Pseudomonas aeruginosa Possesses Two Putative Type I Signal Peptidases, LepB and PA1303, Each with Distinct Roles in Physiology and Virulence

Abstract

Type I signal peptidases (SPases) cleave signal peptides from proteins during translocation across biological membranes and hence play a vital role in cellular physiology. SPase activity is also of fundamental importance to the pathogenesis of infection for many bacteria, including Pseudomonas aeruginosa, which utilizes a variety of secreted virulence factors, such as proteases and toxins. P. aeruginosa possesses two noncontiguous SPase homologues, LepB (PA0768) and PA1303, which share 43% amino acid identity. Reverse transcription (RT)-PCR showed that both proteases were expressed, while a FRET-based assay using a peptide based on the signal sequence cleavage region of the secreted LasB elastase showed that recombinant LepB and PA1303 enzymes were both active. LepB is positioned within a genetic locus that resembles the locus containing the extensively characterized SPase of E. coli and is of similar size and topology. It was also shown to be essential for viability and to have high sequence identity with SPases from other pseudomonads (≥78%). In contrast, PA1303, which is small for a Gram-negative SPase (20 kDa), was found to be dispensable. Mutation of PA1303 resulted in an altered protein secretion profile and increased N-butanoyl homoserine lactone production and influenced several quorum-sensing-controlled phenotypic traits, including swarming motility and the production of rhamnolipid and elastinolytic activity. The data indicate different cellular roles for these P. aeruginosa SPase paralogues; the role of PA1303 is integrated with the quorum-sensing cascade and includes the suppression of virulence factor secretion and virulence-associated phenotypes, while LepB is the primary SPase.

INTRODUCTION

Type I signal peptidases (SPases) are cytoplasmic membrane-bound enzymes that cleave amino-terminal signal peptides from proteins translocated by the general secretory pathway (Sec) of bacteria (reviewed in references 12, 43, 44, 67, and 69). SPases are unique serine proteases that use a catalytic Ser-Lys dyad mechanism. The signal peptides they remove have three conserved domains: a positively charged amino-terminal domain, a central hydrophobic domain, and a carboxy-terminal hydrophilic domain that contains the SPase cleavage site (44). Small neutral residues are found at the −1 and −3 positions relative to the cleavage site, with Ala being the most common amino acid at these positions (41). Substrates of the Sec-independent twin arginine translocation (TAT) pathway also possess similar signal peptides, although they contain a highly conserved twin arginine motif upstream of the hydrophobic region, while the type II signal peptides of lipoproteins are cleaved by a different protease (signal peptidase II) immediately upstream of a Cys residue that is part of the N-terminal lipoprotein box motif (9, 71).

In Gram-negative bacteria, SPase-mediated cleavage of signal peptides releases proteins into the periplasm, from where they may be transported across or into the outer membrane (19). A good illustration of the immense importance of SPases to bacterial cellular physiology was provided by Lewenza and colleagues (36), who performed an in silico analysis to identify protein export signals within the genome of Pseudomonas aeruginosa strain PAO1. This predicted that 801 P. aeruginosa proteins (14.4% of the genome) contain a cleavable type 1 signal peptide (36) and therefore showed that a significant proportion of the proteome of this opportunistic pathogen is targeted to the cell envelope and extracellular milieu. These proteins include components of the general secretory pathway (Sec), iron uptake proteins, outer membrane proteins and porins, flagellar structural proteins, catalase (KatB), cell wall biosynthesis enzymes, components of ABC transporters, and regulatory proteins (36). In addition, many important P. aeruginosa virulence factors possess a type 1 signal peptide, including elastases (LasA and LasB), exotoxin A, β-lactamase (AmpC), and proteins involved in the biosynthesis of alginate (AlgD, AlgE, AlgG, AlgL, and AlgF) (24, 29, 36), and thus reveal a direct role for P. aeruginosa SPase activity, not only in cellular housekeeping, but also in pathogenesis.

Multiple SPases are frequently observed in Gram-positive bacteria. The most extreme case is Bacillus cereus which contains seven paralogous SPases, while Bacillus anthracis, Bacillus subtilis, and Streptomyces lividans have six, five, and four chromosomally located SPases, respectively (69). Other notable examples are Staphylococcus aureus and Listeria monocytogenes, which possess two and three contiguous SPases, respectively (7, 10). However, in the case of S. aureus, SPase activity is performed by only one of the paralogues (SpsB), while the other (SpsA) does not possess the serine-lysine catalytic dyad and is thus devoid of catalytic activity (10). Conversely, L. monocytogenes is a good example of a pathogen that possesses SPases with different functions, as two of the three proteases (SipX and SipZ) have been shown to have different roles in pathogenesis, with SipZ being the major SPase of the organism (7).

In contrast, Gram-negative bacteria usually possess only a single SPase. This has been observed for Escherichia coli, whose SPase (LepB) has been extensively characterized, and documented for many organisms, including Haemophilus influenzae, Legionella pneumophila, Pseudomonas fluorescens, Rhodobacter sphaeroides, Salmonella enterica serovar Typhimurium, and Yersinia pestis (34, 69). However, the possession of multiple SPases is not exclusively a trait of Gram-positive organisms; two SPases have been demonstrated in Synechocystis sp. (80), and the Gram-negative soil bacterium Bradyrhizobium japonicum is known to have at least two functional SPases (SipS and SipF), with a further SPase that has yet to be studied (1, 38, 39). In addition, the list of Gram-negative organisms with multiple SPases is likely to increase as more sequenced genomes are explored for proteins with this important function.

There is urgent clinical need for new therapies against P. aeruginosa, which has a remarkable capacity to infect compromised individuals—it is the principal cause of chronic respiratory infection in cystic fibrosis (CF) patients and is a common nosocomial pathogen, with burn victims, intensive care unit patients, and those with indwelling devices (catheters or ventilators) particularly at risk from infection. In addition, its natural resistance to many frontline antibiotics is compounded by its genetic capacity to express a wide repertoire of resistance mechanisms and acquisition of additional resistance genes and beneficial mutations (33). The need for a functional SPase for viability has been demonstrated in other organisms, such as E. coli, S. aureus, and Streptococcus pneumoniae (10, 13, 79), and several inhibitors, including arylomycin and lipoglycopeptide natural products and β-lactam analogs (penem inhibitors), have been shown to have activity against SPases (3, 5, 31, 34, 37, 43, 51). In this study, we performed a molecular characterization of the two SPase homologues present within the P. aeruginosa genome, LepB and PA1303, in order to determine their physiological roles and their suitability as drug targets. LepB was found to possess all the characteristics of a conventional Gram-negative SPase, to be essential for viability, and to display intra- and interspecies conservation. In contrast, although PA1303 is conserved among P. aeruginosa strains, its mutation shows that it is dispensable in vitro and reveals a potential function in the suppression of virulence factor secretion through involvement in the quorum-sensing (QS) cascade.

MATERIALS AND METHODS

Bacterial strains, culture conditions, and DNA techniques.

The bacterial strains and plasmids used in this study are listed in Table 1. Bacterial strains were grown at 37°C in Luria-Bertani (LB) broth or on LB agar (Invitrogen) or Pseudomonas isolation agar (PIA) (Difco), as indicated. Antibiotics were used at the following concentrations; ampicillin, 100 μg/ml for E. coli; carbenicillin, 100 μg/ml for P. aeruginosa; gentamicin, 10 μg/ml for E. coli and 200 μg/ml for P. aeruginosa; tetracycline, 15 μg/ml for E. coli and 60 or 100 μg/ml for P. aeruginosa. Standard techniques for DNA manipulations were used (53).

Table 1.

