Metagenomic analysis of the biotoxicity of titanium dioxide nanoparticles to microbial nitrogen transformation in constructed wetlands

Journal of Hazardous Materials xxx (xxxx) xxxx

Contents lists available at ScienceDirect

Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat

Metagenomic analysis of the biotoxicity of titanium dioxide nanoparticles to microbial nitrogen transformation in constructed wetlands ⁎

Xiangyu Yanga,b, Yi Chena,b, , Fucheng Guoa,b, Xiaobo Liua,b, Xiaoxuan Suc, Qiang Hea,b a

Key Laboratory of the Three Gorges Region’s Eco-Environment, Ministry of Education, College of Environment and Ecology, Chongqing University, Campus B 83 Shabeijie, Shapingba, Chongqing 400044, China b National Centre for International Research of Low-carbon and Green Buildings, Chongqing University, Chongqing, 400044, China c Key Laboratory of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, 1799 Jimei Road, Xiamen, 361021, China

A R T I C LE I N FO

A B S T R A C T

Editor: R. Debora

Extensive use of titanium dioxide nanoparticles (TiO2 NPs) in various products has increased the release of these particles into wastewater, posing potential environmental risks. As an ecological wastewater treatment facility, constructed wetland (CW) is an important sink of NPs. However, little is known about the effects of NPs on microbial nitrogen transformation and related genes in CWs. In this study, short-term (5 days) and long-term (60 days) exposure experiments were conducted to investigate the effect of TiO2 NPs (0, 1, and 50 mg/L) on microbial nitrogen removal in CWs. The results showed that nitrogen removal efficiency was decreased by 35%–51% after long-term exposure to TiO2 NPs. Metagenomic analysis further confirmed that TiO2 NPs declined the relative abundance of functional genes and those enzyme encoding genes involved in the nitrogen metabolism pathway and glycolysis metabolism process. Furthermore, our data proved that the indigent glycolysis metabolism process resulted in the shortage of electron (NADH) and energy sources (ATP), causing inefficient nitrogen removal. Overall, these results revealed that the accumulation of TiO2 NPs altered the genetic expression of biofilm in CWs, which had significant impacts on biological nitrogen transformation.

Keywords: Constructed wetlands Titanium dioxide nanoparticles (TiO2NPs) Nitrogen transformation Nitrogen metabolism pathway Glycolysis metabolism pathway

1. Introduction Engineered nanomaterials are increasingly used to produce a wide range of industrial and consumer products due to their unique physical and chemical properties (Vance et al., 2015; Kim et al., 2007). In particular, titanium dioxide nanoparticles (TiO2 NPs) are the one of the most common NPs, and generally employed in the production of photocatalyst and sunscreen cream, coatings, and drinking water treatment processes (Robichaud et al., 2009; Kaida et al., 2004). With their increasing usage, TiO2 NPs are inevitably infiltrating into the terrestrial and aquatic environment, causing a series of environmental problems (Johnston et al., 2010; Kocbek et al., 2010; Miller et al., 2010; Adams et al., 2006). Recently, TiO2 NPs have been found in sewage, and have severely permeated into the conventional biological nutrient removal systems (Kaegi et al., 2011; Limbach et al., 2008). While the predicted environmental concentration of TiO2 NPs is at the μg/L level, up to 3 mg/L TiO2 NPs has been detected in wastewater before entering into wastewater treatment plants (WWTPs) (Kiser et al., 2009a) and 27–43 μg/L TiO2 NPs has been noted in effluents (Shi et al., 2016). In particular, the

⁎

NPs concentration may reach as high as dozens of mg per L following swarming of industrial wastewater containing TiO2 NPs into WWTPs (Adams et al., 2006; Cervantes-Aviles et al., 2018). Previous studies have reported that about 91% of TiO2 NPs are trapped in WWTPs (Kiser et al., 2009b), and about 74%–85% of total TiO2 NPs is retained in bioreactors (Shi et al., 2016). Hence, many researchers have tried to explore the effects of TiO2 NPs at mg/L levels on nitrogen removal efficiency. For instance, Zheng et al. (2011a) demonstrated that chronic exposure (70 days) to 50 mg/L TiO2 NPs could evidently decrease the total nitrogen (TN) removal efficiency by 80.3%. Li et al. (Li et al., 2017) found that 1 mg/L TiO2 NPs had no adverse effects on ammonianitrogen (NH4+-N) removal, whereas high TiO2 NPs concentration had significant positive impacts on the denitrification process. The key enzyme activities and microbial community structure in activated sludge have also been examined to explain the adverse effects of TiO2 NPs on nitrogen removal (Li et al., 2017; Yang et al., 2018). Constructed wetlands (CWs), as the biofilm treatment technology, have been widely applied to dispose wastewater containing engineered nanoparticles (Auvinen et al., 2017; Huang et al., 2017). However, after entering into CWs, the NPs could quickly get adsorbed onto the biofilm

Corresponding author. Present address: 174 Shazhengjie Street, Shapingba District, Chongqing 400044, China. E-mail address: [email protected] (Y. Chen).

https://doi.org/10.1016/j.jhazmat.2019.121376 Received 29 August 2019; Received in revised form 30 September 2019; Accepted 30 September 2019 0304-3894/ © 2019 Elsevier B.V. All rights reserved.

Please cite this article as: Xiangyu Yang, et al., Journal of Hazardous Materials, https://doi.org/10.1016/j.jhazmat.2019.121376

Journal of Hazardous Materials xxx (xxxx) xxxx

X. Yang, et al.

powder, with an average particle size of 20 ± 1 nm (BET data), was purchased from Sigma-Aldrich (St. Louis, MO, USA). To prepare 1 and 50 mg/L NPs stock suspension, 12 and 600 mg of TiO2 NPs, respectively, were added into 1 L of artificial wastewater, and the stock suspension was ultrasonicated for 30 min (25 °C, 250 W, 40 kHz) (Zheng et al., 2011a) and mixed with artificial wastewater in a bucket to a volume of up to 12 L. The particle size of approximately 50 TiO2 NPs in stock suspension was determined to be 55 ± 25 nm by transmission electron microscopy (TEM, JEM-1400, JEOL, Tokyo, Japan) operated at 80 kV (Fig. S1).

