Antimicrobial Resistance of Waste Water Microbiome in an Urban Waste Water Treatment Plant

Abstract

Waste water treatment plants (WWTP) are considered as a hotspot for the acquisition and dissemination of antimicrobial resistance (AMR). The present study aimed to assess the AMR rate of the waste water microbiome in a selected WWTP and the treatment efficiency. Culture-dependent and PCR methods were used in the AMR study of raw and treated waste water (TWW) microbiomes. The population proportion of heterotrophic plate count (HPC) bacteria resistant to five antibiotic classes was assessed, as well as the AMR phenotype of a total of 238 Enterobacteriaceae and 259 Enterococcus spp. strains. Waste water treatment increased tetracycline- and ciprofloxacin-resistant bacteria by 67% and 61%, as well as the incidence of Enterobacteriaceae resistant to ciprofloxacin, co-trimoxazole, and cephalosporins. Multiple resistance increased, and 8.8% of TWW isolates exhibited an ESBL-producing phenotype, most often encoded by blaTEM and blaCTX-M genes. The most common resistance among Enterococcus spp. was to erythromycin and tetracycline, and despite the increased AMR rate among TWW isolates, only the increase in tetracycline resistance and the decrease in high-level gentamicin resistance were significant. All parameters analysed demonstrated limited removal of resistant HPC or faecal indicator bacteria in the studied WWTP and a positive selective effect towards some of them, most often to ciprofloxacin.

1. Introduction

Antimicrobial resistance is one of the major threats to human and animal health and is becoming an environmental challenge for water resources as well. Many research studies have identified wastewater treatment facilities as hotspots for the acquisition and spread of AMR [1,2,3]. Urban WWTPs are among the most important receivers and sources of environmental AMR [4,5,6]. There is evidence that biological treatment of waste water (WW) with activated sludge or in biofilm reactors cannot provide complete removal or at least a significant reduction in the content of antibiotics (AB), antibiotic-resistant bacteria (ARB), and antibiotic resistance genes (ARGs), and as a result, WWTPs can act as an important source of pollution for receiving surface waters [7,8,9,10,11].

As generally accepted indicators of faecal contamination, E. coli and Enterococcus spp. were used for assessment of AMR in urban sewage and its dynamics during treatment in WWTPs [8,12,13,14,15]. Regardless of the treatment scheme and the multi-fold reduction in the total number of faecal enterobacteria and enterococci, WWTPs are responsible for throwing significant amounts and diversity of resistant bacteria into the environment [1,14,16]. High resistance rates to aminopenicillins, sulphonamides, and tetracyclines have been reported among E. coli isolates in effluents, as well as resistance to tetracycline and erythromycin among enterococci [2,12,13].

According to numerous studies, WW treatment processes have led to an increased relative proportion of the populations of ARB in WWTP effluents, including of multiple resistant bacteria (MAR) [7,17,18]. However, an opposite trend has also been reported for the resistant coliform bacteria in wastewater of several WWTPs [19,20,21]. The treatment efficiency was not uniform for all ARB and ARGs in WW. Numerous studies determined different degrees of reduction in the individual ARGs in the treated effluents [11,21,22,23]. Despite the moderate efficiency of the WW treatment in WWTPs, the ARB and ARGs undergo various dynamic changes depending on the technological scheme of treatment and the origin and contamination of waste water [24,25]. The differences in the composition of waste water related to the type and frequency of used ABs entering the sewage system, as well as the mixing of urban sewage with hospital or industrial WW, affect the prevalence of AMR in WWTP effluents [6,11,16,17,26].

In conventionally treated waste water, beside the genes encoding resistance to ’older’ ABs, the presence of genes conferring resistance to the last resort ABs (such as carbapenems, colistin, vancomycin, etc.) has been detected. E. coli isolates carrying the genes for ESBL-production were recovered from WWTP effluents [27,28,29]. Carbapenem-resistant coliforms have been identified at all stages of WW treatment in the largest Croatian WWTP [30] and in US wastewater [31], as well among Klebsiella pneumoniae strains collected from an Italian WWTP [32]. Colistin resistance mediated by the mcr-1 gene was detected in wastewater samples from seven WWTPs in Germany [33].

Recently, researchers published data showing that the resistant E. coli in urban waste water were more diverse compared with the resistant E. coli in hospital waste water and that the majority of ARB in WWTPs derived from their presence in the urban population [25,34,35]. Because sewage water in urban WWTPs is strongly influenced by human excreta, it was considered that its resistome represents a picture of the general community’s resistome, influenced by the general antibiotic resistance prevalence and antibiotic consumption in the community [36]. The specific features of urban resistomes were consistent with the human antibiotic consumption in the given countries [6]. Considering that AMR levels of E. coli in wastewater can be representative of the level of AMR in the relevant human population it was assumed that they can be used as an early warning system for changes in resistance patterns in the community [25,34].

WWTP operators should ensure that urban wastewater receives appropriate treatment prior to discharge, aiming to eliminate organic components, nutrients (phosphorus and nitrogen), and suspended solids. However, the total microbial load of waste water, including the abundance of ARB and ARG which enter the environment through direct or indirect discharge of treated water or sludge, is not legally regulated. This is one of the reasons why there are no data available in Bulgaria on the levels of ARB in the wastewater microbiome and the impact of wastewater discharge on the environment.

Considering the need for information on the prevalence of AMR in the effluents of Bulgarian WWTPs, reflecting the influence of antibiotic consumption and waste water treatment effectiveness, we focused our efforts on assessing the level of AMR of the wastewater microbiome and gaining a better understanding of the ecology of ARB and their selection and dissemination. The aims of our study were as follows: (a) to quantify heterotrophic bacteria resistant to five AB groups in the wastewater of the selected WWTP; (b) to assess the prevalence of AMR among Enterobacteriaceae and Enterococcus spp. isolated from raw and treated wastewater through culture-dependent and PCR methods; (c) to study the occurrence of ESBL-producing Enterobacteriaceae in raw and treated wastewater; (d) to assess the treatment efficiency of the studied WWTP in limiting the AMR dissemination.

The results obtained will provide an opportunity to gain knowledge about the AMR of the waste water microbiome of a medium-sized Bulgarian WWTP and the effect of conventional mechanical-biological treatment on the spread of AMR.

2. Materials and Methods

2.1. Description of the Studied WWTP

The selected urban WWTP treats the sewage water of a medium-sized town (about 74,000 inhabitants). In the WWTP, waste water influents (about 38,000 cubic m/day) are subjected to primary sedimentation and biological treatment with activated sludge, followed by secondary sedimentation, being disinfected only in the event of an epidemic threat. The treated waste water is discharged into the Yantra River.

2.2. Enumeration of Total Heterotrophic Bacteria and Antibiotic-Resistant Bacteria

Waste water samples with a volume of 500 mL were taken from the untreated wastewater (UWW) at the inlet of the WWTP and from the treated waste water (TWW) at its outlet. The culturable heterotrophic bacteria in waste water samples were enumerated by the spread plate method with R2A agar (HiMedia Laboratories Pvt. Ltd., Mumbai, India), allowing growth of slow-growing bacteria for 7 days incubation at a temperature of 25 °C. Each water sample was analysed for enumeration of total HPC bacteria (on an AB-free medium) and bacteria resistant to individual ABs (on the same medium supplemented with a particular AB). The HPC bacteria resistant to AB substances from five classes were analysed by the AB-dosed medium: ampicillin, Amp—32 mg/L; tetracycline, TE—8 mg/L; chloramphenicol, C—16 mg/L; ciprofloxacin, CIP—4 mg/L; and sulphamethoxazole, Sul—256 mg/L (HiMedia Laboratories Pvt. Ltd., Mumbai, India).

