Paul Ebner*, Kimberly Garner and Alan Mathew. Food Safety Center of Excellence, University of Tennessee, 2505 River Drive, Knoxville, TN 37996, USA

Transcription

1 Journal of Antimicrobial Chemotherapy (2004) 53, DOI: /jac/dkh192 Advance Access publication 29 April 2004 Class 1 integrons in various Salmonella enterica serovars isolated from animals and identification of genomic island SGI1 in Salmonella enterica var. Meleagridis Paul Ebner*, Kimberly Garner and Alan Mathew Food Safety Center of Excellence, University of Tennessee, 2505 River Drive, Knoxville, TN 37996, USA Received 15 August 2003; returned 13 November 2003; revised 17 February 2004; accepted 22 February 2004 Objectives: To determine the prevalence of integron-mediated antibiotic resistance in a diverse sample set of Salmonella enterica isolated from animals. Materials and methods: Multiplex PCR was used to detect class 1 integron gene sequences, and integron gene cassettes were identified by PCR mapping. Susceptibility to 18 antibiotics or antibiotic combinations commonly used in either human or veterinary medicine was measured using a microdilution method, and statistical comparisons of the frequency of resistance between groups were made using Fisher s two-sided probability test. Genotypic comparisons of isolates were made following pulsed-field gel electrophoresis of genomic DNA. Results: Thirty-two (30.8%) of 104 isolates contained class 1 integron sequences. Integron-positive isolates represented 15 different S. enterica serovars, were obtained from nine different animal species and had a higher frequency of non-integron-mediated antibiotic resistance (P < 0.05) compared with integron-negative isolates. One non-typhimurium isolate (S. enterica Meleagridis) contained an SGI1 genomic island, including the antibiotic resistance gene cluster. Conclusions: These data demonstrate that integron-mediated antibiotic resistance is common among diverse Salmonella serovars, many of them rare. In addition, SGI1 is not limited to Salmonella enterica Typhimurium DT104 or other commonly isolated serovars. Keywords: antibiotic resistance, integrons, Salmonella, SGI1 Introduction Many clinical cases of antibiotic resistance in bacteria involve the acquisition of exogenous DNA and a vast amount of research has led to the characterization of several mobile genetic elements used by bacteria for the horizontal acquisition of antibiotic resistance genes. 1 3 Many of these genes are present on integrons: genetic structures that allow the site-specific incorporation and expression of foreign antibiotic resistance genes. 4 6 Previous studies in our laboratory found that integrons were common among non-haemolytic Escherichia coli isolated from apparently healthy swine (P. D. Ebner, P. Cullen & A. G. Mathew, 2002, unpublished data). The present study examined integrons in various Salmonella enterica serovars isolated from animals or their environments. The isolates were obtained from several different species of animal, representing livestock, companion animals and exotics. As one potential health hazard associated with the veterinary use of antibiotics is the transmission of antibiotic-resistant pathogens from animals to humans, this study examined integron-mediated antibiotic resistance in a genus of bacterium commonly found in animals and whose subtypes are all pathogenic to humans. 2,7,8 Materials and methods Sample isolates Salmonella isolates (n = 104) were kindly provided by Dr F. Ann Draughon of the Food Safety Center of Excellence at the University of Tennessee (Knoxville, TN, USA). Isolates were collected at a veterinary hospital and the following serovars were represented: Agona (1), Albany (1), Anatum (1), Anjona (1), Arizona (5), Bardo (1), Bareilly (1), Bern (1), Berta (1), Bietri (1), Bleadon (1), Blockley (1), Brandenberg (1),... *Corresponding author. Present address: Louisiana State University Health Sciences Center, 1501 Kings Highway, PO Box 33932, Shreveport, LA , USA. Tel: ; Fax: ; pebner@lsuhsc.edu Present address: Department of Biology, Emory University, 1510 Clifford Road, Rollins Research Room 1166, Atlanta, GA 30322, USA JAC vol.53 no.6 The British Society for Antimicrobial Chemotherapy 2004; all rights reserved.

