Reduction of silver nanoparticle toxicity affecting ammonia oxidation using cell entrapment technique

Silver nanoparticles (AgNPs) have been used in many applications, such as textile, medical, and plastic products (Pantic 2014). The large amount of AgNPs used resulted in residue being deposited in wastewater treatment plants (WWTPs) (Mitrano et al. 2011; Hoque et al. 2012; Li et al. 2013). With the superior antimicrobial ability of the particles, microorganisms in WWTPs, including nitrifying microorganisms, might be inhibited (Chopra 2007). Previous works reported that ammonia oxidation (AO) activity of nitrifying sludge could be seriously inhibited by silver species, including AgNPs and liberated Ag⁺ (Choi et al. 2008; Jeong et al. 2012; Giao et al. 2017). The silver species that partially inhibited the AO process followed uncompetitive inhibition-like kinetics with the AgNPs and liberated Ag⁺ inhibitory constants of 73.5 and 0.29 mg/L, respectively (Giao et al. 2017).

The potential of the cell entrapment technique to alleviate the negative impact of toxic substances on microbial activity has been studied. Cell entrapment materials reduce cell exposure to toxic chemicals leading to lower microbial inhibition (Saez et al. 2012; Siripattanakul-Ratpukdi 2012; Siripattanakul-Ratpukdi et al. 2014; Lucchesi et al. 2015). Prior studies have also indicated that the entrapment materials assist living microorganisms to resist adverse environmental conditions (e.g. pH and temperature changes) and to retain higher cell numbers in the system without reducing the microbial activity (Siripattanakul & Khan 2010; Hu & Yang 2015). Siripattanakul-Ratpukdi et al. (2014) found much lower inhibition of nitrification in the entrapped cell system influenced by AgNPs. The previous work preliminarily proved the feasibility of the entrapped cell application. However, the treatment performance and the durability of entrapped cells depended on the entrapment materials (Paul & Vignais 1980; Liu et al. 2012). For example, changing salt solutions (from calcium chloride to barium chloride) during the preparation of gel beads could significantly improve the physical and chemical properties. This, in turn, could lead to higher yield and activity of the entrapped cells (Paul & Vignais 1980). The addition of sodium alginate to polyvinyl alcohol (PVA) could also enhance the stability of the produced beads (Liu et al. 2012).

This study investigated the reduction of AgNP toxicity affecting AO using cell entrapment techniques. The screening of the entrapment materials (barium alginate (BA), polyvinyl alcohol (PVA), and mixture of polyvinyl alcohol and barium alginate (PVA-BA)) was based on nitrification performance and bead durability. The influence of silver species (AgNPs and Ag⁺) on the inhibition of AO was evaluated. The AO kinetics of selected entrapped cells were also investigated. Microbial cell viability and morphology were observed by confocal laser scanning microscopy (CLSM) and scanning electron microscopy coupled with an energy dispersive X–ray (SEM-EDX). This primarily elucidated the mechanism of silver toxicity reduction. The results of this study could be used as fundamental information for AgNP contaminated wastewater treatment in the future.

Silver nanoparticles (5,000 mg/L) were obtained from PrimeNANO technology (Prime Nanotechnology Co. Ltd, Chulalongkorn University, Thailand). Transmission electron microscopy showed spherical AgNP sizes in the range of 2–12 nm (Giao et al. 2017). AgNP suspensions prepared in inorganic synthetic wastewater (SWW) peaked at a wavelength of 400 nm, which indicated that AgNPs were well dispersed in the experimental condition. Additionally, the residue of Ag⁺ from AgNPs of 10 mg/L was found to be less than the method detection limit (3 μg/L). Silver nitrate (AgNO₃, Merck, Darmstadt, Germany) was used as a source of Ag⁺.

