ARTICLE IN PRESS
FOOD
MICROBIOLOGY
Food Microbiology 25 (2008) 22–28
www.elsevier.com/locate/fm
A comparative study between overlay method and selective-differential
media for recovery of stressed Enterobacter sakazakii cells from
infant formula
Murad A. Al-Holya,, Mengshi Linb, Hamzah M. Al-Qadiric, Barbara A. Rascod
a
Department of Clinical Nutrition and Dietetics, Faculty of Allied Health Sciences, Hashemite University, P.O. Box 150459, Zarqa-Jordan, Jordan
b
Food Science Program, 256 William Stringer Wing, Eckles Hall, University of Missouri, Columbia, MO 65211, USA
c
Department of Nutrition and Food Technology, Faculty of Agriculture, The University of Jordan, Amman-Jordan, Jordan
d
Department of Food Science and Human Nutrition, Box 646376, Washington State University, Pullman, WA 99164, USA
Received 19 April 2007; received in revised form 15 August 2007; accepted 5 September 2007
Available online 11 September 2007
Abstract
This study compares the performance of different selective-differential media with the overlay method for recovery of stressed cells of
Enterobacter sakazakii from infant formula milk (IFM). Five different selective-differential media were used in this study: OK medium,
violet red bile agar (VRBA), Druggan–Forsythe–Iversen agar (DFI), Enterobacteriaceae enrichment (EE) agar, and fecal coliform agar
(FCA). Tryptic soy agar supplemented with 0.1% sodium pyruvate (TSAP) was used as a control. The overlay method involved applying
a thin layer (8 ml) of each of the selective media onto TSAP after spreading a sample onto TSAP. Reconstituted IFM was inoculated by
ca 1 107 CFU/ml of a mixture of four strains of E. sakazakii and subjected to different stress conditions: heat (55 1C for 10 min), a
freeze–thaw cycle (20 1C for 24 h, thawed at room temperature, frozen again at 20 1C, and thawed), acidic pH (pH 3.56 for 15 min),
alkaline pH (pH 11.04 for 15 min), and desiccation (E. sakazakii was inoculated onto powdered IFM at a level of ca 1 106 CFU/g, held
at 21 1C, water activity of the inoculated product was 0.29 and examined at 0, 15, and 30 d). No major differences were noticed between
the control (TSAP) and the overlay methods. However, the overlay method recovered significantly higher numbers of stressed
E. sakazakii cells compared to selective-differential media. Also, the selective-differential media exhibited some variability in terms of
their capabilities to recover stressed cells of E. sakazakii. Among all the examined selective-differential media, DFI performed better for
recovering stressed E. sakazakii cells. This study suggests that the overlay method may serve as a potential alternative to direct selective
plating for best recovery of E. sakazakii from IFM.
r 2007 Elsevier Ltd. All rights reserved.
Keywords: Enterobacter sakazakii; Overlay method; Selective-differential media; Infant formula
1. Introduction
Enterobacter sakazakii, a Gram-negative, rod-shaped,
motile bacterium that belongs to the family Enterobacteriaceae, has recently been involved in several cases of fatal
neonatal meningitis (Bar-Oz et al., 2001; Gurtler et al.,
2005; Bowen and Braden, 2006). E. sakazakii is distributed
widely in the environment (Kandhai et al., 2004; Farber,
2004; Arts, 2005). Reconstituted powdered infant formula
milk (IFM) has been implicated as a vehicle of transmisCorresponding author. Tel.: +962 590 3333; fax: +962 390 3350.
E-mail address: [email protected] (M.A. Al-Holy).
0740-0020/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.fm.2007.09.002
sion in several outbreaks and sporadic cases of E. sakazakii
infection (Nazarowec and Farber, 1997; Bar-Oz et al.,
2001). E. sakazakii can be recovered occasionally from
IFM, however, mostly at a very low level (o1 CFU/g)
(Iversen and Forsythe, 2007). Yet, if poor hygienic
practices are used in preparing and handling IFM, the
number of cells may rise quickly, mainly because this
microorganism has a relatively short doubling time at
room temperature (40 min) (Richards et al., 2005). This
becomes very critical since E. sakazakii has a relatively low
infectious dose of 1000 CFU/ml in reconstituted IFM
(Iversen and Forsythe, 2003). The reported fatality rate
associated with E. sakazakii infections is 40–80% (Bowen
ARTICLE IN PRESS
M.A. Al-Holy et al. / Food Microbiology 25 (2008) 22–28
and Braden, 2006). E. sakazakii infection associated with
powdered IFM prompted a recall of a commercial product
in the United States in March 2002 (Center for Disease
Control and Prevention, 2002).
