Theoretical studies of an exceptionally stable RNA tetraloop: observation of convergence from an incorrect NMR structure to the correct one using unrestrained molecular dynamics

Accurate Modeling of RNA Hairpins Through the Explicit Treatment of Electronic Polarizability with the Classical Drude Oscillator Force Field

Molecular dynamics (MD) simulations play a crucial role in modeling biomolecular systems in which the electrostatic interactions are critical in dictating the structural and dynamical properties. Thus, the treatment of the electrostatic interactions defined in the underlying force field (FF) strongly affects the simulation accuracy. Most FFs use fixed partial atomic charges to include electrostatic interactions, and therefore lack the electronic polarization response, representing an intrinsic limitation. To address this limitation, polarizable FFs have been developed that treat atomic polarizabilities explicitly. Here we present the application of the all-atom polarizable (Drude) and non-polarizable (CHARMM) nucleic acid FFs in RNA hairpin systems to investigate the impact of polarization on structural properties, dipole moment distributions, and cation interactions. Results show that the presence of polarizability in the FF significantly improves the stabilization of RNA hairpin structure. As expected, the distributions of dipole moments show more fluctuations when simulated using the polarizable FF, with the variation in dipoles contributing to the stabilization of the structures of the loop regions of the RNAs. Contact map analyses between the bases and cations show that the variation of the ion distribution around the entire hairpin is larger for the polarizable FF and the cations occupy the outer hydration shell to a greater extent. The presented results indicate the importance of the explicit treatment of electronic polarizability in molecular simulations of RNA, including in non-canonical regions.

RNA Structural Dynamics As Captured by Molecular Simulations: A Comprehensive Overview

With both catalytic and genetic functions, ribonucleic acid (RNA) is perhaps the most pluripotent chemical species in molecular biology, and its functions are intimately linked to its structure and dynamics. Computer simulations, and in particular atomistic molecular dynamics (MD), allow structural dynamics of biomolecular systems to be investigated with unprecedented temporal and spatial resolution. We here provide a comprehensive overview of the fast-developing field of MD simulations of RNA molecules. We begin with an in-depth, evaluatory coverage of the most fundamental methodological challenges that set the basis for the future development of the field, in particular, the current developments and inherent physical limitations of the atomistic force fields and the recent advances in a broad spectrum of enhanced sampling methods. We also survey the closely related field of coarse-grained modeling of RNA systems. After dealing with the methodological aspects, we provide an exhaustive overview of the available RNA simulation literature, ranging from studies of the smallest RNA oligonucleotides to investigations of the entire ribosome. Our review encompasses tetranucleotides, tetraloops, a number of small RNA motifs, A-helix RNA, kissing-loop complexes, the TAR RNA element, the decoding center and other important regions of the ribosome, as well as assorted others systems. Extended sections are devoted to RNA-ion interactions, ribozymes, riboswitches, and protein/RNA complexes. Our overview is written for as broad of an audience as possible, aiming to provide a much-needed interdisciplinary bridge between computation and experiment, together with a perspective on the future of the field.

Physics-based all-atom modeling of RNA energetics and structure

The database of RNA sequences is exploding, but knowledge of energetics, structures, and dynamics lags behind. All-atom computational methods, such as molecular dynamics, hold promise for closing this gap. New algorithms and faster computers have accelerated progress in improving the reliability and accuracy of predictions. Currently, the methods can facilitate refinement of experimentally determined nuclear magnetic resonance and x-ray structures, but are 'unreliable' for predictions based only on sequence. Much remains to be discovered, however, about the many molecular interactions driving RNA folding and the best way to approximate them quantitatively. The large number of parameters required means that a wide variety of experimental results will be required to benchmark force fields and different approaches. As computational methods become more reliable and accessible, they will be used by an increasing number of biologists, much as x-ray crystallography has expanded. Thus, many fundamental physical principles underlying the computational methods are described. This review presents a summary of the current state of molecular dynamics as applied to RNA. It is designed to be helpful to students, postdoctoral fellows, and faculty who are considering or starting computational studies of RNA. WIREs RNA 2017, 8:e1422. doi: 10.1002/wrna.1422.

Simultaneous NMR characterisation of multiple minima in the free energy landscape of an RNA UUCG tetraloop

RNA molecules in solution tend to undergo structural fluctuations of relatively large amplitude and to populate a range of different conformations some of which with low populations. It is still very challenging, however, to characterise the structures of these low populated states and to understand their functional roles. In the present study, we address this problem by using NMR residual dipolar couplings (RDCs) as structural restraints in replica-averaged metadynamics (RAM) simulations. By applying this approach to a 14-mer RNA hairpin containing the prototypical UUCG tetraloop motif, we show that it is possible to construct the free energy landscape of this RNA molecule. This free energy landscapes reveals the surprisingly rich dynamics of the UUCG tetraloop and identifies the multiple substates that exist in equilibrium owing to thermal fluctuations. The approach that we present is general and can be applied to the study of the free energy landscapes of other RNA or RNA-protein systems.

Free Energy Landscape of GAGA and UUCG RNA Tetraloops

We report the folding thermodynamics of ccUUCGgg and ccGAGAgg RNA tetraloops using atomistic molecular dynamics simulations. We obtain a previously unreported estimation of the folding free energy using parallel tempering in combination with well-tempered metadynamics. A key ingredient is the use of a recently developed metric distance, eRMSD, as a biased collective variable. We find that the native fold of both tetraloops is not the global free energy minimum using the Amberχ_OL3 force field. The estimated folding free energies are 30.2 ± 0.5 kJ/mol for UUCG and 7.5 ± 0.6 kJ/mol for GAGA, in striking disagreement with experimental data. We evaluate the viability of all possible one-dimensional backbone force field corrections. We find that disfavoring the gauche⁺ region of α and ζ angles consistently improves the existing force field. The level of accuracy achieved with these corrections, however, cannot be considered sufficient by judging on the basis of available thermodynamic data and solution experiments.