The seed sludge was taken from a continuous-flow enriching nitrifier reactor. The continuous reactor had been operating for more than five years with equal hydraulic (HRT) and solid (SRT) retention times of four days. During nitrifying sludge acclimation, SWW was comprised of 1.9821 g/L (NH₄)₂SO₄, 0.2 g/L NaCl, 0.2 g/L K₂HPO₄, 0.4 g/L MgCl₂.6H₂O, 0.1 g/L CaCl₂.2H₂O, 0.5 g/L KCl, and 1.0 g/L NaHCO₃ with an addition of 1 mL of inorganic salt solution. The inorganic salt solution contained 40 g/L MgSO₄.7H₂O, 40 g/L CaCl₂.2H₂O, 200 g/L KH₂PO₄, 1 g/L FeSO₄.7H₂O, 0.1 g/L Na₂MoO₄, 0.2 g/L MnCl₂.4H₂O, 0.02 g/L CuSO₄.5H₂O, 0.1 g/L ZnSO₄.7H₂O, and 0.002 g/L CoCl₂.6H₂O. The NH₄⁺-N concentration of 429 ± 62 mg-N/L was maintained. For the silver exposure experiment, wastewater was comprised of 1.9821 g/L (NH₄)₂SO₄, 1 g/L NaHCO₃, 0.1 g/L K₂HPO₄, 0.025 g/L MgSO₄, and 0.025 g/L NaCl with AgNPs and Ag⁺ at 10 mg/L and 0.1 mg/L, respectively.

The enriched nitrifying sludge was harvested and centrifuged at 5,000 rpm for 20 min and the clear supernatant was removed. The concentrated cells were re-suspended, well-mixed, and centrifuged in SWW as a washing process five times. The washed cells were used for preparing the entrapped cells. The washed cells were added to the entrapment matrices to reach a final cell concentration in the matrix of 2,000 mg-MLSS/L.

Preparation of the BA and PVA-entrapped cells followed previous works (Siripattanakul-Ratpukdi & Tongkliang 2012; Siripattanakul-Ratpukdi et al. 2014). To prepare BA-entrapped cells, 2% (w/v) alginic acid sodium salt and 5% (w/v) barium chloride were used. The alginic acid sodium salt solution and the nitrifying sludge were homogeneously mixed and then dropped into the barium chloride solution using a peristaltic pump at 20 rpm. The gel beads were hardened in the barium chloride solution for 1 h. BA gel beads with diameters of 3–6 mm were formed.

For the PVA-entrapped cells, 10% (w/v) polyvinyl alcohol (PVA, 99%, fully hydrolyzed) was applied. The mixture of the PVA solution and the nitrifying sludge was dropped into a saturated boric acid solution for 30 min before it was transferred to a 1 M sodium phosphate buffer for 2 h to harden the produced gel beads. PVA gel beads of 3–6 mm diameters were formed. The other entrapment material applied in this study was a PVA-BA matrix. The combination of PVA and alginate was developed and successfully applied for entrapping phenol-degrading bacteria (Wu & Wisecarver 1992). The PVA-BA-entrapped cells were prepared by using 10% (w/v) polyvinyl and 1% (w/v) alginic acid sodium salt. The combined gel solution was mixed with the sludge. The preparation protocol for PVA-BA-entrapped cells followed that of the PVA gel, as described above.

AO batch experiments were performed. A glass reactor with 100 mL working volume was applied. The reactor was operated under room temperature and 2 mg/L dissolved oxygen (DO). The study was divided into two parts: (1) inhibited AO by entrapped cells and (2) kinetics of AO by selected entrapped cells. The first part emphasized on selection of appropriate entrapment material (among BA, PVA, and PVA-BA) for the situation. The selection was based on AO percentage, bacterial viability, and material durability. For the second part, AO kinetics by selected entrapped cells was performed. This part focused on AO modeling based on Monod equation. Also, bacterial cell and bead morphology after the experiment was investigated using SEM-EDX. Figure 1 is summarized overall experiments.