Bacterial strains and plasmids

Strain or plasmid	Relevant characteristics and/or genotypea	Reference/source
Strains
E. coli
JM109	endA1 recA1 gyrA96 thi-1 hsdR17(rK− mK−) relA1 supE44 Δ(lac-proAB) [F′ traD36 proAB lacIqZΔM15]	78
S17-1 λpir	thi pro hsdR hsdM+ recA RP4-2-Tc::Mu-KM::Tn7 λpir	58
Rosetta(DE3)pLysS	F− ompT hsdSB(rB− mB−) gal dcm (DE3) pLysSRARE (Camr)	Novagen
P. aeruginosa
PAO1	Wild-type strain; MPAO1	27
ΔlepB/pUCP26-lepB	PAO1 with chromosomal lepB deletion, expressing lepB in trans from pUCP26; Tcr Gmr	This study
ΔPA1303	PAO1 containing chromosomal deletion in PA1303; Gmr	This study
ΔPA1303 pHERD26T	ΔPA1303 mutant carrying pHERD26T; Gmr Tcr	This study
ΔPA1303 pHERD26T-PA1303	ΔPA1303 mutant carrying pHERD26T-PA1303; Gmr Tcr	This study
PAO1 pHERD26T-PA1303-HA	PAO1 carrying pHERD26T-PA1303-HA; Tcr	This study
PAO1 pHERD26T	PAO1 carrying empty vector pHERD26T; Tcr	This study
Plasmids
pUCP18	Broad-host-range vector; Apr	55
pUCP20	pUCP18-derived broad-host-range vector; Apr	76
pUCP26	pUCP18-derived broad-host-range vector; Tcr	76
pHERD26T	pUCP26 Plac replaced with 2.4-kb AdhI-EcoRI fragment of araC-PBAD cassette and oriT; Tcr	47
pUCGm	Source of Gmr cassette; Apr Gmr	54
pEX100T	Gene replacement vector; Apr sacB+	56
pUC18-lepB	pUCP18 carrying lepB and flanking regions; Apr	This study
pUC-lepB::Gmr	pUC18-lepB carrying a Gmr cassette within the cloned lepB gene sequence; Apr Gmr	This study
pEX100T-lepB::Gmr	pEX100T carrying lepB containing a Gmr cassette and flanking regions; Apr Gmr sacB+	This study
pEX100T-ΔlepB::Gmr	pEX100T-lepB::Gmr with 700 bp of lepB removed; deletion construct used for lepB chromosomal mutation analysis; Apr Gmr sacB+	This study
pUCP26-lepB	pUCP26 carrying lepB and flanking regions; Tcr	This study
pUCP20-PA1303	pUCP20 carrying PA1303 and flanking regions; Apr	This study
pUCP20-ΔPA1303::Gmr	pUCP20-PA1303 with 456 bp of PA1303 removed and replaced by a Gmr cassette; Apr Gmr	This study
pEX100T-ΔPA1303::Gmr	pEX100T-containing construct used to generate a chromosomal deletion mutation in PA1303; Apr Gmr	This study
pHERD26T-PA1303	pHERD26T carrying PA1303; Tcr	This study
pHERD26T-PA1303-HA	pHERD26T carrying PA1303 C-terminally tagged with HA; Tcr	This study
pET28	Protein expression vector; Kanr	Novagen
pET28-ΔlepB	pET28 carrying lepB lacking N-terminal transmembrane regions; Kanr	This study
pET28-PA1303	pET28 carrying PA1303; Kanr	This study
pSB1075	AHL reporter plasmid; P. aeruginosa lasRI and luxCDABE from Photorhabdus luminescens	77
pSB536	AHL biosensor; ahyR″:: luxCDABE in pAHP13; Apr	61

^aAp^r, ampicillin resistance; Gm^r, gentamicin resistance; Tc^r, tetracycline resistance; Kan^r, kanamycin resistance; sacB⁺, levansucrase-encoding gene.

Fractionation of P. aeruginosa cells and Western blotting.

Briefly, PAO1 harboring pHERD26T (PAO1 pHERD26T) or pHERD26T-PA1303-HAtag (PAO1 pHERD26T-PA1303-HAtag) was grown overnight at 37°C in LB broth with and without 1% arabinose, and bacterial cells were pelleted by centrifugation (17,000 × g; 3 min). For the total cell lysate analysis, pellets were resuspended in 250 μl SDS (0.2%), and 10-μl aliquots were mixed with 10 μl Nu-PAGE sample buffer (Life Technologies), boiled for 5 min (99°C), resolved on a Nu-PAGE 4 to 12% Bis-Tris gel (Life Technologies), and transferred to polyvinylidene difluoride (PVDF) membranes. PA1303 was detected using monoclonal anti-hemagglutinin (HA) antibodies (Covance) diluted 1:1,000 in TBST (50 mM Tris/HCl, pH 8, 138 mM NaCl, 2.7 mM KCl, 0.1% [vol/vol] Tween 20) and anti-mouse immunoglobulin G peroxidase antibody (Sigma) diluted 1:5,000 in TBST. Blots were developed using the Lumiglo chemiluminescence detection system (Cell Signaling).

For the subcellular localization study, the separation of PAO1 pHERD26T-PA1303-HAtag cells grown overnight at 37°C in LB broth containing 1% arabinose into a soluble fraction (cytoplasm and periplasm) and separation of a total membrane fraction into inner and outer membrane fractions using Sarkosyl solubilization were performed as previously described (26). Six micrograms of each fraction was resolved on a Nu-PAGE 4 to 12% Bis-Tris gel (Life Technologies) and stained with Brilliant Blue G colloidal concentrate (Sigma). After confirmation that differential fractionation had been achieved, protein was transferred to PVDF membranes. The membranes were probed with rabbit polyclonal anti-XcpY and mouse monoclonal anti-RpoA (Neoclone) and anti-HA antibodies. Anti-mouse immunoglobulin G peroxidase antibody (Sigma) and anti-rabbit immunoglobulin G peroxidase antibody (DakoCytomation) were added to probed membranes as appropriate, and blots were developed as described above. The antibody against XcpY was a kind gift from Romé Voulhoux (CNRS, Aix Marseille University). The detection of RpoA and XcpY served as a marker for the detection of cytoplasmic and inner membrane fractions, respectively. Detection of OprF in the outer membrane fraction was achieved through liquid chromatography-tandem mass spectrometry (LC MS-MS) (Centre of Excellence for Mass Spectrometry, King's College London).

RT-PCR.

Expression of lepB was targeted using primers LepB RT-For (5′ GCGCGGCGATGTCATGGTGTT 3′) and LepB RT-Rev (5′ GCGGCTGTCGTTGGAGTTGTCG 3′), while PA1303 was targeted using PA1303 RT-For (5′ CGCGCACGGCTCCTCGG 3′) and PA1303 RT-Rev (5′ CACCGGGCGCTCATTGACGTA 3′). Reverse transcription (RT) was carried out on 250 ng total RNA using a Verso 1-Step RT-PCR Reddy mix kit (ThermoScientific). The following parameters were used: 55°C (30 min) for first-strand cDNA synthesis, followed by inactivation of the enzyme and 30 cycles of 94°C (30 s), 63°C (30 s [LepB]) or 61°C (30 s [PA1303]), and 72°C (1 min). A final extension step was applied at 72°C (5 min). The total RNA used was isolated from P. aeruginosa strain PAO1 planktonic cultures grown to logarithmic and stationary phases in full-strength LB using methodology described previously (74).

Cloning of LepB (truncated) and PA1303 (full length) for protein expression.

Using P. aeruginosa strain PAO1 chromosomal DNA as a template and primers LepB-hydro-del LepB-hydro-del (5′ GGAATTCCATATGCGTTCCTTCCTGGTCGAGCC 3′) and LepB-end-BamH1 (5′ CGGGATCCTCAGTGAATCACGCCGACC 3′), the lepB gene was PCR amplified to generate a DNA fragment lacking the identified N-terminal transmembrane regions to facilitate protein purification. These primers were designed to incorporate NdeI and BamHI restriction sites into the amplified template in order to permit cloning into the corresponding sites of expression vector pET28 (Novagen) to generate plasmid pET28-ΔlepB. This put the truncated lepB gene under the control of a T7 promoter and introduced a polyhistidine tag upstream of the gene to facilitate protein purification. Similarly, the full-length PA1303 DNA sequence was amplified using primers PA1303-up-Nde1 (5′ GGAATTCCATATGGGCCTGCTCGCCGCGAT 3′) and PA1303-end-BamH1 (5′ CGGGATCCTCAGCGCACCGAACCGATGCGC 3′), and the resultant PCR product was cloned into pET-28 to generate pET28-PA1303 using the same procedure. The presence of the correct cloned sequence was verified through DNA sequencing.

Protein production and purification.

PA1303 and LepB were overproduced using the pET system (Novagen) (60). Briefly, E. coli Rosetta(DE3)pLysS transformants containing pET28-PA1303 or pET28-ΔlepB were grown overnight at 18°C and induced with 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside) to express PA1303 and a truncated LepB protein, respectively. Cells from a 1-liter culture were harvested by centrifugation (3,315 × g) and resuspended in 30 ml buffer (20 mM Tris, pH 8.0, 500 mM NaCl, 10 mM imidazole, 1% Triton), and 25 U benzonase (Novagen) was added. The cells were lysed by two passages through an Emulsiflex high-pressure microfluidizer (Avestin) at 22,000 kPa, and cell debris was removed by centrifugation at 35,000 × g for 20 min. The soluble fraction was then incubated with 0.5 ml of Ni-NTA Sepharose (Sigma) for 1 h at 4°C on a rotator, the contents was poured into a column, and unbound material was removed under gravity. The Sepharose was then washed in 30 ml buffer (20 mM Tris, pH 8.0, 500 mM NaCl, 10 mM imidazole), and the protein was eluted in the following buffer: 20 mM Tris, pH 8.0, 500 mM NaCl, 500 mM imidazole. The purity of the protein was determined by SDS-PAGE using 12% gels and Coomassie blue or Simply Blue (Invitrogen) staining. Recombinant protein was then buffer exchanged into 50 mM Tris-HCl, pH 7.5, containing 0.5% Triton X-100 using PD-10 desalting columns (Sephadex G-25 M; GE Healthcare).

In vitro activity of purified LepB and PA1303.

