Dasatinib induces autophagic cell death in human ovarian cancer

Authors

  • Xiao-Feng Le MD, PhD,

    Corresponding author
    1. Department of Experimental Therapeutics, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
    • Department of Experimental Therapeutics, The University of Texas M. D. Anderson Cancer Center, Unit 354, 1515 Holcombe Boulevard, Houston, TX 77030-4009
    Search for more papers by this author
    • Fax: (713) 745-2107

  • Weiqun Mao,

    1. Department of Experimental Therapeutics, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
    Search for more papers by this author
  • Zhen Lu MD,

    1. Department of Experimental Therapeutics, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
    Search for more papers by this author
  • Bing Z. Carter PhD,

    1. Department of Stem Cell Transplantation and Cellular Therapy, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
    Search for more papers by this author
  • Robert C. Bast Jr MD

    1. Department of Experimental Therapeutics, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
    Search for more papers by this author

Abstract

BACKGROUND:

Dasatinib, an inhibitor of Src/Abl family kinases, can inhibit tumor growth of several solid tumors. However, the effect and mechanism of action of dasatinib in human ovarian cancer cells remains unknown.

METHODS:

Dasatinib-induced autophagy was determined by acridine orange staining, punctate localization of GFP-LC3, LC3 protein blotting, and electron microscopy. Significance of beclin 1, AKT, and Bcl-2 in dasatinib-induced autophagy and growth inhibition was assayed by small interfering RNA (siRNA) silencing and/or overexpression of the gene of interest.

RESULTS:

Dasatinib inhibited cell growth by inducing little apoptosis, but substantial autophagy in SKOv3 and HEY ovarian cancer cells. In vivo studies showed dasatinib inhibited tumor growth and induced both autophagy and apoptosis in a HEY xenograft model. Knockdown of beclin 1 and Atg12 expression with their respective siRNAs diminished dasatinib-induced autophagy, whereas knockdown of p27Kip1 with specific siRNAs did not. Small hairpin RNA knockdown of beclin 1 expression reduced dasatinib-induced autophagy and growth inhibition. Dasatinib reduced the phosphorylation of AKT, mTOR, p70S6K, and S6 kinase expression. Constitutive expression of AKT1 and AKT2 inhibited dasatinib-induced autophagy in both HEY and SKOv3 cells. Dasatinib also reduced Bcl-2 expression and activity. Overexpression of Bcl-2 partially prevented dasatinib-induced autophagy.

CONCLUSIONS:

Dasatinib induces autophagic cell death in ovarian cancer that partially depends on beclin 1, AKT, and Bcl-2. These results may have implications for clinical use of dasatinib. Cancer 2010. © 2010 American Cancer Society.

Dasatinib, a dual inhibitor of Src and Abl tyrosine kinase, was approved by the US Food and Drug Administration for second-line treatment of chronic myelogenous leukemia.1, 2 Recently, several preclinical studies have demonstrated that dasatinib inhibits growth of a variety of solid tumors, including breast, prostate, brain, skin, bone, soft tissues, lung, head and neck, colon, and pancreatic cancers.1-7 Several mechanisms underlie dasatinib-induced suppression of leukemia and solid tumors, including G1 arrest of the cell cycle,1, 4, 7 induction of apoptosis,1-4, 7 and inhibition of cell migration/invasion/metastasis.1-8 However, the effect of dasatinib and the role of autophagy, a type II programmed cell death, in dasatinib-treated ovarian cancer cells has not been reported.

Autophagy is an intracellular degradative mechanism for eliminating damaged organelles and long-lived proteins.9, 10 The process of autophagy can be divided into the following steps: signaling initiation, membrane nucleation, vesicle elongation, autophagosome formation, autophagolysosome formation, and content degradation.11 Autophagosomes are defined ultrastructurally as intracellular, double-membraned vesicles that contain damaged organelles and proteins and membrane-bound protein called microtubule-associated protein light chain 3-II (LC3-II). LC3-II is modified from LC3-I by cleavage and phosphatidylethanolamine at the C-terminus and binds tightly to autophagosomal membrane.12 The amount of LC3-II is correlated with the extent of the autophagosome formation.13 Level of p62/SQSTM1 (p62), a multifunctional protein that targets proteins to degradation by proteasomes and autophagy, is also correlated with the extent of the autophagosome formation.14 Intact and increased autophagy function decreases p62 protein.14

