On July 17, 2018, we collected water samples from 4 compartments at a sewage treatment plant in Tokyo. The 4 compartments were 1) in-flow sewage in the sedimentation basin, 2) activated sludge in the aeration tank, 3) treated water, and 4) the final effluent to the river after sterilization with sodium hypochlorite. The samples were rapidly transported to the laboratory of Tokyo University of Marine Science and Technology in sterile bottles on ice and were stored in a refrigerator at 4˚C - 6˚C. The isolation of fecal coliforms was carried out by a modification of the method described by Ham et al. [14]. Briefly, an aliquot from each sample was spread onto MacConkey agar plates (Nissui Pharmaceutical Co., Ltd., Tokyo, Japan) and cultivated at 44.5˚C for 24 h. Growing red-colored colonies were selected because the red colonies were thought to be coliform-like bacteria and because the growth of cells at 44.5˚C was also thought to be characteristic of fecal bacteria. Single-colony isolation was performed on the colonies, and the isolates were again cultivated at 44.5˚C for 24 h and then stored at −80˚C.

An antibiotic susceptibility test was performed on the isolates using the disc diffusion method with the following 14 antibiotics, 8 of which have a β-lactam structure: a penicillin compound (ampicillin, AMPC), 5 cephem compounds (cefotaxime, CTX; cefoxitin, CFX; ceftazidime, CAZ; ceftriaxone, CTRX; and cefpodoxime, CPDX), a monocyclic-lactam compound (aztreonam, ATM), 2 carbapenem compounds (imipenem, IPM; and meropenem, MEPM), 2 aminoglycoside compounds (kanamycin, KM; and gentamicin, GM), a tetracycline compound (tetracycline, TC), a quinolone compound (ciprofloxacin, CPFX), and a chloramphenicol compound (chloramphenicol, CP). Each isolate was suspended in 0.9% (w/v) physiological saline and the 0.5 McFarland standard was used to adjust its turbidity. The suspensions were spread on the surface of Mueller-Hinton agar plates (Nissui Pharmaceutical), antibiotic disks (Japan Becton-Dickinson, Tokyo, Japan) were placed on the plates, and they were incubated at 37˚C for 24 h. Growth inhibitory zones around the disks were interpreted using the Clinical and Laboratory Standards Institute (CLSI) criteria [15]. The antibiotic susceptibilities of the isolates were evaluated for their degrees of sensitivity (S), intermedium (I), and resistance (R).

The identification of each strain was performed by 16S rRNA gene amplification and sequencing, followed by comparison of the sequence with homologous sequences deposited in a database. The total DNA in each fecal coliform isolate was extracted by the alkaline lysis method. The 16S rRNA genes were then amplified by polymerase chain reaction (PCR) using the forward primer 27F (5’-AGAGTTTGATCCTGGCTCAG-3’), and the reverse primer 1492R (5’-GGC TACCTTGTTACGACTT-3’). The sequencing was carried out by Eurofins Genomics (Tokyo, Japan). The 16S rRNA gene sequence of the isolates was subjected to a basic local alignment search tool (BLAST) analysis using the US National Center for Biotechnology Information (NCBI) databases.

First, the ESBL production test was performed for the multi-ARB strains by conducting double disc synergy test (DDST), using a Sensi Disc of AMPC/CVA along with CTX, CAZ, and ATM (Japan Becton-Dickinson). Second, in the ESBL-producing bacteria-like strains, ESBL-encoding genes were amplified by PCR using specific primers for the CTX-M-1 group, CTX-M-2 group, CTX-M-9 group, TEM group, SHV group, and ampC [16] [17] [18] under the conditions shown in Table 1. The sequencing was carried out by Eurofins Genomics. The sequences were subjected to a BLAST analysis of the NCBI databases.

Figure 1 shows the colony forming units (CFUs) of fecal coliform-like bacteria

in the stream water of the sewage treatment plant. Each CFU was approximately 5.1 × 10⁵ in the in-flow sewage in the sedimentation basin (1), approximately 8.0 × 10³ in the activated sludge in the aeration tank (2), approximately 1.4 × 10³ in treated water (3), and approximately 7.0 × 10 in the final effluent to the river after sterilization with sodium hypochlorite (4). The CFUs gradually decreased from (1) to (4) and fecal coliform cells were found to be living at low concentrations in the final effluent even after sterilization with sodium hypochlorite. Hypochlorite sterilization conditions for fecal coliforms have been thoroughly researched [19] and the final effluent from sewage treatment plants is generally sterilized at a strength sufficient to kill bacteria completely. It remains unknown why some cells in the sterilization tank survive, but recent studies have reported that biofilm formation by various kinds of bacteria leads to a remarkable increase in their stability against environmental stress [20] [21] [22]. It is thought that the existence of fecal coliform biofilm in the tank might prevent sterilization of the bacteria by sodium hypochlorite.

