The Virtual Child

SUMMARY

We are building the world's first Virtual Child-a computer model of normal and cancerous human development at the level of each individual cell. The Virtual Child will "develop cancer" that we will subject to unlimited virtual clinical trials that pinpoint, predict, and prioritize potential new treatments, bringing forward the day when no child dies of cancer, giving each one the opportunity to lead a full and healthy life.

Thio-2 inhibits key signaling pathways required for the development and progression of castration resistant prostate cancer

Therapies that abrogate persistent androgen receptor (AR) signaling in castration resistant prostate cancer (CRPC) remain an unmet clinical need. The N-terminal domain (NTD) of the AR that drives transcriptional activity in CRPC remains a challenging therapeutic target. Herein we demonstrate that BAG-1 mRNA is highly expressed and associates with signaling pathways, including AR signaling, that are implicated in the development and progression of CRPC. In addition, interrogation of geometric and physiochemical properties of the BAG domain of BAG-1 isoforms identifies it to be a tractable but challenging drug target. Furthermore, through BAG-1 isoform mouse knockout studies we confirm that BAG-1 isoforms regulate hormone physiology and that therapies targeting the BAG domain will be associated with limited 'on-target' toxicity. Importantly, the postulated inhibitor of BAG-1 isoforms, Thio-2, suppressed AR signaling and other important pathways implicated in the development and progression of CRPC to reduce the growth of treatment resistant prostate cancer cell lines and patient derived models. However, the mechanism by which Thio-2 elicits the observed phenotype needs further elucidation since the genomic abrogation of BAG-1 isoforms was unable to recapitulate the Thio-2 mediated phenotype. Overall, these data support the interrogation of related compounds with improved drug-like properties as a novel therapeutic approach in CRPC, and further highlight the clinical potential of treatments that block persistent AR signaling which are currently undergoing clinical evaluation in CRPC.

SwarmDeepSurv: swarm intelligence advances deep survival network for prognostic radiomics signatures in four solid cancers

Survival models exist to study relationships between biomarkers and treatment effects. Deep learning-powered survival models supersede the classical Cox proportional hazards (CoxPH) model, but substantial performance drops were observed on high-dimensional features because of irrelevant/redundant information. To fill this gap, we proposed SwarmDeepSurv by integrating swarm intelligence algorithms with the deep survival model. Furthermore, four objective functions were designed to optimize prognostic prediction while regularizing selected feature numbers. When testing on multicenter sets (n = 1,058) of four different cancer types, SwarmDeepSurv was less prone to overfitting and achieved optimal patient risk stratification compared with popular survival modeling algorithms. Strikingly, SwarmDeepSurv selected different features compared with classical feature selection algorithms, including the least absolute shrinkage and selection operator (LASSO), with nearly no feature overlapping across these models. Taken together, SwarmDeepSurv offers an alternative approach to model relationships between radiomics features and survival endpoints, which can further extend to study other input data types including genomics.

canSAR: update to the cancer translational research and drug discovery knowledgebase

canSAR (https://cansar.ai) is the largest public cancer drug discovery and translational research knowledgebase. Now hosted in its new home at MD Anderson Cancer Center, canSAR integrates billions of experimental measurements from across molecular profiling, pharmacology, chemistry, structural and systems biology. Moreover, canSAR applies a unique suite of machine learning algorithms designed to inform drug discovery. Here, we describe the latest updates to the knowledgebase, including a focus on significant novel data. These include canSAR's ligandability assessment of AlphaFold; mapping of fragment-based screening data; and new chemical bioactivity data for novel targets. We also describe enhancements to the data and interface.

The Chemical Probes Portal: an expert review-based public resource to empower chemical probe assessment, selection and use

We describe the Chemical Probes Portal (https://www.chemicalprobes.org/), an expert review-based public resource to empower chemical probe assessment, selection and use. Chemical probes are high-quality small-molecule reagents, often inhibitors, that are important for exploring protein function and biological mechanisms, and for validating targets for drug discovery. The publication, dissemination and use of chemical probes provide an important means to accelerate the functional annotation of proteins, the study of proteins in cell biology, physiology, and disease pathology, and to inform and enable subsequent pioneering drug discovery and development efforts. However, the widespread use of small-molecule compounds that are claimed as chemical probes but are lacking sufficient quality, especially being inadequately selective for the desired target or even broadly promiscuous in behaviour, has resulted in many erroneous conclusions in the biomedical literature. The Chemical Probes Portal was established as a public resource to aid the selection and best-practice use of chemical probes in basic and translational biomedical research. We describe the background, principles and content of the Portal and its technical development, as well as examples of its applications and use. The Chemical Probes Portal is a community resource and we therefore describe how researchers can be involved in its content and development.

Individualized Prediction of Drug Response and Rational Combination Therapy in NSCLC Using Artificial Intelligence-Enabled Studies of Acute Phosphoproteomic Changes

We hypothesize that the study of acute protein perturbation in signal transduction by targeted anticancer drugs can predict drug sensitivity of these agents used as single agents and rational combination therapy. We assayed dynamic changes in 52 phosphoproteins caused by an acute exposure (1 hour) to clinically relevant concentrations of seven targeted anticancer drugs in 35 non-small cell lung cancer (NSCLC) cell lines and 16 samples of NSCLC cells isolated from pleural effusions. We studied drug sensitivities across 35 cell lines and synergy of combinations of all drugs in six cell lines (252 combinations). We developed orthogonal machine-learning approaches to predict drug response and rational combination therapy. Our methods predicted the most and least sensitive quartiles of drug sensitivity with an AUC of 0.79 and 0.78, respectively, whereas predictions based on mutations in three genes commonly known to predict response to the drug studied, for example, EGFR, PIK3CA, and KRAS, did not predict sensitivity (AUC of 0.5 across all quartiles). The machine-learning predictions of combinations that were compared with experimentally generated data showed a bias to the highest quartile of Bliss synergy scores (P = 0.0243). We confirmed feasibility of running such assays on 16 patient samples of freshly isolated NSCLC cells from pleural effusions. We have provided proof of concept for novel methods of using acute ex vivo exposure of cancer cells to targeted anticancer drugs to predict response as single agents or combinations. These approaches could complement current approaches using gene mutations/amplifications/rearrangements as biomarkers and demonstrate the utility of proteomics data to inform treatment selection in the clinic.

canSAR chemistry registration and standardization pipeline

Background

Integration of medicinal chemistry data from numerous public resources is an increasingly important part of academic drug discovery and translational research because it can bring a wealth of important knowledge related to compounds in one place. However, different data sources can report the same or related compounds in various forms (e.g., tautomers, racemates, etc.), thus highlighting the need of organising related compounds in hierarchies that alert the user on important bioactivity data that may be relevant. To generate these compound hierarchies, we have developed and implemented canSARchem, a new compound registration and standardization pipeline as part of the canSAR public knowledgebase. canSARchem builds on previously developed ChEMBL and PubChem pipelines and is developed using KNIME. We describe the pipeline which we make publicly available, and we provide examples on the strengths and limitations of the use of hierarchies for bioactivity data exploration. Finally, we identify canonicalization enrichment in FDA-approved drugs, illustrating the benefits of our approach.

