Antibiotic resistance is a global problem impacting both human and animal health. The One Health concept sets forth that the health of people, animals, and the environment is interconnected, which fully applies to antibiotic resistance as well, as exemplified by the relationship between avoparcin usage and the spread of vancomycin-resistant Enterococci in Europe (Bager et al., 1997). Besides spread of resistant strains, gene flow between bacteria of human and animal origin drives the dissemination of resistance genes (Graham et al., 2019). Zoonotic or environmental reservoirs served as sources for emerging resistance genes, e.g., Kluyvera spp. as source for bla_CTX–M genes, Shewanella algae as source for bla_OXA–48, or Acinetobacter radioresistens as source for bla_{OXA–23–like} genes (Livermore et al., 2007; Poirel et al., 2008; Tacão et al., 2018). Resistant bacteria can spread between humans and their households involving their companion animals, and the environment and wildlife. International travel, trade of animal food products, and wildlife migration further contribute to the global dissemination of antibiotic resistance (Guenther et al., 2011; Hussain et al., 2017, 2019; Zendri et al., 2020).

Escherichia coli is a characteristic example linking One Health and antibiotic resistance, being a frequent and abundant member of both human and animal gut microbiome as well as an important pathogen of humans and animals. The massive usage of antibiotics both in human medicine and animal industry led to contamination of natural environments with antimicrobials, antibiotic resistance genes, and resistant human pathogens (Graham et al., 2014, 2019). Wildlife living in contaminated habitats such as landfills, wastewater, sewage sludge of farms, or exposed directly to feces from livestock and companion animals can acquire resistant bacteria or resistance genes (Graham et al., 2014, 2019). These animals, particularly the highly mobile species, may scatter the resistant bacteria. Wild birds were shown to carry antibiotic resistant bacteria; typical carriers are crows (Loncaric et al., 2013; Jamborova et al., 2015) and gulls (Báez et al., 2015; Zendri et al., 2020), which often utilize human waste as food source. These birds are frequently urbanized, and their droppings pollute the cities, potentially reintroducing strains into the human environment. Because of their migration and/or vagrant behavior, these birds may serve as reservoirs and long-distance vectors both for antibiotic-resistant strains and antibiotic resistance genes (Wang et al., 2017).

Our aim was to investigate the prevalence of ESBL-producing E. coli carried by rooks (Corvus frugilegus ssp. frugilegus, Linnaeus 1758) gathering in a university clinic and to compare these isolates with contemporary and geographically related human-derived isolates.

An increasing number of studies reported high prevalence of ESBL-producing E. coli in wild animals (Wang et al., 2017). Birds are the most studied hosts, where the most frequently found genes are bla_CTX–M–1 and bla_CTX–M–15 (Wang et al., 2017). Birds may serve as long distance vectors of strains/genes of human origin. Franklin’s gulls (Leucophaeus pipixcan) sampled in Chile frequently carried pandemic ST131 CTX-M-15-producing strains, which are highly prevalent in humans in the United States but rarely found in Chile (Báez et al., 2015). Similarly, bla_CTX–M–1 and bla_CTX–M–15 were previously dominant in rooks wintering in Europe (Loncaric et al., 2013; Jamborova et al., 2015). A few recent European studies also reported a low prevalence of bla_CTX–M–55 and bla_CTX–M–24 in rooks (Jamborova et al., 2015; Söderlund et al., 2019). Importantly, an earlier study from 2005 reported lack of ESBL producers in wintering rooks in the Czech Republic (Literak et al., 2007).

Rooks wintering in Hungary and in neighboring countries belong to a large population migrating from Russia and Western Asia, belonging to subspecies Western Rook (Corvus frugilegus frugilegus) with a breeding area stretching from the East European Plain to the West Siberian Plain. These populations migrate through the Black Sea–Mediterranean flyway and usually winter in European countries (Madge, 2020). The other rook subspecies Eastern Rook (Corvus frugilegus pastinator) nests in the Central Siberian Plateau and the Manchurian Plain and migrates to China and to Japan through the East Asian–Australasian flyway. On the border of West Siberian Plain and Central Siberian Plateau, there is a hybrid zone of the two subspecies, where birds may intermingle (Madge, 2020).

