bacteriological variation among bordetella bronchiseptica ... · b. bronchiseptica b....

JOURNAL OF CLINICAL MICROBIOLOGY, Apr. 1977, p. 471-480Copyright © 1977 American Society for Microbiology

Vol. 5, No. 4Printed in U.S.A.

Bacteriological Variation Among Bordetella bronchisepticaIsolates from Dogs and Other Species

DAVID A. BEMIS,* HELEN A. GREISEN, AND MAX J. G. APPEL

James A. Baker Institute for Animal Health, Cornell University, Ithaca, New York 14853

Received for publication 13 December 1976

Bacteriological properties of 50 isolates of Bordetella bronchiseptica were

compared. Phase variation, which involved colonial morphology and its associ-ated characters of hemagglutination, hemolysis, acriflavine agglutination, crys-tal violet staining, flagellation, and fimbriation, occurred among these isolates.Organisms representing the three observed morphotypes did not have differentgrowth rates, nor were any differences in their bacteriological characteristicsobserved after repeated subculture on agar. There were also variations inantimicrobial drug susceptibility, especially to sulfonamide-trimethoprim, andin nitrate reduction. The relationships among these variable parameters werenot apparent. None of the observed variations could be attributed to differencesin the species of origin.

Bordetella bronchiseptica has long beenknown as a disease-producing organism of labo-ratory animals and has been implicated in sev-eral disease syndromes of domestic animals. Ithas been isolated from a variety of species in-cluding dogs, cats, guinea pigs, rats, mice, rab-bits, pigs, horses, turkeys, monkeys, humans,and a number of wild animals (9, 10, 11, 19, 21,30, 31, 33, 37). Despite the wide host range ofB.bronchiseptica, taxonomic relationships amongstrains of this organism are generally un-known, and subspecies classification into sero-groups, phage types, or biotypes has not yetbeen accomplished.Although methods for the clinical identifica-

tion ofB. bronchiseptica have been well estab-lished (18, 27), there is considerable disagree-ment regarding some of the classical culturaland biochemical properties ofthe organism. Re-ports describing the colonial (23, 24) and ultra-structural (22, 26) morphologies of B. bronchi-septica have been few, and differences in thesemorphological properties are not well known.

It has been suggested that B. bronchisepticaisolates from different animal species vary inpathogenicity for swine (28); however, this ob-servation has not been confirmed, and otherbacteriological differences between B. bronchi-septica isolates from different species have notbeen reported.To select representative B. bronchiseptica

isolates for further pathogenicity studies indogs, we examined the biochemical, cultural,and morphological characteristics of B. bron-chiseptica isolates from dogs and other species.The purpose of this study was to determine the

degree of bacteriological variation among B.bronchiseptica isolates and to observe to whatextent these variations might be used to distin-guish B. bronchiseptica isolates from differentspecies.

MATERIALS AND METHODSBacterial strains. Fifty strains of B. bronchisep-

tica were studied (Table 1). The species of originwere as follows: 30 from dogs, 10 from swine, 3 fromcats, 2 from guinea pigs, 2 from rats, and 1 from amonkey. The species of origin was undeterminedfor the 2 remaining strains of B. bronchiseptica.One strain each of Bordetella pertussis, Bordetellaparapertussis, and Alcaligenes faecalis, and twostrains of unidentified, nonfermentative, gram-negative bacilli isolated from the upper respiratorytract of dogs, were also studied. The strains studiedrepresented both recent isolates and type cultures.On receipt, all cultures were checked for purity onbrucella agar. Forty-eight-hour bacterial growthwas lyophilized in a suspension of 50% of a stabiliz-ing medium consisting of sucrose, phosphate buffers,glutamate, and albumin (3) and stored at 4°C untilused. Subcultures were kept to a minimum prior totesting, except that strains 87, 110 H and 110 NHwere subcultured at 2- to 5-day intervals for a totalof over 100 passages. These strains were then testedat both high and low passage.

Media. Bacteria were maintained on brucellaagar (Pfizer). Primary isolation was on Levineeosin-methylene blue agar (Difco) and blood agar.The blood agar consisted of brucella agar base plus5% sterile sheep blood and IsoVitaleX enrichment(BBL) at a final concentration of 1:50. Bordet-Gen-gou agar (Difco) was prepared according to the man-ufacturer's instructions with 20% sheep blood. Cul-tures on Bordet-Gengou agar were incubated at 35°Cin an atmosphere of 5% CO2; all other cultures were

471

on January 12, 2021 by guesthttp://jcm

.asm.org/

Dow

nloaded from

http://jcm.asm.org/

472 BEMIS, GREISEN, AND APPEL

TABLE 1. Bacterial strains studied

Strain Species Sourcea Strain Species of Sourceaof origin originStrains received as Strains received as

