A World Health Organization (WHO) International Programme on Chemical Safety (IPCS) Framework outlines a pragmatic approach to the identification and consideration of priorities in the grouping and assessment of combined exposures to multiple chemicals. The Framework includes formal problem formulation followed by step wise consideration of both exposure and hazard in several tiers of increasingly data-informed analyses. These analyses build on recent developments in assessment, incorporating predictive approaches in early tiers and increasingly refined, more data-informed mode of action and probabilistic analyses in later tiers.

Recently, the Framework has been additionally developed in a number of initiatives in both research and application, including guidance of the Organization for Economic Cooperation and Development and the European Food Safety Agency, the development of tools within the European Horizon 2020 Research Project, Euromix and a recent WHO Publication on Chemical Mixtures in Source and Drinking Water.

The Framework also serves as an organizing construct for a “Roadmap” for methodological guidance on chemical risk assessment of the WHO/IPCS Harmonization Initiative. This Roadmap provides the basis for practical advice for the implementation of these various methodologies in tiered assessment and management strategies.

Tiered assessment incorporating methodological advances internationally will be illustrated through presentation of examples and case studies. The implications of increasing international experience in tiered assessment incorporating new and evolving methods will also be addressed.

View PDF

Per- and Polyfluoroalkyl Substances (PFAS) are a group of synthetic substances that were originally developed in the late 1940s for uses as surfactants and surface protectors. The term PFAS includes both polymeric and non-polymeric fully (per-) and partially (poly-) fluoroalkyl substances. The most well studied PFAS are perfluorooctanoic acid (PFOA), a Substance of Very High Concern (SVHC) as designated by the European Union, and perfluorooctane sulfonate (PFOS). Contemporary uses include industrial applications such as polymeric non-stick coatings on cookware, cosmetics, fabric protectors, waxes, paint and firefighting foams. PFAS have made their way into the biosphere in all continents. Some, including PFSAs and PFCAs, are especially persistent due to their strong carbon-fluorine bonds. Due to growing evidence of bioactivity, public health guidance is being developed in numerous countries to limit human exposures through drinking water and food. Also included as PFAS are perfluoroalkanes, other perfluoroalkyl sulfonic acids (PFSA) and perfluoroalkyl carboxylic acids (PFCA), and recently described perfluoroalkyl ether carboxylic acids (PFECA) and sulfonic acids (PFESA), among other proposed subclasses. Information is rapidly emerging on PFAS chemical classification and grouping, environmental chemistry, detection technology, fate and transport, exposure potential, human health toxicity, and ecological toxicity. This presentation will review recent social and public health actions regarding PFAS exposure and toxicity.

Methylmercury is an environmental pollutant that can cause severe central nervous system disorders. However, the mechanism of methylmercury toxicity and the biological defense against it have not been clarified yet. Examining the effects of methylmercury on the concentrations of low molecular weight metabolites may help us to understand how methylmercury affects the brain. We performed an exhaustive metabolomic analysis and found that changing levels of many metabolites in the mouse brain in response to methylmercury. Moreover, administration of methylmercury to mice has been found to increase the levels of putrescine, a type of polyamine, specifically in brains, a phenomenon considered to be a protective response to neuronal toxicity caused by methylmercury.

Gene therapy can be defined as the introduction of nucleic acids into a host cell to result in the replacement or inactivation of a mutated gene, or to introduce something novel. The introduction of the nucleic acids can occur inside or outside of the body, often with the use of a viral vector, such as adeno-associated viruses (AAV). In the past several years, there has been a dramatic increase in the use of gene therapies to develop potential therapies for treating various diseases. However, there are multiple challenges associated with this relatively novel technique; such as antibody or cell mediated responses that limit efficacy, lack of data on the long term safety, and an evolving regulatory environment that establishes the requirements for drug development of gene therapies. This presentation will provide an overview of gene therapy, review the key aspects of the available nonclinical regulatory guidance, discuss things to consider when developing an nonclinical toxicology strategy, and examine case studies of nonclinical packages used for the development of novel therapies.

Perhaps most of the research in toxicology are conducted in the laboratory, today. Considering that the ultimate goal of toxicological research is to promote human health, toxicological research in the field should have equal importance. In this talk, I would like to introduce a couple of field toxicology studies, which focused on general populations in South and Southeast Asian countries. Taking some examples such as (1) groundwater contamination by arsenic found in the Asian countries, (2) life-style transitions in Asian countries and their impact on usage of chemicals , Observations in these field will be discussed regarding their implications for ‘lab toxicology’ and toxicology in general. It will be emphasized that toxic manifestations observed in the field should reflect the interaction between the toxicant and many other chemical and non-chemical factors. In a sense, ‘field’ toxicology is indispensable not only because the field is the only place where the implications of laboratory toxicology can be validated but also because field observations may generate good questions around which new theoretical, experimental studies can be evolved.

Much attention in toxicology has been given to the biotransformation of chemicals and the mechanism of action of toxicants. However, the toxicity of chemicals is also dependent upon the ability to adapt to the presence of the toxicant. One mechanism for adaptation to chemicals that his been examined in detail is the induction of the cytochrome P-450s (CYPs). However, while the induction of the CYPs will increase the biotransformation of many chemicals to less toxic products and results in adaptation, but for other chemicals the CYPs metabolize them to more toxic products and results in more toxicity. These same chemicals that induce the CYPs can also induce uptake and efflux transporters, which have the ability to also adapt cells to the presence of some chemicals, which will be addressed in this lecture. A second mechanism to be discussed is the adaptation to chemicals by the induction of metallothionein (MT). MT is a small protein that is made up of 1/3 cysteines and avidly binds metals such as zinc and cadmium. It is one of the most inducible proteins and plays in a major role in the adaptation to cadmium toxicity. The third mechanism of adaptation to be discussed is the Nrf2-Keap1 transcription factor induction of a plethora of enzymes and transporters that protect against numerous chemicals whose mechanism of toxicity is via electrophiles and oxidative stress. These and other mechanisms of adaptation play a major role in protecting us from the toxicity of chemicals.

