Distinct Roles of Ape1 Protein, an Enzyme Involved in DNA Repair, in High or Low Linear Energy Transfer Ionizing Radiation-induced Cell Killing*

Background: High LET radiation-induced DNA DSBs interfere with only NHEJ. Results: Ape1 enzyme modification affects cell sensitivity to high LET but not to low LET radiation. Conclusion: Ape1 promotes processing of clustered DNA damage into DSBs in high LET-irradiated cells. Significance: The results will help to improve high LET radiotherapy or high LET radiation protection. High linear energy transfer (LET) radiation from space heavy charged particles or a heavier ion radiotherapy machine kills more cells than low LET radiation, mainly because high LET radiation-induced DNA damage is more difficult to repair. Relative biological effectiveness (RBE) is the ratio of the effects generated by high LET radiation to low LET radiation. Previously, our group and others demonstrated that the cell-killing RBE is involved in the interference of high LET radiation with non-homologous end joining but not homologous recombination repair. This effect is attributable, in part, to the small DNA fragments (≤40 bp) directly produced by high LET radiation, the size of which prevents Ku protein from efficiently binding to the two ends of one fragment at the same time, thereby reducing non-homologous end joining efficiency. Here we demonstrate that Ape1, an enzyme required for processing apurinic/apyrimidinic (known as abasic) sites, is also involved in the generation of small DNA fragments during the repair of high LET radiation-induced base damage, which contributes to the higher RBE of high LET radiation-induced cell killing. This discovery opens a new direction to develop approaches for either protecting astronauts from exposure to space radiation or benefiting cancer patients by sensitizing tumor cells to high LET radiotherapy.

Examples of high linear energy transfer (LET) 3 ionizing radiation (IR) include carbon ions used for cancer radiotherapy and cosmic rays present naturally in the space environment (1). Compared with low LET radiation, such as x-rays or ␥-rays, high LET radiation kills more cells at a given dose. The relative biological effectiveness (RBE), a ratio of the biological effects generated by high LET radiation to low LET reference radiation, ranges from 2 to 6 in mammalian cell killing. IR kills cells primarily by inducing DNA double strand breaks (DSBs). Because high LET and low LET radiation generate similar numbers of initial DSBs per unit of absorption, the RBE for high LET radiation in cell killing arises from the fact that DSBs are more difficult to repair (2,3); however, the underlying causes are not fully understood.
NHEJ is one of two major DSB repair pathways in mammalian cells. It is a fast, simple, and cell cycle-independent process (4). The other major pathway, homologous recombination repair (HRR) is slow and more complex. It ordinarily requires a sister chromatid as the donor template, and, therefore, occurs primarily in S and G 2 phases of the cell cycle (5). Interestingly, in an NHEJ-deficient genetic background, where HRR is the main mechanism for repair, the RBE for high LET radiation is close to 1 (6 -8), suggesting that HRR is not subject to inhibition by whatever mechanism interferes with NHEJ in high LETirradiated cells. This has been demonstrated in vivo (9) and confirmed by another group (10).
We have proposed that the small DNA fragments (Յ40 bp) generated by high LET radiation (possibly by penetrating nucleosomes) are too small for Ku to efficiently bind to the two ends of the fragment at the same time and therefore inhibit the NHEJ pathway of DNA DSB repair (8,9). However, the 1.8-fold increase in small DNA fragments in high LET-irradiated cells versus low LET-irradiated cells (8) is insufficient to fully explain an RBE of 2-6 in mammalian cells. In looking for additional mechanisms, we considered the possibility of enzymatic modification at the clustered DNA damage sites because the enzymatic modification contributes to IR-induced total DSBs (11), and clustered DNA damage is more prevalent in high LET-than in low LET-irradiated cells (12,13). * This work was supported, in whole or in part, by National Institutes of Health Grant P30CA138292. This work was also supported by National Aeronautics and Space Administration Grant NNX11AC39G. 1 Both authors contributed equally to this work. 2  The concept of clustered DNA damage was introduced by Goodhead et al. (14) and Ward (15), who proposed that passage of a radiation track through the nucleus results in localized clusters of single strand breaks, DSBs, and oxidized bases (16,17). Repair of base damage is initiated by glycosylase (mainly Ogg1)-induced apurinic/apyrimidinic (AP) sites that are then cleaved by AP endonuclease (Ape1). Ape1 is an essential gene in mammalian cells and plays a protective role in low LET-irradiated cells (18,19), although it can also generate DSBs by dual cutting of closely opposed AP sites (19,20). Therefore, processing of clustered damage sites by Ogg1 and Ape1 may easily produce small DNA fragments. Because no practical physical materials can efficiently shield cells from high LET radiation, the investigation into whether enzyme modification contributes to the RBE of high LET radiation in cell killing will be of importance in reducing cell damage from high LET radiation exposure. In addition, understanding the contribution of enzyme modification to the RBE of high LET radiation in cell killing could elucidate ways to sensitize tumor cells to high LET radiotherapy.

