Introduction

Nitrite accretion was harmful to microbes and it can be reduced to nitric oxide (Ezzine and Ghorbel 2006). Denitrification was a process eliminates the nitrite accumulation in order to reduction of nitrate, nitrite, nitric oxide (NO) and nitrous oxide (N2O) by nitrate reductase, nitrite reductase, NO reductase and N2O reductase, respectively (Alvarez et al. 2014; Tiedje et al. 1988). Denitrifying bacteria are well represented among members of the genus Bacillus (Verbaendert et al. 2011). Bacillus was a classic envoy with not only the strong ability of water colonization and also environmental discharge safety, indicates that has a wide application in nitrogen pollution control (Mania et al. 2014).

Reduction of nitrite to nitric oxide using nitrite reductase was considered to be the rate limiting step in the denitrification process (Santoro et al. 2006). Two types of nitrite reductase with distinct molecular structures and prosthetic metal have been discovered as a copper nitrite reductase encoded by the nirK gene and cytochrome cd1-nitite reductase encoded by the nirS gene that was the key enzyme and forms the unique feature between denitrifies and nitrate reducers (Hoffmann et al. 1998). Denitrifying bacteria contain either nirK or nirS (Jones et al. 2008). The Cu-containing nitrite reductase (CuNiR) formed by nirK gene was distributed widely and molecular structure were studied well in denitrifying gram-negative bacteria (MacPherson and Murphy 2007; Lintuluoto and Lintuluoto 2016; Li et al. 2015; Ezzine and Ghorbel 2006; Berks et al. 1995). CuNiR contains type-1 Cu centre was bounded by two histidine, cysteine and methionine ligand, even as the indistinct tetrahedral type-2 Cu centre was corresponding by three histidine ligands and a single water molecule (Tocheva et al. 2004). The functionality of CU containing nitrite reductase shown in Fig. 1, NO2 − eliminates water and transferred to the CuNIR of Cu(II) centre via its two oxygen atoms. Reduction of the Cu(II) centre, later on occurs as consequence of electron transfer from a Cu(I) centre, it’s probably due to the stimulation of protonation with nearby aspartate residue. Sometimes, the bound nitrite was followed to protonated at the oxygen that is to be distracted, subsequently lysis of N–OH bond. After the N–OH was broken, protonation of the Cu-OH and it leads to NO discharge subsequently formation of a Cu(II)–H2O composite. Later the water composite can be able to re-enter into catalytic system and perform its function in the reduction of nitrite (Timmons and Symes 2015). Fukuda et al. 2011 reported molecular and structure analysis of CuNiR in Gram-positive bacteria of Geobacillus kaustophilus HTA426. However, the report didn’t show the enzyme characteristics of CuNiR in Bacillus. Therefore, it is important to characterize CuNiR in Bacillus.

Fig. 1
figure 1

Source of the figure from Timmons and symes (2015)

The Mechanism of action of nitrite reduction in CuNIR.

Full size image

In the present study Bacillus firmus GY-49 was isolated from the natural environment and it showed denitrification of nitrate as terminal electron acceptor. Here we have studied the biochemical properties of CuNiR encoded by the nirK-homolog gene, which is found in the B. firmus GY-49 genome The gene was cloned and expressed in Escherichia coli, and its molecular and catalytic properties characterized. Further, the structure of CuNiR was studied based on the sequence and structure analysis of well-known CuNiRs. The results showed that the nirK was really an ortholog of nirK in the denitrifying bacteria. This was the first report on the enzyme characteristics of CuNiR in B. firmus.

Materials and methods

Strains and plasmids

Bacillus firmus GY-49 (GenBank: KF982018.1) have been used to amplify the nirK gene. The E.coli strains DH5α and BL21 (DE3) (Invitrogen Co. Ltd., Shanghai, China) were employed for recombinant plasmid construction of the nirK gene. The pMD-19T vector (TaKaRa Biotechnology Co. Ltd., Dalian, China) and pET-28a (+) plasmid (Novagen Co. Ltd., Shanghai, China) were used for gene cloning and heterologous expression of nirK in E. coli BL21 (DE3).

