Chlorpromazine inhibits EAG1 channels by altering the coupling between the PAS, CNBH and pore domains.

EAG1 depolarization-activated potassium selective channels are important targets for treatment of cancer and neurological disorders. EAG1 channels are formed by a tetrameric subunit assembly with each subunit containing an N-terminal Per-Arnt-Sim (PAS) domain and C-terminal cyclic nucleotide-binding homology (CNBH) domain. The PAS and CNBH domains from adjacent subunits interact and form an intracellular tetrameric ring that regulates the EAG1 channel gating, including the movement of the voltage sensor domain (VSD) from closed to open states. Small molecule ligands can inhibit EAG1 channels by binding to their PAS domains. However, the allosteric pathways of this inhibition are not known. Here we show that chlorpromazine, a PAS domain small molecule binder, alters interactions between the PAS and CNBH domains and decreases the coupling between the intracellular tetrameric ring and the pore of the channel, while having little effect on the coupling between the PAS and VSD domains. In addition, chlorpromazine binding to the PAS domain did not alter Cole-Moore shift characteristic of EAG1 channels, further indicating that chlorpromazine has no effect on VSD movement from the deep closed to opened states. Our study provides a framework for understanding global pathways of EAG1 channel regulation by small molecule PAS domain binders.

Multiscale Molecular Dynamics Simulations of the Homodimer Accessory Protein ORF7b of SARS-CoV-2.

The SARS-CoV-2 ORF7b protein has drawn attention for its potential role in viral pathogenesis, but its structural details and lateral membrane associations remain elusive. In this study, we conducted multiscale molecular dynamics simulations to provide detailed molecular insights of the protein's dimerization, which is crucial for unraveling its structural model of protein-protein interface important to regulating cellular immune response. To gain a deeper understanding of homodimer configurations, we employed a machine learning algorithm for structural-based clustering. Clusters were categorized into three distinct groups for both parallel and antiparallel orientations, highlighting the influence of the initial monomer conformation on dimer configurations. Analysis of hydrogen bonding and π-π and π-cation stacking interactions within clusters revealed variations in interactions between clusters. In parallel dimers, weak stacking interactions in the transmembrane (TM) region were observed. In contrast, antiparallel dimers exhibited strong hydrogen bonding and stacking interactions contributing to tight dimeric packing, both within and outside the TM domain. Overall, our study provides a comprehensive view of the structural dynamics of ORF7b homodimerization in both parallel and antiparallel orientations. These findings shed light on the molecular interactions involved in ORF7b dimerization, which are crucial for understanding its potential roles in SARS-CoV-2 pathogenesis. This knowledge could inform future research and therapeutic strategies targeting this viral protein.

Molecular mechanism of EAG1 channel inhibition by imipramine binding to the PAS domain.

Ether-a-go-go (EAG) channels are key regulators of neuronal excitability and tumorigenesis. EAG channels contain an N-terminal Per-Arnt-Sim (PAS) domain that can regulate currents from EAG channels by binding small molecules. The molecular mechanism of this regulation is not clear. Using surface plasmon resonance and electrophysiology we show that a small molecule ligand imipramine can bind to the PAS domain of EAG1 channels and inhibit EAG1 currents via this binding. We further used a combination of molecular dynamics (MD) simulations, electrophysiology, and mutagenesis to investigate the molecular mechanism of EAG1 current inhibition by imipramine binding to the PAS domain. We found that Tyr71, located at the entrance to the PAS domain cavity, serves as a "gatekeeper" limiting access of imipramine to the cavity. MD simulations indicate that the hydrophobic electrostatic profile of the cavity facilitates imipramine binding and in silico mutations of hydrophobic cavity-lining residues to negatively charged glutamates decreased imipramine binding. Probing the PAS domain cavity-lining residues with site-directed mutagenesis, guided by MD simulations, identified D39 and R84 as residues essential for the EAG1 channel inhibition by imipramine binding to the PAS domain. Taken together, our study identified specific residues in the PAS domain that could increase or decrease EAG1 current inhibition by imipramine binding to the PAS domain. These findings should further the understanding of molecular mechanisms of EAG1 channel regulation by ligands and facilitate the development of therapeutic agents targeting these channels.

