Copyright © 1992, AmericanSocietyforMicrobiology
Fluorogenic Substrates for Differentiation of
Gram-Negative
Nonfermentative and Oxidase-Positive Fermentative
Bacteria
PETERKAMPFER,* ISOLDE KULIES, ANDWOLFGANG DOFT
Fachgebiet Hygiene, Technische UniversitatBerlin,Amrumerstrasse 32,
D-1000 Berlin 65, Federal Republic of Gernany
Received 22October 1991/Accepted5 March 1992
A totalof803 strains of gram-negative nonfermentative and oxidase-positive fermentative bacteria (38taxa)
wereinvestigated for theirabilitytohydrolyze 53 different fluorogenic 4-methylumbelliferyl- and
1-naphthyl-amide-linked substrates within 6 h of incubation. The hydrolysis of 16 fluorogenic substrates showed high
separation index valuesamongthe testedtaxa,wasreproducible, and showed goodagreementwith data in the
literature. In combination with other biochemical tests (like carbon substrate utilization tests and classical
biochemical tests), hydrolysisprofiles can improvethe differentiation of gram-negative nonfermentative and oxidase-positivefermentativebacteria.
The useof enzymatic profiling of microorganisms by using
fluorogenic or chromogenic substrates takes advantage of
thepresenceofpreformed enzymes,which can be detected veryrapidly (after
.6
h ofincubation) (4). These qualitative enzyme tests can be used to differentiate a wide range of differentbacteria, and their applications to specific taxa have been reviewed(2, 29). Especially withinthe
familyEntero-bactenaceae, these methods can be used to improve the differentiation of various species (25) ormaybe performed forthespecific detection of
Escherichia
coli(5, 8, 27, 35, 39). Withparticularemphasisongram-negative nonfermentative and gram-negative, oxidase-positive fermentative bacteria,however, these sensitive tests have onlybeen applied to a
restricted set of taxa (1) or have been used as
growth-dependent tests (22, 23). After the development of a com-mercially availablekit(APIZYM;Analytab Products, Plain-ville, N.Y.) (21), some species were studied for their enzymatic profiles (14,31, 33, 34, 36,41,42).Inaddition,the
activitiesof manyspecificenzymeshave also beenrecorded,
for taxonomic purposes, within the generaFlavobacterium (9, 11-14) Weeksella (17, 18), Chryseomonas and
Flavimo-nas (16, 19), and Ochrobactrum (15) by using expanded, experimental API ZYM systemgalleries oftests.However,
the potential of qualitative enzyme tests to differentiate between thenumerous taxa ofgram-negative
nonfermenta-tiveandoxidase-positivefermentative bacteria hasnotbeen
systematicallyevaluated.
This report deals with anextensivestudyon a number of strains representing38 differenttaxaofgram-negative non-fermenters and gram-negative, oxidase-positive fermenta-tive bacteria whichweretested fortheirabilitytohydrolyze
53 different fluorogenic substrates. These testswere evalu-ated fortheirseparationindex valuesandreproducibility.In
addition, thetestresultswerecomparedwith dataavailable in theliterature.
MATERIALSANDMETHODS
Bacteria. A total of 803 strains representing 38 taxa of
gram-negative nonfermenters and gram-negative,
oxidase-positive fermenterswere tested. Of thesestrains, 525 were
obtained -from the National Collection of Type Cultures, *Corresponding author.
London, England, and were identified by the methods used
atthe NationalCollection of Type Cultures(10). The
remain-ing strainswere obtained from different clinical specimens that were identified by different methods. All strains were correctly identified by using the Titertek-NF system (ICN-FlowLaboratories, Meckenheim, Federal Republic of
Ger-many) (24). Beforetesting, mostof the isolatesweregrown
ontryptone soy agar(Oxoid,Basingstoke, England) for 24 to 48 hat30°C.Speciesofthe genera Moraxella andBordetella
wereculturedon sheep blood agar (Oxoid) for 48 h at 30°C. Tests. A total of 15 4-methylumbelliferyl (4-MU)-linked and 38
p-naphthylamide (1-NA)-linked
substrates weretested. All substrates listed below and in Table 1 were
obtained fromServa, Heidelberg, Federal Republicof Ger-many, and Sigma, Deisenhofen, Federal Republic of Ger-many, respectively. The majority of the substrates (final concentration,2x 10-4 M)werepredissolvedinafewdrops
ofdimethyl sulfoxide. L-Glutamic acidy-(4-methoxy-,B-NA) was predissolved in 1 ml of acetate-ethanol (1:1; vol/vol); L-aspartic
acid-1-NA
waspredissolved in 1ml of 1 M HCl;and the 4-MU-fucopyranosides were dissolved in 0.1 M phosphatebuffer,becausefucosidasesareinhibitedby
Tris-HClbuffer(27). Theremainingsubstratesweredissolved at afinal concentration of 2x 10-4Min 0.1 MTris-HCl buffer. The pHwas8.0throughoutthestudy.Alltestsolutionswere
sterilizedbyfiltration(pore size,0.2
pum)
andfilled in100-,u amountsintomicroplates (ICN Flow).Toeachwell, 100pul
ofabacterial suspensionin0.9%NaCl(McFarland tubidity
standard tube no. 8) was added. These
high
cell densitieswere controlled by routine checks
(dilution
experiments).
