Catalase immunization from Pseudomonas aeruginosa enhances bacterial clearance in the rat lung Linda D. Thomas a , Margaret L. Dunkley b , Ryka Moore a , Simone Reynolds a , David A. Bastin a , Jennelle M. Kyd a, *, Allan W. Cripps a a Gadi Research Centre, Division of Science and Design, University of Canberra, Canberra, ACT 2601, Australia b Discipline of Pathology, Faculty of Medicine and Health Sciences, University of Newcastle, Callaghan, NSW 2308, Australia Received 24 December 1999; received in revised form 1 February 2000; accepted 29 March 2000 Abstract Pseudomonas aeruginosa is a common cause of infection in immunocompromised patients and is the major contributor to morbidity in individuals with cystic fibrosis (CF). The antibiotic resistance shown by this pathogen and morbidity in patients with chronic infection has encouraged investigations into the development of a vaccine. This study reports the purification of a 60 kDa protein, isolated from a mucoid strain of P. aeruginosa, identified by amino acid sequence analysis as the catalase protein (KatA). A rat model of acute P. aeruginosa respiratory infection was used to investigate the immunogenicity of KatA and determine the potential of mucosal immunization with KatA to protect against infection. Immunization regimens compared a single intra-Peyer’s patch (IPP) immunization with an IPP primary inoculation followed by an intratracheal boost to the lungs. Mucosal immunization with KatA resulted in significant pulmonary clearance of both homologous p < 0:001and heterologous p < 0:05strains of P. aeruginosa. Both immunization regimens enhanced bacterial clearance, increased the rate of recruitment of phagocytes to the bronchoalveoli and induced KatA-specific antibody. However, the regimen that included a boost induced a more eective immune response that also resulted in better clearance of P. aeruginosa from the lungs. Mucosal immunization induced KatA- specific antibodies in the serum and the bronchoalveolar lavage, and KatA-specific lymphocyte proliferation in vitro in cells isolated from the mesenteric lymph nodes of immunized rats. The data presented suggests that KatA has the potential to aord a protective immune response against pulmonary infection by P. aeruginosa 7 2000 Elsevier Science Ltd. All rights reserved. Keywords: Catalase; KatA; Pseudomonas; Mucosal 1. Introduction Pseudomonas aeruginosa is an environmentally ubi- quitous, extracellular gram-negative bacterium. It is an opportunistic pathogen that is capable of causing fatal systemic disease and is responsible for many hospital- acquired infections, especially in immunocompromised patients such as cancer or burns patients [1,2]. More commonly, however, it is associated with chronic res- piratory infections, especially in patients with cystic fibrosis (CF) where it is the major cause of morbidity [3,4] and in patients with chronic obstructive pulmon- ary disease. Therefore, an eective vaccine against this pathogen must be capable of inducing protection at the respiratory mucosa without enhancing damage evoked by an inflammatory process. There is a significant level of antibiotic resistance among strains of P. aeruginosa [5] prompting novel approaches to therapy and particularly focussing research eorts towards immunization [6]. To date, im- munization has involved either whole killed bacterial cell vaccines that contain lipopolysaccharide (LPS) and Vaccine 19 (2001) 348–357 0264-410X/00/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved. PII: S0264-410X(00)00146-8 www.elsevier.com/locate/vaccine * Corresponding author. Tel.: +61-2-6201-2160; fax: +61-2-6201- 2461. E-mail address: [email protected] (J.M. Kyd).
Transcript
Catalase immunization from Pseudomonas aeruginosa enhancesbacterial clearance in the rat lung
Linda D. Thomasa, Margaret L. Dunkleyb, Ryka Moorea, Simone Reynoldsa,David A. Bastina, Jennelle M. Kyda,*, Allan W. Crippsa
aGadi Research Centre, Division of Science and Design, University of Canberra, Canberra, ACT 2601, AustraliabDiscipline of Pathology, Faculty of Medicine and Health Sciences, University of Newcastle, Callaghan, NSW 2308, Australia
Received 24 December 1999; received in revised form 1 February 2000; accepted 29 March 2000
Abstract
Pseudomonas aeruginosa is a common cause of infection in immunocompromised patients and is the major contributor tomorbidity in individuals with cystic ®brosis (CF). The antibiotic resistance shown by this pathogen and morbidity in patientswith chronic infection has encouraged investigations into the development of a vaccine. This study reports the puri®cation of a60 kDa protein, isolated from a mucoid strain of P. aeruginosa, identi®ed by amino acid sequence analysis as the catalase
protein (KatA). A rat model of acute P. aeruginosa respiratory infection was used to investigate the immunogenicity of KatAand determine the potential of mucosal immunization with KatA to protect against infection. Immunization regimens compareda single intra-Peyer's patch (IPP) immunization with an IPP primary inoculation followed by an intratracheal boost to the lungs.
