Data mining reveals a network of early-response genes as a consensus signature of drug-induced in vitro and in vivo toxicity

Assessment of Drug-Induced Liver Injury through Cell Morphology and Gene Expression Analysis

Drug-induced liver injury (DILI) is a significant concern in drug development, often leading to drug withdrawal. Although many studies aim to identify biomarkers and gene/pathway signatures related to liver toxicity and aim to predict DILI compounds, this remains a challenge in drug discovery. With a strong development of high-content screening/imaging (HCS/HCI) for phenotypic screening, we explored the morphological cell perturbations induced by DILI compounds. In the first step, cell morphological signatures were associated with two datasets of DILI chemicals (DILIRank and eTox). The mechanisms of action were then analyzed for chemicals having transcriptomics data and sharing similar morphological perturbations. Signaling pathways associated with liver toxicity (cell cycle, cell growth, apoptosis, ...) were then captured, and a hypothetical relation between cell morphological perturbations and gene deregulation was illustrated within our analysis. Finally, using the cell morphological signatures, machine learning approaches were developed to predict chemicals with a potential risk of DILI. Some models showed relevant performance with validation set balanced accuracies between 0.645 and 0.739. Overall, our findings demonstrate the utility of combining HCI with transcriptomics data to identify the morphological and gene expression signatures related to DILI chemicals. Moreover, our protocol could be extended to other toxicity end points, offering a promising avenue for comprehensive toxicity assessment in drug discovery.

Antiviral activity of singlet oxygen-photogenerating perylene compounds against SARS-CoV-2: Interaction with the viral envelope and photodynamic virion inactivation

The coronavirus disease 2019 (COVID-19) pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has prompted great interest in novel broad-spectrum antivirals, including perylene-related compounds. In the present study, we performed a structure-activity relationship analysis of a series of perylene derivatives, which comprised a large planar perylene residue, and structurally divergent polar groups connected to the perylene core by a rigid ethynyl or thiophene linker. Most of the tested compounds did not exhibit significant cytotoxicity towards multiple cell types susceptible to SARS-CoV-2 infection, and did not change the expressions of cellular stress-related genes under normal light conditions. These compounds showed nanomolar or sub-micromolar dose-dependent anti-SARS-CoV-2 activity, and also suppressed the in vitro replication of feline coronavirus (FCoV), also termed feline infectious peritonitis virus (FIPV). Perylene compounds exhibited high affinity for liposomal and cellular membranes, and efficiently intercalated into the envelopes of SARS-CoV-2 virions, thereby blocking the viral-cell fusion machinery. Furthermore, the studied compounds were demonstrated to be potent photosensitizers, generating reactive oxygen species (ROS), and their anti-SARS-CoV-2 activities were considerably enhanced after irradiation with blue light. Our results indicated that photosensitization is the major mechanism underlying the anti-SARS-CoV-2 activity of perylene derivatives, with these compounds completely losing their antiviral potency under red light. Overall, perylene-based compounds are broad-spectrum antivirals against multiple enveloped viruses, with antiviral action based on light-induced photochemical damage (ROS-mediated, likely singlet oxygen-mediated), causing impairment of viral membrane rheology.

Harnessing male germline epigenomics for the genetic improvement in cattle

Sperm is essential for successful artificial insemination in dairy cattle, and its quality can be influenced by both epigenetic modification and epigenetic inheritance. The bovine germline differentiation is characterized by epigenetic reprogramming, while intergenerational and transgenerational epigenetic inheritance can influence the offspring's development through the transmission of epigenetic features to the offspring via the germline. Therefore, the selection of bulls with superior sperm quality for the production and fertility traits requires a better understanding of the epigenetic mechanism and more accurate identifications of epigenetic biomarkers. We have comprehensively reviewed the current progress in the studies of bovine sperm epigenome in terms of both resources and biological discovery in order to provide perspectives on how to harness this valuable information for genetic improvement in the cattle breeding industry.

Structure-activity relationship read-across and transcriptomics for branched carboxylic acids

The purpose of this study was to use chemical similarity evaluations, transcriptional profiling, in vitro toxicokinetic data, and physiologically based pharmacokinetic (PBPK) models to support read-across for a series of branched carboxylic acids using valproic acid (VPA), a known developmental toxicant, as a comparator. The chemicals included 2-propylpentanoic acid (VPA), 2-ethylbutanoic acid, 2-ethylhexanoic acid (EHA), 2-methylnonanoic acid, 2-hexyldecanoic acid, 2-propylnonanoic acid (PNA), dipentyl acetic acid or 2-pentylheptanoic acid, octanoic acid (a straight chain alkyl acid), and 2-ethylhexanol. Transcriptomics was evaluated in 4 cell types (A549, HepG2, MCF7, and iCell cardiomyocytes) 6 h after exposure to 3 concentrations of the compounds, using the L1000 platform. The transcriptional profiling data indicate that 2- or 3-carbon alkyl substituents at the alpha position of the carboxylic acid (EHA and PNA) elicit a transcriptional profile similar to the one elicited by VPA. The transcriptional profile is different for the other chemicals tested, which provides support for limiting read-across from VPA to much shorter and longer acids. Molecular docking models for histone deacetylases, the putative target of VPA, provide a possible mechanistic explanation for the activity cliff elucidated by transcriptomics. In vitro toxicokinetic data were utilized in a PBPK model to estimate internal dosimetry. The PBPK modeling data show that as the branched chain increases, predicted plasma Cmax decreases. This work demonstrates how transcriptomics and other mode of action-based methods can improve read-across.

Article title: Transcriptional profiling efficacy to define biological activity similarity for cosmetic ingredients' safety assessment based on next-generation read-across

The objective of this work was to use transcriptional profiling to assess the biological activity of structurally related chemicals to define their biological similarity and with that, substantiate the validity of a read-across approach usable in risk assessment. Two case studies are presented, one with 4 short alkyl chain parabens: methyl (MP), ethyl (EP), butyl (BP), and propylparaben (PP), as well as their main metabolite, p-hydroxybenzoic acid (pHBA) with the assumption that propylparaben was the target chemical; and a second one with caffeine and its main metabolites theophylline, theobromine and paraxanthine where CA was the target chemical. The comprehensive transcriptional response of MCF7, HepG2, A549 and ICell cardiomyocytes was evaluated (TempO-Seq) after exposure to vehicle-control, each paraben or pHBA, CA or its metabolites, at 3 non-cytotoxic concentrations, for 6 h. Differentially expressed genes (FDR ≥0.05, and fold change ±1.2≥) were identified for each chemical, at each concentration, and used to determine similarities. Each of the chemicals is able to elicit changes in the expression of a number of genes, as compared to controls. Importantly, the transcriptional profile elicited by each of the parabens shares a high degree of similarity across the group. The highest number of genes commonly affected was between butylparaben and PP. The transcriptional profile of the parabens is similar to the one elicited by estrogen receptor agonists, with BP being the closest structural and biological analogue for PP. In the CA case, the transcriptional profile elicited of all four methylxanthines had a high degree of similarity across the cell types, with CA and theophylline being the most active. The most robust response was obtained in the cardiomyocytes with the highest transcriptional profile similarity between CA and TP. The transcriptional profile of the methylxanthines is similar to the one elicited by inhibitors of phosphatidylinositol 3-kinase as well as other kinase inhibitors. Overall, our results support the approach of incorporating transcriptional profiling in well-designed in vitro tests as one robust stream of data to support biological similarity driven read-across procedures and strengthening the traditional structure-based approaches useful in risk assessment.

