Evaluation of the RapID ANA II and API ZYM systems for identification of Actinomyces species from clinical specimens

(1)

0095-1137/92/123112-05$02.00/0

Evaluation of

the

RapID ANA

II

and

API

ZYM

Systems

for

Identification

of Actinomyces

Species

from

Clinical

Specimens

MINNA A.BRANDER* ANDHANNELE R. JOUSIMIES-SOMER2t

Department inLappeenranta, National Public HealthInstitute, Kahilanniementie2, 53130Lappeenranta,

and Anaerobe

Reference Laboratory, National

Public Health

Institute,

Helsinki,

2 Finland

Received18May1992/Accepted 31 August 1992

Classification and identification offermentative actinomycetesarelabor-intensive andproblematic. In this study, we evaluated the

applicability

and reliability of the RapID ANA II system (Innovative Diagnostic

Systems, Inc., Atlanta,Ga.) and the discriminatory value of the API ZYM system (Societes Analytab Products Inc., LaBalmeLesGrottes, France) in theidentification ofActinomyces-like bacteria by using conventional methods as a reference. Eighty-five strains, including 71 isolates from mixed anaerobic infections and 14 reference strains, were tested. The RapID ANA IIsystem

correctly

identified all Actinomycesodontolyticus strainsand 65% of Acfinomyces israelii strains.AllArcanobacterium haemolyticum strains were misidentified asActinomyces pyogenes. The most common isolates in the study were Actinomyces meyeri-like organisms, 84% ofwhich, however,wereaerotolerant. Theidentification of these aerotolerant strains thus remains unresolved and warrantsfurther studies. New characteristics and changes to theconventional API ZYM enzyme profiles are suggested. TheAPIZYM enzyme profiles of A.

odontolyticus

and A. israelii wereverysimilar, the

only

discriminating enzyme being at-fucosidase. In differentiation between A. pyogenes and Arcanobacterium

haemolyticum, theproduction of

13-glucuronidase

by theformer and theproduction of acid phosphatase by the latter aresuggested as newhelpful characteristics for useinclinical laboratories.Insummary, theRapID ANA II and APIZYM systems can be used as rapid preliminary methods in theidentification ofActinomyces species but accurateidentification requires supplementary conventionaltestsandgas-liquid chromatography. With the exception of classical actinomycosis due to

Actinomycesisraelii,Actinomyces species areisolated usu-ally from bacterial infections involving mixedfloras. Thus,

the clinical significance of individual species is still

con-troversial (6-8, 15, 17). Conventional testing based on

fer-mentation and biochemical reactions together with the

de-termination of metabolic end products by gas-liquid

chromatography isconsidered themost accurateand reliable

approachforthe identificationofActinomycesspecies (1, 9,

18). However, because these procedures are both

labor-intensiveandtime-consumingandrequirespecial

instrumen-tation, theyarebeyondthecapabilities of manylaboratories

and the isolation and identification of these bacteria have remainedaspecialityofafewlaboratories.

Recently, the trend inclinical anaerobic bacteriologyhas

been towards the useofcommercially available

identifica-tion kits, including those detecting constitutive preformed

enzymes. Theperformancesof thesekitsarenotdependent onactivegrowthandrequire only4 hofaerobic incubation.

Thedifferentialchartsofthesesystems alsoinclude

Actino-mycesspecies.

The purpose ofthisstudywas toevaluate the applicabil-ity, reliability, and accuracy of the RapID ANA II system and thediscriminatory value of the API ZYM system in the

identification ofActinomyces species isolated from clinical

specimens by usingconventional methods as areference.

* Correspondingauthor.

tPresentaddress: Microbial DiseaseResearchLaboratory, 691/ 151J,VAWadsworthMedicalCenter, Los Angeles, CA 90073.

