tasquinimod is an allosteric modulator of hdac4 survival ... · (kd 10–30 nmol/l) to the...
TRANSCRIPT
Therapeutics, Targets, and Chemical Biology
Tasquinimod Is an Allosteric Modulator of HDAC4 SurvivalSignalingwithin theCompromisedCancerMicroenvironment
John T. Isaacs1,2, Lizamma Antony1, Susan L. Dalrymple1, W. Nathaniel Brennen1, Stephanie Gerber1,Hans Hammers1, Michel Wissing1, Sushant Kachhap1, Jun Luo2, Li Xing3, Per Bj€ork4, Anders Olsson4,Anders Bj€ork4, and Tomas Leanderson4,5
AbstractTasquinimod is an orally active antiangiogenic drug that is currently in phase III clinical trials for
the treatment of castration-resistant prostate cancer. However, the target of this drug has remainedunclear. In this study, we applied diverse strategies to identify the histone deacetylase HDAC4 as a targetfor the antiangiogenic activity of tasquinimod. Our comprehensive analysis revealed allosteric binding(Kd 10–30 nmol/L) to the regulatory Zn2þ binding domain of HDAC4 that locks the protein in a confor-mation preventing HDAC4/N-CoR/HDAC3 complex formation. This binding inhibited colocalization ofN-CoR/HDAC3, thereby inhibiting deacetylation of histones and HDAC4 client transcription factors, such asHIF-1a, which are bound at promoter/enhancers where epigenetic reprogramming is required for cancer cellsurvival and angiogenic response. Through this mechanism, tasquinimod is effective as a monotherapeuticagent against human prostate, breast, bladder, and colon tumor xenografts, where its efficacy could befurther enhanced in combination with a targeted thapsigargin prodrug (G202) that selectively kills tumorendothelial cells. Together, our findings define a mechanism of action of tasquinimod and offer a perspectiveon how its clinical activity might be leveraged in combination with other drugs that target the tumormicroenvironment. Cancer Res; 73(4); 1386–99. �2012 AACR.
IntroductionOral dosing with tasquinimod (TasQ; Fig. 1A) results in a
doubling of progression free survival in patients with meta-static castration-resistant prostate cancer (CRPC; ref. 1). Basedon these results, phase III registration trials for metastaticCRPC are ongoing. Preclinical studies show that TasQ sup-presses reciprocal cross-talk between cancer and tumor infil-trating host cells such as endothelial cells, myeloid-derivedsuppressor cells (MDSC), and macrophages (2–7). Such sup-pression inhibits tumor angiogenesis needed for cancer cellsurvival as documented by a variety of assays including endo-thelial capillary tube formation, aortic ring assay, chorio-allantoic membrane assay, real-time tumor blood flow andpO2 measurements, tumor blood vessel density, and tumorhypoxic and apoptotic fractions (2–6). This suppression issignificant because cancers outgrow their blood supply, which
results in the tumor microenvironment becoming acidic, hyp-oxic, and low in nutrients (8). Continued growth in such astressfulmicroenvironment requires cancer and tumor infiltrat-ing host cells to initiate epigenetic reprogramming, leading toan "angiogenic switch," which involves downregulation ofinhibitory genes such as thrombospondin-1 coupled with theupregulation of stimulatory genes such as hypoxia-inducedfactor-1a (i.e., HIF-1a; ref. 9).
TasQ's suppression of this epigenetic angiogenic switch (5)provides a rationale for its efficacy in preclinical models againstprimary cancer as well as its inhibition of metastasis when usedas monotherapy (3, 10) and explains why it enhances efficacywhen combined with androgen ablation (4), taxane-based che-motherapy (4), or fractionated radiation (6), therapies knownto induce a stressful tumor microenvironment. The focus ofthis report is the identification of TasQ's molecular targets touse this information to rationally design combinational appro-aches to enhance therapeutic benefit in the clinic.
Materials and MethodsReagents
Reagents: G202 (GenSpera); Tasquinimod (Active BiotechResearch AB); Trichostatin-A (Sigma-Aldrich); Acetyllysine-7-amino-methylcoumarin (i.e., total HDAC substrate) andtrifluoromethy-acetylysine7-amino-methylcoumarin (i.e.,HDAC4 specific substrate; Bachem Inc.). Antibodies (Ab)against the indicated antigen: acetyl-histone H3: rabbitpolyclonal Ab (Millipore, Cat # 06-599); acetyl-lysine: rabbit
Authors' Affiliations: 1The Sidney Kimmel Comprehensive Cancer Centerat Johns Hopkins; 2The Brady Urological Institute-Department of Urology,The Johns Hopkins University School of Medicine, Baltimore, Maryland;3Inflammation/Immunology Chemistry-Pfizer Worldwide Research andDevelopment, Cambridge, Massachusetts; 4Active Biotech, AB; and5Immunology Group, Lund University, Lund, Sweden
Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).
Corresponding Author: John T. Isaacs, The Johns Hopkins University,1650 Orleans St. - RM 1M44, Baltimore, MD 21287. Phone: 410-955-7777;Fax: 410-614-8397; E-mail: [email protected]
doi: 10.1158/0008-5472.CAN-12-2730
�2012 American Association for Cancer Research.
CancerResearch
Cancer Res; 73(4) February 15, 20131386
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
monoclonal Ab (Millipore, Cat # 06-933); Flag: clone M2:mouse monoclonal Ab (Sigma-Aldrich, Cat # F3165); GFP:mouse monoclonal Ab (Clontech, Cat # 632375); GST: rabbitpolyclonal Ab (Cell Signaling, Cat # 2622); HDAC1(H-11):mouse monoclonal Ab (Santa Cruz, Cat # sc-8410); HDCA2(C-8): mouse monoclonal Ab (Santa Cruz, Cat # sc-9959);HDAC3(B-12): mouse monoclonal Ab (Santa Cruz, Cat # sc-17795); HDAC4: rabbit polyclonal Ab (Active Motif, Cat #40969); HDAC5: rabbit polyclonal Ab (Cell Signaling, Cat #sc-2082); HDAC6(H-300): rabbit polyclonal Ab (Santa Cruz,Cat # 11420); HDAC7(C-18): goat polyclonal Ab (Santa Cruz,Cat # sc-11491); HDAC8(H-145): rabbit polyclonal Ab (SantaCruz, Cat # 11405); HDAC 9 Ab was a gift from Dr. ArthurZelent (Institute of Cancer Research, Sutton, Surrey, UnitedKingdom); HIF-1a(H-206): rabbit polyclonal Ab [Santa Cruz,Cat # sc-10790 (IP)]; HIF-1a: mouse monoclonal Ab [BDTransduction Laboratories, Cat # 610958 (IB)]; HIF-1a(H-206): rabbit polyclonal [Bethyl Laboratories, Cat # A300-286A(IB)]; N-CoR: rabbit polyclonal Ab (Bethyl Laboratories, Cat #A301-146A).
Recombinant wild-type and C669/H675-mutated HDAC4proteinspcDNA vector containing human N-terminal flag-tag full-
length wild-type histone deacetylase 4 (HDAC4) was obtainedfrom Addgene (Cat # 13821). To produce recombinant C669/H675 mutated HDAC4 protein, pcDNA vector containinghuman N-terminal flag-tag full-length wild-type HDAC4 wasmutated using the Quick Change Site Directed Mutagenesis kitfrom Agilent Technologies (Cat # 200523) according to manu-facturer's protocol. HEK-292Tcells were transfected with thesevectors and 48 hours latter, cells washed with phosphataseinhibitor from Roche (Cat# 0490684500); lysed in 50 mmol/LTris HCl pH 7.4 with 150 mmol/L NaCl/1% Triton and proteaseinhibitor cocktail tablet from Roche (Cat# 11836170001). Theresulting lysate was incubated overnight at 4�C with a 50%suspension of flag agarose from Sigma-Aldrich (Cat # A4596);lysate-agarose mixture centrifuged, washed twice, slurry trans-ferred to chromatography column from BioRad (Cat# 7311550);column washed multiple times without Triton; protein elutedwith 100 mg/mL 3X flag peptide from Sigma-Aldrich (Cat#F4799); and dialyzed in PBS and protein concentrated usingultrafiltration centrifugal filter concentrator [Amicon/Millipore(Cat # UFC203024PL)].Three-dimensional (3D)-endothelial sprouting assay was
conducted using a minimum of 5 replicate wells per drug doseper experiment, repeated 3 independent times as describedpreviously (11) using human umbilical vein endothelial cells(HUVECs; Lonza Walkersville, Inc.) with modification that20 mL of 10X concentrated Dulbecco's Modified Eagle'sMedia/10%FCS media conditioned by confluent human pri-mary lung fibroblasts (Lonza) diluted to 200 mL of growthfactor supplemented EGM-2 media (Lonza) was used insteadof fibroblast coculture.
