Department of Cell Toxicology , Helmholtz Centre for Environmental Research - UFZ , Permoserstr. 15 , DE-04318 Leipzig , Germany.
Environmental Toxicology, Center for Applied Geoscience , Eberhard Karls University Tübingen , Hölderlinstr. 12 , DE-72074 Tübingen , Germany.
Chem Res Toxicol. 2019 Aug 19;32(8):1646-1655. doi: 10.1021/acs.chemrestox.9b00182. Epub 2019 Aug 2.
Most studies using high-throughput cell-based bioassays tested chemicals up to a certain fixed concentration. It would be more appropriate to test up to concentrations predicted to elicit baseline toxicity because this is the minimal toxicity of every chemical. Baseline toxicity is also called narcosis and refers to nonspecific intercalation of chemicals in biological membranes, leading to loss of membrane structure and impaired functioning of membrane-related processes such as mitochondrial respiration. In cells, baseline toxicity manifests as cytotoxicity, which was quantified by a robust live-cell imaging method. Inhibitory concentrations for baseline toxicity varied by orders of magnitude between chemicals and were described by a simple quantitative structure activity relationship (QSAR) with the liposome-water partition constant as a sole descriptor. The QSAR equations were remarkably similar for eight reporter gene cell lines of different cellular origin, six of which were used in Tox21. Mass-balance models indicated constant critical membrane concentrations for all cells and all chemicals with a mean of 69 mmol·kg(95% CI: 49-89), which is in the same range as for bacteria and aquatic organisms and consistent with the theory of critical membrane burden of narcosis. The challenge of developing baseline QSARs for cell lines is that many confirmed baseline toxicants are rather volatile. We deduced from cytotoxicity experiments with semi-volatile chemicals that only chemicals with medium-air partition constants >10,000 L/L can be tested in standard robotic setups without appreciable loss of effect. Chemicals just below that cutoff showed crossover effects in neighboring wells, whereas the effects of chemicals with lower medium-air partition constants were plainly lost. Applying the "volatility cut-off" to >8000 chemicals tested in Tox21 indicated that approximately 20% of Tox21 chemicals could have partially been lost during the experiments. We recommend applying the baseline QSARs together with volatility cut-offs for experimental planning of reporter gene assays, that is, to dose only chemicals with medium-air partition constants >10,000 at concentrations up to the baseline toxicity level.
大多数使用高通量细胞生物测定的研究都将化学物质测试到某个固定浓度。更合适的做法是测试到预计会引起基线毒性的浓度，因为这是每种化学物质的最小毒性。基线毒性也称为麻醉作用，是指化学物质在生物膜中的非特异性插入，导致膜结构丧失和膜相关过程（如线粒体呼吸）受损。在细胞中，基线毒性表现为细胞毒性，这可以通过稳健的活细胞成像方法进行定量。基线毒性的抑制浓度在化学物质之间相差几个数量级，并通过简单的定量构效关系（QSAR）用脂质体-水分配常数作为唯一描述符来描述。对于来自不同细胞来源的八个报告基因细胞系，QSAR 方程非常相似，其中六个用于 Tox21。质量平衡模型表明，所有细胞和所有化学物质的临界膜浓度都相同，平均值为 69mmol·kg（95%置信区间：49-89），与细菌和水生生物的范围相同，与麻醉作用的临界膜负担理论一致。为细胞系开发基线 QSAR 的挑战在于，许多已确认的基线毒物相当易挥发。我们从半挥发性化学物质的细胞毒性实验中推断出，只有中等空气分配常数 >10,000 L/L 的化学物质才能在标准机器人设置中进行测试，而不会明显降低效果。刚好低于该截止值的化学物质在相邻孔中显示出交叉效应，而较低的中等空气分配常数的化学物质的作用则明显丧失。将“挥发性截止值”应用于 Tox21 中测试的>8000 种化学物质表明，大约 20%的 Tox21 化学物质在实验过程中可能部分丢失。我们建议在报告基因测定的实验计划中应用基线 QSAR 以及挥发性截止值，即仅在空气分配常数>10,000 的化学物质在达到基线毒性水平的浓度下进行剂量处理。
