Defining a Cancer Dependency Map.

Most human epithelial tumors harbor numerous alterations, making it difficult to predict which genes are required for tumor survival. To systematically identify cancer dependencies, we analyzed 501 genome-scale loss-of-function screens performed in diverse human cancer cell lines. We developed DEMETER, an analytical framework that segregates on- from off-target effects of RNAi. 769 genes were differentially required in subsets of these cell lines at a threshold of six SDs from the mean. We found predictive models for 426 dependencies (55%) by nonlinear regression modeling considering 66,646 molecular features. Many dependencies fall into a limited number of classes, and unexpectedly, in 82% of models, the top biomarkers were expression based. We demonstrated the basis behind one such predictive model linking hypermethylation of the UBB ubiquitin gene to a dependency on UBC. Together, these observations provide a foundation for a cancer dependency map that facilitates the prioritization of therapeutic targets.

ß-Catenin-driven cancers require a YAP1 transcriptional complex for survival and tumorigenesis.

Wnt/ß-catenin signaling plays a key role in the pathogenesis of colon and other cancers; emerging evidence indicates that oncogenic ß-catenin regulates several biological processes essential for cancer initiation and progression. To decipher the role of ß-catenin in transformation, we classified ß-catenin activity in 85 cancer cell lines in which we performed genome-scale loss-of-function screens and found that ß-catenin active cancers are dependent on a signaling pathway involving the transcriptional regulator YAP1. Specifically, we found that YAP1 and the transcription factor TBX5 form a complex with ß-catenin. Phosphorylation of YAP1 by the tyrosine kinase YES1 leads to localization of this complex to the promoters of antiapoptotic genes, including BCL2L1 and BIRC5. A small-molecule inhibitor of YES1 impeded the proliferation of ß-catenin-dependent cancers in both cell lines and animal models. These observations define a ß-catenin-YAP1-TBX5 complex essential to the transformation and survival of ß-catenin-driven cancers.

Cancer vulnerabilities unveiled by genomic loss.

Due to genome instability, most cancers exhibit loss of regions containing tumor suppressor genes and collateral loss of other genes. To identify cancer-specific vulnerabilities that are the result of copy number losses, we performed integrated analyses of genome-wide copy number and RNAi profiles and identified 56 genes for which gene suppression specifically inhibited the proliferation of cells harboring partial copy number loss of that gene. These CYCLOPS (copy number alterations yielding cancer liabilities owing to partial loss) genes are enriched for spliceosome, proteasome, and ribosome components. One CYCLOPS gene, PSMC2, encodes an essential member of the 19S proteasome. Normal cells express excess PSMC2, which resides in a complex with PSMC1, PSMD2, and PSMD5 and acts as a reservoir protecting cells from PSMC2 suppression. Cells harboring partial PSMC2 copy number loss lack this complex and die after PSMC2 suppression. These observations define a distinct class of cancer-specific liabilities resulting from genome instability.

WRN helicase is a synthetic lethal target in microsatellite unstable cancers.

Synthetic lethality-an interaction between two genetic events through which the co-occurrence of these two genetic events leads to cell death, but each event alone does not-can be exploited for cancer therapeutics1. DNA repair processes represent attractive synthetic lethal targets, because many cancers exhibit an impairment of a DNA repair pathway, which can lead to dependence on specific repair proteins2. The success of poly(ADP-ribose) polymerase 1 (PARP-1) inhibitors in cancers with deficiencies in homologous recombination highlights the potential of this approach3. Hypothesizing that other DNA repair defects would give rise to synthetic lethal relationships, we queried dependencies in cancers with microsatellite instability (MSI), which results from deficient DNA mismatch repair. Here we analysed data from large-scale silencing screens using CRISPR-Cas9-mediated knockout and RNA interference, and found that the RecQ DNA helicase WRN was selectively essential in MSI models in vitro and in vivo, yet dispensable in models of cancers that are microsatellite stable. Depletion of WRN induced double-stranded DNA breaks and promoted apoptosis and cell cycle arrest selectively in MSI models. MSI cancer models required the helicase activity of WRN, but not its exonuclease activity. These findings show that WRN is a synthetic lethal vulnerability and promising drug target for MSI cancers.

