**2. DNA damage and biological response to radiation**

Both the hazards and potential beneficial uses of ionizing radiation (IR) were realised soon after the discovery of X-rays by Wilhelm Conrad Roentgen in 1895. Studies of Hiroshima and Nagasaki survivors, reconstructed dosimetry, and unfortunate accidents at nuclear plants documented a pattern of events following a whole-body IR exposure, confirmed by extensive animal experiments. Exposure to high doses of IR (100-150 Gy) leads to death within a few hours which results from neurological and cardiovascular breakdown. At intermediate dose levels (5-12 Gy), death occurs within a few days and is associated with gastrointestinal syndrome. At lower doses (2.5-5 Gy) death occurs within several weeks due to haematopoietic syndrome (Hall, 1973). All these effects are attributable to killing of critical cells, and the question of how IR kills cells has stimulated much research. A key milestone was the identification of DNA as the critical molecular target. This research was prompted by both the potential uses of IR, for example in cancer therapy, and by concerns about effects of IR on health. Potentially damaging exposure may come from diagnostic radiology such as computed tomography as well as from cosmic rays, the sun and radioactive nuclides in the ground (e.g. radon), during high altitude journeys, or in space. Such concerns about occupational and environmental radiation exposure have prompted much scientific and legislative activity, the latter leading to the establishment of the International Commission on Radiological Protection.

It is a commonly recognised concept that two distinct mechanisms are responsible for induction of DNA damage by IR; one involves direct ionisation of atoms in the DNA molecule and usually is referred to as the direct effect, and another that results from DNA attack by free radicals generated as a result of the radiolysis of surrounding water molecules and is referred to as the indirect effect (Hall, 1973; Hall et al., 1988). The major contributor to the indirect effect is the hydroxyl radical (HO•) as evidenced by studies using compounds that scavenge hydroxyl radicals (Roots & Okada, 1972). The direct DNA damage is considered to include the damage that is produced by hydroxyl radicals generated in water molecules intimately associated with DNA water layer since these radicals cannot be scavenged (Ward, 1994a). It is estimated that two-thirds of DNA damage is caused indirectly by scavengeable radicals (Roots & Okada, 1972).

IR induces a wide variety of mainly isolated DNA lesions, including strand breaks (single strand breaks – SSB) and damage that involves DNA bases (modified bases and abasic sites). Isolated DNA lesions are normally easily repaired by cells (Ward, 1994a; Sutherland et al., 2000), without serious biological consequences. It is believed that cytotoxic and mutagenic effects of IR originate from so called clustered DNA damage when two or more lesions occur close to each other in both DNA strands (Goodhead, 1994; Ward, 1994a; Goodhead & Nikjoo, 1997; Nikjoo et al., 1997; Nikjoo et al., 1998; Sutherland et al., 2000). This type of DNA damage was initially termed locally multiply damaged sites (LMDS) (Ward, 1994a), but more recently the term LDMS has been largely replaced by the synonym oxidative clustered DNA lesions (OCDL) (Sutherland et al., 2000; Georgakilas, 2008). Formation of strand breaks in opposite DNA strands results in a double strand break (DSB), that represents a particular case of OCDL (Purkayastha et al., 2007). Assuming that base damage (BD) occurs more frequently than SSB (Ward, 1994b; Ward, 1994a), it is expected that the majority of OCDL are not frank DSB but contain two BD or BD and SSB in opposite strands (Goodhead & Nikjoo, 1997). While isolated DNA lesions are generally repaired efficiently, OCDL are more difficult to resolve (Ward, 1981; Harrison et al., 1999; Georgakilas et al.,

Both the hazards and potential beneficial uses of ionizing radiation (IR) were realised soon after the discovery of X-rays by Wilhelm Conrad Roentgen in 1895. Studies of Hiroshima and Nagasaki survivors, reconstructed dosimetry, and unfortunate accidents at nuclear plants documented a pattern of events following a whole-body IR exposure, confirmed by extensive animal experiments. Exposure to high doses of IR (100-150 Gy) leads to death within a few hours which results from neurological and cardiovascular breakdown. At intermediate dose levels (5-12 Gy), death occurs within a few days and is associated with gastrointestinal syndrome. At lower doses (2.5-5 Gy) death occurs within several weeks due to haematopoietic syndrome (Hall, 1973). All these effects are attributable to killing of critical cells, and the question of how IR kills cells has stimulated much research. A key milestone was the identification of DNA as the critical molecular target. This research was prompted by both the potential uses of IR, for example in cancer therapy, and by concerns about effects of IR on health. Potentially damaging exposure may come from diagnostic radiology such as computed tomography as well as from cosmic rays, the sun and radioactive nuclides in the ground (e.g. radon), during high altitude journeys, or in space. Such concerns about occupational and environmental radiation exposure have prompted much scientific and legislative activity, the latter leading to the establishment of the

