Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Visualizing Interactions along the Escherichia coli Twin-Arginine Translocation Pathway Using Protein Fragment Complementation

  • Jan S. Kostecki,

    Affiliation Department of Biomedical Engineering, Cornell University, Ithaca, New York, United States of America

  • Haiming Li,

    Affiliation Department of Biological Sciences, University of Calgary, Calgary, Alberta, Canada

  • Raymond J. Turner,

    Affiliation Department of Biological Sciences, University of Calgary, Calgary, Alberta, Canada

  • Matthew P. DeLisa

    md255@cornell.edu

    Affiliations Department of Biomedical Engineering, Cornell University, Ithaca, New York, United States of America, School of Chemical and Biomolecular Engineering, Cornell University, Ithaca, New York, United States of America

Visualizing Interactions along the Escherichia coli Twin-Arginine Translocation Pathway Using Protein Fragment Complementation

  • Jan S. Kostecki, 
  • Haiming Li, 
  • Raymond J. Turner, 
  • Matthew P. DeLisa
PLOS
x

Abstract

The twin-arginine translocation (Tat) pathway is well known for its ability to export fully folded substrate proteins out of the cytoplasm of Gram-negative and Gram-positive bacteria. Studies of this mechanism in Escherichia coli have identified numerous transient protein-protein interactions that guide export-competent proteins through the Tat pathway. To visualize these interactions, we have adapted bimolecular fluorescence complementation (BiFC) to detect protein-protein interactions along the Tat pathway of living cells. Fragments of the yellow fluorescent protein (YFP) were fused to soluble and transmembrane factors that participate in the translocation process including Tat substrates, Tat-specific proofreading chaperones and the integral membrane proteins TatABC that form the translocase. Fluorescence analysis of these YFP chimeras revealed a wide range of interactions such as the one between the Tat substrate dimethyl sulfoxide reductase (DmsA) and its dedicated proofreading chaperone DmsD. In addition, BiFC analysis illuminated homo- and hetero-oligomeric complexes of the TatA, TatB and TatC integral membrane proteins that were consistent with the current model of translocase assembly. In the case of TatBC assemblies, we provide the first evidence that these complexes are co-localized at the cell poles. Finally, we used this BiFC approach to capture interactions between the putative Tat receptor complex formed by TatBC and the DmsA substrate or its dedicated chaperone DmsD. Our results demonstrate that BiFC is a powerful approach for studying cytoplasmic and inner membrane interactions underlying bacterial secretory pathways.

Introduction

The bulk of protein transport across the inner membrane of Gram-negative bacteria occurs via the well-characterized Sec export pathway [1][4]. Sec export involves the membrane translocation of polypeptides that are largely unfolded and effectively ratchet their way through the Sec pore in a process requiring ATP hydrolysis [5], [6]. A fundamentally different pathway known as the twin-arginine translocation (Tat) system operates alongside the Sec pathway. The hallmark of the Tat pathway that distinguishes it from the Sec mechanism is the ability to transport proteins of varying dimension that have acquired a largely, if not completely, folded conformation [7][10]. Studies on the Tat mechanism have demonstrated that the integral membrane proteins TatA, TatB, and TatC form the minimal components necessary for exporting folded proteins in E. coli. The TatA and TatB components are single-span integral membrane proteins while TatC has been shown to contain six transmembrane spans [11]. These membrane proteins have been observed to form two distinct complexes: one that is comprised of multiple subunits of TatA and a second that contains predominantly TatB and TatC [12][14]. TatA homo-oligomers form a variable diameter ring structure that may serve as a protein-conducting channel [15] or a patch that facilitates translocation by local destabilization of the bilayer [16]. The TatB and TatC proteins form a complex to which substrates initially bind [17], suggesting that TatBC serves as the twin-arginine signal peptide binding site.

Proteins that transit the Tat pathway do so because they fold too rapidly to remain competent for Sec-dependent export [18] or because they bind protein subunits [9], [19] and/or redox cofactors [20], such as FeS clusters or molybdopterin centers, in the cytoplasm. This raises the important question of how the Tat pathway determines whether a substrate is sufficiently folded, including the assembly of subunits or cofactors, prior to the membrane translocation step. At least three mechanisms operate prior to, or concomitant with, translocation through the Tat pore that serve to prevent wasteful or harmful export of premature or improperly folded substrates. First, a folding quality control mechanism has been proposed on the basis that misfolded or partially folded proteins are not exported via the Tat pathway [7], [21][23]. Recent evidence suggests that the Tat translocase itself apparently “senses” the substrate folded state [24]. Second, Tat export is regulated at an earlier stage by additional “proofreading” factors that recognize specific Tat signal peptides and/or mature domains. These factors include dedicated chaperones such as DmsD and TorD that coordinate the cofactor-insertion and export processes [25], [26] and general molecular chaperones (e.g., DnaK, SlyD) that affect the stability and targeting of certain substrates [26][28]. Third, the Tat apparatus appears to directly initiate the turnover of rejected substrate molecules [29].

Direct visualization of the molecular interactions between proteins can reveal important details about how protein-protein interactions execute and regulate a wide range of events inside living cells. A number of fluorescence-based methods have been developed and widely used for visualizing and identifying interacting proteins including fluorescence resonance energy transfer (FRET) [30], [31] and bimolecular fluorescence complementation (BiFC) [32], [33]. In the case of BiFC, a fluorescent protein is split into two non-fluorescent fragments that are fused to a pair of interacting proteins. Interaction of the two proteins brings the split fragments into close proximity, resulting in reassembly of the fluorescent protein. Hence, reconstituted fluorescence is coupled to the interaction of the two proteins and can be used to conveniently determine how, when and where two proteins interact inside living cells. The power of this technique for capturing interactions along the secretory pathway of mammalian cells was first demonstrated by Michnick and coworkers [34]. In a similar vein, we demonstrate here that BiFC enabled visualization of a wide range of protein-protein interactions that constitute early steps in the Tat translocation cycle. We focused on interactions that had previously been established by alternative techniques or for which previous studies had led to conflicting results. These included: (i) the binding between soluble proteins such as the Tat substrate dimethyl sulfoxide reductase (DmsA) with its dedicated proofreading chaperone DmsD; (ii) the assembly of transmembrane proteins such as TatA with itself or TatB with TatC; and (iii) the targeting of soluble proteins to transmembrane subunits such as DmsA docking on TatC. Our results confirm that BiFC is a powerful tool for molecular dissection of key mechanistic steps of the Tat export process and provide the first robust screening platform of protein-protein interactions along this important pathway.

Results

Development of BiFC for Tat Substrate-Chaperone Interactions

To visualize protein interactions between soluble and transmembrane factors that participate in various steps of the Tat export process, we employed BiFC based on split fragments of enhanced yellow fluorescent protein (YFP). Our first target was the well-characterized interaction between E. coli DmsA and its cognate binding chaperone DmsD (Fig. 1a). The DmsD chaperone recognizes the DmsA twin-arginine signal peptide [26] and helps orchestrate the biogenesis and assembly of the DmsA enzyme [35]. It has been suggested that this interaction serves as a proofreading step that prevents premature export of incompletely folded DmsA [36], [37]. Since the DmsA signal peptide (ssDmsA) alone is sufficient to interact with DmsD [26], we first tested whether ssDmsA fused to the N-terminal YFP fragment (ssDmsA-Y1) interacted with DmsD fused to the C-terminal YFP fragment (DmsD-Y2). As evidenced by fluorescence microscopy, wt TG1 cells expressing these two chimeras emitted strong fluorescence (Fig. 2a) that was nearly 5 times brighter than the background from control cells co-expressing an unfused version of Y1 with DmsD-Y2 (Fig. 2b). The low levels of background fluorescence observed for control cells was likely due to self-assembly of the YFP fragments in the cytoplasm. An equally strong fluorescent phenotype was observed when the same constructs were expressed in a ΔtatC derivative of TG1 that is incapable of Tat-specific transport (Fig. 2a and b), indicating that the interaction was not dependent on a functional Tat system (see below). Importantly, when ssDmsA was replaced with the Sec-dependent PhoA signal peptide (ssPhoA), no fluorescence above background was observed (Fig. 2a and b) verifying that the fluorescence seen following co-expression of ssDmsA-Y1 and DmsD-Y2 was highly specific for the ssDmsA-DmsD interaction. It is noteworthy that replacement of the YFP fragments with similarly designed fragments derived from a monomeric variant of RFP [38] gave nearly identical complementation results for the ssDmsA-DmsD interaction (Fig. S1). This suggests that the BiFC signal seen above was due to the specificity of this tandem chaperone/signal peptide system and was not an artifact of the split reporter protein.

thumbnail
Figure 1. Protein interactions detected via BiFC along the Tat pathway of E. coli.

Splitting YFP into fragments Y1 and Y2 can be used to visualize interactions between: (a) two soluble cytoplasmic proteins; (b) a transmembrane protein with itself; (c) two different transmembrane proteins; and (d) a soluble cytoplasmic protein and a transmembrane protein.

https://doi.org/10.1371/journal.pone.0009225.g001

thumbnail
Figure 2. BiFC illuminates DmsA-DmsD interaction.

