9225 DIFFERENTIATION OF COLIFORM BACTERIA*
9225 A. Introduction
The significance of various coliform organisms in water has
long been a subject of considerable study. Collectively, coliform
bacteria (also called total coliforms) are considered indicators of
possible inadequate sanitation, fecal contamination, and/or
pathogenic and opportunistic pathogenic bacteria. However,
completely and accurately differentiating coliform bacteria re-
mains a challenge.
Until recently, the definition of coliform bacteria did not
include the genera and species of Enterobacteriaceae. The tra-
ditional definition of coliform bacteria, based on lactose fermen-
tation with the production of gas, excluded some strains of
Escherichia coli and included some strains of enteric pathogens
(e.g., Salmonella and Shigella).1 However, as a result of DNA
hybridization tests and other molecular methods, researchers
now classify many species of Enterobacteriaceae as coliform
bacteria.1– 4 There are 18 Enterobacteriaceae genera with spe-
cies or distinct biogroups that ferment lactose and produce gas
during fermentation. Sixty-eight of the 134 named Enterobacte-
riaceae (59%) are coliforms.2 These must be considered along
with traditional coliforms (e.g., Citrobacter, Enterobacter, Esch-
erichia, and Klebsiella) during coliform differentiation. Informa-
tion on nomenclature, classifications, original literature citations,
and detailed information for genera and species in the family
Enterobacteriaceae and newly identified species, with their pub-
lication date, is available.†
Users must decide whether a coliform culture requires a pre-
cise taxonomical identification due to safety, public health, or a
regulatory requirement. This is not a normal practice in the
standard water-testing laboratory.
A history of coliform bacteria identification and differentiation
can be found in the literature.3,4 Other recent reviews1,2,4 detail
coliform differentiation and identification, and discuss the many
variables involved.
Completely and accurately identifying coliform bacteria is a
daunting task; an Enterobacteriaceae reference laboratory may
be needed. A practical solution for a water laboratory is to
consider several coliform-differentiation methods, to word re-
ports precisely, and to include comments qualifying the reported
results (see 9225E for examples).
In the following sections, both cited references and a biblio-
graphy are available for further information.
1. Definition
Total coliforms are historically defined as facultative anaero-
bic, gram-negative, non-spore-forming rods that ferment lactose
* Approved by Standard Methods Committee, 2011.
Joint Task Group: Margo E. Hunt (chair), Michael H. Brodsky, Archie J. Degnan,
Gil Dichter, J.J. Farmer, III, Peter Feng.
† See http://www.bacterio.cict.fr/. Accessed September 2011.
1
with acid and/or gas formation within 48 h at 35°C. However,
there are also slow-growing and anaerogenic (non gas-produc-
ing) lactose-fermenting strains of Escherichia coli and coliforms.
A new definition of coliform bacteria— based on enzyme-
substrate tests—includes those that produce the enzyme ␤-D-
galactopyranosidase to hydrolyze chromogenic substrates, such
as ortho-nitrophenyl-␤-D-galactopyranoside (ONPG) or chloro-
phenol red-␤-D-galactopyranoside (CPRG). This reaction gener-
ally occurs within 24 ⫾ 2 h or less at 35°C. When using
chromogenic media, E. coli are defined as coliform bacteria with
the enzyme ␤-glucuronidase, which hydrolyzes 4-methyl-umbel-
liferyl-␤-D-glucuronide (MUG) to 4-methylum-belliferone, a
fluorescent compound.‡
A few coliform bacteria can grow under elevated temperature
conditions (e.g., 44.5°C), but up to 10% of E. coli cannot grow
at an elevated temperature.5 However, many commercial and
traditional microfiltration methods incubate cultures at 35°C
with various enzyme/substrates to differentiate E. coli from other
coliforms. So, a requirement for thermophilic incubation would
both be problematic and challenge the specificity of these meth-
ods, many of which have been subjected to extensive validation.
2. References
1. FARMER, J.J. III, K.D. BOATWRIGHT & J.M. JANDA. 2007. Enterobac-
teriaceae: Introduction and Identification. In Manual of Clinical
Microbiology, 9th ed., Chapter 42. Amer. Soc. Microbiol. Press,
Washington, D.C.
2. BRENNER, D.J. & J.J. FARMER III. 2005. Family 1. Enterobacteriaceae.
In Bergey’s Manual of Systematic Bacteriology, 2nd ed., Vol. 2: The
Proteobacteria, Part B: The Gammaproteobacteria. Springer, Secau-
cus, N.J.
3. FARMER, J.J. III, M.K. FARMER & B. HOLMES. 2005. The Enterobac-
teriaceae: General Characteristics. In Topley & Wilson’s Microbi-
ology and Microbial Infections, 10th ed., Vol. 2: Bacteriology, Chap-
ter 51. Hodder Arnold, London, U.K.
