Anticancer Drugs Affect the Alternative Splicing of Bcl-X and Other Human Apoptotic Genes

1398

Anticancer drugs affect the alternative splicing of Bcl-x and other human apoptotic genes

Lulzim Shkreta,1 Ulrike Froehlich,2 E´ ric R.Paquet, 2 Introduction 1 1,2 Johanne Toutant, Sherif Abou Elela, Apoptosis or programmed cell death is a major pathway in and Benoit Chabot1,2 the complex homeostatic balance between cellular prolif-

1 eration and cell death. Conserved throughout evolution, RNA Group, De´partement de Microbiologie et d’Infectiologie and apoptosis represents the endpoint of a cascade of events 2Laboratoire de Ge´nomique Fonctionnelle, Faculte´deMe´decine et desSciencesde la Sante´, Universite´de Sherbrooke, triggered by a variety of intracellular or extracellular Sherbrooke, Quebec, Canada stimuli including DNA damage, telomere dysfunction, oncogenic mutations, growth factor depletion, and hypoxia (1–4). Apoptosis is realized through the extrinsic (death Abstract receptor–dependent) or intrinsic (mitochondria–mediated) Inducing an apoptotic response is the goal of most pathway, both routes activating caspases to execute cell current chemotherapeutic interventionsagainstcancer. death. However, little isknown about the effect of chemother- Mainstream anticancer agents act by triggering the apeutic agentson the alternative splicingof apoptotic apoptotic pathway (5–9). However, intrinsic alterations genes. Here, we have tested 20 of the mainstream in the apoptotic pathway are a hallmark of cancer cells anticancer drugsfor their ability to influence the (2, 10, 11) and are considered to be a major cause of drug production of Bcl-x splice isoforms. We find that many inefficacy (12, 13). Alternatively or in addition, the drugs shift splicing toward the proapoptotic Bcl-xS splice promotion of genomic instability by some cytotoxic agents variant in 293 cells. The drugs modulate splicing may provide an opportunity for the emergence of decisions most likely through signaling events because additional lesions in the apoptotic pathway, in turn giving the splicing switch is not compromised by inhibiting rise to even more aggressive cancers that become resistant de novo protein synthesis or the activity of caspases. to subsequent therapies (14). Several drugs also shift Bcl-x splicing in cancer cell lines The majority of the genes involved in apoptosis are (MCF-7, HeLa, PC-3, PA-1, and SKOV-3), but the set of alternatively spliced, in many cases to produce isoforms active drugs varies between cell lines. We also examined with opposing functions. Examples of this type are found in the effect of anticancer agentson the alternative splicing every category of apoptotic regulators including transmem- of 95 other human apoptotic genesin different cell lines. brane receptors (e.g., Fas, Fas ligand, and LARD), adaptor Almost every drug can alter a subset of alternative molecules (e.g., Bcl-x, Bak, Apaf1, survivin, Mcl-1, and splicing events in each cell line. Although drugs of the TRAF2), caspases (e.g., caspase-1, caspase-2, caspase-6, same class often influence the alternative splicing of the caspase-7, caspase-8, and caspase-9), and executors (e.g., same units in individual cell lines, these units differ FLIP and ICAD; ref. 15). The alternative splicing of considerably between cell lines, indicating cell line– apoptotic genes is often altered in cancers (16–18). specific differences in the pathways that control splicing. Although a direct contribution of these changes to [Mol Cancer Ther 2008;7(6):1398–409] carcinogenesis and tumor development is in general poorly documented, the overexpression of specific antiapoptotic splice isoforms can protect cancer cells and contribute to their resistance to chemotherapeutic drugs. For example, the antiapoptotic Bcl-xL isoform is overexpressed in many types of cancers (for examples, see refs. 19–21), and Received 2/26/08; revised 4/10/08; accepted 4/15/08. increasing the expression of the proapoptotic Bcl-xS variant Grant support: Canadian Institute of Health Research (B. Chabot) and can kill cancer cells or improve their sensitivity to Genome Canada and Genome Que´bec. chemotherapeutic agents (17, 22, 23). Likewise, over- Note: S.A. Elela is Chercheur Boursier Senior of the Fonds de la recherche en sante´du Que´bec. B. Chabot is a Canada Research Chair in Functional expressing the proapoptotic variant of caspase-2 in lung Genomics. cancer cells sensitizes them to apoptosis on treatment with The costs of publication of this article were defrayed in part by the cisplatin (24). payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to Although alternative splicing is an important mechanism indicate this fact. regulating apoptosis, the effect of chemotherapeutic agents Requests for reprints: Benoit Chabot, De´partement de Microbiologie on the alternative splicing of apoptotic genes has rarely et d’Infectiologie, Faculte´deMe´decine et des Sciences de la Sante´, been examined. Camptothecin, homocamptothecin deriva- Universite´de Sherbrooke, 3001 12th Avenue North, Sherbrooke, Quebec, Canada J1H 5N4. Phone 819-564-5295; Fax: 819-564-5392. tives, etoposide, amsacrine, doxorubicin, and mitoxantrone E-mail: [email protected] favor the formation of the proapoptotic caspase-2 isoform Copyright C 2008 American Association for Cancer Research. in U937 and HeLa cells (25, 26). A shift toward the doi:10.1158/1535-7163.MCT-08-0192 proapoptotic Bcl-xS isoform occurs when P388 cells are

