New Challenges: Omics Technologies in Ecotoxicology

6 New Challenges: Omics Technologies in Ecotoxicology

6.1. Introduction The life sciences have evolved significantly over the past two decades. The change in scales induced by emerging technologies, most particularly next-generation sequencing (NGS) and high-resolution mass spectrometry, has revolutionized the approach to the exploration and development of living systems. The use of NGS in the environmental sciences presents enormous potential but is only in its infancy. It opens research opportunities that will increase our knowledge of living systems, notably in non-model species in the molecular disciplines, but with an environmental relevance. Moreover, the genomic, proteomic and metabolomic approaches propose new tools to respond to the environmental challenges such as global warming and environmental pollution and their impacts on ecosystems (Faure and Joly 2015). The omics technologies applied to ecotoxicology open up new perspectives for the analysis of the impacts of contaminants in terms of the molecular mechanisms set off by toxic substances. This alliance has given birth to ecotoxicogenomics, a term proposed for the first time by Snape et al. (2004) (Figure 6.1). This field of study encompasses technologies developed at the end of the 1990s and has since then been widely used in medicine and pharmacology, and in toxicology since the 2000s. There is nothing to prevent them from also spreading to ecotoxicology. Chapter written by Odette PRAT and Davide DEGLI-ESPOSTI.

182

Ecotoxicology

This domain nevertheless requires the creation and use of complementary platforms: genomics, proteomics, structural biology, cellular imagery and experimental installations (mesocosms). The massive data generated require implementing adapted bioinformatic software, the infrastructures necessary to rapidly transfer these data, as well as the data scientist skills associated with such data processing. Research communities need these infrastructures to develop ambitious research programs.

Figure 6.1. Conceptual framework for ecotoxicogenomics according to Snape et al. (2004)

An analysis done in the ISI Web of Sciences regarding the number of annual publications and citations associated with the terms “toxicogenomics” and “ecotoxicogenomics” shows that ecotoxicogenomics is much less developed than toxicogenomics, and is currently essentially dedicated to human or rodent studies (Figure 6.2). 6.2. The omics methodologies Toxicogenomics covers the methods referred to as “omics” (see box), i.e. the study of gene expression profiling (transcriptome), protein expression profiling (proteome) or metabolite expression profiling (metabolome) from an organism, a tissue or cells. It aims to discern the links of cause and effect between the exposure of an organism to a chemical or physical agent and the alteration of the expression of these biomolecules. Another purpose is to identify biomarkers of exposure (or effect) to exogenous substances or physical stressors. Finally, it also aims to decipher the molecular mechanisms triggered by exposure to one or several stressors and to the source of physiological disturbances observed using classical methods.

Figure 6.2. Annual number of publications (A, C) and citations (B, D) associated with “toxicogenomics” or “ecotoxicogenomics”, respectively. Data from ISI Web of Science (June 2018)

New Challenges: Omics Technologies in Ecotoxicology 183

184

Ecotoxicology

Figure 6.3. Some of the species used in ecotoxicological assessment for which the genome has been sequenced

NOTES ON FIGURE 6.3 – All images come from Wikimedia commons and have been released into the public domain. Each image is attributed to the authors as follows: Caenorhabditis elegans to the NIH; Danio rerio to Azul; Daphnia magna to the Upper Midwest Environmental Sciences Center – USGS; Drosophila melanogaster to Botaurus; Eurytemora affinis to Natalia Sukhikh, Zoological Institute of RAS; Lymnaea stagnalis to Rex; and Mytilus galloprovincialis to Georges Jansoone. a) Haploid genome sizes are indicated. Genome size data or gene number estimates are from NCBI Genome database. b) Data from EnsemblMetazoa. c) Murgarella et al. (2016).

New Challenges: Omics Technologies in Ecotoxicology

185

However, toxicogenomics is also very useful to study and compare the normal functioning of organisms when the toxic chemical is not used for itself, but instead to disturb systems and thus deduce their functioning and compare the sensitivities or the adaptation mechanisms between phyla. Ecotoxicogenomics is the application of toxicogenomics to organisms that are representative of ecosystems and is used to study the harmful effects of chemicals on individuals and on their ecosystems. Ecotoxicogenomics is therefore intimately related to the knowledge of the genomes of the organisms studied when they are exposed to physicochemical constraints (pressure, temperature, toxic substances) in their environment. Even though genomic data on non-model organisms remain very limited (Figure 6.3), the application of toxicogenomics to a wide variety of organisms could provide a powerful tool to assess the harmful effects on ecosystems. This evolution of knowledge is already underway with the recent development of sequencing methods (see box). Although the problem of the lack of knowledge of the genomes of the organisms studied in ecotoxicology can be circumvented by high-speed gene-sequencing methods (see box), the management of the resources required to provide wide accessibility raises substantial organizational challenges. To solve this problem, a large number of consortia are attempting to coordinate different gene-sequencing project initiatives, and create exchange dynamics and good practices around the genomics of different phyla and subphyla (e.g. arthropods, consortium i5k:http://i5k.github.io; vertebrates Genome 10k Project: genome10k.soe.ucsc.edu). For example, Daphnia magna is a key organism whose habitat covers most of the aquatic areas from the Arctic to dry zones, and which presents a great sensitivity to contaminants. The genome of this crustacean has been sequenced and is accessible to the public (http://wfleabase.org/), which has allowed the development of genetic tools such as cDNA banks and high-density oligonucleotide chips. Poynton et al. (2012) analyzed the impact of silver nanoparticles compared to soluble silver salts on D. magna using gene expression profiles (the measurement of the transcriptome in several experimental conditions) and thus identified mechanisms and stress genes as possible biomarkers of exposure. Many classical studies conducted on this organism have linked the impact of chemical substances and parameters such as embryonic abnormalities, the

