Upregulation of AMPK by 4-O-methylascochlorin promotes autophagy via the HIF-1α expression

2.1 Cells culture and materials

Human non-small lung cancer cell lines A549, H1793, and H23 cells were obtained from the American Type Culture Collection (Rockville, MD, USA). Cells were cultured in RPMI 1640 (Thermo Scientific, Logan, UT, USA), containing 10% of fetal bovine serum and 1% of antibiotic mixture. These were incubated at 5% of CO₂, 37°C. MAC was supplied Chugai Pharmaceutical (Tokyo, Japan).

2.2 Cell viability assay

Human non-small lung cancer cell lines were seeded in 96-well plates at 2 × 10⁴ cells/well in RPMI 1640 media, and allowed to attach for 24 hours. New media containing different concentration treat of MAC and incubate for 24 hours. 3-[4,5-Dimethylthiazol-2-yl]-2,5-diphenyltetra-zolium bromide (MTT) (Roche Molecular Biochemicals, Indianapolis, IN, USA) was added to each well. The amount of formazan deposits was quantified according to the supplier's protocol after 4 hours of incubation with MTT assay in 5% of CO₂ at 37°C.

2.3 Morphology

Cells were cultured in 60 mm dish and incubated until they reached 60% confluence. Then, media were changed fresh RPMI 1640 media and treat MAC for 12 hours, and cells were photographed.

2.4 Western blot analysis

Cells lysates were prepared by lysis buffer [50 mmol/L Tris, 150 mmol/L NaCl, 5 mmol/L EDTA, 1 M DTT, 1% NP-40, 100 μmol/L phenylmethylsulfonyl fluoride, 20 μmol/L aprotinin, and 20 μmol/L leupeptin, adjusted to (pH 8.0)]. Total proteins were electrotransferred to Immobilon-P membranes (Millipore, Billerica, MA, USA). Detection of specific proteins was carried out with an enhanced chemiluminescence Western blotting kit, following the manufacturer's instructions (Amersham, Piscataway, NJ, USA). To determine the activations of HIF-1α nuclear extracts of cells were subjected to describe as follows. Cells were suspended in tubes with 0.4 mL of lysis buffer [10 mmol/L HEPES (pH 7.9), 10 mmol/L KCl, 0.1 mmol/L EDTA, 0.1 mmol/L EGTA, 1 mmol/L DTT, 0.5 mmol/L PMSF, 2 μg/mL leupeptin, and 2 μg/mL aprotinin]. Cells were then allowed to swell on ice for 5 minutes and 25 μL of 10% Nonidet P-40 was added. Homogenates were centrifuged at 4°C for 1 minute at 12 204 g. The nuclear pellets were resuspended in 50 μL of ice-cold nuclear extraction buffer containing [20 mmol/L HEPES (pH 7.9), 0.4 M NaCl, 1 mmol/L EDTA, 1 mmol/L EGTA, 1 mmol/L DTT, 1 mmol/L PMSF, 2 μg/mL leupeptin, and 2 μg/mL aprotinin], and incubated on ice for 15 minutes with intermittent mixing. Nuclear extracts were then centrifuged at 4°C for 5 minutes at 10 000 rpm and supernatants were either used immediately. Specific antibodies were purchased from Cell Signaling Technology (Danvers, MA, USA) for LC3, p62, ATG7, transcription factor EB (TFEB), p70S6K, AMPK, phosphorylation of AMPK and Santa Cruz Biotechnology (Dallas, TX, USA) for BNIP3, Beclin1, HIF-1α, phosphorylation of p70S6K, and phosphorylation of 4EBP1.

2.5 Acridine orange staining

Acridine orange (1 μg/mL) was added to previously treat-cells for 1 hour at 37°C and fixed with 4% of Formaldehyde for 1 minute. Then, cells were washed three times with PBS containing 0.05% v/v Tween-20 and 1% w/v BSA. Coverslips were mounted by using fluorescence mounting medium. Images of cells were observed using confocal microscope.

2.6 siRNA transfection

For transfection experiments with siRNA directed against AMPK, HIF-1α, and ATG7 the medium for A549 and H23 cell was changed to fresh RPMI, and the cell was transfected for 24 hours with 10 μmol/L of siRNA by Trans IT-TKO (Mirusbio, Madison, WI, USA) according to the manufacturer's instruction. After 24 hours, the medium was changed to fresh serum-free RPMI, cell was treated with TPA or 24 hours and total protein in the cell was extracted for Western blot.