What NMR Relaxation Can Tell Us about the Internal Motion of an RNA Hairpin: A Molecular Dynamics Simulation Study

Classical molecular dynamics simulations of a 14-mer UUCG RNA hairpin are performed to study its conformational dynamics and corresponding NMR relaxation parameters. The direct calculation of the relaxation rates from the trajectory yields good agreement with experiment, indicating the validity of the theoretical model. Various ways to provide a link between theory and experiment are considered, including the "model-free approach" of Lipari and Szabo and Gaussian axial fluctuation model of Brüschweilwer. It is studied if the underlying assumptions of these approaches are satisfied in the case of a flexible RNA hairpin. Being consistent with the analysis of the NMR experiments, Lipari-Szabo fits of the first 100 or 1000 ps of the internal correlation functions lead to a nice agreement between calculated and experimental order parameters and internal correlation times. Finally, the relation between NMR order parameters and the underlying internal motion of the RNA hairpin is discussed in detail. A principal component analysis reveals that the principal motions of the molecule account only partially for the measured NMR order parameters, because the latter are insensitive to internal dynamics occurring on a nanosecond time scale due to molecular tumbling.

Propensities for loop structures of RNA & DNA backbones

RNA oligonucleotides exhibit a large tendency to bend and form a loop conformation which is a major motif contributing to their complex three-dimensional structure. This is in contrast to DNA molecules that predominantly form the double-helix structure. In this paper we investigate by molecular dynamics simulation, as well as, by its combination with the replica-exchange method, the propensity of RNA chains containing the GCUAA pentaloop to form spontaneously a hairpin conformation. The results were then compared with those of analogous hybrid oligonucleotides in which the ribose groups in the loop-region were substituted by deoxyriboses. We find that the RNA oligomers exhibit a marginal excess stability to form loop structures. The equilibrium constant for opening the loop to an extended conformation is twice as large in the hybrid than it is in the RNA chain. Analyses of the hydrogen bonds indicate that the excess stability for forming a hairpin is a result of hydrogen bonds the 2'-hydroxyls in the loop region form with other groups in the loop. Of these hydrogen bonds, the most important is the hydrogen bond donated from the 2'-OH at the first position of the loop to N7 of adenine at the forth position. RNA and DNA backbones are characterized by different backbone dihedral angles and sugar puckering that can potentially facilitate or hamper the hydrogen bonds involving the 2'-OH. Nevertheless, the sugar puckerings of all the pentaloop nucleotides were not significantly different between the two chains displaying the C3'-endo conformation characteristic to the A-form double helix. All of the other backbone dihedrals also did not show any considerable difference in the loop-region except of the δ-dihedral. In this case, the RNA loop exhibited bimodal distributions corresponding to, both, the RNA and DNA backbones, whereas the loop of the hybrid chain behaved mostly as that of a DNA backbone. Thus, it is possible that the behavior of the δ-dihedrals in the loop-region of the RNA adopts conformations that facilitate the intra-nucleotide hydrogen bondings of the 2'-hydroxyls, and consequently renders loop structures in RNA more stable.

Ribose 2'-Hydroxyl Groups Stabilize RNA Hairpin Structures Containing GCUAA Pentaloop

The chemical structure of RNA and DNA is very similar; however, the three-dimensional conformation of these two nucleic acids is very different. Whereas the DNA adopts a repetitive structure of a double-stranded helix, RNA is primarily single stranded with a complex three-dimensional structure in which the hairpin is the most common secondary structure. Apart from the difference between uracil and thymine, the difference in the chemical structure between RNA and DNA is the presence of a hydroxyl group at position 2' of the sugar (ribose) instead of a hydrogen (deoxyribose). In this paper, we present molecular dynamics simulations addressing the contribution of 2'-hydroxyls to the stability of a GCUAA pentaloop motif. The results indicate that the 2'-hydroxyls stabilize the hairpin conformation of the GCUAA pentaloop relative to an analogous oligonucleotide in which the ribose sugars in the loop region were substituted with deoxyriboses. The magnitude of the stabilization was found to be 23.8 ± 4.1 kJ/mol using an alchemical mutations free energy method and 4.2 ± 6.5 kJ/mol using potential of mean force calculations. The latter indicates that in addition to its larger thermodynamic stability the RNA hairpin is also kinetically more stable. We find that the excess stability is a result of intrahairpin hydrogen bonds in the loop region between the 2'-hydroxyls and sugars, bases, and phosphates. The hydrogen bonds with the sugars and phosphates involve predominantly interactions with adjacent nucleotides. However, the hydrogen bonds with the bases involve also interactions between groups on opposite sides of the loop or with the middle base of the loop and are therefore likely to contribute significantly to the stability of the loop. Of these hydrogen bonds, the most frequent is observed between the 2'-hydroxyl at the first position of the pentaloop with N6/N7 of adenine at the forth position, as well as between the 2'-hydroxyl at position -1 with N6 of adenine at the fifth position. Our results contribute to the notion that one of the important roles of the ribose sugars in RNA is to facilitate hairpin formation.

A molecular-modeling toolbox aimed at bridging the gap between medicinal chemistry and computational sciences

In the current era of high-throughput drug discovery and development, molecular modeling has become an indispensable tool for identifying, optimizing and prioritizing small-molecule drug candidates. The required background in computational chemistry and the knowledge of how to handle the complex underlying protocols, however, might keep medicinal chemists from routinely using in silico technologies. Our objective is to encourage those researchers to exploit existing modeling technologies more frequently through easy-to-use graphical user interfaces. In this account, we present two innovative tools (which we are prepared to share with academic institutions) facilitating computational tasks commonly utilized in drug discovery and development: (1) the VirtualDesignLab estimates the binding affinity of small molecules by simulating and quantifying their binding to the three-dimensional structure of a target protein; and (2) the MD Client launches molecular dynamics simulations aimed at exploring the time-dependent stability of ligand–protein complexes and provides residue-based interaction energies. This allows medicinal chemists to identify sites of potential improvement in their candidate molecule. As a case study, we present the application of our tools towards the design of novel antagonists for the FimH adhesin.

Free-energy landscape of RNA hairpins constructed via dihedral angle principal component analysis

To systematically construct a low-dimensional free-energy landscape of RNA systems from a classical molecular dynamics simulation, various versions of the principal component analysis (PCA) are compared: the cPCA using the Cartesian coordinates of all atoms, the dPCA using the sine/cosine-transformed six backbone dihedral angles as well as the glycosidic torsional angle chi and the pseudorotational angle P, the aPCA which ignores the circularity of the 6 + 2 dihedral angles of the RNA, and the dPCA(etatheta), which approximates the 6 backbone dihedral angles by 2 pseudotorsional angles eta and theta. As representative examples, a 10-nucleotide UUCG hairpin and the 36-nucleotide segment SL1 of the Psi site of HIV-1 are studied by classical molecular dynamics simulation, using the Amber all-atom force field and explicit solvent. It is shown that the conformational heterogeneity of the RNA hairpins can only be resolved by an angular PCA such as the dPCA but not by the cPCA using Cartesian coordinates. Apart from possible artifacts due to the coupling of overall and internal motion, this is because the details of hydrogen bonding and stacking interactions but also of global structural rearrangements of the RNA are better discriminated by dihedral angles. In line with recent experiments, it is found that the free energy landscape of RNA hairpins is quite rugged and contains various metastable conformational states which may serve as an intermediate for unfolding.

