SK-Hep1, Hep3B, and SNU-354 cells were grown in DMEM with penicillin (100 U/mL)-streptomycin (100 U/mL) and 10% heat-inactivated fetal bovine serum. SNU-387 cells were cultured in RPMI-1640 medium supplemented as above. All cells were cultured in a cell incubator at 37°C, 5% CO₂, and appropriate humidity. Hepatocellular carcinoma cells were purchased from the Shanghai Cell Bank of the Chinese Academy of Sciences. Dimethyl sulfoxide (DMSO) was purchased from Sigma Chemical Co. Carrimycin (batch number:20180625), monomeric isovalerylspiramycin I (batch number:20180517) and, lenvatinib (purity:99.67%) were provided by Shangke Biopharmaceutical (Shanghai) Co., Ltd.

In vitro test solution: carrimycin (3, 10, 30 μg/mlL), monomeric isovalerylspiramycin I (1, 3, 10 μg/ml), and lenvatinib (3, 10 μg/ml).

SK-Hep1, Hep3B, SNU-354, and SNU-387 cells were seeded at a density of 1×10⁵ cells/well in 96-well plates and incubated overnight. The cells were pretreated with carrimycin (0.001, 0.003, 0.01, 0.03, 0.1, 0.3, 1, 3, 10, 30 μg/ml), monomeric isovalerylspiramycin I (0.001, 0.003, 0.01, 0.03, 0.1, 0.3, 1, 3, 10, 30 μg/ml), and lenvatinib (0.001, 0.003, 0.01, 0.03, 0.1, 0.3, 1, 3, 10, 30 μg/ml) for 24, 48, 72 h. At the end of the experiment, the medium was removed and cells were cultured with MTT solution (5 mg/ml) [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] (Sigma, St. Louis, MO, United States) for 4 h. The viable cells converted MTT to formazan, which generated a blue-purple color after dissolving in 100 μl of DMSO. The absorbance at 570 nm was measured by Multiskan GO.

SK-Hep1, Hep3B, SNU-354, and SNU-387 cells were seeded in 6-well plates (5×10³ cells/plate) for colony formation assay. Following 24 h incubation, the cells were treated with various concentrations of carrimycin, monomeric isovalerylspiramycin I, and lenvatinib. Cells treated with DMSO were used as negative controls. After 10 days, the cells were washed with PBS. The colonies were stained with 1% crystal violet for 30 s after fixation with 10% formaldehyde for 5 min. The dishes were rinsed three times with PBS and air-dried.

EdU is a thymidine analog that is incorporated into replicating chromosomal DNA during the S phase of the cell cycle. We used the EdU incorporation assay, which is a highly sensitive and specific method, to investigate the inhibitory effects of carrimycin, monomeric isovalerylspiramycin I, and lenvatinib. Cells were cultured in 96-well plates for 24 h, and then cells were treated with carrimycin, monomeric isovalerylspiramycin I, and lenvatinib for 16 h. Next, the cells were labeled with 5-ethynyl-2′-deoxyuridine (EdU, RIBOBIO; Guangzhou, China) for 6 h, the cells were washed with PBS for 2 min. After being fixed with 4% paraformaldehyde for 30 min, the cells were decolorized with 2 mg/ml glycine for 3 min, treated with 0.5% Triton X-100 for 10 min, and rinsed with PBS three times. Thereafter, the cells were exposed to 100 μl of 1×Apollo^® reaction cocktail for 30 min and incubated with 5 μg/ml of Hoechst 33,342 to stain the cell nuclei for 30 min (Mi et al., 2017). The stained cells were observed with Olympus IX83 inverted fluorescence microscope (Olympus Corporation, Tokyo, Japan).

SK-Hep1, Hep3B, SNU-354, and SNU-387 cells were seeded in 24-well plates (1×10⁵ cells/plate) for wound-healing assay. The cells were cultured to form a monolayer and wounds were made by scratching with a pipette tip and exposed with different treatments for 0, 24, 48, and 72 h. The cell migrations were photographed at identical locations and analyzed by comparing the final gap width to the control group with Olympus IX83 inverted fluorescence microscope (Olympus Corporation, Tokyo, Japan).

