A sustained activation of PI3K/NF-jB pathway is critical for the survival of chroniclymphocytic leukemia B cells

S Cunı1, P Perez-Aciego1, G Perez-Chacon1, JA Vargas2, A Sanchez3, FM Martın-Saavedra4, S Ballester4, J Garcıa-Marco5, J Jorda6

and A Durantez1,2

1Fundacion LAIR, Madrid, Spain; 2Servicio de Medicina Interna I, Hospital Universitario Puerta de Hierro, Madrid, Spain;3Servicio de Neuro-inmunologıa, Hospital Universitario Puerta de Hierro, Madrid, Spain; 4Unidad de Regulacion Genica,Instituto de Salud Carlos III, Madrid, Spain; 5Servicio de Hematologıa, Hospital Universitario Puerta de Hierro, Madrid, Spain; and6Servicio de Hematologıa, Hospital Clınico Universitario San Carlos, Madrid, Spain

The progressive rise of mature CD5þ B lymphocytes, despitethe low proportion of proliferating cells, has led to the notionthat B cell chronic lymphocytic leukemia (B-CLL) is primarilyrelated to defective apoptosis. The microenvironment likelyplays a prominent role because the malignant cells progres-sively accumulate in vivo, whereas they rapidly undergospontaneous apoptosis when cultured in vitro. To assessmicroenvironment-mediated survival signals, B-CLL cells werecultured with a murine fibroblast cell line, Ltk�, with andwithout an agonistic antibody to CD40. Spontaneous apoptosiswas associated with the loss of Akt and NF-jB activities.Interactions with fibroblasts sustained a basal level of Akt andNF-jB activities, which was dependent on phosphatidylinositol-3 kinase (PI3K). Constitutive activity of the PI3K pathway inB-CLL cells when cultured with fibroblasts prevented the down-regulation of the prosurvival Bcl-2 family protein Bcl-xL and thecaspase inhibitor proteins FLIPL and XIAP, and consequentlycaspase-3 activation and apoptosis. CD40 crosslinking in B-CLLcells did not further prevent murine fibroblasts-mediated apop-tosis but induced cell proliferation, which was associated with anincrease of Akt and NF-jB activation compared with cellscultured with fibroblasts alone. The PI3K pathway seems to playa pivotal role in B-CLL cell survival and growth.Leukemia (2004) 18, 1391–1400. doi:10.1038/sj.leu.2403398Published online 3 June 2004Keywords: B-cell chronic lymphocytic leukemia; cell signaling; PI3Kpathway; NF-kB activity; Akt; Bcl-xL expression

Introduction

B-chronic lymphocytic leukemia (B-CLL) is characterized by theaccumulation of long-lived CD5þ B cells.1,2 However, whencultured in vitro, B-CLL cells die quickly by apoptosis.3,4 Severallines of evidence support that nonmalignant cells and other hostelements extrinsic to B-CLL cells provide survival signals tomalignant clones in vivo.5–8 For instance, CD40 stimulation withinlymphoid tissues appears to rescue B-CLL cells from apoptosis.9–13

On the other hand, it has been described that mouse fibroblasts(Ltk� cell line) with addition of IL-4 (the CD40 system) are capableto support the survival and expansion of CLL-B cells in vitro.14,15

However, signaling pathways responsible for the malignant cellsurvival are poorly understood. Exploiting these pathways may beof value in the understanding of the disease pathogenesis anddevelopment of new therapeutic strategies for B-CLL.

The phosphatidylinositol-3 kinase (PI3K)/Akt cascade hasbeen recently identified as a critical component of survival

signaling.16 PI3K-generated phospholipids bind to the serine/threonine kinase Akt/protein kinase B (PKB), leading to itstranslocation from the cytoplasm to the inner surface of theplasma membrane.17 Relocalization of Akt to the plasmamembrane brings Akt in proximity to regulatory kinases(phosphoinositide 3-dependent kinases) that activate Akt byphosphorylation at Thr308 and Ser473.18,19 Once activated, Aktinhibits cell death pathways by directly phosphorylating andinactivating proteins involved in apoptosis, including Bad,procaspase 9, and members of the Forkhead transcriptionfactors family.20–22 In addition, Akt has been implicated as asignaling intermediate upstream of survival gene expression thatis dependent on the transcription factor NF-kB.23–25

NF-kB is a collective name for inducible dimeric transcriptionfactors composed of members of the Rel family of DNA-bindingproteins that recognize a common sequence motif. Mammalsexpress five NF-kB/Rel proteins: p50, p52, p65 (RelA), c-Rel,and RelB.26 NF-kB homo/heterodimers containing differentcombination of these subunits are held in the cytoplasm throughinteraction with specific inhibitors, the IkBs (IkBa, IkBb, andIkBg). The IkBs undergo rapid ubiquitin-dependent degradationafter exposure to a variety of agonists, which activate the IkBkinase (IKK) complex.27 Activation of IKK has been shown to bemediated by upstream acting kinases including NIK, MEKK1, aswell as Akt.28 Once in the nucleus, NF-kB dimers are subjectedto further regulation mainly through phosphorylation of the Relproteins. Several signaling pathways, which induce the phos-phatidylinositol-3 kinase (PI3-K)/Akt pathway, are thought to beinvolved in this process.29 The genes induced by NF-kB topromote survival include cellular inhibitors of apoptosis (c-IAPs),30 FLICE (FADD-like IL-beta-converting enzyme)-inhibi-tory protein (FLIP),31–34 and the Bcl-2 homologous A1 and Bcl-xL.

35–37

Freshly isolated B-CLL cells have been shown to have highconstitutive levels of NF-kB activity,38 which might beattributable to the activation of either the Raf–MEK–ERK or/and PI3-Kinase–Akt pathways. The aim of this study was todetermine signaling pathways responsible for prolonged survivalof B-CLL cells. We tested the hypothesis that a sustained basalPI3K activity is essential for the prevention of the spontaneousapoptosis of B-CLL cells in culture. Murine fibroblasts were usedto mimic in vivo milieu.

Patients, materials, and methods

Patient specimens

Heparinized peripheral blood was obtained from 71 patients (47men and 24 women with a mean age of 67 years (range 40�93))diagnosed of B-CLL by standard clinical and laboratory criteria.1

Received 14 October 2003; accepted 2 April 2004; Published online3 June 2004

Correspondence: Dr P Perez-Aciego, Fundacion LAIR, Avda. deManoteras 50-52, E-28050 Madrid, Spain; Fax: þ 34 91 3830068;E-mail: paloma_perez@cilsp.comAlberto Durantez passed away on 10 January 2003.

Leukemia (2004) 18, 1391–1400& 2004 Nature Publishing Group All rights reserved 0887-6924/04 $30.00

www.nature.com/leu

Patients were either untreated or had not received treatment for3 months before the study. The study protocol was approved bythe institutional review board of the participating centers andwritten informed consent was obtained from all patients.

Purification of B-CLL lymphocytes

Mononuclear cells were isolated from heparinized peripheralblood of CLL patients by Ficoll density-gradient centrifugation.Monocytes and T cells were depleted by magnetic beadseparation using anti-CD2 and anti-CD14 magnetic beads(Dynabeads M450, Dynal Biotech, Oslo, Norway), accordingto the manufacturer’s instructions. After purification, stainingwith selected monoclonal antibodies (MoAbs) verified that theywere composed of greater than 95% of CD19þ CD5þ B cells.

