Dual Inhibition of Canonical and Noncanonical NF-kB...

Cancer Therapy: Preclinical

Dual Inhibition of Canonical and Noncanonical NF-kBPathways Demonstrates Significant AntitumorActivities in Multiple Myeloma

Claire Fabre1,2, NaoyaMimura1,2, KathrynBobb3, Sun-YoungKong1,5, G€ull€uGorgun1,2, DianaCirstea1,2, YiguoHu1,2, Jiro Minami1,2, Hiroto Ohguchi1,2, Jie Zhang4, Jeffrey Meshulam3, Ruben D. Carrasco2,Yu-Tzu Tai1,2, Paul G. Richardson1,2, Teru Hideshima1,2, and Kenneth C. Anderson1,2

AbstractPurpose: NF-kB transcription factor plays a key role in the pathogenesis of multiple myeloma in the

context of the bone marrow microenvironment. Both canonical and noncanonical pathways contribute to

total NF-kB activity. Recent studies have shown a critical role for the noncanonical pathway: selective

inhibitors of the canonical pathway present a limited activity, mutations of the noncanonical pathway are

frequent, and bortezomib-induced cytotoxicity cannot be fully attributed to inhibition of canonical NF-kBactivity.

Experimental Design: Multiple myeloma cell lines, primary patient cells, and the human multiple

myeloma xenograft murine model were used to examine the biologic impact of dual inhibition of both

canonical and noncanonical NF-kB pathways.

Results: We show that PBS-1086 induces potent cytotoxicity in multiple myeloma cells but not in

peripheral blood mononuclear cells. PBS-1086 overcomes the proliferative and antiapoptotic effects of the

bone marrow milieu, associated with inhibition of NF-kB activity. Moreover, PBS-1086 strongly enhances

the cytotoxicity of bortezomib in bortezomib-resistant multiple myeloma cell lines and patient multiple

myeloma cells. PBS-1086 also inhibits osteoclastogenesis through an inhibition of RANK ligand (RANKL)–

induced NF-kB activation. Finally, in a xenograft model of human multiple myeloma in the bone marrow

milieu, PBS-1086 shows significant in vivo anti–multiple myeloma activity and prolongs host survival,

associated with apoptosis and inhibition of both NF-kB pathways in tumor cells.

Conclusions: Our data show that PBS-1086 is a promising dual inhibitor of the canonical and

noncanonical NF-kB pathways. Our preclinical study therefore provides the framework for clinical

evaluation of PBS-1086 in combination with bortezomib for the treatment of multiple myeloma and

related bone lesions. Clin Cancer Res; 1–13. �2012 AACR.

IntroductionMultiple myeloma is a clonal proliferation of malignant

plasma cells that accounts for 1% of all cancers and morethan 10% of all hematologic malignancies. Pure osteolyticbone lesions are pathognomonic of multiple myeloma andaffectmore than 80%of patients (1). Proteasome inhibitors

(bortezomib) have dramatically changed multiple myelo-ma prognosis by overcoming drug resistance to conven-tional treatments (2). However, resistance to bortezomibultimately occurs, highlighting the urgent need for newtherapeutic approaches (3).

NF-kB transcription factors play a key role in the path-ogenesis of cancers, including multiple myeloma, by regu-lating genes involved in proliferation, survival, and drugresistance (4, 5). Constitutive NF-kB activity is present inhuman multiple myeloma cell lines and patient multiplemyeloma cells (6, 7); moreover, adhesion of multiplemyeloma cells to bone marrow stromal cells (BMSC)induces NF-kB–dependent cytokine [interleukin (IL)-6,TNF-a, IL1-b, SDF-1a, and B-cell activating factor (BAFF)]transcription and secretion by BMSCs, which in turn furtheractivates NF-kB and thereby promotes multiple myelomacell growth and survival (6, 8–10). NF-kB also modulatesexpression of antiapoptotic proteins and adhesion mole-cules such as ICAM-1 (CD54) and VCAM-1 (CD106) onboth multiple myeloma and BMSCs, further enhancing

Authors' Affiliations: 1Jerome Lipper Multiple Myeloma Center, 2Depart-ment of Medical Oncology, Dana-Farber Cancer Institute, HarvardMedicalSchool, Boston, Massachusetts; 3Profectus BioSciences Inc.; 4rel�MD,Inc., Baltimore, Maryland; and 5Research Institute and Hospital, NationalCancer Center, Korea

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

Corresponding Author: Kenneth C. Anderson, Jerome Lipper MultipleMyeloma Center, Department of Medical Oncology, Dana-Farber CancerInstitute, 450Brookline Avenue, Boston,MA02215. Phone: 617-632-2144;Fax: 617-632-2140; E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-12-0779

�2012 American Association for Cancer Research.

ClinicalCancer

Research

www.aacrjournals.org OF1

Research. on August 15, 2019. © 2012 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 17, 2012; DOI: 10.1158/1078-0432.CCR-12-0779

http://clincancerres.aacrjournals.org/

adhesion ofmultiplemyeloma cells to BMSCs (11). Finally,NF-kB pathway plays an important role in bone osteolyticlesions, mostly via RANK (receptor activator of NF-kB)/RANK ligand (RANKL)-mediated activation of osteoclasts(12, 13). These studies validate NF-kB pathway as a prom-ising therapeutic target in multiple myeloma.

In multiple myeloma, NF-kB is constitutively present inthe cytoplasm in a latent inactive form through its interac-tion with inhibitory IkB proteins. After stimulation via thecanonical pathway, IkB is phosphorylated by inhibitor ofIkB kinase (IKK) complex at 2 specific N-terminal serineresidues (Ser32 and Ser36), leading to their ubiquitinationand degradation by the 26S proteasome. Rel/NF-kB com-plex is then released and translocates into the nucleus, inwhich it binds to DNA to activate transcription of varioustarget genes. Several studies also show a critical role for thenoncanonical NF-kB pathway in multiple myeloma path-ogenesis (14). Using an 11-gene expression signature forNF-kB activation, recent studies correlated constitutive NF-kB activity with mutations in regulators of NF-kB (CD40,NIK, TRAF2, and TRAF3; refs. 15–17). Overall mutationsinvolving both canonical and noncanonical NF-kB path-ways are present in at least 17% of multiple myelomapatient samples and 40% of multiple myeloma cell lines,enabling multiple myeloma cells to become less dependent

on extrinsic signals from the bone marrow microenviron-ment.Moreover,mutations of the noncanonical pathway in20% of multiple myeloma are associated with resistance tosteroids versus sensitivity to proteasome inhibitors.

To date, the canonical NF-kB pathway can be blocked bysmall-molecule inhibitors of IKKb (e.g., PS-1145 andMLN120B), which inhibit multiple myeloma cell growthin vitro. However, in vivo anti–multiple myeloma activity ofIKKb inhibitors is limited because of the compensatoryactivation of the noncanonical pathway (7, 18). Moreover,bortezomib inhibits inducible NF-kB activity in multiplemyeloma cells but unexpectedly enhances constitutive NF-kB activity via activation of the canonical pathway. There-fore, bortezomib-induced cytotoxicity cannot be fullyattributed to inhibition of canonical NF-kB activity inmultiple myeloma cells (19, 20). Because inhibition ofboth canonical and noncanonical pathways is required toefficiently block total NF-kB activity, we here characterizethe antitumor activity of PBS-1086, an inhibitor of bothcanonical and noncanonical NF-kB pathways (21), in mul-tiple myeloma.

Materials and MethodsReagents

PBS-1086 was provided by Profectus BioSciences Inc.Bortezomib was obtained from Selleck Chemicals. Doxo-rubicin and z-Val-Ala-Asp-fluoromethylketone (z-VAD-fmk) were obtained from Sigma-Aldrich. TNF-a, insulin-like growth factor-I (IGF-I), and recombinant IL-6 werepurchased from R&D Systems

Human multiple myeloma cell linesDexamethasone-sensitive (MM.1S) and dexamethasone-

resistant (MM.1R) cell lines were kindly provided by Dr.Steven Rosen (Northwestern University, Chicago, IL);RPMI-8226 and U266 were purchased from the AmericanType Culture Collection; doxorubicin (Dox)-resistantRPMI-Dox40 (Dox40) and melphalan-resistant RPMI-LR5(LR5) cell lines were provided by Dr. William Dalton(Moffitt Cancer Center, Tampa, FL); KMS18 by the DSMZ;IL-6–dependent INA6 by Dr. Renate Burger (University ofKiehl, Kiehl, Germany); and bortezomib-resistant IL-6–dependent cell lineANBL6-VR5and its parental counterpartANBL6-wt by Dr. Robert Orlowski (MD Anderson CancerCenter, Houston, TX). All multiple myeloma cell lineswere cultured in RPMI-1640 containing 10% FBS (SigmaChemical Co.]; 20%FBS for ANBL6), 2mmol/L L-glutamine,100 U/mL penicillin, and 100 mg/mL streptomycin (fromGIBCO). INA6 andANBL6 cell lines were culturedwith IL-6at 2.5 and 5 ng/mL, respectively.

