The influence of the distance between the anchoring site of the tethered [Ru(TAP)(2)dip](2+) complex (TAP=1,4,5,8-tetraazaphenanthrene; dip=4,7-diphenyl-1,10-phenanthroline) on a probe sequence and the guanines of the complementary target strand was studied by (1) the luminescence quenching of the complex (by electron transfer) and (2) the oligodeoxyribonucleotide adduct (ODN adduct) formation which results in photo-crosslinking of the two strands. Moving the guanine moieties away from the complex induces an important decrease of the efficiency of both processes, but clearly affects the ODN adduct formation more specifically than the quenching process. From these results, we determined the positions of the guanine bases in the duplex ODN that are able to form a photo-adduct with the tethered complex. We also examined the possible competition between a long-range hole migration in the duplex ODN and the formation of a photo-adduct by using a sequence labeled with the complex at the 5'-phosphate end. Such a hole migration appears to be inefficient as compared to the ODN adduct formation. Finally, we studied the influence of the photo-crosslinking on the function of two different DNA polymerases. A 17-mer Ru(II)-labeled ODN was hybridized to its complementary sequence located on the 5'-side of a 40-mer matrix. After illumination, the elongation of a 13-mer DNA primer hybridized to the 3'-extremity of the same matrix was stopped at a position corresponding to the formation of the ODN adduct.

The formation of a photoadduct between a [Ru(1,4,5,8-tetraazaphenanthrene)(2)4,7-diphenylphenanthroline](2+) complex chemically attached to a synthetic oligonucleotide, and a guanine moiety in a complementary targeted single-stranded DNA molecule was studied for ten 17-mer duplexes by denaturing gel electrophoresis. This photoadduct formation leads to photocrosslinking of the two strands. The percentage quenching of luminescence of the complex by electron transfer was compared to the resulting yield of photocrosslinked product. This yield does not only depend on the ionisation potential of the guanine bases, which are electron donors, but also on other factors, such as the position of the guanine bases as compared to the site of attachment of the complex. The photocrosslinking yield is higher when the guanine moieties are towards the 3' end on the complementary strand as compared to the tethering site. Computer modelling results are in agreement with this preference for the 3' side for the photoreaction. Interestingly, the photocrosslink is not alkali labile. Moreover, a type III exonuclease enzyme is blocked at the position of photocrosslinking.

The luminescence quenching by electron transfer of a nonintercalating [Ru(TAP) 2 DIP] 2+ complex, which is covalently linked to different guanine containing oligonucleotide duplexes, is compared to the sequence dependent ionization potentials of guanines as estimated from Hartree-Fock calculations on regular B-DNA by means of Koopmans theorem. From these experimental and theoretical results, it is demonstrated that the ionization potential of the guanines has a major influence on the efficiency of hole injection by the groove binding [Ru(TAP) 2 DIP] 2+ complex in a type I photooxidation of DNA. Because the photoelectron transfer between the complex and DNA takes place on a nanosecond time scale, nanosecond molecular dynamics simulations were performed for two of the sequences in order to check for the possible influence of the dynamical fluctuations in the DNA structure on the ionization potential. IP fluctuations up to 0.5 eV were observed on a picosecond time scale along the molecular dynamics trajectories. However, these fluctuations scatter statistically around the value calculated for regular B-DNA and are thus not relevant for the slow photoelectron transfer with a nonintercalating [Ru(TAP) 2 DIP] 2+ complex.

The optically active cyclometalated Rh(III) complexes, delta[Rh(thpy4,5p(R,R)py)(2)TAP]Cl, lambda[Rh(thpy4,5p(S,S)py)(2)TAP]Cl, and delta[Rh(phpy4,5p(R,R)py)(2)TAP]Cl (where TAP = 1,4,5,8-tetraazaphenanthrene, thpy4,5p(R,R)py = (8R,10R)-2-(2'-thienyl)-4,5-pinenopyridine and phpy4,5p(R,R)py = (8R,10R)-2-(2'-phenyl)-4,5-pinenopyridine) have been prepared and characterized. Their photophysics has been examined in parallel with that of rac[Rh(thpy)(2)TAP]Cl and rac[Rh(phpy)(2)TAP]Cl. Their behaviors have been rationalized from results of TD-DFT calculations. The complexes with thienylpyridine (thpy) as cyclometalating ligands exhibit (3)CT (from thpy to TAP) and (3)LC(pi-pi) (centered on thpy) emissions in a solvent matrix at 77 K and one (3)CT luminescence at room temperature. In contrast, with phenylpyridine (phpy), the complexes show only one (3)CT emission (from phpy to TAP) at both temperatures.