Removal of nitrogen and phosphorus by heterotrophic ...

Ecological Engineering 99 (2017) 199–208

Contents lists available at ScienceDirect

Ecological Engineering journal homepage: www.elsevier.com/locate/ecoleng

Removal of nitrogen and phosphorus by heterotrophic nitrification-aerobic denitrification of a denitrifying phosphorus-accumulating bacterium Enterobacter cloacae HW-15 Wenjie Wan, Donglan He ∗ , Zhijun Xue College of Life Sciences, South-Central University for Nationalities, Wuhan 430070, PR China

a r t i c l e

i n f o

Article history: Received 17 August 2016 Received in revised form 12 October 2016 Accepted 13 November 2016 Available online 19 November 2016 Keywords: Denitrifying phosphate-accumulating bacterium Screening Identification Biological characteristics

a b s t r a c t A new denitrifying phosphorus-accumulating bacterium HW-15 was screened out from phosphorus-rich wastewater and was identified as Enterobacter cloacae sp. based on its physiological-biochemical characteristics and 16S rDNA sequence analysis. The shift in microbial community structure during acclimation was determined through high-throughput sequencing and Enterobacter was found to be the dominant genus. Strain HW-15 exhibited efficient simultaneous denitrifying and phosphorus removal, and achieved both heterotrophic nitrification and aerobic denitrification process, and demonstrated a preference for consuming nitrogen sources in order of NH+ 4-N > NO- 3-N > NO- 2-N. The application of strain HW-15 in real wastewater samples resulted in NH+ 4-N, NO- 3-N, NO- 2-N and PO3- 4-P removal efficiencies of 99%, 88%, 59%, and 73%, respectively. Besides, strain HW-15 was found to simultaneously remove nutrient and arsenic(V) at an arsenic(V) concentration lower than 1.5 mg/L. During anaerobic-aerobic culture period, the polyhydroxyalkanoate (PHA) content in HW-15 cells presented a dynamic change, and resulted in PO3- 4-P removal efficiency of 61.97%. Strain HW-15 could process NH+ 4-N to NO- 2-N and then to gaseous nitrogen. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Nitrogen and phosphorus removal from the wastewater is one of the main purposes for the wastewater treatment. Conventionally, nitrogen and phosphorus removals are function of different microorganisms in a wastewater treatment system (Liu et al., 1996; Lee et al., 2003). Nitrogen is removed by energy and time consuming procedure (Khin and Annachhatre, 2004), which is based on the combination of anaerobic or anoxic denitrification and autotrophic nitrification (Khardenavis et al., 2007). For example, 2.47 g of methanol is required per gram of nitrate nitrogen for complete denitrification (McCarty et al., 1969). In the past 30 years, attempts have been made to use microorganisms to remove nitrogen and phosphorus from wastewater (Kuba et al., 1996; Xia and Liu, 2004; Liu and Li, 2015). However, denitrifying bacteria compete with phosphorus-accumulating organisms (PAO) for limited carbon sources which restrict the growth of bacteria and results in the change of relative abundance of nitrifying bacteria and PAO during the aging of sludge, which presents a challenge in optimizing

∗ Corresponding author. E-mail address: [email protected] (D. He). http://dx.doi.org/10.1016/j.ecoleng.2016.11.030 0925-8574/© 2016 Elsevier B.V. All rights reserved.

sewage treatment systems to achieve simultaneous denitrification and phosphorus removal which need related microorganisms and different reaction conditions (Meinhold et al., 1999). Denitrifying phosphorus-accumulating bacteria (DPAB) are facultative anaerobic bacteria that can utilize O2 , NO- 3-N, or NO2-N as final electron acceptors to take up phosphorus under aerobic conditions or to release phosphorus under anaerobic condition (Kerrn-Jespersen and Henze, 1993; Meinhold et al., 1999; Ahn et al., 2001). During the anaerobic phase, DPAB can hydrolyze intracellular polyphosphate (poly-P) and glycogen to provide energy, take up volatile fatty acids (VFA) to synthesize intracellular polyhydroxyalkanoates (PHAs) and release phosphate to the environment (Zeng et al., 2003a,b,c). In the subsequent aerobic phase, the DPAB are able to absorb more soluble phosphorus into the cell in the form of poly-P than that have been released in the anaerobic phase, so phosphorus is removed from the wastewater (Mino, 2000; Yu et al., 2014). Recently, progress has been achieved in the isolation of DPAB, like Pseudomonas (Li et al., 2015), Acinetobacteria (Tsuneda et al., 2006; Liu and Li, 2015), Aeromonas (Wang et al., 2008), and Planctomycetes (Liu et al., 2013). The discovery of DPAB promise the possibility of solving the problem of the change of bacteria type and relative richness during nutrient removal such that nitrogen and phosphorus removal can be achieved simultaneously (Bortone

