Pasteurella caballi, a new species from equine clinical specimens

(1)

0095-1137/89/102169-06$02.00/0

Pasteurella

caballi,

a

New

Species

from Equine

Clinical Specimens

L. K. SCHLATER,l* DON J. BRENNER,2 A. G. STEIGERWALT,2 C. WAYNE

MOSS,3

MARY ANN LAMBERT,3AND R. A. PACKER4

National Veterinary ServicesLaboratories, Science and Technology, Animal and Plant Health Inspection Service, U.S. DepartmentofAgriculture, Ames, Iowa500101;Molecular Biology2 and Analytical Chemistry3 Laboratories, Centersfor

DiseaseControl, Atlanta, Georgia 30333; and Department of VeterinaryMicrobiology, College of Veterinary Medicine, Iowa State University, Ames, Iowa500114

Received13 March 1989/Accepted19June 1989

ThenamePasteurella caballi is proposedforagroupoforganismsrepresented by 29strainsisolated from

respiratory and other infections in horses. P. caballistrains are gram-negative,oxidase-positive, nonmotile,

fermentative rods with the key characteristics of the genus Pasteurella. These strains differed from other

Pasteurella species in that all were aerogenic and catalase negative, and some strains produced acid from

myo-inositol and L-rhamnose. The levels of DNA relatednessof 28 P. caballi strains withlabeledDNAfromthe

proposedtypestrainaveraged 91 and85% (hydroxyapatite methodat55and70°C).P.caballiwas13to53%

relatedtostrainsrepresenting 22 other species ofthefamily Pasteurellaceae. Theguanine-plus-cytosinecontent ofthe DNA of fourstrains was41 to 42mol%. The type strain is 83851 (=ATCC 49197).

The genusPasteurella is currently classified in the family

Pasteurellaceae, togetherwith the related genera Actinoba-cillus andHaemophilus. Members ofthe genusPasteurella

are gram-negative, oxidase-positive, nonmotile,

fermenta-tive, rod-shaped bacteria(4, 17). All Pasteurella species are

parasitic on the mucous membranes of the respiratory and

digestivetracts of mammals andbirds. Underconditionsof stress, these organisms can become invasive and play a significant roleinthepathogenesis ofavariety ofinfections

in animals, including pneumonia, sinusitis, abortion, masti-tis, andsepticemia (2).

Since 1983, the General Bacteriology Section ofthe

Na-tional Veterinary Services Laboratories has received

cul-tures of an aerogenic, gram-negative bacterium that had many characteristics ofthe genus Pasteurella but could not

beassignedtoanyestablishedspecieswithin the genus. Our

attention wasoriginallydrawn to thesestrains becausethey

were all isolated from horses, an uncommon source for

Pasteurella. Inapreviousreport,thebiochemical

character-istics of13 strains were described (21), but since no DNA

hybridization information wasavailable, itwas notpossible to properly classify them. Subsequent biochemical and

ge-netic studies on these 13 strains, as well as 16 additional strains, showed that the aerogenic equine strains were

genetically similar to each other but distinct from other

phenotypically similar members of the family

Pasteurel-laceae, thereby warranting their designation as a new

spe-cies,Pasteurella caballi.

(Portions of this study were done by L.K.S. in partial

fulfillment of the requirements for the Master of Science

degree from the Department of Veterinary Microbiology,

College

of Veterinary Medicine, Iowa State University,

Ames).

MATERIALS ANDMETHODS

Bacterialstrains.The 29 P. caballi strains studied and their

sources are given in Table 1. Strains

representative

of species within the family Pasteurellaceae used in the DNA

*Corresponding author.

hybridizationstudieswerefrom thesourceslisted in Table 2. Stock cultures of P. caballi and reference strains were stored

at -70°C in 0.4-ml samples of defibrinated rabbit blood. Isolates were subcultured twice on blood agar(37°C, 18 h)

before biochemicaltesting.

Morphology of colonies and cells. Colonial morphologywas determined by observing 24-h growth on blood agar (heart

infusion agar base; Difco Laboratories, Detroit, Mich.) containing 5% defibrinated bovine blood. Growth from the blood agarplate was removed with an inoculatingloop and examined against white filter paper for the presence of

pigment. Hemolytic activity was determined on bovine,

sheep, and rabbit blood agars. Gram staining and acid-fast stainingwere done by the methods ofKoppeloff and Ziehl-Neelson, respectively, on smears prepared from 24-hgrowth onbovine blood agar(6). Motility was determined by

exam-ination of wet mount preparations, using phase-contrast

microscopy. Wet mounts were prepared by using growth

fromchocolate agar (Columbia base; Difco) slants that had been incubated for 24 hat 25 and 37°C.

