Rapid High-Throughput Diagnostic Triage after a Mass Radiation Exposure Event Using Early Gene Expression Changes
Matthias Port,Patrick Ostheim,Matthäus Majewski,Thorsten Voss,Julian Haupt,Andreas Lamkowski,Michael Abend
DOI: https://doi.org/10.1667/RR15360.1
2019-10-07
Radiation Research
Abstract:Radiological exposure scenarios involving large numbers of people require a rapid and high-throughput method to identify the unexposed, and those exposed to low- and high-dose radiation. Those with high-dose exposure, e.g., >2 Gy and depending on host characteristics, may develop severe hematological acute radiation syndrome (HARS), requiring hospitalization and treatment. Previously, we identified a set of genes that discriminated these clinically relevant groups. In the current work, we examined the utility of gene expression changes to classify 1,000 split blood samples into HARS severity scores of H0, H1 and H2–4, with the latter indicating likely hospitalization. In several previous radiation dose experiments, we determined that these HARS categories corresponded, respectively, to doses of 0 Gy (unexposed), 0.5 Gy and 5 Gy. The main purpose of this work was to assess the rapidity of blood sample processing using targeted next-generation sequencing (NGS). Peripheral blood samples from two healthy donors were X-ray irradiated in vitro and incubated at 37°C for 24 h. A total of 1,000 samples were evaluated by laboratory personnel blinded to the radiation dose. Changes in gene expression of FDXR , DDB2 , POU2AF1 and WNT3 were examined with qRT-PCR as positive controls. Targeted NGS (TREX) was used on all samples for the same four genes. Agreement using both methods was almost 78%. Using NGS, all 1,000 samples were processed within 30 h. Classification of the HARS severity categories corresponding to radiation dose had an overall agreement ranging between 90–97%. Depending on the end point, either a combination of all genes or FDXR alone (H0 HARS or unexposed) provided the best classification. Using this optimized automated methodology, we assessed 100× more samples approximately three times faster compared to standard cytogenetic studies. We showed that a small set of genes, rather than a complex constellation of genes, provided robust positive (97%) and negative (97%) predictive values for HARS categories and radiation doses of 0, 0.5 and 5 Gy. The findings of this study support the potential utility of early radiation-induced gene expression changes for high-throughput biodosimetry and rapid identification of irradiated persons in need of hospitalization.
radiology, nuclear medicine & medical imaging,biology,biophysics