The phosphoinositide-3 kinase and mammalian target of rapamycin (mTOR) kinase pathways play major roles in regulating the formation of autophagosomes.9, 10 A series of proteins encoded by autophagy genes (Atg) execute the process of autophagy.9, 10 Beclin 1 (Atg6), a Bcl-2 interacting protein, plays a role in autophagy induction.15 In addition, beclin 1, Atg14, Vps34, and Vps15 form a lipid kinase complex that engages vesicle nucleation.16 Atg5, Atg12, and LC3 promote vesicle elongation.17

Cells induce autophagy as a means of survival by increasing the turnover of intracellular components.18 Under prolonged adverse conditions, progressive cellular atrophy may lead to type II programmed cell death.18 Whether autophagy promotes cancer cell survival or produces type II programmed cell death largely depends on the nature of environmental stress and cancer cell context.9, 10, 19-21 Bcl-2 not only plays a negative role in apoptosis, but also inhibits beclin 1-dependent autophagy.15, 22 Therefore, levels of Bcl-2 may influence autophagy process via beclin 1.

In this report, dasatinib was found to induce significant autophagy, rather than apoptosis, in human ovarian cancer cells. We have documented the evidence of dasatinib-induced autophagy in ovarian cancer cells both in vitro and in vivo by several methods.

MATERIALS AND METHODS

Antibodies and Reagents

Antibodies recognizing phospho-Src (Y416), total Src, and AKT2 were purchased from Upstate-Millipore (Billerica, Mass). p27Kip1 antibody was purchased from BD Biosciences (San Diego, Calif). Antibodies against p-Akt-Ser473, total AKT, p-mTOR, beclin 1, Atg12, and p62 were purchased from Cell Signaling Technology (Beverly, Mass). Anti-LC3 was provided by Drs. N. Mizushima and T. Yoshimori (National Institute for Basic Biology, Okazaki, Japan). An antibody to glyceraldehyde-3-phosphate dehydrogenase was obtained from MBL International (Woburn, Mass). Small interfering RNAs (siRNAs) targeted to p27Kip1, beclin 1, and Atg12 were from Dharmacon (Lafayette, Colo) or Ambion (Austin, Tex). Transfection reagents used were Lipofectamine 2000 from Invitrogen (Grand Island, NY) and DharmaFECT #4 from Dharmacon. Dasatinib (Bristol-Myers Squibb, Princeton, NJ) was purchased from the Pharmacy Division of The University of Texas M. D. Anderson Cancer Center and dissolved in dimethylsulfoxide. A myristoylated AKT1 and AKT2 were provided by Dr. Gordon B. Mills (The University of Texas M. D. Anderson Cancer Center, Houston, Tex). pUC-CAGGS-Bcl-2 plasmid was provided by Dr. Y. Tsujimoto at Osaka University Graduate School of Medicine, Osaka, Japan. A piMARK vector was provided by Dr. Taro Q. P. Uyeda at the National Institute of Advanced Industrial Science and Technology, Ibaraki, Japan. A reporter construct containing the Bcl-2 promoter (pBcl-2-Luc) was provided by Dr. Haojie Huang at the University of Minnesota (Minneapolis, Minn).

Cell Growth Assay

A crystal violet cell growth assay was used to assess anchorage-dependent cell proliferation in 96-well cell culture microplates as described previously.23 Colony formation assay was performed in 6-well cell culture plates. SKOv3 and HEY cells (both cell lines have been verified by the G-banding karyotyping) were seeded at 1000 and 500 cells per well in triplicate and incubated overnight at 37°C. Cells were treated with either dimethylsulfoxide or dasatinib (300 nM for SKOv3 cells and 150 nM for HEY cells) for 14 days. Cells were fixed in 1% glutaraldehyde, and stained with 0.5% crystal violet (Sigma, St. Louis, Mo) in methanol. Colonies with >30 cells were counted under an inverted microscope at low magnification. Growth recovery experiments were carried out in similar conditions to colony formation assay. Washout experiments were done by washing the cells with complete media twice and the cells were refilled with complete media after 72-hour treatment with dasatinib. Instead of counting the colonies, a microplate reader was used to quantify the cell numbers.

