Targeted lipidomics-based study of radiation-induced metabolite profiles changes in plasma of total body irradiation cases

Lipidomics is an important tool for triaging exposed individuals, and helps early adoption of prevention and control strategies. The purpose of this study was to screen significantly perturbed lipids between pre- and post-irradiation of human plasma samples after total body irradiation (TBI) and explore potential radiation biomarkers for early radiation classification. Plasma samples were collected before and after irradiation from 22 hospitalized cases of acute myeloid leukemia (AML) prepared for bone marrow transplantation. Acute total-body γ irradiation was performed at doses of 0, 4, 8, and 12 Gy. Ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) with multiple reaction monitoring (MRM) method was utilized. Self-paired studies before and after irradiation were performed to screen potential lipid categorization markers and markers of dose-response relationships for radiation perturbation in humans. Based on the screened potential markers, a human TBI dose estimation model was developed. In total, 426 individual lipids from 14 major classes were quantified and 152 potential biomarkers with categorical characteristics were screened. A total of 80 lipids (32 TGs, 29 SMs, 9 FAs, 5 CEs, 5 PIs) were upregulated at 4 Gy, and a total of 91 lipids (39 SMs, 18 TGs, 15 HexCers, 7 CEs, 6 Cers, 3 LacCers, 2 LPEs, 1 PI) were upregulated at 12 Gy. Comparison of the ROC curves between the non-exposed and exposed groups at different doses showed AUC values ranging from 0.807 to 0.876. The metabolic pathways of potential lipid markers are mainly sphingolipid and glycerolipid metabolism, unsaturated fatty acid biosynthesis, fatty acid degradation and biosynthesis. Among the 13 dose-dependent radiosensitive lipids, CE (20:5), CE (18:1) and PI (18:2/18:2) were gradually incorporated into the TBI dose estimation model. This study suggested that it was feasible to acquire quantitative lipid biomarker panels using targeted lipidomics platforms for rapid, high-throughput triage. Lipidomics strategies for radiation biodosimetry in humans were established with lipid biomarkers with good dose-response relationship.

脂质组学（Lipidomics）是对受照个体进行分流筛查的重要工具，有助于及早采取防控策略。本研究旨在筛选全身照射（TBI）后人类血浆样本在照射前后发生显著扰动的脂质，并探索可用于早期辐射分类的潜在辐射生物标志物。研究收集了22例准备接受骨髓移植的急性髓系白血病（AML）住院患者照射前后的血浆样本。采用0、4、8、12 Gy剂量的急性全身γ射线进行照射，并采用搭载多反应监测（MRM）模式的超高效液相色谱-串联质谱（UPLC-MS/MS）方法完成检测。通过照射前后的自身配对研究，筛选人类辐射扰动的潜在脂质分类标志物及剂量-响应关系标志物。基于筛选出的潜在标志物，构建了人类TBI剂量估算模型。本研究共定量了14个主要脂质类别中的426种单个脂质，并筛选出152种具有分类特征的潜在生物标志物。在4 Gy剂量组中，共有80种脂质（32种甘油三酯、29种鞘磷脂、9种脂肪酸、5种胆固醇酯、5种磷脂酰肌醇）表达上调；在12 Gy剂量组中，共有91种脂质（39种鞘磷脂、18种甘油三酯、15种己糖基神经酰胺、7种胆固醇酯、6种神经酰胺、3种乳糖基神经酰胺、2种溶血磷脂酰乙醇胺、1种磷脂酰肌醇）表达上调。对不同剂量下未暴露组与暴露组的受试者工作特征（ROC）曲线进行比较，结果显示曲线下面积（AUC）值介于0.807至0.876之间。潜在脂质标志物的代谢通路主要包括鞘脂代谢、甘油脂代谢、不饱和脂肪酸生物合成以及脂肪酸降解与生物合成。在13种剂量依赖性辐射敏感性脂质中，CE（20:5）、CE（18:1）与PI（18:2/18:2）被逐步纳入TBI剂量估算模型。本研究表明，利用靶向脂质组学平台获取定量脂质生物标志物组合以实现快速、高通量分流筛查是可行的。本研究建立了基于具有良好剂量-响应关系的脂质生物标志物的人类辐射生物剂量学脂质组学策略。