Results

We created a chemical registration and standardization pipeline in KNIME and made it freely available to the research community. The pipeline consists of five steps to register the compounds and create the compounds’ hierarchy: 1. Structure checker, 2. Standardization, 3. Generation of canonical tautomers and representative structures, 4. Salt strip, and 5. Generation of abstract structure to generate the compound hierarchy. Unlike ChEMBL’s RDKit pipeline, we carry out compound canonicalization ahead of getting the parent structure, similar to PubChem’s OpenEye pipeline. canSARchem has a lower rejection rate compared to both PubChem and ChEMBL. We use our pipeline to assess the impact of grouping the compounds in hierarchies for bioactivity data exploration. We find that FDA-approved drugs show statistically significant sensitivity to canonicalization compared to the majority of bioactive compounds which demonstrates the importance of this step.

Conclusions

We use canSARchem to standardize all the compounds uploaded in canSAR (> 3 million) enabling efficient data integration and the rapid identification of alternative compound forms with useful bioactivity data. Comparison with PubChem and ChEMBL pipelines evidenced comparable performances in compound standardization, but only PubChem and canSAR canonicalize tautomers and canSAR has a slightly lower rejection rate. Our results highlight the importance of compound hierarchies for bioactivity data exploration. We make canSARchem available under a Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0) at https://gitlab.icr.ac.uk/cansar-public/compound-registration-pipeline.

Target 2035 - update on the quest for a probe for every protein

Twenty years after the publication of the first draft of the human genome, our knowledge of the human proteome is still fragmented. The challenge of translating the wealth of new knowledge from genomics into new medicines is that proteins, and not genes, are the primary executers of biological function. Therefore, much of how biology works in health and disease must be understood through the lens of protein function. Accordingly, a subset of human proteins has been at the heart of research interests of scientists over the centuries, and we have accumulated varying degrees of knowledge about approximately 65% of the human proteome. Nevertheless, a large proportion of proteins in the human proteome (∼35%) remains uncharacterized, and less than 5% of the human proteome has been successfully targeted for drug discovery. This highlights the profound disconnect between our abilities to obtain genetic information and subsequent development of effective medicines. Target 2035 is an international federation of biomedical scientists from the public and private sectors, which aims to address this gap by developing and applying new technologies to create by year 2035 chemogenomic libraries, chemical probes, and/or biological probes for the entire human proteome.

PDBe-KB: collaboratively defining the biological context of structural data

The Protein Data Bank in Europe - Knowledge Base (PDBe-KB, https://pdbe-kb.org) is an open collaboration between world-leading specialist data resources contributing functional and biophysical annotations derived from or relevant to the Protein Data Bank (PDB). The goal of PDBe-KB is to place macromolecular structure data in their biological context by developing standardised data exchange formats and integrating functional annotations from the contributing partner resources into a knowledge graph that can provide valuable biological insights. Since we described PDBe-KB in 2019, there have been significant improvements in the variety of available annotation data sets and user functionality. Here, we provide an overview of the consortium, highlighting the addition of annotations such as predicted covalent binders, phosphorylation sites, effects of mutations on the protein structure and energetic local frustration. In addition, we describe a library of reusable web-based visualisation components and introduce new features such as a bulk download data service and a novel superposition service that generates clusters of superposed protein chains weekly for the whole PDB archive.

Evolution of kinase polypharmacology across HSP90 drug discovery

Most small molecules interact with several target proteins but this polypharmacology is seldom comprehensively investigated or explicitly exploited during drug discovery. Here, we use computational and experimental methods to identify and systematically characterize the kinase cross-pharmacology of representative HSP90 inhibitors. We demonstrate that the resorcinol clinical candidates ganetespib and, to a lesser extent, luminespib, display unique off-target kinase pharmacology as compared with other HSP90 inhibitors. We also demonstrate that polypharmacology evolved during the optimization to discover luminespib and that the hit, leads, and clinical candidate all have different polypharmacological profiles. We therefore recommend the computational and experimental characterization of polypharmacology earlier in drug discovery projects to unlock new multi-target drug design opportunities.

Public resources for chemical probes: the journey so far and the road ahead

High-quality small molecule chemical probes are extremely valuable for biological research and target validation. However, frequent use of flawed small-molecule inhibitors produces misleading results and diminishes the robustness of biomedical research. Several public resources are available to facilitate assessment and selection of better chemical probes for specific protein targets. Here, we review chemical probe resources, discuss their current strengths and limitations, and make recommendations for further improvements. Expert review resources provide in-depth analysis but currently cover only a limited portion of the liganded proteome. Computational resources encompass more proteins and are regularly updated, but have limitations in data availability and curation. We show how biomedical scientists may use these resources to choose the best available chemical probes for their research.

JMJD6 Is a Druggable Oxygenase That Regulates AR-V7 Expression in Prostate Cancer

Endocrine resistance (EnR) in advanced prostate cancer is fatal. EnR can be mediated by androgen receptor (AR) splice variants, with AR splice variant 7 (AR-V7) arguably the most clinically important variant. In this study, we determined proteins key to generating AR-V7, validated our findings using clinical samples, and studied splicing regulatory mechanisms in prostate cancer models. Triangulation studies identified JMJD6 as a key regulator of AR-V7, as evidenced by its upregulation with in vitro EnR, its downregulation alongside AR-V7 by bromodomain inhibition, and its identification as a top hit of a targeted siRNA screen of spliceosome-related genes. JMJD6 protein levels increased (P < 0.001) with castration resistance and were associated with higher AR-V7 levels and shorter survival (P = 0.048). JMJD6 knockdown reduced prostate cancer cell growth, AR-V7 levels, and recruitment of U2AF65 to AR pre-mRNA. Mutagenesis studies suggested that JMJD6 activity is key to the generation of AR-V7, with the catalytic machinery residing within a druggable pocket. Taken together, these data highlight the relationship between JMJD6 and AR-V7 in advanced prostate cancer and support further evaluation of JMJD6 as a therapeutic target in this disease. SIGNIFICANCE: This study identifies JMJD6 as being critical for the generation of AR-V7 in prostate cancer, where it may serve as a tractable target for therapeutic intervention.

Tuning Local Hydration Enables a Deeper Understanding of Protein-Ligand Binding: The PP1-Src Kinase Case

Water plays a key role in biomolecular recognition and binding. Despite the development of several computational and experimental approaches, it is still challenging to comprehensively characterize water-mediated effects on the binding process. Here, we investigate how water affects the binding of Src kinase to one of its inhibitors, PP1. Src kinase is a target for treating several diseases, including cancer. We use biased molecular dynamics simulations, where the hydration of predetermined regions is tuned at will. This computational technique efficiently accelerates the SRC-PP1 binding simulation and allows us to identify several key and yet unexplored aspects of the solvent's role. This study provides a further perspective on the binding phenomenon, which may advance the current drug design approaches for the development of new kinase inhibitors.

canSAR: update to the cancer translational research and drug discovery knowledgebase

canSAR (http://cansar.icr.ac.uk) is the largest, public, freely available, integrative translational research and drug discovery knowledgebase for oncology. canSAR integrates vast multidisciplinary data from across genomic, protein, pharmacological, drug and chemical data with structural biology, protein networks and more. It also provides unique data, curation and annotation and crucially, AI-informed target assessment for drug discovery. canSAR is widely used internationally by academia and industry. Here we describe significant developments and enhancements to the data, web interface and infrastructure of canSAR in the form of the new implementation of the system: canSARblack. We demonstrate new functionality in aiding translation hypothesis generation and experimental design, and show how canSAR can be adapted and utilised outside oncology.

