Pseudomonas aeruginosa is an opportunistic bacterial pathogen that can cause fatal acute lung infections in critically ill individuals. Lung damage due to chronic infections in cystic fibrosis sufferers is the major cause of morbidity and mortality in this group. The bacterium produces various immunomodulatory products that enable it to survive in the lung. Innate and increasing resistance to antibiotic therapy shown by this organism heightens the need for development of a vaccine. This study reports the identification of six non-integral protein antigens; Pa 13, azurin, acyl carrier protein (ACP), amidase, aminopeptidase and KatE, purified from a mucoid strain of P. aeruginosa. N-terminal amino acid sequencing was used to identify these proteins and, based on their ascribed functions, determined that their normal cellular location was cytosolic. A rat model of acute pulmonary infection was used to investigate the ability of these protein antigens to enhance pulmonary clearance of a live P. aeruginosa challenge. Mucosal immunisation with four of the six antigens significantly enhanced bacterial clearance from both the lavage fluid and lung tissue. The greatest level of clearance was demonstrated for the antigens; KatE, aminopeptidase and amidase. Enhanced bacterial clearance was maintained when the antigens amidase and aminopeptidase were produced in recombinant form. When delivered parenterally, aminopeptidase demonstrated its continued efficacy as a vaccine candidate. This study has demonstrated that non-integral outer membrane proteins are antigenic and protective and warrant further investigation as potential components of a vaccine.
Pseudomonas aeruginosa is an important opportunistic pathogen in animals and humans. It is among the most frequently isolated organisms from patients with Gram-negative hospital infections and can be found to induce infections of endemic proportions. P. aeruginosa is also associated with complications in patients with cystic fibrosis (CF), or following surgery, trauma or burns [1,2]. In pulmonary infections in patients with CF, these alginate-producing bacteria begin to grow as microcolonies . As the bacteria survive in the lung, tissue damage occurs, due in part to secreted toxins and enzymes produced by P. aeruginosa, but also because of the host response to infection . The extensive damage to the lungs is a major cause of morbidity for CF patients .
Control against P. aeruginosa is difficult due to its biofilm mode of growth and high resistance to antibiotics. Therefore the development of immunotherapy as an alternative for controlling P. aeruginosa infections is required.
Investigations have shown the potential of mucosal immunisation with a killed whole cell P. aeruginosa vaccine [6,7]. However, strain specificity with a whole cell vaccine limits its usefulness. More recently, it has been shown that mucosal immunisation with the P. aeruginosa catalase, KatA, can enhance clearance of homologous and heterologous P. aeruginosa pulmonary infection . KatA, recognised as a cytosolic protein, appears to become surface-exposed, following cell lysis . Results from animal challenge experiments demonstrate that clearance of bacteria from the lung correlated with increased antigen-specific IgG antibody levels. This protective capacity of an antigen that is not an integral membrane protein represents a comparatively new concept in bacterial vaccinology. Immunisation against Helicobacter pylori using cytosolic proteins adds support to this concept [10,11]. This study investigated the ability of other non-integral OMPs, to confer similar protection in an animal model of acute pulmonary infection. The identification of several protein antigens purified from P. aeruginosa and their efficacy as both mucosal and parenteral immunogens, in both native and recombinant forms, is reported here.
2 Materials and methods
The mucoid P. aeruginosa, strain 385 (Serotype 2) was used for protein purification and has been studied previously [6,8]. This strain was originally isolated from a chronically infected patient with CF. Bacterial stocks were stored at −85°C in nutrient broth (Oxoid, Unipath, Basingstoke, Hampshire, England) supplemented with 10% glycerol (vol/vol).
