We have previously described that BCR engagement induces a survival signal in MCL through an IL6/IL10-dependent activation-loop of STAT3 14 . To further investigate which BCR-induced signaling pathways are critical, we screened purified B cells from primary leukemic MCL for the differential expression of 84 genes upon anti-IgM stimulation using RT² Profiler PCR Arrays (SA Biosciences PAHS-027). Fifteen genes exhibited significant increased or decreased expression as compared to unstimulated cells (Additional file 1: Table S2; average fold-change from 4 individual experiments using UPN 5, 10, 13 and 14). Four genes were down-regulated (APAF1, CCNG2, ATM, BTG2), all corresponding to proapoptotic proteins. Conversely, eleven genes were overexpressed (CDK4, MCL1, NFkB1, IFNB1, SESN2, BCL2A, IL6, TNF, HK2, EGR-1, c-MYC), all of them being involved in cell cycle progression or inhibition of apoptosis. Within this group, three genes encoded for transcription factors; namely NF-kB, c-MYC and EGR-1 the two later being the two most upregulated genes upon anti-IgM stimulation (average of 23.6- and 17.6-fold increase, respectively).

BCR-induced expressions of c-MYC and EGR-1 were then confirmed by kinetic experiments in MCL cell lines (Granta-519 and HBL-2) (Figure 1A and B) and in MCL patients’ samples (Figure 1C). For MCL cell lines, basal levels of EGR-1 mRNA was rapidly increased within 30 min upon BCR ligation, peaked at 1 h and gradually returned to basal level within 3 to 6 hours. Similarly, EGR-1 protein levels increased upon anti-IgM stimulation and returned to basal level within 6 h (Figure 1B). A similar increase was observed for primary cells (UPN7 and UPN10) with EGR-1 proteins still detectable at 6 hours (Figure 1C). C-MYC expression was significantly induced upon BCR engagement in patients’ cells only (Figure 1A for Granta-519 and HBL-2 cells, Figure 1C for UPN7). The pattern of c-MYC mRNA induction differed from that of EGR-1 and displayed a constant increase at least up to 3 h associated with an increase of c-MYC protein (Figure 1C). We next evaluated the impact of BCR stimulation on a series of 7 patients’ samples (Figure 1D). Upon 1 h of anti-IgM stimulation, EGR-1 mRNA expression was highly upregulated in 4 out of 7 cases and clearly with a major extent as compared to induction of c-MYC (median fold increase: 30.1 and 2.98 for EGR-1 and c-MYC respectively). No correlation was observed between IGHV mutational status and intensity of BCR-induced responses (not shown).

Since EGR-1 has been described as a downstream target of JNK activation in various cellular models, we analyzed in MCL the involvement of JNK in the BCR-induced upregulation of EGR-1 and its role on MCL cell survival. In a characteristic patient sample (UPN9), basal JNK phosphorylation was slightly detected and was further enhanced following 5 min of BCR ligation with higher increase of phospho-JNK p46 (Figure 2A). Moreover, increase of BCR-induced phospho-JNK p46 was fully abolished in the presence of a selective inhibitor of JNK (SP600125, 20 μM) (Figure 2A). Inhibition of JNK by SP600125 induced a rapid down-regulation of EGR-1 mRNA expression in HBL-2 and Granta-519 cells associated with a subsequent decrease of EGR-1 protein (Figure 2B). Moreover, treatment with SP600125 upon anti-IgM stimulation also led to a blockade of BCR-induced EGR-1 upregulation in MCL cell lines (HBL-2 and Granta-519) and in primary MCL cells (UPN1) (Figure 2C). To confirm that EGR-1 was a downstream target of JNK in response to BCR activation, anti-IgM-stimulated HBL-2 cells were incubated with 5Z-7-Oxozeanol, an inhibitor of the transforming growth factor-β activated kinase-1 (TAK1) that is critical for BCR-induced JNK activation in B cells 29 . As shown in Additional file 2: Figure S1, treatment with 5Z-7-Oxozeanol (0.5 μM) completely abrogated BCR-induced upregulation of EGR-1. Overall, these results indicate that constitutive and BCR-induced EGR-1 expressions are dependent on JNK activation in MCL cells.