Cloning and construction of recombinet vector

The specific primers for PCR amplification were designed based on gene encoding the putative nitrite reductase genome information of B. firmus GY-49. The putative protein was encoded by 353 amino acids and the N-terminal region from Met1 to Ala26 was considered to be a signal sequence analyzed by SignalP 4.1, which can predict the location of signal peptide cleavage sites in amino acid sequences (http://www.cbs.dtu.dk/services/SignalP/) (Emanuelsson et al. 2007). The nirK gene without signal peptide was amplified with the following primers nirK F: 5′-CCGGAATTCATGCCGGAAAATTCCCTTCA-3′ and nirK R: 5′-CCCAAGCTTTTAATGGCTCATTGTAGCCT-3′ with restriction enzyme sites of EcoRI and HindIII, (Italics). PCR was performed in a thermal cycler programmed with 5 min at 94 °C, 30 s at 94 °C, 30 s at 50 °C, 100 s at 72 °C, and a final elongation 10 min at 72 °C. Then full-length gene was cloned into pMD19T, followed to transform into DH5α and obtained nirK-PCR.

The purified fragment of nirK-PCR was ligated to the pET28a (+) expression vector at corresponding restriction sites and obtained expression vector pET-NirK. To ensure the recombinant protein could express as expected, a Met residue was inserted prior to N-terminus (before the Phe27 residue of NirK). The Phe residue was quoted as first amino acid of the recombinant protein for subsequent clarity. The pET-NirK was transformed into the E. coli BL21 (DE3) competent cells for expression. A final concentration of 0.1 mM isopropyl-β-d-thiogalactoside and 1 mM CuSO4 for the activation of the recombinant protein was added when the optical density at 600 nm (OD600) reached 0.5. To acquire the functional version of cells were further incubated at 4 °C for 72 h after induction at 16 °C for 4 20 h. Next, the cells were collected and washed twice with PBS buffer (0.8% NaCl, 0.02% KCl, 0.142% Na2HPO4, 0.027% KH2PO4; pH 7.4) by centrifugation at 8000 rpm for 10 min and resuspended in PBS buffer and then the cells were disrupted by a French press and the crude enzyme was obtained as supernatant by centrifugation at 12,000 rpm for 40 min at 4 °C, followed to filtration by 0.22 μm filter membrane. The crude enzyme and the precipitation of insoluble protein were used for SDS–PAGE analysis.

Purification of the recombinant protein

The crude enzyme was loaded into Ni–NTA column (QIAGEN, China) pre-equilibrated in ten column volumes of 20 mM Tris–HCl buffer (pH 7.4) containing 250 mM sodium chloride and 1 mM phenylmethylsulfonyl fluriode (buffer A). Then column was washed with five column volumes of same buffer A containing 20 mM imidazole and the buffer A containing 200 mM imidazole was applied to elute the recombinant protein. The fractions containing NirK were analyzed by SDS–PAGE.

In vitro activity assay for purified recombinant protein

Nitrite reductase activity was determined by using methyl viologen (MV) as an electron donor described by Abraham et al. 1993. The reaction was carried out (0.5 mL) in the closed vial containing 100 mM potassium phosphate buffer (pH 6.5), 50 mM methyl viologen and 2.5 M sodium nitrite, 1 M sodium chloride and 0.02 mL purified enzyme (Kataoka et al. 2000). The reaction was initiated with the addition of 50 μL sodium dithionite (1 M) and it was gently swirled. After incubation for 10 min at 37 °C, the reaction was terminated by vigorous shaking for 3 min. The rest of nitrite was determined by spectrophotometrically using diazo-coupling method of Nicholas and Nason (1957). One unit of activity corresponded to the reduction of 1 μmol of nitrite per minute.

