We have analyzed and compared the molecular structures and dynamics of DNA duplexes containing a nick or a gap of one nucleotide where the base in front of the gap is a guanine. The continuous strand has the sequence 5'(CAGAGTCXCTGGCTC) where the residue X is absent for the nick, 14-mer, and where it is a G residue for the gap. Duplexes were formed with the two corresponding 7-mers. Neither of these is phosphorylated adjacent at the nick site, but it is a good model for a single strand break. For the nick structure, the quantitative NMR data show that the global conformation is very close to canonical B-form DNA, but it displays enhanced local flexibility. For the gap structure, we observe only one species in which the extra G is well stacked into the helix. The two half-helices around this residue also show a B-form conformation. As with the nick duplex, the adjacent G imino protons show enhanced exchange with solvent. The gap does not close completely. Using distance constraints, MD calculations show that the nick conformation is very close to a duplex with no lesion but is indeed more flexible in the central part. The gapped structure shows two families of conformations. One is close to B-DNA, the other is significantly kinked at the gap which reduces the size of the cavity. We observe a spine of hydration within the cavities, similar, but of different geometry in the two cases.

In an attempt to target short purine sequences in view of pharmacological application, we have synthesized three new TFO (triple-helix-forming oligonucleotide) conjugates in which an intercalating oxazolopyridocarbazole (OPC) chromophore is linked by a pentamethylene linker to a 7-mer oligonucleotide matching the polypurine/polypyrimidine sequence located in the HIV-1 U3 LTR end region. The TFO moiety of conjugates are 5'CCTTCCC, 5'GGGAAGG, and 5'GGGTTGG. Their ability to bind to double-stranded DNA targets was examined. This binding is demonstrated by a footprinting technique using DNase I as a cleaving agent. The complex involved intermolecular pyr-pur*pyr or pur-pur*pyr triple helix. Pyrimidine TFO-OPC binds in a pH-dependent manner, whereas the others do not. The formation of the complex has been investigated at neutral pH and increasing temperature. We observed that the protection due to the purine and mixed TFO-OPC was pH independent and remained identical up to 40 degrees C. To determine the position of the OPC chromophore, molecular modeling was undertaken on the purine-conjugate/target complex. It has been suggested that the complex involved the intercalation of the OPC at the triplex-duplex junction with a small unwinding at the next excluded site.

No abstract

We investigated the behaviour of a 15mer DNA duplex, [5 H d(CAGAGTCACTGGCTC)3 H ].[5 H d(GAGCCAG)3 H + 5 H d(GACTCTG)3 H ] which contained an adenine opposite the gap. Analysis of the NMR data showed the existence of one major species, which was in equilibrium with two minor species. Their relative concentrations varied as a function of pH with a pK a of <4.5. For the major species, the duplex was globally in B conformation with the central adenine stacked in the helix. The two G´C base pairs adjacent to the central adenine were well formed and a gap was present in front of this adenine. For the minor species, major structural perturbations occurred in the centre of the duplex. At neutral pH, the central adenine was involved in a G´A mismatch with G23 adjacent to the gap. Cytosine C7 was then extrahelical and no gap was observed. Under these conditions, the major neutral species corresponded to 70% of the total and the minor species to 30%. At acidic pH, the central adenine of the minor species was protonated and was involved in a G(syn)´A + (anti) mismatch. The difference is that C9 is now extrahelical and G22 is implicated in the mispair. Three-dimensional models were built to initiate molecular dynamic simulations, which were in good agreement with the NMR data. Their structural stability in terms of hydrogen bonding and their flexibility are discussed and the biological significance for the interaction with DNA polymerase is evoked.Keywords: DNA; gap; molecular dynamics; NMR; repair.DNA duplexes containing gaps of one nucleotide may occur as a result of damaging processes, but also as an intermediate during a DNA repair mechanism. The damage product is the result of an oxidative attack on DNA, which has two main origins: normal aerobic cellular metabolism and ionizing radiation [1]. Oxidative damage to DNA is a frequent event: it occurs up to 150 000 times per day per cell (105 times per min per cell; [2]). In aerobic metabolism and after ionizing radiation, the hydroxyl radical is the most important species to reacts with and modify the DNA structure [3,4]. The nature of the damage product depends on the reaction site of the hydroxyl radical on the DNA [5,6]. Hydroxyl attack on the sugars at the C1 H atom leads to base release (abasic sites); attack at the C4 H atom or on the phosphate backbone may lead to single-strand breaks (nicks) or loss of a single nucleotide (gap). Finally, hydroxyl attack on the bases leads to base damage or, ultimately, base release. In these different cases, O-H, phosphate groups or modified sugars are adjacent to the gap The proteins or protein±DNA complexes involved in this mechanism have been the subject of several recent structural or biochemical reports [16±18]. These reports showed that a particular structure of the DNA duplex during the complexation with DNA polymerase b is required for nucleotide insertion. The base facing the gap must be stacked over its neighbouring base in the 5 H direction. This position of DNA in the active site of DNA polymerase is necessary for plac...

