Discovery of Noninvasive Biomarkers for Radiation Exposure via LC-MS-Based Hair Metabolomics.

Abstract

Ionizing radiation exposure from a potential nuclear energy plant leak or detonation of a nuclear weapon can cause massive casualties to both warfighters and civilians. RNA, proteins, and metabolite biomarkers in biological specimens like blood and tissue have shown potential to determine radiation dose levels. However, these biomarkers in blood and urine are short-lived, typically detectable within hours or a few days. To address the need for stable, long-term radiation exposure biomarkers, we developed two mass spectrometry-based methods using noninvasive hair samples to identify radiation-exposure biomarkers. Our results show that hippuric acid and 5-methoxy-3-indoleacetate significantly increase after higher (4 gray) doses of gamma irradiation compared to lower (1 and 2 Gy) doses or nonexposed hair samples. While 2-aminooctadec-4-ene-1,3-diol, oleoyl ethanolamide, palmitoylcarnitine, 25-hydroxy vitamin D3, vernolic acid, and azelaic acid significantly increased over time after exposure. Trimethylamine N-oxide (TMAO) was found in higher concentrations in female specimens across all time points. Further validation using a machine learning model suggested that these biomarkers can predict differences in the exposure dose and time point. Our findings highlight the potential of noninvasive hair sample analysis for assessing radiation exposure, offering a viable alternative to address critical public health concerns of unexpected radiation exposure.