186

Ecotoxicology

population’s growth rate and phenotypic modifications such as carapax length (Kim et al. 2015). Although such studies are needed to identify toxic substances, they provide only little information on the mechanisms of their toxicity. Yet, understanding these mechanisms is essential to meet the challenges of ecotoxicology, most particularly to predict the toxic responses of numerous organisms that compose an ecosystem and contributing to the establishment of documented regulatory values (Snape et al. 2004). Next-generation sequencing. At the beginning of the 2000s, a set of new technologies able to process millions of reads simultaneously instead of 96 at a time, such as conventional capillary-based Sanger sequencing, came onto the market. These technologies have transformed today’s biology, given the drop in the cost of nucleic acid sequencing and the development of bioinformatic tools and platforms to handle the data generated. These transformations have increased the number of species whose genome sequence has become available, giving rise to emergent model organisms in molecular ecology and ecotoxicology. Recently, new advances have been promised by the development of third-generation sequencing technologies (single-molecule real-time or nanopore sequencing), which offer much longer reads (up to 100 kb) than the high-throughput short-read technologies. Genotyping can identify genetic variants (single-nucleotide polymorphisms, SNPs; structural variants; microsatellite polymorphisms) involved in the susceptibility to various diseases or driving population diversity. NGS technologies have allowed this SNP analysis in non-model organisms, thanks to the development of reduced-representation methods, such as restriction site-associated (RAD) sequencing. RAD sequencing can potentially identify the genetic basis of adaptation or vulnerability of environmentally relevant species to anthropogenic contamination (Laporte et al. 2016). RNA sequencing (RNA-Seq) has dramatically changed our view of the extent and complexity of eukaryotic transcriptomes (Wang et al. 2009). RNA-Seq enables differential gene expression and evolutionary studies in non-model species, without the need for prior genomic resources, although biological variance of the field population requires a higher number of replicates than studies with inbred animal models (Todd et al. 2016). Improvement in study design, especially aiming to prioritize replication over sequencing depth, avoiding technical bias and taking into account confounding factors associated with the variables of interest (such as season, temperature, physicochemical water properties) are required to increase the biological information of comparative transcriptomic analysis in ecotoxicology. Box 6.1. Next-generation sequencing technologies and applications

New Challenges: Omics Technologies in Ecotoxicology

187

Transcriptomics aims to describe the global mRNA profiles of cells, organs or whole organisms in a given condition. Microarray platforms are available for some model organisms (e.g. D. magna, Danio rerio) (Giraudo et al. 2015). However, RNA-sequencing (RNA-Seq) opened the possibility of obtaining transcriptomic data from any organism of interest, including many species of environmental relevance (Trapp et al. 2016a). Both RNA-Seq and microarrays provide relative expression values of thousands of genes per sample that can be analyzed using generalized linear models in order to identify differentially expressed genes in, for instance, different exposure conditions. Some recent developments allowed the RNA-Seq analysis of a selected panel of 1,500 genes to be implemented in high-throughput in vitro screening tests in the context of the US Federal Tox21 Program (Mav et al. 2018). Proteomics, similarly to transcriptomics, aims to describe global protein expression profiles in cells, organs, whole organisms, as well as in body fluids, in a given condition. Advances in mass spectrometry have increased the sensitivity of peptide detection in many biological matrices (Nesatyy and Suter 2008). Coupling RNA-Seq and shotgun proteomics has made it possible to conduct proteomics studies even in organisms without annotated genomes, opening the way for mechanistic ecotoxicoproteomics studies in the laboratory and in the field (Gouveia et al. 2017, Trapp et al. 2014). Moreover, proteomics allows the study of post-translational modification, such as phosphorylation, potentially increasing the understanding of protein activation or inhibition induced by environmental contaminants (Caruso et al. 2014, Lee 2013). Metabolomics aims to describe the whole metabolite repertoire of cells/organs, whole organisms or body fluids. Metabolites that can be addressed are sugars, amino acids, lipids and other small organic molecules. Molecule identification and quantification may be assessed using nuclear magnetic resonance or high-resolution mass spectrometry, notably for untargeted approaches that aim to identify unknown compounds. An advantage of metabolomics compared with proteomics or transcriptomics is the full independence of genomic knowledge and its direct applicability across multiple species (Bahamonde et al. 2016). Because transcriptomics analyses are more standardized and cost-effective than are the two other technologies, so far transcriptomics tends to be more frequently used. Box 6.2. Technologies based on global profiling of mRNAs, proteins and metabolites

188

Ecotoxicology

Bioinformatics and computational biology is a discipline crossing the domains of the life sciences and engineering. Although the term “bioinformatics” was first used in 1970, the strongest emergence of bioinformatics came from the completion of the Human Genome Project (Hogeweg 2011, Vincent and Charette 2015). The field explores biological data acquisition, management and interpretation. The core of the discipline focuses on algorithms, tool conception and designing data resource architectures (i.e. knowledge base databases such as NCBI, UniProt and PDB, to mention just a few), which require robust knowledge in computer programming. However, many biologists from different fields are currently using and combining various bioinformatic tools to address fundamental or applied biological questions (Smith 2015, Vincent and Charette 2015). Bioinformatics has greatly contributed to the omics revolution and has also established the basis for a transformation of biology into a quantitative science (Atwood et al. 2015). In the last decade, NGS technologies, mass spectrometry, microarrays and other high-throughput approaches have led to an increase of omics data by a factor of 10 every year (Berger et al. 2016). An increased amount of data and steady technological developments offer both new challenges and opportunities to the field. Here, we provide a brief summary of the main challenges and opportunities bioinformatics will encounter, notably when applied to disciplines such as ecology and ecotoxicology. Many challenges are similar to those the field has experienced since its infancy. Some of these old problems, such as managing huge data volumes, searching to integrate information from different databases, finding biologically relevant patterns in the data, protein annotation or orthology detection, still require development in the post-genomic and multi-omics era (Atwood et al. 2015). New problems will also surface. In the context of ecology and ecotoxicology, the most probable challenges are expected to come from genome assembly, orphan protein functional annotation, as well as portability and real-time analyses in the field. While technological advances, such as portable real-time DNA sequencing using nanopores or portable mass spectrometers, have taken small genome sequencing and chemical identification in water out of the laboratory, even in remote areas (Brennwald et al. 2016, Pomerantz et al. 2018), progress is needed to access cloud computing at remote field sites (Berger et al. 2016) or real-time in vivo biomarker detection. New opportunities are also expected as a result of the omics revolution in ecology and ecotoxicology. In particular, new algorithmic advances have been stimulated by the increased volume of genomic data. For instance, the concept of compressive acceleration has been applied to genomics (Loh et al. 2012). This method is based on a two-stage search, the first referred to as coarse is performed on subsequences that

New Challenges: Omics Technologies in Ecotoxicology

189

represent unique data, and the second, called fine search, is performed on a subset of the original database showing a certain similarity with any representative sequence identified in the first query. This approach provides orders of magnitude run-time improvements to BLAST searches, with improvements increasing as database grow (Berger et al. 2016). These advances in the field of informatics will improve the chances of success of very ambitious genome sequencing projects within the next 10 years, notably those aiming to sequence, catalog and characterize the genomes of Earth’s eukaryotic biodiversity (i.e. Earth BioGenome Project) (Lewin et al. 2018). These initiatives will provide new fundamental resources for the scientific questions raised by ecology and ecotoxicology. Among these outcomes, we can cite a better understanding of evolutionary laws and relationships among known organisms, and new knowledge on ecosystem composition and functions. Genomics of environmentally relevant species will allow the comparison and provide a better understanding of the response to pollutants in different species. This will help to clarify the effect of anthropogenic activities on biodiversity, leading to a knowledgebased improvement of environmental quality of soil, air and water. Box 6.3. Bioinformatics and computational biology

