Vol. 60, No. 2

INFECTION AND IMMUNITY, Feb. 1992, p. 630-636 0019-9567/92/020630-07$02.00/0 Copyright C 1992, American Society for Microbiology

Comparison of the Cytolysin II Genetic Determinants of Actinobacillus pleuropneumoniae Serotypes RUUD JANSEN,' JAN BRIAIRE,' ELBARTE M. KAMP,2 AND MARI A. SMITS1* Departments of Molecular Biology' and Bacteriology,2 DLO-Central Veterinary Institute, Postbox 65, 8200 AB Lelystad, The Netherlands Received 14 August 1991/Accepted 20 November 1991

Cytolysins (Cly) I, II, and III are toxins secreted by Actinobacillus pleuropneumoniae. These toxins are thought to play an important role in the pathogenesis of porcine pleuropneumonia. Clyl and Clyll are RTX toxins and in general these toxins are encoded by operons consisting of four genes, C, A, B, and D. Our group recently cloned the C and A genes of the ClyII operon (clyIlCA) of serotype 9. We found that this Clyll operon is truncated and lacked intact B and D genes (clyJIBD). B and D genes of the Clyl operon (clyIBD) were present however in serotype 9. In this study we analyzed the ClyIl operons of the reference strains of the 12 A. pleuropneumoniae serotypes and compared them with the Clyll operon of serotype 9. We focused on (i) the presence, (ii) the sequence similarity, and (iii) the genomic environment of the clyIlCA genes. The presence of the clylICA sequences was studied by hybridization analysis of genomic DNA. The sequence similarity was studied by restriction fragment analysis on polymerase chain reaction-amplified DNA. The genomic environment was compared by analysis of the sequences that are located 3' of the clyIICA genes. We demonstrated that the clylICA genes (i) are present in the reference strains of all serotypes, except serotype 10, (ii) have a high degree of sequence similarity, and (iii) are not contiguous with intact clylIBD genes. We conclude that the organization and nucleotide sequence of the Clyll operons of A. pleuropneumoniae are very similar. We also studied the presence of clyIBD sequences and found them to be present in the reference strains of all serotypes, except serotypes 3 and 6. Thus, in most serotypes the clyIBD genes may complement the absent clyllBD genes.

Actinobacillus pleuropneumoniae is the causative reagent of a contagious and often fatal pneumonia in pigs. Extracellular protein toxins are thought to be involved in the development of the pneumonic lesions (19), and immunization with such proteins gave protection in challenge experiments (3). Several hemolytic and cytotoxic activities have been identified in the culture supernatants of this bacterium. Among the 12 serotypes, these activities differ with respect to quantity and quality (6, 13, 14). Frey and Nicolet detected a Ca2'-inducible hemolysin activity (HlyI) in serotypes 1, 5, 9, 10, and 11 and a non inducible hemolysin activity (HlyII) in almost all serotypes (6). Kamp and van Leengoed detected four different cytotoxic activities among the 12 serotypes (14). The proteins responsible for these activities have been poorly identified. Literature on these proteins is confusing and incomplete. To get insight into this matter, we started a program to identify and characterize the secreted toxins and their genes. Since there is growing evidence that strains belonging to one serotype produce the same cytolysins (6, 14), we focussed primarily on the reference strains of the 12 A. pleuropneumoniae serotypes. In the culture supernatants of these reference strains we recently identified three different toxin proteins, cytolysin I (Clyl), cytolysin II (ClyII), and cytolysin III (ClyIII) (13). Some strains secrete only one of these cytolysins, but most secrete two of them. ClyI is a 105-kDa protein and is secreted by the reference strains of serotypes 1, 5, 9, 10, and 11. Clyl is assumed to be equivalent to HlyI. ClyIl is a 103-kDa protein and is secreted by all reference strains, except the serotype 10 reference strain. ClyII is assumed to be equivalent to HlyII. ClyIllI is a 120-kDa protein and is secreted by the reference strains of *

