Export of the Pseudopilin XcpT of the Pseudomonas aeruginosa Type II Secretion System via the Signal Recognition Particle-Sec Pathway

Abstract

Type IV pilins and pseudopilins are found in various prokaryotic envelope protein complexes, including type IV pili and type II secretion machineries of gram-negative bacteria, competence systems of gram-positive bacteria, and flagella and sugar-binding structures in members of the archaeal kingdom. The precursors of these proteins have highly conserved N termini, consisting of a short, positively charged leader peptide, which is cleaved off by a dedicated peptidase during maturation, and a hydrophobic stretch of approximately 20 amino acid residues. Which pathway is involved in the inner membrane translocation of these proteins is unknown. We used XcpT, the major pseudopilin from the type II secretion machinery of Pseudomonas aeruginosa, as a model to study this process. Transport of an XcpT-PhoA hybrid was shown to occur in the absence of other Xcp components in P. aeruginosa and in Escherichia coli. Experiments with conditional sec mutants and reporter-protein fusions showed that this transport process involves the cotranslational signal recognition particle targeting route and is dependent on a functional Sec translocon.

The type II secretion pathway is widely used by gram-negative bacteria to secrete proteins into the extracellular environment (18). Release of exoproteins via this pathway occurs in a two-step process, implicating a periplasmic intermediate stage in the secretion pathway. Type II secretion machineries each encompass 12 to 16 different components (18). Five components show sequence similarity to PilA, the structural subunit of type IV pili, and are therefore named pseudopilins. Their precursors possess a conserved N terminus, which contains a short, positively charged leader peptide followed by a highly hydrophobic domain of approximately 20 amino acid residues (20, 47). The leader peptide is removed during export by a specific prepilin peptidase (4, 36). These prepilin peptidases are polytopic inner membrane proteins that cleave off the leader peptide at the cytoplasmic side of the inner membrane and catalyze the methylation of the new N-terminal amino acid residue (usually phenylalanine) of the mature protein (32, 48).

Type IV pilin-like N-terminal sequences are found not only in type IV pili and in the pseudopilin components of type II secretion systems of gram-negative bacteria, but also in competence systems in gram-positive bacteria and in flagella and sugar-binding proteins of archaea (39). In all these systems, the presence of proteins with prepilin-like N termini coincides with the occurrence of accessory proteins, including a prepilin peptidase, an ATPase, and a multispanning transmembrane protein (39).

How pilins and pseudopilins are transported across the cytoplasmic membrane and recruited by their cognate machinery is unknown. Two possible pathways have been proposed: one via the highly conserved Sec translocon and the other via a dedicated machinery implicating the prepilin peptidase, the ATPase, and/or the multispanning transmembrane protein mentioned above. The Sec system is used generally for protein transport across the cytoplasmic membrane (13). Two targeting pathways, the signal recognition protein (SRP) and SecB pathways, intersect at the Sec translocon (28). SecB interacts with the mature portion of presecretory proteins, and besides targeting the precursor to the translocase, it prevents their folding. SRP binds cotranslationally to hydrophobic sequences in its substrates. The ribosome-nascent chain complex subsequently binds FtsY and is targeted to the Sec translocon. Inner membrane proteins typically depend on a functional SRP pathway, whereas periplasmic and outer membrane proteins predominantly use the SecB route (11). Proteins targeted to the Sec system carry signal sequences that are characterized by a positively charged N-terminal region followed by a 10- to 15-residue-long hydrophobic core and a more polar C terminus containing the signal-peptidase cleavage site (53). Passage of the precursor over the inner membrane is followed by cleavage of the signal sequence by the signal peptidase LepB at the periplasmic side of the inner membrane (10). The N termini of the pilin and pseudopilin [referred to together as (pseudo)pilin] precursors (prepilins) share characteristics with Sec signal sequences, and previous studies have indeed shown that the N-terminal sequences of prepilins from Pseudomonas aeruginosa and Neisseria gonorrhoeae function as export signals for alkaline phosphatase in Escherichia coli (14, 46). However, the N termini of the prepilins are also distinct from Sec signal sequences, since they lack the signal-peptidase cleavage site C terminal to the hydrophobic domain and are, as mentioned above, processed by a specific prepilin peptidase N terminal to the hydrophobic segment at the cytoplasmic side of the membrane. These differences and the fact that the pilin-like proteins are always found in concert with other proteins have led to the proposal that pilins and pseudopilins are exported from the cytoplasm via a dedicated transport route formed by these accessory proteins (9, 18, 35).

To address the question of how (pseudo)pilins are translocated across the inner membrane, we used the major pseudopilin XcpT of the P. aeruginosa type II secretion system as a model protein. The type II secretion machinery of this organism requires at least 12 components, XcpA and XcpP to Z (20). XcpA functions as the prepilin peptidase and is shared by the Xcp system, the type IV piliation (Pil) system, and the Hxc system, which forms a second type II secretion system dedicated to the export of low-molecular-weight alkaline phosphatases (3). We demonstrate that XcpT translocation can occur independently of the presence of other Xcp components in P. aeruginosa and in E. coli and that transport is dependent on a functional Sec apparatus. Furthermore, we show that translocation occurs cotranslationally via the SRP pathway.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Strains used in this study are listed in Table 1. The SRP mutant alleles ffs-69 (50) and ffh-87 (51) were introduced in MC1060 by generalized transduction with phage P1 as described previously (5). After the introduction of construct pJGA03, the transductants were screened for blue coloring on Luria-Bertani (LB) agar plates containing X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside). Out of six transductants, six and three were scored positive for ffs-69 and ffh-87, respectively. P. aeruginosa strains and E. coli strains DH5α, C9, JS7131, and MC1060 and the MC1060 derivatives were grown at 37°C and other E. coli strains at 30°C in a modified LB broth (52), unless otherwise noted. For plasmid maintenance, the following antibiotics were used: for E. coli, 50 μg/ml ampicillin, 25 μg/ml kanamycin, 15 μg/ml tetracycline, and 15 μg/ml gentamicin, and for P. aeruginosa, 40 μg/ml gentamicin.

