Citations

Abstract Malignant transformation is commonly associated with dysregulated ribosome biogenesis and for more than 100 years, pathologists have utilized the increase in size and number of nucleoli, the site of ribosome biogenesis, as a marker for aggressive malignancies (Hein et al., Trends in Molecular Medicine, 2013). Recently we demonstrated that hyperactivated RNA Polymerase I (Pol I) transcription, a rate limiting step in ribosome biogenesis, can be specifically targeted by the small molecule inhibitor, CX-5461 (Drygin et al., Cancer Research, 2011; Bywater et al. Cancer Cell, 2012). When evaluated for its anti-proliferative activity against a panel of genetically diverse human cancer cell lines, those derived from p53 wild type hematological malignancies were the most sensitive (Drygin et al., Cancer Research, 2011). Evaluation of whether Pol I transcription inhibition could serve as therapeutic strategy in vivo revealed that CX-5461 administration significantly prolongs the overall survival of Em-Myc lymphoma bearing mice. This survival advantage was associated with rapid activation of the ribosomal protein (Rp)-MDM2-p53 nucleolar stress response (Deisenroth et al., Oncogene, 2010) and subsequent induction of tumour cell-specific p53-dependent apoptosis; importantly the wild type B-cell population was maintained and did not activate p53 pathways (Bywater et al. Cancer Cell, 2012). To explore the therapeutic potential of Pol I transcription inhibition via CX-5461 in hematological malignancies refractory to standard chemotherapy, we employed mouse models of highly aggressive MLL-driven Acute Myeloid Leukemia (AML) (MLL/AF9 + Nras and MLL/ENL + Nras) (Zuber et al., Genes Dev, 2009) and V*κ-Myc-driven Multiple Myeloma (MM) (Chesi et al. Cancer Cell, 2009). CX-5461 administration significantly increased the overall survival time of AML-bearing mice (MLL/ENL Nras median 17 days for vehicle vs 36 days for drug, P<0.0001 (Figure 1); MLL/AF9 Nras median 15 days for vehicle vs 23 days for drug, P 0.0009) without severe adverse effects. The extended survival of MLL-driven AML bearing mice was associated with acute activation of p53, induction of cell death and an aberrant G2/M cell cycle progression, which consequently delayed tumour progression as monitored by bioluminescence-imaging and white blood cell counts (WBC). In contrast to CX-5461, standard-therapy with a regime containing Cytarabine/Doxorubicin (5:3) failed to induce p53 in MLL/ENL Nras AML-bearing mice, resulting in a significantly lower survival advantage as compared to animals treated with CX-5461 therapy. Figure 1Figure 1. In ongoing experiments with MM-bearing mice, chronic CX-5461 administration robustly reduced the secretion of serum monoclonal Ig, as detected by serum protein electrophoresis (SPEP), and significantly prolonged the overall survival time (Figure 2) (V*κ-Myc median >160 days for drug vs, 103.5 days for vehicle, ongoing) demonstrating that CX-5461 exhibits potent anti-tumour activity in MM model. Analysis of the molecular mechanism responsible for the therapeutic benefit of CX-5461 in MM is ongoing but appears to involve both p53 mediated apoptosis and aberrant cell cycle progression. Figure 2Figure 2. These data demonstrate that hyperactivated Pol I transcription can be successfully targeted through small molecule inhibitors to treat models of human AML and MM that are refractory to standard therapy. Based on these and previously published results (Bywater et al., Cancer Cell, 2012), we have recently initiated a first-in-human, phase I, dose escalation clinical trial of this first-in-class drug, CX-5461 in patients with advanced hematological malignancies. The toxicity, response, pharmacokinetic and co-relative data from this trial will offer valuable insights as to whether inhibition of ribosome synthesis might offer a new non-genotoxic therapeutic approach in the fight against cancers, which are highly refractory to standard therapies. Disclosures: O'Brien: Senhwa Biosciences, Inc: Employment.

