Transition of antimicrobial resistome in wastewater treatment plants: impact of process configuration, geographical location and season

doi:10.21203/rs.3.rs-2011499/v1

Download PDF

Article

Transition of antimicrobial resistome in wastewater treatment plants: impact of process configuration, geographical location and season

https://doi.org/10.21203/rs.3.rs-2011499/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 15 Jun, 2023

Read the published version in npj Clean Water →

You are reading this latest preprint version

Wastewater is the major source of the emergence of antimicrobial resistance (AMR) in water environment. Wastewater treatment plants (WWTPs) are the important barriers for preventing the spread of AMR in wastewater into water environment, as well as the reservoir of AMR, which can be potentially discharged into treatment effluent. In this study, the antimicrobial resistome in WWTP was investigated using systematic sampling and shotgun metagenomic analysis over a variety of geographical locations, seasons, and biological treatment configurations. The results revealed that the transition of antimicrobial resistome occurred at two locations during the course of wastewater treatment process to develop the distinctive antimicrobial resistome in influent wastewater, activated sludge, and treatment effluent regardless of the geographical locations of WWTPs. The antimicrobial resistome in influent wastewater was characterized by higher abundance of antibiotic resistance genes (ARGs) resistant to clinically important drug classes, whereas sludge retained a higher abundance of multidrug ARGs associated with efflux pump. Seasonality was the primary factor to characterize the antimicrobial resistome in influent wastewater, which was partially succeeded to the subsequent resistome of activated sludge and treatment effluent. Importantly, antimicrobial resistome in the treatment effluent was dependent on process configuration of sludge separation. With conventional final sedimentation, antimicrobial resistome in the treatment effluent was partially affected by the resistome in influent wastewater, suggesting some ARGs in influent wastewater bypassed biological treatment and final sedimentation to be retained in the treatment effluent. On the contrary, the resistome of MBR effluent was independent from wastewater resistome, suggesting good reduction of ARG to clinically important drugs originated from influent wastewater.

antibiotic resistance gene (ARG)

shotgun metagenomic sequencing

conventional activated sludge process (CAS)

membrane bioreactor (MBR)

seasonal change

microbial community

Antimicrobial resistance (AMR) is a major threat to human health. A recent study reported that AMR is now the leading cause of death in the world, and it was associated with 1.3 million deaths in 2019 ¹. According to O’ Neil (2014) ², the health burden caused by AMR would continuously increase and result in 10 million annual deaths in 2050. To prevent the spread of AMR, there is an urgent need to develop a “One Health approach,” which is a collaborative effort of multiple disciplines to attain optimal health for people, animals, and the environment ³. Among the environmental pathways of AMR spread, wastewater has been regarded as the primary source of the emergence of AMR into water environment. Particularly, the discharge of untreated wastewater has been reported to result in significant increase in the prevalence of AMR in water environment in various countries ^4–9. Accordingly, wastewater treatment plants (WWTPs) have emerged as an important barrier to prevent the spread of AMR in wastewater into water environment. However, WWTPs can act also as a reservoir for AMR, which is maintained in activated sludge and potentially discharged into treatment effluent ^10–13. Particularly, studies have revealed that WWTP effluent still contains a certain abundance of antimicrobial resistant bacteria (ARB) and antimicrobial resistance genes (ARGs) ^14–17. Hence, treatment effluent from WWTPs can be another major source of AMR in water bodies and drinking water resources receiving the treatment effluent ^18–21.

Although both wastewater and treatment effluent are the major anthropogenic origin of AMR in water environment, wastewater and treatment effluent may exhibit different AMR traits. The major source of AMR in wastewater is gut microbiome, wherein the AMR is enriched via exposure to antimicrobial dose from people in the sewershed community. In contrast, AMR in treatment effluent mainly originates from activated sludge, which is exposed to a variety of stress that induce the co-selection and enhancement of AMR (e.g. heavy metals, reactive oxygen species (ROS), disinfectants, and sub-lethal concentrations of antimicrobials) ^22–24. Hence, the discharge of AMR via wastewater or treatment effluent poses qualitatively different impact on water environment. Several studies have demonstrated the difference in the antimicrobial resistome of influent wastewater and treatment effluent using shotgun metagenomic approach ^25–27. These studies have provided basic understanding as case studies by targeting a single WWTP; however, the factors that generally govern the fate of antimicrobial resistome among various WWTPs with different biological process configuration, and geographic location are yet to be comprehensively understood. Moreover, these studies analyzed the shotgun metagenomic sequence data based on raw reads of the observed ARGs, which are highly biased compared to the original proportion of each ARG, because a longer gene appears to have greater numbers of gene fragments in shotgun metagenomic sequencing. For example, mexB, which exhibit multiple resistance, and tetQ, which exhibit tetracycline resistance, are 3141 and 1974 bp long, respectively, whereas mphB, which exhibit macrolide resistance, and qnrS, which exhibit fluoroquinolone resistance, are 477 and 657 bp long, respectively. This could result in 5 to 10 times overestimation in the proportion of long ARGs in raw reads-based quantification. Therefore, the appropriate normalization of raw reads is essential to acquire accurate snapshots of antimicrobial resistome and to understand the mechanism and the factors involved in the transition of antimicrobial resistome from influent wastewater to treatment effluent in WWTPs.

The objective of this study was to reveal key factors that govern the transition of antimicrobial resistome during wastewater treatment processes. To this end, antimicrobial resistome in influent wastewater, activated sludge, and treatment effluent was investigated using systematic sampling and shotgun metagenomic analysis over a variety of geographical locations, seasons, and biological treatment configurations. In this study, all the ARG data was curated after normalization as reads per kilobase (RPK) to obtain the accurate snapshot of the antimicrobial resistome, which reflects the original proportion of the gene copy number in a sample. In addition, the common characteristics and differences of the antimicrobial resistome (i.e. ARG abundance and composition) in influent wastewater, activated sludge, and treatment effluent, were identified by the introduction of the appropriate normalization. Moreover, according to the differences among the resistome and their association with microbial community, this study revealed the impact of key factors that determine antimicrobial resistome and its transition during wastewater treatment processes.