TABLE 1.

Strains used in this study

Strain	Relevant characteristic(s)	Reference
E. coli
DH5α	thi-1 hsdR17 gyrA96 recA1 endA1 glnV44 relA1 φ80dlacZΔM15 phoA8 λ−	26
C9	Hfr Cav relA1 fhuA22 pitA10 spoT1 ompF627 phoR18 creB501	23
MC4100	F− ΔlacU169 araD139 rpsL thi relA	6
MM52	MC4100 secA(Ts)51	37
IQ86	Tn10 thiA Δlac araD rpsL rpsE relA	45
IQ85	IQ86 secY(Ts)24	45
HPT244	MC4100 Δara-714 ffs-69 zba-3054::Tn10	50
HPT406	MC4100 Δara-714 pheA3141::Tn10Kan ffh-87	51
MC1060	Δ(codB-lacI)3 galK16 galE15(GalS) λ− e14−mcrA relA1 rpsL150(Strr) spoT1 mcrB1 hsdR2	7
JS7131	MC1060 attB::R6Kori ΔyidC ParaBAD-yidC	43
MC1060 ffs-69	MC1060 ffs-69 zba-3054::Tn10	This study
MC1060 ffh-87	MC1060 pheA3141::Tn10Kan ffh-87	This study
P. aeruginosa
PAO25	PAO1 leu arg	25
PAO1ΔT	PAO1 ΔxcpT	3
DZQ40	PAO1ΔxcpP to xcpZ	2
PAOΔhxcΔpilACΔR	Nonpolar deletions of the entire hxc gene cluster, of pilA to part of pilC, and of the xcpR gene	16

Plasmids and DNA manipulations.

Plasmids used in this study are listed in Table 2. Recombinant DNA methods were performed essentially as described previously (42), using E. coli strain DH5α for routine cloning. Plasmids were introduced by the CaCl₂ procedure into E. coli (42) or by electroporation into E. coli and P. aeruginosa (17). PCRs were performed with the proofreading enzyme Pwo DNA polymerase (Roche), and PCR products were cloned into the HincII site of vector pBC18R or into pCRII-TOPO according to the manufacturer's protocol. The lacI gene was PCR amplified using plasmid pET16b as a template and primers PB7 (5′-CTCCTTGCATGCACC-3′) and PB8 (5′-CCCGCGCCCATGGGAAGGAGCTG-3′), thereby introducing an NcoI restriction site (underlined) downstream of the stop codon. The PCR product was cloned into pBC18R, which resulted in the construct pCR-LacI, and subsequently, the SphI-NcoI fragment of pCR-LacI was introduced into the pBBR1-MCS5 vector, resulting in pYRC. The phoA gene without the promoter and signal sequence-encoding part was PCR amplified from pPHO7 with primers PB1 (5′-GATCCCCGGGGATCCGACTCTTATACAC-3′) and PB2 (5′-CGAAAATTCACTGTCTAGAGCGGTTTTATTTC-3′). A BamHI site (underlined) was introduced via PB1 upstream of the fragment encoding signal sequence-less PhoA and an XbaI site (underlined) via PB2 downstream of the stop codon. The PCR product was cloned into pBC18R, which gave pCR-PhoA. Plasmid pYRC-A was constructed by introduction of the BamHI-XbaI fragment of pCR-PhoA into BamHI-XbaI-digested pYRC. Cosmid pAX24 was used as a template to amplify xcpT with the oligonucleotides JAXcpTfor02 (5′-CTTCCGATCCTTCGAATCAACCAACTCGTG-3′) and JAXcpTrev01 (5′-GCCCGCATGTCGGATCCGTTGTCCCAGTTG-3′), and the resulting product was cloned into pCRII-TOPO, resulting in pCR-XcpT1-2. Underlined in JAXcpTrev01 is the BamHI site that replaced the stop codon of xcpT and that allowed for the construction of a translational fusion between xcpT and phoA. The PstI-BamHI fragment of pCR-XcpT1-2 was cloned into pYRC-A, resulting in the construct pJGA01. Plasmid pMPM-K4Ω contains an optimized Shine-Dalgarno sequence upstream of an NcoI site. To clone xcpT in the NcoI site, the gene was PCR amplified with primers pJAXcpTfor03 (5′-CGTGGGGTAATCCCATGGATCAGAGCCGC-3′) and pJAXcpTrev01. Underlined is the NcoI site that replaces the original GTG start codon. Replacement of the start codon also led to the substitution of an aspartic acid for an asparagine at the second residue in XcpT. The PCR product was cloned into pCRII-TOPO, which resulted in pCR-XcpT1-3. The NcoI-BamHI fragment of pCR-XcpT1-3 and the BamHI-XbaI fragment of pYRC-A were ligated into NcoI-XbaI-digested pMPM-K4Ω, resulting in the construct pJGA07. To obtain pT7-T, xcpT as an 850-bp BssH2 fragment from pAX24 was first introduced into SmaI-digested pUC19, resulting in pUX4. Subsequently, the gene was cloned as a KpnI-PstI fragment into KpnI-PstI-digested pSPT19, which resulted in pSX4, and finally, the EcoRI-BamHI fragment from pSX4 was ligated into pT7-6. In construct pT7-T, xcpT expression is under the control of the Φ 10 promoter. Construct pGP1.2A is a derivative of pGP1.2 that contains xcpA introduced as a PstI fragment from pUP2. For construct pJGA03, lacZ was PCR amplified from pUR292 with primers PB3 (5′-CACAGGAAACAGGATCCACCATGATTACGG-3′), which replaces the start codon with a BamHI site (underlined), and PB4 (5′-GGCTCGAGGTCTAGATTACCCCTGACACC-3′), containing an XbaI site (underlined). The resulting product was cloned into pCRII-TOPO. The insert was subsequently excised by BamHI-XbaI digestion and introduced into BamHI-XbaI-digested pJGA01, replacing phoA with lacZ.

