A Small Interference RNA Screen Revealed Proteasome Inhibition as Strategy for Glioblastoma Therapy

Please download to get full document.

View again

of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
A Small Interference RNA Screen Revealed Proteasome Inhibition as Strategy for Glioblastoma Therapy
  C HAPTER   19  A Small Interference RNA Screen Revealed ProteasomeInhibition as Strategy for Glioblastoma Therapy  Kimberly Ng, B.S., Masayuki Nitta, M.D., Ph.D., Lauren Hu, M.S., Santosh Kesari, M.D., Ph.D., Andrew Kung, M.D., Ph.D., Alan D’Andrea, M.D., and Clark C. Chen, M.D., Ph.D. 1,3 G lioblastoma (GBM) is the most common cancer of thecentral nervous system (CNS). 30,41 Despite advances inour understanding of its molecular pathogenesis, GBM re-mains a devastating disease. The current standard of care for GBM consists of surgical resection followed by combined radiation therapy and temozolomide (TMZ) chemotherapy. 52 Unfortunately, this regimen provides only palliative relief  because nearly all patients die of the disease within 3 years of diagnosis. 33,41,52 One strategy for developing new GBM therapeutics isto build on the modest efficacy of ionizing radiation (IR) and TMZ. Traditionally, this is achieved by screening a largenumber of chemicals to identify subsets that enhance thetumor-killing effect of IR and TMZ. 23 Typically, these sub-sets of chemicals exhibit limited biological activity and re-quire structural refinement before acceptable efficacy can beachieved. Because of the potentially infinite number of de-rivatives that can be made from a single candidate chemical,this refinement process is laborious and time-consuming. Toexpedite this refinement process, molecular targets of thecandidate chemical are frequently sought. Unfortunately, be-cause pharmacological agents inevitably affect multiple mo-lecular targets, identification of these targets often requiresyears of investigation.Recent development of small interference RNA (siRNA)technology now allows identification of molecular targets criti-cal for mediating specific biological processes. 18 This capabilityrevolutionizes drug development by allowing for molecular target discovery followed by the identification and synthesis of compounds that selectively inactivate the target. siRNA is aclass of short double-stranded RNA molecules (20–30 nucleo-tides in length) that interferes with gene expression in a se-quence-specific manner. On transfection into mammalian cells,siRNAs are processed by an enzyme called DICER and assem- bled into a complex known as RISC (RNA-inducing silencingcomplex). This complex uses the siRNA sequences as a meansto identify and degrade complementary mRNA sequences, 35 thereby preventing protein synthesis and subsequent phenotypicexpression. With the complete sequencing of the human ge-nome, 11 it is now possible to design siRNA specific to each geneto determine its effect on any biological process of interest.Because both IR and TMZ exert their tumoricidal effect by the induction of DNA damage beyond cellular capacity for repair, 17,49 we hypothesized that silencing selected DNArepair genes may augment tumor killing by these agents. Wetherefore screened a commercially available siRNA librarytargeting the 356 genes previously implicated in DNA repair.We found that genetic silencing of PSMA1 (prosome, mac-ropain subunit, alpha type, 1), a critical component of the proteasome, 16,31,40,61 strongly sensitized the U87MG cell lineto both TMZ and IR.The proteasome consists of a multisubunit protein com- plex that degrades proteins by coordinated activities of chy-motryptic, tryptic, and postglutamyl peptidases. 40,42 The pro-teasome complex is the major mechanism of intracellular  protein degradation. 10 By processing critical proteins in-volved in DNA repair, cell proliferation, and apoptosis, 40 the proteasome complex indirectly regulates these processes.Proteasome inhibitors have been shown to cause cellgrowth arrest and apoptosis in several glioblastoma celllines. 