Mechanism of secretion of TcpF by the Vibrio cholerae toxin-coregulated pilus
Edited by Scott Hultgren, Washington University in St. Louis School of Medicine, St. Louis, MO; received July 25, 2022; accepted February 28, 2023
Significance
Type II secretion (T2S) and Type IV pili (T4P) are widespread bacterial virulence systems that are structurally related. While T2S systems function exclusively to export exoproteins, including toxins and degradative enzymes, from the periplasm to the extracellular space, T4P are multifunctional, with only a subset sharing the secretion function. Substrate recognition in both systems is poorly understood. Here, we show that the disordered N-terminal segment of the Vibrio cholerae exoprotein TcpF is the export signal, which binds to the minor pilin cap at the tip of the T4P for delivery across the outer membrane. Our results have implications for understanding T2S and provide strategies to block this process.
Abstract
Many bacteria possess dynamic filaments called Type IV pili (T4P) that perform diverse functions in colonization and dissemination, including host cell adhesion, DNA uptake, and secretion of protein substrates—exoproteins—from the periplasm to the extracellular space. The Vibrio cholerae toxin-coregulated pilus (TCP) and the enterotoxigenic Escherichia coli CFA/III pilus each mediates export of a single exoprotein, TcpF and CofJ, respectively. Here, we show that the disordered N-terminal segment of mature TcpF is the export signal (ES) recognized by TCP. Deletion of the ES disrupts secretion and causes TcpF to accumulate in the V. cholerae periplasm. The ES alone can mediate export of Neisseria gonorrhoeae FbpA by V. cholerae in a T4P-dependent manner. The ES is specific for its autologous T4P machinery as CofJ bearing the TcpF ES is exported by V. cholerae, whereas TcpF bearing the CofJ ES is not. Specificity is mediated by binding of the ES to TcpB, a minor pilin that primes pilus assembly and forms a trimer at the pilus tip. Finally, the ES is proteolyzed from the mature TcpF protein upon secretion. Together, these results provide a mechanism for delivery of TcpF across the outer membrane and release into the extracellular space.
Sign up for PNAS alerts.
Get alerts for new articles, or get an alert when an article is cited.
Gram-negative bacteria employ an array of complex envelope-spanning machines to export protein substrates from the cytoplasm to the extracellular milieu (1). Secretion can occur in a single step, as seen for the Type III and IV systems that transport substrates directly from the bacterial to the host cell cytoplasm. Other substrates are first translocated across the inner membrane via the Sec or Tat translocons and then across the outer membrane via one of several secretion systems that vary in their components and mechanism (2). Many newly synthesized preproteins are targeted to the Sec translocon via their N-terminal signal peptide (SP), which binds to the cytoplasmic chaperone SecB and is removed by a signal peptidase during translocation across the inner membrane, releasing the mature protein in the periplasm (3). A subset of these mature proteins is exported from the cell in a folded form by the Type II secretion (T2S) system comprising an inner membrane–anchored “endopilus” (also called a pseudopilus) that extrudes periplasmic substrates through an outer membrane secretin channel (4, 5). The endopilus is built from a reservoir of inner membrane major pilin subunits using energy from a cytoplasmic ATPase (6, 7). T2S system substrates include hydrolytic enzymes and toxins, which are critical for colonization, invasion, and dissemination for a number of bacterial pathogens (6, 8, 9). T2S is thought to occur via a piston- or screw-like motion of the endopilus, but the mechanism by which substrates are recognized by the secretion machinery is not well defined.
The T2S system is closely related to Type IV pilus (T4P) systems, some of which are also capable of delivering protein substrates, including proteases and colonization factors, across the outer membrane (10). Like the T2S endopili, T4P are polymers of the major pilin protein that originate in the inner membrane. But rather than being relegated to the periplasm, T4P pass through the secretin channel for display on the bacterial surface. T4P-mediated secretion occurs during pilus extension, driven by a cytoplasmic assembly ATPase, and surface-displayed pili can adhere to host receptors, abiotic surfaces, and substrates such as DNA and bacteriophage (11). T4P are retractile, a process powered by a retraction ATPase motor in some systems and by motor-independent retraction in others (12–17). Retraction can pull the cells along surfaces in a twitching motion and can draw bound substrates into the bacterium (11, 18). Pilus filaments are highly dynamic, with retraction and extension rates of ~1,000 major pilin subunits per second reported for Neisseria gonorrhoeae T4P (19).
Among the simplest of the T4P systems are those classified as Type IVb pili (T4bP), which include the Vibrio cholerae toxin-coregulated pilus (TCP) and the enterotoxigenic Escherichia coli (ETEC) CFA/III pilus. These T4bP systems each secretes a single substrate or “exoprotein”, TcpF, and CofJ, respectively, which is encoded within the T4P operon along with the pilus assembly genes. The functions of these exoproteins are not known, but deletion of tcpF leads to a 5-log decrease in V. cholerae colonization of the infant mouse (20), and CofJ has been implicated in host cell adhesion (21, 22). Some of the more complex Type IVa pili (T4aP) also function as secretion systems. The Francisella tularensis T4aP exports chitinases, a chitin-binding protein, a protease, a β-glucosidase (23), and the T4aP from the sheep pathogen Dichelobacter nodosus exports several proteases (24).
