Detection and Identification of Actinobacillus pleuropneumoniae ...

JOURNAL OF CLINICAL MICROBIOLOGY, June 1998, p. 1704–1710 0095-1137/98/$04.0010 Copyright © 1998, American Society for Microbiology

Vol. 36, No. 6

Detection and Identification of Actinobacillus pleuropneumoniae Serotype 5 by Multiplex PCR TERRY M. LO, CHRISTINE K. WARD,

AND

THOMAS J. INZANA*

Center for Molecular Medicine and Infectious Diseases, Virginia-Maryland Regional College of Veterinary Medicine, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061-0342 Received 3 October 1997/Returned for modification 28 January 1998/Accepted 20 March 1998

Serotyping of Actinobacillus pleuropneumoniae is based on detection of the serotype-specific capsular antigen. However, not all isolates can be serotyped, and some may cross-react with multiple serotyping reagents. To improve sensitivity and specificity of serotyping and for early detection, a multiplex PCR assay was developed for detection of A. pleuropneumoniae and identification of serotype 5 isolates. DNA sequences specific to the conserved export and serotype-specific biosynthesis regions of the capsular polysaccharide of A. pleuropneumoniae serotype 5 were used as primers to amplify 0.7- and 1.1-kb DNA fragments, respectively. The 0.7-kb fragment was amplified from all strains of A. pleuropneumoniae tested with the exception of serotype 4. The 0.7-kb fragment was not amplified from any heterologous species that are also common pathogens or commensals of swine. In contrast, the 1.1-kb fragment was amplified from all serotype 5 strains only. The assay was capable of amplifying DNA from less than 102 CFU. The A. pleuropneumoniae serotype 5 capsular DNA products were readily amplified from lung tissues obtained from infected swine, although the 1.1-kb product was not amplified from some tissues stored frozen for 6 years. The multiplex PCR assay enabled us to detect A. pleuropneumoniae rapidly and to distinguish serotype 5 strains from other serotypes. The use of primers specific to the biosynthesis regions of other A. pleuropneumoniae serotypes would expand the diagnostic and epidemiologic capabilities of this assay. Actinobacillus pleuropneumoniae is the etiologic agent of swine pleuropneumonia, which is highly contagious and may result in high herd mortality (24). Current attempts to control the disease have been unsuccessful and have resulted in large economic losses to the swine industry (1, 12, 16). Because the disease can be transmitted quickly throughout the herd, early detection of this bacterium is important for control and treatment of the disease (3, 30). A. pleuropneumoniae has been classified into two biotypes and 14 serotypes that vary in virulence (15, 17, 20, 21) and predominance in different geographic regions (3, 23). The serotype specificity of A. pleuropneumoniae is determined by the capsular polysaccharide present on its surface (10, 15, 18). Serotyping is important for understanding how the disease is spread, for treatment and prevention, and for epidemiologic monitoring of the serotypes present in herds and/or regions. In the United States, serotype 5 of A. pleuropneumoniae is one of the most commonly isolated (23). Several serologic assays have been developed for serotyping of A. pleuropneumoniae, but the specificities of these assays vary considerably. Cross-reactivities between serotypes 1 and 9, serotypes 3, 6, and 8, serotypes 1 and 5, and serotypes 4 and 7 have been reported (9, 13, 14). These cross-reactions are most likely due to shared species-specific antigens such as lipopolysaccharide or membrane proteins (18). PCR has become a powerful and increasingly popular tool in microbial identification (19). The capability of PCR to detect genetic sequences from minute quantities of DNA is advantageous compared to serologic forms of detection for several reasons: cross-reactions between antigen and antibody are avoided, strains that have been previously characterized as untypeable due to autoagglutination may be typeable by PCR,