(a) Fluorescence microscopy of wt TG1 cells expressing ssDmsA-Y1/DmsD-Y2 (left), TG1 ΔtatC cells expressing ssDmsA-Y1/DmsD-Y2 (center), and wt TG1 cells expressing ssPhoA-Y1/DmsD-Y2 (right). (b) Flow cytometric analysis of cells expressing constructs as indicated. Median fluorescence was obtained for each cell population and normalized to the median fluorescence measured for TG1 cells expressing ssDmsA-Y1/DmsD-Y2 (median fluorescence value for this interaction was M = 2247). Data was reported as the average of 6 replicate experiments (n = 6) and error bars represent the standard error of the mean (sem). (c) Western blot analysis of periplasmic (per) and cytoplasmic (cyt) fractions from wt TG1 or TG1 ΔtatC cells expressing ssDmsA-Y1/DmsD-Y2 or ssPhoA-Y1/DmsD-Y2 as indicated. YFP1 was detected by virtue of a C-terminal FLAG tag using anti-FLAG antibody. YFP2 and GroEL proteins were detected using anti-GFP or anti-GroEL antibodies, respectively.

https://doi.org/10.1371/journal.pone.0009225.g002

To address whether the engineered ssDmsA-Y1 chimera was still faithfully recognized and exported to the periplasm by the Tat translocase, we determined the subcellular location of ssDmsA-Y1 following its co-expression with DmsD-Y2 in wt or ΔtatC cells. As expected, a portion of the ssDmsA-Y1 was localized to the periplasm in wt cells but not in tatC-deficient cells (Fig. 2c). For comparison, DmsD-Y2 was observed exclusively in the cytoplasm of both these strains (Fig. 2c). It should be noted that the Sec-dependent substrate ssPhoA-Y1, like its ssDmsA-Y1 counterpart, accumulated in both the cytoplasm and the periplasm of wt cells (Fig. 2c), indicating that the lack of YFP complementation for the ssPhoA-Y1 construct was not due to poor expression/stability or to highly efficient translocation via the Sec pathway. Taken together, these results indicate that ssDmsA-Y1 is capable of transiting the Tat pathway.

Tat Substrate-Chaperone Interactions Do Not Require the TatABCE Proteins

We next sought to determine whether binding of DmsA by its cognate chaperone DmsD required the TatABCE proteins that comprise the translocase or instead was uncoupled from these components. Following co-expression of ssDmsA-Y1 and DmsD-Y2 in various tat-deficient strain backgrounds, we observed significant binding of ssDmsA by DmsD even when the Tat system was partially (ΔtatE) or completely (ΔtatB, ΔtatC, ΔtatAE and ΔtatABCE) inactivated (Fig. 3a). In addition to the ssDmsA-Y1 reporter protein, we constructed a chimera comprised of the entire DmsA enzyme (DmsA-Y1) to determine if the BiFC could be used to evaluate the binding of full-length Tat substrates by proofreading chaperones. Co-expression of DmsA-Y1 with DmsD-Y2 in wt cells resulted in a fluorescent signal that was significantly above background but only about 50% of that observed for the ssDmsA-Y1/DmsD-Y2 pair (Fig. 3b). We attribute this decrease to the lower level of soluble expression observed for the full-length DmsA-Y1 construct compared to ssDmsA-Y1 (see Fig. S2d). Similar to ssDmsA-Y1, DmsA-Y1 interacted strongly with DmsD in various tat-deficient mutants (Fig. 3b). Overall, our BiFC results are entirely consistent with the view that proofreading chaperones operate at an early stage of Tat export and their substrate binding activity is uncoupled from the membrane translocation step [36].

thumbnail
Figure 3. Specificity determinants of substrate/chaperone interactions.

Co-expression of (a) ssDmsA-Y1/DmsD-Y2 or (b) DmsA-Y1/DmsD-Y2 in various TG1 tat deletion strains as indicated. TG1 (−) indicates cells that co-expressed Y1 lacking the ssDmsA signal peptide and DmsD-Y2. Median cell fluorescence was obtained via flow cytometry and normalized to that for wt TG1 cells co-expressing ssDmsA-Y1/DmsD-Y2. Data was reported as the average of 6 replicate experiments (n = 6) and error bars represent the sem. (c) Co-expression of DmsD-Y2 with wt ssDmsA-Y1, full-length DmsA-Y1, or twin-lysine (KK) variants of ssDmsA-Y1 or DmsA-Y1 in a wt TG1 background. Chaperones DmsD and TorD each fused to Y2 were co-expressed with their cognate or non-cognate signal sequences (ssDmsA-Y1, ssTorA-Y1, ssNarG-Y1). Median cell fluorescence was obtained via flow cytometry and normalized to that for wt TG1 cells co-expressing ssDmsA-Y1/DmsD-Y2. Data was reported as the average of 6 replicate experiments (n = 6) and error bars represent the sem.

https://doi.org/10.1371/journal.pone.0009225.g003

To verify that the BiFC signals from the DmsA/DmsD interactions were due to physical association between the proteins, we performed a co-purification experiment using an 8x polyhistidine-tagged version of DmsD-Y2. Co-expression of this construct with ssDmsA-Y1 or DmsA-Y1 in TG1 ΔtatABCE cells was performed, followed by Ni-NTA chromatography. SDS-PAGE analysis of the elution fractions collected from the column revealed that these fractions contained both the 8xHis-DmsD-Y2 and DmsA-Y1 (Fig. S2a) or ssDmsA-Y1 (data not shown). Moreover, all of the elution fractions were fluorescent, indicating that the recovered 8xHis-DmsD-Y2 was associated with ssDmsA-Y1 or DmsA-Y1 (Fig. S2b). Native PAGE analysis of the elution fractions revealed fluorescent complexes that migrated at the expected sizes for DmsA-Y1/8xHis-DmsD-Y2 and ssDmsA-Y1/8xHis-DmsD-Y2 (Fig. S2c). Western blot analysis of these same fractions confirmed the presence of the DmsA-Y1 or ssDmsA-Y2 fusion proteins in these affinity-captured complexes (Fig. S2d). Similar co-purification results were obtained using versions of DmsA or DmsD that lacked the Y1 fragments (data not shown), although the purified complexes were of course not fluorescent and the yield was lower. We suspect that the higher yield for Y1/Y2-containing complexes was the result of intermolecular stabilization or trapping afforded by the nearly irreversible assembly of the split YFP fragments. As such, the use of YFP fragments may be a convenient strategy for co-purification of interacting proteins, especially those whose association in the cell is short-lived. Overall, these results indicate that the BiFC signals observed above were due to authentic association between the substrate/chaperone pair.

Specificity Determinants of Substrate-Chaperone Interactions

Using our BiFC system, we next explored the substrate specificity of the DmsD proofreading chaperone. First, we tested whether the twin-arginine residues in ssDmsA, which are needed for functional Tat transport, were required for DmsD binding and in turn the BiFC signal. For this, we generated variants of ssDmsA-Y1 and DmsA-Y1 in which the twin-arginine residues in the (S/T)RRxFLK consensus motif were each mutated to lysine, a substitution that completely abolishes export [39], [40]. When ssDmsA(KK)-Y1 was co-expressed with DmsD-Y2 in wt TG1 cells there was no significant difference in cell fluorescence. This result was consistent with in vitro binding results observed for ssTorA(KK), which displayed identical TorD binding characteristics to its twin-arginine counterpart [25]. Cells co-expressing full-length DmsA(KK)-Y1 were also fluorescent although less so than their twin-arginine counterpart (Fig. 3c), suggesting that regions of the mature portion of DmsA may play a role in DmsD specificity. Our results with both the ssDmsA and full-length DmsA constructs support the conclusion that the twin-arginine motif itself is clearly not the overarching signal recognition factor.

To determine whether the DmsD chaperone was specific for its cognate substrate or instead exhibited promiscuity as has been seen previously [26], [41], we cloned the signal peptides from the Tat substrates DmsA, TorA and NarG as fusions with the Y1 fragment. Following co-expression of these constructs with DmsD-Y2 in wt TG1 cells, a strong BiFC signal was observed only for the ssDmsA-Y1/DmsD-Y2 pair (Fig. 3c). It should be noted that the exquisite specificity observed here for DmsD was not observed in earlier studies where DmsD was reported to bind signal peptides derived from DmsA and TorA [26], [41]. However, in vivo complementation assays with authentic DmsA and TorA substrates revealed that DmsD and TorD cannot replace one another [42], suggesting that substrate promiscuity of DmsD may be an artifact of the experimental conditions used to investigate signal peptide-chaperone binding. Even our assay was not immune to this sort of artifact as testing of TorD-Y2 against the same set of signal peptides revealed a BiFC signal for both the cognate ssTorA-Y1 and the non-cognate ssDsmA-Y1 constructs (Fig. 3c). Nonetheless, our results provide further evidence that the BiFC strategy enables direct detection of interactions between different chaperones/signal peptide pairs directly in E. coli without needing to alter the geometry (e.g., linker lengths) or orientation (N- versus C-terminal) of the YFP fragments.

