New, extended biotyping scheme for Campylobacter jejuni, Campylobacter coli, and "Campylobacter laridis"

0095-1137/84/100636-05$02.00/0

Copyright© 1984, American Society for Microbiology

New,

Extended

Biotyping

Scheme for

Campylobacterjejuni,

Campylobacter coli, and "Campylobacter

laridis"

HERMY LIOR

National Reference ServiceforCampylobacters, Division ofEntericBacteriology, Laboratory CentreforDisease Control, Ottawa, Ontario KIA OL2, Canada

Received 6 March 1984/Accepted 6 June 1984

Abiotyping scheme using improved media and methods for the detection ofhippurate hydrolysis, rapid H2S production, and DNA hydrolysis was applied to1,826culturesofCampylobacterjejuni,Campylobactercoli and "Campylobacter larndis" isolates from human and nonhuman sources. Four biotypes were identified among C. jejuni: 57.3% of the isolatesbelonged tobiotypeI;36.0%, tobiotypeII;4.0%, tobiotype III;and2.7%, to biotype IV. C. coli organisms weredifferentiated intobiotype1(67.0%oftheisolates)andbiotype11(33.0%). All"C. laridis"isolatesbelonged to biotype I. The combination of thebiotypingschemewith theserotypingof campylobacters provided additional epidemiological markers by further differentiating the serogroups by species and biotypes.

The sudden surgeininterest in campylobactersashuman

pathogens has led in recent years to therecognition oftwo

species among thermophilic campylobacters: Campylo-bacterjejuni and Campylobacter coli (9, 12). Furthermore,

the recent proposal for a new species, "Campylobacter

laridis" (2), hasfocusedonthe need for thecorrect identifi-cation of the various species. Thenecessity for differentia-tion is essential to both clinicians and epidemiologists to

better understand thepathophysiology andepidemiology of these organisms.

Recently, Harvey (3) usedtherapid hippurate hydrolysis testofHwang and Ederer (6) andwasabletoseparate most

strainsof C.fetus subsp. jejuni, by their abilitytohydrolyze hippurate, from the hippurate hydrolysis-negative strains of C.fetus,C.fecalis, and C. sputorum.

Skirrow and Benjamin (9) proposed a biotyping scheme

based on hippurate hydrolysis, a rapid H2S test in iron-containing media, and resistance to nalidixic acid for the differentiation ofcampylobacters into C. jejuni, C. coli, and athirdgroup, thenalidixic acid-resistant thermophilic cam-pylobacters. Hebert et al. (5) proposed abiotyping scheme

basedonhippurate hydrolysis, DNA hydrolysis, and growth

on charcoal-yeastextractagarby which the C. jejunigroup could be separated into eight biotypes.

Iampresentinga new,extendedbiotyping scheme which

allows the separation of the three species into biotypes and which, when used in conjuction with the serotyping scheme described earlier(8), provides additional significant markers in the study of the epidemiology of these organisms.

MATERIALS ANDMETHODS

Cultures.A total of1,826campylobacter cultures isolated from human (1,451 isolates) and nonhuman (375 strains including isolates from chickens, turkeys, bovine, swine,

water, and others) sources were submitted to the National

Reference Service for Campylobacters for identification, confirmation, and serotyping by provincial laboratories of public health, federal laboratories, hospitals, universities, and investigators from several other countries. Reference strainsforC. jejuni(NCTC 11168 and NCTC 11392), C. coli (NCTC 11353), and "C. laridis" (NCTC 11352) and most

isolates of "C. laridis" werereceived fromM. B. Skirrow.

Growth conditions. All strains were cultured on

Mueller-Hinton agar (Oxoid Ltd., London, U.K.) containing 5% sheep blood and incubatedfor24 hat43°Corfor24to48 hat

36°C,asindicatedfor eachtest, underagas mixture 5%

02-10% C02-85% N2 in anaerobic jars without catalyst. All

cultures grew at 43°C and not at 25°C, were catalase and oxidasepositive, did not produce H2S on triple sugar-iron media, were resistant to cephalothin, and did not grow aerobically.