2 Antibiotic resistance integrons in Salmonella Table 1. Primer pairs for MP-PCR, integron-pcr and SGI1 mapping experiments Name Sequence (5 3 ) Target PCR product size (bp) (1) s407 (f) atcagacgtcgtggatgtcg sul1 346 s753 (r) cgaagaaccgcacaatctcg (2) i965 (f) ccttcgaatgctgtaaccgc inti1 254 i1219 (r) acgcccttgagcggaagtatc (3) q024 (f) gagggctttactaagcttgc qace q224 (r) atacctacaaagccccacgc (4) ntf2 (f) acaccgtggaaacggatgaag integrated gene cassettes n/a qcr2 (r) accgattatgacaacggcgg (5) ntf2 (f) acaccgtggaaacggatgaag inti1 aada 405 antr (r) tatcgctgtatggcttcaggc (6) antf (f) tcagcccgtcttacttgaagc aada qace qcr2 (r) accgattatgacaacggcgg (7) sulf (f) ctggtggttatgcactcagc sul1 flor 919 flor (r) tattccatcgccagtgaagcg (8) flof (f) gcctgatagtcagtatcgtcg flor tetr 655 tetrr (r) gacgcaaatacgctttctctgc (9) tetrf (f) ctcgcttgttctggattagcc tetr tetg 550 tetgr (g) gaatccgaaagctgtccaagc (10) ntf2 (f) acaccgtggaaacggatgaag inti1 pse1 543 pser (r) ttaacgggaagcgctgattgc (11) psef (f) actacgttcagtattgccggc pse1 qace qcr2 (r) accgattatgacaacggcgg n/a, not applicable. California (1), Cerro (1), Cholerasuis (1), Derby (1), Dublin (1), Enteriditis (1), Give (1), Hadar (1), Havana (1), Heidelburg (1), Indiana (1), Infantis (1), Istanbul (1), Java (1), Kentucky (1), Kintambo (1), Lille (1), Loma-Linda (1), Marina (1), Mbdanka (1), Meleagridis (1), Montevideo (1), Muenster (1), Newbrunswick (1), Newington (1), Newport (1), Oranienburg (1), Parera (1), Pomona (1), Reading (1), Rubislaw (1), Sachsenwald (1), St Paul (1), Schwartzengrund (1), Senftenberg (1), Tennessee (1), Thomasville (1), Thompson (1), Typhimurium (1), Typhisuis (1), Waral (1), Weltevrden (2), Widemarsh (1), Uganda (1), 4,5,12:non-motile (1), 4,12 monophasic (1), 9,12:non-motile (1), 45:G,Z51 (1), 47:D-Z39 (1) and untypeable (1). Those isolates that were only serogrouped represented the following serogroups: polya (2), B (18), polyb (1), C1 (3), C2 (3), D (3), D1 (1), E (1), E1 (4). Isolates were obtained from 18 different animal species. When needed, serovars were confirmed by the United States Animal and Plant Health Inspection Service (APHIS; Ames, IA, USA). Class 1 integron detection Integrons were detected using a multiplex PCR (MP-PCR) targeting three conserved sequences of class 1 integron (qace 1, inti1 and sul1). Primer pairs were designed using published sequences (GenBank accession no. AF261825) and purchased from a commercial source (Operon, Inc., Alameda, CA, USA) (Table 1). Template DNA was prepared by boiling overnight cultures. 9 Boiled cultures were cooled on ice for 5 min and 1 µl volumes were used immediately for PCR in a Mastercyler Gradient (Eppendorf, Westbury, NJ, USA) thermocycler with the following cycling: (i) one cycle of 94 C for 4 min; (ii) 10 touchdown cycles of 94 C for 1 min, 65 C for 30 s (decreasing 1 C/cycle), 70 C for 2 min; (iii) 24 cycles of 94 C for 1 min, 55 C for 30 s, 70 C for 2 min; and (iv) one final cycle of 70 C for 5 min. S. enterica Typhimurium DT104, known to contain two class 1 integrons, was used as a positive control. 10 Reaction products were separated by agarose gel electrophoresis and stained with ethidium bromide for visualization. Antibiotic susceptibility testing Each isolate was tested for susceptibility to 18 antibiotics or antibiotic combinations commonly used in either human or veterinary medicine (amikacin, ampicillin, co-amoxiclav, apramycin, cefoxitin, ceftiofur, ceftriaxone, cefalothin, chloramphenicol, ciprofloxacin, gentamicin, imipenem, kanamycin, nalidixic acid, sulfamethoxazole, streptomycin, tetracycline, co-trimoxazole) using the microdilution method as described by the NCCLS. 11 Amplification and mapping of gene cassettes Integrated gene cassettes were amplified by PCR using primers specific for the 5 and 3 conserved ends of class 1 integrons (Table 1), using elongase (1.0 U/µL; Life Technologies, Inc., Gaithersburg, MD, USA) instead of Taq polymerase. S. enterica Typhimurium DT104 was used as a positive control. S. enterica Typhimurium not containing integron gene sequences was used as a negative control. 