Actinobacillus pleuropneumoniae serotype 1 carrying the defined aroA mutation is fully avirulent in the pig

Research in Veterinary Science 2002, 72, 163±167 doi:10.1053/rvsc.2002.0554, available online at http://www.idealibrary.com on

Actinobacillus pleuropneumoniae serotype 1 carrying the defined aroA mutation is fully avirulent in the pig L. H. GARSIDEy, M. COLLINSy, P. R. LANGFORDz, A. N. RYCROFT*y y

Veterinary Bacteriology Group, Department of Pathology and Infectious Disease, Royal Veterinary College, Hawkshead Lane, North Mymms, Herts. AL9 7TA, UK and z Department of Paediatrics, Imperial College School of Medicine (St Mary's), Norfolk Place, London W2 1PG, UK SUMMARY The aroA gene from Actinobacillus pleuropneumoniae serotype 1 reference strain 4074 was isolated and sequenced. The gene complemented the aroA mutation in Escherichia coli AB2829. A kanamycin resistance cassette was inserted into the aroA gene and the mutant gene was reintroduced into A. pleuropneumoniae by allelic replacement. Intratracheal infection of susceptible pigs with A. pleuropneumoniae aroA caused no signs of respiratory disease or lung lesions in any of the animals at a dose 104 times the dose reliably known to induce acute pleuropneumonia; all animals infected with the unaltered control strain developed acute disease. The aroA mutant was rapidly eliminated from the lungs and tonsil of infected animals. The mutant may represent a safely attenuated strain for use in live bacterial vaccination or the delivery of antigen by the intranasal route. However, the residence time of the mutant in the respiratory tract of the pig may be too short for it to be useful in generating a protective immune response. # 2002 Elsevier Science Ltd. All rights reserved.

ACTINOBACILLUS pleuropneumoniae is the causative agent of porcine pleuropneumonia (Shope et al 1964). The organism is highly virulent in growing pigs and in the acute form the disease is characterised by haemorrhagic, fibrinous lesions of the respiratory tract and rapid deaths are common (Nicolet 1993). As few as 103CFU administered by the intratracheal route will induce peracute disease or death through respiratory failure within 48 hours. In animals with a degree of immunity, the acute form is seen sporadically while the chronic form of the disease predominates and may cause economic losses, susceptibility to other respiratory disease and general ill health. Natural immunity to pleuropneumonia in recovered animals is considered to be protective (Cruijsen et al 1995, Haesebrouck et al 1996). Vaccination based either on whole killed A. pleuropneumoniae or on the cytolytic Apx toxins and a common outer membrane protein confers some protection in the field (Rycroft et al 1999). An alternative means of vaccination is to mimic the protection induced by natural infection using live, attenuated bacteria. An important requirement for attenuation is that it should be based on a defined and non-reverting mutation. Mutation of the aroA gene disrupts the essential aromatic biosynthetic pathway, preventing the synthesis of aromatic amino acids and resulting in poor growth in the mammalian host (Hosieth and Stocker 1981). This has been used very *Corresponding author: Dr Andrew N. Rycroft, Department of Pathology and Infectious Diseases, Royal Veterinary College, Hawkshead Lane, North Mymms, Herts. AL9 7TA, UK. Tel.: ‡44 1707 666362; Fax: ‡44 1707 661464; E-mail: [email protected]

