Fate of tetracycline and sulfamethoxazole and their corresponding resistance genes in microbial fuel cell coupled constructed wetlands

Abstract

Pollution by antibiotics and antibiotic resistance genes has become a major health concern. A microbial fuel cell coupled constructed wetland (CW-MFC) is a new installation to simultaneously treat wastewater and produce energy. This study was conducted to assess the removal efficiency of tetracycline (TC) and sulfamethoxazole (SMX) and the development of TC and SMX resistance genes (tetA, tetC, tetO, tetQ, tetW, sulI and sulII) in the CW-MFCs. At the same time, electricity was also produced during the co-metabolism of the antibiotics and glucose. The results indicated that the CW-MFCs could significantly reduce the concentration of TC and SMX in wastewater. Accumulation of antibiotics and the relative abundances of the tet and sul genes in different layers of the CW-MFCs were: anode layer > cathode layer > middle layer > effluent. The relative abundances of ARGs in the substrates ranged from 1.40 × 10⁻⁴ to 6.30 × 10⁻¹ for tet genes and 5.58 × 10⁻³ to 9.27 × 10⁻² for sul genes. The relative abundances of the tetC, tetQ, TetW, sulI and sulII genes were significantly affected by the antibiotics in effluent of the CW-MFCs. In addition, the highest power density obtained under the external resistance of 1000 Ω was 0.0578 W m⁻². However, an obvious drop of power density was observed when the antibiotic concentrations increased from 400 to 1600 μg L⁻¹.

---

1. Introduction

Antibiotics are commonly used medicines in the treatment of human and animal diseases.¹ China leads the world in antibiotic production capacity with approximately 210 000 t of antibiotics manufactured annually, of which 85% is utilized in animal agriculture and medicine.² However, the majority of the antibiotics used by humans and livestock cannot be completely metabolized. Excreted antibiotics are directly discharged into the environment, entering river systems through a variety of pathways, e.g., via point discharges from wastewater treatment plants (WWTPs), animal feeding operations (AFOs), and fish hatcheries or via nonpoint sources, which can cause serious problems in ecosystems.^3,4 It is well known that even low concentrations of antibiotics give rise to a significant risk to environmental security and human health.⁵ An increase in the occurrence of bacterial indicators, antibiotic resistant bacteria (ARB) and resistance genes (ARGs) via overuse and misuse of antibiotics was monitored in wastewater treatment plants, raising major health concerns. In addition, the potential of vertical and horizontal gene transfer can promote ARGs transfer between nonpathogenic and pathogenic bacteria.^3,4,6 Therefore, antibiotics, ARBs and ARGs are regarded as the three worldwide public health problems by the World Health Organization.^3,7

A microbial fuel cell coupled constructed wetland (CW-MFC) system is a new technology that embeds the microbial fuel cell (MFC) into a constructed wetland (CW) to treat wastewater and produce bioelectricity. This technique has attracted considerable interest for its ease of construction and operation as compared to traditional MFCs.⁸ Cathode performance of the embedded MFCs can be promoted using wetland plants.⁹ Previous studies of CW-MFCs have focused on the removal of dye and nutrients in municipal domestic wastewater. Liu et al. (2013) developed CW-MFCs to convert solar energy into electricity by utilizing root exudates of Ipomoea aquatica.¹⁰ Fang et al. (2015) tested CW-MFCs for azo dye removal efficiency and power density.¹¹ Doherty et al. (2015) explored the effects of electrode spacing and flow direction of the wastewater in CW-MFCs for investigating the potential for simultaneous power generation and wastewater treatment.¹²

In recent years, concern about veterinary antibiotics (VAs) and ARGs as potential pollutants has grown significantly after confirmation of their high concentrations in manure, wastewater and soil around livestock farms.^13,14 The target VAs included tetracycline (TC) and sulfamethoxazole (SMX), which have been extensively used in the livestock industry and are always detected in environmental samples.^15,16 Five tet genes (tetA, tetC, tetO, tetQ and tetW) were chosen as representative of three tetracycline resistance mechanisms because of their frequent occurrence in livestock manure and wastewater treatment lagoons.^4,16,17 Two sul genes (sulI, sulII) were chosen as representative of sulfonamide resistance mechanisms for the same reason.^13,16,18,19

