CIP4ControlsCCL19-DrivenCellSteeringandChemotaxisin ... · Corresponding Author: Loïc Dupre,...

Tumor and Stem Cell Biology

CIP4Controls CCL19-DrivenCell Steering andChemotaxis inChronic Lymphocytic Leukemia

Gema Malet-Engra1,4,5,7, Julien Viaud2, Loïc Ysebaert3,6, Manon Farc�e1,4,5, Fanny Lafouresse1,4,5,Guy Laurent3,6, Fr�ed�erique Gaits-Iacovoni2, Giorgio Scita7,8, and Loïc Dupr�e1,4,5

AbstractSolid tumor dissemination relies on the reprogramming of molecular pathways controlling chemotaxis.

Whether the motility of nonsolid tumors such as leukemia depends on the deregulated expression of moleculesdecoding chemotactic signals remains an open question. We identify here the membrane remodeling F-BARadapter protein Cdc42-interacting protein 4 (CIP4) as a key regulator of chemotaxis in chronic lymphocyticleukemia (CLL). CIP4 is expressed at abnormally high levels in CLL cells, where it is required for CCL19-inducedchemotaxis. Upon CCL19 stimulation of CLL cells, CIP4 associates with GTP-bound Cdc42 and is recruited to therear of the lamellipodiumand alongmicrospikes radiating through the lamellipodium. Consistentwith its cellulardistribution, CIP4 removal impairs both the assembly of the polarized lamellipodium and directional migrationalong a diffusible CCL19 gradient. Furthermore, CIP4 depletion results in decreased activation of WASP, butincreased activation of PAK1 and p38 mitogen-activated protein kinase (MAPK). Notably, p38 MAPK inhibitionresults in impaired lamellipodium assembly and loss of directional migration. This suggests that CIP4 modulatesboth theWASP and p38MAPK pathways to promote lamellipodium assembly and chemotaxis. Overall, our studyreveals a critical role of CIP4 in mediating chemotaxis of CLL cells by controlling the dynamics of microspike-containing protrusions and cell steering. Cancer Res; 73(11); 3412–24. �2013 AACR.

IntroductionThere is growing evidence that the dissemination of tumor

cells is governed by a reprogramming of themolecularmachin-ery controlling motility (1). Solid tumors with high dissemi-nation potential have been shown to upregulate actin cyto-skeleton regulators that decode chemotactic cues for theassembly of protrusions supporting migration and invasion(2–4). In particular, high expression of members of the WASPfamily facilitates the dissemination of several cancers bypromoting the assembly of filopodia, lamellipodia, and inva-dopodia (5, 6). The formation of membrane protrusionsinvolved in migration requires a high degree of coordinationbetween actin cytoskeleton remodeling and induction ofmem-brane curvature. One mechanism to achieve this coordination

is mediated by BAR domain-containing proteins, which sensemembrane curvature, mold the plasmamembrane, and recruitactin nucleators such as WASP- and formin-family proteins (7,8). The capacity of tumor cells to harness BARdomain-contain-ing proteins is supported by the finding that invasive breastcancer cells overexpress the F-BAR domain-containing Cdc42-interacting protein 4 (CIP4), which regulates invadopodiaformation and invasion through the extracellularmatrix (9, 10).

The dissemination of hematologic malignancies is thoughtto be derived from the intrinsic ability of hematopoietic cellsand in particular lymphocytes to patrol the body and migratebetween distant tissue compartments (11). However, malig-nant B cells assemble aberrant actin-rich protrusions endow-ing these cells with high migratory and invasive potential (12).Chronic lymphocytic leukemia (CLL), which is a commonlympho-proliferative disorder of mature B lymphocytes (13),is characterized by tumor cells highly responsive to chemo-tactic cues that circulate at high rate through the blood andaccumulate in the bone marrow and lymph nodes (14–18).These aspects of the CLL phenotype may be linked to theirpropensity to assemble actin-rich structures in response tochemotactic signals, thereby altering normal leukocyte hom-ing and homeostasis (12, 19). Accordingly, the actin regulatorHS1 is hyper-phosphorylated in patients with CLL, regulates invivo migration, and is associated with poor clinical outcome(20, 21). The trafficking and homing of hematopoietic cells tothe lymphatic tissues and in particular their migration acrossthe vascular endothelium are controlled in part by CCR7, thereceptor for the CCL19 andCCL21 chemokines (22). CCR7 levelin CLL cells correlates with lymphadenopathy (15, 23) and is

Authors' Affiliations: 1Institut National de la Sant�e et de la RechercheM�edicale (INSERM), U1043, Centre de Physiopathologie de ToulousePurpan; 2INSERM, U1048; 3INSERM, U1037, Centre de Recherche enCanc�erologie de Toulouse; 4Centre National de la Recherche Scientifique(CNRS), U5282; 5Universit�e Toulouse III Paul-Sabatier; 6Department ofHematology, Purpan University Hospital, Toulouse, France; 7IFOM (FIRCInstitute for Molecular Oncology); and 8Dipartimento di Scienze dellaSalute, Universit�a degli Studi di Milano, Milan, Italy

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

Corresponding Author: Loïc Dupr�e, Institut National de la Sant�e et de laRecherche Medicale (INSERM) U1043, Purpan University Hospital, 31300Toulouse, France. Phone: 33-0562744588; Fax: 33-0562744558; E-mail:[email protected]

doi: 10.1158/0008-5472.CAN-12-3564

�2013 American Association for Cancer Research.

CancerResearch

Cancer Res; 73(11) June 1, 20133412

on March 21, 2021. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 3, 2013; DOI: 10.1158/0008-5472.CAN-12-3564

http://cancerres.aacrjournals.org/

higher in ZAP-70þ/CD38þ CLL cells, whose presence is relatedto a poor prognosis (24). Therefore, CLL disease progressionmay correlate with the development of cells that acquireincreased potential to invade lymph nodes via the CCL19/21-CCR7 axis.Our objective was to investigate whether regulators of actin

cytoskeleton and membrane curvature are abnormallyexpressed in CLL to promote assembly of promigratory struc-tures in response to CCL19. We identified the F-BAR domain-containing protein CIP4 as specifically overexpressed in CLLwhen compared with normal B cells or other subtypes of B-cellmalignancies. By combining siRNA/shRNA approaches withreal-time tracking of cells migrating in CCL19 gradients, weshow that CLL cells depend onCIP4 for the assembly of a highlystructured lamellipodium and for directional migration inresponse to CCL19.

Materials and MethodsPatients and cell linesPeripheral blood samples were collected from patients with

CLL after obtaining informed consent and approval from theToulouse University Medical Center Ethic Committee. Afterseparation by Ficoll centrifugation, cells were used freshly orstored as frozen samples (HIMIP collection). According to theFrench law, HIMIP collection has been declared to theMinistryof Higher Education and Research (DC 2008-307 collection 1)and obtained a transfer agreement (AC 2008-129) after appro-bation by the local ethical committee (Comit�e de Protectiondes Personnes Sud-Ouest et Outremer II). Clinical and biologicannotations of the samples have been declared to the nationalcommittee on data processing (Comit�e National Informatiqueet Libert�es). For some experiments, B-leukemic cells werepurified by magnetic separation according to manufacturer'sinstructions (Easy Sep Human BCell Enrichment kit w/o CD43depletion; Stem cell Technologies). Control B cells fromhealthy donors were isolated from buffy coats (EtablissementFrancais du Sang) using a negative selection kit (EasySepHuman B Cell Enrichment Kit; Stem Cell Technologies). B-cellpurity was assessed by flow cytometry. The CLL-derived celllines JVM3 and MEC2 were purchased from DSMZ (n� ACC18and ACC500, respectively) authenticated by flow cytometryanalysis of B-cell surface markers and tested regularly for theabsence of Mycoplasma infection.

Quantitative reverse transcription PCRTotal RNAwas isolated frompurifiedB cells using theTRIzol

reagent (Invitrogen). Two microgram of RNA was used, with100 ng random hexamers, in a reverse transcription reaction(SUPERSCRIPT II; Invitrogen). One-tenth nanogram of cDNAwas amplified, in triplicate, in a reaction volume of 25 mL with10 pmol/L of each gene-specific primer and the SYBR-greenPCR MasterMix (Applied Biosystems). Real-time PCR wasconducted on the ABI/Prism 7700 Sequence Detector System(PerkinElmer/Applied Biosystems), using a pre-PCR step of 10minutes at 95�C, followed by 40 cycles of 15 seconds at 95�Cand 60 seconds at 60�C. Specificity of the amplified productswas confirmed by melting curve analysis (Dissociation Curve;PerkinElmer/Applied Biosystems) and by 6% PAGE. Prepara-

tions with RNA template without reverse transcriptase wereused as negative controls. Samples were amplified with pri-mers for each gene [for details, see quantitative PCR (qPCR)primer list in Supplementary Table S1] and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a housekeeping gene(other housekeeping genes, including rRNA 18S and b-actinwere also tested with comparable results). The Ct values werenormalized to the GAPDH curve and the relative expression ofeach gene was expressed as the ratio relative to control B cellsisolated from the blood of 3 healthy donors.