SOPHIE: Generative Neural Networks Separate Common and Specific Transcriptional Responses

Genome-wide transcriptome profiling identifies genes that are prone to differential expression (DE) across contexts, as well as genes with changes specific to the experimental manipulation. Distinguishing genes that are specifically changed in a context of interest from common differentially expressed genes (DEGs) allows more efficient prediction of which genes are specific to a given biological process under scrutiny. Currently, common DEGs or pathways can only be identified through the laborious manual curation of experiments, an inordinately time-consuming endeavor. Here we pioneer an approach, Specific cOntext Pattern Highlighting In Expression data (SOPHIE), for distinguishing between common and specific transcriptional patterns using a generative neural network to create a background set of experiments from which a null distribution of gene and pathway changes can be generated. We apply SOPHIE to diverse datasets including those from human, human cancer, and bacterial pathogen Pseudomonas aeruginosa. SOPHIE identifies common DEGs in concordance with previously described, manually and systematically determined common DEGs. Further molecular validation indicates that SOPHIE detects highly specific but low-magnitude biologically relevant transcriptional changes. SOPHIE's measure of specificity can complement log2 fold change values generated from traditional DE analyses. For example, by filtering the set of DEGs, one can identify genes that are specifically relevant to the experimental condition of interest. Consequently, these results can inform future research directions. All scripts used in these analyses are available at https://github.com/greenelab/generic-expression-patterns. Users can access https://github.com/greenelab/sophie to run SOPHIE on their own data.

Deriving time-concordant event cascades from gene expression data: A case study for Drug-Induced Liver Injury (DILI)

Adverse event pathogenesis is often a complex process which compromises multiple events ranging from the molecular to the phenotypic level. In toxicology, Adverse Outcome Pathways (AOPs) aim to formalize this as temporal sequences of events, in which event relationships should be supported by causal evidence according to the tailored Bradford-Hill criteria. One of the criteria is whether events are consistently observed in a certain temporal order and, in this work, we study this time concordance using the concept of “first activation” as data-driven means to generate hypotheses on potentially causal mechanisms. As a case study, we analysed liver data from repeat-dose studies in rats from the TG-GATEs database which comprises measurements across eight timepoints, ranging from 3 hours to 4 weeks post-treatment. We identified time-concordant gene expression-derived events preceding adverse histopathology, which serves as surrogate readout for Drug-Induced Liver Injury (DILI). We find known mechanisms in DILI to be time-concordant, and show further that significance, frequency and log fold change (logFC) of differential expression are metrics which can additionally prioritize events although not necessary to be mechanistically relevant. Moreover, we used the temporal order of transcription factor (TF) expression and regulon activity to identify transcriptionally regulated TFs and subsequently combined this with prior knowledge on functional interactions to derive detailed gene-regulatory mechanisms, such as reduced Hnf4a activity leading to decreased expression and activity of Cebpa. At the same time, also potentially novel events are identified such as Sox13 which is highly significantly time-concordant and shows sustained activation over time. Overall, we demonstrate how time-resolved transcriptomics can derive and support mechanistic hypotheses by quantifying time concordance and how this can be combined with prior causal knowledge, with the aim of both understanding mechanisms of toxicity, as well as potential applications to the AOP framework. We make our results available in the form of a Shiny app (https://anikaliu.shinyapps.io/dili_cascades), which allows users to query events of interest in more detail.

Understanding mechanisms from systems-scale biological data is of great relevance in toxicology as well as drug discovery; however how to generate causal hypotheses instead of correlations is by no means clear. In this work, we study the conserved temporal order of events and present an automatable framework to quantify and characterize time concordance across a large set of time-series. We apply this concept to events derived from time-resolved gene expression and histopathology from the TG-GATEs in vivo liver data as a case study. We were able to recover known events involved in the pathogenesis of Drug-Induced Liver Injury (DILI), and identify potentially novel pathway and transcription factors (TFs) which precede adverse histopathology. As complementary sources of evidence for causality, we additionally show how time concordance and prior knowledge on plausible interactions between TFs can be combined to derive causal hypotheses on the TFs’ mode of regulation and interaction partners. Overall, the results derived in our case study can serve as valuable hypothesis-free starting points for the development of Adverse Outcome Pathways for DILI, and demonstrate that our approach provides a novel angle to prioritize mechanistically relevant events.

In vitro profiling of toxic effects of environmental polycyclic aromatic hydrocarbons on nuclear receptor signaling, disruption of endogenous metabolism and induction of cellular stress

Polycyclic aromatic hydrocarbons (PAHs) may interact with multiple intracellular receptors and related signaling pathways. We comprehensively evaluated the toxicity profiles of six environmentally relevant PAHs differing in structure, genotoxicity and their ability to activate the aryl hydrocarbon receptor (AhR). We focused particularly on their impact on intracellular hormone-, xenobiotic- and lipid-sensing receptors, as well as on cellular stress markers, combining a battery of human reporter gene assays and qRT-PCR evaluation of endogenous gene expression in human hepatocyte-like HepaRG cells, with LC/MS-MS analysis of cellular sphingolipids. The effects of PAHs included: activation of estrogen receptor α (in case of fluoranthene (Fla), pyrene (Pyr), benz[a]anthracene (BaA), benzo[a]pyrene (BaP)), suppression of androgen receptor activity (Fla, BaA, BaP and benzo[k]fluoranthene (BkF)), enhancement of dexamethasone-induced glucocorticoid receptor activity (chrysene (Chry), BaA, and BaP), and potentiation of triiodothyronine-induced thyroid receptor α activity (all tested PAHs). PAHs also induced transcription of endogenous gene targets of constitutive androstane receptor (Fla, Pyr), or repression of target genes of pregnane X receptor and peroxisome proliferator-activated receptor α (in case of the AhR-activating PAHs - Chry, BaA, BaP, and BkF) in HepaRG cells. In the same cell model, the AhR agonists reduced the expression of glucose metabolism genes (PCK1, G6PC and PDK4), and they up-regulated levels of glucosylceramides, together with a concomitant induction of expression of UGCG, glucosylceramide synthesis enzyme. Finally, both BaP and BkF were found to induce expression of early stress and genotoxicity markers: ATF3, EGR1, GDF15, CDKN1A/p21, and GADD45A mRNAs, while BaP alone increased levels of IL-6 mRNA. Overall, whereas low-molecular-weight PAHs exerted significant effects on nuclear receptors (with CYP2B6 induction observed already at nanomolar concentrations), the AhR activation by 4-ring and 5-ring PAHs appeared to be a key mechanism underlying their impact on nuclear receptor signaling, endogenous metabolism and induction of early stress and genotoxicity markers.

Tissue heterogeneity is prevalent in gene expression studies

Lack of reproducibility in gene expression studies is a serious issue being actively addressed by the biomedical research community. Besides established factors such as batch effects and incorrect sample annotations, we recently reported tissue heterogeneity, a consequence of unintended profiling of cells of other origins than the tissue of interest, as a source of variance. Although tissue heterogeneity exacerbates irreproducibility, its prevalence in gene expression data remains unknown. Here, we systematically analyse 2 667 publicly available gene expression datasets covering 76 576 samples. Using two independent data compendia and a reproducible, open-source software pipeline, we find a prevalence of tissue heterogeneity in gene expression data that affects between 1 and 40% of the samples, depending on the tissue type. We discover both cases of severe heterogeneity, which may be caused by mistakes in annotation or sample handling, and cases of moderate heterogeneity, which are likely caused by tissue infiltration or sample contamination. Our analysis establishes tissue heterogeneity as a widespread phenomenon in publicly available gene expression datasets, which constitutes an important source of variance that should not be ignored. Consequently, we advocate the application of quality-control methods such as BioQC to detect tissue heterogeneity prior to mining or analysing gene expression data.

