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.

Compound Activity Prediction with Dose-Dependent Transcriptomic Profiles and Deep Learning

Predicting compound activity in assays is a long-standing challenge in drug discovery. Computational models based on compound-induced gene expression signatures from a single profiling assay have shown promise toward predicting compound activity in other, seemingly unrelated, assays. Applications of such models include predicting mechanisms-of-action (MoA) for phenotypic hits, identifying off-target activities, and identifying polypharmacologies. Here, we introduce transcriptomics-to-activity transformer (TAT) models that leverage gene expression profiles observed over compound treatment at multiple concentrations to predict the compound activity in other biochemical or cellular assays. We built TAT models based on gene expression data from a RASL-seq assay to predict the activity of 2692 compounds in 262 dose-response assays. We obtained useful models for 51% of the assays, as determined through a realistic held-out set. Prospectively, we experimentally validated the activity predictions of a TAT model in a malaria inhibition assay. With a 63% hit rate, TAT successfully identified several submicromolar malaria inhibitors. Our results thus demonstrate the potential of transcriptomic responses over compound concentration and the TAT modeling framework as a cost-efficient way to identify the bioactivities of promising compounds across many assays.

Drug target prediction through deep learning functional representation of gene signatures

Many machine learning applications in bioinformatics currently rely on matching gene identities when analyzing input gene signatures and fail to take advantage of preexisting knowledge about gene functions. To further enable comparative analysis of OMICS datasets, including target deconvolution and mechanism of action studies, we develop an approach that represents gene signatures projected onto their biological functions, instead of their identities, similar to how the word2vec technique works in natural language processing. We develop the Functional Representation of Gene Signatures (FRoGS) approach by training a deep learning model and demonstrate that its application to the Broad Institute's L1000 datasets results in more effective compound-target predictions than models based on gene identities alone. By integrating additional pharmacological activity data sources, FRoGS significantly increases the number of high-quality compound-target predictions relative to existing approaches, many of which are supported by in silico and/or experimental evidence. These results underscore the general utility of FRoGS in machine learning-based bioinformatics applications. Prediction networks pre-equipped with the knowledge of gene functions may help uncover new relationships among gene signatures acquired by large-scale OMICs studies on compounds, cell types, disease models, and patient cohorts.

Integrated CRISPR screening and drug profiling identifies combination opportunities for EGFR, ALK, and BRAF/MEK inhibitors

Anti-tumor efficacy of targeted therapies is variable across patients and cancer types. Even in patients with initial deep response, tumors are typically not eradicated and eventually relapse. To address these challenges, we present a systematic screen for targets that limit the anti-tumor efficacy of EGFR and ALK inhibitors in non-small cell lung cancer and BRAF/MEK inhibitors in colorectal cancer. Our approach includes genome-wide CRISPR screens with or without drugs targeting the oncogenic driver ("anchor therapy"), and large-scale pairwise combination screens of anchor therapies with 351 other drugs. Interestingly, targeting of a small number of genes, including MCL1, BCL2L1, and YAP1, sensitizes multiple cell lines to the respective anchor therapy. Data from drug combination screens with EGF816 and ceritinib indicate that dasatinib and agents disrupting microtubules act synergistically across many cell lines. Finally, we show that a higher-order-combination screen with 26 selected drugs in two resistant EGFR-mutant lung cancer cell lines identified active triplet combinations.

Dual ALK and CDK4/6 Inhibition Demonstrates Synergy against Neuroblastoma

Purpose: Anaplastic lymphoma kinase (ALK) is the most frequently mutated oncogene in the pediatric cancer neuroblastoma. We performed an in vitro screen for synergistic drug combinations that target neuroblastomas with mutations in ALK to determine whether drug combinations could enhance antitumor efficacy.Experimental Design: We screened combinations of eight molecularly targeted agents against 17 comprehensively characterized human neuroblastoma-derived cell lines. We investigated the combination of ceritinib and ribociclib on in vitro proliferation, cell cycle, viability, caspase activation, and the cyclin D/CDK4/CDK6/RB and pALK signaling networks in cell lines with representative ALK status. We performed in vivo trials in CB17 SCID mice bearing conventional and patient-derived xenograft models comparing ceritinib alone, ribociclib alone, and the combination, with plasma pharmacokinetics to evaluate for drug-drug interactions.Results: The combination of ribociclib, a dual inhibitor of cyclin-dependent kinase (CDK) 4 and 6, and the ALK inhibitor ceritinib demonstrated higher cytotoxicity (P = 0.008) and synergy scores (P = 0.006) in cell lines with ALK mutations as compared with cell lines lacking mutations or alterations in ALK Compared with either drug alone, combination therapy enhanced growth inhibition, cell-cycle arrest, and caspase-independent cell death. Combination therapy achieved complete regressions in neuroblastoma xenografts with ALK-F1174L and F1245C de novo resistance mutations and prevented the emergence of resistance. Murine ribociclib and ceritinib plasma concentrations were unaltered by combination therapy.Conclusions: This preclinical combination drug screen with in vivo validation has provided the rationale for a first-in-children trial of combination ceritinib and ribociclib in a molecularly selected pediatric population. Clin Cancer Res; 23(11); 2856-68. ©2016 AACR.