serotypes 2, 3, 4, and 8. ClylII is probably equivalent to the pleurotoxin of serotype 2 (18). Like the alpha-hemolysin (Hly) of Escherichia coli and the leukotoxin (Lkt) of Pasteurella haemolytica, the Clyl (HlyI) and Clyll (HlyII) toxins were shown to be members of the RTX toxin family (2, 7, 20). RTX toxins are encoded by operons consisting of four contiguous genes in the order C, A, B, and D (4, 21). The A gene encodes the structural toxin protein, the C gene encodes a protein that activates the toxin by fatty acylation (9), and the B and D genes encode proteins that are involved in the secretion of the toxin. A regulated transcription terminator is present between the A and B genes (15). In a recent study, our group isolated the ClyII-encoding genes clyfIC and clylIA (clyIICA) of the reference strain of serotype 9 (20). Genes encoding hemolytic proteins of serotypes 1 and 5 have been isolated and characterized by Chang et al. (2) and Gygi et al. (7). These hemolysins have only 42% amino acid sequence similarity (5). The hemolysin of serotype 5 appeared to be identical to ClylI of serotype 9 (20). Despite the similarity between the clyJICA sequences of serotypes 5 and 9, major differences were found distal to the clyIIA genes. In serotype 5, an open reading frame of at least 107 bp was present adjacent to clyIIA, coding for the proximal part of a protein similar to RTX B transporter proteins (2). In serotype 9, only the first 37 bp of this putative clyIIB gene were present adjacent to clyIIA (20). Finally, Anderson et al. found in serotype 7 that no sequences similar to lktBD are present adjacent to the gene encoding a 103-kDa toxin (1). In this study, we analyzed the ClyII operons of the reference strains of the 12 A. pleuropneumoniae serotypes and compared them with the ClyII operon of serotype 9. We focused on (i) the presence, (ii) the sequence similarity, and (iii) the genomic environment of the clyIICA genes. The presence of the clyIICA sequences was studied by hybrid-

Corresponding author. 630

VOL. 60, 1992

ization analysis of genomic DNA. The sequence similarity was studied by restriction fragment analysis on polymerase chain reaction (PCR)-amplified DNA. The genomic environment was compared by analysis of the sequences that are located 3' of the clyIICA genes. The results will be discussed in terms of the role of ClyII in pathogenesis and the implications for porcine pleuropneumonia control. MATERIALS AND METHODS Bacterial strains, genomic DNA, plasmids, and oligonucleotides. The reference strains for the 12 serotypes of A. pleuropneumoniae were used as sources of genomic DNA. The reference strains for serotypes 1 to 12 were, respectively, S4047, 1536, 1421, M62, K17, Femo, WF83, 405, 13261, D13039, 56153, and 8329. High-molecular-weight DNA was isolated by proteinase K-sodium dodecyl sulfate (SDS) lysis, phenol-chloroform extractions, and precipitation with ethanol (16). DNA fragments were cloned with the plasmid pGEM7Zf(+) (Promega Corp., Madison, Wis.) in E. coli JM101 by use of standard molecular biology techniques (16). The sequences of the oligonucleotides and their positions relative to the first base of the clylIC open reading frame (20) were as follows: oligonucleotide 158, CGTAGGT GTTGCCCC (2033 to 2052); oligonucleotide 283, CCATTA CAGAACGTTGGTAC (-232 to -208); oligonucleotide 284, ATTAATGQCQQCCCAGGACCAG (1414 to 1435); oligonucleotide 285, ACAAAAGCGGCCGCATCTTACA (1356 to 1377); oligonucleotide 286, CTACAGCTAAACCAAAG ATCCT (3473 to 3493); and oligonucleotide 322, ATTC AATAAGCTTGAGCCGC (3366 to 3385). Recognition sites for the restriction enzymes HindIII in oligonucleotide 322 and NotI in oligonucleotides 284 and 285 are underlined. These sites were introduced for cloning purposes by the modification of one, two, or three bases (oligonucleotide 322, 285, or 284, respectively) of the original serotype 9 clyIICA sequence. In this study we used only the HindIII site. Southern blot and dot blot analyses. Southern blots with restriction fragments of genomic DNA that were separated on 0.8% agarose gels and dot blots with high-molecularweight genomic DNA were made with GeneScreen Plus membranes (New England Nuclear Corp., Boston, Mass.). For the Southern blots, we used 1 jLg of DNA per lane, and for the dot blots, we used 50 ng per dot. The blots were hybridized overnight at 65°C in a solution of 1.5 x SSPE (lx SSPE is 0.18 M NaCl, 10 mM Na H2P04 [pH 7.4], and 1 mM EDTA [pH 7.4])-1% SDS. As probes we used DNA that was labeled with [a-32P]dCTP (Amersham Corp., Arlington Heights, Ill.) by use of a random prime labeling kit (Boehringer GmbH, Mannheim, Germany). The blots were washed with a final stringency of 0.2x SSC (lx SSC is 0.15 M sodium chloride, 0.015 M sodium citrate)-1% SDS for 15 min at 65°C. Radioactivity was detected by autoradiography with X-Omat AR film (Eastman Kodak, Rochester, N.Y.), using intensifying screens. Amplification and radiolabeling of DNA by the PCR. The clylICA sequences of the A. pleuropneumoniae reference strains were amplified by PCR, using genomic DNA as a template. The PCR was done in 50 pJl containing approximately 100 ng of template DNA, 1 ,uM each of two specific oligonucleotides, 1 U of Amplitaq DNA polymerase (PerkinElmer Cetus, Norwalk, Conn.), a 0.2 mM concentration of each of four deoxynucleotide triphosphates (dNTP), 25 mM Tris-HCl (pH 8.7), 2.5 mM MgCl2, and 0.05% gelatin. The reaction mixture was covered with a drop of mineral oil and

