ALBERT

Description: Key Points Knockdown of the sialyl-transferase, ST3GAL6, in MM inhibits in vivo homing and prolongs survival in xenograft mice. In MM patients, high expression of ST3GAL6 is associated with inferior overall survival.

Description: Background. LIN28B regulates developmental processes and cellular reprogramming by suppressing let-7 microRNAs (miRNAs). A role for LIN28B has been reported in cancers, however the LIN28B/let-7 axis has not been studied in multiple myeloma (MM). Methods. LIN28B level expression in MM patients was studied using previously published gene expression profiling (GEP) datasets. Knockdown (KD) of LIN28B was performed on MM cell lines (U266, MOLP-8) using 2 shRNA and validated using 2 sgRNA by CRISPR knockout (KO). Downstream regulations were assessed by qRT-PCR and western blots, as well as RNA sequencing. For RNA sequencing, control and Lin28B CRISPR cells were used for library preparation (NEBNext kit) and sequencing on a HiSeq 2000. Proliferation of KD and KO cells were evaluated in vitro and in vivo in a xenograft mouse model. An LNA-GapmeR technology was used to develop a let-7 mimic in vivo in SCID mice. Findings. Two independent GEP datasets (GSE16558; GSE2658) were analyzed for LIN28B expression, showing a significantly higher level in MM patients compared to healthy controls. In addition, high LIN28B levels correlated with a shorter overall survival (p = 0.0226), along with an enrichment of let-7 target genes by Gene Set Enrichment Analyses (GSEA). LIN28B KD cells had a significantly increased expression level of let-7 family members and were associated with down-regulation of let-7 target genes Myc and Ras at the protein level. We further confirmed downstream regulation of MYC and RAS in a LIN28B CRISPR KO model in MM cells (MOPL-8). We next validated the role of LIN28B in MM in vivo by using a xenograft tumor model showing a decreased tumor burden in LIN28B KD mice compared to scramble control (p =0.0045). In addition, we performed a RNA sequencing from the CRISPR LIN28B KO and control cells and observed a central role by GSEA for both MYC and E2F cell cycle pathways in LIN28B-engineered cells. LIN28B activity in regulating MYC and cell proliferation was further defined to be dependent on let-7 by performing a rescue experiment in MM1S cells. Moreover, we explored the possibility to therapeutically regulate MYC expression through let-7 with an LNA-GapmeR containing a let-7b mimic, in vivo, and showed that high levels of let-7 expression represses tumor growth in SCID mice by regulating MYC expression compared to control GapmeR treated mice (p = 0.0026). Conclusions. These findings reveal the essential role of LIN28B/let-7 in regulating two essential oncogenic pathways in MM, MYC and RAS. Interference with this pathway may represent an efficient option for targeting MYC in cancer. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 1837 Introduction: Current research has demonstrated that cancer cells are not only influenced by their microenvironment, but are also able to drastically alter their surroundings to further accelerate cancer progression. Hematological malignancies create a forward feedback system with local mesenchymal stromal cells within the bone marrow, but the exact mechanisms and cellular changes within the stroma are largely unknown. Our work has aimed at understanding the bidirectional interaction between multiple myeloma (MM) cells and human bone marrow-derived mesenchymal stromal cells (hMSCs) by characterizing hMSCs derived from either MM patients or healthy individuals, using medium-throughput assays and 2D and 3D in vitro bone marrow models. Methods: First, primary human myeloma patient MSCs (MM-MSCs) and normal donor MSCs (ND-MSCs) were characterized in terms of proliferation when cultured with and without MM.1S cancer cells in direct and indirect co-culture. Next, primary ND- and MM-MSCs were profiled for their microRNA (miRNA) (n=3 for normal, n=7 for MM) and mRNA (n=5 for normal, n=5 for MM) expression using Nanostring technologies. We analyzed 800 human miRNAs from miRBase v.18 and 230 human cancer-related genes using the nCounter® Human Cancer Reference Kit, which allowed for much greater specificity and reliability than microarray technologies. Lastly, a more representative culture