Previously frozenorlyophilizedculturesweresubcultured atleastfivetimesonfresh Mueller-Hinton bloodagarplates and incubated as previously stated. Some cultures were

maintained forup to 4 weeks in a semisolid agar transport medium prepared from Wilkins-Chalgren anaerobe broth (Oxoid Ltd.) containing 0.5% Bacto-Agar (Difco Labora-tories, Detroit, Mich.) and 5% difibrinated sheep blood.

Cultures, 24hold,werecollected from theagarplateswitha

swab which was placed into small screw-capped bottles filled with the medium. The caps were tightened, and the bottles wereincubatedaerobicallyat43°C for24 hfollowed by storage at 4°C. All cultures were examined for purity

beforetesting,and when weak orambiguous reactionswere observed, the tests wererepeated with single colonies.

Hippurate hydrolysis test. The rapid hippurate hydrolysis test described by Hwang and Ederer (6) was used. A 1% solution of sodium hippurate (Sigma Chemical Co., St. Louis, Mo.) was prepared in sterile distilled water, dis-pensed in0.4-mlamountsintocorked-capped tubes(10 x 75 mm), and kept frozen at -20°C until needed. A 3.5% solution ofninhydrinwas freshly prepared each time before testing by dissolving 0.350 g of ninhydrin (Sigma Chemical Co.) in a tubecontaining 10ml of a 1:1 mixture of acetone-butanol. A small (1-mm diameter) loopful of a 48-h culture growthfromabloodagar plate was emulsified well in a tube of thawed 1%sodiumhippurateandincubated,withfrequent mixing,for 2 h in a 37°C water bath. After incubation, 0.2 ml ofninhydrinreagent was slowly overlaid on the sides of the tubes. Without mixing, the tubes were returned to the water bathfor another 10 min andexaminedimmediately, without shaking, forcolordevelopment.Apositive test was recorded as a deep purple color, crystal violet-like, indicating the presence ofglycine which resulted from the hydrolysis of hippurate. A pale purple color or colorless tubes were considered negative for hippurate hydrolysis. Cultures 636

on February 8, 2020 by guest

http://jcm.asm.org/

EXTENDED BIOTYPING SCHEME FOR CAMPYLOBACTER SPECIES 637

which showed intermediate blue color intensities were re-tested, using single colonies. Mixing or shaking the tubes after the addition of ninhydrin reagent may result in false-positive readingsand should be avoided. C. jejuni reference strain(NCTC 11168) wasused as the positive control and C. coli reference strain (NCTC 11353) was used as the negative control.

Rapid H2S test. The test originally described by Skirrow andBenjamin (9)was modified as follows. A semisolid agar base medium was prepared which contained the following: brucellabroth (Albimi; GIBCO Canada, Burlington,

Ontar-io), 2.9 g; Na2HPO4 (anhydrous), 0.118 g; KH2PO4

(anhy-drous),0.023g,agar L28(Oxoid),0.2 g;dissolved in 97 ml of

distilledwater. Thebase mediumwas autoclaved for 15 min at 15 lbs and cooled to 48°C. The following were prepared

separately at 10% aqueous solutions, filter sterilized, and

mixed aseptically in the following sequence: 1 ml of the ferrous sulfate * 7H20 solution was added to 1 ml of the sodium metabisulfite solution, mixedwell, and then added to 1mlof the sodiumpyruvate solution, mixed again, and then added to the cooled base medium. The final pH adjusted with 1 NNaOH should beabout 7.3. The complete medium was mixed well, distributed in 3- to 4-ml amounts in sterile

screw-cappedtubes (13 by 100 mm), and stored at 4°C. The

mediumwas prepared freshly every 2 weeks.

Fortesting,only 24-h-old cultureswere used. A large

ball-like inoculum removed from the plate to fill a large 5-mm-diameterloop wasgently suspended, without mixing, in the upper third of the medium. The inoculated tubes were incubated ina37°Cwaterbathfor2h.Ablackening reaction around the bacterialmassrepresented apositive test,

where-as negative reactions show no blackening. C.jejuni NCTC

11392 (H2S positive; biotype 2, Skirrow and Benjamin [9]) and C.jejunireferencestrain C6 (serogroup LIO6) were used

as positive controls.

DNA hydrolysistest. A 100-ml portion ofDNase test agar medium(Difco Laboratories)wasprepared accordingto the manufacturer's instructions and supplemented with1.35 ml ofa0.5%solution ofmethylgreen(certifiedstain catalogno.