4. JANDA, J.M. & S.L. ABBOTT. 2006. The Enterobacteria, 2nd ed. Amer.
Soc. Microbiol. Press, Washington, D.C.
5. EDBERG, S.C., M.J. ALLEN & D.B. SMITH. 1994. Comparison of the
Calilert® Method and Standard Fecal Coliform Methods, Report No.
90647. AWWA Research Foundation & American Water Works
Association, Denver, Colo.
3. Bibliography
Topley & Wilson’s Microbiology and Microbial Infections. 2006. 10th
ed., Wiley-Blackwell, New York, N.Y.
‡ E. coli may be confirmed via broth media (e.g., Sections 9221F and 9222G) or
media incubated at elevated temperatures.
 DIFFERENTIATION OF COLIFORM BACTERIA (9225)/Identification
9225 B. Culture Purification
1. Procedure
A pure culture is essential for accurate identification. “Pre-
sumptive positive” tests for coliforms occur either in a broth
culture tube (acid and/or gas production) or a lactose-positive
colony on a membrane filter. For enzyme substrate tests, use a
test broth, which is positive. Both broth and membrane filter
cultures are often mixed—not pure— cultures. Streak a loopful
of the liquid sample (or a colony) onto a plate containing a
suitable growth medium (e.g., MacConkey agar) to obtain 50 to
100 isolated colonies. Alternatively, produce a pour or spread
plate with the culture to obtain 50 to 100 isolates (for details,
review Sections 9215B and C). MacConkey agar is ideal because
it is commercially available and causes colonies that ferment
lactose rapidly to turn red-pink, while lactose-negative colonies
remain colorless. However, colonies of gram-positive and even
anaerobic organisms may be mixed with or underneath the
lactose-fermenting colony. After an appropriate incubation per-
iod (generally 35 ⫾ 0.5°C for at least 24 h), use a sterile needle
or loop to gently pick up a well-isolated target colony and streak
it on another medium (e.g., MacConkey, tryptic soy, or nutrient
agar plate) to ensure that the isolated colony is a pure culture.
Test one well-isolated colony via gram stain to confirm that
only gram-negative, non-spore-forming rods are present (Section
9225 C. Identification
1. Approaches
There are many ways to identify strains of Enterobacteriaceae
that are coliforms.1– 4 The most practical options for water lab-
oratories seem to be screening tests, biochemical “tube tests,”
validated commercial identification methods (“identification
kits”), and molecular methods. Each has advantages and disad-
vantages.
First, confirm that the presumptive coliform culture is a mem-
ber of the Enterobacteriaceae family. The dyes, bile salts, and
other ingredients in agars and broths are selective when isolating
bacteria. Ensure that the culture is gram-negative, oxidase-
negative, and facultative anaerobic. Doing so will eliminate
species that could be confused with Enterobacteriaceae, such as
Bacillus, Aeromonas, and Vibrio species that can ferment lactose
with gas production. The following approaches ensure quality
control (QC) of the media and testing conditions (e.g., use
known positive and negative culture controls).
2. Traditional Biochemical Tests, Computer Analysis, and
Screening Tests
Many of the most commonly used biochemical tests for dif-
ferentiating coliforms are shown in Table 9225:I. All microbi-
ology laboratories used to perform biochemical testing in tubes,
2
9221B). Also, determine that the culture is oxidase-negative, as
noted in Section 9222.
If the colony yields an unusual or “unidentified” result, a
contaminating organism may be present and the resulting bio-
chemical profile will not be useful.
2. Bibliography
PTAK, D.J., W. GINSBURG & B.F. WILLEY. 1974. Aeromonas, the great
masquerader. In Proc. AWWA Water Quality Technology Conf.,
Dec. 2–3, 1974, Dallas, Tex., p. V-1. American Water Works
Assoc., Denver, Colo.
VAN DER KOOJ, D. 1988. Properties of aeromonads and their occurrence
and hygienic significance in drinking water. Zentralbl. Bacteriol.
Hyg. B 187:1.
HARTMAN, P.A., B. SWAMINATHAN, M.S. CURIALE, R. FIRSTENBERG-EDEN,
A.N. SHARPE, N.A. COX, D.Y.C. FUNG & M.C. GOLDSCHMIDT. 1992.
Rapid methods and automation. In C. Vanderzant & D.F.
Splittstoesser, eds., Compendium of Methods for the Microbiolog-
ical Examination of Foods, 3rd. ed., p. 665. American Public Health
Assoc., Washington, D.C.
STAGER, C.E. & J.R. DAVIS. 1992. Automated systems for identification
of microorganisms. Clin. Microbiol. Rev. 5:302.