MATERIALSANDMETHODS

Bacterial strains. A total of 85 Actinomyces and

Actino-myces-like bacterial strainswereexamined. These included

14 reference strains (13 from the American Type Culture

Collection [ATCC] and 1 from the National Collection of

T7ype

Cultures, Central Public Health Laboratory,

Colin-dale, London, England): A. israelii serotypes 1 (ATCC

10049; NCTC 10236) and 2 (Actinomyces gerencseriae;

ATCC 23860), Actinomyces odontolyticus (ATCC 17929),

Actinomycesnaeslundii (ATCC 12104),Actinomyces visco-susserotype 1 (ATCC 15987),Actinomycesmeyeri (ATCC

35568),Actinomyces georgiae (ATCC 49285),Actinomyces

humiferus

(ATCC 25174),Actinomycesbovis(ATCC 13683),

Actinomyces hordeovulneris (ATCC 35275), Actinomyces sp. serotype WVA 963 (ATCC 49338),Actinomyces pyo-genes (ATCC 19411), andArcanobacterium

haemolyticum

(ATCC 9345).Here,Actinomyces-like bacteriaweredefined

as gram-positive, non-spore-forming rods, most of which

werebranchingand catalasenegative and whichingas-liquid

chromatographyproducedsuccinicacidandwiththeRapID

ANA II system were identified as Actinomyces species.

All organisms except the reference strains were fresh

clinical isolatesobtainedfrompatientspecimensin 1990 and cultured and isolated at the Department of Lappeenranta,

National Public Health Institute, Lappeenranta, Finland,

mainlyaspartof mixed anaerobic floras. Afterisolation,the

strainswereexaminedbyconventionalVirginia Polytechnic

Institute reference methods (9) and the RapID ANA II system. The strains were kept at -70°C in skim milk,

thawed, and subcultured before being tested by the API

ZYMsystem. The conventional methods included aerotol-erancetestingand determinationofGramreaction, cellular

morphology,and colonial characteristicsaswell as

patterns

3112

on April 12, 2020 by guest

http://jcm.asm.org/

(2)

TABLE 1. API ZYM reactions of the study strains identified by the RapID ANA II system as A. meyeri, A.odontolyticus,and A. israelii

No.of No. of strains positive by API ZYM in

microtubea:

strains

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

A. meyerib 31 2 4 1 0 28 0 0 0 0 2 0 0 2 1 19 0 2 0 0

A. odontolyticusc 18 0 12 8 0 18 4 0 0 0 6 1 5 8 0 15 10 5 3 15

A. israeliid 14 0 14 8 0 14 5 0 0 0 1 0 3 6 1 5 3 0 0 0

a Reactionmicrotubes assayed for thefollowingenzymes: 2,alkalinephosphatase; 3, esterase; 4,esterase-lipase; 5, lipase; 6, leucine arylamidase; 7, valine arylamidase; 8, cystine arylamidase; 9, trypsin; 10, chymotrypsin; 11, acid phosphatase; 12, naphthol-AS-BI-phosphohydrolase; 13, a-galactosidase; 14,

3-galactosidase; 15, 3-glucuronidase; 16, a-glucosidase; 17, 3-glucosidase; 18, N-acetyl-,B-glucosaminidase; 19, a-mannosidase; 20,a-fucosidase.

Onlytwostrainswereobligate anaerobes.

CAllstrainskeyed outidenticallyby the conventional methods.

dOnlyeight strains keyed outidenticallyby the conventionalmethods.

of fermentation ofcarbohydrates andproductionofindole,

nitrate reductase, catalase, arginine dehydrolase, and

ure-ase, supplemented by the metabolic end product analysis

with gas-liquid chromatography (20). Some of the strains were also tested for gelatin liquefaction, peptonization of litmus milk, and reverse CAMP reaction.

RapIDANA Hsystem. TheRapIDANA IIsystemconsists

of 10 individual test wells (8 of which are bifunctional)

containing dehydrated reactants designed to test 18

bio-chemical reactions. Substrates included in the 10 wells of the test panel are urea,

p-nitrophenyl-3,D-disaccharide,

p-nitrophenyl-a,L-arabinoside,

o-nitrophenyl-,B-D-galacto-side,p-nitrophenyl-a,D-glucoside,

p-nitrophenyl-3,D-gluco-side,p-nitrophenyl-a,D-galactoside,

p-nitrophenyl-a,L-fuco-side,

p-nitrophenyl-N-acetyl-3,D-glucosaminide,

p-nitro-phenylphosphate, leucine-,-naphthylamide,

glycine-,-naphthylamide,

proline-13-naphthylamide,

phenylalanine-,-naphthylamide,

arginine-p-naphthylamide,

serine-3-naph-thylamide, pyrrolidonyl-,B-naphthylamide, and tryptophane.