Abstract Abstract 4082 N-methyl-2-pyrrolidone (NMP) is considered an inert drug delivery vehicle that is widely used as an industrial organic solvent. NMP is found in a range of pharmaceutical preparations and bio-prosthetic cements. In particular, NMP is used to solubilize kinase inhibitors for pre-clinical testing in experimental mice. The Vk*MYC transgenic mouse is a model of multiple myeloma (MM) that recapitulates the human disease. We have adapted Vk*MYC MM using a syngeneic transplant approach to shorten latency and facilitate testing of novel therapeutics. In this model, we noted delayed disease progression and improved survival in NMP treated mice compared to alternative vehicle controls. To confirm this observation, matched cohorts of mice bearing monoclonal kappa-secreting MM were treated with 20% NMP/80% polyethylene glycol (PEG) (n = 13), PEG alone (n = 10) or PBS (n = 7) at 100uL per 20g mouse daily p.o. Mice were monitored by serial serum protein electrophoresis and quantitative kappa immunoglobulin (IgK) estimation. At the commencement of therapy, cohorts were well matched for disease burden; mean M-spikes 12%, 14% and 14% of total serum protein for NMP, PBS and PEG respectively (p=0.58 NMP vs. PBS, p=0.68 NMP vs. PEG). After 28d treatment, NMP-treated mice had lower paraproteinemia compared to PBS and PEG controls (mean M-spike NMP 31%, PBS 42% and PEG 40%; p < 0.05 NMP vs. PBS or PEG), with concordant suppression of IgK (NMP 18g/L vs. PBS 39g/L, p<0.05). Survival in NMP-treated mice was also significantly longer than that in PBS-treated controls (median survival 50d and 40d respectively; p<0.05 log rank). Survival of PEG-only treated mice did not differ from PBS controls (median survival 39.5d; p=0.49). We compared the structure of NMP to known MM therapeutics, revealing that NMP is a subunit of thalidomide and its analogues. We next tested NMP for cytotoxicity and immune-modulatory drug (IMiD) activity in vitro. NMP (1uM – 10mM) was not directly cytotoxic to healthy donor peripheral blood mononuclear cells (PBMCs) or the U266 myeloma cell line. However, treatment of PBMC with 10uM NMP significantly enhanced NK cell cytotoxicity against the NK cell-sensitive target cell line K562 from 26% to 42% (p < 0.05). Further, NMP favored a Th-1 cytokine profile without directly altering the frequency of CD4, CD8, NK or regulatory T cell subsets in vitro. As compared to the IMiD lenalidomide, NMP inhibited TNF-a production from LPS stimulated monocytes. Given the IMiD-like effects of NMP, we assessed its ability to alter the expression of NK cell ligands on myeloma targets. NMP upregulated both NKG2D and DNAM-1 ligands (MIC-A, MIC-B, ULBP-1, ULBP-2 and CD155) on U266 cells, thereby enhancing the NK cell mediated lysis of the U266 cell line. Our results demonstrate the in vivo anti-myeloma activity and IMiD-like properties of NMP, indicating that it is a putative active thalidomide subunit. The detection of IMiD-like activity was facilitated by a sensitive, immune competent MM model. NMP should no longer be considered inert for the purposes of drug delivery. These findings mandate re-evaluation of drug efficacy data where NMP has been used as the drug delivery vehicle, particularly in immune-competent cancer models. Our findings with NMP may allow the development of new IMiDs. Finally, occupational or iatrogenic exposure to NMP should be further assessed as a potential risk for immunologically mediated adverse effects. Disclosures: No relevant conflicts of interest to declare.

Abstract The success of proteasome inhibition in multiple myeloma highlights the critical role for the ubiquitin-proteasome system (UPS) in this disease. However, there has been little progress in finding more specific targets within the UPS involved in myeloma pathogenesis. In an unbiased screen of de-ubiquitinase activity, we previously found the ubiquitin hydrolase UCH-L1 to be frequently over-expressed in B-cell malignancies, including myeloma. Using a transgenic mouse model, we also found that aberrant expression of UCH-L1 is sufficient to drive the development of spontaneous B-cell lymphomas and plasmacytomas, demonstrating its oncogenic activity. Whether and how UCH-L1 affects the clinical behavior of myeloma, however, is not known. Here we show that UCH-L1 is a potent negative prognostic factor that is essential for the dissemination and progression of myeloma. We found high levels of UCHL1 to predict early progression in newly diagnosed patients; a finding reversed in studies including bortezomib. We also found high UCHL1 levels to be a critical factor in the superiority of bortezomib over high-dose dexamethasone in relapsed patients. High UCHL1 partially overlaps with, but is distinct from, known genetic risks including 4p16 rearrangement and 1q21 amplification. Using an orthotopic mouse model, we found UCH-L1 depletion delays myeloma dissemination and causes regression of established disease. We conclude that UCH-L1 is a biomarker of aggressive myeloma that may be an important marker of bortezomib response, and may itself be an effective target in disseminated disease. Disclosures Galardy: Mission Therapeutics: Research Funding.