Contextualization of causal regulatory networks from toxicogenomics data applied to drug-induced liver injury

In recent years, network-based methods have become an attractive analytical approach for toxicogenomics studies. They can capture not only the global changes of regulatory gene networks but also the relationships between their components. Among them, a causal reasoning approach depicts the mechanisms of regulation that connect upstream regulators in signaling networks to their downstream gene targets. In this work, we applied CARNIVAL, a causal network contextualisation tool, to infer upstream signaling networks deregulated in drug-induced liver injury (DILI) from gene expression microarray data from the TG-GATEs database. We focussed on six compounds that induce observable histopathologies linked to DILI from repeated dosing experiments in rats. We compared responses in vitro and in vivo to identify potential cross-platform concordances in rats as well as network preservations between rat and human. Our results showed similarities of enriched pathways and network motifs between compounds. These pathways and motifs induced the same pathology in rats but not in humans. In particular, the causal interactions "LCK activates SOCS3, which in turn inhibits TFDP1" was commonly identified as a regulatory path among the fibrosis-inducing compounds. This potential pathology-inducing regulation illustrates the value of our approach to generate hypotheses that can be further validated experimentally.

Incorporation of in vitro techniques for botanicals dietary supplement safety assessment - Towards evaluation of developmental and reproductive toxicity (DART)

As complex mixtures, botanicals present unique challenges when assessing safe use, particularly when endpoint gaps exist that cannot be fully resolved by existing toxicological literature. Here we explore in vitro gene expression as well receptor binding and enzyme activity as alternative assays to inform on developmental and reproductive toxicity (DART) relevant modes of action, since DART data gaps are common for botanicals. Specifically, botanicals suspected to have DART effects, in addition to those with a significant history of use, were tested in these assays. Gene expression changes in a number of different cell types were analysed using the connectivity mapping approach (CMap) to identify modes of action through a functional read across approach. Taken together with ligand affinity data obtained using a set of molecular targets customised towards known DART relevant modes of action, it was possible to inform DART risk using functional analogues, potency comparisons and a margin of internal exposure approach.

An Introduction to Machine Learning

In the last few years, machine learning (ML) and artificial intelligence have seen a new wave of publicity fueled by the huge and ever-increasing amount of data and computational power as well as the discovery of improved learning algorithms. However, the idea of a computer learning some abstract concept from data and applying them to yet unseen situations is not new and has been around at least since the 1950s. Many of these basic principles are very familiar to the pharmacometrics and clinical pharmacology community. In this paper, we want to introduce the foundational ideas of ML to this community such that readers obtain the essential tools they need to understand publications on the topic. Although we will not go into the very details and theoretical background, we aim to point readers to relevant literature and put applications of ML in molecular biology as well as the fields of pharmacometrics and clinical pharmacology into perspective.

Gadolinium labelled nanoliposomes as the platform for MRI theranostics: in vitro safety study in liver cells and macrophages

Gadolinium (Gd)-based contrast agents are extensively used for magnetic resonance imaging (MRI). Liposomes are potential nanocarrier-based biocompatible platforms for development of new generations of MRI diagnostics. Liposomes with Gd-complexes (Gd-lip) co-encapsulated with thrombolytic agents can serve both for imaging and treatment of various pathological states including stroke. In this study, we evaluated nanosafety of Gd-lip containing PE-DTPA chelating Gd+3 prepared by lipid film hydration method. We detected no cytotoxicity of Gd-lip in human liver cells including cancer HepG2, progenitor (non-differentiated) HepaRG, and differentiated HepaRG cells. Furthermore, no potential side effects of Gd-lip were found using a complex system including general biomarkers of toxicity, such as induction of early response genes, oxidative, heat shock and endoplasmic reticulum stress, DNA damage responses, induction of xenobiotic metabolizing enzymes, and changes in sphingolipid metabolism in differentiated HepaRG. Moreover, Gd-lip did not show pro-inflammatory effects, as assessed in an assay based on activation of inflammasome NLRP3 in a model of human macrophages, and release of eicosanoids from HepaRG cells. In conclusion, this in vitro study indicates potential in vivo safety of Gd-lip with respect to hepatotoxicity and immunopathology caused by inflammation.

Today's Challenges to De-Risk and Predict Drug Safety in Human "Mind-the-Gap"

Current gaps in drug safety sciences can result from the inability (1) to identify hazard across multiple target organs, (2) to predict and risk assess with certainty against drug safety liabilities for the major target organs, (3) to optimally manage and mitigate against drug safety liabilities, and (4) to apply principles of governance on the generation, integration, and use of experimental data. Translational safety assessment to evaluate several target-organ drug toxicities can only be partially achieved by use of current in vitro and in vivo test systems. What remains to be tackled necessitates the deployment of in vitro-human-relevant test systems to address human specific or selective forms of toxicities. Nevertheless, such models may only address in part some of the requirements in today's armament of biomedical tools essential for improving the discovery of drug candidates. Refinement of in silico tools, Target Safety Assessment and a greater understanding of mechanistic insights of toxicities might provide future opportunities to better identify drug safety liabilities. The increasing diversity of drug modalities present further challenges for nonclinical and clinical development requiring further research to develop suitable test systems and technologies. Our ability to optimally manage and mitigate safety risk will come from the greater refinement of safety margin estimates, provision and use of human-relevant safety biomarkers, and understanding of the translation from in silico, in vitro, and in vivo studies to human. An improvement of governance frameworks and standards at all levels within organizations, national, and international, can only help facilitate drug discovery and development programs.

Immune modulation by chronic exposure to waterpipe smoke and immediate-early gene regulation in murine lungs

OBJECTIVE

We investigated the effects of chronic waterpipe (WP) smoke on pulmonary function and immune response in a murine model using a research-grade WP and the effects of acute exposure on the regulation of immediate-early genes (IEGs).

METHODS

WP smoke was generated using three WP smoke puffing regimens based on the Beirut regimen. WP smoke samples generated under these puffing regimens were quantified for nicotine concentration. Mice were chronically exposed for 6 months followed by assessment of pulmonary function and airway inflammation. Transcriptomic analysis using RNAseq was conducted after acute exposure to characterise the IEG response. These biomarkers were then compared with those generated after exposure to dry smoke (without water added to the WP bowl).

RESULTS

We determined that nicotine composition in WP smoke ranged from 0.4 to 2.5 mg per puffing session. The lung immune response was sensitive to the incremental severity of chronic exposure, with modest decreases in airway inflammatory cells and chemokine levels compared with air-exposed controls. Pulmonary function was unmodified by chronic WP exposure. Acute WP exposure was found to activate the immune response and identified known and novel IEG as potential biomarkers of WP exposure.

CONCLUSION

Chronic exposure to WP smoke leads to immune suppression without significant changes to pulmonary function. Transcriptomic analysis of the lung after acute exposure to WP smoke showed activation of the immune response and revealed IEGs that are common to WP and dry smoke, as well as pools of IEGs unique to each exposure, identifying potential biomarkers specific to WP exposure.

Screening of Cellular Stress Responses Induced by Ambient Aerosol Ultrafine Particle Fraction PM0.5 in A549 Cells

Effects of airborne particles on the expression status of markers of cellular toxic stress and on the release of eicosanoids, linked with inflammation and oxidative damage, remain poorly characterized. Therefore, we proposed a set of various methodological approaches in order to address complexity of PM0.5-induced toxicity. For this purpose, we used a well-characterized model of A549 pulmonary epithelial cells exposed to a non-cytotoxic concentration of ambient aerosol particle fraction PM0.5 for 24 h. Electron microscopy confirmed accumulation of PM0.5 within A549 cells, yet, autophagy was not induced. Expression profiles of various cellular stress response genes that have been previously shown to be involved in early stress responses, namely unfolded protein response, DNA damage response, and in aryl hydrocarbon receptor (AhR) and p53 signaling, were analyzed. This analysis revealed induction of GREM1, EGR1, CYP1A1, CDK1A, PUMA, NOXA and GDF15 and suppression of SOX9 in response to PM0.5 exposure. Analysis of eicosanoids showed no oxidative damage and only a weak anti-inflammatory response. In conclusion, this study helps to identify novel gene markers, GREM1, EGR1, GDF15 and SOX9, that may represent a valuable tool for routine testing of PM0.5-induced in vitro toxicity in lung epithelial cells.

