Evaluation of a Pectin Agar Medium for Isolation of
Yersinia Enterocolitica within 48 Hours
J. HARTLEY BOWEN, M.D., AND SPYROS D. KOMINOS, Sc.D.
Bowen, J. Hartley, and Kominos, Spyros D.: Evaluation of a
pectin agar medium for isolation of Yersinia enterocolitica
within 48 hours., Am J Clin Pathol 72: 586-590, 1979. A
modified pectin agar medium was evaluated for the rapid isolation and presumptive identification of Yersinia enterocolitica. Of
118 isolates of Enterobacteriaceae tested, only the 13 Y. enterocolitica and the three Klebsiella oxytoca strains produced colonies that depressed and sank into the agar. Yersinia enterocolitica was also easily identified in mixed cultures, even from
inocula containing three times as many other Enterobacteriaceae organisms as Y. enterocolitica. The recovery of Y.
enterocolitica was evaluated on Mueller-Hinton, pectin,
Hektoen enteric, xylose lysine desoxycholate, SalmonellaShigella, and MacConkey agars. Compared with MuellerHinton agar, the pectin agar showed a 100% recovery of Y.
enterocolitica, with all strains having depressed colonies, while
the other media showed lesser recoveries of only 5 to 25%,
with no other discriminating colonial characteristic. (Key
words: Pectin agar; Yersinia enterocolitica; Enterobacteriaceae.)
YERSINIA
ENTEROCOLITICA
, a member of the Entero-
bacteriaceae, is being increasingly implicated as an
etiologic agent in disease. Although 11 cases of human
infection with this organism occurred in the United
States between 1933 and 1947,'7 the next case report
was not until 1968.' 9 Since then, worldwide reports of
Y. enterocolitica infection have included an interfamilial,8 a hospital, 22 a ski resort, 10 and two school
outbreaks. 2 2 9
The clinical manifestation of Y. enterocolitica infection in children is an acute enteritis, with fever, malaise,
headache, crampy diffuse abdominal pain, and nonbloody diarrhea. Adults have similar symptoms, but
diarrhea is less common. Several patients operated on
for appendicitis have had normal appendices with terminal ileitis or mesenteric lymphadenitis and positive
Y. enterocolitica cultures. 2 - 6 ' 81215,25 This organism has
also been implicated in a pulmonary abscess,' 8 hepatic
and splenic abscesses, 81314 small-intestinal abscesses, 11
septic arthritis, 20 inflamed mesenteric 3 - 25 and inguinal28
lymph nodes, and Parinaud's oculoglandular syndrome. 5 The mortality rate is as high as 50% in Y.
enterocolitica septicemia. 13
Received April 24, 1978; Received revised manuscript and accepted for publication August 21, 1978.
Dr. Bowen's present address is: Duke University Medical Center,
Department of Pathology, Durham, North Carolina 27710.
Address reprint requests to Dr. Kominos: Department of Pathology, Division of Microbiology, Mercy Hospital, Pittsburgh,
Pennsylvania 15219.
Division of Microbiology, Department of Pathology,
Mercy Hospital, Pittsburgh, Pennsylvania
Pure cultures of Y. enterocolitica are rare, and the
organism is usually recovered from mixed flora from
human, 24 animal,4-23-28 and environmental 910 specimens. Previously used cold-temperature enrichment
isolation technics 716 are not very beneficial, because
the enrichment period is long, requiring sometimes
several weeks. Considering the increased reports of
Y. enterocolitica disease, a selective medium for the
isolation and presumptive identification of this organism in the clinical laboratory is highly desirable.
