Ursolic acid promotes cancer cell death by ... - Wiley Online Library

IJC International Journal of Cancer

Ursolic acid promotes cancer cell death by inducing Atg5-dependent autophagy Shuilong Leng1†, Yanli Hao1†, Daobing Du1, Shanyan Xie1, Lepeng Hong1, Haigang Gu2, Xiao Zhu3, Jinfang Zhang3, Daping Fan4 and Hsiang-fu Kung3 1

Department of Human Anatomy, Guangzhou Medical University, Guangzhou, Guangdong, People’s Republic of China Department of Histology and Embryology, Guangzhou Medical University, Guangzhou, People’s Republic of China 3 Stanley Ho Centre for Emerging Infectious Diseases, Chinese University of Hong Kong, Shatin, Hong Kong 4 Department of Cell Biology and Anatomy, School of Medicine, University of South Carolina, Columbia, SC

Ursolic acid (UA) has been reported to possess anticancer activities. Although some of the anticancer activities of UA have been explained by its apoptosis-inducing properties, the mechanisms underlying its anticancer actions are largely unknown. We have found that UA-activated autophagy induced cytotoxicity and reduced tumor growth of cervical cancer cells TC-1 in a concentration-dependent manner. UA did not induce apoptosis of TC-1 cells in vitro as determined by annexin V/propidium iodide staining, DNA fragmentation, and Western blot analysis of the apoptosis-related proteins. We found that UA increased punctate staining of light chain 3 (LC3), which is an autophagy marker. LC3II, the processed form of LC3I which is formed during the formation of double membranes, was induced by UA treatment. These results were further confirmed by transmission electron microscopy. Wortmannin, an inhibitor of autophagy, and a small interfering RNA (siRNA) for autophagy-related genes (Atg5) reduced LC3II and simultaneously increased the survival of TC-1 cells treated with UA. We also found that LC3II was significantly reduced and that survival was increased in Atg52/2 mouse embryonic fibroblast (MEF) cells compared to Atg51/1 MEF cells under UA treatment. However, silencing BECN1 by siRNA affected neither the expression of LC3II nor the survival of TC-1 cells under UA treatment. These results suggest that autophagy is a major mechanism by which UA kills TC-1 cells. It is Atg5 rather than BECN1 that plays a crucial role in UA-induced autophagic cell death in TC-1 cells. The activation of autophagy by UA may become a potential cancer therapeutic strategy complementing the apoptosis-based therapies. Furthermore, regulation of Atg5 may improve the efficacy of UA in cancer treatment.

Key words: ursolic acid, autophagy, Atg5, apoptosis, TC-1 Abbreviations: Atg: autophagy-related genes; LC3: light chain 3; MEF: mouse embryonic fibroblast; PARP: poly(ADP-ribose) polymerase; PBS: phosphate-buffered saline; PI: propidium iodide; PI3K: phosphoinositide 3-kinase; siRNA: small interfering RNA; 3-MA: 3-methyladenine; UA: ursolic acid Additional Supporting Information may be found in the online version of this article. *S.L. and Y.H. contributed equally to this work Grant sponsor: Guangdong Natural Science; Grant number: 07301608; Grant sponsor: Science and Technology Institute of Guangzhou; Grant numbers: 10A175, 08A036 DOI: 10.1002/ijc.28301 History: Received 5 Sep 2012; Accepted 2 May 2013; Online 4 Jun 2013 Correspondence to: Shuilong Leng: Department of Anatomy, School Basic Science, Guangzhou Medical University, Guangzhou, Guangdong, People’s Republic of China, Tel.: 18620-813401790, Fax: 18620-81340182, E-mail: [email protected]; Hsiang-fu Kung; Stanley Ho Centre for Emerging Infectious Diseases, The Chinese University of Hong Kong; Shatin, Hong Kong; Tel: 18522603-7743; Fax: 1852-2994-4988; E-mail: [email protected]

C 2013 UICC Int. J. Cancer: 133, 2781–2790 (2013) V

Autophagy is a phenomenon that occurs in all eukaryotic cells ranging from yeast to mammals. Like apoptosis, autophagy has been suggested to be an important mechanism in maintaining cellular homeostasis in response to a variety of stimuli such as nutrient deprivation, high temperatures, and hypoxia. Autophagy begins with the formation of a doublemembrane structure called autophagosome which contains subcellular structures. These autophagosomes fuse with lysosomes to become autolysosomes where their contents are degraded into parts that can be reused elsewhere in the cell.1 The molecular mechanism of autophagy was initially studied in yeast. In total, more than 20 different autophagy-related genes (Atgs) have been identified. Among them, BECN1 (the mammalian ortholog of yeast Atg 6) acts during the nucleation and expansion of the autophagosomal membrane.2 The microtubule-associated protein light chain 3, LC3 (Atg8 in yeast), is associated with the double membranes of autophagosomes. The recruitment of LC3 to the membrane occurs via an Atg5-dependent mechanism, and thus Atg5 is essential for autophagosome formation.3,4 Increasing evidence has shown that autophagy is involved in tumorigenesis.5 It has been reported that when organelles and portions of the cytoplasm are degraded beyond a certain threshold, autophagic cell death is induced.6,7 It is also possible that autophagy can kill the cells by selective degradation