amplification of DNA by PCR makes it extremely sensitive, and PCR can be performed directly on samples without a wait for culture of the bacteria. Genes involved in A. pleuropneumoniae serotype 5 capsular polysaccharide export (cpx) and polysaccharide biosynthesis (cps) were recently cloned and sequenced by Ward and Inzana (28, 29). The cpx genes are highly conserved (4), whereas the cps genes are serotype specific (28). Based on the conserved nature of the cpx genes and the serotype specificity of the cps genes, we sought to determine whether DNA primers from these regions could be used in a multiplex PCR assay to reveal the presence of A. pleuropneumoniae in samples and, if present, whether the strain could be identified as serotype 5. MATERIALS AND METHODS Bacterial strains and culture. All bacterial strains used in this study are described in Table 1. K17-C and J45-C are nonencapsulated mutants of A. pleuropneumoniae serotype 5 produced by chemical mutagenesis. J45-100 is a nonencapsulated mutant of serotype 5 strain J45 that lacks a portion of its cps region due to homologous recombination (28). All bacterial strains were grown at 37°C on brain heart infusion (BHI) agar plates (Difco Laboratories, Detroit, Mich.) containing 5 mg of NAD per ml (BHI-NAD) or in BHI-NAD broth. Tissue and nasal swab samples. Lung tissue samples were taken from pigs challenged intratracheally with 5 3 107 CFU of A. pleuropneumoniae serotype 5 (11) and had been stored at 220°C since 1991. DNA isolation. Genomic A. pleuropneumoniae DNA was purified as previously described (29). Briefly, A. pleuropneumoniae log-phase cells were suspended in 10 mM Tris–1 mM EDTA, pH 8.0, and incubated at 37°C for 1 h in 0.66% sodium dodecyl sulfate and 100 mg of RNase per ml. The suspension was then incubated an additional hour at 56°C with 100 mg of proteinase K per ml, and the DNA was purified by repeated phenol-chloroform (Amresco Inc., Solon, Ohio) extraction. Genomic DNA was precipitated from the aqueous phase by adding 0.3 volumes of 3 M sodium acetate and 2.5 volumes of 95% ethanol. The DNA precipitate was then dried and resuspended in sterile water. Multiplex PCR sample preparation. DNA from whole bacterial cells for PCR amplification was extracted by suspending a loopful of colonies in 100 ml of water and lysing the suspended cells by heating them at 100°C for 10 min. The lysed cells were centrifuged, and the supernatant containing the bacterial DNA was removed and frozen at 220°C until needed. For extraction of bacterial DNA from lung samples, the tissue was sectioned into thin slices approximately 2 mm long. The sectioned tissue was then mashed and vortexed in 1 ml of water and

* Corresponding author. Mailing address: 1410 Prices Fork Rd., CMMID, College of Veterinary Medicine, Virginia Tech, Blacksburg, VA 24061-0342. Phone: (540) 231-4692. Fax: (540) 231-3426. E-mail: [email protected]. 1704

A. PLEUROPNEUMONIAE DETECTION BY PCR

VOL. 36, 1998 TABLE 1. Bacterial strains and their sources Species

A. pleuropneumoniae

A. suis H. parasuis S. suis P. multocida B. bronchiseptica

Serotype

Strain

Source a

1 2 3 4 5 5 5

4074 27089 27090 33378 K17 K17-C J45

5 5 5

J45-C J45-100 178

6 7 8

33590 53 405

9

13261

10 11 12 NDb ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND

D13039 56153 8329 39070-95 5091-96 2020-96 1061-96 47592-95 3079-96 42663-95 44398-95 3515-96 42267-95 5372-96 4534-96 340-96 27902-96 4169-94 47549-95 WA447301-95 416694 HP 588-1 721 NJ

ATCC ATCC ATCC ATCC ATCC T. Inzana B. Fenwick, Kansas State University T. Inzana T. Inzana M. Mulks, Michigan State University ATCC M. Mulks K. Mittal, University of Montreal J. Nicolet, University of Bern K. Mittal J. Nicolet K. Mittal K. Post, RDLc K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL K. Post, RDL WADDLd WADDL

NAe ND ND ND NA

33415 Field isolate Field isolate Field isolate PI-53

ATCC K. Post, RDL K. Post, RDL K. Post, RDL H. Veit, VPI&SUf

a

ATCC, American Type Culture Collection, Manassas, Va. ND, not determined. RDL, Rollins Diagnostic Laboratory, Raleigh, N.C. d WADDL, Washington Animal Disease Diagnostic Laboratory, Pullman, Wash. e NA, not applicable. f VPI&SU, Virginia Polytechnic Institute and State University. b c

boiled at 100°C for 10 min. The extract was centrifuged, and the supernatant containing the DNA was removed and frozen at 220°C until used. DNA primers. Oligonucleotide primers were selected by using DNAStar (Madison, Wis.) Primer Select software. Primers A and B were designed from the cps region of A. pleuropneumoniae serotype 5; primers C and D were designed from the cpx region of the same A. pleuropneumoniae strain (28, 29). The