B. bronchiseptica B. bronchisepticaBB Manhattan Dog Appel AMC-Lamazor Dog WilkinsBB 24 II 69 Dog Appel 19705 Dog BemisBB Ithaca Dog Appel D-1 Dog HarrisB. para Dog George SS-232-75 Dog Augst87 Dog Bemis Rat #1 Rat Boyer110 H Dog Bemis Rat #2 Rat Boyer110 NH Dog Bemis BTS Swine Harris138 An Hus Dog Bemis S. madrid Swine HarrisJ696 Dog Bemis Ct. madrid Cat HarrisK704 Dog Bemis Columbus Cat Kahn501 Dog Bemis SHGP-1 Guinea pig Bemis17640 SAC Dog JBS 7-8 NADL Swine Croghan475 An Hus Dog Bemis 495 NADL Swine CroghanDahlund DR64 Dog Tischler 7-11 NADL Swine CroghanHopper DS43 Dog Tischler 2-9 NADL Swine Croghan18731 SAC Dog JBS 5-8 NADL Swine CroghanCD-C47 Dog Bemis 5-4 NADL Swine Croghan19022 SAC Dog JBS Phase-I Tuskegee Swine Jenkins19141 SAC Dog JBS NYS #4 Unknown Smull1388B-3 Dog House NYS #10 Unknown SmullBB-FTC Dog Lauerman ILG Monkey House19395 ATCC Dog ATCC GPA Guinea Pig House780 ATCC D6g ATCC B4 Swine House10580 ATCC Dog ATCC 52190 SAC Cat JBS482-74 Dog Bailie695-74 Dog Bailie Other strains studied

8750 ATCCb Unknown ATCC15237 ATCCC Human ATCC9340 ATCCd Human ATCCH 475e Dog BemisI 98e Dog Bemis

a Abbreviations: Appel, M.J.G. Appel, James A. Baker Institute for Animal Health (JABIAH), CornellUniversity, Ithaca, N.Y.; ATCC, American Type Culture Collection, Rockville, Md.; Augst, V. Augst, LRE,Inc., Kalamazoo, Mich.; Bailie, W. E. Bailie, College of Veternary Medicine, Kansas State University,Manhattan, Kans.; Bemis, D. A. Bemis, JABIAH, Cornell University, Ithaca, N.Y.; JBS, J. Bentinck-Smith, N.Y.S. College of Veternary Medicine, Cornell University, Ithaca, N.Y.; Boyer, C. I. Boyer, Jr.,N.Y.S. College of Veternary Medicine, Cornell University, Ithaca, N.Y.; Croghan, D. Croghan, U.S.D.A.,National Animal Disease Laboratory, Ames, Iowa; George, L. W. George, JABIAH, Cornell University,Ithaca, N.Y.; Harris, D. L. Harris, College of Veternary Nfedicine, Iowa State University, Ames, Iowa;House, J. A. House, Pitman-Moore, Inc., Washington Crossing, N.J.; Jenkins, E. M. Jenkins, School ofVeternary Medicine, Tuskegee Institute, Tuskegee, Ala.; Kahn, D. E. Kahn, College of Veternary Medi-cine, Ohio State University, Columbus, Ohio; Lauerman, L. H. Lauerman, College of Veternary Medicine &Biomedical Science, Colorado State University, Fort Collins, Colo.; Smull, L. L. Smull, Tompkins CountyHospital, Ithaca, N.Y.; Tischler, S.A. Tischler, Valley Veterinary Hospital, Walnut Creek, Calif.; Wilkins,R. J. Wilkins, Animal Medical Center, New York, N.Y.

bReceived as Alcaligenes faecalis.Received as B. parapertussis.

d Received as B. pertussis.e Unidentified nonfermentable gram-negative strains.

incubated at 37°C. Conventional test media wereinoculated with colonies from 48-h brucella agarcultures. Test results were read at varying inter-vals, but each test was completed 7 days after inoc-ulation.

Colonial morphology. Colonial morphologieswere described after observation with direct,oblique, and transmitted lighting. Average colony

size was estimated by measuring at least three well-isolated colonies. Colony type was judged by grossobservation to be either smooth or rough, and thedegree of dissociation was determined by the acri-flavine agglutination and crystal violet-stainingmethods described for brucellae (1). The dissociationrate of a smooth colony-producing strain (110 H) wasdetermined by making repeated plate counts from

J. CLIN. MICROBIOL.


.asm.org/

Dow

nloaded from

http://jcm.asm.org/

VOL. 5, 1977

brucella broth cultures that were incubated at 37°Con a rotating table (250 rpm). Brucella and bloodagar cultures were observed at 48 h, except for B.parapertussis, which grew slowly on this mediumand was therefore observed at 7 days. Growth onBordet-Gengou agar was examined at 1 and 2 days,for most isolates, and at 7 days for the slow-growingstrains.

