Improved gene expression biodosimetry for dose estimation using an expanded panel of radiation-responsive genes

Gene expression analysis provides a minimally invasive approach for biological dosimetry. To advance point-of-care applications, this study aimed to establish and validate an improved gene expression biodosimetry system by employing an expanded panel of radiation-responsive genes in human peripheral blood. Human B lymphoblastoid cells (AHH-1) and peripheral blood from 10 healthy donors were irradiated with ⁶⁰Co γ-rays at doses of 0, 1, 2, 4, 6, and 8 Gy (dose rate: 1 Gy/min). The expression patterns of four candidate transcriptional biomarkers (ZMAT3, SESN1, AEN, and TRIAP1) and a panel of radiation-responsive genes were characterized at 6–48 h post-irradiation. The impact of different dose rates (0.2, 1, and 2 Gy/min) on these gene expressions was also investigated. For each gene, calibration curves were established by fitting a linear regression between the logarithm of absorbed dose and ΔCt values. Gene selection and model construction were performed using stepwise regression to obtain optimized multi-gene models. The accuracy of these dosimetry models for dose prediction was validated in independent ex vivo and in vivo cohorts. The four candidate genes exhibited robust, dose-dependent expression from 6 to 48 h post-irradiation, independent of dose-rate variations (0.2–2 Gy/min). Most genes in the expanded panel, including the candidates, showed strong linear relationships between log2 of dose and ΔCt values across all time points when the 0 Gy point was excluded from regression (R² > 0.90, S < 0.50). Based on these validated genes, optimized multi-gene models achieved high predictive accuracy (R² = 0.81–0.89) with fewer genes. Furthermore, these improved models demonstrated accurate dose estimation capabilities when validated with both ex vivo- and in vivo-irradiated peripheral blood samples. Our study expanded the panel of reliable radiation biomarkers and developed optimized multi-gene models for accurate dose estimation, thereby advancing the standardization and practicality of gene expression biodosimetry.