Conformational Fluctuations in ß2-Microglubulin Using Markov State Modeling and Molecular Dynamics.

Conformational dynamics in proteins can give rise to aggregation prone states during folding, and these kinetically stable states could form oligomers and aggregates. In this study, we investigate the intermediate states and near-folded states of ß2-microglobulin and their physico-chemical properties using molecular dynamics and Markov state modeling. Analysis of hundreds of microseconds simulation show the importance of the edge strands in the misfolded states that give rise to a high exposure of hydrophobic residues in the core of the protein that could initiate oligomerization and aggregate formation. Our study sheds light on the first step of aggregation of ß2m monomers and gave a better picture of the landscape of protein misfolding and aggregation.

Understanding how transmembrane domains regulate interactions between human BST-2 and the SARS-CoV-2 accessory protein ORF7a.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of COVID, replicates at intracellular membranes. Bone marrow stromal antigen 2 (BST-2; tetherin) is an antiviral response protein that inhibits transport of viral particles after budding within infected cells. RNA viruses such as SARS-CoV-2 use various strategies to disable BST-2, including use of transmembrane 'accessory' proteins that interfere with BST-2 oligomerization. ORF7a is a small, transmembrane protein present in SARS-CoV-2 shown previously to alter BST-2 glycosylation and function. In this study, we investigated the structural basis for BST-2 ORF7a interactions, with a particular focus on transmembrane and juxtamembrane interactions. Our results indicate that transmembrane domains play an important role in BST-2 ORF7a interactions and mutations to the transmembrane domain of BST-2 can alter these interactions, particularly single-nucleotide polymorphisms in BST-2 that result in mutations such as I28S. Using molecular dynamics simulations, we identified specific interfaces and interactions between BST-2 and ORF7a to develop a structural basis for the transmembrane interactions. Differences in glycosylation are observed for BST-2 transmembrane mutants interacting with ORF7a, consistent with the idea that transmembrane domains play a key role in their heterooligomerization. Overall, our results indicate that ORF7a transmembrane domain interactions play a key role along with extracellular and juxtamembrane domains in modulating BST-2 function.

Drude Polarizable Lipid Force Field with Explicit Treatment of Long-Range Dispersion: Parametrization and Validation for Saturated and Monounsaturated Zwitterionic Lipids.

Accurate empirical force fields of lipid molecules are a critical component of molecular dynamics simulation studies aimed at investigating properties of monolayers, bilayers, micelles, vesicles, and liposomes, as well as heterogeneous systems, such as protein-membrane complexes, bacterial cell walls, and more. While the majority of lipid force field-based simulations have been performed using pairwise-additive nonpolarizable models, advances have been made in the development of the polarizable force field based on the classical Drude oscillator model. In the present study, we undertake further optimization of the Drude lipid force field, termed Drude2023, including improved treatment of the phosphate and glycerol linker region of PC and PE headgroups, additional optimization of the alkene group in monounsaturated lipids, and inclusion of long-range Lennard-Jones interactions using the particle-mesh Ewald method. Initial optimization targeted quantum mechanical (QM) data on small model compounds representative of the linker region. Subsequent optimization targeted QM data on larger model compounds, experimental data, and dihedral potentials of mean force from the CHARMM36 additive lipid force field using a parameter reweighting protocol. The use of both experimental and QM target data during the reweighting protocol is shown to produce physically reasonable parameters that reproduce a collection of experimental observables. Target data for optimization included surface area/lipid for DPPC, DSPC, DMPC, and DLPC bilayers and nuclear magnetic resonance (NMR) order parameters for DPPC bilayers. Validation data include prediction of membrane thickness, scattering form factors, electrostatic potential profiles, compressibility moduli, surface area per lipid, water permeability, NMR T1 relaxation times, diffusion constants, and monolayer surface tensions for a variety of saturated and unsaturated lipid mono- and bilayers. Overall, the agreement with experimental data is quite good, though the results are less satisfactory for the NMR T1 relaxation times for carbons near the ester groups. Notable improvements compared to the additive C36 force field were obtained for membrane dipole potentials, lipid diffusion coefficients, and water permeability with the exception of monounsaturated lipid bilayers. It is anticipated that the optimized polarizable Drude2023 force field will help generate more accurate molecular simulations of pure bilayers and heterogeneous systems containing membranes, advancing our understanding of the role of electronic polarization in these systems.