Recently, von Riesen demonstrated that species of
Yersinia 27 and some strains of Klebsiella pneumoniae26
had pectinolytic activity, a characteristic formerly believed confined to pathogenic plant bacteria. Starr and
associates 21 quantitated this pectinolytic activity and
proposed a pectin agar medium into which isolated
colonies would sink due to the lysis of pectin. Because
this previously proposed medium contained a low
quantity of agar, making it very soft and difficult to
streak, we have modified it by increasing the agar concentration and at the same time changed the indicator
to provide a suitable color for highlighting the depressed colonies. In these experiments, the modified
pectin agar was used to (1) test the strains of Y.
enterocolitica for colonial depression, (2) survey other
Enterobacteriaceae for pectinolytic activity, (3) determine the feasibility of isolating Y. enterocolitica from
mixed cultures, and (4) compare the recovery rate of
Y. enterocolitica with rates on other media commonly
used for the isolation of enteric pathogens.
Materials and Methods
Medium
Preparation
Five hundred milliliters of distilled water were placed
into a 1-liter Erlenmeyer flask on a stirring hot plate
(Fisher Scientific). Stirring with a magnetic bar and
heating to maximum of 85 C were begun and the following components were added in order: 3.0 ml of
10.0% aqueous CaCl2 solution; 2.5 g of yeast extract
0002-9173/79/1000/0586 $00.75 © American Society of Clinical Pathologists
586
PECTIN AGAR FORY. ENTEROCOLITICA
Vol. 72 . No. 4
(Difco Laboratories, Detroit, Michigan); 15.0 g of
pectin (polygalacturonic acid, sodium salt, Sigma
Chemical Co., St. Louis, Missouri, no. P-1879); 3.5 g
of Bacto-Agar (Difco Laboratories); 0.07 g of phenol
red. The pectin powder was added very slowly by introducing a small amount at a time with a spatula in
order to avoid gelatinous clumps, which, once formed,
are difficult to dissolve. The p H was adjusted to 7.3
with 1 N NaOH until the indicator changed to bright
red (usually requiring 5.0 ml). The p H must not exceed
7.3 since the sodium polypectate is labile at higherpH.
The suspension was then autoclaved at 121 C for 15
min, allowed to cool to 75 C, and poured into petri
dishes. (The medium must be dispensed while still hot,
because it begins to solidify at 50 C, and it cannot be
reliquefied.) Excess moisture was controlled by replacing wet petri dish covers with dry ones. The plates
were left at room temperature overnight, placed in
plastic sleeves, and stored at 4 - 8 C.
Bacteria Tested
The 13 strains of Y. enterocolitica were stock cultures obtained from the following sources: one strain
each from Allegheny General, St. Francis, Veterans
Administration, and West Penn Hospitals, all in Pittsburgh, Pennsylvania; one strain from Hamot Medical
Center, Erie, Pennsylvania; two strains from Mercy
Hospital, Pittsburgh; two strains from Roche Diagnostics, Nutley, New Jersey; four strains from the Center for Disease Control, Atlanta, Georgia. All strains
were motile at 25 C, nonmotile at 35 C, produced
urease and ornithine decarboxylase, lacked lysine decarboxylase, and did not utilize citrate. In triplesugar-iron agar the organisms gave an acid slant and
an acid butt by the fermentation of sucrose and glucose.
Lactose was not fermented. The four strains from the
CDC fermented raffinose, and three of these also fermented rhamnose. The remaining nine strains of Y.
enterocolitica did not ferment raffinose or rhamnose.
The other Enterobacteriaceae were clinical isolates recovered in the microbiology laboratory at Mercy
Hospital.
Pectinolytic
Survey
Each Y. enterocolitica and Enterobacteriaceae isolate was incubated overnight at 35 C in 2 ml of MuellerHinton broth. This was then diluted 1:1,000 in sterile
distilled water and inoculated on the pectin agar plates
by a 0.001 ml calibrated loop. The plates were incubated at 25 C and were observed for the depression
of colonies into the agar at 24 and 48 hours by the
authors and two technologists in microbiology who
were unfamiliar with the identities of the cultures.
587
Y. enterocolitica and Mixtures of
Enterobacteriaceae
One strain of Y. enterocolitica and seven nonpectinolytic isolates of Enterobacteriaceae were each cultured in 2 ml of Mueller-Hinton broth overnight at 35 C.