Cancer modelsThe source, history, and characteristics of the normal
and malignant human epithelial lines used, as well as cell
culture conditions for their in vitro maintenance and thein vivo protocol for their xenografting into immune-deficientnude mice are as described previously (3, 12, 13). All lineswere mycoplasma negative using the MycoSensor PCR Assaykit (Agilent Technologies) and genetically authenticatedwithin the last 6 months using short tandem repeat pro-filing conducted by the Johns Hopkins Genetic ResourceCore Facility. In vitro growth curves were determined asdescribed (12, 13). Animal studies were conducted accord-ing to animal protocol MO09M434 approved by JohnsHopkins Animal Care and Use Committee. Daily orally TasQdosing via the drinking water, tumor volume measurements,tumor blood vessel density determination, and tissue oxy-genation expressed as mmHg were as described previously(3, 4). These experiments were repeated at 3 independenttimes for each xenograft.
Biochemical assaysCoimmunoprecipitation (co-IP) of whole cell lysates or
nuclear fractions was via a kit from Active Motif (NuclearComplex Co-IP kit, Cat# 54001) according to manufacturer'sprotocol. For co-IP with recombinant full length flag-taggedN-CoR (i.e., amino-acid 1-2,440) from Abcam (Cat #ab82239), 100 ng of the protein was incubated with 20 mLantiflag M2 affinity gel beads from Sigma-Aldrich (Cat#A2220) for 2 hours and then the mixture centrifuged, pel-leted beads washed 3 times, resuspended, and 20 ng ofrecombinant glutathione S-transferase (GST)-tagged HDAC4(i.e., amino acid 627-1,085) obtained from BPS Bioscience(Cat # 50004) added to the mixture, which was incubatedovernight at 4�C. The mixture was centrifuged, pellet washed3 times, and then immunoblotted (IB) with anti-GST anti-body. For co-IP with recombinant GST-tagged HDAC4, 20 ngof the protein in 500 mL of pH8 buffer (i.e., 50 mmol/L Tris,150 mmol/L NaCl, 1 mmol/L dithiothreitol, 10 mmol/L Zn,1% Triton-X100) was incubated with 20 mL GlutathioneSepharose 4B Beads from GE Life Sciences (Cat # 17-0756-01) for 2 hours and then the mixture centrifuged,pelleted beads washed 3 times, resuspended inpH 8 buffer, and 100 ng of recombinant flag-tagged-N-CoRadded. The mixture was incubated overnight at 4�C andthen the mixture centrifuged, pellet washed 3 times, andthen sample was IB with antiflag antibody.
For cell-based co-IP, cells transiently transfected withpcDNA vector containing either N-terminal flag-tag full-lengthwild-type HDAC4 from Addgene (Cat # 13821) or HDAC3from Addgene (Cat # 13819) were harvested after 48 hours,antiflag IP conducted as described for isolation of recombinantproteins. IP was then IB with indicated antibody conducted.The biochemical and cell-based co-IP experiment were repeat-ed independently at least 3 times with representative gelspresenting the results.
For detection of acetylated HIF-1a, cells were main-tained under hypoxic conditions (i.e., 2% O2) for 24 hoursand then lysed in radioimmunoprecipitation assay buffer(RIPA) buffer containing added Roche protease and phos-phatase inhibitors. Lysate was centrifuged, supernatantmixed with 4 mg acetylated lysine antibody and 40 mL Protein
HDAC4 and Stress Signaling
www.aacrjournals.org Cancer Res; 73(4) February 15, 2013 1387
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
A agarose (Santa Cruz, Cat # sc-2001), incubated overnightat 4�C, mixture centrifuged and pelleted beads washed4 times with RIPA buffer and then IB with anti-HIF-1aantibody conducted. These experiments were repeated 3independent times with a representative gel presented.
Total HDAC and isotype specific HDAC enzymatic activitywas assayed on a per cell basis using the appropriate substratesas described previously (14, 15). Recombinant human HDACisotypes were obtained commercial as follows: HDAC1 (Milli-pore, Cat#14-838); HDAC3 (BioMol/Enzo Life Sciences cat#SE-507); HDAC4 (Millipore cat#14-828); HDAC6 (BioMol/Enzocat#SE-508); and HDAC8 (Millipore cat#14-609). These experi-mentswere repeated aminimumof 3 independent timeswith 5replicates per time point.
Surface plasmon resonanceSurface plasmon resonance (SPR) analysis was carried out
with the Biacore 3000 system (GE Healthcare) as describedpreviously (16). Sensor chips, amine coupling kit, immobiliza-tion and running buffers, and regeneration solutions were asdescribed previously (16). Binding to TasQ was determined forhuman full length N-terminal GST-tagged HDAC4 (AbnovaCat# H00009759-P01). GST-tagged HDAC4 was immobilizedonto a CM5 chip through an amine linkage. This chip was usedto determine binding of full-length humanN-CoR (AbcamCat#ab82239). These experiments were repeated 3 independenttimes.
Computer docking modelingThe active versus inactive conformations of HDAC4 were
overlaid using the Ca atoms of crystal structures previous re-ported (17). Protein surfaces generated and systematically exam-ined for plausible binding pockets. The revealed bindingpockets were explored by iterative docking, during which bind-ing poses were refined by interplay of manual docking, Glideruns, and visual inspection of the modeled ligand-protein com-plex. More than 20 docking poses were generated. The finalmodel selected based on the conformational energy for theligand as well as the computed ligand-protein energy. The com-putational simulationwas conductedusing Schr€odinger'smolec-ular modeling environment Maestro (www.schrodinger.com).
HDAC4 knockdown and restorationStable shRNA HDAC4 knockdown cells were generat-
ed using pLK0.1 lentiviral vector from Open Biosystems(Thermo Scientific) containing the following specific shRNAsequences: shRNA1 (CCGGCGACTCATCTTGTAGCTTAT-TCTCGAGAATAAGCTACA AGATGAGTCGTTTTT); shRNA2(CCGGC GTGGGTTTCAACGTCAACATCTCGAGATGTTGA-CGTTGAAACCCACGTTTTT); shRNA3 (CCGGGCAGCTCA-AGAACAAGGAGAACTCGAGTTCTCCTTGTTCTTGA GCT-GCTTTTT); and the nonspecific shRNA sequence was (CCT-AAGGTTAAGTCGC CCTCGCTCGAGCGAGGGCGACTTAA-CCTAGGTTTTT). To generate HDAC4 restoration expressionvector resistant to shRNA2 knockdown, 2 nucleotides at posi-tions 9 and 12 in the sHRNA2 targeting region were changedto making it resistant to the HDAC4 shRNA without alteringthe amino acid sequence of wild-type protein.
HIF-1a transcriptional assaysPC-3 human prostate cancer cells stably integrated with
EGFPgene under a promoter containing 5 copies of the HIF-1a-driven hypoxia response element (HRE) of the VEGF geneas described previously (18) were exposed to different doses ofTasQ under both normoxic and hypoxic conditions for 24hours and cell lysates analyzed by IB for GFP from 3 indepen-dent experiments. HUVECs exposed to different doses of TasQunder both normoxic and hypoxic conditions were cotrans-fected with 1 mg of the p2.1 expression plasmid containingfirefly luciferase under a HIF-1a-driven HRE of the enolasegene and 0.1 mg of p-thymidine kinase (pTK)-Renilla controlplasmid containing Renilla luciferase under a TK-driven pro-mote using Fugene6 from Roche. The p2.1 and pTK-Renillaplasmids were a kind gift from Dr. Gregg Semenza (JohnsHopkins School of Medicine) and have been characterizedpreviously (19). After 6 hours, HUVEC cell lysates were ana-lyzed with a Dual Luciferase Reporter Assay kit from Promega(Cat# 1910). These experiments were repeated 3 independenttimeswith representative data reported. HumanHIF-regulatedcDNA plate array from Signosis, Inc. (Cat# AP-0111) wasconducted as per instructions from the manufacturer. Theseexperiments were repeated 3 independent times with repre-sentative data reported.
Gene expression profilingNormal prostate samples from organ donors (n ¼ 23) were
obtained as described previously (20). CRPC specimens (n ¼18) were autopsy specimens from 6 patients who died fromPCa, as previously reported (21). The use of autopsy specimensfor molecular analysis was approved by the Johns HopkinsMedicine Institutional Review Boards. Gene expression pro-filing was conducted 3 independent times according to theguidelines provided by the Agilent Whole Genome ExpressionMicroarray system (Agilent Technologies), using 2-colordesign and the results normalized to a reference samplederived from benign prostatic hyperplasia using the standardlocally weighted least squares regression procedure asdescribed previously (22).
StatisticsAll of the values reported are presented as means � SE of
representative data generated from 1 of a minimum of 3independent experiments in which there were a minimum of5 replicates per data point. Statistical analysis was conductedby a 1-way ANOVA with the Newman–Keuls test for multiplecomparisons.
ResultsTasquinimod inhibition of the adaptive response ofcancer and endothelial cells is enhanced in a stressfulmicroenvironment
Human cancer cells are severely stressed when xenograftedas minced tissue organoids within Matrigel because theseplugs initially lack a blood supply. Thus, cancer organoidsmust initiate an angiogenic switch, both to survive in thisnutrient poor/hypoxic microenvironment (i.e., pO2 in the
Isaacs et al.