Abstract Abstract 3015 Several preclinical models are available to assess the efficacy of novel anti-myeloma therapies, each of them with specific utilities. The most commonly utilized are: 1) Human myeloma cell lines (HMCLs) offer the advantage of being easily manipulated and well genetically characterized and therefore are best used to provide in-vitro target validation and to demonstrate specific target inhibition by a drug. On the other hand, they are highly proliferative and do not recapitulate the complexity of the human disease in an endogenous micro-environment. 2) Xenograft studies, in which HMCLs are injected into immunodeficient mice, serve the purpose of demonstrating that in-vivo target inhibition can be achieved under physiological conditions, but again fail to represent the indolence of human MM and its complex interaction with the BM stromal cells. Drug screenings performed on both these models tend inevitably to overestimate the antimyeloma activity of compounds that simply inhibit proliferation, but are ineffective in targeting the bulk of MM tumor. 3) Direct cytotoxic studies on primary patients cells, alone or in co-culture with stroma cell lines, are useful in assessing a pro-apoptotic activity in a way that is completely independent on proliferation. However, once again this in-vitro model does not capture the complete biology of human MM. The immuno-competent Vk*MYC mouse model of myeloma has already demonstrated high biological fidelity to the human disease, making it an ideal model to study the behavior of myeloma cells in the context of a native microenvironment and immune system. Furthermore, this model offers the advantage of studying both indolent, BM localized, untreated MM (primary MM), and, with the use of transplants into syngeneic hosts, advanced, more proliferative and refractory disease (secondary MM). Using the primary MM model, we have assessed the anti-myeloma activity of 30 known and novel compounds, many of which currently in clinical trials, and found that agents effective in the treatment of patients are active in this model, whereas agents ineffective in the treatment of patients are not. From this study, the most promising class of novel agents in clinical trials are the HDAC inhibitors vorinostat and panobinostat. Based on this promising single agent activity, we sought to model the effects of combination therapy in the secondary myeloma models that we specifically generated to be bortezomib refractory or multidrug resistance. Although transplanted mice treated with full dose single agent bortezomib, vorinostat or panobinostat showed no response and died within three weeks post transplant, recipient mice treated with bortezomib in combination with either vorinostat or panobinostat achieved complete response and are still alive 10 weeks post-transplant. We conclude that the bortezomib + HDAC inhibitors is an active combination that overcomes in vivo bortezomib resistance. Furthermore, the Vk*MYC model provides an excellent platform for the development of novel combinations aimed to the treatment of refractory MM disease. Disclosures: Bergsagel: Amgen: Membership on an entity's Board of Directors or advisory committees; Celgene: Membership on an entity's Board of Directors or advisory committees; Genentech: Membership on an entity's Board of Directors or advisory committees; Millennium: Speakers Bureau; Novartis: Speakers Bureau.