CYTOLYSIN II GENES OF A. PLEUROPNEUMONIAE

631

subjected to 30 PCR cycles of 1 min at 92°C, 1 min at 55°C, and 3 min at 74°C in a Thermal Cycler (Perkin-Elmer Cetus). The amplified DNA fragments were separated by agarose gel electrophoresis and isolated from the gel with GeneClean (Bio 101, La Jolla, Calif.). Two to four nanograms of isolated DNA was radiolabeled in a successive PCR. Now the reaction mixture contained a 0.05 mM concentration of each dNTP and 5 ,uCi of [a-32P]dCTP. The high concentration of dNTPs used in this PCR will decrease the specific activity of the synthesized DNA, but will favor the fidelity and complete extension of the PCR products (8, 10). The reaction mixture was extracted with phenol-chloroform, and the radiolabeled DNA was precipitated with ethanol in the presence of 2.5 M ammonium acetate. Inverse PCR, cloning, and sequence analysis. The reaction conditions for the PCR were also used for the inverse PCR (17), except that the extension reaction was for 90 s at 74°C. The template DNA for the inverse PCR was prepared as follows. Genomic DNA was digested with HindIII, and the fragments were size fractionated by agarose gel electrophoresis. The DNA fragments were isolated from the gel with GeneClean, and approximately 100 ng of DNA was circularized by self-ligation with T4 ligase in a volume of 50 ,ul. One-tenth of the ligation products was used as the template in the inverse PCR. The inverse PCR resulted in a high background of aspecific products. Therefore, we reamplified the amplification products in a subsequent PCR. The reamplification products were digested with HindIII and cloned into pGEM7Zf(+). Sequence analysis of the cloned fragments was done with a T7 sequencing kit (Pharmacia, Uppsala, Sweden) by use of oligonucleotides specific for the SP6 and T7 promoter sequences of pGEM7Zf(+) (Promega). Restriction fragment analysis. Radiolabeled DNA fragments were digested with restriction enzyme AluI, Sau3AI, RsaI, or Hinfll. The resulting DNA fragments were separated on a 5% polyacrylamide gel (acrylamide-bisacrylamide ratio, 19:1) in TBE (0.18 M Tris-boric acid [pH 7.8], 0.5 mM EDTA). The digestion products were visualized by autoradiography of the dried gel. Nucleotide sequence accession number. The GenBankEMBL accession number for primary nucleotide sequence data is X62906.

RESULTS Presence of the clylICA genes in the reference strains of serotypes 1 to 12 and comparison of their nucleotide sequences. We detected clylICA sequences in the genomic DNA of the reference strains of the 12 A. pleuropneumoniae serotypes by dot blot hybridization, using serotype 9 clyIICA sequences as the probe. The probe comprised bases 315 to 3233 of the clyIICA sequence (20) and hybridized to the DNA of all serotypes, except serotype 10 (data not shown). Genomic DNA of the 12 reference strains was subjected to PCR with three different sets of oligonucleotides. Figure 1A shows the position of the oligonucleotides in the clyJICA genes. Oligonucleotide set 283/284 was used for amplification of the 5' region, set 285/286 was used for the 3' region, and set 283/286 was used for full-length clylICA. With these three oligonucleotide sets we obtained for serotypes 1, 2, 3, 4, 5, 7, 8, 11, and 12 amplification products of the same size as for serotype 9. Figure 1B shows the 1.75-kbp fragments obtained with set 283/284. Figure 1C shows the 2.05-kbp fragments obtained with set 285/286, and Fig. 1D shows the 3.2-kbp fragments obtained with set 283/286. The 3.2-kbp fragments of serotypes 1, 2, 3, 4, 5, 7, 8, 11, and 12 had

632

INFECT. IMMUN.

JANSEN ET AL.

A

A uwlu, Hi

N

."'

p

H -I --

..X

2'84

283/284 1 2- 3- 4 5 6 7 -

-

8 -

9-

285/286 2

10111 1- - .-

1

2- 3- 4.

5 -

6 -

7 I

283/286

8- -9 101112 1.-

1 2 3 4 5 6 7 8 9 101112 ...

...

-.O

.1)

5.