M-295; Fisher Scientific Ltd., Ottawa, Ontario, Canada)

which had been extracted repeatedly with chloroform until the chloroform remained colorless (10). The mixture was

autoclaved, cooled, anddistributed intopetri plates(90 mm)

eachcontaining25 mlof medium. Theplateswerewelldried

ina 36or43°Cincubator with the lids partially removed.

A large loopful of 24- or48-h-old culture grown at 36°C

wasusedtoinoculateheavilyacircularareaof about1 cmin diameteron awell-driedplate. Six cultureswereinoculated

onto onepetri plate.Theplateswereincubatedat36°C inthe gasmixture recommendedforcampylobactersandexamined

daily for 3 to 5 days. An area ofgrowth surrounded by a

clear, colorless zone in thegreen-blue agarwasconsidered

positive for DNA hydrolysis (10). No change or a narrow,

hazy zone around the bacterial growth was considered a

negative test. Some strains which displayed intermediate

colorless, clearzones on repeat testingwereconsidered as

weakpositive. A strain ofStaphylococcusaureus was used

initially as a positive control. Subsequently the C. jejuni

reference strainC2074(serogroupLIO36)wasselectedand

used as thepositive control.

Nalidixic acid resistance.Susceptibilitytonalidixic acid(30

jxg)

platesincubatedfor 48 hat 36°C. Absenceofaclearzoneof

inhibition (-10 mm) wasrecorded asresistance.

Growth in the presence of TMAO. A 100-mg amount of

trimethylamine N-oxide

(TMAO)

Chemi-cal Co.) was added to 100 ml of semisolid yeast extract-nutrient broth agar medium as described by Benjamin et al. (2). Themedium was distributed in 4-ml amounts into screw-capped tubes (13 x 100mm) and autoclaved for 15min at 15 lbs.

Cultures, 24 h old, were inoculated by stabbing (two to three times) the semisolid agar to about 1 cm below the surface of the medium. The tubes were incubated anaerobi-cally and examined periodically for up to 7 days for growth throughout the medium in addition to growth just below the surface. "C. laridis" reference strain (NCTC 11352) was used as the positive control and C. jejuni reference strain (NCTC 11168) was used as the negative control. Fresh media were prepared every 2 weeks.

RESULTS

Hippurate hydrolysis test.Of the 1,826 strains of Campylo-bacter tested, 1,407 (77.1%) isolates, including 1,195 strains from human sources and 212 cultures from nonhuman sources and also the C. jejuni reference strain (NCTC 11168),hydrolyzed hippuric acid and were therefore consid-ered as C.jejuni (not tabulated). Over 22% (22.9%) or 421 cultures, including 256 strains from human sources, 163 strainsfromnonhuman sources, and the referencestrains for C. coli (NCTC 11353) and "C. laridis" (NCTC 11352), did nothydrolyze hippurate andwere considered as belonging to the C. coli or "C. laridis" group of organisms (not tabulat-ed). These strainswere further investigated for nalidixic acid resistance and for anaerobicgrowth in TMAO to differenti-ate the "C. laridis" and C. coli.

Rapid H2S test. Of the 1,826 strains investigated, 123 (6.7%) were positive in the rapid H2S test (Table 1). Of these, 25 H2S-positive cultures (including 8 isolated from humans and 17 from nonhumans which were also hippurate hydrolysis negative and resembled the reaction obtained with the "C. laridis" reference strain) were temporarily assigned to the "C. laridis" group and further investigated for nalidixic acid resistance and anaerobic growth in the presence of TMAO as described by Benjamin et al. (2).

Reproducible results in the rapidH2S test were obtained withthe buffered medium only when 24-h-old cultures were used.Cultures which were .48 h old showed false-negative orerratic reactions.

TABLE 1. Rapid H2S test results of Campylobacter isolates

No.positive/

Isolates no.tested

(% positive)

Hippuratepositive (C. jejuni)

Human...80/1,195 (6.7)

Nonhuman...18/212(8.5)

Hippurate negative

Human... 8/256(3.1)a Nonhuman...17/163(10.4)a

Reference strains

C.jejuni(NCTC 11168)... ... 0/1

C.jejuni(NCTC 11392)b... ... 1/1

C.coli(NCTC 11353)... 0/1

"C.laridis" (NCTC 11352)... 1/1

C.jejuni(L1O6) ... ... 1/1

a"C.laridis".

b C.jejuniH2Spositive (biotype 2,Skirrow andBenjamin).