10 Inserted gene cassettes were identified through PCR mapping using primers targeting sequences bridging either inti1 aada or inti1 pse1 (Table 1). Identification of Genomic Island 1 (SGI1)-like cluster Seven pairs of PCR primers were designed to amplify linking sequences of the SGI1 antibiotic resistance gene cluster. Three primer pairs were used to amplify the different left and right SGI1 junction combinations (Table 1). Primers targeting left and right SGI1 junctions were designed 1005

3 P. Ebner et al. Figure 1. Map of SGI1 highlighting targets and amplicon sizes in PCR identification of SGI-like gene cluster in S. enterica Meleagridis. Gene sizes not drawn to scale. as described previously and purchased from a commercial source. 12 Figure 1details the position of each primer pair in relation to the full SGI1 sequence. Macrorestriction profiling Pulsed-field gel electrophoresis of XbaI-digested genomic DNA was performed using the method of Gautom. 13 An S. enterica Newport strain (kindly provided by Barbara Gillespie, University of Tennessee, Knoxville, TN, USA) was used as a reference. All banding patterns were visually analysed and compared using the guidelines of Tenover et al. 14 Statistical analysis The frequency of antibiotic resistance in integron-positive isolates was compared with that of integron-negative isolates using the frequency procedure (proc freq) of SAS. 15 Comparisons were made using Fisher s two-sided probability test. Differences were considered significant at P < Results PCR detection of class 1 integrons Class 1 integrons were detected in 32 (30.8%) of the 104 isolates studied. Integron-positive isolates were obtained from nine species of animal, and represented six different serogroups and 15 different serovars (Table 2). More than half (22/32) of the integron-positive isolates, however, belonged to the B serogroup. Two isolates (435 and 2018) contained non-sul1 type class 1 integrons; isolate 435 contained inti1 gene sequences, but was negative for both sul1 and qace 1; isolate 2018 contained both inti1 and qace 1 gene sequences, but was negative for sul1. Gene cassette characterization PCR was used to determine both the number of integrons in each isolate and the types of cassette contained in each integron. Unknown cassette regions were successfully amplified from 30 of 32 integroncontaining isolates. Isolates 435 and 2018 contained integrons not amplifiable by this reaction, presumably because of the absence of reverse (3 ) primer target sequences. All isolates containing amplifiable integrons produced amplicons of 1000 bp; 12 isolates also produced a second amplicon of 1200 bp. PCR mapping confirmed that the 1000 bp amplicons included an aada gene cassette (streptomycin resistance). Because this reaction was amplification from the 5 CS, we were able to determine that the integrons contained in isolates 435 and 2018 also contained this cassette. PCR mapping also confirmed that the 1200 bp amplicons contained a pse1 gene cassette (ampicillin resistance). There were no empty integrons. Eleven of the 12 isolates containing both aada and pse1 integrons belonged to serogroup B (Table 2). Four serogoup B isolates were serotyped as Typhimurium. Identification of SGI1-like gene cluster in S. enterica Meleagridis Isolate 5567 was originally reported as belonging to serogroup E. Serotyping by APHIS revealed that the isolate belonged to the Meleagridis serovar. As the isolate exhibited an ACSSuT resistance phenotype and contained both aada and pse1 integrons, it was of interest to determine whether the strain contained any other SGI1- like qualities. A PCR was designed to amplify seven interconnecting segments of the SGI1 antibiotic resistance gene cluster, and a second PCR was designed to amplify different combinations of left and right SGI1 junctions. S. enterica Meleagridis isolate 5567 