PIM Kinases Are Essential for Chronic Lymphocytic Leukemia ...mct.aacrjournals.org/content/molcanther/13/5/1231.full.pdfPIM Kinases Are Essential for Chronic Lymphocytic Leukemia Cell

Cancer Biology and Signal Transduction

PIM Kinases Are Essential for Chronic LymphocyticLeukemia Cell Survival (PIM2/3) and CXCR4-MediatedMicroenvironmental Interactions (PIM1)

Sarah Decker1,3, Johannes Finter1, Aaron James Forde1, Sandra Kissel1, Juerg Schwaller5,Thomas Sebastian Mack1, Anabel Kuhn1, Nathanael Gray6,7, Marie Follo1, Hassan Jumaa3,Meike Burger1, Katja Zirlik1, Dietmar Pfeifer1, Chandrasekhar V. Miduturu6,7, Hermann Eibel2,Hendrik Veelken1,8, and Christine Dierks1,4

AbstractOverexpression of the CXCR4 receptor is a hallmark of chronic lymphocytic leukemia (CLL) and is

important for CLL cell survival, migration, and interaction with their protective microenvironment. In acute

myelogenous leukemia (AML), PIM1 was shown to regulate the surface expression of the CXCR4 receptor.

Here, we show that PIM (proviral integration site for Moloney murine leukemia virus) kinases 1–3 are

overexpressed and that the CXCR4 receptor is hyperphosphorylated on Ser339 in CLL compared with normal

lymphocytes. Furthermore, CXCR4 phosphorylation correlates with PIM1 protein expression and PIM1

transcript levels in CLL. PIM kinase inhibition with three different PIM kinase inhibitors induced apoptosis

in CLL cells independent of the presence of protective stromal cells. In addition, PIM inhibition caused

dephosphorylation of the CXCR4 receptor on Ser339, resulting in enhanced ligand-dependent CXCR4

internalization and reduced re-externalization after withdrawal of CXCL12. Furthermore, PIM inhibition in

CLL cells blocked CXCR4 functions, such as migration toward CXCL12- or CXCL12-induced extracellular

signal–regulated kinase (ERK) phosphorylation. In concordance, pretreatment of CLL cells with PIM kinase

inhibitors strongly reducedhomingofCLL cells toward the bonemarrowand the spleen ofRag2�/�gc�/�mice

in vivo. Interestingly, the knockdownof PIMkinases inCLL cells demonstrated diverging functions,with PIM1

regulatingCXCR4 surface expression andPIM2andPIM3as important for the survival ofCLL cells.Our results

show that PIM kinase inhibitors are an effective therapeutic option for CLL, not only by impairing PIM2/3-

mediatedCLL cell survival, but also by blocking the PIM1/CXCR4-mediated interaction ofCLL cellswith their

protective microenvironment. Mol Cancer Ther; 13(5); 1231–45. �2014 AACR.

IntroductionB-cell chronic lymphocytic leukemia (B-CLL) is themost

prevalent adult leukemia inWestern countries and is char-acterized by a progressive accumulation of clonal CD5þ Blymphocytes in the peripheral blood, bone marrow, and

lymphoidorgans (1, 2). Several prognosticmarkers, suchasthe Rai and Binet staging systems (3, 4), immunoglobulinVH genemutational status (5), ZAP70 expression (6, 7), andcytogenetic abnormalities (8) like Del 13q14, Del 17p, andDel 11q, can be used to predict the survival outcomeand direct treatment strategies for patients with CLL (9).

In addition to cell intrinsic alterations, microenviron-mental factors are important for the survival of CLL cells.The cellular components of theCLLmicroenvironment arebone marrow stromal cells (BMSC), mesenchymal stromalcells in secondary lymphoid organs, monocyte-derivednurse-like cells, and T cells (10, 11). Coculture of CLL cellswith BMSCs results in the spontaneous migration of afraction of CLL cells beneath and underneath the BMSCs(12), protecting CLL cells from spontaneous apoptosis invitro (13, 14), and inducing resistance to conventionalchemotherapy in vivo (11, 15). One of the most importantpathways mediating the interaction between BMSCs andCLL cells is the CXCR4/CXCL12 pathway (16). TheG-pro-tein–coupled receptor CXCR4 is highly expressed on thecell surface of CLL cells (12, 17), while BMSCs produceCXCL12 ligands, also known as stroma cell–derived factor

Authors' Affiliations: 1Department of Hematology/Oncology; 2Centre ofChronic Immunodeficiency, University Medical Centre Freiburg; 3Faculty ofBiology; 4BIOSS Centre for Biological Signaling Studies, University of Frei-burg, Freiburg, Germany; 5Department of Biomedicine, University HospitalBasel, Basel, Switzerland; 6Department of Biological Chemistry andMolecularPharmacology, Harvard Medical School; 7Department of Cancer Biology,Dana-Faber Cancer Institute, Boston, Massachusetts; and 8Department ofHematology, Leiden University Medical Centre, Leiden, the Netherlands

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

S. Decker and J. Finter contributed equally to this work.

Corresponding Author: Christine Dierks, Department of Hematology/Oncology, University Medical Centre Freiburg, Hugstetter Strasse 55,79106 Freiburg, Germany. Phone: 49-761-270-71820; Fax: 49-761-270-71770; E-mail: [email protected]

doi: 10.1158/1535-7163.MCT-13-0575-T

�2014 American Association for Cancer Research.

MolecularCancer

Therapeutics

www.aacrjournals.org 1231

on February 27, 2020. © 2014 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst March 21, 2014; DOI: 10.1158/1535-7163.MCT-13-0575-T

http://mct.aacrjournals.org/

(SDF-1; ref. 18). CXCR4 activation by CXCL12 gradientsinducesCLL cell chemotaxis,migration across the vascularendothelium, actin polymerization, and migration underor underneath BMSCs and also shows direct antiapoptoticeffects (12, 19, 20). Binding of CXCL12 to the CXCR4receptor results in an internalization of the CXCR4 recep-tor by receptor endocytosis and pathway activationincluding extracellular signal-regulated kinase/mito-gen-activated protein kinase (ERK/MAPK) cascade acti-vation and intracellular calcium flux (12).

PIM (proviral integration site for Moloney murine leu-kemia virus) kinases were first described as potentialoncogenes by retroviral gene tagging inMoloney leukemiavirus–induced lymphomas (21). Further studies revealedtheir proto-oncogenic activity in cooperation with otheroncogenes such as c-MYC, N-MYC, or BCL2 (22). In mam-mals, three PIM serine/threonine kinases, PIM1, PIM2,and PIM3, are known (23). PIM kinases are constitutivelyactive and are regulated predominantly by transcriptionthrough the JAK–STAT signaling pathway and proteaso-mal degradation (24, 25). PIM1 for example, is a directtarget gene of STAT5 (26), which is frequently activated inhematologicmalignancies. Overexpression of PIM kinaseshas been shown in various hematologicmalignancies suchas diffuse large B-cell lymphoma (27), CLL (PIM2; ref. 28),and FLT3-associated acute myelogenous leukemia (AML;ref. 29) as well as in some solid tumors (30).

PIM kinases regulate pro- and antiapoptotic membersof the BCL2 protein family (31) and can block the proa-poptotic protein BAD (32, 33). Furthermore, PIM1 hasbeen shown to regulate cell-cycle regulators such asp21Cip1/WAF1 (CDKN1A) and p27KIF1 (34–36). Previousstudies on AML have shown an interesting link betweenPIM kinase activity and the surface expression level andfunction of the CXCR4 receptor. Pim1 knockout micedisplayed defects in hematopoietic stem cell (HSC) hom-ing to the bonemarrow and the spleen bydownregulationof the CXCR4 surface expression. Furthermore, PIMkinase inhibition in primary AML blasts induced down-regulation of the CXCR4 receptor on the cell surface andPIM1 transcript levels correlated with CXCR4 surfaceexpression in AML (37).

In the experiments shownhere,we examined the role ofPIM kinases and their effect on regulating the CXCR4receptor in CLL. We found that the CXCR4 receptor ishyperphosphorylated on Ser339, and PIM1 and PIM2 areboth overexpressed in the majority of CLL patient sam-ples. PIMkinase inhibition induced apoptosis inCLL cellseven in the presence of protective stromal cells. Further-more, PIM kinase inhibition reduced CXCR4 surfaceexpression, increased CXCL12-mediated internalizationof the CXCR4 receptor, and abrogated CXCR4 functions,such asmigration towardCXCL12 or extracellular signal–regulated kinase (ERK) phosphorylation, as well as thehoming of CLL cells toward the bone marrow and spleenin vivo. Knockdown of PIM1-3 revealed diverging func-tionswithPIM1 regulatingCXCR4 surface expression andPIM2/3 as prosurvival genes.

