Phosphoinositide Protein Kinase PDPK1 Is a Crucial Cell ...Molecular and Cellular Pathobiology Phosphoinositide Protein Kinase PDPK1 Is a Crucial Cell Signaling Mediator in Multiple

Molecular and Cellular Pathobiology

Phosphoinositide Protein Kinase PDPK1 Is a Crucial CellSignaling Mediator in Multiple Myeloma

Yoshiaki Chinen1, Junya Kuroda1, Yuji Shimura1, Hisao Nagoshi1, Miki Kiyota1, Mio Yamamoto-Sugitani1,ShinsukeMizutani1, Natsumi Sakamoto1, Masaki Ri2, Eri Kawata3, TsutomuKobayashi1, YosukeMatsumoto1,Shigeo Horiike1, Shinsuke Iida2, and Masafumi Taniwaki1

AbstractMultiple myeloma is a cytogenetically/molecularly heterogeneous hematologic malignancy that remains

mostly incurable, and the identification of a universal and relevant therapeutic targetmolecule is essential for thefurther development of therapeutic strategy. Herein, we identified that 3-phosphoinositide–dependent proteinkinase 1 (PDPK1), a serine threonine kinase, is expressed and active in all eleven multiple myeloma–derived celllines examined regardless of the type of cytogenetic abnormality, the mutation state of RAS and FGFR3 genes, orthe activation state of ERK and AKT. Our results revealed that PDPK1 is a pivotal regulator of molecules that areessential for myelomagenesis, such as RSK2, AKT, c-MYC, IRF4, or cyclin Ds, and that PDPK1 inhibition causedthe growth inhibition and the induction of apoptosis with the activation of BIM and BAD, and augmented thein vitro cytotoxic effects of antimyeloma agents in myeloma cells. In the clinical setting, PDPK1 was active inmyeloma cells of approximately 90% of symptomatic patients at diagnosis, and the smaller population of patientswith multiple myeloma exhibiting myeloma cells without active PDPK1 showed a significantly less frequentproportion of the disease stage III by the International Staging System and a significantly more favorableprognosis, including the longer overall survival period and the longer progression-free survival period bybortezomib treatment, than patients with active PDPK1, suggesting that PDPK1 activation accelerates thedisease progression and the resistance to treatment in multiple myeloma. Our study demonstrates that PDPK1 isa potent and a universally targetable signaling mediator in multiple myeloma regardless of the types ofcytogenetic/molecular profiles. Cancer Res; 74(24); 7418–29. �2014 AACR.

IntroductionMultiplemyeloma is amolecularly/symptomatically hetero-

geneous plasma cell malignancy, and is the second mostcommon among hematologic malignancies, constituting 1%of all cancers (1, 2). The recent advent of novel agents, includingimmunomodulatory drugs, such as lenalidomide or pomali-domide, and proteasome inhibitors (PI), bortezomib (BTZ)or carfilzomib, has improved the response to therapy andprolonged the survival of patients with multiple myeloma, andhas thereby dramatically revolutionized the treatment strategyagainst multiple myeloma (3–9). However, multiple myeloma

remainsmostly incurable to date, and the further developmentof an innovative anti-MM therapy is urgently needed.

The emergence and progression of multiple myeloma arecytogenetic/molecular multistep processes. Initially, a myelo-ma-initiating cell acquires a primary cytogenetic abnormalityinvolving cyclin Ds (CCND), FGFR3/MMSET, or MAFs, whichconfers the ability of clonal proliferation and expansion. Dur-ing the long disease course, myeloma cells progressivelyacquire various additional cytogenetic, genetic, and/or epige-netic abnormalities that further augment the proliferativephenotype and the resistance to cytotoxic insults, includinganticancer agents. In addition to such intracellular oncogenicevents, myeloma cells are supported by tumor microenviron-ment components, including bone marrow (BM) stromal cells(BMSC), osteoclasts and soluble factors, such as IL6, IGF1, or B-cell activating factor belonging to the tumor necrosis factorfamily (BAFF), in BM. Many of the intrinsic oncogenic cellularevents and extrinsic stimuli converge to activate two majorsignaling pathways, the RAS/RAF/ERK/RSK2 pathway and thePI3K/AKT pathways, which share critical roles in myelomadevelopment and progression (1, 2). In myeloma cells, forinstance, the RAS/RAF/ERK/RSK2 pathway has frequentlybeen activated by gene mutations of RAS, BRAF, or FGFR3,and by chromosomal translocation involving FGFR3/MMSETorMAFs (1, 2, 10–12), whereas the PI3K/AKT pathway is active

1Division of Hematology and Oncology, Department of Medicine, KyotoPrefectural University of Medicine, Kyoto, Japan. 2Department of MedicalOncology and Immunology, Nagoya City University Graduate School ofMedical Sciences, Nagoya, Aichi, Japan. 3Department of Hematology,Japanese Red Cross Kyoto Daini Hospital, Kyoto, Japan.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

CorrespondingAuthor: JunyaKuroda,Division ofHematologyandOncol-ogy, Department of Medicine, Kyoto Prefectural University of Medicine,465 Kajii-cho, Kamigyo-ku, Kyoto 602-8566, Japan. Phone: 81-75-251-5740; Fax: 81-75-251-5743; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-14-1420

�2014 American Association for Cancer Research.

CancerResearch

Cancer Res; 74(24) December 15, 20147418

on May 14, 2020. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 30, 2014; DOI: 10.1158/0008-5472.CAN-14-1420

http://cancerres.aacrjournals.org/

in approximately half of patients with multiple myeloma, andAKT activity has been suggested to be mediated by DEPdomain-containing mTOR-interacting protein (DEPTOR;refs. 10, 13, 14). Moreover, the abnormal activation of thesetwo pathways is associated with drug resistance and a poorprognosis ofmultiplemyeloma (15–17). Nevertheless, previousstudies have demonstrated that a sole inhibition of either theRAS/MAPK/ERK or PI3K/AKT pathways was not sufficientlyefficacious to eradicate myeloma cells (10, 16, 18–21). Theinhibition of RAS is only effective against myeloma cells withmutated RAS, whereas the inhibition of AKT results in com-pensative activation of the RAS/RAF/ERKpathway inmyelomacells (16, 22–24). These multifactorial and overlapping molec-ular mechanisms underlying tumorigenesis serve to make themolecular architecture of myeloma cells highly complex andheterogeneous among patients with multiple myeloma as wellas even among subclones in a single patient, and thus, make iteven more difficult to identify a universally effective therapeu-tic target in multiple myeloma.We have recently identified that the N-terminal kinase

domain (NTKD) of RSK2, which is the most downstreamsignaling mediator of the RAS/ERK pathway in normaltissues, is universally activated thorough phosphorylation inmyeloma cells with diverse cytogenetic/molecular abnormal-ities, regardless of the activation state of ERK and AKT (25).Importantly, the constitutive activation of RSK2-NTKD isresponsible for the regulation of critical molecules for myelo-magenesis, including c-MYC or CCNDs; however, the under-lying mechanism for RSK2-NTKD in myeloma cells has notbeen identified (25). Because 3-phosphoinositide–dependentprotein kinase 1 (PDPK1), a serine/threonine kinase also calledPDK1, has been shown to be essential for the activationprocesses of both RSK2-NTKD and AKT in physiologic condi-tion (26, 27), in this study, we investigated the role of PDPK1 inmyelomagenesis and the potential of PDPK1 as a therapeutictarget molecule for overcoming the difficulty of treating mul-tiple myeloma due to their interpatient diversity and intraclo-nal molecular heterogeneity.

Materials and MethodsCells and reagentsThis study used human myeloma cell lines AMO-1,

NCI-H929, OPM-2, LP-1 (Deutsche Sammlung von Mikroorga-nismen und Zellkulturen GmbH), PRMI8226, and IM9 (ATCC),which were authenticated by mycoplasma detection, DNAfingerprinting, isozyme detection, and cell vitality detection.KMS-18, KMS-12-BM, KMS-28-PE, KMS-34, and KMS-11 werekind gifts from Dr. T. Ohtsuki (Kawasaki Medical School,Okayama, Japan), and AMU-MM1 was a kind gift from Dr. I.Hanamura (Aichi Medical School, Aichi, Japan). HS-5 is animmortalized human BMSC-derived cell line, which potentlysecretes G-CSF, GM-CSF, M-CSF, Kit ligand, MIP-1a, IL6, IL8,or IL11 (ATCC; ref. 28). OPM-2, KMS-11, and KMS-18 have beenshown to express variants of FGFR3 (K650E, Y373C, andG384D, respectively; ref. 25). Cells were maintained in RPMI-1640 containing 10% FCS, 2 mmol/L L-glutamate, and penicil-lin/streptomycin. Bortezomib-resistant subclones of OPM-2and KMS-11 (OPM-2/BTZ and KMS-11/BTZ, respectively)

were developed under continuous exposure to bortezomib(29). Studies using patient samples were approved by theInstitutional Review Board of Kyoto Prefectural University ofMedicine (Kyoto, Japan; RBMR-G-124-2), and patient-derivedsamples were collected with informed consent in accordancewith the Declaration of Helsinki. BM mononuclear cells werelabeled with anti-CD138 MicroBeads and positively isolatedwith the Mini-MACS separator (Miltenyi Biotec KK). Recom-binant human IL6, IGF1, and BAFF were purchased from R&DSystems. Isolated primary myeloma cells were cultured inconventional culture media supplemented with 50 ng/mL IL6.BX-912 and AR-12, PDPK1 inhibitors, were purchased fromSymansis NZ Limited. BAY11-7085, an NF-kB inhibitor, waspurchased from Calbiochem. Bortezomib, melphalan, andetoposide were provided by the Screening Committee ofAnticancer Drugs.