It is a commonly recognised concept that two distinct mechanisms are responsible for induction of DNA damage by IR; one involves direct ionisation of atoms in the DNA molecule and usually is referred to as the direct effect, and another that results from DNA attack by free radicals generated as a result of the radiolysis of surrounding water molecules and is referred to as the indirect effect (Hall, 1973; Hall et al., 1988). The major contributor to the indirect effect is the hydroxyl radical (HO•) as evidenced by studies using compounds that scavenge hydroxyl radicals (Roots & Okada, 1972). The direct DNA damage is considered to include the damage that is produced by hydroxyl radicals generated in water molecules intimately associated with DNA water layer since these radicals cannot be scavenged (Ward, 1994a). It is estimated that two-thirds of DNA damage is caused

IR induces a wide variety of mainly isolated DNA lesions, including strand breaks (single strand breaks – SSB) and damage that involves DNA bases (modified bases and abasic sites). Isolated DNA lesions are normally easily repaired by cells (Ward, 1994a; Sutherland et al., 2000), without serious biological consequences. It is believed that cytotoxic and mutagenic effects of IR originate from so called clustered DNA damage when two or more lesions occur close to each other in both DNA strands (Goodhead, 1994; Ward, 1994a; Goodhead & Nikjoo, 1997; Nikjoo et al., 1997; Nikjoo et al., 1998; Sutherland et al., 2000). This type of DNA damage was initially termed locally multiply damaged sites (LMDS) (Ward, 1994a), but more recently the term LDMS has been largely replaced by the synonym oxidative clustered DNA lesions (OCDL) (Sutherland et al., 2000; Georgakilas, 2008). Formation of strand breaks in opposite DNA strands results in a double strand break (DSB), that represents a particular case of OCDL (Purkayastha et al., 2007). Assuming that base damage (BD) occurs more frequently than SSB (Ward, 1994b; Ward, 1994a), it is expected that the majority of OCDL are not frank DSB but contain two BD or BD and SSB in opposite strands (Goodhead & Nikjoo, 1997). While isolated DNA lesions are generally repaired efficiently, OCDL are more difficult to resolve (Ward, 1981; Harrison et al., 1999; Georgakilas et al.,

**2. DNA damage and biological response to radiation** 

International Commission on Radiological Protection.

indirectly by scavengeable radicals (Roots & Okada, 1972).

2002; Georgakilas, 2008). An attempt to process by the base excision repair (BER) a modified base that constitute an OCDL with a SSB in opposite strand may result in formation of a DSB, which also can be formed when an unprocessed OCDL interferes with either DNA replication and transcription (Bonner et al., 2008; Sedelnikova et al., 2010). Although DNA DSB can be repaired in cells by non-homologous end joining (NHEJ) and homologous recombination (HR) (Matsumoto et al., 1994; Memisoglu & Samson, 2000; Wilson et al., 2003; Cadet et al., 2010; Hinz, 2010; Lieber, 2010; Mladenov & Iliakis, 2011), DSB are among the most toxic IR-induced DNA lesions. If not properly repaired, both accumulated DSB and OCDL lead to cytotoxicity, genome instability and carcinogenesis (Jeggo & Lobrich, 2007; McKinnon & Caldecott, 2007).

It is established that a substantial level of oxidative DNA lesions may be present in normal cells and tissues, usually a few isolated oxidative DNA lesions per Mbp (Nakamura & Swenberg, 1999; De Bont & van Larebeke, 2004). These lesions are believed to be generated by free radicals that originate from endogenous reactive oxygen species (ROS) (Riley, 1994; Mikkelsen & Wardman, 2003). Two of the biologically important endogenous ROS are superoxide anion radical O2- • and hydrogen peroxide H2O2. Endogenous superoxide is produced in cells mainly as a result of mitochondrial respiration (Mikkelsen & Wardman, 2003), and then is efficiently converted to hydrogen peroxide by cellular superoxide dismutase (SOD). Although the superoxide and hydrogen peroxide are relatively long lived species and are able to diffuse in cells over considerable distance (Riley, 1994), these endogenous species are produced in cytoplasm and also they are not able to damage DNA directly. The genotoxic effect of the endogenous ROS is mainly mediated by their ability to give rise to hydroxyl radicals from hydrogen peroxide by a redox reaction with traces of reduced transitional metal ions, mainly ferrous via Fenton chemistry (Mikkelsen & Wardman, 2003). Since ROS represent potential risk for cells, an antioxidant defense system has been developed in cells to maintain a steady state level of ROS.

