Author's personal copy Genetic diversity of Pasteurella species isolated from European...

transcript

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Research article

Genetic diversity of Pasteurella species isolated from Europeanvespertilionid bats

Kristin Muhldorfer a,*, Stefan Schwarz b, Jorns Fickel a, Gudrun Wibbelt a, Stephanie Speck a,c

a Leibniz Institute for Zoo and Wildlife Research (IZW), Research Group of Wildlife Diseases, Alfred-Kowalke-Str. 17, D-10315 Berlin, Germanyb Institute of Farm Animal Genetics, Friedrich-Loeffler-Institute (FLI), Hoeltystr. 10, D-31535 Neustadt-Mariensee, Germanyc Bundeswehr Institute of Microbiology, Neuherbergstr. 11, D-80937 Munich, Germany

1. Introduction

Members of the genus Pasteurella are importantbacterial pathogens due to their wide distribution andbroad host range. Within this genus, Pasteurella (P.)multocida has been associated with a variety of localizedand systemic infections in mammals and birds, includingdiseases of economic significance in livestock. P. multocida

is also known as a commensal of the normal oropharyngealflora in many animal species. Wound infections following

animal bites from dogs and cats are commonly described(Francis et al., 1975; Westling et al., 2006).

Investigations regarding bacterial diseases in bats aresparse, however, systemic P. multocida infections havebeen described in deceased free-ranging European bats(Simpson, 2000; Daffner, 2001). Most of the bats that diedof pasteurellosis had traumatic injuries and approximately50–65% of these could directly be attributed to catpredation (Simpson, 2000; Routh, 2003). The oral mucosaof cats usually comprises pathogenic Pasteurella strains(Ganiere et al., 1993). Hence a bat attacked by a cat mightdevelop pasteurellosis and die from septicemia, a findingthat has also been described in feral and psittacine birds(Smit et al., 1980; Panigrahy and Harmon, 1985). AlthoughPasteurella infections were reported in up to 22% of

Veterinary Microbiology 149 (2011) 163–171

A R T I C L E I N F O

Article history:

Received 29 July 2010

Received in revised form 30 September 2010

Accepted 4 October 2010

Keywords:

Pasteurella multocida

Wildlife

Chiroptera

16S rDNA

A B S T R A C T

Pasteurella are an important cause of fatal infections in free-ranging bats, but the genetic

diversity of bat-derived strains is unclear. In the current study, 81 Pasteurella strains

associated with pneumonia, severe organ necroses and systemic infection in free-ranging

European vespertilionid bats were characterized by biochemical and molecular typing

methods. Genetic relationships and subspecies status of Pasteurella multocida strains were

determined by comparative 16S rDNA and rpoB gene sequence analysis. In addition, 30

representatives of the bat-derived P. multocida strains were selected based on phenotypic

and genotypic tests to be compared by pulsed-field gel electrophoresis using SmaI. Most

(85%) of the Pasteurella strains obtained from free-ranging bats in this study represented P.

multocida ssp. septica. P. multocida ssp. multocida and Pasteurella species B were also

identified in a small number of isolates. PFGE analysis correlated well with the sequencing

results and revealed a high genetic diversity among bat-derived strains of P. multocida ssp.

septica. Strains sharing identical or closely related SmaI fragment patterns were cultured

from bats of different species, geographic origins, and years of isolation. The presence of

numerous different P. multocida strains allows the assumption that Pasteurella infections

in vespertilionid bats are not solely based on intra- but also on inter-species transmission.

And indeed, our results present evidence of P. multocida infections in bats following cat

predation.

* Corresponding author. Tel.: +49 30 5168227; fax: +49 30 5126104.

E-mail address: muehldorfer@izw-berlin.de (K. Muhldorfer).

Contents lists available at ScienceDirect

Veterinary Microbiology

journa l homepage: www.e lsev ier .com/ locate /vetmic

doi:10.1016/j.vetmic.2010.10.002

deceased bats examined bacteriologically, strains have notbeen further characterized (Simpson, 2000).

We conducted a study to investigate bacterial diseasesand causes of death in free-ranging European bats. Over a6-year period (2003–2009), 394 dead or moribund animalsof eighteen bat species, collected in different geographicregions in Germany, were examined bacteriologically andhisto-pathologically. Pasteurella strains isolated fromvespertilionid bats of eight different species were char-acterized by biochemical and molecular-biological meth-ods. The genetic diversity of P. multocida strains wasinvestigated by 16S rDNA and rpoB gene sequence analysisas well as pulsed-field gel electrophoresis.

2. Materials and methods

2.1. Bacterial strains and growth conditions

A total of 81 Pasteurella strains were cultured fromdifferent organs of 29 deceased free-ranging vespertilio-nid bats (Table 1). Most of the bats (83%) revealedtraumatic injuries, such as fractures and wing lacerations,and 93% had pathological lesions within their internalorgans. The bat carcasses originated from three geo-graphic regions in Germany (Table 1) and were provided

by bat researchers and bat protectionists. Tissue samplesof lung, liver, heart, kidney, and spleen as well as swabsamples from the thoracic cavity of the animals wereplated onto Columbia (5% sheep blood), Chocolate (5%CO2) and MacConkey agar (Oxoid, Wesel, Germany) andincubated at 37 8C for 24–48 h.