VOL.20, 1984

on February 8, 2020 by guest

http://jcm.asm.org/

Nalidixic acid resistanceand anaerobic growth in the pres-enceof TMAO. Of the 1,407 C.jejuni isolates, 52 (3.7%)were foundtoberesistant tonalidixic acid (Table 2). Amongthe 419 cultures which were hippurate hydrolysis negative, 33 (7.9%) were resistant to this antibiotic. These 33 cultures represented 25 isolates and the reference strain for "C. laridis" whichwerealsopositive in the rapid H2Stestand 7 other isolates (including the reference strain for C. coli) which were negative in the rapid H2Stest. The 33 cultures

werefurthertested for anaerobic growth in the presence of TMAO.Only the 26cultures whichwerealsopositiveinthe rapid H2S test wereable togrow in this medium and were considered to be "C. laridis" according to the criteria suggested by Benjamin et al. (2). Reference strains for C. jejuni and C. coli as well as five C.jejuni and two C. coli strains whichwerenalidixic acidsusceptible didnotgrowin thismedium. Theremaining seven(1.8%) cultures negative for growth in the presence of TMAO and negative in the rapid H2S testwere considered presumptively C. coli nali-dixicacid resistant.

DNA hydrolysis test. Of the 1,826 cultures examined for DNA hydrolysis, 37.4% (683 strains) were positive in the DNA hydrolysis test. The results obtained with the three Campylobacterspecies areshown in Table 3.

Biotyping scheme. The results obtained by hippurate hy-drolysis, rapid H2S, and DNAhydrolysistestwereusedina

biotyping scheme which allows the recognition of the three species, C. jejuni, C. coli, and "C. laridis," and also dif-ferentiates eachspecies into biotypes. Using Roman numer-als, four biotypes (I, II, III, and IV) were identified among C.jejuni; two biotypes (I and II), among C. coli; and two

possible biotypes (I and II), among "C. laridis" (Table 4). Analysis of the 1,826 isolates tested by this biotyping scheme shows that 77.1%(1,407) of all isolates belonged to

C. jejuni, 21.5% (393) belonged to C. coli, and 1.4% (26) belongedto "C. laridis" (Table 5).

Biotyping andserotyping of isolates. Strains of C. jejuni, C. coli, and "C. laridis" have been serotyped bya serotyping

scheme based on slide agglutination of live bacteria with absorbed antisera to heat-labile factors, as previously

de-scribed (8). These strains have also been typed with this biotyping schemetoassessthediscriminativeability of each species by thecombined bioserotyping schemes. The results obtained with the 12 mostcommon serogroups arelisted in

Table6. Most human isolates of C. jejuni in thecommonest serogroup 4were found tobelong to biotype I (82 isolates)

andbiotype II (75 isolates). C. jejuni isolated from nonhu-man sourcesinthisserogroupweresubdivided into biotype I

(16 isolates) and biotype II(12 isolates). Similar resultswere

obtained with isolates from each of the other common serogroups; four biotypes were identified among C. jejuni serogroups 7, 8, 11, and 36 and three biotypeswere

identi-TABLE 2. Results of nalidixic acid resistance and anaerobic

growth in TMAO with campylobacter isolates

No.positive/no. tested(% positive)

Isolates Nalidixic acid Anaerobic resistance (30,ug) growth in0.1%

C.jejuni 52/1,407 (3.7) 0/57"

C. coli 7/393 (1.8) 0/9

"C.laridis" 26/26 (100) 26/26 (100)

C.jejuni(NCTC 11168) 0/1 0/1

aInclUdes five C. jejuni nalidixic acid susceptible.

bIncludestwoC. coli nalidixic acid susceptible.

TABLE 3. DNAhydrolysis results of C. jejuni,C.coli,andC. laridis

No.positive/no.