Purification of normal B lymphocytes

Mononuclear cells were isolated from umbilical cord blood bycentrifugation over a Ficoll-Hypaque layer. Immature erythro-cytes were lysed using a solution containing 0.150 mM NH4Cl,1 mM KHCO3, and 0.1 mM Na2EDTA (pH 7.3). B cells werepurified by immunomagnetic selection techniques using a B-cellnegative isolation kit (Dynal Biotech) containing a mixture ofmouse MoAbs for CD2, CD3, CD7, CD14, CD16, and CD56,according to the manufacturer’s instructions. To further depleteadherent cells, additional anti-CD11b antibody (BD Bios-ciences, Erembodegem-Aalst, Belgium) was added. Immuno-phenotyping verified that the cell population was composed ofgreater than 90% of CD19þ , with more than 60% coexpressingCD5þ .

Reagents and antibodies

MoAbs specific for Bcl-xL FLIP and b-actin were purchased fromSigma (Sigma Chemical Co., St Louis, MO, USA). MoAbs againstAkt and phospho473-Akt were obtained from New EnglandBiolabs (Beverly, MA, USA); MoAb specific for X-linkedinhibitor of apoptosis protein (XIAP) from Medical & BiologicalLaboratories (Naka-Ku Nagoya, Japan); and MoAb specific forpolyADP-ribose polymerase (PARP) was purchased from BDBiosciences Pharmingen. The anti-CD40 agonistic MoAb 89was purchased from Immunotech (Marseille, France). Proteinkinase inhibitors Wortmannin (Wm), PD 98059 (Calbiochem, LaJolla, CA, USA), and LY294002 (Cell Signaling Technologies,Beverly, MA, USA) were dissolved in DMSO and further dilutedin cell culture medium before use at 100 nM, 10mM, and 20 mM

final concentrations, respectively (final concentration of DMSOdid not exceed 0.05%).

Cell cultures

Purified normal and leukemic B lymphocytes were cultured inRPMI 1640 medium supplemented with 10% heat-inactivatedfetal bovine serum (Gibco, Paisley, UK), 2 mM L-glutamine,100 U/ml penicillin, and 100 mg/ml streptomycin. For activationthrough the CD40 antigen, freshly purified B-CLL cells werecultured in the presence of the anti-CD40 MoAb 89 presentedby a mouse Ltk� cell line stably expressing CD32 (CD32 L cells),which was kindly provided by J Banchereau. Briefly, 5� 105 Bcells were cultured in 24-well flat-bottom plates under a finalvolume of 1 ml with 5� 104 mitomycin C-treated CD32 L cells

and the anti-CD40 MoAb (0.5mg/ml). In some experiments, Bcells were preincubated with the inhibitors Wm, LY294002, andPD 98059 for 30 min with and without the anti-CD40 MoAb 89.In all the experiments in which PI3K inhibitors were used,DMSO 0.05% was added to the control wells that contained theculture medium alone. Cell activation was assesed by flowcytometry using anti-CD69, anti-CD95, anti-CD25, anti-CD80,and anti-CD38 MoAbs.

Cell surface staining

Cell staining was performed by direct immunofluorescence.Briefly, peripheral blood mononuclear cells or purified Blymphocytes were resuspended in PBS and washed priorincubation for 20 min with the MoAbs at room temperature.Antibodies used for cell surface staining were purchased fromBD Bioscences. Flow cytometric analyses were performed on aFACSort (Becton Dickinson) using the CellQuest softwaresystem.

Proliferation assay

Purified B-CLL cells were cultured on confluent Ltk� monolayersin the presence or absence of anti-CD40 MoAb 89 for 5 days.Bromodeoxyuridine (BrdU) was added for the last 18 h ofculture. B cells were then recovered and the percentage ofproliferating cells was analyzed by flow cytometry using aproliferation kit (BrdU Flow kit, BD Bioscences), according tothe manufacturer’s instructions.

Measurement of apoptosis by annexin V/propidiumiodide double staining

Apoptotic cells were quantified by flow cytometry after annexinV-FITC and propidium iodide (PI) labelling. Briefly, 105 cellswere washed with ice-cold PBS and were incubated for 15 minin the dark in 50ml of binding buffer containing annexin V-FITC/solution (Roche, Barcelona, Spain) and 10 mg/ml PI (Sigma).

Analysis of cytoskeletal morphology

To stain filamentous actin (F-actin) for flow cytometry, cellswere fixed with PBS containing 3.7% formaldehyde for 1 h onice and then left for 2 h in PBS containing 0.1% triton X-100 and200 mU Bodipy 558/568-phalloidin (Molecular Probes, Leiden,Netherlands) at room temperature. The cells were thenresuspended in PBS and analyzed by flow cytometry.

Measurement of caspase-3 activation

Activation of caspase-3 was quantified by flow cytometry.Briefly, 5� 105 cells fixed with 3.7% formaldehyde in PBS for10 min on ice. The fixed cells were washed with PBS and thenincubated for 10 min in PBS containing 0.1% triton X-100, 1.5%fetal bovine serum prior incubation at 41C for 30 min with thespecific antiactive caspase-3 antibodies.

The PI3K/NF-jB pathway and apoptosis of B-CLL cellsS Cunı et al

1392

Leukemia

RNA preparation and reverse transcriptase-polymerasechain reaction (RT-PCR)

Total RNA was extracted from 3 to 5� 106 B-CLL cells usingguanidinium thiocyanate method (TRIZOL, Gibco). The obtainedRNA was reverse transcribed into cDNA using 20U AMVreverse transcriptase (Roche), according to the manufacturer’sinstructions. cDNA was then amplified by PCR in a GeneAmpPCR system 2400 cycler (Perkin Elmer, Norwalk, CT, USA).The PCR reaction was performed in 50ml with 0.2 mM dNTPs,1�GeneAmp buffer (15 mM MgCl2) (Amersham Biosciences,Little Chalfont, UK), 1 U Taq DNA polymerase (AmershamBiosciences), and 0.4 mM of 50 ACT TGT GGC CCA GAT AGGCAC CCA G 30 (forward) and 50 CGA CTT CGC CGA GAT GTCCAG CCA G 30 (reverse) for bcl-2, and 50 TTG GAC AAT GGACTG GTT GA 30(forward) and 50 GTA GAG TGG ATG GTC AGTG 30 (reverse) for bcl-xL (Tib Biomol, Berlin, Germany). 18SrRNA primers and competimers (QuantumRNAt 18S, Ambion,Huntingdon, UK) were added into the PCR reaction as internalstandard for relative quantitation. Cycling conditions were asfollows: 1 cycle at 941C for 5 min, 30 cycles at 941C for 30 s,601C for 1 min, 721C for 30 s, and a final extension at 721C for5 min. Cycling conditions for bcl-xL included 10 previous cycleswith an annealing temperature of 701C and then 20 cycles at601C. In total, 10ml of each sample was subjected to separationby electrophoresis in 2% agarose gel and stained with ethidiumbromide. Densitometric analyses were performed using theQuantity Ones software system (Bio-Rad Laboratories, Her-cules, CA, USA).