2.7 Immunofluorescence microscopy

Cells were cultured and treated on poly-l-lysine-coated coverslips before being fixed in 100% ethanol for 1 minute at room temperature. After two washes with PBS, cells were permeabilized with 0.2% w/v Triton X-100 in PBS for 5 minutes. After two washes with PBS, cells were blocked with 10% v/v normal goat serum in PBS for 1 hour in a humidified chamber. Cells were then incubated with primary antibodies LC3 (1:100, diluted in PBS containing 2% v/v normal goat serum) for 1 hour. The cells were then washed three times before being incubated with FITC, tetramethylrhodamine isothiocyanate-conjugated secondary antibody rabbit immunoglobulin (1:100, diluted in PBS containing 2% v/v normal goat serum) for 1 hour, and five times were washed with PBS containing 0.05% v/v Tween-20 and 1% w/v BSA. Then, cells were incubated with DAPI for 3 minutes at 37°C incubation. Finally, the cells were washed three times with PBS. Coverslips were mounted with 90% of glycerol and sealed with Mounting Media. Slides were examined and scanned on a fluorescence microscope.

2.8 GFP-LC3 assay

After transient transfection with GFP-LC3 plasmid (EGFP-LC3 plasmid was a gift from Karla Kirkegaard; Addgene plasmid #11546), A549 cells were cultured on poly-l-lysine-coated coverslips. After 24 hours, cells were then transfected with 3 μg plasmid of GFP-LC3 using transfection reagent Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. After 24 hours medium incubation, cells were treated with MAC, CoCl₂, and EBSS for 24 hours and fixed with 4% of formaldehyde for 5 minutes and three times washed with PBS. Then, cells were incubated with DAPI for 3 minutes at 37°C. Finally, the cells were washed five times with PBS. Coverslips were mounted by using fluorescence mounting medium. Slides were examined and observed of GFP-LC3 puncta on a fluorescence microscope.

2.9 Statistical analysis

All in vitro results are representative of at least three independent experiments performed in triplicate. *P < 0.05, statistically significant between experimental and control values. Significance of differences between experimental and control values was calculated using ANOVA with Newman–Keuls multicomparison test.

3.1 Effects of ASC and ASC derivatives on autophagosome formation in A549 lung cancer cells

The typical morphological feature of autophagy is the formation of autophagosomes, which fuse with lysosomes and expose cargo to lysosomal enzymes, causing degradation of the contents.19 The effect of ASC and its derivatives AS-6 and MAC on autophagosome formation (Figure 1A) was investigated by examining morphological changes of A549 lung cancer cells. Incubation of A549 cells with 20 μmol/L ASC, AS-6, or MAC for 12 hours induced vesicle formation (Figure 1B). LC3 and p62 (also known as SQSTM1) are the key factors for the autophagy.20 Accumulation of LC3-II allows assessment of autophagy initiation, whereas p62 is degraded by autophagy. The effect of the three chemicals on autophagosome formation was investigated by western blot analysis of LC3-I/II and p62 expression. As shown in Figure 1C, AS-6 and MAC but not ASC upregulated LC3-II. AS-6 and MAC increased the LC3-II/LC3-I ratio by approximately 3- and 10-fold, respectively, whereas ASC had no effect on the LC3-II/LC3-I ratio. MAC also completely degraded p62. These data suggest that MAC induces autophagy more effectively than the other ASC derivatives. Therefore, the mechanism of MAC-induced autophagy was analysed in this study.

3.2 MAC triggered autophagosome formation in various lung cancer cell lines

The effects of MAC on autophagy were examined in three lung cancer cell lines: A549, H1793, and H23. Before investigating the pharmacological effect of MAC on autophagy, the cytotoxic effect of MAC was examined using the MTT assay. The results showed no obvious reduction in A549 cell viability after treatment with MAC at doses lower than 10 μmol/L. MAC at 20 μmol/L decreased cell viability by approximately 20%, and a cytotoxic effect was observed at 30 μmol/L in H1793 and H23 cells (Figure 2A). Based on these results, a concentration of <20 μmol/L MAC was used in subsequent experiments. The effect of MAC treatment on the morphology of the three lung cancer cell lines was examined. As shown in Figure 2B, MAC increased vesicle formation in a dose-dependent manner (at 10 and 20 μmol/L). To detect the development of autophagosome vesicles, cells were stained with acridine orange. A549, H1793, and H23 cells were cultured in the presence of different concentrations of MAC for 24 hours and then stained. The results showed that MAC triggered autophagosome vesicle formation in lung cancer cells (Figure 2C), suggesting that MAC induces autophagy by promoting the formation of autophagosomes.