2.1. ARG abundance in wastewater treatment process

Influent wastewater, activated sludge, and treatment effluent were collected at five municipal WWTPs with a variety of biological treatment process configuration (Fig. 1). Total ARG abundance per 16S bacterial population in influent wastewater significantly decreased in activated sludge, regardless of process configuration and geographical locations of WWTPs. The total ARG abundance in influent wastewater from all the WWTPs was 32–50% (Fig. 2a). There was no remarkable difference of total ARG abundance in influent wastewater among the WWTPs located in different municipality. The total ARG abundance in the activated sludge collected from all the WWTPs decreased substantially to 5–19%. This decrease in the ARG abundance from influent wastewater to activated sludge was observed in all samples regardless of the seasons or process configurations. There was a slight increase or no remarkable change in the total ARG abundance from sludge to the final effluent collected from all WWTPs, although there was a significant reduction in the bacterial population. The ARG abundances observed in the present study are consistent with the typical range of total ARG abundance reported in past literature based on metagenomic approach, which was reported as 20–40% ^28,29. Similarly, the decrease in the ARG abundance from influent wastewater to activated sludge has been reported in previous studies. The metagenome-based total ARG abundance in activated sludge was reported to decrease by 12–13% from influent wastewater to 8–9% ²⁸. Further, studies based on high-throughput qPCR reported a 10–20% decrease in the abundance of major ARGs from influent to sludge ^15,17. However, several studies on culture-based AMR assays have reported no remarkable change or slight increase in the AMR abundance in sludge and final effluent from those in influent wastewater ^14,30−33. This is probably because the culture-based assays target only specific bacteria groups of fecal origin (e.g. coliform, E. coli, Salmonella), whereas metagenomic approaches cover the entire bacterial community. The observation of the total ARG abundance at each treatment stage revealed that the ARG abundance decreased during biological treatment (Fig. 2b). However, there was no change in the total ARG abundance from influent wastewater after primary sedimentation to the equalization tank. This indicates that primary sedimentation and retention in equalization tank had no significant effect on the ARG abundance in influent wastewater. The total ARG abundance decreased to 15–21% in the CAS process and 10–12% in the MBR process when the primary-treated wastewater entered biological treatment.

The total ARG abundances in the treatment effluent were similar to those in the activated sludge, and significantly lower than those in influent wastewater (Fig. 2). Although the 16S bacterial population in the treatment effluent reduced by 1.0–1.7 log compared to that in sludge, there was no significant change in the total ARG abundance. These trends in the ARG abundance were consistent to those reported in past studies ^17,28. In addition, there was no significant change in the total ARG abundance before and after chlorination (Fig. 2b). Previous studies have reported that the log reduction of ARGs in secondary treatment effluent is limited under the typical chlorination intensity of 15 mg×min/L ^34–37, although the population of ARB is substantially reduced. Similarly, in the present study, chlorination reduced the total bacteria population; however, the impacts of chlorination on the reduction of ARG abundance was limited.

In accordance with total ARG abundance, composition of ARG also exhibited a remarkable change from influent wastewater to activated sludge. Only 45–50% of ARGs in influent wastewater remained in activated sludge, then 84–91% of the remained ARGs in the sludge were carried over to treatment effluent (Fig. 3). The proportion of multiple antimicrobial resistance (MAR) genes in the sludge increased, regardless of the process configuration and location of WWTPs (Fig. 4). Particularly, the proportion of ARGs with broad resistance to more than six drug classes increased notably in the activated sludge compared to the influent wastewater. This remarkable change in ARG abundance and composition from wastewater to sludge indicates that activated sludge maintained its unique antimicrobial resistome. In biological treatment, a part of activated sludge was returned back to the beginning of the biological treatment after thickening in the final sedimentation tank. The returned thickened sludge (a.k.a. return sludge) mixed with primary treated wastewater to constitute the activated sludge. Under the typical operating conditions of WWTPs, the flow rate of the return sludge was configurated as 1/2 to 3/4 of the flow rate of primary treated wastewater ³⁸. Meanwhile, the suspended solids (SS) concentration of the return sludge is typically thickened to 6,000–12,000 mg/L, which is 3–5 times larger than the activated sludge in aeration tank. Hence, most of the biosolids in the activated sludge originated from the return sludge. In WWTP B, the flow rate of the return sludge was 25–40% of that of the primary treatment effluent, whereas the SS in the return sludge was 3.4 to 4.4 times higher than that in activated sludge. Thus, antimicrobial resistome in activated sludge was mainly comprised of those in return sludge rather than those in influent wastewater. Consequently, impacts of influent wastewater on the antimicrobial resistome in the activated sludge was relatively small.

Seasonal difference was also observed in both ARG abundance and composition. The increase in the ARGs with broad resistance (< 6) was more notable in summer (+ 0.9–6.5%) than in winter (+ 1.5–11%). In contrast, the proportion of ARGs resistant to two drug classes was remarkably higher in winter. These results indicated the seasonality of ARG compositions in influent wastewater and activated sludge. In addition, the ARG proportion resistant to each drug class differed between influent wastewater and activated sludge (Fig. 5). Particularly, compared to the sludge samples, the influent wastewater samples in all WWTPs were mostly composed of ARGs resistant to tetracycline, fluoroquinolone, and macrolide. Moreover, seasonality was also observed in the proportion of ARGs resistant to some drug classes. For example, the proportion of ARGs resistant to macrolide and streptogramin was larger in winter, whereas those resistant to aminoglycoside exhibited a large proportion regardless of seasons. The comparison of the total ARG abundance and AMR compositions revealed that the following characteristics in the antimicrobial resistome in the WWTPs: (i) total ARG abundance in influent wastewater was approximately 30–50% regardless of seasons and geographical locations of WWTPs; (ii) total ARG abundance significantly reduced in the sludge regardless of the process configurations; (iii) ARG composition and proportion of MAR changed from influent wastewater to sludge and also by season; (iv) chlorination significantly reduced bacterial population, but had no particular effect on the abundance of AMR.