Purecultures of theorganismswereinoculated on chocolate

agar(Oxoid).Plates wereincubatedin ananaerobic

environ-ment at35 to37°C for 48 h. The strainswere

suspended

in

RapID ANA II inoculation fluid, and the inoculum was

adjustedto aturbidity ofMcFarland standard no. 3. The trays

wereinoculated accordingtothe manufacturer'sinstructions

andincubatedaerobicallyat35to37°Cfor 4 h.After

incuba-tion,reactions in the test wells wereinterpretedandrecorded

before and after the addition of reagents. Asix-digit

micro-code was generated, and identification was determined by

usingtheRapIDANA II codecompendium(10).The

Actino-mycesspecies includedinthe codecompendiumareA.bovis,

A. israelii,A. meyeri,A. naeslundii, A. odontolyticus,A.

pyogenes, andA. viscosus. The referencestrainswere exam-ined twice.

API ZYMsystem. The API ZYMgalleryiscomposedof 20

microtubes,which contain 19 substrates in buffer. The first

microtubeserves as acontrol, and the enzymes determined

in the other tubes are listed in Table 1, footnote a. Pure cultures oforganismswere obtained and incubated as with

the RapIDANA IIsystem. The strainswere suspended in

sterile distilledwaterand adjustedtoaturbidity of McFar-land standard no. 5 or greater. The trays were inoculated

accordingtothe manufacturer's instructions and incubated

aerobicallyat35 to37°Cfor 4 h. Afterincubation, reactions

in the testcupuleswereinterpretedand recorded on areport sheetafteraddition of reagents.Acolorintensityvalue from

0to5wasassignedfor each reaction accordingtothe color

chart enclosed with the kit.Onlycolorsshowinganindex of 2or morewererecorded aspositive. The API ZYMsystem

lacks an index book forspecific bacterialidentification. The reference strains were examined twice.

RESULTS

Sixty-three

of the 71clinical isolates studiedwere

identi-fiedby the RapID ANA IIsystem as A. odontolyticus (18

isolates),A. israelii (14isolates), orA. meyeni(31 isolates).

The referencestrains for each werecorrectlyidentified;A.

gerencseriae(A.israelii serotype2)wasalso identified as A.

israelii.

All ofthe 19strainsthat wereidentifiedasA.

odontolyti-cusbyRapID ANA II also wereidentifiedasA.

odontolyti-cus by the conventional methods; all were mannitol

nega-tive, 16 were raffinose and trehalose negative, and 9 were

ribose negative. Twelve different microcodes were found,

noneof which wasoutstanding.Twomicrocodes, however,

gave theidentificationprobability belongingtoA.

odontolyti-cus,Arachnia propionica (Propionibacterium

propionicus),

Lactobacillusacidophilus,orA.israelii.Asthestudy strains

produced succinic acid and nopropionic acid, the

identifi-cationasActinomyces specieswas accepted.Furthermore,

as oneof these twostrainswasa-fucosidasepositiveand the other produced red pigment, the final identification as A.

odontolyticus

instead of A. israeliiwasaccepted.

Only11of the14strains identifiedasA. israeliibyRapID

ANAII wereidentified asA. israelii by conventional

meth-ods.(All strainswerea-fucosidaseand ureasenegative.)Ten

different microcodeswereobtained.

If obligately anaerobic growth (3) were used as a strict

criterion, only 2 of the 31 clinical strains identified as A.

meyeri

in the RapID ANA II system and the type strain would have beenidentifiedasA.meyeri alsobythe

conven-tional methods. Even these three strains (the two clinical

isolates and the type strain) grew slightly in a 5% CO2

atmosphere.Theremaining29clinical isolateswere

aerotol-erant; however, 3 ofthese grew very slowly in air. Three

different microcodes were obtained; of these, 020671 was

produced by 26 (84%) of the 31 strains. According to the

RapID ANA II system code

compendium,

the probable

identification given to microcode 020671 overlapped A.

meyeri

and

Propionibacterium

granulosum. As all the strains were catalase negative and produced succinic acid

but nopropionic acid, the identification asA. meyeri was

accepted.