Solution structure of the Hop TPR2A domain and investigation of target druggability by NMR, biochemical and in silico approaches

Heat shock protein 90 (Hsp90) is a molecular chaperone that plays an important role in tumour biology by promoting the stabilisation and activity of oncogenic 'client' proteins. Inhibition of Hsp90 by small-molecule drugs, acting via its ATP hydrolysis site, has shown promise as a molecularly targeted cancer therapy. Owing to the importance of Hop and other tetratricopeptide repeat (TPR)-containing cochaperones in regulating Hsp90 activity, the Hsp90-TPR domain interface is an alternative site for inhibitors, which could result in effects distinct from ATP site binders. The TPR binding site of Hsp90 cochaperones includes a shallow, positively charged groove that poses a significant challenge for druggability. Herein, we report the apo, solution-state structure of Hop TPR2A which enables this target for NMR-based screening approaches. We have designed prototype TPR ligands that mimic key native 'carboxylate clamp' interactions between Hsp90 and its TPR cochaperones and show that they block binding between Hop TPR2A and the Hsp90 C-terminal MEEVD peptide. We confirm direct TPR-binding of these ligands by mapping 1H-15N HSQC chemical shift perturbations to our new NMR structure. Our work provides a novel structure, a thorough assessment of druggability and robust screening approaches that may offer a potential route, albeit difficult, to address the chemically challenging nature of the Hop TPR2A target, with relevance to other TPR domain interactors.

Correction: Signalling involving MET and FAK supports cell division independent of the activity of the cell cycle-regulating CDK4/6 kinases

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

The kinase polypharmacology landscape of clinical PARP inhibitors

Polypharmacology plays an important role in defining response and adverse effects of drugs. For some mechanisms, experimentally mapping polypharmacology is commonplace, although this is typically done within the same protein class. Four PARP inhibitors have been approved by the FDA as cancer therapeutics, yet a precise mechanistic rationale to guide clinicians on which to choose for a particular patient is lacking. The four drugs have largely similar PARP family inhibition profiles, but several differences at the molecular and clinical level have been reported that remain poorly understood. Here, we report the first comprehensive characterization of the off-target kinase landscape of four FDA-approved PARP drugs. We demonstrate that all four PARP inhibitors have a unique polypharmacological profile across the kinome. Niraparib and rucaparib inhibit DYRK1s, CDK16 and PIM3 at clinically achievable, submicromolar concentrations. These kinases represent the most potently inhibited off-targets of PARP inhibitors identified to date and should be investigated further to clarify their potential implications for efficacy and safety in the clinic. Moreover, broad kinome profiling is recommended for the development of PARP inhibitors as PARP-kinase polypharmacology could potentially be exploited to modulate efficacy and side-effect profiles.

PDBe-KB: a community-driven resource for structural and functional annotations

The Protein Data Bank in Europe-Knowledge Base (PDBe-KB, https://pdbe-kb.org) is a community-driven, collaborative resource for literature-derived, manually curated and computationally predicted structural and functional annotations of macromolecular structure data, contained in the Protein Data Bank (PDB). The goal of PDBe-KB is two-fold: (i) to increase the visibility and reduce the fragmentation of annotations contributed by specialist data resources, and to make these data more findable, accessible, interoperable and reusable (FAIR) and (ii) to place macromolecular structure data in their biological context, thus facilitating their use by the broader scientific community in fundamental and applied research. Here, we describe the guidelines of this collaborative effort, the current status of contributed data, and the PDBe-KB infrastructure, which includes the data exchange format, the deposition system for added value annotations, the distributable database containing the assembled data, and programmatic access endpoints. We also describe a series of novel web-pages-the PDBe-KB aggregated views of structure data-which combine information on macromolecular structures from many PDB entries. We have recently released the first set of pages in this series, which provide an overview of available structural and functional information for a protein of interest, referenced by a UniProtKB accession.

Abstract LB-C01: The kinase polypharmacology landscape of clinical PARP inhibitors

Polypharmacology plays an important role in defining response and adverse effects of drugs. For some mechanisms, experimentally mapping polypharmacology is commonplace, although often this is done within the same protein class. Four PARP inhibitors have been approved by the FDA as cancer therapeutics, yet a precise mechanistic rationale to guide clinicians on which to choose for a particular patient is lacking. The four drugs have largely similar PARP family inhibition profiles, but several differences at the molecular and clinical level have been reported that remain poorly understood. Here, we report the first comprehensive characterization of the off-target kinase landscape of four FDA-approved PARP drugs. We demonstrate that all four PARP inhibitors have a unique polypharmacological profile across the kinome. Niraparib and rucaparib inhibit DYRK1s, CDK16 and PIM3 at clinically achievable, submicromolar concentrations. The most potently inhibited off-targets are CDK16 for rucaparib (IC50 = 381 nM) and DYRK1B for niraparib (IC50 = 254 nM). These kinases represent the most potently inhibited off-targets of PARP inhibitors identified to date and should be investigated further to clarify their potential implications for efficacy and safety in the clinic.

Citation Format: Albert A Antolin, Malaka Ameratunga, Udai Banerji, Paul Clarke, Paul Workman, Bissan Al-Lazikani. The kinase polypharmacology landscape of clinical PARP inhibitors [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 LB-C01. doi:10.1158/1535-7163.TARG-19-LB-C01

A tailored molecular profiling programme for children with cancer to identify clinically actionable genetic alterations

BACKGROUND

For children with cancer, the clinical integration of precision medicine to enable predictive biomarker-based therapeutic stratification is urgently needed.

METHODS

We have developed a hybrid-capture next-generation sequencing (NGS) panel, specifically designed to detect genetic alterations in paediatric solid tumours, which gives reliable results from as little as 50 ng of DNA extracted from formalin-fixed paraffin-embedded (FFPE) tissue. In this study, we offered an NGS panel, with clinical reporting via a molecular tumour board for children with solid tumours. Furthermore, for a cohort of 12 patients, we used a circulating tumour DNA (ctDNA)-specific panel to sequence ctDNA from matched plasma samples and compared plasma and tumour findings.

RESULTS

A total of 255 samples were submitted from 223 patients for the NGS panel. Using FFPE tissue, 82% of all submitted samples passed quality control for clinical reporting. At least one genetic alteration was detected in 70% of sequenced samples. The overall detection rate of clinically actionable alterations, defined by modified OncoKB criteria, for all sequenced samples was 51%. A total of 8 patients were sequenced at different stages of treatment. In 6 of these, there were differences in the genetic alterations detected between time points. Sequencing of matched ctDNA in a cohort of extracranial paediatric solid tumours also identified a high detection rate of somatic alterations in plasma.

CONCLUSION

We demonstrate that tailored clinical molecular profiling of both tumour DNA and plasma-derived ctDNA is feasible for children with solid tumours. Furthermore, we show that a targeted NGS panel-based approach can identify actionable genetic alterations in a high proportion of patients.