2.2. Transition of ARG composition during wastewater treatment process

The ARG composition at five WWTPs were compared using PCA (Fig. 6a). The primary factor distinguishing ARG compositions was treatment stage (i.e. influent wastewater, activated sludge, and treatment effluent), and seasonality was the secondary factor. In contrast, geographic locations and process configurations of WWTPs had no significant effect on the ARG compositions in influent wastewater and activated sludge. The influent wastewater was mostly characterized by negative PC1 scores except WWTP E in summer, where influent wastewater may contain some sludge due to sampling restrictions (as described in Methods). The negative PC1 values indicated the influence by AMR to critically important antimicrobials, including quinolones (qnrS, qnrVC), macrolides (ermG, mel), aminoglycosides (APH(3’), cmlA, aad(6), AAC(6’),), colistins (ICR), cephalosporins, and carbapenem (ESBL genes of MOX, OXA, CfxA, VEB, FOX, GES, CMY) (Fig. 6a; Table S4a). In contrast, the activated sludge samples were characterized by MAR genes, which exhibited highly positive PC1 scores (Fig. 6a). Among the top 30 ARGs with highly positive PC1 loadings, 19 ARGs were efflux pumps associated with broad resistance to five or more antibiotic classes (mexD, medQ, mexW, mexI, mexK, mexN, MuxC, MucB, mexF) (Table S4b). Typically, efflux pump functions as resistance to not only antimicrobials but also a broad range of cell stresses. During biological treatment, sludge bacteria are exposed to various kinds of cell stresses (e.g., antimicrobials, heavy metals, toxic trace chemicals, and oxidative stress by aeration). In an aeration tank, these cell stresses may result in the enhancement or co-selection of antimicrobial resistance ^{23,24, 39–42}. Interestingly, the treatment effluents were plotted at an intermediate range between influent wastewater and sludge groups (Fig. 6a). Treatment effluent is expected to exhibit a similar ARG composition as sludge if each ARG in sludge was equally reduced in the final sedimentation. However, this PCA result demonstrated that the ARG compositions in treatment effluent was not solely affected by activated sludge but also by influent wastewater. This is consistent with the result that the ARGs originated from influent wastewater partly remained in the final effluent of the WWTP (Fig. 3).

The secondary factor affecting ARG compositions was season, which was represented by PC2 (Fig. 6a). Samples collected in winter exhibited positive PC2 scores, which were characterized by ARGs of antibiotic inactivation associated with resistance to β-lactams, such as cephalosporins, carbapenem and ESBL-producing genes of OXA, MOX, GES, SHV, CMY, ADC (Table S5a). In contrast, samples collected in summer mostly exhibited negative PC2 scores, which were mainly characterized by ARGs of efflux pumps associated with MAR. The seasonal difference in ARG composition was larger in influent wastewater than in activated sludge. This may be attributed to the seasonality in the clinical use of antimicrobials in Japan (as discussed below). Treatment effluents in winter and summer were also distinguished by PC2. The PC2 scores of the treatment effluents in winter and in summer were at an intermediate range between those of influent wastewater and activated sludge in winter and summer, respectively. This supports the hypothesis that the treatment effluent was affected not only by activated sludge but also by influent wastewater.

The shift of ARG composition occurred at two locations during the course of the treatment process, according to the PCA scores in each treatment stage of WWTP B (Fig. 6b). There was no notable change in the ARG compositions in the influent wastewater (WW) even after primary sedimentation until equalization tank. However, the PC1 scores turned positive when the primary treated wastewater was subjected to biological treatment. All the sludges in the biological treatment of both CAS and MBR processes exhibited similar ARG composition associated with highly positive PC1 scores. These results indicated that ARG composition remarkably changed at the beginning of biological treatment. This shift in the ARG composition is consistent with the shift in the total ARG abundance and MAR proportion during biological treatment (Figs. 2 and 4). During the CAS process, the second shift in the ARG composition was observed in the secondary treatment effluent after final sedimentation (STE(CAS)). CAS effluent was plotted at an intermediate range between those of the wastewater and sludge groups, suggesting that the ARG composition of the CAS effluents was not only affected by sludge but also by wastewater. This indicated that ARGs of wastewater origin was partly retained in the CAS effluent. In MBR, however, there was no significant change from sludge to MBR effluent. MBR effluents exhibited positive PC1 scores, indicating that their ARG compositions remained similar to those of sludge. This demonstrated that the second shift in the ARG compositions occurred in CAS at final sedimentation, but not in MBR, in which the sludge was eliminated by membrane filtration. These differences between CAS and MBR effluents demonstrated that the ARG composition in the treatment effluent was affected by process configuration, particularly at the sludge separation stage. Moreover, the ARGs of wastewater origin were probably significantly reduced by membrane filtration in MBR, whereas its reduction was limited in final sedimentation in CAS. The second shift in the ARG composition in CAS probably occurred via the overflow of ARGs of wastewater origin at the final sedimentation.

2.3. Transition of microbial community and association with ARG composition

Some phylogenetic classes was found to be associated with specific ARG groups during wastewater treatment. Relative abundance of Bacteroidia and Flavobacteriia classes in Bacteridetes, Bacilli, Clostridia and Erysipelotrichi classes in Firmicutes, Fusobacteriia, Eplilonproteobacteria, Gammaproteobacteria, and Synergistia classes (hereinafter referred as the microbial group ‘A’) had relatively high correlations with some ARGs to macrolide (ErmA, ErmB, ErmF, ErmG, mefC, mefE, and mel), to quinolone (QnrD, QnrS), to tetracycline (tet3, tetM, tetO, tetQ, tetS, tetW) and to vancomycin (vanB, vanW, vanX, vanY), whereas the group ‘A’ had relatively low correlations with most ARGs to sulfonamide (sul1, sul2, sul4) and with multidrug ARGs of efflux pump (most of the mex and Mux genes) (Fig. 7). The microbial group ‘A’ mostly consisted of phylogenetic classes for commensal bacteria and anaerobic bacteria which are often abundant in gut microbiome. These bacteria were abundant in influent wastewater and possibly harbored ARGs to macrolide, quinolone, tetracycline and vancomycin in influent wastewater. Interestingly, some phylogenetic classes (hereinafter referred as the microbial group ‘C’) had the opposite trend of correlations to the microbial group A. The microbial group ‘C’ mostly consisted of phylogenetic classes which are often associated with water and soil environment, i.e. Alphaproteobacteria, Deltaproteobacteria, Cytophagia, Nitrospira and classes in Acidobacteria, Actinobacteria, Chloroflexi, Planctomycetes, Verrumicrobia phyla, etc., which were abundant in sludge samples (Table S6b). The group ‘C’ had relatively high correlations with ARGs to sulfonamide (sul1, sul2, sul4) and to multiple antimicrobials by efflux (most of the mex and Mux genes) (Fig. 7). Hence, these microbes in activated sludge possibly harbored and reserved the ARGs of efflux pump with broad resistance in activated sludge.