Cancer Cell Biology

2

2782

UA induce Atg5-dependent autophagic cell death

Cancer Cell Biology

What’s new? Ursolic acid, a naturally occurring compound found in certain fruits and vegetables, possesses potent anticancer activity. But while capable of undermining cancer cells by inducing their death through apoptosis, other aspects of ursolic acid’s anticancer mechanism are not fully understood. This study shows that the compound promotes cancer cell death by inducing Atg5-dependent autophagy, a process that increasingly has been found to play a role in tumorigenesis. In TC-1 cervical cancer cells, ursolic acid activated autophagy through PI3-K signaling. The findings suggest that the compound could be a potent complementary therapy for cancers resistant to caspase-dependent apoptosis.

of essential proteins. Autophagic cell death is regulated by a molecular mechanism distinct from that of apoptosis although BECN1 and Bcl-2 can cooperate with Atg5 to regulate both autophagy and apoptosis.8 This cell death may prevent neoplasia and thus the activation of the nonapoptotic autophagic cell death program is emerging as a potential cancer therapy complementing apoptosis-based therapies.9 Ursolic acid (UA), a natural pentacyclic triterpenoid carboxylic acid, is a potent anticancer agent. It has been shown to prevent tumorigenesis, inhibit tumor growth10,11 and suppress angiogenesis.12 UA was shown to sensitize cancer cells to chemotherapeutic agents, to induce cancer cell apoptosis and to inhibit the growth of DU145 cells in nude mice.13–15 UA was found to trigger apoptosis in MCF-7 breast cancer cells,16 inhibit the proliferation of mutated human colorectal cancer cells17 and induce apoptosis in cervical carcinoma cells.18 Although the anticancer activities of UA may partially be attributable to its apoptosis-inducing properties, the exact mechanisms of its anticancer actions are largely unknown. Efficacy of anticancer chemotherapeutic drugs relies on their ability to trigger cell death. Some anticancer drugs such as tamoxifen, arsenic trioxide, resveratrol and soybean B-group have been recently reported to induce autophagic cell death.19–22 In our study, we tested the hypothesis that UA also induces autophagic cell death under certain conditions. We demonstrated that UA promotes tumor cell autophagy, resulting in nonapoptotic cell death and a delay of tumor growth in vivo. We further showed that Atg5 plays a crucial role in UA-induced autophagic cell death while BECN1 makes little contribution.

Material and Methods Materials

UA (Sigma-Aldrich, St. Louis, MO) was dissolved in dimethyl sulfoxide, and stored at 220 C. Empirical formula: C30H48O3; molecular weight: 456.70. Primary antibodies including anti-caspase-3, anti-Bax, anti-Atg5 and anti-LC-3 were purchased from Sigma-Aldrich (St. Louis, MO), anticaspase-9, anti-poly(ADP-ribose) polymerase (PARP), antiBcl-2 and anti-BECN1 antibodies were purchased from Cell Signaling Technology (Danvers, MA). Anti-GAPDGH primary antibody and all secondary antibodies were purchased from Santa Cruz Biotechnology (Dallas, Texas). Cell culture and treatments

Two cervical cell lines TC-1 and HeLa cells line were used in our study. TC-1 cells were cultured in RPMI 1640 medium

(Invitrogen, CA, USA). HeLa cells, Atg51/1 and Atg52/2 mouse embryonic fibroblasts (MEFs) (Gifts from Noboru Mizushima, Japan) were maintained in DMEM (Invitrogen, CA, USA). Both media were supplemented with 10% v/v fetal bovine serum and 1% v/v penicillin/streptomycin. Cells were grown at 37 C in a humidified atmosphere consisting of 5% CO2. In some experiments, cells were exposed to various inhibitors for 1 hr prior to UA treatment, such as 10 mM 3methyladenine (3-MA), 400 nM Wortmannin and 50 mM LY294002 to inhibit phosphoinositide 3-kinase (PI3K); 10 mM U0126 to inhibit MEK1/2; 20 lM SB203580 to inhibit the activation of MAPK; 20 lM SP600125 to inhibit the activation of JUN; or 100 nM Bafilomycin, 2 lg/mL E64d and Pepstatin to inhibit degradation of LC3 II. All of these inhibitors were purchased from Sigma-Aldrich (St. Louis, MO). Xenograft tumor models

Female C57BL/6J mice (age, 6–8 weeks) were purchased from the Laboratory Animal Services Center, Sun Yat-sen University, and fed a standard animal diet and water ad libitum. Tumor cells were harvested by trypsinization, washed twice with 13 phosphate-buffered saline (PBS) and finally resuspended in PBS to the designated concentration for injection. Viable cells (1 3 106 in 100 lL of PBS) were inoculated in the left legs of mice. The mice were randomly grouped (six mice per group) and subcutaneously injected with various doses of UA (0, 25 and 50 mg/kg) for 10 weeks. About 2 weeks after inoculation with TC-1 tumor cells, tumors were visible in the left legs of the mice. Tumor growth was monitored by visual inspection and palpation twice a week. The tumor diameters were measured and the tumor volume (mm3) was calculated as follows: volume5 (shortest diameter)2 3 (longest diameter)/2. The animals were sacrificed in the tenth week after the tumor cell injection. All experiments were carried out according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publication No. 80-23) revised 1996. MTT and cell viability assays

Cell proliferation was assessed by methylthiazolyltetrazolium (MTT) assays. Cells (1 3 104) were cultured in 96-well plates in the presence (10, 20 and 30 lM) or absence of UA in a final volume of 200 lL for 5 days. At the end of the treatment, 10 lL of MTT (5 mg/mL in PBS) was added to each well. After 2 hr of incubation, MTT formazan precipitate was C 2013 UICC Int. J. Cancer: 133, 2781–2790 (2013) V

2783

Cancer Cell Biology

Leng et al.

