Shi, Chang‐Xin

Background Bortezomib (BTZ) is highly effective in the treatment of multiple myeloma; however, emergent drug resistance is common. The underlying mechanisms of such proteasome inhibitor resistance are still incompletely understood. Methods To further understand its resistant mechanism, we generated eight multiple myeloma (MM) cell lines resistant to bortezomib (BTZ) by exposure to increasing drug concentration: five of them acquired novel PSMB5 mutations. Given the rarity of similar mutations in over 1,500 analyzed MM patients, we explored in depth the role of the proteasome on MM cell viability and BTZ sensitivity by systematically deleting the major proteasome targets of BTZ by CRISPR. Results We demonstrated that MM cell lines without PSMB5 were surprisingly viable (mutation corresponding yeast gene pre2 is lethal). PSMB5 mutated, BTZ resistant, MM cell lines were re-sensitized to BTZ when PSMB5 was experimentally deleted, implying that this mutation is activating in its drug resistance function. In contrast PSMB6 knockout was lethal to MM cell lines, which were efficiently rescued by re-introduction of wild type PSMB6. Interestingly, reduction in PSMB6 levels also prevented the splicing of the major catalytic subunits PSMB5, PSMB7, PSMB8 and PSMB10. PSMB6 engineered with no splicing function or catalytic activity, also restored viability, inferring that the contribution of PSMB6 to proteasome structure is more important than functional activity. Supporting this observation, BTZ sensitivity was restored in resistant MM cells line by introducing low level expression of mutated PSMB6 lacking splicing function. As with PSMB6, PSMB7 knockout was lethal to MM cell lines. In contrast, loss of immunoproteasome subunits PSMB8 and PSMB9 was neither lethal nor restored sensitivity to BTZ. Our results demonstrate that expression of the three constitutive proteasome subunits PSMB5, PSMB6 and PSMB7 is highly co-dependent. This dependence is relying on the structure, but not the function, of PSMB5 and PSMB6. Conclusions In summary, PSMB5 and PSMB6, but not PSMB8 and PSMB9, are highly relevant for BTZ sensitivity in MM. Absence of PSMB6 or PSMB7, but not PSMB5, was lethal in MM cell lines. Expression of PSMB5, PSMB6 and PSMB7 was highly co-dependent. Together these findings suggest that the modulation of expression rather than function of PSMB5, PSMB6 or PSMB7 may be a new therapeutic strategy. Disclosures No relevant conflicts of interest to declare.

Abstract As multiple myeloma tumors universally dysregulate cyclin D genes we conducted high-throughput chemical library screens for compounds that inhibit signaling pathways driving cyclin D2 promoter transactivation, assaying more than 4,000 compounds. The top-ranked compound from these studies was a natural triterpenoid, pristimerin. Pristimerin markedly suppressed cyclin D2 promoter activity (>90%) in 3T3 fibroblast cells and inhibited cyclin D1, D2 and D3 protein expression in myeloma tumor cells. Strikingly, the early (4 hour) transcriptional response of myeloma cells treated with pristimerin closely resembles cellular responses elicited by proteosome inhibitors (P<10−9) (Connectivity Map Build 2, www.broad.mit.edu/cmap), with rapid induction of heat shock proteins (HSP70 >90-fold), activating transcription factor (ATF) 3 and CHOP. Enzymatic assays performed with purified 20S proteosome, or with total cellular extract, confirm that pristimerin rapidly and specifically inhibits chymotrypsin-like 20S proteosome activity at low concentration (<100nM), causing sustained inhibition lasting >6 hours. Consistent with inhibition of proteosome function, pristimerin causes rapid and sustained accumulation of high molecular weight poly-ubiquitinated protein in myeloma cell lines. Notably, related cytotoxic triterpenoid drugs, such as the methyl ester of 2-cyano-3,12-dioxooleana-1,9(11)-dien-28-oic acid (CDDO-Me, RTA 402) or betulinic acid or the ginsenosides - all of which show promising anti-cancer activities and are currently in clinical trials for advanced lymphoma, leukemia or solid malignancies - commonly inhibit NF-kB activation via direct inhibition of IKKα or IKKβ. In contrast drugs that function as proteosome inhibitors also commonly suppress NF-kB function instead by impairing degradation of ubiquitinated IkB. Immunoblotting for phosphorylated IkB confirms that pristimerin, like other triterpenoids, acts upstream of IkB to inhibit its phosphorylation, although pristimerin simultaneously inhibits proteosome activity with marked potency to diminish the clearance of ubiquitinated IkB. As a result of this two-fold stabilization of IkB, pristimerin causes overt and specific suppression of NF-kB mediated transcription, measured by a panel of transcriptional reporters with synthetic promoters containing 5x repeats of generic binding sites for NF-kB, AP-1, CREB or TCF4. Importantly, specific suppression of constitutive NF-kB transcriptional activity was pronounced in myeloma cells with inherent NF-kB pathway activation resulting from bi-allelic deletion of the TRAF3 tumor suppressor. Constitutive activation of the NF-kB pathway occurs in a significant proportion of primary myeloma tumors, most commonly via inactivation of TRAF3. Selective silencing of NF-kB driven transcription in myeloma cells may mediate the potent suppression of cyclin D proteins induced by this compound. Significantly, multiple myeloma cells are exquisitely sensitive to both proteosome inhibition or NFkB pathway inhibition. Consistent with these twin vulnerabilities, pristimerin is potently and selectively lethal to primary myeloma cells from patients (IC50<100nM) grown in mixed lineage culture and inhibits the growth of xenografted human plasmacytoma tumors grown in mice, providing a strong rationale for pharmaceutical development of triterpenoid dual-function proteosome-plus-NF-kB pathway inhibitors as therapeutics for multiple myeloma and related human malignancies.