Enzyme characterization

The optimum pH was determined by incubating the purified NirK at 35 °C in potassium phosphate buffer (5.7–8.0). The nitrite reductase activity was assayed by incubating 30 μL of enzyme solution in 50 μL of different pH buffers containing sodium nitrite at 35 °C for 10 min. For the effect of temperature, the activities of NirK was determined at different temperatures ranging from 20 to 60 °C. To study its thermostability, the enzyme was incubated at various temperatures (40, 45, 50, and 55 °C) in different time intervals (0–120 min). The enzyme stablility at different pH and temperatures were determined by measuring the residual activity of NirK in potassium phosphate buffer (pH 6.5) at 35 °C for 10 min.

The effects of metal ions and inhibitory additives were determined by pre-incubating the enzyme with 5 mM solution of the following Na+, K+, Cu2+, Ba2+, Mg2+, Zn2+, Mn2+, Sr2+ and EDTA (ethylenediaminetetraacetic acid) and PMSF (phenylmethanesulfonyl fluoride) for 10 min to determine the residual activity. The nitrite reductase activity without additives was defind as 100%.

The kinetic properities of NirK were determined in the potassium phosphate buffer (pH 6.5) with different concentration of sodium nitrite at 35 °C for 10 min. The K m and V max values were calculated with Graphpad Prism software (Graphpad, San Diego, CA) according to non-linear regression plot using Michaelis–Menten equation. The k cat was determined by the equation k cat = V max/[E]. All the assays were carried out in triplicate and against a control assay without enzyme.

Properties of absorption and structural modeling

Spectroscopic measurements were performed in the UV–visible regions using 220A Spectrophotometer (Lambda 25, Perkin Elmer, America) with a 1-cm long light path cuvette. According to Fukuda et al. (2011) and Arnold et al. (2006) structural modeling of NirK amino acid sequence without signal peptide was predicted on the Swiss-Model workspace to build a structure from G. kaustophilus (Accession No. 56419302).

Others

The protein concentration was determined by a modified lowry method using bovine serum albumin as the standard (Dulley and Grieve 1975). Amino acid sequences alignment was performed with the ClustalW program (Thompson et al. 1994). The complete cDNA sequence of nirK in B. firmus GY-49 has been submitted to GenBank database (Accession No. KT899614).

Results

Cloning and sequence analysis of nirK gene

The nucleotide sequence of B. firmus GY-49 was revealed the presence of an open reading frame of 1062-bp encoding CuNiR, composed of 353 amino-acid residues with a signal peptide. The full-length 993-bp cDNA of nirK without signal peptide was obtained through amplification (Figure not shown). The molecular weight of the recombinant protein was 36.41 kDa and predicted isoelectric point (pI) was 5.35. The amino acid sequence of NirK has highest similarity 99 and 78% with B. firmus and Geobacillus stearothermophilus in the NCBI.

Multiple sequence alignment showed T1Cu was integrated in each monomer and coordinated by four amino acid chains (two His imidazoles, one Cys thiolate and one Met thioether). The T1Cu and T2Cu was coordinated with two monomers and harmonized by three amino acid chains (two His imidazoles and one His imidazole from one subunit and the adjacent subunit, respectively). The two Cu sites at a distance of 12.5 Å, but are connected via the His-Cys sequence segment composed of the Cys residue of T1Cu and His residue of T2Cu (Godden et al. 1991). The amino acid sequence of NirK showed high identity to other well-known CuNiRs (Fig. 2). As shown in Fig. 2, these amino-acid sequences have been exposed that obvious conserved domain between each well-known CuNiR. The present results were agreed with previous studies of renowned CuNiRs, which were conserved in the active site of T1Cu and T2Cu formed by His125, Cys165, His173 and Met178 and His130, His164 and His324 residues.

