Search

Bacillus thuringiensis (Bt), belonging to the Bacillus cereus (Bc) group, is commonly used as a biopesticide worldwide, due to its ability to produce insecticidal protein crystals during sporulation. The use of Bt, especially subspecies aizawai and kurstaki, to control pests such as Lepidoptera generally involves spraying mixtures containing spores and crystals on crops intended for human consumption. Recent studies have suggested that the consumption of commercial Bt strains may be responsible for foodborne outbreaks (FBOs). However, its genetic proximity to Bc strains has hindered the development of routine tests to discriminate Bt from other Bc, especially Bacillus cereus sensu stricto (Bc ss), also responsible for FBOs. Here, to develop tools for the detection and the discrimination of Bt in food, we carried out a genome-wide association study (GWAS) on 286 complete genomes of Bc group strains to identify and validate in silico new molecular markers specific to different Bt subtypes. The analyses led to the determination and the validation in silico of 128 molecular markers specific to Bt, its subspecies aizawai, kurstaki and four previously described proximity clusters associated with these subspecies. We developed a command line tool (https://github.com/afelten-Anses/Bt_typing) based on a 14-marker workflow for in silico Bt identification of a putative Bc genome with the aim of facilitating the discrimination of Bt from other Bc and between Bt subspecies, especially in the context of FBOs. Collectively, these data provide key elements for investigating Bc/Bt-associated FBOs and for monitoring Bt in food.

The aim of this task was to select genes over or highly expressed in various strains within a panel of candidate genes suspected to play a role during B. cereus and C. perfringens pathogenesis. In addition, to gain a general knowledge on gene expression and evaluate the potential of new virulence genes, a whole transcriptomic assay was performed (RNAseq technology) in various growth conditions. This task was subdived into 2 subtasks: Sub-Task 2.2.1. Optimization of growth conditions to be used for gene expression analysis, and Sub-Task 2.2.2. Development of RT-PCR assays and transcriptomic analysis. This task was launched on M5. The growth conditions of the strains, the RNA extraction procedures as well as the development of RNAseq assays were defined and optimized. Protocols for the culture of B. cereus and C. perfringens have been implemented and optimized. Following obtention of the strains, RNA were extracted from B. cereus and C. perfringens, to procede with qRT-PCR and RNAseq assays.

International audience; Although relative air humidity (RH) strongly influences microbial survival, its use for fighting surface pathogens in the food industry has been inadequately considered. We asked whether RH control could destroy Listeria monocytogenes EGDe by envelope damage. The impact of dehydration in phosphate-buffered saline (PBS) at 75%, 68%, 43% and 11% RH on the bacterial envelope was investigated using flow cytometry and atomic force microscopy. Changes after rehydration in the protein secondary structure and peptidoglycan were investigated by infrared spectroscopy. Complementary cultivability measurements were performed by running dehydration–rehydration with combinations of NaCl (3–0.01%), distilled water, city water and PBS. The main results show that cell membrane permeability and cell envelope were greatly altered during dehydration in PBS at 68% RH followed by rapid rehydration. This damage led cells to recover only 67% of their initial volume after rehydration. Moreover, the most efficient way to destroy cells was dehydration and rehydration in city water. Our study indicates that rehydration of dried, sullied foods on surfaces may improve current cleaning procedures in the food industry.

Although relative air humidity (RH) strongly influences microbial survival, its use for fighting surface pathogens in the food industry has been inadequately considered. We asked whether RH control could destroy Listeria monocytogenes EGDe by envelope damage. The impact of dehydration in phosphate-buffered saline (PBS) at 75%, 68%, 43% and 11% RH on the bacterial envelope was investigated using flow cytometry and atomic force microscopy. Changes after rehydration in the protein secondary structure and peptidoglycan were investigated by infrared spectroscopy. Complementary cultivability measurements were performed by running dehydration–rehydration with combinations of NaCl (3–0.01%), distilled water, city water and PBS. The main results show that cell membrane permeability and cell envelope were greatly altered during dehydration in PBS at 68% RH followed by rapid rehydration. This damage led cells to recover only 67% of their initial volume after rehydration. Moreover, the most efficient way to destroy cells was dehydration and rehydration in city water. Our study indicates that rehydration of dried, sullied foods on surfaces may improve current cleaning procedures in the food industry.