Based on the obtained HPC data pairs, the percentage ratio of bacteria resistant to each individual AB was calculated.

2.3. Enumeration and Isolation of Enterobacteriaceae, and Antibiotic Susceptibility Testing

2.3.1. Enumeration and Isolation of E. coli and Coliforms

The samples of waste water were analysed for enumeration of E. coli and coliform bacteria by the method of membrane filtration and incubation of the membranes on chromogenic HiChrom^TM coliform agar (HiMedia Laboratories Pvt. Ltd., Mumbai, India) at 35 °C for 24 h. Typical blue colonies of E. coli and pink-coloured of coliforms were isolated, sub-cultured on soybean casein digest agar (HiMedia Laboratories Pvt. Ltd., Mumbai, India), and analysed for oxidase and indole production. Oxidase-negative indole-producing colonies were counted as presumptive E. coli, and those that were indole-negative as coliform bacteria. The pure bacteria cultures were stored at −20 °C.

2.3.2. Isolation and Confirmation of ESBL-Producing Enterobacteriaceae

Waste water samples were analysed for isolation of ESBL-producing Enterobacteriaceae by the method of membrane filtration and incubation on chromogenic HiChrom^TM ESBL agar (HiMedia Laboratories Pvt. Ltd., Mumbai, India) at 35 °C for 24 h. Typical pink or purple colonies of presumptive E. coli and bluish-green colonies of coliforms were isolated, sub-cultured, and tested for oxidase and indole production.

All isolates of ESBL-producing Enterobacteriaceae were subjected to phenotypic confirmation of ESBL-production by the double-disc test for determining susceptibility to cefotaxime (CTX) and ceftazidime (CAZ) alone or in combination with clavulanic acid (cefotaxime and clavulanic acid, CEC or ceftazidime and clavulanic acid, CAC). The test was performed with two pairs of AB discs: CTX 30 μg/CEC 20/10 μg and CAZ 30 μg/CAC 20/10 μg. The isolates were assumed as ESBL-positive if the zone of inhibition of CTX or CAZ in the presence of the inhibitor increased, in accordance with the manufacturer’s instructions (HiMedia Laboratories Pvt. Ltd., Mumbai, India).

2.3.3. Biochemical Identification of Enterobacteriaceae Isolates

Biochemical identification was carried out by the ENTEROtest 24N MICROLATEST^® test (Erba Lachema s.r.o., Brno, Czech Republic) according to the manufacturer’s instructions. Based on the resulting ID score, a taxon can be perfectly distinguished from other taxa at ID ≥ 99%, or very well at ID ≥ 95%, and cannot be sufficiently distinguished without additional tests at ID < 90%.

2.3.4. Determination of Antibiotic Resistance Pattern of Enterobacteriaceae Isolates

The antimicrobial susceptibility of Enterobacteriaceae isolates to fifteen ABs from six classes was tested by the disk diffusion method: nine beta-lactams (Amp 10 μg—ampicillin; AUG 20/10 μg—amoxicillin/clavulanic acid; Cx 30 μg—cefoxitin; CTX 30 μg—cefotaxime; CTR 30 μg—ceftriaxone; CPM 30 μg—cefepime; IMI 10 μg—imipenem; PI 100 μg—piperacillin, and PTZ 100/10 μg—piperacillin/tazobactam); two aminoglycosides (S 10 μg—streptomycin and GEN 10 μg—gentamicin); quinolones (CIP 5 μg—ciprofloxacin); antifolates (COT 1.25/23.75 μg—trimethoprim/sulfamethoxazole, i.e., co-trimoxazole); tetracycline, TE 30 μg; and chloramphenicol, C 30 μg.

Disks with the tested ABs (HiMedia, India) were placed on the surface of a Mueller Hinton agar plate (HiMedia, India) inoculated with a calibrated suspension (0.5 MacFarland) of the tested strain. After incubation at 35 °C for 18 h, the inhibition zone diameter around each AB disk was measured. The strains were classified as sensitive (S), resistant (R), or susceptible with increased exposure (I), and those resistant to at least one AB from at least three different classes were classified as multiple resistant [37,38].

2.4. Enumeration and Isolation of Enterococcus spp., and Antibiotic Susceptibility Testing

Waste water samples were analysed for enumeration of faecal streptococci by the method of membrane filtration according EN ISO 7899-2:2003 [39]. Membranes were incubated on selective Slanetz–Bartley agar with TTC (Merck KGaA, Germany) at 35 °C for 24–48 h. Then, presumptive colonies were confirmed by transferring the membrane filter onto a bile esculin agar plate (Merck KGaA, Germany). Esculin-hydrolysing isolates were counted as presumptive Enterococcus spp.

The antimicrobial susceptibility of Enterococcus spp. isolates towards twelve ABs from nine classes was tested: macrolides (E 15 μg—erythromycin); beta-lactams (Amp 10 μg—ampicillin; IMI 10 μg—imipenem) aminoglycosides (HLG 120 μg—high level gentamicin); quinolones (CIP 5 μg—ciprofloxacin); phenicols (C 30 μg—chloramphenicol); tetracyclines (TE 30 μg—tetracycline; TGC 15 μg—tigecycline (a glycylcycline class drug); antifolates (COT 1.25/23.75 μg—co-trimoxazole; glycopeptides (Va 5 μg—vancomycin; TEI 30 μg—teicoplanin); oxazolidinones (LZ 30 μg—linezolid).

2.5. Antibiotic-Resistant Genes of Isolated Faecal Indicator Bacteria

The qualitative PCR method was applied to determine some genes encoding resistance to the selected ABs. For PCR amplification of the targeted ARGs, primers with known oligonucleotide sequences were used. A list of the primers used (synthesized by Microsynth AG, Switzerland), their characteristics, and the corresponding references are presented in Table S1 (Supplementary Files).

2.5.1. DNA Extraction and Purification

In total, 26 Enterobacteriaceae strains with the antibiotic resistance pattern of ESBL production and 28 Enterococcus spp. strains with the MAR phenotype were selected for genetic analyses. Genomic DNA was isolated and purified by GENE MATRIX Bacterial/Yeast Genomic DNA purification kit (EURx^® Molecular Biology Products, Gdansk, Poland).

2.5.2. Determining ARGs in Enterobacteriaceae Isolates

The biochemical identification of E. coli strains was confirmed by testing carriage of the species-specific uidA gene. The selected isolates were tested for the presence of ESBLs encoding genes blaTEM, blaSHV and blaCTX-M using multiplex PCR assay.

2.5.3. Determining ARGs in Enterococcus spp. Isolates

Species identification of enterococcal isolates was performed by multiplex PCR assay of sodA genes for E. faecalis and E. faecium. The MAR isolates were analysed for the presence of genes encoding resistance to erythromycin (ermB), tetracycline (tetM), ciprofloxacin (emeA), gentamicin (aac6′/aph2″), vancomycin (vanA), chloramphenicol (cat), and trimethoprim (dfrA).

2.5.4. PCR Assays for Detection of ARGs

The PCR assays were carried out using onTaq PCR Master Mix 2x (EURx®Molecular Biology Products, Poland) in a 25 μL volume reaction, with primers in a final concentration of 0.4 μM. The PCR amplifications were performed in a T100 Thermal Cycler (Bio-Rad Laboratory, USA), according to the manufacturer’s instructions of the Master Mix kit. As positive controls, E. coli ATCC 35218, E. coli NBIMCC 1164, and E. coli NBIMCC 1223 were used.

