Aurora-B knockdown inhibits osteosarcoma metastasis by inducing autophagy via the mTOR/ULK1 pathway

Sixty-nine OS tissue samples were obtained by surgical biopsy from the First Affiliated Hospital of Nanchang University, China. Patients who had received radiotherapy or chemotherapy before undergoing biopsy were excluded. Donors included 29 females and 40 males, with a mean age of 26 years (range 5–72 years). All OS tissues were subject to pathological examination, and the expression of Aurora-B and LC3 was evaluated through IHC analysis. The clinical parameters are shown in Table 1; follow-up information was missing for 10 of these patients. Informed study protocols were completed in accordance with the declaration of Helsinki and “Guiding Opinions on the Treatment of Animals” in China. The medical ethics committee of the First Affiliated Hospital of Nanchang University has approved this experimental protocol (Jiangxi, China; NO. Y2019-126).

OS tissue samples were fixed in 4% paraformaldehyde for 20 min, embedded in paraffin, and sectioned to a thickness of ~ 3 µm. These slides were deparaffinized and rehydrated, and then treated with 0.2% Triton X-100PBS for 10 min and blocked with 3% hydrogen peroxide at room temperature for 20 min. These slides were autoclaved in 10 mM citric acid solution to enhance the antigen retrieval for 2 min. The samples were incubated with anti-Aurora-B (Abcam ab45145) and anti-LC3 (Cell signaling technology 2775) overnight at 4 °C. Samples were incubated with appropriate secondary antibodies for 30 min using Histostain Plus kits (Invitrogen, CA, USA). Pictures were captured with the microscope and determined by two pathologists blinded to the specimens.

HOS and 143B cell lines were obtained from the Type Culture Collection of the Chinese Academy of Sciences, Shanghai, China. All these cells were cultured in DMEM medium (Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco, 10099141C) with penicillin 100 U/mL and streptomycin 100 g/mL (Solarbio, Shanghai, China). All these cells were cultured at 37 °C with 5% CO2.

143B and HOS cells (1X104) were cultured in 35 mm cell-culture dish and were infected with 2X105 Lentivirus-Vector (MOI = 20) with Aurora-B (Lv-shAuroraB were inserted into lentivirus vector GV115 (GeneChem Co., Ltd., Shanghai, China), 5′-CCG GCTCCAAACTGCTCAGGCATAACTCGAGTTATGCCTGAGCAGTTTGGAGTTTTTG-3′) for 72 h and puromycin were incubated to the cells for screening. The efficacy of gene knockdown was determined by Western blot and qRT-PCR.

Human osteosarcoma cells were lysed in RIPA buffer containing protease inhibitor (cocktail and PMSF) for 15 min on ice. Protein concentrations were measured by BCA Protein Assay kit (Thermo Fisher Scientific). Total cell lysates were electrophoresed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) on 8%-15% gels and transferred onto PVDF membranes (Millipore). The membranes were blocked with 5% skim milk (BD) for 60 min at room temperature and incubated with primary antibodies overnight at 4 °C. Membranes were washed with 1X TBST 3 times and followed by incubation with secondary antibodies for 2 h. The immune complexes were visualized and measured with an ECL system (Bio-Rad, CA, USA) and the digital gel image analysis system (TANON). The primary antibodies included rabbit anti-human AuroraB (Abcam ab45145), anti-β-tubulin (Abcam ab179513), anti-SQSTM1/P62 (Cell signaling technology 5114), rabbit anti-LC3B (Cell signaling technology 2775), anti-p-mTOR(ser2448)(Cell signaling technology 2971), anti-mTOR(Cell signaling technology 2972), anti-AMPK (Cell Signaling Technology, 2532), anti-pAMPKα (Thr172) (Cell Signaling Technology, 2535), anti-ULK1 (Cell Signaling Technology, 8054) and anti-Pulk1 (Ser555) (Cell Signaling Technology, 5869), mouse anti-human GAPDH (Origene TA802519) and anti-MMP2 (Origene TA806846S). The secondary antibodies included HRP-conjugated Affinipure Goat Anti-Mouse IgG (H + L) (proteintech SA00001-1) and HRP-conjugated Affinipure Goat Anti-Rabbit IgG (H + L) (proteintech SA00001-2).

The total RNA was extracted from the osteosarcoma cell samples by Trizol (Solarbio, Shanghai, China), with the Revert Aid First Strand cDNA Synthesis Kit (Thermo fisher, USA) reverse-transcribed into cDNA. Q-PCR reactions were performed by the TaqMan™ Fast Advanced Master Mix (Thermo fisher, USA), GAPDH was used as a control. The Aurora-B sequence is Forward (5′ → 3′) AGAAGGAGAACTCCTACCCCT, Reverse (5′ → 3′) CGCGTTAAGATGTCGGGTG, GAPHD sequence is (Forward (5′ → 3′)) CCACCCATGGCAAATTCCATGGCA, Reverse (5′ → 3′) TCTAGACGGCAGGTCAGGTCCACC.

