Formosanin C suppresses cancer cell proliferation and migration by impeding autophagy machinery

Abstract

Formosanin C (FC) is a natural compound extracted from Paris formosana Hayata with anticancer activity. FC induces both autophagy and apoptosis in human lung cancer cells. FC‐induced depolarization of mitochondrial membrane potential (MMP) may trigger mitophagy. In this study, we clarified the effect of FC on autophagy, mitophagy, and the role of autophagy in FC‐related cell death and motility. We found FC caused the continuous increase of LC3 II (representing autophagosomes) from 24 to 72 h without degradation after treatment of lung and colon cancer cells, indicating that FC blocks autophagic progression. In addition, we confirmed that FC also induces early stage autophagic activity. Altogether, FC is not only an inducer but also a blocker of autophagy progression. Moreover, FC increased MMP accompanied by overexpression of COX IV (mitochondria marker) and phosphorylated Parkin (p‐Parkin, mitophagy marker) in lung cancer cells, but no colocalization of LC3 with COX IV or p‐Parkin was detected under confocal microscopy. Moreover, FC could not block CCCP (mitophagy inducer)‐induced mitophagy. These results imply that FC disrupts mitochondria dynamics in the treated cells, and the underlying mechanism deserves further exploration. Functional analysis reveals that FC suppresses cell proliferation and motility through apoptosis and EMT‐related pathway, respectively. In conclusion, FC acts as an inducer as well as a blocker of autophagy that results in cancer cell apoptosis and decreased motility. Our findings shed the light on the development of combined therapy with FC and clinical anticancer drugs for cancer treatment.

1. INTRODUCTION

Lung cancer is the most common cause of cancer death worldwide, with an estimated 1.8 million deaths each year. ¹ More than half of patients diagnosed with lung cancer died within 1 year, and the 5‐year survival is around 17.8%. ² Lung cancer is classified into small‐cell lung carcinoma (SCLC) and non‐small‐cell lung carcinoma (NSCLC). NSCLC accounts for about 85% of lung cancers. Among NSCLCs, the most common type of lung cancer is adenocarcinoma which comprises around 40% of all lung cancers worldwide. The risk factors for lung cancer include smoking, air pollution, and genetic susceptibility. The most commonly used treatments for lung cancer are surgery, chemotherapy, and radiotherapy. Because of the high incidence and mortality of lung cancers, it is important to explore novel therapies.

Formosanin C (FC), also called Paris Saponin II, is a naturally occurring compound extracted from Paris formosana Hayata (Liliaceae). It has an anti‐inflammatory effect and is used as a traditional Chinese herbal medicine for snake bite. ³ It has been reported that FC has antitumor activity in diverse human cancer cells, including lung, ovarian, liver, and colorectal cancers. ⁴ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ FC suppressed tumor cells by inducing apoptosis, inhibition of metastasis, or cell cycle arrest. ⁴ , ⁵ , ⁶ , ⁷ Therefore, FC is a potential antitumor drug through its effects on various functions of tumorigenesis.

Autophagy is a cellular degradative pathway for clearing superfluous materials, including microbes, damaged proteins, microRNA, and organelles, to maintain organelle quality and cellular homeostasis. ¹⁰ , ¹¹ , ¹² , ¹³ Autophagy also participates in the cell's metabolic machinery for recycling and energy production. ¹⁴ Autophagy is induced in response to various stresses, including pathogen infection, starvation, hypoxia, and oxidative stress. The progression of autophagy designated as autophagic flux includes (i) induction, (ii) nucleation, (iii) phagophore expansion, autophagosome formation, fusion with the lysosome, and (iv) degradation. Dysfunction of the autophagy machinery leads to diverse diseases including cancers. Autophagy plays a dual role in tumor development, which depends on various factors, including the type of cancer involved and the stage of tumorigenesis. At the initiation stage of tumor development, autophagy may function as a tumor suppressor by recruiting and degrading the oncogenic proteins, damaged DNA, and reactive oxygen species (ROS) to inhibit tumor progression. ¹² In contrast, in the late stages of tumor progression, autophagy switches to promoting tumors by helping cancer cells overcome stressors, including hypoxia and nutrient deficiency. Nonetheless, autophagy is a potential target in anticancer therapy. ¹²

The mitochondria are responsible for energy production and cell apoptosis. Therefore, the homeostasis of mitophagy is crucial for cell survival. Mitophagy is defined as the selective recruitment and degradation of mitochondria by autophagy. Damaged and dysfunctional mitochondria are cleared by mitophagy to maintain the quality of mitochondria as well as to prevent cell apoptosis. One such pathway is PINK1/Parkin‐mediated mitophagy. ¹⁵ Similar to autophagy, mitophagy also plays a dual role in tumorigenesis. Mitophagy regulators are reported to be aberrant in many human cancers, including bladder, lung, ovarian, and breast cancer. ¹⁶ For example, Parkin gene deletion has been detected in bladder, lung, and ovarian cancers. In a mouse model, downregulated mitophagy promotes tumorigenesis. In contrast, mitophagy removes p53 from the mitochondria to maintain the population of hepatic cancer stem cells. ¹⁷ Three‐negative breast cancer cells showed increased invasion and migration through the induction of mitophagy in the presence of TNF‐α. ¹⁸ Therefore, regulation of mitophagy may be an alternative strategy for cancer therapy.