The PCR products were visualized in 1.5% agarose gels in Tris-Acetate-EDTA (TAE) solution.

2.6. Statistical Analyses

The Welch t-test was performed to assess the significance of differences between both waste water sampling locations on the percentage of ARB to a particular AB as the dependent variable and sampling location as the factor. Fisher’s exact test was used for categorical variables. p < 0.05 was considered statistically significant.

3. Results

3.1. Microbiological Characteristics of Waste Water

The study was carried out during a one-year period (summer 2023–spring 2024). Waste water samples were analysed during all seasons, taken in duplicate from the inlet and outlet of the WWTP. The summarized data on the numbers of culturable heterotrophic bacteria and faecal indicator bacteria are presented in

Table 1. Microbiological composition of waste water from the studied WWTP during the seasons.

	HPC, CFU/mL		E. coli, CFU/mL		Coliforms, CFU/mL		Enterococci, CFU/mL
	UWW	TWW	UWW	TWW	UWW	TWW	UWW	TWW
winter	5.0 ± 1.7 × 106	3.8 ± 0.4 × 105	na	na	na	na	na	na
summer	2.2 ± 0.8 × 106	7.6 ± 1.1 × 105	5.1 ± 1.4 × 104	4.0 ± 0.2 × 103	7.0 ± 4.6 × 104	4.3 ± 1.4 × 103	3.4 ± 1.5 × 103	6.3 ± 2.3 × 102
autumn	3.3 ± 0.6 × 106	1.5 ± 0.6 × 105	9.3 ± 1.8 × 104	9.3 ± 5.9 × 102	6.7 ± 1.8 × 104	3.5 ± 0.2 × 102	7.1 ± 2.1 × 103	3.7 ± 2.3 × 102
spring	6.5 ± 1.7 × 106	2.9 ± 0.6 × 105	6.8 ± 1.5 × 104	3.4 ± 0.5 × 103	7.0 ± 0.9 × 104	8.2 ± 1.1 × 103	2.8 ± 0.1 × 104	1.2 ± 0.2 × 103
mean	4.3 ± 1.9 × 106	3.9 ± 2.6 × 105	7.1 ± 2.1 × 104	2.8 ± 1.6 × 103	6.9 ± 1.5 × 104	5.3 ± 2.5 × 103	1.3 ± 1.3 × 103	7.5 ± 4.5 × 102

Note: HPC—heterotrophic bacteria; UWW—untreated waste water; TWW—treated wastewater; na—not analyzed.

. The data show that the WW treatment processes removed 91% of HPC bacteria number, 96% of the E. coli and 92% of coliforms, and 94% of Enterococcus spp. content. The reduction in the number of bacteria in treated waste water was by 1.1–1.4 logs.

The populations of HPC bacteria and faecal indicators demonstrate slight seasonal variations in both types of waste water, but the differences between the log values were statistically insignificant (p > 0.05). The highest fluctuations of population density were registered for enterococci.

3.2. AMR of Heterotrophic Bacteria in Waste Water

The AMR data of HPC bacteria in raw and treated waste water (Figure 1) found that the population proportion of bacteria resistant to Sul was the lowest. Bacteria resistant to Amp and CIP were the most numerous in UWW, while bacteria resistant to TE and CIP dominated in TWW. The WW treatment reduced the number of ARB in TWW by 94.0–96.9%, except in summer, when the removal rate was significantly lower, especially for bacteria resistant to TE and Sul. The various types of ARB underwent greater or lesser population changes. Comparative data on the relative proportion of bacteria resistant to individual antibiotics in the two water types showed a very slight decrease in the population of bacteria resistant to Amp and Sul and an increase in the proportion of those resistant to TE and CIP, of 67% and 61%, respectively.

The data show that the biological treatment of waste water in activated sludge biobasins did not lead to a significant reduction in the relative share of resistant HPC bacteria in TWW but led to selection and an increase in the population proportion of those resistant to TE and CIP. That resulted in a change in the ratio of bacterial populations resistant to the individual ABs in TWW, but no statistically significant differences were proven when comparing with UWW. This may be influenced by temporal variations in the relative proportion of HPC bacteria resistant to individual ABs (Figure 2).

In TWW, greater fluctuations in the abundance of bacteria resistant to CIP and TE were observed, in contrast to the relatively unchanged population proportion of bacteria resistant to Amp and C. A positive selective effect of resistance to CIP and TE was found, but a statistically significant increase in TE resistance was registered only in summer (p = 0.022). The proportion of HPC bacteria resistant to Sul significantly increased (p = 0.006) in summer but decreased in autumn (p = 0.019). In winter and spring, UWW and TWW showed insignificant quantitative differences in the proportion of bacteria resistant to individual antibiotics.

3.3. AMR Phenotype of Enterobacteriaceae Isolates from Waste Water

A total of 238 enterobacteria strains were isolated: 125 strains from the waste water samples collected at the WWTP inlet and 113 strains from those taken at the WWTP outlet. Among Enterobacteriaceae isolates from both types of waste water, the highest levels of resistance were found to the b-lactams Amp, AUG, Pi and Cx, and TE, and the lowest to GEN and PTZ (Figure 3). Antibiotic resistance to 3rd and 4th generation cephalosporins was detected, but carbapenem resistance was found only among the UWW isolates.

The effect of WW treatment on the population density of the bacteria resistant to individual ABs was different. The treatment processes decreased the number of Enterobacteriaceae resistant to aminopenicillins and carbapenems but slightly increased the proportion of those resistant to CIP, C, COT, and cephalosporins. The populations of bacteria resistant to TE and GEN remained relatively unchanged. Despite the established changes, the comparison between the resistance rate to individual ABs in UWW and TWW by Fisher’s exact test shows statistically insignificant differences (p > 0.05). The only exceptions were the levels of resistance to CIP (p = 0.042) and COT (p = 0.035).

Some differences in the fate and abundancy of E. coli and coliforms resistant to the tested ABs were observed as a result of WW treatment: an increased incidence of E. coli resistant to piperacillin and 3rd and 4th generation cephalosporins, as well as of those resistant to CIP, C, and COT (

Table 2. AMR phenotype of E. coli and coliforms, isolated from both WW types.

AB	E. coli, %		Coliforms, %
	UWW	TWW	UWW	TWW
Amp	44	41	55	36
AUG	24	22	18	21
Pi	8	13	23	0
PTZ	1	0	0	0
CTX	2	7	14	14
Cx	8	7	36	32
CPM	3	8	18	7
CTR	3	6	14	4
IMI	1	0	3	0
CIP	3	8	9	7
C	4	11	5	4
COT	8	16	5	7
GEN	1	2	5	0
S	12	14	9	7
TE	27	30	27	21
n	91	97	22	28
Sensitive	47.3	43.3	13.6	14.2
AMR1,2	42.9	37.1	77.3	78.6
MAR3–6	9.9	19.6	9.1	7.1

Note: Sensitive—antibiotic-susceptible isolates; AMR_1,2—isolates resistant to one or two classes of antibiotics; MAR_3–6—multiple resistant isolates; n—number of isolates; Tested antibiotics: Amp—ampicillin; AUG—amoxicillin/clavulanic acid; Pi—piperacillin; PTZ—piperacillin/tazobactam; CTX—cefotaxime; CTR—ceftriaxone; CPM—cefepime; Cx—cefoxitin; CIP—ciprofloxacin; C—chloramphenicol; COT—co-trimoxazole; GEN—gentamicin; IMI—imipenem; S—streptomycin; TE—tetracycline.

