**5. Radioprotection by methylproamine**

#### **5.1 Methylproamine as a DNA binding antioxidant**

Methylproamine is a radioprotector, which belongs to a family of DNA minor groove binders featuring a common bi-benzimidazole structure (Figure 2). Two commercially available bi-benzimidazoles Hoechst 33258 and Hoechst 33342 are widely used as fluorescent DNA binding dyes.

Fig. 2. Chemical structure of bibenzimidazoles. R1 = H and R2 = OCH2CH3 (Hoechst 33342); R1 = CH3 and R2 = N(CH3)2 (methylproamine); R1 = H and R2 = N(CH3)2 (proamine).

Some protective activity against IR was initially discovered for Hoechst 33342 in cultured cells (Smith & Anderson, 1984; Young & Hill, 1989) and followed by reports of radioprotection of isolated DNA (Denison et al., 1992; Martin & Denison, 1992) and *in vivo* radioprotection of mouse lung (Martin et al., 1996) and brain (Lyubimova et al., 2001). New analogues of Hoechst 33342 were designed to improve the radioprotective activity, resulting in synthesis of more efficient compounds proamine (Figure 2) (Martin et al., 1996) and methylproamine (Figure 2) (Martin et al., 2004). Incubation of V79 Chinese Hamster cells in 30 M of methylproamine before and during -irradiation increases clonogenic survival with a DMF of 2.1 (Martin et al., 2004). *In vivo* radioprotection by methylproamine has been demonstrated in mouse jejunum using the Withers assay with a DMF of 1.2 – 1.3.

Methylproamine, like other DNA binding bi-benzimidazoles, has a binding preference for AT-rich sequences, the consensus binding site being 3-4 consecutive AT base pairs as

DNA-Binding Radioprotectors 505

the role of DNA bound methylproamine in radioprotection we have undertaken extensive studies of radioprotection using clonogenic survival of human cultured keratinocytes as an endpoint (Lobachevsky et al., 2011). The dose response curves of clonogenic survival have been established for FEP-1811 keratinocytes pre-incubated with various concentrations of methylproamine from 0.5 to 10 M for 30 min before irradiation with 137Cs -rays. The DMF calculated for each survival curve has increased from 1.01 at 0.5 M to 1.97 at 10 M of

In parallel experiments the uptake of methylproamine in cells and nuclei has been measured by extracting the drug from nuclear and cellular pellets and measurement by liquid chromatography. It was found that while the cellular uptake increased as a linear function of methylproamine concentration (up to 4 fmole/cell at 10 M of methylproamine), the nuclear uptake indicated the presence of the major saturated and minor linear components. The saturated component reflects in our opinion accumulation of DNA bound methylproamine and saturation of high affinity binding sites at high concentrations. The saturation level has been estimated to be 0.173 fmole per nucleus and corresponds to approximately 1 ligand per 58 bp. This value is similar to the size of the binding site calculated from *in vitro* DNA binding studies with methylproamine and analogues (Loontiens et al., 1990; Martin et al., 2004). The fraction of DNA bound methylproamine is not less than 98% as estimated assuming nucleus radius of 5 m (DNA concentration 26 mM bp), binding dissociation constant Kd = 100 nM and methylproamine concentration 450 M (total nuclear uptake 0.234 fmole at 10 M in medium). This result in combination with the presence of the saturated component indicates that the majority of the nuclear methylproamine is in the DNA bound form. The presence of the linear component of the nuclear uptake may result from the heterogeneity in the affinity of binding sites so that

"weaker" sites are occupied at increasing methylproamine concentrations.

**5.4 Methylproamine reduces radiation induced DNA damage in cells** 

Values of DMF obtained at various concentrations of methylproamine have been analysed in conjunction with results of cellular and nuclear uptake studies (Lobachevsky et al., 2011). Correlation has been studied between DMF values and each of the cellular uptake, nuclear uptake and saturated and linear components of nuclear uptake. The best correlation have been achieved for the total nuclear uptake of methylproamine (R2 = 0.97) and the least correlation for the cellular uptake (R2 = 0.87). These results, along with the finding that the majority of the nuclear methylproamine is present in DNA bound form, support the hypothesis that it is the DNA associated drug that is responsible for

In addition to radioprotection at the clonogenic survival endpoint, the effect of methylproamine on the induction by radiation of H2AX foci has been studied (Sprung et al., 2010; Lobachevsky et al., 2011). γH2AX can be detected microscopically using immunofluorescence technique as a distinct focus that is associated with a DNA DSB (Sedelnikova et al., 2002; Sedelnikova et al., 2003). The dose response curves of H2AX focus number per cell were established following irradiation of FEP-1811 keratinocytes preincubated with 20 M of methylproamine for the time intervals of 1, 5 and 15 min before irradiation with 137Cs -rays (Lobachevsky et al., 2011). The results have demonstrated the reduction by methylproamine of the number of radiation induced H2AX foci. The extent of this reduction is consistent with pre-incubation interval as indicated by DMF values

methylproamine.

radioprotection of cells.

established by footprinting (Harshman & Dervan, 1985) and affinity cleavage (Martin & Holmes, 1983; Martin et al., 1990; Murray & Martin, 1994) studies, and confirmed by X-ray crystallography studies (Martin et al., 2004).