Mitochondrial uncoupling and longevity - A role for mitokines?

Aging has been viewed both as a random process due to accumulation of molecular and cellular damage over time and as a programmed process linked to cellular pathway important for growth and maturation. These views converge on mitochondria as both the major producer of damaging reactive oxidant species (ROS) and as signaling organelles. A finite proton leak across the inner mitochondrial membrane leading to a slight uncoupling of oxidative phosphorylation and respiration is an intrinsic property of all mitochondria and according to the "uncoupling to survive" hypothesis it has evolved to protect against ROS production to minimize oxidative damage. This hypothesis is supported by evidence linking an increased endogenous, uncoupling protein (UCP1) mediated, as well as experimentally induced mitochondrial uncoupling to an increased lifespan in rodents. This is possibly due to the synergistic activation of molecular pathways linked to life extending effects of caloric restriction as well as a mitohormetic response. Mitohormesis is an adaptive stress response through mitonuclear signaling which increases stress resistance resulting in health promoting effects. Part of this response is the induction of fibroblast growth factor 21 (FGF21) and growth and differentiation factor 15 (GDF15), two stress-induced mitokines which elicit beneficial systemic metabolic effects via endocrine action.

Assessing Chemical-Induced Liver Injury In Vivo From In Vitro Gene Expression Data in the Rat: The Case of Thioacetamide Toxicity

Consumers are exposed to thousands of chemicals with potentially adverse health effects. However, these chemicals will never be tested for toxicity because of the immense resources needed for animal-based (in vivo) toxicological studies. Today, there are no viable in vitro alternatives to these types of animal studies. To develop an in vitro approach, we investigated whether we could predict in vivo organ injuries in rats with the use of RNA-seq data acquired from tissues early in the development of toxicant-induced injury, by comparing gene expression data from RNA isolated from these rat tissues with those obtained from in vitro exposure of primary liver and kidney cells. We collected RNA-seq data from the liver and kidney tissues of Sprague-Dawley rats 8 or 24 h after exposing them to vehicle (control), low (25 mg/kg), or high (100 mg/kg) doses of thioacetamide, a known liver toxicant that promotes fibrosis; we used these doses and exposure times to cause only mild toxicant-induced injury. For the in vitro study, we treated two cell types from Sprague-Dawley rats, primary hepatocytes (vehicle; low, 0.025 mM; or high, 0.125 mM dose), and renal tube epithelial cells (vehicle; low, 0.125 mM; or high, 0.500 mM) dose) with the thioacetamide metabolite, thioacetamide-S-oxide, selecting in vitro doses and exposure times to recreate the early-stage toxicant-induced injury model that we achieved in vivo. RNA-seq data were collected 9 or 24 h after application of vehicle or thioacetamide-S-oxide. We found that our modular approach for the analysis of gene expression data derived from in vivo RNA-seq strongly correlated (R² > 0.6) with the in vitro results at two different dose levels of thioacetamide/thioacetamide-S-oxide after 24 h of exposure. The top-ranked liver injury modules in vitro correctly identified the ensuing development of liver fibrosis.

Unsupervised identification of disease states from high-dimensional physiological and histopathological profiles

The liver and kidney in mammals play central roles in protecting the organism from xenobiotics and are at high risk of xenobiotic-induced injury. Xenobiotic-induced tissue injury has been extensively studied from both classical histopathological and biochemical perspectives. Here, we introduce a machine-learning approach to analyze toxicological response. Unsupervised characterization of physiological and histological changes in a large toxicogenomic dataset revealed nine discrete toxin-induced disease states, some of which correspond to known pathology, but others were novel. Analysis of dynamics revealed transitions between disease states at constant toxin exposure, mostly toward decreased pathology, implying induction of tolerance. Tolerance correlated with induction of known xenobiotic defense genes and decrease of novel ferroptosis sensitivity biomarkers, suggesting ferroptosis as a druggable driver of tissue pathophysiology. Lastly, mechanism of body weight decrease, a known primary marker for xenobiotic toxicity, was investigated. Combined analysis of food consumption, body weight, and molecular biomarkers indicated that organ injury promotes cachexia by whole-body signaling through Gdf15 and Igf1, suggesting strategies for therapeutic intervention that may be broadly relevant to human disease.

Literature optimized integration of gene expression for organ-specific evaluation of toxicogenomics datasets

The study of drug toxicity in human organs is complicated by their complex inter-relations and by the obvious difficulty to testing drug effects on biologically relevant material. Animal models and human cell cultures offer alternatives for systematic and large-scale profiling of drug effects on gene expression level, as typically found in the so-called toxicogenomics datasets. However, the complexity of these data, which includes variable drug doses, time points, and experimental setups, makes it difficult to choose and integrate the data, and to evaluate the appropriateness of one or another model system to study drug toxicity (of particular drugs) of particular human organs. Here, we define a protocol to integrate drug-wise rankings of gene expression changes in toxicogenomics data, which we apply to the TG-GATEs dataset, to prioritize genes for association to drug toxicity in liver or kidney. Contrast of the results with sets of known human genes associated to drug toxicity in the literature allows to compare different rank aggregation approaches for the task at hand. Collectively, ranks from multiple models point to genes not previously associated to toxicity, notably, the PCNA clamp associated factor (PCLAF), and genes regulated by the master regulator of the antioxidant response NFE2L2, such as NQO1 and SRXN1. In addition, comparing gene ranks from different models allowed us to evaluate striking differences in terms of toxicity-associated genes between human and rat hepatocytes or between rat liver and rat hepatocytes. We interpret these results to point to the different molecular functions associated to organ toxicity that are best described by each model. We conclude that the expected production of toxicogenomics panels with larger numbers of drugs and models, in combination with the ongoing increase of the experimental literature in organ toxicity, will lead to increasingly better associations of genes for organism toxicity.

Embracing the Dark Side: Computational Approaches to Unveil the Functionality of Genes Lacking Biological Annotation in Drug-Induced Liver Injury

In toxicogenomics, functional annotation is an important step to gain additional insights into genes with aberrant expression that drive pathophysiological mechanisms. Nevertheless, there exists a gap on annotation of these genes which often hampers the interpretation of results and limits their applicability in translational medicine. In this study, we evaluated the coverage of functional annotations of differentially expressed genes (DEGs) induced by 10 selected compounds from the TG-GATEs database identified as high- or no-risk in causing drug-induced liver injury (most-DILI or no-DILI, respectively) using in vitro human data. Functional roles of DEGs not present in the most common biological annotation databases – termed “dark genes” – were unveiled via literature mining and via the identification of shared regulatory transcription factors or signaling pathways. Our results demonstrated that there were approximately 13% of dark genes induced by these compounds in vitro and we were able to obtain additional relevant information for up to 76% of those. Using interactome data from several sources, we have uncovered genes such as LRBA, and WDR26 as highly connected in the protein network that play roles in drug response. Genes such as MALAT1, H19, and MIR29C – whose links to hepatotoxicity have been confirmed – were identified as markers for the most-DILI group and appeared as top hits across all literature-based mining methods. Furthermore, we investigated the potential impact of dark genes on liver toxicity by identifying their rat orthologs in combination with their correlation to drug-induced liver pathologies observed in vivo following chemical exposure. We identified a set of important regulatory transcription factors of dark genes for all most-DILI compounds including E2F1 and JUND with supporting evidences in literature and we found Magee1 correlated with chemically induced bile duct hyperplasia and adverse responses at 29 days in rats in vivo. In conclusion, in this study we show the potential role of these poorly annotated genes in mechanisms underlying hepatotoxicity and offer a number of computational approaches that may help to minimize current gaps in gene annotation and highlight their values as potential biomarkers in toxicological studies.