Materials and MethodsCLL patient samples

This study was approved by the Institutional ReviewBoard of the University Medical Center Freiburg (Frei-burg, Germany). Peripheral blood samples were obtainedwith informed consent in accordancewith theDeclarationof Helsinki from patients with B-CLL who were eitheruntreated or off therapy for at least 6months (summary ofpatient data and risk factors; Supplementary Table S1).CLL cases were characterized for IgVH mutational status,disease stage according to the Binet and Rai criteria, andhistory of treatment. Furthermore, genetic aberrationswere analyzed by chromosomal analysis and FISH anal-ysis and copy number changes were verified by single-nucleotide polymorphism (SNP) arrays (ref. 38; patientcharacteristics are shown in Supplementary Table S1).Peripheral blood mononuclear cells (PBMC) were sepa-rated by Ficoll gradient centrifugation and either usedfresh or were cryopreserved in fetal calf serum (FCS) with10% dimethyl sulfoxide (DMSO) until use. Cells weremaintained in RPMI-1640 medium with 10% FCS and1% penicillin–streptomycin. PBMCs for quantitative PCR(qPCR) orWestern blot analysis containedmore than 90%CD20þ/CD5þ CLL cells. Cells were lysed in RLT bufferfor further RNA extraction or deep-frozen for proteinextraction and subsequent immunoblotting.

Isolation of B cells from healthy donorsWhole blood samples from healthy donors were

obtained from the blood tissue bank from the Universityof Freiburg (Freiburg, Germany). Mononuclear cells wereisolated from whole blood samples by Ficoll gradientseparation, followed by CD19 selection with anti-CD19magnetic microbeads (Miltenyi Biotec) and the autoMacsCell Separator (Miltenyi Biotec). Purity was determinedby flow cytometry andwas >90%. Cells were lysed in RLTbuffer for further RNA extraction or deep-frozen forprotein extraction and subsequent immunoblotting.

Cell linesThemurine BMSC lineM2-10B4 (ATCCCRL-1972) and

the human BMSC lineHS-5 (ATCCCRL-11882) were bothacquired fromAmerican Type Culture Collection (ATCC;2010) and cultured in RPMI-1640 supplementedwith 10%FBS. The cell lineswere tested byATCC andwere authen-ticated in our laboratory by growth morphology and bythe expression of laminin and collagen IV (M2-10B4) andof interleukin (IL)-6 and IL-8 (HS-5) by qPCR. MEC-1(ACC-497), a human CLL cell line, was obtained fromDSMZ (in 2012) and cultured inRPMI-1640 supplementedwith 10% FBS and was authenticated by surface stainingand flow cytometry for CD19, CD20, and CD79a.

Small-molecule inhibitors against PIM kinases andCXCR4 antagonists

The PIM kinase inhibitors K00135 and K00486 areimidazo[1,2-b]pyridazines and have been characterized

Decker et al.

Mol Cancer Ther; 13(5) May 2014 Molecular Cancer Therapeutics1232




previously (34, 36). Both inhibitors show specific bindingtoward PIM1 and PIM2 (K00135 IC50 values: PIM1, 0.12mmol/L; PIM2, 1.8mmol/L;K00486 IC50 values: PIM1, 0.04mmol/L; PIM2, 2.5 mmol/L) in in vitro kinase assays. Thesmall-molecule inhibitor A47 is based on another struc-tural background, a furan thiazolidin (39). Besides goodefficacy onPIM1andPIM2, this inhibitor also blocks PIM3kinase activity (IC50 values: PIM1, 3 mmol/L; PIM2, 0.8mmol/L; PIM3, 2.6 mmol/L). The CXCR4 antagonist pler-ixafor (AMD3100; Sigma) is used in the clinic for mobi-lization of HSCs and for the treatment of CLL in earlyclinical trials (16).

Quantitative real-time PCRTotal cellular RNA was isolated from PBMCs from

patients with CLL containing more than 90% of CD20þ

CD5þ CLL cells and from CD19þ cells from healthydonors using the Qiagen RNAeasy Mini Kit. The cDNAwas synthesized using 500 ng of every mRNA samplewithOligo-dT-Primers (Life Technologies), SuperScript IIReverse Transcriptase (Life Technologies), and desoxy-nucleotides (Fermentas) following the manufacturer’sinstructions. The mRNA transcript level was measuredby quantitative TaqMan real-time PCR using a Light-Cycler 480 (Roche). TaqMan primers and probes werepurchased from Applied Biosystems. The following pri-mers were used: PIM1 hs00171473 m1, PIM2 hs00179139m1, PIM3 hs00420511, CXCR4 hs00976734, and GAPDH4310884E-090204. Results were quantified according tothe "delta-delta-CT method" based on the relative expres-sion of the target gene versus a reference gene (GAPDH)and normalized to the median of the control samples. Forquantification of CXCR4 mRNA after PIM kinase inhibi-tion, CLL cells were seeded in 6-well plates and treatedwith the PIM kinase inhibitor 10 mmol/L K00135 for 2hours, after which CLL cells were collected, washed, andRNA was extracted as indicated above.

ImmunoblottingProtein expression of PIM1, PIM2, PIM3, and CXCR4

was determined in 11 human CLL samples and threenormal CD19þ B lymphocyte ("NL") samples using stan-dardized protocols with antibodies against the followingproteins: PIM1 (clone 12H8; Santa Cruz Biotechnology;1:25), PIM2 (clone D1D2; Cell Signaling Technology;1:500), PIM3 (clone RB8591; Abgent; 1:100), CXCR4 (clone2074; Abcam; 1:500; 1:1,000), phospho-CXCR4 (pCXCR4;clone 74012; Abcam; 1:1,000), pERK (9106; Cell SignalingTechnology; 1:1,000), ERK (9102; Cell Signaling Technol-ogy; 1:1,000), pBAD (5284; Cell Signaling Technology;1:250), BAD (9239; Cell Signaling Technology; 1:250), andb-actin (clone AC-15; Sigma; 1:5,000). Western blot anal-yses were analyzed using the ImageJ software (NIH,Bethesda, MD).

Apoptosis assayPBMCs were plated into 96-well plates at a concentra-

tion of 1� 105 cells perwell with orwithout support of the

murine stromal cell line M2-10B4 (ATCC). In the case ofstromal coculture, onday�1, 5� 103 stromal cells perwellwere plated in 100 mL of medium containing RPMI-1640with 10% FBS. After 24 hours, CLL cells were added andtreated with three different PIM kinase inhibitors, theCXCR4 inhibitors AMD3100, or combinations at the indi-cated concentrations. After 24, 48, and 72 hours of incu-bation at 37�C in 5% CO2, cells were stained with a CD19-APC antibody (BD Biosciences), followed by Annexin V/7-AAD staining (BD Biosciences) according to the man-ufacturer’s instructions. Cells were analyzed using theCyAn ADP flow cytometer (Beckmann Coulter). Flowcytometry data were analyzed using the FlowJo 7.6 soft-ware (TreeStar).

CXCR4 surface expressionCLL cells were cultured in 96-well plates treated either

withDMSOor thePIMkinase inhibitor 10mmol/LK00135for 2 hours. CLL cells from 96-well plates were stainedwith allophycocyanin (APC)-conjugated anti-humanCD184 antibodies (clone 12G5; BDBiosciences) forCXCR4surface expression and Annexin V/7-AAD (BD Bios-ciences) for analyzing apoptosis. Nonspecific bindingwasmeasured by APC-conjugated mouse IgG2a, k (cloneG155-178; BD Biosciences) as an isotype control.

Intracellular phospho-ERK stainingCLL cells alone or cocultured with M2-10B4 BMSCs

were plated as above and treatedwith 10 mmol/LK00135,100 mg/mL TN14003, or 20 mmol/L AMD3100. After 2hours of incubation, cells were stained with CD19 anti-body, fixed with 1.85% formalin, and permeabilized with90% methanol. Endogenous levels of p44 and p42 MAPKwere detected with the phospho-p44/42 MAPK mousemonoclonal antibody (#9106; Cell Signaling Technology)and anti-mouse immunoglobulin G (IgG) Alexa Fluor–conjugated antibody (#4408; Cell Signaling Technology).In the case of CXCL12 stimulation, cells were treatedwith 500 ng/mLofCXCL12 (R&DSystems) for 30 secondsbefore fixation. Cells and datawere analyzed as describedabove.

Chemotaxis assaysCLL cells were preincubated for 2 or 24 hours with 10

mmol/L of PIM inhibitor K00135 in a suspension of RPMI-1640 and 0.5% bovine serum albumin. A volume of 100 mLcontaining 5� 106 CLL cells was added to the top chamberofa 6.5-mmdiameterTransfer culture insert (Coster)with apore size of 5 mm. Filters were then transferred tomediumeither with or without 500 ng/mL of CXCL12. Chamberswere incubated for 2 or 24 hours at 37�C and 5% CO2.Then, the lower chamberwas suspended anddivided intoaliquots for counting with the CyAn ADP flow cytometer(Beckmann Coulter) for 40 seconds at 100 mL/min.