produced positive amplicons for each of the seven reactions targeting the SGI1 cluster; and also contained sequences indicative of the left and right SGI1 junctions, as they are found in S. enterica Typhimurium DT It was concluded that S. enterica Meleagridis 5567 contained an SGI1-like island. Neither the integron genotype nor the corresponding antibiotic resistance phenotype was transferable to a susceptible strain by plasmid conjugation. Pulsed-field gel electrophoresis Seventeen isolates belonging to the B serogroup or the B serogroup, Typhimurium serovar were chosen for macrorestriction profiling to determine their genetic relatedness and the number of S. enterica Typhimurium DT104 isolates in the set. Isolate 5567 was included to determine its genetic relatedness to S. enterica Typhimurium DT104. Ten isolates produced banding patterns indistinguishable from the S. enterica Typhimurium DT104 control strain and from the standard S. enterica Typhimurium DT104 pattern commonly seen in our laboratory. 10 These 10 isolates contained both aada and pse1 integrons, showed the penta-drug resistance phenotype (ACSSuT) com- 1006

4 Antibiotic resistance integrons in Salmonella Table 2. Sources, serovars, antibiograms and integron genotypes of integron-positive S. enterica isolates Strain Source Serogroup/serovar Antibiogram No. integrons Integron sizes(s) (kb) Inserted cassette(s) 435 avian necropsy D/Bareilly CeSSuT na a na a aada 905 avian necropsy C/Istanbul AsuT aada 1854 cheetah necropsy B/Agona ACCeKSSuT aada 1878 rodent necropsy B/Loma-Linda SuT aada 1927 bovine faeces D/Berta SuT aada 2018 bovine necropsy B/4, 12 monophasic ASSuTK na a na a aada 2224 reptile faeces D/Arizona SSt aada 2791 feline necropsy B/Typhimurium DT104 ACSSuTK 2 1.0, 1,2 aada, pse canine aspirate B/Typhimurium DT104 ACSSuT 2 1.0, 1,2 aada, pse opossum necropsy C/Bern ASSuT aada 2959 mink necropsy B/Typhimurium DT104 ACSSuT 2 1.0, 1,2 aada, pse canine faeces C/Johannesburg ACeSuT aada 3182 bovine milk B/Typhimurium ACCoKSSuT 2 1.0, 1,2 aada, pse bovine fly B/Typhimurium DT104 ACeSuT 2 1.0, 1,2 aada, pse bovine fly B/Typhimurium DT104 ASuTK 2 1.0, 1,2 aada, pse bovine fly D/nt ACCeSuTK aada 4494 bovine faeces B/Typhimurium DT104 ACSSuT 2 1.0, 1,2 aada, pse bovine faeces B/Typhimurium DT104 ACSSuT 2 1.0, 1,2 aada, pse bovine intestine D/9, 12 non-motile ACSSUTK aada 5528 bovine intestine n/a/untypeable ASSuTK aada 5530 bovine faeces B/Typhimurium DT104 CSSuT 2 1.0, 1,2 aada, pse bovine feed B/Havana SSt aada 5567 bovine faeces E/Meleagridis ACSSuT 2 1.0, 1,2 aada, pse bovine fly trap B/nt ASuTK aada 5564 bovine faeces B/nt ASSuTK aada 5626 bovine faeces B/Typhimurium DT104 ACSSuTK 2 1.0, 1,2 aada, pse bovine faeces B/nt SSu aada 6681 bovine faeces B/nt SuT aada 6766 bovine faeces B/Typhimurium DT104 ACSSuT 2 1.0, 1,2 aada, pse bovine feed B/nt ASuTK aada 6889 bovine faeces B/Typhimurium none aada 6901 bovine necropsy B/Dublin AAuAxCFSSuTTi aada A, ampicillin; Au, co-amoxiclav; Ax, ceftriaxone; C, chloramphenicol; Ce, cefalothin; Co, co-trimoxazole; F, cefoxitin; K, kanamycin; S, streptomycin; Su, sulfamethoxazole; T, tetracycline; Ti, ceftiofur; n/a, not applicable; na, not amplifiable; nt, not tested. a Entire unknown cassette regions were not amplifiable presumably because of lack of 3 target; however, cassettes were identified by PCR mapping. monly associated with S. enterica Typhimurium DT104, and were concluded to belong to this strain (Table 2). Typhimurium isolate 3182, containing both aada and pse1 integrons, produced a distinct banding pattern. PCR showed that like isolate 5567, isolate 3182 also contained an SGI1-like gene cluster. The Typhimurium serotype of isolate 3182, however, was confirmed by APHIS. S. enterica Meleagridis 5567 produced a banding pattern that was distinct from those of isolate 3182 and S. enterica Typhimurium DT104. Antibiotic susceptibility testing and statistical comparison All test isolates were screened for