This project aims to elucidate the molecular mechanisms underlying membrane disruption by antimicrobial peptides. The first mechanism investigated was spontaneous peptide translocation across membranes. We established a methodology to quantify this process [Kabelka I. et al. J Chem Inf Model 2021, 61 (2), 819-830] and demonstrated that translocation can be enhanced by transmembrane proteins/peptides [Bartoš L. et al. Biophys J 2021, 120, 2296-2305]. We identified optimal properties of transmembrane proteins that locally disrupt membranes, facilitating lipid scrambling, i.e. translocation of lipids across the bilayer [Bartoš L. et al. Biophys J 2024, 123, 1-13]. Seeking native proteins that exploit this mechanism, we identified a novel scramblase with a beta-sheet fold, contrasting with the previously reported helical structures [Jahn H. et al. Nat Commun 2023, 14, 8115]. We also demonstrated that insertases use similar membrane-disrupting mechanisms to enhance protein insertion, which also enhance lipid scrambling merging the dual function of these proteins [Bartoš L. et al. Structure 2024, 32, 4, 505-510]. Given the asymmetric lipid composition of most biological membranes, we investigated the effects of lipid asymmetry on translocation, showing that the translocation barrier could differ significantly from symmetric membrane models [Bartoš L. et al. Biophys J 2024, 123, 1-10].
Inspired by natural peptides, we investigated the synergistic effect between the naturally occurring antimicrobial peptides magainin 2 and PGLa. We have shown that peptide association with the membrane is enhanced for the mixture, in which hetero dimers are formed inducing positive membrane curvature [Semeraro E.F. et al. Biophys J 2022, 121, 1-13]. Therefore, we investigated the relationship between peptide sequence and membrane curvature, which is also present in a toroidal pore. We identified the first peptide capable of sensing negative membrane curvature [Pajtinka P. & Vacha R. J Phys Chem Lett 2024, 15, 175-179]. Another natural protein was investigated for its ability to target specific membranes. Using computer simulations with different models, we have shown that the binding is not so lipid specific and that this domain binds to all negatively charged membranes [Falginella F.L. et al. BBA - Biomembr 2022, 1864, 183983].
The second common mechanism of antimicrobial peptides is pore formation. In particular, we focused on barrel-stave pores, where peptides align side by side, forming a pore rim without lipid headgroup participation. We develop a computational approach for the de novo design of α-helical peptides that self-assemble into stable and large transmembrane barrel pores with a central nano-sized functional channel. We address the lack of existing design guidelines for the de novo pore-forming peptides and propose 52 sequence patterns leading to 10 000 000 000 000 000 (10^15) possible pore-forming sequences, each of which can be tailored for different applications using the identified roles of its residues. To validate the pore-forming ability, we employed a variety of biophysical methods including calcein leakage assay with large unilamellar vesicles, translocation of fluorescently labelled dextran of different sizes into giant unilamellar vesicles, atomic force microscopy on surface supported membranes, and conductance measurements on planar lipid membranes. All these methods consistently demonstrated the formation of transmembrane pores of approximately uniform size [Deb R. et al. Biorxiv 10.1101/2022.05.09.491086]. Subsequently, we used the identified roles of all residues to fine-tune these peptides into antimicrobial agents capable of killing even antibiotic-resistant ESKAPE pathogens at nanomolar concentrations while exhibiting low cytotoxicity to human cells and a skin model. The peptides exhibited good stability in human serum and shown even in vivo antibacterial efficacy in a preclinical mouse model of Acinetobacter baumannii infection. In addition, these nanopore-forming peptides can be similarly fine-tuned for other medical and biotechnological applications, as we demonstrated by making the peptides pH-sensitive and even anticancer [Deb R. et al. J Med Chem 2024, in press 10.1021/acs.jmedchem.4c00912].