I -

B

C

D

FIG. 1. PCR amplification of clyIICA sequences. (A) Schematic presentation of the clylICA determinant of serotype 9. The boxes represent the clylIC, clyIIA, and truncated clylIB genes. The locations of recognition sites for the restriction enzymes XbaI (X), PstI (P), and HindIII (H) are indicated. The arrows indicate the locations and orientations of oligonucleotides 283, 284, 285, and 286, used in the PCR to amplify fragments of the Clyll determinant and oligonucleotides 158 and 322, used in the inverse PCR to amplify the 3' flanking sequences. The double-pointed arrow spans the region covered by the clylICA probe used for hybridization. (B, C, and D) Ethidium bromide-stained agarose gels with PCR-amplified DNA fragments of the reference strains of serotypes 1 to 12, using oligonucleotide sets 283/284 (B), 285/286 (C), and 283/286 (D). Serotype numbers and molecular size markers in kilobase pairs are indicated.

identical restriction maps for the enzymes HindIII, XbaI, and PstI as did serotype 9 clyIICA (data not shown). With set 285/286, serotype 6 gave an amplification product of the same size as that of serotype 9, but with sets 283/284 and 283/286 the amplification products were approximately 1.8 kbp longer. Apparently in the reference strain of serotype 6, rearrangements have taken place in the clylICA genes. Therefore, we excluded the reference strain of serotype 6 from further studies. As expected, serotype 10 did not give amplification products with either set of oligonucleotides. The degree of similarity between the clyIICA genes of serotypes 1, 2, 3, 4, 5, 7, 8, 9, 11, and 12 was studied by restriction fragment length polymorphism (RFLP) analysis of the 3.2-kbp amplification products. The DNA fragments were labeled with [t_-32P]dCTP and digested with the restriction enzyme AluI, HinfII, RsaI, HpaII, or Sau3AI. The restriction fragments were analyzed by gel electrophoresis and autoradiography. For each of the four restriction enzymes, the restriction patterns were very similar. As an example, the autoradiogram of the AluI restriction fragments of serotypes 1, 2, 3, 4, 5, 7, 8, 9, and 11 is shown in Fig. 2. The only differences we found were an additional site for RsaI between positions 2818 and 3143 in the reference strain of serotype 7, a small deletion between positions 510 and 690 detected by HindII digestion in the reference strain of serotype 8, and an additional site for Sau3AI at position -135 in the reference strain of serotype 5. The latter difference was expected on the basis of the published sequences (2, 20). Cloning and analysis of the sequences adjacent to clyllA. The proximal parts of putative clyfIB genes were found adjacent to the clyIIA genes of the reference strains of serotypes 5 and 9 (2, 20). In serotype 5, this sequence extended to at least 107 bases, but in serotype 9, it was truncated after 37 bases. To study the presence of this

putative clyfIB gene in the other serotype reference strains, we cloned the sequences distal to clyIIA by inverse PCR. For that purpose, we circularized genomic HindIII fragments, since a unique HindIll site is present in all clyIICA sequences at a position corresponding to base 2008 of the serotype 9 clyIICA sequence. The genomic HindIII fragments that contain the 3' clyIICA sequences were identified by Southern blotting with a probe that comprised bases 2008 to 3493 of the serotype 9 clyIICA sequence. Four HindIlI fragments of different sizes hybridized to this probe (Fig. 3). These are a 2.8-kbp HindIII fragment in the reference strains of serotypes 2, 3, 5, 7, and 8; a 2.3-kbp fragment in the reference strains of serotypes 1, 9, 11, and 12; a 10-kbp fragment in the reference strain of serotype 4; and a 4.3-kb fragment in the reference strain of serotype 6. Inverse PCR on these HindIII fragments is expected to result in amplification products that are approximately 1.3 kbp smaller than the hybridizing HindIll fragments. The expected amplification products were obtained for all serotypes, except serotypes 4 and 6. The failure to get amplification products with the DNA of serotypes 4 and 6 was probably due to the large sizes of the HindlIl fragments (10 and 4.3 kbp; Fig. 3, lanes 4 and 6) that carried the 3' clylIA sequences. We cloned the inverse PCR products of the reference strains of serotypes 1, 2, 3, 5, 7, 8, 9, 11, and 12 into pGEM7Zf(+) and determined their nucleotide sequences. All these strains appeared to have a truncated clyIIB gene adjacent to clyIIA. Two different truncation points were identified, one at base 37 and one at base 501 of the clyIIB open reading frame. Figure 4 shows the nucleotide sequence of the 501-bp long truncated clyIIB and the derived amino acid sequence. This protein has a similarity of 71% with the corresponding sequence of serotype 9 ClyIB (20) and a similarity of 64% with the corresponding sequence of P. haemolytica LktB (21). We identified two major types of nucleotide sequences adjacent

VOL. 60, 1992

1

it'

41

1

ALU

2 3 4 5 ,;1

633

CYTOLYSIN II GENES OF A. PLEUROPNEUMONIAE 2 3 4

5 6

7 8

9 10 11 12

7 8 9 11 C ..

t

w;lo,

509

bor

wv

AW&AL

5.1 --4 2

411

-

.

_.

.