and plant’s root surface (Liu et al., 2019). Unlike activated sludge, there is no the sludge discharging process in CWs, and as a result, the NPs could be significantly retained and accumulated in CWs. In previous studies, Huang et al. (2017) and Liu et al. (2019) confirmed that the amount of NPs retained in CWs was up to 90%, and found that the presence of Ag NPs changed the microbial community structure and decreased the NH4+-N removal efficiency in the vertical flow CWs. Meanwhile, Hu et al. (2018) and Yang et al. (2018) also respectively observed the impacts of CeO2 NPs and TiO2 NPs on nitrification, and noted that both the NPs inhibited the nitrifying enzyme activities in the subsurface flow (SSF) CWs. It is well-known that microbial nitrogen removal process requires ATP and electrons (NADH) produced by degrading organic matter to transform essential ions through the phospholipid bilayers. Microbial nitrification needs to consume ATP produced by degrading extracellular organic matters to transfer NH4+-N (Kartal et al., 2013), and microbial denitrification needs to consume NADH to accomplish the reduction of NO3−-N or NO2−-N to nitrogen (N2) by four key denitrifying enzymes (NO3− reductase (NAR), NO2− reductase (NIR), NO reductase (NOR) and N2O reductase (NOS)) (Wan et al., 2016; Pan et al., 2013). Hence, ATP and NADH behaviors, including production and consumption, play a vital role in nitrogen conversion process. Previous studies have revealed that the ATP and NADH generated through microbial glycolysis metabolism (Li et al., 2016; Zumft, 1997), and the key enzyme activities involved in glycolysis significantly affected ATP and NADH concentration (Su et al., 2019a; Su et al., 2015; Zheng et al., 2011b). However, the influence of NPs on glycolysis has not yet been reported, and there is a lack of studies combining nitrogen and glycolysis metabolism at gene level to reveal the mechanism of NPs impact on CWs performance. Therefore, a comprehensive evaluation of the effects of TiO2 NPs on glycolysis process and nitrogen transformation should be conducted in CWs. With regard to the effects of NPs on nitrogen transformation, previous studies have mainly focused on specific nitrogen removal processes (e.g., nitrification, denitrification, etc.). Although conventional molecular methods (e.g., enzyme activity (Li et al., 2017), PCR (Tan et al., 2015), denaturing gradient gel electrophoresis (DGGE) (Zheng et al., 2011a), etc.) have been used to determine the bacterial species and functional genes responsible for nitrogen removal, they failed to provide a comprehensive insight into the structural and functional variations in microbial nitrogen transformation. Based on next-generation sequencing techniques, metagenomic analysis, including amplicon and shotgun sequencing, presents sufficient sequencing depth and high accuracy to cover complex microbial nitrogen transformation in a micro-ecology system (Zhang et al., 2012; Zhou et al., 2013). This method is ideal to evaluate the response of each step of the nitrogen transformation and glycolysis processes to TiO2 NPs stress. Therefore, in the present study, SSF CWs were selected to investigate the impacts of TiO2 NPs on nitrogen transformation by metagenomic analysis. Given the concentration of TiO2 NPs in the environment and wastewater (Kiser et al., 2009a; Shi et al., 2016), we selected 1 mg/L as the conventional TiO2 NPs concentration in wastewater and 50 mg/L as the industrial effluent TiO2 NPs concentration to investigate the (a) distribution of TiO2 NPs and (b) influences of TiO2 NPs on functional genes and enzyme encoding genes abundances in each step of the entire nitrogen transformation and glycolysis processes under TiO2 NPs stress. Finally, by combining the responses of functional genes and enzyme encoding genes abundance in nitrogen transformation and glycolysis process, the effects of TiO2 NPs on nitrogen transformation in CWs were comprehensively explained.

2.2. Artificial wastewater Artificial wastewater simulated the WWTPs influent, and comprised the following (per 100 L): 264 mL of concentrated feed, 72.6 mL of phosphorus stock solution, 10 g of ammonium chloride, 2 mL of trace element feed, and 10 mL of acetic acid (Supplementary Information (SI) Text S1). The characteristics of the artificial wastewater were as follows: chemical oxygen demand (COD), 200 ± 5 mg/L; TN, 45 ± 0.5 mg/L; NH4+-N, 35 ± 0.5 mg/L; total phosphorus (TP), 10 ± 0.5 mg/L; and no NO3−-N and NO2−-N (Yang et al., 2018). The initial pH of artificial wastewater was adjusted to 7.5 by injecting 4 M NaOH or 4 M HCl solution. 2.3. Laboratory-scale reactors setup and operation In this study, nine sequencing batch SSF CW microcosms (three sets × three replicates; each with a length of 0.3 m, width of 0.3 m, height of 0.5 m, and pore volume of 12 L) were constructed. All the microcosms were filled with gravel (Φ: 8–10 mm, porosity: 0.4) up to a height of 40 cm and planted with Phragmites australis. Three types of reactors were established as follows: no feeding TiO2 NPs control (Ti0), 1 mg/L TiO2 NPs feeding (Ti1), and 50 mg/L TiO2 NPs feeding (Ti50). The reactor design has been described in our previous study (Yang et al., 2018). Each reactor was worked in batch mode, and the hydraulic retention time (HRT) is 5 days. Artificial wastewater was fed into the reactor via the pipe and drained via a tap on the bottom, and the water level was attuned to be 5 cm below the gravel bed surface. The additional details of the culture and operation of CWs are provided in the SI Text S2. The nine microcosms were placed in an air-conditioned greenhouse at 25 C ± 1 °C and applied natural illumination. The formal batch experiments commenced after a 4-month pre-incubation, and the variations in dissolved oxygen (DO) concentration and pH of the artificial wastewater in CWs are provided in Fig. S2. During the short-term exposure experiment and the last batch of long-term exposure experiments, water samples were collected at intervals from the center of a water sampling pipe using a 100-mL syringe and added into 100-mL serum bottles. To ensure the accuracy of results, the water level at the beginning and at the end of each batch was recorded, which was taken into account the deviation caused by evaporation when calculating the effluent concentration. The water quality was determined within 1 h to ensure reliable results (Yang et al., 2018). 2.4. Analytical methods The analysis of water quality, including COD, TN, NH4+-N, NO3−N, and NO2−-N, was conducted according to Standard Methods (APHA, 1998). To determine the distribution of TiO2 NPs on the microcosm, water samples and plant tissues were digested and examined by inductively coupled plasma-optical mass spectrometry (ICP-MS, iCAP6300-Duo, Thermo Fisher Scientific, Waltham, MA, USA) to confirm the Ti concentration, as indicated in our previous study (Yang et al., 2018). The Ti concentration in the substrate was calculated as described in SI Text S3. After the exposure experiment, approximately 220 g of gravel were sampled from a 25-cm depth of each microcosm. A total of 10 g of

2. Materials and methods 2.1. NPs suspension Commercially available P25 TiO2 NPs (purity ≥ 99.5%, anatase) 2

Journal of Hazardous Materials xxx (xxxx) xxxx

X. Yang, et al.

glycolysis pathways were separated from “KEGG viewer” module.

gravel were used for observation under scanning electron microscopyenergy-dispersive X-ray spectroscopy (SEM-EDS) (Hitachi, Japan. equipped with an energy dispersive X-ray spectrometer) (Liu et al., 2019) and TEM (H-7650, Hitachi, Japan) images (Mao et al., 2018), and 200 g of gravel were employed for metagenomic analysis at Majorbio Bio-Pharm Technology Co., Ltd. (Shanghai, China).