200

W. Wan et al. / Ecological Engineering 99 (2017) 199–208

et al., 1999; Lee et al., 2001; Guo et al., 2014). The nitrogen and phosphorus removed by DPAB are mainly carried out in the consortium of anaerobic and aerobic (Li et al., 2015), however, the composition of the consortium was unclear, and the high removal efficiencies for the nitrogen and phosphorus could not be achieved under the same conditions (Li et al., 2015). From the practice view, the achievement of simultaneous nitrogen removal by using heterotrophic nitrification–aerobic denitrification and phosphorus removal by collecting PHA containing bacteria during aerobic period in one reactor will be conceivable. The main purposes of this study were to isolate and identify efficient DPAB strain with high heterotrophic nitrification–aerobic denitrification capacity, and to assess how environmental factors would affect the growth and nutrient removal efficiency, and to determine the possibility of simultaneous removal of nutrient and arsenic(V) which was rich in cultivated soil and agricultural water. Experiments were also conducted to explore the process of nitrogen removal and try to clarify the dynamic change of PHA and phosphorus content in anaerobic-aerobic condition.

TGG CTC AG-3 ; reverse primer, 5 -GGT TAC CTT GTT ACG ACT T-3 ). The amplification mix contained 2.5 ␮L 10 × Ex Taq buffer, 1 ␮L dNTPs (2.5 mmol/L), 1 ␮L forward primer (10 pmol/␮L), 1 ␮L reverse primer (10 pmol/␮L), about 30 ng of genomic DNA, 0.5 ␮L Ex Taq (5 U/␮L) and water to 25 ␮L. The reaction was performed in a Bio-Rad (USA) thermocycler with 5 min at 95 ◦ C, 34 cycles of 50 s at 95 ◦ C, 50 s at 56 ◦ C, 1 min at 72 ◦ C and a final elongation step of 10 min at 72 ◦ C. About 150 ng of 16S rDNA PCR product purified by TaKaRa MiniBEST DNA Fragment Purification Kit (Dalian, China) were ligated with 50 ng of pMD19-T vector (molar ratio was 5:1) in 3 ␮L Solution I and water to 10 ␮L. The sample was incubated at 16 ◦ C for 10 h. Then 5 ␮L of the ligated product was introduced into 200 ␮L of E. Coli DH5␣ competent cells. The mixtures were cultured on LB medium containing ampicillin and clones containing the insert were identified by colony PCR. Positive clone was identified by sequencing (Quintarabio Biotech, Wuhan, China). The 16S rDNA sequence of strain HW-15 was submitted to GenBank and analyzed against known gene sequencing using BLAST. The phylogenetic tree was built using MEGA6 software.

2. Materials and methods 2.4. Physiological-biochemical properties of strain HW-15 2.1. Main media Bacteria selective medium (BSM) contained (per liter) 5 g of CH3 COONa, 0.05 g of K2 HPO4 , 0.2 g of KH2 PO4 , 1 g of NH4 Cl, 0.2 g of MgSO4 ·7H2 O and 2 mL of trace element solution. Basal inorganic medium (BIM) contained (per liter) 0.2 g of K2 HPO4 , 0.2 g of MgSO4 ·7H2 O and 2 mL of trace element solution. Heterotrophic nitrification medium (HNM) contained (per liter) 2.344 g of CH3 COONa, 0.2 g of K2 HPO4 , 0.722 g of KNO3 (HNM-1) or 0.493 g of NaNO2 (HNM-2) or 0.382 g of NH4 Cl (HNM-3), 0.2 g of MgSO4 ·7H2 O and 2 mL of trace element solution. Trace element solution contained (per liter) 10 g of EDTA, 0.2 g of ZnSO4 , 1.2 g of MnCl2 ·4H2 O, 1 g of FeSO4 ·7H2 O, 0.5 g of CuSO4 ·5H2 O, 0.3 g of CoCl2 ·6H2 O, 0.2 g of Na2 MoO4 ·2H2 O, 0.1 g of CaCl2 . LB medium contained (per liter) 10 g/L of tryptone, 5 g/L of yeast extract, 10 g/L of NaCl. The initial pH of all the media mentioned above were adjusted to 7.0–7.2. 2.2. DPAB acclimation and screening Water samples were taken from a phosphorus-rich river in southwestern Hubei Province of China (29◦ 07 N, 108◦ 23 E) and kept at 4 ◦ C. Ten milliliter of the phosphorus-rich wastewater was added to BSM (100 mL). The mixtures were cultured at 30 ◦ C in anaerobic condition (IB-anae), aerobic condition (IB-ae), or a combine of anaerobic/aerobic (12 h/12 h as a cycle) condition (IBanae/ae) for total 72 h. The culture suspension were harvested from IB-anae, IB-ae, and IB-anae/ae and sent for high-throughput sequencing. Meanwhile, the bacteria suspensions were collected for serial dilutions of 10−1 –10−9 . Solid BSM agar was inoculated with 0.05 mL of each dilution in duplicate and cultured in a 30 ◦ C incubator for 1–2 days. Single colonies were sub-cultured by picking and streaking 5 times to isolate pure colonies. The positive strains were designated as DPAB by metachromatic granules and PHA granules staining (Dong and Cai, 2001). 2.3. Molecular identification of strain HW-15 DNA isolation of strain HW-15 was carried out using TIANamp Bacteria DNA Kit (Beijing, China). The total DNA concentration was determined using nucleic acid protein measuring instrument (Q5000, Quawell, USA). The 16S rDNA of strain HW-15 was amplified by PCR using universal bacterial primers (forward primer, 5 -AGA GTT TGA TCC