Hemagglutination. One drop of a heavy bacterialsuspension in phosphate-buffered saline (packed-bacterial-cell volume approximately 8 to 10%) wasmixed with one drop of 20% washed sheep erythro-cytes on a white plastic plate. This mixture wasrotated and observed under direct lighting for hem-agglutinating activity. Final reactions were re-corded after 5 min.

Microscopic morphology. Gram-stained smearsof 24-h brucella broth cultures were examined formicroscopic morphology.

Ultrastructural morphology. Brucella broth cul-tures were incubated on a rotating table (250 rpm) at37°C for 10 to 15 h. Negative-contrast staining wasperformed as follows: Formvar-coated and carbon-stabilized copper grids (200 mesh) were suspendedon a drop of broth culture for 30 s to 1 min, drained,suspended on a drop of 0.1% to 0.5% aqueous potas-sium phosphotungstic acid, pH 7.0, for 30 s to 1 min,and drained until dry. Grids of B. pertussis and B.parapertussis were prepared from 7-day growth onBordet-Gangou agar and required 2% aqueous po-tassium phosphotungstic acid for adequate staining.Grids were examined with a Philips EM 300 electronmicroscope (Philips Electronic Instruments, MountVernon, N. Y.). The overall density of organismsand the staining quality were recorded. In mostcases, a minimum of 20 bacterial cells were ob-served, and their typical ultrastructural morpholo-gies were described. Because the observation of piliwas sometimes obscured in the more poorly stainedspecimens, each negative finding was repeated atleast twice.

Cultural properties. Growth patterns in fluid thi-oglycolate and brucella broths (Pfizer) were de-scribed after 48 h of incubation. The ability ofstrains to grow on salmonella-shigella agar (Difco)and brucella agar containing potassium tellurite(320 mg/liter) was determined after 7 days. Growthrates of three strains, 87, 110 H, and 110 NH, werecompared in brucella broth.

Antibiotic susceptibility tests. Sensitivity to anti-biotics was measured according to the principlesoutlined by Bauer and Kirby et al. (2). The antibioticdisks used (BBL) were as follows: ampicillin (10 ,tg),cephalothin (30 ,ug), chloramphenicol (30 pig), eryth-romycin (15 ,g), gentamicin (15 ,tg), kanamycin (30,ug), nalidixic acid (30 pig), nitrofurantoin (300 ,ug),penicillin (10 U), polymyxin B (300 U), streptomycin(10 ,ug), sulfamethoxazole-trimethoprim (25 gg),and tetracycline (30 ,ug).

Biochemical tests. Biochemical tests were per-formed according to standard methods (8, 36). Allisolates were screened for their inability to fermentglucose on Kligler iron agar (Difco). Ability to uti-lize citrate as a sole carbon source was tested onSimmon citrate medium (Difco). Nitrate reduction

VARIATION OF B. BRONCHISEPTICA 473

was determined after 48 h of incubation in indole-nitrate broth (Difco), and the isolates were tested fornitrate with sulfanilic acid and alpha-naphthylam-ine solutions. Negative reactions were confirmed byaddition of zinc dust. Urea hydrolysis was observedon Christensen urea agar (Difco) after 4 and 18 h.Oxidation of glucose was tested in an open tubecontaining oxidation-fermentation basal medium(Difco) and 1% glucose. This same medium was alsoused to observe motility. Oxidase activity was testedby flooding a 48-h brucella agar slant with a 1%solution of p-aminodimethylaniline oxalate (Difco).Tetrazolium reduction was tested in motility GI me-dium (Difco) containing 75 mg of 2,3,5-triphenylte-trazolium chloride (Eastman) per liter. Cultureswere observed for formation of a red-insoluble for-mazan dye. Gelatin hydrolysis was determined intubes of brucella broth containing 12% gelatin.

RESULTSDetailed bacteriological findings and individ-

ual strain results are on record elsewhere (D.A. Bemis, Ph.D. thesis, Cornell University,Ithaca, N.Y., 1976). All organisms studied weregram-negative rods. None of the strains wereable to ferment or oxidize glucose, nor werethey able to hydrolyze gelatin. Each was aero-bic in its growth requirements and grew only inthe oxidized portion of thioglycollate broth (ifable to grow in this medium). Strains receivedas B. bronchiseptica were motile within 7 days,produced oxidase, rapidly (within 4 h) hydro-lyzed urea, and were capable of reducing tetra-zolium to a red-insoluble formazan. None ofthese strains were able to grow in the presenceof potassium tellurite. Forty-nine strains (98%)were capable of utilizing citrate. By using theabove criteria of motility, oxidase activity, ureahydrolysis, tetrazolium reduction, utilization ofcitrate, and growth on potassium tellurite me-dium, the other related organisms studiedcould be clearly differentiated and excluded asB. bronchiseptica. With one exception, B. bron-chiseptica grew in brucella broth as a heavypellicle adhering to glass with turbiditythroughout the tube. The following representsthe main findings pertaining to variationamong B. bronchiseptica strains.