Planning experiments A classical error would be to carry out first-line biological tests stemming from hypotheses based on the effects frequently induced by toxic chemicals. This often means searching for the oxidizing stress, DNA fragmentation (comet assay), apoptosis and then conducting a toxicogenomic study for verification. Yet the advantage of toxicogenomics is carrying out these tests at the very beginning of the study and with no a priori in terms of the results expected. The toxicogenomic results should first and foremost drive the hypotheses on the toxicity mechanisms, and guide the choice of the biological methods to intersect and validate these hypotheses. This would provide a considerable gain in time. The choice of the experimental conditions is extremely important to avoid biases due to technical errors (number of biological and technical replicates, preparation and preservation of samples, reliability of the method used and robustness of the statistical analysis methods). In this planning phase, the choice of the concentrations and the experiment durations is particularly important. More specifically, the doses tested should always be realistic, i.e. compared to measured or predicted environmental concentrations; if not, if the doses are exaggerated, we will systematically find harmfulness for any substance studied. Too strong doses induce multiple cellular and molecular effects, certainly leading to a high mortality rate and

190

Ecotoxicology

do not shed light on the fine mechanisms, i.e. the initial molecular events that trigger toxicity. One must never lose sight of the idea that it is the combination of toxic effects and the level of exposure that determines the risk of toxicity. A well-designed toxicogenomic study provides a panoramic vision of the state of the organism considered and the metabolic or regulatory pathways involved. The multiplicity of doses and exposure time generally provide the proof of the evidence of the involvement of particular toxicity mechanisms. Biological data mining Finally, the lists of genes or proteins obtained do not necessarily overlap perfectly along a kinetic expression profile, for example. As sometimes sought for biomarkers of effect, it is possible that a gene, a protein or a metabolite be systematically over- or underexpressed in a test group compared to a control group, whatever exposure dose or duration is studied. However, this specific case is not the most frequent one observed. In most cases, genes obey a fine temporal regulation and are expressed over a short lapse of time or are subjected to an oscillatory mode of production (Raj et al. 2006). Other genes, belonging to the same complex, family or reactional cascade take over: it is therefore futile to attempt to cross the results at all costs in terms of specific genes (i.e. the same accession number in a genes or proteins database) along an expression profile. On the other hand, if the data are first regrouped (clustered) by metabolic or regulation pathways (canonical pathways), it often appears evident that it is an entire pathway that is solicited over time, even if the transcripts are not identical (Canovas et al. 2014, Pisani et al. 2015). This type of processing, called “data mining”, allows us to generate highly substantiated mechanistic hypotheses, given that they are based on a large volume of data. Similarly, it is often difficult to make the results of the transcriptome and the proteome correspond on the same organism in response to a given stimulus. Few correspondence is observed in terms of accession number in the genomic and proteomic databases. This is explained by temporal expression windows, which are different between genes and proteins, as well as by post-transcriptional and posttranslational modifications. However, as stated above, the same type of clustering by canonical pathways can be carried out for proteins and metabolites, as well as genes. This is where the value and power of the adverse outcome pathway (AOP)-type methods lie (see text), which ignore one–off comparisons (gene-to-gene or proteinto-protein) to establish an overall picture of the action modes. Box 6.4. Usual pitfalls

New Challenges: Omics Technologies in Ecotoxicology

191

6.3. Post-omic concepts Systems biology is the integrated study of biological systems (cells, tissues, organs or entire organisms) at the molecular level. This involves disturbing a system, recording the modifications in molecular expression of transcripts, proteins and metabolites, integrating and linking the data obtained as interactions between all the elements of a system to model its functioning. By analogy, systems toxicology consists of studying the disturbance of a cell, organelle, tissue, organ or organism in the specified conditions of exposure to biological or chemical substances or stressors. The modifications of conventional toxicological parameters and molecular parameters (mRNA, proteins, etc.) can thus be followed and all these data integrated to describe and understand the general mechanisms of toxicity. Powerful analysis tools exist to accomplish this task (e.g. Ingenuity® Pathway Analysis from Qiagen) to integrate and interpret data from different omics experiments (RNA-Seq, transcriptomic, metabolomic or proteomic expression profiles). Based on rich evolutive databases, as well as the literature, these software tools can link genes or proteins involved in the response to toxic chemicals with known metabolic or regulatory pathways. These analyses, likened to artificial intelligence, reveal the molecular mechanisms involved and identify new targets or potential biomarkers in the context of the biological systems studied. The omics data combined with these biological data mining analyses make it possible to generate mechanistic hypotheses that can then be verified with more classical tests (in situ hybridization, Western blot, gene silencing, CRISPR/Cas9, etc.). The responses to toxic substances and their interactions with environmental factors can occur at different temporal scales and at different levels of biological complexity. Ecotoxicological experiments should integrate temporal exposure models to take into account possible adaptive processes (Fischer et al. 2013). A classical toxicogenomic study consists of comparing biological elements (cells, tissues, individuals or populations) treated with one substance or a blend of potentially toxic substances with untreated control substances that have only received the same quantity of solvent used to vehicle the substance tested. These groups are generally exposed to several doses of the test substance and during variable exposure durations. At the end of the exposure period, the transcripts, proteins or

192

Ecotoxicology

metabolites are isolated and purified, and then analyzed through comparison with the control group. This provides lists of genes, proteins and metabolites that are significantly expressed in a differential manner compared to the control group. The omics methods capture a snapshot of the organism studied in response to a stimulus or a stressor. Consequently, there are advantages of multiplying doses to delineate a concentration threshold for effect and durations to observe the molecular modifications over time, and sometimes the reversibility of these effects. These technologies of course generate a massive quantity of data and require adapted databases, as well as data processing and statistical analysis software (see Box 6). Most of the classical methods used to assess the toxic impact of substances on organisms rely on the fact that they analyze the physiological response of an entire organism such as its growth, reproduction and mortality. These parameters are essential to identify toxic substances, but provide little information on the mechanisms of their toxicity. The AOP (adverse outcome pathway), i.e. “mechanisms that can lead to an adverse effect”, approach is a relatively new concept, which has been discussed since the 2010s (Ankley et al. 2010, Escher et al. 2017, Groh et al. 2015, Villeneuve et al. 2014a, b). This approach gave rise to a series of publications in environmental toxicology in 2011 (Kramer et al. 2011, Perkins et al. 2011, Villeneuve and Garcia-Reyero 2011, Watanabe et al. 2011) as well as OECD activities on the Development and Use of Adverse Outcome Pathways (Sakuratani et al. 2018). The AOP methodology is an approach that provides a conceptual framework (Figure 6.4) to collect, organize and assess relevant information on chemical, biological and toxicological effects – notably of toxic substances. This approach integrates the description of the modes and mechanisms of action of toxic substances through the scales of biological organization: cells, tissues, organisms and populations (Figure 6.5). It organizes the information acquired and allows a better comprehension

New Challenges: Omics Technologies in Ecotoxicology

193

of the progression of toxic events through these different strata. The development of AOPs can improve the comprehension and prediction of the effects related to chronic toxicity. This approach can be relevant in the context of risk assessment (Beyer et al. 2014). By fostering the elucidation of molecular mechanisms, the targeted development of AOPs can also help prioritize the toxicity tests to apply and promote the development of alternative methods to animal experimentation.