CXCR4 recycling assayCLL cells were seeded into 96-well plates with and

without 500 ng/mL CXCL12 and treated with DMSO or

PIM Kinases Regulate CLL Survival and CXCR4 Surface Levels

www.aacrjournals.org Mol Cancer Ther; 13(5) May 2014 1233




10 mmol/L K00135 for 12 hours, followed by three wash-ing steps with PBS to remove CXCL12 and K00135 andresuspension in medium. After treatment and 12 hoursafterwashing out CXCL12 andK00135, the surface of CLLcells was stained with APC-conjugated anti-humanCD184 antibody. Then, cells were fixed with 1.85% for-malin and permeabilizedwith 90%methanol. Total (cyto-plasmic and membrane bound) levels of CXCR4 weredetected with primary anti-human CD184 antibody(2074; Abcam; 1:100) and with secondary antibody AlexaFluor 488 goat anti-rabbit (Life Technologies, 1:100). Fornuclear staining, slides were incubated with 40,6-diami-dino-2-phenylindole (DAPI; Sigma; 1:1,000). CLL cellswere collected for flow cytometry, immunofluorescenceanalysis using confocal microscopy (Leica SP2 AOBSspectral confocal microscope), and High Content Screen-ing using the Olympus ScanR.

CLL cell homing into Rag2�/�gc�/� miceRag2�/�gc�/� mice (40) were bred and handled un-

der sterile conditions and transplanted at 8 to 10 weeksof age after sublethal irradiation (3.0 Gy) 18 hoursbefore transplantation. Primary human CLL cells werestained with 1 mmol/L carboxyfluorescein diacetatesuccinimidyl ester (CFSE; Life Technologies) accordingto the manufacturer’s instructions and treated with10 mmol/L K00135 for 12 hours. Viability of the cellswas assessed by Annexin V/7-AAD staining. Subse-quently, CLL cells were resuspended in Hank’s Bal-anced Salt Solution (HBSS) and irradiated micereceived transplants of 3.5 � 107 PBMCs via tail veininjection. Mice were analyzed 4 hours after transplan-tation. Single-cell suspensions of spleens and bonemarrow were stained with APC-labeled anti-humanCD45 antibody (BD Biosciences) and PE-labeled anti-human CD19 antibody (BD Biosciences) and CD45þ

CD19þ/CFSE cells were detected by flow cytometry.All animal experiments were approved by the Regier-

ungspr€asidium Freiburg andwere in accordance with theU.S. NIH Statement of Compliance with Standards forHumane Care and Use of Laboratory Animals.

siRNA-mediated PIM1–3 knockdownFreshly isolated CLL cells at a concentration of 1 � 107

were resuspended inNucleofection solution (NucleofectorKitV; Lonza). siRNAs for PIM kinases were added at aconcentration of 20mmol/L (PIM1: SI00040978, SI00040985,and SI02629165; PIM2: SI00092869 and SI02224201; PIM3:SI00684915 and SI03084543; Qiagen). Transfection wasstarted using programU-13 of Amaxa Biosystems Nucleo-fector II (AmaxaBiosystems).After 24hours, onepart of thecells was stained with CD184-APC and Annexin V–PE/7-AAD (both BD Biosciences) and analyzed as describedabove. The other part of the cells was used for immuno-blotting for PIM kinases 1–3, phospho-BAD, BAD, andb-actin.Knockdown (KD)efficiencywas calculated relativeto b-actin using ImageJ software and compared with NoSicontrol.

Statistical analysisData are represented as themean� SEM. Comparisons

between parameters were performed using a two-tailed,paired Student t test or the Mann–Whitney test. Correla-tions were assessed with the Spearman correlation coef-ficient. For all analyses, P < 0.05 was considered statisti-cally significant.

ResultsPIM1 and PIM2 are overexpressed and the CXCR4receptor is hyperphosphorylated on Ser339 in CLL

Previous studies have shown an increase in the expres-sion of the PIM2 kinase in CLL cells by using qPCR andintracellular fluorescence-activated cell sorting (FACS)staining [Cohen and colleagues (28); and Chen and col-leagues, (41)]. To verify these results in our CLL cohort(Supplementary Table S1), we investigated the expressionlevel of all three PIM kinases at the transcriptional andprotein levels. Using quantitative real-time PCR (qRT-PCR), we identified a significant increase in PIM2 (P ¼0.0404), but also in PIM1 (P ¼ 0.0060) transcript levels inCLL cells (n ¼ 15) compared with normal B lymphocytes(n¼ 4) from healthy donors (Fig. 1A).Median PIM1 levelswere increased 5.1-fold and median PIM2 levels wereincreased 1.7-fold. There was a tendency toward higherRNA expression levels of PIM3 and CXCR4 in CLL cellscompared with normal B lymphocytes, which was notstatistically significant (Fig. 1A).

To evaluatewhether or not elevated transcript levels aretranslated into increased expression of PIM kinases at theprotein level, we performed Western blot analyses forPIM1, PIM2, PIM3, CXCR4, and pCXCR4 with B-CLLlymphocytes (n ¼ 11) and normal lymphocytes fromhealthy donors (n ¼ 3). Also on protein levels, PIM1 andPIM2 were significantly overexpressed in CLL samplescompared with normal B lymphocytes and relative tob-actin (PIM1, P ¼ 0.0142; PIM2, P ¼ 0.0251; Fig. 1B andC). Relative levels of PIM1 proteinwere increased 2.5-foldwith increased protein levels in 10 of 11 CLL samples,while relative levels of PIM2 were increased 6-fold withincreased protein levels in 11 of 11 CLLs (Fig. 1C). Inter-estingly, the smaller isoform of PIM1 at 37 kDwas presentonly in CLL cells, but not in normal B lymphocytes.Relative levels of PIM3 protein were increased in six of11 CLL samples (Fig. 1B and C). There was no correlationin between PIM transcript levels or PIM protein expres-sion levels and risk factors in CLL, like stage of disease ormutational status.

Previous studies on HSCs have shown that PIM1 canphosphorylate Serine339 on the CXCR4 receptor and thatthis regulates CXCR4 surface expression (37). In CLL, it isknown that the CXCR4 receptor is highly upregulated onthe cell surface compared with normal B lymphocytes,while total CXCR4 protein levels are normal. As previ-ously described, total CXCR4 RNA and protein levelswere not significantly elevated in our CLL cohort com-paredwith normal B cells (Fig. 1A–C). In contrast, CXCR4

Decker et al.





Figure 1. PIM1/2 and pCXCR4 are overexpressed in CLL. A, expression of PIM1, PIM2, PIM3, and CXCR4 mRNA in B-CLL cells (CLL) versus normallymphocytes (NL) from healthy donors by qRT-PCR analysis. The values are normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH)transcript levels and represent each CLL patient or normal lymphocytes donor. Mean value for controls is set to one. Experiments were performed inindependent duplicates or triplicates (Mann–Whitney test). Transcript levels for PIM1 and PIM2 are elevated in CLL patient samples compared with controls(PIM1, P ¼ 0.0060; PIM2, P ¼ 0.0404; patient samples CLL 1–16 from Supplementary Table S1). B, protein expression of pCXCR4, CXCR4, PIM1, PIM2,and PIM3 in B-CLL lymphocytes (CLL#) from patients with CLL and normal lymphocytes from healthy donors analyzed by Western blot analysis andusing b-actin as a loading control. C, protein expression levelswere quantified using the ImageJ software (NIH), and the expression levels of pCXCR4,CXCR4,PIM1, PIM2, and PIM3were normalized to b-actin expression levels. Mean expression levels of specific genes compared with b-actin in normal lymphocyteswere set to 1. PIM1 and PIM2 protein levels and CXCR4 phosphorylation is significantly elevated in CLLs compared with normal lymphocytes (PIM1, 2.5-foldincrease, P ¼ 0.0124; PIM2, 6-fold increase, P ¼ 0.0251; pCXCR4 with 4.6-fold increase). Total levels of PIM3 and CXCR4 are not significantly alteredin CLLs compared with normal lymphocytes. D, Spearman correlation coefficient of pCXCR4 protein and PIM1 protein level (r ¼ 0.6088; P ¼ 0.0209).






phosphorylation on Ser339 was strongly increased inCLLs compared with normal lymphocytes (Fig. 1B andC; 4.6-fold increase). There was a direct positive correla-tion in betweenCXCR4 phosphorylation and PIM1 kinaseprotein expression in CLL cells (Spearman correlationcoefficient, P ¼ 0.0209; Fig. 1D).

