Apx toxins in Pasteurellaceae species from animals

Veterinary Microbiology 74 (2000) 365±376

Apx toxins in Pasteurellaceae species from animals Alain Schallera, Peter Kuhnerta, Victor A. de la Puente-Redondob, Jacques Nicoleta, Joachim Freya,* a

Institute for Veterinary Bacteriology, University of Berne, CH-3012 Berne, Switzerland Section of Microbiology and Immunology, Faculty of Veterinary Medicine, LeÂon, Spain

b

Received 16 December 1999; received in revised form 23 March 2000; accepted 23 March 2000

Abstract Pasteurellaceae species particularly of porcine origin which are closely related to Actinobacillus pleuropneumoniae were analyzed for the presence of analogues to the major A. pleuropneumoniae RTX toxin genes, apxICABD, apxIICA and apxIIICABD and for their expression. Actinobacillus suis contains both apxICABDvar. suis and apxIICAvar. suis operons and was shown to produce ApxI and ApxII toxin. Actinobacillus rossii contained the operons apxIICAvar. rossii and apxIIICABDvar. rossii. However, only the toxin ApxII and not ApxIII could be detected in cultures of A. rossii. The Apx toxins found in A. suis and A. rossi may play a role in virulence of these pathogens. Actinobacillus lignieresii, which was included since it is phylogenetically very closely related to A. pleuropneumoniae, was found to contain a full apxICABDvar. lign. operon which however lacks the ÿ35 and ÿ10 boxes in the promoter sequences. As expected from these results, no expression of ApxI was detected in A. lignieresii grown under standard culture conditions. Actinobacillus seminis, Actinobacillus equuli, Pasteurella aerogenes, Pasteurella multocida, Haemophilus parasuis, and also Mannheimia (Pasteurella) haemolytica, which is known to secrete leukotoxin, were all shown to be devoid of any of the apx toxin genes and did not produce ApxI, ApxII or ApxIII toxin proteins. However, proteins of slightly lower molecular mass than ApxI, ApxII and ApxIII which showed limited cross-reactions with monospeci®c, polyclonal anti-ApxI, anti-ApxII and anti-ApxIII were detected on immunoblot analysis of A. equuli, A. seminis and P. aerogenes. The presence of Apx toxins and proteins that imunologically cross react with Apx toxins in porcine Actinobacillus species other than A. pleuropneumoniae can be expected to interfere with serodiagnosis of porcine pleuropneumonia. # 2000 Elsevier Science B.V. All rights reserved. Keywords: Actinobacillus lignieresii; Actinobacillus pleuropneumoniae; Actinobacillus rossii; Actinobacillus suis; Toxins; Apx genes; Diagnosis Ð Bacteria; Serology

*

Corresponding author. Tel.: ‡41-31-631-2484; fax: ‡41-31-631-2634. E-mail address: [email protected] (J. Frey) 0378-1135/00/$ ± see front matter # 2000 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 1 3 5 ( 0 0 ) 0 0 2 0 4 - 2

