The chain of events leading from a toxic compound exposure to carcinogenicity is still barely understood. With the emergence of high-throughput sequencing, it is now possible to discover many different biological components simultaneously. Using two different RNA libraries, we sequenced the complete transcriptome of human HepG2 liver cells exposed to benzo[a]pyrene, a potent human carcinogen, across six time points. Data were integrated in order to reveal novel complex chemical-gene interactions. Notably, we hypothesized that the inhibition of MGMT, a DNA damage response enzyme, by the over-expressed miR-181a-1 3p induced by BaP, may lead to liver cancer over time.

In order to improve attrition rates of candidate-drugs there is a need for a better understanding of the mechanisms underlying drug-induced hepatotoxicity. We aim to further unravel the toxicological response of hepatocytes to a prototypical cholestatic compound by integrating transcriptomic and metabonomic profiling of HepG2 cells exposed to Cyclosporin A. Cyclosporin A exposure induced intracellular cholesterol accumulation and diminished intracellular bile acid levels. Performing pathway analyses of significant mRNAs and metabolites separately and integrated, resulted in more relevant pathways for the latter. Integrated analyses showed pathways involved in cell cycle and cellular metabolism to be significantly changed. Moreover, pathways involved in protein processing of the endoplasmic reticulum, bile acid biosynthesis and cholesterol metabolism were significantly affected. Our findings indicate that an integrated approach combining metabonomics and transcriptomics data derived from representative in vitro models, with bioinformatics can improve our understanding of the mechanisms of action underlying drug-induced hepatotoxicity. Furthermore, we showed that integrating multiple omics and thereby analyzing genes, microRNAs and metabolites of the opposed model for drug-induced cholestasis can give valuable information about mechanisms of drug-induced cholestasis in vitro and therefore could be used in toxicity screening of new drug candidates at an early stage of drug discovery.

The application of transcriptome analyses in molecular epidemiology studies has become a promising tool in order to evaluate the impact of environmental exposures. These analyses have a great value in establishing the exposome, the totality of human exposures, both by identifying the chemical nature of the exposures and the induced molecular responses. Transcriptomic signatures can be regarded as biomarker of exposure as well as markers of effect which reflect the interaction between individual genetic background and exposure levels. However, the biological interpretation of modulated gene expression profiles is a challenging task and translating affected molecular pathways into risk assessment, for instance in terms of cancer promoting or disease preventing responses, is a far from standardised process. Here, we describe the in-depth analyses of the gene expression responses in a human dietary intervention in which the interaction between genotype and exposure to a blueberry-apple juice containing a complex mixture of phytochemicals is investigated. We also describe how data on differences in genetic background combined with different effect markers can provide a better understanding of gene-environment interactions. Pathway analyses of differentially expressed genes in combination with gene were used to identify complex but strong changes in several biological processes like immune response, cell adhesion, lipid metabolism and apoptosis. These observed changes may lead to upgraded growth control, induced immunity, reduced platelet aggregation and activation, diminished production of reactive oxidative species by platelets, blood glucose homeostasis, regulation of blood lipid levels and increased apoptosis. Our findings demonstrate that applying transcriptomics to well-controlled human dietary intervention studies can provide insight into mechanistic pathways involved in disease prevention by dietary factors.

Arsenic is an established human carcinogen, but the mechanisms through which it contributes to for instance lung cancer development are still unclear. As arsenic is methylated during its metabolism, it may interfere with the DNA methylation process, and is therefore considered to be an epigenetic carcinogen. In the present study, we hypothesize that arsenic is able to induce DNA methylation changes, which lead to changes in specific gene expression, in pathways associated with lung cancer promotion and progression. A549 human adenocarcinoma lung cells were exposed to a low (0.08 µM), intermediate (0.4 µM) and high (2 µM) concentration of sodium arsenite for 1, 2 and 8 weeks. DNA was isolated for whole-genome DNA methylation analyses using NimbleGen 2.1 M deluxe promoter arrays. In addition, RNA was isolated for whole-genome transcriptomic analysis using Affymetrix microarrays. Arsenic modulated DNA methylation and expression levels of hundreds of genes in a dose-dependent and time-dependent manner. By combining whole-genome DNA methylation and gene expression data with possibly involved transcription factors, a large molecular interaction network was created based on transcription factor-target gene pairs, consisting of 216 genes. A tumor protein p53 (TP53) subnetwork was identified, showing the interactions of TP53 with other genes affected by arsenic. Furthermore, multiple other new genes were discovered showing altered DNA methylation and gene expression. In particular, arsenic modulated genes which function as transcription factor, thereby affecting target genes which are known to play a role in lung cancer promotion and progression.

