Ion Linear Energy Transfer Research Articles

Radiation environment at the Moon: Comparisons of transport code modeling and measurements from the CRaTER instrument

AbstractThe Cosmic Ray Telescope for the Effects of Radiation (CRaTER), an instrument carried on the Lunar Reconnaissance Orbiter spacecraft, directly measures the energy depositions by solar and galactic cosmic radiations in its silicon wafer detectors. These energy depositions are converted to linear energy transfer (LET) spectra. High LET particles, which are mainly high‐energy heavy ions found in the incident cosmic ray spectrum, or target fragments and recoils produced by protons and heavier ions, are of particular importance because of their potential to cause significant damage to human tissue and electronic components. Aside from providing LET data useful for space radiation risk analyses for lunar missions, the observed LET spectra can also be used to help validate space radiation transport codes, used for shielding design and risk assessment applications, which is a major thrust of this work. In this work the Monte Carlo transport code HETC‐HEDS (High‐Energy Transport Code‐Human Exploration and Development in Space) is used to estimate LET contributions from the incident primary ions and their charged secondaries produced by nuclear collisions as they pass through the three pairs of silicon detectors. Also in this work, the contributions to the LET of the primary ions and their charged secondaries are analyzed and compared with estimates obtained using the deterministic space radiation code HZETRN 2010, developed at NASA Langley Research Center. LET estimates obtained from the two transport codes are compared with measurements of LET from the CRaTER instrument during the mission. Overall, a comparison of the LET predictions of the HETC‐HEDS code to the predictions of the HZETRN code displays good agreement. The code predictions are also in good agreement with the CRaTER LET measurements above 15 keV/µm but differ from the measurements for smaller values of LET. A possible reason for this disagreement between measured and calculated spectra below 15 keV/µm is an inadequate representation of the light ion spectra in HETC‐HEDS and HZETRN code calculations. It is also clear from the results of this work that Vavilov distributions need to be incorporated into the HETC‐HJEDS code before it will be able to recreate the observed LET spectra measured by the CRaTER instrument.

Distinct mechanisms of the inhibition of vasculogenesis by different species of ionizing particles

The human vasculature includes a vast network of microcapillaries networking the body and is a major target for non-carcinogenic effects of space radiation in the body. The brain microvascular system is crucial to healthy functioning of the brain and its pathology is not only a primary event in a range of neurodegenerative diseases but also an important influencing factor in many others. The vasculature is maintained by angiogenesis regenerating vessels as they are needed, this is particularly relevant if the blood–brain barrier is damaged by agents such as space radiation, thereby creating the need for angiogenic regeneration. The resulting lack of vasculature due to the inhibition of re-growth of vessels can, in turn, lead to a negative feedback loop and further pathologies.Using three-dimensional human vessel cultures with human umbilical vein and brain microvascular endothelial cells, we have developed assays that determine at what stage angiogenesis is inhibited by ionizing radiation. The relative biological effect of high linear energy transfer (LET) 1 GeV Fe ions compared with low LET 1 GeV protons is only one for developing vessels but greater than four for mature vessels. This action of low LET protons on developing vessels was surprisingly effective (50% inhibition with 40 cGy exposure) and together with the effect of high LET ions may represent a significant hazard in the space environment.The morphology of developing vessels 5 days after exposure showed significant differences that suggest distinct mechanisms of inhibition. Cells exposed to protons failed to make connections with other cells. Conversely, cells exposed to Fe ions extended cellular processes and made connections to other cells but did not develop a central lumen. The microtubule and actin cytoskeletons showed differences indicating that motility at the extending tips of endothelial cells is inhibited by protons but not Fe ions. Actin-rich protrusive structures that contain bundled and dynamic microtubules showed a 65% decrease when exposed to high-energy protons but not with the same dose of high-energy Fe ions. Since protein kinase C (PKC) has long been known to stimulate angiogenesis, we hypothesized that rescue of the capillary phenotype after proton exposure would be possible by stimulating PKC before irradiation. One-day-old vessel cultures were treated with 30 and 60 nM phorbol ester (PMA) 15 min before irradiation. Stimulation of PKC restored capillary formation in proton-treated cultures but not in Fe ion-treated cultures. More specifically, stimulation of PKC by PMA was able to restore the tip motility that was inhibited by low LET ions [ 1].Further studies with various charged particles showed that low LET ion particles (Proton and He ions) with an LET lower or equal to 1 keV/μm inhibit vasculogenesis in the same way as protons. Higher LET charged particles (Silicon 1GeV, Oxygen 250 MeV and 1 GeV and Carbon 290 MeV and 1 GeV) with an LET ≥8 keV/μm inhibit vasculogenesis in the same way as Fe ions.In conclusion, we have shown that low and high LET ions inhibit the formation of brain capillaries by different mechanisms. For low LET ions, inhibition involves regulation of PKC-dependent motile tips leading to a failure of cellular processes to migrate through the matrix and meet up with other processes. For high LET ions, the cells fail to complete angiogenesis by not migrating and forming tubular structures. This complexity of response opens up possibilities of greater control over angiogenesis and the resulting pathologies during coincident exposure or therapy. For exposure in space, knowledge of these mechanisms will enable more precise risk assessment and mitigation strategies. For radiotherapy, treatment could be manipulated to utilize the radiation effectively.