4

-

-w

_ ,_b

_

-_

3.5

-w

2.0

260

am

ab

_"

1 4 -m.

" -

-4m

-_

wo.-

144

84

-_ 42 FIG. 2. Autoradiogram of AluI fragments, separated on a 5% polyacrylamide gel of the PCR-amplified and radiolabeled 3.2-kbp clyIICA sequences of the reference strains of serotypes 1, 2, 3, 4, 5, 7, 8, 9, and 11. C (lane 10) is the control PCR product obtained with the cloned serotype 9 clyIICA sequence. Serotype numbers and molecular size markers in base pairs are indicated.

to the truncated clyIIB genes. The type I sequence is present in the reference strains of serotypes 1, 7, 9, 11, and 12, and the type II sequence is present in the reference strains of serotypes 2, 3, 5, and 8. A subtype of type I is present in the reference strains of serotypes 7 and 12, they contain an additional seven bases, AACCACT, at position 3664. The type I nucleotide sequence was published already by our group together with the serotype 9 clyIICA sequence (20). The type II nucleotide sequence is shown in Fig. 4. The type I and type II sequences did not show any similarity to each other or to RTX sequences.

On the basis of Southern blot, RFLP, and sequence analyses, we deduced the genomic structure of the ClyIl operons and the flanking type I and type II sequences of the 12 reference strains. A schematic presentation of these genomic structures is given in Fig. 5. Distribution of type I and type II sequences among A. pleuropneumoniae. We studied the distribution of the type I and II sequences among the reference strains of the 12

0.9

FIG. 3. Autoradiogram of HindIII genomic DNA fragments of the reference strains of serotypes 1 to 12 that were hybridized with radiolabeled clyIICA sequences of serotype 9. The position of the probe in the clyIICA gene is indicated in Fig. 1A and comprises base pairs 2008 to 3493. Serotype numbers and molecular size markers in kilobase pairs are indicated.

serotypes by Southern blot analysis of genomic HindlIl fragments. A 547-bp HindIII-PstI fragment of serotype 9 was used as a type I-specific probe, and a 959-bp XbaI-HindIII fragment of serotype 2 was used as a type II-specific probe (Fig. 6C). The type II probe hybridized with 2.8-kbp HindIII fragments of serotypes 2, 3, 5, and 8 and with a 10-kbp fragment of serotype 4 (Fig. 6B). These HindlIl fragments also carried the clyIICA sequences (Fig. 3, lanes 2, 3, 4, 5, and 8). Although we were unable to clone the sequences distal to serotype 4 clyIIA, we concluded from these data that type II sequences are adjacent to the serotype 4 ClyIl operon, as was shown for serotypes 2, 3, 5, and 8 (Fig. 5). Type II sequences are not present in the reference strains of serotypes 1, 6, 7, 10, 11, and 12. The type I probe showed a more complex hybridization pattern than the type II probe (Fig. 6A). As expected from the data shown in Fig. 3, the type I probe hybridized to a 2.3-kbp fragment of serotypes 1, 9, 11, and 12 and to a 2.8-kbp fragment of serotype 7. Most serotypes, however, showed additional hybridizing fragments. A unique hybridization pattern was found for most serotypes, ranging from no hybridization for serotypes 4, 6, and 10 up to hybridization with four different fragments in serotype 2. Serotype 6 DNA did not hybridize with the type I or type II probes. Presence of clyIBD sequences in the reference strains of the 12 serotypes. None of the 12 reference strains contained a Clyll operon with intact genes for the B and D transporter proteins. We checked the 12 reference strains for the presence of transporter genes by hybridization of their genomic DNA with the serotype 9 clyIBD sequences in a dot blot assay. The probe was a 2,349-bp EcoRI fragment (bases 348 to 2697) that contained most of the clyIBD open reading frames (20). The genomes of all reference strains, except the reference strains of serotypes 3 and 6, hybridized with the

634

INFECT. IMMUN.

JANSEN ET AL. B

M

3436 B

3536

3636_

P

N

K D L G L A A L K I L A Q Y H N I S V N P

K N

T B L K H K F D L D G K G L S L T S W L I A A K S L D L K A K A I N G E L P FV N L P A L I W R E D G K H F I L V K I D K D K K R K A V

L T Y N L E D D A P K I L E B S E F T N L Y Q N K I I L M T S

R Y

3736

_

R A S

I

I

G Q L A K F D .F T W F I

P A V

3836 V

F

Ti Q

_

= ^

I

F

r

y y v

a

v

a

q

r

f

n

I K Y R K I

T L

I

F

1 v f 1 1 h f f 1 1

s

1

V

I

S

ter

3936 4036 4136

4236 4336

4436_ 4536 m

4636

4736=

FIG. 4. Nucleotide sequences adjacent to clyIICA. Nucleotide sequences of the inverse PCR fragments obtained from the reference strains of serotypes 2, 3, 5, and 8. The sequence is numbered from the first nucleotide of clyIIC (20). Position 3956 corresponds to clyIIB position 501. The amino acid sequence of the truncated ClyIIB protein that is homologous to other RTX proteins is indicated by capital letters. The nucleotide sequence of the 37-bp truncated clyIIB, as found in the reference strains of serotypes 1, 9, 11, and 12, is underlined. The type II sequence comprises bases 3957 to 4796. dlA

dyliC

probe. This finding demonstrated the presence of clyIBDrelated sequences in most serotypes. DISCUSSION