Tumor cells and BMSCs from multiple myelomapatients

Blood samples from healthy volunteers were processedby Ficoll Hypaque (GE Healthcare) gradient to obtainperipheral blood mononuclear cells (PBMC) and stimulat-ed by phytohemagglutinin. Tumor cells and BMSCs frompatients with multiple myeloma were obtained from bone

Translational RelevanceNF-kB transcription factor plays a key role in the

pathogenesis of multiple myeloma in the context of thebone marrow microenvironment. Several studies vali-date NF-kB pathway as a promising therapeutic target inmultiple myeloma, with both canonical and noncanon-ical pathways contributing to total NF-kB activity. How-ever, selective inhibitors against these pathways have notyet been developed. The preclinical study presented hereis designed to characterize the antitumor activity of PBS-1086, a dual inhibitor of NF-kB pathways, in vitro inmultiple myeloma cell lines, patient multiple myelomacells, and also in the presence of bone marrow stromalcells. We show that PBS-1086 induces potent selectivecytotoxicity in multiple myeloma cells and overcomesthe prosurvival advantage conferred by the bonemarrowmilieu. We show a synergistic cytotoxicity in bortezo-mib-resistant multiple myeloma cell lines and patientmultiple myeloma cells, suggesting a potential clinicaluse of PBS-1086 in combination with bortezomib. Inaddition, our results indicate that PBS-1086 inhibitsosteoclast activity, suggesting potential benefit in mul-tiple myeloma–related bone disease. Finally, PBS-1086shows significant antitumor activity in ahumanmultiplemyeloma xenograft murinemodel with improvement ofoverall survival. Our preclinical study provides the ratio-nale for clinical evaluation of PBS-1086 in combinationwith bortezomib for the treatment ofmultiplemyeloma.

Fabre et al.

Clin Cancer Res; 2012 Clinical Cancer ResearchOF2




marrow aspirates after informed consent as per the Decla-ration of Helsinki and approval by the Institutional ReviewBoard of the Dana-Farber Cancer Institute (Boston, MA).Mononuclear cells were separated using Ficoll Hypaquedensity sedimentation, and plasma cells were purified(>95% CD138þ) by positive selection with anti-CD138magnetic activation cell separation microbeads (MiltenyiBiotech). BMSCs were generated by culturing bone marrowmononuclear cells for 4 to 6 weeks in Dulbecco’s modifiedEagle’s medium (DMEM) containing 15% FBS, 2 mmol/LL-glutamine, 100 U/mL penicillin, and 100 mg/mLstreptomycin.

Osteoclasts cultivation and differentiation assayPBMCs were isolated as described above and cultured in

a-MEM containing 10% fetal calf serum (FCS), 100 U/mLpenicillin, 100 mg/mL streptomycin, 25 ng/mLmacrophagecolony-stimulating factor (M-CSF; R&D Systems), and 25ng/mL RANKL (PeproTech). After 24 hours, the adherentpopulation was reseeded in 96-well plates. After a cultureperiod of 14 to 21 days, cells were stained for TRAP activity,using a leukocyte acid phosphatase kit (Sigma-Aldrich).Histologic micrographs were taken using a Leica DM200microscope (aperture HC PLANs 10�/22, objective lenses:N PLAN 10�) and a SPOT/insight QE model camera withSPOT advanced acquisition software (Diagnostic Instru-ments. Mature osteoclasts were identified as large multinu-cleated (>2 nuclei) TRAP-positive cells and quantified bylight microscopy.

Growth inhibition assayThe growth-inhibitory effect of PBS-1086 on multiple

myeloma or osteoclast cell growth was assessed by measur-ingMTT (Sigma-Aldrich) dye absorbance. Cells were pulsedwith MTT for the last 4 hours of culture, followed byisopropanol containing 0.04 N HCl. Absorbance was mea-sured at 570/630 nm using a spectrophotometer.

Detection of cytokinesBone serologic marker TRAP5b (tartrate-resistant acid

phosphatase 5b; ref. 22) was measured by ELISA (TRAP5b:Quidel) in culture supernatants from osteoclast, with orwithout PBS-1086.

Detection of apoptosis by Annexin V/propidium iodideDetection of apoptosis was done with the Annexin V-

fluorescein isothiocyanate (FITC)/propidium iodide (PI)Detection Kit (Immunotech/Beckman Coulter), as previ-ously described. Apoptotic cells were enumerated using theFACSCalibur FlowCytometer (BectonDickinson). AnnexinV-FITC single-positive cells were considered to be earlyapoptotic, PI single-positive cells necrotic, and AnnexinV-FITC double-positive cells late apoptotic.

DNA synthesisDNA synthesis was measured by [3H]-thymidine ([3H]-

TdR, Perkin-Elmer) uptake to evaluate growth of multiplemyeloma cells adherent to BMSCs in BMSC-coated 96-well

plates. Cells were pulsed with [3H]-TdR (0.5 mCi/well)during the last 8 hours of culture, harvested onto glassfilters with an automatic cell harvester (Cambridge Tech-nology), and counted using the Betaplate scintillationcounter.

Cytoplasmic and nuclear fractionationCytoplasmic and nuclear fractionation was done using

the nuclear extract kit (Panomics), according to the man-ufacturer’s instructions. For BMSC co-culture, multiplemyeloma cells were incubated in BMSC-coated flasks for12 hours, harvested by pipetting, and subjected to nuclearprotein extraction.

NF-kB DNA-binding ELISADNA-binding activity of NF-kB was quantified by ELISA

using the Trans-AM NF-kB Family Transcription FactorAssay Kit (Active Motif), according to the manufacturer’sinstructions. Briefly, nuclear extracts were transferred to a96-well plate coated with NF-kB consensus oligonucleo-tides. NF-kB proteins bound to the target sequence weredetected with primary antibodies specific for each Relmember and an horseradish peroxidase (HRP)-conjugatedsecondary antibody. Absorbance was measured at 595 nmas a relative measure of protein-bound NF-kB. Specificitywas confirmed by the use of wild-type and mutant NF-kBoligonucleotide competitors.

ImmunoblottingMultiple myeloma cells or mature osteoclasts were trea-

ted with or without therapeutic agents. Cells were thenharvested, washed, and lysed, as in prior studies. Whole-cell lysates were subjected to SDS-PAGE, transferred tomembranes, and immunoblotted with antibodies against:caspase-3, caspase-7, caspase-8, caspase-9, PARP, IkBa,phospho-IkBa (Ser32/36), extracellular signal–regulatedkinase (ERK), c-jun-NH2-terminal kinase (JNK), p38/MAPK, and glyceraldehyde-3-phosphate dehydrogenase(GAPDH; Cell Signaling Technology); as well as p50,p52, p65, RelB, c-Rel, nucleolin, and a-tubulin (Santa CruzBiotechnology). Densitometric analyses of scanned immu-noblotting imagesweredonewith theNIH ImageJ Software.

Osteoclast activity assayThe bone resorption activity of osteoclast was measured

using the Bone Resorption Assay Plate (Cosmo Bio). Oste-oclast precursors obtained from patient bone marrow sam-ples were seeded in the calcium phosphate–coated 24-wellplates in a-MEM containing 10% FCS, 25 ng/mL M-CSF,and 50ng/mLRANKL; PBS-1086was added at the indicateddoses. After a culture period of 14 to 21 days, the 24-wellplate was processed as directed bymanufacturer. Pit resorp-tion areas corresponded to dark areas in which calciumphosphatewas resorbed as opposed to bright areas inwhichcalcium phosphate layer remained intact. Mean pit resorp-tion areaswere analyzed by lightmicroscopy andquantifiedby usingNIH ImageJ program.Histologicmicrographsweretaken using a Leica DM200 microscope (aperture HC

Dual Pathway Inhibition of NF-kB in Myeloma

www.aacrjournals.org Clin Cancer Res; 2012 OF3




B C

D E

PBS-1086 concentration (µmol/L)

Inh

ibit

ion

of

NF

-κB

(in

%)

120

100

80

60

40

20

0Inh

ibit

ion

of

NF

-κB

(in

%)

0.01 0.1 1 10

Time (hours)

CT0.5 1 1.5 2 4 8 12

PBS-1086 5 µmol/L

p65

p50

p52

65 kDa

50 kDa

52 kDa

Nucleolin 110 kDa

37 kDa

Nuclearextracts

Cytoplasmicextracts

3

2.5

2

1.5

1

0.5

0

Ab

sorb

ance

(O

D)