Visualization of cell-cycle modification by ionizing irradiation in single HeLa cells using fluorescent ubiquitination-based cell-cycle indicator

It has been known that cell cycle is retarded or arrested when the cells are exposed to ionizing radiation. The cell-cycle modifications are thought to be controlled by check point mechanisms to ensure the time for DNA repair. Linear energy transfer (LET) dependence of cell-cycle modifications, however, has not been fully revealed. Considerably less is known about detailed cell-cycle arrest for a single-cell level after exposure. Our purpose is to explore high LET radiation effects on the mammalian cell cycle. To examine high LET radiation effects on mammalian cell cycle, it would be essential to track single cells as live cell images observed by time-lapse imaging technique. HeLa cells expressing fluorescent ubiquitination-based cell-cycle indicator (FUCCI) are one of the useful model cell lines to visualize cell cycle because their nuclei show different colors; orange indicating G1 (Cdt1 expression); green indicating S/G2 (Geminin expression) [ 1]. In order to establish a novel assay system to study cell-cycle modification by high LET irradiation such as ion beams, we have developed time-lapse protocol for the HeLa-FUCCI cells irradiated.As a preliminary experiment using conventional X-rays instead of high LET ion beams, we observed the cell cycles of the irradiated HeLa cells. Figure 1 shows a typical time-lapse dynamics of unirradiated cells acquired for 48 h. We establish a new method to decide the time of one cell cycle as shown in Fig. 1. We obtained an evidence that the irradiated (5 Gy) cells show prolonged cell-cycle period when compared with that of control cells (Table 1). We also revealed that the delay is mainly caused in the G2 (Geminin expressing) phase. These results suggest that S/G2, G2/M or M checkpoint mechanism regulate the cell cycle in the irradiated HeLa-FUCCI cells. In conclusion, single FUCCI cell exposure and live cell imaging are a powerful method to trace the single-cell effect of high LET irradiation on the cell cycle in future. Fig. 1.Typical time-lapse dynamics of an unirradiated HeLa-FUCCI cell monitored by geminin fluorescent intensity of the cell nucleus. Table 1.Distribution of the number of cells showing a specific cell cycle period.Cell cycle period, hNumber of cellsControl (unirradiated)5 Gy irradiated>146214–168316–1812918–20221620–2215822–245824–265726–281628–301030 <08Total7567

Technology Scaling Comparison of Flip-Flop Heavy-Ion Single-Event Upset Cross Sections

Heavy-ion experimental results from flip-flops in 180-nm to 28-nm bulk technologies are used to quantify single-event upset trends. The results show that as technologies scale, D flip-flop single-event upset cross sections decrease while redundant storage node flip-flops cross sections may stay the same or increase depending on the layout spacing of storage nodes. As technology feature sizes become smaller, D flip-flop single-event upset cross sections approach redundant storage node hardened flip-flops cross sections for particles with high linear energy transfer values. Experimental results show that redundant storage node designs provide > 100X improvement in single-event upset cross section over DFF for ion linear energy transfer values below 10 MeV-cm2/mg down to 28-nm feature sizes.

In situ investigation of 75 MeV boron and 100 MeV oxygen ion irradiation effects on 50 GHz silicon–germanium heterojunction bipolar transistors

First-generation (50 GHz) silicon–germanium heterojunction bipolar transistors (SiGe HBT) were irradiated with 75 MeV boron ions and 100 MeV oxygen ions. The aim of the present investigation is to study the degradation of current–voltage (I–V) characteristics due to different linear energy transfer ions. The in situ I–V characteristics were measured before and after ion irradiation are forward mode and inverse mode Gummel characteristics, excess base current and current gain. It was found that the oxygen ion-irradiated SiGe HBT showed slightly more degradation when compared with boron ion-irradiated devices. The damage constant was calculated using the Messenger–Spratt equation, which confirmed more degradation in the case of oxygen ion-irradiated HBTs.