Seeing the light: luminescent reporter gene assays

The luminescent reporter gene assay (LRGA) is arguably the most prominent type of reporter gene assay used in biomolecular and pharmaceutical development laboratories. Part of this popularity is due to the high signal associated with luciferases, the foundation of this technology. This feature makes them ideally suited for high throughput screening applications where potentially millions of chemical compounds can be analyzed in a given assay. Recent technical advancements that enhance signal stability of the luciferases along with development and commercialization of multiple forms of luciferases, their respective substrates, and improvements in expression vectors for reporter gene assay (RGA) applications have broadened their use. While the practical challenges related to the application of luminescent technology in a laboratory setting have been overcome, there remains much to do in laying a systematic approach towards the construction of RGAs, which are essential to the elucidation of the basic biology for genes of interest. This mini-review aims at giving a birds-eye view of the available luciferases, substrates and other luminescent technologies available and provides a general blueprint as well as practical considerations for constructing and interfacing RGAs with chemical biology and functional genomics for the elucidation of fundamental biological questions and for biomedical research.

Multiplexed reporter gene assays: monitoring the cell viability and the compound kinetics on luciferase activity

High-throughput screening assays with multiple readouts enable one to monitor multiple assay parameters. By capturing as much information about the underlying biology as possible, the detection of true actives can be improved. This report describes an extension to standard luciferase reporter gene assays that enables multiple parameters to be monitored from each sample. The report describes multiplexing luciferase assays with an orthogonal readout monitoring cell viability using reduction of resazurin. In addition, this technical note shows that by using the luciferin substrate in live cells, an assay time course can be recorded. This enables the identification of nonactive or unspecific compounds that act by inhibiting luciferase, as well as compounds altering gene expression or cell growth.

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.

Large-Scale Annotation of Small-Molecule Libraries Using Public Databases

While many large publicly accessible databases provide excellent annotation for biological macromolecules, the same is not true for small chemical compounds. Commercial data sources also fail to encompass an annotation interface for large numbers of compounds and tend to be cost prohibitive to be widely available to biomedical researchers. Therefore, using annotation information for the selection of lead compounds from a modern day high-throughput screening (HTS) campaign presently occurs only under a very limited scale. The recent rapid expansion of the NIH PubChem database provides an opportunity to link existing biological databases with compound catalogs and provides relevant information that potentially could improve the information garnered from large-scale screening efforts. Using the 2.5 million compound collection at the Genomics Institute of the Novartis Research Foundation (GNF) as a model, we determined that approximately 4% of the library contained compounds with potential annotation in such databases as PubChem and the World Drug Index (WDI) as well as related databases such as the Kyoto Encyclopedia of Genes and Genomes (KEGG) and ChemIDplus. Furthermore, the exact structure match analysis showed 32% of GNF compounds can be linked to third party databases via PubChem. We also showed annotations such as MeSH (medical subject headings) terms can be applied to in-house HTS databases in identifying signature biological inhibition profiles of interest as well as expediting the assay validation process. The automated annotation of thousands of screening hits in batch is becoming feasible and has the potential to play an essential role in the hit-to-lead decision making process.

Learning from the data: mining of large high-throughput screening databases

High-throughput screening (HTS) campaigns in pharmaceutical companies have accumulated a large amount of data for several million compounds over a couple of hundred assays. Despite the general awareness that rich information is hidden inside the vast amount of data, little has been reported for a systematic data mining method that can reliably extract relevant knowledge of interest for chemists and biologists. We developed a data mining approach based on an algorithm called ontology-based pattern identification (OPI) and applied it to our in-house HTS database. We identified nearly 1500 scaffold families with statistically significant structure-HTS activity profile relationships. Among them, dozens of scaffolds were characterized as leading to artifactual results stemming from the screening technology employed, such as assay format and/or readout. Four types of compound scaffolds can be characterized based on this data mining effort: tumor cytotoxic, general toxic, potential reporter gene assay artifact, and target family specific. The OPI-based data mining approach can reliably identify compounds that are not only structurally similar but also share statistically significant biological activity profiles. Statistical tests such as Kruskal-Wallis test and analysis of variance (ANOVA) can then be applied to the discovered scaffolds for effective assignment of relevant biological information. The scaffolds identified by our HTS data mining efforts are an invaluable resource for designing SAR-robust diversity libraries, generating in silico biological annotations of compounds on a scaffold basis, and providing novel target family specific scaffolds for focused compound library design.