MolCancerTher2008;7(6).June2008

treated with a topoisomerase I inhibitor (27) and when Cancer Center; ref. 31). EcR-293 and HeLa S3 cells were cardiomyocytes are treated with daunorubicin (28). The grown in DMEM. MCF-7 cells were grown in Eagle’s MEM gemcitabine-mediated increase in ceramide levels also supplemented with 10 Ag/AL insulin. PC-3 cells were stimulates the production of the proapoptotic splice iso- grown in Ham’s medium/nutrient mixture F-12. PA-1 and forms of Bcl-x and caspase-9 in A549 cells (29). Finally, we SKOV-3 cells were grown in DMEM/F-12. All media were have recently reported that the apoptotic inducer and supplemented with 10% fetal bovine serum. To achieve protein kinase C inhibitor staurosporine can promote a f60% confluence before drug treatment, cells were seeded splicing switch in 293 cells to favor the production of the in six-well plates (35 mm) at the following densities: EcR- Â 5 proapoptotic Bcl-xS isoform (30). 293 cells at 2 10 per well, HeLa and MCF-7 cells at To further assess the relationship between chemothera- 1 Â 105 per well, PC-3 cells at 5 Â 104 per well, PA-1cells at peutic agents and the alternative splicing of apoptotic 7.5 Â 104 per well, and SKOV-3 cells at 6 Â 104 per well. genes, we have examined the effect of 20 mainstream Chemotherapeutic Drugs and Other Compounds anticancer drugs. We find that many anticancer drugs can Drugs used and their modes of action are listed in Bcl-x shift pre-mRNA splicing to the proapoptotic Bcl-xS Table 1. Drugs were obtained from the Service Pharmaceu- splice variant in different cell lines. Interrogating other tique du Centre de Chimiothe´rapie at the Centre Hospital- alternative splicing units reveals global effects of anticancer ier de l’Universite´ de Sherbrooke. Drugs were diluted drugs on alternative splicing. according to instructions from the manufacturers immedi- ately before addition to cell culture medium. Cells were j treated with individual drugs for 24 h at 37 C, 5% CO2. The Materials and Methods final concentrations of drugs used were 1 Amol/L Cell Lines and Cultures gemcitabine, 5 Amol/L cytaralbine, 5 Amol/L capecitabine, The human 293 cell line was from Invitrogen (EcR-293) 5 Amol/L methotrexate, 5 Amol/L 5-fluoruracil, 10 Amol/L and the human carcinomas of breast (MCF-7), prostate (PC- vinorelbine, 10 Amol/L vincristine, 10 Amol/L paclitaxel, 3), ovary (PA-1), and cervix (HeLa S3) were from the 10 Amol/L docetaxel, 0.5 Amol/L topotecan, 10 Amol/L American Type Culture Collection. The ovarian cancer cell etoposide, 25 Amol/L daunorubicin, 5 Amol/L epirubicin, line SKOV-3ip1(herein called SKOV-3) was kindly provid- 5 Amol/L dactinomycin, 10 Amol/L cisplatin, 10 Amol/L ed by Janet Price (The University of Texas M. D. Anderson oxaliplatin, 50 Amol/L cyclophosphamide, 10 Amol/L

Table 1. Classification and characteristics of the anticancer agents used in this study

Class Type Target/mode of action Drug name 293 MCF-7 PC-3 PA-1SKOV-3 Total

Antimetabolites Pyrimidine Gemcitabine 21 10 13 19 9 72 Pyrimidine Cytaralbine 10 8 10 13 14 55 Pyrimidine Capecitabine 15 12 11 12 13 63 DHFR Methotrexate 18 9 5 15 21 68 Pyrimidine Fluorouracil 19 9 13 11 10 62 Mitotic inhibitors Vinca alkaloids Tubulin Vinorelbine 14 8 9 0 14 45 Tubulin Vincristine 14 7 8 14 8 51 Taxanes Tubulin Paclitaxel 12 11 14 11 18 66 Tubulin Docetaxel 13 12 13 10 0 48 Topoisomerase Camptothecin Topoisomerase I Topotecan 25 6 10 12 14 67 inhibitors derivative Podophyllotoxine Topoisomerase II Etoposide 17 11 7 15 13 63 Intercalating agents Topoisomerase I/DNA Daunorubicin 13 10 14 6 9 52 synthesis Topoisomerase I/DNA Epirubicin 16 9 12 5 14 56 synthesis Antitumor DNA/RNA synthesis Dactinomycin 15 9 10 5 12 51 antibiotics Alkylating agents Platinum agents DNA Cisplatin 22 8 10 8 10 58 DNA Oxaliplatin 17 2 12 6 8 45 DNA cross-linkers DNA Cyclophosphamide 19 9 7 5 13 53 DNA Chlorambucil 14 7 9 9 9 48 DNA Dacarbazine 17 6 6 11 12 52 Hormone analogue Estrogen receptor Tamoxifen 16 9 16 8 11 60 Total 327 172 209 195 232 1,135

NOTE: The number of alternative splicing events affected by each drug in each cell line is indicated and is based on the RT-PCR analysis of 96 alternative splicing units in five cell lines.