Imidazo[1,2-b]pyrazole-7-carboxamides Induce Apoptosis in Human Leukemia Cells at Nanomolar Concentrations

Leukemia, the malignancy of the hematopoietic system accounts for 10% of cancer cases with poor overall survival rate in adults; therefore, there is a high unmet medical need for the development of novel therapeutics. Eight imidazo[1,2-b]pyrazole-7-carboxamides have been tested for cytotoxic activity against five leukemia cell lines: Acute promyelocytic leukemia (HL-60), acute monocytic leukemia (THP-1), acute T-lymphoblastic leukemia (MOLT-4), biphenotypic B myelomonocytic leukemia (MV-4-11), and erythroleukemia (K-562) cells in vitro. Imidazo[1,2-b]pyrazole-7-carboxamides hampered the viability of all five leukemia cell lines with different potential. Optimization through structure activity relationship resulted in the following IC50 values for the most effective lead compound DU385: 16.54 nM, 27.24 nM, and 32.25 nM on HL-60, MOLT-4, MV-4-11 cells, respectively. Human primary fibroblasts were much less sensitive in the applied concentration range. Both monolayer or spheroid cultures of murine 4T1 and human MCF7 breast cancer cells were less sensitive to treatment with 1.5⁻10.8 μM IC50 values. Flow cytometry confirmed the absence of necrosis and revealed 60% late apoptotic population for MV-4-11, and 50% early apoptotic population for HL-60. MOLT-4 cells showed only about 30% of total apoptotic population. Toxicogenomic study of DU385 on the most sensitive MV-4-11 cells revealed altered expression of sixteen genes as early (6 h), midterm (12 h), and late response (24 h) genes upon treatment. Changes in ALOX5AP, TXN, and SOD1 expression suggested that DU385 causes oxidative stress, which was confirmed by depletion of cellular glutathione and mitochondrial membrane depolarization induction. Imidazo[1,2-b]pyrazole-7-carboxamides reported herein induced apoptosis in human leukemia cells at nanomolar concentrations.

Machine Learning Based Toxicity Prediction: From Chemical Structural Description to Transcriptome Analysis

Toxicity prediction is very important to public health. Among its many applications, toxicity prediction is essential to reduce the cost and labor of a drug's preclinical and clinical trials, because a lot of drug evaluations (cellular, animal, and clinical) can be spared due to the predicted toxicity. In the era of Big Data and artificial intelligence, toxicity prediction can benefit from machine learning, which has been widely used in many fields such as natural language processing, speech recognition, image recognition, computational chemistry, and bioinformatics, with excellent performance. In this article, we review machine learning methods that have been applied to toxicity prediction, including deep learning, random forests, k-nearest neighbors, and support vector machines. We also discuss the input parameter to the machine learning algorithm, especially its shift from chemical structural description only to that combined with human transcriptome data analysis, which can greatly enhance prediction accuracy.

Sex modulates the association of fibroblast growth factor 21 with end-stage renal disease in Asian people with Type 2 diabetes: a 6.3-year prospective cohort study

AIM

To study whether plasma fibroblast growth factor 21 independently predicts the risk of end-stage renal disease in Asian people with Type 2 diabetes.

METHODS

In this prospective cohort study, 1700 Asian people with Type 2 diabetes were followed for a mean of 6.3 years in a regional hospital in Singapore. Incident end-stage renal disease was identified by linkage with a national renal registry. The association of baseline fibroblast growth factor 21 levels with risk of progression to end-stage renal disease was studied using survival analyses.

RESULTS

Participants were aged 60 ± 10 years, with an average diabetes duration of 12 years. Their estimated GFR was 73 ± 28 ml/min/1.73 m² and 62% had albuminuria at baseline. A total of 179 incident end-stage renal disease cases were identified. Plasma fibroblast growth factor 21 interacted with sex in its association with end-stage renal disease (P_interaction = 0.003). A 1-sd increment in fibroblast growth factor 21 (natural log-transformed) was associated with a 1.32-fold (95% CI 1.05-1.66, P = 0.02) increased hazard for end-stage renal disease in women, after adjustment for traditional risk factors including estimated GFR and albuminuria. Taking death as a competing risk did not materially change the outcome [sub-distribution hazard ratio 1.35 (95% CI 1.11-1.66, P = 0.003)]. Fibroblast growth factor 21 did not predict end-stage renal disease risk in men after adjustment for baseline estimated GFR and albuminuria [hazard ratio 1.07 (95% CI 0.89-1.28, P = 0.49)].

CONCLUSIONS

Plasma fibroblast growth factor 21 level independently predicted risk of progression to end-stage renal disease in women with Type 2 diabetes. The pathophysiological relationships among FGF21, sex and renal progression warrant further study.

A Sensitive In Vitro Approach to Assess the Hybridization-Dependent Toxic Potential of High Affinity Gapmer Oligonucleotides

The successful development of high-affinity gapmer antisense oligonucleotide (ASO) therapeutics containing locked nucleic acid (LNA) or constrained ethyl (cEt) substitutions has been hampered by the risk of hepatotoxicity. Here, we present an in vitro approach using transfected mouse fibroblasts to predict the potential hepatic liabilities of LNA-modified ASOs (LNA-ASOs), validated by assessing 236 different LNA-ASOs with known hepatotoxic potential. This in vitro assay accurately reflects in vivo findings and relates hepatotoxicity to RNase H1 activity, off-target RNA downregulation, and LNA-ASO-binding affinity. We further demonstrate that the hybridization-dependent toxic potential of LNA-ASOs is also evident in different cell types from different species, which indicates probable translatability of the in vitro results to humans. Additionally, we show that the melting temperature (T_m) of LNA-ASOs maintained below a threshold level of about 55°C greatly diminished the hepatotoxic potential. In summary, we have established a sensitive in vitro screening approach for assessing the hybridization-dependent toxic potential of LNA-ASOs, enabling prioritization of candidate molecules in drug discovery and early development.

ATF3-Dependent Regulation of EGR1 in vitro and in vivo

BACKGROUND/AIMS

Activating transcription factor 3 (ATF3) and early growth response protein 1 (EGR1) are reported to interact, but their use as prognostic factors in cancer is discussed controversially.

METHODS

We measured ATF3 and EGR1 gene expression changes in human mini-organ cultures (MOCs) of healthy nasal epithelia, UM-SCC-22B, and FADUDD cells after acid reflux exposure. Next, ATF3 and EGR1 gene expression was analysed in tumour tissues and related to the median expression of autologous reference tissue samples.

RESULTS

ATF3 and EGR1 mRNA expression was significantly reduced after consecutive exposure of MOCs at pH <7.0 to artificial gastric juice (refluxate). In contrast, ATF3 mRNA was upregulated significantly within the first hour of incubation. EGR1 mRNA exhibited no significant changes. The analysed cell lines exhibited a cell line-specific alteration. In FADUDD cells, the upregulation of EGR1 was significant after refluxate exposure, but in HN-SCC 22B, no significant changes were detected. The analysis of the HNSCC samples confirmed the heterogeneous data of the literature.

