Membrane Permeation of Psychedelic Tryptamines by Dynamic Simulations

Renewed scientific interest in psychedelic compounds represents one of the most promising avenues for addressing the current burden of mental health disorders. Classic psychedelics are a group of compounds that exhibit structural similarities to the naturally occurring neurotransmitter serotonin (5-HT). Acting on the 5-HT type 2A receptors (HT2ARs), psychedelics induce enduring neurophysiological changes that parallel their therapeutic psychological and behavioral effects. Recent preclinical evidence suggests that the ability of psychedelics to exert their action is determined by their ability to permeate the neuronal membrane to target a pool of intracellular 5-HT2ARs. In this computational study, we employ classical molecular dynamics simulations and umbrella sampling techniques to investigate the permeation behavior of 12 selected tryptamines and to characterize the interactions that drive the process. We aim at elucidating the impact of N-alkylation, indole ring substitution and positional modifications, and protonation on their membrane permeability. Dimethylation of the primary amine group and the introduction of a methoxy group at position 5 exhibited an increase in permeability. Moreover, there is a significant influence of positional substitutions on the indole groups, and the protonation of the molecules substantially increases the energy barrier at the center of the bilayer, making the compounds highly impermeable. All the information extracted from the trends predicted by the simulations can be applied in future drug design projects to develop psychedelics with enhanced activity.


Membrane Equilibration
As described above, initially, a 200 ns of MD simulation was performed to achieve structural equilibration of the solvated POPC membrane.Two properties were calculated to evaluate the convergence of the equilibration process.The first property to be computed was the area per lipid, defined as the average area that a single lipid molecule occupies on the interface.
Figure 1A shows that after around 20 ns, the area per lipid oscillates around a value of 1 64.4Å 2 , which is in good agreement with the experimental result of 64.3Å 2 at 303 K. 1 The second property was the electron density along the z-axis of the full bilayer.The density profile analysis (Figure 1B) reveals two main peaks approximately at -18 Å and 18 Å from the center of the bilayer, corresponding to the phosphate groups' locations within the POPC membrane.Moving towards the bilayer center, the electron density gradually diminishes, reaching its minimum at the terminal region of the lipid tails, farthest from the polar heads.The electron densities calculated for the first 20 ns and the entire 200 ns of the simulation are very similar, indicating that there are no significant diffusion processes occurring within the lipid bilayer.Both properties suggest that the bilayer is well equilibrated and stable throughout the simulation of 200 ns.and for different time intervals, excluding the initial steps of each window.Specifically, time intervals of 0.5 ns, 1 ns, 1.5 ns, and 2 ns were excluded.No significant changes are observed along the minima and barriers of the PMF, indicating that no initial portions of the simulations need to be discarded.To ensure convergence of the simulation time, a similar procedure was followed, but this time excluding the final steps of each window (Figure 2C).Specifically, the last 2, 4, 6, 8, 10, 20, and 30 ns were removed, which means that the PMF was computed by considering the first 38, 36, 34, 32, 30, 20 and 10 ns of each window.When examining the different profiles one can conclude that convergence is achieved after 30 ns of simulations time per window.Both convergence analyses were performed on the largest molecule, N,N-OME.Having achieved fully converged profiles, the subsequent section will present an analysis of the main features of the permeation profiles.This will be followed by an analysis of the influence of amine alkylation, ring substitution, ring position and protonation on the PMF and permeation for all the compounds under investigation.

Figure 1 :
Figure 1: (A) Area per lipid in 200 ns of membrane equilibration.(B) Electron density profile through the membrane for 20 and 200 ns of equilibration.

Figure 2 :
Figure 2: (A) Probability distribution of the reaction coordinate for TRY across the 45 windows used for umbrella sampling.(B) Potential of mean force (PMF) computed by removing 0.5, 1, 1.5 and 2 ns from the beginning of each window.(C) PMF computed by removing 2, 4, 6, 8, 10, 20 and 30 ns from the end of each window.

Figure
Figure 2B displays the PMF computed for the full simulation time (40 ns per window)