Identification of Permissive Residues in DmsD Binding Pocket

To further demonstrate the utility of our BiFC assay, we attempted to isolate gain-of-function DmsD variants that bind ssDmsA more efficiently. Previous studies identified a “hot pocket” of residues in DmsD that are important for signal peptide binding [43]. In this study, a hyperbinding variant of DmsD carrying a single W87Y substitution and a lower affinity variant, DmsD(R15C/L75S), were reported. When the DmsD(R15C/L75S)-Y2 variant was co-expressed with ssDmsA-Y1, there was a clear decrease in the BiFC signal compared to wt DmsD (Fig. S3a), consistent with the earlier report. However, binding activity of DmsD(W87Y)-Y2 was indistinguishable from wt DmsD (Fig. S3a). Therefore, to experimentally identify residues in this region of DmsD that permitted signal peptide binding, we created 2 random libraries of DmsD variants using an NNK library approach that targeted residues W72/L75/F76 in the putative binding pocket [43]. The resulting DmsD libraries were screened via fluorescence-activated cell sorting (FACS) using either ssDmsA-Y1 or full-length DmsA-Y1 as the co-expressed partner. As seen in Table 1, a strong bias for hydrophobic, uncharged residues in these positions was observed, especially in positions 72 and 76 where a hydrophobic residue was found in 16/21 and 19/21 clones, respectively (7 and 10 of these, respectively, were wt in this position). Position 75 appears to be the most flexible as more than half of the clones carried a hydrophilic residue in this position, and in 2 of these cases the residue was charged (Lys, Asp). It is noteworthy that the BiFC signals emitted by all the isolated clones were comparable to the signal seen for wt DmsD(WLF)-Y2, except for DmsD(HYF) which exhibited a gain-of-function phenotype (Fig. S3a). We attribute this increased fluorescence to improved substrate binding because the expression level of each clone was unchanged relative to wt DmsD (Fig. S3c). Interestingly, much less structural variability was tolerated for these residues in DmsD when full-length DmsA-Y1 was used as substrate (Table 1). This suggests that substrate binding specificity is dependent on the context of the signal peptide and that the sequence determinants for binding of the entire native preprotein are more specific compared to the signal peptide alone. In support of this notion, the most interesting clones in the context of ssDmsA-Y1 (e.g., DmsD(R15C/L75S) and DmsD(HYF)) produced BiFC signals that were indistinguishable from wt DmsD when full-length DmsA-Y1 was the substrate (Fig. S3b).

thumbnail
Table 1. Isolation of permissive residues in the putative binding pocket of DmsD.

https://doi.org/10.1371/journal.pone.0009225.t001

Detection of Interactions between Transmembrane Components of the Tat Translocase

Previous studies have established that each of the Tat components form stable, defined, homo-multimeric complexes [13], [14], [44][48]. Hence, we next tested whether BiFC could be used to detect interactions between the integral TatABC membrane proteins that comprise the translocase and are essential for Tat export (see Fig. 1b and c). For these experiments, each Tat gene was cloned as a fusion to both Y1 and Y2 (e.g., TatA-Y1 and TatA-Y2) and expressed in TG1 cells lacking the tatABCE genes. We chose a strain background lacking all tat genes because previous studies have shown that self-assembly of individual Tat components does not strictly require any of the other Tat components [44][47]. In the case of TatA, we observed a BiFC signal that was 2–3 fold above the negative controls (Fig. 4a), albeit an order of magnitude lower than that seen for the ssDmsA/DmsD interaction described above. Co-expression of a TatA mutant with a substitution in the predicted amphipathic region (F39A) that blocks translocation activity and leads to aberrant TatA oligomers [49] was still able to assemble with wt TatA (Fig. 4a). Interestingly, co-expressed F39A-Y1 and F39A-Y2 were observed to homo-oligomerize very efficiently, with a BiFC signal that was nearly twice as fluorescent as the wt TatA homo-oligomers (Fig. 4a).

thumbnail
Figure 4. Visualizing the formation of TatA homo-oligomers.

(a) Cell fluorescence of TG1 ΔtatABCE cells expressing TatA-Y1, TatA-Y2, F39A-Y1, F39A-Y2, and the negative controls Y1 or Y2. Median fluorescence values were obtained via flow cytometric analysis and reported as the average of 3 replicate measurements (n = 3). Error bars represent the sem. (b) Bright field illumination and fluorescence microscopy for phenotypic analysis of chain complementation and fluorescence localization in TG1 ΔtatAE cells expressing various TatA chimeras as indicated.

https://doi.org/10.1371/journal.pone.0009225.g004

To test assembly of the TatA BiFC constructs under more physiologically relevant conditions, we co-expressed TatA-Y1/TatA-Y2 in TG1 ΔtatAE cells that express native TatB and TatC from the chromosome. This resulted in a clear BiFC signal compared to controls (Fig. 4b) that was quantitatively similar to the BiFC signal seen in ΔtatABCE cells (data not shown). It should be noted that the fluorescence appeared predominantly at the cell poles, consistent with earlier TatA labeling studies [50]. To confirm that the TatA-Y1 and TatA-Y2 chimeras were able to form functional translocases, we examined these cells by light microscopy. It is well known that tat-deficient strains form chains of up to 15 cells [51]. This cell division defect results from the mislocalization of two Tat-dependent amidases, AmiA and AmiC, which have been implicated in cleavage of the septum during cell division [52]. We observed that TatA-Y1 and TatA-Y2 are able to form functional translocases with endogenous TatB and TatC as evidenced by the ability of these constructs to reverse the chain phenotype of ΔtatAE cells (Fig. 4b). In contrast, co-expression of TatA(F39A)-Y1 and TatA(F39)-Y2 did not reverse the chain phenotype of ΔtatAE cells (Fig. 4b), even though these constructs yielded strong BiFC fluorescence that accumulated at the cell poles of ΔtatABCE cells (Fig. 4a). Taken together, these results indicate that the export defect of TatA(F39A) mutants does not arise from an inability of to self-assemble. Similar self-assembly studies were performed for TatB and TatC. Co-expression of TatB-Y1/TatB-Y2, but not TatC-Y1/TatC-Y2, in ΔtatABCE cells resulted in a BiFC signal (Fig. 5a). The lack of BiFC for TatC-Y1/TatC-Y2 was not attributable to instability or inactivity of the TatC fusions, or low production of TatC caused by its overexpression [47], because these chimeras formed functional translocases as evidenced by their ability to reverse the chain phenotype of ΔtatC cells (data not shown).

thumbnail
Figure 5. Assembly of fluorescent TatBC homo- and hetero-oligomers.

(a) Cell fluorescence of TG1 ΔtatABCE cells expressing TatB and TatC BiFC chimeras as indicated. In addition to wt TatC, the TatC variants P48A and E103R were also evaluated. Unfused Y1 and Y2 constructs co-expressed with TatB or TatC chimeras served as negative controls. Median fluorescence values were obtained via flow cytometric analysis and reported as the average of 3 replicate measurements (n = 3). Error bars represent the sem. Bright field illumination and fluorescence microscopy for (b) TG1 ΔtatABCE, (c) TG1 ΔtatB and (d) TG1 ΔtatC cells co-expressing TatB-Y1/TatC-Y2 or TatB-Y2/TatC-Y2 as indicated. Also shown are plasmid-free TG1 ΔtatB and ΔtatC cells (control) to illustrate the chain phenotype of Tat-deficient mutants.

https://doi.org/10.1371/journal.pone.0009225.g005

We next investigated the formation of hetero-multimeric complexes among the various Tat components. Co-expression of different pairs of Tat components (e.g., TatA-Y1 + TatC-Y2) in cells lacking the native tat genes resulted in strong BiFC signals for TatA-Y1/TatC-Y2 and TatB-Y1/TatC-Y2 and a weaker signal for the TatA-Y1/TatB-Y2 that were all 3–4 times more fluorescent than the respective negative controls (Fig. 5a). These results were entirely consistent with earlier findings that TatB and TatC form a complex containing multiple copies of each subunit [12] that serves as the binding site for Tat substrates [17], [53], and that TatB is capable of interacting with TatA even in the absence of TatC [14]. Importantly, switching the Y1 and Y2 fusion partners corroborated the BiFC signals measured for TatA-Y2/TatC-Y1 and TatB-Y2/TatC-Y1 (Fig. 5a). However, the TatA-Y2/TatB-Y1 signal was indistinguishable from the control (data not shown). This result together with the relatively low signal seen above for TatA-Y1/TatB-Y2 suggests that the TatA-TatB interaction may be considerably weaker than that of TatA-TatC and TatB-TatC. We also tested two TatC variants: TatC(P48A) carries a substitution in the first periplasmic loop region that abolishes export and partially impairs TatC interaction with TatB [49] and TatC(E103R) has a substitution in the first cytoplasmic loop between predicted transmembrane helices II and III that blocks export but does not affect TatBC complex formation [54]. In line with these earlier observations, both the TatC(P48A) and TatC(E103R) constructs produced BiFC signals when co-expressed with TatB, however the TatC(P48A) signal was weaker than the signals measured for the interaction between TatB and either wt TatC or TatC(E103R) (Fig. 5a).

Fluorescence microscopy revealed that TatBC assemblies were co-localized at the cell poles in both ΔtatABCE mutants and in the single ΔtatB or ΔtatC deletion strains (Fig. 5b–d). This is the first evidence of a polar location for TatBC complexes and is consistent with earlier findings for the individual TatB and TatC proteins [50]. Also evident in the microscopy analysis is the fact that all TatB and TatC chimeras were able to complement the corresponding single deletion strains (Fig. 5c and d). This complementation required co-expression of both TatB and TatC at nearly equal levels as independent expression of TatB or TatC chimeras was unable to complement the chain phenotype of the ΔtatB or ΔtatC cells, respectively (data not shown). Taken together, these results indicate that Tat function was not impaired under conditions of productive BiFC, but was highly sensitive to the TatBC stoichiometry.

