Abstract
The detection rate of extended-spectrum β-lactamase (ESBL)-producing Enterobacterales, microorganisms associated with health care settings, has significantly increased worldwide. Moreover, their community incidence has increased in several countries. In this study, we investigated the prevalence and genetic diversity of ESBL-producing Escherichia coli isolated from 547 nonduplicated stool specimens from healthy Japanese individuals, between 2015 and 2019. E. coli were isolated on deoxycholate-hydrogen sulfide-lactose (DHL) agar and identified by MALDI-TOF MS, ESBL were screened through disk diffusion method (cefotaxime with or without clavulanate), and genetic detection and genotyping were performed by PCR and DNA sequencing. Clonal similarities between ESBL-producing and nonproducing isolates were assessed by multilocus sequence typing (MLST). The prevalence of ESBL-producing E. coli was 9.7% (53/547). These bacteria harbored CTX-M genes, from which CTX-M-9 (31/53, 58.5%) and CTX-M-1 (13/53, 24.5%) groups were the predominant. The MLST analysis revealed that ST131 genotype prevailed within ESBL-producing E. coli (15/53), whereas ST95 (10/53) and ST73 (8/53) prevailed among non-ESBL producers, with ST131 being present in only four isolates. Overall, a high prevalence rate of CTX-M-type ESBL-producing E. coli was detected. CTX-M-9 group-producing ST131 predominated among healthy Japanese individuals, similar to that observed in hospital isolates. CTX-M-type ESBL may disseminate clonally among hospital patients and subsequently, within the community.
Introduction
Extended-spectrum β-lactamases (ESBLs), which hydrolyze extended-spectrum cephalosporins and monobactams, but not cephamycins and carbapenems, have significantly increased among Enterobacterales worldwide.1−5 CTX-M-type ESBL is the most common ESBL type identified in some Enterobacterales species in several countries.6,7 CTX-M-type ESBLs can be divided in several subgroups based on their amino acid identities, including CTX-M-1, -2, -8, -9, and -25 as the principal groups, and over 240 CTX-M-type genotypes have been identified.8,9
The global distribution of CTX-M-type ESBL-producing Escherichia coli has been reported, and these isolates are mostly obtained from hospitalized patients.1,10 CTX-M-15 (CTX-M-1 group) is the most predominantly detected CTX-M type ESBL produced by E. coli worldwide, and the mode of spread is epidemic in every area.11,12 However, CTX-M-15 is not predominantly detected in Japan, instead CTX-M-14 or -27 (CTX-M-9 group), which mostly belong to the sequence type (ST) 131, were the predominant subgroups isolated from ESBL-producing E. coli. 13
The main challenge is that the ESBL-producing Enterobacterales are not limited to hospital-acquired infections, but spread to community, animals, and the environment.14−16 A high prevalence of ESBL-producing Enterobacterales has been reported in various animals and food products in several countries.15,17−20 The transmission of ESBL-carrying strains through food and animals to humans is suggested to be a potential risk of further spread to the community.21,22
Recent studies have revealed that >50% of the healthy individuals from Thailand are carriers of ESBL-producing Enterobacterales, which primarily comprise E. coli.23–25 In Japan, the prevalence of ESBL-producing E. coli in clinical isolates is 17.1 − 25.5%26,27; however, the prevalence in healthy individuals is unknown. Moreover, fecal carriage of ESBL-producing isolates is now widely studied in hospital settings, but only a few studies have assessed the fecal carriage by healthy individuals in the community.28–30
This study aimed to investigate the prevalence of fecal carriage of ESBL-producing E. coli and characterized the genetic background of the isolates among healthy individuals in Japan. We described the molecular characteristics of 53 CTX-M-type ESBL-producing E. coli isolates in comparison with those of selected non-ESBL producing E. coli.