Mucosal immunization with KatA resulted in signi®cant pulmonary clearance of both homologous �p < 0:001� and heterologous�p < 0:05� strains of P. aeruginosa. Both immunization regimens enhanced bacterial clearance, increased the rate of recruitmentof phagocytes to the bronchoalveoli and induced KatA-speci®c antibody. However, the regimen that included a boost induced a
more e�ective immune response that also resulted in better clearance of P. aeruginosa from the lungs. Mucosal immunizationinduced KatA- speci®c antibodies in the serum and the bronchoalveolar lavage, and KatA-speci®c lymphocyte proliferation invitro in cells isolated from the mesenteric lymph nodes of immunized rats. The data presented suggests that KatA has thepotential to a�ord a protective immune response against pulmonary infection by P. aeruginosa 7 2000 Elsevier Science Ltd. All
rights reserved.
Keywords: Catalase; KatA; Pseudomonas; Mucosal
1. Introduction
Pseudomonas aeruginosa is an environmentally ubi-quitous, extracellular gram-negative bacterium. It is anopportunistic pathogen that is capable of causing fatalsystemic disease and is responsible for many hospital-acquired infections, especially in immunocompromisedpatients such as cancer or burns patients [1,2]. Morecommonly, however, it is associated with chronic res-
piratory infections, especially in patients with cystic
®brosis (CF) where it is the major cause of morbidity
[3,4] and in patients with chronic obstructive pulmon-
ary disease. Therefore, an e�ective vaccine against this
pathogen must be capable of inducing protection at
the respiratory mucosa without enhancing damage
evoked by an in¯ammatory process.
There is a signi®cant level of antibiotic resistance
among strains of P. aeruginosa [5] prompting novel
approaches to therapy and particularly focussing
research e�orts towards immunization [6]. To date, im-
munization has involved either whole killed bacterial
cell vaccines that contain lipopolysaccharide (LPS) and
Vaccine 19 (2001) 348±357
0264-410X/00/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved.
result in a strong immune response, or various cell-as-sociated and secreted antigens of the bacterium [5,7].
Systemic immunization with puri®ed proteins [8] andintratracheal immunization with live P. aeruginosa [9]have both shown protection against acute respiratoryinfection in an animal model. Oral immunization withkilled P. aeruginosa has also demonstrated protectiveproperties in a rat model of acute respiratory infection[10], yet this was not obtained with live P. aeruginosa[11]. However, what is most required in the develop-ment of a vaccine against P. aeruginosa for cystic®brosis patients is a well characterized immunogenthat does not invoke detrimental e�ects. As such, puri-®ed protein candidate antigens o�er signi®cant poten-tial.
Various approaches have been taken in the develop-ment of a vaccine. Most concentrate on surface struc-tures of Pseudomonas and include heptavalent LPS,O-polysaccharide-toxin A conjugates, pili and ¯agella[12±15]. Di�culty with any of these antigens as vaccinecandidates is their genetic variability. The e�cacy ofthese antigens as vaccine preparations has been basedon their ability to elicit protective antibody levels. Themucosal model for pulmonary defence supports the hy-pothesis that cellular immunity is also a major contri-butor to the clearance of P. aeruginosa from the lung[16].
An oral route of immunization is a desirable andpractical approach for a new human vaccine. Com-pared with vaccine delivery by injection, oral vaccineso�er the possibility of more convenient immunizationstrategies, greater compliance and a more practicalmeans of implementing universal vaccination programsthroughout the world. The model currently employedinvolves intra-Peyer's patch (IPP) immunization fortargeting intestinal lymphoid tissue, which induces pro-tection both mucosally and systemically and is analo-gous to oral delivery of vaccine antigens [17]. Usingthis approach, this study examined the ability of a 60kDa protein from P. aeruginosa, identi®ed as the cata-lase, KatA, to provide protection in a rat model ofacute P. aeruginosa respiratory infection.
2. Materials and methods
2.1. Bacterial strains
Two mucoid P. aeruginosa isolates [10] used in thisstudy were strain 385, serotype 2 and strain 423, sero-type 6, phage type 7/44/68/73/F8/F10/109/119X188/1(typed by the Central Public Health Laboratory,London, UK). Both were originally isolated frompatients chronically infected with cystic ®brosis. Bac-terial stocks were stored at ÿ858C in nutrient broth
(Oxoid, Unipath, Basingstoke, Hampshire, England)supplemented with 10% glycerol (v/v).
2.2. Puri®cation of KatA
Bacteria for protein puri®cation and bacterial chal-lenge were grown overnight on nutrient agar (Oxoid)at 378C in 5% CO2. A crude protein preparation wasobtained using a Zwittergent extraction protocol [18].KatA was semi-puri®ed from the extract by usinganion exchange chromatography, conducted using theBio-scale Q5 column (Bio-Rad). The column was equi-librated with 20 mM Tris±HCl, pH 8.5 and the lyophi-lised crude protein extract was resuspended in theequilibration bu�er prior to injection onto the columnand elution with an increasing concentration of 500mM sodium chloride, 20 mM Tris±HCl, pH 8.5. Frac-tions containing KatA, as evidenced by SDS-PAGE,were pooled for further puri®cation by preparativepolyacrylamide gel electrophoresis using sodium dode-cyl sulfate (SDS) [19]. Tris and SDS were electrophor-esis grade reagents purchased from Bio-RadLaboratories, North Ryde, NSW, Australia, and allothers were laboratory grade reagents.