Mol Cancer Ther 2008;7(6).June 2008

chlorambucil, 10 Amol/L dacarbazine, and 10 Amol/L of amplified products was done using the LabChip HT tamoxifen. Staurosporine (Roche Diagnostics) was used at DNA assay on an automated microfluidic station (Caliper). a final concentration of 50 nmol/L. All oligonucleotides used to amplify the set of 96 The inhibitor of caspases z-VAD-fmk (Calbiochem) was alternative splicing units are listed in Supplementary Table used at a final concentration of 50 Amol/L and cells were S23 and the alternative splicing units being monitored are treated for 2 h before anticancer drugs were added. After described in Revil et al. (30). treatments, two thirds of the cells from each well were Bioinformatic Analysis and Data Processing collected for protein extraction and the rest was used for A set of 96 alternative splicing units from human RNA extraction. Whole-cell extracts were prepared by apoptotic genes was selected from the AceView database sonication in Laemmli sample buffer and protein concen- (32). Sets of primers mapping in the exons flanking the tration was determined using the Lowry method. Protein alternative splicing events were designed using Primer3 samples (20 Ag/well) were fractionated on a 9% SDS-PAGE with default variables (33). Analysis of RT-PCR results (29:1acrylamide/bisacrylamide) and transferred onto a using the LabChip HT DNA assay yielded products with Hybond-C nitrocellulose membrane. Immunodetection was raw molarity and size values. We only considered products done according to standard protocols using a dilution of whose size ranged within 10% of the size of expected 1:500 of the rabbit anti-poly(ADP-ribose) polymerase products. In the case of products associated with multiple (PARP) antibody (Biosource) and 1:5,000 of the horseradish isoforms, the closest one, using absolute size difference peroxidase–conjugated anti-rabbit secondary antibody with the expected product, was chosen. The clustering of (Amersham Biosciences). the data of splicing units, where expression of a minimum The transcription elongation inhibitor DRB and the of two and maximum of three splicing isoforms were protein translation inhibitor cycloheximide were used at expected, was done with R (34) and RPy (R for Python) the concentrations indicated. using the function heat map. We used unscaled data and a Reverse Transcription-PCR Analysis hierarchical clustering algorithm that used binary distance For the conventional reverse transcription-PCR (RT-PCR) and ward hierarchical clustering. analysis of Bcl-x products, total RNA was extracted from cells using TRIzol (Invitrogen) and treated with DNase I for 10 min at 37jC followed by a treatment of 5 min at 70jC. Results Reverse transcription was done with a random hexamer Many Anticancer Agents Affect the Alternative using the OmniScript RT kit (Qiagen) at 37jC for 1h. One Splicing of Bcl-x tenth of the cDNA material was used for PCR. The primers We tested 20 drugs that belong to various groups based used to amplify the Bcl-x alternative splicing events were on their structure or putative mechanisms of action (see BX3 (ATGGCAGCAGTAAAGCAAGCG) and BX2 Table 1). The following cell lines were used: the human (TCATTTCCGACTGAAGAGTGA). [a-32P]dCTP (Amer- embryonic kidney 293 cell line, the cervical cancer cell line sham Biosciences) was added to PCR mixtures. PCR HeLa S3, the breast cancer cell line MCF-7, the prostate products were fractionated onto a 4% polyacrylamide gel. cancer cell line PC-3, and the ovarian cancer cell lines PA-1 Dried gels were exposed on screens that were scanned on a and SKOV-3. Our reporter for alternative splicing was the STORM 860 PhosphorImager (Amersham Biosciences) and endogenous Bcl-x pre-mRNA, which is spliced to produce bands were quantitated using the ImageQuant software. primarily the antiapoptotic Bcl-xL isoform in all untreated For the global analysis of splice isoforms, 2 Ag total RNA cell lines (Fig. 1). For each drug, we used concentrations was reverse transcribed using random hexamers and that were inferior to those promoting extensive cell death the Omniscript reverse transcriptase (Qiagen) in a final basedonvisualinspectionafter24hoftreatment. volume of 20 AL. cDNA (40 ng) was amplified with 0.375 Following RNA extraction, changes in the relative abun- A units/15 L HotStarTaq DNA polymerase (Qiagen) in the dance of the Bcl-xL and Bcl-xS splice variants were deter- buffer provided by the manufacturer, without MgCl2 and mined by RT-PCR (Fig. 1A; data not shown). Figure 1B in the presence of the specific primers for each splicing unit plots the average changes in the percentage of Bcl-xS (at concentrations ranging from 0.3 to 0.6 Amol/L) and product in triplicate for all drugs in all cell lines, except for deoxynucleotide triphosphates. The list of genes and the SKOV-3 cells, which were done in duplicate. We consid- sequence of the oligos used are listed in Supplementary ered a drug-induced change significant when it was >2 SD Table S2.3 Reactions were carried out in the GeneAmp PCR from the average values of the corresponding untreated system 9700 (Applied Biosystems). A first cycle of 15 s at samples (P < 0.05). 95jC was followed by 35 cycles of 30 s at 94jC, 30 s at 55jC, Twelve of the 20 anticancer drugs stimulated the j and 1min at 72 C. The reaction was ended with the production of Bcl-xS in 293 cells (Fig. 1B). Previously, we extension step of 10 min at 72jC. Visualization and analysis observed that staurosporine shifted Bcl-x splicing in 293 cells but not in cancer cell lines (30). Here, several compounds affected the levels of Bcl-x splice products in the cancer cell lines MCF-7, HeLa, PC-3, PA-1, and SKOV-3. 3 Supplementary material for this article is available at Molecular Cancer Oxaliplatin and cisplatin were active in all cell lines, Therapeutics Online (http://mct.aacrjournals.org/). whereas epirubicine was active in all cell lines, except PA-1.

