BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, researchlibraries, and research funders in the common goal of maximizing access to critical research.

Development of a PCR for Identification of Bordetella hinziiAuthor(s): Karen B. RegisterSource: Avian Diseases, 57(2):307-310. 2013.Published By: American Association of Avian PathologistsDOI: http://dx.doi.org/10.1637/10433-102212-ResNote.1URL: http://www.bioone.org/doi/full/10.1637/10433-102212-ResNote.1

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, andenvironmental sciences. BioOne provides a sustainable online platform for over 170 journals and books publishedby nonprofit societies, associations, museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance ofBioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercialinquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

Research Note—

Development of a PCR for Identification of Bordetella hinzii

Karen B. RegisterA

Ruminant Diseases and Immunology Research Unit, Agricultural Research Service, United States Department of Agriculture,National Animal Disease Center, 1920 Dayton Avenue, Ames, IA 50010

Received 31 October 2012; Accepted 7 February 2013; Published ahead of print 11 February 2013

SUMMARY. Bordetella hinzii infects primarily poultry and immunocompromised humans. It is closely related to the etiologicagent of turkey coryza, Bordetella avium. Distinguishing between B. avium and B. hinzii is difficult, and there is no method foridentification of B. hinzii suitable for use by diagnostic laboratories. This report details the development of a B. hinzii–specific PCRtargeting the ompA gene. Assay sensitivity is 100% based on analysis of 48 B. hinzii isolates from diverse geographic locationsrepresenting all known ribotypes. Evaluation of 71 isolates of B. avium and 20 other bacterial isolates from poultry, comprisinggram-negative and gram-positive commensals and pathogens of nine genera, demonstrated an assay specificity of 100%. The ompAPCR is a rapid, reliable, and accurate method for identification of B. hinzii and provides a valuable new tool for veterinarydiagnostic laboratories investigating poultry respiratory disease outbreaks.

RESUMEN. Nota de Investigacion—Desarrollo de un metodo de PCR para la identificacion de Bordetella hinzii.La Bordetella hinzii infecta principalmente a las aves de corral y a seres humanos inmunodeprimidos. Esta bacteria esta

estrechamente relacionada con el agente etiologico de la coriza de los pavos, Bordetella avium. La diferenciacion entre B. avium y B.hinzii es difıcil, y no hay un metodo para la identificacion de B. hinzii que sea adecuado para su uso en los laboratorios dediagnostico. Este informe detalla el desarrollo de un metodo de PCR especıfico para B. hinzii dirigida al gene ompA. La sensibilidaddel ensayo fue del 100%, con base en el analisis de 48 aislamientos de B. hinzii aislados de diferentes sitios geograficos querepresentan a todos ribotipos conocidos. La evaluacion de los 71 aislamientos de B. avium y de otros 20 aislamientos bacterianos deaves comerciales, que incluye bacterias comensales y patogenas, gram-negativas y positivas de nueve generos, demostro que elmetodo tenıa una especificidad del 100%. El metodo de PCR basado en el gene ompA es un metodo rapido, confiable y preciso parala identificacion de B. hinzii y proporciona un nuevo y valioso instrumento para los laboratorios de diagnostico veterinario queinvestigan brotes de enfermedades respiratorias en la avicultura.

Key words: Bordetella hinzii, ompA, PCR

Abbreviations: ATCC 5 American Type Culture Collection; NADC 5 National Animal Disease Center

Bordetella hinzii infects primarily poultry and immunocompro-mised humans but it also has been isolated from rabbits (11) andmice (5). Although initially thought to be nonpathogenic in poultry,it was recently shown that some strains cause disease in turkey poultsthat is indistinguishable from the clinical presentation of turkeycoryza caused by Bordetella avium (9). Bordetella hinzii and B. aviumare closely related and share many genetic and phenotypic traits.Only a few phenotypic tests can delineate these species, and resultsmay vary depending on inoculum size, culture conditions, and thespecific procedure used (1,2,6,7,16). Ribotyping and restrictionenzyme analysis reliably distinguish between B. avium and B. hinzii(11,14), but these tests are not readily carried out in most diagnosticlaboratories. There is currently no rapid method for identification ofB. hinzii suitable for use in a diagnostic setting. To address this need,a highly sensitive and specific PCR targeting the ompA gene of B.hinzii was developed.