CT TNF-α PBS-1086

1 h

PBS-1086

5µmol/L

0.5µmol/L

5µmol/L

0.5µmol/L

2 h

CT

CT TNF-α PBS-1086 1 µmol/L

2 h 4 h 6 h 8 h10 h12 h

CT TNF-α PBS-1086 1 µmol/L

2 h 4 h 6 h 8 h10 h12 hp65

p50

p52

GAPDH

65 kDa

50 kDa

52 kDa

100 kDa

105 kDa

F

Nu

clea

r p

rote

in e

xpre

ssio

n

p50 p65 p52

2 4 6 8 10 12

PBS-1086

2 4 6 8 10 12

PBS-1086

2 4 6 8 10 12

PBS-1086

10080

6040

200

- 20

Cyt

op

lasm

ic p

rote

in e

xpre

ssio

n

p50 p65 p52

2 4 6 8 10 12

PBS-1086

2 4 6 8 10 12

PBS-1086 CT TNF-αCT TNF-αCT TNF-αCT TNF-αCT TNF-αCT TNF-α

2 4 6 8 10 12

PBS-1086

2.5

2

1.5

1

0

0.5

1.5

1

0

0.5

A

3.5

O

O

OHHN

HET

R2O(R2)n

Figure 1. PBS-1086 inhibits both canonical and noncanonical NF-kB pathways. A, chemical structure of PBS-1086. B, MM.1S cells were cultured with PBS-1086 at the indicated doses (0.01, 0.1, 1, and 10 mmol/L) for 2 hours. NF-kBDNAbinding activity inMM.1S nuclear extracts was assessed in vitro using Trans-AM NF-kB family transcription factor assay kit. NF-kB canonical activity included p65 ( ), p50 ( ), and c-Rel ( ); NF-kB noncanonicalactivity included p52 ( ) and RelB ( ). C, MM.1S cells were cultured with PBS-1086 (5 mmol/L) at the indicated times (0.5–12 hours). NF-kBDNAbinding activity in MM.1S nuclear extracts was measured by ELISA. NF-kB canonical activity included p65 ( ), p50 ( ), and c-Rel ( );NF-kB noncanonical activity included p52 ( ) and RelB ( ). Results are expressed as percentage inhibition from the quantity of NF-kB proteinbound in the PBS-1086–treated relative to the maximum quantity bound in the control (CT). ELISA data shown are representative of 3 independentexperiments. D, MM.1S cells were treated with PBS-1086 (0.5 and 5 mmol/L) for the indicated times (1 and 2 hours). NF-kB DNA binding activity in MM.1Snuclear extractswasmeasuredbyELISA.NF-kBcanonical activity includedp65 ( ), p50 ( ), andc-Rel ( ); NF-kBnoncanonical activity includedp52 ( ) andRelB ( ). Treatment with TNF-a (10 ng/mL) for 1 hour served as a positive control of NF-kB activity for both time points, 1 and 2 hours. The results ofELISA are expressed as relative absorbance. Data represent mean �SD of 3 independent experiments. OD, optical density. E, MM.1S cells were culturedwith PBS-1086 (1 mmol/L) for the indicated times (2 to 12 hours). Treatment with TNF-a (10 ng/mL) for 1 hour served as a positive control of increased p65,p50, and p52 nuclear translocation. Nuclear and cytoplasmic extracts were subjected to Western blotting using p50, p52, p65, GAPDH, andnucleolin antibodies. GAPDH and nucleolin were used as purity and loading controls for cytoplasmic and nuclear extracts, respectively. Blots arerepresentative of 3 independent experiments. F, the densitometric analysis of scanned immunoblotting images was done with the NIH ImageJ Software.Results of nuclear ( ; left) and cytoplasmic ( ; right) protein expression are expressed as fold change relative to control.

Fabre et al.





B

A

Via

ble

cel

ls (

% o

f co

ntr

ol)

Via

ble

cel

ls (

% o

f co

ntr

ol)

125

100

75

50

25

0

125

100

75

50

25

02.51.250.620.310.08 0.15

Via

ble

cel

ls (

% o

f co

ntr

ol)

40201051.25 2.5

125

100

75

50

25

0Pt 1 Pt 2 Pt 3 Pt 4

CT CT

C

Via

ble

cel

ls (

% o

f co

ntr

ol)

125

100

75

50

25

0105

Donor 4

Donor 3

Donor 1

Donor 2

2.51.250.620.310.15

CT

D

E

100

80

60

40

20

0

120

100

80

60

40

20

0

120

CT PBS-1086 (µmol/L) PBS-1086 (µmol/L)

PBS-1086 (µmol/L)

PBS-1086 (µmol/L)PBS-1086 (µmol/L)

0.31 0.62 1.25

CT

0.31 0.62 1.25

% o

f ce

lls

% o

f ce

lls

F

Via

ble

cel

ls (

% o

f co

ntr

ol)

125

100

75

50

25

0CT CT24 h 48 h 24 h 48 h

PBS-10860.62 µmol/L

PBS-10861.25 µmol/L

PARP

Caspase-3

Caspase-8

Caspase-9

GAPDH

Caspase-7

FLCL

PBS-1086CT

FL

CLFL

CLFL

CLFLCL

37 kDa

1.25 2.5 5 µmol/L

24 h 48 h

Figure 2. PBS-1086 induces cytotoxicity and apoptosis in multiple myeloma cells. A, multiple myeloma cell lines MM.1S ( ), INA6 ( ), and KMS18( ) were cultured with PBS-1086 (0.08–2.5 mmol/L) for 48 hours (left). Multiple myeloma cell lines MM.1R ( ), Dox40 ( ), LR5 ( ), RPMI8226 ( ), and U266 ( ) were treated with PBS-1086 (1.25–40 mmol/L) for 48 hours (right). B, CD138þ multiple myeloma cells from 4 patients weretreated with PBS-1086 at 0 ( ), 0.62 ( ), and 1.25 mmol/L ( ) for 48 hours. C, mononuclear cells isolated from 4 healthy donors and stimulated byphytohemagglutinin were culturedwith PBS-1086 (0.15–10 mmol/L) for 48 hours. Cell viability was assessed byMTT assay of triplicate cultures, expressed aspercentage of untreated control. Data representmean�SDviability. D,MM.1S cellswere treatedwith PBS-1086 at the indicated doses (0.31–1.25mmol/L) for24 and 48 hours. Apoptotic cells were analyzed by flow cytometry using Annexin V/PI staining. Percentages of viable (AV�/PI�; ), early apoptotic

(AVþ/PI�; ), late apoptotic (AVþ/PIþ; ), and necrotic cells (AV�/PIþ; ) are shown as a histogram. Data represent mean � SD of 3 independentexperiments. E, MM.1S cells were culturedwith PBS-1086 at the indicated doses (1.25–5 mmol/L) for 24 hours. Whole-cell lysates were subjected toWesternblotting using anti-caspase-3, caspase-7, caspase-8, caspase-9, PARP, and GAPDH antibodies. CF, cleaved fragment; FL, full length. GAPDH wasused as a loading control. Blots are representative of 3 independent experiments. F, MM.1S cells were cultured with PBS-1086 (0.62 and 1.25 mmol/L) for 24and 48 hours in the presence ( ) or absence ( ) of z-VAD (20 mmol/L). Cell viability was assessed by MTT assay of triplicate cultures, expressed aspercentage of untreated control. Data represent mean � SD viability. CT, control.






PLANs 10�/22, objective lenses:NPLAN10�) and a SPOT/insight QE model camera with SPOT advanced acquisitionsoftware (Diagnostic Instruments). The average pit resorp-tion area is indicated as percentage of control (osteoclastsstimulated with M-CSF and RANKL).

Murine xenograft model of human multiple myelomaCB17 SCIDmice (6- to 8-week-oldmale) were purchased

from Charles River Laboratories, Inc. All animal studieswere conducted according to protocols approved by andconform to the relevant regulatory standards of the Insti-tutional Animal Care and Use Committee of the Dana-Farber Cancer Institute. Mice were injected subcutaneouslywith 5 � 106 MM.1S cells in 100 mL FCS-free RPMI-1640medium.When tumor wasmeasurable, mice were assignedinto 6 treatment groups: PBS-1086 (at 7.5 mg/kg), borte-zomib, PBS-1086 (at 2.5 or 7.5 mg/kg) with bortezomib,vehicle (20% dimethyl sulfoxide/80% Cremophor), andcontrol (100% saline). PBS-1086 was given intraperitone-ally (i.p.) once daily, 5 d/wk for 4 weeks. Control andvehicle were administered i.p. with the same schedule.

Bortezomib was given i.v. at 0.5 mg/kg twice a week for4 weeks. Each group consisted of at least 8 tumor-bearingmice. Tumor volume was calculated from caliper measure-ments 3 times per week, using the following formula: V ¼0.5a� b2 in which a and b are the long and short diametersof the tumor, respectively. Mice were euthanized whentumor volume reached 2 cm3. Survival was evaluated fromthe first day of treatment until death.

In situ detection of apoptosis andimmunohistochemistry

Sections from harvested tumors were subjected to immu-nohistochemical (IHC) staining for terminal deoxynucleo-tidyl transferase–mediated dUTP nick end labeling(TUNEL) for detection of apoptosis. Ki67 was also assessedby IHC staining to quantify proliferation.

