Risk assessment of parabens in a transcriptomics-based in vitro system

Parabens have been used for decades as preservatives in food, drugs and cosmetics. The majority however, were banned in 2009 and 2014 leaving only methyl-, ethyl-, propyl-, and butyl-derivates available for subsequent use. Methyl- and propylparaben have been extensively tested in vivo, with no resulting evidence for developmental and reproductive toxicity (DART). In contrast, ethylparaben has not yet been tested for DART in animal experiments, and it is currently debated if additional animal studies are warranted. In order to perform a comparison of the four currently-approved parabens, we used a previously established in vitro test based on human induced pluripotent stem cells (iPSC) that are exposed to test substances during their differentiation to neuroectodermal cells. EC50 values for cytotoxicity were 906 µM, 698 µM, 216 µM and 63 µM for methyl-, ethyl-, propyl- and butylparaben, respectively, demonstrating that cytotoxicity increases with increasing alkyl chain length. Genome-wide analysis demonstrated that FDR-adjusted significant gene expression changes occurred only at cytotoxic or close to cytotoxic concentrations, for example 1,720 differentially expressed genes (DEG) at 1,000 µM ethylparaben, 1 DEG at 316 µM, and no DEG at 100 µM or lower concentrations. The highest concentration of ethylparaben that did not induce any cytotoxicity nor DEG was 1670-fold above the highest published concentrations reported in biomonitoring studies (60 nM ethylparaben in cord blood). In conclusion, cytotoxicity and gene expression alterations of ethylparaben occurred at concentrations of approximately three orders of magnitude above human blood concentrations; moreover, the substance fitted well into a scenario where toxicity increases with the alkyl chain length, and gene expression changes only occur at cytotoxic or close to cytotoxic concentrations. Therefore, no evidence was obtained suggesting that in vivo DART with ethylparaben would lead to different results as the methyl- or propyl derivates. Human iPSCs were differentiated for 6 days according to a neuroepithelial protocol (UKN1) and incubated with parabens at the same time. Cytotoxicity was measured by a CellTiter-Blue® Cell Viability Assay. RNA was isolated from the cells and a whole transcritpome analysis was performed with Affymetrix HG-U133 Plus 2.0 arrays and Illumina RNA sequencing. Paraben-induced gene expression changes were determined and analyzed with statistical models.

对羟基苯甲酸酯（parabens）作为防腐剂已在食品、药品及化妆品领域应用数十年。然而绝大多数种类已于2009年和2014年被禁用，仅甲基、乙基、丙基及丁基衍生物仍可后续使用。
甲基对羟基苯甲酸酯与丙基对羟基苯甲酸酯已在体内开展广泛测试，未发现其存在发育与生殖毒性（developmental and reproductive toxicity, DART）相关证据。与之相反，乙基对羟基苯甲酸酯尚未在动物实验中开展DART相关测试，目前学界对于是否有必要开展额外动物研究仍存在争议。
为对比四种当前获批的对羟基苯甲酸酯，本研究采用了此前已建立的体外测试模型：以人诱导多能干细胞（human induced pluripotent stem cells, iPSC）为受试对象，使其在向神经外胚层细胞分化的过程中暴露于待测物质。
甲基、乙基、丙基及丁基对羟基苯甲酸酯的细胞毒性半数效应浓度（EC50）分别为906 μM、698 μM、216 μM及63 μM，结果显示细胞毒性随烷基链长度增加而增强。
全基因组分析显示，经错误发现率（false discovery rate, FDR）校正后的显著基因表达变化仅在细胞毒性浓度或接近细胞毒性浓度时出现：例如，在1000 μM乙基对羟基苯甲酸酯处理组中检出1720个差异表达基因（differentially expressed genes, DEG），316 μM组仅检出1个差异表达基因，而100 μM及更低浓度组未检出任何差异表达基因。
未引发任何细胞毒性及差异基因表达的乙基对羟基苯甲酸酯最高浓度，较生物监测研究中报道的最高检出浓度（脐带血中60 nM乙基对羟基苯甲酸酯）高出1670倍。
综上，乙基对羟基苯甲酸酯的细胞毒性及基因表达改变均出现在约高于人体血液浓度三个数量级的浓度下；此外，该物质符合“毒性随烷基链长度增加而增强，且基因表达改变仅在细胞毒性或接近细胞毒性浓度时出现”的规律。因此，未获得任何证据表明乙基对羟基苯甲酸酯的体内DART结果会与甲基或丙基衍生物存在差异。
本研究按照神经上皮细胞分化方案（UKN1）将人诱导多能干细胞诱导分化6天，并同时与对羟基苯甲酸酯共同孵育。
细胞毒性通过CellTiter-Blue®细胞活力检测试剂盒进行测定。
从细胞中提取RNA，并通过Affymetrix HG-U133 Plus 2.0芯片及Illumina RNA测序开展全转录组分析。
采用统计模型对由对羟基苯甲酸酯诱导的基因表达变化进行测定与分析。