Strain characteristics such as Voges-Proskauer test, methyl red test, starch hydrolysis test were determined according to previously described methods (Dong and Cai, 2001). 2.5. Biological characteristics of nitrogen removal A single colony of strain HW-15 was inoculated to 100 mL of LB medium at 30 ◦ C with shaking at 200 rpm for 16 h. Cultures were centrifuged at 4000 rpm for 10 min. The optical density (OD600) was adjusted to 1 by dilution with sterile water after bacteria were washed. The diluted bacteria were used as seed cultures and inoculated to medium with the inoculation rate of 1% in subsequent experiments. To assess how carbon sources, aeration, temperature, C/N ratio, and initial nitrogen concentration affected the nitrogen removal efficiency, single factor experments were conducted in flasks using the BIM supplied with 100 mg/L NO- 3-N (0.722 g/L KNO3 ) or NO2-N (0.493 g/L NaNO2 ) or NH+ 4-N (0.382 g/L NH4 Cl). In the carbon sources experiments, sodium acetate, glucose, sucrose, trehalose, mannitol, and formic acid were added separately to the BIM to yield a C/N of 8. In the aeration experiments, the shaking speeds of 0, 50, 100, 150, and 200 rpm were applied using sodium acetate as carbon source at a C/N ratio of 8. The temperature experiments were performed using the same medium as in the aeration experiments and the medium was incubated at 15, 20, 25, 30, 35, or 40 ◦ C. The effects of C/N ratios were measured using sodium acetate as the sole carbon source, and a C/N ratio of 2, 4, 6, 8, and 10, respectively. In the initial nitrogen concentration experiments, sodium acetate was used as sole carbon source and NO- 3-N or NO- 2-N or NH+ 4-N was added at a final concentration of 25, 50, 100, 200, and 400 mg/L, with a C/N of 8. All of the above experiments were conducted in triplicates with inoculation of 1% (v/v) in 150 mL flasks with 50 mL of sterile medium. During the 24 h incubation, cultures were sampled periodically and filtered by 0.22 ␮m micro filtration membrane to determine the concentration of NO- 3-N, NO- 2-N, and NH+ 4-N. 2.6. Determination of arsenic toxicity to strain HW-15 To explore whether strain HW-15 can remove nutrient in the condition of As(V) which could be a toxicity to bacteria, strain HW-15 was inoculated to HNM-1 prepared with different initial arsenic concentrations ranging from 0.5 mg/L to 3 mg/L, and cul-


201

Fig. 1. Microbial community structure shift during acclimation as determined by Illumina sequencing of 16S rDNA and reflected by relative frequency. IB-anae, IB-ae, and IB-anae/ae were samples that bacteria suspension collected from different culture conditions.

Table 1 NH+ 4-N, NO- 3-N, and NO- 2-N removal efficiency under different carbon sources, shaking speeds, temperatures, C/N ratios, and initial nitrogen concentrations by strain HW-15 in 24 h. The results were the mean values of three replicates. Factors

Level

Removal of efficiency (%)* NO2 − -N

NO3 − -N

NH4 + -N

Carbon source

Glucose Sucrose Trehalose Mannitol Formic acid Sodium acetate

67.54 ± 1.65 59.27 ± 2.04a 66.38 ± 0.67b 52.24 ± 1.28a 49.83 ± 3.45a 77.11 ± 4.56c

2.35 ± 1.03 7.32 ± 2.14a 14.23 ± 0.23a 9.01 ± 1.59a 5.68 ± 0.76a 41.36 ± 1.69b

87.89 ± 0.67c 86.29 ± 1.54c 76.79 ± 1.12b 87.78 ± 2.91c 64.56 ± 1.28a 93.29 ± 2.19c

Shaking speed (rpm)

0 50 100 150 200

94.13 ± 0.89c 95.96 ± 1.16c 85.64 ± 2.11b 70.44 ± 1.38a 65.34 ±0.76a

69.23 ± 1.02b 85.23 ± 2.19c 60.28 ± 3.24a 68.76 ± 1.21b 54.21 ± 1.12a

55.21 ± 1.12a 59.44 ± 1.05a 64.15 ± 3.49ab 85.27 ± 1.15c 93.23 ± 0.21d

Temperature (◦ C)