Western blotting

Cellular lysates equivalent to 5� 106 B cells were preparedusing a lysis buffer containing 50 mM Tris, pH 7.6, 150 mM

NaCl, 5 mM EDTA, 1% NP-40, 2 mM phenylmethylsulfonylfluoride, 10mg/ml leupeptin, 10 mg/ml aprotinin, 10mg/mlpepstatin and, as appropriate, with the phosphatase inhibitorNa orthovanadate (0.1 mM). Protein concentrations were deter-mined using the BCA protein assay kit (Pierce, Rockford, IL,USA). Cell lysates were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and blotting was performedon polyvinylidene difluoride membranes (Amersham Bios-ciences). Blots were developed using an enhanced chemilumis-cence detection system (Amersham Biosciences). For someanalyses the blots were reprobed after exposure to strippingbuffer (100 mM b-mercaptoethanol, 2% SDS, 62.5 mM Tris, pH6.8) for 30 min at 501C. Densitometric analyses were performedusing the Quantity One software system (Bio-Rad).

Electrophoretic mobility shift assay (EMSA)

Nuclear proteins extraction and generation of 32P end-labelledprobe were performed as reported previously.39 Nuclearproteins (2 mg) were incubated for 10 min in a final volume of10ml with 200 ng poly(dIdC) (Amersham Biosciences) beforeaddition of 0.4 ng of 50-end labelled ds-oligonucleotide kB,which contains an NF-kB binding site (50 CAA CGG CAG GGGAAT CTC CCT CTC CTT 30).40 Binding reactions were allowedduring 30 min at room temperature and electrophoresed in a 5%polyacrylamide gel in TBE (0.5� ). DNA–protein complexeswere detected by X-ray film exposition. Every reaction wasassayed twice. For supershift assays, the binding mixtures werepreincubated with the indicated antibody for 3 h before adding

the end-labelled probe. Competition assays were carried out inthe presence of a 50-fold molar excess of cold oligonucleotide10 min before the addition of the probe. An oligonucleotide witha sequence not recognized by NF-kB was used as an unrelatedcompetitor.39

Results

In vitro apoptosis of B-CLL cells is temporally associatedwith downregulation of bcl-xL, which is prevented bycontact with Ltk� mouse fibroblasts

Proclivity of B-CLL cells to spontaneous apoptosis in vitro maybe associated with loss of expression of antiapoptotic proteinsresulting from disruption of microenvironmental survival sig-nals. To study key signaling events involved in the transmissionof these survival signals, we first examined the effect of Ltk�

coculture on malignant cell survival and on bcl-2 and bcl-xL

expression using CD40 ligation as a control for comparison. Forthese experiments, purified B cells from nine CLL patients werecultured with conventional culture medium alone or onmitomycin-treated confluent Ltk� fibroblasts in the presence orabsence of an agonistic MoAb to CD40. B-CLL cells were thenrecovered every 24 h for up to 5 days, and bcl-2 and bcl-xL

mRNA levels were analyzed by RT-PCR. At the same time pointsthe percentage of surviving B-CLL cells was quantified byannexin V/PI double staining followed by flow cytometricanalysis. As illustrated in Figure 1a, B-CLL cells cultured inculture medium alone showed a dramatic increase of apoptosisafter 72 h of culture, which was highly prevented by contact

Figure 1 B-CLL cell spontaneous apoptosis in vitro correlates withdownregulation of bcl-xL, which is prevented by Ltk� coculture.Freshly isolated B cells from nine CLL patients were cultured in theculture medium alone or in coculture with a murine cell line offibroblasts (Ltk�) in the presence or absence of an agonistic antibody toCD40 for different time points (d, days). (a) The percentage of viablecells was determined by flow cytometry analysis of annexin-V and PIdouble staining up to 5 days of culture (mean7s.d.). (b) bcl-2 and bcl-xL, mRNA levels in B cells of a representative patient at the time ofisolation (t0), or cultured for the indicated times. The level of productfrom the gene of interest was then normalized against the constantexpression of the ribosomal RNA 18 s. RT-PCR products wereanalyzed on a 1% agarose gel and stained with ethidium bromide.

The PI3K/NF-jB pathway and apoptosis of B-CLL cellsS Cunı et al

1393

Leukemia

with Ltk� cells. Anti-CD40 presented by Ltk� did not increasesignificantly the effect of Ltk� in promoting B-CLL cell survivalprobably because the presence of Ltk� alone was sufficient forpreventing apoptosis.

On the other hand, B-CLL cells cultured in the culturemedium alone showed stable high expression of bcl-2 over the5-day culture period (Figure 1b, left panel), indicating that thehigh constitutive expression of bcl-2 alone was insufficient toprevent apoptosis in vitro. By contrast, bcl-xL mRNA levels wererapidly reduced in B-CLL cells upon culture in the culturemedium for 24 h (data not shown). After 72 h of culture therewere no detectable levels of bcl-xL while coculture with Ltk�

cells prevented bcl-xL downregulation (Figure 1b, rightpanel). CD40 crosslinking did not further increase the highconstitutive expression of bcl-2 (data not shown), but notablyupregulated bcl-xL expression, as previously described innormal B cells.41,42

Based on the finding that spontaneous apoptosis wasassociated with reduction of bcl-xL expression, we postulatedthat the signaling pathway involved in bcl-xL regulation in CLLcells could be crucial for their survival.

Spontaneous apoptosis of B-CLL cells is prevented byLtk� murine fibroblasts via the PI3K pathway

In many cell types, matrix attachment has been shown to resultin activation of the PI3K/Akt and Raf-MEK-Erk pathways,43,44

protecting epithelial and endothelial cells against matrixdetachment-induced cell death.45,46 We then evaluated thepossibility that the Raf–Mek–Erk and the PI3K/PKB pathwayswould be involved in the promotion of B-CLL cells survivalmediated by Ltk�. To this purpose, B-CLL cells were preincu-bated with the Raf inhibitor PD 98059 and the PI3K inhibitorWm before culture with Ltk� cells. In initial experiments, PD98059 and Wm inhibitors were tested over a range of doses anddetermined that there was no significant effect in cell viability inB-CLL cells cultured in the absence of Ltk� fibroblasts withdoses up to 10 mM and 100 nM, respectively. The inhibition ofMEK did not reverse Ltk�-induced protection of B-CLL cells fromapoptosis, whereas the inhibition of PI3K activity abrogatedalmost completely the Ltk�-mediated B-CLL cell survival in 14cases analyzed, as measured by annexin-V/PI double staining(Figure 2a, left panel). The organization of cytoskeletal isknown to be sensitively affected in cells undergoing apoptosis.47

To confirm the results obtained on annexin V/PI doublestaining, we used the disruption of F-actin organization,measured as F-actin-specific fluorescence from bodipy 558/568-conjugated phalloidin, as a marker of apoptosis. Flowcytometric measurements of F-actin-specific fluorescenceshowed a clear decrease in F-actin staining in those cells thathad been cultured in the culture medium alone for 48 h,whereas Ltk� coculture prevented the decrease in F-actinstaining (Figure 2a, right panels). Preincubation of cells withWm before being cultured on Ltk� monolayers inhibited Ltk�-mediated protection from apoptosis. Therefore, it appears thatinteractions with Ltk� cells protect B-CLL cells from sponta-neous apoptosis through the PI3K but not through the Raf–MEK–Erk pathway.

