Gating topology of the proton-coupled oligopeptide symporters

Abstract

Proton-coupled oligopeptide transporters belong to the major facilitator superfamily (MFS) of membrane transporters. Recent crystal structures suggest the MFS fold facilitates transport through rearrangement of their two six-helix bundles around a central ligand binding site; how this is achieved, however, is poorly understood. Using modeling, molecular dynamics, crystallography, functional assays, and site-directed spin labeling combined with double electron-electron resonance (DEER) spectroscopy, we present a detailed study of the transport dynamics of two bacterial oligopeptide transporters, PepTSo and PepTSt. Our results identify several salt bridges that stabilize outward-facing conformations and we show that, for all the current structures of MFS transporters, the first two helices of each of the four inverted-topology repeat units form half of either the periplasmic or cytoplasmic gate and that these function cooperatively in a scissor-like motion to control access to the peptide binding site during transport.

Graphical abstract

Figure 1

Proton Oligopeptide Symporters Comprise Four Inverted-Topology Repeat Units (A) Each inverted-topology repeat unit (labeled A–D) is made up of three transmembrane α helices (labeled 1–12) (Radestock and Forrest, 2011). (B) The inward-occluded structure of PepT_So rendered using curved cylinders (Dahl et al., 2012) to illustrate their intrinsic kinks and bends and colored according to the same scheme as in (A).

Figure 2

An Inward-Open Structure of the Bacterial Oligopeptide Transporter PepT_So (A) The structure of PepT_So in an inward-open conformation solved to 3.0 Å using X-ray crystallography. The transmembrane helices are colored from red (H1) to blue (H12) as in Figure 1. The two additional helices found in the bacterial proton oligopeptide transporters, HA and HB, are colored light gray. A lateral helix (LH) found between H6 and HA and not seen in the previous structure is highlighted. The data collection and refinement statistics can be found in Table 1. (B) This new structure of PepT_So is broadly similar to that of the lower-resolution inward-occluded structure of PepT_So (PDB: 2XUT) (Newstead et al., 2011). The C_α RMSD, excluding the HA and HB motif, between both structures is 1.7 Å (394 residues). Some differences can, however, be seen. One of these is the positions of the residues that make up the thin gate; in the new structure these are such that the peptide binding site is accessible to the cytoplasm and hence this structure is inward-open. Additional detail can be found in Figure S2. (C) An outward-open model of PepT_So, built using the repeat-swapping method. An image of the outward-open model of PepT_St is shown in Figure S2.

Figure 3

An Electron Paramagnetic Resonance Technique, DEER, Was Used to Measure Distance Distributions for Eight Pairs of Residues of PepT_So (A) The MTSL spin label has a flexible linker with a maximum length of 0.9 nm. (B) Three pairs of residues on the periplasmic face and five on the cytoplasmic face (both in green) of the transporter were labeled with the spin label MTSL. (C) Adding the spin labels requires pairs of cysteines to be introduced. The activity of these double mutants was checked using an uphill transport assay. This showed that while none of the mutants abolished transport, several did decrease the rate at which PepT_So could transport. (D) The DEER distance distributions, p(r). Error bars indicate the standard deviations from triplicate experiments.

Figure 4

The Conformational State of an MFS Transporter Can Be Accurately Determined by Passing a Spherical Probe from One Side of the Protein Structure to the Other (A) The percolation surface through the structure of PepT_So is shown. This was calculated using HOLE (Smart et al., 1996) as described in the Experimental Procedures. The surface is colored according to the maximum radius of the spherical probe; less than 1.15 Å is colored red, greater than 2.30 Å yellow and, in between, orange. The pore profile (the variation in the maximum radius of a spherical probe as a function of z) can be used to identify constrictions. The maximum radius of a probe that can pass any constriction is estimated as the average of the probe radius over a window 4 Å wide centered on the constriction (i.e. the minimum value). The periplasmic and cytoplasmic gate regions in the pore profile are colored light green and cyan and the 4 Å windows colored dark green and dark blue, respectively. (B) The same analysis repeated on the outward-open model of PepT_So. RSM, repeat-swapped models. (C) This analysis has also been repeated for PepT_St and all other known MFS structures (Figure S4). The coordinates of the crystal structures and outward-open models of both PepT_So and PepT_St are shown in red and blue, respectively.

