International Journal of Food Microbiology 73 (2002) 261 – 274

www.elsevier.com/locate/ijfoodmicro

Modelling the competitive growth of Listeria monocytogenes and

Listeria innocua in enrichment broths
Marie Cornu a,b,*, Martin Kalmokoff c, Jean-Pierre Flandrois a
a
CNRS UMR 5558, Laboratoire de bactériologie, Faculté de Médecine Lyon Sud, BP12, 69921 Oullins Cedex, France
b
AFSSA LERPAC, 22 rue Pierre Curie, BP332, 94709 Maisons Alfort Cedex, France
c
Bureau of Microbial Hazards, Food Directorate, Health Protection Branch, Health Canada, Sir Frederick G. Banting Research Centre,
Tunney’s Pasture, P.L. 2204A2, Ottawa, ON, Canada K1A OL2
Received 16 May 2001; accepted 9 August 2001

Abstract

The overgrowth of Listeria innocua in enrichment broths designed for the isolation of Listeria monocytogenes is believed to
result from two factors: a selective growth advantage of L. innocua, and/or an inhibitory interspecies interaction. The generation
times of 13 isolates of L. innocua and L. monocytogenes were determined in Brain Heart Infusion (BHI) and a variety of
enrichment media. No significant differences were found in growth characteristics between either species in the various media,
suggesting that the growth advantage of L. innocua in enrichment media was not as significant as previously described. Kinetic
analysis of mixed cultures of L. monocytogenes and isolates of L. innocua producing a variety of inhibitory activities
demonstrated the possibility of an inhibitory interaction between these two species resulting in the overgrowth of the enrichment
culture with L. innocua. Modelling the evolution of the ratio between two populations in an enrichment process was used to
analyze the impact of a selective growth advantage in L. innocua in an enrichment process for growth of L. monocytogenes.
These findings support the widely held view that an overgrowth of L. innocua in the enrichment process can result from both a
selective growth advantage as well as the production of inhibitory compounds. From a practical perspective, these interactions
can result in an increase in false negatives. Crown Copyright D 2002 Published by Elsevier Science B.V. All rights reserved.

Keywords: Selective advantage; Bacteriocin-like inhibition; Overgrowth

1. Introduction food at the time of consumption (see European

Commission, 1999). In some products, such as soft
Listeria monocytogenes represents an important cheese, there is a zero tolerance for the presence of
food-borne pathogen (Farber and Peterkin, 1991), L. monocytogenes due to the possibility of in-food
and its detection is crucial within the food industry. growth occurring during storage (see European Com-
It is widely recognised that the concentration of L. mission, 1999). In many foods, the initial isolation
monocytogenes should be lower than 100 cfu/g of of L. monocytogenes may pose difficulties due to the
presence of low cell numbers within the larger in-
*
digenous non-pathogenic microflora. Thus, metho-
Corresponding author. AFSSA LERPAC, 22 rue Pierre Curie,
BP332, 94709 Maisons Alfort Cedex, France. Tel.: +33-1-4977-
dologies to detect L. monocytogenes generally
2644; fax: +33-1-4977-2640. involve a selective enrichment step (see Donnelly,
E-mail address: m.cornu@afssa.fr (M. Cornu). 1999).

0168-1605/02/$ - see front matter. Crown Copyright D 2002 Published by Elsevier Science B.V. All rights reserved.
PII: S 0 1 6 8 - 1 6 0 5 ( 0 1 ) 0 0 6 5 8 - 4
262 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274

In the mid-1990s, three publications demonstrated genes) were examined and compared using both a
that the presence of Listeria innocua could mask L. standard growth medium (Brain Heart Infusion, BHI)
monocytogenes following the enrichment procedure and four selective enrichment media. These findings
(Petran and Swanson, 1993; Curiale and Lewus, indicate that interspecies differences could impart a
1994; MacDonald and Sutherland, 1994). Apart from growth advantage of L. innocua over L. monocyto-
being non-pathogenic, L. innocua is physiologically genes in selective media, although these differences
close to L. monocytogenes, and both can occur in the appeared less significant than previously thought. In
same food products. Overgrowth of L. innocua during addition, the effect of inhibitory isolates of L. innocua
enrichment could lead to a higher risk of false-nega- on the growth of L. monocytogenes in enrichment
tives, as there would be little chance to detect L. mono- media was also investigated. Interactions between
cytogenes when high numbers of L. innocua may also these populations could be mathematically modelled
be present. Several explanations for this potential over- and statistically characterised according to the effect
growth have been proposed (Cornu and Flandrois, of each population on the other. Overall, each of these
2000). Most authors proposed that this may result from phenomena (interspecies growth differences and
a higher fitness of L. innocua in selective media inhibitory activity of L. innocua against L. monocy-
(Curiale and Lewus, 1994; MacDonald and Sutherland, togenes) may engender the overgrowth of L. innocua
1994; Beumer et al., 1996). In contrast, others have during enrichment culture and result in the risk of
suggested that this competitive advantage may result false-negatives.
from the production of inhibitory activities in L. inno-
cua (Yokoyama et al., 1998). However, at the present
time, both the circumstances and possible explanations 2. Materials and methods
for this potential overgrowth remain unclear.
Our experiments were designed to evaluate the 2.1. Listeria isolates
effect of interspecies competition in enrichment pro-
cesses and were based on both the ISO 11290-1 method Isolates of L. monocytogenes and L. innocua, as
(Anonymous, 1996) as well as a proprietary culture well as the nature of their respective inhibitors, are
system (BCM L. monocytogenes detection system, listed in Table 1. The inhibitors produced by the
BioSynth, Switzerland; see Restaino et al., 1999). In majority of these isolates have been previously
each case, these enrichment procedures consist of a described (Kalmokoff et al., 1999). Additional strains
primary enrichment in Half Fraser broth or BCM pre- (L. monocytogenes 52 and L. innocua 60), originally
enrichment broth (30
C, 24 h), followed by a second characterised by Begot (1996), were also utilised.
enrichment in Fraser broth or BCM enrichment broth Characterisation of the inhibitor produced by L. inno-
(35
C, 24 h). In the ISO 11290-1 method, enrichment cua 60 was carried out as previously described (Kal-
cultures are then plated onto selective plates (Oxford mokoff et al., 1999).
and Palcam). Those plating media are specific for the Stock cultures were maintained at
196
C in BHI
genus Listeria but it is impossible to distinguish (bioMérieux, France) supplemented with 10% glyc-
between Listeria species. Five presumptive isolates erol. Prior to each experiment, isolates were grown up
are then confirmed as L. monocytogenes by more twice by streaking onto Columbia sheep blood agar
thorough characterisation. The BCM selective/differ- (bioMérieux), and incubated at a temperature of 35
C
ential plating medium is chromogenic: only the virulent for a 24-h period.
species of Listeria (L. monocytogenes and L. ivanovii)
produce the phosphatidylinositol-specific phospholi- 2.2. Growth characteristics of L. innocua and L.
pase C and form turquoise colonies (Restaino et al., monocytogenes isolates
1999). The confirmatory plating medium is based on
the fermentation of rhamnose to distinguish between L. 2.2.1. Growth procedures
monocytogenes and L. ivanovii. Enrichment culturing of Listeria spp. was carried
In this study, the growth characteristics of 13 out using two different methodologies. The first
Listeria isolates (7 L. innocua and 6 L. monocyto- utilised the BCM Listeria detection system (Bio-
 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274 263