The recombinant SPases were screened for enzymatic activity using fluorescence resonance energy transfer (FRET) technology. The intramolecularly quenched fluorescent substrate (Dabcyl)VSPAAFAADL(EDANS) (purchased from Peptide Protein Research Ltd., United Kingdom) contained a decapeptide (based on the cleavage region of the signal peptide sequence of the P. aeruginosa LasB protein) that links a fluorophore [5-((2-aminoethyl)amino)naphthalene-1-sulfonic acid) (EDANS)] with a quencher [4-((4-(dimethylamino)phenyl)azo)benzoic acid (Dabcyl)]. Cleavage of this decapeptide results in fluorescence emission. Reactions using a mixture containing 0.5, 1.0, and 1.5 μM recombinant LepB protein or 0.5 μM recombinant PA1303 protein and 20 μM FRET peptide (final concentration) in assay buffer (50 mM Tris-HCl, pH 7.5) containing 0.25% and/or 0.5% Triton X-100 were carried out at 37°C. Fluorescence intensity readings were taken over a period of 30 to 60 min using a Fluostar Optima microplate reader (BMG Laboratories). The fluorescence emission signal was measured at 520 nm over time, following excitation at 405 nm.

MALDI-TOF MS analysis of FRET peptide substrate cleavage by recombinant PA1303 protein.

Recombinant PA1303 protein in 50 mM Tris-HCl, pH 7.5, containing 0.5% Triton X-100 was dialyzed against 50 mM Tris-HCl, pH 8.0, at 4°C, with four changes of buffer. Reaction mixtures containing 0.5 μM recombinant PA1303 protein and 20 μM FRET peptide were carried out at 37°C. Control reactions with mixtures containing no recombinant PA1303 protein were set up in parallel. After 6 and 15 h of incubation, the reaction mixtures were lyophilized and resuspended in 40 μl in distilled water. Matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) was performed with a Bruker Microflex MALDI-TOF mass spectrometer fitted with a nitrogen laser operating at 337 nm using pulsed extraction in negative linear mode. Peptides were analyzed using 9H-pyrido(3,4)indole (Norharmane; Sigma) in methanol-water (2:1 by volume) at a concentration of 10 mg/ml, used as a matrix. A total of 0.5 μl of peptide/reaction products in water, together with 0.5 μl of matrix solution, was applied to the MALDI plate and allowed to dry in air. The instrument was calibrated using the peptides des-ArgI bradykinin (904.0 Da), angiotensin I (1,296.5 Da), and neurotensin (1,672.3 Da), and average masses were used throughout.

Mutagenesis analysis of lepB.

The P. aeruginosa PAO1 lepB gene and flanking regions was amplified by PCR using primers S26F (5′ GGTCCAGCGAATTACTCAC 3′) and S26R (5′ CGACATGCACTTCCTCAG 3′) and cloned into pUC18 to give plasmid pUC18-lepB. This plasmid was cut with AccIII at nucleotide position 750 of lepB, the sticky ends were polished with Klenow, and a gentamicin cassette was inserted to generate plasmid pUC-lepB::Gm^r. This construct was then excised and inserted into pEX100T to give pEX100T-lepB::Gm^r. pEX100T-ΔlepB::Gm^r was then generated by removing 698 bp of lepB with NarI and KpnI.

Allelic replacement in strain PAO1 was then attempted using standard methodology (56). However, although plasmid pEX100TΔlepB::Gm^r was found to successfully integrate into the PAO1 genome to form a merodiploid strain, excision of the plasmid to yield an inactivated LepB was found not to occur. Chromosomal lepB deletion mutants could be obtained, however, if lepB was provided in trans using plasmid pUCP26-lepB during the mutagenesis procedure. pUCP26-lepB was constructed using primers lepBup-str_for (5′-GCCCTCGCCCTGATCGTCCAC-3′) and lepBdown_rev (5′-CCCTGCGGCGGCGTTCG-3′). Briefly, electrocompetent stocks (16) of merodiploid colonies were made, which were then transformed with pUCP26-lepB (Tc^r), and the mutagenesis protocol was resumed. Using this procedure, colonies containing a chromosomal lepB deletion and harboring lepB in trans were generated. These colonies had the appropriate phenotype; colonies grew on LB agar containing gentamicin (200 μg/ml), sucrose (5%), and tetracycline (60 μg/ml), but not on LB agar containing carbenicillin (100 μg/ml). The production of a chromosomal lepB deletion mutation was confirmed using PCR primers lepB seq for (5′-CCAGGAAATGCGCGAACCGATCT-3′) and lepB seq rev (5′-GCCCGGCGTAGCTGCGATGG-3′), which are located outside the regions cloned into pUC18 and pUCP26.

Generation of a PA1303 chromosomal deletion mutant.

The PA1303 gene and flanking regions was amplified by PCR using primers PA1303 upstream (5′-CGCCAACCTGCTGCTGCTCAAGA-3′) and PA1303 downstream (5′-CCCGCATCGACTTCACCGTGGA-3′) and cloned into pUCP20 to give plasmid pUCP20-PA1303. Plasmid pUCP20-PA1303 was cut with the restriction endonucleases SnaBI and NcoI, removing 456 bp of PA1303 DNA. The resultant sticky ends were then polished with Klenow, and a gentamicin cassette was inserted to generate plasmid pUCP20-ΔPA1303::Gm^r. The PA1303 gentamicin-resistant construct was then excised using the restriction endonucleases EcoRI and SphI and inserted into pEX100T to give pEX100T-ΔPA1303::Gm^r. This plasmid was then introduced into donor strain E. coli S17 by electroporation for conjugal transfer into PAO1, which was performed as described previously (56). In the transformants obtained by this procedure the insertional mutation was confirmed to be in the right location through PCR, using primers flanking the chromosomal region cloned into pUCP20 (PA1303 upstream2, 5′ CCTGCTGGTGGTCGCCGGCTAC 3′; PA1303 downstream2, 5′ CGGCTGGCAGGGCGAGTTCG 3′). PCR with primers S26F and S26R was used to show that no insertion had occurred in the lepB gene of these transformants.

Construction of plasmids pHERD26T-PA1303 and pHERD26T-PA1303-HA.

In order to place PA1303 under the control of a P_BAD arabinose-inducible promoter, the gene was cloned into the shuttle vector pHERD26T (47). This was achieved using the following primers: PA1303-up-EcoRI (5′ AGAATTCCATGGGCCTGCTCGCCGC 3′) and PA1303-end-HindIII (5′ CAAGCTTTCAGCGCACCGAACCGATGCGC 3′), which through PCR introduced terminal restriction sites that were used for ligation into linearized vector. The generated plasmid, pHERD26T-PA1303, was then sequenced using primers that flank the cloning site (pHERD-SF, 5′ ATCGCAACTCTCTACTGTTTCT 3′, and pHERD-SR, 5′ TGCAAGGCGATTAAG TTGGGT 3′) to confirm that no mutation had arisen in the cloned PA1303 sequence. pHERD26T-PA1303 was then introduced into the PA1303 mutant by electroporation.

A similar strategy was used to generate plasmid pHERD26T-PA1303-HA, using primers PA1303-up-NcoI (5′ CGCCATGGGCCTGCTCGCCGCGA 3′) and PA1303-HA-ctag1 (5′ GCAAGCTTTCAAGCGTAGTCTGGGACGTCGTATGGGTAGCGCACCGAACCGATGCGCCGGGT 3′). This replaced the Wild-type (WT) PA1303 sequence with an HA tag (amino acid sequence, YPYDVPDYA) incorporated at the C terminus under the control of the arabinose-inducible promoter. pHERD26T-PA1303-HA was sequenced using primers pHERD-SF and pHERD-SR and introduced into PAO1 by electroporation.

Planktonic-growth analysis.

Overnight LB cultures (2.4 × 10⁹ to 2.5 × 10⁹ CFU/ml) were diluted 100-fold into LB broth, and planktonic growth at 37°C was monitored at 30-min intervals using a Fluostar Optima microplate reader (BMG Laboratories).

TEM.

P. aeruginosa overnight cultures were fixed in 0.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4) for 1 h at 4°C. Samples were prepared for transmission electron microscopy (TEM), as previously described (59), and visualized with a Jeol JEM-1200 EX11 transmission electron microscope at 80 kV.

Two-dimensional (2D) electrophoresis, in-gel digestion, and MALDI-TOF MS analysis of excised spots.

Supernatants from 500-ml overnight cultures of P. aeruginosa strain PAO1 and its isogenic PA1303 mutant were obtained through centrifugation, and a cocktail of protease inhibitors (Roche, Germany) were added. Supernatant (250 ml) was dialyzed against 4 liters of deionized water at 4°C for 2 days with four changes of water. The lysate was then freeze-dried and solubilized in 5 ml buffer (15 mM Tris-HCl, pH 7.5, containing 1% Triton X-100 and a cocktail of protease inhibitors), proteins were precipitated with 5 volumes of ice-cold 10% trichloroacetic acid in acetone containing 0.2% dithiothreitol, and the pellet was washed twice with the same volume of ice-cold acetone. The precipitated protein pellets were then solubilized in 1.0 ml of solubilization buffer (7 M urea, 2 M thiourea, 4% CHAPS {3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate}, 50 mM dithiothreitol, 0.002% bromophenol blue, 2% immobilized pH gradient [IPG] buffer, pH 3 to 10 [GE Healthcare, Bucks, United Kingdom]) at room temperature for 1 h with occasional mixing. After centrifugation of this suspension (100,000 × g), the clear supernatant was stored in 100-μl aliquots at −80°C.