We have studied the binding of the hybrid netropsin-flavin (Net-Fla) molecule onto four sequences containing four A. T base pairs. Molecular mechanics minimizations in vacuo show numerous minimal conformations separated by one base pair. 400 ps molecular dynamics simulations in vacuo have been performed using the lowest minima as the starting conformations. During these simulations, the flavin moiety of the drug makes two hydrogen bonds with an amino group of a neighboring guanine. A 200 ps molecular dynamics simulation in explicit water solution suggests that the binding of Net-Fla upon the DNA substrate is enhanced by water bridges. A water molecule bridging the amidinium of Net-Fla to the N3 atom of an adenine seems to be stuck in the drug-DNA complex during the whole simulation. The fluctuations of the DNA helical parameters and of the torsion angles of the sugar-phosphate backbone are very similar in the simulations in vacuo and in water. The time auto-correlation functions for the DNA helical parameters decrease rapidly in the picosecond range in vacuo. The same functions computed from the water solution molecular dynamics simulations seem to have two modes: the rapid mode is similar to the behavior in vacuo, and is followed by a slower mode in the 10 ps range.

Triple helix formation by oligonucleotides can be extended beyond polypurine tracts with the help of specially designed linkers. In this paper we focus our attention on the integrase-binding site of the HIV-1 virus located on the U5 LTR end which contains two adjacent purine tracts on opposite strands. Two alternate triple helices with a 3'-3' junction in the third strand are considered: 5'-GGTTTTp3'-3'pTGTGT-5' and 5'-GGAAAAp3'-3'pAGAGA-5' The structural plausibility of these triplexes is investigated using molecular mechanics and dynamics simulations, both in vacuo and in aqua. The non-isomorphism of the triplets in the GpT steps in the first sequence, gives rise to non canonical conformations in the torsion angles, hydration appears to be crucial for this triplex. Sugar puckers are predominantly South during in vacuo simulations while they turn East in aqua. In the simulation in aqua the triplexes are shrouded by an hydration shell, however, we have not been able to detect any permanent hydrogen bond bridge between DNA and water. The solvation of ions as well as their radial distribution, appear to be relatively well behaved despite the artifacts known to be generated by the simulation procedure. The experimental feasibility of these structures is discussed.

In an attempt to obtain sequence specific DNA-cleaving molecules, we have synthesized a series of hybrid minor groove binders composed of a photoactiveable isoalloxazine (flavin) chromophore linked through a polymethylenic chain to a bis-pyrrolecarboxamide moiety related to netropsin. Like netropsin, the hybrid derivatives preferentially bind to A+T-rich sequences. Activation of the flavin chromophore by visible light results in the appearance of single strand breaks in the vicinity of the DNA binding site. We have further investigated the cleavage affinity properties of one of these compounds referred to as netropsin-flavin (Net-Fla) and considered as representative of the series. Net-Fla cleaves only one strand at a specific locus downstream of 5'-AAAT-3', upstream of 5'-TAAA-3' and on either side of a 5'-AAAA-3' sequence. Net-Fla cleaves both strands downstream to 5'-AATT-3'. This makes the properties of Net-Fla similar to that of a restriction endonuclease and provides additional insight into establishing the rules for the readout of B-DNA helix by non-nucleotidic compounds. Using molecular modeling, we show that Net-Fla binds to an asymmetric site in one orientation. The values of the energetic minima lie in the same order as expected from the cleavage patterns, which suggests that the oriented cleavage is a consequence of a sequence-oriented binding of Net-Fla in the DNA minor groove.

No abstract