Differences in Signaling Patterns on PI3K Inhibition Reveal Context Specificity in KRAS-Mutant Cancers

It is increasingly appreciated that drug response to different cancers driven by the same oncogene is different and may relate to differences in rewiring of signal transduction. We aimed to study differences in dynamic signaling changes within mutant KRAS (KRAS MT), non-small cell lung cancer (NSCLC), colorectal cancer, and pancreatic ductal adenocarcinoma (PDAC) cells. We used an antibody-based phosphoproteomic platform to study changes in 50 phosphoproteins caused by seven targeted anticancer drugs in a panel of 30 KRAS MT cell lines and cancer cells isolated from 10 patients with KRAS MT cancers. We report for the first time significant differences in dynamic signaling between colorectal cancer and NSCLC cell lines exposed to clinically relevant equimolar concentrations of the pan-PI3K inhibitor pictilisib including a lack of reduction of p-AKTser473 in colorectal cancer cell lines (P = 0.037) and lack of compensatory increase in p-MEK in NSCLC cell lines (P = 0.036). Differences in rewiring of signal transduction between tumor types driven by KRAS MT cancers exist and influence response to combination therapy using targeted agents.

Signalling involving MET and FAK supports cell division independent of the activity of the cell cycle-regulating CDK4/6 kinases

Deregulation of cyclin-dependent kinases 4 and 6 (CDK4/6) is highly prevalent in cancer; yet, inhibitors against these kinases are currently used only in restricted tumour contexts. The extent to which cancers depend on CDK4/6 and the mechanisms that may undermine such dependency are poorly understood. Here, we report that signalling engaging the MET proto-oncogene receptor tyrosine kinase/focal adhesion kinase (FAK) axis leads to CDK4/6-independent CDK2 activation, involving as critical mechanistic events loss of the CDKI p21CIP1 and gain of its regulator, the ubiquitin ligase subunit SKP2. Combined inhibition of MET/FAK and CDK4/6 eliminates the proliferation capacity of cancer cells in culture, and enhances tumour growth inhibition in vivo. Activation of the MET/FAK axis is known to arise through cancer extrinsic and intrinsic cues. Our work predicts that such cues support cell division independent of the activity of the cell cycle-regulating CDK4/6 kinases and identifies MET/FAK as a tractable route to broaden the utility of CDK4/6 inhibitor-based therapies in the clinic.

canSAR: update to the cancer translational research and drug discovery knowledgebase

canSAR (http://cansar.icr.ac.uk) is a public, freely available, integrative translational research and drug discovery knowlegebase. canSAR informs researchers to help solve key bottlenecks in cancer translation and drug discovery. It integrates genomic, protein, pharmacological, drug and chemical data with structural biology, protein networks and unique, comprehensive and orthogonal 'druggability' assessments. canSAR is widely used internationally by academia and industry. Here we describe major enhancements to canSAR including new and expanded data. We also describe the first components of canSARblack-an advanced, responsive, multi-device compatible redesign of canSAR with a question-led interface.

Leveraging Human Genetics to Guide Cancer Drug Development

PURPOSE

The high attrition rate of cancer drug development programs is a barrier to realizing the promise of precision oncology. We have examined whether the genetic insights from genome-wide association studies of cancer can guide drug development and repurposing in oncology.

MATERIALS AND METHODS

Across 37 cancers, we identified 955 genetic risk variants from the National Human Genome Research Institute-European Bioinformatics Institute genome-wide association study catalog. We linked these variants to target genes using strategies that were based on linkage disequilibrium, DNA three-dimensional structure, and integration of predicted gene function and expression. With the use of the Informa Pharmaprojects database, we identified genes that are targets of unique drugs and assessed the level of enrichment that would be afforded by incorporation of genetic information in preclinical and phase II studies. For targets not under development, we implemented machine learning approaches to assess druggability.

RESULTS

For all preclinical targets incorporating genetic information, a 2.00-fold enrichment of a drug being successfully approved could be achieved (95% CI, 1.14- to 3.48-fold; P = .02). For phase II targets, a 2.75-fold enrichment could be achieved (95% CI, 1.42- to 5.35-fold; P < .001). Application of genetic information suggests potential repurposing of 15 approved nononcology drugs.

CONCLUSION

The findings illustrate the value of using insights from the genetics of inherited cancer susceptibility discovery projects as part of a data-driven strategy to inform drug discovery. Support for cancer germline genetic information for prospective targets is available online from the Institute of Cancer Research.

Abstract A067: Targeting the bromodomain and extra-terminal (BET) family proteins and beyond in metastatic castration-resistant prostate cancer (mCRPC): Overcoming aberrant androgen receptor (AR) signaling

Despite robust responses to androgen deprivation therapy and AR targeting therapies (including abiraterone and enzalutamide), nearly all cases of advanced prostate cancer progress to lethal mCRPC and develop therapeutic resistance. This progression is associated with persistent AR signaling, in part due to expression of constitutively active AR splice variants that include AR variant 7 (AR-V7). We show that AR-V7 expression in patient biopsies (protein) and circulating tumor cells (RNA) associates with poor outcome from mCRPC. Therapies that regulate AR-V7 and induce robust anticancer responses are required to confirm the clinical importance of AR-V7 in mCRPC. One such promising approach, currently in clinical trials, is inhibitors of BET family proteins, which include BRD2, BRD3, and BRD4. Preclinical studies have shown that BRD4 binds to AR at the androgen response element and facilitates the recruitment of the transcriptional machinery. We show that BRD4 protein expression increases as patients develop mCRPC and at diagnosis associates with patient outcome and more advanced disease. In addition, through RNAseq analysis we show that expression levels of BRD2, 3, and 4 in mCRPC associated with degree of AR activity. We, and others, have shown that the use of BET inhibitors (BETi) in vitro on AR/AR-V7 expressing cell lines not only decreased AR activity but also preferentially decreased the production on AR-V7 mRNA. In light of BETi having efficacy against BRD2, 3, and 4, and all isoforms being expressed in mCRPC, we explored the effect of genetic knockdown of each isoform. We show that BETi treatment is sufficient to decrease AR-V7 mRNA and protein in CRPC cell lines. Moreover, we demonstrate that BRD4 knockdown, and to a greater extent, the combination of BRD2, 3, and 4 knockdown blocked AR and AR-V7 signaling. Furthermore, C-MYC knockdown did not recapitulate the effect of BETi and led to an increase in AR signaling. Consistent with these findings, BETi and the combination of BRD2, 3, and 4 knockdown reduced the growth of CRPC cell lines. To further investigate whether BETi is sufficient to inhibit AR-V7 production in patients with mCRPC, we treated patient-derived organoids (PDOs) and a mouse xenograft (PDX) grown from patient metastatic biopsies who had progressed on enzalutamide and/or abiraterone. In this study 4 out of 9 PDOs were sensitive to BETi. Consistent with cell culture experiments, BETi treatment of PDO and PDX led to downregulation of both AR-V7 mRNA and protein expression, and growth inhibition. In light of the pleotropic effects of BETi on cancer cell biology and potential for treatment-related toxicities, we explored whether we could identify critical factors for BETi mediated AR-V7 regulation in CRPC. The ability of BETi to regulate AR-V7 may suggest an effect of BETi on pre-mRNA splicing of AR resulting in the observed decrease of AR-V7 expression. RNAseq analysis of the AR-V7 expressing CRPC cell line LNCaP95 treated with BETi demonstrated an increase in total splicing. Despite this, focused analysis of splicing factors and spliceosome components identified a subset of eight splicing factors being downregulated by BETi treatment, including one yet-uncharacterized factor (splicing factor B; Sf-B) that is crucial for AR-V7 expression and LNCaP95 cell growth. In addition, mCRPC patients who express high levels of Sf-B had a significantly poorer outcome and the protein structure of Sf-B is druggable using the drug discovery knowledgebase canSAR. Based on our results, we propose that inhibition of Sf-B may lead to decreased splicing and expression of AR-V7; providing a novel approach to target AR-V7 in mCRPC.

