FhaC takes a bow to FHA in the two-partner do-si-do

Summary

FhaC is an outer membrane transporter from Bordetella pertussis belonging to the two-partner secretion (TPS) pathway with its primary role being the secretion of the virulence factor filamentous haemagglutinin (FHA). FhaC serves as a model transporter of the TPS pathway and significant work has been done to characterize the role of FhaC in FHA secretion. Recent studies characterized interactions between FHA and the POTRA domains of FhaC, suggesting that secretion may involve a successive translocation mechanism mediated by β-augmentation and/or electrostatic interactions. Moreover, it was also shown that reconstituted FhaC is necessary and sufficient to transport FHA into proteoliposomes. While the crystal structure of FhaC clearly suggests a role in transport, the putative transport pore is plugged by an N-terminal α-helix (H1 helix) that occludes access by FHA. Therefore, it has been proposed that the H1 helix must be expelled from the pore in order for secretion of FHA to occur. However, this has yet to be shown experimentally. In this issue of Molecular Microbiology, Guérin et al. report the first direct experimental evidence to show that the FhaC H1 helix is quite dynamic and exchanges between closed and open states upon interaction with FHA.

The secretion of proteins out of a cell can be a daunting task since the secreted protein must typically cross at least one membrane. In Gram-negative bacteria the scenario is even more challenging, since secreted proteins must cross both an inner membrane and an outer membrane. Gram-negative bacteria have developed a number of diverse secretion systems to address this issue (Henderson et al., 2004). Protein translation first occurs in the cytoplasm and the unfolded protein is then transported across the inner membrane by the Sec translocon into the periplasm. Next, periplasmic chaperones typically assist in shuttling the secreted protein to the outer membrane where cargo-specific membrane protein(s) then complete the secretion process by transporting the secreted protein across the outer membrane into the extracellular milieu. The two-partner secretion (TPS) pathway belongs to the Type V secretion system and functions to secrete cargo proteins across the outer membrane of Gram-negative bacteria (Jacob-Dubuisson et al., 2001). As the name suggests, there are two components of the TPS pathway generically termed TpsA (cargo) and TpsB (transporter). TpsA cargo proteins can range in size but are typically very large β-helices with an N-terminal TPS domain of ~250 residues. They often serve as virulence factors in pathogenic strains. TpsB transporter proteins also vary in size but have a core structure consisting of an N-terminal periplasmic domain composed of several polypeptide transport associated (POTRA) repeat domains and a large membrane inserted C-terminal β-barrel domain (Clantin et al., 2007). The archetypal system of the TPS pathway is the secretion of the virulence factor filamentous hæmagglutinin (FHA) by the transporter protein FhaC in Bordetella pertussis, the causative agent in whooping cough (Locht et al., 1993).

The structure of FhaC was reported in 2007 by Clantin et al. and served at the first structure of this family of transporters (Clantin et al., 2007). FhaC contains an N-terminal α-helix (H1) followed by two POTRA domains and a C-terminal 16-stranded membrane integrated β-barrel domain. Upon visualizing the structure of the barrel domain, it was apparent that it could serve as a secretion pore to mediate the transport of FHA across the outer membrane. However, the H1 helix plugged the pore and in this conformation, would surely prevent FHA secretion. Therefore, it was proposed that in a resting state the H1 helix serves as a plug to prevent free diffusion across the outer membrane but it would have to be ejected to allow FHA secretion. It was shown in 2011 by Delattre et al. that FHA communicates with the POTRA domains of FhaC through interactions likely mediated by electrostatics and/or β-augmentation in a successive manner (Delattre et al., 2011). While somewhat expected, these results were intriguing because this study reported distinct interactions which could represent the first step in secretion of FHA across the outer membrane by FhaC. These interactions could also represent the ‘trigger’ required to catalyze the removal of the H1 helix, although this was not shown at the time. Shortly thereafter, Fan et al. in 2012 used an in vitro reconstitution system to show that FhaC is necessary and sufficient to mediate FHA translocation into proteoliposomes, further solidifying the use of the phrase ‘two-partner’ system (Fan et al., 2012). This system has been well characterized over the past 10 years both functionally and structurally, yet there are still aspects of the secretion mechanism that remain unknown.

To address the long standing question of whether the H1 helix is ejected, Guérin et al. combined use of several techniques including electron paramagnetic resonance (EPR) spectroscopy, disulfide crosslinking, electrophysiology, and flow cytometry (Guerin et al., 2014). Initial EPR experiments using singly labeled samples showed a clear shift in conformational dynamics of the H1 helix and the linker connecting the H1 helix to POTRA 1 in the presence of FHA, while the well-conserved loop 6 remained largely immobilized. To gain insight into the possible distance distributions of these conformational changes, particularly in the H1 helix, another EPR technique called pulsed-electron double-resonance (PELDOR) spectroscopy was used. Here, samples are doubly labeled at precise locations with spin labels and flash frozen, which also freezes the protein conformations. A distance distribution between the two labels can then be calculated using PELDOR spectroscopy which yields information on the conformational states within the frozen sample. Labeling the FhaC samples in the N-terminal region of the H1 helix and in both (i) extracellular loop 7 and (ii) POTRA 2, distance distributions using PELDOR spectroscopy revealed a large conformational change of the H1 helix in the presence of FHA. This result is consistent with ejection of the H1 helix from the FhaC barrel domain.

