Programming tumor evolution with selection gene drives to proactively combat drug resistance.

Most targeted anticancer therapies fail due to drug resistance evolution. Here we show that tumor evolution can be reproducibly redirected to engineer therapeutic opportunity, regardless of the exact ensemble of pre-existing genetic heterogeneity. We develop a selection gene drive system that is stably introduced into cancer cells and is composed of two genes, or switches, that couple an inducible fitness advantage with a shared fitness cost. Using stochastic models of evolutionary dynamics, we identify the design criteria for selection gene drives. We then build prototypes that harness the selective pressure of multiple approved tyrosine kinase inhibitors and employ therapeutic mechanisms as diverse as prodrug catalysis and immune activity induction. We show that selection gene drives can eradicate diverse forms of genetic resistance in vitro. Finally, we demonstrate that model-informed switch engagement effectively targets pre-existing resistance in mouse models of solid tumors. These results establish selection gene drives as a powerful framework for evolution-guided anticancer therapy.

Excessive concentrations of kinase inhibitors in translational studies impede effective drug repurposing.

Drug repositioning seeks to leverage existing clinical knowledge to identify alternative clinical settings for approved drugs. However, repositioning efforts fail to demonstrate improved success rates in late-stage clinical trials. Focusing on 11 approved kinase inhibitors that have been evaluated in 139 repositioning hypotheses, we use data mining to characterize the state of clinical repurposing. Then, using a simple experimental correction with human serum proteins in in vitro pharmacodynamic assays, we develop a measurement of a drug's effective exposure. We show that this metric is remarkably predictive of clinical activity for a panel of five kinase inhibitors across 23 drug variant targets in leukemia. We then validate our model's performance in six other kinase inhibitors for two types of solid tumors: non-small cell lung cancer (NSCLC) and gastrointestinal stromal tumors (GISTs). Our approach presents a straightforward strategy to use existing clinical information and experimental systems to decrease the clinical failure rate in drug repurposing studies.

Functional genomic analysis of adult and pediatric brain tumor isolates.

Background: Adult and pediatric tumors display stark differences in their mutation spectra and chromosome alterations. Here, we attempted to identify common and unique gene dependencies and their associated biomarkers among adult and pediatric tumor isolates using functional genetic lethal screens and computational modeling. Methods: We performed CRISRP-Cas9 lethality screens in two adult glioblastoma (GBM) tumor isolates and five pediatric brain tumor isolates representing atypical teratoid rhabdoid tumors (ATRT), diffuse intrinsic pontine glioma, GBM, and medulloblastoma. We then integrated the screen results with machine learning-based gene-dependency models generated from data from >900 cancer cell lines. Results: We found that >50% of candidate dependencies of 280 identified were shared between adult GBM tumors and individual pediatric tumor isolates. 68% of screen hits were found as nodes in our network models, along with shared and tumor-specific predictors of gene dependencies. We investigated network predictors associated with ADAR, EFR3A, FGFR1 (pediatric-specific), and SMARCC2 (ATRT-specific) gene dependency among our tumor isolates. Conclusions: The results suggest that, despite harboring disparate genomic signatures, adult and pediatric tumor isolates share a preponderance of genetic dependences. Further, combining data from primary brain tumor lethality screens with large cancer cell line datasets produced valuable insights into biomarkers of gene dependency, even for rare cancers. Importance of the Study: Our results demonstrate that large cancer cell lines data sets can be computationally mined to identify known and novel gene dependency relationships in adult and pediatric human brain tumor isolates. Gene dependency networks and lethality screen results represent a key resource for neuro-oncology and cancer research communities. We also highlight some of the challenges and limitations of this approach.

Agent-Based Models Help Interpret Patterns of Clinical Drug Resistance by Contextualizing Competition Between Distinct Drug Failure Modes.

