Matrine inhibits cell growth, migration, invasion and promotes autophagy in hepatocellular carcinoma by regulation of circ_0027345/miR-345-5p/HOXD3 axis

Human HCC cell lines Huh-7 and HCCLM3 were purchased from Procell (Wuhan, China). The two cell lines were maintained in Dulbecco’s Modified Eagle Medium (DMEM, Invitrogen, Carlsbad, CA, USA) with 0.1% penicillin/streptomycin and 10% fetal bovine serum (FBS, Invitrogen) in a cell incubator at 37 °C with 5% CO₂.

Overexpression plasmid of circ_0027345 (pcDNA-circ_0027345) and the control (pcDNA-NC), miR-345-5p mimic (miR-345-5p) and the control (miR-NC), inhibitor (anti-miR-345-5p) and the control (anti-miR-NC), small interference RNA targeting HOXD3 (si-HOXD3) and its control (si-NC) were acquired from GenePharma (Shanghai, China). These constructs were transfected into Huh-7 and HCCLM3 cells by using Lipofectamine 3000 (Invitrogen).

96-well plates inoculated with Huh-7 and HCCLM3 cells were placed in a cell incubator overnight. Cells were then stimulated by different doses of matrine (0, 0.4, 0.8 or 1.6 mg/mL) for 48 h. Next, cells were washed and treated with 10 μL cell counting kit-8 (CCK-8, Beyotime, Shanghai, China) for another 2 h, and the optical density (OD) value was estimated by a Wellscan reader (Thermo Labsystems, Santa Rosa, CA, USA) at 450 nm.

Huh-7 and HCCLM3 cells were transfected or treated with matrine for 48 h. After cell collection, apoptotic cells were detected through an Annexin V fluorescein isothiocynate (FITC)/propidium iodide (PI) apoptosis detection kit (Beyotime). The cells were stained with 5 µL FITC and 5 µL PI for 15 min at 37 °C. Then cells were analyzed by flow cytometry (BD Biosciences, Franklin Lake, NJ, USA).

Proteins from HCC cells and nude mouse tumor tissues were extracted by RIPA solution (Beyotime) on the ice. Based on the molecular weight of the protein, different concentrations of sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) was used to separate the proteins, and protein samples were then transferred to polyvinylidene difluoride (PVDF, Beyotime) membranes. Following blocking of 5% non-fat dry milk for 1 h, the membranes were treated with primary antibodies against B cell lymphoma-2 (Bcl-2, 1:1000, ab32124), Bcl-2-associated x (Bax, 1:2000, ab182733), Light chain 3-II/LC3-I (LC3-II/LC3-I, 1:1000, ab128025), Beclin 1 (1:1000, ab210498), HOXD3 (2 µg/mL, ab22840) or glyceraldehyde-3-phosphate dehydrogenase (GAPDH, 1:10,000, ab181602) overnight at 4 °C. Subsequently, membranes were mixed with horseradish peroxidase-conjugated (0.3 µg/mL, ab190492) anti-rabbit antibodies for 1 h at 37 °C. The BeyoECL Plus kit (Beyotime) was used to visualize the protein bands. All the antibodies were obtained from Abcam (Cambridge, MA, USA).

Huh-7 and HCCLM3 cells were collected at 48 h after treatment with matrine or transfection, and fixed in 70% ice-cold ethanol overnight at 4 °C. The next day, cells were first treated with 100 μg/mL RNase A and then stained with PI for 30 min in the dark. The distribution of cells at different phases was analyzed using a FACSCalibur flow cytometer (BD Biosciences).

Cellular migration and invasion analysis were performed by Transwell assay using 24-well transwell plates (Corning Incorporated, Corning, NY, USA). For invasion detection, Matrigel (Corning Incorporated) needed to be laid at the bottom of the upper chambers in advance. HCC cells were resuspended in serum-free medium, and 100 μL cell suspension was seeded into the upper chambers, while the bottom chambers were added to 600  μL completed DMEM with 10% FBS. 24 h later, cells on the lower side of the chambers were stained with 0.1% crystal violet (Beyotime) for 20 min. After washing with PBS, the stained cells were observed and captured by an inverted microscope. For migration detection, there was no need to add Matrigel in upper chambers, the other steps were the same.