We next investigated the impact of JNK inhibition on MCL cell survival. Treatment of HBL-2 and Granta-519 cells with SP600125 for 48 h increased apoptosis (from 27% and 34% to 68% and 61% of apoptotic cells for HBL-2 and Granta-519, respectively) (Figure 3A). A similar increase of apoptosis was observed in MCL primary cells (36 ± 5% and 50 ± 3% for untreated and treated cells respectively, n=4) (Figure 3B). Moreover, BCR engagement induced in most cases a significant inhibition of spontaneous apoptosis (p=0.039) that was abrogated by a treatment with SP600125 (p=0.009) (Figure 3B). To confirm the involvement of EGR-1 in BCR-induced cell survival, MCL primary cells transfected with EGR-1 siRNA were stimulated with anti-IgM. As shown in Figure 3C, a reduction of 20% to 30% of cell survival was observed as compared to transfection with control siRNA. Collectively, these results indicate that EGR-1 is a downstream target of JNK in MCL cells and that JNK promoted constitutive and BCR-induced cell survival in MCL implicating notably EGR-1 induction.

The BCR signal is initially transmitted by LYN kinase leading to activation of various signaling pathways including JNK. We therefore evaluated the activation status of LYN in MCL cells and its involvement in cell survival. Using an anti-phospho-SFK recognizing the catalytic site of several Src kinases among which the Tyr397 of LYN, we detected in 9 out of 10 UPN cases tested (UPN 1,3,5,7,8,9,10,13,14) such a specific signal to variable extents of constitutive phosphorylation forming a 53–56 kDa doublet (Figure 4A and Additional file 3: Figure S2). We confirmed that this doublet corresponded to phospho-LYN by an immunoprecipitation assay using an anti-LYN antibody (Figure 4B). Considering the constitutive activation of LYN in MCL cells, we next evaluated the impact of PP2, a synthetic pyrazolopyrimidine selective inhibitor of SFK, and dasatinib (BMS-354825), an oral multi kinase inhibitor which also inhibits the trans-autophosphorylation of the active Tyr397 residue of LYN 30 . Treatment of primary cells with PP2 or dasatinib led to a dose-dependent decrease of Tyr397 LYN phosphorylation and complete inhibition was achieved up to 10 μM and 100nM for PP2 and dasatinib respectively (Figure 4C). Inhibition of phospho-Tyr397 LYN by PP2 was associated with a significant and dose-dependent increase of apoptosis rate (from 49±4% to 67±3% apoptotic cells for untreated and 24 h treated (10 μM) cells respectively; p=0.006; n=6) (Figure 4D). Treatment with dasatinib for 24 h also led to a significant and dose-dependent increase of apoptosis (from 46±5% to 64±5% apoptotic cells for untreated and treated (100nM) cells, respectively; p=0.0001; n=7) (Figure 4E). Remarkably, dasatinib had little apoptosis effect on phospho-Tyr397 LYN-negative cells (UPN4) at a concentration up to 200nM (Figure 4E, top panel). Altogether, these results indicate that MCL cells display a constitutive phosphorylation of BCR-associated LYN and that treatment with dasatinib or PP2 suppressed LYN activation and increased spontaneous apoptosis.