The lack of an appropriate rapid method to test for the
presence of E. sakazakii might have led to underestimating
the prevalence of E. sakazakii in the environment and
foods and the number of reported cases affected by
E. sakazakii. E. sakazakii has a remarkable capability
to survive in a dry environment such as powdered IFM for
long time periods (2 years) that gives it a competitive
advantage of prevailing in dry IFM (Edelson-Mammel
et al., 2005). Therefore, end product testing of powdered
IFM is necessary. Also, having a rapid method would
make it easier to trace isolates from clinical cases back to
their source.
The US Food and Drug Administration (FDA) uses a
time-consuming method to test for the presence of
E. sakazakii in dry IFM. This method is based on
reconstituting a milk sample and incubating it overnight
followed by selective enrichment in Enterobacteriaceae
enrichment (EE) broth and by streaking on violet red bile
glucose agar (VRBG). Colonies from violet red bile glucose
agar are streaked on tryptic soy agar (TSA) to check for the
presence of yellow-pigmented colonies after 48–72 h
incubation at 25 1C and finally by biochemical testing
using API 20E system (US Food and Drug Administration,
2002). To date, several methods have been developed to
detect E. sakazakii in food and environmental samples such
as a fluorogenic selective medium (OK medium) which is
based on the capability of E. sakazakii strain to produce
a-glucosidase that reacts with 4-methylumbelliferyl-a-Dglucoside to give distinct fluorescent colonies of
E. sakazakii (Oh and Kang, 2004) or by using 5-bromo4-chloro-3-indolyl-a-D-glucopyranoside as a substrate for
a-glucosidase in the Druggan–Forsythe–Iversen agar
(DFI). Distinct blue-green colonies are produced on
DFI medium (Iversen et al., 2004a). Another method
involves a selective enrichment in lauryl sulfate broth
supplemented with NaCl and vancomysin followed by
streaking on tryptone soy agar with the resulting yellowpigmented colonies are identified by API 20E system
(Guillaume-Gentil et al., 2005). However, these methods
are either not sufficiently selective or result in underestimating the presence of the organism in the tested
powdered IFM and food samples because of the presence
of selective and differential ingredients in those media that
prohibit the resuscitation of injured cells. For example,
some E. sakazakii strains, especially stressed cells are
sensitive to sodium deoxycholate in DFI agar and to
crystal violet and bile salts in violet red bile glucose agar
(Gurtler and Beuchat, 2005; Iversen and Forsythe, 2007),
precluding their detection in powdered IFM and other
foods. Therefore, developing a rapid and accurate method
for detection of E. sakazakii is crucial. The objective of this
study was to investigate the capability of overlay methods
coupled with selective-differential media for recovering
23
E. sakazakii cells exposed to different stress conditions
(heat, freeze, acid, alkaline, and desiccation).
2. Materials and methods
2.1. Bacterial cultures
Four E. sakazakii strains (ATCC 12868, ATCC 29004,
FSM 292, and FSM 287) were exposed to different stress
conditions and examined in terms of their recovery on a
general (non-selective) medium which served as a control
and on five other selective-differential media and by the
overlay method, which combine an under general medium
layer with upper selective-differential medium layer. The
four strains of E. sakazakii were transferred individually
from stock aqueous glycerol (15%) solutions stored at
20 1C to slants of TSA (Difco, Becton Dickinson, Spark,
MD) and kept refrigerated. Prior to the experiment, the
cultures were transferred from TSA slants to Brain Heart
Infusion (BHI) broth (Difco) and grown for 24 h at 37 1C;
thereafter, the four strains were inoculated in the reconstituted IFM. The initial level of E. sakazakii was
determined by spread plating on TSA and incubation at
37 1C for 24 h.
2.2. Stress conditions
Stationary phase cells of four strains of E. sakazakii were
exposed to heat, a freeze–thaw treatment, acid, alkaline,
and desiccation stresses as described below. Unstressed
cells served as a control. Ten milliliters samples of
reconstituted IFM were sterilized at 121 1C for 15 min to
be ready for further use. After applying each stress
condition E. sakazakii count was determined by 10-fold
serially diluting in 0.1% peptone water and spread plating
on a general medium, selective-differential media or by the
overlay method.
2.3. Heat stress
A commercial powdered IFM was purchased from a
local grocery store before the experiment. Ten milliliters of
the sterilized reconstituted IFM was inoculated with a
mixture of four strains of E. sakazakii at an initial level of
ca 1 107 CFU/ml. The tubes were heated at 55 1C for
10 min (equivalent to an average of one D-value, this time
is enough to elicit about one log reduction in the initial
E. sakazakii load). Heat treatment was conducted in a
water bath (Iso temp 215, Fisher Scientific, Pittsburgh,
PA). The tubes were submerged completely in the water
bath where the temperature was controlled at the target
temperature 70.5 1C. The temperature of the tubes was
monitored by a type T thermocouple (Barnat Co.,
Barrington, IL) connected with a portable thermometer.
