Search

Accumulating evidence has designated B cells as central players in the pathogenesis of immune diseases. In the late 1990s, anti-CD20 monoclonal antibodies were developed for the treatment of B-cell non-Hodgkin lymphomas, offering the opportunity to efficiently deplete the B-cell compartment for therapeutic immunointerventions. Several studies have since established the beneficial effect of this drug on the course of a wide range of immune diseases. However, paradoxically, it has also been reported that rituximab sometimes worsens the symptoms of the very same conditions. The explanation that reconciles such apparently conflicting results has recently emerged from basic studies, which demonstrate that (1) B cells are also endowed with immune-regulatory properties and (2) the opposing contributions of B cells may overlap during the course of the disease. Caution should therefore be exercised when considering B-cell depletion because the therapeutic effect will depend on the relative contributions of the opposing B-cell activities at the time of the drug administration.

It was recently shown that bacterial thymus-independent (TI) antigens confer long-lasting immunity and generate memory B lymphocytes. However, reactivation of TI memory B cells is repressed in immunocompetent mice, thus raising the issue of the mechanism whereby TI vaccines confer immune protection. Here, we propose an explanation to this apparent paradox by showing that a Streptococcus pneumoniae capsular polysaccharide (PS) generates long-lived bone marrow (BM) plasma cells which frequency can be increased by CpG oligodeoxynucleotides (ODNs). The adjuvant effect of CpG ODNs on the PS3 Ab response is directly targeted to B cells and does not involve B-1a cells. We also demonstrated that BM plasma cells generated in response to the thymus-dependent (TD) form of the PS vaccine have a higher secretion capacity than those produced after immunization with the CpG-adjuvanted PS vaccine. Finally, we show that the PS-specific BM plasma cell compartment is sufficient to confer full protection of vaccinated mice against S pneumoniae infection. Altogether, our results show that TI antigens like their TD counterparts can generate both the lymphoid and the plasma cell component of B-cell memory. They also provide a framework for the improvement and widespread usage of TI vaccines.

We have studied the impact of B-cell receptor (BCR) or CD40 ligation on the in vitro chemotactic response of tonsillar B cells to 4 chemokines: stromal cell–derived factor (SDF)–1α, macrophage inflammatory protein (MIP)–3α, MIP-3β, and B-cell–attracting chemokine (BCA)–1. In the tonsil, SDF-1 and MIP-3α are both expressed in the crypt epithelium, while MIP-3β is found in the T zone and BCA-1 in the follicles. Resting virgin and memory B cells display a similar chemotaxis pattern, and they both have the potential to migrate in vitro to all 4 chemokines studied. This pattern of responsiveness is strongly modified by a surrogate antigen (Ag) but is not altered by CD40 ligand. We report here that surrogate Ag induces a profound and sustained suppression of the response to the crypt chemokines SDF-1α and MIP-3α, while it exacerbates the migratory response to MIP-3β. The effect of surrogate Ag on the response to BCA-1 is biphasic: After an initial phase of suppression, chemotaxis toward BCA-1 is strongly up-regulated. Our results suggest that Ag is primarily responsible for reprogramming the B-cell chemotaxis responsiveness during the humoral response. We propose that it initiates an ordered change of the chemotaxis machinery allowing Ag-activated B cells to relocate in the T zone and B-cell follicles sequentially.