Introduction: Modern targeted cancer therapies are carefully crafted small molecules. These exquisite technologies exhibit an astonishing diversity of observed failure modes (drug resistance mechanisms) in the clinic. This diversity is surprising because back of the envelope calculations and classic modeling results in evolutionary dynamics suggest that the diversity in the modes of clinical drug resistance should be considerably smaller than what is observed. These same calculations suggest that the outgrowth of strong pre-existing genetic resistance mutations within a tumor should be ubiquitous. Yet, clinically relevant drug resistance occurs in the absence of obvious resistance conferring genetic alterations. Quantitatively, understanding the underlying biological mechanisms of failure mode diversity may improve the next generation of targeted anticancer therapies. It also provides insights into how intratumoral heterogeneity might shape interpatient diversity during clinical relapse. Materials and Methods: We employed spatial agent-based models to explore regimes where spatial constraints enable wild type cells (that encounter beneficial microenvironments) to compete against genetically resistant subclones in the presence of therapy. In order to parameterize a model of microenvironmental resistance, BT20 cells were cultured in the presence and absence of fibroblasts from 16 different tissues. The degree of resistance conferred by cancer associated fibroblasts in the tumor microenvironment was quantified by treating mono- and co-cultures with letrozole and then measuring the death rates. Results and Discussion: Our simulations indicate that, even when a mutation is more drug resistant, its outgrowth can be delayed by abundant, low magnitude microenvironmental resistance across large regions of a tumor that lack genetic resistance. These observations hold for different modes of microenvironmental resistance, including juxtacrine signaling, soluble secreted factors, and remodeled ECM. This result helps to explain the remarkable diversity of resistance mechanisms observed in solid tumors, which subverts the presumption that the failure mode that causes the quantitatively fastest growth in the presence of drug should occur most often in the clinic. Conclusion: Our model results demonstrate that spatial effects can interact with low magnitude of resistance microenvironmental effects to successfully compete against genetic resistance that is orders of magnitude larger. Clinical outcomes of solid tumors are intrinsically connected to their spatial structure, and the tractability of spatial agent-based models like the ones presented here enable us to understand this relationship more completely.

Self-assembled ruthenium and osmium nanosystems display a potent anticancer profile by interfering with metabolic activity.

We disclose novel amphiphilic ruthenium and osmium complexes that auto-assemble into nanomedicines with potent antiproliferative activity by inhibition of mitochondrial respiration. The self-assembling units were rationally designed from the [M(p-cymene)(1,10-phenanthroline)Cl]PF6 motif (where M is either RuII or OsII) with an appended C16 fatty chain to achieve high cellular activity, nano-assembling and mitochondrial targeting. These amphiphilic complexes block cell proliferation at the sub-micromolar range and are particularly potent towards glioblastoma neurospheres made from patient-derived cancer stem cells. A subcutaneous mouse model using these glioblastoma stem cells highlights one of our C16 OsII nanomedicines as highly successful in vivo. Mechanistically, we show that they act as metabolic poisons, strongly impairing mitochondrial respiration, corroborated by morphological changes and damage to the mitochondria. A genetic strategy based on RNAi gave further insight on the potential involvement of microtubules as part of the induced cell death. In parallel, we examined the structural properties of these new amphiphilic metal-based constructs, their reactivity and mechanism.

Materials-driven approaches to understand extrinsic drug resistance in cancer.

Metastatic cancer has a poor prognosis, because it is broadly disseminated and associated with both intrinsic and acquired drug resistance. Critical unmet needs in effectively killing drug resistant cancer cells include overcoming the drug desensitization characteristics of some metastatic cancers/lesions, and tailoring therapeutic regimens to both the tumor microenvironment and the genetic profiles of the resident cancer cells. Bioengineers and materials scientists are developing technologies to determine how metastatic sites exclude therapies, and how extracellular factors (including cells, proteins, metabolites, extracellular matrix, and abiotic factors) at metastatic sites significantly affect drug pharmacodynamics. Two looming challenges are determining which feature, or combination of features, from the tumor microenvironment drive drug resistance, and what the relative impact is of extracellular signals vs. intrinsic cell genetics in determining drug response. Sophisticated systems biology tools that can de-convolve a crowded network of signals and responses, as well as controllable microenvironments capable of providing discrete and tunable extracellular cues can help us begin to interrogate the high dimensional interactions governing drug resistance in patients.