15 20 25 30 35 40

58.69 ± 1.14b 61.58 ± 2.03b 70.25 ± 1.69c 65.34 ± 0.76b 70.58 ± 3.45c 40.36 ± 1.25a

45.21 ±1.00a 46.17 ± 0.58a 57.58 ± 3.46b 54.21 ± 1.12b 55.44 ± 2.39b 38.44 ± 1.89a

20.38 ± 0.58a 66.49 ± 5.12c 80.16 ± 2.43d 93.23 ± 0.21e 89.28 ± 2.55e 56.41 ± 3.14b

C/N ratio

2 4 6 8 10

33.55 ± 1.28a 40.41 ± 1.57b 45.89 ± 1.25b 65.34 ±0.76c 47.12 ± 3.67b

17.28 ± 1.25a 24.22 ± 2.58b 36.79 ± 1.25c 54.21 ± 1.12d 26.01 ± 1.52b

30.98 ± 1.28a 66.15 ± 2.55b 80.31 ± 0.82c 93.23 ± 0.21d 70.10 ± 0.59b

Initial nitrogen concentration (mg/L)

25 50 100 200 400

95.24 ± 1.24d 86.21 ± 6.34c 65.34 ± 0.76b 41.02 ± 1.39a 39.44 ± 1.25a

92.89 ± 1.08d 85.24 ± 4.25c 54.21 ± 1.12b 30.24 ± 1.89a 28.16 ± 1.25a

97.32 ± 1.42c 95.87 ± 4.21c 93.23± 0.21c 76.21 ± 5.23b 52.31 ± 1.38a

*

b

a

Values followed by different letters in the same column were significantly different at p < 0.05.

tured at 30 ◦ C and shaken at 200 rpm for 24 h. After incubation, OD600 values, As(V), NH+ 4-N, and PO3- 4-P concentrations were measured.

2.7. Simultaneous removal of nitrogen and phosphorus from real wastewater The wastewater used in the current experiment was collected from a local sewage in Hubei. The wastewater had a pH of

202


NaClO3 or KClO4 was used as control. Samples were taken at the 12th h of the incubation to determine the content of COD, PO34-P, intercellular P, NH+ 4-N, NO- 2-N, NO- 3-N, and intracellular N. 2.9. Determination of PHA content To investigate the dynamic change of PHA content in bacteria cells, strain HW-15 was inoculated to HNM-1 medium and trained in anaerobic-aerobic (12 h/12 h) conditions for 24. After 12 h, NH+ 4-N (final concentration was 100 mg/L) was added to the culture under oxygen flow training for another 12 h. Samples were taken to determine the concentration of PHA, PO3- 4-P, NO- 3-N, NO- 2-N, and NH+ 4-N. 2.10. Analytical methods and calculations

Fig. 2. Poly-P and PHA staining of strain HW-15.

7.15 ± 0.15 and contained (mg/L) chemical oxygen demand (COD), 96.82 ± 2.58; total nitrogen (TN), 49.87 ± 0.62; total phosphorus (TP), 50.58 ± 1.24; NO- 3-N, 5.36 ± 1.02; NO- 2-N, 1.25 ± 0.13 and NH+ 4-N, 42.67 ± 2.12. In treatments, 0.1% (w/v) sodium acetate and 1% (v/v) inoculation were added to the wastewater. Original wastewater without the addition of bacterial inoculum and carbon source was used as blank control group to eliminate the influence of indigenous microbial groups. The shaking speed was set to 200 rpm and the temperature for incubation was 30 ◦ C. The experiments were conducted for 60 h in 1000 mL flasks containing 500 mL of wastewater. During treatment, cultures were sampled periodically to determine the concentration of COD, TN, NO- 3-N, NO- 2-N, NH+ 4-N, and TP. 2.8. Estimation of effects of NaClO3 and KClO4 on heterotrophic nitrification by strain HW-15 To explore if the NH+ 4-N was firstly oxidized to NO- 2-N and then to NO- 3-N in the heterotrophic nitrification process of strain HW-15, mass balance experiment was conducted and NaClO3 or KClO4 was used to block NH+ 4-N being processed into NO- 2-N as previous research reported that NaClO3 was an inhibitor to stop the procedure of NO- 2-N to NO- 3-N (Zhao et al., 2014). NaClO3 or KClO4 (final concentration was 1 g/L) and strain HW-15 were added to HNM-3 simultaneously and incubated under optimal conditions. HNM-3 inoculated with strain HW-15 but without addition