Zhang et al. reported that FC induces apoptosis in lung cancer cells through the induction of autophagy. ⁶ In contrast, we reveal herein that FC is a blocker of autophagy and mitophagy. Moreover, we disclose how FC suppresses lung cancer tumorigenesis through autophagy.

2. MATERIALS AND METHODS

2.1. Cell and culture

Lung carcinoma A549 cells were maintained in Dulbecco's modified Eagle's medium (DMEM; Gibco) containing 10% fetal bovine serum (FBS; Gibco) in a humidified incubator at 37°C with 5% CO₂. Colon cancer SW480 cells were cultured in DMEM medium (Invitrogen) in a humidified incubator at 37°C without CO₂.

2.2. MTT assay

Cells in 96‐well plates were treated with FC (ALB‐RS‐1833, ALB Technology) for 24, 48, and 72 h. MTT [3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide] (Merck KGaA) solution (0.05 mg/mL in DMEM) was added to each well and the plates were incubated at 37°C for 3 h. Then the cells were incubated with DMSO for 15 min. The absorbance was examined on a 96‐well multiscanner autoreader (Thermo Fisher Scientific) at 540 nm.

2.3. Cell death analysis

Cells were seeded in a 96‐well plate and treated with different doses of FC for 0, 24, and 48 h. About 50 mL of the cultured medium was combined with 50 μL reaction mixture of the Cytotoxicity Detection Kit (11644793001, Roche) at RT for 10 min. Cells were cultured in a medium and treated with Triton X, which served as the positive control for 100% cell death. The absorbance of the solution at 490 nm was detected by an ELISA reader.

2.4. Apoptosis analysis

Cells were seeded in a six‐well plate and treated with different doses of FC for 24 h. Cells were collected by centrifugation and suspended in phosphate‐buffered saline (PBS). About 5 μL of Annexin V‐FITC and 5 μL of propidium iodide (K101, BioVision) were added to the cells and incubated at room temperature (RT) for 5 min in the dark.

2.5. Western blotting

Cells were lysed in lysis buffer (1 mL RIPA, 4 μL Na₃SVO₄ [0.5 M], 20 μL EGTA [0.1 M], 10 μL PMSF [0.1 M], 5 μL Aprotinin [2 mg/mL], 5 μL Leupeptin [2 mg/mL], and 2 μL EDTA [0.5 M]), and protein concentration was quantified by Bradford Protein Assay Kit (Thermo Fisher Scientific). The cell lysate was separated by 12% SDS‐polyacrylamide gel followed by transfer to a PVDF membrane (Millipore) in transfer buffer (0.025 M Tris‐Base, 0.2 M Glycine) at 100 V for 1.5 h using an electroblotter (Amersham Pharmacia Biosciences Corp.). The transferred membrane was blocked with 5% milk in PBST buffer (0.137 M NaCl, 0.0023 M KCl, 0.01 M Na₂HPO₄, pH 7.4, and 2.5 mL Tween20 in 1000 mL PBS buffer) for 1 h at RT. The membrane was incubated with the primary antibody (LC3:1:1000, MBL, PM036; GAPDH:1:150,000, GeneTex, GTX109639; p‐Parkin(Ser65):1:1000, biorbyt, orb312554; Parkin:1:1000, Elabscience, E‐AB‐63490; COX‐4: 1:1000, Cell Signaling Technology, #4844; Calreticulin:1:1000, GeneTex, GTX111627; ATG5 1:1000, abcam, ab108327; Vimentin:1:1000, GeneTex, GTX100619; E‐cadherin:1:1000, GeneTex, GTX100443; N‐cadherin, 1:1000, arigo, ARG22587) at 4°C overnight. The membrane was then washed with PBST three times and incubated with secondary antibody anti‐rabbit (Invitrogen, 1:1000, G‐21234) or anti‐mouse (Invitrogen, 1:1000, G‐31430) HRP‐conjugated Secondary Antibody at RT for 1 h followed by washing with PBST three times. The membrane was then rinsed with enhanced chemiluminescence (ECL; WBKLS0500; Millipore) and the intensity of protein expression was measured by BioSpectrum AC (101‐206‐009; UVP).