EXPERIMENTAL PROCEDURES
Plasmid Construction-The mouse Ogg1 cDNA was cloned with the proper primers (Table 1) from C57BL/6J total RNA based on the reference transcript sequence ENSMUST00000032406. The 1.1-kb RT-PCR product was recovered and inserted into the EcoRI and BglII sites of the pCMV-HA expression vector (purchased from Clontech). The wild type, enzymatically overactivated or inactivated human Ape1 was amplified by PCR using plasmid pCMV6-XL5 hAPEX1 (purchased from Origene Inc.) as template with proper primers ( Table 1). The wild type, enzymatically overactivated or inactivated mouse Ape1 was amplified by PCR using plasmid pCMV6-kan/neo mAPEX1 (purchased from Origene Inc.) as template with proper primers ( Table 1). The PCR products were recovered and inserted into the EcoRI and KpnI sites of the pCMV-HA expression vector. The pDR-GFP plasmid that was originally designed by Jasin and co-workers (21) for an HRR assay was modified by inserting a second I-SceI site (5Ј-TAGGGATAACAGGGTAAT-3Ј) upstream of the original I-SceI site (using a second, C-terminal truncated GFP as a donor template). The distance between the two I-SceI sites was 40 bp.
Irradiation-Low LET IR was carried out using an x-ray generator (X-RAD 320 (Precision X-ray, North Branford, CT); 320 kV, 10 mA, 2-mm aluminum filtration) in our laboratory. High LET IR was carried out using an alternating gradient synchrotron to generate 600 MeV/nucleon. 56 Fe ions (180 keV/m; range in water, 27 cm; beam area, 20 ϫ 20 cm; uniformity, Ϯ5%) at Brookhaven National Laboratory. The dose rates for both high LET IR and low LET IR were ϳ1 Gy/min.
Cell Sensitivity to IR-Cell sensitivity to IR was examined using a clonogenic assay as we described previously (8).
Detecting Mre11 in Chromatin DNA-Protein Complexes-The experimental procedure was similar to that described previously (9).
Detecting ␥-H2AX and H2A Levels in Cells-The experimental procedure was as described in our previous publication (26).
Detecting DNA in the Mre11-DNA Complex-The experimental procedure for detecting DNA in Mre11 complexes is similar to that described previously (9). Here, we summarize it in Fig. 4A. Briefly, 10 7 cells were used for paraformaldehyde cross-linking, and the cells were sonicated to produce DNA fragments of Ͻ1000 bp, immunoprecipitated using an anti-Mre11 antibody, and sonicated again. A small portion of each sample was reserved for detecting the Mre11 level. The DNA in the remainder was labeled with [␥-32 P]ATP using polynucleotide kinase. Following exhaustive protease digestion, DNA fragments were analyzed by native PAGE. DNA signals were Ape1 Activity Detection-The Ape1 endonuclease activity assay method was modified based on the published method (27). For Ape1 substrate preparation, two molecular beacons (APSUB and APCTRL) were designed. APSUB is the Ape1 substrate with an abasic site, and APCTRL is the control beacon without an abasic site. Both are labeled with 6-FAM fluorophore at the 5Ј-end and Dabcyl quencher at the 3Ј-end. The sequences are as follows (internal 1Ј2Ј-dideoxyribose spacer is an abasic spacer): APSUB, 5Ј-6-FAM-CCACT/idsp/TTGAAT- When annealed, the beacon formed a stem-loop structure. The lyophilized DNA oligonucleotides were dissolved in oligonucleotide dilution buffer (  or APCTRL in the working solution was allowed to anneal, and the melting curve quality was tested. The substrates were then ready for an Ape1 activity measurement. For nuclear extract preparation, MEF cells were transfected with siRNA against Ape1, control vector, wild type mApe1, mApe1-R176A, or mApe1-D209A expression plasmids in the pCMV-HA expression vector (Clontech, PT3283-5) using Lipofectamine 3000 (Invitrogen). At 24 h after transfection, cells were irradiated and put back in a 37°C incubator. The cells were collected at 1 h after IR, and the nuclear extracts were prepared using an NE-PER kit (Thermo Scientific (Pierce)) and then dialyzed in the buffer (50 mM HEPES-KOH, 100 mM KCl, 0.5 mM EDTA, 20% glycerol, 1 mM DTT) for 4.5 h with three buffer changes. 30 g of nuclear protein from each sample was mixed with 3.2 pmol of APSUB or APCTRL in reaction buffer at 37°C for 60 min. The 6-FAM fluorescence was detected in a Fast 7500 realtime PCR machine. The initial rates of fluorescence in each well were determined to represent the Ape1 endonuclease activity.
