Hydration structure and dynamics of B- and Z-DNA in the presence of counterions via molecular dynamics simulations

The Perfect Club Benchmarks: Effective Performance Evaluation of Supercomputers

This report presents a methodology for measuring the performance of supercomputers. It includes 13 Fortran programs that total over 50,000 lines of source code. They represent applications in several areas of engi neering and scientific computing, and in many cases the codes are currently being used by computational re search and development groups. We also present the PERFECT Fortran standard, a set of guidelines that allow portability to several types of machines. Furthermore, we present some performance measures and a method ology for recording and sharing results among diverse users on different machines. The results presented in this paper should not be used to compare machines, except in a preliminary sense. Rather, they are presented to show how the methodology has been applied, and to encourage others to join us in this effort. The results should be regarded as the first step toward our objec tive, which is to develop a publicly accessible data base of performance information of this type.

Reintroducing electrostatics into macromolecular crystallographic refinement: application to neutron crystallography and DNA hydration

Most current crystallographic structure refinements augment the diffraction data with a priori information consisting of bond, angle, dihedral, planarity restraints, and atomic repulsion based on the Pauli exclusion principle. Yet, electrostatics and van der Waals attraction are physical forces that provide additional a priori information. Here, we assess the inclusion of electrostatics for the force field used for all-atom (including hydrogen) joint neutron/X-ray refinement. Two DNA and a protein crystal structure were refined against joint neutron/X-ray diffraction data sets using force fields without electrostatics or with electrostatics. Hydrogen-bond orientation/geometry favors the inclusion of electrostatics. Refinement of Z-DNA with electrostatics leads to a hypothesis for the entropic stabilization of Z-DNA that may partly explain the thermodynamics of converting the B form of DNA to its Z form. Thus, inclusion of electrostatics assists joint neutron/X-ray refinements, especially for placing and orienting hydrogen atoms.

Effect of ionic strength on PNA-DNA hybridization on surfaces and in solution

Peptide nucleic acids (PNAs) are mimics of oligonucleotides containing a neutral peptidelike backbone and are able to bind complementary DNA targets with high affinity and selectivity. In order to investigate the effect of the ionic strength of the buffer solution, hybridization experiments with PNAs as (catcher) probes and DNAs as target oligonucleotides were performed in different salt solutions. Surface plasmon field-enhanced fluorescence spectroscopy was employed for real-time monitoring of DNA hybridizations to surface bound PNA. Probes with three different strand lengths were immobilized by self-assembly on the sensor surface. By introducing Cy5-labeled DNA targets the affinity constants, K(A)=k(on) (association)/k(off) (dissociation), were determined for fully complementary (MM0) as well as for single base mismatched (MM1) duplexes. Furthermore, the thermal stability of each duplex was determined by measuring melting curves in solution which was then compared to the kinetic and affinity parameters determined for the surface hybridization reactions. The results indicate that ions do not play a significant role for the PNA/DNA hybridization kinetics at surfaces. However, changes in the configuration of the PNA/DNA duplex due to the ionic strength variations influence the fluorescence yield drastically.

Brownian dynamics simulations of ion atmospheres around polyalanine and B-DNA: effects of biomolecular dielectric

We have extended an earlier Brownian dynamics simulation algorithm for simulating the structural dynamics of ions around biomolecules to accommodate dielectric inhomogeneity. The electrostatic environment of a biomolecule immersed in water was obtained by numerically solving the Poisson equation with the biomolecule treated as a low dielectric region and the solvent treated as a high dielectric region. Instead of using the mean-field type approximations of ion interactions as in the Poisson-Boltzmann model, the ions were treated explicitly by allowing them to evolve dynamically under the electrostatic field of the biomolecule. This model thus accounts for ion-ion correlations and the finite-size effects of the ions. For a 13-residue alpha-helical polyalanine and a 12-base-pair bp B-form DNA, we found that the choice of the dielectric constant of the biomolecule has much larger effects on the mean ionic structure around the biomolecule than on the fluctuational and dynamical properties of the ions surrounding the biomolecule.