Colonial morphology. When grown on bru-cella agar, three distinct morphological forms,tentatively called phase I, intermediate phase,and rough phase, were recognized among thestrains of B. bronchiseptica. A summary oftheir properties is given in Table 2. The small,pearllike phase-I colonies were present in only4 of the 50 B. bronchiseptica strains. Thesecolonies were observed to dissociate to rough-phase organisms, especially when cultured inbroth. The rate of dissociation was determinedto be 3 to 4% after 4 days of incubation on a


.asm.org/

Dow

nloaded from

http://jcm.asm.org/


+

+ +

+ +

+1+ +1 o

+

+ +C)) +0 +

+1 0

+1

+

tl

+ Lo O04S-

+ +

+

0

CO

04

+. *; "0

O 0

C,)

O O 4.)

0 bD

C,) _ ;L_ N

Sw c

0- 05 X

J. CLIN. MICROBIOL.

ho0

u:

LO

+++

cd

.

CZ

U

+"0

0

0

0.0

4j

+1

0.

U)̂ *

S cX U)

C.)

a))E;

o *^ .

S..:31

0X +1*

o go 0o M °

0+

rotary shaker. Dissociation to rough phase wasnot reversible in vitro. Colonies in phase I thathad been subcultured over 100 times on agar(each time selecting for original colony type)did not lose any of their morphological proper-ties. The large, rough-phase colonies differedfrom their smooth predecessors in that theywere translucent, raised in the center, and hadan undulating outer margin (Fig. 1 and 2).Only 4 of 50 B. bronchiseptica strains repre-sented rough-phase organisms. The remainingmajority of strains (42 of 50) were of smooth,intermediate colony types. Although six ofthese strains were observed to dissociate torough-phase organisms, it appeared that somestrains might represent stable intermediateforms. Strain 87 was subcultured over 100 timeson agar and numerous times in broth withoutany changes in morphology occurring.

Acriflavine agglutination, hemagglutina-tion, and crystal violet staining of these colo-nies were variable. Each of the four phase-Istrains gave weak- (< 1+) or negative-aggluti-nation reactions in acriflavine, had some hem-agglutinating activity, and stained completelywith crystal violet. The remaining two colonytypes gave variable results on these three tests.All of the strains that gave no hemagglutina-tion reaction were smooth, intermediate-phaseorganisms, as were those that gave strong acri-flavine agglutination. Crystal violet stainingwas found in all phase-I organisms, in 7 of 42intermediate-phase organisms, and in only 1 of4 rough-phase organisms.

Morphological variation was less apparent onblood agar. Only the four strains that producedphase-I colonies on brucella agar were hemo-lytic. Although the two properties were not wellcorrelated, there were more pearl-like colonies(30 of 50) and more hemolytic strains (23 of 50)observed on Bordet-Gengou agar.

Microscopic morphology. Gram staining re-vealed that the B. bronchiseptica strains werethin, gram-negative rods with tapered ends,occurring mostly singly or in pairs. Somestrains also possessed long filamentous forms.

Ultrastructural morphology. When ob-served by negative-contrast staining, B. bron-chiseptica strains were seen to possess a highlyconvoluted cell wall with one to two less elec-tron-dense spots on their surfaces (Fig. 3). Theother species of bacteria examined did not pos-sess these cell wall convolutions. Flagella wereobserved on all B. bronchiseptica strains exceptthe four phase-I strains. When present, theflagella were arranged peritrichously. Numer-ous filamentous appendages (consistent withthe description of pili in other bacteria [5, 7])

cle

to

cI.

Q

U4)

cn

4~

0

ce

.0

C)-

-0

C.DC)

0.

E0

D.o.

LO0

CL))

c4) >0Lo tocD

Cdr.=

obD0. t

E X 8

00-

b0)

r.0

0biD

o

0

0u.0


.asm.org/

Dow

nloaded from

http://jcm.asm.org/


p I wRINFP

_~~~

04..3

FIG. 1. Appearance of smooth- and rough-phase colonies of B. bronchiseptica on brucella agar (25 xmagnification).

i

i

FIG. 2. Rough-phase colony of B. bronchisepticaon brucella agar (100 x magnification).

were found on 82% of the strains tested (Fig. 4).B. parapertussis and B. pertussis were the onlyorganisms other than B. bronchiseptica to pos-sess pili. Most of the piliated strains were capa-ble of producing smooth, phase-I type colonieson Bordet-Gengou agar.