Preparative SDS-PAGE to purify KatA was per-formed with a Bio-Rad 491 Prep Cell as previouslydescribed [19] using 20 ml of 10% polyacrylamide forthe resolving gel and 10 ml of 4% polyacrylamide forthe stacking gel. These conditions separated KatAfrom other proteins and lipopolysaccharide. KatA iso-lated under these conditions contained SDS, whichwas subsequently removed [20].
The presence of lipopolysaccharide (LPS) wasassessed by both silver staining of SDS-PAGE mini-gels and assaying with the E-TOXATE Limulus lysatetest (Sigma, St. Louis, MO).
2.3. Analytical SDS-PAGE
SDS-PAGE was used to analyse fractions for thepresence of KatA [21]. A sample was added to anequal volume of SDS reducing bu�er containing SDSand dithiothreitol (DTT) and boiled for 5 min. Thesame protein sample was also prepared by incubationat RT (238C). Electrophoresis was performed with 10±20% gradient polyacrylamide gel and the Bio-RadMini-Protean II system, followed by silver staining.Molecular weight markers (Amersham Pharmacia Bio-tech, Uppsak, Sweden) were run on the same mini-gelsfor determination of the molecular masses of proteins.
2.4. Protein concentration determination
Protein concentration was determined with thePierce Micro BCA protein assay reagent and the Piercealbumin standard (Rockford, IL, USA).
L.D. Thomas et al. / Vaccine 19 (2001) 348±357 349
2.5. Sequencing of KatA
Both N-terminal and internal amino acid sequenceanalyses were carried out at the BiomolecularResource Facility, Centre for Molecular Structure andFunction, Australian National University. Sequencingwas carried out using the SDS-PAGE compatible S-2-carboxamidoethylation method. Following sequenceanalysis, the data obtained was subjected to BLASTsearch (Basic Local Alignment Search Tool), theNational Centre for Biotechnology Information(Bethesda, MD, USA).
2.6. Immunization
Animal experiments were performed with theapproval of the University of Canberra and the Uni-versity of Newcastle Animal Care and Ethics Commit-tees. Speci®c pathogen free (SPF) dark agouti (DA)male rats aged 8±10 weeks were maintained under SPFconditions until the start of the experiment, andremoved from behind the SPF barrier for immuniz-ations and the ®nal live bacterial challenge. At allother times, the animals remained under SPF con-ditions. The procedure for immunization and challengehas been described previously [10,22]. The immuniz-ation protein was prepared by emulsi®cation of 100 mgof protein per 500 ml in a 1:1 mixture of incompleteFreund's adjuvant (Difco Laboratories, Detroit,Mich.) and PBS, and a total of 10 mg (50 ml) of proteinwas administered via intra Peyer's patch (IPP) inocu-lation to each experimental animal in the immuniz-ation group.
Rats receiving an IT boost were initially immunizedon day 0 as above. On day 14 post-IPP immunization,they were sedated with halothane and 10 mg of theprotein, KatA, at a concentration of 200 mg protein/mlPBS, was introduced into the lungs via a cannula anddispersed with two 5-ml volumes of air. For those ratsundergoing heterologous challenge, a lower IT boostdose of KatA was given at a concentration of 100 mg/ml PBS. Nonimmune rats received 50 ml of PBS only.
2.7. Bacterial challenge
Bacteria were prepared by overnight culture asdescribed above and resuspended in PBS. Bacteriawere washed three times in PBS, using centrifugation.The concentration of the inoculum was estimated bydetermining the optical density at 405 nm and calcu-lated using a previously prepared regression curve.Concentration was con®rmed by counting colonyforming units (CFU) after overnight plating of serialdilutions of the inoculum.
Pulmonary challenge with live bacteria was per-formed on day 14 (single IPP immunization) or day 21
(IPP plus IT immunization). The animals were sedatedwith halothane and 50 ml of a bolus inoculum of 1010
CFU/ml live P. aeruginosa in PBS was introduced intothe lungs via a cannula (as per IT boost) and dispersedby two 5-ml volumes of air. Animals were killed by anoverdose of pentobarbital sodium administered byintraperitoneal injection 4 h after lung inoculation.Blood was collected by heart puncture for preparationof aliquots of serum which were stored at ÿ208C. Thelungs were lavaged with ®ve 2-ml volumes of PBS viathe trachea, which had been exposed through an in-cision in the neck, and the pooled bronchoalveolarlavage (BAL) ¯uid was assessed for bacteria by platingof serial dilutions of the washings for CFU determi-nation. Following lavage, the lungs were removed andhomogenized in 10 ml of PBS for plating serial di-lutions (20 ml in a 10-fold series) onto nutrient agar forCFU determination.