MATERIALS AND METHODS

Bacterial isolates and growth conditions. The host and geographicorigin of bacterial isolates included in this study are indicated inTables 1 and 2. Forty-nine isolates of B. hinzii were evaluated, includingthe American Type Culture Collection (ATCC) type strain, four

reference strains, and 37 strains whose characterization has been reportedpreviously (11,13,14). The 71 B. avium isolates tested include the typestrain, four reference strains, and 49 strains whose characterization hasbeen reported previously (11,13,14). In addition to the ATCC typestrain and one reference strain, 17 avian isolates of Bordetellabronchseptica were analyzed, including 11 described in prior publications(8,10,11,14,15). Bordetella isolates not included in previous reports arewell-characterized strains from a collection maintained at the NationalAnimal Disease Center (NADC). All PvuII ribotypes known to occurin avian isolates (11) are represented within the group tested for eachBordetella species. Bordetella hinzii and B. avium were cultivated for18–36 hr at 37 C on 5% sheep’s blood agar. Although sheep’s bloodagar also supports the growth of B. bronchiseptica, cultures for thisstudy were grown for 24–36 hr at 37 C on Bordet-Gengou agarsupplemented with 10% sheep’s blood to be consistent with laboratoryresearch protocol. Other bacteria represent avian pathogens or normalflora of the turkey respiratory tract, the latter obtained by swabbing theupper third of the trachea of clinically healthy birds with BBLCultureSwabs (Becton Dickinson and Company, Sparks, MD). Swabswere streaked on the day of collection onto duplicate 5% sheep’s bloodagar plates that were incubated at 37 C for 24–48 hr with or without5% CO2. A representative of every distinguishable colony type wasselected for identification. Isolates from tracheal swabs were evaluatedby Gram’s stain and identified using the MicroLog System (BioLog,Inc., Hayward, CA).

PCR. Boiled lysates prepared from a single colony as describedpreviously (12) were used as template. Ten microliters of each PCR wasanalyzed by agarose gel electrophoresis in 3:1 NuSieve (Lonza Rockland,ACorresponding author. E-mail: karen.register@ars.usda.gov

AVIAN DISEASES 57:307–310, 2013

307

Inc., Rockland, ME) containing a 1:10,000 dilution of GelRed (PhenixResearch Products, Candler, NC). Chromosomal DNA purified using acommercially available kit (Promega, Madison, WI) was used to assessthe limit of detection. Purified DNA was quantified with PicoGreen(Invitrogen, Carlsbad, CA).

DNA sequencing. PCR products were purified with spin columns(QIAGEN, Valencia, CA) and sequenced directly at the NADC’sGenomics Unit using BigDyeH Terminator version 3.1 on a 3130 XLGenetic Analyzer sequencer (Applied Biosystems, Foster City, CA).Sequence data were analyzed using Vector NTI Suite software(Invitrogen). Final consensus sequences were derived from a minimumof three sequence reads with at least one from each strand.

RESULTS AND DISCUSSION

The ompA gene is widely distributed among prokaryotes and ishighly polymorphic (3). It has been proposed previously as apotential diagnostic target for B. hinzii (4). An alignment of thethree available B. hinzii ompA sequences in GenBank (accessions

AM748263–AM748265), representing ,540 bp of the 39 portionof the open reading frame, was used to design PCR primers toamplify and sequence ,520 bp of the gene. An alignment of thesequences from PCR products obtained with a subset of the B. hinziiisolates included in Table 1, representing different hosts and PvuIIribotypes, revealed three sequence variants. A representative of eachompA sequence variant was aligned with ompA gene sequences inGenBank from several additional Bordetella species to identify regions

Table 1. Bordetella spp. isolates included in this study.

Organism Host Geographic origin n

B. hinzii Human United States 3Human Switzerland 2Human Spain 1Turkey United States 1Turkey Minnesota 7Turkey Ohio 5Turkey Iowa 4Turkey California 3Turkey New York 1Chicken Australia 1Chicken Belgium 1Turkey or chicken Ohio 12Rabbit Hungary 1Mouse Missouri 1Unknown Minnesota 5

B. avium Turkey North Carolina 1Turkey Iowa 12Turkey California 6Turkey Minnesota 5Turkey Ohio 1Wild turkey New Jersey 1Turkey South Africa 3Turkey Germany 1Mallard New Jersey 7Canada goose New Jersey 1Saw-whet owl Virginia 1Unknown Ohio 15Unknown Minnesota 2Unknown Iowa 7Unknown North Carolina 3Unknown Germany 5

B. bronchiseptica Turkey Ohio 4Turkey Iowa 4Turkey California 3Turkey Minnesota 2Turkey Wisconsin 1Turkey United States 1Turkey Germany 1Turkey Unknown 1Dog United States 1Rabbit United States 1

Table 2. Additional avian isolates included in this study.