39,53,60 When treated with lethal concentrations of pro-teasome inhibitor, glioblastoma cells first develop G2/Marrest before undergoing apoptosis. The G2/M arrest islargely attributable to an accumulation of cyclin B1, p27, and  p21, the degradation of which is required for cell-cycle progression. 2,38,58 The signaling mechanism by which G2/Marrest triggered apoptosis remains unclear.Here we showed that independent proteasome inhibitorssensitized GBM cell lines to IR and TMZ at sublethal concen-trations. This sensitizing effect was independent of the cellular  p53 status. Importantly, the sensitization was observed in neu-rosphere cell lines with stemlike properties, suggesting that thecombined activity is effective against cancer stem cells. Theorderofadditioniscriticalbecausesensitizationisachievedonlywhen proteasome inhibition is delayed after TMZ or IR treat-ment. The mechanism of sensitization is, in part, mediated bymodulation of DNA damage response. METHODSCells and Cell Culture Methods The U87MG, U343MG, and U373MG cell lines wereobtained from American Type Culture Collection (Manassas, Copyright © 2009 by The Congress of Neurological Surgeons0148-703/09/5601-0107 Clinical Neurosurgery   • Volume 56, 2009  107  VA). The U87MG EGFRvIII, 36 U87MG DN p53, 1 BT69, 62 BT74, 62 BT75, 62 and BT78 62 lines were generated as previ-ously described. U87MG H2B-GFP was kindly provided byDr. Yoshinaga Saeki (Ohio State University). Non-neuro-sphere cell lines were propagated at 37°C (humidified atmo-sphere containing 5% CO 2 ) in Dulbecco’s modified Eaglemedium (Gibco, Rockville, MD) supplemented with 15%fetal calf serum (Sigma Aldrich, St. Louis, MO), 2 mM L -glutamine, 100 U/mL penicillin G sodium, and 100   g/mLstreptomycin sulfate (Gibco). Neurosphere cell lines were seri-ally propagated by subcutaneous implantation in NOD-SCIDmice and harvested only for TMZ and IR sensitivity assays. For these assays, the tumors were harvested when they reached approximately 1 cm in size. Tumors cells were disaggregated,counted, and then grown in a serum-free media with epidermalgrowth factor, fibroblast growth factor, and leukemia inhibitoryfactor as previously described. 29,50 TMZ (AK Scientific, Mountain View, CA) was dis-solved in deionized water immediately before use. Velcade(bortezomib) (AK Scientific, Inc., Mountain View, CA) and lactacystin (EMD Chemicals, Gibbstown, NJ) were dissolved in deionized water. ALLN (EMD Chemicals) was dissolved in DMSO (Sigma Aldrich). To test the sensitizing effects of the proteasome inhibitors, each inhibitor was titrated in termsof cytotoxic effects. The concentration of Velcade (5 nM),ALLN (1   M), and lactacystin (10 nM) were selected toensure 80% to 90% cell survival. The concentration of TMZ(100   M) was selected to achieve 30% to 50% cell kill.For the adherent cells (U87MG, U87MG DN p53,U373MG, U343MG, and A172), experiments involving Vel-cade, ALLN, or lactacystin were done in the following manner.Cells were treated with the indicated concentration for 24 hours.The cells were then washed twice with 1x phosphate-buffered saline (PBS) and then incubated in fresh medium. For TMZ,cells were treated for three hours. The cells were then washed twice with 1x PBS and then incubated in fresh medium.For Velcade pretreatment experiments, cells weretreated with 5 nM Velcade for 24 hours. The Velcade-containing medium was then removed, and the cells werewashed and treated with medium containing 100   M TMZfor 3 hours. The TMZ-containing medium was then replaced with fresh medium. For concurrent TMZ/Velcade treatment,cells were treated with medium containing 100  M TMZ and 5 nM Velcade for three hours. This medium was then re-moved. The cells were washed and incubated for an addi-tional 21 hours with medium containing 5 nM Velcade. Thismedium was then replaced with fresh medium. For thedelayed Velcade treatment, cells were incubated in 100   MTMZ-containing medium for three hours. This medium wasreplaced with fresh medium, and the cells were incubated for an additional 21 hours. After this incubation period, Velcadewas added to a final concentration