Technologies to transform toxicity testing in the 21 ...

2015

Banaras Hindu University, Varanasi

Technologies to transform toxicity testing in the 21 =century: In silico toxicology and its Battery of Methods. Bagavathy S. Karthikeyan, Subbiah Parthasarathy and Mohammad A. Akbarsha. Abstract: The traditional toxicity testing practiced in research and drug discovery platforms are archaic, which is 50 to 80 year old, and depends solely on results from animals. From the rapid advancements of 21st century science and research it is realized that human and animals are not same. They are different in physiology and they metabolize drugs/chemicals differently. It is evident that there are differences within species and strains too. The attrition rates and the number of drugs withdrawn from the market have bearing on the species difference. So, there is a pertinent need for a paradigm shift from traditional animal-based toxicity testing to modern rationalized toxicity testing incorporating in silico tools and methods, in vitro toxicology techniques, stem cell applications, various OMICS approaches, the variety of imaging technologies and high throughput screening methods. Due to the implication of 3R's concept worldwide, the researchers and drug discovery industry are in a dire need to modernize and rationalize the testing strategies to produce safer drugs and introduce animal-free testing methods at less time and low cost. Key words: In silico, technology, toxicity.

Toxicology and toxicity end points: The study of the adverse effects of chemical, physical or biological agents on living organisms and the ecosystem is called toxicology (Hayes and Loomis, 1996). Different countries have different laws and regulations to control the marketing of drugs, vaccines, food additives, pesticides, industrial chemicals and other substances of potential toxicological concern. But most of the toxicologicalexperiments are done using animals rather than testing on humans due to the serious ethical concerns. Major toxicity endpoints considered include acute systemic toxicity, carcinogenicity, dermal penetration, ecotoxicity, eye irritation/corrosion, Animals and Alternatives

-.

in Life Science Research

I 91

2015

Banaras Hindu University, Varanasi

genotoxicity, neurotoxicity, pharmacokinetics & metabolism, repeated dose/organ toxicity, reproductive and developmental toxicity (teratogenicity), skin irritation/corrosion and skin sensitization (Mojca and Andrew, 2010). Problems with animal testing: Animal methods are time-consuming (requires 2 year rat bioassay for carcinogenicity), expensive (Jena et al, 2005), give erroneous results (generation of false-positives and false-negatives), and predictive credibility of 20-30% of adverse drug reactions which are not detected during preclinical trials. It does not have real-time situation (exposure condition and dosage vary greatly), loss of valuable chemical and lead substances for industry, and significantly adapting and responding to ethical, moral and practical issues in common (NRC,2007; Bhattacharya et al., 2011). Toxicity testing in the 21st century (TT21C): In 2004 the u.s. EPA and the National ToxicologyProgram of the u.s. National Institute of Environmental Health Sciences commissioned a Committee of the National Academy of Sciences to evaluate the current toxicity testing methods (viz., difficulty in extrapolation of high dose rodent effect to humans, attempts to incorporate pharmacokinetics aspects, and modes of action for certain high-value or high-liability compounds into the risk assessment process). The report "Toxicity Testing in the 21st Century: A Vision and a Strategy (T21Cr by the U.S. National Research Council (NRC,2007) laid out a bold vision for the future of toxicity testing of chemicals based on the explosive changes that have been and are occurring in the basic biological sciences. The vision of T21C includes faster and cheaper methods in risk assessment of more number of chemicals with fewer animals in use. It aims for deeper analysis of more doses, conditions, times and interactions with more mechanistic and fundamental methods. It emphasizes the use of human cells (in vitro) even at lower dose levels. T21C advocates the components of toxicity testing as in figure 1.

921 Animals and Alternatives

in Life Science Research

-.

Banaras Hindu University, Varanasi

2015

Cnemical Charactertzatlon

Toxicity Testing

Figure 1: Components of Toxicity testing in the 21st Century (Adapted from NRC, 2007).