Each broth was then diluted 1:10,000 with sterile distilled water, and a calibrated loop was used to inoculate 0.001 ml from each dilution on a Mueller-Hinton
agar plate. By this procedure, the overnight MuellerHinton broth cultures were determined to have approximately 1 to 2 x 109 organisms per ml. An aliquot
of the single Y. enterocolitica dilution was combined
with an aliquot from each of the seven dilutions of
nonpectinolytic Enterobacteriaceae in a ratio of 1:3.
A calibrated loop was used to inoculate 0.001 ml from
each of the seven binary mixtures onto pectin agar
plates. All plates were incubated at 25 C for 48 hours,
and the depressed Y. enterocolitica and nondepressed
diluent colonies were counted on each of the pectin
agar plates.
Comparison of Media
Each of the 13 isolates of Y. enterocolitica was cultured overnight in 2 ml of Mueller-Hinton broth at 35 C.
A 1:10,000 dilution was made from each overnight
broth, and with a calibrated loop 0.001 ml of each dilution was cultured on Mueller-Hinton, xylose lysine
desoxycholate, Hektoen enteric, Salmonella-Shigella,
MacConkey (BBL, Cockeysville, Maryland) and the
pectin agar plates. The number of colonies per plate
was counted after 48 hours of incubation at 25 C.
Results
Colonial growth, but no colonial depression, was observed on the pectin agar plates after 24 hours incubation at 25 C, and the medium surrounding the colonies
had changed from orange to red. By 48 hours, the
pectinolytic colonies were 2 to 4 mm in diameter and
clearly depressed, as if sinking into the agar, as shown
in Figure 1. Incubating the plates at 35 C did not shorten
the time for the depressions to develop. The pectin
plates were observed daily for seven days, and the
pectinolytic colonies continued to sink deeper into the
agar, while the nonpectinolytic colonies never did show
depressions.
Table 1 lists results obtained with the 118 isolates
of Enterobacteriaceae tested on the pectin agar. All
isolates grew well. All 13 Y. enterocolitica and three
of the 21 Klebsiella pneumoniae isolates showed colonies that depressed the agar, while all the other isolates
grew in the usual fashion on the surface of the agar.
The three isolates of Klebsiella with depressed colonies
were indole-positive strains of the oxytoca group,
588
BOWEN AND KOMINOS
A.J.C.P. • October 1979
FIG. 1. Growth of pectinolytic
colonies of yersinia enterocolitica
and non-pectinolytic colonies of E.
coli (for comparison) after 48 hours
at 25 C. xl.5.
which have previously been shown to be pectinolytic 26
and are now considered to belong to a separate species
—Klebsiella oxytoca.4 The 13 isolates of Serratia
marcescens swarmed on the surface of the agar, preventing individual colony identification; nevertheless,
they did not show any depression of the agar. None of
the Proteus isolates swarmed.
Table 2 shows the results of observations of seven
binary mixtures of Y. enterocolitica suspended with
another nonpectinolytic Enterobacteriaceae strain.
The number of Y. enterocolitica organisms in relation
to the diluent strain inoculated on the pectin agar was
approximately 1:3, or 25% of the total inoculum. After
48 hours of incubation at 25 C, the number of depressed
Y. enterocolitica colonies was easily counted in relation to the nonpectinolytic diluent members of the
Enterobacteriaceae. Depressed colonies of Y. enterocolitica were even observed through the swarming of
S. marcescens. The counts of Y. enterocolitica colonies
exceeded the expected 25% of the total in four mixtures. In the remaining three mixtures, the counts were
satisfactory to aid in easy recovery of Y. enterocolitica
from the mixed cultures. Incubation at 35 C did not
accelerate the growth of the Y. enterocolitica colonies,
but rather obscured them, due to overgrowth by the
diluent Enterobacteriaceae.