2.6. NADH and ATP assays After the long-term exposure experiment, NADH was detected using the enzymatic cycling assay (San et al., 2002). In brief, 10 g of the gravel samples were put into 50-mL conical flasks and washed thrice with 100 mM PBS (pH = 7.8). Then, 10 mL of 0.2 M NaOH were added to the conical flasks, and the flasks were placed in a water bath at 50 °C for 10 min and in ice water to cool to 4 °C. The samples were neutralized by adding 0.2 M HCl, and after centrifugation at 16,000 rpm for 5 min, the supernatants were immediately collected to measure the NADH content (details about the process are presented in SI Text S1). The ATP level was measured using the ATP Assay Kit (Nanjing Jiancheng Biological Engineering, China), according to the manufacturer’s instructions (SI Text S2).

2.5. Metagenomic analysis 2.5.1. Biofilm sample collection and DNA extraction A total of 200 g of gravel sampled from the middle depth of microcosms in Ti0 and Ti50 were suspended in 100 mM phosphate-buffered saline (PBS) (pH = 7.8) and shaken at 200 rpm for 3 h in a conical flask with stopper to remove the attached biofilm (Chen et al., 2016). The suspended liquid was centrifuged and cleaned twice at 5000 rpm for 20 min. Finally, the biomass was divided into two parts, and one part was stored at −20 °C for total DNA extraction and the other was stored at 4 °C to detect NADH and ATP. The DNA was extracted using the E.Z.N.A.® Soil DNA Kit (Omega Bio-tek, Norcross, GA, USA) following the manufacturer’s instructions. The concentration and purity of DNA was quantified with a TBS-380 (Turner BioSystems, USA) and NanoDrop-2000 (Thermo Fisher Scientific, MA, USA), respectively. The quality of DNA was monitored by 1% agarose gel electrophoresis (Fig. S3).

2.7. Statistical analysis All experiments were conducted in triplicate, and the results are presented as the mean ± standard deviation (SD). The statistical differences of the results were confirmed by analysis of variance (ANOVA), and a p < 0.05 indicated statistical significance. 3. Results and discussion

2.5.2. Library preparation and sequencing For paired-end library construction, the DNA extract was split to an average length of about 300 bp using the Covaris M220 (Gene Company Limited, China). The paired-end library was constructed using the TruSeq™ DNA Sample Prep Kit (Illumina, San Diego, CA, USA). Adapters comprising the full complement of sequencing primer hybridization sites were ligated to the blunt-end fragments. Then, pairedend sequencing was performed on the Illumina HiSeq4000 platform (Illumina Inc., San Diego, CA, USA) using the HiSeq 3000/4000 PE Cluster Kit and HiSeq 3000/4000 SBS Kit, according to the manufacturer’s instructions. The adapter sequences were stripped from the 3′ and 5′ ends of paired-end Illumina reads using SeqPrep. To ensure reliability of the gene data, low-quality reads (length < 50 bp or with a quality value of < 20 or having N bases) were removed using Sickle. After sequence quality control, clean reads of the genome dataset were assembled into contigs using the de Bruijn graph-based assembler SOAPdenovo Version 1.06 (Liu et al., 2019). Then, the contigs were uploaded and used for further gene prediction and annotation analysis.

3.1. Impact of TiO2 NPs on nitrogen removal efficiency of CWs Initially, the nitrogen removal efficiency of the first batch was examined (short-term experiment) (Fig. 1). The average TN removal efficiencies of Ti0, Ti1, and Ti50 were 74.05%, 75.26%, and 74.90%, respectively, while those for NH4+-N were 65.17%, 66.01%, and 65.80%, respectively. The variation in the concentration of NO3−-N and NO2−-N in Ti1 and Ti50 was similar to that in Ti0. The results were not statistically different, suggesting that short-term exposure to 1 and 50 mg/L TiO2 NPs had no acute impacts on biological nitrogen removal in CWs. The results might be because that 5-day exposure to TiO2 NPs has no damages to microbial activity (i.e., key enzyme activities, metabolism processes), and then no acute impacts on biological nitrogen removal was observed. Similar findings have also been reported. Zheng et al. (2011a) found that 1-day exposure to 1 and 50 mg/L TiO2 NPs had no acute effects on TN removal in activated sludge, and Li et al. (2014) also observed that the nitrification efficiency of sequencing batch reactors (SBR) was not affected by exposure to 2–200 mg/L TiO2 NPs for 7 days. In general, pH and conductivity did not vary significantly in all the CWs throughout the entire experiment. Under long-term (60 days) exposure to TiO2 NPs, numerous TiO2 NPs (accounting for 83.80%–86.58%, Table 1) intercepted into the CWs. To further investigate the impacts of TiO2 NPs on nitrogen removal in CWs, a batch experiment (120 h) was conducted after 60-day TiO2 NPs feeding (Fig. 1). Exposure to 50 mg/L TiO2 NPs caused severe inhibition in TN and NH4+-N removal, with average TN and NH4+-N removal efficiencies of 60.71% and 54.00%, respectively, which were significantly lower than those observed in the control (Ti0) (74.05% and 63.20%, respectively) (p < 0.05). In addition, NO2−-N and NO3−-N production and consumption rates (indicated by the slope of lines) in Ti1 and Ti50 were lower than those in Ti0, which resulted in the accumulation of NO2−-N and NO3−-N. Furthermore, the removal kinetics of TN and NH4+-N was established and followed first-order kinetics. The highest removal rate constant (k) was noted in Ti0, followed by Ti50 and Ti1. These results revealed that long-term exposure to TiO2 NPs influenced both nitrification and denitrification in CWs, which are in agreement with those reported in some previous studies. Zheng et al. (2011a) demonstrated that 70-day exposure to 50 mg/L TiO2 NPs decreased the NH4+-N removal efficiency of SBR-containing activated sludge, and Li

2.5.3. Gene prediction, taxonomy, and functional annotation Open reading frames (ORFs) from each assembled contig were predicted using MetaGene (Noguchi et al., 2006). The predicted ORFs with length of ≥ 100 bp were retrieved and translated into amino acid sequences using the NCBI translation table. All the predicted genes with 95% sequence identity (90% coverage) were clustered using CD-HIT (Fu et al., 2012), and the longest sequences from each cluster were selected as representative sequences to construct a non-redundant gene catalog. After quality control, the reads were mapped to the representative sequences with 95% identity using SOAP aligner (Li et al., 2008), and the gene abundance in each sample was evaluated. Representative sequences of the non-redundant gene catalog were aligned to the NCBI NR database with e-value threshold of 1e−5 using BLASTP (Version 2.2.28+, http://blast.ncbi.nlm.nih.gov/Blast.cgi) (Altschul et al., 1997) for taxonomic annotations. The Kyoto Encyclopedia of Genes and Genomes (KEGG) annotation was conducted using BLASTP (Version2.2.28+) against the KEGG database (Xie et al., 2011) with an e-value cutoff of 1e−5. The sequences of each sample were mapped to KEGG pathways by importing the BLAST results into MEGAN (Huson et al., 2007), and the “KEGG viewer” module in MEGAN was used to analyze pathways. The nitrogen metabolism and 3