After completion of heating, the tubes were immersed
promptly in a mashed ice bath at 0.070.2 1C.
ARTICLE IN PRESS
24
M.A. Al-Holy et al. / Food Microbiology 25 (2008) 22–28
2.4. Freeze stress
Screw-capped test tubes containing 10 ml of inoculated
(ca 1 107 CFU/ml) reconstituted IFM were frozen at
20 1C for 24 h. Thereafter, reconstituted IFM was thawed
at room temperature (21 1C), frozen again at 20 1C for
2 h, and thawed at 21 1C.
2.5. Acid stress
E. sakazakii inoculated (ca 1 107 CFU/ml) samples
(10 ml) of reconstituted IFM in screw-capped test tubes
were adjusted to pH ¼ 3.56 by adding 5 M HCl. The tubes
were held for 15 min at room temperature. Thereafter, 1 ml
of the acid-treated IFM was added to 9 ml sodium
phosphate–potassium phosphate buffer (pH ¼ 7.00) (Fisher Scientific, Pittsburgh, PA). The pH of the suspension
was determined using Accument AB pH meter (Fisher
Scientific, Pittsburgh, PA).
2.6. Alkaline stress
E. sakazakii inoculated (ca 1 107 CFU/ml) samples
(10 ml) of reconstituted IFM in screw-capped test tubes
were adjusted to pH ¼ 11.04 by adding 5 M NaOH.
Alkaline-treated tubes were held for 15 min at room
temperature. Thereafter, 1 ml of the alkaline-treated IFM
was added to 9 ml sodium phosphate–potassium phosphate
buffer (pH ¼ 7.00).
2.7. Desiccation stress
E. sakazakii cells were desiccation stressed by inoculating
24-h-old cultures of a composite of four strains of
E. sakazakii onto the surface of the powdered IFM.
Powdered IFM (50 g) was placed into 250 ml sterile beaker.
The inoculum (100 ml) was sprinkled onto the formula in a
drop-wise manner. After inoculation, the IFM with the
inoculum was mixed vigorously by a sterile spatula for
3 min. The samples were stored in a desiccator at room
temperature for 30 d. The population of E. sakazakii was
determined at days: 0, 15, and 30. The water activity (aw) of
the powdered IFM was measured before and after the
inoculation at day 0, and after 15 and 30 d of storage at
room temperature using an Aqua lab water activity meter
(Model 3TE, Aqualab, Pullman, WA).
2.8. Media used
2.8.1. TSAP
TSA was supplemented with 0.1% (w/v) of sodium
pyruvate (Acros Organics, NJ) to facilitate resuscitation of
injured E. sakazakii cells caused by the various applied
environmental stresses. Pyruvic acid improves recovery of
injured cells by degrading hydrogen peroxide or blocking
its formation (McDonald et al., 1983). TSAP medium
served as a non-selective control medium to enumerate
intact and injured survivors of E. sakazakii.
2.8.2. OK medium
OK medium was compiled as described by Oh and Kang
(2004). After sterilizing the medium at 121 1C for 15 min,
the medium was cooled to 50 1C and supplemented with
4-methyl-umbelliferyl-a-D-glucoside (50 mg/l) (Sigma, St.
Louis, MO) before being dispensed into Petri dishes.
4-methyl-umbelliferyl-a-D-glucoside is a substrate for
a-glucosidase produced by E. sakazakii strains. Upon
cleavage of this substrate, distinctive brilliant fluorescent
colonies of E. sakazakii appear on the medium.
2.8.3. VRBA
VRBA (Difco) was prepared according to the manufacturer’s instructions.
2.8.4. DFI
DFI agar, also known as chromogenic E. sakazakii agar
(Oxoid, Hampshire, England). DFI was developed by
Iversen et al. (2004a). The medium was prepared according
to the manufacturer’s instructions.
2.8.5. EE agar
EE agar was prepared by adding agar (15 g/l) to EE
broth Mossel (Difco). The medium was prepared according
to the manufacturer’s instructions.
2.8.6. FCA
Fecal coliform agar (FCA) (Accumedia Manufacturers,
Lansing, MI) was supplemented with 10 ml/l of 1% rosolic
acid (Matheson Coleman and Bell, Norwod, OH) in 0.2 N
NaOH. The medium was boiled for 2 min, cooled to 50 1C,
and dispensed into Petri dishes.