Current testing models for predicting drug-induced liver injury are inadequate, as they basically under-report human health risks. We present here an approach towards developing pathways based on hepatotoxicity-associated gene groups derived from two types of publicly accessible hepatotoxicity databases, in order to develop drug-induced liver injury biomarker profiles. One human liver omics-based and four text-mining-based databases were explored for hepatotoxicity-associated gene lists. Over-representation analysis of these gene lists with a hepatotoxicant-exposed primary human hepatocytes data set showed that human liver omics gene lists performed better than text-mining gene lists and the results of the latter differed strongly between databases. However, both types of databases contained gene lists demonstrating biomarker potential. Visualizing those in pathway format may aid in interpreting the biomolecular background. We conclude that exploiting existing and openly accessible databases in a dedicated manner seems promising in providing venues for translational research in toxicology and biomarker development.

A comparison of RNA-seq and microarray data from samples treated with diverse drugs highlights a dependency of cross-platform concordance on treatment effect. The concordance of RNA-sequencing (RNA-seq) with microarrays for genome-wide analysis of differential gene expression has not been rigorously assessed using a range of chemical treatment conditions. Here we use a comprehensive study design to generate Illumina RNA-seq and Affymetrix microarray data from the same liver samples of rats exposed in triplicate to varying degrees of perturbation by 27 chemicals representing multiple modes of action (MOAs). The cross-platform concordance in terms of differentially expressed genes (DEGs) or enriched pathways is linearly correlated with treatment effect size (R2≈0.8). Furthermore, the concordance is also affected by transcript abundance and biological complexity of the MOA. RNA-seq outperforms microarray (93% versus 75%) in DEG verification as assessed by quantitative PCR, with the gain mainly due to its improved accuracy for low-abundance transcripts. Nonetheless, classifiers to predict MOAs perform similarly when developed using data from either platform. Therefore, the endpoint studied and its biological complexity, transcript abundance and the genomic application are important factors in transcriptomic research and for clinical and regulatory decision making.

