Abstract 3371: A DNA-binding priming agent protects cell-free DNA and improves the sensitivity of liquid biopsies

Liquid biopsies using cell-free DNA (cfDNA) enable non-invasive detection and characterization of disease. Advances in sequencing methods have significantly improved the performance of liquid biopsies. Yet, despite these advances, sensitivity remains a fundamental challenge. In oncology, circulating tumor DNA (ctDNA) screening tests only detect 20-40% of stage I tumors and tests for minimal residual disease have only 25-50% sensitivity after surgery. The major barrier to better sensitivity is the intrinsic low level of ctDNA in plasma. Physical absence of tumor DNA molecules in a blood draw from a patient with low disease burden will result in a negative test, no matter the sensitivity of the ex vivo detection platform. To overcome this barrier, here we report a first-in-class intravenous DNA-binding priming agent that is given 2 hours prior to a blood draw to recover more ctDNA, boosting the detection of tumor mutations in plasma by 19-fold and increasing sensitivity from 6% to 84%. Given the rapid clearance of cfDNA from circulation, we reasoned that a priming agent that could bind and protect cfDNA from clearance could increase the tumor DNA recovered from plasma. We selected monoclonal antibodies (mAbs) as the class of molecules to use as cfDNA protectors given their persistence in circulation and ease of engineering. We identify a mAb that binds double-stranded DNA (dsDNA) and find on electrophoretic mobility shift assays that it binds both free and histone-bound dsDNA, the constituent components of cfDNA. We then demonstrate that this mAb can delay the clearance of dsDNA from plasma in vivo through co-injection of the mAb with free- and histone-bound dsDNA in mice. We further identify interactions with Fc-gamma-receptors as a key mediator of early clearance of dsDNA bound to the priming mAb. To address this early clearance and limit potential immune interactions, we engineer the mAb to abrogate its Fc effector function. The engineered variant decreases clearance of injected dsDNA by over 150-fold at one hour post-injection compared to dsDNA alone. We next evaluate the effect of our priming mAb on cancer detection. We use a targeted panel against 1,822 mutations in the MC26 murine colon carcinoma cell line to detect tumor mutations in the plasma of tumor bearing mice. The priming mAb results in 19-fold higher recovery of tumor DNA molecules compared to a control mAb. This improved recovery leads to detection of 77% of targeted sites in plasma compared to only 15% in the control group. In sensitivity analyses, higher recovery of mutant molecules improves sensitivity for cancer detection from 6% to 84% at 0.001% tumor fraction. In summary, we demonstrate an approach to overcome a key barrier in liquid biopsies. We envision that similar to contrast agents in clinical imaging, priming agents could significantly boost the diagnostic sensitivity of liquid biopsies and enable further applications across biomedicine.

Citation Format: Shervin Tabrizi, Carmen Martin-Alonso, Kan Xiong, Timothy Blewett, Sainetra Sridhar, Zhenyi An, Sahil Patel, Sergio Rodriguez-Aponte, Christopher Naranjo, Douglas Shea, Todd Golub, Sangeeta N. Bhatia, Viktor A. Adalsteinsson, J. Christopher Love. A DNA-binding priming agent protects cell-free DNA and improves the sensitivity of liquid biopsies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3371.

Abstract PD11-06: PD11-06 Circulating tumor DNA association with residual cancer burden after neoadjuvant therapy in triple negative breast cancer in TBCRC 030

Background. Patients (pts) with early triple negative breast cancer (eTNBC) are at increased risk of breast cancer recurrence and death. Recent studies have focused on escalation of therapy, with current treatment standard of at least five drugs – and associated toxicities - for eTNBC. Though presence of residual disease after neoadjuvant therapy (NAT) as measured by residual cancer burden (RCB) helps guide addition of adjuvant treatment, more effective tools to tailor therapy are limited. Persistence of circulating tumor DNA (ctDNA) in the setting of residual disease is associated with high risk of distant recurrence. However, more sensitive minimal residual disease (MRD) assays are needed to potentially guide optimization of systemic therapy.

Methods. TBCRC 030 is a phase II randomized study of 12 weeks of NAT single agent cisplatin or paclitaxel for stage II-III TNBC, followed by surgery. The primary objective of the parent study was to correlate baseline biomarker for homologous recombination deficiency and RCB by study arm. From this group, responders (RCB 0/1) and non-responders (RCB 2/3) from both study arms who did not receive additional NAT prior to surgery were selected for analysis from the study cohort, matched on baseline nodal status and tumor size. As a post hoc study amendment, available pts were followed for event free survival (EFS). Plasma samples were collected prior to treatment initiation (W0), at three weeks (W3), and at twelve weeks, prior to surgery (W12). Whole genome sequencing (WGS) was performed on primary tumor tissue to identify somatic mutations and design for each pt a tumor-informed, ctDNA assay tracking up to 1000 mutations to detect MRD. Detection limit was computed for each tested sample as previously described. For each sample assayed, we report tumor fraction (TFx) when MRD was detected and the detection limit at 90% power when MRD was not detected.

Results. Of 139 study pts, 68 had complete tissue and plasma samples and no receipt of additional NAT. Of these, 22 were responders. These responders, and 22 matched non-responders were identified for analysis. Data from 22 pts – 11 responders, 11 non-responders - are described here; full analysis on all 44 pts will be presented at the meeting. Personalized ctDNA assays were designed targeting 434 to 1000 variants (median 1000) and applied to 66 plasma samples. At W0, 100% (22/22) were positive for ctDNA; 73% (16/22) and 55% (12/22) were positive at W3, and W12, respectively. In pts with T1-T2 tumors median TFx was 4.1e-3(7.8e-6, 3.4e-2) and 4.7e-1(4.3e-2, 9.0e-1) in pts with T3-T4 tumors. TFx decreased from W0 to W3 and from W0 to W12 in responders (Table 1). By W12, ctDNA had cleared in 7/8 pts with RCB 0, 1/3 with RCB 1, 2/8 with RCB 2, and 0/3 with RCB 3. Overall, ctDNA levels were broad with median TFx of 1.5e-3 (range 2.9e-6 to 0.90). Detection limit at 90% power for all tested samples was a median of 8.8e-6 (range 9.9e-7 to 6.8e-3).

To investigate whether ctDNA persistence after NAT was associated with BC recurrence, we analyzed a separate group of all 8 pts with known recurrence and with complete data and samples. All pts had persistent ctDNA at W12 (median TFx 6.8e-3, [2.9e-6 to 6.6e-2]).

Conclusions. After 3 weeks of NAT for eTNBC, ctDNA TFx decreased, with a 3900-fold change in responders and 18-fold change in non-responders. By W3, TFx for most pts with RCB 0/1 were below the 1 in 10,000 limit of detection for many currently available assays, emphasizing the need for sensitive tests to potentially guide therapy. Additional studies will determine if ctDNA-guided approaches in eTNBC can improve pt outcomes.

Table 1: Tumer Fraction and Tumer Fraction Fold Change by Response to Neoadjuvant Therapy

Citation Format: Heather A. Parsons, Timothy Blewett, Xiangying Chu, Sainetra Sridhar, Katheryn Santos, Kan Xiong, Vandana Abramson, Ashka Patel, Ju Cheng, Adam M. Brufsky, Justin Rhoades, Jeremy Force, Ruolin Liu, Tiffany A. Traina, Lisa Carey, Mothaffar Rimawi, Ahmed Elkhanany, Vered Stearns, Jennifer M. Specht, Harold Burstein, Antonio C. Wolff, Eric Winer, Nabihah Tayob, Ian Krop, Todd Golub, Erica L. Mayer, Viktor Adalsteinsson. PD11-06 Circulating tumor DNA association with residual cancer burden after neoadjuvant therapy in triple negative breast cancer in TBCRC 030 [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr PD11-06.

Proceedings of the fifth international Molecular Pathological Epidemiology (MPE) meeting

Cancer heterogeneities hold the key to a deeper understanding of cancer etiology and progression and the discovery of more precise cancer therapy. Modern pathological and molecular technologies offer a powerful set of tools to profile tumor heterogeneities at multiple levels in large patient populations, from DNA to RNA, protein and epigenetics, and from tumor tissues to tumor microenvironment and liquid biopsy. When coupled with well-validated epidemiologic methodology and well-characterized epidemiologic resources, the rich tumor pathological and molecular tumor information provide new research opportunities at an unprecedented breadth and depth. This is the research space where Molecular Pathological Epidemiology (MPE) emerged over a decade ago and has been thriving since then. As a truly multidisciplinary field, MPE embraces collaborations from diverse fields including epidemiology, pathology, immunology, genetics, biostatistics, bioinformatics, and data science. Since first convened in 2013, the International MPE Meeting series has grown into a dynamic and dedicated platform for experts from these disciplines to communicate novel findings, discuss new research opportunities and challenges, build professional networks, and educate the next-generation scientists. Herein, we share the proceedings of the Fifth International MPE meeting, held virtually online, on May 24 and 25, 2021. The meeting consisted of 21 presentations organized into the three main themes, which were recent integrative MPE studies, novel cancer profiling technologies, and new statistical and data science approaches. Looking forward to the near future, the meeting attendees anticipated continuous expansion and fruition of MPE research in many research fronts, particularly immune-epidemiology, mutational signatures, liquid biopsy, and health disparities.

Abstract 1028: Phosphate dysregulation as a novel therapeutic strategy in ovarian and uterine cancers

Precision medicine promises to improve the treatment of cancer patients, but a lack of therapeutic targets and associated predictive biomarkers limit this reality. To identify novel strategies, we integrate genome-scale CRISPR viability screens across many cancer models with cellular and molecular features to systematically define The Cancer Dependency Map. Using this data, we have identified that XPR1, an inorganic phosphate exporter protein, is a highly selective dependency gene in ovarian and uterine cancers. These cancers are sensitive to loss of XPR1 due to over-expression of SLC34A2, a phosphate importer protein. These data suggest a synthetic lethal relationship in which intracellular phosphate homeostasis is dysregulated in cancer. As proof-of-concept of pharmacological inhibition of XPR1, we have developed protein ligands based on the receptor binding domain of viruses which use XPR1 for cellular entry. These ligands inhibit XPR1 and kill cancer cells in an on-mechanism manner, but may be limited in their clinical utility. As such, we are deepening our understanding of the mechanisms of XPR1-dependent phosphate efflux, and have identified a novel partner protein that is integral to phosphate efflux, possibly revealing functional domains that small molecule inhibitors might target. Overall, these data highlight a novel mechanism to treat cancers by leveraging cancer-specific phosphate dysregulation and further reinforce the Cancer Dependency Map as a powerful engine to uncover novel therapeutic vulnerabilities.

