ABSTRACT
ABSTRACT
Wild birds are carriers of Escherichia coli. However, little is known about their role as reservoirs for extra-intestinal pathogenic E. coli (ExPEC). In this work we investigated E. coli strains carrying virulence genes related to human and animal ExPEC isolated from free-living wild birds treated in a veterinary hospital. Multidrug resistance was found in 47.4% of the strains, but none of them were extended-spectrum beta-lactamase producers. Not only the virulence genes, but also the serogroups (e.g. O1 and O2) detected in the isolates of E. coli have already been implicated in human and bird diseases. The sequence types detected were also found in wild, companion and food animals, environmental and human clinical isolates in different countries. Furthermore, from the 19 isolates, 17 (89.5%) showed a degree of pathogenicity on an in vivo infection model. The isolates showed high heterogeneity by pulsed-field gel electrophoresis indicating that E. coli from these birds are clonally diverse. Overall, the results showed that wild birds can be reservoirs and/or vectors of highly pathogenic and multidrug-resistant E. coli that have the potential to cause disease in humans and poultry.
Introduction
Avian pathogenic Escherichia coli (APEC) isolates cause several extra-intestinal infections in birds and are responsible for significant economic losses in the poultry industry worldwide (Barnes et al., 2008). APEC is part of a group of E. coli pathotypes known as extra-intestinal pathogenic E. coli (ExPEC), which cause infections outside the intestinal tract and can cause diseases in humans and animals, such as urinary tract infections, neonatal meningitis, bloodstream infections and pyometria (Kaper et al., 2004; Maluta et al., 2014a). There are no definitive genetic markers to define an E. coli strain as APEC, but Johnson et al. (2008) distinguished most APEC from avian faecal E. coli by the possession of five genes (iutA, hlyF, iss, iroN and ompT).
Avian and human pathogenic E. coli strains may share common characteristics, such as serogroups, phylogenetic groups, sequence types (ST) and profiles of virulence genes (Moulin-Schouleur et al., 2007; Maluta et al., 2014c), suggesting that a subset of APEC isolates have the potential to be zoonotic pathogens. The increased interaction among wild animals, domestic animals and humans is very important to veterinary and public health because it has been reported that the intestinal microbiota of wild birds may contain multidrug-resistant (MDR) E. coli (Guenther et al., 2010; Blyton et al., 2015), including strains carrying genes related to APEC (Ewers et al., 2009). In this context, wild birds may serve as reservoirs and mechanical vectors of antimicrobial-resistant ExPEC and could transmit these infectious agents to humans and commercial poultry (Hubálek, 2004).
Although wild birds have only rare contact with antimicrobial agents, they can be infected or colonized by resistant bacteria through contaminated water or food (Cole et al., 2005). A number of workers have examined E. coli associated with human intestinal diseases such as Shiga toxin-producing E. coli and enteropathogenic E. coli in wild birds (Foster et al., 2006; Ahmed et al., 2007; Farooq et al., 2009). However, only few studies have investigated the presence of E. coli harbouring genes related to human and bird ExPEC in these birds (Ewers et al., 2009).
The role of wild birds as reservoirs of APEC may be underestimated. A work presenting a comprehensive characterization of potential APEC isolated from wild birds would be of interest. Therefore, the aim of this work was to characterize potential APEC from wild birds in order to determine if those birds are reservoirs of pathogenic E. coli associated with extra-intestinal diseases.
Materials and methods
Ethics statement
The study is in accordance with the Ethical Principles in Animal Experimentation, adopted by the Brazilian College of Experimentation (COBEA) and was approved by Animal Experimentation Ethics Committee (CEUA) from São Paulo State University, protocol No. 22.222/10.
Sampling and initial procedures
Samples were collected from 123 free-living wild birds that were treated at the Wildlife Veterinary Hospital in UNESP-FCAV, Brazil, from September 2010 to April 2012. These wild birds were taken to the veterinary hospital for many reasons, including found as orphans, brain tumour, bone fractures or heavy metal poisoning. None had any known infectious disease. Faecal and oropharyngeal samples were collected from cloaca (n = 123) and oropharynx (n = 123), respectively, using sterile swabs, as soon as the birds arrived in the hospital, before they received antibiotics. Liver (n = 14) and intestinal (n = 14) samples were taken from wild birds that eventually died. All samples were immediately cultured into tubes containing 5 ml of brain and heart infusion (BHI) broth and transported to the laboratory on ice in thermal boxes. The tubes were incubated aerobically overnight at 37°C, and an aliquot of each culture was stored at −80°C.