The API ZYMreactions of A.

odontolyticus,

A. israelii,

and A.

meyeri

aresummarized in Table 1.

Of thestudystrains,seven wereidentified asA.pyogenes

bytheRapIDANA IIsystem. These strains wereidentified

on April 12, 2020 by guest

http://jcm.asm.org/

(3)

TABLE 2. Reactivities of referencestrains of Actinomycesspp.in the API ZYMsystem

Colorintensity of API ZYM reaction inmicrotubea:

Reference strain

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

A. meyeri - - - - +- - - - +

A. odontolyticus - - - - +. . . +

A. israelii

ATCC10049 - - - - + --- w + - w + -

-NCTC10236 - - - - + . . . w + w + -

-Serotype2, ATCC 23860 - - - - + w . . . w + + +

A. pyogenes - - - - + . . . w w w

Arcanobacteniumhaemolyticum - - - - w + . . . .w

A. naeslundii - - - - + . . . w + ww

-A. viscosus - w + - + w . . . w + + -

-A. hordeovulneris - - w - + + . . . w + +

-A. humiferus - - - - +

A. georgiae - - - - +. . . +--+

Actinomycessp. serotypeWVA963 - - - - + w w + ww

-A. bovis - - - - + w . . . + + + -

-aFormicrotubenumberingsystem, see Table 1, footnotea. -, w, and +, rangesof colorintensity of 0 to 2, 2 to 3, and 3 to 5, respectively.

by the conventional methods as

Arcanobacterium

hae-molyticum.

All strains were reverseCAMP reaction positive

(5), none liquefied gelatin, and only one strain fermented

xyloseslightly and wasabletopeptonize litmusmilk(18).In

the API ZYM system, all these strains produced at least someleucinearylamidase,

P-galactosidase,

and

N-acetyl-p-glucosaminidase, all of which are enzymesproduced byA.

pyogenes accordingtoBergey's Manualof Systematic

Bac-teriology (18). In addition, all the strains produced acid

phosphatase,which hasnotbeenreportedasbeing produced

byA.pyogenes.The API ZYM reactions of the type strain of

Arcanobacterium haemolyticum wereidentical to the

reac-tions of the seven strains; only the production of

3-galac-tosidase was very weak. The type strain of A. pyogenes

produced leucine arylamidase,

P-galactosidase,

,-glucuron-idase, and a-glucosidase and a very small amount of

N-acetyl-,B-glucosaminidase butnoacidphosphatase.

There was only one catalase-positive strain among the clinical strains tested. The RapID ANA II microcode

ob-tainedfor thisorganism,whichbythe conventionalmethods

most closely resembled A. viscosus, was not listed in the

codecompendium.Unlisted microcodeswerealso obtained

for the twocatalase-positive reference strains,A. viscosus

ATCC 15987 and A. hordeovulnenis ATCC 35275. The API ZYM reactions of these two ATCC strains as well as the

otherreference strains aresummarized in Table 2.

DISCUSSION

Thetaxonomy ofActinomyces speciesiscomplicated,and classification and identification of fermentative actino-mycetesstill present problems despite the contributions of modemtechnology. In this study, twocommercially

avail-able kits, the RapID ANA II system and the API ZYM

system, were evaluated for the identification of Actinomy-ces-like bacteria and compared with the conventional

Vir-ginia Polytechnic Institute reference methods. In earlier

evaluations of the RapID ANA II system, about60%of the

Actinomycesstrains could bereliably identified to the genus

level but not always to the species level (8). In those

evaluations,however,onlyafewspecies and strains of these

bacteriawere included and they constituted only a minor fraction of the tested strains(4,8, 12-14, 16, 19).Kilian(13) described the results for 162 strains oftheActinomycetaceae

and related taxa for 20 different enzymatic activities, of

which 10 were the same as in the API ZYM system. Our

resultsconcurwiththose of Kilian.

