Targeting the Deregulated Spliceosome Core Machinery in Cancer Cells Triggers Mtor Blockade and Autophagy

Published OnlineFirst January 28, 2013; DOI: 10.1158/0008-5472.CAN-12-2501

Cancer Therapeutics, Targets, and Chemical Biology Research

Targeting the Deregulated Spliceosome Core Machinery in Cancer Cells Triggers mTOR Blockade and Autophagy

Virginie Quidville1,2,3, Samar Alsafadi1,2,3, Aïcha Goubar1,2,3,Fred eric Commo1,2,3,Veronique Scott1,2,3, Catherine Pioche-Durieu2,3, Isabelle Girault1,2,3, Sonia Baconnais2,3, Eric Le Cam2,3, Vladimir Lazar2,3, Suzette Delaloge1,2, Mahasti Saghatchian1,2, Patricia Pautier2, Philippe Morice2,3, Philippe Dessen2,3, Stephan Vagner1,2,3, and Fabrice Andre1,2,3

Abstract The spliceosome is a large ribonucleoprotein complex that guides pre-mRNA splicing in eukaryotic cells. Here, we determine whether the spliceosome could constitute an attractive therapeutic target in cancer. Analysis of gene expression arrays from lung, breast, and ovarian cancers datasets revealed that several genes encoding components of the core spliceosome composed of a heteroheptameric Sm complex were overexpressed in malignant disease as compared with benign lesions and could also deﬁne a subset of highly aggressive breast cancers. siRNA-mediated depletion of SmE (SNRPE) or SmD1 (SNRPD1) led to a marked reduction of cell viability in breast, lung, and melanoma cancer cell lines, whereas it had little effect on the survival of the nonmalignant MCF-10A breast epithelial cells. SNRPE or SNRPD1 depletion did not lead to apoptotic cell death but autophagy, another form of cell death. Indeed, induction of autophagy was revealed by cytoplasmic accumulation of autophagic vacuoles and by an increase in both LC3 (MAP1LC3A) protein conversion and the amount of acidic autophagic vacuoles. Knockdown of SNRPE dramatically decreased mTOR mRNA and protein levels and was accompanied by a deregulation of the mTOR pathway, which, in part, explains the SNRPE-dependent induction of autophagy. These ﬁndings provide a rational to develop new therapeutic agents targeting spliceosome core components in oncology. Cancer Res; 73(7); 2247–58. 2013 AACR.

Introduction the most enriched pathway in breast cancer samples, when Targeted therapies have been shown to improve outcome in compared with benign lesions (1). breast cancers. Molecular targeted agents usually inhibit key The aim of the present study was to determine whether the oncogenic pathways involved in cancer progression or resis- spliceosome could be a relevant therapeutic target in cancer. tance to conventional treatments. Most of these new molecular The spliceosome is a dynamic macromolecular ribonucleopro- therapies target kinase pathways and are usually highly effec- tein (RNP) complex that catalyses the splicing of nuclear pre- tive in a specific molecular segment. In the recent years, mRNA (precursor mRNA) into mRNAs. Nuclear pre-mRNA accumulative evidence has been reported that beyond kinase splicing is essential to remove internal noncoding regions of and DNA repair defects, many pathways are crucial for cancer pre-mRNA (introns) and to join the remaining segments (exons) progression. For instance, metabolism, immune system, and into mRNA before export to the cytoplasm and translation. The epigenetics have been reported as targetable in oncology. spliceosome does the 2 primary functions of splicing, that is, Looking for novel pathways involved in cancer progression, recognition of the intron/exon boundaries and catalysis of the we previously identified spliceosome assembly components as cut-and-paste reactions that remove introns and join exons. The spliceosomal machinery complex is formed from 5 RNP subunits, termed uridine-rich (U-rich) small nuclear RNP (snRNP), Authors' Affiliations: 1Institut National de la Sante et de la Recherche transiently associated to a range of protein factors (2, 3). Each Medicale (INSERM) U981, Paris; 2Institut de cancerologie Gustave Roussy, Villejuif; and 3Universite Paris-Sud XI, Kremlin-Bicetre,^ France snRNP (U1, U2, U4/U6, and U5) consists of a uridine-rich small nuclear RNA (snRNA) complexed with a set of 7 proteins known Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/). as Sm or SNRP proteins. The Sm proteins (B/B', D1, D2, D3, E, F and G) form a 7-member ring core structure that encompasses V. Quidville and S. Alsafadi contributed equally to this work and are joint first RNA. All Sm proteins share a conserved Sm domain, consisting authors. of 2 sets of conserved sequences (Sm1 and Sm2) that are S. Vagner and F. Andre are co-senior authors. responsible for the assembly of snRNAs in an ordered manner, Corresponding Author: Fabrice Andre, INSERM U981, Institut Gustave to form the Sm core of the spliceosomal snRNPs (4, 5) and are Roussy, 114 rue Edouard Vaillant, Villejuif 94805, France. Phone: 33-1- thereby involved in pre-mRNA processing (6). 42114293; Fax: 33-1-42115217; E-mail: [email protected] The splicing process has a crucial role in the control of the doi: 10.1158/0008-5472.CAN-12-2501 expression of many genes (such as those involved in cell cycle, 2013 American Association for Cancer Research. signal transduction, angiogenesis, apoptosis, and invasion;

www.aacrjournals.org 2247

Quidville et al.

