Enhanced sampling methods and their application in the study of molecular permeation in gram-negative bacteria

Antimicrobial resistance is inhibiting our ability to fight against pathogens. By selectively changing the composition and expression of influx water-filled proteins filling their outer membrane, gram- negative bacteria are able to reduce the rates at which specific polar compounds are able to permeate. A clear comprehension of the mechanism determining substrates diffusion through these pores is still missing. In this thesis, we show how biased computer simulations may offer a unique perspective in the study of molecular permeation through porins, overcoming the intrinsic limitations of both experimental techniques and standard molecular dynamics. The first test-case is Acinetobacter baumannii’s CarO. The use of substrates with varying charge and molecular weight, as well as the creation of a loop-less mutant missing the extracellular domain of the protein, allowed to determine the charge selectivity and the transition rates of polar molecules. We obtained good agreement with the results of liposome swelling assays experiments. Further, we compared the passage of two carbapenem antibiotics in a series of mutated proteins extracted from a patient undergoing long term hospital infection. We connected the mutation of few key residues to a drastic change in the internal electric field of the proteins, showing that the antibiotics follow the choreography of water molecules inside the channels. In the last section, we present a kinetic model that allows to determine for a molecule the relative probability of different conformations and the time required for the translocation through a pore. This approach allowed to connect the results of enhanced sampling MD methods with current blockages in single channel experiments.All these results together show that multiscale MD techniques can offer an exhaustive view on the mechanism of molecular diffusion through pores, helping to understand the most important charac- teristics that determine the rates of translocation of different com- pounds in gram-negative bacteria. We can use these data to com- plement experimental results and to design the next generation of antibiotics.