SK-Hep1, Hep3B, SNU-354, and SNU-387 cells were seeded in a 6 cm Petri dish for matrigel transwell invasion assay. After the cells were attached, the cells were treated with various concentrations of carrimycin, monomeric isovalerylspiramycin I and lenvatinib, and continued to cultivate for 4 days. Matrigel (BD Biocoat, Bedford, MA, United States) was diluted with a serum-free medium to a final concentration of 1 mg/ml, 100 μl of the diluted matrigel was added vertically to the bottom center of the upper chamber of the transwell, and incubated at 37°C for 1 h to make it dry into a gel. After washing the cells with PBS, the cells were suspended in serum-free medium and counted, then adjusted the concentration to 1 × 10⁶/ml, added 100 μl of the cell suspension to the upper chamber and 600 μl of medium containing 10% FBS to the lower chamber (bottom of the 24-well plate). After 24 h of invasion at 37°C, cells passing the filters into bottom wells were fixed with paraformaldehyde and stained with 0.1% crystal violet. Invaded cells were photographed by a microscope.

SK-Hep1, Hep3B, SNU-354, and SNU-387 cells were seeded in 6-well plates (5×10³ cells/plate) for cell apoptosis analysis. After the cells were attached, the cells were treated with various concentrations of carrimycin, monomeric isovalerylspiramycin I, and lenvatinib. After 48 h of treatment, the cells were collected in a 1.5 ml tube, washed with PBS for 3 times, centrifuged to discard the supernatant, washed with binding buffer, centrifuged to discard the supernatant, the cells were resuspended in 50 μl binding buffer, and stained with Annexin V-FITC for 30 min and propidium iodide for 10 min at 37°C in the dark. The samples were analyzed by flow cytometry. The CellQuest software was used to analyze the data (Becton-Dickinson).

SK-Hep1, Hep3B, SNU-354, and SNU-387 cells were seeded in a 6 cm Petri dish for western blotting. After the cells were attached, the cells were treated with various concentrations of carrimycin, monomeric isovalerylspiramycin I, and lenvatinib for 12 h. The cells were washed with pre-cooled PBS and lysed on ice for 30 min. The lysate was collected, centrifuged at 14,000 rpm and 4°C for 30 min, and then the supernatant was collected to obtain cell extracts. The total protein concentrations were measured by the Bradford method. The cell extracts were separated by SDS-polyacrylamide gels (8–10%) and then the protein was transferred to a polyvinylidene difluoride membrane (Millipore, Bedford, MA, United States). The membrane was blocked with 5% skim milk, and incubated with the PD-L1 (Novus Biologicals, NBP1-76769, 1:2,000) and VEGF (Santa Cruz, Cat# sc-152, 1:500) primary antibody (The antibodies were diluted using 1% BSA in PBS). After binding of rabbit secondary antibody (Santa Cruz, Cat# sc-2004, 1:2,000) coupled to horseradish peroxidase, proteins were visualized by enhanced chemiluminescence according to the manufacturer’s instructions (Amersham Pharmacia Biotec, Buckinghamshire, United Kingdom).

Total RNA was extracted using a Trizol reagent (Invitrogen). cDNA synthesis was executed with the Transcriptor First Strand cDNA Synthesis Kit (Roche, Basel, Schweiz) and then amplified by TaKaRa TaqTM RT-PCR kit (TaKaRa). The cycling conditions were 94°C for 4 min, followed by 40 cycles of 94°C for 30 s, 58°C for 30 s, 72°C for 30 s, and final conditions was 72°C for 5 min. The PCR products were electrophoresed on 3% agarose gels and stained with ethidium bromide. The stained bands were visualized under UV light.