Figure 1. The effect of UA on TC-1 cells in vitro and in vivo. (a) UA inhibited the TC-1 cells proliferation in vitro as revealed by MTT assay. (b) Representative images of mice with TC-1 xenograft with or without UA treatment for 10 weeks. Normal mice did not get xenografts. (c) The effects of UA treatment on tumor volume at different time points after TC-1 cell inoculation. (d) Representative images of tumors removed from the xenografted mice. (e) The effects of UA treatment on the tumor weight. TC-1 xenograft mice were treated for 10 weeks and the tumors were removed and weighed. Data were shown as means 6 SD, n 5 6 per group, *p 0.05, **p 0.01. (f) Representative images of H&E-stained tumor sections. White arrows indicate areas of necrosis. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

dissolved in 100 lL of dimethyl sulfoxide. We allowed the plates to stand for 10 min at room temperature and recorded the absorbance at 560 nm. Tests were performed at least three times. Cell viability was tested by first plating wild-type Atg51/ 1 MEFs, Atg52/2 MEFs or TC-1 cells at 2 3 105 cells/well in six-well plates. On the following day, UA (final concentration: 0, 10, 20 or 30 lM) was added and the cells were further cultured for 24 hr at 37 C. Viability was measured by trypan blue exclusion assay. Each data point was the average of three different experiments in duplicates. Flow cytometry analysis

Apoptosis rates were determined by flow cytometry using an annexin V/propidium iodide (PI) kit (Sigma-Aldrich, C 2013 UICC Int. J. Cancer: 133, 2781–2790 (2013) V

St. Louis, MO). Staining was performed according to the manufacturer’s instructions. The percentages of cells staining annexin V positive but PI negative were calculated as early apoptosis rate, whereas those of cells staining annexin V and PI double positive were calculated as percentages of total dead cells. DNA fragmentation assay

DNA fragmentation, which is an indicator of apoptosis, was measured by DNA fragmentation assay using spectrophotometry. The procedures were according to the instructions of Apoptotic DNA Ladder Kit (Roche, Indianapolis, Indiana). Cell extracts were prepared from TC-1 cells (1 3 106 cells) treated with UA at concentrations of 0, 10, 20 or 30 lM for 48 hr. After electrophoresis in a 1.8% agarose gel, the gel was

Cancer Cell Biology

2784


Figure 2. UA did not induce apoptosis in TC-1 cells. (a) Flow cytometry analysis of TC-1 cells treated with various concentrations of UA for 24 hr. (b) The effects of UA treatment on DNA fragmentation in TC-1 cells. M, DNA marker; Lane 1, vacant; 2, 30 lM UA; 3, 20 lM UA; 4, 10 lM UA; 5, 0 lM UA. (c) The effects of UA treatment on DNA fragmentation in Hela cells. M, DNA marker; Lane 1, 0 lM UA; 2, 10 lM UA; 3, 20 lM UA; 4, 30 lM UA; 5, vacant; 6, positive control (lyophilized apoptotic U937 cells). (d) The Western blot analysis of total TC-1 cells lysates for caspase-3, caspase-8, PARP, Bcl-2 and Bax. GAPDH was used to normalize protein loading. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

stained with ethidium bromide and the DNA ladder was visualized under UV light.

sliced into 60-nm sections. Ultrathin sections were stained with uranyl acetate and lead citrate, and examined with a JEM-1230 transmission electron microscope.

Western blotting

Cells grown in 10-cm dishes were treated with UA for 48 hr and then analyzed via Western blotting. Protein samples (total protein/sample, 50 lg) were loaded to 12% SDS-PAGE gels. After separation, the protein bands were transferred to nitrocellulose membranes. The membranes were blocked with TBST buffer containing 5% nonfat milk, incubated with a primary antibody (1:1,000, Sigma) for the detection of caspase-3, PARP, caspase-9, Bcl-2 and Bax proteins at 4 C overnight, followed by the addition of horseradish peroxidase-linked antirabbit antibody (1:200) and enhanced chemiluminescence visualization of the bands. All Western blot experiments were repeated at least three times and the representative images were shown.

Small interfering RNA suppression of gene expression

TC-1 cells (2 3 105 cells/well) cultured in a six-well tissue plate were transiently transfected with 100 nM of Atg5 small interfering RNA (siRNA), BECN1 siRNA or Control siRNA (all from Santa Cruz Biotechnology) according to the manufacturer’s instructions. Seventy-two hours after transfection, cells were subjected to the indicated further treatments. Statistical analysis

All data were expressed as mean 6 SD. Statistical Package for Social Sciences software (SPSS 13.5) was used for oneway ANOVA to compare the experimental groups. Differences were considered significant if p < 0.05.

