A Systems Biology Platform for Predictive Ecotoxicology in Daphnia magna

Lead Research Organisation: University of Liverpool
Department Name: Institute of Integrative Biology

Abstract

The monitoring of the environment for adverse effects of chemical pollution is of paramount importance in maintaining biodiversity and environmental health. This is particularly important for the aquatic environment into which a wide range of pollutants find their way, for example from pesticide run-off, industrial spills and excess nutrients from the release of untreated sewage. Pollution remains a major problem in the UK, with the Environment Agency estimating that up to 82% of rivers are 'at risk' from chemicals such as nitrate and phosphorus. Often adverse impacts to these pollutants can only be identified when they are sufficiently severe so as to affect survival of organisms and thus are identifiable at a late stage, after the damage is done. Attempts have been made to use more sensitive molecular indicators of early change ("biomarkers") but these generally inform only on the levels of chemical exposure and have limited diagnostic or predictive power in relation to toxicity. Therefore the application of molecular biomarkers to environmental monitoring has been very limited to date.
Inspired by the tremendous success of recent technologies in biomedicine, we propose to develop an equivalent system for biomarker discovery in an environmental context, i.e., to develop biomarkers for application to environmental monitoring and diagnostics. These technologies can measure many thousands of biochemicals (including gene products and metabolites) in exposed organisms and by using mathematical and computational tools we can identify the underlying pathways to toxicity. These computational models will enable us to discover a set of genes and metabolites that can be highly predictive of an adverse impact on living organisms and at the same time provide a characteristic fingerprint of the type of pollutant class(es) responsible for such impact. We will study these effects in the water flea (Daphnia magna) which is already commonly used in the testing of contaminated water samples for toxicity. A wide range of chemicals representing major pollutant classes will be assessed and the molecular signatures will be compared to physiological responses in the water fleas. This will allow us to discover molecular signatures that have high diagnostic value in an ecological context, specifically telling us about the health and reproductive fitness of the water fleas. Also, the computational methods will allow us to discover molecular signatures that causally relate to the water fleas' health. This represents a major advance over current molecular biomarkers.
Once these characteristic fingerprints of molecules that are predictive of different toxicities are established, we will test such fingerprints to be predictive of the chemical makeup of water samples taken from polluted environments and with proven environmental impact. Such sampling will be in collaboration with our project partner, the Environment Agency. We will be "blinded" to the nature of these samples, enabling a robust evaluation of our ability to (1) determine the "ecological status" of the water and (2) diagnose the underlying pollutant class, thus enhancing the regulators' ability to target remedial measures. Following this validation of the new predictive biomarkers we will convert them into simple, rapid and economic assays, resulting in the provision of a new generation of environmental monitoring tools. After the project, and through our collaborations with end-users in the UK, Europe and North America, we will seek to pilot our molecular biomarkers alongside conventional biological and chemical monitoring, e.g. as part of the Water Framework Directive. In summary, this exciting project is interdisciplinary, involving fundamental biochemistry and physiology, toxicology, molecular biology and bioinformatics, and promises a significant advance in the tools available to monitor the health of our environment.