366

A. Schaller et al. / Veterinary Microbiology 74 (2000) 365±376

1. Introduction The RTX toxins ApxI, ApxII and ApxIII play a predominant role in pathogenicity of Actinobacillus pleuropneumoniae, the etiological agent of porcine pleuropneumonia. Their presence determines signi®cantly the virulence of strains of A. pleuropneumoniae (Rycroft et al., 1991; Frey et al., 1993; Tascon et al., 1994; Frey, 1995; Reimer et al., 1995; Kamp et al., 1997). A fourth RTX determinant, ApxIVA, has recently been found to be common and speci®c to all serotypes of A. pleuropneumoniae, but its precise role in pathogenicity is not yet fully understood (Schaller et al., 1999). The toxins ApxI, ApxII and ApxIII are encoded on polycistronic operons apxICABD (Gygi et al., 1992; Frey et al., 1993, 1994), apxIICA (Frey et al., 1992, 1993; Jansen et al., 1992, 1993), and apxIIICABD (Frey et al., 1993; Jansen et al., 1993), respectively. In these operons, the A gene encodes the structural protein pretoxin, while C encodes a protein involved in post-translational activation of the pretoxin and the B and D genes encode the speci®c type I secretion functions which actively secrete the Apx toxins. Apx toxins induce strong immunological reactions in infected pigs (Frey and Nicolet, 1988; Komal and Mittal, 1990) and are used as antigens in subunit vaccines to induce protective immunity against porcine pleuropneumonia (Kobisch and Van den Bosch, 1992). Due to the strong antigenic properties of the Apx toxins, they have been used in the serodiagnosis of A. pleuropneumoniae infections. However, sera from healthy pigs of farms that were apparently free of actinobacillosis were frequently found to strongly react on immunoblots with puri®ed ApxI, ApxII and ApxIII (Schaller and Frey, unpublished results). Such reactions might originate from bacterial species other than A. pleuropneumoniae which are capable of producing toxins identical or very similar to ApxI, ApxII and ApxIII of A. pleuropneumoniae. Recently, Actinobacillus suis, which causes infections particularly in early weaned piglets, was shown to contain the genes apxICABDvar. suis and apxIICAvar. suis (also named ash for A. suis hemolysin) (Burrows and Lo, 1992; Kamp et al., 1994; VanOstaaijen et al., 1997). The homologous genes have very similar DNA sequences and certain of the A. suis toxin genes are ampli®ed by PCR methods designed for the identi®cation of the A. pleuropneumoniae apx genes (VanOstaaijen et al., 1997). In order to reveal whether other Pasteurellaceae species which are related to A. pleuropneumoniae and are of importance in veterinary medicine contain apx toxin genes and express RTX toxins that serologically cross react with ApxI, ApxII and ApxIII, we have analyzed the type strains of Actinobacillus rossii, Actinobacillus suis, Actinobacillus lignieresii Actinobacillus seminis, Actinobacillus equuli, Pasteurella aerogenes, Pasteurella multocida, Haemophilus parasuis, and Mannheimia (Pasteurella) haemolytica for the presence of analogues to apxICABD, apxIICA and apxIIICABD by hybridization assays and for the production of ApxI, ApxII and ApxIII using monospeci®c, polyclonal antibodies. 2. Materials and methods 2.1. Bacterial strains, culture conditions and DNA isolation Pasteurellaceae reference strains used in this study are listed in Table 1. A. pleuropneumoniae reference strains serotype 1 Shope 4074T and serotype 2 S1536 were

A. Schaller et al. / Veterinary Microbiology 74 (2000) 365±376

367

Table 1 apx Genotypes and Apx phenotypes of Pasteurellaceae reference strains of veterinary importance Species

Strain

a

Phenotype

Genotype

ApxI ApxII ApxIII

apxI

apxII

apxIII

C

A

B

D

C

A

C

A

B

D

A. rossii A. suis A. lignieresii A. seminis A. equuli P. aerogenes P. multocida H. parasuis serovar 2 H. parasuis serovar 5pf M. (P.) haemolytica

ATCC 27072 ATCC 33415a ATCC 49236a NCTC 10851b ATCC 19392a ATCC 27883a CCUG 17976c Bakos A9 Bakos B26 ATCC 33396a

ÿ ‡ ÿ ÿe ÿe ÿe ÿ ndg nd ÿ

‡ ‡ ÿ ÿe ÿe ÿe ÿ nd nd ÿ

ÿ ÿ ÿ ÿe ÿe ÿe ÿ nd nd ÿ

ÿ ‡ ‡ ÿ ÿ ÿ ÿ ÿ ÿ ÿ

ÿ ‡ ‡ ÿ ÿ ÿ ÿ ÿ ÿ ÿ

ÿ ‡ ‡ ÿ ÿ ÿ ÿ ÿ ÿ ÿ

ÿ ‡ ‡ ÿ ÿ ÿ ÿ ÿ ÿ ÿ

‡ ‡ ÿ ÿ ÿ ÿ ÿ ÿ ÿ ÿ

‡ ‡ ÿ ÿ ÿ ÿ ÿ ÿ ÿ ÿ

‡ ÿ ÿ ÿ ÿ ÿ ÿ ÿ ÿ ÿ

‡ ÿ ÿ ÿ ÿ ÿ ÿ ÿ ÿ ÿ

‡ ÿ ÿ ÿ ÿ ÿ ÿ ÿ ÿ ÿ

‡ ÿ ÿ ÿ ÿ ÿ ÿ ÿ ÿ ÿ

A. pleuropneumoniae serotype 1 A. pleuropneumoniae serotype 2

Shope 4074d

‡

‡

ÿ

‡

‡

‡

‡

‡

‡

ÿ

ÿ

ÿ

ÿ

S1534d

ÿ

‡

‡

ÿ

ÿ

‡

‡

‡

‡

‡

‡

‡

‡

a

American Type Culture Collection, Rockville, MD, USA. National Collection of Type Cultures, London, England. c Culture Collection University of GoÈteborg, GoÈteborg, Sweden. d Used as positive controls for apx genes and their expression. e Protein band of molecular masses lower than ApxI, ApxII and ApxIII crossreacting. f Partial identity. g Not determined. b