Biochemical tests. Dehydrated media from commercial

sources wereusedwhenever possible. Inoculated media for all tests were incubated aerobically at 37°C. Catalase, in-dophenol oxidase, phosphatase, and glucosidase enzyme testresultswererecordedat24 h. Tests for nitrate andnitrite reduction and indole productionwereperformedafter48 h of incubation. The Voges-Proskauer and methyl redtestswere done after 4days of incubation. All other biochemical tests wereobserved daily for 1 week, and resultswereconsidered

negative ifno reaction was observed by that time. Unless otherwiseindicated, methods described by Lanyiwere used for the conventional biochemical tests (12). The test for indoleproduction was done by usingEhrlichreagent follow-ing extraction with xylene. Indophenol oxidase was detected

on filter paper, using the tetramethyl-p-phenylenediamine

dihydrochloride reagent (Marion Scientific, Div. Marion

Laboratories, Inc., Kansas City, Mo.). Porphyrins were detected with a commercial disk reagent (Remel, Lenexa, Kans.). Stab cultures for detecting gelatin hydrolysis were incubated at 37°C and then chilled at4°C for 10min before resultswere recorded (5). Hydrolysis of esculin was tested

on heart infusion agarslants containing 1%(wt/vol) esculin 2169

on April 11, 2020 by guest

http://jcm.asm.org/

(2)

TABLE 1. P. caballisp. nov. strains studied

Strain Location ofsender Source Otherclinical information

83851T Washington Uterus Adultfemale

84314 California Lungabscess" Adult male,respiratory and central nervoussystemsigns preceding death

84513 New Mexico Trachea 1-yr-old female

84532 South Carolina Uterus 12-yr-old female

84678 Indiana Pleuralfluid" 3-yr-old female,pneumonia

84687 Washington Uterus 5-yr-oldfemale, metritis

85256 Kentucky Gutteralpouch 10-yr-old female, sinus infection

85683 Maryland Cervix 19-yr-old female,infertile

86026 Pennsylvania Bone 2-yr-oldmale, Actinomyces sp.also isolated

86200 Montana Leg wound 4-yr-oldfemale

86514 Wisconsin Lunga 4-yr-old female, pneumonia and meningitis

87127 Illinois Lunga 14-yr-old female,inhalationpneumonia

87151 California Lunga 4-yr-oldmale,respiratory disease, centralnervous systemsigns

87658 Alabama Lung' 3-mo-oldfemale, fibrinouspneumonia, pleuritis

87659 Alabama Lung 2-mo-old male, pneumoniaRhodococcusequialsoisolated

88036 Michigan Trachea 5-mo-oldmale,pneumonia, Actinobacillus equuli also isolated 88037 Michigan Mesentericabscess" 11-yr-oldmale

88046 Alabama Infectedincision 7-yr-old female, mixed culture

88176 Kentucky Fistulouswithers 2-yr-oldmale, Streptococcuszooepidemicus also isolated 88180 Michigan Trachea 5-mo-oldmale,Bordetella bronchiseptica also isolated

88181 Michigan Lung 6-mo-oldfemale

88193 Maryland Brain 6-yr-old female

88200 Kentucky Spleen 5-yr-old male

88228 Kentucky Lung Adultfemale,pneumonia,Streptococcuszooepidemicusand

Pseudomonasaeruginosa alsoisolated

88282 Michigan Lung 15-yr-oldfemale, centralnervoussystemsigns, Actinobacillus equuli also isolated

88283 Michigan Growthonback 6-yr-old male, mixed culture

88322 Michigan Uterus Adultfemale,Streptococcuszooepidemicus also isolated 88405 Kentucky Peritonealfluid" 11-yr-old female,colicsurgery

88406 Kentucky Gutteralpouch" 15-yr-old female

"Case histories indicatedthat P.caballiwasisolated inhigh numbersorpurecultureorboth from these tissues.

and 0.5% ferric chloride (19). The test for methylene blue milk reduction was performed as described by MacFaddin (13). Glycosidase enzyme tests were done as described by Kilian (9), using 0.1% (wt/vol) solutions of

2-nitrophenyl-P-D-galactopyranoside

(Calbiochem-Behring, La Jolla, Cal-if.),

4-nitrophenyl-p-D-glucopyranosiduronic

acid (Sigma Chemical Co., St. Louis, Mo.), and

4-nitrophenyl-ot-L-fu-copyranoside (Sigma). For fermentation tests, carbohy-drates (1%, wt/vol) were incorporated into heart infusion broth (Difco) with 1.0% (vol/vol) Andrade's indicator. Gas production from D-glucose was detected in a Durham tube. Cultures were observed daily for 1 week for acid and gas production.