3.3 MAC-induced autophagy-related gene expression

To determine whether autophagy was induced by MAC in lung cancer cells, the expression of LC3-I/II, a key autophagy marker protein, was measured by western blotting. As shown in Figure 3A and B, MAC upregulated LC3-II in a dose- and time-dependent manner in A549 and H1793 cells. In H23 cells, MAC significantly increased the expression of both LC3-I and II. To confirm the relationship between LC3-II expression and autophagosome formation in response to MAC treatment, immunofluorescence was performed using an LC3 antibody. Cells treated with MAC had higher expression of LC3 than untreated cells (Figure 3C). Morphological analysis showed that MAC promoted the accumulation of LC3 around the nucleus. These results indicate that MAC-induced autophagy by upregulating LC3 expression. The human Beclin1 gene is the mammalian homolog of the yeast ATG6/Vps30 gene, which plays an important role in autophagosome formation and autophagy activation.21 ATG proteins have been studied for their roles in autophagosome formation and maturation.22 We showed that MAC upregulated Beclin1 and ATG7 in a time-dependent manner (Figure 3D). These results suggest that MAC promotes autophagy by upregulating LC3-II, Beclin1, and ATG7 protein expression in lung cancer cells. We next evaluated changes in p62 and TFEB expression. p62 is selectively bind to LC3 and degraded by lysosome in the final stages of the autophagy process.23 The expression of p62 was increased at 6 hours after treatment with MAC. However, it began to decline at 12 hours compared with those in untreated (Figure 3E). TFEB, which is a regulator of autophagy to lysosomal biogenesis,24 also began to increase at 12 hours after MAC treatment. These results suggest that MAC promotes autolysosome by regulating p62 and TFEB protein expression in lung cancer cells.

3.4 MAC activated autophagy through the AMPK signalling pathway

AMPK induces autophagy by stimulating the ULK1 and Beclin1 proteins mTOR, a downstream regulator of AMPK, stimulates protein synthesis and inhibits autophagy by disrupting the interaction between ULK1 and AMPK.25 We therefore examined the effect of MAC on the phosphorylation of AMPK and mTOR, the major direct signalling pathways of autophagy. MAC increased the phosphorylation of AMPK and tuberous sclerosis complex 2 (TSC2), an AMPK downstream regulator, in a time-dependent manner in lung cancer cells (Figure 4A). However, MAC had no effect on the levels of phosphorylated mTOR compared with those in control H1793 and H23 cells. The phosphorylation of the regulatory associated protein mTOR (Raptor) at serine 792, which is phosphorylated by AMPK and leads to the inhibition of mTOR complex 1 (mTORC1), induced by MAC was therefore confirmed. MAC inhibited the phosphorylation of P70S6K and 4EBP1, the downstream factors of mTORC1, in a time-dependent manner. These results indicate that mTORC1 downstream signalling was affected by MAC in a manner independent of the regulation of mTOR phosphorylation, and that MAC-induced autophagy is involved in the AMPK/mTORC1 signalling pathway.

To confirm that MAC induced autophagy through the AMPK signalling pathway, LC3 expression was analysed in A549 cells treated with AICAR (AMPK activator). AICAR upregulated both LC3I and LC3II, whereas MAC-induced LC3II expression compared with AICAR (Figure 4B). Transfection of cells with AMPK siRNA suppressed AMPK phosphorylation and expression and reduced MAC induced LC3-II expression (Figure 4C). These results indicate that MAC-induced autophagy is associated with the AMPK signalling pathway. However, the MAC-induced LC3-II expression was not completely suppressed in AMPK siRNA transfected cells, which suggest that the other signalling pathway might be involved in the MAC-regulated autophagy.

3.5 MAC-induced autophagy was related to hypoxia in lung cancer cells

As HIF-1 signalling responds to oxygen deficiency, AMPK, an energy deficiency sensor that responds to stressful environments, may play a fundamental role in cancer control and longevity.26 Therefore, we hypothesized that the MAC-induced dysregulation of LC3-II expression in AMPK siRNA transfected cells contributed to hypoxia-mediated autophagy. To confirm this hypothesis, the effect of MAC on hypoxia-dependent proteins was examined in lung cancer cells. Cells were treated with various concentrations of MAC for 6 hours, and the nuclear translocation of HIF-1α was analysed (Figure 5A). MAC significantly increased HIF-1α expression in nuclear extracts of the three lung cancer cells. Bcl-2/E1B-nineteen kilodalton interacting protein (BNIP3) plays an important role in the hypoxia-induced autophagic death of cancer cells.27, 28 The mechanism by which hypoxia stimulates the expression of BNIP3 most likely involves the direct activation of HIF-1α.29 MAC upregulated BNIP3 expression in a time-dependent manner (Figure 5B). These results suggest that MAC-stimulated HIF-1α and BNIP3 expression is associated with the autophagic activity of MAC.