Afteramixing step, the microplatewassealed withasterile
plastic sealer. Following incubation for 6 h at 30°C, test plateswerereadby usingafluorimeter(Fluoroscan II; ICN
Flow), connected to a personal computer
(Deskpro286;
Compaq, Munich, Federal
Republic
ofGermany).
The fluo-rimeter was autocalibrated prior to measurement, and the fluorescence ofanegative control medium(Tris-HCl buffer[pH 8.0]) containing no
fluorogenic
substrate wasalways
recorded.
Results were measured at a wavelength of 355 nm for
excitation and 460nmforemission. Aftercalculation of the
differences (S = T -
C)
between the relative fluorescence intensities of the test(1)
and the control(C) wells,
a test was coded positive if the difference values(S)
exceeded 1402on April 12, 2020 by guest
http://jcm.asm.org/
20. For the following tests, higher threshold levels were
chosen, because spontaneous hydrolysis of the substrates was observed: 4-MU-phosphate (S > 430); 4-MU-13-D-glu-copyranoside (S > 230); 4-MU-N-acetyl-13-D-galactosamin-ide,4-MU-N-acetyl-1-D-glucosaminide, and 4-MU-13-D-man-nopyranoside (S > 60).
Selection of differentiating characteristics and
reproducibil-ity.
Test selection for calculating the usefulness of the tests for separating the taxa was performed with a computer program that calculated the different separation indices of the CHARSEP program (37): the sum of characterizationmeasure(C) and rank measure (R) defined by Gyllenberg (7), the index of Niemela et al. (30), the variance of strain potential (VSP) index, and the consistency of strain potential
(CSP) index (37). The formulae of the different indices and thefollowingdefinitions are given by Sneath (37). The tests
recommendedfordiagnosticpurposes wereselected, if they showedhigh separation indices in addition to high
reproduc-ibilities. In addition, test results for the most diagnostic
characteristics were compared with those reported in the literature (1, 9, 11-20, 22, 23).
Thereproducibilities of the reactions weredetermined by testing 20 strains threefold on separate days, and an overall probability of erroneous results was calculated as described bySneath and Johnson (38).
RESULTS
Experimental test error was calculated from the data for the 20 duplicate strains. The average probability of an
erroneousresult was4.8% (formula 4 of Sneath and Johnson
[38]). In general, the glycosidase tests were more reliable; they displayed variances of <0.05. The majority of
ami-nopeptidase testswere less reproducible, but most of their variances were <0.1, which has been recommended to be an acceptable level (38).
All test strains were positive for the hydrolysis of
L-ala-nine-1-NA.
Test results for the remaining substrates were different within the numerous taxa. First, the different sep-aration indices(37)werecalculated. Gyllenberg's (7) sum of Crangedfrom 43(DL-methionine-1-NA)
to 53(L-alanine-13-NA) and Gyllenberg's (7) R ranged from 0
(L-alanine-13-NA)
to23,936
(4-MU-N-acetyl-13-D-glucosaminide).
The index of Niemela etal. (30) ranged from 0(L-alanine-13-NA) to 27.68(4-MU-N-acetyl-13-D-glucosaminide).
Eighteen fluorogenictestshadVSP values of less than30% and CSP values of less than 52%.