Table 1
Characteristics of Listeria isolates utilised in this study
Species Serotype Sourcea Phenotypeb
(A) L. monocytogenes isolates
L. monocytogenes ATCC 19112 1/2c C +P
L. monocytogenes 412 1/2a C ND
L. monocytogenes 413 1/2a C +P
L. monocytogenes ATCC 19116 4c F +P
L. monocytogenes 399 1/2a C ND
L. monocytogenes Scott A 4b C +RP
L. monocytogenes 52 4b E +P

(B) L. innocua isolates

L. innocua 246 NRc E ND
L. innocua 854 NR F ND
L. innocua 239 NR E +P
L. innocua 214 NR F +P
L. innocua 227 NR E +P, +D
L. innocua 743 NR F +A
L. innocua 755 NR F +A
L. innocua 60 NR F NR
Cultures of L. monocytogenes were selected on the basis of sensitivity to specific inhibitors produced by various isolates of L. innocua. Results
for the majority of strains are as previously reported (see Kalmokoff et al., 1999).
a
F: food, E: environmental, C: clinical.
b
+P: phage tails, +D: diffuse zone inhibitor, +A: antibiotic, +RP: replicative bacteriophage, ND: no inhibitory activity detected.
c
NR: not recorded.

Synth) and consisted of a 24-h incubation in a pre- 1350 spectrophotometer (bioMérieux). Growth kine-
enrichment broth, followed by a second 24-h incuba- tics were analysed by exponential regression between
tion in enrichment broth (Restaino et al., 1999). The two limits. Generation times were determined by fitting
second method was based on ISO 11290-1 method model (1) to absorbance data between 0.1 and 1:
(Anonymous, 1996) and consisted of a 24-h incuba-
tion in Half Fraser broth, followed by a 24-h incuba- AbsðtÞ ¼ Absi 2t=s ð1Þ
tion in Fraser broth. Half Fraser and Fraser broths where Abs(t) is the absorbance of the culture; t is the
were specially produced without esculine to avoid time elapsed (minutes) since Abs(t) = Absi; and s is the
interference with turbidimetric measures. In addition generation time (minutes).
to these two standard methodologies, a control proce- Analyses were calculated using Mathematica
dure was also carried out that utilised BHI broth. (Wolfram Research, USA). Estimation of confidence
In all cases, pre-enrichment was carried at a tem- intervals were determined with a risk a of 5%. To
perature of 30
C using 3 ml of each respective check between-species differences, a rank analysis
medium inoculated with 30 ml from a cell suspension (Siegel and Castellan, 1988) was performed for each
adjusted to 1 on the McFarland scale (equivalent to ca. broth. To check between-media differences, an anal-
3108 cells/ml). For the second enrichment, each vial ysis of variance was done using the Statistical Anal-
was inoculated with 30 ml of the equivalent culture ysis System (SAS Institute, USA).
and incubated at a temperature of 35
C. Control
experiments using BHI were incubated for a 24-h 2.3. Effect of inhibitory L. innocua on the growth of L.
period only, at a temperature of 30
C. monocytogenes in enrichment media

2.2.2. Monitoring growth 2.3.1. Differential plate counts

Growth within each respective culture was deter- To distinguish L. monocytogenes colonies from L.
mined turbidimetrically (950 nm) using a Urimat ATB innocua, samples were plated on a modified version
264 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274

of the BCM selective/differential plating medium 2.3.4. Growth modelling

(BioSynth). Since the objective of these experiments All growth models were derived from the logistic
was to distinguish L. monocytogenes from L. innocua growth model with delay, i.e. with a breakpoint at the
in a mixed culture, the addition of selective agents transition between the lag phase and the exponential
required to suppress the growth of competing flora phase (Kono, 1968; Baranyi et al., 1993):
found in food samples was not necessary. The modi-
fied plating medium consisted of a non-selective base 8
>
> x0 ; tVlag
(Columbia, bioMérieux) and the supplements of the >
<
BCM identification medium (BioSynth). Colonies of x
xðtÞ ¼
max ; t > lag
L. monocytogenes (turquoise) were readily distin- >
> xmax
>
:1þ
1 expð
lðt
lagÞÞ
guished from L. innocua (white). x0
ð2Þ
2.3.2. Experimental design
Isolates of L. innocua and L. monocytogenes were
selected from among those listed in Table 1 for a series where x(t) is the cell density (cfu/ml) at time t (h); x0
of experiments designed to allow comparison of the and xmax are the initial and maximum cell densities
growth of each strain in pure and mixed cultures (see (cfu/ml); l is the growth rate (h
1) and lag is the lag
Cornu et al., 1999). The procedure was based on ISO time (h).
11290-1 method (Anonymous, 1996). It was carried We used a logarithmic transformation of Eq. (2).
out over a 2-day period and involved a primary enrich- Moreover, a single model was used to fit simultane-
ment in Half Fraser broth at 30
C (24 h), followed by ously couples of growth curves: one in pure culture and
a secondary enrichment in Fraser broth at 35
C (24 h). one in mixed culture. The reference model was then:
For the primary enrichment, two flasks (pure cultures)
were inoculated with 1.0 ml from the 1 McFarland 8 8
suspension of one strain. A third flask (mixed culture) > > y0p ; t V lagp
>
> >
>
ymax 
>
> >
>
> < y0
Log 1 þ 10
> p
was inoculated with 1.0 ml each from both suspen- >
>
> y
1
>
> pure culture : yðtÞ ¼
p
10 0p
sions. For the secondary enrichment, two flasks (pure >
>
> >
> !
>
> >
> 
cultures) were inoculated with 1.0 ml from the equiv- >
> >
> exp
l ðt
lag Þ ; t > lagp
>
> : p p
<
alent pure primary enrichment. The third flask (mixed 8 y ; t V lag
culture) was inoculated with 1.0 ml each from both >
> >
>
0m m
>
> >
>
>
> >
>
ymax 
pure primary enrichments. Each mixed culture was >
> >
< 10 m
>
> y0m
Log 1 þ
1
then carried out with two parallel control pure cultures. >
> mixed culture : yðtÞ ¼ 10 0my
>
> > !
>
> >
>
>
> >
> 
>
> >
>
: : exp
lm ðt
lagm Þ ; t > lagm
2.3.3. Culturing conditions
Cultures were grown in 225 ml of Half Fraser broth ð3Þ
(bioMérieux) during the primary enrichment, or
Fraser broth during the secondary enrichment, con-
tained within a 250-ml Erlenmeyer flask. The flasks
were incubated in a thermostat-controlled water bath where y(t) is the decimal logarithm of cell density (cfu/
at a temperature of 30 F0.1
C during the primary ml) at time t (h) and y0 and ymax are the decimal
enrichment, and 35F0.1
C during the secondary logarithms of the initial and maximum cell densities
enrichment, and agitated using a magnetic stirrer. (cfu/ml). We added the subscript p to growth parame-
Cultures were periodically sampled through a ster- ters in pure cultures and m to growth parameters in
ile spinal needle (Becton Dickinson, USA), and the mixed culture.
number of viable cells determined by plating two 0.1- Different hypotheses were then successively tes-
ml portions of appropriate dilutions on Columbia agar ted: (1) y0p = y0m; (2) lp = lm; (3) lagp = lagm; (4)
(bioMérieux) for the pure cultures, and on the modi- ymaxp= ymaxm. First, the hypothesis y0p=y0m was tested:
fied BCM medium for the mixed cultures. model (4), embedded in model (3) and constrained by
 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274 265