2.8.7. Overlay method
This method was investigated to determine if higher
numbers of E. sakazakii cells exposed to different intrinsic
(acid, alkaline, desiccation) and extrinsic (heat, freeze)
environmental stresses could be recovered compared to the
methods outlined earlier. The control (unstressed) and
stressed E. sakazakii in reconstituted or powdered IFM
samples were 10-fold serially diluted in 0.1% peptone water
and spread-plated onto TSAP and incubated for 2 h at
37 1C to allow for adequate time for injured and debilitated
E. sakazakii cells to recover. Thereafter, a thin layer (8 ml)
of each of the selective media was overlaid onto TSAP and
the plates were incubated for additional 22 h at 37 1C. This
step was taken to preclude the growth of non-coliform
bacteria.
3. Statistical analysis
At least three independent replicate trials were
conducted and standard deviations were determined.
E. sakazakii counts were log transformed and data were
ARTICLE IN PRESS
M.A. Al-Holy et al. / Food Microbiology 25 (2008) 22–28
Log CFU ml-1
7
6.8
6.6
FCA
EE
DFI
VRBA
OK
TSAP + FCA
TSAP + EE
TSAP + DFI
TSAP + VRBA
TSAP + OK
TSAP
6.4
Fig. 1. Comparison of different selective-differential media with the
overlay method to recover inoculated E. sakazakii from untreated
sterilized reconstituted infant formula. The media were incubated at
37 1C for 24 h. TSAP, tryptic soy agar supplemented with 0.1% (w/v)
sodium pyruvate; OK, OK medium; VRBA, violet red bile agar; DFI,
Druggan–Forsythe–Iversen agar; EE, Enterobacteriaceae enrichment agar;
FCA, fecal coliform agar. The values are means of at least three replicates.
5.4
5.2
FCA
EE
DFI
VRBA
OK
TSAP + FCA
TSAP + EE
TSAP + DFI
TSAP + VRBA
TSAP +OK
TSAP
5
Fig. 2. Comparison of different selective-differential media with the
overlay method to recover heat-stressed inoculated E. sakazakii from
sterilized reconstituted infant formula (reconstituted infant formula
samples were heated at 55 1C for 10 min). The media were incubated at
37 1C for 24 h. TSAP, tryptic soy agar supplemented with 0.1% (w/v)
sodium pyruvate; OK, OK medium; VRBA, violet red bile agar; DFI,
Druggan–Forsythe–Iversen agar; EE, Enterobacteriaceae enrichment agar;
FCA, fecal coliform agar. The values are means of at least three replicates.
6.6
6.4
6.2
6
5.8
5.6
FCA
EE
DFI
VRBA
OK
TSAP + FCA
TSAP + EE
TSAP + DFI
5.4
TSAP + VRBA
7.2
5.6
TSAP + OK
7.4
5.8
TSAP
In this study, a mixture of four strains of E. sakazakii
inoculated in IFM was exposed to different stress conditions including, heat, freeze-thaw, acid, alkaline, and
desiccation stresses compared to the control. Fig. 1 shows
counts for the composite population for the control.
E. sakazakii strains exhibit a great disparity in their heat
resistance. Edelson-Mammel and Buchanan (2004) reported that the D58 1C for E. sakazakii in reconstituted
infant formula ranged from about 0.5 to 10.0 min, which
accounts for about a 20-fold difference in heat resistance
among E. sakazakii strains in reconstituted IFM. The
thermal inactivation kinetics of some strains of E. sakazakii
were determined in our lab (Al-Holy et al., 2007) and four
of the most heat resistant strains were chosen for the heatstress trials. The four strains (E. sakazakii ATCC 12868,
E. sakazakii ATCC 29004, E. sakazakii FSM 292, and
E. sakazakii FSM 287) exhibited a D55 1C from 1.9 to
14.8 min. The inoculated reconstituted IFM was heated for
10 min. TSAP performed significantly (Pp0.05) better
than all other media for recovering E. sakazakii cells
followed by the overlay (TSAP+OK) medium. In comparison, fewer cells were recovered on EE agar (Pp0.05)
compared to the other media. For the heat-stressed cells
(Fig. 2), TSAP, TSAP+VRBA, TSAP+IDF, and
TSAP+FCA resulted in the greatest number of recovered
heat-stressed cells of E. sakazakii, with the lowest recovery
in VRBA.
6
Log CFU ml-1
4. Results and discussion
6.2
Log CFU ml-1
analyzed with a computer software package (SAS Institute,
Cary, NC) using analysis of variance and Fisher’s least
significant difference (LSD) test for mean separations
(Pp0.05).