Therefore, the following substrates obtained from Serva
are not recommended for differentiation purposes:
4-MU-P-D-glucuronide,
D-alanine-t3-NA, glycine-,B-NA,glycyl-glycyl-glycyl-phenylalanine-c-NA, glycyl-
phenylalanine-4-methoxy-1 -NA, L-isoleucine-B-NA, L-lysine-1 -NA,
L-lysyl-L-alanine-4-methoxy-o-NA,
andL-methionine-1-NA. In addition, the following substrates from Sigma are
not recommended for differentiation purposes:
4-MU-13-L-fucoside, 4-MU-sulfate,
L-alanine-4-methoxy-13-NA,
L-argi-nine-4-methoxy-1-NA, L-hydroxyproline-13-NA,
D-leucine-,B-NA,
L-leucine-1-NA, L-leucine-4-methoxy-o-NA,
and L-threonine-13-NA.Theremaining 34 substrates had VSP indices of >30% and CSP indices of >53%. However, several of the results obtained from these tests differed from those recorded in the literature (9, 11-20, 28, 34). In addition, some of the tests showed inconsistencies in our reproducibility study. Fur-thermore, some of the tests revealed very similar profiles within the taxa; e.g., the profiles for
L-valine-13-NA
andL-valine 4-methoxy-,B-NA or 1-NA and
L-proline-4-methoxy-1-NAwerealmostidentical. For these numerous reasons, thefollowing substrates obtained from Serva were considered to be not sufficient for differentiation purposes: 4-MU-phosphate, L-aspartic acid-2-methoxy-1-NA, L-glu-tamic acid-1 -NA, L-lysyl-L-proline-methoxy -1 - NA, L-proline-4-methoxy-1 -NA, L - serine - 4 -methoxy - 1 -NA, and L-valine-13-NA. In addition, the following sub-strates obtained from Sigma were considered to be not sufficient for differentiation purposes: 4-MU-a-L-arabinopy-ranoside, 4-MU-13-D-fucoside, 4-MU-13-D-glucopyranoside, 4-MU-13-D-mannopyranoside, N-ao-benzoyl-DL-arginine-1-NA, DL-methionine-1-NA, glycyl-L-prolyl-L-leucine-13-NA,
L-arginine-1-NA, L-arginyl-L-arginine-13-NA,
L-lysyl-L-ala-nine-1-NA, L-pyroglutamicacid-13-NA, and L-serine-13-NA. The remaining 16 substrates, which showed high repro-ducibilities and high separation indices, are given in Table 1 asthe percentageof positivetestsfor each taxon. Within the genus Pseudomonas, we found some species (P. pseudo-mallei, P. cepacia, P. diminuta, P. vesicularis) which hy-drolyzed more of the tested substances than did other species (P. aeruginosa, P. fluorescens, P. putida). Other species, however, were able to cleave only a few of the 16 different chromogenic substrates (P. pickettii, P. stutzeni, P.
mendocina, P.pseudoalcaligenes). Strains of Comamonas
acidovorans and Comamonas testosteroni exhibited reac-tion patterns very similar to those of the inactive Pseudo-monas species. The species of the genus Alcaligenes that we investigated could not be differentiated from each other (Table 1). Representatives of the genera Moraxella and
Acinetobacter were found to be inactive concerning their abilities to hydrolyze fluorogenic substrates, which has also been found to be the case for growth-dependent enzyme tests(23).
Agrobacterium as well asFlavobacteniumspecies showed manypositive resultsconcerningthehydrolysis of the sub-stances tested. Within the genus Flavobactenum, several differentiating features could be found (Table 1); however, taxawhich cannot be easily separated by conventional tests (32), like Flavobacterium group IIb and
Flavobactenum
meningosepticum,were also shown to be very similar.
Othpr
active species were Shewanella putrefaciens, Sphingomo-naspaucimobilis, andXanthomonas maltophilia. Theoxi-dase-positive fermentative species of the genera Vibrio,
Plesiomonas, and Aeromonas also showed more positive results than the majority of the nonfermentative species did, and several differentiating characteristics could be observed (Table 1).
DISCUSSION
Aspointed out by D'Amato et al. (4), the elements for test selection fordiagnostic purposes should consider the prob-ability of a positive or a negative result for each test, the separation index values, and the reproducibility. These
guidelineswere strictly followed in the analysis described here. We used several separation indices in our study; however, aspointed out elsewhere (37), the sum of C and R is not easy to comprehend, nor is the index ofNiemela et al. (30). Therefore, we used the separation indices CSP and VSP (37) to eliminate those tests that showed little separa-tion potential. Sneath (37) recommended that CSP values of less than 50% and VSP values of less than 25% are of little value in an identification matrix.