2.8. Preparation of BAL for cytospin, cell counts andELISA
Cytospin slides were prepared to determine percen-tages of polymorphonuclear neutrophils (PMNs),macrophages, and other cells present in the BAL ¯uid.A 100 ml aliquot of BAL ¯uid was spun for 10 min at4.5 �g (using a Shandon Cytospin 2 by ShandonSouthern Products, Cheshire, UK), onto a microscopeslide. The slides were ®xed and stained in Di� Quick(Veterinary Medical Surgical Supply, Maryville, NSW,Australia), and percentages were determined fromthree di�erential cell counts on each slide. Mean per-centages2the standard error were calculated from thecollected group data.
BAL ¯uid was centrifuged at 200 �g for 10 minwith a Beckman CPR bench top centrifuge. The super-natant was removed, and aliquots were stored atÿ208C until required for enzyme-linked immunosor-bent assay (ELISA) analysis. The pellet was resus-pended in a known volume of PBS, and the totalnumber of cells present in the BAL ¯uid was deter-mined with a haemocytometer after staining withmethylene blue.
2.9. Western blotting
Proteins in a lysate from P. aeruginosa were separ-ated by SDS-PAGE on a 10±20% gradient gel andtransferred to a nitrocellulose membrane (Bio-Rad)using a semi-dry transfer apparatus (Pharmacia). Afterblocking in 1% (w/v) skim milk in Tris bu�ered saline(TBS-skim milk), the membrane was incubated with a1/1000 dilution of pooled sera (from KatA immunizedrats) in TBS-skim milk for 90 min at room tempera-ture. The membrane was washed with TBS and 0.05%Tween 20 and incubated in a 1/1000 dilution of horse
L.D. Thomas et al. / Vaccine 19 (2001) 348±357350
radish peroxidase (HRP) conjugated anti-rat IgG(Nordic Immunological Laboratories, Tilberg, Nether-lands). The blot was developed using HRP develop-ment reagent (Bio-Rad) and scanned using a GS570densitometer (Bio-Rad).
2.10. Flow cytometry
P. aeruginosa was grown to mid logarithmic phasein nutrient broth and harvested by centrifugation at10,000 �g for 10 min at 48C. The bacteria was incu-bated at 378C for 1 h with a 1:50 dilution (in PBS)with either nonimmune serum, KatA immunizedserum from challenge experiments or PBS. The cellswere centrifuged, the supernatant removed and thebacteria resuspended in 200 ml of ¯uorescein isothio-cyanate conjugated anti-rat IgG (Sigma) diluted 1:50in PBS. Following incubation for 30 min at 378C, 1.8ml of PBS was added and the bacterial cells were ana-lysed by ¯ow cytometry (Coulter XL-MCL, CoulterCorporation, Miami, Florida). A total of 20,000 cellswere counted and data were acquired in the instrumentstatus of logarithmic mode for forward scatter, sidescatter and ¯uorescence.
2.11. Anti-KatA antibody ELISAs
Polysorb microtiter wells (Nunc, Roskilde, Den-mark) were coated with 1 mg/ml puri®ed KatA in coat-ing bu�er (15 mM Na2CO3, 35 mM NaHCO3 [pH9.6]) overnight at 48C as described in Ref. [19]. Theplates were washed in PBS containing 0.05% (w/v)Tween 20 and blocked with 5% (w/v) skim milk inPBS-0.05% Tween 20. Serum (1/25 to 1/100 for nonimmune; 1/25 to 1/4000 for immune) or BAL ¯uid (1/2to 1/16) samples serially diluted in blocking bu�er,were incubated at room temperature for 90 min. Afterremoval of the samples by washing, horseradish per-oxidase-conjugated goat anti-rat immunoglobulindiluted in blocking bu�er was added and incubated atroom temperature for 90 min. Conjugated immunoglo-bulins used were IgG (1/2000), IgA (1/1000) and IgM(1/1000) Fc speci®c (Nordic). After washing the wellswere developed with tetramethylbenzidine (Fluka,Buchs, Switzerland) in phosphate-citrate bu�er (pH 5)containing 0.05% (v/v) H2O2 and the reaction stoppedwith 0.5 M H2SO4. Plates were read at 450 nm on aBio-Rad multiplate reader (model 3550). Mean ELISAtiters were calculated by determining the reciprocal ofthe serum or BAL ¯uid dilution that gave an opticaldensity reading of 0.4±0.9 and multiplying by the di-lution factor.
2.12. Anti-KatA lymphocyte assay
The antigen-speci®c lymphocyte assay was per-
formed essentially as previously described [19,23].Brie¯y, lymphocytes were obtained from the mesen-teric lymph nodes (MLN) and were washed in cold,sterile bu�er prepared with PBS containing calciumand magnesium, supplemented with 5% (v/v) foetalcalf serum (heat inactivated at 578C for 30 min), 100U/ml penicillin, 100 mg/ml streptomycin and 0.25 mg/ml amphotericin B. Viable cells were counted by try-pan blue exclusion with a haemocytometer and resus-pended to a ®nal concentration of 106 cells per ml inMulticel RPMI 1640 (Sytosystem, Castle Hill, NSW,Australia) containing 0.01 M HEPES, pH 7.2, 0.05mM b-mercaptoethanol, 2 mM L-glutamine, 5% (v/v)foetal calf serum and penicillin-streptomycin-ampho-tericin B. Sterile ®ltered KatA was added to the cul-ture medium in a 10-fold dilution series to give ®nalconcentrations ranging from 0.1 to 10 mg/ml. Lympho-cyte proliferation was estimated by determining [3H]thymidine (Amersham Pharmacia Biotech, Uppsala,Sweden) incorporation for the last 8 h of a 4-day cul-ture. Results were calculated by subtraction of thebackground radioactivity from the geometric means oftriplicate wells and then from the geometric means2the standard error of the entire treatment group.