25
Fig. 3. Comparison of different selective-differential media with the
overlay method to recover freeze-stressed inoculated E. sakazakii from
sterilized reconstituted infant formula (reconstituted infant formula
samples were frozen at 20 1C for 24 h. Thereafter thawed at room
temperature (21 1C), frozen again at 20 1C for 2 h, and thawed at 21 1C).
The media were incubated at 37 1C for 24 h. TSAP, tryptic soy agar
supplemented with 0.1% (w/v) sodium pyruvate; OK, OK medium;
VRBA, violet red bile agar; DFI, Druggan–Forsythe–Iversen agar; EE,
Enterobacteriaceae enrichment agar; FCA, fecal coliform agar. The values
are means of at least three replicates.
Fig. 3 shows the recovery of E. sakazakii cells that were
exposed to different freezing-thawing cycles in reconstituted IFM. TSAP and the other combinations of the
overlay method achieved significantly better recovery
relative to the other selective-differential media except for
the IDF medium. In comparison, EE agar recovered fewer
freeze-stressed E. sakazakii cells compared to the other
media used. About 0.95 log units reduction in E. sakazakii
ARTICLE IN PRESS
M.A. Al-Holy et al. / Food Microbiology 25 (2008) 22–28
6.6
Log CFU ml-1
6.4
6.2
6
5.8
5.6
FCA
EE
DFI
VRBA
OK
TSAP + FCA
TSAP + EE
TSAP + DFI
TSAP + VRBA
TSAP + OK
TSAP
5.4
Fig. 4. Comparison of different selective-differential media with the
overlay method to recover acid-stressed inoculated E. sakazakii from
sterilized reconstituted infant formula (reconstituted infant formula
samples were adjusted to pH ¼ 3.56 by adding 5 M HCl). The media
were incubated at 37 1C for 24 h. TSAP, tryptic soy agar supplemented
with 0.1% (w/v) sodium pyruvate; OK, OK medium; VRBA, violet red
bile agar; DFI, Druggan–Forsythe–Iversen agar; EE, Enterobacteriaceae
enrichment agar; FCA, fecal coliform agar. The values are means of at
least three replicates.
4
3.8
3.6
3.4
3.2
FCA
EE
DFI
VRBA
OK
TSAP + FCA
TSAP + EE
TSAP + DFI
TSAP + VRBA
TSAP + OK
3
TSAP
count resulted from freezing–thawing cycles compared to
the control (Fig. 1). E. sakazakii is fairly sensitive to
chilling and did not grow in infant cereal reconstituted with
water or milk stored at 4 1C (Richards et al., 2005). In
another study, the growth of E. sakazakii in reconstituted
IFM and tryptic soy broth (TSB) stopped completely after
incubation at 6 1C (Iversen et al., 2004b).
Fig. 4 shows the performance of the media in recovering
acid-stressed E. sakazakii cells. The pH of the inoculated
reconstituted IFM was adjusted to 3.56 and held for 15 min
at room temperature. Here TSAP+IDF exhibited a
greater recovery of the stressed cells than TSAP, other
selective media in overlay methods and the other selectivedifferential media, with FCA being the medium with the
lowest recovery. Approximately 0.90 log units reduction in
E. sakazakii count resulted from the applied acid stress
compared to the control (Fig. 1). Edelson-Mammel et al.
(2006) reported that E. sakazakii is a moderately acidresistant organism and its capability to combat acidic
conditions is enhanced when the cells enter the stationary
phase. Yet, a substantial diversity in the acid resistance was
observed among 12 strains of E. sakazakii tested in
acidified TSB in that study. Edelson-Mammel et al.
(2006) also reported that when the pH of the broth was
adjusted to 3.5 by means of HCl, no significant reduction
was observed after holding the strains for 2 h. Nonetheless,
only two strains showed an appreciable degree of reduction
(1.1–3.5 log units) after 5 h of holding time. The lower
tolerance to acidic condition in our study compared to
Edelson-Mammel et al. (2006) may be attributed to the
type of E. sakazakii strains used, the composition of
medium, age of cells, and time of exposure to acidic
conditions.
Log CFU ml-1
26
Fig. 5. Comparison of different selective-differential media with the
overlay method to recover alkaline-stressed inoculated E. sakazakii from
sterilized reconstituted infant formula (reconstituted infant formula
samples were adjusted to pH ¼ 11.04 by adding 5 M NaOH). The media
were incubated at 37 1C for 24 h. TSAP, tryptic soy agar supplemented
with 0.1% (w/v) sodium pyruvate; OK, OK medium; VRBA, violet red
bile agar; DFI, Druggan–Forsythe–Iversen agar; EE, Enterobacteriaceae
enrichment agar; FCA, fecal coliform agar. The values are means of at
least three replicates.