Figure 6.4. Conceptual diagram of key features of an adverse outcome pathway (AOP) according to Ankley et al. (2010). Each AOP begins with a molecular initiating event in which a chemical interacts with a biological target (anchor 1) leading to a sequential series of higher-order effects to produce an adverse outcome with direct relevance to a given risk assessment context (e.g. survival, development, reproduction; anchor 2). The first three boxes are the parameters that define a toxicity pathway

For example, Ankley et al. (Ankley et al. 2010) showed that hepatic production of vitellogenin in female fish, which is generated by the activation of the estrogenic receptor (ER), can be inhibited by the binding of estrogenic antagonists, leading to a reduction of vitellogenin in the blood circulation. At the same time, the chemicals that induce inhibition of aromatase, an enzyme that catalyzes the synthesis of ethinyl-estradiol (E2) from a precursor of testosterone, reduce the plasma concentrations of E2 and induce a decrease in the production of vitellogenin in female fish. These researchers thus described within an AOP the causal relation between a chain of multiple initial events affecting a common node (the production

194

Ecotoxicology

of vitellogenin in fish) and an adverse effect (reduced fertility), demonstrating the risk assessment at the population level. The definition of these AOPs provides a solid basis for the development of targeted in vitro assays or QSAR models specific to estrogenic regulation to evaluate the danger induced by a multitude of chemical substances in support of risk assessment, as demanded within the REACH program. Knowledge of certain AOPs has already resulted in the elaboration of trial validation methods for estrogenic disturbance, notably in the United States. The level of vitellogenin in female fish has become a classic parameter for detecting and assessing the biological consequences of endocrine-disrupting chemicals (Ankley and Johnson 2004). Box 6.5. Example of AOP use in ecotoxicology

Four fundamental principles guide the use of AOPs: – AOPs do not depend on the chemicals studied; – AOPs include reusable modules, made up of key events and relations between key events. An individual AOP comprising a single sequence of events is a pragmatic unit of development and evaluation of a broader AOP; – Networks made up of multiple AOPs that share events or relations between common events may be the functional unit of prediction for most of the scenarios of the real world; – AOPs are living documents that evolve over time as new knowledge is generated. The key events indicate the progression from one disturbance towards an undesirable result. Each event is defined as an independent measurement taken at a particular level of biological organization. These descriptions contribute to constructing the AOP knowledge base (Villeneuve et al. 2014a, b). Development of the AOP methodology provides a coherent and reliable approach while evolving. Independent of the chemical structure of toxic substances, this methodology can also be used to study the combined effects of multiple stressors or cocktails present in soils or water, a crucial environmental issue.

Figure 6.5. Information flow in ecotoxicological risk assessment: key features and suggestions for further development of adverse outcome pathways framework according to Groh et al. (2015)

New Challenges: Omics Technologies in Ecotoxicology 195

196

Ecotoxicology

NOTES ON FIGURE 6.5.– The top panel depicts the information flow during risk assessment, indicating the AOP position in the process. The farleft column defines several specific aspects, which are discussed in relation to the AOP framework and its potential extension to include additional information. Abbreviations: AOP, adverse outcome pathway; MIE, molecular initiating event; KER, key event relationship; KE, key event; AO, adverse outcome; AOP Xplorer, a tool for AOP visualization within the AOP Wiki resource; QSAR, quantitative structure–activity relationship; PBTK, physiologically based toxicokinetic models. 6.4. The organisms involved Many organisms play the role of sentinel organisms in ecotoxicology. D. magna (Kim et al. 2015), D. rerio (Rhee and Lee 2014, Scholz et al. 2008), Gammarus pulex/Gammarus fossarum (Bertin et al. 2016, Kunz et al. 2010) and Mytilus edulis are frequently used for aquatic species (Beyer et al. 2017). In terrestrial species, earthworms such as the Eisenia spp. (Gong and Perkins 2016) and Lumbricus spp. (Calisi et al. 2019) are considered to be excellent bioindicators for soil ecosystems given their close contact with the environment and their essential roles in soil paedogenesis, structure, fertility and the terrestrial food chain. These earthworms have also been widely used to assess environmental risks and chemical toxicity in the laboratory, and in the field. Recent progress has been made in the toxicogenomics of earthworms, used to assess the ecological impacts of contaminated soils on these organisms. According to Gong Perkins (2016), 25 transcriptomic studies, five proteomic studies and 35 metabolic studies have been conducted on these organisms. These studies have revealed new biomarkers and have detailed the mechanistic knowledge on the impact of contaminants (Gong and Perkins 2016). 6.5. The substances A wide variety of toxic chemicals have been studied using these methods. Since 2010, an exponential use of toxicogenomics has emerged to test the toxicity of metals (As, Cd, Pb, Ti, Zn, Ag), organic contaminants (PCBs, HAPs), insecticides (carbamyl, carbofuran, chlorpyrifos, DDT, endosulfan), herbicides (atrazine, glyphosate), fungicides (epoxiconazole), pharmaceuticals (estrogens, anti-cancer drugs) and nanoparticles. All substances are suitable

New Challenges: Omics Technologies in Ecotoxicology

197

for these analyses. Interestingly, a multiplexed approach using a selected reaction monitoring mass spectrometry methodology was applied to analyze 38 peptides reporting 25 proteins in caged amphipods (G. fossarum) in river waters. The organisms were caged in uncontaminated or reference sites. The responses obtained in contaminated sites included inductions of vitellogeninlike proteins in male organisms, inductions of Na+K+/ATPases and strong inhibitions of molt-related proteins such as chitinase and JHE-carboxylesterase (Gouveia et al. 2017). This proof of principle study showed the feasibility of applying a multi-marker approach in water quality monitoring. Evaluating the potential dangers of mixtures of chemicals is another challenge for research in ecotoxicity, with consequences for environmental risk assessment and regulatory toxicology. A large number of anthropic contaminants are suspected of being involved in combined toxicity phenomena. However, it is highly probable that these are combinations of compounds that present similar modes of action which, when used on a large scale, contribute to locally high exposure concentrations. These are the combinations that present the greatest risk for aquatic organisms. In addition, the compounds that most particularly affect sensitive organisms or developmental stages are also particularly preoccupying. The animal species that are in high ecological or trophic positions that may accumulate persistent organic pollutants (POPs) are generally exposed to a greater risk in the presence of these substances and their mixtures (Pisani et al. 2016). Current research in toxicology, which emphasizes the detailed development of mechanisms leading to adverse effects (AOPs), is particularly relevant for the toxicity of mixtures. The contaminants that act (directly or indirectly) on the disturbance of a biological system (e.g. growth, development and reproduction) by similar AOPs should take top priority in the research on the effects of mixing, because of the additivity of concentrations, which may lead to an amplified effect. However, as new data and analyses on combined toxicity situations emerge, a better description of the detailed mechanisms at play, as well as better predictions of these effects can be expected. This will contribute to improving the methods available to assess the risks for the environment and health of mixtures of chemical substances. Finally, these results can be expected to have an impact on environmental regulations concerning