Antimicrobial Peptide Mechanism Studied by Scattering-Guided Molecular Dynamics Simulation.

In an effort to combat rising antimicrobial resistance, our labs have rationally designed cationic, helical, amphipathic antimicrobial peptides (AMPs) as alternatives to traditional antibiotics since AMPs incur bacterial resistance in weeks, rather than days. One highly positively charged AMP, WLBU2 (+13e), (RRWV RRVR RWVR RVVR VVRR WVRR), has been shown to be effective in killing both Gram-negative (G(-)) and Gram-positive (G(+)) bacteria by directly perturbing the bacterial membrane nonspecifically. Previously, we used two equilibrium experimental methods: synchrotron X-ray diffuse scattering (XDS) providing lipid membrane thickness and neutron reflectometry (NR) providing WLBU2 depth of penetration into three lipid model membranes (LMMs). The purpose of the present study is to use the results from the scattering experiments to guide molecular dynamics (MD) simulations to investigate the detailed biophysics of the interactions of WLBU2 with LMMs of Gram-negative outer and inner membranes, and Gram-positive cell membranes, to elucidate the mechanisms of bacterial killing. Instead of coarse-graining, backmapping, or simulating without bias for several microseconds, all-atom (AA) simulations were guided by the experimental results and then equilibrated for ∼0.5 µs. Multiple replicas of the inserted peptide were run to probe stability and reach a combined time of at least 1.2 µs for G(-) and also 2.0 µs for G(+). The simulations with experimental comparisons help rule out certain structures and orientations and propose the most likely set of structures, orientations, and effects on the membrane. The simulations revealed that water, phosphates, and ions enter the hydrocarbon core when WLBU2 is positioned there. For an inserted peptide, the three types of amino acids, arginine, tryptophan, and valine (R, W, V), are arranged with the 13 Rs extending from the hydrocarbon core to the phosphate group, Ws are located at the interface, and Vs are more centrally located. For a surface state, R, W, and V are positioned relative to the bilayer interface as expected from their hydrophobicities, with Rs closest to the phosphate group, Ws close to the interface, and Vs in between. G(-) and G(+) LMMs are thinned ∼1 Å by the addition of WLBU2. Our results suggest a dual anchoring mechanism for WLBU2 both in the headgroup and in the hydrocarbon region that promotes a defect region where water and ions can flow across the slightly thinned bacterial cell membrane.

Molecular dynamics simulations of the human ocular lens with age and cataract.