2.2. Phenotypic characterization

Primary identification of bacterial strains was based onhaemolysis, Gram-staining, indol production, catalase andoxidase reaction. For identification at the species level,conventional biochemical tests for the utilization ofsucrose, fructose, inositol, galactose, mannose, arabinose,maltose, mannitol, sorbitol, dulcitol, glucose, cellobiose(VWR International GmbH, Dresden, Germany), trehalose,salicin and melibiose (Carl Roth GmbH & CO. KG, Karlsruhe,Germany), for the production of urease (urea 98%, Sigma–Aldrich Chemie GmbH, Steinheim, Germany) and ornithinedecarboxylase (ODC) (L-ornithine hydrochloride 99%, VWRInternational GmbH) as well as for nitrate reduction(nitrate broth, ACILA Dr. Weidner GmbH, Weiterstadt,Germany) and aesculin hydrolysis (aesculin, VWR Inter-national GmbH) were performed according to Olsen et al.(2005).

Table 1

Pasteurella strains isolated from free-ranging vespertilionid bats.

Straina No. of strains (n = 81)/

sample material

Bat species Origin Bt Ct GenBank accession nos.

16S rDNA rpoB

E41/03 4 lu, li, ki, sp Pipistrellus pipistrellus B 4 D HM746965 HM746996

E61/03 1 lu Pipistrellus pipistrellus B 1 A HM746966 HM746997

E64/04 2 li, ki Pipistrellus pipistrellus B 1 A HM746967 HM746998

E5/05 5 lu, li, he, ki, sp Pipistrellus pipistrellus B 2 A, F HM746968 HM746999

E115/07 5 lu, li, he, ki, sp Pipistrellus pipistrellus B 3 A HM746969 HM747000

E142/07 2 li, he Pipistrellus pygmaeus B 3 A HM746970 HM747001

E159/07 1 tc Pipistrellus kuhlii Ba 1 A HM746971 HM747002

E167/07 4 lu, li, he, ki Plecotus auritus Ba 3 A HM746972 HM747003

E195/07 1 he Vespertilio murinus Ba 3 A HM746973 HM747004

E163/08 3 lu, li, he Pipistrellus pipistrellus LS 3 – HM746974 HM747005

E279/08 1 he Pipistrellus pipistrellus LS 3 A HM746975 HM747006

E333/08 2 he, ki Pipistrellus nathusii Ba 3 A HM746976 HM747007

E336/08 4 lu, li, he, ki Pipistrellus nathusii Ba 4 A HM746977 HM747008

E348/08 4 lu, li, he, ki Pipistrellus pipistrellus Ba 2 F HM746978 HM747009

E10/09 4 lu, li, he, ki Pipistrellus kuhlii Ba 2 A HM746979 HM747010

E16/09 2 lu, he Myotis mystacinus Ba 4 A HM746980 HM747011

E22/09 3 lu, li, ki Myotis mystacinus Ba 3 A HM746981 HM747012

E107/09 1 tc Vespertilio murinus Ba 3 – HM746982 HM747013

E135/09/1 3 lu, li, kiPlecotus auritus

Ba n.d. n.d. HM746983 HM747014

E135/09/2 3 lu, li, ki 3 – HM746984 HM747015

E140/09 3 li, he, ki Myotis mystacinus Ba 3 A HM746985 HM747016

E142/09 1 he Pipistrellus pipistrellus Ba 2 A HM746986 HM747017

E171/09/1 1 luPipistrellus pipistrellus

B 1 A HM746987 HM747018

E171/09/2 2 li, ki 3 – HM746988 HM747019

E187/09 2 lu, ki Vespertilio murinus B 3 A HM746989 HM747020

E223/09 3 lu, li, ki Pipistrellus pipistrellus B 1 A HM746992 HM747023

E236/09 2 lu, he Eptesicus serotinus B 1 A HM746994 HM747025

Bt, biochemical type; Ct, capsular type; n.d., not determined; lu, lung; li, liver; he, heart; ki, kidney; sp, spleen; tc, thoracic cavity (swab); B, Berlin; Ba,

Bavaria; LS, Lower Saxony.a Each strain was chosen as representative of all phenotypically and genetically indistinguishable Pasteurella strains obtained from different tissues of

one bat.

K. Muhldorfer et al. / Veterinary Microbiology 149 (2011) 163–171164

2.3. Pasteurella multocida-specific PCR and capsular typing

All strains were confirmed as being P. multocida by amonoplex PCR assay using P. multocida-specific primers(Townsend et al., 1988). The capsular type of the P.

multocida strains was determined by multiplex PCR aspreviously described by Townsend et al. (2001).

