Interactions between polymyxin B and various bacterial membrane mimics: A molecular dynamics study

A molecular overview of the polymyxin-LPS interaction in the context of its mode of action and resistance development

With the rising incidences of antimicrobial resistance (AMR) and the diminishing options of novel antimicrobial agents, it is paramount to decipher the molecular mechanisms of action and the emergence of resistance to the existing drugs. Polymyxin, a cationic antimicrobial lipopeptide, is used to treat infections by Gram-negative bacterial pathogens as a last option. Though polymyxins were identified almost seventy years back, their use has been restricted owing to toxicity issues in humans. However, their clinical use has been increasing in recent times resulting in the rise of polymyxin resistance. Moreover, the detection of "mobile colistin resistance (mcr)" genes in the environment and their spread across the globe have complicated the scenario. The mechanism of polymyxin action and the development of resistance is not thoroughly understood. Specifically, the polymyxin-bacterial lipopolysaccharide (LPS) interaction is a challenging area of investigation. The use of advanced biophysical techniques and improvement in molecular dynamics simulation approaches have furthered our understanding of this interaction, which will help develop polymyxin analogs with better bactericidal effects and lesser toxicity in the future. In this review, we have delved deeper into the mechanisms of polymyxin-LPS interactions, highlighting several models proposed, and the mechanisms of polymyxin resistance development in some of the most critical Gram-negative pathogens.

Martini-3 Coarse-Grained Models for the Bacterial Lipopolysaccharide Outer Membrane of Escherichia coli

The outer lipopolysaccharide (LPS) membrane of Gram-negative bacteria forms the main barrier for transport of antimicrobial molecules into the bacterial cell. In this study we develop coarse-grained models for the outer membrane of Escherichia coli in the Martini-3 framework. The coarse-grained model force field was parametrized and validated using all-atom simulations of symmetric membranes of lipid A and rough LPS as well as a complete asymmetric membrane of LPS with the O-antigen. The bonded parameters were obtained using an iterative refinement procedure with target bonded distributions obtained from all-atom simulations. The membrane thickness, area of the LPS, and density distributions for the different regions as well as the water and ion densities in Martini-3 simulations show excellent agreement with the all-atom data. Additionally the solvent accessible surface area for individual molecules in water was found to be in good agreement. The binding of calcium ions with phosphate and carboxylate moieties of LPS is accurately captured in the Martini-3 model, indicative of the integrity of the highly negatively charged LPS molecules in the outer membranes of Gram-negative bacteria. The melting transition of the coarse-grained lipid A membrane model was found to occur between 300 and 310 K, and the model captured variations in area per LPS, order parameter, and membrane thickness across the melting transition. Our study reveals that the proposed Martini-3 models for LPS are able to capture the physicochemical balance of the complex sugar architecture of the outer membrane of Escherichia coli. The coarse-grained models developed in this study would be useful for determining membrane protein interactions and permeation of potential antimicrobials through bacterial membranes at mesoscopic spatial and temporal scales.

Membrane-Active Peptides Attack Cell Membranes in a Lipid-Regulated Curvature-Generating Mode

Membrane-active peptides (MAPs) exhibit great potential in biomedical applications due to their unique ability to overcome the cell membrane barrier. However, the interactions between MAPs and membranes are complex, and little is known about the possibility of MAP action being specific to certain types of membranes. In this study, a combination of molecular dynamics simulations and theoretical analysis was utilized to investigate the interactions between typical MAPs and realistic cell membrane systems. Remarkably, the simulations revealed that MAPs can attack membranes by generating and sensing positive mean curvature, which is dependent on lipid composition. Furthermore, theoretical calculations demonstrated that this lipid-regulated curvature-based membrane attack mechanism is an integrated result of multiple effects, including peptide-induced membrane wedge and softening effects, the lipid shape effect, the area-difference elastic effect, and the boundary edge effect of formed peptide-lipid nanodomains. This study enhances our comprehension of MAP-membrane interactions and highlights the potential for developing membrane-specific MAP-based agents.