There are a huge number of chemical substances in the environment. Among them, electrophiles are covalently bound to protein nucleophiles, resulting in formation of protein adducts involved in activation/disruption of cellular redox signaling pathways and toxicity. It should be noted that we are exposed to a variety of xenobiotic electrophiles on a daily basis through living environment, lifestyle and dietary habits. In the panel discussion, I introduce our findings to establish a modeling of xenobiotic electrophile exposome.

Part of exogenous environmental chemicals such as electrophiles influences the enzymatic activity of proteins via covalent modification. Whereas, there has not been fully elucidated the effects of these chemicals on cells, tissues, organs, and body throughout life. Exposome is defined as the measure of all exposures of an individual in a lifetime and how those exposures relate to health and disease. Thus, it is very difficult for us to solve this huge issue (exposome) completely at present. We therefore decided to search how adduct formation by chemicals affect the intracellular functions and epigenetics. Emerging results revealed that exposure to some environmental chemicals affects epigenetics. In this symposium, I would present our recent results and discuss how we should deal with exposome.

Lysine residue in proteins contains a highly reactive amino group that is targeted for various posttranslational modifications including acylation. Acyl modification on lysine residues is induced by endogenous carboxylic acids such as fatty acids through chemical reactions in vivo. We are exposed daily to various compounds including carboxylic acids with foods. These exogenous carboxylic acids may be also added to lysine residues of proteins in the body. In this presentation, we would like to introduce our research on lysine acyl modifications and discuss possible influences of exposure to food constituent carboxylic acids from the perspective of lysine adduct exposome.

The need of the regulatory guidelines for nonclinical safety assessment was defined by the thalidomide tragedy. Effects on embryo-fetal development (EFD) should be evaluated using two species (rats and rabbits) because thalidomide was found to cause severe malformations only in a few species including rabbits. Afterward, the internationally harmonized (ICH S5(R2)) guideline was provided for the assessment of nonclinical developmental and reproductive toxicity (DART) testing by the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH). Expert Working Group (EWG) was formed to revise the current ICH S5(R2) guideline in 2015. After the discussion by EWG members, the stage of the ICH S5(R3) guideline has been reached to Step4 in 2020.

The major parts of the revised ICH S5(R3) guideline are not different from the current ICH S5(R2) guideline. The set of studies conducted should encompass observations through one complete life cycle (i.e., from conception in one generation through conception in the following generation). However, there are some key differences between the current and revised versions. One of the differences is that the purpose of the revised guideline is to describe potential strategies and study designs to supplement available data to identify, and assess risk. For example, assessment strategies for pharmaceuticals being developed can be planned in consideration of the target population or therapeutic indication. Another is that alternative approach is acceptable to assess EFD risk under limited circumstances. According to the revised ICH S5(R3) guideline, alternative assays are encouraged and have the potential to defer or replace conventional in vivo studies and reduce animal use.

I will show the general considerations and expectations for detecting DART of human pharmaceuticals, in addition to the lessons learned through the activities of revising the ICH S5 guideline. This presentation will give a hint on future approaches for the assessment of DART risk.

For the food production, young bulls are castrated for the improvement of meat quality through the removal of male hormones. Importantly, these castrated testes are discarded because they are not suitable for food. Thus, over 99% of bull testes are “unused agricultural resources”, although they have a potential to produce numerous spermatozoa. If the method to produce spermatozoa in vitro from castrated testis are established, it will be an important agricultural technology for the future animal production. Additionally, although I am not expert of toxicology, I have received advice on what in vitro spermatogenesis using castrated bovine testes will be good model for the reproductive toxicity as an alternative method over animal experiment. In this meeting, I will show recent our study about bovine spermatogenesis and discuss its potential usefulness in the field of toxicology.

In Japan, the birthrate continues to decline and the average maternal age of which the woman gives birth for the first time is increasing. In such a situation, More and more pregnant ladies are anxious about their fetal health.

In Japan, noninvasive prenatal test (NIPT) has started to be offered since 2013 as a prenatal test for chromosomal abnormalities including trisomy 13, trisomy 18 and trisomy 21. NIPT is available from 10 weeks of gestation and very accurate screening test. In addition, this test can be carried out only by a blood taken from a pregnant lady and noninvasive for a fetus. For that reason, a lot of pregnant ladies want to have this test and the hospitals which are authorized to offer the test by NIPT consortium have not been able to cope with all the request. However, these chromosomal abnormalities which are thought to be appropriate as targets of NIPT are just part of congenital abnormalities.

On the other hand, because of the development of the resolution of ultrasound machines, early diagnosis of fetal diseases is now gathered attention. In our hospital, since December of 2017, we have started first trimester scan not only for assessment of risks of chromosomal abnormalities but also anatomical assessment to detect structural abnormalities which can not be detected by NIPT.

The technology of NIPT is innovative but still limiting. Therefor, by combining NIPT with anatomical assessment by the scan, a prenatal test may be able to be promoted to a fetal health check.

In this symposium, we will show the performance of first trimester anatomical assessment by the scan and discuss fetal medicine following a prenatal diagnosis.