Genetic and transcriptional evolution alters cancer cell line drug response.

Human cancer cell lines are the workhorse of cancer research. Although cell lines are known to evolve in culture, the extent of the resultant genetic and transcriptional heterogeneity and its functional consequences remain understudied. Here we use genomic analyses of 106 human cell lines grown in two laboratories to show extensive clonal diversity. Further comprehensive genomic characterization of 27 strains of the common breast cancer cell line MCF7 uncovered rapid genetic diversification. Similar results were obtained with multiple strains of 13 additional cell lines. Notably, genetic changes were associated with differential activation of gene expression programs and marked differences in cell morphology and proliferation. Barcoding experiments showed that cell line evolution occurs as a result of positive clonal selection that is highly sensitive to culture conditions. Analyses of single-cell-derived clones demonstrated that continuous instability quickly translates into heterogeneity of the cell line. When the 27 MCF7 strains were tested against 321 anti-cancer compounds, we uncovered considerably different drug responses: at least 75% of compounds that strongly inhibited some strains were completely inactive in others. This study documents the extent, origins and consequences of genetic variation within cell lines, and provides a framework for researchers to measure such variation in efforts to support maximally reproducible cancer research.

GeNets: a unified web platform for network-based genomic analyses.

Functional genomics networks are widely used to identify unexpected pathway relationships in large genomic datasets. However, it is challenging to compare the signal-to-noise ratios of different networks and to identify the optimal network with which to interpret a particular genetic dataset. We present GeNets, a platform in which users can train a machine-learning model (Quack) to carry out these comparisons and execute, store, and share analyses of genetic and RNA-sequencing datasets.

Author Correction: Mitochondrial metabolism promotes adaptation to proteotoxic stress.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Mitochondrial metabolism promotes adaptation to proteotoxic stress.

The mechanisms by which cells adapt to proteotoxic stress are largely unknown, but are key to understanding how tumor cells, particularly in vivo, are largely resistant to proteasome inhibitors. Analysis of cancer cell lines, mouse xenografts and patient-derived tumor samples all showed an association between mitochondrial metabolism and proteasome inhibitor sensitivity. When cells were forced to use oxidative phosphorylation rather than glycolysis, they became proteasome-inhibitor resistant. This mitochondrial state, however, creates a unique vulnerability: sensitivity to the small molecule compound elesclomol. Genome-wide CRISPR-Cas9 screening showed that a single gene, encoding the mitochondrial reductase FDX1, could rescue elesclomol-induced cell death. Enzymatic function and nuclear-magnetic-resonance-based analyses further showed that FDX1 is the direct target of elesclomol, which promotes a unique form of copper-dependent cell death. These studies explain a fundamental mechanism by which cells adapt to proteotoxic stress and suggest strategies to mitigate proteasome inhibitor resistance.

PIK3CA mutant tumors depend on oxoglutarate dehydrogenase.

Oncogenic PIK3CA mutations are found in a significant fraction of human cancers, but therapeutic inhibition of PI3K has only shown limited success in clinical trials. To understand how mutant PIK3CA contributes to cancer cell proliferation, we used genome scale loss-of-function screening in a large number of genomically annotated cancer cell lines. As expected, we found that PIK3CA mutant cancer cells require PIK3CA but also require the expression of the TCA cycle enzyme 2-oxoglutarate dehydrogenase (OGDH). To understand the relationship between oncogenic PIK3CA and OGDH function, we interrogated metabolic requirements and found an increased reliance on glucose metabolism to sustain PIK3CA mutant cell proliferation. Functional metabolic studies revealed that OGDH suppression increased levels of the metabolite 2-oxoglutarate (2OG). We found that this increase in 2OG levels, either by OGDH suppression or exogenous 2OG treatment, resulted in aspartate depletion that was specifically manifested as auxotrophy within PIK3CA mutant cells. Reduced levels of aspartate deregulated the malate-aspartate shuttle, which is important for cytoplasmic NAD+ regeneration that sustains rapid glucose breakdown through glycolysis. Consequently, because PIK3CA mutant cells exhibit a profound reliance on glucose metabolism, malate-aspartate shuttle deregulation leads to a specific proliferative block due to the inability to maintain NAD+/NADH homeostasis. Together these observations define a precise metabolic vulnerability imposed by a recurrently mutated oncogene.