refs. 7–15). Mutations of splicing factors or alterations of the of 1 mL. For controls, cells were either exposed to Hiperfect expression of the components of the splicing machinery can reagent alone (mock) or transfected with a nontargeting affect the splicing pattern of genes and contribute to the control siRNA. Transfection medium was replaced by fresh development of tumors (16–19). Deregulated alternative splic- culture medium every 48 hours until the end of experiment. ing can lead to the production of novel mRNA isoforms coding After 96 hours of transfection, cells were harvested, counted for protein variants with key functions in cancer using the trypan blue exclusion method or crystal violet (7, 10, 13, 20, 21). A substantial fraction of alternative splicing staining, and processed for the various biochemical assays. events in humans can also result in mRNA isoforms that harbor siRNA transfection effect and siRNA concentration effect on a premature termination codon (PTC). These transcripts are cell number were estimated by 2-way ANOVA. predicted to be degraded by the nonsense-mediated mRNA For NMD inhibition experiments, cells were transfected with decay (NMD) pathway which constitutes a mechanism of RNA SNRPE siRNA or nontargeting control siRNA. Three days later, surveillance that detects and eliminates aberrantly spliced cells were treated for 8 hours with 100 mg/mL cycloheximide mRNAs harboring a PTC to prevent the synthesis of nonfunc- (Sigma-Aldrich) or with an equivalent amount of dimethyl tional or deleterious truncated proteins (22, 23) and to regulate sulfoxide (DMSO) as a control. RNA was then isolated and the abundance of mRNA transcripts (24). mRNA expression was detected by real-time PCR (RT-PCR). In the present study, we have evaluated whether compo- Quantifications of the RT-PCR signals were calculated from nents of the spliceosome machinery complex were overex- RT-PCR assays conducted on samples from 2 independent pressed in malignant tumors and whether blocking spliceo- transfections. some could lead to antitumor effects. Immunoblot analysis Cells were lysed in radioimmunoprecipitation assay (RIPA) Materials and Methods buffer, and proteins were quantified using a microBCA assay Cell culture and siRNA transfection (Pierce, ThermoScientific). Equal amounts were separated on All cell lines were supplied by American Type Culture SDS-PAGE gels. Proteins were transferred to polyvinylidene Collection (ATCC, LGC Standards) between 2007 and 2011. fluoride (PVDF) membranes followed by immunoblot with Each ATCC cell line was first amplified to generate a cell master antibodies specific for mTOR (Cell Signaling, Ozyme), 4E-BP1 bank. All experiments were carried out from this cell bank. All (Cell Signaling), phospho-4E-BP1 (Ser 65; Cell Signaling), anti- the cell lines used were tested negative for mycoplasma LC3B (Sigma-Aldrich), SNRPD1 (C-15) and SNRPE (L-17; Santa contamination using VenorGeM Advance PCR Kit (Biovalley). Cruz Biotechnology). Immunolabeled proteins were detected Human mammary carcinoma cell lines SKBr-3 (ATCC HTB- by a chemiluminescence detection system (Super Signal West 30), MDA-MB 468 (ATCC HTB-132), HCC1937 (ATCC CRL- Dura Signal, Pierce, ThermoScientific). Membranes were then 2336), lung carcinoma A549 (ATCC CCL-185), and malignant washed in 0.1% TBS-Tween-20 and incubated with rabbit anti- melanoma A375 (ATCC CRL-1619) were cultured as mono- GAPDH or mouse anti-actin (C4; Chemicon, Millipore) to layers at 37 C with 5% CO2 and maintained by regular passage quantify and normalize results. in complete media consisting of RPMI-1640 GlutaMaxI or Dulbeccos' Modified Eagle's Media (DMEM):F12 (Gibco BRL, Cell-cycle analysis Invitrogen Life Technology) supplemented with 10% heat- Cell-cycle distribution was measured in SKBr-3 cells trans- inactivated FBS (PAN Biotech, Dutscher), 1 mmol/L sodium fected with 50 nmol/L SNRPE-specific validated siRNA or pyruvate (Invitrogen), and 10 m mmol/L HEPES (Invitrogen). nontargeting control siRNA. Cells were harvested at 72 hours MCF-7 (ATCC HTB-22) was maintained in completed DMEM/ after transfection, suspended in PBS, and fixed in 70% ethanol. F12–GlutaMaxI medium supplemented with 0.01 mg/mL insu- Then, DNA content was evaluated after propidium iodide lin (Invitrogen). Nontumoral epithelial cell line MCF-10-2A staining. Fluorescence-activated cell-sorting analysis was car- (ATCC CRL-10781) was maintained in DMEM/F12–GlutaMaxI ried out using a FACScan flow cytometer (Beckton Dickinson) medium supplemented with 20 ng/mL EGF (Gibco, Invitro- and CellQuest software. gen), 100 ng/mL cholera toxin (Sigma-Aldrich), 0.01 mg/mL insulin (Invitrogen), 500 ng/mL hydrocortisone (Sigma- Apoptosis detection Aldrich), and 5% heat-inactivated horse serum (Invitrogen). The PE Annexin V Apoptosis Detection Kit I (BD Bio- Forty-eight hours after plating, the cells were transfected with sciences) was used to detect apoptosis by flow cytometry. At specific SNRPE or D1 validated siRNA (see Supplementary 72 hours of siRNA transfection, cells were harvested, washed in Table S1 for sequence information; Qiagen) or nontargeting PBS, and pelleted by centrifugation. They were resuspended at control siRNA (ON-TARGETplus Non-targeting Pool, Dharma- 105 cells/100 mL in a binding buffer to which 5 mL of Annexin-V con) using HiPerfect transfection agent (Qiagen) according to and 5 mL of 7-aminoactinomycin D (AAD) were added and then the manufacturer's instructions. Briefly, cells were plated in 12- incubated in the dark for 15 minutes at room temperature. well plates in 4 replicates and grown to approximately 50% to Then, 400 mL of binding buffer was added and the cells were 60% confluency. siRNA was diluted in OPTI-MEM minimal immediately processed with a FACScan flow cytometer. The culture medium (Invitrogen) at a final concentration of 50 numbers show the percentages of cells in each quadrant nmol/L and then Hiperfect was added to the diluted siRNA and (bottom left: 7AAD/PE, intact cells; bottom right: 7- the mixture was dropwise added to each well in a final volume AADþ/PE, early apoptotic cells; top left: 7-AAD/PEþ,