The possibility that PI3K might regulate the expression ofbcl-xL in B-CLL cells was explored. Densitometric analysis ofRT-PCR products in five CLL cases tested showed that Ltk�

coculture reproducibly induced a clear increase in bcl-xL

expression, and that this increase was arrested by 100 nM Wm(data not shown). The increase in bcl-xL levels mediated by Ltk�

coculture were further augmented in B-CLL cells triggered bycrosslinking of CD40, whereas CD40-mediated bcl-xL inductionwas markedly reduced by PI3K inhibition (Figure 2b upperpanel). Moreover, bcl-xL expression data were then confirmed atprotein level by Western blot in five cases of B-CLL (Figure 2b,lower panel). The use of the second inhibitor of PI3K,LY294002, showed similar results to those obtained with Wm.LY29004 at 20mM dose abrogated almost completely the Ltk�-mediated B-CLL cell survival in eight cases analyzed (29.578.2with culture medium alone; 95.372.2 with Ltk; and 55.5716.1with Ltk and LY294002). In these conditions, LY294002 alsoarrested the increase in bcl-xL expression induced by Ltk�

coculture, which was demonstrated by analysis of RT-PCRproducts and at the protein level by Western blotting (Figure 2cupper and lower panels, respectively). Altogether, PI3K activityis both necessary and sufficient for bcl-xL induction and survivalof B-CLL cells upon contact with Ltk� cells.

The effect of Ltk� coculture on bcl-xL expression andspontaneous apoptosis in normal control B cells was studied.In four samples tested, purified B cells from umbilical cordsamples underwent a high percentage of apoptosis within 24 h(X80% apoptosis). Ltk� fibroblasts neither sustained the survivalof these cells in culture nor the expression of Bcl-xL, as opposedto that observed in B-CLL cells cultured in the same conditions(Figure 2d).

To confirm that Ltk� coculture induces PI3K activity in CLLcells, we analyzed Akt activation by Western blotting usingantibodies specific for the active form, which is phosphoylatedat Ser473. Densitometric analysis from three separate specimensrevealed that Akt phosphorylation was readily evident in freshlyisolated B-CLL cells (no culture, Figure 3a and b). However, thiswas dramatically reduced within 48 h of culture in the culturemedium alone. In agreement with viability data, in six CLLsamples analyzed, Ltk� coculture induced the phosphorylationof Akt. Incubation of B-CLL cells with Wm abrogated theinduction of Akt phosphorylation mediated by Ltk� coculture(Figure 3a and b). Densitometric analysis of Akt phosphorylationusing another PI3K inhibitor, LY294002, showed similar resultsin three cases analyzed (Figure 3a, right panel shows arepresentative patient), demonstrating that Ltk� cocultureinduced PI3K activity in B-CLL cells. No relationship betweenAkt phosphorylation and survival of anti-CD40-stimulated Bcells was observed. There was a three-fold increase in Aktphosphorylation in cells stimulated with anti-CD40 comparedwith cells cultured with Ltk� alone, although the percentage ofviable cells was not significantly affected by the presence ofanti-CD40 (Figure 3b and d). As recent evidence showed thatAkt activity is also involved in cell cycle progression,48–51 weinvestigated the possibility that CD40-triggered increase in Aktphosphorylation might be associated with activation andproliferation of B-CLL cells in vitro. Changes in cell size(Figure 4a) and the upregulation of several lymphocyteactivation markers (data not shown) in response to CD40-mediated signals were firstly analyzed by flow cytometry. Then,the proliferative response of B-CLL cells to CD40 engagementwas assessed by 5-bromo-2-deoxyuridine (BrdU) incorporation.As shown in Figure 4b, CD40 crosslinking increased thepercentage of B cells that incorporate BrdU compared withcells cultured with Ltk� alone (negative control) in all six CLLsamples analyzed. In order to test the role of PI3K in the CD40-mediated B-CLL cell proliferation, cells were incubated withWm before stimulation with anti-CD40. PI3K inhibitionabrogated completely the induction of DNA synthesis (datanot shown), indicating that CD40 crosslinking induces B-CLLcell proliferation via PI3K.

The PI3K/NF-jB pathway and apoptosis of B-CLL cellsS Cunı et al

1394

Leukemia

PI3K/Akt activity prevents caspase-3 activationin B-CLL cells

Given that PI3K has been shown to protect cells from apoptosisin a caspase-dependent manner,52,53 the role of PI3K in caspaseinhibition was examined in seven CLL cases. After 48 h ofculture in medium alone, B-CLL cells presented a highpercentage of activated caspase-3 as measured by flowcytometry using specific antibodies for the active form.However, the percentage of caspase-3 activation was clearlyreduced in those cells that had been cocultured with Ltk�.When cells were incubated with Wm before culturing with Ltk�,the percentage of activated caspase-3 was similar to thatobserved in B-CLL cells cultured in medium alone (Figure 5aand b), indicating that maintenance of the PI3K pathway isrequired to prevent the activation of caspase-3. As PARP iscleaved by caspase-3 to yield an 85-kDa product signature, weanalyzed PARP fragmentation by Western blot to confirm the

findings of caspase-3 activation. The results of PARP cleavagewere always similar to data of flow cytometric analysis(Figure 5c). In the presence of Ltk� all PARP is present in itsintact form, whereas after preincubation with Wm the propor-tion of cleaved PARP is equivalent to that of the cells cultured inthe culture medium alone.

PI3K/Akt activity results in activation of NF-kBin B-CLL cells

It has been shown that B-CLL cells exhibit high constitutivelevels of NF-kB activity.38 Since there were previous reportslinking NF-kB activity to the activation of Akt, and othersdemonstrating bcl-xL as an NF-kB target gene, our findingssuggested that NF-kB could be implicated in the Ltk�-mediatedB-CLL cell survival (downstream of PI3K). On the other hand, itwas previously confirmed that the NF-kB-mediated increase of

Figure 2 PI3K inhibition abolishes Ltk�-mediated protection from apoptosis in B-CLL cells and blocks Ltk�-induced Bcl-xL expression. PurifiedB-CLL cells were cultured for 2 days with medium alone (containing 0.05% DMSO), with Ltk� cells without or upon preincubation with PI3Kinhibitor Wm, and with Ltk� cells in the presence of anti-CD40 MoAb without or upon preincubation with PI3K inhibitors. (a) In 14 patients, thepercentage of viable cells in each circumstance was quantified by flow cytometry after annexin-V and PI double staining (left panel). Each symbolrepresents the percentage of viability for a particular patient, and bars indicate the mean of data point in each condition. A paired t test was used forcomparisons (P-values shown). Right panels show viability data of a representative patient. Dot plots represent annexin-V and PI double staining,and histograms represent the percentage of F-actin content measured as F-actin-specific fluorescence from Biodipy 558/568-conjugated phalloidin.(b) After 48 h, B cells were recovered from cultures and subjected to total RNA isolation and analysis of bcl-xL expression by RT-PCR (upper panel).To confirm RT-PCR results, protein-containing lysates of the same samples were prepared for Western blot analysis (lower panel). In all, 12.5mg ofprotein was subjected to 10% SDS/PAGE and immunoblotting with antibodies specific for Bcl-xL, and for b-actin to ensure equal protein loading (arepresentative case is shown). (c) The same as (b), but using 20mM LY294002 instead of Wm. (d) B cells purified from four umbilical cord bloodsamples and from specimens of nine patients were cultured as indicated. The percentage of viability (upper panel) was measured by doublestaining with annexin-V and PI (mean7s.d.). B cells purified from a representative umbilical cord blood and from two representative patientspecimens were cultured for 2 days with Ltk� cells. B cells were recovered from cultures and lysates were prepared for immunoblotting analysisusing antibodies specific for Bcl-xL and b-actin (lower panel).

