Multidrug-Resistant Escherichia coli and Tetracycline-Resistant Enterococcus faecalis in Wild Raptors of Alabama and Georgia, USA Open Access

Antimicrobial resistance remains one of the largest threats to public health, despite decades of efforts to slow down the selection and transfer of resistance through judicious use of antimicrobials (Davies 2007). Understanding the mode of dissemination of antimicrobial-resistant microbes and their resistance genes is crucial in formulating strategies to prevent and control the rising threat of antibiotic resistance. Free-living raptors fly great distances and inhabit a variety of environments and thus can potentially serve as a reservoir and as a means of dissemination of antimicrobial resistance. However, there is only one report of antimicrobial resistance in raptors in the US, which showed that enterococci were resistant to multiple antibiotics (Marrow et al. 2009). To examine the potential role of wild raptors as a reservoir and a vector of antimicrobial resistance, we investigated the prevalence of antimicrobialresistant Escherichia coli and Enterococcus faecalis, enteric bacteria capable of rapidly acquiring resistance, in wild raptors.

A total of 118 free-living raptors of 17 species were brought to the Southeastern Raptor Center (Auburn, Alabama, US) between January 2013 and April 2014 from 18 Alabama and 15 Georgia counties. The birds were suffering from emaciation or a variety of injuries, including gunshots, severe lacerations, or fractures. At admission, a cloacal swab was obtained from each bird and transferred to the laboratory in Amies transport medium (Thermo Scientific, Grand Island, New York, USA). Swabs were processed within 2 h of collection onto 5% bovine blood agar and MacConkey agar (BD Biosciences, San Jose, California, USA). One colony with morphologies typical for E. coli and E. faecalis was selected from the cultures from each bird and identified by conventional biochemical methods.

For antimicrobial susceptibility testing, we selected antibiotics (ampicillin, chloramphenicol, ciprofloxacin, and tetracycline) representing four major classes of drugs widely used in both medicine and agriculture. All four antibiotics were appropriate for testing against both E. faecalis and E. coli, and neither organism exhibits intrinsic resistance to any of these antibiotics. Minimal inhibitory concentrations (MICs) were determined by using the ETEST (AB Biodisk, bioMérieux, Inc., St. Louis, Missouri, USA) according to the manufacturer's recommendations. Multidrug-resistant bacteria was defined as resistance to three or more classes of antimicrobials (Magiorakos et al. 2012).

Bacterial genomic DNA was extracted by boiling isolates at 100 C, followed by flash cooling. The extracted DNA was stored at –80 C for molecular detection. The representative nucleotide sequences for antimicrobial resistance genes were obtained from GenBank and were aligned by using Clustal multiple alignment (Vector NTI, Invitrogen, Grand Island, New York, USA) to identify regions to design primers and probes (Table 1). The PCRs for β-lactamases, blaTEM-1, blaCTX-M-1, and tetracycline resistance genes, tet(M), tet(L), were run in a SYBR Green PCR system, while PCRs for gyrA and two chloramphenicolresistant genes (cmlA, cat) were in a fluorescence resonance energy transfer–PCR platform, as previously described (Yang et al. 2014; Zhang et al. 2018). All PCRs in this study were performed in a Roche LightCycler 1.5 PCR system (Roche Diagnostics, Indianapolis, Indiana, USA). Sensitivity of PCRs was determined by amplification of standards with 10⁴, 10³, 10², and 10¹ copies per reaction. The specificity of established PCRs was verified by gel electrophoresis and DNA sequencing of PCR products for all positive samples.

The statistical analysis was done by using Statistica version 8.0 (StatSoft, Inc., Tulsa, Oklahoma, USA). A chi-squared test was performed to compare the difference in overall prevalence of antibiotic resistance between E. coli and E. faecalis isolates. Difference was considered significant at P≤0.05.

From the 118 wild raptors, we isolated a total of 112 E. coli and 76 E. faecalis. The rate of isolation for E. coil was 95 and 64% for E. faecalis, with 74 (63%) of the birds being culture positive for both (Table 2). Among the E. coli isolates, the most prevalent resistance was to tetracycline, with 16% of the strains having a MIC>128 µg/mL. Ampicillin resistance was found at a rate of 12% (Table 3), with all of the resistant strains having an MIC>256 µg/mL. Of the four drugs tested, the E. faecalis isolates only exhibited resistance to tetracycline at a rate of 8% (Table 3), with MIC levels ranging from 24 µg/mL to >64 µg/mL. The percentage of isolates resistant to at least one of the antibiotics tested was higher (P=0.052) in E. coli (20/112, 18%) than in E. faecalis (6/76, 8%; Table 2).

Five of the six tested antimicrobial resistance genes, blaTEM-1, blaCTX-M-1, tet(M), cmlA, cat, as well as mutations in gyrA were identified in the resistant E. coli isolates (Table 2). In addition, we identified tet(M) and tet(L) in the tetracycline-resistant isolates of E. faecalis (Table 2). The nucleotide sequences for antimicrobial resistance genes obtained in this study were deposited to the GenBank with the following gene accession numbers: MH033833 and MH033834, bla-TEM-1; MH033835, tet(L); MH033836, tet(M); MH085993, MH085994, and MH085995, cat; and MH085992, cmlA (Table 2).

