INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Bordetella bronchiseptica is a common respiratory pathogen in a number of animal species. It is an etiologic agent of swine atrophic rhinitis and bronchopneumonia, canine tracheobronchitis, and bronchopneumonia in laboratory and companion animals. In rare instances, B. bronchiseptica has been reported to infect humans; a majority of these infections have occurred in patients with underlying conditions such as cystic fibrosis, AIDS, Hodgkin's disease, or leukemia (2, 3, 8, 23). The types of infections seen in these patients have included pneumonia, tracheobronchitis, sinusitis, peritonitis, meningitis, and septicemia (2, 3, 8, 23).

Until recently, there has been a lack of a simple and reliable method for typing of B. bronchiseptica isolates for classification. Ribotyping has been utilized to characterize B. bronchiseptica isolates from several animal species and was shown to provide a basis for grouping these organisms into distinct types (13-15). Keil and Fenwick (5) utilized random amplified polymorphic DNA fingerprinting and ribotyping to evaluate the genetic diversity among 26 canine B. bronchiseptica isolates. Methods such as restriction enzyme analysis (REA) of chromosomal DNA may also have power in discriminating among B. bronchiseptica strains. In fact, REA and ribotyping have been utilized in molecular epidemiologic studies of other bacterial species (1, 4). We have previously reported that REA and ribotyping could be utilized to discriminate Bordetella avium and Bordetella hinzii isolates (17). In the present experiments, REA was utilized as a method for characterizing B. bronchiseptica isolates previously grouped on the basis of ribotyping. This study represents the first examination of the potential usefulness of REA as a method of classifying B. bronchiseptica isolates from several host species.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains. A total of 195 B. bronchiseptica isolates were examined (113 laboratory strains and 82 field isolates). Strains B58, B65, and 5203 (7) were obtained from Tibor Magyar, Veterinary Medical Research Institute of the Hungarian Academy of Sciences, Budapest, Hungary. Strain St. Louis was obtained from Tom Milligan, St. Louis University Hospital, St. Louis, Mo. Strains with the descriptor MBORD were generously provided by David Dyer, University of Oklahoma, Oklahoma City (11). The 113 laboratory strains utilized in the present study were obtained from 11 different host species from diverse geographic locations (Table 1). MBA-4, a bvg isogenic mutant of MBORD846 (produced in the laboratory of Jeff Miller, University of California Los Angeles), was kindly provided by David Dyer. Eighty-two field isolates were included from the following sources. B. bronchiseptica isolates obtained from seals during a phocine morbillivirus outbreak were supplied by Geoff Foster, Scottish Agricultural Colleges Veterinary Science Division, Drummonhill, United Kingdom (15). Thirty turkey isolates were kindly provided by Y. M. Saif, The Ohio State University, Wooster. Swine isolates from a field case of atrophic rhinitis were obtained from the Diagnostic Laboratory, Iowa State University College of Veterinary Medicine, Ames.

REA. (i) Chromosomal DNA isolation. Bacterial strains were grown on blood agar base slants (Difco, Detroit, Mich.) for 48 h at 37°C. Bacterial cells were harvested and adjusted to a similar concentration in 0.85 M NaCl. A 1.5-ml aliquot of the bacterial cells was centrifuged at 16,000 × g for 4 min. The supernatant was decanted; pellets were stored at 70°C. DNA was isolated using a commercially available kit (DNAzol; Gibco BRL, Gaithersburg, Md.) according to recommendations of the manufacturer.