Since PP2 and dasatinib efficiently blocked activation of BCR-associated LYN in MCL cells, we next evaluated the impact of these compounds on JNK phosphorylation, EGR-1 expression and on cell survival upon BCR engagement. As shown in Figure 5A, a strong increase of phospho-Tyr397 LYN was observed in response to BCR ligation (lane 1 compared to lane 4) and treatment with dasatinib (100nM) completely blocked this effect while SP600125 that affect JNK did not. Similarly, PP2 decreased BCR-induced phospho-Tyr397 LYN in primary MCL cells (Figure 5B). Dasatinib also reduced BCR-induced phospho-JNK p46 (Figure 5C, lane 1 compared to lane 3 and lane 4 compared to lane 6), positioning JNK as a downstream target of LYN in response to BCR engagement. We next evaluated the impact of dasatinib on basal and BCR-induced level of EGR-1 as a target of JNK. As shown in Figure 5D (left panels), dasatinib (100nM) decreased basal expression of EGR1 mRNA and totally abrogated its upregulation in response to BCR ligation (UPN3, UPN13). Dasatinib also slightly decreased basal level of EGR1 protein and blocked its BCR-induced upregulation (Figure 5D, right panels and Additional file 4: Figure S3). Finally, we evaluated the impact of PP2 and dasatinib treatment on BCR-induced cell survival. Increasing concentrations of dasatinib abrogated the BCR-induced survival response in a dose-dependent manner and significantly suppressed this survival signal in all UPN cases tested (from 34 ±6% to 54±7% apoptotic cells for untreated and dasatinib-treated BCR-activated cells, respectively; n=6; p<0.001) (Figure 6A). Similarly, PP2 treatment also reduced or abolished BCR-induced cell survival (Figure 6B). Overall, these results highlight the importance of LYN, JNK and EGR1 as intermediates of BCR signaling in mediating survival signals in MCL cells and point out to the efficiency of dasatinib in suppressing cell survival signal emanating from the BCR.

Peripheral blood mononuclear cells were obtained from 14 MCL leukemic patients by Ficoll-Hypaque density gradient (Stem Cell Technologies, Grenoble, France). Lymphocytosis was greater than 8.0×10⁹/L and 10 out of 14 samples contained at least 80% of B lymphocytes (Additional file 5: Table S1). All B lymphocytes are monoclonal tumor B cells as evidenced through flow cytometry phenotyping of the surface immunoglobulin light chain (monotypic kappa or lambda). Seven cases (7/14) showed mutated IGHV and none of them displayed mutation in ITAM sequences of CD79B. The diagnosis of MCL was ascertained by immunophenotyping, cytogenetic and FISH analysis of t(11;14) and overexpression of cyclin D1 was detected by competitive RT-PCR according to the World Health Organization classification. All patients were homogeneously treated in a prospective trial of the GOELAMS group “Lymphome du manteau 2006 SA” 45 . All patients were provided written informed consent, validated by the Ethics Committee from the GOELAMS group, in accordance with the Declaration of Helsinki. Patients usually received treatment very quickly after sampling, making it difficulties to repeat all experiments several times on the same sample. Jeko-1, and Granta-519 cell lines were purchased from the German Collection of Microorganisms and Cell Cultures (DSMZ, Braunshwieig, Germany) and the HBL-2 cell line was a generous gift from Dr B. Sola (Caen, France) 46 .

Patients’ cells were either used freshly isolated or cryopreserved in liquid nitrogen in the presence of 10% dimethyl sulfoxide and 20% heat-inactivated FCS. MCL leukemic cells (3 × 10⁶ cells/ml) were cultured in complete RPMI 1640 medium supplemented with 10% heat-inactivated foetal calf serum. Jeko-1, HBL-2 and Granta-519 cell lines were maintained in culture in the same media. For BCR stimulation, plates were pre-coated with rabbit anti-human IgM antibody (10 μg/mL; Jackson ImmunoResearch, Baltimore, MD) as previously described 14 or the anti-IgM antibody was added to the culture medium at the same concentration for short stimulation time.

Antibodies to EGR-1 (mAb 44D5), c-MYC (mAb 9B11), phospho-Src family (mAb 100 F9) also reactive with phospho-Tyr397 LYN (catalytic site) and phospho-JNK (Thr183/Tyr185) were from Cell Signaling (Beverley, MA). Monoclonal mouse antibody (H-6) and polyclonal rabbit antibody (44) to LYN were from Santa-Cruz (Santa Cruz, CA). Anti-phosphotyrosine monoclonal antibody (clone 4 G10) was from Millipore (Billerica, MA). Dasatinib (Clini Sciences, Montrouge, France) was used at 100nM, unless otherwise stated. JNK inhibitor SP600125 and PP2 (4-Amino-3-(4-chlorophenyl)-1-(t-butyl)-1H-pyrazolo[3,4-d]pyrimidine) was from Sigma (Saint Quentin Fallavier, France) and (5Z)-7-Oxozeaenol (TAK1 inhibitor) was from Tocris Bioscience (Bristol, UK).