Materials and Methods
Study participants and bacterial isolates
The participants enrolled in this study were healthy medical students (20–39 years old, 30% female) who had not been hospitalized nor had taken any antibiotics in the month before sample collection. All of them resided in the Kinki region of Japan. A total of 547 nonduplicated stool specimens were collected between 2015 and 2019. The stool specimens were directly and individually inoculated on deoxycholate-hydrogen sulfide-lactose (DHL) agar and incubated at 37°C for 24 hr. Then, a single colony was randomly selected from the DHL agar plate and each isolate was identified using MALDI-TOF MS (Sysmex bioMérieux Co., Ltd., Tokyo, Japan).
ESBL production by the selected E. coli colonies was confirmed by disk diffusion method using cefotaxime with or without clavulanate, as recommended by the Clinical and Laboratory Standards Institute (CLSI). 29 All ESBL-positive E. coli isolates were tested for further characterization. To compare the characteristics with ESBL-positive strains, the same number of non-ESBL-producing strains were randomly selected from the strains identified as negative.
The protocol was approved by the Ethical Review Committee of the Nara Medical University (project identification code no. 1846). Informed consent was obtained from all healthy individuals enrolled in this study.
Antimicrobial susceptibility testing
The minimum inhibitory concentration (MIC) of the antimicrobial agents, except colistin, was determined by agar dilution method, and the MIC of colistin by broth microdilution method, according to the CLSI guidelines. 31 The antimicrobials tested were ampicillin, piperacillin, cefotaxime, ceftazidime, cefepime, cefmetazole, aztreonam, imipenem, levofloxacin, and amikacin. Clavulanate and tazobactam were used as β-lactamase inhibitors. The results of resistance or susceptibility were interpreted according to the CLSI breakpoints. 31
ESBL genotyping
All ESBL-type E. coli isolates were screened for the presence of blaSHV, blaTEM, and blaCTX-M genes by PCR as described previously. 32 A multiplex PCR was performed to identify five blaCTX-M groups, blaCTX-M-1, blaCTX-M-2, blaCTX-M-8, blaCTX-M-9, and blaCTX-M-25, 32 and the specific identity of the CTX-M was confirmed by DNA sequencing of the PCR products. 33 Sequence alignment and analysis were performed on the NCBI website using the BLAST program.
Genotyping of E. coli
Genotypes of the isolated E. coli, including all the ESBL producers and selected non-ESBL producers, were analyzed by multilocus sequence typing (MLST) using seven housekeeping genes according to the method of Jørgensen et al. 34 The DNA sequence variations were analyzed using MLST database for E. coli.
Transferability of β-lactam resistance gene
The horizontal transferability of blaCTX-M-harboring plasmids was confirmed by conjugation experiment using the broth mating method with sodium azide-resistant recipient strain E. coli J53 as previously described. 35 Exponential-phase Mid-log phase Luria–Bertani (LB) broth cultures of donor strains and recipient strain E. coli J53 were mixed at a volume ratio of 1:10. This mating mixture was incubated for 2 hr at 35°C. Transconjugants were selected on LB agar supplemented with 8 μg/mL cefpodoxime and 100 μg/mL sodium azide, and the presence of blaCTX-M genes was confirmed by PCR. Conjugal transfer frequencies were expressed as transconjugants per recipient cell.
Results
ESBL-producing E. coli and antimicrobial susceptibility
From the 547 total isolates obtained from stool samples of healthy individuals, 411 isolates were identified as E. coli and 53 isolates (9.7%) were confirmed as ESBL-producing E coli. All the ESBL-producing E. coli strains were resistant to third-generation cephalosporins (cefotaxime) and were strongly inhibited by the addition of clavulanate (Table 1). Furthermore, more than half of the isolates were co-resistant to ceftazidime (30/53, 56.6%), and to levofloxacin (31/53, 58.5%), and only one was identified as co-resistant to colistin (1/53, 1.9%). However, all strains were susceptible to cefmetazole, imipenem, and amikacin.