### **5.2 Electron transport is involved in radioprotection by methylproamine**

In the 70's and 80's considerable effort was devoted to the development of hypoxic cell radiosensitisers and it was well established that these agents were electron-affinic. From this dogma, that withdrawing electron density from DNA confers radiosensitivity, it can be inferred that increasing electron density in DNA would have radioprotective effect. It was this simple idea that guided the modification of Hoechst 33342 by the substitution of electron rich groups, and this improved radioprotective activity. Pulse radiolysis studies of methylproamine/DNA complexes provided further information on the movement of an electron from DNA-bound radioprotector to oxidising lesions on DNA (Martin & Anderson, 1998). In these studies, the spectral changes associated with oxidation of the DNA ligand were followed by time resolved spectrophotometry. Moreover, by studying the effect of the drug loading of DNA on the rate of oxidation of the ligand, the range of electron movement from the bound ligand to oxidising lesion could be estimated. The results indicated that for methylproamine, the maximum range was several base pairs (Martin & Anderson, 1998).

The phenomenon of electron donation as the basis for radioprotection also emerged from the studies of tyrosine containing peptides which have been modelled of naturally occurring nuclear proteins involved in the endogenous radioprotection (Tsoi et al., 2010).

From consideration of studies of the chemical mechanism of radioprotection by thiols, the alternative mechanisms of H-atom donation and electron donation have been discussed. The close relationship between these two mechanisms of reduction is also invoked in the combination of electron and proton transfer in a concerted mechanism. Indeed this mechanism is well established in radiation chemistry of DNA (Kumar & Sevilla, 2010). These considerations have lead to the hypothesis that radioprotection by methylproamine involves repair of transient radiation induced species on DNA by electron donation from the DNA bound ligand. An alternative mechanistic concept would invoke the DNA bound ligand as the sink for "holes" produced in irradiated DNA. It is well established that radical cations produced on DNA by powerful oxidants move to the most easily oxidisable base pair, namely GC. The presence of DNA-bound methylproamine would constitute an alternative destination for the hole, thus reducing the yield of oxidised bases. A redox potential of 0.84 - 0.9 volt has been reported for Hoechst 33342 (Adhikary et al., 2000) so it is reasonable to assume that the redox potential for methylproamine is similar and therefore consistent with the hole trapping hypothesis.

#### **5.3 DNA bound methylproamine is responsible for radioprotection of cells**

The results of pulse radiolysis studies of methylproamine-DNA complexes indicated that intramolecular electron transfer from the ligand to radiation induced oxidising species on DNA is involved in the oxidation of methylproamine, and this process can be implicated in radioprotective activity of methylproamine (Martin & Anderson, 1998). Radioprotection *in vivo* however occurs in a different environment than reduction of oxidising species on DNA in pulse radiolysis experiments, and other processes may contribute to radioprotection *in vivo* such as for example scavenging of hydroxyl radicals by free methylproamine. To clarify

established by footprinting (Harshman & Dervan, 1985) and affinity cleavage (Martin & Holmes, 1983; Martin et al., 1990; Murray & Martin, 1994) studies, and confirmed by X-ray

In the 70's and 80's considerable effort was devoted to the development of hypoxic cell radiosensitisers and it was well established that these agents were electron-affinic. From this dogma, that withdrawing electron density from DNA confers radiosensitivity, it can be inferred that increasing electron density in DNA would have radioprotective effect. It was this simple idea that guided the modification of Hoechst 33342 by the substitution of electron rich groups, and this improved radioprotective activity. Pulse radiolysis studies of methylproamine/DNA complexes provided further information on the movement of an electron from DNA-bound radioprotector to oxidising lesions on DNA (Martin & Anderson, 1998). In these studies, the spectral changes associated with oxidation of the DNA ligand were followed by time resolved spectrophotometry. Moreover, by studying the effect of the drug loading of DNA on the rate of oxidation of the ligand, the range of electron movement from the bound ligand to oxidising lesion could be estimated. The results indicated that for methylproamine, the maximum range was several base pairs

The phenomenon of electron donation as the basis for radioprotection also emerged from the studies of tyrosine containing peptides which have been modelled of naturally occurring

From consideration of studies of the chemical mechanism of radioprotection by thiols, the alternative mechanisms of H-atom donation and electron donation have been discussed. The close relationship between these two mechanisms of reduction is also invoked in the combination of electron and proton transfer in a concerted mechanism. Indeed this mechanism is well established in radiation chemistry of DNA (Kumar & Sevilla, 2010). These considerations have lead to the hypothesis that radioprotection by methylproamine involves repair of transient radiation induced species on DNA by electron donation from the DNA bound ligand. An alternative mechanistic concept would invoke the DNA bound ligand as the sink for "holes" produced in irradiated DNA. It is well established that radical cations produced on DNA by powerful oxidants move to the most easily oxidisable base pair, namely GC. The presence of DNA-bound methylproamine would constitute an alternative destination for the hole, thus reducing the yield of oxidised bases. A redox potential of 0.84 - 0.9 volt has been reported for Hoechst 33342 (Adhikary et al., 2000) so it is reasonable to assume that the redox potential for methylproamine is similar and therefore

nuclear proteins involved in the endogenous radioprotection (Tsoi et al., 2010).