Electron microscopy

Results

Cells were initially fixed in 0.1 M of sodium phosphate buffer containing 2.5% glutaraldehyde, pH 7.4, and then fixed in the same buffer containing 1% osmium tetroxide, pH 7.2. Cells were next embedded into Ultracut (Leica, Germany) and

UA promoted TC-1 cell death in vitro and in vivo

To assess the effects of UA on TC-1 cell proliferation in vitro, we treated the cells with increasing concentrations of UA for 5 days and assayed the cell growth by MTT assays. C 2013 UICC Int. J. Cancer: 133, 2781–2790 (2013) V

Figure 3. Representative electron microscopic images of TC-1 cells and HeLa cells exposed to30 lM of UA for 24 hr. (a) A TC-1 cell, the nucleus is intact. (b) An autophagic TC-1 cell after exposed to 30 lM UA for 24 hr, numerous vacuoles are observed. White arrows show the autophagosome. The nucleus is intact. Black arrows show the nucleus. (c) Double-membrane autophagosomes contain subcellular structures shown by white or red arrows. (d) A Hela cell, the nucleus is intact. (e) An apoptotic Hela cell after exposed to 30 lM of UA for 24 hr. The cell as a whole was shrunk, and the nucleus was condensed and fragmented (f). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Dose–response curves were plotted and showed that UA strongly inhibited TC-1 cell proliferation in a dose-dependent manner (Fig. 1a). To evaluate the effects of UA on tumor growth in vivo, we used a well-established TC-1 xenograft mouse model. Tumor growth was significantly inhibited in the UA-treated groups compared to the control group; the decrease in tumor growth was UA dose dependent (Figs. 1b and 1c). After the mice were treated with UA for 10 weeks, only small tumors were macroscopically visible near the left leg; in contrast, much larger tumors are visible in the control mice (Figs. 1d and 1e). Histological examination of tumor sections showed the presence of a significant amount of necrosis in the tumors of UA-treated mice as indicated by the arrows in Figure 1f. Taken together, these results suggested that UA strongly inhibited the proliferation of TC-1 cells in vitro and significantly reduced tumor growth in vivo.

We then performed DNA fragmentation assays. The result showed that UA did not induce any DNA fragmentation under the concentration range used. However, the DNA ladder, typically indicative of apoptosis, became apparent in HeLa cells treated with UA (Fig. 2b). Apoptosis is a highly regulated process that involves activation of a series of molecular events leading to cell death. Regulation of the apoptotic pathway is controlled by apoptosisrelated proteins. Therefore, we assessed the expression of apoptosis-related proteins, including caspase-3, caspase-9, PARP, Bcl-2 and Bax, by Western blot analysis. No significant differences in these proteins were observed in TC-1 cells treated with UA at different concentrations (Fig. 2c). Taken together, these results suggest that the in vitro proliferation suppressing activity and the in vivo tumor growth inhibitory effect of UA are not primarily owing to the induction of apoptosis, rather owing to the induction of nonapoptotic death in TC-1 cells.

UA did not induce apoptosis of TC-1 cells

To explore the mechanism underlying tumor growth inhibition caused by UA, we first examined the effect of UA on apoptotic death of TC-1 by FACS analysis using annexin V/ PI staining. As shown in Figure 2a, the total percentage of dead cells gradually increased in a UA dose-dependent manner. However, the percentage of the early apoptotic cells (annexin positive but PI negative) did not significantly increase. The result revealed that the strong cytotoxic effect of UA may not involve apoptosis of TC-1 cells. C 2013 UICC Int. J. Cancer: 133, 2781–2790 (2013) V

Autophagy was induced by UA in TC-1 cells

Transmission electron microscopy was used to monitor autophagy of TC-1 cells in response to UA (Fig. 3). After exposure to 30 lM of UA for 24 hr, there were a large number of autophagic vacuoles presented in UA-treated cells, but not in untreated cells. The presence of double-membrane containing cellular organelles was observed in TC-1-treated cells at higher magnification. UA-treated HeLa cells exhibited typical apoptosis characteristics in nucleus change.

Cancer Cell Biology

2785

Leng et al.


Cancer Cell Biology

2786

Figure 4. Autophagy was induced by UA in TC-1 cells. (a) Response of TC-1 cells to UA determined by bright-field and fluorescence microscopy. Cells treated with PBS were used as control. (b) Upper panel, the Western blot analysis of total TC-1 cells lysates for LC3. Bottom panel, protein levels of LC3 II were compared to those of GAPDH. (c) TC-1 cells were treated with UA for 24 hr in the presence or absence of U0126, 3MA, Wortmannin, Bafilomycin, E64d or Pepatatin. Whole-cell lysates were obtained and the content of LC3 I and LC3 II was determined by Western blotting. (d) Cell survival of different groups, each point of the cell survival is calculated by the average of different experiments in triplicate as the percent of untreated plated. *p 0.05. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

