Finding of extended-spectrum beta-lactamase (ESBL)-producing Enterobacterales in wild game meat originating from several European countries: predominance of Moellerella wisconsensis producing CTX-M-1, November 2021

Introduction Meat can be a vehicle for food-borne transmission of antimicrobial resistant bacteria and antimicrobial resistance genes. The occurrence of extended‐spectrum beta‐lactamase (ESBL) producing Enterobacterales has been observed in meat from livestock production but has not been well studied in meat from wild game. Aim We aimed to investigate, particularly in central Europe, to what extent ESBL-producing Enterobacterales may be present in wild game meat. Methods A total of 111 samples of different types of game meat supplied by butchers, hunters, retail stores and a large game-processing establishment in Europe were screened for ESBL-producing Enterobacterales using a selective culture medium. Isolates were genotypically and phenotypically characterised. Results Thirty-nine samples (35% of the total) yielded ESBL-producing Enterobacterales, with most (35/39) supplied by the game-processing establishment. Isolates included 32 Moellerella wisconsensis, 18 Escherichia coli and one Escherichia marmotae. PCR screening identified bla CTX-M-1 (n = 31), bla CTX-M-32 (n = 8), bla CTX-M-65 (n = 4), bla CTX-M-15 (n = 3), bla CTX-M-8 (n = 1), bla CTX-M-14 (n = 1), bla CTX-M-55 (n = 1), and bla SHV-12 (n = 2). Most E. coli belonged to phylogenetic group A (n = 7) or B1 (n = 9), but several isolates belonged to extraintestinal pathogenic E. coli (ExPEC) sequence types (ST)58 (n = 4), ST68 (n = 1) and ST540 (n = 1). Whole genome sequencing of six selected isolates localised bla CTX-M-1 on megaplasmids in four M. wisconsensis and bla CTX-M-32 on IncN_1 plasmids in one M. wisconsensis and one E. marmotae. Forty-eight isolates (94%) exhibited a multidrug-resistance phenotype. Conclusion We found a high occurrence of ESBL-producing Enterobacterales in wild game meat, suggesting wildlife habitat pollution and possible microbial contamination events occurring during skinning or cutting carcasses.


Introduction
Bacterial resistance to antimicrobials is one of the most imminent threats to global public health [1]. Enterobacterales that produce extended-spectrum beta-lactamase (ESBLs) are of particular concern, since they are resistant to many beta-lactam antibiotics including extended-spectrum penicillins, monobactams and third generation cephalosporins, all of which are categorised by the World Health Organization (WHO) as critically important antimicrobials for human medicine [2,3]. Since the late 1990s there has been a rapid global emergence and dissemination of Escherichia coli and other Enterobacterales producing CTX-M type ESBLs within the healthcare setting, in livestock, and in food of animal origin [2].
During 2019-2020, specific monitoring of retail livestock meat at European Union (EU)-level showed that on average 23.4% of the broiler meat samples, 5.5% of the pork samples, and 4.3% of the beef samples contained ESBL-producing E. coli [4]. Throughout 2019 and 2020, in the EU and in Switzerland, monitoring of ESBL-producing E. coli in these meat categories was mandatory, but not in hunted wild game meat [4,5]. Wild game meat is gaining popularity in many countries and appeals to a growing demand for foods that are both nutritious and represent an alternative to meat from intensive livestock production [6]. Despite the growing importance of wild game meat, there are few reports addressing antimicrobial resistance (AMR) in this food category and most studies that report the possibility of human exposure to ESBL-producing Enterobacterales and transmissible bla ESBL genes via the consumption of meat are focused on meat from domestic foodproducing animals and do not include wild game meat [7]. Therefore, this study was designed to assess the prevalence of ESBL-producing Enterobacterales in wild game meat collected during November 2021.

Sample collection
The study comprised the analysis of 111 wild game meat samples collected during November 2021. The collection included 38 meat samples of red deer (Cervus elaphus), 42 of roe deer (Capreolus capreolus), 28 of wild boar (Sus scrofa), two of chamois (Rupicapra rupicapra) and one of European fallow deer (Dama dama).
The meat samples (which excluded offal) were either supplied by Swiss hunters from animals shot during the hunting season of 2021 (n = 21) or purchased at Swiss butcher shops (n = 22) and retail stores (n = 18). In addition, 50 meat samples were obtained from a large game meat processing establishment located in Slovenia. The establishment processes domestic and imported hunted game animals and produces game meat and meat products for the European market, including for Switzerland.
Overall, for the 111 game meat samples, the countries of origin included Austria (n = 9), Croatia (n = 1), Germany (n = 10), Hungary (n = 7), New Zealand (n = 1), Poland (n = 16), Slovenia (n = 22) and Switzerland (n = 45). An overview of the countries of origin and the suppliers of the game meat samples is given in Table 1.