Statistical analysisStatistical significance was determined by sample t test

and test for equal variance. Theminimal level of significancewas P < 0.05. Overall survival was analyzed by Kaplan–

Ab

sorb

ance

(O

D)

1.5

1

0.5

0

MM.1S BMSC BMSC/MM.1S

TNF-αPBS-1086 (µmol/L)

--

---

--

- - --

++ + +0.5 0.50.5-5 5 5

DC

3 H-T

hym

idin

e u

pta

ke (

××103

)

BMSC MM.1S BMSC/MM.1S

B

Via

ble

cel

ls (

% o

f co

ntr

ol)

60

50

40

30

20

10

0

A

Via

ble

cel

ls (

% o

f co

ntr

ol)

125

100

75

50

25

0

150

0 1 10 50

IL-6 (ng/mL)

0 1 10 50

IGF-I (ng/mL)

125

100

75

50

25

0

150

Figure 3. PBS-1086 overcomes the proliferative and antiapoptotic effects of BMSCs associated with inhibition of NF-kB. A, MM.1S cells were cultured for 48hours with PBS-1086 0 ( ), 0.15 ( ), 0.31 ( ), 0.62 ( ), 1.25 ( ), and 2.5 mmol/L ( ), in the absence or presence of IL-6 (1, 10, and 50 ng/mL). B,MM.1S cells

were cultured for 48 hours with PBS-1086 0 ( ), 0.15 ( ), 0.31 ( ), 0.62 ( ), 1.25 ( ), and 2.5 mmol/L ( ) in the absence or presence of IGF-I (1, 10, and50 ng/mL). Cell viability was assessed by MTT assay of triplicate cultures, expressed as percentage of untreated control. Data represent mean � SDviability. C, MM.1S cells were treated for 48 hours with PBS-1086 0 ( ), 0.62 ( ), 1.25 ( ), and 2.5 mmol/L ( ) in the presence or absence of BMSCs. Cellviability was assessed by thymidine uptake of quadruplicate cultures, expressed as percentage of untreated control. Data represent mean� SD viability. D,MM.1Scellswere treated for 2 hourswithPBS-1086 (0.5 and5mmol/L), in thepresenceor absenceofBMSCs.NF-kBDNAbinding activity inMM.1S�BMSCsnuclear extracts was measured by ELISA. NF-kB canonical activity included p65 ( ), p50 ( ), and c-Rel ( ); NF-kB noncanonical activity included p52 ( )

and RelB ( ). Treatment with TNF-a (10 ng/mL) for 2 hours served as a positive control of NF-kB activation in MM.1S cells. The results of ELISAwere expressed as relative absorbance. Data represent mean � SD of 3 independent experiments. OD, optical density.

Fabre et al.





Meier method. The Kaplan–Meier curves were constructedfor each group using GraphPad Prism software, and the log-rank (Mantel–Cox) test was used to compare survival curvesamong groups. The combined effect of PBS-1086 withbortezomib was analyzed by isobologram analysis usingthe CompuSyn software program (ComboSyn, Inc).

ResultsPBS-1086 inhibits both canonical and noncanonicalNF-kB pathwaysWe first evaluated the effect of PBS-1086 (chemical

structure in Fig. 1A) on NF-kB activity in multiple mye-

loma cells. PBS-1086 inhibited the binding of all Relproteins to DNA. At 0.1 mmol/L PBS-1086, p65 and p50were inhibited by 62% and 56%, compared with 28%and 22% inhibition of p52 and RelB (Fig. 1B). Toevaluate the kinetics of NF-kB inhibition, MM.1S cellswere treated with PBS-1086 for 0.5 to 12 hours. PBS-1086inhibited NF-kB binding to DNA even after a shortexposure, with potent inhibition of p65, p50 (90%–95%), p52 (63%), and RelB (79%; Fig. 1C). Dose- andtime-dependent NF-kB inhibition after PBS-1086 treat-ment was also confirmed in MM.1S cells (Fig. 1D).Inhibition of NF-kB activity by PBS-1086 in additional

A B C

D

Via

ble

cel

ls (

% o

f co

ntr

ol) 125

100

75

50

25

0

Via

ble

cel

ls (

% o

f co

ntr

ol) 125

100

75

50

25

0

2.5

2

1.5

1

0.5

0

Ab

sorb

ance

(O

D)

CT TNF-α PBS-1086 Bort PBS-1086+ Bort

CT TNF-αPBS-10

86

BortPBS-10

86 +

Bort

CT

TNF-α

PBS-1086 Bort

PBS-1086

+ Bort

p50p52

50 kDa52 kDa

Fo

ld c

han

ge

exp

ress

ion

(rel

ativ

e to

co

ntr

ol)

Nucleolin 110 kDa

1.5

1

0.5

0

Via

ble

cel

ls (

% o

f co

ntr

ol) 125

100

75

50

25

0Pt 1 Pt 2

0 0.31 0.62 1.25 0.15 0.31 0.62 1.250

E

2

2.5

PBS-1086 (µmol/L) PBS-1086 (µmol/L)

Figure 4. PBS-1086 enhancesmultiplemyelomacytotoxicity of bortezomib in bortezomib-resistantmultiplemyeloma cells. A, Dox40 cellswere cultured for 72hours with bortezomib (Bort) at 0 ( ), 10 ( ), and 15 nmol/L ( ) in the presence or absence of PBS-1086 (0.31–1.25 mmol/L). B, ANBL6-VR5 cells werecultured for 48 hours with bortezomib (Bort) 0 ( ), 10 ( ), and 20 nmol/L ( ) in the presence or absence of PBS-1086 (0.15–1.25 mmol/L). Cell viability wasassessed byMTTassay of triplicate cultures, expressed as percentage of untreated control. Data representmean�SDviability. C,CD138þmultiplemyelomacells from 2 patients with bortezomib-resistant multiple myeloma (Pt1 and Pt2) were treated with bortezomib (Bort) at 10 nmol/L ( ) alone, PBS-1086(1.25 mmol/L; ) alone, PBS-1086þBort ( ), or not treated ( ) for 48 hours. Cell viability was assessed by MTT assay of triplicate cultures, expressed aspercentage of untreated control. Data representmean�SDviability. D,ANBL6-VR5cellswere treated for 2 hourswith PBS-1086 (1mmol/L), in thepresenceorabsence of bortezomib (40 nmol/L). NF-kB canonical activity included p65 ( ), p50 ( ), and c-Rel ( ); NF-kB noncanonical activity included p52 ( ) and

RelB ( ). Treatment with TNF-a (10 ng/mL) for 2 hours served as a positive control of NF-kB activation in ANBL6-VR5 cells. The results of ELISA wereexpressed as relative absorbance.Data representmean�SDof 3 independent experiments. E, corresponding nuclear extracts fromANBL6-VR5cells treatedfor 2 hours with PBS-1086 (1 mmol/L) � bortezomib (40 nmol/L) were subjected to Western blotting using p50, p52, and nucleolin antibodies. Blots arerepresentative of 3 independent experiments. The densitometric analysis of scanned immunoblotting images for p50 ( ) and p52 ( ) was done withthe NIH ImageJ Software and expressed as fold change relative to nontreated cells.






(% o

f co

ntr

ol)

(% o

f co

ntr

ol)

(% o

f co

ntr

ol)

(% o

f co

ntr

ol)

A B

C D

E F3 H

-Th

ymid

ine

up

take

(×1

03)

Figure 5. PBS-1086 inhibits osteoclast activity associated with inhibition of NF-kB. A, osteoclasts (OC) andMM.1S cells were cultured for 48 hours with PBS-1086 at 0 ( ), 0.31 ( ), 0.62 ( ), 1.25 ( ), 2.5 ( ), and5mmol/L ( ). Cell viabilitywas assessedbyMTTassayof triplicate cultures, expressed aspercentageofuntreated control. Data represent mean � SD viability. B, cell viability was assessed by thymidine uptake of quadruplicate cultures, expressed aspercentage of untreated control. Data represent mean � SD viability. C, mature osteoclasts were treated for 48 hours with PBS-1086 (0.31–5 mmol/L). Cellswere stained for TRAP activity. Cell density was equal in all samples. TRAP-positive multinucleated osteoclasts after PBS-1086 treatment are quantitated aspercentage of untreated control. For nontreated control as well as PBS-1086 (1.25 and 5 mmol/L)–treated cultures, the corresponding micrographsare shown (�10), with inserts at highermagnification. Data representmean�SDcounts of 3 independent experiments. D,mature osteoclasts were treated for48 hours with PBS-1086 (0.31–5 mmol/L). The presence of TRAP5b in the supernatants of culture was quantified by ELISA, expressed as percentage ofuntreated control. Data represent mean � SD absorbance of triplicate experiments. E, primary human osteoclast precursors derived from 2 patientswith multiple myeloma were seeded on calcium phosphate–coated plates. Cells were treated with PBS-1086 (0.31–5 mmol/L) in the presence of M-CSF(25 ng/mL) andRANKL (50 ng/mL) tomature osteoclasts.Osteoclasts not stimulatedwithM-SCFnor RANKL served as anegative control and failed to resorb.Osteoclasts stimulated with M-CSF and RANKL (positive control) resorbed calcium phosphate (>95%). The average pit resorption area with PBS-1086 wasexpressed as percentage of positive control. For nontreated control, as well as PBS-1086 (1.25 and 5 mmol/L)–treated cultures, the correspondingmicrographs are shown (�10), with inserts at higher magnification. Data represent mean � SD of 4 independent experiments. F, nuclear and cytoplasmicextracts frommature osteoclastswere cultured for 2 hourswithPBS-1086 at 1 and5mmol/L. Control nontreated osteoclastswere only stimulatedwithRANKLandM-CSF. Nuclear extracts were subjected to Western blotting using p50, p52, and nucleolin antibodies. Cytoplasmic extracts were subjected to Westernblotting using IkBa, p-IkBa (Ser 32/36), p38, and a-tubulin antibodies. Nucleolin and a-tubulin were used as purity and loading controls for nuclear andcytoplasmic extracts, respectively. Blots are representative of 3 independent experiments. CT, control.

