Fecal samples were collected from 34 DB and 15 New Zealand White (NZW) Pasturella multocida–free laboratory rabbits acquired at various times from one commercial source. Fecal samples were also collected from six DB rabbits that were acquired from a second vendor and from four rabbits of various breeds that belonged to a local petting zoo.

Fecal pellets were homogenized in tryptic soy broth (Remel, Lenexa, KS) and incubated at 37°C overnight. Each sample was then plated on Rainbow Agar O157 (Biolog, Hayward, CA) and incubated at 37°C for approximately 36 hours. Bacterial colonies were selected on the basis of color. Pink and purple colonies were selected and restreaked on MacConkey or blood agar plates (Remel) or both. All the E. coli isolates were confirmed biochemically and characterized by using API 20E strips (Biomerieux Vitek, Hazelwood, MO). Selected E. coli organisms were serotyped at the Pennsylvania State University E. coli Reference Center (ECRC).

Genomic DNA was extracted from bacteria by using InstaGene Matrix (Bio-Rad Laboratories, Hercules, CA). PCR primers and conditions to detect eae, stx1B, and stx2A and variants have been described previously (6).

Southern blot analysis was performed by using a Shiga toxin 1B (stx1B) probe generated by PCR amplification of EHEC O153:H- DNA to confirm stx1B in EHEC strains isolated from rabbits. Fifteen microliters of amplicon underwent electrophroresis through a 1.3% agarose gel and was transferred onto a Hybond N nylon membrane, as outlined by the manufacturer (Amersham, Piscataway, NJ). DNA was then cross-linked using the UV Stratalinker 1800 (Stratagene, La Jolla, CA). The fixed DNA was subsequently hybridized with the stx1B probe. The probe was labeled with horseradish peroxidase, exposed in the presence of luminol to Hyperfilm-ECL as outlined by the manufacturer (Amersham).

To perform Rep-PCR chromosomal profiling, genomic DNA was extracted from bacteria by using High Pure PCR Template Preparation Kit (Roche Diagnostics GmbH, Mannheim, Germany) and quantified by using a GeneQuant pro (Biochrom Ltd., Cambridge, UK) spectrophotometer. Two primer pairs were used for the amplification reactions (7): REP1R-I (5′-IIIICGICGICATCIGGC-3′), REP2-I (5′-ICGICTTATCIGGCCTAC-3′); and ERIC1R (5′-ATGTAAGCTCCTGGGGATTCAC-3′), ERIC2 (5′-AAGTAAGTGACTGGGGTGAGCG-3′). PCR reaction mixtures were prepared by using puReTaq Ready-To-Go PCR Beads (Amersham) containing 50 pmol of each primer and 1 μL (100 ng) of DNA to a total volume of 25 μL. Rep-PCR reactions were performed by using a Techne Genius (Techne Inc., Princeton, NJ) thermal cycler with the following conditions: for the REP primers, initial denaturation (94°C, 7 min), followed by 30 cycles of denaturation (94°C, 30 s), annealing (40°C, 1 min), and extension (65°C, 8 min). A final extension (65°C, 16 min) completed the cycling protocol. For the ERIC primers, initial denaturation (95°C, 5 min) was followed by 35 cycles of denaturation (92°C, 45 s), annealing (51°C, 1 min), and extension (70°C, 10 min). A final extension (70°C, 20 min) completed the cycling protocol. PCR amplicons were visualized after electrophoresis in a 3% agarose gel and staining with ethidium bromide.

Rabbit fecal samples cultured on Rainbow agar O157 yielded E. coli colonies that ranged in color from pink to purple. No black colonies consistent with E. coli O157 were observed (8).

Biochemical and molecular characterization of the rabbit E. coli strains is summarized in the Table. Most O153 strains were negative for rhamnose and sucrose. Some colonies of the O153:H- EHEC strain isolated from the first outbreak of HUS in DB rabbits (6) and a strain of unknown O serotype that showed autoagglutination were sorbitol negative like most O157:H7 EHEC organisms isolated from humans. In addition, two O145:H- EPEC were sorbitol negative.

Seven (25%) of 28 DB rabbits acquired from one commercial source were positive for eae+, stx1+ E. coli, including serotypes O153:H- and O153:H7. One (14%) of the 7 rabbits was infected with an EHEC of unknown O serotype that showed autoagglutination (DB 02-177, isolate 03-192) and was co-infected with enteropathogenic E. coli (EPEC) O145:H-. Eleven (39%) of the 28 DB rabbits were positive for EPEC O145H- or O145:H7. In addition, 1 (9%) of 11 NZW rabbits from the same source was positive for eae-, stx1+ O153:H-, and O153:H7 E. coli. All the E. coli strains were negative for stx2 and variants. In contrast, six DB rabbits from another commercial source and all four rabbits from the petting zoo were negative for EHEC and EPEC. Almost all of the eae-, stx1- E. coli belonged to serotype O7:H7.

The presence of the stx1B amplicon was confirmed in O153:H- and O153:H7 E. coli. The O145:H- and O145:H7 isolates were confirmed to be negative for stx1B (Figure 1).

Figure 2 shows the results obtained by Rep-PCR by using two different sets of primers, REP and ERIC. All the rabbit O153:H– and O153:H7 Stx–producing E. coli organisms tested, including eae- isolates from a NZW rabbit, produced identical amplification patterns that differed from those produced by two other O153:H- EHEC strains isolated from humans. Similarly, all the rabbit O145:H- and O145:H7 EPEC isolates produced identical amplification patterns that differed from those produced by a human and a bovine O145:H- EHEC. The EHEC isolated from the rabbit infected with EPEC O145:H- had an amplification pattern similar to that produced by the O153 strains.