CONCLUSION

The data maintain the hypothesis that ATF3 and EGR1 are involved in the beginning of inflammatory processes. Whether these two transcription factors act as tumour suppressors or promoters is context dependent and warrants analysis in further studies.

The sbv IMPROVER Systems Toxicology Computational Challenge: Identification of Human and Species-Independent Blood Response Markers as Predictors of Smoking Exposure and Cessation Status

Cigarette smoking entails chronic exposure to a mixture of harmful chemicals that trigger molecular changes over time, and is known to increase the risk of developing diseases. Risk assessment in the context of 21^st century toxicology relies on the elucidation of mechanisms of toxicity and the identification of exposure response markers, usually from high-throughput data, using advanced computational methodologies. The sbv IMPROVER Systems Toxicology computational challenge (Fall 2015-Spring 2016) aimed to evaluate whether robust and sparse (≤40 genes) human (sub-challenge 1, SC1) and species-independent (sub-challenge 2, SC2) exposure response markers (so called gene signatures) could be extracted from human and mouse blood transcriptomics data of current (S), former (FS) and never (NS) smoke-exposed subjects as predictors of smoking and cessation status. Best-performing computational methods were identified by scoring anonymized participants' predictions. Worldwide participation resulted in 12 (SC1) and six (SC2) final submissions qualified for scoring. The results showed that blood gene expression data were informative to predict smoking exposure (i.e. discriminating smoker versus never or former smokers) status in human and across species with a high level of accuracy. By contrast, the prediction of cessation status (i.e. distinguishing FS from NS) remained challenging, as reflected by lower classification performances. Participants successfully developed inductive predictive models and extracted human and species-independent gene signatures, including genes with high consensus across teams. Post-challenge analyses highlighted "feature selection" as a key step in the process of building a classifier and confirmed the importance of testing a gene signature in independent cohorts to ensure the generalized applicability of a predictive model at a population-based level. In conclusion, the Systems Toxicology challenge demonstrated the feasibility of extracting a consistent blood-based smoke exposure response gene signature and further stressed the importance of independent and unbiased data and method evaluations to provide confidence in systems toxicology-based scientific conclusions.

Opportunities and challenges in phenotypic drug discovery: an industry perspective

Phenotypic drug discovery (PDD) approaches do not rely on knowledge of the identity of a specific drug target or a hypothesis about its role in disease, in contrast to the target-based strategies that have been widely used in the pharmaceutical industry in the past three decades. However, in recent years, there has been a resurgence in interest in PDD approaches based on their potential to address the incompletely understood complexity of diseases and their promise of delivering first-in-class drugs, as well as major advances in the tools for cell-based phenotypic screening. Nevertheless, PDD approaches also have considerable challenges, such as hit validation and target deconvolution. This article focuses on the lessons learned by researchers engaged in PDD in the pharmaceutical industry and considers the impact of 'omics' knowledge in defining a cellular disease phenotype in the era of precision medicine, introducing the concept of a chain of translatability. We particularly aim to identify features and areas in which PDD can best deliver value to drug discovery portfolios and can contribute to the identification and the development of novel medicines, and to illustrate the challenges and uncertainties that are associated with PDD in order to help set realistic expectations with regard to its benefits and costs.

A computational toxicogenomics approach identifies a list of highly hepatotoxic compounds from a large microarray database

The liver and the kidney are the most common targets of chemical toxicity, due to their major metabolic and excretory functions. However, since the liver is directly involved in biotransformation, compounds in many currently and normally used drugs could affect it adversely. Most chemical compounds are already labeled according to FDA-approved labels using DILI-concern scale. Drug Induced Liver Injury (DILI) scale refers to an adverse drug reaction. Many compounds do not exhibit hepatotoxicity at early stages of development, so it is important to detect anomalies at gene expression level that could predict adverse reactions in later stages. In this study, a large collection of microarray data is used to investigate gene expression changes associated with hepatotoxicity. Using TG-GATEs a large-scale toxicogenomics database, we present a computational strategy to classify compounds by toxicity levels in human and animal models through patterns of gene expression. We combined machine learning algorithms with time series analysis to identify genes capable of classifying compounds by FDA-approved labeling as DILI-concern toxic. The goal is to define gene expression profiles capable of distinguishing the different subtypes of hepatotoxicity. The study illustrates that expression profiling can be used to classify compounds according to different hepatotoxic levels; to label those that are currently labeled as undertemined; and to determine if at the molecular level, animal models are a good proxy to predict hepatotoxicity in humans.

Inferring Genes and Biological Functions That Are Sensitive to the Severity of Toxicity Symptoms

The effective development of new drugs relies on the identification of genes that are related to the symptoms of toxicity. Although many researchers have inferred toxicity markers, most have focused on discovering toxicity occurrence markers rather than toxicity severity markers. In this study, we aimed to identify gene markers that are relevant to both the occurrence and severity of toxicity symptoms. To identify gene markers for each of four targeted liver toxicity symptoms, we used microarray expression profiles and pathology data from 14,143 in vivo rat samples. The gene markers were found using sparse linear discriminant analysis (sLDA) in which symptom severity is used as a class label. To evaluate the inferred gene markers, we constructed regression models that predicted the severity of toxicity symptoms from gene expression profiles. Our cross-validated results revealed that our approach was more successful at finding gene markers sensitive to the aggravation of toxicity symptoms than conventional methods. Moreover, these markers were closely involved in some of the biological functions significantly related to toxicity severity in the four targeted symptoms.

Use of early phenotypic in vivo markers to assess human relevance of an unusual rodent non-genotoxic carcinogen in vitro

Foci of altered hepatocytes (FAH) are considered putative, pre-neoplastic lesions that can occur spontaneously in aging rodents, but can also be induced by chemicals or drugs. Progression of FAH to hepatocellular neoplasms has been reported repeatedly but increases in foci in rodents do not necessarily lead to tumors in carcinogenicity studies and the relevance for humans often remains unclear. Here we present the case of RG3487, a molecule which induced FAH and, later on, tumors in rats. Because the molecule was negative in genotoxicity assays it was classified as a non-genotoxic carcinogen. In order to assess the potential for liver tumor formation in humans, we analyzed treatment-induced changes in vivo to establish a possible mode of action (MoA). In vivo and in vitro gene expression analysis revealed that nuclear receptor signaling was unlikely to be the relevant MoA and no other known mechanism could be established. We therefore took an approach comparing phenotypic markers, including mRNA changes, proliferation and glycogen accumulation, in vitro using cells of different species to assess the human relevance of this finding. Since the alterations observed in rats were not seen in the liver of mice or dogs in vivo, we could validate the relevance of the cell models chosen by use of hepatocytes from these species in vitro. This ultimately allowed for a cross-species comparison, which suggested that the formation of FAH and liver tumors was rat specific and unlikely to translate to human. Our work showed that phenotypic species comparison in vitro is a useful approach for assessment of the human relevance of pre-clinical findings where no known mechanism can be established.

A Comparative Analysis of Drug-Induced Hepatotoxicity in Clinically Relevant Situations

Drug-induced toxicity is a significant problem in clinical care. A key problem here is a general understanding of the molecular mechanisms accompanying the transition from desired drug effects to adverse events following administration of either therapeutic or toxic doses, in particular within a patient context. Here, a comparative toxicity analysis was performed for fifteen hepatotoxic drugs by evaluating toxic changes reflecting the transition from therapeutic drug responses to toxic reactions at the cellular level. By use of physiologically-based pharmacokinetic modeling, in vitro toxicity data were first contextualized to quantitatively describe time-resolved drug responses within a patient context. Comparatively studying toxic changes across the considered hepatotoxicants allowed the identification of subsets of drugs sharing similar perturbations on key cellular processes, functional classes of genes, and individual genes. The identified subsets of drugs were next analyzed with regard to drug-related characteristics and their physicochemical properties. Toxic changes were finally evaluated to predict both molecular biomarkers and potential drug-drug interactions. The results may facilitate the early diagnosis of adverse drug events in clinical application.