Identification of cancer-cytotoxic modulators of PDE3A by predictive chemogenomics.

High cancer death rates indicate the need for new anticancer therapeutic agents. Approaches to discovering new cancer drugs include target-based drug discovery and phenotypic screening. Here, we identified phosphodiesterase 3A modulators as cell-selective cancer cytotoxic compounds through phenotypic compound library screening and target deconvolution by predictive chemogenomics. We found that sensitivity to 6-(4-(diethylamino)-3-nitrophenyl)-5-methyl-4,5-dihydropyridazin-3(2H)-one, or DNMDP, across 766 cancer cell lines correlates with expression of the gene PDE3A, encoding phosphodiesterase 3A. Like DNMDP, a subset of known PDE3A inhibitors kill selected cancer cells, whereas others do not. Furthermore, PDE3A depletion leads to DNMDP resistance. We demonstrated that DNMDP binding to PDE3A promotes an interaction between PDE3A and Schlafen 12 (SLFN12), suggestive of a neomorphic activity. Coexpression of SLFN12 with PDE3A correlates with DNMDP sensitivity, whereas depletion of SLFN12 results in decreased DNMDP sensitivity. Our results implicate PDE3A modulators as candidate cancer therapeutic agents and demonstrate the power of predictive chemogenomics in small-molecule discovery.

ATARiS: computational quantification of gene suppression phenotypes from multisample RNAi screens.

Genome-scale RNAi libraries enable the systematic interrogation of gene function. However, the interpretation of RNAi screens is complicated by the observation that RNAi reagents designed to suppress the mRNA transcripts of the same gene often produce a spectrum of phenotypic outcomes due to differential on-target gene suppression or perturbation of off-target transcripts. Here we present a computational method, Analytic Technique for Assessment of RNAi by Similarity (ATARiS), that takes advantage of patterns in RNAi data across multiple samples in order to enrich for RNAi reagents whose phenotypic effects relate to suppression of their intended targets. By summarizing only such reagent effects for each gene, ATARiS produces quantitative, gene-level phenotype values, which provide an intuitive measure of the effect of gene suppression in each sample. This method is robust for data sets that contain as few as 10 samples and can be used to analyze screens of any number of targeted genes. We used this analytic approach to interrogate RNAi data derived from screening more than 100 human cancer cell lines and identified HNF1B as a transforming oncogene required for the survival of cancer cells that harbor HNF1B amplifications. ATARiS is publicly available at http://broadinstitute.org/ataris.

Systematic investigation of genetic vulnerabilities across cancer cell lines reveals lineage-specific dependencies in ovarian cancer.

A comprehensive understanding of the molecular vulnerabilities of every type of cancer will provide a powerful roadmap to guide therapeutic approaches. Efforts such as The Cancer Genome Atlas Project will identify genes with aberrant copy number, sequence, or expression in various cancer types, providing a survey of the genes that may have a causal role in cancer. A complementary approach is to perform systematic loss-of-function studies to identify essential genes in particular cancer cell types. We have begun a systematic effort, termed Project Achilles, aimed at identifying genetic vulnerabilities across large numbers of cancer cell lines. Here, we report the assessment of the essentiality of 11,194 genes in 102 human cancer cell lines. We show that the integration of these functional data with information derived from surveying cancer genomes pinpoints known and previously undescribed lineage-specific dependencies across a wide spectrum of cancers. In particular, we found 54 genes that are specifically essential for the proliferation and viability of ovarian cancer cells and also amplified in primary tumors or differentially overexpressed in ovarian cancer cell lines. One such gene, PAX8, is focally amplified in 16% of high-grade serous ovarian cancers and expressed at higher levels in ovarian tumors. Suppression of PAX8 selectively induces apoptotic cell death of ovarian cancer cells. These results identify PAX8 as an ovarian lineage-specific dependency. More generally, these observations demonstrate that the integration of genome-scale functional and structural studies provides an efficient path to identify dependencies of specific cancer types on particular genes and pathways.