Western blot analysis and GST pull-downCells were lysed with Cytobuster Protein Extraction Reagent

(Novagen) containing NaF, Na3VO4 (1 mmol/L each), andprotease inhibitor cocktail (Sigma). Proteins from total celllysates were resolved by gel SDS-PAGE and analyzed by West-ern blotting. The following primary antibodies were used toreveal the membranes: anti-CIP4 (clone 21; BD TransductionLaboratories), anti-GAPDH (clone 374; Chemicon), anti-Cdc42(sc-87; Santa Cruz Biotechnology), anti-total P38 (9212; CellSignaling), anti-pP38 (clone 3D7; Cell Signaling), anti-pErk1/2(E10; Cell Signaling), anti-PAK1/2/3 (2604; Cell Signaling), anti-pPAK1 (Ser199/204)/PAK2 (Ser192/197; 2605; Cell Signaling),anti-WASP (H-250; Santa Cruz Biotechnology), anti-pWASP(phospho Y-290; Abcam), anti-FBP17 (gift from Dr. P. DeCamilli, Yale University School of Medicine, New Haven, CT;ref. 25) and anti-Toca-1 (developed in Dr. G. Scita's laboratory).Horseradish peroxidase-conjugated secondary antibodieswere from Promega. Where indicated, JVM3 cells were pre-treated for 10 minutes with either dimethyl sulfoxide (DMSO)or 50 mmol/L IPA-3 (Sigma). Cells were then stimulated for 5minutes with 100 ng/mL CCL19 before lysis. Recombinantglutathione S-transferase (GST) and GST-Cdc42 were purifiedfrom Escherichia coli and the immobilized GST-Cdc42 proteinwas preloaded with GDP or GTPgS as described previously(26). Cell lysates were prepared for GST pull-down experimentsas described earlier and incubated with 20 mg of each fusionprotein bound to glutathione-sepharose beads for 2 hours at4�C. The beads were washed 4 times with PBST (1% Tween 20)and boiled at 95�C for 5 minutes. Samples were subjected toSDS-PAGE and analyzed byWestern blotting with CIP4 mono-clonal antibody (mAb). Active Cdc42 pull-downwas conductedas previously described (27).

siRNA and shRNA-mediated CIP4 knockdownTransient CIP4 knockdown was obtained by nucleofection

of JVM3 cells and patient CLL cells with 500 nmol/L CIP4-specific siRNA (GGAGAAUAGUAAGCGUAAA) or scrambledcontrol siRNA (GGAATACGAGAGAAATGAT) purchased fromDharmacon. Cells were nucleofected with the Amaxa system(Amaxa Biosystems/Lonza) using Nucleofector Solution V andNucleofector program X-001 for JVM3 cells and U-013 forpatient CLL cells. Stable CIP4 knockdown was obtained bytransduction of JVM3 cells with MISSION lentiviral transduc-tion particles (Sigma) encoding 2 short hairpin RNAs (shRNA)targeting CIP4 (TRCN063186 and TRCN299602). As negativecontrols, MISSION nontarget shRNA control transductionparticles (SHC002V and SHC202V) were used. Stable

CIP4 Regulates Lamellipodium Assembly

www.aacrjournals.org Cancer Res; 73(11) June 1, 2013 3413




knockdown cell lines were generated by selecting transducedcells with 1 mg/mL puromycin. Silencing efficiency was con-trolled by Western blotting.

Transwell chemotaxis assayCells (JVM3 or purified leukemic cells from patients with

CLL) were resuspended in RPMI, 0.5% bovine serum albumin(BSA). A total of 300 mL containing 0.5 � 105 patient cells or0.1 � 105 JVM3 cells was loaded on the top chamber of aTranswell culture PET (polyethylene terephtalate) membraneinsert (BD Falcon) with a pore size of 3 mm. Filters weretransferred into wells containing 1,000 mL RPMI, 0.5% BSAwith varying concentrations of CCL19 (Peprotech). Transwellplates were incubated at 37�C in 5% CO2 for 6 hours for patientcells and overnight for JVM3 cells. Inserts were then removedand cells that transmigrated in the lower wells were resus-pended in 250 mL fluorescence-activated cell sorting (FACS)buffer plus 50 mL calibration beads (BD Biosciences). The samenumber of beads was acquired for each sample and cells werecounted using a FACSCalibur flow cytometer (BDBiosciences).The migration index was calculated as the number of cellstransmigrating toward CCL19 divided by the number of cellstransmigrating spontaneously in the absence of CCL19.

Real-time chemotaxisChemotaxis of JVM3 cells was analyzed in real-time using 2-

dimensional (2D) chemotaxis m-slides coated with Collagen IV(Ibidi). Cells (0.3 � 106 cells in 10 mL) were loaded into thecentral transversal chamber and incubated at 37�C for 50minutes to allow cell attachment. Fresh RPMI was loaded intoadjacent reservoirs and a CCL19 gradient was created follow-ing the manufacturer's instructions. Cell migration wasrecorded at 37�C, 5% CO2 by taking a picture every 2 minutesusing an invertedwide-fieldmicroscope (Zeiss Axiovert) with a2.5� objective and MetaMorph32 software. Images were ana-lyzed with the ImageJ 1.42q software (W.S. Rasband, NationalInstitute of Health, Bethesda, MD), using the Manual Trackingplugin (F. Cordeli�eres, Institut Curie, Orsay, France). Chemo-taxis plots and migration parameters were obtained with theChemotaxis and Migration tool from Ibidi.

Confocal laser scanning microscopyTo visualize and measure lamellipodium formation, JVM3

cells or primary CLL cells were seeded for 5 minutes on glassslides precoated with 4 mg/mL ICAM-1, in the presence of 100ng/mL soluble CCL19. Cells were fixed with 3% paraformal-dehyde, permeabilized in 3% BSA/0.1% saponin, and stained in3% BSA/0.1% saponin with either anti-a-tubulin rabbit pAb(Abcam) followed by goat anti-rabbit mAb labeled with AlexaFluor 488 or phalloidin labeled with Alexa Fluor 633. They wereadditionally stained with 40,6-diamidino-2-phenylindole(DAPI) before mounting. Pictures of individual JVM3 cellscollected with a �63 objective mounted on a Confocal Zeiss710 microscope were analyzed on ImageJ by delineating man-ually the borders of the lamellipodium to calculate its surface.To investigate the effect of p38 inhibition on CCL19-inducedlamellipodium assembly, lamellipodium surfacewasmeasuredas above in JVM3 cells that were treated for 1 hour with 10

mmol/L SB203580 (Calbiochem) or DMSO, before CCL19 expo-sure. To study the subcellular localization of CIP4 and Cdc42,JVM3 cells activated for 5 minutes with immobilized ICAM-1and soluble CCL19 were fixed, permeabilized (as above), andstained with anti-Cdc42 rabbit pAb (Santa Cruz Biotechnolo-gy) and anti-CIP4 mAb (clone 21; BD Transduction Laborato-ries), followed by goat anti-rabbitmAb labeledwithAlexa Fluor488 and goat anti-mouse mAb labeled with Alexa Fluor 546.The relative localization of Cdc42 and CIP4 was analyzed usingthe Linescan function of ImageJ. To study the subcellularlocalization of CIP4 in migrating cells, 3� 106 JVM3 cells weretransiently nucleofected with 1 to 3 mg of a plasmid encodingGFP–CIP430. After 24 hours, dead cells were removed using aDead Cell Removal Kit (Miltenyi). For time-lapse video record-ing, cells were placed in Ibidi chemotaxis chambers and aConfocal Zeiss 710 microscope was used to track cells every 8seconds using a �63 objective.