DNA relatedness. The procedures used for preparing un-labeled DNA, the hydroxyapatite method for DNA hybrid-ization, and the calculations used to determine relatedness have been described elsewhere (3). DNA from P. caballi

83851T (type strain) was labeled with

[32P04]dCTP

by nick translation, using a commercial kit (catalog no. 8160SB;

Bethesda Research Laboratories, Inc., Gaithersburg, Md.). Because of the relatively low guanine-plus-cytosine (G+C) content of P. caballi DNA, hybridization reactions were

done at 55°C (optimal reassociation) and 70°C (stringent reassociation).

Determination of DNA base composition. The G+C content of DNA prepared from four P. caballi strains was

deter-mined

by

denaturingDNAdissolved in lx SSC

(lx

SSC is 0.15 MNaCl

plus

0.015 M sodium

citrate)

in a model 2600

spectrophotometer

(Gilford

Instrument

Laboratories,

Inc.,

Oberlin, Ohio) equipped

withamodel2527

thermoprogram-mer. The G+C contents were calculated from the thermal

denaturation

midpoint

temperatures

(T,,)

by

using

the

for-mula % G+C = (Tm -

69.3)/0.41 (15).

Proteus mirabilis

ATCC 14273 was used as acontrol.

Determinationof cellularfattyacids.Thecellular

fatty

acid

compositions

of the P. caballi and reference strains of

Pasteurellaceae were determined

by

gas-liquid

chromatog-raphy

by

methods described

by

Moss and co-workers

(16).

Cellsfor

fatty

acid

analysis

were obtained after

growth

on

heart infusion agar with 5% rabbit blood.

Fatty

acids

(as

methyl

esters)were

analyzed by

gas-liquid

chromatography,

using

the HP 5898A microbial identification system

(Hewlett-Packard

Co.,

Avondale, Pa.).

Identifications of

individual

fatty

acids were confirmed

by

hydrogenation,

acetylation,

and mass spectrometry.

Antimicrobial susceptibility tests.

Antibiograms

were de-termined

by

the disk diffusion method of Bauer et al.

(1),

using

Mueller-Hinton agar

(Difco) supplemented

with 5%

sheep

blood and 1% IsoVitaleX

(BBL

Microbiology

Sys-tems,

Cockeysville,

Md.).

The antibiotics and the

concen-trations used are listed in Table 3. Zone sizes were

inter-preted

as

resistant,

intermediate,

or

susceptible

according

to

on April 11, 2020 by guest

http://jcm.asm.org/

(3)

TABLE 2. Strains used in DNA relatedness studies

Organismand strain" _designationOther Source Sender'

Pasteurella multocida subsp. multicida NCTC 10322T ATCC 43137T Swine NCTC

P. multocida subsp. gallicida NCTC10204T Bovine NCTC

P. multocida subsp. septicaNCTC 11619 Human NCTC

P.multocida ATCC27883T Swine ATCC

P. anatisNCTC 11413T Duck NCTC

P. aviumATCC 29546T Chicken ATCC

P. canisNCTC 11621T Dog NCTC

P. dagmatis NCTC11617T Human NCTC

P.gallinarumP-913 ATCC 13361T Chicken K. Rhoades

P.haemolytica ATCC33396T NCTC9380T Sheep ATCC

P. langaa NCTC 11411T Chicken NCTC

P.pneumotropica ATCC 35149T NCTC8141T Mouse ATCC

P. stomnatis NCTC11623T Dog NCTC

P.testudinis ATCC33688T Tortoise ATCC

Actinobacillus equuli ATCC19392T NCTC 8529 Horse ATCC

A.lignieresii ATCC 19393T NCTC 4976 Bovine ATCC

A. suis88441 Horse Va.

A.ureae NCTC 10219T ATCC25976 Human NCTC

Actinobacillus sp. Ross 63' ATCC 27073 Swine R. F. Ross

Haemophilus influenza ATCC 33391T NCTC 8143 Human ATCC

H. aphrophilus F3436 Human R. E. Weaver

'H. somnus" ISU 8025 Bovine M. L. Kaeberle

"Straindesignations arethoseof the sender. T, Type strain. NCTC,National Collection of Type Cultures, Central Public Health Laboratory, London, England;ATCC, AmericanTypeCulture Collection, Rockville, Md.

h R. E.Weaver, Special Bacteriology Reference

Laboratory.

