Xiaotong Wang, Christine E Crute, Ashley Allemang, Jiri Aubrecht, Florence Burleson, Yasmin Dietz-Baum, Lena Dorsheimer, Albert Fornace, Roland Froetschl, Ulrike Hemmann, Constance A Mitchell, Stefan Pfuhler, Andrew Williams, Lorreta Yun-Tien Lin, Syril Pettit, Carole Yauk, Henghong Li
{"title":"TGx-DDI (Toxicogenomic DNA Damage-Inducing) Biomarker Validation: Multi-Site Ring Trial Supporting Regulatory Use.","authors":"Xiaotong Wang, Christine E Crute, Ashley Allemang, Jiri Aubrecht, Florence Burleson, Yasmin Dietz-Baum, Lena Dorsheimer, Albert Fornace, Roland Froetschl, Ulrike Hemmann, Constance A Mitchell, Stefan Pfuhler, Andrew Williams, Lorreta Yun-Tien Lin, Syril Pettit, Carole Yauk, Henghong Li","doi":"10.1093/toxsci/kfaf138","DOIUrl":null,"url":null,"abstract":"<p><p>Standard in vitro genotoxicity assays often suffer from low specificity, leading to irrelevant positive findings that require costly in vivo follow-up studies. The TGx-DDI (Toxicogenomic DNA Damage-Inducing) transcriptomic biomarker was developed to address this limitation by identifying DNA damage-inducing compounds through gene expression profiling in human TK6 lymphoblastoid cells. To qualify TGx-DDI as a reliable, reproducible biomarker for augmenting genotoxicity hazard assessment, a multi-site ring-trial was conducted across four laboratories using 14 blinded test compounds and standardized protocols. TK6 cells were exposed to three concentrations of each compound, followed by RNA extraction and digital nucleic acid counting using the NanoString nCounter® platform. A three-pronged bioinformatics approach-Nearest Shrunken Centroid Probability Analysis, Principal Component Analysis, and Hierarchical Clustering-was used to assign DDI or non-DDI classifications. TGx-DDI demonstrated 100% sensitivity, 86% specificity, and 91% accuracy in distinguishing DDI from non-DDI compounds under validated test conditions. High inter-laboratory concordance was observed (agreement coefficients ≥ 0.61), and transcriptomic data showed strong cross-site correlation (Pearson r > 0.84). The biomarker reproducibly classified test agents even when conducted across study sites. These results demonstrate that TGx-DDI is a robust and reproducible transcriptomic biomarker that enhances the specificity of genotoxicity testing by distinguishing biologically relevant DNA damage responses. Its integration into genotoxicity testing strategies can support regulatory decision-making, reduce unnecessary animal use, and improve the assessment of human health risks.</p>","PeriodicalId":23178,"journal":{"name":"Toxicological Sciences","volume":" ","pages":""},"PeriodicalIF":4.1000,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Toxicological Sciences","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1093/toxsci/kfaf138","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"TOXICOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Standard in vitro genotoxicity assays often suffer from low specificity, leading to irrelevant positive findings that require costly in vivo follow-up studies. The TGx-DDI (Toxicogenomic DNA Damage-Inducing) transcriptomic biomarker was developed to address this limitation by identifying DNA damage-inducing compounds through gene expression profiling in human TK6 lymphoblastoid cells. To qualify TGx-DDI as a reliable, reproducible biomarker for augmenting genotoxicity hazard assessment, a multi-site ring-trial was conducted across four laboratories using 14 blinded test compounds and standardized protocols. TK6 cells were exposed to three concentrations of each compound, followed by RNA extraction and digital nucleic acid counting using the NanoString nCounter® platform. A three-pronged bioinformatics approach-Nearest Shrunken Centroid Probability Analysis, Principal Component Analysis, and Hierarchical Clustering-was used to assign DDI or non-DDI classifications. TGx-DDI demonstrated 100% sensitivity, 86% specificity, and 91% accuracy in distinguishing DDI from non-DDI compounds under validated test conditions. High inter-laboratory concordance was observed (agreement coefficients ≥ 0.61), and transcriptomic data showed strong cross-site correlation (Pearson r > 0.84). The biomarker reproducibly classified test agents even when conducted across study sites. These results demonstrate that TGx-DDI is a robust and reproducible transcriptomic biomarker that enhances the specificity of genotoxicity testing by distinguishing biologically relevant DNA damage responses. Its integration into genotoxicity testing strategies can support regulatory decision-making, reduce unnecessary animal use, and improve the assessment of human health risks.
标准的体外遗传毒性检测通常特异性较低,导致不相关的阳性结果,需要昂贵的体内随访研究。TGx-DDI(毒物基因组DNA损伤诱导)转录组生物标志物是为了解决这一限制而开发的,通过人类TK6淋巴母细胞样细胞的基因表达谱鉴定DNA损伤诱导化合物。为了证明TGx-DDI是一种可靠的、可重复的生物标志物,可用于增强遗传毒性危害评估,研究人员在4个实验室进行了一项多地点环试验,使用14种盲法测试化合物和标准化方案。将TK6细胞暴露于每种化合物的三种浓度下,然后使用NanoString nCounter®平台进行RNA提取和数字核酸计数。一种三管齐下的生物信息学方法——最近缩小质心概率分析、主成分分析和分层聚类——被用于分配DDI或非DDI分类。在验证的测试条件下,TGx-DDI在区分DDI和非DDI化合物方面表现出100%的灵敏度,86%的特异性和91%的准确性。实验室间高度一致(一致系数≥0.61),转录组学数据显示出很强的跨位点相关性(Pearson r = 0.84)。即使在跨研究地点进行试验时,生物标志物也可重复分类试验剂。这些结果表明,TGx-DDI是一个强大的、可重复的转录组生物标志物,通过区分生物学相关的DNA损伤反应,增强了遗传毒性测试的特异性。将其纳入遗传毒性测试战略可以支持监管决策,减少不必要的动物使用,并改进对人类健康风险的评估。
期刊介绍:
The mission of Toxicological Sciences, the official journal of the Society of Toxicology, is to publish a broad spectrum of impactful research in the field of toxicology.
The primary focus of Toxicological Sciences is on original research articles. The journal also provides expert insight via contemporary and systematic reviews, as well as forum articles and editorial content that addresses important topics in the field.
The scope of Toxicological Sciences is focused on a broad spectrum of impactful toxicological research that will advance the multidisciplinary field of toxicology ranging from basic research to model development and application, and decision making. Submissions will include diverse technologies and approaches including, but not limited to: bioinformatics and computational biology, biochemistry, exposure science, histopathology, mass spectrometry, molecular biology, population-based sciences, tissue and cell-based systems, and whole-animal studies. Integrative approaches that combine realistic exposure scenarios with impactful analyses that move the field forward are encouraged.