Global scientific and engineering simulations on scalar, vector and parallel LCAP-type supercomputers

We present here an example of the global simulation approach to complex systems, selecting, as example, the study of a liquid, specifically water. We start by building the molecules of the liquid from nuclei and electrons using quantum mechanics. Next we obtain the interaction potentials (two, three and four body), again by quantum mechanics. Then we use Monte Carlo and molecular dynamics to study the motions of a water molecule within its Onsager sphere, and the collective properties; subsequently we overlap fluid dynamics by considering a flow along a channel with or without obstacles; finally we extend further and report a preliminary simulation of a Benard problem, using Newton’s equations. These simulations are performed on a parallel supercomputer which we have recently assembled; the system is briefly described in the second part of this work. A number of applications in science and engineering are analysed with attention to the degree of parallelization achievable with and without special hardwares such as busses and bulk shared memory. To conclude the implication of the global simulation methodology and supercomputers evolution is discussed in terms of productivity of information.

Molecular dynamics simulations of poly(dA).poly(dT): comparisons between implicit and explicit solvent representations

The program AMBER 3.0 has been used to generate molecular dynamics trajectories of a poly(dA).poly(dT) decamer. The simulations were performed using different methods to treat solvent effects. Results of a simulation including 18 counterions NH4+ and 4109 water molecules under (N, P, T) conditions were compared to simulation runs with implicit solvent representation in which solvent screening effects were represented by the use of a sigmoidal distance-dependent dielectric function. In the latter case, the system was simulated under microcanonical (N, V, E) and canonical (N, V, T) conditions. For the fully hydrated system simulation, a preequilibration protocol was developed since it was observed that long and progressive periods of heating and equilibration on the overall system were necessary in order to avoid energetic collisions between the solute and the solvent molecules, leading to severe irreversible deformation of the solute. A detailed analysis of DNA conformations, sugar puckers, and stability of the hydrogen bonds, Watson-Crick and three-center H bonds, is reported. The results show that DNA remains essentially in the B conformer with a tendency in the hydrated model to adopt a slightly distorted, unwound, and stretched conformation in comparison to standard B-DNA. Concerning sugar puckers, the mean pseudorotation phases of the adenine residues are systematically higher than those of the thymine residues, except in the case of the hydrated model for which a articular behavior is observed for the adenine strand. In this case, the terminal bases oscillate between C2'-endo and O4'-endo and the central ones stay in the C3'-endo domain. The mean lifetimes of the internal Watson-Crick H-bond (A) HN6...O4(T) are also dependent on the base pairs included in the calculation, excepted for the implicit solvent simulation at constant temperature. The three-center H bonds have very small mean lifetimes in all three cases of MD simulation. In the minor groove of the hydrated model, a spine of hydration is found as observed by x-ray crystallography and other theoretical simulations. On the basis of the rms deviations, it appears that the fully hydrated simulation has not reached a plateau at the end of the run, while the implicit simulation at constant energy seems to have converged. At constant temperature, very large oscillations in rms deviations are observed.

A molecular dynamics study of conformational changes and hydration of left-handed d(CGCGCGCGCGCG)2 in a nonsalt solution

Twelve dinucleotides (one complete turn) of left-handed, flexible, double-helix poly(dG-dC) Z-DNA have been simulated in aqueous solution with K+ counterions for 70 ps. Most of the d(GpC) phosphates have rotated in accordance with a ZI----ZII transition. The ZII conformation was probably partly stabilized by counterions, which coordinate one of the anionic oxygens and the guanine-N7 of the next (5'----3' direction) base. The presence of base-coordinating ions close to the helical axis rotated and pulled about half of the d(CpG) phosphates further into the groove. These ions also gave rise to rather large deviations from the crystal structure (ZI) with their tendency of pulling the bases closer toward the helical axis. A flipping of the orientation about the glycosyl bond from the +sc to the -sc region was observed for one guanosine, also leading to deviations from the crystal structure. Many bridges containing one or two water molecules were found, with a dominance for the latter. They essentially formed a network of intra- and interstrand bridges between anionic and esterified phosphate oxygens. A "spine" of water molecules could be distinguished as a dark zig-zag pattern in the water density map. The lifetime of a bridge containing one water was about twice as long as that of a two-water bridge and it lasted 5-15 times longer than a hydrogen bond in water. The lifetimes were also calculated for a selection of bridge types, in order of decreasing stability: O1P/O2P ... W ... O'4 much greater than O1P/O2P ... W ... guanine-N2 greater than O1P/O2P ... W ... O1P/O2P. The reorientational motion of water molecules in the first hydration shell around selected groups was slowed down considerably compared to bulk water and the decreasing order of correlation times was guanine-N2 greater than O'4 greater than O'3/O'5 greater than O1P/O2P.