DmsA and DmsD Interact with the TatB and TatC Proteins

As mentioned above, the TatBC complex has been implicated as the substrate-binding site [17], [53] and also as a possible docking site for the DmsD chaperone [55]. Accordingly, we next investigated interactions between the TatABC inner membrane proteins and soluble cytoplasmic factors (see Fig. 1d). Following co-expression of ssDmsA-Y1 with TatB-Y2 and TatC-Y2 in TG1 ΔtatABCE cells, a BiFC signal was observed that was 3.5- and nearly 7-fold above background, respectively (Fig. 6a). A much weaker but still significant BiFC signal was observed for full-length DmsA-Y1, especially when co-expressed with TatC-Y2 (data not shown). No signal above background was observed when ssDmsA-Y1 was co-expressed with TatA-Y2 (data not shown). To independently confirm the ssDmsA-Y1/TatC-Y2 interaction, we isolated membrane fractions from ΔtatABCE cells and analyzed these by Western blotting. When ssDmsA-Y1 was expressed alone, we detected the fusion protein in the soluble fraction but not in the membrane fraction (Fig. S4). However, when TatC-Y2 was co-expressed, the ssDmsA-Y1 construct was found to co-localize in the membrane fraction (Fig. S4) presumably due to its association with TatC-Y2.

thumbnail
Figure 6. BiFC reveals substrate and chaperone “docking” on TatB or TatC.

(a) Cell fluorescence of TG1 ΔtatABCE cells co-expressing ssDmsA-Y1 with either TatB-Y2 or TatC-Y2 as indicated. Also shown are data for the ssDmsA twin-lysine (KK) variant and the TatC variants P48A and E103R. (b) TG1 ΔtatABCE cells co-expressing DmsD with either TatB or TatC chimeras as indicated. Unfused Y1 and Y2 constructs co-expressed with TatB or TatC chimeras served as negative controls. Median fluorescence values were obtained via flow cytometric analysis and reported as the average of 3 replicate measurements (n = 3). Error bars represent the sem.

https://doi.org/10.1371/journal.pone.0009225.g006

Interestingly, when the twin arginines in ssDmsA were substituted with lysines, the BiFC signal following co-expression of TatB-Y2 and TatC-Y2 increased to 5.4- and 10.7-fold above background, respectively (Fig. 6a). This corroborates the recent observation that the twin-arginine residues of the Tat consensus motif are not essential for binding of precursor to the TatBC complex [54]. This result is also consistent with the observation above that cytoplasmic accumulation of non-exported substrates (e.g., ssDmsA(KK)) resulted in a stronger BiFC signal. When either TatC(P48)-Y2 or TatC(E103R)-Y2 was co-expressed with ssDmsA-Y1, we observed a BiFC signal that was nearly 2-fold less fluorescent than that seen for wt TatC. In the case of TatC(E103R), the reduced BiFC signal was in close agreement with recent data indicating that this variant exhibits reduced substrate binding [54]. To test whether TatBC also interacts with the DmsD chaperone, we co-expressed TatB-Y1 or TatC-Y1 with DmsD-Y2 in ΔtatABCE cells. DmsD interaction with TatB or TatC resulted in a ∼2-4-fold increase in the BiFC signal above background that was independent of the orientation of the split fragments (Fig. 6b). Interestingly, the P48A substitution resulted in a reduced BiFC signal while the E103R mutation blocked interaction with DmsD (Fig. 6b), suggesting that these residues, especially E103R, are important for the interaction of DmsD with the TatC subunit of the translocase.

Discussion

Bacterial protein export requires a wide range of protein interactions between soluble and transmembrane proteins, many of which have been difficult to detect using traditional approaches especially in the context of living cells. Here, we show that the YFP BiFC overcomes these limitations and enables a detailed analysis of numerous protein-protein interactions along the bacterial Tat pathway of live cells. Among these were interactions between (i) soluble cytoplasmic proteins, (ii) transmembrane proteins with themselves or a different transmembrane protein; and (iii) soluble cytoplasmic proteins and transmembrane proteins. Although not tested here, we anticipate that specific interactions between two soluble periplasmic proteins or a periplasmic protein with a transmembrane protein will also be detectable by protein fragmentation analysis. The challenge with detecting interactions on the periplasmic side of the inner membrane is that YFP and its relatives (e.g., GFP, CFP) do not attain a fluorescent conformation in the periplasm [56] unless delivered there in an already folded conformation via the Tat system [40], [57], [58]. Hence, assembly of split fluorescent proteins in the periplasm may not yield a fluorescent signal. One solution is to use split mRFP [38] (and Fig. S1) since full-length mRFP can fold into a fluorescent conformation in the E. coli periplasm (our unpublished observations). Alternatively, one could employ other protein fragment complementation systems such as split β-lactamase that are compatible with assembly and folding in the periplasm [59]. It is also noteworthy that, even though not a problem in our studies, the BiFC system could be further improved by increasing the solubility of the split YFP fragments, especially Y1, using protein engineering strategies. The reduced solubility of Y1 fusion proteins can be partially offset by expressing these from a high-copy vector while co-expressing Y2 chimeras from a low-copy vector, a strategy that was used here and elsewhere [60].

We have shown that the BiFC system can be an effective tool for confirming hypotheses regarding the Tat mechanism as well as for generating new experimental insights on how the Tat system functions. For instance, it is now generally accepted that several layers of quality control regulate the export of Tat substrate proteins [61]. The first layer, which we were able to visualize, is the association of specific molecular chaperones (e.g., DmsD, TorD) with Tat substrates. These interactions are thought to be important in substrate folding as well as in preventing premature export of improperly or incompletely folded proteins [26][28], [62][64]. Our results with ssDmsA-Y1 support the notion that substrate specificity of Tat chaperones is governed by the signal peptide, however our data also indicate that the mature domain of DmsA makes an important contribution to chaperone binding. In fact, there were a few cases where we observed measurable differences for interactions involving ssDmsA versus full-length DmsA, highlighting that care should be taken when interpreting data from chaperone binding experiments where signal peptides are used as surrogates for the full-length preprotein substrate. The involvement of chaperones has also led to the interesting hypothesis that these proteins guide their substrates to the translocase. In support of this hypothesis, biochemical studies revealed that DmsD interacted tightly with the E. coli inner membrane and that the TatB and TatC subunits were important for this interaction [55]. Our BiFC results confirm that DmsD interacts specifically with TatB and TatC (Fig. 6b), but not TatA (our unpublished observations). We also observed that a small fraction of the ssDmsA/DmsD complexes co-localized to the cell poles (Fig. 2a and also our unpublished observations), which is also where the TatBC receptor was observed to co-localize (Fig. 5b–d). These findings provide the first genetic evidence that DmsD may play a role as a targeting factor that delivers substrates to the TatBC receptor complex. To confirm this, we are currently developing a three-way BiFC-based FRET interaction system [65] to investigate whether DmsA/DmsD/TatB (or TatC) form a ternary complex in living cells. A final step prior to substrate export appears to be evaluation of a substrate's folding state by the Tat apparatus. Indeed, mounting evidence indicates that the Tat system generally discriminates against unfolded substrates [7], [10], [23], [24] (although at least two exceptions exist [16], [66]) and it has been suggested that this folding quality control may be performed directly by the Tat translocase [24], [29]. Thus, although not directly investigated here, we anticipate that our BiFC system will enable genetic dissection of this poorly understood aspect of Tat protein export and should provide some insights into the path of a Tat precursor following its recognition by TatBC up to a step where it is brought into close vicinity of TatA.

Materials and Methods

Bacterial Strains, Plasmids, Growth and Induction Conditions

The bacterial strains and plasmids used in this study are described in Table 2. For cloning purposes, E. coli MC4100 cells were grown aerobically in either liquid LB media or on solid LB media with agar (LBA). For the BiFC assay, TG1 cells were made electrocompetent by standard methods [67], transformed with equal plasmid concentrations, and grown overnight on solid LB media and antibiotics (BD Diagnostic Systems) at 37°C. The next morning individual colonies were picked from the plates, placed into 3 mL of liquid LB with antibiotics in 16–18 mm culture tubes, and grown aerobically for 4 hrs at 37°C and 200 rpm until the optical density reached OD600 ∼0.5. Isopropyl β-D-thiogalactoside (IPTG) was added to a final concentration of 1 mM for induction of protein expression, the culture was then moved to a room temperature incubator (20–24°C) at 200 rpm for the next 8 hrs. Fluorescence was only measured for cells grown at room temperature. All single knockout TG1 Tat mutants were generated by P1 transduction from the Keio collection [68]. Strain TG1 ΔtatABCE was first created by P1 transduction of ΔtatE::KanR from the Keio collection; the kanamycin resistance was removed as described previously [69], and P1 transduction was performed again from BW25113 ΔtatABC::aac [70], however the apramycin resistance was not removed. Antibiotic selection was maintained for all markers on plasmids at the following concentrations: ampicillin (Amp), 100 µg/mL; chloramphenicol (Cam), 20 µg/mL; kanamycin (Kan), 50 µg/mL; and tetracycline (Tet), 10 µg/mL.

