Substrate-triggered position switching of TatA and TatB during Tat transport in Escherichia coli

Abstract

The twin-arginine protein transport (Tat) machinery mediates the translocation of folded proteins across the cytoplasmic membrane of prokaryotes and the thylakoid membrane of plant chloroplasts. The Escherichia coli Tat system comprises TatC and two additional sequence-related proteins, TatA and TatB. The active translocase is assembled on demand, with substrate-binding at a TatABC receptor complex triggering recruitment and assembly of multiple additional copies of TatA; however, the molecular interactions mediating translocase assembly are poorly understood. A 'polar cluster' site on TatC transmembrane (TM) helix 5 was previously identified as binding to TatB. Here, we use disulfide cross-linking and molecular modelling to identify a new binding site on TatC TM helix 6, adjacent to the polar cluster site. We demonstrate that TatA and TatB each have the capacity to bind at both TatC sites, however in vivo this is regulated according to the activation state of the complex. In the resting-state system, TatB binds the polar cluster site, with TatA occupying the TM helix 6 site. However when the system is activated by overproduction of a substrate, TatA and TatB switch binding sites. We propose that this substrate-triggered positional exchange is a key step in the assembly of an active Tat translocase.

Keywords: Tat pathway; protein transport; transport mechanism; twin-arginine signal peptide.

Figure 1.

Development of an in vivo disulfide cross-linking protocol. Cells of strain MC4100ΔBC (ΔtatBC) harbouring plasmid p101C*BC producing TatB^L9C alongside TatC^M205C were incubated with either LB medium (control, C), or LB supplemented with 10 mM DTT (reduced; R) or (a) the indicated concentrations of CuP for 15 min or (b) with 1.8 mM CuP for 1–15 min. The reaction was quenched by addition of 8 mM NEM/12 mM EDTA, membranes were prepared and proteins were separated by SDS-PAGE (10% polyacrylamide). Cross-linked products were visualized by immunoblotting using anti-TatB_FL or anti-TatC antibodies, as indicated. (c) Aliquots of cells from the oxidized and control samples in (b) were spread on LB plates containing chloramphenicol and the number of colonies enumerated following growth at 37°C for 24 h. The y-axis shows the ratio of the number of colony forming units (cfu) obtained after incubation with 1.8 mM CuP compared to the number after incubation in LB medium only; n = 3 biological replicates, error bars are ±s.d.

Figure 2.

TatB^L9C interacts with TatC^M205C in vivo. (a) Homology model of E. coli TatC showing positions of the residues used for disulfide cross-linking analysis in yellow. The side-chains of M205 and F213 are indicated. (b,c) Western blot analysis (separated on 10% polyacrylamide gels) of membranes from E. coli strain MC4100ΔBC producing TatB^L9C alongside the indicated Cys substitutions in TatC (from plasmid p101C*BC) following exposure of whole cells to 1.8 mM CuP (oxidizing) or 10 mM DTT (reducing) for 1 min. Cross-linked products were visualized by immunoblotting using (b) an anti-TatB peptide antibody or (c) an anti-TatC antibody. The asterisks indicate likely TatBC cross-links. (d) Structural model of TatB interacting with TatC at the polar cluster site (adapted from [15]). The backbone distances between TatB^L9/TatC^M205 and TatB^G16/TatC^V198 are shown. (e) Whole cells of strain MC4100ΔBC producing TatB^L9C alongside TatC^{F94A,E103,/M205C} or TatC^{F94A/E103A/F213C} from plasmid p101C*BC (annotated TatC^FEA,M205C or TatC^FEA,F213C, respectively) were left untreated (C) or incubated for 1 min with 1.8 mM CuP (O) as indicated. Following membrane preparation, cross-links were detected with an anti-TatB peptide antibody.

Figure 3.