**5.3 DNA bound methylproamine is responsible for radioprotection of cells** 

The results of pulse radiolysis studies of methylproamine-DNA complexes indicated that intramolecular electron transfer from the ligand to radiation induced oxidising species on DNA is involved in the oxidation of methylproamine, and this process can be implicated in radioprotective activity of methylproamine (Martin & Anderson, 1998). Radioprotection *in vivo* however occurs in a different environment than reduction of oxidising species on DNA in pulse radiolysis experiments, and other processes may contribute to radioprotection *in vivo* such as for example scavenging of hydroxyl radicals by free methylproamine. To clarify

**5.2 Electron transport is involved in radioprotection by methylproamine** 

crystallography studies (Martin et al., 2004).

(Martin & Anderson, 1998).

consistent with the hole trapping hypothesis.

the role of DNA bound methylproamine in radioprotection we have undertaken extensive studies of radioprotection using clonogenic survival of human cultured keratinocytes as an endpoint (Lobachevsky et al., 2011). The dose response curves of clonogenic survival have been established for FEP-1811 keratinocytes pre-incubated with various concentrations of methylproamine from 0.5 to 10 M for 30 min before irradiation with 137Cs -rays. The DMF calculated for each survival curve has increased from 1.01 at 0.5 M to 1.97 at 10 M of methylproamine.

In parallel experiments the uptake of methylproamine in cells and nuclei has been measured by extracting the drug from nuclear and cellular pellets and measurement by liquid chromatography. It was found that while the cellular uptake increased as a linear function of methylproamine concentration (up to 4 fmole/cell at 10 M of methylproamine), the nuclear uptake indicated the presence of the major saturated and minor linear components. The saturated component reflects in our opinion accumulation of DNA bound methylproamine and saturation of high affinity binding sites at high concentrations. The saturation level has been estimated to be 0.173 fmole per nucleus and corresponds to approximately 1 ligand per 58 bp. This value is similar to the size of the binding site calculated from *in vitro* DNA binding studies with methylproamine and analogues (Loontiens et al., 1990; Martin et al., 2004). The fraction of DNA bound methylproamine is not less than 98% as estimated assuming nucleus radius of 5 m (DNA concentration 26 mM bp), binding dissociation constant Kd = 100 nM and methylproamine concentration 450 M (total nuclear uptake 0.234 fmole at 10 M in medium). This result in combination with the presence of the saturated component indicates that the majority of the nuclear methylproamine is in the DNA bound form. The presence of the linear component of the nuclear uptake may result from the heterogeneity in the affinity of binding sites so that "weaker" sites are occupied at increasing methylproamine concentrations.

Values of DMF obtained at various concentrations of methylproamine have been analysed in conjunction with results of cellular and nuclear uptake studies (Lobachevsky et al., 2011). Correlation has been studied between DMF values and each of the cellular uptake, nuclear uptake and saturated and linear components of nuclear uptake. The best correlation have been achieved for the total nuclear uptake of methylproamine (R2 = 0.97) and the least correlation for the cellular uptake (R2 = 0.87). These results, along with the finding that the majority of the nuclear methylproamine is present in DNA bound form, support the hypothesis that it is the DNA associated drug that is responsible for radioprotection of cells.

#### **5.4 Methylproamine reduces radiation induced DNA damage in cells**

In addition to radioprotection at the clonogenic survival endpoint, the effect of methylproamine on the induction by radiation of H2AX foci has been studied (Sprung et al., 2010; Lobachevsky et al., 2011). γH2AX can be detected microscopically using immunofluorescence technique as a distinct focus that is associated with a DNA DSB (Sedelnikova et al., 2002; Sedelnikova et al., 2003). The dose response curves of H2AX focus number per cell were established following irradiation of FEP-1811 keratinocytes preincubated with 20 M of methylproamine for the time intervals of 1, 5 and 15 min before irradiation with 137Cs -rays (Lobachevsky et al., 2011). The results have demonstrated the reduction by methylproamine of the number of radiation induced H2AX foci. The extent of this reduction is consistent with pre-incubation interval as indicated by DMF values

DNA-Binding Radioprotectors 507

Early experiments with plasmid DNA model have demonstrated ability of methylproamine analogues Hoechst 33342 and Hoechst 33258 to protect DNA from radiation induced SSB (Denison et al., 1992; Martin & Denison, 1992) with a linear increase in the DMF from approximately 2 to more than 10 with the ligand concentration changing from 5 to 50 M (Martin & Denison, 1992). Two modes of protection have been suggested: the site-specific and global protection. The site-specific protection occurs at the site of the ligand binding on DNA and has been suggested to involve direct block of the radical attack by the ligand occupying DNA minor groove and/or electron or H-atom transfer from the ligand to DNA. However, since only limited DNA regions (less than 20%) can by occupied by bound ligand, the site-specific protection can not account completely for the observed extent in the SSB reduction (up to DMF of 10). Therefore the idea of global protection has been suggested that implies radioprotection between ligand binding sites. Given that the linear relationship between the extent of radioprotection and the ligand concentration has been observed, the most likely mechanism of global protection detected in these experiments is the scavenging of hydroxyl radicals by free ligand in solution. This suggestion is supported by the much smaller extent of radioprotection by 25 M of Hoechst 33258 in the presence of 100 mM mannitol that efficiently scavenges hydroxyl radicals: a DMF of 1.4 as compared to 9.6 for 25 M of Hoechst 33258 in phosphate buffer (Martin & Denison, 1992), and the fact that majority of the ligand binding sites (94-100%) are occupied within the range of the studied concentrations of Hoechst 33258 (5 – 50 M) indicating that the radioprotection by DNA

