Suppression of Tumor Growth In vivo by the Mitocan α-tocopheryl Succinate Requires Respiratory Complex II

Purpose: Vitamin E (VE) analogs are potent novel anti-cancer drugs. The purpose of this study was to elucidate the cellular target by which these agents, represented by α-tocopoheryl succinate (α-TOS), suppress tumors in vivo, with the focus on the mitochondrial complex II (CII). Experimental Design: Chinese hamster lung fibroblasts with functional, dysfunctional and reconstituted CII were transformed using H-Ras. The cells were then used to form xenografts in immunocompromized mice, and response of the cells and the tumors to α-TOS was studied. Results: The CII-functional and –reconstituted cells, unlike their CII-dysfunctional counterparts, responded to α-TOS by ROS generation and apoptosis execution. Tumors derived from these cell lines reciprocated their responses to α-TOS. Thus, growth of CIIfunctional and –reconstituted tumors was strongly suppressed by the agent, and this was accompanied by high level of apoptosis induction in the tumor cells. On the other hand, αTOS did not inhibit the CII-dysfuntional tumors. Conclusions: We document in this report a novel paradigm, according to which the mitochondrial CII, which rarely mutates in human neoplasias, is a plausible target for anticancer drugs from the group of vitamin E analogs, providing support for their testing in clinical trials. Dong et al.: Complex II as a new cancer therapy target 4 Mitochondria, organelles vital for cellular energy homeostasis as well as being central purveyors of cell death, have recently come into focus as promising targets for anti-cancer therapies (1-3). Mitocans are small compounds that exhibit anti-cancer activity by selectively inducing apoptosis by mitochondrial interference, comprising 7 classes of agents with different modes of action (4). The class 5 mitocans include drugs that act by targeting the mitochondrial electron transport chain (ETC), of which the prime examples are vitamin E (VE) analogs, epitomized by the ester α-tocopheryl succinate (α-TOS) (5, 6), 3bromopyruvate (3BP) (7), and adaphostine (8). VE analogs also belong to the class 2 mitocans since, in addition to targeting the ETC, these agents act as Bcl-2 homology-3 (BH3) mimetics (9). VE analogs hold substantial promise as selective anti-cancer drugs since they suppress tumor growth in several pre-clinical models, including mice with experimental breast (10, 11), lung (12), prostate (13) and colon carcinomas (6, 14), as well as mesotheliomas (15, 16). Due to its selectivity for cancer cells and low toxicity for non-malignant cells (5), the ester analog α-TOS is an agent that has considerable clinical potential. A similar level of anti-cancer efficacy and selectivity has also been reported for the ether analog of VE αtocopheryloxyacetic acid (α-TEA) (17, 18). The importance of mitochondria as target by which α-TOS induces apoptosis was proposed earlier (19-22). Although it is now established that the VE analog causes mitochondrial destabilization through the formation of Bax/Bak channels resulting in cytochrome c translocation and caspase activation (20-22), the precise molecular target for apoptosis induction by the agent was not fully understood. α-TOS has been recently identified as a compound that induces apoptosis by targeting the ubiquinone (UbQ)-binding sites in the mitochondrial complex II (CII; succinate dehydrogenase, SDH) (23). In this respect, we recently reported that the VE analog induces apoptosis by a novel mechanism via generation Dong et al.: Complex II as a new cancer therapy target 5 of reactive oxygen species (ROS) upon displacing UbQ from the membrane subunits of CII (23). Therefore, electrons that are generated during the conversion of succinate to fumarate by subunit A of CII (SDHA, within the matrix part of the complex; 24) diffuse from CII and reduce molecular oxygen to generate ROS (23, 25) that, in turn, induce apoptosis (21, 26, 27). In this communication, we show the importance of CII as the major target for the mitocan, α-TOS, using a mouse model with genetically modified tumors expressing either wild type, defective or reconstituted CII. This study highlights the importance of CII as a target for anticancer drugs as well as its essential role in the activity of the mitocan, α-TOS, for promoting apoptosis. Materials and methods Cell culture, transfection and treatment. Parental Chinese hamster (ch) lung fibroblasts (B1 cells) (28), CII-defective cells (B9 cells with mutant CybL, SDHC) (28), CII-reconstituted cells (B9SDHC cells) (23), and CI-defective cells (B10 cells) (29), were grown in DMEM with 10% FCS, antibiotics, 5 mg/ml glucose and 1% (w/v) non-essential amino acids. B1, B9 and B10 cells were transformed with H-Ras. Briefly, the cells were transfected with pEGFP-C3-H-Ras (30) and the positive cells selected with G418. Single cell-derived colonies were obtained by limiting dilution except for B10Ras cells. Clone 6 of B1Ras cells and clone 1 of B9Ras cells were chosen for further experiments based on the level of the H-Ras protein. SDHC-mutant B9Ras cells were reconstituted by transfection with a vector encoding a functional human (h) SDHC gene. First, the pEFIRES-P-SDHC plasmid was constructed. SDHC was digested from pCR3.1-SDHC (31) using Nhe1 and Not1. SDHC and pEFIRES-P (32) were separately extracted from the gel using the QIAEX II kit (Qiagen). SDHC was then inserted into pEFIRES-P and the resulting pEFIRES-P-SDHC digested with BamHI to Dong et al.: Complex II as a new cancer therapy target 6 confirm the correct orientation of the cloned SDHC. The construct was amplified and purified using the QIAfilter plasmid purification kit. B9Ras cells were then reconstituted with hSDHC by transfection with pEFIRES-P-SDHC. G418 (100 μg/ml) and puromycin (4 μg/ml) were used for double selection; single clones were obtained by limiting dilution. Clone 5 of B9RasSDHC cells with high SDHC and H-Ras-EGFP expression was chosen for further studies. For treatment, cells were grown to ~60% confluency before exposure to α-TOS or thenoyltrifluoroacetone (TTFA) (both Sigma). In some experiments, cells were pre-treated with MitoQ (33, 34). Colony-forming assay. The colony-forming activity of the non-transformed B1, B9 and B9SDHC cells and their H-Ras-transformed counterparts was assessed as described (35). The number of colonies >50 μm in diameter in each dish was then counted using light microscopy. Assessment of apoptosis, ROS accumulation and oxygen consumption. Apoptosis was estimated routinely with flow cytometry using the annexin V-FITC or annexin V-PE binding method(s) as detailed elsewhere (21). Cellular ROS were detected by electron paramagnetic resonance (EPR) spectroscopy in cells loaded with the radical scavenger 5,5-dimethyl-1-pyrroline N-oxide (10 mM, DMPO; Sigma) (21, 23) or by flow cytometry using the redox-sensitive probe dihydroethidium (DHE) (21). Western blotting and RT-PCR. Western blotting was performed using anti-SDHC IgG (clone 3E2; Novus Biologicals), anti-H-Ras IgG (BD Biosciences) and anti-GFP IgG with anti-β-actin IgG (both Santa Cruz) as a loading control. For RT-PCR, the primers were as follows; hSDHC primers (31): sense 5'-CAC TTC CGT CCA GAC CGG AAC-3’, anti-sense 5'-ATG CTG GGA GCC TCC TTT CTT CA-3'; chSDHC primers: sense 5’-CGT CCT GTT TCT CCC CAC CTC-3’, anti-sense 5’-CAG CAA GCA TCA AGA CAG CCA C-3’; chGAPDH primers (36): sense 5’-GCA AGT TCA AAG Dong et al.: Complex II as a new cancer therapy target 7 GCA CAG TCA A-3’; anti-sense 5’-CGC TCC TGG AAG ATG GTG AT-3’. For RT-PCR ex vivo, total RNA was prepared from mouse tumors (see below) by homogenizing with TRIZOL in a glass mortar placed on ice (Dual Size 22/5 ml, Kontes Glass). MTT and SDH activity analysis. The MTT viability assay was used to assess the proliferative activity of the parental and H-Ras-transformed B1, B9, B10 and B9SDHC cell lines. SDH activity was measured using a short term (2 h) modified MTT assay with succinate as the sole source of electrons, driving the respiratory system specifically via CII, as described (11, 23). Animal experiments. Immunocompromized, athymic Balb c/nu-nu mice were injected subcutaneously with either B1Ras, B9Ras or B9Ras-SDHC cells at 5x10 cells per animal. After 2-3 weeks (B1Ras and B9Ras-SDHC cells) and 3-4 weeks (B9Ras cells), tumors were observed in the animals and further inspected using the Vevo770 ultrasound imaging (USI) device equipped with the RMV708 probe (frequency, 80 MHz; resolution, 30 μm) (VisualSonics) (10, 11, 23). Mice with tumors were treated by intraperitoneal (i.p.) injection of α-TOS dissolved in corn oil/4% ethanol (v/v) twice a week, at 10 μmol α-TOS for the first two doses and 6 μmol αTOS subsequently. The kinetics of tumor progression was quantified non-invasively twice a week by measuring the tumor size using the USI device and the volumetric analysis software. Animal studies were performed according to the guidelines of the Australian and New Zealand Council for the Care and Use of Animals in Research and Teaching and were approved by the local Animal Ethics Committee. Histochemistry. Tumors were excised and either snap-frozen (-80°C) or fixed in neutralbuffered formalin. The frozen tissue was used for RT-PCR and the fixed tissue was paraffinembedded and cut into 5 μm-thick serial sections, which were evaluated following hematoxylin-eosin (H&E) staining. For fluorescence microscopy, nuclei were stained with DAPI present in the Vectashield mounting medium (Vector Laboratories), and the tumor cells Dong et al.: Complex II as a new cancer therapy target 8 detected on the basis of green fluorescence resulting from expression of H-Ras-EGFP. The confocal images were taken using the Olympus IX81 microscope operated by the Fluoroview v. 1.6A software. Tumor sections were also inspected for apoptotic nuclei. Three sections were evaluated for every type of tumor and condition. Statistical analyses. Between-group comparisons were made using mean ± SD and the unpaired Student’s t test. Differences in the mean relative tumor size (± SEM) were examined using analyses of covariance (ANCOVA) with days as the covariate. Statistical analyses were performed using SPSS 10.0 analytical software. Statistical significance was accepted at