CW-MFCs possess a large effective area to drive the reductive degradation of toxic and refractory compounds to produce bioelectricity and have recently attracted attention for wastewater treatment. However, less attention has been paid to the potential for removal of trace antibiotics from domestic sewage using CW-MFCs. In this study, the performance of CW-MFCs was evaluated for its effect on removal of antibiotics TC and SMX and their corresponding antibiotic resistant genes from domestic sewage over a period of six months. The main objectives of this work were to (1) study the removal efficiencies of CW-MFCs on TC and SMX from the domestic sewage, (2) explore the possible removal mechanisms, and (3) assess the risk of TC and SMX accumulation in the substrate on electricity production and ARG development.

2. Materials and methods

2.1. Experimental design of CW-MFCs

Three CW-MFCs were constructed using polyacrylic plastic chambers (the internal diameter was 20 cm and the height was 55 cm) containing four layers from bottom to top: the bottom layer (20 cm high) was filled with soil and gravel (volume ratio of sand/soil = 40 : 1, 2–3 mm diameter of gravel), the anode layer (8 cm high) was made of granular-activated carbon (GAC, 3–5 mm diameter and 500–900 m² g⁻¹ specific area), the middle layer (20 cm high) was filled with soil and gravel (volume ratio of sand/soil = 40 : 1, 2–3 mm diameter of gravel), the cathode layer (6 cm high) was made of GAC (3–5 mm diameter and 500–900 m² g⁻¹ specific area) and the same Oenanthe javanica was transplanted into each cathode layer (Fig. 1). The sampling ports were arranged at the same intervals to collect samples from the bottom to the top: anode layer, middle layer 1, middle layer 2 and cathode layer. A titanium wire (1.0 mm diameter) was used to connect the external circuit with an external resistance of 1000 Ω. Epoxy was used to seal metals from exposure to the solution.¹¹

Fig. 1 The structure of the CW-MFC (1 computer; 2 data acquisition module; 3 resistance; 4 Oenanthe javanica; 5 cathode; 6 anode; 7 peristaltic pump; 8 tank).

2.2. Inoculation and operation of CW-MFCs

Anaerobic sludge (mixed liquor suspended solids (MLSS): 60 g L⁻¹) was obtained from the South City Municipal Wastewater Treatment Plant of Nanjing, China. The sludge was pretreated by mixing with GAC (sludge volume ratio of 3 : 1) and introduced into CW-MFCs for microbial inoculation. Synthetic domestic wastewater was compiled that contained glucose (0.225 g L⁻¹), KH₂PO₄ (0.004 g L⁻¹), NH₄Cl (0.020 g L⁻¹), NaCl (0.15 g L⁻¹) and 0.20 mL of a trace essential elements solution (contained per liter: 15 g MgSO₄·7H₂O, 2.2 g MnSO₄·H₂O, 5 mg ZnCl₂, 2 g ZnSO₄·7H₂O, 15 g CaCl₂, 1 g FeSO₄, 0.24 g CoCl₂·6H₂O, 10 mg FeCl₃·6H₂O, 2 mg NiCl₂·6H₂O, 0.4 mg Na₂MoO₄·2H₂O and 1 mg CuCl₂·2H₂O).²⁰

After the inoculations were complete, the nutrient solution was fed continuously into the CW-MFCs using peristaltic pumps (BT100-1L, Baoding Longer Precision Pump Co., Ltd., China). The concentrations of TC and SMX in the three CW-MFCs were 200, 500 and 800 μg L⁻¹. The hydraulic retention time was 2.5 d. Experiments were conducted in the dark at 28 ± 2 °C continuously for six months from April to September 2015. All the cycles were repeated 3 times and error bars were used to represent standard errors of each parallel experiment.