Cell density was measured using a spectrophotometer UV1750 (Shimadzu, Japan) at an absorbance of 600 nm (OD600). COD, PO3- 4-P, NO- 3-N, NO- 2-N, and NH+ 4-N were determined by potassium dichromate method, antimony molybdenum blue spectrophotometry, ultraviolet spectrophotometric method, ␣-naphthylamine spectrophotometry, and nessler’s reagent spectrophotometry, respectively. As(V) was determined by heteropoly arsenomolybdicacid-crystal violet spectrophotometric method (APAH, 1998). TN was detected by the alkaline persulfate oxidation with an ultraviolet spectrophotometric method. PHA was detected using gas chromatography (Randall and Liu, 2002). Nitrogen and phosphorus removal efficiency and removal rate were calculated as: ␩ = 100% × (Ci − Cf )/Ci and ␦ = (Ci − Cf )/t, respectively, where ␩ was removal efficiency, Ci was the initial concentration (mg/L), Cf was the final concentration (mg/L), ␦ was removal rate (mg/(Lh)) and t was the incubation time (h). PHA concentration was calculated as: C (mg/L) = m/v, where C was the PHA concentration, m was the PHA mass (mg), v was the volume of bacterial suspension (L). The T-test was applied for statistical analysis of nitrogen and phosphorus removal efficiency and p < 0.05 or p < 0.01 was taken as the significant level determination. Standard deviations of three replicates in the experiment were calculated by Microsoft excel. Sequencing data were optimized by removing low-quality sequences, unrecognized reverse primers, and any ambiguous base calls, with a length < 200. High-quality sequences were clustered into 97% similarity operational taxonomic units using UCLUST software (http://drive5.com/usearch/manual/uclust algo.html). A representative sequence from each operational taxonomic unit was classified and phylogenetically assigned to a taxonomic identity (genus levels) using RDP classifier (Wang et al., 2007). Shannon diversity indices and species richness estimators of Chao 1 were generated for each sample using Quantitative Insights Into Microbial Ecology (QIIME) pipeline version 1.3.0 (http://qiime.org). The shift of microbial community structure was determined by measuring the type and relative abundance of bacteria in the three samples (IB-anae, IB-ae, and IB-anae/ae). The relative abundance was determined by high-throughput sequencing based on the summed reads per strain. 3. Results and discussion 3.1. Compositional shift in the microbial community during acclimation In the genus level, three samples (IB-anae, IB-ae, and IB-anae/ae) exhibited significant different about bacteria community richness (Fig. 1). The dominant genera in IB-anae were main anaerobic bacteria and facultative bacteria, like Vibrio, Bacillus, Enterobacter, and

W. Wan et al. / Ecological Engineering 99 (2017) 199–208 Table 2 The nitrogen and phosphorus removal efficiency from wastewater by strain HW-15. The results were the mean values of three replicates.

NO3 − -N NO2 − -N NH4 + -N TN TP COD

Control (mg/L)

Treatment (mg/L)

% Removal efficiency

5.12 ± 0.12 1.24 ± 0.09 41.48 ± 1.01 47.12 ± 2.14 48.76 ± 0.48 93.26 ± 1.69

0.61 ± 0.05 0.51 ± 0.06 0.56 ± 0.08 1.84 ± 0.23 13.34 ± 0.11 17.41 ± 0.56

88.09 ± 1.67 58.87 ± 2.13 98.65 ± 0.27 96.16 ± 1.01 72.64 ± 2.39 81.33 ± 3.02

Propionibacterium. By contrast, dominant genera in IB-ae were main aerobic bacteria facultative bacteria, such as Bacillus, Enterobacter, Acetobacte, Vibrio, Actinomyces, and Pseudomonas. While the genus with the highest relative frequency in IB-anae/ae was Enterobacter, followed by Vibrio and Actinomyces, they were facultative bacteria. Rhodospirillum, which presented low abundance in IB-anae and IBae, displayed significantly higher abundance in IB-anae/ae. While Halobacterium exhibited low abundance in these three samples. The microbial community structures of these three samples were different, which was due to the diverse culture condition. Among these genera mentioned above, the denitrifying phosphorus-accumulating bacteria of Bacillus and Pseudomonas have been reported (Kim et al., 2005; Li et al., 2015). In the acclimated community, Vibrio and Bacillus were dominant in IB-anae, Bacillus and Acetobacte were dominant in IB-ae, whereas their abundances dramatically decreased to low values in IB-anae/ae. However, Enterobacter obviously showed high abundance in the three samples. These phenomena demonstrated that oxygen was a key factor for bacterial growth, which resulted in aerobic bacteria dominating under aerobic condition and anaerobic bacteria dominating in anaerobic condition. To harvest facultative anaerobic DPAB, the anaerobic-aerobic training was feasible. At the end of acclimation, a single colony growing well was picked out from plate and the bacterium was named as HW-15 and stored for subsequent experiments. 3.2. Identification of strain HW-15 Granule staining demonstrated the accumulation of poly-P (Fig. 2a) and PHA particles (Fig. 2b) within bacteria cells. Strain HW15 was a gram negative stain and showed positive results in the tests for Voges-Proskauer, methyl red, and production of ammonia, utilization of sodium citrate, urea, tyrosine, and malonate. It was negative for production of H2 S, indole, and hydrogen peroxide, starch hydrolysis, and gelatin liquefaction. Based on the literature and physiological-biochemical characteristics, strain HW-15 was suggested to belong to the genus of Enterobacter (Dong and Cai, 2001). The result was consistent with that of Enterobacter was the dominant genus in anaerobic-aerobic acclimation. The 16S rDNA sequence of strain HW-15 was submitted to GenBank and analyzed against known gene sequences using BLAST. The 1503 bp fragment of 16S rDNA (Fig. S1) exhibited 99% similarity to the known Enterobacter cloacae (accession number: NR009854.1). Thus, strain HW-15 was identified as Enterobacter cloacae sp. by its physiological-biochemical characteristics and 16S rDNA sequence analysis. 3.3. Assessment of heterotrophic nitrification-aerobic denitrification of strain HW-15 Nitrogen removal capacity of strain HW-15 varied depending on exogenous carbon sources (Table 1). Sodium acetate was the most efficient carbon source for all three types of nitrogen removal and phosphorus removal, presenting a removal efficiency of 87.11%, 11.36%, 93.29%, and 65.56%, for NO- 3-N, NO- 2-N, NH+ 4-N, and