1705

primers were selected based on the following properties: primer length, product length, product location, hairpin formations, dimer formations, and annealing temperature. The sequences for the oligonucleotide primers are given in Table 2. PCR. PCRs were performed in a total volume of 50 ml and were based on the procedure described by Saiki et al. (22). Master mixes for PCR were made fresh in batches of 750 ml. Each final reaction mix contained 10 mM Tris; 50 mM KCl; 2 mM MgCl2; 400 mM (each) dATP, dCTP, dGTP, and dTTP; and 480 mM (each) primers A and B or primers C and D. Five microliters of template DNA thawed at room temperature was added to 45 ml of the master mix for each reaction. The samples were then overlaid with 50 ml of mineral oil to prevent evaporation. The PCR assays were performed in an Omnigene thermal cycler (Hybaid Unlimited). The template DNA was denatured at 94°C for 2 min, and then 2.5 U of Taq polymerase (Fisher Scientific, Atlanta, Ga.) was added. A total of 30 cycles of PCR were performed, with each cycle consisting of 1 min of denaturation at 94°C, 2 min of annealing at 54°C, and 2 min of extension at 72°C. Twelve microliters of each PCR mixture was then loaded into a 0.7% agarose gel containing 0.5% ethidium bromide. Following electrophoresis the products were visualized by exposure to UV light. PCR sensitivity. Ten microliters of strain J45 log-phase broth culture (harvested at 109 CFU/ml) was added to 990 ml of sterile water and vortexed. Log10 serial dilutions of bacteria were made to obtain dilutions containing 106 to 102 CFU/ml. Bacterial concentrations were confirmed by viable plate counts. The diluted samples were then boiled for 10 min, and 5 ml of each dilution was used as a DNA template for PCR amplification. Southern blotting. Southern blotting was done as described previously (26). DNA was covalently linked to the nylon membranes by UV irradiation with a UV Stratalinker (Stratagene, La Jolla, Calif.). The hybridizations were performed under high stringency at 68°C and in the presence of 53 SSC (13 SSC is 0.15 M NaCl plus 0.015 M sodium citrate). Five microliters of a cpx probe was added to 25 ml of prehybridization buffer for use as the hybridization solution for cpx products. The cpx probe was manufactured by PCR and labeled with digoxigenin11-UTP (Genius System; Boehringer Mannheim, Indianapolis, Ind.). The same cpx primers that were used for the multiplex PCR assay were used to produce an identical 0.7-kb fragment. PCR conditions were as described above, except that 1.5 ml of 20 mM cpx forward primer, 1.5 ml of 20 mM cpx reverse primer, and 10 ml of A. pleuropneumoniae template DNA were used. The PCR product was verified by gel electrophoresis. Following Southern transfer, the membranes were washed and developed as recommended by the manufacturer. Latex agglutination test. The latex agglutination test was used to identify serotypes 1, 5, and 7 from A. pleuropneumoniae field isolates, as described by Inzana (8).

RESULTS PCR standardization and optimization. Two pairs of primers were designed from the sequenced DNA of the A. pleuropneumoniae serotype 5 cpx and cps capsule locus. The four primers are listed in Table 2. Primers A and B were designed to amplify a portion of the serotype-specific cps region, while primers C and D amplified part of the serotype-conserved cpx region. Amplification of purified A. pleuropneumoniae J45 genomic DNA with cps primers A and B or with cpx primers C and D resulted in a single 1.1-kb band and a 0.7-kb band, respectively. When J45 genomic DNA was amplified with both the cpx and cps sets of primers, both the 1.1- and 0.7-kb bands were detected (Fig. 1). Both products were consistent with the sizes predicted from the sequenced data of those regions. The multiplex PCR assay was optimized by using extracts of serotype 5 bacteria. The magnesium concentrations used for the assay varied from 1 to 7.5 mM; annealing temperatures varied from 49 to 55°C. These conditions did not seem to affect the amplification of the 1.1-kb cps or 0.7-kb cpx band, as no differences in the intensity of the bands were observed when

TABLE 2. Primer sequences used for multiplex PCR Name

A B C D

Function

Forward primer (biosynthesis region) Reverse primer (biosynthesis region) Forward primer (export region) Reverse primer (export region)

Sequence

Primer size (bp)

Product size (bp)

59-TTTATCACTATCACCGTCCACACCT-39 59-CATTCGGGTCTTGTGGCTACTAAA-39 59-TGGCGATACCGGAAACAGAGTC-39 59-GCGAAAGGCTATGGTATGGGTATGG-39

25 23 22 24

1,114 715

1706

LO ET AL.

FIG. 1. Agarose gel electrophoresis of PCR products from A. pleuropneumoniae serotype 5 genomic DNA. Lanes: 1, 1-kb DNA ladder; 2, cps primers A and B; 3, cpx primers C and D; 4, primers A, B, C, and D.