used as controls and their properties have been reported previously (Frey and Nicolet, 1990). Actinobacillus, Pasteurella and Mannheimia species were grown on Columbia broth or on Columbia broth agar (BBL Microbiology Systems, Cockeysville, MD, USA) supplemented with 0.01% b-NAD (Sigma, MO, USA). H. parasuis was grown on trypticase soy agar medium (BBL Microbiology Systems) containing 5% sheep blood in a 5% CO2 atmosphere. The tests for hemolysis were performed on trypticase soy agar medium containing 5% washed sheep erythrocytes. Genomic DNA was extracted by the guanidine thiocyanate method (Pitcher et al., 1989). Rabbit polyclonal antiserum against polyhistidine tailed ApxI, ApxII and ApxIII was the generous gift of R. Segers, Intervet International, Boxmeer, NL. 2.2. Construction of non-radioactive labelled probes Digoxigenin-11-dUTP (DIG)-labelled probes for the apxI and apxII operons were obtained by PCR ampli®cation using genomic DNA from A. pleuropneumoniae serotype 1 strain Shope 4074T and serotype 2 reference strain S1534 as templates. Chromosomal regions coding for each of the individual toxin genes of the apxICABD and apxIICA operons were ampli®ed using primer pairs APX1C-L/APX1C-R, APX1A-L/APX1A-R,

368

A. Schaller et al. / Veterinary Microbiology 74 (2000) 365±376

APX1B-L/APX1B-R, APX1D-L/APX1D-R, APX2C-L/APX2C-R and APX2A-L/ APX2A-R, respectively (Table 1). The PCR ampli®cation gave rise to a 507-bp APX1C, a 3067-bp APX1A, a 2116-bp APX1B and a 1415-bp APX1D probe for the detection of analogues to apxIC, apxIA, apxIB, and apxID, respectively. The probes APX2C and APX2A for the detection of apxIIC and apxIIA analogues had a size of 469-bp and 2835bp, respectively, and were constructed with primer pairs APX2C-L/APX2C-R and APX2A-L/APX2A-R, respectively. The probes for the apxIII operon were produced by PCR ampli®cation using genomic DNA of A. pleuropneumoniae serotype 2 S1536. Oligonucleotide primer pairs APX3C-L/APX3C-R, APX3A-L/APX3A-R, APX3B-L/ APX3B-R and APX3D-L/APX3D-R were used to construct probes for each of the individual toxin genes of the apxIIICABD operon. The presence of analogues to apxIIIC, apxIIIA, apxIIIB and apxIIID was tested with a 508-bp APX3C, a 3056-bp APX3A, a 1964-bp APX3B, and a 1354-bp APX3D probe, respectively. PCR reactions were run in a DNA thermal cycler (GeneAmp 9600; Perkin±Elmer Cetus) in a 50 ml volume containing 10 mM Tris-HCl, pH 8.3, 1.5 mM MgCl2, 50 mM KCl, 0.005% Tween 20, 170 mM of each dNTP, 0.25 mM forward and reverse primers, 0.5 unit Taq polymerase, 5 ng template DNA, and 50 mM DIG-dUTP. The DNAs were subjected to 35 cycles of ampli®cation (30 s at 948C, 30 s at 528C, and 3 min at 728C). The DIG-probes were detected by using phophatase-labelled antidigoxigenin antibodies according to the manufacturer's instructions (Boehringer Mannheim). 2.3. Southern blot analysis and restriction endonuclease mapping The presence of analogues to apx operons in the different type- or reference-strains (Table 1) was assessed by Southern blot hybridization at high stringency conditions (Ausubel et al., 1990) of ClaI digested genomic DNA using DIG-probes constructed from each of the 10 individual toxin genes of the apxICABD, apxIICA, and apxIIICABD operon. H. parasuis serovar 2 and serovar 5 were identi®ed to belong to two different clusters based on 16S rRNA sequence analysis (data not shown). Chromosomal DNA of A. pleuropneumoniae serotypes 1 and 2 reference strains were used as controls for the hybridization. Restriction enzyme mapping was performed using Southern blot hybridization with genomic DNA digested singly or in combination pair wise with restriction endonuclases BamHI, ClaI, EcoRI, HindIII, KpnI, NdeI, NruI, PstI, and SpeI. 2.4. SDS-PAGE and immunoblotting analysis Bacterial cultures of the different reference strains were grown in liquid medium to mid-exponential phase at an optical density A650ˆ0.8. Total culture proteins were boiled for 10 min in sample buffer (62.2 mM Tris-HCl pH 6.8, 2% SDS, 5% b-mercaptoethanol, 10% glycerol, 0.005% bromophenol blue) and then electrophoresed on 10% SDSpolyacrylamide gels according to the manufacturer's instructions. Immunoblotting analysis was performed as described (Ausubel et al., 1990) with hyperimmune rabbit sera (Schaller et al., 1999) used at a dilution of 1:1000. Bound antibodies were visualized by using phosphatase-labelled goat antibodies directed against rabbit IgG (Kirkegaard Perry, Gaithersburg, MD, USA).

