Enhancing the Small-Scale Screenable Biological Space beyond Known Chemogenomics Libraries with Gray Chemical Matter─Compounds with Novel Mechanisms from High-Throughput Screening Profiles

Phenotypic assays have become an established approach to drug discovery. Greater disease relevance is often achieved through cellular models with increased complexity and more detailed readouts, such as gene expression or advanced imaging. However, the intricate nature and cost of these assays impose limitations on their screening capacity, often restricting screens to well-characterized small compound sets such as chemogenomics libraries. Here, we outline a cheminformatics approach to identify a small set of compounds with likely novel mechanisms of action (MoAs), expanding the MoA search space for throughput limited phenotypic assays. Our approach is based on mining existing large-scale, phenotypic high-throughput screening (HTS) data. It enables the identification of chemotypes that exhibit selectivity across multiple cell-based assays, which are characterized by persistent and broad structure activity relationships (SAR). We validate the effectiveness of our approach in broad cellular profiling assays (Cell Painting, DRUG-seq, and Promotor Signature Profiling) and chemical proteomics experiments. These experiments revealed that the compounds behave similarly to known chemogenetic libraries, but with a notable bias toward novel protein targets. To foster collaboration and advance research in this area, we have curated a public set of such compounds based on the PubChem BioAssay dataset and made it available for use by the scientific community.

Exquisite Sensitivity to Dual BRG1/BRM ATPase Inhibitors Reveals Broad SWI/SNF Dependencies in Acute Myeloid Leukemia

Various subunits of mammalian SWI/SNF chromatin remodeling complexes display loss-of-function mutations characteristic of tumor suppressors in different cancers, but an additional role for SWI/SNF supporting cell survival in distinct cancer contexts is emerging. In particular, genetic dependence on the catalytic subunit BRG1/SMARCA4 has been observed in acute myeloid leukemia (AML), yet the feasibility of direct therapeutic targeting of SWI/SNF catalytic activity in leukemia remains unknown. Here, we evaluated the activity of dual BRG1/BRM ATPase inhibitors across a genetically diverse panel of cancer cell lines and observed that hematopoietic cancer cell lines were among the most sensitive compared to other lineages. This result was striking in comparison to data from pooled short hairpin RNA screens, which showed that only a subset of leukemia cell lines display sensitivity to BRG1 knockdown. We demonstrate that combined genetic knockdown of BRG1 and BRM is required to recapitulate the effects of dual inhibitors, suggesting that SWI/SNF dependency in human leukemia extends beyond a predominantly BRG1-driven mechanism. Through gene expression and chromatin accessibility studies, we show that the dual inhibitors act at genomic loci associated with oncogenic transcription factors, and observe a downregulation of leukemic pathway genes including MYC, a well-established target of BRG1 activity in AML. Overall, small molecule inhibition of BRG1/BRM induced common transcriptional responses across leukemia models resulting in a spectrum of cellular phenotypes. Implications: Our studies reveal the breadth of SWI/SNF dependency in leukemia and support targeting SWI/SNF catalytic function as a potential therapeutic strategy in AML.

CYP27A1-dependent anti-melanoma activity of limonoid natural products targets mitochondrial metabolism

Three limonoid natural products with selective anti-proliferative activity against BRAF(V600E) and NRAS(Q61K)-mutation-dependent melanoma cell lines were identified. Differential transcriptome analysis revealed dependency of compound activity on expression of the mitochondrial cytochrome P450 oxidase CYP27A1, a transcriptional target of melanogenesis-associated transcription factor (MITF). We determined that CYP27A1 activity is necessary for the generation of a reactive metabolite that proceeds to inhibit cellular proliferation. A genome-wide small interfering RNA screen in combination with chemical proteomics experiments revealed gene-drug functional epistasis, suggesting that these compounds target mitochondrial biogenesis and inhibit tumor bioenergetics through a covalent mechanism. Our work suggests a strategy for melanoma-specific targeting by exploiting the expression of MITF target gene CYP27A1 and inhibiting mitochondrial oxidative phosphorylation in BRAF mutant melanomas.

Systematic Chemogenetic Library Assembly

Chemogenetic libraries, collections of well-defined chemical probes, provide tremendous value to biomedical research but require substantial effort to ensure diversity as well as quality of the contents. We have assembled a chemogenetic library by data mining and crowdsourcing institutional expertise. We are sharing our approach, lessons learned, and disclosing our current collection of 4,185 compounds with their primary annotated gene targets (https://github.com/Novartis/MoaBox). This physical collection is regularly updated and used broadly both within Novartis and in collaboration with external partners.