2.6. Immunofluorescent staining

Cells were seeded onto a chamber slide and treated with different drugs for specific times. The sample was fixed in 3.7% formaldehyde for 30 min, followed by PBS wash three times. The sample was then incubated in 0.1% Triton X‐100 in PBS at RT for 30 min. After washing with PBS, cells were incubated with anti‐LC3 antibody (MBL, M152‐3) at 4°C overnight. Cells were incubated with anti‐rabbit Alexa Fluor® 594‐conjugated secondary antibody (1:200, Invitrogen, A‐11012) and anti‐mouse Alexa Fluor® 488‐conjugated secondary antibody (1:200, Invitrogen, A‐11001) at RT for 1 h. Hoechst (5 mg/mL; Sigma‐Aldrich) at a dilution of 1:1000 in PBS was used for nucleus staining. The fluorescent signal was detected using a fluorescence microscope.

2.7. JC‐1 staining

Cells were treated with different doses of FC for 24 h and stained with JC‐1 dye (T3168, Thermo Fisher Scientific). Green and red fluorescence were detected under flow cytometry (CytoFLEX, Beckman).

2.8. Transwell™ assay

Cells were treated with 2 or 4 μM FC in the medium with 10% FBS for 24 h. Cells were seeded in the upper chamber of the Transwell® (Corning) without FBS, and the medium containing 10% FBS was added to the lower chamber. Cells were fixed with 3.7% paraformaldehyde for 20 min and stained with 0.5% crystal violet (V5265, Sigma‐Aldrich) for 30 min. Migrated cells were visualized under microscopy.

2.9. Statistical analysis

Data are presented as the mean ± standard deviation (SD) values (error bars). Differences between the experimental and control groups were analyzed by two‐tailed Student's t test. p Values <0.05 were considered statistically significant: *p < 0.05, **p < 0.01, and ***p < 0.001.

3. RESULTS

3.1. FC suppresses human lung cancer cell growth by inducing apoptosis

Lung cancer A549 cells were treated with different doses of FC for various times to investigate FC inhibition of cancer cell growth. The cell numbers were gradually decreased either by increase of FC concentration (Figure 1) or by extension of exposure time (Figure S1) using MTT assay, indicating that FC suppresses cell numbers in a time‐ and dose‐dependent manner. The value of 50% inhibitory concentration (IC50) of FC is 4.2 μM at 24 h. Similarly, FC inhibits colon cancer SW480 cells in a dose‐dependent manner, and the IC50 of FC is 0.06 μM at 24 h (Figure S2A). Therefore, we used 4 μM of FC in the following study.

FIGURE 1.

FC suppresses human lung cancer cell growth by inducing apoptosis. A549 cells were treated with different doses of FC for 24 h. (A) Cell number was determined by MTT assay. (B) Cell viability was determined using LDH Cytotoxicity Assay Kit. (C) Apoptosis was determined by Annexin V/PI staining followed by flow cytometry. The red box encompasses the apoptotic cell population. The data are shown as mean ± SEM from three independent experiments. Data were analyzed using one‐way ANOVA. ANOVA, analysis of variance; FC, formosanin C; PI, propidium iodide; SEM, standard error of the mean. ***p < 0.001.

Initially, we clarified whether FC suppression of A549 cell number was due to cell death or inhibition of cell proliferation using the lactate dehydrogenase (LDH) cytotoxicity assay. Our data showed that FC decreased cell viability in a dose‐dependent manner compared to the untreated group (Figure 1B), indicating cell death after FC treatment. We then conducted Annexin V/PI staining followed by flow cytometry to determine whether FC‐induced cell death was caused by apoptosis. Our data showed that FC caused apoptosis of A549 cells in a dose‐dependent manner (Figure 1C, red box). Previous studies have shown that FC increased the levels of cleaved caspase‐3, ‐8, ‐9, and pro‐apoptotic protein BAX as well as decreased anti‐apoptotic protein Bcl‐2 in lung cancer cell H460. ¹⁹ In addition, FC also increased the levels of cleaved caspase‐2 and ‐9 in colon cancer cell HT‐29. ⁴ Moreover, caspase inhibitors suppressed FC‐induced apoptosis of these two cancer cell lines. ⁴ , ¹⁹ Altogether, these data imply that FC suppresses the lung cancer cell population through caspase activation‐mediated apoptosis.