equality between growth rates in pure and mixed where shift is the time of the shift into decay (h) and d is
cultures, was compared
8 8
to model (3). the decay rate (h
1).
> > y ; t Vlagp All analyses were done using Mathematica (Wolf-
>
> > 0 >
ymax 
>
> >
>
< y0
Log 1 þ 10 ram Research). The risk factor a was 5%.
p
>
> >
1
>
> y
>
> pure culture : yðtÞ ¼ 10 0
>
> > !
>
> >
> 
>
> >
> 2.4. Characterising the impact of a selective advantage
>
> >
: exp
lp ðt
lagp Þ ; t > lagp
>
<
8 y ; t Vlag
>
> >
>
0 m
2.4.1. Culturing conditions
>
> >
>
>
> >
>
ymax 
>
> >
< 10 m A final experiment was carried out with L. innocua
>
> y0
Log 1 þ
1
>
> mixed culture : yðtÞ ¼ 10 0y 239 and L. monocytogenes ATCC 19116. The proce-
>
> >
> !
>
> >
>
>
>
> >
>  dure was similar to the former (see above) except that
>
: >
: exp
lm ðt
lagm Þ ; t > lagm
only mixed cultures were utilised, and the initial
ð4Þ population densities were closer to what would be
where y0 is a common parameter for both curves. found in a food sample. For the primary enrichment,
Such comparisons were based on visual evaluation the flask was inoculated with 1.0 ml from a suspen-
and on an F-test based on the likelihood ratio (Bates sion of each isolate adjusted to 10
3 on the McFarland
and Watts, 1988), as described by Cornu et al. (1999). scale; the initial density populations in Half Fraser
If the result of this first test was that y0p was significant- broth were then 103 cfu/ml instead of 106 cfu/ml. For
ly different from y0m, the second hypothesis lp = lm the secondary enrichment, the flask was inoculated
was tested without taking into account the first hypoth- with 2.25 ml from the primary enrichment (10
2
esis. A model embedded in model (3) and constrained dilution, as in the ISO method).
by equality between growth rates in pure and mixed
cultures was compared to model (3). On the contrary, if 2.4.2. Modelling the relative evolution of both
the first hypothesis had not been rejected, it was populations
accepted for the second test. A model embedded in Growth parameters for each curve were determined
model (4) and constrained by equality between both by fitting the logarithmic transformation of model (2)
initial populations and growth rates in pure and mixed to each growth curve:
cultures was compared to model (4). The same prin- 8 8
ciple was applied to test the third and fourth hypoth- >
>
>
>
>
>
y0mono ; t V lagmono
>
> >
>
eses. So, these tests were conditional as each one was >
>
>
>
>
>

ymax 
>
> < 10 mono
dependent of the result(s) of previous one(s). >
> y0mono
Log 1 þ
1
>
> ymono ðtÞ ¼ 10 y0mono

In case of obvious inhibition, model (3) was not >

> >
> !
>
> >
>
>
> >
> 
adapted. A new model (5) was built, taking into account >
> >
>
>
< : exp
lmono ðt
lagmono Þ ; t > lagmono
>
a shift into exponential
8 8
decay for the mixed culture. 8
> y0innocua ; t V laginnocua
>
> >
> y0p ; t V lagp >
> >
>
>
> >
> >
> >
>

>
> >
>
ymax  >
> >
>
>
> >
> >
> >
> 10 ymaxinnocua
>
> < 10 p >
> < y0
Log 1 þ
1
>
> y
Log 1 þ
1 >
> y
10 0innocua
>
> pure culture : yðtÞ ¼ 0p
10 y0p > yinnocua ðtÞ ¼
> !
>
> > ! >
> >
>
>
> >
> >
> >
> 
>
> >
>  >
> >
> exp
l ðt
lag Þ ;
>
> >
> exp
l ðt
lag Þ ; t > lagp >
> >
> innocua innocua
>
> >
: p p >
> >
:
>
< :
8 t > laginnocua
> y0 ; t V lagm
> > m
> ð6Þ
>
> >
>
!
>
> >
> 
>
> >
>
>
> >
< y0m þ Log exp lm t
lagm Þ ; Such models are based on a logistic deceleration
>
>
>
> mixed culture : yðtÞ ¼
>
> > lagm < t V shift based on the ratio between the density population and
>
> >
>
>
> >
> a maximal density population. To check that both
> >
>
>
>
> > y þ Logexpðl ðshift
lag ÞÞ
>
>
>
: >
: 0m  m m decelerations were simultaneous and resulted from the

Log expðdðt
shiftÞÞ ; t > shift
competition for a common limiting resource, we
ð5Þ proposed a new differential model, based on the
266 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274

hypothesis of a logarithmic deceleration based on the 3. Results

total density population:
8 8
> > 0; t V lagmono 3.1. Growth characteristics of L. innocua and L.
>
> >
>
>
> >
> monocytogenes isolates
>
> >
< l
>
> d x mono mono xmono ðtÞ
>
> ¼
>
> d t > 
>
> >
> Generation times for all 13 isolates in BHI (30
C),
>
> >
> xmono ðtÞ þ xinnocua ðtÞ
>
> >
:  1
; t > lagmono
>
> xmaxtot BCM pre-enrichment (30
C), BCM enrichment (35
<

