Análise da estabilidade termodinâmica e de parâmetros estruturais de DNA e RNA por modelos mesoscópicos

Abstract

Mesoscopic models, like the one proposed by Peyrard and Bishop, are important for the study of the thermal stability of DNA and RNA. Its computational simplicity allows to analyse these oligonucleotides over longer time scales of physiological and experimental importance which are not accessible by more sophisticated techniques. Recently, there has been progress in the parametrization of canonical base pairs for this model. However, as they are more stable than non-canonical pairs, these bases are relatively easy to model. Our study aimed at applying the PB model beyond the canonical bases, as well as to investigate structural parameters. The first system we analysed was guanine-uracil (GU) in RNA. GU has an important biological role, acting as a recognition site for biomolecules and is also the most common non-complementarybase pair in RNA. Due to the non-isostericity of GU, its stability is influenced by the sequence context. GU has several conformational possibilities and may even have different hydrogen bonds depending on neighbouring bases. To consider all context possibilities for GU in RNA it becomes necessary to deal with a very large number of parameters. This problem is dealt with by optimizing the parameters correlatingwith melting temperature data from the literature. We developed a strategy to reduce the parameter search space which allowed us to determine GU configuration groups sorted by hydrogen bond intensity. When comparing with experimental results from the literature we obtain a good agreement for the hydrogen bonds, especially from NMR. Therefore, we not only were able to conclude the parametrization of GU inRNA but also to demonstrate that the technique allows to predict hydrogen bonds for non-canonical base pairs. In the second subject we approached a limitation of the 2D Peyrard-Bishop Hamiltonian which is the lack of structural parameters. We showed that it is possible to start from a helicoidal Hamiltonian and obtain a modified 2D Hamiltonian which includes information about the helix step, technically known asrise. In this study we also use the optimization method where we correlate melting temperature data from the literature to obtain the values for rise in DNA. To perform the comparison of our results with those obtained from experimental measurements we developed a query tool which accesses the Nucleic Acids Database (NDB) and selectssequences of interest to obtain values of rise coming from X-ray and NMR. In general our results agreed well with those from the NDB except for AT followed by TA in DNA. Furthermore, we were able to study the variation of rise with salt concentration. Our results show the possibility of performing structural studies in oligonucleotides using melting temperatures.