RICE, E.W., M.J. ALLEN, T.C. COVERT, J. LANGEWIS & J. STANDRIDGE.
1993. Identifying Escherichia species with biochemical test kits
and standard bacteriological tests. J. Amer. Water Works Assoc.
85(2):74.
and many reference and public health laboratories still do.2
Although some laboratories prepare their own media from com-
mercial dehydrated powders, most of the commonly used media
are available in commercially prepared, ready-to-inoculate glass
tubes. Such tubes are recommended for QC purposes. The test
typically involves inoculating each tube with one colony’s
growth, incubating it at 35°C, and reading the results at 24 and
48 h. In many reference laboratories, analysts often keep the
tubes for 7 d to detect delayed reactions. Unfortunately, the
media and tests are not completely standardized, and few labo-
ratories use the exact same formulations or procedures. The
results for 68 “potential coliform bacteria” of Enterobacteri-
aceae in 47 tests can be found in the literature.2
Analysts then compare the culture’s biochemical reactions
results to charts or tables1–5—such as Table 9260:I—to deter-
mine which Enterobacteriaceae species are compatible with the
results.
The next step could be a computer software and database
analyses of the coliform culture’s biochemical profile. Computer
analysis can be a useful supplement to manual analysis because
of the “potential coliform species” in the Enterobacteriaceae
family. PIBWin* is a free identification program that can be
* See http://www.som.soton.ac.uk/research/sites/pibwin/. Accessed September
2011.
 DIFFERENTIATION OF COLIFORM BACTERIA (9225)/Identification

TABLE 9225:I. BIOCHEMICAL REACTIONS OF SEVERAL SPECIES OF THE FAMILY ENTEROBACTERIACEAE*

Percentage Isolates Positive in 1 to 2 Days†

Ornithine Decarboxylase

Cellobiose Fermentation

Yellow Pigment, 25oC

Lysine Decarboxylase

Sorbitol Fermentation
Lactose Fermentation

Simmons’s Citrate
ONPG Hydrolysis

Motility, 35–37°C
Indole Production

Voges-Proskauer
Methyl Red
Species
Citrobacter diversus 35 96 99 100 0 99 99 0 99 99 95 0
Citrobacter freundii 50 95 5 100 0 95 20 0 98 55 95 0
Enterobacter aerogenes 95 100 0 5 98 95 98 98 100 100 97 0
Enterobacter agglomerans 40 90 20 50 70 50 0 0 30 55 85 75
Enterobacter cloacae 93 99 0 5 100 100 96 0 95 99 95 0
Escherichia coli 95 95 98 99 0 1 65 90 94 2 95 0
Escherichia coli, variant 25 45 80 95 0 1 20 40 75 2 5 0
Escherichia fergusonii 0 83 98 100 0 17 100 95 0 96 93 0
Escherichia hermannii 45 98 99 100 0 1 100 6 0 97 99 98
Escherichia vulneris 15 100 0 100 0 0 0 85 1 100 100 50
Hafnia alvei 5 90 0 40 85 10 98 100 0 15 85 0
Klebsiella oxytoca 100 100 99 20 95 95 0 99 99 100 0 1
Klebsiella ozaenae 30 80 0 98 0 30 3 40 65 92 0 0
Klebsiella pneumoniae 98 99 0 10 98 98 0 98 99 98 0 0
Klebsiella rhinoscleromatis 0 0 0 100 0 0 0 0 100 100 0 0
Serratia fonticola 97 100 0 100 9 91 97 100 100 6 91 0
Serratia marcescens 2 95 1 20 98 98 99 99 99 5 97 0
* Modified after Farmer, J.J. III, 1985. Clinical identification of new species and biogroups of Enterobacteriaceae. J. Clin. Microbiol. 21:46. This list is not all inclusive.
† Reactions that become positive after 2 d are not considered.

downloaded and used with the “Colindale data matrix of Entero- preting results (sometimes read via machine with simultaneous
bacteriaceae”6 (a current option in the program). Although the computer analysis). Many of these miniaturized biochemical
matrix is now outdated, it still contains the most common coli- tests are AOAC International-validated. Most clinical and food
form organisms found in water. The Centers for Disease Control microbiology laboratories now use miniaturized identification
and Prevention’s (CDC’s) Enterobacteriaceae data matrix, kits to identify common Enterobacteriaceae cultures, including
which includes the more recently described species, will be coliform cultures isolated from water, but these assays may be
added to PIBWin in the future. less accurate for newly described species. Check the instruction
Screening tests can differentiate coliforms or confirm an iden- manual to determine which organisms are included in the data-
tification made via a commercial identification system or molec- base and the number of strains used to define each organism.
ular method. Table 9225:I lists useful screening tests, key reac- The main problem with identification kits is that the test panels
tions, and the properties of some of the most common coliform used (usually about 20) are insufficient to differentiate all of the
bacteria. Additional Enterobacteriaceae screening tests that may relevant Enterobacteriaceae species. (This is also becoming a
be useful for differentiating coliforms are available and de- problem with conventional tube tests, even when 40 to 50 tests
scribed elsewhere.2 are done.) Unusual or “unidentified” results may be tested with
another kit, which may have similar limitations. Alternatively,
3. Commercial Media, Reagents, and Test Kits the culture may be sent to a reference laboratory for identifica-
tion using molecular tests or 16S rRNA sequencing.
