ﬁling bortezomib resistance identiﬁes secondary therapies ... · bzr cells were generated by...

Tools and Technologies

Profiling Bortezomib Resistance Identifies SecondaryTherapies in a Mouse Myeloma Model

Holly A.F. Stessman1, Linda B. Baughn1, Aaron Sarver4, Tian Xia2, Raamesh Deshpande2,Aatif Mansoor1, Susan A. Walsh5, John J. Sunderland5, Nathan G. Dolloff8, Michael A. Linden3,Fenghuang Zhan6, Siegfried Janz7, Chad L. Myers2, and Brian G. Van Ness1

AbstractMultiple myeloma is a hematologic malignancy characterized by the proliferation of neoplastic plasma

cells in the bone marrow. Although the first-to-market proteasome inhibitor bortezomib (Velcade) has

been successfully used to treat patients with myeloma, drug resistance remains an emerging problem. In

this study, we identify signatures of bortezomib sensitivity and resistance by gene expression profiling

(GEP) using pairs of bortezomib-sensitive (BzS) and bortezomib-resistant (BzR) cell lines created from the

Bcl-XL/Myc double-transgenic mouse model of multiple myeloma. Notably, these BzR cell lines show

cross-resistance to the next-generation proteasome inhibitors, MLN2238 and carfilzomib (Kyprolis) but

not to other antimyeloma drugs. We further characterized the response to bortezomib using the Connec-

tivity Map database, revealing a differential response between these cell lines to histone deacetylase

(HDAC) inhibitors. Furthermore, in vivo experiments using the HDAC inhibitor panobinostat confirmed

that the predicted responder showed increased sensitivity to HDAC inhibitors in the BzR line. These

findings show that GEP may be used to document bortezomib resistance in myeloma cells and predict

individual sensitivity to other drug classes. Finally, these data reveal complex heterogeneity within

multiple myeloma and suggest that resistance to one drug class reprograms resistant clones for increased

sensitivity to a distinct class of drugs. This study represents an important next step in translating

pharmacogenomic profiling and may be useful for understanding personalized pharmacotherapy for

patients with multiple myeloma. Mol Cancer Ther; 12(6); 1140–50. �2013 AACR.

IntroductionMultiple myeloma is a hematopoietic neoplasm charac-

terized by theproliferationofmalignant plasmacells in thebonemarrow (1). Each year about 22,000 new cases arise inthe United States, accounting for approximately 2% of allcancer deaths (2). Standard treatments for patients withmultiple myeloma use combination chemotherapies (i.e.,alkylating agents and corticosteroids) along with autolo-

gous stem cell transplants. However, in the past decade, anumber of novel classes of agents have beendeveloped forthe treatment of multiple myeloma, including the protea-some inhibitor, bortezomib (Velcade; Millennium Phar-maceuticals, Inc.), which is approved for the treatment ofmultiplemyelomaand relapsedmantle cell lymphoma (3).Despite the initial success of bortezomib therapy, multiplemyeloma remains incurable due in part to the emergenceof bortezomib-resistant (BzR) cells in the majority ofpatients (4, 5).

The primary target of bortezomib, the proteasome, ispart of the highly regulated ubiquitin–proteasome system(UPS) necessary for intracellular proteolysis. The UPSplays a critical role in cellular homeostasis, cell-cycleprogression, and DNA repair (6, 7). The constitutiveproteasome, a primary UPS player, is composed of thecatalytic 20S core barrel and 19S regulatory caps (togethercalled the 26S proteasome). Bortezomib is a boronic aciddipeptide that is highly selective for inhibition of thechymotryptic activity of the 26S proteasome via reversiblebinding of its target, PSMB5, a subunit of the 20S catalyticcore (8, 9). Bortezomib treatment has been shown toinhibit the transcriptional activity of NF-kB as well astrigger the unfolded protein response (UPR), leadingto cell stress and apoptosis (10–12). With the advent ofnext-generation proteasome inhibitors, it has become

Authors' Affiliations: Departments of 1Genetics, Cell Biology and Devel-opment, 2Computer Science and Engineering, and 3Laboratory Medicineand Pathology; 4Masonic Cancer Center Bioinformatics Support andServices, University of Minnesota, Minneapolis, Minnesota; Departmentsof 5Radiology, 6Internal Medicine, and 7Pathology, University of Iowa, IowaCity, Iowa; and 8Department of Medicine, Penn State Hershey MedicalCenter, Hershey, Pennsylvania

Note: Supplementary data for this article are available at Molecular CancerTherapeutics Online (http://mct.aacrjournals.org/).

Current address for T. Xia: Department of Electronics and InformationEngineering, Huazhong University of Science and Technology, Wuhan,China.

CorrespondingAuthor:BrianG. VanNess, University ofMinnesota, 6-160Jackson Hall, 321 Church Street SE, Minneapolis, MN 55455. Phone: 612-624-9944; Fax: 612-625-4294; E-mail: [email protected]

doi: 10.1158/1535-7163.MCT-12-1151

�2013 American Association for Cancer Research.

MolecularCancer

Therapeutics

Mol Cancer Ther; 12(6) June 20131140

on March 7, 2020. © 2013 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst March 27, 2013; DOI: 10.1158/1535-7163.MCT-12-1151

http://mct.aacrjournals.org/

imperative that the bortezomib response and signaturesthat are associated with bortezomib-resistance can befurther defined to identify those patients who (i) willbenefitmost fromproteasome inhibitor treatment, (ii) willshow signs of emerging resistance, and (iii) will benefitfrom selective secondary therapies.Double-transgenic Bcl-XL/Myc mice develop plasma

cell tumors (mean onset of 135 days) with full (100%)penetrance that possess many of the karyotypic, pheno-typic, and gene expression features of human multiplemyeloma (13, 14). Furthermore, malignant plasma cellscan be isolated from these animals, expanded, modifiedin vitro, and subsequently transferred back into syngeneicmice for drug treatment in the presence of an activeimmune system,which includes a complete bonemarrowmicroenvironment (13, 14). We have chosen to use thissystem to model bortezomib response to better under-stand acquired bortezomib-resistance within malignantplasma cells for potential application to the human mul-tiple myeloma patient population. Although therapeuticresponse in humanmultiplemyeloma is likely dependenton a combination of tumor genetic variation and inherit-ed population variation, the mouse model provides acommon strain background, allowing us to focus ontumor variation and tumor evolution that lead to drug-resistance.In this study, we have used gene expression profiling

(GEP) in the mousemodel system to identify bortezomib-responsive genes in vitro. To identify signatures of borte-zomib-resistance, we created BzR mouse cell lines andexamined their response to bortezomib compared withsensitive controls by GEP, revealing signatures of borte-zomib-resistance over the course of bortezomib treatment.Finally, we used the Connectivity Map (CMAP) databaseto further identify individual differences in bortezomibresistance in our representative bortezomib-sensitive(BzS) and BzR lines, which identified a unique responseto histone deacetylase (HDAC) inhibitors, a class of drugscurrently in early clinical trials for the treatment of mul-tiple myeloma (15–21).