A. Schaller et al. / Veterinary Microbiology 74 (2000) 365±376

369

Table 2 Oligonucleotide primers Name

Sequence (50 to 30 )

Location

Temperaturee

APX1C-L APX1C-R APX1A-L APX1A-R APX1B-L APX1B-R APX1D-L APX1D-R APX2C-L APX2C-R APX2A-L APX2A-R APX3C-L APX3C-R APX3A-L APX3A-R APX3B-L APX3B-R APX3D-L APX3D-R LIG1-L LIG1-R LIG3-L LIG3-R

GGATTTGAGGTTTTAGGAGA GTCTCCTTAGCTATTTACTAAT TGGCTAACTCTCAGCTCGAT GCTGCTTGTGCTAAAGAATA CGGGAAGAAGACTACGGATT CCTCCTATTCCGATTGTAAT CATGGCTAATGGGTTTATAT ACTCTCCGAAACGGATTCTT ATGGGAGGGATGATGCTAAA ACTCTTGCTCATACTGAAGAA CGTCCTTACAACAAGGATT CGGCTCTAGCTAATTGAAT CGGTTCTTAAAGTGGATAA GTTCTTCCTGATATTGTTGT ACGGGCTGAAGAAGCCAAA TCCCTGAGCCACCCAATTGT CGGATTACATGCATTGGTAA GACGGTGGGCAATAATGATT GGACTTGGGGAATTTTTTCA CACTGCGTTCTCCTGTTTTA CGAAAGGCGATGATGAAATC GCCTGAACGTCCTACGATAC CGGTTGAAACCGCTAAAGAA CCATGCCACCTCTCCTAAAA

235±254a 741±720a 746±765a 3812±3793a 3909±3928a 6024±6005a 6036±6055a 7450±7431a 260±279b 728±707b 818±836b 3652±3634b 161±179c 668±649c 754±772c 3809±3790c 3972±3991c 5935±5916c 6099±6118c 7452±7433c 2966±2985a 5411±5392a 4411±4430d 4784±4765d

52 52 52 52 52 52 52 52 52 52 52 52 52 52 52 52 52 52 52 52 52 52 52 52

a

Nucleotide position in X68595. Nucleotide position in M30602. c Nucleotide position in X80055. d Nucleotide position in U05042. e Annealing temperature in 8C. b

2.5. DNA sequence analysis DNA sequencing was performed by using an Applied Biosystems DNA Sequenator AB373 with the Taq Dye Deoxy Terminator Cycle Kit (Applied Biosystems/Perkin±Elmer, Norwalk, CT) and the oligonucleotide primers listed in Table 2. The nucleotide sequence of the promotor region of the apxICABDvar. lign.operon in A. lignieresii upstream of apxICvar. lign. analogue has been deposited at the GenBank/EMBL under accession number AF1188870. 3. Results 3.1. Southern blot hybridization analysis ClaI-digested genomic DNA of A. rossii, A. suis, A. lignieresii, A. seminis, A. equuli, P. aerogenes, P. multocida, H. parasuis, and M. (P.) haemolytica strains were screened

370

A. Schaller et al. / Veterinary Microbiology 74 (2000) 365±376

Fig. 1. Schematic presentation of apxICABD, apxIICA and apxIIICABD operons in A. pleuropneumoniae serotype reference strains and locations of the hybridization probes. The individual toxin genes of the apx operons are shown as open boxes. The locations of the hybridization probes and corresponding primers labelled with their name are indicated with solid bars and white triangles, respectively.