The human ocular lens consists primarily of elongated, static fibers characterized by high stability and low turnover, which differ dramatically in their composition and properties from other biological membranes. Cholesterol (Chol) and sphingolipids (SL) are present at high concentrations, including saturated SLs, such as dihyrosphingomyelin (DHSM). Past molecular dynamics simulations demonstrated that the presence of DHSM and high Chol concentration contributes to higher order in lipid membranes. This current study simulated more complex models of human lens membranes. Models were developed representing physiological compositions in cataractous lenses aged 74 ± 6 years and in healthy lenses aged 22 ± 4, 41 ± 6, and 69 ± 3 years. With older age, Chol and ceramide concentrations increase and glycerophospholipid concentration decreases. With cataract, ceramide concentration increases and Chol and glycerophospholipid concentrations decrease. Surface area per lipid, deuterium order parameters (SCD), sterol tilt angle, electron density profiles, bilayer thickness, chain interdigitation, two-dimensional radial distribution functions (2D-RDF), lipid clustering, and hydrogen bonding were calculated for all simulations. All systems exhibited low surface area per lipid and high bilayer thickness, indicative of strong vertical packing. SCD parameters suggest similarly, with saturated tails in the hydrophobic core of the membrane having elevated order. Vertical packing and acyl tail order increased with both age and cataract condition. Lateral diffusion decreased with age and cataracts, with the older and cataractous models demonstrating increased long-range structure by the 2D-RDF analysis. In future work examining the membrane proteins of the lens, these models can serve as a physiologically accurate representation of the lens lipidome.

Computational Study of the Allosteric Effects of p5 on CDK5-p25 Hyperactivity as Alternative Inhibitory Mechanisms in Neurodegeneration.

The cyclin-dependent kinase (CDK5) forms a stable complex with its activator p25, leading to the hyperphosphorylation of tau proteins and to the formation of plaques and tangles that are considered to be one of the typical causes of Alzheimer's disease (AD). Hence, the pathological CDK5-p25 complex is a promising therapeutic target for AD. Small peptides, obtained from the truncation of CDK5 physiological activator p35, have shown promise in inhibiting the pathological complex effectively while also crossing the blood-brain barrier. One such small 24-residue peptide, p5, has shown selective inhibition toward the pathological complex in vivo. Our previous research focused on the characterization of a computationally predicted CDK5-p5 binding mode and of its pharmacophore, which was consistent with competitive inhibition. In continuation of our previous work, herein, we investigate four additional binding modes to explore other possible mechanisms of interaction between CDK5 and p5. The quantitative description of the pharmacophore is consistent with both competitive and allosteric p5-induced inhibition mechanisms of CDK5-p25 pathology. The gained insights can direct further in vivo/in vitro tests and help design small peptides, linear or cyclic, or peptidomimetic compounds as adjuvants of orthosteric inhibitors or as part of a cocktail of drugs with enhanced effectiveness and lower side effects.

GraphVAMPNet, using graph neural networks and variational approach to Markov processes for dynamical modeling of biomolecules.

Finding a low dimensional representation of data from long-timescale trajectories of biomolecular processes, such as protein folding or ligand-receptor binding, is of fundamental importance, and kinetic models, such as Markov modeling, have proven useful in describing the kinetics of these systems. Recently, an unsupervised machine learning technique called VAMPNet was introduced to learn the low dimensional representation and the linear dynamical model in an end-to-end manner. VAMPNet is based on the variational approach for Markov processes and relies on neural networks to learn the coarse-grained dynamics. In this paper, we combine VAMPNet and graph neural networks to generate an end-to-end framework to efficiently learn high-level dynamics and metastable states from the long-timescale molecular dynamics trajectories. This method bears the advantages of graph representation learning and uses graph message passing operations to generate an embedding for each datapoint, which is used in the VAMPNet to generate a coarse-grained dynamical model. This type of molecular representation results in a higher resolution and a more interpretable Markov model than the standard VAMPNet, enabling a more detailed kinetic study of the biomolecular processes. Our GraphVAMPNet approach is also enhanced with an attention mechanism to find the important residues for classification into different metastable states.

A reengineered common chain cytokine augments CD8+ T cell-dependent immunotherapy.