Abstract LCL161 is a IAP antagonist targeting the cellular inhibitor of apoptosis proteins cIAP1 and -2 (cIAP1/2). We previously reported that loss of the ubiquitin ligase cIAP1/2 results in the stabilization of NIK and activation of the non-canonical NFkB pathway, a pathway we found frequently activated by a promiscuous array of mutations in multiple myeloma (MM), including the homozygous deletion of cIAP1/2 themselves. This suggests that LCL161 treatment will have no direct anti-tumor activity, and may even increase the already high level of NFkB in MM. However, this pathway is also a key regulator of the host immune response and LCL161 has been shown to induce some activation of the immune system that can potentiate the effects of anti-tumor vaccines. Like IMiDs, which target another ubiquitin ligase complex containing cereblon, we found that LCL161 inhibits TNFa secretion by LPS-stimulated monocytes in vitro. Also like IMiDs, LCL161 has little direct anti-tumor activity against primary MM cells and MM cell lines in vitro, or immunodeficient xenograft models in vivo, which, contrary to preclinical studies in other tumor types, is not potentiated by the addition of TNFa. At concentrations ten times higher than the ones required for target modulation (cIAP1/2 degradation, achieved at nanomolar doses), direct cytotoxicity was observed, as reported by other investigators, most likely derived from off target inhibition of the caspase inhibitor XIAP. LCL161 has been studied in a phase I study in solid tumors where it was well tolerated with a cytokine release syndrome as the dose limiting toxicity and is now being studied in a phase II study in combination with paclitaxel. Surprisingly we found that LCL161 has dramatic activity in vivo against MM that develops spontaneously in an immunocompetent genetically engineered mouse model of MM (Vk*MYC). This model has been rigorously validated and found to have a positive predictive value for clinical activity of 67%, and a negative predictive value for clinical inactivity of 88%. Among nearly 50 drugs tested as single agents, only LCL161, proteasome inhibitors and alkylating agents induce complete response in this model. We found that LCL161 has no activity in vitro against these same tumor cells, suggesting an important role for the host in mediating the anti-tumor effect. MM that develops in congenic recipient mice transplanted with unfractionated bone marrow from Vk*MYC donor mice with MM, but not from those transplanted with purified CD138+ MM cells, is sensitive to LCL161. Cell fractionation and “add back” transplantation experiments in Vk*MYC mice have identified plasmacytoid dendritic cells (pDCs) as the key mediators of LCL161 anti-MM activity, which is completely independent of adaptive immunity. pDCs infiltrate the BM of MM patients, and similarly of Vk*MYC mice, where they stimulate tumor growth through release of inflammatory cytokines and NFkB activation. It has been shown that engagement of TLR with CPG abrogates this effect. MM is a tumor exquisitely dependent on pro-inflammatory cytokines for survival, but also sensitive to the cytotoxic effects of type-I IFN. We found that the dramatic anti-MM activity of LCL161 is directly mediated by pDCs. In normal condition pDCs support MM growth by providing a cytokine rich environment. LCL161, however, turns them into killers by suppressing the MYD88 pathway responsible for IL6 and TNFa production while simultaneously activating IFNa production. Although the response rate to single agent IFNa in MM patients is approximately 20%, its use in the treatment of MM has been discontinued because of toxicity. LCL161 may represent a novel way of administering high level of IFNa directly to the tumor bed, limiting toxicity and increasing efficacy. Interestingly, IMiDs have been recently shown to augment IFNb production downstream of MYD88 in lymphoma, and their anti-lymphoma activity can be inhibited by antibodies to IFNb. We conclude that LCL161 is a novel immunomodulator with very potent anti-MM activity in vivo and well-defined molecular and cellular targets. A phase II clinical trial of LCL161 in MM is ongoing. Disclosures: No relevant conflicts of interest to declare.

Abstract Dexamethasone is a powerful anti-inflammatory, immunosuppressive agent widely used in the treatment of MM with pleiotropic effects on both the tumor, and host immune cells. We report on a completely novel approach to immune modulation using an agent with almost directly opposite effects to glucocorticoids on innate immune cells. LCL161 is an IAP antagonist targeting the cellular inhibitor of apoptosis proteins cIAP1 and -2 (cIAP1/2) that in a phase I study in solid tumors was well tolerated with a cytokine release syndrome as the dose limiting toxicity. We previously reported that loss of cIAP1/2 in MM cells results in the stabilization of NIK and activation of the non-canonical NFkB pathway, a pathway we found frequently activated by a promiscuous array of mutations in multiple myeloma (MM). As expected, we found that LCL161 has no direct anti-tumor activity against MM cells in vitro and ex vivo. However, the non-canonical NFkB pathway is also a key activator of the innate immune response. Remarkably, we found that in vivo LCL161 has dramatic, apoptosis-independent activity associated with