Improved Angle Potentials for Coarse-Grained Molecular Dynamics Simulations.

Potentials routinely used in atomistic molecular dynamics simulations are not always suitable for modeling systems at coarse-grained resolution. For example, in the calculation of traditional torsion angle potentials, numerical instability is often encountered in the case of very flexible molecules. To improve the stability and accuracy of coarse-grained molecular dynamics simulations, we propose two approaches. The first makes use of improved forms for the angle potentials: the restricted bending (ReB) potential prevents torsion angles from visiting unstable or unphysical configurations and the combined bending-torsion (CBT) potential smoothly flattens the interactions when such configurations are sampled. In the second approach, dummy-assisted dihedral (DAD), the torsion potential is applied differently: instead of acting directly on the beads, it acts on virtual beads, bound to the real ones. For simple geometrical reasons, the unstable region is excluded from the accessible conformational space. The benefits of the new approaches are demonstrated in simulations of polyethylene glycol (PEG), polystyrene (PS), and polypeptide molecules described by the MARTINI coarse-grained force field. The new potentials are implemented in an in-house version of the Gromacs package, publicly available.

In silico design of robust bolalipid membranes.

The robustness of microorganisms used in industrial fermentations is essential for the efficiency and yield of the production process. A viable tool to increase the robustness is through engineering of the cell membrane and especially by incorporating lipids from species that survive under harsh conditions. Bolalipids are tetraether lipids found in Archaea bacteria, conferring stability to these bacteria by spanning across the cytoplasmic membrane. Here we report on in silico experiments to characterize and design optimal bolalipid membranes in terms of robustness. We use coarse-grained molecular dynamics simulations to study the structure, dynamics, and stability of membranes composed of model bolalipids, consisting of two dipalmitoylphosphatidylcholine (DPPC) lipids covalently linked together at either one or both tail ends. We find that bolalipid membranes differ substantially from a normal lipid membrane, with an increase in thickness and tail order, an increase in the gel-to-liquid crystalline phase transition temperature, and a decrease in diffusivity of the lipids. By changing the flexibility of the linker between the lipid tails, we furthermore show how the membrane properties can be controlled. A stiffer linker increases the ratio between spanning and looping conformations, rendering the membrane more rigid. Our study may help in designing artificial membranes, with tunable properties, able to function under extreme conditions. As an example, we show that incorporation of bolalipids makes the membrane more tolerant toward butanol.

Immobilization of the plug domain inside the SecY channel allows unrestricted protein translocation.

The SecYEG complex forms a protein-conducting channel in the inner membrane of Escherichia coli to support the translocation of secretory proteins in their unfolded state. The SecY channel is closed at the periplasmic face of the membrane by a small re-entrance loop that connects transmembrane segment 1 with 2b. This helical domain 2a is termed the plug domain. By the introduction of pairs of cysteines and crosslinkers, the plug domain was immobilized inside the channel and connected to transmembrane segment 10. Translocation was inhibited to various degrees depending on the position and crosslinker spacer length. With one of the crosslinked mutants translocation occurred unrestricted. Biochemical characterization of this mutant as well as molecular dynamics simulations suggest that only a limited movement of the plug domain suffices for translocation.

Areas of monounsaturated diacylphosphatidylcholines.

We have studied the structural properties of monounsaturated diacylphosphatidylcholine lipid bilayers (i.e., diCn:1PC, where n = 14, 16, 18, 20, 22, and 24 is the number of acyl chain carbons). High-resolution x-ray scattering data were analyzed in conjunction with contrast-varied neutron scattering data using a technique we recently developed. Analyses of the data show that the manner by which bilayer thickness increases with increasing n in monounsaturated diacylphosphatidylcholines is dependent on the double bond's position. For commonly available monounsaturated diacylphosphatidylcholines, this results in the nonlinear behavior of both bilayer thickness and lipid area, whereas for diC18:1PC bilayers, lipid area assumes a maximum value. It is worthwhile to note that compared to previous data, our results indicate that lipid areas are smaller by approximately 10%. This observation highlights the need to revisit lipid areas, as they are often used in comparisons with molecular dynamics simulations. Moreover, simulators are encouraged to compare their results not only to x-ray scattering data, but to neutron data as well.