A comparison of the recovery of Y. enterocolitica
by pectin agar and other commonly used media for
isolating enteric pathogens is shown in Table 3. Defining growth as at least one colony per plate, all 13
isolates of Y. enterocolitica grew well on the MuellerHinton and pectin agars, while on Hektoen enteric,
xylose-lysine desoxycholate, and Salmonella-Shigella
agars, growth was poor. Only one strain failed to grow
on the MacConkey agar. All colonies were depressed
into the pectin agar, while the colonies of the strains
that did grow on Hektoen enteric and xylose-lysine
desoxycholate agars appeared lactose-positive due to
the fermentation of sucrose present in these media.
The colonies on the Salmonella-Shigella plate were 1
mm or less in diameter, and the colonies on the
MacConkey agar were 1 to 2 mm in diameter, although
on both media they clearly appeared lactose-negative.
Using the Mueller-Hinton agar plate as a control,
the efficacy of recovering Y. enterocolitica, as indicated by the average of the colony counts from the
13 strains, is also shown in Table 3. Compared with
122 colonies on the Mueller-Hinton agar plate, the numbers of colonies from the same quantity of inoculum
(0.001 ml) ranged from six on the Hektoen enteric agar
to 124 on the pectin agar. Thus, 100% recovery was
obtained on the pectin agar plate, while the other media
showed only 5 to 25% recoveries.
Discussion
There is a definite need for a medium to aid the
clinical microbiologist in the isolation of Y. enterocolitica from cultures with mixed flora, such as those
from stool specimens. This modified pectin agar medium is very useful, since the pectinolytic colonies of
Y. enterocolitica have the distinctive property of sinking into the agar and appearing depressed. This property of sinking into the agar enables the microbiologist
Vol. 72 • No. 4
PECTIN AGAR FOR Y. ENTEROCOLITICA
presumptively to differentiate Y. enterocolitica from
other common Enterobacteriaceae. The depressed
colonies may then be confirmed as Y. enterocolitica
by the expected biochemical tests, such as production
of urease and ornithine decarboxylase, lack of production of lysine decarboxylase, nonutilization of
citrate, nonmotility at 35 C, and motility at 25 C.
The pectin agar medium described here is a modification of the one proposed by Starr and co-workers21 for
the detection of pectinolytic bacteria. By increasing
the agar content from 0.3 to 0.7%, the medium is
more solidified to facilitate streaking. In addition, the
indicator phenol red replaces bromothymol blue in
order to provide a color contrast for enhancing the
depressions formed by the colonies of Y. enterocolitca
as they lyse pectin and sink into the medium.
All strains of Y. enterocolitica tested thus far clearly
showed depressed colonies, which were easily recognized in either pure or mixed cultures even by laboratory personnel unfamiliar with the identity of the cultures. Also, as shown from the data, Y. enterocolitica
may be isolated from mixed cultures even when there
are three times as many nonpectinolytic Enterobacteriaceae as Y. enterocolitica in the inoculum. From
all the other Enterobacteriaceae tested, only three isolates of Klebsiella oxytoca showed depressed colonies.
These were easily differentiated from Y. enterocolitica
by utilizing citrate, lacking ornithine decarboxylase,
producing lysine decarboxylase, and being nonmotile.
Although all the isolates of 5. marcescens tested
were swarming on the pectin agar, this did not interfere with the observation of the depressed colonies of
Y. enterocolitica in mixed cultures (Table 2). We offer
no explanation as to the cause of swarming by 5.
marcescens and the absence of swarming by all the
species of Proteus tested in the pectin agar.
Stool specimens for the recovery of enteric patho-
589
Table 2. Isolation of Yersinia enterocolitica from 1:3
Mixtures of Some Common Enterobacteriaceae
by the Pectin Agar*
Diluent Bacteria
No. of Colonies
with Depressiont
Escherichia coli
Enterobacter cloacae
Klebsiella pneumoniae
Citrohacter freundii
Proteus mirabilis
Serratia marcescens
Providencia stuartii
No. of Colonies
with No Depression
30
25
12
20
20
It
8
50
50
23
95
40
—§
60
* Cultures observed at 48 h at 25 C.
+ Isolates were confirmed as Y. enterocolitica.
t Depressed colonies observed through swarming of 5. marcescens.