Cancer Res; 73(4) February 15, 2013 Cancer Research1388
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
startingMatrigel plug is 10� 2mmHg vs. 40� 9mmHg in fullyoxygenated normal tissue) and to stimulate host angiogenicresponse. If a daily oral dose of TasQ (i.e., 10 mg/kg/d), whichmaintains serum drug concentration at 0.5 to 1 mmol/L (3), isinitiated upon inoculation of Matrigel containing 15 to 20 mgsof CWR22-RH human CRPC tissue suspension into castratedmale nude mice, 80% of animals (n¼ 10) are cured and tumorgrowth in the remaining 20% is profoundly inhibited (P <0.05; Fig. 1B). In contrast, 100% of untreated animals (n ¼ 10)develop an adequate tumor blood vessel density and thecancers grow eventually killing the host. If animals are initiatedon daily TasQ after tumors grow to a starting size of 0.250 cc,the cancers do not regress, but subsequent growth is pro-foundly inhibited (P < 0.05; Fig. 1B). Growth inhibition isassociatedwith 50% lower (P< 0.05) tumor blood vessel density(2.6 � 1.4% vs. 4.8 � 0.9% of tumor area in TasQ treated vs.untreated, respectively).
To clarify why TasQ is maximally effective when the tumormicroenvironment is most compromised, in vitro experimentswere used. Under standard culture conditions (5%CO2/21%O2/pH7.2 media), human PC cells (LNCaP, LAPC-4, and CWR-22Rv1) andHUVECs grow exponentially with doubling times ofapproximately 2 days. Under these standard conditions, TasQinhibits growth (P < 0.05) with IC50 values of approximately 50mmol/L for cancer and endothelial cells, values 50- to 100-foldhigher than therapeutic blood levels (Fig. 1C). This raises theissue of whether TasQ's potency and/or efficacy for growthinhibition and/or cell death are increased in a stressful micro-environment. To evaluate this, LNCaPs and HUVECs weretreated with 1 mmol/L TasQ (a dose that had no effect understandard in vitro conditions, but was therapeutic in vivo) in lowglucose/high CO2/low O2/acidic media to mimic the stressfulmicroenvironment within cancer sites. Under these compro-mised conditions, untreated LNCaPs and HUVECs survive, but
Figure 1. A, chemical structure ofTasQ (N-ethyl-N-phenyl-5-chloro-1,2-dihydro-4-hydroxy-1-methyl-2-oxo-3-quinoline-carboxamide. B,comparison of anticancer efficacy ofTasQ (10mg/kg/d via drinkingwater)against CWR22-RH human prostatecancer xenografts (n ¼ 10/group)growing in castrated hosts initiatedimmediately at time of tumorinoculation versus delayedtreatment starting on day 49posttumor inoculation when tumorswere 0.250 cc in size. Each symbolpoint is the mean tumor volume atthe indicated time. SEM is less than15%of themean for each time point.P < 0.05 for all treatment time pointsbeyond day 50 compared withcontrols. C, growth inhibitoryresponse of HUVEC and indicatedhuman prostate cell lines to 1-weektreatment with indicatedconcentration of TasQ. Results arenormalized to number of viable cellsin control cultures not exposed toTasQ. Statistical difference (P < 0.05)for TasQ-treated cells comparedwith untreated control cells. D, TasQdose–response inhibition of 3D ECsprouting in fibrin gels over a 7-day(D7) observation period versusuntreated (control) cells.
A BControl
Tasquinimod
Tasquinimod delayed
10 20 30 40 50
Days posttumor inoculation
60 70 80 90
Me
an
tu
mo
r vo
lum
e (
cc)
C D
0
50
100
1 µmol/L 10 µmol/L 50 µmol/L
Perc
en
t o
f co
ntr
ol
TasQ
HUVEC
D0 control D7 control
D7 0.5 µmol/L
D7 10 µmol/L
D7 1 µmol/L
2.50
2.00
1.50
1.00
0.50
0.00
LAPC-4
CWR22Rv1
LnCaP
Inhibition of EC sprouting
TasQ 0.5 µmol/L 50%
TasQ 1 µmol/L 75%
TasQ 10 µmol/L 90%
HDAC4 and Stress Signaling
www.aacrjournals.org Cancer Res; 73(4) February 15, 2013 1389
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
slow their growth (doubling time increases from 42� 5 to 144� 12 hours for LNCaP and 47 � 7 to 64 � 8 hours for HUVEC;P < 0.05). Addition of 1 mmol/L TasQ to such a compromisedcondition further decreases doubling time of HUVECs (88 � 9hours; P < 0.05). In contrast, addition of 1 mmol/L TasQ to sucha compromised conditions results in 52 � 9% (P < 0.05) ofLNCaPs dying within 1 week.
To evaluate whether TasQ also inhibits the functional re-sponse of endothelial cells, 3D in vitro angiogenic sprouting/tube formation assays were used. When HUVECs grown onmicrocarrier beads are embedded in fibrin gels with mediaconditioned by normal lung fibroblasts to provide angiogenicfactors (e.g., VEGF), typically secreted by cancer cells in ahypoxic microenvironment, cells "sprout" producing canalizedneovessel tubes within 1 week (Fig. 1D). If TasQ is added atculture initiation, sprouting is suppressed (P < 0.05) with an
IC50 value of 0.5 mmol/L (Fig. 1D). In contrast, if HUVECs areallowed to preform 3D neovessels and then TasQ is added, noregression occurs even at 50 mmol/L. These results documentthat TasQ's inhibition of cancer cells survival and endothelialangiogenic responses are profoundly enhanced by a stressfulmicroenvironment.
Tasquinimod suppresses hypoxia-induced decrease inhistone acetylation without lowering HDAC expressionor directly inhibiting HDAC activity
Survival in a stressful hypoxic microenvironment requiresepigenetic reprogramming in which global transcription isdecreased via decreased global histone acetylation, coupledwith increased transcription of a select group of survival genes(23–26). Decreased global histone acetylation stimulated byhypoxia (2% O2) was confirmed in human PC lines and TasQ
A B
E
C
D
LNCaP
Hypoxic
957E
/hTER
T
LNCaP
PC3
DU14
5
LAPC4
CWR22
- Rv1
Normoxic
Acetyl-H3
Acetyl-H3
Acetyl-H3
HDAC 1
HDAC 2
HDAC 3
HDAC 4
HDAC 6
HDAC 8
HDAC 1
HDAC 2
HDAC 3
HDAC 4
HDAC 6
HDAC 5
HDAC 7
HDAC 8
HDAC 9
Vincullin
LNCaP LAPC-4 CWR22-
Rv1
HDAC5
Co
ntr
ol
Co
ntr
ol
Co
ntr
ol
Co
ntr
ol
+1
µm
ol/
L T
as
Q
+1
µm
ol/
L T
as
Q
+1
µm
ol/
L T
as
Q
+1
µm
ol/
L T
as
Q
LNCaP LAPC4 CWR22R
Contr
ol
TasQ
Contr
ol
Tas
Q
Contr
ol
Tas
QC
ontr
ol
Tas
Q
HUVEC
β-Actin
β-Actin
β-Actin
β-Actin
β-Actin
β-Actin
HDAC1
HDAC2
HDAC3
HDAC4
HDAC6
HDAC8
HUVEC
HDAC7
Figure 2. A, Tas-Q (1 mmol/L) andTSA (200 nmol/L) prevent lysinedeacetylation in H3-histone atposition 9 and19 induced byhypoxia in all of the humanprostate cancer lines tested.b-actin was used as a loadingcontrol. B,Western blot analysis ofindicated HDAC in normal humanprostate epithelial (i.e., 957E/hTERT) cells and LNCaP, PC-3,and DU-145 human PCs. Vincullinwas used as a loading control.C–E, TasQ (1 mmol/L) does notdecrease HDAC isotypesexpressed by human PC lines (i.e.,LNCaP, LAPC-4, and CWR22-Rv1) or HUVECs. b-actin was usedas a loading control. In D, Westernblot analysis of HDAC5 wasdetected using SuperSignal WestFemto Maximum SensitivitySubstrate.
Isaacs et al.