HRR Detection-The digestion efficiencies of I-SceI for the plasmid were examined in vitro using I-SceI enzyme (purchased from New England Biolabs). Briefly, 250 ng of plasmid were digested with 2 units of I-SceI for 30 min at room temperature. 10 ng of the digested products were used as the template for PCR, and the undigested plasmid was used as a control. The PCRs were run for 15 cycles with proper primers ( Table 1) that were used to check the downstream I-SceI site, which generated a 694-bp PCR product when the I-SceI site was undigested. The PCR products were separated by 1% (w/v) agarose. The digestion efficiency was analyzed by a ratio of digested DNA to total plasmid DNA (including undigested substrate). The HRR efficiency was detected with GFP signals using flow cytometry after the cells were transfected with I-SceI.

Up-regulation of Ape1 Promotes High LET IR-induced Cell
Killing-To test the hypothesis that Ape1 may affect cell survival through enzymatic digestion following high LET radiation, we first chose the approach of enhancing the Ape1 function by up-regulating the normal Ape1 protein because the Ape1 knockdown approach generated side effects on cell survival that masked the enzymatic digestion effects of Ape1 (19). Transiently transfecting a wild type Ape1 cDNA vector into wild type or Ku80 Ϫ/Ϫ cells (NHEJ-deficient MEF lines (28)) increased expression of Ape1 2-3-fold relative to cells transfected with a control vector (Fig. 1A). Under such conditions, the up-regulated Ape1 did not affect the sensitivity of the cells to low LET IR but significantly sensitized the wild type and Ku80 Ϫ/Ϫ cells to high LET IR (Fig. 1B). To further study whether the sensitization of Ape1-overexpressing cells to high LET IR was due to the AP site digestion function of Ape1, we examined the effects of Ape1 in Ogg1 Ϫ/Ϫ MEFs, which are deficient in the generation of AP sites (29). Overexpression of Ape1 in Ogg1 Ϫ/Ϫ cells (Fig. 1C) did not change the cell sensitivity to either low or high LET IR (Fig. 1D), which is different from the results obtained from wild type or NHEJ-deficient MEF cells with DNA glycosylase activity (Fig. 1B). Re-expression of Ogg1 in the Ogg1 Ϫ/Ϫ cells (Fig. 1C) restored the overexpression of Ape1-induced cell sensitization to high LET radiation (Fig. 1D).
