Escherichia fergusonii, an Underrated Repository for Antimicrobial Resistance in Food Animals

ABSTRACT A total of 1,400 samples of food animals (pigs, chickens, and ducks) were collected between July and September 2019 in China to uncover the prevalence of E. fergusonii and its potential role in the evolution of antimicrobial resistance (AMR). An isolation of E. fergusonii was performed and pulsed-field gel electrophoresis (PFGE) was used to uncover the genetic relationship. The AMR of E. fergusonii isolates was comprehensively characterized using broth microdilution-based antimicrobial susceptibility testing, S1-PFGE, southern hybridization, whole-genome sequencing, and in-depth bioinformatics analysis. As a result, a total of 133 E. fergusonii isolates were obtained. These isolates could be grouped into 41 PFGE subclades, suggesting a diverse genetic relationship. The resistance phenotypes of sulfafurazole (97.74%) and tetracycline (94.74%) were the most frequently found. Of the E. fergusonii isolates, 51.88% were extended spectrum beta-lactamase (ESBL)-positive. Forty-three different AMR genes were revealed based on 25 genome sequences harboring mcr-1. Briefly, aph(6)-Id, aph(3′′)-Ib and tet(A) genes were the most frequently observed, with the highest rate being 76.00% (19/25). Three mcr-1-harboring plasmids were identified after Nanopore sequencing, including pTB31P1 (IncHI2-IncHI2A, 184,652 bp), pTB44P3 (IncI2, 62,882 bp), and pTB91P1 (IncHI2-IncHI2A, 255,882 bp). Additionally, 25 E. fergusonii isolates harboring mcr-1 were clustered together with other E. fergusonii isolates from different regions and sources available in GenBank, suggesting a possible random process of mcr-1 transmission in E. fergusonii. In conclusion, E. fergusonii is widespread in food animals in China and might be an important reservoir of AMR genes, especially mcr-1, and facilitate the evolution of AMR. IMPORTANCE E. fergusonii, a member of the genus Escherichia, has been reported to transmit via the food chain and cause diseases in humans. However, the prevalence of multidrug-resistant E. fergusonii, especially mcr-1-positive E. fergusonii isolates, has rarely been reported. Here, we collected 1,400 samples from food animals in three provinces of China and obtained 133 E. fergusonii isolates (9.5%). We found that the prevalence of E. fergusonii isolates was diverse, with high levels of antimicrobial resistance. Among them, 18.8% E. fergusonii isolates carried the colistin resistance gene mcr-1. Thus, E. fergusonii may facilitate the evolution of colistin resistance as a reservoir of mcr-1. As far as we know, the prevalence and AMR of E. fergusonii in the food animals in this study was first reported in China. These findings increase our understanding of the role of E. fergusonii in public health and the evolution of antibiotic resistance.

E scherichia is an important genus of Enterobacteriaceae bacteria. To date, there are nine species in the Escherichia genus, including Escherichia coli, Escherichia fergusonii, Escherichia albertii, Escherichia blattae, Escherichia hermanii, Escherichia vulneris, Escherichia adecarboxylata, Escherichia marmotae, and Escherichia ruysiae (1) (https:// lpsn.dsmz.de/genus/escherichia). E. fergusonii, as a new species distinguished from E. coli, was first detected in a clinical blood sample in 1985 (2). Subsequently, E. fergusonii has been isolated from other clinical cases, including wound infections, urinary tract infections, bacteremia, and diarrhea, indicating its ability to cause diseases and infections (3)(4)(5). In addition, E. fergusonii was also found to be prevalent in foods (6,7) and food animals (8)(9)(10), indicating that it is potentially widespread in the food chain. Based on the above characteristics, E. fergusonii could pose a potential risk to food safety and public health.
It has been reported that multidrug-resistant E. fergusonii has been isolated from both clinical cases and food animals, which suggests that antimicrobial resistance (AMR) might occur in this species (10)(11)(12). In addition, it was reported that diverse AMR genes were detected in E. fergusonii, including some notorious genes such as extended spectrum beta-lactamases (ESBLs) (12), further indicating that E. fergusonii might be an important reservoir of AMR genes. Therefore, it was speculated that E. fergusonii might play an important role in AMR transmission. However, the characteristics of antibiotic resistance in E. fergusonii have not been comprehensively studied.