For the kinetic experiments, the 60-h AO tests with NH₄⁺-N (10 to 180 mg-N/L) and silver (AgNPs of 10 mg/L, Ag⁺ of 0.1 mg/L, and no silver addition) were performed. The NH₄⁺-N concentrations were selected to represent the concentrations reported in municipal and industrial wastewater, such as coking and rubber production wastewater (Limpiyakorn et al. 2005; Chaiprapat & Sdoodee 2007; Bai et al. 2012). The selected silver concentrations previously reported a similar AO inhibition percentage from the free nitrifying sludge (Giao et al. 2017). 100 mL of the reactors and the selected entrapped and free cells were aerated for 60 h. The aqueous samples were collected at 0, 5, 17, 42, and 60 h to monitor AO rate. The concentration of NH₄⁺-N was measured using the salicylate–hypochlorite method (Bower & Holm-Hansen 1980). The NH4⁺-N measurement method was selected because it is more suitable when several ions, such as Ca²⁺, Mg²⁺, and Cl⁻, are present in the sample (Le & Boyd 2012). This method is frequently used for the analysis of ammonia in seawater.

Free and entrapped cell viability was investigated. Ten entrapped cell beads after the experiment were collected. The bead sample was washed five times with de-ionized water and five times with a NaCl solution (0.85%). The washed beads were cut using a pair of tweezers. The tweezed beads were dipped in 10 mL of NaCl (0.85%) and then gently squeezed to let the cells out of the entrapment matrices. The suspended cells (free and de-entrapped cells) were centrifuged at 5,000 rpm for 5 min. The supernatant was discarded, and the pellets were collected. The pellets were re-suspended and concentrated to obtain the free and de-entrapped cell samples for the live/dead assay.

The staining procedure was performed using a staining kit (LIVE/DEAD^® BacLight™ Bacterial Viability, Molecular Probes, Invitrogen). The samples were observed under CLSM (FluoView FV10i, Olympus, Japan). The following two samples were compared: (1) microbial cells directly taken from the nitrifying reactor and (2) killed cells (autoclaved at 121 °C, at 100 kPa for 15 min), stained for estimation of microbial viability. The percentage of live and dead cells was estimated based on 30 CLSM-images. The results were presented as % dead cells (±standard deviation).

After the kinetic experiment, microbial cells and bead physiology were observed using SEM-EDX. The BA-entrapped cells (selected matrix) were prepared for SEM observations according to the procedure by Siripattanakul-Ratpukdi et al. (2014). Briefly, the entrapped cell beads from the experiment were collected and washed with de-ionized water. The collected gel beads were fixed using 1% osmium tetroxide for 1 h. The fixed gel beads were then rinsed three times with de-ionized water for 10 min. The beads were dehydrated with a series of ethanol that ranged from 30 to 95%. The absolute ethanol was applied as the last step of the dehydration. The dehydrated beads were dried using a critical point dryer (Balzers, CPD 020, Liechtenstein). The beads were divided into two parts using a razor blade in liquid nitrogen, attached to a stub, and coated with gold. The dried BA, PVA, and PVA-BA beads were observed using SEM with an energy-dispersive spectroscopy attachment (JEOL, JSM-5410LV, Tokyo, Japan). Elemental compositions of selected areas were analyzed using an energy dispersive X–ray (EDX, Oxford Instruments, Model X–MaxN, UK) to identify aggregates of silver.

The percentage of AO inhibition and membrane compromised cells caused by AgNPs and Ag⁺ was presented as an averaged percent inhibition (%) ±standard deviation (SD), regardless of the initial NH₄⁺–N concentrations. A statistical comparison was performed using IBM SPSS statistics for Windows, Version 19.0 (IBM Corp., Armonk, NY, USA). The significance of the differences (p < 0.05) was determined by a one-way analysis of variance (ANOVA) followed by Duncan's test (Ahrari et al. 2015).

An AO experiment under various conditions was performed. Blank AO test and AO tests by only entrapment matrices (no cells) were conducted (results presented in Figure S2 in the supplementary material, available with the online version of this paper). It was found that natural ammonia removal (based on blank experiment) and ammonia adsorption on entrapment matrices did not obviously occur. For the tests with nitrifying sludge, under the absence of silver contamination, the nitrifying sludge reduced ammonia. Percentages of AO by the free (94.5%) and entrapped (90.2–95.5%) nitrifying sludge were similar (Figure 2). The entrapped cells did not run into limited substrate diffusion problem (Siripattanakul-Ratpukdi et al. 2014). This showed potential of entrapped cell utilization in the future.