Partial gene suppression improves identification of cancer vulnerabilities when CRISPR-Cas9 knockout is pan-lethal.

BACKGROUND: Hundreds of functional genomic screens have been performed across a diverse set of cancer contexts, as part of efforts such as the Cancer Dependency Map, to identify gene dependencies-genes whose loss of function reduces cell viability or fitness. Recently, large-scale screening efforts have shifted from RNAi to CRISPR-Cas9, due to superior efficacy and specificity. However, many effective oncology drugs only partially inhibit their protein targets, leading us to question whether partial suppression of genes using RNAi could reveal cancer vulnerabilities that are missed by complete knockout using CRISPR-Cas9. Here, we compare CRISPR-Cas9 and RNAi dependency profiles of genes across approximately 400 matched cancer cell lines. RESULTS: We find that CRISPR screens accurately identify more gene dependencies per cell line, but the majority of each cell line's dependencies are part of a set of 1867 genes that are shared dependencies across the entire collection (pan-lethals). While RNAi knockdown of about 30% of these genes is also pan-lethal, approximately 50% have selective dependency patterns across cell lines, suggesting they could still be cancer vulnerabilities. The accuracy of the unique RNAi selectivity is supported by associations to multi-omics profiles, drug sensitivity, and other expected co-dependencies. CONCLUSIONS: Incorporating RNAi data for genes that are pan-lethal knockouts facilitates the discovery of a wider range of gene targets than could be detected using the CRISPR dataset alone. This can aid in the interpretation of contrasting results obtained from CRISPR and RNAi screens and reinforce the importance of partial gene suppression methods in building a cancer dependency map.

Genome-scale functional genomics identify genes preferentially essential for multiple myeloma cells compared to other neoplasias.

Clinical progress in multiple myeloma (MM), an incurable plasma cell (PC) neoplasia, has been driven by therapies that have limited applications beyond MM/PC neoplasias and do not target specific oncogenic mutations in MM. Instead, these agents target pathways critical for PC biology yet largely dispensable for malignant or normal cells of most other lineages. Here we systematically characterized the lineage-preferential molecular dependencies of MM through genome-scale clustered regularly interspaced short palindromic repeats (CRISPR) studies in 19 MM versus hundreds of non-MM lines and identified 116 genes whose disruption more significantly affects MM cell fitness compared with other malignancies. These genes, some known, others not previously linked to MM, encode transcription factors, chromatin modifiers, endoplasmic reticulum components, metabolic regulators or signaling molecules. Most of these genes are not among the top amplified, overexpressed or mutated in MM. Functional genomics approaches thus define new therapeutic targets in MM not readily identifiable by standard genomic, transcriptional or epigenetic profiling analyses.

Sparse dictionary learning recovers pleiotropy from human cell fitness screens.

In high-throughput functional genomic screens, each gene product is commonly assumed to exhibit a singular biological function within a defined protein complex or pathway. In practice, a single gene perturbation may induce multiple cascading functional outcomes, a genetic principle known as pleiotropy. Here, we model pleiotropy in fitness screen collections by representing each gene perturbation as the sum of multiple perturbations of biological functions, each harboring independent fitness effects inferred empirically from the data. Our approach (Webster) recovered pleiotropic functions for DNA damage proteins from genotoxic fitness screens, untangled distinct signaling pathways upstream of shared effector proteins from cancer cell fitness screens, and predicted the stoichiometry of an unknown protein complex subunit from fitness data alone. Modeling compound sensitivity profiles in terms of genetic functions recovered compound mechanisms of action. Our approach establishes a sparse approximation mechanism for unraveling complex genetic architectures underlying high-dimensional gene perturbation readouts.

