L10 broth  amended with 10% rumen fluid was used for culturing chicken intestinal microbiota and L10 agar was used for plating and colony screening. The anaerobic incubation medium (AIM)  amended with 10% chicken cecal digesta extract (CecExt) was used for selection of DON-transforming microorganisms. The CecExt was prepared by adding 10 g cecal digesta into 90 ml distill water. The resulting mixture was shaken at 110 rpm at 22°C for 30 minutes and then the supernatant recovered from the mixture was filtrated through a filter (Corning Inc., Corning, New York, USA) with the pore size of 0.22 μm. The media of MRS , RB , VL , and DAM  were tested for the selection of DON-transforming bacteria.
Sample collection and microbial cultures
Intestinal digesta was obtained from Leghorn hens. The chickens were housed on floor with free access to water and a layer diet. All research procedures for using chickens complied with the University of Guelph Animal Care Committee Guidelines. To collect digesta samples, the chickens were euthanized by cervical dislocation and their intestines were removed, placed in plastic bags, and immediately brought into an anaerobic chamber (Coy Laboratory Products Inc., Grass Lake, Michigan, USA) with atmosphere of 95% CO2 and 5% H2. Digesta was removed from the small and large intestine of individual birds and kept separately for selecting bacteria. The crop content was also collected and each sample was generated by combining the crop content from three chickens in the same treatment group. Microbial cultures were established by adding 0.2 g digesta into 1 ml L10 broth and incubated at 37°C for 72 hrs in the anaerobic chamber. This incubation condition was used throughout all experiments unless described otherwise. Microbial subcultures were obtained from inoculation of a fresh medium with 10% initial culture followed by incubation. DON (100 μg ml-1) was included in the media (broth) for all experiments unless otherwise indicated.
DNA extraction, PCR amplification, and DNA sequence analysis
QIAamp® DNA Stool Mini Kit (QIAGEN Canada, Mississauga, Ontario, Canada) was used to extract genomic DNA from digesta or mixed microbial cultures following the manufacturer's instructions. Qiagen DNeasy Tissue Kit was used to extract genomic DNA from pure cultures of bacterial isolates.
The 16S rRNA genes were amplified from genomic DNA of the isolates by PCR using eubacterial primers F8 (5'-AGAGTTTGATCCTGGCTCAG-3') and R1541 (5'-AAGGAGGTGATCCAAGCC-3') as described previously . PCR amplicons were sequenced using primer 16S1100r (5'-AGGGTTGCGCTCGTTG-3'). Partial 16S rDNA sequences corresponding to Escherichia coli 16S rRNA bases 300 to 1050 were compared with the GenBank, EMBI, and DBJI nonredundant nucleotide databases using BLAST analysis. The sequences were also submitted to Ribosomal Database Project (RDP) Classifier for identification of the isolates.
PCR-DGGE bacterial profile analysis
The V3 region of the 16S rRNA genes (position 339 to 539 in the E. coli gene) of bacteria was amplified using primers HDA1-GC (5'-CGC CCG GGG CGC GCC CCG GGC GGG GCG GGG GCA CGG GGG G AC TCC TAC GGG AGG CAG CAG T-3'; the GC clamp is in boldface) and HDA2 (5'-GTA TTA CCG CGG CTG CTG GCA C-3') as described by Walter et al. . PCR reaction mixtures (50 μl) contained 1× PCR buffer (ThermoPol reaction buffer, New England Biolabs, Inc., Pickering, Ontario, Canada), 200 μM of each dNTPs, 0.5 μM of each forward and reverse primers, 4% (v v-1) dimethylsulfoxide (DMSO), 2.5 units of Taq polymerase (New England Biolabs, Inc.), and an appropriate amount of template DNA. The 1× PCR buffer (pH 8.8) is composed of 10 mM KCl, 10 mM (NH4)2SO4, 20 mM Tris-HCl, 2 mM MgSO4, and 0.1% (v v-1) Triton X-100. PCR amplification program consisted of preheating at 94°C for 4 min and 30 cycles of denaturing (94°C, 30 sec), annealing (56°C, 30 sec), and extension (72°C, 2 min) followed by final extension at 72°C for 10 min. The DGGE analysis of PCR amplicons was performed using the Bio-Rad DCode Universal Mutation Detection System (Bio-Rad Canada, Mississauga, ON, Canada). The amplicons were separated in 10% polyacrylamide (acrylamide/bisacrylamide 35.7:0.8) gels containing a 35 to 65% gradient of urea and formamide increasing in the direction of electrophoresis. A 100% denaturing solution consisted of 7 M urea and 40% (v v-1) deionized formamide. The electrophoresis was conducted in 1× TAE buffer with 100 V at 60°C for 16 hr. DNA bands in gels were visualized by silver staining . The number of DNA bands, including the presence and density, were used to determine the richness of bacterial populations. The BioNumerics software (version 3.0, Applied Maths, Sint-Martens-Latem, Belgium) was used for similarity analyses of the profiles as described previously .