TT21C envisions a paradigm shift from traditional toxicity testing to the emerging technologies. It aims to develop testing strategies to assess the potential health risks of large numbers of environmental agents. Also, it aims to exploit scientific advances in biology and toxicologyto achieve risk assessments that are more relevant to humans rather than depending on animals. The use of high-throughput technologies like OMICS approaches, in vitro methods (2 dimensional and 3 dimensional), Systems biology, in silica methods and invertebrate animal models embraces to potentially reduce the use of animals from toxicity procedures and could enhance the rate of production of safer drugs as well as strengthen testing strategies for chemicals from environment (NRC,2007; Bhattacharya et al., 2011). In silico toxicology: In the broadest sense in silica toxicologyis defined as "anything that we can do with a computer intoxicology"(Raunio, 2011). The United States Environmental Protection Agency (US EPA)defines in silica toxicologyas the "integration of modern computing and information technology with molecular biology to improve agency prioritization of data requirements and Animals and Alternatives

in Life Science Research

I 93

2015

Banaras Hindu University, Varanasi

risk assessment of chemicals" (US EPA, 2003). Many different types of in silico methods have been developed to characterize, risk assess or predict toxic outcomes in humans as well as environment. This review summarizes different methods and their applications and how toxicity testing procedures can be rationalized to reduce the animal use and, at the same time, scaleup drug discovery rates or toxicity prediction rates.

Relevance of in silico toxicology to 3R's concept: Computational or in silica or predictive toxicology is an interdisciplinary area, which embraces its applications from mathematics, chemical sciences, physics, and computational strategies. It enormously contributes to 3R's concept (Figure 2), especially, the 'Reduction' principle. The use of computational toxicology techniques substantially reduces the reliance on animals in the early phase of risk assessment of chemicals/drugs since it predicts adverse effects and so animal use. may be limited to potentially toxic chemicals as revealed in in silica analysis.

Reduction

~

In silico toxicology

Refinem Less painful procedures

Figure 2: Schematic diagram showing 3R's principles and contribution of in silico toxicology to it. 941 Animals and Alternatives

in Life Science Research

-.

Banaras Hindu University, Varanasi

2015

Advantages of in silica approaches over in vivo oranimal-based methods: In silica toxicology approaches consume less time, and are easy to use and adopt. It is economical (not like in viva which requires large amount of compounds/animals) and can be incorporated in any stage of pharmacology or toxicology research. The data generated from in silica experiments are interoperable, and have good correlation with in vitro data. In silica toxicologyapproaches greatly help in scrutinizing the problem or goal which significantly reduces the number of animals used during the early phase of the toxicologyexperiments.

In silico toxicology - battery of methods In silica toxicology methods are broadly classified into 'Reductionist' and 'Holistic' approaches which deal from molecular level to system level. ,

A. Reductionist approaches 1) Data driven systems/QSAR Relationship) method

(Quantitative

Structure-Activity

2) Expert Systems 3) Molecular modeling and docking methods

B. Holistic approaches 1) Pharmacokinetics and PBPK (Physiologically-BasedPharmacokinetic) modeling 2) Network Biologyand Network Pharmacology approach 3) Systems biology and Systems Toxicology approach response modeling) 4) Virtual modeling - Virtual tissues/organ.

A: Reductionist

(e.g., Dose-

approaches:

1) Data driven systems/QSAR Activity Relationship) method:

(Quantitative

Structure-

The baseline of this method is that 'function/ property / activity of the compound/molecule depends upon the structure of the molecule (known)'. Animals and Alternatives

in Life Science Research

I 95

2015

Banaras Hindu University, Varanasi

This method derives predictions from a training set of experimentally determined data available from the databases. It finds a mathematical relationship between a set of descriptors of molecules and their activity (Katritzky et al., 1995). The descriptors can be experimentally or computationally derived. The experimental data includes results derived from in vivo (mostly) and in vitro experiments (Figure 3). There are certain limitations of this method which include lack of relevant human data, results are often laboratory-specific (differences is dose, distribution characteristics, animal strains), and functions/activity can be predicted only for structures that exist within the training set (Helma, 2005).

eActivity/Toxicology eRules/Toxicophores

-Structure· Reaction mechanisms

QSAR

-Analytical

methods

Figure 3: Schematic diagram depicting QSAR and contribution of various fields to it.

This method predicts basic physiochemical properties of a compound such as molecular weight, polar surface area, logP, logD, number of hydrogen bond - donors and acceptors, etc. REACH (Registration, 961 Animals and Alternatives

in Life Science Research

2015

Banaras Hindu University, Varanasi

Evaluation, Authorization and Restriction of Chemicals) is a regulation of the European Union, adopted to improve the protection of human health and the environment from the risks that can be posed by chemicals, while enhancing the competitiveness of the EU chemicals industry. It had come up with a "Guidance on information requirements and chemical safety assessment" detailing on "QSARsand grouping of chemicals" which explains about validity, applicability and acceptance of QSARs,its reporting formats, case studies, etc. The coverage on link http://qsars-inreach.researchdissemination.eu/ summarize about QSARs in REACH Uses, issues and priorities. For example - The prediction of partition coefficient log P is an important measure used in identifying "drug-likeness" according to Lipinski's Rule of Five in Pharmacology. Also in toxicology, the 10gPgives information about solubility of the compound. Even for compounds available in the environment, the solubility is significant for its persistence. E.g. Pesticides. Databases - To retrieve chemical information: Pharmabase, Drug Bank, Duke's Phytochemical & Ethnobotanical Database, KEGG Ligand Database, PubChem Database. Tools: OSIRIS Property Explorer, OSRA, MODEL, LogP Calculator, MOLINSPIRATIONtool.

Expert Systems: This is the method which formalizes the knowledge of experts who assessed the toxicity of compounds in a computer program. Many of the most successful predictive software are in fact expert systems (Helma, 2005). By using these methods all possible endpoints ranging from mutagenicity, carcinogenicity, skin sensitization, irritancy, teratogenicity, hepatotoxicity, MTD -maximum tolerated dose, biodegradation, acute toxicity, LD50-lethal dose, environmental toxicity, immunotoxicity, neurotoxicity, endocrine disruption, cardio-toxicity, etc., can be predicted (Mojca and Andrew, 2010). These methods are intuitively appealing, easy to access topological knowledge prediction of biotransformation and metabolites, and provides specificityand integration of very diverse chemical and biological information, but requires accurate and regular update of database and at last it suffers from moderate sensitivity. Animals and Alternatives

in Life Science Research

I 97

Banaras Hindu University, Varanasi

2015

Software/Tools - LAZAR, OECD QSAR tool box, PASS, T.E.S.T, TOPKAT,DEREK, MultiCASE, etc. (Mojcaand Andrew, 2010).

2) Molecular modeling and docking methods: This method assesses the interaction of small molecules with biological macromolecules (predominately proteins), by fitting the ligand into the active site of the receptor. It is mainly used for pharmaceutical research, also applied for toxicologicalscreening purposes. It can be used to elucidate the mechanism and biotransformation and to predict receptor-mediated toxicity. The limitation of this method is, it can be used only for receptormediated mechanisms (e.g., Cytochrome P450, Estrogen receptor) where in the receptor structure is available; complex and partially unknown mechanisms are beyond its scope (Jacob et al., 2009; Raunio, 2011). The three dimensional structure of ligand and receptor (protein/DNA/ carbohydrate/lipid) is required to carry out this method. The structural targets can be determined by X-ray crystallography and Nuclear Magnetic Resonance methods. The already determined structures can be retrieved from Protein Data Bank (Archiveof experimentally determined 3D structures of biologicalmacromolecules). http://www.rcsb.org/pdb/home/ home.do The three dimensional structures of targets can be modeled using Homology Modeling -a Bioinformatics-based method. This method of docking relies on Lock and Key principle demonstrated by Hermann Emil Fischer in 1894 (Figure 4), where the lock represents the macro-molecule (Protein/Carbohydrate/Lipid) and the key represents the small molecule (Drug / Xenobiotic). Virtual screening methods are applied in drug designing to screen plethora of ligands from the various chemical databases towards risk assessment or designing a drug for therapeutic purpose. The binding/interaction efficiency can be ascertained by binding energy/interaction energy expressed as Kcal/rnol as well as formation of hydrogen bond and Van der Waal's interactions and its length is expressed in Angstrom unit (A).

981 Animals and Alternatives

in Life Science Research

Banaras Hindu University, Varanasi

2015

Figure 4: Representation of lock and key principle.

Software/Tools

- Autodock (Scripps), Glide (Schrodinger), Hex, Argus

Lab, Discovery Studio, Gold, FlexX.

B: Holistic approaches: Reductionism vs holism molecules to System:

in pharmacology/toxicology

-

From

The four centuries old biology believed that life resides inside the cells or sub-cellular components like genes/proteins. The fields ranging from cell biology, molecular biology to advanced OMICS technologies are examples of reductionism. The antithesis of reductionism is holism which believes that life resides in interactions and organization of biological componentsand describes function/behavior which arises from its emergent properties. In the recent times many complex diseases and real problems of biology ranging from cancer to diabetes are focused to this approach. Toxicologyis also a multi-faceted, multi-wired phenomenon which requires Animals and Alternatives

in Life Science Research

I 99

Banaras Hindu University, Varanasi

2015

diseases are regulated by complex biological networks rather than a single gene/protein or a single target.

Software/Tools - Cytoscape, VisANT, NeAT, Biological Networks, Pathway Studio, OSPREY,Pajek and Patika. 2)

Systems Biology and Systems Toxicology approach:

Systems Biology aims at understanding biological systems at system level. Biological systems are interconnected, and the study of the interactions between the components of biological systems is significant. Systems Biology allows one to study how these interactions give rise to the function and behavior of the system by mathematical/kinetic modeling and simulating the system with defined parameters (FigureS). Systems Biology, along with OMICS technologies, is applied in toxicological research to predict the fate of living systems when a chemical/drug is administered (Bhattacharya et al., 2011).

Systems Toxicology or S Figure 5: Schematic diagram depicting contribution of different fields to Systems Biology and Systems Toxicology.

Animals and Alternatives

in Life Science Research 1101

2015

Banaras Hindu University, Varanasi

holistic understanding of toxicological pathways and effect of chemicals on it.

Pharmacokinetics and Pharmacokinetic) modeling:

PBPK

(Physiologically-Based

In silico methods can be also be applied in predicting ADME properties of a drug in the form of Pharmacokinetics (PK) and PBPK modeling. PK focuses on what a drug does to the body or how a drug gets metabolized. PBPK modeling is a mathematical modeling technique for predicting the absorption, distribution, metabolism and excretion (ADME) of synthetic or natural chemical substances in humans and other animal species (Bouzom, 2012). It is used in pharmaceutical research and drug development, and in health risk assessment for cosmetics or general chemicals. It relies on chemical property prediction models (QSARmodels or predictive chemistry models) on one hand and they also extend into Systems Biologymodels of metabolic pathways.

Software/Tools Berkeley Madonna, ModelMaker. 1)

- BioDMET, PK-Sim, Simcyp Simulator, ADAPT 5, GNUMCSIM, GNU Octave, Matlab-

PottersWheel,

Network Biology and Network Pharmacology approach:

Networks representing any biological systems can be termed 'Network biology'. It enables understanding of various biological processes and their interactions such as protein-protein interaction, gene-gene interaction, metabolic pathways, biochemical pathways, etc. Network Biology involves the integration of biological processes, network theory or graph theory, and statistical approaches to analyze a biological network (Panagiotou, 2012). Network Pharmacology deals with network analysis of drug action. Networks can be used to understand the relationship between drugs, their targets and diseases. Here, the first step is the identification of appropriate targets for complex diseases (Panagiotou, 2012). The complex diseases include cancers, cardiovascular diseases, neurodegenerative diseases, etc. The reason behind need for network biology in pharmacology is that these 1001 Animals and Alternatives

in Life Science Research

Banaras Hindu University, Varanasi

2015

Databases/Models/Formats - Biomodels, CellML, Panther, Reactome, BioCyc, KEGG Pathway, BioChemWeb,BioPAX,SBML,BRENDA,LibSBML, MathML. Software/Tools

- Cell Designer, JDesigner, E-Cell, Gepasi, MATLAB, SBML

Toolbox, Systems Biology Workbench (SBW), Pathway Analyzer, JigCell.

Virtual modeling - Virtual tissues/organs It focuses on developing virtual tissues / organs in silico with advanced mathematical and computational methods to risk assess thousands of chemicals. Virtual tissues/organs aim to predict histopathological outcomes from alterations of cellular phenotypes that are controlled by chemical-induced perturbations in molecular pathways. These models/ agents mimic cell interactions and cellular responses to the microenvironment. The behavior of models is constrained by physical laws and biological rules derived from experimental evidence. There areIots of research groups focusing on developing virtual liver, virtual lung, etc., for toxicity testing and risk assessment worldwide (Shah, 2010). Virtual liver models from Virtual Liver Network, Germany (http://www.virtual-liver.de/). and Heptox® from Strand life sciences, India, are the virtual liver models to name a few that have been accomplished till date. Human Toxome Project: Human Toxome Project (HTP) involves mapping Pathways of Toxicity (PoT) by systems toxicology approach. HTP uses the most predictive, high-throughput human-derived cell-based assays to evaluate perturbations in key toxicity pathways, and to find the targets present in those pathways. This project provides a holistic approach for finding the accurate changes or damages that occur in a cell due exposure to the toxic substance. Although a number of toxicity pathways have already been identified, most are only partially known, and no common annotation exists (Hartung and McBride, 2011). Pathways of Toxicity (PoT) are the cellular pathways that may be perturbed by the chemicals such as xenobiotics which include drugs and environmental chemicals toxic to human beings. PoT may be divided into 1021 Animals and Alternatives

in Life Science Research

Banaras Hindu University, Varanasi

2015

two categories: receptor-mediated and non-receptor mediated (Figure 6). Receptor-mediated PoT involves the nuclear receptors that are involved in Phase-I, and Phase-II metabolizing enzymes (AHR, CAR/R.XR, PXR/R.XR, PPAR/R.XR) for detoxification of xenobiotics and other stress enzymes. Nonreceptor mediated PoT include genes of Phase-II metabolizing enzymes and other cellular defensive enzymes. The five crucial PoT are Wnt/ -catenin, the TGF-, the Notch, the Hedgehog, and the receptor kinase/ ras pathway that are involved in cancer (Hartung and McBride, 2011).

Xenobiotlcs: Drugs and Environ,""nt.1 chemkals

Non-Receptor Mediated

Receptor Mediated

Figure 6: Pharmacogenomic/toxicogenomic regulation and signaling of Phase -I and PhaseII enzymes participating in hepatic metabolism and detoxification processes for xenobiotics

(Adapted from Rushmore and Kong, 2002).

Animals and Alternatives

in Life Science Research 1103

BanarasHinduUniversity,Varanasi

2015

The team, led by toxicologist Dr. Thomas Hartung from Center for Alternatives to Animal Testing (CAAT),aims to map the molecular pathways of toxicity (PoT)in cells which will help in understanding the pathways of diseases and also to find the toxicity of the substance. "Mapping the entirety of these pathways, which I've termed the 'Human Toxome", will be a large-scale effort, perhaps on the order of the Human Genome Project" - Thomas Hartung.

Conclusion: These various in silico methods have their own advantages and limitations, and can be applied individually or integrated conceptually to help in potential risk assessment of plethora of chemicals/drugs. It reduces the use of animals and saves lives of voiceless animals being used in early phase of risk assessment. Thus, these methods establish a rationalized way of toxicity testing when coupled with in vitro and other omics approaches as envisioned in Tox21c.

References: Bhattacharya

S, Zhang Q, Carmichael PL, Boekelheide K, Andersen ME. Toxicity testing in the 21 century: defining new risk assessment approaches based on perturbation of intracellular toxicity pathways, PLoS One, 2011,6, e20887. Bouzom F, Ball K, Perdaems N, Walther B, Physiologically based pharmacokinetic (PBPK) modeling tools: how to fit with our needs? Biopharm Drug Dispos, 2012, 33, 55-71. Hartung T, McBride M, Food for Thought ... on mapping the human toxome, ALTEX,2011, 28, 83-93. Hayes WA, Loomis TA, Essentials of Toxicology, Academic Press, London, New York, 4th edition, 1996. Helma C, In silico predictive toxicology: the state-of-the-art and strategies to predict human health effects, Curr Opin Drug Discov Dev, 2005, 8,27-31. Jacob A, Pratuangdejkul J, Buffet S, Launay JM, Manivet P, In silico platform for xenobiotics ADME-T pharmacological properties modeling and prediction. Part II: The body in a Hilbertian space, Drug Discov Today, 2009, 14,406-412.

1041 Animalsand Alternativesin LifeScienceResearch

Banaras Hindu University, Varanasi

2015 Jena

GB, Kaul CL, Ramarao P, Regulatory requirements and ICH guidelines on carcinogenicity testing of pharmaceuticals: A review on current status, Indian J Pharmacol, 2005,37, 209-22. Katritzky R, Lobanov VS, Karelson M, QSPR: The correlation and quantitative prediction of chemical and physical properties from structure, Chern

Soc Rev, 1995,279-287 Mojca FJ, Andrew W, Review of Software Tools for Toxicity Prediction, EUR - Sci Tech Res Rep, 2010. National Research Council (NRC), Toxicity Testing in the 21st Century: A Vision and A Strategy, National Academy Press, Washington, DC, 2007. Panagiotou G, Taboureau 0, The impact of network biology in pharmacology and toxicology, SAR QSAR Environ Res, 2012, 23, 221-235. Raunio H, In silico toxicology - non-testing methods, Front Pharmacol, 2011, 30, 33. Rushmore TH, Kong AN, Pharmacogenomics, regulation and signaling pathways of phase I and II drug metabolizing enzymes, Curr Drug Metab, 2002, 3, 481-490. Shah I, Wambaugh J, Virtual tissues in toxicology, J Toxicol Environ Health B Crit Rev, 2010, 13,314-328. US EPA, A Framework for a Computational Toxicology Research Program, Office of Research and Development, U.S. Environmental Protection Agency, Washington,

DC, 2003.

Animals and Alternatives

in Life Science Research 1105