Isolates tested(%

positive)

C.jejuni

Human...463/1,195(38.7)

Nonhuman ... 96/212(45.3)

C. coli

Human ... 82/248(33.1)

Nonhuman...,... 42/145 (29.0)

"C. laridis"

Human... 0/8

Nonhuman... 0/18

fied among serogroups 7, 2, and 4, whereas the remaining serogroups were subdivided into two biotypes each. All isolates of serogroup 6 belonged to biotypes III and IV (biotype 2of Skirrow andBenjamin [9]).

C. coli isolates belonging to these common serogroups were further differentiated into biotypes I and II. Most human isolates of this species in serogroup 8 belong to biotype II, whereasin serogroup 20, which comprisesonly isolates of C. coli, most human and nonhuman isolates belong tobiotype I (Table 6). No "C. laridis" strains were found to belongtothese common serogroups.

DISCUSSION

The new, extended biotyping scheme I am proposing is based on therapid hippurate hydrolysis test of Hwang and Ederer (6), the rapid H2S production in a buffered iron-bisulfate-pyruvate semisolidmedium,and the DNA

hydroly-sis testina modifiedDNase test agarmedium.This biotyp-ing scheme will not only allow the recognition of three species among thermophilic campylobacters, but will also subdividecommon C.jejuni into fourbiotypes, C. coli into two biotypes, and the "C. laridis" into two possible bio-types.

Skirrow andBenjamin (9)werefirsttopropose abiotyping scheme for enteropathogenic campylobacters by using the hippurate hydrolysistest,therapid H2Stestin iron-contain-ing media, and resistance to nalidixic acid by which three groupsweredifferentiated: C.jejuni, C. coli, and the nalidix-icacid-resistant thermophilic campylobacters for whichthe name"C.

laridis"

(2). Hdbertet al. (5) recently proposed abiotyping scheme

for C.jejuni based on hippurate hydrolysis, DNA hydroly-sis, and growth on charcoal-yeast extract agar. C. jejuni isolatesweredifferentiated into eight biotypes, withbiotypes 1 to 4 comprising hippurate hydrolysis-positive strains and biotypes 5 to 8 representing hippurate hydrolysis-negative isolateswhichtheauthorsreferto as C. jejuni. Theinclusion ofthe hippurate-negative strains into the C.jejuni species

TABLE 4. Biotyping scheme for C. jejuni, C. coli,

and"C.laridis"

C.jejuni C.coli "C.laridis"

Test

Ia II III IV I II I II

Hippurate hydrolysis + + + + - - -

-Rapid H2S test - - + + - - + +

DNAhydrolysis - + - + - + - +

aBiotype.

on February 8, 2020 by guest

http://jcm.asm.org/

EXTENDED BIOTYPING SCHEME FOR CAMPYLOBACTER SPECIES 639

TABLE 5. Resultsofbiotypingof campylobacter isolates No. (%) ofisolatesof:

C.jejuni C. coli ".C.laridis"

Source (n= 1,407;77.1%) (n =393;21.5%) (n=26;1.4%)

I II III IV I II I II

Human 685 (57.3) 430 (36.0) 47 (4.0) 33 (2.7) 166 (67.0) 82 (33.0) 8 (100) Nonhuman 109 (51.4) 85 (40.1) 7 (3.3) 11 (5.2) 103 (71.0) 42 (29.0) 18 (100) Total 794 (56.5) 515 (36.6) 54 (3.8) 44 (3.1) 269 (68.4) 124 (31.6) 26 (100)

cannot bejustified even more so as the hippurate-negative ed by the use of ninhydrinto produce a deep crystal violet-strains included known reference violet-strains ofC.coli (CIP 7080 like color. Strains showing doubtful, intermediate blue

and Skirrow's4620). reactionsmay be retested by the gas-liquid chromatography

Therapid hippurate hydrolysistestis basedontheability method described by Kodaka et al. (7).

ofsome bacteriato showhippuricase activity which hydro- Thevalidity ofthehippurate hydrolysis test proposed by lyzes hippurate to benzoic acid and glycine, which is detect- Skirrow and Benjamin (9) for the differentiation of C. jejuni

TABLE 6. Distribution of commonCampylobacter serogroups by species, biotype, and source

C.jejuni C. coli

Ia II III IV I II

Serogroup

Human _humanNon- Human _humanNon- Human _humanNon- Human _humanNon- Human _humanNon- Human _human