During the experiments with silver, AO performance was significantly reduced. Under the presence of AgNP (10 mg/L), the free and entrapped cells reduced ammonia by 31.3 and 55.0–70.2%, respectively. With regards to the influence of Ag⁺ (0.1 mg/L), the free and entrapped cells oxidized ammonia by 19.4 and 54.2–75.5%, respectively. Previously, Giao et al. (2017) reported that the AO inhibition by silver caused by microbial cell damage. In this study, the entrapment matrix led to tortuosity of silver to reach to microbial cells resulting better AO by the entrapped cells. Figure 2 clearly shows that AgNPs at 10 mg/L and Ag⁺ at 0.10 mg/L (represented liberated dissolved silver from AgNPs) partially inhibited AO of the BA-entrapped cells by 25.4 and 19.4%, the PVA-entrapped cells by 37.0 and 38.2%, and the PVA-BA-entrapped cells by 20.7 and 30.4%, respectively. The result revealed that the BA and PVA-BA-entrapped cells performed better than the PVA-entrapped cells. The different entrapment materials produced different gel structures resulting in altered AO inhibition. When compared to the result of free cells reported previously, inhibition percentages of AO by AgNPs at 10 mg/L and Ag⁺ at 0.10 mg/L were 63.3% and 75.3%, respectively (Table 1). The results clearly showed lower AO inhibition in the entrapped cell experiment.

We preliminarily monitored NO₂⁻ plus NO₃⁻-N concentration during the experiments as shown in Table S1 (in supplementary material, available online). The result stated that the entrapment material (BA) did not show the ammonia removal and nitrite plus nitrate production. Nitrite plus nitrate accumulated in all tests with microbial cells under the presence or the absence of silver. However, it was obvious that silver influenced AO and nitrite plus nitrate production. At AgNPs 10 mg/L and Ag+ 0.10 mg/L, inhibition of nitrite plus nitrate production was 41% and 56%, respectively.

In this sub-section, the stability of the entrapped cell beads was the focus. Only the durability of the entrapped beads after 30 days of exposure to AgNPs and Ag⁺ was observed. Figure 3 preliminarily indicates the durability of the gel beads. During the experiment, damaged beads were quantified based on bead swelling and abrasion compared to beads at the beginning. After 30 days, no damaged BA beads (0%) were observed while the PVA (56%) and PVA-BA (10%) beads were obviously swollen and damaged. This phenomenon was also reported in a previous study (Bach & Dinh 2014). The PVA beads deformed because of the hydrophilic nature. The cross-linking of PVA and boric acid could be reversed, resulting in bead swelling. The results revealed that the addition of BA into the PVA gel could improve stability in SWW. This result was in line with other studies that showed that a combination of PVA and BA led to higher stability and storage longevity of the produced gel beads (Paul & Vignais 1980; Liu et al. 2012). Among the cell entrapment techniques, the BA-entrapped cells were the most stable beads. Similarly, a previous study reported that BA had high gel strength and flexibility (Taweetanawanit et al. 2017). The BA-entrapped cells were stable during long-term operation and supported the growth of the entrapped microbes (Siripattanakul-Ratpukdi & Tongkliang 2012).

The cell viability assay (after the 60-h experiment), based on bacterial cell membrane damage is shown in Figure 4. The proportions of dead cells between the free cells (collected directly from the enriching reactor) and the controlled BA- and PVA-entrapped cells (entrapped cells without AgNPs or Ag⁺ exposure) were insignificantly different (p > 0.05). The result revealed that BA and PVA were not toxic to bacterial cells. Prior studies reported that BA and PVA were not harmful to microbes (Siripattanakul & Khan 2010; Wadhawan et al. 2011; Vecchiatini et al. 2015).

Conversely, the percentage of dead cells from the PVA-BA-entrapped cell test (no silver) was significantly higher than free cells (p < 0.05). As discussed earlier, BA and PVA are non-toxic to microbial cells, and likewise the mixture of PVA-BA should not cause toxicity to the entrapped cells. The damaged cells in the PVA-BA test might be due to the viscosity of the material (PVA-BA) resulting in less diffusion of oxygen and ammonia (substrate) into the beads. Among the entrapped cells, BA was the most promising entrapment material for mitigation of the toxicity of AgNPs and Ag⁺. The BA matrix promoted high treatment performance (low inhibition) and cell viability.

