Structural conservation of the B subunit in the ammonia monooxygenase/particulate methane monooxygenase superfamily

Abstract

The ammonia monooxygenase (AMO)/particulate methane monooxygenase (pMMO) superfamily is a diverse group of membrane-bound enzymes of which only pMMO has been characterized on the molecular level. The pMMO active site is believed to reside in the soluble N-terminal region of the pmoB subunit. To understand the degree of structural conservation within this superfamily, the crystal structure of the corresponding domain of an archaeal amoB subunit from has been determined to 1.8 Å resolution. The structure reveals a remarkable conservation of overall fold and copper binding site location as well as several notable differences that may have implications for function and stability.

INTRODUCTION

The copper membrane-associated monooxygenases (CuMOs) are a diverse superfamily of multisubunit prokaryotic enzymes that oxidize a range of compounds, including ammonia and hydrocarbons.¹ The founding members are ammonia monooxygenase (AMO)² and particulate methane monooxygenase (pMMO),³ both produced by gram-negative bacteria. AMO converts ammonia to hydroxylamine in nitrifiers, and pMMO oxidizes methane to methanol in methanotrophs. Hydroxylamine and methanol are then further metabolized and serve as the primary energy source for the host organisms.^4,5 A recent explosion in genomic data has led to the detection of numerous other homologs, including archaeal AMOs,^6,7 putative butane monooxygenases (BMOs),⁸ hydrocarbon monooxygenases from gram-positive Mycobacteria^9,10 and marine isolates (pHMOs),^11,12 and pXMO systems, for which the substrate is unknown (Fig. 1A phylogenetic tree).¹

Figure 1.

Members of the AMO/pMMO superfamily. A: Phylogenetic tree. β-AOB, ammonia oxidizing β-proteobacteria; γ-AOB, ammonia oxidizing γ-proteobacteria; γ-MOB, methane oxidizing γ-proteobacteria; γ-bacteria, non methanotrophic γ-proteobacteria. B: Multiple sequence alignment of representative B subunits from each phylogenetic cluster. C: Gene cluster arrangement typically found in AOA and bacteria. The gene shown in grey is a conserved hypothetical protein of unknown function. The break between amoA and amoC indicates these genes are not always clustered in AOA.

Because of difficulties cultivating CuMO-producing organisms and isolating purified, active enzymes, little structural data are available for this superfamily. To date, the best-characterized CuMO is pMMO. Crystal structures of pMMO from three different methanotrophs show it is an α₃β₃γ₃ heterotrimer of which each individual protomer comprises single copies of the pmoC, pmoA, and pmoB gene products.^13–15 pMMO activity is dependent on copper, and the catalytic site is believed to be located at a crystallographically modeled dinuclear copper center in the pmoB subunit.^16–18 The pmoB subunit consists of two cupredoxin-like domains connected by two transmembrane helices and a long linker region. The copper active site is housed in the N-terminal, periplasmic cupredoxin-like domain with three histidines and the N-terminal amino group of the polypeptide within coordinating distance. These three histidines are conserved in all pmoB, amoB (homologous subunit from AMO), and related sequences except those from the methanotrophic Verrucomicrobia (Fig. 1B sequence alignment).¹⁹

Interestingly, the degree of sequence conservation between methanotrophic pmoB subunits and some homologs is quite low with < 20% identity. The AMOs from ammonia oxidizing archaea (AOA) are the most phylogenetically distant. Not only is there a low degree of sequence conservation (Fig. 1B sequence alignment), but these amoB subunits lack the C-terminal cupredoxin domain; only an N-terminal cupredoxin domain followed by a single transmembrane helix is present (Fig. 1C operon).^6,7 The impact of these sequence variations on the proposed location of the active site within this subunit is not clear. To address this question, we determined the 1.8 Å resolution crystal structure of the soluble region of an amoB subunit from the thermophilic AOA Nitrosocaldus yellowstonii.⁷