Citation Format: Daniel Bondeson, Brenton Paolella, Adhana Asfaw, Michael Rothberg, Thomas Skipper, Gabriel Mesa, Alfredo Gonzalez, Lauren E. Surface, Kentaro Ito, Mariya Kazachkova, William N. Colgan, Allie Warren, Joshua Dempster, J Michael Krill-Burger, Maria Ericsson, Andrew Tang, Iris Fung, Emily S. Chambers, Mai Abdusamad, Nancy Dumont, John Doench, Federica Piccioni, David Root, Jesse Boehm, William C. Hahn, Michael Mannstadt, James McFarland, Francisca Vazquez, Todd Golub. Phosphate dysregulation as a novel therapeutic strategy in ovarian and uterine cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1028.

Abstract 2682: Small molecule targeting the lipoic acid post-translational modification impacts proliferation of colorectal and PIK3CA-mutant cell lines

To identify novel therapeutic targets, we utilize the PRISM platform, a multiplexed cell line viability technology of 500 solid tumor cell lines and correlate responses to functional genomic and baseline genetic data. We describe ESD0140656, a small molecule with selective anti-proliferative effect on colorectal and PIK3CA-mutant cell lines. Response to ESD0140656 is correlated to sensitivity to CRISPR/Cas9 KO of components of the protein lipoylation pathway and OGDH complex members, which catalyze a step of the TCA cycle. Lipoylation is a rare post-translational modification attached to just four enzymes in humans, including the OGDH complex. Knockout of the protein that transfers lipoic acid to these four enzymes (LIPT1) sensitizes cells to ESD0140656, and ESD0140656 treatment leads to reduction of lipoic acid in cells. These results suggest ESD0140656 targets the lipoylation pathway and may represent a novel therapeutic angle for colorectal and PIK3CA-mutant tumors.

Citation Format: Laura Doherty, Tenzin Sangpo, Peter Tsvetkov, John Davis, Navid Dianati, Wolfgang Schwede, Katja Zimmermann, Laura Evans, Aldo Amatucci, Henrik Seidel, Atanas Kamburov, Gizem Akcay, Todd Golub, Ashley Eheim, Nils Burkhardt, Knut Eis, Sven Christian, Matt Rees, Jennifer Roth. Small molecule targeting the lipoic acid post-translational modification impacts proliferation of colorectal and PIK3CA-mutant cell lines [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2682.

Abstract 3088: The efficient utilization of paracrine support from established cell lines for breast/ovarian cancer model generation

Ex vivo cancer cell models provide the starting material for in-depth mechanistic studies of cancer. However, the clinical/histopathologic, biomarker, and genetic heterogeneity of breast cancer has not been well represented in the current breast cancer cell model collection. While published breast cancer model generation protocols have been helpful, their high failure rates indicate the urgent need to improve model derivation efficiency. Here, the Broad Cancer Cell Line Factory (CCLF) project presents a novel model derivation technology to generate breast and ovarian cancer organoids with a high success rate by leveraging the paracrine support from historical cancer cell lines.

We observed that most established breast cancer cell lines can grow in a simple basal media with 10% fetal bovine serum; We hypothesized that historical cell lines may secrete vital growth factors that support breast cancer cells' survival and growth. To test this, we randomly selected a pool of 20 breast cancer cell lines, collected its conditioned media (CM20) and incorporated the CM20 as a supplement into our empirical rich media matrix (HYBRID, 16 mixed media conditions) with a Matrigel culturing system. Three-dimensional (3D) structures formed at Day 14-21 in the CM20 supplementary conditions compared to conditions without CM20 and only organoids with the CM20 supplement could be propagated to passage 5 and beyond. We performed pan-cancer targeted sequencing to evaluate tumor content of these organoids at passage 5 with paired tumor tissues. In our first 10 attempts, 95% of organoid cultures were genomically verified as high purity tumor models, indicating the CM20 is essential to enrich breast cancer cell growth in an in vitro culturing setting.

We applied the CM20 to ovarian cancers and observed a similar success rate suggesting a tissue-specific supporting manner. We tested conditioned media collected from other historical cancer cell lines but the breast/ovarian cancer organoid growth effect was not recapitulated. Importantly, when testing the individual breast cancer cell lines from the pool of 20, we discovered one cell line to be supporting the effect. More biochemistry work is needed to dissect the possible factors secreted by the line and molecular mechanisms of cancer cell survivors but preliminary data suggests the secretion factors are most likely proteins.

We generated 27 breast/ovarian cancer cell models using this technology and RNAseq data shows the breast cancer organoids still express their expected molecular subtype markers. 22 breast/ovarian cancer organoids have been propagated long-term with 17(out of 22) deposited to ATCC. Overall, this method provides an efficient model generation rate for female cancers. We anticipate that this method will not only allow us to quickly increase breast cancer cell model diversity but shed light on a new direction for breast cancer dependencies

Citation Format: Rebecca Deasy, Xin Jin, Pierre Michel Jean Beltran, Adel Atari, Madison Liistro, Cheryl Thompson, Stefanie Avril, Julie Boerner, Payal Pradhan, Samuel Klempner, Keith Ligon, William Sellers, Steven Carr, Todd Golub, Yuen-Yi (Moony) Tseng. The efficient utilization of paracrine support from established cell lines for breast/ovarian cancer model generation [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3088.

Genome-scale screens identify factors regulating tumor cell responses to natural killer cells

To systematically define molecular features in human tumor cells that determine their degree of sensitivity to human allogeneic natural killer (NK) cells, we quantified the NK cell responsiveness of hundreds of molecularly annotated 'DNA-barcoded' solid tumor cell lines in multiplexed format and applied genome-scale CRISPR-based gene-editing screens in several solid tumor cell lines, to functionally interrogate which genes in tumor cells regulate the response to NK cells. In these orthogonal studies, NK cell-sensitive tumor cells tend to exhibit 'mesenchymal-like' transcriptional programs; high transcriptional signature for chromatin remodeling complexes; high levels of B7-H6 (NCR3LG1); and low levels of HLA-E/antigen presentation genes. Importantly, transcriptional signatures of NK cell-sensitive tumor cells correlate with immune checkpoint inhibitor (ICI) resistance in clinical samples. This study provides a comprehensive map of mechanisms regulating tumor cell responses to NK cells, with implications for future biomarker-driven applications of NK cell immunotherapies.

Abstract NG01: Synthetic lethal interaction between the ESCRT paralog enzymes VPS4A and VPS4B in cancers harboring loss of chromosome 18q or 16q

Background: Discovery of new biomarker-linked cancer therapeutic targets may enable novel drug development and ultimately lead to advances in clinical care. Somatic copy number alterations (CNAs) leading to loss of tumor suppressor gene function constitute important driver events in tumorigenesis. Unfortunately, there are few existing therapeutic options to target the oncogenic processes evoked by tumor suppressor inactivation. However, developing drugs that target tractable synthetic lethal interactions with common somatic CNAs represents a promising approach to attain cancer-selective therapeutics. Synthetic lethality refers to the observation that for certain gene pairs, inactivation of either gene is tolerated but combined loss-of-function of both genes results in decreased cell viability. Synthetic lethal relationships in cancer have been defined in several different contexts, including among paralog genes for which dependency on one paralog is conferred by loss of a second functionally redundant paralog gene. Since targeting synthetic lethal relationships in cancer may yield a wide therapeutic window of efficacy between tumor and normal cells, identification of pharmacologically tractable synthetic lethal targets remains a priority for oncology drug development programs. Results and Discussion: To systematically define synthetic lethal vulnerabilities associated with genomic loss of established tumor suppressor genes, we analyzed genome-scale CRISPR-SpCas9 and RNA interference loss-of-function screening data from over 600 cancer cell lines. We identified and prioritized 193 synthetic lethal interactions with genomic loss of one or more of 51 common tumor suppressor genes. In particular, we discovered that the paralog genes encoding vacuolar protein sorting 4 homolog A and B (VPS4A and VPS4B) are selective genetic vulnerabilities for tumors harboring genomic copy loss of SMAD4 or CDH1 due to co-deletion of VPS4B or VPS4A, respectively. VPS4B is located on the long arm (q) of chromosome 18, 12.3 Mb away from SMAD4, while VPS4A is located 0.476 Mb downstream of CDH1 (encoding E-cadherin) on chromosome 16q. Thus, cancer cells with genomic loss of VPS4B selectively depend on expression of VPS4A for survival, and tumors with loss of VPS4A depend on VPS4B expression. Co-deletion of SMAD4 and VPS4B is commonly observed in approximately 33% of human cancer, with particularly high rates of loss in pancreatic cancers (68%), colorectal (71%) and renal cell carcinomas (17%) and to a lesser extent in cancers of the bile duct, lung, prostate, esophagus, uterus, cervix and ovary. Meanwhile, loss of CDH1 and VPS4A occurs frequently in cancers of the stomach, breast, skin, colon and prostate. VPS4A and B function as AAA ATPases which are critical for the regulation of endosomal sorting complex required for transport (ESCRT), a multimeric protein complex essential for inverse membrane remodeling. The ESCRT machinery is involved in a range of cellular processes, including cytokinesis, membrane repair, autophagy and endosomal processing. VPS4A/B are believed to form asymmetric hexameric complexes that are recruited to ESCRT-III filaments to drive ESCRT-mediated membrane fission and sealing. Here, we demonstrate that suppression of VPS4A in cancer cells with reduced copy number of VPS4B leads to accumulation of CHMP4B-containing ESCRT-III filaments, cytokinesis defects, nuclear membrane abnormalities and micronucleation, ultimately resulting in G2/M cell cycle arrest and apoptosis. We also observed that VPS4 suppression leads to defects in endosomal and endoplasmic reticulum structure. Furthermore, upon VPS4A suppression, we observed potent in vivo tumor regressions, which led to markedly prolonged survival in mouse xenograft models of pancreatic cancer and rhabdomyosarcoma harboring genomic loss of VPS4B. To understand regulators of VPS4A dependency, we performed a CRISPR-SpCas9 genome-scale screen in a pancreatic cancer cell line in the context of VPS4A suppression. We identified multiple genes that promote or suppress VPS4A dependency. Cancer cell sensitivity to VPS4A suppression was potently enhanced by disruption of regulators of the abscission checkpoint, including genes encoding the ULK3 kinase and the ESCRT-III proteins CHMP1A and CHMP1B. The abscission checkpoint is a genome protection mechanism that relies on Aurora B kinase (AURKB) and ESCRT-III subunits to delay abscission in response to chromosome mis-segregation to avoid DNA damage and aneuploidy. These findings suggest that inhibition of the ESCRT pathway and blockade of the abscission checkpoint could provide strategies to further enhance sensitivity of cancer cells to VPS4A suppression. Moreover, through CRISPR-SpCas9 screening and integrative transcriptomic and proteomic analysis, we also identified a strong correlation between baseline interferon response gene expression and VPS4A dependency. Indeed, when we treated VPS4B-deficient cells with interferon-β and interferon-γ to induce interferon signaling, we observed a pronounced sensitization of these cells to VPS4A depletion, thus suggesting that immune signals from the tumor microenvironment may influence VPS4 dependency. These data collectively suggest potential future therapeutic strategies for combination with VPS4A inhibition. Finally, we demonstrate through mutant rescue experiments that the ATPase domain is critical for the function of VPS4A in mediating survival of cells with partial copy loss of VPS4B. Furthermore, we provide data that elucidate the degree to which VPS4A and VPS4B cooperate and form functional complexes in human cancer cells. Although VPS4A and B demonstrate 80.5% homology, the development of small molecules that differentially target VPS4A in cells with VPS4B loss or VPS4B in cells with VPS4A loss remains a tractable possibility due to small structural differences near the ATP-binding pocket. Moreover, combined inhibition of VPS4A and VPS4B may also prove effective and clinically tolerable given a potential therapeutic window arising from gene dosage alterations and differences in total VPS4A/B levels in tumor versus normal cells.

Citation Format: Jasper E. Neggers, Brenton Paolella, Adhana Asfaw, Michael V. Rothberg, Tom A. Skipper, Radha Kalekar, Michael Burger, Neekesh Dharia, Guillaume Kugener, Jeremie Kalfon, Nancy Dumont, Yvonne Li, Liam Spurr, Annan Yang, Wenbo Wu, AndrewAdam Durbin, Brian M. Wolpin, David E. Root, Jesse Boehm, Andrew D. Cherniack, Aviad Tsherniak, Andrew L. Hong, William C. Hahn, Kimberly Stegmaier, Todd Golub, Francisca Vazquez, Andrew J. Aguirre. Synthetic lethal interaction between the ESCRT paralog enzymes VPS4A and VPS4B in cancers harboring loss of chromosome 18q or 16q [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr NG01.

BIOL-03. PROTEIN TRANSLATION FROM NON-CODING GENOMIC LOCI PRODUCE BIOLOGICALLY-ACTIVE PROTEINS IMPLICATED IN CANCER CELL SURVIVAL IN PEDIATRIC BRAIN TUMORS

Protein translation is both a fundamental cellular process essential for life as well as an oncogenic mechanism employed by tumors to enact cancer cell biology. While protein translation is most readily manifest in the ~20,000 known human protein coding genes, there are, in fact, several thousand additional regions of the cancer genome that are translated and contribute the complexity of the molecular milieu of cancer. Here, we systematically addressed the question of whether such uncharacterized genomic regions encode truly biologically active proteins and applied these findings to pediatric brain tumors. We experimentally interrogated 553 candidates selected from non-canonical open reading frame (ORF) datasets. Of these, 57 induced viability defects when knocked out in a broad array of human cancer cell lines. Upon ectopic expression, 257 showed evidence of protein expression and 401 induced gene expression changes. CRISPR tiling and start codon mutagenesis indicated that their biological effects required translation as opposed to RNA-mediated effects. We characterized several of these in the context of pediatric brain tumors, where dense CRISPR tiling screens revealed unique functional relevance of dozens of non-canonical ORFs in pediatric brain cancer cell survival. We found that one of these ORFs, ASNSD1 uORF, encodes a well-folded protein whose translation is a selective genetic dependency distinct from the adjacent ASNSD1 annotated protein. In vitro molecular biology assays confirmed the MYC-amplified medulloblastoma cell lines had a heightened dependency on this protein, and that MYC binds to the promoter of this gene, with MYC expression correlating with ASNSD1 in patient tumors. Co-immunoprecipitation assays defined ASNSD1 uORF as a novel member of the prefoldin complex of cytoplasmic protein stability regulators. Overall, our experiments suggest that the abundant protein translation found in the “non-coding” genome may produce biologically active non-canonical ORFs that are potential therapeutic targets.

DDRE-07. FATTY ACID SYNTHESIS IS REQUIRED FOR BREAST CANCER BRAIN METASTASIS

Brain metastases are refractory to therapies that otherwise control systemic disease in patients with human epidermal growth factor receptor 2 (HER2+) breast cancer, and the unique brain microenvironment contributes to this therapy resistance. Nutrient availability can vary across tissues, therefore metabolic adaptations required for breast cancer growth in the brain microenvironment may also introduce liabilities that can be exploited for therapy. Here, we assessed how metabolism differs between breast tumors growing in the brain versus extracranial sites and found that fatty acid synthesis is elevated in breast tumors growing in the brain. We determine that this phenotype is an adaptation to decreased lipid availability in the brain relative to other tissues, which results in a site-specific dependency on fatty acid synthesis for breast tumors growing at this site. Genetic or pharmacological inhibition of fatty acid synthase (FASN) reduces HER2+ breast tumor growth in the brain, demonstrating that differences in nutrient availability across metastatic sites can result in targetable metabolic dependencies.

Abstract PO041: Landscape of molecular events regulating tumor cell responses to natural killer cells

Natural killer (NK) cells exhibit potent activity in pre-clinical models of diverse hematologic malignancies and solid tumors and infusion of high numbers of NK cells, either autologous or allogeneic, after their ex vivo expansion and activation, has been feasible and safe in clinical studies. To systematically define molecular features in human tumor cells which determine their degree of sensitivity to human allogeneic NK cells, we quantified the NK cell responsiveness of hundreds of molecularly-annotated “DNA-barcoded” solid tumor cell lines in multiplexed format (PRISM; Profiling Relative Inhibition Simultaneously in Mixtures approach), correlating cytotoxicity scores for each cell line with the CCLE transcriptional data (RNA-seq), to reveal genes that are associated with resistance or sensitivity to NK cells. In addition, we applied genome-scale CRISPR-based gene editing screens in several solid tumor cell lines to interrogate, at a functional level, which genes regulate tumor cell response to NK cells. Based on these orthogonal studies, NK sensitive tumor cells tend to exhibit high levels of the NK cell-activating ligand B7-H6 (NCR3LG1); low levels of the inhibitory ligand HLA-E; microsatellite instability (MSI) status; high transcriptional signature for chromatin remodeling complexes and low antigen presentation machinery genes. Treatment with an HDAC inhibitor reduced the sensitivity of SW620 colon cancer cells, increased antigen presentation machinery, including HLA-E, and reduced B7-H6. Importantly, we observe that transcriptional signatures of NK cell-sensitive tumor cells correlate with immune checkpoint inhibitor resistance in clinical samples. Strikingly, comprehensive analysis of the CCLE transcriptional signatures revealed that cell lines with mesenchymal-like program tend to be more sensitive to NK cells treatment, compared with cell lines of epithelial-like program. Indeed, mesenchymal tumors tend to have lower expression of antigen presentation machinery in both CCLE and TCGA, suggesting a link between these two machieneries. This study provides a comprehensive map of mechanisms regulating tumor cell responses to NK cells, with implications for future biomarker-driven applications of NK cell immunotherapies.

Citation Format: Michal Sheffer, Emily Lowry, Nicky Beelen, Minasri Borah, Suha Naffar-Abu Amara, Chris C. Mader, Jennifer Roth, Aviad Tsherniak, Olga Dashevsky, Sara Gandolfi, Samantha Bender, Jordan Bryan, Cong Zhu, Li Wang, Ricardo De-Matos Simoes, Channing Yu, Yiguo Hu, Olli Dufva, Marios Giannakis, Todd Golub, Rizwan Romee, Satu Mustjoki, Aedin C. Culhane, Lotte Wieten, Constantine S. Mitsiades. Landscape of molecular events regulating tumor cell responses to natural killer cells [abstract]. In: Abstracts: AACR Virtual Special Conference: Tumor Immunology and Immunotherapy; 2020 Oct 19-20. Philadelphia (PA): AACR; Cancer Immunol Res 2021;9(2 Suppl):Abstract nr PO041.

TBIO-26. NON-CANONICAL OPEN READING FRAMES ENCODE FUNCTIONAL PROTEINS ESSENTIAL FOR CANCER CELL SURVIVAL

The brain is the foremost non-gonadal tissue for expression of non-coding RNAs of unclear function. Yet, whether such transcripts are truly non-coding or rather the source of non-canonical protein translation is unknown. Here, we used functional genomic screens to establish the cellular bioactivity of non-canonical proteins located in putative non-coding RNAs or untranslated regions of protein-coding genes. We experimentally interrogated 553 open reading frames (ORFs) identified by ribosome profiling for three major phenotypes: 257 (46%) demonstrated protein translation when ectopically expressed in HEK293T cells, 401 (73%) induced gene expression changes following ectopic expression across 4 cancer cell types, and 57 (10%) induced a viability defect when the endogenous ORF was knocked out using CRISPR/Cas9 in 8 human cancer cell lines. CRISPR tiling and start codon mutagenesis indicated that the biological impact of these non-canonical ORFs required their translation as opposed to RNA-mediated effects. We functionally characterized one of these ORFs, G029442—renamed GREP1 (Glycine-Rich Extracellular Protein-1)—as a cancer-implicated gene with high expression in multiple cancer types, such as gliomas. GREP1 knockout in >200 cancer cell lines reduced cell viability in multiple cancer types, including glioblastoma, in a cell-autonomous manner and produced cell cycle arrest via single-cell RNA sequencing. Analysis of the secretome of GREP1-expressing cells showed increased abundance of the oncogenic cytokine GDF15, and GDF15 supplementation mitigated the growth inhibitory effect of GREP1 knock-out. Taken together, these experiments suggest that the non-canonical ORFeome is surprisingly rich in biologically active proteins and potential cancer therapeutic targets deserving of further study.

EXTH-37. TARGETING EPIGENETIC VULNERABILITIES IDENTIFIED FROM A CRISPR SCREEN IN H3.3K27M DIPG

Children diagnosed with diffuse intrinsic pontine glioma (DIPG), a type of high grade glioma in the brainstem, currently have a dismal 5-year overall survival of only 2%. The majority of DIPG patients harbor a K27M mutation in histone 3.3 encoding genes (H3.3K27M). To understand if the aberrant epigenetic landscape induced by H3.3K27M provides an opportunity for novel targeted therapies, we conducted the first CRISPR/Cas9 screen using a focused library of 1,350 epigenetic regulatory and cancer related genes in six H3.3K27M DIPG patient-derived primary neurosphere cell lines. We identified gene dependencies in chromatin regulators, polycomb repressive complexes 1 and 2 (PRC1 and PRC2), histone demethylases, acetyltransferases and deacetylators as novel tumor cell dependencies in DIPG. We hypothesized that targeting dysregulated functions of chromatin regulators by genetically deleting and chemically targeting these epigenetically induced vulnerabilities, we could ameliorate, or even reverse the downstream oncogenic effects of the aberrant epigenetic landscape of DIPG. In our secondary CRISPR nanoscreen, we first used six single guide RNAs (sgRNA) to knockout each gene using CRISPR/Cas9 ribonucleoprotein nucleofections, followed by use of three best sgRNAs combined with homology directed repair templates. Compared to lentiviral delivery, nucleofection is a rapid method, with reduced off-target toxicity, suitable for single gene knockouts in DIPG neurospheres. Secondary CRISPR validations confirmed dependencies in BMI1, CBX4, KDM1A, EZH2, EED, SUZ12, HDAC2, and EP300. Next, we conducted a chemical screen using 20 inhibitors and degraders to target the aberrant activity of HDAC, KDM1A, P300/CBP, PRC1 and PRC2. We identified eight chemical compounds that were effective in H3.3K27M DIPG neurosphere cell lines at low drug concentrations. Among these, an inhibitor and degrader targeting P300/CBP activity indicates a novel strategy of epigenetic therapy in DIPG. Through our combinatorial testing, we will identify a synergistic combination of epigenetic therapy for treating children diagnosed with H3.3K27M DIPG.