§ No isolated colonies due to swarming.
Table 3. Recoveries of 13 Isolates of Yersinia
enterocolitica on Pectin Agar and on Other
Commonly Used Media
Medium
No. of Isolates
Recovered*
No. of Colonies
per Platet
Mueller-Hinton
Pectin agar
Hektoen enteric
Xylose-lysine desoxycholate
Salmonella-Shigella
MacConkey
13
13
5
11
7
12
122
124(100)!
6 (5)
18(15)
9 (7)
30(25)
* At least one colony per plate.
t Average number of colonies from 13 isolates tested.
t Parentheses indicate percentage of colonies recovered in relation to Mueller-Hinton
agar.
gens are usually cultured on selective media such as
Hektoen enteric, xylose-lysine desoxycholate, and
salmonella-shigella, or combinations of these. Hektoen enteric and xylose-lysine desoxycholate media
are definitely inadequate for isolating Y. enterocolitica,
since they have low recovery rates, and this organism
appears as a lactose fermenter (due to sucrose fermentation) and thus is discarded as a non-pathogen. ReTable I. Colonial Depression of Yersinia enterocolitica
covery on Salmonella-Shigella medium is also poor,
and Commonly Recovered Enterobacteriaceae
as the colonies appear "pinpoint" and do not show
on Pectin Agar*
any particular characteristic to aid in differentiation
from the common flora. Cold-enrichment technics, alNo. of Isolates
% Isolates with
Organism
Tested
Depressed Colonies
though they enhance the yield of the Y. enterocolitica
isolates, 718 are not practical due to the prolonged delay
Yersinia enterocolitica
13
100
necessary for the isolation of the pathogen.
Escherichia colt
48
0
Serratia marcescens
13
0
This improved pectin agar medium makes possible
Klebsiella pneumoniae
21
14t
the
presumptive identification of Y. enterocolitica from
Proteus species!
13
0
mixed cultures in 48 hours. It can easily be prepared,
Enterobacter species§
7
0
Citrohacter freundii
1
0
and has a shelf life of at least two weeks in a refrigerator.
Providencia stuartii
2
0
This pectin medium is currently being used in the
Clinical Microbiology Laboratory at Mercy Hospital in
* Cultures observed at 48 h at 25 C.
+ Confirmed as indole-positive Klebsiella oxytoca.
X Proteus mirahilis 9, P. mor^anii 2. P. vulgaris I, P. rettfieri I.the routine examination of stool specimens for enteric
§ Enterobacter cloacae 7, £. aero^enes I, E. antilomerans 2. pathogens. Such a field trial will aid in an actual clini-
590
BOWEN AND KOMINOS
A.J.C.P. • Oclober 1979
13. Rabson AR, Hallet AF, Koorhhof HJ: Generalized Yersinia
enteroeoliliea infection. J Infect Dis 131:447-451, 1975
14. Rabson AR, Koomhof HJ, Notman J, et al: Hepatosplenic
abscess due to Yersinia enteroeoliliea. Br Med J 4:341, 1972
15. Randall C, Bannatyne RM: Experience with Yersinia enteroAeknonledfiinenl. Dr. Charles E. Copeland provided use of the
eoliliea at The Hospital for Sick Children 1972-4. Can Med J
General Surgery Laboratory.