Establishment of Stable Ovarian Cancer Subclones

For overexpressing GFP-LC3 and Bcl-2, HEY or SKOv3 cells were cotransfected with pGFP-LC3,24 or pUC-CAGGS-Bcl-2,25 and pcDNA3.1 (Invitrogen) by Lipofectamine 2000 according to the manufacturer's instructions. Stable clones were selected with G418. For expressing shBECN1, 2 oligonucleotides designed to target the beclin 1 sequence at agaactacaaacgctgttt were synthesized (Sigma Genosys, The Woodlands, Tex), annealed, and cloned into a piMARK vector.26 The sequences of the 2 oligonucleotides were: forward, 5′-caccagaattatagacgctgtttacgtgtgctgtccgtaaacagcgtttgtagttcttttt-3′; reverse, 5′-gcataaaaagaactacaaacgctgtttacggacagcacacgtaaacagcgtct ataattct-3′. Resultant pishBECN1 construct was verified by DNA sequencing. Stable clones were selected with blasticidin.

siRNA and Plasmid Transfection

HEY or SKOv3 cells were seeded on 6-well culture plates and transfected with control, beclin 1, or Atg12 siRNAs using the DharmaFECT #4 reagent (Dharmacon). A mixture of siRNA (50 nM final concentration) and transfection reagents were incubated for 20 minutes. Empty vector pcDNA3.1, or pGFP-LC3, or myristoylated AKT1 and AKT2 was transfected with Lipofectamine 2000.

Cell Cycle Distribution and Immunoblot Analysis

Distribution of cells in the sub-G1, G1, S, and G2/M phases of the cell cycle and Western blot analysis were measured as described previously.27

Quantitative Reverse-Transcription Polymerase Chain Reaction Analysis

Quantitative reverse-transcription polymerase chain reaction (PCR) analysis was performed with an ABI Prism 7900HT Sequence Detection System using SYBR Green universal PCR master mix (ABI, Foster City, Calif) as described previously.28 Oligonucleotide sequences of the primer sets used were: human Bcl-2 (forward, 5′-gggtacgataaccgggagat-3′; reverse, 5′-ctgaagagctcctccaccac-3′); human glyceraldehyde-3-phosphate dehydrogenase (forward, 5′-cgtcttcaccaccatggaga-3′; reverse, 5′-cggccatcacgccacagttt-3′). The melting curves were used to ensure there was no nonspecific amplification.

Apoptotic Assay

Caspase 3/7 activity was measured after dasatinib treatment with a Caspase-Glo 3/7 kit from Promega (Madison, Wis) according to the manufacturer's instructions. POLARstar OPTIMA (BMG Labtech, Offenburg, Germany) was used to determine the luminescent units of caspase activity.

Luciferase Reporter Assays

HEY cells were seeded on 12-well culture plates in triplicate and first transfected with a reporter construct containing the Bcl-2 promoter (pBcl-2-Luc)29 for 48 hours. Cells were then treated with dasatinib at different concentrations for another 24 hours. Relative luminescence units (RLU) were normalized with protein concentrations in each sample, and final values of RLU were expressed as RLU per migrogram of protein per milliliter.

HEY Xenografts in Nude Mice

HEY cells (106 in 0.1 mL phosphate-buffered saline [PBS]) were subcutaneously implanted into the flanks of 4-week-old female athymic nu/nu mice (ERO Animal Facility, The University of Texas M. D. Anderson Cancer Center). Five mice per group were used for each treatment. Once the tumors became palpable (mean size of 0.3 mm3 on Day 4 after inoculation), mice were divided into 2 groups and treated intraperitoneally with either dasatinib (10 mg/kg) or vehicle dimethylsulfoxide 5×/week. Treatment was continued for 2½ weeks. Tumors were collected immediately after sacrifice and fixed for electron microscopy as described below. The tumor volume in cubic millimeters was calculated as described previously.30 Experiments with nude mice were repeated twice with similar results. The results are presented as the mean ± standard error for all values. Experiments with nu/nu mice were reviewed and approved by The University of Texas M. D. Anderson Cancer Center Institutional Animal Care and Use Committee.

Acridine Orange Staining

Acridine orange (AO) flow cytometry was used to detect the development of acidic vesicular organelles as described before, with minor modification.31 Briefly, SKOv3 and HEY cells were treated with or without dasatinib for 24 hours. Cells were then stained with 0.5 μg/mL AO (Molecular Probes, Eugene, Ore) in complete medium at 37°C for 15 minutes, and examined by flow cytometry within 2 hours.

Confocal Microscopic Analysis of GFP-LC3 Spots

HEY subclones that stably expressed GFP-LC3 or cells that were transiently transfected with pGFP-LC3 were fixed with 4% formaldehyde after treatment with dasatinib (150 nM) for 24 hours. Cells were then washed with PBS, mounted, and examined using a confocal microscope (Olympus FluoView 1000, Olympus, Melville, NY). Digital images were obtained using FluoView 1000 software (Olympus). 200 GFP-LC3 cells that had >10 bright punctate GFP-LC3 spots were counted. The percentage of these cells among dimethylsulfoxide- or dasatinib-treated groups was calculated.