Tumor B-lymphocytes from MCL patients were purified by the RosetteSep® Human B Cell Enrichment Cocktail (Stemcell technologies, Grenoble, France). Cells were cultured for 3 hours upon BCR stimulation or left untreated. Total RNA were extracted and analyzed with “p53 signaling pathway” array (SA Biosciences PAHS-027) according to the manufacturer’s instructions (SA Biosciences, Frederick, MD) with an Applied Biosystems 7500 Fast Real-Time PCR Systems. Each gene expression was normalized to the mean Ct values from the four housekeeping genes available in the PCR array (β2-microglobulin, β-actin ribosomal protein L13A, hypoxanthine phosphorybosyl transferase-1 and β-actin), then normalized to unstimulated control cells to determine the fold-change. Relative fold change of expression was calculated by the ΔΔCt method and the values are expressed as 2^-ΔΔCt. All points were done in duplicate.

Cell apoptosis was evaluated using flow cytometry (FACSCantoTM II Becton Dickinson) on leukemic MCL PBMC after gating on CD19+ cells using Annexin V-FITC and propidium iodide staining (BD Biosciences, San Jose, CA). Percentage of apoptotic cells corresponded to% of annexin V-positive, including PI-negative and PI-positive cells. All measurements were done in duplicate and the mean is indicated.

RNA from unstimulated or anti-IgM stimulated cells were extracted using RNeasy Mini kit (QIAGEN) and EGR-1 and c-MYC expressions were analyzed by qRT-PCR using SYBR Green reagents. Results were normalized to the mean Ct values from cyclophilin A housekeeping gene then normalized to unstimulated control cells to determine the fold change. Relative fold change of expression was calculated by the ΔΔCt method and the values are expressed as 2 ^-ΔΔCt. All points were done in duplicate. The primers used for amplification (Sigma-Aldrich) were as follows:

EGR-1 forward primer (5^′CGAGCAGCCCTACGAGCACCTGAC3^′),

EGR-1 reverse primer (5^′TGCGCAGCTCAGGGGTGGGCTCTG3^′),

c-MYC forward primer (5^′GCTGCTTAGACGCTGGATTTTT3^′),

c-MYC reverse primer (5^′ACCGAGTCGTAGTCGAGGTCAT3^′),

cyclophilin A forward primer (5^′GCACTGGAGAGAAAGGATTTGG3^′) and

cyclophilin A reverse primer (5^′AGTGCCATTATGGCGTGTGA3^′).

Total protein extracts from 3×10⁶ MCL cells (PBMC) were separated on 10% polyacrylamide denaturing gel, transferred to a nitrocellulose membrane and incubated overnight with the appropriate antibody followed by a secondary horseradish peroxidase-conjugated antibody (Bio-Rad). Detection was performed using ECL (Amersham Biosciences, Buckinghamshire, UK) and autoradiography (Biomax, Kodak, Rochester, USA). For LYN immunoprecipitation, HBL-2 cells were lysed in 1% Nonidet P-40 (NP-40) lysis buffer for 30 minutes on ice. Insoluble material was removed by centrifugation at 27 000 g for 10 minutes at 4°C and soluble proteins were immunoprecipitated with a rabbit anti-LYN (H-6) antibody for 2 hours at 4°C. Immunocomplexes were solubilized in SDS sample buffer, analyzed on SDS-PAGE, transferred and subjected to immunoblotting as described above using either a mouse anti-phosphotyrosine antibody (clone 4 G10) or a mouse anti-LYN (44) antibody.

Three million cells were resuspended in 100 μL of Human B Cell Lymphoma Nucleofector® Kit (Amaxa) containing either 1 μM of EGR-1 siRNA (On-Target plus SMART pool Human EGR-1, Dharmacon) or 1 μM of control siRNA (Ctrl1: On-Target plus non-targeting pool; Ctrl2: On-target plus non-targeting siRNA#1; Dharmacon). Cells were transfected in a Nucleofector II device (Lonza) by using U-015 program, transferred to culture plates and western blot and apoptosis assays were performed as described above.

Differences between groups were determined using the Student’s t test. Statistical analyses were performed using GraphPad Prism software (San Diego, CA). P values below 0.05 were considered statistically significant.