Assays for growth inhibition and apoptosisAssays for growth inhibition and apoptosis were performed

as described in the Supplementary Information.

Drug combination assaysCells were treated with various concentrations of bortezo-

mib, melphalan, etoposide, or BAY11-7085 alone or in combi-nation with BX-912 for 48 hours. The fractional effect con-centrations (i.e., a fractional effect of 0.25 equals a 25% growthinhibitory effect) and the combination index (CI) were calcu-lated with CalcuSyn (Biosoft). This method facilitates thequantification of synergism (CI < 1) and antagonism (CI >1) at different doses and effect levels. CI calculations wereconducted on the assumption that the drug mechanisms werenot mutually exclusive.

RNAiRNAi of PDPK1 was performed by transfecting shRNA

plasmid into KMS-18 cells. Briefly, 2.5� 106 KMS-18 cells wereresuspended in 100 mL of transfection buffer and transfectedwith shRNA plasmid (Qiagen) for PDPK1 or with controlshRNA plasmid by means of the CLB-Transfection Kit (LonzaAG) using protocol 9. Two different shRNA plasmids targetingPDPK1 were utilized independently.

Western blot analysisWestern blot analysis was conducted as described elsewhere

(25). The primary antibodies used are described in the Sup-plementary Information.

Assay for NF-kB activityNF-kB activity was assessed using NF-kB (p65) Transcrip-

tion Factor Assay Kit according to the manufacture's instruc-tion (Cayman Chemical Company).

IHC staining of phospho-PDPK1 in patient-derived BMclot sections and statistical analysis

Approval was obtained from the Institutional Review Boardof our institute for the investigation of patient-derived samplesand their medical records (RBMR-G-124-2). Formalin-fixedand paraffin-embedded tissues were examined by IHC staining

PDPK1 in Multiple Myeloma

www.aacrjournals.org Cancer Res; 74(24) December 15, 2014 7419




with an anti-phosphorylated (p-) PDPK1Ser241 monoclonal Ab.The activity of p-PDPK1 in myeloma cells was evaluated asnegative (�), no staining at all; positive (þ), moderate stainingin myeloma cells; or strong positive (þþ), strong staining. Thep-PDPK1 activity in myeloma cells was independently evalu-ated by three hematologists who were blinded to the patients'clinical records. The International Myeloma Working Groupcriteria were used for the assessment of treatment response,and the disease stage was assessed according to the interna-tional staging system (ISS). Progression-free survival (PFS) wasdefined as the period from the date of therapy initiation to thedate of thefirst assessment of disease progression or the date ofdeath in patients who did not exhibit disease progression. Allpatients below 65 years old received up-front high-dose mel-phalan (HDT) supported by autologous stem cell transplan-tation (ASCT; HDT/ASCT) following induction chemotherapy.Overall survival (OS) was calculated from the date of diagnosisto the last date of patient follow-up or the date of death fromany cause. OS and PFS were estimated using the Kaplan–Meiermethod. The confidence interval was 95% for all analyses and P< 0.05 was considered to be statistically significant. The log-rank test was used to compare differences between survivalcurves.

Analysis of NRAS, KRAS, and BRAF mutationsThe cDNA from themyeloma cell lines was amplified by PCR

with primers spanning themutation hotspot codons 12, 13, and61 for NRAS and KRAS, and codon 600 for BRAF. Analyses ofPCR products were conducted as described in the Supplemen-tary Information.

FISHCytogenetic abnormalities were examined by means of

interphase FISH analysis (30). The probes used are describedin the Supplementary Information.

ResultsPDPK1 and RSK2-NTKD are constitutively activatedthrough phosphorylation in multiple myeloma–derivedcell lines

We first examined the activation state of PDPK1 in associ-ation with the activity of ERK, RSK2, and AKT in 11 myeloma-derived cell lines. Phosphorylated (p)-PDPK1Ser241 was used asa surrogate marker for the active confirmation. We simulta-neously examined the expression of FGFR3 and DEPTOR,cytogenetic abnormalities that are highly associated withmultiple myeloma, and the gene mutation status of NRAS,KRAS, and BRAF in all cell lines examined. As shown in Fig. 1,seven of 11 cell lines exhibited chromosomal translocationt(4;14) involving FGFR3/MMSET, whereas both t(11;14) involv-ing CCND1 and t(14;16) involving c-MAFwere identified in onecell line each. Bothmutations ofNRAS andKRASwere detectedin three cell lines each, and their existence was mutuallyexclusive. None of the cell lines possessed the BRAFV600Emutation. PDPK1Ser241 and RSK2Ser227-NTKD were phosphor-ylated in all 11 multiple myeloma–derived cell lines examined,whereas ERK, AKT, and RSK2Tyr529-C-terminal kinase domain(CTKD) were activated by phosphorylation in 9, 6, and 3 of 11

multiple myeloma–derived cell lines, respectively (Fig. 1).These results indicated that both PDPK1 and RSK2-NTKD areuniversally activated through phosphorylation in all 11 mul-tiple myeloma–derived cell lines examined, regardless of thetype of cytogenetic abnormalities or the status of upstreamsignaling molecules, and that myeloma cell lines can be largelysubdivided into two types, namely, those that are active forPDPK1, RSK2-NTKD, and AKT and those that are active forPDPK1 and RSK2-NTKD, but inactive for AKT. In addition, allseven cell lines with NRAS, KRAS, or FGFR3 mutations wereshown to possess p-ERK, while we found no positive correla-tion between the expression of DEPTOR and p-AKT status inour series.

PDPK1 inhibition exerts anti-MM effect via G2–M cell-cycle blockade and the induction of apoptosis

Next, we investigated the effects of PDPK1 inhibition onmyeloma-derived cell lines using small-molecule inhibitors forPDPK1, BX-912, and AR-12. Although BX-912 is a highly selec-tive inhibitor for PDPK1 (27, 31), AR-12 inhibits not onlyPDPK1, but also other off target molecules, such as COX-2(32). Forty-eight–hour exposure of either BX-912 or AR-12resulted in the time- and dose-dependent growth inhibitionin various myeloma cell lines (Supplementary Fig. S1). The IC50

values for the 11 myeloma cell lines were 2.5–12.8 mmol/L withBX-912, compared with 3.0 to 12.5 mmol/L with AR-12. Thegrowth inhibitory effects of BX-912 and AR-12 were not sig-nificantly affected by the p-AKT state of myeloma cells (Fig. 2Aand B). To further explore the mechanisms underlying thegrowth inhibition by PDPK1 inhibition, AMO-1 cells (repre-sentative of myeloma cells with active RSK2-NTKD but lackingAKT activity) and NCI-H929 cells (representative of myelomacells with active RSK2-NTKD and AKT) were subjected to flow-cytometric analyses. As the results, treatment with the PDPK1-specific inhibitor, BX-912 or AR-12, caused G2–M cell-cycleblockade and the induction of apoptosis in both AMO-1 cellsand NCI-H929 cells (Fig. 2C andD; Supplementary Fig. S2A andS2B).

Molecular sequelae following PDPK1 inhibition inmultiple myeloma cell lines

We next analyzed the molecular sequelae following p-PDPK1 inhibition in AMO-1 and NCI-H929 cells. The block-ade of p-PDPK1 by BX-912 caused dephosphorylation ofRSK2-NTKD in AMO-1 cells and dephosphorylation of bothRSK2-NTKD and AKT in NCI-H929 cells (Fig. 3A). We nextexamined the effects of PDPK1 inhibition on the down-stream target molecules of RSK2 and AKT. As shownin Fig. 3B, inhibition of PDPK1 caused the downregulationof c-MYC, IRF4, and CCNDs, the activation of proapoptoticBH3-only proteins, namely induction of BIMEL and BIML anddephosphorylation of BAD, and the inhibition of PLK1through dephosphorylation.