). Unlike E. coli, the proportion of resistant coliforms decreased, especially of those resistant to the beta-lactams Amp, Pi, cephalosporins, and IMI.

Despite the changes in the population density and diversity of resistant Enterobacteriaceae during the WW treatment, the comparison between UWW and TWW isolates, based on the mean values of the total number of antibiotic classes to which a resistant phenotype was manifested (

Table 3. Number of antibiotic substances or classes to which phenotypic AMR was manifested among the resistant Enterobacteriaceae isolates.

Type of AMR to	UWW		TWW
	Total Number of Resistance, n	Number of R Strains	Total Number of Resistance, n	Number of R Strains
Classes of ABs	1.84 (1.26)	67	2.04 (1.28)	79
Number of ABs	2.97 (2.39)	67	3.04 (2.39)	79

), shows an insignificant difference (p > 0.05). The comparative data on the total number of ABs to which the isolates from both WW types were resistant also showed insignificant change due to the treatment processes.

A total of 32 strains were multiple resistant, 11 of them isolated from UWW (9.7%), and 21 (16.8%) from TWW (Figure 4). The WW treatment processes provided a higher proportion of MAR isolates resistant to up to five classes of ABs, as well as a higher diversity of AMR patterns. Among UWW isolates, two strains resistant to six classes of AB were detected.

All MAR isolates from UWW were resistant to beta-lactams and streptomycin, exhibiting six various phenotypes: three MAR-3 profiles; one MAR-4 profile (with sensitivity to CIP and COT); two MAR-5 profiles (with susceptibility to C or COT); and MAR-6 pattern of resistance to all tested AB classes.

All MAR isolates from TWW were resistant to beta-lactams and exhibited fourteen phenotypic patterns: four different MAR-3 profiles; five different MAR-4 profiles (each with sensitivity to two different AB classes); and four MAR-5 profiles (with sensitivity to one of the six tested AB classes).

Despite the phenotypic differences between the isolates from both waste water types, insignificant differences (p > 0.05) between their values of MAR to individual ABs or classes of ABs were found. MAR isolates from UWW were resistant to 4.27 (±1.19) AB classes or 6.73 (±3.07) ABs, while those from TWW were resistant to 3.90 (±0.83) classes or 6.14 (±2.18) ABs.

It is important to note that three MAR isolates from UWW (6.2%) and nine isolates from TWW (8.8%) expressed the ESBL phenotype, confirmed by the double-disk test. The increased incidence of ESBL-producing isolates in TWW was associated with an increased resistance to other classes of non-beta-lactam ABs, most often CIP, COT, or TE.

3.4. AMR Phenotype of ESBL-Producing Enterobacteriaceae Isolates

In order to obtain a more comprehensive AMR assessment of ESBL-producing Enterobacteriaceae, triplicate WW samples were collected and screened for ESBL-production. As a result, a total of 39 strains of ESBL-producers were isolated, and phenotypic confirmation by the double-disc test with CTX and CAZ was performed (

Table 4. Summary data from the double-disc test for production of ESBL.

Phenotypic Profile of the Isolates	Number of Isolates, n
	UWW	TWW
CTX/CEC (+)	22	10
CTX/CEC (−)	4	3
CAZ/CAC (+)	0	0
CAZ/CAC (−)	26	13

). The data show that 32 of the isolates were sensitive to the inhibitory action of clavulanic acid against CTX and negative against CAZ, and 7 strains were non-ESBL-producers. Ten of the confirmed ESBL-producers were of TWW origin, and twenty-two of them were of UWW origin. All but three strains were identified as E. coli by determining the uidA gene.

All confirmed ESBL-producing strains were resistant to CTX (a cephalosporin of the 3rd generation) and 50% of them were resistant to CPM (a cephalosporin of the 4th generation) (Figure 5) but were susceptible to the carbepenems IMI and MEM. A higher level of resistance to AUG and Cx was found among ESBL-isolates from UWW, and to CAZ among the isolates from TWW. In addition to beta-lactams resistance, a high level of resistance to four classes of non-beta-lactam ABs was found, ranging from 30% to 90%. The lowest resistance was to aminoglycosides (GEN < 10%).

The isolates of UWW demonstrated a higher level of resistance to COT and TE compared to TWW, while those from TWW had an increased resistance to fluoroquinolones (CIP and MOX) and amphenicols (C). The acquisition of resistance to fluoroquinolones by ESBL-strains in the WW treatment is not unexpected, especially when the validity of that trend was already found for the tested Enterobacteriaceae. The same trend was observed for the resistance to C, in contrast to that found for resistance to TE and COT.

ESBL isolates from TWW demonstrated a higher percentage of MAR compared to those from UWW—7 out of 10 strains from TWW (70%) and 13 out of 22 isolates from UWW (59%) exhibited a MAR phenotype (Figure 6). The number of isolates with MAR-3 and MAR-5 increased in TWW, while the MAR-4 isolates had a greater number in UWW.

MAR isolates from UWW were resistant to 3.10 (±0.99) antibiotic classes or 6.30 (±2.50) antibiotics, while those from TWW were resistant to 2.91 (±1.08) or 6.09 (±1.08), respectively. The statistical comparison between the MAR of the isolates from UWW and TWW, based on these parameters, did not show a significant difference (p > 0.05).

3.5. Prevalence of ESBL Genes

The Enterobacteriaceae strains with the ESBL phenotype were tested for genes encoding production of extended-spectrum beta-lactamases, namely blaTEM, blaSHV, and blaCTX-M. Twenty-six strains were tested, with fifteen of them having UWW origin, and the rest isolated from TWW. The genes encoding ESBL production were detected in 13 strains of UWW and 9 strains of TWW. Four strains did not harbour the tested genes.

Six genotypes of ESBL production were established (

Table 5. AMR genotype pattern of ESBL-producing Enterobacteriaceae strains.

Genotype	UWW (n, 15)	TWW (n, 11)	Total n, (%)
blaTEM	3	2	5 (22.7)
blaSHV	0	1	1 (4.5)
blaCTX-M	3	0	3 (13.6)
blaTEM, blaSHV	0	1	1 (4.5)
blaTEM, blaCTX-M	7	4	11 (50.0)
blaTEM, blaSHV, bla CTX-M	0	1	1 (4.5)
non-ESBL	2	2

). The ESBL production was encoded by carriage of one gene alone (mainly blaTEM or blaCTX-M) or together with the second one. The blaTEM gene occurred more often alone than the blaCTX-M gene, and the blaSHV gene was found alone in one strain. The most common ESBL genotype (50% of the isolates) was related to the carriage of blaTEM and blaCTX-M genes. All tested bla genes were harboured together in only one strain.

The blaCTX-M was detected alone only among UWW isolates and found together with the blaTEM gene in 7 out of a total of 11 ESBL isolates. The blaSHV gene was found only among TWW isolates—alone in one strain, Klebsiella, or in combination with the blaTEM gene.

3.6. AMR Phenotype of Enterococcus spp. Isolated from Waste Water

A total of 259 strains of enterococci were isolated from the studied WWTP—134 of them were from UWW and another 125 were from TWW. Sensitivity to ABs’ action was 57.5% among UWW isolates, while among those from TWW, the AB susceptibility decreased to 44.8%. The data show a slight increase in the AMR rate of TWW isolates (Figure 7). Resistance to up to two classes of AB was 33.6% among the isolates from UWW and 44.0% among those from TWW, and MAR strains represented 9.0% and 11.2%, respectively.