There are many commercial products (e.g., dehydrated media, When using identification kits to differentiate coliforms, keep
ready-to-use media, and reagents) and commercial identification in mind that their databases may not include some coliform
products (identification kits) for differentiating Enterobacteri- bacteria likely to occur in water. These kits and databases were
aceae and similar bacteria.7,8 Identification kits typically consist originally developed for use in clinical microbiology, hence may
of a plastic strip that contains a set of miniaturized biochemical not include all the coliform bacteria isolated from water.8 Note
tests, often those traditionally used for identification (indole that some semi-automated and automated identification kits are
production, Voges-Proskauer, citrate utilization, lactose fermen- now being developed for environmental samples.†
tation, etc.). The approach is similar to biochemical testing in
glass tubes, with the main differences being in miniaturization,
simplicity, ease of use, speed of analysis, number of tests avail- † Vitek, bioMérieux Inc., Durham, NC; MicroLog Micro station, Biolog, Bremen,
able, suspending medium, and the method of reading and inter- Germany; or equivalent.

3
 DIFFERENTIATION OF COLIFORM BACTERIA (9225)/Media, Reagents, and Procedures
4. Molecular Methods
Molecular methods can be used for taxonomic identification of
a coliform, including family, genus, species, serotype, clone, and
strain and even potential virulence. These methods are imprac-
tical for routine coliform differentiation, but can be extremely
useful in research projects, epidemiological studies, or when a
definitive identification is required.
One of the most accurate methods for identifying a coliform
isolate is a complete or partial 16S rRNA gene sequence that can
be done by fee-for-service laboratories, which will provide stra-
tegic sequence data.9 Such labs often can identify a coliform
species based on the sequence of the first 500 base pairs in its
16S rRNA gene. A complete 16S rRNA gene sequence is more
accurate, but much more expensive. Typically, the commercial
lab will process the culture, sequence its genes, and put together
an identification report with technical details and a figure show-
ing the position of the submitted culture in relation to reference
strains in the database. The per-test cost may depend on the total
number of strains submitted and the “turnaround time” re-
quested.
The advantage of 16S rRNA gene sequencing is that it con-
siders all Enterobacteriaceae, not just those found in human
clinical species. When a definitive identification is required (e.g.,
in a judicial proceeding), this may be the best option.
5. References
1. BRENNER, D.J. & J.J. FARMER, III. 2005. Family 1. Enterobacteri-
aceae. In Bergey’s Manual of Systematic Bacteriology, 2nd ed., Vol.
2: The Proteobacteria, Part B: The Gammaproteobacteria. Springer,
Secaucus, N.J.
2. FARMER, J.J. III, K.D. BOATWRIGHT & J.M. JANDA. 2007. Enterobac-
teriaceae: Introduction and Identification. In Manual of Clinical
9225 D. Media, Reagents, and Procedures
In routine testing of drinking and waste waters, identification
down to the species or strain is not a common practice. However,
many laboratories are concerned with bacterial identification.
There is no “standard method” for identifying Enterobacteri-
aceae (coliform differentiation); all of the existing identification
methods simply evolved over 100 years. Likewise, the commer-
cial media and reagents available are also changing or being
modified, so keep the following in mind when choosing media,
reagents, and procedures.
If identifying coliform cultures via biochemical tests in tubes,
remember that there are many variables involved. When possi-
ble, use the media and methods that accompany the chosen
“biochemical data matrix” or “identification schema.” For ex-
ample, one reference1 notes the reactions of more than 10 000
strains of Enterobacteriaceae tested at CDC in specific media
using specific test conditions described in a second reference.2
On the other hand, the Enterobacteriaceae biochemical data
matrix generated by England’s national reference laboratories at
4
Microbiology, 9th ed., Chapter 42. Amer. Soc. Microbiol. Press,
Washington, D.C.
3. FARMER, J.J. III, M.K. FARMER & B. HOLMES. 2005. The Enterobac-
teriaceae: General Characteristics. In Topley & Wilson’s Microbi-
ology and Microbial Infections, 10th ed., Vol. 2: Bacteriology, Chap-
ter 51. Hodder Arnold, London, U.K.
