Science of the Total Environment 682 (2019) 756–764
Contents lists available at ScienceDirect
Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv
Fate of organic micropollutants and their biological effects in a drinking water source treated by a ﬁeld-scale constructed wetland Shuhui Xu a, Sicong Zhou a,c, Liqun Xing a,b, Peng Shi a,⁎, Wei Shi a, Qing Zhou a, Yang Pan a, Mao-Yong Song d, Aimin Li a a
State Key Laboratory of Pollution Control and Resource Reuse, School of the Environment, Nanjing University, Nanjing 210023, China Nanjing University, Yancheng Academy of Environmental Protection Technology and Engineering, Yancheng 224000, China Jiangsu Province Key Laboratory of Environmental Engineering, Jiangsu Provincial Academy of Environmental Science, Nanjing 210036, China d State Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China b c
H I G H L I G H T S
G R A P H I C A L
A B S T R A C T
• Puriﬁcation effects of constructed wetland on drinking source water. • The approach integrating of bioassays and instrumental analysis was applied. • Conventional water quality indexes were effectively controlled. • Comprehensive assessment of relevant toxicities and health risks of sources water. • Cytotoxicity, ROS and anti-androgen activity of outﬂow were all decreased.
a r t i c l e
i n f o
Article history: Received 22 March 2019 Received in revised form 9 May 2019 Accepted 11 May 2019 Available online 17 May 2019 Editor: Jay Gan Keywords: Organic micropollutants In vitro bioassays Constructed wetland Drinking water source Health risk
a b s t r a c t The safety of drinking water is directly related to the occurrence and concentrations of numerous organic micropollutants (OMPs) in source water. In this study, an approach integrating in vitro bioassays and chemical analyses was used to assess the puriﬁcation effects of a ﬁeld-scale constructed wetland on the fates of OMPs and their relevant toxicities and health risks in both summer and winter. Overall, 45 of 86 OMPs, including polycyclic aromatic hydrocarbons (PAHs), polychlorinated biphenyls (PCBs), organochlorine pesticides (OCPs), organophosphorus pesticides (OPPs), and phthalates (PAEs), were detected in at least one of the water samples. The constructed wetland signiﬁcantly decreased the concentrations of most types of OMPs, while showed negative effects on the PAEs and OPPs. Toxicological evaluation of water samples indicated that the cytotoxicity, reactive oxygen species (ROS) level and anti-androgen (Ant-AR) activity were all dramatically decreased after the constructed wetland treatment. PAEs and PAHs were the dominant contributors and accounted for 75.12–97.48% of the predicted Ant-AR potencies, while the total predicted Ant-AR potencies only contributed 3.13–15.97% of the observed Ant-AR potencies in the examined water samples, suggesting more OMPs that pose toxic effects are still undetected. The human health risk assessment demonstrated that noncarcinogenic risks of the water samples were acceptable. However, potential carcinogenic risks that were mainly induced by 2, 6dinitrotoluene, 2, 4-dinitrotoluene, pentachlorophenol and PAEs cannot be ignored. This study can help to understand the role of constructed wetlands in removing OMPs and biological effects from drinking water sources. © 2019 Elsevier B.V. All rights reserved.
⁎ Corresponding author at: State Key Laboratory of Pollution Control and Resource Reuse, School of the Environment, Nanjing University, 163 Xianlin Avenue, Nanjing 210023, China. E-mail address: [email protected]
https://doi.org/10.1016/j.scitotenv.2019.05.151 0048-9697/© 2019 Elsevier B.V. All rights reserved.
S. Xu et al. / Science of the Total Environment 682 (2019) 756–764
1. Introduction Drinking water safety is a particularly sensitive issue because it is directly and closely related to the public health (World Health Organization, 2011). With the rapid development of industrialization, urbanization and agricultural modernization, thousands of organic micropollutants (OMPs) are continuously discharged into aquatic environments and thus pose a threat to drinking water. In the past years, adverse health effects caused by OMPs, such as neurotoxicity, developmental and reproductive toxicity, and metabolic interference, have been frequently reported by previous literatures (Jie et al., 2013; Gaspar et al., 2014). For example, numerous OMPs that exist in municipal wastewater, river water and drinking water have been demonstrated to show obvious neurotoxicity, phytotoxicity, endocrinedisrupting effects and genotoxicity (Watson et al., 2012; Sun et al., 2017). Among these OMPs, polycyclic aromatic hydrocarbons (PAHs), polychlorinated biphenyls (PCBs), organochlorine pesticides (OCPs), organophosphorus pesticides (OPPs) and phthalates (PAEs) are persistent contaminants and have been ubiquitously detected in aquatic environments (Karalius et al., 2014; Peng et al., 2018; Škrbić et al., 2018), especially in the drinking water of cities along the Chinese coastland of the Yellow Sea according to our previous studies (Shi et al., 2017; Li et al., 2018b). Although these detected OMPs are generally at low ng/L levels, these contaminants could pose risks through additive/synergistic interactions after long-term exposure, accumulate through the food chain and further negatively affect human health (Mir-Tutusaus et al., 2018). In addition, the occurrences and distributions of OMPs in source water have become a crucial issue for drinking water safety due to the difﬁculty of removal and the mixture effect (Stamm et al., 2016; Ingold et al., 2018). Thus, constructed wetlands have been developed rapidly as advanced water treatment technology because of their economy and sustainability (Vymazal et al., 2005). It is reported that constructed wetlands exhibited high removal efﬁciency for drugs (e.g., ibuprofen, gemﬁbrozil and naproxen) and herbicides (e.g., metolachlor and terbuthylazine) (Pappalardo et al., 2016; Zhang et al., 2018). Moreover, these wetlands have an excellent effect on removing nitrogen, phosphorus and particulate matter, especially nitrate, which accounts for approximately 90% of total nitrogen loss (Billy et al., 2013; Schoumans et al., 2014; Sun et al., 2016). To the best of our knowledge, constructed wetlands are mostly used as an advanced treatment technology for the further treatment of efﬂuents of municipal or industrial wastewater treatment plants or the remediation of blackodor rivers, but there are few reports on the puriﬁcation of drinking water sources with multiple OMPs and biological effects. The demand for high-quality water resources is increasingly becoming an issue of global concern in the 21st century with rapid population growth, economic development and an increased demand for clean and healthy living environments (Simonovic, 2002). Bioassay-based evaluation of water quality and identiﬁcation of organic toxicants has