2.4. 16S rDNA and rpoB gene sequence analysis and

phylogenetic calculations

The phylogenetic relationship of the Pasteurella strainswas investigated by sequencing segments of their 16SrDNA and rpoB genes, and comparing them with sequencesdeposited in GenBank (http://www.ncbi.nlm.nih.gov/).Three to four bacterial colonies from overnight cultureswere transferred to 200 ml of nuclease-free water (Pro-mega, Mannheim, Germany). Suspended colonies wereheated at 98 8C for 10 min, stored on crushed ice andsubsequently subjected to PCR. Amplification of the 16SrDNA gene fragment was carried out as reported by Davieset al. (1996) with minor modifications. PCR mixtures(50 ml) contained 0.5 mM each of primer 27F (50-AGA GTTTGA TCM TGG CTC AG-30) and 1492R (50-TAC GGY TAC CTTGTT ACG ACT T-30), 10 ml 5� GoTaq reaction buffer(1.5 mM MgCl2 f.c.; Promega), 200 mM dNTPs (Promega),1.25 U of GoTaq DNA polymerase (Promega), and 2 ml oftemplate DNA. Cycling conditions on a T-Gradientthermocycler (Biometra, Gottingen, Germany) were:10 min 94 8C, 30 cycles (40 s 94 8C, 30 s 61 8C, 80 s 72 8C),10 min 72 8C. Determination of the partial rpoB genesequence was performed as described by Korczak et al.(2004) and Christensen et al. (2004) using primersPasrpob-L (50-GCA GTG AAA GAR TTC TTT GGT TC-30)and Rpob-R (50-GTT GCA TGT TIG IAC CCA T-30). PCRproducts were purified with the QIAquick PCR purificationkit (Qiagen, Hilden, Germany), and directly sequencedapplying the BigDye Terminator Cycle Sequencing Kit 1.1(ABI, Darmstadt, Germany) in a peqStar 96 UniversalGradient thermocycler (PeqLab, Erlangen, Germany) fol-lowed by fragment separation and analysis on an ABImodel A3130xl Genetic Analyzer using the softwareSEQUENCING ANALYSIS v.5.2 (ABI). Sequences werealigned with corresponding sequences of the P. multocida

reference strains retrieved from the GenBank databaseusing CLUSTAL X v.2.0 (Larkin et al., 2007).

Since often more than one P. multocida strain wasobtained from different organs of the same bat, all P.

multocida strains of one host were compared for theirresults of the aforementioned phenotypic and genotypictests. To avoid the multiple inclusion of the same strain, 30representative individual P. multocida strains wereselected for phylogenetic analyses. Of all phenotypicallyand genotypically indistinguishable P. multocida strainsobtained from the same bat, only a single strain wasincluded in this set of 30 representative strains. Sequencedistance calculations were performed using the MEGAsoftware v.4.0 (Tamura et al., 2007). Parameters requiredfor tree construction, such as a model of nucleotidesubstitution and side mutation rate heterogeneity, wereestimated using a combination of PAUP* v.4.b10 (Swofford,

2002) and MODELTEST v.3.7 (Posada and Crandall, 1998;Posada, 2003). The model selected for both data sets wasthe general time-reversible (GTR) model (Lanave et al.,1984; Tavare, 1986) with an allowance both for invariantsites (I) and a gamma (G) distribution shape parameter afor among-site rate variation (GTR + I + G) (Rodriguez et al.,1990). Phylogenetic trees for 16S rDNA and rpoB geneswere constructed applying the Maximum-Likelihood (ML)algorithm implemented in the software TREEPUZZLE v.5.2(Schmidt et al., 2002), the Bayesian Inference (BI)algorithm included in the program MrBAYES v.3.1(Ronquist and Huelsenbeck, 2003), as well as theNeighbor-Joining (NJ) using the TN93 substitution model(Tamura and Nei, 1993) and Maximum-Parsimony (MP)algorithms implemented in the software MEGA v.4.0(Tamura et al., 2007). Gamma distribution parameterswere a = 0.011 (16S rDNA) and 0.017 (rpoB). Statisticalsupport was obtained by bootstrapping (1000 replicates)for ML, NJ and MP and by Monte Carlo Markov chains(100,000 burnins, 100,000 runs) for BI. All newly obtainedsequences were deposited into GenBank (16S rDNAaccession nos.: HM746965–HM746995; rpoB accessionnos.: HM746996–HM747026).

2.5. Pulsed-field gel electrophoresis (PFGE)

Whole-cell DNA for pulsed-field gel electrophoresiswas prepared as described previously (Kehrenberg andSchwarz, 2000). Slices of the DNA containing agarose plugswere incubated for 4 h in the presence of 20 U of SmaI(Boehringer, Mannheim, Germany). The respective DNAfragments were separated by agarose gel electrophoresis(SeaKem GTG, 1%, w/v; Biozym, Hess. Oldendorf, Germany)in a CHEF DR III system (Bio-Rad, Munchen, Germany)using 0.5� TBE as running buffer. The running conditionswere 5.6 V/cm at 14 8C for 24 h. The pulse times wereramped from 2 to 5 s over 24 h. The gel was stained withethidium bromide (2 mg/ml; Sigma) and photographedunder UV illumination. The dendrogram was constructedusing GELCOMPAR II v.6.0 (Applied Maths, Sint-Martens-Latem, Belgium). The chromosomal SmaI fragments ofStaphylococcus aureus 8325 served as size standard.