Colistin, a peptide antibiotic, was one of the antibiotics of last resort for the treatment of multidrug-resistant, especially carbapenem-resistant, Gram-negative pathogen infections (13,14). However, there has been an increasing trend of Enterobacteriaceae resistance to colistin in recent years (15)(16)(17). In 2016, mcr-1, a plasmid-borne resistance gene, was identified as the main factor contributing to the widespread colistin resistance which poses a serious risk to patient health (13,18). An mcr-1-positive E. fergusonii was first detected in 2018 in Guangdong, China (14). Subsequently, an E. fergusonii plasmid with a complete sequence that harbored both mcr-1 and ESBLs was identified from chicken feces in 2019 in Zhejiang, China, suggesting that E. fergusonii might play an important role in the transmission of mcr-1 and pose a serious threat to clinical infection treatment (19). Therefore, a comprehensive examination of the characteristics of mcr-1-positive E. fergusonii is also highly warranted.
In this study, a comprehensive study of the prevalence and characterization of E. fergusonii in food animals was performed. The PFGE profiles and AMR of E. fergusonii isolates were uncovered and discussed. In addition, novel insight into the features and routes of phylogenetic relationship of mcr-1-positive E. fergusonii and mcr-1-harboring plasmids was obtained. These results highlight the serious threat E. fergusonii poses to public health and greatly improve our understanding of its role in the transmission and development of antibiotic resistance.

RESULTS
The prevalence and PFGE profile of Escherichia fergusonii. In total, 133 E. fergusonii isolates were isolated from 1,400 samples at 23 different locations (Fig. 1), with a total isolation rate of 9.50% (Table S1 in the supplemental material). Among these isolates, 84 isolates were isolated from chicken-source samples, with the highest isolation rate being 11.67%. The isolation rates of duck and pig samples were 7.41% (20/270) and 7.07% (29/410), respectively. In addition, the distribution of isolation rates in different locations ranged from 0% to 36%. E. fergusonii was not detected in 9 sampling sites.
To identify the genetic relatedness of E. fergusonii isolates recovered from different animal sources and locations, all 133 isolates were subtyped using pulsed-gel electrophoresis (PFGE). As shown in Fig. 2, at similarity cutoff values of 70% and 85%, DNA profiles of the 133 isolates were grouped into 12 PFGE types and 41 subclades, which included 93 PFGE fingerprints with a similarity of 63.1% ; 100.0%. Among the 12 PFGE types, group B (24/133, 18.05%) was the most common PFGE type, followed by group J (18/133, 13.53%).
The PFGE types of isolates from different animal sources showed different features, and there were many PFGE types in isolates from the same animal source. In detail, there were 30 (30/41, 73.17%) PFGE subclades in chicken-source isolates, and the prevalent subclades were B1, K1, and J1. There were 24 PFGE subclades only found in chicken-source isolates. In addition, there were 12 subclades specifically identified in pig-source isolates and 6 subclades identified in duck-source isolates. It was found that the isolates from different animal sources could share the same PFGE subclade. For example, each of the subclades B4, C1, F4, I5, J2, K1, and L3 included isolates from two different animal sources (Fig. 2). However, there were no subclade present in three different animal source isolates in this study.
Multiple PFGE subclades also could be identified in the isolates from a single region. Among these regions, the number of PFGE subclades in Shaoxing was the greatest (14/41, 34.15%), followed by Xiamen (7/41, 17.07%). In addition, it was found that the   Table S1 in the supplemental material. same PFGE subclade could be prevalent in different regions. For example, K1 was identified in different isolates from Shaoxing, Jiaxing, Fuqing, Xiamen, and Ningbo, and B1 was identified in isolates from Jiande, Shaoxing, Fuqing, and Zhuzhou. Moreover, multiple PFGE subclades could be identified in the isolates from single location (Fig. 2). For example, A2, C1, D1, F4, H1, H2, I3, I4, K2, and K3 were identified in different isolates from farm ZCS2, suggesting that there might be diverse transmission sources in a single farm. Meanwhile, it was also found that different E. fergusonii isolates from a single location had the same PFGE subclade. For example, strains EF68 to EF71, belonging to H4, were isolated from FPS2, and strains EF131 to EF133, belonging to B1, were isolated from HCF3. This phenomenon suggested that clone spread may be occurring. Thus, the PFGE profile results showed that there was a complex transmission mode of E. fergusonii isolates in a single sampling location.