In both WW types, Enterococcus spp. isolates resistant to erythromycin and tetracycline predominated, followed by those resistant to imipenem. The level of chloramphenicol and co-trimoxazole resistance was the lowest. All tested enterococci were susceptible to vancomycin, teicoplanin, tigecycline, and linezolid.

Comparative data showed a higher proportion of the majority of resistant phenotypes in TWW compared to UWW but a decrease in resistance to Amp and high-level gentamicin (HLG). Despite the increased AMR rate among the isolates from TWW, only the change in tetracycline resistance was statistically significant (p = 0.002). The decrease in resistance to HLG was significant as well (p = 0.029).

UWW isolates were resistant to 1.72 (±1.11) AB classes or 1.86 (±1.33) ABs, while those from TWW were resistant to 1.71 (± 1.09) or 1.75 (±1.25), respectively, with no significant difference (p > 0.05) between the isolates from UWW and TWW regarding these parameters.

A total of 26 MAR strains were determined—12 of them isolated from UWW and 14 from TWW. Among the TWW isolates, a slight increase in the relative proportion of resistant and MAR isolates was found (Figure 8).

The MAR isolates from UWW were resistant to 3.57 (±0.90) number of AB classes or 4.00 (±1.25) number of ABs, and those from TWW—to 3.58 (±0.85) or 3.86 (±1.23), respectively, as the difference between the isolates from both WW types based on these values was insignificant (p > 0.05).

A broad diversity of MAR phenotypes was found among the isolates from UWW and TWW. A total of 13 MAR phenotypes were detected; three of them manifested among the enterococci isolated from both types of waste water, and five MAR phenotypes were specific to each particular waste water type (

Table 6. Prevalence of MAR phenotypes patterns among Enterococcus spp. isolated from WW.

Phenotypic Pattern	MAR, n of ABs	UWW (n, 12)		TWW (n, 14)
		n	%	n	%
E-TE-C	3	2	16.7	3	21.4
E-TE-b-lactam	3	1	8.3	2	14.3
E-TE-HLG	3	3	25.0	1	7.1
E-TE-CIP	3	0	0	2	14.3
E-b-lactam-COT	3	1	8.3	0	0
TE-b-lactam-CIP	3	1	8.3	0	0
TE-b-lactam-C	3	0	0	1	7.1
E-b-lactam-CIP-HLG	4	1	8.3	0	0
E-TE-b-lactam-CIP	4	0	0	1	7.1
E-TE-CIP-HLG	4	0	0	1	7.1
E-TE-b-lactam-CIP-HLG	5	2	16.7	0	0
E-TE-b-lactam-CIP-COT	5	0	0	3	21.4
E-TE-C-CIP-HLG	5	1	8.3	0	0

Note: UWW—untreated waste water; TWW—treated waste water; MAR, multiple resistance to number of antibiotics; n—number of isolates; %—percentage proportion of a particular phenotypic pattern; Antibiotics: E—erythromycin; b-lactam—beta-lactams (ampicillin and/or imipenem); CIP—ciprofloxacin; C—chloramphenicol; COT—co-trimoxazole; HLG—high level gentamicin; TE—tetracycline.

).

Among UWW isolates, the MAR pattern E-TE-HLG appeared alone in 25% of isolates or in co-occurrence with resistance to other AB classes in another 25% of isolates. That phenotype pattern was predominant, being found in a total of 50% of UWW isolates, in contrast to 14.3% of isolates from TWW. The phenotype E-TE-C was exhibited in 16.7% of UWW isolates, in contrast to its highest frequency (21.4%) among TWW isolates.

3.7. Occurrence of ARGs Among Enterococcus spp. Isolates

A total of 28 strains were tested for detection of six genes encoding resistance to different classes of ABs—11 of them were of UWW origin and 17 were isolated from TWW. Species identification of the tested enterococcal strains was performed by determining the species-specific sodA gene. As might be expected, the enterococcal isolates from the urban sewerage belonged to E. faecalis and E. faecium. Among the UWW isolates, 55% were identified as E. faecalis and 45% as E. faecium, while the number of those identified as E. faecium increased to 59% in TWW.

The comparison in terms of the tested ARGs shows a higher prevalence of tetM, aac(6′)/aph(2″), cat, and dfrA genes among TWW isolates, while emeA and ermB were more common among those from UWW (Figure 9). These data show a significant correspondence between the established phenotypic and genotypic levels of AMR. The highest levels of phenotypic resistance to E and TE were found to correspond to the highest frequencies of the ermB gene for resistance to E and the tetM gene for resistance to TE. However, the prevalence of the gentamicin resistance gene aac(6′)/aph(2″) was high, although the phenotypic expression of high-level gentamicin resistance was less common. The lowest levels of resistance to C or COT corresponded to the lowest frequency of the genes for resistance to chloramphenicol (cat) or trimethoprim (dfrA). The carriage of the emeA gene, encoding efflux removal of ciprofloxacin, was found more often than the manifested resistant phenotype to CIP. All enterococcal strains were sensitive to vancomycin, and no vanA gene was detected.

Comparing the frequency of carriage of the individual ARGs among both enterococcal species, a slightly higher incidence among E. faecalis isolates from both types of wastewater was found, especially of the genes conferring resistance to E, TE, and CIP, while cat and dfrA genes were more frequent among E. faecium isolates (

Table 7. Distribution of antibiotic resistance genes among the studied isolates from wastewater.

Species	Type of WW	sodA	aac(6′)/aph(2″)	emeA	ermB	tetM	cat	dfrA	vanA
E. faecalis	UWW	6	2	4 (4)	6	4	0	0	0
E. faecium	UWW	5	2	2	4	2	0	1	0
E. faecalis	TWW	7	6	6	7	7	1	1	0
E. faecium	TWW	10	6	4	7	7	4	2	0
Total enterococci	UWW	11	4	6 (9)	10	6	0	1	0
Total enterococci	TWW	17	12	10	14	14	5	3	0

).

Although the AMR of UWW and TWW isolates was conferred by 2.45 (±0.93) and 3.63 (±1.67) ARGs, respectively, an insignificant difference (p = 0.38) was observed. The AMR of E. faecalis and E. faecium isolates from UWW was encoded by a similar average number of ARGs, 2.83 (±1.47) and 2.40 (±1.14), respectively, while that of TWW isolates was encoded by 4.43 (±1.40) and 3.40 (±2.67) ARGs, respectively, and the differences between E. faecalis or E. faecium isolates from the compared waste water types were insignificant (p > 0.05).

4. Discussion

The AMR assessment of the waste water microbiome of the selected urban WWTP was based on analyses of three culture-dependent parameters—number of resistant HPC bacteria and antibiotic susceptibility of Enterobacteriaceae or Enterococcus spp. isolates, as well as PCR analyses of ARGs encoding antibiotic resistance in Enterococcus spp. and ESBL production in Enterobacteriaceae. The population study assessed the overall AMR of the HPC bacteria involved in WW treatment, while the AMR analyses of faecal indicators in the WWTP influents revealed the contribution of the respective urban population to the AMR prevalence. The AMR assessment of the WWTP effluents took into account the influence and efficiency of WW treatment.

