERK and CHOP at both the mRNA and protein levels ( Figure S1 , Figures 5A–D ). These results indicate that the specific KOR agonist U50488h may promote apoptosis of HCC cells by upregulating PERK pathway.

To further explore the relationship between ER stress and apoptosis, the PERK inhibitor GSK2606414 was used. Hep3B and Huh7 cells were co-treated with 2 µM GSK2606414 and the selective KOR agonist U50488h (10 µM) for 48 h, and flow cytometric analysis of apoptosis and cell proliferation assay were performed. As shown in Figures S2A, B , GSK2606414 reduced the apoptosis rate induced by U50488h. Also, GSK2606414 mitigated the inhibitory effect of U50488h on HCC cell proliferation ( Figure S2C ). Western blotting verified that the expression of GRP78, PERK and CHOP was decreased by treatment of GSK2606414 ( Figures S2D–G ). Although GSK2606414 was unable to completely reverse the effects of U50488h on HCC apoptosis and cell proliferation, these findings indicated that the effects were partially via the up-regulated PERK signaling.

For observation of the effects of the 3 OR agonists on HCC growth in vivo, Balb/C nude mice were injected with 5◊10⁶ Hep3B cells into the right axilla. The drugs (U50488h, oxycodone, or morphine) with equivalent analgesic potency were intraperitoneally injected every two days until the tumor grew to about 30-60mm³ and PBS was used as the control. U50488h and oxycodone, but not morphine, reduced the tumor weight and volume compared with the PBS group ( Figures 6A–D ). The OR agonists used in our study did not decrease the body weight of the mice ( Figure 6D ). IHC staining indicated that expression of PERK and caspase 3 was enhanced in the HCC tissues of the mice-treated with U50488h ( Figures 7A, B ), but not with morphine. Similar results were observed in western blotting analysis ( Figure 7C ).