2.13. Statistical analysis
Data were expressed as the mean and standard errorof the mean (SE). Pulmonary clearance, total number
Fig. 1. SDS-PAGE analysis of KatA. The KatA was heated at di�er-
ent temperatures for 5 min. Analysis was conducted using a 10±20%
gradient gel that was visualised by silver staining. Approximately 2
mg of puri®ed KatA was loaded in each lane: lane 1, molecular mass
markers; lane 2, 238C (RT); lane 3, 378C; lane 4, 568C; lane 5, 708Cand lane 6, 1008C.
L.D. Thomas et al. / Vaccine 19 (2001) 348±357 351
of phagocytic cells, di�erential cell counts, antibodydata and lymphocyte proliferation in immune andnonimmune animals were compared using one-wayanalysis of variance with Bonforoni's assumption ofequal variance. Linear correlation between variableswas determined using Pearson's correlation coe�cient.Statistics were analysed using the WINKS 4.5 statisticsprogram by Texasoft.
3. Results
3.1. Puri®cation of a 60 kDa protein and itsidenti®cation as KatA
A protein of approximately 60 kDa was isolatedfrom P. aeruginosa strain 385, and is shown in Fig. 1.Assessment of the protein for the presence of endo-toxin using the E-TOXATE assay found it to be freeof LPS contamination (detection limit of assay was0.015 endotoxin units/ml). The presence of LPS wasalso assessed by silver staining and none was detected(Fig. 1). This 60 kDa protein was found not to be aheat modi®able protein and migration of the proteinon SDS-PAGE was uniform for sample preparation inreducing bu�er at incubation temperatures rangingfrom RT to 1008C.
The N-terminal sequence of the 60 kDa proteinwas determined to be -EEKTPLT±A-VV-NA±. Thissequence gave a perfect match for six of seven resi-dues with the amino acid sequence of P. aeruginosacatalase isoenzyme A (KatA), accession number gi|2896139 (AF047025). This sequence was locatedbetween residues 2 and 8, from a total length of 482residues. Internal sequencing of this protein producedthe sequence LAHFDREVIPERR. These 13 residuesgave 100% match with a segment of P. aeruginosaKatA that was located between residues 41 and 53.These 13 residues produced high-scoring segmentpairs with an additional six catalases. A secondsequence derived from internal sequencing EAATYR-YNPFDL again gave a 100% match with P. aerugi-nosa and corresponding to residues from 268 to 279.These 12 amino acids formed a unique sequence pro-ducing no additional high-scoring segment pairs.Hence the 60 kDa protein was identi®ed as P. aerugi-nosa KatA.
The location of KatA on P. aeruginosa was deter-mined by ¯ow cytometry analysis of whole bacteriaand western blot assessment of bacterial fractions fordetection of KatA antisera binding. Fig. 2 shows theresults of the ¯ow cytometry assessment and indicatesthat KatA can be detected on the surface of P. aerugi-nosa. Fluorescence with either PBS or nonimmuneserum resulted in the same peak (peak 1, Fig. 2),whereas incubation with antiserum to KatA resulted ina signi®cant shift in the curve (peak 2). Western blotassessment of P. aeruginosa bacterial cell lysate showedbinding of KatA antiserum to a 60 kDa band (Fig. 3)which was absent on incubation with antisera fromunimmunized animals.
Fig. 3. Western blot analysis of KatA. A P. aeruginosa lysate was
transferred to a nitrocellulose membrane and probed with antisera
from KatA-immunized rats: lane 1, molecular mass markers; lane 2,
P. aeruginosa lysate demonstrating KatA recognition.
Fig. 2. Results of ¯ow cytometry with antiserum to KatA assayed
with P. aeruginosa. Vertical axis indicates the events detected by the
¯ow cytometer and the horizontal axis indicates the ¯uorescence.
Peak 1 marked NC represents P. aeruginosa assayed with either non-
immune serum or PBS; peak 2 is the result when P. aeruginosa was
assayed with antiserum to KatA.