In drug development, embryo-fetal risk assessment is critical to ensure safety in patients. Recently, the modality of drugs has been diversified to antibody, nucleic acids, and gene therapy; while species differences in epitopes or nucleic acid sequences limit the appropriate safety evaluations in animals. In this presentation, a convenient approach to evaluate human-specific embryo-fetal risks in vitro using human iPS cells will be introduced. Also, current in vitro systems for evaluation of embryo-fetal risk will be summarized and the usefulness of human-specific systems will be discussed.

As a base of safety assessment of pharmaceuticals, the initial data management requires accurate toxicological data acquisition, which is based on regulatory safety studies according to guidelines, and computational systems have been developed under the application of GLP. In addition to these regulatory toxicology studies, investigative toxicological study data for the selection of lead compound and candidate compound for clinical trials are directed to the estimation by computational systems such as QSAR and related expert systems.

Furthermore, in the “Go”, “No-Go” decision of drug development, supportive utilization of a scientifically interpretable computational toxicology system is required for human safety evaluation. Pharmaceutical safety evaluator as a related toxicologist who is facing to practical decision does not need a data-driven AI (Artificial Intelligence) system that calls for the final consequence, rather requires an explainable AI that can provide comprehensive information necessary for evaluation and can help decision making. Through the explication and suggestion of information on the mechanism of toxic effects to safety assessment scientists, ultimately a subsidiary partnership system for risk assessment is to be a powerful tool that can indicate project-vector with data weight for the corresponding counterparts.

To bridge the gaps between the big-data and the knowledge, multi-dimensional thinking based on philosophical ontology theory is necessary to handle heterogeneous data such as interpretable computational toxicology related to drug safety assessment.

The safety of pharmaceuticals, general chemicals, and food materials has been evaluated in an integrated manner mainly by combining in vitro and in vivo experimental data, in silico toxicity predictions, knowledge of experts in various fields, and cases reported in the literature.

For safety assessments using in silico systems, both the AI system that outputs the final results and the use of scientifically explainable AI techniques and computer systems based on the mechanism of action are useful.

Pathway ontology and experimental information considering each hierarchy of molecular, cellular, organelle, and organ levels is useful as data on the mechanism of toxicity. This information includes the molecular initiating event involving the binding of a compound to biomolecules based on the adverse outcome pathway concept. This information also includes the key event connected to the mechanism of toxicity and adverse outcomes comprising adverse events such as toxicity.

In addition, to evaluate in vivo safety in humans and rats, it is necessary to incorporate information on the mechanism of action of based on the ADME concept (absorption, distribution, metabolism, and elimination) in the body, the mechanism of drug metabolism, and kinetics such as the formation, conjugation, and production of reactive metabolites. Because the onset of the toxicity of compounds distributed to organs is largely influenced by the amount and duration of exposure in the organs, it is important to consider a range of action mechanisms from pharmacokinetics to toxicity by utilizing information including the concentration profiles of each organ based on physiological models.

The current status and future prospects of computer systems and AI technologies for safety assessments based on the mechanism of action from pharmacokinetics to toxicity and experimental information are presented together with examples of current activities.

Some R & D in this presentation was supported by the Ministry of Economy, Trade and Industry "Development of evaluation technology for energy-saving electronic device materials" (Development of high-speed, high-efficiency safety evaluation technology to support social implementation of functional materials and development of next-generation safety prediction methods using artificial intelligence using toxicological big data).

Computer science has significantly contributed to toxicology. Data-driven research is indispensable for omics technologies, pathway analysis, and Quantitative Structure-Activity Relationship (QSAR) models for toxicity prediction. However, in order to safety evaluation, we need not only an informatics-oriented approach but also a philosophy-oriented ontological approach for accountability from a scientific viewpoint [1]. Furthermore, safety management requires support for sharing knowledge across disciplines and interoperating on how to reduce risk.

This presentation illustrates how ontology can be contributed to the drug safety evaluation with examples of a toxicity process ontology, TXPO, and a toxic process interpretable knowledge system, TOXPILOT (https://toxpilot.nibiohn.go.jp). Ontology identifies the intrinsic nature of the target from a domain-independent viewpoint and gives weighting and directions appropriately according to the user’s purpose. Knowledge infrastructure based on humans and computers interoperable serves heterogeneous bridging data and contributes to the progress of computational toxicology.

[1] Yamagata Y., Yamada H., and Horii I., Current status and future perspective of computational toxicology in drug safety assessment under ontological intellection, J Toxicol Sci. 44(11):721-735, (2019)

Immune checkpoint inhibitors, bispecific T-cell engagers and CAR-T therapy have provided unique therapeutic effects. Currently, great efforts are underway to develop various therapeutics of these platforms. These therapeutics have excellent therapeutic effects and severe adverse effects. The most frequent and common adverse effect is cytokine release syndrome (CRS) that often becomes dose-limiting toxicity. Recent mechanistic investigations have suggested possibility of prevention of severe CRS without reducing anti-tumor activities. Here, I would like to review the findings of the recent CRS investigations suggesting potential measures to control severe CRS.

Recently, various modalities of cancer immunotherapy products have been developed, such as antibody, peptide vaccine, gene therapy product which expresses the cancer antigen in vivo, and chimeric antigen receptor (CAR)-engineered T cells. Preclinical safety evaluation of these products should be conducted on case-by-case basis and is delineated in “Guidance on cancer immunotherapy development” (PSFB/ELD Notification No.0308-1).

In this presentation, our preclinical consideration for cancer immunotherapy products will be provided based on our experience.

Development of cancer immunotherapeutics creates some unique challenges for nonclinical safety assessment. These can range from lack of pharmacological activity in standard toxicology models to exaggerated pharmacology that can be dose limiting. This presentation will describe two case studies, one checkpoint inhibitor (atezolizumab) and one T cell dependent bispecific (CD20/CD3 TDB), to serve as an example of each scenario. Furthermore, nonclinical regulatory guidances and strategies to support clinical combinations containing cancer immunotherapeutics will be discussed.

