INTRODUCTION

Top Abstract Introduction Materials and Methods References

Dissimilatory nitrite reductase is a key enzyme in the denitrification process, in which nitrate undergoes stepwise reduction to the gaseous products nitrous oxide and dinitrogen (42). There are two distinct classes of periplasmic nitrite reductase: one containing cd₁ heme as the prosthetic group and the other containing copper (42). The copper centers in nitrite reductases comprise type 1 centers, giving rise to the blue or green color of these enzymes, and type 2 centers, which do not contribute significantly to the visible spectrum. Copper-containing nitrite reductases can be distinguished as either green enzymes or blue enzymes depending on the electronic absorbance of their type 1 copper centers. Thus, the enzymes that have been isolated from Achromobacter cycloclastes, Alcaligenes faecalis, Pseudomonas sp., and Rhodobacter sphaeroides are green enzymes that have absorbance maxima at ~460, 595, and 700 to 750 nm, while Alcaligenes xylosoxidans and Pseudomonas aureofaciens (42) have blue enzymes that show little absorbance in the 460-nm range. The X-ray structures of the Nir enzymes from A. cycloclastes (3, 11), A. faecalis (15), and A. xylosoxidans (6, 7, 14) have been determined at 2- to 2.1-Å resolution. This information, together with the deduced amino acid sequences of a number of nitrite reductases from various sources derived from the gene sequences, has been important in elucidating the type 1 and type 2 copper ligands (5, 9, 10, 22, 35, 36, 41). The availability of the high-resolution structures of the blue Nir enzyme from A. xylosoxidans has allowed comparison of the structures of the type 1 centers of the blue and green enzymes. The structures of the type 1 sites show the same ligands to the copper ions and little difference in the Cu ligand distances, including that of the Cu-S (Met) bond, which had been suggested to be responsible for the differences in color (3). The major difference is in the His-Cu-Met angle of 17° between the two classes of type 1 Cu sites (7), a finding consistent with a theoretical analysis of the factors likely to affect the electronic structure of such Cu sites (16). In addition, it has been proposed that the difference in color may result from a shift in the polypeptide backbone around the Met caused by a Tyr residue in the neighborhood of the type 1 copper center of the green enzymes being replaced with a Thr in the blue enzymes (14).

Work originally done by Libby and Averill (19) with the A. cycloclastes enzyme and later studies with the enzyme from A. xylosoxidans (1, 2) indicated depletion of the type 2 copper center (type 2-depleted [T2D] enzyme) was associated with a dramatic reduction in enzyme activity. Indeed, both native and recombinant enzymes appear to lose the type 2 Cu during isolation (42). Reconstitution of the center by incubation with exogenous copper restores enzyme activity and type 2 Cu electron paramagnetic resonance (EPR) signals (1, 2, 15). These results are commensurate with the type 2 copper center being involved directly in catalysis (13, 23). Evidence has been presented that indicates that the function of the type 1 Cu center is to receive electrons from the physiological electron donor pseudoazurin (15) and to transfer them to the type 2 center (8, 31, 32, 38).

In order to dissect the molecular events involved in electron transfer between the copper centers and the reduction of nitrite at the type 2 copper center, it would be a significant advantage to have an expression system that permits ready manipulation of the gene. This would facilitate the introduction of specific mutations and thus provide a rapid means by which the enzyme could be recovered in large amounts for biochemical analysis. We describe here the cloning of the nirA gene from A. xylosoxidans, the purification of the enzyme from the periplasmic fraction of Escherichia coli, and the characterization of the recombinant enzyme. Among the many advantages presented by this heterologous expression system is that the recombinant enzyme on a biochemical basis is essentially indistinguishable from the native enzyme. Moreover, recombinant Nir can be readily isolated in a form that is wholly deficient in type 2 copper centers. Significantly, the enzyme can also be isolated with a full complement of type 1 and type 2 centers simply by incubation with copper sulfate prior to purification.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods References