Fabre et al.





multiple myeloma cell lines is shown in SupplementaryFig. S1A (KMS18) and S1B (INA6). We examined theexpression of p65, p50, and p52 in cytoplasmic andnuclear extracts from MM.1S cells treated with PBS-1086 by Western blotting (Fig. 1E). A time-dependentdecrease in expression of NF-kB proteins was observed innuclear extracts, whereas cytoplasmic protein remainedstable (Fig. 1F), confirming that PBS-1086 acts throughan inhibition of NF-kB translocation into the nucleus.Furthermore, PBS-1086 did not inhibit the phosphory-lation and degradation of IkBa in multiple myeloma cells(Supplementary Fig. S2). Altogether, these data confirmthat PBS-1086 blocks both canonical and noncanonicalNF-kB pathways in multiple myeloma cell lines in a dose-and time-dependent manner.

PBS-1086 induces cytotoxicity and apoptosis inmultiple myeloma cellsWe next investigated the growth-inhibitory effect of

PBS-1086 in vitro. Treatment of multiple myeloma cellswith PBS-1086 for 48 hours induced a dose-dependentdecrease in cell viability, with IC50 values of PBS-1086ranging from 0.31 to 5 mmol/L (Fig. 2A). Cells vary intheir baseline NF-kB activity (19) and included sensitivecells (MM.1S, INA6, and KMS18 with IC50 0.31–0.62mmol/L), intermediate sensitive cells (MM.1R, Dox40,and RPMI-8226 with IC50 1.25–2.5 mmol/L), and resis-tant cells (LR5 and U266 with IC50 2.5–5 mmol/L).However, growth-inhibitory effect of PBS-1086 wasobserved in multiple myeloma cell lines irrespective oftheir sensitivity or resistance to conventional treatmentsor their genetic background. A similar growth-inhibitoryeffect was observed in 4 patient multiple myeloma cellstreated for 48 hours with PBS-1086 (Fig. 2B). NormalPBMCs from 4 healthy volunteers stimulated by phyto-hemagglutinin were treated for 48 hours with PBS-1086and similarly analyzed for cytotoxicity. PBS-1086 showedonly modest cytotoxicity on normal PBMCs, with a max-imum viability loss of 10% at 1.25 mmol/L (Fig. 2C) andIC50 0.62 to 1.25 mmol/L, indicating a higher sensitivityof tumor cells than PBMCs to PBS-1086. These resultssuggest that PBS-1086 potently inhibits the growth ofmultiple myeloma cell lines in a dose-dependent mannerwith a favorable therapeutic window. To analyze molec-ular mechanisms whereby PBS-1086 induces cytotoxicityin multiple myeloma cells, MM.1S cells treated with PBS-1086 were analyzed for apoptosis using Annexin V-PIstaining. PBS-1086 significantly increased Annexin V(þ)/PI(�) in MM.1S cells in a time- and dose-dependentmanner (Fig. 2D). Moreover, PBS-1086 triggered cleavageof caspase-8, caspase-3/7, and PARP indicating that theintrinsic apoptotic pathway was predominantly activatedby PBS-1086 (Fig. 2E). Conversely, pan-caspase inhibitorz-VAD-fmk markedly abrogated PBS-1086–induced apo-ptosis, confirming that PBS-1086–induced cytotoxicity ismediated, at least in part, via caspase-dependent apopto-sis (Fig. 2F). Taken together, these results show that PBS-1086 induced apoptosis in multiple myeloma cells.

PBS-1086 overcomes the proliferative andantiapoptotic effects of BMSCs associated withinhibition of NF-kB

Because IL-6 and IGF-I are major growth and/or survivalfactors in the bonemarrowmilieu (8, 23–28), we examinedtheir effect on PBS-1086–induced apoptosis. Even withexogenous IL-6 or IGF-I, PBS-1086 induced cytotoxicity(Fig. 3A and B). In a dose-dependent manner, PBS-1086also inhibited growth of MM.1S cells cocultured withBMSCs (Fig. 3C). We next investigated whether PBS-1086could inhibit NF-kB inducible activity in the bone marrowmicroenvironment. Co-culture of MM.1S cells with BMSCsactivated NF-kB, predominantly via the canonical pathway,which was significantly inhibited by PBS-1086 (Fig. 3D).We also investigated the effect of PBS-1086 on BMSCs.Importantly, PBS-1086 inhibited both canonical and non-canonical NF-kB pathways in BMSCs. Altogether, theseresults show that PBS-1086 targets not only multiple mye-loma cells but also the bone marrow microenvironment toovercome the proliferative and antiapoptotic effects ofBMSCs.

PBS-1086 enhances cytotoxicity of bortezomib inbortezomib-resistant multiple myeloma cells

We next evaluated the combination of PBS-1086 withbortezomib in bortezomib-resistant multiple myeloma celllines. Combining PBS-1086 with bortezomib triggered syn-ergistic cytotoxicity against Dox40 (Fig. 4A and Supplemen-tary Table S3) and ANBL6-VR5 (Fig. 4B and SupplementaryTable S3) cells. For example, PBS-1086 (0.31 mmol/L) andbortezomib (20 nmol/L) triggered 8% and 35% cytotoxic-ity, whereas PBS-1086 with bortezomib at the same con-centrations induced 52% cytotoxicity (Fig. 4B). In bortezo-mib-resistant patient multiple myeloma cells, combiningPBS-1086 with bortezomib markedly decreased multiplemyeloma cell viability (Fig. 4C), suggesting that PBS-1086overcomes bortezomib resistance (Supplementary Fig. S4).We further investigated the effect of PBS-1086/bortezomibcombination on NF-kB activity in bortezomib-resistantANBL6-VR5 cells. The binding activity of Rel proteins wassimilar in bortezomib-treated and control cells (Fig. 4D).Moreover, bortezomib-induced activation of canonical NF-kB pathway was inhibited by PBS-1086 (Fig. 4D and E). Nooverexpression in hsp27, which has been associated withbortezomib resistance (ref. 29; data not shown), nor changein p38, an upstream effector of hsp27 (30, 31), wasobserved in ANBL6-VR5 cells after PBS-1086 treatment(Supplementary Fig. S4).

PBS-1086 inhibits osteoclast activity associated withinhibition of NF-kB

We also investigated the effect of PBS-1086 on osteo-clasts. PBS-1086 did not induce cytotoxicity in osteoclasts,as evidenced by both MTT and 3H-thymidine uptake(Fig. 5A and B). To examine the effect of PBS-1086 onosteoclast differentiation, mature osteoclasts were treatedwith PBS-1086, followed by TRAP assay. PBS-1086 inhib-ited osteoclast differentiation in a dose-dependent manner






(Fig. 5C). To evaluate the effect of PBS-1086 on osteoclastactivity, we quantified TRAP5b in the culture supernatantsof mature osteoclasts treated with PBS-1086: 15% and 25%reduction in TRAP5b activity was observed at 1.25 and 5mmol/L, respectively (Fig. 5D). Functional osteoclast activ-ity resorbing bone substrate was also assessed using calciumphosphate–coated plates. PBS-1086 inhibited the boneresorption activity of mature osteoclasts in a dose-depen-dent manner, with 75% and 100% decrease in pit area at1.25 and 5 mmol/L PBS-1086, respectively (Fig. 5E). Finally,we investigated whether the inhibitory effect of PBS-1086on osteoclastogenesis was due to inhibition of NF-kB.Mature osteoclasts exhibited high baseline p50 expressionin the nucleus (Fig. 5F), suggesting that NF-kB activity inmature osteoclasts ismaintained via the canonical pathway.Importantly, PBS-1086 inhibited both p50 and p52 NF-kBin a dose-dependentmanner. No change of p38MAPK, JNK,and ERK, or other pathways mediating osteoclastogenesis(32, 33), was observed (Supplementary Fig. S5). Takentogether, our results indicate that PBS-1086 inhibitsRANKL-induced osteoclastogenesis, associated with inhibi-tion of NF-kB.