Screening for extended-spectrum beta-lactamase-producing Enterobacterales
Of each meat sample, 10-20 g were placed in a sterile blender bag (Seward, Worthing, United Kingdom (UK)), homogenised at a 1:10 ratio in Enterobacteriaceae enrichment (EE) broth (BD, Franklin Lakes, United States), and incubated at 37 °C for 24 hours.

Detection of bla ESBL genes
Bacterial species found, which included E. coli, E. marmotae and Moellerella wisconsensis were characterised by identification of bla genes. Bacterial DNA was extracted using a standard heat lysis protocol. Screening for bla TEM and bla SHV was carried out using primers described previously [8]. Screening for bla CTX-M alleles belonging to CTX-M groups 1, 2, 8, 9, and 25 was performed as described by Woodford et al. [9]. Amplicons for sequencing bla CTX-M genes were generated using primers described previously [10]. Synthesis of primers and DNA custom sequencing was carried out by Microsynth (Balgach, Switzerland

Antimicrobial susceptibility testing
Antimicrobial susceptibility testing (AST) was conducted on the isolates using the disk-diffusion method according to the guidelines of the Clinical and Laboratory Standards Institute (CLSI) [11]. Antimicrobial substances included ampicillin, amoxicillin/clavulanic acid, cefazolin, cefotaxime, cefepime, nalidixic acid, ciprofloxacin, sulfamethoxazole-trimethoprim, fosfomycin, azithromycin, nitrofurantoin, streptomycin, kanamycin, gentamicin, chloramphenicol, and tetracycline (Becton, Dickinson, Heidelberg, Germany). Results were interpreted according to CLSI breakpoints for human clinical isolates [11]. Multidrug resistance (MDR) was defined as resistance to three or more classes of antimicrobials, counting beta-lactams as one class.

Extended-spectrum beta-lactamaseproducing Escherichia coli phylogenetic analysis and typing
The distribution of phylogenetic groups among the ESBL-producing E. coli was determined by PCR targeting the genes chuA, yjaA, arpA and TspE4.C2, as described by Clermont et al. [12]. Isolates were assigned to one of the following eight phylogenetic groups including the seven (A, B1, B2, C, D, E, F) belonging to E. coli sensu stricto, and one (the eighth), which is the Escherichia clade I.
For multilocus sequence typing (MLST) of E. coli isolates, internal fragments of the seven housekeeping genes (adk, fumC, gyrB, icd, mdh, purA, and recA) were amplified by PCR as described by Wirth et al. [13]. The amplification products were custom sequenced and sequence types (STs) were determined using the E. coli MLST database website (https://enterobase.warwick.ac.uk).

Whole genome sequencing of Moellerella wisconsensis and Escherichia marmotae
Whole-genome sequences of six isolates were determined using short-read (Illumina MiniSeq) and long-read sequencing (MinION, Oxford Nanopore Technologies). The isolates were selected based on the following criteria; they (i) were constituted of Enterobacterales species for which data on genomic location of bla ESBL genes are scarce (E. marmotae and M. wisconsensis), (ii) produced different ESBLs (CTX-M-1 or CTX-M-32), and (iii) were of relatively diverse origins, such as from the Croatian, Hungarian and Polish game meat samples obtained via the Slovenian processing plant or the Swiss sample obtained from a hunter. The isolates were grown on sheep blood agar at 37 °C overnight before DNA isolation. For short-read sequencing, genomic DNA was extracted using the DNA blood and tissue kit (Qiagen, Hombrechtikon, Switzerland). The DNA libraries were prepared using a Nextera DNA Flex Sample Preparation Kit (Illumina, San Diego, CA, US), and sequencing was done on an Illumina MiniSeq (Illumina, San Diego, CA, US). For long-read sequencing, genomic DNA was extracted using the MasterPure Complete DNA and RNA Purification Kit (Lucigen LubioScience, Zürich, Switzerland). Multiplex libraries were prepared using the SQK-LSK109 ligation sequencing kit with the EXP-NBD104 native barcoding expansion kit (ONT, Oxford, UK) and sequenced on a MinION Mk1B device using the FLO-MIN106 (R9) flow cell (ONT). Base calling, demultiplexing, and barcode trimming was performed with Guppy v4.2.2 (Oxford Nanopore Technologies, Oxford, UK) and quality assessed with LongQC v1.2.0 [14].