3. Results

3.1. Phenotypic characterization of Pasteurella strains

Most (88%) of the bat-derived Pasteurella strains weregrown in pure cultures from one or more organs of thesame animal. Four strains were isolated from mixedcultures with Klebsiella and Streptococcus species, and twophenotypically different Pasteurella strains (E135/09/1,E135/09/2) were cultured from lung, liver and kidney of abrown long-eared bat (Plecotus auritus). All strains werecatalase-, indol- and oxidase-positive and stained as Gram-negative coccobacilli. They demonstrated non-haemolytic,aerobic growth on Columbia sheep-blood agar while nogrowth was observed on MacConkey agar. According tobiochemical results, all strains were characterized as P.

multocida with the exception of three strains from batE135/09 that met all phenotypic criteria of Pasteurella

K. Muhldorfer et al. / Veterinary Microbiology 149 (2011) 163–171 165

species B (Table 2). Among the 78 P. multocida strains, fourdifferent biochemical types (Bt) were recognized based ondifferences observed for ODC production as well astrehalose and sorbitol utilization. The most prevalentbiochemical type was Bt 3 (55%) which represents P.

multocida strains positive for trehalose metabolism andODC production, but negative for sorbitol utilization (Table2).

3.2. Pasteurella multocida-specific PCR and capsular typing

All 78 P. multocida strains yielded the expectedamplicon (460 bp) in the P. multocida-specific PCR assay.PCR-directed analysis of the capsular types resulted in asingle band representing the serogroup-specific regions ofthe respective biosynthetic loci in 64 (82%) P. multocida

strains. Each of these PCR fragments corresponded in sizeto the amplicons obtained from the reference strains ofcapsular types A (1044 bp), D (657 bp) or F (851 bp).Among them, 56 (72%) strains belonged to capsular type A,while four (5.1%) strains belonged to capsular type D (allfrom bat E41/03) and another four strains to capsular typeF (all from bat E348/08). All five strains isolated from batE5/05 showed two PCR fragments in repeated experi-ments, one corresponding to capsular type A, the other totype F. The remaining nine strains failed to produce anamplicon.

3.3. 16S rDNA and rpoB sequence comparison

Stretches of 1321 bp (16S rDNA) and 520 bp (rpoB gene)were used in a multiple sequence alignment for sequence

distance analysis and phylogenetic tree construction.Based on nucleotide variations, 12 different 16S rDNAsequence types and 11 rpoB sequence types were observedamong the newly derived P. multocida sequences (Table 3).Two distinct groups, A and B, were identified by visualcomparison of the polymorphic nucleotide positionswithin the 16S rDNA and rpoB sequences (SupplementalFigs. 1 and 2).

For phylogenetic analysis, the 16S rDNA and rpoB

sequences of the 30 representative P. multocida strains ofbat origin were compared with those of the type strains ofP. multocida ssp. multocida CCUG 17976T, P. multocida ssp.septica CCUG 17977T (cat bite origin), and P. multocida ssp.gallicida CCUG 17978T, as well as with those of different P.

multocida strains obtained from rabbits (Stahel et al., 2009)and cats (Kuhnert et al., 2000). The 16S rDNA sequencesgrouped in two distinct clusters, A and B (Fig. 1), whichcorresponded to groups A and B of the 16S rDNA sequencedistance analysis (Supplemental Fig. 1). Most of the P.

multocida strains obtained from bats were located incluster A that also contained the type strain P. multocida

ssp. septica CCUG 17977T (accession no. AF294411) as wellas P. multocida ssp. septica strains from rabbits (e.g.accession no. EF579833). Three strains (E348/08, E10/09,E142/09) were located in cluster B along with the typestrain P. multocida ssp. multocida CCUG 17976T (accessionno. AF294410) and P. multocida ssp. gallicida CCUG 17978T

(accession no. AF294412). 16S rDNA sequence comparisonof the bat-derived cluster A strains and GenBank retrievedP. multocida strains revealed a high degree of similarity(99–100%) with the P. multocida ssp. septica type strainCCUG 17977T as well as with P. multocida ssp. septica from

Table 2

Biochemical characteristics of bat-derived Pasteurella strains in comparison to the type strains of Pasteurella multocida and Pasteurella species B (Olsen et al.,

2005).

Test P. multocida isolated from bats P. multocida type strains Pasteurella species B

(n = 11)

(n = 14)

(n = 43)

(n = 10)

ssp. multocida

CCUG 17976T

ssp. septica

CCUG 17977T

ssp. gallicida

CCUG 17978T

Bat strain E135/09/1 Type strain CCUG 19794T

Indole + + + + + + + + +

Dulcitol � � � � � � + + +

Sorbitol + + � � + � + � �Mannitol + + + + + + + � �Trehalose + � + + + + � + +

Maltose � � � � � � � + +

ODC + + + � + + + + +

Table 3

16S rDNA and rpoB sequence distances among bat-derived Pasteurella multocida strains (2).