Background The precise molecular mechanism and targets of lenalidomide specific anti-Multiple Myeloma (MM) activity remain largely unknown, although, cereblon (CRBN) has been identified as a central mediator of activity with IRF-4 acting as a downstream molecule. Lenalidomide resistance in MM cells which, despite CRBN depletion, are able to restore their IRF-4 levels suggests the existence of alternative pathways. Methods To measure apoptosis and gene expression analysis 1x 106cells were incubated with 10μm to 50μm of lenalidomide for 24 to 96 hours before they were examined by annexin-PI and FACS analysis. Gene and protein expression were measured by RT-PCR, western blot, and immunohistochemistry. For lenalidomide followed bortezomib serial studies, cells were incubated with 10μm of lenalidomide for 48 hours and later treated with bortezomib for 24 to 48 hrs. CRBN and immunoglobulin light chain knock-down were performed by lentiviral mediated shRNA. Results We found lenalidomide induces Endoplasmic reticulum stress (ERS) via accumulation of intracellular immunoglobulin proteins, but only in the presence of CRBN in a panel of MM cell lines. This accumulation of intracellular proteins resulted in ERS as documented by western blot and RT-PCR in OPM2 and MMIS cell lines. We next utilized OPM2 cells knocked down for CRBN and control, non-target (NT), and treated with different doses of lenalidomide. After three days of treatment an ERS response was occurring in OPM2-NT cells but not in CRBN knockdown cells. Using Western blot we found decreased XBP-1u and increased XBP-1s, ERS marker proteins, in CRBN positive (NT) cells after 72 hours of lenalidomide treatment, but not in CRBN knock-down cells. Another ERS marker GRP78/BiP was also clearly accumulated after lenalidomide treatment in CRBN positive cells. Inactivation of the tumor suppressor p53 by degradation is a mechanism utilized by cells to adapt to ERS has been reported. We also found that p53 protein down regulated after lenalidomide treatment in CRBN positive OPM2 and MMIS cells. We have analyzed other isogenic cell lines MM1.S (lenalidomide-sensitive) and isogenic MM1.Sres (lenalidomide-resistant) HMCLs and found that lenalidomide induces ERS in MM1.S but not MM1.Sres cells. CRBN is important for lenalidomide to induce immunoglobulin protein accumulation and ERS was further proved by over-expressing CRBN in OCI-My5 MM cell line and subsequent lenalidomide treatment. To test the effect of light chain accumulation as ER stressors, we knocked down immunoglobulin light chain lambda in human MM cell line OPM2. Knock-down of lambda light chain in OPM2 cells resulted in lenalidomide resistance and it induced endoplasmic reticulum stress mediated anti-myeloma activity. Our preliminary data reveals that lenalidomide mediated progressive ERS can positively enhance bortezomib induced apoptosis in an in-vitro MM model. We pre-treated myeloma cell lines with lenalidomide and subsequently treated with bortezomib. MM cells pre-treated with lenalidomide for two days and there after treated with bortezomib clearly shown increased sensitivity to bortezomib induced apoptosis than non-pre-lenalidomide treated cells. Conclusion Here we show that lenalidomide causes accumulation of immunoglobulin light chains, inducing progressive ER stress, resulting in anti-myeloma activity and that effect depends on the presence of CRBN. Proteasome inhibitors such as bortezomib also sensitize MM cells to endoplasmic reticulum stress-mediated apoptosis. Our data demonstrated the synergistic effect of lenalidomide and proteasome inhibitors, likely due to the augmented protein stress. This data support the use of serial therapy utilizing pre-lenalidomide treatment followed by bortezomib or carfilzomib treatment. Disclosures: Fonseca: Medtronic: Consultancy; Otsuka: Consultancy; Celgene: Consultancy; Genzyme: Consultancy; BMS: Consultancy; Lilly: Consultancy; Onyx: Consultancy, Research Funding; Binding Site: Consultancy; Millennium: Consultancy; AMGEN: Consultancy; Cylene: Research Funding; Prognostication of MM based on genetic categorization of the disease: Prognostication of MM based on genetic categorization of the disease, Prognostication of MM based on genetic categorization of the disease Patents & Royalties.