Background Ponatinib, a third generation TKI, induces high rates of cytogenetic and molecular responses in heavily pre-treated CML patients (pts) resistant to ≥2 TKI and/or with a T315I mutation, especially in CP (J. Cortes at al., NEJM 2013). This agent induces diverse non-severe adverse events (AEs), a substantial proportion of pts experience severe arterial thrombotic events [(ATE) 17% by 3 years, J. Cortes et al., Haematologica 2015]. This agent is now licensed and it is mandatory to explore the rates of responses and ATE in the real-life setting. Aims We took the opportunity of the French ponatinib compassionate use program between May 2013 (end of PACE inclusions) and January 2014 (ponatinib license) to evaluate different outcomes in real-life conditions. Methods This is a multicenter observational retrospective study, designed to examine safety and efficacy of ponatinib in CML (any phase) resistant/intolerant to prior TKIs, in university and non-university hospitals, benefiting from the national ponatinib compassionate use program. Data were captured and validated following the rules and regulations of observational studies in France. Pts were analyzed in intention-to-treat, Molecular biology tests were performed according to ELN guidelines and BCR-ABL/ABL are expressed as % IS in all centers. Standard clinical [gender, age, weight, body mass index (BMI), cardiovascular risk factors (previous events, tobacco abuse, high blood pressure, diabetes)], onset of any CV events before and during ponatinib treatment) and metabolic biological parameters (cholesterol, triglycerides) were collected. Results Thirty-five observations were collected in CP, 4 in AP and 5 in BC. We focused our analysis on the 35 CP pts only. There were 16 (46%) males and 19 females, with a median age of 53 (18-76) years at CML diagnosis and 59 (20-82) years at ponatinib initiation. Sokal scores were high in 13 (37%), intermediate in 12 (34%), low in 4 (12%) and unknown in 6 (17%); Euro scores were high in 6 (17%), intermediate in 15 (43%), low in 3 (9%) and unknown in 11 (31%); Eutos scores were high in 7 (20%), low in 20 (57%) and unknown in 8 (23%) pts. All pts harbored "major" BCR-ABL transcripts except one (e19a2). Regarding cardiovascular risk factors (CVRF) prior to ponatinib, 12 pts (34%) were treated for hypertension, 1 was diabetic, 6 (17%) had dyslipidemia (all on statins). Tobacco abuse was present in 8 (23%) pts and 14 pts had some pre-existing CVRF in total. Median weight just prior ponatinib was 66 (48-107) kg, and BMI was 24.2 (17.85-33) kg/m2. Thirteen (37%) pts were on anti-aggregants or anti-coagulants (AAG/C) prior to ponatinib. All pts had received imatinib first-line for a median of 29.5 (4-123) months, 20 (57%) dasatinib and 14 (40)% nilotinib as second-line, one pt developed a T315I after imatinib only; 21 (60%) had received all 3 TKIs prior to ponatinib. At ponatinib initiation, 7 (20%) harboured a T315I mutation, 7 (20%) other mutation(s) 20 (57%) none. In one case mutation screen was not performed. The trigger for ponatinib was resistance (hematologic, cytogenetic or molecular progression) in 26 (74%) pts, intolerance in 7 (20%) pts and both in 2 (6%) pts. Pts were initiated at a median of 45 (30-45) mg daily after a median of 79.5 (12-217) months of disease duration. The median follow-up on ponatinib was 26 (2-37) months. The cumulative incidence of major molecular responses was 29% at 3, 42% at 6, 56% at 12 and 70% at 18 months. The overall survival is shown in figure 1. Four pts died, 3 of disease progression and 1 of myocardial infarction. Six (17%) pts had grade 3-4 hematologic AEs imposing transient ponatinib withhold, and 14 (40%) had diverse grade 1-2 