of 5 nM. The cells wereincubated in this medium for another 24 hours. The Velcade-containing medium was then replaced with fresh medium.Because the neurospheres were nonadherent, serialwashes and medium changes were not practical. As such, the protocol for the neurosphere experiments was modified to thefollowing. The neurospheres were treated with various con-centrations of TMZ for 24 hours. Velcade was then added toa final concentration of 1 nM. The cells were incubated in thismixture for seven days before viability determination by theMTT Viability Assay kit (Biotium, Hayward, CA). siRNA Screening Transfection of cell lines in a 96-well format with theQIAGEN DNA repair library was performed as previouslydescribed. 26 For the radiation sensitization screen, the cellswere irradiated with 5 Gy of IR 24 hours after transfectionusing the Gamma40 (cesium-137 source) irradiator (BestTheratronics Ltd, Ottawa, Canada). For the TMZ sensitiza-tion screen, the cells were treated with 50   M of TMZ 24hours after transfection. The IR and TMZ doses were empir-ically determined to achieve 50% to 70% cell kill in our assay. Viability was assessed five days after treatment usingthe CellTiter-Glo Luminescent Cell Viability Assay kit (Pro-mega, Madison, WI). At this time point, the cells wereapproximately 60% to 80% confluent in each well. Theexperiment was performed twice for each cell line to allowfor statistical analysis. Statistical analysis was performed as previously described. 26 In brief, the corrected viability for each siRNA oligonucleotide was calculated as a percentageof the mean viability of the 16 control wells on each plate.The corrected viability of the treated cells was then divided  by the corrected viability of the nontreated cells to calculatethe relative viability after irradiation for each respective genetarget. The mean viability of the treated cells relative to thenontreated cells (also referred to as the mean percentage of viability after treatment), along with the standard deviation,was calculated from four individually corrected viabilityvalues (averaging the two distinct siRNA oligonucleotidesdirected against the same target gene in each of the twoexperiments). The candidates were ranked based on the mean percentage of viability after treatment. The sequences of thesiRNAs are available on request. Viability Assays For the neurosphere assay, the MTT Viability Assay kit(Biotium) was used in accordance to the manufacturer’s instruc-tions. For clonogenic assays, the cells were seeded in 10-cmdishes and treated with the various concentrations of Velcade,TMZ, and IR in the orders as described previously. Colonieswere counted after crystal violet staining at 14 days. 7 Flow Cytometry U87MG cells (1  10 5 ) were plated onto 6-well platesand cultured overnight. The drug treatment was done as  Ng et al.  Clinical Neurosurgery   • Volume 56, 2009 © 2009 The Congress of Neurological Surgeons 108  indicated in  Figure 19.2B . The cells were harvested at eachtime point and fixed with 70% ethanol and stored at  20°C.After the collection of all the samples, the cells were washed with PBS twice, treated with 1 mg/mL of RNAseA/PBS for 15 minutes, and stained with propidium iodide (25   g/mL).Fluorescence-activated cell sorting was performed with FAC-SCalibur (BD Biosciences), and cell cycle distribution wasanalyzed with CellQuest software (BD Bioscience, San Jose,CA); 1  10 4 cells were analyzed for each sample. Live Cell Imaging U87MG stably expressed H2B-GFP cells were grownon 12-well glass-bottom dishes (MatTek, Ashland, MA)overnight (5  10 4 cells per well). Drug treatment was doneas described in  Figure 19.2B . Images were acquired automat-ically from each well using a Nikon TE2000E PFS inverted microscope fitted with a 20   Nikon Plan Fluor objective(Nikon, Melville, NY), a linearly encoded stage (Prior ProScan, Prior Scientific, Rockland, MA) and a HamamatsuOrca-ER CCD camera (Hamamatsu, Bridgewater, NJ). Themicroscope was controlled using NIS Element (Nikon). Themicroscope was housed in a custom-designed 