Global computational alignment of tumor and cell line transcriptional profiles.

Cell lines are key tools for preclinical cancer research, but it remains unclear how well they represent patient tumor samples. Direct comparisons of tumor and cell line transcriptional profiles are complicated by several factors, including the variable presence of normal cells in tumor samples. We thus develop an unsupervised alignment method (Celligner) and apply it to integrate several large-scale cell line and tumor RNA-Seq datasets. Although our method aligns the majority of cell lines with tumor samples of the same cancer type, it also reveals large differences in tumor similarity across cell lines. Using this approach, we identify several hundred cell lines from diverse lineages that present a more mesenchymal and undifferentiated transcriptional state and that exhibit distinct chemical and genetic dependencies. Celligner could be used to guide the selection of cell lines that more closely resemble patient tumors and improve the clinical translation of insights gained from cell lines.

Chronos: a cell population dynamics model of CRISPR experiments that improves inference of gene fitness effects.

CRISPR loss of function screens are powerful tools to interrogate biology but exhibit a number of biases and artifacts that can confound the results. Here, we introduce Chronos, an algorithm for inferring gene knockout fitness effects based on an explicit model of cell proliferation dynamics after CRISPR gene knockout. We test Chronos on two pan-cancer CRISPR datasets and one longitudinal CRISPR screen. Chronos generally outperforms competitors in separation of controls and strength of biomarker associations, particularly when longitudinal data is available. Additionally, Chronos exhibits the lowest copy number and screen quality bias of evaluated methods. Chronos is available at https://github.com/broadinstitute/chronos .

Integrated cross-study datasets of genetic dependencies in cancer.

CRISPR-Cas9 viability screens are increasingly performed at a genome-wide scale across large panels of cell lines to identify new therapeutic targets for precision cancer therapy. Integrating the datasets resulting from these studies is necessary to adequately represent the heterogeneity of human cancers and to assemble a comprehensive map of cancer genetic vulnerabilities. Here, we integrated the two largest public independent CRISPR-Cas9 screens performed to date (at the Broad and Sanger institutes) by assessing, comparing, and selecting methods for correcting biases due to heterogeneous single-guide RNA efficiency, gene-independent responses to CRISPR-Cas9 targeting originated from copy number alterations, and experimental batch effects. Our integrated datasets recapitulate findings from the individual datasets, provide greater statistical power to cancer- and subtype-specific analyses, unveil additional biomarkers of gene dependency, and improve the detection of common essential genes. We provide the largest integrated resources of CRISPR-Cas9 screens to date and the basis for harmonizing existing and future functional genetics datasets.

An Embryonic Diapause-like Adaptation with Suppressed Myc Activity Enables Tumor Treatment Persistence.

Treatment-persistent residual tumors impede curative cancer therapy. To understand this cancer cell state we generated models of treatment persistence that simulate the residual tumors. We observe that treatment-persistent tumor cells in organoids, xenografts, and cancer patients adopt a distinct and reversible transcriptional program resembling that of embryonic diapause, a dormant stage of suspended development triggered by stress and associated with suppressed Myc activity and overall biosynthesis. In cancer cells, depleting Myc or inhibiting Brd4, a Myc transcriptional co-activator, attenuates drug cytotoxicity through a dormant diapause-like adaptation with reduced apoptotic priming. Conversely, inducible Myc upregulation enhances acute chemotherapeutic activity. Maintaining residual cells in dormancy after chemotherapy by inhibiting Myc activity or interfering with the diapause-like adaptation by inhibiting cyclin-dependent kinase 9 represent potential therapeutic strategies against chemotherapy-persistent tumor cells. Our study demonstrates that cancer co-opts a mechanism similar to diapause with adaptive inactivation of Myc to persist during treatment.

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.