non-hematologic, non-CV AEs (pancreatic, hepatic, skin toxicities, no grade 3-4). ATEs occurred in 19 (54%) pts after a median of 5.8 (0.7-21.7) months of ponatinib, without CVRF in 10 (53%) pts and without AAG/C in 13 (68%) pts. CVRF and AAG/C had no significant influence here on ATEs onset in univariate Cox model (p=0.76 and 0.37 respectively). Lipids and HbA1c were not modified on ponatinib. Overall 43% of pts stopped ponatinib for toxicity. Conclusion In the French compassionate use program in CP-CML patients resistant or intolerant to previous TKIs, ponatinib displayed strong efficacy, as previously described. In this unselected population of patients, ATEs were confirmed to represent the main tolerance concern in the real life setting. Disclosures Nicolini: Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Bristol-Myers Squibb: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Ariad Pharmaceuticals: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau. Coiteux:Novartis: Speakers Bureau; BMS: Speakers Bureau; ARIAD: Speakers Bureau. Charbonnier:Novartis: Speakers Bureau; BMS: Speakers Bureau; ARIAD: Speakers Bureau. Huguet:Novartis: Consultancy, Research Funding; PFIZER: Consultancy, Speakers Bureau; BMS: Consultancy, Speakers Bureau; ARIAD: Consultancy, Speakers Bureau. Etienne:ARIAD: Consultancy, Honoraria, Speakers Bureau; BMS: Consultancy, Honoraria, Speakers Bureau; Novartis: Consultancy, Honoraria, Other: Congress Travel/Accomodations, Research Funding, Speakers Bureau. Legros:Novartis: Research Funding, Speakers Bureau; ARIAD: Speakers Bureau; BMS: Speakers Bureau. Rousselot:Pfizer: Consultancy; BMS: Consultancy, Speakers Bureau; Novartis: Speakers Bureau. Amé:BMS: Speakers Bureau; Novartis: Speakers Bureau. DeFrance:Ariad: Consultancy. Mahon:ARIAD: Consultancy; Pfizer: Consultancy; Bristol-Myers Squibb: Consultancy, Honoraria; Novartis: Consultancy, Honoraria. Rea:Novartis: Honoraria; Bristol-Myers Squibb: Honoraria; Ariad: Honoraria; Pfizer: Honoraria.

Recent studies performed in the laboratory have established that interleukin-4 (IL-4) used in combination with anti-CD40 monoclonal antibody (MoAb) 89 presented on Ltk- mouse fibroblasts stably expressing human Fc gamma RII/CDw32 (referred to as the CD40 system) sustains long-term proliferation of normal human B cells. In the present study, B-cell chronic lymphocytic leukemias (B-CLLs) activated through slgs or CD40 were examined for their capacity to proliferate and differentiate in response to various cytokines. Our results indicate that the outcome of IL-4 stimulation on the in vitro growth of B-CLL depends on the signalling pathway used for their activation. Whereas IL-4 did not display any growth-stimulatory effect on B-CLL activated by Ig cross-linking agents, it could stimulate DNA synthesis and enhance the viable cell recovery when leukemic B cells were cultured in the CD40 system. Most B-CLL samples were induced for IgM synthesis upon Staphylococcus aureus strain Cowan I stimulation. This Ig response was potentiated by IL-2 and antagonized by IL-4. Anti-CD40 MoAb used alone or in combination with cytokines (IL-1 alpha to IL-6, interferon gamma, tumor necrosis factor gamma, and transforming growth factor beta) failed to induce Ig secretion from B-CLL cells. No evidence for Ig isotype switching was obtained with the cytokines listed above, regardless of the mode of activation. Taken together, our results suggest that B-CLL cells can be partially released from their apparent maturation block by IL-2 and Ig cross-linking agents. In contrast, combinations of IL-4 and cross-linked anti-CD40 antibodies induced entry of B-CLL cell into cycle, but poorly stimulated their differentiation into Ig secreting cells.