Cancer Res; 73(4) February 15, 2013 Cancer Research1390
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
inhibits this decrease (Fig. 2A). Trichostatin A (TSA) a knownpan-class I/II HDAC inhibitor (27) likewise inhibits decreasein lysine 9/19 acetylated H3-histone (Fig. 2A). These resultssuggest TasQ might be an HDAC inhibitor.There are multiple isotypes of class I (HDAC1, 2, 3, and 8),
class IIa (HDAC4, 5, 7, and 9) and IIb HDACs (HDAC6; 27).Normal human prostate epithelial cells (957E/hTERT) ex-press all of the class I and IIb (HDAC6) isotypes, but HDAC4is the only class IIa isotype consistently detectable in these cells(Fig. 2B). The lack of detectable expression of the other class IIaisotypes by normal prostate epithelial cells is not due to thequality of the antibodies used for detection, as a positivecontrolHDAC5 is detected inCWR22-Rv1 cell (Fig. 2C), HDAC7in MCF-7 (data not shown), and HDAC9 is detected in LNCaP
cells (Fig. 2B), using the appropriate antibodies. Likewise,human prostate cancer lines characteristically express all ofthe class I isotypes and HDAC4, with most also expressingHDAC6, but only CWR22-Rv1 cells express HDAC5 and onlyLNCaP cells express HDAC9, respectively (Fig. 2B and C).HUVEC cells express HDAC2, 3, and 8 of the class I and HDAC4and 7of the class IIa and HDAC6 of the class IIb isotypes (Fig.2D and E). Interestingly, human prostate cancer lines express3- to 5-fold higher total HDAC activity than endothelial cells(Supplementary Fig. S1). At 1mmol/L, TasQ treatment does notlower the expression of any isotypes expressed by human PClines (Fig. 2C and D) or HUVECs (Fig. 2D and E). TasQ even at100 mmol/L does not inhibit total HDAC enzymatic activity inHUVECs or a series of human PC lines (LNCaP, CWR22-Rv1,
Figure 3. A, schematic organizationof HDAC4 protein. Numbers refer toamino acid position. Black boxes areknown sites of transcription factorbinding in N-terminal adapterdomain. S-P refers to serines, which ifphosphorylated, allow binding to14-3-3 protein. B, computer-baseddocking of TasQ to inactive (nonN-CoR binding) conformation ofregulatory ZRD within the catalyticdomain (amino acids 648–1,051) ofhuman HDAC4. Upper magentacolored ball is Zn2þ in the ZRD andlower magenta ball is Zn2þ in thecatalytic domain. C, interaction ofTasQ with various amino acids in theinactive conformation of ZRD. D,surfaceplasmon resonanceof humanfull-length HDAC4 binding to TasQimmobilized onto a Biacore chip.
A
D
NLS ZBD NES
HDAC domain
S
244 279
S S
648 751 1051 1084
P P P
802 950
Catalytic domain
HIF-1α binding
N-terminal adapter domain
N-CoR binding
14-3-3
-10
0
10
20
30
40
50
60
70
50 150 250 350 450
Time (s)
Resp
on
se (
RU
)
10
20
30
40
50
60
70
20 60
HDAC4 nmol/L
40
B
C
HDAC4 and Stress Signaling
www.aacrjournals.org Cancer Res; 73(4) February 15, 2013 1391
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
PC3, or PAC-MetUT1; Supplementary Fig. S1). This inabilitywas confirmed using recombinant human proteins of type I(HDAC1, 3, and 8), type IIa (HDAC4), and type IIb (HDAC6)isotypes in biochemical assays.
Tasquinimod binds allosterically within the regulatoryZinc domain of HDAC4
Previous studies document an enhanced level of HDAC4protein within PC nuclei in patients with CRPC and thatsuppression of HDAC4 expression causes in vitro growthinhibition of a variety of solid malignancies, including PC,particularly during hypoxic stress (28–30). Gene expressionarray data document that HDAC4 is overexpressed by morethan 3-fold (P < 0.05) in metastatic sites in patients withCRPC (n ¼ 18) compared with normal prostate from organdonors (n ¼ 23; Supplementary Fig. S2A). PC lines expressmore than 5-fold higher levels of HDAC4 protein thannormal prostate epithelial cells, the majority of which isin the nucleus (Supplementary Fig. S2B). HDAC4 is 1,084amino acids (AA) with a nuclear localization signal (NLS) atAA244-279, a nuclear export signal (NES) at AA1051-1084,and 3 serines at AA246, 467,and 632 whose phosphorylationis required for binding 14-3-3, which restricts localiza-tion to the cytoplasm (Fig. 3A; ref. 31). While HDAC4 lacksintrinsic DNA-binding activity, sequences within eitherthe N-terminal adaptor domain (AA1-648) or the C-terminalHDAC domain (AA648-1051) selectively bind a subset ofDNA anchoring transcription factors as part of eitherrepressive or stimulatory complexes at specific promoter/enhancers (31).
Within the HDAC domain, there is a zinc-bound "cata-lytic domain" involving AA802-950, however HDAC4 isenzymatically inactive against classic acetylated peptidesubstrates (14, 15). Also within the HDAC domain is aregulatory zinc-binding domain (ZRD; AA648-751; Fig.3A). Crystallography studies document that the ZRD has2 alternative conformations (17). When Zn2þ is coordinatedby C667, C669, H675, and C751, the ZRD is in an activeconformation allowing binding to the transcriptional core-pressor, N-CoR/HDAC3 complex via RD3 domain of N-CoR(14, 17). In the HDAC4/N-CoR/HDAC3 complex, HDAC3 isactive and deacetylates proteins tethered to the complex viabinding to HDAC4 (15). Alternatively, when H665, C667,H678, and C751 are coordinated with the structural Zn2þ,the ZRD is in an inactive conformation unable to bindN-CoR/HDAC3 (17).
Using published crystal structures (17), TasQ dockingwas conducted based upon energy miminization modeling.This modeling identified a specific allosteric binding sitefor TasQ within the inactive conformation of the ZRD,distinct from the catalytic domain (Fig. 3B). Critically,specific amino acid interactions involved in this bindingare only possible in inactive, not active, conformation of theZRD (Fig. 3C). TasQ was immobilized onto a Biacore chipvia an amino linker and binding of human recombinant full-length HDAC4 protein was determined using SPR. Thesestudies showed that TasQ binds HDAC4 with a Kd of 10 to30 nmol/L (Fig. 3D). Additional SPR studies determined that
TasQ binds to human recombinant full length HDAC4protein immobilized onto a Biacore chip in a 1:1 manner.SPR analysis documented that recombinant human HDAC4in which the C669 and H675 in the ZRD are mutated toalanine preventing formation of the active conformation(17) retains high affinity binding for TasQ (SupplementaryFig. S3A).
Tasquinimod phenocopies HDAC4 knockdown inendothelial and prostate cancer cells
A series of specific and nonspecific HDAC4 shRNA lenti-viral constructs were tested for their ability to downregulateHDAC4 in LNCaP cells (Supplementary Fig. S3B). The bestconstruct (shRNA2) was then used to transduce LNCaP andHUVEC cells. In multiple transductions, cells that grew outfollowing such shRNA2 transduction had only approximately60% HDAC4 knockdown (Fig. 4A). Even as little as a 60%reduction in HDAC4 completely inhibits HUVECs fromsprouting in 3D assays (Fig. 4B), which is an identicalresponse induced in wild type HUVECs by 1 mmol/LTasQ (Fig. 1D). With regard to LNCaP cells, such limitedHDAC4 reduction had little effect upon growth rate underthese standard nonstressed conditions (Fig. 4C). In contrast,however, under even the limited stress of exposure to 2% O2,without increasing CO2 and decreasing pH, 60% knockdownof HDAC4 decreases survival of LNCaPs, which is a responseidentical to treatment with 1 mmol/L TasQ under suchlimited stress (Fig. 4C, bottom).
TasQ also phenocopies the response to HDAC4 knock-down in vivo. Similar to the situation where PCs are non-tumorigenic when TasQ treatment is initiated at inocula-tion (Fig. 1B), HDAC4 shRNA2 knockdown LNCaPs are alsonontumorigenic in mice (Supplementary Fig. S3C and S3D).The specificity of this response was documented by thefact that restoration of HDAC4 expression in shRNA2 knockdown LNCaPs (Fig. 4D) restores their tumorigenicity (Fig. 4Eand Supplementary Fig. S3E).
Tasquinimod blocks the formation of HDAC4/N-CoR/HDAC3 complexes
SPR analysis documents that full-length HDAC4 proteinimmobilized onto a Biacore chip binds human recombinantfull length N-CoR with a Kd of 1 nmol/L and that TasQinhibits this binding in a dose-dependent manner (Fig. 5A).In co-IP/IB assays, HDAC4/N-CoR complex formation isZnþ2 dependent because it is blocked by EDTA (Fig. 5B).TasQ also blocks such complex formation with an IC50
value of less than 50 nmol/L (Fig. 5C). To evaluate thisblockade in a cell-based system, HEK-293T human embry-onic kidney cells were transfected with HDAC4. Co-IP/IBanalyses documented that TasQ inhibits N-CoR/HDAC3binding to HDAC4 with an IC50 < 1 mmol/L in normoxia(20% O2; Fig. 5D and E), and hypoxia (2% O2; Fig. 5E). Toevaluate whether these results are unique to HEK-293Ts,LNCaPs were transfected with HDAC3. Co-IP/IB analysesdocument that TasQ inhibits the binding of HDAC4 toN-CoR/HDAC3 complexes within PCs with an IC50 of lessthan 1 mmol/L (Fig. 5F).
Isaacs et al.