Abstract Abstract 3513 To identify targets for rational combination therapies with 5-Azacytidine (5-Aza) in myeloid malignancies, we utilized high-throughput RNA-interference (RNAi) viability screening. A siRNA library targeting 572 kinases and a custom collection of 289 putative cancer targets, including cell cycle and apoptosis regulatory genes, were screened alone and in combination with 5-Aza in TF-1 and ML-2 myeloid leukemia cell lines to identify synergistic interactions in reducing cell viability. Of the 572 kinases that were individually silenced, less than 1% significantly increased sensitivity to 5-Aza. The kinase library was also screened in combination with 5-Aza in a third myeloid cell line, THP-1, confirming that few kinases sensitize to 5-Aza when inhibited. While few kinases sensitized to 5-Aza, the anti-apoptotic Bcl-2 family of genes emerged as potent sensitizers to 5-Aza from RNAi screens. Therefore, silencing by siRNA of BCL-XL, BCL-2, BCL-W, MCL-1 and BFL-1 was evaluated in combination with 5-Aza treatment in an expanded panel of myeloid cell lines including TF-1, HEL, THP-1, ML-2 and MDS-L. BCL-XL validated as a vulnerability and potent sensitizer to 5-Aza in erythroid leukemia cell lines TF-1 and HEL, whereas MCL-1 was a strong vulnerability in the monocytic leukemia cell line THP-1 and also a moderate sensitizer to 5-Aza in ML-2, THP-1, TF-1 and HEL. Published proteomics data from our group indicate that M6 and M7 leukemias exhibit higher levels of BCL-XL, while additional unpublished data suggest elevated levels of MCL-1 in M4 and M5 leukemias, supporting our functional observations. Additionally, data from the public database Oncomine suggest that BCL-XL expression is elevated in M6 and M7 leukemias while MCL-1 shows a trend towards elevation in M4 and M5 leukemias. Based on RNAi screening results, siRNA validation experiments and proteomic/mRNA expression data, we evaluated the BCL-2/BCL-XL/BCL-W inhibitor ABT-737 in combination with 5-Aza. ABT-737 resulted in dose-dependent sensitization to 5-Aza in all AML-derived cell lines examined (including M7, M6, M5, M4 and M2 FAB subtypes) and in the MDS cell line MDS-L; however, no sensitization was observed in the CML cell line K562. In extensive ex vivo experiments with 17 primary specimens, potent synergy between 5-Aza and ABT-737 was observed in AML, MDS and MPN samples, but not in most CML samples examined. Calculations with CalcuSyn software demonstrate synergy, with combination index values as low as 0.2, between 5-Aza and ABT-737 both ex vivo and in vitro. The combination of 5-Aza with ABT-737 resulted in substantial induction of apoptosis, measured by the induction of cleaved caspase 3 in TF-1 and HL-60 cells, as compared to either compound alone. Interestingly, although siRNA silencing of MCL-1 in combination with 5-Aza was potent across several cell lines, and silencing of BCL-XL preferentially in an erythroid differentiation background, ABT-737 with 5-Aza sensitized across a variety of cell lines and all myeloid primary specimens ex vivo. We suggest that inhibition of anti-apoptotic Bcl-2 family members is a most promising rational combination strategy with 5-Aza for the treatment of leukemias. Our results also highlight the potential utility of more specific anti-apoptotic Bcl-2 family inhibitors in the lineage-specific treatment of myeloid malignancies. Disclosures: Off Label Use: AraC in AML. Experimental Agent MK1775. Mesa:Incyte: Research Funding; Lilly: Research Funding; SBio: Research Funding; Astra Zeneca: Research Funding; NS Pharma: Research Funding; Celgene: Research Funding.