Abstract B121: An oncogene-linked prodrug strategy in lung cancer

KEAP1-mutant non-small cell lung cancer is a high prevalence indication that responds poorly to conventional chemotherapy owing to constitutive activation of NRF2 and its associated drug metabolism target genes. Using an approach to identify compounds that selectively kill cancer cell lines in a biomarker-driven manner, we identify BRD-K19050021 (K1905), a compound displaying strong toxicity in cells expressing CYP4F11 - a cytochrome p450 family member and NRF2 transcriptional target. Using genome-wide pooled CRISPR screens, we show that CYP4F11 activity is necessary and sufficient for response to K1905, and in acquired resistance models developed from multiple cell lines, we show a minor subpopulation of cells mechanistically converge on suppressed CYP4F11 expression to bypass K1905 response. CYP4F11 converts K1905 from a prodrug into a covalent active metabolite that alkylates several cellular targets, triggering all three canonical arms of the unfolded protein response pathway and culminating in cell death. We propose that CYP4F11 and similar metabolic enzyme activities promoted by oncogenic drivers represent a unique opportunity to restrict prodrug activation within tumors, provided that normal tissue expression of such prodrug-converting enzymes do not diminish therapeutic index.

Citation Format: Aaron Boudreau, Jennifer Roth, James Rong, Niclas Olsson, Chris Mader, Jordan Byran, Jordan Rossen, Li Wang, Kevin Larpenteur, Amy Goodale, Colin Trepicchio, Samantha Bender, Aviad Tsherniak, Aravind Subramanian, Mustafa Kocak, Federica Piccioni, Josh Bittker, Cong Zhu, Frank Li, Nick Eriksson, Daphne Koller, Fiona McAllister, Todd Golub, Jeff Settleman, Ari Firestone, David Stokoe. An oncogene-linked prodrug strategy in lung cancer [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2019 Oct 26-30; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2019;18(12 Suppl):Abstract nr B121. doi:10.1158/1535-7163.TARG-19-B121

Abstract 2690: Massively parallel multiplexed methods to screen hundreds of barcoded cancer cell line models with small molecules or genetic perturbations using next-generation sequencing

Phenotypic screening is a valuable tool to identify compounds to treat cancer, but is limited as it is time and resource intensive to screen hundreds of cancer cell lines. In order to increase the throughput of phenotypic screening, we set out to expand and further develop the PRISM method (Profiling Relative Inhibition Simultaneously in Mixtures) [Yu et al., 2016]. PRISM is a method to barcode cell lines with a unique 24 nucleotide barcode and mix them together to screen simultaneously in a pool, decreasing the time and cost of screening. The previously described PRISM method was limited in that it used only 100 adherent cell lines of one cancer model context, it required the use of a highly specialized Luminex bead-based detection system, and it was only applicable to small molecule screening.

Here we report on the expansion of our barcoded cell line collection to 800 cancer cell lines of over 45 lineages, an improved method called PRISMseq using next-generation sequencing, and a novel method to perform genetic perturbations in 500 cancer cell lines simultaneously called PRISPR (PRISM/CRISPR). PRISMseq allows for assaying compound cell line sensitivity profiles in a large pool of hundreds of cell lines and detecting the DNA barcodes using next-generation sequencing (NGS). We recapitulate the known biology for established oncology drugs like the BRAF, EGFR, BCR-ABL and MDM2 inhibitor classes using this method with our cell line panel. We also further extended this method to be applicable to genetic perturbation using CRISPR/Cas9 knockout of individual genes concurrently in hundreds of cancer cell lines. We are able to recover expected biomarkers or genetic dependencies for our CRISPR/Cas9 validation reagents.

This expanded PRISM profiling approach increases statistical power due to the addition of cancer contexts in our cell line collection, and improves versatility by enabling the screening of both small molecules and genetic perturbations. We believe that this method has overcome the limitations of the original PRISM method. It has also overcome the limitations of phenotypic screening, as the resources required to screen hundreds of cell lines has been decreased by orders of magnitude.

Citation Format: Samantha Bender, Cong Zhu, Li Wang, Michael Rothberg, Joshua Dempster, Brenton Paolella, Mustafa Kocak, Massami Laird, Jordan Rossen, Karolina Stumbraite, Jordan Bryan, Vickie Wang, John Doench, Francisca Vazquez, Aviad Tsherniak, Todd Golub, Jennifer Roth. Massively parallel multiplexed methods to screen hundreds of barcoded cancer cell line models with small molecules or genetic perturbations using next-generation sequencing [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2690.

Abstract 5384: The Angiosarcoma Project: Generating the genomic landscape of an exceedingly rare cancer through a nationwide patient-driven initiative

Angiosarcoma (AS) is an exceedingly rare soft tissue sarcoma, with an incidence of 300 cases/yr and a 5-year disease-specific survival of 30%. The low incidence has impeded large-scale research efforts that may lead to improved clinical outcomes. To address this, we launched a nationwide clinical-genomics study in order to empower patients to accelerate research by sharing their normal and tumor samples and clinical information remotely. Patients can access the study through an online portal (ASCproject.org). Enrolled patients are mailed saliva and blood draw kits. The study team obtains medical records and stored FFPE tumor samples. All received FFPE samples are examined by an expert pathologist to confirm a diagnosis of angiosarcoma. In order to validate that our processes would enable the generation of a robust dataset from tissues acquired from multiple institutions, we sought to characterize previously described genes known to be altered in angiosarcoma (e.g., TP53, NF1, KDR, BRCA2, MET, ARID1A, POT1, BRCA1, ASXL1, KDM6A, BRAF, SETD2, PTPRB, NRAS). A total of 251 patients have enrolled since the project launched in March of 2017. Primary locations of AS are primary breast 59 (25%), breast with prior radiation 45 (19%), head/face/neck/scalp 52 (22%), bone/limb 26 (11%), abdomen 5 (2%), heart 5 (2%), lung 2 (1%), liver 1 (1%), lymph 1 (0.4%), multiple locations 25 (11%), and other locations 12 (5%); 107 (52%) reported being disease free at the time of enrollment. To date, we have received 129 saliva kits, 106 medical records, 19 blood samples, and 36 tissue samples. Whole-exome sequencing (WES) was performed on 21 FFPE/saliva matched pairs with a goal mean target coverage of 150x for tumors. Ultra-low pass whole-genome sequencing (0.1x) was performed on cell free DNA (cfDNA) from plasma in order to determine tumor fraction. Of 10 cfDNA samples sequenced, 4 samples met criteria to perform WES. Additionally, transcriptome sequencing was performed on 9 FFPE samples. Sequence data processing and analysis has been completed on the first 10 samples and is in progress for the subsequent samples. Alterations were detected in genes previously described to be affected in angiosarcoma. Recurrent mutations in TP53 were detected in 50% (5/10) of analyzed samples, comprising 3 missense mutations, 1 frameshift deletion, and 1 frameshift insertion. Alterations were seen in at least one sample in all other genes selected for this initial analysis. This initiative demonstrates the feasibility of studying tissues from geographically dispersed patients and serves as proof of concept that patient-driven genomics efforts can democratize research for exceedingly rare cancers. Enrollment is still in progress, and additional samples will be sequenced and analyzed at scale. The data generated from these studies will be deposited into the public domain in six-month intervals.

Citation Format: Michael Dunphy, Esha Jain, Elana Anastasio, Mary McGillicuddy, Rachel Stoddard, Beena Thomas, Sara Balch, Kristin Anderka, Katie Larkin, Niall Lennon, Yen-Lin Chen, Andrew Zimmer, Esme O. Baker, Simone Maiwald, Jen Hendrey Lapan, Jason Hornick, Chandrajit Raut, George Demetri, Eric Lander, Todd Golub, Nikhil Wagle, Corrie Painter. The Angiosarcoma Project: Generating the genomic landscape of an exceedingly rare cancer through a nationwide patient-driven initiative [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 5384.

The Metastatic Prostate Cancer project (MPCproject): Translational genomics through direct patient engagement

279

Background: While there has been substantial advancement in the genomic understanding of metastatic prostate cancer (MPC), there is still much to be discovered. Additional progress is dependent upon obtaining a large amount of clinically-annotated genomic data. Therefore, we piloted a direct-to-patient nationwide research initiative where patients can contribute their medical records and biospecimens to accelerate research ( mpcproject.org ). Methods: In collaboration with patients and advocacy groups, we have developed a website ( mpcproject.org ). Participants are asked to complete a 17-question survey about their experiences with prostate cancer and an electronic informed consent. All participants receive a saliva kit for germline DNA and blood kit for circulating tumor DNA (ctDNA). Additionally, medical records are collected and archived tissue samples are requested if available. Ultra low pass whole genome sequencing (ULP-WGS) and whole exome sequencing (WES) are performed on the whole blood samples. WES is performed on saliva samples. Genomic, clinical, and patient-reported data will be shared widely with the research community. Aggregate study results will be reported to patients. Results: As of October 2017, 12 pilot patients aged 47-74 from 7 states, provided informed consent. 7 saliva kits, 4 blood kits, and 2 medical records were received. 4 patients were diagnosed with de novo metastatic disease, 8 reported a family history of breast and/or prostate cancer, 6 reported a secondary malignancy. All blood kits were submitted for ULP-WGS and WES. Updated genomic, clinical, and patient-reported data will be presented. Conclusions: We have provided preliminary evidence that partnering directly with MPC patients enabled the remote collection of saliva and blood samples, medical records, and patient-reported data. At the conclusion of the pilot phase, the MPC Project will open enrollment for all men with metastatic and advanced prostate cancer in the US and Canada.

Abstract IA05: Genomic approaches to drug discovery

It is now becoming possible to bring powerful genomic methods to bear on the process of drug discovery. This talk will illustrate 1) the use of the Connectivity Map, a large-scale effort to develop a compendium of gene expression signatures of cellular state; 2) the use of PRISM, a new multiplexed, high-throughput approach to screening cancer cell lines for anticancer activity; and 3) the use of cancer models (cell lines and patient-derived xenografts) in discovery, with particular focus on clonal drift in these models and the impact of that drift on drug sensitivity.

Citation Format: Todd Golub. Genomic approaches to drug discovery [abstract]. In: Proceedings of the AACR Conference on Advances in Sarcomas: From Basic Science to Clinical Translation; May 16-19, 2017; Philadelphia, PA. Philadelphia (PA): AACR; Clin Cancer Res 2018;24(2_Suppl):Abstract nr IA05.