203

PO3- 4-P, respectively. These results were consistent with previous findings that sodium acetate has been used as external carbon source to remove nitrogen in some bacteria, such as Agrobacterium sp. LAD9 (Chen and Ni, 2012) and Acinetobacter sp. Y16 (Huang et al., 2013). In addition to the selection of proper carbon source, the data in this assay also demonstrated that the removal efficiency for the three nitrogen compounds was NH+ 4-N > NO- 3-N > NO- 2-N, and NH+ 4-N was significantly (p < 0.01) more consumed than the other two types of nitrogen sources. To investigate the effect of dissolved oxygen content on nitrogen removal, aeration was controlled by adjusting the shaking speed. As shown in Table 1, the maximum NO- 2-N removal efficiency (85.23%) was observed in the culture with the shaking speed of 50 rpm, similar with that of Bacillus subtilis A1 (Yang et al., 2011) and Pseudomonas stutzeri YG-24 (Li et al., 2015). The maximum NH+ 4-N removal efficiency was 93.23% at a shaking speed of 200 rpm. This result was consistent with those of Microbacterium sp. SFA13 (Zhang et al., 2013) and Marinobacter sp. F6 (Zheng et al., 2012) which showed higher NH+ 4-N removal as dissolved oxygen increased. The data showed that over 85% of NO- 3-N could be removed at the shaking speeds ≤ 100 rpm. Besides, the highest phosphorus removal efficiency (67.44%) was achieved at the shaking speed of 200 rpm, and phosphorus removal efficiency increased as shaking speed increased. These results were reasonable because the NO- x-N reduction demanded lower oxygen concentration compared with NH+ 4-N removal. Besides, these results demonstrated that, to remove various types of nitrogenous compounds effectively, the aeration condition must be adjusted according to different sources of wastewater in practical application when using strain HW-15. Nitrogen removal efficiency of strain HW-15 occurred at a wide range of temperature profile (Table 1). Similar and high NO- 3-N, NO- 2-N, NH+ 4-N, and PO3- 4-P removal efficiencies were achieved between 25 ◦ C and 35 ◦ C and these efficiencies were decreased at 20 ◦ C and 40 ◦ C. This phenomenon might be due to the fact that strain HW-15 was a mesophilic bacterium. It was reported that P. stutzeri YZN-001 (Zhang et al., 2011) could remove NH+ 4-N at 4–37 ◦ C, but the removal of NO- 3-N and NO- 2-N only occurred at 30 ◦ C. Strain HW-15 showed stronger adaptability to temperature and could maintain high nitrogen removal efficiency when the external temperature was constantly changing. The C/N ratio significantly (p < 0.05 or p < 0.01) affected the nitrogen and phosphorus removal efficiency (Table 1). The removal efficiency of nitrogen increased with the C/N ratio going up from 2 to 8, but a little decreased when at a C/N ratio of 10. The highest removal efficiency of 65.34%, 54.21%, and 93.23% for NO- 3-N, NO- 2-N, and NH+ 4-N, respectively, was observed at a C/N ratio of 8. The residual of nitrogen at low C/N ratio (e.g., 2 and 4) was mainly due to the exhaustion of the carbon source of heterotrophs. The optimal C/N ratio should be part of the metabolic characters, such as the pathway of NH+ 4-N oxidation and NO- 3-N, NO- 2N reduction, thus strain HW-15 achieved higher nitrogen removal efficiency at C/N ratio of 8 than that of at C/N ratio of 10. This optimal C/N ratio was similar to that of Bacillus strain (Kim et al., 2005) and Acinetobacter junii YB (Ren et al., 2014). Over 85% of NO- 3-N, NO- 2-N, and NH+ 4-N were removed in 24 h by strain HW-15 using initial nitrogen concentration of 25 mg/L and 50 mg/L (Table 1). At a nitrogen concentration of 100 mg/L, the nitrogen removal efficiencies by strain HW-15 were significantly decreased, but still maintained at acceptable levels: 65.34% for NO- 3-N, 54.21% for NO- 2-N, and 93.23% for NH+ 4N. Above all, strain HW-15 could effectively remove NO- 3-N and NO- 2-N at concentrations ≤ 50 mg/L, and NH+ 4-N at 15–100 mg/L. Wan et al. (2011) reported that Pseudomonas sp. yy7 could survive and remove up to 40 mg/L of NO- 2-N, but the cell growth was poor when 50 mg/L of NO- 2-N was added. In comparison, strain HW-15