Cancer Res; 73(4) February 15, 2013 Cancer Research1392
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
Tasquinimodphenocopies loss ofHIF-1a transcriptionalstress response induced by HDAC4 knockdownTasQ should inhibit HDAC4 client protein deacetylation
via blocking colocalization of N-CoR associated HDAC3 dea-cetylase activity. One client protein is HIF-1a, which bindsvia its ID domain (AA603-785) to the catalytic domain (AA802-950) of HDAC4 (Fig. 3A; refs. 30, 32–34). Because of cessation ofhydroxylation of a series of prolines and an asparagine, Hif-1aaccumulates in the nucleus during hypoxia, where it is acet-ylated by p300/CBP associated factor (PCAF) acetyltransferaseon a series of lysines between AA10–21 and on AA674(30, 35, 36). The N-terminal acetylated lysines are within DNAbinding domain of HIF-1a (37) and are deacetylated by anHDAC4 dependent mechanism (30). Such HDAC4 binding tothe ID of HIF-1a also competitively inhibits binding of factorinhibiting HIF-1a and instead facilitates binding of p300HATand dimerization with HIF-1b to form transcriptionally activeHIF-1 at HREs within promoter/enhancers of hypoxic respon-sive genes (19, 34). shRNA2 knockdown of HDAC4 preventsdeacetylation of N-terminal lysines of HIF-1a enhancing its
degradation, inhibiting its transcriptional activity, anddecreasing cell survival under stressful conditions (30, 33).
To confirm these observations, HIF-1a was comparedbetween control, HDAC4 shRNA2 knockdown, and knock-down-restored LNCaPs. As expected, HDAC4 knockdowndecreases HIF-1a and this is reversed when HDAC4 isrestored (Fig. 4D). If TasQ inhibits HDAC4/N-CoR/HDAC3-dependent N-terminal lysine deacetylation of HIF-1a, then it should decrease HIF-1a that suppresses HIF-1-dependent transcription. When HEK-293T cells were trans-fected with HDAC4, HIF-1a is decreased by TasQ (Fig. 6A).To determine if this is due to inhibition of HIF-1a/HDAC4complex formation, LNCaPs transfected with HDAC4 weretreated with TasQ under both normoxic (21% O2) andhypoxic (2% O2) conditions. IB of total cell extracts showedthat hypoxia increases HIF-1a coupled with decreasing its(Fig. 6B). Decreased HIF-1a mobility is due to phosphory-lation that enhances its transcriptional activity (38). Co-IP/IB of nuclear extracts documents that 1 mmol/L TasQdecreases HIF-1a/HDAC4 complex formation by more than
Figure 4. A, HDAC4 protein inHDAC4shRNA2knockdownLNCaPand HUVEC endothelial cells.Number under lane is relative level ofexpression per cell normalized tountreated control cells. B, HDAC4shRNA knockdown inhibitssprouting ofHUVECs in 3Dassay. C,top, growth of control versusHDAC4 shRNA2 knockdown versus1 mmol/L TasQ-treated LNCaPgrowth under normoxic conditions;bottom, under hypoxic conditions.D, HDAC4 and HIF-1a proteinexpression in control versus HDAC4shRNA2 knockdown versusknockdown-restored LNCaPs.b-actin was used as a loadingcontrol. E, tumorigenicity of HDAC4shRNA2 knockdown versusrestored LNCaPs when xenograftedinto male nude mice (n ¼ 8/group).
BA
C
0.000
0.020
0.040
0.060
0.080
0.100
0.120
0.140
Control
shRNA
Tasquinimod
Control
shRNA
Tasquinimod
0 2 4 6Days
8
Tu
mo
r v
olu
me
(c
c)
LnCaP-HDAC 4
knockdown
LnCaP-HDAC 4
knockdown restored
D
E
IB: HDAC4
Normoxic conditions
Ce
ll n
um
be
rC
ell
nu
mb
er
Hypoxic conditions
IB: HIF-1ββ
LNCaP
1.00
350,000
300,000
250,000
200,000
150,000
100,000
50,000
0
350,000
300,000
250,000
200,000
150,000
100,000
50,000
0
0.44 0.48 1.00
LN
Ca
P
LN
Ca
P H
DA
C4
Kn
oc
kd
ow
n
IB: HDAC4
Hu
ve
c
Hu
vec H
DA
C4
Kn
oc
kd
ow
n
IB: β-Actin
IB: β-Actin
0 2 4 6Days
8
HDAC4 and Stress Signaling
www.aacrjournals.org Cancer Res; 73(4) February 15, 2013 1393
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
50% under both normoxic and hypoxic conditions (Fig. 6B).The inhibitory effect of TasQ is not dependent upon overexpression of HDAC4, because it suppresses HIF-1a by morethan 50% under hypoxic conditions in nuclear extracts fromnontransfected LNCaPs (Fig. 6C). Nuclear extracts fromnontransfected LNCaPs were co-IP with an antiacetyl lysineantibody and the IP analyzed by IB with anti-HIF-1a anti-body. Even though TasQ decreases nuclear HIF-1a by morethan 2-fold in the LNCaPs (Fig. 6C), acetylated HIF-1aincreases by 40% (Fig. 6D). Thus, TasQ inhibits the nuclearfraction of N-terminal HIF-1a lysines deacetylated underhypoxic conditions by more than 3-fold.
On the basis of its inhibition of HIF-1a deacetylation,TasQ should inhibit HIF-1a-dependent transcription.TasQ's dose–response inhibition of the adaptive stressresponse of PC-3 human prostate cancer cells that have
stably integrated a HIF-1a-dependent HRE-driven GFP con-struct was evaluated to test this prediction. Like LNCaP,PC-3 predominantly has nuclear HDAC4 (SupplementaryFig. S2B) and HIF-1a even under normoxic conditions.In both conditions, PC-3s sense "stress" as documented bytheir expression of GFP. While this stress response is muchgreater (i.e., >100-fold increase in GFP production) underhypoxic conditions, TasQ inhibits GFP expression underboth conditions (Fig. 6E).
To further evaluate the generality of TasQ's ability toinhibit HIF-1a-driven survival signaling in prostate cancercells, lactate dehydrogenase-A (LDH-A) protein expressionwas determined in LNCaP cancer cells exposed to 1 mmol/LTasQ under both normoxic and hypoxic conditions. Theseresults document that expression of LDH-A, a knownHIF-1a transcriptional target (19), is increased by hypoxia
E
C D
A
IB:
HDAC4
1µmol/L TasQ
Total
Extract
IP: Flag
-HDAC3
__ + _ +
_ +
IB:
Flag-
HDAC3
B
HDAC4
40
80
120
160
Resp
on
se (
RU
)
NCoR/No Comp
NCoR/1 µmol/L TasQ
NCoR/10 µmol/L TasQ
NCoR/100 µmol/L TasQ
F
1 mmol/L EDTA
50 n
mol/L
Tas
Q10
0 nm
ol/L
Tas
Q
Contr
ol
1 µm
ol/L
Tas
Q10
µm
ol/L
Tas
Q25
µm
ol/L
Tas
Q
Contr
ol
-ve
contr
ol
1 µm
ol/L
Tas
Q10
µm
ol/L
Tas
Q
Contr
ol
1 µm
ol/L
Tas
Q10
µm
ol/L
Tas
Q
Contr
ol
-ve
contr
ol
IB:GST-HDAC4
IP: Flag-NCoR
IB:Flag-NCoR
IP: GST-HDAC4
IB:
Flag-NCoR
IB:
GST-HDAC4
IB:
HDAC3
IB:
NCoR
IP: Flag-HDAC4
IB:
Flag-
HDAC4
IB:
HDAC3
IP: Flag-HDAC4
Normoxic Hypoxic
IB:
Flag-
HDAC4
Figure 5. A, SPR determinedTasQ dose–response inhibition ofN-CoR binding to HDAC4immobilized onto a Biacore chip.B, Zn2þ-dependent binding of full-length N-CoR protein binding torecombinant full-length HDAC4protein is prevented by chelationwith EDTA. Flag-NCoR was usedas loading control. C, binding ofrecombinant full-length HDAC4binding to full-length N-CoRprotein in the presence of Zn2þ isinhibited by TasQ. GST-HDAC4was used as loading control. TasQinhibits HDAC4 binding to N-CoR/HDAC3 complexes in HEK-293cells in both normoxic (D and E)and hypoxic (E) conditions. Flag-HDAC4 was used as loadingcontrol. F, TasQ inhibits HDAC4binding to N-CoR/HDAC3complexes in LNCaPs. Flag-HDAC3 was used as loadingcontrol.
Isaacs et al.
Cancer Res; 73(4) February 15, 2013 Cancer Research1394
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
and that TasQ inhibits such upregulation (SupplementaryFig. S4A).To evaluate whether TasQ's inhibition of the HIF-1a
transcriptional response also occurs in endothelial cells,HUVECs were cotransfected with a HIF-1a-dependentHRE-driven firefly luciferase vector plus a cytomegalovi-rus-driven renilla luciferase control vector and thecells exposed to normoxia versus hypoxia. These resultsshow that hypoxia induced HIF-1a-dependent transcrip-tion in HUVECs is completely inhibited by 1 mmol/L
TasQ (Supplementary Fig. S4B). Likewise, a commercialHIF-1a-dependent hypoxia cDNA microarray documentedthat 1 mmol/L TasQ prevents HUVECs from adaptivelyupregulating transcription of stress genes under hypoxia(Fig. 6F).