A mutational signature reveals alterations underlying deficient homologous recombination repair in breast cancer

Biallelic inactivation of BRCA1 or BRCA2 is associated with a pattern of genome-wide mutations known as signature 3. By analyzing ∼1,000 breast cancer samples, we confirmed this association and established that germline nonsense and frameshift variants in PALB2, but not in ATM or CHEK2, can also give rise to the same signature. We were able to accurately classify missense BRCA1 or BRCA2 variants known to impair homologous recombination (HR) on the basis of this signature. Finally, we show that epigenetic silencing of RAD51C and BRCA1 by promoter methylation is strongly associated with signature 3 and, in our data set, was highly enriched in basal-like breast cancers in young individuals of African descent.

Abstract A02: Expanding tumor chemical-genetic interaction map using next-generation cancer models

The development of new cancer therapeutics requires sufficient genetic and phenotypic diversity of cancer models. Current collections of human cancer cell lines are limited and for many rare cancer types, zero models exist that are broadly available. Here, we report results from the pilot phase of the Cancer Cell Line Factory (CCLF) project that aims to overcome this obstacle by systematically creating next-generation in vitro cancer models from adult and pediatric cancer patients' specimens and making these models broadly available.

We first developed a workflow of laboratory, genomics and informatics tools that make it possible to systematically compare published ex vivo culture conditions for each individual tumor to enable the scientific community to iterate towards disease-specific culture recipes. Based on sample volume and rarity, 4-100 conditions were applied to each sample and all data was captured in a custom Laboratory Information Management System to enhance subsequent predictions. We developed a $150, 5-day turnaround genomics panel to validate cultures based on genomics. Importantly, we show that tumor genomics can be retained in such patient-derived models and tumor genomics are generally stable across 20 passages. Since the inception of this project, we have processed over 600 patient cancer specimens from 450 patients across 16 tumor types and report the successful generation of over 100 genomically characterized adult and pediatric cancer and normal models.

We next hypothesized that novel patient-derived cell models could be used to enhance dependency predictions. To do so, we tested 72 cell lines against the informer set of 440 compounds developed by the Broad Cancer Target Discovery and Development (CTD2) Center. We show that generating cell lines and testing their sensitivities within 3 months is feasible and the high-throughput drug responses are reproducible. Moreover, to strengthen relationships between drug sensitivities and cellular features, we compared results with recently published data on the identical compounds tested against 860 existing cell lines. With this approach, we show that many chemical-genetic interaction vulnerabilities can be rapidly assessed. Importantly, adding more cancer models with the dimensions of quantity and diversity increases the predictive power of chemical-genetic interaction map. We are currently evaluating these drug sensitivity predictors for novel co-dependencies. Overall, our proof-of-concept framework demonstrates initial feasibility of rapidly generating cancer models at scale and expanding the chemical-genetic interaction map to identify new cancer vulnerability.

Citation Format: Yuen-Yi (Moony) Tseng, Andrew Hong, Shubhroz Gill, Paula Keskula, Srivatsan Raghavan, Jaime Cheah, Aviad Tsherniak, Francisca Vazquez, Sahar Alkhairy, Anson Peng, Abeer Sayeed, Rebecca Deasy, Peter Ronning, Philip Kantoff, Levi Garraway, Mark Rubin, Calvin Kuo, Sidharth Puram, Adi Gazdar, Nikhil Wagle, Adam Bass, Keith Ligon, Katherine Janeway, David Root, Stuart Schreiber, Paul Clemons, Todd Golub, William Hahn, Jesse Boehm. Expanding tumor chemical-genetic interaction map using next-generation cancer models [abstract]. In: Proceedings of the AACR Precision Medicine Series: Opportunities and Challenges of Exploiting Synthetic Lethality in Cancer; Jan 4-7, 2017; San Diego, CA. Philadelphia (PA): AACR; Mol Cancer Ther 2017;16(10 Suppl):Abstract nr A02.

Abstract 2028: PDE3A modulation for cancer therapy

In a differential cytotoxicity screen, we identified a novel small molecule modulator of phosphodiesterase 3A (PDE3A) that kills cancer cells expressing elevated levels of PDE3A and SLFN12 (de Waal, Nat Chem Biol, 2016). Treatment with this cell-selective cytotoxic small molecule, DNMDP, induces complex formation between PDE3A and SLFN12, resulting in apoptosis. Inhibition of PDE3A enzymatic activity is not sufficient for cancer cell killing, and expression of both PDE3A and SLFN12 are required. Although the mechanism of signaling to the apoptosis machinery remains unclear, we examined more closely the role of the PDE3A-SLFN12 complex in cancer cell killing mediated by DNMDP. We found that cancer cell lines made resistant to DNMDP by persistent exposure downregulated SLFN12 expression and that re-expression of SLFN12 was sufficient to restore sensitivity. Furthermore, ectopic expression of PDE3A and SLFN12 are sufficient to sensitize cancer cells to DNMDP. These data underscore the tight correlation of PDE3A-SLFN12 complex formation and cancer cell killing mediated by DNMDP.

Citation Format: Xiaoyun Wu, Timothy Lewis, Luc de Waal, Galen Gao, Jian Zhang, Monica Schenone, Colin Garvie, Brett Diamond, Selena Lorrey, Andrew Cherniack, Steven Corsello, Alex Burgin, Todd Golub, Stuart Schreiber, Matthew Meyerson, Heidi Greulich. PDE3A modulation for cancer therapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2028. doi:10.1158/1538-7445.AM2017-2028

Abstract 1953: Accelerating prediction of pediatric and rare cancer vulnerabilities using next-generation cancer models

Ongoing pre-clinical efforts aim to deploy genome-scale CRISPR/Cas9 technology and large collections of small molecules to catalog maps of cancer vulnerabilities at scale. However, such efforts in pediatric and rare cancers have lagged behind comparable efforts in more common cancer types due to the dearth of cell models. Here, we present an update from our “Cancer Cell Line Factory” project on efforts to overcome key laboratory and biologistics challenges precluding progress in pediatric and rare cancers. This effort, now in it’s 3rd year, represents an industry scale pipeline aiming to generate, characterize and share novel cancer models of many tumor types with the scientific community. Overall, we have processed 1153 samples from 818 patients across over 16 cancer types through this pipeline with a 28% success rate overall, including over 350 patient samples from rare and pediatric cancers. To optimize conditions for each tumor type, we have systematically compared published methods including (1) next-generation 2-dimension, (2) organoid and (3) standard approaches and have captured all information with a data management system that should enhance the ability to predict optimal ex vivo propagation conditions for future samples. Among the successful cell models verified already as part of this effort, we have generated a series of over 30 unique pediatric and rare cancer models, many of which represent the first of their kind. We screened these and other models against a library of highly annotated 440 small molecules that were previously tested against 860 existing cancer cell lines. Our results suggest that dependency data generated with novel next-generation cell cultures is potentially backwards-compatible with existing small molecule dependency datasets. Furthermore, we tested the novel Broad Institute Drug Repurposing library consisting of 4100 approved therapeutics, or those under investigation for any disease, against the first cell line models of several of these rare next generation models including angioimmunoblastic T-cell lymphoma and renal medullary carcinoma, leading to several novel drug repurposing hypotheses for rare cancers. Given these proof-of-concept studies, in partnership with the Rare Cancer Research Foundation, we launched an online matchmaking platform to connect patients with rare cancers to available research studies, facilitate online consent and provide biologistics support to enable fresh tissue donation to support cancer model generation from any clinical site in the United States. We will present results from this novel direct-to-patient approach to facilitate the generation of even larger numbers of next generation models from rare and pediatric cancers, propelling the generation of pre-clinical dependency maps of these tumors for the scientific community.

Citation Format: Yuen-Yi Tseng, Andrew Hong, Paula Keskula, Shubhroz Gill, Jaime Cheah, Grigoriy Kryukov, Aviad Tsherniak, Francisca Vazquez, Glenn Cowley, Sahar Alkhairy, Coyin Oh, Anson Peng, Rebecca Deasy, Abeer Sayeed, Peter Ronning, Samuel Ng, Steven Corsello, Corrie Painter, David Sandak, Levi Garraway, Mark Rubin, Calvin Kuo, Sidharth Puram, David Weinstock, Adam Bass, Nikhil Wagle, Keith Ligon, Katherine Janeway, David Root, Stuart Schreiber, Paul Clemons, Aly Shamji, Aly Shamji, William Hahn, Todd Golub, Jesse Boehm. Accelerating prediction of pediatric and rare cancer vulnerabilities using next-generation cancer models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 1953. doi:10.1158/1538-7445.AM2017-1953

The Angiosarcoma Project: Generating the genomic landscape of a rare cancer through a direct-to-patient initiative

1519

Background: Angiosarcoma (AS) is a rare soft tissue sarcoma, with an incidence of 300/yr and a 5-year DSS of 30%. The low incidence has impeded large-scale research efforts that may lead to improved clinical outcomes. To address this, we launched a nationwide study, which seeks to empower patients (pts) to accelerate research by sharing their samples and clinical information remotely. Methods: With pts and advocacy groups we developed a website to allow AS pts to participate across the US. Pts are mailed a saliva and blood draw kit for germline and cell free (cf) DNA analysis. We then obtain medical records and stored tumor samples. Whole exome sequencing will be performed on tumor, cfDNA and saliva samples. Transcriptome analysis will be performed on tumor samples. A clinically annotated genomic database will be generated and shared widely to identify genomic drivers and mechanisms of response and resistance to therapies. Study updates will be shared with pts regularly. Results: We conducted a 3-week pilot study to test the feasibility of enrolling geographically dispersed AS pts through a direct-to-patient (DTP) approach. Through social media, we identified 100+ pts willing to participate, 90 within the first day of outreach. We enrolled 15 pts from 10 states to test our ability to remotely obtain pt reported data, online consent, and samples. The average age of pts is 48, ranging 23-71 yrs. Primary locations of AS are breast 6 pts (40%), cardiac 4 pts (27%), scalp 2 pts (13%), liver 1 pt (6%), bladder 1 pt (6%), forehead 1 pt (6%). 9 pts (60%) reported being disease free, 4 pts (27%) reported having AS spread to lung, lymph, bone, and hip. Requests for medical records and tissue samples are underway, and initial saliva samples have been received. We are now opening this study to all AS pts in the USA. Conclusions: A DTP approach enabled rapid identification of an initial cohort of AS pts willing to share tumors, saliva, blood and medical records. We were able to obtain detailed clinical experiences and samples to perform genomic analysis. This study serves as proof of principle that DTP genomics efforts can democratize cancer research for exceedingly rare cancers, which to date have been disproportionately understudied.