0034-5288/02/020163 ‡ 05 $35.00/0

successfully in many Gram-negative bacteria, and in particular Salmonella sp. aroA mutants of respiratory pathogens including Bordetella pertussis (Roberts et al 1990), Pasteurella multocida (Homchampa et al 1992) and Pasteurella haemolytica (Homchampa et al 1994, Tatum et al 1994), have been reported and show reduced virulence in animal models of infection and potential as attenuated living vaccines. In this communication we report the construction of a defined aroA mutant of A. pleuropneumoniae serotype 1 by allelic exchange and show that this causes virulent A. pleuropneumoniae to become non-virulent in the natural target species and unable to colonise the respiratory tract or tonsil. MATERIALS AND METHODS Bacteria, plasmids and media A. pleuropneumoniae strain 4074 serotype 1 (Shope et al 1964) was used in the study. Strain HK361, serotype 2 (Macdonald and Rycroft, 1992), was used as control in the toxin typing. E coli plasmid host strain JM109 and the E coli aroA mutant AB2829 (G. Dougan, London) were used. A. pleuropneumoniae 4074 was routinely cultured in Tryptone Soya broth (TSB), (Oxoid) or on solid Tryptone Soya agar (TSA) (Oxoid) supplemented with 2 mg mlÿ1 NAD (Sigma) at 37 C. Kanamycin (25 mg mlÿ1) was added to media where appropriate. E coli JM109 and AB2829 were grown in LB medium or on LB plates at 37 C supplemented with ampicillin at 100 mg mlÿ1 or kanamycin at 50 mg mlÿ1 where appropriate. Minimal medium (Davis and Mingioli 1950) was # 2002 Elsevier Science Ltd. All rights reserved.

164

L. H. Garside, M. Collins, P. R. Langford, A. N. Rycroft

All DNA probes were digoxigenin (DIG) labelled and hybridisation visualised using the DIG labelling and detection system (Boehringer). Membranes were hybridised overnight at 45 C in a standard hybridisation mix with formamide. The aroAP1 probe consisted of a gel-purified, 930 bp PCR fragment produced from degenerate primers AROPCR1 and AROPCR2 (Fig 1). To produce the aroAP2 probe, the entire 3
5 kb EcoRI fragment containing the aroA gene was isolated and gel-purified from pARO18. The plasmid pUC-N was digested with EcoRI and labelled to produce the DNA probe pUCP3.

used to demonstrate complementation by plasmidborne aroA gene sequences. Table 1 lists the plasmids used in this work. DNA

manipulations

Plasmid DNA and A. pleuropneumoniae genomic DNA were prepared using standard techniques (Maniatis et al 1982). Ultrapure plasmid DNA for electroporation or sequencing was prepared using the Qiaprep spin miniprep kit (Qiagen, Crawley, UK). All excised DNA bands were purified from agarose gel using the QIAquick gel extraction kit. All DNA modifying and restriction enzymes were from Promega (Southampton) and used according to manufacturer's instructions. DNA

Electroporation Electrocompetent cells were prepared by the method of Jansen et al (1995) and stored at ÿ70 C. Frozen cells were thawed on ice and incubated on ice for 10 minutes before electroporation. Cuvettes were chilled on ice before the addition of 50 ml of thawed cells and 0
5±1 mg of plasmid DNA. Pulse conditions used for all transformations were: voltage 2
5 kV; capacitance 25 mF; resistance 200 ohms. Immediately after electroporation, the A. pleuropneumoniae cells were transferred into 1 ml of ÿ1 TSB supplemented with 2 mg ml NAD and incubated at 37 C for 3±4 hours; E coli cells were incubated for 1 hour with 1 ml of SOB medium (Maniatis et al 1982).

hybridisation

Restriction digested A. pleuropneumoniae genomic was size fractionated on 0.7 per cent agarose gels and blotted onto nitrocellulose membrane. Bacterial colonies were transferred to nitrocellulose membranes. DNA

TABLE 1: Plasmids used in this study Plasmid

Relevant characteristics

Source

pUC18

Cloning vector (colEI origin of replication) PUC18 with the NdeI site removed PUC18 with the kanr gene cassette 3
5 kb EcoRI fragment including aroA gene in pUC18 3
5 kb EcoRI fragment including aroA gene in pUC-N PARO18 with kanr cassette insertion at the NdeI site in the aroA gene