Tasquinimod's efficacy against solid malignancies andits enhancement by a combinational approach
The previous results identify HDAC4 as a molecular targetfor TasQ's suppression of epigenetic signaling required for
A B
Co
ntr
ol
1 µ
mo
l/L
TasQ
IP:
IgG
IP:Acetyl-lysine
IB:
HIF-1α1.0 1.4
C
F
D E
IB: GFP
Normoxic
25,000 cells
Hypoxic
1,250 cells
IB:
β-actin
Total extract Normoxic Hypoxic
Nuclear extract
IP: Flag-HDAC4
Flag-HDAC4
Tasquinimod
IB:HIF-1α
+–
– – +
+
IB:β-Actin
IB:HIF-1α
IB:β-Actin
IB:HIF-1α
1 µ
mo
l/L
T
as
Q
10
µm
ol/
L
Ta
sQ
Co
ntr
ol
IB:
H3-histone
IB:H3-histone
0
0.5
1
1.5
2
2.5
3Normoxic
Hypoxic
Norm
oxic
Hyp
oxic
Contr
ol
1 µm
ol/L T
asQ
Contr
ol
1 µm
ol/L T
asQ
10 µ
mol/L
Tas
Q
Contr
ol
1 µm
ol/L T
asQ
Hypoxic + 1 umol/L TasQ
Ex
pre
ss
ion
re
lati
ve
to
β-a
cti
n
Figure 6. A, HDAC4 increases and 1 mmol/L TasQ decreases HIF-1a protein in HEK-293T cells. b-actin was used as a loading control. B, TasQ lowers level ofHDAC4 in the nuclei of LNCaPs in both normoxia and hypoxia. Upper arrow denotes phosphorylated form. b-actin was used as a loading control for totalextracts, whereas H3-histone was used as loading control for nuclear extract. C, TasQ lowers endogenous HDAC4 in the nuclei of LNCaPs in hypoxia. H3-histone was used as loading control. D, TasQ increases the proportion of HIF-1a that is lysine acetylated within the nuclei of LNCaPs under hypoxia.Number under lane is the relative level of expression per cell normalized to untreated control cells. b-actin was used as a loading control. E, TasQ inhibits HIF-1a–dependent GFP expression in PC-3 human prostate cancer cells that have stably integrated a HRE/GFP construct in normoxia (signal from 25,000 cells)versus 24 hours of hypoxia (sign from 1,250 cells). F, TasQ inhibits hypoxia-induced HIF-1a–dependent transcriptional stress response of HUVECsdetected using a hypoxia cDNAmicroarray with the results normalized to b-actin. Abbreviations for the genes probed are as follows: CA9, carbonic anhydraseIX; CTSD, cathepsin D; Glut-1, glucose transporter-1; Glut-3, glucose transporter-3; HIF-1, hypoxia-induced factor-1a; HIF-2, hypoxia-induced factor-2a;MMP2, matrix metalloproteinase 2; MXi-1, Max interactor-1; NDRG, N-myc downstream regulated-1; NDRG2, N-myc downstream regulated-2; and PAI-1,plasminogen activator inhibitor-1.
HDAC4 and Stress Signaling
www.aacrjournals.org Cancer Res; 73(4) February 15, 2013 1395
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
survival in a stressful microenvironment by both cancer andendothelial cells. Such suppression is via TasQ's allostericbinding to the ZRD of HDAC4 locking it in an inactive con-
formation in both cancer and endothelial cells. TasQ's alloste-ric binding prevents deacetylase activity of N-CoR/HDAC3from colocalizing with HDAC4/client transcription factors
Human cancer
xenografts
Percent
inhibition of
cancer growth
by tasquinimod
LAPC-4 (prostate) 78 ± 11
LNCaP (prostate) 57 ± 8
CWR22-RH (prostate) 84 ± 12
MCF-7 (mammary) 52 ± 7
TSU (bladder) 59 ± 10
HCT-116 (colon) 50 ± 5
B
A
0.0
0.5
1.0
1.5
2.0
2.5
35 40 45 50 55 60
Rela
tive t
um
or
siz
e
Days
Control G202
TasQ Combo
C
Cytoplasm
nucleus
Basal generepression
Survival genetranscription
Transcriptionfactor
Figure 7. A, overview of TasQ's mechanism of action. Under hypoxia/stressful microenvironment conditions, HDAC4/N-CoR/HDAC3 complex binds viaHDAC4 to DNA-bound transcription factors, allowing HDAC3 to deacetylate histones locally repressing basal transcription within the nucleus ofendothelial and cancer cells. Also during such stress, HIF-1a accumulates in the nuclei where PCAF acetylates its N-terminal lysines betweenAA11–21 and atposition 674 in its inhibitory domain (ID). HDAC4/N-CoR/HDAC3complexesbind viaHDAC4 to the acetylated lysine 674 in the IDofHIF-1 colocalizingN-CoR/HDAC3, resulting in deacetylation of N-terminal lysines between AA11–21 of HIF-1, facilitating binding of p300HAT andHIF-1b needed for formation of HIF-1complex, which via p300 acetylates histones in cell survival/angiogenesis genes, stimulating their transcription. TasQ binding allosterically locks ZRD ofHDAC4 in an inactive conformation, preventing basal gene repression and survival/angiogenesis gene transcription needed for the angiogenic switch. B,TasQ's efficacy against a diverse series of human solid cancer xenografts. Results are expressed as percent inhibition of cancer growth based uponcomparison of tumor volume (N¼5–10 cancer-bearingmice per cancer subtype) in TasQ-treated (10mg/kg/d via drinkingwater) versus untreatedmice over amonth observation period. C, therapeutic response of established (0.8 cc) MCF-7 human breast cancers growing in mice given 2 daily intravenousinjections at 56mg/kg of a tumor endothelial targeted cytolytic prodrug (G202) alone and in combinationwith 10mg/kg/d oral TasQ (N¼ 8/group). Results arepresented as relative tumor size normalized to tumor volume at initiation of treatment. P < 0.05 for combination (combo) group versus eithermonotherapies after day 49.
Isaacs et al.
Cancer Res; 73(4) February 15, 2013 Cancer Research1396
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
bound at the promoter/enhancers of genes (Fig. 7A). HDAC4binding with the majority of its client DNA bound-transcrip-tion factors results in decreased histone acetylation due tocolocalization of HDAC3 repressing their transcription (31;(Fig. 7A). This explains how TasQ inhibits hypoxia stimulatedglobal histone deacetylation (Fig. 1D). TasQ binding to HDAC4also prevents HIF-1a/HDAC4/N-CoR/HDAC3 complex forma-tion. This is critical because the stability and thus the tran-scriptional activity of HIF-1a needed for both cancer cellssurvival and endothelial cells angiogenesis is inhibitedwhen itsN-terminal lysines are acetylated and HDAC4 is required fortheir full deacetylation (Fig. 7A).On the basis of this HDAC4 allosteric binding, TasQ should
be effective when used as monotherapy against multipletypes of solid malignancies because ablating HDAC4 signal-ing not only inhibits endothelial tumor angiogenesis, but alsogrowth/survival of a variety of human organ-specific cancers(29, 30, 39, 40). To evaluate this possibility, a series of humanprostate, breast, bladder, and colon cancer xenografts wereinoculated into appropriate immune-suppressed hosts andcancers allowed to reach 0.250 cc before daily oral TasQ(10 mg/kg/d) was initiated. As predicted, TasQ mono-therapy inhibited (P < 0.05) subsequent tumor growth bya least 50% in all of these cancer types (Fig. 7B).Because TasQ's potency and efficacy for inhibition of
cancer cell survival and endothelial angiogenic response areenhanced in a stressful tumor microenvironment, combin-ing TasQ with agents that increase tumor microenviron-mental stress should enhance therapeutic efficacy. To testthis prediction, TasQ was combined with G202, which is aThapsigargin prodrug engineered to be hydrolyzed to acytotoxin by a protease uniquely expressed by cancer asso-ciated endothelial cells to selectively kill tumor endothelialcells enhancing a more stressful tumor microenvironment(12). When estrogenized female nude mice are inoculatedwith MCF-7 human breast cancer cells and cancers allowedto grow to a starting size of approximately 0.2 cc beforeanimals are given a single cycle of 2 daily IV injections of 56mg/kg of G202, cancers regress due to the death of tumorassociated endothelial and cancer cells and in a subset ofanimals, the cancers does not regrow (12). If such breast,cancers are allowed to grow to a starting size of approxi-mately 0.8 cc before initiating a single 2-day cycle with G202,cancers regress by more than 50% followed by regrowth (Fig.7C). While daily TasQ treatment (10 mg/kg/d) initiated whenthe breast cancers are approximately 0.8 cc inhibits subse-quent tumor growth, no regression of these large cancersoccurs (Fig. 7C). In contrast, combining daily TasQ withG202 treatment results in regression of these large cancers,but now regrowth is inhibited (P < 0.05; Fig. 7C).