Cultural properties. All but three strains of

B. bronchiseptica were able to grow on salmo-nella-shigella agar, although some grewpoorly. Growth rates of three strains (110 H,110 NH, and 87) representing each of the ob-served morphological phases were not different.Near the end of the logarithmic phase ofgrowth, cultures of 110 H autoagglutinated. Se-quential Gram stains made during the growthstudies showed that the incidence and length offilamentous forms increased during the laterstages of log-phase growth. In strain 110 H,these filaments were arranged in palisades andformed lattice structures when the density be-came great enough. Macroscopically, these lat-tice structures were observed as granular ag-glutination throughout the entire broth cul-ture.

Antibiotic susceptibility. As determined byexisting clinical standards (19), B. bronchisep-tica strains were all sensitive to polymyxin B,chloramphenicol, and tetracycline. Ninety-eight percent of the strains were susceptible togentamicin, and 96% were susceptible to kana-mycin. Nalidixic acid, cephalothin, ampicillin,and sulfonamide-trimethoprim were somewhatless effective (70, 76, and 66%, respectively,were susceptible). All strains were resistant tostreptomycin and penicillin, and only onestrain was partially susceptible to nitrofuran-toin. The zone diameters were distributed over

IVOL. 5, 1977


.asm.org/

Dow

nloaded from

http://jcm.asm.org/


.I;,'1,IFIG. 3. Negative-contrast staining ofB. bronchiseptica (6,000 x magnification) .

FIG. 4. Negative-contrast staining ofB. bronchiseptica (61,000 x magnification).

a narrow range for all antibiotics except sulfon-amide-trimethoprim. Zone diameters for thisdrug ranged from 9 to 63 mm, and a bimodaltype of distribution was suggested.Biochemical tests. Only the test for nitrate

reduction gave variable results. Thirty-threestrains reduced nitrate to nitrite.

Effect of serial passage. Strains 87, 110 H,

and 110 NH were subcultured over 100 times onbrucella agar with care taken to select for theoriginal colony type. The bacteriological char-acteristics of these high-passage strains wereidentical to those observed at low passage.B. bronchiseptica variation between differ-

ent host species. Table 3 summarizes the varia-bility of B. bronchiseptica isolates from differ-

J. CLIN. MICROBIOL.


.asm.org/

Dow

nloaded from

http://jcm.asm.org/


O

C 40

'0_i, .+i . =

ci: CUY

b_0WDY*

..C ° c.

0 0--

CCO

to o

., CO ^ Oto

~cio 0, U)0 Z-

CU

CU

U 0

0'QCUI_

0:

0.

CI)

o) ccne1 r-4

C)

t- 0C14 C-I

~C

014

CO CZ>

01

C-1 C14 1- e

00 c CS0 1 1

00 CeD _1- _ O

L:CfO_ 0_-

CDO 1 _ 00C1

0 C DC0>0

0 C >010100

0" -40100

1-

CZ~~~~~~C

O CU O CU 0 O

oz

VOL. 5, 1977

to

.I

C.)cn

SU

C..

0

C..

C.)

CU

2-CC.)SU2..-o

..I

04

CO

CU

CD0

C.)

CD"0CDiCU

C.)

CU

CZ

20

e )n

CU

0

0.

0CU0

C.)cUCU

a-CU

CU0

.ob1CUaz

o 5bwC

CUCUCCUC*bDUC--=

CU CUblc: Yc CU

00101 qCUCUCI

CUocU CU C^ciH;

-4) 4. a--OCU0d 0 UOc-c,-

CI) CI Z


.asm.org/

Dow

nloaded from

http://jcm.asm.org/


ent species. The same degree of bacteriologicalvariation was observed in canine and noncan-ine isolates of B. bronchiseptica.