4. JANDA, J.M. & S.L. ABBOTT. 2006. The Enterobacteria, 2nd ed. Amer.
Soc. Microbiol. Press, Washington, DC.
5. FARMER, J.J. III, M.A. ASBURY, E.W. HICKMAN, D.J. BRENNER & THE
ENTEROBACTERIACEAE STUDY GROUP. 1980. Enterobacter sakazakii: A
new species of “Enterobacteriaceae” isolated from clinical speci-
mens. Int. J. Syst. Bacteriol. 30:569.
6. HOLMES, B. & M. COSTAS. 1992. Identification of Enterobacteriaceae
by computerized methods. In Identification Methods in Applied and
Environmental Microbiology. Blackwell Scientific Publ., Oxford,
England.
7. O’HARA, C.M. 2005. Manual and automated instrumentation for
identification of Enterobacteriaceae and other aerobic gram-negative
bacilli. Clin. Microbiol. Rev. 18:147.
8. CARROLL, K.C. & M.P. WEINSTEIN. 2007. Manual and automated
systems for the detection and identification of microorganisms. In
Manual of Clinical Microbiology, 9th ed., Chapter 15. Amer. Soc.
Microbiol., Washington, DC.
9. WOO, P.C., K.H. NG, S.K. LAU, K.T. YIP, A.M. FUNG, K.W. LEUNG,
D.M. TAM, T.L. QUE & K.Y. YUEN. 2003. Usefulness of the MicroSeq
500 16S ribosomal DNA-based bacterial identification system for
identification of clinically significant bacterial isolates with ambigu-
ous biochemical profiles. J. Clin. Microbiol. 41:1996.
6. Bibliography
IVERSEN, C., L. LANCASHIRE, M. WADDINGTON, S. FORSYTHE & G. BALL.
2006. Identification of Enterobacter sakazakii from closely related
species: the use of artificial neural networks in the analysis of
biochemical and 16sDNA data. BMC Microbiol. 6:28.
Colindale3,4 is based on a different set of media and methods (the
Colindale methods).
The following are traditionally used media and methods; how-
ever, there is no biochemical data matrix or “coliform identifi-
cation schema” that accompanies them. Use commercial media
and reagents to improve media consistency, reduce labor costs,
and increase quality. Include negative and positive controls using
known stock cultures to ensure continued accuracy and reliabil-
ity. Follow the QC guidance in Section 9020B for media prep-
aration, handling, and storage conditions.
1. Lactose, Sorbitol, and Cellobiose Fermentation Tests
Suspend 16 g phenol red broth base and 5 g of the selected
carbohydrate (lactose, sorbitol, or cellabiose) in 1 L reagent-
grade water, and stir to dissolve completely. Dispense in tubes,
filling each tube to one-third its length. Then place a small
inverted vial (Durham tube) in the tube to determine gas pro-
 DIFFERENTIATION OF COLIFORM BACTERIA (9225)/Media, Reagents, and Procedures
duction. Close tubes and sterilize at 121°C for 12 to 15 min.
Store tubes in the dark (refrigeration preferred), and discard if
medium becomes discolored or evaporation exceeds 10% of
volume.
To conduct a test, inoculate a tube with a loopful of growth
from a well-isolated colony or slant and incubate for 24 to 48 h
at 35 ⫾ 0.5°C. Carbohydrate fermentation (acid production) will
lower pH, causing the phenol red pH indicator to change from
red-orange to yellow (pH ⬍6.6). Alternatively, for lactose fer-
mentation, lauryl tryptose broth (Section 9221B.2a) may be
used.
2. Ortho-nitrophenyl-␤-D-galactopyranoside (ONPG)
Hydrolysis
This test looks for the ␤-galactopyranosidase enzyme by hy-
drolyzing the substrate o-nitrophenyl-␤-D-galactopyranoside.
Use one of the numerous commercial test kits and disks avail-
able, or an ONPG-containing medium (Section 9223).
To conduct the test, inoculate 10 mL ONPG broth with a
heavy loopful of growth from a slant and incubate at 35 ⫾ 0.5°C
for up to 24 h. If the medium develops a yellow color—
compared to an uninoculated tube or (preferably) a tube inocu-
lated with an ONPG-negative culture—results are positive. In-
terpret tests of yellow-pigmented organisms with caution.
3. Other Chromogenic Assays for ␤-galactopyranosidase
The chlorophenol red-␤-D-galactopyranoside (CPRG) hydro-
lysis test is similar to the ONPG test, except the color of a
positive result is red to magenta. Similarly, X-Gal (5-bromo-4-
chloro-3-indolyl-␤-D-galactopyranodise) is hydrolyzed by ␤-ga-
lactopyranosidase to yield a blue product.
