Millisecond-Long Simulations of Antibiotics Transport through Outer Membrane Channels

Abstract

To reach their target site inside Gram-negative bacteria, almost all antibiotics need to cross the outer membrane. Computational modeling of such processes can be numerically demanding due to the size of the systems and especially due to the timescales involved. Recently, a hybrid Brownian and molecular dynamics approach, i.e., Brownian dynamics including explicit atoms (BRODEA), has been developed and evaluated for studying the transport of monoatomic ions through membrane channels. Later on, this numerically efficient scheme has been applied to determine the free energy surfaces of the ciprofloxacin and enrofloxacin translocation through the porin OmpC using temperature-accelerated simulations. To improve the usability and accuracy of the approach, schemes to approximate the position-dependent diffusion constant of the molecule while traversing the pore had to be established. To this end, we have studied the translocation of the charged phosphonic acid antibiotic fosfomycin through the porin OmpF from Escherichia coli devising and benchmarking several diffusion models. To test the efficiency and sensitivity of these models, the effect of OmpF mutations on the permeation of fosfomycin was analyzed. Permeation events have been recorded over millisecond-long biased and unbiased simulations, from which thermodynamics and kinetics quantities of the translocation processes were determined. As a result, the use of the BRODEA approach, together with the appropriate diffusion model, was seen to accurately reproduce the findings observed in electrophysiology experiments and all-atom molecular dynamics simulations. These results suggest that the BRODEA approach can become a valuable tool for screening numerous compounds to evaluate their outer membrane permeability, a property important in the development of new antibiotics.

How to Enter a Bacterium: Bacterial Porins and the Permeation of Antibiotics

“…Additionally, in the case of the enrofloxacin molecule, it was possible to qualitatively reproduce the FES as well as the higher energy barrier, compared to ciprofloxacin, for crossing the eyelet region with the usage of the implicit model (Figure ). More recently, this implicit model together with additional diffusion models has made it possible to determine the energetics of the fosfomycin permeation through OmpF and its earlier mentioned mutants using millisecond long unbiased simulations . Moreover, the study also reports permeation rates of fosfomycin from these simulations as well as from applied-field simulations.…”

“…More recently, this implicit model together with additional diffusion models has made it possible to determine the energetics of the fosfomycin permeation through OmpF and its earlier mentioned mutants using millisecond long unbiased simulations. 351 Moreover, the study also reports permeation rates of fosfomycin from these simulations as well as from applied-field simulations. Thus, the BRODEA scheme recommends itself for fast screening campaigns of many compounds in order to find potential hits having improved properties for translocating through bacterial pores.…”

Permeation of Fosfomycin through the Phosphate-Specific Channels OprP and OprO of Pseudomonas aeruginosa

“…coli channel OmpF shows a barrier of about 2 kcal/mol but much shallower wells due to less specific binding. For the translocation of fosfomycin through wild-type OmpF, applied-field flux rates in the order of 10 5 molecules per second have been estimated from MD and Brownian dynamics simulations at voltages of about 50 mV as well as from reversal potential measurements . For the translocation through OprP and OprO these rates are likely lower due to the more attractive and specific interactions in the interior of the channels.…”

“…Furthermore, for the monophosphate, fosfomycin, and fosmidomycin molecules previously reported CHARMM force field parameters were employed. , A grid spacing of 1.2 Å was set to compute the long-range electrostatic interactions using the particle mesh Ewald (PME) method, and the short-range nonbonded interactions were calculated using a switching distance of 10 Å and a cutoff of 12 Å. In addition, the LINCS algorithm was used to constraint all bonds . The molecular figures were generated using VMD and PyMOL…”

What have molecular simulations contributed to understanding of Gram-negative bacterial cell envelopes?

“…In general, the simulations have been able to provide details of antibiotic-protein interactions, hydration levels of protein channels, flexibility of internal protein cavities as well as other conformational details that are not available from static structures. Due to space limitations we are unable to describe these studies in detail here, but we refer the reader to a number of key recent papers and a review [39][40][41][42].…”