DISCUSSIONMany authors have stressed the use of

"smooth" or "phase-I" strains of Bordetellabronchiseptica for conducting pathogenicity ex-periments or producing antigen. However, acomparative description of such strains and thesignificance of their use are not generallyknown. Nakase (23, 24) described four differentcolonial morphologies of B. bronchiseptica,which he called, "phase I, II, III and roughphase." Colonial morphologies similar to thoseobserved by Nakase appeared to represent themajor variables among the B. bronchisepticastrains included in the present study. Althoughmost produced smooth colony types, the inci-dence of phase-I isolates was only 8%. Contraryto the observation of Nakase, the dissociationrate of phase-I organisms was low, and morpho-logical characteristics could be maintained forover 100 subcultures on agar. Nakase notedthat organisms producing the highly convex,entire-edged phase-I colonies possessed a fewflagella and were encapsulated; he also notedthat the organisms of the flat, irregular-edgedrough-phase colonies were not encapsulatedand did not possess flagella. These findings donot represent the usual relationship betweencolonial morphology, possession of capsules,and possession of flagella. In the present study,organisms producing phase-I colonies did notpossess flagella, while organisms producingrough-phase colonies with spreading perime-ters possessed numerous flagella. Capsules,though they may be important components ofB. bronchiseptica pathogenicity, were not read-ily demonstrated in any of the strains studied.The tendency for phase-I organisms to be

stained more heavily by crystal violet stain isthe converse of the reaction observed with bru-cella organisms. Kang et al. (14) described acidagglutination as a property of phase-I B. bron-chiseptica organisms. Acriflavine agglutina-tion, a property of rough brucella organisms,was positive to some degree in most strainsstudied. Although variation is usually unidi-rectional in vitro, passing from smooth torough, rough- to smooth-phase variation hasbeen suggested in some species (6, 29). Thespecificity of the above reactions for distin-guishing all smooth- or rough-phase organismsmust be questioned.

Hemolysis on blood agar was observed onlywith phase-I organisms; however, some addi-tional strains that were not hemolytic on blood

J. CLIN. MICROBIOL.

agar were hemolytic on Bordet-Gengou agar.This difference in growth media may accountfor the reported differences in the hemolyticcharacter of B. bronchiseptica. There was alsoa tendency to form more phase-I type colonieson Bordet-Gengou agar; so, as suggested byLacey (17), the composition of the medium maydetermine to some extent the colonial structureof an organism.

Bacterial hemagglutination has been de-scribed as a characteristic of certain bacterialsurface structures such as capsules or pili (7,12, 15, 25). Hemagglutination was observed formost strains of B. bronchiseptica and was notinhibited by the presence of D-mannose (0.7%concentration). This finding, at least in somespecies, is indicative of nonfimbrial hemagglu-tination (35). However, fimbrial-associated he-magglutination by Neisseria species has re-cently been shown to be unaffected by D-man-nose (16, 38). Hemagglutination tended to bestronger among the phase-I isolates but wasalso observed for some strains producing rough-and intermediate-type colonies. Kang et al. (14)suggested that differences in hemagglutinatingability may be due to differences in origin, ashis single canine isolate reacted very stronglyin the hemagglutination test. No such differ-ences could be found in the present study.

In other bacterial species, especially thoseorganisms that colonize mucosal surfaces, thepossession of pili or fimbriae has been associ-ated with colonial morphology (4, 22, 29, 32).There has been only one published electronmicroscope study ofB. bronchiseptica (26), andpili were not observed on these organisms.Morse and Morse (22) suggested that pilus-likestructures similar to those found on phase-I B.pertussis organisms were also present on twostrains of B. bronchiseptica. Although 82% ofthe B. bronchiseptica strains studied possessednumerous thin filamentous appendages resem-bling pili, 82% were hemagglutinating and 92%represented smooth-colony types; an absoluterelationship between these properties could notbe established.

Antibiotic susceptibility and nitrate reduc-tion, two additional characteristics that couldnot be readily associated with colonial morphol-ogy, also indicated that there was no differencebetween B. bronchiseptica isolates from differ-ent species. Of those antimicrobials agents hav-ing variable effects, sulfonamide-trimethoprimwas of greatest interest. Interpretation of sus-ceptibility relies on the correlation betweenzone diameters and minimum inhibitory con-centrations of the drugs against many differentbacteria (20). Strains producing zone diameters


.asm.org/

Dow

nloaded from

http://jcm.asm.org/

VOL. 5, 1977

smaller than the smallest zone observed forbacteria considered to be susceptible are re-ported as resistant. Conversely, strains produc-ing zone diameters larger than the largest zoneobserved for bacteria considered to be resistantare reported as susceptible. An area of overlapexists between these two ranges of zone diame-ters and presents difficulty in estimating thetrue susceptibility of some strains. Interpreta-tion of susceptibility of B. bronchisepticastrains by the diffusion test may be furtherinfluenced by their somewhat slower growthrates than most bacteria used to standardizethe test. Regardless of the determined suscepti-bility, it is of importance to note that two popu-lations of organisms emerged with regard tozones of inhibition produced by sulfonamide-trimethoprim. There are likely to be some addi-tional characters associated with these two pop-ulations. R-factors conferring resistance to sul-fonamides and other antimicrobial agents havebeen found in isolates from swine (13, 34).There was no associated species distributionnor any accompanying antibiotic resistance inthese sulfonamide-trimethoprim-reactive pop-ulations.