High-resolution NMR structure of an RNA model system: the 14-mer cUUCGg tetraloop hairpin RNA

We present a high-resolution nuclear magnetic resonance (NMR) solution structure of a 14-mer RNA hairpin capped by cUUCGg tetraloop. This short and very stable RNA presents an important model system for the study of RNA structure and dynamics using NMR spectroscopy, molecular dynamics (MD) simulations and RNA force-field development. The extraordinary high precision of the structure (root mean square deviation of 0.3 A) could be achieved by measuring and incorporating all currently accessible NMR parameters, including distances derived from nuclear Overhauser effect (NOE) intensities, torsion-angle dependent homonuclear and heteronuclear scalar coupling constants, projection-angle-dependent cross-correlated relaxation rates and residual dipolar couplings. The structure calculations were performed with the program CNS using the ARIA setup and protocols. The structure quality was further improved by a final refinement in explicit water using OPLS force field parameters for non-bonded interactions and charges. In addition, the 2'-hydroxyl groups have been assigned and their conformation has been analyzed based on NOE contacts. The structure currently defines a benchmark for the precision and accuracy amenable to RNA structure determination by NMR spectroscopy. Here, we discuss the impact of various NMR restraints on structure quality and discuss in detail the dynamics of this system as previously determined.

RNA phosphodiester backbone dynamics of a perdeuterated cUUCGg tetraloop RNA from phosphorus-31 NMR relaxation analysis

We have analyzed the relaxation properties of all (31)P nuclei in an RNA cUUCGg tetraloop model hairpin at proton magnetic field strengths of 300, 600 and 900 MHz in solution. Significant H, P dipolar contributions to R (1) and R (2) relaxation are observed in a protonated RNA sample at 600 MHz. These contributions can be suppressed using a perdeuterated RNA sample. In order to interpret the (31)P relaxation data (R (1), R (2)), we measured the (31)P chemical shift anisotropy (CSA) by solid-state NMR spectroscopy under various salt and hydration conditions. A value of 178.5 ppm for the (31)P CSA in the static state (S (2) = 1) could be determined. In order to obtain information about fast time scale dynamics we performed a modelfree analysis on the basis of our relaxation data. The results show that subnanosecond dynamics detected around the phosphodiester backbone are more pronounced than the dynamics detected for the ribofuranosyl and nucleobase moieties of the individual nucleotides (Duchardt and Schwalbe, J Biomol NMR 32:295-308, 2005; Ferner et al., Nucleic Acids Res 36:1928-1940, 2008). Furthermore, the dynamics of the individual phosphate groups seem to be correlated to the 5' neighbouring nucleobases.

Conformationally restricted nucleotides as a probe of structure-function relationships in RNA

We introduce the use of commercially available locked nucleic acids (LNAs) as a functional probe in RNA. LNA nucleotides contain a covalent linkage that restricts the pseudorotation phase of the ribose to C3'-endo (A-form). Introduction of an LNA at a single site thus allows the role of ribose structure and dynamics in RNA function to be assessed. We apply LNA probing at multiple sites to analyze self-cleavage in the lead-dependent ribozyme (leadzyme), thermodynamic stability in the UUCG tetraloop, and the kinetics of recognition of U1A protein by U1 snRNA hairpin II. In the leadzyme, locking a single guanosine residue into the C3'-endo pucker increases the catalytic rate by a factor of 20, despite the fact that X-ray crystallographic and NMR structures of the leadzyme ground state reported a C2'-endo conformation at this site. These results strongly suggest that a conformational change at this position is critical for catalytic function. Functional insights obtained in all three systems demonstrate the highly general applicability of LNA probing in analysis of the role of ribose orientation in RNA structure, dynamics, and function.

NMR and MD studies of the temperature-dependent dynamics of RNA YNMG-tetraloops

In a combined NMR/MD study, the temperature-dependent changes in the conformation of two members of the RNA YNMG-tetraloop motif (cUUCGg and uCACGg) have been investigated at temperatures of 298, 317 and 325 K. The two members have considerable different thermal stability and biological functions. In order to address these differences, the combined NMR/MD study was performed. The large temperature range represents a challenge for both, NMR relaxation analysis (consistent choice of effective bond length and CSA parameter) and all-atom MD simulation with explicit solvent (necessity to rescale the temperature). A convincing agreement of experiment and theory is found. Employing a principle component analysis of the MD trajectories, the conformational distribution of both hairpins at various temperatures is investigated. The ground state conformation and dynamics of the two tetraloops are indeed found to be very similar. Furthermore, both systems are initially destabilized by a loss of the stacking interactions between the first and the third nucleobase in the loop region. While the global fold is still preserved, this initiation of unfolding is already observed at 317 K for the uCACGg hairpin but at a significantly higher temperature for the cUUCGg hairpin.

The conformational landscape of the ribosomal protein S15 and its influence on the protein interaction with 16S RNA

The interaction between the ribosomal protein S15 and its binding sites in the 16S RNA was examined from two points of view. First, the isolated protein S15 was studied by comparing NMR conformer sets, available in the PDB and recalculated using the CNS-ARIA protocol. Molecular dynamics (MD) trajectories were then recorded starting from a conformer of each set. The recalculation of the S15 NMR structure, as well as the recording of MD trajectories, reveals that several orientations of the N-terminal alpha-helix alpha1 with respect to the structure core are populated. MD trajectories of the complex between the ribosomal protein S15 and RNA were also recorded in the presence and absence of Mg(2+) ions. The Mg(2+) ions are hexacoordinated by water and RNA oxygens. The coordination spheres mainly interact with the RNA phosphodiester backbone, reducing the RNA mobility and inducing electrostatic screening. When the Mg(2+) ions are removed, the internal mobility of the RNA and of the protein increases at the interaction interface close to the RNA G-U/G-C motif as a result of a gap between the phosphate groups in the UUCG capping tetraloop and of the disruption of S15-RNA hydrogen bonds in that region. On the other hand, several S15-RNA hydrogen bonds are reinforced, and water bridges appear between the three-way junction region and S15. The network of hydrogen bonds observed in the loop between alpha1 and alpha2 is consequently reorganized. In the absence of Mg(2+), this network has the same pattern as the network observed in the isolated protein, where the helix alpha1 is mobile with respect to the protein core. The presence of Mg(2+) ions may thus play a role in stabilizing the orientation of the helix alpha1 of S15.