All test strains were positive for the hydrolysis of
L-ala-nine-1-NA,
which has been demonstrated to beon April 12, 2020 by guest
http://jcm.asm.org/
>4
0
0.
0
0A
0) 0
oCo
_ 4; 00Et LL3
0)
0 rA
0)
C's
o,°eo°N'IC t000
"- -4 1- P- r- -4 -4
- o ccoQor Q ci N ooO^c_O_sooroOOOr^db>ooO
0 ci 0000 I~ ~O ON 00 0 0 00 ) QOt t 0_0rOC0,O rO m 00 0
OiN O all 0 00 0 0 o ) a o CDo 0 OoI "t o 0
0000000000 0000 cl~~~~~~~~~~~~~~~~~0 0000000000000~~~~~~~~~~~~~~~o0100 IC00 00
--- - -
--~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~l-0t0ci~ o o oooooc00ooOOOO~ 0 00o1ooooo-oi0D -4\0c asO00
---,-4- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~-4
-4-ci
'q
000C 0 0 N O O700 000
0 0 0 0N* _ O£ci0000800O\00OD00~000 0 0O00
- -4 -4- -f -4 ---- - -
---OtOOONOO 00 N Q O O O O O _ O O O O O O O O O O O N b 908-40 00tnN V: O O oo 0-4 0
0 ON 00 00 N 0000 0 0 0 O 0 O 0 0 0 0 0 000
X _ _ c o ~o o O 0 O -00oo£t
O--QOO -4
00ct'r00 00000000 0 P-40000000ci000000c~ 0 _0 cD
r-~4 0000C) O CDO00 t
--4-4- 4 1- -4 -4
rm 000 0~O000000 0(0 0 0 00_ IO 8 0 all t
00 en Cs0 0 0 0"C 0 OOO
--4--_ __ _- -4_
0 o000o oi o0 o o ooooooooo00000000000000000 0 C> CD CD 0 oO0 O O O Oco
- -~~~~~~~~~~~~~,-4 -
-r-00u~~0000.O~~-0ooc4000000 0000000 c-
or-t 0 O 4 00=0 -O 0 0 00 00 00 00
- - - -- -
-00000000~~~~~~~n000 0 000-4000000000c0 0 0 00 )0Q ciOOO0 =o8OCo= 1O00=" 00 00
\°00 £0 00°i - \O - c^ c 0N 0° 0°~U ObN - ft c~ 00%^ O0N Ci 0 '-4 -0 000% '-40
c* -__^N~t___d4_1__-4r----_____O__ -_--_--r- -_
0 0
'.-..'0
. .
-0 c..
. 0 0
.
Q
L)
Q
ts
0: Q)
0n
005
0l) 03
t4--0z
on April 12, 2020 by guest
http://jcm.asm.org/
tic for gram-negative bacteria (3), but it has no diagnostic
value among the taxa.
Results of the 16 tests given in Table 1 were, furthermore, compared in detail with data in the literature. In the study of Algorta et al. (1), the hydrolysis of 18 p-nitrophenyl-linked
compounds in addition toonep-nitroanilide-linked substrate (paper disk method) after 18 h of incubation at 37°C was tested for 22 species (211 strains). For the majority of tests, ourresults were in line (85% averageagreement) with those reported by Algorta et al. (1); however, less agreement could be found for the substrates 4-MU-N-acetyl-3-D-glu-cosaminide and 4-MU-N-acetyl-,-D-galactosaminide.
Be-cause the tests were reproducible in our study and showed good agreement with the results of all other studies reporting these enzyme activities (9, 11-19, 22, 23, 28), the results obtained by Algorta et al. (1) were probably caused by the unstablenatureof these substrates, which has been reported in the literature (31).
A comparison of the results for five substrates, which were obtained in growth-dependent tests (22, 23), revealed 88% average agreement for 38 taxa of nonfermentative bacteria and 87% average agreement for 11 oxidase-positive fermentative species.