In order to track autophagosomes, HOS and 143B cells were transfected with lentiviral vectors harboring GFP-RFP-LC3 (GeneChem Co., Ltd., Shanghai, China) to obtain cells with stable expression of the GFP-RFP-LC3B protein. Then, these cells were treated with Lv-shAurora-B and control groups for the indicated duration. Images were captured by laser scanning confocal microscopy (ZEISS/LSM 800, Germany) to observe the fluorescence spots mark autophagosomes in these cells.

In brief, 143B and HOS cells were collected, fixed with 2.5% glutaraldehyde and encased. Ultrathin 60–80 nm sections were prepared with an Ultramicrotome (Leica UC7; Germany) and stained with uranyl acetate (15 min) and Reynolds lead citrate (15 min). The images were captured by a transmission electron microscope (HITACHI HT7700; Japan).

The Millipore 8 µm 24-well transwell chamber (Millipore) was used in the osteosarcoma cells migration and invasion. Cells (3 × 10⁴ cells per well) were loaded on FBS-free DMEM in the upper chamber of well coated with or without Matrigel (100 µl; 1:20 dilution; BD Biosciences), The lower chambers were filled with 500 μl complete medium mixed with 12 µm CQ or 2 µm MHY-1485 and a corresponding dose of PBS as control. After 12 h, the lower chambers mixtures were removed and replaced with 500 μl new complete medium. Another 12 h later, cells on the upper surface of the well were removed, and invasion cells passed through the well on the bottom were stained with 1% crystal violet. Cells in six randomly selected fields were counted and photographed (magnification 10X).

143B and HOS cells were grown to confluence in 6-well plates in the density of 5–8 × 10⁶ cells per well. When the cells were grown to 90% confluence, we used 20 µl pipette tip to scratch wounds through the center of the plate. The cells were washed three times with PBS to remove the floated cells and then incubated with 1% FBS MDEM mixed with 12 µm CQ or 2 µm MHY-1485 and a corresponding dose of PBS as control at 37 °C for 8 h. The mixtures were removed and replaced by new 1% FBS MDEM. Images were captured at different time points (0 and 24 h), and the migration distance was measured by ImageJ compared with the time zero.

All experimental protocols were approved by the Institutional Animal Care and Use Committee of Nanchang University (Jiangxi, China; NO. Y2019-126). We purchased female BALB/C nude mice at 6 weeks of age from the Nanjing Biomedical Research Institute of Nanjing University (NBRI, Nanjing, China); mice were housed in the SPF (Specific Pathogen Free) Transgenic Animal Facility of Nanchang University. The protocol for generation of a spontaneous metastasis/orthotopic osteosarcoma mouse model was applied as reported in our previous study [20]. The mice were randomly divided into four groups (Ctrl group, AZD2811 groups, ADZ2811 + 3BDO group, and AZD2811 + CQ groups; n = 6 in each group). Drugs (normal saline, AZD2811 (150 mg/kg), ADZ2811 (150 mg/kg), and 3BDO (80 mg/kg) mixture, AZD2811 (150 mg/kg), and CQ (80 mg/kg) mixture) were separately administered to mice via intraperitoneal injection, twice a week. After 6 weeks, the tumors were dissected and fixed in 10% formalin. The lung tissues were dissected to evaluate pulmonary metastasis by optical microscope using the Vivo imaging system (Berthold LB983; Germany). The tissues were fixed in 10% formalin for further detection.

To investigation the relationship of Aurora-B expression with prognosis in OS. We examined the expression of Aurora-B by immunohistochemical staining and collected the follow-up information from OS patients. We found that the expression of Aurora-B protein is closely correlated with Enneking stage (Table 1; Fig. 1a, b) (P < 0.01). Moreover, Kaplan–Meier analysis showed that high levels of Aurora-B expression were negative correlated with poor overall survival (Fig. 1c) (P < 0.01) and lower metastasis-free survival (Fig. 1d) (P < 0.01) in these OS patients. These results collectively indicate that the expression of Aurora-B protein was negatively correlated with prognosis in OS.