2

In this report we demonstrated that clylIC and clyIIA genes are present in the reference strains of the 12 A. pleuropneumoniae serotypes, except in the reference strain of serotype 10. This result is in agreement with the observation that serotype 10 is the only serotype that does not secrete ClyIl (13). PCR amplification of the clyIlCA sequences resulted in equally sized products, except for serotype 6. The clyIICA genes of the serotype 6 reference strain contain an insertion of approximately 1.8 kbp in the 5' region. This insertion probably influences the expression of the ClyIIC protein, because this strain produces a normally sized Clyl protein with a low hemolytic activity (13). The clyIICA sequences of the reference strains of serotypes 1, 2, 3, 4, 5, 7, 8, 9, 11, and 12 were compared with each other by extensive RFLP studies. For these studies we used four different restriction enzymes, which together have 57 recognition sites in the clyIICA sequence of serotype 9. The RFLP studies showed very similar restriction patterns for the clyIICA sequences of the 10 serotypes that were examined. This finding indicates that the clyIICA genes of the reference strains of serotypes 1, 2, 3, 4, 5, 7, 8, 9, 11, and 12 have very similar primary structures and encode very similar Clyll proteins. The degree of similarity is indicated by the small number of differences we found (only three) among the 10 clyIICA sequences tested. Comparison of the reported clyIICA sequences of serotypes 5 and 9 also revealed in serotype 5 the presence of an additional HpaII site at position 209, a 3-bp deletion at position 51, and a single base pair deletion at position 44. With the RFLP analysis we found no evidence for the additional HpaII site and for the deletions. This finding indicated that the clylICA sequences of serotypes 5 and 9 are even more similar to each other than would be expected from their reported sequences. In contrast to the similarity of the clyIICA sequences is the presence of two different sequences adjacent to the ClylI

3 |-w

A

H

dyliB

H

H

H

H

H

|~a

7H

5?o>rr s

4 *

HI H

X

H

I

5

6 7

8 9

*- I

,=?

l

I

H H

H

H

H

H

H H

H

10 11

12

H

s

7bp W~~~s

H

FIG. 5. Schematic presentation of the organization of the Clyll determinant of the reference strains of the 12 A. pleuropneumoniae serotypes. Open boxes represent the clyIIC, clyIIA, and truncated clyIIB genes. Hatched boxes represent the type I sequences, and stippled boxes represent the type II sequences. The HindilI (H) sites used for the inverse PCR are indicated. For serotypes 4 and 6, these HindIII sites are not indicated, since no inverse PCR and sequence data were obtained from these serotypes. The presence of the type II sequence adjacent to clyIICA of serotype 4 was established by the data from the genomic Southern blot shown in Fig. 6B, lane 4. No data were obtained with respect to the nucleotide sequence and the length of the serotype 4 truncated clyIIB, as is indicated by the question mark. The position of the 7-bp insertions in the type I sequences of serotypes 7 and 12 is indicated by the triangle. The black box in the serotype 6 ClyIl operon represents the 1.8-kb insertion.

CYTOLYSIN II GENES OF A. PLEUROPNEUMONIAE

VOL. 60, 1992 2 3 4 5 6 7 8 9 10 11 12

1

*0-p

1

2 3

635

4 5 6 7 8 9 10 11 12

-

L8. I

40

*

5.1 _

rl -w

3.5

0

2.0

*@

* -_ 1.4

A

B cIvI/C 2

clyllA

A clyllB lix

x

II _ . ...

..2 .S

.I ..

..............

------ I I

9

I)

II

..

'---

II

-

C

_ I_

FIG. 6. Distribution of type I and type II sequences among the reference strains of the 12 serotypes. (A) Autoradiogram of genomic HindIII fragments of serotypes 1 to 12 that were probed with a type I-specific PstI-HindIII fragment, comprising bases 3953 to 4499, of serotype 9. (B) Autoradiogram of genomic HindlIl fragments of serotypes 1 to 12 that were probed with a type II-specific 1-kbp XbaI fragment of serotype 2. Serotype numbers and molecular size markers in kilobase pairs are indicated. (C) Schematic presentation of the Clyll determinants of the reference strains of serotypes 2 and 9. Indicated are the HindIll (H), XbaI (X), and PstI (P) sites. Open boxes represent the clylIC, clylIA, and truncated clyfIB genes. Hatched and stippled boxes represent the type I and type II sequences, respectively. The positions of the type I and type II probes are indicated by double-pointed arrows.