2248 Cancer Res; 73(7) April 1, 2013 Cancer Research

Spliceosome as a Target for Anticancer Treatment

necrotic cells; top right: 7-AADþ/PEþ, late apoptotic or ed sample amplification and labeling, hybridization, array necrotic cells). Staurosporine (5 mmol/L)-treated SKBr-3 cells scanning, and data normalization served as positive control of apoptosis induction. Reverse transcription and RT-quantitative PCR Autophagy analysis by transmission electron microscopy Total RNA was extracted using RNA isolation RNeasy micro SKBr-3 cells were transfected with SNRPE-specific validated Kit (Qiagen) at 96 hours after transfection. mTOR mRNA siRNA or nontargeting control siRNA as described above. expression was analyzed by real-time quantitative PCR (RT- Forty-eight hours after transfection, cells were fixed with qPCR) in SKBr-3 cells transfected by SNRPE or control siRNA. 1.6% glutaraldehyde in 0.1 mol/L phosphate buffer (pH 7.4) One microgram of total RNA from each sample was reverse- at room temperature for 1 hour. Samples were rinsed in the transcribed by superscript II reverse transcriptase (Invitrogen) same buffer then postfixed with osmium tetroxyde and 1.5% in the presence of random primers (Applied Biosystems, Life potassium ferrocyanide in 0.1 mol/L cacodylate buffer (pH 7.4) technologies). Quantitative PCR (qPCR) was carried out on an for 1 hour at room temperature. Cells were rinsed with distilled equivalent amount of 12.5 ng total RNA per tube in a final water, dehydrated in increasing concentrations of ethanol, and volume of 25 mL. Oligonucleotide primers and TaqMan probes embedded in epoxy resin. Embedded samples were then were designed using the PrimerExpress computer software conventionally processed for thin sectioning (70 nm), con- (Applied Biosystems) and purchased from MWG Biotech. The trasted by uranyl acetate and lead citrate, and observed with primers and probe sequences used for the qPCR for mTOR and Zeiss 902 electron microscopy in filtered zero loss mode using housekeeping genes are detailed in Supplementary Table S2. CDD megaviewIII camera and SIS system (Olympus). Autop- The relative expression of mTOR mRNA was determined by DDCt hagic cells were counted on 60 cells series per sample. The using the Ct value and the 2 method. The data are identification of autophagy was based on the observation of presented as the fold change in gene expression normalized macro-autophagic vacuoles, that is, multiple membranes vesi- using housekeeping genes (TBBP, RPLPO). cles engulfing a portion of the cytoplasm. Expression profiling of spliceosome-related genes Detection and quantification of acidic vesicular Two unpublished datasets of gene expression arrays were organelles with acridine orange staining using flow used to determine a correlation between spliceosome expres- cytometry sion and clinical characteristics as well as the proliferation To detect the presence of acidic vesicular organelles (AVO), biomarker MKI67. These two studies used Agilent 44K plat- tumor cells staining with acridine orange (Sigma-Aldrich) was form. The first study compared gene expression levels between conducted as described previously (25). Briefly, 96 hours after ovarian adenocarcinomas (n ¼ 21) and borderline (nonmalig- siRNA transfection, SKBr-3 cells were washed with PBS and nant) lesions (n ¼ 21). The second one evaluated gene expres- stained with 1 mg/mL acridine orange in serum-free medium sion in 148 breast cancer samples including 27 triple-negative (Sigma Aldrich) for 20 minutes at 37 C. Acridine orange– breast cancer (TNBC). In this latter study, the reference for stained cells were trypsinized, washed, and analyzed with flow gene expression was normal breast (pool of 3 samples). In cytometer. Green (510–530 nm, FL1-H channel) and red (>650 addition to these 2 studies, we used publically available gene nm, FL3-H channel) fluorescence emissions from 104 cells expression data available from Landi and colleagues (26) and illuminated with blue (488 nm) excitation light are measured Andre and colleagues (1). Lung dataset from Landi and col- with a FACScanto from BD Biosciences using DIVA software. leagues included 107 non–small cell lung carcinomas (n ¼ 58) Autophagy was quantified as a ratio between geomean fluo- and nontumor tissues (n ¼ 49). Breast dataset from Andre and rescence intensity of red versus green fluorescence (FL3/FL1). colleagues included 165 samples (120 cancers and 45 benign Treated cells with nutrient-free medium EBSS (Invitrogen) for lesions) obtained by fine-needle aspiration of breast lesion. 24 hours before the addition of acridine orange served as Spliceosome complex components (from Biocarta database) positive controls. were extracted from these datasets respectively. The log2 gene expression was analyzed and when one gene is represented by Gene expression microarray more than one probe set, the one with the highest variance was Transcriptome analysis following SNRPE depletion was used. To illustrate the gene expression distribution of the conducted using exon array (ExonHit Therapeutics). SKBr-3 spliceosome complex components, boxplot graphs depict this cells were transfected with SNRPE-specific validated siRNA or information according to sample types. nontargeting control siRNA as previously described. Total RNA was extracted using RNA isolation RNeasy micro Kit (Qiagen) Statistical analysis at 96 hours after transfection. The concentration of RNA was Effects of siRNA transfections on cell lines proliferation were measured using Nanodrop spectrophotometer. The quality of estimated by 2-way ANOVA. The effects of transfection con- RNA preparations was assessed using Lab-on-a-Chip Bioana- ditions were first analyzed by considering all cell lines as a first lyser 2000 technology (Agilent Technologies). Aliquots of factor and the transfection conditions as a second factor extracted total RNA from the samples were sent to ExonHit (nontransfected cells, mock, unspecific siRNA). Here, and for Therapeutics for array processing. ExonHit staff processed to a each cell line, counting values were initially transformed into human Genome-Wide SpliceArray microarray on the Affy- proportions of the maximum value in each corresponding cell metrix platform using single color labeling. Processing includ- line and then linearized with a logit link. Next, the effects of the

www.aacrjournals.org Cancer Res; 73(7) April 1, 2013 2249

Quidville et al.