L.D. Thomas et al. / Vaccine 19 (2001) 348±357352
3.2. E�ect of immunization on pulmonary clearance
Table 1 shows the bacterial clearance obtained inimmunized and control rats. Results demonstrate a sig-ni®cant di�erence in the level of clearance in both theBAL �p < 0:001� and lung homogenate �p < 0:001�between groups. A single IPP immunization withKatA resulted in a signi®cantly increased level of pul-monary clearance compared with nonimmune animalsin both the BAL and lung homogenate against P. aer-uginosa strain 385 �p < 0:05), the strain from whichKatA was isolated. Homologous bacterial clearance ofP. aeruginosa strain 385 was further enhanced in theBAL and lung homogenate when IPP immunizationwas followed by an IT boost at day 14 �p < 0:05).Increasing the concentration of KatA by four-fold to40 mg for both single IPP immunization or IPP immu-nization with an IT boost regimens did not furtherincrease the bacterial clearance (data not shown).
The ability of KatA to protect against other strainsof P. aeruginosa was tested by heterologous challengeusing strain 423. Animals were immunized and boostedas above with the exception of a lower boost dose of 5mg (previously 10 mg) of KatA. Table 2 shows signi®-cantly increased clearance of bacteria from both theBAL �p < 0:01� and lung homogenate �p < 0:05� ofimmunized animals against challenge with a di�erentP. aeruginosa serotype. This enhanced clearance wasmaintained against the heterologous challenge despitea reduced boost dosage.
3.3. Recruitment of phagocytic cells
BAL ¯uid was assessed for both the total number ofcells and di�erential cell populations (Table 3). There
was a signi®cant increase in the total number of cellsrecruited to the lungs in all the immunized groups,with the greatest number of recruited cells in the ani-mals given an IT boost �p < 0:01). Counts of thedi�erential cell populations in the BAL showed thatalong with an increase in the total number of cellsrecruited, there was also an increase in the proportionof PMNs in the immune BAL. There was a signi®cantcorrelation between bacterial clearance and recruitmentof phagocytic cells �r � ÿ0:74).
3.4. Anti-KatA antibody in serum and BAL ¯uid
Antibody responses following IPP immunizationwith KatA were assessed by ELISA. The data inTable 4 show increases in anti-KatA IgG, IgM andIgA in the immunized groups compared with the non-immune rats. Addition of an IT boost signi®cantly el-evated antibody in comparison with the single IPPimmunization regime, with the exception of BAL IgMand serum IgA. The increases in anti-KatA IgG andIgA in the BAL correlated with the increased bacterialclearance �r � ÿ0:83 and ÿ0.81, respectively).
3.5. Antigen-speci®c lymphocyte proliferation
Lymphocytes from the MLN of groups of ratsimmunized with KatA were cultured with KatA atthree di�erent concentrations to assess levels of pro-liferation (see Fig. 4). Antigen-speci®c proliferationwas observed in response to KatA in immunizedgroups at 1 mg/ml and 10 mg/ml concentrations�p < 0:01). Proliferation was greater (at the same sti-mulating concentrations) in the immunized groupgiven an IT boost. Crude P. aeruginosa extract, usedas a control to assess in vitro proliferation (data notshown) demonstrated some mitogenic activity but wassigni®cantly lower than the level of stimulation by
Table 2
Live bacteria recovered from BAL ¯uid or lung homogenate of rats
immunized with KatA and challenged postimmunization with live P.
aeruginosa strain 423
Rat group N Log10 CFU of P. aeruginosa at 4 h postchallengea
BAL ¯uid Lung homogenate
Nonimmune 5 7.6120.08 7.9120.09
IPP+ITb 6 6.3720.24c 7.4820.14d
a The values shown represent the mean2SEM for the BAL ¯uid
or lung homogenate from rats at 4 h postpulmonary challenge with
a heterologous strain of P. aeruginosa.b IPP immunization with 10 mg KatA on day 0, IT boost with 5 mg
KatA on day 14 and bacterial challenge on day 21 post-IPP.c p < 0:01 compared with nonimmune group.d p < 0:005 compared with nonimmune group.
Table 1
Live bacteria recovered from BAL ¯uid and lung homogenate of rats
immunized with KatA and challenged postimmunization with live P.
aeruginosa strain 385
Group N Log10 CFU of P. aeruginosa at 4 h post challengea
BAL ¯uid Lung homogenate
Nonimmune 10 7.8820.11 8.8420.12
IPPb 6 6.9520.07d 8.4420.09d
IPP+ITc 5 6.320.08e 7.620.08e
a The values shown represent the mean2SEM for the BAL ¯uid
or lung homogenate from rats at 4 h postpulmonary challenge with
homologous live P. aeruginosa.b IPP immunization only with 10 mg KatA and bacterial challenge
on day 14 post-IPP.c IPP immunization with 10 mg KatA on day 0, IT boost with 10
mg KatA on day 14 and bacterial challenge on day 21 post-IPP.d p < 0:001 compared with nonimmune group.e p < 0:001 compared with nonimmune group and p < 0:05 com-
pared with IPP group.
L.D. Thomas et al. / Vaccine 19 (2001) 348±357 353
KatA at all concentrations. The crude extract also con-tained KatA, although at a low concentration, andthus some response by cells from KatA-immunizedrats was expected.