The RNA from HCC cells or nude mouse tumor tissues was isolated by TRIzol^® (Invitrogen). Reverse transcription was performed using the All-in-One™ First-Strand cDNA Synthesis Kit (FulenGen, Guangzhou, China). Then, qRT-PCR was conducted using SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA, USA) on the 7900HT Fast Real-Time PCR System (Applied Biosystems). The 2^−ΔΔct method was utilized to calculate the relative expression, and GAPDH was the internal control for circ_0027345 and HOXD3, while U6 was the internal control for miR-345-5p. The primer sequences were as follows: circ_0027345, F: 5′-TCACTGGTTTGGATGCATTG-3′, R: 5′-AAGGTGGCTCATGGAACTTG-3′. GAPDH, F: 5′-TGATGACATCAAGAAGGTGGTGAAG-3′, R: 5′-TCCTTGGAGGCCATGTGGGCCAT-3′. miR-345-5p, F: 5′-TGAGGGGCAGAGAGCGAGACTTT-3′, R: 5′-CTCAACTGGTGTCGTGGA-3′. U6, F: 5′-ACCCTGAGAAATACCCTCACAT-3′, R: 5′-GACGACTGAGCCCCTGATG-3′. HOXD3, F: 5′-CCATAAATCAGCCGCAAGGAT-3′, R: 5′-GATGGGTCTCAGACTTACCTTTGG-3′.

The sequences of circ_0027345 containing miR-345-5p wild-type (WT) binding sites or mutant (MUT) were cloned into the dual-luciferase reporter vector pmirGLO (Promega, Fitchburg, WI, USA), named as circ_0027345 WT and circ_0027345 MUT. Similarly, the 3′untranslated regions (3′UTRs) of HOXD3 containing miR-345-5p binding sites or mutant were cloned into the pmirGLO vector, named as HOXD3 3′UTR WT and HOXD3 3′UTR MUT. Huh-7 and HCCLM3 cells were harvested at 24 h after co-transfection with miR-345-5p or miR-NC and these vectors, respectively. The luciferase activity was determined using a dual-luciferase reporter assay kit (Promega).

Briefly, suspensions of HCCLM3 tumor cells were inoculated into the 4-week-old male nude mice, and the mice were divided into two groups (n = 5 each group). The mice in treatment group were treated with matrine (50 mg/kg) every day, and mice in control group were treated with same amount of normal saline. The tumor volumes were measured every 5 days. The mice were sacrificed after 30 days of inoculation and tumors were weighed. Furthermore, tumor tissues were preserved at − 80 °C for RNA and protein extraction. The animal experiments were permitted by the Animal Care and Use Committee of Fujian Provincial Hospital.

Data analysis was performed using SPSS v19.0 software, and results were shown as the mean ± standard deviation (SD) from at least three times independently. Student’s t-test was utilized to compare the differences between the two groups and one-way analysis of variance (ANOVA) was employed to analyze the differences for multiple comparisons. P < 0.05 was considered statistically significant.