Ampicillin resistance in E. coli was primarily mediated by blaTEM-1 and blaCTX-M-1. Five of the resistant E. coli possessed both genes, while four only carried the blaTEM-1 gene, and another three only contained blaCTX-M-1. Two ampicillin-resistant E. coli strains were negative for both blaTEM-1 and blaCTX-M-1, suggesting a different β-lactamase is involved. Of the 18 tetracyclineresistant E. coli, only five possessed the tet(M) gene. Chloramphenicol resistance in E. coli was mediated by cat (n=8), with three also possessing cmlA. Of interest is the observation that all of the chloramphenicol-resistant E. coli were also resistant to tetracycline. In the six ciprofloxacin-resistant E. coli isolates, we identified five isolates with double mutations of gyrA (serine-83_leucine and aspartic acid-87_asparagine). Three of six tetracyclineresistant E. faecalis isolates harbored both tet(M) and tet(L) genes, but only tet(M) was found in the remaining three isolates (Table 2).

Wild birds of prey can cover great distances, have wide dietary choices, and inhabit many diverse environments. Because of their predatory nature, the discovery of any drug resistance in bacteria recovered from their gut flora is most likely acquired from their food sources or environmental exposure. As such, wild raptors can signal environmental contamination with antimicrobial-resistant bacteria, but of even more concern, they may also serve as a reservoir for commingling of drugresistant bacteria and resistance genes and, ultimately, as a means of dispersal of such agents.

Multiple studies have been conducted in Europe and elsewhere surveying the level of antimicrobial resistance in various avian species (Bonnedahl and Järhult 2014). Limited information is available in the US to indicate if there is a similar pattern of drug resistance among the normal flora of wild birds. A majority of the E. coli (82%) and E. faecalis (92%) isolates obtained from raptors in this study were susceptible to all four antibiotics tested (Table 2). However, of the 20 (18%) E. coli isolates expressing some level of resistance, nine expressed a multidrug-resistant phenotype of which four were resistant to all four classes of antibiotics. Despite a low overall prevalence of antimicrobial-resistant bacteria, finding nearly half of the resistant E. coli to be multidrug resistant is alarming, especially in birds that should be antibiotic naive.

Among 13 ampicillin-resistant E. coli, the resistant genes, blaTEM-1 and blaCTX-M-1 were relatively common: blaTEM-1 and blaCTX-M-1 were each found in eight (62%) isolates. The gene blaTEM-1 encodes the TEM type β-lactamase and confers resistance to penicillin and first-generation cephalosporin and is widely found among gram-negative bacteria (Palzkill 2018). The blaCTX-M-1 provides high-level resistance to third-generation cephalosporins, and they confer variable levels of resistance to fourth-generation cephalosporins (Cantón et al. 2012). The presence of these two genes only among birds that have mammals in their diet in this study is supportive of the role food source has in spreading antimicrobial resistance in remote areas.

Tetracycline resistance was prominent among the E. coli isolates, with all 18 isolates expressing very high MICs. Of the two common genes tet(M) and tet(L) tested, we identified only tet(M) in just five isolates, indicating presence of other resistance mechanism. Previous publications reported that tet(M) confers only low levels of resistance in E. coli (Bryan et al. 2004), whereas our five tet(M)-positive isolates had exceptionally high MIC values (16 times the resistant break point). Most resistance mechanisms to chloramphenicol are efflux pumps (cmlA), as well as inactivating enzymes, such as chloramphenicol acetyltransferase (cat; Frye and Jackson 2013). Chloramphenicol resistance was detected in eight birds; the cat gene was found in all eight, while the cmlA gene was found in only three.

Ample evidence suggests that wild birds can host antimicrobial-resistant bacteria (Bonnedahl and Järhult 2014); however, few studies are from the US. In a study conducted in the midwestern region of the US (Marrow et al. 2009), E. faecalis was isolated from 21 wild and four captive raptors, including hawks, owls, and kestrels. Multiple isolates were recovered from each bird and demonstrated resistance to several antibiotics, including chloramphenicol (39%), enrofloxacin (41%), and tetracycline (33%), but no resistance to ampicillin was detected. The current study differed from Marrow et al. (2009) in that single isolates of E. coli and E. faecalis were collected from wild raptors, and the sampled population included birds with more diverse feeding habits. By selecting birds with limited human contact, these birds were more likely to represent the natural population, which allowed us to determine the level of environmental contamination due to drug resistant bacteria. Furthermore, by testing only one isolate per bird in a limited number of population over a 15-mo period, we assured that the level of drug resistance in the wild population was not artificially inflated, and we may even have underestimated the level of antimicrobial resistance.

Our findings indicated that wild raptors could be an important indicator of environmental contamination with antimicrobial-resistant bacteria, and even more importantly, that they could be potential vectors in the dissemination of antibiotic resistance. Future studies should expand the area for sampling, including wild raptors from different geographic areas, distinguishing between congested urban areas versus remote locations and birds with different dietary choices. During surveillance, it will be important to expand gene testing to determine what new resistance genes wild birds are already absorbing and the degree with which they are shedding such genes. Our findings reflected the blurring of lines separating human and animal populations and the environment. These results should initiate a discussion into the development of practices that will slow the spread of drug resistance, reducing its impact on human and animal health.

This work was supported by Alabama Agricultural Experimental Station and the US Department of Agriculture, National Institute of Food and Agriculture, Hatch Project (ALA052-1-17026).