The TatC component of the twin-arginine protein translocase functions as an obligate oligomer

Abstract

The Tat protein export system translocates folded proteins across the bacterial cytoplasmic membrane and the plant thylakoid membrane. The Tat system in Escherichia coli is composed of TatA, TatB and TatC proteins. TatB and TatC form an oligomeric, multivalent receptor complex that binds Tat substrates, while multiple protomers of TatA assemble at substrate-bound TatBC receptors to facilitate substrate transport. We have addressed whether oligomerisation of TatC is an absolute requirement for operation of the Tat pathway by screening for dominant negative alleles of tatC that inactivate Tat function in the presence of wild-type tatC. Single substitutions that confer dominant negative TatC activity were localised to the periplasmic cap region. The variant TatC proteins retained the ability to interact with TatB and with a Tat substrate but were unable to support the in vivo assembly of TatA complexes. Blue-native PAGE analysis showed that the variant TatC proteins produced smaller TatBC complexes than the wild-type TatC protein. The substitutions did not alter disulphide crosslinking to neighbouring TatC molecules from positions in the periplasmic cap but abolished a substrate-induced disulphide crosslink in transmembrane helix 5 of TatC. Our findings show that TatC functions as an obligate oligomer.

Figure 1

Phenotypic characterisation of TatC single amino‐acid variants in a tatC ⁻ and tatC ⁺ strain background. Spot tests of strain DADE (Δtat ABCD, ΔtatE; top panel) or MC4100 (tat ⁺; bottom panel) harbouring the pTAT1d vector encoding TatA, TatB and the indicated amino‐acid variant in TatC. Strain MC4100 also harboured the additional screening plasmid pTTC1 (since we noted that dominant negative phenotypes were more clear‐cut in the presence of the overproduced TorA–CAT fusion protein). tatC ⁺ indicates strain DADE harbouring wild‐type pTAT1d (top panel) or strain MC4100 harbouring pTTC1 and wild‐type pTAT1d (bottom panel) and tatC ⁻ is strain DADE harbouring the pTAT1d parental plasmid, pUniprom. Strains were grown overnight in liquid media, diluted to give an OD _600nm of 0.0001 and 5 μl aliquots were replica spotted onto: LB medium (control), LB medium containing 2% SDS or M9 medium containing glycerol and TMAO. Strains were incubated for 3 days aerobically at room temperature for growth on LB and LB–SDS containing medium and anaerobically at 37°C for growth on M9–glycerol–TMAO medium.

Figure 2

Topological organisation of TatC and location of TatC variants analysed in this study. The diagram was generated using TEXtopo (Beitz, 2000) and the positions of the transmembrane helices are indicated and single amino‐acid residues that when substituted give rise to a dominant negative phenotype are shown in red. The red boxes indicate the positions of truncations that, in the context of a TatB–TatC fusion, give rise to dominant negative activity. Residues G144 and M205 that were mutated to cysteine and used for disulphide crosslinking analysis are shown in yellow.

Figure 3

Phenotypic characterisation of TatBC fusions and TatC truncations in a tatC ⁻ and tatC ⁺ strain background. Spot tests of strain DADE (Δtat ABCD, ΔtatE; top panel) or MC4100 (tat ⁺; bottom panel) harbouring the pTAT1d vector that contains the mutated tat ABC operon encoding TatC variants or TatB‐C fusion proteins, as indicated. Strain MC4100 also harboured the additional screening plasmid pTTC1. tatC ⁺ indicates strain DADE harbouring wild‐type pTAT1d (top panel) or strain MC4100 harbouring pTTC1 wild‐type pTAT1d (bottom panel) and tatC ⁻ is strain DADE harbouring the empty vector pUniprom. Strains were grown overnight in liquid media, diluted to give an OD _600nm of 0.0001 and 5 μl aliquots were replica spotted onto: LB medium (control), LB medium containing 2% SDS or M9 medium containing glycerol and TMAO. Strains were incubated for 3 days aerobically at room temperature for growth on LB and LB–SDS containing medium and anaerobically at 37°C for growth on M9–glycerol–TMAO medium.

Figure 4

Dominant negative point substituted variants of TatC interact with TatB and TatA. Crude membrane fractions of the E . coli strain DADE (Δtat ABCD ΔtatE) harbouring pREP4 (Zamenhof and Villarejo, 1972) and over‐producing TatA, TatB and hexa‐histidine‐tagged wild‐type or amino‐acid substituted TatC, as indicated were solubilised with digitionin and the TatC‐his protein purified using nickel‐charged beads as described in Experimental Procedures. In each case a sample that was loaded onto the beads (load) along with the sample that was eluted from the beads (elute) were separated by SDS–PAGE (12% acrylamide), electroblotted and immunoreactive bands were detected with either anti‐his, anti‐TatB or anti‐TatA antisera. As a control, solubilised membranes from strain DADE/pREP4 overproducing TatA, TatB and a non‐tagged variant of TatC (lanes labelled No his) were used to show that there was no unspecific binding of TatA and TatB to the beads. Five microlitres of sample was loaded in each lane.