The Metastatic Breast Cancer (MBC) project: Accelerating translational research through direct patient engagement

1076

Background: The Metastatic Breast Cancer Project is a nationwide research study, launched in Oct 2015 in collaboration with patients (pts) and advocacy groups, that directly engages pts through social media and seeks to empower them to share their experiences, clinical information, and samples to accelerate research. Methods: MBC pts enroll by providing their information at mbcproject.org. Pts are sent a saliva kit and asked to mail back a sample which is used to extract germline DNA. We contact pts medical providers and obtain medical records (MRs) and stored tumor samples. Pts may also submit a blood sample, used to extract cell free DNA (cfDNA). Whole exome sequencing (WES) is performed on tumor, germline, and cfDNA; transcriptome sequencing is performed on tumor. Clinical and genomic data are used to generate genomic landscapes in pt subgroups and to identify mechanisms of response and resistance to therapies. Data are shared widely through public databases. Pts receive regular study updates. Results: In 12 months, 2908 MBC pts from 50 states enrolled. 95% completed the 16-question survey about their cancer, treatments, and demographics. 1730 (60%) completed the online consent form. 100-200 pts continue to enroll monthly. To date, 1539 saliva kits were mailed and 1120 samples were received (73%). 992 unique treating institutions were reported by pts, including 733 institutions reported by only 1 pt each and 5 institutions reported by more than 40 pts each. We have obtained MRs from 253 patients (67% yield) and tumor samples from 85 pts (67% yield). WES was successfully completed for 79 tumors of 88 attempted (90%). WES has been performed on initial cfDNA samples. Conclusions: A direct-to-patient approach enabled rapid identification of thousands of MBC pts willing to share MRs, saliva, and tumor samples, including many with rare phenotypes. Remote acquisition of MRs, saliva, tumor, and blood for pts located throughout the US is feasible. We estimate that for ~33% of consenting patients, we can obtain medical records, saliva, and tumor tissue. Genomic analysis of tumor and cfDNA from subgroups including young pts, pts with extraordinary responses, and pts with de novo MBC will be presented.

The Prostate Cancer Project (PC Project): Translational genomics through direct patient engagement

199

Background: While there has been substantial advancement in the genomic understanding of prostate cancer (PCa), there is still much to be discovered. Additional progress is dependent upon obtaining a large amount of clinically annotated genomic data. As PCa is often treated in a community setting, where research samples are not collected, we are starting a direct-to-patient nationwide research initiative where patients can donate their medical records and biospecimens to accelerate research. Previously, we launched the metastatic breast cancer project (MBCproject; mbcproject.org) that leverages social media to engage the MBC community. Based on the initial success with this approach, we now aim to build out the PCproject. Methods: In collaboration with patients, we are developing a website to enable participation in the PCproject. Enrolled patients will be sent a saliva kit, used for germline DNA. We will also obtain medical records. Metastatic patients will also be sent a blood draw kit for circulating tumor DNA (ctDNA). Whole exome sequencing of the ctDNA will be performed. We will use the recruitment infrastructure, clinical record abstraction, and biospecimen processing developed for the MBC project. The data will be shared widely with the research community. Aggregate study results will be reported to patients. Results: In the first year of the MBCproject, 2912 MBC patients from all 50 states enrolled. 2766 (95.0%) completed the 16-question survey about their cancer, treatments, and demographic information. 1716 (58.9%) completed the online consent form permitting acquisition and analysis of medical records, tumor tissue, and saliva samples. 936 (68.8% success rate) saliva samples have been received. To date, we have obtained medical records from 155 patients (72.1% success rate) and tumor samples from 60 patients (72.3% success rate). Based on initial recruitment and surveys among PCa patients, we estimate that 500 patients will enroll in 2017. Conclusions: Based on experience from the MBC project, we will partner directly with patients to recruit and drive the PCproject forward. Remote acquisition of medical records, saliva samples, and tumor tissue for patients located throughout the US is feasible.

Abstract P1-05-13: The metastatic breast cancer project: Translational genomics through direct patient engagement

Background: The Metastatic Breast Cancer Project is a nationwide research initiative that directly engages patients through social media and advocacy groups and seeks to empower them to share their samples and clinical information to accelerate research. Because the vast majority of patients are treated in the community setting, we sought to determine the feasibility of remotely obtaining tumor and saliva samples as well as medical records from a large cohort of metastatic breast cancer (MBC) patients who receive their care in diverse settings around the country.

Methods: In collaboration with patients and advocacy groups, we developed a website to allow MBC patients across the U.S. to participate. Enrolled patients are sent a saliva kit and asked to mail back a saliva sample, which is used to extract germline DNA. We contact participants' medical providers and obtain medical records and part of their tumor biopsy. Whole exome and transcriptome sequencing is performed on tumor and germline samples. Clinically annotated genomic data are used to identify mechanisms of response and resistance to therapies. The database will be shared widely with researchers. Study updates and discoveries are shared with participants regularly.

Results: In the first 8 months, 2285 MBC patients from all 50 states enrolled. 2163 (95%) completed the 16-question survey about their cancer, treatments, and demographic information. 1232 completed the online consent form permitting acquisition and analysis of medical records, tumor tissue, and saliva samples. 556 saliva samples have been received. Initial medical record and tumor sample requests have been made for patients who have provided saliva samples. To date, we have obtained medical records from 102 patients (93% success rate) and tumor samples from 32 patients (77% success rate). Whole exome and transcriptome sequencing has been successfully completed on initial samples received and is ongoing for additional samples.

Conclusions: Partnering directly with patients through social media and advocacy groups enables rapid identification of thousands of patients willing to share tumors, saliva, and medical records to accelerate research. This approach allows for rapid identification of patients with rare phenotypes such as extraordinary responders, who have been challenging to identify with traditional approaches. Remote acquisition of medical records, saliva samples, and tumor tissue for patients located throughout the U.S. is feasible. Genomic analysis and medical record abstraction for these patients is underway. As data is generated, a clinically annotated database will be shared widely with the research community.

Citation Format: Painter CA, Anastasio E, Krevalin M, Kim D, Larken K, Lennon N, Frank E, Winer EP, Lander ES, Golub T. The metastatic breast cancer project: Translational genomics through direct patient engagement [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P1-05-13.

Abstract B43: Towards a Cancer Dependency Map

This abstract is being presented as a short talk in the scientific program. A full abstract is printed in the Proffered Abstracts section (PR02) of the Conference Proceedings.

Citation Format: Aviad Tsherniak, Francisca Vazquez, Barbara Weir, Philip Montgomery, Glenn Cowley, Stanley Gill, Gregory Kryukov, Sasha Pantel, Will Harrington, Mike Burger, Robin Meyers, Levi Ali, Amy Goodale, Yenarae Lee, Levi Garraway, Jesse Boehm, David Root, Todd Golub, William Hahn. Towards a Cancer Dependency Map. [abstract]. In: Proceedings of the AACR Precision Medicine Series: Targeting the Vulnerabilities of Cancer; May 16-19, 2016; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2017;23(1_Suppl):Abstract nr B43.

Abstract PR02: Towards a Cancer Dependency Map

The mapping of cancer genomes is rapidly approaching completion. The genomic information encoded by individual patients' tumors should, in principle, provide a guide for predicting acquired cancer dependencies. Unfortunately, while the success of precision cancer genomics hinges on the decoding of such dependencies, we lack the ability to predict dependencies for most individual tumors. The challenge stems from the absence of clinical data relating genotypes with dependencies since most cancer mutations are rare and our arsenal of cancer drugs is incomplete.

A comprehensive Cancer Dependency Map comprised of a catalog of genetic and small molecule vulnerabilities across a diverse set of cancers, along with robust statistical models able to predict these vulnerabilities from molecular and genomic features, would provide a roadmap of targets ripe for therapeutic development and would help reveal the mechanisms underlying the emergence of these vulnerabilities.

Here, we report progress in creating a Cancer Dependency Map consisting of the following components: 1) Systematic genetic perturbation (RNAi/CRISPR) of over 600 cancer cell models representing a wide range of human cancers and cell lineages using massively parallel genome scale loss-of-function screens. 2) Computational segregation of on- from off-target effects of RNAi enabling the discovery of outlier dependencies. 3) Predictive modeling to discover biomarkers for each dependency.

Our results demonstrate that our analytical approach (DEMETER) that models both gene and miRNA-based seed sequence effects effectively segregates on- from off-target effects of shRNAs. We discover 768 preferential dependencies whose suppression decreases viability at a level greater than six standard deviations in at least one of 503 cancer models and 105 such dependencies each present in at least 15 models. We find that 95% of the cancer models screened are strongly sensitive to the suppression of at least one of these dependencies, and that many models have common dependencies so that all models harbor at least one six-sigma dependency out of a set of only 76. Using a custom random forest based predictive modeling framework (ATLANTIS), we discover predictive biomarkers for hundreds of dependencies. These include known and novel vulnerabilities specified by somatic oncogenic alterations, overexpression of genes that specify lineage and differentiation, copy-number driven essentiality, and loss of functionally redundant paralogs.

These observations provide a rigorous computational and experimental foundation for the creation of a comprehensive Cancer Dependency Map. Subsampling and projection analyses suggest that over 10,000 genomically characterized cancer cell models will be needed to achieve this important goal.

This abstract is also being presented as Poster B43.

Citation Format: Aviad Tsherniak, Francisca Vazquez, Barbara Weir, Philip Montgomery, Glenn Cowley, Stanley Gill, Gregory Kryukov, Sasha Pantel, Will Harrington, Mike Burger, Robin Meyers, Levi Ali, Amy Goodale, Yenarae Lee, Levi Garraway, Jesse Boehm, David Root, Todd Golub, William Hahn. Towards a Cancer Dependency Map. [abstract]. In: Proceedings of the AACR Precision Medicine Series: Targeting the Vulnerabilities of Cancer; May 16-19, 2016; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2017;23(1_Suppl):Abstract nr PR02.

Abstract B44: Emerging targets from Cancer Dependency Map v0.1

Precision Cancer Medicine requires the identification of vulnerabilities linked to genetic features of tumors. Recent studies utilizing highly annotated small molecule collections to assess dependencies across hundreds of genomically annotated cell lines have demonstrated the potential for such large-scale preclinical “Dependency Map” projects. We have undertaken a complementary approach using genetic perturbation tools (RNAi and CRISPR-Cas9 based loss-of-function viability screens), to systematically catalog preferential genetic dependencies and markers that predict response. These efforts are providing a foundation for the discovery of novel targets poised for early therapeutic discovery projects together with patient populations that may be enriched for responders to such therapies.

Here, we present results from our initial Cancer Dependency Map consisting of RNAi loss-of-function screens across 503 cell lines, including both solid and hematopoietic tumors. We discovered 43 genes whose mutation or copy number creates a cancer dependency (oncogene addiction) including a novel dependency on the small GTPase, GNAI2 in Diffuse large B-cell Lymphoma. We discovered 142 genes in which elevated levels of expression create a dependency (gene addiction), a group of genes highly enriched for master regulator transcription factors such as SOX10, SPDEF, PAX8 and HNF1B. We discovered 474 genes for which hemizygous copy number creates a dependency (CYCLOPS genes), a group of genes highly enriched for members of macromolecular protein complexes including the spliceosome and proteasome. Finally, we discovered 171 genes that become a dependency when a redundant functional paralog is lost in cancer cells (redundant essentials). We demonstrate the mechanistic basis behind one such redundant essential dependency relationship in which promoter methylation of the UBB ubiquitin gene eliminates a compensatory mechanism leading to a novel vulnerability on the suppression of the UBC ubiquitin gene.