Planned Impact

The impact of this work relates to both the academic community and to a range of end users with interests in the maintenance of a healthy environment. Regarding the former, the provision of the first comprehensive integrated "molecular-physiology effects" dataset representing the response of an organism of relevance for environmental and toxicological assessment of chemicals will be an invaluable resource for academic researchers in ecotoxicology. The academic community within toxicology will benefit specifically from improved understanding of mechanisms of toxicity and by gaining a crucial knowledge for progressing the safety of chemicals such as pesticides and drugs. The pipeline we will develop and validate will represent an important reference for establishing a gold standard for 2nd generation biomarker discovery in the NERC and environmental biology international communities. Furthermore, the methodologies to be developed will help the progression of systems biology as an academic discipline well beyond the environmental biology area. The research will play a role in enhancing the integration of the disciplines of mathematics, computer science and biology. The early career researchers on the project will also benefit from expertise developed within a multidisciplinary research environment including interaction with end-users thus equipping them with skills appropriate for a future career at the interface between academia and industry or regulators in an environmental context.
The primary objective of this project is to develop, within the 3-year period, a proof-of-principle diagnostic approach for predictive ecotoxicology that is anticipated to have considerable value, in the future, within the context of the EU Water Framework Directive. Upon successfully demonstrating the value of this "2nd generation biomarker discovery" approach we plan to develop this into a practical and implementable tool for environmental monitoring and diagnostics. These tools will be directly relevant to "Good Environmental Status" (GES) defined as "Concentrations of contaminants at levels not giving rise to pollution effects". Hence the EA represents the major beneficiary within the UK. Other organisations involved in monitoring of the marine environment (Cefas and FRS within the UK and international organisations such as OECD, ICES and OSPAR will also benefit in the same way. Other organisations that would benefit from the novel diagnostic tools developed here are those associated with ecological risk assessment of chemicals not just confined to the aquatic environment, including industry (e.g. pharmaceutical companies such as AstraZeneca, Brixham Environmental Laboratory) and non-governmental organisations (e.g. Cefic, representing the European chemical industry) and the EC Joint Research Centre (JRC) for sustainable development. The novel biomarker sets will also be of value to international organisations beyond the EU such as the US Environmental Protection Agency and Environment Canada.

Ultimately there will be benefits to the wider public through improved and efficient monitoring of the environment for pollutant impact and thus maintenance of diversity in the countryside and enhancement of recreational sports such as fishing. Overall, the implementation of more efficient, more predictive and high-throughput assessment of the impact of pollutants in the environment, that should be achieved within 3 years of completion of the project, will be of benefit to regulators and policy makers, aiding more effective management of our water resources and the prioritisation of sites requiring remediation and thus offering benefits relating to health and wealth to society.

Publications

10 25 50
 
Description The overall purpose of the study was to improve our understanding of toxicity for environmentally relevant chemicals and to identify potential biomarkers that may be used in environmental monitoring. More precisely, the project aimed at linking molecular signatures represented by genome wide transcriptional and metabolic response, to chemicals mechanism of action and organism toxicity in Daphnia magna.

Our finding represent a considerable increase in scientific knowledge and useful biomarkers that, once validated, may improve our ability to do water quality assessment.

Our work proves that we can indeed identify biologically interpretable transcriptional signatures predictive of chemical class [1]. This is an important result since the biomarkers we have identified are able to detect with a high degree of certainty the presence of important class of contaminants in the water without the need of targeted chemical analysis.

We succeeded in attracting further interest in these findings from an important UK company in the water sector. These findings have in fact attracted the interest of Thames Water (http://www.thameswater.co.uk) that have made available a small amount of funding (£45K) for a proof of concept study demonstrating the applicability of our system to monitor their water purification plants. This work, which is a direct output of this project grant is currently under way.

Moreover, a more in depth analysis of the expression profiling data mentioned above did show that alterations in the intracellular concentration of calcium may be the molecular initiating event underlying basal toxicity in this important organism [2]. We have also generated data representing feeding and respiration rates in Daphnia control and exposed animals. We have developed statistical models linking molecular response to organism physiology and developed an approach to predict the effects of mixture exposures in laboratory conditions as well as using environmental exposures from 2 rivers in UK.

We have also developed models that link molecular networks to specific aspects of Daphnia physiology (oxygen consumption and feeding rates). This has helped us to identify putative biomarkers that can be used to monitor sub-lethal effects of pollutants with potential effects on the ecosystem, thus improving our monitoring capabilities.

Using a similar strategy we have successfully attracted the interest of Unilever (http://www.unilever.co.uk) that has financed a project aiming at expanding these findings to two additional species with the purpose of demonstrating that calcium alterations may be the general mechanism for basal toxicity. This project (£400K) involved both Prof. Falciani (PI in this grant) and Prof. Mark Viant (Co-I in this grant) and it is just been extended for another year with an additional £100K.