To clarify the role of matrine in HCC, Huh-7 and HCCLM3 cells were treated with various concentrations of matrine for 48 h, following CCK-8 assay was performed. The results showed that matrine repressed cell viability in a dose-dependent manner (Fig. 1a, b). To investigate whether matrine regulated apoptosis of HCC cells, we detected the apoptosis rate by Flow cytometry. As presented in Fig. 1c, the apoptosis ratio of Huh-7 and HCCLM3 cells was significantly increased as the concentration of matrine increasing from 0  mg/mL to 1.6  mg/mL. Bcl-2 and Bax have been widely reported as markers of apoptosis [23]. Western blot data revealed that matrine decreased Bcl-2 expression and increased the expression of Bax in a dose-dependent manner, suggesting that matrine could facilitate apoptosis of Huh-7 and HCCLM3 cells (Fig. 1d, e). Furthermore, Flow cytometry results showed that after the treatment of Huh-7 and HCCLM3 cells with different concentrations of matrine, cell cycle was arrested at the G1/G0 checkpoint, while the cells in S and G2/M checkpoints were decreased (Fig. 1f, g), indicating that matrine could block the cell cycle. Metastasis is the main cause of poor prognosis in cancer patients. To examine whether matrine might repress the migration and invasion of HCC cells, Transwell assay was performed. As presented in Fig. 1h, i, matrine obviously hampered the migration and invasion of Huh-7 and HCCLM3 cells in a dose-dependent manner. Then, the expression of major autophagy regulatory proteins LC3-II, LC3-I and Beclin 1 [24, 25] was measured in matrine-treated Huh-7 and HCCLM3 cells by western blot. The results showed that the expression of Beclin 1 and ratio of LC3-II/I were enhanced in cells treated with matrine in a dose-dependent manner (Fig. 1j, k). These results therefore demonstrated that matrine restrained cell growth, migration, invasion and raised cell apoptosis and autophagy in HCC cells in vitro.

Next, we explored the effect of circ_0027345 on matrine-induced HCC cells, and circ_0027345 expression was first measured. The results of qRT-PCR analysis showed that circ_0027345 expression was evidently decreased in matrine-induced Huh-7 and HCCLM3 cells in a dose-dependent manner (Fig. 2a). As the concentration of 1.6 mg/mL matrine had a high inhibitory effect on circ_0027345 expression, this concentration was used for subsequent experiments. We then overexpressed circ_0027345 in matrine-treated Huh-7 and HCCLM3 cells by transfection of pcDNA-circ_0027345, and the overexpression efficiency was determined by qRT-PCR (Fig. 2b). Subsequently, CCK-8, Flow cytometry and Transwell results showed that overexpression of circ_0027345 could reverse the inhibitory effects of matrine on cell viability (Fig. 2c), cycle (Fig. 2g, h), migration (Fig. 2i) and invasion (Fig. 2j) in Huh-7 and HCCLM3 cells, and also alleviate the promoting impact of matrine on cell apoptosis (Fig. 2d). Besides, the decreased effect of matrine on Bcl-2 and the increased effects on the levels of Bax, cleaved-caspase 9 (Fig. 2e, f), Beclin 1 and LC3-II/LC3-I (Fig. 2k, l) could also be offset by overexpressing circ_0027345. Therefore, we concluded that matrine might inhibit the progression of HCC cells by down-regulating circ_0027345.

Increasing evidence has confirmed that circRNAs perform their functions primarily by sponging the target miRNAs [26]. As searched by StarBase v2.0 prediction tool, miR-345-5p contains the binding sites of circ_0027345 (Fig. 3a). Dual luciferase reporter assay was then carried out to confirm the prediction, and the results indicated that the luciferase activity was strikingly reduced in Huh-7 and HCCLM3 cells co-transfected with circ_0027345 WT and miR-345-5p compared to that cells co-transfected with circ_0027345 MUT and miR-345-5p (Fig. 3b, c). Moreover, overexpression of circ_0027345 silenced miR-345-5p expression (Fig. 3d, e). Meanwhile, miR-345-5p expression in matrine-induced Huh-7 and HCCLM3 cells was detected, and qRT-PCR data showed that miR-345-5p expression was elevated in a dose-dependent manner (Fig. 3f, g). These findings supported that circ_0027345 could act as a sponge for miR-345-5p.

To investigate whether circ_0027345 could regulate HCC progression by targeting miR-345-5p, pcDNA-circ_0027345 and miR-345-5p mimic (miR-345-5p) were co-transfected into the Huh-7 and HCCLM3 cells exposed to matrine. As shown in Fig. 4a, pcDNA-circ_0027345 and miR-345-5p were transfected into HCC cell lines successfully with high transfection efficiency. CCK-8 and Flow cytometry results revealed that overexpression of circ_0027345 increased cell viability (Fig. 4b) and inhibited cell apoptosis (Fig. 4c) in matrine-treated Huh-7 and HCCLM3 cells, while these effects were reversed by up-regulation of miR-345-5p. As expected, circ_0027345 overexpression enhanced the expression of Bcl-2 and decreased the expression levels of Bax and cleaved-caspase 9, however, miR-345-5p could overturn the trends (Fig. 4d, e). In addition, overexpression of circ_0027345 could accelerate cell cycle progression (Fig. 4f, g), promote cell migration (Fig. 4h) and invasion (Fig. 4i), and inhibit autophagy by down-regulating the levels of Beclin 1 and LC3-II/LC3-I (Fig. 4j, k), all of which could be attenuated by co-transfection of miR-345-5p in matrine-treated Huh-7 and HCCLM3 cells. All the findings implied that circ_0027345 regulated viability, apoptosis, cell cycle, migration, invasion and autophagy by binding to miR-345-5p in matrine-treated HCC cells.

