A Simulation Approach for Determining the Spectrum of DNA Damage Induced by Protons

TL;DR

This study uses Geant4-DNA Monte Carlo simulations to analyze DNA damage caused by protons, revealing how damage frequency varies with proton energy and aligning well with experimental data.

Contribution

It introduces a detailed simulation framework for DNA damage by protons, incorporating both direct and indirect damage mechanisms with a comprehensive DNA model.

Findings

Over 50% of energy depositions cause strand breaks below 2 MeV

Double-strand break frequency is highly sensitive to proton energy

Simulation results align well with experimental data and other models

Abstract

In order to study the molecular damage induced in the form of single-strand and double-strand breaks by the ionizing radiation at the DNA level, Geant4-DNA Monte Carlo simulation code for complete transportation of primary protons and other secondary particles in liquid water have been employed in this work. To this aim, a B-DNA model and a thorough classification of DNA damage concerning their complexity were used. Strand breaks were assumed to have been primarily originated by direct physical interactions via energy depositions, assuming a threshold energy of 17.5 eV, or indirect chemical reactions of hydroxyl radicals, assuming a probability of 0.13. The simulation results on the complexity and frequency of various damage are computed for proton energies of 0.5 to 20 MeV. The yield results for a cell (Gy.cell)-1 are presented, assuming 22 chromosomes per DNA and a mean number of 245 Mbp per chromosome. The results show that for proton energies below 2 MeV, more than 50% of energy depositions within the DNA volume resulted in strand breaks. For double-strand breaks (DSBs), there is considerable sensitivity of DSB frequency to the proton energy. A comparison of DSB frequencies predicted by different simulations and experiments is presented as a function of proton LET. It is shown that, generally, our yield results (Gy.Gbp)-1 are comparable with various experimental data and there seems to be a better agreement between our results and a number of experimental studies, as compared to some other simulations.