system was designed to better mimic myeloma growth within the bone marrow. We developed a 3D in vitro model using hMSCS and fluorescent, luciferase-labeled MM cell lines seeded into porous, autofluorescent silk scaffolds. Scaffolds were made using silk fibroin protein isolated from silkworm cocoons and were formed into biocompatible cylinders with pores of 500–600 microns in diameter. Both ND- and MM-MSCs were labeled with fluorescent cell-tracker dyes and cultured on scaffolds with or without MM1S-GFP+/Luc+ cells. These were non-destructively assessed using confocal microscopy and bioluminescent imaging (BLI) for their ability to promote cancer cell growth and protect cancer cells from chemotherapeutics. Comparison with a conventional 2D in vitro model was performed. Scaffolds were also assessed for their ability to support in vitro culture of primary MM cells using confocal microscopy. Results: MM-MSCs differed from ND-MSCs at miRNA, mRNA, and functional levels. MM-MSCs proliferated slower than ND-MSCs and direct, but not indirect, co-culture of ND-MSCs with MM.1S was able to recapitulate this slowed proliferation in vitro. Indeed, we found that 22 microRNAs were significantly dis-regulated (14 up-regulated and 8 down-regulated in MM patients versus healthy individuals). These included up-regulation of miRNA-222, -181, -146, and -382; together with down-regulation of miRNA-15a, -143, and -199a, in MM-derived hMSCs, compared to ND-MSCs (p

Description: Multiple myeloma (MM) is a malignancy of plasma cells and is characterized by unrestricted tumor cell growth in bone marrow (BM). MM causes destructive osteolytic lesions causing bone fracture, bone pain, hypercalcaemia, and nerve-compression, resulting from increased bone resorption and suppressed bone formation. Despite the introduction of agents to inhibit bone resorption, such as bisphosphonates, which prevent further bone loss, approaches to preventing osteoblast suppression and repair bone lesions are limited and there are no agents available clinically. The wnt/β-catenin pathway plays a critical role in the regulation of bone formation. Production of the soluble wnt antagonist dickkopf1 (Dkk1) by MM cells has been implicated in MM inhibition of bone formation. As such, Anti-Dkk1 treatment prevents bone disease in pre-clinical models of MM and is in early clinical development. However, Dkk1 is not expressed by all myeloma cells; hence only a proportion of patients may respond to anti-Dkk1 therapy. Sclerostin (Scl) is a soluble wnt antagonist whose expression, unlike Dkk1, is restricted to osteocytes; therefore Scl targeted agents may have less off target effects. Anti-Sclerostin (Anti-Scl) treatment increased bone formation and bone volume in experimental models of osteoporosis, and increased bone mineral density in phase II osteoporosis clinical trials. However, Anti-Scl treatment effects on myeloma bone disease are unknown. Further, cells of the BM such as osteoblasts have been implicated in the regulation of MM cell survival and growth. Thus, in the present study we explored the potential for Anti-Scl therapy to prevent MM induced bone loss and inhibit MM growth in both murine and human xenograft MM models. Female C57BLKalwRij mice (n=8) were injected i.v, with 5TGM1/eGFP murine MM cells (1×106) and female SCID/beige mice (n=10) were injected i.v. with MM1S/Luc/eGFP human MM cells (4 × 106). 24 hours later, naïve mice (without tumor cells) or mice bearing MM cells were treated with anti-sclerostin antibody (Anti-Scl) (100mg/kg i.v) or control antibody. Mice were sacrificed at day 21 (MM1S) or day 28 (5TGM1) and the effect of Anti-Scl on bone structure in the femora and vertebrae were determined by microCT analysis. The effect of Anti-Scl on MM burden was determined by bioluminescent imaging (BLI) performed twice weekly from week 1 using a Xenogen IVIS system, whereas MM burden in 5TGM1/eGFP bearing mice was examined by FACS analysis. Anti-Scl treatment in naïve C57BLKalwRij mice increased trabecular bone volume fraction (BV/TV, 39%, p

