DNA is considered a major target for ionizing radiation (IR)-based therapies [1], [2], [3]. Exposure of cells to IR has long been known to induce double-strand breaks (DSBs), a particularly cytotoxic form of DNA damage [2], but it is only in the mid 1990’s that this was visualized in situ within the nucleus. Immunofluorescence (IF) detection showed that irradiation of mammalian cells with 137Cs induces the accumulation, into discrete intranuclear foci (also called IRIF for ionizing radiation-induced foci), of RAD51 [4] or RAD50 and MRE11 [5] proteins, all involved in DSB repair (DSBR). These results led to the proposal of the formation of repair centers in response to DNA damage induced by γ-rays. A major advance in the understanding of DNA repair processes of the IR-induced DNA damage arose with the possibility of irradiating subnuclear regions with soft X-rays and monitoring the recruitment of DNA repair proteins by IF [6]. This approach confirmed the hypothesis that proteins involved in the DNA damage response (DDR) [7] are specifically recruited to the irradiated regions, corresponding to DNA damage sites, and opened the door to a wealth of information on the mechanisms underlying the repair of DSBs. More recently, the possibility of inducing localized DNA damage combined with the monitoring of nuclear protein behaviors in living cells by fluorescently tagging them allowed analysis of the recruitment kinetics of DNA repair factors to IR-induced DSBs [8], [9], [10].

Due to its advantages over other IR types, proton beam therapy is increasingly used for the treatment of a variety of cancers [11], [12]. Indeed, deposition of the bulk of the beam energy in a defined region, the Bragg peak, minimizes its impact on the healthy tissues found in its path, therefore reducing toxicity [13]. It is therefore important to understand the mechanisms underlying the biological response triggered by such IR. This should allow not only better assessment of the risk associated with human exposure but also to identify ways of improving the efficacy of proton therapy. As expected, irradiation with protons also induces DSBs. However, when compared with photon irradiation, at similar doses, irradiation with protons produced larger and more irregularly shaped γH2AX and 53BP1 foci [14], [15], [16]. Interestingly, experiments comparing different types of irradiation showed delayed kinetics of repair for some of the lesions generated by protons [14], [17], supporting the idea that the DSBs induced by this type of radiation are more often in close proximity with other radiation-induced DNA lesions and therefore more difficult for the cell to repair [18]. While there is an increasingly large literature on the formation and repair of proton-induced DSBs, less is known on what other kinds of lesions are formed at the proximity of DSBs, potentially altering their repair of efficiency. Early work following the localized synthesis of poly(ADP-ribose) after proton beam micro-irradiation [19] suggested the possibility of the induction of single-strand breaks (SSBs), later confirmed experimentally [17]. Generation of clustered lesions, comprising, beside SSBs and DSBs, a mixture of abasic sites and damaged bases, is assumed as well [18]. In particular, the excess of reactive oxygen species (ROS) observed in different cell types upon proton irradiation when compared to photon irradiation [20], [21], [22] points to the potential presence of oxidized bases at the site of irradiation. Due to its low oxidative potential, 7,8-dihydro-8-oxoguanine (8-oxoG) is the predominant lesion in DNA exposed to ROS [23], [24]. Another oxidized base abundantly formed on DNA following irradiation is thymine glycol (TG) [25], [26], [27]. In mammalian cells, 8-oxoG and TG are recognized by specific DNA glycosylases, OGG1 and NTH1, respectively, that initiate the base excision repair (BER) by excising the oxidized base and generating an abasic site. Although both OGG1 and NTH1 are bifunctional DNA glycosylases possessing also AP lyase activities, at least in the case of OGG1, the abasic site is usually cleaved by APE1 [28], [29]. The resulting AP site provides a substrate to a DNA polymerase and a ligase that can then complete the repair process. Other proteins like PARP1 and XRCC1 have also been shown to be involved in both pathways, SSB repair (SSBR) and BER [30], [31]. The use of photon microbeams has allowed the study of the recruitment of DNA glycosylases to laser micro-irradiated regions of cell nuclei [32], [33], [34], [35], [36]. However, while there is increasing information on the recruitment of DSBR proteins to the site of DNA damage induced by proton beams [37], [38], [39], the formation of other types of lesions and involvement of cognate repair mechanisms has not been explored.

In recent years, microbeam irradiation has been applied as a new approach aimed at deciphering the molecular mechanisms of IR-induced DNA damage recognition, signaling, and repair. Indeed, the ion microbeams provide powerful experimental tools allowing strict control over the site and time of damage and the study of recruitment kinetics of proteins involved in these processes [9], [19], [40], [41], [42], [43]. Here we used the IRSN’s MIRCOM facility designed for targeted irradiation with a focused ion microbeam extracted in air [44] to induce DNA damage by localized proton irradiation within nuclei of living cells with a high spatio-temporal resolution. We show that such a treatment locally induces, in addition to DSBs, 8-oxoG and TG. Consistent with the formation of oxidative base damage, we detected the recruitment of the DNA glycosylases OGG1 and NTH1, capable of excising 8-oxoG and TG, respectively, and therefore initiating the BER pathway. To our knowledge, this is the first evidence indicating that proton beams induce oxidative DNA damage, and thus involving BER in the repair of DNA lesions induced by protons. Considering the recent developments of BER inhibitors, this observation opens the door to new therapeutic avenues to improve the use of proton therapy.