II). The type I and type II in all reference strains of A. pleuropneumoniae. The type II sequences are exclusively present in the reference strains of serotypes 2, 3, 4, 5, and 8 that carry the large (501-bp) truncated form of clyIIB. The type II sequences are absent in the other reference strains. The type I sequences are, in contrast to the type II sequences, not exclusively linked to a certain form of truncated clyIIB. The type I sequences are found in most reference strains and often are present as multicopies at variable positions in the genomes. Such a random distribution is often seen for mobile genetic elements and may indicate that the type I sequence is part of such a mobile element. However, no sequence homology could be demonstrated between the type I sequence and other mobile elements in the GenBank-EMBL data base. The genetic organization of the ClyII operons, as shown in Fig. 5, may shed some light on the evolutionary origin of this operon in A. pleuropneumoniae. In our view, the sequence similarity of the clyIlCA genes and of the inactive truncated clyIIB genes reflects a single transfer of the ClyII operon to operons (type I and type sequences are not present

this bacterium. The transferred operon may already have carried the 501-bp truncated clyIIB gene linked to the type II sequence, as is still found in the reference strains of the serotypes 2, 3, 4, 5, and 8. In the reference strains of the serotypes 1, 9, 11, and 12, a deletion event, mediated by the type I flanking sequence, has further reduced the truncated clyfIB to 37 bp and coeliminated the type II sequence from these strains. The genetic organization in the reference strains of serotypes 6 and 7 cannot be fully explained by the proposed sequence of events and indicates a more complex evolutionary history of the Clyll operon in these serotypes. The Clyll operon of A. pleuropneumoniae is strikingly different from other RTX operons because it does not include intact genes for the B and D transporter proteins. This conclusion confirms the results of Chang et al. (2a) and Anderson et al. (1), who provided evidence that in the reference strains of serotypes 5 and 7 transporter BD genes are not linked to CA genes. DNA hybridization studies with the genomes of the 12 reference strains indicate that sequences similar to the clylBD genes are present in all reference strains, except in the reference strains of serotypes

636

JANSEN ET AL.

3 and 6. It might be very well that in all serotypes, except 3 and 6, translation products of the ClyI operon act in trans and account for the transmembrane transport of Clyll. Detailed Southern hybridization indicated that the reference strains of serotypes 1, 5, 9, 10, and 11 do contain intact ClyI operons consisting of the four genes C, A, B, and D, whereas the reference strains of serotypes 2, 4, 6, 7, 8, and 12 contain Clyl operons in which a major part of the C and A genes have been deleted (11). The transporter proteins for ClylI of serotypes 3 and 6, however, remain to be identified. To our knowledge the proposed complementation of the Clyll transporter genes by those of the Clyl operon is the first indication that RTX transporter genes are exchangeable in a naturally occurring organism. The data presented in this report enhance our insight into the structure and the genetic organization of the Clyll determinants of A. pleuropneumoniae. We believe that Clyll plays an important role in porcine pleuropneumonia, since preliminary results from our group indicate that pneumonic lesions can be induced in pig lungs by culture supernatants of recombinant E. coli cells that secrete ClyII (12). Additional evidence for the importance of ClyIl is the observation that all serotypes, except serotype 10, secrete this cytolysin and that serotypes 7 and 12 secrete Clyll as the only cytolysin. If Clyll is important for pathogenesis, it might be an important target for protective antibodies. The high degree of similarity between the ClyII proteins of the various serotypes might therefore facilitate the development of a subunit vaccine; immunization with the Clyll protein of one serotype will induce antibodies directed against ClyII of all serotypes. The high degree of similarity between the clyIICA genes may also be of value in diagnosing A. pleuropneumoniae infections; with the aid of DNA probes it might be possible to detect all A. pleuropneumoniae serotypes, except serotype 10. In view of these prospects, it should be kept in mind that we have analyzed only the reference strains of the 12 serotypes, but there is good evidence that field strains produce the same cytolysins as the reference strains (6, 14). ACKNOWLEDGMENTS We thank Johan Popma for culturing A. pleuropneumoniae and Arno Gielkens and Hilde Smith for their contributions to the

manuscript. REFERENCES 1. Anderson, C., G. F. Gerlach, A. A. Potter, S. Klashinsky, and P. J. Willson. 1991. Molecular characterization of an Actinobacillus pleuropneumoniae serotype 7 cytolysin gene, abstr. B-46, p. 33. Abstr. 91st Gen. Meet. Am. Soc. Microbiol. 1991. American Society for Microbiology, Washington, D.C. 2. Chang, Y., R. Young, and D. K. Struck. 1989. Cloning and characterization of a hemolysin gene from Actinobacillus (Haemophilus) pleuropneumoniae. DNA 8:635-647. 2a.Chang, Y., R. Young, and D. K. Struck. 1991. The Actinobacillus pleuropneumoniae hemolysin determinant: unlinked appCA and appBD loci flanked by pseudogenes. J. Bacteriol. 173:51515158. 3. Devenish, J., S. Rosendal, and J. T. Bosse. 1990. Humoral

INFECT. IMMUN.