198

Ecotoxicology

discharges of chemicals and effluents in Europe’s aquatic environments (Beyer et al. 2014). 6.6. The applications 6.6.1. Epigenetics Adapted responses to environmental changes (food availability, drought conditions, etc.) can set off phenotypic changes in plants and animals that affect their viability and their reproductive ability. Using sequencing to identify changes in DNA methylation, the structure of chromatin and the expression of small RNA, researchers can better understand how epigenetic factors contribute to controlling these traits and others in a space of interest (Burggren 2014, Kamstra et al. 2015, Vandegehuchte and Janssen 2014). However, the transmissibility of these acquired characteristics is far from being systematic (Vandegehuchte and Janssen 2011). This field of research is in full development, but remains limited for organisms with an environmental interest (Willett 2018, Wang et al. 2018). 6.6.2. Biomarkers in ecotoxicology Integrating toxicogenomic analyses as well as bioinformatic tools into ecotoxicology could a priori improve risk assessment procedures, through better knowledge of the modes of action and comparison. Measuring the levels of gene and/or protein expression, during exposure to chemical agents or stressors, could be used to develop robust molecular biomarkers, allowing early detection of environmental stress caused by chronic exposure, for example. Demonstrating the mechanisms of action of toxic agents makes it possible to focus interest on the biomolecules involved in the key events of these mechanisms. These biomolecules, particularly impacted by the substances tested, can form robust molecular biomarkers substituting phenotypic end points of exposure (Calzolai et al. 2007). In this way, alteration of the sperm quality of a crustacean, G. fossarum, a key sentinel species for freshwater biomonitoring, was studied after laboratory exposure to a variety of concentrations of xenobiotics (cadmium, methoxyfenozide and pyriproxyfen) (Trapp et al. 2015). The integrity of the reproduction process was assessed using sperm quality markers. A semi-quantitative proteomic study was conducted on male gonads to observe

New Challenges: Omics Technologies in Ecotoxicology

199

the biological impact of each of these substances. The changes in a total of 871 proteins were monitored in response to the toxic pressure. A drastic effect was observed on spermatozoid production, with a dose–response relation. Among the crustaceans, G. fossarum is the first case documented on the involvement of diverse protein families in reproduction and egg yolk formation processes (Trapp et al. 2016b). Most particularly, the proteins from comparative proteomic studies can also provide valuable panels of markers, capable of marking a toxic effect caused by one or a mixture of substances. More broadly, these panels could differentiate environmental sites subjected to varying degrees of pollution (Gouveia et al. 2017). 6.7. The challenges To respond to the challenges raised by ecotoxicology, several problems remain to be solved – some technological, but others fundamental concerning intra- (multigenerational sensitivity) and inter-species (species biodiversity) variability, as well as the relation between models such as AOPs applicable to both the organism and population levels. 6.7.1. Big data storage Huge quantities of data are generated by the omics approaches, as well as by imaging and geolocalization in ecology. A major challenge will be analyzing, synthesizing and storing these data sets, which largely surpass everything that has been stored up to the 2010s. Solutions to methodological interfaces will have to be promoted, which will allow the comparison of interdisciplinary results so that the complexity of biological systems can be described. Artificial intelligence systems should emerge, and there is an urgent need for public repository sites that are capable of combining omics data sets associated with biological, chemical and toxicological observations (Waters and Fostel 2004). 6.7.2. In silico modeling and prediction Dose–response modeling is one of the most important steps in assessing ecological risks. It requires relations between the concentrations of the

200

Ecotoxicology

substances studied and the effects observed for the species considered. The absence of toxicity data often makes it impossible to obtain dose–response relations for lethal (CL50), sublethal and carcinogenic effects. These shortcomings can be overcome by using diverse in silico methods. The mechanistic approaches such as toxicokinetic analysis (Chen et al. 2016) and toxicodynamics suit interspecific extrapolation better than statistical methods, such as structure–activity quantitative relations (Bradbury et al. 2003) and other empirical models such as Haber’s rule (Haber 1924). For the last 100 years, Haber’s rule has been successfully used to extrapolate from reported exposure times to other exposure times that may be needed for setting standards, health risk assessments and other applications. It has limitations, however, particularly in environmental applications where exposure levels are low and exposure times are relatively long. The Reduced Life Expectancy (RLE) model overcomes these problems and can be utilized under all exposure conditions (Connell et al. 2016). A new approach has been proposed in which the effects of exposure to toxic substances are quantified as a function of the probability of cell damage and interactions between metabolites. This approach recommends modeling cell damage with a toxicodynamic model and physiological or metabolite interactions with a toxicokinetic model. For example, this model, proposed to study aquatic species in the Arctic, provides more reliable estimations of toxicity, which will facilitate assessment of ecological risks in the Arctic environment (Fahd et al. 2017). – Ecological risk assessment quantifies the likelihood of undesirable impacts of stressors, primarily at high levels of biological organization. Data used to inform ecological risk assessments come primarily from tests on individual organisms or from suborganismal studies, indicating a disconnect between primary data and protection goals (Murphy et al. 2018). We know how to relate individual responses to population dynamics using individualbased models, and there are emerging ideas on how to make connections to ecosystem services. However, there is no established methodology to connect effects seen at higher levels of biological organization with suborganismal dynamics, despite progress made in identifying AOPs that link molecular initiating events to ecologically relevant key events. AOP models quantify explicit molecular-, cellular- or organ-level processes, but