The phenotype of antimicrobial resistance. The MICs for all antimicrobial agents (Table S2) of 133 strains were obtained by microbroth dilution method (Table S3), and the MIC distributions are shown in Fig. S1 in the supplemental material. Among 133 E. fergusonii isolates, the highest level of resistance was observed for sulfafurazole (SUL), with 97.74% isolates, followed by tetracycline (TET) (94.74%), ampicillin (AMP) (84.21%), and sulfamethoxazole (SXT) (83.46%) (Fig. 3A). 18.80% of isolates were resistant to colistin (CT), while none of the E. fergusonii isolates showed resistance to meropenem (MEM) or imipenem (IMI). In addition, 51.88% of E. fergusonii isolates were ESBL-positive. All 133 E. fergusonii isolates were resistant to at least one antimicrobial agent, while 128 isolates (96.24%) showed resistance to four or more antimicrobial agents. In particular, one isolate was found to be resistant to 20 antimicrobials (Fig. S2).
Characterization of the genome features of colistin-resistant E. fergusonii isolates. Whole-genome sequencing was performed on all 25 colistin-resistant E. fergusonii isolates. As shown in Fig. S3, there were 11 plasmid replicons predicted across all isolates, with an average of 5.2 plasmid replicons in each isolate. The IncI2 and IncX1 plasmids were both predicted in 22 of the 25 isolates (88.00%) and were the most prevalent plasmid types in colistin-resistant E. fergusonii.
Acquired AMR genes can be 100% consistent with the phenotypic susceptibility, such as streptomycin, gentamicin, ampicillin, colistin, chloromycetin, sulfafurazole, and trimethoprim. In addition, genome sequences can explain 92% of the AMR profiles for cephalosporin, cefotaxime, tetracycline, and nalidixic acid (including gyrA:p.S83L mutations) (Fig. 4). All 25 mcr-1-positive E. fergusonii isolates were multidrug resistant, which might be caused by a variety of acquired AMR genes and chromosomal point mutations.
Phylogenetic and comparative analysis of mcr-1-harboring E. fergusonii. The complete sequences of 25 mcr-1-harboring E. fergusonii in this study were compared with all complete sequences of E. fergusonii available in the GenBank database (up to June 2021), isolated from a variety of sources and countries, and a phylogenetic tree was built based on a single-nucleotide polymorphism (SNP) analysis using the Maximum Likelihood (ML) method. As shown in Fig. 5, some isolates identified from different hosts, sources, locations, and collection dates could be clustered together. At the same time, isolates from the same region could also be in different lineages. In particular, the E. fergusonii isolates prevalent in different countries could be clustered in the same lineage, suggesting spread between countries. Although the E. fergusonii isolates in this study had a potential phenomenon of clustering together, several isolates showed a closed phylogenetic relationship with isolates from different hosts, sources, locations, and collection dates, such as EF31, EF32, and EF90. Combined with the cluster results of previously reported mcr-1-positive isolates WYDM02000000, CP040806, PUBX01000000, and PUBY01000000, and the 25 mcr-1-positive isolates in this study, it was found that there is no obvious relationship between mcr-1 and strain features, suggesting that the mcr-1 transmission in E. fergusonii might be a random process.
Based on the results above, 3 E. fergusonii isolates (EF31, EF44, and EF91) with the mcr-1 gene carried in different-sized plasmids were further chosen for complete genome sequencing. In addition, these strains were selected for survival verification under colistin pressure by the agar dilution method. It was obvious that they grew well at a colistin concentration of less than 4 mg/mL compared with the control (Fig. 3B). Subsequently, mcr-1-harboring plasmids were identified in the 3 E. fergusonii isolates above, including pTB31P1 (accession no. CP070952) in strain EF31, pTB44P3 (accession no. CP070957) in strain EF44, and pTB91P1 (accession no. CP070874) in strain EF91 (Fig. S5). The plasmid pTB31P1, which is a combination of IncHI2-, IncHI2A-, and IncNtype plasmids, was determined to be 184,652 bp in length and have an average GC content of 46.4%. Except for the mcr-1 gene, 12 antibiotic resistance genes, such as bla TEM-1B , oqxA, and oqxB, were harbored by pTB31P1. The plasmid pTB44P3, which is a IncI2-type plasmid, was determined to be 62,882 bp in length and have an average GC content of 42.9%. This plasmid only harbored the mcr-1 gene. The plasmid pTB91P1, which is a combination of IncHI2-and IncHI2A-type plasmids, was determined to be 255,882 bp in length and have an average GC content of 46.3%. Except the mcr-1 gene, 14 AMR genes, such as floR, fosA3, and bla CTX-M-137 , were harbored by pTB91P1.