We further validated the effect of PDPK1 inhibition bygene knockdown of PDPK1 on myeloma cells. Two differentshRNA vectors were utilized independently to knockdownPDPK1. Regardless of the type of shRNA vector for PDPK1utilized, transient partial knockdown of PDPK1 expression

Chinen et al.

Cancer Res; 74(24) December 15, 2014 Cancer Research7420




caused the downregulation of c-MYC, IRF4, and CCNDs andthe upregulation of BIMEL and BIML in the growth inhibitorystage in KMS-18 cells. As the result, the knockdown ofPDPK1 was accompanied by the accumulation of cleavedform of caspase-3 and growth inhibition, indicating thatPDPK1 knockdown caused the growth inhibition at leastpartly via the induction of apoptosis (Fig. 4A and B). Incontrast, the transfection of control RNAi vector showed noeffect on the expression and/or the activities of PDPK1, aswell as the respective downstream molecules, or on theproliferation of KMS-18 cells. These indicate that PDPK1 isa crucial regulator of both RSK2-NTKD and AKT, and,thereby, regulates various molecules essential for diseasedevelopment, cell cycling, cell proliferation, and survival ofmultiple myeloma (1, 2, 33, 34).

PDPK1 activity is critical for the viability of patient-derived primary myeloma cells and the long-termoutcome of multiple myelomaCD138-positive primary myeloma cells isolated from 17 BM

samples from patients with multiple myeloma were availablefor the protein preparation forWestern blot analyses of PDPK1,RSK2, and AKT, and 11 of those 17 samples were also availablefor cytotoxicity assays by BX-912. Seventeen patients includedfour asymptomatic, so-called smoldering, multiple myeloma(aMM), four symptomatic newly diagnosed multiple myeloma(NDMM), and nine refractory/relapsed multiple myeloma(RRMM; Supplementary Table S1). PDPK1 was active in all17 primary myeloma cell populations regardless of the diseasestate, whereas RSK2-NTKD and AKT were active in all NDMM

and RRMMand in only two (Pts. 2 and 3) of four aMM (Fig. 5A).In addition, BX-912 treatment resulted in the reduced numbersof viable myeloma cells from 11 patients at concentrationssimilar to those used for the multiple myeloma–derived celllines (Fig. 5B). Because untreated primarymyeloma cells couldsurvive but did not proliferate significantly in IL6 containingmedia in our system, our results indicate that the exposure toBX-912 induced cell death in patient-derived primarymyelomacells.

We also retrospectively evaluated the association betweenPDPK1 activity and treatment outcome of multiple myelomaby IHC staining of p-PDPK1Ser241 in formalin-fixed and paraf-fin-embedded BM clot specimens from 65 symptomaticNDMM patients that were consecutively diagnosed at ourinstitute from January 2005 to August 2013. The median ageof the 65 patients analyzed was 66 years (range, 14–88). Thepatients' backgrounds, including gender, M-protein type, anddisease stage according to the ISS, are summarized in Table 1.Thirty-six patients were also evaluated for chromosomalabnormalities, including 13q-, t(4;14), and t(11;14). Forty-sixpatients had a history of treatment with bortezomib plusdexamethasone (BD), 31 had undergone treatment with lena-lidomide and weekly dexamethasone (Rd), and 31 had beentreatedwithHDT/ASCT. Among the 65 patients, 8 (12.3%)werep-PDPK1 negative [PDPK1(�); i.e., PDPK1 inactive], whereas57 (87.3%) were positive for p-PDPK1, including 45 patientswith myeloma cells exhibiting moderately active p-PDPK1[PDPK1(þ)] and 12 (18.5% of 65) patients with rather higheractivity of p-PDPK1 [PDPK1(þþ)] in their myeloma cells. In asingle patient from among the PDPK1(þ) and PDPK1(þþ)

t(4;14)

t(11;14)

t(14;16)

t(8;14)

13q-

1q21+

NRAS mutation

KRAS mutation

FGFR3 mutation

FGFR3

ERK1/2

p-ERK1/2Thr202/Tyr204

RSK2

p-RSK2Tyr529

p-RSK2Ser227

PDPK1

p-PDPK1Ser241

AKT

p-AKTThr308

DEPTOR

AMO-1KMS-18

NCI-H929

RPMI8226

LP-1 OPM-2KMS-

12-BMKMS-28-PE

KMS-34

IM9AMU-MM1

Normal lymphocyte

ββ-Actin

G12A G12A Q61R

G13DG13D Q61K

G384D K650E

Figure 1. Phosphorylated (p)-PDPK1Ser241 is active in multiplemyeloma cell lines. Expression andactivity of PDPK1, RSK2, FGFR3,AKT, ERK1/2, and DEPTOR wereexamined by Western blotting in11 myeloma-derived cell lines. Themyeloma-related chromosomalabnormalities andNRAS, KRAS, orFGFR3 gene mutations are alsoshown (26). Normal peripherallymphocytes were used as thecontrol.






cohorts, the staining intensity of p-PDPK1 was similar amongmyeloma cells (Fig. 5C). Patients with the disease stage IIIaccording to ISS were significantly more frequent in thePDPK1(þ) and/or PDPK1(þþ) cohorts compared withPDPK1(�) cohort, while patients' backgrounds, such as age,gender, or paraprotein type, were not significantly differentamong the PDPK1(�), PDPK1(þ), and PDPK1(þþ) cohorts(Table 1). Although the overall response rates for BD or HDT/ASCT were not significantly different among the PDPK1(�),PDPK1(þ), and PDPK1(þþ) cohorts, all PDPK1(�) patientsattained at least a partial response with some kind of therapy.With a median follow-up period of 1,310 days, the median OSperiod of the smaller population of PDPK1(�) cohort wassignificantly longer than that of the PDPK1(þþ) cohort (2,925

days vs. 1,694 days, P¼ 0.012), whereas those of the PDPK1(þ)and PDPK1(þþ) cohorts were not significantly different. Inaddition, the 8-year OS rate was significantly higher in thePDPK1(�) cohort compared with the PDPK1(þ) plus PDPK1(þþ) cohorts (83% vs. 17%; P ¼ 0.041). In response to BD, themedian PFS period of the PDPK1(�) cohort was significantlylonger than the PDPK1(þ) or PDPK1(þ) plus PDPK1(þþ)cohorts (not reached vs. 378 days; P ¼ 0.049 vs. 167 days; P ¼0.049). With HDT/ASCT, the median PFS period of the PDPK1(�) cohort was significantly longer than that of the PDPK1(þþ) cohort (972 days vs. 305 days; P ¼ 0.046; Fig. 5D). Theimpact of PDPK1 activity on the long-term outcome with Rdwas not assessable by statistical analysis, because only 2patients were treated by Rd in the PDPK1(�) cohort.

A

Via

ble

cell

num

ber

(rel

ativ

e to

unt

reat

ed c

ells

)

p-AKT-negative

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20

AMO-1LP-1KMS-12-BMIM9AMU-MM1

BX912 concentration (μmol/L)

p-AKT-positive

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20

KMS-18NCI-H929RPMI8226OPM-2KMS-28-PEKMS-34

BX912 concentration(μmol/L)

B

Via

ble

cell

num

ber

(rel

ativ

e to

unt

reat

ed c

ells

)

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20

AMO-1LP-1KMS-12-BMIM9AMU-MM1

p-AKT-negative

AR12 concentration(μmol/L)

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20

KMS-18NCI-H929RPMI8226OPM-2KMS-28-PEKMS-34

p-AKT-positive

AR12 concentration(μmol/L)