KEAP1-NRF2 system is a sulfur-based cytoprotection mechanism. KEAP1 utilizes multiple cysteine residues for sensing electrophiles, whereas NRF2 activates various genes regulating sulfur-involving reactions. Although NRF2 enhances mitochondrial activity, a precise mechanism has not been fully understood. Based on our recent discovery that mitochondrial sulfur metabolism makes a substantial contribution to generation of the mitochondrial membrane potential, we hypothesized that NRF2 promotes the mitochondrial activity through promoting the sulfur metabolism. Since SLC7A11 encoding a cystine transporter, xCT, is a well-known NRF2 target gene, NRF2 was likely to enhance the mitochondrial sulfur metabolism to generate persulfides potentially by increasing the supply of their substrate, cysteine. Indeed, xCT inhibition decreased the mitochondrial membrane potential. In addition, we found that SQRDL, encoding sulfide-quinone oxidoreductase (SQR), is directly activated by NRF2. To examine the in vivo contribution of SQR, we generated SQR mutant mice, which express mutant SQR lacking a mitochondrial localization signal. The mutant mice exhibited compensatory facilitation of β-oxidation. Although the mutant mice died by 10 weeks after birth with normal diet, high-fat diet extended their lifespan. Our detailed sulfur metabolome analysis revealed that SQR effectively catalyzes proton and electron transfer from hydropersulfides in addition to sulfide to the mitochondrial electron transport chain (ETC). Thus, SQR makes an essential contribution to the maintenance of the mitochondrial membrane potential by mediating proton/electron supply to the ETC. These results suggest that one of the mechanisms for NRF2-dependent mitochondrial activation is the facilitation of sulfur respiration in mitochondria following the increased cystine uptake.

Electrophiles such as MeHg modulate cellular functions by covalent modification with endogenous thiols. Recently we found that exposure of cells to MeHg elevated intracellular 8-nitro-cGMP. Exposure to MeHg also induced S-guanylation and activation of H-Ras followed by injury to neurons. The electrophile-induced activation of redox signaling and the consequent cell damage were attenuated by pretreatment with a RSS donor. Thus, exogenous electrophiles impair the redox signaling regulatory mechanism, particularly of intracellular RSS, and therefore lead to cellular pathogenesis.

Cysteine persulfide (CysSSH) and polysulfides (CysS[S]nH, n > 2) are cysteine derivatives having sulfane sulfur atoms bound to cysteine thiol. Accumulating evidence has suggested that cysteine per/polysulfides are abundantly present in prokaryotes and eukaryotes, and play important roles in diverse biological processes such as antioxidative host defense. Reagents that can increase intracellular per/polysulfide levels become powerful chemical tools for understanding of the physiological and pathological roles of persulfide species in biological systems. We here describe the synthesis of N-acetyl-L-cystine polysulfides (NAC polysulfides) and application of them as potent per/polysulfide donors in cell culture and in animal models. Mass spectrometry-based sulfur metabolomics revealed that NAC polysulfides very efficiently donate their sulfane sulfur atoms to acceptor thiols such as reduced glutathione, leading to generation of hydroper/polysulfides in test tubes as well as in cultured cells. Importantly, NAC polysulfides exhibited strong anti-inflammatory activities against macrophages stimulated with lipopolysaccharide as evidenced by marked suppression of pro-inflammatory cytokine production including tumor necrosis factor-α and interferon-γ. Inhibitions of IKK/NF-κB axis as well as TRAM-TRIF pathway were suggested to be involved in anti-inflammatory activities of NAC polysulfides. Furthermore, we demonstrated that NAC polysulfide treatment protected mice from lethal endotoxin shock diseases. Taken together, NAC polysulfides can become powerful chemical tools to study biological processes that are sensitive to endogenous per/polysulfide levels and to develop new strategies to treat pathological conditions associated with dysregulated inflammatory responses.

Recently, it has been clarified that polysulfides have an antioxidant activity and are involved in intracellular protein functions by techniques such as analysis utilizing mass spectrometry and development of fluorescent probes. Additionally, when sulfide was added to serum in vitro system, it bound to serum proteins and was not observed as a sulfide, suggesting that the serum proteins have a sulfur metabolism mechanism. However, the presence of polysulfides in serum and other biological fluids has almost not reported because of lacks a method for quantifying oxidized polysulfides. Therefore, to develop a comprehensive understanding of oxidized polysulfide in serum, we developed a novel assay for measuring oxidized polysulfide and clarified its physiological functions. This assay revealed that polysulfides were present in biological fluids such as serum and semen, and that human serum albumin (HSA) maintained high concentrations of oxidized polysulfides in human blood. Interestingly, HSA could convert oxidized polysulfide to reduced polysulfide by the stimulation of oxidative stress, and has a reaction mechanism that temporarily increases antioxidant activity. This oxidative stress response was also observed in clinical serum samples from diabetic nephropathy and acute renal impairment, which are oxidative stress-related diseases. These findings will lead to the elucidation of the function control of serum proteins with sulfur metabolism mechanisms and the development of markers for prevention and diagnosis of oxidative stress diseases. In this symposium, we would like to introduce about the latest findings on biological fluids other than the polysulfide pool in blood described above.

Selenium, an essential trace element, is highly reactive and toxic, but the living body takes in the properties of selenium and uses it for biological defense. Selenium is mainly incorporated into the proteins in the form of selenocysteine (Sec: an analog of cysteine that contains selenium instead of sulfur) and forms the active site of glutathione peroxidase and thioredoxin reductase. Selenium is essential for the antioxidative defense. However, in recent years, it has been revealed that the disorder of selenium metabolism is involved in lifestyle-related diseases such as diabetes. The plasma selenium-containing protein selenoprotein P (SeP) induced by high glucose and high fat worsens insulin resistance and insulin secretion, acting as a “bad guy” in the onset and development of diabetes.

