Comparison of ribotyping and sequence-based typing for discriminating among isolates of Bordetella bronchiseptica

Authors: Register, Karen; Nicholson, Tracy; Brunelle, Brian

Submitted to: Journal of Microbiological Methods
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/15/2016
Publication Date: 8/16/2016
Publication URL: http://handle.nal.usda.gov/10113/5430224
Citation: Register, K.B., Nicholson, T.L., Brunelle, B.W. 2016. Comparison of ribotyping and sequence-based typing for discriminating among isolates of Bordetella bronchiseptica. Journal of Microbiological Methods. 129:117-126. doi: 10.1016/j.mimet.2016.08.008.

Interpretive Summary: Bordetella bronchiseptica is a widespread bacterial pathogen that infects a variety of domesticated and wildlife animal species. It causes tracheobronchitis (kennel cough) in dogs, bronchopneumonia in numerous laboratory, companion and wild animals, neonatal pneumonia in piglets and is an important contributor to swine atrophic rhinitis and porcine respiratory disease complex. The ability to identify isolates of B. bronchiseptica that are genetically distinct from one another is essential to establish its population structure and to understand how the bacterium spreads throughout and within the various hosts it infects. Among the methods that have been used for this purpose are multilocus sequence typing (MLST) and ribotyping. Both methods have been shown to perform well but they have never been directly compared to one another to determine which of them might be optimal. An accessory approach sometimes also applied is the comparison of DNA sequence from two repeat regions of the pertactin gene, which encodes one of the important virulence factors of B. bronchiseptica. The purpose of this study was to directly compare the discriminatory power of MLST and ribotyping using a single set of diverse B. bronchiseptica isolates and to ascertain whether subtyping based on DNA sequence from the pertactin gene repeat regions provides an additional level of discrimination. These three methods were used to evaluate 122 isolates of B. bronchiseptica sourced from the United States, Europe, Israel and Australia, representing 11 mammalian or avian hosts. MLST and ribotyping were found to be equally discriminatory and, in both cases, isolates could be further differentiated when DNA sequences from their pertactin gene repeat regions were compared. Because MLST offers several technical and practical advantages as compared to ribotyping, we conclude that MLST is the preferred method for typing isolates of B. bronchiseptica. Isolates can frequently be further differentiated when MLST is paired with analysis of the pertactin gene repeat region sequences. Data reported here demonstrate that this approach provides a highly discriminatory tool suitable for evaluating the population structure and investigating the epidemiology of the bacterium, thereby facilitating efforts to control its spread and to minimize its impact on human and animal health and animal production.

Technical Abstract: Aims: Our goal was to compare the discriminatory power of PvuII ribotyping and MLST using a single set of diverse Bordetella bronchiseptica isolates and to determine whether subtyping based on repeat region sequences of the pertactin gene (prn) provides additional resolution. Methods and Results: One hundred twenty-two isolates were analyzed, representing 11 mammalian or avian hosts, sourced from the United States, Europe, Israel and Australia. Thirty-two ribotype patterns were identified; one isolate could not be typed. In comparison, all isolates were typeable by MLST and a total of 30 sequence types was identified. The discrimination indices for ribotyping and MLST are nearly identical, 0.920 and 0.919, respectively. Nonetheless, for ten ribotypes and eight MLST sequence types, the alternative method discriminates among isolates that otherwise type identically. Pairing prn repeat region typing with ribotyping yielded 534 genotypes and increased the discrimination index to 0.954. Repeat region typing combined with MLST resulted in 457 genotypes and a discrimination index of 0.9443. Conclusion: Given the technical and practical advantages of MLST over ribotyping, we conclude that MLST is preferable as a primary typing tool. Isolates can frequently be further differentiated on the basis of prn repeat region sequences. Significance and Impact of the Study: This is the first direct comparison of the discriminatory power of ribotyping and MLST applied to B. bronchiseptica. Our data reveals MLST combined with prn repeat region typing to be a highly discriminatory tool suitable for evaluating the population structure and investigating the epidemiology of the bacterium.