Isolation of extended‐spectrum β‐lactamase‐producing Escherichia coli from Japanese red fox (Vulpes vulpes japonica)

Abstract Antimicrobial resistance is a global concern requiring a one‐health approach. Given wild animals can harbor antimicrobial‐resistant bacteria (ARB), we investigated their presence in 11 fecal samples from wild animals using deoxycholate hydrogen sulfide lactose agar with or without cefotaxime (CTX, 1 mg/L). Thus, we isolated CTX‐resistant Escherichia coli from two Japanese red fox fecal samples. One strain was O83:H42‐ST1485‐fimH58 CTX‐M‐55‐producing E. coli carrying the genes aph(3″)‐Ib, aph(3′)‐Ia, aph(6)‐Id, mdf(A), sitABCD, sul2, tet(A), and tet(B), whereas the other was O25:H4‐ST131‐fimH30 CTX‐M‐14‐producing E. coli carrying mdf(A) and sitABCD and showing fluoroquinolone resistance. Thus, the presence of extended‐spectrum β‐lactamase producers in wild foxes suggests a spillover of ARB from human activities to these wild animals.


| INTRODUCTION
Antimicrobial resistance (AMR) in bacteria represents a common global issue in humans, domestic animals, and the environment. The global action plan on AMR published by the World Health Organization (World Health Organization, 2015) states that multiple sectors, comprising humans, animals, and the environment, including wild animals, should be corroborated globally to control the emergence and prevalence of antimicrobial-resistant bacteria (ARB).
Multisectoral approaches have been implemented according to the national action plan on AMR in Japan, as described in the Nippon AMR One Health Report (The AMR One Health Surveillance Committee, 2021). Furthermore, in Japan, ARB in wild animals has been addressed as a component of the environment since 2020, and the prevalence of AMR among Escherichia coli found in several wild animals has been recently reported (Asai et al., 2020;Fukuda et al., 2021;Tamamura-Andoh et al., 2021).
Wild animals are potential sentinels of ARB in the environment, with infections among humans living in wildlife areas posing potential risks for transmitting pathogens, including ARB, to wild animals.
Recent studies in Japan revealed a low prevalence of ARB in freeliving wild animals, depending on their proximity to human activities, feeding habits, behavioral patterns, and habitats (Asai et al., 2020;Tamamura-Andoh et al., 2021). However, wild animals in urban parks (Ikushima et al., 2021) and around animal facilities serve as reservoirs of ARB (Yossapol et al., 2021). Therefore, wild animals in areas close MicrobiologyOpen. 2022;11:e1317. www.MicrobiologyOpen.com to human activities may carry ARB transmitted from humans and domestic animals.
As AMR is a growing public health and pharmacological concern worldwide, it should be recognized that ARB in the excrement of wild animals may also pose a health risk to human beings. In this study, to clarify the possible health risk of ARB associated with wild animals living around humans, we analyzed fecal samples from unidentified animals and their feed habitats. We also estimated the prevalence of ARB in these unidentified animals.

