Structural insight in the toppling mechanism of an energy-coupling factor transporter

Abstract

Energy-coupling factor (ECF) transporters mediate uptake of micronutrients in prokaryotes. The transporters consist of an S-component that binds the transported substrate and an ECF module (EcfAA'T) that binds and hydrolyses ATP. The mechanism of transport is poorly understood but presumably involves an unusual step in which the membrane-embedded S-component topples over to carry the substrate across the membrane. In many ECF transporters, the S-component dissociates from the ECF module after transport. Subsequently, substrate-bound S-components out-compete the empty proteins for re-binding to the ECF module in a new round of transport. Here we present crystal structures of the folate-specific transporter ECF-FolT from Lactobacillus delbrueckii. Interaction of the ECF module with FolT stabilizes the toppled state, and simultaneously destroys the high-affinity folate-binding site, allowing substrate release into the cytosol. We hypothesize that differences in the kinetics of toppling can explain how substrate-loaded FolT out-competes apo-FolT for association with the ECF module.

Figure 1. Crystal structure of folate-bound FolT1.

(a) Cartoon representation of FolT1 coloured from blue (N terminus) to red (C terminus), with folate shown in stick representation. H1-6 indicates α-helices 1-6. The carbon atoms of folate are shown in deep salmon, the oxygen and nitrogen atoms are shown in red and blue, respectively. The colour coding for oxygen and nitrogen atoms remained throughout the article. The membrane orientation was derived from the hydrophobicity of the surface, the positive inside rule and the location of aromatic residues shown in b, in which the transparent surface is coloured according to the surface electrostatic potential (negative potential in red and positive potential in blue), and tryptophan residues are shown in sticks with their carbon atoms coloured yellow. The bottom of the protein faces the cytoplasm, while the top faces the extracellular side of the membrane.

Figure 2. Transport and ATPase activity of ECF–FolT2.

(a) Transport activity of ECF–FolT2 in proteoliposomes loaded with 5 mM of MgATP (circles), 5 mM MgADP (inverted triangles), 5 mM MgAMP–PNP (triangles) or 5 mM Na₂ATP plus 5 mM EDTA (squares). (b) Efflux activity of ECF–FolT2 from proteoliposomes loaded with 5 mM of MgATP, after accumulation of radiolabelled folate for 16 min. At t=16 min, 5 mM of MgATP (circles), 5 mM MgADP (inverted triangles), 5 mM Na₂ATP plus 5 mM EDTA (squares) or 100 μM non-radiolabelled folate (diamonds) was added to the reactions. (c) ATPase activity of ECF–FolT2 reconstituted in proteoliposomes (black bars) and background ATPase activity by empty liposomes (white bars). When indicated, folate was present both in the lumen of the liposomes and in the environment. The error bars show the s.d.'s from three independent measurements.

Figure 3. Crystal structure of AMP–PNP bound ECF–FolT2.

(a) Cartoon representation of ECF–FolT2 viewed from the plane of the membrane, with EcfA and EcfA′ coloured in two shades of red, EcfT in cyan and FolT2 in yellow. AMP–PNP molecules are shown in sticks representation, with the carbon and phosphor atoms coloured grey and orange, respectively. The transmembrane helices of EcfT are indicated by TMH1-5 and the three coupling helices by CH1-3. (b) Surface representation using the subunit colours from a. Loop L1 of FolT2 and proline 71 (P71) in TMH3 of EcfT are coloured orange. The pathway leading from the open folate-binding cavity to the cytosol is indicated by the dashed arrow. (c) View along an axis perpendicular to the membrane.

Figure 4. Comparison of folate-bound FolT1 and substrate-free FolT2.

(a) Structural alignment of folate-bound FolT1 (pale yellow) and substrate-free FolT2 (bright yellow). The viewpoint is from the extracytoplasmic side of the membrane. The carbon atoms of folate are shown in pale yellow and loops L1 and L3 are coloured in pale and bright orange for FolT1 and FolT2, respectively. (b) Repositioning of loop L3 of FolT2 (grey) compared with FolT1 (yellow) disrupts the interactions with folate, of which the carbon atoms are coloured yellow. Binding-site residues are indicated using the one-letter code for amino acids. (c) Surface representation showing how loop L1 (orange) of substrate-bound FolT1 would clash with P71 (orange) of EcfT when forming a complex with the ECF module. Loop L1 of FolT2 is shown in pale orange for comparison. The dashed arrow shows the movement of loop L1 between FolT2 and substrate-bound FolT1. (d) Slice-through representation of ECF–FolT2 at the level of the dashed line in c. FolT2 is shown in surface representation with loops L1 and L3 in orange, and P71 of EcfT in orange surface representation. (e) The same slice as shown in d but with substrate-bound FolT1 replacing FolT2, with the carbon atoms of folate shown in grey. Loops L1 and L3 would clash with P71.

Figure 5. Sites of interaction between EcfT and FolT2.

(a) Slice-through of the ECF module in surface representation viewed from the extracellular side of the membrane. FolT2 has been deleted to show the hydrophobic platform on top of the coupling domain of EcfT, indicated by the oval. Colouring according to hydrophobicity, from red (hydrophobic) to grey (hydrophilic). (b) Surface of FolT2 interacting with the surface shown in a, using the same colour-coding. (c) Structural flexibility in the EcfT subunit. The apo and AMP–PNP-bound structures were superimposed by structural alignment of the ATPase subunits (dark grey surface representation). The resulting positions of the EcfT subunits are shown in cartoon representation with a viewpoint from the membrane plane. The coupling domains with coupling helices CH1-3 are in almost identical positions in the two complexes, but the membrane domains differ. The membrane domain in the apo-structure (grey) has hinged away compared with the AMP–PNP-bound structure (cyan). Proline 71 on TMH3 of EcfT as well as loops L1 and L3 in FolT2 are shown in orange.

Figure 6. AMP–PNP binding to the EcfAA′ heterodimer.

(a) Surface representation AMP–PNP-bound EcfAA′ heterodimer coloured as in Fig. 3 with the AMP–PNP molecules in stick representation. The viewpoint is from the extracytoplasmic side (b) Cartoon representation of the AMP–PNP-bound EcfAA′ heterodimer and the coupling helices CH2 and CH3 of EcfT. The conserved arginines at the C-terminal end of the coupling helices, and the interacting aspartates in the EcfAA′ heterodimer are shown in sticks. The interactions are indicated by the dashed lines. (c) AMP–PNP-binding site on EcfA′ (coloured light red). The conserved motifs found in ABC transporter ATPases are indicated. (d) 2Fo–Fc electron density at 2σ in blue contouring AMP–PNP (same viewpoint as in c).

Figure 7. Working model for the transport mechanism of group II ECF transporters.

Colours of the subunits as in Fig. 3. Starting with the complex trapped in the post-translocation state, binding of ATP is needed to release the empty S-component by disruption of the hydrophobic interface (red) (step (1)). The S-component will reorient to the outward-facing state (2) and can bind substrate on the extracellular side of the membrane (3). ATP hydrolysis in the ECF module regenerates the binding platform for the S-component (4). Possibly futile ATP hydrolysis takes place in this stage. The substrate-bound S-component can topple over in the membrane possibly aided by the vicinity of the ECF module (5). The toppled S-component binds to the ECF module via the complementary hydrophobic surfaces, coloured in red (6). Binding of the S-component to the ECF module forces the disruption of the substrate-binding site and release of the substrate into the cytoplasm.