1 Effect of local stress on accurate modeling of bacterial outer membranes using all -atom molecular dynamics

2025-04-30 0 0 1.13MB 29 页 10玖币

侵权投诉

Effect of local stress on accurate modeling of bacterial outer membranes using all-atom

molecular dynamics

Emad Pirhadi1, Juan M. Vanegas†2, Mithila N. Farin1, Jeffrey W. Schertzer3, Xin Yong1*

1Department of Mechanical Engineering, Binghamton University, Binghamton, New York

2Department of Biochemistry and Biophysics, Oregon State University, Corvallis, Oregon

3Department of Biological Sciences, Binghamton University, Binghamton, New York

ABSTRACT

Biological membranes are fundamental components of living organisms that play an undeniable

role in their survival. Molecular dynamics (MD) serves as an essential computational tool for

studying biomembranes on molecular and atomistic scales. The status quo of MD simulations of

biomembranes studies a nanometer-sized membrane patch periodically extended under periodic

boundary conditions (PBC). In nature, membranes are usually composed of different lipids in their

two layers (referred to as leaflets). This compositional asymmetry imposes a fixed ratio of lipid

numbers between the two leaflets in a periodically constrained membrane, which needs to be set

appropriately. The widely adopted methods of defining leaflet lipid ratio suffer from the lack of

control over the mechanical tension of each leaflet, which could significantly influence research

findings. In this study, we investigate the role of membrane-building protocol and the resulting

initial stress state on the interaction between small molecules and asymmetric membranes. We

model the outer membrane of Pseudomonas aeruginosa bacteria using two different building

protocols and probe their interactions with the Pseudomonas Quinolone Signal (PQS). Our results

show that differential stress could shift the position of free energy minimum for the PQS molecule

between the two leaflets of the asymmetric membrane. This work provides critical insights into

the relationship between the initial per-leaflet tension and the spontaneous intercalation of PQS.

*Email: xyong@binghamton.edu

†Email: vanegasj@oregonstate.edu

INTRODUCTION

Biological membranes are present in all domains of life, separating the cell interior from the

external environment or defining different cell organelles. As selectively permeable barriers, they

are involved in various cellular processes such as compartmentalization, signaling, transport and

trafficking, sensing, metabolism, and overall regulation of many biological functions.1–3 These

indispensable thin films are comprised of various lipids and proteins, asymmetrically distributed

between the two bilayer leaflets.4 An important and widespread example is the outer membrane

(OM) of Gram-negative bacteria, which is responsible for many of their biological traits.5,6 While

the inner leaflet of the OM is composed of common phospholipids (PLs), the abundance of

lipopolysaccharides (LPS) in the outer leaflet renders the membrane highly asymmetric in leaflet

lipid composition.7 The complex structure of LPS helps shield Gram-negative bacteria against

environmental threats. The structure consists of a diverse and variable length O-antigen

polysaccharide in the outermost region, a more conserved oligosaccharide “core” region proximal

to the membrane, and most importantly, an endotoxic Lipid A region anchored in the outer leaflet

of the OM.8 Bacterial membranes have been studied extensively in the last few decades.9–12

Nevertheless, the undeniable role of OM in the survival of Gram-negative bacteria, including its

interaction with the human immune system and contributions to bacterial pathogenesis, requires

that we continuously develop and refine tools to investigate how the membrane will interact with

important molecules.13–15

Unfortunately, state-of-the-art experimental tools still face numerous challenges in probing

biomembranes and their interactions with small molecules, proteins, and nanomaterials on

decreasing length and time scales.16–18 This limitation stimulates the rapid advance in

computational modeling of biomembranes, particularly molecular dynamics (MD) simulations,

which offer nanoscale spatiotemporal resolutions and provide detailed information on molecular

interactions not easily accessible to physical experiments.19,20 Rapid advances in high-performance

computing made it possible to use MD to investigate realistic cell membranes with diverse and

complex compositions. However, the information obtained from MD requires discreet

interpretation since the accuracy and reliability of these models are highly dependent on the

modeling process and system parameterization. Examples of the essential parameters include the

number of molecules, the representation of chemical structures (e.g., all-atom, united-atom, or

coarse-grained), force fields, and electrical charge distributions.21–23 A poorly chosen set of system

parameters could significantly influence the results of MD simulations, leading to defective

conclusions.24,25 The sensitivity of MD systems could bias the understanding of physical

phenomena of interest and compromise the outcomes of a research project. Therefore, it is crucial

to understand the influence of critical parameters in the MD modeling of biomembranes.

Differential stress, defined as the difference between the tension of membrane leaflets, has

attracted considerable attention in the community in the past few years. Hossein and Deserno

reported that the differential stress could significantly influence the mechanical properties of a

compositionally asymmetric membrane.26 This behavior has also been demonstrated in recent

experiments on free-standing asymmetric membranes.27 Realistic MD models of biological

membranes need to include lipid asymmetry and thus could possess pre-existing differential stress

due to asymmetric lipid packing or suppressed spontaneous curvature in the periodic simulation

domain.28 Thus, the initial conditions would significantly influence the differential stress and,

consequently, the mechanical properties of the model membrane. The initial stress state of the

membrane due to the coupling between intrinsic bending and asymmetric lipid packing can also

influence the interactions with membrane-active components such as small molecules,

nanoparticles, peptides, and proteins. This emerging issue has prompted the development of new

methods for rationally building an asymmetric membrane with controlled per-leaflet tension.29,30

This work aims to elucidate the effects of building protocol as a system parameter often overlooked

for accurately modeling physiological relevant OMs. Specifically, we characterize the stress states

of a model OM of Pseudomonas aeruginosa bacteria constructed using different membrane-

building protocols and systematically probe how differential stress affects the membrane

interaction with signaling molecules. The findings will provide guidelines for future simulation

studies of asymmetric membranes.