The pattern of in vitro growth response of freshly isolated non-Hodgkin malignant lymphoma B cells (NHML) to cytokines was investigated. Ten tumor specimens of low- or intermediate-grade malignancy were selected for study. To assess their proliferative capacity in vitro, B-lymphoma cells were activated through ligation of their surface Ig receptor with insolubilized anti-IgM antibodies or Staphylococcus aureus strain Cowan I (SAC). In the great majority of cases, interleukin-2 (IL-2) was the sole factor that significantly and reproducibly stimulated DNA synthesis in NHML activated through their surface Igs. Other B-cell tropic factors, including IL-4, IL-5, IL-6, and tumor necrosis factor- alpha (TNF-alpha), failed to elicit a growth response in most of the IL- 2-responsive neoplastic samples. However, one specimen among 10 exhibited the opposite pattern of response and proliferated following culture with IL-4 and anti-Ig reagents, but not after IL-2 stimulation. Three specimens could also be induced for DNA synthesis on cross- linking of their surface Igs in the absence of exogenous growth factors. Although IL-4 could not support the in vitro growth of the majority of NHML cases, it strongly suppressed the proliferative signals delivered to these cells by anti-Ig reagents used alone or in combination with IL-2. Our data suggest that, in most cases, IL-4 essentially provides growth-inhibitory signals to NHML when they are activated through their surface Ig receptors and as such may be considered to be a valid candidate for future therapy of this type of mature B-cell malignancy.

Background. Peripheral T-cell lymphomas (PTCLs) originate from post-thymic T cells but compared to B-cell lymphomas the exact cell of origin is usually unknown except for angioimmunoblastic T-cell lymphoma arising from a follicular helper T-cell. Furthermore, no recurrent cytogenetic or molecular abnormalities are identified in PTCLs. Recently, recurrent impairment of the p53 pathway has been pointed out in PTCLs. However, p53 knockout (KO) mice are known to develop immature thymic T-cell lymphomas and solid tumors but surprisingly PTCLs have not been reported for more than 20 years in those mice. NKT cells are a T-cell subset responsive to glycolipids presented by CD1d, a major histocompatibility complex (MHC) class I-like antigen-presenting molecule, in contrast to conventional T cells, which recognize peptide antigens. Two types of NKT cells have been described so far: type I or invariant NKT cells (iNKT) that express a Valpha14-Jalpha18 (in mice) or Valpha24-Jalpha18 (in humans) constant chain and type II NKT cells that express a variable TCR but are CD1d-dependent as well. Most type II NKT cells are of alpha/beta phenotype but CD1d-restricted gamma/delta T cells have also been described in mice and humans. Methods. The development of PTCLs in p53 KO mice (B6.129S2-Trp53tm1Tyj/J) was studied. Identification of PTCLs was made by immunohistochemistry and flow cytometry analysis. Gene expression profile analysis (GeneChip Mouse Genome 430 2.0 array, Affymetrix) was performed to characterize lymphomas developed in the mouse model. Transfer experiments were done by intravenously retro-orbital injection into syngeneic, immunocompetent C57Bl/6J WT animals or immunocompromised CD3ε-/- mice. Therapeutic trials in mice were performed with the use of blocking anti-CD1d monoclonal antibodies (mAb) (clone HB323; BioXcell). Results. We found that p53 KO mice developed well-known and characterized thymic T-cell lymphomas and solid tumors as previously described. However, about 20% of p53 KO mice spontaneously developed a previously unrecognized entity of PTCLs originating from CD1d-restricted iNKT cells (ie type I NKT cells) referred to as NKTLs for NKT lymphomas thereafter. Both alpha-galactosylceramide-CD1d tetramer staining and unique Valpha14-Jalpha18 TCR rearrangement confirmed the iNKT nature of these lymphomas. Chronic injection of Streptococcus pneumoniae (Spn), reported to express glycolipid antigens activating NKT cells, significantly increased the incidence of these NKTLs compared to a control group of p53 KO mice injected with PBS (P=0.03). Gene expression profile analysis indicated a significant down-regulation of genes in the TCR signaling pathway of NKTLs (false discovery rate q-value=0.01 by gene set enrichment analysis) suggesting an underlying antigenic chronic stimulation as previously reported in chronically activated T cells (Figure 1). Moreover, NKTLs were characterized by upregulation of PD-1 and loss of NK1.1 expression compared to resting NKT cells (P We did not identify human PTCLs arising from type I iNKT cells by using alphaGalCer-CD1d tetramer staining. However, using sulfatide-loaded CD1d tetramers (ie another type of glycolipid-CD1d tetramer identifying type II NKT cells), we identified CD1d-restricted human PTCLs among gamma/delta hepatosplenic T-cell lymphomas (HSTLs) and PTCL-not otherwise specified (PTCL-NOS) expressing the Vd1 TCR but not the Vd2 TCR (Figure 3). Conclusion. This demonstrates for the first time the existence of human PTCLs arising from gamma/delta CD1d-restricted type II NKT cells. These results refine the classification of PTCLs in humans by identifying a new cell of origin and pave the way for the development of blocking anti-CD1d antibodies for therapeutic purposes. Figure 1. Figure 1. Figure 2. Figure 2. Figure 3. Figure 3. Disclosures No relevant conflicts of interest to declare.