3.2. FC acts as an autophagy inducer and a blocker in human lung cancer cells

Although FC has been reported to be an autophagy inducer, its role in autophagy remains contradictory because no LC3 degradation was detected in FC‐treated lung cancer A549 cells. ⁶ To clarify the role of FC treatment in autophagy progression, the same A549 cells were treated with FC, and the level of LC3 was measured by Western blotting and immunofluorescence assay. The expression level of LC3 II increased without degradation from 4 to 16 h after FC treatment (Figure 2A). The number of LC3 puncta in the cytosol was significantly increased in FC‐treated cells compared to untreated cells at 24 h post‐treatment (Figure 2B). We clarified whether FC induces the autophagic flux in lung cancer cells by extending the time of FC treatment to 72 h. The expression level of LC3 II steadily increased without degradation at 72 h in FC‐treated cells (Figure 2C). As a positive control, the cells treated with amiodarone (AD, an autophagy inducer), a complete autophagic flux was demonstrated within 72 h (Figure 2D). These results indicate that FC blocks autophagic flux in lung cancer cells. We further compared the effects of FC, AD, and chloroquine (CQ, autophagy blocker) on autophagy progression. Cells were treated with these three reagents for 24, 48, and 72 h. A steady increase in LC3 II level was detected in FC‐ and CQ‐treated cells. In contrast. A gradual increase followed by a decrease in LC3 II protein level was only seen in AD‐treated cells (Figure 2E). We also investigated LC3 puncta formation of FC‐, CQ‐, and AD‐treated cells plus control cells (Ctrl) at 72 h post‐treatment under fluorescent microscopy. Similarly, only FC‐ and CQ‐treated cells showed accumulation of green LC3 puncta but not AD‐treated cells (Figure 2F). These data imply that FC, similar to CQ, blocks autophagic flux. To confirm FC blockage of autophagy progression, A549 cells were treated with AD (inducer) for 48 h, followed by adding FC for another 24 h. In the AD group, complete autophagic flux was demonstrated at 72 h. LC3 II was evidently accumulated in the FC + AD group compared to the AD alone group (Figure 2G, 2.4 vs. 8.6), indicating that FC, like CQ, functions as a blocker of autophagic flux. Moreover, to clarify whether FC can also induce autophagic activity, cells were treated with FC, CQ, or FC + CQ for 48 h (Figure 2H). The level of LC3 II in FC + CQ group cells was higher than the level of LC3 II in FC‐ or CQ‐treated cells (Figure 2H, 11 vs. 8.7 or 7.8), indicating that FC could induce autophagic progression at the early stage (increased LC3 II accumulation) but fail to proceed to the degradation stage. Similarly, a study reported that FC suppresses PI3K/AKT/mTOR signaling pathway to induce autophagy in myeloma and lung cancer cells. ⁶ , ²⁰ Zhang et al. also reported that FC is an autophagy inducer in A549 cells. ⁶ Altogether, we reveal that FC treatment can induce early stage autophagy progression and proceed to blockage of autophagic flux in lung cancer cells.

FIGURE 2.

FC acts as an autophagy inducer and a blocker in human lung cancer cells. (A) A549 cells were treated with 4 μM FC for 0, 4, 8, 12, and 16 h. LC3 protein level in the cell lysate was evaluated using anti‐LC3 antibody by Western blotting. GAPDH was used as the internal control. (B) Cells were treated with or without 4 μM FC for 24 h, followed by labeling with FITC‐conjugated anti‐LC3 antibody, and DAPI was used to label the nuclei. Cells were investigated under a confocal microscope. Scale bar = 5 μm. For quantification, 10 cells were counted for the puncta number in each cell. (C) Cells were treated with or without 4 μM FC for 0, 24, 48, and 72 h. The LC3 protein level in the cell lysate was evaluated using anti‐LC3 antibody by Western blotting. GAPDH was used as the internal control. (D) Cells were treated with 5 μM amiodarone for 0, 24, 48, and 72 h. LC3 protein level was investigated. (E) Cells were treated with 4 μM FC, 5 μM AD, or 50 μM CQ for 0, 24, 48, and 72 h. LC3 protein level was investigated. (F) Cells were treated with 4 μM FC, 5 μM amiodarone, or 50 μM CQ for 72 h, followed by labeling with FITC‐conjugated anti‐LC3 antibody, and DAPI was used to label the nuclei. Cells were investigated under a confocal microscope. Scale bar = 5 μm (G) Cells were treated with or without FC (4 μM) and/or AD (5 μM) for 72 h. (H) Cells were treated with or without FC (4 μM) and/or CQ (50 μM) for 48 h.  The LC3 protein level of cell lysates was examined by Western blotting. GAPDH was used as the internal control. The data are shown as mean ± SEM from three independent experiments. Data were analyzed using one‐way ANOVA. ANOVA, analysis of variance; FC, formosanin C; SEM, standard error of the mean. ***p < 0.001.

