Shrestha, Vanasse, Cooper, and Antosh (2017) profiled 15,222 Drosophila melanogaster genes at days 2, 10, and 20 after irradiation at exposures of 10 to 20,000 roentgen and identified a small panel of genes whose expression fold-change responds linearly (R2>0.9R^2 > 0.9) with dose at all sampled times — including Irbp (human homologue XRCC6), CG7277/COQ6, CG12264/NFS1, Pex14/PEX14, and Dyrk3/DYRK2 — and proposed the panel, particularly the genes with human homologues, as a candidate gene-expression radiation biodosimeter. The paper explicitly leaves three questions open: whether the Drosophila results translate to humans through the homologous genes; whether the panel genes remain linearly dose-responsive at postirradiation times other than days 2, 10, and 20, especially earlier than 2 days; and how the responses depend on the energy and type of irradiation.

Background

After a large-scale radiation event — a nuclear detonation, reactor accident, or occupational mishap — many exposed people carry no physical dosimeter, so radiation biodosimetry (estimating an individual's absorbed dose from biological markers rather than instruments) becomes a medical necessity for triage. One long-developed approach is gene-expression dosimetry: ionizing radiation alters the transcript levels of genes involved in DNA-damage response and stress signaling in accessible cells such as peripheral blood, and the magnitude of these changes can, in principle, be inverted to estimate dose. In humans, the mature candidate signatures are built on strongly inducible genes such as FDXR (ferredoxin reductase), DDB2 (DNA damage-binding protein 2), and CDKN1A (which encodes the cyclin-dependent kinase inhibitor p21) in blood, and recent reviews (Amundson 2023; Abend et al. 2023) describe the approach as promising but not yet deployable, with individual variation, confounders, and realistic exposure scenarios among the open challenges.

Shrestha et al. (2017) approached the same goal from an unusual direction: the fruit fly Drosophila melanogaster. They measured expression of 15,222 fly genes at 2, 10, and 20 days after whole-body irradiation at exposures of 10, 1000, 5000, 10,000, and 20,000 roentgen (R; 1 R ≈ 0.01 Gy, where the gray, Gy, is the SI unit of absorbed dose). The paper's abstract describes these as X-ray exposures, though its methods state that the irradiation came from a chamber containing cesium-137, a gamma-ray source. The paper treats 10,000 R as roughly 100 Gy, so the range spans about 0.1–200 Gy; flies tolerate doses far above the 0–10 Gy window relevant to human casualty triage, and a companion lifespan study of the same fly cohort found no longevity effect below about 10,000 R (Antosh et al. 2014). A gene was called dose-linear when its expression fold-change (irradiated divided by control) fit a straight line in dose with R2>0.9R^2 > 0.9 (the coefficient of determination of the linear fit). Six genes were linear at all three time points with every dose included, and thirteen when the lowest, near-noise 10 R dose was excluded; the thirteen retain five of the original six (CG7277 drops out because its day-20 linear fit falls to R2=0.88R^2 = 0.88 once the 10 R point is removed), and four genes of each set have known human homologues (genes in different species descended from a common ancestral gene, here taken from the NCBI HomoloGene database). The flagship candidate is Irbp (inverted repeat-binding protein, FlyBase ID FBgn0011774; FlyBase is the standard Drosophila gene database), homologous to human XRCC6, which encodes the Ku70 subunit of the Ku70/Ku80 heterodimer that initiates non-homologous end joining (NHEJ) — the main pathway repairing radiation-induced DNA double-strand breaks. The other homologous pairs are CG7277/COQ6 (ubiquinone biosynthesis), CG12264/NFS1 (iron–sulfur cluster assembly), Pex14/PEX14 (peroxisomal protein import), and Dyrk3/DYRK2 (protein phosphorylation). Notably, some linear genes switch response direction across the dose range (fold-change below 1 at low doses, above 1 at high doses), and several have no known function.

The plausibility of carrying these fly results to humans is genuinely uncertain rather than routine. Cross-species conservation of radiation-responsive genes is imperfect even between closely related mammals: Amundson's 2023 review highlights that FDXR, one of the most robustly induced genes in irradiated human blood, shows no significant response in macaques and is down-regulated in baboons. There is partial prior evidence that the flagship homologue responds in humans — Ku70/XRCC6 transcript and protein are up-regulated in irradiated human peripheral blood mononuclear cells (the resting white-blood-cell fraction commonly used for ex vivo irradiation assays) at early times up to about 4 hours (Shelke and Das 2015) — but no one has tested XRCC6, let alone the rest of the panel, for the specific property the fly study found attractive: a linear dose-response that persists across many days after exposure. The major human dose-reconstruction signatures (for example Ghandhi et al. 2019) contain none of the five homologues. Separately, radiation quality is known to matter: in human blood, neutron versus X-ray exposure and even the neutron energy spectrum itself change which genes respond and by how much (Basavaraj et al. 2026), and Drosophila radiobiology as a whole still lacks systematic dosimetry, dose-rate, and photon-energy studies (Trinca and de Navascués 2025). Against this backdrop, the source paper closes with three explicit future questions bearing on whether its panel can become a practical biodosimeter.