Selenium contained in food is digested, absorbed from the gastrointestinal tract, and then enters the synthesis pathway of selenium-containing proteins. The metabolic pathway differs depending on the form of selenium. Sec is recognized by the living body as "selenium", cleaved by Sec lyase, and then inorganic selenium enters the Sec synthesis pathway. On the other hand, selenomethionine (SeMet: an analog of methionine that contains selenium instead of sulfur) absorbed in the body is metabolized in vivo without distinction between selenium and sulfur, and a part is incorporated into proteins, directly. In this presentation, the interaction between selenium and sulfur metabolism, especially the metabolic pathway of amino acids containing each element and the discrimination mechanism of each element in vivo, is outlined. Further, the cross-talk between selenium and sulfur metabolism, especially the biological response against electrophile and its detoxication, are discussed.

In this presentation, I introduce detecting potential of developmental neurotoxicity by analysis of hippocampal neurogenesis, and then, induction of aberrant neurogenesis by developmental exposure to neurotoxicants targeting mature neurons. Next, I present availability of hippocampal neurogenesis for detecting developmental neurotoxicity in a framework of 28-day toxicity study. Finally, I show examples of biomarkers for detection of developmental neurotoxicity obtained by global gene screening techniques, in particular, irreversible markers downregulated by hypermethylation.

Determination of the potential for developmental neurotoxicity (DNT) is necessary for assessing and evaluating the toxic characteristics of a chemical substance. Our aim is to study an electrophysiological developmental index at the lactation period to predict adverse effects after the development. Using hippocampal slices obtained from rat pups at lactation period, we found a sudden enhancement of stimulation/response relationship in the CA1 area on the day of eye-opening. Then, we investigated prenatally exposed to valproic acid (VPA) or tributyltin (TBT), known as a developmental neurotoxicant, and found that VPA accelerated but TBT decelerated the congenital development of hippocampal excitability programmed on the day of eye-opening. Thus, our experimental model can be verified to be useful for DNT evaluation.

Environmental chemicals have potential effects on children's neurodevelopmental toxicity, leading to neurobehavioral outcomes such as autism. Since the developing brain is much more vulnerable to chemicals compared with the adult brain, exposure to chemicals during prenatal and/or postnatal period can cause delayed neural disorders.

Current developmental neurotoxicity (DNT) guidelines (OECD #426, EPA #OPPTS 870.6300) using in vivo studies comprise in vivo behavioral observational tests and other tests to assess DNT. To overcome the disadvantages of high cost and time consuming of in vivo DNT tests, we have established the human iPSC-based platform. The differentiation in vitro is considered to be similar to the stages of brain development in human. We have evaluated the effect of DNT compounds on neural differentiation capability and functional neural network recorded by multi-electrode assay system. The iPSC technology would replace the traditional black box animal models by providing mechanistic data at the cellular and molecular levels.

In the symposium, I would like to talk about a possible role of iPSC on in vitro DNT testing. In addition, I would like to share the current status and future perspectives of in vitro DNT testing, which has been proposed at OECD.

Micro-electrode array (MEA) assay using human iPSC-derived neurons are expected to one of in vitro assessment to predict the toxicity and predict the mechanism of action of chemical compounds. However, the analytical method that can predict the toxicity from MEA data are not established. In this study, we attempted to detect the risk ranking of pesticides from MEA data in cultured human iPSC-derived neurons. Human iPSC-derived neurons (Neucyte inc.) were cultured on Micro-electrode array (MEA) plate, and 15 pesticides were tested at 5 concentrations from 0.01 to 100 µM. Using multivariate analysis of parameters for synchronized burst firings, we have succeeded in distinguishing the dose-dependent responses to pesticides into low, middle, and high risk. In addition, we found that deep learning using the divvied image data of raster plots can separate dose-dependent responses into low, middle, and high risk. Although there is a problem of in vitro to in vivo extrapolation, analytical methods using multivariate analysis and deep learning are useful for the detection of risk ranking of pesticides from MEA data in cultured hiPSC-derived neuronal networks.

Computational toxicology techniques can be used to construct ultra-high-throughput prediction systems capable of distinguishing toxic substances from a large number of candidate compounds. However, in the creation of high-accuracy in silico prediction models, problems are encountered, such as the low number of actual measured toxicity values. In this presentation, I will introduce a drug hepatotoxicity/pulmonary toxicity prediction system built to address these problems in the AMED project: “Construction of drug discovery informatics system.” In addition, I will introduce the AI-SHIPS project in the Ministry of Economy, Trade and Industry.