(ii) Restriction enzyme digestion, electrophoresis, photography, and analysis. The following restriction enzymes (Gibco BRL) were examined: AluI, BglII, ClaI, DraI, DdeI, EcoRI, EcoRV, HaeIII, HhaI, HindIII, HinfI, HpaI, HpaII, MvaI, NciI, PvuII, PstI, RsaI, TaqI, and XbaI. Digestion of chromosomal DNA with each restriction enzyme was carried out via the recommendations of the manufacturer. The reactions were stopped by the addition of 5 µl of stop solution (0.25% bromophenol blue, 0.25% xylene cyanole, 25% Ficoll 400) to 21 µl of reaction mixture. The digested DNA fragments were electrophoresed in 0.7% agarose gels using TBE buffer (0.089 M Tris, 0.089 M boric acid, 2 mM EDTA, pH 8.0). A HindIII digest of lambda phage DNA was used as a molecular size marker. Gels were stained and photographed as previously described (17). Photographs were scanned for computer analysis using a Scanjet IIcx with DeskScan software (Hewlett-Packard, Boise, Idaho). GelCompar software (Applied Maths, Kortrijk, Belgium) was used for comparison of fingerprint profiles. Similarity between all possible pairs of fingerprint profiles using the coefficient of Dice (18) was calculated by the cluster analysis module of the software. Dendrograms were derived from a matrix of similarity values by the unweighted pair-group method using arithmetic averages.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

REA. Twenty restriction endonucleases were evaluated for use in REA of B. bronchiseptica isolates. Of the endonucleases examined, digestion of B. bronchiseptica chromosomal DNA with HinfI or AluI resulted in well-separated and distinct bands in the 4 to 10 kb molecular size range. Use of the other endonucleases resulted in bands which could not be readily distinguished, especially in the 3 to 23.1 kb molecular size range, where optimum resolution occurs under the electrophoresis conditions used in this study. Forty-eight distinct fingerprint profiles were found among the 195 B. bronchiseptica isolates following HinfI digestion. These isolates had previously been characterized into 19 distinct PvuII ribotypes (13-15). An example of the REA profiles of selected ribotype 3 B. bronchiseptica isolates is shown in Fig. 1. Based on HinfI restriction enzyme digestion patterns, dendrograms were constructed and similarity between the fingerprint profiles was calculated using the coefficient of Dice by the cluster analysis module of GelCompar software. Genetic diversity among B. bronchiseptica isolates was considerable, with similarity ranging from 68 to 97% (Fig. 2). Even within a host species, the diversity among isolates was striking. For example, there was less than 70% similarity between some swine isolates. Interestingly, the two human isolates clustered with B. bronchiseptica isolates obtained from birds.

View larger version (109K): [in this window] [in a new window]   FIG. 1.   Representative REA profiles of selected B. bronchiseptica isolates following HinfI restriction enzyme digestion of chromosomal DNA. These isolates were previously characterized as ribotype 3. Note that REA further differentiated these isolates which were previously characterized by ribotyping. Lanes M, molecular size marker (lambda phage HindIII digest).

View larger version (39K): [in this window] [in a new window]   FIG. 2.   Dendrogram showing percent similarity among B. brochiseptica isolates using HinfI restriction endonuclease digestion of chromosomal DNA. Representatives of each of the 48 profiles observed for the 195 B. bronchiseptica isolates are shown. Similarity between fingerprint profiles based on the coefficient of Dice was calculated by the cluster analysis module of GelCompar software.

View larger version (67K): [in this window] [in a new window]   FIG. 3.   Representative REA profiles of selected B. bronchiseptica isolates following AluI restriction enzyme digestion of chromosomal DNA. These isolates were previously characterized as ribotype 3. Note that REA further differentiated isolates which were previously characterized by ribotyping. Lanes M, molecular size marker (lambda phage HindIII digest).

View larger version (32K): [in this window] [in a new window]   FIG. 4.   Dendrogram showing percent similarity among B. brochiseptica isolates using AluI restriction endonuclease digestion of chromosomal DNA. Representatives of each of the 39 profiles observed for the 195 B. bronchiseptica isolates are shown. Similarity between fingerprint profiles based on the coefficient of Dice was calculated by the cluster analysis module of GelCompar software.

