|Home | About | Journals | Submit | Contact Us | Français|
Twenty-eight Escherichia coli isolates from various animal and environmental sources with defined tetracycline-resistance genotypes for tet(A), tet(B), and tet(C) were tested for their susceptibility to tetracycline by means of both broth microdilution and Etest. All tet(C)-positive isolates had tetracycline minimum inhibitory concentrations clustering around an intermediate susceptibility range of 2 to 16 μg/mL. Detecting tet(C)-positive isolates by means of susceptibility testing may therefore be difficult with use of the current breakpoint for tetracycline of the Clinical and Laboratory Standards Institute guidelines.
Vingt-huit isolats d’Escherichia coli provenant de diverses espèces animales et de l’environnement possédant les génotypes définis de résistance à la tétracycline pour tet(A), tet(B) et tet(C) ont été testés pour leur sensibilité à la tétracycline par micro-dilution en bouillon et Etest. Tous les isolats tet(C) positifs avaient des concentrations minimales inhibitrices regroupées autour des valeurs de sensibilité intermédiaire variant de 2 à 16 μg/mL. La détection des isolats tet(C) positifs à l’aide d’épreuve de sensibilité pourrait ainsi s’avérer difficile en utilisant les valeurs seuils actuelles pour la tétracycline telles qu’indiquées dans les recommandations du «Clinical and Laboratory Standards Institute».
(Traduit par Docteur Serge Messier)
Tetracycline is a broad-spectrum bacteriostatic agent that binds to the 30S ribosomal subunit and inhibits protein synthesis in bacteria (1). Because of its extensive use, resistance to this antimicrobial agent is one of the most frequently observed in bacteria from animals, including Salmonella enterica and indicator bacteria such as Escherichia coli (2). A variety of mechanisms cause resistance to tetracycline, but in Enterobacteriaceae the main cause is tetracycline-specific efflux pumps that reduce the intracellular concentration of the antibiotic (1). From work with DNA–DNA hybridizations (3) and, more recently, amino acid sequences (4), many tetracycline-efflux pumps in bacteria have been described (1,5). However, previous studies have shown that 3 types of efflux pumps, encoded by the genes tet(A), tet(B), and tet(C), are responsible for most of the tetracycline resistance observed in E. coli from animals (6–8).
While investigating E. coli isolates in collaboration with the Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS), we recently noticed that despite being classified as susceptible to tetracycline (2,9) a number of isolates were positive for the tet(C) gene. We therefore selected a larger sample from our collection of E. coli isolates, which came from a variety of sources, to investigate further the relationship between tet gene variants and tetracycline minimum inhibitory concentrations (MICs). We wanted in particular to assess whether tet(C) is consistently associated with a low tetracycline MIC and thus to determine the association’s general relevance.
Twenty-eight E. coli isolates were investigated. They were obtained between 2003 and 2008 from various sources in Canada, including cattle (n = 8), swine (n = 9), raccoons (n = 4), a house mouse (n = 1), a field mouse (n = 1), and surface waters (n = 5) (Table I). These isolates, previously tested by polymerase chain reaction (PCR) for the presence of tetracycline-resistance genes (10), were selected on the basis of their genotypes to represent the major resistance genes tet(A) (n = 4), tet(B) (n = 3), and tet(C) (n = 16), as well as fully susceptible isolates without any of these genes (n = 5). For a representative sample, isolates with each tet gene were selected randomly within each source, without regard to their tetracycline MIC.
Susceptibility to 15 antimicrobial agents was tested with the use of broth microdilution, according to CIPARS protocols (2). Tetracycline MICs were confirmed for each isolate through broth microdilution with extended dilution series ranging from 0.125 to 512 μg/mL, according to the standards of the Clinical and Laboratory Standards Institute (CLSI) (9). In addition, tetracycline MICs were determined by means of Etest strips, according to the instructions of the manufacturer (AB Biodisk, Solna, Sweden), in parallel with the broth microdilution.