As observed in ARG composition, the microbial community was also differentiated among the influent wastewater, activated sludge, and treatment effluent samples (Fig. 6c). Differently from the ARG composition, however, seasonal change did not exhibit a significant effect on the microbial community. In contrast, process configuration exhibited a more significant effect on the microbial community in the sludge samples. The influent wastewater samples exhibited negative PC1 scores, while the activated sludge samples exhibited highly positive PC1 scores. The microbial community of influent wastewater was characterized by microbes with highly negative PC1 loadings, which were mainly composed of anaerobic enteric bacteria, such as Bacteroidetes, Firmicutes, and Gammaproteobacteria, including Enterobacteriaceae, Aeromonadaceae (Table S6a). In contrast, the sludge samples were characterized by microbes with highly positive PC1 loadings, which largely consisted of aerobic bacteria including, Deltaproteobacteria and Planctomycetia (Table S6b). Similarly, the PC1 scores of treatment effluent were plotted at an intermediate range between those of influent wastewater and activated sludge, as observed in ARG composition. This suggested that the microbial community in the treatment effluent was not only affected by activated sludge but also by influent wastewater. Hence, some microbes in influent wastewater were not reduced, but were retained in the final treatment effluent. If these microbes harbored ARGs and remained in treatment effluent, ARG composition in treatment effluent would be partly dependent on influent wastewater. In the present study, most of the microbes characterizing influent wastewater belonged to the microbial group ‘A’, which had high correlations with ARGs to macrolide, quinolone, tetracycline and vancomycin was found abundantly in influent wastewater (Fig. 7). These bacteria were possible carriers of ARGs to clinically important drugs remaining in treatment effluent.

Transition of the microbial community also occurred at two locations during the treatment process, as observed in ARG compositions (Fig. 6d). The first shift occurred when primary treated wastewater entered biological treatment, wherein the PC1 scores turned from negative to positive. The sludge group was divided into CAS and MBR groups, which were characterized by negative and positive PC2 scores, respectively. The second shift in the microbial community occurred at sludge separation stage after biological treatment. Treatment effluents exhibited the intermediate PC1 scores of those of the wastewater and sludge groups. Therefore, the shift in the microbial community occurred at the same timing as that of the ARG composition. Importantly, the second shift in the microbial community after sludge separation stage was observed both in CAS and MBR. This indicates that process configuration, particularly sludge separation, exhibited no significant effect on the microbial community of the treatment effluent, but exhibited a notable effect on the ARG composition.

In summary, microbial community and antimicrobial resistome had the following relations during wastewater treatment: (i) anaerobic enteric bacteria abundant in influent wastewater was associated with ARGs to macrolide, fluoroquinolone and tetracycline; (ii) microbial community drastically changed at biological treatment, where microbes were associated with multidrug ARG became more abundant; (iii) microbial community in CAS effluent and MBR effluent did not have remarkable difference, while antimicrobial resistome were distinctively different.

2.4. Determinative factors of antimicrobial resistome in wastewater treatment

The transition of antimicrobial resistome in wastewater treatment process exhibited a common tendency regardless of geographical locations of WWTPs. In summary, the transition in the ARG composition occurred at two locations during the course of wastewater treatment process. Consequently, distinctive ARG compositions were developed in influent wastewater, activated sludge, and treatment effluent. In the present study, seasonality was the primary factor to characterize the “wastewater resistome” (i.e., antimicrobial resistome in influent wastewater and primary treated wastewater). The wastewater resistome is likely to reflect the clinical use of antimicrobials in the sewershed community. Thus, the wastewater resistome was affected by seasonality of the antimicrobial use. The wastewater resistome was mainly characterized by ARGs resistant to aminoglycoside, cephalosporins, macrolides, quinolones and tetracyclines (Fig. 5; Table S4a), which are commonly-used antimicrobials in clinics in Japan ⁴³ (Fig. 8). In Japan, the prescription of 3rd -generation cephalosporins, macrolides and other antimicrobials has seasonal increase in winter (Dec – Feb) ^43–45. Accordingly, the increased abundance of ARGs to cephalosporins and macrolides was observed in wastewater resistome (Fig. 5; Table S5a). This seasonality of the wastewater resistome was likely to partially succeeded to the subsequent resistome of activated sludge and treatment effluent, in accordance with its dependence on the wastewater resistome.

The “sludge resistome” was less diverse and had no apparent determinative factors besides the seasonality succeeded from influent wastewater. The sludge resistome was not apparently affected by process configuration, operating parameter nor geographical location of the WWTPs, but probably developed its unique resistome while sludge was retained in the biological treatment system by circulation of return sludge. In the biological treatment, microbes which harbor resistance genes to broad virulence could be gradually enriched by co-selection through exposure to various virulent substances while circulated and retained in the biological treatment. Consequently, all the observed resistome in activated sludge was characterized by a higher abundance of ARGs associated with broad AMR by efflux pump (Table S4b). High abundance of multidrug ARGs in sludge were also reported in past studies ^26,27. Ng et al. (2019) ²⁶ reported that sludge in MBR exhibited a higher abundance of multidrug ARGs of the mex gene family, which exhibits broad resistance to most drug classes. The increased abundance of the mex gene family in “sludge resistome” was also observed in the present study. Hence, activated sludge should be noted as the important reservoir and potential source of ARGs with multiple resistance, which is possibly discharged from WWTP effluent into water environment.