SU-E-T-297: Modeling the Cell Inactivation Process for High LET Particles by Using the Law of Mass Action

Purpose: To develop a method to model the cell inactivation caused by photons or light ions irradiation that accurately describes the survival curves for various linear energy transfer (LET) as well as high dose survival curves. Methods: We adapted the law of mass action widely used in physical chemistry to model the cell inactivation process through a differential equation. Its solution depends on two parameters, D0 and p. D0 is positive and represents the dose that causes a survival of 37%; p is a dimensionless parameter equal or greater to 1. We evaluated our model using data available in the literature for four human cancer cell lines (DU‐145, CP3, U1690 and H460) exposed to 137Cs and doses up to 20 Gy; and to V79 cells exposed to 240 kVp x‐rays and to light ions (protons, deuterons and 3He) of various LET ranging from 10 to 105 keV um−1. Results: Our model was capable to fit survival curves for the human cancer cell lines up to 20 Gy as good as the linear quadratic model. D0 was determined with uncertainty better than 4% and ranged from 2.7 to 3.8 Gy; p was determined with uncertainty better than 6% and ranged from 1.3 to 1.7. For the V79 data, the parameters Do and p exhibited a well‐defined exponential dependence on the LET. Also, the p was found to approach 1 asymptotically, as predicted by the theory. The RBE as a function of LET was also accurately determined. Conclusion: The adapted law of mass action for the radiobiology provides an accurate model to study the cell survival curves under various conditions. Furthermore, it sets useful limits to the parameters that could be verified experimentally. NCI grant P01CA021239

High-energy He ion bombardment-induced modifications of surface properties of UHMWPE

Ultra high-molecular-weight polyethylene (UHMWPE) is the material of choice for the load-bearing surfaces of total joint implants. This is primarily due to its good wear resistance. However, failure is often observed as a result of the UHMWPE brittleness. In order to improve its mechanical properties, UHMWPE was bombarded with 4-MeV He ions to fluence ranging from 1012 to 5 × 1014 ions/cm2. Changes of surface layer composition produced by ion bombardment of UHMWPE samples were studied. Polymer radiolysis due to the ion bombardment leading to hydrogen release was studied by means of nuclear reaction analysis using the 1H(15N, αθ)12C reaction. Hydrogen release increasing with the ion fluence was observed and correlated to the ion linear energy transfer. Moreover, important oxidation of bombarded samples was also noticed. These effects resulted in important modifications of the surface properties of bombarded material such as increase of hardness and improved wettability.

Biodamage via shock waves initiated by irradiation with ions

Radiation damage following the ionising radiation of tissue has different scenarios and mechanisms depending on the projectiles or radiation modality. We investigate the radiation damage effects due to shock waves produced by ions. We analyse the strength of the shock wave capable of directly producing DNA strand breaks and, depending on the ion's linear energy transfer, estimate the radius from the ion's path, within which DNA damage by the shock wave mechanism is dominant. At much smaller values of linear energy transfer, the shock waves turn out to be instrumental in propagating reactive species formed close to the ion's path to large distances, successfully competing with diffusion.

Comparison of cellular lethality in DNA repair-proficient or -deficient cell lines resulting from exposure to 70 MeV/n protons or 290 MeV/n carbon ions