4. Indole Test
Indole is a product of tryptophan metabolism.
a. Reagents:
1) Medium: Use tryptophan broth. Dissolve 10.0 g tryptone or
trypticase per L reagent-grade water. Dispense medium in 5-mL
portions in test tubes, and sterilize.
2) Kovac’s reagent: Dissolve 5 g p-dimethylaminobenzalde-
hyde in 75 mL isoamyl (or normal amyl) alcohol, ACS grade,
and add 25 mL conc HCl. CAUTION: Reagents are both flam-
mable and corrosive. The reagent should be yellow.
The amyl alcohol solution’s pH should be less than 6.0.
Purchase both amyl alcohol and benzaldehyde in amounts as
small as consistent with the volume of work to be done.
b. Procedure: Inoculate 5-mL portions of tryptophan broth
medium from a pure culture and incubate at 35 ⫾ 0.5°C for 24
⫾ 2 h. Add 0.2 to 0.3 mL Kovac’s test reagent and gently shake.
Let stand for about 10 min, and observe results.
Results are positive if the amyl alcohol surface layer is a dark
red color; results are negative if the layer remains yellow. An
orange color probably indicates the presence of skatole (a break-
down product of tryptophan), but it is not indole. Record as a
variable result.
5
5. Methyl Red Test
The methyl red test measures organisms’ ability to produce
stable acid end products via glucose fermentation.
a. Reagents:
1) Medium: Use buffered glucose broth. Dissolve 7.0 g pro-
teose peptone or equivalent peptone, 5.0 g glucose, and 5.0 g
dipotassium hydrogen phosphate (K2HPO4) in 1 L reagent-grade
water. Dispense in 5-mL portions in test tubes, and sterilize by
autoclaving at 121°C for 12 to 15 min, making sure that total
heat-exposure time is no longer than 30 min.
2) Indicator solution: Dissolve 0.1 g methyl red in 300 mL
95% ethyl alcohol and dilute to 500 mL with reagent-grade
water. CAUTION: Ethyl alcohol is flammable.
b. Procedure: Inoculate 10-mL portions of medium from a
pure culture. Incubate at 35 ⫾ 0.5°C for 5 d. Then, add 5 drops
methyl red indicator solution to 5 mL of the culture and read
immediately.
A 48-h incubation period is adequate for most cultures, but do
not incubate for less than 48 h. If test results are equivocal at
48 h, repeat with cultures incubated for 4 or 5 d. In such cases,
incubate duplicate cultures at 22 to 25°C. Testing culture por-
tions at 2, 3, 4, and 5 d may provide positive results sooner.
Results are positive if the sample is distinctly red, negative if
it is distinctly yellow, and questionable if the color is a mixed
shade.
6. Voges-Proskauer Test
The Voges-Proskauer test measures organisms’ ability to pro-
duce a neutral product (acetoin or acetyl methylcarbinol) via
glucose fermentation.
a. Reagents:
1) Medium: See 9225D.4a1).
2) Naphthol solution: Dissolve 5 g purified ␣-naphthol (melt-
ing point 92.5°C or higher) in 100 mL absolute ethyl alcohol.
When stored at 5 to 10°C, this solution is stable for 2 weeks.
CAUTION: Ethyl alcohol is flammable.
3) Potassium hydroxide, 7N: Dissolve 40 g KOH in 100 mL
reagent-grade water.
b. Procedure: Inoculate 5 mL medium and incubate for 48 h
at 35 ⫾ 0.5°C. To 1 mL of culture, add 0.6 mL naphthol solution
and 0.2 mL KOH solution. Shake well after adding each reagent.
Results are positive if a pink to crimson color develops at the
surface within 5 min. Do not read after 10 min. Disregard tubes
developing a copper color as this represents a negative result.
7. Simmons’ Citrate Test
The citrate test measures a bacterium’s ability to use citrate as
the sole carbon source, using bromothymol blue as the pH
indicator.
a. Medium: Use Simmons’ citrate agar. To make Simmons’
citrate agar, add 0.2 g MgSO4 䡠 7H2O, 1.0 g ammonium dihy-
drogen phosphate (NH4H2PO4), 1.0 g K2HPO4, 2.0 g sodium
citrate dihydrate, 5.0 g NaCl, 15.0 g agar, and 0.08 g bromthy-
mol blue to 1 L reagent-grade water. Autoclave at 121°C for 12
to 15 minutes and cool tubes at an angle to form an agar slant.
Final pH after sterilization is 6.8 ⫾ 0.2.
 DIFFERENTIATION OF COLIFORM BACTERIA (9225)/Media, Reagents, and Procedures
b. Procedure: Inoculate agar medium via the streak technique
using a light inoculum.