Total Group A Group B

Nucleotide variations of 16S rDNA sequences (1321 bp)

Number of sequence types 12 9 3

Number of polymorphic nucleotide positionsa 24 (1.74) 8 (0.61) 2 (0.15)

Number of pairwise sequence differencesb 1–22 (9.58� 1.91) 1–4 (2.22� 0.87) 1–2 (1.33� 0.98)

Number of pairwise group differencesb 17–22 (20.30� 4.09)

Nucleotide variations of rpoB sequences (520 bp)

Number of sequence types 11 8 3

Number of polymorphic nucleotide positionsa 43 (8.27) 7 (1.35) 21 (4.04)

Number of pairwise sequence differencesb 1–35 (15.64� 2.40) 1–7 (3.46� 1.31) 10–17 (14.0� 3.0)

Number of pairwise group differencesb 23–35 (30.04� 4.68)a The percentage per given sequence length is given in parentheses.b The mean number of nucleotide differences is given in parentheses.

different hosts (i.e. cats, rabbits). In contrast, the remainingstrains E348/08, E10/09 and E142/09 showed 99.8–100%sequence similarity to the type strain of P. multocida ssp.multocida and P. multocida ssp. gallicida, respectively.

The rpoB gene provided a higher resolution compared to16S rDNA and subdivided the strains of the P. multocida

ssp. multocida/P. multocida ssp. gallicida cluster B of the 16SrDNA-based tree into two distinct clusters, B1 and B2(Fig. 2). In accordance with the 16S rDNA analysis, most ofthe bat-derived strains represented rpoB cluster A. Theyrevealed a high sequence similarity (99–100%) to P.

multocida ssp. septica CCUG 17977T (accession no.AY362970) whereas less than 96% and 94% sequencesimilarity was observed to P. multocida ssp. multocida

CCUG 17976T (accession no. AY170216) and P. multocida

ssp. gallicida CCUG 17978T (accession no. AY362969),respectively. Strain E348/08 was located in cluster B1together with the type strain P. multocida ssp. multocida

CCUG 17976T and a clinical strain of P. multocida ssp.multocida from a rabbit (accession no. EF579854), andrevealed 98.5% and 96.5% sequence similarity to the typestrain of P. multocida ssp. multocida and P. multocida ssp.gallicida, respectively. The remaining strains, E10/09 andE142/09, were placed in cluster B2 together with the typestrain of P. multocida ssp. gallicida CCUG 17978T andshowed sequence similarities of >98% and >97% to P.

multocida ssp. gallicida CCUG 17978T and P. multocida ssp.multocida CCUG 17976T, respectively.

[()TD$FIG]

Fig. 1. 16S rDNA gene-based Maximum-Likelihood tree showing the phylogenetic relationships of 30 representative Pasteurella multocida strains obtained

from bats in relation to different P. multocida reference strains. Strain numbers are listed and the GenBank accession numbers of the reference strains are

given in square brackets. Distinct clusters are indicated with A and B. Statistical support for nodes recognized in the Neighbor-Joining (NJ), Maximum-

Parsimony (MP), Bayesian Inference (BI) and Maximum-Likelihood (ML) analysis are indicated as follows NJ/MP/BI/ML.

3.4. PFGE analysis

The 30 representative P. multocida strains underwentSmaI macrorestriction analysis to be compared by PFGE.DNA of two strains however (E159/07, E140/09), couldnot be digested by SmaI. Among the remaining 28 strains,25 distinct PFGE patterns were identified. Indistinguish-able fragment patterns were observed for the strainsE187/09 and E22/09, and the strains E163/08, E107/09and E171/09/2. Strains sharing more than 80% patternsimilarity formed four distinct clusters (A–D) (Fig. 3).

4. Discussion

Pasteurella strains retrieved from free-ranging vesper-tilionid bats predominantly represented P. multocida (96%)which is in accordance with results of previous studies infree-ranging bats (Simpson, 2000; Daffner, 2001). Three

strains of bat E135/09 were identified as Pasteurella speciesB with 99.6% 16S rDNA sequence similarity to thecorresponding type strain CCUG 19794T (accession no.M75052). Phenotypic classification to the subspecies levelof bat-derived P. multocida strains was in part not inagreement with the results obtained by DNA-basedmethods. However, variations in P. multocida phenotypeshave also been described by different authors. Kuhnertet al. (2000) found phenotypic heterogeneity in strainsderived from cats classified as P. multocida ssp. septica

which reacted negative for ODC (usually positive for all P.