As previously described, CXCR4 surface expressionlevels were significantly elevated in our CLL cohort (Fig.2A). PIM1 kinase expression on transcriptional level andalso on protein level both positively correlated with thesurface expression of the CXCR4 receptor (Fig. 2B and C),pointing toward PIM1 as a potential regulator of CXCR4surface expression in CLL.

PIM kinase inhibition with three different PIMinhibitors induces apoptosis in CLLs independent ofcoculture with stromal cells

To verify the importance of PIM kinase activity for thesurvival of CLL cells, we investigated the potential ofthree different small-molecule PIM kinase inhibitors onapoptosis induction in CLL cells (34, 36, 39). K00135 andK00486, two imidazo[1,2-b]pyridazines, show low IC50

values for inhibition of PIM1 and PIM2 kinase activity,andhave shownproapoptotic effects inAMLcells (37, 42).The third PIM kinase inhibitor, A47, is based on anotherchemical structure, a furan thiazolidin, and inhibits allthree PIM kinases in low micromolar concentrationsincluding PIM3 (39). Treatment of CLL cells (n¼ 16) withPIM inhibitors for 48 hours showed a dose-dependentincrease in apoptosis (Fig. 3A for K00135, for example,CLL10) and a significant reduction in viable cells (Fig. 3B).PIM1/PIM2 kinase inhibitors K00135 and K00486 at 10mmol/L reduced themean cell viability to about 50% after48 hours of incubation (Fig. 3C). Treatment with the A47PIM1–3 kinase inhibitor also showed a significant induc-tion of apoptosis, although its efficacywas lower than thatof the two other compounds.

Previous studies have demonstrated the importanceof the microenvironment, and especially the presence ofstromal cells, for the survival of CLL cells (10, 42). The

interaction between BMSCs and CLL cells induces aprotective microenvironment blocking the proapoptoticeffect of various chemotherapeutical agents and spon-taneous apoptosis in vitro. Pathways involved in thisprocess are the CXCL12–CXCR4 axis, Hedgehog (HH)ligands activating the HH signaling pathway in CLLcells, and various other chemokines and integrins(12, 16, 43). To evaluate the effect of protective stromalcells on the effect of PIM kinase inhibitors in CLL, wecocultured CLL cells with the BMSC line M2-10B4 (baseline viability with and without stroma is shown inSupplementary Fig. S1). Interestingly, M2-10B4 cellscould not sufficiently protect the CLL cells from PIMinhibitor–induced apoptosis, although the effect of thePIM inhibitors was stronger in the absence of stromalcells (Fig. 3C). Mean IC50 values without stroma werelowest for K00486 (2.40 mmol/L) compared with K00135(2.96 mmol/L) and A47 (4.10 mmol/L). The presence ofstromal cells significantly increased the IC50 values forK00135 (mean IC50, 4.61 mmol/L; P ¼ 0.0029) and A47(mean IC50, 5.71 mmol/L; P¼ 0.0148; Fig. 3D), but not forK00486 (mean IC50, 3.36 mmol/L; P ¼ 0.0571; Fig. 3D).

Two possible mechanisms could be responsible for thislow protective effect. Either PIM inhibitors can directlyblock the protective pathways coming from the stromalcells, or the stromal cells are directly affected by the PIMinhibitors.

To address both possibilities, we first investigated thepresence of the PIM–CXCR4 signaling pathway in themurine stromal cell line M2-10B4 (MS) and in the humanstromal cell line HS-5 (HS). Interestingly, besides CLLcells, also BMSCs express all three PIM kinases and theCXCR4 receptor (Fig. 4A). High PIM inhibitor concentra-tions of K00486 and K00135 (10 mmol/L) could induceapoptosis in murine stromal cells with IC50 valuesbetween 8 and 10 mmol/L (stroma IC50 values forK00135, 8.2 mmol/L; K00486, 9.65 mmol/L; and A47,>10 mmol/L; Fig. 3B andD), while therewas no significantapoptosis induction in M2-10B4 cells after treatment withthe A47 inhibitor.

Figure 2. CXCR4 surface expression is upregulated on CLL cells and correlates with PIM1 transcript and protein levels. A, MFI of CXCR4 surfaceexpression onprimaryCLL samples (n¼16) comparedwith normal lymphocytes (NL; n¼3; unpaired t test;P¼0.0319). B, Spearman correlation coefficient ofCXCR4 surface expression measured as MFI versus relative PIM1 transcript levels as assessed by qPCR and relative to glyceraldehyde-3-phosphatedehydrogenase (GAPDH; r ¼ 0.6054; P ¼ 0.0100). C, Spearman correlation coefficient of CXCR4 surface expression (MFI) versus PIM1 protein expressionrelative to b-actin (r ¼ 0.6209; P ¼ 0.0235).

Decker et al.





To investigate whether PIM kinases and the impor-tant CXCR4–CXCL12 axis share the same survival sig-nals, we investigated the effect of a combination of thePIM kinase inhibitor K00135 and the potent CXCR4inhibitor plerixafor on CLL cells (n ¼ 5) in the presenceof M2-10B4 BMSCs. Treatment of CLL cells with 20mmol/L plerixafor could significantly induce moderateapoptosis in CLL cells (Fig. 4B). The treatment of CLLcells with 5 mmol/L K00135 showed a significant and

strong induction of apoptosis as measured by AnnexinV/7-AAD staining after 48 hours. The addition of pler-ixafor to the PIM inhibitor K00135 could not enhanceapoptosis induction, which indicates that proapoptoticeffects induced by plerixafor might already be blockedby the PIM kinase inhibitor (Fig. 4B). Taken together,the lack of CLL protection by stromal cells might becaused by a dual effect of PIM inhibitors, with directlyaffecting stromal cell viability at least at high inhibitor

Figure 3. PIM inhibitors induceapoptosis inCLL cells independentof coculture with protectiveBMSCs. A, flow cytometryanalyses of CLL#10 cellscocultured with M2-10B4 stromalcells and treated with K00135 atdifferent concentrations for 48hours. Apoptosis was assessed byAnnexin V–PE/7-AAD staining. B,cell viability of CLL cells with orwithout stromal support after 48hours of treatment with K00135(n ¼ 11), K00486 (n ¼ 16), and A47(n ¼ 16). Viable cells were definedas negative for Annexin V–PE/7-AAD staining and graphs showviable cells relative to the DMSOcontrol. Viability of stromal cellswas assessed using CellTiter-Gloassay of M2-10B4 stromal cellsafter 48 hours of PIM inhibition (��,�, paired t test). C, averageresponses shown by viable cellcounts of CLL samples #1–#16after 48 hours treatment with 10mmol/L of PIM inhibitor (��, �, pairedt test). D, IC50 values for apoptosisinduction of three different PIMinhibitors in CLL cells with andwithout stromal support or stromalcells alone (��, �, paired t test). n.s.,not significant. �, 0.01 < P < 0.05;��, 0.001 < P < 0.01; ��, P < 0.001.






concentrations on the one hand, and direct interferencewith CLL survival pathways activated by stromal cells,like CXCR4, on the other.

PIM kinase inhibitors reduce CXCR4 surfaceexpression and CXCR4 phosphorylation in CLL

To further investigate the second possibility, we inves-tigated the effect of PIM inhibitors on CXCR4 surfaceexpression andCXCR4phosphorylation. InAML,Grund-ler and colleagues identified PIM1 as a regulator for thesurface expression of the CXCR4 receptor (37). Treatmentof CLL cellswith theK00135 PIMkinase inhibitor resultedin a significant decrease in CXCR4 surface expressionlevels (n ¼ 5), indicating that PIM kinases can act asregulators for CXCR4 surface expression also in CLL (Fig.5A). In contrast to the downregulation of CXCR4 surfaceexpression by PIM inhibition, we found that the CXCR4

mRNA levels were upregulated in the same samples afterPIM inhibition,which suggests a counteracting regulatoryloop in CLL cells after downregulation of CXCR4 surfaceexpression (Fig. 5B).

Grundler and colleagues previously identified phos-phorylation of Ser339 to be critical for CXCR4 surfaceexpression on AML cells (37). Therefore, we also investi-gated the phosphorylation level of Ser339 in CLL cellstreated with PIM kinase inhibitors and found a strongreduction of Ser339 phosphorylation within 6 hours oftreatment (Fig. 5C).

PIM kinase inhibition enhances CXCL12-mediatedinternalization of the CXCR4 receptor in CLL cells

To further characterize the effect of PIM kinase inhibi-tion on the surface expression of theCXCR4 receptor, CLLcells (four independent experiments with four different

Figure 4. Combination treatmentof PIM inhibitors with CXCR4inhibitors does not showsynergistic effects. A, proteinexpression of pCXCR4, CXCR4,PIM1, PIM2, and PIM3 in M2-10B4andHS-5 stromal cells analyzed byWestern blot analysis and usingb-actin as a loading control. B, flowcytometry analyses for AnnexinV/7-AAD of CLL cells (n ¼ 5)cocultured with M2-10B4 stromalcells and treated with K00135 (5mmol/L) and AMD3100 (20 mmol/L)for 48 hours. Graph shows viablecell counts relative to DMSOcontrol.