International audience; Listeria monocytogenes, a bacterium that is responsible for listeriosis, is a very diverse species. Desiccation resistance has been rarely studied in L. monocytogenes, although it is a stress that is largely encountered by this microorganism in food-processing environments and that could be managed to prevent its presence. The objective of this study was to evaluate the resistance of 30 L. monocytogenes strains to moderate desiccation (75% relative humidity) and evaluate the correlation of such resistance with the strains' virulence, serotype and genotype. The results showed a great heterogeneity of strains regarding their ability to survive (loss of cultivability between 0.4 and 2.0 log). Strains were classified into three groups according to desiccation resistance (sensitive, intermediate, or resistant), and the strain repartition was analyzed relative to serotype, virulence level and environmental origin of the strains. No correlation was found between isolate origin and desiccation resistance. All serotype 1/2b strains were classified into the group of resistant strains. Virulent and hypovirulent strains were distributed among the three groups of desiccation resistance. Finally, a genomic comparison was performed based on 31 genes that were previously identified as being involved in desiccation resistance. The presence of those genes was localized among the genomes of some strains and compared regarding strain-resistance levels. High nucleotide conservation was identified between resistant and desiccation-sensitive strains. In conclusion, the findings regarding the strains of serotype 1/2b indicate potential serotype-specific resistance to desiccation, and thus, to relative humidity fluctuations potentially encountered in food-related environments. The genomic comparison of 31 genes associated to desiccation tolerance did not reveal differences among four strains which have different level of resistance to desiccation.

Clostridium perfringens is both an ubiquitous environmental bacterium and the fourth most common causative agent of foodborne outbreaks (FBOs) in France and Europe. These outbreaks are known to be caused by C. perfringens enterotoxin (CPE) encoded by the cpe gene. However, additional information on the toxin/virulence gene content of C. perfringens has become available in the last few years. Therefore, to understand the enteropathogenicity of this bacterium, we need to describe the toxin and virulence genes content of strains involved in FBOs. In this study, we used a new real-time PCR typing technique based on a comprehensive set of 17 genes encoding virulence factors. The analysis was performed on a collection of 141 strains involved in 42 FBOs in the Paris region. It was combined with whole genome sequence (WGS) phylogenomic reconstruction, based on the coregenome single nucleotide polymorphisms (SNPs) of 58 isolates, representatives of the identified virulence gene profiles. Two or three different virulence gene profiles were detected in 10 FBOs, demonstrating that C. perfringens FBOs may be associated with heterogeneous strains. cpe-positive strains were isolated in 23 outbreaks, confirming the prominent role of CPE in pathogenicity. However, while C. perfringens was the sole pathogen isolated from the incriminated food, the cpe gene was not detected in strains related to 13 outbreaks. This result indicates either that the standard method was not able to isolate cpe+ strains or that the cpe gene may not be the only determinant of the enterotoxigenic potential of C. perfringens strains. Using phylogenomic reconstruction, we identified two clades distinguishing chromosomal cpe-positive from cpe-negative and plasmid-borne cpe. Important epidemiological information was also garnered from this phylogenomic reconstruction that revealed unexpected links between different outbreaks associated with closely related strains (seven SNP differences) and having common virulence gene profiles. This study provides new insight into the characterization of foodborne C. perfringens and highlights the potential of WGS for the investigation of FBOs.

The ability of Listeria monocytogenes to adhere to and persist on surfaces for months or even years may be responsible for its transmission from contaminated surfaces to food products. Hence the necessity to find effective means to prevent the establishment of L. monocytogenes in food processing environments. The aim of this study was to assess, through a fractional experimental design, the environmental factors that could affect the survival of L. monocytogenes cells on surfaces to thereby prevent the persistence of this pathogen in conditions mimicking those encountered in food processing plants: culture with smoked salmon juice or meat exudate, use of two materials with different hygiene status, biofilm of L. monocytogenes in pure-culture or dual-culture with a Pseudomonas fluorescens strain, application of a drying step after cleaning and disinfection (C&D) and comparison of two strains of L. monocytogenes. Bacterial survival was assessed by culture, qPCR to quantify total cells, and propidium monoazide coupled with qPCR to quantify viable cells and highlight viable but non-culturable (VBNC) cells. Our results showed that failure to apply C&D causes cell persistence on surfaces. Moreover, the sanitation procedure leads only to a loss of culturability and appearance of VBNC populations. However, an additional daily drying step after C&D optimises the effectiveness of these procedures to reduce culturable populations. Our results reinforce the importance to use molecular tools to monitor viable pathogens in food processing plants to avoid underestimating the amounts of cells using only methods based on cell culture.

