Item 8. Project Report to Defra.
· The Scientific Objectives as set out in the contract
1. Determine the proportion of C. coli and C. jejuni sub-types isolated from broilers before during and after treatment with the three most commonly used antibiotics in commercial poultry production (e.g. amoxycillin, tylosin and chlortetracycline).
2. Determine whether certain types of C. jejuni or C. coli predominate amongst each species and each sample.
3. Determine the proportion of Campylobacter species and their sub-types isolated from broilers that are resistant to the treatment antibiotic, resistant to a fluoroquinolone (FQ) or resistant to multiple antibiotics before, during and after treatment with three commonly used veterinary antibiotics.
4. Determine the prevalence of FQ-resistant or MAR Campylobacter spp. in broilers from non-treated flocks.
5. Determine the prevalence of FQ-resistant or MAR Campylobacter in samples associated with the broiler house environment of non-treated flocks.
6. To evaluate the likelihood of FQ-resistant strains entering the food chain, determine the proportion of ciprofloxacin-resistant strains that persist for over two weeks after FQ treatment.
7. To evaluate the likelihood of MAR-Campylobacter spp strains entering the food chain, determine the proportion of MAR strains that persist for over two weeks after FQ treatment.
8. Determine resilience of MAR strains isolated from treated and non-treated flocks (obj. 3-5) to triclosan and DEFRA-approved quaternary ammonium compounds and phenolics, and tolerance to heat and drying.
· The extent to which the objectives set out in the contract have been met
All objectives, except 6 and 7, have been met. In objective 6, it had been hoped that the persistence of FQ-resistant Campylobacter for longer than two weeks after FQ treatment could be assessed. Unfortunately, we were unable to assess this objective as the birds grew too large to carry their own weight and had to be euthanased. In addition, the persistence of CipR strains in another flock could not be assessed as the farmer withdrew access to obtain samples. Objective 7 could not be addressed as few MAR campylobacters were isolated in any of samples from any of the flocks and so the proportion of MAR strains that persist post-antibiotic treatment could not be assessed.
· Details of methods used and the results obtained, including statistical analysis (if appropriate)
METHODS
Treatment of commercial flocks with amoxycillin and sampling of chicken flocks. An alert system was put in place whereby the Bristol laboratory was informed when a flock reared in the United Kingdom was about to be treated with amoxycillin for a clinically relevant infection. The laboratory was alerted to the possibility of a flock needing treatment with amoxycillin on two occasions. Flock 4 comprised 5,100 free-range broilers that were treated with a therapeutic dose of amoxycillin (Vetremox; Alpharma Animal Health Ltd., Hampshire, UK) and was initially sampled at 50 days of age. Sampling of this flock was performed until one week post-treatment only, as the birds were sent for slaughter at that time. Flock 10 was a layer flock of 5,600 birds, treated as above and were first sampled at 119 days of age. During the study, alerts were received from a further seven flocks that required antibiotic therapy. All of these were negative for campylobacter. One further flock was tested, but the birds were not treated with antibiotics.
Treatment of experimental flocks with antibiotics and sampling of chicken flocks. Due to the lack of alerts received, birds were purchased from commercial poultry flocks that had previously reared campylobacter-positive birds. Three Flocks, 5, 12 and 13, were experimentally treated with antibiotics at the School of Clinical Veterinary Science, University of Bristol under Category 2 conditions. Flock 5 consisted of 40 birds at 21 d of age, purchased from a free-range organic broiler flock consisting of 7,500 birds in September 2004, the birds were housed as two independent groups of 20 birds. The birds in each group were either treated with a therapeutic dose of chlortetracycline at 40mg/per kg/per day for 6 days as recommended by the manufacturer (Aurogran; Novartis, Hertfordshire, UK) or amoxycillin (Amoxinsol 100; Vétoquinol UK Ltd., Buckingham, UK). Flock 12 consisted of 70 birds brought on-site in October 2005 purchased from a free-range broiler flock consisting of 6,300 birds at 22 d separated into five groups of 14 birds, each housed independently. Flock 12a comprised 14 birds not treated with any antibiotic, Flock 12b comprised birds treated with a therapeutic does of amoxycillin, Flock 12c comprised birds treated with a therapeutic dose of chlortetracycline, Flock 12d comprised birds treated with a therapeutic dose of enrofloxacin (Baytril; Bayer plc, Animal Health Division) and Flock 12e comprised birds treated with a therapeutic dose of tylosin (Tylan; Elanco Animal Health, Basingstoke, UK). Flock 13 consisted of 60 birds brought on-site in May 2006 purchased from a housed corn fed ‘Freedom Foods’ flock consisting of 31,000 birds at 49 d. The birds were separated into four groups of 15 birds, each housed independently. Flock 13a comprised birds not treated with any antibiotic, Flock 13b comprised birds treated with a therapeutic dose of chlortetracycline, Flock 13c comprised birds treated with a therapeutic dose of tylosin, Flock 13d comprised birds treated with a therapeutic dose of enrofloxacin (Baytril).