DiscussionSolid malignancies outgrow their blood supply producing a
hypoxic tumor microenvironment. Such tumor hypoxiaenhances glycolysis and amino-acid catabolism increasinglactic acid secretion producing an acidic extracelluar micro-environment coupled with nutrient limitation (e.g., glucoseand amino-acid deficiency; ref. 8). To survive this stressful
microenvironment, an epigenetic adaptive survival response,termed the angiogenic switch, is activated involving reciprocalinteractions between cancer cells and tumor host infiltratingcells including MDSC, macrophages, and endothelial cells,(7–9). This angiogenic switch facilitates global repression oftranscription coupled with selective expression of specificstress–response genes via type IIa HDAC4 dependent coloca-lization of type I HDAC3 deacetylase activity at specific pro-moter/enhancers in cancer and host tumor infiltrating cells(refs. 30, 33; Fig. 7A). The importance of such deacetylaseactivity is documented by ability of pan- and type 1 isotypeselective HDAC inhibitors to suppress cancer growth viareduced tumor angiogenesis (30, 33). Limitation with usingpan- or type 1 isotype selective HDAC inhibitors, however, isthat they upregulate, instead of repress, transcription of asubset of unwanted survival genes [e.g., GRP78/BiP, Cox-2,fibroblast growth factor (FGF2), and Slit2; refs. 41–46)]. Thus,selective inhibition of HDAC4 signaling is a more optimalapproach for blocking the angiogenic switch without activat-ing unwanted survival signaling within cancer sites.
The present studies identify TasQ as such a high affinityHDAC4 selective negative allosteric modulator. Besides allo-sterically binding HDAC4, TasQ also binds with high affinityto the calcium binding protein, S100A9 (7, 16). In the presenceof zinc, S100A9 undergoes a conformational change becominga ligand for the proinflammatory Toll-like receptor 4 (TLR4)and TasQ binding to S100A9 inhibits this binding to TLR4 (7).This is significant because MDSCs from the bone marrowcirculate in the blood and express S100A9 protein and TLR4and S100A9 binding to TLR4 stimulates tumor infiltration ofMDSCs (7, 47, 48). Under hypoxic conditions, tumor infiltrat-ing DSC differentiate via HIF-1a-dependent transcriptioninto tumor-associated macrophages (TAM; ref. 49). Once dif-ferentiated, TAMs secrete angiogenic factors such as VEGF,FGF, TNF-a, and TGF-b (50). Thus, TasQ's can suppress tumorangiogenesis due to inhibition of S100A9/TLR4-dependentMDSCs tumor infiltration and/or to inhibition of HDAC4/N-CoR/HDAC3-dependent deacetylation of HIF-1a by MDSCssuppressing their differentiation into TAMs. Presently, therelative importance of both of these complementary TasQtargets for anticancer efficacy is being evaluated in preclinicalmodels. Regardless of the results of these additional studies,the present studies document that TasQ's therapeutic anti-cancer efficacy is enhanced when combined with an additionaldrug (e.g., G202), which produces a more stressful tumormicroenvironment.
Disclosure of Potential Conflict of InterestNo potential conflicts of interest were disclosed.
Authors' ContributionsConception and design: J.T. Isaacs, A. Bj€ork, T. LeandersonDevelopment of methodology: J.T. Isaacs, M.D. Wissing, S.K. Kachhap, L. XingAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): J.T. Isaacs, L. Antony, S. Dalrymple, S. Gerber,H. Hammers, J. Luo, P. Bj€orkAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): J.T. Isaacs, H. Hammers, J. Luo, L. Xing, P. Bj€ork,A. Olsson, T. LeandersonWriting, review, and/or revision of the manuscript: J.T. Isaacs, W.N.Brennen, M.D. Wissing, S.K. Kachhap, L. Xing, P. Bj€ork, A. Olsson, T. Leanderson
HDAC4 and Stress Signaling
www.aacrjournals.org Cancer Res; 73(4) February 15, 2013 1397
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
Administrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): J.T. Isaacs, L. Antony, S. Dalrymple,W.N. Brennen, M.D. Wissing, S.K. Kachhap, A. OlssonStudy supervision: J.T. Isaacs
AcknowledgmentsThe authors thank Georges H. Ndikuyeze for expert technical assistance in
conducting the in vivo experiments. The authors thank Drs. Gregg Semenza,Venu Raman, and Robert Getzenberg (Johns Hopkins School of Medicine,Baltimore, MD), Dr. Arthur Zelent (Institute of Cancer Research, Sutton, Surrey,United Kingdom), and Dr. Eric Verdin (Gladstone Institute of Virology andImmunology, San Francisco, CA) for their generosity in supplying reagents and
tissue samples. The authors thank Dr. Craig Dionne (GenSpera, San Antonio, TX)for supplying G202.
Grant SupportThis study was partially supported by NIH grants P50CA058236 and
P30CA006973 and a sponsored research agreement between The Johns HopkinsUniversity School of Medicine (J.T. Isaacs, PI) and Active Biotech Research AB.
The costs of publication of this article were defrayed in part by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received July 10, 2012; revised September 27, 2012; accepted October 15, 2012;published OnlineFirst November 13, 2012.
References1. Pili R, HaggmanM, Stadler W, Gingrich JR, Assikis VJ, Bjork A, et al. A
randomized multicenter international phase II study of tasquinimod inchemotherapy-naive patients with metastatic castration-resistantprostate cancer. J Clin Oncol 2010;28:75.
2. Isaacs JT. The long and winding road for the development of tasqui-nimod as an oral second-generation quinoline-3-carboxamide anti-angiogenic drug for the treatment of prostate cancer. Expert OpinInvestig Drugs 2010;19:1235–43.
3. Isaacs JT, Pili R, Qian DZ, Dalrymple SL, Garrison JB, Kyprianou N,et al. Identification of ABR-215050 as lead second generation quin-oline-3-carboxamide anti-angiogenic agent for the treatment of pros-tate cancer. Prostate 2006;66:1768–78.
4. Dalrymple SL, Becker RE, Isaacs JT. The quinoline-3-carboxamideanti-angiogenic agent, tasquinimod, enhances the anti-prostate can-cer efficacy of androgen ablation and taxotere without effecting serumPSA directly in human xenografts. Prostate 2007;67:790–7.
5. Olsson A, Bjork A, Vallon-Christersson J, Isaacs JT, Leanderson T.Tasquinimod (ABR-215050), a quinoline-3-carboxamide anti-angio-genic agent,modulates the expression of thrombospondin-1 in humanprostate tumors. Mol Cancer 2010;9:107.
6. Dalrymple SL, Becker RE, Zhou H, DeWeese TL, Isaacs JT. Tasqui-nimod prevents the angiogenic rebound induced by fractionatedradiation resulting in an enhanced therapeutic response of prostatecancer xenografts. Prostate 2012;72:638–48.
7. Kallberg E, Vogl T, Liberg D, Olsson A, Bjork P, Wilkstrom P, et al.S100A9 interaction with TLR4 promotes tumor growth. PLoS One2012;7:e34207.
8. Lunt SJ, Chaudary N, Hill RP. The tumor microenvironment andmetastatic disease. Clin Exp Metastasis 2009;26:19–34.
9. Hanahan D, Folkman J. Patterns and emerging mechanisms of theangiogenic switch during tumorigenesis. Cell 1996;86:353–64.
10. Jennbacken K, Welen K, Olsson A, Axelsson B, Torngren M, DamberJE, et al. Inhibition of metastasis in a castration resistant prostatecancer model by quinoline-3-carboxamide tasquinimod (ABR-215050). Prostate 2012;72:913–24.
11. Nakatsu MN, Sainson RC, Aoto JN, AikenheadM, Perez-del-Pulgar S,Carpenter PM, et al. Angiogenic sprouting and capillary lumen forma-tion modeled by human umbilical vein endothetial cells (HUVEC) infibrin gels: the role of fibroblasts and Angiopoietin-1. Microvasc Res2003;66:102–12.
12. Denmeade SR, Mhaka AM, Rosen DM, Brennen WN, Dalrymple S,Dach I, et al. Engineering a prostate-specific membrane antigen–activated tumor endothelial cell prodrug for cancer therapy. Sci TransMed 2012;4:140–54.
13. Litvinov IV, VanderGriendDJ, Xu Y, Antony L, Dalrymple SL, Isaacs JT.Low-calcium serum-free definedmedium selects for growth of normalprostatic epithelial stem cells. Cancer Res 2006;66:8598–607.
14. Fischle W, Dequiedt F, Hendzel MJ, Guenther MG, Lazar MA, VoelterW, et al. Enzymatic activity associated with class II HDACs is depen-dent on amultiprotein complex containing HDAC3 and SMRT/N-CoR.Mol Cell 2002;9:45–57
15. Lahm A, Paolini C, Pallaoro M, Jones P, Neddermann P, SambuciniS, et al. Unraveling the hidden catalytic activity of vertebrate classIIa histone deacetylases. Proc Natl Acad Sci USA 2007;104:17335–40.
16. Bjork P, Bjork A, Vogl T, Stenstrom M, Liberg D, Olsson A, et al.Identification of human S100A9 as a novel target for treatment ofautoimmune disease via binding to quinoline-3-carboxamides. PLoSBiol 2009;7:e97.
17. BottomleyMJ, Surdo PL, Di Giovine P, Cirillo A, Scarpelli R, Ferrigno F,et al. Structural and functional analysis of the human HDAC4 catalyticdomain reveals a regulatory structural zinc-binding domain. J BiolChem 2008;283:26694–704.