4. Discussion
In this study, the vaccine potential of the P. aerugi-nosa catalase, KatA, was examined in a rat model ofacute P. aeruginosa respiratory infection in which vac-cine e�cacy is seen as enhanced clearance. Mucosalimmunization regimens compared single intestinal im-munization against primary intestinal immunizationcoupled with a pulmonary boost. In addition, antigen-speci®c immune responses were assessed.
Results demonstrate the value of KatA as a vaccinecandidate. Signi®cantly greater clearance of live bac-teria in immune rats compared with nonimmune ratswas noted for both homologous (strain 385) and heter-ologous (strain 423) P. aeruginosa challenges. A singleintestinal immunization appeared to induce a subopti-mal immune response, which, while signi®cant,
resulted in a lower level of clearance. The addition ofan intratracheal boost to the immunization regimenincreased the immune response to KatA and furtherenhanced bacterial clearance. The amount of antigenrequired to boost the response was lower than initiallyanticipated, as determined from the e�ectiveness of im-munization to clear the heterologous P. aeruginosastrain 423.
The e�ectiveness of mucosal immunization to pro-tect against pulmonary challenge with P. aeruginosahas been demonstrated previously [10]. Presentation ofantigen to Peyer's patches induces antigen-speci®c Band T cells which are then disseminated to othermucosal sites, including the respiratory tract [24]. Theprevious rat studies, using whole killed P. aeruginosacells as immunogens, have shown that mucosal immu-nization with killed P. aeruginosa invokes a responsethat clears pulmonary infection, whereas in nonim-mune animals, bacterial numbers increase until deathat 10±12 h [10].
Clearance of most bacteria from the lungs involvesrecruitment of phagocytic cells [25,26]. Along withmacrophages, the recruitment and activation of PMNs
Table 3
Phagocytic cell counts in BAL ¯uid at 4 h postpulmonary challenge with live P. aeruginosa
Rat group N Mean di�erential cell count (105)2SEM Total no. of cells in BAL ¯uid (105).
Challenge strain PMNs Macrophages Other cells
P. aeruginosa 385
Nonimmune 8 31.721.9 3.420.9 3.521.4 39.4212.8
IPP 6 59.920.6 2.420.6 0.220.06 62.5216.8a
IPP+IT 4 303.724.3b 14.122.9b 11.523.0 329251.2c
P. aeruginosa 423
Nonimmune 5 93.320.8 3.620.6 2.120.9 99.2235.1
IPP+IT 6 226.721.3a 5.822.4 10.522.7 243216d
a Signi®cantly di�erent to nonimmune group at p < 0:05.b Signi®cantly di�erent to both nonimmune and single immunization groups at p < 0:05.c Signi®cantly di�erent to both nonimmune and single immunization groups at p < 0:01.d Signi®cantly di�erent to nonimmune group at p < 0:01.
Table 4
Comparison of KatA-speci®c antibody levels in serum and BAL ¯uid following immunization
Treatment Mean titer of antibody to KatA (ELISA titera)2SEM
a ELISA titer calculated as described in Section 2.b p < 0:05 compared with nonimmune group.c p < 0:05 compared with nonimmune and single IPP immunization groups.d ND: not detectable.
L.D. Thomas et al. / Vaccine 19 (2001) 348±357354
have been shown to contribute to pulmonary immu-nity following intestinal immunization against P. aeru-ginosa infection [27,28]. The data presented here ofPMN recruitment and correlation with enhanced clear-ance, are in agreement with these previous ®ndings. Inthis rat model, the percentage of PMNs in BAL ¯uidwas greater in the immunization groups at 4 h post-challenge with live bacteria (Table 3). Furthermore,the total number of cells recruited to the lungs at thistime was signi®cantly increased by immunization andcorrelated with clearance. Greater numbers of cellswere recruited to the lungs when IPP immunizationwas followed by a pulmonary (IT) boost. Immunizedanimals also cleared bacteria associated with the lungtissue indicating that immunization protected againstany localised invasion by P. aeruginosa. The import-ance of recruitment and activation of PMNs to thebronchoalveolar spaces in protection has also beendemonstrated in studies of P. aeruginosa lung infectionin susceptible and resistant strains of mice [29].
The role of anti-Pa antibodies in protection againstP. aeruginosa infection is yet to be clari®ed. Studies inpatients with cystic ®brosis suggest that a good prog-nosis may be determined by IgG subclass response toProtein H [30] and mucoid exopolysaccharide (MEP)antibodies [31]. However, there is also a suggestionthat the hyperimmune state and antibody subclass ofthe response may increase the severity of the disease[32±34]. Longitudinal studies have found that antibodytiters to MEP rise with disease progression but thatopsonic activity decreases as infection becomes estab-lished [35]. In this study, higher levels of antibodywere associated with greater rates of clearance. Itwould appear, therefore, that in our model of acuterespiratory infection, antibody levels provide a markerfor protection and may be important for e�ectiveimmunity. Passive transfer of immune serum [36] andIgG antibody [37] has been shown to confer protectionagainst P. aeruginosa challenge in the rat model,implying that antibody, and in particular IgG, may beimportant in protecting against infection.