Incubate 48 h at 35 ⫾ 0.5°C. Uninoculated medium is green in
color. Results are positive when the medium develops a blue
color; this indicates the metabolism of citrate and production of
alkaline byproducts. Results are negative if there is no color
change.
8. Motility Test
The motility test measures whether an organism is motile in a
semi-solid medium.
a. Medium: Add 3.0 g beef extract, 10.0 g peptone, 5.0 g NaCl,
and 4.0 g agar to 1 L reagent-grade water. Adjust pH to 7.4,
dispense in 3-mL portions in 13- ⫻ 100-mm tubes or 8-mL
portions in 16- ⫻ 125-mm tubes, and sterilize.
b. Procedure: Inoculate by stabbing an inoculating needle
5 mm deep into the center of the medium. Incubate for 1 to 2 d at
35°C. If results are negative, incubate another 5 d at 22 to 25°C.
Results are positive if there is diffuse growth through the
medium from the inoculation point. Results are negative if
growth is visible only along the stab line, and the surrounding
medium remains clear.
Alternatively, prepare the medium without agar and examine
a young culture via the hanging drop slide technique for motile
organisms. Examine the wet-mount slide immediately; do not let
it dry.
9. Lysine and Ornithine Decarboxylase Tests
This procedure tests bacteria’s ability to metabolize (decar-
boxylate) the amino acids lysine and ornithine. Use of commer-
cially prepared medium is preferred for QC purposes.
a. Reagents:
1) Media: Use a basal medium made according to the Moeller
or Falkow methods. The Moeller method involves dissolving
5.0 g peptone (Orthana special, thiotone, or equivalent), 5.0 g
beef extract, 0.625 mL bromcresol purple (l.6%), 2.5 mL cresol
red (0.2%), 0.5 g glucose, and 5.0 mg pyridoxal in 1 L reagent-
grade water and adjusting to pH 6.0 to 6.5. The Falkow method
involves dissolving 5.0 g peptone, 3.0 g yeast extract, 1.0 g
glucose, and 1.0 mL bromcresol purple (1.6%) in 1 L reagent-
grade water and adjusting to pH 6.7 to 6.8.
For either decarboxylase test, divide into three portions: make
no addition to the first portion (the control), add enough L-lysine
dihydrochloride to the second portion to make a 1% solution, and
add enough L-ornithine dihydrochloride to the third to make a
1% solution (for the Falkow method, add only 0.5% of the
L-amino acid). After adding ornithine, readjust the medium’s pH
to 6.0 ⫾ 0.2. Dispense in 3- to 4-mL portions in screw-capped
test tubes and sterilize via autoclaving at 121°C for 10 min. A
floccular precipitate in the ornithine medium does not interfere
with its use.
2) Mineral oil: Use mineral oil sterilized via autoclaving at
121°C for 30 to 60 min, depending on container size.
b. Procedure: Lightly inoculate each medium, add an approx-
imately 10-mm-thick layer of mineral oil to promote fermenta-
tion, tighten caps, and incubate at 35 to 37°C for up to 4 d.
Examine tubes daily. Results are positive if the color changes
6
from yellow to violet or reddish-violet, weakly positive if the
color changes to bluish gray, and negative if the color remains
yellow (unchanged). For a weakly positive culture, a drop of
bromocresol purple can be added. If the color does not change,
this is considered a negative test.
10. Oxidase Test
The oxidase test determines the presence of oxidase enzymes.
Coliform bacteria are oxidase-negative.
a. Reagents:
1) Media: Use either nutrient agar or tryptic soy agar plates to
streak cultures and produce isolated colonies. From these, obtain
the inoculum for oxidase testing on impregnated filter paper.
Cultures should be less than 24 h old. Do not use any medium
that includes a carbohydrate in its formulation. Use only tryptic
soy agar if reagent is dropped on colonies.
Tryptic soy agar:
Tryptone......................................................................................15.0 g
Soytone .........................................................................................5.0 g
Sodium chloride, NaCl.................................................................5.0 g
Agar ............................................................................................15.0 g
Reagent-grade water.....................................................................1.0 L
pH should be 7.3 ⫾ 0.2 after sterilization.
2) Tetramethyl p-phenylenediamine dihydrochloride, 1%
aqueous solution, freshly prepared or refrigerated for no longer
than 1 week. Impregnate a filter paper strip* with this solution.
