Bergsagel, P. Leif

Abstract Multiple myeloma (MM) is the second most common hematologic cancer, accounting for roughly 2% of all cancer deaths. Unfortunately, MM remains incurable and nearly all patients experience relapse. Recently, we have identified a novel cIAP1 antagonist, LCL161, with remarkable anti-MM effects. Interestingly, unlike in other tumor types, our previous work found that LCL161 does not directly induce apoptosis of MM cells, but rather drives activation and infiltration of macrophages to the tumor bed and phagocytosis of tumor cells. Furthermore, in vitro experiments demonstrate that the anti-MM activity of the agent relies on the presence of the tumor, with no observed increase in MM phagocytosis when macrophages are treated in isolation. However, LCL161 is a strong and indiscriminate transcriptional activator and thus, the molecular mechanisms underlying this immune response in MM remain poorly understood. In an effort to characterize molecular signatures associated with this response, we collected whole spleens from seven WT mice and eleven mice with VK12598 transplanted myeloma for single cell RNA-seq. Of the eleven tumor-bearing mice, three remained untreated, three were treated with LCL161, three with bortezomib, and three with cyclophosphamide. Additionally, one of the WT mice was treated with LCL161. Using the 10X genomics chromium platform, we measured mRNA levels in over 90,000 individual cells. We identified groups of cells with shared expression profiles, corresponding to known cell types, using graph-based clustering. Consistent with our previous work, we observed a marked reduction in tumor cells and activation of the noncanonical NFkB pathway in mice treated with LCL161. Turning our attention to cells identified as monocytes, with a particular focus on putative macrophages, we sought to identify molecular signatures associated with LCL161 treatment. Indeed, using a binomial model we identified over 500 upregulated genes in LCL161-treated macrophages. However, our previous data suggested that the efficacy of LCL161 was dependent on the presence of tumor cells. Thus, we refined these results by comparing the expression profile of macrophages treated with LCL161 in WT and tumor-bearing mice. These contrasts identified dozens of genes that were significantly upregulated in macrophages from LCL161 treated, tumor-bearing mice. Of note, we identified a small number of genes (including Spp1, Irg1, and Lipg) that were active only in macrophages from mice treated with LCL161 and bearing a transplanted tumor and unexpressed in all other cells. We believe these genes—and the pathways in which they are active—hold promise for the development of novel targeted therapies to better treat MM. Citation Format: Nicholas Banovich, Martin Boateng, Marta Chesi, Leif Bergsagel, Jonathan Keats. Single-cell RNA sequencing identifies macrophage-specific expression signatures associated with phagocytosis of multiple myeloma after treatment with cIAP antagonist [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3135.

Abstract Multiple myeloma (MM) is characterized by clonal expansion of malignant plasma cells (PCs) and aberrant production of monoclonal immunoglobulin detected as an M spike using serum protein electrophoresis. In the United States, MM represents ~15% of hematologic malignancies and is one of the few cancers increasing in incidence (e.g., 14,400 in 1996 to 30,330 in 2016, from SEER). Previously, a Vk*MYC mouse model was described, in which AID-dependent activation of MYC transgene in germinal center (GC) B cells catalyzes a highly penetrant, indolent MM after a prolonged latency. However, the indolent MM developed in Vk*MYC mice rarely progresses to a malignant stage, suggesting that additional genetic mutations are required for the malignant MM progression. Recent sequencing of paired tumor/normal samples from advanced or refractory MM patients identified that constitutive activation of Ras signaling pathway (KRAS: 23%; NRAS:20%; BRAF: 8%) associates with MM progression and therapy resistance. To determine whether oncogenic Nras promotes the progression of Myc-induced indolent MM to a malignant stage, we generated NrasLSL Q61R/+; Vk*MYC; IgG1-Cre (VQ) mice along with single mutant mice Vk*MYC; IgG1-Cre (Vk) or NrasLSL Q61R/+; IgG1-Cre (Q61R). To boost NrasQ61R expression in GC B cells, 6-7 weeks old mice were immunized with NP-CGG. A significant fraction of VQ mice developed M-spike after immunization and subsequently died of a highly malignant MM. The MM developed in some mice were restricted in bone marrow and spleen and thus mimicked majority of MM patients, while MM developed in other mice was primarily located in lymph nodes and represented patients with extramedullary MM. Both types of MM cells displayed hyperproliferation and hyperactivation of AKT and ERK pathways. They could be cultured for several weeks with their original stromal cells and in the presence of Il-6. The cultured MM cells could be easily infected or transfected. Upon transplantation, both types of MM cells caused a rapid lethality (within 16 and 4 weeks respectively) in secondary and tertiary recipients and preserved the initial locations of donor MM cells. MM recipient mice developed bone lesions and/or fractures characteristic of MM patients. Preliminary preclinical studies indicated that both types of MM cells were resistant to Bortezomib, a proteasome inhibitor commonly used in treating MM patients, and AZD6244, a potent MEK inhibitor. Additional tests of combinatorial therapies are ongoing. Taken together, we generated a novel mouse model in which activation of Vk*MYC and oncogenic Nras in GC B cells results in a highly malignant, transplantable MM. This model can serve as a platform to investigate pathogenesis of highly malignant multiple myeloma and test the efficacy of novel therapeutic agents. Citation Format: Zhi Wen, Adhithi Rajagopalan, Erik A. Ranheim, Remington Finn, Adam Pagenkopf, Grant Yun, Yun Zhou, Yidan Wang, Demin Wang, Marta Chesi, P. Leif Bergsagel, Fotis Asimakopoulos, Jing Zhang. Expression of oncogenic Nras and a MYC transgene in germinal center B cells induces a highly malignant multiple myeloma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3018.