An efficient rapid system for profiling the cellular activities of molecular libraries

Rapid quantitative methods for characterizing small molecules, peptides, proteins, or RNAs in a broad array of cellular assays would allow one to discover new biological activities associated with these molecules and also provide a more comprehensive profile of drug candidates early in the drug development process. Here we describe a robotic system, termed the automated compound profiler, capable of both propagating a large number of cell lines in parallel and assaying large collections of molecules simultaneously against a matrix of cellular assays in a highly reproducible manner. To illustrate its utility, we have characterized a set of 1,400 kinase inhibitors in a panel of 35 activated tyrosine-kinase-dependent cellular assays in dose-response format in a single experiment. Analysis of the resulting multidimensional dataset revealed subclusters of both inhibitors and kinases with closely correlated activities. The approach also identified activities for the p38 inhibitor BIRB796 and the dual src/abl inhibitor BMS-354825 and exposed the expected side activities for Glivec/STI571, including cellular inhibition of c-kit and platelet-derived growth factor receptor. This methodology provides a powerful tool for unraveling the cellular biology and molecular pharmacology of both naturally occurring and synthetic chemical diversity.

Genome-scale functional profiling of the mammalian AP-1 signaling pathway

Large-scale functional genomics approaches are fundamental to the characterization of mammalian transcriptomes annotated by genome sequencing projects. Although current high-throughput strategies systematically survey either transcriptional or biochemical networks, analogous genome-scale investigations that analyze gene function in mammalian cells have yet to be fully realized. Through transient overexpression analysis, we describe the parallel interrogation of approximately 20,000 sequence annotated genes in cancer-related signaling pathways. For experimental validation of these genome data, we apply an integrative strategy to characterize previously unreported effectors of activator protein-1 (AP-1) mediated growth and mitogenic response pathways. These studies identify the ADP-ribosylation factor GTPase-activating protein Centaurin alpha1 and a Tudor domain-containing hypothetical protein as putative AP-1 regulatory oncogenes. These results provide insight into the composition of the AP-1 signaling machinery and validate this approach as a tractable platform for genome-wide functional analysis.

Yb modulates the divisions of both germline and somatic stem cells through piwi- and hh-mediated mechanisms in the Drosophila ovary

The coordinated division of distinctive types of stem cells within an organ is crucial for organogenesis and homeostasis. Here we show genetic interactions among fs(1)Yb (Yb), piwi, and hedgehog (hh) that regulate the division of both germline stem cells (GSCs) and somatic stem cells (SSCs), the two constituent stem cell populations of the Drosophila ovary. Yb is required for both GSC and SSC divisions; loss of Yb function eliminates GSCs and reduces SSC division, while Yb overexpression increases GSC number and causes SSC overproliferation. We also show that Yb acts via the piwi- and hh-mediated signaling pathways that emanate from the same signaling cells to control GSC and SSC division, respectively. hh signaling also has a minor effect in GSC division.

Somatic signaling mediated by fs(1)Yb is essential for germline stem cell maintenance during Drosophila oogenesis

Drosophila oogenesis starts when a germline stem cell divides asymmetrically to generate a daughter germline stem cell and a cystoblast that will develop into a mature egg. We show that the fs(1)Yb gene is essential for the maintenance of germline stem cells during oogenesis. We delineate fs(1)Yb within a 6.4 kb genomic region by transgenic rescue experiments. fs(1)Yb encodes a 4.1 kb RNA that is present in the third instar larval, pupal and adult stages, consistent with its role in regulating germline stem cells during oogenesis. Germline clonal analysis shows that all fs(1)Yb mutations are soma-dependent. In the adult ovary, fs(1)Yb is specifically expressed in the terminal filament cells, suggesting that fs(1)Yb acts in these signaling cells to maintain germline stem cells. fs(1)Yb encodes a novel hydrophilic protein with no potential signal peptide or transmembrane domains, suggesting that this protein is not itself a signal but a key component of the signaling machinery for germline stem cell maintenance.