TABLE 2.

Plasmids used in this study

Plasmid	Relevant characteristic(s)a	Source or reference
pAX24	xcpP to xcpZ cluster in pLAFR3	19
pBBR1MCS5	Gmr; cloning vector; Plac	30
pBC18R	Apr; cloning vector	8
pPHO7	Apr; phoA without ss-encoding part	24
pET16b	AprlacI	Novagen
pCR-LacI	pBC18R lacI	This study
pCR-PhoA	pBC18R; phoA without ss-encoding part	This study
pYRC	pBBR1MCS-5 lacI	This study
pYRC-A	pYRC; phoA without ss-encoding part	This study
pCRII-TOPO	Apr Kmr; TOPO TA cloning vector	Invitrogen
pCR-XcpT1-2	pCRII-TOPO xcpT	This study
pJGA01	pYRC Plac-xcpT-phoA	This study
pMPM-K4Ω	Kmr; cloning vector; ParaBAD	34
pCR-XcpT1-3	pCRII-TOPO xcpT (start codon replaced by NcoI site)	This study
pJGA07	pMPM-K4Ω ParaBAD-xcpT-phoA	This study
pUC19	Apr; cloning vector	55
pUX4	pUC19 xcpT	This study
pSPT19	Apr; cloning vector	Pharmacia
pSX4	pSPT19 xcpT	This study
pT7-6	Apr; cloning vector; T7 Φ10 promoter	49
pT7-T	pT7-6 PΦ10-xcpT	This study
pGP1.2	Kmr; T7 polymerase gene	49
pUP2	pUC19 xcpA	12
pGP1.2A	pGP1.2 xcpA	This study
pUR292	Apr; cloning vector; promoterless lacZ	41
pJGA03	pYRC xcpT-lacZ	This study

^aGm, gentamicin; Ap, ampicillin; Km, kanamycin; ss, signal sequence.

Enzyme assays.

Secretion of elastase was analyzed qualitatively on LB plates with a top layer containing 1% elastin (Sigma). After overnight growth, the plates were screened for the presence of halos around the colonies. For alkaline phosphatase activity assays, overnight cultures were diluted to an optical density at 600 nm (OD₆₀₀) of 0.3 (E. coli) or 0.6 (P. aeruginosa) in fresh LB and grown for 2.5 (E. coli) or 3 (P. aeruginosa) h. E. coli cultures were subsequently split into two and incubated for another 60 min either at 30°C (permissive) or at 42°C (restrictive) to induce the sec phenotype of IQ85. The constructs were then induced with 0.01% l-arabinose (E. coli) (Sigma) or 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG) (P. aeruginosa) (Sigma). After 90 min (E. coli) or 3 h (P. aeruginosa) of growth in the presence of the inducer, the alkaline phosphatase activity in 1 ml of cell suspension was assayed with the substrate para-nitrophenyl phosphate (pNPP; J. T. Baker). Cells were pelleted by centrifugation, resuspended in 1 ml 0.9% NaCl, and permeabilized by the addition of 20 μl of chloroform and 20 μl of 0.05% sodium dodecyl sulfate (SDS). After the cells were vortexed, 420 μl of 0.5 M Tris-HCl (pH 8.0) was added and the suspension was incubated for 10 min at 30°C. The reaction was started by the addition of 50 μl of pNPP (30 μg/ml in 0.5 M Tris-HCl, pH 8.0). When significant yellow coloring was observed, the reaction was stopped by the addition of 500 μl 0.5 M NaOH. After centrifugation, the amount of released pNPP was determined by measuring the OD₄₂₀ of the supernatant. The alkaline phosphatase activity in AP units was calculated as OD₄₂₀ × 1,000/reaction time (min) × culture volume (ml) × OD₆₀₀. The activity of β-galactosidase was analyzed qualitatively on LB plates containing IPTG and X-Gal (Sigma) as the substrate. For quantitative β-galactosidase activity assays, overnight-grown cells were harvested from LB plates containing the appropriate antibiotic and 0.1% glucose and suspended in 0.9% NaCl. The cells were pelleted by centrifugation, resuspended in 950 μl PM2 buffer (40 mM Na₂HPO₄, 26 mM KH₂PO₄, pH 7.0) and permeabilized by the addition of 50 μl chloroform. After the addition of 50 μl o-nitrophenyl-β-d-galacto pyranoside (ONPG) (4 mg/ml in PM2 buffer), the hydrolysis reaction was performed at 30°C for 65 min. The reaction was stopped by the addition of 300 μl of 1 M Na₂CO₃. After centrifugation, the amount of released ONPG was determined by measuring the OD₄₂₀ of the supernatant. The β-galactosidase activity was calculated as OD₄₂₀ × 1,000/reaction time (min) × culture volume (ml) × OD₆₀₀. In the case of strain JS7131, quantitative β-galactosidase activity assays were performed on cells grown in liquid to allow for YidC depletion. The cells were cultured in LB broth containing 0.2% l-arabinose to an OD₆₀₀ of 0.6. The cells were collected by centrifugation and suspended in LB containing 0.2% l-arabinose or 0.2% glucose. To induce xcpT-lacZ expression, 0.5 mM IPTG was added and cultures were grown for an additional 90 min. β-Galactosidase activities were determined as described above.

