Coarse-Grained Simulations of RNA and DNA Duplexes

Abstract

Although RNAs play many cellular functions, little is known about the dynamics and thermodynamics of these molecules. In principle, all-atom molecular dynamics simulations can investigate these issues, but with current computer facilities, these simulations have been limited to small RNAs and to short times. HiRe-RNA, a recently proposed high-resolution coarse-grained RNA that captures many geometric details such as base pairing and stacking, is able to fold RNA molecules to near-native structures in a short computational time. So far, it had been applied to simple hairpins, and here we present its application to duplexes of a couple dozen nucleotides and show how with replica exchange molecular dynamics (REMD) we can easily predict the correct double helix from a completely random configuration and study the dissociation curve. To show the versatility of our model, we present an application to a double stranded DNA molecule as well. A reconstruction algorithm allows us to obtain full atom structures from the coarse-grained model. Through atomistic molecular dynamics (MD), we can compare the dynamics starting from a representative structure of a low temperature replica or from the experimental structure, and show how the two are statistically identical, highlighting the validity of a coarse-grained approach for structured RNAs and DNAs.

A Coarse-Grained DNA Model Parameterized from Atomistic Simulations by Inverse Monte Carlo

“…A number of theoretical coarse-graining models of DNA [24][25][26][27][28][29][30][31][32][33][34][35] and chromatin [13][14][15][36][37][38][39][40][41][42][43][44][45][46][47][48][49] have been reported recently. In recent works from the Papoian group, a CG model of a single DNA molecule was developed [24][25][26]50].…”

Multiscale simulation of DNA

“…Due to the reduction in the number of particles, and hence in the degrees of freedom, systems containing dozens to thousands of bp have been successfully simulated ( Table 1, phenomenological properties of a particular system. Under both approximations, the vast majority of the models use classical terms found in all-atom force fields, namely harmonic potentials for bonded terms [124•,127-130,142], and one or more expression (mostly Lenard-Jones, Coulomb, but also ad hoc potentials [129]) to reproduce non-bonded interactions (see Table 1 for a global view).…”

“…Pasquali and Derreumaux extended their RNA model named HiRE-RNA to dsDNA [129], model to reproduce DNA folding starting from two short single-stranded oligomers [128]; and Aksimentiev and coworkers [148] who published a model for ssDNA based on IBI from MD simulations, refined by fine tuning the parameters to reproduce experimentally measured radii of gyration.…”

“…In this regard, several coarse-grained models were applied to understand certain processes of biological interest, like duplex formation starting from short ssDNA oligomers [127,131,132,142]. Other models have shown to be useful in reproducing the structure of DNA/RNA duplexes [129], or structural properties such as the sequencedependent/salt-dependent persistence length [127,131,132], or the DNA curvature. The model from Vercauteren and coworkers was able to reproduce the topology of several DNA mini-circles with different link numbers, simulated using explicit K + Cl -ions [126].…”

RNA structure prediction: Progress and perspective

“…[89,90] The base pairing is modeled by hydrogen bonding interactions consisting of 2-, 3-and 4-body terms and the involved interactions include canonical Watson-Crick and wobble base pairs as well as the relatively rare A-C, A-G and U-C pairs. The model has been built in molecular dynamics simulations to study the stability of RNA hairpins [89] and duplexes [90] . However, the parameters of the model may need further adjustment to match the predicted thermodynamic properties of RNAs to the experimental data.…”