these parameters were varied. Initially, the assay conditions for amplifying serotype 5 DNA were 5 mM MgCl2 and an annealing temperature of 54°C. The multiplex PCR assay was then used to amplify DNA from whole cells of the type strains of all 12 A. pleuropneumoniae serotypes. The following parameters were examined to optimize the multiplex PCR assay for all serotypes: annealing temperature, primer concentration, Taq polymerase concentration, and MgCl2 concentration. Nonspecific products were observed from some serotypes when the conditions described above for serotype 5 were used. For serotypes 2, 3, and 6, a prominent band of about 1.2 kb was produced. Amplification of serotype 4 DNA produced a prominent band that was approximately 1.3 kb in size, but the 0.7-kb cpx product was not amplified (Fig. 2). In addition, faint nonspecific bands of various sizes were often amplified from these and some other serotypes. Serotypes 5, 9, 11, and 12 appeared to produce more intense bands than the other serotypes. Lowering the concentration of MgCl2 resulted in progressively fainter to nondetectable cpx bands for serotypes 1, 2, 3, 6, 7, 8, and 10, whereas the effects of lowering the MgCl2 concentration on serotypes 5, 9, 11, and 12 were less noticeable (data not shown). The 0.7-kb cpx band was not amplified from serotype 4 under any conditions. Annealing temperatures varied from 49 to 56°C. Increasing the annealing temperature to eliminate the nonspecific products also resulted in the loss of specific PCR products and therefore did not help to optimize the assay. In order to eliminate the nonspecific bands amplified from serotypes 2, 3, 4, 6, and others without eliminating the 0.7-kb cpx band, the concentrations of cps primers A and B were systematically reduced from starting concentrations of 480 mM (each) to final concentrations of 25 mM (each). Although the 1.2-kb band in serotypes 2, 3, and 6 showed a marked reduction of amplification, the intensity of the 1.1-kb band from serotype 5 was also substantially reduced (data not shown). The inability to eliminate the nonspecific bands by reducing the cps primer concentration suggested that the nonspecific bands may have been amplified by the cpx primers. To determine which primers were responsible for generating the nonspecific bands, DNA from serotypes 2 and 4 were amplified with the following combinations of primers: A and B, C and D, A and D, and C and B. A band similar in size to the 1.1-kb cps product was amplified from serotype 2 with primer combinations A and B, C and D, and C and B. Similarly, the 1.3-kb

J. CLIN. MICROBIOL.

band produced from serotype 4 was amplified from one cps primer, primer A, and one cpx primer, primer D (data not shown). This indicated that the nonspecific band amplified from serotypes 2 and 4 DNA was being amplified from cpx primers as well as cps primers. The effect of the Taq polymerase concentration on PCR products was examined in volumes of 50 and 100 ml. At 2.5 U of Taq polymerase/100 ml of reaction volume, no amplified products were visible. However, at 7.5 U of Taq polymerase/ 100 ml of reaction volume, the cpx band was visible as well as a faint nonspecific band similar in size to the 1.1-kb cps product. When 2.5 U of Taq polymerase/50 ml of reaction volume was used, the cpx band was amplified without amplification of nonspecific products. At 7.5 U of Taq polymerase/50 ml of reaction volume, both the cpx band and nonspecific bands were clearly visible (data not shown). These results indicate that the concentration of Taq polymerase had a pronounced effect on nonspecific amplification of PCR products. Increasing the MgCl2 concentration also resulted in an increase of nonspecific products for serotypes 2 and 4. At a concentration of 2 mM MgCl2, only the cpx band was visible. However, as the MgCl2 concentration was increased from 2 to 5 mM, there was a noticeable increase in nonspecific banding. When the concentration of MgCl2 was greater than 5 mM, the amount of nonspecific bands began to decrease. Thus, the optimum specificity of the multiplex PCR assay with bacterial samples required 2 mM MgCl2, while maximum sensitivity required 5 mM MgCl2. The optimum conditions for the multiplex PCR assay with samples of bacterial cells were therefore determined to be a mixture of 2.5 U of Taq DNA polymerase, 10 mM Tris-HCl, 50 mM KCl, 2 mM MgCl2, 400 mM (each) deoxynucleoside triphosphates, 480 mM (each) primer, and 5 ml of bacterial sample in a final volume of 50 ml. Assay sensitivity. The sensitivity of the multiplex PCR was determined for strain J45 whole bacterial cells. The bacteria were harvested in broth at mid-log phase to minimize the number of dead cells present in the sample. Both the 0.7-kb cpx and 1.1-kb cps PCR products were visualized by gel electrophoresis from at least 102 CFU/reaction (data not shown). There was not any detectable increase or decrease in sensitivity or quantity of amplified product when reagents such as sodium

FIG. 2. Agarose gel electrophoresis of PCR products amplified from whole cells of serotypes 1 through 6 at a MgCl2 concentration of 5 mM. Lanes: 1, 1-kb ladder; 2 through 7, PCR products from serotypes 1 through 6, respectively, amplified with primers A, B, C, and D.

VOL. 36, 1998

FIG. 3. Agarose gel electrophoresis of PCR products amplified from whole cells of encapsulated and nonencapsulated serotype 5 strains. Lanes: 1, 1-kb ladder; 2, encapsulated strain K17; 3, chemically induced nonencapsulated mutant K17-C; 4, nonencapsulated recombinant mutant J45-100 containing a large deletion in cps; 5, chemically induced nonencapsulated mutant J45-C; 6, encapsulated strain J45.