Structure and dynamics of an RNA tetraloop: a joint molecular dynamics and NMR study

Molecular dynamics simulations of the RNA tetraloop 5'-CGCUUUUGCG-3' with high melting temperature and significant conformational heterogeneity in explicit water solvent are presented and compared to NMR studies. The NMR data allow for a detailed test of the theoretical model, including the quality of the force field and the conformational sampling. Due to the conformational heterogeneity of the tetraloop, high temperature (350 K) and locally enhanced sampling simulations need to be invoked. The Amber98 force field leads to a good overall agreement with experimental data. Based on NMR data and a principal component analysis of the 350 K trajectory, the dynamic structure of the tetraloop is revealed. The principal component free energy surface exhibits four minima, which correspond to well-defined conformational structures that differ mainly by their base stacking in the loop region. No correlation between the motion of the sugar rings and the stacking dynamics of the loop bases is found.

Iron responsive element RNA flexibility described by NMR and isotropic reorientational eigenmode dynamics

The first example of the application of reorientational eigenmode dynamics (RED) to RNA is shown here for the small and floppy Iron Responsive Element (IRE) RNA hairpin. Order parameters calculated for bases and riboses from a 12 ns molecular dynamics trajectory are compared to experimentally determined order parameters from 13C-1H NMR relaxation experiments, and shown to be in qualitative agreement. Given the small size of the IRE hairpin and its very flexible loop, isotropic RED (iRED) was also used to analyze the trajectory in order to describe its dynamic motions. iRED analysis shows that the global and internal dynamics of the IRE are not rigorously separable, which will result in inaccurate experimental order parameters. In addition, the iRED analysis described the many correlated motions that comprise the dynamics of the IRE RNA. The combined use of NMR relaxation, RED, and iRED provide a uniquely detailed description of IRE RNA dynamics.

Modified replica exchange simulation methods for local structure refinement

Parallel tempering, also known as replica exchange molecular dynamics (REMD), has recently been successfully used to study the structure and thermodynamic properties of biomolecules such as peptides and small proteins. For large systems, however, applying REMD can be costly since the number of replicas needed increases as the square root of the number of degrees of freedom in the system. Often, enhanced sampling is only needed for a subset of atoms, such as a loop region of a large protein or a small ligand binding to a receptor. In such applications, it is often reasonable to assume a weak dependence of the structure of the larger region on the instantaneous conformation of the smaller region of interest. For these cases, we derived two variant replica exchange methods, partial replica exchange molecular dynamics (PREMD) and local replica exchange molecular dynamics (LREMD). The Hamiltonian for the system is separated, with replica exchange carried out only for terms involving the subsystem of interest while the remainder of the system is maintained at a single temperature. The number of replicas required for efficient exchange thus depends on the number of degrees of freedom in the fragment needing refinement rather than on the size of the full system. The method can be applied to much larger systems than was previously practical. This also provides a means to preserve the integrity of the structure outside the refinement region without introduction of restraints. LREMD takes this weak coupling approximation a step further, employing only a single representation of the large fragment that simultaneously interacts with all of the replicas of the subsystem of interest. This is obtained by combining replica exchange with the locally enhanced sampling approximation (LES), reducing the computational expense of replica exchange simulations to near that of a single standard molecular dynamics (MD) simulation. Use of LREMD also permits the use of LES without requiring the specification of a single temperature, a known difficulty for standard LES simulations. We tested these two methods on the loop region of an RNA hairpin model system and find significant advantages over standard MD and REMD simulations.

Effects of base substitutions in an RNA hairpin from molecular dynamics and free energy simulations

Contributions of individual interactions in the GGCGCAAGCC hairpin containing a GCAA tetraloop were studied by computer simulations using base substitutions. The G in the first tetraloop position was replaced by inosine (I) or adenosine (A), and the G in the C-G basepair closing the tetraloop was replaced by I. These substitutions eliminate particular hydrogen bonds proposed in the nuclear magnetic resonance model of the GCAA tetraloop. Molecular dynamics simulations of the GCAA tetraloop in aqueous solvent displayed a well-defined hydrogen pattern between the first and last loop nucleotides (G and A) stabilized by a bridging water molecule. Substitution of G-->I in the basepair closing the tetraloop did not significantly influence the loop structure and dynamics. The ICAA loop maintained the overall structure, but displayed variation in the hydrogen-bond network within the tetraloop itself. Molecular dynamics simulations of the ACAA loop led to conformational heterogeneity of the resulting structures. Changes of hairpin formation free energy associated with substitutions of individual bases were calculated by the free energy perturbation method. The calculated decrease of the hairpin stability upon G-->I substitution in the C-G basepair closing the tetraloop was in good agreement with experimental thermodynamic data. Our theoretical estimates for G-->I and G-->A mutations located in the tetraloop suggest larger loop destabilization than corresponding experimental results. The extent of conformational sampling of the structures resulting from base substitutions and its impact on the calculated free energy was discussed.

Generation of accurate protein loop conformations through low-barrier molecular dynamics

Prediction and refinement of protein loop structures are important and challenging tasks for which no general solution has been found. In addition to the accuracy of scoring functions, the main problems reside in (1) insufficient statistical sampling and (2) crossing energy barriers that impede conformational rearrangements of the loop. We approach these two issues by using "low-barrier molecular dynamics," a combination of energy smoothing techniques. To address statistical sampling, locally enhanced sampling (LES) is used to produce multiple copies of the loop, thus improving statistics and reducing energy barriers. We introduce a novel extension of LES that can improve local sampling even further through hierarchical subdivision of copies. Even though LES reduces energy barriers, it cannot provide for crossing infinite barriers, which can be problematic when substantial rearrangement of residues is necessary. To permit this kind of loop residue repacking, a "soft-core" potential energy function is introduced, so that atomic overlaps are temporarily allowed. We tested this new combined methodology to a loop in anti-influenza antibody Fab 17/9 (7 residues long) and to another loop in the antiprogesterone antibody DB3 (8 residues). In both cases, starting from random conformations, we were able to locate correct loop structures (including sidechain orientations) with heavy-atom root-mean-square deviation (fit to the nonloop region) of approximately 1.1 A in Fab 17/9 and approximately 1.8 A in DB3. We show that the combination of LES and soft-core potential substantially improves sampling compared to regular molecular dynamics. Moreover, the sampling improvement obtained with this combined approach is significantly better than that provided by either of the two methods alone.