Figure 5. PIM inhibition reducesCXCR4 surface expression andCXCR4 Ser339 phosphorylation inCLL. A, median CXCR4 surfaceexpression quantified by flowcytometry after 2 hours oftreatment with K00135 andCXCL12þ stromal support (n ¼ 5,paired t test, P ¼ 0.0004; CLL#6,#9, #13, #15, and #16). B, CXCR4mRNA values in qRT-PCR of thesame CLL samples after 2 hours oftreatment with PIM inhibitorK00135 and stromal support (n¼ 5,paired t test, P ¼ 0.0399). C,Western blot analysis for pCXCR4and CXCR4 versus b-actin fromCLL cells treated either with DMSOor the PIM inhibitor K00135 for 6and 24 hours (CLL#19).

Decker et al.





patients with CLL) were incubated for 12 hours with 500ng/mLof CXCL12 to induce internalization of the CXCR4receptor andwere treatedwith eitherDMSOor 10mmol/LK00135. Viability of cells as assessed by Annexin V/7-AAD staining after this treatment was more than 90%in both treatment groups (data not shown). CLL cellstreated with CXCL12 showed a nearly complete down-regulation of CXCR4 from the cell surface, as shown byflow cytometry (Fig. 6A, second lane) and by dual colorconfocal microscopy (APC-labeled CXCR4 membranestaining in red and Alexa Fluor 488–labeled intracellularCXCR4 staining in green; Fig. 6B, mid lane). Cells treatedwith the PIM inhibitor K00135 showed a slight decrease inbaseline CXCR4 surface expression (Fig. 6A) as measuredby flow cytometry. Upon treatment with CXCL12, CLLcells with PIM inhibition also showed CXCR4 internali-zation (Fig. 6A and B) and showed clustering of theCXCR4 receptor in the cytoplasm (Fig. 6B). In the controlgroup,washout ofCXCL12 for 12 hours resulted in a slow,but measurable re-externalization of the CXCR4 receptor

to the cell surface (Fig. 6A–D), whereas CLL cells withprevious PIMkinase inhibition showed a further decreasein surface CXCR4 levels and accumulation of CXCR4 inclusters in the cytoplasm (Fig. 6B). Flow cytometry datawere quantitated as mean fluorescence intensity (MFI)in Fig. 6C and for a CLL cell line (MEC-1) in Supplemen-tary Fig. S2. In addition to the confocal microscopy, wealso performed high-resolution imaging and quantitatedcells with mainly membrane staining (red ring) versuscells with mainly intracellular staining (green). By usingthis method, we clearly see the internalization of theCXCR4 receptor after CXCL12 stimulation (Fig. 6D).Furthermore, we can detect re-externalization of theCXCR4 receptor to the membrane after CXCL12 wash-out, which is totally absent after PIM kinase inhibitortreatment (Fig. 6D).

Taken together, these results indicate an importantregulatory role for PIM kinases on CXCR4 surfaceexpression, and especially the re-externalization of thereceptor in CLL.

Figure 6. PIM kinase inhibitorsblock CXCR4 re-externalization onCLL cells. A, CXCR4 surfaceexpression measured by flowcytometry of CLL#6 CLL cells afterincubation with and withoutCXCL12 and/or K00135 for 12hours (top and middle row) and 12hours after washout of CXCL12and K00135 (bottom row). Datashow one of two independentexperiments. B, confocalmicroscopy from anotherexperiment with CLL#24 using adual color CXCR4 staining tovisualize the localization of theCXCR4 receptor. Red, membrane-boundCXCR4; green, total CXCR4after permeabilization includingcytoplasmic CXCR4. Data showone of four individual experiments.C, MFI for CXCR4 surfaceexpression from A. D, ratio ofsurface stained (red) versusintracellularly stained (green)CXCR4 in CLL cells (CLL#24) wasanalyzed by High ContentScreening using Olympus ScanR.��, P value 0.001 < P < 0.01.






PIM kinase inhibitors block the function of theCXCR4 receptor regarding migration and ERKphosphorylation

Previous studies have shown that CXCR4-expressingcells canmigrate toward aCXCL12gradient (12). Toverifythat PIMkinases can not only affect the surface expressionof the CXCR4 receptor, but can also affect the function ofCXCR4, we performed migration assays of CLL cellstoward a CXCL12 gradient in the presence of either PIMinhibitor or DMSO as a control. The amount of migratingcells without CXCL12 was below 1%. Upon addition ofCXCL12, CLL cells started to migrate into the lower

chamber (CLL#16, 3.8% after 2 hours and 30% after 24hours; Fig. 7A), yet this effect was nearly completelyabolisheduponPIMkinase inhibition (withPIM inhibitor,2.2% after 2 hours and 3% after 24 hours; Fig. 7A). Threeother CLL samples showed very similar results (mean ofmigrating cells, DMSO41%andK00135 15%), indicating afunctional loss of the CXCR4 receptor upon treatmentwith PIM kinase inhibitors (Fig. 7A, third panel).

To further examine the effect of PIM kinase inhibitiononCXCR4 receptor signaling,we investigated the effect ofPIM inhibition on the MAPK pathway, which is a knownCXCR4 downstream signaling pathway. Stimulation of

Figure 7. PIM kinase inhibitionblocks CXCR4 phosphorylationand function in vitro. A, migrationassay using a Transwell chamberandCXCL12 as a chemoattractant.Percentage of migrated cells forCLL sample #16 after 2 (left) and24 (middle) hours and meanpercentage of migrated cells forthree CLL samples (CLL#6, #15,and #16; left) after 24-hourpretreatment with 10 mmol/L ofK00135 toward a CXCL12 gradientand without (��, 0.001 < P < 0.01;paired t test). B, Western blotanalysis for phospho-ERK andERK versus b-actin of CLL cellscocultured with or without CXCL12(right) or M2-10B4 stromal cellsand treated with PIM inhibitorK00135 (K) or the CXCR4 inhibitorAMD3100 (A) for 3 or 6 hours(CLL#13). C, MFI for phospho-ERKmeasured by flow cytometry inCLL cells cultivated with or withoutCXCL12 or M2-10B4 stromal cellsand treated with the PIM inhibitorK00135 or the CXCR4 inhibitorAMD3100 (n ¼ 7; paired t test;analyzed patients top left, top right,and bottom left: 4, 13, 15, 17, 19,20, and 21; bottom right: 4, 13, 15,17, 18, 20, and 21). D, Western blotanalysis for phospho-ERK andERK after 3 hours and forphospho-BAD, BAD, pCXCR4,and CXCR4 after 24 hours oftreatment with PIM inhibitorsK00486 and A47 (CLL#19).

Decker et al.





CLL cells with recombinant CXCL12 or coculture of CLLcells with CXCL12-producing stromal cells (M2-10B4)induced ERK phosphorylation as shown by intracellularphospho-flowcytometry andWesternblot analysis (Fig. 7Band C). Treatment of CXCL12-stimulated CLL cells withthe CXCR4 inhibitor AMD3100 (plerixafor) stronglyreduced CXCL12-induced ERK phosphorylation (Fig. 7BandC). Interestingly,PIMkinase inhibitionwith10mmol/LK00135 had the same effect on ERK phosphorylation,indicating an interference with the CXCL12–CXCR4 sig-naling pathway in CLL cells (Fig. 7B and C). In addition,ERK phosphorylation in CLL cells induced by coculturewith CXCL12-producing M2-10B4 stromal cells could beabrogated by CXCR4 inhibitors or PIM inhibitors (Fig. 7B).Also the other two PIM kinase inhibitors (K00486 and

A47) were effective in reducing ERK phosphorylationeven in the absence of CXCL12 stimulation (Fig. 7D) andcould reduce phosphorylation of the known PIM kinasetargets BAD and the CXCR4 receptor (Fig. 7D).