Similar results were also obtained from human wild type (MRC5SV) and NHEJ-deficient 180BRM cells (Fig. 1, E and F). When Ogg1 was knocked down in MRC5SV cells, the Ape1 overexpression-induced radiosensitization effect disappeared (Fig. 1, E and F), confirming that the Ape1 overexpression-induced radiosensitization depends on Ogg1. Notably, without overexpression of Ape1, the RBE of cell killing is 1 for NHEJdeficient MEF (Ku80 Ϫ/Ϫ ) and human (180BRM) cells (Fig. 1G) (8), but overexpression of Ape1 results in the RBE increasing to 2 (Fig. 1G), suggesting that additional DNA DSBs occurred in these Ape1-overexpressing cells as well. Up-regulation of Ape1 Produces More Unrepaired DNA DSBs in High LET-than in Low LET-irradiated Cells-To confirm whether overexpression of Ape1 generated additional DNA DSBs in high LET-irradiated cells but not in low LET-irradiated cells, we used a ␥-H2AX focus assay to compare the remaining DNA DSBs because exposure to 1 Gy of high LET or low LET radiation resulted in almost 100% of the cells having ␥-H2AX signals (Fig. 2). The ␥-H2AX foci were more plentiful in wild type MEF (Ku80 ϩ/ϩ ) or NHEJ-deficient MEF (Ku80 Ϫ/Ϫ ) cells with Ape1 overexpression than in the cells transfected with control vectors at 4 and 24 h after high LET IR, but they did not show a difference in low LET-irradiated cells (Fig. 2, A  and B). Notably, overexpression of Ape1 in Ogg1 Ϫ/Ϫ MEF cells made no difference in the number of ␥-H2AX foci after exposure to high LET or low LET radiation (Fig. 2C). Re-expression of Ogg1 restored the difference in the number of ␥-H2AX foci-positive cells with and without Ape1 overexpression after exposure to high LET radiation (Fig. 2C). Similarly, overexpression of Ape1 in human cells (wild type MRC5SV or NHEJ-deficient 180BRM) increased the remaining ␥-H2AX foci-positive cells at 4 or 24 h after exposure to high LET IR but did not affect those exposed to low LET IR (Fig. 2, D and E). These data match the survival data (Fig. 1) and indicate that the Ape1 up-regulation results in additional DNA DSBs only in high LET-irradiated cells when Ogg1 is present. These results also suggest that Ape1 can generate small DNA fragments in clusters of damaged DNA by cleaving closely opposed Ogg1-induced AP sites.

Up-regulation of Ape1 Generates More Small Fragments of DNA in High LET-than in Low LET-irradiated Cells-
To demonstrate that up-regulation of Ape1 generated more small DNA fragments in high LET-irradiated cells than in low LET-irradiated cells, we examined the amounts of Mre11 bound to chromatin in cells overexpressing Ape1, as described in our previous report (9). Mre11, as the initial factor involved in the HRR pathway, requires a shorter DNA sequence for efficient DNA end binding (30,31) than Ku (32), which is the initial factor involved in NHEJ. We used the amount of ␥-H2AX binding to chromatin in the cells as a reference because the ␥-H2AX signal levels indicate the protein amount bound to broken DNA, but ␥-H2AX could not bind to any DNA fragment of Ͻ150 bp (33). The levels of ␥-H2AX in cells (with H2A as the internal control) were examined as described previously (26). The results showed that the levels of ␥-H2AX increased in irradiated wild type and NHEJ-deficient (Ku80 Ϫ/Ϫ ) MEF cells, regardless of the type of IR (low or high LET) or the expression levels of Ape1 (basal or overexpressed) (Fig. 3, A and B). On the other hand, overexpression of Ape1 increased the amount of Mre11 bound to chromatin in high LET-irradiated wild type and NHEJ-deficient (Ku80 Ϫ/Ϫ ) MEF cells when compared with that in low LET-irradiated cells (Fig. 3, A and B). These results suggest that overexpression of Ape1 generated additional DNA fragments of Ͻ150 bp in high LET-irradiated cells. Interestingly, overexpression of Ape1 did not change the level of Mre11 bound to chromatin in Ogg1 Ϫ/Ϫ cells after exposure to high LET radiation (Fig. 3C). However, re-expression of Ogg1 restored the  OCTOBER  increased amount of Mre11 bound to chromatin in high LETirradiated cells (Fig. 3D). These results suggest that overexpression of Ape1 generated additional DNA fragments of Ͻ150 bp in high LET-irradiated cells dependent on Ogg1.

Ape1 Contributes to RBE on Cell Killing
Next, to confirm that overexpression of Ape1-induced increased Mre11 bound to chromatin is due to increased amounts of DNADSBs, we examined the yield of DNA fragments bound to Mre11 in the immunoprecipitated Mre11-DNA complex at 1 h after exposure to high LET radiation (Fig.  4A). Once again, overexpression of Ape1 did not change the amount of DNA bound by Mre11 in Ogg1 Ϫ/Ϫ cells at 1 h after exposure to high LET IR (Fig. 4, B and C), but re-expressing Ogg1 restored the increase in DNA bound to Mre11 in the high LET-irradiated cells (Fig. 4, B and C). Combining these results shown in Figs. 2 and 3, we can conclude that up-regulation of Ape1 sensitized cells to high LET radiation cells by producing additional DNA fragments via digestion of DNA at Ogg1-dependet AP sites. These results also suggest that under physiological conditions, Ape1 may contribute to a higher RBE of high LET IR in cell killing.