Non-1 35 3 bovine 51 7 bovine 2 4 2 turkey 2

2N/Sb 1turkey 1N/S

2 19 4turkey 20 6bovine 1 N/S 1 1 river 2 1 river

2canine 1chicken water water

3bovine 1 turkey 2turkey

4 82 1canine 75 1mink 1 1 1 sea 1

12bovine 9bovine water

2chicken 2N/S

1 N/S

5 13 9bovine 10 2bovine 3 N/S

1N/S

6 23 2 river 12 1 river

water water

7 96 1 milk 28 11bovine 2bovine 2 1 bovine 1

32bovine 1N/S

1canine 1ovine

8 6 3bovine 17 15bovine 1 1 canine 4 2chicken 28 2turkey

9turkey 1chicken 1 turkey 1N/S

1chicken

9 13 1chicken 9 3bovine 1 5river 6 river

1 ovine 1chicken water water

1 N/S

11 10 1 chicken 8 3chicken 1 1goat

X~

17 13 2 canine 4 1

20 7 4turkey 2 1 swine

2 chicken 2 N/S 1 N/S

36 23 3N/S 25 2N/S 1bovine 1 1 3 N/S 1 2 N/S

1 swine

aBiotype.

b N/S,Exact sourcenotknown.

VOL. 20, 1984

on February 8, 2020 by guest

http://jcm.asm.org/

and C. coli has been confirmed by the DNA homology studies of Belland and Trust (1), Harvey and Greenwood (4), and Ursing et al. (11).

The enzymes responsible for the blackening caused by

some campylobacter strains in therapid H2S testare notyet known. Reproducibility of results of the Skirrowand Benja-min medium (2) was poor and erratic, depending on the age of the medium or the incubation time of the cultures. Reproducibility of results was obtainedonly when the medi-um was buffered and 24-h-old cultures were used.Consistent results were obtained with previously frozen or lyophilized strains when these had been subcultured at least five times before being tested. All C. coli isolates were negative inthis test; 6.7% of the human and 8.5% of the nonhuman isolates of C. jejuni and all "C. laridis" cultures examined were positive for rapid H2S production.

In the DNA hydrolysis test, organisms which showDNase activity will hydrolyze DNA resulting in the release of methyl green from the complex with DNA to produce clear, colorless areas around the bacterial growth (10). By increas-ing the concentration of the methyl green in the medium from0.005% (10) to 0.006%, a better contrast was provided, thus reducing the questionable narrow hazy zones which were difficult to interpret. Best results were obtained when the readings were made by holding the plates against a white background and indirect light.

The overall results obtained by this biotyping scheme show that, among C. jejuni, biotype I is most common among human and nonhuman isolates, with 57.3 and 51.4%, respectively (Table 5), followed by biotype II with 36.0 and 40.1%, respectively. The differentiation among C. coli was not possible by the Skirrow and Benjamin biotyping scheme, but by the use of this scheme 67.0%of the human isolates of C. coli and 71.0% of the nonhuman isolates were found to belong to biotype I and the remaining 33.0% of the human

and 29.0%of the nonhuman isolates were found to belong to

biotypeII (Table 5).

The routine screening of campylobacter isolates with nalidixic acid disks showed that resistances were encoun-tered in 3.7% of the C. jejuni isolates tested and in 1.8% of the hippurate-negative, H2S-negative, TMAO-negative strains which I have identified presumptively as C. coli. Occasional strains of C. jejuni have been found to be resistant to nalidixic acid (2, 11), but I am not aware of any report regarding nalidixic acid resistance among hippurate-negative, rapid H2S-hippurate-negative, TMAO-negative strains. Some laboratories may use the susceptibility to nalidixic acid as one of the criteria for the identification of C. jejuni and C. coli, and it is possible that strains resistant to this antibiotic may be erroneously discarded. Further studies are in pro-gress to determine the genetic basis of the variability of susceptibility to nalidixic acid observed with some strains of C. jejuni and C. coli.