2 specific siRNA transfections were analyzed independently. consistently found to be among the top differentially expressed For each cell line, both nonspecific and specific siRNA counting genes in the 3 series of samples. We first used siRNA-mediated values were initially transformed into proportions of the depletion of SNRPE to knockdown SNRPE expression in the corresponding mock counting value. After a logit transforma- SKBr-3 and MDA-MB 468 breast cancer cell lines that were tion, cell lines and transfection (specific siRNA vs. nonspecific chosen because they are respectively HER2-amplified and siRNA) were considered as the 2 factors in a 2-way ANOVA. triple-negative. We also used the nontumoral MCF-10A breast cell line to compare the effects of depleting a spliceosome Results component in cancer cells with nontumoral cells. As revealed Genes involved in spliceosome machinery are by Western blot analysis, SNRPE protein level was efficiently overexpressed in malignant disease and define reduced by siRNA in the 3 cell lines (Fig. 3A). SNRPE knock- aggressive phenotypes down induced a major decrease of cell viability as compared We evaluated the expression of 15 genes related to spliceo- with mock in the 2 tumor cell lines [MDA-MB 468 (88.6% vs. somal assembly, according to the BioCarta database, including 102.1% in siControl cells) and SKBr-3 (58.1% vs. 104.9% in U1snRNP-specific protein (SNRNP70 or U1-70K), constitutive siControl cells)], whereas it had a much weaker effect (24.2% vs. splicing factors (U2AF1, U2AF2), splicing regulator (SFRS2), and 95.1% in siControl cells) in the nontumoral MCF-10A cells ¼ core spliceosomal components (SNRP-A/A1, B/B2, C, D1/D2/ (Pinteraction 0.00022). D3, E, F, and G). Differential gene expression analysis between To ascertain that this effect is due to spliceosome targeting, malignant and normal tissues was first conducted on (i) a we knocked down the expression of another core spliceosomal microarray dataset from Landi and colleagues comparing gene component. We chose SNRPD1, another upregulated core expression in 107 non–small cell lung carcinomas (n ¼ 58) and spliceosomal protein in malignant tumors (Fig. 3B). Here again, nontumor tissues (n ¼ 49) and (ii) an unpublished microarray following SNRPD1 knockdown, we observed a major reduction dataset comparing 148 breast cancers and normal breast of cell viability in the 2 breast tumoral cell lines MDA-MB 468 samples. We found that genes related to spliceosome assembly (78.4%) and SKBr-3 (57.7%) and a weaker effect (24.1%) in the ¼ components are overexpressed in lung adenocarcinoma (Fig. nontumoral MCF-10A cells (Pinteraction 0.00224). 1A) and breast cancer (Fig. 1B) as compared with nontumoral Of note expression of the genes encoding the other core tissue. Next, we used an unpublished ovary microarray dataset spliceosomal components (SNRP-A/A1, B2, C, D2/D3, F, and G) to compare gene expression between 21 invasive ovarian was assessed by RT-qPCR in SNRPE- or SNRPD1- depleted cancers and 21 borderline lesions. The borderline lesions are cells. No or only weak changes were observed in the expression defined as a subset of epithelial ovarian tumors of low malig- of these genes (Supplementary Fig. S2), showing that the nant potential representing an intermediate stage between siRNAs used in this study and targeting SNRPE or SNRPD1 fully benign and fully malignant adenocarcinomas. Strikingly, had no off-target effects on related components of the spliceosome assembly components are overexpressed in ovar- spliceosome. ian cancer as compared with borderline tumors (Fig. 1C), We next evaluate whether spliceosome depletion impacts on suggesting that overexpression of spliceosomal machinery cell proliferation of additional breast, melanoma, and lung components was related to malignancy. No major correlation tumoral cell lines. A significant decrease in cell viability was was found between the proliferation marker MKI67 and the observed in all tested tumoral cell lines following SNRPE and spliceosome-related genes in the 3 datasets described (Sup- SNRPD1 knockdown (Supplementary Fig. S3). plementary Fig. S1). Altogether, these results revealed that all tumoral cell lines We finally assessed whether overexpression of spliceosome were sensitive to spliceosome depletion, independently of their assembly components defines a more aggressive disease. To do proliferation rate (Supplementary Fig. S4). Moreover, a non- this, we looked at the correlation between the expression of the tumoral breast cell line (MCF-10A) was less sensitive to spli- 15 genes and clinical characteristics. For this analysis, we used ceosome depletion than tumoral cell lines, suggesting that the a public dataset of 120 breast cancer samples and 45 benign antiproliferative effect of spliceosome depletion was specificto breast lesions (1). Interestingly, unsupervised clustering malignant cells. defined 2 populations, one of which overexpressed the spliceosome assembly components (Fig. 2). This spliceosome positive SNRPE or SNRPD1 targeting mediates cancer cell death cluster was enriched in patients presenting with a TNBC (lack through autophagy expression of estrogen receptor, progesterone receptor, and To further investigate the cellular mechanism that led to a HER2) and Her2-overexpressing breast cancer (47% of the decrease in cell viability, we assessed whether the spliceosome spliceosome-positive breast cancers). These findings sug- regulated apoptosis or cell cycle. As revealed by fluorescence- gested that overexpression of spliceosome assembly compo- activated cell-sorting (FACS) analysis, SNRPE knockdown nents were not only overexpressed in malignant diseases but induced a marked increase in the sub-G1 population in also could define a subset of highly aggressive cancers. SNRPE-depleted cells (19.32%) as compared with control cells (2.62%), indicative of cell death occurrence (Fig. 4A). Apoptotic Knockdown of components of the spliceosome decreases cell death was analyzed by using double cell labeling with viability of cancer cells, but not of nonmalignant cells Annexin V-PE and 7-AAD. SNRPE depletion did not lead to a To investigate the role of the spliceosome, we focused on the significant increase in the amount of Annexin V–positive cells core spliceosomal components SNRPE and SNRPD1 that were (31.44% as compared with 39.7% for control cells; Fig. 4B) and

2250 Cancer Res; 73(7) April 1, 2013 Cancer Research

Spliceosome as a Target for Anticancer Treatment

A Lung dataset (scaled log2 values) log2 (scaled Spliceosome components Spliceosome components gene expression

Figure 1. Genes involved in spliceosome assembly are overexpressed in malignant disease. B Gene expression of spliceosome 4 Breast dataset components related genes (from Biocarta database) using publically available gene expression data sourced from Landi and colleagues (25) non–small cell lung carcinomas (n ¼ 58) and nontumor tissues (n ¼ 49). A, unpublished breast dataset from Gustave Roussy

Cancer Institute of 148 samples 02 including 27 TNBCs and 121 ERþ and/or HER2-overexpressing breast cancers B, unpublished ovarian dataset from Gustave Roussy Cancer Institute including 21 invasive -2 ovarian cancers and 21 borderline tumors (C). In each panel, box plots (scaled log2 ratio tumors vs reference) ratio tumors log2(scaled represent gene expression (log2 values) between malignant and expression gene components Spliceosome -4 normal tissue [lung (A) and breast (B) datasets] or borderline tumors [ovarian (C) dataset]. C Ovary dataset (scaled log2 values) log2 (scaled Spliceosome components gene expression gene components Spliceosome

www.aacrjournals.org Cancer Res; 73(7) April 1, 2013 2251

Quidville et al.

Figure 2. Expression of spliceosome assembly components deﬁnes aggressive disease. Unsupervised clustering of the 15 genes involved in spliceosome assembly (from BioCarta database) and correlated clusters to clinical characteristics Cluster Cluster using a public breast dataset Spliceosome negative Spliceosome positive including 120 breast cancer (n = 93) (n = 72) samples and 45 benign breast Histologic diagnosis: lesions (1). One cluster 39 (42%) 6 (8%) Benign lesions overexpressed spliceosome assembly components. This Malignant lesions 54 (58%) 66 (92%) spliceosome-positive cluster was Molecular class of breast enriched in aggressive breast cancer 3 (5% of BC) 12 (18% of BC) diseases including triple-negative Triple-Negative BC and HER2-overexpressing breast cancers (BC). HER2-overexpressing BC 6 (10% of BC) 19 (29% of BC)