multocida) and trehalose (normally positive for P. multocida

ssp. septica) and positive for sorbitol (expected negative forP. multocida ssp. septica). ODC-negative P. multocida ssp.multocida strains obtained from rabbits have been reportedby Stahel et al. (2009) and Christensen et al. (2004)characterized sorbitol-negative variants of P. multocida

ssp. multocida derived from bovids. Bt 4 strains in the

[()TD$FIG]

Fig. 2. RpoB gene-based Maximum-Likelihood tree showing the phylogenetic relationships of 30 representative Pasteurella multocida strains obtained from

bats in relation to different P. multocida reference strains. Strain numbers are listed and the GenBank accession numbers of the reference strains are given in

square brackets. Distinct clusters are indicated with A, B1 and B2. Statistical support for nodes recognized in the Neighbor-Joining (NJ), Maximum-

Parsimony (MP), Bayesian Inference (BI) and Maximum-Likelihood (ML) analysis are indicated as follows NJ/MP/BI/ML.

current study appeared to represent an ODC-negativevariant of P. multocida ssp. septica. According to results of16S rDNA and rpoB sequence analyses, 69 out of 78 P.

multocida strains from free-ranging bats were clearlyassigned to P. multocida ssp. septica. More than half of thesestrains corresponded in their biochemical characteristics(Bt 3) to those expected for P. multocida ssp. septica,whereas strains of Bt 1 (16%) phenotypically represented P.

multocida ssp. multocida (Olsen et al., 2005). All strainslocated in cluster B of the phylogenetic trees (Figs. 1 and 2)were assigned to Bt 2. Among them, four strains obtainedfrom the same bat (E348/08) were classified as P. multocida

ssp. multocida based on 99.2% rpoB sequence similarity tothe rabbit-derived strain Clin 35 (accession no. EF579854).The remaining P. multocida strains (bats E10/09, E142/09)

could not be clearly assigned to a subspecies, but appearedto be closely related to the type strain of P. multocida ssp.gallicida. None of these strains were positive for dulcitol asexpected for this subspecies (Olsen et al., 2005).

Three different capsular types (A, D, F) were observedamong P. multocida strains of bat origin. Capsular type Awas the most prevalent type found here. It was detected in75% of P. multocida ssp. septica and in all strains ofunclassified subspecies. These results are in accordancewith previous observations, as P. multocida strains ofcapsular type A were found to be predominant in mostanimal species (Rhoades and Rimler, 1989; Ewers et al.,2006). It has long been recognized that host predilectionand disease symptoms are linked to certain P. multocida

capsular types (Rhoades and Rimler, 1989). In our study,

[()TD$FIG]

Fig. 3. SmaI-generated macrorestriction patterns visualized by pulsed-field gel electrophoresis showing the genetic relationships of 30 representative

Pasteurella multocida strains obtained from free-ranging vespertilionid bats. PFGE profiles of two strains (E159/07, E140/09) are not given as DNA could not

be digested by SmaI. Node information represents percentage of pattern similarity. Distinct clusters of strains sharing band similarities of more than 80% are

indicated from A to D at the respective branches. Biochemical types (Bt) and capsular types (Ct) are listed behind the strain numbers as follows [Bt/Ct].

expression of specific capsule biosynthesis genes appearedunrelated to pathological changes found in P. multocida-infected bats. Moreover, strains of capsular types A and Fwere detected in one single host (E5/05), which has notbeen reported so far. It should also be noted that capsularPCR results of bat-derived P. multocida strains corre-sponded to those described for strains derived fromdomestic cats (Ewers et al., 2006).

Genetic heterogeneity as well as remarkable geneticstability of P. multocida has been described among strainsfrom different livestock and avian sources (Hirsh et al.,1990; Pedersen et al., 2003; Marois et al., 2009). In ourstudy, SmaI macrorestriction analysis demonstrated highlydiverse PFGE patterns among P. multocida strains fromfree-ranging bats which is in contrast to results of Pedersenet al. (2003). These authors found P. multocida isolatedfrom wild birds in Denmark appears to be clonal,irrespective of bird species, geographic origin, and yearof isolation. However, evidence of species-specific P.

multocida clones in migratory waterfowl was describedearlier (Hirsh et al., 1990). In the present study, all P.

multocida ssp. septica strains retrieved from vespertilionidbats grouped together which corresponded well to resultsof 16S rDNA and rpoB gene analysis. Among them, a fewstrains (clusters B and D) exhibited indistinguishable orclosely related (similarity >90%) PFGE profiles suggestingthe occurrence of one P. multocida ssp. septica clone amongthe respective group of bats, despite the fact that thesestrains were isolated from bats of different species,geographic origins, and years of isolation. While strainsE167/07, E22/09 and E187/09 (cluster B) revealed identicalbiochemical and PCR typing results, strains E163/08, E107/09 and E171/09/2 (cluster D) showed differences of one ortwo nucleotides among their 16S rDNA sequences.Interestingly, strains E171/09/1 and E171/09/2 obtainedfrom different tissues of one common pipistrelle (Pipis-

trellus pipistrellus) differed markedly in their PFGE patternsas well as in their phenotypic and genotypic character-istics. This observation provides evidence of co-infectionwith two different strains of P. multocida ssp. septica. Theremaining P. multocida strains (E10/09, E142/09, E348/08)showed distinct PFGE profiles, while strains E10/09 andE142/09 demonstrated higher pattern similarity in com-parison to the banding profile of strain E348/08 identifiedas P. multocida ssp. multocida.