Construction of Plasmids

Plasmid pDmsDT25 was constructed previously by amplifying E. coli dmsD from genomic DNA via PCR and cloning into the XbaI and KpnI sites of pKNT25 [43]. The resulting plasmid harbors a chimeric gene encoding dmsD fused to the T25 fragment of the catalytic domain of Bordetella pertussis adenylate cyclase. Similarly, plasmid pDmsALT18 was constructed previously by cloning a PCR fragment encoding the signal peptide of E. coli dmsA (excluding the signal peptide cleavage site) into the PstI and KpnI sites of pUT18 [43]. To establish the BiFC assay system, PCR fragments encoding the N- (1–154 aa) and C-terminal (155–238 aa) halves of the enhanced yellow fluorescent protein (YFP), abbreviated as Y1 and Y2 respectively, were amplified from pIAF817YFP (a gift from Dr. Rolf Morosoli). Plasmids pDmsD-Y2 and pssDmsA-Y1 were constructed by replacing the T25 and T18 fragments in plasmids pDmsDT25 and pDmsALT18 with Y2 and Y1, respectively. The linker sequences used for the fusion proteins were designed based on those used by Hu et al. [32]. All further plasmid constructions used in this study were based on these two initial plasmids. All plasmid DNA constructs were verified by sequencing.

Fluorescence Analysis

After induction of protein expression, flow cytometric data was collected on a FACSCalibur System (Becton Dickinson) at 0 and 8 hrs post induction. Samples for flow cytometry readings were prepared by diluting 50 µL of live bacterial cells directly from culture in 1 mL of 1x PBS. Median fluorescence was determined from histograms of the cell fluorescence emitted by 30,000 viable cells collected using the FACSCalibur flow cytometer in scan mode. For microscopy, 15 µL of live bacterial cells directly from culture were placed onto a microscope slide with cover slip. All images were taken under oil immersion microscopy using a Zeiss 100x/1,30 lens. Microscopy was performed on a Zeiss Axioskop 40 equipped with a Zeiss 100x/1,30 Oil Plan-NEOFLUAR lens, an X-Cite light source (EXFO, Mississauga, Ontario), a Semrock Brightline filter cube for YFP emission (YFP-2427A-ZHE) (Rochester, NY), digitally imaged with a SPOT FLEX digital camera (Diagnostic Instruments, Inc.) and controlled with Spot Imaging Software. All images captured under 100x-oil immersion microscopy using the Zeiss 100x/1,30 Oil Plan-NEOFLUAR lens were under bright field illumination (exposure 150 ms) or under UV illumination (exposure 500 ms). For RFP analysis, see Supplemental Methods S1.

DmsD NNK library construction and testing.

Random DmsD libraries were constructed by introducing diversity to the W72/L75/F76 residues of the DmsD protein. Briefly, site-directed random mutagenesis (Stratagene QuickChange® Site-Directed Mutagenesis Kit) of these residues was performed using degenerate NNK primers to amplify dmsD from plasmid pDmsD-Y2. The resulting DNA library was transformed into XL-1 Blue cells and ∼105 clones (>3x coverage) were obtained. Library cells were harvested, grown in liquid culture and plasmid DNA was isolated. The isolated plasmid DNA library was digested with SphI and KpnI to excise the diversified dmsD genes, which were subsequently ligated into pDmsD-Y2 that had been similarly digested with SphI and KpnI to remove wild-type (wt) dmsD. This was done to avoid any potential mutations in the plasmid backbone that may have been introduced during the site-directed mutagenesis reaction. This library was isolated from cells and electroporated into TG1 cells that contained either pssDmsA-Y1 or pDmsA-Y1. This library was spread on LBA plates supplemented with Amp and Kan and incubated overnight at 37°C for resolution of transformants. The cells were pooled into a 500 mL culture, allowed to grow to OD600 ∼0.5 at 37°C and 200 rpm, induced with 1 mM IPTG at room temperature and 200 rpm for 8 hrs. Aliquots of the culture were taken and resuspended in 1 mL of 1x PBS and run through a FACSCalibur flow cytometer set for cell recovery mode. The gate used on the FACSCalibur was set to recover cells with a fluorescent signal greater than the ssDmsA-Y1/DmsD-Y2 or DmsA-Y1/DmsD-Y2 BiFC signal. The recovered cells were concentrated on 0.45 µm sterile membrane filters (Whatman) and the membrane filters were transferred to LBA + Amp/Kan plates to allow single colonies to grow overnight at 37°C. Isolates from the overnight incubation were then picked and grown in 96-well plates to an OD600 of ∼0.5 at 37°C and 200 rpm, induced with 1 mM IPTG at room temperature and 200 rpm for 8 hrs and then checked for fluorescence using a fluorescence microplate reader (Biotek Synergy HT) with excitation filter 485/20 and emission filter 528/20. Cells with a fluorescent signal greater than the ssDmsA-Y1/DmsD-Y2 or DmsA-Y1/DmsD-Y2 signal were grown and plasmid was harvested for DNA sequencing. Selected sequences are listed in Table 1.

Cell fractionation and protein analysis.

After 8 hrs of induction, 1 mL of cells was collected and the OD600 was measured using a spectrophotometer (Thermo Scientific Biomate3). The cells were spun down for 2 min at 13,000×g and the supernatant was removed. The periplasmic fraction from the E. coli cells was isolated using a modified protocol of the Epicentre Biotechnologies PeriPreps™ Periplasting Kit (Madison, WI), where the periplasting buffer did not contain any Ready-Lyse Lysozyme. The soluble protein fraction from the periplasted E. coli cells was isolated with BugBuster® Master Mix (Novagen) according to the manufacturer's protocol. For Western blotting, an equal volume of 2x SDS-PAGE buffer was added to the periplasmic and soluble protein fractions and then boiled for 15 min at 100°C. Samples were loaded onto 4–20% iGels (NuSep Ltd, Australia) where protein amount was normalized to the optical density of the cells taken before fractionation. After SDS-PAGE, the proteins were transferred to Immobilon-P PVDF 0.45 µm membrane (Millipore, MA) and probed for the epitope FLAG (DYKDDDK) tag on all Y1 constructs using the primary antibody anti-FLAG® M2 (Stratagene, CA). To detect the Y2 fragment, the primary antibody was anti-GFP (Roche, IN). As a cytoplasmic fractionation marker, the primary antibody anti-GroEL (Sigma) was used. The secondary antibody was always anti-mouse IgG-HRP (Promega, WI). HRP detection was via chemiluminescence using the Immun-Star HRP Chemiluminescent Kit (BioRad) and captured on X-Omat Film (Kodak). For substrate/chaperone co-purification and membrane co-localization protocols, see Supplemental Methods S1.

Supporting Information

Figure S1.

RFP BiFC reports DmsA-DmsD interaction. (a) BiFC analysis using split mRFP1Q66T in wt TG1 cells. Shown are fluorescence microscopy images of TG1 cells co-expressing ssDmsA-R1 and DmsD-R2, as well as controls co-expressing unfused R1 and/or R2 as indicated. (b) Quantification of mRFP1Q66T BiFC signals using a fluorescence microplate reader for the same cells as in (a). Whole cell fluorescence values were normalized to the fluorescence emission from wt TG1 cells expressing ssDmsA-R1/DmsD-R2 and reported as the average of 3 replicate measurements (n = 3). Error bars represent the sem.

https://doi.org/10.1371/journal.pone.0009225.s001

(1.62 MB TIF)

Figure S2.

Co-purification of substrate/chaperone pairs from the cytoplasm of E. coli. (a) Purification of 8xHis-DmsD-Y2 from TG1 ΔtatABCE cells co-expressing DmsA-Y1. Lanes were loaded with (from left to right): MW, molecular weight ladder; 1, cell lysate; 2, 50k MWCO filtrate; 3, flow-through; 4, 5 mM imidazole; 5, 60 mM imidazole; 6, 80 mM imidazole; 7, 100 mM imidazole; 8, 150 mM imidazole; 9, 1000 mM imidazole. Numbers to the left correspond to the MW of the ladder proteins. Gel was stained with BioRad BioSafe Coomassie Blue and imaged on a BioRad ChemiDoc. (b) UV illumination of elution fractions corresponding to lanes 5–9 in (a). (c) Native PAGE analysis of elution fractions corresponding to lanes 6–9 in (a) from DmsA-Y1 expressing cells. Also shown are similar elution fractions generated from cells co-expressing ssDmsA-Y1 with 8xHis-DmsD-Y2. PAGE gel was illuminated using UV transilluminator. (d) Western blot analysis of samples in (c) using anti-FLAG antibodies that recognize the C-terminal FLAG tag on DmsA-Y1 and ssDmsA-Y1.

https://doi.org/10.1371/journal.pone.0009225.s002

(3.37 MB TIF)

Figure S3.

Isolation of gain-of-function chaperones. (a) Cell fluorescence of DmsD-Y2 library isolates (HYF, YLF, FYL, IVT) following co-expression with ssDmsA-Y1 in TG1 cells. Two previously characterized mutants (R15C/L75S and W87Y) were included for comparison. Unfused Y2 co-expressed with ssDmsA-Y1 served as a negative control. (b) Cell fluorescence of the same library isolates described in (a) but co-expressed with full-length DmsA-Y1 in TG1 cells. All median fluorescence values obtained via flow cytometric analysis were normalized to the signal obtained for ssDmsA-Y1/DmsD-Y2 signal. These normalized values are reported as the average of 3 replicate measurements (n = 3). Error bars represent the sem. (c) Western blot analysis of the cytoplasmic (c) or periplasmic (p) fractions isolated from cells co-expressing ssDmsA-Y1 with the DmsD-Y2 variants as indicated. GroEL served as a fractionation marker for cytoplasmic protein.

https://doi.org/10.1371/journal.pone.0009225.s003

(1.00 MB TIF)

Figure S4.