In addition to varying in the number and type of substrates secreted, there are key differences in the export machineries of T4aP, T4bP, and T2S systems that have implications for secretion. First, whereas the T4aP possess a retraction ATPase that would facilitate a piston-like motion to efficiently and processively export substrates, this counter motor is absent in the T4bP and T2S systems. Nonetheless, retraction has been demonstrated for V. cholerae TCP and other T4P lacking a retraction ATPase and is in fact necessary for their functions, including secretion (15, 17, 25). Second, these systems differ in the low-abundance minor pilins that nucleate T4P polymerization and presumably cap the growing pilus (26). The V. cholerae TCP and ETEC CFA/III T4bP utilize a single minor pilin, TcpB and CofB, respectively, which forms a homotrimeric priming complex at the pilus tip (15, 27–29). In contrast, the T2S systems and the T4aP have four minor pilins that form a heteromultimeric priming complex (30–33). Additionally, many T4aP systems have a nonpilin protein called PilY1 or PilC that forms part of the priming complex and is a putative tip protein (31, 32, 34). These pilus caps may interact directly with exoproteins as they are extruded through the secretin channel and may also be critical for regulating passage of the pilus through this channel.
The V. cholerae TcpB trimer is located at the pilus tip and is the receptor for the cholera toxin phage CTXφ (27). The ETEC minor pilin CofB bears strong structural homology to TcpB and also crystallizes as a trimer (28, 29). Thus, CofB is also expected to form a trimeric cap on the CFA/III pilus. Oki and colleagues revealed a direct interaction between the CofB trimer and the N-terminal segment of the ETEC exoprotein CofJ: Purified CofJ was pulled down by His-tagged CofB using nickel Sepharose beads, and a crystal structure was determined for the CofB trimer in complex with three copies of an N-terminal CofJ peptide (21). The peptide, which corresponds to residues 1 to 24 of mature CofJ, binds to the groove between two C-terminal domains of the CofB trimer in a 1:1 stoichiometry. Only residues 5 to 15 are resolved in this structure. The central aromatic Phe10 side chain is inserted deep into the domain interface. Substitution of Phe10 with Ala abolishes binding of the N-terminal peptide to CofB. These data suggested that ETEC CFA/III exports its substrate by binding to it via its tip-associated minor pilin. Consistent with these findings, a segment corresponding to residues 6 to 29 of mature TcpF was shown to be essential for export of this exoprotein by the V. cholerae T4bP (35), and deletion of residues 5 to 8 disrupts export and leads to accumulation of TcpF in the periplasm (36). Like Phe10 on CofJ, an aromatic residue on TcpF is involved as variants with a conserved Tyr5Phe substitution were exported from V. cholerae, but Tyr5Ser and Tyr5Arg variants were not. These results support a mechanism of secretion whereby the exoprotein N-terminal segment is bound by the minor pilin trimer at the pilus tip for delivery through the secretin channel.
TcpF and CofJ are similar in size (318 and 326 residues, respectively) and dimensions and are encoded syntenically on their respective T4P operons, yet they share neither sequence nor structural homology. TcpF is an elongated bilobed protein with an α-helical N-terminal domain and a β-sandwich C-terminal domain connected via a short linker (37). CofJ is also elongated but has a single domain comprising a large β-sandwich decorated by short α-helices (22). Yet both TcpF and CofJ possess an ~24-residue N-terminal segment that is not resolved in the crystal structures and is proteolytically labile, suggesting flexibility. In both proteins, this flexible segment is followed by a disulfide bond that secures the flanking N-terminal region to the globular body of the mature exoprotein. The TcpF disulfide bond between Cys34 and Cys47 links the first two α-helices in the structured portion of the protein. The CofJ disulfide bond between Cys26 and Cys323 links the N terminus to the C terminus of the protein. The disulfide bonds stabilize the fold of the exoprotein body leaving a flexible N-terminal “handle” to bind to the T4P machinery for delivery across the outer membrane.
Here, we examine T4P-mediated secretion of V. cholerae TcpF across the bacterial outer membrane. We show that the N-terminal segment of TcpF is necessary and sufficient for export of this exoprotein and that it interacts specifically with the tip-associated minor pilin TcpB. Our data suggest that TcpF is transiently tethered via this export signal (ES) to the TcpB trimer at the tip of the pilus for transport through the secretin channel. These results provide a mechanistic understanding of T4bP-mediated secretion with implications for understanding secretion in the more complex T4aP and T2S systems.
Results
The N-Terminal Segment of Mature TcpF Is Necessary and Sufficient for Export into the Extracellular Space.