susceptibility to 18 antibiotics/ antimicrobials resulting in several diverse antibiograms (Table 2). The frequency of resistance was compared between integronpositive and integron-negative isolates to determine whether there was any correlation between integron genotype and non-integronmediated antibiotic resistance phenotypes. Isolates containing integrons had a higher incidence of resistance to tetracycline, chloramphenicol and kanamycin; none of these phenotypes could be explained by the presence of integrons (P < 0.05; Table 3). When S. enterica Typhimurium DT104 isolates and isolates 3182 and 5567 were removed from the data set, the frequency of resistance to tetracycline, chloramphenicol and kanamycin remained statistically higher in the integron-containing isolates when compared with integron-negative isolates (P < 0.05; Table 3). No statistical differences were found, however, when comparing other antibiotics. Discussion Integron gene sequences have been reported in many Gram-negative bacteria, including E. coli isolated from both diseased and apparently healthy animals, Gram-negative bacteria isolated from hospitals and natural aquatic environments, 20,21 and, rarely, in Gram-positive bacteria. 22 Previous reports described the prevalence of integron gene sequences among S. enterica var. Enteriditis isolates. 23 Here we report the widespread prevalence of integrons in several S. enterica serovars, many of them rare. The frequency with which they were 1007

5 P. Ebner et al. Table 3. Frequency (%) of resistance to six antimicrobials in integron-positive and integron-negative S. enterica isolates ( ) Integron C T Su S A K a 87.5 a 87.5 a 90.6 a 75.0 a 34.4 a 7.0 b 26.8 b 22.5 b 23.9 b 16.0 b a 80.0 a 80.0 a 85.0 a 65.0 a 40.0 a C, chloramphenicol; T, tetracycline; Su, sulfamethoxazole; S, streptomycin; A, ampicillin; K, kanamycin; ++, all integron-positive isolates (n = 32); +, integron-positive isolates with SGI1-positive isolates removed (n = 20);, integron-negative isolates (n = 72); comparisons made using Fisher s two-sided probability test; numbers with different superscripts are different at P < 0.05; comparisons are within antibiotic; no statistical differences were found with other antibiotics tested. found in our diverse sample set was similar to that found by those working with less diverse bacterial populations. 23 In addition, integronpositive isolates were obtained from various species of animal representing exotics, domestic pets and livestock; and from different sources within each species or type of animal (internal organs versus faeces, etc.). Thus, these data support the notion that integron gene sequences are ubiquitous in several bacterial populations and, in this case, the Salmonella genus. There was very little diversity among the integron gene cassettes in our sample set. Integron-containing isolates exhibited assorted antibiograms, but all contained an integron containing an aada gene cassette. This was the only cassette found in 20 isolates that lacked the SGI1 genomic island. Interestingly, the therapeutic use of streptomycin has declined in both human and veterinary medicine. Resistance to streptomycin is largely integron mediated, however, and the link demonstrated by these data between integron genotypes and non-integron-mediated antibiotic resistance phenotypes illustrates one mechanism by which streptomycin resistance could be maintained in the absence of a selection pressure. 3 Regardless, we agree with others that the high level of streptomycin resistance could serve as an interesting model in illustrating how antibiotic susceptibility does not always return upon the withdrawal of the antibiotic from the bacterial environment. 24 Ampicillin resistance was found in 34.6% of isolates studied. Resistance to ampicillin in Salmonella is usually mediated by TEM β-lactamases, the genes for which are not associated with integrons. 