2.2 Purification of native protein antigens
Bacteria were grown and proteins purified using Zwittergent extraction, anion chromatography and gel electrophoresis as previously described . Purification of azurin required the additional step of hydrophobic interaction chromatography (HIC) following the anion-exchange chromatography. HIC was performed using an Econo-pac 5 ml cartridge (Bio-Rad) with the support matrix composed of methyl Sepharose (–OCH3) groups. The column was equilibrated with 2.4 M ammonium sulfate, 100 mM sodium phosphate, pH 6.8, and proteins eluted by a reduction in the ionic strength using 100 mM sodium phosphate, pH 6.8. Proteins were distinguished by molecular mass on sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) at each stage of purification and desalted by dialysis. N-terminal amino acid sequencing was carried out on the final purification products for identification of the proteins prior to immunisation. The presence of lipopolysaccharide (LPS) was assessed by the E-toxate Limulus lysate test (Sigma, St. Louis, MO, USA). Protein concentration was determined using the Pierce Micro BCA protein assay reagent and albumin standards (Rockford, IL, USA).
2.3 Analytical SDS—PAGE
SDS–PAGE was performed under reducing conditions in SDS-reducing buffer containing dithiothreitol and heating the samples for 5 min at 100°C. Electrophoresis was performed using the Bio-Rad mini-Protean II system followed by staining with Coomassie blue. Molecular mass markers (Amersham Pharmacia Biotech, Uppsala, Sweden) were run on the same gels for determination of molecular masses of the proteins.
2.4 Cloning of protein antigens; amidase and aminopeptidase
Chromosomal DNA was prepared from the clinical isolate, P. aeruginosa 385 . The polymerase chain reaction (PCR) product and plasmid DNA were digested using the restriction endonucleases, Bam HI and Hin DIII as per the manufacturer's instructions. The oligonucleotides used for PCR amplification of the amiE gene from P. aeruginosa chromosomal DNA were: forward, 5′-GGCGGATCCCGTCACGGCGATATTTCCAGC-3′; reverse, 5′-GGCAAGCTTGGCCTCCTTCTCCAGTCCCTC-3′, containing engineered restriction sites of Bam HI and Hin DIII, respectively.
The oligonucleotides used for PCR amplification of the Pa 3247 gene from P. aeruginosa chromosomal DNA were: forward, 5′-GGCGGATCCCGCGCAGAACTCAACCAGGGCCTG-3′; reverse, 5′-GGCCTCGAGGGGCAGCTCGCAGGCGTAGAAGGC-3′, and containing engineered restriction sites for Bam HI and Hin DIII, respectively.
Amplification was carried out using 1 ng Pa385 DNA, 5 pmol primer (each), 10 µM dNTP (each) and 1 U DNA polymerase (Qiagen, Hilden, Germany) in a 20 µl volume. Conditions for amplification comprised 30 cycles of denaturation at 94°C, annealing at 50°C and extension at 72°C using a Corbett FTS 4000 capillary thermal cycle sequencer (Corbett Research, Australia).
The amplification products were digested using the restriction enzymes Bam HI and Hin DIII and visualised on agarose gel electrophoresis. The digested products were subsequently ligated into a Bam HI and Hin DIII-digested pQE30A Type IV high-expression vector (Qiagen). Ligated DNA was transformed into Escherichia coli JM109 by the CaCl2 method .
2.5 Expression of the recombinant amiE and Pa 3247 gene products
E. coli JM109 pQE30. amiE and E. coli JM109 pQE30. Pa 3247 were grown to an optical density of 0.6 at 600 nm. Uninduced samples were taken prior to the addition of isopropyl thiogalactose (Promega, Madison, WI, USA) to give a final concentration of 1 mM. The bacteria were incubated for a further 4 h and harvested by centrifugation at 4000×g for 10 min. Cells were resuspended in reducing buffer (62.5 mM Tris [pH 6.8], 10% [v/v] glycerol, 2% [w/v] SDS, 5% [v/v] β-mercaptoethanol, 1.2×10−3% [w/v] bromophenol blue) and were heated to 100°C for 10 min. The lysates were then centrifuged again at 4000×g for 10 min and the supernatants used for purification of recombinant amidase and aminopeptidase.
2.6 Purification of recombinant amidase and aminopeptidase
Recombinant amidase and aminopeptidase were purified using Ni–NTA affinity chromatography under non-denaturing conditions as per manufacturer's instructions (Qiagen).