PBS-1086 inhibits tumor growth in amurine xenograftmodel of human multiple myeloma

Finally, we investigated the effect of PBS-1086 in com-bination with bortezomib on MM.1S cell growth in amurine xenograft model of humanmultiplemyeloma cells.Mice were injected subcutaneously with MM.1S cells thatexhibit both canonical and noncanonical NF-kB pathways.In combination with bortezomib, PBS-1086 significantlyinhibited tumor growth versus control (2.5 mg/kg, P ¼0.00039 and 7.5 mg/kg, P ¼ 0.00084; Fig. 6A). Tumorgrowth was also significantly reduced in PBS-1086 withbortezomib versus bortezomib alone (2.5 mg/kg, P ¼0.00325 and 7.5 mg/kg, P ¼ 0.0057) treated cohorts.Importantly, treatment with PBS-1086 and bortezomibprolonged overall survival versus control (2.5 mg/kg, P ¼0.0133 and 7.5 mg/kg, P ¼ 0.0121). Overall survival wasalso significantly increased in PBS-1086 and bortezomibgroups versusbortezomib alone (2.5mg/kg,P¼0.0151 and7.5mg/kg, P¼ 0.0307; Fig. 6B).With amedian follow-up of140 days (range, 28–170), 50% mice receiving PBS-10867.5 mg/kg and bortezomib were alive with no detectabletumor. Overall, treatment with PBS-1086, alone or in com-bination, was well tolerated, with no significant bodyweight changes compared with control group (Fig. 6C).IHC staining for NF-kB on excised tumor showed time-dependent inhibition of both canonical and noncanonicalNF-kB pathways, with significantly decreased expression ofp65, p50, and p52 in the nucleus in tumor cells harvestedfrom PBS-1086–treated mice (data not shown). PBS-1086also induced time-dependent apoptosis (TUNEL staining)and decrease in proliferation (Ki67 staining) in vivo (datanot shown). Altogether, our in vivo study shows that inhi-bitionofNF-kBbyPBS-1086 is associatedwith significant invivo anti–multiple myeloma activity and improved overallsurvival.

A

B

C 125

100

75

50

25

0Bo

dy

wei

gh

t ch

ang

es(i

n %

)

Days from treatment start



Su

rviv

al (

%)

Tum

or

volu

me

(mm

3 )

3,000

2,000

1,000

01 8 15 22 29 36 43 50 57

CT

VehiclePBS-1086

(IP)

(IV) Bort

Wk 1

20 40 60 80 100 120 1400

20

40

60

80

100

Treatment schedule

1 3 5 8 10 12 15 1719 22 24 26 29

Wk 2 Wk 3 Wk 4

Figure 6. PBS-1086 inhibits tumor growth in a murine xenograft model ofhumanmultiplemyeloma.A,SCIDmicewere injected subcutaneouslywith5 � 106 MM.1S cells and treated with 7.5 mg/kg PBS-1086 i.p. daily for 4weeks ( ;N¼ 8); 0.5mg/kg bortezomib i.v. twice a week for 4 weeks( ; N ¼ 8); PBS-1086 at 2.5 mg/kg i.p. and bortezomib (Bort) i.v.( ;N¼ 9); or PBS-1086 at 7.5mg/kg i.p. and bortezomib i.v. ( ;N ¼ 9). A vehicle control group ( ; N ¼ 8) received i.p. injections ofvehicle alone daily for 4 weeks. The nontreated control (CT) group ( ;N¼ 8) received i.p. injectionsof salinedaily for 4weeks. Tumor volumewascalculated from caliper measurements 3 times per week, and error barsrepresent�SE.Tumorvolumecurveextends to57days fromthefirst dayoftreatment, when half of the mice per group have died. B, survival rate wasevaluated from the first day of treatment using Kaplan–Meier curves. C,body weight of mice treated with PBS-1086, bortezomib (Bort), PBS-1086þ bortezomib (PBS-1086 þ Bort), vehicle (Veh), or control (CT) wasexpressed as percentage of baseline. Body weight curve extends to 29days from the first day of treatment, when the death in any of the groupswas first observed. Error bars represent �SD body weight.

Fabre et al.





DiscussionNF-kB transcription factors play a critical role in the

pathogenesis of multiple myeloma because constitutiveNF-kB activation promotes cell growth, survival, anddrug resistance (6). NF-kB is involved in the intimatecrosstalk among multiple myeloma cells, the bone mar-row microenvironment, and the bone matrix (34). Con-sequently, targeting direct NF-kB is a promising treat-ment strategy in multiple myeloma. The importance ofNF-kB noncanonical pathway in multiple myeloma hasbeen confirmed both by the frequency of mutationsaffecting this pathway (15, 16, 18) and the limitedactivity of IKKb inhibitors of the canonical pathway(18). Therefore, inhibition of both canonical and non-canonical NF-kB pathways is necessary to achieve com-plete blockade of NF-kB activity. In the present study, weinvestigated the effect of PBS-1086, a dual NF-kB inhib-itor. PBS-1086 binds to Rel proteins to form covalentbonds with cysteine 38 in RelA (p65), cysteine 144 inRelB, or cysteine 67 in c-Rel to inhibit binding of Rel tothe kB sites in DNA (35).We first confirmed that PBS-1086 strongly inhibits both

canonical and noncanonical NF-kB pathways in multiplemyeloma cells. MTT evaluation in multiple myeloma celllines and patient CD138þ multiple myeloma cells showsa selective effect on multiple myeloma, with a favorabletherapeutic index. PBS-1086 induces apoptosis in multi-ple myeloma cells through activation of the intrinsicapoptotic pathway. Partial reversibility of cell killing inthe presence of a pan-caspase inhibitor z-VAD-fmk sug-gests that PBS-1086 might also trigger signaling pathwaysother than caspases. PBS-1086 induces synergistic cyto-toxicity in combination with bortezomib in bortezomib-resistant multiple myeloma cells and in patient multiplemyeloma cells refractory to bortezomib. Overexpressionof hsp27 confers bortezomib resistance (29); however,this mechanism was not observed in ANBL6-VR5 cell line.We did not determine whether PBS-1086 might interferewith additional mechanisms contributing to bortezomibresistance, such as mutations of proteasome subunits orincreased activity of the aggresome pathway (36, 37). Incombination with bortezomib, PBS-1086 induced potentanti–multiple myeloma activity in vivo in a murine xeno-graft model of human multiple myeloma, with significanttumor growth reduction. Importantly, significantly pro-longed overall survival was observed in PBS-1086/borte-zomib combination treatment groups. IHC analysis ofharvested tumors has shown that PBS-1086 anti–multiplemyeloma activity in vivo was associated with dual inhi-bition of NF-kB pathway in tumor cells. Altogether, ourresults suggest a broad clinical applicability of PBS-1086to overcome bortezomib resistance not only in multiplemyeloma but also in other hematologic malignancies andsolid tumors (38).Several studies have emphasized the role of the bone

marrow microenvironment in promoting drug resistance,showing the need for anti–multiple myeloma agents totarget not only multiple myeloma cells but also BMSCs

(39). In the bone marrow milieu, TNF-a is secreted bymultiple myeloma cells and induces NF-kB–dependentexpression of adhesion molecules on both multiple mye-loma cells and BMSCs, further increasing cell adhesion(11). Enhanced binding in turn confers resistance toapoptosis and triggers NF-kB–dependent secretion ofcytokines (11). Our data show that PBS-1086 inhibitsboth constitutive and inducible NF-kB activity in multiplemyeloma cells and in BMSCs. Our laboratory has previ-ously shown that bortezomib inhibits TNF-a–inducedNF-kB activation (18). However, some multiple myelomacells have constitutive activation of NF-kB through pro-teasome inhibitor–resistant pathways, leading to borte-zomib resistance due to both constitutive and inducibleNF-kB activities (40). Our data show, unlike bortezomib,that PBS-1086 might be efficacious even when protea-some inhibitor–resistant pathways are activated. More-over, during multiple myeloma progression, mutations ofthe NF-kB pathway lead to decreased dependence onextrinsic signals from the bone marrow microenvironment(41). Importantly, our data further suggest that PBS-1086might preventmultiplemyeloma progression by inhibitionof TNF-a–induced NF-kB activation. Finally, we did notshow any significant inhibitory effect of PBS-1086 on IL-6secretion within the bone marrowmilieu, even though NF-kB is involved in the transcriptional regulation of IL-6expression in multiple myeloma cells (8). Recently, someauthors have reported constitutive expression of IL-6 regu-lated by several transcription factors besidesNF-kB, with norequirement for NF-kB binding activity to DNA (42, 43).The effect of PBS-1086, at least on paracrine IL-6 secretion,requires further investigation.