Ape1 at Physiological Levels Contributes to High LET IR-induced RBE in Cell Killing-To directly demonstrate whether Ape1 contributes to high LET IR-induced RBE in cell killing at physiological levels, we generated three vectors that encode three different HA-tagged Ape1 constructs and are resistant to siRNA (Fig. 5A): wild type, overactivated Ape1, and inactivated Ape1. The overactivated Ape1 was generated by mutating amino acid 176 from Arg to Ala (34), and the inactivated Ape1 was generated by mutating amino acid 209 from Asp to Ala (35) (Fig. 5B). When the endogenous Ape1 was knocked down with the specific siRNA, the vector encoding HA-Ape1 was expressed in the cells (Fig. 5C). The Ape1 enzyme activities in these cells were verified (Fig. 5D). Notably, compared with cells expressed with wild type Ape1, cells expressed with Ape1 inactivation were more resistant to high LET IR, and cells expressed with Ape1 overactivation were more sensitive to high LET IR (Fig. 5E); however, there was no difference in the sensitivity of these cells to low LET IR (Fig. 5E). These results demonstrate that the enzymatic activity of Ape1 plays an important role in high LET IR-induced cell killing, suggesting that at a physiological level, Ape1 can contribute to the RBE on high LET radiation-induced cell killing.
Combining the 1.8-fold greater amount of small DNA fragments generated directly by high LET IR when compared with low LET IR (8) and the subsequent Ape1-generated DNA DSBs at the clustered DNA damage sites, we can better explain the RBE, 2-6, for general cell survival. It is believed that enzymatic modification contributes to 30 -50% of IR-induced total DSBs (11), and thus the initial 1.8-fold more small DNA fragments generated by high LET radiation (8) would increase after Ape1 functions in high LET-irradiated cells, which results in more cell killing and a higher RBE.
Small DNA Fragments That Affect NHEJ Do Not Affect HRR-The fact that RBE in cell killing is equivalent for low and high LET radiation in NHEJ-deficient cells indicates that high LET radiation only interferes with NHEJ and not with HRR. This conclusion has been verified by our group (8,9,36) and another group (10). The mechanism involves high LET IR-generated small DNA fragments (Յ40 bp) whose size prevents Ku from simultaneously binding to the two ends of one DNA fragment to initiate efficient NHEJ (8). We have also shown that small DNA fragments (Յ40 bp) do not affect the DNA binding efficiency of Mre11 (an initial step for HRR) (9,36). These differences arise from the different structures, DNA binding characteristics, and functions of the two proteins: Ku (32,37) and Mre11 (30,38).
To provide direct evidence that small DNA fragments (Յ40 bp) do not affect HRR efficiency in vivo, we used the I-SceI-HRR-reporter system by modifying the pDR-GFP construct that was originally designed by Jasin and co-workers (21).
The modified plasmid contains two sites for I-SceI digestion separated by 40 nucleotides within the GFP expression frame (Fig. 6A). We first compared the I-SceI digestion efficiency for the plasmids containing one digestion site and two digestion sites in vitro by PCR to amplify the digestion products with specific primers. The results showed that the I-SceI digestion was more efficient (about 3-fold) for the plasmids with one digestion site than for the plasmid with two digestion sites (Fig.  6B), suggesting that the I-SceI enzyme requires Ͼ20 bp at each side of the digestion site for efficient digestion. After confirming that the plasmids had integrated into transfected human 293FT cells using the same primers to amplify PCR products from the genomic DNA, we measured the HRR efficiency by transfecting I-SceI in these cells. The results showed that the difference in HRR efficiency (ϳ3-fold difference) in the cells containing the plasmid DNA with one I-SceI-digestion site and those with two I-SceI-digestion sites (Fig. 6C) was similar to that in the original comparison of I-SceI digestion efficiency (ϳ3-fold difference). These results indicate that a 40-bp DNA fragment does not decrease HRR efficiency, although DNA The pink sites are mutated nucleotides. C, endogenous or exogenous Ape1 was detected by immunoblotting with the Ape1 or HA antibody in wild type MEF cells transfected with wild type (WT), inactivated (In), or overactivated (Ac) Ape1 combined with control (Ct) RNA or siRNA against Ape1 (siApe1). D, the relative Ape1 enzymatic activity in the MEF cells expressing different types of Ape1 was analyzed by comparing with the enzymatic activity of the cells transfected with wild type Ape1 when the endogenous Ape1 was knocked down without IR (NR). The cells were collected at 1 h after exposure to 4 Gy of low LET (L4 Gy) or 2 Gy of high LET (H2 Gy) radiation. The data were derived from three independent experiments. *, p Ͻ 0.05; **, p Ͻ 0.01; ***, p Ͻ 0.001. E, the cell sensitivities to low LET (L) or high LET (H) irradiation at the indicated dose as labeled were examined using a clonogenic assay. Data are the mean and S.D. (error bars) obtained from three independent experiments. *, p Ͻ 0.05; **, p Ͻ 0.01. OCTOBER 31, 2014 • VOLUME 289 • NUMBER 44 fragments of this size affect efficient NHEJ. These results strongly suggest that the small DNA fragments (Յ40 bp) generated by either heavy ions directly penetrating a nucleosome or consequent Ape1 digestion within clustered DNA damage in high LET-irradiated cells do not affect HRR (Fig. 6D).