Listeria monocytogenes is one of the main targets of hygiene procedures in the ready-to-eat food industry due to its ability to persist for months or even years in processing plants, where it can contaminate food during processing. The factors associated with persistence are often those that foster growth, which itself depends on food contamination of surfaces. It is therefore essential to experiment by using food soils or media modelling these soils to understand the behaviour of L. monocytogenes on surfaces of food processing plants. Thus, we set up an experimental plan including three physiological parameters characteristic of the behaviour of cells on surfaces, namely spatial distribution, adhesion forces and the physiological state of sessile L. monocytogenes. These were recorded in two food soils: smoked salmon juice and meat exudate. According to our results, the behaviour of L. monocytogenes on stainless steel surfaces is highly dependent on the food soil used. The presence of viable but non-culturable (VBNC) cells was demonstrated using meat exudate, while all viable cells were recovered using smoked salmon juice. Moreover, on the basis of our criteria and after validation with three strains of L. monocytogenes, we showed that smoked salmon juice can be substituted by a modified culture medium, demonstrating that drawbacks associated with the use of food soils can be overcome.

We assessed the impact of industrial cleaning and disinfection (CD) on colony-forming units (CFUs), viable (culturable and viable but non-culturable) cells and on total cells (viable and dead cells). Bacterial loads on polyvinyl chloride (PVC) and stainless steel surfaces in a cutting room at a beef processing plant were determined before and after CD by real-time PCR to quantify cells from successive swabs from surfaces with or without an ethidium monoazide pre-treatment and by CFU counts on tryptone soy agar. Agar contact plates were also applied after CD for comparison. Before CD, total cells reached 5.4 and 4.7 log cells/cm(2), viable cells 4.0 and 4.4 log cells/cm(2) and CFUs 3.1 and 2.9 log CFU/cm(2) on PVC and stainless steel surfaces, respectively. Although CD left surfaces visually clean, it did not lead to a significant reduction in total cells. Significant reductions were only observed on PVC for CFUs: 0.8 log and on stainless steel surfaces for viable cells and CFUs: 0.8 and 1.5 log, respectively. Our results show that CFUs were both more easily detached and killed on stainless steel surfaces than on PVC surfaces. Other important results include the following observations: 1) a single swabbing detached only between 2 and 27% of the actual bacterial load; 2) after CD, the difference between the actual culturable population and the one assessed by one agar contact plate was 1.9 and 2.7 log CFU/cm(2) on PVC and stainless steel surfaces, respectively.

Background: The intestinal microbiota is suspected to play a role in colitis and particularly in inflammatory bowel disease (IBD) pathogenesis. The aim was to compare the fecal microbiota composition of patients with colitis to that of healthy subjects (HS). Methods: fecal samples from 22 active Crohn's disease (A-CD) patients, 10 CD patients in remission (R-CD), 13 active ulcerative colitis (A-UC) patients, 4 UC patients in remission (R-UC), 8 infectious colitis (IC) patients, and 27 HS were analyzed by quantitative real-time polymerase chain reaction (PCR) targeting the 16S rRNA gene. Bacterial counts were transformed to logarithms (Log10 CFU) for statistical analysis. Results: Bacteria of the phylum Firmicutes (Clostridium leptum and Clostridium coccoides groups) were less represented in A-IBD patients (9.7; P = 0.004) and IC (9.4; P = 0.02), compared to HS (10.8). Faecalibacterium prausnitzii species (a major representative of the C. leptum group) had lower counts in A-IBD and IC patients compared to HS (8.8 and 8.3 versus 10.4; P = 0.0004 and P = 0.003). The Firmicutes/Bacteroidetes ratio was lower in A-IBD (1.3; P = 0.0001) and IC patients (0.4; P = 0.002). Compared to HS, Bifidobacteria were less represented in A-IBD and IC (7.9 and 7.7 versus 9.2; P = 0.001 and P = 0.01). Conclusions: The fecal microbiota of patients with IBD differs from that of HS. The phylum Firmicutes and particularly the species F. prausnitzii, are underrepresented in A-IBD patients as well as in IC patients. These bacteria could be crucial to gut homeostasis since lower counts of F. prausnitzii are consistently associated with a reduced protection of the gut mucosa. (Inflamm Bowel Dis 2009)