Pharmacia

Polymerase chain reaction

This study Pharmacia This study

Degenerate primers AROPCR1 and AROPCR2, based on a `back-translation' of conserved amino acid sequences in the AroA protein were used (Maskell, 1993). The primers AroKF 5 0 -GGACGACTAAGGTTACCAAT-3 0 and AroKR 5 0 -TGACGAAGCGCATCAACCAA-3 0 hybridise with A. pleuropneumoniae 4074 aroA nucleotides 110 to 129 and to the

pUC-N pUC4K pARO17 pARO18 pARO2K

This study This study

1.55 kb aroKF

aroKR

kanr

aroAP1 930bp aroPCR1

EcoRI pUC18

Nde I

aroA aroK

aroK

aroPCR2

EcoRI pUC18

350 bp

aroAP2 3.5 kb FIG 1: Construction of the aroA gene mutation in A. pleuropneumoniae. The aroA gene from A. pleuropneumoniae was cloned as a 3
5 kb EcoRI fragment into a modified pUC18 vector to form pARO18.The kanr cassette from pUC4K was inserted into the aroA gene at a blunted NdeI site. PCR primers AROPCR1 and AROPCR2 amplified a 930 bp aroA fragment which was labelled to produce the DNA probe aroAP1.The PCR primers aroKF and aroKR amplifya fragment of1
55 kb if the aroAgene is disrupted with the kanr cassette, and of 350 bp if no cassette ispresent.With EcoRI-digested genomic DNA, probe aroAP2 hybridises to a fragment of 3
5 kb if the aroA gene is not disrupted, and of 4
7 kb if the kanr cassette isinserted in the gene.

aroA mutant of A. pleuropneumoniae

complementary strand of nucleotides 394 to 413 (Fig 1). For amplification, 30 cycles of 1 minute at 94 C, 1 minute at 55 C and 2 minutes at 72 C were used. Sequence analysis Plasmid DNA was sequenced on a Pharmacia ALF Express sequencer using labelled primers and a Thermosequenase cycle sequencing kit (Amersham). Both strands were sequenced at least twice to achieve a consensus sequence. Sequence analysis was carried out using the GCG package of programs (Devereux et al 1984) at the SEQNET facility. Apx toxin gene detection The presence of apxCA genes in A. pleuropneumoniae was determined by the method of Frey et al (1995). Animals and experimental infection Nine high-health status piglets of 12 weeks of age were housed in three groups of three animals which were treated identically. Barrier measures were instituted to prevent cross-infection between groups. A. pleuropneumoniae cultures were grown for 16 hours in Columbia broth (Difco) with 2 mg mlÿ1 NAD. This was subcultured 1:50 into fresh prewarmed medium and grown with aeration for 5 hours before dilution in HEPES saline [1
3 mM N-2-hydroxyethyl piperazine-N 0 -2-ethane sulphonic acid (pH 7
3); 140 mM NaCl; 3 mM KCl; 1 mM CaCl2; 1 mM MgCl2] to give a suspension of approximately 2.0  103CFU mlÿ1 (low dose) or 2
0  107CFU mlÿ1 (high dose). Animals were lightly anaesthetised and infected by the intratracheal route with 5.0 ml of A. pleuropneumoniae suspension. Animals in group A were given the aroA mutant at low dose; group B received the aroA mutant at high dose. Group C received the parent organism (4074) at low dose. Animals were observed for clinical disease and euthanased if clearly showing signs of respiratory disease. Surviving animals were euthanased after 36 hours. Pulmonary lesions were recorded at post-mortem examination and a score of 0 to 5 was assigned to each lung lobe (Hannan et al 1982) according to the extent of the lesions (maximum 35). Bacterial culture from lungs was performed by swab from three independent sites within the lung and from the cut surface of the tonsil on TYE agar. RESULTS Isolation of the aroA gene Degenerate PCR primers AROPCR1 and AROPCR2 were used to amplify part of the aroA sequence from A. pleuropneumoniae genomic DNA template. A single band of 930 bp was amplified, DIG labelled (DNA probe aroAP1) and used to detect the complete aroA gene. The aroAP1 probe was hybridised with a Southern blot of digested A. pleuropneumoniae genomic DNA and a single band of approximately 3
5 kb was