Extraction and quantification of DON and DOM-1
The detailed procedures of DON extraction and quantification were described previously . Briefly, DON was extracted from a bacterial culture using acetonitrile. The extracts were dissolved in methanol/water (1:1 in volume) and filtered through a C18 SPE cartridge (Phenomenex, Torrance, CA, USA). The extracts were analyzed for DON and DOM-1 by injecting 20 μl aliquot into an Agilent Zorbax Eclipse XDB-C18 column (4.6 × 150 mm, 3.5 μm) followed by detection with a ThermoFinnigan SpectraSystem UV6000LP detector and a ThermoFinnigan LCQ Deca MS spectrometer. The MS was operated in the positive APCI mode. DON or DOM-1 were quantified on the basis of integrated peak areas using absorbance units (UV) at 218 nm or multiple ion counts (MS) at m/z 231, 249, 267, 279, and 297 for DON and m/z 215, 233, 245, 251, 263, and 281 for DOM-1. These values were compared against UV and MS values taken from calibration curves of authentic DON and DOM-1. The ratio of DON to DOM-1 transformation was calculated as:
Transformation ratio = (DOM-1)/(DON + DOM-1) × 100.
Selection of DON-transforming bacterial isolates
An integrated approach was designed to select DON-transforming bacterial isolates from intestinal digesta samples (Fig. 2). The approach consisted of in vivo enrichment and in vitro selection with combined use of conventional microbiological selection strategies and PCR-DGGE bacterial profiling techniques. The activity of DON transformation and the richness of bacterial populations were the two criteria used for the selection. During the entire process for selecting DON-transforming bacteria, PCR-DGGE bacterial profiles were analyzed after each treatment and used to guide the selection of the bacteria. When a sample exhibited a high activity of DON transformation and a significant reduction in the richness of bacterial populations (species) after a particular treatment, the sample was then used for further bacterial selection.
The first subcultures from LIC and SIC digesta samples of the chickens fed DON-contaminated wheat in the in vivo experiment (Step 1 in Fig. 2) were used as the start cultures (Step 2 in Fig. 2) for the bacterial selection. The LIC start cultures were initially subjected to the antibiotic treatment (Step 3 in Fig. 2). The resulting cultures through the antibiotic selection were then grown in the AIM+CecExt medium to further eliminate unwanted bacteria (Step 4 in Fig. 2). The SIC start cultures were, however, treated only with AIM+CecExt before single colony isolation (Step 5 in Fig. 2).
In vivo enrichment
Twelve 69-week-old Leghorn hens were divided into 2 groups. One group (6 chickens) was fed a layer diet supplemented with clean wheat, the other group with contaminated wheat containing 10 ppm (μg g-1) DON. The trial lasted for two weeks with digesta samples collected on day 7 and 14, respectively.
Bacitracin, carbadox, gentamicin, lincomycin, penicillin G, salinomycin, streptomycin, tylosin, vancomycin, and virginiamycin at different concentrations (Table 1) were used to suppress unwanted bacterial populations during the in vitro selection for DON-transforming bacterial isolates. The antibiotics were initially tested individually, and then in different combinations for their effect on the activity of DON transformation and on the richness of bacterial populations determined by the PCR-DGGE analysis. The concentrations of each antibiotic were selected based on their level in feeding practice for prophylactic use in food animal production. The tested antibiotics were included in L10 broth during the incubation of microbial cultures for the selection.
AIM + CecExt medium-based selection
The AIM+CecExt medium offered an advantage in retaining the activity of DON transformation with a minimum support for the growth of bacterial populations. The medium was therefore used after the antibiotic selection to further reduce unwanted bacterial populations. Briefly, the cultures completely transformed DON to DOM-1 through the antibiotic selection were diluted 10-fold in series in AIM + CecExt followed by incubation for 72 hrs and examined for the activity of DON transformation. The cultures with a highest level of dilution and full activity of DON transformation were further diluted in AIM+CecExt. An aliquot of 0.1 ml of each dilute was plated onto L10 agar for screening colonies and mixed with 0.9 ml of L10 broth to grow a culture for PCR-DGGE bacterial profiles, respectively. Sixty eight single colonies from SIC and 128 colonies from LIC were screened for their ability to transform DON to DOM-1.