113:542-543, 1975
16. Schiemann DA, Toma S: Isolation of Yersinia enteroeoliliea
References
from raw milk. Appl Environ Microbiol 35:54-58, 1978
1. Bisset ML: Yersinia enteroeoliliea isolates from humans in Cali- 17. Schleifstein JI. Clark M: Interesting observations (Bacterium
fornia. J Clin Microbiol 4:137-144. 1976
enterocolitieitin). Ann Rep Div Lab Res NY State Dept Health
2. Black RE. Jackson RJ, Tsai T, et al: Epidemic Yersinia entero71-72, 1947
eoliliea due to contaminated chocolate milk. N Engl J Med 18. Sebes JI, Mabry EH, Rabinowitz JG: Lung abscess and osteo298:76-79, 1978
myelitis of rib due to Yersinia enteroeoliliea. Chest 69:5463. Braunstein H. Tucker EB. Gibson BC: Mesenteric lymphade548. 1976
nitis due to Yersinia enteroeoliliea. Am J Clin Pathol 55: 19. Sonnenwirth AC: Yersinia enteroeoliliea as an etiologic agent
506-510, 1971
in meningitis. Bacteriol Proc 87:128, 1969
4. Brenner DJ, Farmer JJ, Hickman FW, et al: Taxonomic and
20. Spira TJ. Kabins SA: Yersinia enteroeoliliea septicemia with
nomenclature changes in Enterobaeleriaeeae. HEW Pub No
septic arthritis. Arch Int Med 136:305-308, 1976
(CDC) 78-8356. 1977
21. Starr MP, Chatterjee AK, Starr PB, et al: Enzymatic degrada5. Chin GN. Noble RC: Ocular involvement in Yersinia enterotion of polygalacturonic acid by Yersinia and Klebsiella speeoliliea infection presenting as Parinaud's oculoglandular syncies in relation to clinical laboratory procedures. J Clin Midrome. Am J Ophthalmol 83:19-23, 1977
crobiol 6:379-386, 1977
6. Doraiswamy NV, Currie ABM, Gray J, et al: Terminal ileitis.
22. Toivanen P, Toivanen A, OIkkonen L, et al: Hospital outbreak
Yersinia enteroeoliliea isolated from faeces. Br Med J 2:23,
of Yersinia enteroeoliliea infection. Lancet 1:801-803, 1973
1977
23. Toma S, Deidrick VR: Isolation of Yersinia enteroeoliliea from
7. Greenwood JR. Flanigan SM, Pickett MJ, et al: Clinical isolaswine. J Clin Microbiol 2:478-481, 1975
tion of Yersinia enteroeoliliea: Cold temperature enrichment. 24. Toma S, LaFleur L: Survey on the incidence of Yersinia enleroJ Clin Microbiol 2:559-560, 1975
colitica infection in Canada. Appl Microbiol 28:469-473. 1974
8. Gutman LT, Ottesen EA. Quan TJ, et al: An interfamilial out25. Vilinskas J, Tilton RC, Kriz JJ: A new clinical entity: Human
break of Yersinia enteroeoliliea enteritis. N Engl J Med 288:
infection with Yersinia presenting as an acute abdomen. Am
1372-1377, 1973
Surg 37:568-572, 1971
9. Harvey S, Greenwood JR. Pickett MJ. et al: Recovery of
26. von Riesen VL: Pectinolytic indole positive strains of Klebsiella
Yersinia enteroeoliliea from streams and lakes of California.
pneumoniae. Int J Syst Bacteriol 26:143-145, 1976
Appl Environ Microbiol 32:352-354. 1976
27. von Riesen VL: Polypectate digestion by Yersinia. J Clin Mi10. Highsmith AK, Feeley JC, Skaliy P. et al: Isolation of Yersinia
crobiol 2:552-553, 1975
enteroeoliliea from well water and growth in distilled water. 28. Wilson HD, McCormick JB, Felley JC: Yersinia enteroeoliliea
Appl Environ Microbiol 34:745-750. 1977
infection in a 4 month old infant associated with infection
11. Kohl S, Jacobson JA. Nahmias A. et al: Yersinia enteroin household dogs. J Pediatr 89:767-769, 1976
eoliliea infections in children. J Pediatr 89:77-79, 1976
29. Zen-Yoji H, Maruyama T, Saki S, et al: An outbreak of en12. Leino R, Kalliomaki JL: Yersiniosis as an internal disease. Ann
teritis due to Yersinia enteroeoliliea occurring at a junior high
Intern Med 81:458-461, 1974
school. Jpn J Microbiol 17:220-222. 1973
cal evaluation of this medium, and thus may establish
it as an integral part of the search for Y. enterocolitica in clinical specimens.