165

detected with EcoRI digested DNA. EcoRI-digested A. pleuropneumoniae genomic DNA was separated in a 0
7 per cent agarose gel, and fragments in the size range 3 to 5 kb were purified and ligated to EcoRI-digested pUC18 and transformed into E coli JM109. Resulting colonies were screened by colony hybridisation with the aroAP1 DNA probe. Sequence analysis of plasmid DNA from one positively hybridising colony (pARO17) showed a high level of identity to the aroA gene sequence from related bacteria including Bordetella pertussis Pasteurella multocida and Mannheimia haemolytica. The aroA mutant E coli AB2829 was electroporated with the plasmid (pARO17) and colonies containing pARO17 were able to grow on minimal medium, indicating that the aroA gene in pARO17 was functional and complemented the mutation in AB2829. The whole of the aroA gene was subsequently sequenced. The DNA sequence accession number is AJ012748. Mutation of the aroA gene The single NdeI site of vector plasmid pUC18 was deleted by treatment of the digested NdeI site with Klenow fragment and religation to form the modified vector pUC-N. The aroA gene from pARO17 was then subcloned into pUC-N to create pARO18. This construct also complemented the aroA mutation of AB2829. The kanamycin-resistance cassette from pUC4K was then inserted into the NdeI site of the aroA gene in pARO18 to form the plasmid pARO2K (Fig 1). Insertional inactivation of the aroA gene in pARO2K was confirmed when pARO2K failed to complement the aroA mutation of AB2829. Allelic replacement of the aroA gene in A. pleuropneumoniae Since colE1-based replicons are not maintained in A. pleuropneumoniae, the plasmid pARO2K was used as a suicide vector. Following electroporation of A. pleuropneumoniae with pARO2K, kanamycinresistant transformants were selected that harboured the kanr gene in their chromosome as a result of homologous recombination between aroA sequences of chromosome and plasmid. Colonies in which the wildtype aroA gene had been replaced by the inactivated aroA gene (a double crossover event) were identified by PCR and Southern hybridisation. Template DNA from 45 colonies was examined by PCR (primers AROKF and AROKR, spanning the NdeI site of the aroA gene) and one isolate (AP20) produced a single fragment of approximately 1
5 kb, consistent with the presence of the kanr cassette at the NdeI site (Fig 2). Wild type A. pleuropneumoniae produced a single fragment of 350 bp; all other clones produced fragments of both sizes, indicating that a single crossover event had occurred. This result was confirmed by Southern blot analysis (Fig 3). To verify that the mutant remained unaltered in carriage of the toxin genes apxCAI and apxCAII, known to be crucial for the pathogenicity of A. pleuropneumoniae strains, a PCR-based method was

166

L. H. Garside, M. Collins, P. R. Langford, A. N. Rycroft

M 1

2

3

4

5

6 1

4074 2

3

4

AP20 5 6

7

HK361 8 9

10

M

aroA + kanr

1500 bp

apxICA apxIICA apxIIICA

aroA

350 bp

FIG 2: PCR analysis of genomic DNA from kanamycin-resistant transformants. Amplification of the A. pleuropneumoniae aroA gene using primers aroKF and aroKR. M: 1kb molecular weight marker (Promega), lane 1; pARO18, lane 2; pARO2K, lane 3; A. pleuropneumoniae 4074 genomic DNA, lane 4; clone 7 genomic DNA, lane 5; clone 9 genomic DNA, lane 6; clone 20 genomic DNA (AP20). pARO18 and A. pleuropneumoniae 4074 only show the 350 bp fragment of the unmutated aroA gene; pARO2K has the 1
5 kb mutated aroA gene only. Clones 7 and 9 show both the 350 bp and 1
5 kb bands, indicating that a single crossover event occurred. AP20 has only the 1
5 kb band from the mutant gene, consistent with a double-crossoverevent.