DEF-1, a novel Src SH3 binding protein that promotes adipogenesis in fibroblastic cell lines

The Src homology 3 (SH3) motif is found in numerous signal transduction proteins involved in cellular growth and differentiation. We have purified and cloned a novel protein, DEF-1 (differentiation-enhancing factor), from bovine brain by using a Src SH3 affinity column. Ectopic expression of DEF-1 in fibroblasts resulted in the differentiation of a significant fraction of the culture into adipocytes. This phenotype appears to be related to the induction of the transcription factor peroxisome proliferator-activated receptor gamma (PPARgamma), since DEF-1 NIH 3T3 cells demonstrated augmented levels of PPARgamma mRNA and, when treated with activating PPARgamma ligands, efficient induction of differentiation. Further evidence for a role for DEF-1 in adipogenesis was provided by heightened expression of DEF-1 mRNA in adipose tissue isolated from obese and diabetes mice compared to that in tissue isolated from wild-type mice. However, DEF-1 mRNA was detected in multiple tissues, suggesting that the signal transduction pathway(s) in which DEF-1 is involved is not limited to adipogenesis. These results suggest that DEF-1 is an important component of a signal transduction process that is involved in the differentiation of fibroblasts and possibly of other types of cells.

Differentiation and reversal of malignant changes in colon cancer through PPARgamma

PPARgamma is a nuclear receptor that has a dominant regulatory role in differentiation of cells of the adipose lineage, and has recently been shown to be expressed in the colon. We show here that PPARgamma is expressed at high levels in both well- and poorly-differentiated adenocarcinomas, in normal colonic mucosa and in human colon cancer cell lines. Ligand activation of this receptor in colon cancer cells causes a considerable reduction in linear and clonogenic growth, increased expression of carcinoembryonic antigen and the reversal of many gene expression events specifically associated with colon cancer. Transplantable tumors derived from human colon cancer cells show a significant reduction of growth when mice are treated with troglitazone, a PPARgamma ligand. These results indicate that the growth and differentiation of colon cancer cells can be modulated through PPARgamma.

Increased bone mineral content of preterm infants fed with a nutrient enriched formula after discharge from hospital

Bone disease with persistent reduced bone mineralisation is common in premature infants. To test the hypothesis that enhancement of nutritional intake after discharge from hospital improves bone mineralisation, 31 formula fed preterm infants were randomly assigned to receive standard or multinutrient enriched milk from the time of discharge. The calcium and phosphorus contents of the enriched milk were 70 and 35 mg/100 ml v 35 and 29 mg/100 ml for the standard formula. Bone mineral content was measured before discharge from hospital in 21 of the infants; there was no difference in the bone mineral content between the groups at that time (35 mg/cm for the two groups). There was a significant increase in bone mineral content for those infants receiving the enriched v standard formula at 3 and 9 months corrected postnatal age: at 3 months the bone mineral content was 83 v 63 mg/cm and at 9 months 115 v 95 mg/cm. The difference between the groups was thus maintained although not increased at a corrected age of 9 months, when the bone mineral content of infants fed the enriched but not the standard formula was no longer significantly different from that of normal infants after adjusting for body size. The difference was not explained by the larger body size in infants fed the enriched formula. The results suggest that the use of a special nutrient enriched postdischarge formula has a significant positive effect on bone growth and mineralisation during a period of rapid skeletal development.

Randomised trial of nutrition for preterm infants after discharge

In a randomised double blind trial, the effect on growth and clinical status of a nutrient enriched 'post-discharge' milk formula versus a standard term formula, was compared in 32 exclusively bottle fed preterm infants. The formulas were used as the sole milk intake up to a postnatal age of 9 months. Significant increases in linear growth and weight gain were observed in the infants who received the enriched diet. There were no differences in vomiting, posseting, or bowel habit between the groups. Formula volumes ingested were similar between diet groups, indicating that the difference in formula composition did not affect the infants' regulation of intake. These preliminary data suggest that there is a role for specially designed formulas for preterm infants after discharge from hospital.