BALB/c athymic nude mice (male, 4–5 weeks old, 22 ± 2 g and specific-pathogen-free) were purchased from BEIJING HFK BIOSCIENCE CO., LTD (Beijing, China). The animals were treated humanely, and all efforts were made to minimize the animal suffering and numbers of experimental animals. The animals were given standard pellet food and water ad libitum and subjected to a 12 h light/12 h dark cycle. The room temperature was 19–23°C, and humidity was 35–55%. Lenti-EGFP virus particles were used to transfect SK-Hep1 cells (5×10⁷ cells/mL). The transfected cells (0.1 ml) were subcutaneously injected into the mice on the left side of the back. After 7 days of injection, the tumors formed on the back of the mice. The mice were randomly assigned to nine groups (n = 9/group). The volume of each administration is 200 μl, the carrimycin concentration is 2, 4, 8 μg/ml. The monomeric isovalerylspiramycin I concentration is 1, 2, 4 μg/ml. The lenvatinib concentration is 1 μg/ml. Carrimycin, monomeric isovalerylspiramycin I, and lenvatinib were orally given to the mice once a day for 28 days. The blank control group and the model group were given the same volume of solvent dissolved in carrimycin. The sizes of the tumors were measured every 2 days. The volume of the tumor was calculated by using the equation: volume of the tumor (mm³)=(width)² × length/2. After 28 days, mice died peacefully after intraperitoneal injection of barbiturates, tumors were excised and fixed to conduct further experiments. All the protocols have obtained approval from the Committee of Animal Experiments Ethics in Yanbian University [(SPF, SCXK [JI] 2017-0003), Jilin, China].

All the procedures and operations in animal experiments had adhered to IACUC guidelines. C57BL/6J mice (male, 4–5 weeks old, 20 ± 2 g and specific-pathogen-free) were purchased from BEIJING HFK BIOSCIENCE CO., LTD (Beijing, China) and kept in the experimental animal center of Yanbian University authorized by the Chinese Association for Laboratory Animal Sciences (approval ID: SCXK (Jinlin) 2017-0003) during the experiment. The animals were treated humanely, and all efforts were made to minimize the animal suffering and numbers of experimental animals. The animals were given standard pellet food and water ad libitum and subjected to a 12 h light/12 h dark cycle. The room temperature was 19–23°C, and humidity was 35–55%.The mice were randomly assigned to nine groups (n = 9/group). H22 tumor cells were cultured, collected, and centrifuged to adjust the cell concentration to 1 × 10⁷/ml. After disinfecting the abdomen of C57BL/6J mice with alcohol cotton balls, the mice were injected intraperitoneally with 0.1 ml H22 cell suspensions, and the changes of the mice abdomen were observed every day. After 7–10 days, a large amount of ascites was formed in the abdomen of the mice. We used disposable syringes to extract ascites cells from mice, the cell concentration was adjusted to 1 × 10⁷/ml with sterile PBS buffer, and cells suspension were injected intraperitoneally to second-generation mice. After repeating the operation three times, the ascites cells were collected and washed twice with saline, adjusted the cell concentration to 2 × 10⁷/ml, three generations of ascites cells were collected for subsequent use.

C57BL/6J healthy mice fasted for 10 h and water-free for 4 h before surgery. The mice were anesthetized intraperitoneally with 4% chloral hydrate at a dose of 0.01 ml/g. The mice were fixed, and their skins were routinely disinfected and flattened. An incision of about 1 cm was made under the xiphoid process in the middle of the upper abdomen, and the skin and peritoneum were cut layer by layer, arranged the gauze soaked with saline under the incision, gently squeezed both sides of the costal arch. Exposed the left lobe of the mouse liver to wet gauze, and used 50 μl micro syringes to insert the needle along the long axis of the liver lobe at an angle of 20°C parallel to the liver surface. After inserting the needle, moved about 0.5 cm in parallel, 50 μl containing 3 × 10⁶ H22 cells were slowly injected into the mouse liver. With drawn the needle, gently pressed the needle eye with a cotton swab, and gently wiped the area around the needle eye with an alcohol cotton ball to remove a small number of leaked tumor cells and prevented tumor cells from escaping. Gently returned the liver to the abdominal cavity. After checking for active bleeding, sutured the abdominal wall layer by layer with No. One silk thread, disinfected the skin again, and closed the operation. After 4 h of fasting and water prohibition, 0.3% sucrose water was given, and 12 h later, they were given with normal food and water.