Liver toxicity may occur at drug levels above the therapeutic range and is thus a crucial problem in clinical care. However, the cellular changes induced by drug administration of therapeutic and toxic doses in humans are still not well understood. We here coupled patient-specific drug concentration-time profiles following oral administration of therapeutic and toxic doses with in vitro drug response data to predict toxic changes that quantitatively reflect the transition from desired drug effects to undesired toxic reactions. These toxic changes were comparatively evaluated for fifteen hepatotoxic drugs to identify subsets of drugs, which show similar drug effects on key cellular processes, functional classes of genes, and individual genes, respectively. In addition, analyzing toxic changes for individual genes allowed the prediction of molecular biomarkers and potential drug-drug interactions. Our results may hence support the early diagnosis of liver toxicity in clinical care in the future and may, moreover, help to assess potential risks of drug combination therapies.

Patient-specific hepatocyte-like cells derived from induced pluripotent stem cells model pazopanib-mediated hepatotoxicity

Idiosyncratic drug-induced hepatotoxicity is a major cause of liver damage and drug pipeline failure, and is difficult to study as patient-specific features are not readily incorporated in traditional hepatotoxicity testing approaches using population pooled cell sources. Here we demonstrate the use of patient-specific hepatocyte-like cells (HLCs) derived from induced pluripotent stem cells for modeling idiosyncratic hepatotoxicity to pazopanib (PZ), a tyrosine kinase inhibitor drug associated with significant hepatotoxicity of unknown mechanistic basis. In vitro cytotoxicity assays confirmed that HLCs from patients with clinically identified hepatotoxicity were more sensitive to PZ-induced toxicity than other individuals, while a prototype hepatotoxin acetaminophen was similarly toxic to all HLCs studied. Transcriptional analyses showed that PZ induces oxidative stress (OS) in HLCs in general, but in HLCs from susceptible individuals, PZ causes relative disruption of iron metabolism and higher burden of OS. Our study establishes the first patient-specific HLC-based platform for idiosyncratic hepatotoxicity testing, incorporating multiple potential causative factors and permitting the correlation of transcriptomic and cellular responses to clinical phenotypes. Establishment of patient-specific HLCs with clinical phenotypes representing population variations will be valuable for pharmaceutical drug testing.

DTNI: a novel toxicogenomics data analysis tool for identifying the molecular mechanisms underlying the adverse effects of toxic compounds

Unravelling gene regulatory networks (GRNs) influenced by chemicals is a major challenge in systems toxicology. Because toxicant-induced GRNs evolve over time and dose, the analysis of global gene expression data measured at multiple time points and doses will provide insight in the adverse effects of compounds. Therefore, there is a need for mathematical methods for GRN identification from time-over-dose-dependent data. One of the current approaches for GRN inference is Time Series Network Identification (TSNI). TSNI is based on ordinary differential equations (ODE), describing the time evolution of the expression of each gene, which is assumed to be dependent on the expression of other genes and an external perturbation (i.e. chemical exposure). Here, we present Dose-Time Network Identification (DTNI), a method extending TSNI by including ODE describing how the expression of each gene evolves with dose, which is supposed to depend on the expression of other genes and the exposure time. We also adapted TSNI in order to enable inclusion of time-over-dose-dependent data from multiple compounds. Here, we show that DTNI outperforms TSNI in inferring a toxicant-induced GRN. Moreover, we show that DTNI is a suitable method to infer a GRN dose- and time-dependently induced by a group of compounds influencing a common biological process. Applying DTNI on experimental data from TG-GATEs, we demonstrate that DTNI provides in-depth information on the mode of action of compounds, in particular key events and potential molecular initiating events. Furthermore, DTNI also discloses several unknown interactions which have to be verified experimentally.

Detecting cognizable trends of gene expression in a time series RNA-sequencing experiment: a bootstrap approach

Study of temporal trajectory of gene expression is important. RNA sequencing is popular in genome-scale studies of transcription. Because of high expenses involved, many time-course RNA sequencing studies are challenged by inadequacy of sample sizes. This poses difficulties in conducting formal statistical tests of significance of null hypotheses. We propose a bootstrap algorithm to identify 'cognizable' 'time-trends' of gene expression. Properties of the proposed algorithm are derived using a simulation study. The proposed algorithm captured known 'time-trends' in the simulated data with a high probability of success, even when sample sizes were small (n < 10). The proposed statistical method is efficient and robust to capture 'cognizable' 'time-trends' in RNA sequencing data.

Microfluidic-Based Multi-Organ Platforms for Drug Discovery

Development of predictive multi-organ models before implementing costly clinical trials is central for screening the toxicity, efficacy, and side effects of new therapeutic agents. Despite significant efforts that have been recently made to develop biomimetic in vitro tissue models, the clinical application of such platforms is still far from reality. Recent advances in physiologically-based pharmacokinetic and pharmacodynamic (PBPK-PD) modeling, micro- and nanotechnology, and in silico modeling have enabled single- and multi-organ platforms for investigation of new chemical agents and tissue-tissue interactions. This review provides an overview of the principles of designing microfluidic-based organ-on-chip models for drug testing and highlights current state-of-the-art in developing predictive multi-organ models for studying the cross-talk of interconnected organs. We further discuss the challenges associated with establishing a predictive body-on-chip (BOC) model such as the scaling, cell types, the common medium, and principles of the study design for characterizing the interaction of drugs with multiple targets.

Model-based contextualization of in vitro toxicity data quantitatively predicts in vivo drug response in patients

Understanding central mechanisms underlying drug-induced toxicity plays a crucial role in drug development and drug safety. However, a translation of cellular in vitro findings to an actual in vivo context remains challenging. Here, physiologically based pharmacokinetic (PBPK) modeling was used for in vivo contextualization of in vitro toxicity data (PICD) to quantitatively predict in vivo drug response over time by integrating multiple levels of biological organization. Explicitly, in vitro toxicity data at the cellular level were integrated into whole-body PBPK models at the organism level by coupling in vitro drug exposure with in vivo drug concentration–time profiles simulated in the extracellular environment within the organ. PICD was exemplarily applied on the hepatotoxicant azathioprine to quantitatively predict in vivo drug response of perturbed biological pathways and cellular processes in rats and humans. The predictive accuracy of PICD was assessed by comparing in vivo drug response predicted for rats with observed in vivo measurements. To demonstrate clinical applicability of PICD, in vivo drug responses of a critical toxicity-related pathway were predicted for eight patients following acute azathioprine overdoses. Moreover, acute liver failure after multiple dosing of azathioprine was investigated in a patient case study by use of own clinical data. Simulated pharmacokinetic profiles were therefore related to in vivo drug response predicted for genes associated with observed clinical symptoms and to clinical biomarkers measured in vivo. PICD provides a generic platform to investigate drug-induced toxicity at a patient level and thus may facilitate individualized risk assessment during drug development.