3'-Exonuclease resistance of DNA oligodeoxynucleotides containing O6-[4-oxo-4-(3-pyridyl)butyl]guanine

Tobacco-specific nitrosamine, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), is a chemical carcinogen thought to be involved in the initiation of lung cancer in smokers. NNK is metabolically activated to methylating and pyridyloxobutylating species that form promutagenic adducts with DNA nucleobases, e.g. O(6)-[4-oxo-4-(3-pyridyl)butyl]guanine (O(6)-POB-dG). O(6)-POB-dG is a strongly mispairing DNA lesion capable of inducing both G-->A and G-->T base changes, suggesting its importance in NNK mutagenesis and carcinogenesis. Our earlier investigations have identified the ability of O(6)-POB-dG to hinder DNA digestion by snake venom phosphodiesterase (SVPDE), a 3'-exonuclease commonly used for DNA ladder sequencing and as a model enzyme to test nuclease sensitivity of anti-sense oligonucleotide drugs. We now extend our investigation to three other enzymes possessing 3'-exonuclease activity: bacteriophage T4 DNA polymerase, Escherichia coli DNA polymerase I, and E.coli exonuclease III. Our results indicate that, unlike SVPDE, 3'-exonuclease activities of these three enzymes are not blocked by O(6)-POB-dG lesion. Conformational analysis and molecular dynamics simulations of DNA containing O(6)-POB-dG suggest that the observed resistance of the O(6)-POB-dG lesion to SVPDE-catalyzed hydrolysis may result from the structural changes in the DNA strand induced by the O(6)-POB group, including C3'-endo sugar puckering and the loss of stacking interaction between the pyridyloxobutylated guanine and its flanking bases. In contrast, O(6)-methylguanine lesion used as a control does not induce similar structural changes in DNA and does not prevent its digestion by SVPDE.

DNA tri- and tetra-loops and RNA tetra-loops hairpins fold as elastic biopolymer chains in agreement with PDB coordinates

The biopolymer chain elasticity (BCE) approach and the new molecular modelling methodology presented previously are used to predict the tri- dimensional backbones of DNA and RNA hairpin loops. The structures of eight remarkably stable DNA or RNA hairpin molecules closed by a mispair, recently determined in solution by NMR and deposited in the PDB, are shown to verify the predicted trajectories by an analysis automated for large numbers of PDB conformations. They encompass: one DNA tetraloop, -GTTA-; three DNA triloops, -AAA- or -GCA-; and four RNA tetraloops, -UUCG-. Folding generates no distortions and bond lengths and bond angles of main atoms of the sugar-phosphate backbone are well restored upon energy refinement. Three different methods (superpositions, distance of main chain atoms to the elastic line and RMSd) are used to show a very good agreement between the trajectories of sugar-phosphate backbones and between entire molecules of theoretical models and of PDB conformations. The geometry of end conditions imposed by the stem is sufficient to dictate the different characteristic DNA or RNA folding shapes. The reduced angular space, consisting of the new parameter, angle Omega, together with the chi angle offers a simple, coherent and quantitative description of hairpin loops.

Conformational heterogeneity observed in simulations of a pyrene-substituted DNA

NMR studies previously carried out for a DNA system with a pyrene-substituted base did not observe NOEs involving the adenine located 5' to the pyrene, and thus the conformation of the adenine was poorly defined in the resulting family of refined structures. However, chemical shift data suggested that an AT base pair may be present. We have carried out fully unrestrained molecular dynamics simulations starting from several members of the family of structures, and these simulations support the existence of an AT base pair for this region. Simulations in both explicit and implicit solvent were carried out, with each converging to either anti or syn conformation for adenine and base pairing in all cases. During these simulations, large and dramatic conformational changes are observed that suggest pathways for complex conformational changes in the highly packed DNA interior. Our analysis reveals little difference in the energies of these syn and anti conformations, in contrast to control calculations carried out for standard DNA (in the absence of a neighboring pyrene). While no interconversion between the conformations was observed in standard simulations, reversible anti/syn exchange was directly simulated using the locally enhanced sampling approach. No exchange was seen in the non-pyrene control sequence. Together, these results suggest that an increased flexibility is introduced as a consequence of the pyrene substitution, offering an explanation that is consistent with the available experimental data. These results increase our optimism that simulations in atomic detail may provide accurate models for experimental observations in complex systems.

Molecular dynamics applied to nucleic acids

In this Account, we focus on molecular dynamics (MD) simulations involving fully solvated nucleic acids. Historically, MD simulations were first applied to proteins and several years later to nucleic acids. The first MD simulations of DNA were carried out in vacuo, but nowadays fully solvated systems are common practice. Recently, technical improvements have made it possible to conduct accurate MD simulations of highly charged nucleic acids. The state-of-the-art of MD simulations and a number of applications on various nucleic acid systems are discussed.

Molecular dynamics simulation of nucleic acids: successes, limitations, and promise

In the last five years we have witnessed a significant increase in the number publications describing accurate and reliable all-atom molecular dynamics simulations of nucleic acids. This increase has been facilitated by the development of fast and efficient methods for treating the long-range electrostatic interactions, the availability of faster parallel computers, and the development of well-validated empirical molecular mechanical force fields. With these technologies, it has been demonstrated that simulation is not only capable of consistently reproducing experimental observations of sequence specific fine structure of DNA, but also can give detailed insight into prevalent problems in nucleic acid structure, ion association and specific hydration of nucleic acids, polyadenine tract bending, and the subtle environmental dependence of the A-DNA-B-DNA duplex equilibrium. Despite the advances, there are still issues with the methods that need to be resolved through rigorous controlled testing. In general, these relate to deficiencies of the underlying molecular mechanical potentials or applied methods (such as the imposition of true periodicity in Ewald simulations and the need for energy conservation), and significant limits in effective conformational sampling. In this perspective, we provide an overview of our experiences, provide some cautionary notes, and provide recommendations for further study in molecular dynamics simulation of nucleic acids.