Figure 5

Dominant negative point substituted variants of TatC do not prevent his‐tagged wild‐type TatC from interacting with TatB. Crude membrane fractions of E . coli strain MC75CH (as MC4100, tat C _his) harbouring pTTC1 and over‐producing TatA, TatB and wild‐type or amino‐acid substituted TatC, from pTAT1d as indicated, were solubilised with digitonin and the chromosomally encoded TatC‐his protein from the chromosome of strain MC75CH was purified using nickel‐charged beads as described in Experimental Procedures. In each case, a sample that was loaded onto the beads (load) along with the sample that was eluted from the beads (elute) were separated by SDS–PAGE (12% acrylamide), electroblotted and immunoreactive bands were detected with either anti‐His or anti‐TatB. As a control, solubilised membranes from strain MC4100[pTTC1] overproducing TatA, TatB and a non‐tagged variant of TatC (lane labelled ‐his) were used to show that there was no unspecific binding of TatB to the beads.

Figure 6

Bacterial two hybrid analysis of TatC–TatC interaction. A and B. Interactions between the indicated variants of TatC fused to either the T18 or T25 fragments of Bordetella pertussis adenylate cyclase, as indicated. Error bars represent the standard error of the mean (n = 6; two technical replicates of three biological replicates). Significance was assessed using Student's t‐test, where asterisk signifies P < 0.005 and double asterisks signifies P < 0.001 relative to the value for T18‐TatC + T25‐TatC.

Figure 7

Dominant negative TatC proteins can be co‐purified with the Tat substrate, SufI. Membrane fractions were prepared from strain DADE (Δtat ABCD, Δtat E) harbouring pREP4 and either pFAT75ΔASufIhisKK or pFAT75ΔASufIhis coding for wild‐type or variant TatC (along with wild‐type TatB and his‐tagged SufI) as indicated, and resuspended in buffer to give an equivalent protein concentration. Following solubilisation with 1% digitonin, detergent‐extracted proteins were loaded onto nickel‐charged resin, washed with buffer containing 50 mM imidazole and finally eluted from the column in the presence of 10 mM EDTA. Samples (5 μl) of the load and eluate fractions were separated by SDS–PAGE (12% acrylamide) and analysed by immunoblotting with anti‐his, anti‐TatC and anti‐TatB antibodies.

Figure 8

Substrate‐induced assembly of TatA‐YFP is blocked by the TatC S66P and D150Y mutations. The Tat substrate CueO was overproduced from plasmid pQE80‐CueO (Leake et al., 2008) in strain MΔABC‐A λAry p101C*TatBCflag encoding TatB along with the wild‐type or dominant negative TatC variants as indicated, with a C‐terminal FLAG tag on TatC. A. Fluorescence micrographs showing the distribution of TatA‐YFP following overproduction of CueO for the indicated TatC variants. B. Isolated membranes of the same strains used for microscopy analysed by Western blot with antibodies against TatA, TatB or FLAG, or whole cell extracts were analysed similarly with antibodies against CueO. Eighteen micrograms of membranes were loaded per lane. CueO antibodies were raised in rabbits against the mature protein, and affinity purified.

Figure 9

Blue native (BN)‐PAGE analysis of TatBC complexes containing dominant negative TatC substitutions. Crude membrane fractions of the E . coli strain DADE (Δtat ABCD Δtat E) overproducing wild‐type or P48L, M59K, S66P, V145E or D150Y amino‐acid substituted TatC along with TatA and TatB from the pTAT1d plasmid were solubilised using 2% digitonin and samples (approximately 200 μg protein) were analysed by BN‐PAGE. The arrows to the right indicate TatBC‐containing complexes. The bottom panels show SDS–PAGE analysis of the same solubilised samples.

Figure 10

A TatC self‐crosslink in the periplasmic cap of TatC is not affected by dominant negative substitutions. A. Periplasmic view of a homology model of E. coli TatC showing positions of the P48L, M59K, S66P, V145E and D150Y dominant negative amino‐acid substitutions (red) and G144C substitution (yellow) used for disulphide crosslinking analysis. B. Intact cell suspensions (50 ml of culture of OD_600nm = 0.15) of strain DADE (ΔtatABCD ΔtatE) co‐producing TatA, TatB and the single cysteine variant G144C of TatC together with additional TatC substitutions, as indicated, from pTat101 were left untreated (control sample; C), or subjected to either oxidising (O) or reducing (R) conditions as described in Experimental Procedures. Samples (50 μg of membrane protein) were resolved by SDS–PAGE (12% acrylamide), and TatC monomers and dimers visualised by western blotting.

Figure 11

A substrate‐induced TatC self‐crosslink in transmembrane helix 5 of TatC. A. Whole cells of strain DADE (ΔtatABCD ΔtatE) harbouring pTat101 co‐producing TatA, TatB and the single cysteine variant M205C of TatC, alone or with additional plasmid pQE80‐CueO were cultured as described in Experimental Procedures and subjected to oxidising (O) or reducing (R) conditions, or left untreated (control; C). Samples (50 μg of membrane protein) were resolved by SDS–PAGE (12% acrylamide), and TatC monomers and dimers visualised by western blotting using an anti‐TatC antibody. B. Cells of the same strain harbouring pTat101 co‐producing TatA, TatB and the single cysteine variant M205C of TatC together with the indicated additional TatC substitutions alone (top panel) or with pQE80‐CueO (middle and bottom panels) were treated as in part A and samples (50 μg of membrane protein) were resolved by SDS–PAGE (12% acrylamide), and blotted with anti‐TatC (middle panel) or anti‐his antibody (bottom panel). Note that the fast migrating band detected on the anti‐his blot for the oxidised sample most probably represents a crosslinked form of CueO containing an intramolecular disulphide.