Phase II Study of Ponatinib in Advanced Gastrointestinal Stromal Tumors: Efficacy, Safety, and Impact of Liquid Biopsy and Other Biomarkers.

PURPOSE: The purpose of this study is to evaluate ponatinib for advanced gastrointestinal stromal tumors (GIST). PATIENTS AND METHODS: This single-arm phase II trial enrolled patients with metastatic and/or unresectable GIST with failure of prior tyrosine kinase inhibitor (TKI) treatment into two cohorts based on presence or absence of KIT exon 11 (ex11) primary mutations. Patients initially received ponatinib 45 mg once daily. Following a temporary clinical hold in October 2013, dose reductions were implemented to reduce risk of arterial occlusive events (AOE). Primary endpoint was 16-week clinical benefit rate (CBR) in KIT ex11-positive cohort. KIT mutations in circulating tumor DNA (ctDNA) were assessed. RESULTS: Forty-five patients enrolled (30 KIT ex11-positive and 15 KIT ex11-negative); median follow-up was 14.7 and 13.6 months, respectively, as of August 1, 2016. Sixteen-week CBR was 36% (KIT ex11-positive; primary endpoint) and 20% (KIT ex11-negative). ctDNA analyses (n = 37) demonstrated strong concordance of primary KIT mutations between plasma and tumor. At least two secondary mutations were detected in 35% of patients overall and 54% of KIT ex11-positive patients. Changes from baseline in mutated ctDNA levels were consistent with clinical activity. Ponatinib was ineffective in patients with KIT exon 9 primary mutations. Resistance was associated with emergence of V654A. AOEs and venous thromboembolic events occurred in three and two patients, respectively. Six patients died; two deaths (pneumonia and pulmonary embolism) were considered possibly ponatinib-related. CONCLUSIONS: Ponatinib demonstrated activity in advanced GIST, particularly in KIT ex11-positive disease. ctDNA analysis confirmed heterogeneous resistance mutations in TKI-pretreated advanced GIST. Safety was consistent with previous studies.

Genomic and experimental evidence that ALKATI does not predict single agent sensitivity to ALK inhibitors.

Genomic data can facilitate personalized treatment decisions by enabling therapeutic hypotheses in individual patients. Mutual exclusivity has been an empirically useful signal for identifying activating mutations that respond to single agent targeted therapies. However, a low mutation frequency can underpower this signal for rare variants. We develop a resampling based method for the direct pairwise comparison of conditional selection between sets of gene pairs. We apply this method to a transcript variant of anaplastic lymphoma kinase (ALK) in melanoma, termed ALKATI that was suggested to predict sensitivity to ALK inhibitors and we find that it is not mutually exclusive with key melanoma oncogenes. Furthermore, we find that ALKATI is not likely to be sufficient for cellular transformation or growth, and it does not predict single agent therapeutic dependency. Our work strongly disfavors the role of ALKATI as a targetable oncogenic driver that might be sensitive to single agent ALK treatment.

A global cancer data integrator reveals principles of synthetic lethality, sex disparity and immunotherapy.

BACKGROUND: Advances in cancer biology are increasingly dependent on integration of heterogeneous datasets. Large-scale efforts have systematically mapped many aspects of cancer cell biology; however, it remains challenging for individual scientists to effectively integrate and understand this data. RESULTS: We have developed a new data retrieval and indexing framework that allows us to integrate publicly available data from different sources and to combine publicly available data with new or bespoke datasets. Our approach, which we have named the cancer data integrator (CanDI), is straightforward to implement, is well documented, and is continuously updated which should enable individual users to take full advantage of efforts to map cancer cell biology. We show that CanDI empowered testable hypotheses of new synthetic lethal gene pairs, genes associated with sex disparity, and immunotherapy targets in cancer. CONCLUSIONS: CanDI provides a flexible approach for large-scale data integration in cancer research enabling rapid generation of hypotheses. The CanDI data integrator is available at https://github.com/GilbertLabUCSF/CanDI .

Synthesis, structure and anticancer properties of new biotin- and morpholine-functionalized ruthenium and osmium half-sandwich complexes.