The differentiation of the thermophilic campylobacters into species and into biotypes can provide valuable markers for epidemiological investigations, but by itself the informa-tion provided is rather general in nature. However, when used in conjunction with serotyping results as an extension of the serotyping scheme developed by us (8), the discrimi-nating ability of the system is greatly enhanced by the further differentiation of the common serogroups into biotypes (Table 6). The epidemiological significance of the combined schemes becomes more evident in the interpretation of data from common serogroups which comprise large numbers of

isolates. Serogroup 4, the most common serogroup, repre-sents 158 isolates fromhuman and 28 isolates from

nonhu-man sources such as dogs, cows, and chickens. As human infections may be acquired from these sources, determina-tion of the serogroups alone may not always provide suffi-cient evidence for linking the humaninfection to ananimal

or food source. A more precise correlation can be estab-lished with additional markers in those cases where the strains under investigation are found to belong not only to the same serogroup but also to the same biotype. Excellent correlation of biotyping results has been obtained with

epidemiologically linked isolates from several large

out-breaks (H. Lior, manuscript in

preparation).

The discriminatory ability ofthe biotyping scheme,

espe-cially when used as an extension ofthe serotyping

scheme,

will provide additional epidemiological information in the study of outbreaks and in the study of the ecology, the

transmissibility, and the pathogenicity of Campylobacter

species in humans and other animals.

ACKNOWLEDGMENTS

I thank D. L. Woodward and M.J. Larose for excellent assist-ance, W. P. Winter for the medium preparation, F. Ashton for critical comments, and D. Cathcart for typing. I am indebted to

M. B. Skirrow for the reference strains and "C. laridis" isolates.

LITERATURE CITED

1. Belland, R.J. and T. J. Trust. 1982. Deoxyribonucleic acid sequence relatedness between thermophilic members of the genus Campylobacter. J. Gen. Microbiol. 128:2515-2522. 2. Benjamin, J., S. Leaper,R. J. Owen,and M. B.Skirrow. 1983.

Description of 'Campylobacter laridis' anew species compris-ing the nalidixic acid resistant thermophilic campylobacter (NARTC) group. Curr. Microbiol. 8:231-238.

3. Harvey, S. M. 1980 Hippurate hydrolysis by Campylobacter fetus. J. Clin. Microbiol. 11:435-437.

4. Harvey,S. M., and J. R. Greenwood. 1983. Relationshipamong

catalase-positive campylobacters determined by deoxyribonu-cleic acid-deoxyribonucleic acid hybridization. Int. J. Syst. Bacteriol. 33:275-284.

5. Hebert,G.A., D. G. Hollis, R. E. Weaver, M. A. Lambert, M.J. Blaser, andC. W. Moss. 1982. Thirty years ofcampylobacters: biochemical characteristics and a biotyping proposal for Cam-pylobacterjejuni. J. Clin. Microbiol. 15:1065-1073.

6. Hwang,M.N.,andG. M.Ederer. 1975. Rapid hippurate hydro-lysis method for presumptive identification of group B strepto-cocci. J. Clin. Microbiol. 1:114-115.

7. Kodaka, H., G. L.Lombard, and V. R. Dowell, Jr. 1982. Gas-liquid chromatography technique for detection of hippurate hydrolysis and conversion of fumarate to succinateby microor-ganisms. J. Clin. Microbiol. 16:962-964.

8. Lior, H., D.L. Woodward, J. A. Edgar, L. J. Laroche, and P.

Gill. 1982. Serotyping of Campylobacterjejuni by slide

aggluti-nation basedonheat-labile antigenicfactors. J. Clin. Microbiol. 15:761-768.

9. Skirrow, M. B., and J. Benjamin. 1980. Differentiation of

en-teropathogenic campylobacter. J. Clin. Pathol. 33:1122.

10. Smith, P. B., G. A. Hancock, and D. L. Rhoden. 1969.

Im-proved medium for detectingdeoxyribonuclease-producing

bac-teria. Appl. Microbiol. 18:991-993.

11. Ursing, J., M. Walder, and K. Sandstedt. 1983. Base composi-tion and sequence homology ofdeoxyribonucleic acid of ther-motolerant Campylobacter from human and animal sources. Curr. Microbiol. 8:307-310.

12. Veron, M., and R. Chatelain. 1973. Taxonomic study of the

genusCampylobacter, Sebald and Veron, and designation of the

neotype strain for type species, Campylobacterfetus (Smith and Taylor) Sebald and Veron.Int. J. Syst. Bacteriol. 23:124-134.

on February 8, 2020 by guest

http://jcm.asm.org/