In the present work, bla_CTX–M–55 and bla_CTX–M–27 were predominant in rooks; bla_CTX–M–55 is rarely reported in Europe from humans but highly prevalent in Southeast Asia (Lupo et al., 2018). It has been suggested that bla_CTX–M–55 in humans in Asia arose from food animal sources, highlighting the importance of One Health (Bevan et al., 2017). Previously, bla_CTX–M–14 and, to a lesser extent, bla_CTX–M–15 were dominant ESBL genes in Asia; recently, bla_CTX–M–55 emerged as the most common ESBL gene in human and animal isolates, while bla_CTX–M–27 started to outcompete bla_CTX–M–14 (Bevan et al., 2017). As bla_CTX–M–55 is dominant in livestock in Asia (Bevan et al., 2017), and manure is often used to fertilize crop fields and may contain ESBL-producing E. coli, rooks foraging in these may acquire bla_CTX–M–55 producers. This shift in the epidemiology of ESBL genes in Asia may be the cause of the alteration of ESBL genes in rooks as compared with earlier studies (Loncaric et al., 2013; Jamborova et al., 2015).

We hypothesized that Eastern Rooks acquire strains carrying resistance genes prevalent in animals and humans in China and transmit them to Western Rook individuals interacting with carrier Eastern Rooks in the hybrid area. Thus, intermingling rooks may become long-distance vectors mediating spread of strains/genes from Asia to Europe. Similarly, this may have been the route for clade C1-M27 described first in Japan in 2006 then in Korea in 2008 spreading since to Europe and to America (Matsumura et al., 2016). Similar spread routes of H5N1 avian influenza virus was reported extensively in different bird species (Gauthier-Clerc et al., 2007). High similarity between ST131 C1-M27 isolates of rook and human origin may also be explained by acquisition of these strains by rooks in Hungary, suggesting bidirectional transfer. Food importation may be another potential pathway where bla_CTX–M–55 can spread from animals to humans toward Europe as exemplified by the dissemination mcr-1 resistance gene since a lot of poultry and pork are imported from China to Europe (Hasman et al., 2015) and the detection of these genes can often be traced back there (Lupo et al., 2018).

The majority of rook isolates of phylogroup B2 had indistinguishable macrorestriction profile (Supplementary Figure 2) identified as ST24 carrying bla_CTX–M–27. ST24 was reported rarely, mainly from diarrheic rabbits, cattle, and humans (Moura et al., 2009; Xiong et al., 2012). All ST24 isolates carried the eae gene (Supplementary Table 1), which encodes a major virulence gene of enteropathogenic E. coli (EPEC); lack of the bfp gene indicates that these isolates are atypical EPEC strains, which are reported both from humans and animals (Moura et al., 2009). A study found that atypical EPEC strains of animal origin have potential to cause diarrhea in humans and revealed a close clonal relationship between human and animal isolates (Moura et al., 2009).

The main phylogroup was B1 among rook isolates, forming a large cluster belonging to ST162 and carrying bla_CTX–M–55 (Supplementary Table 1). ST162 was found previously in rooks wintering in Europe with low occurrence (Loncaric et al., 2013; Jamborova et al., 2015). This emerging multiresistant E. coli lineage is now found worldwide colonizing different hosts including livestock, wild animals, humans, rivers, and sewage (Fuentes-Castillo et al., 2020). ST162 was reported from dairy cows with mastitis (Tahar et al., 2020) and from human clinical samples, even associated with bla_NDM–5 in humans (Yoon et al., 2018). These associations raise the concern of dissemination of commensal multiresistant strains in human populations and the diffusion of the antibiotic resistance carried by these strains to other non-commensal, pathogenic strains (Zhuge et al., 2019). Moreover, ST162 E. coli recovered from poultry was identified as a highly virulent clone, despite belonging to phylogroup B1, capable of causing bloodstream infections and meningitis in animal models (Zhuge et al., 2019). In our study, ST162 and ST24 isolates seemed to have clonally expanded in rooks, which is notable as the clonal spread of ESBL-producing E. coli is scarcely documented in wild animals (Lupo et al., 2018).

Sequence types with human importance ST744 and ST131 C1-M27 were found seldom. ST744 is an international high-risk clone identified in our rook, human fecal, and clinical isolates (Table 1). ST744 carrying bla_CTX–M–55 was previously reported from diseased pigs in the United States and from healthy and diseased bovines in France, from wastewater, birds of prey (Guenther et al., 2012; Lupo et al., 2018; Hayer et al., 2020) as well as from healthy and diseased companion animals, and humans (Tacão et al., 2017; Zhong et al., 2017; Zogg et al., 2018). Besides ESBLs, ST744 was sporadically associated with mcr-1, mcr-3, bla_KPC–3, and bla_NDM genes from patients, healthy individuals, and livestock worldwide (Tacão et al., 2017; Zhong et al., 2017; Lupo et al., 2018; Zogg et al., 2018; Li et al., 2019). ST744 isolates of phylogroup A are not as virulent as those belonging to phylogroup B2 and D, but our ST744 rook isolate carried 15 virulence factors (Supplementary Table 1) commonly found in extraintestinal pathogenic E. coli (ExPEC) isolates.