Ruthenium (Ru) and osmium (Os) complexes are of sustained interest in cancer research and may be alternative to platinum-based therapy. We detail here three new series of ruthenium and osmium complexes, supported by physico-chemical characterizations, including time-dependent density functional theory, a combined experimental and computational study on the aquation reactions and the nature of the metal-arene bond. Cytotoxic profiles were then evaluated on several cancer cell lines although with limited success. Further investigations were, however, performed on the most active series using a genetic approach based on RNA interference and highlighted a potential multi-target mechanism of action through topoisomerase II, mitotic spindle, HDAC and DNMT inhibition.

Multi-scale Predictions of Drug Resistance Epidemiology Identify Design Principles for Rational Drug Design.

Rationally designing drugs that last longer in the face of biological evolution is a critical objective of drug discovery. However, this goal is thwarted by the diversity and stochasticity of evolutionary trajectories that drive uncertainty in the clinic. Although biophysical models can qualitatively predict whether a mutation causes resistance, they cannot quantitatively predict the relative abundance of resistance mutations in patient populations. We present stochastic, first-principle models that are parameterized on a large in vitro dataset and that accurately predict the epidemiological abundance of resistance mutations across multiple leukemia clinical trials. The ability to forecast resistance variants requires an understanding of their underlying mutation biases. Beyond leukemia, a meta-analysis across prostate cancer, breast cancer, and gastrointestinal stromal tumors suggests that resistance evolution in the adjuvant setting is influenced by mutational bias. Our analysis establishes a principle for rational drug design: when evolution favors the most probable mutant, so should drug design.

The primary mechanism of cytotoxicity of the chemotherapeutic agent CX-5461 is topoisomerase II poisoning.

Small molecules can affect many cellular processes. The disambiguation of these effects to identify the causative mechanisms of cell death is extremely challenging. This challenge impacts both clinical development and the interpretation of chemical genetic experiments. CX-5461 was developed as a selective RNA polymerase I inhibitor, but recent evidence suggests that it may cause DNA damage and induce G-quadraplex formation. Here we use three complimentary data mining modalities alongside biochemical and cell biological assays to show that CX-5461 exerts its primary cytotoxic activity through topoisomerase II poisoning. We then show that acquired resistance to CX-5461 in previously sensitive lymphoma cells confers collateral resistance to the topoisomerase II poison doxorubicin. Doxorubicin is already a frontline chemotherapy in a variety of hematopoietic malignancies, and CX-5461 is being tested in relapse/refractory hematopoietic tumors. Our data suggest that the mechanism of cell death induced by CX-5461 is critical for rational clinical development in these patients. Moreover, CX-5461 usage as a specific chemical genetic probe of RNA polymerase I function is challenging to interpret. Our multimodal data-driven approach is a useful way to detangle the intended and unintended mechanisms of drug action across diverse essential cellular processes.

Evolution of the nonsense-mediated decay pathway is associated with decreased cytolytic immune infiltration.

The somatic co-evolution of tumors and the cellular immune responses that combat them drives the diversity of immune-tumor interactions. This includes tumor mutations that generate neo-antigenic epitopes that elicit cytotoxic T-cell activity and subsequent pressure to select for genetic loss of antigen presentation. Most studies have focused on how tumor missense mutations can drive tumor immunity, but frameshift mutations have the potential to create far greater antigenic diversity. However, expression of this antigenic diversity is potentially regulated by Nonsense Mediated Decay (NMD) and NMD has been shown to be of variable efficiency in cancers. Here we studied how mutational changes influence global NMD and cytolytic immune responses. Using TCGA datasets, we derived novel patient-level metrics of 'NMD burden' and interrogated how different mutation and most importantly NMD burdens influence cytolytic activity using machine learning models and survival outcomes. We find that NMD is a significant and independent predictor of immune cytolytic activity. Different indications exhibited varying dependence on NMD and mutation burden features. We also observed significant co-alteration of genes in the NMD pathway, with a global increase in NMD efficiency in patients with NMD co-alterations. Finally, NMD burden also stratified patient survival in multivariate regression models in subset of cancer types. Our work suggests that beyond selecting for mutations that elicit NMD in tumor suppressors, tumor evolution may react to the selective pressure generated by inflammation to globally enhance NMD through coordinated amplification and/or mutation.