G1 = 2.6

S = 28.8G

1 = 41.4

S = 37.8G

2 = 12.4

BX12 hVehicle

12.3%

G1 = 1.0

S = 25.4G

2 = 19.1

BX 24 h

44.5%14.4%

6.2%

G1 = 31.5

S = 37.0G

2 = 28.6

28.3%

BX 12 hVehicle

51.1%

G1 = 0.7

S = 25.5G

2 = 19.5

BX 24 hG

1 = 4.7

S = 35.2G

2 = 26.8

AMO-1

NCI-H929

C

Vehicle

Vehicle

BX 6 h

BX 6 h

BX 12 h

BX 12 h

10.60.5

12.276.7

13.61.9

8.476.1

11.51.2

18.269.2

14.91.3

28.155.7

15.81.8

32.150.4

18.51.2

22.957.5

AMO-1

NCI-H929

D

G2 = 43.6

Figure 2. Effects of PDPK1 inhibition on myeloma-derived cell lines. A and B, myeloma cell lines were treated with various concentrations of BX-912 (A) orAR-12 (B) for 48 hours and cell proliferation was determined by themodifiedMTT assay. Results are expressed as themean of triplicate experiments. The IC50

values of BX-912 for AMO-1, KMS-18, NCI-H929, RPMI8226, LP-1, OPM-2, KMS-12-BM, KMS-28-PE, KMS-34, IM9, and AMU-MM1 were 3.0, 12.4, 6.6,12.8, 12.3, 4.0, 8.9, 5.9, 8.5, 2.5, and 4.8 mmol/L, respectively, whereas those of AR-12 for AMO-1, KMS-18, NCI-H929, RPMI8226, LP-1, OPM-2, KMS-12-BM, KMS-28-PE, KMS-34, IM9, and AMU-MM1 were 3.0, 11.0, 2.9, 5.9, 12.5, 3.9, 6.4, 12.5, 7.4, 3.3, and 4.3 mmol/L, respectively. C, DNA content wasexamined by propidium iodide (PI)-staining of formalin-fixed cells using flow-cytometric analysis. The x-axis represents DNA content and the y-axisrepresents the relative cell numbers. BX-912 treatment induced G2–M cell-cycle arrest (light blue fraction) and increased the sub-G1 fraction, a hallmark ofapoptosis induction (indicated by arrow), in the myeloma AMO-1 and NCI-H929 cell lines. AMO-1 and NCI-H929 cells were treated with BX-912 at theirIC80 concentrations (4.8 mmol/L for AMO-1 and 12.0 mmol/L for NCI-H929) for 12 and 24 hours. D, AMO-1 and NCI-H929 cells were treated with 4.8 and12.0 mmol/L, respectively, of BX-912 for 6 to 12 hours, and were subjected to double staining for Annexin-V (x-axis) and PI (y-axis). Fraction I, Annexin-V(AV; �)/PI (�; viable cells); fraction II, AV (þ)/PI (�; cells undergoing early apoptosis); fraction III, AV(þ)/PI (þ; cells undergoing late apoptosis); and fractionIV, AV (�)/PI (þ) populations (necrotic cells). BX-912 treatment caused the increase of fraction II and then fraction III, indicating that BX-912 induced apoptosisin both AMO-1 cells and NCI-H929 cells.

Chinen et al.





The clinical implication of targeting p-PDPK1 inthe anti-MM treatmentWeexamined the utility of the targeting p-PDPK1 in the anti-

MM therapy. First, we examined the interaction betweenPDPK1 activity and the stimuli from cell extrinsic tumor

microenvironment components, that is, soluble factors IL6(100 ng/mL), BAFF (200 ng/mL), IGF1 (100 ng/mL), and BMSC(35, 36). As shown in Fig. 6A, the addition of IL6, BAFF, or IGF1mostly enhanced the phosphorylation states of RSK2 and AKT,but not PDPK1, in both AMO1 cells (RSK2 active, but AKT

AMO-1

NCI-H929

AKT

p-AKTThr308

PDPK1

p-PDPK1Ser241

RSK2

p-RSK2Ser227

AKT

p-AKTThr308

PDPK1

p-PDPK1Ser241

RSK2

p-RSK2Ser227

A12 h6 hCtl.

AMO-1 NCI-H929 KMS-3412 h6 hCtl.12 h6 hCtl.12 h6 hCtl.

B

c-MYC

CCND1

CCND2

BAD

p-BADSer112

BIMEL

BIML

PLK1

p-PLK1Thr210

β-Actin

IRF4

Figure 3. Molecular sequelae following PDPK1 inhibition by BX-912 in myeloma cells. AMO-1, NCI-H929, and KMS-34 cells were treated with BX-912 at theirIC80 concentrations (4.8 mmol/L for AMO-1; 12.0 mmol/L for NCI-H929; and 12.0 mmol/L for KMS-34) for 6 to 12 hours (h), respectively. A, PDPK1 inhibition ledto dephosphorylation of bothRSK2Ser227 andAKTThr308 in NCI-H929 cells (activeRSK2-NTKDandAKT), and dephosphorylation of RSK2Ser227 in AMO-1 cells(active RSK2-NTKD but inactive AKT). B, PDPK1 inhibition resulted in the reduction of c-MYC, IRF4, and CCNDs, induction of BIM and the activation of BADthrough dephosphorylation, and inactivation of PLK1 through dephosphorylation in all three multiple myeloma cell lines examined. Ctl., untreated control.

Days

0

2

4

6

8

10

54321

BA

Via

ble

cell

num

ber

( × 1

05

cells

/mL)

BIMEL

p-RSK2Ser227

PDPK1

CCND1

MYC

p-AKTThy308

ββ-Actin

p-PDPK1Ser241

CCND2

IRF4

RSK2

AKT

BIML

Cleaved caspase-3

1.0

KMS-1

8/ctl

.

KMS-1

8/sh

-1

KMS-1

8/sh

-2

0.56 0.60

1.0 0.73 0.63

1.0 0.80 1.05

1.0 0.78 0.71

1.0 0.84 0.73

1.0 0.50 0.77

1.0 0.60 0.37

1.0 0.55 0.23

1.0 0.15 0.22

1.0 0.37 0.49

1.0 1.54 1.43

1.0 1.85 2.73

Figure 4. Gene knockdown ofPDPK1 by RNAi. A, knockdown ofPDPK1 caused dephosphorylationof phospho (p)-RSK2Ser227 andp-AKTThr308, downregulation ofMYC, IRF4, CCND1, and CCND2,and induction of BIML in KMS-18cells. Two different shRNA vectors(sh-1 and sh-2) for PDPK1 wereexamined. Expression levels oftarget proteins relative to thosetreated with control RNAi vectorwere calculated with ImageJsoftware and are described beloweach band. B, cell proliferation ofKMS-18 cells was inhibited bytransfection of sh-1 againstPDPK1, but not by the controlvector. Cells were stained byTrypan blue, and viable cellnumbers were counted under theinverted microscope. Solid linesrepresent the growth of mock-transfected cells and dotted linesrepresent that of cells transfectedwith the sh-1 for PDPK1. Triplicateexperiments were carried out foreach condition, and cell counts areexpressed as the mean � SD.






inactive) andNCI-H929 cells (bothRSK2 andAKTactive), whilethe enhanced phosphorylation of RSK2 or AKT by solublefactors was inhibited by BX-912. These indicate that PDPK1phosphorylation is independent of soluble factors, while is theprerequisite for the activation of RSK2 and AKT by soluble

factors. In addition, BX-912 failed in the complete dephos-phorylation of PDPK1 and its downstream kinase(s) underIGF1 stimulation, suggesting that IGF1 may have the strongerpotency in enhancing PDPK1-mediated signaling, comparedwith IL6 or BAFF (Fig. 6A). In contrast, the coculture of

PDPK1

p-PDPK1Ser241

p-RSK2Ser227

p-AKT Thr308

β-Actin

APt. 1 2 3 4 5 6 7 8 N.L. 9 10 11 12 13 14 15 16 17

aMM RRMMNDMM

B

BX-912 concentration (μmol/L)

Via

ble

cell

num

ber

(rel

ativ

e to

unt

reat

ed c

ells

)

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 15 20

N.L. Pt.2 Pt.3Pt.4 Pt.5 Pt.6Pt.7 Pt.8 Pt.10Pt.12 Pt.16 Pt.17

p-PDK1(–) myeloma cells

p-PDK1(+) myeloma cells

p-PDK1(++) myeloma cells

C

D

0 500 1,000 1,500

0.0

0.2

0.4

0.6

0.8

1.0

Days0 100 200 300 400 500 600 700

0.0

0.2

0.4

0.6

0.8

1.0

Days0 500 1,000 1,500 2,000 2,500 3,000

0.0

0.2

0.4

0.6

0.8

1.0

Days

0 500 1,000 1,500

0.0

0.2

0.4

0.6

0.8

1.0

Days0 100 200 300 400 500 600 700

0.0

0.2

0.4

0.6

0.8

1.0

Days0 500 1,000 1,500 2,000 2,500 3,000

0.0

0.2

0.4

0.6

0.8

1.0

Days

PFS with HDT/ASCTPFS with BDOS

p-PDK1 (–)

p-PDK1 (–)

p-PDK1 (+) plus (++)

p-PDK1 (–)

OS PFS with BD PFS with HDT/ASCT

p-PDK1 (+)

p-PDK1 (++)

p-PDK1 (–)

p-PDK1 (+)

p-PDK1 (++)

p-PDK1 (–)

p-PDK1 (+)

p-PDK1 (++)

p-PDK1 (–)

p-PDK1 (+) plus (++) p-PDK1 (+) plus (++)

(%) (%) (%)

(%) (%) (%)