The kinetics of AO from the BA-entrapped cells are demonstrated in Figure 5. Table 2 is a summary of the AO kinetic parameters. Figure 5 shows that the experimental results well fitted Monod model. Under absence of silver, self-substrate inhibition did not occur, resulting in the steady specific AO rates for the tests with NH₄⁺ higher than 100 mg/L. Based on free ammonia calculation, only 0.56 mg/L free ammonia may form (Anthonisen et al. 1976). Free ammonia of 10–150 mg/L could inhibit AO. At the concentration related to this study, no inhibition by free ammonia would take place. The result from this study revealed that the tested NH₄⁺ concentrations (0–180 mg-N/L) did not show adverse effect from nitrogenous species.

Results from the tests with no silver indicated that the entrapped cell system encountered the substrate diffusion problem resulting in much lower q_max values compared to that of the free cells. Under the presence of silver, both and values of the free cell systems were much lower than those without silver. This result implied that inhibitors (AgNPs or Ag+) obviously influenced AO in the free cell system. The entrapped cell tests with silver (both species) had similar kinetic parameter values (Table 2). This highlights the success of the cell entrapment application for reducing AgNPs and liberated Ag⁺ toxicity. Table 1 confirms that the entrapped cells reduced the influence of AgNPs and Ag⁺ by three and five times, respectively, as compared to those from the suspended cells at the same conditions.

Figure 6 presents SEM images of the BA-entrapped cells. As shown in Figure 5(a) and 5(b), several micro-colonies laid on the BA-matrices. A previous study indicated that cross-linking between barium and alginate created a dense network with an abundance of fine pores (Siripattanakul-Ratpukdi & Tongkliang 2012). This study also found numerous small pores (Figure 6(c)). The pore sizes ranged from 0.1 to 0.3 μm, which were much smaller than the single cells (1–2 μm), resulting in no micro-colonies in the pores. The small porous structure was related to the low maximum AO rates found in the kinetic experiment of the entrapped cells. This structure was torturous for the substrate to reach the microbial cells inside the BA beads. The dense structure protected the cells from silver species, leading to similar AO rates in the experiments by the entrapped cells under absence and presence of silver species.

It was noted by Giao et al. (2017) that the microbial cells changed from long rod-shaped to shorter rod-shaped in the presence of AgNPs. In this study, most of the cells remained long rod-shaped (Figure 6(d)). This could primarily indicate that the microbial cells were not stressed from the toxic substance; therefore, cell morphological adaptation was not observed. During SEM observation, element (silver) analysis using an EDX technique was also performed. In the entrapped cell experiment, silver clusters were not detected inside the beads. This confirmed that the entrapment matrices reduced AgNPs and liberated Ag⁺ transporting into the beads.

The entrapment materials, including BA, PVA, and PVA-BA, were non-toxic to microbial cells. In the short-term AO test, BA was the most promising entrapment material for reduction of AgNPs and Ag⁺ toxicity. The inhibitory kinetics results showed that the cells entrapped in BA maintained high microbial cell viability (approximately 20% dead cells) and AO up to 90% in the presence of AgNPs or Ag⁺. The kinetic parameters confirmed low influence of AgNPs and Ag⁺ on AO by the entrapped cells. Microscopic observation of the entrapped cells suggested that the porous structure of matrices limited penetration of silver, resulting in lower cell damage and AO inhibition. Future study on the long-term impact of AgNPs and Ag⁺ on AO activity and microbial community change in entrapped cells should be investigated.

The authors acknowledge financial and equipment support from the Research Center for Environmental and Hazardous Substance Management (Khon Kaen University) and Research Career Development Grant RSA6080054 (Thailand Research Fund (TRF) and Khon Kaen University). This work was supported by the National Nanotechnology Center (NANOTEC), NSTDA, Ministry of Science and Technology, Thailand, through its program of Research Network NANOTEC (RNN). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the grant agencies.