dodecyl sulfate or lysozyme were used in the extraction process. Centrifugation of the crude extract after the cells were boiled was important to avoid decreased sensitivity in this PCR assay. In addition, detection of PCR products was diminished from cells that had been grown in BHI broth unless they were washed prior to preparation. Other methods used to improve the specificity of the primers, such as varying the annealing temperature, were found to be ineffective. Serotype and species specificity. Amplification of serotype 5 DNA from strains J45, J45-C, K17, and K17-C by multiplex PCR produced both cps and cpx bands. However, only the cpx band was amplified from DNA of J45-100, a nonencapsulated knockout mutant lacking part of the cps region (28), confirming that the cps primers were amplifying DNA from the cps locus (Fig. 3). The multiplex PCR assay was used to amplify DNA from the reference strains of 12 A. pleuropneumoniae serotypes (Fig. 4). With the exception of serotype 4, a band of 0.7 kb, consistent with the cpx product, was amplified from all serotypes. However, the 1.1-kb band was amplified only from serotype 5 DNA. No amplified products were observed for serotype 4 under optimum conditions. Previously untyped field isolates were assayed by both PCR


1707

and the latex agglutination test (Table 3). All of the eight strains typed as serotype 5 were also identified by multiplex PCR as serotype 5 strains, determined by amplification of the distinct cps and cpx products. PCR assays were also performed on six strains that were typed as serotype 1 and six strains that were typed as serotype 7 by latex agglutination. The cpx product was amplified by all 12 strains, identifying them as A. pleuropneumoniae, but the cps product was not amplified from any of these strains, confirming they were not serotype 5. In addition, one strain that was not typed as serotype 1, 5, or 7 by latex agglutination was determined to be A. pleuropneumoniae by amplification of only the cpx product by multiplex PCR. Species specificity of the multiplex PCR was examined by applying the assay to five other swine respiratory pathogens (Fig. 5): Actinobacillus suis, Bordetella bronchiseptica, Haemophilus parasuis, Pasteurella multocida, and Streptococcus suis. No amplified products were made from DNA of any of these species when the optimum conditions for specificity were used. However, when 5 mM MgCl2 was used in the assay, nonspecific bands were amplified from H. parasuis, P. multocida, and S. suis DNA. These bands were distinct in size from the 0.7-kb cpx and 1.1-kb cps products (data not shown). Southern hybridization was performed on amplified products from H. parasuis, P. multocida, and S. suis DNA, as well as from A. pleuropneumoniae serotypes 2 and 4 DNA, to further examine the specificity of products amplified when 5 mM MgCl2 was used in the multiplex PCR assay (Fig. 6a). The bands were probed under high-stringency conditions with a labeled PCR product generated by cpx primers C and D from A. pleuropneumoniae serotype 5. The probe did not hybridize to any of the bands produced by serotype 4 or the three nonA. pleuropneumoniae species (Fig. 6b), but it did hybridize to the 0.7-kb cpx bands produced by serotypes 2 and 5. Amplification of DNA from clinical specimens. The multiplex PCR was used on lung tissue from swine infected with A. pleuropneumoniae serotype 5 to determine if the assay could be used for rapid diagnosis of clinical disease. Both the cpx and the cps products were amplified from lung tissue samples of two pigs that had recently been infected with strain J45 (data not shown). Frozen lung tissue samples from eleven serotype 5-challenged pigs that had been used for an immunization

FIG. 4. Agarose gel electrophoresis of bacterial samples of serotypes 1 through 12 at a MgCl2 concentration of 2 mM. Lanes: 1, 1-kb DNA ladder; 2 through 7, PCR products from serotypes 1 through 6, respectively; 8 through 14, PCR products from serotypes 5 and 7 through 12, respectively; 15, 1-kb DNA ladder. All products were amplified with primers A, B, C, and D.

1708

LO ET AL.

J. CLIN. MICROBIOL.

TABLE 3. Serotyping of A. pleuropneumoniae isolates Serotype by latex agglutinationa

Strain

5091-96 2020-96 3079-96 44398-95 27902-96 47549-95 721 NJ 178 3515-96 340-96 4169-94 WA447301-95 416694 HP 588-1 39070-95 1061-96 47592-95 42663-95 42267-95 5372-96 4534-96

Multiplex PCR product

1

5

7

cpx

cps

2 2 2 2 2 2 2 2 41 41 41 41 41 41 2 2 2 2 2 2 2

31 41 41 41 31 31 21 31 2 2 2 2 2 2 2 2 2 2 2 2 2

2 2 2 2 2 2 2 2 2 2 2 2 2 2 31 31 31 31 31 31 2

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2

a 41, rapid, 100% agglutination of beads with ring formation; 31, about 75% agglutination of beads with some ring formation; 21, about 50% agglutination of beads with little ring formation; 2, no detectable agglutination of beads compared to negative control.