Figure 5

The Minimum C_α-C_α Distance between the Tips of H1 & H2 and H7 & H8 Correlates Best with the State of the Periplasmic Gate and the Minimum C_α-C_α Distance between the Tips of H4 & H5 and H10 & H11 Correlates Best with the State of the Cytoplasmic Gate (A) There are three possible contiguous helix pairs in each half of the protein on the periplasmic side and only two possible contiguous helix pairs in each half of the protein on the cytoplasmic side. To determine which pairs of helix tip pairs constitute the gates of the transporter, the minimum C_α-C_α distance between the tips of all possible helix tip pairs was calculated. (B) The distance between H1 & H2 and H7 & H8 correlated most closely with the state of the periplasmic gate, as determined by HOLE (r = 0.88). (C) The distance between H4 & H5 and H10 & H11 correlated most closely with the state of the cytoplasmic gate, as determined by HOLE (r = 0.78).

Figure 6

During the Molecular Dynamics Simulations PepT_So Explores Inward-Facing, Occluded, and Some Partially Outward-Facing Conformations, as Defined by the Minimum Distance between the C_α Atoms of the Relevant Pairs of Helix Tips The density of states explored during the simulations starting from the 2XUT PepT_So structures are plotted in pink, and two representative inward-facing and outward-facing structures are shown. The coordinates of known MFS structures are plotted to provide some context, and the different quadrants of the coordinate space are labeled. The coordinates of the PepT_So and PepT_St crystal structures and repeat-swapped models (RSM) are labeled in red and blue, respectively. The results from the PepT_St simulations can be found in Figure S6.

Figure 7

Two Salt Bridges Are Predicted to Stabilize Outward-Facing Conformations of PepT_So (A) The ensemble of conformations produced by the molecular dynamics simulations were analyzed for salt bridges and the conformation classified as defined in Figure 6. Seven salt bridges in total were found. The three whose propensities are a function of the conformation of the transporter, and therefore may stabilize one of more conformational states, are shown here; the others are described in the Supplemental Information (Figure S7). (B) Two of the salt bridges (K84-D79 and D136-K439) are found on the cytoplasmic side of PepT_So, while the third (R52-D328) occurs on the periplasmic side. Since these are found in different conformations, the pull-out figures are taken from different parts of the molecular dynamics trajectories. (C) All the alanine mutants were either inactive in transport or had significantly reduced function. Error bars indicate the standard deviations from triplicate experiments.

Figure 8

The Kink Produced by the Conserved Prolines in H8 Is Important for Transport (A) In the inward-occluded experimental structure of PepT_So, H8 (in red) is kinked because of two prolines, P345 and P353 (in pink). We measured the relative motion of the C-terminal ends of H8 and H6 (pink) by attaching MTSL spin labels to the E364C R201C mutant of PepT_So. (B) The same features are highlighted in blue on the outward-open model of PepT_So, demonstrating that this model predicts a shorter distance between positions 201 and 364. (C) There is moderate overlap between the R201C E364C spin-spin distance distributions measured experimentally (black line) and those predicted from the inward-occluded crystal structure of PepT_So (filled red bars). The outward-open model instead predicts a shorter distance between the ends of H6 and H8 (filled blue bars). (D) Mutating both prolines to alanine results in a more complex spin-spin distance distribution. We suggest that H8 in the R201C E364C P345A P353A is straighter than wild-type. Consistent with this, there is now reasonable agreement between the spin-spin distance distributions measured experimentally (black line) and those predicted from the model of the outward-facing conformation (filled blue bars). (E) Mutating either or both prolines in PepT_So or PepT_St either reduces or abolishes proton-driven active transport. Error bars indicate the standard deviations from triplicate experiments.

Figure 9

The Periplasmic and Cytoplasmic Gates of Proton Oligopeptide Transporters Are Formed from Two Bundles of Four α helices (A) The periplasmic gate is formed by H1, H2, H7, and H8 (colored green) and the cytoplasmic gate is formed by H4, H5, H10, and H11 (cyan). (B) The motion of the helices is described by a pair of scissors; this captures the concerted scissoring movement of the two bundles of helices and the increased motion of H7–H12, compared with H1–H6. Also shown are schematic salt bridges that stabilize the closed gates. (C) The first two helices from each repeat unit therefore contribute to either the periplasmic gate (green) or the cytoplasmic gate (cyan). The third helix in each repeat-swapped unit is located at the periphery, and does not play a direct role in gating the transporter.