Charge grouping approaches to calculation of electrostatic forces in molecular dynamics of macromolecules

Calculation of long-range electrostatic interactions is the most time-consuming step in theoretical simulation of the structure and dynamics of macromolecules. In practice very short cutoff distances are used, which may distort the behavior of the model system. We describe two accurate approaches to calculation of electrostatic forces based on hierarchical grouping of charges into cubes. The first is similar to the O(NlogN) algorithm developed by Barnes, J. and Hut, P., Nature (London) 324, 446-449 (1986), for simulation of a gravitational motion of N bodies. The second approach we formulate for a system with periodic boundary conditions in the nearest image approximation. The calculation of electrostatic interactions and a charge grouping procedure are faster than O(N2). The average inaccuracy in the force introduced by the grouping does not exceed 1%. We describe a small modification of the same approach which makes it suitable for long strongly charged polymers as well. This accurate approach to calculation of electrostatic interactions is illustrated with an example of the dynamics of ions near DNA. Quick equilibration of the ionic distribution is observed during molecular dynamics simulation if electrostatic forces are properly calculated, while the behavior and distribution of ions are less realistic when the conventional cutoff distances are used.

A molecular dynamics study of the effect of G.T mispairs on the conformation of DNA in solution

The effect of G.T mispair incorporation into a double-helical environment was examined by molecular dynamics simulation. The 60-ps simulations performed on the two hexanucleotide duplexes d (G3C3)2 and d(G3TC2)2 included 10 Na+ counterions and first hydration shell waters. The resulting backbone torsional angle trajectories were analyzed to select time spans representative of conformational domains. The average backbone angles and helical parameters of the last time span for both duplexes are reported. During the simulation the hexamers retained B-type DNA structures that differed from typical A- or B-DNA forms. The overall helical structures for the two duplexes are vary similar. The presence of G.T mispairs did not alter the overall helical structure of the oligonucleotide duplex. Large propeller twist and buckle angles were obtained for both duplexes. The purine/pyrimidine crossover step showed a large decrease in propeller twist in the normal duplex but not in the mismatch duplex. Upon the formation of wobble mispairs in the mismatched duplex, the guanines moved into the minor groove and the thymines moved into the major groove. This helped prevent purine/purine clash and created a deformation in the relative orientation of the glycosidic bonds. It also exposed the free O4 of the thymines in the major groove and N2 of the guanines in the minor groove to interactions with solvent and counterions. These factors seemed to contribute to the apparently higher rigidity of the mismatched duplex during the simulation.

Molecular dynamics simulation of the hydration shell of a B-DNA decamer reveals two main types of minor-groove hydration depending on groove width

The conformation of the self-complementary B-DNA decamer C-C-A-A-C-G-T-T-G-G is known from a high-resolution x-ray crystal structure analysis. Molecular dynamics simulation of the hydration shell of the decamer has revealed two main types of minor-groove hydration, depending on groove width. The narrow part of the minor groove has a spine of hydration analogous to that described for the A + T-rich center of the minor groove in the dodecamer C-G-C-G-A-A-T-T-C-G-C-G [Drew, H. R. & Dickerson, R. E. (1981) J. Mol. Biol. 151, 535-556], the first hydration layer of which contains one water molecular per base pair. In contrast, in the wide part of the minor groove, each base is hydrated individually, water molecules lying predominantly in the base plane. In intermediate-width regions, preferred water-molecule sites are shifted away from the base plane in a 3'-to-5' direction. This shift becomes more pronounced as the minor groove narrows, until the two water molecules lie approximately midway between base pairs. If the minor groove is narrowed still further, it accommodates only one water molecule, and the hydration transforms to the well-known water spine. The observed pattern agrees with available crystallographic data and with our earlier calculations. The results confirm the assumption that preferred positions of water oxygens in the minor groove depend predominantly on groove width rather than on base sequence. However, the location of water hydrogens, and the network of hydrogen bonding, can depend on base sequence. We suggest a simple explanation of water-spine formation in the narrow minor groove of a random DNA sequence. The spine of hydration may be a property of the minor groove of overwound variants of B-DNA, the C and D forms, for which the middle part of the decamer C-C-A-A-C-G-T-T-G-G can serve as a model.