Description: Background Extramedullary disease (EMD) in patients with multiple myeloma occurs mostly in advanced disease or relapse. EMD seems to have a different pathogenesis from medullary myeloma and is often characterized by a more aggressive clinical course. To date molecular mechanisms of development of EMD have not been fully understood. Methods: Human MM cell lines, IM-9 and MM1S were serially selected in immune-deficient mice. IM-9 and MM1S cells were inoculated intravenously and harvested from the tumors developed in the bone marrow (BM) and liver (site of extramedullary disease), aiming to establish BM-prone and liver-prone clones. Tumor progression was periodically checked by bioluminescence (BLI) and in vivo live confocal imaging. After three rounds of in vivo selections, the cells of both BM- and liver-prone were characterized by gene and protein expression and cellular functional assays. Results: We obtained three liver-prone sub-clones in both IM-9 and MM1S after serial in vivo selections. These cells had equal proliferation rates in vitro compared to the original or BM-prone cells, but exhibited more aggressive phenotype in vivo. Liver-prone clones had significantly higher migration ability than BM-prone clones (11.6% vs 6.1% migration, respectively p=0.018). Gene and protein expression analysis revealed that each liver-prone clone had a higher expression of sets of chemokine receptors specifically CXCR4. Using thein vivo metastatic model CXCR4-over expressing myeloma cells exhibited higher metastatic property to the extramedullary organs whereas CXCR4-knockdown cells has less tumor metastasis to the liver. Discussion: We established EMD-prone human multiple myeloma cell lines that reproducibly developed liverinvolvement consistent with human extramedullary disease. These cells exhibited higher migratory ability and increased expression of several chemokine receptors, specifically CXCR4. We validated the effect of CXCR4 on developing extramedullary myeloma using our established in vivo mode. Further studies to determine the role of CXCR4 for therapeutic targeting of extramedullary disease in MM are ongoing. Disclosures: Ghobrial: Sanofi: Research Funding; Noxxon: Research Funding; BMS: Advisory board, Advisory board Other, Research Funding; Onyx: Advisoryboard Other.

Description: Background. Endothelial progenitor cells (EPCs) are circulating precursors with the capacity to differentiate into endothelial lineage cells through a process known as “vasculogenesis” thus contributing to vessel formation. We studied the role of endothelial progenitor cells in multiple myeloma (MM) pathogenesis. Methods. EPC levels were evaluated in peripheral blood (PB) of patients with smoldering MM (SMM), and active myeloma (MM) and in PB of healthy controls, by using flow-cytometry (CD34+VEGFR2+ cells), and by performing endothelial colony forming assays. EPC levels were studied in PB from transgenic Vk*MYC mice harboring either early MM (smoldering-like stage); or late (active MM-like stage); as well as in mice injected intravenously with either murine MM 5TGM1-turboRFP+ cells or human MM1S-GFP+/luc+ cells. Healthy syngeneic mice were used as controls. GFP-bone marrow (BM) transplantation and sub-cutaneous femur implantation were performed in recipient mice to study EPC BM mobilization (GFP+CD34+VEGFR2 cells) and EPC BM recruitment to the implanted femurs during 5TGM1 MM model progression. Transgenic ID1+/-ID3-/- mice (with a defect in EPC mobilization and differentiation ability) were injected with transplantable Vk*MYC cells and followed for survival. Wild type littermates were used as controls. Similar experiments were performed using transgenic ID1+/-ID3-/- mice transplanted with BM of wild type littermates. Therapeutic activity of DC101 anti-murine VEGFR2 Ab was evaluated in the MM1S human orthotopic xenograft model, using both early and late treatment approaches. Results. EPC levels were significantly increased in PB of MM patients, including SMM, compared to healthy controls (P