Citation Format: Adam Sharp, Jon Welti, Wei Yuan, Ines Figueiredo, Veronica Gil, Daniel Nava Rodrigues, Maryou Lambros, Eleanor Knight, Jian Ning, Jeff Francis, David Dolling, Lorna Pope, Antje Neeb, Gunther Boysen, Yezi Zhu, Mateus Crespo, Alec Paschalis, Jun Luo, Stephen Plymate, Bissan Al-Lazikani, Amanda Swain, Johann de Bono. Targeting the bromodomain and extra-terminal (BET) family proteins and beyond in metastatic castration-resistant prostate cancer (mCRPC): Overcoming aberrant androgen receptor (AR) signaling [abstract]. In: Proceedings of the AACR Special Conference: Prostate Cancer: Advances in Basic, Translational, and Clinical Research; 2017 Dec 2-5; Orlando, Florida. Philadelphia (PA): AACR; Cancer Res 2018;78(16 Suppl):Abstract nr A067.

Abstract 776: Utilising genetic susceptibility and big data to inform novel cancer therapies

Despite a move towards personalised medicine, attrition rates for new cancer drugs remain unacceptably high. The pharmaceutical industry has also shown a preference for well-studied targets and pathways, as evidenced by ‘me too' drugs. Together with the challenge of inadequate pre-clinical models, this indicates a need for novel, evidence-based therapeutic targets. Genome-wide association studies (GWAS) have identified over 450 robust genetic variants associated with increased cancer risk. Genes implicated through GWAS are often mutated somatically and therefore represent attractive therapeutic targets. Examples include the target of venetoclax in chronic lymphocytic leukaemia, BCL2.

We exploit this principle more generally by integrating genetic associations for common cancers with drug target data and druggability using the canSAR drug discovery knowledgebase (https://cansar.icr.ac.uk). By harnessing the power of Big Data we aim to both identify opportunities for repurposing of existing drugs, and prioritise novel targets for cancer drug discovery.

We mined the NHGRI-EBI Catalog of published GWAS for all cancer risk SNPs. We annotated candidate target genes through overlapping topologically associating domains (TADs), a more sensitive technique than previously published methods using linkage disequilibrium. We used canSAR to identify target genes for which there is no FDA-approved small molecule drug, and the resource Probe Miner to identify targets for which high-quality chemical probes exist. We also utilised canSAR's machine learning algorithms to assess the druggability of target genes by structure-, ligand-, precedence- and network-based approaches.

We additionally analysed results from cancer drug databases to ascertain whether there is an enrichment of ‘drug target-indication pairs' at successive stages of the drug development pathway for which supporting evidence from GWAS exists: this indicates potential ‘stumbling blocks' that may present a risk for future drug development projects.

7 257 protein-coding genes mapped within TADs overlapping cancer risk SNPs. Of these, 98 were pre-existing targets for which there is an FDA-approved small molecule drug. For the remaining 7 159 genes we performed multi-faceted druggability analyses incorporating assessments of the 3D structure of the target and any protein complexes it exists in, chemical properties of known ligands of the target, and the target's position and role within the human interactome. We comprehensively rank our target-indication pairings by criteria including novelty relative to existing targets and predicted attrition risk.

Mapping approved drug targets back to cancer GWAS signals enables identification of both novel drug targets and patient populations. Collectively our findings show the value of investigating germline cancer genetics as part of interdisciplinary, data-driven approaches to inform drug discovery.

Citation Format: Elizabeth A. Coker, Ben Kinnersley, Amit Sud, Patrizio Di Micco, Bissan Al-Lazikani, Richard Houlston. Utilising genetic susceptibility and big data to inform novel cancer therapies [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 776.

Sequencing of prostate cancers identifies new cancer genes, routes of progression and drug targets

Prostate cancer represents a substantial clinical challenge because it is difficult to predict outcome and advanced disease is often fatal. We sequenced the whole genomes of 112 primary and metastatic prostate cancer samples. From joint analysis of these cancers with those from previous studies (930 cancers in total), we found evidence for 22 previously unidentified putative driver genes harboring coding mutations, as well as evidence for NEAT1 and FOXA1 acting as drivers through noncoding mutations. Through the temporal dissection of aberrations, we identified driver mutations specifically associated with steps in the progression of prostate cancer, establishing, for example, loss of CHD1 and BRCA2 as early events in cancer development of ETS fusion-negative cancers. Computational chemogenomic (canSAR) analysis of prostate cancer mutations identified 11 targets of approved drugs, 7 targets of investigational drugs, and 62 targets of compounds that may be active and should be considered candidates for future clinical trials.

Objective, Quantitative, Data-Driven Assessment of Chemical Probes

Chemical probes are essential tools for understanding biological systems and for target validation, yet selecting probes for biomedical research is rarely based on objective assessment of all potential compounds. Here, we describe the Probe Miner: Chemical Probes Objective Assessment resource, capitalizing on the plethora of public medicinal chemistry data to empower quantitative, objective, data-driven evaluation of chemical probes. We assess >1.8 million compounds for their suitability as chemical tools against 2,220 human targets and dissect the biases and limitations encountered. Probe Miner represents a valuable resource to aid the identification of potential chemical probes, particularly when used alongside expert curation.

Abstract B096: canSAR, a cancer research and drug discovery knowledgebase

canSAR (http://cansar.icr.ac.uk) is a freely available, multidisciplinary, cancer-focused knowledgebase developed to bring together information from genomic, transcriptomic, protein, pathway, chemical, pharmacologic, and 3D structural data. canSAR provides a powerful, uniqu,e and user-friendly portal to enable translational research and to help generate and test hypotheses and support scientific decision-making in drug discovery both before and after target selection. With its three alternative approaches to examine druggability, canSAR represents the most comprehensive public druggability assessment resource. canSAR provides 3D-structure-based druggability assessment for more than 3,100,000 cavities on more than 391,000 protein chains; ligand-based druggability assessment for 8,197 human proteins; and, more recently, protein network-based druggability results for 13,345 human proteins. Together these provide a powerful enabler for target selection and validation for drug discovery. Druggability assessments are presented alongside data from resources including 224,000 clinical trials from ClinicalTrials.gov, drug indications from Cancer.gov, target gene expression from TCGA for cell lines, and patient samples to provide a detailed picture of the target’s biologic context. Recent updates to canSAR include integration of ChEMBL 23 and additional curation of a protein-protein interaction network of 13,500 nodes. canSAR is currently used by more than 65,000 users annually from both academia and industry, and we will illustrate how canSAR can empower decision making in translational drug discovery.

Citation Format: Elizabeth A. Coker, Patrizio Di Micco, Joesph E. Tym, Costas Mitsopoulos, Angeliki Komianou, Albert A. Antolin, Bissan Al-Lazikani. canSAR, a cancer research and drug discovery knowledgebase [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2017 Oct 26-30; Philadelphia, PA. Philadelphia (PA): AACR; Mol Cancer Ther 2018;17(1 Suppl):Abstract nr B096.