2.3. Quantification of antibiotics

Effluents were collected in triplicate at seven sampling periods: 1^st May, 16^th May, 31^st May, 1^st August, 16^th August, 31^st August, and 15^th September 2015. The antibiotics in the effluents and media (soil and GAC) were detected by LC-MS analysis (LCQAD-60000, Thermo Fisher Scientific, Waltham, MA, USA). Water samples were filtered through 0.45 μm fiber filters and extracted using Oasis HLB (6 mL, Waters, USA) extraction cartridges (Huang et al., 2013). Antibiotics analysis in the samples was based on the published method.^16,21,22

2.4. Quantification of ARGs

Samples of anode, middle layer, cathode layer and effluents were collected in triplicate on 15 September 2015. Genomic DNA was extracted using a Power Soil DNA isolation kit (MoBio, Carlsbad, CA, USA). DNA concentration and quality were examined by a UV-9100 spectrophotometer (Lab Tech Ltd, Beijing, China) and gel electrophoresis. Seven tet genes (tetA, tetC, tetO, tetQ, tetW, sulI and sulII) were quantified using CFX Connect Real-Time PCR System (Bio-Rad, Shanghai, China) and SYBR Green qPCR mix (Bio-Rad, Shanghai, China). Primer sequences targeting these genes and PCR protocol were based on the previous methods.^4,15,16 Each reaction was conducted in 25 μL on 0.1 mL qPCR strip tubes kit (Thermo Fisher Scientific, Taiwan) containing 12 μL SYBR Green qPCR Mix (Bio-Rad, Shanghai, China), primers (forward and reverse primers, 0.6 μL of 10 μM each), 0.8 μL DNA templates and 11 μL ddH₂O.^16,23 All PCR reactions were conducted in triplicate. Plasmids carrying the target ARGs were created (Sangon Biotech, Shanghai, China) to generate standard curves (R² > 0.990).

2.5. Instrumentation and analyses

A data acquisition module (DAM-3057, Art Technology Co. Ltd., China) was used to monitor the voltage (V) data. The volumetric power density (P_v, W m⁻²) was measured by the effective volume of the anode chamber (m²) for the electrons used for power generation in anode.¹¹ The internal resistance of the CW-MFCs was estimated by polarization curve based on the published methods.¹¹

2.6. Statistical analysis

The averages and standard deviations of all the data were calculated by Microsoft Excel 2010. Statistical analyses were carried out using SPSS Version 19.0, and all the figures were plotted by Sigmaplot 11.0. Student–Newman–Keuls tests (p < 0.05) were used to compare treatment means.

3. Results and discussion

3.1. Removal efficiencies of TC and SMX in CW-MFCs

The target antibiotic concentrations in the effluent of the three CW-MFCs are shown in Table 1. During an operation period of more than six months, the average effluent concentrations ranged from 0 to 0.66, 0 to 1.1 and 0 to 1.65 μg L⁻¹ for TC in the three CW-MFCs at total influent antibiotic concentrations of 400, 1000 and 1600 μg L⁻¹ respectively. Moreover, the average effluent concentrations ranged from 0 to 0.90, 0 to 1.70 and 0 to 2.40 μg L⁻¹ for SMX at total influent antibiotic concentrations of 400, 1000 and 1600 μg L⁻¹ in CW-MFCs respectively. Compared with SMX, TC showed better removal efficiency in all CW-MFCs; however, no statistically significant differences (p < 0.05) were found. Among various types of CWs most frequently used for pharmaceutical removal,⁴ those used in the current study showed comparable removal efficiencies.

Table 1 Concentrations of antibiotics TC and SMX in the effluent water of CW-MFCs (initial concentrations, 200, 500 and 800 μg L⁻¹, mean ± SD, n = 3)

Time	Effluent (μg L^−1)
	TC (200)	SMX (200)	TC (500)	SMX (500)	TC (800)	SMX (800)
1 May	—	—	—	—	—	—
16 May	—	0.14 ± 0.05	—	—	0.21 ± 0.03	0.26 ± 0.03
31 May	—	—	0.23 ± 0.03	0.56 ± 0.05	0.43 ± 0.05	0.90 ± 0.08
1 Aug	0.66 ± 0.07	0.90 ± 0.10	1.11 ± 0.12	1.17 ± 0.12	1.65 ± 0.25	2.29 ± 0.15
16 Aug	0.53 ± 0.08	0.88 ± 0.11	0.82 ± 0.11	1.70 ± 0.25	1.25 ± 0.11	2.40 ± 0.21
31 Aug	0.51 ± 0.09	0.48 ± 0.05	0.57 ± 0.05	1.08 ± 0.11	0.53 ± 0.04	0.41 ± 0.04
15 Sep	0.52 ± 0.06	0.18 ± 0.02	0.54 ± 0.06	0.29 ± 0.04	0.64 ± 0.06	0.87 ± 0.09