SDS-PAGE and immunoblot analysis.

Bacterial cells were suspended in SDS-polyacrylamide gel electrophoresis (PAGE) sample buffer (2% SDS, 5% β-mercaptoethanol, 10% glycerol, 0.02% bromophenol blue, 0.1 M Tris-HCl, pH 6.8). Extracellular proteins were precipitated using 5% trichloroacetic acid (TCA) and washed with acetone. The amounts of proteins loaded were equivalent to an OD₆₀₀ of 1 of bacterial cells. Samples were heated for 10 min at 95°C and separated on SDS-PAGE gels (8% gels for XcpT-LacZ, 11% gels for XcpT-PhoA, and 14% gels for XcpT). Proteins were stained with Coomassie brilliant blue or transferred to nitrocellulose membranes by semidry electroblotting for immunodetection. The primary antisera used were anti-XcpT at 1:1,000 and anti-PhoA and anti-LacZ at 1:10,000 dilutions. Either alkaline phosphatase- or peroxidase-conjugated goat anti-rabbit immunoglobulin G antibodies (Biosource International) were used as secondary antibodies. Detection of the latter was performed with chemiluminescence (Pierce). In the case of pulse-labeled samples, polyacrylamide gels were incubated in Amplify (Amersham) after electrophoresis, vacuum dried, and exposed to X-ray films at −80°C.

Pulse-labeling experiments.

E. coli cells carrying plasmid pT7-T in combination with either pGP1.2 or pGP1.2A were grown overnight at 30°C in M9-Casamino Acids medium (M9 salts [42] supplemented with 1% Casamino Acids, 0.2% glucose, 1 mM MgSO₄, and 0.0001% thiamine). Overnight cultures were diluted to an OD₆₀₀ of 0.3 in M9 medium supplemented with all amino acids except for methionine and cysteine (all at 0.12%, except aspartic acid, glutamic acid, and tyrosine at 0.05% and tryptophan at 0.01%) and grown for 2.5 h. Then, cells were incubated at 42°C for 60 min to induce the synthesis of T7 RNA polymerase, as well as the sec phenotype in the case of strains MM52 and IQ85. Cells were pelleted and resuspended in one-fifth of the original volume of the growth medium. Samples were pulse-labeled for 1 min with 3 μCi Redivue l-[³⁵S]methionine (Amersham) at 42°C. Incorporation of label was stopped by placing the samples on ice and by rapidly adding one volume of a solution of 10% TCA.

Immunoprecipitation.

After removal of TCA from the pulse-labeled samples by washing with acetone, the pellets were dissolved in SDS buffer (2% SDS, 50 mM Tris-HCl [pH 8.0], 1 mM EDTA) and incubated for 10 min at 100°C. Subsequently, Triton buffer (2% Triton X-100, 50 mM Tris-HCl [pH 8.0], 0.15 M NaCl) was added and insoluble material was removed by centrifugation. Polyclonal antiserum was added, and after 3 h of incubation at room temperature, protein A CL-4B Sepharose beads (Amersham) were added. After incubation for 1 h at room temperature under gentle rocking, immunocomplexes were collected by centrifugation and washed with Triton buffer. Boiling for 10 min in SDS-PAGE sample buffer eluted antigens from the Sepharose beads.

Proteinase K accessibility.

Cells carrying plasmid pJGA07 were grown as described in the pulse-labeling experiments. After 1.5 h of growth, arabinose was added to a final concentration of 0.01% to induce XcpT-PhoA expression. After 1 h of induction, the cells were incubated for 1 h at 42°C to induce the sec phenotype. Pulse-labeling was performed as described above. Directly after the pulse-labeling, an excess of nonradioactive methionine/cysteine was added, and the cells were collected by centrifugation. For conversion to spheroplasts, cells were resuspended in ice-cold buffer A (40% [wt/vol] sucrose, 1.5 mM EDTA, 33 mM Tris-HCl [pH 8.0]) and incubated with lysozyme (final concentration, 5 μg/ml). After 10 min on ice, incubation was continued at 37°C for 10 min, followed by the addition of 10 mM MgCl₂. Aliquots of the spheroplast suspension were incubated on ice for 1 h in the presence or absence of proteinase K (final concentration, 50 μg/ml). Subsequently, 2 mM phenylmethylsulfonyl fluoride was added to the cell suspension and incubation was continued for 5 min on ice. Proteins were precipitated with 5% TCA and analyzed by SDS-PAGE and autoradiography.