Charged particle therapy utilizing protons or carbon ions has been rapidly intensifying over recent years. The present study was designed to jointly investigate these two charged particle treatment modalities with respect to modeled anatomical depth-dependent dose and linear energy transfer (LET) deliveries to cells with either normal or compromised DNA repair phenotypes. We compared cellular lethality in response to dose, LET and Bragg peak location for accelerated protons and carbon ions at 70 and 290 MeV/n, respectively. A novel experimental live cell irradiation OptiCell™ in vitro culture system using three different Chinese hamster ovary (CHO) cells as a mammalian model was conducted. A wild-type DNA repair-competent CHO cell line (CHO 10B2) was compared to two other CHO cell lines (51D1 and xrs5), each genetically deficient with respect to one of the two major DNA repair pathways (homologous recombination and non-homologous end joining pathways, respectively) following genotoxic insults. We found that wild-type and homologous recombination-deficient (Rad51D) cellular lethality was dependent on both the dose and LET of the carbon ions, whereas it was only dependent on dose for protons. The non-homologous end joining deficient cell line (Ku80 mutant) showed nearly identical dose-response profiles for both carbon ions and protons. Our results show that the increasingly used modality of carbon ions as charged particle therapy is advantageous to protons in a radiotherapeutic context, primarily for tumor cells proficient in non-homologous end joining DNA repair where cellular lethality is dependent not only on the dose as in the case of more common photon therapeutic modalities, but more importantly on the carbon ion LETs. Genetic characterization of patient tumors would be key to individualize and optimize the selection of radiation modality, clinical outcome and treatment cost.

Yield of Single- and Double-Strand Breaks and Nucleobase Lesions in Fully Hydrated Plasmid DNA Films Irradiated with High-LET Charged Particles

We measured the yield and spectrum of strand breaks and nucleobase lesions produced in fully hydrated plasmid DNA films to determine the linear energy transfer (LET) dependence of DNA damage induced by ion-beam irradiation in relation to the change in the atomic number of ions. The yield of isolated damage was revealed as a decrease in prompt SSBs with increasing LET of He(2+), C(5+,6+) and Ne(8+,10+) ions. On the other hand, the yields of prompt DSBs increased with increasing ion LET. SSBs were additionally induced in ion-irradiated DNA film by treatment with two kinds of base excision repair proteins (glycosylases), Nth and Fpg, indicating that base lesions are produced in the hydrated DNA film. This result shows that nucleobase lesions are produced via both chemical reactions with diffusible water radicals, such as OH radicals, and direct energy deposition onto DNA and the hydrated water layer. Nth-sensitive sites deduced to be pyrimidine lesions, such as 5,6-dihydrothymine (DHT), showed a relatively larger yield than Fpg-sensitive sites deduced to be purine lesions, such as 7,8-dihydro-8-oxo-2'deoxyguanine (8-oxoGua), for all ion exposures tested. The yield of SSBs or DSBs observed by enzyme treatment decreased noticeably with increasing LET for all tested ions. These results indicated that higher-LET ions preferentially produce a complex type of damage that might compromise the activities of the glycosylases used in this study. These findings are biologically important since, under cell mimicking conditions, persistent DNA damage occurs in part due to direct energy deposition on the DNA or hydrated water shell that is specifically induced by dense ionization in the track.

Relative biological effectiveness of high linear energy transfer α-particles for the induction of DNA-double-strand breaks, chromosome aberrations and reproductive cell death in SW-1573 lung tumour cells

Ionizing radiation-induced foci (IRIF) of DNA repair-related proteins accumulated at DNA double-strand break (DSB) sites have been suggested to be a powerful biodosimetric tool. However, the relationship between IRIF induction and biologically relevant endpoints, such as cell death and formation of chromosome rearrangements is less clear, especially for high linear energy transfer (LET) radiation. It is thus not sufficiently established whether IRIF are valid indicators of biological effectiveness of the various radiation types. This question is more significant in light of the recent advancements in light ion-beam and radionuclide therapy. Dose-effect relationships were determined for the induction of DNA-DSBs, chromosome aberrations and reproductive cell death in cultured SW-1573 cells irradiated with γ-rays from a Cs-137 source or with α-particles from an Am-241 source. Values of relative biological effectiveness (RBE) of the high LET α-particles were derived for these effects. DNA-DSB were detected by scoring of γ-H2AX foci, chromosome aberrations by fragments and translocations using premature chromosome condensation and cell survival by colony formation. Analysis of dose-effect relations was based on the linear-quadratic model. Except for the survival curves, for other effects no significant contribution was derived of the quadratic term in the range of doses up to 2Gy of γ-rays. Calculated RBE values derived for the linear component of dose-effect relations for γ-H2AX foci, cell reproductive death, chromosome fragments and colour junctions are 1.0±0.3, 14.7±5.1, 15.3±5.9 and 13.3±6.0, respectively. RBE values calculated at a certain biological effect level are 1, 4, 13 and 13, respectively. The RBE values derived from the LQ model are preferred as they are based on clinically relevant doses. The results show that with low LET radiation only a small fraction of the numerous DNA-DSBs yield chromosome damage and reproductive cell death. It is concluded that many of the chromosomal aberrations detected by premature chromosome condensation do not cause reproductive cell death. Furthermore, RBE values for DNA-DSB detectable by γ-H2AX foci shortly after irradiation, provide no information relevant to applications of high LET radiation in radiotherapy. The RBE values of chromosome aberrations assessed by premature chromosome condensation are close to the value for reproductive cell death. This suggests possible relevance to assess RBE values for radiotherapy with high LET ions.