| MATERIALS AND METHODS
Eleven fecal samples from unidentified animals were collected around the parking space of a university (Gifu City, Japan), in the bushes around a pig farm (Gifu City), and in the garden close to a house (Yamagata City, Japan), as shown in Figure 1 and Table 1.
For animal identification, fecal DNA was extracted from the samples using a QiaAmp Fast DNA Stool Mini Kit (ID: 51604; Qiagen) following the manufacturer's instructions, which was then subjected to a polymerase chain reaction (PCR) using a previously reported primer set (MiMammal-U) (Ushio et al., 2017). The PCR products were sequenced at the Life Science Research Center of Gifu University, and the sequences were analyzed using a basic local alignment search tool (BLAST; National Center for Biotechnology Information). The animal species were identified based on genetic and ecological information.
Escherichia coli was isolated using deoxycholate hydrogen sulfide lactose (DHL) agar (Nissui Pharmaceutical) with or without antimicrobials The extracted DNA from five fox fecal samples was amplified using MiMammal-U for mammals (Ushio et al., 2017) and rbcl (Primer 3) for plants (Aziz et al., 2017). Thereafter, the PCR products were purified using an AMPure XP Kit (Beckman Coulter Inc.) and used to prepare sequencing libraries using Nextera XT Index Kit v2 (Illumina). Before sequencing, the library was purified via agarose gel electrophoresis using the Wizard ® SV Gel and PCR Clean-Up System (Promega Co.). The purified library samples were then sequenced using Illumina iSeq with 2 × 150-bp paired-end kits (Illumina), whereas the raw reads were analyzed using QIIME2. For this, the forward and reverse reads were merged, and low-quality (less than 98%) tails were excluded. After the removal of both primer sequences from the assembled reads, quality filtering was performed to remove reads with an expected error rate of 1% or more and short reads of 100 bp or less. Amplicon sequence variants (ASVs) were generated, and denoising was performed using the "unoise" algorithm. Thereafter, ASV identification was performed using  Specifically, multiple genes, (aph(3″)-Ib, aph(3′)-Ia, and aph(6)-Id) encoding aminoglycoside phosphotransferases present in this isolate were responsible for its KAN resistance, whereas both tet(A) and tet(B) were responsible for its TET resistance. The extended-spectrum β-lactamase (ESBL) producer was phenotypically and genotypically defined as multidrug-resistant. Reportedly, CTX-M-15-producing ST1485 E. coli has been observed in community dwellers in Japan (Nakamura et al., 2016). The other isolate from WL211207-2 (with AMP, CFZ, CTX, NAL, and CIP resistance) was O25:H4-ST131-fimH30 CTX-M-14producing E. coli carrying the genes mdf(A) and sitABCD, with quinolone resistance-determining regions mutations in gyrA (S83L, D87N), parC (S80I, E84V), and parE (I529L). Further analysis revealed that the O25:H4-ST131 strain could be classified as H30R/non-Rx. Reportedly, O25:H4-ST131 E. coli is a pandemic clone that is disseminated among humans (Kawamura et al., 2018) and companion animals (Kawamura et al., 2017) in Japan and other countries (Banerjee & Johnson, 2014).
Although CTX-M-27-producing E. coli is dominant among ST131 H30R in humans and companion animals in Japan, CTX-M-14-producing E. coli has also been reported in these populations (Kawamura et al., 2018(Kawamura et al., , 2017. In this study, though both ESBL producers were first isolated from Japanese red foxes, they may have spilled over from humans and domestic animals, including companion animals. However, it does not appear to be a case of continuous carriage in fox(es), as identical ESBL producers were not isolated from two foxes sampled at almost the same location. Thus, identical ESBL producers were not isolated from foxes thereafter, implying that foxes may be transient carriers.
To elucidate the relationship between wild animals and human activity, the DNA metabarcoding method was applied to analyze the feeding habits of the five identified foxes. The presence of a domestic animal, domesticated plant, and forage crop genes in the feces of these wild animals implied possible direct or indirect connection with human activities (agricultural and animal industries). Regarding the fecal samples from the five animals, four contained 99% or more reads identified as fox genes, whereas one sample (WL210405) contained reads corresponding to fox (94.4%) and mole cricket genes (5.6%), as shown in Table 2 were detected in three samples (WL201209, WL211207-1, and WL211207-2) ( Table 2). Notably, oriental persimmons are cultivated in Gifu, while some are wild. Unfortunately, these two types of persimmons could not be distinguished using the methods employed in this study.
Additionally, the genes of domesticated plants, such as cucumber, sunflower, and rapeseed, were detected in WL210121. We also observed the genes of undomesticated plants in sample WL210405.
Of the five fox samples studied, the hosts of two samples (WL210405 and WL211207-2), from which the ESBL producer was isolated, fed on undomesticated plants and fruits, respectively. Even though there are no reports regarding food retention time in foxes, we reasoned that meals, including plant fiber, may be excreted within two days, as is the case in dogs (Burrows et al., 1982). The effects of seasonality and the issue of retention time duration should be fully considered in the subsequent studies. As different types of ESBL producers were found at the same location, it is also necessary to consider the excretion period of the resistant bacteria. Since this study did not observe the intestinal contents associated with domestic animals and plants in the two foxes carrying the ESBL producers, it was impossible to infer the origin of the ARB they carried.

| CONCLUSION
Taken together, this study identified ESBL-producing E. coli in Japanese red foxes for the first time. This finding suggests that appropriate and careful handling of feces dropped around areas of human activities is essential. Notwithstanding, this study had some limitations. For example, it was conducted using a small number of fecal samples from wild animals. Therefore, continuous estimation of AMR pollution among wild animals is required as ESBL producers may spill over from human activities to wild foxes.