3.3. FC increases MMP, blocks autophagy progression, and has no influence on CCCP‐induced mitophagy

It has been reported that FC decreases the mitochondrial membrane potential (MMP) of colon cancer cells. ⁴ Depolarized MMP is a marker of damaged mitochondria, which are then degraded by the autophagy degradation machinery in a process known as mitophagy. Loss of MMP may activate mitophagy to eliminate depolarized or damaged mitochondria. To explore the effect of FC on mitophagy of cancer cells, MMP and mitophagy‐related proteins (Parkin, phosphorylated Parkin (p‐Parkin), and COX IV) were investigated. Initially, we stained FC‐treated cells with JC‐1 to measure the level of MMP in the cell. JC‐1 aggregates in the mitochondrion and emits a red signal while MMP is normal, whereas in a depolarized mitochondrion, JC‐1 becomes a monomer and emits a green signal. CCCP (carbonyl cyanide m‐chlorophenylhydrazone)‐treated cells are the positive control of depolarized mitochondria. Unexpectedly, FC treatment increased but did not decrease MMP in A549 cells compared to CCCP‐treated and non‐treated cells (Figure 3A). Second, in FC‐treated cells, the expression levels of LC3 II, COX IV (mitochondrial intramembrane protein), and p‐Parkin (docking on damaged mitochondria) were increased (Figure 3B). Altogether, FC treatment increased MMP accompanied by increased COX IV (mitochondrion marker) and p‐Parkin, suggesting disruption of mitochondrial dynamics of A549 cells. It is possible that FC might block the recruitment of the mitochondrion by the autophagosome. To confirm whether FC impedes the interaction between the autophagosome and the mitochondrion, the distribution of LC3 (green, an autophagosome marker), COX IV (red, a mitophagy marker), and p‐Parkin (red, a mitophagy marker) in the cells with FC treatment was investigated under confocal microscopy. COX IV and LC3 were widely distributed in the cytosol of FC‐treated cells (Figure 3C). Similarly, p‐Parkin in the nucleus and LC3 in the cytoplasm were not colocalized in FC‐treated cells (Figure 3C). In conclusion, our image data sustain that autophagosome recruitment of mitochondria harboring either p‐Parkin or COX IV protein was interrupted by FC treatment. To validate the effect of FC on mitophagy, A549 cells were pretreated with FC for 24 h, followed by CCCP treatment for another 24 h. CCCP, a mitochondrial uncoupler and mitophagy inducer, increases the proton permeability of the mitochondrial inner membrane accompanied by decreased MMP. CCCP‐induced mitophagy in A549 was demonstrated by suppression of COX IV (Figure 3D). In contrast, FC‐treated cells showed higher MMP (Figure 3A, column 2 vs. 3) and increased COX IV level (Figure 3B) compared to CCCP treatment (Figure 3D, lane 3 vs. 2). Moreover, FC failed to affect CCCP‐related mitophagy (Figure 3D, lane 2 vs. 4). Taken together, our results imply that FC treatment blocks autophagy progression, increases MMP, COX IV, and p‐Parkin without affecting CCCP‐induced mitophagy.

FIGURE 3.

FC increases MMP, blocks autophagy progression, and has no influence on CCCP‐induced mitophagy. (A) A549 cells were treated with 4 μM FC for 24 h or 50 μM CCCP for 30 min. MMP of these cells after JC‐1 staining was analyzed by flow cytometry. Quantification of MMP change was determined from the value encompassed by the red box. (B) Cells were treated with or without 4 μM FC for 24, 48, and 72 h. Specific protein expression levels in the cell lysate were evaluated using specific antibodies by Western blotting. GAPDH was used as the internal control. (C) Cells were treated with 4 μM FC for 24 h. The distribution and co‐localization of LC3 (green) and COX IV (red) or LC3 and p‐Parkin proteins (red) in the cells with FC treatment were visualized by a confocal microscope. FITC‐labeled anti‐LC3 antibody, rhodamine‐labeled anti‐COX IV or p‐Parkin, and DAPI nuclei stain were used to label the cell. Scale bar = 5 μm. (D) Cells were treated with or without 4 μM FC and/or 40 μM CCCP for 24 h. Specific protein levels in the cell lysate were evaluated using specific antibodies by Western blotting. GAPDH was used as the internal control. FC, formosanin C; MMP, mitochondrial membrane potential.

3.4. FC treatment suppresses the motility of lung cancer cells through EMT pathway

We further investigated the effect of FC on cancer cell motility using Transwell™ assay. The concentration of FC was decreased to 2 and 4 μM to minimize FC‐induced cell death. We maintained A549 cells with or without FC (2 and 4 μM) for 24 h in 6‐cm dishes, followed by harvesting and seeding equal amounts of the cells in Transwells, and parallelly in 96‐well plates for another 7 h for migration and MTT assay, respectively. Migrated cells in Transwells (Figure 4A) and cell viability in 96‐well plates (Figure 4B) were measured. Figure 4B showed that 4 μM FC suppressed 34% of cell growth; however, 4 μM FC showed 71% suppression of cell migration in Figure 4A, indicating that FC indeed suppresses cell migration. It is known that epithelial–mesenchymal transition (EMT) participates in cell migration. Cells might downregulate epithelial‐related proteins (E‐cadherin) and upregulate mesenchymal‐related proteins (N‐cadherin and vimentin) to increase migratory ability. We found that FC treatment increased the level of E‐cadherin and decreased the levels of N‐cadherin and vimentin in A549 cells (Figure 4C). Besides, FC treatment decreased the levels of N‐cadherin and vimentin without changing the level of E‐cadherin in colon cancer SW480 cells (Figure S2D). The discrepancy between these two cell lines may relate to cell type difference and deserves further exploration. Altogether, the above results imply that FC suppresses cell migration possibly through EMT‐related pathways.