study 6 years earlier (11) were also assayed by PCR. The cpx product was amplified from each of these samples, indicating that A. pleuropneumoniae DNA was present. However, the cpx band from five of these samples was weak, and the cps product was not amplified from four samples (Fig. 7). DISCUSSION Serotypes of A. pleuropneumoniae are distinguished by their unique capsular polysaccharide (10, 18). Because cross-reactions often occur in serotyping with traditional serologic assays (9, 13, 14). PCR offers a practical alternative that does not employ antigens or antibodies. This attribute makes capsule genes an ideal target for typing by PCR. Furthermore, culture of A. pleuropneumoniae may be successful for 50% or fewer of the specimens submitted. When pigs die before specimens can be collected, contamination usually prevents any isolation of A. pleuropneumoniae (27). PCR, however, is more sensitive than culture for this bacterium (5, 6), and the multiplex PCR described here is specific enough not to be affected by the presence of contaminants. Failing to isolate the agent for susceptibility testing should not be of concern because A. pleuropneumoniae is susceptible to most broad-spectrum antibiotics (3). Falla et al. (2) previously reported using primers from the capsular DNA region as a reliable method for typing of Haemophilus influenzae by PCR. To date, a reliable method for serotyping of A. pleuropneumoniae by PCR has not yet been established. Hennessy et al. (7) have reported using an arbitrarily primed PCR assay for serotyping of A. pleuropneumoniae, but this method has the disadvantage of requiring pure bacterial samples for testing and is highly susceptible to contamination. Sirois et al. (25) described the use of uncharacterized primers for amplification of A. pleuropneumoniae DNA. Although this assay was specific for swine pathogens, it could not discriminate between A. pleuropneumoniae biotype 1, biotype 2, and Actinobacillus lignieresii, and the sensitivity of the assay was not

determined. The same primers were later used to detect A. pleuropneumoniae from tonsils, with a sensitivity of 103 CFU/reaction tube (5). Gram et al. (6) recently described a species-specific PCR assay for amplification of an A. pleuropneumoniae outer membrane lipoprotein that has improved specificity. However, culture is still required with either assay to identify the serotype, which is needed for epidemiology, herd management, and monitoring of the spread or introduction of new strains. The present study describes the first use of primers to amplify conserved and serotype-specific capsular DNA regions to simultaneously identify A. pleuropneumoniae and the serotype. Both the cps and cpx regions of serotype 5 DNA were successfully amplified with samples of purified DNA, bacterial colonies, and lung tissue. Combining the primers did not require any change in PCR conditions. The use of multiplex PCR provided the advantage of using multiple primer sets in a single reaction and simultaneously determining both the species and the serotype, in this case A. pleuropneumoniae and serotype 5. The detectable limit of the PCR products of serotype 5 by agarose gel electrophoresis was less than 102 CFU/reaction. The presence of nonspecific bands amplified from some serotypes, particularly serotypes 2, 3, and 6, while maintaining the cpx band in all serotypes, was initially problematic. Although the cpx region appears to be highly conserved, the primers selected from that region were designed from the sequence of serotype 5 capsular DNA. Because there is no sequence data available on the capsular export regions of other serotypes, the amount of homology of these primers to other serotypes is unknown. The MgCl2 concentration was the single most important parameter involved in the specificity of PCR amplification and was successfully used to control the presence of nonspecific bands. The successful application of the multiplex PCR assay to bacterial colonies provided an effective method of identifying and serotyping A. pleuropneumoniae. With the exception of the rare serotype 4, a distinct 0.7-kb band was amplified from all

FIG. 5. Agarose gel electrophoresis of PCR products of bacterial samples of respiratory swine pathogens amplified with primers A, B, C, and D. Lanes: 1, A. pleuropneumoniae serotype 5; 2, A. suis; 3, B. bronchiseptica; 4, H. parasuis; 5, P. multocida; 6, S. suis; 7, 1-kb DNA ladder.

VOL. 36, 1998


1709

FIG. 6. (a) Agarose gel electrophoresis of PCR products from bacterial samples used for Southern hybridizations. Lanes: 1, 1-kb ladder; 2, serotype 2; 3, serotype 4; 4, serotype 5; 5, S. suis; 6, P. multocida; 7, H. parasuis. (b) Southern blot of PCR products from bacterial samples hybridized with a cpx probe. Lanes: 1, serotype 2; 2, serotype 4; 3, serotype 5; 4, S. suis; 5, P. multocida; 6, H. parasuis.

serotypes. This PCR assay also confirmed results obtained by latex agglutination. Of 21 field isolates serotyped by latex agglutination, all 21 isolates were identified as A. pleuropneumoniae by the multiplex PCR assay. In addition, all strains that were identified as serotype 5 by latex agglutination were also identified as serotype 5 by multiplex PCR. Because results from agglutination tests can often be inconclusive or subject to interpretation, PCR can potentially be used to definitively type strains that are difficult to assay by agglutination. Furthermore, the multiplex PCR can confirm that nontypeable isolates are A. pleuropneumoniae. No amplification of DNA under conditions for optimal specificity was observed from any of the other swine respiratory pathogens tested, indicating that the multiplex PCR assay can be used to determine whether an infection is due to A. pleuropneumoniae. PCR is also relatively simple and can be performed in under 5 h. The amplification of the cpx product in all of the serotypes, with the exception of serotype 4, supports the existing evidence that the capsular export region is highly conserved among A. pleuropneumoniae serotypes (29). Southern blotting also indicated that the 0.7-kb band of serotype 2 contains homology to the 0.7-kb band of serotype 5. Although the 0.7-kb product was not amplified