MolCancerTher2008;7(6).June2008

Mol Cancer Ther 2008;7(6).June 2008

Gemcitabine, methotrexate, etoposide, daunorubicin, and the caspase-dependent degradation of components of the dactinomycin improved Bcl-xS production in at least four splicing machinery. Indeed, spliceosomal components and cell lines. Many drugs shifted the abundance of Bcl-x RNA-binding proteins can be targets for caspases (35–38). products in only one or two cell lines. For example, To ask whether caspase activation is required for the effect fluorouracil and cyclophosphamide had a significant effect on Bcl-x splicing in 293 cells, drug treatments were done only in 293 and SKOV-3 cells, respectively. Except for with or without the pancaspase inhibitor z-VAD-fmk. docetaxel in HeLa cells, drugs that interfere with microtu- Caspase activation was monitored through the cleavage bule formation did not affect Bcl-x splicing. This is unlikely of PARP, a preferential substrate for caspase-3 and caspase- to be due to suboptimal concentrations or poor drug intake 7. PARP cleavage was induced by all drugs, except because these drugs promoted PARP cleavage in 293 cells dactinomycin, dacarbazine, and fluorouracil (Fig. 2C). The (Fig. 2C). Moreover, concentrations of drugs that increase level of PARP cleavage did not correlate with the shift in cell death did not promote a switch toward Bcl-xS for drugs Bcl-x splicing, an indication that this shift might not be that failed to do so at lower concentrations (data not controlled by caspases. Although z-VAD-fmk itself had no shown). Thus, although drugs changed the relative significant effect on the splicing profile of Bcl-x in 293 cells, proportion of Bcl-x splice isoforms, the set of active drugs it prevented drug-induced PARP cleavage. In addition, vary between cell lines. z-VAD-fmk did not prevent the Bcl-x splicing shift The switch in the relative abundance of Bcl-x products induced by gemcitabine, methotrexate, topotecan, etopo- could reflect a change in alternative splicing or in the side, daunorubicin, epirubicine, cisplatin, and oxaliplatin differential stability of the Bcl-x mRNA isoforms. For (Fig. 2D). Thus, the effect of anticancer agents on staurosporine, we previously ruled out an effect on Bcl-x Bcl-x splicing in 293 cells does not require the activation mRNA stability because active transcription was required of caspases. de novo for the drug-mediated increase in Bcl-xS (30). To determine Given that protein synthesis and protein if this conclusion also applied to the set of anticancer drugs degradation are not required to obtain the drug-induced Bcl-x that increase the relative abundance of Bcl-xS in 293 cells, switch in splicing, the drugs are most likely activating cells were pretreated with the transcriptional elongation endogenous signaling pathways. Notably, a switch toward inhibitor DRB (Fig. 2A) before treatment with the antican- Bcl-xS was obtained by inhibiting protein kinase C or by cer drugs. The results indicate that all drug-induced activating Fyn kinase or protein phosphatase 1(29, 30, 39). increases in Bcl-xS require active transcription, supporting Whereas staurosporine and tamoxifen are known protein the view that the shifts are caused primarily by alterations kinase C inhibitors (40, 41), the signaling events mediating in splice site selection. the effect of the other drugs on Bcl-x alternative splicing Drug-Mediated Shifts in Bcl-x Splicing Do Not remain to be identified. Require New Protein Synthesis or the Activation of Global Effect of Chemotherapeutic Drugs on the Caspases Alternative Splicing ofApoptotic Genes To assess if de novo protein synthesis was required for the The effect of anticancer drugs on Bcl-x splicing led us to effect of the drugs on Bcl-x splicing in 293 cells, we tested ask if other alternative splicing units were sensitive to the effect of the active drugs in the presence of the protein anticancer agents. A collection of 96 splicing units from synthesis inhibitor cycloheximide. As reported previously apoptotic genes was interrogated in the 293, MCF-7, PC-3, (30), cycloheximide by itself stimulates the production of PA-1, and SKOV-3 cell lines. The relative abundance of the Bcl-xS mRNA product in a concentration-dependent man- RT-PCR products was determined by fractionation on ner (Fig. 2B, control lanes). However, when the stimulation microfluidic capillary workstations (42). Based on size obtained with an anticancer drug was superior to that of estimation, all products corresponded to known splicing cycloheximide alone, cycloheximide never reduced the events and no novel splicing products were uncovered in level of stimulation (Fig. 2B), suggesting that the drug- this analysis. RT-PCR yielding products that were below an induced splicing switches did not require de novo protein arbitrary threshold of sensitivity (2 nmol/L) were not synthesis. considered. Percentage values for splicing products were Chemotherapeutic agents may also exert their effect on compared with the values obtained with corresponding Bcl-x splicing by inducing apoptosis, which can promote mock-treated samples. Given that 95% of the variations

Figure 1. Alternative splicing of Bcl-x in cell lines treated with chemotherapeutic drugs. A, examples of RT-PCR amplification of endogenous Bcl-x splicing products. Total RNA from 293, MCF-7, and HeLa cells was analyzed by RT-PCR to obtain 32P-labeled products that were fractionated in native acrylamide gels. For PC-3, PA-1, and SKOV-3 cells, unlabeled RT-PCR products were fractionated on a Caliper microfluidic station. B, effect of anticancer agents on the relative abundance of the Bcl-x splice variants. Average difference (in percentage) in the abundance of the Bcl-xS isoform between drug- treated and the corresponding untreated (control) sample for each drug treatment in each cell line. These experiments were done in triplicates for all cell lines, except in SKOV-3, where they were done in duplicate. Based on the RT-PCR assays, the percentage of the Bcl-xS product relative to the sum of Bcl-xS and Bcl-xL products in various cell lines was 20.4 F 3.7% in EcR-293 cells, 11.3 F 2.8% in HeLa cells, 17.9 F 2.9% in PC-3 cells, 9.8 F 3.2% in PA-1 cells, 10.3 F 2.4% in MCF-7 cells, and 12.5% F 3.1% in SKOV-3 cells. We consider that a drug affected the production of Bcl-xS if the change corresponds to at least half the percentage value of Bcl-xS in untreated cells. For example, if the Bcl-xS product represented 10% of the total amount of Bcl- x products in untreated cells, a reduction of the Bcl-xS product to <5% or an increase to >15% was considered significant. Vertical dotted line, the threshold for each cell line. The average values above this threshold were always >2 SD from the average values of the corresponding untreated samples.