In the taxonomic studies of Holmes and colleagues (9, 11-19) and Kiredjian et al. (28), a total of 58 chromogenic substrates were tested on nine newly described species. A
comparison of 14 tests gave an overall agreement of 84%. However, it should be noted that the results of several glycosidase tests were listed as discrepant. For the 3-galac-tosidase reaction, which is a key test in the differentiation of gram-negative nonfermentative bacteria (6, 32), our results
were in line with those obtained by the classical method
(o-nitrophenyl-13-D-galactopyranoside
test). In the studies of Holmes and colleagues (9, 11-14, 17-19) on different species of the generaFlavobacterium, Weeksella, Flavimonas, andChryseomonas, these discrepancies have already been noted. All five glycosidase substrates that showed high
discrepancies when compared with our results were 3-NA-linked or 6-bromo-,B-NA-3-NA-linked compounds. These results
are a reflection of the different assay methods that we used. Hsu et al. (20) investigated cultures of four Pseudomonas
species (P. fluorescens, P. putida, P. pickettii, and P.
stutzeri) for their aminopeptidase profiles and found higher activities for L-aspartic acid aminopeptidase and -y-L-glu-tamic acidaminopeptidase in P. pickettii and P. stutzen than in P. fluorescens and P. putida, which is in line with our results(Table 1). Alcaligenes species and Moraxella species could not be separated by fluorogenic enzyme tests; how-ever, they canbe differentiated by carbon substrate utiliza-tion tests (28) or from their different physiological features (32). For the oxidase-positive fermentative species of the genera
Vibrio,
Plesiomonas, and Aeromonas, some fluoro-genictests(Table 1) can be used in addition to those reported inthe literature (26, 40).As pointed out elsewhere (25), various parameters like incubation temperature, inoculum size, and duration of incubation may influence the testresults. In addition, when using commercially available enzyme tests (e.g., the API
ZYM system), a gradual color change is seen when the
hydrolysis of the substrate occurs, and the choice of the cutoff point between positive and negative is largely subjec-tive. Reading of test results via a fluorimeter can minimize these subjectivities, and further developments, including
morestandardized andreproducible methods, may even use
quantitative enzyme activities fordifferentiation purposes. The identification ofgram-negative nonfermentative
bac-teria in the routine laboratory is still difficult in some cases. Although several classical biochemical tests have been de-scribed and performed for theidentificationof these bacteria (6, 32), many problems still frequently arise because of the diversity of these microorganisms. The use of qualitative
enzyme tests is not sufficient for the identificationof alltaxa
of gram-negative nonfermentative bacteria, but they can be
used as supplementary tests inconnection with some other
test groups (e.g., carbon substrate assimilation tests) in order to achieve a more precise identification of these
organisms.
REFERENCES
1. Algorta, G., V. Monteil, and M. Y. Popoff. 1987. Profils d'ac-tivites en enzymatiques des genresAgrobacterium,Alcaligenes, Alteromonas,Flavobacteriumet Pseudomonas. Ann. Inst. Pas-teur/Microbiol. 138:297-302.
2. Bascomb, S. 1987. Enzyme tests in bacterial identification. Methods Microbiol. 19:105-160.
3. Cerny, G. 1978. Studies on the aminopeptidase test for the distinction of gram-negative bacteria from gram-positive bacte-ria. Eur. J. Appl. Microbiol. Biotechnol. 5:113-122.
4. D'Amato, R. F., E. J. Bottone, and D. Amsterdam. 1991. Substrate profile systems for the identification of bacteria and yeasts by rapid and automated approaches, p. 128-136. In A. Balows, W. J. Hausler, Jr., K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (ed.), Manual of clinical microbiology, 5th ed. American Society for Microbiology, Washington, D.C. 5. Edberg, S. C., and C. M. Kontnick 1986. Comparison of
,B-glucuronidase-based substrate system for identification of Escherichia coli. J. Clin. Microbiol. 24:368-371.
6. Gilardi, G. L. 1991. Pseudomonas and related genera, p. 429-441. In A. Balows, W. J. Hausler, Jr., K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (ed.), Manual of clinical microbiology, 5th ed. American Society for Microbiology, Washington, D.C.
7. Gyllenberg, H. G. 1963. A general method for deriving determi-nation schemes for random collections of microbial isolates. Ann. Acad. Sci. Fenn. Ser. A. IV Biol. 69:1-23.