After the model was established, the mice were randomly divided into nine groups: blank control group, model group, carrimycin 20, 40, 80 mg/kg group, monomeric isovalerylspiramycin I 10, 20, 40 mg/kg group and the positive control drug lenvatinib 10 mg/kg group (n = 12/group). Carrimycin, monomeric isovalerylspiramycin I, and lenvatinib were orally given to the mice once a day for 14 consecutive days. The volume of administration was 5 ml/kg. The blank control group and the model group were given the same volume of solvent dissolved in carrimycin.

After 13 days, mice died peacefully after intraperitoneal injection of barbiturates, the livers were dissected and taken out, the tumor tissues were separated, weighed, and photographed. Tumors were fixed to conduct further experiments. All the protocols have obtained approval from the Committee of Animal Experiments Ethics in Yanbian University.

Liver tissues and colon tissues were fixed in 4% formaldehyde, embedded in paraffin blocks, and cut into 5 μm tissue specimens. The sections were stained with hematoxylin and eosin (H&E) according to a standard protocol, and examined under a light microscope (ECLIPSE Ni-U, Nikon, Japan).

The concentrations of mouse VEGF were measured using ELISA kits (Shanghai Enzyme-Linked Biotechnology Co., Ltd. LOT: 07/2019) according to the manufacturer’s instructions. The blood in mice was collected, and the serum was separated after centrifugation at 3,000 rpm for 20 min within 2 h after self-coagulation. The microwell plate was coated with mouse vascular endothelial growth factor VEGF to make a solid phase antibody. The sample was added to the coated microwell and combined with the HRP-labeled detection antibody. After washing, the substrate TMB was added to coloring. TMB is converted into blue under the catalysis of HRP enzyme, and change to yellow under the action of nucleic acid. Measure the OD value with a microplate reader at 450 nm wavelength.

All values are expressed as mean ± standard deviation. A comparison of the results was performed with one-way ANOVA and Tukey’s multiple comparison tests (GraphPad Prism software 5.0, La Jolla, CA, United States). Statistically significant differences between groups were defined as p-values less than 0.05.

To detect the effect of carrimycin, monomeric isovalerylspiramycin I, and lenvatinib on hepatocellular carcinoma cells, we used the SK-Hep1, Hep3B, SNU-354, and SNU-387 cell lines to perform cell viability assay. Cells were treated with increasing concentrations of drug for 24, 48, and 72 h, and the cell viability was determined by MTT assay. At 24 h, compared with 0.001 μg/ml drug treatment groups, carrimycin, monomeric isovalerylspiramycin I, and lenvatinib have a significant inhibitory effect on cell viability. At 48 h and 72 h, the inhibitory effect of the three drugs was more obvious in SK-Hep1, Hep3B, SNU-354, and SNU-387 cell lines (Figures 1A–D). In vitro activities of the carrimycin, monomeric isovalerylspiramycin I, and lenvatinib were summarized in Supplementary Figure S1E.

We used colony formation experiments to further study cell proliferation. Compared with the control group, 30 μg/ml carrimycin, 10 μg/ml monomeric isovalerylspiramycin I, and 10 μg/ml lenvatinib can significantly inhibit cell colony formation, and the inhibitory effects of the three drugs are similar (Figures 2A,B). In SK-Hep1, Hep3B, and SNU-354 cells, the inhibition rates of monomeric isovalerylspiramycin I, carrimycin, and lenvatinib were greater than 80%. In SNU-387 cells, the inhibition rates of monomeric isovalerylspiramycin I and carrimycin were greater than 80%, and the inhibition rate of lenvatinib was approximately 60%.