marked phagocytosisagainst MM that develops spontaneously in an immunocompetent genetically engineered mouse model of MM (Vk*MYC). Cyclophosphamide (Cy), which has been shown to induce an acute secretory activating phenotype in tumor-associated macrophages, markedly potentiated the anti-tumor effects of LCL161. We show that phagocytes (depleted by liposomal clodronate), but not adaptive immune cells, are required for the anti-tumor effect. Consistently we show that LCL161 activates dendritic cells and macrophages in vitro, and in vivo, and upregulates production of inflammatory cytokines. A phase 2 clinical trial of LCL161 was conducted in 24 patients with relapsed MM previously exposed to both IMiDs and PIs, but less than 5 prior lines of therapy. Patients received LCL161 1200mg po weekly, with Cy 500mg po weekly added for progressive disease or lack of response. Single agent LCL161 was well-tolerated, with grade 3 or 4 non-hematologic toxicity in four patients, and following the addition of cyclophosphamide in 9 patients. No patients responded to single agent LCL161, and the best response was stable disease in 6/24. In 21 patients, Cy was added after a median of 2 cycles, in whom the best response was stable disease or better in 18/21 (p<0.001 compared to LCL161 alone), and a PR or better (including 1 VGPR and 1 CR) in 5/21 (p<0.001 compared to LCL161 alone). Six months following the initiation of Cy, 35% of patients remain event-free. Two patients had stable disease for more than nine months after stopping treatment. One, relapsing after VDT-PACE, HDM200, RVd x 3 years with t(11;14) and del17p, received only three months of LCL161, the last month in combination with Cy. He then achieved a PR, reconstitution of his uninvolved IgG to normal levels for the first time in more than 3 years and still has not progressed one year after stopping therapy. Similar changes in serum and bone marrow inflammatory cytokines and gene expression were observed in patients following treatment with LCL161 to those observed in the mice. We conclude that LCL161 is able to modulate the tumor microenvironment to exert long-term disease control in combination with Cy, and deserves further study in combination with agents that more directly depend on phagocytic cell activity (e.g., monoclonal antibodies, inhibitors of "Don't eat me" signals). Disclosures Bergsagel: Novartis: Research Funding; Constellation Pharmaceuticals: Research Funding; Incyte: Consultancy; Jannsen: Consultancy; Mundipharma: Consultancy. Off Label Use: Use of LCL161 to treat multiple myeloma. Kumar:Celgene: Research Funding; Millenium/Takeda: Research Funding; Onyx: Research Funding; AbbVie: Research Funding; Janssen: Research Funding; Sanofi: Research Funding; Celgene, Millenium, Sanofi, Skyline, BMS, Onyx, Noxxon,: Other: Consultant, no compensation,; Skyline, Noxxon: Honoraria. Stewart:Oncospire Genomics: Research Funding.

Abstract The time in cellular differentiation at which the expression of an oncogene is dysregulated is thought to determine the phenotype of the resulting neoplasia. We tested this assumption in C57BL6/J transgenic mice expressing an HA-tagged c-myc oncogene in B cells under the control of kappa light chain regulatory elements. Two transgenes were engineered that differed by only one nucleotide. The first set of mice carries a wild type HA-MYC transgene, develops pro-B cell tumors, and succumbs rapidly to pronounced lymphosplenomegaly. In the second set of mice, a point mutation in the HA-tag portion of the transgene creates a stop codon that abrogates HA-MYC translation. This stop codon was engineered to be a hotspot for somatic hypermutation. Thus sporadically, in a germinal center B cell, somatic hypermutation may revert the stop codon, allowing translation of HA-MYC. In two independently derived lines, mice spontaneously develop monoclonal gammopathies (50% incidence at 30 weeks (n=35), 80% at 40 weeks (n=35)). Serum protein electrophoresis detects M-spikes that increase in intensity over time, and 6 out of 7 cases tested so far were of IgG1 isotype. Unlike other mouse models of plasma cell neoplasia, no lymphosplenomegaly nor ascites were detected. In analogy to human multiple myeloma, that is a disease of isotype-switched and hypermutated cells homing to the bone, large populations of plasma cells were found in the bone marrow. Bone marrow (but not spleen) lysates reacted for HA and human MYC proteins by western blot and immunohistochemistry, thus indicating that reversion of the stop codon occurred. Remarkably, in two young mice, monoclonal spikes appeared 2 weeks after vaccination with NP-CGG. These spikes were sustained and the monoclonal protein was reactive to the NP antigen. Together, comparison of the two sets of mice demonstrates that activation of the same transgene at distinct B cell developmental stages results in dramatically different tumor phenotypes. In addition, and opposite to most engineered mice, we have created a system where the oncogene is turned on sporadically, as occurs in the human disease. Furthermore, we have developed a model that faithfully reproduces important clinical aspects of monoclonal gammopathy of undetermined significance (MGUS) and its progression to multiple myeloma. This model will be useful to develop immunologic and chemoterapeutic approaches.