Abstract Introduction: We employed a customized Multiple Myeloma (MM) specific Mutation-Panel (M3 P) to screen a homogenous cohort of refractory MM (rMM) patients in a timely and cost-effective manner. rMM was defined as progressive disease (PD) during therapy or within 60 days after discontinuation of therapy. The M3 P provides a practical alternative to whole genome/exome sequencing approaches and includes 88 genes selected for being either actionable targets, potentially related to drug response or part of known key pathways in MM biology. Patients and Methods: Tumor and germline samples of 50 rMM patients were screened using M3 P. Patients received a median of five lines of treatment at the time of tumor sampling (range 2-15 lines). All patients received at least one immunomodulatory agent (IMiD) and proteasome inhibitor (PI), and 88%, 76% and 66% were refractory to either an IMiD, PI or both. All patients had received lenalidomide and bortezomib, 48% and 34% were additionally exposed to pomalidomide and thalidomide and 18% to carfilzomib at any time prior to sampling. The majority of patients had a progressive and refractory disease immediately prior to sampling (82% and 78%), with 43% of them being IMiD-refractory and 46% being PI-refractory in the most recent line of therapy. Cytogenetic data at time of sampling were available from 78% of patients. Adverse cytogenetics were detectable in 82% of patients with gain 1q21 > 2 copies (62%), deletion 17p (33%), t(4;14) (13%) and t(14;16) (8%) being the most common. Results: An average sequencing depth of 750x was obtained across the samples. Our screening revealed an increased prevalence of mutations in the MAPK pathway (72%), including 34% of KRAS, 26% of NRAS, 18% of BRAF (with one patient carrying two BRAF mutations and 8% in actionable p.V600E). Additionally, two mutations (4%) were found in RASA2, a RAS inhibitor. Other potentially actionable mutations were seen in p.R132H IDH1 and p.R248C FGFR3 (2% each). Most prominently, the CRBN pathway was found mutated in 22% of cases, including CRBN (12%, one case with concomitant CUL4B mutations), CUL4B (4%), IRF4 (4%) and IKZF1 (2%). In CRBN we found four cases with mutations leading to a truncated protein (frameshift, nonsense and splicing) and two cases with missense mutations located within the thalidomide-binding domain, suggesting a causal effect on IMiD resistance. All CRBN mutated patients and 91% of the CRBN pathway mutated patients were at least once unresponsive to IMiD based treatment. In one patient with IRF4 mutation sensitivity to IMiDs remained unknown. A majority of patients were also resistant to PI treatment (76%), however mutations in proteasome subunits (PSMD1 and PSMB8) and XBP1 wererare (4%), but only a limited number of this gene family were represented in our panel. Other recurrently mutated genes were TP53 (26%), FAM46C (12%), ATM (10%) and TRAF3 (8%). Notably, DIS3 mutations were identified in only 6% of cases. As previously reported, TP53 mutations were associated with deletion 17p (8/13 cases, 62%). A correlation between the absolute number of mutations and previous lines of treatment was not observed (r=0.05, p=0.73). Conclusion: Targeted sequencing on a homogenous cohort of heavy pretreated rMM patients gives key insights in the landscape of the rMM genome, uncovering an increasing frequency of MAPK, TP53 and CRBN mutations. Importantly, this is the first study showing recurrent mutations in CRBN in patients unresponsive to IMiD treatment, supporting a potential association with resistance to IMiD based therapies. Disclosures Mai: Onyx: Other: Travel Grant; Janssen-Cilag: Other: Travel Grant; Celgene: Other: Travel Grant; Mundipharma: Other: Travel Grant. Merz:Janssen: Other: Travel grants; Celgene: Other: Travel grants. Hillengass:Takeda: Honoraria, Other: Travel support; Novartis: Membership on an entity's Board of Directors or advisory committees, Research Funding; Sanofi: Research Funding; Janssen-Cilag: Honoraria, Other: Travel support; Celgene: Honoraria, Other: Travel support. Goldschmidt:Celgene: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Janssen-Cilag: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Millennium: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Onyx: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Novartis: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Chugai: Research Funding, Speakers Bureau. Stewart:Oncospire Genomics: Research Funding.