The primary factor to determine the “effluent resistome” was process configuration of sludge separation stage. The second shift in the ARG composition from “sludge resistome” to “effluent resistome” occurred during the final sedimentation. However, this second shift was not observed in MBR, suggesting that the configuration of sludge separation is the key factor to characterize the “effluent resistome”. Understanding resistome in treatment effluent is essential because a WWTP is an important barrier to prevent spread of AMR into environment. Past studies reported that treatment effluent had different antimicrobial resistome from influent wastewater ^47,48. However, the present study revealed that resistome of treatment effluent is not fully independent from that of influent wastewater. On the contrary, almost 90% of ARGs in treatment effluent was common to influent wastewater (Figs. 3). These facts demonstrated that a certain portion of ARGs in influent wastewater bypassed biological treatment and final sedimentation to be retained in the treatment effluent. Raza et al. (2022) ⁴⁹ reported that some ARGs which were commonly present in influent wastewater and treatment effluent were associated with opportunistic pathogenic bacteria. In the present study, phylogenetic classes including opportunistic pathogenic bacteria were associated with ARGs of wastewater origin. These findings also support that ARGs of wastewater origin possibly passed through the wastewater treatment and remain in effluent. In contrast, the MBR effluent had similar resistome to sludge, which is definitely different from wastewater resistome. This suggests that sludge separation is the key step to exclude ARGs of wastewater origin from treatment effluent. Membrane filtration in MBR has good reduction of microbes and extracellular ARGs in the supernatant, whereas the final sedimentation tank is mechanistically unable to eliminate microbes and extracellular ARGs which are present in the supernatant. Therefore, the independence of resistome in MBR effluent from wastewater resistome implies that ARGs of wastewater origin were abundant in the supernatant of activated sludge. In fact, extracellular ARGs are reportedly abundant in treatment effluent of CAS ⁵⁰. Elimination ARGs in the supernatant of activated sludge is important to reduce ARGs of wastewater origin in treatment effluent. The results of the present study suggested that application of MBR could be effective for reduction of ARGs of wastewater origin discharged via WWTP effluent. Actually, MBR reportedly exhibits a good reduction in extracellular genes even with MF membrane due to cake layer filtration ⁵¹. As observed in the present study, “effluent resistome” from conventional final sedimentation was not only composed of ARGs from sludge but also ARGs from influent wastewater. Hence, treatment effluent is the possible source of a large variety of ARGs, including both multidrug ARGs with broad AMR from activated sludge as well as ARGs to clinically important drugs from influent wastewater. Therefore, reduction of ARG in WWTP effluent is essential to reduce the variety of ARG discharged into water environment.

3.1. Sampling at wastewater treatment plants

Influent wastewater, activated sludge, and treatment effluent samples were collected at five municipal WWTPs located in different municipality in Japan. The target WWTPs were selected to exhibit varying biological treatment process configuration: conventional activated sludge (CAS) process at WWTP A, CAS and membrane bioreactor (MBR) processes treating identical wastewater in parallel at WWTP B, enhanced biological phosphorus removal (EBPR) process at WWTP C, anaerobic–anoxic–oxic (A2O) process at WWTP D, and oxidation–ditch (OD) process at WWTP E (Fig. 1). Treatment capacity and operating conditions of the target wastewater treatment processes are shown in Table S1. The influent wastewater samples were collected before primary sedimentation except WWTP E, where the influent wastewater was collected at the inflowing point of wastewater into an aeration tank because influent wastewater directly flowed into the aeration tank without primary sedimentation. The sludge samples were collected from an aeration tank (or oxic tank). The treatment effluent samples were collected after chlorination before discharge to the environment. Samples were collected at each WWTP in dry weather for three times both in winter (Dec 2018 – Feb 2019) and in summer (Jun – Aug 2019) in 2019, with at least two weeks interval. In February and August, samples were collected from WWTP B at each treatment stage (Fig. 1). All samples were transported to the laboratory on ice and processed within 24 h after sampling.

3.2. Sample processing

For the influent wastewater and sludge samples, first, 50 mL of the sample was centrifuged at 10,000 ×g for 15 min. Next, DNA was extracted from the centrifuged pellets using a FastDNA Spin Kit for Soil (MPBiomedicals, USA). For the treatment effluent samples, first, the samples were concentrated using membrane filtration. The sample volume of the treatment effluent was 200 mL except the MBR effluent collected at WWTP B, in which 2 L of the sample was filtrated. The solid particles in the treatment effluent samples were collected on 0.2-µm polycarbonate membrane filter (K020G047A, Advantec-Toyo, Japan) using vacuum filtration. Thereafter, DNA was extracted directly from the membrane filter using the FastDNA Spin Kit for Soil (MPBiomedicals, USA). The concentration of each DNA extract was measured using a spectrophotometer (µCuvette Biophotometer, Eppendorf, Hamburg, Germany). A seasonal composite extract of DNA was prepared to obtain the average snapshot of each season (winter or summer). To prepare the composite DNA extract, the three samples collected in the same season (winter or summer) for each sample type (wastewater, sludge, or treatment effluent) at each WWTP were mixed together to achieve an equal mass (ng) of DNA. A total of 43 DNA extracts of the samples were prepared for the subsequent analyses, as listed in Table S2.

3.3. Microbial population and community analysis based on 16S rRNA gene

The microbial community of each sample was analyzed by targeting the V3-V4 regions of 16S rRNA gene. To this end, a sequence library was prepared by 2-step tailed polymerase chain reaction (PCR) using ExTaq Hot Start (Takara-bio, Japan). The primer sequences and PCR amplification conditions are shown in Table S3. The prepared library was purified using a Wizard SV Gel and PCR Clean-UP System (Promega, Madison, USA). After the validation of the library quality using a Fragment Analyzer with a dsDNA 915 Reagent Kit (Agilent Technologies, USA), the paired-end sequence of 2 × 300 bp was acquired using a MiSeq System (Illumina, USA) with MiSeq Reagent Kit v3 (Illumina, USA). Raw sequencing data were trimmed and filtered using Fastx toolkit version 0.0.14 and sickle version 1.33. To obtain the paired-end library, trimmed reads were merged using FLASH version 1.2.11. Quantitative Insights into Microbial Ecology 2 (QIIME2) version 2020.2 was employed for taxonomy analysis. Chimeric sequences were removed using dada2 plugin prior to classifying into operational taxonomic units (OTUs) under reference library of Greengene (ver. 13_8) with criteria of 97% identity. The sequence data were deposited in DRA database of DNA Data Bank of Japan (DDBJ) as accession numbers of SAMD00281909–00281924, SAMD00281876–00281891, SAMD00206914–00206922, SAMD00233324–00233332, SAMD00282801 and SAMD00282798 (Table S2).