Isoelectric focusing with immobilized pH gradients (IPG strips) in the IPGphor Isoelectric Focusing System (GE Healthcare) was carried out according to the manufacturer's instructions using 13-cm IPG strips in the pH range 3 to 11 (nonlinear) loaded with 250 μg protein per strip. The IPG strips were incubated in 10 ml of equilibration buffer (23) with gentle shaking for 15 min to denature the proteins and reduce disulfide bridges, followed by alkylation of cysteine residues (23). SDS-PAGE was performed in an SE600 standard vertical system without the use of a stacking gel. The gels were stained with colloidal Coomassie brilliant blue as previously described (40).

Protein spots were excised, destained, and digested with sequencing-grade trypsin (13 μg/ml) overnight. Peptide extracts were desalted and concentrated using Zip-Tip C₁₈ resin. Peptide mass fingerprinting was performed using MALDI-TOF MS and/or quadrupole time of flight tandem mass spectrometry (Q-TOF MS-MS), and peptides were identified using the complete translated open reading frames (ORFs) for sequenced P. aeruginosa strain PAO1 (MRC Clinical Sciences Centre, Imperial College Faculty of Medicine, London, United Kingdom).

Phenotypic assays.

Two methods were used to assess elastase production; the qualitative elastin-nutrient agar method and the quantitative elastin Congo red (ECR) assay (Sigma) (52, 64). Elastin-nutrient agar plates were prepared as previously described and contained an insoluble-elastin (Sigma; elastin from bovine neck ligament)-containing agar overlay and a nutrient agar base. Zones of elastin clearing were observed after overnight incubation at 37°C and storage at 4°C for 24 h. For the ECR assay, 100-μl aliquots of sterile supernatant from triplicate P. aeruginosa cultures grown for 15 h in LB broth or LB broth containing 1% arabinose were added to 2 ml of ECR buffer (0.1 M Tris, pH 7.0) containing 20 mg of ECR and incubated with shaking at 37°C for 18 h. Insoluble ECR was removed by centrifugation, and the absorption of the supernatant was measured at an optical density at 495 nm (OD₄₉₅) against a buffer blank.

Swarming motility plates contained constituents that have been described previously (49). Overnight LB broth cultures (2 μl) were inoculated onto the agar surface, and the plates were incubated at both 30 and 37°C for 15 h. Swimming motility was detected using medium containing 0.3% Bacto agar, 1% tryptone, 1% NaCl, and 0.2% glucose (22). Bacterial cells from agar plates were stab inoculated into the center of the swimming agar, and motility was observed after incubation at 30°C and 37°C for 15 h. Twitching motility was observed using LB agar (Invitrogen) containing 2,3,5-triphenyltetrazolium chloride (Sigma). Wild-type and mutant strains were stab inoculated into the agar and incubated at 37°C for 24 h, and bright-red twitching zones were observed after a further 2 to 3 days of incubation at room temperature.

Rhamnolipid production was observed using biosurfactant detection agar plates, which contained constituents that have been described previously (8). Forty microliters of overnight LB broth cultures were inoculated onto the agar surface, and the plates were incubated at 37°C for 24 h, followed by a further 24 h at room temperature. Rhamnolipid production was observed through the precipitation of cetyltrimethylammonium bromide (CTAB) and the formation of a white-outlined halo that surrounded the colonies that had grown (57).

AHL detection.

N-Acylhomoserine lactones (AHLs) were detected using E. coli harboring the following reporter plasmids: pSB1075 [for N-(3-oxododecanoyl)-l-homoserine lactone (3-oxo-C₁₂-HSL)] and pSB536 (for N-butanoyl homoserine lactone [C₄-HSL]) using methodology similar to that previously described (15). Briefly, sterile supernatant from triplicate P. aeruginosa cultures grown to exponential (4.5 h; OD₆₀₀ = 0.7) and stationary (8 to 8.5 h; OD₆₀₀ = 1.05 to 1) phases was obtained by membrane filtration. The supernatant was diluted 1:10 in LB broth, and 100 μl was added to a microtiter plate containing 100 μl 1:1,000-diluted E. coli harboring pSB1075. The same procedure was also followed for reaction mixtures containing E. coli harboring pSB536. The plates were incubated at 30°C, and luminescence was determined hourly and turbidity was determined after 4 and 6 h for reaction mixtures containing pSB536 and pSB1075, respectively, using a Fluostar Optima microplate reader (BMG Laboratories).

RESULTS

P. aeruginosa has two signal peptidase homologues.

The sequenced genome of P. aeruginosa strain PAO1 contains two putative SPases, PA0768 (annotated as LepB) and PA1303, which have 41% and 32% amino acid sequence identity, respectively, with the prototype bacterial SPase—E. coli LepB. The literature to date suggests that the presence of multiple SPases is uncommon for Gram-negative bacteria, and multiple SPases are often contiguous in Gram-positive bacteria; however, the P. aeruginosa paralogues are located in different regions of the PAO1 genome (LepB, nucleotides 837328 to 838182; PA1303, nucleotides 1414147 to 1414686). The locus in which P. aeruginosa lepB is present has an arrangement similar to that of the equivalent locus in E. coli. The only difference from the lepA-lepB-rnc-era-recO-pdxJ gene sequence present in E. coli is that immediately downstream of lepB in P. aeruginosa is a gene encoding a protein of unknown function (PA0769). However, the genes immediately up- and downstream of PA1303 appear to be a novel combination of genes flanking a Gram-negative SPase; PA1302 and PA1304 are annotated as a probable heme utilization protein precursor and a probable oligopeptidase, respectively (http://www.Pseudomonas.com).

Although both LepB and PA1303 contain the SPase catalytic Ser/Lys dyad (LepB, S90 and K145; PA1303, S40 and K83), the protein LepB is much larger (LepB, 284 amino acids, predicted molecular weight, 32,103; PA1303, 179 amino acids, predicted molecular weight, 20,067) and is similar in size to the LepB protein of E. coli (323 amino acids; molecular weight, 35,988), while SPases of approximately 20 kDa are more commonly found in Gram-positive organisms. An alignment of the amino acid sequences of LepB and PA1303 shows that the two proteins have 43.2% identity (58.9% similarity) (Fig. 1).

Fig 1.

Alignment of PAO1 lepB and PA1303 amino acid sequences. The sequences were aligned using CLUSTAL X (1.83). Conserved boxes A to E (shaded) were identified using sequence alignments and the following consensus motifs, in which uppercase underlined letters denote absolutely conserved, uppercase letters denote conserved, and X denotes not conserved (box B, IPSGSMXPTLX; box C, RGDIVVFXXP; box D, YIKRXXGXPGDXV; and box E, VPXGXYFXMGDNRDNSXDSR) (12, 34). Box A is labeled as previously identified (34). The first transmembrane helix identified for LepB using two prediction methods (aa 7 to 28, HMMTOP; aa 5 to 28, DAS) and the single short transmembrane region identified for PA1303 (aa 6 to 14, DAS) are underlined. The catalytic serine and lysine residues are boxed. Identical residues are also indicated (asterisks), as are conserved residues (colons).

The topologies of both LepB and PA1303 were predicted using two methods: Hidden Markov Model for TOpology Prediction (HMMTOP) software (65, 66) and the dense alignment surface (DAS) (transmembrane server) method (11). As expected for a Gram-negative SPase, two transmembrane helices (amino acids [aa] 7 to 28 and aa 57 to 76, HMMTOP; aa 5 to 28 and aa 65 to 81, DAS) were identified for LepB using both prediction methods. However, only the DAS prediction method identified a single short transmembrane region for PA1303 (aa 6 to 14).

As the identification of an N-terminal transmembrane region for PA1303 was inconclusive using bioinformatics topology prediction methods, a subcellular location analysis was performed. For this study, PA1303 was introduced in trans into strain PAO1 on plasmid pHERD26T-PA1303-HA. Plasmid pHERD26T-PA1303-HA carries the PA1303 gene with an HA tag (amino acid sequence, YPYDVPDYA) incorporated at the C terminus (to facilitate detection by Western analysis) under the control of a P_BAD arabinose-inducible promoter. Total cell lysates obtained from PAO1 pHERD26T-PA1303-HA and PAO1 pHERD26T (negative control for background and cross-reactivity) grown overnight in LB broth with and without 1% arabinose were analyzed by Western blotting with anti-HA antibodies. This analysis showed that the anti-HA antibodies have no cross-reactivity with proteins from PAO1 containing the empty vector (pHERD26T) and that PA1303 (predicted molecular weight, 20,067) could be detected only in total cell lysates derived from PAO1 pHERD26T-PA1303-HA grown in LB broth containing 1% arabinose (data not shown). Similar results were obtained using LB broth containing 0.5% arabinose.