Strains and growth conditions. The bacterial strains used in this study were A. xylosoxidans subsp. xylosoxidans (NCIMB 11015) and E. coli JM109 (40) and BL21(DE3) (30). The E. coli strains were grown in Luria-Bertani medium. A. xylosoxidans was grown in a medium containing the following (per liter): nutrient broth, 8 g; NaCl, 0.5 g; yeast extract, 1 g; Na₂CH₃O₂, 5 g; NaNO₃, 5 g; and 1 µM concentrations of CuSO₄, Na₂MoO₄, MnSO₄, and FeCl₃. Growth was performed in standing cultures at 30°C, while E. coli strains were grown aerobically at 37°C in vigorously shaking conical flasks filled to a maximum of 10% of their volumes with medium. When used, glucose was added to a final concentration of 10 mM, and isopropyl--D-thiogalactopyranoside (IPTG) was used at a final concentration of 0.25 mM. Antibiotics were used at the following final concentrations: ampicillin, 50 mg · liter¹; kanamycin, 50 mg · liter¹.

Cloning of the A. xylosoxidans nirA gene. Two degenerate primers based on the amino acid sequence of the A. xylosoxidans NirA protein (36a) were designed. The forward primer corresponded to amino acid positions 33 to 38 of the protein and had the sequence 5'-AAGGARTTCACNATGAC-3', and the reverse primer, which corresponded to amino acid positions 254 to 259, had the sequence 5'-CCANACCCARTCNCCRTG-3'. N represents any nucleotide, while R represents an A or a G. A PCR was performed with 20 pmol of each of the above primers and 1 ng of chromosomal DNA isolated from A. xylosoxidans. The resulting ~680-bp DNA fragment was subcloned into SmaI-digested pUC19, yielding plasmid pUAX-1, and the authenticity of the DNA sequence of the insert was confirmed (27). In order to clone the wild-type nirA gene, 10-µg aliquots of chromosomal DNA from A. xylosoxidans were initially digested to completion with either BamHI or SalI and, after separation in a 0.8% (wt/vol) agarose gel, the DNA fragments were blotted onto a nitrocellulose membrane and hybridized with the ³²P-labelled DNA insert from pUAX-1 (26). Two signals of 3.0 and 2.1 kb were detected after BamHI digestion, and fragments of 1.4 and 1.2 kb were detected after SalI digestion. The 2.1-kb BamHI DNA fragment was successfully cloned into pUC19 digested with BamHI, yielding pUB1.

Overproduction and subcellular localization of NirA. Overexpression of the nirA gene was achieved by introducing pEnirsp-1 into BL21(DE3) (30). Four 0.5-liter cultures of the transformed strain were grown at 30°C until an optical density at 600 nm of 0.5 was attained. Subsequently, IPTG was added to a final concentration of 0.25 mM, and the culture was incubated for a further 90 min with vigorous agitation. The cells were harvested by centrifugation at 8,000 × g for 20 min at 4°C. The cell pellet was resuspended in 15 ml of 50 mM potassium phosphate buffer (pH 7.0) and centrifuged again at 8,000 × g for 20 min. Spheroplasts and the periplasmic fraction were prepared according to the method of Osborn et al. (24). Briefly, the cell pellet was resuspended in 10 mM potassium phosphate buffer (pH 7.0) at a concentration of 1 ml/g (wet weight) of cells. To each milliliter of cell suspension 5 ml of 1 M sucrose, 0.4 ml of 1 M Tris-HCl (pH 8.0), 0.4 ml of 0.1 M EDTA (pH 8.0), and 1.2 ml of a freshly prepared 5-mg/ml solution of lysozyme were added. The mixture was stirred gently for 2 min, and then 1 ml of 0.1 M MgCl₂ and 10 ml of H₂O were slowly added. The mixture was gently stirred at 30°C for 30 min. Spheroplasts were separated from the periplasmic fraction by centrifugation at 12,000 × g for 30 min. The spheroplasts were resuspended in 20 ml of 10 mM potassium phosphate buffer and disrupted by two passages through a French press at 16,000 lb/in² (1.03 × 10² MPa). The crude extract was prepared by centrifugation at 10,000 × g for 30 min.