Tumor necrosis factor (TNF) superfamily members BAFF, or B-cell activation factor of the TNF family, and APRIL, a proliferation-inducing ligand, are involved in normal B-cell survival and differentiation. They interact with 3 receptors: BAFF-R, specific to BAFF; and TACI and BCMA, which are shared by BAFF and APRIL. We tested the potential role of these proteins in B-cell chronic lymphocytic leukemia (B-CLL) resistance to apoptosis. TACI and BAFF-R mRNAs were found in leukemic B cells. BAFF and APRIL mRNAs and proteins were detected in B-CLL leukemic cells and normal blood or tonsil-derived B lymphocytes. Yet, in contrast to normal B lymphocytes, BAFF and APRIL were expressed at the membranes of leukemic cells. Adding soluble BAFF or APRIL protected B-CLL cells against spontaneous and drug-induced apoptosis and stimulated NF-κB activation. Conversely, adding soluble BCMA-Fc or anti-BAFF and anti-APRIL antibodies enhanced B-CLL apoptosis. Moreover, a soluble form of BAFF was detected using surface-enhanced laser desorption/ionization–time-of-flight mass spectrometry (SELDI-TOF MS) in the sera of B-CLL patients but not of healthy donors. Taken together, our results indicate that B-CLL cells can be rescued from apoptosis through an autocrine process involving BAFF, APRIL, and their receptors. Inhibiting BAFF and APRIL pathways may be of therapeutic value for B-CLL treatment.

The expression of 3 lymphoid chemokines-macrophage inflammatory protein-3alpha (MIP-3alpha), stromal cell-derived factor-1 (SDF-1), and B-cell-attracting chemokine-1 (BCA-1)-in the tonsil and the possible correlation between their sites of expression and B-cell localization within this tissue were studied. The results show that all 3 chemokines are produced in the crypts but differ by the nature of the cells that produce them and their location within the crypt. SDF-1 and MIP-3alpha are produced by epithelial cells, but their secretion is mutually exclusive. Both MIP-3alpha- and SDF-1-expressing cells are in close contact with memory B cells. By contrast, BCA-1-producing cells in the crypt are not epithelial and form clusters colocalized with plasma cells. Altogether, these data suggest that the chemokines produced in the tonsillar crypt may (1) attract memory B cells to antigen and (2) recruit and retain plasma cells and memory B cells within the supportive epithelial microenvironment of the crypt. (Blood. 2001;97:3992-3994)

Two subtypes of G-protein–coupled cannabinoid receptors have been identified to date: the CB1 central receptor subtype, which is mainly expressed in the brain, and the CB2 peripheral receptor subtype, which appears particularly abundant in the immune system. We investigated the expression of CB2 receptors in leukocytes using anti-CB2 receptor immunopurified polyclonal antibodies. We showed that peripheral blood and tonsillar B cells were the leukocyte subsets expressing the highest amount of CB2 receptor proteins. Dual-color confocal microscopy performed on tonsillar tissues showed a marked expression of CB2 receptors in mantle zones of secondary follicles, whereas germinal centers (GC) were weakly stained, suggesting a modulation of this receptor during the differentiation stages from virgin B lymphocytes to memory B cells. Indeed, we showed a clear downregulation of CB2 receptor expression during B-cell differentiation both at transcript and protein levels. The lowest expression was observed in GC proliferating centroblasts. Furthermore, we investigated the effect of the cannabinoid agonist CP55,940 on the CD40-mediated proliferation of both virgin and GC B-cell subsets. We found that CP55,940 enhanced the proliferation of both subsets and that this enhancement was blocked by the CB2 receptor antagonist SR 144528 but not by the CB1 receptor antagonist SR 141716. Finally, we observed that CB2 receptors were dramatically upregulated in both B-cell subsets during the first 24 hours of CD40-mediated activation. These data strongly support an involvement of CB2 receptors during B-cell differentiation.