Abstract A024: Probe Miner: objective, quantitative, data-driven assessment of chemical probes for target validation

Chemical probes are important, widely used reagents for understanding biologic systems and for target validation. However, selection of chemical probes is largely subjective and prone to historical and commercial biases. Despite many publications discussing the aspirational properties of chemical probes and the proposal of "fitness factors" to be considered when assessing chemical tools, scientists often select probes through web-based searchers or previous literature that are heavily biased towards older and often flawed probes or use vendor catalogues that do not discriminate between probes. Here, we analyze the scope and quality of published bioactive molecules and uncover large biases and limitations of chemical tools in public databases that need to be urgently addressed and should be always considered when using chemical tools. We also provide the online Probe Miner resource (http://probeminer.icr.ac.uk) capitalizing on the plethora of public pharmacologic data to enable quantitative, unbiased, objective, Big Data-driven assessment of chemical probes and complement expert-curated approaches. We assess &gt;1.8m compounds for their suitability as chemical tools against 2,220 human targets, demonstrating that large-scale public data can contribute to improving chemical probe assessment and prioritization to empower researchers in the selection of chemical tools for biomedical research and target validation.

Citation Format: Albert A. Antolin, Joe E. Tym, Angeliki Komianou, Ian Collins, Paul Workman, Bissan Al-Lazikani. Probe Miner: objective, quantitative, data-driven assessment of chemical probes for target validation [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2017 Oct 26-30; Philadelphia, PA. Philadelphia (PA): AACR; Mol Cancer Ther 2018;17(1 Suppl):Abstract nr A024.

Development of Bag-1L as a therapeutic target in androgen receptor-dependent prostate cancer

Targeting the activation function-1 (AF-1) domain located in the N-terminus of the androgen receptor (AR) is an attractive therapeutic alternative to the current approaches to inhibit AR action in prostate cancer (PCa). Here we show that the AR AF-1 is bound by the cochaperone Bag-1L. Mutations in the AR interaction domain or loss of Bag-1L abrogate AR signaling and reduce PCa growth. Clinically, Bag-1L protein levels increase with progression to castration-resistant PCa (CRPC) and high levels of Bag-1L in primary PCa associate with a reduced clinical benefit from abiraterone when these tumors progress. Intriguingly, residues in Bag-1L important for its interaction with the AR AF-1 are within a potentially druggable pocket, implicating Bag-1L as a potential therapeutic target in PCa.

Abstract 996: A translational phosphoproteomic approach to study differences in KRAS signaling in pancreatic, colorectal and lung cancers

Aims To understand any context-dependent differences in signaling pathways between pancreatic (PAN), colorectal (CR) and lung (LU) cancers with KRAS mutations using a targeted phosphoproteomic approach in cell lines and patient-derived cancer cells exposed to targeted anticancer drugs ex-vivo.

Materials and Methods We studied a panel of 30 KRAS mutant cell lines: 10 PAN, 10 CR and 10 LU cell lines. Cancer cells were also immuno-magnetically isolated from pleural effusions and ascites of patients with KRAS mutant CR and LU cancer and exposed to a DMSO control and clinically relevant concentrations of PI3K (pictilisib), AKT (AZD5363), mTOR (everolimus), EGFR (gefitinib), BRAF (vemurafenib), MEK (trametinib) and HSP90 (luminespib) inhibitors for 1 hr. Dynamic changes in a panel of 52 relevant phosphoproteins were studied using the Luminex 200 platform. Hierarchical clustering and logistic regression were used to find differences in dynamic changes in phosphoproteins between KRAS mutant, PAN, CR and LU cancer cells.

Results Supervised clustering studying exposure to different drugs revealed that when exposed to the PI3K inhibitor, pictilisib, KRAS mutant LU cancers did not significantly cluster together; p=0.008, p=0.104 following Benjamini-Hochberg correction. Independently, logistic regression showed significant differences in signaling of KRAS mutant cells when exposed to the PI3K inhibitor, pictilisib. PAN and CR cancers showed an increase in p-MEK while LU cancer cells did not; p=0.0195. LU cancer cell lines showed significantly more reduction of p-AKT compared to PAN and CR cell lines when exposed to the PI3K inhibitor; p=0.0423. As expected, exposure to vemurafenib increased p-MEK levels across the majority of the KRAS mutant cell lines, however compensatory reductions in p-mTOR levels were seen significantly more in PAN and CR cell lines and not in LU cell lines; p=0.0084. The dynamic phosphoprotein changes caused by pictilisib were validated in cancer cells isolated from serous effusions of 3 KRAS mutant LU and 4 KRAS mutant CR cancer patients. Validation of these findings using multiple other inhibitors and time-points is ongoing.

Interpretation/conclusions We hypothesise that the significantly greater reduction in p-AKT and less increase of compensatory p-MEK caused by PI3K inhibition in KRAS mutant LU cells compared to KRAS mutant PAN and CR cell lines represents preferential signaling of these cells through the PI3K pathway. Increase in p-MEK driven by BRAF inhibitors caused a reduction in p-mTOR in PAN and CR but not in LU cell lines also indicating preferential dependence of signaling in KRAS mutant lung cancer cells through the PI3K pathway. These findings are important while designing clinical trials of KRAS mutant cancers and more broadly to precision medicine where mutation status independent of tissue context is often used.

Citation Format: Adam Stewart, Elizabeth A. Coker, Anna Minchom, Sebastian Pölsterl, Alexandros Georgiou, Paul Huang, Bissan Al-Lazikani, Udai Banerji. A translational phosphoproteomic approach to study differences in KRAS signaling in pancreatic, colorectal and lung cancers [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 996. doi:10.1158/1538-7445.AM2017-996

SiGNet: A signaling network data simulator to enable signaling network inference

Network models are widely used to describe complex signaling systems. Cellular wiring varies in different cellular contexts and numerous inference techniques have been developed to infer the structure of a network from experimental data of the network's behavior. To objectively identify which inference strategy is best suited to a specific network, a gold standard network and dataset are required. However, suitable datasets for benchmarking are difficult to find. Numerous tools exist that can simulate data for transcriptional networks, but these are of limited use for the study of signaling networks. Here, we describe SiGNet (Signal Generator for Networks): a Cytoscape app that simulates experimental data for a signaling network of known structure. SiGNet has been developed and tested against published experimental data, incorporating information on network architecture, and the directionality and strength of interactions to create biological data in silico. SiGNet is the first tool to simulate biological signaling data, enabling an accurate and systematic assessment of inference strategies. SiGNet can also be used to produce preliminary models of key biological pathways following perturbation.

Rational design of non-resistant targeted cancer therapies

Drug resistance is one of the major problems in targeted cancer therapy. A major cause of resistance is changes in the amino acids that form the drug-target binding site. Despite of the numerous efforts made to individually understand and overcome these mutations, there is a lack of comprehensive analysis of the mutational landscape that can prospectively estimate drug-resistance mutations. Here we describe and computationally validate a framework that combines the cancer-specific likelihood with the resistance impact to enable the detection of single point mutations with the highest chance to be responsible of resistance to a particular targeted cancer therapy. Moreover, for these treatment-threatening mutations, the model proposes alternative therapies overcoming the resistance. We exemplified the applicability of the model using EGFR-gefitinib treatment for Lung Adenocarcinoma (LUAD) and Lung Squamous Cell Cancer (LSCC) and the ERK2-VTX11e treatment for melanoma and colorectal cancer. Our model correctly identified the phenotype known resistance mutations, including the classic EGFR-T790M and the ERK2-P58L/S/T mutations. Moreover, the model predicted new previously undescribed mutations as potentially responsible of drug resistance. Finally, we provided a map of the predicted sensitivity of alternative ERK2 and EGFR inhibitors, with a particular highlight of two molecules with a low predicted resistance impact.