Specific pathogen-free (SPF) Dark agouti (DA) male rats aged between 8 and 10 weeks were maintained and prepared for immunisation as previously described . The procedure for mucosal immunisation and challenge have been described previously [6,13]. Antigens for mucosal immunisation were prepared by emulsification of 100 µg of protein per 500 µl in a 1:1 ratio of phosphate-buffered saline (PBS) with incomplete Freund's adjuvant (IFA; Sigma). A total of 10 µg of protein in 50 µl was administered via intestinal Peyer's patch (IPP) inoculation to each experimental animal in the immunisation group with the exception of amidase and KatE-immunised rats, where a reduced dose of 5 µg of protein was delivered IPP due to protein availability. On day 14 post-IPP, animal groups were intra-tracheally (IT) boosted with either 10 µg acyl carrier protein (ACP), 2 µg KatE or 5 µg for all other proteins, in 50 µl of PBS as previously described . Non-immune rats received 50 µl of PBS only.
For parenteral immunisation, animals were sedated with halothane to facilitate immunisation. The immunisation protein was prepared by emulsification of 20 µg of protein per 100 µl in a 1:1 mixture of IFA (Difco Laboratories, Detroit, MI, USA) and PBS. The inoculum was injected subcutaneously into the neck scruff of each rat using a 23-gauge needle. On day 14 post primary immunisation, animals received a boost with the same antigen dosage and inoculation procedure.
2.8 Live bacterial challenge, bacterial clearance and cell counts
Pulmonary challenge with live P. aeruginosa 385 (5×108 colony forming units (CFU)) was performed on day 21 post-IPP or primary parenteral immunisation. Pulmonary challenge, bacterial clearance and cell counts were conducted as previously described .
2.9 Statistical analysis
All data are expressed as the mean±standard error of the mean. Statistical significance between groups was evaluated using either an independent t test or one-way analysis of variance with a test of equality of variance on log10 transformed data. All analyses were carried out using WINKS 4.5 statistics programme by Texasoft.
3.1 Purification and identification of protein antigens
Sufficient protein was purified from P. aeruginosa to carry out a study of immune-specific responses to six proteins of molecular masses equivalent to 13, 16, 20, 40, 45 and 80 kDA, as shown in Fig. 1. Protein purity was assessed by SDS–PAGE, which suggested homogeneous protein preparations. Purity was further confirmed by N-terminal amino acid sequencing. Contamination by LPS in protein preparations showed no detectable levels of endotoxin as assessed by the E-toxate kit Limulus assay (detection limit of assay was 0.015 endotoxin units ml−1).
SDS–PAGE analysis of P. aeruginosa antigens used in this study. Samples were run on a 4–15% gradient polyacrylamide gel and Coomassie-stained. Lanes: 1, molecular mass standards (values on left are in kDa); 3, Pa13; 4, azurin; 5, ACP; 6, amidase; 7, aminopeptidase; 8, KatE.
Following N-terminal amino acid sequencing, an electronic database search of GenBank sequences using ‘Entrez’ was conducted in order to identify the proteins. The 13-kDa protein yielded a sequence of AETIVNTTKA. This 10-residue sequence had no P. aeruginosa match, suggesting that this protein either may be unique to Pa 385, may encode for a variable N-terminus or may be a novel protein. It has been assigned as Pa 13. The 16-kDa sequence was determined to be AEESVDIQENDQMQFNTNAI, which gave an 85% identity in a 20 amino acid overlap with azurin from P. aeruginosa, accession number X07317. The 20-kDa sequence was determined to be STIEERVKKIVAEQL, which gave a 100% identity in a 15 amino acid overlap with an ACP from P. aeruginosa, accession number U91631. The 40-kDa sequence was determined to be MRHGDISSSNDTVG which gave a 100% identity in a 13 amino acid overlap with aliphatic amidase from P. aeruginosa, accession number M27612. The 45-kDa sequence was determined to be MRAELNQGLIDFLKA, which gave a 100% identity in a 15 amino acid overlap with aspartyl aminopeptidase from P. aeruginosa, accession number A83240. The 80-kDa sequencing revealed identity with KatE from P. aeruginosa, accession number AE004642.