Obtaining isolates and determining pathotype using PCR
After the sample incubation in BHI broth (37°C for 24 h under static conditions) DNA extraction was performed by the thermal lysis procedure.1 After preparation of the DNA templates, polymerase chain reaction (PCR) was performed as described previously (Borges et al., 2012) using primers listed in the supplemental table. Each DNA extract was screened for six virulence-associated genes (VAGs) that had been linked to APEC virulence, iroN, iss, iutA, ompT, hlyF (Johnson et al., 2008) and cvaC (Rodriguez-Siek et al., 2005). Two sets of primer pools were used for multiplex PCR: one set contained primers specific for iroN, ompT and iss, and the other contained primers specific for iutA, hlyF and cvaC. The primers, annealing temperatures and controls are described in the supplemental table. All PCR reactions were carried out using negative (EcL3463) and positive control strains of E. coli (supplemental table). Samples that were positive for at least one of the genes by PCR were cultured onto MacConkey agar plates and incubated at 37°C for 24 h. From each plate, 10 randomly selected colonies typical of E. coli were tested by PCR as described above. Colonies positive for at least one gene were stored at −80°C as pure cultures for subsequent experiments. This methodology is from the Reference Laboratory for Escherichia coli (EcL) Université de Montréal (Maluta et al., 2014b).
Detection of extended-spectrum beta-lactamases (ESBL) and additional virulence genes
All isolates were tested for ESBL and 14 additional virulence genes by PCR amplification. The primers used are listed in the supplemental table. One triplex and four duplex PCR procedures were performed for bulk genotyping studies of the following genes: (1) irp2 and fyuA; (2) tsh and sitA; (3) traT and iucC; (4) fimH and iucD and (5) papGI, papGII and papGIII. The cnf1, papC, sfa and blaCTX−M genes were detected using monoplex PCR assays.
Determination of phylogenetic group by PCR
Strains were assigned to one of the four main phylogenetic groups of E. coli (A, B1, B2 or D) by using the triplex PCR targeting chuA, yjaA and the DNA fragment TSPE4.C2 (Clermont et al., 2000).
Pathogenicity test in one-day-old chicks
One-day-old chicks were used to determine the pathogenicity of each isolate as described previously (Guastalli et al., 2013). The chicks were observed daily for 10 days, and the strains were classified according to the following mortality index: high pathogenicity (mortality > 80%), intermediate pathogenicity (mortality > 50% but <80%), low pathogenicity (mortality ≤ 50%), and non-pathogenic (no mortality). E. coli isolated from broiler chickens with chronic respiratory disease (CRD) – EC55 (serogroup O1, iss+ papC+ iucD+ cva/cvi+) was used as a positive control for virulence. The birds used as a negative control were inoculated with BHI broth.
Serotyping
Strains were serotyped at the E. coli Reference Center, Pennsylvania State University, USA. The O antigen typing was performed as described previously (Ørskov & Ørskov, 1984). The H antigen typing was performed by the PCR – Restriction Fragment Length Polymorphism of the fliC gene which encodes flagella (Machado et al., 2000).
Antimicrobial susceptibility test
Antimicrobial disc susceptibility tests were performed using the disc diffusion method (Clinical and Laboratory Standards Institute [CLSI] 2014). The antimicrobials tested were amikacin (30 µg), amoxicillin-clavulanic acid (30 µg), ampicillin (10 µg), cefoxitin (30 µg), ceftiofur (30 µg), ceftriaxone (30 µg), chloramphenicol (30 µg), ciprofloxacin (5 µg), gentamicin (10 µg), kanamycin (30 µg), nalidixic acid (30 µg), nitrofurantoin (300 µg), norfloxacin (10 µg), sulfamethoxazole + trimethoprim (25 µg) and tetracycline (30 µg).