Figure 5. Clinical significance of PDPK1 activity in multiple myeloma. A, Western blot analyses of the phosphorylation state of PDPK1Ser241, RSK2Ser227, andAKTThr308 in CD138-positive patient-derived primarymyeloma cells fromaMMpatients (Pt. 1–4), symptomatic NDMM (Pt. 5–8), andRRMMpatients (Pt. 9–17;Supplementary Table S1). Normal peripheral lymphocytes (N.L.) were utilized as a negative control. B, in vitro cytotoxic effect of BX-912 on patient-derivedprimary myeloma cells. Myeloma cells were exposed to various concentrations of BX-912 in IL-6–containingmedium for 48 hours and the relative numbers ofviable cellsweredeterminedby themodifiedMTTassay. The IC50 valueswere 11.3 (Pt.2), 2.2 (Pt. 3), 8.6 (Pt. 4), 7.9 (Pt. 5), 8.9 (Pt. 6), 6.1 (Pt. 7), 6.5 (Pt. 8), 1.9 (Pt.10), 3.6 (Pt. 12), 2.6 (Pt. 16), 4.1 (Pt. 17), and 29.3 mmol/L (normal lymphocytes). Results are expressed as themean of triplicate experiments. C, IHC staining ofphospho (p)-PDPK1Ser241 of patient-derived myelomatous BM clot specimens. The top panel, the middle panel, and the bottom panel are representativefigures of five patients with p-PDPK1(�), p-PDPK1(þ), and p-PDPK1(þþ) myeloma cells, respectively. Small windows of each figure are the higher-magnification views of the myeloma cells. D, long-term outcomes of patients with multiple myeloma according to p-PDPK1Ser241 status. Two cohorts ofp-PDPK1(�; dotted lines) and p-PDPK1(þ) plus p-PDPK1(þþ; solid lines) were compared in the top panels, whereas three cohorts of p-PDPK1(�; largerdotted lines), p-PDPK1(þ; solid lines), and p-PDPK1(þþ; smaller dotted lines) were compared in the bottom panels.

Chinen et al.





Tab

le1.

Patient

bac

kgroun

dan

dtrea

tmen

tou

tcom

ein

asso

ciationwith

PDPK1S

er241pho

spho

rylatio

n,as

asse

ssed

bytheIHC

staining

Total

PDPK1(�

)PDPK1(þ

)plus(þ

þ)

P(�

)vs

.(þ

)plus(þ

þ)PDPK1(þ

)P

(�)v

s.(þ

)PDPK1(þ

þ)P

(�)vs

.(þþ

)

Patient

number

658

5745

12

Age

,med

ian(ra

nge)

66(14–

88)

63(51–

82)

66(14–

88)

NS

64(14–

88)

NS

71(62–

83)

NS

Gen

der;M

:F32

:33

6:2

26:31

NS

19:26

NS

7:5

NS

M-protein;

Ig-G

/A/BJP

/others

32/18/12

/33/3/2/0

29/15/10

/3NS

25/11/6/3

NS

4/4/4/0

NS

ISS

I/II/III/NA

21/20/23

/12/5/0/1

19/15/23

/0NS

16/11/18

/0NS

3/4/5/0

NS

IþII/III

41/23

7/0

34/23

0.03

627

/18

0.03

97/5

0.04

7

FISH

13q-/t(4

;14)/t(11;14

)16

/12/8

2/1/1

14/11/7

NS

10/10/7

NS

4/1/0

NS

OS

Med

ian,

day

2155

2925

2,15

5NS

2155

NS

1694

0.01

24-yrate

(%)

100%

67%

NS

68%

NS

63%

NS

8-yrate

(%)

83%

17%

0.04

125

%NS

0%NS

Res

pon

seto

BD

(N¼

46)

CR/VGPR/PR/SD/PD

4/11

/17/10

/42/0/3/0/0

[40/0/60

/0/0

(%)]

2/11

/14/10

/4[4.9/26.8/34

.1/24.4/9.8(%

)]NS

2/7/12

/8/3

NS

0/4/2/2/1

NS

PFS

with

BD

Med

ian,

day

378

Not

reac

hed

167

0.04

937

80.04

999

NS

0.5-yrate

(%)

100%

49%

NS

51%

NS

44%

NS

1-yrate

(%)

100%

49%

NS

44%

NS

44%

NS

Res

pon

seto

Rd(N

¼31

)CR/VGPR/PR/SD/PD

10/3/11/6/0

1/1/0/0/0

[50/50

/0/0/0

(%)]

9/3/11

/4/2

[31/10

.3/38/13

.8/6.9

(%)]

NS

8/2/8/4/1

NS

1/1/3/0/1

NS

Res

pon

seto

HDT/ASCT

(N¼

31)C

R/VGPR/PR/SD/PD

13/4/8/2/0

1/0/5/0/0

[16.7/0/83

.3/0/0

(%)]

12/4/7/2/0

[48/16

/28/8/0(%

)]NS

10/4/7/1/0

NS

2/0/0/1/0

NS

PFS

with

ASCT

Med

ian,

day

681

972

567

NS

567

NS

305

0.04

61-yrate

(%)

83%

59%

NS

64%

NS

33%

NS

2-yrate

(%)

83%

37%

NS

37%

NS

33%

NS

Abbreviations

:M,m

ale;F,

female;Ig,immun

oglobulin;B

JP,B

ence

Jone

sProtein;B

D,b

ortezo

mibplusdex

ametha

sone

;CR,com

plete

resp

onse

;VGPR,verygo

odpartia

lres

pon

se;P

R,p

artia

lres

pon

se;S

D,

stab

ledisea

se;P

D,p

rogres

sive

disea

se;Rd,len

alidom

idepluswee

klydex

ametha

sone

;NA,n

otav

ailable;N

S,no

sign

ifica

ntdifferen

ce.






myeloma cells with HS-5 cells showed no prominent effects onthe phosphorylation status of PDPK1, RSK2, or AKT in ourhands (Supplementary Fig. S3).

We next examined the interaction between PDPK1-medi-ated signaling and NF-kB signaling, because the NF-kB path-way is frequently activated by various kinds of gene mutations,as well as soluble factors in tumor microenvironment, such asBAFF, and, thereby, plays critical roles in promoting cellsurvival and drug resistance in myeloma cells (1, 2, 17). Asshown in Fig. 6B, neither NF-kB inhibitor BAY11-7085 norbortezomib has inhibitory effect on PDPK1 activity. In addi-tion, although the basal NF-kB activity was much higher in

AMO-1 cells compared with that in NCI-H929 cells, BX-912 hasno inhibitory effect onNF-kB activity in both cell lines (Fig. 6C).These indicate that PDPK1-mediated signaling and NF-kBsignaling promote myeloma cell survival as independent path-ways. Given these results, we examined the combinatory effectsof BX-912 and various anticancer drugs, including BAY11-7085and bortezomib, in the NCI-H929 cells and the AMO-1 cells.The combined use of BX-912 and clinically available agents,such as melphalan, etoposide, or bortezomib, showed additiveor synergistic effects in both NCI-H929 cells and AMO-1 cells(Fig. 6D). We also examined the cytotoxic effects of BX-912on the bortezomib-resistant multiple myeloma cell lines,

0 0.2 0.4 0.6 0.8 1.00

0.5

1.0

1.5

2.0

Fractional effect

BX+NFkB - Algebraic estimate

CI ±

1.9

6 s.

d.

A

NCI-H929

AMO1 FBS-

FBS+ +IL-6ctl – + – + – ++BAFF +IGF-1

p-PDPK1Ser241

p-AKTThr308

p-RSK2Ser227

PDPK1

RSK2

AKTActin

Actin

1.0 1.1 0.1 1.0 0.30.11.01.2

1.0 1.5 0.0 1.5 0.00.01.30.6

1.0 1.0 0.2 2.7 0.30.21.21.0

1.0 0.8 0.0 0.9 0.40.10.81.5

1.0 2.0 0.1 2.7 0.70.11.31.0

1.0 2.3 0.0 3.2 0.50.01.21.1

AMO1 NCI-H929

p-PDPK1Ser241

PDPK1

Actin

0 5 10 20BAY11-7085 (μmol/L)

p-PDPK1Ser241

PDPK1

Actin

0 5 10 20BTZ (nmol/L)

0 5 10 20

0 5 10 20

FBS-

FBS+ +IL-6ctl – + – + – ++BAFF +IGF-1

p-PDPK1Ser241

p-AKTThr308

p-RSK2Ser227

PDPK1

RSK2

AKT

B

C

0

0.2

0.4

0.6

0.8

1

0 20 40

NCI-H929+BX-912NCI-H929+BAY11-7085AMO1+BX-912AMO1+BAY11-7085

Flu

ore

scen

ce (

O.D

. 450

nm

ol/L

)

Drug concentration (mmol/L)

D AMO1NCI-H929

0 0.2 0.4 0.6 0.8 1.00

0.5

1.0

1.5

Fractional effect

BX+NFkB - Algebraic estimate

CI ±

1.9

6 s.

d.

0 0.2 0.4 0.6 0.8 1.00

0.5

1.0

1.5

2.0

Fractional effect

BX+Vel - Algebraic estimate

CI ±

1.9

6 s.

d.