RNA-seq facilitates unbiased genome-wide gene-expression profiling. However, its concordance with the well-established microarray platform must be rigorously assessed for confident uses in clinical and regulatory application. Here we use a comprehensive study design to generate Illumina RNA-seq and Affymetrix microarray data from the same set of liver samples of rats under varying degrees of perturbation by 27 chemicals representing multiple modes of action (MOA). The cross-platform concordance in terms of differentially expressed genes (DEGs) or enriched pathways is highly correlated with treatment effect size, gene-expression abundance and the biological complexity of the MOA. RNA-seq outperforms microarray (90% versus 76%) in DEG verification by quantitative PCR and the main gain is its improved accuracy for low expressed genes. Nonetheless, predictive classifiers derived from both platforms performed similarly. Therefore, the endpoint studied and its biological complexity, transcript abundance, and intended application are important factors in transcriptomic research and for decision-making. Emerging technologies facilitate basic science research, but their value in clinical and regulatory settings requires rigorous assessment and consensus within the research community. The U.S. Food and Drug Administration’s (FDA’s) initiative on advancing regulatory science embraces collaborations among various stakeholders to expedite translation of advancement in basic science to regulatory application1. In the past decade, microarrays have been a principal technology for analyzing transcriptomes to support drug development and safety evaluation2. The FDA launched the community-wide MicroArray Quality Control (MAQC) consortium to investigate the reliability and utility of microarrays in identifying differentially expressed genes (DEGs) and predicting patient/toxicity outcomes based on gene-expression data in the first (MAQC-I)3, 4 and second (MAQCII)5, 6 phases of the project, respectively. MAQC-I and MAQC-II demonstrated the critical Wang et al. Page 2 Nat Biotechnol. Author manuscript; available in PMC 2014 November 25. N IH -P A A uhor M anscript N IH -P A A uhor M anscript N IH -P A A uhor M anscript roles of a comprehensive study design and crowd sourcing model to reach community-wide consensus on the fit-for-purpose use of emerging technologies. High-throughput sequencing technologies provide new methods for whole-transcriptome analyses of gene expression7. Recently published studies have compared data obtained from microarrays and RNA-seq in terms of technical reproducibility, variance structure, absolute expression and detection of DEGs or gene isoforms8–20 (Supplementary Table 1). Some of these studies suggested that RNA-seq exhibits lower precision for weakly expressed genes owing to the nature of sampling21, 22, whereas others found higher sensitivity of RNA-seq for gene detection23, 24. The varied conclusions can be attributed to the fact that they used few treatment conditions and hence they do not cover a wide range of biologic complexity. Furthermore, the question has not been adequately addressed about whether predicting toxicity outcomes based on gene-expression data could be enhanced with RNA-seq over microarray. Under the umbrella of the third phase of the MAQC consortium3–6, also known as the SEquencing Quality Control (SEQC) project, we conducted a comprehensive study to evaluate RNA-seq in its differences and similarities to microarrays in terms of identifying DEGs and developing predictive models. In contrast to data generated as part of the SEQC project using reference RNA samples25, our study design provides a comparison of the transcription response for rat livers that each platform detects in terms of extensive chemical treatments, biologic replication and breath of shared mode of action (MOA) of the chemicals beyond simply monitoring performance metrics. Specifically, we report the results of a comparative analysis of gene expression responses profiled by Affymetrix microarray and Illumina RNA-seq in liver tissue from rats exposed to diverse chemicals. We used either microarray or RNA-seq data to generate DEGs and predictive models of MOA of each chemical. This allowed us to assess the influence of the chemical (referred hereafter as the ‘treatment effect’) on the concordance between RNA-seq and microarrays and on the performance of predictive models generated using each technology. Treatment effect is characterized by the number of DEGs and the overexpressed pathways underlying MOA of the chemical. We found that (i) the concordance between array and sequencing platforms for detecting the number of DEGs was positively correlated with the extensive perturbation elicited by the treatment, (ii) RNA-seq performed better than microarrays at detecting weakly expressed genes, and (iii) gene expression–based predictive models generated from RNA-seq and microarray data were similar. The experimental design also allowed us to identify positive correlations in differentially expressed RNA elements (mRNA, splice variants, non-coding RNA and exon-exon junction) with the extensive perturbation elicited by the treatment, and to examine treatment-induced alternative splicing and shortening of 3’ untranslated regions (UTRs).

Quality control (QC) is crucial for any scientific method producing data. Applying adequate QC introduces new challenges in the genomics field where large amounts of data are produced with complex technologies. For DNA microarrays, specific algorithms for QC and pre-processing including normalization have been developed by the scientific community, especially for expression chips of the Affymetrix platform. Many of these have been implemented in the statistical scripting language R and are available from the Bioconductor repository. However, application is hampered by lack of integrative tools that can be used by users of any experience level. To fill this gap, we developed a freely available tool for QC and pre-processing of Affymetrix gene expression results, extending, integrating and harmonizing functionality of Bioconductor packages. The tool can be easily accessed through a wizard-like web portal at http://www.arrayanalysis.org or downloaded for local use in R. The portal provides extensive documentation, including user guides, interpretation help with real output illustrations and detailed technical documentation. It assists newcomers to the field in performing state-of-the-art QC and pre-processing while offering data analysts an integral open-source package. Providing the scientific community with this easily accessible tool will allow improving data quality and reuse and adoption of standards.