The original tetracycline-resistance genotypes of the 28 isolates were confirmed with a 2nd tet PCR (11), and the identity of 4 representative tet(C) amplicons was confirmed by DNA sequencing. For this purpose, additional primers were designed on the basis of the tet(C) sequence of pAPEC-O1-R (GenBank accession no. DQ517526; www.ncbi.nlm.nih.gov) to obtain a set of overlapping PCR products and to sequence the entire tet(C) gene of all 4 isolates. The tetR gene and the region between tet(C) and tetR containing the tet(C) promoter were also sequenced in 3 of the 4 isolates (isolates 1, 4, and 23) with a combination of PCR and primer walking with plasmid preparations. The resulting sequences were compared with other sequences available in GenBank by means of the Basic Logical Alignment Search Tool (BLASTn) (12).
Plasmid preparations from the 16 tet(C)-positive isolates were obtained with use of a QIAGEN Plasmid Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions. To confirm the plasmid location of tet(C), a tet(C)-specific probe was synthesized by means of the PCR DIG Probe Synthesis Kit (Roche Diagnostics, Mannheim, Germany), a tet(C) PCR product being used as the template (10). The probe was subsequently used on Southern blots of the plasmid preparations according to standard protocols (13). Seven of the plasmid preparations (from isolates 1, 3, 4, 5, 16, 23, and 24), from isolates with broth MICs between 2 and 16 μg/mL, were electroporated into E. coli DH10B (Invitrogen, Carlsbad, California, USA) according to standard protocols (Bio-Rad Laboratories, Hercules, California, USA). Transformants were selected on Mueller–Hinton agar with 2 μg/mL of tetracycline (Becton Dickinson Microbiology Systems, Sparks, Maryland, USA). Successful transformation of the tet(C) gene was confirmed by PCR (11), and the plasmids from these transformants were sized by gel electrophoresis, with use of the BAC-Tracker Supercoiled DNA Ladder (EPICENTRE Biotechnologies, Madison, Wisconsin, USA) as a molecular weight marker.
The tetracycline MICs obtained by broth microdilution for the 16 tet(C)-positive isolates clearly clustered between 2 and 16 (mode 8) μg/mL and were distinct from the MICs for the tet(A)-positive, tet(B)-positive, and tet-negative isolates, irrespective of origin (Figure 1). The tetracycline MICs obtained by Etest for the tet(C)-positive isolates ranged from 3 to 12 (average 5.8) μg/mL. The MICs obtained by the 2 methods were highly correlated across all genotypes (Figure 2). The MICs for the tet(C) reference sequence of pBR322 (4) (GenBank accession no. J01749) in E. coli K12 TB1 (New England Biolabs, Ipswich, Massachusetts, USA) were slightly higher than those for the field isolates (64 μg/mL by broth micro-dilution and 14 μg/mL by Etest). However, this plasmid lacks the tetR repressor gene.
Three of the 4 sequenced tet(C) genes (GenBank accession nos. EU751610 to EU751612) were identical to pBR322 except for a single substitution at position 1049, resulting in an amino acid change at position 384 of the Tet C protein. However, these 3 sequences were identical to numerous other tet(C) sequences in GenBank, including both cloning vectors and sequences from wild-type bacteria (data not shown). The 4th tet(C) sequence (EU751613) from our study showed a frameshift deletion of 15 base pairs (bp) near its 3′-end, resulting in truncation of 5 amino acids. This particular tet(C) allele was associated with the highest MIC among the tet(C)-positive wild-type isolates. Therefore, this truncation seems to have no deleterious effect on the tetracycline-resistance function of the Tet C protein. Sequencing of the tetR gene and the promoter region between tet(C) and tetR in 3 of our isolates (1, 4, and 23) showed sequences (EU751614, EU751615) identical to previous GenBank entries (DQ517526, AJ639924, and others).
