Optimal conditions for the detection of heat-injured cells of Listeria mono
cytogenes in modified Pennsylvania State University (mPSU) broth were deter
mined using a response surface design generated by a computer program, EChi
p. Different combinations of incubation temperatures and lithium, magnesium
, and D-serine concentrations were evaluated to determine the optimum condi
tions for the detection of heat-injured L. monocytogenes in filter-steriliz
ed whole milk inoculated with selected problematic background microflora. A
concentration of 212 mM lithium chloride completely inhibited the growth o
f Enterococcus faecium while permitting recovery and detection of L. monocy
togenes. A concentration of 15.8 mM MgSO4 was found to be optimum for the r
ecovery and detection of L. monocytogenes. A concentration of 140.2 mM D-se
rine was found to completely inhibit the germination of Bacillus subtilis v
ar. globii spores but not recovery and detection of L. monocytogenes. Under
optimum concentrations of LiCl. MgSO4, and D-serine and in the absence of
background microflora, the effect of incubation temperature on percentage d
etection was described by a second-order polynomial model, and 28 degreesC
was determined to be optimal. In the presence of background microflora, the
effect of incubation temperature on percentage detection of heat-injured c
ells was described by a third-order polynomial model, and 30 degreesC was f
ound to be optimal. Optimizing the levels of highly specific and selective
agents, nutrients, and incubation temperature in one recovery enrichment sy
stem dramatically increased the Listeria/background microflora ratio. This
resulting medium, optimized PSU (oPSU) broth, greatly improved the detectio
n of heat-injured and nonheat-injured L. monocytogenes by both conventional
and molecular methods (Oxoid's Listeria Rapid Test, Gen-Probes Accuprobe L
isteria monocytogenes Culture Identification Test, and Qualicon's BAX for s
creening Listeria monocytogenes).