3.2 Mucosal immunisation with native P. aeruginosa antigens enhances pulmonary clearance
A mucosal method of immunisation was employed where rats were administered single antigens to the Peyer's patches followed by a lung boost. All experimental animals received a pulmonary challenge with live bacteria of a homologous strain for 4 h on day 21 post-IPP immunisation. This model has been used to evaluate the protective efficacy of the antigens. Bacterial recovery obtained in immunised and non-immunised rats is shown in Table 1. Results demonstrate significant differences in the level of clearance in the bronchoalveolar lavage (BAL) between groups for Pa 13 (P<0.05), amidase (P<0.01), aminopeptidase (P<0.01) and KatE (P<0.01) immunised rats compared with their non-immunised controls. Clearance in the lung homogenate corresponded to clearance in the BAL, albeit at a slightly lower level of significance, for Pa 13 (P<0.05), amidase (P<0.05), aminopeptidase (P<0.05) and KatE (P<0.05) immunised rats compared with non-immunised rats.
Live bacteria recovered from BAL fluid and lung homogenate of rats immunised with native protein antigens and challenged post-immunisation with live P. aeruginosa
↵All animals received a primary IPP immunisation with 10 µg protein, except amidase and KatE groups, which received 5 µg doses. Upon boosting, the ACP group received a 10 µg dose, KatE was 2 µg, and all others received 5 µg doses each.
↵The values are the means±the standard error of n=4–5 rats per group.
↵Recovery values are percentage comparisons with the non-immune rats.
↵Significantly different from non-immune group at P<0.05.
↵Significantly different from non-immune group at P<0.01.
Azurin and ACP immunisation did not enhance clearance at the P<0.5 level. ACP-immunised rats showed some enhancement of clearance but this was non-significant in both BAL and lung homogenate. Azurin immunisation resulted in no significant difference in the level of bacterial clearance from the BAL but a significantly greater recovery of bacteria from the lung homogenate (P<0.05) (Table 1).
3.3 Mucosal immunisation using recombinant amidase and aminopeptidase maintains protection
Studies were conducted to determine the protective capacity of amidase and aminopeptidase as recombinant proteins. Mucosal immunisation using recombinant amidase and aminopeptidase demonstrated that the immunogenicity of these antigens was maintained. Bacterial recovery obtained in immunised and non-immunised rats is shown in Table 2. Results demonstrate significant differences in the level of clearance in both the BAL and lung homogenate between groups for amidase and aminopeptidase immunised rats compared with their non-immunised controls (P<0.05). The level of clearance for both these recombinant antigens is commensurate with that found for the native protein antigens.
Live bacteria recovered from BAL fluid and lung homogenate of rats mucosally immunised with recombinant amidase or aminopeptidase
↵The CFU (log10 transformed) values shown are means±the standard error for n=4–5 rats per group.
↵Recovery values are percentage comparisons with the non-immune rats.
↵Significantly different from the non-immune group at P<0.05.
3.4 Pulmonary clearance is also achieved following parenteral immunisation with recombinant amidase and recombinant aminopeptidase
Recombinant forms of amidase and aminopeptidase were tested to determine their effectiveness when delivered subcutaneously. Results for rats parenterally immunised with recombinant amidase demonstrate no significant difference in the level of clearance in either the BAL or lung homogenate at the P<0.05 level, although reduced bacterial numbers were recovered compared with non-immunised animals. However, rats parenterally immunised with recombinant aminopeptidase demonstrate a significant difference in the level of clearance in both the BAL and lung homogenate at the P<0.05 level, commensurate with mucosal delivery (Table 3).
Live bacteria recovered from BAL fluid and lung homogenate of rats parenterally immunised with recombinant protein antigens and challenged with live P. aeruginosa
↵The values shown are means±the standard errors for n=5 rats per group.