Although the site-specific protection by methylproamine analogues partially prevents formation of SSB, and the global protection by DNA bound ligand in cells can also potentially reduce the yield of SSB, it is unlikely that these mechanisms of radioprotection by methylproamine will affect the yield of DSB as to account for the observed radioprotection at the clonogenic survival and H2AX induction endpoints with DMF of 2 and more. A more likely mechanism is the chemical reduction by DNA bound methylproamine of the initial oxidative DNA lesion that results in base damage constituting a part of OCDL. Such OCDL represent a difficult challenge for the cellular DNA repair machinery and potentially can result in enzymatic DNA DSB. However, following reduction by methylproamine of the radical precursor of the lesion that otherwise would constitute a part of an OCDL, the formation of this OCDL can be prevented, thus preventing potential formation of an enzymatic DSB. The ability of methylproamine to reduce oxidative base

In our experiments pBR322 plasmid DNA has been irradiated with 137Cs -rays in the PTP buffer (Figure 3) containing thiocyanate ions, as described by (Milligan et al., 2000b; Milligan et al., 2001). In this buffer, the thiocyanate ion (SCN-) is the main scavenger of radiation induced hydroxyl radicals (HO•) that otherwise would be responsible for the

Interaction of SCN- with HO• results in formation of highly reactive radical species SCN•

buffer. In contrast to HO• radicals that are efficient in induction of DNA breaks,

the most frequently damaged site oxidation of which results in formation of guanyl radical that eventually is converted mainly to 8-oxoG. Compared to HO• radicals, with a reduction potential (E) of 2.3 V, SCN•/(SCN)2•- are more moderate oxidants (E = 1.62/1.32 V), but

•- that are the major mediators of the radiation induced DNA damage in this

•- radicals produce oxidative lesions of DNA bases. Guanine is considered as

bound ligand wouldn't change significantly at this condition.

lesions has also been demonstrated using plasmid DNA model.

induction of the majority of DNA breakage.

and/or (SCN)2

SCN•/(SCN)2

of 1.4, 1.9 and 3.5 for 1, 5 and 15 min respectively. The efficient reduction of the number of radiation induced H2AX foci by pre-incubation with methylproamine has been also demonstrated with three lymphoblast cell lines derived from the blood of the radiotherapy patients with different DNA repair capacity (Sprung et al., 2005; Sprung et al., 2008). In these experiments (Sprung et al., 2010), cells have been irradiated with 137Cs -rays following 15 min pre-incubation with 20 M methylproamine. Radioprotection has been observed in all three cell lines including those obtained from a radiosensitive patient and with a defective DNA ligase IV that is critical for DNA DSB repair pathway. This finding demonstrates the ability of methylproamine to reduce the amount of radiation induced DNA damage.

To further investigate the effect of methylproamine on the radiation induced DNA damage in cells, the pulsed field gel electrophoresis (PFGE) assay has been exploited (Sprung et al., 2010). For this assay, lymphoblast cells have been irradiated with -ray doses of 20, 40 and 80 Gy in the presence or without 20 M methylproamine. DNA was extracted from the cells, analysed on PFGE and the fraction of lower molecular weight DNA released from the wells has been quantified (Sprung et al., 2010). The results demonstrate substantial decrease of the low molecular weight fraction in all irradiated samples pre-treated with methylproamine as compared to the irradiated only samples thus indicating prevention by methylproamine of DNA fragmentation due to radiation induced DSB.

#### **5.5 Methylproamine protects against breaks and base damage in plasmid DNA**

A series of observations such as the role of DNA bound methylproamine in the radioprotection of cells, the requirement for methylproamine to be present in cells at the time of irradiation and demonstration that electron transport is involved in radioprotection support the hypothesis that the chemical reduction of transient radiation induced oxidative species on DNA by donation of an electron from methylproamine constitutes the main mechanism of radioprotection. This hypothesis however, in conjunction with the observation that methylproamine prevents formation of radiation induced DSB in cells, as demonstrated by pulsed field gel electrophoresis studies and the reduction of the yield of H2AX foci, prompts the question of what are those oxidative DNA species that are reduced by methylproamine and how the chemical reduction can repair or prevent the formation of a DNA DSB. A further insight into the mechanisms of radioprotection can be obtained from investigation of DNA damage of isolated DNA using a plasmid model.

Plasmid DNA is a convenient tool to assay DNA strand breakage. It exploits conformational changes of the supercoiled plasmid to the relaxed open circle form following induction of a SSB and to the linear form following induction of a DSB (Freifelder & Trumbo, 1969; Cowan et al., 1987; Lobachevsky et al., 2004). Three plasmid forms can be separated using agarose gel electrophoresis and numbers of SSB and DSB calculated from fractions of the linear and relaxed forms (Cowan et al., 1987). Combination of the plasmid DNA breakage assay and treatment of DNA with base excision repair enzymes (endonucleases) that recognise various DNA base lesions and convert them to strand breaks allows quantitation of radiation induced base lesions (Milligan et al., 2000a). One of such enzymes is the endonuclease formamidopyrimidine-DNA N-glycosylase (FPG) from Escherichia coli (O'Connor & Laval, 1989). FPG recognises oxidised purines, in particular 8-oxoG (Chetsanga & Lindahl, 1979; Milligan et al., 2002) and possesses both glycosylase and endonuclease activity to excise the modified base and then produce a nick at this abasic site (O'Connor & Laval, 1989), thus converting the base damage to a SSB.