Molecular dynamics simulations of the denaturation and refolding of an RNA tetraloop

Tetraloops are very abundant structural elements of RNA that are formed by four nucleotides in a hairpin loop which is closed by a double stranded helical stem with some Watson-Crick base pairs. A tetraloop r(GCGAAGGC) was identified from the crystal structure of the central domain of 16S rRNA (727-730) in the Thermus thermophilus 30S ribosomal complex. The crystal structure of the 30S complex includes a total of 104 nucleotides from the central domain of the 16S rRNA and three ribosomal proteins S6, S15 and S18. Independent biochemical experiments have demonstrated that protein S15 plays the role in initiating the formation of the central domain of this complex. In the crystal, the tetraloop interacts with the protein S15 at two sites: one of them is associated with hydrogen bond interactions between residue His50 and nucleotide G730, and the other is associated with the occurrence of residue Arg53 beside A728. This paper uses molecular dynamics (MD) simulations to investigate the protein-dependent structural stability of the tetraloop and demonstrates the folding pathway of this tetraloop via melting denaturation and its subsequent refolding. Three important results are derived from these simulations: (i) The stability of nucleotide A728 appears to be protein dependent. Without the interaction with S15, A728 flips away from stacking with A729. (ii) The melting temperature demonstrated by the simulations is analogous to the results of thermodynamic experiments. In addition, the simulated folding of the tetraloop is stepwise: the native shape of the backbone is formed first; this is then followed by the formation of the Watson- Crick base pairs in the stem; and finally the hydrogen bonds and base stacking in the loop are formed. (iii) The tetraloop structure is similar to the crystal structure at salt concentrations of 0.1 M and 1.0 M used for the simulations, but the refolded structure at 0.1 M salt is more comparable to the crystal structure than at 1.0 M. The results from the simulations using both the Generalized Born continuum model and explicit solvent model (Particle Mesh Ewald) generate a similar pathway for unfolding/refolding of the tetraloop.

Influence of the dynamic positions of cations on the structure of the DNA minor groove: sequence-dependent effects

Different models for minor groove structures predict that the conformation is essentially fixed by sequence and has an influence on local ion distribution or alternatively that temporal positions of ions around the minor groove can affect the structure if they neutralize cross-strand phosphate charges. Our previous studies show that the minor groove in an AATT dodecamer responds to local sodium ion positions and is narrow when ions neutralize cross-strand phosphate-phosphate charges [J. Am. Chem. Soc. 2000, 122, 10513-10520]. Previous results from a number of laboratories have shown that G-tracts often have a wider minor groove than A-tracts, but they do not indicate whether this is due to reduced flexibility or differences in ion interactions. We have undertaken a molecular dynamics study of a d(TATAGGCCTATA) duplex to answer this question. The results show that the G-tract has the same amplitude of minor groove fluctuations as the A-tract sequence but that it has fewer ion interactions that neutralize cross-strand phosphate charges. These results demonstrate that differences in time-average groove width between A- and G-tracts are due to differences in ion interactions at the minor groove. When ions neutralize the cross-strand phosphates, the minor groove is narrow. When there are no neutralizing ion interactions, the minor groove is wide. The population of structures with no ion interactions is larger with the GGCC than with the AATT duplex, and GGCC has a wider time-average minor groove in agreement with experiment.

RNase-stable RNA: conformational parameters of the nucleic acid backbone for binding to RNase T1

An RNA sequence showing high stability with respect to digestion by ribonuclease T1 (RNase T1) was isolated by in vitro selection from an RNA library. Although ribonuclease T1 cleaves single-stranded RNA specifically after guanosine residues, secondary structure calculations predict several guanosines in single-stranded areas. Two of these guanosines are part of a GGCA-tetraloop, a recurring structure element in the secondary structure predictions. Molecular dynamics simulations of the conformation space of the nucleotides involved in this tetraloop show on the one hand that the nucleic acid backbone of the guanosines cannot realise the conformation required for cleavage by RNase T1. On the other hand, it could be shown that an RNA molecule not forced into a tetraloop occupies this conformation several times in the course of the simulation. The simulations confirm the GGCA-tetraloop as an RNase-stable secondary structure element. Our results show that, besides the known prerequisite of a single-stranded RNA, RNase T1 has additional demands on the substrate conformation.

Molecular dynamics simulation of nucleic acids

We review molecular dynamics simulations of nucleic acids, including those completed from 1995 to 2000, with a focus on the applications and results rather than the methods. After the introduction, which discusses recent advances in the simulation of nucleic acids in solution, we describe force fields for nucleic acids and then provide a detailed summary of the published literature. We emphasize simulations of small nucleic acids ( approximately 6 to 24 mer) in explicit solvent with counterions, using reliable force fields and modern simulation protocols that properly represent the long-range electrostatic interactions. We also provide some limited discussion of simulation in the absence of explicit solvent. Absent from this discussion are results from simulations of protein-nucleic acid complexes and modified DNA analogs. Highlights from the molecular dynamics simulation are the spontaneous observation of A B transitions in duplex DNA in response to the environment, specific ion binding and hydration, and reliable representation of protein-nucleic acid interactions. We close by examining major issues and the future promise for these methods.

Molecular dynamics simulation reveals conformational switching of water-mediated uracil-cytosine base-pairs in an RNA duplex

A 4 ns molecular dynamics simulation of an RNA duplex (r-GGACUUCGGUCC)(2 )in solution with Na+ and Cl- as counterions was performed. The X-ray structure of this duplex includes two water-mediated uracil-cytosine pairs. In contrast to the other base-pairs in the duplex the water-mediated pairs switch between different conformations. One conformation corresponds to the geometry of the water-mediated UC pairs in the duplex X-ray structure with water acting both as hydrogen-bond donor and acceptor. Another conformation is close to that of a water-mediated UC base-pair found in the X-ray structure of the 23 S rRNA sarcin/ricin domain. In this case the oxygen of the water molecule is linked to two-base donor sites. For a very short time also a direct UC base-pair and a further conformation that is similar to the one found in the RNA duplex structure but exhibits an increased H3(U)...N3(C) distance is observed. Water molecules with unusually long residence times are involved in the water-mediated conformations. These results indicate that the dynamic behaviour of the water-mediated UC base-pairs differs from that of the duplex Watson-Crick and non-canonical guanine-uracil pairs with two or three direct hydrogen bonds. The conformational variability and increased flexibility has to be taken into account when considering these base-pairs as RNA building blocks and as recognition motifs.