Both V. cholerae TcpF and ETEC CofJ exoproteins are expressed as preproteins with an N-terminal Sec-dependent signal peptide (SP) for translocation across the inner membrane into the periplasm (Fig. 1 A and B) (38). SP is removed from the preprotein during translocation, leaving the mature protein. To examine the role of the disordered N-terminal segment of mature TcpF as a putative T4P-dependent export signal (ES, residues 1 to 24, Fig. 1 A and B), we deleted this segment and expressed TcpF-ΔES (Fig. 1C) ectopically in a V. cholerae strain lacking tcpF. The SP was retained to ensure that TcpF was translocated to the periplasm. V. cholerae O395 (wild type, wt) cells were grown overnight in pilus-inducing conditions along with a ΔtcpF strain expressing either wt TcpF or TcpF-ΔES. Bacterial cells were separated from the culture supernatant by centrifugation. Whole cells were then treated with polymyxin B to release the periplasmic fraction, which was separated by centrifugation from the spheroplast fraction containing the cytoplasm and membranes. Cell fractions were analyzed by SDS-PAGE and immunoblotting to assess TcpF export (Fig. 1D), and the distribution of TcpF within the fractions was quantified by densitometry and plotted as a fraction of the total TcpF (SI Appendix, Fig. S1). The majority of both endogenously and ectopically expressed TcpF is found in the culture supernatant as a doublet representing full-length protein (36 kDa) and an ~34 kDa proteolyzed form. Of the cellular TcpF, the majority is periplasmic, indicating efficient secretion via the Sec machinery. TcpF lacking the ES (TcpF-ΔES) is retained in the whole-cell fraction, primarily in the periplasm, demonstrating that the N-terminal segment of mature TcpF is required for its export from the periplasm into the culture supernatant.
Fig. 1.
To further test whether the ES is sufficient for T4P-mediated export in V. cholerae, the TcpF SP and ES were fused to the N terminus of an unrelated periplasmic protein from N. gonorrhoeae, FbpA, in place of its own SP (Fig. 2A). A hexahistidine tag was appended to the C terminus of FbpA for detection with an anti-His antibody (mature protein, FES-FbpA-His). As controls, the SP and ES of CofJ (JES-FbpA-His) as well as the TcpF SP alone (FbpA-His) were also tested for their ability to mediate export. All constructs were expressed in V. cholerae ΔtcpF. The majority of FES-FbpA-His is exported into the culture supernatant as a doublet representing full length mature protein and a proteolytic fragment (Fig. 2B and SI Appendix, Fig. S2). Proteolysis occurs at the N-terminal ES end of the protein as the anti-His antibody binds to a C-terminal His tag. The majority of JES-FbpA-His and FbpA-His are found in the whole-cell fraction comparable to the localization of FES-FbpA-His in a non-piliated mutant, V. cholerae ΔtcpA. Together, these data demonstrate that the TcpF ES alone drives TCP-mediated export across the V. cholerae outer membrane.
Fig. 2.
Mass spectrometry was used to identify the proteolytic cleavage site in the ES of TcpF and FbpA-His. Proteins were purified from the V. cholerae periplasm and incubated at room temperature to allow proteolysis and then analyzed by SDS-PAGE to visualize the extent of proteolysis and by liquid chromatography/MALDI-TOF mass spectrometry to identify the products. Full-length mature TcpF (35,843 Da) is converted to a product with a mass of 34,292 Da, which lacks residues 1 to 14 of the ES (TcpFΔ1-14, SI Appendix, Fig. S3A). Full-length mature FES-FbpA-His (36,862 Da) is proteolyzed at the same location, between residues 14 and 15 of the TcpF ES, resulting in a 35,312-Da product (SI Appendix, Fig. S3B). Thus, the TcpF ES is preferentially cleaved between residues Thr14 and Ser15 which may contribute to release of the exoprotein from the pilus.
The V. cholerae Type IV Pilus Specifically Recognizes the TcpF ES.
The disordered N-terminal segments of both TcpF and CofJ (ES) are similar in length, are rich in serines, and rely on an aromatic residue for binding to the T4P (Fig. 1A): TcpF Tyr5 is necessary for secretion (36), and CofJ Phe10 is necessary for CofB binding (21). To investigate whether the V. cholerae T4P system specifically recognizes its cognate SP, both TcpF and CofJ were expressed in V. cholerae ΔtcpF with their endogenous SP/ES or with their SP/ES segments swapped (Fig. 3A), and expression and localization of these hybrid substrates were assessed (Fig. 3B and SI Appendix, Fig. S4). Approximately 50% of wt TcpF is exported from V. cholerae ΔtcpF, whereas the majority (80%) of TcpF bearing the CofJ SP/ES (TcpF-JSP/ES) is retained in the cell fraction. The export defect is not due to an inability of the V. cholerae Sec translocon to recognize the CofJ SP as this hybrid exoprotein is located primarily in the periplasm rather than the spheroplast. WT CofJ is expressed in V. cholerae ΔtcpF but remains primarily in the periplasm, whereas ~85% of CofJ containing TcpF SP/ES (CofJ-FSP/ES) is exported into the culture supernatant. These results demonstrate that the V. cholerae T4P machinery is specific for the TcpF ES.
Fig. 3.
Residue Tyr5 but not Tyr12 Is Critical for TcpF Export.