Cell fractionation.

Overnight cultures of DH5α carrying pJGA03 were diluted into fresh, prewarmed LB medium to an OD₆₀₀ of 0.15 and grown for 2 h at 37°C with shaking at 200 rpm. Subsequently, IPTG was added to a final concentration of 1 mM. After 90 min of incubation, the cells were harvested and spheroplasts were prepared as described previously (38). After the spheroplasts were pelleted by centrifugation, the supernatant was kept as the periplasmic fraction. The spheroplasts were washed in 0.9% NaCl, resuspended in 2 mM EDTA-50 mM Tris-HCl (pH 8.5), and frozen at −20°C. After being thawed, they were disrupted by sonication and the membranes were collected by centrifugation for 1 h at 150,000 × g at 4°C. The supernatant contained the cytoplasmic proteins.

RESULTS

Inner membrane translocation of XcpT-PhoA occurs independently of other Xcp components.

To study pseudopilin transport across the inner membrane, the periplasmic reporter enzyme alkaline phosphatase (PhoA) without its signal sequence was C terminally fused to the complete major pseudopilin XcpT of P. aeruginosa. Plasmid pJGA01, carrying the xcpT-phoA gene fusion, was introduced into wild-type PAO25 and into the xcpT mutant PAO1ΔT, and immunoblot analysis with an antiserum directed against PhoA confirmed production of XcpT-PhoA (Fig. 1A). Some breakdown products migrating at the same position as wild-type PhoA were detected as well. Remarkably, based on halo formation on elastin plates (Fig. 1B) and analysis of extracellular proteins (data not shown), the fusion protein appeared to restore the secretion of the Xcp substrate elastase in the xcpT mutant strain, indicating that the PhoA moiety did not interfere with the functionality of the protein.

FIG. 1.

Production and functionality of XcpT-PhoA fusion protein. (A) Immunoblot analysis of whole-cell extracts of P. aeruginosa strain PAO25 and an xcpT mutant, PAO1ΔT, expressing XcpT-PhoA from pJGA01 or containing the vector pYRC-A. Immunodetection was carried out with PhoA-specific antibodies. The position of XcpT-PhoA is indicated by an arrow. Total cell lysate of E. coli strain C9 constitutively producing PhoA (indicated by an asterisk) was included for reference. (B) PAO25 and PAO1ΔT expressing XcpT-PhoA from pJGA01 or containing the vector pYRC-A were grown on LB agar containing 1% elastin. Secretion of elastase is visualized by clearance of elastin around the colonies.

PAO25 cells carrying pJGA01 displayed alkaline phosphatase activity (Fig. 2). Since alkaline phosphatase is active only when transported out of the cytoplasm, this result shows that the XcpT-PhoA fusion protein is transported across the inner membrane. Pseudopilin transport has been proposed to occur via a dedicated transport pathway formed by the other components of the secretion machinery. To determine whether other Xcp components indeed were required for inner membrane passage of XcpT-PhoA, pJGA01 was introduced into P. aeruginosa strain D40ZQ lacking the xcp gene cluster. D40ZQ cells carrying pJGA01 still displayed alkaline phosphatase activity, although the level was somewhat lower than that found in the wild-type strain (Fig. 2). Thus, cytoplasmic membrane passage of XcpT-PhoA can occur independently of the Xcp system. Durand et al. (15) have reported that the Pil system and the Hxc system also can function in the assembly of the XcpT protein into a pilus-like structure. To study whether one of these two systems was responsible for the inner membrane transport of XcpT-PhoA in the absence of the Xcp apparatus, pJGA01 was introduced into a strain mutated in all three pathways (PAOΔhxcΔpilACΔR). This strain still showed alkaline phosphatase activity, similar to that of the xcp deletion strain (Fig. 2). Hence, inner membrane passage of XcpT-PhoA can occur independently of the Xcp, Hxc, and Pil systems.

FIG. 2.

Enzyme activity of XcpT-PhoA in P. aeruginosa. Alkaline phosphatase activities were measured in P. aeruginosa strains expressing XcpT-PhoA from pJGA01 or containing the vector pYRC-A. The strains used were the wild-type strain PAO25; strain D40ZQ, which lacks the xcp cluster; and strain PAOΔhxcΔpilACΔR (Triple), which lacks hxcP to hxcZ, pilA to pilC, and xcpR. The alkaline phosphatase activity in AP units was calculated as OD₄₂₀ × 1,000/reaction time (min) × culture volume (ml) × OD₆₀₀. The bars represent the averages of three independent assays, and standard deviations are indicated.

XcpT-PhoA activity in E. coli depends on a functional Sec system.