MolCancerTher2008;7(6).June2008

Figure 2. Splicing activity of anticancer agents in EcR-293 cells requires transcription but not de novo protein synthesis or caspase activation. A, effect of the transcription elongation inhibitor DRB. Drugs that shifted Bcl-x splicing in Fig. 1 were tested for activity in the presence of DRB. RT-PCR assays were carried after treating cells with the indicated concentrations of DRBfor 1 h before adding anticancer drugs. B, effect of the translation inhibitor cycloheximide, which was added 1 h before the addition of the anticancer drug. The concentrations of cycloheximide used are indicated. The intermediate concentration of 1.5 Ag/AL cycloheximide reduced by >50% the incorporation of [35S]methionine in bulk proteins (data not shown). C and D, caspase inhibition and Bcl-x splicing. Cells were pretreated for 1 h with the general inhibitor of caspases z-VAD-fmk (50 Amol/L) and then with chemotherapeutic agents. For the PARP cleavage assay in C, proteins were extracted after various treatments, fractionated on gels, and transferred to a nitrocellulose membrane. Western analysis was done with the anti-PARP antibody. Bcl-x splicing was assessed by RT-PCR and the results are shown in D. The relative abundance of the Bcl-xS isoform is plotted with the percentage of Bcl-xS product indicated below each histogram.

Mol Cancer Ther 2008;7(6).June 2008

seen for all controls were within 15%, a change was very good fit (84% overall concordance) between the two declared significant when the difference was equal or independent sets of experiments. superior to 15% (P < 0.05). The direction of the shift (e.g., The number of significant shifts for each drug in each cell exon inclusion or skipping) was not considered in this line is shown in Table 1. Overall, treatment of all cell lines analysis. Splicing units that did not shift once in any cell with 20 anticancer drugs shifted alternative splicing in 12% lines are not included in the figures. of the cases. Except for two cases (vinorelbine in PA-1and First, we compared the Bcl-x (BCL2L1) portion of docetaxel in SKOV-3), all the drugs shifted at least one this experiment with our previous manual assessment. alternative splicing event in all cell lines. Splicing decisions Twenty-seven of the 29 cases of drugs affecting Bcl-x in 293 cells were more responsive to drugs than in the splicing in different cell lines had been detected in our cancer-derived cell lines. first analysis (Fig. 1). Because more hits had been To analyze the effect of different drugs in individual cell identified manually, the automated procedure was lines, we did a hierarchical two-dimensional clustering therefore highly specific (0.91) but less sensitive (0.55). analysis (Fig. 3). The x axis of the clustered maps plots drug Of the 15 cases of splicing modulation that did not match response according to alternative splicing units in the between the two analyses, this number was reduced to y axis. This clustering analysis reveals a tendency for drugs 8 when we included differences of z12% rather than of the same group to be neighbors. For example, in z15%. Thus, based on the analysis of Bcl-x, there is a 293 cells, antimetabolites and mitotic inhibitors formed

Figure 3. Clustering for individual cell lines. Hierarchical clustering of units that displayed a significant difference in alternative splicing on drug treatment. X axis, all drugs; y axis, splicing units that were expressed in each cell line. For access to individual results, please link to http://lgfus.ca/annotate/luli/.

MolCancerTher2008;7(6).June2008

relatively distinct clusters, whereas alkylating agents (pla- for all cell lines are plotted relative to drugs (Fig. 4). Two tinum agents and DNA cross-linkers) clustered together. distinct clusters were observed: one made of the anti- In PA-1cells, platinum agents and several antimetabolites metabolites and mitotic inhibitors and the other one formed distinct clusters. Various topoisomerase inhibitors containing the alkylating agents and the topoisomerase and intercalating agents were also clustering closely in inhibitors. Thus, drug identity is making a significant MCF-7, PC-3, and SKOV-3 cells. Thus, drugs with common contribution toward determining alternative splicing hits. modes of action had a tendency to affect the same Although drugs of the same class were often neighbors, alternative splicing events in individual cell lines. This the hierarchical relationship between distinct clusters characteristic was also detected globally when target units appeared to vary considerably between cell lines (Fig. 3).

Figure 4. Hierarchical clustering of units that displayed a significant difference in alternative splicing on drug treatment. X axis, all 20 drugs; y axis, splicing unit/cell line combinations.

Mol Cancer Ther 2008;7(6).June 2008

Figure 5. Hierarchical clustering of units that displayed a significant difference in alternative splicing on drug treatment. X axis, all drug/cell line combinations; y axis, splicing units that were expressed in all cell lines.