Transmission Electron Microscopy

Tissue or cells were fixed overnight at 4°C with a solution containing 3% glutaraldehyde plus 2% paraformaldehyde in 0.1 M cacodylate buffer (pH 7.3). Samples were treated with 0.1% cacodylate buffered tannic acid, postfixed with 1% buffered osmium tetroxide for 30 minutes, and stained en bloc with 1% uranyl acetate. Samples were then dehydrated in increasing concentrations of ethanol, infiltrated, and embedded in LX-112 medium (Electron Microscopic Sciences, Fort Washington, Pa). The samples were polymerized in a 70°C oven for 2 days. Ultrathin sections were cut in a Leica ultracut microtome (Leica, Deerfield, Ill), stained with uranyl acetate and lead citrate in a Leica EM Stainer, and examined in a JEOL 1010 transmission electron microscope (JEOL USA, Peabody, Mass) at an accelerating voltage of 80 kV. Digital images were obtained using an AMT Imaging System (Advanced Microscopy Techniques, Danvers, Mass).

Statistical Analysis

The results are presented as the mean ± 95% confidence intervals for all values except in vivo studies. A paired Student t test or analysis of variance testing was used to compare the differences among groups, with statistical significance considered if P ≤ .05.

RESULTS

Dasatinib Inhibits Tyrosine Phosphorylation of Src Kinase, Anchorage-Dependent Cell Proliferation, and Colony Formation of Ovarian Cancer Cells, but Induces Minimal Apoptosis

Tyrosine phosphorylation of Src kinase at site 416 (p-Y416 Src) is critical for its kinase activity.32 Dasatinib profoundly inhibited p-Y416 Src (Fig. 1A). Dasatinib also inhibited anchorage-dependent cell growth of HEY (Fig. 1B) and SKOv3 (Fig. 1C) ovarian cancer cells in a dose-dependent manner. At levels of dasatinib that can be attained in human plasma (approximately 250-300 nM), short-term treatment (72 hours) with dasatinib inhibited growth of HEY cells by 46% (Fig. 1B) and SKOv3 cells by 34% (Fig. 1C). Long-term treatment (14 days) with dasatinib significantly suppressed the ability of both HEY and SKOv3 cells to form colonies by 70% (Fig. 1D). Surprisingly, no significant apoptosis was detected in dasatinib-treated HEY and SKOv3 cells, as shown in Figure 1E to G.

Figure 1.

Dasatinib inhibits Src tyrosine phosphorylation, cell proliferation, and colony formation of human ovarian cancer cells with minimal induction of apoptosis in vitro. (A) Effect of dasatinib (Das) on tyrosine phosphorylation of Src at tyrosine 416 is shown. HEY cells were treated with dasatinib or dimethylsulfoxide (DMSO) for 24 hours, and total protein was then extracted for Western blot analysis. (B) Effect of dasatinib on HEY growth is shown. HEY cells were treated with dasatinib for 72 hours. Crystal violet staining expressed as optical density (OD) was used to measure cell proliferation. *P < .05 compared with untreated group. **P < .01 compared with untreated group. (C) Effect of dasatinib on SKOv3 growth is shown. SKOv3 cells were treated and evaluated as described in (B). *P < .05. (D) Effect of dasatinib on colony formation is shown. HEY and SKOv3 cells were treated with dasatinib for 14 days. **P < .01. (E) Effect of dasatinib on apoptosis in HEY and SKOv3 cells as determined by Sub-G1 fraction is shown. HEY and SKOv3 cells were treated with dasatinib (150 nM for HEY, 300 nM for SKOv3) for 24 hours. (F, G) Effect of dasatinib on caspase activity and nuclear morphology is shown. HEY cells were treated with dasatinib for 24 hours. (F) Caspase 3 and 7 activity was expressed in relative luminescence units (RLUs). (G) 4,6-Diamidino-2-phenylindole was used to stain nuclei. CTRL indicates control.