References

[1] Antczak P, Jo HJ, Woo S, Scanlan L, Poynton H, Loguinov A, Chan S, Falciani F*, Vulpe C*. Molecular toxicity identification evaluation (mTIE) approach predicts chemical exposure in Daphnia magna. Environ Sci Technol. (2013) 15;47(20):11747-56. (* joint senior and correspondent authors)

[2] Philipp Antczak, Thomas A. White, Anirudha Giri, Mark T. D. Cronin, Francesco Michelangeli, Mark R. Viant, Chris Vulpe and Francesco Falciani. A systems biology approach reveals a novel calcium- dependent mechanism for basal toxicity in Daphnia magna. Under revision in Environ Sci Technol.
Exploitation Route The primary objective of this project was to develop a proof-of-principle diagnostic approach for predictive ecotoxicology. The framework we have developed allows the identification of a chemical toxicity pathway from large scale functional genomics datasets when no other knowledge on a chemical MOA is available. With DEFRA we have then expanded the framework to identify interactions between chemicals in a mixture (6+ Papers in preparation). We actively promoted the development of this approach by organizing conference sessions, workshops and by coordinating international working groups with academics and stakeholders. Many of these have led to opinion papers that have been published in top peer reviewed journals in the field. These papers outline the way that our research can be used to promote impact a several levels.
The first level is at the stage of development of a Systems based risk assessment approach within the OECD and other similar organisations. A great example of this is a new Adverse Outcome Pathway (AOP) for narcosis that is currently an integral part of the recommended risk assessment procedure for this very important class of chemicals.
In the Industrial setting, AstraZeneca is developing biosensor technology to simultaneously predict chemical toxicity, phenotype outcomes, and the molecular basis for these changes. We are also aware that Unilever are keen to establish our pipelines in their own workflows to enable a more comprehensive analysis suite of their products.
Sectors Chemicals

Environment

 
Description Since the initial phase of the project we have already proven 'the demonstrable contribution that excellent research makes to society and the economy'. More specifically, we have shown that the diagnostic tools and the technology that we have developed (computational biology strategy) have the potential to transform R&D in the water and home product industry. The research contracts with Thames Water and Unilever mentioned in the first section of this document are the direct outcome of the science developed in this project. In the first case, the biomarker discovery approach has been used to identify residual chemical contamination in water samples from pilot water reprocessing plants in UK. In the second case the computational framework developed in this project has been applied to additional species of interest to Unilever. Interestingly we have just been granted a one year extension of the Unilever project, which allow us to further progress in the application of the findings. This is an extremely important example of how advance last generation biomarkers can really be developed in industry settings to improve our ability to estimate health and environmental risk of chemical contamination. Such achievement, especially considering the financial commitment of the companies involved in supporting research in our group is not very common in this area. Therefore our achievements have great transformative value. We also held an international workshop exploring the impact of our research in risk assessment. This was held at the University of Liverpool (home institution of Prof. Falciani, PI in this grant) in conjunction of the second iEOS conference, sponsored by the NERC (http://environmentalomics.org/ieos2014/). The outcome of the workshop will be shortly published as an opinion paper, which is currently in preparation. The ideas developed in this workshop are in line with the efforts of the OECD in developing recommendations to use complex approaches for the definition of a comprehensive framework for organizing knowledge and support risk assessment. This framework has been originally defined by the U.S. Environmental Protection Agency and its further development has the potential to change the way we perform risk assessment. This workshop has been co-chaired by one of the leading scientists in the U.S. EPA (Prof. Gary Anckley) a Unilever scientist (Dr. Geoff Hodges) and by Prof. Falciani (PI of this grant). We have then recently organised an NERC/STFC funded workshop expanding the concepts of Systems Ecotoxicology beyond single organisms and discussed the application of modelling approaches that include populations and ecosystem.e
First Year Of Impact 2013
Sector Chemicals,Environment
Impact Types Economic

 
Description Industry Funding Thames Water
Amount £45,000 (GBP)
Organisation Thames Water Utilities Limited 
Sector Private
Country United Kingdom
Start 08/2014 
End 08/2015
 
Description Industry Funding Unilever
Amount £380,000 (GBP)
Organisation Unilever 
Department Unilever Research and Development
Sector Private
Country United Kingdom
Start 05/2014 
End 05/2016
 
Description Unilever Research Contract
Amount £90,000 (GBP)
Organisation Unilever 
Department Unilever Research and Development
Sector Private
Country United Kingdom
Start 03/2016 
End 03/2018
 
Title RiskAssesmentTool01 
Description The science developed with this award allowed us to secure further funding from Unilever for the development of an environmental risk assessment tool based on mRNA expression profiling (see further funding section). The overarching objective of this approach is to be able to rapidly assess the mechanism of action of toxicity and possible toxicity from a gene expression or metabolomics readout following a short-term exposure. The hart of the method is a computational model that integrate gene expression data with chemical structure descriptors to predict chemical class (polar or non-polar narcotic, non-specific reactivity, etc.) and to provide a possible biological effect, for example the inhibition of acetylcholine signalling caused by a narcotic. More precisely, the method uses a statistical model predictive of chemical class. Briefly, a genetic algorithm selects the most informative combination of gene expression profiles while a statistical engine based on decision tree approach (random forest) assign each compound to a given class and outcome. Since the pipeline is implemented in the statistical environment R, it is widely applicable and prone to be interfaced by a web-based front end. The method has been tested on two datasets. The first, developed by our group, represents the transcriptional response of C. elegans to a short term sub-lethal exposure as a readout. This dataset is relatively small (50 chemicals and encompasses polar- and non-polar narcotics and acetylcholine esterase inhibitors). The second, dataset we have used is publicly available and has been developed as an NIH Common Fund program. It represents how human cells globally respond to chemical, genetic, and disease perturbations. Currently, it includes about 19.000 different compounds. So far, we have successfully experimentally validated some of the predictions. Unilever has expressed a strong interest in these results and asked us to deploy our pipeline in a usable software tool for internal use. At present we have a prototype. Deployment is expected in 5-month time. 
Type Of Material Computer model/algorithm 
Year Produced 2017 
Provided To Others? No  
Impact Unilever is also undergoing an assessment of the potential impact of such tool in their manufacturing process. The expectation is that it would be responsible for significant cost savings at the risk assessment stage. 
 
Company Name Omic Analytics Limited 
Description  
Year Established 2014 
Impact The company has developed software supporting Mass Spectrometry instrumentation. This is currently licensed to the world leading company Waters (http://www.waters.com/waters/home.htm?locale=en_GB) and supports their instrumentation. The company provides contract research support to several other companies. The know-how developed in the associated grants has allowed the company to deliver research contracts in the area of Systems Biology.
Website http://www.omicanalytics.com
 
Description Conference Atheneum 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Public/other audiences
Results and Impact I presented a historical perspective on our understanding of chemical toxicity, starting from ancient Egypt to modern day predictive toxicology. The audience (approximatively 50 people) were very engaged, particularly in respect to the ethical implications of using animals in toxicology testing. At the ed of my presentation I summarised the results of our research, including the project supported by this grant. This was very well received and after the talk I received a considerable amount of feedback.

The Athenaeum is a private members club in Liverpool, England. It has been in continuous use since the end of the 18th century. The club was founded to ensure the up-to-date provision of newspapers and pamphlets, and to create a library for the use of the merchants and professional men in the city. The Athenaeum contains a large library, and it is also used by the Proprietors for social functions. It can be hired for use by outside individuals and organisations.
Year(s) Of Engagement Activity 2017
 
Description Stakeholder/Scientists interaction Workshop 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Policymakers/politicians
Results and Impact We also held an international workshop exploring the impact of our research in risk assessment. This was held at the University of Liverpool (home institution of Prof. Falciani, PI in this grant) in conjunction of the second iEOS conference, sponsored by the NERC (http://environmentalomics.org/ieos2014/).

The outcome of the workshop will be published in a opinion paper, which is currently in preparation. The ideas developed in this workshop are in line with the efforts of the OECD in developing recommendations to use complex approaches for the definition of a comprehensive framework for organizing knowledge and support risk assessment. This framework has been originally defined by the U.S. Environmental Protection Agency and its further development has the potential to change the way we perform risk assessment. This workshop has been co-chaired by one of the leading scientists in the U.S. EPA (Prof. Gary Anckley) a Unilever scientist (Dr. Geoff Hodges) and by Prof. Falciani (PI of this grant).
Year(s) Of Engagement Activity 2014
URL http://environmentalomics.org/ieos2014-workshops/