To get a better appreciation of this cell line–specific The dendrogram clearly reveals that cell line identity was signature, we did a hierarchical clustering that helps see driving the clustering. Distinct clusters included a majority how individual splicing units reacted to individual drugs of the hits in 293, SKOV-3, PC-3, and MCF-7. For PA-1, hits in all cell lines (Fig. 5). In this case, we only considered were dispersed in three major subclusters, with one cluster splicing events that could be monitored in all five cell lines positioned near the SKOV-3 hits, possibly reflecting their (38 alternative splicing units) to ignore possible cell line– common ovarian origin. Overall, the coherence by cell line specific expression biases that might drive the clustering. was good as indicated by a j statistic of 0.65 (two-tailed

MolCancerTher2008;7(6).June2008

95% confidence interval, 0.53-0.76). Thus, despite a clear protein synthesis with cycloheximide did not block the relationship between drug classes and splicing targets, cell drug-induced shift in Bcl-x splicing. Our results are line identity is also making an important contribution to consistent with the notion that the drugs might alter determine which splicing units react to drugs. signaling pathways that impinge on Bcl-x splicing control. Our global analysis can also be examined from the Recently, we have observed that protein kinase C is part of ¶ perspective of alternative splicing events because distinct a network that represses the use of the 5 splice site of Bcl-xS units may share a common mode of regulation if their in 293 cells (30). Although protein kinase C inhibitors like splicing is affected by the same drugs. For example, the staurosporine, Go¨6976, and tamoxifen improve the pro- TIA1 TIAF1 Bcl- and splicing units belong to two related duction of Bcl-xS in 293 cells, these drugs did not affect genes and both reacted to cisplatin and topotecan in 293 x splicing in the cancer cell lines HeLa, SKOV-3, U-87, cells (Fig. 5). Likewise, TRAF2 and TRAF4 reacted to the MCF-7, and HCT 116 (30). It was therefore of interest to test same three drugs in SKOV-3 cells (Fig. 5). The case for a wider collection of anticancer drugs to see if Bcl-x splicing common regulatory pathways is also supported when remained resistant in cancer cell lines. Notably, for each distinct units respond to the same drugs in different cell cancer cell line tested (MCF-7, HeLa, PC-3, PA-1, and lines. For example, the response of alternative splicing SKOV-3), we identified drugs that increase the production BCL2L11 BCL2L12 units in genes of the Bcl-2 family ( and ; of Bcl-xS. However, the number of drugs active in cancer BCLAF1 and BAX) and tumor necrosis factor receptors cell lines was smaller than those active in 293 cells and the (TNFRSF6 and TNFRSF10B) was sufficiently similar in composition of the active set varied considerably between different cell lines to form distinct sets of adjacent or close cell lines. This cell line specificity was unlikely to be due to neighbors (Fig. 5), suggesting that members in each set deficiencies in drug uptake because almost every drug share common regulatory pathways. The other end of the affected the alternative splicing of other units in each cell spectrum is exemplified by BRCA1 and CASP9, which line. Hence, our results indicate that the regulatory features were affected by 18 and 19 drugs in SKOV-3 and 293 of Bcl-x splicing are largely cell line specific. This may cells, respectively (Fig. 5; Supplementary Table S1),3 reflect cell line–specific differences in the nuclear abun- suggesting that numerous pathways may converge to dance of splicing regulatory proteins. Alternatively or in regulate the alternative splicing of these units in SKOV-3 addition, cell line–specific differences in the signaling and 293 cells. pathways may alter the localization and activity of proteins that control Bcl-x splicing. Analyzing the effect of drugs on other apoptotic splicing Discussion units revealed that different drugs affected a common Alterations in apoptosis are associated with malignant subset of splicing events in each cell line. However, the growth and are a characteristic of nearly all types of cancers composition of these targets varied considerably between (2, 43). Likewise, aberrant profiles of alternative splicing cell lines. Thus, the networks of splicing regulation affected occur frequently in cancer, but the precise contributions of by the drugs appear largely cell line specific. Overall, this these alterations to malignancy and metastasis remain cell type specificity means that if a drug can shift splicing to poorly understood (reviewed in refs. 44, 45). Although the proapoptotic products in some cancer cell types, it may not expression of specific splice isoforms likely influences necessarily do so in others. cancer progression and the success of anticancer regimens, In contrast to the effect of drugs on Bcl-x splicing, which little is known about the effect of anticancer drugs on always shifted splicing toward the proapoptotic isoform, alternative splicing. To assess the general propensity of the effects of drugs on other alternative splicing units in anticancer drugs to modulate alternative splicing, we have apoptotic genes was not as clear cut. In 293 cells, most studied the effect of 20 anticancer agents on the alternative drugs activated the proapoptotic splicing event in Casp9 splicing of the important apoptotic regulator Bcl-x in 293 and survivin, but the reverse effect was seen on Fas and cells and five cancer cell lines. The analysis was further Siva. This profile did not consistently hold up in other cell expanded to cover a 95 alternative splicing events occur- lines. The ability of chemotherapeutic drugs to activate ring in apoptotic genes. signals that simultaneously induce and antagonize apopto- Our first conclusion is that a variety of anticancer agents sis has been noted previously (25, 46–48), and our analysis increased the relative abundance of Bcl-xS transcripts in 293 expand these observations to splice isoforms with opposing cells. Our experiments indicate that the drugs affected activities. splicing and not the differential stability of Bcl-x mRNA The effect of anticancer drugs on the alternative splicing variants because blocking transcription prevented the of Bcl-x and other apoptotic genes provides a novel drug-induced shifts. Although we have used concentra- rationale for using these compounds. The general ability tions of drugs that did not promote massive apoptosis, we of most drugs to favor the production of the proapoptotic nevertheless tested the possibility that this splicing shift Bcl-xS isoform may help promote apoptosis in cancer cells might be due to a caspase-dependent degradation of or may sensitize cancer cells to other cytotoxic effects of splicing factors. Clearly, the drug-induced Bcl-x splicing the drugs (22, 49). Yet, a drug that favors the production shift continued to be observed in the presence of a pan- of Bcl-xS may not always encourage the production of inhibitor of caspases. Likewise, compromising de novo proapoptotic variants in other genes. This may partly