Cytokine therapy is limited by undesirable off-target side effects as well as terminal differentiation and exhaustion of chronically stimulated T cells. Here, we describe the signaling properties of a potentially unique cytokine by design, where T cell surface binding and signaling are separated between 2 different families of receptors. This fusion protein cytokine, called OMCPmutIL-2, bound with high affinity to the cytotoxic lymphocyte-defining immunoreceptor NKG2D but signaled through the common γ chain cytokine receptor. In addition to precise activation of cytotoxic T cells due to redirected binding, OMCPmutIL-2 resulted in superior activation of both human and murine CD8+ T cells by improving their survival and memory cell generation and decreasing exhaustion. This functional improvement was the direct result of altered signal transduction based on the reorganization of surface membrane lipid rafts that led to Janus kinase-3-mediated phosphorylation of the T cell receptor rather than STAT/AKT signaling intermediates. This potentially novel signaling pathway increased CD8+ T cell response to low-affinity antigens, activated nuclear factor of activated T cells transcription factors, and promoted mitochondrial biogenesis. OMCPmutIL-2 thus outperformed other common γ chain cytokines as a catalyst for in vitro CD8+ T cell expansion and in vivo CD8+ T cell-based immunotherapy.

Modeling the membrane binding mechanism of a lipid transport protein Osh4 to single membranes.

All-atom (AA) molecular dynamics simulations are used to unravel the binding mechanism of yeast oxysterol binding protein (Osh4) to model membranes with varying anionic lipid concentration using AA and the highly mobile membrane mimetic (HMMM) representations. For certain protein-lipid interactions, an improved forcefield description is used (CUFIX) to accurately describe lipid-protein electrostatic interactions. Our detailed computational studies have identified a single, ß-crease oriented, membrane-bound conformation of Osh4 for all anionic membranes. The penetration of the PHE-239 residue below the membrane phosphate plane is the characteristic signature of the membrane-bound state of Osh4. As the phenylalanine loop anchors itself deeply in the membrane; the other regions of the Osh4, namely, ALPS motif, ß6- ß7 loop, ß14- ß15 loop, and ß16- ß17 loop, maximize their contact with the membrane. Furthermore, loose lipid packing and higher mobility of HMMM enable stronger association of the ALPS motif with the membrane lipids through its hydrophobic surface. After the HMMM is converted to AA and equilibrated, the binding is two to three times stronger compared with simulations started with the AA representation, yielding the major importance of the ALPS motif to binding. Quantitative estimation of binding energy revealed that the phenylalanine loop plays a crucial role in stable membrane attachment of Osh4 and contributes significantly toward overall binding process. The CUFIX parameters provide a more balanced picture of hydrophobic and electrostatic interactions between the protein and the membrane, which differs from our past work that showed salt bridges alone stabilized Osh4-membrane contact. Our study provides a comprehensive picture of the binding mechanism of Osh4 with model single membranes and, thus, understanding of the initial interactions is important for elucidating the biological function of this protein to shuttle lipids between organelles.

All-Atom Modeling of Complex Cellular Membranes.

Cell membranes are composed of a variety of lipids and proteins where they interact with each other to fulfill their roles. The first step in modeling these interactions in molecular simulations is to have reliable mimetics of the membrane's lipid environment. This Feature Article presents our recent efforts to model complex cellular membranes using all-atom force fields. A short review of the CHARMM36 (C36) lipid force field and its recent update to incorporate the long-range dispersion is presented. Key examples of model membranes mimicking various species and organelles are given. These include single-celled organisms such as bacteria (E. coli., chlamydia, and P. aeruginosa) and yeast (plasma membrane, endoplasmic reticulum, and trans-Golgi network) and more advanced ones such as plants (soybean and Arabidopsis thaliana) and mammals (ocular lens, stratum corneum, and peripheral nerve myelin). Leaflet asymmetry in composition has also been applied to some of these models. With the increased lipid diversity in the C36 lipid FF, these complex models can better reflect the structural, mechanical, and dynamic properties of realistic membranes and open an opportunity to study biological processes involving other molecules.

Leaflet Asymmetry Modeling in the Lipid Composition of Escherichia coli Cytoplasmic Membranes.