Having shown that the interaction between the exoprotein ES and the T4P is specific, we sought to delineate residues in the TcpF ES that are important for binding. Megli and Taylor showed that a deletion of residues 1 to 5 disrupts TcpF export, as does a Tyr5Ala substitution alone (36), whereas individual alanine substitutions at residues 6 to 8 have no effect on export. Building on this analysis, we substituted TcpF residues 1, 2, 3, 4, and 10 individually with alanine in V. cholerae, and mutants were tested for TcpF expression and export. We also made alanine and phenylalanine substitutions in Tyr5 to confirm the results of Megli and Taylor and in Tyr12 to test whether this aromatic residue is important for binding as Phe10 is for CofJ (21). Of these residues, Phe1 contributes modestly to export, with less TcpF observed in the supernatant than in the whole-cell fraction for the F1A variant compared to wt TcpF (Fig. 4 and SI Appendix, Fig. S5). Alanine-substituted variants for residues 2 to 4 (N2A, D3A, and N4A) and T10A are exported at wt levels. As shown previously, Tyr5 (Y5A) is critical for export, with the majority of this variant remaining in the cells. A Y5F substitution had no effect on secretion, consistent with the aromatic ring and not the hydroxyl group being the key interaction moiety. Both Y12F and Y12A are secreted at wild-type levels, indicating that Tyr12 is not critical to the TcpF–pilus interaction. Thus, while both TcpF and CofJ bind to the T4P via their N-terminal residues, with aromatic residues mediating these interactions, TcpF uses its terminal amino acids, whereas CofJ binds via residues 5 to 15.
Fig. 4.
TcpF Binds via Its N-Terminal ES to the Minor Pilin TcpB In Vitro and In Vivo.
We next sought to identify the component of the V. cholerae TCP that recognizes the TcpF ES. The N-terminal segment of ETEC CofJ was shown previously to bind to the C-terminal domain of the trimeric minor pilin CofB (21). The V. cholerae minor pilin TcpB is highly similar in structure to CofB, crystallizes as a homotrimer, and localizes as a trimer at the pilus tip (27). Thus, we hypothesized that periplasmic TcpF binds via its ES to the TcpB trimer at the pilus tip for delivery through the secretin channel as the pilus extends. To test this hypothesis, we incubated purified forms of TcpF and TcpB, one with a His tag and one without, and examined the ability of tagged bait protein to pull down the nontagged prey partner protein using Ni-NTA beads. Nontagged and C-terminally His-tagged TcpF was expressed in V. cholerae ΔtcpF and purified from the periplasm (37). Two soluble forms of TcpB were used, His-TcpB-C, which represents the C-terminal domain (residues 243 to 423) and forms a trimer in solution and in the crystal structure, and a larger construct, ΔN-TcpB, which lacks only the N-terminal a1N segment (residues 1 to 24) and also forms a trimer in solution (27). Where necessary, the N-terminal His tags were removed from the recombinant TcpB proteins by thrombin digestion. Tagged bait and nontagged prey proteins were mixed and added to nickel-NTA beads, and bound proteins were eluted with imidazole. Fractions were analyzed by SDS-PAGE and Coomassie staining. His-tagged TcpF-His pulls down nontagged TcpB-C (Fig. 5A), and in the reciprocal assay, His-tagged TcpB-C pulls down nontagged TcpF (Fig. 5B). In contrast, ETEC CofJ-His is not able to pull down TcpB-C (Fig. 5C). These results demonstrate a direct and specific interaction between TcpF and TcpB. The proteolytic product of His-tagged TcpF, TcpF(Δ1-14)-His binds to the Ni-NTA beads (Fig. 3A) as it retains the C-terminal His tag. However, this product is not pulled down by His-TcpB, consistent with the ES mediating binding of TcpF to TcpB.
Fig. 5.
TcpF-His is also able to pull down the larger version of TcpB, ΔN-TcpB, which contains both the C-terminal trimerization domain and the pilin domain (Fig. 5D). We used this TcpB construct to examine the ability of N. gonorrhoeae FbpA bearing the TcpF ES to bind to TcpB. FES-FbpA-His was purified from the V. cholerae periplasm, mixed with ΔN-TcpB, and captured with Ni-NTA beads. As predicted, His-tagged FES-FbpA pulls down ΔN-TcpB (Fig. 5E) via the TcpF ES.
We next sought evidence for a TcpF–TcpB interaction in the V. cholerae periplasm. Soluble His-tagged TcpB-C was expressed in the periplasm of V. cholerae by adding the TcpF SP such that it is translocated across the inner membrane via the Sec machinery. The ability of periplasmic His-TcpB-C to bind to TcpF was assessed by incubating the periplasmic fraction with Ni-NTA beads and looking for TcpF. TcpB, along with a small amount of TcpF, is observed in the fraction eluted from the Ni-NTA beads, whereas no TcpF is found when His-TcpB-C is absent (Fig. 6A). These results, which are interpreted in Fig. 6B, reinforce those of the in vitro pull-down assays demonstrating a direct interaction between TcpF and TcpB and imply that this interaction occurs in the periplasm, where TcpF would encounter the TcpB trimer at the tip of the growing pilus. Of note, TcpF export into the V. cholerae supernatant is only slightly reduced when His-TcpB is present in the periplasm, and only a small proportion of the TcpF is pulled down with His-TcpB-C, suggesting that binding to TcpB at the pilus tip is preferred to that of soluble recombinant TcpB.
Fig. 6.