Mol Cancer Ther 2008;7(6).June 2008

explain the variable success of these mainstream antican- 19. Lebedeva I, Rando R, Ojwang J, Cossum P, Stein CA. Bcl-xL in prostate cancer cells: effects of overexpression and down-regulation on cer agents. Based on the information gathered in this chemosensitivity. Cancer Res 2000;60:6052 – 60. study, it may be worth testing the cell-killing effect of 20. Cho HJ, Kim JK, Kim KD, et al. Upregulation of Bcl-2 is associated drug combinations that favor the production of proapop- with cisplatin-resistance via inhibition of Bax translocation in human totic isoforms from several genes. It will also be important bladder cancer cells. Cancer Lett 2006;237:56 – 66. 21. Gobe G, Rubin M, Williams G, Sawczuk I, Buttyan R. Apoptosis and to continue assessing the function of the many apoptotic expression of Bcl-2, Bcl-XL, and Bax in renal cell carcinomas. Cancer isoforms produced by alternative splicing and to deter- Invest 2002;20:324 – 32. mine how chemotherapeutic agents affect their production 22. Taylor JK, Zhang QQ, Wyatt JR, Dean NM. Induction of endogenous in different types of cancers. When taken in consideration Bcl-xS through the control of Bcl-x pre-mRNA splicing by antisense oligonucleotides. Nat Biotechnol 1999;17:1097 – 100. with other factors, such as p53 status, this information 23. Hayward RL, Macpherson JS, Cummings J, Monia BP, Smyth JF, may help improve the efficacies of future anticancer Jodrell DI. Antisense Bcl-xl down-regulation switches the response to therapies. topoisomerase I inhibition from senescence to apoptosis in colorectal cancer cells, enhancing global cytotoxicity. Clin Cancer Res 2003;9: 2856 – 65. Disclosure ofPotential Conflicts ofInterest 24. Fujimura M, Kasahara K, Shirasaki H, et al. Up-regulation of ICH-1L protein by thromboxane A2 antagonists enhances cisplatin-induced No potential conflicts of interest were disclosed. apoptosis in non-small-cell lung-cancer cell lines. J Cancer Res Clin Oncol 1999;125:389 – 94. Acknowledgments 25. Solier S, Lansiaux A, Logette E, et al. Topoisomerase I and II inhibitors control caspase-2 pre-messenger RNA splicing in human cells. Mol Cancer We thank Claudine Rancourt and Janet Price for the cell lines. Res 2004;2:53 – 61. 26. Wotawa A, Solier S, Logette E, Solary E, Corcos L. Differential References influence of etoposide on two caspase-2 mRNA isoforms in leukemic cells. Cancer Lett 2002;185:181 – 9. 1. Adams JM. Ways of dying: multiple pathways to apoptosis. Genes Dev 27. Pilch B, Allemand E, Facompre M, et al. Specific inhibition of serine- 2003;17:2481 – 95. and arginine-rich splicing factors phosphorylation, spliceosome assem- 2. Nelson DA, Tan TT, Rabson AB, Anderson D, Degenhardt K, White E. bly, and splicing by the antitumor drug NB-506. Cancer Res 2001;61: Hypoxia and defective apoptosis drive genomic instability and tumorigen- 6876 – 84. esis. Genes Dev 2004;18:2095 – 107. 28. Rohrbach S, Muller-Werdan U, Werdan K, Koch S, Gellerich NF, Holtz 3. Gelinas C, White E. BH3-only proteins in control: specificity regulates J. Apoptosis-modulating interaction of the neuregulin/erbBpathway with MCL-1 and BAK-mediated apoptosis. Genes Dev 2005;19:1263 – 8. anthracyclines in regulating Bcl-xS and Bcl-xL in cardiomyocytes. J Mol 4. Hengartner MO. The biochemistry of apoptosis. Nature 2000;407: Cell Cardiol 2005;38:485 – 93. 770 – 6. 29. Chalfant CE, Rathman K, Pinkerman RL, et al. De novo ceramide 5. Schmitt CA, Lowe SW. Apoptosis is critical for drug response in vivo. regulates the alternative splicing of caspase 9 and Bcl-x in A549 lung Drug Resist Updat 2001;4:132 – 4. adenocarcinoma cells. Dependence on protein phosphatase-1. J Biol Chem 2002;277:12587 – 95. 6. Lowe SW, Lin AW. Apoptosis in cancer. Carcinogenesis 2000;21: 485 – 95. 30. Revil T, Toutant J, Shkreta L, Garneau D, Cloutier P, Chabot B. Protein kinase C-dependent control of Bcl-x alternative splicing. Mol Cell Biol 7. Herr I, Debatin KM. Cellular stress response and apoptosis in cancer 2007;27:8431 – 41. therapy. Blood 2001;98:2603 – 14. 31. Yu D, Wolf JK, Scanlon M, Price JE, Hung MC. Enhanced c-erbB-2/ 8. Brown JM, Attardi LD. The role of apoptosis in cancer development and neu expression in human ovarian cancer cells correlates with more treatment response. Nat Rev Cancer 2005;5:231 – 7. severe malignancy that can be suppressed by E1A. Cancer Res 1993; 9. Reed JC. Drug insight: cancer therapy strategies based on restoration 53:891 – 8. of endogenous cell death mechanisms. Nat Clin Pract Oncol 2006;3: 32. Thierry-Mieg D, Thierry-Mieg J. AceView: a comprehensive cDNA- 388 – 98. supported gene and transcripts annotation. Genome Biol 2006;7 Suppl 1: 10. Kaufmann SH, Vaux DL. Alterations in the apoptotic machinery and S12.1 – 4. their potential role in anticancer drug resistance. Oncogene 2003;22: 33. Rozen S, Skaletsky H. Primer3 on the WWW for general users and for 7414 – 30. biologist programmers. Methods Mol Biol 2000;132:365 – 86. 11. White E. Mechanisms of apoptosis regulation by viral oncogenes in 34. Ihaka R, Gentleman R. A language for data analysis and graphics. infection and tumorigenesis. Cell Death Differ 2006;13:1371 – 7. J Comput Graph Stat 1996;5:299 – 314. 12. Klein S, McCormick F, Levitzki A. Killing time for cancer cells. Nat Rev 35. Thiede B, Dimmler C, Siejak F, Rudel T. Predominant identification of Cancer 2005;5:573 – 80. RNA-binding proteins in Fas-induced apoptosis by proteome analysis. 13. Waxman DJ, Schwartz PS. Harnessing apoptosis for improved J Biol Chem 2001;276:26044 – 50. anticancer gene therapy. Cancer Res 2003;63:8563 – 72. 36. Waterhouse N, Kumar S, Song Q, et al. Heteronuclear ribonucleopro- 14. Johnstone RW, Ruefli AA, Lowe SW. Apoptosis: a link between teins C1 and C2, components of the spliceosome, are specific targets of cancer genetics and chemotherapy. Cell 2002;108:153 – 64. interleukin 1h-converting enzyme-like proteases in apoptosis. J Biol Chem 15. Schwerk C, Schulze-Osthoff K. Regulation of apoptosis by alternative 1996;271:29335 – 41. pre-mRNA splicing. Mol Cell 2005;19:1 – 13. 37. Back SH, Shin S, Jang SK. Polypyrimidine tract-binding proteins 16. Li C, Kato M, Shiue L, Shively JE, Ares M, Jr., Lin RJ. Cell type and are cleaved by caspase-3 during apoptosis. J Biol Chem 2002;277: culture condition-dependent alternative splicing in human breast cancer 27200 – 9. cells revealed by splicing-sensitive microarrays. Cancer Res 2006;66: 38. Fischer B, Coelho D, Dufour P, et al. Caspase 8-mediated cleavage of 1990 – 9. the pro-apoptotic BCL-2 family member BID in p53-dependent apoptosis. 17. Mercatante DR, Bortner CD, Cidlowski JA, Kole R. Modification of Biochem Biophys Res Commun 2003;306:516 – 22. alternative splicing of Bcl-x pre-mRNA in prostate and breast cancer 39. Paronetto MP, Achsel T, Massiello A, Chalfant CE, Sette C. The RNA- cells. analysis of apoptosis and cell death. J Biol Chem 2001;276: binding protein Sam68 modulates the alternative splicing of Bcl-x. J Cell 16411 – 7. Biol 2007;176:929 – 39. 18. Hayes GM, Carrigan PE, Beck AM, Miller LJ. Targeting the RNA 40. Gundimeda U, Chen ZH, Gopalakrishna R. Tamoxifen modulates splicing machinery as a novel treatment strategy for pancreatic carcinoma. protein kinase C via oxidative stress in estrogen receptor-negative breast Cancer Res 2006;66:3819 – 27. cancer cells. J Biol Chem 1996;271:13504 – 14.

