Emmanuelle Houard

Prevalence and characterization of Campylobacter jejuni from chicken meat sold in French retail outlets.

Campylobacter was detected in 76% of broiler meat products collected in retail outlets during a monitoring plan carried out in France throughout 2009. Campylobacter jejuni was the most prevalent species (64.7% of products being contaminated). The 175 C. jejuni isolates collected were characterized. MLST typing results confirmed substantial genetic diversity as the 175 C. jejuni isolates generated 76 sequence types (STs). The ST-21, ST-45 and ST-464 complexes predominated accounting for 43% of all isolates. A class-specific PCR to screen the sialylated lipooligosaccharide (LOS) locus classes A, B and C showed that 50.3% of the C. jejuni isolates harbored sialylated LOS. The antimicrobial resistance profiles established using a subset of 97 isolates showed that resistance to tetracycline was the most common (53.6%), followed with ciprofloxacin and nalidixic acid (32.9%, and 32.0% respectively). All the tested isolates were susceptible to erythromycin, chloramphenicol and gentamicin. Clear associations were demonstrated between certain clonal complexes and LOS locus classes and between certain clonal complexes and antimicrobial resistance. This work paints a representative picture of C. jejuni isolated from poultry products circulating in France, providing data on STs, LOS locus classes and antibiotic resistance profiles in isolates recovered from products directly available to the consumer.

A simplified method for detecting pathogenic Yersinia enterocolitica in slaughtered pig tonsils

The aim of this study was to collect preliminary data on the carriage of pathogenic Yersinia enterocolitica in slaughtered pigs in France and to test a simplified method for detecting these strains from tonsils. From January to March 2009, 900 tonsil swabs were taken from pigs at one slaughterhouse in Brittany, France. The swabs were vortexed in 10 ml PSB broth, then 1 ml was added to 9 ml ITC broth. The media were incubated for 48 h at 25°C. The PSB enrichment broth was streaked on CIN plates and the ITC enrichment broth on SSDC plates. In addition to the ISO 10273 method, we also streaked ITC enrichment broth on CIN plates. The plates were incubated for 24h at 30°C, and we then streaked a maximum of four typical colonies per plate onto a plate containing chromogenic medium (YeCM), for the isolation of pathogenic Y. enterocolitica isolates. In parallel, biochemical assays were carried out to confirm the identification of the isolates as Yersinia and to determine biotype. After passage on a YeCM plate and biochemical tests, 380 strains were confirmed to be pathogenic Y. enterocolitica. Finally, with the ISO 10273 method, 9.1% (CI(95%) [5.8-12.4]) of tonsil swabs and 60% (CI(95%) [45.4-74.6]) of the batches were positive. With the ITC-CIN method, 14.0% (CI95% [10.7-17.3]) of the tonsil swabs and 68.9% (CI(95%) [54.3-83.5]) of the batches were positive. Identification as pathogenic Y. enterocolitica was confirmed for 97.0% of the typical colonies obtained on the chromogenic medium, YeCM. The most prevalent biotype was biotype 4 (80.5% of the isolates), followed by biotype 3. This study demonstrates that the ITC-CIN method, followed by streaking on YeCM, may be an effective approach to the isolation of pathogenic Y. enterocolitica from tonsil swabs and the recovery of positive samples. This method is less time-consuming than the ISO 10273 method and reduces the number of biochemical tests required for the confirmation of Yersinia identification, through the use of YeCM.

Characterization of Campylobacter spp. transferred from naturally contaminated chicken legs to cooked chicken slices via a cutting board

Campylobacter represents the leading cause of gastroenteritis in Europe. Campylobacteriosis is mainly due to C. jejuni and C. coli. Poultry meat is the main source of contamination, and cross-contaminations in the consumers kitchen appear to be the important route for exposure. The aim of this study was to examine the transfer of Campylobacter from naturally contaminated raw poultry products to a cooked chicken product via the cutting board and to determine the characteristics of the involved isolates. This study showed that transfer occurred in nearly 30% of the assays and that both the C. jejuni and C. coli species were able to transfer. Transfer seems to be linked to specific isolates: some were able to transfer during separate trials while others were not. No correlation was found between transfer and adhesion to inert surfaces, but more than 90% of the isolates presented moderate or high adhesion ability. All tested isolates had the ability to adhere and invade Caco-2 cells, but presented high variability between isolates. Our results highlighted the occurrence of Campylobacter cross-contamination via the cutting board in the kitchen. Moreover, they provided new interesting data to be considered in risk assessment studies.