All samples were transported to the laboratory within 3 h of collection. Freshly voided faecal samples were collected before, during and up to 4 weeks post-treatment and Campylobacter spp. isolated as previously described (Humphrey et al. 2005). For Flock 12, at 4 weeks post-treatment, the caeca were removed post-mortem for analysis of individual chicken’s flora for Campylobacter spp.
Enumeration of antibiotic-resistant campylobacters. The proportion of antibiotic-resistant campylobacters was determined in the faecal samples by replica-plating (using sterile furniture grade velvet). Colonies from the ‘master plate’ of bacteria grown on modified charcoal cefoperazone deoxycholate agar (mCCDA) were transferred onto Mueller-Hinton agar containing CCDA selective supplement (Oxoid), 5 % defibrinated horse blood and 8 µg/mL amoxycillin, 8 µg/mL tetracycline, 8 µg/mL tylosin, 8 µg/mL erythromycin, or 1 µg/mL ciprofloxacin. Replica plates were incubated at 37 °C in a microaerobic atmosphere (5-6 % O2, 3-7 % CO2 and 7 % H2, in a balance of nitrogen (Bolton et al., 1992). for 48 h. The number of colonies growing on the replica plate were compared with those on the ‘master’ plate, and the proportion of resistant strains in the faecal samples was determined by subtraction.
Isolation of Campylobacter spp. from environmental samples. Campylobacter spp. were isolated from environmental samples by enrichment culture using modified Exeter Broth (mEB) as previously described (Humphrey et al., 2005, Jørgensen et al. 2002).
Identification of presumptive colonies as Campylobacter spp. Presumptive campylobacter colonies (up to three per sample) were sub-cultured from mCCDA to Columbia blood agar with 5 % defibrinated horse blood (Oxoid) and incubated at 37 °C for 48 h in a microaerobic atmosphere. Isolates were confirmed as campylobacter using light microscopy for motility and cell morphology and lack of growth in air at 37 °C after 48 h. Confirmed isolates were sent to the Campylobacter and Helicobacter Reference Unit (CHRU) on Amies charcoal transport swabs (Technical Service Consultants Ltd., Heywood, UK) for speciation and typing.
Species identification. Speciation of C. jejuni and C. coli was performed using the method of Best et al. (2003) employing a real-time PCR assay based on the ABI PRISM 7700 Sequence Detection System (Taqman) platform. Isolates that were not identified as Campylobacter spp were subcultured under aerobic conditions and any isolates that were aerotolerant were speciated as either Arcobacter butzleri, A. cryaeroplilus or A. skirrowii according to the multiplex PCR assay of Houf et al. (2000).
Flagellin gene (flaA SVR) sequence typing. Genotyping based on flaA SVR type was performed as follows. DNA templates were generated from chromosomal DNA recovered from boiled whole cell suspensions using PCR primers and protocols described by Nachamkin et al. (1993). Template preparations were purified using the Whatmanâ 96 Well PCR Cleanup Kit, and quantification of the products was performed by electrophoresis on a 1% (w/v) agarose gel. Sequencing primers and protocols for the SVR analysis were as described by Meinersmann et al. (1997). Sequencing was performed in house by standard protocols using a Beckman CEQ800 Genetic Analysis System, and also commercially (K-Biosciences, Hoddesdon, Herts, England). Forward and reverse sequences were aligned and trimmed to a 321 bp region covering the flaA SVR of genome sequenced strain C. jejuni NCTC 11168 using Bionumerics V4.5 software (Applied Maths, Kortrijk, Belgium). The 321 bp sequences were compared with flaA sequences (870 alleles at the time of writing) held in the online Campylobacter FlaA Variable Region Database (http://hercules.medawar.ox.ac.uk/flaA/) and corresponding type names assigned accordingly. The flaA SVR types were designated based on 100% homology to the reference sequences.. SVR sequences that did not correspond to one of the reference sequences in the database were submitted to the database Curator for assignation of a novel sequence type.
Pulsed-field gel electrophoresis (PFGE) and profile analysis. PFGE was performed according to the protocol of Gibson et al. (1994) when it was considered necessary to check any anomalous results. For example, when isolates had the same flaA SVR type but different antibiotic susceptibility patterns (resistance phenotype), or when an isolate was untypable by flaA SVR. PFGE gel profiles were arbitrarily assigned numbers by comparison (visually and by using BioNumerics V4.5) with profiles of other isolates tested as part of a larger collaborative study examining additional flocks (data not shown).