Microbial population in each sample was quantified based on the 16S rRNA gene using real-time PCR assay with a universal primer set of 341F (5′-CCTACGGGNGGCWGCAG-3′) and 805R (5′-GACTACHVGGGTATCTAATCC-3′) ⁵². For the PCR assay, a PCR reaction mixture (20 µL) of each sample was prepared using Brilliant III Ultra-Fast SYBR Green QPCR Master Mix Kit (Agilent Technologies, USA), after which amplification was detected using a real-time PCR system (Mx3000P, Agilent Technologies, USA). PCR conditions were: denaturation at 95°C for 3 min followed by 40 cycles of annealing at 55°C for 45 s and extension at 72°C for 1 min.

3.4. Shotgun metagenomic sequence analysis

Shotgun metagenomic sequences of the sample DNA extracts were analyzed using HiSeq X system. The DNA in a sample was fragmented to 500 bp using Covaris S220 (Covaris Inc., Massachusetts, USA). The sequence library was prepared using KAPA HyperPlus Kit with FastGene Adapter Kit (FastGene, Japan). After validation of the library quality using a Fragment Analyzer with dsDNA 915 Reagent Kit (Agilent Technologies, USA), the paired-end sequence of 2 × 151bp was acquired using HiSeq X system (Illumina, USA). The raw paired-end sequences were filtered and trimmed using Enveomics collection pipeline ⁵³. After quality trimming, the ARGs for each fragment sequence were identified using BLASTn against Comprehensive Antibiotic Resistance Database (CARD) ver. 3.0.7 ⁵⁴ with an E-value cutoff of 1 × 10^− 5. According to the BLAST output, the read count of each identified ARG was aggregated to obtain the “raw ARG profile,” which was composed of the list of ARGs and their read count present in each sample. The read count of a gene fragment tends to become larger when the original gene length is longer. To exclude the bias caused by the gene length, the read count of each ARG were normalized as reads per kilobase (RPK), according to the subject sequence length (slen) in the CARD database. The read count of the 16S rRNA gene in each sample was counted using Parallel-META 3 ⁵⁵. The read count of 16S rRNA gene was also normalized as RPK according to the 16S rRNA gene length of E. coli as 1,541 bp ⁵⁶. The normalized ARG profile (i.e. the list of ARGs and their RPK in a sample) was called “ARG profile.” The relative proportion of each ARG in a sample was calculated as the ratio of RPK of the ARG to the sum of RPK of all ARGs. The list of ARGs and their proportion in a sample was called “ARG composition.” The total ARG abundance of a sample was calculated as the ratio of the sum of RPK of all ARGs to RPK of 16S rRNA gene. The shotgun metagenomic sequence data were deposited in DRA database of DNA Data Bank of Japan (DDBJ) as accession numbers of SAMD00282059–00282074, SAMD00206783–00206785, SAMD00206872–00206884, SAMD00206896–00206913, and SAMD00282809–00282810 (Table S2).

3.5. Multivariate analysis

The antimicrobial resistome and microbial community in the samples were compared using principal component analysis (PCA) using R version 4.0.0. The ARG composition data and genus-level OTU abundance data were used in the PCA to compare the antimicrobial resistome and microbial community, respectively. In both PCAs, scaling was performed to focus on the relative change of each component rather than absolute change of abundant components.

DATA AVAILABILITY

All sequence data are available at DRA database of DNA Data Bank of Japan (DDBJ). The accession numbers of the sequence data are listed in the Methods and Supplementary information.

ACKNOWLEDGMENTS

This study was financially supported by JST MIRAI Program (Grant No. JPMJMI18DC), JSPS KAKENHI fund for Fostering Joint International Research (Grant No.16KK0146, 18KK0114, 21KK0073), Grant-in-Aid for Scientific Research B (Grant No. 19H02272, 21H03617) and Kurita Water and Environment Foundation (Grant No. 22T007). Computations were partially performed on the NIG supercomputer at ROIS National Institute of Genetics.

AUTHOR CONTRIBUTIONS

Ryo Honda: Conceptualization, Methodology, Formal analysis, Writing – Original Draft, Supervision; Norihisa Matsuura: Methodology, Software, Resource, Writing – Review & Editing; Sorn Sovannlaksmy: Investigation, Data Curation; Sawako Asakura, Yuta Morinaga, Than Van Huy: Investigation, Data Curation, Formal analysis; Muhammad Adnan: Data Curation, Writing – Review & Editing; Yalkhin Masakke: Methodology, Software; Hiroe Hara-Yamamura: Investigation, Formal analysis, Project administration, Writing – Review & Editing; Toru Watanabe: Supervision, Project administration, Funding acquisition.

COMPETING INTERESTS

The authors declare no competing interests.