Antimicrobial Susceptibilities of 53 CTX-M-Type ESBL-Producing Escherichia coli Isolates
Interpreted according to the Clinical and Laboratory Standards Institute guidelines. 31
Genotyping of ESBL genes
All the 53 ESBL-producing E. coli isolates were found to be positive for the blaCTX-M gene (Table 2). Genotyping of blaCTX-M revealed that most of the CTX-M producers harbored genes belonging to the CTX-M-9 group (30 strains, 56.6%), including CTX-M-14 (10 strains), -24 (1 strain), -27 (18 strains), and -65 (1 strain); followed by the CTX-M-1 group (13 strains, 24.5%), which included CTX-M-15 (8 strains) and -55 (5 strains). Six (11.3%) and three (5.7%) strains of the blaCTX-M gene-positive strains carried the CTX-M-2 group and CTX-M-8 group, respectively. One E. coli isolate carried both CTX-M-14 and CTX-M-15. Among the 53 CTX-M producers, five isolates co-harbored other β-lactamase genes (one with TEM-1b, three with TEM104, and one with SHV-2a).
Characteristics of blaCTX-M Gene-Carrying Plasmids from ESBL-Producing Escherichia coli Isolated from Healthy Individuals
Described as the number of transconjugants obtained per isolate.
Transferability of blaCTX-M-harboring plasmids
To evaluate the transferability of the blaCTX-M-harboring plasmids obtained in this study, conjugation experiments were performed. Transconjugants were isolated from 21 of the 53 CTX-M-producing E. coli strains that were individually cultured with recipient E. coli J53. The frequency of transfer ranged from 10–1 to 10–9 per recipient cells.
Multilocus sequence typing
All the 53 ESBL-producing E. coli and randomly selected 53 non-ESBL-producing E. coli isolates were genotyped using MLST analysis (Fig. 1). ESBL-producing E. coli belonged to 27 different STs and mainly to ST131 (14/53, 26.4%), followed by ST38 (9/53, 17.0%). The selected 53 non-ESBL-producing E. coli were identified as belonging to 28 different STs. ST131 (4/53, 7.5%) was not predominant among the non-ESBL producers, whereas ST73 (8/53, 15.1%) and ST95 (10/53, 18.9%) were found to be predominant. A comparison of the STs of ESBL producers and non-ESBL producers revealed that only five STs (ST38, 58, 131, 550, and 1193) were commonly shared.

Distribution of STs in ESBL-producing and non-ESBL-producing Escherichia coli. The molecular strain types of the isolated E. coli were analyzed by MLST using a standard protocol for E. coli identification. ▪, number of ESBL-producing E. coli; □, number of non-ESBL-producing E. coli; ESBL, extended-spectrum β-lactamase; MLST, multilocus sequence typing; STs, sequence types. Color images are available online.
Discussion
In this study, we isolated E. coli strains from the fecal specimens of healthy individuals, revealed the characteristics of ESBL-producing and non-ESBL-producing E. coli, examined their genotypes, and analyzed their antimicrobial susceptibilities. We found that the prevalence of fecal carriage of ESBL-producing E. coli among the healthy individuals in Japan was 9.2%, which was slightly higher than previously reported results (6.4% in 2011).28,29
One half of the ESBL-producing E. coli isolates in this study were susceptible to levofloxacin. The resistance rate of levofloxacin observed in this study (54.7% vs. 87.1%) was lower when compared with that of previously reported isolates. 36 In addition, it has been reported that ST131 E. coli tends to be resistant to levofloxacin, and a study has shown that ∼95% of ST131 E. coli clinical isolates were resistant to levofloxacin. 37 In this study, we reported that 26.4% of the ESBL producers belonged to ST131 type, which is much lower than that reported for clinical isolates. 37 This fact could explain the lower resistance rate observed in our study.