C), Half Fraser (30
C) and Fraser (35
C) broths
>
> 8
>
> > 0; t V laginnocua were determined by turbidimetric monitoring of pure
>
> >
>
>
> >
> cultures (see Section 2.2). Statistical results are pre-
>
> >
< l
>
> d x mono innocua xinnocua ðtÞ
>
> ¼ sented in Table 2.
>
> d t > !
>
> >
> At 30
C, growth in both BCM pre-enrichment and
>
> >
> xmono ðtÞ þ xinnocua ðtÞ
>
: >
:  1
; t > laginnocua
x
maxtot
Half Fraser broths was significantly slower on average
than in BHI. With the exception of the L. monocyto-
ð7Þ genes strains in Fraser broth, growth on average
where xmaxtot is a common parameter for both curves. improved for all isolates following transfer from pri-
This is the simplest model for joint growth of two mary enrichment broths to BCM enrichment and Fraser
populations (for a review, see Vereecken et al., 2000). broth, which is likely a result of the increased temper-
This differential system cannot be symbolically ature of incubation in these media (35 vs. 30
C).
solved. However, it can be numerically solved, with To detect interspecies differences in terms of gen-
particular values of growth parameters, using the pro- eration times, a non-parametric statistical approach
cedure NDSolve of Mathematica (Wolfram Research). was chosen. For each broth, the 13 generation times

Table 2
Analysis of turbidimetric kinetics: generation times (minutes) and 95% confidence intervals for L. monocytogenes and L. innocua isolates
Strain Brain Heart Pre-enrichment Enrichment Half Fraser Fraser broth
Infusion BCM BCM broth
(A) L. monocytogenes isolates
L. monocytogenes ATCC 19112 43.7F1.0 55.5F0.5 34.4F6.8 86.2F4.7 65.9F8.1
L. monocytogenes 412 40.7F1.4 50.3F1.3 32.2F3.6 73.4F4.4 60.9F5.1
L. monocytogenes 413 43.8F1.0 49.8F1.8 30.2F2.5 76.7F4.3 60.7F6.3
L. monocytogenes ATCC 19116 39.7F1.3 49.8F1.4 30.6F2.6 81.0F3.5 75.7F3.2
L. monocytogenes 399 39.1F1.0 54.0F2.9 35.2F3.7 72.0F3.4 55.9F3.5
L. monocytogenes Scott A 42.0F1.5 61.4F0.8 NR 98.9F1.4 101.6F5.7
Mean (L. monocytogenes) 41.5 53.5 32.5 81.4 70.1
Median (L. monocytogenes) 41.4 52.2 32.2 78.9 63.4

(B) L. innocua isolates

L. innocua 246 41.9F1.4 48.7F1.8 31.4F1.7 78.7F4.3 61.0F5.1
L. innocua 854 42.3F1.9 52.8F1.1 30.9F1.9 74.8F3.5 58.2F4.5
L. innocua 239 41.0F1.5 50.6F2.0 28.8F2.0 74.6F4.9 55.4F5.3
L. innocua 214 40.7F1.2 48.9F2.0 31.0F4.1 71.1F2.1 56.6F3.4
L. innocua 227 43.8F1.0 58.0F2.2 38.5F3.0 83.1F4.0 58.6F5.9
L. innocua 743 47.0F1.1 68.7F1.3 NR 84.4F2.1 63.3F3.2
L. innocua 755 49.4F1.2 66.6F1.5 48.0F1.8 85.5F2.0 65.0F2.0
Mean (L. innocua) 43.7 56.3 34.8 78.9 59.7
Median (L. innocua) 42.3 52.8 31.2 78.7 58.6
Control procedure: 24 h in brain heart infusion at 30
C. ‘‘BCM Listeria (Biosynth)’’ procedure: 24 h in Biosynth pre-enrichment broth at 30
C,
and 24 h in Biosynth enrichment broth at 35
C. ISO procedure: 24 h in Half Fraser broth without esculine at 30
C, and 24 h in Fraser Broth
without esculine at 35
C.
NR: not recorded.
 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274 267

were ordered from the lower (fastest strain) to the entire growth cycle of these isolates in mixed culture
upper (slowest strain), and the sum of L. monocyto- conditions (one isolate of L. monocytogenes plus one
genes ranks calculated. The higher the sum of ranks, isolate of L. innocua), using Half Fraser and Fraser
the slower the L. monocytogenes strains were relative broth. The objective of these experiments was to
to the L. innocua strains. The highest sum of ranks evaluate the growth interaction of inhibitory isolates
was found for Fraser broth, and a rank test detected a of L. innocua on the growth of sensitive isolates of L.
nearly significant ( p = 0.075) difference between the monocytogenes in both primary and secondary enrich-
two species. In all other broths, differences between ment broths.
the generation times between the two species were not For each set of two kinetics (same strain, same
significant. However, based on the sum of ranks broth, one curve in pure culture and one curve in
criterion, we could order the broths from the less to mixed culture), an 8-parameter model (3), modelling
the more advantageous for L. innocua (or disadvanta- both curves simultaneously, was fitted. Following
geous for L. monocytogenes): BHI < BCM pre-enrich- application of this model, a number of different
ment < BCM enrichment < Half Fraser < Fraser. hypotheses were successively tested: (1) y0p = y0m; (2)
lp = lm; (3) lagp = lagm; (4) ymaxp = ymaxm. Growth
3.2. Effect of inhibitory L. innocua on the growth of L. kinetics, as well as fitted models obtained in these
monocytogenes in enrichment media experiments, are presented in Figs. 1 – 4. Growth
parameters and statistical results for all four experi-
Isolates of L. innocua producing a variety of ments are presented in Table 3.
known inhibitors and isolates of L. monocytogenes
sensitive to inhibition by these inhibitors were 3.2.1. L. innocua 743 and L. monocytogenes 399
selected from among those listed in Table 1. A series L. innocua 743 produces a peptide antibiotic dur-
of four experiments were carried out to model the ing exponential growth in both BHI and Fraser broth

Table 3
Growth parameters and 95% confidence intervals for pure cultures (subscript p) of L. innocua (L.i.) and L. monocytogenes (L.m.) in Half Fraser
broth (HFB) and Fraser broth (FB)
y0p lp (h
1) lagp (h) ymaxp y0m lm (h
1) lagm (h) ymaxm
HFB L.i. 743 5.99F0.09 0.62F0.06 2.51F0.63 9.03F0.15 =y0p 0.48F0.06 3.05F0.77 =ymaxm
FB L.i. 743 6.48F0.10 0.72F0.08 2.25F0.45 8.88F0.14 =y0p =lp =lagp =ymaxm
HFB L.m. 399 6.36F0.14 0.72F0.10 3.17F0.62 9.03F0.27 cf. text
FB L.m. 399 6.66F0.08 0.78F0.08 1.00F0.36 8.79F0.10 cf. text

HFB L.i. 227 5.68F0.10 0.62F0.06 3.34F0.65 8.81F0.26 =y0p =lp =lagp 7.82F0.14a
FB L.i. 227 6.44F0.16 0.87F0.07 1.76F0.25 8.72F0.12 =y0p =lp =lagp 8.33F0.08
HFB L.m. 399 5.95F0.08 0.67F0.04 3.50F0.46 8.98F0.16 =y0p =lp 2.81F0.44 8.76F0.12
FB L.m. 399 6.59F0.08 0.89F0.09 1.37F0.34 8.94F0.13 =y0p =lp =lagp 8.41F0.09

HFB L.i. 239 6.08F0.08 0.79F0.07 2.13F0.47 8.90F0.10 =y0p =lp 2.55F0.47 8.54F0.10
FB L.i. 239 6.37F0.10 0.82F0.11 1.64F0.47 8.62F0.18 =y0p =lp =lagp =ymaxm
HFB L.m. 399 6.28F0.08 0.71F0.06 3.04F0.32 8.97F0.14 =y0p =lp =lagp 8.46F0.10
FB L.m. 399 6.62F0.12 0.77F0.10 1.11F0.56 8.97F0.21 =y0p =lp =lagp =ymaxm

HFB L.i. 60 6.42F0.08 0.66F0.04 1.43F0.40 9.19F0.10 =y0p =lp =lagp =ymaxm
FB L.i. 60 7.25F0.10 0.92F0.12 1.51F0.40 9.15F0.10 =y0p =lp =lagp 8.84F0.09
HFB L.m. 52 6.49F0.10 0.65F0.06 1.71F0.49 8.97F0.17 cf. text
FB L.m. 52 6.97F0.12 0.98F0.20 1.52F0.48 8.64F0.12 cf. text
Corresponding parameters in mixed cultures (subscript m) are listed if significantly different from that found in pure cultures. Growth
parameters which do not vary significantly from those in pure culture are shown using an equal (=) sign.
a
All parameters are jointly non-significantly different between pure and mixed cultures.
268 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274

(results not shown). In Half Fraser broth, growth of L. isolate used in the mixed culture (L. monocytogenes
innocua 743 was slower in a mixed culture with L. 399) was sensitive to the 1/2-serospecific inhibitor,
monocytogenes 399 than in pure culture, although the but not to the phage tail (results not shown). The
same was eventually reached (Fig. 1, panel A). In behaviour of both isolates in mixed culture conformed
Fraser broth, the growth of L. innocua 743 was to model (3), and results are presented graphically in
identical in pure and mixed cultures (Fig. 1, panel Fig. 2 (panels A to D). There was no difference within
B). In contrast, the growth of L. monocytogenes 399 any set between growth parameters except that the
was inhibited by L. innocua 743 in both media (see final population densities of both strains were lower in
Fig. 1, panels C and D). The growth kinetics of L. mixed culture than in pure culture (Table 3). The
monocytogenes 399 in mixed cultures (both in Half interaction between L. innocua 227 and L. monocy-
Fraser and Fraser broths) did not satisfactorily fit togenes 399 appeared purely competitive in Half
model (3) but model (5). It was assumed that lag Fraser and Fraser broths.
times and growth rates in mixed cultures were equal to
equivalent parameters derived from pure culture con- 3.2.3. L. innocua 239 and L. monocytogenes 399
trols (Table 3). These results demonstrated in both L. innocua 239 produces a phage tail with a wide
media an obvious inhibition of L. monocytogenes 399 spectrum of activity against isolates of L. monocyto-
by L. innocua 743, without competition. genes. Although the inhibition is readily detectable
using a deferred antagonism plate test (Kalmokoff et
3.2.2. L. innocua 227 and L. monocytogenes 399 al., 1999) in liquid medium, production of detectable
L. innocua 227 produces both a phage tail as well levels requires induction of the SOS response (results
as a yet to be characterised proteinateous inhibitor not shown). There was no difference within any set
with a specificity limited to certain isolates within the between growth parameters except that the final
1/2-serogroup of L. monocytogenes. Both inhibitors population densities of both strains in Half Fraser
are produced in low concentrations during growth in broth were lower in mixed culture than in pure culture
liquid medium (Kalmokoff et al., 1999). The second (Table 3). Mixed cultures of L. innocua 239 and L.

Fig. 1. Growth curves for pure cultures (.) and mixed cultures ( ) in Half Fraser broth (panels A and C) and Fraser broth (panels B and D):
L. innocua 743 (top panels), and L. monocytogenes 399 (bottom panels). y(t) is the decimal logarithm of cell density (cfu/ml).
 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274 269

Fig. 2. Growth curves for pure cultures (.) and mixed cultures ( ) in Half Fraser broth (panels A and C) and Fraser broth (panels B and D):
L. innocua 227 (top panels), and L. monocytogenes 399 (bottom panels). y(t) is the decimal logarithm of cell density (cfu/ml).

monocytogenes 399 in Half Fraser broth demonstrated 3.2.4. L. innocua 60 and L. monocytogenes 52
a competition (Fig. 3, panels A and C), whereas in L. innocua 60 was determined to produce a phage
Fraser broth, there appeared to be no interaction tail in a deferred antagonism plate test (results not
between the two isolates (Fig. 3, panels B and D). shown). No inhibitory activity was detected by direct

Fig. 3. Growth curves for pure cultures (.) and mixed cultures ( ) in Half Fraser broth (panels A and C) and Fraser broth (panels B and D):
L. innocua 239 (top panels), and L. monocytogenes 399 (bottom panels). y(t) is the decimal logarithm of cell density (cfu/ml).
270 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274

spot testing sterile spent culture fluids (BHI) onto an two isolates (one of each species) having obvious
overlay of sensitive isolates, suggesting that prior differences in growth rate in enrichment broths were
induction was required for production of the phage selected. L. monocytogenes ATCC 11916 was deter-
tail in liquid cultures. mined to be the slowest growing L. monocytogenes
For modelling the growth of L. monocytogenes 52, strain, apart from the outlier L. monocytogenes Scott
model (5) had to be chosen since model (3) was not A (see Table 2). Co-culture of this isolate along with
satisfactory. L. monocytogenes 52 was inhibited by L. L. innocua 239 using conditions similar to those
innocua 60 both in Half Fraser broth and in Fraser found in a complete enrichment procedure was carried
broth (Fig. 4, panels C and D). In Half Fraser broth, out. Results are presented in Fig. 5.
growth of L. innocua 60 was identical in pure and It appeared that both species were still in growing
mixed cultures (Table 3 and Fig. 4, panel A). At the phase at the end of the primary enrichment in Half
end of mixed growth in Fraser broth, the final pop- Fraser (0– 24 h, see Fig. 5, panel A), which explained
ulation density of L. monocytogenes 52 was signifi- that no significant lag phase was apparent following
cantly lower as compared to the pure culture (Table 3 transfer into the secondary enrichment (24 –36 h, see
and Fig. 4, panel B). The interaction was then a pure Fig. 5a). Model (6) was fitted without deceleration
inhibition in Half Fraser broth without competition, ( ymaxmono = ymaxinnocua = 1) for the primary enrichment
whereas there was a competition in addition to the and without lag phase (lagmono = laginnocua = 0) for the
inhibition in Fraser broth. secondary enrichment.
Based on estimations of growth parameters of
3.3. Impact of a growth advantage in enrichment model (6) in secondary enrichment, we integrated
medium numerically model (7). A theoretical evolution of
(xmono(t))/(xmono(t)+xinnocua(t)) was then obtained from
In order to assess the possible impact of a known the numerical integration of model (7).
selective advantage on the evolution of two popula- A good correlation was found between data and
tions in an enrichment procedure (anon, ISO, 1999), these theoretical models, indicating that modelling the

Fig. 4. Growth curves for pure cultures (.) and mixed cultures ( ) in Half Fraser broth (panels A and C) and Fraser broth (panels B and D):
L. innocua 60 (top panels), and L. monocytogenes 52 (bottom panels). y(t) is the decimal logarithm of cell density (cfu/ml).
 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274 271

(1993) reported that L. innocua grew faster than L.

monocytogenes in BHI broth, at temperatures less than
40
C. Similarly, Curiale and Lewus (1994) found that
L. innocua was faster than L. monocytogenes in three
different media: Trypticase soy broth and 0.6% yeast
extract, University of Vermont medium, and Fraser
broth. However, both studies were carried out using
only a single strain from each species, and both
neglected to address questions regarding interstrain
growth variability (see Barbosa et al., 1994; Begot et
al., 1996). Studies using a wider selection of isolates
(six of each species; MacDonald and Sutherland, 1994)
detected significant differences between both species in
Listeria enrichment broth, in which the slowest L.
innocua strain was faster than the fastest L. monocyto-
genes strain, but found no differences using Listeria
resuscitation broth or Trypticase soy broth.
We determined the generation times for 13 isolates
of L. monocytogenes and L. innocua using BHI and
four different enrichment media (see Table 2). There
were obvious differences in terms of generation times
among each of the isolates, within the different media.
The longest generation times for both species occurred
in Half Fraser and Fraser broths. This was particularly
Fig. 5. Mixed culture of L. innocua 239 ( ) and L. monocytogenes
evident for L. monocytogenes Scott A, which would be
17 (.) in Half Fraser Broth (0 – 24 h) and Fraser Broth (24 – 36 h).
Panel A: Kinetics of both strains, where y(t) is the decimal logarithm an outlier among L. monocytogenes strains (Barbosa et
of cell density (cfu/ml). Panel B: Evolution of L. innocua pro- al., 1994; MacDonald and Sutherland, 1994; Begot et
portion. Dots: experimental data. Curves: numerical integration of al., 1997). Isolates of L. innocua which also displayed
model (7). slower growth in all of the media included L. innocua
743 and L. innocua 755. However, both isolates pro-
deceleration by a logistic function on the total pop- duce a peptide antibiotic during the exponential growth
ulation was satisfactory, and that both populations phase (results not shown), which may impart some
may have been in competition for a common limiting metabolic cost over the course of the growth cycle.
resource. The relative proportion of L. innocua 239 Differences found in the generation times among
increased dramatically (Fig. 5, panel B), and the both species have been suggested to result from differ-
growth advantage of L. innocua appeared to be ences in susceptibility to inhibition by components
sufficient to explain this evolution. within enrichment broths (MacDonald and Sutherland,
1994). Beumer et al. (1996) observed that increasing
acriflavine concentrations affected both lag and gen-
4. Discussion eration times in isolates of L. monocytogenes, whereas
less significant effects were observed against isolates of
4.1. Selective advantage of L. innocua in enrichment L. innocua. Recent enrichment procedures, such as ISO
media 11290-1, may reduce the inhibitory effect of selective
agents by using a primary enrichment broth (Half
It has become widely accepted that isolates of L. Fraser broth) containing half the amount of acriflavine
innocua have a selective advantage over L. monocyto- (12 mg/l) as found in Fraser broth (24 mg/l).
genes during enrichment culturing, which is generally In contrast to previously reported findings (see
attributed to faster rates of growth. Duh and Schafner above), there were no significant differences found
272 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274

in generation times between the L. monocytogenes 1999). Similar findings have also been reported in
and L. innocua isolates in any given media, although both mitomycin-C and UV-light induced cultures
the L. monocytogenes strains did tend to grow slower (Ortel, 1989; Wilhems and Sandow, 1989; Yokoyama
in Fraser broth (although not significant at an a = 0.05, et al., 1998). This high incidence of inhibitory activity
see Table 2). Although, there are combinations from has indicated the importance of evaluating the impact
among the two species that might result in a selective of these inhibitory factors on the enrichment process
advantage in growth rate for L. innocua during an (Yokoyama et al., 1998).
enrichment process, it would be difficult to assess the In this study, four isolates of L. innocua producing
impact of this given the fact that these determinations three distinct inhibitory activities were utilised to
were carried out using pure cultures without the determine whether the production of these inhibitors
presence of sample material and/or a competing may impart a selective advantage during co-culture
microbial flora (see Beumer et al., 1996). Whether with L. monocytogenes in enrichment broths. Two
significant differences do generally occur between examples of inhibitions were observed and mathe-
both species might be more completely addressed by matically modelled. In the first case, L. innocua 743
examining the growth characteristics of a much clearly inhibited the growth of L. monocytogenes 399
broader number of isolates. in both Half Fraser and Fraser broth (see Fig. 1). This
finding that was not unexpected, as this isolate pro-
4.2. Inhibitory effect of L. innocua against L. mono- duces an inhibitory peptide during exponential growth
cytogenes phase, to which L. monocytogenes 399 is quite sensi-
tive (Kalmokoff et al., 1999).
A variety of inhibitory activities have been reported The second example of inhibition occurred in co-
among isolates of Listeria spp. (Tagg et al., 1976; cultures involving L. innocua 60, a producer of phage
Mollerach et al., 1988; Ortel, 1989; Curtis and Mitch- tails, and L. monocytogenes 52. These findings were
ell, 1992; Lebek et al., 1993; Zink et al., 1995; Yoko- in marked contrast to that found using L. innocua 239
yama et al., 1998; Kalmokoff et al., 1999), and in some (phage tail producer) in co-culture with L. monocyto-
cases, the agents responsible for these activities char- genes 399. In the latter case, the interaction was
acterized (see Zink et al., 1995; Loessner et al., 1995; purely competitive in Half Fraser broth, with no
Yokoyama et al., 1998; Kalmokoff et al., 1999). In interaction in Fraser broth. In both instances, the
most cases, these activities result from the production inhibitors produced by both L. innocua isolates were
of defective bacteriophage particles (phage tails; see characterised using deferred antagonism plate testing,
Zink et al., 1995), which despite being non-replicative and no accompanying inhibitory activity could be
can still kill a sensitive cell. These inhibitors are also detected in spent culture supernatants (BHI, results
referred to as monocins, listeriocins or bacteriocin-like not shown). The production of detectable levels of
inhibitors (Zink et al., 1995). Generally, the production phage tails in culture supernatants requires induction
of phage tails requires induction of the SOS response; of the SOS response in the producer cells, generally
however, in Listeria spp., these are also produced using UV-light or mitomycin-C (Zink et al., 1995;
during growth on plates, and in certain instances in Yokohama et al., 1998). We have no explanation for
liquid media (Bradley and Dewar, 1966; Kalmokoff et the induction of phage tails during the co-culturing of
al., 1999). In addition to phage tails, other isolates of L. innocua 60, nor do we have any further data
both L. innocua and L. monocytogenes also produce indicating whether this may be of common occurrence
replicative bacteriophages (see Loessner et al., 1995) as among other isolates of L. innocua.
well as additional inhibitory proteins and peptides The larger question posed by these findings con-
(Mollerach et al., 1988; Kalmokoff et al., 1999). cerns the possible impact of these inhibitory activities
Many of these inhibitors demonstrate wide inhib- on a given enrichment process. In the case of anti-
itory activities against other Listeria isolates. Despite biotic production, the impact may be lower since only
the nature of the inhibitory activity, a recent survey of approximately 1% of 300 isolates surveyed produced
300 Listeria isolates revealed that 75% produced at this type of inhibitory activity (Kalmokoff et al.,
least one type of inhibitory activity (Kalmokoff et al., 1999). However, many Listeria spp. isolates produce
 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274 273

phage tails, and these could have a significant impact recovery rate of L. monocytogenes was 100% (14
during the enrichment process. Further work in this laboratories out of 14), but dropped to an average of
area is required. 5% (0 to 2 laboratories out of 14) in samples also
containing L. innocua. Our findings have demonstrated
4.3. Overgrowth of L. innocua along the enrichment that a selective advantage resulting from either an
procedure of L. monocytogenes inhibitory activity and/or increased growth rate in L.
innocua may result at the end of an enrichment process
These results have demonstrated that interspecies in the final proportion of L. innocua exceeding greater
inhibition can impact on the overgrowth of L. innocua than 80% of the culture. With usual plating media
in co-culture using enrichment media. In the case of the media (Oxford and Palcam), on which only five pre-
inhibitory interactions in Half Fraser broth, the propor- sumptive isolates are characterised, such an over-
tion of L. innocua 743 and L. innocua 60 increased growth of L. innocua in enrichment cultures certainly
from 28% to 99% and 45% to 99.8%, respectively. In impacts the recovery rate of L. monocytogenes. Newer
Fraser broth, the proportions increased from 54% to plating media, such as ALOA (AES Laboratories,
98% and from 66% to 95%, respectively. In a real France), RAPID L’MONO (Sanofi Diagnostic Pasteur,
enrichment process, this phenomenon might be further France), or BCM (BioSynth), now make it possible to
amplified since the initial ratio in Fraser broth would be distinguish L. monocytogenes colonies from L. innocua
the final population ratio present in the Half Fraser colonies. However, even on these newer plating media,
broth culture. Moreover, since the initial L. monocyto- false negatives remain a possibility for enrichments
genes population densities would be lower, the growth where the final L. innocua level exceeds 99%.
phase would be significantly increased.
In order to assess the impact of a growth advantage
in an enrichment process, an experiment which more 5. Conclusions
closely simulated an actual enrichment process was
designed, and utilised two isolates demonstrating The overgrowth of L. innocua in enrichment broths
differences in terms of generation times in Half Fraser designed for the isolation of L. monocytogenes results
and Fraser broths. The evolution between the ratio of in an increased risk of false negatives. We have
the two populations (L. innocua: 40% of inoculum in demonstrated using the modelling techniques of pre-
primary enrichment) and final ratio (L. innocua: 95% dictive microbiology that this overgrowth can result
at the end of the primary enrichment and 99.5% at the from two strain-dependant phenomena: (i) a growth
end of the secondary enrichment) were explained by advantage of L. innocua over L. monocytogenes; (ii)
the selective growth advantage of L. innocua 239 over the inhibition of L. monocytogenes by L. innocua re-
L. monocytogenes 17. Indeed, kinetic modelling con- sulting from the production of an inhibitory compound.
firmed that the interaction in the enrichment process Although further experiments are required to better
was of a competitive nature. However, it should be characterise these interactions, it is clear that optimisa-
mentioned that the inverse case might also be true for tion of an enrichment/detection procedure must con-
enrichment process where the L. monocytogenes sider the impact of both factors to ensure the success of
present may have a selective growth advantage over any given process. Such an application of predictive
L. innocua strains. microbiology to evaluate a microbial protocol has, to
our knowledge, never been illustrated before.
4.4. Consequences on the recovery rate of L. mono-
cytogenes
Acknowledgements
An interlaboratory study examined the relative
recovery rates of single and mixed L. innocua and L. The authors would like to thank G. Fardel for
monocytogenes cultures from artificially contaminated technical assistance, Drs. I. and A. Lebert who pro-
beef following a two-step enrichment procedure (Curi- vided the strains characterised by Begot (1996), and Dr.
ale and Lewus, 1994). In the absence of L. innocua, the J.M. Farber for access to his culture collection.
274 M. Cornu et al. / International Journal of Food Microbiology 73 (2002) 261–274