been widely applied in environmental monitoring (Hu et al., 2015; Shi et al., 2017) and has been proven to be useful, reliable, rapid, and reproducible when determining the toxicological proﬁles of environmental samples. Moreover, through the integration of chemical analyses, the possible priority contributors causing toxicity can also be identiﬁed. For example, the major contributors to androgen receptor-antagonistic potencies in drinking water sources from the eastern cities of China were identiﬁed as diisobutyl phthalate (DIBP) and dibutyl phthalate (DBP) (Hu et al., 2013). Chlorpyrifos was identiﬁed as the major contributor to toxicity in surface water (Hu et al., 2015). Among the biological toxicity tests, in vitro bioassays as a part of effect-direct analysis, especially for the diagnosis of organic extract in environmental samples, were considered as excellent tools due to their high sensitivity and time efﬁciency (Hu et al., 2013; Shi et al., 2017). The purpose of this study was to investigate the fate and biological effects of OMPs in drinking water sources during the treatment process
of a ﬁeld-scale constructed wetland. A variety of chemical and biological tests are used to measure the puriﬁcation effect of constructed wetlands on drinking water sources. The key toxic substances of the source water were identiﬁed by effect-directed analysis. The non-carcinogenic and carcinogenic risks of health-related OMPs detected in drinking water sources were further assessed. Meanwhile, the fates of conventional water quality indexes including chemical oxygen demand (CODMn), ammonia nitrogen (NH3-N), total nitrogen (TN), and total phosphorus (TP) were simultaneously investigated. This study can fully evaluate the performance of the constructed wetland on OMPs and biological effects in drinking water source while also providing theoretical support for the selection of a proper treatment technology achieving safer drinking water. 2. Materials and methods 2.1. Chemicals and reagents The standard agents M-525.2-SV-ASL including semivolatile organic compounds (SVOCs), S-21827-R1-H including sixteen phthalic acid esters (PAEs), S-15402-R1 including nine organochlorine pesticides (OCPs), and M-8140M including twenty organophosphorus pesticides (OPPs) were obtained from J&K Scientiﬁc (Shanghai, China). Estradiol (E1), 17β-estradiol (βE2), Estriol (E3), 17-α-ethinylestradiol (EE2), dihydrotestosterone (DHT), phenol (NP), and octyl phenol (OP) and bisphenol A (BPA) were purchased from Sigma-Aldrich (St. Louis, MO, USA). HPLC grade solvents and G.R. grade reagents were used in this study. The abbreviations of all chemicals and CAS numbers are listed in Table S1. 2.2. Sample collection and pretreatment A large-scale constructed wetland, which is located in Yancheng city, Jiangsu province in eastern China (33°20′ N; 120°1′ E), has steadily operated for N5 years. Its area is approximately 222.87 ha with a daily treatment capacity of 300,000 tons. The Mangshe River is the source that ﬂows into the constructed wetland, which have been slightly polluted. In general, this constructed wetland includes three puriﬁcation zones: pretreatment zone (Pz, 1.97 × 105 m2, the depth is about 2 m, with ceramsite and aeration device), submerged-plant zone (Sz, 7.85 × 105 m2, the depth is 0.3–0.5 m, mainly planted with Phragmites austrails, Iris pseudacorus and Zizania caduciﬂora), ecological pond zone (Ez, 1.05 × 106 m2, the depth is about 3–5 m, with aquatic plants and animals) function as the water storage section accounts for half of the total area of the constructed wetland. The hydraulic retention times (HRTs) of the three zones were approximately 4 h, 30 h and 20 days, respectively. The inﬂuent water (In) and efﬂuents of the three puriﬁcation zones (Pz, Sz, Ez) were collected according to the HRTs of the different zones (Fig. 1). Grabbing samples at 1–2 m from the edge of the wetland with a sample collection apparatus. Especially, water samples in the ecological pond were ﬁrst mixed and then collected according to the grid point method due to its large area. In detail, water samples were collected every month according to the HRTs and taken in triplicate for conventional water quality indexes determination from January to December, 2016. Furthermore, 5 L of each water sample was collected at each site for OMPs and toxicity determination at the beginning, middle, and end of June and December, and then mixed to represent summer and winter samples. The water samples were collected into glass bottles which were prewashed with chromic acid, distilled water, ultrapure water and methanol in sequence, the pH was adjusted to below 2 with HCl, and 0.1–0.2 g of ascorbic acid was added. The samples were stored on ice during transportation, stored at 4 °C in the laboratory and extracted within 24 h. The water samples were extracted according to EPA 525.2 (USEPA., 1995) and our previous studies (Zhou et al., 2015; Shi et al., 2017; Li
S. Xu et al. / Science of the Total Environment 682 (2019) 756–764
Fig. 1. Overview map of sampling points of the Yanlong Lake constructed wetland. Red points indicate sampling locations of inﬂuent, pretreatment zone, submerged-plant zone and ecological pond zone. Especially, the grid point method is applied to gain representative water sample of point S4.
et al., 2018b). Brieﬂy, one liter of each water sample was acidiﬁed and ﬁltered through a 1 μm glass ﬁber ﬁlter and a 0.45 μm acetate cellulose ﬁlter to remove particles, and extracted using solid-phase extraction (SPE) followed by cartridge activation. For instrumental analyses, the water samples (1 L) were passed through Oasis HLB cartridges (200 mg, 6 mL, Waters) preactivated with 5 mL of hexane and dichloromethane (4:1, v/v), 5 mL of dichloromethane and methanol (1:1, v/v), and 10 mL of ultrapure water at a ﬂow rate of approximately 10 mL/min. Then, the cartridges were washed with 5 mL of hexane, 5 mL of hexane and dichloromethane (4:1, v/v), and 5 mL of dichloromethane and methanol (1:1, v/v) at a ﬂow rate of approximately 1 mL/min. The eluents were then concentrated to near dryness with a gentle stream of nitrogen. The ﬁnal extracts were reconstituted in 1 mL of dichloromethane for gas chromatography–mass spectrometry (GC–MS) analysis (TRACE 1300 GC-ISQ MS, Thermo Fisher, USA) and were reconstituted in 1 mL of methanol for ultra-performance liquid chromatography-mass spectrometry (UPLC-MS/MS) analysis (UPLC ICLASS & Xevo TQ-MS, Waters, USA). For bioassays analyses, the same SPE method was used, but the water samples (2 L) and Oasis HLB cartridges (500 mg, 6 mL, Waters) were replaced, and the ﬁnal extracts were reconstituted in 1 mL of dimethyl sulfoxide (DMSO). Ultrapure water was taken as the blank which was treated identically to the samples, and all samples were run in three parallels. All the extracts were stored at −20 °C for further analyses. 2.3. Chemical analyses CODMn, NH3-N, TN and TP were analyzed according to the standard wastewater monitoring and analysis method (The State Environmental Protection Administration of China, 2002). Quantitative analyses of OMPs were performed by GC–MS and UPLC-MS/MS according to our previous studies (Zhou et al., 2015; Shi et al., 2017; Li et al., 2018b). Among them, PAHs, PCBs, PAEs, OCPs and OPPs were quantiﬁed using GC–MS with a TG-5 MS capillary column (Thermo Scientiﬁc, 30 m × 0.25 mm × 0.25 μm), while E1, βE2, EE2, E3, DHT, NP, OP and BPA were quantiﬁed using UPLC-MS/MS. In brief, 1 μL of each sample was injected into the column with an injector temperature of 250 °C