Southern blots showed that tet(C) was located on the plasmids prepared from all 16 isolates tested (data not shown). The size of the plasmids carrying tet(C) in the 7 tet(C) transformants ranged from 10 to 100 kbp. The tetracycline MICs of these transformants obtained by broth microdilution ranged from 8 to 16 μg/mL (Table I). Susceptibility testing showed cotransfer of sulfonamide and streptomycin resistance in 2 of the 7 transformants; PCR (14) demonstrated sul1 and aadA in these 2 transformants, whose parent strains were from pigs of unrelated origin. No cotransfer of resistance to antimicrobial agents other than tetracycline was observed with the other 5 transformants.
These results contrast strongly with those obtained in another study, in which an average MIC of 143.9 μg/mL was obtained by Etest in tet(C)-positive E. coli isolates from Scottish swine (15). However, others had already observed lower MICs (8 to 64 μg/mL) for tet(C)-positive coliform bacteria from swine (16). The potential reasons for these discrepancies between studies of tet(C)-positive isolates may be numerous, but the diversity of sources for the isolates in this study suggests that the low MICs we observed in association with tet(C) represent a potentially widespread phenomenon in animal and environmental E. coli isolates in Canada. Although different testing procedures were used in the other studies (15,16), our results showing a close correlation between MICs determined by broth microdilution and Etest suggest that the reasons for these discrepancies are not related to methodologic issues in susceptibility testing. Since the MICs for our tet(A)- and tet(B)-positive isolates were in the same range as those found by Blake et al (15), the discrepancies observed seem to affect only tet(C). In contrast to the present study, Blake et al used selective media containing tetracycline (4 μg/mL) for the isolation of porcine fecal E. coli from intensive farms. This may have eliminated most of the tet(C)-positive isolates with low tetracycline MICs and selected for a subset of tet(C)-positive isolates with particularly high MICs, which were not detected in our study on nonselective media.
Although further testing is needed, the single-substitution difference between the reference tet(C) from pBR322 and the tet(C) from our isolates is unlikely to be the reason for the low observed MICs. The regulation of tet(C) expression may have differed between our isolates and those of Blake et al (15). Nevertheless, the identity of the regulatory and promoter regions in the isolates in the present study and sequences published by others suggests that the findings of this study are of general relevance. The tet(C) determinant associated with low-level resistance to tetracycline in the present study was found on a variety of plasmids, which suggests horizontal spread across E. coli populations. Similarly, the isolates in the study of Lee et al (16) all carried tet(C) on plasmids, but it is unclear whether the tet(C) genes from the isolates in the study of Blake et al (15) were chromosomal or located on plasmids. Unfortunately, no sequence data are available from these 2 earlier studies. Thus, further hypotheses based on the potential effects of tet(C) location, gene variants, and regulation mechanisms to explain the differences between tetracycline MICs in these studies are purely speculative.
In conclusion, our results show that tet(C)-positive isolates recovered in Canada may often be considered intermediate in resistance or even fully susceptible according to the CLSI guidelines (9). A dilution series extended beyond that used by CIPARS and the National Antimicrobial Resistance Monitoring System in the United States (17), down to 0.25 μg/mL, may be appropriate in a full epidemiologic examination of tetracycline resistance. In addition, epidemiologic breakpoints lower than those used for clinical purposes should be considered for tetracycline-resistance surveillance. Many tet(C)-positive isolates may otherwise be misclassified as susceptible and may not be tested further for the presence of tetracycline-resistance genes. As suggested by Blake et al (15), such isolates with low tetracycline MICs may behave differently from those carrying tet(A) or tet(B) when exposed to tetracycline. The linkage of tet(C) with integrons (sul1 and aadA) in some of these isolates also suggests that tetracycline use at levels currently considered subinhibitory may still select for multiresistance through unnoticed tetracycline-resistance determinants.
We thank Emily Weir and Laura Martin at the Laboratory for Foodborne Zoonoses, Public Health Agency of Canada, Guelph, Ontario, for their technical assistance with this work. This research was funded by the Public Health Agency of Canada.