DISCUSSION
Ape1 was independently identified as a reductive activator of the AP-1 (c-Jun/Fos heterodimer) transcription factor and named redox effector factor 1 (Ref-1) (39). Ape1 is involved in both the base excision repair pathway and the regulation of gene expression as a redox co-activator of different transcription factors, such as early growth response protein-1 (Egr-1), p53, HIF1-␣, and AP-1 (40,41). These activities are controlled by two functionally distinct domains; the N terminus is mainly devoted to the gene regulation, whereas the C terminus has the enzymatic activity on the AP sites of DNA (34,42).
The survival threat of knockout of Ape1 to mice (43) or knockdown of Ape1 to cells (19) may depend mainly on the gene regulation function of Ape1 because Ogg1 (the main enzyme to generate the AP site in base excision repair) deficiency did not affect mouse survival (29). Up-or down-regulating the Ape1 activity without IR did not affect cell sensitivity to low IR induced killing, which also supports the above prediction. In addition, exogenous expression of enzyme-inactivated Ape1 as compared with wild type Ape1 when the endogenous Ape1 in the cells was knocked down decreased IR-induced cell killing (Fig. 5E) and provides additional evidence to suggest that the survival threat of knockout/knockdown of Ape1 depends mainly on the gene regulation function of Ape1. Modifying the Ape1 activity primarily affects cell sensitivity to high LET radiation but does not affect cell sensitivity to low LET radiation, which is mainly because of the difference in the structure of damaged DNA generated by high LET IR and low LET IR. High LET radiation with the dense ionizing events generates more cluster DNA damage than low LET radiation (13). At clustered DNA damage sites, it is much easier for Ape1 to generate small DNA fragments (Fig. 6D).
The same amount of ionizing events occur in the same doses of high LET-and low LET-irradiated cells, which results in the High LET radiation-induced small DNA fragments are generated from a direct energy transfer track to break DNA strands and from subsequent Ape1 enzymatic modification in the clustered damage DNA sites. These small DNA fragments interfere with Ku-dependent NHEJ but do not affect HRR. same yield of DNA DSBs. However, it is known that high LET does kill more wild type, HRR-deficient cells than low LET IR but not the Ku-dependent NHEJ-deficient cells (6 -8, 10), indicating that high LET IR interferes with only Ku-dependent NHEJ. We believe that the small DNA fragment model is one of the major reasons for the phenotypes because there are no other models supported by convincing evidence to explain why high LET IR interferes with only the Ku-dependent NHEJ. One question concerning the small DNA fragment model is why the damaged DNAs do not simply lose the small fragments and directly join the remaining ends. In fact, the radiation-induced small DNA fragments are very likely lost if they are located between nucleosomes without histone attachments. However, histone attachments protect the damaged DNAs from losing small fragments when they are located within nucleosomes. The observation of more DNA fragments (Յ40 bp) in high LET-irradiated cells than in low LET-irradiated cells (8) and the RBE of 1 for NHEJ-deficient cells strongly support this explanation. Based on our results, it would be possible to reduce high LET IR-induced damage in normal cells by down-regulating the Ape1 activity or to sensitize tumor cells to high LET radiotherapy by up-regulating the Ape1 activity in the near future.
Taken together, these results reveal that the base excision repair function of Ape1 plays different roles in low LET-and high LET-irradiated cells due to the dense clustered DNA damage generated by high LET radiation. This new discovery could contribute to improve either high LET radiation protection or high LET radiotherapy.