To determine the relationship between autophagy and HIF-1α expression, autophagy in the presence of MAC was compared with that in the presence of CoCl₂ (mimic hypoxia) and EBSS media (amino acid deprivation). Microscopic analysis of cell morphology showed that MAC, CoCl₂, and EBSS increased vesicle formation. Acridine orange staining showed that MAC, CoCl₂, and EBSS induced autophagosome vesicle formation. In transfected A549 cells with GFP-LC3, moreover, MAC, CoCl₂, and EBSS increased the formation of GFP-LC3 puncta, a marker that allows for the assessment of autophagic flux (Figure 5C). Autophagosome formation was increased by hypoxia mimic, suggesting that HIF-1α expression induces autophagy.

Next, the protein expression of autophagy marker genes was analysed in cells cultured in under MAC, CoCl₂, and EBSS-containing media. MAC and CoCl₂ upregulated LC3-II, Beclin1, ATG7, and BNIP3 in the different lung cancer cells (Figure 5D). However, EBSS-containing media weakly induced LC3-II expression and had no effect on the other factors. MAC and CoCl₂ upregulated HIF-1α expression, whereas EBSS treatment had no effect on HIF-1α expression (Figure 5E). These results indicated that MAC induces HIF-1α expression and autophagy similar to the hypoxia environment.

3.6 MAC promotes autophagy by regulating AMPK/HIF-1α expression in lung cancer cells

To determine whether MAC-induced autophagy by affecting HIF-1α expression in lung cancer, the effect of YC-1 (HIF-1α inhibitor) on autophagy-related gene expression was examined. YC-1 downregulated BNIP3 in A549 and H23 cells, suggesting that HIF-1α regulates BNIP3 expression in the normoxia condition. MAC upregulated the expression of both LC3 and BNIP3 in the same manner. YC-1 inhibited the MAC-induced LC3-II/LC3-I ratio and BNIP3 expression (Figure 6A). To determine whether MAC-induced autophagy was mediated by the regulation of HIF-1α expression, cells were transfected with siRNA, targeting HIF-1α. As shown in Figure 6B, MAC-induced HIF-1α expression in the control siRNA transfected cells, whereas this effect was suppressed by HIF-1α siRNA. Transfection with HIF-1α siRNA substantially abolished the effect of MAC on upregulating autophagy-associated proteins including LC3, ATG7, and BNIP3. These results suggest that MAC induces autophagy by upregulating HIF-1α expression.

Previously, we showed that MAC stabilizes HIF-1α by activating AMPK. To determine whether MAC-induced autophagy mediated by the regulation of HIF-1α expression was associated with AMPK activation, the regulatory effect of AMPK on HIF-1α, and BNIP3 expression was examined using AICAR and AMPK siRNA. As shown in Figure 6C and D, AICAR upregulated HIF-1α and BNIP3 in a dose-dependent manner. In AMPK siRNA transfected cells, MAC weakly upregulated HIF-1α and BNIP3. These results indicated that MAC induces autophagy through AMPK-dependent and AMPK-independent mechanisms of MAC-induced HIF-1α signalling pathway.

To elucidate the role of selective MAC-induced autophagy, autophagy was inhibited using the ATG7 siRNA. After the exposure of cells to MAC (20 μmol/L) in the control siRNA- or ATG7 siRNA-transfected cells, the ATG7 protein expression and cytotoxic effects were assessed. Transfection of cells with ATG7 siRNA significantly suppressed the MAC-induced ATG7 expression in A549 cells (Figure 6E). Inhibition of basal autophagy by ATG7 siRNA was not change cellular viability. MAC at 20 μmol/L in the control siRNA-transfected cells decreased cell viability by approximately 19%, whereas treatment of cells with MAC in the ATG7 siRNA-transfected cells led to approximately 28% decrease in cell viability (Figure 6F). These results indicate that inhibition of autophagy sensitizes lung cancer cells to MAC-induced cytotoxicity, and the combination treatments with autophagy inhibitors might improve the cytotoxic effects of MAC alone.