Materials and MethodsMouse and human plasma cell linesThemouse cell lines 595, 589, and 638were isolated from

3 individual Bcl-XL/Myc double transgenic mice andcultured in Cell Signaling Technologies (CST) media aspreviously described (14) and cultured for greater than 40passages before any experimentation, after which nochanges in rate of growth, clonal drift, and response todrug or gene expression were observed. The humanmye-loma cell linesMM1.S andU266 (obtained fromAmericanType Culture Collection) were maintained in humanmyeloma cell line (HMCL) media: RPMI-1640 (Lonza)supplemented with 15% FBS (Cellgro), 50 mmol/L b-mer-captoethanol (Sigma-Aldrich), 50 U/mL of penicillin andstreptomycin (Hyclone), and 2 mmol/L L-glutamine (LifeTechnologies). Cell lines were not authenticated.

Drugs and treatment conditionsBortezomib (Millennium Pharmaceuticals, Inc.) was

dissolved in serum-free RPMI-1640 (Lonza) and storedat �80�C. MLN2238 (Millennium Pharmaceuticals, Inc.),MG-132 (American Peptide Company), trichostatin A(Cell Signaling Technology), and vorinostat (SAHA; LCLaboratories) were dissolved in dimethyl sulfoxide(DMSO; Sigma-Aldrich), melphalan (Sigma-Aldrich) wasdissolved in EtOH, and panobinostat (LC Laboratories)was dissolved in double-distilled water (ddH2O); alldrugs were stored at �20�C.

BzR cells were generated by dose escalation of borte-zomib over 6 months by once weekly treatment withbortezomib starting at 16 nmol/L; the bortezomib con-centration was doubled every 3 weeks until a finalgrowth concentration of 64 nmol/L was reached. Cul-tures were removed from bortezomib for 14 days or 6months before analysis and cultured in a manner con-sistent with the parental lines.

Preparation of RNA for GEPThe mouse cell lines 595 BzS, 595 BzR, 638 BzS, and 638

BzR, removed frombortezomib selection for 14days,wereplated at a density of 4 � 105 cells per mL in CST mediawith mIL-6. In addition, 595 BzS (595.2), 595 BzR (BzR595.2), 589 BzS, and 589 BzR, removed from bortezomibselection for 6 months, were plated at a density of 4 � 105

cells per mL in CST media with mIL-6. After 24-hourincubation, 66 nmol/L bortezomib was added to eachwell, and the cellswere collected at 0, 2, 8, 16, and 24 hoursafter treatment. Similarly, MM1.S and U266 were plated24 hours before 33nmol/Lbortezomib treatment and cellswere collected at 0, 16, and 24 hours. Cells were lysed andtotal RNAwas extracted using QIAshredder and RNeasyRNA purification columns (Qiagen). RNA concentrationand integrity were analyzed using the Nanodrop-8000(Thermo Scientific) and 2100 Bioanalyzer (Agilent Tech-nologies). cDNA was prepared and labeled with theIllumina TotalPrep-96 RNA Amplification Kit (LifeTechnologies). Samples were hybridized to the IlluminaMouseWG-6 v2.0 Expression BeadChip according tomanufacturer’s protocols and read on the Illumina iScan(Illumina). Microarray data have been made available atthe Gene Expression Omnibus (GEO) website (accessionno. GSE41930).

Cell viability assayCells were seeded at a concentration of 4� 105 cells per

mL. After 24 hours, the cells were treated with the indi-cated concentrations of drug for 48 hours. Cell viabilitywasmeasured by CellTiter-Glo Luminescent cell viabilityassay according tomanufacturer’s instructions (Promega)using the Synergy 2 Microplate Reader (Biotek). Valueswere normalized to untreated controls and IC50 valueswere estimated by calculating the nonlinear regressionusing the sigmoidal dose–response equation (variableslope) in GraphPad (Prism). In some experiments, cell

Identifying Secondary Therapies in Refractory Myeloma

www.aacrjournals.org Mol Cancer Ther; 12(6) June 2013 1141




growthwas calculated byTrypanblue (Life Technologies)exclusion with a hemocytometer.

CMAP drug predictionGenes were assigned a binary distinction of "up" or

"down," respectively, by selecting genes with at least 2-fold higher or lower expression in the BzS cell linesrelative to the derived BzR lines at the 24-hour time point(post-bortezomib treatment). These "up" and "down" setswere queried against the CMAP database (22, 23). Thegiven connectivity score estimates the similarity betweenthe input and database expression signatures. A positivescore indicates that the input signature is similar to, and anegative score indicates that the input signature is oppo-site to what would be expected by the predicted com-pound. Predicted compounds were ranked in ascendingorder of P value and viable compounds were chosen onthe basis of significance of correlation with the inputsignature (P < 0.05).

Animal care, tumor injection, and drug treatmentFVBN/Bl6 recipientmicewere generated as previously

described (13, 14). Mice were maintained in a controlledenvironment receiving food and water ad libitum. Fresh,ficolled (Ficoll-Paque PLUS, GE Healthcare) 595 BzS andBzR cells (1 � 106) suspended in 100 mL of serum-freeRPMI-1640 media were injected into 6 mice per group(3 groups/cell line) via the lateral tail vein of age-matchedrecipient mice. Mice were weighed and treated by intra-peritoneal injection twice per week starting 5 days aftercell transfer with vehicle (5% dextrose in water), borte-zomib (1.2 mg/kg dissolved in 0.9% saline), or panobino-stat (10 mg/kg suspended in vehicle) until moribund.Kaplan–Meier curves were generated using GraphPad(Prism). Statistical significance was determined using aStudent t test. Tumor cell homing was monitored bypositron emission tomography (PET) imaging (see Sup-plementary Methods). All mouse veterinary care, colonymaintenance, and PET imaging experiments were car-ried out in accordance with University of Iowa Institu-tional Animal Care and Use Committee guidelines andapprovals.

ResultsBcl-XL/Myc transgenicmouseplasma cell tumor linesshow similar significant shifts in gene expressionupon bortezomib treatment as human myeloma

In this study, 3 representative clonal cell lines isolatedfrom the Bcl-XL/Myc double transgenic mouse model ofplasma cell malignancy were used (13, 14) to identifytranscriptional responses to bortezomib in BzS and BzRcells in vitro. The 595, 589, and 638 plasma cell linesderived from individual mice showed IC50 values withina 22 to 32 nmol/L range to bortezomib by cell viabilityassay following 48 hours of drug treatment (Supplemen-tary Fig. S1A). Consistent with previous reports (24),bortezomib treatment of these cell lines resulted in pro-

grammed cell death as evidenced by caspase-3 cleavageand annexin V-positive/propidium iodide-negativestaining by flow cytometry (data not shown). This cyto-toxic profile was similar to 2 representative, well-described BzSHMCLs,MM1.S andU266 (SupplementaryFig. S1B; refs. 25–27). Taken together, these data suggestthat, such as in humanmultiplemyeloma, the treatment ofthese mouse cell lines in vitro with bortezomib induces acytotoxic response.