0 0.2 0.4 0.6 0.8 1.00

0.5

1.0

1.5

2.0

2.5

Fractional effect

BX+Vel - Algebraic estimate

CI ±

1.9

6 s.

d.

BX-912+ BAY11-7085

BX-912+ BTZ

0 0.2 0.4 0.6 0.8 1.00

2.0

4.0

6.0

Fractional effect

BX+MEL - Algebraic estimate

CI ±

1.9

6 s.

d.

0 0.2 0.4 0.6 0.8 1.00

0.5

1.0

1.5

2.0

2.5

Fractional effect

BX+MEL - Algebraic estimate

CI ±

1.9

6 s.

d.

0 0.2 0.4 0.6 0.8 1.00

2.0

4.0

6.0

Fractional effect

BX+ETP - Algebraic estimate

CI ±

1.9

6 s.

d.0 0.2 0.4 0.6 0.8 1.0

0

1.0

2.0

3.0

4.0

Fractional effect

BX+ETP - Algebraic estimate

CI ±

1.9

6 s.

d.

BX-912+ L-PAM

BX-912+ ETP

Figure 6. The utility of PDPK1 inhibition by BX-912 in the anti-MM therapy. A, AMO-1 cells and NCI-H929 cells were cultured with IL6 (100 ng/mL), BAFF(200 ng/mL), or IGF1 (100 ng/mL) in serum-starved (FBS-) media for 48 hours, and were further cultured for 3 hours with (þ) or without (�) BX-912. Solublefactors or BX-912 were not added in control sample (ctl). Intensities of phospho (p)-PDPK1Ser241, p-RSK2Ser227, and p-AKTThr308 relative to those total levelswere calculated with ImageJ software and described on individual bands. The activity of each kinase in serum-starved untreated control cells (ctl) wasconsidered to 1.0. B, AMO-1 cells and NCI-H929 cells were treated by various concentrations of either BAY11-7085 or bortezomib for 3 hours. The activity ofp-PDPK1Ser241 was examined by Western blotting. C, AMO-1 cells and NCI-H929 cells were treated by various concentrations of either BAY11-7085 or BX-912 for 3 hours. NF-kB activity was examined by ELISA. D, the combinatory effects of BX-912 and various anticancer agents (melphalan, etoposide,bortezomib, orBAY11-7085)wereexamined in twomultiplemyeloma–derived cell lines,NCI-H929andAMO-1.Cellswere treatedwith various concentrationsof agents for 48hours. x-axis, fractional effect concentrations; y-axis,CI. CI is expressedasCI�1.96SD.Shadedareas show thesynergistic or additive effectsof the two agents. In particular, areas below the bidirectional arrows indicate synergism of the two agents.

Chinen et al.





OPM-2/BTZ and KMS-11/BTZ cells (29). Although both OPM-2/BTZ and KMS-11/BTZ cells were highly resistant to borte-zomib compared with their respective parental cell lines, theyexhibited the same sensitivities to BX-912 as their respectiveparental cells (Supplementary Fig. S4A). Thus, we found nocross-resistance between the bortezomib and PDPK1 inhibi-tors. We also examined the effect of combinatory blockade ofPDPK1 and NF-kB in AMO-1 cells, NCI-H929 cells, and thebortezomib-resistant cells, using BX-912 and the anti-NF-kBagent, BAY11-7085. This combination therapy exhibited addi-tive effects against all multiple myeloma cell lines examined(Fig. 6D and Supplementary Fig. S4B).

DiscussionThe overlapping acquisitions of various cytogenetic/genetic

abnormalities as well as the stimulation by the tumor micro-environment are essential for the development and progres-sion of multiple myeloma. Because the types of cell intrinsiccytogenetic/genetic mutations and tumor environmental con-ditions varywidely among patients withmultiplemyeloma andeven among subclones in a single patient (1, 2), it is unlikelythat the therapeutic targeting of individual mutations or cellextrinsic stimulation would be equally effective for eliminatingtumor cells of all patients. For the development of a newstrategy to treat a broad range of patients with multiplemyeloma, it would be ideal to identify a molecule that isresponsible for the regulation of a signaling pathway uponwhich the stimuli by cell intrinsic and extrinsic abnormalitiesconverge. In this study, we demonstrated that PDPK1 isgenerally activated through phosphorylation in multiple mye-loma cell lines, regardless of their types of cytogenetic abnor-malities, the existence of RAS or FGFR3 mutations, or theactivation state of ERK, RSK2-CTKD, and AKT. In addition,PDPK1 was also found to be active in myeloma cells inapproximately 90% of NDMM patients. The phosphorylationof amino acid residue at serine 241 of PDPK1 is essential for theactivation of multiple AGC family protein kinases, such asRSK2, AKT, PAK, ILK, or PKC (26, 31, 37). Under physiologicconditions, PDPK1 is essential for B-cell development andsurvival (38, 39), whereas high PDPK1 activity is required forthe proliferation and survival of tumor cells, including putativeleukemia-initiating cells, and is also associated with diseaseprogression and metastasis in various solid tumors and acutemyeloid leukemias (40–42). However, its pathogenetic role hasnot been clarified in multiple myeloma.On the basis of data presented herein, we propose that

PDPK1 could be a rational therapeutic target molecule for abroad range of patients with multiple myeloma for the follow-ing reasons. First, PDPK1 regulates two pivotal cell signalingmolecules, RSK2-NTKD and AKT, and, thereby, controls theactivity of molecules that are essential for disease develop-ment, proliferation, and survival in myeloma cells. Forinstance, PDPK1 regulates the activity of PLK1, which may beassociated with cell division and the expression of c-MYC (43).Indeed, PDPK1 inhibition caused G2–M cell-cycle arrest inmyeloma cells. Furthermore, our data also revealed thatPDPK1 regulates the MYC/IRF4 axis, which is essential for

myeloma cell survival (44). Presumably, PDPK1 regulates theexpression of c-MYC and IRF4 inmyeloma cells, at least in partvia PLK1, RSK2, and AKT (25, 36, 43, 44). CCNDs, which arecritically important for myelomagenesis, are also regulated byPDPK1, presumably through RSK2-NTKD (1, 2, 25, 33), whereasproapoptotic BIM and BAD, which are essential for apoptosisof hematopoietic cells, are negatively regulated by PDPK1through the activation of ERK/RSK2 with or without AKTactivation. Presumably through the comprehensive regulationof those molecules critical for myeloma pathophysiology,PDPK1 inhibition exerted antiproliferative effects via theblockade of cell cycling and the induction of apoptosis in allprimary myeloma cells and myeloma-derived cell lines exam-ined, regardless of their molecular features. Second, PDPK1was generally active in primary myeloma cells from patientswith early-stage (aMM) to advanced-stage (RRMM) disease,and in multiple myeloma–derived cell lines, whereas RSK2-NTKD and AKT are also generally active in RRMM and in celllines, but were less active or inactive in myeloma cells fromsome aMM and NDMM patients in our series. These findingsimplicate that PDPK1 activation may be involved in thepathophysiology of multiple myeloma from the early diseasephase to the aggressive phase. Third, our IHC analysis revealedthat small numbers of patients with inactive PDPK1 hadsignificantly more favorable prognoses compared with thelarger population of PDPK1-active patients. Because theresponse rates to the anti-MM therapies were not significantlydifferent between the PDPK1(�) and the PDPK1(þ) plusPDPK1(þþ) cohorts, the shorter PFS and OS of the PDPK1-active cohorts suggest that PDPK1 promotes the earlier acqui-sition of resistance to therapy and the expansion of residualtumor cells. It would be anticipated that inhibition of PDPK1would improve the long-termoutcome of themajor populationof patients with multiple myeloma with active PDPK1 similarto that of the PDPK1-inactive cohort. Because the numbers ofpatients with multiple myeloma analyzed were limited in thisstudy, the clinical data presented here should be taken withcaution. However, considering the significant prognosticimpacts of PDPK1 activity in other cancers (40–42), the clinicalsignificance of PDPK1 activity in multiple myeloma also needsto be addressed in the larger cohorts in future. Finally, PDPK1inhibition exhibited additive and/or synergistic antimyelomaeffects when combined with several clinically availabledrugs. In addition, the cytotoxic effect of PDPK1 inhibitionwas not hampered by the acquired resistance to bortezomib.Because the prognoses of patients with multiple myelomawho have acquired resistance toward novel agents, especial-ly bortezomib, continue to be dismal, our results suggestingthat PDPK1 inhibition is useful for overcoming resistance tobortezomib are encouraging. It is also interesting that theinhibition of PDPK1 exerts an additive cytotoxic effect withthe inhibition of NF-kB on myeloma cells. Because nearly50% of multiple myeloma cell lines and most primarypatient-derived primary myeloma cells have been shown topossess high levels of NF-kB activity due to both cellintrinsic and extrinsic mechanisms, the combination ofinhibitors for NF-kB and PDPK1 may be an attractive futuretherapeutic strategy for multiple myeloma.