The role of NF-kB signaling in bone pathogenesis isprimarily via RANKL/RANK-induced activation of NF-kBpathway (12).Mice deficient in both p50 and p52 subunits,which are deficient in total NF-kB activity, develop severeosteopetrosis due to the absence of osteoclast formation(44–46). In contrast, deletion of either p50or p52 causes nodetectable bone phenotype, with intact osteoclast forma-tion (47). Dominant-negative mutant IKKb leads to inhi-bition of NF-kB canonical pathway and results in decreasedosteoclast differentiation (48). Here, we show that PBS-1086 inhibits both the differentiation and function ofosteoclast by inhibition of both canonical and noncanon-ical NF-kB pathways. The inhibitory effect on bone resorp-tion was more potent than on osteoclast differentiation,consistent with dual inhibition of NF-kB and more potentinhibition of the canonical pathway. As RANKL/RANKbinding can also activate other signaling pathways, includ-ing p38/MAPK (32), JNK, and ERK (49), we confirmed thatthe inhibitory effect of PBS-1086 on osteoclastogenesis wasspecific toNF-kB inhibition. Bortezomib inhibits osteoclas-togenesis through different pathways dependent on osteo-clast differentiation status, with inhibition of p38/MAPK atearly stages and inhibition of other pathways including NF-kB at later stages (49, 50). Therefore, we cannot exclude thatPBS-1086 might also indirectly affect NF-kB–independentpathways at earlier stages of osteoclastogenesis in vivo.






Importantly, besides its direct cytotoxicity on multiplemyeloma cells, PBS-1086 exerts an indirect inhibitory effecton the bone marrow milieu and bone matrix, therebydisrupting tumor–bone marrowmilieu interactions, whichcontribute to multiple myeloma progression.

In conclusion, our study shows that PBS-1086 is a prom-ising dual inhibitor of the canonical and noncanonical NF-kB pathways. Besides its potent and selective cytotoxicity onmultiple myeloma cells, PBS-1086 also targets the bonemarrowmicroenvironment andovercomes the proliferativeand antiapoptotic effects of BMSCs, associated with aninhibition of TNF-a–inducible NF-kB activation. In addi-tion, we show that PBS-1086 is synergistic with bortezomibagainst bortezomib-resistant multiple myeloma cells andpatient multiple myeloma cells refractory to bortezomib.Finally, PBS-1086 inhibits RANKL-induced osteoclastogen-esis, associated with an inhibition of NF-kB pathway. Ourdata therefore provide the framework for clinical evaluationof PBS-1086 in combination with bortezomib for the treat-ment of multiple myeloma and related bone lesions.

Disclosure of Potential Conflicts of InterestK. Bobb, J. Zhang, and J.Meshulam(as the VP andCOO) are employees of

Profectus and rel�MD, which produce PBS-1086. P.G. Richardson is aconsultant/advisory board member for Millennium, Celgene, and Johnson&Johnson.T.Hideshima is a consultant/advisoryboardmember forAcetylonPharmaceuticals. K.C. Anderson has ownership interest (including patents)in Acetylon as the Scientific Founder and is a consultant/advisory boardmember for Celgene, Millennium, Bristol Myers Squibb, Onyx, and Merck.No potential conflicts of interest were disclosed by the other authors.

Authors' ContributionsConception and design: C. Fabre, J. Zhang, J. Meshulam, Y.-T. Tai,T. Hideshima, K.C. AndersonDevelopment of methodology: C. Fabre, S.-Y. Kong, Y. Hu, J. Zhang,J. Meshulam, Y.-T. Tai, K.C. AndersonAcquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.): C. Fabre, N. Mimura, K. Bobb, S.-Y. Kong, J.Zhang, J. Meshulam, R.D. Carrasco, Y.-T. Tai, P.G. Richardson, T. HideshimaAnalysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis):C. Fabre, N.Mimura, K. Bobb, S.-Y. Kong,G. Gorgun, D. Cirstea, Y. Hu, H. Ohguchi, J. Zhang, J. Meshulam, R.D.Carrasco, T. Hideshima, K.C. AndersonWriting, review, and/or revision of themanuscript:C. Fabre, G. Gorgun,J. Zhang, P.G. Richardson, T. Hideshima, K.C. AndersonAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): C. Fabre, K. Bobb, D. Cirstea,J. Minami, H. Ohguchi, Y.-T. Tai, K.C. AndersonStudy supervision: J. Meshulam, K.C. AndersonConduct immunohistochemical studies: R.D. CarrascoDesigned dosing regimen: J. Zhang, J. Meshulam

AcknowledgmentsThe authors thank Dharminder Chauhan, Catriona Hayes, Constantine

Mitsiades, and Loredana Santo for helpful suggestions.

The Editor-in-Chief of Clinical Cancer Research is an author of thisarticle. In keeping with the AACR’s Editorial Policy, a member of theAACR’s Publications Committee had the article reviewed independentlyof the journal’s review process and made the decision concerningacceptability.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received March 6, 2012; revised May 14, 2012; accepted July 10, 2012;published OnlineFirst July 17, 2012.

References1. Roodman GD. Pathogenesis of myeloma bone disease. Leukemia

2009;23:435–41.2. Richardson PG, Barlogie B, Berenson J, Singhal S, Jagannath S, Irwin

D, et al. Aphase 2studyof bortezomib in relapsed, refractorymyeloma.N Engl J Med 2003;348:2609–17.

3. Laubach J, Richardson P, Anderson K. Multiple myeloma. Annu RevMed 2011;62:249–64.

4. Hideshima T, Anderson KC. Molecular mechanisms of novel thera-peutic approaches for multiple myeloma. Nat Rev Cancer 2002;2:927–37.

5. Gilmore TD. Multiple myeloma: lusting for NF-kappaB. Cancer Cell2007;12:95–7.

6. Hideshima T, Chauhan D, Richardson P, Mitsiades C, Mitsiades N,Hayashi T, et al. NF-kappa B as a therapeutic target in multiplemyeloma. J Biol Chem 2002;277:16639–47.

7. Hideshima T, Neri P, Tassone P, Yasui H, Ishitsuka K, Raje N, et al.MLN120B, a novel IkappaB kinase beta inhibitor, blocks multiplemyeloma cell growth in vitro and in vivo. Clin Cancer Res 2006;12:5887–94.

8. Chauhan D, Uchiyama H, Akbarali Y, Urashima M, Yamamoto K,Libermann TA, et al. Multiple myeloma cell adhesion-induced inter-leukin-6 expression in bonemarrow stromal cells involves activation ofNF-kappa B. Blood 1996;87:1104–12.

9. Jourdan M, Moreaux J, Vos JD, Hose D, Mahtouk K, Abouladze M,et al. Targeting NF-kappaB pathway with an IKK2 inhibitor inducesinhibition of multiple myeloma cell growth. Br J Haematol 2007;138:160–8.

10. Landowski TH, Olashaw NE, Agrawal D, Dalton WS. Cell adhesion-mediated drug resistance (CAM-DR) is associated with activation ofNF-kappa B (RelB/p50) in myeloma cells. Oncogene 2003;22:2417–21.

11. Hideshima T, ChauhanD, SchlossmanR,RichardsonP, Anderson KC.The role of tumor necrosis factor alpha in the pathophysiology of

human multiple myeloma: therapeutic applications. Oncogene 2001;20:4519–27.

12. Boyle WJ, Simonet WS, Lacey DL. Osteoclast differentiation andactivation. Nature 2003;423:337–42.

13. Novack DV, Yin L, Hagen-Stapleton A, Schreiber RD, Goeddel DV,Ross FP, et al. The IkappaB function of NF-kappaB2 p100controls stimulated osteoclastogenesis. J Exp Med 2003;198:771–81.

14. Sun SC. Non-canonical NF-kappaB signaling pathway. Cell Res2011;21:71–85.

15. Annunziata CM, Davis RE, Demchenko Y, Bellamy W, Gabrea A, ZhanF, et al. Frequent engagement of the classical and alternative NF-kappaB pathways by diverse genetic abnormalities in multiple mye-loma. Cancer Cell 2007;12:115–30.

16. Keats JJ, Fonseca R, Chesi M, Schop R, Baker A, Chng WJ, et al.Promiscuous mutations activate the noncanonical NF-kappaBpathway in multiple myeloma. Cancer Cell 2007;12:131–44.

17. Demchenko YN,GlebovOK, ZingoneA, Keats JJ, Bergsagel PL, KuehlWM. Classical and/or alternative NF-kappaB pathway activation inmultiple myeloma. Blood 2010;115:3541–52.

18. Hideshima T, Chauhan D, Kiziltepe T, IkedaH, Okawa Y, Podar K, et al.Biologic sequelae of I{kappa}B kinase (IKK) inhibition in multiplemyeloma: therapeutic implications. Blood 2009;113:5228–36.

19. Hideshima T, Ikeda H, Chauhan D, Okawa Y, Raje N, Podar K, et al.Bortezomib induces canonical nuclear factor-kappaB activation inmultiple myeloma cells. Blood 2009;114:1046–52.

20. Li C, Chen S, Yue P, Deng X, Lonial S, Khuri FR, et al. Proteasomeinhibitor PS-341 (bortezomib) induces calpain-dependent IkappaB(alpha) degradation. J Biol Chem 2010;285:16096–104.