of 1.4, 1.9 and 3.5 for 1, 5 and 15 min respectively. The efficient reduction of the number of radiation induced H2AX foci by pre-incubation with methylproamine has been also demonstrated with three lymphoblast cell lines derived from the blood of the radiotherapy patients with different DNA repair capacity (Sprung et al., 2005; Sprung et al., 2008). In these experiments (Sprung et al., 2010), cells have been irradiated with 137Cs -rays following 15 min pre-incubation with 20 M methylproamine. Radioprotection has been observed in all three cell lines including those obtained from a radiosensitive patient and with a defective DNA ligase IV that is critical for DNA DSB repair pathway. This finding demonstrates the ability of methylproamine to reduce the amount of radiation

To further investigate the effect of methylproamine on the radiation induced DNA damage in cells, the pulsed field gel electrophoresis (PFGE) assay has been exploited (Sprung et al., 2010). For this assay, lymphoblast cells have been irradiated with -ray doses of 20, 40 and 80 Gy in the presence or without 20 M methylproamine. DNA was extracted from the cells, analysed on PFGE and the fraction of lower molecular weight DNA released from the wells has been quantified (Sprung et al., 2010). The results demonstrate substantial decrease of the low molecular weight fraction in all irradiated samples pre-treated with methylproamine as compared to the irradiated only samples thus indicating prevention by methylproamine of

**5.5 Methylproamine protects against breaks and base damage in plasmid DNA** 

investigation of DNA damage of isolated DNA using a plasmid model.

A series of observations such as the role of DNA bound methylproamine in the radioprotection of cells, the requirement for methylproamine to be present in cells at the time of irradiation and demonstration that electron transport is involved in radioprotection support the hypothesis that the chemical reduction of transient radiation induced oxidative species on DNA by donation of an electron from methylproamine constitutes the main mechanism of radioprotection. This hypothesis however, in conjunction with the observation that methylproamine prevents formation of radiation induced DSB in cells, as demonstrated by pulsed field gel electrophoresis studies and the reduction of the yield of H2AX foci, prompts the question of what are those oxidative DNA species that are reduced by methylproamine and how the chemical reduction can repair or prevent the formation of a DNA DSB. A further insight into the mechanisms of radioprotection can be obtained from

Plasmid DNA is a convenient tool to assay DNA strand breakage. It exploits conformational changes of the supercoiled plasmid to the relaxed open circle form following induction of a SSB and to the linear form following induction of a DSB (Freifelder & Trumbo, 1969; Cowan et al., 1987; Lobachevsky et al., 2004). Three plasmid forms can be separated using agarose gel electrophoresis and numbers of SSB and DSB calculated from fractions of the linear and relaxed forms (Cowan et al., 1987). Combination of the plasmid DNA breakage assay and treatment of DNA with base excision repair enzymes (endonucleases) that recognise various DNA base lesions and convert them to strand breaks allows quantitation of radiation induced base lesions (Milligan et al., 2000a). One of such enzymes is the endonuclease formamidopyrimidine-DNA N-glycosylase (FPG) from Escherichia coli (O'Connor & Laval, 1989). FPG recognises oxidised purines, in particular 8-oxoG (Chetsanga & Lindahl, 1979; Milligan et al., 2002) and possesses both glycosylase and endonuclease activity to excise the modified base and then produce a nick at this abasic site (O'Connor & Laval, 1989), thus

induced DNA damage.

DNA fragmentation due to radiation induced DSB.

converting the base damage to a SSB.

Early experiments with plasmid DNA model have demonstrated ability of methylproamine analogues Hoechst 33342 and Hoechst 33258 to protect DNA from radiation induced SSB (Denison et al., 1992; Martin & Denison, 1992) with a linear increase in the DMF from approximately 2 to more than 10 with the ligand concentration changing from 5 to 50 M (Martin & Denison, 1992). Two modes of protection have been suggested: the site-specific and global protection. The site-specific protection occurs at the site of the ligand binding on DNA and has been suggested to involve direct block of the radical attack by the ligand occupying DNA minor groove and/or electron or H-atom transfer from the ligand to DNA. However, since only limited DNA regions (less than 20%) can by occupied by bound ligand, the site-specific protection can not account completely for the observed extent in the SSB reduction (up to DMF of 10). Therefore the idea of global protection has been suggested that implies radioprotection between ligand binding sites. Given that the linear relationship between the extent of radioprotection and the ligand concentration has been observed, the most likely mechanism of global protection detected in these experiments is the scavenging of hydroxyl radicals by free ligand in solution. This suggestion is supported by the much smaller extent of radioprotection by 25 M of Hoechst 33258 in the presence of 100 mM mannitol that efficiently scavenges hydroxyl radicals: a DMF of 1.4 as compared to 9.6 for 25 M of Hoechst 33258 in phosphate buffer (Martin & Denison, 1992), and the fact that majority of the ligand binding sites (94-100%) are occupied within the range of the studied concentrations of Hoechst 33258 (5 – 50 M) indicating that the radioprotection by DNA bound ligand wouldn't change significantly at this condition.