Centers for DiseaseControl,Atlanta, Ga;K.Rhoades, National Animal DiseaseCenter, Ames, Iowa.;R. F.Ross, Veterinary Medical Research Institute,IowaStateUniversity,Ames; M. L. Kaeberle. College of VeterinaryMedicine, Ames, Iowa.; Va.,

stateof origin ofculturesubmittedto theNationalVeterinary Services Laboratories,Ames,Iowa. ' Seereference20.

the

guidelines

of the National Committee for Clinical

Labo-ratory Standards

(18).

RESULTSANDDISCUSSION

Sources and habitats. All P. caballi strains were isolated

from

equine

clinical

specimens

and representhorsesfrom13

states. The

larger

number of strains submitted from

Ken-tucky

and

Michigan

may

reflect

an interest on the part of

colleagues

in this

study

rather than an

unusually high

inci-TABLE 3. Antimicrobialsusceptibilityresults for 29

P. caballi strainsby agar diffusion

Antimicrobial agent No. of strains

(diskconcn)" Resistant Intermediate Susceptible

Amikacin(30) 0 3 26

Ampicillin (10) 0 0 29

Carbenicillin(100) 0 0 29

Cephalothin(30) 0 0 29

Chloramphenicol (30) 0 0 29

Erythromycin(15) 0 3 26

Gentamicin(10) 0 0 29

Kanamycin(30) 0 2 27

Lincomycin(2) 29 0 0

Nalidixicacid(30) 0 1 28

Neomycin (30) 4 il 14

Penicillin G (10) 0 8 21

Streptomycin(10) 12 6 il

Sulfadiazine (300) 26 3 0

Tetracycline(30) 0 0 29

Tobramycin (10) 0 0 29

Trimethoprim (5) il 0 18

'Alldisk concentrationsexcept forpenicillinGareinmicrograms. Peni-cillin G concentration isgiven inunits.

dence of this organism. Of the 29 strains, 15 were isolated from respiratory sources, suggesting that P. caballi, like other Pasteurella spp., is an inhabitant of the mucous

membranes of the upperrespiratory tract. P. caballi strains were isolated in high numbers and in pure culture from horses with upper-respiratory-tract infection, pneumonia, andperitonitis and from a mesenteric abscess. This evidence suggests that P. caballi played a significant role in the pathogenesis of those infections. The clinical significance of P. caballi is less certain in instances inwhich theorganism was presentin mixed cultures.

Phenotypic characteristics. The 29 equine strains studied

were small, nonmotile, bipolar-stained, gram-negative rods TABLE 4. Variablephenotypic characteristics

of P. caballi strains

Cumulative %of

strainspositive Reaction oftype

Test on day: strain 83851"

1 2 7

Acidfrom:

mvo-Inositol 14 14 14

-Lactose 0 14 100 (+)

Maltose 76 83 93 +

D-Mannitol 93 97 97 +

Raffinose 0 7 72 (+)

L-Rhamnose 21 21 21

D-Sorbitol 14 14 14

D-Xylose 59 59 59 +

Growthat42°C 93 93 93 +

L-Ornithinedecarboxylase 41 59 66 +

Yellowpigment 76 76 76 +

-,Negative after 7 days;+. positive within 2days;t+),positive in 3 to 7days.

on April 11, 2020 by guest

http://jcm.asm.org/

(4)

TABLE 5. Differentiation of P. caballi from selected aerogenic species of Pasteurellaceae'

Characteristic P. P.aero- P.dag- H.aphro- Actino-caballi genes matis philus bacillusSp."

Gas fromD-glucose + + + + +

Growth on - + - - +

MacConkey agar

Catalase - + + - +

C02enhances growth - - - +

Indole - - +

Betahemolysis - - - - +

L-Ornithine d' d

decarboxylase

Urease - + + - +

Acidfrom:

L-Arabinose - d - - +

myo-Inositol d d - d +

L-Rhamnose d d - - +

D-Sorbitol d d - - +

Trehalose - - + +

D-Xylose d d - - +

Usualhost

Human - - - +

Horse +

Pig - + - - +

Dog or cat - - +

aDatafor previously described speciesareadaptedfromreferences4, 10,

11,17,and 20.

bSeereference20.