Alternatively, prepare a 1% solution of dimethyl p-phenylene-
diamine hydrochloride. Commercially available, single-use re-
agent ampules are convenient and economical, but use them with
caution. This reagent is also available in little dropper bottles, with
which analysts can add 1 to 2 drops to filter paper. Use tryptic soy
agar when the reagent will be dropped directly on colonies, because
nutrient agar plates give inconsistent results; when smearing a
portion of a picked colony on reagent-impregnated filter paper, do
not transfer any medium with the culture material.
b. Procedure: Remove some of a colony from agar plate with a
platinum wire, a wooden or plastic applicator stick, or a glass rod,
and smear on the test strip. Do not use iron or other reactive wire
because it will cause false positive reactions. Results are positive if
a dark purple color develops within 10 s. Test positive and negative
cultures concurrently. If liquid reagent is used, drop it on colonies
on the culture plate. Oxidase-positive colonies develop a pink color
that successively becomes maroon, dark red, and finally, black.
11. Yellow Pigment
Observe isolated colonies on tryptic soy agar slants or plates
(nutrient agar may be an acceptable substitute) incubated at 25 ⫾
0.5°C for up to 48 h. Pigmentation often intensifies as incubation
time proceeds.
12. References
1. FARMER, J.J., III, K.D. BOATWRIGHT & J.M. JANDA. 2007. Enterobac-
teriaceae: Introduction and Identification. In Manual of Clinical
* Whatman No. 1 or equivalent.
 DIFFERENTIATION OF COLIFORM BACTERIA (9225)/Reporting Results
Microbiology, 9th ed., Chapter 42. Amer. Soc. Microbiol. Press,
Washington, DC.
2. FARMER, J.J., III, M.A. ASBURY, E.W. HICKMAN, D.J. BRENNER & THE
ENTEROBACTERIACEAE STUDY GROUP. 1980. Enterobacter sakazakii: A
new species of “Enterobacteriaceae” isolated from clinical speci-
mens. Int. J. Syst. Bacteriol. 30:569.
3. LAPAGE, S.P., S. BASCOMB, W.R. WILLCOX & M.A. CURTIS. 1973.
Identification of bacteria by computer: general aspects and perspec-
tives. J. Gen. Microbiol. 77:273.
4. HOLMES, B. & M. COSTAS. 1992. Identification of Enterobacteriaceae
by computerized methods. In Identification Methods in Applied and
Environmental Microbiology. Blackwell Scientific Publ., Oxford,
England.
9225 E. Reporting Results
Precisely worded reports are essential because of the identifi-
cation methods’ limitations. Specify the method(s) used and
provide comments. Note that careful wording of results will not
help an unqualified test. Below are some possibilities, with
less-precise identifications given first.
1. Identification: Klebsiella–Raoultella complex
Klebsiella sp. identification is difficult. The two species of
Raoultella were formerly identified as Klebsiella sp.
Comment: This culture gave a “possible identification” based
on its biochemical test profile determined by the XYZ commer-
cial identification system (47 biochemical tests). Further testing
would be required for a definitive identification; please contact
the laboratory if this is needed.
2. Identification: Klebsiella species
Comment: Identification was based on the following screen-
ing-test results:
• grew as large, pink, mucoid colonies on MacConkey agar
plus 30 ␮g per mL ampicillin;
• was a gram-negative rod with large capsules around the cells;
and
• was non-motile, lysine decarboxylase-positive, arginine di-
hydrolase-negative, and ornithine decarboxylase-negative.
13. Bibliography
O’HARA, C.M. 2005. Manual and automated instrumentation for iden-
tification of Enterobacteriaceae and other aerobic Gram-negative
bacilli. Clin. Microbiol. Rev. 18:147.
HEALTH PROTECTION AGENCY. 2007. Identification of Enterobacteriaceae.
National Standard Method BSOP ID 16:2.
MURRAY, P.R., ed. 2007. Manual of Clinical Microbiology. Amer. Soc.
Microbiol. Press, Washington, D.C.
HEALTH PROTECTION AGENCY. 2010. Identification of Enterobacteriaceae,
National Standard Method BSOP ID 16:3. http://www.hpastandard
methods.org.uk/documents/bsopid/pdf/bsopid16.pdf. Accessed
September 2011.
These are all typical and defining characteristics of Klebsiella
species.
3. Identification: Klebsiella oxytoca
Comment: “High likelihood identification” based on biochem-
ical test profile determined by the ABC commercial identifica-
tion system (21 biochemical tests).
NOTE: This system is not sensitive and specific for differenti-
ating all species in the genus Klebsiella.
4. Identification: Klebsiella varicola
Comment: This was a “probable identification” based on a
biochemical test profile determined via the CDE commercial
identification system (95 biochemical tests). It was later con-
firmed by a partial 16S rRNA gene sequence (451 base) done by
a commercial laboratory (report attached).
5. Identification: Klebsiella terrigena
Comment: This identification is based on a PCR test our
laboratory developed based on a sequence of the rpoB gene. This
research test is sensitive, specific for Klebsiella terrigena, and
differentiates other Klebsiella species. These methods and results
will be published. Technical details are attached to this report.