Crystal structures of an Extracytoplasmic Solute Receptor from a TRAP transporter in its open and closed forms reveal a helix-swapped dimer requiring a cation for alpha-keto acid binding

Abstract

Background: The import of solutes into the bacterial cytoplasm involves several types of membrane transporters, which may be driven by ATP hydrolysis (ABC transporters) or by an ion or H+ electrochemical membrane potential, as in the tripartite ATP-independent periplasmic system (TRAP). In both the ABC and TRAP systems, a specific periplasmic protein from the ESR family (Extracytoplasmic Solute Receptors) is often involved for the recruitment of the solute and its presentation to the membrane complex. In Rhodobacter sphaeroides, TakP (previously named SmoM) is an ESR from a TRAP transporter and binds alpha-keto acids in vitro.

Results: We describe the high-resolution crystal structures of TakP in its unliganded form and as a complex with sodium-pyruvate. The results show a limited "Venus flytrap" conformational change induced by substrate binding. In the liganded structure, a cation (most probably a sodium ion) is present and plays a key role in the association of the pyruvate to the protein. The structure of the binding pocket gives a rationale for the relative affinities of various ligands that were tested from a fluorescence assay. The protein appears to be dimeric in solution and in the crystals, with a helix-swapping structure largely participating in the dimer formation. A 30 A-long water channel buried at the dimer interface connects the two ligand binding cavities of the dimer.

Conclusion: The concerted recruitment by TakP of the substrate group with a cation could represent a first step in the coupled transport of both partners, providing the driving force for solute import. Furthermore, the unexpected dimeric structure of TakP suggests a molecular mechanism of solute uptake by the dimeric ESR via a channel that connects the binding sites of the two monomers.

Figure 1

Binding of 2-keto acids to TakP. A: The amplitude of the integrated fluorescence band (normalized to its extent in the absence of ligand) was plotted as a function of the concentration of added pyruvate. The concentration of binding sites (monomeric unit) was 50 nM. The solid line is a fit with the equation given in the text, yielding K_d ≈ 270 nM. The dashed line is a numerical simulation of a (slightly) cooperative model, assuming that the first binding event occurs with K_d(1) = 270 nM and the second one with Kd(2) = 0.75 × Kd(1). When fitted by a Hill equation, this model corresponds to a Hill number n ≈ 1.1. Any larger cooperativity would increase the sigmoidal character of the binding curve and could not be consistent with the data. B-D: Fluorescence amplitude change, molecular structure and dissociation constant obtained using oxobutyrate (B), oxovalerate (C) and 4-Methyl-2-Oxovalerate (D).

Figure 2

Oligomeric state of TakP in solution. A: Elution profile of a gel filtration experiment with TakP in 50 mM NaCl, 20 mM Tris HCl pH 8.0. A standard curve is superimposed and was calculated using the known molecular weight of four protein standards (blue circles). The theoretical molecular weight of TakP is 39 kDa. B: Denaturing gel electrophoresis before (Control) and after 3 hours incubation with the cross-linker glutaraldehyde (50 mM). Two concentrations of TakP were used as indicated. Each lane contains 2 μg of protein.

Figure 3

Overall structure of TakP. A: View of the TakP monomer colored as a ramp from blue to red from the N- to C-Terminus. B: View of the TakP monomer with colors according to the different structural domains. C, D: Two different orientations of the TakP dimer. Both monomers are colored as in B but one is slightly transparent and result in a paler coloring.

Figure 4

Comparison of the TakP ad SiaP structure. Stereo figure of a superimposition of the Cα positions of TakP (blue) and SiaP (magenta), both in their unliganded form.

Figure 5

A conserved dimeric interface. A: The structure of one monomer is represented as a surface that is colored according to the sequence conservation pattern generated using 100 homologous sequences and as found by ConSurf [42]. B: Weblogo representation of the terminal swapped helix using the same set of sequences. In this representation the overall height of a stack indicates the sequence conservation at that position, while the height of symbols within the stack indicates the relative frequency of each amino acid at that position.

Figure 6

Cation-pyruvate binding to TakP and structural changes upon ligand binding. A: View of the binding region with the electron density omit map (omitting the pyruvate and the sodium ion from the calculation) together with a stick representation of the protein residues involved in the binding of the ligand. The sodium ion is represented as a purple sphere. The helix in red, visible at bottom of the panel, belongs to the other monomer. B: View of the overall changes induced by ligand binding. The unliganded protein is displayed as a cyan ribbon and the liganded protein is gray. For clarity, only the bound state of the other monomer is shown (gray surface). The distance between the two molecules of pyruvate from each monomer is 35Å. C, D: View of the interdomain closing (C, no ligand; D, liganded protein). TakP is represented in CPK, with the same color coding as in A, except for the residues interacting with sodium pyruvate, which are pictured dark blue and orange for residues belonging to Domain I and II, respectively. The pyruvate molecule (black, barely visible) is completely buried.

Figure 7

View of the binding pockets and connecting channel. A: a long water channel (red mesh) located at the interface of the two monomers opens onto both pyruvate binding sites. B, C, D: Models showing how various ligands may fit into the binding pocket : oxobutyrate (B), oxovalerate (C) and 4-methyl-2 oxovalerate (D). The K_d's found for these molecules is indicated in Figure 1.

Figure 8

Model of interaction of the TakP dimer with its membrane partners TakQ and TakM. In this model, a membrane-bound pathway is hypothesized (heavy arrows). The channel at the dimer interface would facilitates the passage of the substrate from the solvent-exposed monomer to the other monomer that interacts with a membrane subunit. This model does not exclude the possibility that binding of the loaded ESR may happen (light arrows).