Triple negative or HER2-overerexpressing 9 (15% of BC) 31 (47% of BC) BC

therefore did not induce typical apoptosis. Similar effects were increase of orange red fluorescence as compared with control obtained in SNRPD1-depletd cells (Supplementary Fig. S5). cells, indicating formation of acidic autophagolysosomal We next analyzed autophagy, another form of cell death, by vacuoles following spliceosome depletion (Fig 5C). using transmission electron microscopy (TEM) approach that Overall, these data indicated that targeting core spliceoso- allowed detecting the presence of autophagic vacuoles in cell mal component such as SNRPE or SNRPD1 could mediate cytoplasm. Autophagic features such as intracellular vacuoli- cancer cell death through autophagy. zation and double-membrane structures referred as autophagosomes (early autophagic compartment) were observed in SNRPE knockdown leads to mTOR downregulation SNRPE-depleted cells (Fig. 5A). The amount of cells containing To identify the molecular mechanisms leading to autophagy, autophagosomes was 2-fold higher in SNRPE-depleted cells we analyzed gene expression profiles in SNRPE-depleted SKBr- (9%) than in control cells (4%) at 48 hours posttransfection. 3 cell line. Transcriptome analysis on exon microarrays The cytoplasmic accumulation of autophagosomes following revealed that 51 genes were overexpressed and 41 genes were SNRPE depletion is consistent with an induction of the autop- downregulated (Supplementary Fig. S1) following SNRPE hagic pathway. depletion. Looking for genes involved in autophagy, we found We next evaluated the expression of the microtubule-asso- that mTOR was among the most downregulated gene in ciated protein 1 light chain 3 (LC3), commonly used as a SNRPE-depleted cells (Supplementary Fig. S6). Importantly, marker of the autophagic process that directly correlates with this decrease was robust as evidenced by the similar behavior the extent of autophagosome formation (27, 28). LC3 protein of all the 142 probes covering the mTOR mRNA (Fig. 6A). This exists in 2 forms: a cytosolic 18-kDa LC3-I and a 16-kDa LC3-II decrease (5-fold) in mTOR mRNA abundance in SNRPE-deplet- that binds to autophagosomes membranes during autophagy. ed cells was confirmed by RT-qPCR (Fig. 6B). Furthermore, a SNRPE or SNRPD1 depletion led to an increase in the LC3-II decrease in mTOR protein abundance was observed in SNRPE- form that indicated an autophagosome formation and an depleted cells (Fig. 6C). induction of autophagy (Fig. 5B). We next hypothesized that NMD could mediate the mTOR To further reinforce the data concerning autophagy, we loss induced by SNRPE-knockdown, as a recent study has evaluated the presence of late (autolysosome) markers of shown that knockdown of SNRPB, another core SNRP spliceo- autophagy using cell staining with acridine orange, a lysotropic somal protein, resulted in the inclusion of many PTC-contain- dye that accumulates in acidic organelles formed during ing alternative exons (29). These aberrant PTC-containing autophagy. The quantitative analysis of acridine orange– mRNAs are subsequently degraded by the NMD machinery stained cells showed that knockdown of SNRPE or SNRPD1 (29). To determine whether NMD was involved in the disap- protein expression in SKBr-3 cells induced a significant pearance of the mTOR mRNA upon SNRPE depletion, we

2252 Cancer Res; 73(7) April 1, 2013 Cancer Research

Spliceosome as a Target for Anticancer Treatment

A siRNA siRNA siRNA Mock Mock Control Mock Control Control SNRPE SNRPE SNRPE

SNRPE

SNRPE relative level 0.91 0.84 0.36 0.88 0.88 0.61 0.95 1.00 0.10

β-Actin GAPDH GAPDH

MCF-10A MDA-MB 468 SKBr-3 nontumoral triple-negative Her2-amplified 1.2

1.0 Figure 3. Effect of siRNA-mediated depletion of SNRPE or SNRPD1 on 0.8 cell viability in malignant and nonmalignant breast cell lines. 0.6 Breast tumor cell lines (MDA-MB 468, SKBr-3) and nontumoral breast 0.4 cell line MCF-10A were transfected Viability (vs. mock) with siRNAs targeting SNRPE and SNRPD1 or control siRNA as 0.2 indicated. Depletion of SNRPE or D1 was analyzed by Western blotting. 0.0 siControl siSNRP-E means Cell viability was assessed 96 hours after transfection by counting using the Trypan blue exclusion method or B siRNA siRNA siRNA crystal violet staining. The effect of SNRPE or SNRPD1 depletion on cell viability versus mock was estimated Mock Control SNRPD1 Mock Mock Control SNRPD1 Control SNRPD1 by 2-way ANOVA. Results for SNRPE and SNRPD1 are shown in A and B, SNRPD1 respectively. The levels of SNRPE SNRPD1 relative level 1.09 1.01 0.35 0.6 0.91 0.22 0.93 1.20 0.10 and SNRPD1 were quantiﬁed by densitometry using ImageJ software β-Actin GAPDH GAPDH and expressed relative to actin or GAPDH signals. MCF-10A MDA-MB 468 SKBr-3 nontumoral triple-negative Her2-amplified 1.2

1.0

0.8

0.6

0.4 Viability (vs. mock) 0.2

0.0 siControl siSNRP-D1 means disrupted NMD using cycloheximide in SNRPE-depleted cells. increased upon cycloheximide treatment (Supplementary Fig. The translation inhibitor cycloheximide inhibits mRNA deg- S7). These data indicated that the cycloheximide-dependent radation by the NMD pathway, as this process requires increase in mTOR mRNA expression was speciﬁc to SNRPE ongoing translation. Following SNRPE depletion, NMD inhi- knockdown. bition by cycloheximide resulted in an approximately 2.5-fold To next evaluate whether SNRPE depletion could affect increase in the relative abundance of mTOR mRNA, as downstream targets of the mTOR signaling pathway, we compared with 1.5-fold in control cells (Fig. 6D), suggesting assessed the phosphorylation of 4E-BP1 that lies just down- that a substantial fraction of the mTOR pre-mRNA was stream of mTOR. SNRPE depletion induced a dephosphoryla- aberrantly spliced into one or more mRNA isoforms that tion of 4E-BP1 (Fig. 6C) that indeed indicated a blockage of were degraded by NMD. Our results also showed that mRNA mTOR signaling. Overall, these data indicated a global SNRPE- expression for 2 control genes (GUSB and RPLPO) was not dependent deregulation of mTOR signaling.

www.aacrjournals.org Cancer Res; 73(7) April 1, 2013 2253

Quidville et al.

A Untransfected SKBr-3 Control siRNA SNRPE siRNA

M1=G1: 58.88% M1=G1: 58.50% M1=G1: 46.15% M2=S: 24.01% M2=S: 22.88% M2=S: 16.50%

M3=G2–M: 12.81% M3=G2–M: 10.28% M3=G2–M: 15.17%

M4=SubG1: 2.62% M4=SubG1: 2.69% M4=SubG1: 19.32%

Figure 4. Cell-cycle analysis and apoptosis detection following SNRPE depletion. A, cell-cycle M4 M4 analysis was used to monitor changes in the DNA of SKBr-3 DNA content DNA content DNA content cells harvested at 72 hours posttransfection with 50 nmol/L of SNRPE targeting siRNA or Untransfected Staurosporine B nontargeting control siRNA as indicated. B, apoptosis detection 4.08% 11.69% 3.83% 73.33% was quantitated using PE Annexin V/7-AAD staining followed by ﬂow cytometric analyses in cells at 72 hours posttransfection with 50

7-AAD nmol/L SNRPE targeting siRNA or 7-AAD nontargeting control siRNA. The numbers show the percentages of cells in each quadrant (bottom left, 69.99% 15.25% 1.04% 21.80% 7AAD/PE, intact cells; bottom right: 7-AADþ/PE, early PE Annexin V PE Annexin V apoptotic cells; top left, 7-AAD / PEþ, necrotic cells; top right, Control siRNA SNRPE siRNA 7-AADþ/PEþ, late apoptotic or necrotic cells). Staurosporine 4.37% 18.79% 2.82% 17.46% (5 mmol/L)-treated SKBr-3 cells served as positive control of apoptosis induction. 7-AAD 7-AAD