Single-Molecule Sequencing Reveals Patterns of Preexisting Drug Resistance That Suggest Treatment Strategies in Philadelphia-Positive Leukemias.

Purpose: Sequential treatment with targeted therapies can result in complex combinations of resistance mutations in drug targets. This mutational complexity has spurred the development of pan-target inhibitors, i.e., therapies for which no single target mutation can cause resistance. Because the propensity for on- versus off-target resistance varies across cancer types, a deeper understanding of the mutational burden in drug targets could rationalize treatment outcomes and prioritize pan-target inhibitors for indications where on-target mutations are most likely.Experimental Design: To measure and model the mutational landscape of a drug target at high resolution, we integrated single-molecule Duplex Sequencing of the ABL1 gene in Philadelphia-positive (Ph+) leukemias with computational simulations.Results: A combination of drug target mutational burden and tumor-initiating cell fraction is sufficient to predict that most patients with chronic myeloid leukemia are unlikely to harbor ABL1 resistance mutations at the time of diagnosis, rationalizing the exceptional success of targeted therapy in this setting. In contrast, our analysis predicts that many patients with Ph+ acute lymphoblastic leukemia (Ph+ ALL) harbor multiple preexisting resistant cells with single mutants. The emergence of compound mutations can be traced to initial use of an ABL1 inhibitor that is susceptible to resistance from single point mutations.Conclusions: These results argue that early use of therapies that achieve pan-inhibition of ABL1 resistance mutants might improve outcomes in Ph+ ALL. Our findings show how a deep understanding of the mutational burden in drug targets can be quantitatively coupled to phenotypic heterogeneity to rationalize clinical phenomena. Clin Cancer Res; 24(21); 5321-34. ©2018 AACR.

A subset of platinum-containing chemotherapeutic agents kills cells by inducing ribosome biogenesis stress.

Cisplatin and its platinum analogs, carboplatin and oxaliplatin, are some of the most widely used cancer chemotherapeutics. Although cisplatin and carboplatin are used primarily in germ cell, breast and lung malignancies, oxaliplatin is instead used almost exclusively to treat colorectal and other gastrointestinal cancers. Here we utilize a unique, multi-platform genetic approach to study the mechanism of action of these clinically established platinum anti-cancer agents, as well as more recently developed cisplatin analogs. We show that oxaliplatin, unlike cisplatin and carboplatin, does not kill cells through the DNA-damage response. Rather, oxaliplatin kills cells by inducing ribosome biogenesis stress. This difference in drug mechanism explains the distinct clinical implementation of oxaliplatin relative to cisplatin, and it might enable mechanistically informed selection of distinct platinum drugs for distinct malignancies. These data highlight the functional diversity of core components of front-line cancer therapy and the potential benefits of applying a mechanism-based rationale to the use of our current arsenal of anti-cancer drugs.

Inherited Disease Genetics Improves the Identification of Cancer-Associated Genes.

The identification of biologically significant variants in cancer genomes is critical to therapeutic discovery, but it is limited by the statistical power needed to discern driver from passenger. Independent biological data can be used to filter cancer exomes and increase statistical power. Large genetic databases for inherited diseases are uniquely suited to this task because they contain specific amino acid alterations with known pathogenicity and molecular mechanisms. However, no rigorous method to overlay this information onto the cancer exome exists. Here, we present a computational methodology that overlays any variant database onto the somatic mutations in all cancer exomes. We validate the computation experimentally and identify novel associations in a re-analysis of 7362 cancer exomes. This analysis identified activating SOS1 mutations associated with Noonan syndrome as significantly altered in melanoma and the first kinase-activating mutations in ACVR1 associated with adult tumors. Beyond a filter, significant variants found in both rare cancers and rare inherited diseases increase the unmet medical need for therapeutics that target these variants and may bootstrap drug discovery efforts in orphan indications.

Exploiting Temporal Collateral Sensitivity in Tumor Clonal Evolution.