Counterion effects on the physical properties and the A to B transition of calf-thymus DNA films

We report measurements of the water content and swelling of wet-spun films of Na-, K-, Rb-, and Cs-DNA as a function of relative humidity (rh). The water contents (number of water molecules per base pair of DNA) of these films are found to be quite similar, indicating that the identity of the counterion species is unimportant for the water content. Since the A to B transition for these DNAs occurs at different rhs, the corresponding water contents of the A to B transition are found to be quite different. These films swell as a function of rh in a very similar manner, implying that the changes in the intermolecular bonds in the different DNAs are similar. Dramatic changes are observed in the dimensions of the films above 84% rh for all types of DNA. Combining the water content and swelling measurements yields the dependence of the volume per base pair on water content. The volume per base pair is observed to be a nonlinear function of water content, indicating nonideal mixing of the water with the DNA.

Dielectric study on hydration of B-, A-, and Z-DNA

Dielectric relaxation peak due to bound water was found around 100 MHz in poly(dG-dC).poly(dG-dC) and calf thymus DNA in water-ethanol mixtures with NaCl buffer. Relaxation time and strength show a transition for poly(dG-dC).poly(dG-dC) at an ethanol composition Cw = 0.45 (w/w) where the structural transition from B- to Z-DNA takes place. It has been suggested that the transition is caused by removal of the bound water molecules preferentially from the phosphate groups. If the bound water molecules are removed equally from the phosphate groups and the grooves, the structural transition from B to A takes place. By analogy with hydration of tropocollagen, it was found that 19 water molecules per one nucleotide are at least necessary to keep B-DNA. Thirteen molecules are bound to A-DNA and 9 molecules to Z-DNA. Stringlike multimers are proposed as available structures of the bound water.

Effects of nucleotide bromination on the stabilities of Z-RNA and Z-DNA: a molecular mechanics/thermodynamic perturbation study

The structures of ZI- and ZII-form RNA and DNA oligonucleotides were energy minimized in vacuum using the AMBER molecular mechanics force field. Alternating C-G sequences were studied containing either unmodified nucleotides, 8-bromoguanosine in place of all guanosine residues, 5-bromocytidine in place of all cytidine residues, or all modified residues. Some molecules were also energy minimized in the presence of H2O and cations. Free energy perturbation calculations were done in which G8 and C5 hydrogen atoms in one or two residues of Z-form RNAs and DNAs were replaced in a stepwise manner by bromines. Bromination had little effect on the structures of the energy-minimized molecules. Both the minimized molecular energies and the results of the perturbation calculations indicate that bromination of guanosine at C8 will stabilize the Z forms of RNA and DNA relative to the nonbrominated Z form, while bromination of cytidine at C5 stabilizes Z-DNA and destabilizes Z-RNA. These results are in agreement with experimental data. The destabilizing effect of br5C in Z-RNAs is apparently due to an unfavorable interaction between the negatively charged C5 bromine atom and the guanosine hydroxyl group. The vacuum-minimized energies of the ZII-form oligonucleotides are lower than those of the corresponding ZI-form molecules for both RNA and DNA. Previous x-ray diffraction, nmr, and molecular mechanics studies indicate that hydration effects may favor the ZI conformation over the ZII form in DNA. Molecular mechanics calculations show that the ZII-ZI energy differences for the RNAs are greater than three times those obtained for the DNAs. This is due to structurally reinforcing hydrogen-bonding interactions involving the hydroxyl groups in the ZII form, especially between the guanosine hydroxyl hydrogen atom and the 3'-adjacent phosphate oxygen. In addition, the cytidine hydroxyl oxygen forms a hydrogen bond with the 5'-adjacent guanosine amino group in the ZII-form molecule. Both of these interactions are less likely in the ZI-form molecule: the former due to the orientation of the GpC phosphate away from the guanosine ribose in the ZI form, and the latter apparently due to competitive hydrogen bonding of the cytidine 2'-hydroxyl hydrogen with the cytosine carbonyl oxygen in the ZI form. The hydrogen-bonding interaction between the cytidine hydroxyl oxygen and the 5'-adjacent guanosine amino group in Z-RNA twists the amino group out of the plane of the base. This may be responsible for differences in the CD and Raman spectra of Z-RNA and Z-DNA.