An SNP-based phylogenetic analysis was conducted using 124 complete sequences (46 IncHI2-type plasmid sequences and 78 IncI2-type plasmid sequences) of mcr-1-harboring plasmids available in GenBank. As shown in Fig. S6 and S7, pTB31P1 was clustered together with CP034788, KX856066, KY990887, MG656414, MG825373, MH459020, MK355502, and MK477611; pTB44P3 was clustered together with KU934208, KX592672, KX772778, KY363994, KY363995, KY405001, KY471312, KY795977, and KY802014; and pTB91P1 was clustered together with CP033224, KU341381, KX084394, KX856065, MK477608, and MK477609. As shown in Fig. S5, the pTB31P1, pTB44P3, and pTB91P1 plasmids shared the same backbone with their similar plasmids. The main difference were some insertion sequences, AMR genes, and type IV secretion system (T4SS). In particular, plasmid pTB31P1 lacked some T4SSs necessary for conjugation, suggesting that its ability to transfer might be impaired. The highly similar plasmids above were isolated from different hosts, including clinical patients, foods, and food animals in different countries, and were harbored by multiple kinds of Enterobacteriaceae, including E. coli, Shigella sonnei, Salmonella enterica, and Klebsiella pneumoniae, which demonstrates a diverse and complex transmission mode.
To uncover the genetic context of mcr-1 in the different plasmids in this study, a comparative analysis was performed. As shown in Fig. 6, both pTB31P1 and pTB91P1 had a structure of ISApl1-mcr-1-pap2, which showed a high similarity with the plasmids pHNSHP45-2 and pTBMCR421. In the plasmid pTB44P3, the genetic structure of mcr-1-pap2 was similar to that of plasmids pFORC82_3, pHNSHP45, pWH09-3, and p974-MCR, which lacked an ISApl1 upstream of mcr-1 compared with pTB31P1 and pTB91P1.

DISCUSSION
It was reported that clinical cases caused by E. fergusonii have occurred in many countries (4,5,20), posing a serious risk to public health. In particular, E. fergusonii could potentially have an important role in antibiotic resistance transmission, and there might be a complex and diverse transmission chain similar to that of E. coli and other Enterobacteriaceae strains. It has been reported that food animals play a key role in the transmission of antibiotic resistance genes (13,21). However, the prevalence of E. fergusonii in food animals, and its potential role in the transmission of antibiotic resistance, are poorly understood. In this study, a comprehensive investigation of the prevalence and antibiotic resistance of E. fergusonii in food animals was performed. To the best of our knowledge, this is the first report that reveals the features of E. fergusonii in food animals and increases our understanding of the diversity and complexity of E. fergusonii in China.
From the results, there was a 9.50% isolation rate of E. fergusonii in food animals, which suggests that food animals are an important reservoir of E. fergusonii. Chickensource samples had the highest isolation rate, followed by duck-source samples and pig-source samples, which had similar isolation rates. In the past, ducks have often been ignored as a potential reservoir of pathogens; however, there have been increasing reports about the prevalence of pathogens, including multidrug pathogens, from duck sources (22,23). The results in this study are consistent with previous reports which highlight the important role of ducks in the transmission and development of antimicrobial resistance in pathogens, which could pose a serious threat to food safety and public health. In addition, the E. fergusonii isolates in this study had a variety of PFGE profiles, and based on the diversity of PFGE types from different sources and regions, the prevalence of E. fergusonii is diverse. This increases the difficulty of controlling E. fergusonii and highlights the need for long-term monitoring.
The E. fergusonii isolates in this study demonstrated serious antibiotic resistance, especially the colistin-resistant E. fergusonii isolates. Plasmids, a type of mobile genetic element, facilitate the development of antibiotic resistance (24). In this study, a variety of Inc-type plasmids harboring many antibiotic resistance genes were identified in the E. fergusonii isolates. These plasmids are prevalent in different antibiotic-resistant pathogens (16). To our knowledge, this is the first report that describes the characteristics of plasmids harbored by a range of E. fergusonii isolates. Although the database in this study was limited, our results suggest that E. fergusonii could be the host of plasmids carrying AMR genes and mediate the development of antibiotic resistance.