4.

5.

6. 7.

8.

9.

10. 11. 12. 13.

14.

15.

16. 17. 18.

19.

20. 21.

antibody response and protective immunity in swine following immunization with the 104-kilodalton hemolysin of Actinobacillus pleuropneumoniae. Infect. Immun. 58:3829-3832. Felmlee, T., S. Pellett, and R. A. Welch. 1985. Nucleotide sequence of an Escherichia coli chromosomal hemolysin. J. Bacteriol. 163:94-105. Frey, J., R. Meier, and J. Nicolet. 1991. DNA sequence analysis of the hemolysin I gene of Actinobacillus pleuropneumoniae serotype 1 strain 4074, abstr. 266, p. 48. Annu. Meet. Conference of Research Workers in Animal Disease. Frey, J., and J. Nicolet. 1990. Hemolysin patterns of Actinobacillus pleuropneumoniae. J. Clin. Microbiol. 28:232-236. Gygi, D., J. Nicolet, J. Frey, M. K. Cross, V. Koronakis, and C. Hughes. 1990. Isolation of the Actinobacillus pleuropneumoniae hemolysin gene and the activation and secretion of the prohemolysin by the HlyC, HlyB and HlyD proteins of Escherichia coli. Mol. Microbiol. 4:123-128. Innis, M. A., and D. H. Gelfand. 1990. Optimization of PCR's, p. 3-12. In M. A. Innis, D. H. Gelfand, J. J. Sninsky, and T. J. White (ed.), PCR protocols, a guide to methods and applications. Academic Press, Inc., San Diego, Calif. Issartel, J. P., V. Koronakis, and C. Hughes. 1991. Activation of Escherichia coli prohaemolysin to the mature toxin by acyl carrier protein-dependent fatty acylation. Nature (London) 351: 759-761. Jansen, R., and F. D. Ledley. 1989. Production of discrete high specific activity DNA probes using the polymerase chain reaction. Gene Anal. Technol. 6:79-83. Jansen, R., et al. Unpublished data. Jansen, R., et al. Unpublished data. Kamp, E. M., J. K. Popma, J. Anakotta, and M. A. Smits. 1991. Identification of hemolytic and cytotoxic proteins of Actinobacillus pleuropneumoniae by using monoclonal antibodies. Infect. Immun. 59:3079-3085. Kamp, E. M., and L. A. M. G. van Leengoed. 1990. Serotyperelated differences in production and type of heat-labile hemolysin and heat-labile cytotoxin of Actinobacillus (Haemophilus) pleuropneumoniae. J. Clin. Microbiol. 27:1187-1191. Koronakis, V., M. Cross, and C. Hughes. 1988. Expression of the E. coli hemolysin secretion gene hlyB involves transcript anti-termination within the Hly operon. Nucleic Acids Res. 16:4789-4800. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Ochman, H., A. S. Gerber, and D. L. Hartl. 1988. Genetic applications of an inverse polymerase chain reaction. Genetics 120:621-625. Rycroft, A. N., D. Williams, J. M. Cullen, and J. Macdonald. 1991. The cytotoxin of Actinobacillus pleuropneumoniae (pleurotoxin) is distinct from the haemolysin and is associated with a 120-kDa polypeptide. J. Gen. Microbiol. 137:561-568. Sebunya, T. N. K., and J. R. Saunders. 1983. Haemophilus pleuropneumoniae infection in swine: a review. J. Am. Vet. Med. Assoc. 182:1331-1337. Smits, M. A., J. Briaire, R. Jansen, H. Smith, E. M. Kamp, and A. Gielkens. 1991. Cytolysins of Actinobacillus pleuropneumoniae serotype 9. Infect. Immun. 59:4497-4504. Strathdee, C. A., and R. Y. C. Lo. 1989. Cloning, nucleotide sequence, and characterization of genes encoding the secretion function of the Pasteurella haemolytica leukotoxin. J. Bacteriol. 171:916-928.

Comparison of the cytolysin II genetic determinants of Actinobacillus pleuropneumoniae serotypes.

Cytolysins (Cly) I, II, and III are toxins secreted by Actinobacillus pleuropneumoniae. These toxins are thought to play an important role in the path...
2MB Sizes 0 Downloads 0 Views