CTX-M-15–Producing Escherichia coli in Dolphin, Portugal

To the Editor: The global emergence and pandemic spread of sequence type (ST) 131 CTX-M-15–producing Escherichia coli among humans and its detection in livestock, companion animals, and wildlife is a major cause for concern (1,2). Hence, it is imperative to identify and explore its dissemination traits. If CTX-M-15–producing E. coli continues to spread among different environments, therapeutic options in veterinary and human medicine will be greatly narrowed (1). E. coli is one of the gram-negative bacteria most frequently isolated from bottlenose dolphins (3). However, few studies about antimicrobial drug–resistant bacteria in dolphins have been published (4–6). We explored dissemination linkages between CTX-M-15–producing E. coli isolated from a marine dolphin (Tursiops truncatus) andclinical isolates collected during the same period from humans all over Portugal. 
 
In 2009, E. coli strain LV143, isolated from respiratory exudate collected through the spiracle of a female dolphin from a zoo, was sent to the National Institute for Agricultural and Veterinary Research in Lisbon, Portugal, for bacteriological and mycological analysis and antimicrobial drug susceptibility testing. No clinical history for the animal was available. Mycologic examination detected no fungi or yeasts. 
 
Drug susceptibility testing of the dolphin E. coli strain (LV143), performed by the agar dilution method and interpreted according to European Committee of Antimicrobial Susceptibility Testing (http://www.eucast.org/), revealed a non–wild-type phenotype to cefotaxime (MIC >8 μg/mL); it also showed a synergy toward clavulanic acid, suggesting production of extended-spectrum β-lactamase (ESBL). LV143 was also non–wild-type to ampicillin (MIC >64 µg/mL), nalidixic acid (MIC >512 µg/mL), ciprofloxacin (MIC >8 µg/mL), gentamicin (MIC >32 μg/mL), and tetracycline (MIC >64 μ /mL). This isolate remained wild-type to chloramphenicol (MIC 4 μg/mL), florfenicol (MIC 8 μg/mL), sulfamethoxazole (MIC 32 μg/mL), trimethoprim (MIC ≤0.25 μg/mL), and streptomycin (MIC 4 μg/mL). 
 
To analyze the zoonotic potential of the dolphin isolate, we selected 61 human clinical E. coli isolates, previously recovered from different specimens during 2004–2009 in 7 geographically separated hospitals in Portugal (Figure), from the National Reference Laboratory of Antibiotic Resistances and Healthcare Associated Infections collection. Inclusion criteria for the clinical isolates were 1) non–wild-type susceptibility to cefotaxime, 2) presumptive phenotypic ESBL production, and 3) genetic similarity by pulsed-field gel electrophoresis. Analysis of the genetic relatedness of human and dolphin isolates, determined by pulsed-field gel electrophoresis that used XbaI digested DNA (7), revealed 1 major cluster, which included 22 (35%) clinical isolates from 3 regions in Portugal and the isolate from the dolphin (Figure). 
 
 
 
Figure 
 
Dendogram of pulsed-field gel electrophoresis (PFGE) profiles showing the relationship between a clonal strain of Escherichia coli of animal origin (LV143, in boldface), and 22 E. coli isolates from humans. We used the unweighted pair group method and ... 
 
 
 
The genetic characterization of the 1 dolphin and 22 clinical isolates was performed by PCR and sequencing selective for the most prevalent ESBL-mediated genes (blaTEM, blaSHV, blaOXA-G1, blaCTX-M) and genes encoding plasmid-mediated quinolone resistance (qnrA, qnrB, qnrC, qnrD, qnrS, qepA, aac(6’)Ib-cr), as previously described (7). Specifically, the strain recovered from the dolphin contained blaCTX-M-15, blaTEM-1, and blaOXA-30, associated with a plasmid-mediated quinolone resistance gene, aac(6’)-Ib-cr (Figure). All clinical isolates were also positive for blaCTX-M-15 and blaOXA-30 genes; 18 isolates contained the blaTEM-1 gene and 3 blaSHV-1, 5 blaSHV-12, 8 qnrB, and 16 aac(6’)-Ib-cr genes. The presence of class 1 integron, ISEcp1, IS26, and IS903 elements was also investigated, as has been done previously (8). The LV143 strain was positive for the insertion sequence ISEcp1, associated with blaCTX-M-15 (Figure), and was negative for the class 1 integron (data not shown). In 2 clinical isolates, we identified ISEcp1, and in 1 isolate we identified IS903. PCR-based replicon typing (9) revealed the presence of IncF plasmid group in the 1 animal and 9 human isolates (a selected sample to evaluate PCR-based replicon typing) (Figure). 
 
Multilocus sequence typing (MLST) was performed for 9 of the 23 E. coli isolates. According to E. coli MLST website (http://mlst.ucc.ie/mlst/dbs/Ecoli), clones from the dolphin and from the humans exhibited the same combination of alleles across the 7 sequenced loci, corresponding to the epidemic ST131, associated with CTX-M-15 and widely disseminated among hospitals in Portugal (2,7). Within-ST subclones were analyzed on the basis of sequence variation of the E. coli fimbrial adhesin gene fimH, as previously described (10). The fimH30-Rx lineage was identified in all 23 E. coli isolates (fluoroquinolone-resistant and CTX-M-15–positive isolates), which clustered together on the dendrogram, regardless of MLST result (Figure). It is worth noting that the blaCTX-M-type gene has been detected in ESBL-positive E. coli isolates from healthy mammals (1). 
 