Discussion
We show here that T4P-mediated export of the V. cholerae exoprotein depends on an interaction between a flexible, protease-sensitive N-terminal segment of TcpF and the pilus tip-associated minor pilin TcpB. This interaction delivers TcpF from the periplasm through the outer membrane secretin channel and into the extracellular milieu where it is released (Fig. 7A). Our results build on those of Oki et al., who reported a crystal structure of an N-terminal 24-residue peptide from ETEC CofJ exoprotein bound to the CofB minor pilin trimer (21) (Fig. 7B). In this complex, residues 5 to 15 of the CofJ peptide are resolved and are bound in an extended conformation in the cleft between each CofB monomer, burying the Phe10 side chain. The N terminus of the peptide is oriented toward the CofB pilin domains, which would sit at the tip of the pilus, placing the body of CofJ distal to the tip of the pilus. The TcpF ES length is comparable to that of CofJ, with a critical aromatic residue [Tyr5, (36)]. Proteolysis of the first 14 amino acids disrupts its ability to bind to TcpB, suggesting that it interacts with the TcpB trimer in a manner similar to that of CofJ with CofB, but the minor pilin binding site is shifted toward the N terminus for the TcpF ES and centered on Tyr5.
Fig. 7.
While this manuscript was in revision, Oki et al. published a remarkable crystal structure of the TcpB trimer in complex with three TcpF molecules (39). In this complex, the N-terminal 10 residues of each TcpF molecule bind in the cleft between the TcpB C-terminal domains, and the TcpF bodies extend away from the C-terminal tip of the TcpB trimer like the petals of a flower (Fig. 7C). The TcpF N-terminal domains (NTDs) form a symmetrical trimer, but instead of this symmetry matching that of the TcpB trimer, the TcpF NTD trimer is rotated such that its plane is almost parallel to the long axis of the TcpB and the pilus filament. This arrangement, made possible by the flexible linker between the TcpB binding site (residues 1 to 10) and the TcpF body (residues 28 to 318), causes each C-terminal domain of TcpF to splay in a different direction. The authors showed using isothermal titration calorimetry that a 10–amino acid TcpF peptide corresponding to the TcpB binding site binds only weakly to TcpB (Kd 506 μM), whereas a slightly longer peptide, 1 to 15, binds with a higher affinity (5.2 μM), and full-length TcpB binds at 0.11 μM. These results are consistent with the NTD trimer formed by TcpF in the TcpB:TcpF crystal structure and suggest that the TcpF trimer interface may contribute to complex formation. However, this trimeric interaction may not be significant in vivo. A single amino acid change in TcpF at Tyr5 disrupts export (shown here and in ref. 36), and we show here that the ES alone is sufficient to induce export of ETEC CofJ and N. gonorrhoeae FbpA from V. cholerae, demonstrating that trimerization is not required for export in the living organism. Furthermore, the splayed flower-like arrangement of TcpF, induced by trimerization of the NTDs, appears ill-suited for passage through the gated secretin channel. We propose that TcpF bound to the TcpB trimer via the ES folds down toward the TcpB trimer to form a compact pilus tip that can readily pass through the secretin channel (Fig. 7A).
We examined the TcpB–TcpF interaction in the crystal structure (39) in the context of the mutagenesis results reported here and by Megli and Taylor (36). Residues 1 to 5 of TcpF form a tight hook, with the side chains of Phe1 and Tyr5 buried in the cleft between the TcpB monomers in an offset stack. Although this region of the cleft is not particularly hydrophobic, the Phe1 phenyl ring is within 4 Å of TcpB Phe345 and the acyl chain of Lys351 (Fig. 7D). The Tyr5 phenol is sandwiched between the Phe1 phenyl ring of TcpF and TcpB residues Leu406 on one subunit and the His366 imidazole ring on another. The Tyr5 hydroxyl forms a hydrogen bond with a Ser370 side chain on TcpB. Given the burial of both Phe1 and Tyr, it is surprising that the Phe1Ala substitution has such a modest effect on TcpF export. Burial of the Phe1 phenyl ring may shield it from solvent without it being necessary for binding. The side chains of residues 2 to 4 and 6 are solvent-exposed, yet Asp3 forms a salt bridge with Arg408 of TcpB. The hydroxyl groups of Ser 7, Thr8, and Thr10 hydrogen bond with TcpB oxygens in the cleft. Substitution of these residues does not impact secretion, suggesting that none of these interactions is critical to TcpF binding to TcpB. Hydrogen bonds also occur between backbone atoms of Tyr5, Ser6, Ser7, Thr8, and TcpB side chains on either side of the cleft, which may explain why single and even triple amino acid changes in this short binding segment do not impact TcpF export. Interestingly, Tyr12 of one TcpF monomer forms a hydrogen bond with Ser393 of TcpB but is oriented toward the body of TcpF in the second TcpF monomer and is unresolved in the third, consistent with this residue not being critical for TcpF binding and export. Thus, binding of the TcpF ES to TcpB is mediated by burial of Tyr5 together with backbone hydrogen bonds.
Both TcpF and CofJ form stable complexes with their respective minor pilins in vitro, yet these interactions must be transient so that exoproteins are released from the extending pilus, allowing it to retract into the periplasm to pick up more exoprotein “cargo”. Release would also be necessary for V. cholerae TCP to bind to the cholera toxin phage CTXφ, which uses TcpB as its receptor (27). Release may be facilitated by a relatively weak interaction between TcpF and TcpB, which is more likely to occur in the periplasm where TcpF is concentrated than in the lumen of the small intestine where TcpF would be present in low concentrations. TcpF release in the extracellular space may also be facilitated by proteolysis, which would free the body of the exoprotein from the ES attached to the pilus. We identified a preferred protease cleavage site between Thr14 and Ser15 of the ES for both TcpF and FbpA bearing the TcpF ES. Proteolysis is also observed for CofJ (Fig. 4B), as was noted previously (21). However, only a portion of the soluble exoprotein in the supernatant is proteolyzed, implying that proteolysis is not essential for release. Furthermore, proteolysis would potentially leave the N terminus of the ES attached to the pilus tip, which would prevent subsequent loading of new exoprotein cargo. Proteolysis may simply be the consequence of an exposed and flexible N terminus on the newly released exoprotein.