Cd, Strainsvarywithin species.

that gave positive reactions in tests for oxidase, phos-phatase, and nitrate reduction. In addition, all strains were

facultative anaerobes, and all produced acid from the

fer-mentation of carbohydrates. Additional characteristics that were common to all strains are given below in the species

description. Variable characteristics for the 29 strains are

given in Table 4. The 29 equine strains studied had the

characteristics of the family Pasteurellaceae (14). Of the three existinggenera in this family(i.e., Pasteurella,

Hae-mophilus, andActinobacillus), theseequinestrains are best

placed in the genusPasteurella (4, 17). A lack of a

require-ment for preformed growth factors excludes these strains

from the genus Haemophilus. Members of the genus

Acti-nobacillus typically produce sticky colonies, grow well on

MacConkeyagar, andhydrolyzeurea,characteristics absent in these strains. Several characteristics unusual for Pas-teurella spp. were observed: a negative catalase test,

pro-duction ofgas from D-glucose, and the production of acid

frommyo-inositol and L-rhamnose by aminority of strains.

However, we do not believe these differences warrant exclusion of these strains from thegenuswhen thenumerous morphological and biochemical characteristics in common with Pasteurella spp. are considered. The tests that are

useful fordifferentiating P. caballi from selected aerogenic members of the familyPasteurellaceaearegiven in Table 5.

Negative testsfor catalase and urease as well as a lack of

growth on MacConkey agar distinguish P. caballi from

Pasteurella aerogenes and theActinobacillus sp. described

by Ross et al. (20). P. caballi can be distinguished from Pasteurella dagmatis by its negative reactions for catalase, indole, urease, andfermentation of trehalose. P. caballi is phenotypicallymostsimilartoHaemophilus aphrophilus but

can be differentiated by its negative test for trehalose fer-mentation and its ability to grow well without additional

carbon dioxide. The source of an isolate would be of

particular importance in differentiating P. caballi from H. aphrophilus,since eachappearstohavealimited hostrange. Antibiotic susceptibility. The responses of the P. caballi strains to several antibiotics are shown in Table 3. The

antibioticsusceptibilities of these strainsaregenerally

con-sistent with those reported for other members of thegenus Pasteurella (7).

Fatty acid analyses. The fatty acid composition of 29 strains ofP.caballiwassimilartothoseof strains of 12 other species of Pasteurellaceae examined (Table 6). Relatively large amounts of tetradecanoic (C14:0), hexadecenoic (C16:1

w7c; palmitoleic), and hexadecanoic (C16:0; palmitic) acids

werefound. Alsopresent,but inlowerconcentrations,were 3-hydroxy-tetradecanoic

(3-OHC14:0),

cis-9,12-octadecadie-noic (C18:2; linoleic), cis-9-octadecenoic (C181 9c; oleic), and octadecanoic (C18:0; stearic) acid. A small amount (2%

or less)ofdodecanoic (C12:0; lauric) acid and an isomer of

tetradecenoic (C14:1) were also present in all strains

exam-ined. These results are in general agreement with those of Jantzen and co-workers (8), who examined 18 species of Pasteurellaceae,exceptthatweobtained smalleramountsof

TABLE 6. Cellularfatty acidcompositionof P.caballi and otherspeciesofPasteurellaceae

%of totalfattyacidcomposition (SD)" Species No. ofstrains

C14:0 3-OHC14:0 CI6:,w7c C16:0 C18:2 C18:1w9c C18:0

Pasteurella caballi 29 30(2) 9(2) 28(3) 16(2) 5(1) 4(1) 2(1)

P.multocida il 19(3) 5(1) 29(2) 32(3) 4(1) 2(1) 3(1)

P.dagmatis 4 25(3) 8(1) 29(5) 22(2) 6(2) 3(1) 3(1)

P.pneumotropica 3 25(5) 5(1) 31 31(1) 1 2(1) 2(2)

P.aerogenes 2 22 7(1) 31(2) 28 3 2 2

P.canis 2 21(7) 6 24(5) 28(8) 9 4(1) 4(3)

P.stomatis 2 25(4) 10(3) 22(5) 21(3) 9(1) 6(1) 3

Actinobacillusequuli 4 25(5) 7(1) 27(4) 19(2) 8(3) 5(2) 3(2)