Presence of key virulence genes (iss, iroN, hlyF and ompT, iutA and cvaC) and phylogroup A indicates that our rook ST744 is an avian pathogenic E. coli (APEC) strain, which is the main cause of avian colibacillosis (Johnson et al., 2008). Therefore, wild birds carrying APEC strains might pose a potential economic risk toward poultry; these genes are also frequently found among human ExPEC strains, raising the possibility of zoonotic transmission (Johnson et al., 2008; Zhuge et al., 2019). These suggest that ST744 may be a zoonotic strain capable of colonizing and infecting multiple host species including humans. Moreover, its potential to carry multiple plasmids predisposes it to be involved in transmission of resistance plasmids to other E. coli STs.

Subclade ST131 C1-M27 is associated with clonal spread in humans, and was also reported from great cormorants (Phalacrocorax carbo), mallards (Anas platyrhynchos) (Tausova et al., 2012), gulls (Zendri et al., 2020), companion animals, freshwater, and wastewater (Bevan et al., 2017). Similarly, it occurred in rooks and was also prevalent in human isolates in our study.

Although bla_CTX–M–15 remained the predominant ESBL gene among asymptomatically carried human isolates in this study, the prevalence of ST131-CTX-M-15 E. coli was lower compared with earlier findings (Ebrahimi et al., 2014, 2016a,b), suggesting a slow replacement of C2 subclade carrying bla_CTX–M–15 by the C1-M27 subclade (Merino et al., 2018). ST131 C1-M27 had a higher transmission rate than CTX-M-15-producing ST131 C2 (Merino et al., 2018). C1-M27 isolates often show lower co-resistance to other antimicrobial agents than C2 isolates (Jamborova et al., 2018), which may be advantageous in an antibiotic landscape dominated by beta-lactams as seen in many European countries including Hungary, and particularly the setting where this study was conducted (Tóth et al., 2019).

Both bla_CTX–M–27 and bla_CTX–M–55 are associated with a wide range of plasmid replicons in animal isolates and certain plasmids showed epidemic spread in Asia in humans (Bevan et al., 2017; Lupo et al., 2018). Unlike bla_CTX–M–27 associated with ST131 C1-M27, horizontal transmission is considered to be the main factor driving the dissemination of bla_CTX–M–55 in China (Ho et al., 2013). In our work, bla_CTX–M–55 have been associated with IncN replicon type (Table 1), which often harbors various ESBL genes but rarely bla_CTX–M–55². This association of bla_CTX–M–55 with IncN plasmids carried by ST162 may open a new way to the dissemination of bla_CTX–M–55 in and from Asia toward Europe by bird migration or vagrancy as it may have earlier happened in the case of ST131 C1-M27 (Matsumura et al., 2016).

In summary, increased carriage of ESBL-producing E. coli was found in rooks than reported in previous years. Despite the possibilities for contact, birds and humans shared a low proportion of genotypes. The presence of high-risk clones (ST131 and ST744), high prevalence of Asia-related ESBL genes (bla_CTX–M–55 and bla_CTX–M–27) together with the epidemiological history of ST131 C1-M27 clone suggests that rooks are among the potential vectors for the dissemination of antibiotic resistance genes and resistant strains.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

Ethical review and approval was not required for the animal study because the birds were captured for the purpose of bird ringing. Bird ringing permission number 309 (Birdlife Hungary MME).

GK, LK, ÁL-K, and ÁT: conceptualization. BN, BB, and PG: data curation. BN, EK, KBl, and ÁT: formal analysis. BN, BB, IB, ID, and KBl: investigation. PG, LK, ÁL-K, ID, BN, LM, ÁT, and KBn: methodology. LK, ÁT, KBn, and GK: resources. EK, KBl, ID, and ÁT: software. ÁT and GK: supervision. PG, ÁL-K, ID, LM, ÁT, and KBn: validation. ID and ÁT: visualization. BN and GK: writing—original draft. BN, ÁT, and GK: writing—review and editing. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.