Dasatinib Induces Autophagy in Human Ovarian Cancer Cells In Vitro

AO is a lysosomotropic agent and is able to stain the acidic vesicular organelles.31 Although AO staining is not restricted to autophagic vesicles, this technique offers a rapid and quantitative method to measure induction of autophagy. Consequently, we have first used the AO staining and flow cytometric analysis to evaluate dasatinib-treated HEY cells for acidic vesicular organelles. As shown in Figure 2A, dasatinib treatment for 72 hours dramatically increased red fluorescence in HEY cells from 5.1% to 81.0%, indicating the induction of acidic vesicular organelles. Similar results were obtained in SKOv3 cells (Data not shown). The effect of dasatinib on endogenous LC3 protein was checked by Western blot analysis. As shown in Figure 2B, dasatinib decreased LC3-I protein and increased LC3-II in both SKOv3 and HEY cells. Dasatinib concurrently decreased p62 levels (Fig. 2B), consistently with its correlation with autophagy.14 These results were further confirmed by GFP-LC3 fluorescence microscopic analysis. After dasatinib treatment, punctate GFP-LC3 fluorescent spots dramatically increased, whereas diffuse fluorescence of GFP-LC3 in the cytoplasm and the nucleus disappeared (Fig. 2C). Dasatinib treatment resulted in 57% of HEY cells with punctate LC3, as opposed to only 8.4% cells with punctuate LC3 in dimethylsulfoxide-treated cells (Fig. 2D). Similar results were also found in SKOv3 cells (Fig. 2E). Finally, dasatinib-induced autophagy was confirmed by transmission electron microscopy. Dasatinib induced a dramatic increase in autophagosomes (blue arrows) and autophagolysosomes (green arrows) in the cytoplasm of both HEY cells (Fig. 2F) and SKOv3 cells (Fig. 2G). In contrast, dimethylsulfoxide-cells exhibited only normal mitochondria and endoplasmic reticulum without autophagic vesicles (Fig. 2F, G). No apoptotic features were found in dasatinib-treated HEY and SKOv3 cells. Thus, dasatinib induces typical autophagy in human ovarian cancer cells in vitro.

Figure 2.

Dasatinib induces autophagy in human ovarian cancer cells in vitro. (A) Measurement of dasatinib (Das)-induced autophagy with acridine orange (AO) staining is shown. HEY cells were treated with dasatinib for 72 hours. AO-stained cells were analyzed by flow cytometry. *P < .05. (B) Measurement of dasatinib-induced autophagy with Western blot analysis is shown. HEY and SKOv3 cells were treated as described in (A). GAPDH indicates glyceraldehyde-3-phosphate dehydrogenase. (C) Measurement of dasatinib-induced autophagy with stable subclones that express GFP-LC3 is shown. HEY and SKOv3 stable cells were treated with dasatinib for 72 hours. GFP-LC3 cells with >10 bright punctate GFP-LC3 spots were counted and shown as the percentage among dimethylsulfoxide (DMSO)- or Das-treated groups in HEY (D) and SKOv3 (E) clones. *P < .05. (F) Measurement of dasatinib-induced autophagy with electron microscopy (EM) in HEY cells. Cells were processed for EM examination as described in Materials and Methods. Two magnifications (×4000 and×25,000) of ultrastructure are shown. Autophagosomes are indicated by blue arrows and autophagolysosomes by green arrows. N indicates the nucleus. (G) Measurement of dasatinib-induced autophagy with EM in SKOv3 cells is shown as described in (F).

Dasatinib Inhibits Ovarian Tumor Growth and Induces Autophagy and Apoptosis in Ovarian Cancer Xenografts

To further confirm the ability of dasatinib to induce autophagy, a HEY ovarian cancer xenograft model in nu/nu mice was used to test the effects of dasatinib in vivo. As shown in Figure 3A, treatment with dasatinib at a dose of 10 mg/kg significantly suppressed the growth of HEY tumors. Ultrastructural analysis of the xenograft tumors collected at Day 21 showed autophagosomes (blue arrows) and autophagolysosomes (green arrows) in dasatinib-treated tumors (Fig. 3C), but not in dimethylsulfoxide-treated tumors (Fig. 3B). In contrast to the in vitro results, electron microscopic examination of xenograft tumors also showed typical signs of apoptosis such as nuclear condensation and nuclear body in dasatinib-treated tumors (Fig. 3E, red arrows), but not in dimethylsulfoxide-treated tumors (Fig. 3D). These data demonstrate that dasatinib inhibits ovarian tumor growth and induces autophagy and apoptosis in vivo.

Figure 3.

Dasatinib inhibits ovarian tumor growth and induces autophagy and apoptosis in vivo. (A) Effect of dasatinib on tumor growth of HEY xenografts is shown. The arrow indicates the beginning of dasatinib treatment. *P < .05, **P < .01 compared with the dimethylsulfoxide (DMSO)-treated group. (B) Ultrastructural examination in DMSO-treated HEY xenograft tumors is shown. Two magnifications (×4000 and×25,000) of ultrastructure are shown. N indicates the nucleus. (C) Ultrastructural examination in dasatinib-treated HEY tumors is shown at 2 magnifications. Autophagosomes are indicated by blue arrows and autophagolysosomes by green arrows. (D) Ultrastructural examination in DMSO-treated HEY xenograft tumors is shown. One magnification (×4000) of ultrastructure is shown. (E) Ultrastructural examination in dasatinib-treated HEY xenograft tumors is shown.