for the presence of each of the individual apx toxin genes from A. pleuropneumoniae with the DIG-probes as described (Fig. 1). A. lignieresii and A. suis hybridized strongly on Southern blot with the probes APX1C, APX1A, APX1B, and APX1D probes indicating the presence of the complete apxI operon in these two species. This con®rmed the previous observation of apxICABDvar. suis in A. suis where this operon seems to be uniformly present in this species (VanOstaaijen et al., 1997) and revealed a new apxICABDvar. lign. operon in A. lignieresii. Hybridization with the APX2C and APX2A probes was observed speci®cally with genomic DNA of A. rossii and A. suis, indicating the presence of the typically incomplete apxIICA operon in these two species. Furthermore, genomic DNA of A. rossii hybridized with the probes APX3C, APX3A, APX3B, and APX3D which showed that this species contained a full apxIIICABDvar. rossii operon. The genomic DNA of A. pleuropneumoniae reference strains serotypes 1 and 2 which were used as controls showed the expected gene pro®le; all the probes for apxI and apxII genes hybridized strongly with genomic DNA of serotype 1, while serotype 2 hybridized strongly with the probes APX1B and APX1D and with all the apxII and apxIII gene probes. The other species listed in Table 1 did not show hybridization signals to any of the apx probes with the given conditions of hybridization stringency. 3.2. Western blot analysis In order to investigate biosynthesis of the protein toxins ApxI, ApxII and ApxIII in A. rossii, A. suis, A. lignieresii, A. seminis, A. equuli, P. aerogenes, P. multocida, and M. (P.)

A. Schaller et al. / Veterinary Microbiology 74 (2000) 365±376

371

haemolytica, we applied immunoblotting analysis using monospeci®c polyclonal rabbit antiserum against ApxI, ApxII and ApxIII, respectively (Table 1). In antigen preparation of A. suis, a protein in the 105 kDa range was detected using both anti-ApxI and antiApxII antibodies. In contrast, no reaction was observed with anti-ApxIII antibodies in this species. Anti-ApxII antibodies bound to a protein of ca. 105 kDa in total antigen preparation of A. rossii. No protein band reacting to polyclonal anti-ApxI and anti-ApxIII was detected in this species. No reactions with anti-ApxI, anti-ApxII and anti-ApxIII antibodies were observed when total culture antigens of A. lignieresii, M. (P.) haemolytica, or P. multocida were tested by immunoblotting (data not shown). However, A. seminis, A. equuli and P. aerogenes yielded in proteins in the 80±100 kDa range that showed reactions with one or several antibodies. In control blots, anti-ApxI and antiApxII antibodies detected proteins in the size range of 105 kDa characteristic for ApxI and ApxII, respectively, in total culture of A. pleuropneumoniae serotype 1, while total culture of A. pleuropneumoniae serotype 2 reacted with anti-ApxII and anti-ApxIII antibodies. 3.3. Restriction endonuclease maps Restriction endonuclease mapping for the newly detected operons apxICABDvar. lign. in A. lignieresii (Fig. 2a), and to apxIICAvar. rossii and apxIIICABDvar. rossii in A. rossii (Fig. 2b and c) was made by Southern blotting using the DIG-probes constructed from each individual toxin genes of apxIC, apxIA, apxIB, apxID, apxIIC, apxIIA, apxIIIC, apxIIIA, apxIIIB, apxIIID. The results of the Southern blots made with genomic DNA of A. lignieresii and A. rossii using different combinations of restriction enzymes are given in Table 3 which resumes the molecular masses of the bands hybridizing to the various probes. Restriction endonucleases NruI, PstI, and SpeI did not cleave apxICABDvar. lign. Restriction endonucleases BamHl and KpnI did not cleave apxIICAvar. rossii. Partial sequencing using primers APX1A-R and LIG1-L (Table 2) revealed differences in the 30 -terminal region from the apxIAvar. lign. gene of A. lignieresii compared to that in A. pleuropneumoniae. Partial sequencing using primer APX3A-R (Table 2) revealed minor differences in the 30 -terminal region of the apxIIIAvar. rossii gene. All the results were veri®ed using supplementary simple and double digestions and partial sequencing with each of the primers used to amplify individual apx toxin genes. 3.4. Promoter sequence of A. lignieresii Since A. lignieresii contains an entire apxICABDvar. lign. operon but apparently does not produce ApxI antigen, we analyzed by DNA sequencing the promotor region of this operon. Primers LIG3-L and LIG3-R were designed to amplify the apxICABDvar. lign. promotor region based on the DNA sequence of apxICABD in A. pleuropneumoniae serotype 1 strain Shope 4074T. These primers ampli®ed a 346-pb chromosomal fragment of the 50 part of the apxICABDvar. lign. operon from A. lignieresii ATCC 49236. Sequence

372 A. Schaller et al. / Veterinary Microbiology 74 (2000) 365±376

Fig. 2. Restriction endonuclease maps of (a) apxICABDvar. lign., (b) apxIICAvar. lign. and (c) apxIIICABDvar. lign..The open boxes represent the imprecise locations of the individual toxin genes of the operons.

