Hsuan-Yi Chen

Coarse-grain model for lipid bilayer self-assembly and dynamics: Multiparticle collision description of the solvent

A mesoscopic coarse-grain model for computationally efficient simulations of biomembranes is presented. It combines molecular dynamics simulations for the lipids, modeled as elastic chains of beads, with multiparticle collision dynamics for the solvent. Self-assembly of a membrane from a uniform mixture of lipids is observed. Simulations at different temperatures demonstrate that it reproduces the gel and liquid phases of lipid bilayers. Investigations of lipid diffusion in different phases reveals a crossover from subdiffusion to normal diffusion at long times. Macroscopic membrane properties, such as stretching and bending elastic moduli, are determined directly from the mesoscopic simulations. Velocity correlation functions for membrane flows are determined and analyzed.

Internal States of Active Inclusions and the Dynamics of an Active Membrane

A theoretical model of a two-component fluid membrane containing lipids and two-state active inclusions is presented. This model predicts several nonequilibrium morphology transitions. (i) Active pumping of the inclusions can drive a long-wavelength undulation instability. (ii) Active excitation of the inclusions can induce aggregation of high-curvature excited inclusions. (iii) Active inclusion conformation changes can produce finite-size domains. The resulting steady state domain size depends on inclusion activities. For a stable membrane the height fluctuation spectrum in the long-wavelength limit is similar to previous studies which neglected the inclusion internal states.

Diffusion, sedimentation equilibrium, and harmonic trapping of run-and-tumble nanoswimmers

The diffusion of self-propelling nanoswimmers is explored by dissipative particle dynamics in which a nanoswimmer swims by forming an instantaneous force dipole with one of its nearest neighboring solvent beads. Our simulations mimic run-and-tumble behavior by letting the swimmer run for a time τ, then it randomly changes its direction for the next run period. Our simulations show that the swimming speed (ν(a)) of a nanoswimmer is proportional to the propulsion force and the mobility of a pusher is the same as that of a puller. The effective diffusivity is determined by three methods: mean squared displacement, velocity autocorrelation function, and sedimentation equilibrium. The active colloid undergoes directed propulsion at short time scales but changes to random motion at long time scales. The velocity autocorrelation function decreases with time and becomes zero beyond the run time. Under gravity, the concentration profile of active colloids follows Boltzmann distribution with a sedimentation length consistent with that acquired from the drift-diffusion equation. In our simulation, all three methods yield the same result, the effective diffusivity of an active colloid is the sum of the diffusivity of a passive colloid and ν(a)²τ/6. When the active colloids are confined by a harmonic well, they are trapped within a confinement length defined by the balance between the swimmer active force and restoring force of the well. When the confinement length is large compared to the run length, the stationary density profile follows the Boltzmann distribution. However, when the run length exceeds the confinement length, the density distribution is no longer described by Boltzmann distribution, instead we found a bimodal distribution.

Nano-swimmers in biological membranes and propulsion hydrodynamics in two dimensions.

Active protein inclusions in biological membranes can represent nano-swimmers and propel themselves in lipid bilayers. A simple model of an active inclusion with three particles (domains) connected by variable elastic links is considered. First, the membrane is modeled as a two-dimensional viscous fluid and propulsion behavior in two dimensions is examined. After that, an example of a microscopic dynamical simulation is presented, where the lipid bilayer structure of the membrane is resolved and the solvent effects are included by multiparticle collision dynamics. Statistical analysis of data reveals ballistic motion of the swimmer, in contrast to the classical diffusion behavior found in the absence of active transitions between the states.Graphical abstract

Coarse-grain simulations of active molecular machines in lipid bilayers

A coarse-grain method for simulations of the dynamics of active protein inclusions in lipid bilayers is described. It combines the previously proposed hybrid simulations of bilayers [M.-J. Huang, R. Kapral, A. S. Mikhailov, and H.-Y. Chen, J. Chem. Phys. 137, 055101 (2012)], based on molecular dynamics for the lipids and multi-particle collision dynamics for the solvent, with an elastic-network description of active proteins. The method is implemented for a model molecular machine which performs active conformational motions induced by ligand binding and its release after reaction. The situation characteristic for peripheral membrane proteins is considered. Statistical investigations of the effects of single active or passive inclusions on the shape of the membrane are carried out. The results show that the peripheral machine produces asymmetric perturbations of the thickness of two leaflets of the membrane. It also produces a local saddle in the midplane height of the bilayer. Analysis of the power spectrum of the fluctuations of the membrane midplane shows that the conformational motion of the machine perturbs these membrane fluctuations. The hydrodynamic lipid flows induced by cyclic conformational changes in the machine are analyzed. It is shown that such flows are long-ranged and should provide an additional important mechanism for interactions between active inclusions in biological membranes.