It was apparent from this study that nitratereduction, though frequent, was not a consist-ent feature ofB. bronchiseptica strains. A ten-uous relationship between nitrate reductionand the previously mentioned pattern of sus-ceptibility to sulfonamide-trimethoprim wasnoted. All strains that produced low-zone diam-eters against the sulfonamide drug were nitratepositive, and, conversely, all strains that werenitrate negative produced high-zone diametersagainst the sulfonamide.The degree of bacteriological variation

among strains ofB. bronchiseptica was surpris-ingly small considering the parasitic natureand wide host range of this organism. Becausethe same degree of variation was observed incanine and noncanine isolates, there does notappear to be any promise of differentiating B.bronchiseptica isolates from different species.Serological studies on B. bronchiseptica iso-lates from different species are needed; how-ever, initial studies on isolates from guineapigs indicate that serological differences areagain related to colonial phase variation (23).The significance of capsules, pili, and otherproperties associated with colony morphologyin the pathogenicity of B. bronchiseptica re-quires further study.

ACKNOWLEDGMENTSWe wish to thank Patricia Barker for her technical as-

sistance and Ann Signore for her help in preparing thismanuscript.


This work was supported by the Whitehall Foundationand the John M. Olin Foundation.

LITERATURE CITED

1. Alton, G. G., and L. M. Jones. 1967. Laboratory tech-niques in brucellosis. World Health Organization,Geneva.

2. Bauer, A. W., W. M. M. Kirby, J. C. Sherris, and M.Turck. 1966. Antibiotic susceptibility testing by astandardized single disk method. Am. J. Clin. Pa-thol. 45:493-496.

3. Bavarnick, M. R., J. C. Miller, and J. C. Snyder. 1950.The influence of certain salts, amino acids, sugarsand proteins on the stability of Rickettsiae. J. Bacte-riol. 59:509-522.

4. Bovre, K., and L. 0. Froholm. 1972. Variation of colonymorphology reflecting fimbriation in Moraxella bovisand two reference strains ofM. nonliquefaciens. ActaPathol. Microbiol. Scand. Sect. B. 80:629-640.

5. Brinton, C. C. 1965. The structure, function, synthesisand genetic control of bacterial pili and a molecularmodel for DNA and RNA transport in gram-negativebacteria. Trans. N. Y. Acad. Sci. 27:1003-1054.

6. Carmichael, L. E. 1968. Canine brucellosis: isolation,diagnosis and transmission. Proc. U. S. LivestockSan. Assoc. 71:517-527.

7. Duguid, J. P., I. W. Smith, G. Dempster, and P. N.Edmunds. 1955. Non-flagellar filamentous append-ages ("fimbriae") and hemagglutinating activity inBacterium coli. J. Pathol. Bacteriol. 70:335-348.

8. Ewing, W. H., and B. R. Davis. 1970. Media and testsfor differentiation of enterobacteriaceae. Health Ser-vices and Mental Health Administration, Centerfor Disease Control, Atlanta, Ga.

9. Filion, P. R. 1967. Infection respiratoire du dindonneaucausee par un microbe apparente au Bordetella bron-chiseptica. Can. J. Comp. Med. Vet. Sci. 31:129-134.

10. Fisk, S. K., and 0. A. Soave. 1973. Bordetella bronchi-septica in laboratory cats from central California.Lab. Anim. Sci. 23:33-35.

11. Graves, I. L. 1970. Agglutinating antibodies for Borde-tella bronchiseptica in sera before, during and afteran epizootic of pneumonia in caged monkeys. Lab.Anim. Care 20:246-250.

12. Griffitts, J. J. 1948. Hemagglutination by bacterial sus-pensions with special reference to Shigella alkales-cens. Proc. Soc. Exp. Biol. Med. 67:358-362.

13. Hedges, R. W., A. E. Jacob, and J. T. Smith. 1974.Properties of an R factor from Bordetella bronchisep-tica. J. Gen. Microbiol. 84:199-204.

14. Kang, B. K., K. Koshimizu, and M. Ogata. 1970. Stud-ies on the etiology of infectious atrophic rhinitis ofswine. II. Agglutination test on Bordetella bronchi-septica infection. Jpn. J. Vet. Sci. 32:295-306.

15. Keogh, E. V., E. A. North, and M. F. Warburton. 1947.Hemagglutinins of the hemophilus group. Nature(London) 160:63.

16. Koransky, J. R., R. W. Scales, and S. J. Kraus. 1975.Bacterial hemagglutination by Neisseria gonor-rhoeae. Infect. Immun. 12:495-498.

