The radiation adaptive response and priming dose influence: the quantification of the Raper-Yonezawa effect and its three-parameter model for postradiation DNA lesions and mutations

Abstract

The priming dose effect, called also the Raper-Yonezawa effect or simply the Yonezawa effect, is a special case of the radiation adaptive response phenomenon (radioadaptation), which refers to: (a) faster repair of direct DNA lesions (damage), and (b) DNA mutation frequency reduction after irradiation, by applying a small priming (conditioning) dose prior to the high detrimental (challenging) one. This effect is observed in many (but not all) radiobiological experiments which present the reduction of lesion, mutation or even mortality frequency of the irradiated cells or species. Additionally, the multi-parameter model created by Dr. Yonezawa and collaborators tried to explain it theoretically based on experimental data on the mortality of mice with chronic internal irradiation. The presented paper proposes a new theoretical approach to understanding and explaining the priming dose effect: it starts from the radiation adaptive response theory and moves to the three-parameter model, separately for two previously mentioned situations: creation of fast (lesions) and delayed damage (mutations). The proposed biophysical model was applied to experimental data-lesions in human lymphocytes and chromosomal inversions in mice-and was shown to be able to predict the Yonezawa effect for future investigations. It was also found that the strongest radioadaptation is correlated with the weakest cellular radiosensitivity. Additional discussions were focussed on more general situations where many small priming doses are used.

Keywords: Adaptive response; Cancer physics; Challenging dose; Lymphocytes; Priming dose; Radiation; Radiation biophysics; Radioadaptation; Radiosensitivity; Yonezawa effect.

Fig. 1

The scheme of the Yonezawa effect (also called the priming dose effect or the Raper-Yonezawa effect): the single small dose (D₁) generates much less mutations than the single high dose (D₂). However, when D₂ follows D₁ with some time distance between them (Δt), the mutation (or lesion) frequency for that D₁ + D₂ total dose is lower than for single D₂ by the exemplary value of δ = 0.73. This exemplary result was obtained using the input data: D₁ = 1 UAD (Unit of Absorbed Dose), D₂ = 5 UAD, Δt = 3 UT (Unit of Time), α₀ = 1 [UT⁻³ UAD⁻²], α₁ = 1 [UAD⁻¹], α₂ = 0.7 [UT⁻¹]. The parameter δ is therefore showing the percentage difference between the number of mutations (or lesions) generated by the single dose D₂ (without the priming dose) and the combination of D₁ + D₂

Fig. 2

The time-dependent relations of the probability distribution of the adaptive response, p_AR, and the number of postradiation lesions, N(t), given by Eq. (2) and Eqs. (21)–(22), respectively. Plot a presents a scenario where a single (reference) dose D₂ is applied; one can observe the hunchbacked shape of the p_AR function and the decrease in the number of lesions, N(t). Plot b presents a scenario with a combination of doses D₁ + D₂, where the priming dose D₁ is given prior to the challenging dose D₂ (the latter received in the same moment as in the plot a)); one can observe that the priming dose increased the probability of successful repair (enhanced value of p_AR) which causes a stronger decrease in the number of lesions, N(t). All input data were the same as in Fig. 1. The variable t is the global time where t = 0 corresponds to the moment of D₂

Fig. 3

The residual DNA damage (lesions) represented by double-strand breaks (DSB) in human lymphocytes related to time, after X-ray irradiation of 2 Gy. Four different relationships represents four different sensitivities to radiation, from hyper-radiosensitivity (upper grey curve) to radioresistance (lower black curve)—this last case represents the strongest adaptive response effect (Fornalski 2019). The figure was created based on the paper by Müller et al. (2001) and presentation of Feinendegen (2012)

Fig. 4

The case from the Fig. 2 with additional priming dose D₂ = D₁, received Δt₁ = 3 UT after D₁. The challenging dose D₃ was received Δt₂ = 2 UT after D₂

Fig. 5

The relationship between the delta (δ) parameter for DNA lesions and the priming dose (D₁) for human lymphocytes (Shadley and Wolff 1987), where D₂ = 1.5 Gy, Δt = 16 h, and T = 22 h. The Yonezawa effect disappears above approx. D₁ = 200 mGy. The dashed line represents the potential purely empirical trend (best fit with R² ≈ 0.997) given here by unsymmetrical Gaussian function: f(x) = exp(− (x − a)²/b) + c + (x − a)^(1/d), where a = − 0.036, b = 0.067, c = − 1.12, and d = 3.668

Fig. 6

The relationship between the delta (δ) parameter for DNA lesions and the time interval between priming and challenging dose (Δt) for human lymphocytes (Shadley et al. 1987), where D₁ = 10 mGy, D₂ = 1.5 Gy, and T − Δt = 6 h. The dashed line represents the assumed potential linear trend (the best fit by δ = − 0.005 [h⁻¹] Δt + 0.487 with R² ≈ 0.60) according to which the Yonezawa effect disappears above Δt ≈ 100 h. The straight line was selected as the simplest best fit according to the existing scatter of data points. Due to potential outliers, this fitting was also tested by the robust Bayesian regression method (Fornalski et al. 2010) but the result was practically the same (δ = − 0.005 [h⁻¹] Δt + 0.483)

Fig. 7

The non-normalized probability functions of radiation adaptive response in human lymphocytes (Shadley et al. 1987) in phase G₀–G₁ (blue solid line), in phase G₁–S (orange dashed line) and after phase S (green dotted line), related to time (hours). The orange dashed line corresponds to the lowest radiosensitivity of the cell and therefore the strongest radioadaptation (color figure online)

Fig. 8

a The shape of delta (δ) parameter function from Eq. (9) for the exemplary sets of parameters and their ranges: α₀ = 36.21 Gy⁻² h⁻³, α₁ from 20 to 1200 Gy⁻¹, α₂ from 0.02 to 0.23 h⁻¹, D₁ = 10 mGy, D₂ = 1.5 Gy, Δt = 34 h; plot b represents the same situation but two parameters’ ranges were narrowed: α₁ from 50 to 550 Gy⁻¹, and α₂ from 0.08 to 0.1 h⁻¹

Fig. 9

The flow chart of the Simplified Genetic Algorithm (SGA) used to evaluate parameters α₀, α₁ and α₂