We observed that TC-1 cells treated with UA accumulated vacuoles in cytoplasm and displayed punctuated staining of LC3, a marker of autophagosome (Fig. 4a). Induction of autophagy was confirmed by Western blotting for LC3. LC3 II, a processed form of LC3I formed during the formation of the double membrane, was increased by UA treatment in a concentration-dependent manner (Fig. 4b). TC-1 cells treated with UA in the presence or absence of inhibitors were used to assess both the induction of autophagy and the autophagic flux caused by UA. Figure 4c shows that the increased expression of LC3II was abrogated by Wortmannin and LY294002, two PI3K inhibitors. Although 3-MA did not have a significant effect, the results indicated that PI3K signaling may be involved in UA-induced autophagy. In contrast, Bafilomycin, E64d and Pepstatin decreased the degradation of LC3 in lysosomes, which in turn, caused both LC3I and LC3II accumulation in TC-1 cells. The MEK inhibitor U0126, the p38 inhibitor SB203580 and the JUN inhibitor SP600125 had little effects on autophagy induced by

UA, indicating that MAPK signaling is not involved in UA-induced autophagy in TC-1 cells (Supporting Information Fig. S1A). To determine whether autophagy induced by UA was responsible for the reduced viability of TC-1 cells, we measured the viability after 24 hr of UA treatment. The viability of TC-1 cells was significantly restored by Wortmannin, but not by LY294002 or 3-MA. Moreover, the increase in autophagy induced by Bafilomycin A, E64d and Pepstatin was accompanied by a decrease in TC-1 viability (Fig. 4d and Supporting Information Fig. S2). This data pointed out a previously unseen role of UA in inducing autophagy that leads to cell death in TC-1 cells. Atg5, but not BECN1 is required in autophagosome formation in UA-treated cells

We hypothesized that if autophagy was important in UA cytotoxicity to tumor cells, Atg52/2 cells would be less sensitive to UA. In our study, we found significantly more cell C 2013 UICC Int. J. Cancer: 133, 2781–2790 (2013) V

2787

Cancer Cell Biology

Leng et al.

Figure 5. Atg5, but not Beclin 1 is required in autophagosome formation in UA-treated cells. (a) Atg51/1 MEFs and Atg52/2 MEFs were treated with increasing concentration of UA. Scale bar 5 50 lm. (b) Cell viability assay in Atg51/1 MEFs and Atg52/2 MEFs treated with increasing concentrations of UA. Viability was measured by trypan blue exclusion and was represented by average of three different experiments in duplicate. All error bars represent SEM, *p 0.05. (c) Western bolt for LC3 in Atg51/1 and Atg52/2 MEFs treated with different concentrations of UA. GAPDH was used as loading control. (d) Western blot for Atg5 in Atg51/1, Atg52/2 MEFs and TC-1 cells transiently transfected with siRNA against Atg5. (e) Western blot for LC3 in TC-1 cells transiently transfected with siRNA against Atg5. (f) Protein levels of LC3 II were compared to those of GAPDH in (e). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

viability during UA treatment in Atg52/2 cells as compared to Atg51/1 cells (Figs. 5a and 5b). This effect correlates with the LC3 II processing that was observed only in Atg51/ 1 cells in response to UA and was concentration dependent (Fig. 5c). Silencing of Atg5 in TC-1 tumor cells by siRNA decreased the expression of LC3 II (Figs. 5d–5f) and reduced the UA cytotoxicity in TC-1 cells (Fig. 6a). BECN1, an essential mediator of autophagy, plays pivotal role during the nucleation and expansion of the autophagosomal membrane, we explored whether BECN1 is required in autophagosome formation in UA-treated cells. However, reduction of BECN1 levels by siRNA affected neither the expression of LC3II nor the survival of TC-1 cells during UA treatment (Figs. 6b and 6c). Taken together, our results suggested that cell death of TC-1 cells induced by UA is dependent on Atg5, but not C 2013 UICC Int. J. Cancer: 133, 2781–2790 (2013) V

BECN1. The results indicated that Atg5, but not BECN1, is involved in UA-induced autophagy and toxicity in TC-1 tumor cells.

Discussion The results of our study have shown that UA is highly effective in inducing death of TC-1 cells and inhibiting tumor growth in a TC-1 xenograft mouse model (Fig. 1). Mechanistic study showed that the apoptotic pathway was not activated by UA in TC-1 cells (Fig. 2). Instead, autophagy may be a major mechanism for TC-1 cell death induced by UA and UA-induced autophagy was dependent on Atg5, but not on BECN1. TC-1 cells were derived from primary lung epithelial cells and immortalized by HPV 16 E6/E7 and then transformed

Cancer Cell Biology

2788


Figure 6. Silencing of Atg5 reduced the cytotoxicity of UA to TC-1 cells. (a) Bright-field microscopic imaging of TC1 cells treated with UA for 24 hr after transiently transfected with siRNA against Atg5 and BECN1. (b) Cell survival assay in TC1 cells treated with UA for 24 hr after transiently transfected with siRNA against Atg5 or BECN1, *p 0.05. (c) Western blot for BECN1 and LC3 in TC-1 cells transiently transfected with siRNA against BECN1, and then exposed to UA for 24 hr. (d) Schematic diagram shows the mechanism by which UA induces cell death in TC-1 cells. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