To investigate the subcellular localization of PA1303, PAO1 pHERD26T-PA1303-HA cells grown overnight in LB broth containing 1% arabinose were separated into a soluble (cytoplasm/periplasm) fraction and membrane (inner and outer) fractions. When resolved by one-dimensional (1D) SDS-PAGE, different protein profiles were obtained for each fraction, as expected (Fig. 2A). A Western analysis using these fractions and anti-HA antibodies showed PA1303 to be located in the inner membrane (Fig. 2B), which is a feature consistent with an SPase. Antibodies specific for XcpY (inner membrane) and RpoA (cytoplasmic) were used as fractionation markers (Fig. 2B), while liquid chromatography (LC) MS-MS analysis of a band shown by 1D SDS-PAGE to be prominent only in the outer membrane fraction identified OprF (37 kDa; 14 unique peptides; 47% coverage), which is a common marker for this fraction (Fig. 2A).

Fig 2.

Subcellular localization of PA1303 in P. aeruginosa. Total cell (TC), soluble (S), inner membrane (IM), and outer membrane (OM) fractions from PAO1 harboring pHERD26T-PA1303-HA (PAO1 pHERD26T-PA1303-HA) grown overnight in LB broth containing 1% arabinose were used. (A) Fractions resolved on a Nu-PAGE 4 to 12% Bis-Tris gel (Life Technologies) and stained with Brilliant Blue G colloidal concentrate (Sigma). LC MS-MS analysis of a band only prominent in the outer membrane fraction (indicated by the arrow) identified OprF (14 unique peptides were detected, giving 47% coverage of OprF), an OM fraction marker. (B) Localization of PA1303 through immunoblotting of fractions with antibody specific for HA (detection of PA1303), XcpY (inner membrane marker), and RpoA (cytoplasmic marker).

SPases possess five regions of significant sequence homology referred to as boxes A to E. Box A is a transmembrane region, and boxes B to E (box B and D contain the catalytic serine and lysine residues, respectively) all lie near the signal peptidase active site and are part of the catalytic SPase protein fold (12, 43). Both LepB and PA1303 possess all the boxes of the catalytic region (boxes B to E), while the presence of box A is clear only for LepB (Fig. 1), which, similar to other Gram-negative SPases, corresponds to the second transmembrane helix identified by HMMTOP.

In order to determine whether both lepB and PA1303 are expressed, an RT-PCR analysis was performed. LepB and PA1303 transcripts were detected in RNA extracted from P. aeruginosa PAO1 cells grown to both logarithmic and stationary phases in planktonic culture (Fig. 3).

Fig 3.

Gene expression analysis by RT-PCR. Lanes 1 to 4, PA1303-specific primers used; lanes 5 to 8, lepB-specific primers used. +, reverse transcriptase added; −, no reverse transcriptase control included to exclude possible contamination of the RNA sample by chromosomal DNA. Shown are planktonic-culture RNA samples; L, logarithmic phase; S, stationary phase.

Intra- and interspecies comparisons of LepB and PA1303.

When the amino acid sequences of strain PAO1 LepB and PA1303 were compared with their orthologues in the six sequenced genomes at http://www.Pseudomonas.com for which these loci are complete (PA14, 2192, PA7, PACS2, LESB58, and 39016), at least 98.9% and 96% sequence identity was found for LepB and PA1303, respectively, showing that both proteins are highly conserved among different sequenced P. aeruginosa isolates.

A BLASTP search of the sequenced members of the family Pseudomonadaceae (P. fluorescens SBW25, Pf0-1 and Pf-5; Pseudomonas putida W619, F1, GB-1, and KT2440; Pseudomonas fulva 12-X; Pseudomonas syringae pv. syringae B728a; P. syringae pv. phaseolicola 1448A; Pseudomonas stutzeri A1501; Pseudomonas entomophila L48; P. syringae pv. tomato strain DC3000; Pseudomonas brassicacearum subsp. brassicacearum NFM421; and Pseudomonas mendocina ymp and NK-01 [http://www.Pseudomonas.com]) using the P. aeruginosa PAO1 LepB amino acid sequence also revealed orthologues with high sequence similarity (≥78.17% sequence identity; E values ≤ 2.0E−120) (Table 2) harbored within similar genetic loci (i.e., with the lepA-lepB-rnc-era-recO-pdxJ arrangement of genes).

Table 2.

Amino acid sequence identity of orthologues of LepB and PA1303 found within the genomes of sequenced members of the family Pseudomonadaceae

Pseudomonas species	LepB		PA1303
	%a	E value	%a	E value
P. fluorescens
SBW25	82.69b	6.00E−140	38.17b	5.00E−27
	33.0c	2.0E−14	38.16c	6.00E−23
Pf0-1	83.39	3.00E−136	40.54	4.00E−28
Pf-5	82.33	6.00E−134	40.54	4.00E−28
P. putida
W619	81.34	2.00E−136	40.54	3.00E−29
F1	82.04	6.00E−136	41.08	9.00E−29
GB-1	81.69	6.00E−136	40	1.00E−28
KT2440	81.69	1.00E−135	41.08	8.00E−29
P. fulva 12-X	82.04	4.00E−139	37.36	3.00E−28
P. syringae pv. syringae B728a	81.63	2.00E−138	41.21	1.00E−27
P. syringae pv. phaseolicola 1448A	81.27	8.00E−138	40.66	1.00E−27
P. stutzeri A1501	81.69	2.00E−137	41.21	2.00E−28
P. entomophila L48	81.69	6.00E−137	39.46	5.00E−28
P. syringae pv. tomato DC3000	80.57	8.00E−137	40.66	5.00E−28
P. brassicacearum subsp. brassicacearum NFM421	83.75	9.00E−137	41.08	2.00E−28
P. mendocina
ymp	79.23	2.00E−126	39.01	4.00E−28
NK-01	78.17b	2.00E−120	38.92b	5.00E−28
	32.51c	4.0E−19	48.89c	4.00E−41

^aPercent amino acid sequence identity obtained using a BLASTP search (http://www.Pseudomonas.com).

^bOrthologue 1.

^cOrthologue 2.

In contrast to the LepB analysis, a BLASTP search using the PA1303 amino acid sequence revealed much lower sequence identities (38.16 to 48.89%; E values, 6.0E−23 to 4.0E−41) (Table 2) with SPases from these related organisms. Therefore, these data show that P. aeruginosa LepB has more evolutionary conservation in this family. Interestingly, this analysis also revealed that only P. mendocina NK-01 and P. fluorescens SBW25 possess an additional SPase, annotated as MDS_1606 (33% identity to LepB, E value 4.0E−19; 49% identity to PA1303, E value 4.0E−41) and PFLU0154 (33% identity to LepB, E value 2.0E−14; 38% identity to PA1303, E value 6.0E−23), respectively (Table 2).

Overproduction and purification of LepB and PA1303.

Full-length PA1303 and truncated LepB (Δ2-76, LepB lacking N-terminal transmembrane regions) proteins were overexpressed in E. coli Rosetta(DE3)pLysS using the pET system, and purification was facilitated through His tags present at the N termini of both proteins. Both LepB and PA1303 were induced by IPTG and were observed with apparent molecular masses of 26 and 22 kDa, respectively, on Coomassie-stained SDS-PAGE (Fig. 4A and B).

Fig 4.

Purification of P. aeruginosa LepB (A) and PA1303 (B) under native conditions. Lane 1, lysed cells; lane 2, cell debris removed by centrifugation; lane 3, soluble fraction after centrifugation; lane 4, removal of unbound material—flowthrough; lane 5, wash fraction; lanes 6 to 9, eluted purified recombinant protein.

In vitro activity of purified LepB and PA1303.

The recombinant LepB protein generated under native conditions (Fig. 4A and B) was screened for enzymatic activity using FRET technology, an approach that has previously been successfully used to assess the in vitro activity of SPases of Staphylococcus epidermidis and S. aureus (4, 48). For this assay, an internally quenched fluorescent peptide substrate based on the signal peptide sequence cleavage region of the P. aeruginosa elastase (LasB) preproenzyme was used. The signal peptide of elastase (LasB) was chosen, as it is a well-characterized secreted virulence factor of P. aeruginosa that has a signal peptide cleavage site that has been experimentally identified as the peptide bond between the alanine residues at positions 23 and 24 relative to the N-terminal methionine (29). Thus, the peptide substrate (Dabcyl)VSPAAFAADL(EDANS) was designed to span 10 amino acids of the LasB signal peptide cleavage region, with alanine residues 7 and 8 corresponding to positions 23 and 24 of the preproenzyme. Dose-dependent cleavage of this peptide by recombinant P. aeruginosa LepB enzyme (0.5, 1.0, and 1.5 μM) in buffer containing 0.25% Triton X-100 was observed through the emission of fluorescence, which confirmed the enzymatic activity of this recombinant protein (Fig. 5A). As the yield of recombinant PA1303 protein was low during purification (Fig. 4B), the maximum concentration of the enzyme that could be used in the FRET assay was 0.5 μM. Recombinant PA1303 was also found to be active in the assay, and using buffer containing 0.5% Triton X-100, fluorescence emitted through cleavage of the FRET peptide with 0.5 μM PA1303 was found to be higher than that obtained using the highest concentration of LepB recombinant protein (1.5 μM) (Fig. 5B).