In order to further study the downstream targets of circ_0027345/miR-345-5p axis, Targetscan bioinformatics analysis tool was used to predict the target mRNAs of miR-345-5p. As displayed in Fig. 5a, there were binding sites between miR-345-5p and the 3′UTR of HOXD3. Dual-luciferase reporter assay showed that the luciferase activity was reduced in Huh-7 and HCCLM3 cells co-transfected with miR-345-5p and HOXD3 3′UTR WT compared to miR-345-5p and HOXD3 3′UTR MUT co-transfected group, suggesting the interaction between miR-345-5p and HOXD3 (Fig. 5b, c). Additionally, the mRNA and protein levels of HOXD3 were enhanced by anti-miR-345-5p in Huh-7 and HCCLM3 cells (Fig. 5d, e). As expected, HOXD3 expression was dwindled in matrine-stimulated Huh-7 and HCCLM3 cells in a dose-dependent manner at mRNA and protein levels (Fig. 5f–h). More importantly, in matrine-treated Huh-7 and HCCLM3 cells, circ_0027345 overexpression could augment the mRNA (Fig. 5i) and protein levels of HOXD3 (Fig. 5j, k), while this effect could be weakened by overexpressing miR-345-5p. The above results supported that circ_0027345 acted as a ceRNA of miR-345-5p to regulate HOXD3 expression.

After confirming the targeted relationship between miR-345-5p and HOXD3, we further explored whether HOXD3 was involved in the regulation of miR-345-5p on HCC cell progression. As shown in Fig. 6a, b, co-transfection of si-HOXD3 counteracted the promotion effect of anti-miR-345-5p on mRNA and protein levels of HOXD3 in matrine-treated Huh-7 and HCCLM3 cells. As expected, repression of miR-345-5p potentiated cell viability (Fig. 6c) and impaired cell apoptosis (Fig. 6d), and these effects were restored by silencing HOXD3. Similarly, the effect of miR-345-5p knockdown on expression of apoptotic markers Bcl-2, Bax and cleaved-caspase 9 (Fig. 6e, f) could also be rescued by interfering with HOXD3. More than that, the facilitated effects of anti-miR-345-5p on cycle (Fig. 6g, h), migration (Fig. 6i) and invasion (Fig. 6j) were also abrogated by HOXD3 deficiency in matrine-stimulated Huh-7 and HCCLM3 cells. Meanwhile, the inhibitory effects of anti-miR-345-5p on levels of Beclin 1 and LC3-II/LC3-I were reversed by repression of HOXD3 (Fig. 6k, l). The results indicated that miR-345-5p could regulate cell development in matrine-stimulated HCC cells through targeting HOXD3.

According to the above results, matrine could hamper the development of HCC cells in vitro, we further evaluated whether matrine had the same inhibitory effect in vivo. To verify this guess, a mouse xenograft model was constructed. The results were shown in Fig. 7a, b, the tumors of mice treated with matrine had smaller volume and weight. Consistent with the above results in vitro, circ_0027345 expression (Fig. 7c) and the mRNA and protein levels of HOXD3 (Fig. 7e, f) were substantially decreased in tumor tissues of matrine-treated group, and miR-345-5p expression was markedly increased (Fig. 7d). The data in vivo proved that matrine hindered the tumor growth of HCC by up-regulating miR-345-5p and down-regulating circ_0027345 and HOXD3.