Gene Expression Ratios Lead to Accurate and Translatable Predictors of DR5 Agonism across Multiple Tumor Lineages

Death Receptor 5 (DR5) agonists demonstrate anti-tumor activity in preclinical models but have yet to demonstrate robust clinical responses. A key limitation may be the lack of patient selection strategies to identify those most likely to respond to treatment. To overcome this limitation, we screened a DR5 agonist Nanobody across >600 cell lines representing 21 tumor lineages and assessed molecular features associated with response. High expression of DR5 and Casp8 were significantly associated with sensitivity, but their expression thresholds were difficult to translate due to low dynamic ranges. To address the translational challenge of establishing thresholds of gene expression, we developed a classifier based on ratios of genes that predicted response across lineages. The ratio classifier outperformed the DR5+Casp8 classifier, as well as standard approaches for feature selection and classification using genes, instead of ratios. This classifier was independently validated using 11 primary patient-derived pancreatic xenograft models showing perfect predictions as well as a striking linearity between prediction probability and anti-tumor response. A network analysis of the genes in the ratio classifier captured important biological relationships mediating drug response, specifically identifying key positive and negative regulators of DR5 mediated apoptosis, including DR5, CASP8, BID, cFLIP, XIAP and PEA15. Importantly, the ratio classifier shows translatability across gene expression platforms (from Affymetrix microarrays to RNA-seq) and across model systems (in vitro to in vivo). Our approach of using gene expression ratios presents a robust and novel method for constructing translatable biomarkers of compound response, which can also probe the underlying biology of treatment response.

The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity

The systematic translation of cancer genomic data into knowledge of tumour biology and therapeutic possibilities remains challenging. Such efforts should be greatly aided by robust preclinical model systems that reflect the genomic diversity of human cancers and for which detailed genetic and pharmacological annotation is available. Here we describe the Cancer Cell Line Encyclopedia (CCLE): a compilation of gene expression, chromosomal copy number and massively parallel sequencing data from 947 human cancer cell lines. When coupled with pharmacological profiles for 24 anticancer drugs across 479 of the cell lines, this collection allowed identification of genetic, lineage, and gene-expression-based predictors of drug sensitivity. In addition to known predictors, we found that plasma cell lineage correlated with sensitivity to IGF1 receptor inhibitors; AHR expression was associated with MEK inhibitor efficacy in NRAS-mutant lines; and SLFN11 expression predicted sensitivity to topoisomerase inhibitors. Together, our results indicate that large, annotated cell-line collections may help to enable preclinical stratification schemata for anticancer agents. The generation of genetic predictions of drug response in the preclinical setting and their incorporation into cancer clinical trial design could speed the emergence of 'personalized' therapeutic regimens.

The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity

The systematic translation of cancer genomic data into knowledge of tumour biology and therapeutic possibilities remains challenging. Such efforts should be greatly aided by robust preclinical model systems that reflect the genomic diversity of human cancers and for which detailed genetic and pharmacological annotation is available. Here we describe the Cancer Cell Line Encyclopedia (CCLE): a compilation of gene expression, chromosomal copy number and massively parallel sequencing data from 947 human cancer cell lines. When coupled with pharmacological profiles for 24 anticancer drugs across 479 of the cell lines, this collection allowed identification of genetic, lineage, and gene-expression-based predictors of drug sensitivity. In addition to known predictors, we found that plasma cell lineage correlated with sensitivity to IGF1 receptor inhibitors; AHR expression was associated with MEK inhibitor efficacy in NRAS-mutant lines; and SLFN11 expression predicted sensitivity to topoisomerase inhibitors. Together, our results indicate that large, annotated cell-line collections may help to enable preclinical stratification schemata for anticancer agents. The generation of genetic predictions of drug response in the preclinical setting and their incorporation into cancer clinical trial design could speed the emergence of 'personalized' therapeutic regimens.

Pathway Reporter Assays Reveal Small Molecule Mechanisms of Action

Cell-based, phenotypic screening of small molecules often identifies compounds with provocative biological properties. However, determining the cellular target(s) and/or mechanism of action (MoA) of lead compounds remains an extremely challenging and time-consuming exercise. To provide insights into a compound's cellular action and greatly reduce the time required for MoA determination, we have developed a screening platform consisting of an extensive series of reporter gene assays (RGAs). A collection of > 11,000 compounds of known MoA (e.g., World Drug Index entries) were screened against the entire panel. The output provided evidence that an RGA signature could be ascribed to numerous, biologically diverse MoAs. The reference database generated suggested novel biological activity for particular compounds. For example, the profiling data led to the prediction that the cellular target of the natural product terprenin was dihydroorotate dehydrogenase (DHODH), which was confirmed experimentally. The screening methodology developed for this endeavor renders it amenable to the future examination of compounds with unknown MoA, in an automated, inexpensive, and time-efficient manner.