Fig 5.

Detection of in vitro activity of recombinant LepB and PA1303 protein using fluorescence spectroscopy. (A and B) Cleavage of the LasB FRET decapeptide (20 μM) by recombinant enzyme, resulting in fluorescence emission. LepB, 0.5 (diamonds), 1.0 (squares), and 1.5 (triangles) μM; PA1303, 0.5 μM (broken line, diamonds). The buffer conditions used were 50 mM Tris-HCl, pH 7.5, containing 0.25% Triton X-100 (A) and 0.5% Triton X-100 (B). FI, fluorescence intensity. A no-enzyme control was also performed, and the background values were subtracted from test samples. (A) Average background control units subtracted from each test replicate at time zero and 15 and 30 min were 1,162, 1,191, and 1,155 FI units, respectively. (B) Average background control units subtracted from each test replicate at time zero and 15, 30, and 60 min were 1,336, 1,330, 1,324, and 1,282 FI units, respectively. The error bars show the standard deviations of three replicates.

MALDI-TOF MS analysis of FRET peptide substrate cleavage by recombinant PA1303 protein.

FRET peptide was treated with recombinant PA1303 enzyme, and the hydrolysis products were analyzed by MALDI-TOF MS in order to determine the site of cleavage. The FRET peptide gave a signal at m/z 1,462 (Fig. 6A), which is consistent with the predicted molar weight. An identical mass spectrometry profile was obtained for FRET peptide diluted in distilled water (data not shown). After 6 h of incubation at 37°C with recombinant enzyme PA1303, the peak at m/z 1,462 and two additional signals at m/z 557 and m/z 913 were observed (Fig. 6B). The last two signals are consistent with SPase activity, which would cleave the FRET peptide between Ala residues 7 and 8 and thus would yield products of approximately m/z 566 and m/z 913. During the MALDI-TOF MS of the FRET peptide, we observed an additional signal at m/z 1,329 (Fig. 6A), which differs from the m/z of the FRET peptide (m/z 1,462) by 133 m/z mass units. During the analysis of products of the enzyme reaction, we observed a peak at m/z 780 (Fig. 6B and C), which differs from the peak at m/z 913 by 133 mass units. We believe the peak at m/z 1,329 is a contaminant/shortened FRET peptide produced during synthesis that is nevertheless cleaved by the recombinant PA1303 enzyme to yield two peaks at m/z 557 and m/z 780. After overnight incubation (15 h) at 37°C, no signals of FRET peptide at m/z 1,462 were observed (indicating complete digestion), whereas the signals at m/z 557 and m/z 913 were present (Fig. 6C).

Fig 6.

MALDI-TOF MS of cleavage of FRET peptide by recombinant PA1303. (A) FRET peptide (20 μM) in 50 mM Tris-HCl, pH 8.0 (assay buffer), incubated for 15 h at 37°C. (B and C) FRET peptide (20 μM) was incubated with recombinant 0.5 μM PA1303 in assay buffer for 6 h (B) and 15 h (C) at 37°C. The m/z at which major peaks occur are labeled.

LepB is essential for viability.

In order to determine the phenotypic effect of a deletion mutation in lepB, a lepB gene replacement plasmid (pEX100TΔlepB) was constructed. This plasmid contains PAO1 lepB and flanking regions with 698 bp of the 855-bp lepB gene removed (amino acids 18 to 252 are deleted) and replaced by a gentamicin cassette. This deletion removes the five signal peptidase conserved regions (boxes A to E), which include the catalytic dyad. Attempts to obtain a lepB deletion mutant using standard gene replacement methodology were unsuccessful (56). During the gene replacement process, plasmid pEX100TΔlepB was found to successfully integrate into the PAO1 genome to form a merodiploid strain, but excision of this plasmid to yield an inactivated LepB was found not to occur. However, colonies containing a chromosomal deletion in lepB were generated if the gene was provided in trans on plasmid pUCP26-lepB by transformation of merodiploid strains during the mutagenesis procedure. The production of a chromosomal lepB deletion mutation was confirmed using PCR primers lepBseqfor and lepBseqrev, which amplified the expected product of 2,106 bp, and the presence of the inserted gentamicin cassette was confirmed by digestion with two different restriction endonucleases. The presence of an intact copy of lepB in trans (pUCP26lepB) was confirmed by purifying plasmid DNA and digesting it with two different restriction endonucleases. Therefore, these data suggest that a functional lepB is essential for viability, and this is consistent with previous studies that have shown that chromosomal mutations can be made in essential genes of P. aeruginosa when a wild-type copy is provided in trans (17, 75).

PA1303 is not essential for viability in vitro.

Unlike lepB, we found that by using a standard mutagenesis procedure for P. aeruginosa we could generate PA1303 chromosomal deletion mutants in the absence of expression in trans. The mutants constructed had 456 bp of the PAO1 chromosomal PA1303 gene removed (this deletes amino acids 2 to 153) and replaced by a gentamicin resistance cassette. This deletion removes the four SPase conserved catalytic regions of PA1303 (boxes B to E). The generation of these mutants shows that, unlike LepB, PA1303 is not essential for viability in vitro. In addition, when growth as a planktonic culture was evaluated, similar growth rates were observed for the PA1303 mutant and the wild-type organism, and when stationary-phase cells were evaluated by TEM, no obvious changes in cell morphology were found to be induced through the mutation (data not shown).

Proteomic studies show that a chromosomal deletion mutation in PA1303 leads to the increased secretion of extracellular proteins.

2D gel electrophoresis of extracellular proteins prepared from wild-type P. aeruginosa strain PAO1 and isogenic PA1303 chromosomal deletion mutant stationary-phase cultures was performed using Immobiline Dry IPG strips in the pH range 3 to 11 (nonlinear) essentially according to the instructions of the manufacturer (GE Healthcare) (Fig. 7). Ten spots with obviously higher intensity in the PA1303 mutant extracellular fraction were analyzed by peptide mass fingerprinting using MALDI-TOF and/or Q-TOF MS-MS. Significant matches with the complete translated ORFs for the sequenced P. aeruginosa strain PAO1 were obtained for 9 out of 10 spots (Table 3). Six spots were identified as extracellular proteins: spot M-C, alkaline protease (AprA); spot M-D, aminopeptidase PA2939; spot M-G, LasB elastase PA3724; and spots M-F, M-H, and M-I, identified as the chitin-binding protein PA0852. All four identified proteins are commonly secreted by P. aeruginosa (42), with both AprA and LasB elastases possessing active roles in pathogenesis, including working in synergy to interfere with the host defense (46). In addition, LasB, together with the LasA protease, degrades elastin, a major structural component of human lung tissue and blood vessels (18, 68).

Fig 7.

Two-dimensional electrophoresis of extracellular proteins of PAO1 (A) and PA1303 mutant (ΔPA1303) (B). Samples were prepared as described in the text. Isoelectric focusing with immobilized pH gradients (IPG strips) in the IPGphor Isoelectric Focusing System (GE Healthcare) was carried out according to the manufacturer's instructions using 13-cm IPG strips in the range pH 3 to 11 (nonlinear), followed by SDS-PAGE in 12.5% acrylamide gels and staining with colloidal Coomassie brilliant blue. Protein spots with obviously greater intensity in the extracellular fraction of the PA1303 mutant (arrows, MA to MJ) were cut out from the gel, digested with trypsin, and subjected to peptide mass fingerprinting by MALDI-TOF MS and Q-TOF MS-MS in certain cases, as described in the text. The molecular masses of marker proteins are indicated.

Table 3.