Beclin 1 and Atg12 are Critical for Dasatinib-Induced Autophagy

Dasatinib was able to induce p27Kip1-dependent G1 arrest of the cell cycle (Data not shown). p27Kip1 has been reported to play a role in autophagy induction in some settings.33, 34 Therefore, we sought to test whether dasatinib-induced autophagy is p27Kip1-dependent. HEY cells were transiently transfected with siRNA targeted to p27Kip1, control siRNA, or siRNAs targeted to well-known mediators that are required for autophagy induction such as beclin 1 and Atg 12. As shown in Figure 4A, knockdown of p27Kip1 with siRNA targeted to p27Kip11 did not reduce dasatinib-induced autophagy. Similar results were achieved with a second siRNA targeted to p27Kip12 (data not shown). As expected, down-regulation of beclin 1 or Atg12 with respective siRNAs significantly blocked dasatinib-induced autophagy (Fig. 4A). Effects of siRNA targeted to p27Kip1NA, beclin 1 siRNA, and Atg12 siRNA used in this study on respective proteins have been demonstrated by Western blot analysis (Fig. 4B-D). These results indicate that beclin 1 and Atg12, but not p27Kip1, are critical for dasatinib-induced autophagy.

Figure 4.

Beclin 1 and Atg12, but not p27Kip1, are required for dasatinib (Das)-induced autophagy. Dasatinib-induced growth inhibition partially depends on autophagy induction. (A) Effects of p27Kip1, beclin 1, and Atg12 small interfering RNAs (siRNAs) on dasatinib (Das)-induced autophagy determined by acridine orange (AO) staining are shown. HEY cells were transfected with negative control (Neg siR), p27Kip1#1 (siRNA targeted to p27Kip11), beclin 1 (BECN1 siR), or Atg12 (Atg siR) siRNAs (50 nM final concentration). Cells were treated with either dimethylsulfoxide or Das for 24 hours. AO staining was analyzed by flow cytometry. *P < .05 compared with Neg siR/Das-treated group; **P < .01 compared with Neg siR/Das-treated group. Data shown were averaged from 4 independent experiments (siRNA targeted to p27Kip1 data from 6 independent experiments). (B-D) Effect of p27Kip1, beclin 1, and Atg12 siRNAs on their respective protein expression is shown by Western blot analysis. Mock was the transfection reagent-treated cells. (E) Beclin 1 levels in stable subclones #46 and #69 that express beclin 1 shRNA (shBECN1) detected by Western blot analysis are shown. (F) Effects of shBECN1 on dasatinib-induced growth inhibition are shown. HEY stable shBECN1 subclones #46 and #69 and their control piMARK cells were treated with Das (150 nM) for 72 hours. Cell viability was measured with crystal violet staining. Dimethylsulfoxide-treated control (CTRL) was set as 1 for each subclone of piMARK, shBECN1#46, and shBECN1#69. Das-treated groups in these 3 subclones were expressed as ratio of respective CTRL. **P < .01 compared with piMARK CTRL group. (G) Effects of shBECN1 on dasatinib-induced autophagy are shown. shBECN1 subclones #46 and #69 and control piMARK cells were treated with Das for 24 hours, and AO staining was performed. GAPDH indicates glyceraldehyde-3-phosphate dehydrogenase.

Dasatinib-Induced Growth Inhibition Partially Depends on Autophagy Induction

To determine whether dasatinib-induced autophagy is responsible for growth inhibition in ovarian cancer cells, we have chosen to down-regulate beclin 1 as a testing model because it was required in dasatinib-induced autophagy (Fig. 4A). Beclin 1 siRNA sequences were subcloned into piMARK small hairpin RNA vector. Two stable sublines that consistently express reduced beclin 1 levels were selected (Fig. 4E). As shown in Figure 4F, shBECN #46 and #69 cells were more resistant to dasatinib-induced inhibition of cell growth as compared with piMARK control cells. shBECN #46 and #69 cells had lower autophagy induction than the piMARK control in response to dasatinib treatment (Fig. 4G), consistent with the beclin 1 siRNA data. These results suggest that dasatinib-induced autophagy accounts for dasatinib-induced growth inhibition in ovarian cancer cells.