Blueberries contain relatively large amounts of different phytochemicals, which are suggested to have chemopreventive properties, but little information is available on the underlying molecular modes of action. This study investigates whole genome gene expression changes in lymphocytes of 143 humans after a 4-week blueberry-apple juice dietary intervention. Differentially expressed genes and genes correlating with the extent of antioxidant protection were identified in four subgroups. The magnitude of the preventive effect after the intervention differed between these four subgroups. Furthermore, subjects in two groups carried genetic polymorphisms that were previously found to influence the chemopreventive response. Pathway analysis of the identified genes showed strong but complex gene expression changes in pathways signaling for apoptosis, immune response, cell adhesion, and lipid metabolism. These pathways indicate increased apoptosis, upgraded growth control, induced immunity, reduced platelet aggregation and activation, blood glucose homeostasis, and regulation of fatty acid metabolism. Based on these observations, we hypothesize that combining transcriptomic data with phenotypic markers of oxidative stress may provide insight into the relevant cellular processes and genetic pathways, which contribute to the antioxidant response of complex mixtures of phytochemicals, such as found in blueberry-apple juice.

Acetaminophen is the primary cause of acute liver toxicity in Europe/USA, which led the FDA to reconsider recommendations concerning safe acetaminophen dosage/use. Unfortunately, the current tests for liver toxicity are no ideal predictive markers for liver injury, i.e. they only measure acetaminophen exposure after profound liver toxicity has already occurred. Furthermore, these tests do not provide mechanistic information. Here, 'omics techniques (global analysis of metabolomic/gene-expression responses) may provide additional insight. To better understand acetaminophen-induced responses at low doses, we evaluated the effects of (sub-)therapeutic acetaminophen doses on metabolite formation and global gene-expression changes (including, for the first time, full-genome human miRNA expression changes) in blood/urine samples from healthy human volunteers. Many known and several new acetaminophen-metabolites were detected, in particular in relation to hepatotoxicity-linked, oxidative metabolism of acetaminophen. Transcriptomic changes indicated immune-modulating effects (2g dose) and oxidative stress responses (4g dose). For the first time, effects of acetaminophen on full-genome human miRNA expression have been considered and confirmed the findings on mRNA level. 'Omics techniques outperformed clinical chemistry tests and revealed novel response pathways to acetaminophen in humans. Although no definitive conclusion about potential immunotoxic effects of acetaminophen can be drawn from this study, there are clear indications that the immune system is triggered even after intake of low doses of acetaminophen. Also, oxidative stress-related gene responses, similar to those seen after high dose acetaminophen exposure, suggest the occurrence of possible pre-toxic effects of therapeutic acetaminophen doses. Possibly, these effects are related to dose-dependent increases in levels of hepatotoxicity-related metabolites.

Whole-genome transcriptome measurements are pivotal for characterizing molecular mechanisms of chemicals and predicting toxic classes, such as genotoxicity and carcinogenicity, from in vitro and in vivo assays. In recent years, deep sequencing technologies have been developed that hold the promise of measuring the transcriptome in a more complete and unbiased manner than DNA microarrays. Here, we applied this RNA-seq technology for the characterization of the transcriptomic responses in HepG2 cells upon exposure to benzo[a]pyrene (BaP), a well-known DNA damaging human carcinogen. Based on EnsEMBL genes, we demonstrate that RNA-seq detects ca 20% more genes than microarray-based technology but almost threefold more significantly differentially expressed genes. Functional enrichment analyses show that RNA-seq yields more insight into the biology and mechanisms related to the toxic effects caused by BaP, i.e., two- to fivefold more affected pathways and biological processes. Additionally, we demonstrate that RNA-seq allows detecting alternative isoform expression in many genes, including regulators of cell death and DNA repair such as TP53, BCL2 and XPA, which are relevant for genotoxic responses. Moreover, potentially novel isoforms were found, such as fragments of known transcripts, transcripts with additional exons, intron retention or exon-skipping events. The biological function(s) of these isoforms remain for the time being unknown. Finally, we demonstrate that RNA-seq enables the investigation of allele-specific gene expression, although no changes could be observed. Our results provide evidence that RNA-seq is a powerful tool for toxicology, which, compared with microarrays, is capable of generating novel and valuable information at the transcriptome level for characterizing deleterious effects caused by chemicals.