Journal of Hazardous Materials xxx (xxxx) xxxx

X. Yang, et al.

Fig. 1. Variations in TN, NH4+-N, NO2−-N, and NO3−-N during short-term (empty symbol) and the last batch of long-term (fill symbol) exposure to different concentrations of TiO2 NPs. “P > C” and “P < C” indicate the degree of production beyond and below the consumption of nitrate or nitrite, respectively. All the standard deviations of triplicate measurements are less than 10%.

adsorption capacity and are more sensitive to TiO2 NPs. Moreover, Qiang et al. (Qian et al., 2017) noted that exposure to TiO2 NPs, resulting in high amount of extracellular polymeric substances (EPS) secretion by microorganisms, could block nitrogen ion exchange channels

et al. (2017), by performing microbial enzyme activities analysis in SBR, found that the presence of TiO2 NPs inhibited nitrogen removal. Overall, when compared with activated sludge system, CW microcosms with a combination of biofilm-coated gravel and plants have strong NPs 4

Journal of Hazardous Materials xxx (xxxx) xxxx

X. Yang, et al.

Table 1 Proportions of Ti in each part of CW microcosms exposed to 1 and 50 mg/L TiO2 NPs for 60 days, including substrates, plants, and effluents, respectively. Data are presented as the means ± SD, n = 3. Material Position

Substrate Effluent Plant

TiO2 NPs Ti1 (%)

Ti50 (%)

83.80% ± 1.33% 14.64% ± 1.20% 1.56% ± 0.13%

86.56% ± 1.19% 12.14% ± 1.12% 1.31% ± 0.07%

(e.g., NO2−-N and NO3−-N), which might be responsible for inefficient nitrogen transformation.

Fig. 3. Nitrogen removal functional annotation of microbial DNA sequences in the biofilm (KEGG module classification of top BLASTP hit; e-value ≤ 10−5).

3.2. Comparative metagenomic analysis

abundance of nitrogen fixation (M00175) (N2 to NH4+-N) and dissimilatory and assimilatory nitrate reduction (M00530 and M00531) (NO3−-N to NH4+-N) modules increased under long-term 50 mg/L TiO2 NPs stress. All these results indicated that the presence of TiO2 NPs in the intracellular region produced a negative disturbance to the main nitrogen conversion processes and subsequently affected nitrogen transformation.

Microbial nitrogen transformation is a complex interaction among various microorganisms catalyzing different conversion processes (Tang et al., 2018). The core nitrogen transformation involves four reduction pathways (nitrogen fixation: M00175, assimilatory nitrate reduction: M00531, dissimilatory nitrate reduction: M00530, and denitrification: M00529) and two oxidation pathways (nitrification: M00528 and anaerobic ammonium oxidation (anammox)). Following long-term exposure to TiO2 NPs, the NPs were found in the intracellular region (Fig. 2). Functional annotations of nitrogen conversion processes were assigned to the KEGG pathway database to analyze the change in microbial internal metabolism. As indicated in Fig. 3, the gene abundance of nitrification (M00528) and denitrification (M00529) modules declined from 2.02% and 54.64% in Ti0 to 1.18% (p < 0.05) and 53.29% (p > 0.05) in Ti50, respectively. However, the gene

3.2.1. Alterations in nitrogen metabolism Metagenomic analysis provided a detailed classification and quantification of the metabolic processes. Figs. 4 and 5 present the schematic diagrams of the canonical microbial metabolism (nitrogen and glycolysis metabolism pathways) and related gene abundances according to KEGG and KO database under the same BLAST search conditions (evalue < 10−5).

Fig. 2. SEM images of biofilm (A, a) and TEM images of microorganisms (B, b) on gravel after exposure to 0 (A, B) and 50 mg/L (a, b) TiO2 NPs for 60 days. The dark spots pointed by red arrows are TiO2 NPs. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article). 5

Journal of Hazardous Materials xxx (xxxx) xxxx

X. Yang, et al.

Fig. 4. Relative abundances of key enzyme classifications (identified by their EC number) for each step in the nitrogen pathway (A) and relative abundances of functional genes in the nitrogen pathway (B) in the KEGG Orthology database (e-value < 10−5) following 60 days of exposure to 0 and 50 mg/L TiO2 NPs in CWs.

Responses of functional genes in nitrogen transformation As shown in Fig. 4B, the functional genes of nitrogen transformation, including nitrification (amoABC, hao, narG, narH), denitrification (narI, napAB, nirKS, norBC, nosZ), dissimilatory nitrate reduction (nrfAH, nirBD), assimilatory nitrite reduction (NR, nirA, narB), and nitrogen fixation (nifDKH), were quantified. For nitrification, we quantified the expression of ammonia monooxygenase (AMO) genes (amoABC), catalyzing oxidation of NH3-N to hydroxylamine (NH2OH), and hydroxylamine dehydrogenase (HAO) gene (hao), catalyzing oxidation of NH2OH to NO2−-N (Levy-Booth et al., 2014). Our data consistently affirmed that TiO2 NPs feeding obviously decreased amoABC abundance by 50% in Ti50 (Fig. 4B), while the hao abundance in Ti50 was also less than 50% of that noted in Ti0. With respect to the denitrification process, the abundance of functional genes (napB, nirK, norB, and norC) (Levy-Booth et al., 2014) in Ti50 declined by 4.0%, 14.8%, 2.8%, and 14.1% of that observed in the control, respectively, whereas no negative effects were found on narI, napA, nirS, and nosZ (Fig. 4B). In addition, the abundance of nirA and NR declined by 46.5% and 49.3% (p < 0.05), respectively, in Ti50. Conversely, the genes involved in the dissimilatory nitrate reduction (nirBD and nrfAH) and nitrogen fixation (nifDHK) processes were slightly increased in Ti50 (Fig. 4B). Numerous similar studies had been performed with other engineered nanomaterials, such as Ag NPs, ZnO NPs, and CuO NPs. For instance, Liu et al. (2019) applied metagenomic analysis to show that Ag NPs feeding significantly decreased the relative abundances of almost all the functional genes involved in nitrogen cycling in CWs, especially the nitrification genes (amoABC and hao). Similarly, Gruen