in splitless injection mode, and ultra-pure helium was used as the carrier gas at 1.2 mL/min. The quadrupole mass spectrometer was operated at 70 eV in electron ionization (EI) mode, and the temperatures of the ion source and transfer line were set at 280 °C and 250 °C, respectively. Multiple reaction monitoring (MRM) mode was applied for quantitative determination, and the concentration of each component was calculated according to its calibration curve. And the detection methods are described in Supporting Information Text S1 and Table S2. The limit of detection (LOD) was deﬁned as 3 times the standard deviation (SD) of blanks (n = 3). Limit of quantiﬁcation (LOQ), procedural recovery and matrix spike recovery of each compound were shown in Supporting Information Table S3. 2.4. Cytotoxicity and reactive oxygen species (ROS) assays The cytotoxicity of extracts from the efﬂuents of different zones of the constructed wetland was assessed using the human hepatocarcinoma (HepG2) cell line. HepG2 cells were cultured using the Dulbecco's Modiﬁed Eagle Medium (DMEM) with 10% fetal bovine serum (FBS, Gibco) at 37 °C in an incubator containing 5% CO2 (Mcguigan and Li, 2014). A Cell Counting Kit-8 Assay (CCK-8; Nanjing KeyGen Biotech. Co. Ltd.) was used to evaluate the cell proliferation according to the method described elsewhere (Shi et al., 2012). Brieﬂy, 100 μL of HepG2 cells were seeded into 96-well plants at a density of 1 × 105 cells/mL and cultured at 37 °C for 24 h. Then, the concentrated extracts were added into the experimental wells, serially diluted 2 times column by column. After 24 h of incubation, the exposure medium was removed from each well, washed with phosphate buffered saline (PBS), and 10 μL of CCK-8 solution was added to each well and continuously incubated at 37 °C for 1.5 h. 0.5% phenol, DMEM with HepG2 cells and DMEM without HepG2 cells were used as positive, negative and blank controls, respectively. The ﬁnal concentration of DMSO in all media did not exceed 1%. The absorbance was measured at 450 nm using a microplate reader (Synergy H4, BioTek). To avoid the edge effect, the edge of the 96-well plants was not used. To assess the ROS, 10 mM DCFH solution was prepared by DMSO and D-Hanks solution. The cell inoculation and exposure are the same as
S. Xu et al. / Science of the Total Environment 682 (2019) 756–764
described above. HepG2 cells with D-Hanks solution and HepG2 cells with DMSO were used as negative and blank controls, respectively. After 24h, the medium was removed and the cells were washed twice with D-Hanks. 50 μL of DCFH solution was added into each well and continuously incubated at 37 °C for 25 min; then, the exposure medium was removed from each well, washed twice with D-Hanks, and 50 μL of D-Hanks was added into each well. After 10–20 min, 100 μL of Hoechst 33342 (5 μg/L) prepared by DMSO and D-Hanks was added into each well and continuously incubated at 37 °C for 20 min. Then, the exposure medium was removed from each well, washed twice with D-Hanks, and 50 μL of D-Hanks was added into each well. Next, the ﬂuorescence intensities were detected with a microplate reader (Synergy H4, BioTek) at an excitation wavelength of 488 nm and emission wavelength of 525 nm. Then, the ﬂuorescence intensities were detected with a microplate reader (Synergy H4, Bio Tek) at an excitation wavelength of 350 nm and emission wavelength of 460 nm. The ROS was calculated by the ratio of the two ﬂuorescence intensities.
where RfDi represents the reference dose value of each OMP (ng/kg bw/ day). It is considered to pose a health risk if HI is higher than 1. The carcinogenic risk (CR) was calculated using the following Eq. (3): CR ¼
where SFOi is the oral slope factor of each OMP (per ng/kg bw/day). In this study, the carcinogenic risks of OMPs with available SFOs were estimated. A CR value below 1.0 × 10−6 indicates negligible cancer risk, whereas a value between 1.0 × 10−6 and 1.0 × 10−4 suggests potential cancer risk, and a value above 1.0 × 10−4 is an indication of highpotential risk (Li et al., 2018a; Škrbić et al., 2018). Detailed information regarding the RfD and SFO of OMPs are shown in Supporting Information Table S4. 2.7. Mass balance and statistical analysis
2.5. MDA-kb2 and MLVN cell culture and reporter gene assays MDA-kb2 cells were used to determine the potencies of androgen receptor (AR) agonists and antagonists. 76 μL of cell suspension with 1 × 105 cells/mL which was diluted with L-15 medium (Gibco) (10% activated carbon-treated fetal bovine serum (HyClone)) was inoculated into a 384-well plate and incubated at 37 °C without CO2 for 24 h. Then, 4 μL of a series of concentrations of test extracts were added to a 384-well plate, and different concentrations of ﬂutamide (FLU) and DHT were taken as positive controls for the antiandrogen effect; blank and solvent control (0.1% DMSO) were also performed in this study. The concentration of DMSO in the medium did not exceed 0.1%. Then, the cells were cultured in an incubator at 37 °C for 24 h without CO2. After that, the medium in each well was discarded, 10 μL of 1 × cell lysate was added into each well, and the 384-well plate was shaken for 10 min to ensure sufﬁcient cell lysis. Finally, the luciferase intensity was measured by a microplate reader after the addition of 25 μL of luciferase assay reagent. MVLN cells were used to determine the potency of estrogen in the water samples. Similarly, MVLN cells were cultured in DMEM medium at 37 °C with 5% CO2, and a series of concentrations of test extracts were added to the 384-well plate. Different concentrations of E2 were taken as positive controls, and blank and solvent control (0.1% DMSO) were also performed. The measured method is the same as the MDAkb2 cells. All exposures were performed in ﬁve parallels on each plate and each sample was tested in triplicate. 2.6. Health risk assessment The average daily dose (ADD) of each OMP was calculated using Eq. (1) (Ding et al., 2015; Li et al., 2018a):
ðC IR APÞ BW