The PI3K/NF-jB pathway and apoptosis of B-CLL cellsS Cunı et al

1395

Leukemia

CD6954 was abrogated in our cultures by Wm treatment (culturewith culture medium alone, 22715% of CD69 positive cells;culture with Ltk�, 60715%; culture with Ltk� before Wmtreatment, 25710%).

To rule out this possibility, we analyzed NF-kB bindingactivity by EMSA in B-CLL cells, immediately after purification(Figure 6a, left panel) and after 48 h of culture using the culturemedium alone, with Ltk� cells in the presence or absence ofWm, and with Ltk� cells in the presence of anti-CD40 (Figure 6a,right panels). Densitometric analysis in three CLL specimenstested showed that immediately upon isolation (no culture)B-CLL cells presented high levels of NF-kB activity. However,culture with the culture medium alone resulted in the loss of NF-kB activity within 12 h (data not shown). Consistent with ourprevious findings, Ltk� coculture sustained a moderate level ofNF-kB activity in B-CLL cells and Wm abrogated the NF-kBactivity mediated by Ltk� (Figure 6a, right panels). Furthermoreand in agreement with other studies,38 stimulation of B-CLL cellswith anti-CD40 induced high levels of NF-kB activity. Toconfirm the specificity of the NF-kB signal, a cold oligonucleo-tide competition assay with 50-fold molar excess of anunlabeled and unrelated competitor39 was used (Figure 6b).Supershift assay showed that NF-kB complexes translocatedpredominantly p50–p65 heterodimers to the nuclear extracts(Figure 6c), as previously reported in normal and B-CLL cells.55

These observations indicate that Ltk�-mediated activation ofPI3K leads to NF-kB activity and that such activity may beinvolved in the protection of B-CLL cells from spontaneousapoptosis.

According to the findings of a correlation between PI3Kactivation and caspase-3 inhibition as well as of NF-kBactivation by PI3K activity, it was postulated that caspase-3inhibition could be mediated by NF-kB target proteins known toinhibit caspase activity. To test this hypothesis we determinedthe expression of X-linked inhibitor of apoptosis protein (XIAP)and FLIPL proteins by Western blotting in B-CLL cells after 48 hof culture in the culture medium alone, with Ltk� without orupon preincubation with Wm, and with Ltk� and anti-CD40.Both FLIPL and XIAP were upregulated in cells cultured with

Figure 3 Ltk� coculture prevents the loss of PI3K/Akt activity in B-CLL cells. Freshly purified B-CLL cells were cultured with media alone(containing 0.05% DMSO), with Ltk� cells without or upon preincubation with the PI3K inhibitor Wm, and with Ltk� cells in the presence of anti-CD40 MoAb. (a) Analysis of Akt activation of three representative patients. Whole-cell lysates were prepared from cells immediately upon isolation(no culture) and from cells cultured at the above mentioned conditions for 48 h. Protein-containing lysates were then resolved by SDS-PAGE,followed by Western blotting with antiphosphoserine Akt MoAb. Blots were stripped and reprobed with total anti-Akt MoAb. (b) Densitometricevaluation of Western blotting results. Data are expressed as Akt fold activation in each circumstance relative to B-CLL cells cultured in the culturemedium alone (the total Akt band was used as control for the normalization of samples). The mean7s.d. of six separate experiments is represented,except in the circumstance of freshly isolated cells (no culture), which was assayed only in three CLL cases. (c) The same as (a) using LY294002instead of Wm as PI3K inhibitor (a representative case is shown). (d) Percentage of viable cells in each circumstance quantified by annexin-V andPI double staining.

Figure 4 CD40 crosslinking induces DNA synthesis in B-CLLcells. Freshly purified B-CLL cells were cultured for 5 days with Ltk�

fibroblasts in the presence or absence (negative control) of anti-CD40MoAb. (a) Flow cytometric analysis of forward and side scatter ofB-CLL cells of a representative patient. (b) BrdU incorporation inresponse to anti-CD40 assayed in six CLL patients.

The PI3K/NF-jB pathway and apoptosis of B-CLL cellsS Cunı et al

1396

Leukemia

Ltk� compared with cells cultured in the culture medium alone(Figure 7a and b). Consistent with the results obtained by EMSAanalysis, this finding was not observed when cells wereincubated with Wm. On the other hand, CD40 crosslinking

clearly increased FLIPL but not XIAP expression when comparedwith cells cultured with Ltk� cells without anti-CD40. These resultssuggest that the maintenance of NF-kB activity in B-CLL cells maybe essential for preventing caspase-3 activity and apoptosis.

Figure 5 PI3K inhibition abrogate Ltk�-mediated prevention of caspase activity in B-CLL cells. Purified B-CLL cells from seven CLL sampleswere cultured in culture medium alone (containing 0.05% DMSO) or in coculture with Ltk� cells without or upon preincubation with Wm. After48 h of culture, cells were recovered and the percentage of active caspase-3-positive cells was determined by flow cytometry using specificantibodies: (a) representative flow cytometric analysis and (b) mean values7s.d. (c) A representative Western blot analysis showing the cleavage ofthe caspase-3 substrate PARP. The MoAb used detected both the 116 kd intact and 85 kd cleaved forms of PARP.

Figure 6 Ltk�-mediated PI3K activity leads to NF-kB activation in B-CLL cells. B-CLL cells from specimens of three patients were cultured inthe culture medium alone (containing 0.05% DMSO), with Ltk� cells without or upon preincubation with Wm, and with Ltk� cells and anti-CD40MoAb. (a) Nuclear extracts were prepared from B-CLL cells immediately after purification (no culture, left panel) or after 48 h of culture (rightpanels, ns denotes nonspecific bands). A total of 5 mg of protein from each nuclear extract were analyzed for NF-kB activity by EMSA using alabelled probe for NF-kB binding activity. (b) To confirm the specificity of the signals for NF-kB activity the nuclear extracts were incubated withthe labelled oligonucleotide in the presence or absence of a 50-fold molar excess of an unrelated and unlabelled competitor or NF-kB-bindingoligonucleotides. (c) Nuclear extracts were analyzed for specific NF-kB proteins by supershift EMSA using polyclonal rabbit antibodies to p50,p52, p65, and c-Rel, n specific NF-kB signal).

The PI3K/NF-jB pathway and apoptosis of B-CLL cellsS Cunı et al

1397

Leukemia

Discussion

Malignant B-CLL cells in vivo display an enhanced survival andtheir level of apoptosis remains low compared with that ofnormal B lymphocytes. By contrast, in vitro B-CLL cells diespontaneously and are difficult to keep alive. This observationsuggests that the malignant cell survival depends uponsupportive interactions within specific microenvironments. Ithas been reported that B-CLL cells accumulate in the bonemarrow and in neoplastic follicules, where they receive survivalor growth signals from bystander cells.56,57 One fundamentalstimulus in the bone marrow microenvironment is that mediatedby CD40. Owing to its antiapoptotic properties, this molecule isbelieved to exert profound influences on the growth of Blymphocytes and to direct T-lymphocyte-mediated rescue ofgerminal center B cells.9,10,58 Furman et al38 showed that CD40ligation inhibits apoptosis in CLL cells by inducing thetranscription factor NF-kB. On the other hand, human bonemarrow stromal cells have been described to prevent apoptosisand support the survival of B-CLL cells in vitro.5,6 Experimentalevidences have demonstrated that the murine Ltk� cell line cansubstitute for human bone marrow stromal feeder cells invitro.15 In an attempt to determine the molecular mechanismsinvolved in the survival and growth pattern of B-CLL cells, weused the Ltk� cell line in the presence or absence of an anti-CD40 MoAb, as a model to mimic in vivo tumor microenviron-ment.