PIM kinase inhibition in CLL cells blocks migrationof CLL cells into the bone marrow and spleens ofRag2�/�gc�/� miceThe CXCR4 receptor is known to be important for the

homing of mature B cells and CLL cells into the bonemarrow and spleens of xenotransplanted mice. Toinvestigate whether the observed internalization of theCXCR4 receptor induced by PIM inhibition is alsofunctional in vivo, we pretreated CFSE-stained CLL cellswith the PIM kinase inhibitor K00135 (10 mmol/L) orwith DMSO for 12 hours. Then, CLL cells were collectedfor in vivo experiments and some cells were separated toassess apoptosis. Apoptosis was measured usingAnnexin V/7-AAD staining and cell viability was great-er than 90% for both groups (Supplementary Fig. S3).Four irradiated Rag2�/�gc�/� mice per group wereeach transplanted with 3.5 � 107 CLL cells via tail vein

injection. After 4 hours, the mice were sacrificed and thehoming of CLL cells into spleens and bone marrow wasassessed by flow cytometry. CLL cells that were pre-treated with the PIM inhibitor showed a significantdecrease in their homing capacity for both organs (Fig.8). The amount of CLL cells (CD45þCFSEþCD19þ) in thebone marrow was 2.3% in the control group versus 1%in the treatment group (P ¼ 0.0293) and in the spleen23.3% versus 9.3%, respectively (P ¼ 0.0104; Fig. 8),indicating that interference with the CXCR4 receptorfunction also occurs in vivo.

siRNA knockdown of PIM1–3 in CLLs showsdiverging functions regarding regulation of CXCR4and apoptosis induction

To further dissect the PIM kinase–dependent functionsregarding apoptosis induction and CXCR4 receptor reg-ulation, we performed siRNA knockdown with two tothree siRNAs per PIM kinase in five different CLL sam-ples using the Amaxa technology as previously described(Fig. 9A; refs. 44, 45). siRNA knockdown of PIM1 induceda significant downregulation of the CXCR4 surfaceexpression with three different siRNAs targeting PIM1,but did not induce apoptosis in CLL cells (Fig. 9B and C).In contrast, PIM2 knockdown did not significantly altertheCXCR4 surface expression, but could induce apoptosisin CLL cells. PIM3 knockdown had the strongest proa-poptotic effect in CLL cells, but did not alter the CXCR4surface expression (Fig. 9B and C). To further dissect thediverging functions of PIM kinases, we also investigatedthe effect of the PIMknockdownonBADphosphorylationand total BAD protein levels. Interestingly, while PIM1did not alter BAD phosphorylation levels, both PIM2 andPIM3 reduced BAD phosphorylation and total BADlevels after PIM2 knockdown (Fig. 9D), which mightexplain the proapoptotic effects seen after knockdown ofPIM2 and PIM3 (32, 46–48).

Figure 8. PIM kinase inhibitionreduces migration of CLL cells intothe spleen and bone marrow ofRag2�/�gc�/� mice. CLL cellswere stained with CFSE and thentreated with K00135 versusDMSO for 12 hours. Cells werecollected and 3.5 � 107 CLL cellswere injected per irradiatedRag2�/�gc�/� mouse. Figureshows the percentage of CFSE/CD45þ/CD19þ cells in the spleen(P ¼ 0.0104) and the bone marrow(P ¼ 0.0293) 4 hours after injection(n ¼ 4, patient 4; paired t test).






Taken together, the proapoptotic effects of PIM inhibi-tors and their effect on CXCR4 downregulation from thecell surface are mediated by different PIM kinases, withPIM1 regulating the CXCR4 surface expression andPIM2/PIM3 supporting survival of CLL cells.

DiscussionThe CXCR4 receptor is crucial for homing, migration,

and retention of various hematopoietic cell types, such asHSCs, B cells, and also CLL cells, into the bone marrow(12, 49). Furthermore, CXCR4 signaling provides directprosurvival signals and protects CLL cells from the effectof chemotherapeutical agents in vivo (11, 15). The functionof the CXCR4 receptor is strongly dependent on its local-ization on the cell surface and is activated upon ligandbinding of CXCL12 (16). Therefore, current efforts arefocused ondirect CXCR4 inhibitors (TN14003, plerixafor),blocking the interaction of the CXCR4 receptor with theCXCL12 ligand (50, 51). Plerixafor is currently used for themobilization of HSCs from the bonemarrow and is testedin ongoing clinical trials for its effect on mobilizing CLLcells for chemosensitization away from their protectivemicroenvironment (52). Unfortunately, the use of CXCR4inhibitors induces a counter-regulatory upregulation ofCXCR4 on the cell surface of the target cells, resulting in ashort-lived (8–12 hours) and only partial mobilization of

CXCR4-expressing cells into the peripheral blood. Incontrast to CXCR4 antagonists blockingCXCR4 signaling,PIM inhibitors block the externalization of the CXCR4receptor and might therefore be able to circumvent thecounter-regulatory upregulation of CXCR4 upon plerix-afor treatment. Our data suggest that at least variousaspects of CXCR4 signaling in CLL cells can be affectedby PIM kinase inhibition. PIM kinase inhibition stronglyaffected the surface expression of the CXCR4 receptor onCLL cells, and induced dephosphorylation of the Ser339residue in the C-terminal tail of the receptor, resulting inreduced recycling of the CXCR4 receptor to the cell sur-face. Furthermore, PIM kinase inhibition was able toeffectively block various functions of the CXCR4 receptor,such as migration toward CXCL12, CXCL12-mediatedERK phosphorylation, and even in vivo homing of CLLcells into the bone marrow and the spleens of Rag2�/�

gc�/� mice. Therefore, single use of PIM inhibitors mightalready be sufficient for CLL-cell mobilization or could bean intelligent addition to the plerixafor treatment to avoidcounter-regulatoryCXCR4upregulation. Treatment stud-ies combining plerixafor and PIM inhibitors in CLL xeno-graft mouse models will be performed to proof thisconcept in vivo.

In concordance with those findings, the presence ofprotective stromal cells could not efficiently block apo-ptosis of CLL cells treated with PIM inhibitors. By using

Figure 9. PIM1 knockdowndownregulates CXCR4 surfaceexpression, whereas PIM2/3induces apoptosis in CLL cells. A,after 24 hours of siRNA-mediatedPIM kinase knockdown, CLL cellswere lysed and Western blotanalysis was stained for PIM1 (CLL17), PIM2 (CLL 19), and PIM3 (CLL17) versus b-actin. Knockdownefficiencywascalculated relative tob-actin using ImageJ software andcompared with NoSi (PIM1: si1,45%; si2, 73%; si3, 58%; PIM2:si1, 79%; si2, 68%; PIM3: si1,57%; si2, 65%). B, MFI of CXCR4surface expression after 24 hoursof PIM1, PIM2, and PIM3knockdown compared with NoSicontrol (n ¼ 3, paired t test; PIM1:si1, P ¼ 0.0349; si2, P ¼ 0.0117) inthree CLLs (n ¼ 3; CLL#17, #19,and #20). C, after 24 hours of PIMkinase knockdown, CLL cells werestained with Annexin V–PE/7-AAD.Viable cells were defined asnegative for Annexin V–PE/7-AAD staining and graph showsviable cells relative to the NoSicontrol (n ¼ 3, paired t test; PIM3:si1,P¼ 0.0482; si2,P¼ 0.0285). D,after 24 hours of siRNA-mediatedknockdown of PIM kinases 1–3,Western blot analysis was stainedfor pBAD and BAD versus b-actin.

Decker et al.





the PIM inhibitor K00135, we observed apoptosis rateshigher than 50% in CLL cells that were cultivated alone.Interestingly, PIM inhibition in CLL cells cocultivatedwith protective stromal cells showed similar rates ofapoptosis, indicating that the stromal protection, whichis at least partially mediated by CXCL12/CXCR4, is notfunctional in the presence of PIM inhibitors. As stromalcells have an intact PIM–CXCR4 axis, they can be directlyaffected by PIM kinase inhibitor treatment. Although theapoptosis induced inM2-10B4 stromal cells only occurredat very high inhibitor concentrations, we cannot excludethat the secretion of cytokines/integrins is already affect-ed at lower concentrations and might therefore indirectlyreduce the stroma-protective effect on CLL cells.The role of the three Pim kinases for normal and malig-

nant hematopoiesis shows strongly overlapping functionsat least in various murine lymphoma models. Pim3 wasable to compensate for the loss ofPim1andPim2 inMuLV-induced lymphomagenesis (23). Furthermore, Em-Myc-Em-Pim1double-transgenicmice showedenhanced tumordevelopment similar to that seen in Em-Myc-Em-Pim2transgenic mice, pointing toward redundant functions ofPIM kinases in Myc-dependent B-cell lymphoma devel-opment (22, 53).Previous work on PIM kinases in CLL has identified

PIM2 as a proapoptotic target in CLL regulating C-MYCandMCL-1 (41),whilePIM1andPIM3or the interaction ofPIM kinases with the CXCR4 receptor were not investi-gated. As previously described (28), we could confirmoverexpression of PIM2 in our patient cohort, but alsofound strong overexpression of PIM1 inmost CLLs and ofPIM3 in a portion of more than 50% of CLL samples asmeasured by qPCR andWestern blot analysis. As severalPIM kinases are overexpressed in CLL, we wanted todissect their functions by using siRNA knockdown ofeach PIM kinase in primary CLL samples. Interestingly,only knockdown of PIM1 could significantly alter surfaceexpression of the CXCR4 receptor in CLLs, while PIM3and PIM2 were not able to affect CXCR4 surface expres-sion inCLL. In linewith those findings, only PIM1 proteinlevels correlated with pCXCR4 protein levels and PIM1transcript and protein levels positively correlated withCXCR4 surface expression levels. Therefore, high PIM1expression might contribute to increased CXCR4 surfaceexpression identified on CLL cells, and can affect theinteraction of CLL cells with their microenvironment.