Drug metabolism and disposition of industrial chemicals with toxicity concern such as phthalates were mediated by humanized-livers in chimeric mice in the similar manner as humans in terms of biliary/urinary excretions of their glucuronides. A significant difference in the concentration-time profiles of the primary metabolite 7-hydroxywarfarin in humanized-liver mice treated with tienilic acid that cytochrome P450 2C9 was metabolically inactivated. These toxicological studies were extensively investigated by humanized-model animals. Nonetheless, in general, toxicity studies for chemicals may require significant cost and time. The main aim of the present study was to model the plasma, hepatic, and renal pharmacokinetics of disparate types of chemicals and drugs after virtual oral administrations in rats based on reported rat plasma values and experimental pharmacokinetics determined after oral administration to rats for evaluating toxicological potential. The calculated absorbance rates of the chemicals based on cell permeability studies were found to be inversely correlated to the no-observed-effect (NOEL) levels for hepatoxicity after oral administration taken from the Hazard Evaluation Support System Integrated Platform in Japan. The lowest-observed-effect level (LOEL) values in rats for oral administrations given in the database and these hepatic area under the curve values that we obtained from modeling were inversely correlated (r = -0.78, p < 0.05, n = 7). This significant example indicated the importance of simulating the hepatic levels of general chemicals, because those with high hepatic concentrations are more likely to be potent hepatotoxic compounds. These predicted effective concentrations can then be used to determine the range of test concentrations to be applied in vitro or in vivo. Where NO(A)EL/LO(A)EL and pharmaco/toxicokinetic data from in vivo studies are available, a PBPK model can be used to predict the effective concentrations in plasma and target tissues. The present simplified models could estimate the relationships between hepatic/plasma concentrations and oral doses of general chemicals using both forward and reverse dosimetry. These methods are therefore valuable for estimating hepatic toxicity.

Drug-induced liver injury (DILI) can cause hepatic failure and result in drug withdrawal from the market. Preclinical prediction of DILI risk is very challenging and safety assessments based on animals inadequately forecast human DILI risk. In contrast, human-derived in vitro cell culture-based models could improve DILI risk prediction accuracy. Here, we developed and validated a method to assess DILI risk associated with various compounds. Fifty-four marketed and withdrawn drugs classified as DILI risks of “most concern”, “less concern”, and “no concern” based on Liver Toxicity Knowledge Base were tested using a combination of four assays addressing mitochondrial injury, intrahepatic lipid accumulation, inhibition of biliary network, and bile acid-dependent toxicity. Using these in vitro testings, an algorithm with the highest available DILI risk prediction power was built by artificial neural network (ANN) analysis. The optimal combination of assays (or descriptors in general) may not yet be achieved, but the strategy we employed here may be one of the strategies to predict DILI risk whose precise pathogenesis is unknown.

We used chromosome engineering technology to develop human artificial chromosomes (HAC) and mouse artificial chromosomes (MAC) that can introduce large human genes, multiple human genes in a stable manner. In this presentation, we will introduce new drug discovery tools (fully human antibody-producing animal, human drug metabolism model animal, metabolic enzyme expressing cell, toxicity monitor cell, etc.) developed by HAC/MAC technology, and further introduce new combined technologies of DNA synthesis and HAC/MAC.

We observed that acetaminophen induced CYP by inhibition of the proteolysis, and phospholipid accumulation by metabolic activation of loratadine in hepatocyte spheroids. On the other hand, we have evaluated drug-induced phospholipidosis by amiodarone, and cholestasis by rifampicin in chimeric mice with humanized liver. It may also be useful for assessment of drug-induced steatosis. Furthermore, it can be used for estimation of safety margin because PK could be predicted by chimeric mice. It is expected that combination of the in vitro and in vivo evaluation can contribute to drug discovery.

Regarding the in vivo drug-induced toxicity, the production of reactive metabolite (s) is considered to be an initiation reaction for both intrinsic- and idiosyncratic-type of toxicity. GSH content in kidney, heart and muscle is much lower than in liver, lung and spleen in normal rodents. GSH levels in the kidneys and muscles decrease rapidly by drug administration and recover slowly. A series of animal models that have established and utilized for drug-induced toxicity by reducing the scavenge ability by administering GSH synthase inhibitor L-buthionine-(S, R)-sulfoximine (BSO) to experimental animals will be introduced. (1) BSO was administered to normal mice for 7 days, an acute kidney injury model was established, and this model was able to detect the renal damage of the drug with high sensitivity. (2) A mouse model of rhabdomyolysis was established using a combination of a new quinolone antibiotic and a statin and a combination of a fibrate and a statin. This required the use of BSO, and became a test system that could assess the risk of drug-interaction with statins. (3) In the liver injury model in mice, the involvement of immune/inflammation-related factors is clear, but in rats it is unclear and liver damage is unlikely to occur. Therefore, in rats, combined use of BSO is necessary to induce liver injury. (4) GSH-conjugation of acyl glucuronide metabolites was inhibited by BSO, indicating that acyl glucuronide metabolites were involved in vivo in renal injury. Although there are still unclear about idiosyncratic drug-induced organ damage in humans, information from in vivo animal models will be useful for future research.

Today, gene therapy using gene editing tool (genome editing therapy) is becoming potential new modality treatment. But we need to consider it comes with unconventional safety issues. AMED research group and taskforce established in Japan Pharmaceutical Manufacturers Association have been cooperating in searching of genome editing therapy products and establishing a safety guideline for genome editing therapy. In this session, we will overview the landscape of genome editing therapy products development as the outcome of this co-working activity.

When genome editing is performed in mouse fertilized eggs to induce DSB, insertion of unintended sequence, such as retrotransposons, endogenous genes, and vectors, occurs in approximately 10% of DSB sites (on target). The generation of fusion genes of these unintended sequences and endogenous genes can be a new risk event for clinical applications using genome editing (on-target risk). We also showed that exosome-mediated horizontal gene transfer occurs in DSB repair during genome editing, and is the driving force behind mammalian genome evolution.

CRISPR/Cas9 system has opened the new era for reverse genetics. In 2013, we developed the efficient gene knockout system in mice by injecting the plasmids expressing humanized Cas9 (hCas9) and single-guide RNA (sgRNA) into zygotes. Now it is replaced by electroporation of oocytes with CAS9/crRNA/tracrRNA ribonucleoprotein complex for simple genome editing (knockout, point mutation, tag insertion, etc). Altogether, to date, we have knocked out 272 testis abundant genes and analyzed the phenotypes in vivo. Whereas 168 of the KO mouse lines were fertile and did not show any drastic phenotypes, 9 KO mouse lines showed lethality. The remaining 103 KO mouse lines showed infertility or subfertility and propelled our research. I would like to introduce recent findings in mammalian fertilization. In addition, I would also like to discuss on the risk and benefit of CRISPR/Cas9 mediated genome editing in mice.