We next asked whether the transcriptional profileinduced over time by exposure to bortezomibwas similaracross both species. The 3 BzS mouse lines (595 shownin duplicate as 595.2) were treated with a sublethal 66nmol/L dose of bortezomib over a time course of 24 hoursand analyzed using GEP. This dose was chosen because itresulted in less than 20% death at 24 hours (Supplemen-tary Fig. S1C) but greater than 50%death at 48 hours (datanot shown), suggesting that it was an optimal concentra-tion for collecting kinetic data within a 24-hour timeframe. In addition, the HMCLs, U266 and MM1.S, werealso treated with a sublethal 33 nmol/L dose (Supple-mentary Fig. S1D) of bortezomib (equitoxic to the doseused on themouse cell lines) and analyzed by time courseGEP. The variant transcripts identified from these human(genes ¼ 1,421; variance > 0.1) and mouse (genes ¼ 1,021;variance > 0.1) time course data shared 132 genes incommon, 58% of which also shared a common, kineticpattern of response to bortezomib across all cell linesanalyzed across both species (Fig. 1B and SupplementaryTable S1).

This transcriptional response was further validatedusing GEP data from a recently published humanMMTT3 drug trial by Shaughnessy and colleagueswhereRNA was collected from newly diagnosed patients withmultiple myeloma before and following a single, 48-hourtest dose of bortezomib (28). Gene set enrichment anal-ysis (GSEA) of themouse in vitro transcriptional responseto bortezomib (24 vs. 0 hours) showed remarkable enrich-ment [nominal P value (NOM), P < 0.05; false discoveryrate (FDR) < 25%] for the Shaughnessy and colleagues 80-gene model. This 80-gene model included those genesthat were (i) most changed among all patients by thebortezomib treatment at 48 hours and (ii) were associ-ated with progression-free survival (PFS), of which 46geneswere unique and shared a gene symbolwithmouse(Supplementary Fig. S2A; ref. 28). Twenty-nine geneswere shown to significantly contribute to the enrichmentbetween the mouse in vitro and the human in vivo data(Supplementary Fig. S2B) many of which are knowncomponents of the proteasome ubiquitination pathway(28), which was found to be enriched in this dataset byIngenuity PathwayAnalysis (IPA). In addition, the kinet-ic GEP response in both the human and mouse cell lineswas enriched for downstream targets of the transcriptionfactor NFE2L2 (NRF2; P < 1 � 10�10; z score > 4; IPA) atranscription factor known to be involved in theoxidative stress response to proteasome inhibition (29).Taken together, these data show a robust transcriptional

Stessman et al.

Mol Cancer Ther; 12(6) June 2013 Molecular Cancer Therapeutics1142




response to bortezomib that is conserved not onlybetween mouse and human plasma cell malignancies,but also within isolated tumor cells in vitro and tumorcells in their nativemicroenvironment in humanpatients.

Generation of clonally related bortezomib-sensitiveand -resistant cell line pairsBecause the mouse model provides a drug-na€�ve sys-

tem with a common strain background and the abilityto transfer cells back into syngeneic animals with a com-petent immune system, we chose to further profile thetranscriptional response associated with bortezomib-resistance within the tumor cell using this model system.The 3 BzS mouse cell lines characterized earlier, weredose-escalated with bortezomib over 6 months to obtainresistant (BzR), clonally related (data not shown), daugh-ter cell lines with a 2- to 5-fold increase in IC50 in responseto bortezomib (Fig. 2A andTable 1). Doubling rates for thepairs of cell lines were similar (data not shown), suggest-ing that the increased IC50 in BzR lines is not due to anincrease in cell growth. To determine whether these BzRmouse cell lines were capable of maintaining their borte-zomib-resistance over time, the cells were removed fromdrug selection, and bortezomib cell viability assays wereconducted after 1 year. The BzR lines maintained theirresistant phenotype showing virtually identical IC50

values (data not shown).

Others have reported that bortezomib-resistance isoften associated with increases in chymotrypsin-like pro-teasome activity (30–32). Indeed, we find that 2 represen-tative BzR cell lines have significantly increased chymo-trypsin-like activity, whereas trypsin- and caspase-likeactivity is decreased (Supplementary Fig. S3A and S3B)compared with their BzS counterparts. This observationdirectly correlated with a similar increase in PSMB5 pro-tein expression by Western blotting in the BzR cells (datanot shown) in the absence ofPsmb5mutations by sequenc-ing (data not shown), suggesting that higher baselineproteasome activity (via increased PSMB5 protein expres-sion), not Psmb5 inactivating mutations, likely contributeto the resistance observed in these BzR lines. In addition,these BzR lines showed cross-resistance to the boronicacid next-generation proteasome inhibitor, MLN2238, aswell as the epoxyketone next-generation proteasomeinhibitor, carfilzomib (Kyprolis). Although the 595 BzRline also showed cross-resistance to the classical aldehydeproteasome inhibitor,MG-132, this line showed increasedsensitivity to the multiple myeloma drug, melphalan,whereas the 589 BzR line maintained sensitivity to thesecompounds (MG-132, melphalan; Table 1; Fig. 1A). Nei-ther the BzS nor the BzR lines responded to drugs knownto be ineffective as single agents against multiple myelo-ma: vincristine, hydroxyurea, and fludarabine (data notshown). These data suggest that not only is the resistance

Figure 1. Identification of bortezomib-responsive genes by GEP. A, chemical structures of the proteasome inhibitors bortezomib, MLN2238, and carfilzomiband the alkylating agent melphalan. B, heatmap of bortezomib-responsive genes common to mouse and human plasma cell malignancy in vitrousing 3 BzS mouse lines (595 in duplicate as 595.2) treated with 66 nmol/L bortezomib and 2 HMCLs treated with 33 nmol/L bortezomib. Mouse cells werecollected at 0, 2, 8, 16, and 24 hours and human cells at 0, 16, and 24 hours after bortezomib treatment. Columns represent time points and rowsrepresent genes. Columns are ordered first by cell line and second by ascending time; genes are ordered by hierarchical cluster analysis. Color indicates foldchange, which was determined for each gene by comparing the average of the 16- and 24-hour time points to the 0-hour time point in each time course.A full gene list has been provided in Supplementary Table S1.






observed in the BzR lines specific to bortezomib and itsnext-generation derivative (MLN2238) and sustainedover time but that the resistant phenotype may be over-come with other drugs.