Pollution of the environment by human and animal faecal pollution affects the safety of shellfish, drinking water and recreational beaches. To pinpoint the origin of contaminations, it is essential to define the differences between human microbiota and that of farm animals. A strategy based on real-time quantitative PCR (qPCR) assays was therefore developed and applied to compare the composition of intestinal microbiota of these two groups. Primers were designed to quantify the 16S rRNA gene from dominant and subdominant bacterial groups. TaqMan® probes were defined for the qPCR technique used for dominant microbiota. Human faecal microbiota was compared with that of farm animals using faecal samples collected from rabbits, goats, horses, pigs, sheep and cows. Three dominant bacterial groups (Bacteroides/Prevotella, Clostridium coccoides and Bifidobacterium) of the human microbiota showed differential population levels in animal species. The Clostridium leptum group showed the lowest differences among human and farm animal species. Human subdominant bacterial groups were highly variable in animal species. Partial least squares regression indicated that the human microbiota could be distinguished from all farm animals studied. This culture-independent comparative assessment of the faecal microbiota between humans and farm animals will prove useful in identifying biomarkers of human and animal faecal contaminations that can be applied to microbial source tracking methods.

The objective of this study was to determine i) if Camembert cheese micro-organisms could be detected in fecal samples after regular consumption by human subjects and ii) the consequence of this consumption on global metabolic activities of the host colonic microbiota. An open human protocol was designed where 12 healthy volunteers were included: a 2-week period of fermented products exclusion followed by a 4-weeks Camembert ingestion period where 2x40 g/day of Camembert cheese was consumed. Stools were collected from the volunteers before consumption, twice during the ingestion period (2nd and 4th week) and once after a wash out period of 2 weeks. During the consumption of Camembert cheese, high levels of Lactococcus lactis and Leuconostoc mesenteroides were measured in fecal samples using real-time quantitative PCR, reaching median values of 8.2 and 7.5 Log(10) genome equivalents/g of stool. For Ln. mesenteroides, persistence was observed 15 days after the end of Camembert consumption. The survival of Geotrichum candidum was also assessed and the fecal concentration reached a median level of 7.1 Log(10) CFU/g in stools. Except a decreasing trend of the nitrate reductase activity, no significant modification was shown in the metabolic activities during this study.

The aim of this study was to evaluate the survival of Lactobacillus rhamnosus R11 and Lactobacillus acidophilus R52 in the human digestive tract and their effects on the microbiota homeostasis. We designed an open human trial including 14 healthy volunteers. A 3-week exclusion period of fermented products was followed by a 12-day consumption period of 4 capsules daily containing 2 × 109L. rhamnosus R11 and 1 × 108L. acidophilus R52, and a 12-day wash-out period. The 2 strains and dominant bacterial groups of the microbiota were quantified by real-time polymerase chain reaction. At the end of the capsule consumption period, high levels of L. rhamnosus R11 were detected in faecal samples from all volunteers, reaching a mean value of 7.1 log10 colony-forming unit (CFU) equivalents/g of stool. L. acidophilus R52 was detected in the stools of only 1 volunteer, reaching a maximum level of 6.1 log10 CFU equivalents/g of stool. Dilution plating enumerations performed in parallel provided less consistent and generally lower levels. No significant effect of capsule consumption was observed on microbiota homeostasis for the dominant faecal populations. Mean values of 8.8, 9.2, 9.9 and 10.6 log10 CFU equivalents/g of stool were obtained for the Clostridium coccoides, Bifidobacterium sp., Bacteroides sp. and Clostridium leptum groups, respectively.