Orientational and rotational velocity correlation functions for water-hydrating B- and Z-DNA double helices

The molecular dynamics simulations reported earlier for the structure and dynamics of water molecules hydrating B- and Z-DNA double helices are analyzed for the orientational correlation functions and the proton rotational velocity autocorrelation functions. The spectra of the rotational velocity autocorrelation functions obtained from the simulation results are compared with the neutron inelastic scattering experiments on hydrated Na-DNA samples. The results predict a small frequency component associated with water molecules bound to the double helices that disappears for waters away from the double helix.

A theoretical study of the aqueous hydration of canonical B d(CGCGAATTCGCG): Monte Carlo simulation and comparison with crystallographic ordered water sites

Monte Carlo computer simulation is described for the dodecamer d(CGCGAATTCGCG) together with 1777 water molecules at an environmental density of 1 gm/cc in a cubic cell under periodic boundary conditions. Water-water interactions were treated using the TIP4P potential and the solute water interactions by TIP4P spliced with the non-bonded interactions from the AMBER 3.0 force field. The stimulation was subjected to proximity analysis to obtain solute coordination numbers and pair interaction energies for each solute atom. Hydration density distributions partitioned into contributions from the major groove side, the minor groove side and the sugar-phosphate backbone were examined, and the probabilities of occurence for one- and two-water bridges in the simulation were enumerated. The results were compared with observations of crystallographic ordered water sites from x-ray diffraction studies on the native dodecamer by Dickerson and coworkers.

Hydration forces between parallel DNA double helices: computer simulations

We have performed two molecular dynamics computer simulations of two 10-base-pair segments of DNA molecules immersed in water. The goal of these simulations is to study the structural and dynamical properties of water between the DNA molecules. We have observed water ordering next to DNA surfaces. Existence of such ordering was proposed earlier by Marcelja and Radic [Marcelja, S. & Radic, N. (1976) Chem. Phys. Lett. 42, 129-130] to explain strong hydration forces between macromolecular surfaces.

Conformational and helicoidal analysis of 30 PS of molecular dynamics on the d(CGCGAATTCGCG) double helix: "curves", dials and windows

A new procedure for the analysis of the structure and molecular dynamics of duplex DNA is introduced, in which comprehensive visualization of results and pattern recognition is greatly facilitated. The method involves determining the values of the conformational and helicoidal parameters for each structure entering the analysis using the method "Curves" developed by Lavery and Sklenar, J. Biomol. Str. Dyn. 6, 63 (1988), followed by a novel computer graphic display of the results. The graphic display is organized systematically using conformation wheels, or "dials", for each IUPAC torsional parameter and "windows" on the range of values assumed by the linear and angular helicoidal parameters, and is presented in a form isomorphous with the structure per se. The complete time evolution of the conformational and helicoidal parameters of a DNA double helix can then be depicted in a set of six composite figures. Dynamical aspects of helix bending are also subsumed in this analysis. The procedure is illustrated with an analysis of the structures of canonical A and B forms of DNA and the 300 degrees K native dodecamer duplex d(CGCGAATTCGCG). The "dials and windows" are then used for a comprehensive analysis of 30 psec of molecular dynamics on the dodecamer in the vicinity of a canonical B-DNA energy minimum. This involves presentation of the time evolution of 206 conformational and 230 helicoidal parameters for the dodecamer. A number of interesting structural features can be recognized in the analysis, including crankshaft motions, BI - BII transitions, sugar repuckerings, and a description of spontaneous helix bending at what corresponds to the 1 degrees and 2 degrees "hinge points" indicated in the crystal structure. Our approach is expected to be directly useful for critical analysis of the effects of various assumptions about force field parameters, hydration and electrostatic effects and thus contribute to the development of reliable simulation protocols for nucleic acid systems. Extension of the method to present differential changes in conformational and helicoidal parameters is expected to be valuable for the analysis of structural and molecular dynamics studies of the reorganization and adaptation of DNA on complexation with various drugs and regulatory proteins.

Solvent bridging sites in A- and B-DNA helices

Nucleotide hydration is important for the understanding of the stability of and the transitions between the different helical conformations of DNA. We have used energy minimization and geometric criteria in order to look for possible sites for solvent which can bridge more than one polar or charged atomic group on a nucleotide. Such bridging sites between phosphate groups have been seen experimentally and used to explain the A to B transition. We show that these phosphate bridging sites occur at energy minima around A-DNA but do not occur around B-DNA. We also find that there are further low energy bridging sites which depend on sequence and which enable the more economical hydration of the A form.