↵Recovery values are percentage comparisons with the non-immune rats.
↵Significantly different from the non-immune group at P<0.05.
Cellular and secreted products of P. aeruginosa contain numerous antigens with different immunogenicity. Enhanced bacterial clearance has been shown following mucosal immunisation with the cytosolic protein KatA . This study reports an assessment as vaccine candidates of an additional six antigens similarly purified from P. aeruginosa. Results show significant clearance of live bacteria with four of the six antigens following mucosal immunisation and assessment in the rat model of acute respiratory infection. Pa 13, amidase, aminopeptidase and KatE all demonstrate the ability to enhance bacterial clearance, thereby suggesting their potential protective capacity. Further characterisation of amidase and aminopeptidase demonstrates that this protective capacity is maintained in the recombinant form when these antigens are delivered by the mucosal route. For amidase, some loss of immunogenicity is apparent when this antigen is delivered subcutaneously.
The identification of these proteins has enabled their cellular location to be proposed and all are reported to be periplasmic proteins, with the exception of Pa 13, which awaits further characterisation. All proteins represent previously untested antigens for P. aeruginosa immunisation.
Traditionally, integral membrane structures have been tested as vaccine antigens. Periplasmic proteins were considered unable to be recognised by the immune system and have thus not received the same attention in immunological studies. More recently, proteins which are not embedded in the outer membrane have demonstrated immunogenicity [8,10,11]. The periplasm of P. aeruginosa possesses a variety of macromolecules according to the environmental status. The periplasm contains enzymes, trafficking proteins, secreted materials and newly synthesised outer membrane proteins. Thus, periplasmic proteins may become associated with the bacterial membrane through a number of mechanisms including transporter systems, autolysis and micro-vesicles [10,14–16]. Demonstration here of the bacterial clearance-enhancing properties of identified periplasmic antigens suggests that the periplasm may represent a rich source of vaccine antigens.
One advantage of choosing periplasmic proteins is their greater solubility and thus ease of purification. More importantly, the essential function of many of these proteins increases the likelihood of conservation across serotypes. The protective periplasmic protein amidase has been identified in this study as the product of the amiE gene . P. aeruginosa uses amide as both a carbon and nitrogen source. Aminopeptidase, also a protective periplasmic protein, has been identified as the product of the Pa 3247 gene, belonging to the peptidase family, M18. KatE is a catalase, the product of the katE gene, used by P. aeruginosa to protect cells from the toxic effects of aerobically produced hydrogen peroxide. The function of the remaining protective antigen, Pa 13, awaits further study.
The degree of protection afforded by these new antigens can be compared with that of killed whole cell vaccination, which has also been tested in the rat model of mucosal immunisation with acute respiratory challenge presented here . Whole cells, with their array of surface-exposed antigens and the adjuvant effect of LPS, have the potential to provide greater immune responses. However, there is a similar level of efficacy evident following immunisation with these protein antigens.
Proteins derived from the cytosol may in fact be more immunogenic than some outer-membrane proteins. In CF, substantial bacterial burdens can occur over many years despite the pressures of antibiotic challenge. It may be that selective pressures for persistence of the bacterium have resulted in a lowered immunogenicity of continuously exposed surface epitopes.
Protein antigens, with the possibility of determining protective epitopes and thus the manufacture of synthetic peptides, are of great interest as potential vaccines. They provide the possibility of inexpensive large-scale purification and low-priced vaccines. This study reports on the potential of four new protein antigens as vaccine candidates against P. aeruginosa respiratory infection. The ability of these antigens to protect against heterologous challenge will be a focus of future research. In addition, the potential of a composite vaccine including the most immunogenic proteins should be investigated.
This research was supported by a grant from the National Health and Medical Research Council (Grant No. 940492). We are grateful to Clare Batum, Nancy Fisher, Melissa Musicka, Catherine Delahunty, Corrina Oszko and Amanda McCue for expert technical assistance.