Molecular dynamics studies of the HIV-1 TAR and its complex with argininamide

The dynamic behavior of HIV-1 TAR and its complex with argininamide is investigated by means of molecular dynamics simulations starting from NMR structures, with explicit inclusion of water and periodic boundary conditions particle mesh Ewald representation of the electrostatic energy. During simulations of free and argininamide-bound TAR, local structural patterns, as determined by NMR experiments, were reproduced. An interdomain motion was observed in the simulations of free TAR, which is absent in the case of bound TAR, leading to the conclusion that the free conformation of TAR is intrinsically more flexible than the bound conformation. In particular, in the bound conformation the TAR-argininamide interface is very well ordered, as a result of the formation of a U.A.U base triple, which imposes structural constraints on the global conformation of the molecule. Free energy analysis, which includes solvation contributions, was used to evaluate the influence of van der Waals and electrostatic terms on formation of the complex and on the conformational rearrangement from free to bound TAR.

Structure of the Sm binding site from human U4 snRNA derived from a 3 ns PME molecular dynamics simulation

A molecular dynamics simulation of the Sm binding site from human U4 snRNA was undertaken to determine the conformational flexibility of this region and to identify RNA conformations that were important for binding of the Sm proteins. The RNA was fully-solvated (>9,000 water molecules) and charge neutralized by inclusion of potassium ions. A three nanosecond MD simulation was conducted using AMBER with long-range electrostatic forces considered using the particle mesh Ewald summation method. The initial model of the Sm binding site region had the central and 3' stem-loops that flanked the Sm site co-axial with one another, and with the single-stranded Sm binding site region ([I] conformation). During the course of the trajectory, the axes of the 3' stem-loop, and later the central stem-loop, became roughly orthogonal from their original anti-parallel orientation. As these conformational changes occurred, the snRNA adopted first an [L] conformation, and finally a [U] conformation. The [U] conformation was more stable than either the [I] or [L] conformations, and persisted for the final 1 ns of the trajectory. Analysis of the structure resulting from the MD simulations revealed the bulged nucleotide, U114, and the mismatched Ag91-G110 base pair provided distinctive structural features that may enhance Sm protein binding. Based on the results of the MD simulation and the available experimental data, we proposed a mechanism for the binding of the Sm protein sub-complexes to the snRNA. In this model, the D1/D2 and E/F/G Sm protein sub-complexes first bind the snRNA in the [U] conformation, followed by conformational re-arrangement to the [I] conformation and binding of the D3/B Sm protein sub-complex.

The crystal structure of UUCG tetraloop

All large structured RNAs contain hairpin motifs made of a stem closed by several looped nucleotides. The most frequent loop motif is the UUCG one. This motif belongs to the tetraloop family and has the peculiarity of being highly thermodynamically stable. Here, we report the first crystal structure of two UUCG tetraloops embedded in a larger RNA-protein complex solved at 2.8 A resolution. The two loops present in the asymmetric unit are in a different crystal packing environment but, nevertheless, have an identical conformation. The observed structure is globally close to that obtained in solution by nuclear magnetic resonance. However, subtle differences point to a more detailed picture of the role played by 2'-hydroxyl groups in stabilising this tetraloop.

Simulation of the structure and dynamics of nonhelical RNA motifs

Computer simulation methods are increasingly being used to study possible conformations and dynamics of structural motifs in RNA. Recent results of molecular dynamics simulations and continum solvent studies of RNA structures and RNA-ligand complexes show promising agreement with experimental data. Combined with the ongoing progress in the experimental characterization of RNA structure and thermodynamics, these computational approaches can help to better understand the mechanism of RNA structure formation and the binding of ligands.

Experimental and computational studies of the G[UUCG]C RNA tetraloop

In prokaryotic ribosomal RNAs, most UUCG tetraloops are closed by a C-G base-pair. However, this preference is greatly reduced in eukaryotic rRNA species where many UUCG tetraloops are closed by G-C base-pairs. Here, biophysical properties of the C[UUCG]G and G[UUCG]C tetraloops are compared, using experimental and computational methods. Thermal denaturation experiments are used to derive thermodynamic parameters for the wild-type G[UUCG]C tetraloop and variants containing single deoxy substitutions in the loop. A comparison with analogous experiments on the C[UUCG]G motif shows that the two RNA species exhibit similar patterns in response to the substitutions, suggesting that their loop structures are similar. This conclusion is supported by NMR data that suggest that the essential UUCG loop structure is maintained in both tetraloops. However, NMR results show that the G[UUCG]C loop structure is disrupted prior to melting of the stem; this behavior is in contrast to the two-state behavior of the C[UUCG]G molecule. Stochastic dynamics simulations using the GB/SA continuum solvation model, run as a function of temperature, show rare conformational transitions in several G[UUCG]C simulations. These results lead to the conclusion that substitution of a G-C for a C-G closing base-pair increases the intrinsic flexibility of the UUCG loop.

Nucleic acids: theory and computer simulation, Y2K

Molecular dynamics simulations on DNA and RNA that include solvent are now being performed under realistic environmental conditions of water activity and salt. Improvements to force-fields and treatments of long-range interactions have significantly increased the reliability of simulations. New studies of sequence effects, axis bending, solvation and conformational transitions have appeared.

Experimental and theoretical studies of the effects of deoxyribose substitutions on the stability of the UUCG tetraloop

Experimental and theoretical thermodynamic studies of the consequences of 2'-hydroxyl substitution in the RNA UUCG tetraloop show distinct position dependence consistent with the diverse structural contexts of the four-loop ribose hydroxyls in this motif. The results suggest that even for simple substitutions, such as the replacement of the ribose hydroxyl (2'-OH) with hydrogen (2'-H), the free energy change reflects a complex interplay of hydrogen bonding and solvation effects and is influenced by the intrinsic pucker preferences of the nucleotides. Furthermore, theoretical studies suggest that the effect of these mutations in the single-strand state is sequence dependent, in contrast to what is commonly assumed. Free energy perturbation simulations of ribose-deoxyribose mutations in a single-strand dodecamer and in trinucleotide models suggest that in the denatured state, the magnitude of the free energy change for deoxyribose substitutions is determined to a larger extent by the identity of the nucleotide (A, C, G or U) rather than its structural context. Single-strand mutational effects must be considered when interpreting mutational studies in molecular terms.