Exogenous cytotoxic agents can lead to the increase above the steady state in the ROS level thus creating an oxidative stress that can result in induction of additional oxidative DNA damage (Sedelnikova et al., 2010). Exposure to IR is also known to cause the oxidative stress however, interestingly, the level of ROS generated directly from radiolysis of water at biologically relevant doses is much less that the level of the endogenous ROS. This follows from the estimation that for example a 100 Gy radiation dose would be required to double the endogenous level of one of the major types of DNA base damage 7,8-dihydro-8 oxoguanine (8-oxoG) (Ward, 1994b). There is also experimental evidence that the transient increase in the cellular ROS level following irradiation is dependent on mitochondria respiration, however it is dose independent in the range of biologically relevant doses (1 – 10 Gy) with the fraction of cells exhibiting the increased ROS level being dose dependent (Leach et al., 2001). These observations underline the minimal impact of isolated DNA lesions and the critical role of clustered lesions for cyto- and genotoxic consequences of IR. The major difference between endogenous ROS and those generated by IR is that while the spatial distribution of hydroxyl radicals produced from endogenous ROS is random, IR is also able to generate clusters of hydroxyl radicals within a nanometre scale resulting in the multiple radical attack on DNA within small volume from a single track of a charged particle (Goodhead, 1994; Goodhead & Nikjoo, 1997; Nikjoo et al., 1997; Nikjoo et al., 1998). As a result, the relative frequency of OCDL is much higher for damage induced by IR as compared to endogenous oxidative DNA damage for which OCDL are very rare. The frequency of endogenous OCDL is estimated to be a few per Gbp in normal tissues (Bennett

DNA-Binding Radioprotectors 501

The US National Cancer Institute Workshop have developed recommendations for the terminology and classification of the agents used to ameliorate the biological consequences of the exposure to IR (Stone, 2003). The classification implies that there are different mechanisms of action of these agents, and therefore they may be efficient when administered appropriately with regard to the time of the exposure to IR. Accordingly, there are three groups of such agents. Prophylactic agents/protectors are administered before exposure to IR and mainly act by chemically preventing the initial radiochemical damage; mitigators are given during or soon after exposure to IR to prevent development of tissue damage; and treatment agents are administered after exposure to IR to reduce symptoms

Apart from normal tissue damage, another major concern associated with cancer radiotherapy is the potential for emergence of secondary radiation-induced cancers, affecting more than 1% of patients (Hall, 2006). Such an outcome can arise in two ways, the first being the induction of mutagenic DNA damage in nearby normal tissues. The second is associated with a phenomenon similar to RIBE in *in vitro* settings that has been reported by cancer radiotherapists more than 50 years ago and termed the abscopal effect (Mole, 1953; Kaminski et al., 2005). It is defined as a change in an organ or tissue distant from the irradiated region. Since these non-targeted effects include malignant transformation (Hall & Hei, 2003; Mancuso et al., 2008), the abscopal effect represents a serious risk factor in

Therefore, efforts to reduce radiation toxicity in normal tissues and/or in a whole organism are of significant clinical importance and an area of active research. The development of

Of the thousands of compounds synthesised and tested at the Walter Reed Army Institute of Research in the 1960's search for radioprotectors, aminothiols emerged as the most promising compounds. The persistent motif associated with radioprotective activity of aminothiols is a thiol separated from an aliphatic amino group by a two carbon chain (Brown et al., 1982). The simplest example is cysteamine (chemical formula H2N-CH2-CH2- SH). One of the most studied aminothiols is the radioprotector WR1065 (2- [aminopropyl)amino]ethanethiol, Figure 1), which is the active thiol metabolite of

WR1065 protects cultured cells against radiation induced clonogenic death. A dose modification factor (DMF) of 1.9 is achieved for V79 cells pre-incubated 30 min with 4mM of WR1065 before irradiation (Grdina et al., 1985). DMF is defined as the ratio of radiation doses producing the same degree of radiation effect, in the presence and absence of the radiomodifier. In the context of radioprotection, and particularly for *in vivo* endpoints, DRF, dose reduction factor is often used. It has been shown using neutral elution technique that

radioprotectors can be regarded as an important strategy to achieve these objectives.