Murray, C. J. et al. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet 399, 629–655 (2022).
O’ Neil, J. Review on Antibiotic resisitance. Antimicrobial Resistance: Tackling a crisis for the health and wealth of nations. Health and Wealth Nations (2014).
Rousham, E. K., Unicomb, L. & Islam, M. A. Human, animal and environmental contributors to antibiotic resistance in low-resource settings: integrating behavioural, epidemiological and One Health approaches. Proc. R. Soc. B Biol. Sci. 285, 20180332 (2018).
Rosas, I. et al. Characterization of Escherichia coli isolates from an urban lake receiving water from a wastewater treatment plant in Mexico city: Fecal pollution and antibiotic resistance. Curr. Microbiol. 71, 490–495 (2015).
Honda, R. et al. Impacts of urbanization on the prevalence of antibioticresistant Escherichia coli in the Chaophraya River and its tributaries. Water Sci. Technol. 73, 362–374 (2016).
Honda, R. et al. Estimated discharge of antibiotic-resistant bacteria from combined sewer overflows of urban sewage system. npj Clean Water 3, 15 (2020).
Kumar, M. et al. Prevalence of antibiotic resistance in the tropical rivers of Sri Lanka and India. Environ. Res. 188, 109765 (2020).
Xu, J. et al. Occurrence of antibiotics and antibiotic resistance genes in a sewage treatment plant and its effluent-receiving river. Chemosphere 119, 1379–1385 (2015).
Suzuki, S. & Hoa, P. T. P. Distribution of quinolones, sulfonamides, tetracyclines in aquatic environment and antibiotic resistance in indochina. Front. Microbiol. 3, 67 (2012).
Guo, J., Li, J., Chen, H., Bond, P. L. & Yuan, Z. Metagenomic analysis reveals wastewater treatment plants as hotspots of antibiotic resistance genes and mobile genetic elements. Water Res. 123, 468–478 (2017).
Pazda, M., Kumirska, J., Stepnowski, P. & Mulkiewicz, E. Antibiotic resistance genes identified in wastewater treatment plant systems – A review. Sci. Total Environ. 697, 134023 (2019).
Zhang, T., Zhang, X.-X. & Ye, L. Plasmid Metagenome Reveals High Levels of Antibiotic Resistance Genes and Mobile Genetic Elements in Activated Sludge. PLoS One 6, e26041 (2011).
Yang, Y., Li, B., Ju, F. & Zhang, T. Exploring Variation of Antibiotic Resistance Genes in Activated Sludge over a Four-Year Period through a Metagenomic Approach. Environ. Sci. Technol. 47, 10197–10205 (2013).
Honda, R., Tachi, C., Noguchi, M., Yamamoto-Ikemoto, R. & Watanabe, T. Fate and seasonal change of Escherichia coli resistant to different antibiotic classes at each stage of conventional activated sludge process. J. Water Health 18, 879–889 (2020).
Tong, J. et al. Microbial community evolution and fate of antibiotic resistance genes along six different full-scale municipal wastewater treatment processes. Bioresour. Technol. 272, 489–500 (2019).
Threedeach, S., Chiemchaisri, W. & Chiemchaisri, C. Fate of antibiotic resistant E. coli in anoxic/aerobic membrane bioreactor treating municipal solid waste leachate. Int. Biodeterior. Biodegradation 113, 57–65 (2016).
An, X. L. et al. Tracking antibiotic resistome during wastewater treatment using high throughput quantitative PCR. Environ. Int. 117, 146–153 (2018).
Chu, B. T. T. et al. Metagenomics Reveals the Impact of Wastewater Treatment Plants on the Dispersal of Microorganisms and Genes in Aquatic Sediments. Appl. Environ. Microbiol. 84, (2018).
Kotlarska, E., Łuczkiewicz, A., Pisowacka, M. & Burzyński, A. Antibiotic resistance and prevalence of class 1 and 2 integrons in Escherichia coli isolated from two wastewater treatment plants, and their receiving waters (Gulf of Gdansk, Baltic Sea, Poland). Environ. Sci. Pollut. Res. 22, 2018–2030 (2015).
Port, J. A., Cullen, A. C., Wallace, J. C., Smith, M. N. & Faustman, E. M. Metagenomic frameworks for monitoring antibiotic resistance in aquatic environments. Environ. Health Perspect. 122, 222–228 (2014).
Kasuga, I., Nagasawa, K., Suzuki, M., Kurisu, F. & Furumai, H. High-Throughput Screening of Antimicrobial Resistance Genes and Their Association With Class 1 Integrons in Urban Rivers in Japan. Front. Environ. Sci. 10, 1–15 (2022).
Rodríguez-Rojas, A., Rodríguez-Beltrán, J., Couce, A. & Blázquez, J. Antibiotics and antibiotic resistance: A bitter fight against evolution. Int. J. Med. Microbiol. 303, 293–297 (2013).
Chapman, J. S. Disinfectant resistance mechanisms, cross-resistance, and co-resistance. Int. Biodeterior. Biodegradation 51, 271–276 (2003).
Wales, A. & Davies, R. Co-Selection of Resistance to Antibiotics, Biocides and Heavy Metals, and Its Relevance to Foodborne Pathogens. Antibiotics 4, 567–604 (2015).
Lira, F., Vaz-Moreira, I., Tamames, J., Manaia, C. M. & Martínez, J. L. Metagenomic analysis of an urban resistome before and after wastewater treatment. Sci. Rep. 10, 1–9 (2020).
Ng, C. et al. Metagenomic and resistome analysis of a full-scale municipal wastewater treatment plant in Singapore containing membrane bioreactors. Front. Microbiol. 10, 1–13 (2019).
Yang, Y., Li, B., Zou, S., Fang, H. H. P. & Zhang, T. Fate of antibiotic resistance genes in sewage treatment plant revealed by metagenomic approach. Water Res. 62, 97–106 (2014).
Bengtsson-Palme, J. et al. Elucidating selection processes for antibiotic resistance in sewage treatment plants using metagenomics. Sci. Total Environ. 572, 697–712 (2016).
Ping, Q. et al. The prevalence and removal of antibiotic resistance genes in full-scale wastewater treatment plants: Bacterial host, influencing factors and correlation with nitrogen metabolic pathway. Sci. Total Environ. 827, 154154 (2022).
Zhang, S. et al. Fate of antibiotic resistant cultivable heterotrophic bacteria and antibiotic resistance genes in wastewater treatment processes. Chemosphere 135, 138–145 (2015).
Reinthaler, F. et al. Antibiotic resistance of E. coli in sewage and sludge. Water Res. 37, 1685–1690 (2003).
Novo, A. & Manaia, C. M. Factors influencing antibiotic resistance burden in municipal wastewater treatment plants. Appl. Microbiol. Biotechnol. 87, 1157–1166 (2010).
Ferreira da Silva, M., Vaz-Moreira, I., Gonzalez-Pajuelo, M., Nunes, O. C. & Manaia, C. M. Antimicrobial resistance patterns in Enterobacteriaceae isolated from an urban wastewater treatment plant. FEMS Microbiol. Ecol. 60, 166–176 (2007).
Stange, C., Sidhu, J. P. S., Toze, S. & Tiehm, A. Comparative removal of antibiotic resistance genes during chlorination, ozonation, and UV treatment. Int. J. Hyg. Environ. Health 222, 541–548 (2019).
Zhang, Y. et al. Inactivation of antibiotic resistance genes in municipal wastewater effluent by chlorination and sequential UV/chlorination disinfection. Sci. Total Environ. 512–513, 125–132 (2015).
Yuan, Q.-B., Guo, M.-T. & Yang, J. Fate of Antibiotic Resistant Bacteria and Genes during Wastewater Chlorination: Implication for Antibiotic Resistance Control. PLoS One 10, e0119403 (2015).
Munir, M., Wong, K. & Xagoraraki, I. Release of antibiotic resistant bacteria and genes in the effluent and biosolids of five wastewater utilities in Michigan. Water Res. 45, 681–693 (2011).
Tchobanoglous, G. et al. Suspended growth bioological treatment processes. in Wastewater Engineering: Treatment and Resource Recovery 697–940 (McGraw-Hill Education, 2014).
Sulfikar, Honda, R., Noguchi, M., Yamamoto-Ikemoto, R. & Watanabe, T. Effect of sedimentation and aeration on antibiotic resistance induction in the activated sludge process. J. Water Environ. Technol. 16, 94–105 (2018).
Baker-Austin, C., Wright, M. S., Stepanauskas, R. & McArthur, J. V. Co-selection of antibiotic and metal resistance. Trends Microbiol. 14, 176–182 (2006).
Aminov, R. I. Horizontal Gene Exchange in Environmental Microbiota. Front. Microbiol. 2, 1–19 (2011).
Jong, M., Harwood, C. R., Blackburn, A., Snape, J. R. & Graham, D. W. Impact of Redox Conditions on Antibiotic Resistance Conjugative Gene Transfer Frequency and Plasmid Fate in Wastewater Ecosystems. Environ. Sci. Technol. 54, 14984–14993 (2020).
Okubo, Y. et al. Change in clinical practice variations for antibiotic prescriptions across different pediatric clinics: A Japan’s nationwide observational study. J. Infect. Chemother. 27, 1621–1625 (2021).
Shishido, A. et al. Effects of a nudge-based antimicrobial stewardship program in a pediatric primary emergency medical center. Eur. J. Pediatr. 180, 1933–1940 (2021).
Sugawara, T. et al. Estimation of Amount of Antimicrobials Used by Pharmacy Surveillance. Japanese J. Infect. Prev. Control 27, 195–198 (2012).
AMR Clinical Reference Center. Surveillance of antibiotic sales in Japan. https://amrcrc.ncgm.go.jp/surveillance/020/archives_en.html (2021).
Conco, T. et al. Profiling of emerging pathogens, antibiotic resistance genes and mobile genetic elements in different biological wastewater treatment plants. J. Environ. Chem. Eng. 10, (2022).
Calderón-Franco, D. et al. Metagenomic profiling and transfer dynamics of antibiotic resistance determinants in a full-scale granular sludge wastewater treatment plant. Water Res. 219, 118571 (2022).
Raza, S., Shin, H., Hur, H. G. & Unno, T. Higher abundance of core antimicrobial resistant genes in effluent from wastewater treatment plants. Water Res. 208, 117882 (2022).
Yuan, Q.-B. et al. Redistribution of intracellular and extracellular free and adsorbed antibiotic resistance genes through a wastewater treatment plant by an enhanced extracellular DNA extraction method with magnetic beads. Environ. Int. 131, 104986 (2019).
Wang, R., Matsuura, N., Hara-Yamamura, H., Watanabe, T. & Honda, R. Initial behaviors and removal of extracellular plasmid gene in membrane bioreactor. J. Environ. Manage. 298, 113541 (2021).
Klindworth, A. et al. Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Res. 41, 1–11 (2013).
Rodriguez-R, L. M. & Konstantinidis, K. T. The enveomics collection: a toolbox for specialized analyses of microbial genomes and metagenomes. PeerJ Prepr. 4, e1900v1 (2016).
Jia, B. et al. CARD 2017: expansion and model-centric curation of the comprehensive antibiotic resistance database. Nucleic Acids Res. 45, D566–D573 (2017).
Jing, G. et al. Parallel-META 3: Comprehensive taxonomical and functional analysis platform for efficient comparison of microbial communities. Sci. Rep. 7, 40371 (2017).
Brosius, J., Palmer, M. L., Kennedy, P. J. & Noller, H. F. Complete nucleotide sequence of a 16S ribosomal RNA gene from Escherichia coli. Proc. Natl. Acad. Sci. 75, 4801–4805 (1978).