where C is the concentration of each OMP in surface water (ng/L), IR is the ingestion rate of water (L/day), AP is the absorption percent of intake which is assumed to be 100%, and BW is body weight (kg). In the exposure risk calculation, age and gender-speciﬁc intake of water were also included (Ding et al., 2015). The non-carcinogenic and carcinogenic risks of each OMP and ΣOMPs were estimated according to the methods described by previous studies (Ding et al., 2015; Li et al., 2018a). The non-carcinogenic risk was assessed using the Hazard Index (HI) shown in Eq. (2): n ADDi
HI ¼ ∑1
In this study, all the extracts did not show signiﬁcant pseudoandrogenic (Pse-AR) and estrogen-like effects, so we mainly discuss their anti-androgen effects (Ant-AR). The anti-androgenic activity equivalent (Ant-AR-EQs) of the environmental samples was used to characterize their anti-androgen effects according to previous studies (Hu et al., 2013; Shi et al., 2017). The anti-androgenic activity of the sample is expressed by the equivalent value of the standard (DHT/ FLU) as follows: Ant1AR1EQ ¼
EC50 of FLU enrichment factors of tested samples
GraphPad Prism 5 was used to calculate the EC50 of the standard and enrichment factor of the environmental sample identical to the standard EC50. The RePs of individual detected OMPs to FLU originate from previous reported literatures and are shown in Supporting Information Table S5. The signiﬁcant differences were performed by use of one-way analysis of variance (ANOVA) followed by Duncan's multiple comparison test using SPSS 11 statistical software (SPSS Inc., Chicago, IL). In the bioassay, no signiﬁcant differences were observed between the ﬁeld blanks and the solvent control. 3. Results and discussion 3.1. Long-term removal of conventional water quality indexes The constructed wetland has a good reduction effect on CODMn, NH3-N, TN and TP in the drinking water source (Fig. 2). It showed that CODMn, NH3-N and TP were mostly below the limit of class III of the environmental quality standards for drinking source water in China (GB 3838-2002) after the constructed wetland treatment, while TN was still higher than the standard although it has already dropped signiﬁcantly. The average removal rates of the constructed wetland for CODMn, NH3-N, TN and TP were 13.22%, 49.32%, 35.91% and 47.53%, respectively (Fig. 2). For different water quality indicators, the removal efﬁciency of each zone is different. In general, the principle puriﬁcation unit for CODMn removal was the ecological pond zone, accounting for approximately 94% of the total reduction. This may be caused by the longest HRT of the ecological pond zone, which enables sufﬁcient biodegradation of organic matter by microorganisms. Previous studies have demonstrated that the HRT of the constructed wetland is the key factor that determines the COD removal (Saeed and Sun, 2012). The pretreatment and ecological pond zones contribute mostly to the reduction of NH3-N in the inﬂuent, with the highest removal efﬁciencies of 45% and 33%, respectively. (Fig. 2) The nitrogen form of NH3-N can be oxidized to nitrate through the synergistic effects of aeration, microbes and plants in the pretreatment and submerged-plant zones, which have
S. Xu et al. / Science of the Total Environment 682 (2019) 756–764
Fig. 2. Concentration of chemical oxygen demand (CODMn), ammonia nitrogen (NH3-N), total nitrogen (TN), total phosphorus (TP) in different water samples. Red dot dash line presented the standard of the environmental quality standards for drinking water source in China, blue solid cube represents the average concentration of each sample. The samples number of each plot-box is twelve which includes one composite sample in each month in a year. In: inﬂuent water; Pz: pretreatment zone; Sz: submerged-plant zone; Ez: ecological zone.
been reported in previous studies (Lee et al., 2010). In addition, approximately 47% and 45% of the TP was removed by the pretreatment and ecological pond zones of the constructed wetland. Generally, the removal efﬁciency of the construction wetland is high for CODMn and NH3-N, but low and variable for the nutrients of nitrogen and phosphate (Song et al., 2006; García et al., 2010). It was reported that the removal efﬁciencies of a constructed wetland system for COD, NH3-N and TP were 62.2 ± 10.1%, 40.6 ± 15.3% and 29.6 ± 12.8% for sewage treatment, respectively (Song et al., 2006), while the removal efﬁciencies of BOD, TN and TP reached 90%, 50% and 60% in the tropics, respectively (Shuval, 1986; Sperling, 1996). Another study demonstrated that the removal efﬁciencies for COD and NH3-N were approximately 35% and 80%, respectively, with an 18.5 km ﬁeld-scale wetland (Sun et al., 2016). The removal efﬁciency for COD in our study was lower than those in other reported studies, while the removal efﬁciencies for NH3-N, TN and TP were generally consistent with previous studies, a ﬁnding which may be due to different inﬂow concentrations, designs, operating conditions, maintenance parameters, climate conditions and anthropogenic factors (Maine et al., 2007; Saeed and Sun, 2012; Wei and Ji, 2012). Moreover, most of the constructed wetlands were designed to treat urban wastewater or synthetic urban wastewater, while the object of this study is the drinking water source, which is much cleaner than sewage. The low-concentration pollutants in the inﬂuent also make deep puriﬁcation of the constructed wetland more difﬁcult.
OMPs in the Huaihe River, Yangtze River and Taihu Lake reported in our previous study (Shi et al., 2017), but were consistent with the emerging organic pollutants in surface water from the Yangtze River Delta reported by Peng et al. (2018). In addition, the PAEs, especially di-n-butyl phthalate (DBP), diisobutyl phthalate (DIBP), bis(2ethylhesy) phthalate (BEHP) and diethyl phthalate (DEP), were the dominant pollutants, accounting for 58–84% of the total OMPs in summer and 55–83% of the total OMPs in winter, respectively, which were similar with previous studies (Li et al., 2017; Shi et al., 2017).
3.2. Fates of OMPs in the constructed wetland As shown in Fig. 3, a total of 45 OMPs including 10 PAHs, 10 PAEs, 6 OCPs, 7 OPPs, 6 PCBs and 6 other pollutants were detected in all the water samples. The total concentrations of OMPs in In, Pz, Sz and Ez were respectively 2.43 ± 0.18 μg/L, 11.39 ± 0.70 μg/L, 2.74 ± 0.21 μg/L and 4.28 ± 0.56 μg/L in summer, while they were respectively 4.56 ± 0.35 μg/L, 6.28 ± 0.34 μg/L, 8.23 ± 0.53 μg/L and 6.91 ± 0.87 μg/L in winter, which were higher than the total concentrations of detected
Fig. 3. Cumulative concentrations and composition proﬁles of different types of organic micro-pollutants (OMPs) in water samples in summer and winter. The center circle comprehensively displays the detected OMPs, and the fan shaped graph at four corners represents the detected OMPs at each sampling point. Different color represents different OMPs and the size of the fan indicates the level of concentration. Full names of abbreviated compounds can be found in Table S1. In: inﬂuent water; Pz: pretreatment zone; Sz: submerged-plant zone; Ez: ecological zone.