A. Schaller et al. / Veterinary Microbiology 74 (2000) 365±376

373

Table 3 Hybridization results with apx gene probes Enzyme probe

HindIII

HindIII KpnI

HindIII ClaI

HindIII NruI

HindIII NdeI

6.0

n.d.a

6.0

n.d.

APX2A

6.0 3.0

n.d.

n.d.

APX3C

5.5

4.2

6.0 2.2 0.8 n.d.

4.3 1.7 1.7 1.3

5.5

n.d.

APX3A

5.5 3.2 0.6

4.2 3.2 0.6

n.d.

5.5 3.2 0.6

n.d.

APX3B

3.2 1.6 6.5 1.6

3.2 1.6 6.5 1.6

n.d.

3.2 0.3 6.5 1.3

n.d.

Probe\enzyme

HindIII

HindIII ClaI

ClaI KpnI

ClaI

ClaI EcoRI

A. lignieresii APX1C

3.8

3.8

1.8

7.5

7.5

1.9

3.9

3.8 1.8 3.8 3.5 3.5

7.5

APX1D

3.8 2.6 2.6 1.3 2.2 1.3

7.5

APX1B

3.9 3.8 3.9

7.5 3.5 3.5

7.5 1.2 2.3 1.2

A rosii APX2C

APX3D

APX1A

a

n.d.

n.d.

HindIII EcoRI

PstI

PstI HindIII

EcoRI KpnI

4.7

4.8

4.8

n.d.

4.7 3.0 1.3 3.8 1.7 3.2 1.7 0.6

4.8 2.0 0.90.8 >20.0

1.2 0.9

n.d.

5.5

>20.0 0.6 >20.0 0.6

3.2 1.6 6.5 1.6

5.5

KpnI

>20.0 5.5

6.5 5.5 KpnI NdeI

5.5 3.2 0.4 0.2 3.2 1.6 1.6 0.3

>20.0 >20.0

NdeI

>20.0 1.9 >20.0

1.1 0.8 6.3 0.8 6.3

14.0 7.2 7.2

>20.0

6.3

7.2

7.2

Not done.

analysis of this fragment (Fig. 3) revealed 82.1% amino acid identity compared to the analogous segment in apxICABD of A. pleuropneumoniae. Characteristically, a ÿ44 box CAAAAT site (Galas et al., 1985) was found at the same location as in the A. pleuropneumoniae apxICABD operon (Frey et al., 1994) (Fig. 3). Such sequences, which have a potential for DNA bending, have been shown to activate transcription initiation (Bracco et al., 1989). However, in A. lignieresii the ÿ10 and ÿ35 consensus sequence of promotors were not found in contrast to apxI in A. pleuropneumoniae where these sequences are immediately downstream the ÿ44 box (Frey et al., 1994). A consensus sequence for a putative ribosome-binding site (RBS) (Shine and Dalgarno, 1974) located directly upstream of the apxICvar. lign. was found in A. lignieresii operon preceding the start codon ATG. The nucleotide sequence of the promotor region of the apxICABDvar. lign. operon in A. lignieresii upstream of apxICvar. lign. analogue has been deposited at the GenBank/EMBL under accession number AF1188870.

374

A. Schaller et al. / Veterinary Microbiology 74 (2000) 365±376

Fig. 3. Sequence of the apxICABDvar. lign. promotor in A. lignieresii ATCC 49236. The 346-pb DNA sequenced region located upstream of the apxI analogue is shown. Proposed ribosome binding site (RBS) and the hÿ44i box are underlined. The EMBL/GenBank accession is AF188870.

4. Discussion Our study on the presence and expression of genes homologous to those that encode the three major RTX toxins of A. pleuropneumoniae, ApxI, ApxII and ApxIII revealed the presence of such genes in the three closely related species A. lignieresii, A. rossii and A. suis, while other related species were devoid of apx genes. We con®rmed the presence of apxICABDvar. suis and apxIICAvar. suis in A. suis as described previously (Burrows and Lo, 1992; VanOstaaijen et al., 1997). The study further showed that both apxICABDvar. suis and apxIICAvar. suis are expressed and that A. suis secretes both ApxI and ApxII. These two toxins confer on A. suis the hemolytic phenotype. In A. lignieresii, an entire apxICABDvar. lign. was found, but no biosynthesis of ApxI could be detected with monospeci®c polyclonal anti-ApxI antibodies. This explains the non-hemolytic phenotype of A. lignieresii on blood agar growth medium. The lack of the canonical promotor sequences ÿ35 and ÿ10 upstream of the ®rst gene, apxICvar. lign., explains the lack of expression under standard growth conditions. It remains to be determined further if the ÿ44 box could be involved in the expression of apxICABDvar. lign. under certain special growth conditions. A. rossii appears on blood agar plates without particular hemolysis, in spite of the biosynthesis of ApxII encoded by the gene apxIICAvar. rossii. This might be explained by the lack of secretion genes, e.g. analogues to apxIBD which were shown to be necessary for secretion of ApxII in A. pleuropneumoniae (Frey, 1995). The function of the apxIIICABDvar. rossii operon remains unclear at present since we have so far not detected ApxIII in cultures of A. rossii grown in standard media. Whether

