Following earlier reports of radioprotection of cells by Hoechst 33342, we have investigated radioprotection of isolated DNA by the minor groove binders Hoechst 33258 and Hoechst 33342. Analysis of radiation-induced single strand breakage in plasmid DNA (pBR322) showed concentration-dependant protection, up to a dose-modifying factor of 9.3 for 25 microM Hoechst 33258, at which the ligand: bp ratio was 0.67. Since the ligands bind at discrete sites along DNA, sequencing gel analysis was used to investigate the radioprotective effects of the ligands both at and between the ligand-binding sites. These experiments showed that although protection was more pronounced at the binding sites, there was also some reduction in strand-breakage between binding sites. Detailed analysis at a particular site, the EcoR1 site in a 3'-32P-endlabelled 100bp restriction fragment from pBR322, showed that protection was most pronounced at the 'inner T': GAATTC. Irradiation of a synthetic oligodeoxynucleotide containing a single ligand-binding site, and labelled at the 5'-end, gave the expected doublet bands in high resolution gels, corresponding to fragments with 3'-phosphoryl- and 3'-phosphorylglycollate terminii. In the Hoechst 33258-protected sample, the 3'-phosphorylglycollate band was preferentially suppressed within the binding site. These results, together with published crystal structure data for a Hoechst 33258/dodecamer complex, suggest that the site-specific radioprotection may be due to H-atom donation from the benzimidazole NH groups in the ligand to radiation-induced radicals on 4'-deoxyribosyl carbons. In contrast to the experiments with purified DNA, in which the two ligands yielded similar results, Hoechst 33342 was a much more active radioprotector in experiments with intact cells. For 20 microM Hoechst 33342, the dose-modifying factor was 1.7 at 1% survival and 1.3 at 10% survival, whereas the same level of Hoechst 33258 yielded barely detectable protection, perhaps due to a demonstrably lower cellular uptake. Presumably the radioprotection of cells by Hoechst 33342 is due to suppression of DNA strand breakage, and further investigation of the protection mechanism(s) should enable development of improved radioprotectors.

The rationale of boron (10B) neutron capture therapy (BNCT) is based on the high thermal neutron capture cross section of 10B and the limited maximum range (about one cell diameter) of the high LET fission products of the boron neutron capture (NC) reaction. The resulting radiochemical damage is confined to the cell containing the BNC reaction. Although other nuclides have higher thermal neutron capture cross sections than 10B, NC by such nuclides results in the emission of highly penetrating gamma rays. However, gadolinium-157 (157Gd) n-gamma reaction is also accompanied by some internal conversion and, by implication, Auger electron emission. Irradiation of Gd3+-DNA complexes with thermal neutrons results in the induction of DNA double-strand (ds) breaks, but the effect is largely abrogated in the presence of EDTA. Thus, by analogy with the effects of decay of Auger electron-emitting isotopes such as 125I, the Gd NC event must take place in the close proximity of DNA in order to induce a DNA ds break. It is proposed that 157Gd-DNA ligands therefore have potential in NCT. The thermal neutron capture cross section of 157Gd, a nonradioactive isotope, is more than 50 times that of 10B.

Irradiation of plasmid DNA/Gd3+ mixtures with thermal neutrons induces DNA double-strand breaks (dsb). However, the extent of breakage is markedly reduced by sequestering the Gd3+ from DNA by addition of EDTA. Since the 157Gd neutron capture event involves some internal conversion, we suggest that the DNA dsb induction results from Auger electron emission.

A double stranded oligodeoxynucleotide containing a single 125I-dC in a defined location was used to investigate DNA strand breakage resulting from 125I decay. Samples of a 41 bp oligodeoxynucleotide were incubated in 20 mM phosphate buffer (PB), or PB plus 2 M dimethylsulphoxide (DMSO), at 4 degrees C during 18-20 days. The 32P-5'-end labelled DNA fragments produced by 125I decays were separated on denaturing polyacrylamide gels, and the 32P activity in each fragment was determined by scintillation counting after elution of fragments from gel. Most of the breaks, around 90%, occurred within 4-5 nucleotides of the 125I-dC, but DNA breaks were detected up to 16 nucleotides from the decay site. The 125I-dC was located at the 21st nucleotide from the 32P-5'-end label, and since 32P was not detected in fragments longer than 20 nucleotides, it was assumed that all 125I decay events produce at least one break in the 125I-labelled DNA strand. The results show a considerable protection effect of DMSO on DNA breaks at sites >5-6 nucleotides from the 125I location. The probability of breaks in this region was decreased with DMSO by a factor of 2 to 8-fold, suggesting significant role for radical-mediated DNA breaks at the more distant sites. However, the total protection effect of DMSO is rather small: 1.1, because of the small contribution of breakage at distant sites to the total yield.

An iodinated DNA ligand, iodo Hoechst 33258, which binds in the minor groove of DNA, enhances DNA strand breakage and cell killing by UV-A irradiation. The sites of UV-induced strand breaks reflect the known sequence specificity of the ligand.