with an activated ras oncogene. This transformation process imitates the natural tumor progression of cervical cancer.23 We found that UA did not change the level of HPV 16 E6 gene and Ras expression (Supporting Information Fig. S2). But UA decreased the expression of HPV18 E6/E7 gene in HeLa cells,24 suggesting that UA exerts effects on TC-1 cells and HeLa cells by different mechanisms. UA has been reported to possess anticancer activities through inducing apoptosis of tumor cells, such as gastric cancer cell line BGC-803, hepatocellular cancer cell HepG2 and MCF-7 breast cancer cells.25–27 However, TC-1 cell death induced by UA lacked the features of apoptosis, such as nuclear fragmentation, apoptotic body formation and chromatin condensation (Fig. 2). In the absence of these characteristics of apoptosis, mitochondrial swelling (data not shown) and autophagic vacuoles were observed instead (Fig. 4a). Transmission electron microscopy showed a marked activity of autophagy with accumulation of autophagosomes (Fig. 3b). The formation of autophagosomes was confirmed by using autophagy marker LC3. Immunofluorescence studies revealed a clear accumulation of LC3 in the autophagosomes of UA-treated cells (Fig. 4a). LC3 is a microtubule-associated protein that is associated with the formation of autophagosomes through the conversion of LC3I (cytosolic) to LC3II

(membrane bound). Our studies showed that LC3II was increased by the treatment of UA in a concentrationdependent manner (Fig. 4b), indicating that autophagic process is tightly linked to UA cytotoxicity. Several signaling transduction pathways have been reported to be involved in autophagy, such as the phosphatidylinositol 3-kinase (PI3K) pathway and the Ras/MEK/ERK pathway.28–30 It appears that upstream signals to induce autophagy are stimulus dependent. In our study, the increase in expression of LC3II was abrogated by Wortmannin and LY294002, two PI3K inhibitors (Fig. 4c). The results indicate that PI3K is required as an initial signal in UA-induced autophagy. However, 3-MA, another PI3K inhibitor, did not significantly decrease the expression of LC3II. The reason may be that 3-MA can inhibit both Class I PI3K activity (which inhibits autophagy) and Class III PI3K activity (which is required for autophagy), and thus the overall effect of 3MA may not be apparent.31 Interestingly, the viability of TC1 cells was significantly restored by Wortmannin. However, similar results were not observed by using LY294002. LY294002 may have some additional effects on cell survival by targeting its ATP-binding site.32 Autophagic flux can be measured by inferring degradation of LC3II in autolysosomes. Inhibiting the degradation can be achieved by using C 2013 UICC Int. J. Cancer: 133, 2781–2790 (2013) V

2789

Leng et al.

treatment.42,43 Pathological stimulation of BECN1independent autophagy is associated with neuronal cell death.44 Arsenic trioxide induced a BECN1-independent autophagic pathway in ovarian carcinoma cells.45,46 In our study, inhibition of BECN1 by siRNA did not change the expression of LC3II and cell death in response to UA, suggesting that UA induced a BECN1-independent autophagic pathway in TC-1 cells. Therefore, for those human cancer cells that are resistant to drugs that induce caspasedependent apoptotic cell death owing to the deletion of BECN1, UA may still be effective. Furthermore, the regulation of autophagy via Atg5 can increase the levels of cell death and enhance the efficiency of UA. Thus, the combination of drugs that induces apoptotic cell death and UA may yield synergistic clinical benefit in cancer treatment. These results also suggest that UA may be a potent complementary therapy for cancers that are resistant to caspase-dependent apoptotic cell death.

Conclusions In summary, our study demonstrated for the first time that UA induced autophagy in TC-1 cervical cancer cells through PI3-K signaling pathways that have been previously implicated in the control of the induction of autophagy. The autophagy induction is a major mechanism for UA-induced TC-1 cell death, and UA-induced autophagy is dependent on Atg5, but not BECN1. This new anticancer mechanism of UA suggests that UA may be a potent complementary drug for cancer treatment when patients are unresponsive to drugs that induce caspase-dependent apoptotic cell death.

Acknowledgement The authors thank Professor Noboru Mizushima (Tokyo Medical and Dental University, Japan) for donating the Atg51/1 and Atg52/2 MEFs.

References 1.

2.

3.

4.

5.

6.

Mizushima N, Levine B, Cuervo AM, et al. Autophagy fights disease through cellular selfdigestion. Nature 2008;451:1069–75. Arsov I, Li X, Matthews G, et al. BAC-mediated transgenic expression of fluorescent autophagic protein Beclin 1 reveals a role for Beclin 1 in lymphocyte development. Cell Death Differ 2008; 15:1385–95. Mizushima N, Yamamoto A, Hatano M, et al. Dissection of autophagosome formation using Apg5-deficient mouse embryonic stem cells. J Cell Biol 2001;152:657–68. Kuma A, Hatano M, Matsui M, et al. The role of autophagy during the early neonatal starvation period. Nature 2004;432: 1032–6. Mathew R, Kongara S, Beaudoin B, et al. Autophagy suppresses tumor progression by limiting chromosomal instability. Genes Dev 2007;21: 1367–81. McPhee CK, Logan MA, Freeman MR, et al. Activation of autophagy during cell death requires the engulfment receptor Draper. Nature 2010;465:1093–6.

7.