Purification of mature, recombinant nitrite reductase. Mature, recombinant NirA was purified to apparent homogeneity from the periplasmic fraction of BL21(DE3) in a single chromatographic step with a carboxymethyl cellulose CM52 matrix (Whatman). A 110-ml (14 by 2.5 cm) column of CM52 initially equilibrated with 20 mM Tris-HCl (pH 7.0) was equilibrated with H₂O and 95 ml (~40 mg of protein) of periplasmic fraction was applied. Unbound proteins were removed by washing the column with H₂O. NirA appeared as a tight, dark blue band at the top of the column. Pure NirA was eluted with 20 mM Tris-HCl (pH 7.0) containing 50 mM NaCl. Subsequent to elution, NirA was dialyzed against 20 mM MES (morpholineethanesulfonic acid; pH 6.0) and concentrated to approximately 2 mg ml¹ (total volume, ~0.5 ml) by using a Centricon-10 concentrator (Amicon).

Reconstitution of type II copper sites in recombinant nitrite reductase. The periplasmic fraction was dialyzed at 4°C for 16 h against 1,000 volumes of 20 mM MES buffer (pH 6.0) containing 0.1 mM CuSO₄. The periplasmic fraction was then dialyzed exhaustively against three changes of 20 mM MES buffer (pH 6.0) with a minimum equilibration period of 5 h between buffer changes.

Analysis of nitrite reductase enzyme activity. Nitrite reductase enzyme activity was determined by using the discontinuous assay described previously (1). The thermostability of native and recombinant nitrite reductase was determined by measuring enzyme activity after incubation of 1 ml of a solution of the enzyme (40 µg of protein in 20 mM MES [pH 6.0]) in a water bath at 80°C and after 5-µl aliquots were withdrawn at various time intervals.

Spectroscopic methods. UV-visible spectra were recorded at room temperature on a Hewlett-Packard 8452A diode array spectrophotometer. EPR spectra were recorded at 60 K, at a microwave power of 20 mW, a microwave frequency of 9.3 GHz, and a modulation amplitude of 4.637 G on a Bruker ER ER200 D-SRC spectrometer fitted with an ER042 MRH microwave bridge and by using an ER033C field frequency lock and an Oxford Instruments Ite⁵⁰³ temperature controller. Integrated intensities were obtained by comparison with aqueous Cu-EDTA with a g value correction applied as described previously (1). Simulations were performed by using a computer program based on that of Lowe (20).

Other methods. The protein concentration was determined by the method of Lowry et al. (21). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis of proteins was performed as described previously (17). Specific radioactive labelling of polypeptides with [³⁵S]methionine was performed with strain BL21(DE3) as described previously (18, 33). Aliquots were removed different time points, and labelled polypeptides were analyzed by autoradiography after separation by SDS-PAGE. The copper content of purified nitrite reductase was determined on wet-ash samples by using induced coupled plasma emission spectroscopy exactly as described previously (1). Amino acid alignments were performed with the CLUSTAL W program (34).

	RESULTS AND DISCUSSION

Cloning and sequence analysis of the nirA gene. The amino acid sequence of A. xylosoxidans nitrite reductase has recently been determined by chemical analysis and mass spectrometry (37). To facilitate cloning of the complete nirA gene for a blue Nir, a DNA probe of the nirA gene was generated by PCR by using degenerate oligonucleotides, which were designed based on the Nir amino acid sequence, as described in Materials and Methods. Southern blot analysis of A. xylosoxidans chromosomal DNA failed to identify a suitably sized restriction fragment that carried the complete nirA gene (data not shown). Portions of the nirA gene could, however, be shown to reside on 3.0- and 2.1-kb BamHI fragments or on 1.4- and 1.2-kb SalI DNA fragments. Both the 2.1-kb BamHI fragment and the 1.4-kb SalI fragment could be readily cloned in pUC19. DNA sequence analysis of the BamHI fragment in pUB1 revealed that only the initial 548 bp of the nirA gene were present on the plasmid. It was not possible to clone either the 3-kb BamHI fragment or the 1.2-kb SalI DNA fragment in either a high- or a low-copy-number vector due to problems with plasmid instability. Instead, an inverse PCR approach with minicircles generated from the 1.2-kb SalI DNA fragment proved successful in amplifying the 3' portion of the nirA gene. The complete nirA gene was subsequently cloned on a single DNA fragment, and determination of the nucleotide sequence on both strands failed to reveal a discrepancy in the sequence of several clones. Moreover, all plasmids constructed that contained the complete nirA gene were completely stable. This strongly suggests that the inability to clone the 3-kb BamHI fragment or the 1.2-kb SalI DNA fragment must be due to DNA sequences on these fragments that lie outside the nirA coding sequence.