Fig. 5 shows the recovery of alkaline-stressed
E. sakazakii from reconstituted IFM in different media.
The pH of the reconstituted IFM was adjusted with NaOH
to 11.04 and samples were held for 15 min at room
temperature. Generally, TSAP and overlay methods
performed significantly better than the selective-differential
media alone. The greatest recovery of the alkaline-stressed
cells was in TSAP and TSAP+OK; with the lowest
recovery in the OK medium.
Table 1 shows recovery of E. sakazakii cells exposed to
desiccation stress over a 30-d period. The initial water
activity of the powdered IFM was about 0.20 and increased
immediately to about 0.29 after the inoculum was applied
to the powdered IFM. The samples were stored in a
desiccator, and after 15–30 d, the water activity increased
slightly possibly resulting from moisture migration from
the air to the sample during storage. E. sakazakii has a
remarkable capability to survive in dry environment for
long time periods that give it a competitive advantage to
prevail in a dry environment such as powdered IFM
(Edelson-Mammel et al., 2005). E. sakazakii produces,
trehalose, a compatible solute that plays an important role
in the protection of E. sakazakii against drying by
stabilizing phospholipids membranes and proteins
(Breeuwer et al., 2003). A significant reduction in the
number of cell numbers occurred after 30 d of storage of
the powdered IFM. In TSAP, the counts decreased by
about 1.3 log units after 15 d and to about 1.7 log units
after 30 d. Performance of the tested media for recovering
desiccation-stressed E. sakazakii cells was in general
agreement with that observed for the other treatments
with the highest recovery in TSAP, TSAP+VRBA,
TSAP+DFI, and TSAP+FCA and the lowest recovery
ARTICLE IN PRESS
M.A. Al-Holy et al. / Food Microbiology 25 (2008) 22–28
Table 1
Comparison of different selective-differential media with the overlay
method to recover desiccation-stressed E. sakazakii from powdered infant
formula
Media
Log CFU/ga
Day 0
TSAP
TSAP+OK
TSAP+VRBA
TSAP+DFI
TSAP+EE
TSAP+FCA
OK
VRBA
DFI
EE
FCA
6.0370.07
5.9270.18
5.9770.11
6.0370.05
6.0070.10
6.0270.10
5.9070.07
5.8070.05
5.8770.12
5.9170.11
5.8570.06
Day 15
Aa
ABCa
ABa
Aa
Aba
Aba
ABCa
Ca
ABCa
ABCa
BCa
4.7970.07
4.5770.05
4.6570.03
4.6870.06
4.6070.05
4.6770.09
4.4470.12
4.4070.04
4.4970.06
4.2170.04
4.3170.06
Day 30
Ab
CDb
Bb
Bb
BCb
Bb
EFb
Fb
DEb
Hb
Gb
4.3370.01
4.3370.02
4.2370.04
4.2670.15
4.1170.11
4.1970.11
4.1170.12
4.3670.21
4.1870.18
3.4770.16
4.0770.36
Ac
Ac
Abc
Abc
Bc
Abc
Bc
Ab
ABc
Cc
Bb
The water activities of the infant formula were 0.197, 0.285, 0.348, and
0.351 for the un-inoculated, inoculated at day 0, 15, and 30, respectively.
TSAP, tryptic soy agar supplemented with 0.1% (w/v) sodium pyruvate;
OK, OK medium; VRBA, violet red bile agar; DFI, Druggan–Forsythe–
Iversen agar; EE, Enterobacteriaceae enrichment agar; FCA, fecal
coliform agar.
a
Data represent means7standard deviations of three measurements.
Means with the same capital letter in the same column are not significantly
different (PX0.05). Means with the same lowercase letter in the same row
are not significantly different (PX0.05).
in EE medium. This is further evidence that supports the
use of the overlay method for detecting low levels of
injured E. sakazakii in powdered IFM without compromising the selectivity of the medium. Also, as evident from
Table 1, the agar made from EE broth was less effective in
recovering desiccation-stressed E. sakazakii cells compared
to other treatments as the storage time increased and even
the overlay medium (TSAP+EE) recovered significantly
lower number of desiccation-stressed cells after 30 d
compared to TSAP and other overlay media combinations.