et al. (Gruen et al., 2018) also found the strongest decline in nitrification genes (amoA) and nitrogen fixation genes (nifH) following Ag NPs treatment for 1 year in loamy soil, which ultimately affected nitrogen cycling. Liu et al. (2018) reported that denitrification was inhibited by the decrease in nirK abundance following exposure to Ag NPs in sediments. Zheng et al. (2014) only focused on the impacts of ZnO NPs on denitrification and demonstrated that 50 mg/L ZnO NPs substantially downregulated all denitrifying genes expressions and altered the denitrification process. Conversely, Reyes et al. (2018) reported that 10- and 100-day exposure to 100 mg/L CuO NPs caused significant increase in nirB and norB abundance, and had no negative effects on the number of nitrification genes in wetland. Although the findings of these studies are in agreement with those obtained in the present study, the impact mechanisms of Ag NPs, CuO NPs, and ZnO NPs generally rely on the release of Ag+, Cu2+, and Zn2+ from the NPs, whereas that of TiO2 NPs depends on the NPs owing to the low solubility of Ti. Responses of key enzymes in nitrogen transformation It has been reported that nitrogen transformation in WWTPs is closely related to various genes abundances (Zheng et al., 2014; Chen et al., 2012). For example, amoA and hao are the essential genes encoding AMO and HAO to catalyze the conversion of NH3 to NO2−-N (Yang et al., 2013); The narGI and napAB genes are most often utilized to express cytoplasmic NAR and periplasmic NAP and then accomplish NO3− reduction (Stolz and Basu, 2002; Arnoux et al., 2003; Philippot, 2002); Nitrite is converted to NO or N2O by nitrite reductase, which is generally presumed to be encoded by nirK and nirS (Su et al., 2019a; Zheng et al., 2014), and nitrogenase (NG) is encoded by nifHDK operon (Kneip et al., 2007).

Fig. 5. Relative abundances of key enzyme classifications (identified by their EC number) for each step in the glycolysis process (A) and relative abundances of functional genes in the glycolysis process (B) in KEGG Orthology database (e-value < 10−5) following 60 days of exposure to 0 and 50 mg/L TiO2 NPs in CWs. 6

Journal of Hazardous Materials xxx (xxxx) xxxx

X. Yang, et al.

In the present study, metagenomic analysis classified the abundance of the functional enzymes encoding genes (indicated as “the enzyme abundance”) involved in nitrogen metabolism, and expressed them by their Enzyme Commission (EC) number (Fig. 4A). In particular, the major seven functional enzymes were focused, including AMO (EC 1.14.99.39), HAO (EC 1.7.2.6), nitrate reductase (NAR, EC 1.7.5.1 and EC 1.7.7.2), NIR (EC 1.7.2.1), nitric oxide reductase (NOR, EC 1.7.2.5), nitrous oxide reductase (N2OR, EC 1.7.2.4), and NG (EC 1.18.6.1) (Fig. 4A). Fig. 4A shows the abundances of vital enzymes involved in nitrogen metabolism after exposure to 50 mg/L TiO2 NPs. It can be observed that the relative abundances of majority of these enzymes were reduced, with the exception of NG (EC 1.18.6.1) (Fig. 4A). In other words, exposure to TiO2 NPs evidently reduced the relative abundance of AMO (EC 1.14.99.39) and HAO (EC 1.7.2.6) involved in the nitrification process by 69.64% and 58.25%, respectively. The relative abundance of NAR (EC 1.7.5.1 and EC 1.7.7.2), NIR (EC 1.7.2.1), NOR (EC 1.7.2.5), and N2OR (EC 1.7.2.4) involved in the denitrification process declined by 3.47%, 21.08%, 19.35%, 44.44%, and 23.62%, respectively. The increase in NG (EC 1.18.6.1) abundance was related to the fixation of nitrogen to NH4+-N. Except for the seven main enzymes, the abundances of nitrite oxidoreductase (EC 1.7.99.-) and NIR (NADH) large subunit (EC 1.7.1.15) increased by 13.93% and 32.55%, respectively, when compared with those in Ti0. Similarly, Liu et al. (2019) also reported that the abundances of functional enzymes involved in nitrogen removal, except for that of NG, were always lower than those in the control under long-term Ag NPs stress in CWs, resulting in a decline in nitrogen removal. Although the effects of Ag NPs are similar to those of TiO2 NPs under long-term exposure in CWs, there are no other evidences to document the interaction between NPs and microorganisms, except the SEM images to confirm that NPs can cover the biofilm. In the present study, TiO2 NPs were found in the intracellular region (Fig. 2b), which can explain the variation in microbial internal metabolism. Similarly, previous studies have demonstrated the genotoxicity of TiO2 NPs in animal cells (e.g., in nephrocyte, neurocyte, and Caenorhabditis elegans) using the standard Comet assay, which indicated that functional gene damage mainly resulted from DNA strand breaks and oxidized nucleotides under TiO2 NPs stress (Asare et al., 2016; Rui et al., 2013; Ze et al., 2014). Accordingly, in the present study, we speculated that the presence of TiO2 NPs in the intracellular region produced similar genotoxicity to microbial functional genes, and ultimately decreased the genes abundances and even disrupted the expression of functional genes. Moreover, the variation in functional genes also indicated that nitrification is more susceptible to TiO2 NPs stress than denitrification in CWs. These results are in agreement with the finding that nitrifying bacteria were more sensitive to environmental factors than denitrifying bacteria (Grunditz et al., 1998; LopezFiuza et al., 2002). Comprehensive influence analysis results indicated that the presence of TiO2 NPs significantly disrupted nitrogen metabolism, thereby reducing the nitrogen transformation efficiency of CWs.

Fig. 6. Generation of electron donor (NADH) and energy sources (ATP), and responses of the critical steps in microbial glycolysis metabolism to TiO2 NPs accumulations. Error bars denote the standard deviations of three independent tests. Asterisks (*) indicate significant differences compared to the control (oneway ANOVA, p < 0.05).

figure that TiO2 NPs significantly reduced the NADH contents by 40.8%–53.1% (p < 0.05), while ATP levels declined from 0.0197 μmol/g of gravel in Ti0 to 0.097 μmol/g of gravel in Ti50 (p < 0.05). It is well-known that nitrifiers and denitrifiers require ATP and NADH to complete physiological metabolism (Tiquia et al., 2002; Bhandari et al., 1984). In the present study, TiO2 NPs affected the replication and transcription of genes encoding functional enzymes (pfkBC, PK, PGK, and GAPDH) (Fig. 5B), and inhibited the vital step of glycolysis (Fig. 5A), which might be responsible for inefficient COD removal (Fig. S4). Consequently, the production of ATP and NADH was blocked (Fig. 6), resulting in the decline in microbial activity. Microbial nitrification requires ATP and NADH to transform NH4+N (Kartal et al., 2013), and microbial denitrification needs electrons to sufficiently complete the reduction of nitrate and nitrite to nitrogen (Zumft, 1997; Su et al., 2019a). Thus, the replication and transcription of key genes involved in glycolysis metabolism were inhibited, resulting in a low NH4+-N removal efficiency, and nitrate and nitrite accumulation (Fig. 1). Similar results have also been reported by Zheng et al. (2014) and Su et al. (2019b), who found that feeding of ZnO NPs in activated sludge and chlorothalonil in soil resulted in the blocking of functional genes involved in the production and transformation of ATP and NADH, and a decline in the denitrification rate, respectively. The results of the present study also revealed that low ATP and NADH contents were one of the main reasons for the ineffective nitrogen removal efficiency in Ti50 (Figs. 6 and 1). Overall, this study is the first to confirm that TiO2 NPs can disturb the glycolysis process and decrease the production of ATP and NADH, resulting in an indirect decline in nitrogen transformation efficiency.