Although the site-specific protection by methylproamine analogues partially prevents formation of SSB, and the global protection by DNA bound ligand in cells can also potentially reduce the yield of SSB, it is unlikely that these mechanisms of radioprotection by methylproamine will affect the yield of DSB as to account for the observed radioprotection at the clonogenic survival and H2AX induction endpoints with DMF of 2 and more. A more likely mechanism is the chemical reduction by DNA bound methylproamine of the initial oxidative DNA lesion that results in base damage constituting a part of OCDL. Such OCDL represent a difficult challenge for the cellular DNA repair machinery and potentially can result in enzymatic DNA DSB. However, following reduction by methylproamine of the radical precursor of the lesion that otherwise would constitute a part of an OCDL, the formation of this OCDL can be prevented, thus preventing potential formation of an enzymatic DSB. The ability of methylproamine to reduce oxidative base lesions has also been demonstrated using plasmid DNA model.

In our experiments pBR322 plasmid DNA has been irradiated with 137Cs -rays in the PTP buffer (Figure 3) containing thiocyanate ions, as described by (Milligan et al., 2000b; Milligan et al., 2001). In this buffer, the thiocyanate ion (SCN-) is the main scavenger of radiation induced hydroxyl radicals (HO•) that otherwise would be responsible for the induction of the majority of DNA breakage.

Interaction of SCN- with HO• results in formation of highly reactive radical species SCN• and/or (SCN)2 •- that are the major mediators of the radiation induced DNA damage in this buffer. In contrast to HO• radicals that are efficient in induction of DNA breaks, SCN•/(SCN)2 •- radicals produce oxidative lesions of DNA bases. Guanine is considered as the most frequently damaged site oxidation of which results in formation of guanyl radical that eventually is converted mainly to 8-oxoG. Compared to HO• radicals, with a reduction potential (E) of 2.3 V, SCN•/(SCN)2•- are more moderate oxidants (E = 1.62/1.32 V), but

DNA-Binding Radioprotectors 509

Methylproamine at concentration as low as 1.25 M protects plasmid DNA against enzymatic SSB with a DMF of 10 while against frank SSB with a moderate DMF of 1.4 (Table 1), thus demonstrating much higher extent of protection against base damage than against frank SSB. A possible candidate lesion for repair by methylproamine is a guanyl radical cation that results, if not repaired, in formation of 8-oxoG, a modified base that is recognised and converted to SSB by FPG. The most critical question with regard to mechanisms of such efficient protection against base damage is whether radioprotection is mediated by DNA bound or free methylproamine and is achieved via the reduction by methylproamine of

The results presented in Table 1 for radioprotection at 1.25 and 5 M of methylproamine indicate a moderate decrease in the yield of base damage from 3.16x10-2 to 2.08x10-2 (34%) that resulted from 4-fold increase in methylproamine concentration. While addition of 1.25 M methylproamine to PTP buffer prevents formation of 90% of base damage, second addition of 1.25 M (from 1.25 to 2.5 M) prevents formation of only 22% of the remaining base damage. It is important to note that the change in the fraction of pBR322 DNA binding sites occupied by methylproamine is minimal (estimated from 85 to 97% as methylproamine concentration changes from 1.25 to 2.5 M). The results therefore are consistent with the hypothesis that the radioprotection against base damage is mainly mediated by DNA bound methylproamine. In general, these results demonstrate the ability of methylproamine to protect against radiation induced base damage and therefore support the hypothesis that reduction of the oxidative DNA lesions by methylproamine accounts for radioprotection of

The strong high affinity binding of bibenzimidazoles in the minor groove of DNA is a factor that determines the high radioprotective potency of methylproamine: substantial radioprotection of clonogenic survival is achieved at concentrations as low as a few M in cell culture medium (a DMF 1.6 at 2 M) (Lobachevsky et al., 2011), and at even lower concentrations for base damage in the plasmid model (a DMF of 10 at 1.25 M) (Table 1). On the other hand, it is logical to expect that the tightly DNA associated bisbenzimidazole molecule will interfere with normal DNA metabolic processes such as replication, transcription, repair etc and such an interaction may result in adverse cytotoxic and mutagenic effects. While no effect of methylproamine on the clonogenic survival of human keratinocytes has been detected following 60 min incubation with 10 M of the drug, at 20 M of methylproamine the clonogenic survival has been reduced to 80% (Lobachevsky et al., 2011). Mechanisms of this cytotoxicity have not been fully investigated. The cytotoxicity of Hoechst 33342 has prompted consideration and further development of bibenzimidazoles as potential antitumour agents (Baraldi et al., 2004). Inhibition of DNA synthesis has been demonstrated in V79 cells exposed to 5 and 10 M of Hoechst 33342. This inhibition results in substantial changes in the progression of cells through cell cycle as manifested by appearance of increased S-phase population and S/G2 block (Durand & Olive, 1982). One of the potential mechanisms of the Hoechst 33342 cytotoxicity is its interaction with topoisomerase I. Inhibition of the topoisomerase I activity by Hoechst 33342 and 33258 has been demonstrated using both the relaxation and cleavage assays (Chen et al., 1993). The

•- radicals in solutions that cause

oxidative lesions on DNA or scavenging SCN•/(SCN)2

**6. Cytotoxicity of DNA binding ligands** 

DNA lesions.

cells.

nevertheless powerful enough to oxidise guanines in DNA (E = 1.29 V). Thus irradiation in PTP buffer results in selective damage; primarily oxidation of guanine.