In the present study, A. meyeri was the most common identificationgiven bytheRapID ANA II system(31 of 71 study strains). By the conventionalmethods,all these strains

wouldhave received the same identification had 26 strains

notgrownaerobically.Ifthe criterionofnoaerobicgrowth

(3) is obligatory, the identification ofthese strains remains

unresolved. Whether these aerotolerant A. meyeri-like

strains represent variants of A. meyeri or a new species

warrantsfurther studies. Recently, three of thestrains have

been investigated byLillian V. H. Moore atVirginia

Poly-technicInstitute and StateUniversity, Blacksburg, and the strains were unlike any described species that she has

studied (1Sa). It is also noteworthy that none of the 31

clinicalA. meyeri-like strainsnortheA. meyeri typestrain

produced valine arylamidase or cystine arylamidase,

al-thoughthesearelisted inBergey'smanualascharacteristic

products of A. meyeri.

All19 strains identifiedby theRapID ANA II system asA.

odontolyticuswerealso identifiedidentically bythe

conven-tionalmethods. Theproductionof a-fucosidasewas aunique

reaction for A. odontolyticus strains, being produced byno otherspeciesinthisstudy,aresult similartothoseofearlier studies (18). Onlythree of the A.

odontolyticus

strains did

notproduceanya-fucosidase.AlthoughKilian(13) reported

thatA. viscosus serotype 1produced a-fucosidaseintheAPI ZYM system,inthepresentstudy this reactionwasnegative

for the type strain. The colonies of only 37% ofA.

odon-tolyticus

strains developed redpigment. This percentage is somewhat lowerthan thosereportedearlier(2, 11,15). Thus,

the demonstration of pigment formation might not be as

importantinthe identification ofthisActinomycesspeciesas

has beenthought. Thespecimensin the earlier studies have

beenmainlyof dentalorigin,butthiswas notthecasein the

presentstudy,as83%werefrom nonoralsources.However,

this difference inspecimensourcemaynotsufficetoexplain

thedivergentresults. Somebrownishred colorcouldbeseen

inmostof the ATCCstrains included in thisstudy,provided

the strainswere incubatedlong enough(5 to7 days). Also,

Johnsonetal. could demonstratetypicalred colonies inonly 58% of the serotype I strains and 43% of the serotype II strains of A.

odontolyticus (11).

on April 12, 2020 by guest

http://jcm.asm.org/

(4)

According to Bergey's manual, there are only three

en-zymes produced by A.

odontolyticus

in the API ZYM system: leucine arylamidase, 1-glucosidase, and a-fucosi-dase. The A. odontolyticus strains in the present study produced more enzymes, a result concordant with that of

Kilian (13). More than halfofourA.

odontolyticus

strains

producedesteraseand a-glucosidase, andatleast27% ofthe

strains produced esterase-lipase, 3-galactosidase, and

N-acetyl-fi-glucosaminidase.TwoA. meyeri-likestrainsalso

producedsomeN-acetyl-,-glucosaminidase, an enzyme

re-portedasbeingproducedonly byA. bovis and A.pyogenes

accordingtoSchaal (18). Of the reference strains, A. bovis

did not produce that enzyme at all and A. pyogenes

pro-duced onlya verysmallamountofit. The referencestrains which produced the enzyme were A. hordeovulneris and

Arcanobacterium haemolyticum.

Wepropose addingesterase, esterase-lipase,

a-galactosi-dase, 1-galactosidase, a-glucosidase, and

N-acetyl-f3-glu-cosaminidase to the list of enzymes produced by some

strains of A. odontolyticus.

Atotal of 11 (all 3 reference strains and 8 clinical strains) of the 17strains, including the referencestrainsidentified as

A. israelii by the RapID ANA II system, were identified

correctlytothespecies level. Only three clinical strainsgave

thesamereactionsasthereference strains in the APIZYM system. Theremaining five clinical strains didnot produce

any ,-galactosidase or a- or 3-glucosidase, enzymes

pro-duced byat least 90% of A. israelii strains (18). However, thecarbohydratefermentationresults andanegative urease

reaction were consistent with the characteristics of A.

is-raelii. Most of the A. israelii strains in the study were

smooth-colony variants and were aerotolerant. Only two

study strains and ATCC strain 23860 did not grow

aerobi-cally.