On the other hand, knockdown of PIM3 and to a lowerdegree also of PIM2 induced apoptosis inCLL cells,whichcould not be observed upon PIM1 knockdown. In linewith those findings, PIM2 and PIM3 knockdown couldreduce phospho-BAD levels and total BAD in CLL cells,which was not observed after PIM1 knockdown. Thoseresults strongly argue for nonredundant functions ofthose three PIM kinases in CLL and separate survivalfrom CXCR4 surface expression. Consequently, pan-PIMinhibitors should be developed for the treatment of CLL,to block PIM1/CXCR4–dependent CLL cell migrationand to induce PIM2/3–dependent apoptosis.

In general, our results show that PIM kinase inhibitorsare an effective treatment option for CLL not only bydirectly inducing proapoptotic pathways in CLL cells, butalso by blocking the interaction of CLL cells with theirprotectivemicroenvironment through theCXCR4 receptor.

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

Authors' ContributionsConception and design: S. Decker, J. Finter, C. DierksDevelopment of methodology: S. Decker, J. Finter, C. DierksAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): J. Finter, T.S. Mack, N. Gray, M. Follo, K. ZirlikAnalysis and interpretation of data (e.g., statistical analysis, biostatis-tics, computational analysis): S. Decker, J. Finter, M. Follo, K. Zirlik,D. Pfeifer, C. DierksWriting, review, and/or revision of the manuscript: S. Decker, J. Finter,T.S. Mack, M. Burger, H. Eibel, H. Veelken, C. DierksAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): J. Finter, A.J. Forde, S. Kissel,J. Schwaller, A. Kuhn, M. Burger, H. Veelken, C. DierksStudy supervision: H. Jumaa, C. DierksInhibitor provision: C.V. Miduturu

AcknowledgmentsThe authors thank RolandMertelsmann for critically reading the article

and helpful advice.

Grant SupportThis work was supported by the Deutsche Krebshilfe (grant 110670) to

C. Dierks. H. Jumaa and C. Dierks are supported by the "BIOSS Centre forBiological Signalling Studies." Theposition ofC.Dierks is supported by theEmmy-Noether program of the DFG (DI 1664/1-1).

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 July 22, 2013; revised February 24, 2014; accepted February 26,2014; published OnlineFirst March 21, 2014.

References1. Munk Pedersen I, Reed J. Microenvironmental interactions and sur-

vival of CLL B-cells. Leuk Lymphoma 2004;45:2365–72.2. Chiorazzi N, Rai KR, Ferrarini M. Chronic lymphocytic leukemia. N Engl

J Med 2005;352:804–15.3. Rai KR, Sawitsky A, Cronkite EP, Chanana AD, Levy RN, Pasternack

BS. Clinical staging of chronic lymphocytic leukemia. Blood 1975;46:219–34.

4. Binet JL, Auquier A, Dighiero G, Chastang C, Piguet H, Goasguen J,et al. A new prognostic classification of chronic lymphocytic leukemiaderived fromamultivariate survival analysis.Cancer 1981;48:198–206.

5. DamleRN,Wasil T, Fais F, Ghiotto F, Valetto A, Allen SL, et al. Ig V genemutation status andCD38expression as novel prognostic indicators inchronic lymphocytic leukemia. Blood 1999;94:1840–7.

6. Wiestner A, Rosenwald A, Barry TS, Wright G, Davis RE, HenricksonSE, et al. ZAP-70 expression identifies a chronic lymphocyticleukemia subtype with unmutated immunoglobulin genes, inferiorclinical outcome, and distinct gene expression profile. Blood 2003;101:4944–51.

7. Rassenti LZ, Huynh L, Toy TL, Chen L, Keating MJ, Gribben JG, et al.ZAP-70 compared with immunoglobulin heavy-chain gene mutation






status as a predictor of disease progression in chronic lymphocyticleukemia. N Engl J Med 2004;351:893–901.

8. D€ohner H, Stilgenbauer S, Benner A, Leupolt E, Kr€ober A, Bullinger L,et al. Genomic aberrations and survival in chronic lymphocytic leuke-mia. N Engl J Med 2000;343:1910–6.

9. Hallek M, Cheson BD, Catovsky D, Caligaris-Cappio F, Dighiero G,D€ohner H, et al. Guidelines for the diagnosis and treatment ofchronic lymphocytic leukemia: a report from the International Work-shop on Chronic Lymphocytic Leukemia updating the NationalCancer Institute-Working Group 1996 guidelines. Blood 2008;111:5446–56.

10. Caligaris-Cappio F. Role of the microenvironment in chronic lympho-cytic leukaemia. Br J Haematol 2003;123:380–8.

11. Burger M, Hartmann T, KromeM, Rawluk J, Tamamura H, Fujii N, et al.Small peptide inhibitors of the CXCR4 chemokine receptor (CD184)antagonize the activation, migration, and antiapoptotic responses ofCXCL12 in chronic lymphocytic leukemia B cells. Blood 2005;106:1824–30.

12. Burger JA, Burger M, Kipps TJ. Chronic lymphocytic leukemia B cellsexpress functional CXCR4 chemokine receptors that mediate spon-taneous migration beneath bone marrow stromal cells. Blood 1999;94:3658–67.

13. Panayiotidis P, Jones D, Ganeshaguru K, Foroni L, Hoffbrand AV.Human bone marrow stromal cells prevent apoptosis and support thesurvival of chronic lymphocytic leukaemia cells in vitro. Br J Haematol1996;92:97–103.

14. Lagneaux L, Delforge A, Bron D, De Bruyn C, Stryckmans P. Chroniclymphocytic leukemic B cells but not normal B cells are rescued fromapoptosis by contact with normal bone marrow stromal cells. Blood1998;91:2387–96.

15. Kurtova AV, Balakrishnan K, Chen R, DingW, Schnabl S, Quiroga MP,et al. Diversemarrow stromal cells protect CLL cells from spontaneousand drug-induced apoptosis: development of a reliable and reproduc-ible system to assess stromal cell adhesion-mediated drug resistance.Blood 2009;114:4441–50.

16. Burger JA. Chemokines and chemokine receptors in chronic lympho-cytic leukemia (CLL): from understanding the basics towards thera-peutic targeting. Semin Cancer Biol 2010;20:424–30.

17. Oberlin E, Amara A, Bachelerie F, Bessia C, Virelizier JL, Arenzana-Seisdedos F, et al. TheCXC chemokine SDF-1 is the ligand for LESTR/fusin and prevents infection by T-cell-line–adapted HIV-1. Nature1996;382:833–5.

18. Bleul CC, Fuhlbrigge RC, Casasnovas JM, Aiuti A, Springer TA. Ahighly efficacious lymphocyte chemoattractant, stromal cell–derivedfactor 1 (SDF-1). J Exp Med 1996;184:1101–9.

19. Till KJ, Lin K, Zuzel M, Cawley JC. The chemokine receptor CCR7 andalpha4 integrin are important for migration of chronic lymphocyticleukemia cells into lymph nodes. Blood 2002;99:2977–84.

20. Trentin L, Cabrelle A, Facco M, Carollo D, Miorin M, Tosoni A, et al.Homeostatic chemokines drive migration of malignant B cells inpatients with non-Hodgkin lymphomas. Blood 2004;104:502–8.

21. Cuypers HT, Selten G, Quint W, Zijlstra M, Maandag ER, Boelens W,et al. Murine leukemia virus-induced T-cell lymphomagenesis: inte-gration of proviruses in a distinct chromosomal region. Cell 1984;37:141–50.

22. van LohuizenM, Verbeek S, Krimpenfort P, Domen J, Saris C, Radasz-kiewicz T, et al. Predisposition to lymphomagenesis in pim-1 trans-genic mice: cooperation with c-myc and N-myc in murine leukemiavirus-induced tumors. Cell 1989;56:673–82.

23. Mikkers H, Allen J, Knipscheer P, Romeijn L, Hart A, Vink E, et al. High-throughput retroviral tagging to identify components of specific sig-naling pathways in cancer. Nat Genet 2002;32:153–9.

24. Qian KC, Wang L, Hickey ER, Studts J, Barringer K, Peng C, et al.Structural basis of constitutive activity and a unique nucleotidebinding mode of human Pim-1 kinase. J Biol Chem 2005;280:6130–7.