---

The accumulation of antibiotics in the different layer of the CW-MFCs was analyzed in Fig. 2. Clearly, the concentration of TC in anode layer was higher, followed by the cathode layer, middle layer; the concentration of SMX had the same trend (p < 0.05). A strong performance of antibiotics adsorption onto activated carbon was reported at several studies.^24,25 The difference in accumulation of TC and SMX in the different layers was mainly caused by their adsorption characteristics on GAC. In addition, the difference in accumulation of TC and SMX in the anode layers and cathode layers was considered of statistical significant (p < 0.05) at different influent concentrations. This implies that the accumulation is mainly affected by the influent antibiotic content. A series of biological and physicochemical processes, including substrate absorption, microbial degradation, plant uptake, hydrolysis and electrochemistry, may be involved in antibiotics removal in CW-MFCs.^4,26

Fig. 2 Accumulation of antibiotics in the different layer of CW-MFCs. Data are mean ± SD (n = 3). Bars with different letters are significantly different at p < 0.05 according to Duncan's multiple tests.

3.2. Typical tet and sul genes in the samples

The relative abundances of tet and sul genes in the anode layer, middle layer 1, middle layer 2, cathode layer and effluents of the reactors are shown in Fig. 3. As with SMX and TC, the level of the corresponding ARGs in the substrates were detected and ranged from 1.40 × 10⁻⁴ to 6.30 × 10⁻¹ for tet genes and 5.58 × 10⁻³ to 9.27 × 10⁻² for sul genes.

Fig. 3 Quantities of tet and sul genes (tetA, tetC, tetO, tetQ, tetW, sulI and sulII) normalized to 16S rRNA genes in the cathode, middle, anode layer and effluents of CW-MFCs, respectively. Data are mean ± SD (n = 3). Bars with different letters are significantly different at p < 0.05 according to Duncan's multiple tests.

A greater relative abundance of ARGs is commonly detected in biological systems during wastewater treatment, probably due to direct input of ARGs.¹⁴ Previous studies have reported similar relative abundances of tet and sul genes in CWs.¹³ In this study, a clear increasing trend of relative abundances for tet and sul genes was acquired with the enhancement of antibiotics (p < 0.05). This implies that the relative abundance of tet and sul genes in the reactors is mainly affected by the antibiotic content. Note the agreement between this result and the previous observation that higher influent TC concentration led to more copy numbers of tet genes in the effluent.⁴ In general, the relative abundances of tet and sul genes in CW-MFC layers increased in the following order: anode layer > cathode layer > middle layer 2 ≥ middle layer 1. These results imply that the fate of ARGs in the different layers of CW-MFCs was mainly attributable to the substrate accumulation of antibiotics (Fig. 2). In addition, the relatively high abundance of tet and sul genes in the cathode layer was probably caused by the difference in oxygen capacity in CW-MFCs. Previous studies have indicated that oxygen could increase in the quantities of ARGs in CWs, yet oxygen was not considered as a controlling factor in ARGs development.²⁷ However, high relative abundance of ARGs in the cathode layer may increase the risk posed by wide distribution of the host bacteria.

Some present surveys indicated that tetM is the dominant tetracycline resistance gene in most cases.⁴ However, the relative abundances of tetA were slightly higher than those of the other tet genes in the samples. Similar to a previous report,¹³ the relative abundances of sulI and sulII, genes were similar at each sampling site under the same conditions of exogenous sul genes input. This implies that the abundance of different types of tet and sul genes in the CW-MFCs was mainly affected by the content antibiotics, structure and operating conditions.