204


Fig. 3. Cell growth and nitrogen, phosphorus removal characteristics of strain HW-15 in the heterotrophic nitrification medium containing ammonium (a) and aerobic denitrification medium containing nitrate (b) or nitrite (c). The results were the mean values of three replicates, error bars represented standard error.


205

Table 3 Mass balance during heterotrophic ammonium removal by strain HW-15 in the basic inorganic medium with and without NaClO3 or KClO4 addition. The results were the mean values of three replicates. Control

With NaClO3

With KClO4

Initial (mg/L)

NH+ 4-N NO- 3-N COD P

99.76 ± 1.13 2.01 ± 0.49 1479.2 ± 5.87 36.63 ± 1.19

96.73 ± 2.15 2.45 ± 0.32 1458 ± 7.33 36.42 ± 1.03

96.97 ± 3.24 2. 87 ± 0.59 1446 ± 8.97 36.57 ± 1.31

Final (mg/L)

NH+ 4-N NO- 3-N COD P Intracellular P Intracellular N

29.11 ± 1.02 17.14 ± 0.98 87.34 ± 3.28 15.36 ± 0.54 19.79 ± 0.97 15.11 ± 0.76

32.45 ± 1.24 14.39 ± 0.37 101.12 ± 4.23 17.69 ± 1.61 17.77 ± 0.53 13.54 ± 0.67

33.76 ± 0.58 12.58 ± 1.12 112.03 ± 3.34 18.48 ± 1.46 16.75 ± 1.08 14.8 ± 1.16

Initial Final

OD600 OD600 N lost (%)

0.05 0.54 ± 0.01 39.71 ± 2.47

0.05 0.51 ± 0.01 39.12 ± 0.53

0.05 0.48 ± 0.02 38.76 ± 1.31

Fig. 4. Cell growth and nutrient, arsenic(V) removal efficiency of strain HW-15 in different initial arsenic(V) concentration. The results were the mean values of three replicates, error bars represented standard error.

demonstrated high tolerance towards 400 mg/L of NO- 3-N, NO2-N, and NH+ 4-N.

3.4. Nitrogen and phosphorus removal in synthetic and real wastewater NH+ 4-N, NO- 3-N, or NO- 2-N together with PO3- 4-P in synthetic wastewater water was simultaneously removed by strain HW-15 (Fig. 3). In this analysis, the growth curves of strain HW15 presented by OD600 were similar in HNM media: a lag phase occurred in the first 4 h, followed by the exponential phase from the 4th to 12th h, then the stationary phase. The removal curves of NH+ 4-N, NO- 3-N, and NO- 2-N showed similar trend. NH+ 4-N got the highest removal rare (3.91 mg/(Lh)) (Fig. 3a), then followed by NO- 3-N (2.73 mg/(Lh)) (Fig. 3b) and NO- 2-N (2.34 mg/(Lh)) (Fig. 3c), after incubated for 24 h. Related to the growth of strain HW-15 and removal of nitrogen, the highest phosphorus removal efficiency was achieved in HNM-3 (67.44%), followed by 63.75% in HNM-1 and 49.75% in HNM-2. Strain HW-15 also had different removal efficiency in the removal of NH+ 4-N, NO- 3-N, and NO- 2-N (Fig. 3). During heterotrophic nitrification, 17.14 mg/L of NO- 3-N and 3.98 mg/L of