Dasatinib-Induced Autophagy Partially Depends on AKT

To investigate the mechanisms of dasatinib-induced autophagy, we have studied the effects of dasatinib on phosphorylation of AKT and mTOR by Western blot analysis. As shown in Figure 5A, dasatinib treatment dramatically decreased phosphorylation of AKT and mTOR in both HEY and SKOv3 cells. To confirm the value of decreasing AKT in dasatinib-induced autophagy, HEY cells that stably express AKT1 and AKT2 were produced. AKT1 and AKT2 cells were transiently transfected with GFP-LC3 expression vector, and then treated with dasatinib for 24 hours. GFP-LC3 punctae were measured by confocal microscope. Overexpression of both AKT1 and AKT2 partially attenuates dasatinib-induced LC3 spots (Fig. 5B). This attenuated effect was associated with decreased inhibition of AKT and mTOR phosphorylation by dasatinib (Fig. 5C). Effects of AKT1 and AKT2 overexpression on dasatinib-induced autophagy was further confirmed in SKOv3 cells by using flow cytometric AO staining (Fig. 5D). A reverse phase protein array analysis was performed to check the signaling pathways affected by dasatinib. Results confirmed the inhibitory effects of dasatinib on p-AKT and p-mTOR, and further showed that downstream targets of AKT and mTOR, p70S6K and S6, were significantly suppressed (Fig. 5E). Therefore, the above data indicate that dasatinib-induced autophagy partially depends on AKT pathway.

Figure 5.

Dasatinib (Das)-induced autophagy partially depends on AKT. (A) Effects of Das on phosphorylation of AKT and mammalian target of rapamycin (mTOR) as determined by Western blot analysis are shown. (B) Effects of dominant positive AKT1 and AKT2 on Das-induced autophagy in HEY cells as determined by GFP-LC3 fluorescence microscopy are shown. HEY cells were transiently cotransfected with AKT1, AKT2, and pGFP-LC3 for 48 hours and then treated with dasatinib for another 24 hours. *P < .05 compared with Das-treated group. (C) AKT1, AKT2, and phosphorylation of AKT and mTOR expression was determined by Western blot analysis. (D) Effects of AKT1 and AKT2 on Das-induced autophagy in SKOv3 cells as determined by flow cytometric AO staining are shown. SKOv3 cells were treated as described in (B). *P < .05 compared with vector/Das-treated group. (E) Effects of dasatinib on phosphorylation of p70S6K and S6 kinase determined by a reverse phase protein array (RPPA). Phosphor-HER2 was used an internal control. DMSO indicates dimethylsulfoxide; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; AO, acridine orange.

Dasatinib Inhibits Bcl-2 Levels and Its Promoter Activity

Antiapoptotic Bcl-2 binds beclin 1 and inhibits beclin 1-dependent autophagy.15, 22, 35 Because dasatinib-induced autophagy is beclin 1-dependent (Fig. 4), we sought to check the effect of dasatinib on Bcl-2 levels. As shown in Figure 6A, dasatinib markedly suppressed Bcl-2 protein levels illustrated by Western blot analysis. Dasatinib also reduced Bcl-2 mRNA level as shown by quantitative PCR (Fig. 6B), indicating that dasatinib-induced Bcl-2 down-regulation occurred at the level of transcription. Consistent with this conclusion, dasatinib decreased Bcl-2 promoter activity in a dose-dependent manner (Fig. 6C). To further determine the role of Bcl-2 down-regulation in dasatinib-induced autophagy, HEY cells that stably express Bcl-2 were generated (Fig. 6D). As shown in Figure 6E, overexpression of Bcl-2 partially but significantly blocked dasatinib-induced autophagy in 2 independent sublines that overexpressed Bcl-2 as demonstrated by AO staining. This result was further confirmed by GFP-LC3 fluorescence analysis. As shown in Figure 6F, Bcl-2–overexpressing HEY cells exhibited fewer LC3 punctate spots induced by dasatinib treatment than control cells, indicating partial blockade of dasatinib-induced autophagy by Bcl-2 expression. Collectively, these results show that dasatinib-induced autophagy partially depends on Bcl-2.

Figure 6.

Dasatinib (Das)-induced autophagy partially depends on Bcl-2. (A) Effect of Das on Bcl-2 protein as detected by Western blot analysis is shown. HEY cells were treated with Das for 24 hours. (B) Effect of Das on Bcl-2 mRNA as detected by quantitative reverse-transcription polymerase chain reaction (QRTPCR) is shown. HEY cells were treated with Das for 24 hours. *P < .05. (C) Effect of Das on Bcl-2 promoter activity as detected by luciferase reporter assay. Relative luminescence units (RLUs) were normalized with protein concentrations in each sample, and final values of RLUs were expressed as RLU/μg protein/mL. (D) Validation of Bcl-2 stable subclones in HEY cells by Western blot analysis. (E) Effect of Bcl-2 overexpression on Das-induced autophagy as detected by acridine orange (AO) staining. HEY Bcl-2 stable subclones #26 and #27 were treated with Das for 24 hours. AO staining was measured with flow cytometry. *P < .05. (F) Effect of Bcl-2 overexpression on Das-induced autophagy as determined by GFP-LC3 fluorescence microscopy is shown. Bcl-2 stable sublines #26 and #27 were transiently transfected with pGFP-LC3 and then treated with Das for another 24 hours. *P < .05. GAPDH indicates glyceraldehyde-3-phosphate dehydrogenase; DMSO, dimethylsulfoxide.