It was observed that B-CLL cell proclivity to apoptosis wastemporarily associated with downregulation of bcl-xL. Ltk�

coculture prevented both spontaneous apoptosis and Bcl-xL

downregulation by a PI3K-dependent mechanism. The expres-sion of this antiapoptotic protein decreased rapidly upon cellculturing explaining why our data probably differ from thosepreviously reported. The inhibition of the MAPK pathway had

no significant effect in the Ltk�-mediated survival signals,whereas the inhibition of PI3K abrogated almost completelythe protector effect in cell survival mediated by Ltk�.Accordingly, Ltk� coculture is an adequate method forstudying the effect of constitutive PI3K activity on B-CLL cells.Unlike B-CLL, normal CD5þ B cells from the umbilical cordwere unable to survive when cocultured with Ltk� cells,suggesting a specific sensitivity of B-CLL cells to the antiapop-totic signals provided by Ltk� fibroblasts compared withnonmalignant B cells. An abnormal signaling regulation or theabnormal expression of a surface molecule leading to theenhanced response to survival signals observed in B-CLLcells may be postulated. Although the mechanism by whichLtk� cells induce the PI3K survival pathway in B-CLL cellsremains to be elucidated.

We detected constitutive Akt phosphorylation and NF-kBactivation in freshly purified B-CLL cells. However, both Akt andNF-kB activity decreased quickly once cultured in vitro,suggesting that in vivo activation may be maintained bymicroenvironmental factors. These findings are in agreementwith previous reports showing that B-CLL cells exhibit highconstitutive levels of NF-kB activity compared with nonmalig-nant human B cells.38 Bernal et al59 have reported that NF-kBactivity was reduced in B-CLL cells after 4 h of culture inconventional culture medium. In the same study, these authorsshowed that together with CD40 ligation, engagement of the Bcell receptor promotes both, PI3-K/Akt and NF-kB activity, andconsequently B-CLL survival.

Bcl-xL has been identified as an NF-kB target gene in othercellular types.36,60 We have shown that interaction with Ltk�

fibroblasts resulted in Akt and NF-kB activation in B-CLL cells ina PI3-K dependent manner. On the other hand, we observed thatinhibition of PI3K with Wm and LY294002 resulted in aninhibition of the bcl-xL gene (as shown by RT-PCR) and the Bcl-

Figure 7 PI3K-dependent NF-kB activity in B-CLL cells induces the expression of the XIAP and FLIPL proteins. Freshly purified B-CLL cellswere cultured for 2 days in the culture medium alone, with Ltk� cells without or upon preincubation with Wm, and with Ltk� cells in the presenceof anti-CD40 MoAb. (a) Representative analysis of XIAP and FLIPL expression by Western blotting of B-CLL cells after 48 h of culture in theindicated conditions. (b) Densitometric analysis of Western blot results. Data are expressed as protein fold induction in each circumstance relativeto B-CLL cells cultured in the culture medium alone after normalization for b-actin and represent the mean7s.d. (seven cases analyzed for XIAPand five for FLIPL).

The PI3K/NF-jB pathway and apoptosis of B-CLL cellsS Cunı et al

1398

Leukemia

xL protein (as shown by Western blot) (Figure 2b). Sinceinhibition of PI3K with Wm also inhibited NF-kB activity, itmay be postulated that PI3K induces Bcl-xL expression in B-CLLcells via NF-kB activation.

The Akt and NF-kB activity induced by Ltk� correlated withcaspase-3 inhibition and promotion of B-CLL cell survival.Recent evidence has implicated Akt as a signaling intermediateupstream of survival gene expression that is dependent on NF-kB activation.16 Moreover, in our study, the PI3K-dependentNF-kB activity mediated by Ltk� resulted in the expression ofXIAP and FLIPL, which are known to bind and inhibit the activityof the effector caspases 7 and 361 and caspase 8,62 respectively,probably preventing apoptosis of B-CLL cells. The levels of Aktphosphorylation and NF-kB activity, as well as the expression ofthe antiapoptosis proteins Bcl-xL and FLIPL, were furtherincreased in B-CLL cells stimulated by anti-CD40. However,interaction with Ltk� cells alone resulted in PI3K- and Akt-mediated survival signs, which were sufficient to maintain thecells viable. Additional CD40 crosslinking would furtherenhance PI3K and NF-kB activity levels, which may lead toB-CLL cell proliferation.

The presence of proliferation centers in lymphoid tissues frommost CLL cases has been described. These are follicle-likestructures which have been suggested to be the centers ofantigen-driven proliferation.63 In addition, the presence ofCD4þ T cells expressing CD40L in the proliferation centersand in the microenvironment of the bone marrow has beenshown.13,64 All these findings suggest that B-CLL cells in vivomight have special niches, where they receive growth signalsprobably through CD40 and through the B cell receptor, leadingto a high PI3K and NF-kB activity levels and, consequently, tothe expansion of the malignant clone. The survival of theresulting cells might be dependent on a basal level of PI3K/Aktand NF-kB activities, which probably are maintained byinteractions with the extracellular matrix and stromal cells. Thisbasal level of activity would be sufficient for their extendedsurvival, allowing them to reach again a proliferation promotingmicroenvironment. Finally, our finding that the PI3K-mediatedsignaling pathway is critical in the survival and growth of B-CLLcells suggests that chemotherapeutic intervention at variouspoints along the pathway may facilitate the entry of B-CLL cellsinto apoptosis. Blockade of the PI3K pathway might also preventthe expansion of the malignant clone of B cells. Precisedefinition of the components of the PI3K transduction pathwaycritical for the survival of B-CLL cells may potentially offerseveral attractive targets for new therapies in this disease. Forexample, it has been recently demonstrated that Akt stimulatesNF-kB predominantly by upregulating the transactivationpotential of p65, and that activated Akt requires IKK toefficiently stimulate the transactivation domain of the p65subunit of NF-kB.29 Then, small molecule inhibitors of the NF-kB-inducing kinases, IKKa and IKKb,65 and proteasome inhibi-tors that prevent IkB degradation66,67 could find applicability forthe treatment of the disease.

Acknowledgements

We thank the contribution to the study of B Contreras, I Losada, TMartın, I Martın-Larrauri, and S Rosado from Fundacion LAIR; MJCitores and C Puerta from Hospital Puerta de Hierro; and AMaraver and JF Rodrıguez from Centro Nacional de Biotecnolo-gıa; and Marta Pulido, MD, for editing the manuscript andeditorial assistance. This work was supported by Fundacion LAIR,Madrid, Spain.

References

1 Rozman C, Montserrat E. Chronic lymphocytic leukemia. N Engl JMed 1995; 333: 1052–1057.

2 Freedman AS, Boyd AW, Bieber FR, Daley J, Rosen K, Horowitz JCet al. Normal cellular counterparts of B cell chronic lymphocyticleukemia. Blood 1987; 70: 418–427.