The aim of this study was to determine how quickly the surface of a refrigerated supermarket serve over counter becomes loaded with bacteria. New material made of polyvinyl chloride or stainless steel was placed on the surface on which foodstuffs are displayed for sale. One to three samples per week for 7 weeks were collected on gauze pads. CFUs were counted and total cells were quantified by real-time PCR. “Viable” cells using real-time PCR following pre-treatment with ethidium monoazide were quantified on stainless steel. Attachment strengths were assessed at the end of the experiment by constructing detachment curves. Whatever the material, on day 1 the microbial load reached values near those observed in the following weeks i.e. 103–104 log total cells/cm2. The number of cells deposited in one week was compensated for by the small reduction obtained by cleaning and disinfection (C&D). The mean difference between total and viable cells was 0.54 log CFUs/cm2. A big drop in CFUs following C&D was observed at the beginning of the experiment, despite no visible decrease in the number of viable cells, but the CFU reduction decreased over time. Nevertheless, the low efficiency of C&D on the dominant microbiota did not indicate the fate of pathogenic bacteria on these materials. Our data suggest that dead cells do not adhere quite so well as viable cells. Although no growth was observed and the attached bacterial community cannot therefore be considered a biofilm, attached cells shared certain properties attributed to biofilms i.e. their resistance to C&D increased over time and they followed a biphasic detachment curve.

Our aim was to assess the potential of Escherichia coli O157:H7 to persist in a processing environment. We studied E. coli behaviour under conditions modelling those of meat plants to establish one initial bacterial load that allows persistence and another that does not. Polyurethane coupons (3.5 cm²) were contaminated once with E. coli in meat exudate before being subjected daily to a cleaning product and a disinfectant, both at half the recommended in-use concentrations, and a further soiling with the exudate. This procedure aimed to model what occurs in harbourage sites. Because previous experiments showed that persistence could not be achieved at 15°C (temperature of slaughter halls), we incubated the coupons at 20°C. Viable cells were determined by ethidium monoazide-qPCR (EMA-qPCR). When the first chemical treatment (CT) was applied to 24-hour biofilms with 5.4 log CFU/cm² cells were no longer detectable after the first week. However, on 66-hour biofilms with 6.7 log CFU/cm², after initially decreasing, E. coli numbers reached 6.6 log CFU/cm² and 8.3 log viable cells/cm² on the 11th day. When E. coli was cultured with a Comamonas testosteroni previously shown to increase E. coli biofilm formation, and subjected to CT on alternate days, E. coli stabilized at 4.6 log CFU/cm² before the CT, from the 5th day of the experiment. The killing and detachment effects of the CT decreased over time and PCR quantification detected a resumption of growth after 2 days (CT on alternate days) or 3 days (daily CT). Intracellular pH (pHi) of individual cells was determined during an experiment in which the CT was applied on alternate days. The proportion of cells with no proton gradient towards the environment (pHi ≤ 5.4) increased after the CT as expected. But during the first week of the experiment only, a further increase in this proportion occurred 24 h after the CT, suggesting that some of the surviving viable but non-culturable cells finally died. This study shows that conditions leading to E. coli O157:H7 persistence are not likely to arise when good refrigeration and hygiene practices are applied, and highlights the usefulness of EMA or PMA-qPCR as a complement to CFU determination in studying bacterial survival after cleaning and disinfection.

