Chirality-Dependent Adsorption between Amphipathic Peptide and POPC Membrane

Abstract

The interactions between chiral molecules and cell membranes have attracted more and more attention in recent decades, due to their importance in molecular science and medical applications. It is observed that some peptides composed of different chiral amino acids may have distinct interactions with a membrane. How does the membrane exhibit a selective behavior related to the chirality of the peptides? Microscopically, the interactions between the peptides and the membrane are poorly understood. In this work, we study the interactions between an amphipathic peptide (C6) and POPC membrane with simulations. The kinetics and thermodynamics of peptide enantiomers during the adsorption to the membrane are characterized with direct simulations and umbrella sampling. It is observed that there are slow kinetics for the peptide composed of D-type amino acids. Along the observed pathways, the free energy landscapes are determined with umbrella sampling techniques. A free-energy barrier for the peptide composed of D-amino acids is observed, which is consistent with the kinetic observations. The results indicate the concurrent adsorption and rotation of the peptide helix. The local interactions between the peptides and the membrane are examined in detail, including the contact interactions between the peptides and the membrane, and the distributions of the lipids around the peptide. There are observable differences of the local interactions for the cases related to different peptide enantiomers. These results further demonstrate the importance of the rotation of peptide helix during the adsorption. More interestingly, all these kinetic differences between peptide enantiomers can be explained based on the conformations of the residue Trp and interactions between Trp and lipid molecules. These results give us a molecular understanding of the mechanism of the chirality-dependent peptide-membrane interactions, and may provide clues to designing systems which are sensitive to the chirality of membranes.

Keywords: POPC; chirality; membrane; molecular dynamics; protein; tryptophan.

Figure 1

(a) Wheel projection of peptide LC6. The sizes of the residue labels (circles) are scaled according to the location of the residue along the helical axis (larger at the N-terminus and smaller at the C-terminus). DC6 is the enantiomer of LC6. (b) Diagram of angle $\theta$ which characterizes the rotation of C6. The residues are colored in the same manner as in the wheel projection (blue for polar residues and yellow for nonpolar residues). The rotation angle $\theta$ is measured with the dihedral composed of the axis of the helix, the side chain of Trp (atom CG) and the positive direction of z-axis. See Materials and Methods for further elaborations. The arrows marked by “rotation” denote the rotation directions of C6 enantiomers. Small wheel projections of the enantiomers are shown to demonstrate the mirror symmetry. The wheel projection is plotted with NetWheels [51].

Figure 2

(a) The CCD (circles) and exponential fittings (dashed lines) of FPT at $d_z=1.5 \mathrm{nm}$. The CCD of DC6 is fitted with a hybrid of two exponential distributions with normalized weights. $d_z$ records for typical trajectories are shown in the inset cell. The dot-dashed line indicates the location of $d_z=1.5 \mathrm{nm}$. (b) The typical pre-adsorption and post-adsorption snapshots of monomeric simulations. Coloring of the peptide follows the scheme in Figure 1. The phosphorus of the membrane are shown as orange spheres. The lipid tails are shown in gray and those in the foreground of the peptides are removed for clarity.

Figure 3

The PMF landscapes for the adsorption of LC6 (a) and of DC6 in the slow mode (b). The landscapes are reweighted from the umbrella sampling simulations. The PMF values are in $k_{\mathrm{B}}T$ with $T=303.15K$.

Figure 4

Averaged score for rotation of pre- and post-adsorption states. The standard error of the mean (SEM) is plotted as an error bar. The SEM is a measure of the dispersion of sample means around the true mean and is estimated as ${\sigma }_{S_{rot}}\approx s/\sqrt{n}$, where s is the standard deviation of the sample, and n is the size of the sample (number of observed frames). The size of sample for each state in each window is typically 800 to 15,000. Relatively larger SEMs are observed for pre-adsorption states in window 8 and 9 of DC6 due to the relatively smaller sample sizes. Every pair of samples of the pre-adsorption state in each window have different means at the significance level of 0.1, according to Welch’s t-test. Guidelines are plotted as signs of the trends.

Figure 5

(a) The reweighted expectation and standard deviation of $\Delta n_{des-asc}$ (the difference of lipid numbers between the descending and ascending sides). The standard deviation is given only to show the fluctuation of $\Delta n_{des-asc}$, which reflects the flexibility of lipids. (b) The number difference ($\Delta n_{switch}=n_{switch}^L-n_{switch}^D$, where $n_{switch}^X$ is the number of lipids in the case of XC6 that experience switching between the $des$ and $asc$ sides during the simulations.) of lipids switching between descending and ascending sides.

Figure 6

The diagram of surface helix fragments and the vectors employed to characterize the Trp indole orientation of LC6 (a) and DC6 (b). The four helix fragments are colored alternately in yellow and white. The diagrams are visualized with PyMOL [56]. (c) The distribution of the orientation of Trp indole measured as $\overrightarrow{t}·\overrightarrow{l}$ for pre-adsorption states of kinetic trajectories. The pre-adsorption mean value for umbrella sampling windows of L/DC6 is marked as an unfilled star in corresponding color. (d) The distribution of contact number between Trp and lipid head groups for pre-adsorption states of kinetic trajectories. The means of the distributions in (d) are (in the order of LC6, DC6normal and DC6slow) 0.96, 0.92 and 1.54 with the sizes of the corresponding samples 50443, 70085 and 64324, respectively. (e) The averaged number of contacts between Trp and lipid head groups in the windows of umbrella sampling.