Fig. 2
figure 2

Alignment of amino-acid sequences of CuNiR from different bacterial sources. The red section represents amino acid identity, and the similar amino acid including aromatic amino acid, basic amino acid, acidic amino acid, and hydrophobic acids. The filled triangle and circles under the sequence alignment signify the T1Cu and T2Cu ligands, respectively. The sequence for CuNiR was aligned with the following proteins (Accession numbers KT899614, Q5L1 × 8, 1KBWA, 1HAUA); Open circles indicate amino acid residues involved in a hydrogen bonding network around T2Cu. The putative linker, loop tower and extra loop regions are in the box. The amino acid alignment was performed by Clustal-W package and the figure was generated using the program ESPript (Abraham et al. 1997; Gouet et al. 1999; Ladunga 2002)

Full size image

Expression and purification of nirK gene

The CuNiR from B. firmus GY-49 was successfully expressed heterologously in E. coli BL21 (DE3) and the recombinant protein was purified by Ni2+-NTA affinity chromatography. The CuNiR was analyzed by SDS–PAGE (Fig. 3). The Fig. 3a showed a band of crude enzyme with molecular weight of approximately 37 kDa (amino acid sequences of the mature protein). The purified CuNiR was appeared two protein bands whose molecular weights to be 37 and 34 kDa with an approximately 1:1 stoichiometry, as shown in Fig. 3b.

Fig. 3
figure 3

a SDS-PAGE analysis of the recombinant crude enzyme. M: The protein molecular weight marker; Lane 1 The crude enzyme of recombinant strain E. coli BL21 (DE3) with pET-28a (+) as control; Lane 2 The crude enzyme recombinant strain E. coli BL21 (DE3) with pET-28a-NirK, b SDS-PAGE gel of the purified recombinant enzyme. Lane 1 Purified recombinant protein

Full size image
Fig. 4
figure 4

UV and visible absorption spectrum of recombinant nitrite reductase. Purified NirK (0.91 mg/mL) was dissolved in 20 mM Tris–HCl buffer (pH 8.0) at room temperature. The absorbance of the enzyme at around 280 nm was reduced to a scale of 1/10 the natural intensity

Full size image

Characterization of purified recombinant protein.

The purified enzyme appears bluish green in oxidized state and it becomes lighter color in reduction with Na2S2O4, indicates presence of T1Cu in the recombinant enzyme. The absorbance spectrum displayed two peaks at 456 and 608 nm in the spectrum of the visible region in Fig. 4. Similarly, the enzyme produced by Nitrosococcus oceani was showed two absorption peaks at 455 and 575 nm (Kondo et al. 2012).

Optimum pH, temperature and metal ions

Nitrite reductase activity was measured by using MV and sodium nitrite act as an electron donor and substrate. More prominently the amalgamation of Cu ions was essential for increasing enzyme activity since Cu ions could contribute to the reconstruction of T2Cu (Kondo et al. 2012; Prudencio et al. 1999). The optimum pH of NirK was found to be at 6.5 and exhibited 93 and 78% of its retaining activity at 5.7 and 8.0, respectively (Fig. 5). Purified NirK showed broad activity range of a temperature from 20 to 60 °C (Fig. 6a). The enzyme showed maximum activity at 35 °C and retained more than 83 and 41% activity at 20 and 60 °C. In order to further study, the thermostability was studied in the range of a temperature from 40 to 55 °C (Fig. 6b). After 2 h incubation, NirK could retain more than 89.7% activity at 40 °C, but only about 24.35% activity was retained at 50 °C. However, it was almost inactive after 2 h incubation at 55 °C. The results showed that the enzyme has high catalytic efficiency at standard temperature, but it would become unstable at high temperatures. The purified recombinant protein had a specific activity of 98.4 U/mg in the periplasmic fraction under the optimal conditions. The present results were agreed with the existing of previous reports (Hira et al. 2012; Prudencio et al. 1999). In case of metal ions Na+, Mg2+, Mn2+, Cu2+, Sr2+ showed increase the activity of NirK by 3.26–34.91% (Table 1). NirK was less affected and retained about 89.09% of the original activity with certain concentration of Ba2+, but it could inhibited by same concentration of Zn2+. Interestingly, NirK was not sensitive to inhibitory additives (EDTA and PMSF) and it exhibits more than 80% activity (Table 1).