Bortezomib resistance is associated with changes intranscription that are predictive of patient outcomes

We were particularly interested in determining thetranscriptional differences in gene expression between

Figure 2. Kinetic similarities and differences in bortezomib (Bz) response in BzS and BzRmouse lines. A, kill curve analysis of 3 representative pairs of BzS andBzRmouse cell lines (595, left; 589, middle; 638, right). Cell viability data were collected over a range of 0 to 128 nmol/L bortezomib. Open squares representBzS lines and filled squares represent BzR lines. All data points shown were normalized to untreated controls. Error bars, which are shown but aresmaller than the data symbols, represent theSDof triplicate reads over each experiment. B, Kaplan–Meier curves showing significant differences in event-freesurvival (EFS; left) and OS (right) in patients from the MMTT3 drug trial clustered on the basis of the expression of the genes that most distinguishedmouse BzS and BzR cell lines (Supplementary Fig. S3C). Heatmaps of bortezomib-responsive genes across the 3 BzS mouse lines and their BzRcounterparts (595 in duplicate) treated with 66 nmol/L bortezomib (C) and collected at 0, 2, 8, 16, and 24 hours after treatment, highlighting those geneswhose responses are most significantly different between BzS and BzR lines (D). Columns represent time points and rows represent the genes. Columns areordered first by sensitivity and second by sample; genes are ordered by hierarchical cluster analysis. Colors indicate fold change, which was determined foreach gene by comparing the average of the 16- and 24-hour time points to the 0-hour time. A full gene list for B has been provided in Supplementary Table S2.

Stessman et al.





these BzS and BzR cells lines and, thus, examined timecourses of response by GEP following the same, sublethalbortezomib treatment in BzR lines (Supplementary Fig.S1E) as used on their BzS counterparts (Fig. 1). To globallyvisualize these expression patterns, we analyzed the BzRtranscriptional response to bortezomib as we had the BzScell lines described earlier.First, a pairwise comparisonof theBzS andBzRbaseline

gene expression in the absence of bortezomib treatmentwas conducted revealing a 51-gene expression signaturethat statistically distinguished (fold change > 2; P < 0.05;Student t test) sensitive and resistant lines (Supplemen-tary Fig. S3C). Of these 51 mouse genes, 23 had humanhomologs. To test the predictive power of this 23-genemodel, we queried the gene expression signatures of 210patients from the MMTT3 human drug trial (28). Unsu-pervised clustering of these patients based on the 23-genemodel identified 2 groupswhose PFS and overall survival(OS) were significantly different (log-rank test; Fig. 2B).These results suggest that our in vitro mouse model ofbortezomib-resistance has predictive value in humanmultiple myeloma drug trials, which include bortezomib.To further define the differences in the transcriptional

response to bortezomib in BzS and BzR cell lines, weconducted a combined analysis of the BzS and BzR borte-zomib transcriptional profiles identifying 219 genes thatchanged significantly in both the sensitive or the resistantgroup in response to drug (P < 0.001; jfold changej > 2;pairwise 2 group t test; Supplementary Table S2). Exam-inationof heatmapclusters showed that themajority of theresponse to bortezomib in the sensitive cell lines wasconserved in the resistant cell lines (Fig. 2C and Supple-mentary Table S2), illustrating a common programmedtranscriptional response of cell lines to bortezomib regard-less of their overall sensitivity (measured by IC50) to thecompound.However, 29 genesweredifferentially respon-sive to bortezomib in vitro between BzS and BzR lines(P<0.001; jfoldchangej>2; Fig. 2D).Among thesewere theupregulation of a cluster of NRF-2–mediated oxidativestress response genes:Hspb1,Dnajb1,Hspa1a,Hspa1b, andDdit3 [also known as CCAAT/enhancer-binding proteinhomologous protein (CHOP); ref. 29], which has beenobserved in other BzS human cell lines (12, 29), suggesting

that these may serve as a signature of bortezomib-medi-ated cell death, which is unique to BzS compared withderived BzR cells.

Using CMAP to identify drugs with high correlationto bortezomib-associated death by GEP

Biomarkers that are associated with emerging resis-tance to bortezomib and yet may define sensitivity toalternative therapies, remain ill defined (33). Althoughwe do see a common gene signature of response tobortezomib and have identified similar biomarkers acrossBzR cell lines compared with their sensitive counterparts,we have also observed differences in response to second-ary therapies across BzR cell lines (Table 1). Indeed, notall patients with bortezomib-refractory multiple mye-loma respond similarly to secondary therapies. Giventhese differences, we aimed to use the individual cell lineGEP data described earlier to identify non-proteasomeinhibitor drugs that could target BzR cell populationsusing CMAP (Broad Institute, Boston, MA; ref. 22). TheCMAP database contains treatment-induced transcrip-tional signatures from 1,309 bioactive compounds in 4human cancer cell lines. An input signature can be used toquery the database for correlated drug signatures. Tocreate these input signatures, GSEA was used to identifyin individual paired mouse GEPs those genes that weremost different in BzS versus BzR lines following drugtreatment (24 vs. 0 hour data comparison; jfold changej> 2). These genes were then used to query CMAP toidentify drugs that induced expression signatures thatwere similar (positive correlation) or dissimilar (negativecorrelation) to the differential expression response of eachBzS lines relative to its BzR line when treated with borte-zomib. We hypothesized that such drugs, particularlythose with positive correlations to the differentialresponse, may have the potential to kill the BzR lines.

The CMAP query produced a number of compoundswith signatures that were predicted as significantly cor-related (positive) or anticorrelated (negative) with theBzS/BzR differential response (Table 2 and Supplemen-tary Table S3). Broadly, drugs promoting inhibition of theproteasome–ubiquitin pathway, NF-kB, HSP90, proteinsynthesis, and microtubules showed positive correlation,

Table 1. Drug IC50a table for BzS and BzR mouse lines

Drug595BzS

595BzR

Foldincreaseb

589BzS

589BzR

Foldincreaseb

638BzS

638BzR

Foldincreaseb

Bortezomib, nmol/L 23 112 4.9 22 96 4.4 36 102 2.8MLN2238, nmol/L 28 154 5.5 39 180 4.6 N/A N/A N/ACarfilzomib, nmol/L 28 71 2.5 41 77 1.9 N/A N/A N/AMG-132, nmol/L 42 95 2.3 80 85 1.1 N/A N/A N/AMelphalan, mmol/L 240 50 0.2 80 100 1.3 N/A N/A N/A

Abbreviation: N/A, not available.aIC50 calculations were determined by a sigmoidal dose–response equation in triplicate over multiple experiments.bFold increase is representative of the BzR line normalized to its BzS control.






whereas drugs that inhibit the cell cycle showed negativecorrelation with the differential responses in all pairs(Table 2 and Supplementary Table S3). These predictionswere not surprising given the known modes of action forbortezomib (34). Interestingly, several HDAC inhibitors(HDACi) were significantly positively correlatedwith thedifferential bortezomib response in 1of 3pairs of cell lines,595 (P < 2 � 10�5; Table 2; Supplementary Table S3), butwere not predicted in the other 2 highlighting someheterogeneity among these cell line pairs. In fact, this wasthe only drug family with consistent drug predictions foreach drug across the 3 cell lines queried (SupplementaryTable S3). Because HDACis have been reported as havingsynergistic effects when combined with other multiplemyeloma treatments (34) andare currently in clinical trialsfor refractorymultiplemyeloma (35),we chose to evaluatethis result further. We hypothesized that the 595 pair oflines would respond differently toHDACi treatment thanthe other pairs of lines based on the CMAP results.