Organism Host Geographic origin n

Alcaligenes faecalis Turkey North Carolina 2Escherichia coli Turkey Iowa 1

Chicken unknown 1Enterococcus gallinarum Turkey Iowa 1Enterococcus columbae Turkey Iowa 1Enterococcus avium/faecalis Turkey Iowa 1Ornithobacterium rhinotracheale Turkey Minnesota 2Pasteurella multocida Turkey West Virginia 1

Turkey California 1Pseudomonas aeruginosa Turkey Iowa 1Staphylococcus aureus Turkey Iowa 1Staphylococcus hyicus Turkey Iowa 1Staphylococcus delphini Turkey Iowa 1Staphylococcus xylosis Turkey Iowa 1Staphylococcus intermedius Turkey Iowa 1

Fig. 1. Representative results of ompA-5/ompA-6 PCR with B.hinzii (A), B. avium (B), or other bacterial isolates from turkeys (C). InA, lanes 1–11, B. hinzii; lane 12, negative control. In B, lanes 1–14, B.avium; lane 15, B. hinzii control. In C, lanes 1–15, non–Bordetellabacterial isolates; lane 16, B. hinzii control.

308 K. B. Register

suitable for B. hinzii–specific primers. On the basis of preliminarytesting of several primer pairs with a subset of the isolates listed inTable 1 (including eight B. hinzii isolates, seven B. avium isolates, andthree B. bronchiseptica isolates), primers ompA-5 (59-CGTTC-CGGTTGCCCAGAAG-39) and ompA-6 (59-GCCTTCGGCG-GTCTTGTTGGTC-39) were selected for additional analysis.

After comparison of various reaction and cycling conditions, thoseselected as optimal for use with the ompA-5/ompA-6 primer set are1 U of AmpliTaq polymerase (Applied Biosystems), 2.5 ml of 103

buffer II, 2.5 ml of dimethyl sulfoxide, 1.0 mM MgCl2, 0.5 mMprimers, and 200 mM dNTPs in a final volume of 25 ml subjected to15 min at 95 C, 35 cycles of 30 sec at 95 C, 30 sec at 56 C, and60 sec at 72 C, and a final extension step of 7 min at 72 C.

Assay sensitivity was determined by testing 48 B. hinzii isolatesrepresenting all known hosts and a variety of geographic locations(Table 1). A single band of the predicted size (340 bp) was obtainedwith all isolates (representative results are shown in Fig. 1A).

Assay specificity was assessed with 71 B. avium isolates, 19 B.bronchiseptica isolates (including 17 of avian origin), and 20 bacterialisolates of avian origin comprised of both gram-positive and gram-negative commensals and pathogens representing seven genera(Tables 1 and 2). A PCR product was detected from only one isolate(representative results are shown in Fig. 1B, C). Amplicons wereobtained from all isolates when a universal 16S rRNA-specific primerpair (12) was used in a positive control reaction (data not shown).

The B. avium strain yielding an apparent false positive result,Stille F1 (Fig. 1B, lane 1), was further evaluated. The DNAsequence of the Stille F1 ompA-5/ompA-6 amplicon was determinedand found to be 100% identical to one of the three known B. hinziiamplicon variants and only 91% identical to available B. aviumompA sequences. Negative PCR results were obtained with thisisolate using a B. avium-specific PCR (12); a control reaction usingthe same template with universal 16S rRNA primers (12) yielded anamplicon of the expected size (data not shown). Finally, results ofPvuII ribotyping provide further support that isolate Stille F1 wasinitially mistakenly identified as B. avium and is, in fact, a B. hinzii

ribotype 2 isolate (Fig. 2). With the proper identification of thisisolate, assay specificity for the ompA-5/ompA-6 primer pair is100% under the conditions of this study.

The limit of detection for the optimized ompA-specific PCRwas determined using purified chromosomal DNA. As shown inFig. 3, the limit of detection is ,5–10 pg. The genome size of B.hinzii is currently unknown, but assuming a size similar to that ofB. avium, the PCR would be expected to detect ,1250–2500 ge-nome copies.

The ompA PCR described has 100% sensitivity and 100%specificity for B. hinzii, with a limit of detection of ,10 pg. Thisassay can be expected to generate clinically significant results andprovides an important new diagnostic tool for specific identificationof B. hinzii in poultry and perhaps in additional host species,including humans.

Fig. 2. PvuII ribotype of isolate Stille F1 (A, lane 2) compared with the four known B. hinzii ribotypes (B) or the eight known B. aviumribotypes (C). For comparison, B. avium ribotype 1 is shown in lanes 1 (A, C).