FhaC has previously been shown to form ion-permeable channels, a presumed consequence of the conformational dynamics of the H1 helix within the pore of the barrel domain (Meli et al., 2006). To further verify that the H1 helix was being ejected, Guérin et al. performed electrophysiology and disulfide crosslinking experiments (Guerin et al., 2014). Using a FhaC construct where the tip of the H1 helix is crosslinked along extracellular loop 7 to prevent ejection, electrophysiology experiments showed that channel formation could be significantly reduced compared to controls. This indicated that the crosslink prevented ejection of the H1 helix plug which in turn blocked channel formation. Elimination of the crosslink by treatment with TCEP, which reduces disulfide bonds, restored channel properties to control levels in this FhaC crosslink variant, suggesting the H1 helix was now free to leave the pore. In vivo disulfide crosslinking experiments provided further evidence that the H1 helix was actually exiting the pore and residing within the periplasm. Here, pairs of cysteines were engineered along (i) the H1 helix and (ii) POTRA 1 or POTRA2. SDS-PAGE and Western blot analyses were then used to determine whether spontaneous crosslinks were formed as evident by an observable gel shift, which could be eliminated by reduction of the disulfide crosslink. The experiments showed that a number of spontaneous crosslinks could form between the H1 helix and the POTRA domains, indicating that the H1 helix was in fact moving out of the barrel domain and into the periplasm. Interestingly, these crosslinks also formed in the absence of FHA, suggesting that the H1 helix is actually quite dynamic irrespective of the presence of substrate.

As final proof that the H1 helix of FhaC must be removed from the pore of the barrel domain prior to FHA secretion, Guérin et al. used a clever approach where they placed a Myc-tag at the tip of the H1 helix, which is exposed to the surface, and then co-expressed either (i) a native-like substrate (Fha30) which would be fully secreted, or (ii) a chimera substrate that included a large folded domain at the C-terminus called BugE (Fha30-BugE) which would stall during secretion (Guerin et al., 2014). They used flow cytometry to monitor the presence of the Myc-tag at the surface of the cell, which indicated whether the H1 helix was in the pore of the barrel domain or displaced into the periplasm. The results showed that the presence of Fha30 only slightly reduced the percentage of cells presenting the Myc-tag compared to FhaC alone (89.3% → 85.8%), however, the presence of the chimera Fha30-BugE drastically reduced Myc-tag presentation (89.3% → 10.3%), consistent with the H1 helix being trapped in the periplasm in the presence of the chimeric substrate which stalls during secretion across the outer membrane. In separate but related experiments, while verifying the directionality of the chimeric substrate during secretion, it was also found that only when a Myc-tag is placed at the N-terminus of Fha30-BugE (rather than the C-terminus) is it presented at the surface. This observation is consistent with the hypothesis that the N-terminal TPS domain of the FHA substrate is likely the first to be transported across the outer membrane during secretion, rather than being the last as has been proposed (Mazar & Cotter, 2006), and could serve as the folding catalyst which drives secretion.

Guérin et al. have provided experimental evidence to convincingly demonstrate that removal of the H1 helix plug of FhaC is required for FHA secretion (Guerin et al., 2014). Precisely how far the H1 helix must be ejected from the barrel domain remains to be determined, but as shown here, it seems likely that it could assume a stable conformation in close proximity to the periplasmic face of the barrel domain by interacting directly with the POTRA domains, particularly POTRA 2. This would position the H1 helix at an ideal location to quickly reinsert and plug the pore once secretion is complete. Now that it seems clear the H1 helix plug must be removed for secretion, other mechanistic questions can be addressed. For example, does the barrel domain of FhaC truly serve at the secretion pore and if so, can substrate be trapped inside the barrel domain? Concerning the secretion mechanism of FHA, does the TPS domain truly exit first or remain anchored in the periplasm until secretion is complete? The study presented here addresses the long standing question H1 helix movement and will serve as a springboard to decipher remaining details on the secretion of FHA by FhaC, the model TPS system for Type V secretion.

Figure 1. Conformational dynamics of the H1 helix plug of FhaC during FHA secretion.

FhaC (green) is an outer membrane (OM) transporter from Bordetella pertussis belonging to the TPS pathway with its primary role being the secretion of the virulence factor filamentous hæmagglutinin (FHA) (purple). Shown here is a depiction of the recent work by Guérin et al. where they show for the first time experimentally that the H1 helix plug (H1, red) is indeed removed upon interaction with FHA. Removal of the plug then allows for secretion of FHA presumably through the barrel domain of FhaC. Once secretion is complete, the H1 helix again plugs the pore preventing free diffusion across the outer membrane. The linker, which connects the H1 helix to POTRA 1 (P1), is indicated by ‘L1’ and POTRA 2 by ‘P2’.