718 Part VII: Food Safety
foods (Donnelly 2001). These methods vary only in
the type of selective media that are used (Hayes et
al. 1992). They are species specific, but because of
the strong selective media used, which may reduce
growth of L. monocytogenes, the methods tend to
lose some of their sensitivity. For example, Hayes et
al. (Hayes et al. 1992) showed that the USDA-FSIS
method was only able to detect Listeriain 65% of
foods known to be contaminated with the bacteria,
while the NGFIS method only gave positive results
for 74% of the same cases. If any two of the above
three methods were used together, the success in
detecting Listeriacontamination jumped to between
87 and 91%. This is still not an acceptable rate of
detection, when deaths may be caused if the patho-
gen is present but not detected.
Some enrichment-based methods have been mod-
ified or automated to increase throughput. For exam-
ple, automated dilution of the samples before ho-
mogenization, and automated homogenization in
sterile bags is faster than sample preparation by
hand (de Boer and Beumer 1999). Flow cytometry
has also been used to speed up counting times, but it
cannot be used to distinguish between living and
dead cells (de Boer and Beumer 1999).
There are a few problems with enrichment-based
methods. First, they are time-consuming. For exam-
ple, International Organization for Standardization
method 11290-1 (Wan et al. 2003), the official
method used by Australian food inspectors, requires
5 days for determination of a negative result for Lis-
teriacontamination. If a positive test result occurs,
additional days are required for biochemical tests to
identify the strain (Wan et al. 2003). Second, enrich-
ment methods do not account for the recovery of
sublethally injured bacteria that may be present as a
result of heating, freezing, or acidification of the
foods. These bacteria may not be able to form
colonies under selective pressure during recovery,
but are still virulent and able to grow in food. For
example, sublethally injured L. monocytogenescells
are capable of repair and growth at storage tempera-
tures of 4°C or higher (Donnelly 2001). Third, after
enrichment, selection is often possible only to the
genus level. In Listeria, to distinguish between the
pathogenic L. monocytogenesand the other five non-
pathogenic species, additional biochemical charac-
terization must still be done. This adds more time to
the detection procedure. Finally, there is some evi-
dence (Kathariou 2002, Loncarevic et al. 1996,
Ryser et al. 1996) that some L. monocytogenes
strains are more sensitive to enrichment and selec-
tion procedures than others. For instance, serotype
4b is prevalent in 50–70% of listeriosis outbreaks
but is rarely found in contaminated food, thereby
inviting speculation that its occurrence is underesti-
mated using this method.
To avoid some of these problems, many newer
methods rely on the initial use of selective, chro-
mogenic substrates that change color in the presence
of some enzymes present only in certain species,
which makes these species easier to identify. For
instance, L. monocytogenesand L. ivanoviicontain
the plcA gene, encoding a phosphatidylinositol-
specific phospholipase C, which is not found in any
of the other Listeriastrains. When plated on selec-
tive BCM plates, positive colonies are turquoise,
while all others are white (Jinneman et al. 2003).
The two strains can then be separated based on sug-
ar utilization or other plating techniques, thus reduc-
ing the number of tests that must be performed to
detect L. monocytogenesin food samples (Jinneman
et al. 2003).AUTOMATEDMETHODS FORDETECTION OF
PATHOGENS(NON-NUCLEICACID-BASED)There are a number of commercially available,
semiautomated systems for the detection of various
foodborne pathogens, including the MicroLog Sys-
tem (Biolog, Inc., Hayward, California), the Micro-
bial Identification System (MIS) (MIDI, Inc., Ne-
wark, Delaware), the VITEK System (bioMerieux
Vitek, Hazelwood, Missouri), and the Replianalyzer
system (Oxoid Inc., Nepean, Ontario) (Odumeru et
al. 1999). The Microlog System is based on carbon
oxidation profiles, MIS on lipid analysis, and the
VITEK and Replianalyzer systems on biochemical
selection. According to Odumeru et al. (1999), these
systems are able to detect Listeriareliably at the
genus level 90–100% of the time, depending on the
system being used, but biochemical tests are still
required for detection at the species level.NUCLEICACID-BASEDMETHODS OF
PATHOGENDETECTIONOne of the biggest problems associated with detec-
tion of Listeriais the low numbers at which the bac-
teria are normally found in contaminated food sam-