37°C chamber with a secondary internal chamber that delivered humidified 5% CO 2 . Fluorescence and differential interference con-trast images were obtained every 15 minutes for a period of 72 hours.For analysis of abnormal mitosis, cells were analyzed if they entered mitosis at least 20 hours before the end of imaging. Mitotic chromosomes were analyzed for the pres-ence of chromosome segregation error. We also documented whether cells completed cytokinesis. The cells were scored ashaving chromosome segregation error if H2B-GFP–positivechromatin was found outside its mother nucleus during an-aphase/telophase. Cells were considered to be in prolonged mitotic arrest if they arrested for more than 24 hours. Mouse Xenograft Model and In Vivo Imaging The U87 cell line harboring a luciferase expressionconstruct was constructed as previously described. 43 Cellsderived from this cell line were harvested in mid-logarithmicgrowth phase and resuspended in PBS. Homozygous NCR nude mice (Charles River, Wilmington, MA) were anesthe-tized with ketamine hydrochloride at 150 mg/kg and xylazineat 12 mg/kg (Phoenix Pharmaceuticals, St. Joseph, MO)intraperitoneally before head fixation in a stereotactic frame(Stoelting, Wood Dale, IL). The cranium was exposed. Onehundred thousand U87-Luciferase cells were resuspended in10   L of PBS and injected through a 27-gauge needle for two minutes at 2 mm lateral and posterior to the bregma. Thedepth of the implant was 3 mm below the dura. The incisionwas closed with Vetbond (3M Co., St Paul, MN).Uptake of the implanted U87 cells were assessed 15 daysafter implant using in vivo imaging (see below). The mice withsuccessful uptake were divided into four groups, with four tofive mice in each group. The TMZ-only group was treated with15 mg/kg TMZ (orally) on days one, two, and three. TheVelcade-only group was treated with 1 mg/kg of Velcade (in-travenously) on day one. The TMZ    Velcade group wastreated with 15 mg/kg of TMZ (orally) on days one, two, threeand 1 mg/kg of Velcade (intravenously) on day five. Weekly invivo imaging was performed thereafter. The control group wasmock-treated with vehicle medium at the time points described above.For in vivo imaging, the mice were anesthetized asdescribed above and injected with D-luceferin at 50 mg/mLintraperitoneally (Xenogen, Alameda, CA) and imaged withthe IVIS Imaging System (Xenogen) for 10 to 120 seconds, bin size 2. To quantify bioluminescence, identical circular regions of interest were drawn to encircle the entire head of each animal, and the integrated flux of photons (photons per second) within each region of interest was determined usingthe LIVING IMAGES software package (Xenogen). Datawere normalized to bioluminescence at the initiation of treat-ment for each animal. The experiment was repeated twice. RESULTSPSMA1 Silencing Sensitized U87MG to BothTMZ and IR We selected the U87MG cell line for our screen be-cause of the ease of transfection and gene silencing in this cellline. Because TMZ and IR causes tumor killing by theinduction of DNA damage beyond cellular capacity for re- pair, we reasoned that inactivation of selected DNA repair genes may augment tumor killing by these agents. We used the QIAGEN DNA repair siRNA library that targets 356DNA repair and damage response genes (for detailed desc-ription of the library, please visit http://www1.qiagen.com/default.aspx). All critical genes involved in the major DNArepair pathways and damage response pathways known todate are included in this collection. 26 The library is gridded into a 96-well format (  Fig. 19.1A ) such that each wellcontains a single siRNA. Each gene target is represented bytwo distinct siRNAs. The screening process is outlined in  Figure 19.1A  and detailed in the Methods section.The 15 genes that, when silenced by independent siRNAs,consistently and efficiently sensitized U87MG to TMZ and IR.Many of the gene targets represented by these siRNAs have previously been shown to mediate radiation response(ATM, RAD51L3, ADPRT1, RPA 6,46 ) and TMZ response(BRCA1, FANCG, ADPRT2). 