1

2

3

4

4.7 kb 3.5 kb

1.6 kb

FIG 3: Southern analysis of EcoRI digested A. pleuropneumoniae genomic DNA. The EcoRI insert from pARO18 was labelled as a probe (aroAP2) and hybridised with a blot of EcoRI-digested A. pleuropneumoniae genomic DNA. Lane 1,1
632 kb marker from pBR322; lane 2, representative clone AP7 shows both intact and mutant aroA fragments (3
5 and 4
7 kb respectively), consistent with a single crossover; lane 3, clone AP20 shows only a mutant aroA sequence, consistent with allelic exchange; lane 4 shows the intact aroA fragment in wild-type 4074.

used. The toxin profile of mutant AP20 was identical to that of the parent 4074 (Fig 4). Effect of aroA mutation on the virulence of A. pleuropneumoniae in the pig The wild-type A. pleuropneumoniae 4074 and its aroA mutant AP20 showed similar growth rates in vitro

FIG 4: PCR analysisof apxgenesfrom 4074 and AP20 (aroA).Lanes1to 3, DNA amplifiedfrom 4074; lanes 4 to 6, DNA amplified from AP20; lanes 7 to 9, DNA amplified from serotype 2 controlstrain HK361.Primersused to amplify apxIAC, lanes1, 4 and 7; apxIIAC, lanes 2,5 and 8; apxIIIAC, lanes 3, 6 and 9.Lane10, control; M,1kb markers (LifeTechnologies).

and could not be distinguished in colony morphology or other cultural characteristics. To compare the virulence of the aroA mutant with the wild-type A. pleuropneumoniae, groups of pigs were infected with either wild-type 4074 or the mutant strain by the intratracheal route. Two dose levels of the mutant were used. All three animals receiving the wild-type A. pleuropneumoniae (group C, 1
25  104CFU) developed clinical signs of respiratory disease within 36 hours. Haemorrhagic pulmonary lesions and fibrinous pleurisy were demonstrated in these animals at post-mortem, and a mean lung score of 14
6 out of 35 was recorded. Reisolation of A. pleuropneumoniae from the lung lesions and the tonsil confirmed the aetiology of the lesions in each animal. No signs of respiratory disease were observed in any animals in group A (low dose 1
15  104CFU) or in group B (high dose, 1
40  108CFU) infected with the aroA mutant strain. No lesions were present in any of these animals at post-mortem (lung score 0) and no A. pleuropneumoniae could be recovered from the lungs or from the cut surface of the tonsil. DISCUSSION The sequence of aroA in A. pleuropneumoniae serotype 2 (Moral et al 1999) shows 94 per cent sequence identity with the serotype 1 sequence determined in this study. Serotypes 1 and 2 fall into different groups based on toxin typing (Frey et al 1995) and PCR-RFLP of aroA (Moral et al 1999). This is consistent with the level of sequence difference recognised here. The aroA gene sequence shared a high level of sequence similarity with the published aroA gene sequences of other organisms. In many Gram-negative bacteria, aroA is transcribed in an operon with the serC gene (Duncan and Coggins 1986, Hosieth and Stocker 1985). The sequence immediately upstream of the A. pleuropneumoniae aroA gene shared no sequence similarity with published