Toxicological studies assessing the safety of compounds for humans frequently use in vitro systems to characterize toxic responses in combination with transcriptomic analyses. Thus far, changes have mostly been investigated at the mRNA level. Recently, microRNAs have attracted attention because they are powerful negative regulators of mRNA levels and, thus, may be responsible for the modulation of important mRNA networks implicated in toxicity. This study aimed to identify possible microRNA-mRNA networks as novel interactions on the gene expression level after a genotoxic insult. We used benzo[a]pyrene (BaP), a polycyclic aromatic hydrocarbon, as a model genotoxic/carcinogenic compound. We analyzed time-dependent effects on mRNA and microRNA profiles in HepG2 cells, a widely used human liver cell line that expresses active p53 and is competent for the biotransformation of BaP. Changes in microRNA expression in response to BaP, in combination with multiple alterations of mRNA levels, were observed. Many of these altered mRNAs are targets of altered microRNAs. Using pathway analysis, we evaluated the relevance of such microRNA deregulations to genotoxicity. This revealed eight microRNAs that appear to participate in specific BaP-responsive pathways relevant to genotoxicity, such as apoptotic signaling, cell cycle arrest, DNA damage response, and DNA damage repair. Our results particularly highlight the potential of microRNA-29b, microRNA-26a-1*, and microRNA-122* as novel players in the BaP response. Therefore, this study demonstrates the added value of an integrated microRNA-mRNA approach for identifying molecular mechanisms induced by BaP in an in vitro human model.

UNLABELLED:We introduce GO-Elite, a flexible and powerful pathway analysis tool for a wide array of species, identifiers (IDs), pathways, ontologies and gene sets. In addition to the Gene Ontology (GO), GO-Elite allows the user to perform over-representation analysis on any structured ontology annotations, pathway database or biological IDs (e.g. gene, protein or metabolite). GO-Elite exploits the structured nature of biological ontologies to report a minimal set of non-overlapping terms. The results can be visualized on WikiPathways or as networks. Built-in support is provided for over 60 species and 50 ID systems, covering gene, disease and phenotype ontologies, multiple pathway databases, biomarkers, and transcription factor and microRNA targets. GO-Elite is available as a web interface, GenMAPP-CS plugin and as a cross-platform application. AVAILABILITY:http://www.genmapp.org/go_elite

In the present study, the effect of Trichostatin A (TSA), a histone deacetylase inhibitor, was investigated on the microRNA (miR, miRNA) expression profile in cultured primary rat hepatocytes by means of microarray analysis. Simultaneously, albumin secretory capacity and morphological features of the hepatocytes were evaluated throughout the culture time. In total, 25 out of 348 miRNAs were found to be differentially expressed between freshly isolated hepatocytes and 7-day cultured cells. Nineteen of these miRNAs were connected with 'general metabolism'. miR-21 and miR-126 were shown to be the most up and down regulated miRs upon cultivation and could be linked to the proliferative response triggered in the hepatocytes upon their isolation from the liver. miR-379 and miR-143, on the other hand, were found to be the most up and down regulated miRs upon TSA treatment. Together with the higher expression of miR-122 observed in TSA-treated versus non-treated cultures, we hypothesize that the changes observed for miR-122, miR-143 and miR-379 could be related to the inhibitory effects of TSA on hepatocellular proliferation.

Current high-throughput technology in genomics creates a large amount of biological data. Bioinformatics approaches are directed towards understanding such data on a systems biology level. Advanced mathematical methods like principal component analysis, clustering, neural networks, support vector machine (SVM) approaches and neural networks can help to find patterns in the data. However, to really understand the data the patterns must be combined with existing knowledge. One of the approaches to do so is to associate these data to functional classifications such as can be found in the Gene Ontology. Other methods focus on using biological pathways coming from both public and private pathway databases like KEGG, WikiPathways, Reactome, and MetaCore. Some of these pathway databases contain rather aspecific information about which genes are involved in reactions. A single identifier, like an Enzyme Code, may correspond to a full family of enzymes, whereas only one family member is responsible for the reaction in the given biological context. Another problem arises in the availability of relevant biological pathways for both rodents and humans. This is especially true in the field of toxicology, where many pathways are still lacking. The focus of this project was to create phase I and phase II detoxification pathways using PathVisio, the pathway editor of WikiPathways. New pathway content was generated using information obtained from several trusted resources: a) handbooks, b) manual/automated PubMed literature searches; c) online pathway resources; d) UniProt Knowledgebase; e) EnsEMBL; f) GeneCards. In addition, the content of some existing toxicology pathways was improved. To illustrate the usefulness of the new pathways data from experiments with carcinogenic compounds were retrieved from the main online microarray data repositories Gene Expression Omnibus (GEO) and ArrayExpress. The resulting data were visualized on the new pathways using PathVisio. All these pathways are made available to the community at WikiPathways, where they can be further used for statistical pathway analysis and visualization in the (toxico)genomics field.

