Highlight: Track-structure based evaluation of DNA damage by light ions

Mission: Assessment of personalized health risks from radiation exposure for clinics and population

The Research Group Radiation Risk investigates health risks from exposures to ionizing radiation. Models of carcinogenesis and cardiovascular diseases are developed and applied for assessment of radiation-induced risk. These models can be used in all areas of exposure to ionising radiation.  Research is focused on risk assessment for medical therapeutic and diagnostic procedures.

Highlight: Track structure based evaluation of DNA damage by light ions in the energy range used for radiation therapy

Radiation effects of ions over a wide range of energies are of paramount interest in particle radiotherapy (RT). The treatment planning has to account primarily for the distribution of the deposited dose. For ions also the increased relative biological effectiveness (RBE) has to be taken into account; clear understanding of RBE is essential for optimal use of ion RT. Radiation transport model calculations have proven their suitability for dose determination in ion RT on spatial scales down to mm. The RBE of ion radiation qualities, however, is essentially related to damage distributions on nm- and µm-scales. In this domain track-structure based simulations are the method of choice, and PARTRAC is there one of the leading biophysical Monte Carlo modelling tools. At low energies where the stopping power of ions gives rise to a maximum (the so-called Bragg peak), however, the availability of cross sections has limited such calculations to energies above about 1 MeV/u; simulations below this limit have been performed for H, He and C only. To overcome this limitation, PARTRAC has been extended by a new cross-section scaling scheme. It accounts for the effect of charge-changing processes, and yields track structures with stopping power values corresponding to established standards, as shown in Figure 1, down to energies around 10 keV/u where interactions with nuclei in the target material become important.


Figure 1: Stopping power of ions with 0.25 MeV/u (thin histogram lines) or 0.5 MeV/u initial energy (thick histogram lines) in PARTRAC calculations compared to standard data from ICRU73 represented by symbols (0.25 MeV/u: empty squares, 0.5 MeV/u: filled squares). PARTRAC results for C and Ne refer to the new scaling scheme, data for H and He are based on element-specific cross sections by Dingfelder et al. Data for H and 0.5 MeV/u He show the so-called Bragg peak, i.e. a maximum in the stopping power near the track end. After the maximum the stopping power decreases forming the distal branch of the Bragg peak; correspondingly, the proximal branch refers to the increase of the stopping power up to its maximum.

Initial DNA damage has been assessed with PARTRAC both qualitatively and quantitatively in models of human lymphocyte and fibroblast nuclei for beams of H, He, C, N, O and Ne ions over the RT relevant energy range from 256 MeV/u down to 0.25 MeV/u; recently supplemented with calculations for Li, Be and B ions. This energy range corresponds to penetration depths from about 30 cm down to a few micrometers where the ions are stopped inside the cell nucleus. Figure 2 illustrates the setup of the simulations for a spherical lymphocyte nucleus.


Figure 2: Setup of the DNA damage simulations. Ions start from random positions of a circular source in normal direction; the source layer is tangential to the spherical cell nucleus and randomly rotated around it. DNA in the cell nucleus is represented by chromosomes in different colours. Arrows represent carbon ions starting with 0.25 MeV/u energy; they are stopped inside the nucleus after ~6 µm penetration depth (see Figure 1). The orange part of the arrows contributes to the dose absorbed by the cell nucleus. The investigation of the radiation damage by slow ions (see Figure 4) refers to slabs of 200 nm thickness.

Event-by-event calculations of energy depositions by ionizations and excitations within the penetrated medium are superimposed with a comprehensive model of the genomic DNA in a human cell nucleus to determine DNA damage by direct effects. Moreover, the formation of reactive species, their diffusion and their reactions in water surrounding the DNA are traced to evaluate DNA damage from indirect effects via OH radical attack. DNA damage from both direct and indirect effects has been analysed in terms of DNA strand breaks, double-strand breaks (DSB), clustering of DSB and formation of DNA fragments from a few ten up to millions of base pairs. Comparisons between measurements and simulation results have to take into account the experimental limitations in resolving DSB in close vicinity. Figure 3 shows measured DSB yields from analysis of DNA fragments in the size range 6 kbp – 5 Mbp and PARTRAC simulation results for DSB associated with DNA fragments in this length interval.      

Figure 3: Yield of DNA double-strand breaks (DSB) in dependence of the LET of the incident ions. Experimental data from Höglund et al. 2000 (symbols) and calculated results (lines) refer both to DSB associated with DNA fragments in the size interval from 6 kbp up to 5 Mbp. Measurements and simulations are in good agreement which is even better regarding LET-dependent relative biological effectiveness: all experimental results are slightly below corresponding simulations.

For the lowest energies, where ion’s stopping power largely varies along the transport distance and eventually decreases inside the nucleus (see Figure 1), the DNA damage pattern has been analyzed differentially in dependence on local LET (which is equal to the stopping power) in slabs of 200 nm thickness (see Figure 2). Local LET values on the distal branch of the Bragg peak have a counterpart on the increasing (proximal) branch of the Bragg peak which is for higher ion energies largely invariant over nuclear scales and can be determined as nucleus-averaged LET. Thus, DNA damage as a function of local LET on the distal branch of the Bragg peak at the very track end can be compared to damage at the same LET for the same ion on the proximal branch at higher particle energies. This comparison is shown in Figure 4 for DSB induction due to H, He, C and Ne ions with 0.25 MeV/u initial energy for effects on the distal branch. For H the local increase of the DSB yields corresponding to maxima for heavier ions is seen which is smeared out to about 20% lower values when averaged over the nucleus. For He ions, the LET-dependent DSB yields are largely equal on both sides of the Bragg peak, for the other ions the effects on the distal branch tend to be lower than on the proximal branch.

Figure 4: Yield of total DSB in dependence of the LET of the incident ions at low energies (on the distal side of the Bragg peak, solid lines) compared to the yields at higher energies (on the proximal side of the Bragg peak, dashed lines). For He the DSB yields are largely equal at the same LET value on both sides of the Bragg peak. For the other ions lower DSB yields have been found for lower particle energies on the distal side of the Bragg peak.


Friedland, W., Schmitt, E., Kundrát, P., Dingfelder, M., Baiocco, G., Barbieri, S., Ottolenghi, A. (2017) Comprehensive track-structure based evaluation of DNA damage by light ions from radiotherapy-relevant energies down to stopping. Scientific Reports 7:45161