Diffusion and surface excess of a confined nanoswimmer dispersion

The diffusivity and surface excess of nanoswimmers which are confined in two plates with the separation H are explored by dissipative particle dynamics. Both mean squared displacement and velocity autocorrelation function methods are used to study the diffusive behavior of nanoswimmers with the Brownian diffusivity D0 and the results obtained from both methods are consistent. The active diffusivity of confined nanoswimmers (D - D0) depends on the wall separation, swimming speed v(a), and run time τ. Our simulation results show that (D-D0)/v(a)(2)τ is a function of v(a)τ/H. The reduction in the diffusivity of active colloids is more significant than that of passive particles. The distribution of nanoswimmers between two parallel walls is acquired and two regions can be identified. The accumulation of nanoswimmers near walls is quantitatively described by the surface excess Γ. It is found that Γ grows as the nanoswimmer concentration c(b), swimming speed v(a), and run time τ are increased. The coupling between the ballistic trajectory of nanoswimmers and the walls results in nanoswimmer accumulation. The simulation outcomes indicate that Γ/Hc(b) is a function of H/v(a)τ.

Forced dissociation of a biomolecular complex under periodic and correlated random forcing

The dissociation of a biomolecular complex under the action of periodic and correlated random forcing is studied theoretically. The former is characterized by the period tau p and the latter by the correlation time tau r. The rupture rates are calculated by overdamped Langevin dynamics and three distinct regimes are identified for both cases by comparison to local relaxation time tau R and bond lifetime T. For periodic forcing, the adiabatic approximation cannot be applied in the regime tau p<<tau R and the bond lifetime is determined by the average pulling. As tau R<<tau p<<T, the rupture rate is enhanced by periodic forcing but is tau(p) independent. Analytical expressions are obtained for small and large force amplitudes. As T<<tau p, the rupture rate depends on the phase lag and the process behaves like it is under constant force or loading rate. The result of correlated random forcing is similar to that of periodic forcing. Since the fluctuating forces greater than the average force F contribute more than the fluctuating forces less than F, the force fluctuations enhance the rupture rate. As T<tau r, the pulling felt by the bond before rupture cannot follow the random forcing protocol and, thus, force fluctuations decline with increasing tau r.

Dynamics of biomembranes with active multiple-state inclusions

Nonequilibrium dynamics of biomembranes with active multiple-state inclusions is considered. The inclusions represent protein molecules which perform cyclic internal conformational motions driven by the energy brought with adenosine triphosphate (ATP) ligands. As protein conformations cyclically change, this induces hydrodynamical flows and also directly affects the local curvature of a membrane. On the other hand, variations in the local curvature of the membrane modify the transition rates between conformational states in a protein, leading to a feedback in the considered system. Moreover, active inclusions can move diffusively through the membrane so that their surface concentration varies. The kinetic description of this system is constructed and the stability of the uniform stationary state is analytically investigated. We show that, as the rate of supply of chemical energy is increased above a certain threshold, this uniform state becomes unstable and stationary or traveling waves spontaneously develop in the system. Such waves are accompanied by periodic spatial variations of the membrane curvature and the inclusion density. For typical parameter values, their characteristic wavelengths are of the order of hundreds of nanometers. For traveling waves, the characteristic frequency is of the order of a thousand Hz or less. The predicted instabilities are possible only if at least three internal inclusion states are present.

Finite-size domains in membranes with active two-state inclusions.

The distribution of inclusion-rich domains in membranes with active two-state inclusions is studied by simulations. Our study shows that typical size of inclusion-rich domains (L) can be controlled by inclusion activities in several ways. When there is effective attraction between state-1 inclusions, we find: (i) Small domains with only several inclusions are observed for inclusions with time scales (approximately 10(-3)s) and interaction energy [approximately Omicron(kBT)] comparable to motor proteins. (ii) L scales as 13 power of the lifetime of state-1 for a wide range of parameters. (iii) L shows a switch-like dependence on state-2 lifetime k12(-1). That is, L depends weakly on k12 when k12k12*, the crossover k12* occurs when the diffusion length of a typical state-2 inclusion within its lifetime is comparable to L. (iv) Inclusion-curvature coupling provides another length scale that competes with the effects of transition rates.

Adhesion-induced lateral phase separation of multicomponent membranes: The effect of repellers and confinement

We present a theoretical study of adhesion-induced lateral phase separation for a membrane with short stickers, long stickers, and repellers confined between two hard walls. The effects of confinement and repellers on lateral phase separation are investigated. We find that the critical potential depth of the stickers for lateral phase separation increases as the distance between the hard walls decreases. This suggests confinement-induced or force-induced mixing of stickers. We also find that repellers with stronger repulsive potential tend to enhance, while repellers with weaker repulsive potential tend to suppress adhesion-induced lateral phase separation.