Proton coupling and the multiscale kinetic mechanism of a peptide transporter

Abstract

Proton-coupled peptide transporters (POTs) are crucial for the uptake of di- and tripeptides as well as drug and prodrug molecules in prokaryotes and eukaryotic cells. We illustrate from multiscale modeling how transmembrane proton flux couples within a POT protein to drive essential steps of the full functional cycle: 1) protonation of a glutamate on transmembrane helix 7 (TM7) opens the extracellular gate, allowing ligand entry; 2) inward proton flow induces the cytosolic release of ligand by varying the protonation state of a second conserved glutamate on TM10; 3) proton movement between TM7 and TM10 is thermodynamically driven and kinetically permissible via water proton shuttling without the participation of ligand. Our results, for the first time, give direct computational confirmation for the alternating access model of POTs, and point to a quantitative multiscale kinetic picture of the functioning protein mechanism.

Figure 1

PepT_Sh structure (PDB: 6EXS) and essential ionizable residues. (A) Side view parallel to the cellular membrane. The N-terminal bundle is rendered in green, the C-terminal bundle is in blue while flanking helices A and B are in red. Black dashed line indicate a salt bridge. (B) Top view from periplasm, perpendicular to the cellular membrane. The helices are indicated with labels in cycles.

Figure 2

Role of E418 in ligand binding. (A) The inward-facing holo crystal structure (PDB: 6EXS). The pore radius profile was computed by the HOLE program (69). The region that forbids water (pore radius < 1.15 Å) is colored red, the region that allows single water permeation (1.15 Å < pore radius < 2.30 Å) is colored yellow, and orange indicates pore radius > 2.30 Å. This color scheme will be used for the following molecular figures. (B) The position of the ligand and the TM7-TM10 pocket as well as TM9 in the crystal structure. (C) The minimum distance between the ligand and TM9 backbone atoms in E33^H/E311^–/E418^H (run no. 1-1) and E33^H/E311^–/E418^– (run no. 2-1) simulations. The value in the crystal structure is indicated by the dashed horizontal line.

Figure 3

The binding mode of the ligand when E418 is deprotonated. (A) A superposition of the crystal structure (gray) and the equilibrated structure with a deprotonated E418. (B) The minimum distance between the E418 carboxyl and the ligand N-terminus in E33^H/E311^–/E418^H (run no. 1-1) and E33^H/E311^–/E418^– (run no. 2-1) simulations. The value in the crystal structure is indicated by the dashed horizontal line.

Figure 4

Proton-induced ligand release in E33^H/E311^–/E418^H (run no. 3) simulations. (A) Time evolution of the z coordinate of ligand geometric center with respect to the center of the membrane (ΔZCOG). (B) The contact map between ligand functional groups and binding site residues (B) in the bound state with a deprotonated E418 (run no. 2-1), (C) in the first 200-ns metastable state in runs no. 3-1 and no. 3-3, and (D) in the first 20-ns of run no. 3-2.

Figure 5

Proton-induced conformational change. (A) Two-dimensional histogram of the minimum distance between E311 and R44 heavy atoms, and the extracellular gate size of simulations E33^H/E311^–/E418^– (run no. 2-1, blue), E33^H/E311^H/E418^– initiated from an inward-facing occluded conformation (run no. 4-1, black), and E33^H/E311^–/E418^– initiated from an inward-facing conformation (run no. 5-1, red). (B) Two-dimensional histogram of the intracellular and extracellular gate sizes. (C) The gate sizes and the water density around the intracellular gate in the E33^H/E311^–/E418^– simulation (run no. 2-1). The region corresponding to the inward-facing occluded state is highlighted by gray. A running average with a 20-ns window was performed on the time series. (D) The pore radius profile of the MD-sampled inward-facing occluded state. (E) The pore radius profile of the MD-sampled outward-facing state.

Figure 6

Structural comparison between PepT_Sh and PepT2. (A) Superposition of the MD-sampled outward-facing PepT_Sh (green) with a cryo-EM outward-facing PepT2 (gray; PDB: 7NQK). The lipid phosphorus atoms are represented by dark yellow spheres. Comparison of (B) key residue, (C) extracellular gate, and (D) intracellular gate between PepT_Sh and PepT2. The helices that form the gates are labeled in (C) and (D).

Figure 7

Characterization of proton transport between TM7 and TM10. (A) Potential of mean force (free energy profile) for proton transport between E311 and E418. The error bars were estimated from independent runs for mteadynamics or block averaging for umbrella sampling as detailed in the Methods section. (B) Two-dimensional potential of mean force of proton transport between E311 and E418 with the minimum distance between E418 carboxyl and ligand N-terminal nitrogen atoms. The inset is a zoom-in showing the strongly coupled region between proton and ligand. (C) and (D) Molecular figures showing Grotthuss proton shuttling mechanism when the ligand is present. The most probable hydronium oxygen is highlighted in purple. The ligand is shown in the van der Waals representation. The N-terminal bundle of the protein is shown in transparent yellow and the C-terminal bundle is shown in transparent green.

Figure 8

Schematic diagram of the transporter functional conformational and kinetic cycle. The dashed arrows represent the transitions whose reaction rates are not known yet from the performed simulations. The N-terminal bundle is represented by yellow sticks and the C-terminal bundle is colored in green. Note the PT rates were computed without a pH gradient.