The prevalence of AMR to five ABs with long-term use (such as Amp, C, TE, CIP, and Sul) tested among the populations of HPC bacteria, Enterobacteriaceae, or Enterococcus spp. showed the following specific features:

-

The highest AMR level was registered among Enterobacteriaceae;

-

High levels of resistance of all studied bacteria to Amp and TE, and low levels of resistance to Sul;

-

Wastewater treatment resulted in the greatest increase in resistance to CIP and TE among HPC bacteria; CIP, COT, and C among Enterobacteriaceae; TE and E among Enterococcus spp.;

-

There was an increased resistance to CIP among all bacteria under study in the treated waste water, although the change was only statistically confirmed for Enterobacteriaceae, and for HPC bacteria only in the summer.

Comparing between the AMR level of UWW and TWW microbiomes on the base of the studied parameters reveals an inefficiency of the conducted WW treatment in the selected WWTP for complete removal of resistant HPC or faecal indicator bacteria and demonstrates a positive selective effect of increasing AMR to some ABs (most often CIP, COT, and TE).

4.1. Prevalence of AMR Among HPC Bacteria

Our investigations on the AMR of the waste water microbiome target ABs widely used to treat infections, such as ampicillin, tetracycline, ciprofloxacin, and sulfamethoxazole. These pharmaceuticals are categorized by the World Health Organization as high-priority ‘critically important antimicrobial agents’ (e.g., ampicillin), as ‘highly important antimicrobials’ (e.g., sulfamethoxazole and tetracycline), or as the highest-priority critically important antimicrobials in human medicine (e.g., fluoroquinolones) [40]. Because AB substances are not completely utilized in the human body, their residues or metabolites leave the organism and end up in sewage or water intakes [7,41]. The amount of excreted beta-lactams, quinolones, tetracyclines, phenicols, and trimethoprim can exceed half of the administered dose [41] and vary from nanograms to micrograms per litre of WW depending on usage intensity, chemical stability, WW treatment effectiveness for ABs removal, etc. [16,42]. AB residues can act as a driving force for selection of ARB and acquisition of AMR.

The data obtained show occurrence of bacteria resistant to the tested ABs in the raw and biologically treated waste water but with different abundances depending on WW type and AB substance. In both WW types, the level of resistance to Sul was the lowest, followed by resistance to C. However, the HPC bacteria in UWW were most commonly resistant to Amp or CIP, while those in TWW demonstrated the most common resistance to TE and CIP. These populational changes in the antibiotic resistance of HPC bacteria are illustrated by the shifts in the ascending order of AMR (Sul < C < TE < CIP/Amp (for UWW)) to the following one: Sul < C < Amp/TE < CIP (for TWW). Despite the increase in the relative proportion of HPC bacteria resistant to CIP or TE in TWW, the trend was not statistically confirmed.

Our data on the increased proportion of HPC bacteria resistant to various ABs in TWW are consistent with the trends established by other researchers [15,22,43]. A significantly high level of heterotrophic bacteria resistant to penicillin, ampicillin, cephalothin, chloramphenicol, tetracycline, and rifampicin have been reported for the secondary effluents of a WWTP of Beijing, China, with relative proportions of 63%, 47%, 55%, 69%, 2.6%, and 11%, respectively [22]. It has been also reported that the increased percentage share of ARB in the total number of bacteria after the WW treatment depended on the technology used or its modifications, with the highest effectiveness of ARB reduction being obtained in WWTPs with A₂O systems and the lowest in WWTPs with mechanical-biological treatment systems [15].

We observed a higher variability in the AMR rates in UWW than TWW, which lets us to suppose that this was related to fluctuations in the ABs content in UWW depending on sampling time. Moreover, the ABs content might occasionally reach concentrations capable of selecting for ARB. Unfortunately, we do not have data available on the AB content in wastewater. In UWW, ciprofloxacin resistance of HPC bacteria appeared to vary the most, with an exceptionally high rate of resistance (26%) detected in summer and the lowest value of 5% in winter; tetracycline resistance was in the range 2–13%, also with an increased value in summer; the proportion of Sul-resistant bacteria increased in summer but decreased in autumn. In TWW, CIP resistance varied in the broadest range (3–33%), followed by TE resistance (2–21%), while the resistance rate to Amp and C was relatively unchanged.

There is a scarcity of data regarding the seasonal changes in the abundance of resistant HPC bacteria in waste water. A study of tetracycline-resistant bacteria in a Polish WWTP has found fluctuation in the HPC population share, with the highest values in April regardless of the type of waste water, while the percentage share of oxytertacycline-resistant bacteria in UWW had the highest values in February and April, but in TWW, this was in April and October [43]. The population proportion of TE-resistant bacteria found in our study (7.0% and 11.8% for UWW and TWW, respectively) was higher than the values of 2.6% or 1.6% determined in the previous studies [22,43]. It should be noted that these studies [22,43] were conducted at higher concentrations of TE (16 μg/mL or 32 μg/mL), which may explain, to some extent, the difference from our data obtained at 8 μg/mL TE. Despite the limited data, it can be assumed that the observed fluctuations are mainly related to the selective pressure of ABs, whose residues probably vary over time as a result of their use, as well as to the different treatment efficiency in the WWTPs, related to ambient temperature, amount of precipitation entering the sewer, etc.

A comparison of the AMR data of HPC bacteria in WW with those from our previous AMR study of the water of the Yantra River (recipient of the effluents from the studied WWTP) reveals significant differences in the levels of resistance to individual AB. In contrast to WW, the population proportion of TE-resistant bacteria in the river water was the lowest, followed by CIP-resistant bacteria, and the highest values were that of Amp and C-resistant bacteria [44]. The quantity of HPC bacteria resistant to CIP, TE, and Amp was greater in the WWTP effluents compared to Yantra River water, which may suggest involvement of the discharged effluents in spreading of resistance to these ABs in the receiving waters.

Our comparative data on the different types of ARB in the two types of WW clearly show the changes that occur in the levels of AMR in wastewater treatment processes, as well as the fact that, despite the rather high level of reduction in the number of ARB, the wastewater discharged from the WWTP significantly loads the aquatic environment with various ARB, thus contributing to the spread of AMR.

4.2. Prevalence of AMR Among Enterobacteriaceae Isolates from Waste Water of the Studied WWTP

E. coli and coliforms are commensal bacteria that enter sewage system with faeces. Some species are pathogenic. Due to their common habitats with other pathogens or ARB in living organisms or during WW treatment, they can acquire resistance by horizontal gene transfer. The increased frequency of resistant Enterobacteriaceae in discharged waste water suggests the possibility of increased prevalence of AMR in the receiving water body.

The AMR assessment of Enterobacteriaceae isolates found the highest resistance rate to aminopenicillins (Amp—up to 46%), ureidopenicillins (PI—20%), and TE (up to 28%) and the lowest to GEN (2%). Resistance to cephalosporins of the 3rd and 4th generations was 9% or 8%, respectively, and a carbapenem resistance of 3% was found only among UWW isolates. The AMR rate of Enterobacteriaceae found in our study was lower or within the range of some previous studies [8,15,18,45]. The data reported for 10 countries from different European regions (3 from western, 4 from northern, and 3 from southern Europe) on the resistant rates to aminopenicillines, fluoroquinolones, 3rd generation cephalosporins, and amynoglycosides found predominantly lower AMR rates than those determined in our study, excepting a higher resistance to Amp [34]. The AMR rates reported for the urban effluents of a large Romanian WWTP were higher in terms of Amp (61.9%), CTX and CTR (42.86%), and COT (76.19%) but similar for CIP, C, and TE [45]. A high occurrence of antibiotic-resistant faecal coliforms has been reported in two urban WWTPs in Ireland: more than 90% of isolates were resistant to Amp; to TE—up to 39.82%; to CIP—up to 31.42%; and to IMI—up to 15.93% [18]. Meanwhile, a much lower level of AMR prevalence was found in another Irish WWTP [14].