55.93% 20.91% 65.75% 13.98%

PE Annexin V PE Annexin V

Discussion malignant diseases. Indeed, mutations in genes encoding the In the present study, we report that components of the Sm U2AF35, ZRSR2, SRSF2, and SF3B1 splicing factors have been core spliceosomal machinery are overexpressed in cancer, found in hematologic malignancies (16). Sampath and collea- regulate mTOR, and impact on cell proliferation and autop- gues reported that the splicing factor SPF45 was overexpressed hagy (Fig. 6E). in carcinoma from various origins including breast, lung, ovary, Indeed, gene expression proﬁling using breast and lung colon, and prostate (30). This group also reported that SPF45 datasets revealed that genes related to the core spliceosomal overexpression was associated with resistance to cyclophos- components are highly expressed in lung and breast cancer in phamide (30). More recently, Niu and colleagues reported that contrast to nontumoral tissue. Moreover, gene expression proteins involved in splicing machinery were overexpressed in analyses using 2 other datasets of ovarian and breast cancers 8 samples of bladder cancer, as compared with normal tissue indicated that overexpression of genes encoding spliceosome (31). The present study based on larger number of samples complex components could be associated with a more aggres- suggests that genes encoding the spliceosome core machinery sive phenotype of cancers and could deﬁne a subset of highly are overexpressed in solid tumors. Overexpression of spliceo- malignant cancers. Among these genes involved in spliceosomal component is not only a hallmark of cancer but also some complex, those encoding the core spliceosomal proteins could be a therapeutic target. Indeed, we have shown here that SNRPE and SNRPD1 were consistently found to be among the depletion of several components of the core spliceosome (i.e., top differentially expressed genes in the 4 series of samples. SNRPE or SNRPD1) has a selective action on the proliferation Several studies have previously suggested that components of transformed cell lines but not of an immortalized nontu- of the splicing machinery are mutated or overexpressed in morigenic cell line (MCF-10A). In addition, SNRPE depletion

2254 Cancer Res; 73(7) April 1, 2013 Cancer Research

Spliceosome as a Target for Anticancer Treatment

A Control cells SNRPE-depleted cells

Figure 5. Autophagy analysis following SNRPE or SNRPD1 depletion. A, SKBr-3 cells were transfected with SNRPE siRNA or nontargeting control siRNA. Autophagic events characterized by TEM are mainly observed in SNRPE- depleted cells (c, d, e, f) as compared with control cells (a, b). Cells are shown at 2 magnifications, 3,000 (a, c) and 20,000 (b, d). Two characteristic multimembrane vesicles (autophagosomes) are presented in e and f at high magnification (50,000). The double membrane is pointed out (white arrow). B, immunoblot analysis of LC3 protein conversion. SKBr-3 cells B siRNA siRNA were transfected with SNRPE, Control SNRPE Control SNRPD1 SNRPD1 siRNA, or nontargeting kDa kDa control siRNA and then whole-cell 15 SNRPE 11 SNRPD1 protein lysates were immunoblotted with LC3B, SNRPE, SNRPD1, and b LC3B I -actin antibodies. C, autophagy 18 LC3B I 18 fi 16 was quanti ed as a ratio between LC3B II LC3B II 16 geomean fluorescence intensity of red versus green fluorescence β β-Actin (FL3/FL1), showing an increase of -Actin 42 42 red/green (FL3/FL1) fluorescence ratio following depletion of SNRPE or SNRPD1 as compared with control C cells, as previously reported (25). 6 Nutrient-free medium EBSS treated siRNA siRNA cells served as positive controls. Control SNRPE Control SNRPD1 Depletion of SNRPE or D1 protein 5 SNRPE SNRPD1 level was analyzed by Western 4 blotting. Results were analyzed using ANOVA (GraphPad Prism 5.04 β-Actin β-Actin software). Differences between 3 means were tested by Mann– Whitney tests (, P < 0.01 vs. 2

control). Acridine orange (FL3/FL1) 1 Mock Control SNRPE SNRPD1 EBSS

siRNA Untransfected

leads to inhibition of mTOR expression. This effect may, in part, stress (33). Consistent withthis observation, wehave shown here be the consequence of the production of aberrantly spliced that targeting SNRPE or SNRPD1 leads to cell death through mTOR mRNA isoforms degraded by the NMD pathway, as we autophagy. Interestingly, a recent report using a short hairpin have found here that NMD inhibition by cycloheximide results in RNA screening procedure in Drosophila melanogaster identiﬁed an increase in the relative abundance of mTOR mRNA. Further- the spliceosome machinery as one of the main pathway regu- more, a recent study reports that regulation of mTOR alternative lating the mTORC1 signaling pathway (34). This observation, splicing leads to the production of a short transcript degraded by together with our study, highlights the importance of the NMD during adipogenesis (32). Strikingly, by regulating mTOR, spliceosome–mTOR axis and reinforces the interest of the core SNRPE impacts on one of the most often deregulated signaling splicecosome as an interesting target in the control of eukaryotic pathways in cancer. mTOR is a key component that coordinately cell growth and death. regulates the balance between growth and autophagy in These ﬁndings open up new avenues for drug development. response to cellular physiologic conditions and environmental Several drugs have already been developed to target the

www.aacrjournals.org Cancer Res; 73(7) April 1, 2013 2255

Quidville et al.

A 10 Control cells SNRPE-depleted cells log2) 8 6 4 mTOR exon expression ( exon mTOR 2

0 20406080100120140 142 probes corresponding to mTOR gene on the array B C siRNA Control SNRPE kDa

1.2 SNRPE 11

1 mTOR 289 0.8

0.6 p-4E-BP1 15-20 4E-BP1 0.4

mTOR mRNA level mRNA mTOR 19 % p-4E-BP1 15-20 0.2 (ser65)

0 β-Actin Untransfected Control SNRPE 42

siRNA D E snRNPs (U1, U2, U4, U5, U6)

3 siRNAs E B B2 2,5 D1 Sm Core Proteins D2 G 2 F D3

1,5 Up in breast cancer Up in lung cancer 1 Up in ovarian cancer

mRNA expressionratio 0,5 CHX-treated cell/untreated cells 0 Survival Proliferation siControl SNRPE siRNA mTOR Autophagy