Antibiotics significantly affected the relative abundances of the tetC, tetQ, tetW, sulI and sulII genes in the effluent of the CW-MFCs. The ARG content of the influent was considered negligible for the synthetic wastewater made from tap water; therefore, the relative abundances of the target ARGs in the effluents were probably caused by SMX and TC (Table 1). Moreover, sul and tet genes exhibited a relatively low abundance in effluent water compared with the substrates, which may led to the occurrence and spread of ARB and ARGs. However, the reactors performed well in inhibiting ARGs in the effluent compared with the techniques reported in other studies.⁴

3.3. Effects of antibiotics on electricity production

To explore the performance of CW-MFCs under three different concentrations of antibiotics, current density studies were performed (Fig. 4A). The current density was affected by the concentration of antibiotics, and higher concentration led to a lower current density. For example, the current density was 25 mA m⁻² for the CW-MFC containing 1600 μg L⁻¹ antibiotics; however, it was 30 mA m⁻² for one that contained 400 μg L⁻¹ antibiotics. Thus, an increase in antibiotic concentration resulted in a decrease in current density or voltage. The polarization curves portrayed here visualized the variation with different external resistance. Power density over time was slightly increased when the antibiotics level increased from 400 to 1000 μg L⁻¹ in CW-MFCs (Fig. 4B). This implies that the power density was not strongly affected by the antibiotic content in such a complicated system. In addition, vertex cell voltages were 0.62, 0.68 and 0.41 V at total influent antibiotic concentrations of 400, 1000 and 1600 μg L⁻¹, respectively (Fig. 4B). This phenomenon was probably due to the impedance of the biological activity of the systems by antibiotics.

Fig. 4 Electricity research of CW-MFCs under different concentrations of antibiotics (antibiotics concentrations were 400, 1000 and 1600 μg L⁻¹ in CW-MFCs respectively): (A) current density; (B) polarization and power density curves; (C) electrode potential.

An apparent deviation of power density trend between the 1600 μg L⁻¹ and lower concentrations reflected that the higher concentration of antibiotics actually brought about a lower output of power. Furthermore, details about electrode polarization are described below (Fig. 4C). Anode polarization profiles with identical large slopes were detected by changing the currents in large scope. However, the reverse situation occurred in all the cathodes, which maintained voltages at around 0.2 V. Performance in all the cathodes proved that oxygen reduction reversibility in cathodes was sufficient to support a big current output. Therefore, the structure with their huge surface areas of activated carbon and oxygen filling plants in the cathodes enhanced the electrochemical reaction.^28–30 The anodes, which played a role in electron output by electrically activated bacteria, had bad polarization in all installations, maybe because of weak electron production capacitance in anaerobic conduction competing with other strains. Small systems with intensive electrode potential induction often owe their performance to the anode.^31,32

4. Conclusions

This study clearly demonstrated the strong performance of CW-MFCs for TC and SMX removal from wastewater. The concentrations of TC and SMX in the different layers of the CW-MFCs decreased in the order of anode layer > cathode layer > middle layer. Moreover, the relative abundances of corresponding ARGs in all the samples were detected and ranged from 1.40 × 10⁻⁴ to 6.30 × 10⁻¹ for tet genes and 5.58 × 10⁻³ to 9.27 × 10⁻² for sul genes. The fate of the ARGs in the different layers of the CW-MFCs is mainly attributed to the substrate accumulation of antibiotics. The relative abundances of the tetC, tetQ, TetW, sulI and sulII genes in the effluent were significantly affected by antibiotics. In addition, 0.0578 W m⁻² was the highest power density obtained under the external resistance of 1000 Ω. However, an obvious drop in power density was observed as the concentration of antibiotics increased from 400 to 1600 μg L⁻¹. Further study is required to better understand the fate of potential pathogens caring tet and sul genes.

Acknowledgements

The authors gratefully acknowledge the financial supports from the National Natural Science Foundation of China (41571476), the Provincial Natural Science Foundation of Jiangsu, China (BK20141117) and the Provincial Key Technologies R&D Program of Jiangsu, China (BE2015358).