New Challenges: Omics Technologies in Ecotoxicology

201

do not offer a route to linking suborganismal damage to adverse effects on individual growth, reproduction and survival, which can be propagated to the population level through individual-based models. The National Center for Mathematical and Biological Synthesis (NIMBioS) assessed the feasibility of using dynamic energy budget (DEB) models of individual organisms as a “pivot” connecting suborganismal processes to higher-level ecological processes. DEB models describe these processes, but use abstract variables with undetermined connections to suborganismal biology. Recently, this group proposed linking DEB and quantitative AOP models by interpreting AOP key events as measures of damage-inducing processes in a DEB model, using existing modeling tools available for both AOP and DEB (Murphy et al. 2018). 6.7.3. Variability and environmental genomics In recent years, one overarching objective of many has been to address fundamental questions concerning experimental design and the robustness of data collected under the broad umbrella of environmental genomics. One lesson learned from intergenomics studies is that there are core molecular networks that can be identified by multiple laboratories using the same platform. This supports the idea that “omics networks” defined a priori may be a viable approach moving forward for evaluating environmental impacts over time. Both spatial and temporal variability in ecosystem structure is expected to influence molecular responses to environmental stressors, and it is important to recognize how these variables, as well as individual factors (i.e. sex, age and maturation), may confound interpretation of network’s responses to chemicals. As AOPs become more defined and their potential in environmental monitoring assessments becomes more widely recognized, the AOP framework may prove to be the conduit between the molecular dimension and more complex biological scales such as individual organisms, or populations to improve environmental risk assessments (Martyniuk 2018). 6.8. Conclusion These new tools may make it possible to link genotypes and phenotypes, conduct interspecies comparisons, predict the structures and functions of genes, reconstruct regulation networks, and use genotypes and phenotypes to

202

Ecotoxicology

predict the toxicity of a xenobiotic, all at the cellular level up through the population level. A thorough understanding of these new tools in all their dimensions, in vitro and in silico, will provide useful data to solve environmental problems in several domains such as climate change; the assessment and management of environmental risks; the effects of anthropic disturbances on organisms, populations, communities and the functioning of ecosystems; erosion of biodiversity; the prevention and treatment of pollution; the management of aquatic resources; and the protection of natural environments. This knowledge will require sustained collaboration between disciplines that will be required to drop their delimitation, from ecology to genomics to biochemistry, biology and bioinformatics. These new approaches could therefore provide a modernized framework to advance the predictive capacity of the environmental and ecotoxicological sciences. 6.9. References Ankley, G.T., Bennett, R.S., Erickson, R.J., Hoff, D.J., Hornung, M.W., Johnson, R.D., Mount, D.R., Nichols, J.W., Russom, C.L., Schmieder, P.K., Serrrano, J.A., Tietge, J.E., Villeneuve, D.L. (2010). Adverse outcome pathways: A conceptual framework to support ecotoxicology research and risk assessment. Environ Toxicol Chem, 29(3), 730–741. Ankley, G.T., Johnson, R.D. (2004). Small fish models for identifying and assessing the effects of endocrine-disrupting chemicals. Ilar Journal, 45(4), 469–483. Atwood, T.K., Bongcam-Rudloff, E., Brazas, M.E., Corpas, M., Gaudet, P., Lewitter, F., Mulder, N., Palagi, P.M., Schneider, M.V., Van Gelder, C.W. (2015). GOBLET: The global organisation for bioinformatics learning, education and training. PLoS Computational Biology, 11(4), e1004143. Bahamonde, P.A., Feswick, A., Isaacs, M.A., Munkittrick, K.R., Martyniuk, C.J. (2016). Defining the role of omics in assessing ecosystem health: Perspectives from the Canadian environmental monitoring program. Environ Toxicol Chem, 35(1), 20–35. Berger, B., Daniels, N.M., Yu, Y.W. (2016). Computational biology in the 21st century: Scaling with compressive algorithms. Communications of the ACM, 59(8), 72–80.

New Challenges: Omics Technologies in Ecotoxicology

203

Bertin, D., Labadie, P., Ferrari, B.J., Sapin, A., Garric, J., Geffard, O., Budzinski, H., Babut, M. (2016). Potential exposure routes and accumulation kinetics for poly- and perfluorinated alkyl compounds for a freshwater amphipod: Gammarus spp. (Crustacea). Chemosphere, 155, 380–387. Beyer, J., Green, N.W., Brooks, S., Allan, I.J., Ruus, A., Gomes, T., Bråte, I.L.N., Schøyen, M. (2017). Blue mussels (Mytilus edulis spp.) as sentinel organisms in coastal pollution monitoring: A review. Marine Environmental Research, 130, 338–365. Beyer, J., Petersen, K., Song, Y., Ruus, A., Grung, M., Bakke, T., Tollefsen, K.E. (2014). Environmental risk assessment of combined effects in aquatic ecotoxicology: A discussion paper. Marine Environmental Research, 96, 81–91. Bradbury, S.P., Russom, C.L., Ankley, G.T., Schultz, T.W., Walker, J.D. (2003). Overview of data and conceptual approaches for derivation of quantitative structure – activity relationships for ecotoxicological effects of organic chemicals. Environmental Toxicology and Chemistry, 22(8), 1789–1798. Brennwald, M.S., Schmidt, M., Oser, J., Kipfer, R. (2016). A portable and autonomous mass spectrometric system for on-site environmental gas analysis. Environ Sci Technol, 50(24), 13455–13463. Burggren, W.W. (2014). Epigenetics as a source of variation in comparative animal physiology - or - Lamarck is lookin’ pretty good these days. J Exp Biol, 217(Pt 5), 682–689. Calisi, A., Latino, M.E., Corallo, A., Grimaldi, A., Ferronato, C., Antisari, L.V., Dondero, F. (2019). Biomarkers in Soil Organisms. Bioremediation of Agricultural Soils, 262. Calzolai, L., Ansorge, W., Calabrese, E., Denslow, N., Part, P., Lettieri, T. (2007). Transcriptomics and proteomics. Applications to ecotoxicology. Comparative Biochemistry and Physiology Part D: Genomics and Proteomics, 2(3), 245–249. Canovas, A., Reverter, A., Deatley, K.L., Ashley, R.L., Colgrave, M.L., Fortes, M.R. (2014). Multi-tissue omics analyses reveal molecular regulatory networks for puberty in composite beef cattle. PLoS One, 9. Caruso, J.A., Stemmer, P.M., Dombkowski, A., Caruthers, N.J., Gill, R., Rosenspire, A.J. (2014). A systems toxicology approach identifies Lyn as a key signaling phosphoprotein modulated by mercury in a B lymphocyte cell model. Toxicol Appl Pharmacol, 276(1), 47–54.