Membrane-Specific Binding of 4 nm Lipid Nanoparticles Mediated by an Entropy-Driven Interaction Mechanism

Lipid nanoparticles (LNPs) are a leading biomimetic drug delivery platform due to their distinctive advantages and highly tunable formulations. A mechanistic understanding of the interaction between LNPs and cell membranes is essential for developing the cell-targeted carriers for precision medicine. Here the interactions between sub 10 nm cationic LNPs (cLNPs; e.g., 4 nm in size) and varying model cell membranes are systematically investigated using molecular dynamics simulations. We find that the membrane-binding behavior of cLNPs is governed by a two-step mechanism that is initiated by direct contact followed by a more crucial lipid exchange (dissociation of cLNP's coating lipids and subsequent flip and intercalation into the membrane). Importantly, our simulations demonstrate that the membrane binding of cLNPs is an entropy-driven process, which thus enables cLNPs to differentiate between membranes having different lipid compositions (e.g., the outer and inner membranes of bacteria vs the red blood cell membranes). Accordingly, the possible strategies to drive the membrane-targeting behaviors of cLNPs, which mainly depend on the entropy change in the complicated entropy-enthalpy competition of the cLNP-membrane interaction process, are investigated. Our work unveils the molecular mechanism underlying the membrane selectivity of cLNPs and provides useful hints to develop cLNPs as membrane-targeting agents for precision medicine.

Synergistic Membrane Disturbance Improves the Antibacterial Performance of Polymyxin B

Drug-resistant Gram-negative bacteria pose a serious threat to public health, and polymyxin B (PMB) is clinically used as a last-line therapy for the treatment of infections caused by these pathogens. However, the appearance of PMB resistance calls for an effort to develop new approaches to improve its antibacterial performance. In this work, a new type of nanocomposite, composed of PMB molecules being chemically decorated on the surface of graphene oxide (GO) nanosheets, was designed, which showed potent antibacterial ability through synergistically and physically disturbing the bacterial membrane. The as-fabricated PMB@GO nanocomposites demonstrated an enhanced bacterial-killing efficiency, with a minimum inhibitory concentration (MIC) value half of that of free PMB (with an MIC value as low as 0.5 μg mL^-1 over Escherichia coli), and a bacterial viability less than one fourth of that of PMB (with a bacterial reduction of 60% after 3 h treatment, and 90% after 6 h incubation). Furthermore, the nanocomposite displayed moderate cytotoxicity or hemolysis effect, with cellular viabilities over 85% at concentrations up to 16 times the MIC value. Studies on antibacterial mechanism revealed that the synergy between PMB molecules and GO nanosheets greatly facilitated the vertical insertion of the nanocomposite into the lipid membrane, leading to membrane disturbance and permeabilization. Our results demonstrate a physical mechanism for improving the antibacterial performance of PMB and developing advanced antibacterial agents for better clinic uses.

Polymyxins induce lipid scrambling and disrupt the homeostasis of Gram-negative bacteria membrane

Polymyxins are increasingly used as the last-line therapeutic option for the treatment of infections caused by multidrug-resistant Gram-negative bacteria. However, efforts to address the resistance in superbugs are compromised by a poor understanding of the bactericidal modes because high-resolution detection of the cell structure is still lacking. By performing molecular dynamics simulations at a coarse-grained level, here we show that polymyxin B (PmB) disrupts Gram-negative bacterial membranes by altering lipid homeostasis and asymmetry. We found that the binding of PmBs onto the asymmetric outer membrane (OM) loosens the packing of lipopolysaccharides (LPS) and induces unbalanced bending torque between the inner and outer leaflets, which in turn triggers phospholipids to flip from the inner leaflet to the outer leaflet to compensate for the stress deformation. Meanwhile, some LPSs may be detained on the inner membrane (IM). Then, the lipid-scrambled OM undergoes phase separation. Defects are created at the boundaries between LPS-rich domains and phospholipid-rich domains, which consequently facilitate the uptake of PmB across the OM. Finally, PmBs target LPSs detained on the IM and similarly perturb the IM. This lipid Scramble, membrane phase Separation, and peptide Translocation model depicts a novel mechanism by which polymyxins kill bacteria and sheds light on developing a new generation of polymyxins or antibiotic adjuvants with improved killing activities and higher therapeutic indices.