A.Iignieresii 3 25(2) 7(1) 31(2) 22(2) 5 (1) 3 2

A. ureae 3 29(5) 6 23(5) 24(5) 6(2) 5(2) 3(1)

Actinobacillus sp.c 2 32(6) 7(1) 32 18(5) 4 (1) 2 1

Haemophilusaphrophilus 7 21(40) 6(2) 26(9) 33(3) S (3) 4(2) 2(1)

H. influenzae 4 18 (6) 4(2) 35(5) 35 (3) 1 T 2(1)

aThe first number is thearithmeticmeanof the relative percentage of thefattyacid;absence ofanumber inparenthesesindicatesthat thestandarddeviation

wasless than0.6%. Allvalues were rounded to the nearest whole number.T,less than0.6%detected.

bNumbers before the colon indicate the numberofcarbon atoms, and numbers after the colonrefertothe numberof double bonds. 3-OH indicateshydroxy group atcarbon3; w7c and w9c indicatedouble-bondpositionfromhydrocarbonendof carbonchain;cis cis isomer.

CSeereference20.

on April 11, 2020 by guest

http://jcm.asm.org/

(5)

TABLE 7. DNA relatedness of P. caballi strains

Source(s) ofunlabeledDNA Pasteurella caballi 83851T 85683 88176 84679 88282 84513 88228 87127 84532 85256 87658 88283 88405 86514 88200 87151 86026 88180 84314 84687 87659 88193 88322 88046 86200 88036 88181 88037 88406

P.langaa NCTC11411T

Actinobacillus sp. ATCC 27073

PasteurellapneumotropicaATCC35149TandP.multocida subsp.gallicida NCTC 10204T

P. canisNCTC 11621T,P. multocida subsp.septicaNCTC 11619,P. multocida subsp. multocida NCTC10322T,P. avium ATCC29546', and P.dagmatis NCTC11617T

P. anatisNCTC 11413', P.stomatisNCTC 11623',P. testudinis ATCC

33688',

Actinobacillus equuli ATCC

19396',

A. lignieresiiATCC 19393T,

A.ureaeNCTC 10219T, Haemophilusaphrophilus F3436,H.influenza

ATCC33391T,and "H. somnus" ISU8025

Pasteurellagallinarum P-913andActinobacillussuis88441 P. aerogenesATCC27882Tand P. haemolytica ATCC33396T

Relatednessto labeled DNAfromP.caballi 83851T

RBR at55°C" oDb RBRat700C

100 99 98 97 97 94 94 94 93 93 93 93 93 92 92 91 90 90 89 89 89 89 89 88 87 87 87 85 83 53 36 30-35 0.0 1.0 1.0 1.0 1.0 1.0 2.0 1.5 1.0 1.0 1.5 1.0 1.5 1.0 1.0 1.5 1.0 1.0 1.0 1.0 1.0 1.5 1.0 1.0 1.0 1.0 1.0 1.0 1.0 11.5 14.0 100 85 77 86 88 91 89 87 86 77 82 85 86 87 87 85 88 85 85 85 84 89 84 87 82 85 88 82 82 21 il 25-29 20-24 15-19 10-14

aRBR, Relativebinding ratio (calculatedas100x percent DNA bound tohydroxyapatiteinheterologousreactions/percentDNA boundtohydroxyapatitein

homologousreactions.

b%D,Percentdivergence (calculatedontheassumptionthat1%decrease in thermalstabilityofaheterologousDNAduplexwascausedby1% unpaired bases). Itisexpressedtothenearest0.5%.

3-OH

C14:0

acid. This result is due todifferencesinmethods,

because our base hydrolysis procedure does not

quantita-tively liberate all amide-linked hydroxy acids. Althoughfatty

acid composition is oflittle value in differentiating among generaand species of Pasteurellaceae, all members of the

family are easily distinguished from other small,

gram-negative, rod-shaped bacteria by their characteristic fatty

acid composition.

DNAstudies.The mean G+C contentofDNAfrom four P. caballi strains, including the type strain, was 41.2 to 42.4

mol%.These values are within the range (37.7 to 45.9) given

for the genus Pasteurella (4). The results of DNA-DNA hybridization experiments are shown in Table 7. Strains

designated P. caballi formed a single DNA relatedness group.The levelsof relatedness of labeled P. caballi DNA to DNA from 28 other P. caballi strains were 83 to 99% (average, 91%) in reactions at the optimal reassociation

temperature. DivergenceinrelatedDNAsequenceswas1 to

2%. Relatedness remained high (average, 85%) under

strin-gentreassociation conditions. Strain

83851'

was53%orless

related to other species ofPasteurellaceae, indicating that theequine strainsrepresenta new species. Closest related-nessoccurred with Pasteurellalangaa. Relatednesstoother species of Pasteurellaceae was 13to 34%.