23 Accordingly, there was a low level of integron-mediated ampicillin resistance in our sample set when SGI1-containing isolates were removed from the analysis (this island includes a pse β-lactamase gene). Resistance to other antibiotics commonly associated with integron gene cassettes (e.g. chloramphenicol, trimethoprim) was much lower in comparison with streptomycin resistance. In no case could these phenotypes be attributed to an integron gene cassette. In this study, we identified the S. enterica Typhimurium DT104- associated antibiotic resistance gene cluster SGI1 in a low-prevalence S. enterica serovar (Meleagridis). Similar clusters have been identified in other S. enterica Typhimurium phage types, in a strain of S. enterica Agona and in a strain of S. enterica Paratyphi B 12,25 but, to our knowledge, this is the first report in the Meleagridis serovar. SGI1-positive S. enterica Meleagridis isolate 5567 was obtained from cattle faeces in the State of Washington in the year Outbreaks caused by this serovar were documented throughout Western North America in 1997, primarily involving contaminated sprouts, 26 and it has been isolated from wild birds and other objects in dairy environments. 27 SGI1 gene sequences were also identified in isolate 3182, which was confirmed to have a Typhimurium serotype, but yielded a macrorestriction banding pattern distinct from DT104 strains. There are similar reports of non-dt104 Typhimurium isolates containing SGI1. 12 Identification of an SGI1 cluster in a new serovar (Meleagridis) indicates that this island could be, at least in part, mobile. Boyd et al. 12 have characterized the 43 kb island and the antibiotic resistance gene cluster and determined that, like pathogenicity islands, SGI1 is surrounded by inverted repeats, which could indicate some type of recombination event. Moreover, the island contains many cryptic and functional genes associated with mobile elements. The S. enterica Typhimurium DT104 penta-resistant phenotype can be transduced through P22-like phages ES18 and PDT17, which can be released by the S. enterica Typhimurium DT104 strain, thereby illustrating one possible route by which the cluster could be disseminated. 28 In conclusion, while several diverse antibiograms were found in integron-containing Salmonella isolates, very few resistance phenotypes were integron mediated; only resistance to streptomycin and/or ampicillin was directly attributed to gene cassettes. Nevertheless, the high prevalence of integron-containing strains in this study indicates that these genetic structures are common among varied S. enterica serovars. In addition, genomic island SGI1, once thought to be indigenous to one clonal Salmonella strain, has now been identified in at least five different strains of S. enterica representing four serovars. Acknowledgements This research was funded by the University of Tennessee Food Safety Center of Excellence and Hatch funds allocated to the Tennessee Agricultural Experiment Station. References 1. Aarestrup, F. (1999). Association between the consumption of antimicrobial agents in animal husbandry and the occurrence of resistant bacteria among food animals. International Journal of Antimicrobial Agents 12, Levy, S. (1987). Antibiotic use for growth promotion in animals: Ecological and public health consequences. Journal of Food Protection 50, Ochman, H., Lawrence, J. & Groisman, E. (2000). Lateral gene transfer and the nature of bacterial innovation. Nature 405, Rowe-Magnus, D. & Mazel, D. (2002). The role of integrons in antibiotic resistance gene capture. International Journal of Medical Microbiology 292, Hall, R., Collis, C., Kim, M. et al. (1998). Mobile gene cassettes and integrons in evolution. Annals of the New York Academy of Sciences 870, Hall, R. & Stokes, H. (1993). Integrons: novel DNA elements which capture genes by site-specific recombination. Genetica 90, Lipsitch, M., Singer, R. & Levin, B. (2002). Antibiotics in agriculture: when is it time to close the barn door? Proceedings of the National Academy of Sciences, USA 99, Schwartz, K. (1991). Salmonellosis in swine. Compendium on Continuing Education for the Practicing Veterinarian 13,