Given the similarity between the N termini of prepilins and Sec signal sequences, the most likely inner membrane translocation pathway for XcpT-PhoA, since it is not transported via accessory Xcp components, would be via the Sec machinery. To investigate this possibility, XcpT-PhoA was produced in the heterologous host E. coli, of which several well-characterized sec mutant strains are available. Construct pJGA07, which contains xcpT-phoA under the control of ParaBAD, was introduced into the temperature-sensitive secY mutant IQ85 and its parental strain, IQ86. Expression of the XcpT-PhoA protein was verified by immunoblot analysis with an antiserum against PhoA (results not shown), and inner membrane transport of the fusion protein was studied by measuring alkaline phosphatase activity. At the permissive temperature of 30°C, IQ85 produced alkaline phosphatase activity similar to that of the parental strain (Fig. 3). However, at the restrictive temperature of 42°C, PhoA activity was drastically lower in the secY mutant than in the parental strain. This result shows that XcpT-PhoA is transported over the inner membrane of E. coli in a SecY-dependent manner.

FIG. 3.

Enzyme activity of XcpT-PhoA is dependent on SecY in E. coli. Alkaline phosphatase activities were measured in the temperature-sensitive secY mutant strain IQ85 and its parent, IQ86, both expressing XcpT-PhoA from pJGA07. Activities were measured in cultures incubated for 2.5 h at 30°C (white bars) or at 42°C (gray bars). The alkaline phosphatase activity in AP units was calculated as OD₄₂₀ × 1,000/reaction time (min) × culture volume (ml) × OD₆₀₀. The bars represent the averages of three independent assays, and standard deviations are indicated.

Processing and inner membrane transport of XcpT requires a functional Sec system.

To study the inner membrane translocation of native XcpT in E. coli, pulse-labeling experiments were performed. In these experiments, processing of XcpT by the prepilin peptidase XcpA was used as an indicator of inner membrane transport. Processing of XcpT results in the removal of 8 residues from the N terminus and can be visualized by a small mobility shift on acrylamide gels. To allow for simultaneous induction of the sec phenotype and XcpT expression, the xcpT gene was introduced into pT7-6 under the control of the T7 promoter. The resulting construct, pT7-T, was combined either with helper plasmid pGP1.2, which contains the gene for T7 RNA polymerase under the control of the λ P_L promoter and the cI857 gene encoding a temperature-sensitive λ repressor, or with pGP1.2A, which additionally contains xcpA. These constructs were introduced into the temperature-sensitive secA and secY mutant strains MM52 and IQ85, respectively, and their cognate parental strains. To induce the synthesis of T7 RNA polymerase and the Sec phenotype, cells were shifted to 42°C for 1 h before pulse-labeling was performed. In these experiments, processed XcpT was detected only upon coexpression of XcpA (data not shown). In the XcpA-expressing parental strains, MC410 and IQ86, processing of the pseudopilin XcpT was nearly complete after 1 min of pulse-labeling (Fig. 4). However, the maturation of XcpT was clearly reduced in the secA and secY mutants, which shows that its processing requires a functional Sec translocase.

FIG. 4.

Inhibitory effect of sec mutations on XcpT processing. XcpT was expressed from plasmid pT7-T in E. coli strains MC4100 and IQ86 and their isogenic, temperature-sensitive secA and secY mutants MM52 and IQ85, respectively, both coexpressing XcpA from pGP1.2A. After being pulse-labeled at 42°C and immunoprecipitated with anti-XcpT, samples were analyzed by SDS-PAGE, followed by autoradiography. The locations on the autoradiogram of mature XcpT and its precursor are indicated by m and p, respectively. WT, wild type.

The above results suggest that the Sec translocase is required for the transport of XcpT and that transport is a prerequisite for processing to occur. Lack of transport in the sec mutants was further evaluated in proteinase K accessibility experiments on spheroplasts. Since XcpT is intrinsically very resistant to the protease (results not shown), the XcpT-PhoA fusion protein was employed in these experiments. When the fusion protein was expressed from pJGA07 in the parental strains MC4100 and IQ86, it was almost completely degraded and thus accessible for proteinase K in spheroplasts (Fig. 5A). In these experiments, proteinase K treatment did not yield the stable PhoA moiety, likely because the kinetics were not fast enough to allow PhoA to obtain its mature conformation. When produced under the restrictive conditions, the XcpT-PhoA protein was completely protected from the protease in spheroplasts of strain IQ85, showing that the protein remains on the cytoplasmic side of the membrane in the absence of a functional Sec system. In these spheroplasts, pre-OmpA, the precursor of a well-known Sec substrate, was also found to accumulate in a form inaccessible to proteinase K (Fig. 5B). In spheroplasts of MM52 cells grown at the restrictive temperature, XcpT-PhoA was also protected against the proteinase K treatment, although not completely (Fig. 5A). In these cells also, some mature OmpA was detected, indicating that the SecA phenotype was not complete. As expected, the proteinase K treatment resulted in the degradation of the mature form only of OmpA (Fig. 5B).

FIG. 5.

Proteinase K accessibility of XcpT-PhoA and OmpA in spheroplasts. (A) Cultures of temperature-sensitive secA and secY mutants MM52 and IQ85, respectively, and their parental strains, MC4100 and IQ86, were induced for XcpT-PhoA expression from pJGA07 and incubated at the restrictive temperature. Cells were subsequently pulse-labeled with [³⁵S]methionine, converted to spheroplasts, and where indicated (−, absent; +, present), treated with proteinase K (prot K). After immunoprecipitation with anti-PhoA serum, samples were analyzed by SDS-PAGE, followed by autoradiography. Because of the large size of XcpT-PhoA (∼65 kDa), removal of the short XcpT leader peptide during processing is not visible as a mobility shift. (B) The same samples were also used for immunoprecipitation with anti-OmpA serum. Samples were analyzed by SDS-PAGE and autoradiography. The locations on the autoradiogram of mature OmpA and its precursor are indicated by m and p, respectively.