ResultsCIP4 is expressed at high levels in CLL cell lines and isrequired for chemotaxis toward CCL19

Given the involvement of WASP-family proteins in theregulation of tumor cell migration, expression of a selectedset of WASP-family proteins (WASP, N-WASP, and WAVE1-3)and direct interactors (WIP, FBP17, and Toca-1, CIP4) wasanalyzed by quantitative reverse transcription PCR (qRT-PCR)in the JVM3 and MEC2 CLL cell lines. Levels of mRNA in theCLL cell lines were normalized to those measured in normalperipheral blood B cells. This analysis revealed that the mRNAencoding the F-BAR domain-containing CIP4 was expressed atelevated levels in both CLL cell lines (Fig. 1A). The other mRNAanalyzed in the CLL cell lines was either mildly overexpressed(WASP, WIP, WAVE1-2, and FBP17), expressed at levels com-parable with those of control B cells (N-WASP, Toca-1), orundetected (WAVE3). Accordingly, the CIP4 protein was alsooverexpressed in JVM3 and MEC2 when compared with nor-mal peripheral blood B cells (Fig. 1B). CIP4 overexpressionseemed to be restricted to the CLL cell lines as CIP4 expressionin cell lines corresponding to different B-cell malignancies wascomparable or even lower to that in normal B cells (Fig. 1B).Further analysis of the levels of the CIP4-related proteins Toca-1 and FBP17 showed that CIP4 overexpression is a specificfeature of this protein as there no concomitant increase norcompensation by the othermembers of the CIP4 family, Toca-1and FBP17 (Fig. 1B). Because of its high expression in CLL celllines and its membrane–actin interface regulatory role, weformulated the hypothesis that CIP4 might regulate CLL cellmigration. The contribution of CIP4 to CLL cell migration wastested by knocking down CIP4 expression in JVM3 cells andmeasuring the ability of those cells to migrate toward CCL19.Nucleofection with CIP4 siRNA efficiently and reproduciblydepleted JVM3 cells in CIP4 (Fig. 1C). Transwell assays showedthat the migration of cells knocked down for CIP4 toward aCCL19 gradient was significantly reduced (Fig. 1D). To testwhether this decrease in chemotactic response could be theresult of either defective motility or defective orientationwithin the chemokine gradient, single cell trackswere recordedin Ibidi chemotaxis m-slides. Most CIP4-expressing JVM3 cells

Malet-Engra et al.

Cancer Res; 73(11) June 1, 2013 Cancer Research3414




(scrambled siRNA) followed migratory paths directed towardthe source of CCL19 (Fig. 1E). A similar pattern of migrationwas obtained for mock-nucleofected JVM3 cells (data notshown). Conversely, JVM3 cells knocked down for CIP4 movedalong tracks that followed random directions, similarly to cellsnot exposed to any CCL19 gradient (Fig. 1E). Measurement ofthe forward migration index along the y-axis (coincident with

the direction of the CCL19 gradient) confirmed that CIP4-expressing JVM3 cells migrated directionally toward CCL19(Fig. 1F). In contrast, CIP4 knocked down cells displayed noforward migration, supporting the finding that cells lackingCIP4 displayed a random migration pattern (Fig. 1F). CIP4knocked down cells remained motile both in basal conditionsand in response to CCL19, with migration speed comparable

BA

mR

NA

level (A

U)

0

10

20

30

40

50

B cells MEC2JVM3

DO

HH

2

LIB

GR

AN

TA

RL

DE

AU

ME

C2

RE

C1

JV

M3

DLBCLMCLFLCLL

B c

ells

CIP4

GAPDH

Vinculin

Toca-1

FBP17

C

0

10

20

30

40

50

% M

igra

ting c

ells

+–

Scr

+–

CIP4

CCL19 :

SiRNA :

* P = 0.03D

JVM3

CIP4ScrSiRNA :

CIP4

GAPDH

CIP4Scr

CC

L19 (

ng/m

L)

0

100

SiRNA :

No C

CL19

E

Velo

city (

µm

/min

)

0

1

2

3

4

5

Fo

rwa

rd

mig

ration index

–0.2

0

0.2

0.4

CCL19 :

SiRNA :

+–

Scr

+–

CIP4

ns

* P = 0.01F

Number of tracks: 29 counts up: 2 counts down: 27

Center of mass (μm): x = -1.42 y = -327.13

x axis [unit]

–1,000 1,000–500 5000 –1,000 1,000–500 5000

–1,000

1,000

800

600

400

200

0

–200

–400

–600

–800

–1,000

1,000

800

600

400

200

0

–200

–400

–600

–800

–1,000

1,000

800

600

400

200

0

–200

–400

–600

–800

–1,000

1,000

800

600

400

200

0

–200

–400

–600

–800

–1,000

1,000–500 5000 –1,000 1,000–500 5000

x axis [μm]

x axis [μm]

y axis

[μm

]

y axis

[μm

]y

axis

[μm

]

y axis

[unit]

x axis [μm]


Center of mass (μm): x = -130.16 y = 27.48


Center of mass (μm): x = -14.62 y = 52.53Number of tracks: 30 counts up: 23 counts down: 7

Center of mass (μm): x = -244.5 y = 91.16

Figure 1. CIP4 is overexpressed in CLL cell lines and regulates the directed migration of JVM3 cells within a CCL19 gradient. A, gene expression levels of9 selected regulators of actin cytoskeleton and membrane remodeling were measured by qRT-PCR. RNA was isolated from 2 CLL-derived lines (JVM3and MEC2). To normalize the relative gene expression levels, a value of 1 was attributed to the expression levels of normal peripheral blood B cells (n ¼ 3).B, Western blot analysis of CIP4, Toca-1, and FBP17 in normal peripheral blood B cells and in cell lines, models of CLL (JVM3 and MEC2), follicularlymphoma (FL; RL and DOHH2), mantle cell lymphoma (MCL; GRANTA and REC1) and diffuse large B-cell lymphoma (DLBCL; DEAU and LIB). GAPDH andvinculin are shown as loading controls. C, CIP4 expression level analyzed by Western blotting in JVM3 cells 72 hours following nucleofection with theindicated siRNA (control scrambled sequence or sequence targeting CIP4). D, CCL19-driven chemotaxis of control and CIP4 knocked down JVM3 cells wasassessed in a Transwell assay by measuring the migration index in response to 100 ng/mL CCL19. E, comparison of the chemotactic tracks of JVM3nucleofected with the control scrambled siRNA or the CIP4 siRNA. Tracks correspond to 16-hour trajectories of individual cells moving in a 2D chemotacticchamber loaded with 100 ng/mL CCL19. The data are from one representative experiment out of 3 carried out under the same conditions. F, analysisofY forwardmigration indexandspeed fromCIP4 knockeddownandcontrol JVM3cells. The data represent themeanandSEMof 3 independent experiments(each experiment including the analysis of 10–30 tracks per cell type). ns, nonsignificant.






with that of scrambled siRNA-treated cells (Fig. 1F). Therefore,those results show that CIP4 is dispensable for JVM3 cellmotility, but is required for CCL19-evoked cell steering andchemotaxis.

CIP4 is required for CCL19-driven lamellipodiumassembly in JVM3 cells

Because the best-characterized function of CIP4 is to reg-ulate receptor endocytosis, the abrogation of chemotaxisconsecutive to CIP4 knockdown might be linked to an abnor-mal control of CCL19-induced CCR7 internalization. However,JVM3 cells depleted for CIP4 by shRNA-encoding lentiviralvectors expressed normal basal levels of surface CCR7, whichwas equally internalized in CIP4 knocked down and controlcells (Supplementary Fig. S1). This result indicates that CIP4control of CCL19-driven chemotaxis does not result fromaltered CCR7 endocytosis. Through its F-BAR domain, CIP4might also associate to membrane protrusions in migrating

cells.We tested this possibility by comparing themorphologiesof JVM3 cells depleted or not for CIP4. Independently of CIP4expression, JVM3 cells plated on ICAM-1 displayed variousmorphologies and emitted numerous types of membraneprotrusions. A hallmark of CCL19 stimulation in CIP4-expres-sing JVM3 cells was the assembly of a lamellipodium, whichconsisted of a wide flat protrusion made of a veil embeddedwith microspikes. This structure was observed both in fixedcells following short exposure to ICAM-1 plus CCL19 (Fig. 2A)and in live cells exposed to a CCL19 gradient (SupplementaryFig. S2A). Real-time imaging revealed that this lamellipodiumis a dynamic structure leading the cells toward the source ofCCL19 (Supplementary Movie S1). At the opposite side of thelamellipodium, JVM3 cells emitted a thinner protrusion resem-bling a uropod. Furthermore, the microtubule organizingcenter (MTOC) was usually localized along the lamellipo-dium–uropod axis, mostly between the lamellipodium andthe nucleus. In contrast, cells depleted for CIP4 (Fig. 2B) failed

B

0

100

200

300

400

La

me

llipodiu

msu

rfa

ce

(µ

m2)

CCL19 ++++ ––––

Untr. Sh-

Contr.

Sh-

CIP4/1

Sh-

CIP4/2

ns

ns

*** ******

***

CIP4

GAPDH

Un

tr

Sh

-Co

ntr

.

Sh

-CIP

4/1

Sh

-CIP

4/2

A

C

α-Tub. α-Tub./DAPIDIC

ICAM-1 + CCL19ICAM-1

Untr.