Identification of proteins in spots of greater intensity in PA1303 mutant culture supernatant

Spot ID	PA no.	Gene	Protein identification	No. of peptides matched	% Coverage	MOWSE scoree	Predicted total mass (Da)c	pIc	Signal sequence	PA1303 mutant/ WT ratiod
M-A	PA4385	groEL	GroEL protein (60-kDa chaperonin)a	28	27	792	57,085	4.76	No	4.15
	PA2939	PA2939	Probable aminopeptidasea	12	22	606	57,511	4.76	Yes
	PA1092	fliC	Flagellin type Ba	8	18	552	49,242	5.18	No
M-B			No significant hits							23.21
M-C	PA1249	aprA	Alkaline protease (AprA)b	12	51	125	50,432	4.09	No	3.14
	PA1249	aprA	Alkaline protease (AprA)a	63	39	891	50,432	4.09	No
M-D	PA2939	PA2939	Probable aminopeptidaseb	14	41	159	57,511	4.76	Yes	4.59
			Probable aminopeptidasea	43	44	1301	57,511	4.76	Yes
M-E	PA0958	oprD	Outer membrane porin proteina	10	19	396	48,360	4.75	Yes	6.06
M-F	PA0852	cbpD	Chitin-binding proteina	27	41	573	41,917	6.85	Yes	3.47
M-G	PA3724	lasB	Elastase LasBb	16	55	145	53,687	6.74	Yes	2.39
	PA3724	lasB	Elastase LasBa	52	50	870	53,687	6.74	Yes
M-H	PA0852	cbpD	Chitin-binding proteina	18	20	379	41,917	6.85	Yes	6.23
M-I	PA0852	cbpD	Chitin-binding proteina	13	16	280	41,917	6.85	Yes	6.37
M-J	PA1094	fliD	Flagellar capping protein; FliDb	11	33	80	49,449	7.07	No	1.83
	PA1094	fliD	Flagellar capping protein; FliDa	66	30	871	49,449	7.07	No

^aIdentified by Q-TOF MS-MS.

^bIdentified by MALDI-TOF MS.

^cValues obtained from http://cmr.jcvi.org.

^dProteins matched between the PA1303 mutant and the WT organism and the ratio of normalized spot volume calculated using Image Master 2D Elite v3.1 software from Amersham Pharmacia Biotech. Spots M-A, M-B, M-H, M-I, and M-E were not identified among the proteins derived from the WT organism; therefore, the values quoted are the normalized volumes of these spots derived from the PA1303 mutant.

^eCalculated by MS-Fit using peptide mass fingerprint data.

Of the three remaining spots, one was shown to contain a cytoplasmic component and a surface component (spot M-A; GroEL PA4385 and FliC PA1092), while two spots were identified as surface components (spot M-E, outer membrane protein OprD; spot M-J, flagellar capping protein FliD). The presence of surface components, such as OprD and flagellin proteins, within the extracellular fraction could be a result of degradation either by elastase or naturally by protein turnover during the bacterial growth cycle, as suggested previously (42), while the presence of GroEL could be a result of cellular lysis of a fraction of the bacterial population. In addition, the identification of flagellar components could also be due to shedding of these filaments during cell culture.

Deletion of PA1303 affects multiple phenotypes.

Given the proteomic identification of the LasB elastase, one of the major virulence factors of P. aeruginosa, as an abundant extracellular protein of the PA1303 mutant, we next chose to perform a more detailed analysis of elastase activity produced by the mutant using elastin nutrient agar plates and the quantitative ECR assay. Both P. aeruginosa elastases (LasA and LasB) contribute to elastolysis on elastin nutrient agar plates, while the LasB elastase cleaves elastin in the ECR assay, and LasA augments this activity (63). In agreement with the proteomic data, increased elastinolytic activity was detected for the PA1303 mutant using both methods of detection (Fig. 8A and B). We also found that the elastase activity produced by the PA1303 mutant could be reduced when PA1303 was provided in trans on plasmid pHERD26T (Fig. 8A and B). However, the results in Fig. 8A and B were generated in the absence of arabinose and thus show that the basal transcription of PA1303 from the P_BAD promoter is enough to alter this phenotype of the PA1303 mutant. This phenomenon has been observed in the complementation of an efflux pump mutation in Burkholderia pseudomallei using a similar pHERD vector (pHERD30T) (47). In addition, a further reduction in elastase activity was observed for the complemented PA1303 mutant when 1% arabinose was added to the medium (Fig. 8C).

Fig 8.

Mutation of PA1303 in PAO1 results in increased elastase activity. PAO1, wild-type organism; ΔPA1303, PA1303 deletion mutant; ΔPA1303 pHERD26T, PA1303 deletion mutant containing empty vector pHERD26T; ΔPA1303 pHERD26T-PA1303, PA1303 deletion mutant containing PA1303 in trans (pHERD26T-PA1303). (A) Elastase activity detected using elastin-nutrient agar. (B) Elastase activity in LB broth detected using ECR. (C) Elastase activity in LB broth containing 1% arabinose detected using ECR. Standard deviations of triplicate cultures are shown.

We also investigated whether other important virulence-associated traits and factors were affected by a mutation in PA1303. P. aeruginosa is capable of surface translocation through swimming, swarming, and twitching motility. In the case of swarming motility, bacteria migrate as defined groups (tendrils) on semisolid surfaces (8). Interestingly, we found that the mutation in PA1303 led to a hyperswarmer phenotype, with tendrils covering much more of the semisolid agar surface than the wild-type organism after incubation at both 30 and 37°C (Fig. 9A and B). Indeed, at 37°C, the PA1303 mutant had covered the whole plate. The hyperswarmer phenotype was found to be reversed when PA1303 was expressed in trans on plasmid pHERD26T-PA1303 (Fig. 9A and B). In contrast to the swarming motility observations, the PA1303 mutant showed swimming and twitching motility phenotypes similar to that of the wild-type organism under the conditions used (results not shown).

Fig 9.

Mutation of PA1303 in PAO1 results in a hyperswarmer phenotype. Shown are swarming motility plates incubated at 30°C (A) and 37°C (B). PAO1, wild-type organism; ΔPA1303, PA1303 mutant; ΔPA1303 pHERD26T, PA1303 mutant containing empty vector pHERD26T; ΔPA1303 pHERD26T-PA1303, PA1303 mutant containing PA1303 in trans (pHERD26T-PA1303).

Swarming motility is known to require propulsion using a polar flagellum and also the surfactant rhamnolipid, which reduces friction between the cell and a semisolid surface for swarming motility (30). Therefore, we postulated that increased rhamnolipid production could be responsible for the observed increase in swarming motility. Indeed, using indicator agar plates in which a halo is produced when a cationic dye (methylene blue) present in the medium is precipitated by an anionic surfactant, such as rhamnolipid, we observed that the PA1303 deletion mutant produced a larger halo than the wild-type organism (Fig. 10A). In addition, the halo surrounding the complemented PA1303 mutant (ΔPA1303 pHERD26T-PA1303) was found to be smaller than that produced by the PA1303 mutant with and without empty vector (ΔPA1303 and ΔPA1303 pHERD26T) (Fig. 10A). Thus, this is also evidence of the increased production of another virulence factor, as rhamnolipid is associated with a number of biological activities, including inactivation of tracheal cilia of mammalian cells and solubilization of lung surfactant phospholipids, which facilitates their cleavage by phospholipase C (25, 32). It is also involved in the maintenance of biofilm architecture and biofilm dispersal (6, 14).

Fig 10.

Mutation of PA1303 in PAO1 results in increased rhamnolipid production, decreased pigmentation on PIA, and an increase in production of C₄-HSL. PAO1, wild-type organism; ΔPA1303, PA1303 mutant; ΔPA1303 pHERD26T, PA1303 mutant containing empty vector pHERD26T; ΔPA1303 pHERD26T-PA1303, PA1303 mutant containing PA1303 in trans (pHERD26T-PA1303). (A) Rhamnolipid production was detected through precipitation of CTAB and halo formation on agar plates using a method described previously (8). (B) Growth and pigmentation after overnight growth (37°C) on PIA. (C and D) Detection of AHL production in exponential-phase (shaded bars) and stationary-phase (white bars) cultures using reporter plasmids pSB1075 for 3-oxo-C₁₂-HSL (C) and pSB536 for C₄-HSL (D). RLU, relative light units (test minus no-supernatant plasmid controls) after 6 h (C) and 4 h (D) of incubation at 30°C. Standard deviations of triplicate cultures are shown.

Another noticeable difference between the WT and the PA1303 mutant was the amount of pigmentation on PIA, which is formulated to enhance pyocyanin production. On this agar, the PA1303 mutant was found to be less pigmented than the wild-type organism (Fig. 10B). Interestingly, however, the PA1303 mutant appeared to be more pigmented on elastin-nutrient (Fig. 8A) and swarming agar (Fig. 9B) at 37°C than the wild-type strain. Thus, it appears that the role of PA1303 in the secretion of pyocyanin is dependent on culture conditions. This is not the first observation of culture condition-dependent activation and repression of pyocyanin production, as a P. aeruginosa vfr mutant has been shown to produce increased and decreased amounts of pyocyanin on PIA and LB agar, respectively, compared to the wild-type organism (2).

Deletion of PA1303 alters the QS cascade.