A comprehensive map of molecular drug targets

The success of mechanism-based drug discovery depends on the definition of the drug target. This definition becomes even more important as we try to link drug response to genetic variation, understand stratified clinical efficacy and safety, rationalize the differences between drugs in the same therapeutic class and predict drug utility in patient subgroups. However, drug targets are often poorly defined in the literature, both for launched drugs and for potential therapeutic agents in discovery and development. Here, we present an updated comprehensive map of molecular targets of approved drugs. We curate a total of 893 human and pathogen-derived biomolecules through which 1,578 US FDA-approved drugs act. These biomolecules include 667 human-genome-derived proteins targeted by drugs for human disease. Analysis of these drug targets indicates the continued dominance of privileged target families across disease areas, but also the growth of novel first-in-class mechanisms, particularly in oncology. We explore the relationships between bioactivity class and clinical success, as well as the presence of orthologues between human and animal models and between pathogen and human genomes. Through the collaboration of three independent teams, we highlight some of the ongoing challenges in accurately defining the targets of molecular therapeutics and present conventions for deconvoluting the complexities of molecular pharmacology and drug efficacy.

Blocking the survival of the nastiest by HSP90 inhibition

It is now recognised that genetic, epigenetic and phenotypic heterogeneity within individual human cancers is responsible for therapeutic resistance – knowledge that is having a profound impact on current thinking and experimentation. There has been concern that molecularly targeted therapy is doomed to failure, with resistant clones emerging in response to the Darwinian selective pressure of any drug treatment. However, two studies have shown that the evolution of drug resistance can be restrained by co-administration of a pharmacologic inhibitor of the HSP90 molecular chaperone.

Abstract 4383: SOCRATES: integrating ex vivo and in silico analysis to identify optimal drug combinations for patients

Recent work on cancer heterogeneity and evolution has greatly enhanced our understanding of drivers of drug resistance, but this has also highlighted the need to move away from generic, “one-size-fits-all” treatment regimes. It is clear that we should tailor therapy to individual patients at different stages of disease based on the current behaviour of their tumour cells, yet detailed experimental characterisation and analysis of patient tissue is impractical and costly. Predictive computational models that can produce patient-specific recommendations of drug combinations would therefore be of great value to the field of personalised medicine. In silico modelling has already been used to successfully predict synergistic drug combinations in single cell lines in the lab, but such approaches have not yet been translated to dynamic, real patient data in the clinic.

The SOCRATES project (Signalling output to rationalise combinations of targeted anticancer therapies) integrates experimental data and computational analysis to predict context-specific synergistic drug combinations. Here we present ambitious computational modelling of dynamic signalling network changes in response to ex vivo exposure of seven targeted drugs used singly at clinically appropriate concentrations. We also apply our adaptive modelling to genomics and proteomics data from 54 proteins in over 30 non small-cell lung cancer cell lines and ten patient samples to predict and experimentally validate personalised, context specific drug combinations. We will describe our adaptive, in silico modelling for predicting response to drug combinations, including novel methods utilizing network topology. We have predicted over 50 synergistic drug combinations and will present our predictions alongside initial results of experimental validation in the lab. We will also illustrate the differences observed in various cell lines and patient derived cells depending on genetic context.

SOCRATES is, to our knowledge, the first project of its kind and is a prototype for the future of adaptive, individualised cancer treatment.

Citation Format: Elizabeth A. Coker, Adam Stewart, Anna Minchom, Mary O’Brien, Timothy Yap, Paul Workman, Udai Banerji, Bissan Al-Lazikani. SOCRATES: integrating ex vivo and in silico analysis to identify optimal drug combinations for patients. [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 4383.

Abstract 3099: KRAS and clinical context: Differential dynamic signaling output of KRAS mutant lung, colorectal and pancreatic cancer cell lines when exposed to targeted anticancer drugs

Background

Clinical trials have shown that cancers originating from different tissues driven by the same oncogene respond differently to targeted anticancer drugs. We aimed to understand different signaling patterns in KRAS mutant cells derived from non-small cell lung cancer (NSCLC), colorectal cancer (CRC) and pancreatic cancer.

Materials and methods

We optimized a 50 phosphoprotein antibody-based assay on the Luminex 200 platform. We then exposed a panel of 15 KRAS mutant cell lines (5 cell lines each originating in the lung, pancreas and colon) to a DMSO control (n = 3) and clinically significant concentrations (Cmax achieved in humans adjusted for protein binding in culture medium) of a PI3K (GDC-0941), AKT (AZD5363), m-TOR (everolimus), BRAF (vemurafenib), EGFR (gefitinib), MEK (trametinib) and an HSP90 inhibitor (luminespib) for 1 hr. We quantified the change in phosphorylation of proteins for each drug compared to control. Logistic regression analysis was used to analyse differences between KRAS-driven cell lines originating from different anatomical sites.

Results

There were changes in phosphorylation related to the pharmacodynamic effects of the drug independent of cell line of origin; however, there were interesting differences between KRAS mutant cells originating from different anatomical sites. In NSCLC cell lines, p-EGFR levels changed significantly less when exposed to PI3K, AKT and m-TOR inhibitors (p = 0.047, 0.022 and 0.047, respectively) when compared to cells originating from CRC and pancreatic cancer. CRC cell lines, when compared to NSCLC and pancreatic cancer cell lines, showed significantly less changes in phosphorylation of key cell cycle regulators such as CHK1 when exposed to PI3K, AKT and m-TOR inhibitors, (p = 0.001, 0.047 and 0.047, respectively) and RB when exposed to an AKT and m-TOR inhibitor (p = 0.047 and 0.047, respectively). Interestingly, pancreatic cell lines showed significantly more changes in p-m-TOR compared to CRC and NSCLC cell lines following exposure to PI3K and AKT inhibitors (p = 0.0095 and 0.022, respectively). Of note, drugs not directly targeting the PI3K pathway differentially regulated different nodes in the PI3K pathway, for example, BRAF inhibitors significantly differentially changed levels of phosphorylation at different nodes in the PI3K pathway such as AKT in NSCLC cell lines, p = 0.047, p70S6K in CRC cell lines, p = 0.0472 and PRAS40 in the pancreatic cancer cell lines, p = 0.022.

Conclusion

These results suggest that there are significant differences in signaling patterns caused by PI3K pathway inhibitors in KRAS mutant NSCLC, CRC and pancreatic cancer cell lines. Our findings shed light on the putative use of PI3K pathway inhibitors in KRAS mutant cancers. They also question the universal application of solely using genetic mutations to stratify patients in ‘basket’ clinical studies.

Citation Format: Adam Stewart, Elizabeth Coker, Anna Minchom, Parames Thavasu, Alexandros Georgiou, Anguraj Sadanandam, Timothy A. Yap, Johann S. de Bono, Bissan Al-Lazikani, Udai Banerji. KRAS and clinical context: Differential dynamic signaling output of KRAS mutant lung, colorectal and pancreatic cancer cell lines when exposed to targeted anticancer drugs. [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 3099.