Presenilin-based genetic screens in Drosophila melanogaster identify novel notch pathway modifiers

Presenilin is the enzymatic component of gamma-secretase, a multisubunit intramembrane protease that processes several transmembrane receptors, such as the amyloid precursor protein (APP). Mutations in human Presenilins lead to altered APP cleavage and early-onset Alzheimer's disease. Presenilins also play an essential role in Notch receptor cleavage and signaling. The Notch pathway is a highly conserved signaling pathway that functions during the development of multicellular organisms, including vertebrates, Drosophila, and C. elegans. Recent studies have shown that Notch signaling is sensitive to perturbations in subcellular trafficking, although the specific mechanisms are largely unknown. To identify genes that regulate Notch pathway function, we have performed two genetic screens in Drosophila for modifiers of Presenilin-dependent Notch phenotypes. We describe here the cloning and identification of 19 modifiers, including nicastrin and several genes with previously undescribed involvement in Notch biology. The predicted functions of these newly identified genes are consistent with extracellular matrix and vesicular trafficking mechanisms in Presenilin and Notch pathway regulation and suggest a novel role for gamma-tubulin in the pathway.

High-throughput gene mapping in Caenorhabditis elegans

Positional cloning of mutations in model genetic systems is a powerful method for the identification of targets of medical and agricultural importance. To facilitate the high-throughput mapping of mutations in Caenorhabditis elegans, we have identified a further 9602 putative new single nucleotide polymorphisms (SNPs) between two C. elegans strains, Bristol N2 and the Hawaiian mapping strain CB4856, by sequencing inserts from a CB4856 genomic DNA library and using an informatics pipeline to compare sequences with the canonical N2 genomic sequence. When combined with data from other laboratories, our marker set of 17,189 SNPs provides even coverage of the complete worm genome. To date, we have confirmed >1099 evenly spaced SNPs (one every 91 +/- 56 kb) across the six chromosomes and validated the utility of our SNP marker set and new fluorescence polarization-based genotyping methods for systematic and high-throughput identification of genes in C. elegans by cloning several proprietary genes. We illustrate our approach by recombination mapping and confirmation of the mutation in the cloned gene, dpy-18.

Single Nucleotide Polymorphism Markers for Genetic Mapping in Drosophila melanogaster

For nearly a century, genetic analysis in Drosophila melanogaster has been a powerful tool for analyzing gene function, yet Drosophila lacks the molecular genetic mapping tools that recently have revolutionized human, mouse, and plant genetics. Here, we describe the systematic characterization of a dense set of molecular markers in Drosophila by using a sequence tagged site-based physical map of the genome. We identify 474 biallelic markers in standard laboratory strains of Drosophila that span the genome. Most of these markers are single nucleotide polymorphisms and sequences for these variants are provided in an accessible format. The average density of the new markers is one per 225 kb on the autosomes and one per megabase on the X chromosome. We include in this survey a set of P-element strains that provide additional use for high-resolution mapping. We show one application of the new markers in a simple set of crosses to map a mutation in the hedgehog gene to an interval of <1 Mb. This new map resource significantly increases the efficiency and resolution of recombination mapping and will be of immediate value to the Drosophila research community.

Single nucleotide polymorphism markers for genetic mapping in Drosophila melanogaster

For nearly a century, genetic analysis in Drosophila melanogaster has been a powerful tool for analyzing gene function, yet Drosophila lacks the molecular genetic mapping tools that recently have revolutionized human, mouse, and plant genetics. Here, we describe the systematic characterization of a dense set of molecular markers in Drosophila by using a sequence tagged site-based physical map of the genome. We identify 474 biallelic markers in standard laboratory strains of Drosophila that span the genome. Most of these markers are single nucleotide polymorphisms and sequences for these variants are provided in an accessible format. The average density of the new markers is one per 225 kb on the autosomes and one per megabase on the X chromosome. We include in this survey a set of P-element strains that provide additional use for high-resolution mapping. We show one application of the new markers in a simple set of crosses to map a mutation in the hedgehog gene to an interval of <1 Mb. This new map resource significantly increases the efficiency and resolution of recombination mapping and will be of immediate value to the Drosophila research community.

Transferrin receptor mutation analysis in hereditary hemochromatosis patients

The Cys282-->Tyr mutation in the HFE gene is carried by the majority of hereditary hemochromatosis patient chromosomes, yet some patients do not seem to harbor any mutation in this gene. This suggests a possibility that these patients may have a mutation in other genes in the same pathway as HFE. We analyzed the cDNA sequences of transferrin receptor (TFR), which was recently shown to interact with HFE, in twenty-one hereditary hemochromatosis patients including sixteen individuals who did not carry a Cys282-->Tyr mutation. A nucleotide substitution (424A-->G), which resulted in the Ser142-->Gly amino acid substitution, was the only amino acid polymorphism detected in the open reading frame of the TFR gene in these patients. This amino acid substitution was a rather common polymorphism in the general population (49%) and its frequency did not significantly differ in the hereditary hemochromatosis (HH) patients regardless of the HFE genotype. Thus, amino acid changes in the TFR gene do not appear to play a role in HH even when the patients do not have a HFE mutation. However, this study does not rule out the possibility of the involvement of mutations in non-coding regions.