MolCancerTher2008;7(6).June2008

41. Ruegg UT, Burgess GM. Staurosporine, K-252 and UCN-01: potent Garrido C. Positive and negative regulation of apoptotic pathways by but nonspecific inhibitors of protein kinases. Trends Pharmacol Sci 1989; cytotoxic agents in hematological malignancies. Leukemia 2000;14: 10:218 – 20. 1833 – 49. 42. Klinck R, Bramard A, Inkel L, et al. Multiple alternative splicing 47. Bottero V, Busuttil V, Loubat A, et al. Activation of nuclear factor nB markers for ovarian cancer. Cancer Res 2008;68:657 – 63. through the IKK complex by the topoisomerase poisons SN38 and 43. Gerl R, Vaux DL. Apoptosis in the development and treatment of doxorubicin: a brake to apoptosis in HeLa human carcinoma cells. Cancer cancer. Carcinogenesis 2005;26:263 – 70. Res 2001;61:7785 – 91. 44. Pajares MJ, Ezponda T, Catena R, Calvo A, Pio R, Montuenga LM. 48. Guichard S, Arnould S, Hennebelle I, Bugat R, Canal P. Combination Alternative splicing: an emerging topic in molecular and clinical oncology. of oxaliplatin and irinotecan on human colon cancer cell lines: activity Lancet Oncol 2007;8:349 – 57. in vitro and in vivo. Anticancer Drugs 2001;12:741 – 51. 45. Skotheim RI, Nees M. Alternative splicing in cancer: noise, functional, 49. Mercatante DR, Mohler JL, Kole R. Cellular response to an antisense- or systematic? Int J Biochem Cell Biol 2007;39:1432 – 49. mediated shift of Bcl-x pre-mRNA splicing and antineoplastic agents. J Biol 46. Solary E, Droin N, Bettaieb A, Corcos L, Dimanche-Boitrel MT, Chem 2002;277:49374 – 82.

Mol Cancer Ther 2008;7(6).June 2008

Lulzim Shkreta, Ulrike Froehlich, Éric R. Paquet, et al.

Mol Cancer Ther 2008;7:1398-1409.

Updated version Access the most recent version of this article at: http://mct.aacrjournals.org/content/7/6/1398

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2008/07/18/7.6.1398.DC1

Cited articles This article cites 48 articles, 23 of which you can access for free at: http://mct.aacrjournals.org/content/7/6/1398.full#ref-list-1

Citing articles This article has been cited by 6 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/7/6/1398.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 Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/7/6/1398. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.