Description: Background The extracellular matrix (ECM) is a major component of the tumor microenvironment, contributing to the regulation of cell survival, proliferation, differentiation and metastasis. In multiple myeloma (MM), interactions between MM cells and the bone marrow (BM) microenvironment, of which the ECM forms a major component, are critical to the pathogenesis of the disease and the development of drug resistance. To date, despite some knowledge of the composition of the ECM in tumors, detailed profiling of the composition of the ECM in MM has not been carried out. Until recently ECM proteins have proven difficult to characterize due to their biochemical properties and large size. Recent advances in proteomics have led to the characterization of the ECM and ECM-associated proteins (“matrisome”) in normal human tissues and tumors using a systematic and comprehensive approach. Methods Tumor Xenograft models; MM1S-GFP-Luc+ cells (5x106) were injected intravenously into SCID-Bg mice (n=4/group) and animals underwent weekly bioluminescent imaging (BLI). Mice were sacrificed after 2 weeks in order to mimic early tumor development (luminescence = 1x105 p/sec/cm2/sr) or 5 weeks (1x108 p/sec/cm2/sr) to model more advanced MM. Human bone marrow aspirates; Whole bone marrow was obtained from newly diagnosed MM patients (n=9) and healthy human donors (ND) (n=4) following written informed consent. ECM proteins were enriched from bone marrow samples obtained from MM patients, NDs and mice according to previously published methods.Tandem Mass Spectrometry (LC-MS/MS): Peptides were run using reversed-phase microcapillary liquid chromatography – tandem mass spectrometry (LC-MS/MS) on a high resolution hybrid Orbitrap Elite mass spectrometer. MS/MS data were searched against the UniProt Human database using MASCOT to identify proteins. Spectral counts were used as a semi-quantitative measure of abundance. ECM proteins were defined according to the in-silico definition of the matrisome. Validation of expression of ECM mRNA in MM cell lines (MM1s, RPMI-8226 and U266) and in CD138+ cells and bone marrow stromal cells (BMSC’s) from MM patients in comparison to NDs was performed using qRT-PCR. Results Primary myeloma sample ECM; Using a spectral count of 2 as a cutoff of peptide abundance we identified a total of 536 unique proteins in ND bone marrow of which 35 are defined as matrisome proteins. 982 unique proteins were enriched from whole bone marrow samples of newly diagnosed MM patients of which 26 are defined as matrisome proteins, 7 unique proteins were identified as ECM or ECM-associated in newly diagnosed patients which were not detected in the ND samples including PRG3, FGG, LEG10, TLN1 and PLEC. Critical ECM components such as laminins, matrix metalloproteinases and collagens were also found to be significantly altered in newly diagnosed MM with evidence of destruction of ECM components in active disease. Tumor Xenograft ECM; In mice with an earlier phase of human MM1s cell tumor burden we detected a total of 329 unique proteins of which 48 were defined as matrisome proteins, 23 of these proteins were unique to the earlier phase of MM in these mice. Mice with more advanced tumor development had unique ECM proteins which were not detected in the earlier disease stage including collagens, laminins and matrix metalloproteinases, indicating that these ECM components may be critical for re-modelling the ECM in MM. Interestingly, in our xenograft model of MM we were able to detect both human and mouse ECM components indicating that the tumor ECM is secreted from both the murine stroma and the human MM cells and allowing delineation of the source of individual ECM components. This indicates that as MM progresses certain ECM components, including FBN1, which were initially derived from stroma are later derived from MM cells. Differential expression of ECM components, including FBN1 between normal and malignant plasma cells was confirmed using qRT-PCR. Conclusions We have performed proteomic profiling of the unique tumor ECM in MM using mass spectrometry with a view to determining the specific components that may be altered with disease progression. Through this approach plasma-cell-derived ECM can be identified with a view to developing therapeutic strategies in this disease. Disclosures Glavey: BMS: Consultancy, Research Funding. Palumbo:Bristol-Myers Squibb: Consultancy, Honoraria; Genmab A/S: Consultancy, Honoraria; Celgene: Consultancy, Honoraria; Janssen-Cilag: Consultancy, Honoraria; Millennium Pharmaceuticals, Inc.: Consultancy, Honoraria; Onyx Pharmaceuticals: Consultancy, Honoraria; Array BioPharma: Honoraria; Amgen: Consultancy, Honoraria; Sanofi: Honoraria. Ghobrial:Onyx: Advisory board Other; BMS: Advisory board, Advisory board Other, Research Funding; Noxxon: Research Funding; Sanofi: Research Funding.