Pulsed-field gel electrophoresis (PFGE)
The isolates were subtyped by the standardized rapid PFGE protocol used by laboratories participating in the PulseNet scheme as described previously (Ribot et al., 2006). The E. coli chromosomal DNA was digested with XbaI. Electrophoresis conditions consisted of an initial pulse time of 2.2 s and a final pulse time of 54.2 s at a gradient of 6 V/cm and an included angle of 120° in a CHEF-DRIII PFGE system (Bio-Rad Laboratories, Hercules, CA, USA). The gels were electrophoresed for 22 h. The similarities of fragments were compared using a Dice coefficient at 1% tolerance and 0.5% optimization, and dendrograms were constructed using the UPGMA clustering method using the BioNumerics version 7.1 (Applied Maths, Austin, TX, USA).
Multilocus sequence typing (MLST)
MLST was performed following the Achtman scheme (http://mlst.ucc.ie/mlst/dbs/Ecoli). This scheme is based on the sequencing of the PCR amplification products of adk, fumC, gyrB, icd, mdh, purA and recA (supplemental table). DNA template preparation and PCR were done as described in the previous section. Sequencing was performed at the University of California Berkeley DNA Sequencing Facility.
Results
Isolates harbouring ESBL and virulence genes
A total of 274 samples from wild birds were tested by PCR for the detection of the APEC VAGs (iroN, iss, iutA, ompT and hlyF) and an additional gene (cvaC); 72 (30.6%) samples were positive for at least one of these six genes (37 from cloaca, 23 from oropharynx, seven from liver and five from intestine). From these 72 positive samples, 19 (6.9%) isolates were obtained (12 from cloaca, four from oropharynx, two from liver and one from intestine). These isolates were subjected to a new round of PCR to detect ESBL and 14 additional virulence genes (Figure 1). Within the isolates, iss was the most prevalent gene (94.7%), followed by sitA (84.2%), traT (63.1%), ompT (47.4%), fyuA (47.4%), irp2 (47.4%), iroN (42.1%), hlyF (42.1%), fimH (21.0%), iutA (10.5%), iucC (10.5%), iucD (5.3%), tsh (5.3%) and cvaC (5.3%). None of the isolates were positive for papGI, papGII, papGIII, papC, cnf1, sfa or blaCTX−M genes.
Published online:
18 October 2016![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/cavp20/2017/cavp20.v046.i01/03079457.2016.1209298/20161223/images/medium/cavp_a_1209298_f0001_b.gif"})
Figure 1. Wild bird paper dendrogram.
Determination of the phylogenetic group by PCR
Phylogenetic group testing showed that 8/19 (42.1%) isolates belonged to the B1 group, 8/19 (42.1%) isolates to the B2 group, 2/19 (10.5%) to the D group and 1/19 to the A group (5.2%) groups (Figure 1).
Pathogenicity test in day-old chicks
Pathogenicity testing in one-day-old chicks revealed that six (31.6%), three (15.8%), eight (42.1%) and two (10.5%) isolates were classified as highly pathogenic, intermediate pathogenic, low pathogenicity or non-pathogenic, respectively. Of the six highly pathogenic isolates, four belonged to the phylogenetic group B2, one to B1 and one to D. Regarding the intermediate pathogenic isolates, two belonged to B2 and one to B1, while five low pathogenic isolates belonged to B1, two to B2 and one to D. The two non-pathogenic isolates belonged to B1 and A groups (Figure 1).
O and H typing
The serogroups more frequently detected in the E. coli isolates were O2 (n = 3, 15.8%), O83 (n = 3, 15.8%), O75 (n = 2, 10.5%) and O106 (n = 2, 10.5%). Three strains presented a non-typable O antigen. Serotypes O2:HNT (three isolates), O83:H6 (two isolates) and O106:HNT (two isolates) were detected in more than one strain (Figure 1).
Antimicrobial susceptibility test
All the isolates in this study were tested against 15 antimicrobial agents. Multidrug resistance, which was defined as resistance against three or more classes of antimicrobials, was associated with 47.4% of the E. coli isolated from wild birds. Two isolates were susceptible to all antimicrobials. The isolates were more resistant to tetracycline (52.6%), nalidixic acid (52.6%), sulfamethoxazole/trimethoprim (36.8%), kanamycin (36.8%), ciprofloxacin (31.6%), amikacin (26.3%), nitrofurantoin (26.3%), ampicillin (21.0%), cefoxitin (21.0%), gentamicin (21.0%), norfloxacin (21.0%), amoxicillin-clavulanic acid (15.8%), ceftriaxone (10.5%), chloramphenicol (5.2%) and ceftiofur (0%). Table 1 shows the resistant profile of each isolate.