Statistical analysis
Comparisons of proportions of samples containing ESBL-producing Enterobacterales from meat of different animal species, from different countries of origin of the meat samples, and different meat suppliers (butchers, hunters, production plant or retail stores) were performed by Fisher's exact test. For each comparison, the proportion of meat containing ESBL-producing bacteria in one item was compared to the proportion of ESBLproducers in all other items. The significance criterion was set at p ≤ 0.05. Calculations were performed using GraphPad (https://www.graphpad.com).

What did you want to address in this study?
Wild game meat consumption is gaining popularity in many countries, but the prevalence of antibioticresistant bacteria in this type of meat it is not well known. In November 2021, we collected wild game meat samples from various European suppliers (butchers, hunters, retail stores and a game-meat-processing plant) to investigate for the presence of such bacteria.

What have we learnt from this study?
Wild game meat may contain antimicrobial resistant germs, including bacteria producing enzymes called extended-spectrum beta-lactamases (ESBLs). ESBL-producing bacteria are resistant to many beta-lactam antibiotics, such as third-generation cephalosporins which are used to treat a variety of infections, for example pneumonia and meningitis.

What are the implications of your findings for public health?
The potential transfer of ESBL-producing bacteria to humans constitutes a public health risk, as such bacteria may transmit their antimicrobial-resistance genes to pathogenic bacteria within the human gut, making infections harder to treat.

Extended-spectrum beta-lactamase-producing Enterobacterales' distribution among game meat samples
An overview of the number of samples collected per animal species, country of origin, and supplier is given in Table 2. Overall, 39 (35%) of the 111 game meat samples yielded ESBL-producing Enterobacterales. They included 24 of the 38 samples from red deer, nine of the 42 samples from roe deer, and six of the 28 samples from wild boar. The proportion of samples containing ESBL-producing Enterobacterales was significantly higher among meat from red deer and from roe deer than from other types of meat ( Table 2).
The 39 contaminated meat samples included six of the nine samples from Austria, the single sample from Croatia, five of the seven samples from Hungary, 11 of the 16 samples from Poland, 15 of the 22 from Slovenia, and one of the 45 samples from Switzerland. Proportions of samples with ESBL-producing Enterobacterales from Poland and Slovenia were significantly higher than those from other countries ( Table 2).
The ESBL-producing Enterobacterales were detected in one of the 21 samples supplied by hunters, in 35 of the 50 samples originating from the game meat processing establishment and in three of the 18 samples purchased in retail stores. The proportion of meat samples containing ESBL-producing Enterobacterales was significantly higher among samples originating from the production plant than among samples from other suppliers ( Table 2).

Identification of bla ESBL genes
All 51 isolates were characterised with respect to their ESBL genotypes. Of the 32 M. wisconsensis, 28 Table 1 Country of origin and type of supplier of wild game meat samples, November 2021 (n = 111)

Antimicrobial susceptibility phenotypes
Phenotypic AST revealed that all 51 isolates were resistant to ampicillin and to the first generation cephalosporin cefazolin. Further, 49 (96%) exhibited resistance to the third generation cephalosporin cefotaxime.

Phylogenetic analysis of extended-spectrum beta-lactamase-producing Escherichia coli
Phylogenetic typing allocated 16 of the 18 E. coli isolates to group A or B1, which typically contain commensal E. coli strains. One isolate each (1/18) belonged to extraintestinal pathogenic phylogroup B2 and D, respectively (Figure 1). A total of 14 different STs were identified by MLST. Thereof, only two STs occurred more than once: four E. coli ST58 were found in two red deer samples from Austria and two wild boar samples respectively from Poland and Slovenia; two E. coli ST2179 were identified in one red deer sample from Slovenia and one wild boar sample from Poland ( Figure  1). Of note, ST58 represents an extraintestinal pathogenic E. coli (ExPEC), as do ST68 and ST540, which were found in one wild boar sample from Poland and one roe deer sample from Slovenia in the present study ( Figure 1).