Forced reptation revealed by chain pull-out simulations.

We report computation results obtained from extensive molecular dynamics simulations of tensile disentanglement of connector chains placed at the interface between two polymer bulks. Each polymer chain (either belonging to the bulks or being a connector) is treated as a sequence of beads interconnected by springs, using a coarse-grained representation based on the Kremer-Grest model, extended to account for stiffness along the chain backbone. Forced reptation of the connectors was observed during their disentanglement from the bulk chains. The extracted chains are clearly seen following an imaginary "tube" inside the bulks as they are pulled out. The entropic and energetic responses to the external deformation are investigated by monitoring the connector conformation tensor and the modifications of the internal parameters (bonds, bending, and torsion angles along the connectors). The work needed to separate the two bulks is computed from the tensile force induced during debonding in the connector chains. The value of the work reached at total separation is considered as the debonding energy G. The most important parameters controlling G are the length (n) of the chains placed at the interface and their areal density. Our in silico experiments are performed at relatively low areal density and are disregarded if chain scission occurs during disentanglement. As predicted by the reptation theory, for this pure pull-out regime, the power exponent from the scaling G proportional, variant n(a) is a approximately 2, irrespective of chain stiffness. Small variations are found when the connectors form different number of stitches at the interface, or when their length is randomly distributed in between the two bulks. Our results show that the effects of the number of stitches and of the randomness of the block lengths have to be considered together, especially when comparing with experiments where they cannot be controlled rigorously. These results may be significant for industrial applications, such reinforcement of polymer-polymer adhesion by connector chains, when incorporated as constitutive laws at higher time/length scales in finite element calculations.

Molecular-dynamics simulation study of the glass transition in amorphous polymers with controlled chain stiffness.

We report computation results obtained from extensive coarse-grained molecular-dynamics simulations of amorphous ensembles of polymer chains at constant density. In our polymer model, we use bending and torsion potentials acting along the polymer backbone to control the chain stiffness. The static and dynamic properties of the polymer bulk have been analyzed over a large temperature interval in search for the onset of the glass transition. The glass transition temperatures Tg, for different types of chain stiffness, have been determined from the dependence of the self-diffusion coefficient D on the temperature T as the limiting value where the diffusion vanishes. Increasing the chain stiffness induces an increase of the glass transition temperature. The Tg values estimated from diffusion are confirmed by analyzing the relaxation times of the autocorrelation functions for the torsion angle and for the end-to-end vector. The dependence of the diffusion coefficient D on the chain length N is strongly affected by temperature for chains with bending and torsion stiffness. For systems with relatively short chains (N

Effect of bending and torsion rigidity on self-diffusion in polymer melts: a molecular-dynamics study.

Extensive molecular-dynamics simulations have been performed to study the effect of chain conformational rigidity, controlled by bending and torsion potentials, on self-diffusion in polymer melts. The polymer model employs a novel torsion potential that avoids computational singularities without the need to impose rigid constraints on the bending angles. Two power laws are traditionally used to characterize the dependence of the self-diffusion coefficient on polymer length: D proportional to N(-nu) with nu=1 for NNe (reptation regime), Ne being the entanglement length. Our simulations, at constant temperature and density, up to N=250 reveal that, as the chain rigidity increases, the exponent nu gradually increases towards nu=2.0 for NNe. The value of Ne is slightly increased from 70 for flexible chains, up to the point where the crossover becomes undefined. This behavior is confirmed also by an analysis of the bead mean-square displacement. Subsequent investigations of the Rouse modes, dynamical structure factor, and chain trajectories indicate that the pre-reptation regime, for short stiff chains, is a modified Rouse regime rather than reptation.