Table 2 shows all the isolates in the sewage plant and their strain numbers followed by their antibiotic susceptibility. There were 64 strains of isolates and all were found to be fecal coliforms. In (1), we found 13 strains including 2 genera and 3 species (Escherichia coli, Klebsiella pneumoniae, and K. quasipneumoniae);

(1) In-flow sewage in the sedimentation basin; (2) Activated sludge in the aeration tank; (3) Treated water; (4) Final effluent to the river after sterilization with sodium hypochlorite; AMPC, ampicillin; CTX, cefotaxime; CAZ, ceftazidime; CFX, cefoxitin; IPM, imipenem; MEPM, meropenem; ATM, aztreonam; CTRX, ceftriaxone; CPDX, cefpodoxime; KM, kanamycin; GM, gentamicin; TC, tetracycline; CPFX, ciprofloxacin; CP, chloramphenicol; S, sensitivity; I, intermedium; R, resistance.

in (2), we found 20 strains including 4 genera and 7 species (Citrobacter spp., Enterobacter cloacae, E. coli, K. oxytoca, K. pneumoniae, K. quasipneumoniae and K. variicola); in (3), we found 16 strains containing 3 genera and 4 species (C. freundii, E. coli, K. pneumoniae, and K. quasipneumoniae); and in (4), we found 15 strains containing 2 genera and 3 species (E. coli, K. pneumoniae, and K. quasipneumoniae). There were a total of 25 E. coli strains (39% of 64 isolates) and 26 K. pneumoniae strains (41% of 64) from the sewage treatment plant. Seven strains showed sensitivity against all antibiotics, 37 showed mono-resistance or mono-intermedium, and 20 showed multi-resistance or multi-intermedium. At the sewage plant in the present study, there were 11 multi-ARB strains (17% of all fecal coliforms).

Figure 2 shows the numbers of ARFCs isolated from the stream water of the sewage treatment plant. The ratios of multi-ARFCs to all antibiotic-resistant and intermedium FCs were 33% in (1), 5% in (2), 25% in (3), and 33% in (4). Multi-ARFCs were also found to survive even after sterilization treatment, and then seemed to increase in the metropolitan aquatic environment.

Antibiotic susceptibility testing showed the following candidate strains of ESBL-producing bacteria: Klebsiella pneumoniae A-12 having CTX-, CFX-, and ATM-resistance in (1); Escherichia coli D-18 having AMPC-, CTX-, CTRX-, and CPDX-resistance in (3); K. quasipneumoniae B-7 having AMPC-, CAZ-, CTX-, CTRX-, CPDX-, and ATM-resistance in (4); K. pneumoniae B-8 having AMPC-, CTX-, CTRX-, CPDX-, and CFX-resistance in (4); and K. pneumoniae B-11 having AMPC- and ATM-resistance in (4), as shown in Table 3. These strains were examined using the DDST and amplification of ESBL-encoding genes in order to ascertain their ESBL production.

Table 4 shows the DDST results for the candidate strains of ESBL-producing bacteria. K. pneumoniae A-12 was pseudo-positive for the CTX-M-2, TEM, and SHV groups. E. coli D-18 was positive for the CTX-M-1 and TEM groups, as was

AMPC, ampicillin; CAZ, ceftazidime; CTX, cefotaxime; CTRX, ceftriaxone; CPDX, cefpodoxime; CFX, cefoxitin; ATM, aztreonam; MEPM, meropenem; IPM, imipenem; KM,(kanamycin; GM, gentamicin; TC, tetracycline; CP, chloramphenicol; CPFX, ciprofloxacin; S, sensitivity; I, intermedium; R, resistance.

+, positive; ±, pseudo-positive; −, negative.

K. quasipneumoniae B-7. K. pneumoniae B-8 was positive for the CTX-M-9 and TEM groups. And K. pneumoniae B-11 was pseudo-positive for the TEM group. All five strains were assayed by the amplification of ESBL-encoding genes.

Table 5(a) and Table 5(b) show the results of the amplification of ESBL-encoding genes (AEEG) for ESBL-producing bacteria-like strains. The homology of AEEG sequences in the BLAST analysis is summarized in Table 5(a). In K. pneumoniae A-12, the product amplified by the TEM primer showed 93% homology to TEM β-lactamase in Burkholderia sp. LLH-Slr-7 and that amplified by the SHV primer showed 83% homology to Class A β-lactamase SHV-152 in K. pneumoniae, but that amplified by the CTX-M-2 primer showed only 54% homology to lipoate-protein ligase Lp1A in K. pneumoniae. In E. coli D-18, the product amplified by the TEM primer showed 97% homology to blaTEM-84_1_ AF427130 and that amplified by the CTX-M-1 primer showed 92% homology to β-lactamase CTX-M-1 in K. pneumoniae. In K. quasipneumoniae B-7, the product amplified by the CTX-M-1 primer showed 98% homology to β-lactamase CTX-M-15 and that amplified by the TEM primer showed 96% homology to class A ESBL TEM-143 in E. coli. In K. pneumoniae B-8, the product amplified by the CTX-M-9 primer showed 94% homology to ESBL CTX-M-14 in K. pneumoniae while that amplified by the TEM primer showed only 61% homology to β-lactamase TEM in E. coli. In K. pneumoniae B-11, the product amplified by the TEM primer showed 98% homology to TEM β-lactamase in E. coli.

Therefore, the five strains of the ESBL-producing candidates were found to have ESBL-encoding genes as follows: K. pneumoniae A-12 was positive for the TEM and SHV groups; E. coli D-18 and K. quasipneumoniae B-7 were positive for the CTX-M-1 and TEM groups; K. pneumoniae B-8 was positive for the CTX-M-9 group; and K. pneumoniae B-11 was positive for the TEM group (Table 5(b)). The positive and pseudo-positive groups identified by DDST were partially different from the groups identified as positive in AEEG. In conclusion, all five strains were found to be ESBL-producing bacteria having one or two kinds of ESBL-encoding genes.