Description: Introduction.Exosomes are small vesicles (50-100 nm) of endocytic origin, which are released in the extra-cellular milieu by several cell types. Exosomes play a role in tumor progression since they have been shown to carry and transfer microRNAs (miRNAs) to recipient cells. In this study, we sought to characterize circulating exosomes in Multiple Myeloma (MM) patients and to assess the prognostic value of circulating exosomal miRNAs in a cohort of 112 newly diagnosed MM patients. Methods. Exosomes were isolated from peripheral blood (serum samples) using an ultracentrifugation protocol and Exoquick® solution. They were characterized using electron microscopy with immunogold labeling for the detection of CD63 and CD81, as well as for size using Nanosight® analysis. MiRNAs were isolated using miRNeasy micro kit (Qiagen®) and profiled using nCounter miRNA expression assay (Nanostring® Technologies) in 5 healthy donors (HD) and 10 MM patients. We next studied a cohort of 112 newly diagnosed MM patients uniformly treated with Bortezomib-Dexamethasone (4 cycles) followed by high dose Melphalan and autotransplant on the IFM 2005-01 Phase III Trial. After exosome isolation and miRNA extraction, we performed a low density qRT-PCR miRNA array using Taqman Array Micro Fluidic Cards for a panel of 23 specific miRNAs. Data were normalized by robust median global normalization and conditional inference tree was used to select the cut-point for miRNA levels. Kaplan-Meier estimation, Log-Rank test and Cox regression were used to assess progression-free survival (PFS). Results. Circulating exosomes were first characterized from 5 HD and 10 MM patients. CD63 and CD81 expression was demonstrated by immunogold labeling and electron microscopy. The size (approx. 100nm) and concentration of peripheral blood exosomes did not differ between HD and MM patients using Nanosight® analysis. We next performed a miRNA array (Nanostring) and identified differentially expressed miRNAs in circulating exosomes from MM patients compared to HD. Notably, the miR-17/92, -106b/25, -106a/363 clusters were more highly expressed and the cluster miR-15a/16 and let-7 family members were down regulated in circulating exosomes from MM patients as compared to HD. We then designed a panel of 23 specific miRNAs from circulating exosomes in MM and performed a Taqman Low Density Array on a cohort of 112 patients who were newly diagnosed and therapy-naive in the IFM 2005-01 Phase III Trial or treated as such. The mean patient’s age was 61 years, 36% of patients had a 13q deletion, 7% a t(4;14) and 4% a 17p deletion. ISS stage I, II and III distribution was 37%, 39% and 21% respectively. The median follow up was 5.4 years. Among tumor suppressor miRNAs, we found that patients with lower let-7e levels before treatment had significantly shorter PFS compared to patients with higher levels of let-7e (P

Description: Background. Proline-rich tyrosine kinase (Pyk2) is a non-receptor tyrosine kinase which belongs to the focal adhesion kinase (FAK) family. It is known to facilitate the TNFα induced EMT process in solid tumors, but this has not been investigated in the field of hematologic malignancies. We therefore dissected the role of PyK2 in multiple myeloma (MM) by looking at its ability to modulate MM biology both in vitro and in vivo. Methods. Lentiviral packaged small hairpin RNA (shRNA), overexpression plasmid, related scrambled shRNA probe and empty vector were introduced into the MM1.S (GFP+/Luc+) cell line, to generate stable Pyk2 knock down (K.D.; #A2 and #A4), Pyk2 over-expression (Pyk2+), and control cells, respectively. The efficiency of K.D. or over-expression was validated by qPCR and immunoblotting. Cell viability and cell proliferation were detected by using CellTiter-Glo® luminescent assay and thymidine uptake, respectively. Gain- and loss-of function studies were also performed on MM cells in the presence of primary bone marrow stromal cells isolated from MM patients (MM-BMSCs). Adhesion of Pyk2 stable cells to fibronectin was measured by using an ECM cell adhesion assay kit. The synergistic effects of Pyk2 with Bortezomib were determined through calculating the DNA synthesis of Pyk2 K.D. cells treated with Bortezomib (2.5-5 µM), using Calcusyn software and Chou-Talalay method. Pyk2 K.D. stable cells were intravenously injected into SCID-Biege mice to generate a xenograft model. In vivo tumor growth was monitored by Bioluminescent Imaging. Pyk2-dependent-modulation of the Wnt/β-catenin pathway signaling was assessed using immunoblotting. The FAK/Pyk2 kinase inhibitor, VS-4718, was tested in vitro using both MM cell lines and primary bone marrow-derived MM cells; and in vivo using bioluminescence imaging. Results. Knockdown of Pyk2 in MM cells significantly repressed