167

aroA mutant of A. pleuropneumoniae

serC gene sequences, but a high level of sequence similarity was found with corresponding regions in P. haemolytica (Homchampa et al 1994) and P. multocida (Homchampa et al 1992). This suggests that, like P. haemolytica and P. multocida, the aroA gene of A. pleuropneumoniae is transcribed independently, and not in an operon with the serC gene. In preliminary (unpublished) studies in this laboratory, a deletion mutant of the pARO18 plasmid insert with only 98 nucleotides upstream of the ATG transcriptional start codon was able to complement the E coli aroA mutant strain AB2829, suggesting that the minimal promoter elements are contained within this 98 bp region. Insertional inactivation of the aroA gene in AP20 had no detectable effect on the in-vitro growth characteristics. To determine the effect on virulence, pigs were challenged with either the wild-type or mutant A. pleuropneumoniae strains. While A. pleuropneumoniae 4074 induced severe clinical signs in all animals within 36 hours of the challenge, animals infected with the mutant A. pleuropneumoniae AP20 showed no signs of clinical disease. Since the apxCA gene profile had not been altered by the manipulation, it was concluded that recombination or inactivation of these known virulence determinants was not responsible for the loss of virulence. This confirmed that inactivation of the aroA gene had markedly attenuated the virulence of A. pleuropneumoniae 4074: essentially AP20 was unable to cause lung lesions in a very severe infection model. To facilitate direct comparison of the pathogenicity of the attenuated and wild-type strains, the six pigs challenged with AP20 were also euthanased 36 hours post-challenge. No evidence of bacterial infection of the lungs was detected at this stage. While the group size used here is small, the profound difference in pathogenicity between parent and mutant organisms and the use of a high-dose inoculation group means that larger group sizes are not necessary and could be considered unethical. These findings are in agreement with previous studies using riboflavin auxotrophs of A. pleuropneumoniae (Fuller et al 1996), in which the challenge strain was not recovered from the lungs 48 hours post-challenge. The aroA mutant constructed in this study may have potential as a vaccine candidate delivered into the respiratory tract. However, a live vaccine against A. pleuropneumoniae infection would probably need to persist for some time to evoke a substantial protective immune response. AP20 was eliminated very rapidly from the airway and could not be detected by culture from the tonsil. In contrast, the well documented Salmonella typhimurium aroA mutants can be detected in the liver and spleens of mice for several weeks after infection. This may be a result of the protected intracellular habitat of Salmonella as opposed to the extracellular site of invasion of A. pleuropneumoniae. A mutation in A. pleuropneumoniae that eliminates virulence, such as that reported here, but that does not cause such rapid elimination of the organism from the host respiratory tract surface may be critical to the development of a novel live vaccine.