7,8,12 The finding that thesegenes appeared as top sensitizers served to validate our screen. We identified one common gene target that whensilenced consistently sensitized U87MG to both TMZ and IR.This gene, PSMA1 (prosome, macropain subunit, alpha type1), encodes a critical component of the proteasome com- Clinical Neurosurgery   • Volume 56, 2009  Glioblastoma, Proteasome Inhibition © 2009 The Congress of Neurological Surgeons  109   plex. 16,31,40,61 This result suggests that proteasome inhibitionmay be a strategy for simultaneous radiation and TMZsensitization. Proteasome Inhibition Sensitized U87MG toTMZ and IR As a first step to translate the results of our siRNAscreen into a clinical strategy, we tested whether proteasomeinhibitors sensitized the U87MG cell line to TMZ. As shownin  Figure 19.2A , Velcade exhibits significant tumoricidalactivity in the absence of TMZ or IR treatment at a 10-nMconcentration. We therefore selected 5 nM as a sublethal dosefor testing TMZ sensitization.TMZ is an alkylating agent that methylates DNA at avariety of positions. 13,14,17 When unrepaired, these alkylated  products cause the formation of DNA strand breaks thatultimately result in cell death or cell senescence. 21,22,51 The process by which alkylated DNA products are converted intoDNA strand breaks requires progression through one entirecell cycle. 51 Because Velcade has been shown to induceG2/M arrest, 54 we reasoned that the order of addition betweenVelcade and TMZ may be critical and tested the variousorders of addition.As shown in  Figure 19.2C  , treatment of U87MG with5 nM Velcade for 24 hours resulted in 5% to 10% loss inclonogenic survival. Treatment with TMZ for 3 hours caused an approximately 50% reduction in clonogenic survival. Clo-nogenic survival was not significantly altered by pre- or simultaneous treatment with Velcade. However, delayed Vel-cade treatment 24 hours after TMZ treatment significantlyaugmented the tumoricidal activity of TMZ, with clonogenicsurvival of 5% to 10%. When fit into the Chou-Talalaymutually nonexclusive model, 9 the combination index of thisregimen was 0.3, suggesting that the interaction betweendelayed Velcade treatment and TMZ was synergistic.Similar results were observed with IR. Pretreatment or simultaneous treatment with Velcade did not significantlyaugment the tumoricidal activity of IR. Delayed Velcadetreatment, conversely, augmented IR-induced tumor killing ina synergistic manner.We wished to further confirm that the effect of Velcadewas entirely due to proteasome inhibition rather than nonspe-cific activity related to the compound. We therefore tested two other proteasome inhibitors of chemical structure thatdiffer from Velcade (ALLN 47 and lactacystin 24 ). As shown in  Figure 19.2D , the TMZ sensitizing effect of Velcade wasrecapitulated by ALLN and lactacystin. TMZ Sensitization by Proteasome Inhibition IsIndependent of the Cellular p53 Status The U87MG line harbors a wild-type p53 gene with anintact p53 axis. 25 Because p53 was previously shown to be an FIGURE 19.1.  Small interference RNA (siRNA)screen to identify gene silencings that sensitized theU87MG line to ionizing radiation (IR) and temozo-lomide (TMZ).  A,  Schematic depiction of thescreen. Cells were seeded into 96-well plates on dayone. On day two, each well was transfected with asiRNA oligonucleotide directed against one DNAdamage response/repair gene. On day three, oneset of cells remained untreated, and another set wastreated with TMZ at a concentration of 50   M. Athird set was irradiated with 5 Gy of IR. On day six,the viability of the cells in each well was measuredusing the Cell Titer-Glo Luminescent Cell Viability Assay kit. ATP, adenosine triphosphate.  B,  The top15 gene targets from the genetic screen areshown. The percentage of viability and standarddeviations were calculated as described in Meth-ods. PSMA1 is indicated in red because silencingof this gene by independent siRNAs sensitized theU87MG line to both TMZ and IR.  Ng et al.  