Fig. 5
figure 5

The optimal pH and stability of the recombinant NirK. a The effect of pH on NirK activity. The activity was assayed at 35 °C in buffers containing a pH range of 5.7–8.0. b The effect of pH on the stability of NirK. The activity was measured under the standard procedure after the enzyme was incubated in different pH buffers at 35 °C for 2 h. Each value represents the average of triplicate experiments. Errors bars represent the standard deviation

Full size image
Fig. 6
figure 6

The optimal temperature and stability of the recombinant NirK. a The effect of temperature on NirK activity. The activity was assayed at pH 6.5 at 20–60 °C. b The effect of temperatrue on the stability of NirK. The activity was measured under the standard procedure after the enzyme was incubated at 40, 45, 50, 55 °C in different time intervals and the residual activity was measured according to the standard conditions. Each value represents the average of triplicate experiments. Errors bars represent the standard deviation

Full size image
Table 1 Effects of various metal ions and inhibitory additives on NirK
Full size table

Kinetic studies

NirK kinetic parameters, K m and K cat were studied under standard method and conditions at pH 6.5 and 35 °C. The values of K m and K cat were 0.27 mM and 0.36 × 103 s−1, respectively (Fig S1). The catalytic efficiency of NirK could be counted by K m/K cat with a computed value of 1.3 × 103 s−1 mM−1. The present Km values were similar to a green-type of NirK from Alcaligenes  faecalis and sinorhizobium meliloti 2011 (Kukimoto et al. 1994; Ferroni et al. 2014).

Structural evolution

The structural modeling of CuNiR from B. firmus GY-49 was build with the G. kaustophilus nitrite reductase as template. The overall structure for NirK in B. firmus GY-49 could be folded into two cupredoxin domains of the well-known CuNiR structure and each monomer consist of α-helix, β-pleated sheet and random helix. Generally, these conserved residues (H125, H173, C164, M178, H164, H130, H324, H274, D128) shown as firewood in Fig. 7a, were hypothetically to create the active catalytic sites of NirK. The confirmed structure was located in the southern surface of T1Cu and functioned as the form of α-helix structure (Fig. 7a). In addition to that space occupied by the structure was corresponding to the interaction surface with the physiological redox-partner proteins (Nojiri et al. 2009). Previously reported structure analysis of G. kaustophilus suggested that the N-terminal α-helix could play an important role for effective electron transfer in physiological redox-partners pairs (Fukuda et al. 2014). However, similar residues composed of N-terminal α-helix in B. firmus GY-49 are shorter than those in G. kaustophilus, which is colored in gray (Fig. 7b). Superimposing the predicted structure of CuNiR from this study on the established for comparison, lead to following observations. N-terminal region of the T1Cu and overall structure of CuNiR from B. firmus GY-49 were well relation with tertiary structure of CuNiR from G. kaustophilus. Moreover it shows that there is no connection between the folding of the core region of CuNiR and length of N-terminal region in the B. firmus GY-49. The results implied that the novel structure may be existing in CuNiR in all Bacillus, which may be the subject of further attempts at crystallizing the enzyme.

Fig. 7
figure 7

Homology modeling of the NirK structure. Homology modeling of NirK from G. kaustophilus. a Protein secondary structure showed α-helix, β-pleated sheet and random coil were colored by red, yellow and blue, respectively. In particularly residues (H125, H173, C164, M 178, H164, H130, H324) are involved in the formation of T1Cu and T2Cu site of NirK. Moreover, the “Linker”, “Tower” and “Extra” loops are colored by blue, orange and pink, respectively. N-terminal region formed α-helix structure is signed by light-gray and the dashed-line circle indicates the space occupied by the N-terminal region in other well-known CuNiRs. This figure was generated using the PYMOL program (DeLano 2002). b Superimpose of the CuNiR from B. firmus GY-49 (green) and G. kaustophilus (gray). The structure of the main chain is almost the same between the two structures. The red line represents the difference in α-helix structure between two Bacillus groups

Full size image

Discussion

To date, CuNiR produced by B. firmus was the first report and it belongs to Gram-positive bacteria. The literature showed only one eminent CuNiR from G. kaustophilus HTA426 (Fukuda et al. 2011) which was functionally verified and present results showed 72.73% similarity. All CuNiRs has trimeric structure and each monomer consisted of two types of Cu that plays a very important role in catalysis, which was defined as type 1 (T1Cu) and type 2 (T2Cu), respectively. Generally the active site of T1Cu could accept an electron from external electron transfer protein and was accepted by T2Cu in electron transfer process of CuNiRs (Iwasaki et al. 1975; Kakutani et al. 1981).

