Introduction: Here, we describe the generation of hypotheses for grouping nanoforms (NFs) after inhalation exposure and the tailored Integrated Approaches to Testing and Assessment (IATA) with which each specific hypothesis can be tested. This is part of a state-of-the-art framework to support the hypothesis-driven grouping and read-across of NFs, as developed by the EU-funded Horizon 2020 project GRACIOUS. Development of Grouping Hypotheses and IATA: Respirable NFs, depending on their physicochemical properties, may dissolve either in lung lining fluid or in acidic lysosomal fluid after uptake by cells. Alternatively, NFs may also persist in particulate form. Dissolution in the lung is, therefore, a decisive factor for the toxicokinetics of NFs. This has led to the development of four hypotheses, broadly grouping NFs as instantaneous, quickly, gradually, and very slowly dissolving NFs. For instantaneously dissolving NFs, hazard information can be derived by read-across from the ions. For quickly dissolving particles, as accumulation of particles is not expected, ion toxicity will drive the toxic profile. However, the particle aspect influences the location of the ion release. For gradually dissolving and very slowly dissolving NFs, particle-driven toxicity is of concern. These NFs may be grouped by their reactivity and inflammation potency. The hypotheses are substantiated by a tailored IATA, which describes the minimum information and laboratory assessments of NFs under investigation required to justify grouping. Conclusion: The GRACIOUS hypotheses and tailored IATA for respiratory toxicity of inhaled NFs can be used to support decision making regarding Safe(r)-by-Design product development or adoption of precautionary measures to mitigate potential risks. It can also be used to support read-across of adverse effects such as pulmonary inflammation and subsequent downstream effects such as lung fibrosis and lung tumor formation after long-term exposure.
Introduction: Investigation of normal human thyroid function using in vitro culture systems is dependent on cells that recapitulate physiology of differentiated thyrocytes. Primary thyrocytes retain features of the native organ but have limited lifespan in culture. Immortalized thyrocytes offer an alternative if challenges maintaining phenotypic stability can be overcome to retain functional features of primary cells.
Materials and methods: CI-SCREEN immortalization technology was applied to normal human thyroid tissue to generate four cell line variants. The lines were characterized for transgene integration, biomarker expression, genomic stability, and proliferation rates. Thyroid Stimulating Hormone (TSH)-dependent morphology, thyroglobulin production, thyroxine hormone synthesis, and viability were assessed using conventional 2D monolayer and 3D microtissue culture formats in huThyrEC or h7H medium.
Results: Despite differential transgene profiles, the lines had similar biomarker expression patterns and proliferation rates. In 2D culture there was no thyroxine synthesis or changes in viability, but TSH-dependent thyroglobulin production was more significant for several lines in h7H than huThyrEC medium. Comparatively, in 3D microtissues, TSH-dependent thyroglobulin induction was greater for cell lines in h7H medium. Synthesis of thyroxine in one cell line was higher than background with TSH exposure, but not significantly different than control.
Discussion: Immortalization of primary human thyrocytes yielded transgenic lines of epithelial origin. When evaluated in 2D or 3D culture formats, h7H medium supported thyroglobulin production to a greater magnitude than huThyrEC medium. One cell line cultured in 3D microtissue format marginally recapitulated T4 synthesis under continuous TSH exposure.
Conclusion: Select human thyroid cell lines exhibited morphological and functional features of primary thyrocytes and are a novel resource for in vitro disease modeling and toxicity testing that will enable reproducible culture models more representative of normal human thyroid function.
[This corrects the article DOI: 10.1089/aivt.2019.0005.].