Abstract Multiple myeloma (MM) is a tumor of terminally differentiated plasma cells that underwent affinity maturation and class switch recombination processes in germinal centers before homing to the bone marrow (BM), where they secrete large amounts of monoclonal immunoglobulins, clinically detected as an M-spike by serum protein electrophoresis (SPEP). Clinical manifestations of MM also include renal impairment, anemia, and bone disease, the latter resulting from abnormal osteoclast activation, which highlights the complex crosstalk between MM cells and their BM microenvironment. Chromosomal translocations into the immunoglobulin loci are the primary oncogenic events in ~50% of MM cases and are thought to occur as a result of errors during somatic hypermutation or class switch recombination. The other half of MM is characterized by trisomies of odd-numbered chromosomes. Both of these genomic abnormalities are present in the premalignant phase of MM called MGUS, which progresses to MM at a rate of 1% per year. Progression events include mutations of genes in the RAS/MAPK pathways; chromosomal rearrangements of the MYC locus; and mutations, amplifications, or biallelic deletions of genes in the NFkB pathway. Among the latter, the most common are deletions of TRAF3 or cIAP1 and -2 (BIRC2 and -3) and amplification or translocations of NIK (Map3k14), all resulting in NIK stabilization, phosphorylation of IKKα, processing of p100 to p52, NFKB2 nuclear translocation, and activation of the noncanonical NFKB pathway. This can be induced in MM cell lines in vitro by treatment with SMAC mimetics or IAP antagonists (IAPa), which induce cIAP1 auto-ubiquitination and degradation. We postulated that the resulting constitutive activation of the noncanonical NFKB pathway would render MM cells independent on the BM-localized survival signals provided by BAFF and APRIL, leading to extramedullary dissemination. We tested this hypothesis in vivo by exposing Vk*MYC mice with BM-localized MM to the IAP antagonist LCL161 and monitoring disease progression. In the immunocompetent Vk*MYC transgenic mouse model MYC expression is sporadically induced in germinal center cells by AID-dependent somatic hypermutation, leading to progressive accumulation of monoclonal plasma cells restricted to the BM that can be monitored over time by SPEP and M-spike quantization, as is done clinically in MM patients. Through a very rigorous testing of drugs whose clinical activity in MM has been assessed in clinical trials, we have also validated Vk*MYC mice as a faithful model to predict drug activity in MM, with a positive predictive value of 73% and a negative predictive value of 92%. To our surprise, treatment of Vk*MYC mice with the IAPa LCL161 did not promote MM extramedullary dissemination as we predicted, but rather induced a significant reduction in M-spike levels, indicative of reduction of tumor burden. Based on these preclinical data, we conducted a phase 2 clinical trial of LCL161 in combination with cyclophosphamide in relapsed/refractory MM patients. Results from 25 patients enrolled in the study demonstrated a response in five patients (1CR, 1 VGPR, 2PR, 1MR) and a median PFS of 10 months. Consistent with in vitro data on MM cell lines, LCL161 treatment did not promote direct apoptosis of Vk*MYC cells but phagocytosis of live cells, detectable by intravital microscopy within four hours after treatment. While LCL161 treatment of BM-derived macrophages (MΦ) in vitro promoted the acquisition of an inflammatory phenotype, with increased expression of IL6, IL12, and CD86 that resembled activation with CD40 agonists, it did not increase their phagocytic ability. In contrast, exposure of MΦ to LCL161 treated tumor cells or to the conditioned medium collected after LCL161 treatment enhanced phagocytosis. We therefore performed gene expression analysis on MM cells collected from Vk*MYC mice and MM patients before and after LCL161 treatment and identified marked upregulation of NFKB target genes, as expected, but also a type-I interferon (IFN) signature. Interestingly, the presence of blocking antibody to the IFN receptor subunit 1 (aIFNAR1) during LCL161 treatment of tumor cells inhibited their phagocytosis in vitro, and pretreatment of Vk*MYC mice with aIFNAR antibody abrogated LCL161 activity in vivo. Pretreatment of Vk*MYC mice with liposomal clodronate, known to deplete phagocytic cells, also abrogated LCL161 activity, indicating that MΦ-dependent phagocytosis, stimulated by the LCL161 induced tumor cell autonomous IFN response, is required for the anti-MM effects of LCL161. LCL161 has been shown to promote T-cell function in vitro, but in our hand LCL161 retained full anti-MM activity in T-cell deficient mice. However, while ~15% of immunocompetent mice were cured by just two weeks of LCL161 treatment, none of the T-cell deficient mice were, indicating a role for T cells in preventing disease recurrence. Consistently, rechallenge of cured mice with specific MM tumors induced expansion of IFNγ-producing T cells, indicative of reactivation of immunologic memory. We were surprised by the hyporesponsiveness of T cells to LCL161 in vivo and attributed it to the high levels of expression of the coinhibitory molecule PD1 we noted on both CD4 and CD8 cells irrespective of treatment. We also detected high levels of PDL1 on most CD11b+ myeloid cells as well as on a fraction of MM cells, which increased following LCL161 treatment. We therefore sought to combine LCL161 with an antibody against PD1. Remarkably, the combination of LCL161 with aPD1 was curative in all the mice that completed the two-week treatment, and we are in the process of translating this discovery to the clinic by opening a new trial of LCL161 in combination with aPD1. Citation Format: Marta Chesi, David Fooksman, Peter Leif Bergsagel. Immunomodulatory therapy of multiple myeloma with IAP antagonists [abstract]. In: Proceedings of the Second AACR Conference on Hematologic Malignancies: Translating Discoveries to Novel Therapies; May 6-9, 2017; Boston, MA. Philadelphia (PA): AACR; Clin Cancer Res 2017;23(24_Suppl):Abstract nr IA21.