S. Xu et al. / Science of the Total Environment 682 (2019) 756–764
Meanwhile, sterols were found as the dominant OMPs, accounting for 38–92% in the surface water system of northern Serbia (Škrbić et al., 2018). Remarkably, the total concentrations of OMPs conversely increased after treatment of the constructed wetland without new emission source. Moreover, the composition proﬁles of the OMPs also varied greatly during the treatment processes of the constructed wetland (Fig. 3). PAHs and OCPs were greatly reduced by 57.23% and 34.76% in summer and 25.53% and 27.30% in winter, respectively. PCBs were slightly reduced by 5.79% and 4.84% in summer and winter, respectively, while the others were slightly reduced by 1.87% in summer, but 80.83% in winter. Conversely, the negative effects of the constructed wetland on PAEs and OPPs were found in both summer and winter. OPPs and PAEs were increased by 847.5% and 60.26% in summer and 126.79% and 26.43% in winter, respectively. Obviously, removal performance for PAEs and OPPs was worse in summer than its in winter. Previous studies showed that direct uptake, accumulation and translocation of OMPs by plants and microbial degradation have been considered as important mechanisms for phytoremediation technology, especially for the degradation of organic compounds which are relatively recalcitrant in soil (Onesios et al., 2009). Therefore, it can be speculated that the increase of OPPs in the submerged-plant section was mainly due to the releases from the plants and sediments. The PAEs were mainly increased by the pretreatment section. The fates of PAEs in constructed wetlands can be strongly inﬂuenced by their adsorption to soil and sediment (García et al., 2010). It must be mentioned that the initial concentration of PAEs in inﬂuent water is relatively high, the PAEs which was adsorbed on the suspended media and sediments surface may be released under the action of air agitation. It can be speculated that the increase of PAEs in the pretreatment section was mainly due to the releases from the artiﬁcial media and sediments. Other detected OMPs moderately decreased after the treatment by the constructed wetlands. The synergism among the interaction among three factors (natural physical, chemical, and biological phenomena) was the main removal mechanism. Temperature and water quality also exerted signiﬁcant impacts on the removal of OMPs (Reyescontreras et al., 2012), which may be the main reason for the differences between summer and winter. 3.3. Cytotoxicity and ROS of water samples The cytotoxicity of the water extracts was examined using the CCK-8 method in this study. The results showed that all the samples exhibited obvious cytotoxicity after 24-h incubation with the HepG2 cell line, which was more toxic than the other source and drinking water in the Huai River, Yangtze River and Taihu Lake (Shi et al., 2016; Shi et al., 2017). It was reported that complex organic mixtures usually showed cytotoxicity but were difﬁcult to effectively evaluate (Jemimah et al., 2016; Morgado et al., 2017). However, the EC50 of enrichment factors
was increased from 50 to 101 in summer and from 26 to 74 in winter, respectively, meaning that the cytotoxicities were signiﬁcantly reduced by the constructed wetland (Table S6). The cytotoxicities of water samples in winter were higher than those in summer, respectively. The total concentration in winter was also higher than it in summer which was consistent with the results of cytotoxicity test. Interestingly, almost all puriﬁcation sections of the constructed wetland showed reduction effects on the cytotoxicity. That seemed not to agree with the results of the fates of OMPs, which may be related to the composition proﬁles variety and undetected toxic substances causing cytotoxicity. Although this study detected tens of OMPs in the water samples, they are still only a tip of the iceberg. For instance, hundreds of emerging organic pollutants including industrial chemicals, pesticides, pharmaceuticals and personal care products were detected in surface water from the Yangtze River Delta (Peng et al., 2018), indicating that the cytotoxicity cannot be completely explained by instrumentally quantiﬁed OMPs. ROS is a category of derivatives of oxygen (e.g., •O− 2 , H2O2, HO2•, •OH and ROO•) that are chemically reactive and signs of oxidative damage to organisms, and they can increase dramatically under different environmental stresses such as nutrient deﬁciency, UV radiation and pollutant exposure (Ralph et al., 2010; Zhe et al., 2018). Under physiological concentrations, ROS can act as signaling molecules mediating cell growth, migration and differentiation, whereas at higher concentrations, they may cause cell death and apoptosis through oxidative damage of macromolecules including proteins, lipids and DNA (Helﬁnger and Schröder, 2018). The results of ROS levels showed that the ROS levels increased with the enrichment factors of the tested samples (Fig. 4). The ROS levels were nearly stable when the logarithm of the enrichment factors were b1, while they exhibited a remarkable decrease after treatment by the constructed wetland when the logarithm of the enrichment factors reached up to N1. This indicated that the constructed wetland efﬁciently eliminated the ROS accumulation, which was consistent with the results of cytotoxicity in the water samples. 3.4. Antiandrogen (Ant-AR) activities and possible causes of water samples In this study, none of the water samples exhibited AR agonistic potency, which indicated the AR agonistic compounds in these water samples were less than the limit of detection by use of the bioassay. Previous studies also found similar results, claiming an undetectable AR agonistic effect in water samples from the Huai River (Shi et al., 2017), Yangtze River (Shi et al., 2016) and Tai Lake (Hu et al., 2013). However, the average responses of the Ant-AR activity were signiﬁcantly suppressed to 25.41–47.58% of the activity induced by 1.0 × 10−9 mol/L DHT in summer, and 24.40–66.75% in winter at the maximal exposure concentration (16 times the concentration in the source water), respectively (Fig. 5). Especially, the extracts of the inﬂuent water showed a signiﬁcant Ant-AR activity when the concentrated times reached 4, but the
Fig. 4. The reactive oxygen species (ROS) levels of different water samples with different enrichment factors in summer (a) and winter (b). The abscissa represents the logarithm of the water samples enrichment factors. In: inﬂuent water; Pz: pretreatment zone; Sz: submerged-plant zone; Ez: ecological zone.
S. Xu et al. / Science of the Total Environment 682 (2019) 756–764
Fig. 5. Androgen receptor (AR) antagonistic potencies of different water samples measured by use of reporter gene assays based on MDA-kb2 cell at 1, 2, 4, 8, and 16 times the original concentration. The levels of signiﬁcance were set as ⁎⁎⁎ (p b 0.001), ⁎⁎ (p b 0.01) and ⁎ (p b 0.05). In: inﬂuent water; Pz: pretreatment zone; Sz: submerged-plant zone; Ez: ecological zone.