Co-localization of ssDmsA-Y1 with TatC in E. coli membranes. Western blot analysis of soluble and membrane fractions isolated from TG1 ΔtatABCE cells expressing ssDmsA-Y1 alone or co-expressing ssDmsA-Y1 with TatC-Y2. Blot was probed with anti-FLAG antibodies for detection of ssDmsA-Y1. Numbers to the left indicate the molecular weight (MW) of the ladder proteins. Two separate aliquots from the fraction collected from the top of the 70% sucrose layer (total membrane fraction) were analyzed side-by-side on the blot. An equivalent amount of soluble or membrane proteins was added to each lane.

https://doi.org/10.1371/journal.pone.0009225.s004

(0.58 MB TIF)

Acknowledgments

We thank Rolf Morosoli (INRS-Institut Armand-Frappier, Université du Québec, Laval, QC, Canada) for his generous gift of plasmid pIAF817YFP used in this study.

Author Contributions

Conceived and designed the experiments: JSK HL RJT MPD. Performed the experiments: JSK HL. Analyzed the data: JSK RJT MPD. Wrote the paper: JSK MPD.

References

  1. 1. Pugsley AP (1993) The complete general secretory pathway in gram-negative bacteria. Microbiol Rev 57: 50–108.AP Pugsley1993The complete general secretory pathway in gram-negative bacteria.Microbiol Rev5750108
  2. 2. Mori H, Ito K (2001) The Sec protein-translocation pathway. Trends Microbiol 9: 494–500.H. MoriK. Ito2001The Sec protein-translocation pathway.Trends Microbiol9494500
  3. 3. Driessen AJ, Fekkes P, van der Wolk JP (1998) The Sec system. Curr Opin Microbiol 1: 216–22.AJ DriessenP. FekkesJP van der Wolk1998The Sec system.Curr Opin Microbiol121622
  4. 4. Driessen AJ, Manting EH, van der Does C (2001) The structural basis of protein targeting and translocation in bacteria. Nat Struct Biol 8: 492–8.AJ DriessenEH MantingC. van der Does2001The structural basis of protein targeting and translocation in bacteria.Nat Struct Biol84928
  5. 5. Schatz G, Dobberstein B (1996) Common principles of protein translocation across membranes. Science 271: 1519–26.G. SchatzB. Dobberstein1996Common principles of protein translocation across membranes.Science271151926
  6. 6. Stuart RA, Neupert W (2000) Making membranes in bacteria. Nature 406: 575, 577.RA StuartW. Neupert2000Making membranes in bacteria.Nature406575, 577
  7. 7. DeLisa MP, Tullman D, Georgiou G (2003) Folding quality control in the export of proteins by the bacterial twin-arginine translocation pathway. Proc Natl Acad Sci U S A 100: 6115–20.MP DeLisaD. TullmanG. Georgiou2003Folding quality control in the export of proteins by the bacterial twin-arginine translocation pathway.Proc Natl Acad Sci U S A100611520
  8. 8. Hynds PJ, Robinson D, Robinson C (1998) The sec-independent twin-arginine translocation system can transport both tightly folded and malfolded proteins across the thylakoid membrane. J Biol Chem 273: 34868–74.PJ HyndsD. RobinsonC. Robinson1998The sec-independent twin-arginine translocation system can transport both tightly folded and malfolded proteins across the thylakoid membrane.J Biol Chem2733486874
  9. 9. Rodrigue A, Chanal A, Beck K, Muller M, Wu LF (1999) Co-translocation of a periplasmic enzyme complex by a hitchhiker mechanism through the bacterial Tat pathway. J Biol Chem 274: 13223–8.A. RodrigueA. ChanalK. BeckM. MullerLF Wu1999Co-translocation of a periplasmic enzyme complex by a hitchhiker mechanism through the bacterial Tat pathway.J Biol Chem274132238
  10. 10. Sanders C, Wethkamp N, Lill H (2001) Transport of cytochrome c derivatives by the bacterial Tat protein translocation system. Mol Microbiol 41: 241–6.C. SandersN. WethkampH. Lill2001Transport of cytochrome c derivatives by the bacterial Tat protein translocation system.Mol Microbiol412416
  11. 11. Behrendt J, Standar K, Lindenstrauss U, Bruser T (2004) Topological studies on the twin-arginine translocase component TatC. FEMS Microbiol Lett 234: 303–8.J. BehrendtK. StandarU. LindenstraussT. Bruser2004Topological studies on the twin-arginine translocase component TatC.FEMS Microbiol Lett2343038
  12. 12. Bolhuis A, Mathers JE, Thomas JD, Barrett CM, Robinson C (2001) TatB and TatC form a functional and structural unit of the twin-arginine translocase from Escherichia coli. J Biol Chem 276: 20213–9.A. BolhuisJE MathersJD ThomasCM BarrettC. Robinson2001TatB and TatC form a functional and structural unit of the twin-arginine translocase from Escherichia coli.J Biol Chem276202139
  13. 13. Porcelli I, de Leeuw E, Wallis R, van den Brink-van der Laan E, de Kruijff B, et al. (2002) Characterization and membrane assembly of the TatA component of the Escherichia coli twin-arginine protein transport system. Biochemistry 41: 13690–7.I. PorcelliE. de LeeuwR. WallisE. van den Brink-van der LaanB. de Kruijff2002Characterization and membrane assembly of the TatA component of the Escherichia coli twin-arginine protein transport system.Biochemistry41136907
  14. 14. de Leeuw E, Granjon T, Porcelli I, Alami M, Carr SB, et al. (2002) Oligomeric properties and signal peptide binding by Escherichia coli Tat protein transport complexes. J Mol Biol 322: 1135–46.E. de LeeuwT. GranjonI. PorcelliM. AlamiSB Carr2002Oligomeric properties and signal peptide binding by Escherichia coli Tat protein transport complexes.J Mol Biol322113546
  15. 15. Gohlke U, Pullan L, McDevitt CA, Porcelli I, de Leeuw E, et al. (2005) The TatA component of the twin-arginine protein transport system forms channel complexes of variable diameter. Proc Natl Acad Sci U S A 102: 10482–6.U. GohlkeL. PullanCA McDevittI. PorcelliE. de Leeuw2005The TatA component of the twin-arginine protein transport system forms channel complexes of variable diameter.Proc Natl Acad Sci U S A102104826
  16. 16. Cline K, McCaffery M (2007) Evidence for a dynamic and transient pathway through the Tat protein transport machinery. EMBO J 26: 3039–49.K. ClineM. McCaffery2007Evidence for a dynamic and transient pathway through the Tat protein transport machinery.EMBO J26303949
  17. 17. Alami M, Luke I, Deitermann S, Eisner G, Koch HG, et al. (2003) Differential interactions between a twin-arginine signal peptide and its translocase in Escherichia coli. Mol Cell 12: 937–46.M. AlamiI. LukeS. DeitermannG. EisnerHG Koch2003Differential interactions between a twin-arginine signal peptide and its translocase in Escherichia coli.Mol Cell1293746
  18. 18. Ribnicky B, Van Blarcom T, Georgiou G (2007) A scFv antibody mutant isolated in a genetic screen for improved export via the twin arginine transporter pathway exhibits faster folding. J Mol Biol 369: 631–9.B. RibnickyT. Van BlarcomG. Georgiou2007A scFv antibody mutant isolated in a genetic screen for improved export via the twin arginine transporter pathway exhibits faster folding.J Mol Biol3696319
  19. 19. Waraho D, DeLisa MP (2009) Versatile selection technology for intracellular protein-protein interactions mediated by a unique bacterial hitchhiker transport mechanism. Proc Natl Acad Sci U S A 106: 3692–7.D. WarahoMP DeLisa2009Versatile selection technology for intracellular protein-protein interactions mediated by a unique bacterial hitchhiker transport mechanism.Proc Natl Acad Sci U S A10636927
  20. 20. Berks BC (1996) A common export pathway for proteins binding complex redox cofactors? Mol Microbiol 22: 393–404.BC Berks1996A common export pathway for proteins binding complex redox cofactors?Mol Microbiol22393404
  21. 21. Fisher AC, DeLisa MP (2009) Efficient isolation of soluble intracellular single-chain antibodies using the twin-arginine translocation machinery. J Mol Biol 385: 299–311.AC FisherMP DeLisa2009Efficient isolation of soluble intracellular single-chain antibodies using the twin-arginine translocation machinery.J Mol Biol385299311
  22. 22. Fisher AC, Kim W, DeLisa MP (2006) Genetic selection for protein solubility enabled by the folding quality control feature of the twin-arginine translocation pathway. Protein Sci 15: 449–58.AC FisherW. KimMP DeLisa2006Genetic selection for protein solubility enabled by the folding quality control feature of the twin-arginine translocation pathway.Protein Sci1544958
  23. 23. Richter S, Bruser T (2005) Targeting of unfolded PhoA to the Tat translocon of Escherichia coli. J Biol Chem 280: 42723–30.S. RichterT. Bruser2005Targeting of unfolded PhoA to the Tat translocon of Escherichia coli.J Biol Chem2804272330