A molecular dynamics simulation of the flavin mononucleotide-RNA aptamer complex

We report on an unrestrained molecular dynamics simulation of the flavin mononucleotide (FMN)-RNA aptamer. The simulated average structure maintains both cross-strand and intermolecular FMN-RNA nuclear Overhauser effects from the nmr experiments and has all qualitative features of the nmr structure including the G10-U12-A25 base triple and the A13-G24, A8-G28, and G9-G27 mismatches. However, the relative orientation of the hairpin loop to the remaining part of the molecule differs from the nmr structure. The simulation predicts that the flexible phosphoglycerol part of FMN moves toward G27 and forms hydrogen bonds. There are structurally long-lived water molecules in the FMN binding pocket forming hydrogen bonds within FMN and between FMN and RNA. In addition, long-lived water is found bridging primarily RNA backbone atoms. A general feature of the environment of long-lived "structural" water is at least two and in most cases three or four potential acceptor atoms. The 2'-OH group of RNA usually acts as an acceptor in interactions with the solvent. There are almost no intrastrand O2'H(n) vertical ...O4'(n + 1) hydrogen bonds within the RNA backbone. In the standard case the preferred orientation of the 2'-OH hydrogen atoms is approximately toward O3' of the same nucleotide. However, a relatively large number of conformations with the backbone torsional angle gamma in the trans orientation is found. A survey of all experimental RNA x-ray structures shows that this backbone conformation occurs but is less frequent than found in the simulation. Experimental nmr RNA aptamer structures have a higher fraction of this conformation as compared to the x-ray structures. The backbone conformation of nucleotide n + 1 with the torsional angle gamma in the trans orientation leads to a relatively short distance between 2'-OH(n) and O5'(n + 1), enabling hydrogen-bond formation. In this case the preferred orientation of the 2'-OH hydrogen atom is approximately toward O5'(n + 1). We find two relatively short and dynamically stable types of backbone-backbone next-neighbor contacts, namely C2'(H)(n) vertical ...O4'(n + 1) and C5'(H)(n + 1) vertical ...O2'(n). These interactions may affect both backbone rigidity and thermodynamic stability of RNA helical structures.

Molecular dynamics simulations of the complex between human U1A protein and hairpin II of U1 small nuclear RNA and of free RNA in solution

RNA-protein interactions are essential to a wide range of biological processes. In this paper, a 0.6-ns molecular dynamics simulation of the sequence-specific interaction of human U1A protein with hairpin II of U1 snRNA in solution, together with a 1.2-ns simulation of the free RNA hairpin, is reported. Compared to the findings in the x-ray structure of the complex, most of the interactions remained stable. The nucleotide U8, one of the seven conserved nucleotides AUUGCAC in the loop region, was unusually flexible during the simulation, leading to a loss of direct contacts with the protein, in contrast to the situation in the x-ray structure. Instead the sugar-phosphate backbone of nucleotide C15 was found to form several interactions with the protein. Compared to the NMR structure of U1A protein complexed with the 3'-untranslated region of its own pre-mRNA, the protein core kept the same conformation, and in the two RNA molecules the conserved AUUGCAC of the loop and the closest CG base pair were located in very similar positions and orientations, and underwent very similar interactions with the protein. Therefore, a common sequence-specific interaction mechanism was suggested for the two RNA substrates to bind to the U1A protein. Conformational analysis of the RNA hairpin showed that the conformational changes of the RNA primarily occurred in the loop region, which is just involved in the sites of binding to the protein and in agreement with experimental observation. Both the loop and stem of the RNA became more ordered upon binding to the protein. It was also demonstrated that the molecular dynamics method could be successfully used to simulate the dynamical behavior of a large RNA-protein complex in aqueous solution, thus opening a path for the exploration of the complex biological processes involving RNA at a molecular level.

Unrestrained stochastic dynamics simulations of the UUCG tetraloop using an implicit solvation model

Three unrestrained stochastic dynamics simulations have been carried out on the RNA hairpin GGAC[UUCG] GUCC, using the AMBER94 force field (Cornell et al., 1995. J. Am. Chem. Soc. 117:5179-5197) in MacroModel 5.5 (Mohamadi et al., 1990. J. Comp. Chem. 11:440-467) and either the GB/SA continuum solvation model (Still et al., 1990. J. Am. Chem. Soc. 112:6127-6129) or a linear distance-dependent dielectric (1/R) treatment. The linear distance-dependent treatment results in severe distortion of the nucleic acid structure, restriction of all hydroxyl dihedrals, and collapse of the counterion atmosphere over the course of a 5-ns simulation. An additional vacuum simulation without counterions shows somewhat improved behavior. In contrast, the two GB/SA simulations (1.149 and 3.060 ns in length) give average structures within 1.2 A of the initial NMR structure and in excellent agreement with results of an earlier explicit solvent simulation (Miller and Kollman, 1997. J. Mol. Biol. 270:436-450). In a 3-ns GB/SA simulation starting with the incorrect UUCG tetraloop structure (Cheong et al., 1990. Nature. 346:680-682), this loop conformation converts to the correct loop geometry (Allain and Varani, 1995. J. Mol. Biol. 250:333-353), suggesting enhanced sampling relative to the previous explicit solvent simulation. Thermodynamic effects of 2'-deoxyribose substitutions of loop nucleotides were experimentally determined and are found to correlate with the fraction of time the ribose 2'-OH is hydrogen bonded and the distribution of the hydroxyl dihedral is observed in the GB/SA simulations. The GB/SA simulations thus appear to faithfully represent structural features of the RNA without the computational expense of explicit solvent.

A modified version of the Cornell et al. force field with improved sugar pucker phases and helical repeat

We have examined some subtle parameter modifications to the Cornell et al. force field, which has proven quite successful in reproducing nucleic acid properties, but whose C2'-endo sugar pucker phase and helical repeat for B DNA appear to be somewhat underestimated. Encouragingly, the addition of a single V2 term involving the atoms C(sp3)-O-(sp3)-C(sp3)-N(sp2), which can be nicely rationalized because of the anomeric effect (lone pairs on oxygen are preferentially oriented relative to the electron withdrawing N), brings the sugar pucker phase of C2'-endo sugars to near perfect agreement with ab initio calculations (W near 162 degrees). Secondly, the use of high level ab initio calculations on entire nucleosides (in contrast to smaller model systems necessitated in 1994-95 by computer limitations) lets one improve the chi torsional potential for nucleic acids. Finally, the O(sp3)-C(sp3)- C(sp3)-O(sp3) V2 torsional potential has been empirically adjusted to reproduce the ab initio calculated relative energy of C2'-endo and C3'-endo nucleosides. These modifications are tested in molecular dynamics simulations of mononucleosides (to assess sugar pucker percentages) and double helices of DNA and RNA (to assess helical and sequence specific structural properties). In both areas, the modified force field leads to improved agreement with experimental data.