These observations begin to provide an initial census, categorization and prioritization of robust cancer dependencies and support the potential impact for expanding early efforts to develop dependency maps of cancer.

Citation Format: Francisca Vazquez, Aviad Tsherniak, Barbara Weir, Phil Montgomery, Glenn Cowley, Stanley Gill, Gregory Kryukov, Sasha Pantel, Will Harrington, Mike Burger, Robin Meyers, Levi Ali, Amy Goodale, Yenarae Lee, Levi Garraway, Jesse Boehm, David Root, Todd Golub, William Hahn. Emerging targets from Cancer Dependency Map v0.1. [abstract]. In: Proceedings of the AACR Precision Medicine Series: Targeting the Vulnerabilities of Cancer; May 16-19, 2016; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2017;23(1_Suppl):Abstract nr B44.

Abstract 4367: Accelerating prediction of tumor vulnerabilities using next-generation cancer models

The mapping of cancer genomes is rapidly approaching completion. The genomic information encoded by individual patients’ tumors should, in principle, provide a guide for predicting dependencies, but our ability to do so is suboptimal. The challenge stems from the absence of clinical data relating genotypes with dependencies since most cancer mutations are rare and our arsenal of cancer drugs is incomplete. If it was possible to build a preclinical ‘cancer dependency map’ at a scale that captured the genomic diversity of cancer (for instance, models of all genotypes tested for genetic and small-molecule dependencies), it should be feasible to improve dependency predictions. New technologies (e.g. CRISPR/Cas9 libraries) make such an effort now feasible. However, we lack a sufficient diversity of cancer models derived directly from patient samples to reflect the genetic diversity of cancer and the ability to systematically create functional data for each cancer patient to expand the map.

In an attempt to overcome these obstacles, we have established an industry-scale pipeline to generate new cancer models directly from patient samples, a “Cancer Cell Line Factory”. We have processed over 620 samples from 400 patients across 16 cancer types through this pipeline with a 25% success rate overall. To optimize conditions for each tumor type, we have systematically compared published cell line generation methods with standard approaches and captured all information with a data management system that will enhance the ability to predict optimal ex vivo propagation conditions for future samples. In all, we report the successful derivation of over 100 new genomically confirmed cancer and normal cell lines, including a series of unique pediatric cancer models derived from rare tumors.

We hypothesized that novel patient-derived cultures could be used to enhance dependency predictions. To test this hypothesis, we tested dependencies of 65 of these novel cultures against an identical set of 440 small molecules that were previously tested against 860 existing cancer cell lines. Our results suggest that dependency data generated with novel cell cultures is potentially backwards-compatible with existing small molecule dependency datasets. Finally, we demonstrate proof-of-concept that such new models can successfully used in CRISPR-Cas9 screens and integrate results with small molecule sensitivities to uncover CDK4 and XPO1 dependencies in a rare pediatric undifferentiated sarcoma. In aggregate, these proof-of-concept studies demarcate a path by which pre-clinical dependency maps may enhance clinical dependency predictions from genomic data alone.

Citation Format: Yuen-Yi Tseng, Andrew Hong, Paula Keskula, Shubhroz Gill, Jaime Cheah, Grigoriy Kryukov, Aviad Tsherniak, Francisca Vazquez, Glenn Cowley, Coyin Oh, Anson Peng, Abeer Sayeed, Rebecca Deasy, Peter Ronning, Philip Kantoff, Levi Garraway, Mark Rubin, Calvin Kuo, Sidharth Puram, Adi Gazdar, Filemon Dela Cruz, Adam Bass, Nikhil Wagle, Keith Ligon, Katherine Janeway, David Root, Stuart Schreiber, Paul Clemons, Aly Shamji, William Hahn, Todd Golub, Jesse S. Boehm. Accelerating prediction of tumor vulnerabilities using next-generation cancer models. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4367.

The Metastatic Breast Cancer Project: A national direct-to-patient initiative to accelerate genomics research

LBA1519

Background: The challenge in studying tumors from patients (pts) with metastatic breast cancer (MBC) has been that most tumors are not available for research, largely because most pts are cared for in community settings where genomics studies are not conducted. To address this, we launched a nationwide study, The Metastatic Breast Cancer Project, which seeks to empower patients to accelerate research by sharing their samples and clinical information. Methods: In collaboration with pts and advocacy groups, we developed a website to allow MBC pts to participate across the U.S. Enrolled pts are sent a saliva kit and asked to mail back a saliva sample, which is used to extract germline DNA. We contact participants’ medical providers and obtain medical records and part of their tumor biopsy. Whole exome and transcriptome sequencing is performed on tumor and germline. Clinically annotated genomic data are used to identify mechanisms of response and resistance to therapies. The database is shared widely with researchers. Study updates and discoveries are shared with participants regularly. Results: In the first 3 months, 1227 MBC pts enrolled. 1178 (96%) completed the 16-question survey about their cancer and treatments. Median age was 54 years (yrs) (range 25-91). Median time between initial diagnosis (dx) of breast cancer and MBC was 2 yrs; 424 pts were dx’d with de novo MBC. 1022 (87%) reported having a biopsy at or following their dx of MBC. Median time since MBC dx was 3 yrs; 87 reported having MBC >10 yrs. 436 (37%) reported being on a therapy for >2 yrs; 672 (57%) reported an “extraordinary response” to a therapy. For example, 77 reported long and/or extraordinary responses to capecitabine ; 44 to platinums, and 20 to everolimus. Initial medical records, saliva, and tumors have been received. Conclusions: A direct-to-patient approach enabled rapid identification of large numbers of MBC pts willing to share tumors, saliva, and medical records. This includes many with rare phenotypes, a group that has been challenging to identify with traditional approaches. Genomic analysis of pts with extraordinary responses and with de novo MBC are underway. Pt reported data has also identified unanticipated research questions.

Abstract PR04: High-throughput gene expression profiling as a generalizable assay for determination of mutation impact on gene function

Recently, the decline in the cost of genome sequencing has led to the rapid identification of thousands of cancer-associated somatic mutations. However, progress in characterization of these genetic events has lagged significantly behind. Understanding mutation function is critical not only for research purposes but also for determining targeted treatment strategies based on individual tumor genetic profiles, yet determination of mutation impact remains a significant bottleneck. Here we describe a high-throughput approach to classify somatic mutations that is robust, scalable, and requires no prior information of gene function.

We generated a lentiviral cDNA expression library of ~550 mutated and wild-type alleles of genes mutated in lung adenocarcinoma and introduced these alleles into four human lung cell lines. 96 hours post-infection, gene expression profiles were generated using Luminex-based L1000 profiling. In total, more than 2000 gene expression signatures were generated.

We discovered that gain-of-function mutants induce expression signatures with a greater signal strength or different identity than the corresponding wild-type gene signature. In contrast, loss-of-function mutants could be identified by their incapability to induce strong signatures. Based on these features of signature strength and signature identity, we developed a decision-tree approach to classify mutations as either dominant, loss-of-function, or likely inert. An orthogonal functional approach, an EGFR inhibitor resistance screen, was used as validation. The gene expression approach correctly classified known gain-of-function mutations in KRAS (13/13), EGFR (6/7), and ARAF (2/2) and identified dozens of never-characterized gain-of-function and loss-of-function missense mutations. In addition to rare, dominant mutations in clinically-actionable oncogenes such as PIK3CA and AKT1, we identified unexpected dominant mutations in the transcription factor MAX and the phosphatase subunit PPP2R1A, among others. We also observed a substantial enrichment of loss-of-function mutations in tumor suppressor genes such as STK11, KEAP1, FBXW7, and CASP8 as well as in genes not previously connected to lung adenocarcinoma, including GPR137B and MAPK7. Most genes assayed also harbored variants that are likely inert, further underscoring the importance of characterizing individual variant alleles.

The method developed here can, in principle, characterize any genetic variant, independent of prior knowledge of gene function, and should significantly advance the pace of functional characterization of mutations identified from genome sequencing.

Citation Format: Alice Berger, Angela Brooks, Xiaoyun Wu, Larson Hogstrom, Itay Tirosh, Federica Piccioni, Mukta Bagul, Cong Zhu, Yashaswi Shretha, David Root, Pablo Tamayo, Ryo Sakai, Bang Wong, Aravind Subramanian, Todd Golub, Matthew Meyerson, Jesse Boehm. High-throughput gene expression profiling as a generalizable assay for determination of mutation impact on gene function. [abstract]. In: Proceedings of the AACR Special Conference on Computational and Systems Biology of Cancer; Feb 8-11 2015; San Francisco, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(22 Suppl 2):Abstract nr PR04.

The genomic landscape of juvenile myelomonocytic leukemia

Juvenile myelomonocytic leukemia (JMML) is a myeloproliferative neoplasm (MPN) of childhood with a poor prognosis. Mutations in NF1, NRAS, KRAS, PTPN11 and CBL occur in 85% of patients, yet there are currently no risk stratification algorithms capable of predicting which patients will be refractory to conventional treatment and therefore be candidates for experimental therapies. In addition, there have been few other molecular pathways identified aside from the Ras/MAPK pathway to serve as the basis for such novel therapeutic strategies. We therefore sought to genomically characterize serial samples from patients at diagnosis through relapse and transformation to acute myeloid leukemia in order to expand our knowledge of the mutational spectrum in JMML. We identified recurrent mutations in genes involved in signal transduction, gene splicing, the polycomb repressive complex 2 (PRC2) and transcription. Importantly, the number of somatic alterations present at diagnosis appears to be the major determinant of outcome.

Abstract 957: Towards precision functional genomics via next-generation functional mapping of cancer variants

For the vast majority of cancer patients, existing knowledge of the function(s) of the mutant genes harbored by their tumor and the dependencies they induce is incomplete or non-existent since most cancer mutations are exceedingly rare. As a result, we now have long lists of candidate alleles but a paucity of targets whose biology is sufficiently well understood to guide therapeutics. Here we present an interim progress report on a pilot effort aiming to create a generalizable framework to systematically map the molecular consequences of cancer variants at scale (Target Accelerator). First, we created an efficient pipeline to generate cancer variants and generated an initial library of 1300 mutant cDNA clones corresponding to variants in lung cancer and diffuse large B-cell lymphoma as well as those nominated by “pan-cancer” computational analyses. Second, we established an industry-scale, next-generation pipeline to generate new cancer models (Cell Line Factory) directly from patient samples. We have leveraged this pipeline to process over 330 samples from 208 patients across 16 cancer types, with over 60% growing through at least 5 population doublings. We show that tumor genomics can be retained in such patient-derived models and that drug testing within 3 months is feasible. In addition, we use combinatorial molecular barcoding to rapidly generate a panel of pathway-primed human tumorigenesis models that are suitable for massively parallel multiplexed tumorigenesis assays in vivo (TumorPlex). We hypothesized that this integrated framework could be utilized to generate meaningful functional hypotheses in a high-throughput manner. To test this hypothesis, we introduced over 1000 cancer mutations into cell models and created gene expression signatures together with phenotypic data. In lung cancer, we show that the mutational impact of mutant alleles with known and unknown functions can be rapidly assessed by comparing signatures of wild-type and mutant alleles. We show that this generalizable approach, which does not require prior knowledge, can place variants of unknown significance into dominant gain-of-function and loss-of-function categories. As a complementary approach, we have used TumorPlex assays to test the tumorigenic potential of 550 mutant alleles nominated by Pan-Cancer computational analyses and discovered unexpected variants in the KRAS, AKT1, MAP2K1, ERBB2, PIK3CB, NFE2L2, FAM200A and POT1 genes as being potently tumorigenic. These proof-of-concept studies demonstrate initial feasibility of mapping cancer variant function at scale. Importantly, they demarcate a path by which mapping variant function and predicting vulnerabilities might soon be possible on a patient-by-patient basis, achieving the promise of precision functional genomics.

Citation Format: Alice Berger, Eejung Kim, Angela Brooks, Yashaswi Shrestha, Yuen-Yi Tseng, Xiaoyun Wu, Nina Ilic, Lihua Zou, Atanas Kamburov, Xiaoping Yang, Cong Zhu, Paula Keskula, Sara Seepo, Andrew Hong, John Doench, Aravind Subramanian, Keith Ligon, Philip Kantoff, Katherine Janeway, Levi Garraway, David Root, Todd Golub, Matthew Meyerson, William Hahn, Gad Getz, Jesse Boehm. Towards precision functional genomics via next-generation functional mapping of cancer variants. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 957. doi:10.1158/1538-7445.AM2015-957

Expression profiles of 151 pediatric low-grade gliomas reveal molecular differences associated with location and histological subtype

BACKGROUND

Pediatric low-grade gliomas (PLGGs), the most frequent pediatric brain tumor, comprise a heterogeneous group of diseases. Recent genomic analyses suggest that these tumors are mostly driven by mitogene-activated protein kinase (MAPK) pathway alterations. However, little is known about the molecular characteristics inherent to their clinical and histological heterogeneity.

METHODS

We performed gene expression profiling on 151 paraffin-embedded PLGGs from different locations, ages, and histologies. Using unsupervised and supervised analyses, we compared molecular features with age, location, histology, and BRAF genomic status. We compared molecular differences with normal pediatric brain expression profiles to observe whether those patterns were mirrored in normal brain.

RESULTS

Unsupervised clustering distinguished 3 molecular groups that correlated with location in the brain and histological subtype. "Not otherwise specified" (NOS) tumors did not constitute a unified class. Supratentorial pilocytic astrocytomas (PAs) were significantly enriched with genes involved in pathways related to inflammatory activity compared with infratentorial tumors. Differences based on tumor location were not mirrored in location-dependent differences in expression within normal brain tissue. We identified significant differences between supratentorial PAs and diffuse astrocytomas as well as between supratentorial PAs and dysembryoplastic neuroepithelial tumors but not between supratentorial PAs and gangliogliomas. Similar expression patterns were observed between childhood and adolescent PAs. We identified differences between BRAF-duplicated and V600E-mutated tumors but not between primary and recurrent PLGGs.

CONCLUSION

Expression profiling of PLGGs reveals significant differences associated with tumor location, histology, and BRAF genomic status. Supratentorial PAs, in particular, are enriched in inflammatory pathways that appear to be tumor-related.

Abstract A41: Complementary genomic approaches highlight the PI3K/mTOR pathway as a common vulnerability in osteosarcoma

Osteosarcoma is the most common primary bone tumor and yet there have been no substantial advances in treatment or survival in over 2 decades. We examined 59 tumor/normal pairs by whole-genome, whole-exome and RNA-Sequencing. Only TP53 was mutated at significant frequency across the 59 samples. The mean non-silent somatic mutation rate was 1.2 mutation per megabase and there were a median of 231 somatic rearrangements per tumor. Complex chains of rearrangements and localized hypermutation were detected in almost all cases. Given the inter-tumor heterogeneity, the extent of genomic instability and the difficulty in acquiring a large sample size in a rare tumor we used several methods to identify genomic events contributing to osteosarcoma proliferation and survival. Pathway analysis, a heuristic analytic algorithm, a comparative oncology approach and a genome-wide, pooled short hairpin RNA (shRNA) screen all point to the PI3K/mTOR pathway as a potential central vulnerability for therapeutic exploitation in osteosarcoma. Osteosarcoma cell lines are responsive to pharmacologic and genetic inhibition of the PI3K/mTOR pathway both in vitro and in vivo.

Citation Format: Adam Kiezun, Jennifer Perry, Peter Tonzi, Eliezer Van Allen, Scott L. Carter, Sylvan Baca, Ami Bhatt, Michael Lawrence, Loren Walensky, Nikhil Wagle, Jaume Mora, Carmen deTorres, Cinzia Lavarino, Liliana Velasco-Hidalgo, Rocio Cardenas-Cardos, Simone dos Santos Aguiar, Jose A. Yunes, Gabriela Mercado, Jorge Melendez-Zajgla, Charles Roberts, Levi Garraway, Carlos Rodriguez-Galindo, Todd Golub, Stuart Orkin, Gad Getz, Katherine Janeway. Complementary genomic approaches highlight the PI3K/mTOR pathway as a common vulnerability in osteosarcoma. [abstract]. In: Proceedings of the AACR Special Conference on Pediatric Cancer at the Crossroads: Translating Discovery into Improved Outcomes; Nov 3-6, 2013; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2013;74(20 Suppl):Abstract nr A41.

Abstract 999: The genomic landscape of pediatric Ewing sarcoma

The sequencing of aggressive pediatric solid tumors is revealing remarkably stable genomes. In the cases of malignant rhabdoid and retinoblastoma, there is a paucity of recurrently mutated genes, and oncogenesis appears to be driven, at least in part, by epigenetic deregulation. It has been suggested that pediatric tumors characterized by oncogenic fusions will exhibit relatively few additional somatic driver aberrancies. Ewing sarcoma, the second most common pediatric bone tumor, is characterized by rearrangements of the EWS gene and ETS-family transcription factor genes, most commonly FLI and ERG. In experimental models, Ewing sarcoma demonstrates dependency on the expression of the resulting chimeric fusion products. As such, Ewing sarcoma represents a paradigm for studying the genomic landscape of fusion-driven cancers. To this end, we performed whole-exome sequencing of 96 Ewing sarcoma tumors and 11 Ewing sarcoma cell lines, as well as whole-genome sequencing, transcriptome sequencing, and copy-number analysis of a subset of these samples. We found that Ewing sarcoma is one of the most genetically normal cancers sequenced to date, but that treatment, which generally employs genotoxic chemotherapy and radiation, is associated with an increase in mutation rate and single nucleotide substitutions associated with DNA damage. There was a marked absence of recurrent mutations in immediately druggable targets, such as tyrosine kinases, calling into question the feasibility of utilizing tumor sequencing to nominate targeted therapies for patients with Ewing sarcoma. Rather, these results highlight the importance of directly targeting the EWS/ETS fusion events or identifying synthetic lethal dependencies. To this end, we clarified a number of outstanding questions regarding the EWS/ETS fusions. We found that reciprocal ETS/EWS fusions are not expressed in Ewing sarcoma and therefore unlikely to play a role in Ewing pathogenesis as is seen with reciprocal fusions of PML-RARα in acute promyelocytic leukemia. We also found that wild-type FLI and wild-type ERG are not expressed in Ewing sarcoma tumors. However, there appears to be a role for ETS gene deregulation in this disease beyond the expression of EWS/ETS fusion proteins because we found recurrent somatic events in ERF and ETS1. We also identified a small number of other recurrently mutated genes that likely collaborate with EWS/ETS fusions in a minority of cases and confirmed that loss of STAG2 occurs in approximately 15% of Ewing sarcoma tumors. Thus, massively parallel sequencing of a large collection of Ewing sarcoma tumors supports the notion that fusion-driven pediatric malignancies bear quiet genomes, underscores the importance of identifying new treatment approaches targeting EWS/ETS fusions, and also identifies new genetic abnormalities that warrant further biological validation.

Citation Format: Brian Crompton, Chip Stewart, Amaro Taylor-Weiner, Gabriela Alexa, Kyle Kurek, Monica Calicchio, Adam Kiezun, Scott Carter, Sachet Shukla, Swapnil Mehta, Aaron Thorner, Carmen de Torres, Cinzia Lavarino, Mariona Sunol, Aaron McKenna, Andrey Sivachenko, Kristian Cibulskis, Michael Lawrence, Lauren Ambrogio, Daniel Auclair, Ivan Rosshandler, Angela Schwarz-Cruz y Celis, Miguel Rivera, Carlos Rodriguez-Galindo, Mark Fleming, Todd Golub, Gad Getz, Jaume Mora, Kimberly Stegmaier. The genomic landscape of pediatric Ewing sarcoma. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 999. doi:10.1158/1538-7445.AM2014-999

CXCR4 pathway associated with family history of melanoma

PURPOSE

Genetic predisposition plays a major role in the etiology of melanoma, but known genetic markers only account for a limited fraction of family-history-associated melanoma cases. Expression microarrays have offered the opportunity to identify further genomic profiles correlated with family history of melanoma. We aimed to distinguish mRNA expression signatures between melanoma cases with and without a family history of melanoma.

METHODS

Based on the Nurses' Health Study, family history was defined as having one or more first-degree family members diagnosed with melanoma. Melanoma diagnosis was confirmed by reviewing pathology reports, and tumor blocks were collected by mail from across the USA. Genomic interrogation was accomplished through evaluating expression profiling of formalin-fixed paraffin-embedded tissues from 78 primary cutaneous invasive melanoma cases, on either a 6K or whole-genome (24K) Illumina gene chip. Gene set enrichment analysis was performed for each batch to determine the differentially enriched pathways and key contributing genes.

RESULTS

The CXC chemokine receptor 4 (CXCR4) pathway was consistently up-regulated within cases of familial melanoma in both platforms. Leading edge analysis showed four genes from the CXCR4 pathway, including MAPK1, PLCG1, CRK, and PTK2, were among the core members that contributed to the enrichment of this pathway. There was no association between the enrichment of CXCR4 pathway and NRAS, BRAF mutation, or Breslow thickness of the primary melanoma cases.

CONCLUSIONS

We found that the CXCR4 pathway might constitute a novel susceptibility pathway associated with family history of melanoma in first-degree relatives.