Waste water treatment in the studied WWTP reduced the proportion of Enterobacteriaceae resistant to Amp and IMI but increased that of those resistant to cephalosporins, CIP, C, and COT. However, the populations of E. coli and coliforms underwent different changes during the WW treatment: the incidence of resistant E. coli increased, especially those resistant to cephalosporins, CIP, C, and COT, while the abundance of resistant coliforms decreased, especially those resistant to Pi and the cephalosporins CTR and CPM, but the reduction in piperacillin resistance was most significant (p < 0.05). Despite the AMR changes experienced during WW treatment, statistically insignificant differences in resistance determinants were found between UWW and TWW, except for resistance to CIP and COT (p < 0.05). However, the treatment process exerted selective pressure, especially on E. coli, increasing resistance to cephalosporins, CIP, COT, and C. The obtained data on the increased resistance to cephalosporins, CIP, and COT in TWW are consistent with previous studies [8,16].

The greatest fractions of resistant E. coli in UWW were to Amp (44%) and TE (30%), and the lowest were to GEN (1%) and IMI (1%). Resistance rates to CTX and CIP were 2% and 3%, respectively. Comparing E. coli resistance phenotypes with WHO/ECDC data on clinical isolates from Bulgaria, a much lower (from 0.9 to 3.3 times) level of AMR among waterborne E. coli was found [46]. According to a WHO/ECDC report, in Bulgaria, the invasive E. coli had a population-weighted mean value of 54.6% for ampicillin resistance, 14.9% for third-generation cephalosporin resistance, 23.8% to fluoroquinolones, 10.2% to aminoglycosides, and 0.2% to carbapenems [46]. Moreover, E. coli isolates from UWW samples showed a resistance rate strongly correlated with the reported data on clinical isolates (r² = 0.94). These data support the findings of Huijbers et al. (2020) that compared resistance prevalence in E. coli isolates from sewage water to that in clinical E. coli for 10 EU countries (based on data of the European Antibiotic Resistance Surveillance Network) and found a significant relationship (r² = 0.94). Based on these data, the possibility of using wastewater monitoring for the prediction of clinical resistance levels and provision of clinically relevant AMR data has been assumed [34]. This assumption has been supported by comparative AMR data of E. coli from a regional, primary care centre with urban wastewater data, which showed a significant relationship regarding the prevalence of resistance to eleven antibiotics (r² = 0.82) [36], as the AMR rates in municipal sewage isolates were about half of those measured in primary care isolates.

The comparison between the AMR level of Enterobacteriaceae in the WWTP effluents and in the receiving water of the Yantra River showed a slightly higher or similar AMR values in TWW. Moreover, the most common resistance among Enterobacteriaceae isolates from TWW and the river water was towards the ‘old’ ABs—Amp, TE, and COT—with the resistance rates in riverine bacteria being 36%, 25%, and 16%, respectively [44], vs. 41%, 30%, and 16% in TWW. In both types of water, the lowest resistance to GEN (2%) and susceptibility to imipenem was found. However, the river water contained a lower number of Enterobacteriaceae resistant to Cx, AUG, C, CIP, S, and TE, suggesting a possible transitory negative effect of the TWW discharge.

The WW treatment provided a higher proportion of multiple resistant isolates in TWW (16.8%) than in UWW (9.7%). A similar trend of a significant increase in the MAR level of the WWTP effluent compared to the influent water was reported [45], but with many-times-higher MAR values. Besides their higher abundance, the MAR strains from TWW show a greater diversity of resistant phenotypes than UWW, exhibiting 14 MAR phenotypes vs. 6 resistant patterns of UWW isolates.

It is also important to note that 12 MAR strains (3 isolated from UWW and 9 from TWW) expressed the ESBL phenotype. The increased incidence of ESBL producers in TWW was associated with an increased co-resistance to non-beta-lactam ABs, most often CIP, COT, or TE. The occurrence of ESBL-producing Enterobacteriaceae isolates in TWW was 8.8%, which is consistent with the findings of 6% [14] or 11.8% [18] for urban effluents but is lower compared with the reported values of 18% [25] or 19.8% [28]. Data of a much higher prevalence of ESBL-producing Enterobacteriaceae in WWTP effluents have also been reported, including rates of 56% [47], 56.3% [32], 56.3% [48], and about 73% [27], which is probably due to ineffective or insufficient WW treatment, including absence of biological treatment and/or disinfection.

4.3. AMR of ESBL-Producing Enterobacteriaceae Strains

To enrich the assessment of the prevalence of cephalosporins resistance, an additional study was carried out among 32 Enterobacteriaceae strains screened and confirmed as ESBL producers and tested for AB susceptibility and genes encoding ESBL production. As might be expected, ESBL-producing Enterobacteriaceae isolates exhibited high resistance rates to beta-lactams—Amp (100%), CTX (100%), CAZ (94%), and CPM (50%). The ESBL phenotype was accompanied by a high resistance rate to other classes of ABs: TE—up to 60%; fluoroquinolones—CIP (up to 38%) or MOX (up to 60%); and COT—up to 45%. The resistance rate to non-beta-lactams was much higher than that of the total Enterobacteriaceae isolates, which was associated with the high level of MAR among the ESBL producers (59% of UWW isolates and 70% of TWW ones). Similar high MAR values of ESBL isolates have been reported [8,18]. Furthermore, it is important to note that an increase in resistance to CIP and MOX among ESBL isolates from TWW was registered. Moreover, a study based on in-depth statistical analysis of the probability of detecting ARB or ARGs at the simultaneous presence of AB residues indicated CIP in waste water as a good indicator of the presence of ESBL-producing Enterobacteriaceae [49].

Among ESBL-producing Enterobacteriaceae, six different genotypes encoded by the bla-genes TEM, SHV, or CTX-M, alone or in association, were determined. The blaTEM gene occurred more often alone (22.7%) than the blaCTX-M gene (13.6%), and the most common ESBL genotype was associated with co-carriage of blaTEM and blaCTX-M genes (found in 50% of isolates). A similar trend for high prevalence of the blaCTX-M gene has been reported for several other WWTPs [27,28,50], while in other studies, blaTEM has been detected as the most prevalent gene encoding the ESBL phenotype [18,47]. The blaSHV gene occurred rarely only among TWW isolates. Only one TWW isolate harboured all bla-genes encoding ESBL production. However, most importantly, multidrug-resistant ESBL-producing E. coli strains are present in the WWTP effluents, which may contribute to their spread in the receiving water and pose an increasing health risk in water intended use.

4.4. Prevalence of AMR Among Enterococcus spp. Isolates from Waste Water

Enterococci are ubiquitous Gram-positive bacteria inhabiting various ecological niches—humans, animals, food, water, and soil. As commensal bacteria in the gastrointestinal tracts of mammals, birds, fishes, and invertebrates, they enter waste water, environmental water, and soil through faeces [51,52]. They can cause urinary tract infections and are associated with nosocomial pathogenicity worldwide, particularly in immune-compromised peoples. Therefore, Enterococcus spp., along with other important pathogens, are the focus of AMR research in the One Health humans–animals–environment continuum [53,54].