Abstract Abstract 2801 Poster Board II-777 To provide a rationale basis for targeted drug development for multiple myeloma patients, we have conducted high-throughput genome-scale RNA interference (RNAi) synthetic lethality studies in human myeloma cell lines (HMCL) to generate a comprehensive map of critical genes and molecular vulnerabilities in human myeloma tumor cells. KMS11 human myeloma cells were screened with a 13,982-oligo library targeting the ‘druggable' genome (6,991 genes) using optimized conditions that resulted in >95% transfection efficiency. Each gene was screened with 2 or more distinct oligos, in duplicate, using a single-siRNA-per-well format, testing >34,000 wells. Replicate high throughput experiments yielded highly reproducible results (R2=0.82). Viability was measured at 96h by ATP-dependent luminescence. Universally lethal siRNA and non-targeted siRNA were employed as controls. The specificity (FDR) of lethal RNAi was evaluated by custom-developed statistical methods based on RNAi result concordancy. From screening, 5.8% of druggable genome siRNA caused statistically relevant reductions in HMCL viability (greater than three standard deviations from control samples treated with non-silencing siRNA, compared with an anticipated rate due to chance of only 0.135%). Of these, two hundred and nineteen genes, targeted by the most lethal siRNA, were forwarded to validation studies. Validation high-throughput RNAi studies, using 4 oligos per gene, were conducted in order to verify target gene vulnerability. Ultimately, seventy two genes were validated as highly critical for myeloma cell survival (each with multiple concordant siRNA hits plus high reproducibility of R2 =0.94). Among top-ranked lethal molecular vulnerabilities in KMS11 myeloma cells we recurrently identified the proteasome (8 PSM subunits were independently identified as highly vulnerable); and BCL2 family member MCL1; both known therapeutic targets, validating a functional genomics approach to target discovery. In addition, a number of novel, equally lethal, molecular vulnerabilities were identified in KMS11 including WEE1, kinetochore complex component KNTC2, the aurora kinases, polo-like kinase 1, a previously uncharacterized DNA methyl transferase, ribonucleotide reductase, ribosomal protein L38, leptin receptor overlapping transcript LEPROT and SLC25A23 (a carrier responsible for mitochondrial adenine nucleotide flux). Various ubiquitous cellular proteins involved in RNA or protein processing (e.g. SF3a, SNW1, SNRPA1, EIF3s8, amongst others) were also identified as non redundant and critical for cellular viability. Thirty nine top-ranked molecular vulnerabilities in KMS11 were evaluated for comparative vulnerability in a second myeloma cell line, JJN3, and in A549 and 293 epithelial cells. A majority were vulnerable in JJN3. While 25 genes proved equally lethal on silencing in A549 and 293 epithelial cells, several, notably including MCL1 and three proteasome subunits, appear differentially susceptible in myeloma cells versus epithelial cells. From this genome scale study, MCL1 ranks third as a target capable of inducing cell death in myeloma cells, highlighting the potent vulnerability of myeloma cell viability to alterations in this anti-apoptosis regulator. Moreover, from active comparative studies of molecular vulnerability in myeloma versus epithelial cell lines, MCL1 as an RNAi target demonstrates perhaps the greatest cytotoxic selectivity between myeloma and epithelial cell lines of all targets examined. Gene expression studies indicate that MCL1 is highly expressed in myeloma tumor cells but is absent or only weakly expressed in human primary somatic tissues. Overall, therefore, from an unsupervised genome-scale screening approach, our data support MCL1 as an optimal Achilles heel molecular target in multiple myeloma, potentially offering greater specificity and deeper cytotoxic effect than proteasome inhibition. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 1807 We have recently demonstrated that cereblon (CRBN) mediates the direct anti-myeloma activity of immunomodulatory drugs (IMiDs). However, the genes/pathways downstream of CRBN associated with anti-myeloma activity remain unclear. We, and others, identified interferon regulatory factor 4 (IRF4) as one of the downstream targets of CRBN-associated signaling. Both lenalidomide treatment and CRBN knockdown downregulate IRF4. IRF4 levels return to baseline in IMiD resistant cells surviving CRBN silencing. To determine whether IMiD-induced IRF4 downregulation is critical to anti-MM activity, we overexpressed IRF4 in two IMiD-sensitive human MM cell lines (HMCLs), KMS11 and MM1.S, followed by lenalidomide treatment. Lenalidomide-induced cytotoxicity was greatly impaired in both HMCLs overexpressing IRF4 compared with the control virus infected cells. Further analysis indicated that IRF4 over-expression does not completely prevent lenalidomide-induced growth