This finding questions the feasibility of using EE broth as
a step in the procedure adopted by the FDA for detecting
E. sakazakii in powdered IFM. DFI agar performed better
than all other selective-differential media in recovering
desiccation-stressed E. sakazakii cells supporting the
finding of Iversen and Forsythe (2007) who reported that
DFI agar was more favorable relative to the FDA method
for the detection of E. sakazakii in powdered IFM,
especially, when the microorganism is present at low
concentrations in the product. DFI agar sensitivity can
be enhanced and the number of false-positive colonies on
DFI can be minimized if the E. sakazakii enrichment broth
is supplemented with 100 g/l of sucrose. This is because the
majority of a-glucosidase positive Enterobacteriaceae
cannot ferment sucrose and because sucrose is a humectant
that lowers the water activity of the broth and hence,
precluding the growth of the majority of interfering
Enterobacteriaceae (Iversen and Forsythe, 2007). Additionally, the capability of DFI agar to recover injured
27
E. sakazakii cells may be further enhanced by supplementation of the medium with sodium pyruvate.
TSAP is a general medium that contains no selective or
differential ingredients that may compromise the resuscitation of stressed E. sakazakii cells. Additionally, it is
supplemented with pyruvate that may trigger the recovery
of injured cells (McDonald et al., 1983; Czechowicz et al.,
1996). Notwithstanding, this medium cannot be used alone
for the presumptive detection of E. sakazakii cells from
powdered IFM or other foodstuffs because TSAP lacks
selective ingredients. The FDA recommends the use of a
method for the detection of E. sakazakii in IFM based on a
selective enrichment step in EE broth followed by selective
plating on violet red bile glucose agar (US Food and Drug
Administration, 2002). However, the results of our study
indicate that both of these media inhibit the recovery of the
injured E. sakazakii cells. Also, it was pointed out that
about 2% of E. sakazakii strains do not produce yellow
pigmentation on TSA after 3 d of incubation at 25 1C.
Additionally, most a-glucosidase positive Enterobacteriaceae other than E. sakazakii give yellow pigmentation on
TSA at 25 1C, a criterion, which has been recommended for
the presumptive identification of E. sakazakii from
powdered IFM (Iversen and Forsythe, 2007).
On the other hand, selective-differential media contain
certain selective ingredients that may prevent the resuscitation of the debilitated E. sakazakii cells. This is possibly the
reason behind the modest performance of the selectivedifferential media in recovering stressed E. sakazakii cells.
For example, some E. sakazakii strains are unable to grow
in lauryl sulfate broth or brilliant green bile broth (Iversen
et al., 2004b; Iversen and Forsythe, 2007). Additionally, EE
broth contains oxgall and brilliant green, VRBA contains
bile salts no. 3 and crystal violet, and DFI contains
sodium deoxycholate that may prevent repair of injured
E. sakazakii cells, precluding their detection. Therefore, we
hypothesize that combining TSAP with other selectivedifferential media in an overlay method may be a prudent
idea to recover E. sakazakii cells from food samples,
especially, in case of cells exposed to different kinds of
stresses that may be used in a food processing environment.
This is particularly significant because E. sakazakii usually
occurs at very low levels in the powdered IFM (o1 CFU/g).
The results of the current study reveal the inferiority of
the selective-differential media in recovering stressed cells
of E. sakazakii regardless of the type of the stress condition
applied unless coupled with a non-selective medium such as
TSA in an overlay method.
This study as well as other studies question the feasibility
of using the current FDA method to detect E. sakazakii in
powdered IFM. Generally, TSAP and the combinations of
TSAP with other selective-differential media in an overlay
method resulted in a greater recovery of E. sakazakii cells
exposed to heat, freeze–thaw cycles, acid, alkaline, and
desiccation stresses. However, because of the lack of
selectivity of TSAP, it is preferable to use the overlay
method to recover E. sakazakii from IFM powder. IDF
ARTICLE IN PRESS
28
M.A. Al-Holy et al. / Food Microbiology 25 (2008) 22–28
agar generally outperformed the other selective-differential
media used in recovering stressed cells of E. sakazakii.
Further research is needed to optimize the incubation
conditions for the recovery of stressed E. sakazakii cells in
different selective media and to validate the overlay
method against a standard reference method to determine
E. sakazakii count in dry IFM.
Acknowledgments
This research was supported by USDA special grants,
Washington State University and Hashemite University,
Jordan. Special thanks are given to Dr. Dong-Hyun Kang
and Mr. Peter Gray at Washington State University for
providing bacterial strains.
References
Al-Holy, M.A., Lin, M., Abu-Ghoush, M., Rasco, B.A., 2007. Thermal
resistance, growth and inactivation of Enterobacter sakazakii in dry
and reconstituted infant formula. In: Proceedings of the Fifth
International Congress on Food Technology, vol. 3, Thessalonica,
Greece, pp. 303–313.