Monte Carlo simulation of energy-deposit clustering for ions of the same LET in liquid water

This work presents a Monte Carlo study of energy depositions due to protons, alpha particles and carbon ions of the same linear-energy-transfer (LET) in liquid water. The corresponding track structures were generated using the Geant4-DNA toolkit, and the energy deposition spatial distributions were analyzed using an adapted version of the DBSCAN clustering algorithm. Combining the Geant4 simulations and the clustering algorithm it was possible to compare the quality of the different radiation types. The ratios of clustered and single energy depositions are shown versus particle LET and frequency-mean lineal energies. The estimated effect of these types of radiation on biological tissues is then discussed by comparing the results obtained for different particles with the same LET.

Variable Depth Bragg Peak Method for Single Event Effects Testing

The Variable Depth Bragg Peak (VDBP) method for measuring the Single Event Effects (SEE) cross-section of an integrated circuit (IC) in a closed package as a function of ion linear energy transfer (LET) is described. The method uses long-range, high-energy heavy ions that can penetrate the package and deposit charge in the device's sensitive volume (SV), the depth of which is not known. A series of calibrated energy degraders is used to vary the depth of the Bragg peak relative to the device's sensitive volume. When the Bragg peak is located at the sensitive volume, the measured SEE cross-section is a maximum, as is the LET, which is calculated using a Monte Carlo-based program, TRIM that takes both straggling and spread in beam energy and angle into account. Degrader thickness is varied and the change in LET is calculated while the corresponding cross-section is measured. Good agreement was obtained between the LET-dependence of the single event upset (SEU) cross-section for a 4 Mbit memory in an unopened package using the above method and that for an identical de-lidded part previously measured.

A Monte Carlo Study of the Radiation Quality Dependence of DNA Fragmentation Spectra

We simulated the irradiation of human fibroblasts with gamma rays, protons and helium, carbon and iron ions at a fixed dose of 5 Gy. The simulations were performed with the biophysical Monte Carlo code PARTRAC. From the output of the code, containing in particular the genomic positions of the radiation-induced DNA double-strand breaks (DSBs), we obtained the DNA fragmentation spectra. Very small fragments, in particular those related to "complex lesions" (few tens of base pairs), are probably very important for the late cellular consequences, but their detection is not possible with the common experimental techniques. We paid special attention to the differences among the various ions in the production of these very small fragments; in particular, we compared the fragmentation spectra for ions of the same specific energy and for ions of the same LET (linear energy transfer). As found previously for iron ions, we found that the RBE (relative biological effectiveness) for DSB production was considerably higher than 1 for all high-LET radiations considered. This is at variance with the results obtainable from experimental data, and it is due to the ability to count the contribution of small fragments. It should be noted that for a given LET this RBE decreases with increasing ion charge, due mainly to the increasing mean energy of secondary electrons. A precise quantification of the DNA initial damage can be of great importance for both radiation protection, particularly in open-space long-term manned missions, and hadrontherapy.

Radiobiologic Parameters and Local Effect Model Predictions for Head-and-Neck Squamous Cell Carcinomas Exposed to High Linear Energy Transfer Ions

To establish the radiobiologic parameters of head-and-neck squamous cell carcinomas (HNSCC) in response to ion irradiation with various linear energy transfer (LET) values and to evaluate the relevance of the local effect model (LEM) in HNSCC. Cell survival curves were established in radiosensitive SCC61 and radioresistant SQ20B cell lines irradiated with [33.6 and 184 keV/n] carbon, [302 keV/n] argon, and X-rays. The results of ion experiments were confronted to LEM predictions. The relative biologic efficiency ranged from 1.5 to 4.2 for SCC61 and 2.1 to 2.8 for SQ20B cells. Fixing an arbitrary D(0) parameter, which characterized survival to X-ray at high doses (>10 Gy), gave unsatisfying LEM predictions for both cell lines. For D(0) = 10 Gy, the error on survival fraction at 2 Gy amounted to a factor of 10 for [184 keV/n] carbon in SCC61 cells. We showed that the slope (s(max)) of the survival curve at high doses was much more reliable than D(0). Fitting s(max) to 2.5 Gy(-1) gave better predictions for both cell lines. Nevertheless, LEM could not predict the responses to fast and slow ions with the same accuracy. The LEM could predict the main trends of these experimental data with correct orders of magnitude while s(max) was optimized. Thus the efficiency of carbon ions cannot be simply extracted from the clinical response of a patient to X-rays. LEM should help to optimize planning for hadrontherapy if a set of experimental data is available for high-LET radiations in various types of tumors.