  24. 24. Panahandeh S, Maurer C, Moser M, DeLisa MP, Muller M (2008) Following the path of a twin-arginine precursor along the TatABC translocase of Escherichia coli. J Biol Chem 283: 33267–75.S. PanahandehC. MaurerM. MoserMP DeLisaM. Muller2008Following the path of a twin-arginine precursor along the TatABC translocase of Escherichia coli.J Biol Chem2833326775
  25. 25. Hatzixanthis K, Clarke TA, Oubrie A, Richardson DJ, Turner RJ, et al. (2005) Signal peptide-chaperone interactions on the twin-arginine protein transport pathway. Proc Natl Acad Sci U S A 102: 8460–5.K. HatzixanthisTA ClarkeA. OubrieDJ RichardsonRJ Turner2005Signal peptide-chaperone interactions on the twin-arginine protein transport pathway.Proc Natl Acad Sci U S A10284605
  26. 26. Oresnik IJ, Ladner CL, Turner RJ (2001) Identification of a twin-arginine leader-binding protein. Mol Microbiol 40: 323–31.IJ OresnikCL LadnerRJ Turner2001Identification of a twin-arginine leader-binding protein.Mol Microbiol4032331
  27. 27. Graubner W, Schierhorn A, Bruser T (2007) DnaK plays a pivotal role in Tat targeting of CueO and functions beside SlyD as a general Tat signal binding chaperone. J Biol Chem 282: 7116–24.W. GraubnerA. SchierhornT. Bruser2007DnaK plays a pivotal role in Tat targeting of CueO and functions beside SlyD as a general Tat signal binding chaperone.J Biol Chem282711624
  28. 28. Perez-Rodriguez R, Fisher AC, Perlmutter JD, Hicks MG, Chanal A, et al. (2007) An essential role for the DnaK molecular chaperone in stabilizing over-expressed substrate proteins of the bacterial twin-arginine translocation pathway. J Mol Biol 367: 715–30.R. Perez-RodriguezAC FisherJD PerlmutterMG HicksA. Chanal2007An essential role for the DnaK molecular chaperone in stabilizing over-expressed substrate proteins of the bacterial twin-arginine translocation pathway.J Mol Biol36771530
  29. 29. Matos CF, Robinson C, Di Cola A (2008) The Tat system proofreads FeS protein substrates and directly initiates the disposal of rejected molecules. EMBO J 27: 2055–63.CF MatosC. RobinsonA. Di Cola2008The Tat system proofreads FeS protein substrates and directly initiates the disposal of rejected molecules.EMBO J27205563
  30. 30. Nguyen AW, Daugherty PS (2005) Evolutionary optimization of fluorescent proteins for intracellular FRET. Nat Biotechnol 23: 355–60.AW NguyenPS Daugherty2005Evolutionary optimization of fluorescent proteins for intracellular FRET.Nat Biotechnol2335560
  31. 31. Tsien RY (1998) The green fluorescent protein. Annu Rev Biochem 67: 509–44.RY Tsien1998The green fluorescent protein.Annu Rev Biochem6750944
  32. 32. Hu CD, Chinenov Y, Kerppola TK (2002) Visualization of interactions among bZIP and Rel family proteins in living cells using bimolecular fluorescence complementation. Mol Cell 9: 789–98.CD HuY. ChinenovTK Kerppola2002Visualization of interactions among bZIP and Rel family proteins in living cells using bimolecular fluorescence complementation.Mol Cell978998
  33. 33. Kerppola TK (2006) Visualization of molecular interactions by fluorescence complementation. Nat Rev Mol Cell Biol 7: 449–56.TK Kerppola2006Visualization of molecular interactions by fluorescence complementation.Nat Rev Mol Cell Biol744956
  34. 34. Nyfeler B, Michnick SW, Hauri HP (2005) Capturing protein interactions in the secretory pathway of living cells. Proc Natl Acad Sci U S A 102: 6350–5.B. NyfelerSW MichnickHP Hauri2005Capturing protein interactions in the secretory pathway of living cells.Proc Natl Acad Sci U S A10263505
  35. 35. Ray N, Oates J, Turner RJ, Robinson C (2003) DmsD is required for the biogenesis of DMSO reductase in Escherichia coli but not for the interaction of the DmsA signal peptide with the Tat apparatus. FEBS Lett 534: 156–60.N. RayJ. OatesRJ TurnerC. Robinson2003DmsD is required for the biogenesis of DMSO reductase in Escherichia coli but not for the interaction of the DmsA signal peptide with the Tat apparatus.FEBS Lett53415660
  36. 36. Jack RL, Buchanan G, Dubini A, Hatzixanthis K, Palmer T, et al. (2004) Coordinating assembly and export of complex bacterial proteins. EMBO J 23: 3962–72.RL JackG. BuchananA. DubiniK. HatzixanthisT. Palmer2004Coordinating assembly and export of complex bacterial proteins.EMBO J23396272
  37. 37. Jack RL, Dubini A, Palmer T, Sargent F (2005) Common principles in the biosynthesis of diverse enzymes. Biochem Soc Trans 33: 105–7.RL JackA. DubiniT. PalmerF. Sargent2005Common principles in the biosynthesis of diverse enzymes.Biochem Soc Trans331057
  38. 38. Jach G, Pesch M, Richter K, Frings S, Uhrig JF (2006) An improved mRFP1 adds red to bimolecular fluorescence complementation. Nat Methods 3: 597–600.G. JachM. PeschK. RichterS. FringsJF Uhrig2006An improved mRFP1 adds red to bimolecular fluorescence complementation.Nat Methods3597600
  39. 39. Cristobal S, de Gier JW, Nielsen H, von Heijne G (1999) Competition between Sec- and Tat-dependent protein translocation in Escherichia coli. EMBO J 18: 2982–90.S. CristobalJW de GierH. NielsenG. von Heijne1999Competition between Sec- and Tat-dependent protein translocation in Escherichia coli.EMBO J18298290
  40. 40. DeLisa MP, Samuelson P, Palmer T, Georgiou G (2002) Genetic analysis of the twin arginine translocator secretion pathway in bacteria. J Biol Chem 277: 29825–31.MP DeLisaP. SamuelsonT. PalmerG. Georgiou2002Genetic analysis of the twin arginine translocator secretion pathway in bacteria.J Biol Chem2772982531
  41. 41. Chan CS, Chang L, Rommens KL, Turner RJ (2009) Differential interactions between Tat-specific redox enzyme peptides and their chaperones. J Bacteriol 191: 2091–101.CS ChanL. ChangKL RommensRJ Turner2009Differential interactions between Tat-specific redox enzyme peptides and their chaperones.J Bacteriol1912091101
  42. 42. Ilbert M, Mejean V, Iobbi-Nivol C (2004) Functional and structural analysis of members of the TorD family, a large chaperone family dedicated to molybdoproteins. Microbiology 150: 935–43.M. IlbertV. MejeanC. Iobbi-Nivol2004Functional and structural analysis of members of the TorD family, a large chaperone family dedicated to molybdoproteins.Microbiology15093543
  43. 43. Chan CS, Winstone TM, Chang L, Stevens CM, Workentine ML, et al. (2008) Identification of residues in DmsD for twin-arginine leader peptide binding, defined through random and bioinformatics-directed mutagenesis. Biochemistry 47: 2749–59.CS ChanTM WinstoneL. ChangCM StevensML Workentine2008Identification of residues in DmsD for twin-arginine leader peptide binding, defined through random and bioinformatics-directed mutagenesis.Biochemistry47274959
  44. 44. De Leeuw E, Porcelli I, Sargent F, Palmer T, Berks BC (2001) Membrane interactions and self-association of the TatA and TatB components of the twin-arginine translocation pathway. FEBS Lett 506: 143–8.E. De LeeuwI. PorcelliF. SargentT. PalmerBC Berks2001Membrane interactions and self-association of the TatA and TatB components of the twin-arginine translocation pathway.FEBS Lett5061438
  45. 45. Leake MC, Greene NP, Godun RM, Granjon T, Buchanan G, et al. (2008) Variable stoichiometry of the TatA component of the twin-arginine protein transport system observed by in vivo single-molecule imaging. Proc Natl Acad Sci U S A 105: 15376–81.MC LeakeNP GreeneRM GodunT. GranjonG. Buchanan2008Variable stoichiometry of the TatA component of the twin-arginine protein transport system observed by in vivo single-molecule imaging.Proc Natl Acad Sci U S A1051537681
  46. 46. Behrendt J, Lindenstrauss U, Bruser T (2007) The TatBC complex formation suppresses a modular TatB-multimerization in Escherichia coli. FEBS Lett 581: 4085–90.J. BehrendtU. LindenstraussT. Bruser2007The TatBC complex formation suppresses a modular TatB-multimerization in Escherichia coli.FEBS Lett581408590
  47. 47. Orriss GL, Tarry MJ, Ize B, Sargent F, Lea SM, et al. (2007) TatBC, TatB, and TatC form structurally autonomous units within the twin arginine protein transport system of Escherichia coli. FEBS Lett 581: 4091–7.GL OrrissMJ TarryB. IzeF. SargentSM Lea2007TatBC, TatB, and TatC form structurally autonomous units within the twin arginine protein transport system of Escherichia coli.FEBS Lett58140917