Although PDPK1 ismost likely to be an attractive therapeutictarget for multiple myeloma, it remains unknown whether ornot PDPK1 inhibition is safe in the clinical setting, such ashyperglycemia (45–47). Our study showed that normal lympho-cytes were much less sensitive to BX-912 compared withmyeloma cells, and PDPK1 inhibition led to no life-threateningdeleterious effect in several clinical trials for human andpreclinical studies with animals (40, 43, 47). For instance,previous study has shown that PDPK1 hypomorphic miceexpressing only 10% of the normal level of PDPK1 were viable,fertile, and exhibited no obvious harmful phenotype (48). Nev-ertheless, cautions for adverse events should be needed whenPDPK1 inhibitor is used for human. Although rare, unforesee-able severe adverse events, such as cardiac events, have beenoccasionally reported with various molecular targeted agents,whenutilized for humanpatients. It is expected todevelopmorepowerful and selective inhibitors for PDPK1 that can induceanti-MM effects at much lower concentrations without pro-found toxicities. In addition, the utility of PDPK1 activity as abiomarker for predicting prognosis may be of value, because10% of all patients with multiple myeloma with inactive PDPK1can be isolated as having the favorable prognostic "PDPK1-inactive variant" myeloma. Our study also provides a newresearch topic in multiple myeloma. PDPK1 potently activatesRSK2-NTKD and AKT. However, although PDPK1 and RSK2-NTKD are generally active in myeloma-derived cell lines, AKTwas only active in limited cell lines. This suggests that althoughRSK2-NTKD activity is largely regulated by PDPK1, someunknown mechanisms are also involved in AKT activity inmyeloma cells. Finally, the underlying cell intrinsic mechanismfor PDPK1 activation in myeloma cells remains to be verified inthis study. Recent studies have indicated that the deregulationofmiRNA-375 (miR375) associates with the deregulated expres-sion and activity of PDPK1 in various cancers (49). Moreover,our preliminary data coincide with the previous finding thatreported the decreased expression ofmiR375 in a largemajorityof patients with multiple myeloma regardless of their cyto-genetic features (50). We are currently investigating the func-

tional role of miR375 deregulation for PDPK1 activation inmyeloma cells.

In conclusion, PDPK1 is generally phosphorylated in mul-tiple myeloma, which is a highly molecularly heterogeneoushematologic disorder that continues to be incurable. Inacti-vation of PDPK1 is a promising and attractive therapeutictarget for multiple myeloma.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: Y. Chinen, J. Kuroda, M. Kiyota, Y. MatsumotoDevelopment of methodology: Y. Chinen, J. KurodaAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): Y. Chinen, J. Kuroda, Y. Shimura, H. Nagoshi,N. Sakamoto, M. Ri, E. KawataAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): Y. Chinen, J. Kuroda, Y. Shimura, H. Nagoshi,M. Yamamoto-Sugitani, N. Sakamoto, T. KobayashiWriting, review, and/or revision of the manuscript: Y. Chinen, J. Kuroda,H. Nagoshi, S. Horiike, S. IidaAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): Y. Chinen, J. Kuroda, Y. Shimura,N. Sakamoto, S. Horiike, S. IidaStudy supervision: Y. Chinen, S. Mizutani, M. Taniwaki

AcknowledgmentsThe authors thank Drs. C Shimazaki, Y Kobayashi, and S Kobayashi for their

scientific supports and Y. Yamane for her excellent technical assistance.

Grant SupportThisworkwas supported in part byGrants-in-Aid for Scientific Research from

the National Cancer Center Research and Development Fund of Japan (23-A-17),the Ministry of Education, Culture, Sports, Science and Technology of Japan(J. Kuroda and M. Taniwaki), and the Award in Aki's Memory from theInternational Myeloma Foundation (J. Kuroda). SCAD was supported by aGrant-in-Aid for Scientific Research on Innovative Areas, Scientific SupportPrograms for Cancer Research, from The Ministry of Education, Culture, Sports,Science and Technology, Japan.

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 May 12, 2014; revised August 22, 2014; accepted September 13, 2014;published OnlineFirst September 30, 2014.

References1. Kuehl WM, Bergsagel PL. Molecular pathogenesis of multiple mye-

loma and its premalignant precursor. J Clin Invest 2012;122:3456–63.

2. MorganGJ,Walker BA, Davies FE. The genetic architecture ofmultiplemyeloma. Nat Rev Cancer 2012;12:335–48.

3. Harousseau JL. Ten years of improvement in the management ofmultiple myeloma: 2000–2010. Clin Lymphoma Myeloma Leuk2010;10:424–42.

4. Gay F, Larocca A,Wijermans P, Cavallo F, Rossi D, SchaafsmaR, et al.Complete response correlates with long-term progression-free andoverall survival in elderly myeloma treated with novel agents: analysisof 1175 patients. Blood 2011;117:3025–31.

5. Kumar SK, Rajkumar SV, Dispenzieri A, Lacy MQ, Hayman SR, BuadiFK, et al. Improved survival in multiple myeloma and the impact ofnovel therapies. Blood 2008;111:2516–20.

6. Kuroda J, Shimura Y, Ohta K, Tanaka H, Shibayama H, Kosugi S, et al.Limited value of the International Staging System for predicting long-term outcome of transplant-ineligible, newly diagnosed, symptomaticmultiple myeloma in the era of novel agents. Int J Hematol 2014;99:441–9.

7. Kuhn DJ, Chen Q, Voorhees PM, Strader JS, Shenk KD, Sun CM, et al.Potent activity of carfilzomib, a novel, irreversible inhibitor of theubiquitin-proteasome pathway, against preclinical models of multiplemyeloma. Blood 2007;110:3281–90.

8. Lacy MQ, Allred JB, Gertz MA, Hayman SR, Short KD, Buadi F,et al. Pomalidomide plus low-dose dexamethasone in myelomarefractory to both bortezomib and lenalidomide: comparison of 2dosing strategies in dual-refractory disease. Blood 2011;118:2970–5.

9. Kuroda J, Nagoshi H, Shimura Y, Taniwaki M. Elotuzumab and dar-atumumab: emerging new monoclonal antibodies for multiple myelo-ma. Expert Rev Anticancer Ther 2013;13:1081–8.

10. Steinbrunn T, St€uhmer T, Gattenl€ohner S, Rosenwald A, Mottok A,Unzicker C, et al. Mutated RAS and constitutively activated Aktdelineate distinct oncogenic pathways, which independently con-tribute to multiple myeloma cell survival. Blood 2011;117:1998–2004.

11. Chng WJ, Gonzalez-Paz N, Price-Troska T, Jacobus S, Rajkumar SV,OkenMM, et al. Clinical andbiological significanceofRASmutations inmultiple myeloma. Leukemia 2008;22:2280–4.

Chinen et al.





12. Chapman MA, Lawrence MS, Keats JJ, Cibulskis K, Sougnez C,Schinzel AC, et al. Initial genome sequencing and analysis of multiplemyeloma. Nature 2011;471:467–72.

13. Peterson TR, LaplanteM, ThoreenCC, Sancak Y, KangSA, KuehlWM,et al. DEPTOR is an mTOR inhibitor frequently overexpressed inmultiple myeloma cells and required for their survival. Cell 2009;137:873–86.

14. Z€ollinger A, St€uhmer T, ChatterjeeM, Gattenl€ohner S, Haralambieva E,M€uller-Hermelink HK, et al. Combined functional and molecular anal-ysis of tumor cell signaling defines 2 distinctmyeloma subgroups: Akt-dependent and Akt-independent multiple myeloma. Blood 2008;112:3403–11.

15. Liu P, Leong T,QuamL, BilladeauD, KayNE, Greipp P, et al. Activatingmutations of N- and K-ras in multiple myeloma show different clinicalassociations: analysis of the Eastern Cooperative Oncology GroupPhase III Trial. Blood 1996;88:2699–706.

16. Ramakrishnan V, Kimlinger T, Haug J, Painuly U, Wellik L, Halling T,et al. Anti-myeloma activity of Akt inhibition is linked to the activationstatus of PI3K/Akt andMEK/ERKpathway. PLoSONE2012;7:e50005.

17. 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.

18. Ocio EM,MateosMV,Maiso P, Pandiella A, San-Miguel JF. Newdrugsin multiple myeloma: mechanisms of action and phase I/II clinicalfindings. Lancet Oncol 2008;9:1157–65.