On the basis of the HDACi prediction, 3 HDAC inhi-bitorswere chosen to test 2 representative pairs of BzS andBzR mouse lines (595 and 589) for in vitro sensitivity. The595 BzR line showed enhanced sensitivity to the HDACistrichostatin A and vorinostat (SAHA; Fig. 3A and B, top),which were chosen from the CMAP-predicted drug list(Supplementary Table S3), as well as panobinostat (Fig.3C, top). This was a cytotoxic response as evidenced byCHOPinduction (SupplementaryFig. S4A; refs. 12, 34, 36).In contrast, the 589 BzR line (not predicted for enhanced

sensitivity toHDACis byCMAP) showed cross-resistanceto all 3 HDACi compounds compared with the BzS line(Fig. 3A–C, bottom). This panobinostat-sensitive pheno-type has been observed in an additional BzR mouse cellline (bortezomib IC50s: 34 nmol/L (BzS) and 63 nmol/L(BzR); panobinostat IC50s: 37 nmol/L (BzS) and 22 nmol/L (BzR)) and indicates that CMAPmay have the ability toidentify differences in secondary drug response in thesecell line models.

A subset of bortezomib-resistantmouse cell lines hasgreater in vivo sensitivity to HDAC inhibitors

It is well documented that the myeloma response tosome chemotherapeutic agents is dependent not only onthe stromal microenvironment of the bone marrow (37),but also on a complete immune system (38) highlightingthe use of our immunocompetent mouse model system.To determine whether the differential in vitro drugresponses were maintained in vivo, the 595 BzS and BzRcell lines were adoptively transferred back into syngeneicrecipients. The untreated BzS mouse phenotype seemedto be significantly less severe compared with BzR mice,which reachedmoribundity quickly at amedian timeof 16days (P ¼ 0.033).

To better compare the disease burden in these animals,both 2[18F]fluoro-2-deoxy-D-glucose (FDG)- and fluoro-L-thymidine (FLT)-PET imaging were used. FDG-PETimaging of representative animals from each groupshowed that the BzS cells home to the bone marrow,whereas the BzR cells are dispersed in moribund animalswith fewer characteristic "hot spots" in the long bones(Fig. 4A). Representative histopathologic analyses of bonemarrows and soft tissues from these animals showed thatthemalignant plasma cellswere present in themarrows ofboth animals but were notably absent or minimal in thesoft tissues of BzS mice (Supplementary Fig. S4B). ThoseBzS and BzR cells that home to the bone marrow hadsimilar metabolic activity (FDG-PET); however, BzS cellshad significantly higher rates of proliferation (FLT-PET) invivo even though the BzS and BzR growth rates weresimilar in vitro (Fig. 4B). Treatment of BzS and BzR micewith bortezomib in vivo (n ¼ 6 mice/group) showed thatBzS mice received a significantly greater survival advan-tage with bortezomib treatment than their BzR counter-parts (Fig. 4C) whose survival did not differ from BzRvehicle-treated mice (data not shown) in agreement withour in vitro data (Fig. 2A). This was directly correlatedwith significantly decreased tumor burden (FDG-PET) inBzS mice treated with bortezomib compared with vehicle(Fig. 4D). These results indicate thatmyeloma cell homingmay play a role in bortezomib sensitivity and that theimmunophenotype associated with extramedullary hom-ing may be associated with bortezomib resistance in vivo.

To specifically determinewhether 595 BzRmice receivea survival advantage from panobinostat treatment aspredicted in vitro, mice were injected with BzR cells andtreated with either vehicle (n¼ 5) or panobinostat (n¼ 6),and the time to death for each group was compared.

Table 2. Families of drugs with expressionpatterns significantly correlated to thebortezomib response in mouse lines usingCMAP

Predictiona

Drug family 595spb 638bm 589bm

Proteasome inhibitors andUPS modulators

þ þ þ

NF-kB inhibitors þ þ þAlkylating agents þ þ þHSP90 inhibitors þ þ þProtein synthesis inhibitors þ þ þMicrotubule inhibitors þ þ þCDK/TopoI/TopoII/cell-cycleinhibitors

� � �

HDAC inhibitors þ np np

aSignificant predictions (P < 0.05) were made for each of themouse line pairs individually. A positive prediction (þ) indi-cates that the drug family induces a similar expressionsignature to the BzS signature. A negative prediction (�)indicates that the drug family induces an opposite expres-sion signature to theBzS signature. An "np" indicates that nodrug prediction was made.b595 is a combination of the 595 and 595.2 datasets.

Stessman et al.





Consistent with our in vitro findings (Figs. 2A and 3C),panobinostat treatment significantly increased the OS ofBzR mice (Fig. 4E) and decreased the tumor burden(Fig. 4F) in these animals. These in vivo results not onlyconfirm our in vitro findings but further suggest that theremaybe agreater benefit fromHDAC inhibitor therapyas asalvage therapy in some refractory myelomas.

DiscussionIn this study, we defined the malignant plasma cell

response to bortezomib and characterized signatures ofbortezomib-sensitivity and -resistance by GEP using celllines derived from the Bcl-XL/Myc double transgenicmouse model. Although a number of bortezomib treat-ment studies usingHMCLs, primary patient samples, andnonmyelomaprimarypatient samples havebeen reported(25, 28, 39–41), they did not provide a kinetic analysis ofBzS and BzR cells exposed to bortezomib over time nordid they explore individual differences in bortezomib-resistance that may contribute to secondary drug efficacy.In addition, the availability of published bortezomib datahas allowed us to validate the Bcl-XL/Myc cell lines’

response to drugs in vitro further illustrating their use asa preclinical tool for asking pharmacogenomic questions.

Malignant plasma cells isolated in cell culture fromthese animals were generally sensitive (nmol/L concen-trations) to bortezomib. In addition, we found evidence ofa robust, conserved, and likely complex response to bor-tezomib in our mouse cell lines consistent with recentreports (28, 29, 39). This responsewas highly conserved inwell-characterized HMCLs and in human patient clinicaltrial samples. Perhaps, the most striking trend was thecoordinated upregulation of the majority of the constitu-tive proteasomal subunits in response to bortezomibtreatment in both BzS and BzR cell lines, which has beenpreviously described (28, 29, 39) further validating ourmodel system.