3 Bannerji R, Byrd JC. Update on the biology of chronic lymphocyticleukemia. Curr Opin Oncol 2000; 12: 22–29.

4 Collins RJ, Verschuer LA, Harmon BV, Prentice RL, Pope JH, KerrJF. Spontaneous programmed death (apoptosis) of B-chroniclymphocytic leukaemia cells following their culture in vitro. Br JHaematol 1989; 71: 343–350.

5 Panayiotidis P, Jones D, Ganeshaguru K, Foroni L, Hoffbrand AV.Human bone marrow stromal cells prevent apoptosis and supportthe survival of chronic lymphocytic leukaemia cells in vitro. Br JHaematol 1996; 92: 97–103.

6 Lagneaux L, Delforge A, Bron D, De BC, Stryckmans P. Chroniclymphocytic leukemic B cells but not normal B cells are rescuedfrom apoptosis by contact with normal bone marrow stromal cells.Blood 1998; 91: 2387–2396.

7 Burger JA, Tsukada N, Burger M, Zvaifler NJ, Dell’Aquila M, KippsTJ. Blood-derived nurse-like cells protect chronic lymphocyticleukemia B cells from spontaneous apoptosis through stromal cell-derived factor-1. Blood 2000; 96: 2655–2663.

8 Pedersen IM, Kitada S, Leoni LM, Zapata JM, Karras JG, Tsukada Net al. Protection of CLL B cells by a follicular dendritic cell line isdependent on induction of Mcl-1. Blood 2002; 100: 1795–1801.

9 Banchereau J, Bazan F, Blanchard D, Briere F, Galizzi JP, VanKooten C et al. The CD40 antigen and its ligand. Annu RevImmunol 1994; 12: 881–922.

10 Liu YJ, Johnson GD, Gordon J, MacLennan IC. Germinal centres inT-cell-dependent antibody responses. Immunol Today 1992; 13:17–21.

11 Fluckiger AC, Rossi JF, Bussel A, Bryon P, Banchereau J, DefranceT. Responsiveness of chronic lymphocytic leukemia B cellsactivated via surface Igs or CD40 to B-cell tropic factors. Blood1992; 80: 3173–3181.

12 Osorio LM, Aguilar-Santelises M. Apoptosis in B-chronic lympho-cytic leukaemia. Med Oncol 1998; 15: 234–240.

13 Granziero L, Ghia P, Circosta P, Gottardi D, Strola G, Geuna Met al. Survivin is expressed on CD40 stimulation and interfacesproliferation and apoptosis in B-cell chronic lymphocytic leuke-mia. Blood 2001; 97: 2777–2783.

14 Banchereau J, de Paoli P, Valle A, Garcia E, Rousset F. Long-termhuman B cell lines dependent on interleukin-4 and antibody toCD40. Science 1991; 251: 70–72.

15 Buske C, Gogowski G, Schreiber K, Rave FM, Hiddemann W,Wormann B. Stimulation of B-chronic lymphocytic leukemia cellsby murine fibroblasts, IL-4, anti-CD40 antibodies, and the solubleCD40 ligand. Exp Hematol 1997; 25: 329–337.

16 Datta SR, Brunet A, Greenberg ME. Cellular survival: a play inthree Akts. Genes Dev 1999; 13: 2905–2927.

17 Durie FH, Fava RA, Foy TM, Aruffo A, Ledbetter JA, Noelle RJ.Prevention of collagen-induced arthritis with an antibody to gp39,the ligand for CD40. Science 1993; 261: 1328–1330.

18 Alessi DR, James SR, Downes CP, Holmes AB, Gaffney PR, ReeseCB et al. Characterization of a 3-phosphoinositide-dependentprotein kinase which phosphorylates and activates protein kinaseBalpha. Curr Biol 1997; 7: 261–269.

19 Stokoe D, Stephens LR, Copeland T, Gaffney PR, Reese CB, PainterGF et al. Dual role of phosphatidylinositol-3,4,5-trisphosphate inthe activation of protein kinase B. Science 1997; 277: 567–570.

20 Datta SR, Dudek H, Tao X, Masters S, Fu H, Gotoh Y et al. Aktphosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell 1997; 91: 231–241.

21 Cardone MH, Roy N, Stennicke HR, Salvesen GS, Franke TF,Stanbridge E et al. Regulation of cell death protease caspase-9 byphosphorylation. Science 1998; 282: 1318–1321.

22 Kops GJ, Burgering BM. Forkhead transcription factors are targetsof signalling by the proto-oncogene PKB (C-AKT). J Anat 2000; 4:571–574.

23 Beraud C, Henzel WJ, Baeuerle PA. Involvement of regulatory andcatalytic subunits of phosphoinositide 3-kinase in NF-kappaBactivation. Proc Natl Acad Sci USA 1999; 96: 429–434.

The PI3K/NF-jB pathway and apoptosis of B-CLL cellsS Cunı et al

1399

Leukemia

24 Romashkova JA, Makarov SS. NF-kappaB is a target of AKT in anti-apoptotic PDGF signalling. Nature 1999; 401: 86–90.

25 Kane LP, Shapiro VS, Stokoe D, Weiss A. Induction of NF-kappaBby the Akt/PKB kinase. Curr Biol 1999; 9: 601–604.

26 Ghosh S, May MJ, Kopp EB. NF-kappa B and Rel proteins:evolutionarily conserved mediators of immune responses. AnnuRev Immunol 1998; 16: 225–260.

27 Franke TF, Kaplan DR, Cantley LC. PI3K: downstream AKTionblocks apoptosis. Cell 1997; 88: 435–437.

28 Karin M, Ben Neriah Y. Phosphorylation meets ubiquitination: thecontrol of NF-[kappa]B activity. Annu Rev Immunol 2000; 18:621–663.

29 Madrid LV, Wang CY, Guttridge DC, Schottelius AJ, Baldwin Jr S,Mayo MW. Akt suppresses apoptosis by stimulating the transacti-vation potential of the RelA/p65 subunit of NF-kappaB. Mol CellBiol 2000; 20: 1626–1638.

30 Deveraux QL, Roy N, Stennicke HR, Van Arsdale T, Zhou Q,Srinivasula SM et al. IAPs block apoptotic events induced bycaspase-8 and cytochrome c by direct inhibition of distinctcaspases. EMBO J 1998; 17: 2215–2223.

31 Yeh JH, Hsu SC, Han SH, Lai MZ. Mitogen-activated proteinkinase kinase antagonized fas-associated death domain protein-mediated apoptosis by induced FLICE-inhibitory protein expres-sion. J Exp Med 1998; 188: 1795–1802.

32 Panka DJ, Mano T, Suhara T, Walsh K, Mier JW. Phosphatidyli-nositol 3-kinase/Akt activity regulates c-FLIP expression in tumorcells. J Biol Chem 2001; 276: 6893–6896.

33 Kazama H, Yonehara S. Oncogenic K-Ras and basic fibroblastgrowth factor prevent Fas-mediated apoptosis in fibroblaststhrough activation of mitogen-activated protein kinase. J Cell Biol2000; 148: 557–566.

34 Micheau O, Lens S, Gaide O, Alevizopoulos K, Tschopp J. NF-kappaB signals induce the expression of c-FLIP. Mol Cell Biol2001; 21: 5299–5305.

35 Wang CY, Guttridge DC, Mayo MW, Baldwin ASJ. NF-kappaBinduces expression of the Bcl-2 homologue A1/Bfl-1 to preferen-tially suppress chemotherapy-induced apoptosis. Mol Cell Biol1999; 19: 5923–5929.