Whereas a flexible N-terminal segment is sufficient to signal export of TcpF and CofJ, the secretion signals of T4aP and T2S exoproteins are more complex. The T4bP systems of V. cholerae and ETEC are unique in having threefold symmetry at the tip that provides three identical minor pilin binding sites. Furthermore, they appear to export a single exoprotein. In contrast, the T4aP and T2S pilus tips are formed by four distinct minor pilins, plus PilY1 for some T4aP, which must recognize multiple diverse exoprotein substrates (8). Common features allowing specific secretion of exoproteins via these systems have not been identified. It has been suggested that the recognition motif of substrates exported by the T2S system is discontinuous, formed by the folded exoprotein (5, 6). Indeed, secretion determinants of pullulanase, the substrate for the Klebsiella oxytoca T2S system, map to multiple sites on the protein (40, 41). Additionally, interaction with the bacterial inner membrane is required for pullulanase to be recognized and transported by the T2S system (41). The secretin channel itself may also be involved in exoprotein recognition. The exoprotein cholera toxin is exported by the V. cholerae T2S. This AB5 toxin was shown by surface plasmon resonance studies to interact via its B subunit with the periplasmic domain of the secretin subunit GspD (42), and electron microcopy analysis of negative stained preparations of the V. cholerae secretin channel revealed density for the B subunit pentamer within the vestibule of the channel, consistent with the holotoxin homing to this vestibule to facilitate its export. We observed export of CofJ-FSP/ES from V. cholerae, suggesting that recognition of the exoprotein by the cognate secretin channel is not a prerequisite for secretion in this system. Further work is needed to understand the mechanism of substrate recognition in the more complex T2S and T4aP systems and to determine the feasibility of targeting these systems as antimicrobial strategies.
Materials and Methods
Detailed descriptions of the methods used in this study as well as tables listing the bacterial strains, plasmids, primers, and antibodies are provided in SI Appendix and summarized below.
Construction of Vectors for Exoprotein Expression.
Plasmids encoding TcpF, CofJ, TcpB, and FbpA constructs were prepared by cloning the respective genes from genomic DNA from V. cholerae O395, ETEC 31-10, and N. gonorrhoeae MS11 into plasmid pJMA10.1 and subcloning (SI Appendix, Table S1).
Analysis of Exoprotein Secretion and Export.
Cells were grown overnight and separated from the supernatant containing secreted protein by centrifugation. Periplasmic fractions were separated from the spheroplasmic fractions by treating the cell pellets with polymyxin B followed by centrifugation. Fractions were analyzed by SDS-PAGE and immunoblotting and protein levels were quantified by densitometry.
Expression and Purification of Soluble Proteins.
Proteins used for mass spectrometry and pull-down assays were expressed in the E. coli expression strain Shuffle T7 Express lysY and purified from the cytoplasm by affinity chromatography using a Ni-NTA column. TcpF was isolated from the V. cholerae periplasm by ammonium sulfate precipitation and purified by size exclusion chromatography. TcpF-His, CofJ-His, and FES-FbpA-His were also isolated from the V. cholerae periplasm and purified using a Ni-NTA column.
Mass Spectrometry to Identify TcpF, FbpA Proteolytic Products.
TcpF and FES-FbpA-His proteins and proteolytic fragments were separated on a Zorbax 300SB-C8 HPLC column (Agilent Technologies) and eluted with a gradient of acetonitrile (10 to 60% W/V with 0.1% formic acid) and analyzed by liquid chromatography–mass spectrometry (LC-MS) in the ESI mode using a Bruker maXis Impact Quadrupole Time-of-Flight LC/MS System consisting of an Agilent 1200 HPLC and a Bruker maXis Impact Ultra-High Resolution tandem TOF (UHR-Qq-TOF) mass spectrometer.
In Vitro Pull-Down Assays.
Proteins were incubated in a 1:1 molar ratio and added to Ni-NTA agarose beads. Unbound protein was washed from the beads, and bound protein was eluted using 300 mM imidazole. Fractions were analyzed by SDS-PAGE and Coomassie blue staining.
Pull Down of Periplasmic TcpF by Soluble Periplasmic TcpB.
His-TcpB was expressed in a soluble form in the periplasm of V. cholerae. Cells were grown overnight, and the periplasmic fraction was isolated and incubated with Ni-NTA agarose beads to capture His-TcpB. Unbound protein was washed from the beads, and bound protein was eluted using 300 mM imidazole. Fractions were analyzed by SDS-PAGE and immunoblotting.
Data, Materials, and Software Availability
All study data are included in the article and/or SI Appendix.
Acknowledgments
We thank Hongwen Chen and the Simon Fraser University Mass Spectrometry Facility for analysis of TcpF and FbpA samples and Dixon Ng for valuable discussions. Tony Harn generated the ptcpFSP-His-tcpB-C plasmid. L.C. was supported by a Discovery Grant from the Natural Sciences and Engineering Research Council (RGPIN-2017-05757).