To determine whether carrimycin, monomeric isovalerylspiramycin I, and lenvatinib could affect DNA replication, the EdU incorporation assay was performed. For all 4 cell lines, incorporation of EdU in DNA was significantly reduced after treatment with 30 μg/ml carrimycin, 10 μg/ml monomeric isovalerylspiramycin I, or 10 μg/ml lenvatinib compared with the negative control groups (Figure 3A). The inhibitory effect of carrimycin and monomeric isovalerylspiramycin I was similar to that of lenvatinib. As shown in Figure 3B, in the SK-Hep1, Hep3B, SNU-354 and SNU-387 lines, compared with the control group, the monomer isovalerylspiramycin I, carrimycin and levatinib all had inhibitory effects. The inhibitory activity of carrimycin, monomeric isovalerylspiramycin I and lenvatinib were similar.

Inhibiting the migration of hepatocellular carcinoma cells can effectively inhibit the spread of tumors. Therefore, we evaluated the migration potential of the hepatocellular carcinoma lines in the presence of the three agents. Figures 4A–D show that migration was significantly suppressed in the 4 cell lines after treatment with 30 μg/ml carrimycin, 10 μg/ml monomeric isovalerylspiramycin I, and 10 μg/ml lenvatinib. The histograms show that the inhibitory effects of the agents on migration were similar.

Inhibiting the invasion of hepatocellular carcinoma cells can effectively inhibit the spread of tumors, too. Therefore, we also evaluated the invasion potential of the hepatocellular carcinoma lines in the presence of the three agents. Figure 5A indicates that invasion potential was significantly inhibited in the 4 lines when treated with 30 μg/ml carrimycin, 10 μg/ml monomeric isovalerylspiramycin I, and 10 μg/ml lenvatinib compared with the negative control groups. The histograms show that the inhibitory effects of monomeric isovalerylspiramycin I, carrimycin, and lenvatinib on invasive potential were similar in all 4 cell lines (Figure 5B).

Promoting apoptosis of tumor cells can effectively alleviate the development and reduce the size of tumors. Figure 6 shows that the rate of apoptosis increased in each cell line after treatment with carrimycin, monomeric isovalerylspiramycin I, and lenvatinib compared with the negative control group. Figure 6 shows the apoptosis rates were 38.2, 30.3 and 15.7% in SK-Hep1 cells (Figure 6A), 16.1, 21.4, and 28.5% in Hep3B cells (Figure 6B), 23.7, 14.5 and 11.3% in SNU-354 cells (Figure 6C), and 34.5, 28.9, 17% in SNU-387 cells (Figure 6D) after treatment with 30 μg/ml carrimycin, 10 μg/ml monomeric isovalerylspiramycin I and 10 μg/ml lenvatinib, respectively. As detected by flow cytometry, the apoptosis rate of SK-Hep1, SNU-354, Hep3B and SNU-387 cells was promoted with the treament with carrimycin, monomeric isovalerylspiramycin I and lenvatinib compared with the control groups. The pro-apoptotic rates of carrimycin, monomeric isovalerylspiramycin I and lenvatinib were similar. The histograms show that carrimycin and monomeric isovalerylspiramycin I could promote apoptosis of hepatocellular carcinoma cells.

VEGF is an important marker of angiogenesis, and PD-L1 is an immune cell checkpoint-related protein. Therefore, we analyzed the expression of these two proteins by western blot. Figure 7 indicates that the expressions of PD-L1 protein levels were significantly lower after treatment with the three agents in all 4 lines compared with the negative control groups. And the expressions of VEGF protein levels were significantly lower after treatment with the monomer isovalerylspiramycin I and levatinib in all 4 cell lines compared with the negative control groups. However, carrimycin has no significant effect on VEGF in Sk-Hep1 cells (Figures 7A–D).

Next, we test the effect of carrimycin, monomeric isovalerylspiramycin I, and lenvatinib on VEGF and PD-L1 mRNA levels in SK-Hep1, SNU-354, Hep3B and SNU-387 cells. As showed in Figures 8A–D, compared with the control group, carrimycin, monomeric isovalerylspiramycin I, and lenvatinib can inhibit VEGF and PD-L1 mRNA levels in a dose-dependent manner, and the inhibitory effects of 30 μg/ml carrimycin, 10 μg/ml monomeric isovalerylspiramycin I, and 10 μg/ml lenvatinib on VEGF and PD-L1 mRNA levels were similar. The primer sequence in Supplementary Table S1.

