Limitations of Atomistic Molecular Dynamics to Reveal Ejection of Proteins from Charged Nanodroplets

Abstract

Atomistic molecular dynamics (MD) is frequently used to unravel the mechanisms of macroion release from electrosprayed droplets. However, atomistic MD is currently feasible for only the smallest window of droplet sizes appearing at the end steps of a droplet's lifetime. The relevance of the observations made to the actual droplet evolution, which is much longer than the simulated sizes, has not been addressed yet in the literature. Here, we perform a systematic study of the desolvation mechanisms of poly(ethylene glycol) (PEG), protonated peptides of different compositions, and proteins, to (a) obtain insight into the charging mechanism of macromolecules in larger droplets than those that are currently amenable to atomistic MD and (b) examine whether currently used atomistic MD modeling can establish the extrusion mechanism of proteins from droplets. To mimic larger droplets that are not amenable to MD modeling, we scale down the systems, by simulating a large droplet size relative to the macromolecule. MD of PEG charging reveals that, above a critical droplet size, ions are available near the backbone of the macromolecule, but charging occurs only transiently by transfer of ions from the solvent to the macroion, while below the critical size, the capture of the ion from PEG has a lifetime sufficiently long for the extrusion of a charged PEG from the aqueous droplet. This is the first report of the role of droplet curvature in the relation between macroion conformation and charging. Simulations of protonated peptides with a high degree of hydrophobicity show that partial extrusion of a peptide from the droplet surface is rare relative to desolvation by drying-out. Different from what has been presented in the literature, we argue that atomistic MD simulations have not sufficiently established the extrusion mechanism of proteins from droplets and their charging mechanism. We also argue that release of highly charged proteins can occur at an earlier stage of a droplet's lifetime than predicted by atomistic MD. In this earlier stage, we emphasize the key role of jets emanating from a droplet at the point of charge-induced instability in the release of proteins.

MD Simulations of Peptide-Containing Electrospray Droplets: Effects of Parameter Settings on the Predicted Mechanisms of Gas Phase Ion Formation

“…ESI droplets produced at the Taylor cone undergo evaporation and fission events, culminating in nanometer-sized progeny droplets . Size limitations dictate that MD simulations can only probe these final nanodroplets, while earlier stages of the process are inaccessible. ,,,− , As a result, the exact atomistic history of ESI nanodroplets is unknown, and one has to make certain assumptions regarding their properties. One of these factors is the initial analyte position within the nanodroplet.…”

“…To our knowledge, force field effects on ESI simulations have not been examined yet. Following several earlier ESI studies, ,, all data discussed above were generated with CHARMM36 . Some other ESI simulations employed an AMBER force field, ,, prompting us to also perform 20 bradykinin runs using AMBER99SB-ILDN .…”

“…MD simulations were performed using GROMACS 2020.4 with leapfrog integration in 2 fs time steps with CUDA GPU acceleration . Unless noted otherwise, all simulations used the CHARMM36 force field, reflecting its use for earlier ESI simulations. ,, Additional runs were performed using the AMBER99SB-ILDN force field . TIP4P/2005 was the default water model, while a subset of simulations employed TIP3P .…”