18. Raman V, Artemov D, Pathak AP, Winnard PT, McNutt S, Yudina A,et al. Characterizing vascular parameters in hypoxic regions:a combination magnetic resonance and optical imaging studyof a human prostate cancer model. Cancer Res 2006;66:9929–36.
19. Semenza GL, Jiang BH, Leung SW, Passantino R, Concordet JP,Maire P, et al. Hypoxia response elements in the aldolase A, enolase1, and lactate dehydrogenase A gene promoters contain essentialbinding sites for hypoxia-inducible factor 1. J Biol Chem 1996;271:32529–37.
20. Prakash K, Pirozzi G, Elashoff M, Munger W, Waga I, Dhir R, et al.Symptomatic and asymptomatic benign prostatic hyperplasia: molec-ular differentiation by using microarrays. Proc Natl Acad Sci U S A2002;99:7598–603.
21. Liu W, Laitinen S, Vihinen M, Kowalski J, Yu G, Chen L, et al. Copynumber analysis indicates monoclonal origin of lethal metastaticprostate cancer. Nat Med 2009;15:559–65.
22. Dunn TA, Chen S, Faith DA, Hicks JL, Platz EA, Chen Y, et al. A novelrole of myosin VI in human prostate cancer. Am J Pathol 2006;169:1843–54.
23. Denko N, Wernke-Dollries K, Johnson AB, Hammond E, Chiang CM,Barton MC. Hypoxia actively represses transcription by inducingnegative cofactor 2 (Dr1/DrAP1) and blocking preinitiation complexassembly. J Biol Chem 2003;278:5744–9.
24. Manalo D, Rowan A, Lavoie T, Natarajan L, Kelly BD, Ye SQ, et al.Transcriptional regulation of vascular endothelial cell responses tohypoxia by HIF-1. Blood 2005;105:659–69.
25. Johnson AB, Denko N, Barton MC. Hypoxia induces a novel signatureof chromatin modifications and global repression of transcription. MutRes 2008;640:174–9.
26. Li Q, Costa M. c-Myc mediates a hypoxia-induced decrease in acet-ylated histone H4. Biochimie 2009;91:1307–10.
27. Haberland M, Montgomery RL, Olson ER. The many roles of histonedeacetylases in development and physiology: implications for diseaseand therapy. Nat Rev/Genet 2009;10:32–42.
28. Halkidou K, Cook S, Leung HY, Neal DE, Robson CN. Nuclear accu-mulation of histone deacetylase 4 (HDAC4) coincides with the loss ofandrogen sensitivity in hormone refractory cancer of the prostate. EurUrol 2004;45:382–9.
29. Cadot B, Brunetti M, Coppari S, Fedeli S, de Rinaldis E, Dello Russo C,et al. Loss of histone deacetylase 4 causes segregation defects duringmitosis of p53-deficient human tumor cells. Cancer Res 2009;69:6074–82.
30. Geng H, Harvey CT, Pittsenbarger J, Liu Q, Beer TM, Xue C, et al.HDAC4 protein regulates HIF-1a protein lysine acetylationand cancer cell response to hypoxia. J Biol Chem 2011;286:38095–102.
Isaacs et al.
Cancer Res; 73(4) February 15, 2013 Cancer Research1398
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
31. Martin M, Kettmann R, Dequiedt F. Class IIa histone deacetylases:regulating the regulators. Oncogene 2007;26:5450–67.
32. KatoH, Tamamizu-KatoS, Shibasaki F. HistoneDeacetylase 7 associ-ates with hypoxia-induced factor-1a and increases transcriptionalactivity. J Biol Chem 2004;279:41966–74.
33. Ellis L, Hammers H, Pili R. Targeting tumor angiogenesis with histonedeacetylase inhibitors. Cancer Lett 2009;280:145–53.
34. Seo HW, Kim EJ, Na H, Lee MO. Transcriptional activation ofhypoxia-inducible factor-1 alpha by HDAC4 and HDAC5 involvesdifferential recruitment of p300 and FIH-1. FEBS Lett 2009;583:55–60.
35. Xenaki G, Ontikatze T, Rajendran R, Stratford IJ, Dive C, Krstic-Demonacos M, et al. PCAF is an HIF-1a cofactor that regulates p53transcriptional activity in hypoxia. Oncogene 2008;27:5785–96.
36. Lim JH, Lee YM, Chun YS, Chen J, Kim JE, Park JW. Sirtuin 1modulates cellular responses to hypoxia by deacetylating hypoxia-induced factor 1a. Mol Cell 2010;38:864–78.
37. Michel G, Minet E, Mottet D, Remacle J, Michiels C. Site-directedmutagenesis of the hypoxia-induced factor-1a DNA-binding domain.Biochim Biophys Acta 2002;1578:73–83.
38. Richard DE, Berra E, Gothie E, Roux D, Pouyssegur J. p42/p44Mitogen-activated protein kinases phosphorylate hypoxia-induciblefactor 1a (HIF-1a) and enhance the transcriptional activity of HIF-1a.J Biol Chem 1999;274:32631–7.
39. Liu R, Wang L, Chen G, Katoh H, Chen C, Liu Y, et al. FOXP3 up-regulates p21 expression by site-specific inhibition of histone deace-tylase 2/histone deacetylatase 4 association to the locus. Cancer Res2009;69:2252–9.
40. Mottet D, Pirotte S, Lamour V, HagedornM, Javerzat S, Bikfalvi A, et al.HDAC4 represses p21WAF1/Cip1 expression in human cancer cellsthrough a Sp1-dependent, p53-independent mechanism. Oncogene2009;28:243–56.
41. Baumeister P, Dong D, Fu Y, Lee AS. Transcriptional induction ofGRP78/BiPbyhistone deacetylase inhibitors and resistance to histone
deacetylase inhibitor-induced apoptosis. Mol Cancer Ther 2009;8:1086–94.
42. Halili M, Andrews MR, Labzin LI, Schroder K, Matthias G, Cao C, et al.Differential effects of selectiveHDAC inhibitors onmacrophage inflam-matory responses to the Toll-like receptor 4 agonist LPS. J LeukocBiol2010;87:1103–14.
43. Wang X, Li G, Wang A, Zan Z, Merchan JR, Halmos B. Combinedhistone deacetylase and cyclooxigenase inhibition achieves enhancedantiangiogenic effects in lung cancer cells. Mol Carcinogenesis. 2012.[Epub ahead of print].
44. Urbich C, Rossig L, Kaluza D, Potent M, Boeckel JN, Knau A, et al.Hdac5 is a repressor of angiogenesis and determines the angiogenicgene expression pattern of endothelial cells. Blood2009;113:5669–79.
45. Stronach EA, Alfraidi A, Rama N, Datler C, Studd JB, Agarwal R, et al.HDAC4-regulated STAT1 activation mediates platinum resistance inovarian cancer. Cancer Res 2011;71:4412–22.
46. Chu F, Chou P, Mirkin BL, Mousa SA, Rebbaa A. Cellular conditioningwith trichostatin A enhances the anti-stress response through up-regulation of HDAC4 and down-regulation of the IGF/Akt pathway.Aging Cell 2008;7:516–25.
47. Zhao F, Hoechst B, Duffy A, Gamrekelashvili J, Fiorvanti S,MannsMP,et al. S100A9 a new marker for monocytic human myeloid-derivedsuppressor cells. Immunol 2012;136:176–83.
48. Bunt SK, Clements VK, Hanson EM, Sinha P, Ostrand-Rosenberg S.Inflammation enhances myeloid-derived suppressor cell cross-talkby signaling through Toll-like receptor 4. J Leukoc Biol 2009;85:996–1004.
49. Corzo CA, Condamine T, Lu L, Cotter MJ, Youn JI, Cheng P, et al. HIF-1a regulates function and differentiation of myeloid-derived sup-pressor cells in the tumor microenvironment. J Exp Med 2010;207;2439–53.
50. Squadrito ML, De Palma M. Macrophage regulation of tumor angio-genesis: implications for cancer therapy. Mol Aspects Med 2011;32:123–45.
HDAC4 and Stress Signaling
www.aacrjournals.org Cancer Res; 73(4) February 15, 2013 1399
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730
2013;73:1386-1399. Published OnlineFirst November 13, 2012.Cancer Res John T. Isaacs, Lizamma Antony, Susan L. Dalrymple, et al. Signaling within the Compromised Cancer MicroenvironmentTasquinimod Is an Allosteric Modulator of HDAC4 Survival
Updated version
10.1158/0008-5472.CAN-12-2730doi:
Access the most recent version of this article at:
Material
Supplementary
http://cancerres.aacrjournals.org/content/suppl/2012/11/12/0008-5472.CAN-12-2730.DC1
Access the most recent supplemental material at:
Cited articles
http://cancerres.aacrjournals.org/content/73/4/1386.full#ref-list-1
This article cites 49 articles, 17 of which you can access for free at:
Citing articles
http://cancerres.aacrjournals.org/content/73/4/1386.full#related-urls
This article has been cited by 7 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
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://cancerres.aacrjournals.org/content/73/4/1386To request permission to re-use all or part of this article, use this link
on April 26, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Published OnlineFirst November 13, 2012; DOI: 10.1158/0008-5472.CAN-12-2730