Abstract Deletions of chromosome 13 occur in about 50% of multiple myeloma (MM) patients, with monosomy being the most common (85%), and small interstitial deletions, often involving the Rb gene, representing only 15% of cases. In contrast to CLL, where a minimal deleted region (MDR) has been mapped to the DLEU2/miR-15a/16-1 locus, no MDR has been identified in MM. The prognostic significance of monosomy 13 has been overshadowed by its strong association with high-risk MM, mostly t(4;14) and t(14;16) translocations. Genomic analysis of 28 MM tumors from Vk*MYC mice identified recurrent (27%) monosomy of the syntenic chromosome 14, with focal bi-allelic deletions of Rb1 in three independent tumors. One tumor presented a focal monoallelic interstitial deletion of the mirR15a-16.1 cluster. Reasoning that the biological significance of monosomy 13 would be conserved between human and murine MM, we carried out genetic crosses between Vk*MYC and mice heterozygous for Rb1 (MYC/Rb1), for the DLEU2/miR-15a/16-1 locus (MYC/MDR) or for the miR-15a/16-1 cluster only (MYC/MIR). All of the MYC/Rb1 mice died around one year of age of pituitary tumors without a significant plasma cell expansion compared to the Vk*MYC mice alone. In contrast, both the MYC/MDR+/- and MYC/MIR+/- mice exhibited a progressive monoclonal plasmacytosis affecting secondary lymphoid organs, with associated shorter overall survival. The development of extra-medullary MM was further accelerated in MYC/MDR -/- and MYC/MIR -/- mice, and was reminiscent of that observed in Vk*MYC crossed with Emu-BCL2 mice. As previously shown for the CLL-like disease that develops in the MIR and MDR mice, the MM phenotype is more pronounced in the MYC/MDR than the MYC/MIR, and in the homozygous than the heterozygous mice. Although BCL2 has been proposed as a target of the miR-15a/16-1, it has not been identified among genes whose expression is altered in MIR null B lymphocytes, where instead miR-15a/16-1 deletion promotes proliferation. In contrast, we found significant BCL2 upregulation in some MM tumors from MYC/MIR and MYC/MDR, suggesting that miR-15a/16-1 deletion in MM provides a survival signal that renders plasma cells less dependent on the BM microenvironment. We propose that selection for monosomy 13 in MM is driven in at least in part by loss of one copy of the miR cluster. Disclosures: No relevant conflicts of interest to declare.

Abstract The search for novel and clinically effective anti-myeloma therapies has been hampered by a paucity of good pre-clinical animal models. Most established animal models of multiple myeloma (MM) fail in one or more crucial features to resemble human MM, frequently exhibiting extramedullary tumors or lacking a competent immune system. In addition, competing models are either too expensive or time consuming to generate and maintain. Finally, no model has been shown to reliably predict both drug response and drug inactivity. We have previously described a C57Bl6/J transgenic mouse model in which the expression of the human c-myc oncogene is activated in post-germinal B cells by somatic hypermutation. These mice (Vk*myc) spontaneously develop monoclonal gammopathies and plasma cell expansion beginning at 20 weeks of age. However, in contrast to other models, their plasma cells cannot be found in secondary lymphoid tissues. Now, we report that this model closely reproduces the clinical behavior of human MM. Serum protein electrophoresis (SPEP) as well as ELISA for serum IgG demonstrate that the incidence and quantity of monoclonal paraproteinemia are greatly elevated in the Vk*myc mice as compared to age matched wild type C57Bl6/J controls (50 weeks mean IgG 1.92g/dL Vk*myc vs 0.2g/dL controls; peak IgG levels up to 7.5g/dL). As these mice age, their paraproteins continue rising and remain far higher than that of controls at every time point. Mice with significant paraproteinemia demonstrate marrow with up to 50% plasma cells with evidence of a low proliferative index, similar to what is observed in human MM and unlike what is seen in other models of MM. Anemia is observed, with a mean hemoglobin of 8.9g/dL in the Vk*myc mice vs 13.4g/dL in the wild type. Vk*myc mice analyzed show marked bone thinning with a 20% reduction in total bone volume and in the number of trabeculae per unit area as shown by microCT. In addition a 15% reduction in bone mineral density was demonstrated in affected mice. We next demonstrated the efficacy of 3 drugs, used commonly in clinical