Figure 6. Knockdown of SNRPE alters mTOR expression and downstream signaling pathway. A, analysis of mTOR expression by exon arrays in SNRPE- depleted or control (nontargeting siRNA) SKBr-3 cells. mTOR expression was analyzed by exon arrays. mTOR expression is represented in a green–red heat color coding from low expression (light green) to high expression (red color) for transfected (~) and control (*) cells separately. The x-axis corresponds to the rank of mTOR probes in the control cells sorted by their expression in an increasing order. B, mTOR mRNA expression was analyzed by RT-qPCR in SNRPE-depleted or control (nontargeting siRNA) SKBr-3 cells. The graph represents mTOR mRNA level (fold change) normalized using housekeeping genes. C, expression of SNRPE and different mTOR downstream proteins was analyzed by Western blotting in SNRPE-depleted or control (nontargeting siRNA) SKBr-3 cells. D, for NMD inhibition experiments, cells were transfected with specific SNRPE siRNA or nontargeting control siRNA. Three days later, cells were treated for 8 hours with 100 mg/mL cycloheximide (CHX) or with an equivalent amount of DMSO as a control. mTOR RNA expression was detected by RT-PCR. Quantifications of the RT-PCR signals were calculated from RT-PCR assays carried out on samples from 2 independent transfections. The graph represents the ratio of mTOR mRNA expression in cells treated with cycloheximide as compared with untreated cells and shows that NMD inhibition resulted in an approximately 2.5-fold increase in the relative abundance of mTOR mRNA in SNRPE-depleted cells as compared with 1.5-fold increase in control cells. E, schematic representation of the sequence of processes leading from targeting SNRPE and SNRPD1 by specific siRNAs to the inhibition of mTOR survival pathway. The spliceosomal machinery complex is formed from 5 RNP subunits, termed uridine-rich (U-rich) small snRNPs. Each snRNP (U1, U2, U4, U5, and U6) consists of a uridine-rich snRNA complexed with a set of Sm proteins (B/B2, D1, D2, D3, E, F, and G) forming a ring core structure that encompasses RNA. The colored circles refer to the most significantly overexpressed Sm core proteins in the studied datasets (see Materials and Methods). Pink, breast cancer; blue, lung cancer; and green, ovarian cancer.

2256 Cancer Res; 73(7) April 1, 2013 Cancer Research

Spliceosome as a Target for Anticancer Treatment

splicing process. Using a chemical library, Soret and colleagues kinase inhibitors that would eventually lead to the inhibition of have reported that ellipticine inhibits several splicing events in spliceosome activity. vitro through targeting splicing regulatory factors such as SR Overall, our results suggest that core spliceosome machinery protein (35). This drug has been used in the last 30 years as could be an attractive therapeutic target in malignant solid chemotherapy agent in breast cancer. Consistent studies have tumors. These findings provide a rationale to design new drugs shown that ellipticine was associated with a strong antitumor targeting SNRPE, SNRPD1, and to develop spliceostatin A and effect in patients with breast cancer. More recently, 2 com- the pladienolide derivatives for patients with SNRPE/SNRPD1- pounds spliceostatin A and the pladienolide derivatives, were overexpressing cancers. independently reported to interact with SF3b, a subcomplex of U2 snRNP, which is an essential component of the spliceosome Disclosure of Potential Conflicts of Interest (25, 26). The binding of pladienolide and spliceostatin A to SF3b The sponsors did not participate in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and prepara- leads to an inhibition of the spliceosomal functions resulting in tion, review, or approval of the manuscript. Its contents are solely the respon- impaired splicing and thus altered gene expression patterns. sibility of the authors. No potential conflicts of interest were disclosed Pladienolide derivative compounds exhibited potent anti-cancer activity in several cancer cell lines and human tumor Authors' Contributions xenografts (36). Treatment of a drug screen panel of 39 cancer Conception and design: S. Vagner, F. Andre Development of methodology: V. Quidville, S. Alsafadi, S. Vagner, F. Andre cell lines with pladienolides resulted in major growth inhibi- Acquisition of data (provided animals, acquired and managed patients, tion with greatest activity against breast and lung cancer cell provided facilities, etc.): V. Quidville, S. Alsafadi, C. Durieu, E. Le Cam, S. Delaloge, M. Saghatchian, P. Morice, S. Vagner, F. Andre lines (37). Spliceostatin A affects also proliferation of numerous Analysis and interpretation of data (e.g., statistical analysis, biostatistics, cancer cell lines and inhibits tumor growth in various xenograft computational analysis): V. Quidville, S. Alsafadi, A. Goubar, F. Commo, C. models (38). On the basis of the promising preclinical data with Durieu, S. Baconnais, E. Le Cam, P. Dessen, S. Vagner, F. Andre Writing, review, and/or revision of the manuscript: V. Quidville, S. Alsafadi, respect to the potential anticancer activity of pladienolide, F. Commo, C. Durieu, E. Le Cam, S. Delaloge, S. Vagner, F. Andre phase I dose-escalation trials with pladienolide D were initi- Administrative, technical, or material support (i.e., reporting or orga- ated in Europe and United States. The previously mentioned nizing data, constructing databases): S. Alsafadi, A. Goubar, V. Scott, I. Girault, V. Lazar, P. Morice, S. Vagner, F. Andre drugs target splicing regulators [i.e., SRSF/serine-arginine-rich Study supervision: V. Lazar, S. Vagner, F. Andre (SR) splicing factors] or constitutive splicing factors (e.g., SF3b) Preliminary biologic results before the conception of the study: P. Pautier but not the core spliceosomal machinery. Also, as they are derived from natural compounds, they do not specifically Grant Support This work was supported by the donation program of "Operation Parrain- target splicing machinery. On the basis of the present study, Chercheur" and grants from INCA and Dassault foundation. there is a rationale to design chemical compounds that spe- The costs of publication of this article were defrayed in part by the payment of cifically target SNRPE and SNRPD1. Other therapeutic strat- page charges. This article must therefore be hereby marked advertisement in egies are possible to inhibit spliceosome activity. It has been accordance with 18 U.S.C. Section 1734 solely to indicate this fact. reported that spliceosome activity is regulated by kinases, Received June 25, 2012; revised December 5, 2012; accepted January 11, 2013; including SRPK1 (39). This could provide a rationale to develop published OnlineFirst January 28, 2013.