Large-Scale Profiling of Kinase Dependencies in Cancer Cell Lines

One approach to identifying cancer-specific vulnerabilities and therapeutic targets is to profile genetic dependencies in cancer cell lines. Here, we describe data from a series of siRNA screens that identify the kinase genetic dependencies in 117 cancer cell lines from ten cancer types. By integrating the siRNA screen data with molecular profiling data, including exome sequencing data, we show how vulnerabilities/genetic dependencies that are associated with mutations in specific cancer driver genes can be identified. By integrating additional data sets into this analysis, including protein-protein interaction data, we also demonstrate that the genetic dependencies associated with many cancer driver genes form dense connections on functional interaction networks. We demonstrate the utility of this resource by using it to predict the drug sensitivity of genetically or histologically defined subsets of tumor cell lines, including an increased sensitivity of osteosarcoma cell lines to FGFR inhibitors and SMAD4 mutant tumor cells to mitotic inhibitors.

canSAR: an updated cancer research and drug discovery knowledgebase

canSAR (http://cansar.icr.ac.uk) is a publicly available, multidisciplinary, cancer-focused knowledgebase developed to support cancer translational research and drug discovery. canSAR integrates genomic, protein, pharmacological, drug and chemical data with structural biology, protein networks and druggability data. canSAR is widely used to rapidly access information and help interpret experimental data in a translational and drug discovery context. Here we describe major enhancements to canSAR including new data, improved search and browsing capabilities, new disease and cancer cell line summaries and new and enhanced batch analysis tools.

Polypharmacology in Precision Oncology: Current Applications and Future Prospects

Over the past decade, a more comprehensive, large-scale approach to studying cancer genetics and biology has revealed the challenges of tumor heterogeneity, adaption, evolution and drug resistance, while systems-based pharmacology and chemical biology strategies have uncovered a much more complex interaction between drugs and the human proteome than was previously anticipated. In this mini-review we assess the progress and potential of drug polypharmacology in biomarker-driven precision oncology. Polypharmacology not only provides great opportunities for drug repurposing to exploit off-target effects in a new single-target indication but through simultaneous blockade of multiple targets or pathways offers exciting opportunities to slow, overcome or even prevent inherent or adaptive drug resistance. We highlight the many challenges associated with exploiting known or desired polypharmacology in drug design and development, and assess computational and experimental methods to uncover unknown polypharmacology. A comprehensive understanding of the intricate links between polypharmacology, efficacy and safety is urgently needed if we are to tackle the enduring challenge of cancer drug resistance and to fully exploit polypharmacology for the ultimate benefit of cancer patients.

Distinctive Behaviors of Druggable Proteins in Cellular Networks

The interaction environment of a protein in a cellular network is important in defining the role that the protein plays in the system as a whole, and thus its potential suitability as a drug target. Despite the importance of the network environment, it is neglected during target selection for drug discovery. Here, we present the first systematic, comprehensive computational analysis of topological, community and graphical network parameters of the human interactome and identify discriminatory network patterns that strongly distinguish drug targets from the interactome as a whole. Importantly, we identify striking differences in the network behavior of targets of cancer drugs versus targets from other therapeutic areas and explore how they may relate to successful drug combinations to overcome acquired resistance to cancer drugs. We develop, computationally validate and provide the first public domain predictive algorithm for identifying druggable neighborhoods based on network parameters. We also make available full predictions for 13,345 proteins to aid target selection for drug discovery. All target predictions are available through canSAR.icr.ac.uk. Underlying data and tools are available at https://cansar.icr.ac.uk/cansar/publications/druggable_network_neighbourhoods/.

The need for well-validated targets for drug discovery is more pressing than ever, especially in cancer in view of resistance to current therapeutics coupled with late stage drug failures. Target prioritization and selection methodologies have typically not taken the protein interaction environment into account. Here we analyze a large representation of the human interactome comprising almost 90,000 interactions between 13,345 proteins. We assess these interactions using an extensive set of topological, graphical and community parameters, and we identify behaviors that distinguish the protein interaction environments of drug targets from the general interactome. Moreover, we identify clear distinctions between the network environment of cancer-drug targets and targets from other therapeutics areas. We use these distinguishing properties to build a predictive methodology to prioritize potential drug targets based on network parameters alone and we validate our predictive models using current FDA-approved drug targets. Our models provide an objective, interactome-based target prioritization methodology to complement existing structure-based and ligand-based prioritization methods. We provide our interactome-based predictions alongside other druggability predictors within the public canSAR resource (cansar.icr.ac.uk).

Therapeutic opportunities within the DNA damage response

The DNA damage response (DDR) is essential for maintaining the genomic integrity of the cell, and its disruption is one of the hallmarks of cancer. Classically, defects in the DDR have been exploited therapeutically in the treatment of cancer with radiation therapies or genotoxic chemotherapies. More recently, protein components of the DDR systems have been identified as promising avenues for targeted cancer therapeutics. Here, we present an in-depth analysis of the function, role in cancer and therapeutic potential of 450 expert-curated human DDR genes. We discuss the DDR drugs that have been approved by the US Food and Drug Administration (FDA) or that are under clinical investigation. We examine large-scale genomic and expression data for 15 cancers to identify deregulated components of the DDR, and we apply systematic computational analysis to identify DDR proteins that are amenable to modulation by small molecules, highlighting potential novel therapeutic targets.

Abstract 2730: RNAi knockdown or chemical inhibition of anaphase-promoting complex components is synthetic lethal with HSP90 inhibition

The molecular chaperone heat shock protein 90 (HSP90) maintains the conformation, stability and function of oncogenic client proteins, many of which are mutated or overexpressed. Therefore, HSP90 is an important cancer therapeutic target. To further increase the efficacy of HSP90 inhibitors, combinatorial therapeutic strategies may be beneficial. Here we report an unbiased global screening approach to identify genes that encode potentially druggable proteins that modulate cellular responses to HSP90 inhibition. From the 7,593 genes that were tested, three components of the anaphase-promoting complex (APC/C) were among the genes which when silenced caused significant HSP90 inhibitor sensitisation. siRNA knockdown of ANAPC1, -8 or -4 produced up to three-fold sensitisation to HSP90 inhibitors 17-AAG and AUY922 in human colon cancer cells. Intriguingly, we show that combinatorial siRNA-mediated ANAPC1 knockdown and HSP90 inhibition induced an accumulation of mitotic cells exhibiting a tri-polar spindle formation. A marked increase in cyclin B1 and aurora A protein levels was observed in response to combinatorial ANAPC1 knockdown and HSP90 inhibition. These effects may have been responsible for the observed formation of aberrations in the mitotic spindle architecture. Furthermore, after a period of prolonged mitotic arrest, cancer cells treated with combinational ANAPC1 knockdown and HSP90 inhibition were committed to apoptosis. Mimicking siRNA knockdown, treatment with the APC/C inhibitor proTAME synergistically sensitised human colon carcinoma cells to HSP90 inhibition. These exciting results provide a mechanistic explanation underlying the HSP90 inhibitor sensitisation phenotype, and suggest that the APC/C may represent an interesting therapeutic target to exploit in combination with HSP90 inhibitors. Furthermore, therapeutic approaches may be developed which exploit the vulnerability of mitotic cancer cells in the context of HSP90 inhibition.

Citation Format: Jennifer Howes, Bing-Feng Lu, Marissa Powers, Costas Mitsopoulos, Bissan Al-Lazikani, Spiros Linardopoulos, Paul Clarke, Paul Workman. RNAi knockdown or chemical inhibition of anaphase-promoting complex components is synthetic lethal with HSP90 inhibition. [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 2730. doi:10.1158/1538-7445.AM2014-2730