Arts, M., 2005. Enterobacter sakazakii in factories and household. Lancet
364, 414–415.
Bar-Oz, B., Perminger, A., Peleg, O., Block, C., Arad, I., 2001. Enterobacter
sakazakii infection in the newborn. Acta Paediatr. 90, 356–358.
Bowen, A.B., Braden, C.R., 2006. Invasive Enterobacter sakazakii in
infants. Emerg. Infect. Dis. 12, 1185–1189.
Breeuwer, P., Lardeau, A., Peterz, M., Joosten, H.M., 2003. Desiccation
and heat tolerance of Enterobacter sakazakii. J. Appl. Microbiol. 95,
967–973.
Center for Disease Control and Prevention, 2002. Enterobacter sakazakii
infections associated with the use of powdered infant formula—
Tennessee, 2001. Morb. Mortal. Wkly. Rep. 51, 297–300.
Czechowicz, S.M., Santos, O., Zottola, E.A., 1996. Recovery of thermallystressed Escherichia coli O157:H7 by media supplemented with
pyruvate. Int. J. Food Microbiol. 33, 275–284.
Edelson-Mammel, S.G., Buchanan, R.L., 2004. Thermal inactivation of
Enterobacter sakazakii in rehydrated infant formula. J. Food Prot. 67,
60–63.
Edelson-Mammel, S.G., Porteous, M.K., Buchanan, R.L., 2005. Survival
of Enterobacter sakazakii in a dehydrated powdered infant formula.
J. Food Prot. 68, 1900–1902.
Edelson-Mammel, S.G., Porteous, M.K., Buchanan, R.L., 2006. Acid
resistance of Enterobacter sakazakii, and impact of habituating the
cells to an acidic environment. J. Food Sci. 71, M201–M207.
Farber, J.M., 2004. Enterobacter sakazakii new foods for thought. Lancet
363, 5–6.
Guillaume-Gentil, O., Sonnard, V., Kandhai, M.C., Marugg, J.D.,
Joosten, H., 2005. A simple and rapid cultural method for detection
of Enterobacter sakazakii in environmental samples. J. Food Prot. 68,
64–69.
Gurtler, J.B., Beuchat, L.R., 2005. Performance of media for recovering
stressed cells of Enterobacter sakazakii as determined using spiral
plating and ecometric techniques. Appl. Environ. Microbiol. 71,
7661–7669.
Gurtler, J.B., Kornacki, J.L., Beuchat, L.R., 2005. Enterobacter sakazakii:
a coliform of increased concern to infant health. Int. J. Food
Microbiol. 104, 1–34.
Iversen, C., Forsythe, S.J., 2003. Risk profile of Enterobacter sakazakii, an
emergent pathogen associated with infant milk formula. Trends Food
Sci. Technol. 14, 443–454.
Iversen, C., Forsythe, S.J., 2007. Comparison of media for the isolation of
Enterobacter sakazakii. Appl. Environ. Microbiol. 73, 48–52.
Iversen, C., Druggan, P., Forsythe, S.J., 2004a. A selective differential
medium for Enterobacter sakazakii. Int. J. Food Microbiol. 96,
133–139.
Iversen, C., Lane, M., Forsythe, S.J., 2004b. The growth profile,
thermotolerance and biofilm formation of Enterobacter sakazakii
grown in infant formula milk. Lett. Appl. Microbiol. 38, 378–382.
Kandhai, M.C., Reij, M.W., Gorris, L.G.M., Guillaume-Gentil, O.,
Schothrost, M.V., 2004. Occurrence of Enterobacter sakazakii
in food production environments and household. Lancet 363,
39–40.
McDonald, LC., Hackney, C.R., Ray, R., 1983. Enhanced recovery
of injured Escherichia coli by compounds that degrade hydrogen
peroxide or block its formation. Appl. Environ. Microbiol. 45,
360–365.
Nazarowec-White, M., Farber, J.M., 1997. Enterobacter sakazakii: a
review. Int. J. Food Microbiol. 34, 103–113.
Oh, S.-W., Kang, D.-H., 2004. Flurogenic selective and differential
medium for isolation of Enterobacter sakazakii. Appl. Environ.
Microbiol. 70, 5692–5694.
Richards, G.M., Gurtler, J.B., Beuchat, L.R., 2005. Survival and growth
of Enterobacter sakazakii in infant rice cereal reconstituted with water,
milk, liquid infant formula, or apple juice. J. Appl. Microbiol. 99,
844–850.
US Food and Drug administration, 2002. Isolation and enumeration of
Enterobacter sakazakii from dehydrated powdered infant formula
/http://www.cfsan.fda.gov/comm/mmesakaz.htmlS.