The arms race between parasitic sequences and hosts is a major driving force for evolution of gene control systems. Since transposable elements (TEs) are potentially deleterious, eukaryotes silence them by epigenetic mechanisms such as DNA methylation. Little is known about how TEs counteract the silencing to propagate during evolution. Here we report behavior of sequence-specific anti-silencing proteins used by Arabidopsis TEs and coevolution of those proteins and their target sequences. Through this coevolution, these TEs propagate with minimum host damage. Our findings provide insight into the evolutionary dynamics of these apparently "selfish" sequences.

About 360 million years ago (the late Devonian period), the first terrestrial vertebrate, amphibian emerged from water and adapted to life on land. These animals evolved their skin epidermis into a keratinized stratified squamous epithelium to prevent water loss and as protection from sunlight. Stratum corneum (SC) is the uppermost dead cell layer of skin epidermis and provides these protective barrier function. Previously, we reported that mammalian skin-specific retroviral-like aspartic protease SASPase are the key regulator of SC moisturization. It cleaves profilaggrin to produce filaggrin monomer in the lower SC. As its optimum pH of SASPase protease activity is low pH, we hypothesized that lower SC has an acidic environment. To determine the pH distribution in the SC, we fused two fluorescent proteins, pH-sensitive Venus and insensitive mCherry and generated mice expressing the fusion protein from the uppermost stratum granulosum (SG1 cells). Confocal microscopic analysis revealed that SC has three distinct pH layers, lower-slightly acidic, middle highly acidic and upper-neutral SC layers, rather than gradual change over SC layers. SC pH imaging in various parts of the body confirmed that the lower-acid SC layer consistently exists in tail, foot pad, ear and back skin irrespective of their thickness and compactness of SC. These findings collectively indicate that SC indeed has an environment to activate SASPase protease. Thus, co-option (exaptation) of SASPase gene function was achieved not only the acquirement of SASPase gene but also the generation of acidic SC, result in moisturization of mammalian skin.

Drug-induced lung injury is defined as a lung injury that results from the specific use of a drug including over-the-counter drugs, herbal medicines, supplements, and illegal narcotics. Drug-induced interstitial lung disease that is one of the most common types of drug -induced lung injury occasionally causes fatal outcome. Recent anti-cancer drugs, such as tyrosine kinase inhibitors, immune checkpoint inhibitors and antibody drug conjugates, are not only highly effective for tumors but also common causative medications of drug induced interstitial lung disease. Pathogenetic mechanisms of drug induced interstitial lung disease have not been well known but direct toxic effects on alveolar type I epithelial cells, airway epithelial cells, or vascular endothelial cells and activation immune cells as a hapten or an antigen are suspected one of the pathogenetic mechanisms. Radiological and pathological patterns are classified based on the morphologic patterns of idiopathic interstitial pneumonia, hypersensitivity pneumonitis and sarcoidosis and so on. These patterns are not specific and different with each case. But these patterns are useful to differentiate drug induced interstitial lung disease from other lung diseases and identify diffuse alveolar damage pattern as a poor prognostic factor. In this presentation, radiological/pathological findings of drug induced pneumonitis and clinical significance of the morphologic patterns will be reviewed.

Drug-induced lung injury (DLI) in cancer patients is a possibility with almost all anticancer drugs, including cytotoxic anticancer drugs, molecular targeted agents, and immune checkpoint inhibitors. Although the precise mechanisms involved in DLI are not fully understood, it is thought to be caused by direct toxicity to pulmonary tissue and/or immune-mediated effects. Furthermore, this condition is influenced by various host and environment factors. The risk factors for the development of DLI are as follows: preexisting pulmonary lesions (interstitial lung disease / pulmonary fibrosis / chronic inflammatory diseases), hyperoxia, postoperative acute lung injury, combination therapy with anticancer drugs, radiation therapy, genetic factors, aging, or smoking. Although preclinical safety studies indicated no evidence of injury to intact lungs, some clinical cases showed DLI. Furthermore, despite the observation of immune-related adverse events (irAEs) in animal models, the quality and quantity of irAEs in clinical, including prediction of target organs, has not been fully established. In this presentation, I will outline pathological features and issues affecting various preclinical lung injury models, and then introduce preclinical evaluations of DLI using the bleomycin-induced lung injury model. In addition, translational issues, which is the problem of difference between preclinical and clinical, will be discussed.

Trastuzumab deruxtecan (T-DXd; DS-8201) is a HER2-targeting antibody-drug conjugate composed of a humanized anti-HER2 antibody and an exatecan derivative (DXd), a topoisomerase I inhibitor, which are bound together by a cleavable peptide-based linker. Before clinical trials, 6-week toxicity studies (every 3 weeks totaling 3 doses) with T-DXd were conducted in cynomolgus monkeys (cross-reactive species) and in rats (non cross-reactive species). The major target organs/tissues in rats and monkeys were the intestines and bone marrow. This effect seemed to be attributable to the cytotoxic effects of DXd and typical dose-limiting factors in the clinical use of topoisomerase I inhibitors. T-DXd caused pulmonary toxicity in monkeys at ≥30 mg/kg, although it was not observed in rats. In a 3-month monkey toxicity study (T-DXd every 3 weeks for a total of 5 doses), pulmonary toxicity was observed at 30 mg/kg (the highest dose). An extended dosing period did not increase the severity of lesions. While comprehensive mechanisms of the pulmonary toxicity remain unclear, this finding in monkeys could be relevant to the understanding of mechanism of interstitial lung disease (ILD) in patients treated with T-DXd. In this presentation, nonclinical toxicity data are reviewed with an emphasis on relevant safety findings.