3.2.2. Changes in glycolysis pathway Glycolysis is a vital process to supply ATP and NADH for microbial nitrogen transformation (Su et al., 2015). In the present study, the key metabolic enzymes of the glycolysis pathway involved in energy substance (ATP) and electron donor (NADH) transformation were identified after long-term TiO2 NPs exposure. As shown in Fig. 5A and B, under long-term 50 mg/L TiO2 NPs stress, a relatively low abundance of 6-phosphofructose kinase PFK (EC 2.7.1.11) and pyruvate kinase PK (EC 2.7.1.40), reaching 9.75% (p < 0.05) and 1.47%, respectively, was noted. As a pivotal enzyme to catalyze glyceraldehyde-3phosphate to 1,3-biphosphoglycerate and produce NADH for denitrification (Gerald, 1996), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (EC 1.2.1.59) exhibited a decline in its abundance, reaching a value of 28.3% (p < 0.05). The amount of NADH and ATP in response to the decreasing abundance of energy metabolism genes under long-term 50 mg/L TiO2 NPs stress is presented in Fig. 6. It can be noted from the

3.3. Environmental significance CWs, which are usually controlled by microbial groups, are employed to treat rainwater and secondary and tertiary wastewater (Vymazal, 1998), and are the hotspots of NPs. In the present study, TiO2 NPs accumulation was found to negatively affect microbial nitrogen removal, resulting in inefficient nitrogen conversion, which might facilitate eutrophication of the receiving water. The results obtained highlighted the importance of the control of NPs in the ecosystem, which could be achieved through the following three strategies: (1) NPs-containing wastewater from industrial raw material factories must be treated to meet standard criteria before discharge and disposal; (2) 7

Journal of Hazardous Materials xxx (xxxx) xxxx

X. Yang, et al.

the pretreatment system must be positioned in front of the CWs to gather and remove NPs through coagulation and sedimentation, instead of directly discharging wastewater into the wetland system; and (3) culture and improve the ability of defending against NPs and the adaptability to nanotoxicity for the biofilm used in CWs.

Gruen, A.-L., Straskraba, S., Schulz, S., Schloter, M., Emmerling, C., 2018. Long-term effects of environmentally relevant concentrations of silver nanoparticles on microbial biomass, enzyme activity, and functional genes involved in the nitrogen cycle of loamy soil. J. Environ. Sci. China 69, 12–22. Grunditz, C., Gumaelius, L., Dalhammar, G., 1998. Comparison of inhibition assays using nitrogen removing bacteria: application to industrial wastewater. Water Res. 32, 2995–3000. Hu, X., Liu, X., Yang, X., Guo, F., Su, X., Chen, Y., 2018. Acute and chronic responses of macrophyte and microorganisms in constructed wetlands to cerium dioxide nanoparticles: implications for wastewater treatment. Chem. Eng. J. 348, 35–45. Huang, J., Cao, C., Yan, C., Liu, J., Hu, Q., Guan, W., 2017. Impacts of silver nanoparticles on the nutrient removal and functional bacterial community in vertical subsurface flow constructed wetlands. Bioresour. Technol. 243, 1216–1226. Huson, D.H., Auch, A.F., Qi, J., Schuster, S.C., 2007. Megan analysis of metagenomic data. Genome Res. 17, 377–386. Johnston, B.D., Scown, T.M., Moger, J., Cumberland, S.A., Baalousha, M., Linge, K., van Aerle, R., Jarvis, K., Lead, J.R., Tyler, C.R., 2010. Bioavailability of nanoscale metal oxides TiO2, CeO2, and ZnO to fish. Environ. Sci. Technol. 44, 1144–1151. Kaegi, R., Voegelin, A., Sinnet, B., Zuleeg, S., Hagendorfer, H., Burkhardt, M., Siegrist, H., 2011. Behavior of metallic silver nanoparticles in a pilot wastewater treatment plant. Environ. Sci. Technol. 45, 3902–3908. Kaida, T., Kobayashi, K., Adachi, M., Suzuki, F., 2004. Optical characteristics of titanium oxide interference film and the film laminated with oxides and their applications for cosmetics. J. Cosmet. Sci. 55, 219–220. Kartal, B., de Almeida, N.M., Maalcke, W.J., Op den Camp, H.J.M., Jetten, M.S.M., Keltjens, J.T., 2013. How to make a living from anaerobic ammonium oxidation. FEMS Microbiol. Rev. 37, 428–461. Kim, J.S., Kuk, E., Yu, K.N., Kim, J.H., Park, S.J., Lee, H.J., Kim, S.H., Park, Y.K., Park, Y.H., Hwang, C.Y., Kim, Y.K., Lee, Y.S., Jeong, D.H., Cho, M.H., 2007. Antimicrobial effects of silver nanoparticles. Nanomed. Nanotechnol. 3, 95–101. Kiser, M.A., Westerhoff, P., Benn, T., Wang, Y., Perez-Rivera, J., Hristovski, K., 2009a. Titanium nanomaterial removal and release from wastewater treatment plants. Environ. Sci. Technol. 43, 6757–6763. Kiser, M.A., Westerhoff, P., Benn, T., Wang, Y., Perez-Rivera, J., Hristovski, K., 2009b. Titanium nanomaterial removal and release from wastewater treatment plants. Environ. Sci. Technol. 43, 6757–6763. Kneip, C., Lockhart, P., Voss, C., Maier, U.-G., 2007. Nitrogen fixation in eukaryotes - New models for symbiosis. BMC Evol. Biol. 7. Kocbek, P., Teskac, K., Kreft, M.E., Kristl, J., 2010. Toxicological aspects of long-term treatment of keratinocytes with ZnO and TiO2 nanoparticles. Small 6, 1908–1917. Levy-Booth, D.J., Prescott, C.E., Grayston, S.J., 2014. Microbial functional genes involved in nitrogen fixation, nitrification and denitrification in forest ecosystems. Soil Biol. Biochem. 75, 11–25. Li, Z., Wang, X., Ma, B., Wang, S., Zheng, D., She, Z., Guo, L., Zhao, Y., Xu, Q., Jin, C., Li, S., Gao, M., 2017. Long-term impacts of titanium dioxide nanoparticles (TiO2 NPs) on performance and microbial community of activated sludge. Bioresour. Technol. 238, 361–368. Li, M., Su, Y., Chen, Y., Wan, R., Zheng, X., Liu, K., 2016. The effects of fulvic acid on microbial denitrification: promotion of NADH generation, electron transfer, and consumption. Appl. Microbiol. Biotechnol. 100, 5607–5618. Li, R., Li, Y., Kristiansen, K., Wang, J., 2008. SOAP: short oligonucleotide alignment program. Bioinformatics 24, 713–714. Li, D., Cui, F., Zhao, Z., Liu, D., Xu, Y., Li, H., Yang, X., 2014. The impact of titanium dioxide nanoparticles on biological nitrogen removal from wastewater and bacterial community shifts in activated sludge. Biodegradation 25, 167–177. Limbach, L.K., Bereiter, R., Mueller, E., Krebs, R., Gaelli, R., Stark, W.J., 2008. Removal of oxide nanoparticles in a model wastewater treatment plant: influence of agglomeration and surfactants on clearing efficiency. Environ. Sci. Technol. 42, 5828–5833. Liu, X., Yang, X., Hu, X., He, Q., Zhai, J., Chen, Y., Xiong, Q., Vymazal, J., 2019. Comprehensive metagenomic analysis reveals the effects of silver nanoparticles on nitrogen transformation in constructed wetlands. Chem. Eng. J. 358, 1552–1560. Liu, S., Wang, C., Hou, J., Wang, P., Miao, L., Fan, X., You, G., Xu, Y., 2018. Effects of Ag and Ag2S nanoparticles on denitrification in sediments. Water Res. 137, 28–36. Lopez-Fiuza, J., Buys, B., Mosquera-Corral, A., Omil, F., Mendez, R., 2002. Toxic effects exerted on methanogenic, nitrifying and denitrifying bacteria by chemicals used in a milk analysis laboratory. Enzyme Microb. Technol. 31, 976–985. Mao, Y., Ai, H., Chen, Y., Zhang, Z., Zeng, P., Kang, L., Li, W., Gu, W., He, Q., Li, H., 2018. Phytoplankton response to polystyrene microplastics: perspective from an entire growth period. Chemosphere 208, 59–68. Miller, R.J., Lenihan, H.S., Muller, E.B., Tseng, N., Hanna, S.K., Keller, A.A., 2010. Impacts of metal oxide nanoparticles on marine phytoplankton. Environ. Sci. Technol. 44, 7329–7334. Noguchi, H., Park, J., Takagi, T., 2006. MetaGene: prokaryotic gene finding from environmental genome shotgun sequences. Nucl. Acid. Res. 34, 5623–5630. Pan, Y., Ni, B.-J., Bond, P.L., Ye, L., Yuan, Z., 2013. Electron competition among nitrogen oxides reduction during methanol-utilizing denitrification in wastewater treatment. Water Res. 47, 3273–3281. Philippot, L., 2002. Denitrifying genes in bacterial and Archaeal genomes. Biochim. Et Biophys. Acta-Gene Struct. Expression 1577, 355–376. Qian, J., Li, K., Wang, P., Wang, C., Shen, M., Liu, J., Lu, B., Tian, X., 2017. Toxic effects of three crystalline phases of TiO2 nanoparticles on extracellular polymeric substances in freshwater biofilms. Bioresour. Technol. 241, 276–283. Reyes, V.C., Gedalanga, P.B., Merino, N., Van Nostrand, J.D., Keely, S.P., De Long, S.K., Zhou, J., Mahendra, S., 2018. Differential sensitivity of wetland-derived nitrogen cycling microorganisms to copper nanoparticles. ACS Sustain. Chem. Eng. 6, 11642–11652.