8. Heizmann,W., P. C.Doller, B. Gutbrod, and H. Werner. 1988. Rapid identification ofEscherichia coli by fluorocult media and positive indole reaction. J. Clin. Microbiol. 26:2682-2684. 9. Holmes, B. 1983. The taxonomy of the genusFlavobacterium,
p. 273-294. In H. Leclerc (ed.), Les bacilles a gram-negatif d'interet medical et en sante publique. Leseditions INSERM, InstitutNational de laSante et de la RechercheMedicale, Paris. 10. Holmes, B., C. A. Dawson, and C. A. Pinning. 1986. A proba-bility matrix for the identification of gram-negative, aerobic, non-fermentative bacteria that grow on nutrient agar. J. Gen. Microbiol. 132:1827-1842.
11. Holmes, B., D. G. Hollis, A. G. Steigerwalt, M. J. Pickett, and D. J. Brenner. 1983. Flavobacterium thalpophilum, a new species recovered from human clinical material. Int. J. Syst. Bacteriol. 33:677-682.
12. Holmes, B., R. J.Owen,and D. G. Hollis. 1982.Flavobacterium
spiritivorum, anewspecies isolated from human clinical speci-mens. Int. J. Syst. Bacteriol.32:157-165.
13. Holmes, B., R. J. Owen, A. G.Steigerwalt,and D. J. Brenner. 1984. Flavobacterium gleum, a new species found in human clinical specimens. Int. J. Syst. Bacteriol. 34:21-25.
14. Holmes, B.,R. J. Owen, and R. E. Weaver. 1981. Flavobacte-num multivorum, a new species isolated from human clinical specimens and previously known as group Ilk. Int. J. Syst. Bacteriol. 31:21-34.
15. Holmes, B., M. Popoff, M. Kiredjian, and K. Kersters. 1988. Ochrobactrumanthropi gen. nov.,sp. nov. from human clinical specimens and previously known as group V-d. Int. J. Syst. Bacteriol. 38:406-416.
16. Holmes, B., A. G.Steigerwalt,R. E.Weaver, andD. J.Brenner. 1986.Chryseomonaspolytrichagen. nov., sp. nov., a Pseudo-monas-like organism from human clinical specimens and for-merly known as group Ve-1. Int. J. Syst.Bacteriol.36:161-165.
on April 12, 2020 by guest
http://jcm.asm.org/
17. Holmes,B.,A. G. Steigerwalt,R. E. Weaver, and D.J.Brenner. 1986. Weeksellavirosagen. nov., sp. nov. (formerlyIIf),from
human clinicalspecimens. Syst.Appl. Microbiol. 8:185-190. 18. Holmes, B., A. G. Steigerwalt, R. E. Weaver, and D. J. Brenner.
1986.Weeksella zoohelcumsp. nov.(formerlyIIj), from human
clinicalspecimens. Syst. Appl. Microbiol. 8:191-196.
19. Holmes, B., A. G. Steigerwalt, R. E. Weaver, and D. J. Brenner. 1987. Chryseomonas luteolacomb.nov. and Flavimonas
oryz-ihabitans gen. nov., comb. nov., Pseudomonas-like species
from humanclinical specimens and formerly known, respectively,
as groupsVe-1and Ve-2. Int. J.Syst. Bacteriol. 37:245-250. 20. Hsu,E.J., J.H. Godsey, E. K. Chang, and S. L. Landuyt. 1981.
Differentiation ofpseudomonads by amplification of metabolic profiles. Int. J. Syst. Bacteriol. 31:43-55.
21. Humble, M. W.,A. King,and I. Phillips. 1977. API ZYM: a
single rapid systemfor the detection ofbacterial enzymes. J. Clin. Pathol.30:275-277.
22. Kainpfer, P., W. Bette, and W. Dott. 1987. Phenotypic differen-tiation of members of the familyVibrionaceae using miniatur-ized biochemicaltests.Zentralbl. Bakteriol. Parasitenkd. Infek-tionskr.Hyg. Abt. 1Orig. ReiheA266:425-437.
23. Kampfer, P., andW.Dott.1988. Growth dependent enzymatic profiles of some gram-negative nonfermentative bacteria of
clinical significance. Zentralbl. Bakteriol. Parasitenkd. Infek-tionskr.Hyg.Abt. 1Orig. ReiheA269:460-467.
24. Kiimpfer, P., and W. Dott. 1989.Evaluationof theTitertek-NF systemforidentification ofgram-negativenonfermentative and oxidase-positive fermentative bacteria. J. Clin. Microbiol. 27: 1201-1205.
25. Kampfer, P., 0. Rauhoff, and W. Dott. 1991. Glycosidase
profiles of members of the family Enterobacteriaceae. J. Clin. Microbiol. 29:2877-2879.