The hemochromatosis founder mutation in HLA-H disrupts beta2-microglobulin interaction and cell surface expression

We recently reported the positional cloning of a candidate gene for hereditary hemochromatosis (HH), called HLA-H, which is a novel member of the major histocompatibility complex class I family. A mutation in this gene, cysteine 282 --> tyrosine (C282Y), was found to be present in 83% of HH patient DNAs, while a second variant, histidine 63 --> aspartate (H63D), was enriched in patients heterozygous for C282Y. The functional relevance of either mutation has not been described. Co-immunoprecipitation studies of cell lysates from human embryonic kidney cells transfected with wild-type or mutant HLA-H cDNA demonstrate that wild-type HLA-H binds beta2-microglobulin and that the C282Y mutation, but not the H63D mutation, completely abrogates this interaction. Immunofluorescence labeling and subcellular fractionations demonstrate that while the wild-type and H63D HLA-H proteins are expressed on the cell surface, the C282Y mutant protein is localized exclusively intracellularly. This report describes the first functional significance of the C282Y mutation by suggesting that an abnormality in protein trafficking and/or cell-surface expression of HLA-H leads to HH disease.

A 1.1-Mb transcript map of the hereditary hemochromatosis locus

In the process of positionally cloning a candidate gene responsible for hereditary hemochromatosis (HH), we constructed a 1.1-Mb transcript map of the region of human chromosome 6p that lies 4.5 Mb telomeric to HLA-A. A combination of three gene-finding techniques, direct cDNA selection, exon trapping, and sample sequencing, were used initially for a saturation screening of the 1.1-Mb region for expressed sequence fragments. As genetic analysis further narrowed the HH candidate locus, we sequenced completely 0.25 Mb of genomic DNA as a final measure to identify all genes. Besides the novel MHC class 1-like HH candidate gene HLA-H, we identified a family of five butyrophilin-related sequences, two genes with structural similarity to a type 1 sodium phosphate transporter, 12 novel histone genes, and a gene we named RoRet based on its strong similarity to the 52-kD Ro/SSA lupus and Sjogren's syndrome auto-antigen and the RET finger protein. Several members of the butyrophilin family and the RoRet gene share an exon of common evolutionary origin called B30-2. The B30-2 exon was originally isolated from the HLA class 1 region, yet has apparently "shuffled" into several genes along the chromosome telomeric to the MHC. The conservation of the B30-2 exon in several novel genes and the previously described amino acid homology of HLA-H to MHC class 1 molecules provide further support that this gene-rich region of 6p21.3 is related to the MHC. Finally, we performed an analysis of the four approaches for gene finding and conclude that direct selection provides the most effective probes for cDNA screening, and that as much as 30% of ESTs in this 1.1-Mb region may be derived from noncoding genomic DNA.

A novel MHC class I-like gene is mutated in patients with hereditary haemochromatosis

Hereditary haemochromatosis (HH), which affects some 1 in 400 and has an estimated carrier frequency of 1 in 10 individuals of Northern European descent, results in multi-organ dysfunction caused by increased iron deposition, and is treatable if detected early. Using linkage-disequilibrium and full haplotype analysis, we have identified a 250-kilobase region more than 3 megabases telomeric of the major histocompatibility complex (MHC) that is identical-by-descent in 85% of patient chromosomes. Within this region, we have identified a gene related to the MHC class I family, termed HLA-H, containing two missense alterations. One of these is predicted to inactivate this class of proteins and was found homozygous in 83% of 178 patients. A role of this gene in haemochromatosis is supported by the frequency and nature of the major mutation and prior studies implicating MHC class I-like proteins in iron metabolism.

One hundred and one new simple sequence repeat-based markers for the canine genome

One hundred and one new dinucleotide repeat polymorphisms specific for the canine genome have been identified and characterized. Screening of both primary libraries and marker-selected libraries enriched for simple sequence repeats led to the isolation of large numbers of genomic clones that contained (CA)n repeats. Over 200 of these clones were sequenced, and PCR primers that bracket the repeat were developed for those that contained ten or more continuous (CA)n units. This effort led to the production of 101 polymorphic markers, which were assigned to one of four categories depending on their degree of polymorphism. Fifty-four markers were found to be highly or very highly polymorphic as they had four or more alleles when tested on a panel of unrelated dogs. This group of markers will be useful for following inheritance of traits in crosses between dogs.