cell viability and proliferation, as well as their adhesive ability to BMSCs, compared to scrambled shRNA control cells. Moreover, Pyk2 knockdown induced de-adhesion of MM cells from BMSCs thus inducing chemosensitivity of tumor cells to Bortezomib. We next corroborated our findings by studying Pyk2+ MM cells, and showed that stably upregulated Pyk2 expression promoted MM cell growth as measured by either ATP quantitation or DNA synthesis. Upregulation of Pyk2 expression also stabilized the adhesion of MM cells to BMSCs, leading to drug resistance of MM cells to Bortezomib, compared with vector control cells. Pyk2-regulated tumor growth was further validated by establishing a xenograft mouse model. By using bioluminescence imaging, we found a significantly lower tumor burden in mice injected with Pyk2 K.D. cells, compared to control mice (injected with scrambled shRNA cells). We next dissected the effect of Pyk2 in modulation of cellular signaling in MM cells by using immunoblotting, and demonstrated that Pyk2 played an important role in regulating β-catenin signaling. Indeed, knockdown of Pyk2 induced GSK3β phosphorylation, leading to increased β-catenin phosphorylation, thus resulting in β-catenin degradation and inhibited translocation to the nucleus. Importantly, Pyk2 K.D. cells presented with reduced expression of c-myc and cyclin D1 at the protein level. Conversely, Pyk2 overexpression enhanced β-catenin expression together with c-myc and cyclin D1 up-regulation, thus confirming the role of Pyk2 in modulating Wnt/β-catenin signaling activity in MM. We finally determined that VS-4718, a potent FAK/Pyk2 kinase inhibitor currently in a phase I clinical study, induced apoptosis in MM cell lines and primary MM cells, and inhibited MM tumor growth in vivo. Conclusion. These findings indicate that Pyk2 exhibits pro-oncogenic properties in MM through modulation of Wnt/β-catenin signaling. The findings that the FAK/Pyk2 kinase inhibitor VS-4718 exerts anti-MM activity in cellular and in vivo models support MM as a potential clinical direction for this agent. Disclosures Ring: Verastem: Employment. Tam:Verastem: Employment. Xu:Verastem: Employment. Pachter:Verastem Inc.: Employment, Equity Ownership. Ghobrial:Onyx: Advisory board Other; BMS: Advisory board, Advisory board Other, Research Funding; Noxxon: Research Funding; Sanofi: Research Funding.

Description: Introduction. Growing evidence suggests that immune cells that reside within the tumor microenvironment are dysregulated and functionally impaired, leading to defective anti-tumor immunity of the host. One of the major immunosuppressive mechanisms during tumor progression is expansion of regulatory immune cells. Here, we analyzed the immune cells within the bone marrow (BM) and the peripheral blood (PB) of 2 immunocompetent multiple myeloma (MM) mouse models. We next studied the role of regulatory T cells (Tregs) in MM pathogenesis. Materials and methods. To study the immune cell populations of the BM and PB, we used two immuncompetent mouse models and transplanted VK*MYC cells or 5TGM1 cells into C57BL/6 and C57BL/Kalwrij mice respectively. The immune cell populations and checkpoint receptor expressions were analyzed by CyTOF mass cytometer or flow-cytometry. Treg induction assay was performed in vitro to study the mechanism of Treg increase in the BM of myeloma injected mice. CD4+ CD25- cells were obtained from C57BL/Kalwrij mice and were co-cultured with 5TGM1 cells or B cells from C57BL/Kalwrij mice in vitro. Treg induction was compared by flow-cytometry. Transplantable VK*MYC cells were injected into "depletion of regulatory T cell" (DEREG) mice, which expresses a simian diphtheria toxin (DT) receptor-enhanced GFP fusion protein under the control of the FOXP3 gene locus, or their wild type littermates. DT injection into these mice leads to depletion of Tregs as previously described (J Exp Med. 2007; 204: 57-63). DT was given once every week for a total 3 times i.p to the DEREG mice or the littermate controls to specifically deplete Tregs and to study the role of Tregs during MM progression. Tregs (CD4+ FOXP3-GFP+ cells) were sorted from VK*MYC injected mice or non-injected DEREG mice BM using FACSAria cell sorter. Cells isolated were subjected to RNA sequencing. Gene Set Enrichment Analysis (GSEA) was performed to define differences in molecular signatures between MM-associated and normal Tregs. Results. The Treg proportion was significantly increased within the CD4+ T cells in the BM of myeloma cell injected mice from the early stage of disease compared to control mice, while in the PB, the increase was observed only at the late stages of disease progression. The effector T cell (Teff)/Treg ratio was significantly decreased in the BM at the end-stage myeloma bearing mice (P