It is well established that many P. aeruginosa virulence factors are regulated in a cell density-dependent manner by the LasR-LasI and RhlR-RhlI QS systems, which use AHLs as signaling molecules. As mutation of PA1303 was shown to influence several quorum-sensing-controlled phenotypic traits (elastinolytic activity, pyocyanin and rhamnolipid production, and swarming motility), we used the AHL reporter plasmids pSB1075 and pSB536 to detect the signals synthesized by the products of the lasI and rhlI genes (3-oxo-C₁₂-HSL and C₄-HSL, respectively) to determine whether the changes in these multiple phenotypes were due to an alteration in AHL levels. These bioassays showed that the levels of 3-oxo-C₁₂-HSL were similar in exponential and stationary-phase cultures of the WT organism and the PA1303 mutant (Fig. 10C), while the level of C₄-HSL was found to be increased by 3.4- and 1.7-fold in the PA1303 mutant in exponential- and stationary-phase cultures, respectively (Fig. 10D). Thus, a mutation in PA1303 was found to affect the level of production of the signaling molecule (C₄-HSL) of the RhlR-RhlI QS system but not the signaling molecule (3-oxo-C₁₂-HSL) of the LasR-LasI system.

DISCUSSION

In this study, we carried out a molecular characterization of the two P. aeruginosa SPases (LepB and PA1303). We showed that recombinant protein generated for both paralogues was active and examined their inter- and intraspecies genomic distribution, their physical properties and gene expression, and the consequences of their mutation.

LepB was shown to have features similar to those of the prototype SPase, E. coli LepB: a similar genetic locus organization, a similar topology (two transmembrane helices), the five regions of significant sequence homology (boxes A to E) that contain the catalytic Ser/Lys dyad associated with SPase activity, and a similar size. In addition, we found that LepB is essential for viability and is highly conserved among all pseudomonads. Previously, we generated transcriptomic data sets that profiled global gene expression of P. aeruginosa as (i) a surface-associated community (biofilm) grown under aerobic conditions and as a free-living (planktonic) culture and (ii) an anaerobic biofilm (72–74). When the values for LepB were observed in these data sets, it was found to be expressed under all conditions (planktonic culture, logarithmic and stationary phases; biofilms, developing, mature, and anaerobic), with the highest expression levels found in actively growing cells (logarithmic-phase planktonic culture and the developing biofilms sampled at 8 h) (data not shown). This information suggests an active role for the protease in both the planktonic and biofilm modes of growth. Therefore, our genomic, transcriptomic, and phenotypic evidence strongly suggests that LepB is the primary signal peptidase of P. aeruginosa.

The role of PA1303, however, is at present less clear. Although it possesses the four conserved boxes associated with the catalytic activity of SPases (boxes B to E) and was detected in the inner membrane fraction in the subcellular localization analysis (Fig. 2B), the predicted molecular weight of PA1303 (20,067) is very small for an SPase of Gram-negative bacteria and is a size more associated with SPases of Gram-positive bacteria (43). PA1303 also displays much lower sequence similarity (38 to 49%) than LepB (≥78%) with orthologues found in other pseudomonads (Table 2). Although the RT-PCR analysis in this study showed that PA1303 is expressed in planktonic culture (logarithmic and stationary phases), expression of PA1303 was barely detectable by microarray probes under all conditions in our previously generated transcriptomic biofilm and planktonic data sets (72–74). A possible explanation for the inconsistency between this study and previous studies could be that expression of PA1303 was below the threshold of detection by microarray analysis, due to the fact that weaker medium was used for the microarray study (1/5-strength LB medium) than for the RT-PCR analysis (full-strength LB medium). Although we know that PA1303 is expressed at least in planktonic culture, the observation that deletion mutants in PA1303 could be easily obtained shows that it is dispensable and its physiological role is not essential to the viability of P. aeruginosa in vitro. However, our studies showed that PA1303 has an important role in the control of virulence and virulence-associated phenotypes of this important opportunistic pathogen.

In species with multiple functional SPases, the individual enzymes are usually not essential for viability, and although they can differ in their contributions to the process of secretion, this suggests that they can at least partly complement each other (21, 69). Overlapping substrate specificities between different SPases have been commonly observed in Gram-positive bacteria with multiple SPases, while the efficiency of preprotein processing of individual SPases within an organism is known to vary (20, 21, 62). A good illustration of a pathogen that has SPases with distinct activities is L. monocytogenes, which possesses three contiguous SPases (SipX, SipY, and SipZ) (7). Bonnemain et al. found that SipX was not essential for bacterial growth but played a role in virulence; inactivation of sipY had no detectable effect, while SipZ was found to be the dominant SPase, being required for normal growth, efficient protein secretion, intracellular survival, and virulence. It is logical to hypothesize that the insertional inactivation of a nonessential SPase would be detrimental to the secretory capacity of an organism. This was observed in the L. monocytogenes study, where a mutant expressing only SipX showed reduced secretion of virulence factors (listeriolysin O and phosphatidylcholine phospholipase C), and a reduced secretory capacity has been observed in an S. lividans sipY mutant (7, 45). However, it has also been demonstrated that the insertional inactivation of a nonessential SPase can improve the processing of a preprotein substrate. For example, in B. subtilis, whose multiple SPases have overlapping substrate specificities, the deletion of the SipS or SipU SPase leads to more efficient cleavage of the α-amylase preprotein substrate (62). Therefore, the observed phenotypic changes found in the PA1303 mutant could be the result of decreased or more efficient processing of preprotein(s) in the absence of the PA1303 protease.

In this study, we show that both recombinant LepB and PA1303 can process a FRET decapeptide substrate based on the signal peptide cleavage region of LasB (Fig. 5A and B). This is in contrast to S. aureus, where only one (SpsB) of its two paralogues has been shown to have SPase activity (10, 28). We also show that fluorescence generated through cleavage of the FRET peptide with 0.5 μM recombinant PA1303 was approximately 4.5- to 7-fold higher than that obtained with the same concentration of recombinant LepB. However, although this result suggests that LepB and PA1303 have overlapping substrate specificities, further studies must be performed to determine whether this apparently more efficient processing of the FRET decapeptide is due to the reaction conditions used (50 mM Tris-HCl, pH 7.5, 0.5% Triton X-100) favoring the recombinant PA1303 enzyme or whether the two enzymes have different preferred substrates.

Four secreted P. aeruginosa proteins (LasB, AprA, aminopeptidase PA2939, and CbpD) were all in greater abundance in the extracellular fraction of the PA1303 mutant (average ratios of normalized spot volume, 2.39, 3.14, 4.37, and 4.85, respectively) (Table 3), while swarming motility, rhamnolipid production, and elastinolytic activity were also found to be increased. As all four proteins and rhamnolipid production are regulated through cell-cell signaling and AprA does not harbor a signal sequence and is not translocated by the general secretory pathway, this suggested that an alteration in the QS cascade would also be a consequence of chromosomal deletion of PA1303 and would be responsible for the observed altered phenotypes. Using reporter plasmids to detect levels of 3-oxo-C₁₂-HSL and C₄-HSL in exponential- and stationary-phase cultures, we did indeed observe an increase in the level of C₄-HSL for the PA1303 mutant, but not in the level of 3-oxo-C₁₂-HSL (Fig. 10C and D). Control of QS is known to be multilayered and hierarchical, with the Las system exerting control over the Rhl system (35). In addition, quinolone signaling molecules and a plethora of other transcriptional and posttranscriptional regulators are integrated into the QS circuit (15, 70). Thus, our data suggest that the protein(s) that is targeted for processing by PA1303 is involved (directly or indirectly) in the control of the RhlI-RhlR system. However, the exact molecular events that lead to the observed phenotypic changes are at present unclear. Possible explanations are the enhanced cleavage of signal peptides of a positive regulator(s) of the RhlI-RhlR system by LepB in the absence of PA1303 or that PA1303 has a specific role in the cleavage of signal peptides of a negative regulator(s) of this system. The identification of substrate proteins that are cleaved by LepB and PA1303 (801 P. aeruginosa proteins are predicted to possess type 1 signal peptides) and the determination of the efficiency of these cleavage reactions will be key to understanding the phenotypes generated by a PA1303 mutation. A role for an SPase in QS has only previously been reported for S. aureus, where the SPase SpsB was shown to cleave the N-terminal leader of AgrD, which is the peptide precursor of the autoinducing peptide (AIP) (28).

As LepB appears to be essential for viability of P. aeruginosa, it is an attractive target for a protease inhibitor. The carboxy-terminal catalytic domains of Gram-negative SPases are present on the outer surface of the cytoplasmic membrane facing into the periplasm and therefore would be accessible to any inhibitor that could traverse the outer membrane. SPases have been shown to be viable drug targets in other bacteria, with lipoglycopeptides showing antibacterial activity against E. coli and S. pneumoniae while S. epidermidis is sensitive to the related arylomycins (5, 31, 51). Unfortunately, the type I SPase inhibitors identified to date have not shown antibacterial activity against P. aeruginosa. Recently however, P. aeruginosa was rendered susceptible to the antibiotic activity of arylomycin derivatives through mutation of a resistance-conferring proline residue (P84) in LepB (50). This not only supports our observation that P. aeruginosa LepB is essential for viability, but also inspires confidence in the suitability of the protease as a drug target. In addition, it emphasizes the need for new and more potent rationally designed synthetic type I SPase inhibitors optimized for the highly conserved P. aeruginosa LepB catalytic region, which can bind regardless of the resistance-conferring proline residue.