Ant-AR activity was mostly eliminated after the constructed wetland treatment (Fig. 5). The relationship of the suppressed expression of the reporter gene with a concentration-dependent manner could be still constructed in this study. The Ant-AR-EQs of the water samples ranged from 33.5 μg FLU/L in the inﬂuent water to b15.3 μg FLU/L (detection limit) in the efﬂuent of the constructed wetland in summer, and 45.21 μg FLU/L in the inﬂuent to b15.3 μg FLU/L in the efﬂuent in winter (Table 1). The Ant-AR-EQs of the water samples in this study were comparable with the water samples from the Huai River (b15.3–43 μg FLU/L), Yangtze River (41–140 μg FLU/L) and Tai Lake (16–76 μg FLU/L) (Hu et al., 2013), but lower than the levels of the water samples from the Pearl River system (20.4–935 μg FLU/L) (Zhao et al., 2011). Overall, the Ant-AR activity of the inﬂuent was markedly reduced by N50% in both summer and winter, which indicated that the constructed wetland posed a good puriﬁcation effect on the Ant-AR activity in the drinking water source. This ﬁnding is consistent with the results of other researchers, who reported that constructed wetland can effectively remove emerging pollutants such as steroid hormones and biocides (Chen et al., 2019). The contribution of the detected OMPs to the total endocrine activity of these water samples was further investigated in this study. The predicted Ant-AR-EQs based on the detected OMPs were 1.05–4.06 μg
FLU/L. Among the detected OMPs, PAEs and PAHs accounted for 13.99–83.75% and 0.97–73.80% of the total predicted Ant-AR-EQs, respectively. DBP and benzo(α)pyrene (BaP) followed by benzo(k)ﬂuoranthene (BkF) and benzo(β)ﬂuoranthene (BbF) were the dominant contributors to the predicted Ant-AR-EQs. Early studies have also identiﬁed PAHs, OCPs and PAEs (especially DIBP and DBP) as major contributors to AR antagonistic potencies in source waters (Qu et al., 2011; Zhang et al., 2012; Hu et al., 2013). However, the predicted Ant-AREQs accounted for only 3.13–15.97% of the measured Ant-AR-EQs, indicating the presence of undetected AR antagonists or synergistic effects. Similarly, our study also indicated that N50% of the measured Ant-AREQs in surface water of this area cannot be explained by the detected OMPs (Shi et al., 2016). Additionally, the reduction of Ant-AR-EQs generally agreed with the preceding results of the cytotoxicity assay. 3.5. Health risk assessment As shown in Fig. 6a, the HIs of 18 health-related OMPs were all lower than 0.10 for diverse groups (age and gender) in each treatment section, suggesting that the ADDs of these compounds were at least 1–2 orders lower than their RfD values and within acceptable levels. The HIs of ΣOMPs in the efﬂuent of the constructed wetland for men, women,
Table 1 Anti-androgenic activity equivalents (Ant-AR-EQs), predicted Ant-AR-EQs of each compound and the sum of the predicted Ant-AR-EQs (ΣAnt-AR-EQs) and percentages of the ΣAnt-AREQs to the Ant-AR-EQs in water samples from different sections in summer and winter. Full names of abbreviated compounds can be found in Table S1. Summer
Predicted Ant-AR-EQs (μg ﬂutamide/L)
Ant Phe Pyr BaP CHR BkF BbF Bap BBP DBP DIBP DEP DMP p,p′-DDE p,p′-DDD Dz MP PCB88 ΣAnt-AR-EQs Ant-AR-EQ of water samples (μg ﬂutamide/L) Contribution of Σpredicted to water samples (%)
1.95E−03 3.81E−03 1.11E−02 3.40E−02 1.18E−01 1.58E−01 1.90E−01 2.58E−01 1.09E−03 1.07E−01 3.43E−02 4.13E−03 1.43E−02 7.92E−03 0.00E+00 0.00E+00 8.45E−03 9.60E−02 1.05E+00 3.35E+01
3.38E−03 3.85E−03 5.16E−01 3.34E−02 1.30E−01 3.17E−01 1.78E−01 5.01E−01 9.84E−04 2.07E+00 1.87E−01 2.78E−03 1.44E−02 3.94E−03 1.35E−02 0.00E+00 8.46E−03 7.63E−02 4.06E+00 2.54E+01
2.48E−03 5.64E−03 1.51E−02 5.30E−02 2.24E−01 3.03E−01 3.03E−01 4.78E−01 0.00E+00 6.66E−01 1.57E−01 9.11E−03 2.72E−02 0.00E+00 5.17E−03 1.65E−01 7.75E−02 1.68E−01 2.66E+00 1.76E+01
1.48E−03 3.90E−03 8.37E−03 0.00E+00 0.00E+00 1.45E−01 1.83E−01 2.45E−01 5.16E−04 4.88E−01 1.21E−01 1.31E−02 1.51E−02 5.71E−03 0.00E+00 1.12E−01 2.08E−01 7.69E−02 1.63E+00 b1.53 E +01 ND
2.12E−03 5.54E−03 1.46E−02 0.00E+00 0.00E+00 0.00E+00 0.00E+00 4.84E−01 0.00E+00 6.46E−01 1.08E−01 4.78E−03 9.32E−02 0.00E+00 0.00E+00 1.62E−01 4.09E−02 2.18E−01 1.78E+00 4.51E+01
1.30E−03 3.32E−03 1.04E−02 3.69E−02 1.27E−01 1.91E−01 2.07E−01 2.92E−01 0.00E+00 1.09E−01 5.99E−02 1.39E−02 1.03E−02 1.05E−02 0.00E+00 1.10E−01 1.17E−01 8.33E−02 1.38E+00 b1.53 E +01 ND
2.32E−03 7.17E−03 1.50E−02 0.00E+00 0.00E+00 3.12E−01 3.08E−01 4.84E−01 0.00E+00 2.26E+00 1.64E−01 7.04E−03 1.98E−02 0.00E+00 0.00E+00 1.61E−01 4.78E−02 1.64E−01 3.95E+00 b1.53 E +01 ND
2.22E−03 5.96E−03 1.43E−02 0.00E+00 0.00E+00 0.00E+00 0.00E+00 0.00E+00 0.00E+00 1.75E+00 1.69E−01 5.98E−03 1.81E−02 0.00E+00 1.33E−03 1.56E−01 4.78E−02 1.54E−01 2.32E+00 b1.53 E +01 ND
S. Xu et al. / Science of the Total Environment 682 (2019) 756–764
Fig. 6. Non-carcinogenic (HI) and carcinogenic risks (CR) of different types of OMPs in water samples in summer and winter (M: man; W: woman; B: boy; G: girl). In: inﬂuent water; Pz: pretreatment zone; Sz: submerged-plant zone; Ez: ecological zone.