Enhanced Degradation in Power MOSFET Devices Due to Heavy Ion Irradiation

Large, unexpected shifts in the current-voltage (IV) characteristics of commercial power MOSFETs irradiated with heavy ions have been observed. The shifts can be more than sixty-five times larger than the shifts resulting from total dose irradiation with gamma rays or electrons, and are shown to strongly depend on both the irradiation bias and the ion linear energy transfer (LET). These large shifts are a significant concern for devices intended to operate in low power space applications because it is shown that they can lead to off-state leakage currents greater than 1 A. The data are consistent with the formation of parasitic transistors resulting from the microdose deposited in the gate oxides of these devices by the heavy ions. These results have significant implications for hardness assurance testing of MOS devices for use in space.

Model for Cumulative Solar Heavy Ion Energy and Linear Energy Transfer Spectra

A probabilistic model of cumulative solar heavy ion energy and LET spectra is developed for spacecraft design applications. Spectra are given as a function of confidence level, mission time period during solar maximum and shielding thickness. It is shown that long-term solar heavy ion fluxes exceed galactic cosmic ray fluxes during solar maximum for shielding levels of interest. Cumulative solar heavy ion fluences should therefore be accounted for in single event effects rate calculations and in the planning of space missions.

Spectroscopic properties of novel fluorescent nuclear track detectors for high and low LET charged particles

A new optical, non-destructive method of detecting and imaging individual heavy charged particle tracks using fluorescent nuclear track detectors (FNTD) is investigated as a possible spectroscopic technology for heavy charged particles of low and high linear energy transfer (LET). The technique uses new luminescent aluminum oxide single crystals having aggregate oxygen vacancy defects and doped with Mg ( Al 2 O 3 : C , Mg ) in combination with laser scanning confocal fluorescence microscopy. Spectroscopic capabilities of this new method were demonstrated for energetic heavy ions of LET ∞ H 2 O ranging from 2.2 to 8767 keV / μ m . The new technology is intended for use in neutron detection and dosimetry, proton and heavy ion radiobiology, and space radiation dosimetry as well as in nuclear and particle physics research.

Single-ion dosemeter based on floating gate memories

Floating Gate (FG) nonvolatile memories are based on a tiny polysilicon layer (the FG) which can be permanently charged with electrons or holes, thus changing the threshold voltage of a MOSFET. Every time a FG is hit by a high energy ion, it experiences a charge loss, depending on the ion linear energy transfer (LET) and on the transistor geometrical and electrical characteristics. This paper discusses the opportunities to use this devices as single an ion dosemeter with sub-micrometer spatial resolution and capable of distinguish the impinging ion LET.

Contribution of Indirect Action to Radiation-Induced Mammalian Cell Inactivation: Dependence on Photon Energy and Heavy-Ion LET

The contribution of indirect action mediated by OH radicals to cell inactivation by ionizing radiations was evaluated for photons over the energy range from 12.4 keV to 1.25 MeV and for heavy ions over the linear energy transfer (LET) range from 20 keV/microm to 440 keV/microm by applying competition kinetics analysis using the OH radical scavenger DMSO. The maximum level of protection provided by DMSO (the protectable fraction) decreased with decreasing photon energy down to 63% at 12.4 keV. For heavy ions, a protectable fraction of 65% was found for an LET of around 200 keV/microm; above that LET, the value stayed the same. The reaction rate of OH radicals with intracellular molecules responsible for cell inactivation was nearly constant for photon inactivation, while for the heavy ions, the rate increased with increasing LET, suggesting a reaction with the densely produced OH radicals by high-LET ions. Using the protectable fraction, the cell killing was separated into two components, one due to indirect action and the other due to direct action. The inactivation efficiency for indirect action was greater than that for direct action over the photon energy range and the ion LET range tested. A significant contribution of direct action was also found for the increased RBE in the low photon energy region.