Fig. 3. Limit of detection for the ompA-5/ompA-6 PCR. Theamount of purified DNA used as template in each reaction is indicatedabove each corresponding lane. NT 5 no template.

Bordetella hinzii PCR 309

REFERENCES

1. Berkhoff, H. A., and G. D. Riddle. Differentiation of Alcaligenes-likebacteria of avian origin and comparison with Alcaligenes spp. referencestrains. J. Clin. Microbiol. 19:477–481. 1984.

2. Blackall, P. J., and J. G. Farrah. An evaluation of two methods ofsubstrate alkalinization for the identification of Bordetella avium and othersimilar organisms. Vet. Microbiol. 11:301–306. 1986.

3. Confer, A. W., and S. Ayalew. The OmpA family of proteins: roles inbacterial pathogenesis and immunity. Vet. Microbiol. http://dx.doi.org/10.1016/j.bbr.2011.03.031. 2012.

4. Fry, N. K., J. Duncan, M. T. Edwards, R. E. Tilley, D. Chitnavis, R.Harman, H. Hammerton, and L. Dainton. A UK clinical isolate ofBordetella hinzii from a patient with myelodysplastic syndrome. J. Med.Microbiol. 56:1700–1703. 2007.

5. Hayashimoto, N., H. Morita, M. Yasuda, T. Ishida, S. Kameda, A.Takakura, and T. Itoh. Prevalence of Bordetella hinzii in mice inexperimental facilities in Japan. Res. Vet. Sci. 93:624–626. 2012.

6. Hinz, K.-H., and G. Glunder. Identification of Bordetella avium sp.nov. by the API 20 NE system. Avian Pathol. 15:611–614. 1986.

7. Kattar, M. M., J. F. Chavez, A. P. Limaye, S. L. Rassoulian-Barrett,S. L. Yarfitz, L. C. Carlson, Y. Houze, S. Swanzy, B. L. Wood, and B. T.Cookson. Application of 16S rRNA gene sequencing to identify Bordetellahinzii as the causative agent of fatal septicemia. J. Clin. Microbiol.38:789–794. 2000.

8. Raffel, T. R., K. B. Register, S. A. Marks, and L. Temple. prevalenceof Bordetella avium infection in selected wild and domesticated birds in theeastern USA. J. Wildl. Dis. 38:40–46. 2002.

9. Register, K. B., and R. A. Kunkle. Strain-specific virulence ofBordetella hinzii in poultry. Avian Dis. 53:50–54. 2009.

10. Register, K. B., and T. L. Nicholson. Misidentification of Bordetellabronchiseptica as Bordetella pertussis using a newly described real-time PCRtargeting the pertactin gene. J. Med. Microbiol. 56:1608–1610. 2007.

11. Register, K. B., R. E. Sacco, and G. E. Nordholm. Comparison ofribotyping and restriction enzyme analysis for inter- and intraspeciesdiscrimination of Bordetella avium and Bordetella hinzii. J. Clin. Microbiol.41:1512–1519. 2003.

12. Register, K. B., and A. G. Yersin. Analytical verification of a PCRassay for identification of Bordetella avium. J. Clin. Microbiol.43:5567–5573. 2005.

13. Rimler, R. B., and D. G. Simmons. Differentiation among bacteriaisolated from turkeys with coryza (rhinotracheitis). Avian Dis. 27:491–500. 1983.

14. Sacco, R. E., K. B. Register, and G. E. Nordholm. Restriction enzymeanalysis and ribotyping distinguish Bordetella avium and Bordetella hinziiisolates. Epidemiol. Infect. 124:83–90. 2000.

15. van der Zee, A., F. Mooi, J. Van Embden, and J. Musser. Molecularevolution and host adaptation of Bordetella spp. : phylogenetic analysis usingmultilocus enzyme electrophoresis and typing with three insertion sequences.J. Bacteriol. 179:6609–6617. 1997.

16. Vandamme, P., J. Hommez, M. Vancanneyt, M. Monsieurs, B.Hoste, B. Cookson, C. H. Wirsing von Konig, K. Kersters, and P. J.Blackall. Bordetella hinzii sp. nov., isolated from poultry and humans. Int. J.Syst. Bacteriol. 45:37–45. 1995.

ACKNOWLEDGMENTS

I acknowledge the excellent technical assistance of William Boat-wright and thank Lea Ann Hobbs and David Alt at the NADCGenomics Unit for DNA sequence data.

310 K. B. Register