View larger version (64K): [in this window] [in a new window]   FIG. 5.   REA profile for bvg+ and bvg mutant B. bronchiseptica strains. Molecular size markers are in lanes marked with an M. Note: MBA-4 is a bvg isogenic mutant of MBORD846, whereas B65 is a spontaneous bvg mutant of B58. Lanes 1 to 4, MBA-4, MBORD846, B58, and B65, respectively.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

REA is a highly discriminatory method for determining phylogenetic relationships and has been utilized by previous investigators in examining the molecular epidemiology of genetically diverse strains. In our experiments, REA was utilized in the examination of 195 B. bronchiseptica isolates from 12 host animal species that had been previously characterized by ribotyping. Digestion of B. bronchiseptica chromosomal DNA with HinfI or AluI resulted in DNA fragments in the 4 to 10 kb molecular size range which were more readily distinguishable than fragments generated by digestion with the other restriction enzymes examined. Furthermore, we found that REA is superior to previously described techniques for distinguishing B. bronchiseptica isolates. In fact, as was shown for canine isolates (5), our results indicate that there is more genetic diversity among B. bronchiseptica isolates than previously appreciated. The fingerprint profiles generated by HinfI or AluI restriction enzyme digestion were stable following 25 in vitro passages and were not affected by differences in bvgAS expression.

Techniques utilized to distinguish among Bordetella isolates have included biochemical and physiological characteristics, whole-cell protein profiles, and fatty acid analysis (10, 16, 21, 22). Additional methods have relied on examining stable genetic elements for characterization of Bordetella isolates, which should be more reproducible than expression-based methods. Indeed, the utility of ribotyping in discriminating among B. bronchiseptica isolates has been proven (5, 13-15). In addition, Keil and Fenwick (5) examined random amplified polymorphic DNA fingerprinting, but this method has its limitations (20). Moreover, macrorestriction fingerprinting using the rare-cutting enzyme XbaI and pulsed-field gel electrophoresis has been utilized by investigators to characterize isolates of B. bronchiseptica (6, 19). However, pulsed field gel electrophoresis protocols typically involve time-consuming procedures for purification of genomic DNA in agarose, and lengthy restriction enzyme digests and electrophoresis times. In their study, Khattak and Matthews (6) had examined restriction fragment length polymorphism analysis of Bordetella species and found that it failed to discriminate among Bordetella pertussis, Bordetella parapertussis, or B. bronchiseptica isolates when chromosomal DNAs were cut with the frequently cutting enzyme EcoRI. Evidently, these investigators did not examine other restriction enzymes for use in restriction fragment length polymorphism analysis. In agreement with these investigators, we found that restriction enzyme digestion with EcoRI produced numerous bands in the 3 to 23.1 kb molecular size range such that discrimination among Bordetella isolates was not possible. Nonetheless, in our experiments we examined 20 different restriction enzymes and were able to demonstrate that digestion of chromosomal DNA using HinfI or AluI restriction endonucleases is useful in discriminating B. bronchiseptica isolates.

An inherent problem in the classification of Bordetella spp. based on phenotypic characteristics is the extensive alterations in expression that can occur depending upon Bvg phase. Thus, while mutations in bvgAS could influence the characterization of Bordetella isolates using expression-based methods, we have shown that alterations in bvgAS expression as a result of mutation do not affect the REA fingerprint profiles of B. bronchiseptica isolates following either HinfI or AluI restriction endonuclease digestion.

As stated above, previous investigators have shown the utility of ribotyping using PvuII to classify B. bronchiseptica isolates. We have shown in the present study that REA can also be utilized in characterizing B. bronchiseptica isolates. While REA generally provided more discriminatory power than ribotyping, there were examples where the use of ribotyping was more discriminatory than REA. Thus, since neither method is technically difficult, the combination of REA and ribotyping should prove useful in molecular epidemiological studies of Bordetella species and in the development of a reference typing system. We propose that B. bronchiseptica isolates be assigned a descriptive identification epithet based on fingerprint profiles generated by REA and ribotyping. Numerous fingerprint profiles could be analyzed and used to generate a database from which individual isolates could be easily assigned a descriptive identification epithet code.