Bat species roosting in human habitation frequently fallprey to domestic cats (Routh, 2003). We found that 83%(n = 24) of Pasteurella-infected bats had traumatic injuriessuggestive of cat predation. In seven cases cat predationwas confirmed. As there is a high incidence of Pasteurella ingingival scrapings of cats (Ganiere et al., 1993) it seemslikely that at least some of the bats in our study have beeninfected by P. multocida strains of cat origin. Most bat-derived strains in this study represented P. multocida ssp.septica and capsular type A. Such strains have beendescribed as the predominant P. multocida in domesticcats (Kuhnert et al., 2000; Ewers et al., 2006). The presenceof a high number of different P. multocida strains might alsoindicate that Pasteurella infections in free-ranging Eur-opean vespertilionid bats are not based on direct bat-to-bat transmission, but further investigations are required to

clarify the genetic relationships of P. multocida isolatedfrom bats.

Acknowledgements

The authors are grateful to M. Kistler, S. Morgenroth, M.Kredler, E. Muhlbach, K. Muller, S. Rosenau, and W. and H.Zoels for providing the bat carcasses, and to R. Becker, N.Jahn, D. Krumnow and D. Lieckfeldt for their excellenttechnical assistance. We also thank J. Kelemen and G. B.Michael for their assistance. The study was supported bythe Adolf and Hildegard Isler Stiftung, the FAZIT Stiftungand the Klara Samariter Stiftung.

Appendix A. Supplementary data

Supplementary data associated with this article can

be found, in the online version, at doi:10.1016/j.vetmic.

2010.10.002.

References

Christensen, H., Kuhnert, P., Olsen, J.E., Bisgaard, M., 2004. Comparativephylogenies of the housekeeping genes atpD, infB, and rpoB and the16S rRNA gene within the Pasteurellaceae. Int. J. Syst. Evol. Microbiol.54, 1601–1609.

Daffner, B.L., 2001. Causes of morbidity and mortality in British batspecies and prevalence of selected zoonotic pathogens. Thesis forMSc in Wild Animal Health. University of London.

Davies, R.L., Paster, B.J., Dewhirst, F.E., 1996. Phylogenetic relationshipsand diversity within the Pasteurella haemolytica complex based on16S rRNA sequence comparison and outer membrane protein andlipopolysaccharide analysis. Int. J. Syst. Bacteriol. 46, 736–744.

Ewers, C., Lubke-Becker, A., Bethe, A., Kießling, S., Filter, M., Wieler, L.H.,2006. Virulence genotype of Pasteurella multocida strains isolatedfrom different hosts with various disease status. Vet. Microbiol.114, 304–317.

Francis, D.P., Holmes, M.A., Brandon, G., 1975. Pasteurella multocida.Infections after domestic animal bites and scratches. JAMA 233,42–45.

Ganiere, J.P., Escande, F., Andre, G., Larrat, M., 1993. Characterization ofPasteurella from gingival scrapings of dogs and cats. Comp. Immunol.Microbiol. Infect. Dis. 16, 77–85.

Hirsh, D.C., Jessup, D.A., Snipes, K.P., Carpenter, T.E., Hird, D.W., McCapes,R.H., 1990. Characteristics of Pasteurella multocida isolated fromwaterfowl and associated avian species in California. J. Wildl. Dis.26, 204–209.

Kehrenberg, C., Schwarz, S., 2000. Identification of a truncated, butfunctionally active tet(H) tetracycline resistance gene in Pasteurellaaerogenes and Pasteurella multocida. FEMS Microbiol. Lett. 188, 191–195.

Korczak, B., Christensen, H., Emler, S., Frey, J., Kuhnert, P., 2004. Phylogenyof the family Pasteurellaceae based on rpoB sequences. Int. J. Syst. Evol.Microbiol. 54, 1393–1399.

Kuhnert, P., Boerlin, P., Emler, S., Krawinkler, M., Frey, J., 2000. Phyloge-netic analysis of Pasteurella multocida subspecies and molecularidentification of feline P. multocida subsp. septica by 16S rRNA genesequencing. Int. J. Med. Microbiol. 290, 599–604.

Lanave, C., Preparata, G., Sacone, C., Serio, G., 1984. A new method forcalculating evolutionary substitution rates. J. Mol. Evol. 20, 86–93.

Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A.,McWilliam, H., Valentin, F., Wallace, I.M., Wilm, A., Lopez, R., Thomp-son, J.D., Gibson, T.J., Higgins, D.G., 2007. Clustal W and Clustal Xversion 2.0. Bioinformatics 23, 2947–2948.

Marois, C., Fablet, C., Gaillot, O., Morvan, H., Madec, F., Kobisch, M., 2009.Molecular diversity of porcine and human isolates of Pasteurellamultocida. J. Appl. Microbiol. 107, 1830–1836.

Olsen, I., Dewhirst, F.E., Paster, B.J., Busse, H.-J., 2005. Family I. Pasteur-ellaceae Pohl 1981b, 382VP (Effective publication: Pohl 1979, 81). In:Brenner, D.J., Krieg, N.R., Staley, J.T. (Eds.), Bergey’s Manual of Sys-tematic Bacteriology. 2nd edition. Springer, New York, pp. 851–912.

Panigrahy, B., Harmon, B.G., 1985. Bacterial septicemias in two psittacinebirds. J. Am. Vet. Med. Assoc. 186, 983–984.

Pedersen, K., Dietz, H.H., Jørgensen, J.C., Christensen, T.K., Bregnballe, T.,Andersen, T.H., 2003. Pasteurella multocida outbreaks of avian cholerain wild and captive birds in Denmark. J. Wildl. Dis. 39, 808–816.

Posada, D., Crandall, K.A., 1998. MODELTEST: testing the model of DNAsubstitution. Bioinform. Appl. Note 14, 817–818.

Posada, D., 2003. Chapter 6. 5: using Modeltest and Paup* to select amodel of nucleotide substitution. In: Baxevanis, A.D., Davison, D.B.,Page, R.D.M., Petsko, G.A., Stein, L.D., Stormo, G.D. (Eds.), CurrentProtocols in Bioinformatics. John Wiley & Sons Inc., (approx. 1900 pp),pp. 6.5.1–6.5.14.

Rhoades, K.R., Rimler, R.B., 1989. Fowl cholera. In: Adlam, C., Rutter, J.M.(Eds.), Pasteurella and Pasteurellosis. Academic Press, London, pp.95–113.

Rodriguez, F., Oliver, J.F., Marin, A., Medina, J.R., 1990. The generalstochastic model of nucleotide substitution. J. Theor. Biol. 142,485–501.

Ronquist, F., Huelsenbeck, J.P., 2003. MrBayes 3: Bayesian phylogeneticinference under mixed models. Bioinformatics 19, 1572–1574.

Routh, A., 2003. Chapter 9: bats. In: Mullineaux, E., Best, D., Cooper, J.E.(Eds.), Manual of Wildlife Casualities. BASVA Publishers, Wareham,Dorset, UK, pp. 95–108.

Schmidt, H.A., Strimmer, K., Vingron, M., von Haeseler, A., 2002. Tree-Puzzle: maximum likelihood phylogenetic analysis using quartetsand parallel computing. Bioinformatics 18, 502–504.

Simpson, V.R., 2000. Veterinary advances in the investigation of wildlifediseases in Britain. Res. Vet. Sci. 69, 11–16.

Smit, T., Moraal, L.G., Bakhuizen, T., 1980. Pasteurella multocida infectionsin birds following cat bites. Tijdschr. Diergeneeskd. 105, 327–329Dutch.

Stahel, A.B.J., Hoop, R.K., Kuhnert, P., Korczak, B.M., 2009. Phenotypic andgenetic characterization of Pasteurella multocida and related isolatesfrom rabbits in Switzerland. J. Vet. Diagn. Invest. 21, 793–802.

Swofford, D.L., 2002. PAUP*. Phylogenetic Analysis Using Parsimony (*andother methods). Version 4 (b10) Sinauer Associates, Sunderland,Massachusetts.

Tamura, K., Dudley, J., Nei, M., Kumar, S., 2007. MEGA4: MolecularEvolutionary Genetics Analysis (MEGA) software version 4.0. Mol.Biol. Evol. 24, 1596–1599.

Tamura, K., Nei, M., 1993. Estimation of the number of nucleotide sub-stitutions in the control region of mitochondrial DNA in humans andchimpanzees. Mol. Biol. Evol. 10, 512–526.

Tavare, S., 1986. Some probabilistic and statistical problems in theanalysis of DNA sequences. Lect. Math. Life Sci. 17, 57–86.

Townsend, K.M., Frost, A.J., Lee, C.W., Papadimitriou, J.M., Dawkins, H.J.,1988. Development of PCR assays for species- and type-specificidentification of Pasteurella multocida isolates. J. Clin. Microbiol. 36,1096–1100.

Townsend, K.M., Boyce, J.D., Chung, J.Y., Frost, A.J., Adler, B., 2001. Geneticorganization of Pasteurella multocida cap loci and development of amultiplex capsular PCR typing system. J. Clin. Microbiol. 39, 924–929.

Westling, K., Farra, A., Cars, B., Ekblom, A.G., Sandstedt, K., Settergren, B.,Wretlind, B., Jorup, C., 2006. Cat bite wound infections: a prospectiveclinical and microbiological study at three emergency wards in Stock-holm. Sweden J. Infect. 53, 403–407.

Author's personal copy Genetic diversity of Pasteurella species isolated from European...

Documents