A colony-stimulating factor 1 (CSF-1) receptor/platelet-derived growth factor-beta receptor gene fusion confers CSF-1 independence and tumorigenicity on a c-myc-immortalized monocyte cell line

Monocytes and macrophages express the receptor for the hematopoietic growth factor colony-stimulating factor 1 (CSF-1) and require this factor for growth in culture. A murine monocyte tumor cell line that lacks the usual requirement for CSF-1 was isolated. On the basis of the similarity of the structures of the CSF-1 and platelet-derived growth factor (PDGF) receptors and because monocytes normally secrete PDGF, we analyzed the tumor cell line for anomalous expression of the PDGF-R beta gene. Two different cDNAs that each contain sequences corresponding to the complete coding sequence of PDGF-R beta fused (in frame) to the amino-terminal half of the CSF-1 receptor were isolated. Introduction of these PDGF-R beta-related cDNAs into two partially transformed, CSF-1-dependent monocyte cell lines resulted in autonomous growth and cell transformation. These monocyte cell lines exhibit a novel form of growth factor receptor activation that can lead to oncogenic growth in collaboration with the c-myc oncogene.

Linear growth in the early neonatal period

To test the hypothesis that early linear growth is independent of changes in weight we undertook took simple anthropometry in 45 term infants daily to day 7 after birth. Linear growth proceeded rapidly and independently of changes in weight variations from the first day after birth; we suggest that this implies 'programmed' continuity of skeletal growth, possibly fuelled at the expense of other body tissues.

Paradoxical bone mineralisation in the twin to twin transfusion syndrome

We report twin preterm infants with the twin to twin transfusion syndrome, exhibiting grossly different bone densities on chest radiographs. Photonabsorptiometry showed the polycythaemic twin was osteopenic and the anaemic twin osteosclerotic; bone mineral contents were 0.028 g/cm and 0.074 g/cm respectively (normal mean (SD) 0.041 (0.006) g/cm. We speculate that alterations in macrophage derived osteoclastic activity contribute to these previously unreported findings.

Molecular cloning and sequencing of the murine c-fgr gene

The c-fgr gene is a member of the tyrosine kinase family of proto-oncogenes and is expressed exclusively in hemopoietic cells. We found that c-fgr was expressed at high levels in a limited subset of murine monocyte/macrophage tumors that were induced by the c-myc oncogene, in cells representing late stages of monocyte differentiation. A cDNA clone was isolated from a library made from a monocyte tumor cell lines using a human c-fgr and v-fgr probe. The composite nucleotide and predicted amino acid sequence of the cDNA indicates that it represents the complete coding sequence for the murine c-fgr gene. Comparison of the N-terminal human and mouse c-fgr amino acid sequences indicates regions of identity that are not homologous to other tyrosine kinases. Interestingly, these regions share a limited but significant homology to two viral proteins, adenovirus E1b and HIV nef. In addition, there are other regions of homology that are shared by several tyrosine kinases and other non-kinase proteins which may be important for subcellular localization.

Detection of Parasites in Fish Muscle by Candling Technique

This study determined optimum conditions for detecting codworms in skinned cod fillets by using candling tables under commercial conditions. The best balance of factors was sought for obtaining maximum lighting conditions, reducing operator fatigue, retaining natural fillet color, and having a high contrast between parasites and fish flesh. Based on the results obtained, a recommended procedure has been adopted by AOAC.

Procedure for Cooking Seafood Products

As a result of experience gained using procedure 18.003, a revision is proposed. The revised procedure has been adopted.

Collaborative Study of Determination of Amount of Seafood in Seafood Cocktails

Method 18.B02–18.B03 for shrimp in shrimp cocktail was used to determine the amount of seafood ingredient in 6 seafood cocktails: clam, crab body meat, crab leg meat, fish, oyster, and shrimp. Duplicate samples of each kind of seafood cocktail were studied by 10 collaborators. Their results indicate that the method is suitable for a variety of seafood cocktails if 2 revisions are made. A revised version of 18.B02— 18.B03 was adopted as a procedure for seafood in seafood cocktails other than shrimp.

ACCEPTABILITY OF MAIN DISHES (ENTREES) BASED ON MIXTURES OF GROUND BEEF WITH GROUND FISH OBTAINED FROM UNDER-USED SOURCES

Technological, economic, and nutritional considerations suggest that ground fish can be better utilized if a greater variety of acceptable food items are developed. Using well-known main dish recipes, the feasibility of mixing ground beef and ground fish was tested. Several types of main dishes received favorable acceptance ratings. Some of these evaluations were based on ground fish that had been stored in a freezer for up to a year.

Significance of Dimethyl Sulfide to the Odor of Soft-Shell Clams

In 12 samples of fresh soft-shell clam meats which were collected over their spawning period, the average concentration of dimethyl sulfide (DMS) was 3 ppm. Odor evaluations were made to determine the range of concentration of pure DMS that a panel recognized as a clamlike odor. Since the concentration of DMS in our clam samples was within this range (0.08–40 ppm), it is suggested that DMS dominates the odor of fresh soft-shell clams.