Lipid composition asymmetry between leaflets is important to cell function and plays a key role in the "positive inside" rule in transmembrane proteins. In this work, Escherichia coli inner plasma membrane models reflecting this asymmetry have been investigated at the early-log and stationary stages during the bacterial lifecycle using all-atom molecular dynamics simulations. The CHARMM36 lipid force field is used, and selected membrane properties are tested for variations between two leaflets and whole membranes. Our models include bacterial lipids with a cyclopropane moiety on the sn-2 acyl chain in the stationary membrane model. The PE/PG ratio for two leaflets reflects the "positive inside" rule of membrane proteins, set to 6.8 and 2.8 for the inner and outer leaflets of the two models, respectively. We are the first to model leaflet asymmetry in the lipid composition of E. coli cytoplasmic membranes and observe the effect on membrane properties in leaflets and whole membranes. Specifically, our results show that for the stationary phase bilayer, the surface area per lipid (SA/lipid) is larger, the thickness (2DC and DB) is smaller, the tilt angle is larger, the tilt modulus is smaller, and the deuterium order parameters (SCD) of sn-1 and sn-2 tails are lower, compared to the early-log stage. Moreover, the stationary stage bilayer has a positive spontaneous curvature, while the early-log stage has a near flat spontaneous curvature. For leaflet asymmetry, the inner leaflet has a larger SA/lipid, a smaller thickness, a smaller elastic tilt modulus (a larger tilt angle), and lower SCD, compared to the outer leaflet in both stages. Moreover, an asymmetric membrane involves a lipid tilt and a lateral extension, varying from a reference state of a pre-equilibrium membrane. This work encourages a more profound exploration of leaflet asymmetry in various other membrane models and how this might affect the structure and function of membrane-associated peptides and proteins.

Variational embedding of protein folding simulations using Gaussian mixture variational autoencoders.

Conformational sampling of biomolecules using molecular dynamics simulations often produces a large amount of high dimensional data that makes it difficult to interpret using conventional analysis techniques. Dimensionality reduction methods are thus required to extract useful and relevant information. Here, we devise a machine learning method, Gaussian mixture variational autoencoder (GMVAE), that can simultaneously perform dimensionality reduction and clustering of biomolecular conformations in an unsupervised way. We show that GMVAE can learn a reduced representation of the free energy landscape of protein folding with highly separated clusters that correspond to the metastable states during folding. Since GMVAE uses a mixture of Gaussians as its prior, it can directly acknowledge the multi-basin nature of the protein folding free energy landscape. To make the model end-to-end differentiable, we use a Gumbel-softmax distribution. We test the model on three long-timescale protein folding trajectories and show that GMVAE embedding resembles the folding funnel with folded states down the funnel and unfolded states outside the funnel path. Additionally, we show that the latent space of GMVAE can be used for kinetic analysis and Markov state models built on this embedding produce folding and unfolding timescales that are in close agreement with other rigorous dynamical embeddings such as time independent component analysis.

Symmetric and Asymmetric Models for the Arabidopsis thaliana Plasma Membrane: A Simulation Study.

Arabidopsis thaliana is an important model organism, which has attracted many biologists. While most research efforts have been on studying the genetics and proteins of this organism, a systematic study of its lipidomics is lacking. Here, we present a novel, asymmetric model of its cell membrane with its lipid composition consisting of five glycerophospholipids, two glycolipids, and sitosterol determined from multiple independent experiments. A typical lipid type in plant membranes is glycosyl inositol phosphoryl ceramide (GIPC), which accounts for about 10% of the total lipids in the outer leaflet in our model. Two symmetric models representing the inner and outer leaflets of the membrane were built and simulated until equilibrium was reached and then combined to form the asymmetric model. Our results indicate that the outer leaflet is more rigid and tightly packed compared to the inner leaflet. Pressure profiles for the two leaflets are overall similar though the outer leaflet exhibits larger oscillations. A special focus on lipid organization is discussed and the interplay between glycolipids and sitosterols is found to be important. The current model provides a baseline for future modeling of similar membranes and can be used to study partitioning of small molecules in the membrane or further developed to study the interaction between plant membrane proteins and lipids.