Insummary, resultsofphenotypiccharacterization, fatty acid analyses, and DNA-DNA hybridization studies show

that these equinestrainsform anew, homogenous taxon in the genus Pasteurella. We propose that these strains be

designated Pasteurella caballi.

Description of P. caballi sp. nov. Pasteurella caballi

(ca.bal'li. L.n. caballus, a horse; L.gen.n. caballi, of a horse; Pasteurella caballi, thePasteurella ofa horse) cells

arerod-shaped (0.8to 1.0by1.3 to 1.9F.m), bipolar stained,

and arranged singly or in pairs with occasional swollen,

on April 11, 2020 by guest

http://jcm.asm.org/

(6)

curved, or filamentous forms observed. The bacterium is gram negative, not acid fast, nonmotile, and facultatively anaerobic. Endospores are notformed. Colonies are 1.0 to 1.5 mm in diameter, smooth, slightly raised, and grayish yellow on blood agar after 24 h of incubation. Colonies grow atroomtemperature and at 37°C butnotat 4°C. Colonies are nothemolyticon bovine, sheep, or rabbit blood agar but do impart a brownish discoloration to these media under areas

ofconfluent growth. Growth scraped up from the surface of the blood agar is cream (less frequently) to bright yellow (usual). Growth on MacConkey agar is scant or absent. The

bacterium is X factor and V factor independent and positive

for oxidase, phosphatase, and

P-galactosidase.

Nitrate is reduced to nitrite. Methylene blue milk is reduced. Porphy-rins are produced from 8-aminolevulinic acid. Catalase,

indole,lysinedecarboxylase,arginine dihydrolase,

a.-fucosi-dase, and 3-glucuronidase are not produced. Esculin,

gela-tin, and urea are not hydrolyzed. Citrate is not utilized. Voges-Proskauer and methyl red tests are negative.

D-Glucose is fermented, with the production of acid and gas.

Acid is produced from D-fructose, D-galactose, lactose, D-mannose, and sucrose. Fermentation of lactose is delayed. Acid is not produced from adonitol, L-arabinose, cellobiose,

dulcitol, inulin, salicin, starch, or trehalose. The bacterium

contains 3-OH

C14:0

and approximately 25% each of

C14:0,

C16,

7c, and

C16:0

acids, which is the characteristic fatty

acid profile of all members of the family Pastuerellaceae. The G+C content of DNA is 41 to 42mol%

(T,,,).

P. caballi

strains have been isolated fromequine clinical specimens.

The type strain of P. caballi is 83851 (=ATCC 49197), and it

exhibits all of the characteristics described above.

Addi-tional variable characteristicsare givenin Table 4. The type

strain was isolated in 1983 from the uterus ofa mare. The

G+C contentofthe DNA was 42 mol%.

ACKNOWLEDGMENTS

WethankJ. M.Donohue, R.Walker,A.Render, andothers who submittedisolates.

LITERATURE CITED

1. Bauer, A. W., W. M. M. Kirby, J. C. Sherris, and M. Turck. 1966. Antibiotic susceptibility testing by standardized single diskmethod. Am. J. Clin. Pathol. 45:493-496.

2. Biberstein, E. L. 1981. Haemophilus-Pasteurella-Actinobacil-lus: their significance in veterinary medicine, p. 61-74. In M. Kilian,W. Fredericksen, andE. L.Biberstein(ed)., Haemoph-ilus, Pasteurella, and Actinobacillus. Academic Press, Inc. (London), Ltd., London.

3. Brenner, D. J., A. C. McWhorter, J. K. Leete Knutson, and A.G. Steigerwalt. 1982. Escherichiavulneris:a newspecies of Enterobacteriaceae associated with human wounds. J. Clin. Microbiol. 15:1133-1140.

4. Carter,G. R. 1980.Genus I.Pasteurella,p. 552-557. In N. R. Krieg and J. G. Holt (ed.), Bergey's manual of systematic bacteriology, vol. 1. TheWilliams & WilkinsCo.,Baltimore. 5. Clark, W.A., D. G. Hollis,R.E. Weaver, and P.Riley. 1984.