A. Schaller et al. / Veterinary Microbiology 74 (2000) 365±376

375

apxIIICABD is expressed by A. rossii under special conditions such as during infection or colonization of pigs needs to be determined further. The presence of the different apx genes and the production of ApxI and ApxII in various Actinobacillus species which are apparently not signi®cant pathogens is of major concern in determination of the serological status of pigs. ApxI, ApxII, and ApxIII are strongly recognised antigens. Marked serological reactions to ApxI and ApxII are currently found in sera of healthy pigs from farms which are free of porcine pleuropneumonia. Antibodies against ApxII are very often detected in sera of healthy swine. It must be considered that infections of swine with species of Pasteurellaceae other than A. pleuropneumoniae can severely interfere with serodiagnosis of porcine pleuropneumonia, and that results may be modi®ed by the nature of the antigens employed. Although no apx genes were found in A. seminis, A. equuli, and P. aerogenes, it must be noted that we were able to detect antigens in the size range of 80±100 kDa which cross reacted with monospeci®c, polyclonal anti-ApxI, anti-ApxII or anti-ApxIII sera, thus indicating the possible presence of RTX toxins other than Apx in these species. In this respect, it has to be noted that most recently a new RTX toxin named Pax has been found in P. aerogenes strains that were isolated from porcine abortion or neonatal septicemia (Kuhnert et al., 2000). Acknowledgements We gratefully thank Yvonne Schlatter for her technical assistance and R. Segers, Intervet International, Boxmeer, The Netherlands, for providing us with anti-ApxI, antiApxII and anti-ApxIII antibodies. This work was supported by grant No. 5002.045027 of the Swiss National Science Foundation, SPP Biotechnology program. V. de la Puente was a fellow from the `Ministerio de Educacion y Cultura' from Spain, funded by Project CICYT AGF96-0489.

References Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., Struhl, K., 1990. Current Protocols in Molecular biology. Wiley, New York. Bracco, L., Kotlarz, D., Kolb, A., Diekmann, S., Buc, H., 1989. Synthetic curved DNA sequences can act as transcriptional activators in Escherichia coli. EMBO J. 8, 4289±4296. Burrows, L.L., Lo, R.Y., 1992. Molecular characterization of an RTX toxin determinant from Actinobacillus suis. Infect. Immun. 60, 2166±2173. Frey, J., 1995. Virulence in Actinobacillus pleuropneumoniae and RTX toxins. Trends Microbiol. 3, 257±261. Frey, J., Bosse, J.T., Chang, Y.F., Cullen, J.M., Fenwick, B., Gerlach, G.F., Gygi, D., Haesebrouck, F., Inzana, T.J., Jansen, R., Kamp, E.M., Macdonald, J., MacInnes, J.I., Mittal, K.R., Nicolet, J., Rycroft, A.N., Segers, R.P.A.M., Smits, M.A., Stenbaek, E., Struck, D.K., Vandenbosch, J.F., Willson, P.J., Young, R., 1993. Actinobacillus pleuropneumoniae RTX-toxins Ð Uniform designation of haemolysins, cytolysins, pleurotoxin and their genes. J. Genet. Microbiol. 139, 1723±1728. Frey, J., Haldimann, A., Nicolet, J., Bof®ni, A., Prentki, P., 1994. Sequence analysis and transcription of the apxI operon (hemolysin I) from Actinobacillus pleuropneumoniae. Gene 142, 97±102.