arrest, but reduces cell death by 70% after lenalidomide treatment. Immunoblotting analysis of KMS11 cells indicated that IRF4 over-expression blocks lenalidomide-induced activation of caspase 8, reduces up-regulation of p21waf and increases CDK6 expression but does not significantly affect lenalidomide-induced MYC down-regulation. Although cereblon and IRF4 are broadly expressed in MM, baseline levels of expression are only weakly correlated (r=0.22) in primary MM patient gene expression analysis. Gene expression studies revealed statistical changes in 1,368 genes when comparing high versus low CRBN expression in primary myeloma samples. Interestingly genes associated with high CRBN expression included cyclin D2, SOCS3 and IL4 while genes associated with low cereblon expression included cyclin D1, FRZB and CD200. In order to understand how CRBN is connected with downstream anti-myeloma signaling, a structure-function study was performed to determine which CRBN domain is required for lenalidomide-induced IRF4 down-regulation and cytotoxicity. Lentiviral constructs expressing wild-type CRBN and a series of mutated CRBN were generated, including mutations at thalidomide binding site (Y384A/W386A), deletion of DDB1 binding region (ΔMid) and truncations at N-terminal and C-terminal. Lentiviruses from these constructs were used to infect IMiDs resistant HMCLs, OCI-MY5 and MM1.S res. Both of these cell lines have very low endogenous CRBN expression and they became sensitive to lenalidomide after introduction of wild-type CRBN. Conversely, introduction of CRBN with mutated thalidomide-binding site or with DDB1 binding region depletion failed to mediate lenalidomide toxicity and down-regulation of IRF4. OCI-MY5 cells expressing either N-terminal or C-terminal truncated CRBN showed substantial reduced responses (more than 50%) to lenalidomide compared with wild-type CRBN expressing cells. Deletion of only 20–30 amino acids at either ends of CRBN greatly impaired the protein function, suggesting that protein folding might be important for CRBN-mediated IMiD response. Our data indicate that IMiD induced myeloma cytotoxicity is largely mediated by modifying CRBN associated E3 ubiqutin ligase and subsequent IRF4 downregulation, suggesting the CRBN-IRF4 axis is a potential target for development of new anti-myeloma drugs. Disclosures: Schmidt: Karyopharm: Research Funding. Stewart:Millenium: Consultancy, Honoraria, Research Funding; Onyx: Consultancy; Celgene: Consultancy.

Abstract Previous work in our laboratory has identified the importance of the NFB signaling pathway for plasma cell (PC) survival and accumulation in Multiple Myeloma (MM). Although up to 43% of patient PCs appear to have activated NFkB by gene expression profiling (GEP), the remainder must rely on an alternate signaling pathway for their survival. We examined our MM GEP datasets and supervised the clustering of patients according to an index of NFkB activity. The majority of patients not activating NFkB were associated with a high IL-6/Stat3 signature, which included the protein phosphatase of regenerating liver-3 (PRL-3). Interestingly, analysis of the PRL-3 promoter revealed competing binding sites for both NFkB and STAT3 within the first intron, suggesting that these play a role in the transcriptional regulation of this gene. PRL-3 has been previously demonstrated to be absent in most normal tissues while its expression in MM cells is quite elevated. As normal PCs have low levels of PRL-3 expression, we determined the upper limit of normal to be 3SD above the mean of these cells. Using this cutoff we report that elevated PRL-3 expression is observed in plasma cells from 1/15 normal individuals, 4/22 MGUS patients and 47/126 patients with either smoldering or overt MM (6.7%, 18.2% and 45.6% respectively). Less than 10% of patients were seen to have both an elevated NFkB index and high PRL-3 expression. Patients with higher expression of PRL-3 were found to have statistically significant higher mean values of creatinine, C-reactive protein, lower albumin levels and a lower overall survival in our Mayo Clinic MM dataset. To examine the role PRL-3 in MM biology we first stimulated MM cell lines with IL-6 and observed PRL-3 up-regulation in two cell lines (INA6, OPM1) but not in JK6L (an NFkB activated cell line). Using a luciferase reporter we have determined that STAT3 induction of PRL3 expression requires the first intron of the PRL3 gene which contains the putative STAT3 binding site. Finally we show that PRL-3 inhibition by shRNA leads to decreased cell growth and loss of IL-6 responsiveness in vitro in at least one cell line. Based on these results, we propose that the majority of MM patients depend on either the NFkB or STAT3 signaling pathways for their survival and that these are to a large degree mutually exclusive. Our experimental results suggest that PRL-3 plays a role in IL-6/STAT3 signaling and that it is a promising target for drug development.