View larger version (69K): [in this window] [in a new window]   FIG. 1.   Alignment of the deduced amino acid sequences of copper-containing nitrite reductases from different bacterial sources. The amino acid alignment was generated with the CLUSTAL W package (34). Asterisks represent amino acid identity, and dots indicate similar amino acids. Similar amino acids include aromatic amino acids, basic amino acids, acidic amino acids, and hydrophobic amino acids. The numerals 1 and 2 above the sequence alignment signify the type 1 and type 2 copper ligands, respectively. ALCXY, A. xylosoxidans accession number AF051831; ACHCY, A. cycloclastes accession number Z48635 (9); ALCFA, A. faecalis accession number D13155 (22); PSEsp, Pseudomonas sp. strain G-179 accession number M97294 (41); PSEAR, P. aureofaciens accession number Z21945 (10); RHIHE, R. hedysari accession number U65658 (35); RHOSPH, R. sphaeroides accession number U62291 (36).

Overproduction and subcellular localization of Nir in E. coli. The nirA gene was subcloned from pUnirsp-1 into pET28a on a SacI-HindIII fragment (see Materials and Methods). The resulting plasmid (pEnirsp-1) was transformed into BL21(DE3), and the polypeptides derived from T7 RNA polymerase-directed transcription of the nirA gene were detected (Fig. 2A). After electrophoretic separation in a 12.5% acrylamide gel, two polypeptides with apparent molecular masses of 43 and 35 kDa were observed. The N-terminal amino acid sequence of the smaller polypeptide was determined to be Q-D-A-D-K-L, which is in perfect agreement with that predicted for the mature Nir protein lacking the signal sequence. Thus, mature recombinant Nir (35 kDa) migrated with a molecular mass that was in close agreement with the DNA-deduced molecular mass of 36.5 kDa. Clearly, E. coli, unlike A. xylosoxidans, is unable to modify the N-terminal glutamine residue of Nir to pyroglutamate.

View larger version (56K): [in this window] [in a new window]   FIG. 2.   Nitrite reductase from A. xylosoxidans is exported to the periplasm in the heterologous host E. coli BL21(DE3). (A) Subcellular localization of recombinant Nir in E. coli BL21(DE3). Polypeptides (50 µg of protein per lane) were separated by SDS-PAGE in a gel containing 12.5% (wt/vol) acrylamide. Lanes: 1, whole-cell extracts; 2, insoluble material from the crude extract; 3, the periplasmic fraction. The minus sign indicates subcellular fractions derived from BL21(DE3)/pET28a, and the plus sign indicates subcellular fractions derived from BL21(DE3)/pEnirsp-1. The location of the mature Nir protein is indicated. The migration positions of the molecular mass markers are shown on the left of the diagram. (B) Pulse-chase experiment showing that mature Nir production is independent of the His-tagged Nir fusion protein. Samples were removed at the time points indicated, and the [35S]methionine-labelled polypeptides were separated by SDS-PAGE in 12.5% (wt/vol) acrylamide gels. After being dried, the gels were exposed to X-ray film. The migration positions of the insoluble His6-Nir fusion protein, the Nir precursor, and mature Nir are shown.

Purification and spectroscopic analysis of mature recombinant Nir. Nir was purified from the periplasmic fraction prepared from approximately 2 g (wet weight) of cells. The specific nitrite reductase activity in the periplasmic fraction was 1.35 µmol of nitrite reduced min¹ mg of protein¹. After activation with copper sulfate the specific activity increased to 5.48 U mg of protein¹ for the periplasmic fraction. Purification of nitrite reductase from the periplasmic fraction was achieved in a single step after chromatography on carboxymethyl cellulose (Fig. 3). Purified nitrite reductase was dark blue and had a specific activity of 167.7 U mg of protein¹. It was not possible to determine the yield of Nir recovered after purification because there was a discrepancy between the total activity of the periplasmic fraction (18.3 U) and the total activity of the purified enzyme (160 U). It appears that an unidentified activity present in the periplasmic fraction of E. coli BL21(DE3) interferes with the accurate determination of nitrite reductase enzyme activity as measured by the methyl viologen-linked assay.