Sh-

Contr.

Sh-

CIP4/1

Sh-

CIP4/2

α-Tub. α-Tub./DAPIDIC

Figure 2. CIP4 is required forCCL19-driven lamellipodiumassembly in JVM3 cells. A,representative confocal images ofJVM3 cells either untransduced ortransducedwith a control vector or2 CIP4 shRNA-encoding vectorsthat were plated on ICAM-1 in theabsence of presence of 100 ng/mLsoluble CCL19. Cells were stainedwith an antitubulin antibody andDAPI. To visualize the microtubulenetwork and the MTOC, maximumintensity projections of z-stackswere built with the Zen software(Carl Zeiss). B, CIP4 expressionlevel analyzed by Western blottingin untransduced JVM3 cells or inJVM3 cells transduced withlentiviral particles encoding theindicated shRNA sequences(1 control sequence and 2sequences targeting CIP4 mRNA).C, lamellipodium surface of theindicated JVM3 cells measuredfollowing incubation on ICAM-1in the absence of presence ofCCL19. The mean � SEM of thelamellipodium surface withineach group of cells is represented(n ¼ 19–62 cells/group). Anunpaired Student t test wasapplied to compare the groups,��, P � 0.001; ns, nonsignificant.

Malet-Engra et al.





to assemble a polarized lamellipodium upon CCL19 exposureand lost the MTOC axial positioning. These cells formedmultiple filopodia and flat protrusions devoid of microspikes,which extended in different directions (Fig. 2A). The morpho-logic alteration induced by CIP4 depletion was further docu-mented by a loss in lamellipodium surface in CCL19-stimu-lated JVM3 cells (Fig. 2C). The role of CIP4 in controlling JVM3cell lamellipodium assembly was confirmed in live cellsexposed to a CCL19 gradient. Most cells treated with a CIP4siRNA failed to assemble the microspike-containing lamelli-podia predominantly observed in control cells (SupplementaryFig. S2A). This failure was associated with frequent changes indirections of migration and reduced directional persistence(Supplementary Fig. S2B). As observed by time-lapse videorecording, instead of assembling a lamellipodium, CIP4knocked down JVM3 cells emitted transiently and sequentiallymultiple types of protrusions (Supplementary Movie S2). Inagreement with the observation in fixed cells, these cellsemitted flat protrusions and short filopodia, as well as narrowlamellipodia containing 1 to 2 microspikes only, or blebs(Supplementary Fig. S2C). Clearly, these cells failed to orientatealong the chemokine gradient (Supplementary Fig. S2D). Tak-en together, these findings show that CIP4 is crucial for thepolarized assembly of the lamellipodium in response to CCL19in JVM3 cells.

CIP4 dynamically distributes to the lamellipodiumTo further investigate the mechanism by which CIP4 con-

trols lamellipodium dynamics, JVM3 cells expressing a CIP4–GFP fusion protein were observed by time-lapse microscopyduring directional migration toward CCL19. JVM3 cells placedin media without chemokine were either round or displayednarrow cytoplasmic extensions emitting short filopodia. Thesecells displayed a cytoplasmic distribution of CIP4 with localenrichment in discrete patches (Fig. 3A). Exposure to a CCL19gradient led to lamellipodium assembly and to the polarizationof CIP4 toward the lamellipodium rear and along microspikes(Fig. 3B and Supplementary Movie S3). GFP–CIP4 intensityprofile along microspikes revealed that CIP4 was enrichedeither at the rear of microspikes or along these structures(Fig. 3C–E). Parallel analysis of GFP–actin and GFP–CCR7 inmigrating JVM3 cells indicated thatmicrospikes were enrichedin actin and contained CCR7 (Supplementary Movies S4 andS5). Higher magnification examination showed that themicro-spikes extended in length up to 6 mm from the lamella until thetip of the lamellipodium and harbored pendularmovements atan angular speed of 1 to 6.5 rad/s. Notably, CIP4 accumulatedalong microspikes as they converged and fused, an event thatwas associatedwithmicrospike shortening and retraction (Fig.3F and Supplementary Movies S6–S8). The study of CIP4localization during microspike dynamics revealed that coin-cidently with microspike contact (interdistance < 2 mm), CIP4was recruited at the rear and then along the contactingmicrospikes. As the microspikes started to retract (within0–30 seconds following maximal CIP4 accumulation), CIP4intensity decreased in a parallel manner (Fig. 3G–I and Sup-plementary Movies S6–S8). The dynamic patterns of CIP4distribution suggest that this molecule is associated with the

assembly of lamellipodia and with the turnover of radialmicrospikes.

CCL19-driven activation of Cdc42, WASP, and MAPK inCIP4-depleted JVM3 cells

CIP4 was originally described as an effector of Cdc42 (28), aRhoGTPase that plays a key role in polarizedmigration (28, 29).We therefore analyzed whether CIP4 interacted with Cdc42 inJVM3 cells and whether CIP4 depletion might affect Cdc42activation in these cells. To test directly whether active Cdc42and CIP4 interacted in JVM3 cells exposed to CCL19, celllysates were subjected to pull-down experiments withCdc42-coated beads. Active GST-Cdc42 (GTP form) boundCIP4 very efficiently, whereas no binding was detected witheither inactive GST-Cdc42 (GDP form) orGST alone, indicatingthat CIP4 interacted with active Cdc42 in our model (Fig. 4A).The association of CIP4 to Cdc42 in our model raised thepossibility that CIP4 might influence Cdc42 activation orstability. However, CIP4-depleted JVM3 cells displayed bothnormal total Cdc42 levels and normal CCL19-evoked Cdc42activation (Fig. 4B). Next, we testedwhether CIP4, as a bona fideeffector of Cdc42 (28), was required to mediate signalingdownstream of this GTPase. Most notably, biochemical recon-stitution experiments showed that TOCA-1, and presumablyalso CIP4 and FBP17, were critical to promote optimal acti-vation of the otherwise inhibited actin nucleation-promotingfactors WASP and N-WASP (30). Hence, we tested the activa-tion status of WASP, which is specific of the hematopoieticsystem. Loss of CIP4 significantly impairedWASP activation inresponse to CCL19 stimulation as indicated by its tyrosinephosphorylation status (Fig. 4C; ref. 31), suggesting that onemechanism underlying protrusion deficiency consecutive toCIP4 knockdown might be impaired WASP activation andsubsequent WASP-dependent actin polymerization. The mito-gen-activated protein kinase (MAPK) p38 and Erk1/2 areadditional, important regulators of cell migration (32) andhave been shown to act downstream of Cdc42 (33). We firstassessed the involvement of the p38 pathway on chemotaxis ofprimary CLL cells by using the p38 pharmacologic inhibitorSB203580. Leukemic cells isolated from the blood of a patientwith CLL displayed a reduced migration toward CCL19 uponSB203580 treatment (Fig. 4D). We then investigated whetherCIP4 depletion had an impact on p38 and Erk1/2 phosphor-ylation. As expected from the reported constitutive expressionof phosphorylated p38 in CLL (34), nonstimulated controlJMV3 cells displayed a high p38 phosphorylation status. Inter-estingly, in these cells, CCL19 treatment did not modify p38phosphorylation. CIP4-depleted JMV3 cells also displayed ahigh basal phosphorylation of p38, which, however, displayed a2-fold increase upon CCL19 treatment (Fig. 4E and F). p38phosphorylation has been linked to the activation PAK1, aknown effector of Cdc42 (35). Therefore, we assessed PAK1phosphorylation status in control and CIP4-depleted JMV3cells. Group 1 PAKs (PAK1/2/3) were not activated in non-stimulated cells and were activated upon CCL19 treatment(Fig. 4E). As observed for p38, depletion of CIP4 in JMV3 cellsmarkedly increased PAK1 phosphorylation as compared withcontrol cells. To evaluate the role of PAK in p38 activation






following CIP4 depletion, cells were treated with IPA-3, aselective PAK1/2/3 inhibitor that binds covalently to the PAK1regulatory domain and prevents binding to the upstream

activator Cdc42 (26). As expected, IPA-3 treatment beforeCCL19 stimulation inhibited PAK phosphorylation andreduced p38 phosphorylation back to control levels, indicating

B

1.0

2.0

CC

L1

9 g

rad

ien

t

34” 3’15” 3’25”1’32”44”

EDC4’26”

0

200

400

600

800

1,000

1,200

1,400

8072645648403224168000.511.5

22.533.544.55

0

500

1,000

1,500

2,000

2,500

104968880726456484032168000.51

1.5

2

2.5

3

3.5

4

4.5

Dis

tance (

µm

)