In recent decades, our knowledge of the genetics and functional genomics of drug-metabolizing enzymes has increased and a wealth of data on drug-related 'omics' has become available. Despite the availability of large amounts of biological information on xenobiotic biotransformation, the number of available biotransformation pathway maps that can easily be used for visualization of multiple omics data is limited. Here, we created integrated biotransformation pathway maps suitable for multiple omics analysis using PathVisio. The ease of visualizing data on these maps was demonstrated by using published microarray data from human hepatocyte-like cell models, exemplifying - where a sufficient capacity for metabolizing chemicals is a prerequisite for a suited model - how the biotransformation pathway maps can be used for model selection.

Inflammatory bowel diseases (IBD) are immunomediated ailments affecting millions of individuals. Although diet is regarded as an important factor influencing IBD, there are no accepted dietary recommendations presently available. We administered 7·6 % lyophilised apples obtained from two cultivars (Golden Delicious and Marie Ménard, low and high in polyphenols, respectively) to HLA-B27 transgenic rats which develop spontaneous IBD. After 3 months feeding, rats fed Marie Ménard apples had reduced myeloperoxidase activity (3·6 (sem 0·3) v. 2·2 (sem 0·2) U/g tissue; P < 0·05) and reduced cyclo-oxygenase-2 (P < 0·05) and inducible NO synthase gene expression (P < 0·01) in the colon mucosa and significantly less diarrhoea (P < 0·05), compared with control rats. Cell proliferation in the colon mucosa was reduced significantly by feeding Golden Delicious apples, with a borderline effect of Marie Ménard apples. Gene expression profiling of the colon mucosa, analysed using the Whole Rat Genome 4 × 44 K Agilent Arrays, revealed a down-regulation of the pathways of PG synthesis, mitogen-activated protein kinase (MAPK) signalling and TNFα–NF-κB in Marie Ménard-fed rats. In the stools of the animals of this group we also measured a significant reduction of bacteria of the Bacteriodes fragilis group. In conclusion, the administration of Marie Ménard apples, rich in polyphenols and used at present only in the manufacturing of cider, ameliorates colon inflammation in transgenic rats developing spontaneous intestinal inflammation, suggesting the possible use of these and other apple varieties to control inflammation in IBD patients.

Selenium is an essential micronutrient. Its recommended daily allowance is not attained by a significant proportion of the population in many countries and its intake has been suggested to affect colorectal carcinogenesis. Therefore, microarrays were used to determine how both selenoprotein and global gene expression patterns in the mouse colon were affected by marginal selenium deficiency comparable to variations in human dietary intakes. Two groups of 12 mice each were fed a selenium-deficient (0.086 mg Se/kg) or a selenium-adequate (0.15 mg Se/kg) diet. After 6 wk, plasma selenium level, liver, and colon glutathione peroxidase (GPx) activity in the deficient group was 12, 34, and 50%, respectively, of that of the adequate group. Differential gene expression was analysed with mouse 44K whole genome microarrays. Pathway analysis by GenMAPP identified the protein biosynthesis pathway as most significantly affected, followed by inflammation, Delta-Notch and Wnt pathways. Selected gene expression changes were confirmed by quantitative real-time PCR. GPx1 and the selenoproteins W, H, and M, responded significantly to selenium intake making them candidates as biomarkers for selenium status. Thus, feeding a marginal selenium-deficient diet resulted in distinct changes in global gene expression in the mouse colon. Modulation of cancer-related pathways may contribute to the higher susceptibility to colon carcinogenesis in low selenium status.