25. Yip-Schneider MT, Horie M, Broxmeyer HE. Transcriptional inductionof pim-1 protein kinase gene expression by interferon gamma andposttranscriptional effects on costimulation with steel factor. Blood1995;85:3494–502.

26. Mui AL, Wakao H, Kinoshita T, Kitamura T, Miyajima A. Suppres-sion of interleukin-3–induced gene expression by a C-terminaltruncated Stat5: role of Stat5 in proliferation. EMBO J 1996;15:2425–33.

27. Alizadeh AA, EisenMB, Davis RE,MaC, Lossos IS, Rosenwald A, et al.Distinct types of diffuse large B-cell lymphoma identified by geneexpression profiling. Nature 2000;403:503–11.

28. Cohen AM, Grinblat B, Bessler H, Kristt D, Kremer A, Schwartz A, et al.Increased expression of the hPim-2 gene in human chronic lympho-cytic leukemia and non-Hodgkin lymphoma. Leuk Lymphoma 2004;45:951–5.

29. Mizuki M, Schwable J, Steur C, Choudhary C, Agrawal S, Sargin B,et al. Suppression of myeloid transcription factors and induction ofSTAT response genes by AML-specific Flt3 mutations. Blood 2003;101:3164–73.

30. ValdmanA, FangX,PangS-T, EkmanP, Egevad L. Pim-1 expression inprostatic intraepithelial neoplasia and human prostate cancer. Pros-tate 2004;60:367–71.

31. LillyM, Sandholm J, Cooper JJ, KoskinenPJ, Kraft A. ThePIM-1 serinekinase prolongs survival and inhibits apoptosis-related mitochondrialdysfunction in part through a bcl-2-dependent pathway. Oncogene1999;18:4022–31.

32. Yan B, Zemskova M, Holder S, Chin V, Kraft A, Koskinen PJ, et al. ThePIM-2 kinase phosphorylates BAD on serine 112 and reverses BAD-induced cell death. J Biol Chem 2003;278:45358–67.

33. Aho TLT, Sandholm J, Peltola KJ, Mankonen HP, Lilly M, Koskinen PJ.Pim-1 kinase promotes inactivation of the pro-apoptotic Bad proteinby phosphorylating it on the Ser112 gatekeeper site. FEBS Lett 2004;571:43–9.

34. Wang Z, Bhattacharya N, Mixter PF, Wei W, Sedivy J, Magnuson NS.Phosphorylation of the cell cycle inhibitor p21Cip1/WAF1 by Pim-1kinase. Biochim Biophys Acta 2002;1593:45–55.

35. Zhang Y,Wang Z,Magnuson NS. Pim-1 kinase-dependent phosphor-ylation of p21Cip1/WAF1 regulates its stability and cellular localizationin H1299 cells. Mol Cancer Res 2007;5:909–22.

36. Morishita D, Katayama R, Sekimizu K, Tsuruo T, Fujita N. Pim kinasespromote cell cycle progression by phosphorylating and down-regu-lating p27Kip1 at the transcriptional and posttranscriptional levels.Cancer Res 2008;68:5076–85.

37. Grundler R, Brault L, Gasser C, Bullock AN, DechowT,Woetzel S, et al.Dissection of PIM serine/threonine kinases in FLT3-ITD–induced leu-kemogenesis reveals PIM1 as regulator of CXCL12-CXCR4–mediatedhoming and migration. J Exp Med 2009;206:1957–70.

38. Pfeifer D, Pantic M, Skatulla I, Rawluk J, Kreutz C, Martens UM, et al.Genome-wide analysis of DNA copy number changes and LOH in CLLusing high-density SNP arrays. Blood 2007;109:1202–10.

39. Miduturu CV, Deng X, Kwiatkowski N, Yang W, Brault L, Filippako-poulos P, et al. High-throughput kinase profiling: a more efficientapproach toward the discovery of new kinase inhibitors. Chem Biol2011;18:868–79.

40. Goldman JP, Blundell MP, Lopes L, Kinnon C, Di Santo JP, ThrasherAJ. Enhanced human cell engraftment in mice deficient in RAG2 andthe common cytokine receptor gamma chain. Br J Haematol 1998;103:335–42.

41. Chen LS, Redkar S, Bearss D, Weirda WG, Gandhi V. Pim kinaseinhibitor, SGI-1776, induces apoptosis in chronic lymphocytic leuke-mia cells. Blood 2009;114:4150–7.

42. Pogacic V, Bullock AN, Fedorov O, Filippakopoulos P, Gasser C,Biondi A, et al. Structural analysis identifies imidazo[1,2-b]pyridazinesas PIM kinase inhibitors with in vitro antileukemic activity. Cancer Res2007;67:6916–24.

43. Decker S, Zirlik K, Djebatchie L, Hartmann D, Ihorst G, Schmitt-GraeffA, et al. Trisomy 12 and elevated GLI1 and PTCH1 transcript levels arebiomarkers for Hedgehog-inhibitor responsiveness in CLL. Blood2012;119:997–1007.

44. Gresch O, Engel FB, Nesic D, Tran TT, England HM, Hickman ES, et al.New non-viral method for gene transfer into primary cells. Methods2004;33:151–63.

45. AdamM,PogacicV,BenditM,ChappuisR,NawijnMC,Duyster J, et al.Targeting PIM kinases impairs survival of hematopoietic cells


Decker et al.




transformedby kinase inhibitor-sensitive and kinase inhibitor-resistantforms of Fms-like tyrosine kinase 3 and BCR/ABL. Cancer Res 2006;66:3828–35.

46. Santio NM, Vahakoski RL, Rainio EM, Sandholm JA, Virtanen SS,Prudhomme M, et al. Pim-selective inhibitor DHPCC-9 reveals Pimkinases as potent stimulators of cancer cell migration and invasion.Mol Cancer 2010;9:279.

47. Chen LS, Redkar S, Taverna P, Cortes JE, Gandhi V. Mechanisms ofcytotoxicity to Pim kinase inhibitor, SGI-1776, in acute myeloid leu-kemia. Blood 2011;118:693–702.

48. Xu D, Allsop SA, Witherspoon SM, Snider JL, Yeh JJ, Fiordalisi JJ,et al. The oncogenic kinase Pim-1 is modulated by K-Ras signalingand mediates transformed growth and radioresistance in humanpancreatic ductal adenocarcinoma cells. Carcinogenesis 2011;32:488–95.

49. Nagasawa T. A chemokine, SDF-1/PBSF, and its receptor, CXCchemokine receptor 4, as mediators of hematopoiesis. Int J Hematol2000;72:408–11.

50. Fujii N, Nakashima H, Tamamura H. The therapeutic potential ofCXCR4 antagonists in the treatment of HIV. ExpertOpin Investig Drugs2003;12:185–95.

51. TamamuraH,OmagariA,HiramatsuK,GotohK,KanamotoT,XuY,etal.Development of specific CXCR4 inhibitors possessing high selectivityindexes as well as complete stability in serum based on an anti-HIVpeptide T140. Bioorg Med Chem Lett 2001;11:1897–902.

52. Calandra G, Bridger G, Fricker S. CXCR4 in clinical hematology. CurrTop Microbiol Immunol 2010;341:173–91.

53. Allen JD, Verhoeven E, Domen J, van der Valk M, Berns A. Pim-2transgene induces lymphoid tumors, exhibiting potent synergy withc-myc. Oncogene 1997;15:1133–41.






2014;13:1231-1245. Published OnlineFirst March 21, 2014.Mol Cancer Ther Sarah Decker, Johannes Finter, Aaron James Forde, et al. Interactions (PIM1)Survival (PIM2/3) and CXCR4-Mediated Microenvironmental PIM Kinases Are Essential for Chronic Lymphocytic Leukemia Cell

Updated version

10.1158/1535-7163.MCT-13-0575-Tdoi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2014/03/21/1535-7163.MCT-13-0575-T.DC1

Access the most recent supplemental material at:

Cited articles

http://mct.aacrjournals.org/content/13/5/1231.full#ref-list-1

This article cites 53 articles, 24 of which you can access for free at:

Citing articles

http://mct.aacrjournals.org/content/13/5/1231.full#related-urls

This article has been cited by 2 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://mct.aacrjournals.org/content/13/5/1231To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-13-0575-T

http://mct.aacrjournals.org/content/suppl/2014/03/21/1535-7163.MCT-13-0575-T.DC1

http://mct.aacrjournals.org/content/13/5/1231.full#ref-list-1

http://mct.aacrjournals.org/content/13/5/1231.full#related-urls

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

mailto:[email protected]

http://mct.aacrjournals.org/content/13/5/1231


Documents

PIM Kinases Are Essential for Chronic Lymphocytic Leukemia ...mct.aacrjournals.org/content/molcanther/13/5/1231.full.pdfPIM Kinases Are Essential for Chronic Lymphocytic Leukemia Cell