Colistin has been banned as a growth-promoting agent in livestock and poultry breeding in China since 2017 (25), and the colistin resistance rate of E. coli of food animal origins decreased to 5% in 2018, compared with 18.1% to 34.0% before 2017 (17). Surprisingly, the isolation rate of mcr-1-positive E. fergusonii in this study was 18.80%, higher than that of other zoonotic bacteria, such as E. coli (1.55%) in 2016 (26) and Salmonella enterica (1.60%) in 2019 (16) in Zhejiang, which suggests that the prevalence of mcr-1 in E. fergusonii is a serious concern. In addition, comparative analysis of mcr-1-harboring plasmids and the genetic context of mcr-1 suggests that there is longterm and complex communication involving mcr-1 between E. fergusonii and other colistin-resistant pathogens. Particularly, the comparative analysis results of mcr-1-harboring E. fergusonii showed that E. fergusonii was prevalent in various hosts, sources, and FIG 6 Genetic environment of mcr-1 in the plasmids pTB31P1, pTB44P3, and pTB91P1. (A) Alignment of pTB31P1 and pTB91P1 with pTBMCR421 and pHNSHP45-2 which harbored a "ISApl1-mcr-1-pap2" genetic structure. (B) Alignment of pTB44P3 with pFORC82_3, pHNSHP45, pWH09-3, and p974-MCR which harbored a "mcr-1-pap2" genetic structure.
locations, and that mcr-1-negative E. fergusonii had the potential ability to carry mcr-1. This suggests that the role of E. fergusonii in the transmission and development of colistin resistance is more important than previously believed. Therefore, the importance of E. fergusonii in the evolution of the mcr-1 gene should be appreciated, and longterm monitoring of mcr-1-positive E. fergusonii is necessary.
In summary, the prevalence and characterization of E. fergusonii in food animals in China was investigated, as well as its important role in the evolution of antimicrobial resistance. This work suggests that E. fergusonii has a high prevalence in food animals and contributes to serious antibiotic resistance. In particular, E. fergusonii might facilitate the development of antibiotic resistance as a host of plasmids which harbor resistance genes. Given the potential threat to food safety and public health posed by E. fergusonii, long-term monitoring is necessary.

MATERIALS AND METHODS
Sample collection and identification of isolates. A total of 1,400 samples from 410 pigs, 720 chickens, and 270 ducks (feces and anal swabs) were collected between July and September 2019 from 32 different animal farms and slaughterhouses of 14 districts in the Zhejiang, Fujian, and Hunan provinces in China. Detailed information on sample collection is shown in Table S1.
The isolation and identification of E. fergusonii was performed based on previously reported methods (27)(28)(29)(30). In brief, samples were suspended in buffered peptone water (Land Bridge Technology, Beijing, China) and incubated at 37°C for 18 h. The homogenates were then streaked onto Simmons Citrate Agar plates (Land Bridge Technology, Beijing, China) supplemented with ribitol (5 mL/100 mL; J&K Scientific Ltd., Beijing, China) and incubated at 37°C for 36 to 48 h. The presumptive colony of E. fergusonii would be yellow with orange sediment in the center because of fermenting ribitol and D-arabinitol (30,31). Subsequently, a single typical colony (yellow with orange sediment in the center) on each plate was selected to streak onto a Sorbitol MacConkey Agar plate (Land Bridge Technology, Beijing, China) and incubated at 37°C for 24 h. The presumptive colony would be colorless because it could not ferment sorbitol, while E. coli can ferment sorbitol (32). For each sample, a single presumptive colony (colorless, translucent, and flat) was streaked onto a Luria-Bertani (LB) agar plate (Land Bridge Technology, Beijing, China) and incubated at 37°C for 12 to 18 h. The DNA of presumptive E. fergusonii isolates was then prepared using a Bacterial DNA Extraction Kit (Generay Biotech, Shanghai, China). To identify positive E. fergusonii, the obtained DNA was used as a template for PCR amplification using the specific primers EF-F (59-AGATTCACGTAAGCTGTTACCTT-39) and EF-R (59-CGTCTGATGAAAGATTTGG GAAG-39) with 100% identity as previously described (33). E. coli strain ATCC 25922 and a previously isolated E. fergusonii strain were used as quality controls.