Accordingtoourresultsand those of Kilian, the APIZYM enzymeprofilesof A. israelii and A. odontolyticus are very

similar, the onlydiscriminatingenzymebeinga-fucosidase.

Allsevenstrainsidentified asA.pyogenes bythe RapID

ANA II system were identified asArcanobacterium

hae-molyticumby theconventionalmethods. The cell

morphol-ogy and biochemical reactions of these two species very

muchresemble each other, anddifferentiationbetweenthem is difficult. Because the two species are nowadays

consid-ered to belong to two distinct taxa (18) and because A. pyogenesisolates areoften ofanimal origin, the

identifica-tion ofArcanobacterium haemolyticum asA. pyogenes is

misleading in terms of both the source and the disease association of the bacterium and ought to be corrected. Comparing the typestrains of thesetwo species, we found

that thehemolysisproducedbyArcanobactenium

haemolyti-cumwasmuch strongerand developedmorequickly(1-day

incubation) thanthat produced by A.pyogenes(3-day

incu-bation).Arcanobacterium haemolyticum also grew equally

well in allthreeatmospheres, butA.pyogenesgrewbetter in

anaerobic and CO2 atmospheres. Other important discrimi-native reactions were the reverse CAMP reaction, gelatin

liquefaction, and xylosefermentation (5). In theAPI ZYM

system, the most distinctive difference was between the

production ofacid phosphatase by Arcanobacterium

hae-molyticum and the production of ,3-glucuronidase by A.

pyogenes, characteristics which could beofhelptoclinical laboratories in differentiating thesetwospecies.

The reactions observed with the RapID ANA II system

weremostlyeasytointerpret, although borderline reactions occurred in the hydrolysisofnaphthylamine substrates. In the APIZYMsystem, minorerrorsininterpretingthe color

intensities were possible. Although some of the reactions

were included in both systems, the substrates were not

consistently the same, nor were the inoculum densities

equal.Therefore, directcomparison of the enzyme reactions

was not considered justified.

In thisstudy, theRapID ANA II system correctly

identi-fied to thespecies level all the A.

odontolyticus

strains and 65% oftheA.

israelii

strains. However, allArcanobactenum

haemolyticum

strains were misidentified as A. pyogenes.

Theenzyme reactions inthe API ZYM system aided in the

identification of the strains but did not always concur with

those given in Bergey's manual for equivalent species;

accordingly, additions totheenzyme profiles arepresented.

API ZYM profiles of some newer Actinomyces spp. not

tested before are reported here. In conclusion, both rapid

methodsareuseful aids in the identification of

Actinomyces-like bacteria, although accurate identification will often

requiresupplementary conventional methods.

ACKNOWLEDGMENTS

Wethank SeppoPaltemaa, RailiHaanpaa,and TuulaRasanenfor excellentmicrobiological assistance andLillianV. H.Moore, Vir-ginia Polytechnic Institute and State University, Blacksburg, for analyzing threerepresentatives of the aerotolerantA. meyeri-resem-bling strains. We thank P. H. Makela, director ofInfectious Dis easesDivision, National Public HealthInstitute, Helsinki, Finland, for criticalreview of the manuscript.

REFERENCES

1. Allen, S. D. 1985. Gram-positive, nonsporeforming anaerobic bacilli, p. 461-472. In E. H. Lennette, A. Balows, W. J. Hausler, Jr., and H. J. Shadomy (ed.), Manual of clinical microbiology, 4th ed. American Society for Microbiology, Washington, D.C.

2. Batty, I. 1958. Actinomyces odontolyticus, a new species of actinomycete regularly isolated from deep carious dentine. J. Pathol. Bacteriol. 75:455-459.

3. Cato, E. P., W. E. C.Moore, G.Nygaard,and L. V.Holdeman. 1984.Actinomycesmeyeri sp. nov., specific epithet rev. Int. J. Syst.Bacteriol. 34:487-489.