FIGURE 4.

FC treatment suppresses the motility of lung cancer cells through the EMT pathway. (A) A549 cells were maintained with or without FC (2 and 4 μm) for 24 h in 6‐cm dishes, followed by harvesting and seeding of equal amounts of the cells in Transwells (5*10⁴ cells), and parallelly in 96‐well plates (1*10⁴ cells) for another 7 h in the presence or absence of FC. The medium in the Transwell contained no FBS, and 10% FBS was used as a chemoattractant. Migrated cells were stained with 0.5% crystal violet. Migrated cells were calculated under a light microscope. (B) The cells in the six‐well plates were analyzed by MTT assay for cell survival rate after FC treatment for 7 h. (C) A549 cells were treated with or without FC (2 and 4 μm) for 24 h. Specific protein expression levels in the cell lysate were evaluated using specific antibodies by Western blotting. GAPDH was used as the internal control. The data are shown as mean ± SEM from three independent experiments. Data were analyzed using one‐way ANOVA. ANOVA, analysis of variance; EMT, epithelial–mesenchymal transition; FBS, fetal bovine serum; FC, formosanin C; SEM, standard error of the mean. ***p < 0.001.

3.5. The roles of autophagy in FC suppression of survival and migration of lung and colon cancer cells

To clarify the role of autophagy in FC‐related cancer cell tumorigenesis. Autophagy essential gene Atg5 was knocked down by lentiviral Atg5 shRNA in lung cancer A549 cells or knocked out (KO) by the CRISPR‐Cas9 system in colon cancer SW480 cells. The levels of both Atg5 and LC3 II protein were decreased in Atg5‐silenced cells (shAtg5) compared to non‐treated and shGFP A549 cells (Figure 5A) as well as in Atg5‐KO SW480 cells compared to the wild‐type (WT) cells (Figure S3C). We found that FC suppression of cell survival was further enhanced by about 10% in Atg5 knockdown cells compared to shGFP control cells (Figure 5B, column 2 vs. 4). In contrast, the knockdown of Atg5 alone did not affect cell viability (Figure 5B, column 1 vs. 3). However, FC‐induced apoptosis was not different between shGFP and shAtg5 cells, which indicates that autophagy is independent of FC‐induced apoptosis (Figure 5E, column 2 vs. 4). Moreover, FC suppression of cell migration (Figure 5C, column 2 vs. 4) and cell survival (Figure 5D, column 2 vs. 4) were no difference in Atg5 knockdown cells compared to the control shGFP cells, implying FC suppressed A549 cell proliferation and migration in an autophagy‐independent manner. Moreover, colon cancer SW480 Atg5 WT and KO cells were investigated for the role of autophagy in cell survival and migration (Figure S3). Similar to A549 cells, low concentrations of FC (0.5–1 μM) were used to treat SW480 for cell migration. We found that FC significantly suppressed WT SW480 cell migration compared to untreated control cells (Figure S3A,B). SW480 Atg5‐KO cells showed increased cell migration but no effect on cell survival in the presence of FC. However, SW480 Atg5‐KO cells showed increased cell migration but decreased cell survival without FC (Figure S3D,E), indicating that autophagy affects SW480 cell migration and survival differently in the presence or absence of FC. Altogether, autophagy plays different roles in FC‐related suppression of cell survival and migration depending on cancer cell types.

FIGURE 5.

The roles of autophagy in FC suppression of survival and migration of lung and colon cancer cells. (A) A549 was transduced with lentivirus shAtg5 or shGFP for 24 h. Cell lysates were harvested and analyzed by Western blotting for the expression levels of Atg5 and LC3 using specific antibodies. GAPDH was used as the internal control. (B) Cells were treated with or without 4 μM FC in the presence or absence of shAtg5 or shGFP for 24 h. The cell population was determined by MTT assay. (C) Cells were treated with or without 2 μM FC in the presence of shAtg5 or shGFP for 24 h. Cell motility was determined by Transwell™ assay. Migrated cells were stained with 0.5% crystal violet and counted under a light microscope. (D) The cell survival in the parallel 96‐well tray was determined by MTT assay. (E) Cells were treated with or without 4 μM FC in the presence of shAtg5 or shGFP for 24 h. Apoptosis was determined by Annexin V/PI staining followed by flow cytometry. The red box encompasses the apoptotic cell population. The data are shown as mean ± SEM from three independent experiments. Data were analyzed using one‐way ANOVA. ANOVA, analysis of variance; FC, formosanin C; PI, propidium iodide; SEM, standard error of the mean. ***p < 0.001.