17. Lacey, B. W. 1960. Antigenic modulation ofBordetellapertussis. J. Hyg. 58:57-93.

18. Lautrop, H., and B. W. Lacey. 1960. Laboratory diag-nosis of whooping cough or bordetella infections.Bull. W.H.O. 23:15-35.

19. McGowan, J. P. 1911. Some observations on a labora-tory epidemic, principally among dogs and cats, inwhich the animals affected present the symptoms ofthe disease called "distemper." J. Pathol. 15:372-426.

20. Matsen, J. M., and A. L. Barry. 1974. Diffusion testprocedures, p. 418-427. In E. H. Lennette, E. H.Spaulding, and J. P. Truant (ed.), Manual of clinical


.asm.org/

Dow

nloaded from

http://jcm.asm.org/


microbiology. The Williams and Wilkins Co., Balti-more.

21. Monteverde, J. J., H. J. Monteverde, and E. J. Chialvo.1972. Bordetella bronchiseptica en neumonia de Po-trillos. Rev. Med. Vet. (Buenos Aires) 53:3-8.

22. Morse, J. H., and S. I. Morse. 1970. Studies on theultrastructure ofBordetella pertussis. I. Morphology,origin, and biological activity of structures present inthe extracellular fluid and liquid cultures of Borde-tella pertussis. J. Exp. Med. 131:1342-1357.

23. Nakase, Y. 1957. Studies on Hemophilus bronchisepti-cus. I. The antigenic structures ofH. bronchisepticusfrom guinea pig. Kitasato Arch. Exp. Med. 30:57-72.

24. Nakase, Y. 1957. Studies on Hemophilus bronchisepti-cus. II. Phase variation ofH. bronchisepticus. Kitas-ato Arch. Exp. Med. 30:73-78.

25. Neter, E. 1956. Bacterial hemagglutination and hemol-ysis. Bacteriol. Rev. 20:166-188.

26. Richter, G. W., and Y. Kress. 1967. Electron micros-copy of a strain ofBrodetella bronchiseptica. J. Bacte-riol. 94:1216-1224.

27. Ross, R. F. 1964. Methods of identification of certainetiologic agents of atrophic rhinitis. Proc. 67th Annu.Meet. U. S. Anim. Health Assoc., p. 512-515.

28. Ross, R. F., W. P. Switzer, and J. R. Duncan. 1967.Comparison of pathogenicity of various isolates ofB.bronchiseptica in young pigs. Can. J. Comp. Med.Vet. Sci. 31:53-57.

29. Sandhu, T. S., F. H. White, and C. F. Simpson. 1974.Association of pili with rough colony type of Morax-ella bovis. Am. J. Vet. Res. 35:437-439.

30. Smith, T. 1913. Some bacteriological and environmen-tal factors in the pneumonias of lower animals withspecial reference to the guinea pig. J. Med. Res.24:291-324.

31. Snyder, S. B., S. K. Fisk, J. G. Fox, and 0. A. Soave.1973. Respiratory tract disease associated withBorde-tella bronchiseptica infection in cats. J. Am. Vet.Med. Assoc. 163:293-294.

32. Swanson, J., S. J. Kraus, and E. C. Gotschlich. 1971.Studies on gonococcus infection. I. Pili and zones ofadhesion: their relation to gonococcal growth pat-terns. J. Exp. Med. 134:886-906.

33. Switzer, W. P., C. J. Mare, and E. 0. Hubbard. 1966.Incidence of Bordetella bronchiseptica in wildlife andman in Iowa. Am. J. Vet. Res. 27:1134-1136.

34. Terakado, N., and S. Mitsuhashi. 1974. Properties of Rfactors from Bordetella bronchiseptica. Antimicrob.Agents Chemother. 6:836-840.

35. Tweedy, J. M., R. W. A. Park, and W. Hodgkiss. 1968.Evidence for the presence of fimbriae (pili) on vibriospecies. J. Gen. Microbiol. 51:235-244.

36. Vera, H. D., and M. Dumoff, 1974. Culture media, p.879-929. In E. H. Lennette, E. H. Spaulding, and J.P. Truant (ed.), Manual of clinical microbiology. TheWilliams and Wilkins Co., Baltimore.

37. Winsser, J. 1960. A study of Bordetella bronchiseptica.Proc. Anim. Care Panel 10:87-104.

38. Wistreich, G. A., and R. F. Baker. 1971. The presenceof fimbriae (pili) in three species of Neisseria. J. Gen.Microbiol. 65:167-173.

J. CLIN. MICROBIOL.


.asm.org/

Dow

nloaded from

http://jcm.asm.org/

bacteriological variation among bordetella bronchiseptica ... · b. bronchiseptica b....

Documents