Identification ofpathogenic gram-negative aerobicand faculta-tively anaerobic bacteria, p. 14. Centers for Disease Control, PublicHealthService, U.S. Departmentof Health and Human

Services, Atlanta.

6. Cottral, G. E. (ed). 1978. Manual of standardized methods for veterinary microbiology, p. 697-698, 701. Cornell University Press, Ithaca, N.Y.

7. Fales, W. H., L. A. Selby, J. J. Webber, L. J. Hoffman, L. D. Kinter, S. L. Nelson, R. B. Miller, J. G. Thorne, J. T. McGinity, and D. K. Smith. 1982. Antimicrobial resistance among Pas-teurella spp. recovered from Missouri and Iowa cattle with bovine respiratory disease complex. J. Am. Vet. Med. Assoc. 181:477-479.

8. Jantzen, E., B. P. Berdal, and T.Omland. 1981. Cellularfatty acidtaxonomy ofHaemophilus, Pasteurella, and Actinobacil-lus, p. 197-204. In M. Kilian, W. Fredericksen, and E. L. Biberstein (ed.), Haemophilus,Pasteurella, andActinobacillus. Academic Press, Inc. (London), Ltd., London.

9. Kilian, M.1978. Rapid identificationof Actinomycetaceae and relatedbacteria.J.Clin. Microbiol. 8:127-133.

10. Kilian, M., and E. L. Biberstein. 1980.Genus Il. Haemophilus, p.558-569. In N. R.Kriegand J. G. Holt(ed.), Bergey'smanual ofsystematic bacteriology, vol. 1.TheWilliams & Wilkins Co., Baltimore.

11. Kilian, M., and W. Frederickson. 1981. Identificationtablesfor theHaemophilus-Pasteurella-Actinobacillus group, p.281-290. In M. Kilian, W. Frederickson, and E. L. Biberstein (ed.), Haemophilus, Pasteurella, and Actinobacillus. Academic Press, Inc. (London), Ltd.,London.

12. Lanyi, B. 1987. Classical andrapid identification methods for medicallyimportantbacteria, p. 1-67. In R. R. Colwell and R. Grigorova (ed.), Methods in microbiology, vol. 19. Academic Press, Inc.(London), Ltd., London.

13. MacFaddin, J. F. 1980. Biochemical tests for identification of medicalbacteria, 2nd ed.,p.205-209. TheWilliams& Wilkins Co., Baltimore.

14. Mannheim, W. 1980. Family lII. Pasteurellaceae Pohl, p. 550-552. In N. R.Krieg andJ.G. Holt(ed.), Bergey'smanual ofsystematic bacteriology, vol.1.The Williams &WilkinsCo., Baltimore.

15. Marmur, J., and P. Doty. 1962. Determination of the base composition of deoxyribonucleic acid from its denaturation temperature. J. Mol. Biol.5:109-118.

16. Moss, C. W.,P. L. Wallace, D. G. Hollis, and R. E. Weaver. 1988. Cultural and chemical characterization of CDC groups EO-2, M-5, andM-6, Moraxella (Moraxella) species, Oligella

urethralis, Acinetobacter species, andPsychrobacter immobi-lis. J.Clin. Microbiol. 26:484-492.

17. Mutters, R., P. Ihm, S. Pohl,W. Frederiksen, and W. Mann-heim. 1985. Reclassification of thegenusPasteurellaTrevisan 1887 on the basis of deoxyribonucleic acid homology, with proposalsforthe newspecies Pasteurella dagmatis, Pasteurella canis,Pasteurellastomatis,Pasteurellaanatis, and Pasteurella langaa. Int.J. Syst. Bacteriol. 35:309-322.

18. National Committee for Clinical Laboratory Standards. 1975. Performance standards for antimicrobial disc susceptibility tests. National Committee for Clinical Laboratory Standards, Villanova, Pa.

19. Phillips,E., and P. Nash.1985.Culturemedia,p. 1051-1092. In E. H. Lennette, A. Balows, W. J. Hausler, Jr., and H. J. Shadomy (ed.), Manual of clinicalmicrobiology,4thed. Amer-icanSociety forMicrobiology,Washington, D.C.

20. Ross, R. F., J. E. Hall, A. P. Orning, and S. E. Dale. 1972. Characterization of an Actinobacillus isolated from the sow

vagina. Int. J. Syst. Bacteriol. 22:39-46.

21. Schlater, L. K. 1989. Anaerogenic Pasteurella-like organism isolated from horses. J. Vet. Diagn. Invest. 1:3-5.