0

200

400

600

800

1,000

1,200

1,400

88807264564840322416800

1

2

3

4

5

6

F

IHG

Distance between 2 microspikesMicrospike length CIP4 intensity

Time (s)

1.0

2.0

CC

L1

9 g

rad

ien

t

CIP

4 inte

nsity (

AF

U)

” 48”

0

50

100

150

200

250

300

543210

50

100

150

200

250

300

54321

0

50

100

150

200

250

300

54321

CIP

4 inte

nsity (

AF

U)

Length (µm)

00 0

CIP4 intensity along microspike Mean CIP4 intensity within lamella

A

400” 8” 16” 24” 64”

Figure 3. CIP4 dynamics at the lamellipodium of migrating JVM3 cells. A and B, time-lapse images of JVM3 cells nucleofected with GFP–CIP4 and placed in achemokine-free chamber (A) or in a chamber with a CCL19 chemokine gradient (B). The pseudo-color scale highlights the local intensity of CIP4,from blue (background) to red (maximum intensity). C–E, quantification of CIP4 intensity profile along representative microspikes of a cell migrating in aCCL19 gradient. F, time-lapse images of microspike dynamics within the lamellipodium of a cell migrating in a CCL19 gradient. G–I, quantification of 3representative events of microspike turnover. Measurements in time of CIP4 intensity at microspikes, distance between microspikes, and length ofmicrospikes are shown in parallel.

Malet-Engra et al.





that PAK is responsible for p38 phosphorylation in CLL19-stimulated JVM3 cells. IPA-3 treatment also inhibited Erkactivation in control and CIP4-depleted JMV3 cells suggestingthat PAK is also regulating Erk activation in this model.Altogether, these data suggest that in our CLL model, CIP4regulates CCL19-directed cell migration by modulating theactivation of WASP and PAK/p38 as central motility-relatedpathways.

p38 and CIP4 are present within lamellipodia andcontrol their formationGiven the recent identification of p38 as a major driver of

chemotaxis in neutrophils (36), we addressed whether thep38 pathway might cooperate with CIP4 in the control oflamellipodium assembly in CLL cells. We first studied thelocalization of p38 and CIP4 in cells emitting CCL19-evokedlamellipodia. As shown in previous figures, JVM3 cells platedon ICAM-1 and stimulated with CCL19 formed lamellipodia(Fig. 5A). CIP4 displayed a punctuate distribution, reminis-cent of intracellular vesicles. CIP4 was particularly enrichedin large patches at the rear of the lamellipodium. CIP4 wasalso detected within the lamellipodium itself, albeit at lower

levels. The distribution of phospho-p38 within the cell bodydiffered from that of CIP4 and was mainly nuclear, asrevealed by the DAPI staining. Differently from CIP4, phos-pho-p38 could not be detected at the rear of the lamellipo-dium. However, phospho-p38 was detected within the lamel-lipodium extension. Along a cross-section of the lamellipo-dium, CIP4 and phospho-p38 displayed parallel intensityprofiles, suggesting that this structure might be a sitefavoring interaction between these 2 molecules (Fig. 5B).The parallel distribution of CIP4 and p38 over the breath ofthe lamellipodium extension was further strengthened by arelatively high Pearson's coefficient, which applied both forphospho-p38 and total p38 stainings, when calculated for thewhole lamellipodium area (Fig. 5C). We then tested whetherp38 might control lamellipodium formation in JVM3 cells.Blockade of the p38 pathway with the specific p38 inhibitorSB203580 inhibited lamellipodia extension in control, butnot in JVM3 CIP4-depleted JVM3 cells (Fig. 5D). Takentogether, our data establish a link between CIP4 and p38as both molecules are colocalized to the lamellipodium andas they are both necessary to the assembly of this structurepromoting directional migration.

Figure 4. CIP4 depletion results inincreased p38 activation. A, bindingof GST, GST-Cdc42GDP, and GST-Cdc42GTP immobilized onglutathione-sepharose beads toendogenous CIP4 from JVM3 cells.B, Cdc42 activation analyzed bypull-down using GST-PBD fromPAK1 and Western blotting in CIP4-expressing and CIP4-depletedJVM3 cells stimulated or not with100 ng/mL CCL19. C, CIP4-expressing and -depleted JVM3cells were stimulated for 5 or 20minutes with 100 ng/mL CCL19.Stimulation was stopped by additionof ice-cold medium and cell lysateswere immunoblotted with CIP4,WASP, P-WASP (phosphoY-290),and vinculin antibodies. D, primaryleukemic cells were pretreated for 1hour with 2.5 mmol/L SB203580 orDMSO and assessed for CCL19-driven chemotaxis in a Transwellassay with the indicated CCL19concentrations. The mean � SD oftriplicate values is represented. E,analysis from 5 independentexperiments of the fold induction ofp38 MAPK phosphorylationfollowing CCL19 stimulation of theindicated cell lines. F,phosphorylation of p38, Erk1/2, andPAK1 analyzed by Western blottingfollowing CCL19 stimulation ofCIP4-expressing and -depletedJVM3 cells. Where indicated, thePAK inhibitor IPA-3 was added.Expression of CIP4, total PAK1/2/3,and actin are shown as controls.

A

E

GS

T

GS

T-C

dc4

2G

DP

GS

T-C

dc4

2G

TP

Total CIP4

PD CIP4

CIP4

Total Cdc42

PD Cdc42

CCL19 : – + – +

Sh-

Contr.

Sh-

CIP4/1

B

D

F

C

% M

igra

tin

g c

ells

SB203580

0

20

40

60

80

1,0005001000

DMSO

CCL19 (ng/mL)

0 5 20 0 5 20

CIP4

Sh-

Contr.

Sh-

CIP4

Vinculin

CCL19(min)

P-WASP

WASP

Re

lative

p3

8

ph

osp

hory

latio

n

Untr. Sh-

Contr.

Sh-

CIP4/1

Sh-

CIP4/2

1.0

1.5

2.0

2.5

0.0

IPA-3

P-PAK1/PAK2

+––

CCL19

CIP4

P-p38

Actin

P-Erk1/2

PAK1/2/3

+––+––+––

++–++–++–++–

PAK1/3PAK2

PAK1/3PAK2

Untr.Sh-

Contr.

Sh-

CIP4/1

Sh-

CIP4/2






CIP4 is highly expressed in primary CLL leukemic cellsand regulates CCL19-driven chemotaxis

To investigate the physiopathologic relevance of our find-ings, the expression of CIP4 and its contribution to CCL19-driven migration were investigated in primary leukemic cellsisolated from the peripheral blood of patients with CLL. Asshown by the Western blot analysis conducted on a represen-tative patient series, CIP4 was expressed in a majority ofsamples at higher levels than in control B cells (Fig. 6A).Further analysis of samples from a total of 43 patients withCLL and densitometric quantification indicated that CIP4 wasexpressed in primary CCL cells in an average of 2.5-fold overthe levels of control B cells (range, 0.85–5.45). This analysisreinforced the notion that CLL is associated to high CIP4expression. The analysis of CIP4 localization in primary CLLcells stimulated by ICAM-1/CCL19 showed that CIP4 wasassociated to actin-rich protrusions, including lamellipodia(Fig. 6B), thereby confirming our findings from JVM3 cells.To further explore the role of CIP4 in primary CLL cells, itsexpression was knocked down by siRNA in samples collectedfrom 3 patients (Fig. 6C). CIP4 siRNA-treated leukemic cellsfrom patients 10 and 15 displayed a reduced migration whencompared with the corresponding control cells treatedwith scrambled siRNA. In contrast, the leukemic cells from

patient 22, in which levels of CIP4 remained high upon CIP4siRNA treatment, retained their ability to migrate efficientlytoward the CCL19 gradient, reinforcing the idea that CLLcell chemotaxis toward CCL19 is dependent on the presenceof a high CIP4 level (Fig. 6C). Given the relative variability ofCIP4 expression in primary CLL cells, we investigatedwhether CIP4 expression levels may correlate with knowndisease markers by subdividing the 43 patients analyzed intoCIP4-high (>2.5-fold increase over control B cells) and CIP4-low (<2.5-fold increase over control B cells) groups ofcomparable size. As reported in Table 1, CIP4-high patientsseemed to be biased toward the IgVH-unmutated status,which is associated to an aggressive form of the disease andto a shortened survival (37, 38). Further studies will benecessary to consolidate this association and to investigatehow the molecule CIP4 identified here as a key regulator ofCLL cell directed migration might be involved in the path-ogenesis of the disease.

DiscussionThis study provides novel molecular insights into the reg-

ulation of the high migratory potency of CLL leukemic cells.Real-time imaging of CLL cell directional migration in a CCL19

A

D

Sh-CIP4/1Sh-Contr.