ACKNOWLEDGEMENTS This study was supported by the BBSRC (48/CEL04595) and the Wellcome Trust (052275). Support of this laboratory by the Meat and Livestock Commission is also gratefully acknowledged. REFERENCES CRUIJSEN, T. L. M., VAN LEENGOED, L. A. M. G., HAM-HOFFIES, M. & VERHEIJDEN, J. H. M. (1995) Convalescent pigs are protected completely against infection with a homologous Actinobacillus pleuropneumoniae strain but incompletely against a heterologous-serotype strain. Infection and Immunity 63, 2341±2343 DAVIS, B. D. & MINGIOLI, E. S. (1950) Mutants of Escherichia coli requiring methionine or vitamin B12. Journal of Bacteriology 60, 17±28 DEVEREUX, J., HAEBERLI, P. & SMITHIES, O. (1984) A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Research 12, 387±395 DUNCAN, K. & COGGINS, J. R. (1986) The serC-aroA operon of Escherichia coli. Biochemical Journal 234, 49±57 FREY, J., BECK, M., VAN DEN BOSCH, J. F., SEGERS, R. P. A. M. & NICOLET, J. (1995) Development of an efficient PCR method for toxin typing of Actinobacillus pleuropneumoniae strains. Molecular and Cellular Probes 9, 277±282 FULLER, T. E., THACKER, B. J. & MULKS, M. H. (1996) A riboflavin auxotroph of Actinobacillus pleuropneumoniae is attenuated in swine. Infection and Immunity 64, 4659±4664 HAESEBROUCK, F., VAN DE KERKHOF, A., DOM, P., CHIERS, K. & DUCATELLE, R. (1996) Cross-protection between Actinobacillus pleuropneumoniae biotypes±serotypes in pigs. Veterinary Microbiology 52, 277±284 HANNAN, P. C. T., BHOGAL, B. S. & FISH, J. P. (1982) Tylosin tartrate and tiamutilin effects on experimental piglet pneumonia induced with pneumonic pig lung homogenate containing mycoplasmas, bacteria and viruses. Research in Veterinary Science 33, 76±88 HOMCHAMPA, P., STRUGNELL, R. A. & ADLER, B. (1992) Molecular analysis of the aroA gene of Pasteurella multocida and vaccine potential of a constructed aroA mutant. Molecular Microbiology 6, 3585±3593 HOMCHAMPA, P., STRUGNELL, R. A. & ADLER, B. (1994) Construction and vaccine potential of an aroA mutant of Pasteurella haemolytica. Veterinary Microbiology 42, 35±44 HOSIETH, S. K. & STOCKER, B. A. D. (1981) Aromatic-dependant Salmonella typhimurium are non-virulent and effective as live vaccines. Nature 291, 238±239 HOSIETH, S. K. & STOCKER, B. A. D. (1985) Genes aroA and serC of Salmonella typhimurium constitute an operon. Journal of Bacteriology 163, 355±361 JANSEN, R., BRIAIRE, J., SMITH, H. E., DOM, P., HAESEBROUCK, F., KAMP, E. M., GIELKENS, A. L. & SMITS, M. A. (1995) Knockout mutants of Actinobacillus pleuropneumoniae serotype 1 that are devoid of RTX toxins do not activate or kill porcine neutrophils. Infection and Immunity 63, 27±37 MACDONALD, J. & RYCROFT, A. N. (1992) Molecular cloning and expression of ptxA, the gene encoding the 120-kilodalton cytotoxin of Actinobacillus pleuropneumoniae serotype 2. Infection and Immunity 60, 2726±2732 MANIATIS, T., FRITCH, E. F. & SAMBROOK, J. (1982) Molecular Cloning: a Laboratory Manual. New York: Cold Spring Harbour Laboratory, Cold Spring Harbour MASKELL, D. (1993) Cloning and sequencing of the Haemophilus influenzae aroA gene. Gene 129, 155±156 MORAL, C. H., SORIANO, A. C., SALAZAR, M. S., MARCOS, J. Y., RAMOS, S. S. & CARRASCO, G. N. (1999) Molecular cloning and sequencing of the aroA gene from Actinobacillus pleuropneumoniae and its use in a PCR assay for rapid identification. Journal of Clinical Microbiology 37, 1575±1578 NICOLET, J. (1993) Actinobacillus pleuropneumoniae. In: Diseases of Swine, 7th edn. Eds A. D. Leman, B. E. Straw, W. L. Mengeling, S. D'Allaire and D. J. Taylor. London: Wolfe Publishing Ltd, pp. 401±408 ROBERTS, M., MASKELL, D., NOVOTONY, P. & DOUGAN, G. (1990) Construction and characterisation in vivo of Bordetella pertussis aroA mutants. Infection and Immunity 58, 732±739 RYCROFT, A. N., PENNINGS A. & VAN DEN BOSCH, J. (1999) A. pleuropneumoniae subunit vaccine: compliance with European Pharmacopoeia efficacy testing in pigs. Proceedings of 3rd International Conference on Haemophilus, Actinobacillus and Pasteurella, September 1999 SHOPE, R. E., WHITE, D. C. & LEIDY, G. (1964) Porcine contagious pleuropneumonia. II. Studies of the pathogenicity of the etiological agent Haemophilus pleuropneumoniae. Journal of Experimental Medicine 119, 369±375 TATUM, F. M., BRIGGS, R. E. & HALLING, S. M. (1994) Molecular gene cloning and nucleotide sequencing and construction of an aroA mutant of Pasteurella haemolytica serotype A1. Applied and Environmental Microbiology 60, 2011±2016

Accepted January 25, 2002