In our study, all Enterococcus spp. isolates were analysed to determine their AMR phenotype and some of them were analysed for ARGs encoding resistance. The predominant species found in the raw and treated WW tested were E. faecalis and E. faecium. Among them, the most common resistance was to erythromycin (up to 39%) and tetracycline (up to 46%), followed by imipenem (up to 15%). The least common was resistance to chloramphenicol (up to 4%) and co-trimoxazole (up to 3%). The high levels of E and TE resistance found are consistent with the findings for a Portuguese WWTP [8], but a lower resistance to CIP was detected in our study (8% vs. 33%). Prevalence of resistance mostly to E, TE, and CIP has been reported for several WWTPs, with E. faecium exhibiting the highest AMR rate [55]. A large-scale study on AMR of Enterococcus spp. in the scope of a One Health continuum comprising 8430 isolates from sectors such as humans, agriculture, food, natural streams, and waste water found prevalent phenotypic resistance to tetracyclines and macrolides encoded by tetM and ermB, respectively [53,54]. It has been also clarified that urban waste water and human clinical isolates have a greater diversity of resistance phenotypes and exhibit higher levels of MAR compared to other sectors.

The prevalence of tetracycline and erythromycin resistance in the studied WWTP confirms the data reported for many other WWTPs [8,53,55], but the lower level of CIP resistance and higher beta-lactam resistance identified in our study are probably related to differences in the type and amount of ABs used by local communities. In the present study, all enterococcal isolates were susceptible to newer generations of ABs, such as glycopeptides, glycylcycline, and oxazilidinones. The susceptibility to vancomycin and teicoplanin of the tested strains is in contrast with the data for occurrence of vancomycin-resistant enterococci (VRE) in the effluents of some other WWTPs [9,56,57]. VRE strains of E. faecium have been isolated from a Czech WWTP, and all but one were of the vanA phenotype [56]. Some VRE E. faecium isolates from WWTPs located in the North Spain–South France region were resistant to vancomycin and teicoplanin, expressing the vanB phenotype [9]. VanA-harbouring E. faecium, as well as vanC1/vanC2-containing enterococcal strains, have been detected in a Portuguese WWTP [57].

A comparison between the AMR phenotypes of waste water isolates obtained in our study and clinical isolates from Bulgarian hospitals [58] found various resistance patterns. Among WW isolates, the E-TE-HLG phenotype predominated, found in 50% of enterococcal strains from UWW and in 14.2% of TWW isolates, while the Amp-CIP-HLG phenotype was the most common (79%) among the clinical isolates. Similarly, the incidence of MAR isolates in UWW and TWW of the studied WWTP amounted to 9.0% and 11.2%, respectively, while a higher MAR (39%) was reported for the clinical isolates [58].

The AMR rate of Enterococcus spp. in waste water was higher than that found for food and environmental enterococci. A variable but low level of AMR was found among enterococcal strains isolated from Bulgarian food and environmental habitats [52]. Among them, the ermB gene also was most frequently found but at a lower level (15.2%). In contrast with our data, the prevalence of the aac(6′)-aph(2″) gene was lower (1.4% vs. 57%), and the vanA gene was detected as well [52].

According to our data on gentamicin resistance, 58% of the tested enterococcal strains contained the aac(6′)/aph(2″) gene, but only 3% exhibited a high-level gentamicin-resistant phenotype. Such a discrepancy was also established for clinical isolates, 98% of whom harboured the aac(6′)/aph(2″) gene but just 58% of whom exhibited resistance to HLG by the disc diffusion method [58]. However, the resistance rate to HLG was reduced in the studied WWTP, and the reduction effect was statistically significant.

In our study, the efflux pump gene emeA was detected in all ciprofloxacin-resistant enterococcal strains, suggesting that the drug efflux is the main mechanism involved in resistance to fluoroquinolones. Despite the reduced incidence of ermB and emeA genes and an increased prevalence of the rest of the tested ARGs, ermB and tetM remained the most common ARGs among the Enterococcus spp. in TWW. Therefore, all data obtained reveal that the WW treatment did not completely limit the incidence of resistant Enterococcus spp. in the WWTP effluents and could contribute to AMR spread in water environments.

The results of present study demonstrate a limited efficiency of the mechanical-biological treatment processes in the monitored WWTP with regard to HPC bacteria, Enterobacteriaceae, and Enterococcus spp. resistant to the studied antibiotics. The various effectiveness of removal of ARB and ARGs in WWTPs depending on the type of pollutants and operational characteristics of the treatment facilities [1,24] indicates a need for information on each different pollutant and particular waste water treatment facility. Usage of advanced WW treatment methods, including WWTP effluents disinfection with chlorine, ozone, or UV, has been explored as an approach to reduce AMR emissions into the environment [11,21,59,60]. Advanced WW treatment is a tool for removal of not only the target pollutants but also a range of additional ones, in this way providing collateral benefits. Therefore, when choosing measures to limit transmission of ARB in the environment, it is important to assess the risk level and the amount of costs to achieve certain benefits. Our study is a first step in assessing a regional situation of AMR spread, which reveals the need to assess AMR not only in urban WWTPs but also in other critical points of selection and transmission of AMR, such as urban sewage and hospital wastewater facilities. Considering the limitations of our study (a single WWTP studied, analysis of replicate samples, not pooled daily samples) we will enrich our research by directing it to other WWTPs, which differ in degree of treatment and type of treatment facilities, and by improving the sampling method as well. Continuing our investigations in this line will allow us to clarify the risks of selection of ARB in certain compartments of the urban water cycle and to prioritize measures to limit the spread of AMR in the receiving waters.

5. Conclusions

To the best of our knowledge, the present study is the first assessing the prevalence of AMR among HPC bacteria and faecal indicators Enterobacteriaceae and Enterococcus spp. in a Bulgarian WWTP. To achieve a more complete clarification, the AMR prevalence in the waste water of the selected WWTP was assessed by three culture-dependent parameters and conventional PCR for ARGs detection.

The data on the resistant HPC bacteria and faecal indicators provide an opportunity to assess the AMR prevalence among the regional community, as far as the AMR rate of urban sewage mirrors the contribution of local population as a result of ABs usage. The higher AMR prevalence found among faecal indicators Enterobacteriaceae and Enterococcus spp. in waste water than among HPC bacteria confirms the notion that the AMR rate of faecal bacteria in urban sewage is indicative of the level of AMR prevalence in the local population. High levels of resistance to Amp and TE among Enterobacteriaceae strains isolated from untreated sewage, or to E, TE, and beta-lactams among Enterococcus spp. strains, reveal the AMR rate in the local community. Also of concern is the finding of clinically relevant antibiotic resistance among E. coli, associated with ESBL production and carriage of blaTEM and blaCTX-M genes, as well as co-resistance to CIP, COT, and C.

Comparing between the AMR levels of raw and treated waste water based on the tested microbiological parameters reveals the insufficient efficiency of the performed treatment in the studied WWTP for complete removal of the resistant heterotrophic or faecal indicator bacteria and demonstrates a positive selective effect for increasing resistance to some ABs, most often to CIP. Waste water treatment resulted in the greatest increase in resistance to CIP and TE among the HPC bacteria, to CIP, COT, and C among Enterobacteriaceae, and to E and TE among Enterococcus spp.

Elucidating the AMR levels of the wastewater microbiome and the risk of ARB and ARG selection during urban waste water treatment may be a warning of the need for measures to limit the transmission of AMR in the environment.