SB203580 ++ ––

0

100

200

300

400

La

me

llipodiu

m

su

rfa

ce

(µ

m2)

***

ns

P-p38

CIP4

P-p38

DAPI

CIP4

DIC

30

B

0

200

400

600

800

20100

Sta

inin

g in

ten

sity (

AU

)

P-p38

CIP4

Position along lamellipodium (µm)

C

Pe

ars

on

’s c

oe

ffic

ient

0.2

0.4

0.6

0.8

0

1.0

P-p38/CIP4p38/CIP4

Colocalization in lamellipodium

Figure 5. p38 localizes with CIP4 tothe lamellipodium and is requiredfor its assembly. A, representativeconfocal image of a JVM3 celladhering to ICAM-1 in thepresence of 100 ng/mL CCL19stained with anti-phospho-p38antibodies, anti-CIP4 antibodies,and DAPI. B, representative linescan analysis of phospho-p38 andCIP4 localization in the JVM3 cellsrepresented in A, along thelamellipodium section depicted bythe white line in A. C, meanPearson's coefficient used toevaluate the total p38 and CIP4colocalization or the phospho-p38and CIP4 colocalization, from thelamellipodium area of 6representative polarized cells.D, lamellipodium surface of theindicated JVM3 cells pretreated for1 hourwith 10mmol/LSB203580 orDMSOand incubated for 5minuteson ICAM-1 in the presence of100 ng/mL soluble CCL19. Themean � SEM of the lamellipodiumsurface within each group of cellsis represented. An unpairedStudent t test was applied tocompare the groups, ��, P �0.001; ns, nonsignificant.

Malet-Engra et al.





gradient revealed that these cells assemble a polarized, par-ticularly large anddynamic actinmicrospike-containing lamel-lipodium. More importantly, the assembly and turnover of thisprotrusive structure that drives the cells toward the source ofCCL19 is dependent on CIP4, a F-BAR domain protein knownto regulate endocytosis by sensing and modifying membranecurvature (25, 39, 40).The lamellipodium of CLL cells seems as a broad veil

bearing radially arranged microspikes that contain actinbundles. To our knowledge, these structures have neverbeen reported in leukocytes. In addition, very little isknown on the dynamics of lamellipodium-associated micro-spikes as their early description in fibroblasts (41). In CLLleukemic cells, microspike formation and dynamics aredependent on chemokine exposure, suggesting that thesestructures are involved in chemokine gradient sensing.This possibility is further reinforced by the observationthat CCR7 is present along these structures. CIP4-depletedcells display multiple, short-lived, and variably oriented

protrusions largely devoid of microspikes. As a consequenceof these alterations, cells lose polarity and chemotacticability. Notably, CIP4 is dynamically and transientlyenriched at the rear of microspikes just before these struc-tures fuse and disassemble, highlighting a novel role forCIP4 in regulating microspike turnover in highly migratingtumoral cells.

CIP4 has been shown to control endocytosis of numerousreceptors, including receptors triggering cell migration (42).However, in our model CCR7 internalization and cell surfaceexpression are not affected by removal of CIP4, suggestingthat its endocytic function might not account for its controlover polarized lamellipodia extension, microspike dynamics,and chemotactic cell migration. In the context of prenatalneuronal development, CIP4 has recently been discovered tonegatively regulate neurite outgrowth by inducing lamelli-podia and veil-like protrusions rich in actin bundles (43).Notably, CIP4 activity in developing neurons was also inde-pendent from its role in endocytosis. CIP4 function, instead,

Figure 6. CIP4 is expressed at highlevels in primary CLL cells and isrequired for CCL19-drivenmigration. A, Western blot analysisof CIP4 in normal peripheral blood Bcells and in leukemic cells from 12representative patients with CLL. B,representative confocal image ofCLL primary cells plated on ICAM-1in the presence of 100 ng/mL solubleCCL19. Cells were stained with ananti-CIP4 antibody, fluorescentphalloidin, and DAPI. The 3-dimensional (3D) view was obtainedfrom a z-stack. C, CIP4 expressionlevel analyzed byWestern blotting inprimary leukemic cells from 3patients with CLL 72 hours followingnucleofection with the indicatedsiRNA (control scrambled sequenceor sequence targeting CIP4).CCL19-driven chemotaxis of theobtained CIP4 knocked down andcontrol primary leukemic cells wasassessed in a Transwell assaycontaining the indicatedconcentrations of CCL19.

AP27 P28 P29 P30 P31 P32 P33 P34 P35 P36 P37 P38B

ce

lls CLL patients

CIP4

GAPDH

B

C

CCL19 (ng/mL)

% M

igra

tin

g c

ells

0

5

10

15

20

25

30

1,0005001000

0

5

10

15

20

25

30

1,000500100 1,0000 5001000

0

10

20

30

40

50

Scr

CIP4

siRNA

P15

CIP4Scr

P10

CIP4Scr

P22

CIP4ScrSiRNA :

CIP4

GAPDH

F-actinCIP4 Overlay (3D)DIC

P32






was restricted to the migratory and developmental phase ofneuronal cells, and controlled the emission of neuronalprotrusion along neurites and axons. Remarkably, this func-tional role was attributed to the ability of CIP4 to mediatelarge and extended lamellipodia, a finding similar to what wereport in this study in a dramatically different cellularcontext. Collectively, these results support the notion thatone major role of CIP4 is to regulate the dynamics oflamellipodia protrusions, further impacting on cell motility,a central function that might have been highjacked bytumoral cells to promote spreading.

A still unanswered question remains: how does CIP4 pro-mote themembrane and actin dynamics underlying the assem-bly of the microspike-bearing lamellipodium? Because of theirconcave membrane-bending activity, F-BAR proteins are usu-ally associated to the regulation of endocytosis. However, someF-BAR proteins includingCIP4 have also been associated to thepromotion of membrane protrusions such as filopodia andinvadopodia (9, 28, 44, 45). Given thatmicrospikes ofmigratingCLL cells contain actin bundles, it is very likely that additionalmolecular partners involved in actin polymerization are work-ing together with CIP4. In keeping with this notion, we showthat CCL19 stimulates Cdc42 activation and relocalization tothe rear of the lamellipodium together with CIP4. CIP4 andCdc42, in turn, have been shown to interact withWASP/WAVEproteins thereby promoting actin nucleation and filamentelongation (7, 9, 46–48). In this context, it is worth noting thatCdc42 is primarily implicated in microspike/filopodia forma-tion, rather than directly in the extension of lamellipodia.However, in several cell types, filopodia and sheet-like lamelli-podia rapidly interchange during protrusive activity. On onehand, filopodia are embedded into or arise from preexistinglamellipodia (49). On the other hand, filopodia may becomeanchoring sites from where lamellipodial veils first form. This

latter notion is consistent with our observations and recentdata in neurons (43), showing that CIP4 controls microspikedynamics, favoring the assembly of veil-like structures.

Our study highlights a further level of molecular control oflamellipodium assembly in CLL cells by linking for the firsttime CIP4 to PAK1 and p38MAPK activation. Our observationsstrengthen the notion that p38, a known regulator of stressresponses, also regulates directedmigration (32). Interestingly,p38 MAPK has been shown to be constitutively active in CLL,where it regulates MMP-9 production, suggesting a role ininvasion (34). We report here that p38 activation is furtherincreased upon CLL19 exposure in CIP4-depleted CLL cells,suggesting that control of its activation is important in regu-lating migration. In agreement with our data, it has beenreported that over-activation of PAK1 compromised the main-tenance of cell polarity in the migrating mesendoderm duringXenopus gastrulation (50). Similarly to CIP4 knockdown, p38inhibition prevents lamellipodium assembly, an effect that isnot additive with CIP4 depletion. The fact that both p38 hyper-activation or inhibition are detrimental to lamellipodiumassembly suggests that a balanced p38 activity is required forthis process. It therefore indicates that themodulation by CIP4levels and subsequent PAK activation downstream of Cdc42can be key regulators of directed migration. Future work willelucidate themechanism of CIP4 and p38MAPK collaborationin CLL cells.