Clinical Neurosurgery   • Volume 56, 2009 © 2009 The Congress of Neurological Surgeons 110  0204060801001200.1 nM1 nM10 nM100 nM 020406080100120   u  n  t  r e  a  t e  d V   T   M  Z   V  /   T   M  Z   V  +   T   M  Z   T   M  Z  /   V I   R    V  /   I   R    V  +   I   R    I   R   /   V Velcade% survival TMZ (3 hr)TMZ (3 hr)Velcade(24 hr)Velcade(24 hr)No treatmentVelcadeTMZTMZ/VelcadeTMZ (3 hr)Velcade(24 hr)Velcade/TMZTMZ+VelcadeVelcadeVelcade/TMZTMZ+VelcadeTMZ (3 hr) + Velcade (24 hr) % survival 020406080100120   u  n  t  r e  a  t e  d   T   M  Z V   V  /   T   M  Z   V  +   T   M  Z   T   M  Z  /   V  A   L   L   N  A   L   L   N  /   T   M  Z  A   L   L   N  +   T   M  Z   T   M  Z -  >  A   L   L   N   L  a c   L  a c  /   T   M  Z   L  a c  +   T   M  Z   T   M  Z -  >   L  a c A BCD FIGURE 19.2.  DelayedVelcade(bortezomib)treatmentsensitizedthetumoricidaleffectoftemozolomide(TMZ)andionizingradiation(IR). A,  Dose-response curves demonstrating the response of U87MG cells to Velcade. Cells were incubated with the various concentrations of  Velcade for 24 hours. The Velcade-containing medium was then removed. The cells were washed twice with 1x phosphate-buffered salinebefore fresh medium was added. The cells were then incubated in this medium for 14 days before colonies were counted. The percentageof survival after treatment was calculated by dividing viability of treated cells relative to nontreated cells. The results represent the average of triplicates.Theexperimentwasrepeatedthreetimes.Arepresentativeexperimentisshown. B, RegimenforcombiningVelcadeandTMZor IR.ControlexperimentswereperformedbytreatingU87MGcellswithvehicle(dimethylsulfoxide),100  MTMZfor3hoursor5nMVelcade for24hours.ForVelcadepretreatmentexperiments,cellsweretreatedwith5nMVelcadefor24hours.TheVelcade-containingmediumwasthenremoved,andthecellswashedandtreatedwithmediumcontaining100  MofTMZforthreehours.TheTMZ-containingmediumwasthenreplacedwithfreshmedium.ForconcurrentTMZ/Velcadetreatment,cellsweretreatedwithmediacontaining100  MTMZand5nM Velcadefor3hours.Thismediawasthenremoved.Thecellswerewashedandincubatedforanadditional21hourswithmediumcontaining5 nM Velcade. This medium was then replaced with fresh medium. For the delayed Velcade treatment, cells were incubated in 100   MTMZ-containingmediumfor3hours.Thismediumwasreplacedwithfreshmedium.Thecellswereincubatedforanadditional21hoursinthismedia.Afterthisincubationperiod,Velcadewasaddedtoafinalconcentrationof5nM.Thecellswereincubatedinthismediaforanother 24 hours. The Velcade-containing medium was then replaced with fresh medium. For the IR experiments, the pretreatment, concurrenttreatment,anddelayedtreatmentwithVelcadewereperformedinananalogousmannerexceptthatTMZwasreplacedwithIR. C, Delayed Velcade treatment sensitized U87MG cells to TMZ and IR. Clonogenic survival assay was performed after treatment described in B. Thepercentageofsurvivalaftertreatmentwascalculatedbydividingtheviabilityoftreatedcellsrelativetonontreatedcells.Theresultsrepresentthe average of triplicates. The experiment was repeated three times. A representative experiment is shown. V, Velcade; V/TMZ, Velcadepretreatment before TMZ; V   TMZ, simultaneous Velcade treatment with TMZ; TMZ/V, delayed Velcade treatment after TMZ.  D,  Indepen-dent proteasome inhibitors sensitized U87MG to the tumoricidal effect of TMZ. To confirm that the TMZ-sensitizing effect of Velcade wasdirectly related to proteasome inhibition, we tested the effect of two other proteasome inhibitors (ALLN 47 and lactacystin (Lac) 24 ). ThetreatmentregimenswereasdescribedforVelcade( B )exceptthatALLN(1  M)andlactacystin(10nM)replacedVelcadeintheseregimens.The percentage of survival after treatment was calculated by dividing the viability of treated cells relative to nontreated cells. The resultsrepresent the average of triplicates. The experiment was repeated three times. A representative experiment is shown. Clinical Neurosurgery   • Volume 56, 2009  Glioblastoma, Proteasome Inhibition © 2009 The Congress of Neurological Surgeons  111
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks