Download Report

Letters in Applied Microbiology ISSN 0266-8254
ORIGINAL ARTICLE
The effect of the native bacterial community structure
on the predictability of E. coli O157:H7 survival in
manure-amended soil
L.S. van Overbeek1, E. Franz2, A.V. Semenov3, O.J. de Vos4 and A.H.C. van Bruggen5
1
2
3
4
5
Plant Research International BV, Wageningen University and Research Centre, Wageningen, the Netherlands
RIKILT-Institute of Food Safety, Wageningen University and Research Centre, Wageningen, the Netherlands
Department of Microbial Ecology, Centre of Ecological and Evolutionary Studies, Groningen University, Haren, the Netherlands
Biological Farming Systems Group, Wageningen University, Wageningen, the Netherlands
Department of Plant Pathology, Emerging Pathogens Institute, IFAS, University of Florida, Gainesville, FL, USA
Keywords
Escherichia coli O157, microbial community,
predictive microbiology.
Correspondence
Eelco Franz, Laboratory for Zoonoses and
Environmental Microbiology, Centre for
Infectious Diseases Control Netherlands,
National Institute for Public Health and the
Environment (RIVM), PO Box 1, 3720 BA
Bilthoven, the Netherlands.
E-mail: [email protected]
2009 ⁄ 2004: received 19 November 2009,
revised 28 January 2010 and accepted 28
January 2010
doi:10.1111/j.1472-765X.2010.02817.x
Abstract
Aims: The survival capability of pathogens like Escherichia coli O157:H7 in
manure-amended soil is considered to be an important factor for the likelihood
of crop contamination. The aim of this study was to reveal the effects of the
diversity and composition of soil bacterial community structure on the survival
time (ttd) and stability (irregularity, defined as the intensity of irregular
dynamic changes in a population over time) of an introduced E. coli O157:H7
gfp-strain were investigated for 36 different soils by means of bacterial PCRDGGE fingerprints.
Methods and Results: Bacterial PCR-DGGE fingerprints made with DNA
extracts from the different soils using bacterial 16S-rRNA-gene-based primers
were grouped by cluster analysis into two clusters consisting of six and 29 soils
and one single soil at a cross-correlation level of 16% among samples per
cluster. Average irregularity values for E. coli O157:H7 survival in the same soils
differed significantly between clusters (P = 0Æ05), whereas no significant difference was found for the corresponding average ttd values (P = 0Æ20). The irregularity was higher for cluster 1, which consisted primarily of soils that had
received liquid manure and artificial fertilizer and had a significant higher
bacterial diversity and evenness values (P < 0Æ001).
Conclusions: Bacterial PCR-DGGE fingerprints of 36 manure-amended soils
revealed two clusters which differed significantly in the stability (irregularity) of
E. coli O157 decline. The cluster with the higher irregularity was characterized
by higher bacterial diversity and evenness.
Significance and Impact of the Study: The consequence of a high temporal
irregularity is a lower accuracy of predictions of population behaviour, which
results in higher levels of uncertainty associated with the estimates of model
parameters when modelling the behaviour of E. coli O157:H7 in the framework
of risk assessments. Soil community structure parameters like species diversity
and evenness can be indicative for the reliability of predictive models describing the fate of pathogens in (agricultural) soil ecosystems.
Introduction
Fresh fruits and vegetables are now recognized to be
important routes of entry for zoonotic human pathogens
into the human food chain (Brandl 2006; Doyle and
Erickson 2008; Franz and van Bruggen 2008b; Lynch et al.
2009). Escherichia coli O157:H7 and Salmonella enterica
are among the zoonotic pathogens most frequently
ª 2010 The Authors
Journal compilation ª 2010 The Society for Applied Microbiology, Letters in Applied Microbiology 50 (2010) 425–430
425
Predictability of E. coli O157:H7 survival
L.S. van Overbeek et al.
associated with foodborne diseases resulting from the
consumption of fresh produce (Sivapalasingam et al.
2004). Although contamination of produce with pathogens like E. coli O157:H7 can occur at various stages
throughout the production and distribution chain,
contamination most likely occurs during the primary production phase when crops are grown in fields amended
with contaminated animal manure and ⁄ or are irrigated
with contaminated surface water (Franz and van Bruggen
2008b). Escherichia coli O157:H7 may become associated
with crop plants via the soil (Solomon et al. 2002; Islam
et al. 2004a,b; Franz et al. 2007) and therefore survival in
soil should be considered as an important factor for the
likelihood of crop contamination.
The fate of micro-organisms introduced into a soil
system most likely depends on both biotic and abiotic
factors (van Veen et al. 1997). Few attempts have been
made to link survival of E. coli O157:H7 with soil physicochemical and biological variables. Nutrient availability
seems to be a key issue in the survival of introduced E. coli
O157:H7 in soil (Mubiru et al. 2000; Vidovic et al. 2007;
Franz et al. 2008a). In addition, pathogen survival capabilities are thought to be a function of the total microbial
community composition because of the competition for
nutrients, the production of inhibitory substances, and
overall density. The best demonstration of the importance
of the native microbial community in pathogen survival is
the significant enhanced persistence and even further
outgrowth of E. coli O157:H7 and Salm. enterica in sterilized manure and soil (Jiang et al. 2002; Semenov et al.
2007). It was concluded that the most likely factor
involved in the enhanced survival of E. coli O157:H7 was
the lowering of soil microbiota complexity, resulting in a
lower functional redundancy (van Elsas et al. 2007).
Besides the determination of the survival time, the
intensity of variation around pathogen survival curves
also is important for risk assessment purposes, as this
variation determines the reliability associated with the
predicted survival time. Recently, the Approximate
Entropy procedure, which calculates the irregularity in the
survival pattern, was used and evaluated for E. coli
O157:H7 survival in manure-amended soil (Semenov
et al. 2008). In predictive microbiology, mathematical
models are used to predict the behaviour of a microbial
population in a particular substrate (environmental or
food), by making use of detailed knowledge about the
type of micro-organism and locally prevailing environmental conditions. These models can subsequently be
used to estimate food safety risks. Although environmental and food substrates are considered to be complex
microbial systems, consisting of various heterogeneous
microbial populations that interact with each other, the
complexity of microbial interactions and implications for
426
pathogen growth and survival have been frequently overlooked (Leroy 2007). Considering this full complexity,
detailed knowledge of these substrates would be required
such as: microbial composition, inoculum levels and the
factors affecting competitive interactions (Powell et al.
2004).
It seems logical that bacteriostasis against invading
species depends on the availability of ecological niches
(competition for nutrients and ⁄ or habitable places) in
soils and ⁄ or on variation in the presence of antagonizing
or predating microbial populations (van Elsas et al. 2007;
Franz and van Bruggen 2008b). We hypothesize that variation in measured survival time and irregularity of an
E. coli O157:H7 strain introduced into soils can be
explained by variation in soil community structure. If
there is a relationship, this would indicate that microbial
community composition is important in risk modelling
of pathogens like E. coli O157:H7 in environmental substrates, something that so far has not been included in
any existing risk model.
Materials and methods
Soil collection
The soils (n = 36) were collected and stored according to
the procedures described earlier (Franz et al. 2008a; Semenov et al. 2008). Fields were treated under different
management regimes, 16 were organically farmed and 20
conventionally, and the soils were categorized into two
classes: sand and loam. Samples were taken at 24 different
locations, from which 12 were paired, i.e. samples from
neighbouring organically and conventionally farmed fields
with the same crop (potato, grassland, sugar beet, wheat or
maize). The water content of all soils was set at 60% of the
water holding capacity before inoculation with E. coli
O157:H7 B6-914 gfp-91 (Fratamico et al. 1997) cells.
Survival of introduced Escherichia coli O157:H7 B6-914
gfp-91 in soils, manure-amended soil DNA extraction
and PCR-DGGE.
The survival of E. coli O157:H7 B6-914 gfp-91 in 36
manure-amended soils was studied as described previously (Franz et al. 2008a). In short, 50 g manure portions
was amended with E. coli O157:H7 B6-914 gfp-91cells
suspended in sterile water to about 108 cells per gram
manure (inoculated manure) or was amended with the
same amount of water (noninoculated manure). Then,
all manure samples were added to 450 g soil portions
and thoroughly mixed reaching final densities of
approximately 107 CFU per gram of dry soil for
inoculated soil-manure mixtures. All soil manure
ª 2010 The Authors
Journal compilation ª 2010 The Society for Applied Microbiology, Letters in Applied Microbiology 50 (2010) 425–430
L.S. van Overbeek et al.
mixtures were transferred to 1–l pots after which pots
were covered with lids that allow ambient air exchange,
and all pots were incubated at 16C in darkness.
No significant difference in survival between gfp-modified and nonmodified derivatives of E. coli O157:H7
strain B6-914 under growth-constraining conditions was
observed (Fratamico et al. 1997). Also, no evidence for
instability of the gfp marker and ⁄ or lack of expression in
the modified host strain under the same conditions was
found (Fratamico et al. 1997; Tombolini et al. 1997;
Vialette et al. 2004).
Escherichia coli O157:H7 B6-914 gfp-91 survival in 36
soil samples was determined by counting fluorescent CFU
numbers on Sorbitol MacConkey (SMAC, Oxoid) agar
with 50 mg ml)1 ampicillin at regular time intervals. Logtransformed survival data of E. coli O157:H7 B6-914 gfp91 CFUs in all 36 soils were used for calculation of the
time to reach the detection limit (ttd) according to a
Weibull decline pattern (Franz et al. 2008a) and stability
of the population decline (irregularity) according to the
Approximate Entropy (ApEn) procedure (Semenov et al.
2008). With respect to the latter, a higher irregularity
value means a less stable decline (i.e. a higher temporal
variation along the decline curve).
Polymerase chain reaction-denaturing gradient gel
electrophoresis (PCR-DGGE) analysis was performed on
noninoculated manure soil DNA extracts. The obtained
PCR-DGGE fingerprints were normalized and then digitized as previously described (Franz et al. 2008a). In
short, DNA was extracted from 300 mg soil portions of
all 36 noninoculated soil-manure mixtures using the
Bio101 Systems FastDNA SPIN kit for Soil (Qbiogene,
Carlsbad, CA, USA) according to the specifications provided by the manufacturer Bacterial 16S rRNA genes were
PCR-amplified from these DNA extracts with primers
U968-GC and L1401 (Felske et al. 1996) using a touchdown thermocycle program (Rosado et al. 1998; Janse
et al. 2004), including a final extension step at 72C for
30 min. PCR bands were separated in a 6% polyacrylamide gel containing a 45–65% denaturing gradient
(100% consists of 7 mol l)1 urea and 40% formamide)
using a DCode DGGE system (Bio-Rad Laboratories,
Hercules, CA, USA), and gels were run for 16 h. Gels
included a marker loaded at three different positions for
normalization of fingerprints in gelcompare II (Applied
maths, St Martens-Latem, Belgium) and later comparisons between gels. Normalized fingerprints were digitized
and used for further statistical analyses.
PCR-DGGE fingerprint analyses and statistics
A dendrogram based on PCR-DGGE band position and
relative intensity was constructed using the Pearson
Predictability of E. coli O157:H7 survival
correlation algorithm of gelcompare II. Significance of
differences of average Shannon diversity (H¢), evenness
(both calculated in CANOCO 4Æ5; Biometris, Wageningen, the Netherlands), ttd (Franz et al. 2008a) and irregularity (Semenov et al. 2008) values between samples that
clustered separately in the dendrogram were calculated by
a two-paired t-tests (Genstat, 10th edn; Rothamsted
Experimental Station, Harpenden, UK). Linear correlation
between Shanon diversity and irregularity values was
calculated by regression analysis.
Results
Bacterial PCR-DGGE fingerprints of in total 36 samples
taken from different agricultural fields across the Netherlands were compared. A total of 118 separate bands from
all 36 PCR-DGGE fingerprints were taken into account,
and the number of bands varied between 31 and 96 per
fingerprint with an average of 77, and band intensities
varied between 0–50% of the total band intensity per
lane. A dendrogram constructed on the basis of PCRDGGE band positions and intensities revealed that there
were three clusters of fingerprints present, distinguishable
at a correlation percentage of 16 (Fig. 1). Cluster 1
comprised six fingerprints, cluster 2 twenty-nine and cluster 3 one. Twenty-five PCR-DGGE fingerprints of cluster
2 had 60% or higher correlations with each other and
four showed lower correlations. PCR-DGGE fingerprints
of group 1 always had lower correlations (<60%) with
each other, indicating that cluster 1 fingerprints diverged
more from each other than those of cluster 2. Some consistency was present in soil type and management regime
in cluster 1 fingerprints, namely four were from sandy
soils and five were from soils farmed under conventional
agricultural regimes. This consistency was not found
among cluster 2 fingerprints (14 from sandy soils and 15
from conventionally farmed soils). Significant differences
(P < 0Æ001) between the averages in Shannon diversity
(3Æ93 for cluster 1 and 3Æ35 for cluster 2) and evenness
values (0Æ74 for cluster 1 and 0Æ55 for cluster 2) were
present between clusters 1 and 2. Species diversity in cluster 1 soils was higher and more evenly distributed than in
cluster 2 soils. The PCR-DGGE fingerprint of cluster 3
(Shannon diversity = 3Æ78; evenness = 0Æ64; irregularity = 0Æ21 and ttd = 98) must be considered as a singleton; i.e. the bacterial community in this soil was
dissimilar from all others. A strong positive correlation
(P < 0Æ001) between Shannon diversity and irregularity
values in 10 samples; three from cluster 1 and seven from
cluster 2 soils was observed. This correlation was not
found for the other 26 samples.
Escherichia coli O157:H7 B6-914 gfp-91 irregularity
values differed significantly (P = 0Æ05) between both
ª 2010 The Authors
Journal compilation ª 2010 The Society for Applied Microbiology, Letters in Applied Microbiology 50 (2010) 425–430
427
Predictability of E. coli O157:H7 survival
L.S. van Overbeek et al.
100
80
60
40
20
Pearson correlation (0·0%–100·0%)
DGGE
29 Vredepeel, organic, sand
7 Veghel, conventional, sand
9 Meterik, conventional, sand
20 Marknesse, conventional, sand
5 Ens, conventional, loam
28 Wieringerwerf, conventional, loam
19 Marknesse, organic, sand
8 Berghem, conventional, sand
24 Oostkapelle, conventional, loam
Cluster 1
27 Wieringerwerf, organic, loam
1 Scherpenzeel, organic, sand
16 Eastermar, conventional, sand
26 Langeweg, conventional, loam
12 Dronten, conventional, loam
3 Middachten, organic, sand
14 Lunteren, conventional, sand
33 Lelystad, conventional, sand
2 Rossum, organic, sand
11 Slootdorp, conventional, loam
30 Vredepeel, conventional, sand
31 Nagele, organic, loam
25 Langeweg, organic, loam
23 Oostkappelle, organic, loam
17 Leimuiden, organic, loam
18 Leimuiden, conventional, loam
13 Lunteren, organic, sand
15 Eastermar, organic, sand
36 Orvelte, conventional, sand
21 Ens, organic, loam
32 Nagele, conventional, loam
10 Meterik, conventional, sand
35 Orvelte, organic, sand
22 Ens, conventional, loam
6 Zwaagdijk, organic, loam
34 Lelystad, conventional, loam
4 Duiven, organic, sand
Cluster 2
Cluster 3
Figure 1 Dendrogram based on Pearson correlation of PCR-amplified bacterial 16S rRNA gene fingerprints (DGGE) from 36 soils. Three clusters
can be distinguished diverging at a 16% correlation level.
dendrogram clusters, with means of 0Æ43 for cluster 1 and
0Æ27 for cluster 2, but ttd values did not differ significantly (P = 0Æ20) between the two clusters (73Æ5 days for
cluster 1 and 80Æ6 days for cluster 2). Lower irregularity
values of samples in cluster 2 vs those in cluster 1 indicate
that the E. coli O157:H7 B6-914 gfp-91 decline in cluster
2 soils was better predictable and less irregular than in
cluster 1 soils. Cluster 1, characterized by a higher irregularity, consisted of 83% (5 ⁄ 6) of conventional soils, where
both liquid manure and synthetic fertilizers had been
applied. No correlation was observed between the survival
428
time (ttd) and the level of irregularity (irregularity)
(r = )0Æ16, P = 0Æ36).
Discussion
A high temporal irregularity can be interpreted as a high
level of entropy, i.e. population behaviour not following a
specific pattern that can be modelled with existing quantitative microbiological models. The consequence of a high
temporal irregularity is a lower accuracy of predictions of
population behaviour. This may subsequently be
ª 2010 The Authors
Journal compilation ª 2010 The Society for Applied Microbiology, Letters in Applied Microbiology 50 (2010) 425–430
L.S. van Overbeek et al.
translated to higher levels of uncertainty associated with
the estimates of model parameters when modelling the
behaviour of E. coli O157:H7 in the framework of risk
assessments. Currently, the mechanism responsible for the
observed relationship between the level of irregularity associated with the population dynamics of E. coli O157:H7
and the microbial composition of the manure-amended
soil is unknown. However, for a subset of soils a positive
correlation between irregularity and Shannon diversity
index was observed. This indicates that a higher bacterial
diversity occasionally result in higher irregularity. It was
previously observed that an ecosystem depending on more
species in food chains of a longer length could be less stable (May 1988). Nevertheless, there seems to be consensus
that a minimum number of species are essential for ecosystem functioning. A larger number of species are probably necessary for maintaining the stability of an ecosystem
with constantly changing environments (De Ruiter et al.
1995; Loreau 2001; Botton et al. 2006). The evident example of a system with changing environments is soil used
for cultivation. Regular changes in the soil characteristics
(application of fertilizers, crop harvesting and ploughing)
can lead to irregular changes in microbial community
structure (Botton et al. 2006). This situation may also
allow new species to successfully survive in a system with
frequent changes in competing species. Therefore, soils
that are under constant intensive pressure by farming procedures more likely display unpredictable behaviour of
microbial communities and will show a lower predictability for enteropathogen survival. Previously, it was
concluded that the level of irregularity in E. coli O157:H7
decline in manure-amended soil was significantly higher
in conventional soils compared to organic soils and that
the variation in irregularity could be well explained (52%)
by the variation in the ratio between oligotrophic and
copiotrophic soil micro-organisms (Semenov et al. 2008).
Possibly, these differences correspond to the differences in
microbial community structure between the two clusters
identified in the present study. Future research could focus
on the identification of microbial populations or species
responsible for the observed differences. The interaction
between microbial groups that are correlated with E. coli
O157:H7 stability and survival should be tested with
E. coli O157:H7 in an experimental set up under the same
or comparable chemical ⁄ physical conditions as present in
soils.
A high temporal irregularity results in a lower accuracy
of predictions of population behaviour. This means a
higher level of uncertainty associated with the estimates
of model parameters, e.g. when modelling the behaviour
of E. coli O157:H7 in the framework of risk assessments.
Soil community structure parameters produced with
molecular fingerprinting techniques, like Shannon
Predictability of E. coli O157:H7 survival
diversity and evenness values, could aid to improve predictive models that describe the fate of pathogens in
(agricultural) soil ecosystems.
Acknowledgements
We thank Ilya Senechkin for assistance in the analyses of
PCR-DGGE fingerprints. This work was supported by the
Technology Foundation STW, applied science division of
NWO and by strategic funding (KB6 programme on Food
safety) from the Dutch ministry of Agriculture, Nature
Conservation and Food safety.
References
Botton, S., van Heusden, M., Parsons, J.R., Smidt, H. and van
Straalen, N. (2006) Resilience of microbial systems towards
disturbances. Crit Rev Microbiol, 32, 101–112.
Brandl, M.T. (2006) Fitness of human enteric pathogens on
plants and implications for food safety. Ann Rev Phytopath
44, 367–392.
De Ruiter, P.C., Neutel, A.M. and Moore, J.C. (1995) Energetics, patterns of interaction strengths, and stability in real
ecosystem. Science 269, 1257–1260.
Doyle, M.P. and Erickson, M.C. (2008) Summer meeting 2007
– The problems with fresh produce: an overview. J Appl
Microbiol 105, 317–330.
van Elsas, J.D., Hill, P., Chronakova, P., Grekova, M., Topalova, Y., Elhottova, D. and Kristuvek, V. (2007) Survival of
genetically marked Escherichia coli O157:H7 in soil as
affected by soil microbial community shifts. ISME J 1,
204–214.
Felske, A., Engelen, B., Nu¨bel, U. and Backhaus, H. (1996)
Direct ribosome isolation from soil to extract bacterial
rRNA for community analysis. Appl. Environ Micobiol 62,
4162–4167.
Franz, E. and van Bruggen, A.H.C. (2008b) Ecology of E. coli
O157:H7 and Salmonella entercia in the primary vegetable
production chain. Crit Rev Microbiol 34, 143–161.
Franz, E., Visser, A.A., Van Diepeningen, A.D., Klerks, M.M.,
Termorshuizen, A.J. and van Bruggen, A.H.C. (2007)
Quantification of contamination of lettuce by GFPexpressing Escherichia coli O157:H7 and Salmonella enterica
serovar Typhimurium. Food Microbiol 24, 106–112.
Franz, E., Semenov, A.V., Termorshuizen, A.J., de Vos, O.J.,
Bokhorst, J.G. and van Bruggen, A.H.C. (2008a)
Manure-amended soil characteristics affecting the survival
of E. coli O157:H7 in 36 Dutch soils. Environ Microbiol 10,
313–327.
Fratamico, P.M., Deng, M.Y., Strobaugh, T.P. and Palumbo,
S.A. (1997) Construction and characterization of Escherichia coli O157:H7 strains expressing firefly luciferase and
green fluorescent protein and their use in survival studies.
J Food Prot 60, 1167–1173.
ª 2010 The Authors
Journal compilation ª 2010 The Society for Applied Microbiology, Letters in Applied Microbiology 50 (2010) 425–430
429
Predictability of E. coli O157:H7 survival
L.S. van Overbeek et al.
Islam, M., Doyle, M.P., Phatak, S.C., Millner, P. and Jiang, X.
(2004a) Persistence of enterohemorrhagic Escherichia coli
O157:H7 in soil and on leaf lettuce and parsley grown in
fields treated with contaminated manure composts or
irrigation water. J Food Prot 67, 1365–1370.
Islam, M., Morgan, J., Doyle, M.P., Phatak, S.C., Millner, P.
and Jiang, X. (2004b) Persistence of Salmonella enterica
serovar typhimurium on lettuce and parsley and in
soils on which they were grown in fields treated with
contaminated manure composts or irrigation water. Foodborne Pathog Dis 1, 27–35.
Janse, I., Bok, J. and Zwart, G. (2004) A simple remedy against
artifactual double bands in denaturing gel electrophoresis.
J Microbiol Meth 57, 279–281.
Jiang, X., Morgan, J. and Doyle, M.P. (2002) Fate of Escherichia coli O157:H7 in manure-amended soil. Appl Environ
Microbiol 68, 2605–2609.
Leroy, F.a.D.V.L. (2007) Modelling microbial interactions in
foods. In Modelling Micro-organisms in Food ed. Brul, S.,
van Gerwen, S. and Zwietering, M. Cambridge, England:
Woodhead Publishing Limited.
Loreau, M. (2001) Microbial diversity, producer-decomposer
interactions and ecosystem processes: a theoretical model.
Proc R Soc Lond B Biol Sci 268, 303–309.
Lynch, M.F., Tauxe, R.V. and Hedberg, C.W. (2009) The
growing burden of foodborne outbreaks due to contaminated fresh produce: risks and opportunities. Epidem Inf
137, 307–315.
Mubiru, D.N., Coyne, M.S. and Grove, J.H. (2000) Mortality of
Escherichia coli O157:H7 in two soils with different physical
and chemical properties. J Environ Qual 29, 1821–1825.
Powell, M., Schlosser, W. and Ebel, E. (2004) Considering the
complexity of microbial community dynamics in food
safety risk assessment. Int J Food Microbiol 90, 171–179.
Rosado, A.S., Duarte, G.F., Seldin, L. and Van Elas, J.D.
(1998) Genetic diversity of niftl gene sequence in Paenibacillus azotofixans strains and soil samples analysed by dena-
430
turing gradient gel elctrophoresis of PCR-amplified gene
fragments. Appl Environ Microbiol 64, 2770–2779.
Semenov, A.V., van Bruggen, A.H.C., van Overbeek, L.,
Termorshuizen, A.J. and Semenov, A.M. (2007) Influence
of temperature fluctuations on Escherichia coli O157:H7
and Salmonella enterica serovar Typhimurium in cow
manure. FEMS Microbiol Ecol 60, 419–428.
Semenov, A.V., Franz, E., Van Overbeek, L., Termorshuizen,
A.J. and Van Bruggen, A.H.C. (2008) Estimating the stability of Escherichia coli O157:H7 survival in manureamended soils with different management histories. Environ Microbiol 10, 1450–1459.
Sivapalasingam, S., Friedman, C.R., Cohen, L. and Tauxe, R.V.
(2004) Fresh produce: a growing cause of outbreaks of
foodborne illness in the United States, 1973 through 1997.
J Food Prot 67, 2342–2353.
Solomon, E.B., Yaron, S. and Matthews, K.R. (2002) Transmission of Escherichia coli O157:H7 from contaminated manure and irrigation water to lettuce plant tissue and its
subsequent internalization. Appl Environ Microbiol 68,
397–400.
Tombolini, R., Unge, A., Davey, M.E., De Bruijn, F.J. and
Jansson, J.K. (1997) Flow cytometric and microscopic
analysis of GFP-tagged Pseudomonas fluorescens bacteria.
FEMS Microbiol Ecol 22, 17–28.
van Veen, J.A., van Overbeek, L.S. and van Elsas, J.D. (1997)
Fate and activity of micro-organisms introduced into soil.
Microbiol Mol Biol Rev 61, 121–135.
Vialette, M., Jandos-Rudnik, A.M., Guyard, C., Legeay, O.,
Pinon, A. and Lange, M. (2004) Validating the use of
green fluorescent-marked Escherichia coli O157:H7 for
assessing the organism behaviour in foods. J Appl Microbiol 96, 1097–1104.
Vidovic, S., Block, H.C. and Korber, D.R. (2007) Effect of soil
composition, temperature, indigenous microflora, and
environmental conditions on the survival of Escherichia coli
O157:H7. Can J Microbiol 53, 822–829.
ª 2010 The Authors
Journal compilation ª 2010 The Society for Applied Microbiology, Letters in Applied Microbiology 50 (2010) 425–430