204

Ecotoxicology

Chen, X., Li, H., Zhang, J., Ding, Y., You, J. (2016). Does cadmium affect the toxicokinetics of permethrin in Chironomus dilutus at sublethal level? Evidence of enzymatic activity and gene expression. Environmental Pollution, 218, 1005–1013. Connell, D.W., Yu, Q.J., Verma, V. (2016). Influence of exposure time on toxicity—An overview. Toxicology, 355–356, 49–53. Escher, B.I., Hackermüller, J., Polte, T., Scholz, S., Aigner, A., Altenburger, R., Böhme, A., Bopp, S.K., Brack, W., Busch, W., Chadeau-Hyam, M., Covaci, A., Eisenträger, A., Galligan, J.J., Garcia-Reyero, N., Hartung, T., Hein, M., Herberth, G., Jahnke, A., Kleinjans, J., Klüver, N., Krauss, M., Lamoree, M., Lehmann, I., Luckenbach, T., Miller, G.W., Müller, A., Phillips, D.H., Reemtsma, T., Rolle-Kampczyk, U., Schüürmann, G., Schwikowski, B., Tan, Y.M., Trump, S., Walter-Rohde, S., Wambaugh, J.F. (2017). From the exposome to mechanistic understanding of chemical-induced adverse effects. Environment International, 99, 97–106. Fahd, F., Khan, F., Veitch, B., Yang, M. (2017). Aquatic ecotoxicological models and their applicability in Arctic regions. Marine Pollution Bulletin, 120(1), 428–437. Faure, D., Joly, D. (2015). Next-generation sequencing as a powerful motor for advances in the biological and environmental sciences. Genetica, 143(2), 129–132. Fischer, B.B., Pomati, F., Eggen, R.I.L. (2013). The toxicity of chemical pollutants in dynamic natural systems: The challenge of integrating environmental factors and biological complexity. Science of The Total Environment, 449, 253–259. Giraudo, M., Douville, M., Houde, M. (2015). Chronic toxicity evaluation of the flame retardant tris (2-butoxyethyl) phosphate (TBOEP) using Daphnia magna transcriptomic response. Chemosphere, 132, 159–165. Gong, P., Perkins, E.J. (2016). Earthworm toxicogenomics: A renewed genomewide quest for novel biomarkers and mechanistic insights. Applied Soil Ecology, 104, 12–24. Gouveia, D., Chaumot, A., Charnot, A., Almunia, C., Francois, A., Navarro, L., Armengaud, J., Salvador, A., Geffard, O. (2017). Ecotoxico-proteomics for aquatic environmental monitoring: First in situ application of a new proteomicsbased multibiomarker assay using caged amphipods. Environ Sci Technol, 51(22), 13417–13426.

New Challenges: Omics Technologies in Ecotoxicology

205

Groh, K.J., Carvalho, R.N., Chipman, J.K., Denslow, N.D., Halder, M., Murphy, C.A., Roelofs, D., Rolaki, A., Schirmer, K., Watanabe, K.H. (2015). Development and application of the adverse outcome pathway framework for understanding and predicting chronic toxicity: I. Challenges and research needs in ecotoxicology. Chemosphere, 120, 764–777. Haber, F. (1924). Fünf Vorträge aus den jahren 1920–1923. Verlag von Julius Springer, 76–92. Hogeweg, P. (2011). The roots of bioinformatics in theoretical biology. PLoS Comput Biol, 7(3), e1002021. Kamstra, J.H., Alestrom, P., Kooter, J.M., Legler, J. (2015). Zebrafish as a model to study the role of DNA methylation in environmental toxicology. Environ Sci Pollut Res Int, 22(21), 16262–16276. Kim, H.J., Koedrith, P., Seo, Y.R. (2015). Ecotoxicogenomic approaches for understanding molecular mechanisms of environmental chemical toxicity using aquatic invertebrate, Daphnia model organism. Int. J. Mol. Sci., 16(6), 12261–12287. Kramer, V.J., Etterson, M.A., Hecker, M., Murphy, C.A., Roesijadi, G., Spade, D.J., Spromberg, J.A., Wang, M., Ankley, G.T. (2011). Adverse outcome pathways and ecological risk assessment: Bridging to population‐level effects. Environmental Toxicology and Chemistry, 30(1), 64–76. Kunz, P.Y., Kienle, C., Gerhardt, A. (2010). Gammarus spp. in aquatic ecotoxicology and water quality assessment: toward integrated multilevel tests. Rev Environ Contam Toxicol, 205, 1–76. Laporte, M., Pavey, S.A., Rougeux, C., Pierron, F., Lauzent, M., Budzinski, H., Labadie, P., Geneste, E., Couture, P., Baudrimont, M., Bernatchez, L. (2016). RAD sequencing reveals within-generation polygenic selection in response to anthropogenic organic and metal contamination in North Atlantic Eels. Mol Ecol., 25(1), 219–237. Lee, S. (2013). Post-translational modification of proteins in toxicological research: focus on lysine acylation. Toxicol Res, 29(2), 81–86. Lewin, H.A., Robinson, G.E., Kress, W.J., Baker, W.J., Coddington, J., Crandall, K.A., Durbin, R., Edwards, S.V., Forest, F., Gilbert, M.T.P., Goldstein, M.M., Grigoriev, I.V., Hackett, K.J., Haussler, D., Jarvis, E.D., Johnson, W.E., Patrinos, A., Richards, S., Castilla-Rubio, J.C., Van Sluys, M.A., Soltis, P.S., Xu, X., Yang, H., Zhang, G. (2018). Earth biogenome project: Sequencing life for the future of life. Proc Natl Acad Sci U S A, 115(17), 4325–4333.

206

Ecotoxicology

Loh, P.R., Baym, M., Berger, B. (2012). Compressive genomics. Nat Biotechnol, 30(7), 627–630. Martyniuk, C.J. (2018). Are we closer to the vision? A proposed framework for incorporating omics into environmental assessments. Environmental Toxicology and Pharmacology, 59, 87–93. Mav, D., Shah, R.R., Howard, B.E., Auerbach, S.S., Bushel, P.R., Collins, J.B., Gerhold, D.L., Judson, R.S., Karmaus, A.L., Maull, E.A., Mendrick, D.L., Merrick, B.A., Sipes, N.S., Svoboda, D., Paules, R.S. (2018). A hybrid gene selection approach to create the S1500+ targeted gene sets for use in highthroughput transcriptomics. PLoS One, 13(2), e0191105. Murgarella, M., Puiu, D., Novoa, B., Figueras, A., Posada, D., Canchaya, C. (2016). A first insight into the genome of the filter-feeder mussel Mytilus galloprovincialis. PLoS One, 11(3), e0151561. Murphy, C.A., Nisbet, R.M., Antczak, P., Garcia-Reyero, N., Gergs, A., Lika, K., Mathews, T., Muller, E.B., Nacci, D., Peace, A., Remien, C.H., Schultz, I.R., Watanabe, K.H. (2018). Linking adverse outcome pathways to dynamic energy budgets: A conceptual model. In A Systems Biology Approach to Advancing Adverse Outcome Pathways for Risk Assessment, Garcia-Reyero, N. Murphy, C.A. (eds). Springer International Publishing, Cham, 281–302. Nesatyy, V.J., Suter, M.J. (2008). Analysis of environmental stress response on the proteome level. Mass Spectrom Rev, 27(6), 556–574. Perkins, E.J., Chipman, J.K., Edwards, S., Habib, T., Falciani, F., Taylor, R., Van Aggelen, G., Vulpe, C., Antczak, P., Loguinov, A. (2011). Reverse engineering adverse outcome pathways. Environmental Toxicology and Chemistry, 30(1), 22–38. Pisani, C., Gaillard, J.C., Nouvel, V., Odorico, M., Armengaud, J., Prat, O. (2015). High-throughput, quantitative assessment of the effects of low-dose silica nanoparticles on lung cells: grasping complex toxicity with a great depth of field. BMC Genomics, 16, 315. Pisani, C., Voisin, S., Arafah, K., Durand, P., Perrard, M.-H., Guichaoua, M.-R., Bulet, P., Prat, O. (2016). Ex vivo assessment of testicular toxicity induced by carbendazim and iprodione, alone or in a mixture. ALTEX-Alternatives to Animal Experimentation, 33(4), 393–413. Pomerantz, A., Penafiel, N., Arteaga, A., Bustamante, L., Pichardo, F., Coloma, L.A., Barrio-Amoros, C.L., Salazar-Valenzuela, D., Prost, S. (2018). Real-time DNA barcoding in a rainforest using nanopore sequencing: Opportunities for rapid biodiversity assessments and local capacity building. Gigascience, 7(4).