References

1. C. Bouki, D. Venieri and E. Diamadopoulos, Ecotoxicol. Environ. Saf., 2013, 91, 1–9.
2. Y. Luo, L. Xu, M. Rysz, Y. Wang, H. Zhang and P. J. Alvarez, Environ. Sci. Technol., 2011, 45, 1827–1833.
3. B. Li, X. Zhang, F. Guo, W. Wu and T. Zhang, Water Res., 2013, 47, 4207–4216.
4. X. Huang, C. Liu, K. Li, J. Su, G. Zhu and L. Liu, Water Res., 2015, 70, 109–117.
5. K. Kümmerer, Chemosphere, 2009, 75, 417–434.
6. J. L. Martinez, Environ. Pollut., 2009, 157, 2893–2902.
7. J. Davies and D. Davies, MMBR, Microbiol. Mol. Biol. Rev., 2010, 74, 417–433.
8. S. R. Higgins, C. Lau, P. Atanassov, S. D. Minteer and M. J. Cooney, ACS Catal., 2011, 1, 994–997.
9. J. Villaseñor, P. Capilla, M. A. Rodrigo, P. Cañizares and F. J. Fernández, Water Res., 2013, 47, 6731–6738.
10. S. T. Liu, H. L. Song, S. Z. Wei, F. Yang and X. N. Li, Bioresour. Technol., 2014, 166, 575–583.
11. Z. Fang, H. L. Song, N. Cang and X. N. Li, Biosens. Bioelectron., 2015, 68, 135–141.
12. L. Doherty, Y. Q. Zhao, X. H. Zhao and W. K. Wang, Chem. Eng. J., 2015, 266, 74–81.
13. L. Liu, Y. H. Liu, Z. Wang, C. X. Liu, X. Huang and G. F. Zhu, J. Hazard. Mater., 2014, 278, 304–310.
14. C. W. McKinney, K. A. Loftin, M. T. Meyer, J. G. Davis and A. Pruden, Environ. Sci. Technol., 2010, 44, 6102–6109.
15. S. Rodriguez-Mozaz, S. Chamorro, E. Marti, B. Huerta, M. Gros, A. Sanchez-Melsio, C. M. Borrego, D. Barcelo and J. L. Balcazar, Water Res., 2015, 69, 234–242.
16. D. Wu, Z. T. Huang, K. Yang, D. Graham and B. Xie, Environ. Sci. Technol., 2015, 49, 4122–4128.
17. A. L. Barkovskii and C. Bridges, Water, Air, Soil Pollut., 2011, 223, 49–62.
18. A. Naquin, A. Shrestha, M. Sherpa, R. Nathaniel and R. Boopathy, Bioresour. Technol., 2015, 188, 79–83.
19. A. Dan, Y. Yang, Y. N. Dai, C. X. Chen, S. Y. Wang and R. Tao, Bioresour. Technol., 2013, 146, 363–370.
20. X. Cao, H. L. Song, C. Y. Yu and X. N. Li, Bioresour. Technol., 2015, 189, 87–93.
21. H. Storteboom, M. Arabi, J. G. Davis, B. Crimi and A. Pruden, Environ. Sci. Technol., 2010, 44, 1947–1953.
22. L. Liu, C. Liu, J. Zheng, X. Huang, Z. Wang, Y. Liu and G. Zhu, Chemosphere, 2013, 91, 1088–1093.
23. Z. Shuai, H. L. Song, X. L. Yang, K. Y. Yang and X. Y. Wang, Chemosphere, 2016, 164, 113–119.
24. L. Nielsen, M. J. Biggs, W. Skinner and T. J. Bandosz, Carbon, 2014, 80, 419–432.
25. E. K. Putra, R. Pranowo, J. Sunarso, N. Indraswati and S. Ismadji, Water Res., 2009, 43, 2419–2430.
26. A. V. Dordio and A. J. Carvalho, J. Hazard. Mater., 2013, 252–253, 272–292.
27. N. Wu, M. Qiao, B. Zhang, W.-D. Cheng and Y.-G. Zhu, Environ. Sci. Technol., 2010, 44, 6933–6939.
28. Z. Fang, H.-L. Song, N. Cang and X.-N. Li, Bioresour. Technol., 2013, 144, 165–171.
29. Y. Zhao, S. Collum, M. Phelan, T. Goodbody, L. Doherty and Y. Hu, Chem. Eng. J., 2013, 229, 364–370.
30. J. Villasenor, P. Capilla, M. Rodrigo, P. Canizares and F. Fernandez, Water Res., 2013, 47, 6731–6738.
31. C. I. Torres, A. Kato Marcus and B. E. Rittmann, Biotechnol. Bioeng., 2008, 100, 872–881.
32. C. I. Torres, A. K. Marcus and B. E. Rittmann, Appl. Microbiol. Biotechnol., 2007, 77, 689–697.

---