36 Sevilla L, Zaldumbide A, Pognonec P, Boulukos KE. Transcrip-tional regulation of the bcl-x gene encoding the anti-apoptotic Bcl-xL protein by Ets, Rel/NFkappaB, STAT and AP1 transcriptionfactor families. Histol Histopathol 2001; 16: 595–601.

37 Lee HH, Dadgostar H, Cheng Q, Shu J, Cheng G. NF-kappaB-mediated up-regulation of Bcl-x and Bfl-1/A1 is required for CD40survival signaling in B lymphocytes. Proc Natl Acad Sci USA 1999;96: 9136–9141.

38 Furman RR, Asgary Z, Mascarenhas JO, Liou HC, Schattner EJ.Modulation of NF-kappa B activity and apoptosis in chroniclymphocytic leukemia B cells. J Immunol 2000; 164: 2200–2206.

39 Dorado B, Portoles P, Ballester S. NF-kappaB in Th2 cells: delayedand long-lasting induction through the TCR complex. Eur JImmunol 1998; 28: 2234–2244.

40 Kahn-Perles B, Lipcey C, Lecine P, Olive D, Imbert J. Temporaland subunit-specific modulations of the Rel/NF-kappaB transcrip-tion factors through CD28 costimulation. J Biol Chem 1997; 272:21774–21783.

41 Zhang XH, Li L, Choe J, Krajewski S, Reed JC, Thompson C et al.Up-regulation of Bcl-xL expression protects CD40-activated hu-man B cells from Fas-mediated apoptosis. Cell Immunol 1996;173: 149–154.

42 Ishida T, Kobayashi N, Tojo T, Ishida S, Yamamoto T, Inoue J.CD40 signaling-mediated induction of Bcl-XL, Cdk4, and Cdk6.Implication of their cooperation in selective B cell growth.J Immunol 1995; 155: 5527–5535.

43 Giancotti FG, Ruoslahti E. Integrin signaling. Science 1999; 285:1028–1032.

44 Schlaepfer DD, Hauck CR, Sieg DJ. Signaling through focaladhesion kinase. Prog Biophys Mol Biol 1999; 71: 435–478.

45 Khwaja A, Rodriguez VP, Wennstrom S, Warne PH, Downward J.Matrix adhesion and Ras transformation both activate a phosphoi-nositide 3-OH kinase and protein kinase B/Akt cellular survivalpathway. EMBO J 1997; 16: 2783–2793.

46 Le Gall M, Chambard JC, Breittmayer JP, Grall D, Pouyssegur J,Obberghen-Schilling E. The p42/p44 MAP kinase pathwayprevents apoptosis induced by anchorage and serum removal.Mol Biol Cell 2000; 11: 1103–1112.

47 Chow SC, Orrenius S. Rapid cytoskeleton modification inthymocytes induced by the immunotoxicant tributyltin. ToxicolAppl Pharmacol 1994; 127: 19–26.

48 Shin I, Yakes FM, Rojo F, Shin NY, Bakin AV, Baselga J et al. PKB/Akt mediates cell-cycle progression by phosphorylation ofp27(Kip1) at threonine 157 and modulation of its cellularlocalization. Nat Med 2002; 8: 1145–1152.

49 Liang J, Zubovitz J, Petrocelli T, Kotchetkov R, Connor MK, Han Ket al. PKB/Akt phosphorylates p27, impairs nuclear import ofp27 and opposes p27-mediated G1 arrest. Nat Med 2002; 8:1153–1160.

50 Medema RH, Kops GJ, Bos JL, Burgering BM. AFX-like Forkheadtranscription factors mediate cell-cycle regulation by Ras and PKBthrough p27kip1. Nature 2000; 404: 782–787.

51 Ahmed NN, Grimes HL, Bellacosa A, Chan TO, Tsichlis PN.Transduction of interleukin-2 antiapoptotic and proliferativesignals via Akt protein kinase. Proc Natl Acad Sci USA 1997; 94:3627–3632.

52 Berra E, Diaz-Meco MT, Moscat J. The activation of p38 andapoptosis by the inhibition of Erk is antagonized by thephosphoinositide 3-kinase/Akt pathway. J Biol Chem 1998; 273:10792–10797.

53 Gibbs BF, Grabbe J. Inhibitors of PI 3-kinase and MEK kinasedifferentially affect mediator secretion from immunologicallyactivated human basophils. J Leukocyte Biol 1999; 65: 883–890.

54 Lopez-Cabrera M, Munoz E, Blazquez MV, Ursa MA, Santis AG,Sanchez-Madrid F. Transcriptional regulation of the gene encodingthe human C-type lectin leukocyte receptor AIM/CD69 andfunctional characterization of its tumor necrosis factor-alpha-responsive elements. J Biol Chem 1995; 270: 21545–21551.

55 Romano MF, Lamberti A, Tassone P, Alfinito F, Costantini S,Chiurazzi F et al. Triggering of CD40 antigen inhibits fludarabine-induced apoptosis in B chronic lymphocytic leukemia cells. Blood1998; 92: 990–995.

56 Ghia P, Granziero L, Chilosi M, Caligaris-Cappio F. Chronic B cellmalignancies and bone marrow microenvironment. Semin CancerBiol 2002; 12: 149–155.

57 Ghia P, Caligaris-Cappio F. The indispensable role of microenvir-onment in the natural history of low-grade B-cell neoplasms. AdvCancer Res 2000; 79: 157–173.

58 Foy TM, Aruffo A, Bajorath J, Buhlmann JE, Noelle RJ. Immuneregulation by CD40 and its ligand GP39. Annu Rev Immunol1996; 14: 591–617.

59 Bernal A, Pastore RD, Asgary Z, Keller SA, Cesarman E, Liou HCet al. Survival of leukemic B cells promoted by engagement of theantigen receptor. Blood 2001; 98: 3050–3057.

60 Chen C, Edelstein LC, Gelinas C. The Rel/NF-kappaB familydirectly activates expression of the apoptosis inhibitor Bcl-x(L).Mol Cell Biol 2000; 20: 2687–2695.

61 Yang YL, Li XM. The IAP family: endogenous caspase inhibitorswith multiple biological activities. Cell Res 2000; 10: 169–177.

62 Hennino A, Berard M, Krammer PH, Defrance T. FLICE-inhibitoryprotein is a key regulator of germinal center B cell apoptosis. J ExpMed 2001; 193: 447–458.

63 Schmid C, Isaacson PG. Proliferation centres in B-cell malignantlymphoma, lymphocytic (B-CLL): an immunophenotypic study.Histopathology 1994; 24: 445–451.

64 Caligaris-Cappio F, Hamblin TJ. B-cell chronic lymphocyticleukemia: a bird of a different feather. J Clin Oncol 1999; 17:399–408.

65 Rossi A, Kapahi P, Natoli G, Takahashi T, Chen Y, Karin M et al.Anti-inflammatory cyclopentenone prostaglandins are direct in-hibitors of IkappaB kinase. Nature 2000; 403: 103–108.

66 Adams J, Palombella VJ, Elliott PJ. Proteasome inhibition: anew strategy in cancer treatment. Invest New Drugs 2000; 18:109–121.

67 Murray RZ, Norbury C. Proteasome inhibitors as anti-canceragents. Anticancer Drugs 2000; 11: 407–417.

The PI3K/NF-jB pathway and apoptosis of B-CLL cellsS Cunı et al

1400

Leukemia