Given that carrimycin and monomeric isovalerylspiramycin I inhibited the proliferation of tumor cells in vitro, we investigated whether these results could be translated in vivo. Thus, we developed an HCC xenograft model using SK-Hep1 cells in nude mice. The tumor volumes of the treated groups were significantly reduced compared with the mock group (Figures 9A,B). Compared with the positive control drug lenvatinib (10 mg/kg dose), there were similarities results with crimycin at the 80 mg/kg dose and the monomeric isovalerylspiramycin I at the 40 mg/kg dose. As shown in Figure 9C, after 28 days, carrimycin dose of 20, 40, and 80 mg/kg resulted in tumor inhibition rates of 59.31, 75.88, 90.62%, respectively, and monomeric isovalerylspiramycin I dose of 10, 20, and 40 mg/kg resulted in 62.15, 78.74, 90.47% tumor inhibition, respectively. The positive control group treated with 10 mg/kg lenvatinib showed a tumor inhibition rate of 92.61%. Figure 9D indicates that the body weights of the mice did not change. The tumor tissues were embedded in paraffin and stained with hematoxylin and eosin for histological examination and compared with the mock control group. After staining the tumor tissue with HE (Figure 10A), the tumor tissues in the model group were poor differentiation, large atypia, invasive or exogenous growth, and the boundary between them and surrounding tissue was not clear, as shown by the arrows. After drug treatment, the tumor tissues were well differentiation, with small atypia, rare mitotic, without pathological mitotic figures, expansion or exogenous growth, the boundary with the surrounding tissue is clear. The tumor tissues in the 20 mg/kg group of carrimycin and the 10 mg/kg group of monomeric isovalerylspiramycin I showed peripheral invasive growth. However, compared with the mock mice, the infiltrating growth around the tumor tissues was weakened by carrimycin (40 mg/kg and 80 mg/kg doses), monomeric isovalerylspiramycin I (20 mg/kg and 40 mg/kg doses), and lenvatinib (10 mg/kg dose), the boundary with the periphery became clear and the tumor size was significantly reduced. Using the ELISA, it was observed from Figure 10B that carrimycin, monomeric isovalerylspiramycin I, and lenvatinib treatment significantly reduced the concentration of VEGF in mouse serum. Compared with the positive control drug lenvatinib, the effect of carrimycin and monomer isovalerylspiramycin I on the concentration of VEGF was similar.

We examined the in vivo efficacy of carrimycin and monomeric isovalerylspiramycin I in an orthotopic liver transplantation model for HCC. Figures 11A,B show the representative mice and the transplanted tumor pictures of orthotopic liver transplantation model. Compared with the positive control drug lenvatinib (10 mg/kg dose), there were similarities results with crimycin at the 80 mg/kg dose and the monomeric isovalerylspiramycin I at the 40 mg/kg dose. There were no differences in body weight between the groups of mice (Figure 11C). After staining the tumor tissue with HE (Figure 12A), it was observed that the tumors in the mock group increased in mass and developed large. The transplanted tumor was completely or partially enveloped with the surrounding liver tissue, and there were nodules with different sizes on the surface, mainly with invasive growth, as shown by the arrows. The tumor tissues in the 20 mg/kg group of carrimycin and the 10 mg/kg group of monomeric isovalerylspiramycin I showed peripheral invasive growth. However, compared with the mock mice, the infiltrating growth around the tumor tissues was weakened by carrimycin (40 mg/kg and 80 mg/kg doses), monomeric isovalerylspiramycin I (20 mg/kg and 40 mg/kg doses), and lenvatinib (10 mg/kg dose), the boundary with the periphery became clear and the tumor size was significantly reduced. Furthermore, Figure 12B shows that carrimycin, monomeric isovalerylspiramycin I, and lenvatinib significantly reduced the concentration of VEGF in mice serum compared with the mock mice.