New Challenges: Omics Technologies in Ecotoxicology

207

Poynton, H.C., Lazorchak, J.M., Impellitteri, C.A., Blalock, B.J., Rogers, K., Allen, H.J., Loguinov, A., Heckman, J.L., Govindasmawy, S. (2012). Toxicogenomic responses of nanotoxicity in Daphnia magna exposed to silver nitrate and coated silver nanoparticles. Environmental Science & Technology, 46(11), 6288–6296. Raj, A., Peskin, C.S., Tranchina, D., Vargas, D.Y., Tyagi, S. (2006). Stochastic mRNA synthesis in mammalian cells. PLoS Biol, 4(10), e309. Rhee, J.S., Lee, J.S. (2014). Whole genome data for omics-based research on the self-fertilizing fish Kryptolebias marmoratus. Mar Pollut Bull, 85(2), 532–541. Sakuratani, Y., Horie, M., Leinala, E. (2018). Integrated approaches to testing and assessment: OECD activities on the development and use of adverse outcome pathways and case studies. Basic & Clinical Pharmacology & Toxicology, 123, 20–28. Scholz, S., Fischer, S., Gundel, U., Kuster, E., Luckenbach, T., Voelker, D. (2008). The zebrafish embryo model in environmental risk assessment–applications beyond acute toxicity testing. Environ Sci Pollut Res Int, 15(5), 394–404. Smith, D.R. (2015). Broadening the definition of a bioinformatician. Front Genet, 6, 258. Snape, J.R., Maund, S.J., Pickford, D.B., Hutchinson, T.H. (2004). Ecotoxicogenomics: The challenge of integrating genomics into aquatic and terrestrial ecotoxicology. Aquatic Toxicology, 67(2), 143–154. Todd, E.V., Black, M.A., Gemmell, N.J. (2016). The power and promise of RNAseq in ecology and evolution. Mol Ecol, 25(6), 1224–1241. Trapp, J., Almunia, C., Gaillard, J.C., Pible, O., Chaumot, A., Geffard, O., Armengaud, J. (2016a). Proteogenomic insights into the core-proteome of female reproductive tissues from crustacean amphipods. J Proteomics, 135, 51–61. Trapp, J., Armengaud, J., Gaillard, J.-C., Pible, O., Chaumot, A., Geffard, O. (2016b). High-throughput proteome dynamics for discovery of key proteins in sentinel species: Unsuspected vitellogenins diversity in the crustacean Gammarus fossarum. Journal of Proteomics, 146, 207–214. Trapp, J., Armengaud, J., Pible, O., Gaillard, J.-C., Abbaci, K., Habtoul, Y., Chaumot, A., Geffard, O. (2015). Proteomic investigation of male gammarus fossarum, a freshwater crustacean, in response to endocrine disruptors. Journal of Proteome Research, 14(1), 292–303. Trapp, J., Armengaud, J., Salvador, A., Chaumot, A., Geffard, O. (2014). Nextgeneration proteomics: Toward customized biomarkers for environmental biomonitoring. Environ Sci Technol, 48(23), 13560–13572.

208

Ecotoxicology

Vandegehuchte, M.B., Janssen, C.R. (2011). Epigenetics and its implications for ecotoxicology. Ecotoxicology, 20(3), 607–624. Vandegehuchte, M.B., Janssen, C.R. (2014). Epigenetics in an ecotoxicological context. Mutation Research/Genetic Toxicology and Environmental Mutagenesis, 764–765, 36–45. Villeneuve, D.L., Crump, D., Garcia-Reyero, N., Hecker, M., Hutchinson, T.H., Lalone, C.A., Landesmann, B., Lettieri, T., Munn, S., Nepelska, M. (2014a). Adverse outcome pathway (AOP) development I: Strategies and principles. Toxicological Sciences, 142(2), 312–320. Villeneuve, D.L., Crump, D., Garcia-Reyero, N., Hecker, M., Hutchinson, T.H., Lalone, C.A., Landesmann, B., Lettieri, T., Munn, S., Nepelska, M. (2014b). Adverse outcome pathway development II: Best practices. Toxicological Sciences, 142(2), 321–330. Villeneuve, D.L., Garcia-Reyero, N. (2011). Vision & strategy: Predictive ecotoxicology in the 21st century. Environmental Toxicology and Chemistry, 30(1), 1–8. Vincent, A.T., Charette, S.J. (2015). Who qualifies to be a bioinformatician? Frontiers in Genetics, 6, 164. Wang, W.-X., Meng, J., Weng, N. (2018). Trace metals in oysters: Molecular and cellular mechanisms and ecotoxicological impacts. Environmental Science: Processes & Impacts. Wang, Z., Gerstein, M., Snyder, M. (2009). RNA-Seq: A revolutionary tool for transcriptomics. Nat Rev Genet, 10(1), 57–63. Watanabe, K.H., Andersen, M.E., Basu, N., Carvan, M.J., Crofton, K.M., King, K.A., Suñol, C., Tiffany‐Castiglioni, E., Schultz, I.R. (2011). Defining and modeling known adverse outcome pathways: Domoic acid and neuronal signaling as a case study. Environmental Toxicology and Chemistry, 30(1), 9–21. Waters, M.D., Fostel, J.M. (2004). Toxicogenomics and systems toxicology: Aims and prospects. Nat Rev Genet, 5(12), 936–948. Willett, K.L. (2018). Considering epigenetics in adverse outcome pathways. A Systems Biology Approach to Advancing Adverse Outcome Pathways for Risk Assessment, Garcia-Reyero, N. Murphy, C.A. (eds). Springer International Publishing, Cham, 219–234.