from serotype 4, it is not surprising that even within a highly conserved region there may be some areas of nonhomology. However, cps primers A and B were very specific in amplifying a 1.1-kb band from only serotype 5 isolates. The multiplex PCR was also applied to testing of clinical specimens. The 1.1-kb cps and 0.7-kb cpx bands were amplified from lung tissue samples of swine infected with A. pleuropneumoniae serotype 5. The cpx product was amplified from all samples, although much less product was made from five of the samples. However, the cps product was not amplified from 4 of 11 samples. It is not clear why the 1.1-kb cps product was not amplified from these samples, but it may have been due to the samples being 6 years old. The DNA in this region may have degraded over time to the point that it could not be amplified. Alternatively, it should be noted that these samples were taken from a vaccination study. It is likely that efficacious vaccination enabled the host to clear the challenge infection quickly, substantially lowering the number of bacteria present in the lung. In conclusion, the multiplex PCR assay was effective in detecting A. pleuropneumoniae and identifying serotype 5 from whole bacterial cells and infected lung tissues. At present, autoagglutinating, cross-reacting, and nontypeable strains are dif-

FIG. 7. Agarose gel electrophoresis of PCR products from lung tissue samples taken from swine that had been infected with serotype 5. Lanes and strain numbers: 1, 204; 2, 205; 3, 207; 4, 209; 5, 211; 6, 215; 7, 216; 8, 220; 9, 221; 10, 223; 11, 228. Lane 12, 1-kb DNA ladder.

1710

LO ET AL.

J. CLIN. MICROBIOL.

ficult to identify by serologic assays. The use of multiplex PCR with primers directed to the capsular DNA regions can potentially be used for detection and serotyping of A. pleuropneumoniae with both high specificity and high sensitivity, avoiding the problems associated with serologic assays. Once the sequences for the capsular regions of other serotypes have been determined, the assay can be expanded to serotype any strain of A. pleuropneumoniae. ACKNOWLEDGMENTS We thank Eric Wong and N. Sriranganathan for helpful suggestions and advice; K. Mittal, Martha Mulks, Karen Post, Jacques Nicolet, and Brad Fenwick for providing bacterial strains; and Gretchen Glindemann, Mark Lawrence, and John McQuiston for technical assistance. This work was supported, in part, by grants from Solvay Animal Health and by HATCH formula funds to the Virginia State Agricultural Experiment Station. REFERENCES 1. Brandreth, S. R., and I. M. Smith. 1985. Prevalence of pig herds affected by pleuropneumonia associated with Haemophilus pleuropneumoniae in eastern England. Vet. Rec. 117:143–147. 2. Falla, T. J., D. M. Crook, L. N. Brophy, D. Maskell, J. S. Kroll, and E. R. Moxon. 1994. PCR for capsular typing of Haemophilus influenzae. J. Clin. Microbiol. 32:2382–2386. 3. Fedorka-Cray, P. J., L. Hoffman, W. C. Cray, J. T. Gray, S. A. Breish, and G. A. Anderson. 1993. Actinobacillus (Haemophilus) pleuropneumoniae. Part I. History, epidemiology, serotyping, and treatment. Compend. Contin. Ed. Practic. Vet. 15:1447–1455. 4. Frosch, M., U. Edwards, K. Bousset, B. Krause, and C. Weisgerber. 1991. Evidence for a common molecular origin of the capsule gene loci in gramnegative bacteria expressing group II capsular polysaccharides. Mol. Microbiol. 5:1251–1263. 5. Gram, T., P. Ahrens, and J. P. Nielsen. 1996. Evaluation of a PCR for detection of Actinobacillus pleuropneumoniae in mixed bacterial cultures from tonsils. Vet. Microbiol. 51:95–104. 6. Gram, T., P. Ahrens, and J. P. Nielsen. 1998. Improved diagnostic PCR assay for Actinobacillus pleuropneumoniae based on the nucleotide sequence of an outer membrane lipoprotein. J. Clin. Microbiol. 36:443–448. 7. Hennessy, K. J., J. J. Iandolo, and B. W. Fenwick. 1993. Serotype identification of Actinobacillus pleuropneumoniae by arbitrarily primed polymerase chain reaction. J. Clin. Microbiol. 31:1155–1159. 8. Inzana, T. J. 1995. Simplified procedure for preparation of sensitized latex particles to detect capsular polysaccharides: application to typing and diagnosis of Actinobacillus pleuropneumoniae. J. Clin. Microbiol. 33:2297–2303. 9. Inzana, T. J., G. F. Clark, and J. Todd. 1990. Detection of serotype-specific antibodies or capsular antigen of Actinobacillus pleuropneumoniae by a double-label radioimmunoassay. J. Clin. Microbiol. 28:312–318. 10. Inzana, T. J., and B. Mathison. 1987. Serotype specificity and immunogenicity of the capsular polymer of Haemophilus pleuropneumoniae. Infect. Immun. 55:1580–1587. 11. Inzana, T. J., J. Todd, and H. P. Veit. 1993. Safety, stability, and efficacy of

12. 13. 14. 15. 16. 17. 18.