A role for antigen-driven stimulation has been proposed in the pathogenesis of monoclonal gammopathy of undetermined significance (MGUS) and multiple myeloma (MM) based largely on the binding properties of monoclonal Ig. However, insights into antigen binding to clonal B cell receptors and in vivo responsiveness of the malignant clone to antigen-mediated stimulation are needed to understand the role of antigenic stimulation in tumor growth. Lysolipid-reactive clonal Ig were detected in Gaucher disease (GD) and some sporadic gammopathies. Here, we show that recombinant Ig (rIg) cloned from sort-purified single tumor cells from lipid-reactive sporadic and GD-associated gammopathy specifically bound lysolipids. Liposome sedimentation and binding assays confirmed specific interaction of lipid-reactive monoclonal Ig with lysolipids. The clonal nature of lysolipid-binding Ig was validated by protein sequencing. Gene expression profiling and cytogenetic analyses from 2 patient cohorts showed enrichment of nonhyperdiploid tumors in lipid-reactive patients. In vivo antigen-mediated stimulation led to an increase in clonal Ig and plasma cells (PCs) in GD gammopathy and also reactivated previously suppressed antigenically related nonclonal PCs. These data support a model wherein antigenic stimulation mediates an initial polyclonal phase, followed by evolution of monoclonal tumors enriched in nonhyperdiploid genomes, responsive to underlying antigen. Targeting underlying antigens may therefore prevent clinical MM.

Abstract Inhibition of the AAA ATPase, p97, was recently shown to be a novel method for targeting the ubiquitin proteasome system, and CB-5083, a first-in-class inhibitor of p97, has demonstrated broad antitumor activity in a range of both hematologic and solid tumor models. Here, we show that CB-5083 has robust activity against multiple myeloma cell lines and a number of in vivo multiple myeloma models. Treatment with CB-5083 is associated with accumulation of ubiquitinated proteins, induction of the unfolded protein response, and apoptosis. CB-5083 decreases viability in multiple myeloma cell lines and patient-derived multiple myeloma cells, including those with background proteasome inhibitor (PI) resistance. CB-5083 has a unique mechanism of action that combines well with PIs, which is likely owing to the p97-dependent retro-translocation of the transcription factor, Nrf1, which transcribes proteasome subunit genes following exposure to a PI. In vivo studies using clinically relevant multiple myeloma models demonstrate that single-agent CB-5083 inhibits tumor growth and combines well with multiple myeloma standard-of-care agents. Our preclinical data demonstrate the efficacy of CB-5083 in several multiple myeloma disease models and provide the rationale for clinical evaluation as monotherapy and in combination in multiple myeloma. Mol Cancer Ther; 16(11); 2375–86. ©2017 AACR.

Key Points PVD is an active combination in relapsed lenalidomide-refractory MM patients. PVD with weekly bortezomib offers a simpler, more convenient, and well-tolerated regimen option.