  48. 48. Oates J, Barrett CM, Barnett JP, Byrne KG, Bolhuis A, et al. (2005) The Escherichia coli twin-arginine translocation apparatus incorporates a distinct form of TatABC complex, spectrum of modular TatA complexes and minor TatAB complex. J Mol Biol 346: 295–305.J. OatesCM BarrettJP BarnettKG ByrneA. Bolhuis2005The Escherichia coli twin-arginine translocation apparatus incorporates a distinct form of TatABC complex, spectrum of modular TatA complexes and minor TatAB complex.J Mol Biol346295305
  49. 49. Barrett CM, Mangels D, Robinson C (2005) Mutations in subunits of the Escherichia coli twin-arginine translocase block function via differing effects on translocation activity or Tat complex structure. J Mol Biol 347: 453–63.CM BarrettD. MangelsC. Robinson2005Mutations in subunits of the Escherichia coli twin-arginine translocase block function via differing effects on translocation activity or Tat complex structure.J Mol Biol34745363
  50. 50. Berthelmann F, Bruser T (2004) Localization of the Tat translocon components in Escherichia coli. FEBS Lett 569: 82–8.F. BerthelmannT. Bruser2004Localization of the Tat translocon components in Escherichia coli.FEBS Lett569828
  51. 51. Stanley NR, Findlay K, Berks BC, Palmer T (2001) Escherichia coli strains blocked in Tat-dependent protein export exhibit pleiotropic defects in the cell envelope. J Bacteriol 183: 139–44.NR StanleyK. FindlayBC BerksT. Palmer2001Escherichia coli strains blocked in Tat-dependent protein export exhibit pleiotropic defects in the cell envelope.J Bacteriol18313944
  52. 52. Ize B, Stanley NR, Buchanan G, Palmer T (2003) Role of the Escherichia coli Tat pathway in outer membrane integrity. Mol Microbiol 48: 1183–93.B. IzeNR StanleyG. BuchananT. Palmer2003Role of the Escherichia coli Tat pathway in outer membrane integrity.Mol Microbiol48118393
  53. 53. Cline K, Mori H (2001) Thylakoid DeltapH-dependent precursor proteins bind to a cpTatC-Hcf106 complex before Tha4-dependent transport. J Cell Biol 154: 719–29.K. ClineH. Mori2001Thylakoid DeltapH-dependent precursor proteins bind to a cpTatC-Hcf106 complex before Tha4-dependent transport.J Cell Biol15471929
  54. 54. McDevitt CA, Buchanan G, Sargent F, Palmer T, Berks BC (2006) Subunit composition and in vivo substrate-binding characteristics of Escherichia coli Tat protein complexes expressed at native levels. FEBS J 273: 5656–68.CA McDevittG. BuchananF. SargentT. PalmerBC Berks2006Subunit composition and in vivo substrate-binding characteristics of Escherichia coli Tat protein complexes expressed at native levels.FEBS J273565668
  55. 55. Papish AL, Ladner CL, Turner RJ (2003) The twin-arginine leader-binding protein, DmsD, interacts with the TatB and TatC subunits of the Escherichia coli twin-arginine translocase. J Biol Chem 278: 32501–6.AL PapishCL LadnerRJ Turner2003The twin-arginine leader-binding protein, DmsD, interacts with the TatB and TatC subunits of the Escherichia coli twin-arginine translocase.J Biol Chem278325016
  56. 56. Feilmeier BJ, Iseminger G, Schroeder D, Webber H, Phillips GJ (2000) Green fluorescent protein functions as a reporter for protein localization in Escherichia coli. J Bacteriol 182: 4068–76.BJ FeilmeierG. IsemingerD. SchroederH. WebberGJ Phillips2000Green fluorescent protein functions as a reporter for protein localization in Escherichia coli.J Bacteriol182406876
  57. 57. Santini CL, Bernadac A, Zhang M, Chanal A, Ize B, et al. (2001) Translocation of jellyfish green fluorescent protein via the Tat system of Escherichia coli and change of its periplasmic localization in response to osmotic up-shock. J Biol Chem 276: 8159–64.CL SantiniA. BernadacM. ZhangA. ChanalB. Ize2001Translocation of jellyfish green fluorescent protein via the Tat system of Escherichia coli and change of its periplasmic localization in response to osmotic up-shock.J Biol Chem276815964
  58. 58. Thomas JD, Daniel RA, Errington J, Robinson C (2001) Export of active green fluorescent protein to the periplasm by the twin-arginine translocase (Tat) pathway in Escherichia coli. Mol Microbiol 39: 47–53.JD ThomasRA DanielJ. ErringtonC. Robinson2001Export of active green fluorescent protein to the periplasm by the twin-arginine translocase (Tat) pathway in Escherichia coli.Mol Microbiol394753
  59. 59. Wehrman T, Kleaveland B, Her JH, Balint RF, Blau HM (2002) Protein-protein interactions monitored in mammalian cells via complementation of beta-lactamase enzyme fragments. Proc Natl Acad Sci U S A 99: 3469–74.T. WehrmanB. KleavelandJH HerRF BalintHM Blau2002Protein-protein interactions monitored in mammalian cells via complementation of beta-lactamase enzyme fragments.Proc Natl Acad Sci U S A99346974
  60. 60. Wilson CG, Magliery TJ, Regan L (2004) Detecting protein-protein interactions with GFP-fragment reassembly. Nat Methods 1: 255–62.CG WilsonTJ MaglieryL. Regan2004Detecting protein-protein interactions with GFP-fragment reassembly.Nat Methods125562
  61. 61. Fisher AC, DeLisa MP (2004) A little help from my friends: quality control of presecretory proteins in bacteria. J Bacteriol 186: 7467–73.AC FisherMP DeLisa2004A little help from my friends: quality control of presecretory proteins in bacteria.J Bacteriol186746773
  62. 62. Chan CS, Howell JM, Workentine ML, Turner RJ (2006) Twin-arginine translocase may have a role in the chaperone function of NarJ from Escherichia coli. Biochem Biophys Res Commun 343: 244–51.CS ChanJM HowellML WorkentineRJ Turner2006Twin-arginine translocase may have a role in the chaperone function of NarJ from Escherichia coli.Biochem Biophys Res Commun34324451
  63. 63. Pommier J, Mejean V, Giordano G, Iobbi-Nivol C (1998) TorD, a cytoplasmic chaperone that interacts with the unfolded trimethylamine N-oxide reductase enzyme (TorA) in Escherichia coli. J Biol Chem 273: 16615–20.J. PommierV. MejeanG. GiordanoC. Iobbi-Nivol1998TorD, a cytoplasmic chaperone that interacts with the unfolded trimethylamine N-oxide reductase enzyme (TorA) in Escherichia coli.J Biol Chem2731661520
  64. 64. Vergnes A, Pommier J, Toci R, Blasco F, Giordano G, et al. (2006) NarJ chaperone binds on two distinct sites of the aponitrate reductase of Escherichia coli to coordinate molybdenum cofactor insertion and assembly. J Biol Chem 281: 2170–6.A. VergnesJ. PommierR. TociF. BlascoG. Giordano2006NarJ chaperone binds on two distinct sites of the aponitrate reductase of Escherichia coli to coordinate molybdenum cofactor insertion and assembly.J Biol Chem28121706
  65. 65. Shyu YJ, Suarez CD, Hu CD (2008) Visualization of ternary complexes in living cells by using a BiFC-based FRET assay. Nat Protoc 3: 1693–702.YJ ShyuCD SuarezCD Hu2008Visualization of ternary complexes in living cells by using a BiFC-based FRET assay.Nat Protoc31693702
  66. 66. Richter S, Lindenstrauss U, Lucke C, Bayliss R, Bruser T (2007) Functional Tat transport of unstructured, small, hydrophilic proteins. J Biol Chem 282: 33257–64.S. RichterU. LindenstraussC. LuckeR. BaylissT. Bruser2007Functional Tat transport of unstructured, small, hydrophilic proteins.J Biol Chem2823325764
  67. 67. Cold Spring Harbor, NY: Cold Spring Harbor Lab Press. Cold Spring Harbor, NYCold Spring Harbor Lab PressRussell DW and Sambrook J, Molecular Cloning - A Laboratory Manual 3rd Edition ed. 2001,. Russell DW and Sambrook J, Molecular Cloning - A Laboratory Manual 3rd Edition ed. 2001,.
  68. 68. Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, et al. (2006) Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2: 2006 0008.T. BabaT. AraM. HasegawaY. TakaiY. Okumura2006Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection.Mol Syst Biol22006 0008
  69. 69. Datsenko KA, Wanner BL (2000) One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A 97: 6640–5.KA DatsenkoBL Wanner2000One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products.Proc Natl Acad Sci U S A9766405
  70. 70. Lee PA, Orriss GL, Buchanan G, Greene NP, Bond PJ, et al. (2006) Cysteine-scanning mutagenesis and disulfide mapping studies of the conserved domain of the twin-arginine translocase TatB component. J Biol Chem 281: 34072–85.PA LeeGL OrrissG. BuchananNP GreenePJ Bond2006Cysteine-scanning mutagenesis and disulfide mapping studies of the conserved domain of the twin-arginine translocase TatB component.J Biol Chem2813407285