Extended-spectrum betalactamase-producing Moellerella wisconsensis and Escherichia marmotae
Genomes of five M. wisconsensis (four producing CTX-M-1 and one producing CTX-M-32) isolated from meat from different countries including Croatia, Hungary, Poland, and Switzerland were completely resolved to determine the genomic location (plasmid or chromosome) of the bla CTX-M . In addition, the CTX-M-1-producing E. marmotae was included for whole genome sequencing analysis. ESBL-encoding plasmids were identified in each isolate (Table 3). In four of the five M. wisconsensis isolates, bla CTX-M-1 was identified on megaplasmids (pW18-2-a, pW51-a, pW1-a, and pW65-a; ca 270-350 kb) which contained six to nine additional antimicrobial resistance genes (Table 3) and   were largely homologous (67-88% alignment coverage, > 99% sequence identity).
In all instances, the bla CTX-M-1 genes were found on IS26 flanked transposition units containing bla CTX-M-1 -∆wbuCmrx(A)-mph(A) (Figure 2). A search with the megaplasmids as queries did not identify any corresponding hits in the bacterial plasmid database PLSDB. M. wisconsensis W17-3 and E. marmotae W49-2 encoded bla CTX-M-32 on IncN_1 plasmids. Further, each isolate contained between one and four additional non-bla ESBL -encoding plasmids (data not shown).