Fig. 3. Loss of supercoiled plasmid with increasing radiation dose. A solution of 7.5 g/mL of pBR322 (11.4 M bp) in a buffer containing 5 mM sodium phosphate, pH 7.0, 1 mM sodium thiocyanate and 110 mM sodium perchlorate (PTP buffer) has been irradiated with 137Cs -rays without (circles) or with 5 M of methylproamine (squares). Irradiated samples were analysed by agarose gel electrophoresis to assay frank SSB (open symbols) or after treatment with FPG to assay total frank SSB and enzymatic SSB (base damage) (closed symbols).

The yield of FPG sensitive lesions (FPG enzymatic SSB) in pBR322 following irradiation in the thiocyanate buffer is more than 10-fold higher than the yield of frank SSB (Figure 3, Table 1).


Table 1. Effect of methylproamine on the yield of frank and enzymatic SSB in irradiated pBR322 plasmid DNA.

nevertheless powerful enough to oxidise guanines in DNA (E = 1.29 V). Thus irradiation in

Fig. 3. Loss of supercoiled plasmid with increasing radiation dose. A solution of 7.5 g/mL of pBR322 (11.4 M bp) in a buffer containing 5 mM sodium phosphate, pH 7.0, 1 mM sodium thiocyanate and 110 mM sodium perchlorate (PTP buffer) has been irradiated with 137Cs -rays without (circles) or with 5 M of methylproamine (squares). Irradiated samples were analysed by agarose gel electrophoresis to assay frank SSB (open symbols) or after treatment with FPG to assay total frank SSB and enzymatic SSB (base damage) (closed

The yield of FPG sensitive lesions (FPG enzymatic SSB) in pBR322 following irradiation in the thiocyanate buffer is more than 10-fold higher than the yield of frank SSB (Figure 3,

Yield of SSB and BD

Buffer Frank SSB BD

PTP 2.45 ± 0.05 31.8 ± 1.2

PTP+1.25 M methylproamine 1.76 ± 0.03 3.16 ± 0.29 1.39 10.1 PTP+2.5 M methylproamine 1.67 ± 0.01 2.45 ± 0.13 1.47 13.0 PTP+5 M methylproamine 1.52 ± 0.01 2.08 ± 0.09 1.61 15.3 Table 1. Effect of methylproamine on the yield of frank and enzymatic SSB in irradiated

10-2 per plasmid per Gy DMF

(enzymatic SSB) Frank SSB BD

(enzymatic SSB)

symbols).

Table 1).

pBR322 plasmid DNA.

PTP buffer results in selective damage; primarily oxidation of guanine.

Methylproamine at concentration as low as 1.25 M protects plasmid DNA against enzymatic SSB with a DMF of 10 while against frank SSB with a moderate DMF of 1.4 (Table 1), thus demonstrating much higher extent of protection against base damage than against frank SSB. A possible candidate lesion for repair by methylproamine is a guanyl radical cation that results, if not repaired, in formation of 8-oxoG, a modified base that is recognised and converted to SSB by FPG. The most critical question with regard to mechanisms of such efficient protection against base damage is whether radioprotection is mediated by DNA bound or free methylproamine and is achieved via the reduction by methylproamine of oxidative lesions on DNA or scavenging SCN•/(SCN)2 •- radicals in solutions that cause DNA lesions.

The results presented in Table 1 for radioprotection at 1.25 and 5 M of methylproamine indicate a moderate decrease in the yield of base damage from 3.16x10-2 to 2.08x10-2 (34%) that resulted from 4-fold increase in methylproamine concentration. While addition of 1.25 M methylproamine to PTP buffer prevents formation of 90% of base damage, second addition of 1.25 M (from 1.25 to 2.5 M) prevents formation of only 22% of the remaining base damage. It is important to note that the change in the fraction of pBR322 DNA binding sites occupied by methylproamine is minimal (estimated from 85 to 97% as methylproamine concentration changes from 1.25 to 2.5 M). The results therefore are consistent with the hypothesis that the radioprotection against base damage is mainly mediated by DNA bound methylproamine. In general, these results demonstrate the ability of methylproamine to protect against radiation induced base damage and therefore support the hypothesis that reduction of the oxidative DNA lesions by methylproamine accounts for radioprotection of cells.

### **6. Cytotoxicity of DNA binding ligands**

The strong high affinity binding of bibenzimidazoles in the minor groove of DNA is a factor that determines the high radioprotective potency of methylproamine: substantial radioprotection of clonogenic survival is achieved at concentrations as low as a few M in cell culture medium (a DMF 1.6 at 2 M) (Lobachevsky et al., 2011), and at even lower concentrations for base damage in the plasmid model (a DMF of 10 at 1.25 M) (Table 1). On the other hand, it is logical to expect that the tightly DNA associated bisbenzimidazole molecule will interfere with normal DNA metabolic processes such as replication, transcription, repair etc and such an interaction may result in adverse cytotoxic and mutagenic effects. While no effect of methylproamine on the clonogenic survival of human keratinocytes has been detected following 60 min incubation with 10 M of the drug, at 20 M of methylproamine the clonogenic survival has been reduced to 80% (Lobachevsky et al., 2011). Mechanisms of this cytotoxicity have not been fully investigated. The cytotoxicity of Hoechst 33342 has prompted consideration and further development of bibenzimidazoles as potential antitumour agents (Baraldi et al., 2004). Inhibition of DNA synthesis has been demonstrated in V79 cells exposed to 5 and 10 M of Hoechst 33342. This inhibition results in substantial changes in the progression of cells through cell cycle as manifested by appearance of increased S-phase population and S/G2 block (Durand & Olive, 1982). One of the potential mechanisms of the Hoechst 33342 cytotoxicity is its interaction with topoisomerase I. Inhibition of the topoisomerase I activity by Hoechst 33342 and 33258 has been demonstrated using both the relaxation and cleavage assays (Chen et al., 1993). The