376

A. Schaller et al. / Veterinary Microbiology 74 (2000) 365±376

Frey, J., Nicolet, J., 1988. Regulation of hemolysin expression in Actinobacillus pleuropneumoniae serotype 1 by Ca2‡. Infect. Immun. 56, 2570±2575. Frey, J., Nicolet, J., 1990. Hemolysin patterns of Actinobacillus pleuropneumoniae. J. Clin. Microbiol. 28, 232±236. Frey, J., van den Bosch, H., Segers, R., Nicolet, J., 1992. Identi®cation of a second hemolysin (HlyII) in Actinobacillus pleuropneumoniae serotype 1 and expression of the gene in Escherichia coli. Infect. Immun. 60, 1671±1676. Galas, D.J., Eggert, M., Waterman, M.S., 1985. Rigorous pattern-recognition methods for DNA sequences. Analysis of promoter sequences from Escherichia coli. J. Mol. Biol. 186, 117±128. Gygi, D., Nicolet, J., Hughes, C., Frey, J., 1992. Functional analysis of the Ca(2‡)-regulated hemolysin I operon of Actinobacillus pleuropneumoniae serotype 1. Infect. Immun. 60, 3059±3064. Jansen, R., Briaire, J., Kamp, E.M., Gielkens, A.L.J., Smits, M.A., 1993. Cloning and characterization of the Actinobacillus pleuropneumoniae RTX-Toxin III (ApxIII) gene. Infect. Immun. 61, 947±954. Jansen, R., Briaire, J., Kamp, E.M., Smits, M.A., 1992. Comparison of the cytolysin II genetic determinants of Actinobacillus pleuropneumoniae serotypes. Infect. Immun. 60, 630±636. Kamp, E.M., Stockhofezurwieden, N., vanLeengoed, L.G., Smits, M.A., 1997. Endobronchial inoculation with Apx toxins of Actinobacillus pleuropneumoniae leads to pleuropneumonia in pigs. Infect. Immun. 65, 4350±4354. Kamp, E.M., Vermeulen, T.M.M., Smits, M.A., Haagsma, J., 1994. Production of Apx toxins by ®eld strains of Actinobacillus pleuropneumoniae and Actinobacillus suis. Infect. Immun. 62, 4063±4065. Kobisch, M., Van den Bosch, J.F., 1992. Ef®cacy of an Actinobacillus pleuropneumoniae subunit vaccine. International Pig Veterinary Society, 12th Congress, The Hague, 17±20 August, pp. 216±217. Komal, J.P., Mittal, K.R., 1990. Studies on the interactions of two different serotypes of Actinobacillus pleuropneumoniae. Comp. Immunol. Microbiol. Infect. Dis. 13, 25±34. Kuhnert, P., Heyberger-Meyer, B., Nicolet, J., Frey, J., 2000. Characterization of PaxA and its operon: a cohemolytic RTX Toxin Determinant from Pathogenic Pasteurella aerogenes. Infect. Immun. 68, 6±12. Pitcher, D.G., Saunders, N.A., Owen, R.J., 1989. Rapid extraction of bacterial genomic DNA with guanidium thiocyanate. Lett. Appl. Microbiol. 8, 151±156. Reimer, D., Frey, J., Jansen, R., Veit, H.P., Inzana, T.J., 1995. Molecular investigation of the role of ApxI and ApxII in the virulence of Actinobacillus pleuropneumoniae serotype 5. Microb. Pathog. 18, 197±209. Rycroft, A.N., Williams, D., McCandlish, I.A., Taylor, D.J., 1991. Experimental reproduction of acute lesions of porcine pleuropneumonia with a haemolysin-de®cient mutant of Actinobacillus pleuropneumoniae. Vet. Rec. 129, 441±443. Schaller, A., Kuhn, R., Kuhnert, P., Nicolet, J., Anderson, T.J., MacInnes, J.I., Segers, R.P.A.M., Frey, J., 1999. Characterization of apxIVA, a new RTX determinant of Actinobacillus pleuropneumoniae. Microbiology 145, 2105±2116. Shine, J., Dalgarno, L., 1974. The 30 -terminal sequence of Escherichia coli 16S ribosomal RNA: complementarity to nonsense triplets and ribosome binding sites. Proc. Natl. Acad. Sci. USA 71, 1342±1346. Tascon, R.I., Vazquez-Boland, J.A., Gutierrez-Martin, C.B., Rodriguez-Barbosa, I., Rodriguez-Ferri, E.F., 1994. The RTX haemolysins ApxI and ApxII are major virulence factors of the swine pathogen Actinobacillus pleuropneumoniae: evidence from mutational analysis. Mol. Microbiol. 14, 207±216. VanOstaaijen, J., Frey, J., Rosendal, S., MacInnes, J.I., 1997. Actinobacillus suis strains isolated from healthy and diseased swine are clonal and carry apxICABD(var. suis) and apxIICA(var. suis) toxin genes. J. Clin. Microbiol. 35, 1131±1137.