We also thank bioMérieux for supporting this European Commission. European Commission. Health and Con-
work. sumer Protection Directorate, 1999. Opinion of the Scientific
Committee on Veterinary Measures Relating to Public Health
on Listeria monocytogenes, http://europa.eu.int/comm/food/fs/
sc/scv/out25
en.pdf.
References Farber, J.M., Peterkin, P.I., 1991. Listeria monocytogenes, a food-
borne pathogen. Microbiol. Rev. 55, 476 – 511.
Kalmokoff, M.L., Daley, E., Austin, J.W., Farber, J.M., 1999. Bac-
Anonymous, 1996. Microbiology of Food and Animal Feeding
teriocin-like inhibitory activities among various species of Lis-
Stuffs — Horizontal Method for the Detection and Enumeration
teria. Int. J. Food Microbiol. 50, 191 – 201.
of Listeria monocytogenes. ISO 11290-1. International Organ-
Kono, T., 1968. Kinetics of microbial cell growth. Biotechnol. Bio-
ization for Standardization, Geneva, Switzerland.
eng. 10, 105 – 131.
Baranyi, J., Roberts, T.A., McClure, P., 1993. A non-autonomous
Lebek, G., Teysseire, P., Baumgartner, A., 1993. A method for
differential equation to model bacterial growth. Food Microbiol.
typing Listeria monocytogenes strains by classification of lister-
10, 43 – 59.
iocins and phage receptors. Zbl. Bakt. 278, 58 – 68.
Barbosa, W.B., Cabedo, L., Wederquist, H.J., Sofos, J.N., Schmidt,
Loessner, M.J., Wendlinger, G., Scherer, S., 1995. Heterologous
G.R., 1994. Growth variation among species and strains of Lis-
endolysins in Listeria monocytogenes bacteriophages: a new
teria in culture broth. J. Food Prot. 57, 765 – 775.
class of enzymes and evidence for conserved holin genes within
Bates, D.M., Watts, D.G., 1988. Nonlinear Regression Analysis and
the siphoviral lysis cassettes. Mol. Microbiol. 16, 1231 – 1241.
Its Applications. Wiley, Chichester.
MacDonald, F., Sutherland, A.D., 1994. Important differences
Begot, C., 1996. Mesure et modélisation de la croissance de Listeria
between the generation times of Listeria monocytogenes and
en milieu liquide et en surface de viande. PhD Thesis of Cler-
List. innocua in two Listeria enrichment broths. J. Dairy Res.
mont Ferrand University.
61, 433 – 436.
Begot, C., Lebert, I., Lebert, A., 1997. Variability of the response of
Mollerach, M.E., Ogueta, S.B., De Torres, R.A., 1988. Production
66 Listeria monocytogenes and Listeria innocua strains to dif-
of linnociucina 819, a bacteriocin produced by Listeria innocua.
ferent growth conditions. Food Microbiol. 14, 403 – 412.
Microbiologica 11, 219 – 224.
Beumer, R.R., Te Giffel, M.C., Anthonie, S.V.R., Cox, L.J., 1996.
Ortel, S., 1989. Listeriocins, monocins. Int. J. Food Microbiol. 8,
The effect of acriflavine and nalidixic acid on the growth of
249 – 250.
Listeria spp. in enrichment media. Food Microbiol. 13, 137 – 148.
Petran, R.L., Swanson, K.M.J., 1993. Simultaneous growth of Lis-
Bradley, D.E., Dewar, C.A., 1966. The structure of phage-like ob-
teria monocytogenes and Listeria innocua. J. Food Prot. 56,
jects associated with non-induced bacteriocinogenic bacteria. J.
616 – 618.
Gen. Microbiol. 45, 399 – 408.
Restaino, L., Frampton, E.W, Irbe, R.M., Schabert, G., Spitz, H.,
Cornu, M., Flandrois, J.P., 2000. Modelling the competitive growth
1999. Isolation and detection of Listeria monocytogenes using
of Listeria monocytogenes and Listeria innocua in enrichment
fluorogenic and chromogenic substrates for phosphatidylinosi-
broths. In: Van Impe, J.F.M., Bernaerts, K. (Eds.), Predictive
tol-specific phospholipase C. J. Food Prot. 62, 244 – 251.
Modelling in Foods — Conference Proceedings. KULeuven/
Siegel, S., Castellan, N.J., 1988. Nonparametric Statistics. McGraw-
BioTeC, Belgium, pp. 172 – 174 (ISBN 90-804818-3-1).
Hill, New York.
Cornu, M., Delignette Muller, M.L., Flandrois, J.P., 1999. Charac-
Tagg, J.R., Dajani, A.S., Wannamaker, L.W., 1976. Bacteriocins of
terization of unexpected growth of Escherichia coli O157:H7 by
gram-positive bacteria. Bacteriol. Rev. 40, 722 – 756.
modeling. Appl. Environ. Microbiol. 65, 5322 – 5327.
Vereecken, K.M., Dens, E.J., Van Impe, J.F., 2000. Predictive mod-
Curiale, M.S., Lewus, C., 1994. Detection of Listeria monocyto-
eling of mixed microbial populations in food products: evalua-
genes in samples containing Listeria innocua. J. Food Prot.
tion of two-species models. J. Theor. Biol. 205, 53 – 72.
57, 1048 – 1051.
Wilhelms, D., Sandow, D., 1989. Preliminary studies on monocine
Curtis, G.W.D., Mitchell, R.G., 1992. Bacteriocin, monocin inter-
typing of Listeria monocytogenes strains. Acta Microbiol. Hung.
actions among Listeria monocytogenes strains. Int. J. Food
36, 235 – 238.
Microbiol. 16, 283 – 292.
Yokoyama, E., Maruyama, S., Katsube, Y., 1998. Production of
Donnelly, C.W., 1999. Conventional methods to detect and isolate
bacteriocin-like-substance by Listeria innocua against Listeria
Listeria monocytogenes. In: Ryser, E.T., Marth, E.H. (Eds.),
monocytogenes. Int. J. Food Microbiol. 40, 133 – 137.
Listeria, Listeriosis, and Food Safety. 2nd edn. Marcel Dekker,
Zink, R., Loessner, M.J., Scherer, S., 1995. Characterization of
New York, USA.
cryptic prophages, monocins. in Listeria and sequence analysis
Duh, Y.S., Schaffner, D.W., 1993. Modeling the effect of temper-
of a holin/endolysin gene. Microbiology 141, 2577 – 2584.
ature on the growth rate and lag time of Listeria innocua and
Listeria monocytogenes. J. Food Prot. 56, 205 – 210.