19. 20. 21. 22. 23. 24. 25. 26.

27.

28.

29. 30.

noncapsulated mutants of Actinobacillus pleuropneumoniae for use in live vaccines. Infect. Immun. 61:1682–1686. MacInnes, J. I., and S. Rosendal. 1988. Prevention and control of Actinobacillus (Haemophilus) pleuropneumoniae infection in swine: a review. Can. Vet. J. 29:572–573. Mittal, K. R. 1990. Cross-reactions between Actinobacillus (Haemophilus) pleuropneumoniae strains of serotypes 1 and 9. J. Clin. Microbiol. 28:535– 539. Mittal, K. R., R. Higgins, and S. Lavriviere. 1987. An evaluation of agglutination and coagglutination techniques for serotyping of Haemophilus pleuropneumoniae isolates. Am. J. Vet. Res. 48:219–226. Nicolet, J. 1988. Taxonomy and serological identification of Actinobacillus pleuropneumoniae. Can. Vet. J. 29:578–579. Nicolet, J. 1992. Actinobacillus pleuropneumoniae, p. 401–408. In A. D. Leman, B. E. Straw, W. L. Mengeling, S. D’Allaire, and D. J. Taylor (ed.), Diseases of swine. Iowa State University Press, Ames. Niven, D. F., and M. Le´vesque. 1988. V-factor-dependent growth of Actinobacillus pleuropneumoniae biotype 2 (Bertschinger 2008/76). Int. J. Syst. Bacteriol. 38:319–320. Perry, M. B., E. Altman, J.-R. Brisson, L. M. Beynon, and J. C. Richards. 1990. Structural characteristics of the antigenic capsular polysaccharides and lipopolysaccharides involved in the serological classification of Actinobacillus pleuropneumoniae strains. Serodiagn. Immun. Infect. Dis. 4:299–308. Rodu, B. 1990. Molecular biology in medicine. The polymerase chain reaction: the revolution within. Am. J. Med. Sci. 299:210–216. Rogers, R. J., L. E. Eaves, P. J. Blackall, and K. F. Truman. 1990. The comparative pathogenicity of four serovars of Actinobacillus (Haemophilus) pleuropneumoniae. Aust. Vet. J. 67:9–12. Rosendal, S., D. Boyd, and K. A. Gilbride. 1985. Comparative virulence of porcine Haemophilus bacteria. Can. J. Comp. Med. 49:68–74. Saiki, R. K., D. H. Gelfand, S. Stoffel, S. J. Scharf, R. Higuchi, G. T. Horn, K. B. Mullis, and H. A. Erlich. 1988. Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239:487–491. Sebunya, T. K., and J. R. Saunders. 1983. Haemophilus pleuropneumoniae infections in swine: a review. J. Am. Vet. Med. Assoc. 182:1331–1337. Shope, R. R. 1964. Porcine contagious pleuropneumonia. I. Experimental transmission, etiology and pathology. J. Exp. Med. 1119:357–358. Sirois, M., E. G. Lemire, and R. C. Levesque. 1991. Construction of a DNA probe and detection of Actinobacillus pleuropneumoniae by using polymerase chain reaction. J. Clin. Microbiol. 29:1183–1187. Straus, W. M. 1993. Hybridization with radioactive probes, p. 6.3.1–6.3.6. In F. M. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl (ed.), Current protocols in molecular biology, vol. 1. John Wiley & Sons, New York, N.Y. Torrelorell, M., C. Pijoan, K. J. Janni, R. Walker, and H. S. Joo. 1997. Airborne transmission of Actinobacillus pleuropneumoniae and porcine reproductive and repiratory syndrome virus in nursery pigs. Am. J. Vet. Res. 58:828–832. Ward, C. K., and T. J. Inzana. 1996. Mutations in capsular polysaccharide biosynthesis render Actinobacillus pleuropneumoniae avirulent, abstr. 25, p. 34. Haemophilus, Actinobacillus, Pasteurella 96 International Conference, Acapulco, Mexico. Ward, C. K., and T. J. Inzana. 1997. Identification and characterization of a DNA region involved in the export of capsular polysaccharide by Actinobacillus pleuropneumoniae serotype 5a. Infect. Immun. 65:2491–2496. Willson, P. J., G. Falk, and S. Klashinsky. 1987. Detection of Actinobacillus pleuropneumoniae infection in pigs. Can. Vet. J. 28:111–116.