Background In humans, the intestinal microbiota plays an important role in the maintenance of host health by providing energy, nutrients, and immunological protection. Applying current molecular methods is necessary to surmount the limitations of classical culturing techniques in order to obtain an accurate description of the microbiota composition. Results Here we report on the comparative assessment of human fecal microbiota from three age-groups: infants, adults and the elderly. We demonstrate that the human intestinal microbiota undergoes maturation from birth to adulthood and is further altered with ageing. The counts of major bacterial groups Clostridium leptum, Clostridium coccoides, Bacteroidetes, Bifidobacterium, Lactobacillus and Escherichia coli were assessed by quantitative PCR (qPCR). By comparing species diversity profiles, we observed age-related changes in the human fecal microbiota. The microbiota of infants was generally characterized by low levels of total bacteria. C. leptum and C. coccoides species were highly represented in the microbiota of infants, while elderly subjects exhibited high levels of E. coli and Bacteroidetes. We observed that the ratio of Firmicutes to Bacteroidetes evolves during different life stages. For infants, adults and elderly individuals we measured ratios of 0.4, 10.9 and 0.6, respectively. Conclusion In this work we have confirmed that qPCR is a powerful technique in studying the diverse and complex fecal microbiota. Our work demonstrates that the fecal microbiota composition evolves throughout life, from early childhood to old age.

Enterococci are natural inhabitants of the human gastrointestinal tract and the main Gram-positive and facultative anaerobic cocci recovered in human faeces. They are also present in a variety of fermented dairy and meat products, and some rare isolates are responsible for severe infections such as endocarditis and meningitis. The aim of the present study was to evaluate the effect of Camembert cheese consumption by healthy human volunteers on the faecal enterococcal population. A highly specific real-time quantitative PCR approach was designed and used to type enterococcal species in human faeces. Two species were found, Enterococcus faecalis and Enterococcus faecium, and only the Enterococcus faecalis population was significantly enhanced after Camembert cheese consumption, whereas Escherichia coli population and the dominant microbiota remained unaffected throughout the trial.

International audience; To date, there is significant controversy as to the survival of yogurt bacteria (namely, Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus) after passage through the human gastrointestinal tract. Survival of both bacterial species in human feces was investigated by culture on selective media. Out of 39 samples recovered from 13 healthy subjects over a 12-day period of fresh yogurt intake, 32 and 37 samples contained viable S. thermophilus (median value of 6.3 x 10(4) CFU g(-1) of feces) and L. delbrueckii (median value of 7.2 x 10(4)CFU g(-1) of feces), respectively. The results of the present study indicate that substantial numbers of yogurt bacteria can survive human gastrointestinal transit.

Although the main causes of pathogenicity for toxin-producing bacteria such as Bacillus cereus and Clostridium perfringens are well known today, the mechanisms responsible for the expression of this virulence need, however, more investigations. The emergence of B. cereus as a foodborne pathogen and as an opportunistic pathogen has intensified the need to distinguish strains of public health concern. The pathogenic potential of B. cereus is extremely variable, with some strains being harmless and others lethal. B. cereus possesses several toxin genes located on the chromosome, such as nhe, hbl and cytK. These toxins provide an indication of the strain toxicity potential but are not sufficient, alone, to discriminate hazardous from harmless strains. Indeed, several studies have shown that Nhe production by hazardous strains is variable and that non-pathogenic strains can also produce it in large quantities. B. cereus produces other toxins such as haemolysin II (HlyII), the metalloproteases InhA1, InhA2 and the cell wall peptidase FM (CwpFM), which may also be involved in pathogenicity. To date, the above described determinants were not sufficient to completely explain the virulence of B. cereus and there are currently no markers, either alone or in combination, to differentiate pathogenic from non-pathogenic strains. For C. perfringens, even if 17 different toxins have been described, this bacterial pathogen is currently classified in five toxinic types (A to E) only based on the production of 4 major toxins (alpha, beta, epsilon and iota). The C. perfringens type A that produces CPE toxin is a major cause of food-borne gastroenteritis (FBGE) whereas the role of other virulence factors remains poorly investigated. Several data also indicate that, for these two pathogens, the level of toxin production could play a major role in determining the pathogenic potential of a particular strain, while the presence of specific toxin genes appears to be less discriminating. In addition, although cytotoxicity assays have already been used for classifying bacterial strains according to their toxic potency, further investigation on the toxicity mechanisms involved has been rarely carried out. The objective of this task was to analyse the RNAseq, PCR and cytotoxic data and study potential correlations between gene presence or expression with strain toxicity and virulence. The growth conditions of the strains, the cytotoxicity assays, the RNA extraction procedures as well as the development of RNAseq assays were defined and optimized. The data were analysed to allow gene expression and strain toxicity comparaison.