Current opinion in toxicology最新文献

The lung is constantly exposed to a myriad of exogenous stressors. Ground-level ozone represents a ubiquitous and extremely reactive anthropogenic toxicant, impacting the health of millions across the globe. While abundant, epidemiological, in vivo, and in vitro data focuses the ozone toxicity in individual cell types (e.g. epithelial type II, alveolar macrophages) or signaling pathways involved in the injury (e.g. akt, glutathione). When appropriately used, bulk and single-cell RNA sequencing techniques have the potential to provide complete, and in certain cases unbiased, information of the molecular events taking place in the steady-state and injured lung, and even capture the phenotypic diversity of neighboring cells. To this end, this review compiles information pertaining to the latest understanding of lung cell identity and activation in the steady-state and ozone-exposed lung. In addition, it discusses the value and benefits of multi-omics approaches and other tools developed to predict cell–cell communication and dissect spatial heterogeneity.

Toxicogenomics is a subdiscipline of toxicology that assesses the toxicity using a combination of genomics, advanced computational technologies, and other high-throughput techniques. In the last few years, the rapid evolution of technologies such as transcriptomics, proteomics, metabolomics, and computational technologies allowed the application of a more flexible strategy for the risk assessment due to exposure to chemical and xenobiotic substances. The integration of bioinformatics approaches with novel genomic technologies permits extraordinary advancements in different fields of biology, toxicology, medicine, and, in particular, in occupational medicine, making accessible a “personalized” approach to this discipline. The scope of this paper is to investigate the recent trend of genomic toxicology in combination with bioinformatics approaches and its novel application in toxicological studies with reference to our experience and findings. In particular, it is shown how the dose biomarkers, the pollutants urine metabolites in our case, allow a precise exposure assessment whilst effect biomarkers can be individuated, significantly associated with the former. Among the effect biomarkers, our experience is mainly focused on oxidative stress biomarkers and miRNA profiles. Finally, it is shown how the early effect biomarkers are dysregulated when specific clinical outcomes are occurring. This early dysregulation can be used as individual susceptibility biomarker.

Translation of biological knowledge from animal models to humans is an important step in the development of therapeutics, but there remain limitations for effective translation. Systems biology offers approaches to understand the limitations for translation between species through data-driven models, such as methods that rely on learning patterns from data, and mechanism-driven models of biological processes, such as pharmacokinetic models. Here, we describe recent advances in both data-driven and mechanism-driven systems biology approaches to better understand limitations to translation from animal models to humans. Both approaches to modeling have their strengths and weaknesses but still provide key biological insight for translating between model systems and humans (Fig. 1). The presented methods not only identify differences between different model organisms but also provide opportunities to identify shared biomarkers and unique biological insight.

The nematode Caenorhabditis elegans offers great potential to address the need for faster and more reliable testing methods for predictive toxicology. The duration and cost of running C. elegans assays is comparable to cell-based in vitro testing, yet allows for toxic exposure information in a whole animal with many genetic, developmental, neuronal, and toxic mode of action processes that are conserved with mammals. Demonstrated areas of concordance for toxic response include aging, aneuploidy and germ cell genome abnormalities, growth and development, mammalian LD50 prediction, and neurotoxicity. Newer avenues of exploration, such as epigenetic regulation, innate immunity effects, and mutagenicity via DNA damage responses, also show promise. For predictive toxicology, the C. elegans model is most likely to prove useful as a complementary tool for early toxicity screening, as well as for the identification of conserved modes of toxic action.

New approach methodologies (NAMs) refer to any non-animal technology, methodology, approach, or combination thereof that can be used to provide information on chemical hazard and risk assessment that avoids the use of intact animals. A spectrum of in silico models is needed for the integrated analysis of various domains in toxicology to improve predictivity and reduce animal testing. This review focuses on in silico approaches, computer models, and computational intelligence for developmental and reproductive toxicity (predictive DART), providing a means to measure toxicodynamics in simulated systems for quantitative prediction of adverse outcomes phenotypes.

Translational safety biomarkers are responsive across nonclinical species where toxicity is directly correlated with histopathology; in humans, where histopathological assessment is not feasible, biomarker response is correlated with current standard biomarker response and clinical adjudication for tissue injury. Clinical scientists are nearly exclusively reliant on safety biomarkers to assess drug-induced tissue injury in clinical trials. It is therefore critical that safety biomarkers are deemed reliable for clinical decision-making and patient safety. Biomarker qualification is the formal regulatory acceptance of a biomarker for a specific context of use. Qualification results in certainty as to how the biomarker can be applied and how the biomarker data generated in drug development programs should be interpreted by both drug developers and regulators.

A major goal of translational toxicology is to identify adverse chemical effects and determine whether they are conserved or divergent across experimental systems. Translational toxicology encompasses assessment of chemical toxicity across multiple life stages, determination of toxic mode of action, computational prediction modeling, and identification of interventions that protect or restore health after toxic chemical exposures. The zebrafish is increasingly used in translational toxicology because it combines the genetic and physiological advantages of mammalian models with the higher-throughput capabilities and genetic manipulability of invertebrate models. Here, we review the recent literature demonstrating the power of the zebrafish as a model for addressing all four activities of translational toxicology. Important data gaps and challenges associated with using zebrafish for translational toxicology are also discussed.

The advances in the immunotherapy field in the last decade have enabled efficient treatment of previously intractable cancers. These advances have also exposed the limitations of commonly used animal models for safety prediction of such drugs. In fact, 90% biological immunotherapies entering clinical trials fail because of low efficacy and/or unmanageable toxicity despite having undergone rigorous preclinical safety assessment. This low predictability is due to challenges related to the poor conservation of key immune-biological aspects between man and standard preclinical safety workhorses. To void this gap, new models are urgently needed. Here, we discuss the growing area of advanced human physiological in vitro and in silico models and propose future directions needed to enable more accurate and faster prediction of drug responses.