References 1. Andre F, Michiels S, Dessen P, Scott V, Suciu V, Uzan C, et al. Exonic 11. Skotheim RI, Nees M. Alternative splicing in cancer: noise, functional, expression profiling of breast cancer and benign lesions: a retrospec- or systematic? Int J Biochem Cell Biol 2007;39:1432–49. tive analysis. Lancet Oncol 2009;10:381–90. 12. Srebrow A, Kornblihtt AR. The connection between splicing and 2. Jurica MS, Moore MJ. Pre-mRNA splicing: awash in a sea of proteins. cancer. J Cell Sci 2006;119:2635–41. Mol Cell 2003;12:5–14. 13. Venables JP. Unbalanced alternative splicing and its significance in 3. Zhou Z, Licklider LJ, Gygi SP, Reed R. Comprehensive proteomic cancer. Bioessays 2006;28:378–86. analysis of the human spliceosome. Nature 2002;419:182–5. 14. Kalnina Z, Zayakin P, Silina K, Line A. Alterations of pre-mRNA splicing 4. Salgado-Garrido J, Bragado-Nilsson E, Kandels-Lewis S, Seraphin B. in cancer. Genes Chromosomes Cancer 2005;42:342–57. Sm and Sm-like proteins assemble in two related complexes of deep 15. Brinkman BM. Splice variants as cancer biomarkers. Clin Biochem evolutionary origin. EMBO J 1999;18:3451–62. 2004;37:584–94. 5. Hermann H, Fabrizio P, Raker VA, Foulaki K, Hornig H, Brahms H, et al. 16. Yoshida K, Sanada M, Shiraishi Y, Nowak D, Nagata Y, Yamamoto R, snRNP Sm proteins share two evolutionarily conserved sequence et al. Frequent pathway mutations of splicing machinery in myelodys- motifs which are involved in Sm protein-protein interactions. EMBO plasia. Nature 2011;478:64–9. J 1995;14:2076–88. 17. Wang L, Lawrence MS, Wan Y, Stojanov P, Sougnez C, Stevenson K, 6. Wahl MC, Will CL, Luhrmann R. The spliceosome: design principles of et al. SF3B1 and other novel cancer genes in chronic lymphocytic a dynamic RNP machine. Cell 2009;136:701–18. leukemia. N Engl J Med 2011;365:2497–506. 7. Dutertre M, Vagner S, Auboeuf D. Alternative splicing and breast 18. Graubert TA, Shen D, Ding L, Okeyo-Owuor T, Lunn CL, Shao J, et al. cancer. RNA Biol 2010;7:403–11. Recurrent mutations in the U2AF1 splicing factor in myelodysplastic 8. Pettigrew CA, Brown MA. Pre-mRNA splicing aberrations and cancer. syndromes. Nat Genet 2012;44:53–7. Front Biosci 2008;13:1090–105. 19. Quesada V, Conde L, Villamor N, Ordonez GR, Jares P, Bassaganyas 9. Fackenthal JD, Godley LA. Aberrant RNA splicing and its functional L, et al. Exome sequencing identifies recurrent mutations of the consequences in cancer cells. Dis Model Mech 2008;1:37–42. splicing factor SF3B1 gene in chronic lymphocytic leukemia. Nat 10. Pajares MJ, Ezponda T, Catena R, Calvo A, Pio R, Montuenga LM. Genet 2012;44:47–52. Alternative splicing: an emerging topic in molecular and clinical oncol- 20. David CJ, Manley JL. Alternative pre-mRNA splicing regulation in can- ogy. Lancet Oncol 2007;8:349–57. cer: pathways and programs unhinged. Genes Dev 2010;24:2343–64.

www.aacrjournals.org Cancer Res; 73(7) April 1, 2013 2257

Quidville et al.

21. Schwerk C, Schulze-Osthoff K. Regulation of apoptosis by alternative 31. Niu HT, Yang CM, Chen B, Dong Q. Biomarker research and some pre-mRNA splicing. Mol Cell 2005;19:1–13. deduction in superficial bladder cancer cells combined with corre- 22. Baker KE, Parker R. Nonsense-mediated mRNA decay: terminating sponding stroma. Cancer Biomark 2011;10:109–16. erroneous gene expression. Curr Opin Cell Biol 2004;16:293–9. 32. Huot ME, Vogel G, Zabarauskas A, Ngo CT, Coulombe-Huntington J, 23. Maquat LE. Nonsense-mediated mRNA decay. Curr Biol 2002;12: MajewskiJ,etal.TheSam68STARRNA-bindingproteinregulatesmTOR R196–7. alternative splicing during adipogenesis. Mol Cell 2012;46:187–99. 24. Lareau LF, Brooks AN, Soergel DA, Meng Q, Brenner SE. The coupling 33. Jung CH, Ro SH, Cao J, Otto NM, Kim DH. mTOR regulation of of alternative splicing and nonsense-mediated mRNA decay. Adv Exp autophagy. FEBS Lett 2010;584:1287–95. Med Biol 2007;623:190–211. 34. Lindquist RA, Ottina KA, Wheeler DB, Hsu PP, Thoreen CC, Guertin 25. Traganos F, Darzynkiewicz Z. Lysosomal proton pump activity: supra- DA, et al. Genome-scale RNAi on living-cell microarrays identifies vital cell staining with acridine orange differentiates leukocyte sub- novel regulators of Drosophila melanogaster TORC1-S6K pathway populations. Methods Cell Biol 1994;41:185–94. signaling. Genome Res 2011;21:433–46. 26. Landi MT, Dracheva T, Rotunno M, Figueroa JD, Liu H, Dasgupta A, 35. Soret J, Bakkour N, Maire S, Durand S, Zekri L, Gabut M, et al. Selective et al. Gene expression signature of cigarette smoking and its role in modification of alternative splicing by indole derivatives that target lung adenocarcinoma development and survival. PLoS One 2008;3: serine-arginine-rich protein splicing factors. Proc Natl Acad Sci U S A e1651. 2005;102:8764–9. 27. Kabeya Y, Mizushima N, Ueno T, Yamamoto A, Kirisako T, Noda T, 36. Iwata M, Ozawa Y, Uenaka T, Shimizu H, Niijima J, et al. E7107, a new et al. LC3, a mammalian homologue of yeast Apg8p, is localized in 7-urethane derivative of pladienolide D, displays curative effect against autophagosome membranes after processing. EMBO J 2000;19: several human tumor xenografts. Proc Am Assoc Cancer Res 2004;45. 5720–8. Abstract nr 2989. 28. Barth S, Glick D, Macleod KF. Autophagy: assays and artifacts. 37. Mizui Y, Sakai T, Iwata M, Uenaka T, Okamoto K, Shimizu H, et al. J Pathol 2010;221:117–24. Pladienolides, new substances from culture of Streptomyces platensis 29. Saltzman AL, Pan Q, Blencowe BJ. Regulation of alternative splicing by Mer-11107. III. In vitro and in vivo antitumor activities. J Antibiot (Tokyo) the core spliceosomal machinery. Genes Dev 2011;25:373–84. 2004;57:188–96. 30. Sampath J, Long PR, Shepard RL, Xia X, Devanarayan V, Sandusky 38. Kaida D, Motoyoshi H, Tashiro E, Nojima T, Hagiwara M, Ishigami K, GE, et al. Human SPF45, a splicing factor, has limited expression in et al. Spliceostatin A targets SF3b and inhibits both splicing and normal tissues, is overexpressed in many tumors, and can confer a nuclear retention of pre-mRNA. Nat Chem Biol 2007;3:576–83. multidrug-resistant phenotype to cells. Am J Pathol 2003;163: 39. Ghosh G, Adams JA. Phosphorylation mechanism and structure of 1781–90. serine-arginine protein kinases. FEBS J 2011;278:587–97.

2258 Cancer Res; 73(7) April 1, 2013 Cancer Research

Targeting the Deregulated Spliceosome Core Machinery in Cancer Cells Triggers mTOR Blockade and Autophagy

Virginie Quidville, Samar Alsafadi, Aïcha Goubar, et al.

Cancer Res 2013;73:2247-2258. Published OnlineFirst January 28, 2013.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-12-2501

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2013/01/28/0008-5472.CAN-12-2501.DC1

Cited articles This article cites 38 articles, 8 of which you can access for free at: http://cancerres.aacrjournals.org/content/73/7/2247.full#ref-list-1

Citing articles This article has been cited by 10 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/73/7/2247.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/73/7/2247. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.