Pulsed-field gel electrophoresis and S1-pulsed-field gel electrophoresis. Pulsed-field gel electrophoresis (PFGE) was performed to determine the genotypic relatedness of E. fergusonii isolates, and S1-PFGE was performed to demonstrate the approximate size of plasmids harbored by E. fergusonii. The restriction endonucleases Xba I and S1 nuclease were used in PFGE analysis and S1-PFGE analysis, respectively. Briefly, after E. fergusonii cells were fixed by SeaKem Gold Agarose (Lonza Group AG, USA) and subsequently lysed, the embedded DNAs were digested using Xba I or S1 nuclease (Takara Bio, China) in a 37°C water bath for 3 h. The restricted DNA fragments were separated in 0.5Â Tris-Borate-EDTA buffer (Sangon Biotech, Shanghai, China) at 14°C for 18 h using a CHEF Mapper electrophoresis system (Bio-Rad, USA) with pulse times of 2.2 to 63.8 s. PFGE images were processed by cluster analysis and band-matching using BioNumerics software (Applied Maths, Sint-Martens-Latem, Belgium) to determine the E. fergusonii genotypes and kinships. The band matching was performed with optimization and tolerance settings of 1.0% and 1.5%, respectively. Homology cutoff values of 70% and 85% were used to group the related isolates within the same PFGE-type. Salmonella Braenderup H9812 was used as a standard DNA ladder strain.
Antimicrobial susceptibility testing. For all E. fergusonii isolates, broth microdilution-based antimicrobial susceptibility testing of Gram-negative pathogens (Biofosun, Fosun Diagnostics, Shanghai, China) was used for testing. A total of 30 antimicrobial agents were used for susceptibility testing (Table S2). The breakpoints for each antimicrobial agent were set by the Clinical and Laboratory Standards Institute (CLSI, 2020: M100-S30) (Table S2). E. coli ATCC 25922, E. coli AR Bank no. 0349, and K. pneumoniae ATCC 700603 served as quality control strains, and the acceptable MIC ranges are shown in Table S2.
The agar dilution method was used to confirm the survival situation of mcr-1-harboring E. fergusonii strains under colistin pressure (26). The suspensions were prepared from E. coli strain ATCC 25922 and mcr-1-harboring E. fergusonii strains cultured for 12 to 18 h at 37°C on LB agar. The McFarland standard of the suspensions was adjusted to 0.5 with a densimat (bioMérieux, France). The suspensions were serially diluted 10-fold to 10 26 , after which 10-mL aliquots of the 10 22 -to-10 26 dilutions for each strain were added to colistin-LB agar at concentrations of 0.25 to 16 mg/mL colistin; LB agar without colistin was used as a negative control. The LB agar with the suspensions was then further cultured at 37°C for 16 h.
Southern hybridization. Southern hybridization was performed as described previously (16). The mcr-1-specific probe was labeled using a DIG High Prime DNA Labeling and Detection Starter Kit I (Roche, Sant Cugat del Vallés, Spain) following the manufacturer's instructions.
Whole-genome sequencing and sequence analysis. All mcr-1-positive E. fergusonii were used to perform whole-genome sequencing. An Illumina sequencing library was generated using a NEXTflex DNA sequencing kit (Bioo Scientific, USA). Genome sequencing was carried out using an Illumina HiSeq X10 platform. Paired-end reads were checked for quality and trimmed with Trimmomatic v0.36. All low-quality (Q , 20) data were filtered out. Finally, raw reads were assembled using SPAdes v3.12.0.
Three strains carrying mcr-1 plasmids with different sizes were further sequenced to obtain the complete genome sequences. Libraries were prepared by SQK-LSK109 kit (Oxford Nanopore Technologies [ONT], United Kingdom) and sequenced using R9.4 flow cell technology on a GridION sequencer (ONT). All software systems were operated on their default settings. Guppy v3.2.4 was used for base calling of raw fast5 data and removal of adapter sequences. The reads were then assembled using Unicycler. The assembly was corrected by Illumina reads using Pilon v1.22 software and circularized using Circulator v1.5.1. Gene prediction and genome annotation were performed using the NCBI Prokaryotic Genome Annotation Pipeline.
For convenient sampling, fecal samples were collected for bacterial isolation after the farmer's verbal consent. All the activities in our study were approved by the institutional review board of the Zhejiang Academy of Agricultural Sciences.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 2.5 MB. SUPPLEMENTAL FILE 2, XLSX file, 0.1 MB.

ACKNOWLEDGMENTS
This work was financially supported by the Key Research and Development Program of Zhejiang Province (2020C02031), the National Natural Science Foundation for the Youth (31700007), and the State Key Laboratory for Managing Biotic and Chemical Threats to the Quality and Safety of Agro-products (2010DS700124-ZZ2008).
We report no conflicts of interest relevant to this article.