Problem Statement

The source paper fixed a concrete panel of Drosophila melanogaster genes whose expression responds linearly with radiation dose across days 2, 10, and 20 postirradiation: the six genes linear with all doses from 10 to 20,000 roentgen (Irbp/FBgn0011774, CG2909, CG7277, CG12264, Pex14, Qtzl) and the thirteen genes linear when the 10 roentgen point is excluded (these thirteen comprise five of the original six — all except CG7277, whose day-20 linear fit falls to R2=0.88R^2 = 0.88 without the 10 roentgen point — together with agt, Dyrk3, Arc1, CR10102, CG9646, CG10638, CG14695, and CG17278). Determine, for this fixed panel, the three questions the source left open, which bear on whether the panel can serve as a practical gene-expression radiation biodosimeter:

  1. Cross-species translation: do the human homologues of the panel genes — XRCC6 (homologue of Irbp), COQ6 (CG7277), NFS1 (CG12264), PEX14 (Pex14), and DYRK2 (Dyrk3) — reproduce the dose-response behavior found in the fly, in particular a linear response of expression fold-change with dose that persists across postirradiation times, in human tissues (for example, accessible blood cells)?

  2. Temporal generality: do the panel genes continue to respond linearly with dose at postirradiation times other than days 2, 10, and 20 — in particular at times earlier than 2 days — thereby establishing the time window over which such a dosimeter would be reliable?

  3. Radiation energy and type: how do the panel's responses change with the energy and type of irradiation (for example, different photon energies or particle radiations rather than the source's single radiation quality — irradiation in a cesium-137 gamma chamber, reported by the paper's abstract as X-ray exposures in roentgen)? Does the dose-linearity of individual genes and the composition of the responsive panel persist, or do linearity and gene selection change with radiation modality?

Whether the published literature settles each of these three questions, or leaves it open, is itself part of what is to be determined.

References

  1. Shrestha S, Vanasse A, Cooper LN, Antosh MP. Gene Expression as a Dosimeter in Irradiated Drosophila melanogaster. Journal of Computational Biology 24(12):1265–1274, 2017. DOI: 10.1089/cmb.2017.0170. https://doi.org/10.1089/cmb.2017.0170
  2. Antosh M, Fox D, Hasselbacher T, Lanou R, Neretti N, Cooper LN. Drosophila melanogaster show a threshold effect in response to radiation. Dose-Response 12(4):551–581, 2014. DOI: 10.2203/dose-response.13-047.Antosh. https://doi.org/10.2203/dose-response.13-047.Antosh
  3. Trinca TM, de Navascués J. Drosophila melanogaster: an old and future ally to radiobiology. Journal of Radiation Research 66(6):579–593, 2025. DOI: 10.1093/jrr/rraf060. https://doi.org/10.1093/jrr/rraf060
  4. Basavaraj R, Vanasse A, Broustas C, Garty G, Deoli NT, Harken A, Nam S, Goodwin C, Hemme CL, Amundson SA, Antosh M. Gene expression as a biological dosimeter: effects of different neutron energies. Radiation and Environmental Biophysics 65(3):957–966, 2026. DOI: 10.1007/s00411-026-01237-4. https://doi.org/10.1007/s00411-026-01237-4
  5. Amundson SA. Transcriptomics for radiation biodosimetry: progress and challenges. International Journal of Radiation Biology 99(6):925–933, 2023. DOI: 10.1080/09553002.2021.1928784. https://doi.org/10.1080/09553002.2021.1928784
  6. Abend M, Ostheim P, Port M. Radiation-Induced Gene Expression Changes Used for Biodosimetry and Clinical Outcome Prediction: Challenges and Promises. Cytogenetic and Genome Research 163(3–4):223–230, 2023. DOI: 10.1159/000530947. https://doi.org/10.1159/000530947
  7. Ghandhi SA, Shuryak I, Morton SR, Amundson SA, Brenner DJ. New Approaches for Quantitative Reconstruction of Radiation Dose in Human Blood Cells. Scientific Reports 9:18441, 2019. DOI: 10.1038/s41598-019-54967-5. https://doi.org/10.1038/s41598-019-54967-5
  8. Shelke S, Das B. Dose response and adaptive response of non-homologous end joining repair genes and proteins in resting human peripheral blood mononuclear cells exposed to γ radiation. Mutagenesis 30(3):365–379, 2015. DOI: 10.1093/mutage/geu081. https://doi.org/10.1093/mutage/geu081
  9. Koval L, Proshkina E, Shaposhnikov M, Moskalev A. The role of DNA repair genes in radiation-induced adaptive response in Drosophila melanogaster is differential and conditional. Biogerontology 21(1):45–56, 2020. DOI: 10.1007/s10522-019-09842-1. https://doi.org/10.1007/s10522-019-09842-1
  10. Yushkova E, Bashlykova L. Transgenerational effects in offspring of chronically irradiated populations of Drosophila melanogaster after the Chernobyl accident. Environmental and Molecular Mutagenesis 62(1):39–51, 2021 (published online 5 December 2020). DOI: 10.1002/em.22416. https://doi.org/10.1002/em.22416
  11. Zarubin M, Pikalov V, Gangapshev A. BioMuon experiment (Bioμ): review of studies on biological impact of high-energy muons and research prospects. International Journal of Radiation Biology 101(12):1169–1180, 2025. DOI: 10.1080/09553002.2025.2560882. https://doi.org/10.1080/09553002.2025.2560882