6 Antibiotic resistance integrons in Salmonella 9. Khan, A., Nawaz, M., Khan, S. et al. (2000). Detection of multidrug-resistant Salmonella Typhimurium DT104 by multiplex polymerase chain reaction. FEMS Microbiology Letters 182, Ebner, P. & Mathew, A. (2001). Three molecular methods to identify Salmonella enterica serotype Typhimurium DT104: PCR fingerprinting, multi-plex PCR and rapid PFGE. FEMS Microbiology Letters 205, National Committee for Clinical Laboratory Standards. (1997). Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically Fourth edition: Approved Standard M7-A4 (M100- S7). NCCLS, Wayne, PA, USA. 12. Boyd, D., Peters, G., Cloeckaert, A. et al. (2001). Complete nucleotide sequence of a 43-kilobase genomic island associated with the multidrug resistance region of Salmonella enterica serovar Typhimurium DT104 and its identification in phage type DT120 and serovar Agona. Journal of Bacteriology 183, Gautom, R. (1997). Rapid pulsed-field gel electrophoresis protocol for typing Escherichia coli O157:H7 and other Gram-negative organisms in 1 day. Journal of Clinical Microbiology 35, Tenover, F., Arbeit, R., Goering, R. et al. (1995). Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. Journal of Clinical Microbiology 33, SAS. (2000). User s Guide. Statistics, 5th edn. SAS Institution, Gary, NC. 16. Bass, L., Liebert, C. A., Lee, M. D. et al. (1999). Incidence and characterization of integrons, genetic elements mediating multiple-drug resistance, in avian Escherichia coli. Antimicrobial Agents and Chemotherapy 43, Mazel, D., Dychinco, B., Webb, V. et al. (2000). Antibiotic resistance in the ECOR collection: integrons and identification of novel aad gene. Antimicrobial Agents and Chemotherapy 44, Sunde, M. & Sorum, H. (1999). Characterization of integrons in Escherichia coli of the normal intestinal flora of swine. Microbial Drug Resistance 5, Zhao, S., White, D. G., Ge, B. et al. (2001) Identification and characterization of integron-mediated antibiotic resistance among Shiga toxin producing Escherichia coli isolates. Applied and Environmental Microbiology 67, Martinez-Freijo, P., Fluit, A. C., Schmitz, F. et al. (2002) Class 1 integrons in Gram-negative isolates from different European hospitals and association with decreased susceptibility to multiple antibiotic compounds. Journal of Antimicrobial Chemotherapy 42, Rosser, S. & Young, H. (1999). Identification and characterization of class 1 integrons in bacteria from an aquatic environment. Journal of Antimicrobial Chemotherapy 44, Clark, N., Olsvik, O., Swenson, J. et al. (1999). Detection of a streptomycin adenyltransferase gene (aada) in Enterococcus faecalis. Antimicrobial Agents and Chemotherapy 43, Brown, A., Rankin, S. & Platt, S. (2000). Detection and characterization of integrons in Salmonella enterica serotype Enteriditis. FEMS Microbiology Letters 191, Chiew, Y., Yeo, S., Hall, L. et al. (1998). Can susceptibility to an antimicrobial be restored by halting its use: the case of streptomycin versus Enterobacteriaceae? Journal of Antimicrobial Chemotherapy 41, Meunier, D., Boyd, D., Mulvey, M. et al. (2002). Salmonella enterica serotype Typhimurium DT104 antibiotic resistance genomic island I in serotype Paratyphi B. Emerging Infectious Diseases 8, Warner, B. (1997). Proceedings of the Third Annual Federal/State Conference on Food Safety. (1997). [Online.] OPPDE/animalprod/ Proceedings/Sprouts.html (12 December 2003, date late accessed). 27. Kirk, J. H., Holmberg, C. A. & Jeffrey, J. S. (2002). Prevalence of Salmonella spp. in selected birds captured on California dairies. Journal of the American Veterinary Medicine Association 220, Schmieger, H. & Schikelmaier, P. (1999). Transduction of multiple-drug resistance of Salmonella enterica serovar Typhimurium DT104. FEMS Microbiology Letters 170,