DISCUSSION

As this report was being prepared for publication, a study was published showing that dasatinib induces autophagy, but not G1 arrest or apoptosis, in glioma cells.36 Our study presented here not only showed dasatinib-induced autophagy by AO staining and LC3 Western blot analysis, but also documented it by in vivo study and additional methods including electron microscopy. Our report has also revealed several key molecules and pathways such as beclin 1, AKT, and Bcl-2 that are required for dasatinib-induced autophagy. More importantly, we show that dasatinib-induced autophagy was functionally required for dasatinib-induced inhibition of cell growth.

Results shown by Milano indicate that dasatinib-induced autophagy is independent of the cell cycle G1 arrest,36 consistent with our observations that dasatinib-induced autophagy depended on beclin 1 and Atg12, but not p27Kip1 in ovarian cancer cells (Fig. 4A). However, there are reports that indicate the role of p27Kip1 in autophagy.33, 34 LKB1-AMPK–dependent phosphorylation of p27Kip1 at Thr 198 is reported to increase p27Kip1 and autophagy induction.34 Our data do not support the role of p27Kip1 in dasatinib-induced autophagy. Interestingly, knockdown of p27Kip1 seems to somehow increase dasatinib-induced autophagy (Fig. 4A). We do not know why p27Kip1 depletion enhances dasatinib-induced autophagy in HEY cells. We do know that dasatinib does not increase phosphorylation of p27Kip1 at Thr 198 or AMPK phosphorylation (data not shown). If phosphorylation of p27Kip1 at Thr 198 is required for p27Kip1 to induce autophagy, dasatinib-induced p27Kip1 is unable to induce autophagy. Another possibility is that p27Kip1 is expressed parallel or downstream of Src/Abl kinase inhibition. Src/Abl kinases have been indicated in directly regulating p27Kip1 expression.37, 38 Dasatinib-induced p27Kip1 could result directly from inhibition of Src/Abl kinases.

Dasatinib can induce both autophagy and apoptosis in vivo, but only autophagy in vitro in ovarian cancer cells (Figs. 2 and 3). The mechanism(s) underlying this disparity are unknown. It is possible that in vivo environmental conditions such as hypoxia, nutrition insufficiency or deprivation, and the immunodefensive system permit dasatinib to induce both type I and type II programmed cell death. Apoptosis induced in an HEY xenograft model by dasatinib may result from an autophagy-dependent mechanism. Prolonged and massive induction of autophagy is known to lead to autophagic cell death.18 Indeed, Konecny et al reported that dasatinib treatment with higher dose (1 μM) and longer duration (5 days) induced significant apoptosis in ovarian cancer cell lines.39 Resveratrol-induced apoptosis was reported to depend on resveratrol-induced autophagy.40 Therefore, dasatinib-induced autophagy and apoptosis may be related and contribute to dasatinib-induced growth inhibition.

Our in vitro and in vivo data showed that dasatinib induces autophagic cell death that accounts for dasatinib-induced growth inhibition in human ovarian cancer cells. Beclin 1, AKT, and Bcl-2 are critical mediators of dasatinib-induced autophagy.

Acknowledgements

We thank Drs. N. Mizushima and T. Yoshimori for supplying LC3 antibody and GFP-LC3 vector, Dr. Gordon B. Mills for providing AKT1/AKT2 plasmids, Dr. Y. Tsujimoto for providing Bcl-2 plasmid, Dr. Taro QP Uyeda for supplying the piMARK vector, and Dr. Haojie Huang for providing Bcl-2 promoter construct.

CONFLICT OF INTEREST DISCLOSURES

Supported by the Anne and Henry Zarrow Foundation; Mr. Stuart Zarrow; and National Foundation for Cancer Research. The High Resolution Electron Microscopy Core Facility, Flow Cytometry Core Facility South Campus, Media Preparation Core Facility, and Animal Core Facility were utilized. These facilities were funded by Core Grant CA16672 from the National Cancer Institute.

Ancillary