Prevalence and Distribution of Listeria monocytogenes Serotypes and Pulsotypes in Sows and Fattening Pigs in Farrow-to-Finish Farms (France, 2008)

This study was undertaken to acquire new data on the prevalence of Listeria monocytogenes in sows and fattening pigs in farrow-to-finish pig farms, and to analyze distribution of serotypes and genotypes of the bacterium within farms. Detection of L. monocytogenes was carried out on 730 pooled feces samples from sows in 73 pig farms and on 172 pooled feces samples from fattening pigs in 43 of these farms. Isolates were serotyped and typed by pulsed-field gel electrophoresis. For sows, 46% of the farms and 11% of the samples were positive for L. monocytogenes. A total of 124 isolates were collected and distributed in four serotypes: 1/2a (41%), 1/2b (36%), 4b (21%), and 1/2c (2%). Positive farms harbored one to three serotypes. The genetic diversity was high; 51 genetic profiles were obtained with 25, 16, 9, and 1 for the serotypes 1/2a, 1/2b, 4b, and 1/2c, respectively. Positive farms harbored 1 to 6 genetic profiles. Isolates showing similar genotypes occurred in several farms. For fattening pigs, 25% of the farms and 14.5% of the samples were positive for L. monocytogenes. The 34 isolates belonged to four serotypes: 1/2a (32%), 1/2b (41%), 4b (24%), and 1/2c (3%). They were distributed in 20 genotypes: 6 for 1/2a; 8 for 1/2b, 5 for 4b, and 1 for 1/2c. Similar serotypes and pulsotypes were recovered in sows and fattening pigs from the same farms, suggesting common sources of contamination.

Distribution of serotypes and genotypes of Salmonella enterica species in French pig production

The population of Salmonella found at various stages of pig production in France was characterised to analyse the distribution and spread of Salmonella in the pig production chain. We serotyped and genotyped by PFGE 174 isolates collected from breeding pigs from breeding farms, 163 collected from breeding pigs from production farms, and 325 collected from fattening pigs. Forty-seven serovars and 110 genotypes were identified. The major serovars were S Derby (263 isolates) and S Typhimurium (162 isolates). The percentage of S Derby isolates decreased slightly through the production system (44.3, 41.1 per cent and 36.5 per cent) and 79.1 per cent of the S Derby isolates were distributed in the five genotypes common to all three stages. The percentage of S Typhimurium isolates was high for slaughter pigs (40.8 per cent) and 43 of the 46 S Typhimurium genotypes were only identified at this stage. Distributions of S Derby and S Typhimurium between breeding and fattening pigs were different. S Derby was found throughout the pig production pyramid, suggesting that this serotype may be transmitted by the transfer of animals between herds. The presence of multiple S Typhimurium genotypes in fattening pigs suggests that there were many sources of contamination at this stage, with fattening pigs having higher levels of exposure and/or sensitivity to this serotype.

Investigation of a type C/D botulism outbreak in free-range laying hens in France

ABSTRACT In 2014, a botulism outbreak in a flock of laying hens was investigated in France. In the flock of 5020 hens, clinical signs of botulism occurred at 46 weeks of age. A type C/D botulism outbreak was confirmed using the mouse lethality assay for detection of botulinum toxin in serum and a real-time PCR test to detect Clostridium botulinum in intestinal contents. The disease lasted one week with a mortality rate of 2.6% without recurrence. Botulism in laying hens has rarely been reported. Five monthly visits were made to the farm between December 2014 and May 2015 for a longitudinal study of the persistence of C. botulinum in the poultry house after the outbreak, and to assess egg contamination by C. botulinum. Several samples were collected on each visit: in the house (from the ventilation circuit, the egg circuit, water and feed, droppings) and the surrounding area. Thirty clean and 30 dirty eggs were also swabbed at each visit. In addition, 12 dirty and 12 clean eggs were collected to analyse eggshell and egg content. The samples were analysed using real-time PCR to detect type C/D C. botulinum. The bacterium was still detected in the house more than 5 months after the outbreak, mostly on the walls and in the egg circuit. Regarding egg contamination, the bacteria were detected only on the shell but not in the content of the eggs. Control measures should therefore be implemented throughout the egg production period to avoid dissemination of the bacteria, particularly during egg collection.

Development and Validation of a New Reliable Method for the Diagnosis of Avian Botulism

Liver is a reliable matrix for laboratory confirmation of avian botulism using real-time PCR. Here, we developed, optimized, and validated the analytical steps preceding PCR to maximize the detection of Clostridium botulinum group III in avian liver. These pre-PCR steps included enrichment incubation of the whole liver (maximum 25 g) at 37°C for at least 24 h in an anaerobic chamber and DNA extraction using an enzymatic digestion step followed by a DNA purification step. Conditions of sample storage before analysis appear to have a strong effect on the detection of group III C. botulinum strains and our results recommend storage at temperatures below -18°C. Short-term storage at 5°C is possible for up to 24 h, but a decrease in sensitivity was observed at 48 h of storage at this temperature. Analysis of whole livers (maximum 25 g) is required and pooling samples before enrichment culturing must be avoided. Pooling is however possible before or after DNA extraction under certain conditions. Whole livers should be 10-fold diluted in enrichment medium and homogenized using a Pulsifier® blender (Microgen, Surrey, UK) instead of a conventional paddle blender. Spiked liver samples showed a limit of detection of 5 spores/g liver for types C and D and 250 spores/g for type E. Using the method developed here, the analysis of 268 samples from 73 suspected outbreaks showed 100% specificity and 95.35% sensitivity compared with other PCR-based methods considered as reference. The mosaic type C/D was the most common neurotoxin type found in examined samples, which included both wild and domestic birds.