4. Conclusion In the present study, short-term exposure to TiO2 NPs had no significant inhibitory effects on the nitrogen removal efficiency of CWs. However, long-term exposure to TiO2 NPs caused accumulation and even infiltration of the NPs into the microbial cell membrane, which disturbed microbial physiological metabolism. Metagenomic analysis further indicated that exposure to TiO2 NPs caused decreases in the abundance of functional genes and genes encoding enzymes involved in nitrogen transformation, as well as ATP and NADH production processes, and fundamentally explained the decline in nitrogen conversion efficiency of CWs fed with TiO2 NPs. Overall, this is the first study of the responses of the nitrogen transformation process to NPs in CWs, combined with evaluation of energy metabolism. Thus, this study provides new insights into the effects of emerging pollutants produced by human activities on the nitrogen cycle in ecological wastewater treatment systems. Acknowledgments This study was supported by the National Natural Science Foundation of China (Grant No.51708056); the National Major Project of Pollution Control and Treatment Science and Technology (Grant No. 2017ZX07401003-4); and Fundamental Research Funds for the Central Universities of Chongqing University (Grant No. 2019CDXYCH0027). We would also like to thank Analytical and Testing Center of Chongqing University for SEM and TEM images. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.jhazmat.2019.121376. References Adams, L.K., Lyon, D.Y., Alvarez, P.J., 2006. Comparative eco-toxicity of nanoscale TiO2, SiO2, and ZnO water suspensions. Water Res. 40, 3527–3532. Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J.H., Zhang, Z., Miller, W., Lipman, D.J., 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucl. Acid. Res. 25, 3389–3402. APHA, 1998. Standard Methods for the Examination of Water and Wastewater. American Public Health Association, Washington, DC. Arnoux, P., Sabaty, M., Alric, J., Frangioni, B., Guigliarelli, B., Adriano, J.M., Pignol, D., 2003. Structural and redox plasticity in the heterodimeric periplasmic nitrate reductase. Nat. Struct. Biol. 10, 928–934. Asare, N., Duale, N., Slagsvold, H.H., Lindeman, B., Olsen, A.K., Gromadzka-Ostrowska, J., Meczynska-Wielgosz, S., Kruszewski, M., Brunborg, G., Instanes, C., 2016. Genotoxicity and gene expression modulation of silver and titanium dioxide nanoparticles in mice. Nanotoxicology 10, 312–321. Auvinen, H., Gagnon, V., Rousseau, D.P.L., Du Laing, G., 2017. Fate of metallic engineered nanomaterials in constructed wetlands: prospection and future research perspectives. Rev. Environ. Sci. Biotechnol. 16, 207–222. Bhandari, B., Naik, M.S., Nicholas, D.J.D., 1984. ATP production coupled to the denitrification of nitrate in rhizobium-japonicum, grown in cultures and in bacteroids from glycine-max. FEBS Lett. 168, 321–326. Cervantes-Aviles, P., Ida, J., Toda, T., Cuevas-Rodriguez, G., 2018. Effects and fate of TiO2 nanoparticles in the anaerobic treatment of wastewater and waste sludge. J. Environ. Manage. 222, 227–233. Chen, Y., Wen, Y., Zhou, Q., Huang, J., Vymazal, J., Kuschk, P., 2016. Sulfate removal and sulfur transformation in constructed wetlands: the roles of filling material and plant biomass. Water Res. 102, 572–581. Chen, Y., Su, Y., Zheng, X., Chen, H., Yang, H., 2012. Alumina nanoparticles-induced effects on wastewater nitrogen and phosphorus removal after short-term and longterm exposure. Water Res. 46, 4379–4386. Fu, L., Niu, B., Zhu, Z., Wu, S., Li, W., 2012. CD-HIT: accelerated for clustering the nextgeneration sequencing data. Bioinformatics 28, 3150–3152. Gerald, K., 1996. Cell and Molecular Biology: Concepts and Experiments (3e).