The transcriptional response to bortezomib that weobserved in all cells regardless of their sensitivity to thedrug included the induction of the gene Psmd4, which ispart of the 19S proteasomal cap complex that selectivelytargets and binds ubiquitinated substrates for degradationby the 20S proteasome and whose amplification has beenassociated with high-risk myeloma (28). We did not

Figure 3. Mouse cell lines respond differently to HDAC inhibitors. A–C, kill curve analysis of 2 representative pairs of BzS and BzR mouse cell lines(595, top; 589, bottom). Cell viability data were collected over a range of 0 to 100 nmol/L trichostatin A (A), 0 to 4 mmol/L vorinostat (B), or 0 to 128 nmol/Lpanobinostat (C). Open squares represent BzS lines and filled squares represent BzR lines. All data points shown were normalized to untreated controls.Error bars, which are shownbut are smaller than the data symbols, represent theSDof triplicate reads over each experiment. Drug structures are shownbelowcorresponding kill curves. TSA, trichostatin A.






identify higher baseline Psmd4 expression in our BzR celllines, nor did we identify increases in copy number at themousePsmd4 locus by array comparative genomic hybrid-ization (mouse 3qF2.1 syntenic to human 1q21; data notshown). Therefore, although 1q21 amplification has beenassociated with a high-risk multiple myeloma phenotypein patients, our data would suggest variations in bortezo-mib responsemaybemore complex thanPsmd4 expressionlevels alone. Interestingly, we do observe differences inbaseline expression of Eno1 and Cxcr4 in BzR comparedwithBzS cell lines, bothofwhichhavebeenassociatedwitha poor prognostic outcome in patients with multiple mye-loma (42, 43). We believe that these data further highlightthe use of this system and underscore the need for com-panion analyses to validate these potential biomarkersusing clinical samples,whichwe are currently developing.

We found it particularly interesting that HDACis werepredicted by CMAP in this study because bortezomib isknown to downregulate the expression of class I HDACs(44), and HDACis have been shown to decrease the 20Schymotryptic activity of the proteasome (45), both medi-

ating cell death through the induction of CHOP andNOXA (46). In addition, HDAC6 has been shown to bea key regulator of aggresome activity, an alternative path-way for protein degradation in the absence of proteasomeactivity (47). Bortezomib response is correlated to theHDAC inhibitor response in only 1 of 3 pairs of mousecell lines, suggesting that (i) we might further identify asmall, very specific gene signature that is associated withan HDAC inhibitor response in vitro in these mouse celllines, and (ii) not all drug-resistantmyelomaswill respondsimilarly to HDACis highlighting some heterogeneitywithin our in vitro system that is also likely present in thebortezomib-refractory patient population. Although thereis a conserved transcriptional response to bortezomibbetween these 2 pairs of BzS and BzR cell lines, theyrespond very differently to HDACi compounds, suggest-ing that amore-favorable responsemightbe achievedwithHDACis suchaspanobinostat or vorinostat as a secondarytherapy in some, but not all relapsing myelomas. Indeed,wehave identified those genes that uniquely predicted theHDACi response in the 595 cell lines (Supplementary

Figure 4. BzR cells respondfavorably to panobinostat (Pan)in vivo. A, representative FDG-PETimaging of untreated 595 BzS andBzR cell-injected FVBN/Bl6syngeneic mice (1 mouse shownfrom each group). Red arrowsindicate tumor "hotspots." B,quantification of FDG- (blue) andFLT-PET (red) imaging of untreatedBzS and BzRmice [n¼ 2 per group(4 tissue sites scored from each);�, P < 0.05]. BM, bone marrow;SUV, standardized uptake value.C, Kaplan–Meier curve of BzS(black) and BzR (red) cell-injectedmice treated with bortezomib (Bz).Statistical significance wasdetermined using a one-tailedStudent t test. D, quantification ofFDG-PET–imaged BzS micetreated with either vehicle (left;white bar) or bortezomib [right; graybar; n ¼ 2 per group (4 tissue sitesscored from each); ��,P < 0.0005].E, Kaplan–Meier curve of BzR cell-injected mice treated with vehicle(n ¼ 5; black) or panobinostat (n ¼6; blue) until death. Statisticalsignificance of both Kaplan–Meiercurves was determined using aone-tailed Student t test. F,quantification of FDG-PET–imagedBzR mice treated with eithervehicle (left; white bar) orpanobinostat [right; gray bar; n¼ 2per group (4 tissue sites scoredfrom each); �, P < 0.05]. Statisticalsignificance of all PET imaging wasdetermined using a two-tailed,nonparametric Mann–Whitney test[95% confidence interval (CI)].

Stessman et al.





Table S4). Interestingly, these genes were enriched forbiosynthetic andmetabolic pathways (IPA), which we arecurrently investigating in BzR disease.We did not includea dual treatment (bortezomib and panobinostat) arm inour in vivo study; however, dual treatment of the cell linesin vitro indicates that there is synergy between these 2compounds (data not shown), which has been shown inanother mouse model of multiple myeloma (48). Themechanism of this synergy may be further informed bythese studies.Although CMAP produced additional predictions to

other classes of drugs, the true predictive power of thisapproach remains unclear. For example, alkylating agentswerepredictedbyCMAPbutdidnot includemelphalan forwhichwe have observed differential response between ourBzR cell lines. In fact, HDACis were the only drug familythat had a consistent prediction pattern across all 3 cell linepairs for all drugs predicted. This suggests that perhaps themost viable secondary drug candidates will be those withconsistent predictions. It remains to be seen if there aresimilar predictions of response in subsets of patients receiv-ing bortezomib/panobinostat combination therapy; how-ever, our data suggest that refinement of thismodelmay beeffective for predicting these cases in advance.With the initial success of combination bortezomib and

HDACi treatments for refractory patients in the clinic (49)and our indication of differential response to HDACis invitro and in vivo, our model system provides additionalopportunities to characterize sensitivity and resistance toHDACis in vitro and in vivo. It is apparent that theapproach taken in this study identifies a novel use forGEP in the repurposing of drugs as secondary therapies inmyeloma. The prediction of HDACis as a secondarytherapyhere suggests that apanobinostat responseprofileof emerging resistance might be further elucidated that

could provide preclinical support for personalized med-icine approaches in multiple myeloma.

Disclosure of Potential Conflicts of InterestH.A.F. Stessman has a commercial research grant from Millennium

Pharmaceuticals, Inc: The Takeda Oncology Company. No potential con-flicts of interest were disclosed by the other authors.

Authors' ContributionsConception and design: H.A.F. Stessman, B.G. Van NessDevelopment of methodology: H.A.F. Stessman, A. Sarver, A. Mansoor,J.J. Sunderland, C.L. Myers, B.G. Van NessAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): H.A.F. Stessman, A. Mansoor, S.A. Walsh, J.J.Sunderland, S. JanzAnalysis and interpretation of data (e.g., statistical analysis, biostatis-tics, computational analysis): H.A.F. Stessman, L.B. Baughn, A. Sarver,T. Xia, R. Deshpande, A. Mansoor, S.A. Walsh, M.A. Linden, F. Zhan,S. Janz, C.L. MyersWriting, review, and/or revision of themanuscript:H.A.F. Stessman, L.B.Baughn, A. Sarver, A. Mansoor, N.G. Dolloff, M.A. Linden, C.L. Myers,B.G. Van NessAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): H.A.F. Stessman, A. MansoorStudy supervision: H.A.F. Stessman, B.G. Van Ness

AcknowledgmentsThe authors thank Dr. Heather Nelson, Jacquelyn Kuriger, and

Anthony Rizzardi, University of Minnesota, for use of laboratory equip-ment and supplies, LingHu,University of Iowa, formouse husbandry andtechnical assistance,Michael R.Acevedo,C.N.M.T.,University of Iowa, forPET imaging and quantification assistance, and Tsu Wu, University ofMinnesota, for laboratory support. The GEP was conducted at the Uni-versity of Minnesota Genomic Center.