Abstract Background: The proteasome inhibitor Bortezomib (BTZ) is an efficient treatment option for both, newly diagnosed and relapsed/refractory multiple myeloma (MM). Despite the effectiveness, most patients eventually acquire drug resistance for reasons not fully understood Materials and methods: To better understand BTZ resistance mechanism we used a CRISPR library (GeCKO V2) targeting 19052 human genes, trying to identify genes responsible for BTZ resistance. CRISPR sgRNAs targeting the ERN1-XBP1 pathway were used as positive controls. We first infected RPMI 8226 myeloma cell line expressing Cas9 with the CRISPR library packaged into lenti-vectors and selected for resistance to BTZ. Surviving cells were subjected to next generation sequencing. Based on the initial screen results we constructed a second CRISPR sgRNA library including 31 genes, each gene targeted with four sgRNAs. After the second round screening and subsequent sequencing, we selected the top 20 genes for individual validation. Results: Proteasome regulatory gene PSMC6 was identified as the only reproducible gene conferring BTZ resistance. Interestingly, the same gene was independently found by a second group using a CRISPR approach (Sheffer M et al. ASH Abstract 273, 2014). Resistance was reproducible using a PSMC6 knockout by three individual CRISPR sgRNAs targeting exonic regions and one pair of SgRNA targeting intron region flanking exon for the deletion of exon one. PSMC6 knockout was verified by PCR and Sanger sequencing. Sensitivity to BTZ was rescued by over expression of PSMC6 cDNA in RPMI 8226 cells harboring a deletion of PSMC6 exon 1. MM cells lacking PSMC6 also developed resistance against Carfilzomib. Resistance was reproduced on a second MM cell line, KMS11. We did not see any significant difference of toxicity in PSMC6 deleted cells for other chemicals tested (tunicamycin, staurosporine, dexamethasone and melphalan). We demonstrated that the sensitivity of chymotrypsin-like activity of proteasome against BTZ was significantly reduced in cells lacking PSMC6. Consequently, MM cells without the PSMC6 gene were relatively resistant to apoptosis induced by BTZ, which was verified by Western blot for caspase 8 degradation and luminescent assay for caspase 3/7 activities. Clinically we could not correlate the PSMC6 expression level with the outcome of BTZ treatment in BTZ naive patients using publically available gene expression data. We initially used CRISPR sgRNAs targeting ER stress pathway (ERN1 and XBP1) as a positive control. However we could not derive any resistant cells from the experiment. We also could not identify any sgRNAs targeting the ERN1-XBP1 pathway from our whole exome screen and next generation sequencing. Since this contradicts published data, we decided to knockout ERN1 and XBP1 genes individually. Clones of cells with successful knockout of ERN1 in three cell lines (RPMI 8226, KMS11 and JJN3) and XBP1 in two cell lines (RPMI 8226 and KMS11) were tested for response to BTZ and Carfilzomib , however we did not find any drug response difference between the knockout and parent cells. We also found that the ERN1-XBP1 knockout cells did not show difference in response to ERN1 specific inhibitors (4u8C and STF-038010) and ER stress inducers (tunicamycin and thapsigargin) compared to parental cells. Conclusions: Human multiple myeloma cells lacking the PSMC6 gene develop significant resistance to apoptosis induced by BTZ. We have however not found a correlation of PSMC6 expression levels with outcome to BTZ treatment in BTZ naïve patients. We are therefore currently investigating the PSMC6 mutation rate in relapsed MM patients after proteasome treatment. In contrast to previous reports showing that progenitor MM cells lacing XBP-1 or ERN-1 invoked BTZ resistance, we were not able to demonstrate a change in sensitivity after full CRISPR knock out of either ERN1 or XBP1. It has long been believed that ERN1-XBP1 pathway plays an important role for MM treatment, leading to the development of ERN1 specific inhibitors. However, we demonstrated that the toxicity of two ERN1-specific inhibitors appears independent of the ERN1-XBP1 pathway. We also demonstrated that the toxicity of two important ER stress inducers, tunicamycin and thapsigargin, is independent of the ERN1-XBP1 pathway. Disclosures Stewart: Celgene: Consultancy; Oncospire Inc.: Equity Ownership; BMS: Membership on an entity's Board of Directors or advisory committees; Novartis: Consultancy.