Hira et al. (2012) reported that formation of T1Cu by His122, Cys159, His168 and Met173 conserved sequence and His127, His158 and His321 for T2Cu in Anammox organism KSU-1. In addition Asp128 and His274 were identified in the sequence of CuNiR, suspecting that these residues may involve forming hydrogen bond during proton transform to NO−. The SDS–PAGE analysis showed two bands, in which second lower band was probably a degradation product of the larger subunit was similar to other nitrite reductase (Inatomi and Hochstein 1996; Rosa et al. 2001). The present results were related with previous report of Suzuki et al. 1999. Based on the T1Cu absorption characteristics CuNiRs has been divided into green and blue enzyme. This was due to the geometrical distortion of the T1Cu (LaCroix et al. 1996). In contrast to the nitrite reductase from Alcaligenes xylosoxidans showed a single absorption peak at 595 nm indicates blue enzyme (Prudencio and Sawers 1999). As well as the ratio (A455/A608) of NirK from anammox bacterium KSU-1 and A.faecalis was 1.03 which was reasonable to NirK (Wijma et al. 2006). The results implied that the CuNiR from B. firmus GY-49 resemble that green-type nitrite reductase belong to class II (Kobayashi and Shoun 1995; Suzuki et al. 1999; Leferink et al. 2012; Nojiri et al. 2009; Hoffmann et al. 1998).

Moreover CuNiR can be divided into two distinct groups, class-I and Class-II (Boulanger and Murphy 2002). A characteristic difference between the class I and class II groups was the length of the linker loop and tower loop. The tower loop in class I CuNIRs extends to the upper surface of the type 1 Cu site and was important for docking with the electron transfer partners. There are partial deletions in the linker loop and tower loop of class II (Boulanger and Murphy 2002). Sequence alignment analysis of NirK from B. firmus GY-49 showed linker loops and tower loops of Class-II CuNiR were shorter than those found in Class-I CuNiR. Interestingly, the Class-II CuNiR from B. firmus GY-49 and G. kaustophilus exhibited similar characteristics of linker loops and tower loops, its due to the related or close physiological and genetic relationship between these two Gram-positive bacterium. Furthermore a distinctive extra loop was found in B. firmus GY-49 as well as in G. kaustophilus. The extra loop consisted of 12 residues were located in the N-terminal region of CuNiRs (Fukuda et al. 2014). Likewise presents of N-terminal region (extra loop) in B. firmus GY-49 was a novel and unique structure for Gram-positive bacteria. Moreover the interaction and electron transfer mechanisms for CuNiR and relevant physiological electron donor protein was well studied in denitrifying bacteria (Dodd 1995; Kabayashi and Shoun 1995; Liu et al. 1986). The physiological electron donor of CuNiR from B. firmus GY-49 could not be defined, However, it is worth mentioning that we have found the presence of Cytochrome c551 instead of pseudoazurin or azurin analogues in the genome of the B. firmus GY-49. Therefore CuNiR protein acts as a physiological electron-donor between the electron transfer reactions through its acidic residues located around the T1Cu site. In order to confirm the above statement, further research was continued about mechanism of action of NirK.

Conclusion

The spectroscopic and molecular properties of the enzyme imply that NirK is a green-type CuNiR. Predominantly, the recombinant enzyme displayed a higher activity than those in the G. kaustophilus HAT426, and the enzyme was a moderate and organic solvents tolerant reductase with remarkable stability, it was proposed to be a potential microecologics for wastewater treatment. Furthermore, homology analysis of CuNiR in B. firmus GY-49 shown a few new characteristics, including extra loop and N-terminal α-helix around the T1Cu site. CuNiR possessed an N-terminal α-helix, which can greatly contribute to the interaction between the partner protein and CuNiR. The present results of the CuNiR have great value to the application of nitrite reduction in polluted water.