Author contributions
T.-H.W. and L.C. designed research; M.N., T.-H.W., K.J.D., N.M.K., and J.Z.Z. performed research; M.N., T.-H.W., K.J.D., N.M.K., J.Z.Z., and L.C. analyzed data; and L.C. wrote the paper.
Competing interests
The authors declare no competing interest.
Supporting Information
Appendix 01 (PDF)
- Download
- 1.04 MB
References
1
T. R. D. Costa et al., Secretion systems in gram-negative bacteria: Structural and mechanistic insights. Nat. Rev. Microbiol. 13, 343–359 (2015).
2
P. Natale, T. Bruser, A. J. Driessen, Sec- and Tat-mediated protein secretion across the bacterial cytoplasmic membrane–distinct translocases and mechanisms. Biochim. Biophys. Acta 1778, 1735–1756 (2008).
3
A. J. Driessen, N. Nouwen, Protein translocation across the bacterial cytoplasmic membrane. Ann. Rev. Biochem. 77, 643–667 (2008).
4
J. L. Thomassin, J. S. Moreno, I. Guilvout, G. Tran Van Nhieu, O. Francetic, The trans-envelope architecture and function of the type 2 secretion system: New insights raising new questions. Mol. Microbiol. 105, 211–226 (2017).
5
M. Nivaskumar, O. Francetic, Type II secretion system: A magic beanstalk or a protein escalator. Biochim. Biophys. Acta 1843, 1568–1577 (2014).
6
K. V. Korotkov, M. Sandkvist, Architecture, function, and substrates of the type II secretion system. EcoSal Plus 8 (2019), https://doi.org/10.1128/ecosalplus.ESP-0034-2018.
7
C. Lu et al., Hexamers of the type II secretion ATPase GspE from Vibrio cholerae with increased ATPase activity. Structure 21, 1707–1717 (2013).
8
N. P. Cianciotto, R. C. White, Expanding role of type II secretion in bacterial pathogenesis and beyond. Infect. Immun. 85, e00014-17 (2017).
9
A. E. Sikora, R. A. Zielke, D. A. Lawrence, P. C. Andrews, M. Sandkvist, Proteomic analysis of the Vibrio cholerae type II secretome reveals new proteins, including three related serine proteases. J. Biol. Chem. 286, 16555–16566 (2011).
10
D. Nunn, Bacterial type II protein export and pilus biogenesis: More than just homologies? Trends Cell Biol. 9, 402–408 (1999).
11
L. Craig, K. T. Forest, B. Maier, Type IV pili: Dynamics, biophysics and functional consequences. Nat. Rev. Microbiol. 17, 429–440 (2019).
12
M. Clausen, M. Koomey, B. Maier, Dynamics of type IV pili is controlled by switching between multiple states. Biophys. J. 96, 1169–1177 (2009).
13
A. J. Merz, M. So, M. P. Sheetz, Pilus retraction powers bacterial twitching motility. Nature 407, 98–102 (2000).
14
J. M. Skerker, H. C. Berg, Direct observation of extension and retraction of type IV pili. Proc. Natl. Acad. Sci. U.S.A. 98, 6901–6904 (2001).
15
D. Ng et al., The Vibrio cholerae minor pilin TcpB initiates assembly and retraction of the toxin-coregulated pilus. PLoS Pathog. 12, e1006109 (2016).
16
C. K. Ellison et al., Retraction of DNA-bound type IV competence pili initiates DNA uptake during natural transformation in Vibrio cholerae. Nat. Microbiol. 3, 773–780 (2018).
17
J. L. Chlebek, R. Denise, L. Craig, A. B. Dalia, Motor-independent retraction of type IV pili is governed by an inherent property of the pilus filament. Proc. Natl. Acad. Sci. U.S.A. 118, e2102780118 (2021).
18
L. L. Burrows, Weapons of mass retraction. Mol. Microbiol. 57, 878–888 (2005).
19
S. Kraus-Romer, I. Wielert, I. Rathmann, J. Grossbach, B. Maier, External stresses affect gonococcal type 4 pilus dynamics. Front. Microbiol. 13, 839711 (2022).
20
T. J. Kirn, N. Bose, R. K. Taylor, Secretion of a soluble colonization factor by the TCP type 4 pilus biogenesis pathway in Vibrio cholerae. Mol. Microbiol. 49, 81–92 (2003).
21
H. Oki et al., Interplay of a secreted protein with type IVb pilus for efficient enterotoxigenic Escherichia coli colonization. Proc. Natl. Acad. Sci. U.S.A. 115, 7422–7427 (2018).
22
A. S. Yuen, S. Kolappan, D. Ng, L. Craig, Structure and secretion of CofJ, a putative colonization factor of enterotoxigenic Escherichia coli. Mol. Microbiol. 90, 898–918 (2013).
23
A. J. Hager et al., Type IV pili-mediated secretion modulates Francisella virulence. Mol. Microbiol. 62, 227–237 (2006).
24
X. Han, R. M. Kennan, D. Parker, J. K. Davies, J. I. Rood, Type IV fimbrial biogenesis is required for protease secretion and natural transformation in Dichelobacter nodosus. J. Bacteriol. 189, 5022–5033 (2007).