boys and girls were respectively 4.50 × 10−2, 3.86 × 10−2, 4.42 × 10−2 and 4.33 × 10−2 in summer and 6.23 × 10−2, 5.35 × 10−2, 6.13 × 10−2 and 6.00 × 10−2 in winter, results which are much higher than the previous results of drinking water along the coastland of the Yellow Sea (Li et al., 2018b) and northern Serbia (Škrbić et al., 2018), but lower than the results of the water samples from the Huai River (Zhou et al., 2015). The relative sensitivity order was man N boy N girl N woman, which agreed with the ﬁnding of Ding et al. (2015). Moreover, 2, 2′, 4, 4′-tetrachlorobiphenyl (TtCB) was found as the dominant OMP causing noncarcinogenic risk, accounting for 48.57–68.63% of the HIs in summer, and 69.15–75.63% of the HIs in winter, respectively, which is consistent with the results of the Huai River in our previous study (Zhou et al., 2015). In addition, the HI values in winter were higher than those in summer, and a noncarcinogenic risk reduction effect was found during the treatment process of the constructed wetland. In addition to the noncarcinogenic risk, the carcinogenic risks of 22 detected OMPs with available SFO were also assessed (Fig. 6b). The results indicated that potential cancer risks (between 1.0 × 10−6 and 1.0 × 10−4) still existed in the efﬂuents of all the treatment sections in both summer and winter. The CRs of ΣOMPs for men, women, boys and girls were respectively 7.85 × 10−6, 6.75 × 10−6, 7.73 × 10−6 and 7.56 × 10−6 in summer and 8.32 × 10−6, 7.15 × 10−6, 8.19 × 10−6 and 8.02 × 10−6 in the ﬁnal efﬂuent of the constructed wetland, results which were also higher than the previous results of drinking water along the coastland of the Yellow Sea (Li et al., 2018b) and northern Serbia (Škrbić et al., 2018) but lower than the results of the Huai River (Zhou et al., 2015). The relative sensitivity order was similar with the noncarcinogenic risk and study by Ding et al. (2015). However, 2,6dinitrotoluene, 2,4-dinitrotoluene, pentachlorophenol and PAEs were found as the dominant OMPs, accounting for approximately 45.23–76.90% and 52.51–69.10% of the CRs in summer and winter, respectively. In the present study, a ﬁeld-scale and steady-operated constructed wetland was selected to investigate the fates of OMPs and biological effects during the treatment. In general, the constructed wetland showed excellent performance on removing conventional water quality indexes, cytotoxicity, ROS and Ant-AR activities. However, the total concentration of detected OMPs inversely increased after the treatment by the constructed wetland, and this ﬁnding was not consistent with the change of the biological effects. The discrepancy indicates that
bioassays may be more suitable than chemical analysis when determining the removal effects of the treatment units. Although the toxic effects of the inﬂuent were signiﬁcantly lowered after the constructed wetland, potential cancer risks still existed in the efﬂuent, effects which were mainly caused by several persistent OMPs. Overall, this study may help to understand the function of constructed wetlands regarding their removal of conventional pollutants, OMPs and biological effects from source water; the study also provides insights into the subsequent drinking water treatment in waterworks. Acknowledgments This study was ﬁnancially supported by National Natural Science Foundation of China (Grant No. 51608253, 51678290 and 51438008), Natural Science Foundation of Jiangsu Province, China (BK20160656) and the Open Research Fund of Jiangsu Province Key Laboratory of Environmental Engineering (ZX2017008). Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.scitotenv.2019.05.151. References Billy, C., Birgand, F., Ansart, P., Peschard, J., Sebilo, M., Tournebize, J., 2013. Factors controlling nitrate concentrations in surface waters of an artiﬁcially drained agricultural watershed. Landsc. Ecol. 28, 665–684. Chen, J., Liu, Y., Deng, W., Ying, G., 2019. Removal of steroid hormones and biocides from rural wastewater by an integrated constructed wetland. Sci. Total Environ. 660, 358–365. Ding, J., Shen, X., Liu, W., Covaci, A., Yang, F., 2015. Occurrence and risk assessment of organophosphate esters in drinking water from Eastern China. Sci. Total Environ. 538, 959–965. García, J., Rousseau, Dpl, Morató, J., Lesage, E., Matamoros, V., 2010. Contaminant removal processes in subsurface-ﬂow constructed wetlands: a review. Environ. Sci. Technol. 40, 561–661. Gaspar, F.W., Castorina, R., Maddalena, R.L., Nishioka, M.G., Mckone, T.E., Bradman, A., 2014. Phthalate exposure and risk assessment in California child care facilities. Environ. Sci. Technol. 48, 7593–7601. Helﬁnger, V., Schröder, K., 2018. Redox control in cancer development and progression. Mol. Asp. Med. 63, 88–98. Hu, X., Shi, W., Wei, S., Zhang, X., Feng, J., Hu, G., Chen, S., Giesy, J.P., Yu, H., 2013. Occurrence and potential causes of androgenic activities in source and drinking water in China. Environ. Sci. Technol. 47, 10591–10600.
S. Xu et al. / Science of the Total Environment 682 (2019) 756–764
Hu, X., Shi, W., Yu, N., Jiang, X., Wang, S., Giesy, J.P., Zhang, X., Wei, S., Yu, H., 2015. Bioassay-directed identiﬁcation of organic toxicants in water and sediment of Tai Lake, China. Water Res. 73, 231–241. Ingold, K., Moser, A., Metz, F., Herzog, L., Bader, H.P., Scheidegger, R., Stamm, C., 2018. Misﬁt between physical affectedness and regulatory embeddedness: the case of drinking water supply along the Rhine River. Glob. Environ. Chang. 48, 136–150. Jemimah, N.S., Subathra, D.C., Mohanasrinivasan, V., George, P.D.C., 2016. Bioactivity of marine Streptomyces sp. VITJS4: interactions of cytotoxic phthalate derivatives with human topoisomerase II α: an in silico molecular docking analysis. Interdiscip. Sci. 10, 1–10. Jie, X., Zheng, F., Lei, G., Biao, Z., Wochenschrift, Y.J., 2013. Neurotoxic effects of nonylphenol: a review. Wien. Klin. Wochenschr. 125, 61–70. Karalius, V.P., Harbison, J.E., Plange-Rhule, J., Breemen, R.B.V., Li, G., Ke, H., Durazo-Arvizu, R.A., Mora, N., Dugas, L.R., Vail, L., 2014. Bisphenol A (BPA) found in humans and water in three geographic regions with distinctly different levels of economic development. Environ. Health Insights. 8, 1–3. Lee, C-g., Fletcher, T.D., Sciences, G.S., 2010. Nitrogen removal in constructed wetland systems. Eng. Life Sci. 9, 11–22. Li, R., Liang, J., Gong, Z., Zhang, N., Duan, H., 2017. Occurrence, spatial distribution, historical trend and ecological risk of phthalate esters in the Jiulong River, Southeast China. Sci. Total Environ. 580, 388–397. Li, J., Zhang, Z., Ma, L., Zhang, Y., Niu, Z., 2018a. Implementation of USEPA RfD and SFO for improved risk assessment of organophosphate esters (organophosphate ﬂame retardants and plasticizers). Environ. Int. 114, 21–26. Li, Z., Chang, F., Shi, P., Chen, X., Yang, F., Zhou, Q., Pan, Y., Li, A., 2018b. Occurrence and potential human health risks of semi-volatile organic compounds in drinking water from cities along the Chinese coastland of the Yellow Sea. Chemosphere 206, 655–662. Maine, M.A., Suñe, N., Hadad, H., Sánchez, G., Bonetto, C., 2007. Removal efﬁciency of a constructed wetland for wastewater treatment according to vegetation dominance. Chemosphere 68, 1105–1113. Mcguigan, C.F., Li, X.F., 2014. Cytotoxicity and genotoxicity of phenazine in two human cell lines. Toxicol. in Vitro 28, 607–615. Mir-Tutusaus, J.A., Baccar, R., Caminal, G., Sarrà, M., 2018. Can white-rot fungi be a real wastewater treatment alternative for organic micropollutants removal? A review. Water Res. 138, 137–151. Morgado, P.I., Jose, S., Wanke, R., Amm, A., Soﬁa, C.A., Jordao, L., 2017. Integration of cellular and molecular endpoints to assess the toxicity of polycyclic aromatic hydrocarbons in hepg2 cell line. Environ. Toxicol. Chem. 36, 3404–3414. Onesios, K.M., Yu, J.T., Bouwer, E.J., 2009. Biodegradation and removal of pharmaceuticals and personal care products in treatment systems: a review. Biodegradation 20, 441–466. Pappalardo, S.E., Otto, S., Gasparini, V., Zanin, G., Borin, M., 2016. Mitigation of herbicide runoff as an ecosystem service from a constructed surface ﬂow wetland. Hydrobiologia 774, 193–202. Peng, Y., Fang, W., Krauss, M., Brack, W., Wang, Z., Li, F., Zhang, X., 2018. Screening hundreds of emerging organic pollutants (EOPs) in surface water from the Yangtze River Delta (YRD): occurrence, distribution, ecological risk. Environ. Pollut. 241, 484–493. Qu, C.S., Chen, W., Bi, J., Huang, L., Li, F.Y., 2011. Ecological risk assessment of pesticide residues in Taihu Lake wetland, China. Ecol. Model. 222, 287–292. Ralph, S.J., Rodríguez-Enríquez, S., Neuzil, J., Saavedra, E., Moreno-Sánchez, R., 2010. The causes of cancer revisited: “mitochondrial malignancy” and ros-induced oncogenic transformation–why mitochondria are targets for cancer therapy. Mol. Asp. Med. 31 (2), 145–170. Reyescontreras, C., Hijosavalsero, M., Sidrachcardona, R., Bayona, J.M., Bécares, E., 2012. Temporal evolution in PPCP removal from urban wastewater by constructed wetlands of different conﬁguration: a medium-term study. Chemosphere 88, 161–167. Saeed, T., Sun, G., 2012. A review on nitrogen and organics removal mechanisms in subsurface ﬂow constructed wetlands: dependency on environmental parameters, operating conditions and supporting media. J. Environ. Manag. 112, 429–448. Schoumans, O.F., Chardon, W.J., Bechmann, M.E., Gascuelodoux, C., Hofman, G., Kronvang, B., Dorioz, J.M., 2014. Mitigation options to reduce phosphorus losses from the agricultural sector and improve surface water quality: a review. Sci. Total Environ. 468469, 1255–1266.