Grant SupportB.G. Van Ness received research funding for this project from Millen-

nium Pharmaceuticals, Inc. and Onyx Pharmaceuticals.The costs of publication of this article were defrayed in part by the

payment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received November 27, 2012; revised March 20, 2013; accepted March20, 2013; published OnlineFirst March 27, 2013.

References1. Kubagawa H, Vogler LB, Capra JD, Conrad ME, Lawton AR, Cooper

MD. Studies on the clonal origin of multiple myeloma. Use of individ-ually specific (idiotype) antibodies to trace the oncogenic event to itsearliest point of expression in B-cell differentiation. J Exp Med 1979;150:792–807.

2. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA CancerJ Clin 2012;62:10–29.

3. Chen D, Frezza M, Schmitt S, Kanwar J, Dou QP. Bortezomib as thefirst proteasome inhibitor anticancer drug: current status and futureperspectives. Curr Cancer Drug Targets 2011;11:239–53.

4. Richardson PG, Barlogie B, Berenson J, Singhal S, Jagannath S, IrwinD, et al. A phase2studyof bortezomib in relapsed, refractorymyeloma.N Engl J Med 2003;348:2609–17.

5. Mahindra A, Laubach J, Raje N, Munshi N, Richardson PG, AndersonK. Latest advances and current challenges in the treatment of multiplemyeloma. Nat Rev Clin Oncol 2012;9:135–43.

6. Orlowski RZ, Kuhn DJ. Proteasome inhibitors in cancer therapy:lessons from the first decade. Clin Cancer Res 2008;14:1649–57.

7. Ciechanover A. Proteolysis: from the lysosome to ubiquitin and theproteasome. Nat Rev Mol Cell Biol 2005;6:79–87.

8. Kisselev AF, van der Linden WA, Overkleeft HS. Proteasome inhi-bitors: an expanding army attacking a unique target. Chem Biol2012;19:99–115.

9. LeBlanc R, Catley LP, Hideshima T, Lentzsch S, Mitsiades CS, Mit-siades N, et al. Proteasome inhibitor PS-341 inhibits human myelomacell growth in vivo andprolongs survival in amurinemodel. Cancer Res2002;62:4996–5000.

10. Adams J. The proteasome: a suitable antineoplastic target. Nat RevCancer 2004;4:349–60.

11. Lee AH, Iwakoshi NN, Anderson KC, Glimcher LH. Proteasome inhi-bitors disrupt the unfolded protein response in myeloma cells. ProcNatl Acad Sci U S A 2003;100:9946–51.

12. Obeng EA, Carlson LM, Gutman DM, Harrington WJ Jr, Lee KP, BoiseLH.Proteasome inhibitors induce a terminal unfoldedprotein responsein multiple myeloma cells. Blood 2006;107:4907–16.

13. Cheung WC, Kim JS, Linden M, Peng L, Van Ness B, Polakiewicz RD,et al. Novel targeted deregulation of c-Myc cooperates with Bcl-X(L) tocause plasma cell neoplasms inmice. J Clin Invest 2004;113:1763–73.

14. Boylan KL, GosseMA, Staggs SE, Janz S, Grindle S, Kansas GS, et al.A transgenic mouse model of plasma cell malignancy shows pheno-typic, cytogenetic, and gene expression heterogeneity similar tohuman multiple myeloma. Cancer Res 2007;67:4069–78.

15. Atadja P. Development of the pan-DAC inhibitor panobinostat(LBH589): successes and challenges. Cancer Lett 2009;280:233–41.

16. Pei XY, Dai Y, Grant S. Synergistic induction of oxidative injuryand apoptosis in human multiple myeloma cells by the proteasome






inhibitor bortezomib and histone deacetylase inhibitors. Clin CancerRes 2004;10:3839–52.

17. Hideshima T, Bradner JE, Wong J, Chauhan D, Richardson P, Schrei-ber SL, et al. Small-molecule inhibition of proteasome and aggresomefunction induces synergistic antitumor activity in multiple myeloma.Proc Natl Acad Sci U S A 2005;102:8567–72.

18. Catley L, Weisberg E, Kiziltepe T, Tai YT, Hideshima T, Neri P, et al.Aggresome induction by proteasome inhibitor bortezomib and alpha-tubulin hyperacetylation by tubulin deacetylase (TDAC) inhibitorLBH589 are synergistic in myeloma cells. Blood 2006;108:3441–9.

19. Campbell RA, Sanchez E, Steinberg J, Shalitin D, Li ZW, Chen H, et al.Vorinostat enhances the antimyeloma effects of melphalan and bor-tezomib. Eur J Haematol 2010;84:201–11.

20. Ocio EM, Vilanova D, Atadja P, Maiso P, Crusoe E, Fernandez-LazaroD, et al. In vitro and in vivo rationale for the triple combination ofpanobinostat (LBH589) and dexamethasone with either bortezomib orlenalidomide in multiple myeloma. Haematologica 2010;95:794–803.

21. Sanchez E, Shen J, Steinberg J, Li M, Wang C, Bonavida B, et al. Thehistone deacetylase inhibitor LBH589 enhances the anti-myelomaeffects of chemotherapy in vitro and in vivo. Leuk Res 2011;35:373–9.

22. LambJ,CrawfordED,PeckD,Modell JW,Blat IC,WrobelMJ, et al. TheConnectivity Map: using gene-expression signatures to connect smallmolecules, genes, and disease. Science 2006;313:1929–35.

23. Wei G, Twomey D, Lamb J, Schlis K, Agarwal J, Stam RW, et al. Geneexpression-based chemical genomics identifies rapamycin as a mod-ulator of MCL1 and glucocorticoid resistance. Cancer Cell 2006;10:331–42.

24. Hideshima T, Mitsiades C, Akiyama M, Hayashi T, Chauhan D,Richardson P, et al. Molecular mechanisms mediating antimyelomaactivity of proteasome inhibitor PS-341. Blood 2003;101:1530–4.

25. Altun M, Galardy PJ, Shringarpure R, Hideshima T, LeBlanc R,Anderson KC, et al. Effects of PS-341 on the activity and compo-sition of proteasomes in multiple myeloma cells. Cancer Res 2005;65:7896–901.

26. Lombardi L, Poretti G, Mattioli M, Fabris S, Agnelli L, Bicciato S, et al.Molecular characterization of human multiple myeloma cell lines byintegrative genomics: insights into the biology of the disease. GenesChromosomes Cancer 2007;46:226–38.

27. Greenstein S, Krett NL, Kurosawa Y, Ma C, Chauhan D, Hideshima T,et al. Characterization of the MM.1 humanmultiple myeloma (MM) celllines: a model system to elucidate the characteristics, behavior, andsignaling of steroid-sensitive and -resistant MM cells. Exp Hematol2003;31:271–82.