Low molecular weight compounds such can interact with multiple targets, which often causes harmful unrecognized effects, even by the developers. To understand the multiple effects of a compound, omics analysis is a suitable choice because it converts the biological information of a specimen into the numeric one without bias, based on its comprehensiveness.

We have developed a novel analysis method, orthogonal linear separation analysis (OLSA), whose advantage is easiness for biologists to understand the outcome of the unbiased data analysis. We introduce the framework and the application.

In the new drug discovery, it has been general to perform the quantitative analysis of drug pharamacokinetics using mathematical modeling. Physiologically-based pharamacokinetic (PBPK) model, which directly uses physiological and anatomical parameters such as human blood flow and organ volume, has been focused in recent years. The parameters of PBPK model correspond a certain physiological and physicochemical meaning. Theses parameters values have inter-individual variability by genetic polymorphism, ethnicity, individual difference, etc. The accumulation of those variability may cause unexpected blood concentraion-time profile, pharmacological and toxicological effects of drugs It is difficult to detect such unexpected results in the early stage of clinical study because of the insufficient number of subject. On the other hand, the information of inter-individual variability of PBPK model parameters is obtained. We established virtual clinical study (VCS), which uses computational simulation and prediction of drug pharamacokinetics, pharmacological and toxicological effects in the specific population with virtual patients generated by the variability information. In this presentation, I will introduce some examples of VCS to predict results of actual clinical studies.

Health foods served with tablets or capsules contain concentrated functional substances and may cause health damage due to potent bioactivities. In 2005, Ministry of Health, Labour and Welfare in Japan (MHLW) provided the guidelines to manufacture health foods using the method of Good Manufacturing Practice (GMP) and to ensure safety of raw materials and products. The amendment of the Food Sanitation Act in 2018 in MHLW has introduced a designation system for imposing mandatory GMP of foods containing the ingredients and components that require particular care. These issues will be discussed.

The EU Cosmetics Directive ban on animal in vivo testing and its marketing ban have led the cosmetics industry to develop alternatives to animal testing, and Shiseido has established a weight-of-evidence based safety assurance system that doesn’t test on animals. As for photo-sensitization, the current non-animal method is not enough to test cosmetic materials, and thus we have been developing several in vitro tests (e.g. photo-DPRA, photo-KeratinoSens^TM and photo-h-CLAT) corresponding to the Adverse Outcome Pathway. In this lecture, we will introduce our efforts to develop a photo-sensitization assessment system and the challenges for the safety assurance of cosmetics.

Nanoparticles (NPs) that are smaller than 100 nm are widely used in various products including cosmetics, foods, and medicines, Because of their beneficial properties such as enhanced tissue permeability and surface reaction compared to other micro or larger sized materials. In contrast, the risk analysis of NPs is needed, since NPs exhibit different pharmacokinetics from other micro or larger sized materials and may induce unknown biological responses. From this respect, hazard identification of NP has been analyzed worldwide. On the other hand, few studies have examined the exposure situations and kinetics for NPs. Especially, it was recently reported that particles were re-formed from metal ions as well as that metal NPs ionized. Furthermore, we reported that the difference of their physical properties, such as particles size or distinguish between particles and ions, affected biological responses. Therefore, it is necessary to evaluate their physical properties in the body, in order to understand the safety of NPs. However, it is difficult to quantitatively analyze their physical properties in the body in the conventional methods, so that the development of a new technology which establish them, has been required. In this regard, we have developed the technology that can quantitatively analyze their physical properties in the body by optimizing the pretreatment method for biological samples.

In this symposium, I would like to introduce an overview of the technology and discuss how to ensure the safety of pharmaceuticals, foods, and cosmetics in humans.

In recent years, tetrodotoxin (TTX) has drawn international attention as a shellfish toxin. “Mouse Unit (MU)” has been used for a long period for evaluation of shellfish toxin in Japan; MU is determined by “Mortality” considering the lethal time in an acute toxicity study (ATS). On the other hand, EFSA has reported evaluation of TTX in which the “Apathy” was used as endpoint (EP) in an ATS. ATSs are also essential for classification and labeling of substances, but are criticized both scientific and ethical grounds; not to provide information on possible target organs and possible mechanisms of toxicity, and Mortality is not a desirable EP from the viewpoint of animal welfare. It is considered an excellent way to improve ATSs by accurately measuring and quantifying the signs of toxicity across multiple items and creating reasonable criteria for the acute toxicity of a substance. However, this will be a complicated and expensive test compared to ATSs, which uses Mortality as the endpoint. For modernization of ATSs, we are focusing on vital signs and wearable devices with cutting-edge sensors, furthermore, to consider validity of the vital signs for acute toxicity, comprehensive gene expression analysis of central nervous system was conducted. In other words, it is the development of a test method that obtains vital signs from animals corresponding to tests that human acutely poisoned patients undergo in emergencies. In this symposium, we would like to share the effect on central nervous system of TTX as a case evaluated by modernization of ATS.

Domoic acid (DA) is a kainic acid-type neurotoxin that causes amnesic shellfish poisoning. Previously, we demonstrated brain dysfunction with behavioral impairments by the disturbance of neural signals at prenatal period by using DA. In this symposium, we will show the effects on emotional/cognitive behavior of male adult mice induced by administration of DA at postnatal period. Furthermore, we also introduce the experimental trials to develop countermeasures against DA poisoning.