Determination of antimicrobial resistance. Initially, all isolates were screened by the breakpoint screening method of Thwaites et al. (1997) to determine antibiotic susceptibility (no growth) or resistance (growth) using the following concentrations: ampicillin (Amp) 8 µg/mL and 32 µg/mL; chloramphenicol (Chl) 8 µg/mL; gentamicin 4 µg/mL; kanamycin (Kan) 16 µg/mL; neomycin 8 µg/mL; tetracycline (Tet) 8 and 128 µg/mL; nalidixic acid (Nal) 16 µg/mL; ciprofloxacin (Cip) 1 µg/mL; and erythromycin (Ery) 4 µg/mL. Isolates were selected for further study on the basis of species, flaA SVR type and resistance phenotype determined by breakpoint screening. One strain from each campylobacter-positive sample at each treatment phase in each flock was selected, except when more than one phenotype was present within a sample, when one strain of each phenotype was chosen. Bacteria were grown on Mueller-Hinton agar plus 5% horse blood at 37 °C in 7.5% CO2. The agar doubling dilution procedure recommended by the NCCLS Campylobacter Working Group (McDermott et al., 2004) was used throughout the study, as described previously (Griggs et al., 2005), to determine the MICs of selected antibacterial agents and dyes chlortetracycline (CTC), tetracycline , ampicillin, amoxycillin (Amx), ciprofloxacin, nalidixic acid, erythromycin, chloramphenicol, kanamycin, lincospectin, triclosan and ethidium bromide (EtBr). C. jejuni NCTC 11168 and C. coli NCTC 11366 were used as control strains. In addition, the susceptibility of tetracycline and erythromycin +/- the efflux pump inhibitor PAßN (20 µg/mL) was determined for selected strains. Designation of strains as antibiotic susceptible or resistant was made with reference to the guidelines of the British Society for Antimicrobial Chemotherapy and NCCLS, as described previously (Griggs et al., 2005).
PCR detection of bla gene encoding a putative periplasmic β-lactamase Ampicillin-resistant strains were tested for the presence of the gene (Cj0299) shown to encode a putative periplasmic beta-lactamase in C. jejuni NCTC 11168. Bacteria were grown as described above for 48 h. Bacterial colonies were harvested from the agar plate and a suspension prepared in sterile distilled water. The cells were lysed by boiling for 5 min and the lysate centrifuged to remove cell debris. PCR of bla was performed with a reaction mixture volume of 25 mL using PCR Mastermix (ABgene, Epsom, UK), 0.5 mL boiled cell lysate, and 250 nM primers 469 (5’-GAGTATAATACAAGCGGCAC-3’) and 470 (5’-CCAATTCTTCTTGCCACTTC-3’). An initial denaturation was carried out at 94 °C for 5 min, followed by 30 cycles of 94 °C for 30 s, 56 °C for 45 s, and 72 °C for 30 s, with a final extension at 72 °C for 10 min. A DNA amplimer of the correct size (281 bp) observed on a 1 % (w/v) agarose gel indicated the strain was positive for the bla gene. Primers were synthesised commercially (Invitrogen, Paisley, UK).
Detection of b-lactamase activity Representative ampicillin-resistant isolates of each flaA type at each treatment stage, for each flock, were tested for production of b-lactamase. Bacteria were grown as described above for 48 h. Bacterial colonies were harvested from the agar plate and a turbid suspension prepared in 1 mL sterile distilled water. The cell suspension was lysed using a MSE Soniprep 150 microprobe (Sanyo Biomedical, Loughborough, UK) for 2 min, pausing for 30 s after each 30 s sonication. 50 mL of sonicate was added to 10 mL 500 mg/mL nitrocefin (Oxoid) in the well of a microtitre tray, a color change from yellow to red within 5 min indicated a b-lactamase positive strain.
PCR detection of tet(O) and transfer of tetracycline resistance. All isolates inhibited by 16 mg/mL of tetracycline were examined for the presence of tet(O) in cell lysates using primers based upon the sequence published by Manavathu et al., (1998) and amplified from nt513-1146. Furthermore, 13 isolates representative of resistant strains from each flock were examined for their ability to transfer tetracycline resistance to C. jejuni 81116 as described by Taylor et al. (1981). These were examined in parallel to 11 tetracycline resistant isolates from our previous study (Humphrey et al. 2005; Griggs et al. 2005). The DNA sequence of tetO of four isolates from our previous studies and three isolates from Flock 4 were sequenced. These represented isolates with a range of MIC values from 16 - >128 mg/mL of tetracycline and examples of transferable and non-transferable tet(O).
RESULTS: OBJECTIVES 1 - 3.
1. EFFECT OF AMOXYCILLIN TREATMENT
Commercial Flock 4: Campylobacter spp. were isolated from this broiler flock before and during treatment and replica plating indicated that all isolates were Amp resistant (R) (Figure 1). Before treatment 18/19 isolates were C. jejuni flaA34/PFGE profile 1 (hereafter referred to as P1) and two resistotypes were seen: AmpRNalRCipR and AmpRTetRNalRCipR, which were confirmed by MIC testing (Table 1). C. jejuni flaA34 P1 (AmpRTetRNalRCipR) persisted at low numbers during treatment. Four representative isolates were positive for the bla gene and produced b-lactamase. One C. jejuni flaA191/P2 identified before treatment was susceptible to Amp, Amx, and Nal, but TetR by MIC testing. During treatment this type pre-dominated (3/4 isolates) and MIC analysis identified that these isolates were multiple drug resistant (MDR, resistant to three or more unrelated antibacterial agents), with resistance to Amp, Amx, Tet, Nal and Ery. Another isolate, C. jejuni flaA191/AT/P2 was negative for bla but produced b-lactamase. From one week post treatment no campylobacters were detected and only A. cryaerophilus was isolated.