Li P, Du Q, Cao Z, et al. Interferon-gamma induces autophagy with growth inhibition and cell death in human hepatocellular carcinoma (HCC) cells through interferon-regulatory factor-1 (IRF1). Cancer Lett 2012;314:213–22. 8. Zhou F, Yang Y, Xing D. Bcl-2 and Bcl-xL play important roles in the crosstalk between autophagy and apoptosis. FEBS J 2011;278:403–13. 9. Chi S, Kitanaka C, Noguchi K, et al. Oncogenic Ras triggers cell suicide through the activation of a caspase-independent cell death program in human cancer cells. Oncogene 1999;18: 2281–90. 10. Kim KH, Seo HS, Choi HS, et al. Induction of apoptotic cell death by ursolic acid through mitochondrial death pathway and extrinsic death receptor pathway in MDA-MB-231 cells. Arch Pharm Res 2011;34:1363–72. 11. Ikeda Y, Murakami A, Ohigashi H. Ursolic acid: an anti- and pro-inflammatory triterpenoid. Mol Nutr Food Res 2008;52:26–42. 12. Sohn KH, Lee HY, Chung HY, et al. Anti-angiogenic activity of triterpene acids. Cancer Lett 1995;94:213–8.


13. Shyu MH, Kao TC, Yen GC. Oleanolic acid and ursolic acid induce apoptosis in HuH7 human hepatocellular carcinoma cells through a mitochondrial-dependent pathway and downregulation of XIAP. J Agric Food Chem 2010;58: 6110–8. 14. Li Y, Xing D, Chen Q, et al. Enhancement of chemotherapeutic agent-induced apoptosis by inhibition of NF-kappaB using ursolic acid. Int J Cancer 2010;127:462–73. 15. Shanmugam MK, Rajendran P, Li F, et al. Ursolic acid inhibits multiple cell survival pathways leading to suppression of growth of prostate cancer xenograft in nude mice. J Mol Med 2011;89: 713–27. 16. Kassi E, Sourlingas TG, Spiliotaki M, et al. Ursolic acid triggers apoptosis and Bcl-2 downregulation in MCF-7 breast cancer cells. Cancer Invest 2009;27:723–33. 17. Xavier CP, Lima CF, Preto A, et al. Luteolin, quercetin and ursolic acid are potent inhibitors of proliferation and inducers of apoptosis in both KRAS and BRAF mutated human colorectal cancer cells. Cancer Lett 2009;281:162–70.

Cancer Cell Biology

bafilomycin, which blocks the fusion of autophagosomes with lysosomes, and protease inhibitors (e.g., E64d and pepstatin).33,34 Bafilomycin, E64d and pepstatin resulted in accumulation of both LC3I and LC3II in the TC-1 cells, and hence increased the total LC3 level induced by UA while simultaneously decreasing the viability of TC-1 cells (Fig. 4d and Supporting Information Fig. S1B). This is owing to the fact that when LC3 levels are beyond a certain threshold autophagic cell death may be enhanced. Taken together, these data revealed a previously unseen role for UA in inducing autophagy-mediated death of TC-1 cells. We test the effect of UA on other two gynecological tumor cell lines: ID8 and 4T1. ID8, a cell line derived from spontaneous in vitro malignant transformation of C57BL/6 mouse ovarian surface epithelial cells. 4T1 cells were the mouse breast cancer cell line.35,36 We also found that UA increased the level of autophagy in both 4T1 and ID8 with dose-dependent manner, meanwhile significantly decreased the viability of the two cell lines (Supporting Information Fig. S3). Autophagy and autophagic cell death may share common machinery utilizing Atg5. Atg5 plays a crucial role in IFN-ginduced autophagic cell death by interacting with FADD as a necessary step leading to cell death,37 whereas upregulation of Atg5 is necessary for the oncogenic H-ras-induced autophagic cell death.38 Our study showed that Atg5 was required for LC3-II expression in UA-treated cells, and was thus involved in autophagic cell death in response to UA. These findings implicate a role for Atg5 in promoting UA-mediated autophagic cell death. BECN1 acts as a tumor suppressor gene in mice and is frequently deleted in human cancers.39–41 BECN1 is involved in some autophagic cell deaths but not in others. The BECN1-independent autophagic death is attracting increasing interests in the development of new strategies for cancer

Cancer Cell Biology

2790

18. Yim EK, Lee KH, Namkoong SE, et al. Proteomic analysis of ursolic acid-induced apoptosis in cervical carcinoma cells. Cancer Lett 2006;235:209–20. 19. Bursch W, Ellinger A, Kienzl H, et al. Active cell death induced by the anti-estrogens tamoxifen and ICI 164 384 in human mammary carcinoma cells (MCF-7) in culture: the role of autophagy. Carcinogenesis 1996;17:1595–607. 20. Kanzawa T, Kondo Y, Ito H, et al. Induction of autophagic cell death in malignant glioma cells by arsenic trioxide. Cancer Res 2003;63:2103–8. 21. Opipari AW, Jr, Tan L, Boitano AE, et al. Resveratrol-induced autophagocytosis in ovarian cancer cells. Cancer Res 2004;64:696–703. 22. Ellington AA, Berhow M, Singletary KW. Induction of macroautophagy in human colon cancer cells by soybean B-group triterpenoid saponins. Carcinogenesis 2005;26:159–67. 23. Sharma RK, Elpek KG, Yolcu ES, et al. Costimulation as a platform for the development of vaccines: a peptide-based vaccine containing a novel form of 4-1BB ligand eradicates established tumors. Cancer Res 2009;69:4319–26. 24. Yim EK, Lee MJ, Lee KH, et al. Antiproliferative and antiviral mechanisms of ursolic acid and dexamethasone in cervical carcinoma cell lines. Int J Gynecol Cancer 2006;16:2023–31. 25. Prasad S, Yadav VR, Kannappan R, et al. Ursolic acid, a pentacyclin triterpene, potentiates TRAILinduced apoptosis through p53-independent upregulation of death receptors: evidence for the role of reactive oxygen species and JNK. J Biol Chem 2011;286:5546–57. 26. Yang L, Liu X, Lu Z, et al. Ursolic acid induces doxorubicin-resistant HepG2 cell death via the release of apoptosis-inducing factor. Cancer Lett 2011;298:128–38. 27. Huang CY, Lin CY, Tsai CW, et al. Inhibition of cell proliferation, invasion and migration by ursolic acid in human lung cancer cell lines. Toxicol In Vitro 2011;25:1274–80.


