Statistical method for A-RNA and B-DNA

TL;DR

This paper presents a mesoscopic Hamiltonian model for nucleic acids that captures local flexibility and fluctuations, enabling accurate predictions of their elastic and thermodynamic properties at short scales, including twist-stretch behavior and cyclization.

Contribution

It introduces a three-dimensional discrete model with path integral formalism to analyze short-scale flexibility and responses of ds-DNA and ds-RNA, improving upon elastic rod models.

Findings

DNA overtwists under force, RNA untwists, due to structural differences.

Model accurately predicts twist-stretch relations for DNA and RNA.

Enhanced estimation of molecule bendability at short length scales.

Abstract

Nucleic acids have been regarded as stiff polymers with long-range flexibility and generally modeled using elastic rod models of polymer physics. Notwithstanding, investigations carried out over the past few years on single fragments of order $\sim 100$ base pairs have revealed remarkable flexibility properties at short scales and called for theoretical approaches that emphasize the role of the bending fluctuations at single sites along the molecule stack. Here, we review a three dimensional mesoscopic Hamiltonian model which assumes a discrete representation of the double stranded (ds) molecules at the level of the nucleotides. The model captures the fundamental local interactions between adjacent sugar-phosphate groups and the pairwise interactions between complementary base pair mates. A statistical method based on the path integral formalism sets the ensemble of the base pair breathing fluctuations which are included in the partition function and permits to derive the thermodynamics and the elastic response of single molecules to external forces. We apply the model to the computation of the twist-stretch relations for fragments of ds-DNA and ds-RNA, showing that the obtained opposite pattern (DNA overtwists whereas RNA untwists versus force) follows from the different structural features of the two helices. Moreover, we focus on the DNA stretching due to the confinement in nano-pores and, finally, on the computation of the cyclization probability of open ends molecules of $\sim 100$ base pairs under physiological conditions. The mesoscopic model shows a distinct advantage over the elastic rod model in estimating the molecule bendability at short length scale.