NO- 2-N were formed in 12 h because of the oxidation of NH+ 4-N and decreased to 3.22 mg/L and 0.34 mg/L, respectively, after 24 h of incubation. In the aerobic denitrification of NO- 3-N, NO- 2-N was observed, which reached to the maximum of 5.23 mg/L at the 12th h, and gradually decreased to 1.65 mg/L at the end of incubation. In the aerobic denitrification of NO- 2-N, 14.54 mg/L of NO3-N was detected at the 12th h because NO- 2-N was oxidized, and the concentration of NO- 3-N decreased to 3.15 mg/L. Although higher removal rate of NH+ 4-N (28.9 mg/(Lh)) was reported of Alcaligenes faecalis No. 4 (Joo et al., 2005), it could not remove NO- 3-N and NO- 2-N. Besides, the NH+ 4-N removal rate (3.91 mg/(Lh)) of strain HW-15 was higher than that of Rhodococcus sp. CPZ24 (3.1 mg/(Lh)) (Chen et al., 2012). Though NH+ 4-N removal rate was lower than P. stutzeri YZN-001 (5.53 mg/(Lh)) (Zhang et al., 2011) and P. stutzeri YG-24 (8.75 mg/(Lh)) (Li et al., 2015), the removal efficiency of phosphorus was 67.44%. The average NO- 3-N removal rate (2.73 mg/(Lh)) of strain HW-15 was much higher than Bacillus methylotrophicus strain L7 (5.81 mg/(Ld)) (Zhang et al., 2012) and Pseudomonas sp. yy7 (18.20 mg/(Ld)) (Wan et al., 2011). In addition, the average NO- 2-N removal rate (2.34 mg/(Lh)) was higher than that of Klebsiella pneumoniae CF-S9 (2.2 mg/(Lh)) (Padhi et al., 2013).

206


Fig. 5. Total nitrogen (TN) and total phosphorus (TP) profiles in real wastewater incubated with strain HW-15 (treatment) compared with the control (without inoculation). The wastewater was incubated at 30 ◦ C with shaking of 200 rpm. The results were the mean values of three replicates, error bars represented standard error.

Fig. 6. Nutrient removal characteristics and change in PHA content of strain HW-15. Strain HW-15 was cultured in heterotrophic nitrification medium under anaerobic-aerobic (12 h/ 12 h) condition. The results were the mean values of three replicates, error bars represented standard error.

The nitrogen, phosphorus, and COD removal efficiencies in real wastewater were shown in Fig. 5 and Table 2. In this assay, almost no NH+ 4-N, TP, and TN were removed in the control during 60 h, which illustrated that indigenous microbial population did not show significant (p > 0.05) impact on nitrogen and phosphorus removal. In contrast, TN and NH+ 4-N got higher removal efficiencies, while NO- 3-N and NO- 2-N achieved lower removal efficiencies in treatment (Table 2). Most of organic compounds were eliminated after 40 h in treatment and TP removal efficiency was acceptable. After 40 h, there were slight TN, TP, and NH+ 4-N increases (Fig. 5) caused by dead bacterial cells. The composition of real wastewater was both complex and variable, in which numerous types of ions and indigenous microbial groups affect the nitrogen and phosphorus efficiency. It was reported that P. stutzeri

YZN-001 (Zhang et al., 2011) could remove 85.91% of NH+ 4-N from domestic sewage in 72 h. The application of strain HW-15 in wastewater samples not only resulted in higher TN and NH+ 4-N removal efficiencies, but also contributed to obvious NO- 3-N, NO2-N, PO3- 4-P, and COD decrease. Therefore, strain HW-15 had a broad application prospect in terms of wastewater treatment. 3.5. Determination of effects of NaClO3 and KClO4 on nitrogen metabolism Wehrfritz et al. (1993) proposed the classical heterotrophic nitrification and aerobic denitrification (HNAD) coupling model with T. pantotropha, in which ammonium was sequentially oxidized to nitrite and nitrate, and then denitrified (NH+ 4 → NH2 OH → NO-


2 → NO- 3 → NO- 2 → N2 ). In order to understand whether the proposed pathway was applied to ammonia metabolism of strain HW-15, mass balance experiment was conducted. As shown in Table 3, the lost nitrogen as gaseous nitrogen was almost the same between treatments without and with NaClO3 or KClO4 added, and the concentration of NO- 3-N was slightly increased, indicating that the procedure from NO- 2-N to NO- 3-N was involved in the removal of NH+ 4-N by strain HW-15 indeed.

3.6. Simultaneous removal of nutrient and As(V) More than 80% nitrogen and 60% phosphorus were removed at As(V) concentrations lower than 1.5 mg/L (Fig. 4). Meanwhile, bacterial growth of strain HW-15 didn’t exhibit significant decrease (p > 0.05). At As(V) concentrations higher than 1.5 mg/L, nitrogen and phosphorus removal efficiencies sharply decreased and decreased to the minimum values of 27.48% and 19.76%, respectively, at an As(V) concentration of 3 mg/L. The As(V) removal efficiency was dramatically decreased and got the highest value at an As(V) concentration of 2.0 mg/L, and decreased to 14.12% at an As(V) concentration of 3 mg/L. The results demonstrated that both nutrient and As(V) were removed when As(V) concentrations lower than 1.5 mg/L. High As(V) concentration (higher than 1.5 mg/L) significantly (p