In Vitro Model for Hepatotoxicity Studies Based on Primary Human Hepatocyte Cultivation in a Perfused 3D Bioreactor System

Accurate prediction of the potential hepatotoxic nature of new pharmaceuticals remains highly challenging. Therefore, novel in vitro models with improved external validity are needed to investigate hepatic metabolism and timely identify any toxicity of drugs in humans. In this study, we examined the effects of diclofenac, as a model substance with a known risk of hepatotoxicity in vivo, in a dynamic multi-compartment bioreactor using primary human liver cells. Biotransformation pathways of the drug and possible effects on metabolic activities, morphology and cell transcriptome were evaluated. Formation rates of diclofenac metabolites were relatively stable over the application period of seven days in bioreactors exposed to 300 µM diclofenac (300 µM bioreactors (300 µM BR)), while in bioreactors exposed to 1000 µM diclofenac (1000 µM BR) metabolite concentrations declined drastically. The biochemical data showed a significant decrease in lactate production and for the higher dose a significant increase in ammonia secretion, indicating a dose-dependent effect of diclofenac application. The microarray analyses performed revealed a stable hepatic phenotype of the cells over time and the observed transcriptional changes were in line with functional readouts of the system. In conclusion, the data highlight the suitability of the bioreactor technology for studying the hepatotoxicity of drugs in vitro.

Assessing Concordance of Drug-Induced Transcriptional Response in Rodent Liver and Cultured Hepatocytes

The effect of drugs, disease and other perturbations on mRNA levels are studied using gene expression microarrays or RNA-seq, with the goal of understanding molecular effects arising from the perturbation. Previous comparisons of reproducibility across laboratories have been limited in scale and focused on a single model. The use of model systems, such as cultured primary cells or cancer cell lines, assumes that mechanistic insights derived from the models would have been observed via in vivo studies. We examined the concordance of compound-induced transcriptional changes using data from several sources: rat liver and rat primary hepatocytes (RPH) from Drug Matrix (DM) and open TG-GATEs (TG), human primary hepatocytes (HPH) from TG, and mouse liver / HepG2 results from the Gene Expression Omnibus (GEO) repository. Gene expression changes for treatments were normalized to controls and analyzed with three methods: 1) gene level for 9071 high expression genes in rat liver, 2) gene set analysis (GSA) using canonical pathways and gene ontology sets, 3) weighted gene co-expression network analysis (WGCNA). Co-expression networks performed better than genes or GSA when comparing treatment effects within rat liver and rat vs. mouse liver. Genes and modules performed similarly at Connectivity Map-style analyses, where success at identifying similar treatments among a collection of reference profiles is the goal. Comparisons between rat liver and RPH, and those between RPH, HPH and HepG2 cells reveal lower concordance for all methods. We observe that the baseline state of untreated cultured cells relative to untreated rat liver shows striking similarity with toxicant-exposed cells in vivo, indicating that gross systems level perturbation in the underlying networks in culture may contribute to the low concordance.

Gene expression studies in model systems are widely used for understanding the mechanism of drugs and other perturbations in biological systems. Other researchers have examined the reproducibility of microarray studies between laboratories, or comparing microarrays and/or RNA sequencing. However, no large scale studies have compared results from protocols which differ in minor details, or results generated in vivo vs. in vitro culture systems thought to serve as useful models. The rat liver is by far the most extensively studied model evaluating effects of drugs and other perturbations, and existing data allowed us to assess the level of concordance between rat liver and rat primary hepatocytes cultured in collagen-coated plates (i.e. “flat” culture) for hundreds of drugs. We found that the mouse liver serves as a better model of the rat liver than do rat primary hepatocytes, even after allowing for differences due to pharmacokinetics. The low concordance observed between rat liver and rat hepatocytes suggests that validating the utility of ‘omics data generated on emerging cell culture approaches (e.g. “organ-on-a-chip”, 3D-printed tissues) using rat cells and comparison to the rat liver may be necessary in order to gain confidence these approaches substantially improve on traditional culture models of human cells.

Characterization of Conserved Toxicogenomic Responses in Chemically Exposed Hepatocytes across Species and Platforms

BACKGROUND

Genome-wide expression profiling is increasingly being used to identify transcriptional changes induced by drugs and environmental stressors. In this context, the Toxicogenomics Project-Genomics Assisted Toxicity Evaluation system (TG-GATEs) project generated transcriptional profiles from rat liver samples and human/rat cultured primary hepatocytes exposed to more than 100 different chemicals.

OBJECTIVES

To assess the capacity of the cell culture models to recapitulate pathways induced by chemicals in vivo, we leveraged the TG-GATEs data set to compare the early transcriptional responses observed in the liver of rats treated with a large set of chemicals with those of cultured rat and human primary hepatocytes challenged with the same compounds in vitro.

METHODS

We developed a new pathway-based computational pipeline that efficiently combines gene set enrichment analysis (GSEA) using pathways from the Reactome database with biclustering to identify common modules of pathways that are modulated by several chemicals in vivo and in vitro across species.

RESULTS

We found that some chemicals induced conserved patterns of early transcriptional responses in in vitro and in vivo settings, and across human and rat genomes. These responses involved pathways of cell survival, inflammation, xenobiotic metabolism, oxidative stress, and apoptosis. Moreover, our results support the transforming growth factor beta receptor (TGF-βR) signaling pathway as a candidate biomarker associated with exposure to environmental toxicants in primary human hepatocytes.

CONCLUSIONS

Our integrative analysis of toxicogenomics data provides a comprehensive overview of biochemical perturbations affected by a large panel of chemicals. Furthermore, we show that the early toxicological response occurring in animals is recapitulated in human and rat primary hepatocyte cultures at the molecular level, indicating that these models reproduce key pathways in response to chemical stress. These findings expand our understanding and interpretation of toxicogenomics data from human hepatocytes exposed to environmental toxicants.

Precision multidimensional assay for high-throughput microRNA drug discovery

Development of drug discovery assays that combine high content with throughput is challenging. Information-processing gene networks can address this challenge by integrating multiple potential targets of drug candidates' activities into a small number of informative readouts, reporting simultaneously on specific and non-specific effects. Here we show a family of networks implementing this concept in a cell-based drug discovery assay for miRNA drug targets. The networks comprise multiple modules reporting on specific effects towards an intended miRNA target, together with non-specific effects on gene expression, off-target miRNAs and RNA interference pathway. We validate the assays using known perturbations of on- and off-target miRNAs, and evaluate an ∼700 compound library in an automated screen with a follow-up on specific and non-specific hits. We further customize and validate assays for additional drug targets and non-specific inputs. Our study offers a novel framework for precision drug discovery assays applicable to diverse target families.

Progress in drug discovery can be hampered by a limited exploration of chemical space and the difficulty in assessing the full range of drug candidates' effects on living cells. Here the authors describe a cell-based assay to distinguish between off-target and specific effects of candidate compounds targeting micro RNAs.

Meta-Analysis of Large-Scale Toxicogenomic Data Finds Neuronal Regeneration Related Protein and Cathepsin D to Be Novel Biomarkers of Drug-Induced Toxicity

Undesirable toxicity is one of the main reasons for withdrawing drugs from the market or eliminating them as candidates in clinical trials. Although numerous studies have attempted to identify biomarkers capable of predicting pharmacotoxicity, few have attempted to discover robust biomarkers that are coherent across various species and experimental settings. To identify such biomarkers, we conducted meta-analyses of massive gene expression profiles for 6,567 in vivo rat samples and 453 compounds. After applying rigorous feature reduction procedures, our analyses identified 18 genes to be related with toxicity upon comparisons of untreated versus treated and innocuous versus toxic specimens of kidney, liver and heart tissue. We then independently validated these genes in human cell lines. In doing so, we found several of these genes to be coherently regulated in both in vivo rat specimens and in human cell lines. Specifically, mRNA expression of neuronal regeneration-related protein was robustly down-regulated in both liver and kidney cells, while mRNA expression of cathepsin D was commonly up-regulated in liver cells after exposure to toxic concentrations of chemical compounds. Use of these novel toxicity biomarkers may enhance the efficiency of screening for safe lead compounds in early-phase drug development prior to animal testing.