25
C. K. Ellison et al., Obstruction of pilus retraction stimulates bacterial surface sensing. Science 358, 535–538 (2017).
26
T. Jacobsen, B. Bardiaux, O. Francetic, N. Izadi-Pruneyre, M. Nilges, Structure and function of minor pilins of type IV pili. Med. Microbiol. Immunol. 209, 301–308 (2020).
27
M. Gutierrez-Rodarte, S. Kolappan, B. A. Burrell, L. Craig, The Vibrio cholerae minor pilin TcpB mediates uptake of the cholera toxin phage CTXphi. J. Biol. Chem. 294, 15698–15710 (2019).
28
S. Kolappan, D. Ng, G. Yang, T. Harn, L. Craig, Crystal structure of the minor pilin CofB, the initiator of CFA/III pilus assembly in enterotoxigenic Escherichia coli. J. Biol. Chem. 290, 25805–25818 (2015).
29
K. Kawahara et al., Homo-trimeric structure of the type IVb minor pilin CofB suggests mechanism of CFA/III pilus assembly in human enterotoxigenic Escherichia coli. J. Mol. Biol. 428, 1209–1226 (2016).
30
D. A. Cisneros, P. J. Bond, A. P. Pugsley, M. Campos, O. Francetic, Minor pseudopilin self-assembly primes type II secretion pseudopilus elongation. EMBO J. 31, 1041–1053 (2012).
31
Y. Nguyen et al., Pseudomonas aeruginosa minor pilins prime type IVa pilus assembly and promote surface display of the PilY1 adhesin. J. Biol. Chem. 290, 601–611 (2015).
32
A. Treuner-Lange et al., PilY1 and minor pilins form a complex priming the type IVa pilus in Myxococcus xanthus. Nat. Commun. 11, 5054 (2020).
33
K. V. Korotkov, W. G. Hol, Structure of the GspK-GspI-GspJ complex from the enterotoxigenic Escherichia coli type 2 secretion system. Nat. Struct. Mol. Biol. 15, 462–468 (2008).
34
T. Rudel, I. Scheurerpflug, T. F. Meyer, Neisseria PilC protein identified as type-4 pilus tip-located adhesin. Nature 373, 357–359 (1995).
35
S. J. Krebs, T. J. Kirn, R. K. Taylor, Genetic mapping of secretion and functional determinants of the Vibrio cholerae TcpF colonization factor. J. Bacteriol. 191, 3665–3676 (2009).
36
C. Megli, R. K. Taylor, Secretion of TcpF by the Vibrio cholerae toxin-coregulated pilus biogenesis apparatus requires an N-terminal determinant. J. Bacteriol. 195, 2718 (2013).
37
C. J. Megli et al., Crystal structure of the Vibrio cholerae colonization factor TcpF and identification of a functional immunogenic site. J. Mol. Biol. 409, 146–158 (2011).
38
T. J. Kirn, R. K. Taylor, TcpF is a soluble colonization factor and protective antigen secreted by El Tor and classical O1 and O139 Vibrio cholerae serogroups. Infect. Immun. 73, 4461–4470 (2005).
39
H. Oki et al., Structural basis for the toxin-coregulated pilus-dependent secretion of Vibrio cholerae colonization factor. Sci. Adv. 8, eabo3013 (2022).
40
N. Sauvonnet, A. P. Pugsley, Identification of two regions of Klebsiella oxytoca pullulanase that together are capable of promoting beta-lactamase secretion by the general secretory pathway. Mol. Microbiol. 22, 1–7 (1996).
41
A. East et al., Structural basis of pullulanase membrane binding and secretion revealed by X-ray crystallography, molecular dynamics and biochemical Analysis. Structure 25, 92–104 (2016).
42
S. L. Reichow et al., The binding of cholera toxin to the periplasmic vestibule of the type II secretion channel. Channels (Austin) 5, 215–218 (2011).
Information & Authors
Information
Published in
Classifications
Copyright
Copyright © 2023 the Author(s). Published by PNAS. This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).
Data, Materials, and Software Availability
All study data are included in the article and/or SI Appendix.
Submission history
Received: July 25, 2022
Accepted: February 28, 2023
Published online: April 11, 2023
Published in issue: April 18, 2023
Keywords
Acknowledgments
We thank Hongwen Chen and the Simon Fraser University Mass Spectrometry Facility for analysis of TcpF and FbpA samples and Dixon Ng for valuable discussions. Tony Harn generated the ptcpFSP-His-tcpB-C plasmid. L.C. was supported by a Discovery Grant from the Natural Sciences and Engineering Research Council (RGPIN-2017-05757).
Author contributions
T.-H.W. and L.C. designed research; M.N., T.-H.W., K.J.D., N.M.K., and J.Z.Z. performed research; M.N., T.-H.W., K.J.D., N.M.K., J.Z.Z., and L.C. analyzed data; and L.C. wrote the paper.
Competing interests
The authors declare no competing interest.
Notes
This article is a PNAS Direct Submission.
Authors
Metrics & Citations
Metrics
Citation statements
Altmetrics
Citations
Cite this article
120 (16) e2212664120,
Export the article citation data by selecting a format from the list below and clicking Export.
Cited by
Loading...
View Options
View options
PDF format
Download this article as a PDF file
DOWNLOAD PDFLogin options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginRecommend to a librarian
Recommend PNAS to a LibrarianPurchase options
Purchase this article to access the full text.