Abstract Background: It is now well established that the immune modulator (IMiD) drugs thalidomide (Thal), lenalidomide (Len) and pomalidomide (Pom) bind a specific pocket in the protein, cereblon (CRBN) that is required for MM cytotoxicity after IMiD treatment. Recently, we reported 46 CRBN binding proteins which were altered by Len, two of which were the transcription factors Ikaros (IKZF1) and Aiolis (IKZF3), important regulators of B and T cell development. Others had previously reported the same finding and went further by demonstrating that following binding of IMiD drugs, CRBN accumulates and accentuates the ubiquitination and subsequent proteasomal degradation of these two binding proteins. The resulting degradation and depletion of IKZF proteins has the simultaneous effect of downregulating IRF4 and MYC leading to MM cell death while simultaneously enhancing transcription of interleukin 2 and activating the immune compartment. Methods: To further explore this phenomenon and its relationship to drug sensitivity we first created two stable MM cell lines (H929IKZF1Luc, 8226IKZF1Luc) expressing an IKZF1-luciferase (luc) fusion gene and tested for sensitivity to Len alone and in combination with dexamethasone (Dex), bortezomib (Bor) or the histone deacetylase inhibitor SAHA. Next, we constructed an adenoviral vector expressing IKZF1-luc and used this to transfect 16 MM cell lines, confirmed expression and tested sensitivity to Len by MTT (Figure 1). Results: We first examined the stable cell lines and measured luc expression after exposure to all three IMiD drugs. With all IMiD drugs luc dramatically decreased in sensitive H929 cells. IKZF1 depletion was both time and dose dependent. Using this surrogate marker cell line assay, luc is most efficiently degraded by Pom which is twice as potent as Len, but a thousand times more potent than Thal (Concentration of drug required to reduce luc expression by 50%, Pom: 4.9nM, Len: 10.2 nM, and Thal: 4,795 nM). We then tested all 16 cell lines with Len 2uM for 5 hours and show a strong correlation of IKZF-luc depletion with sensitivity. Figure 1: Luc levels 5 hours after Len treatment at 2uM. We divided the cell lines into three groups: 1) Len resistant: 8226, KMS12PE, My5, SKMM2, JJN3, FR4, H112; 2) Len intermediate response: EJM, OPM2, U266, KMS18; 3) Len sensitive: H929, MM1.S, MM1.R, KMS11, XG1; Figure 1:. Luc levels 5 hours after Len treatment at 2uM. We divided the cell lines into three groups: 1) Len resistant: 8226, KMS12PE, My5, SKMM2, JJN3, FR4, H112; 2) Len intermediate response: EJM, OPM2, U266, KMS18; 3) Len sensitive: H929, MM1.S, MM1.R, KMS11, XG1; We next performed western blots on 9 MM cell lines and demonstrate that IKZF1 levels were lowest in 3 of 5 resistant MM cell lines and normal in all 4 sensitive lines studied. The cell line (FR4) with the lowest levels of IKZF1 harbors a damaging mutation of IZKF1 as well as a translocation which upregulates IRF4, an IKZF target. Of the remaining two resistant lines one has low levels of Cereblon. In one line the mechanism of IMiD resistance could not be established. Finally, we were curious about the dichotomy of the IMiD biology which seemingly has an absolute requirement for a functional proteasome for successful IMiD cytotoxicity which is at variance with clinical reports of synergy between the IMiDs and the proteasome inhibitors (PIs) Bor and carfilzomib (Car). To explore this, we treated the 8226IKZF1Luc cell line with Len alone or in combination with increasing concentrations of Bor (6 nM) or Car (40nM) and found that both totally blocked IKZF1 fusion protein degradation promoted by Len, beginning as soon as 1 hour at 100nM Bor or after 3 hours at 40nM treatment. Conversely, in low concentration of Bor (<4nM) no IKZF1 degradation blocking was observed in the Luc assay and Bor remained lethal to the MM cells. In that light with lower levels of Bor (<3nM) even varying the sequencing of MM cell exposure to the two drug classes did not block cytotoxicity. Other MM active drug Dex or SAHA, blocking the aggresome, had no significant inhibition for Len-induced IKZF1 fusion protein degradation. Conclusion: These results validate the central role for the CRBN-IKZF-IRF4 axis proteins in IMiD response and show a strong correlation of IZKF1 expression and the ability to degrade IKZF1 with sensitivity. Our data provide interesting and perhaps disconcerting biologic data suggesting that the timing and dosing of clinical use of IMiDs and PI treatment in combination should be carefully evaluated to prevent antagonism of effect but suggest in vitro that a therapeutic window exists at which proteasome inhibitors can be active without blocking IMiD activity. Disclosures Stewart: Sanofi: Consultancy; Array BioPharma: Consultancy; Bristol Myers Squibb: Consultancy; Celgene: Consultancy; Novartis: Consultancy; Takeda Pharmaceuticals International Co.: Research Funding.