Preliminary Study Of The Radiation Biological Dose Estimation Plasma Markers Based On GC-MS Metabolomics

Rapid radiation injury early triage after a radiological or nuclear exposure is vitalfor treatment of a large number of wounded people and limited medical resourcesallocation，it is critical for improving the treatment and survival efficiency. Here，ametabolomic methods based on gas chromatography-mass spectrometer（GC-MS）wasdeveloped for the analysis of plasma metabolic phenotype changes in rat bodyexploratory and select the potential radiation biomarkers. Finally，the radiation injurytriage was studied combined with non-linear kernel partial least squares（KPLS）. Inaddition, the biological significance and relevant metabolic pathways of thesebiomarkers were explored. The findings provide experimental evidence for developingfast, simple, sensitive and accurate radiation biodosimetry and understanding thepathogenesis of radiation injury and repair.1) A GC-MS method was developed to analyze plasma metabolites. Based on thetwo step derivative of MSTFA, the sample preparation process (such as the selection ofextraction solvent, the dosage of extraction solvent and the derivatization condition) andGC-MS analysis conditions including temperature programs were optimized. In addition,the analysis method characteristics including precision, reproducibility and stabilitywere investigated. With the NIST mass spectral database, the plasma metabolites wereconfirmed by comparing with the standard spectrometry.2) The GC-MS plasma metabolites method was applied to investigate the effects ofionizing radiation on rat plasma metabolites from the dose-response effect after differenttime points (5h,24h,48h and72h) post-exposure. Combined with the KPLS model, theradiation injury triage was studied. The classification accuracy of all levels of radiationinjury was all achieved100%at5h,48h and72h except that of24h with above92%classification accuracy. Moreover, the potential radiation biological dose estimationmarkers were screened based on genetic algorithm (GA) and random forest (RF) feature method.3) The biological significance and the key metabolic pathways involved in thesepotential radiation biological dose estimation markers were explored. The perturbedmetabolic pathways in response to radiation damage included amino acid metabolism,energy metabolism and lipid metabolism.