Table 1. Antimicrobial resistance phenotypes of E. coli isolates from wild birds (n = 19).
Pulsed-field gel electrophoresis (PFGE)
Genetic diversity was analysed in the 18 strains of E. coli that were isolated from wild birds by PFGE (one strain was non-typable with XbaI), which revealed 17 distinct restriction patterns. Accordingly, the 18 typable isolates were clustered into 15 groups and demonstrated high heterogeneity. Isolates 14F and 14i shared the same PFGE fingerprint pattern and represented one cluster, whereas two other clusters contained two closely related isolates each. The remaining 12 E. coli isolates demonstrated distinct PFGE profiles (Figure 1).
Multilocus sequence typing (MLST)
MLST was performed with 19 E. coli isolates and revealed nine distinct STs. Those more frequently found were ST58 (n = 2), ST789 (n = 2), ST1858 (n = 2) and ST2567 (n = 2), while ST156, ST963, ST1287, ST3310 and ST5829 were detected in single isolates. Five isolates contained unknown STs.
Discussion
Due to the limited information about the characteristics of ExPEC isolated from wild birds, we performed this study where a number of virulence genes and the patterns of antimicrobial resistance were determined. Wild birds are reservoirs of poultry pathogens, such as avian influenza A (H5N1) and Newcastle disease (Cardenas Garcia et al., 2013; Gerloff et al., 2013). The presence of virulence genes that are reported to be frequent in APEC from commercial poultry and free-ranging frigates (Rodriguez-Siek et al., 2005; Johnson et al., 2006; Zhao et al., 2009; Saviolli et al., 2016) associated with the pathogenicity test suggest that wild birds are a reservoir for E. coli that is pathogenic to poultry. Moreover, the VAGs found in this study were also detected in wild mammals and human ExPEC (Frömmel et al., 2013; Maluta et al., 2014c), suggesting that wild birds might be reservoirs for virulence genes that are transmissible to other ExPEC strains.
The majority of E. coli isolates of high and intermediate pathogenicity from the present study were found in birds of prey and psittacine species, belonging to the B2 phylogroup, similar to other studies where B2 strains from frigate birds, seagulls and broiler chickens with colisepticemia had the highest pathogenicity score (Simões et al., 2010; Barbieri et al., 2014; Saviolli et al., 2016) agreeing with studies that proposed that B2 is the most predominant and virulent in most cases of ExPEC infections (Clermont et al., 2000; Smith et al., 2007). Most of the low pathogenic strains belonged to the B1 group, the non-pathogenic strains belonged to A and B1 groups, while the D group contained one highly pathogenic isolate and one of low pathogenicity. Based on these results, we can infer that the ECOR typing system also gives a good correlation between phylogenetic background and virulence for E. coli from wild birds.
All B2 phylogenetic group isolates showed pathogenicity in the in vivo infection model and belonged to O2:HNT, O83:H6, O32:H6, O75:H2, and O75:H14 serotypes. This fact is of great importance because these serogroups presented zoonotic potential and were implicated in extra-intestinal human diseases such as urinary tract infections, newborn meningitis and septicaemia (Ewers et al., 2007; Maluta et al., 2014c). Besides that, O2 is one of the serogroups most associated with APEC strains worldwide (Moulin-Schouleur et al., 2007) and this serogroup was also found in free-ranging frigates belonging to the B2 phylogroup (Saviolli et al., 2016), suggesting that either wild birds are reservoirs for E. coli pathogenic for poultry or that poultry could transmit this pathogen to wild birds, since there is environmental pollution with MDR bacteria from poultry farms (Ferreira et al., 2014).
Almost 50.0% (n = 9) of the E. coli strains presented MDR, they were obtained from psittacine species (n = 4), birds of prey (n = 3), passerine species (n = 1) and waterfowl (n = 1); however studies conducted in Australia and Bangladesh detected only 1.5% and 22.7% of MDR E. coli from wild birds, respectively (Hasan et al., 2012; Blyton et al., 2015). Diet seems to be the main factor influencing the exposure of these birds to MDR bacteria, for example birds of prey may eat animals that are in close contact with humans, while the other birds may ingest contaminated food and water in landfills, livestock farms, wastewater treatment facilities or sewage systems (Allen et al., 2010).