Simulations of Diabetic and Non-Diabetic Peripheral Nerve Myelin Lipid Bilayers.

The multilayered myelin sheath is a critical component of both central and peripheral nervous systems, forming a protective barrier against axonal damage and facilitating the movement of nervous impulses. It is primarily composed of cholesterol (CHL1), phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS), phosphatidylinositol (PI), sphingomyelin (SM), and galactosylceramide (GalCer) lipids. For rat sciatic nerve myelin (part of the peripheral nervous system, PNS), it has been found that cholesterol and unsaturated fatty acid contents are significantly lower in diabetic than in non-diabetic conditions. In this study, lipid compositions from experimental data are used to create four model rat sciatic nerve myelin lipid bilayers: PI-containing (non-diabetic and diabetic) and PS-containing (non-diabetic and diabetic), which were then simulated using the all-atom CHARMM36 force field. Simulation results of diabetic membranes indicate less rigid, more laterally expansive, and thinner bilayers as well as potentially reduced interactions between GalCer on opposing myelin leaflets, supporting a direct role of the cholesterol content decrease in instigating myelin deterioration and diabetic peripheral neuropathy. Compared to PI-lipids, PS-lipids were found to cause higher inter-lipid spacing and decreased order within membranes as a result of their smaller headgroup size and higher inter-lipid hydrogen bonding potential, which allow them to more frequently reside deeper in the membrane plane and produce pushing effects on other lipids. GalCer deuterium order parameters and non-diabetic headgroup-to-headgroup bilayer thicknesses were compared to experimental data, exhibiting close alignment, supporting the future usage of these models to study the PNS myelin sheath.

A replica exchange umbrella sampling (REUS) approach to predict host-guest binding free energies in SAMPL8 challenge.

In this study, we report binding free energy calculations of various drugs-of-abuse to Cucurbit-[8]-uril as part of the SAMPL8 blind challenge. Force-field parameters were obtained from force-matching with different quantum mechanical levels of theory. The Replica Exchange Umbrella Sampling (REUS) approach was used with a cylindrical restraint to enhance the sampling of host-guest binding. Binding free energy was calculated by pulling the guest molecule from one side of the symmetric and cylindrical host, then into and through the host, and out the other side (bidirectional) as compared to pulling only to the bound pose inside the cylindrical host (unidirectional). The initial results with force-matched MP2 parameter set led to RMSE of 4.68 [Formula: see text] from experimental values. However, the follow-up study with CHARMM generalized force field parameters and force-matched PM6-D3H4 parameters resulted in RMSEs from experiment of [Formula: see text] and [Formula: see text], respectively, which demonstrates the potential of REUS for accurate binding free energy calculation given a more suitable description of energetics. Moreover, we compared the free energies for the so called bidirectional and unidirectional free energy approach and found that the binding free energies were highly similar. However, one issue in the bidirectional approach is the asymmetry of profile on the two sides of the host. This is mainly due to the insufficient sampling for these larger systems and can be avoided by longer sampling simulations. Overall REUS shows great promise for binding free energy calculations.

Considerations of Recent All-Atom Lipid Force Field Development.

Molecular simulations of biological molecules require an accurate description of molecular interactions through a force field (FF). The focus of this Perspective is on all-atom lipid FFs. Recent additions to the CHARMM36 lipid FF continue to expand a researcher's ability to probe membrane structure and function with a wide variety of biologically important lipids. Currently, there is an effort to reduce the assumptions in all-atom lipid FFs. The inclusion of long-range dispersion interaction through particle-mesh Ewald is allowing for more accurate descriptions of lipid bilayer and monolayer properties without additional computational cost. Soon, simulations with lipid FFs will no longer depend on short-range cutoffs and will accurately represent long-range dispersion. This requires efficient FF parametrization with an automated approach due to FF complexity. In addition, polarizable FFs for lipids will be important for the next generation of simulations that accurately represent how molecule interactions respond to a varied environment.