28. Shaughnessy JD Jr, Qu P, Usmani S, Heuck CJ, Zhang Q, Zhou Y,et al. Pharmacogenomics of bortezomib test-dosing identifieshyperexpression of proteasome genes, especially PSMD4, as novelhigh-risk feature in myeloma treated with total therapy 3. Blood2011;118:3512–24.

29. Weniger MA, Rizzatti EG, Perez-Galan P, Liu D, Wang Q, Munson PJ,et al. Treatment-induced oxidative stress and cellular antioxidantcapacity determine response to bortezomib in mantle cell lymphoma.Clin Cancer Res 2011;17:5101–12.

30. Suzuki E, DemoS, Deu E, Keats J, Arastu-Kapur S, Bergsagel PL, et al.Molecularmechanisms of bortezomib resistant adenocarcinoma cells.PloS One 2011;6:e27996.

31. de Wilt LH, Jansen G, Assaraf YG, van Meerloo J, Cloos J, SchimmerAD, et al. Proteasome-based mechanisms of intrinsic and acquiredbortezomib resistance in non–small cell lung cancer. Biochem Phar-macol 2012;83:207–17.

32. Perez-Galan P, Mora-Jensen H, Weniger MA, Shaffer AL III, RizzattiEG, Chapman CM, et al. Bortezomib resistance in mantle cell lym-

phoma is associated with plasmacytic differentiation. Blood 2010;117:542–52.

33. Morabito F, Recchia AG, Mazzone C, Gentile M. Targeted therapy ofmultiple myeloma: the changing paradigm at the beginning of the newmillennium. Curr Cancer Drug Targets 2012;12:743–56.

34. Hideshima T, Richardson PG, Anderson KC. Mechanism of action ofproteasome inhibitorsanddeacetylase inhibitorsandthebiologicalbasisof synergy in multiple myeloma. Mol Cancer Ther 2011;10:2034–42.

35. Stewart AK. Novel therapeutics in multiple myeloma. Hematology2012;17(Suppl 1):S105–8.

36. Rao R, Nalluri S, Kolhe R, Yang Y, Fiskus W, Chen J, et al. Treatmentwith panobinostat induces glucose-regulated protein 78 acetylationand endoplasmic reticulum stress in breast cancer cells. Mol CancerTher 2010;9:942–52.

37. Hideshima T, Anderson KC. Novel therapies in MM: from the aspect ofpreclinical studies. Int J Hematol 2011;94:344–54.

38. Chauhan D, Singh AV, Brahmandam M, Carrasco R, Bandi M, Hide-shima T, et al. Functional interaction of plasmacytoid dendritic cellswith multiple myeloma cells: a therapeutic target. Cancer Cell 2009;16:309–23.

39. Mulligan G, Mitsiades C, Bryant B, Zhan F, Chng WJ, Roels S, et al.Gene expression profiling and correlation with outcome in clinical trialsof the proteasome inhibitor bortezomib. Blood 2007;109:3177–88.

40. Goy A, Bernstein SH, McDonald A, Pickard MD, Shi H, Fleming MD,et al. Potential biomarkers of bortezomib activity in mantle cell lym-phoma from the phase 2 PINNACLE trial. Leuk Lymphoma 2010;51:1269–77.

41. JiaH,GeF, Lu X, ZengH, Li L, Chen Z, et al. Proteomics of apoptosis ofmultiple myeloma cells induced by proteasome inhibitor PS-341.Zhong Nan Da Xue Xue Bao Yi Xue Ban 2010;35:784–91.

42. Shaughnessy JD Jr, Zhan F, Burington BE, Huang Y, Colla S, Hana-mura I, et al. A validated gene expression model of high-risk multiplemyeloma is defined by deregulated expression of genes mapping tochromosome 1. Blood 2007;109:2276–84.

43. VandeBroek I, Leleu X, Schots R, Facon T, Vanderkerken K, VanCampB, et al. Clinical significance of chemokine receptor (CCR1, CCR2 andCXCR4) expression in human myeloma cells: the association withdisease activity and survival. Haematologica 2006;91:200–6.

44. Kikuchi J, Wada T, Shimizu R, Izumi T, Akutsu M, Mitsunaga K, et al.Histone deacetylases are critical targets of bortezomib-induced cyto-toxicity in multiple myeloma. Blood 2010;116:406–17.

45. Catley L, Weisberg E, Tai YT, Atadja P, Remiszewski S, HideshimaT, et al. NVP-LAQ824 is a potent novel histone deacetylase inhibitorwith significant activity against multiple myeloma. Blood 2003;102:2615–22.

46. Rao R, Nalluri S, Fiskus W, Savoie A, Buckley KM, Ha K, et al. Role ofCAAT/enhancer binding protein homologous protein in panobinostat-mediated potentiation of bortezomib-induced lethal endoplasmicreticulum stress in mantle cell lymphoma cells. Clin Cancer Res2010;16:4742–54.

47. Kawaguchi Y, Kovacs JJ, McLaurin A, Vance JM, Ito A, Yao TP. Thedeacetylase HDAC6 regulates aggresome formation and cell viabilityin response to misfolded protein stress. Cell 2003;115:727–38.

48. Chesi M, Matthews GM, Garbitt VM, Palmer SE, Shortt J, Lefebure M,et al. Drug response in a genetically engineered mouse model ofmultiple myeloma is predictive of clinical efficacy. Blood 2012;120:376–85.

49. Badros A, Burger AM, Philip S, Niesvizky R, Kolla SS, Goloubeva O,et al. Phase I study of vorinostat in combination with bortezomib forrelapsed and refractory multiple myeloma. Clin Cancer Res 2009;15:5250–7.

Stessman et al.





2013;12:1140-1150. Published OnlineFirst March 27, 2013.Mol Cancer Ther Holly A.F. Stessman, Linda B. Baughn, Aaron Sarver, et al. a Mouse Myeloma ModelProfiling Bortezomib Resistance Identifies Secondary Therapies in

Updated version

10.1158/1535-7163.MCT-12-1151doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2013/07/02/1535-7163.MCT-12-1151.DC1

Access the most recent supplemental material at:

Cited articles

http://mct.aacrjournals.org/content/12/6/1140.full#ref-list-1

This article cites 49 articles, 25 of which you can access for free at:

Citing articles

http://mct.aacrjournals.org/content/12/6/1140.full#related-urls

This article has been cited by 3 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://mct.aacrjournals.org/content/12/6/1140To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-12-1151

http://mct.aacrjournals.org/content/suppl/2013/07/02/1535-7163.MCT-12-1151.DC1

http://mct.aacrjournals.org/content/12/6/1140.full#ref-list-1

http://mct.aacrjournals.org/content/12/6/1140.full#related-urls

http://mct.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://mct.aacrjournals.org/content/12/6/1140


ﬁling bortezomib resistance identiﬁes secondary therapies ... · bzr cells were generated by...

Documents