* Environmental Toxicology Graduate Program, University of California, Riverside, California 92521;
Toxicol Sci. 2013 Nov;136(1):205-15. doi: 10.1093/toxsci/kft182. Epub 2013 Aug 22.
The ability to anchor chemical class-based gene expression changes to phenotypic lesions and to describe these changes as a function of dose and time informs mode-of-action determinations and improves quantitative risk assessments. Previous global expression profiling identified a 330-probe cluster differentially expressed and commonly responsive to 3 hepatotumorigenic conazoles (cyproconazole, epoxiconazole, and propiconazole) at 30 days. Extended to 2 more conazoles (triadimefon and myclobutanil), the present assessment encompasses 4 tumorigenic and 1 nontumorigenic conazole. Transcriptional benchmark dose levels (BMDL(T)) were estimated for a subset of the cluster with dose-responsive behavior and a ≥ 5-fold increase or decrease in signal intensity at the highest dose. These genes primarily encompassed CAR/RXR activation, P450 metabolism, liver hypertrophy- glutathione depletion, LPS/IL-1-mediated inhibition of RXR, and NRF2-mediated oxidative stress pathways. Median BMDL(T) estimates from the subset were concordant (within a factor of 2.4) with apical benchmark doses (BMDL(A)) for increased liver weight at 30 days for the 5 conazoles. The 30-day median BMDL(T) estimates were within one-half order of magnitude of the chronic BMDLA for hepatocellular tumors. Potency differences seen in the dose-responsive transcription of certain phase II metabolism, bile acid detoxification, and lipid oxidation genes mirrored each conazole's tumorigenic potency. The 30-day BMDL(T) corresponded to tumorigenic potency on a milligram per kilogram day basis with cyproconazole > epoxiconazole > propiconazole > triadimefon > myclobutanil (nontumorigenic). These results support the utility of measuring short-term gene expression changes to inform quantitative risk assessments from long-term exposures.
将基于化学类别的基因表达变化与表型病变联系起来的能力,并将这些变化描述为剂量和时间的函数,可以为作用机制的确定提供信息,并提高定量风险评估的准确性。先前的全基因表达谱分析确定了一个 330 个探针的簇,该簇在 30 天时对 3 种肝肿瘤形成的唑类化合物(环丙唑醇、环氧氯丙烷和丙环唑)表现出差异表达和共同的反应性。将范围扩展到另外 2 种唑类化合物(三唑酮和咪菌腈),本评估涵盖了 4 种致肿瘤性和 1 种非致肿瘤性唑类化合物。对具有剂量反应性的簇的子集进行了转录基准剂量(BMDL(T))的估计,该子集的信号强度在最高剂量下有≥5 倍的增加或减少。这些基因主要包括 CAR/RXR 激活、P450 代谢、肝肥大-谷胱甘肽耗竭、LPS/IL-1 介导的 RXR 抑制以及 NRF2 介导的氧化应激途径。来自子集的中位数 BMDL(T)估计值与 5 种唑类化合物在 30 天时肝重增加的顶极基准剂量(BMDL(A))相一致(相差不超过 2.4 倍)。30 天中位数 BMDL(T)估计值与慢性肝细胞肿瘤的 BMDLA 相差一个数量级。某些 II 期代谢物、胆汁酸解毒和脂质氧化基因的转录反应的剂量反应性差异反映了每种唑类化合物的致瘤效力。30 天 BMDL(T)与基于毫克/千克/天的致瘤效力相对应,环丙唑醇>环氧氯丙烷>丙环唑>三唑酮>咪菌腈(非致瘤性)。这些结果支持了测量短期基因表达变化以告知长期暴露的定量风险评估的实用性。