Cotranslational transport of XcpT-LacZ.

Transport of XcpT across the inner membrane is affected greatly in the temperature-sensitive secY mutant and, albeit to a lesser extent, also in the secA mutant. The cytoplasmic protein β-galactosidase (LacZ) can be used as a reporter to distinguish between the SRP and SecB pathways (5), which intersect at the Sec translocon. Cells expressing LacZ fused to a SecB substrate display a Lac⁺ phenotype because of the cytoplasmic accumulation of β-galactosidase. They are also sensitive to induction of the expression of the fusion protein because of jamming of the Sec translocon due to rapid folding of the LacZ moiety of the fusion protein. In contrast, when LacZ is fused to an SRP substrate, cells are Lac⁻ because β-galactosidase is exported. Since export is cotranslational, the enzyme cannot fold in the cytoplasm and no jamming of the translocon occurs. To some extent, the cells can still be sensitive to induction of the expression of the fusion protein, because of the formation of toxic aggregates of LacZ in the periplasm (5). To make use of this system, plasmid pJGA03 was constructed, encoding an XcpT-LacZ fusion protein. Production of this chimeric protein from pJGA03 in DH5α resulted in a Lac⁻ phenotype, as indicated by the white-colony phenotype on plates containing the β-galactosidase substrate X-Gal (Fig. 6A), even after two days of incubation. Immunoblot analysis with antisera against LacZ and XcpT revealed that the fusion protein was produced (Fig. 6B), although the majority was detectable as a smaller degradation product recognized only by the LacZ antibodies. The degradation product was not always observed (see Fig. 7B) and probably reflects degradation during sample preparation. However, the appearance of the degradation product allowed us to determine the localization of XcpT-LacZ. Cell fractionation showed that the fusion protein was associated with the membranes (Fig. 6B), whereas the degradation product corresponding to the LacZ part was found mostly in the periplasmic fraction, showing that the fusion protein is transported over the inner membrane. DH5α cells carrying pJGA03 were not sensitive to induction of the expression of the chimeric gene with IPTG. The efficient export of the XcpT-LacZ fusion protein across the inner membrane without jamming the Sec pathway is consistent with cotranslational targeting via the SRP pathway. To further substantiate this conclusion, pJGA03 was introduced into MC1060 and its derivatives containing the SRP mutant alleles ffs-69 (50) and ffh-87 (51), which confer mild defects on SRP-dependent secretion. The strains with the SRP mutant alleles formed light blue colonies on X-Gal-containing plates, while the parental strain remained white (Fig. 6A), showing that the SRP mutations interfere with efficient translocation of XcpT-LacZ to the periplasm. This effect was quantified by measuring β-galactosidase activity. As shown in Fig. 7A, the strains with the mutant alleles displayed a fivefold increase in β-galactosidase activity compared to that in the wild-type strain, whereas the production of XcpT-LacZ was the same in all the strains (Fig. 7B). These results underscore the SRP dependency of XcpT targeting to the Sec translocon. To test the effect of YidC on the translocation of XcpT-LacZ, pJGA03 was introduced into strain JS7131, which carries a ParaBAD-yidC operon. No effects on β-galactosidase activities were measured upon the removal of arabinose, which results in YidC depletion (43; data not shown).

FIG. 6.

XcpT-LacZ production and localization in E. coli. (A) E. coli strains DH5α and MC1060 and MC1060 derivatives with alleles ffs-69 and ffh-87 producing XcpT-LacZ from pJGA03 were grown overnight on LB plates with X-Gal. White colonies correspond to a Lac⁻ phenotype and blue colonies to a Lac⁺ phenotype. (B) DH5α cells carrying pJGA03 were induced for the expression of XcpT-LacZ, and whole-cell extracts or cell fractions were analyzed by SDS-PAGE and Western blotting. Immunodetection was performed with antibodies against LacZ and XcpT. The positions of XcpT-LacZ and of LacZ are indicated with an arrowhead and an asterisk, respectively. p, periplasmic fraction; m, membrane fraction; c, cytoplasmic fraction; wc, whole-cell extract; b, whole-cell extract of plasmidless DH5α cells (the weak band detected by the anti-LacZ serum in this extract likely corresponds to the inactive LacZΔM15 protein encoded from the chromosome [31]).

FIG. 7.

XcpT-LacZ production in the presence of mutant SRP alleles. (A) β-Galactosidase activity (in Miller units) of cells grown on LB plates. The activity values represent the means of results from four independent experiments, with standard deviations indicated in parentheses. The values of β-galactosidase activity are normalized against the wild-type strain and given as percentages. (B) Immunoblot analysis of whole-cell lysates of E. coli strain MC1060 and MC1060 derivatives with alleles ffs-69 and ffh-87, producing XcpT-LacZ from pJGA03. Immunodetection was carried out with LacZ-specific antibodies. The position of XcpT-LacZ on the immunoblot is indicated by an arrowhead.