DNA-Binding Radioprotectors 511

Brizel, D. M. (2005). *Strategies for protecting normal tissues in the treatment of head and neck* 

Brown, D. Q., Pittock, J. W., 3rd & Rubinstein, J. S. (1982). Early results of the screening

Cadet, J., Douki, T. & Ravanat, J. L. (2010). Oxidatively generated base damage to cellular

Chen, A. Y., Yu, C., Bodley, A., Peng, L. F. & Liu, L. F. (1993). A new mammalian DNA

Chetsanga, C. J. & Lindahl, T. (1979). Release of 7-methylguanine residues whose imidazole

Citrin, D., Cotrim, A.P., Hyodo, F., Baum, B.J., Krishna, M.C., Mitchell, J.B. (2010).

Cowan, R., Collis, C. M. & Grigg, G. W. (1987). Breakage of double-stranded DNA due to

Crosbie, J. C., Anderson, R. L., Rothkamm, K., Restall, C. M., Cann, L., Ruwanpura, S.,

De Bont, R. & van Larebeke, N. (2004). Endogenous DNA damage in humans: a review of

Denekamp, J., Michael, B. D., Rojas, A. & Stewart, F. A. (1982). Radioprotection of mouse

Denison, L., Haigh, A., D'Cunha, G. & Martin, R. F. (1992). DNA ligands as radioprotectors:

Dickey, J. S., Baird, B. J., Redon, C. E., Sokolov, M. V., Sedelnikova, O. A. & Bonner, W. M.

Dilmanian, F. A., Button, T. M., Le Duc, G., Zhong, N., Pena, L. A., Smith, J. A., Martinez, S.

Dilmanian, F. A., Morris, G. M., Zhong, N., Bacarian, T., Hainfeld, J. F., Kalef-Ezra, J.,

program for radioprotectors. *Int J Radiat Oncol Biol Phys*, Vol. 8, No. 3-4, pp. 565-

topoisomerase I poison Hoechst 33342: cytotoxicity and drug resistance in human

rings have been opened from damaged DNA by a DNA glycosylase from

Radioprotectors and Mitigators of Radiation-Induced Normal Tissue Injury. *The* 

Meachem, S., Yagi, N., Svalbe, I., Lewis, R. A., Williams, B. R. & Rogers, P. A. (2010). Tumor cell response to synchrotron microbeam radiation therapy differs markedly from cells in normal tissues. *Int J Radiat Oncol Biol Phys*, Vol. 77, No. 3,

skin by WR-2721: the critical influence of oxygen tension. *Int J Radiat Oncol Biol* 

molecular studies with Hoechst 33342 and Hoechst 33258. *International Journal of* 

(2009). Intercellular communication of cellular stress monitored by gamma-H2AX

R., Bacarian, T., Tammam, J., Ren, B., Farmer, P. M., Kalef-Ezra, J., Micca, P. L., Nawrocky, M. M., Niederer, J. A., Recksiek, F. P., Fuchs, A. & Rosen, E. M. (2002). Response of rat intracranial 9L gliosarcoma to microbeam radiation therapy. *Neuro* 

Brewington, L. J., Tammam, J. & Rosen, E. M. (2003). Murine EMT-6 carcinoma: high therapeutic efficacy of microbeam radiation therapy. *Radiat Res*, Vol. 159, No.

*cancer*. Totowa, NJ, Humana Press.

*Oncologist*, Vol. 15, No., pp. 360-371.

*Phys*, Vol. 8, No. 3-4, pp. 531-534.

*Oncol*, Vol. 4, No. 1, pp. 26-38.

5, pp. 632-641.

*Radiation Biology*, Vol. 61, No., pp. 69-81.

DNA. *Free Radic Biol Med*, Vol. 49, No. 1, pp. 9-21.

cell cultures. *Cancer Res*, Vol. 53, No. 6, pp. 1332-1337.

Escherichia coli. *Nucleic Acids Res*, Vol. 6, No. 11, pp. 3673-3684.

single-stranded nicking. *J Theor Biol*, Vol. 127, No. 2, pp. 229-245.

quantitative data. *Mutagenesis*, Vol. 19, No. 3, pp. 169-185.

induction. *Carcinogenesis*, Vol. 30, No. 10, pp. 1686-1695.

570.

pp. 886-894.

inhibition activity of the Hoechst ligands in the relaxation assay might indicate their interaction with the binding of topoisomerase I with DNA. In the cleavage assay, a DNA SSB is induced that is attributed to the trapping by Hoechst ligands of topoisomerase I cleavable complexes (Chen et al., 1993; Bailly, 2000).