Wild birds should have limited contact with antimicrobial agents, but due to the expansion of urban areas and loss of wildlife habitats the amount of contact that wild birds have with contaminated environments increases, favouring the transfer of antimicrobial-resistant bacteria not only to wild birds, but also to other wild animals, since studies in Europe found wild mammals carrying antimicrobial-resistant E. coli (Costa et al., 2006; Literak et al., 2010b; Smith et al., 2014).
Although MDR E. coli were found in the present study, none of them were ESBL-producers. There are no scientific reports from Brazil in the available literature with which to compare our data; however, the same results were reported in Nigeria where they did not detect ESBL-producing E. coli isolates from wild birds and bats (Iroha et al., 2015). These results contrast with studies conducted in Germany, Mongolia, Russia, Poland and the USA that reported ESBL-producing E. coli in wild birds (Hernandez et al., 2010; Literak et al., 2010a; Poirel et al., 2012; Guenther et al., 2013).
MDR E. coli from the present study belonged to ST58 (phylogroup B1), ST1287 (phylogroup B1), ST1858 (phylogroup B2) and ST789 (phylogroup B2), which have already been found in wild, companion and food animals, environmental and human clinical isolates in different countries (Simões et al., 2010; Dierikx et al., 2012, 2013; Vredenburg et al., 2014; Manges et al., 2015; Hertz et al., 2016).
STs found in E. coli that did not show MDR, such as ST963 (phylogroup D), ST156 (phylogroup B1) and ST2567 (phylogroup B2), were also detected in E. coli from wild birds in North America and Europe, in broiler chickens from Brazil and in human clinical samples in Korea (Poirel et al., 2012; Maluta et al., 2014c; Vredenburg et al., 2014; Alcalá et al., 2015; Yun et al., 2015). The present results highlight the difficulty in determining the human or animal origin of the E. coli strains infecting wild birds.
The PFGE results and profiles of virulence genes demonstrated high heterogeneity among the isolates, indicating the lack of a predominant clone. Some isolates recovered from the same bird (10C/10 T, 3C/3 T and14F/14i) but at different body sites (cloaca and oropharynx or liver and intestine) showed more than 80% similarity and had the same virulence profile. This correlation indicates that E. coli may have migrated from the intestine into other body sites, supporting the idea that the avian gut may be a potential ExPEC reservoir.
In conclusion, our data show that wild birds can be a reservoir for MDR ExPEC that are potentially pathogenic for humans and poultry. This is even more important considering the fact that these birds can shed the bacteria with the faeces intermittently for months or years, thus transmitting this pathogen within the environment and posing a risk to human and bird health.
| Isolates | Antimicrobial resistance |
|---|---|
| 1C | NAL + SXT + TET + NIT + GEN + CIP + NOR + FOX + CRO + |
| 2C | NAL + SXT + TET + CHL + CIP + NOR + |
| 3C | No resistance |
| 3T | NAL + TET + GEN + KAN + AMK + |
| 4C | NAL + SXT + TET + AMP + KAN + AMK + |
| 5T | NAL + SXT + TET + NIT + AMP + KAN + |
| 6C | NAL + SXT + TET + GEN + CIP + AMP + NOR + FOX + KAN + AMK + AMC + |
| 7C | NAL + SXT + TET + NIT + CIP + |
| 8C | NAL + CIP + AMP + FOX + AMC + |
| 9C | AMC + |
| 10C | NAL + SXT + TET + NIT + FOX + KAN + CRO + |
| 10T | KAN + |
| 11T | KAN + |
| 12C1 | No resistance |
| 12C2 | AMK + |
| 13C | TET + AMK + |
| 14F | GEN + |
| 14i | NIT + |
| 15F | NAL + TET + CIP + NOR + |
Abbreviations: AMC, amoxicillin-clavulanic acid; AMP, ampicillin; AMK, amikacin; CHL, chloramphenicol; CIP, ciprofloxacin; CTF, ceftiofur; CRO, ceftriaxone; FOX, cefoxitin; GEN, gentamicin; KAN, kanamycin; NAL, nalidixic acid; NIT, nitrofurantoin; NOR, norfloxacin; SXT, trimethoprim/sulfamethoxazole; TET, tetracycline.
Supplemental table
Download MS Word (183 KB)Related Research Data
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1. See: http://www.apzec.ca/en/APZEC/Protocols/pdfs/ECL_PCR_Protocol.pdf (accessed 9 June 2015).