DISCUSSION

Sequences resembling the N termini of prepilins are frequently detected in proteins from gram-negative and -positive bacteria, as well as from archaea (39). Although these sequences mediate transport across the inner membrane, the pathway involved in this process was largely unknown. To study this transport process, alkaline phosphatase was fused to XcpT, the major pseudopilin of the type II secretion machinery of P. aeruginosa, and alkaline phosphatase activity was used to monitor transport. Export of XcpT-PhoA did occur in a P. aeruginosa strain lacking functional Xcp, Hxc, and Pil systems, as well as in E. coli. These findings argue against the hypothesis that accessory Xcp components or their homologues form a dedicated inner membrane transport system for the pseudopilins. This idea was based on the observation that proteins with prepilin-like N termini are always found in concert with an ATPase (XcpR in P. aeruginosa), an integral inner membrane protein (XcpS in P. aeruginosa), and a prepilin peptidase. In support of this hypothesis, Chung and Dubnau (9) reported that the prepilin peptidase ComC of the Bacillus subtilis competence system is required for translocation of the pilin-like protein ComGC. Moreover, Kagami et al. (27) showed that a conditional mutation in XcpT could be suppressed by a secondary mutation in the cytoplasmic ATPase XcpR, indicating that these proteins interact during the assembly and/or functioning of the machinery.

However, our conclusion that the pseudopilin XcpT is not transported via accessory Xcp proteins is consistent with the observation that a mutant of the archaeon Sulfolobus solfataricus in which the genes encoding the accessory ATPase (homologue of XcpR) and the integral inner membrane component (homologue of XcpS) are deleted is not affected in its ability to insert sugar-binding proteins with prepilin-like N termini into the membrane (B. Zolghadr, personal communication). Moreover, although pilins and pseudopilins are unable to functionally replace each other, it is possible to exchange their leader peptides and hydrophobic domains without loss of function (29). Even in P. aeruginosa, where the Pil and Xcp systems function side by side, the N termini of PilA and XcpT could be exchanged without affecting function (our unpublished results). Thus, apparently, the N termini do not contain the information for the targeting of the (pseudo)pilins to the cognate machinery, which is in agreement with the export of these proteins via a general route. An explanation for the aforementioned results described by Kagami et al. (27) and Chung and Dubnau (9) might be offered by the possibility that the accessory proteins, although not required for the transport of the pseudopilins across the inner membrane, are required for their assembly into a pilus-like structure (16, 44). In the case of ComGC, translocation was assayed in NaOH solubility studies and it is conceivable that this method distinguishes ComGC assembled in a pilus-like structure from membrane-embedded ComGC rather than showing translocation.

Since the N termini of (pseudo)pilins to a certain degree resemble Sec signal sequences, we reasoned that inner membrane translocation of XcpT would occur via the Sec machinery. The involvement of the Sec translocon was studied in the heterologous host E. coli. In a temperature-sensitive secY mutant, translocation of XcpT-PhoA was blocked at the restrictive temperature, indicating that indeed the Sec pathway is involved. The requirement for SecA and SecY in XcpT transport could also be demonstrated in pulse-labeling experiments. Both XcpT processing and protease accessibility were affected in the secA and secY mutants at the restrictive temperature. Maturation of XcpT in these experiments was strictly dependent on the presence of the P. aeruginosa prepilin peptidase XcpA. Nonetheless, immunoblot analysis with anti-XcpT serum of extracts of E. coli cells expressing XcpT grown overnight showed partial processing (unpublished observation). Apparently, previously described E. coli prepilin peptidase homologues (22, 54) are functional, but their ability to process XcpT was not sufficiently efficient to be detected in the short pulse-labeling experiments. These experiments showed that processing of the pseudopilin occurs after transport, despite the fact that the prepilin peptidase has its active center at the cytoplasmic side of the inner membrane (32). Apparently, membrane insertion of XcpT is required to expose its processing site accurately to the XcpA peptidase.

In order to distinguish between targeting via the SecB or the SRP route, XcpT was fused to the reporter LacZ, which was previously shown to be a useful reporter for this purpose (5). This fusion protein was efficiently transported to the periplasm without jamming the Sec translocon, consistent with cotranslational transport via the SRP pathway. Moreover, efficient inner membrane translocation of the XcpT-LacZ fusion was dependent on fully functional Ffh and Ffs, but not on YidC. Taken together, our results demonstrate that XcpT is targeted to the Sec system via the SRP route, which is common for inner membrane proteins. The precise requirements for SRP signals are still not completely clear, but hydrophobicity and secondary structure have been shown to be important determinants (1, 5, 33). Indeed, the N termini of prepilins are highly hydrophobic and contain an extended N-terminal α-helix. Since pseudopilins have a high tendency to associate (40), cotranslational translocation may be important to circumvent premature interactions.

This report shows for the first time the Sec and SRP dependency of a pseudopilin for its export. Given the strong conservation of their N-terminal sequences, it is likely that other pilins and pseudopilins also use the SRP/Sec pathway. Indeed, similar conclusions were reached by Francetic et al. in the accompanying paper on the export of the pseudopilin PulG of the type II secretion system of Klebsiella oxytoca (21). We propose that, after translocation, the (pseudo)pilins leave the Sec translocon laterally, after which they become recruited by their cognate machinery. The polytopic inner membrane component will subsequently act as a platform for the assembly of the subunits into a piluslike structure, using the energy provided by the ATPase.