Occurrence of C. botulinum in healthy cattle and their environment following poultry botulism outbreaks in mixed farms

Ten cattle farms located in an area with a recent history of poultry botulism outbreaks were investigated to evaluate the occurrence of toxigenic C. botulinum in healthy cattle. Environmental samples in the 10 cattle farms and bovine fecal contents in farms with a confirmed environmental contamination were collected. Detection of C. botulinum toxin genes C, D, C/D, D/C and E was performed using real-time PCR. 4.9% (7/143) of the environmental samples collected in the 10 investigated cattle farms were positive for C. botulinum type C/D. Theses samples (boot-swabs in stalls and on pasture and water of a stream) were collected in 3 different farms. One cow dung sample and 3 out of 64 fecal contents samples collected in a single farm were also positive for C. botulinum type C/D. This study demonstrates that cattle are probably indirectly contaminated via poultry botulism in the area and that they can be intermittent carrier of C. botulinum type C/D after poultry botulism outbreaks in mixed farms.

A representative overview of the genetic diversity and lipooligosaccharide sialylation in Campylobacter jejuni along the broiler production chain in France and its comparison with human isolates

Campylobacter jejuni is the most common cause of bacterial gastroenteritis worldwide and is associated with post-infectious neuropathies. Moreover, the chicken reservoir is described as the main source of human infection and C. jejuni sialylated lipooligosaccharides seem to play an important role in the pathogenesis of neuropathies. In this study, MultiLocus Sequence Typing (MLST) and Comparative Genomic Fingerprinting using 40 assay genes (CGF40) were used to describe C. jejuni populations within clinical isolates and a representative collection of isolates from French poultry production. In addition, the sialylation of C. jejuni LOS was assessed. Here, we report high levels of genetic diversity among both chicken and human disease C. jejuni populations. The predominance of the ST-21, ST-45, and ST-464 complexes in chicken isolates and of the ST-21, ST-206, and ST-48 complexes in the clinical isolates was observed as were correlations between some MLST and CGF40 genotypes. Furthermore, some C. jejuni genotypes were frequently isolated among clinical cases as well as all along the broiler production chain, suggesting a potentially high implication of chicken in human campylobacteriosis in France. Finally, the LOS classes A, B and C were predominant within clinical C. jejuni isolates supporting the hypothesis of a benefit in infectivity for C. jejuni isolates showing sialylated LOS.

Diversity of Yersinia enterocolitica isolated from pigs in a French slaughterhouse over 2 years

The pig is one of the main reservoirs of Yersinia enterocolitica strains pathogenic to humans. A description of the Y. enterocolitica population in this reservoir, and accurate discriminatory techniques for typing isolates are needed for prevention, outbreak investigation, and surveillance. This study investigates the genetic diversity of pathogenic Y. enterocolitica isolates obtained from pig tonsils in a French pig slaughterhouse in 2009 (S1) and 2010 (S2). The use of Pulsed‐Field Gel Electrophoresis (PFGE) and MLVA as typing techniques was also compared and evaluated. First, a total of 167 isolates (12 of biotype 3 recovered during S1, and 155 of biotype 4 recovered during S1 and S2) were typed by PFGE using the XbaI enzyme. MLVA was then tested on all the biotype 3 isolates in addition to 70 selected biotype 4 isolates recovered over the 2 years. PFGE generated two specific XbaI‐PFGE profiles for biotype 3 isolates. Nine XbaI profiles were obtained for biotype 4, with a higher diversity (ID = 0.599) than biotype 3 (ID = 0.167). Two out of the nine XbaI profiles were reported during both surveys and at different months. MLVA improved the differentiation between isolates; the index of diversity reached 0.621 and 0.958, respectively, for biotype 3 (three MLVA types) and biotype 4 (32 MLVA types). The MLVA types for biotype 4 differed over the two surveys, but some isolates with different MLVA types were genetically closely related. This study provides an initial evaluation of the genetic diversity of Y. enterocolitica strains isolated from pigs in France. We show that some PFGE profiles are maintained in the pig production sector, and, through MLVA, that part of the Y. enterocolitica population remained genetically close over the two years. MLVA proved its effectiveness as a tool for investigating pathogenic Y. enterocolitica strains isolated from pigs.