26. Kelly, M. T., F.W. Hickman-Brenner, and J. J.Farmer III. 1991.Vibric,p.384-395.In A.Balows,W.J.Hausler, Jr.,K. L.
Herrmann, H.D.Isenberg, and H. J. Shadomy (ed.), Manualof clinicalmicrobiology,5th ed. American Societyfor
Microbiol-ogy,Washington,D.C.
27. Kilian,M., andP.Bulow. 1976.Rapid diagnosisof
Enterobac-tenaceae. I. Detection of bacterialglycosidases. ActaPathol.
Microbiol. Scand. Sect.B84:245-251.
28. Kiredjian, M., B. Holmes, K. Kersters,I. Guilvout, andJ. De Ley.1986.Alcaligenespiechaudii, anew species fromhuman clinicalspecimensand the environment. Int. J.Syst.Bacteriol. 36:282-287.
29. Manafi, M., W. Kneifel, andS.Bascomb.1991.Fluorogenicand chromogenic substrates used in bacterial diagnostics. Micro-biol.Rev.55:335-348.
30. Niemela,S.I.,J. W.Hopkins, and C. Quadling. 1968.Selection aneconomicalbinary testbatteryfor a setofmicrobialcultures. Can.J.Microbiol. 14:271-279.
31. O'Brien, M., and G. H. G. Davis. 1982. Enzymatic profile of Pseudomonas maltophilia. J. Clin. Microbiol. 16:417-421. 32. Pickett,M. J., D. G. Hollis, and E. J. Bottone. 1991.
Miscella-neousgram-negative bacteria,p.410-428.In A.Balows, W. J. Hausler, Jr., K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (ed.), Manual ofclinicalmicrobiology, 5th ed. Amer-icanSociety for Microbiology,Washington, D.C.
33. Poo, C. L., and G. K.Loh. 1988.Enzymatic characterizationof Pseudomonas cepacia by API ZYM profile. J. Clin. Microbiol. 26:607-608.
34. Rossau, R., K. Kersters, E. Falsen, E. Jantzen, P. Segers, A. Union, L. Nehls, and J. De Ley. 1987. Oligella, a new genus including Oligella urethralis comb. nov. (formerly Moraxella urethralis) and Oligella ureolytica sp. nov. (formerly CDC groupIVe): relationship toTaylorellaequigenitalis and related taxa.Int.J. Syst. Bacteriol.37:198-210.
35. Sarhan, H. R., and H. A. Foster. 1991. A rapid fluorogenic method for thedetection of Escherichia coli by the production of,3-glucuronidase.J. Appl. Bacteriol.70:394-400.
36. Smalley,D. L.1982. Enzymaticprofileof Pseudomonaspauci-mobilis.J. Clin. Microbiol. 16:564-565.
37. Sneath,P.H. A.1979. BASICprogramforcharacterseparation indices from anidentification matrixof percentpositive charac-ters.Comput. Geosci. 5:349-357.
38. Sneath, P. H. A., and R. Johnson. 1972. The influence on numerical taxonomic similarities of errors in microbiological tests. J.Gen. Microbiol. 72:377-392.
39. Trepeta, R. W., and S. C. Edberg.1984.
Methylumbelliferyl-13-D-glucuronide based mediumfor rapidisolation and identifica-tion ofEscherichia coli. J. Clin. Microbiol.19:172-174. 40. Von Graevenitz, A., and M. Altwegg. 1991. Aeromonas and
Plesiomonas, p. 396-401. In A. Balows, W. J. Hausler, Jr., K. L. Herrmann, H. D. Isenberg, and H. J. Shadomy (ed.), Manualofclinicalmicrobiology, 5th ed. American Societyfor Microbiology, Washington, D.C.
41. Waltman, W. D., II,E. B. Shofts, andT.-C. Hsu. 1982. Enzy-matic characterization ofAeromonas hydrophila complex by theAPI ZYM system. J. Clin. Microbiol. 16:692-696. 42. Willems, A., B. Pot, E. Falsen, P. Vandamme, M. Gillis, K.
Kersters, andJ.DeLey. 1991. Polyphasic taxonomicstudy of theemendedgenusComamonas:relationshiptoAquaspirillum aquaticum, E.Falsen group10,and other clinical isolates. Int. J. Syst.Bacteriol. 41:427-444.