(Not answered)

SupplementaryMaterialsv4.pdf

Download PDF

Journal Publication

published 15 Jun, 2023

Read the published version in npj Clean Water →

Editorial decision: revise
27 Jan, 2023
Review #4 received at journal
14 Jan, 2023
Reviewer #5 agreed at journal
06 Jan, 2023
Reviewer #4 agreed at journal
03 Jan, 2023
Reviewer #3 agreed at journal
21 Oct, 2022
Review #1 received at journal
30 Sep, 2022
Reviewer #2 agreed at journal
25 Sep, 2022
Reviewer #1 agreed at journal
21 Sep, 2022
Reviewers invited by journal
21 Sep, 2022
Submission checks completed at journal
30 Aug, 2022
First submitted to journal
29 Aug, 2022
Editor assigned by journal
29 Aug, 2022

You are reading this latest preprint version

Transition of antimicrobial resistome in wastewater treatment plants: impact of process configuration, geographical location and season

Status:

Journal Publication

Version 1

Abstract

Figures

1. Introduction

2. Results And Discussion

2.1. ARG abundance in wastewater treatment process

2.2. Transition of ARG composition during wastewater treatment process

2.3. Transition of microbial community and association with ARG composition

2.4. Determinative factors of antimicrobial resistome in wastewater treatment

3. Methods

3.1. Sampling at wastewater treatment plants

3.2. Sample processing

3.3. Microbial population and community analysis based on 16S rRNA gene

3.4. Shotgun metagenomic sequence analysis

3.5. Multivariate analysis

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1