Abstract 2401 Background. Compelling evidence suggests that chronic infections by various pathogens are linked to lymphoma development. The transformation process is supposed to be either direct (eg for EBV) or indirect. The association between Helicobacter pylori chronic infection and gastric marginal zone lymphoma (MZL) is the best characterized example for an indirect transformation. Several other pathogens such as hepatitis C virus, Campylobacter jejuni or Streptococcus Pneumoniae (Spn) are also suspected to promote B-cell lymphoma development through repeated stimulations of the BCR and/or inflammation. However, except for gastric MZL, animal models are lacking to study potential correlations between chronic infections and lymphoma development. Methods. To amplify and precipitate lymphomagenesis by precluding repair of DNA lesions potentially generated during chronic immune response, p53-deficient mice were used as a permissive model. Given the suspected role of carbohydrates from encapsulated bacteria such as Spn in promoting chronic lymphocytic leukemia, p53−/− (n=15) and p53+/− (n=53) mice were chronically injected with heat-killed Spn until disease development. P53-deficient mice chronically injected with PBS were used as control. Results. Unexpectedly, chronic injections of Spn promoted T-cell rather than B-cell lymphomagenesis in both p53−/− and p53+/− mice and shortened survival in p53+/− mice (P=.004). Whereas mostly thymic CD4+CD8+ double positive T-cell lymphomas have been described in p53-deficient mice, a vast majority of lymphomas observed following chronic Spn injections were of peripheral origin (TdT−) and exhibited an effector memory phenotype (CD44hiCD62LloCCR7−CD25−). Clonality and transferability of those peripheral T-cell lymphomas (PTCL) were established. Furthermore, lymphoma cells showed features of chronically stimulated T cells such as TCR, CD3 or CD4/CD8 co-receptor down-regulations along with PD-1 up-regulation. Several lines of evidence suggested a contribution of the TCR to the development of these PTCL: 1/all PTCL following Spn injections exhibited a Vß repertoire usage bias (Vß8 in 100%) consistent with a transformation process originating from a chronically-stimulated T cell by a pathogen-specific immunodominant peptide; 2/cyclosporin A, a strong TCR signaling inhibitor, decreased cell survival in vitro, and prolonged mice survival following transfer of lymphoma cells into recipient mice; 3/engraftment of CD8+ PTCL in MHC class I KO mice was significantly reduced compared to wild type mice (P The absence of B-cell lymphoma development in p53−/− mice and its very late onset in p53+/− mice (ie >450 days) compared to T-cell lymphoma prompted us to dissect the potential role of p53 in mature T-cell response in a context of chronic stimulation. WT and p53−/− negatively selected CD4+ and CD8+ T cells were repeatedly (ie every 7–10 days) stimulated in vitro using anti-CD3/anti-CD28-coated beads. In agreement with previous reports, no significant difference of cell viability was observed after the first or the second stimulation confirming the minor role of p53 in initial activation and proliferation as well as in activation-induced cell death. Nonetheless, a dramatic increase in cell viability was observed 48h after the third stimulation of p53−/− T cells, indicating a crucial function of p53 in deletion of chronically activated T cells. Conclusion. Chronic stimulations with heat-killed Spn unexpectedly increased peripheral T-cell lymphoma development in p53-deficient mice. Phenotypic characterization was consistent with a transformation process occurring in a pathogen-specific chronically-stimulated T cell. The incidence of p53 mutations is higher in T-cell than in B-cell mature malignancies in humans and the p53 pathway is functionally impaired in virtually all enteropathy-associated T-cell lymphomas, which are supposed to be a key model for an antigen-driven process. Therefore, aside from its known role in immature T-cell lymphoma development and in progression of B-cell malignancies, our work sheds light on a previously unsuspected physiopathological role of the p53 pathway in peripheral T-cell lymphomagenesis. Disclosures: No relevant conflicts of interest to declare.

