The shape of the radiation dose response for DNA double-strand break induction and repair

Abstract

DNA double-strand breaks are among the most deleterious lesions induced by ionising radiation. A range of inter-connected cellular response mechanisms has evolved to enable their efficient repair and thus protect the cell from the harmful consequences of un- or mis-repaired breaks which may include early effects such as cell killing and associated acute toxicities and late effects such as cancer. A number of studies suggest that the induction and repair of double-strand breaks may not always occur linearly with ionising radiation dose. Here we have aimed to identify and discuss some of the biological and methodological factors that can potentially modify the shape of the dose response curve obtained for these endpoints using the most common assays for double-strand breaks, pulsed-field gel electrophoresis and microscopic scoring of radiation-induced foci.

Figure 1

Schematic dose response and time course for different classes of signals that contribute to DSB measurements. Top: pulsed-field gel electrophoresis (PFGE). Bottom: fluorescence microscopic gamma-H2AX foci scoring. Note that the graphs are for illustrative purposes only. The values shown should not be taken as representative of the ‘typical’ contributions as they depend on numerous experimental factors.

Figure 2

The impact of S-phase DNA on pulsed-field electrophoretic DSB measurements. Nocodazole-synchronised chicken DT40 pre-B cells were analysed 0–18 hours after removal of the drug. (A) Flow cytometric estimates of the S phase fraction. (B) Pulsed-field gel images of DNA migration following 40 and 10 Gy X-irradiation without repair incubation. (C): Fraction of DNA released (FAR) as a function of time after nocodazole removal. FAR values are inversely correlated to the fraction of S phase cells shown in the top diagram.

Figure 3

gamma-H2AX and 53BP1 foci induction by X-rays. (A) Immunofluorescence microscopy images were taken at 0.5h following X-irradiation of normal human fibroblasts. Each image is 20 μm wide. In the merged images 53BP1 is red, gamma-H2AX green and the nuclear margins are shown in blue. Co-localising foci appear yellow or orange. (B) Colocalisation analysis of gamma-H2AX and 53BP1 foci. Pearson’s correlation coefficients were calculated as described in [43]. A value of one represents total co-localisation. The significance of correlation coefficients was determined for individual cells using Costes’ spatial statistics approach [44]. Each point represents one cell. Filled red circles: non-significant, open green circles: significant correlation. Blue triangles, connected by blue line: mean correlation coefficient; error bars are standard errors from the analysis of 10–20 cells for each dose. (C) Gamma-H2AX versus 53BP1 foci count per cell, manually scored in the same 1,000 double-immunostained cells following exposure to a range of X-ray doses. Each data point corresponds to one cell. Shading of data point symbols reflects the number of coinciding points. The blue line indicates a 1:1 ratio.