The prevailing approach to addressing secondary drug resistance in cancer focuses on treating the resistance mechanisms at relapse. However, the dynamic nature of clonal evolution, along with potential fitness costs and cost compensations, may present exploitable vulnerabilities-a notion that we term "temporal collateral sensitivity." Using a combined pharmacological screen and drug resistance selection approach in a murine model of Ph(+) acute lymphoblastic leukemia, we indeed find that temporal and/or persistent collateral sensitivity to non-classical BCR-ABL1 drugs arises in emergent tumor subpopulations during the evolution of resistance toward initial treatment with BCR-ABL1-targeted inhibitors. We determined the sensitization mechanism via genotypic, phenotypic, signaling, and binding measurements in combination with computational models and demonstrated significant overall survival extension in mice. Additional stochastic mathematical models and small-molecule screens extended our insights, indicating the value of focusing on evolutionary trajectories and pharmacological profiles to identify new strategies to treat dynamic tumor vulnerabilities.

A genome-scale in vivo loss-of-function screen identifies Phf6 as a lineage-specific regulator of leukemia cell growth.

We performed a genome-scale shRNA screen for modulators of B-cell leukemia progression in vivo. Results from this work revealed dramatic distinctions between the relative effects of shRNAs on the growth of tumor cells in culture versus in their native microenvironment. Specifically, we identified many "context-specific" regulators of leukemia development. These included the gene encoding the zinc finger protein Phf6. While inactivating mutations in PHF6 are commonly observed in human myeloid and T-cell malignancies, we found that Phf6 suppression in B-cell malignancies impairs tumor progression. Thus, Phf6 is a "lineage-specific" cancer gene that plays opposing roles in developmentally distinct hematopoietic malignancies.

Addressing genetic tumor heterogeneity through computationally predictive combination therapy.

UNLABELLED: Recent tumor sequencing data suggest an urgent need to develop a methodology to directly address intratumoral heterogeneity in the design of anticancer treatment regimens. We use RNA interference to model heterogeneous tumors, and demonstrate successful validation of computational predictions for how optimized drug combinations can yield superior effects on these tumors both in vitro and in vivo. Importantly, we discover here that for many such tumors knowledge of the predominant subpopulation is insufficient for determining the best drug combination. Surprisingly, in some cases, the optimal drug combination does not include drugs that would treat any particular subpopulation most effectively, challenging straightforward intuition. We confirm examples of such a case with survival studies in a murine preclinical lymphoma model. Altogether, our approach provides new insights about design principles for combination therapy in the context of intratumoral diversity, data that should inform the development of drug regimens superior for complex tumors. SIGNIFICANCE: This study provides the first example of how combination drug regimens, using existing chemotherapies, can be rationally designed to maximize tumor cell death, while minimizing the outgrowth of clonal subpopulations.

Predicting cancer drug mechanisms of action using molecular network signatures.

Molecular signatures are a powerful approach to characterize novel small molecules and derivatized small molecule libraries. While new experimental techniques are being developed in diverse model systems, informatics approaches lag behind these exciting advances. We propose an analysis pipeline for signature based drug annotation. We develop an integrated strategy, utilizing supervised and unsupervised learning methodologies that are bridged by network based statistics. Using this approach we can: 1, predict new examples of drug mechanisms that we trained our model upon; 2, identify "New" mechanisms of action that do not belong to drug categories that our model was trained upon; and 3, update our training sets with these "New" mechanisms and accurately predict entirely distinct examples from these new categories. Thus, not only does our strategy provide statistical generalization but it also offers biological generalization. Additionally, we show that our approach is applicable to diverse types of data, and that distinct biological mechanisms characterize its resolution of categories across different data types. As particular examples, we find that our predictive resolution of drug mechanisms from mRNA expression studies relies upon the analog measurement of a cell stress-related transcriptional rheostat along with a transcriptional representation of cell cycle state; whereas, in contrast, drug mechanism resolution from functional RNAi studies rely upon more dichotomous (e.g., either enhances or inhibits) association with cell death states. We believe that our approach can facilitate molecular signature-based drug mechanism understanding from different technology platforms and across diverse biological phenomena.