21. Oh U, McCormick MJ, Datta D, Turner RV, Bobb K, Monie DD, et al.Inhibition of immune activation by a novel nuclear factor-kappa Binhibitor in HTLV-I-associated neurologic disease. Blood 2011;117:3363–9.

Fabre et al.





22. Halleen JM, Alatalo SL, Suominen H, Cheng S, Janckila AJ, VaananenHK. Tartrate-resistant acid phosphatase 5b: a novel serum marker ofbone resorption. J Bone Miner Res 2000;15:1337–45.

23. HideshimaT,NakamuraN, ChauhanD, AndersonKC.Biologic sequel-ae of interleukin-6 induced PI3-K/Akt signaling in multiple myeloma.Oncogene 2001;20:5991–6000.

24. Catlett-Falcone R, Landowski TH, Oshiro MM, Turkson J, Levitzki A,Savino R, et al. Constitutive activation of Stat3 signaling confersresistance to apoptosis in human U266 myeloma cells. Immunity1999;10:105–15.

25. Ogata A, Chauhan D, Teoh G, Treon SP, UrashimaM, SchlossmanRL,et al. IL-6 triggers cell growth via the Ras-dependent mitogen-acti-vated protein kinase cascade. J Immunol 1997;159:2212–21.

26. Mitsiades CS, Mitsiades N, Poulaki V, Schlossman R, Akiyama M,Chauhan D, et al. Activation of NF-kappaB and upregulation of intra-cellular anti-apoptotic proteins via the IGF-1/Akt signaling in humanmultiple myeloma cells: therapeutic implications. Oncogene 2002;21:5673–83.

27. Hideshima T, Mitsiades C, Tonon G, Richardson PG, Anderson KC.Understanding multiple myeloma pathogenesis in the bone marrowto identify new therapeutic targets. Nat Rev Cancer 2007;7:585–98.

28. Tai YT, Podar K, Catley L, Tseng YH, AkiyamaM, Shringarpure R, et al.Insulin-like growth factor-1 induces adhesion and migration in humanmultiple myeloma cells via activation of beta1-integrin and phospha-tidylinositol 30-kinase/AKT signaling. Cancer Res 2003;63:5850–8.

29. Chauhan D, Li G, Shringarpure R, Podar K, Ohtake Y, Hideshima T,et al. Blockade of Hsp27 overcomes Bortezomib/proteasome inhibitorPS-341 resistance in lymphoma cells. Cancer Res 2003;63:6174–7.

30. Hideshima T, Podar K, Chauhan D, Ishitsuka K, Mitsiades C, Tai YT,et al. p38 MAPK inhibition enhances PS-341 (bortezomib)-inducedcytotoxicity against multiple myeloma cells. Oncogene 2004;23:8766–76.

31. Yasui H, Hideshima T, Ikeda H, Jin J, Ocio EM, Kiziltepe T, et al. BIRB796 enhances cytotoxicity triggeredbybortezomib, heat shockprotein(Hsp) 90 inhibitor, and dexamethasone via inhibition of p38 mitogen-activated protein kinase/Hsp27 pathway inmultiplemyeloma cell linesand inhibits paracrine tumour growth. Br J Haematol 2007;136:414–23.

32. Ishitsuka K, Hideshima T, Neri P, Vallet S, Shiraishi N, Okawa Y, et al.p38 mitogen-activated protein kinase inhibitor LY2228820 enhancesbortezomib-induced cytotoxicity and inhibits osteoclastogenesis inmultiple myeloma; therapeutic implications. Br J Haematol 2008;141:598–606.

33. MatsumotoM, Sudo T, Saito T, Osada H, Tsujimoto M. Involvement ofp38 mitogen-activated protein kinase signaling pathway in osteoclas-togenesis mediated by receptor activator of NF-kappa B ligand(RANKL). J Biol Chem 2000;275:31155–61.

34. Abe M, Hiura K, Wilde J, Shioyasono A, Moriyama K, Hashimoto T,et al. Osteoclasts enhance myeloma cell growth and survival via cell-cell contact: a vicious cycle between bone destruction and myelomaexpansion. Blood 2004;104:2484–91.

35. Ouk S, Liou ML, Liou HC. Direct Rel/NF-kappaB inhibitors: structuralbasis for mechanism of action. Future Med Chem 2009;1:1683–707.

36. Oerlemans R, Franke NE, Assaraf YG, Cloos J, van Zantwijk I, BerkersCR, et al. Molecular basis of bortezomib resistance: proteasomesubunit beta5 (PSMB5) gene mutation and overexpression of PSMB5protein. Blood 2008;112:2489–99.

37. Ri M, Iida S, Nakashima T, Miyazaki H, Mori F, Ito A, et al. Bortezomib-resistant myeloma cell lines: a role for mutated PSMB5 in preventingthe accumulation of unfolded proteins and fatal ER stress. Leukemia2010;24:1506–12.

38. McConkey DJ, Zhu K.Mechanisms of proteasome inhibitor action andresistance in cancer. Drug Resist Updat 2008;11:164–79.

39. Anderson KC. Targeted therapy of multiple myeloma based upontumor-microenvironmental interactions. Exp Hematol 2007;35 Suppl1:155–62.

40. Markovina S, Callander NS, O'Connor SL, Kim J, Werndli JE, RaschkoM, et al. Bortezomib-resistant nuclear factor-kappaB activity in mul-tiple myeloma cells. Mol Cancer Res 2008;6:1356–64.

41. Markovina S, Callander NS, O'Connor SL, Xu G, Shi Y, Leith CP, et al.Bone marrow stromal cells from multiple myeloma patients uniquelyinduce bortezomib resistant NF-kappaB activity in myeloma cells. MolCancer 2010;9:176.

42. Xiao W, Hodge DR, Wang L, Yang X, Zhang X, Farrar WL. NF-kappaBactivates IL-6 expression through cooperationwith c-Jun and IL6-AP1site, but is independent of its IL6-NFkappaB regulatory site in autocrinehuman multiple myeloma cells. Cancer Biol Ther 2004;3:1007–17.

43. XiaoW, HodgeDR,Wang L, Yang X, Zhang X, FarrarWL. Co-operativefunctions between nuclear factors NFkappaB and CCAT/enhancer-binding protein-beta (C/EBP-beta) regulate the IL-6 promoter in auto-crine human prostate cancer cells. Prostate 2004;61:354–70.

44. Franzoso G, Carlson L, Xing L, Poljak L, Shores EW, Brown KD, et al.Requirement for NF-kappaB in osteoclast and B-cell development.Genes Dev 1997;11:3482–96.

45. Zheng H, Yu X, Collin-Osdoby P, Osdoby P. RANKL stimulatesinducible nitric-oxide synthase expression and nitric oxide productionin developing osteoclasts. Anautocrine negative feedbackmechanismtriggered by RANKL-induced interferon-beta via NF-kappaB thatrestrains osteoclastogenesis and bone resorption. J Biol Chem2006;281:15809–20.

46. Iotsova V, Caamano J, Loy J, Yang Y, Lewin A, Bravo R. Osteopetrosisin mice lacking NF-kappaB1 and NF-kappaB2. Nat Med 1997;3:1285–9.

47. Xing L, Carlson L, Story B, Tai Z, Keng P, Siebenlist U, et al. Expressionof either NF-kappaB p50 or p52 in osteoclast precursors is required forIL-1-induced bone resorption. J Bone Miner Res 2003;18:260–9.

48. YamamotoA,Miyazaki T, KadonoY, Takayanagi H,Miura T,NishinaH,et al. Possible involvement of IkappaB kinase 2 and MKK7 in osteo-clastogenesis induced by receptor activator of nuclear factor kappaBligand. J Bone Miner Res 2002;17:612–21.

49. vonMetzler I, KrebbelH,HechtM,ManzRA, FleissnerC,MiethM, et al.Bortezomib inhibits human osteoclastogenesis. Leukemia 2007;21:2025–34.

50. Hongming H, Jian H. Bortezomib inhibits maturation and function ofosteoclasts from PBMCs of patients with multiple myeloma by down-regulating TRAF6. Leuk Res 2009;33:115–22.






Published OnlineFirst July 17, 2012.Clin Cancer Res Claire Fabre, Naoya Mimura, Kathryn Bobb, et al. Multiple MyelomaPathways Demonstrates Significant Antitumor Activities in

BκDual Inhibition of Canonical and Noncanonical NF-

Updated version

10.1158/1078-0432.CCR-12-0779doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2012/07/18/1078-0432.CCR-12-0779.DC1Access the most recent supplemental material at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://clincancerres.aacrjournals.org/content/early/2012/08/08/1078-0432.CCR-12-0779To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-12-0779

http://clincancerres.aacrjournals.org/content/suppl/2012/07/18/1078-0432.CCR-12-0779.DC1

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/early/2012/08/08/1078-0432.CCR-12-0779


Dual Inhibition of Canonical and Noncanonical NF-kB...

Documents

Transcript of Dual Inhibition of Canonical and Noncanonical NF-kB...