4. Celig, D. M., and P. C.Schreckenberger. 1991. Clinical evalua-tion oftheRapID-ANA II panel for identification of anaerobic bacteria. J.Chin. Microbiol. 29:457-462.

5. Clarmdge, J. E. 1989. The recognition and significance of Arcanobacterium haemolyticum. Clin. Microbiol. Newsl. 11:

41-45.

6. Finegold, S. M. 1989. General aspects of anaerobic infection, p. 137-153. In S. M. Finegold and W. L. George (ed.),Anaerobic infections in humans. Academic Press, Inc., San Diego, Calif. 7. George, W. L. 1989. Actinomycosis, p. 529-539. In S. M.

Finegold and W. L. George (ed.), Anaerobic infections in humans.Academic Press, Inc., San Diego, Calif.

8. Hillier, S., and B. J. Moncla. 1991. Anaerobic gram-positive nonsporeforming bacilli and cocci, p. 522-537. In A. Balows, W. J.Hausler, Jr., K. L.Herrmann,H. D.Isenberg, and H. J.

Shadomy(ed.),Manual of clinicalmicrobiology, 5th ed.

Amer-icanSociety forMicrobiology, Washington, D.C.

9. Holdeman, L.V., E. P. Cato, and W. E. C.Moore (ed.). 1977. Anaerobe laboratory manual, 4th ed. Virginia Polytechnic In-stitute and StateUniversity, Blacksburg.

10. Innovative Diagnostic Systems, Inc. 1989. RapID ANA II code compendium. Innovative Diagnostic Systems, Inc., Atlanta. 11. Johnson, J.L., L. V. H. Moore, B. Kaneko, and W. E. C. Moore.

1990.Actinomyces

geoigiae

sp. nov.,Actinomyces

gerencse-riaesp. nov., designation of two genospecies of Actinomyces naeslundii, andinclusion of A. naeslundii serotypes II and III andActinomyces viscosus serotype II in A. naeslundii geno-species 2. Int. J. Syst. Bacteriol.40:273-286.

12. Karachewski, N. O., E. L. Busch, and C. L. Wells. 1985. Comparison of PRAS II, RapID ANA, and API 20A systems for

on April 12, 2020 by guest

http://jcm.asm.org/

(5)

identification of anaerobic bacteria. J. Clin. Microbiol. 21:122-126.

13. Kilian, M. 1978. Rapididentification ofActinomycetaceae and related bacteria. J. Clin. Microbiol. 8:127-133.

14. Marler, L. M., J. A. Siders, L. C. Wolters, Y.Pettigrew, B. L. Skitt,and S.D.Allen.1991. Evaluation ofthenewRapID-ANA II system for theidentification of clinical anaerobic isolates. J. Clin. Microbiol. 29:874-878.

15. Mitchell, R. G., and M. R. Crow. 1984.Actinomyces odontolyti-cus isolated from the female genital tract. J. Clin. Pathol. 37:1379-1383.

15a.Moore, L. V. H. Personal communication.

16. Murray, P. R., C. J. Weber, and A. C. Niles. 1985. Comparative evaluation ofthree identification systems for anaerobes. J. Clin. Microbiol.22:52-55.

17. Pordy, R. C. 1988. Lumpyjaw due toActinomyces meyeni: report ofthefirstcaseand review of theliterature. Mt. Sinai J. Med.55:190-193.

18. Schaal,K. P. 1986. GenusActinomyces Harz 1877, 133-', p. 1383-1418. In P. H. A.Sneath,N.S.Mair, M. E. Sharpe, and J. G.Holt (ed.), Bergey's manual of systematicbacteriology, vol. 2.Williams &Wilkins,Baltimore.

19. Summanen, P., and H. Jousimies-Somer. 1988. Comparative evaluation of RapID-ANA and API 20A for identification of anaerobic bacteria. Eur. J. Clin. Microbiol. Infect. Dis. 7:771-775.

20. Sutter,V. L., D. M. Citron, M. A. C. Edelstein, and S. M. Finegold (ed.). 1985. Wadsworth anaerobic bacteriology man-ual, 4th ed. Star Publishing Co., Belmont, Calif.