Food Protection Trends - May/June 2018 - 181

10 × 20 cm sterile bags, and inoculated to reach target
bacterial concentrations of 2.0, 3.0, and 4.0 log CFU/g.
The meat was thoroughly hand mixed for about 1 min and
then refrigerated for 30 min to allow bacterial attachment to
the ground beef. Bacterial concentrations in the inoculated
beef samples were confirmed prior to IMS by plating on
TSA and mRBA and incubating for 24 h at 37°C. Colonies
were enumerated and reported as CFU/g. Log CFU/g
conversion was done as needed.
IMS sensitivity
Assays were conducted using cultures in enrichment
broth. IMS was performed individually for each of the seven
target serogroups, using the five-strain cocktails previously
mentioned. Bacterial concentrations of 1.0, 2.0, 3.0, 4.0, or 5.0
log CFU/ml were tested. Results were defined as positive when
at least one colony was recovered after IMS, and as negative
when no colonies were recovered after IMS. Recovered
colonies were always confirmed via latex agglutination assays
per serogroup. Sensitivity was estimated as: positive results/
(positive results + negative results) × 100 (20).
Testing magnetic bead volumes
Reduced volumes of magnetic beads were tested with
different bacterial concentrations. For each serogroup, bead
volumes of 20, 16, 14, 12, and 10 µl were evaluated at each
bacterial concentration (1.0, 2.0, 3.0, 4.0, or 5.0 log CFU/
ml), and IMS was performed to recover target cells in broth.
For the inoculated ground beef, 20 and 10 µl of anti-O157
beads were evaluated at each bacterial concentration (2.0,
3.0, and 4.0 log CFU/g). Automated IMS was performed
individually for each serogroup, following the bead retriever
manufacturer's standardized protocol. As recommended in
the standard procedure, 5-wells tube strips were used, and
each volume of beads tested was divided in two and added
to the corresponding wells (e.g., for 20 µl, aliquots of 10 µl
were dispensed into wells 1 and 2, for 16 µl, aliquots of 8 µl
were dispensed, etc.). After completion of IMS, 50 µl of the
bacteria-beads complex were spread-plated onto TSA for
cultures tested in enrichment broth and mRBA for ground
beef, and incubated for 24 h at 37°C. Recovered colonies
were confirmed using agglutination tests specific to each O
group, ensuring the presence of the correct serogroup. Also,
colonies were enumerated and the CFU/ml estimated.
Statistical analyses
For the evaluation of the magnetic beads capture efficiency
and IMS detection limit, a full factorial experimental design
was used. Experiments carried out with the pure culture
had a 5 by 5 factorial design, with the factors of bacterial
concentration (1.0 to 5.0 log CFU/ml) and bead volume
(20, 16, 14, 12, and 10 µl). For the experiments carried out
with inoculated ground beef samples, the factorial design was
3 by 5, with the factors of bacterial concentration (2.0, 3.0,

and 4.0 log CFU/g) and bead volume (20, 16, 14, 12, and 10
µl). All plate counts obtained after IMS were log transformed
prior to analysis. Duplicate samples were processed for
each repetition, and five repetitions of the experiment were
conducted. Correlations between factors were determined
through analysis of variance (R Version 2.15.0, 2012. R Core
Team). Means were separated by use of Tukey's test. Means
were considered significantly different when P ≤ 0.05.
RESULTS
IMS sensitivity and limits of detection
Sensitivity was evaluated at the detection limit indicated
by the magnetic bead manufacturers, which is 100 cells/ml.
Findings from this study indicate that the lowest bacterial
concentration detected by IMS varied among serogroups.
When the bacterial concentration in the sample was 2 log
CFU/ml and 20 µl of beads (recommended by manufacturer) were used, sensitivity per serogroup was found to be 62.5,
87.5, 87.5, 50.0, 87.5, and 87.5, and 0.0% for O157, O103,
O26, O145, O45, O121, and O111, respectively. Furthermore, other bacterial concentrations were tested to identify
the detection limit at 100% sensitivity. Table 2 shows the
proportion of samples from which target cells were recovered
after IMS. After conducting IMS on STEC cultures at 2 log
CFU/ml, magnetic beads were not always effective in recovering target cells, indicating that 100% sensitivity was not
achieved with the suggested 100 cells. For serogroups O157,
O103, O26, O145, O45, and O121, the lowest bacterial concentration effective in recovering cells from 100% of the samples tested was 3.0 log CFU/ml. E. coli O111 was detected
through IMS in all repetitions only when the concentration
was at least 5.0 log CFU/ml.
Testing a reduced magnetic bead volume
To evaluate whether the bead volume was related to
IMS sensitivity, the assays were performed individually
for each STEC serogroup. Cell recovery was tested using
the aforementioned cultures in enrichment broth. Samples
contained bacterial concentrations of 1.0, 2.0, and 3.0 log
CFU/ml (prior to IMS). Varying magnetic bead volumes
of 20, 16, 14, 12, and 10 µl per culture concentration were
evaluated to determine the optimum volume of beads
required during IMS to separate target cells. The volume of
beads recommended by the manufacturer for the standard
automated IMS protocol is 20 µl, and lowering this volume
to 10 µl did not affect IMS efficiency in recovering bacterial
target cells from enrichment broth. No significant difference
(P > 0.05) in cells recovered was found among the different
bead volumes tested for any of the target STEC serogroups
(Table 3). While all tested serogroups were evaluated with
reduced volumes of beads, using bacterial concentrations
from 1 to 3 log CFU/ml, E. coli O111 was tested also, with
additional concentrations of 4 and 5 log CFU/ml. With 4
log CFU/ml, the average CFU recovered was 2.13 and

May/June Food Protection Trends

181

Table of Contents for the Digital Edition of Food Protection Trends - May/June 2018