To investigate the correlation between Aurora-B and autophagy, we first applied immunohistochemical staining to examine the Aurora-B and LC3 expression in OS tissues. In the tissues with Aurora-B-positive expression, LC3 protein was poorly expressed, and the opposite was observed in the tissue samples that tested negative for Aurora-B expression (Table 1; Fig. 2a, b) (P < 0.05). These results suggested that a potential correlation between Aurora-B and autophagy in OS. Further investigate the effect of Aurora-B on autophagy in OS cells, we knocked down endogenous Aurora-B with shRNA (short-hairpin RNA) in 143B and HOS cells by lentiviral infection (Fig. 2c, d). The RFP-GFP-LC3 fusion assay was performed to observe the formation of autophagosomes and autolysosomes in OS cells. The number of yellow LC3 puncta (representing autophagosomes) and red LC3 puncta (representing autolysosomes) were both higher in Aurora-B silenced cells than in cells infected with scrambled lentivirus, which represented the negative control (Fig. 2e). Similar results were observed by transmission electron microscopy (Fig. 2f). Furthermore, the ratio of LC3-II to LC3-I was significantly increased, whereas p62 expression was decreased, in cells with Aurora-B-knockdown relative to that in cells infected with scrambled lentiviruses (Fig. 2g). The results indicate that inhibition of Aurora-B enhanced autophagy in OS cells. Furthermore, to distinguish whether the accumulation of LC3 is due to the induction of autophagy or the depletion of autophagy flux, cells were treated with the specific autophagy inhibitor chloroquine (CQ), which inhibits autophagosome-lysosome fusion, thus preventing LC3 degradation. As shown in Fig. 1f, the LC3 II protein level was significantly increased following treatment with CQ in both control and Aurora-B-silenced cells compared with that in cells that were not treated with CQ (Fig. 2h). These results indicate that Aurora-B knockdown more likely enhances autophagic flux than inhibits autophagy in OS cells.

The mTOR/ULK1 pathway is known to regulate autophagy, and the inactivation of this pathway has been shown to enhance autophagy levels [21, 22]. AMPK is a critical sensor of stress and energy metabolism that negatively regulates mTOR, activating autophagy initiation factor ULK1 and promoting autophagic flux [23, 24]. To explore the molecular mechanisms involved in the enhancement of autophagy by Aurora-B knockdown, we examined the expression of proteins involved in the mTOR/ULK1 pathway in Aurora-B-knockdown 143B and HOS cells. Interestingly, as shown in Fig. 3a, the phosphorylation levels of AMPK and ULK1 were upregulated and those of mTOR were downregulated in Aurora-B knockdown 143B and HOS cells, which was consistent with the observed changes in total AMPK, ULK1, and mTOR levels. These results suggest that Aurora-B knockdown enhances autophagy by decreasing mTOR/ULK1 pathway activity. Recent evidence has additionally demonstrated that the suppression of the malignant phenotype of OS cells by Aurora-B silencing might be achieved via autophagy-mediated activation of the mTOR/ULK1 signaling pathway. Thus, to further investigate the role of the mTOR/ULK1 pathway during Aurora-B silencing-induced autophagy, we investigated the expression of the autophagy-related protein p62 after treatment with MHY-1485, a well-known mTOR activator [25], or not, in Aurora-B-knockdown or scrambled OS cells. In the Aurora-B-knockdown OS cells, MHY-1485 treatment activated mTOR/ULK1 signaling and increased p62 levels. In the scrambled groups, after treatment with MHY-1485, the expression of p62 and phosphorylation of mTOR increased (Fig. 3b). These results reveal that inhibition of Aurora-B induced autophagy in OS cells via decreasing the mTOR/ULK1 pathway.

Autophagy is known to be involved in tumor progression [26, 27]. In order to elucidate the effect of Aurora-B inhibition induced autophagy on the metastasis of OS cells, wound healing, transwell migration and invasion assays were performed to detect migration and invasion ability of OS cells. We surprisingly found that Aurora-B inhibition significantly decreased the motility and invasive ability of 143B and HOS cells. After treatment with CQ, the attenuated ability could be reversed. (Fig. 4a–f). These findings indicate that autophagy plays an essential role in the process of Aurora-B affecting the ability of motility and invasive in OS cells. Autophagy is reported to inhibit cell migration by accelerating the degradation of MMP family proteins [26, 27], therefore, we sought to determine the MMP2 protein levels in Aurora-B-knockdown OS cells with or without CQ treatment. Results showed that the suppression of MMP2 expression due to Aurora-B knockdown could be restored by CQ treatment (Fig. 4g). These findings demonstrate that inhibitio of Aurora B suppresses the migration and invasion of OS cells by stimulating autophagy.

To confirm whether the mTOR/ULK1 pathway contributes to Aurora-B-induced suppression of the malignant phenotype in OS cells, we performed wound healing, transwell migration, and invasion assays to evaluate the migration and invasion ability of Aurora-B-knockdown OS cells treated with MHY1485 or not. The results revealed that the inhibition of migration and invasion ability induced by Aurora-B knockdown could be reversed by MHY1485 treatment (Fig. 5a–f). And it indicates that activation of the mTOR/ULK1 pathway reverses the induction of Aurora-B knockdown-induced autophagy, which inhibits migration and invasion in OS cells.

In order to investigate how the inhibition of Aurora-B affects the metastasis of OS, we established an orthotopic xenograft model using the luciferase-labeled 143B cell line (Fig. 6a). Bioluminescent imaging analysis revealed that the number of metastatic foci in the lung was significantly decreased when treated with AZD2811, a specific Aurora-B inhibitor in these mice models. Furthermore, 3BDO, a new mTOR activator, or CQ injection reversed these effects. H&E staining and lung anatomy confirmed this result (Fig. 6b–d). These results suggested that Aurora-B knockdown inhibits OS metastasis via mTOR-mediated autophagy in vivo.