Discussion
In this study we investigated 111 wild game meat samples obtained locally or imported into Switzerland from the main European game meat-producing countries which are primarily located in Central Europe. We isolated 51 ESBL-producing Enterobacterales including 18 E. coli, 32 M. wisconsensis and one E. marmotae. To the best of our knowledge, this is the first report of MDR ESBL-producing M. wisconsensis and E. marmotae from wild game meat, although CTX-M-1-producing M. wisconsensis was recently described in wild birds in Greece [21], and the isolation of CTX-M-1-producing E. marmotae from an Alpine marmot (Marmota marmotae) in a Belgian zoo was reported in 2021 [22].
The overall prevalence of ESBL-producing Enterobacterales among the game meat samples in this study was 35%, which amounted to a prevalence of ESBL-producing E. coli of 16%. By comparison, for meat derived from domestic food-producing animals in Europe, the prevalence of ESBL-producing E. coli in broiler, pig, and bovine meat was 23.4%, 5.5%, and 4.3%, respectively in 2019-2020 [4]. However, with the majority (32/51) of the ESBL-producing Enterobacterales represented by M. wisconsensis, our data indicate that using E. coli as an indicator may underestimate the true prevalence of ESBL-producing Enterobacterales among samples originating from wildlife and associated environments. M. wisconsensis is a rare opportunistic pathogen whose potential role in the dissemination of bla ESBL genes in wildlife and in the food chain remains to be to be clarified [21].
Among the samples examined in this work, the occurrence of ESBL-producing Enterobacterales in meat differed between animal species, as well as between countries of origin and suppliers. Meat from red deer showed a higher prevalence (24 of 38 ESBL-producing Enterobacterales) than meat from other animals. Conversely, previous studies published respectively in 2012 and 2018 have shown that faecal carriage of ESBL-producing Enterobacterales in wild red deer is very low in Switzerland (0%) and Poland (0.4%) [23,24]. Thus, the high prevalence of ESBL-producing Enterobacterales found in meat from red deer was unexpected. Similarly, the high proportion of meat containing ESBL-producing Enterobacterales which originated from Austria (6/9), Hungary (5/7), Poland (11/16), and Slovenia (15/22) [23,25,26]. Notably, the majority (35/38) of the ESBL positive samples from these countries were provided by the same game processing establishment. By contrast, for meat originating from New Zealand, Germany, and Switzerland the prevalence of ESBL-producing Enterobacterales was very low (0/1, 0/10, and 1/45, respectively). For Germany, these results are supported by a similar study published in 2017, that reported a low prevalence of ESBL-producing E. coli (0.4%) in locally sourced and processed wild game meat [7]. Meat samples from these three countries were all supplied by butchers, hunters, or retailers, indicating that practices in handling, processing or distribution of wild game carcasses and meat may have an impact on the prevalence of ESBL-producing Enterobacterales.
The most common ESBL genotype among our isolates was bla CTX-M-1 , which is an animal-associated bla ESBL gene found frequently in isolates from domestic animals as well as in wildlife [27]. E. coli carrying bla CTX-M-1 have been recovered from faecal samples collected from red deer in Spain in 2013 and in Switzerland in 2011 [24,28] as well as from Enterobacterales isolated from wild boars in Portugal and Czechia described in 2009 and 2010, respectively [26,29]. While the association of bla CTX-M genes with plasmids is well documented for E. coli and other Enterobacterales [30], this is to our knowledge, the first description of bla CTX-M-1 on large (> 270 kb) plasmids in M. wisconsensis. Interestingly, the bla CTX-M-1 -carrying megaplasmids (pW18-2-a, pW51a, pW1-a, and pW65-a, respectively) identified in M. wisconsensis were largely homologous, although the isolates had intentionally been selected to be from diverse origins. This suggests that these plasmids are common in CTX-M-1-producing M. wisconsensis. In addition, the plasmid sequences had no close match in the PLSDB database, raising the possibility that they are not widespread among other Enterobacterales species. However, since the number of sequenced isolates was limited, further investigations that include a higher number of M. wisconsensis are needed to confirm these observations and to assess the molecular features and potential role of these megaplasmids in the dissemination of antimicrobial resistance determinants in wildlife and elsewhere. By contrast, the bla CTX-M-32 carrying plasmid pW17-3-a identified in M. wisconsensis was related to Proteus mirabilis plasmid pJPM24 (query coverage: 86%, identity: 99.95%) isolated from chicken faeces in China (NZ_CP053895.1) [31], and the bla CTX-M-32 harbouring plasmid pW49-2-b from E. marmotae was identical with pWE_H_2, a multidrug resistance plasmid which was found among a complex bacterial community taken from wastewater effluent in Sweden MW574948.1 [32], indicating wide dissemination of these ESBL-encoding plasmids.
Based on MLST, no high-risk clones such as E. coli ST131 or ST69 were detected [33], however, six of the 18 E. coli belonged to STs which have previously been associated with extra-intestinal disease (urinary tract infections) in humans, including E. coli ST58, ST68 and ST540 [34]. The E. coli STs were to a large extent typical of those found in animals and in humans worldwide, therefore the presence of ESBL genes in isolates from wild game meat may be a consequence of environmental exposure of wildlife to antibiotic residues, resistant bacteria, or resistance genes. AMR Enterobacterales and other microbes from shot animals may be transferred to the carcasses.
Our study has some limitations. The first limitation is that the wild game meat from four of the six the countries investigated (Croatia, Hungary, Poland and Slovenia) came from a single meat processing plant. Hence, it was not possible to distinguish if the occurrence of ESBL-producing Enterobacterales in the game meat from these countries represented carriage acquired in the country of origin by the wild animals, or if it was due to potential cross-contamination at the plant. Second, this study did not include environmental or human samples from the meat processing plant which could provide information on potential crosscontamination among the meat samples, the equipment, the environment, and the workers who handle the meat. Third, while all samples were obtained in Switzerland in November 2021 (some imported and some of local origin), the exact date and geographical location (habitat) of when and where the animals were hunted was unavailable. Therefore, the contribution of environmental exposure (e.g. contact with farmed land) could not be assessed. Finally, due to the considerable differences in sampling numbers, the statistical analysis has limited power. In particular the number of samples from some countries was very small, leading to low precision.
Although in general not consumed raw, game meat represents a risk of exposure to ESBL-producing Enterobacterales through direct contact with the raw meat, through consumption of undercooked meat, and through cross-contamination or re-contamination processes between raw meat and foods that are consumed raw. The potential transfer of ESBL-producing Enterobacterales to humans constitutes a public health risk, as such bacteria may transfer antimicrobial-resistance genes to pathogenic bacteria within the human gut.

Conclusions
This study revealed an occurrence of ESBL-producing Enterobacterales in about a third of wild game meat samples from central European countries. Our data further indicate that the prevalence of ESBL-producing Enterobacterales in wildlife may be higher than previously anticipated, and that microbial contamination and cross-contamination events associated with harvesting of wild game carcasses, as well as crosscontamination within game meat processing establishments may contribute to a high contamination of the final meat product.
Raised awareness for adequate handling of wild game meat should be considered to mitigate the risk of transmission of antibiotic resistant Enterobacterales to the consumer.

Ethical statement
No animal was killed for the purpose of providing samples. Ethical approval was not required for this study.

Funding statement
This work was partly supported by the Swiss Federal Office of Public Health, Division Communicable Diseases.

Data availability
Sequencing read data and genome assemblies have been deposited under BioProject accession number PRJNA814053.