Multiple Myeloma (MM) is a plasma cell (PC) malignancy producing large amounts of monoclonal Igs requiring a highly developed endoplasmic reticulum (ER). A coordinated adaptative program called the unfolded protein response (UPR) allows cells to survive the stress consecutive to a transient protein overload of the ER. This process is referred to as "physiological UPR". If the stress is excessive or sustained, the UPR leads to cell death ("terminal UPR"). We have recently documented that the spontaneous death of non transformed human PC is initiated by an ER stress. Moreover the group of R. Sitia has established that the proteasomal activity decreases in the late phases of PC differentiation. These findings suggest that the imbalance between the high-rate of Igs secretion and the impaired proteolytic machinery predisposes PC to apoptosis. We have postulated that a dysregulated adaptative UPR contributes to the increased survival capacity of MM cells. We have undertaken a study aimed at promoting death of MM cells by disruption of ER homeostasis. Our approach has consisted in using RNA interference to knock-down the UPR sensors PERK, ATF6 and IRE1, in two MM cell lines (U 266 and NCI-H929). We show that the anti-UPR siRNAs induce a 70 to 80% reduction of the levels of expression of PERK, ATF6 and IRE1 and promote cell death of these two human MM cell lines. The PERK-targeted siRNA was consistently found to be the best apoptosis inducer. We found that invalidation of PERK leads to induction of a terminal UPR as revealed by the upregulation of GADD153/CHOP and of the spliced form of XBP-1. The death elicited by invalidation of the UPR sensors was accompanied by phosphatidylserine (PS) exposure and early loss of the mitochondrial transmembrane potential. By contrast, we found no evidence of DNA fragmentation and electronic microscopy revealed only a partial nuclear chromatin condensation. Cell death induced by UPR-targeted siRNAs operated in a caspase -independent fashion since it was not inhibited by the pan-caspase inhibitor z-VAD-fmk. Accordingly, the activated forms of caspase-3 and-4 were undetectable. However, calpain inhibitors and calcium chelators prevented the loss of the mitochondrial transmembrane potential and partially inhibited cell death. This suggests that the death signal initiated at the level of the ER is transmitted to the mitochondria via the calcium cysteine protease calpain. Transmission electron microscopy revealed that the PERK siRNA induces formation of numerous autophagic vacuoles, a result which was confirmed by the positive staining with monodansylcadaverine (MDC). Appearance of mature autophagic vacuoles preceeded acquisition of death morphology which was characterized by partial nuclear condensation and protein accumulation. The anti-UPR siRNAs induced the presence of immature and mature autophagic vacuoles but, in contrast, when acquiring death morphology, cells were devoid of autophagic vacuoles although they showed partial nuclear condensation. The autophagic specific inhibitor 3 methyladenine reduced both PS exposure and MDC staining induced by PERK siRNA. This suggests that UPR-targeted siRNAs induce death of MM cells by autophagy. Altogether, our data demonstrate that disruption of ER homeostasis by extinction of UPR sensors induce death of MM cells. They also reveal that a connection exists between ER stress and autophagy in MM cells.