4. DISCUSSION

FC and other types of saponin have been reported to function as autophagy inducers of cancer cells. ²¹ , ²² Saponin induces autophagy through the AMPK, Akt, JNK, and Ras/MEK/ERK signaling pathways, and some of them are ROS‐dependent. ²¹ , ²² , ²³ , ²⁴ In this study, FC induces a steady increase of LC3 II level (representing autophagosome formation) from 24 to 72 h without degradation after treatment, which was similar to the effect of treatment with the autophagosome fusion blocker chloroquine (CQ). This indicates that FC, like CQ, is an autophagy blocker, which differs from Zhang et al.'s report that FC is an autophagy inducer in lung cancer A549 cells. ⁶ In this study, we found that FC could induce an early autophagic activity because the level of LC3 II in CQ + FC‐treated cells was higher than the levels of LC3 II in CQ‐ and FC‐treated cells (Figure 2H, 11 vs. 8.7 and 7.8). Altogether, our data imply that FC could induce an early autophagic activity followed by blocking the degradation of LC3 II by the lysosome. However, whether FC blocks the fusion of the autophagosome with the lysosome or inhibits the degradation of LC3 protein in the autolysosome warrants further exploration.

LC3, designated as microtubule‐associated protein light chain 3, is widely expressed in cytoplasm and nucleus. Cytosolic LC3 I is converted to LC3 II when autophagy is induced. ²⁵ However, the LC3 I signal was low or even undetectable in some of our data after FC treatment (Figures 2C–E,G,H, 3B,D, 4C, and 5A and Figures S2E  and S3C). The possible reasons are: (1) LC3 II is more detectable than LC3 I using most of the commercial antibodies ²⁶ ; (2) FC treatment leads to induced autophagic activity followed by LC3 II accumulation. The band intensity of the overexpressed LC3 II masked the level of LC3 I under normal exposure (Figure S4); (3) Despite LC3 gene is ubiquitously expressed in many cells and tissues, the expression levels of LC3 I and II are cell type dependent.

Saponin‐induced autophagy may promote or suppress cancer cell growth. Generally, autophagy plays a cytotoxic role in tumorigenesis including apoptosis, ²¹ cell cycle arrest, ²⁷ and increasing the sensitivity to chemotherapy. ²⁸ , ²⁹ Herein, we reveal that knockdown of autophagy Atg5 gene increased FC‐induced cell death but not apoptosis (Figure 5B,E). This indicates that FC‐related autophagy is independent of FC‐induced cell death. Cui et al. also showed that polyphyllin VII (a pennogenin) inhibits FC‐induced autophagy by caspase‐dependent cleavage of Beclin1 to increase the cytotoxicity of lung cancer cells. ¹⁹ In contrast, blockage of autophagy degradation by CQ reduced FC‐induced apoptosis. ⁶ Lee et al. reported that FC‐induced apoptosis requires depolarization of MMP. ⁴ Loss of MMP leads to the release of cytochrome c and other components to induce apoptosis. ³⁰ In contrast, we revealed that FC increased MMP compared to CCCP‐treated and control cells (Figure 3A). MMP results from the electron transport chain via a serial redox reaction. H+ is transported by proton pumps from the mitochondrial matrix to intermembrane space, which causes the membrane potential between the matrix and intermembrane space. ATP production through ATP synthase is dependent on MMP. ³¹ Low MMP decreases ATP production, which might lead to energy deprivation. High MMP increases ATP production accompanied by ROS production. However, excessive ROS is harmful to the cell. Thus, maintaining an appropriate level of MMP is crucial for the cell. Mitophagy machinery is executed to clean abnormal mitochondria. Our data showed that FC increased MMP (Figure 3A) accompanied by increased p‐Parkin and COX IV (Figure 3B), but no colocalization of LC3 and p‐Parkin or LC3 and COX IV could be detected in A549 cells (Figure 3C), indicating no interaction between autophagosomes and mitochondria, which was supported by the data that FC treatment has no effect on CCCP‐induced mitophagy in A549 cells (Figure 3D). In addition, Lee et al. reported that FC treatment decreased MMP in HT‐29 cells. ⁴ The discrepancy may be caused by the difference in the experimental procedure. Notably, we removed the medium together with floating cells, and only the attached cells were analyzed. It is possible that the cells with the decreased MMP were removed together with the floating fraction. Instead, the attached cells contain high MMP, p‐Parkin, and COX IV. In summary, our findings imply that FC treatment disrupts mitochondrial dynamics. The underlying mechanism requires further exploration.