Shi, M., Yang, Y., Guan, D., Zhang, Y., Zhang, Z., 2012. Bioactivity of the crude polysaccharides from fermented soybean curd residue by Flammulina velutipes. Carbohydr. Polym. 89, 1268–1276. Shi, W., Deng, D.Y., Wang, Y.T., Hu, G.J., Guo, J., Zhang, X.W., Wang, X.R., Giesy, J.P., Yu, H.X., Wang, Z.H., 2016. Causes of endocrine disrupting potencies in surface water in East China. Chemosphere 144, 1435–1442. Shi, P., Zhou, S., Xiao, H., Qiu, J., Li, A., Zhou, Q., Pan, Y., Hollert, H., 2017. Toxicological and chemical insights into representative source and drinking water in eastern China. Environ. Pollut. 233, 35–44. Shuval, Hillel I., 1986. Wastewater Irrigation in Developing Countries. World Bank. Simonovic, S.P., 2002. World water dynamics: global modeling of water resources. J. Environ. Manag. 66, 249–267. Škrbić, B.D., Kadokami, K., Antić, I., 2018. Survey on the micro-pollutants presence in surface water system of northern Serbia and environmental and health risk assessment. Environ. Res. 166, 130–140. Song, Z., Zheng, Z., Li, J., Sun, X., Han, X., Wang, W., Xu, M., 2006. Seasonal and annual performance of a full-scale constructed wetland system for sewage treatment in China. Ecol. Eng. 26, 272–282. Sperling, M.V., 1996. Comparison among the most frequently used systems for wastewater treatment in developing countries. Water Sci. Technol. 33, 59–72. Stamm, C., Räsänen, K., Burdon, F.J., Altermatt, F., Jokela, J., Joss, A., Ackermann, M., Eggen, R.I.L., 2016. Unravelling the impacts of micropollutants in aquatic ecosystems: interdisciplinary studies at the interface of large-scale ecology. Adv. Ecol. Res. 55, 183–223. Sun, J., Ji, X., Zhang, R., Huang, Y., Liang, Y., Du, J., Xie, X., Li, A., 2016. Endocrine disrupting compounds reduction and water quality improvement in reclaimed municipal wastewater: a ﬁeld-scale study along Jialu River in North China. Chemosphere 157, 232–240. Sun, J., Wang, J., Zhang, R., Wei, D., Long, Q., Huang, Y., Xie, X., Li, A., 2017. Comparison of different advanced treatment processes in removing endocrine disruption effects from municipal wastewater secondary efﬂuent. Chemosphere 168, 1–9. The State Environmental Protection Administration of China, 2002. Water and Wastewater Monitoring and Analysis Methods. fourth ed. China Environmental Science Press, Beijing. USEPA, 1995. Method 525.2: Determination of Organic Compounds in Drinking Water by Liquid-Solid Extraction and Capillary Column Gas Chromatography/Mass Spectrometry, Revision 2.0 (Cincinnati, OH: U.S). Vymazal, J., Greenway, M., Tonderski, K., Brix, H., Mander, Ü., 2005. Constructed wetlands for wastewater treatment, in: Verhoeven (Eds.), Ecological Studies. Springer Berlin Heidelberg, Berlin, pp. 69–96. Watson, K., Shaw, G., Leusch, F., Knight, N.L., 2012. Chlorine disinfection by-products in wastewater efﬂuent: bioassay-based assessment of toxicological impact. Water Res. 46, 6069–6083. Wei, Z., Ji, G., 2012. Constructed wetlands, 1991–2011: a review of research development, current trends, and future directions. Sci. Total Environ. 441, 19–27. World Health Organization (WHO), 2011. Guidelines for Drinking-water Quality. 4th ed. World Health Organization, Geneva, Switzerland. Zhang, L., Dong, L., Ren, L., Shi, S., Zhou, L., Zhang, T., Huang, Y.J., 2012. Concentration and source identiﬁcation of polycyclic aromatic hydrocarbons and phthalic acid esters in the surface water of the Yangtze River Delta, China. J. Environ. Sci. 24, 335–342. Zhang, X., Jing, R., Feng, X., Dai, Y., Tao, R., Vymazal, J., Cai, N., Yang, Y., 2018. Removal of acidic pharmaceuticals by small-scale constructed wetlands using different design conﬁgurations. Sci. Total Environ. 639, 640–647. Zhao, J.L., Ying, G.G., Yang, B., Liu, S., Zhou, L.J., Chen, Z.F., Lai, H.J., 2011. Screening of multiple hormonal activities in surface water and sediment from the Pearl River system, South China, using effect-directed in vitro bioassays. Environ. Toxicol. Chem. 30, 2208–2215. Zhe, H., Qing, L., Shuya, S., Wei, Z., Liying, S., 2018. Inactivated Sendai virus strain Tianjin induces apoptosis and autophagy through reactive oxygen species production in osteosarcoma mg-63 cells. J. Cell. Physiol. 234, 4179–4190. Zhou, S., Qiu, J., Liang, Y., Ma, Y., Wang, W., Zhou, Q., Chen, X., Shi, P., Li, A., 2015. Development of a magnetic solid-phase extraction coupled with gas chromatography and mass spectrometry method for the analysis of semivolatile organic compounds. J. Sep. Sci. 38, 3295–3303.