practice to treat myeloma (melphalan, dexamethasone and bortezomib). These were given as single courses of daily IP injections for 5 days, for the first two drugs, and bi-weekly IP for 4 weeks for the third. As early as two weeks post treatment, a statistically significant reduction in serum paraprotein levels was observed in the treated mice versus vehicle treated controls. The reductions were maximal at 21d post treatment in the melphalan (−77.2%±39 p<0.02) and dexamethasone (−67.4%±29 p<0.02) groups and at 28d in the bortezomib group (−68.9%±9 p<.001). No reduction was seen in the vehicle treated controls at any time point. As with human myeloma, the effect of one course of treatment was transient in nature. In contrast, treatment of VK*myc mice with 3 drugs that have not shown clinical activity against myeloma as single agents (hydroxyurea, vincristine and fludarabine) did not reduce serum monoclonal proteins. We conclude that our Vk*myc model closely reproduces the clinical characteristics of human MM. We have also shown that this model responds to drugs known to be clinically active, while showing no response to drugs that have little or no effect against human MM. To our knowledge, this is the first time that this level of fidelity to human MM has been demonstrated in a pre-clinical animal model. Our model therefore represents a unique and powerful tool for pre-clinical anti-myeloma drug development.

Abstract In an attempt to identify novel genetic abnormalities in multiple myeloma (MM) we analyzed 68 MM patient samples and 42 human myeloma cell lines (HMCL) with high-resolution (60kb) array-based comparative genomic hybridization (aCGH). We concentrated on bi-allelic genomic deletions with the expectation that these regions might contain novel tumor suppressor genes and identified a common abnormality at 14q32. Fine mapping of the 14q32 microdeletion by PCR demonstrated a 45 kb minimal deletion encompassing two genes; TRAF3, a negative regulator of the non-canonical NF-kB pathway, and AMN. To determine the frequency of copy number abnormalities of this region a cohort of 161 patients were screened with a FISH probe mapping to the minimally deleted region. This identified 7 (4.4%) patients with bi-allelic deletion and 19 (11.8%) patients with a single copy of the target region. Sequencing of all TRAF3 coding exons in 62 patients and 42 HMCL identified inactivating mutations in 9 patient samples and 4 HMCL. Overall, the incidence of TRAF3 inactivation, bi-allelic deletion or mutation, is 7/42 (16.7%) in HMCL and 12/62 (19.4%) in patients, making it one of the most common genetic abnormalities identified in MM. Importantly, dysregulation of NF-kB pathways is a common event in MM, but the mechanism for this dysregulation is largely unknown. Although TRAF3 is a negative regulator of the non-canonical NF-kB pathway it has not previously been implicated in cancer. Therefore, we tested the tumor suppressor function of TRAF3 by reintroducing TRAF3 into HMCL harboring TRAF3 abnormalities using an adenovirus gene transfer system. Expression of TRAF3 in HMCL with TRAF3 abnormalities rapidly reversed cellular processing of NFKB2 from p100 (an inactive isoform) to p52 (active) and additionally led to a 60% reduction in cellular proliferation (range, 10–89%) associated with Go/G1 cell cycle arrest and induction of cell death by apoptosis. At 48 hrs post infection the proportion of cells entering S-phase or G2/M declined by 31% (range, 8.1–56.1) and the proportion undergoing apoptosis increased by 28% (range 11.7–47.4%). In contrast, over-expression of TRAF3 in control HMCL without TRAF3 abnormalities had no effect. Overall, TRAF3 is identified as a novel tumor suppressor that regulates the non-canonical NF-kB pathway and is inactivated in ~ 20% of MM patients. Given the involvement of the NF-kB pathway in multiple human malignancies and the near constitutive expression of TRAF3 we propose that inactivation of TRAF3, a novel tumor suppressor identified for the first time in this study, may occur in other cancers. In this regard, we have recently identified a Burkitt lymphoma cell line that does not express TRAF3 and thus has constitutive processing of p100 to p52. The clinical and therapeutic consequences of this finding and its frequency in other tumor types are currently the focus of further investigation. In conclusion, the abnormalities identified in this study represent the first mutations identified that cause constitutive NF-kB activation in MM.