Our study illustrated clonality among clinical isolates and a dolphin strain with common antimicrobial drug–resistance genes, specifically blaCTX-M-15 and aac(6')-Ib-cr, and common plasmids, such as those from group IncF. These bacteria have gone through identical evolutionary genetic events, which ultimately led to the establishment of the same allelic diversity pattern (ST131 fimH30-Rx). The linkage between these 2 reservoirs highlights the zoonotic potential of this isolate from the dolphin.

Escherichia coli among humans and its detection in livestock, companion animals, and wildlife is a major cause for concern (1,2). Hence, it is imperative to identify and explore its dissemination traits. If CTX-M-15-producing E. coli continues to spread among different environments, therapeutic options in veterinary and human medicine will be greatly narrowed (1). E. coli is one of the gram-negative bacteria most frequently isolated from bottlenose dolphins (3). However, few studies about antimicrobial drug-resistant bacteria in dolphins have been published (4)(5)(6). We explored dissemination linkages between CTX-M-15-producing E. coli isolated from a marine dolphin (Tursiops truncatus) and clinical isolates collected during the same period from humans all over Portugal.
In 2009, E. coli strain LV143, isolated from respiratory exudate collected through the spiracle of a female dolphin from a zoo, was sent to the National Institute for Agricultural and Veterinary Research in Lisbon, Portugal, for bacteriological and mycological analysis and antimicrobial drug susceptibility testing. No clinical history for the animal was available. Mycologic examination detected no fungi or yeasts.
To analyze the zoonotic potential of the dolphin isolate, we selected 61 human clinical E. coli isolates, previously recovered from different specimens during 2004-2009 in 7 geographically separated hospitals in Portugal (Figure), from the National Reference Laboratory of Antibiotic Resistances and Healthcare Associated Infections collection. Inclusion criteria for the clinical isolates were 1) non-wild-type susceptibility to cefotaxime, 2) presumptive phenotypic ESBL production, and 3) genetic similarity by pulsed-field gel electrophoresis. Analysis of the genetic relatedness of human and dolphin isolates, determined by pulsed-field gel electrophoresis that used XbaI digested DNA (7), revealed 1 major cluster, which included 22 (35%) clinical isolates from 3 regions in Portugal and the isolate from the dolphin (Figure).
The selective for the most prevalent ESBL-mediated genes (bla TEM , bla SHV , bla OXA-G1 , bla CTX-M ) and genes encoding plasmid-mediated quinolone resistance (qnrA, qnrB, qnrC, qnrD, qnrS, qepA, aac(6')Ib-cr), as previously described (7). Specifically, the strain recovered from the dolphin contained bla CTX-M-15 , bla TEM-1 , and bla OXA-30 , associated with a plasmid-mediated quinolone resistance gene, aac(6')-Ib-cr (Figure). All clinical isolates were also positive for bla CTX-M-15 and bla OXA-30 genes; 18 isolates contained the bla TEM-1 gene and 3 bla SHV-1 , 5 bla SHV-12, 8 qnrB, and 16 aac(6')-Ib-cr genes. The presence of class 1 integron, ISEcp1, IS26, and IS903 elements was also investigated, as has been done previously (8). The LV143 strain was positive for the insertion sequence ISEcp1, associated with bla CTX-M-15 ( Figure), and was negative for the class 1 integron (data not shown). In 2 clinical isolates, we identified ISEcp1, and in 1 isolate we identified IS903. PCR-based replicon typing (9) revealed the presence of IncF plasmid group in the 1 animal and 9 human isolates (a selected sample to evaluate PCR-based replicon typing) (Figure). Multilocus sequence typing (MLST) was performed for 9 of 23 E. coli isolates. According to the E. coli MLST website (http://mlst.ucc.ie/mlst/dbs/Ecoli), clones from the dolphin and from the humans exhibited the same combination of alleles across the 7 sequenced loci, corresponding to the epidemic ST131, associated with CTX-M-15 and widely disseminated among hospitals in Portugal (2,7). Within-ST subclones were analyzed on the basis of sequence variation of the E. coli fimbrial adhesin gene fimH, as previously described (10). The fimH30-Rx lineage was identified in all 23 E. coli isolates (fluoroquinolone-resistant and CTX-M-15positive isolates), which clustered together on the dendrogram, regardless of MLST result (Figure). It is worth noting that the bla CTX-M-type gene has been detected in ESBL-positive E. coli isolates from healthy mammals (1).
Our study illustrated clonality among clinical isolates and a dolphin strain with common antimicrobial drug-resistance genes, specifically bla CTX-M-15 and aac(6')-Ib-cr, and common plasmids, such as those from group IncF. These bacteria have gone through identical evolutionary genetic events, which ultimately led to the establishment of the same allelic diversity pattern (ST131 fimH30-Rx). The linkage between these 2 reservoirs highlights the zoonotic potential of this isolate from the dolphin.