Fig. 1. Structure of WR1065 (R = H) and its prodrug amifostine (R = H2PO3).

developed as a result of this exposure.

**4. Aminothiols as radioprotectors** 

radiotherapy.

amifostine (WR2721).

et al., 2005), and can probably be formed as a result of two isolated oxidative lesions occurring close to each other spatially and temporally. The radiation induced OCDL can also be more complex since consisting of more than two individual lesions in a cluster (Ward, 1994a; Bennett et al., 2005).

Recently, the dogma that cells subjected to IR are killed solely through direct energy deposition within a cell, with the effect being proportional to dose, has been reconsidered in view of the discovery of the radiation-induced bystander effect (RIBE). The neighbours of irradiated cells respond as if they themselves have been irradiated. The RIBE is now a wellestablished consequence of exposure to IR and is manifested as increased genomic abnormalities and loss of viability in unirradiated ("bystander") cells associated with the targeted cells. Affected bystander cells exhibit increased levels of micronuclei, apoptosis, mutations, altered DNA damage and repair, and senescence arrest (Sokolov et al., 2005; Sedelnikova et al., 2007; Prise & O'Sullivan, 2009). DSB appear in the DNA of bystander cells, invoking the existence of some kind of biological "danger signal" that is sent from irradiated to bystander cell. Possible mediators of the RIBE include various inflammationrelated cytokines and ROS including nitric oxide that have been found at elevated levels in medium conditioned by irradiated cells (Dickey et al., 2009; Prise & O'Sullivan, 2009; Hei et al., 2010; Ivanov et al., 2010).

### **3. Radiation therapy of tumours and the role of radioprotectors**

The detrimental consequences of IR for cells and tissues can be harnessed in cancer radiation therapy. Radiation therapy exploits the cytotoxic effect of IR on cancer cells (Lawrence et al., 2008) and represents one of the three major treatment modalities for cancer, along with surgery and cytotoxic chemotherapy. Radiotherapy is used in approximately half of all patients diagnosed with cancer at some stage of their illness. Technological advances in the physical targeting of radiation to the tumour are extensively exploited, reflecting the simple idea that the most efficient radioprotection strategy is to exclude normal tissues from irradiated volume. These include such techniques as conformal radiotherapy, intensity modulated therapy, image guided radiotherapy etc (Brizel, 2005). Although still in the preclinical stage of development, microbeam radiotherapy (MRT), in which the X-ray beam is split into an array of planar parallel microbeams, shows much greater therapeutic index than conventional radiotherapy. Studies of synchrotron MRT in animal models indicated that tumours can be ablated by MRT at radiation levels that spare normal tissues (Dilmanian et al., 2002; Dilmanian et al., 2003; Crosbie et al., 2010). Nevertheless, whether used alone or in combination with other treatment modalities, the dose of radiation that can be safely delivered is limited by radiation induced injury to the normal tissues in the irradiated volume. This gives rise to the concept of "treatment to tolerance" i.e. the administration of the maximally tolerated dose (MTD) imposed by the normal tissues which is often less than required to effect a high probability of tumour eradication. Clearly, any strategy that selectively increases the MTD improves the chances of tumour cure, and one of such strategies is the selective pharmacologic modification of the normal tissue response with radiation modifiers/protectors. These agents alter the response of normal tissues to irradiation when present in tissues prior or after exposure to IR (Citrin, 2010). This approach can also be viewed as an attractive countermeasure for possible nuclear/radiological terrorism and radiation accidents, but without the important constraint to avoid protection of the tumour that is imperative in the use of radioprotectors in radiation oncology.

et al., 2005), and can probably be formed as a result of two isolated oxidative lesions occurring close to each other spatially and temporally. The radiation induced OCDL can also be more complex since consisting of more than two individual lesions in a cluster