19. RichardsonPG,Wolf J, Jakubowiak A, Zonder J, Lonial S, Irwin D, et al.Perifosine plus bortezomib and dexamethasone in patients withrelapsed/refractory multiple myeloma previously treated with borte-zomib: results of a multicenter phase I/II trial. J Clin Oncol 2011;29:4243–9.

20. Richardson PG, Mitsiades CS, Laubach JP, Lonial S, Chanan-KhanAA, Anderson KC. Inhibition of heat shock protein 90 (HSP90) as atherapeutic strategy for the treatment of myeloma and other cancers.Br J Haematol 2011;152:367–79.

21. Ghobrial IM, Weller E, Vij R, Munshi NC, Banwait R, Bagshaw M,et al. Weekly bortezomib in combination with temsirolimus inrelapsed or relapsed and refractory multiple myeloma: a multicen-tre, phase 1/2, open-label, dose-escalation study. Lancet Oncol2011;12:263–72.

22. Will M, Qin AC, Toy W, Yao Z, Rodrik-Outmezguine V, Schneider C,et al. Rapid induction of apoptosis by PI3K inhibitors is dependentupon their transient inhibition of RAS-ERK signaling. Cancer Discov2014;4:334–47.

23. Zimmermann S, Moelling K. Phosphorylation and regulation of Raf byAkt (protein kinase B). Science 1999;286:1741–4.

24. Steinbrunn T, St€uhmer T, Sayehli C, Chatterjee M, Einsele H, BargouRC. Combined targeting of MEK/MAPK and PI3K/Akt signalling inmultiple myeloma. Br J Haematol 2012;159:430–40.

25. Shimura Y, Kuroda J, Ri M, Nagoshi H, Yamamoto-Sugitani M,Kobayashi T, et al. RSK2(Ser227) at N-terminal kinase domain is apotential therapeutic target for multiple myeloma. Mol Cancer Ther2012;11:2600–9.

26. AnjumR, Blenis J. The RSK family of kinases: emerging roles in cellularsignalling. Nat Rev Mol Cell Biol 2008;9:747–58.

27. Peifer C, Alessi DR. Small-molecule inhibitors of PDK1. Chem MedChem 2008;3:1810–38.

28. Roecklein BA, Torok-Storb B. Functionally distinct human marrowstromal cell lines immortalized by transduction with the human pap-illoma virus E6/E7 genes. Blood 1995;85:997–1005.

29. 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.

30. Kiyota M, Kobayashi T, Fuchida S, Yamamoto-Sugitani M, Ohshiro M,Shimura Y, et al. Monosomy 13 in metaphase spreads is a predictor ofpoor long-term outcome after bortezomib plus dexamethasone treat-ment for relapsed/refractory multiple myeloma. Int J Hematol 2012;95:516–26.

31. Feldman RI,Wu JM, Polokoff MA, KochannyMJ, Dinter H, Zhu D, et al.Novel small molecule inhibitors of 3-phosphoinositide-dependentkinase-1. J Biol Chem 2005;280:19867–74.

32. Zhang S, Suvannasankha A, Crean CD, White VL, Johnson A, ChenCS, et al. OSU-03012, a novel celecoxib derivative, is cytotoxic tomyeloma cells and acts through multiple mechanisms. Clin CancerRes 2007;13:4750–8.

33. Bergsagel PL, Kuehl WM, Zhan F, Sawyer J, Barlogie B, ShaughnessyJ Jr. Cyclin D dysregulation: an early and unifying pathogenic event inmultiple myeloma. Blood 2005;106:296–303.

34. Nakamura K, SakaueH, Nishizawa A,Matsuki Y, Gomi H,Watanabe E,et al. PDK1 regulates cell proliferation and cell cycle progressionthrough control of cyclin D1 and p27Kip1 expression. J Biol Chem2008;283:17702–11.

35. Moreaux J, Cremer FW, Reme T, Raab M, Mahtouk K, Kaukel P, et al.The level of TACI gene expression inmyelomacells is associatedwith asignature of microenvironment dependence versus a plasmablasticsignature. Blood 2005;106:1021–30

36. Huo J, Xu S, Lin B, Chng WJ, Lam KP. Fas apoptosis inhibitorymolecule is upregulated by IGF-1 signaling and modulates Akt acti-vation and IRF4 expression in multiple myeloma. Leukemia2013;27:1165–71.

37. Casamayor A, Morrice NA, Alessi DR. Phosphorylation of Ser-241 isessential for the activity of 3-phosphoinositide-dependent proteinkinase-1: identification of five sites of phosphorylation in vivo. BiochemJ 1999;342:287–92.

38. Venigalla RK, McGuire VA, Clarke R, Patterson-Kane JC, NajafovA, Toth R, et al. PDK1 regulates VDJ recombination, cell-cycleexit and survival during B-cell development. EMBO J 2013;32:1008–22.

39. Park SG, Long M, Kang JA, Kim WS, Lee CR, Im SH, et al. The kinasePDK1 is essential for B-cell receptormediated survival signaling. PLoSONE 2013;8:e55378.

40. Eser S, Reiff N, Messer M, Seidler B, Gottschalk K, Dobler M, et al.Selective requirement of PI3K/PDK1 signaling for Kras oncogene-driven pancreatic cell plasticity and cancer. Cancer Cell 2013;23:406–20.

41. Lu Z, Cox-Hipkin MA, Windsor WT, Boyapati A. 3-phosphoinositide-dependent protein kinase-1 regulates proliferation and survival ofcancer cells with an activated mitogen-activated protein kinase path-way. Mol Cancer Res 2010;8:421–32.

42. Zabkiewicz J, Pearn L, Hills RK,MorganRG, Tonks A, Burnett AK, et al.The PDK1 master kinase is overexpressed in acute myeloid leukemiaand promotes PKC-mediated survival of leukemic blasts. Haemato-logica 2014;99:858–64.

43. Tan J, Li Z, Lee PL, Guan P, Aau MY, Lee ST, et al. PDK1 signalingtoward PLK1-MYC activation confers oncogenic transformation,tumor-initiating cell activation, and resistance to mTOR-targeted ther-apy. Cancer Discov 2013;3:1156–71.

44. Shaffer AL, Emre NC, Lamy L, Ngo VN, Wright G, Xiao W, et al. IRF4addiction in multiple myeloma. Nature 2008;454:226–31.

45. Mora A, Davies AM, Bertrand L, Sharif I, Budas GR, Jovanovi�c S, et al.Deficiency of PDK1 in cardiac muscle results in heart failure andincreased sensitivity to hypoxia. EMBO J 2003;22:4666–76.

46. Ellwood-Yen K, Keilhack H, Kunii K, Dolinski B, Connor Y, Hu K, et al.PDK1 attenuation fails to prevent tumor formation in PTEN-deficienttransgenic mouse models. Cancer Res 2011;71:3052–65.

47. Hotte SJ, Oza A, Winquist EW, Moore M, Chen EX, Brown S, et al.Phase I trial of UCN-01 in combination with topotecan in patients withadvanced solid cancers: a Princess Margaret Hospital Phase II Con-sortium study. Ann Oncol 2006;17:334–40.

48. Lawlor MA, Mora A, Ashby PR, Williams MR, Murray-Tait V, Malone L,et al. Essential role of PDK1 in regulating cell size and development inmice. EMBO J 2002;21:3728–38.

49. Yan JW, Lin JS, He XX. The emerging role of miR-375 in cancer. Int JCancer 2014;135:1011–8.

50. Guti�errezNC, SarasqueteME,Misiewicz-Krzeminska I, DelgadoM,DeLas Rivas J, Ticona FV, et al. Deregulation of microRNA expression inthedifferent genetic subtypesofmultiplemyelomaandcorrelationwithgene expression profiling. Leukemia 2010;24:629–37.






2014;74:7418-7429. Published OnlineFirst September 30, 2014.Cancer Res Yoshiaki Chinen, Junya Kuroda, Yuji Shimura, et al. Mediator in Multiple MyelomaPhosphoinositide Protein Kinase PDPK1 Is a Crucial Cell Signaling

Updated version

10.1158/0008-5472.CAN-14-1420doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2015/03/18/0008-5472.CAN-14-1420.DC2

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/74/24/7418.full#ref-list-1

This article cites 50 articles, 25 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/74/24/7418.full#related-urls

This article has been cited by 4 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://cancerres.aacrjournals.org/content/74/24/7418To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-14-1420

http://cancerres.aacrjournals.org/content/suppl/2015/03/18/0008-5472.CAN-14-1420.DC2

http://cancerres.aacrjournals.org/content/74/24/7418.full#ref-list-1

http://cancerres.aacrjournals.org/content/74/24/7418.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/74/24/7418


Documents

Phosphoinositide Protein Kinase PDPK1 Is a Crucial Cell ...Molecular and Cellular Pathobiology Phosphoinositide Protein Kinase PDPK1 Is a Crucial Cell Signaling Mediator in Multiple