View larger version (63K): [in this window] [in a new window]   FIG. 3.   Purification of recombinant nitrite reductase. The photograph shows polypeptides separated in a 12.5% (wt/vol) acrylamide gel and stained with Coomassie brilliant blue R-250. Lanes: 1, molecular mass markers (Sigma wide molecular-weight range); 2, crude extract from BL21(DE3)/pET28a (60 µg of protein); 3, crude extract from BL21(DE3)/pEnirsp-1 (50 µg of protein); 4, periplasmic fraction from BL21(DE3)/pET28a (20 µg of protein); 5, purified Nir after carboxymethyl cellulose chromatography (5.5 µg of protein).

1

View larger version (15K): [in this window] [in a new window]   FIG. 4.   UV and visible absorption spectrum of recombinant nitrite reductase. Spectra were recorded in 20 mM MES buffer (pH 6.0) at room temperature. The spectrum was recorded at a 0.75-mg/ml protein concentration.

View larger version (12K): [in this window] [in a new window]   FIG. 5.   EPR spectra of type 2-deficient and reconstituted recombinant nitrite reductase. The experimentally obtained spectra are shown and were recorded at 60 K and at a 10-mW microwave frequency of 9.312 GHz. The simulation (not shown) determined the type 1 Cu to have g"  1.24, g  2.05, and A"  6.3 mT and the type 2 Cu to have g"  2.38, g  2.05, and A"  12.7 mT. Approximately equal amounts of type 1 and type 2 copper were determined to be present in the reconstituted sample. The protein concentration of the sample was 9.4 mg/ml for the "as-purified" enzyme (255 µM monomer) and 4.8 mg/ml for the reconstituted enzyme (135 µM monomer). The location of the features corresponding to the type 1 and type 2 Cu centers is shown by the stick diagram.

Thermostability of recombinant and native nitrite reductases. Nir when purified from A. xylosoxidans is heat stable and when heated at 80°C retains 50% of the initial activity after a 20-min exposure to these potentially denaturing conditions. This resistance to denaturation may arise from the relatively large area of the monomer-monomer interface, which the crystal structure shows to be 4,500 Å², and is stabilized by extensive interactions (7). The purified recombinant enzyme showed a similar temperature stability (data not shown). This, together with the similarity of the apparent K_m for nitrite (35 µM at pH 7.5 [1]) and activity pH profile, indicates that the properties of the catalytic sites and the global features responsible for the thermostability of the enzyme are retained in the recombinant enzyme.

Conclusions. We were able to isolate approximately 1 mg of pure recombinant Nir from the periplasmic fraction of E. coli BL21(DE3) cells derived from a 2-liter culture. Kinetic and thermostability analyses demonstrated that the recombinant enzyme was indistinguishable from the enzyme isolated directly from A. xylosoxidans. Based on EPR spectroscopy and metal analysis the enzyme that had not been activated with CuSO₄ prior to purification was found to possess only type 1 copper. As a consequence, this enzyme had low nitrite reductase enzyme activity. Reconstitution of the type 2 copper sites was achieved by incubation of the periplasmic fraction with CuSO₄ prior to enzyme purification. This enzyme was trimeric, had six copper atoms per trimer, exhibited an EPR spectrum with characteristics of type 1 and type 2 copper centers and had a high nitrite reductase specific activity. Taken together with the fact that processing of the signal peptide in the heterologous host occurred at the same site as in A. xylosoxidans, these data indicate that the recombinant enzyme is indistinguishable from the native protein (1). This newly developed isolation procedure with E. coli will greatly facilitate the analysis of the ligands to, and the pathway of electron transfer between, the copper centers. Moreover, the ability to isolate "clean" T2D enzyme will permit analyses of the electron transfer reactions to the type 1 center.