Recently, the dogma that cells subjected to IR are killed solely through direct energy deposition within a cell, with the effect being proportional to dose, has been reconsidered in view of the discovery of the radiation-induced bystander effect (RIBE). The neighbours of irradiated cells respond as if they themselves have been irradiated. The RIBE is now a wellestablished consequence of exposure to IR and is manifested as increased genomic abnormalities and loss of viability in unirradiated ("bystander") cells associated with the targeted cells. Affected bystander cells exhibit increased levels of micronuclei, apoptosis, mutations, altered DNA damage and repair, and senescence arrest (Sokolov et al., 2005; Sedelnikova et al., 2007; Prise & O'Sullivan, 2009). DSB appear in the DNA of bystander cells, invoking the existence of some kind of biological "danger signal" that is sent from irradiated to bystander cell. Possible mediators of the RIBE include various inflammationrelated cytokines and ROS including nitric oxide that have been found at elevated levels in medium conditioned by irradiated cells (Dickey et al., 2009; Prise & O'Sullivan, 2009; Hei et

**3. Radiation therapy of tumours and the role of radioprotectors** 

The detrimental consequences of IR for cells and tissues can be harnessed in cancer radiation therapy. Radiation therapy exploits the cytotoxic effect of IR on cancer cells (Lawrence et al., 2008) and represents one of the three major treatment modalities for cancer, along with surgery and cytotoxic chemotherapy. Radiotherapy is used in approximately half of all patients diagnosed with cancer at some stage of their illness. Technological advances in the physical targeting of radiation to the tumour are extensively exploited, reflecting the simple idea that the most efficient radioprotection strategy is to exclude normal tissues from irradiated volume. These include such techniques as conformal radiotherapy, intensity modulated therapy, image guided radiotherapy etc (Brizel, 2005). Although still in the preclinical stage of development, microbeam radiotherapy (MRT), in which the X-ray beam is split into an array of planar parallel microbeams, shows much greater therapeutic index than conventional radiotherapy. Studies of synchrotron MRT in animal models indicated that tumours can be ablated by MRT at radiation levels that spare normal tissues (Dilmanian et al., 2002; Dilmanian et al., 2003; Crosbie et al., 2010). Nevertheless, whether used alone or in combination with other treatment modalities, the dose of radiation that can be safely delivered is limited by radiation induced injury to the normal tissues in the irradiated volume. This gives rise to the concept of "treatment to tolerance" i.e. the administration of the maximally tolerated dose (MTD) imposed by the normal tissues which is often less than required to effect a high probability of tumour eradication. Clearly, any strategy that selectively increases the MTD improves the chances of tumour cure, and one of such strategies is the selective pharmacologic modification of the normal tissue response with radiation modifiers/protectors. These agents alter the response of normal tissues to irradiation when present in tissues prior or after exposure to IR (Citrin, 2010). This approach can also be viewed as an attractive countermeasure for possible nuclear/radiological terrorism and radiation accidents, but without the important constraint to avoid protection

of the tumour that is imperative in the use of radioprotectors in radiation oncology.

(Ward, 1994a; Bennett et al., 2005).

al., 2010; Ivanov et al., 2010).

The US National Cancer Institute Workshop have developed recommendations for the terminology and classification of the agents used to ameliorate the biological consequences of the exposure to IR (Stone, 2003). The classification implies that there are different mechanisms of action of these agents, and therefore they may be efficient when administered appropriately with regard to the time of the exposure to IR. Accordingly, there are three groups of such agents. Prophylactic agents/protectors are administered before exposure to IR and mainly act by chemically preventing the initial radiochemical damage; mitigators are given during or soon after exposure to IR to prevent development of tissue damage; and treatment agents are administered after exposure to IR to reduce symptoms developed as a result of this exposure.

Apart from normal tissue damage, another major concern associated with cancer radiotherapy is the potential for emergence of secondary radiation-induced cancers, affecting more than 1% of patients (Hall, 2006). Such an outcome can arise in two ways, the first being the induction of mutagenic DNA damage in nearby normal tissues. The second is associated with a phenomenon similar to RIBE in *in vitro* settings that has been reported by cancer radiotherapists more than 50 years ago and termed the abscopal effect (Mole, 1953; Kaminski et al., 2005). It is defined as a change in an organ or tissue distant from the irradiated region. Since these non-targeted effects include malignant transformation (Hall & Hei, 2003; Mancuso et al., 2008), the abscopal effect represents a serious risk factor in radiotherapy.

Therefore, efforts to reduce radiation toxicity in normal tissues and/or in a whole organism are of significant clinical importance and an area of active research. The development of radioprotectors can be regarded as an important strategy to achieve these objectives.