MATERIALS AND METHODS

Outer membrane and signaling molecule

Lipid A has a disaccharide backbone acylated with four to eight fatty acid tails. This potent

activator of the innate immune system is also accountable for the toxic effects of Gram-negative

bacteria.31 In this model OM, the outer leaflet was composed of a hexa-acylated P. aeruginosa

Lipid A corresponding to the PA14 strain (see Figure 1). The lipid composition of the inner leaflet

was selected based on previous experimental lipidome analysis, which reported that the PLs in P.

aeruginosa OM mostly have one saturated and one unsaturated acyl chain.32 The fatty acid profiles

showed that the predominant pair is palmitic acid (C16:0) and oleic acid (C18:1), and their molar

ratio is approximately 1:1.33,34 From the reported PL distributions, phosphoethanolamine and

phosphoglycerol (PG) are the most abundant head groups with the PE / PG ratio of 2.0 ~ 2.2.32,34

Thus, the inner leaflet of our model OM was constructed by mixing 1-palmitoyl-2-oleoyl-sn-

glycero-3-phosphoethanolamine (POPE) and 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoglycerol

(POPG) at a ratio of 2.2. The bilayer was solvated in water. The system also contained

corresponding numbers of calcium (Ca2+) and sodium (Na+) ions as counter ions to neutralize Lipid

A and POPG, respectively. Additional NaCl at 150 mM concentration was added to mimic

physiological conditions.35 The 2-heptyl-3-hydroxy-4-quinolone (Figure 1 (d)) named

Pseudomonas quinolone signal (PQS) is a self-produced quorum sensing molecule with multiple

functionalities in the survival of P. aeruginosa bacteria. This study explores PQS interaction with

the model OM and its quantitative effect on membrane stress.

Figure 1. Chemical structures of (a) the P. aeruginosa PA14 Lipid A, (b) POPE, (c) POPG, and

(d) PQS in the model system.

Force field, Lipid A protonation state, and equilibration process

The CHARMM36 force field36,37 was selected to perform all-atom molecular dynamics simulation

(AAMD). We modified the standard CHARMM36 parameter set using Ca2+ NBFIX35,38 and Na+

CUFIX.39 These nonbonded interaction Fixes improve the accuracy of ion-ion and ion-lipid

interactions by changing the minimum distance of the Lennard-Jones potential defined between

specific atom pairs. The net charge of Lipid A phosphate group was set to be -1, following the

suggestion of a recent study by Rice et al.40 To be consistent with the membrane force field, we

employed the CHARMM General Force Field41,42 to parameterize the force field and partial charge

distribution of PQS.

The CHARMM-GUI membrane builder43,44 was used to generate the initial configuration of the

model membrane and PQS. All simulations in this study were carried out using the GROMACS

2021.3 package.45,46 The membrane geometry was optimized using the steepest descent energy

minimization algorithm, followed by an isothermal–isochoric (NVT) equilibration of 100 ps at

310.15 K with the velocity-rescaling thermostat.47 The lipids and solvent were separately coupled.

After that, an isothermal–isobaric (NPT) equilibration was performed for 1 ns at 1 bar with semi-

isotropic pressure coupling using the Berendsen barostat while maintaining the temperature

through the Berendsen thermostat.48 The final equilibration was performed for 500-1000 ns in the

NPT ensemble at 310.15 K and 1 bar with semi-isotropic pressure coupling using the Parrinello-

Rahman algorithm49 together with the Nosé-Hoover thermostat.50 All bonds were constrained by

the LINCS algorithm.51 The cutoff of short-range van der Waals and electrostatic interactions was

set to 1.2 nm with the neighbor list updated every 20 time steps. Long-range electrostatic

interactions were calculated using the particle-mesh Ewald method.52 The time step for the leap-

frog integrator was set to 2 fs.

Barostat: Parrinello-Rahman versus Stochastic Cell Rescaling

Performing MD simulations at a constant pressure requires a numerical algorithm to control the

system stress, referred to as a barostat. It has been shown that the most commonly used Parrinello-

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

1Effectoflocalstressonaccuratemodelingofbacterialoutermembranesusingall-atommoleculardynamicsEmadPirhadi1,JuanM.Vanegas†2,MithilaN.Farin1,JeffreyW.Schertzer3,XinYong1*1DepartmentofMechanicalEngineering,BinghamtonUniversity,Binghamton,NewYork2DepartmentofBiochemistryandBiophysics,OregonStateUniversit...

展开>> 收起<<

1 Effect of local stress on accurate modeling of bacterial outer membranes using all -atom molecular dynamics.pdf

共29页,预览5页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

1 Effect of local stress on accurate modeling of bacterial outer membranes using all -atom molecular dynamics

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: