Mesoscopic models for RNA salt dependence and for oncogene probe design with locked nucleic acids

Abstract

The melting process of nucleic acids is suggested to be a first-order transition and has been described by different approaches over the years. One such approach is the Peyrard-Bishop (PB) model which mainly consists of describing the helix through the hydrogen bonds connecting each base pair and the intra-helix stacking interaction between adjacent bases. The PB model exhibited success and experimental accordance when applied to modified nucleic acids, such as Inosine and threose nucleic acid (TNA). Its computational feasibility and capability to derive intra-molecular parameters from melting temperatures provide us with a robust tool to reinterpret published experimental data and achieve new insights over hydrogen bonds and stacking interactions in oligonucleotides. Moreover, the PB model parameterization also allows us to use those derived parameters to predict the melting temperature of non-measured sequences. One missing parameterization for the PB model is RNA at low salt concentrations, due to the limited amount of published melting temperatures. Although the PB model was found to be largely independent of strand concentrations, it requires that all temperatures are provided at the same strand concentrations. We adapted the PB model to handle multiple strand concentrations and in this way, we were able to make use of an experimental set of temperatures to model the hydrogen bond and stacking interactions at low and intermediate sodium concentrations. For the parameterizations, we make a distinction between terminal and internal base pairs, and the resulting potentials were qualitatively similar as we obtained previously for DNA. The main difference from DNA parameters, was the Morse potentials at low sodium concentrations for terminal r(AU) which is stronger than d(AT), suggesting higher hydrogen-bond strength. Another open problem is the source of the stabilization provided by chemically modified base pairs. One such modification is the locked nucleic acid (LNA), which due to the methylene bridge addition in the sugar moiety mimics the conformation of an RNA helix. This change improves the overall stability of the helix and has been used in several applications such as in polymerase chain reaction (PCR), small interfering RNA (siRNA), and antisense oligonucleotides (ASOs). However, the source of the stability improvement is not clearly delineated yet. Studies have suggested a favorable change either in the entropy, enthalpy, or both, of the modified helix, which is mainly driven by an enhancement in the stacking interactions of neighboring base pairs. The major challenge lies in compile sets containing sufficient measurements of melting temperature at similar buffer conditions and in the model parameterization. Fortunately, due to LNA popularity, we were able to collect a data set of over 300 temperature measurements. We have derived a complete set of parameters for the LNA insertion in DNA sequences and contrarily from the previous assumptions we have found stronger hydrogen bonding in the modified base pairs and their stacking interactions have shown little change. A complete parameterization of LNA base pairs allows the optimization of their use in oncogene probes and other types of applications. Therefore, we used the parameters to predict all possible LNA insertions in oncogene variants of BRAF, KRAS and EGFR. The probes were selected from the pool of temperature predictions, synthesized and their melting temperature measured, the accuracy of the measurements with the predictions was of 1 ◦C.