Our data also suggest that high CIP4 expression is ahallmark of tumoral CLL cells and is associated to the highchemotactic activity of these cells. Breast cancer cell linesalso express high levels of CIP4, which regulates invasiveness(9, 10). Very recently, CIP4 expression has also been asso-ciated to the invasiveness of osteosarcoma (51). Together,these findings suggest that the tight regulation of CIP4expression observed during neuronal development mightbe lost in some tumors, thereby participating to the trans-formation process. Consistent with this notion, we show thatprimary leukemic cells from patients with CLL display highexpression of CIP4, which is required for CCL19-drivenchemotaxis. Analysis of CIP4 protein expression in restingand activated normal B cells and various subtypes of B-cellmalignancies suggests that high CIP4 expression may bespecific for CLL leukemic cells within the B-cell lineage. Inthe cohort of 43 patients with CLL analyzed in this study, wenoted a bias toward an IgVH-unmutated status in thepatient group with the highest CIP4 expression. Furtherstudies on larger cohorts will be necessary to confirm thatCIP4-high expression is linked to disease severity. It remainsalso to be clarified how CIP4 expression and its control overdirected cell migration is participating to CLL pathogenesis.Importantly, our data indicate that lowering CIP4 expressionto levels comparable with those of normal B cells can impairthe chemotactic ability of leukemic cells from patients withCLL. This observation raises the possibility that CLL cellsbecome addicted to CIP4 to migrate toward CCL19. Alongthat view, CIP4 may have an important therapeutic valueand could be considered a molecular target of pharmaco-logic inactivation to block CLL cells dissemination to lymphnodes, a strategy already pursued with the use of anti-CCR7

Table 1. CLL patients

CLL patientsubgroupsa

CIP4-low(n ¼ 21)

CIP4-high(n ¼ 22) P

First line/relapse 95/5% 82/18% nsGender M/F 62/38% 77/23% nsAge (median) 66y 66y nsLymphocytosis (median) 86,400 84,000 nsBinet stage A/B/C 52/29/19% 64/18/18% nsIgVH unmutated 38% 63.6% 0.09Deletion 11q 9.5% 18.2% nsDeletion 17p 9.5% 4.5% nsBulky disease (>5 cm) 23.8% 13.6% nsCD38 > 20% 62.3% 60% nsIndication for therapy 47.6% 59% ns

Abbreviation: ns, nonsignificant.aDisease parameters of 43 patients with CLL divided asCIP4-low andCIP4-high (lower/higher than 2.5-fold increaseof CIP4 expression in leukemic cells as compared withnormal B cells).

Malet-Engra et al.





antibodies (52). CLL clinical spectrum is wide, ranging frommild lymphocytosis to lymph node enlargement and multi-organ failure. Disease progression may be related to thedevelopment of cells acquiring an exacerbated potential toenter and invade lymph nodes via the CCL19/21-CCR7 axis.This is supported by the increased CCR7 expression andCCR7-dependent migration reported in ZAP-70þ/CD38þ

CLL cells, whose presence is related to a bad prognosis(24, 53). In that context, further studies will be required toelucidate how disease progression may be related to theCIP4-driven chemotactic activity characterized here.

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

Authors' ContributionsConception and design:G. Malet-Engra, J. Viaud, G. Laurent, G. Scita, L. Dupr�eDevelopment of methodology: G. Malet-Engra, J. Viaud, M. Farc�e,F. LafouresseAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): G. Malet-Engra, J. Viaud, L. Ysebaert, M. Farc�e,F. Lafouresse, L. Dupr�eAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): G. Malet-Engra, J. Viaud, F. Lafouresse, F. Gaits-Iacovoni, L. Dupr�eWriting, review, and/or revision of the manuscript: G. Malet-Engra,L. Ysebaert, F. Lafouresse, F. Gaits-Iacovoni, G. Scita, L. Dupr�e

Administrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): G. Malet-Engra, L. Ysebaert, M. Farc�e,L. Dupr�eStudy supervision: L. Dupr�e

AcknowledgmentsThe authors thank V�eronique De Mas and C�ecile Demur, in charge of the

HIMIP cell bank, for preparing and providing primary cells from patients withCLL; Daniel Sapede and Sophie Allart from the Institut National de la Sant�e et dela Recherche Medicale (INSERM) U1043 imaging facility for assistance in imageprocessing; Daniel Legler for providing the GFP-tagged CCR7 construct; AnneQuillet-Mary, Emilie Gross, and Salvatore Valitutti for helpful discussions; andJosipa Spoljaric for article editing.

Grant SupportThis work was supported by the European Community (Marie Curie

Excellence grant, contract MEXT-CT-2005-025032 to L. Dupr�e), by the LigueContre le Cancer (grant from the Comit�e R�egional de Haute-Garonne to L.Dupr�e) and by grants from the Associazione Italiana per la Ricerca sul Cancro(AIRC), by a grant from the Fondation ARC pour la Recherche sur le Cancer(SL220100601347), and the Italian Ministries of Education-University-Research (MIUR-PRIN) grant to G. Scita. J. Viaud is a fellow of "FondationARC pour la Recherche sur le Cancer."

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 indicate thisfact.

Received September 13, 2012; revised March 7, 2013; accepted March 14, 2013;published OnlineFirst May 3, 2013.

References1. RoussosET,Condeelis JS,PatsialouA.Chemotaxis in cancer.NatRev

Cancer 2011;11:573–87.2. Yamaguchi H, Wyckoff J, Condeelis J. Cell migration in tumors. Curr

Opin Cell Biol 2005;17:559–64.3. Yamazaki D, Kurisu S, Takenawa T. Regulation of cancer cell motility

through actin reorganization. Cancer Sci 2005;96:379–86.4. Gimona M. The microfilament system in the formation of invasive

adhesions. Semin Cancer Biol 2008;18:23–34.5. Sarmiento C, Wang W, Dovas A, Yamaguchi H, Sidani M, El-Sibai

M, et al. WASP family members and formin proteins coordinateregulation of cell protrusions in carcinoma cells. J Cell Biol 2008;180:1245–60.

6. KurisuS, TakenawaT.WASPandWAVE family proteins: friends or foesin cancer invasion? Cancer Sci 2010;101:2093–104.

7. Takano K, Toyooka K, Suetsugu S. EFC/F-BAR proteins and the N-WASP-WIP complex induce membrane curvature-dependent actinpolymerization. Embo J 2008;27:2817–28.

8. Suetsugu S, Toyooka K, Senju Y. Subcellular membrane curvaturemediatedby theBARdomain superfamily proteins. SeminCell DevBiol2010;21:340–9.

9. Pichot CS, Arvanitis C, Hartig SM, Jensen SA, Bechill J, Marzouk S,et al. Cdc42-interacting protein 4 promotes breast cancer cell invasionand formation of invadopodia through activation of N-WASp. CancerRes 2010;70:8347–56.

10. Hu J, Mukhopadhyay A, Truesdell P, Chander H, Mukhopadhyay UK,Mak AS, et al. Cdc42-interacting protein 4 is a Src substrate thatregulates invadopodia and invasiveness of breast tumors by promot-ing MT1-MMP endocytosis. J Cell Sci 2011;124:1739–51.

11. Pals ST, de Gorter DJ, Spaargaren M. Lymphoma dissemination: theother face of lymphocyte homing. Blood 2007;110:3102–11.

12. Caligaris-Cappio F, Bergui L, Tesio L, Corbascio G, Tousco F, March-isio PC. Cytoskeleton organization is aberrantly rearranged in the cellsof B chronic lymphocytic leukemia and hairy cell leukemia. Blood1986;67:233–9.

13. Zenz T, Mertens D, Kuppers R, Dohner H, Stilgenbauer S. Frompathogenesis to treatment of chronic lymphocytic leukaemia. Nat RevCancer 2010;10:37–50.

14. Mohle R, Failenschmid C, Bautz F, Kanz L. Overexpression of thechemokine receptor CXCR4 in B cell chronic lymphocytic leukemia isassociated with increased functional response to stromal cell-derivedfactor-1 (SDF-1). Leukemia 1999;13:1954–9.

15. L�opez-Giral S, Quintana NE, Cabrerizo M, Alfonso-P�erez M, Sala-Vald�es M, De Soria VG, et al. Chemokine receptors that mediate B cellhoming to secondary lymphoid tissues are highly expressed in B cellchronic lymphocytic leukemia and non-Hodgkin lymphomas withwidespread nodular dissemination. J Leukoc Biol 2004;76:462–71.

16. Redondo-Munoz J, Escobar-Diaz E, Samaniego R, Terol MJ, Garcia-Marco JA, Garcia-Pardo A. MMP-9 in B-cell chronic lymphocyticleukemia is up-regulated by alpha4beta1 integrin or CXCR4 engage-ment via distinct signaling pathways, localizes to podosomes, and isinvolved in cell invasion and migration. Blood 2006;108:3143–51.

17. Redondo-Munoz J, Jose Terol M, Garcia-Marco JA, Garcia-Pardo A.Matrix metalloproteinase-9 is up-regulated by CCL21/CCR7 interac-tion via extracellular signal-regulated kinase-1/2 signaling and isinvolved in CCL21-driven B-cell chronic lymphocytic leukemia cellinvasion and migration. Blood 2008;111:383–6.