8

Journal of Hazardous Materials xxx (xxxx) xxxx

X. Yang, et al.

Hull, M.S., 2015. Nanotechnology in the real world: redeveloping the nanomaterial consumer products inventory. Beilstein J. Nanotechnol. 6, 1769–1780. Vymazal, J., 1998. Removal Mechanisms and Types of Constructed Wetlands. Wan, R., Chen, Y., Zheng, X., Su, Y., Li, M., 2016. Effect of CO2 on microbial denitrification via inhibiting Electron transport and consumption. Environ. Sci. Technol. 50, 9915–9922. Xie, C., Mao, X., Huang, J., Ding, Y., Wu, J., Dong, S., Kong, L., Gao, G., Li, C.-Y., Wei, L., 2011. KOBAS 2.0: a web server for annotation and identification of enriched pathways and diseases. Nucl. Acid. Res. 39, W316–W322. Yang, X., Chen, Y., Liu, X., Guo, F., Su, X., He, Q., 2018. Influence of titanium dioxide nanoparticles on functionalities of constructed wetlands for wastewater treatment. Chem. Eng. J. 352, 655–663. Yang, Y., Wang, J., Xiu, Z., Alvarez, P.J.J., 2013. Impacts of silver nanoparticles on cellular and transcriptional activity of nitrogen-cycling bacteria. Environ. Toxicol. Chem. 32, 1488–1494. Ze, Y., Hu, R., Wang, X., Sang, X., Ze, X., Li, B., Su, J., Wang, Y., Guan, N., Zhao, X., Gui, S., Zhu, L., Cheng, Z., Cheng, J., Sheng, L., Sun, Q., Wang, L., Hong, F., 2014. Neurotoxicity and gene-expressed profile in brain-injured mice caused by exposure to titanium dioxide nanoparticles. J. Biomed. Mater. Res. A 102, 470–478. Zhang, T., Shao, M.-F., Ye, L., 2012. 454 Pyrosequencing reveals bacterial diversity of activated sludge from 14 sewage treatment plants. ISME J. 6, 1137–1147. Zheng, X., Su, Y., Chen, Y., Wan, R., Liu, K., Li, M., Yin, D., 2014. Zinc oxide nanoparticles cause inhibition of microbial denitrification by affecting transcriptional regulation and enzyme activity. Environ. Sci. Technol. 48, 13800–13807. Zheng, X., Chen, Y., Wu, R., 2011a. Long-term effects of titanium dioxide nanoparticles on nitrogen and phosphorus removal from wastewater and bacterial community shift in activated sludge. Environ. Sci. Technol. 45, 7284–7290. Zheng, X., Wu, R., Chen, Y., 2011b. Effects of ZnO nanoparticles on wastewater biological nitrogen and phosphorus removal. Environ. Sci. Technol. 45, 2826–2832. Zhou, J., Jiang, Y.-H., Deng, Y., Shi, Z., Zhou, B.Y., Xue, K., Wu, L., He, Z., Yang, Y., 2013. Random sampling process leads to overestimation of beta-diversity of microbial communities. Mbio 4, 00313–00324. Zumft, W.G., 1997. Cell biology and molecular basis of denitrification. Microbiol. Mol. Biol. Rev. 61, 533–616.

Robichaud, C.O., Uyar, A.E., Darby, M.R., Zucker, L.G., Wiesner, M.R., 2009. Estimates of upper bounds and trends in nano-TiO2 production as a basis for exposure assessment. Environ. Sci. Technol. 43, 4227–4233. Rui, Q., Zhao, Y., Wu, Q., Tang, M., Wang, D., 2013. Biosafety assessment of titanium dioxide nanoparticles in acutely exposed nematode Caenorhabditis elegans with mutations of genes required for oxidative stress or stress responset. Chemosphere 93, 2289–2296. San, K.Y., Bennett, G.N., Berrios-Rivera, S.J., Vadali, R.V., Yang, Y.T., Horton, E., Rudolph, F.B., Sariyar, B., Blackwood, K., 2002. Metabolic engineering through cofactor manipulation and its effects on metabolic flux redistribution in Escherichia coli. Metab. Eng. 4, 182–192. Shi, X., Li, Z., Chen, W., Qiang, L., Xia, J., Chen, M., Zhu, L., Alvarez, P.J.J., 2016. Fate of TiO2 nanoparticles entering sewage treatment plants and bioaccumulation in fish in the receiving streams. Nanoimpact 3-4, 96–103. Stolz, J.F., Basu, P., 2002. Evolution of nitrate reductase: molecular and structural variations on a common function. ChemBioChem 3, 198–206. Su, Y., Zheng, X., Chen, A., Chen, Y., He, G., Chen, H., 2015. Hydroxyl functionalization of single-walled carbon nanotubes causes inhibition to the bacterial denitrification process. Chem. Eng. J. 279, 47–55. Su, X., Chen, Y., Wang, Y., Yang, X., He, Q., 2019a. Impacts of chlorothalonil on denitrification and N2O emission in riparian sediments: microbial metabolism mechanism. Water Res. 148, 188–197. Su, X., Chen, Y., Wang, Y., Yang, X., He, Q., 2019b. Disturbances of electron production, transport and utilization caused by chlorothalonil are responsible for the deterioration of soil denitrification. Soil Biol. Biochem. 134, 100–107. Tan, M., Qiu, G., Ting, Y.P., 2015. Effects of ZnO nanoparticles on wastewater treatment and their removal behavior in a membrane bioreactor. Bioresour. Technol. 185, 125–133. Tang, Y., Yu, G., Zhang, X., Wang, Q., Ge, J., Liu, S., 2018. Changes in nitrogen-cycling microbial communities with depth in temperate and subtropical forest soils. Appl. Soil Ecol. 124, 218–228. Tiquia, S.M., Wan, J.H.C., Tam, N.F.Y., 2002. Dynamics of yard trimmings composting as determined by dehydrogenase activity, ATP content, arginine ammonification, and nitrification potential. Process Biochem. 37, 1057–1065. Vance, M.E., Kuiken, T., Vejerano, E.P., McGinnis, S.P., Hochella Jr., M.F., Rejeski, D.,

9