Amino acid sequencing has identi®ed this protein asthe P. aeruginosa catalase, KatA. P. aeruginosa pos-sesses two heme-containing catalases [38]. KatAdetoxi®es hydrogen peroxide produced by the uni-valent reduction of dioxygen during aerobic metab-olism, while KatB acts as a virulence factor against therespiratory bursts of macrophages [38]. Catalases arewell documented as having these antioxidant proper-ties but their role as vaccine candidates has not beenevaluated. P. aeruginosa is a strict aerobe and pos-sesses catalase in order to neutralise potentially hazar-dous oxygen reduction products. KatA, the`housekeeping' or `isoenzyme A' form catalase, hasbeen reported as being located in both the cytoplasmand periplasm [38]; however, our study has detected
the protein in the media (suggesting it may be secreted)and on the bacterial surface. The estimated molecularmass of KatA isolated in this study was close to thepredicted molecular mass of 54 kDa, and with that ofthe major catalase (Cat4) of Streptomyces coelicolorMuÈ ller [39].
There is a high degree of homology between regionsof catalase proteins distributed throughout the phyla[40]. Comparison of amino acid sequences for KatArevealed signi®cant conservation of this catalasebetween P. aeruginosa and numerous other bacteriasuch as Haemophilus in¯uenzae, Proteus, Bordetella, E.coli, Vibrio, Neisseria and Rizobium. Such homologyincreases con®dence in the ability of this immunogento protect across other P. aeruginosa strains. Indeed,another study currently being completed has foundthat antibody to KatA isolated from P. aeruginosastrain 385 was widely detected in both serum andbronchoalveolar lavage ¯uid from cystic ®brosispatients (unpublished data).
Immunization with KatA appears to be able toinduce a protective immune response despite its pre-dicted periplasmic and cytoplasmic localisation [38].Traditional vaccine candidates have been outer mem-brane proteins which interact with the correspondingmembrane receptors of the host immunocompetentcell, stimulating humoral and cellular immunity [5].Precedents exist, however, which demonstrate that pro-teins which are not embedded in the outer membranecan be highly immunogenic. Catalase [41] and urease[42] from Helicobacter pylori, both cytoplasmic pro-teins, are reported to provide protection in murinemodels of Helicobacter infection. The acquisition of
Fig. 4. KatA-speci®c proliferation of lymphocytes isolated from the
MLN of IPP immunized (R), IPP� IT immunized (.) and nonim-
mune (Q) rats. The values shown represent the mean2 the standard
errors of the mean for triplicate cultures of lymphocytes from each
of ®ve or six rats per group at three concentrations of protein.�p < 0:01:
L.D. Thomas et al. / Vaccine 19 (2001) 348±357 355
protective immunity with these enzymes suggests thatthey may be either periplasmic, exported to the cellsurface or secreted. Both urease and catalase from H.pylori have been found on the outer membrane and ithas been proposed that these proteins have reachedthis site via bacterial lysis [43] or some other unknownmechanism [41,44]. More recently, it has beensuggested that H. pylori proteins, such as the ureasesubunits UreA and UreB, are speci®cally and selec-tively secreted and that programmed autolysis doesnot make a major contribution [44]. Since KatA con-tains no recognizable signal sequence, if it is secreted,it must be via type III or type I (ABC transporter) sys-tems. Detection of KatA on P. aeruginosa by ¯owcytometry indicates that it must become surface-exposed, but the mechanism by which it ends up onthe bacterial surface has yet to be determined.
Whole cell vaccines, with their array of surfaceexposed antigens and adjuvant-e�ect of LPS, have thepotential to provide greater immune responses whencompared with puri®ed protein antigens. Although sys-temic immunization with puri®ed proteins [8] has beenshown to protect against acute respiratory infection, ashas IT immunization with live P. aeruginosa [9], oralimmunization with live P. aeruginosa has not demon-strated the same protective properties [11]. However,oral immunization with killed P. aeruginosa has shownprotection in this rat model [10] with comparable e�-cacy to KatA immunization.
In conclusion, mucosal immunization with KatAcatalase was able to induce a protective immune re-sponse in a rat model of acute respiratory P. aerugi-nosa infection. This antigen induced both cellular andhumoral immune responses. The rat model has beenuseful in identifying the vaccine potential of this anti-gen but further studies are required to investigate thefull potential of KatA, including investigation of mech-anisms for antigen delivery in a vaccine. With the longterm limitations in the e�ectiveness of antibioticagents, the ability to generate protective immunity byvaccination may prove useful in high-risk populationssuch as cystic ®brosis patients. These results suggestthat KatA warrants further investigation to determineits potential as a vaccine candidate for prevention ofacute pseudomonal infection and resolution of chronicPseudomonas infections.
Acknowledgements
This research was supported by a grant from theNational Health and Medical Research Council (GrantNo. 940492). We are grateful to Mojca Keglovic,Melissa Musicka, Catherine Delahunty, CorrinaOszko, Amanda McCue and Heather Domaschenz forexpert technical assistance.
References
[1] Baltch AL, Smith RP. Pseudomonas aeruginosa infections and