Increasing evidence has shown that inhibition of mitophagy promotes an anti‐cancer effect. B5G1, a derivative of betulinic acid, induces cell death in multidrug‐resistant cancer cells with PINK1/Parkin‐mediated mitophagy. ³² In addition, inhibition of mitophagy increases B5G1‐induced apoptosis in cancer cells. Blockage of doxorubicin (DXR)‐induced mitophagy increases the sensitivity to DXR in cancer stem cells. ³³ Resveratrol, a natural phenol, inhibits the growth of HeLa cells through impairment of mitophagy and excessive ROS production. ³⁴ Mitochondria play a key role in the induction of apoptosis. There are many apoptosis‐related proteins, including cytochrome c and apoptosis‐inducing factors, which are stored in the mitochondria. ³⁰ Apoptosis is induced once these compounds are released into the cytosol. Excessive ROS leakage from mitochondria into the cytosol also triggers apoptosis or necrosis. ³⁵ Thus, the regulation of mitochondrial is crucial for cell survival.

Maintaining a normal level of MMP can prevent apoptosis‐related materials from leaking into the cytosol. It is probable that cancer cells evade FC‐related cell death by increasing MMP (Figure 3A). Photobiomodulation suppresses apoptosis of the auditory cell line by enhancing MMP and ATP production. ³⁶ Besides increased MMP, colocalization of mitochondrial markers (COX IV and p‐Parkin) and LC3 protein was not seen in FC‐treated cells (Figure 3C), indicating that FC‐related autophagy was not involved in mitophagy. Altogether, we demonstrate that FC induces apoptosis accompanied by disruption of mitophagy progression of lung cancer cells (Figures 1C and 3).

Many mitophagy receptors have been reported, including OPTN (optineurin), NDP52 (also known as Calcoco2, calcium binding and coiled‐coil domain 2), FUNDC1 (FUN14 Domain Containing 1), Nix, BNIP3 (BCL2 Interacting Protein 3), and PHB2 (prohibitin‐2). Moreover, Xian et al. discovered that the interaction between mitochondrial fission protein 1 (Fis1) and Syntaxin 17 (STX17), a SNARE (soluble Nethylmaleimide‐sensitive factor attachment protein receptor) protein is involved in autophagy to prevent mitophagy. ³⁷ FC abolishes mitophagy possibly through manipulating the above proteins. Nonetheless, how FC disrupts mitophagy progression warrants further study.

Many papers have reported that FC has a broad‐range antitumor capacity, including the ability to induce apoptosis by activating caspase‐2, enhancing mitochondrial dysfunction, and promoting the cleavage of caspase‐3, caspase‐9, and poly (ADP‐ribose) polymerase (PARP). In addition, FC could decrease the expression of anti‐apoptotic protein Bcl‐XL and increase the levels of pro‐apoptotic proteins Bax and Bak. ⁴ FC also caused cell cycle arrest in ovarian cancer cells. ⁵ In addition, FC could suppress lung tumor growth, ovarian tumor angiogenesis, and metastasis in vivo. FC suppresses metastasis by inhibiting broad‐spectrum metalloproteinases (MMPs), ⁷ which are critical enzymes for cell invasion and are related to poor prognosis in many cancers. FC can also improve immune reactions to kill cancer cells in vivo. FC has the potential to become an antitumor reagent due to its effects on diverse functions of tumorigenesis. Therefore, combined therapy using FC and present anticancer reagents might be a solution for drug‐resistant patients.

5. CONCLUSION

FC is a well‐known, naturally occurring compound with anticancer activity. In contrast to Zhang et al.'s report that FC is an autophagy inducer, we demonstrate that FC function as an inducer as well as a blocker of autophagy in cancer cells, which leads to cancer cell apoptosis and decreased motility. FC‐related suppression of cell migration is possible through EMT‐related pathway.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

Supporting information

FIGURE S1. A549 cells were treated with different doses of FC for 0, 24, 48, and 72 h. Cell number was determined by MTT assay.

FIGURE S2. SW480 cells were treated with different doses of FC for 24 h.

FIGURE S3. SW480 cells were treated with or without 0.5 μM FC for 24 h.

FIGURE S4. A549, SW480, and 293 t cells were treated with or without different doses of FC for 24, 48, and 72 h. Cell lysates were harvested and analyzed by Western blotting for the expression level of LC3 using anti‐LC3 antibodies. GAPDH was used as the internal control.

Chu M‐L, Lin P‐W, Liu Y‐W, Wu S‐Y, Lan S‐H, Su C‐L, et al. Formosanin C suppresses cancer cell proliferation and migration by impeding autophagy machinery. Kaohsiung J Med Sci. 2023;39(5):489–500. 10.1002/kjm2.12658