TatA^L9C interacts with TatC^F213C in vivo. (a,e,g,h) Western blot analysis (separated on 12.5% polyacrylamide gels) of whole cells of E. coli strain DADE-P producing the indicated Cys substitutions in TatA and TatC (and wild-type TatB, from plasmid pUNITATCC4) following exposure to 1.8 mM CuP (oxidizing) or 10 mM DTT (reducing) for 1 min. Cross-linked products were visualized by immunoblotting using anti-TatC antibodies. The asterisk in (h) indicates a faint TatAC cross-link. (b) The TatA^L9C–TatC^F213C oxidized (O) and reduced (R) samples from (a) were separately probed with an anti-TatA antibody (note that the TatA monomer that is in large excess has been run off the bottom of the gel). (c,f) Cells of strain DADE harbouring plasmid pTAT101 producing (c) TatA^L9C and wild-type TatB along with either TatC^V212C or TatC^F213C, or (f) TatA^S5C, wild-type TatB and TatC^F213C, as indicated, were incubated with 1.8 mM CuP (O) or 10 mM DTT (R) for 1 min. (d) Structural model of TatA interacting with TatC at the TatA constitutive binding site. The backbone distances between TatA^S5/L9/TatC^F213, TatA^I6/TatC^V212, TatA^A13/TatC^I220 and TatA^V17/TatC^E227 are shown. (i) Cells of strain DADE producing TatA^L9C and wild-type TatB alongside TatC^{F94A,E103A,M205C} or TatC^{F94A,E103A,F213C} (annotated TatC^FEA,M205C or TatC^FEA,F213C, respectively) from pTAT101 were left untreated (C) or incubated with 1.8 mM CuP (O) for 1 min. For (c,d), following quenching, membranes were prepared, samples separated by SDS-PAGE (12.5% polyacrylamide) and immunoblotted using an anti-TatC antibody.

Figure 4.

Models of the TatABC trimer in the resting and activated state. Three views of (a) the resting-state TatABC complex and (b) the substrate-activated TatABC complex. TatA is shown in silver, TatB gold and TatC green. Note that in (b) the substrate signal peptide is not shown as it is currently unclear precisely where it binds in the activated state.

Figure 5.

TatA and TatB can each occupy both binding sites on TatC. (a) Western blot analysis (separated on 12.5% polyacrylamide gels) of whole cells of E. coli strain DADE-P producing TatA^L9C alongside the indicated Cys substitutions in TatC (in the absence of TatB, from plasmid pUNITATCC4ΔB) following exposure to 1.8 mM CuP (oxidizing) or 10 mM DTT (reducing) for 1 min. Cross-linked products were visualized by immunoblotting using anti-TatC antibodies. (b) Western blot analysis (separated on 10% polyacrylamide gels) of membranes from E. coli strain DADE producing TatB^L9C alongside the indicated Cys substitutions in TatC (from plasmid p101C*BC) following exposure of whole cells to 1.8 mM CuP (oxidizing) or 10 mM DTT (reducing) for 1 min. Cross-linked products were visualized by immunoblotting using an anti-TatB peptide antibody. The asterisks indicate likely TatBC cross-links.

Figure 6.

TatA and TatB cross-linking patterns are altered in the presence of an overproduced Tat substrate. (a) Strain MC4100ΔBC harbouring plasmid p101C*BC producing TatB^L9C alongside TatC^M205C and plasmid pQE80-CueO where indicated, were left untreated (control, C), or incubated with 1.8 mM CuP for 1 min (O). Membrane fractions were prepared, separated by SDS-PAGE (10% polyacrylamide) and immunoblotted with anti-TatB_FL, anti-TatC as indicated. An aliquot of the soluble fraction following membrane preparation was retained and analysed by immunoblotting with an anti-Histag antibody. (b) Strain MC4100ΔBC producing the indicated Cys variants of TatB and TatC from plasmid p101C*BC and his-tagged CueO from pQE80-CueO were incubated with 1.8 mM CuP for 1 min. Following quenching, membrane fractions were separated by SDS-PAGE (10% polyacrylamide) and immunoblotted with an anti-TatC antibody. A non-oxidized sample of membranes harbouring TatB^L9C–TatC^MF213C is shown in the left-most lane. An aliquot of the soluble fraction from each sample was retained and analysed by immunoblotting with an anti-Histag antibody. (c) Whole cells of strain MC4100ΔBC producing TatB^L9C alongside TatC^F213C or TatC^{F94A,E103A,M205C} (annotated TatC^FEA,M205C) from plasmid p101C*BC, and his-tagged CueO (from pQE80-CueO) were incubated for 1 min with 1.8 mM CuP. Following membrane preparation, cross-links were detected with an anti-TatB peptide antibody or an anti-TatC antibody. An aliquot of the soluble fraction from each sample was retained and analysed by immunoblotting with an anti-Histag antibody. (d) Strain DADE harbouring plasmid pTAT101 producing wild-type TatB, TatA^L9C and either TatC^M205C or TatC^F213C along with plasmid pQE80-CueO were left untreated (control, C), or incubated with 1.8 mM CuP for 1 min (O). Membrane fractions were separated by SDS-PAGE (12.5% polyacrylamide) and immunoblotted with an anti-TatC antibody. An aliquot of the soluble fraction from each sample was retained and analysed by immunoblotting with an anti-Histag antibody. p: precursor, m: mature forms of substrate CueO-His * indicates an oxidation product of CueO.

Figure 7.

Models of the multimeric resting-state TatABC complex. Models based on (a) three or (b) four heterotrimers. Modified from Alcock et al. [15].