The dominance of CTX-M among the ESBL producers was consistent with the global trend. The CTX-M-9 and CTX-M-1 groups are chiefly detected in European and Asian countries, and the presence of CTX-M-2 group is mainly reported in South America.6,38 Recent surveillance has shown that the CTX-M-9 group is predominant among the ESBL-producing E. coli in Japan, 26 whereas other publication reported a similar dominance in isolates from healthy Japanese individuals. 28
The DNA sequence analysis performed in our study revealed that blaCTX-M-27 (CTX-M-9 group) was the most predominant CTX-M-type ESBL gene in healthy individuals (18/53, 34.0%). Accordingly, CTX-M-9 was also predominantly isolated from hospitals in Japan. 26 Moreover, blaCTX-M-8, which is very rarely isolated from hospitals in Japan, was identified in healthy individuals of this study.
According to the MLST analysis, 15 out of the 53 ESBL-producing E. coli were identified as ST131 (28.3%), of which 12 isolates harbored CTX-M-9 (CTX-M-14, -27), two contained CTX-M-1 group (CTX-M-15), and one contained both CTX-M-14 and CTX-M-15. These CTX-M-type enzymes have been reported to be dispersed widely through various Enterobacterales by the transmission of plasmids through different food products and animals.15,17−20
Furthermore, ST38 (9 strains, 17.0%) was the dominant genotype detected. Isolates belonging to ST38 harbored the CTX-M-9 (CTX-M-14: 6, CTX-M-24: 1, and CTX-M-27: 1 strain) and CTX-M-1 groups (CTX-M-15). ST38, ST131, ST405, and ST648 reportedly play an important role in the successful worldwide distribution of CTX-M-producing E. coli. 39 Furthermore, ST38 seems to have a higher preference for CTX-M-14 enzymes, 40 and clones are spread both in hospital settings and in the community.
Among the non-ESBL-producing E. coli, ST131 was found in only four isolates, whereas ST73 (8/53, 15.1%) and ST95 (10/53, 18.9%) were more predominant. Thus, there were evident differences between the ST of ESBL-producing E. coli and that of non-ESBL-producing E. coli. ST73 and ST95 are reported as dominant in extraintestinal pathogenic E. coli lineages that continue to be the common causes of urinary tract infection. Strains of these genotypes are rarely multidrug resistant and seldom associated with ESBL production,41,42 and these clones seem to be found predominantly in intestinal E. coli strains harbored by healthy individuals.
Conjugation experiments revealed that 21 of the 53 ESBL-producing E. coli were able to transfer the CTX-M-coding plasmids to E. coli J53. A rise in the incidence of multidrug-resistant E. coli ST73 with ESBL positivity was reported in the United Kingdom, 43 suggesting a potential threat of ESBL-producing E. coli such as ST73 and ST95 spreading among healthy individuals.
In conclusion, this study revealed a high prevalence of CTX-M-type ESBL-producing E. coli among healthy individuals in Japan. The STs of ESBL-producing E. coli and CTX-M type in the community were similar to those observed in hospitals, indicating that clones similar to those found in hospitals may be spreading around the city. Pursuant to the One Health concept, it is thus important to examine trends throughout the city to deepen the understanding on how resistant bacteria are spread. However, the STs of the ESBL-producing and non-ESBL-producing E. coli in the community were different. Further surveillance and public health endeavors are needed to control the dissemination of ESBL-producing E. coli in Japan.
Ethical Approval
The protocol was approved by the Ethical Review Committee of the Nara Medical University (project identification code no. 1846).
Footnotes
Acknowledgments
The authors thank Kokoro Nakagawa, Tetsuya Matsumoto, and Kazuhiro Yoshimura from the Department of Microbiology and Infectious Diseases, Nara Medical University (Nara, Japan) for their excellent technical assistance.
Disclosure Statement
No competing financial interests exist.
Funding Information
This study was supported by Japan Society for the Promotion of Science KAKENHI (grant numbers 17K10027 and 21K10403).