18. Messmer BT, Messmer D, Allen SL, Kolitz JE, Kudalkar P, Cesar D,et al. In vivomeasurements document the dynamic cellular kinetics ofchronic lymphocytic leukemia B cells. J Clin Invest 2005;115:755–64.

19. 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.

20. Scielzo C, Ghia P, Conti A, Bachi A, Guida G, Geuna M, et al. HS1protein is differentially expressed in chronic lymphocytic leukemiapatient subsets with good or poor prognoses. J Clin Invest 2005;115:1644–50.

21. ScielzoC,BertilaccioMT, Simonetti G, Dagklis A, tenHackenE, Fazi C,et al. HS1 has a central role in the trafficking and homing of leukemic Bcells. Blood 2010;116:3537–46.

22. Forster R, Davalos-Misslitz AC, Rot A. CCR7 and its ligands: balancingimmunity and tolerance. Nat Rev Immunol 2008;8:362–71.

23. 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.






24. Richardson SJ, Matthews C, Catherwood MA, Alexander HD, CareyBS, Farrugia J, et al. ZAP-70 expression is associated with enhancedability to respond to migratory and survival signals in B-cell chroniclymphocytic leukemia (B-CLL). Blood 2006;107:3584–92.

25. Itoh T, Erdmann KS, Roux A, Habermann B, Werner H, De Camilli P.Dynamin and the actin cytoskeleton cooperatively regulate plasmamembrane invagination by BAR and F-BAR proteins. Dev Cell 2005;9:791–804.

26. Viaud J, Peterson JR. An allosteric kinase inhibitor binds the p21-activated kinase autoregulatory domain covalently. Mol Cancer Ther2009;8:2559–65.

27. Colomba A, Courilleau D, Ramel D, Billadeau DD, Espinos E, Delsol G,et al. Activation of Rac1 and the exchange factor Vav3 are involved inNPM-ALK signaling in anaplastic large cell lymphomas. Oncogene2008;27:2728–36.

28. Aspenstrom P. A Cdc42 target protein with homology to the non-kinase domain of FER has a potential role in regulating the actincytoskeleton. Curr Biol 1997;7:479–87.

29. Etienne-Manneville S. Polarity proteins in migration and invasion.Oncogene 2008;27:6970–80.

30. Ho HY, Rohatgi R, Lebensohn AM, Le Ma, Li J, Gygi SP, et al. Toca-1mediates Cdc42-dependent actin nucleation by activating the N-WASP-WIP complex. Cell 2004;118:203–16.

31. Torres E, Rosen MK. Contingent phosphorylation/dephosphorylationprovides a mechanism of molecular memory in WASP. Mol Cell 2003;11:1215–27.

32. Huang C, Jacobson K, Schaller MD. MAP kinases and cell migration.J Cell Sci 2004;117:4619–28.

33. Szczur K, Xu H, Atkinson S, Zheng Y, Filippi MD. Rho GTPase CDC42regulates directionality and random movement via distinct MAPKpathways in neutrophils. Blood 2006;108:4205–13.

34. Ringshausen I, DechowT, Schneller F,Weick K, OelsnerM, Peschel C,et al. Constitutive activation of theMAPkinase p38 is critical forMMP-9production and survival of B-CLL cells on bone marrow stromal cells.Leukemia 2004;18:1964–70.

35. Rousseau S, Dolado I, Beardmore V, Shpiro N, Marquez R, NebredaAR, et al. CXCL12 and C5a trigger cell migration via a PAK1/2-p38alpha MAPK-MAPKAP-K2-HSP27 pathway. Cell Signal 2006;18:1897–905.

36. Liu X, Ma B, Malik AB, Tang H, Yang T, Sun B, et al. Bidirectionalregulation of neutrophil migration by mitogen-activated proteinkinases. Nat Immunol 2012;13:457–64.

37. Hamblin TJ,Davis Z,GardinerA,Oscier DG,StevensonFK.UnmutatedIg V(H) genes are associated with a more aggressive form of chroniclymphocytic leukemia. Blood 1999;94:1848–54.

38. 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.

39. Feng Y, Hartig SM, Bechill JE, Blanchard EG, Caudell E, Corey SJ. TheCdc42-interacting protein-4 (CIP4) gene knock-out mouse revealsdelayed and decreased endocytosis. J Biol Chem 2010;285:4348–54.

40. HuJ, Troglio F,MukhopadhyayA, EveringhamS,KwokE,ScitaG, et al.F-BAR-containing adaptor CIP4 localizes to early endosomes andregulates epidermal growth factor receptor trafficking and downregu-lation. Cell Signal 2009;21:1686–97.

41. Wang YL. Exchange of actin subunits at the leading edge of livingfibroblasts: possible role of treadmilling. JCell Biol 1985;101:597–602.

42. Toguchi M, Richnau N, Ruusala A, Aspenstrom P. Members of theCIP4 family of proteins participate in the regulation of platelet-derivedgrowth factor receptor-beta-dependent actin reorganization andmigration. Biol Cell 2010;102:215–30.

43. Saengsawang W, Mitok K, Viesselmann C, Pietila L, Lumbard DC,Corey SJ, et al. The F-BAR protein CIP4 inhibits neurite formation byproducing lamellipodial protrusions. Curr Biol 2012;22:1–8.

44. Lee K, Gallop JL, Rambani K, Kirschner MW. Self-assembly of filopo-dia-like structures on supported lipid bilayers. Science 2010;329:1341–5.

45. Shimada A, Takano K, Shirouzu M, Hanawa-Suetsugu K, Terada T,Toyooka K, et al. Mapping of the basic amino-acid residues respon-sible for tubulation andcellular protrusionby theEFC/F-BARdomainofpacsin2/Syndapin II. FEBS Lett 2010;584:1111–8.

46. Linder S, Hufner K, Wintergerst U, Aepfelbacher M. Microtubule-dependent formation of podosomal adhesion structures in primaryhuman macrophages. J Cell Sci 2000;113:4165–76.

47. Tian L, Nelson DL, Stewart DM. Cdc42-interacting protein 4 mediatesbindingof theWiskott-Aldrich syndromeprotein tomicrotubules. JBiolChem 2000;275:7854–61.

48. Fricke R, Gohl C, Dharmalingam E, Grevelh€orster A, Zahedi B, HardenN, et al. Drosophila Cip4/Toca-1 integrates membrane trafficking andactin dynamics through WASP and SCAR/WAVE. Curr Biol 2009;19:1429–37.

49. Small JV, Kaverina I, KrylyshkinaO, Rottner K. Cytoskeleton cross-talkduring cell motility. FEBS Lett 1999;452:96–9.

50. NagelM, LuuO,BissonN,MacanovicB,Moss T,WinklbauerR. Role ofp21-activated kinase in cell polarity and directional mesendodermmigration in the Xenopus gastrula. Dev Dyn 2009;238:1709–26.

51. Koshkina NV, Yang G, Kleinerman ES. Inhibition of Cdc42-interactingprotein 4 (CIP4) impairs osteosarcoma tumor progression.CurrCancerDrug Targets 2013;13:48–56.

52. Alfonso-Perez M, Lopez-Giral S, Quintana NE, Loscertales J, Martin-JimenezP,MunozC.Anti-CCR7monoclonal antibodies as anovel toolfor the treatment of chronic lymphocyte leukemia. J Leukoc Biol2006;79:1157–65.

53. Deaglio S, Vaisitti T, Aydin S, Bergui L, D'Arena G, Bonello L, et al.CD38 and ZAP-70 are functionally linked and mark CLL cells with highmigratory potential. Blood 2007;110:4012–21.

Malet-Engra et al.





2013;73:3412-3424. Published OnlineFirst May 3, 2013.Cancer Res Gema Malet-Engra, Julien Viaud, Loïc Ysebaert, et al. Chronic Lymphocytic LeukemiaCIP4 Controls CCL19-Driven Cell Steering and Chemotaxis in

Updated version

10.1158/0008-5472.CAN-12-3564doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2013/05/06/0008-5472.CAN-12-3564.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/73/11/3412.full#ref-list-1

This article cites 53 articles, 21 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/73/11/3412.full#related-urls

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

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://cancerres.aacrjournals.org/content/73/11/3412To 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-12-3564

http://cancerres.aacrjournals.org/content/suppl/2013/05/06/0008-5472.CAN-12-3564.DC1

http://cancerres.aacrjournals.org/content/73/11/3412.full#ref-list-1

http://cancerres.aacrjournals.org/content/73/11/3412.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/73/11/3412


CIP4ControlsCCL19-DrivenCellSteeringandChemotaxisin ... · Corresponding Author: Loïc Dupre,...

Documents

Transcript of CIP4ControlsCCL19-DrivenCellSteeringandChemotaxisin ... · Corresponding Author: Loïc Dupre,...