28. Pozuelo-Rubio M. Regulation of autophagic activity by 14-3-3zeta proteins associated with class III phosphatidylinositol-3-kinase. Cell Death Differ 2010;18:479–92. 29. Manov I, Pollak Y, Broneshter R, et al. Inhibition of doxorubicin-induced autophagy in hepatocellular carcinoma Hep3B cells by sorafenib—the role of extracellular signal-regulated kinase counteraction. FEBS J 2011;278:3494–507. 30. Wu YT, Tan HL, Huang Q, et al. Activation of the PI3K-Akt-mTOR signaling pathway promotes necrotic cell death via suppression of autophagy. Autophagy 2009;5:824–34. 31. Ito S, Koshikawa N, Mochizuki S, et al. 3-Methyladenine suppresses cell migration and invasion of HT1080 fibrosarcoma cells through inhibiting phosphoinositide 3-kinases independently of autophagy inhibition. Int J Oncol 2007;31:261–8. 32. Alaimo PJ, Knight ZA, Shokat KM. Targeting the gatekeeper residue in phosphoinositide 3-kinases. Bioorg Med Chem 2005;13:2825–36. 33. Tanida I, Minematsu-Ikeguchi N, Ueno T, et al. Lysosomal turnover, but not a cellular level, of endogenous LC3 is a marker for autophagy. Autophagy 2005;1:84–91. 34. Mizushima N, Yoshimori T. How to interpret LC3 immunoblotting. Autophagy 2007;3:542–5. 35. Ganapathy E, Su F, Meriwether D, et al. D-4F, an apoA-I mimetic peptide, inhibits proliferation and tumorigenicity of epithelial ovarian cancer cells by upregulating the antioxidant enzyme MnSOD. Int J Cancer 2011;130:1071–81. 36. Przybyla BD, Shafirstein G, Koonce NA, et al. Conductive thermal ablation of 4T1 murine breast carcinoma reduces severe hypoxia in surviving tumour. Int J Hyperthermia 2012;28:156–62. 37. Pyo JO, Jang MH, Kwon YK, et al. Essential roles of Atg5 and FADD in autophagic cell death: dissection of autophagic cell death into vacuole formation and cell death. J Biol Chem 2005;280: 20722–9.

38. Byun JY, Yoon CH, An S, et al. The Rac1/ MKK7/JNK pathway signals upregulation of Atg5 and subsequent autophagic cell death in response to oncogenic Ras. Carcinogenesis 2009;30: 1880–8. 39. Jiang ZF, Shao LJ, Wang WM, et al. Decreased expression of Beclin-1 and LC3 in human lung cancer. Mol Biol Rep 2012;39:259–67. 40. Won KY, Kim GY, Lim SJ, et al. Decreased Beclin-1 expression is correlated with the growth of the primary tumor in patients with squamous cell carcinoma and adenocarcinoma of the lung. Hum Pathol 2012;43:62–8. 41. Cheng HY, Zhang YN, Wu QL, et al. Expression of beclin 1, an autophagy-related protein, in human cervical carcinoma and its clinical significance. Eur J Gynaecol Oncol 2012;33: 15–20. 42. Grishchuk Y, Ginet V, Truttmann AC, et al. Beclin 1-independent autophagy contributes to apoptosis in cortical neurons. Autophagy 2011;7: 1115–31. 43. Scarlatti F, Maffei R, Beau I, et al. Role of noncanonical Beclin 1-independent autophagy in cell death induced by resveratrol in human breast cancer cells. Cell Death Differ 2008;15: 1318–29. 44. Zhu JH, Horbinski C, Guo F, et al. Regulation of autophagy by extracellular signal-regulated protein kinases during 1-methyl-4phenylpyridinium-induced cell death. Am J Pathol 2007;170:75–86. 45. Smith DM, Patel S, Raffoul F, et al. Arsenic trioxide induces a beclin-1-independent autophagic pathway via modulation of SnoN/SkiL expression in ovarian carcinoma cells. Cell Death Differ 2010;17:1867–81. 46. Raffoul F, Campla C, Nanjundan M. SnoN/SkiL, a TGFbeta signaling mediator: a participant in autophagy induced by arsenic trioxide. Autophagy 2010;6:955–7.