MATERIALS AND METHODS

The DNA encoding for amino acids 31 to 186 of Nitrosocaldus yellowstonii amoB (Ny_amoB) was amplified from genomic DNA and cloned into the vector pASK-IBA2 (IBA GmbH) using the BsaI restriction sites. Residue 31 corresponds to the predicted N-terminus of the mature polypeptide following signal sequence cleavage, and residue 186 marks the beginning of the predicted transmembrane region. The pASK-IBA2 vector was selected because it encodes a C-terminal Strep-tag for affinity chromatography and an N-terminal OmpA signal peptide that targets the protein for export to the periplasm where the signal peptide is cleaved, yielding the native N-terminus. Rosetta-2 (DE3) (EMD Millipore) cells transformed with the pASK-IBA2 vector containing the Ny_amoB gene were grown in Luria broth at 37 °C to an optical density at 600 nm of 0.6–1.0 and then induced with 200 μg L⁻¹ anhydrotetracycline and 0.5 mM CuSO₄ (final concentrations). Following induction, cells were grown overnight at 18 °C, harvested by centrifugation at 3,100 g for 10 min, and frozen as a solid pellet for later use.

For protein purification, cells were resuspended in buffer containing 100 mM Tris, pH 8.0, 500 mM NaCl, and 10% (v/v) glycerol, and lysed either by multiple passes through a 15,000 PSI microfluidizer or by sonication for a total of 8 min in 1 s pulses with 3 s resting time. To remove insoluble material, the cell extract was centrifuged for 30 min to 1 hr at 125,000 g. The supernatant was then applied to a 20 mL streptactin affinity column (Qiagen) equilibrated with buffer containing 500 mM NaCl, 100 mM Tris, pH 8.0, and 10% glycerol. Following extensive washing, Ny_amoB was eluted with buffer containing 100 mM Tris, pH 8.0, 500 mM NaCl, 2.5 mM d-desthiobiotin, and 10% glycerol. The eluted protein was then concentrated in an Amicon concentrator with 10 kDa nominal molecular weight cut off and applied to a 120 mL Superdex 75 (GE Healthcare) size exclusion column equilibrated with 20 mM Pipes, pH 7.2, 20 mM NaCl, and 10% glycerol. Notably, Ny_amoB is extremely stable and can be subjected to temperatures of > 65 °C for extended periods of time without precipitating.

Crystals of Ny_amoB were obtained using the sitting drop vapor diffusion method by mixing 1 μL of solution containing 5 mg mL⁻¹ protein and 3.5 μL of well solution containing 1 M NH₄SO₄, 100 mM sodium formate pH 4.0, and 2% polyethylene glycol 8000. Football shaped crystals appeared within 2–3 days. Data collection was performed at sector 21 (Life Sciences Collaborative Access Team, LS-CAT) of the Advanced Photon Source at Argonne National Laboratory. HKL2000 was used to process and integrate all data sets.²⁰ Copper single wavelength anomalous dispersion (SAD) was used to determine phases with a high redundancy data set (Table 1), and an initial model was built using the program COOT.²¹ This model was then used as a molecular replacement search model for a 1.8 Å resolution data set (Table 1). Phaser was used to obtain molecular replacement solutions, and the structure was refined with REFMAC.^22,23 The final coordinates were deposited in the Protein Data Bank (PDB accession code 4O65). Structural comparisons to pMMO were performed using the secondary structure matching (SSM) algorithm at http://www.ebi.ac.uk/msd-srv/ssm.²⁴ The pMMO crystal structures were inspected manually to determine which portions of the pmoB sequences correspond to the N-terminal cupredoxin domain, and sequence comparisons to Ny_amoB were carried out using the program EMBOSS Needle.²⁵

Table I.

Data Collection and Refinement Statistics

	Cu SAD	High resolution
Space group	P43212	P43212
Resolution (Å)	2.5 (2.5–25.0)	1.8 (1.80–22.43)
Rsym	0.058 (0.084)	0.056 (0.384)
I/σI	120.0 (41.0)	33.5 (3.5)
Completeness (%)	99.8 (100)	99.5 (98.4)
Redundancy	27 (28)	7(7.7)
Rwork‡/Rfree§ (%)		18.1/22.9
Average B-factor		20.9
r.m.s deviations
Bond lengths (Å)		0.019
Bond angles (°)		2.200

The phylogenetic tree found in Figure 1A was constructed using software available at www.phylogeny.fr. In sum 61 CuMO sequences were curated to contain domain 1 only and to remove the putative signal peptides, Muscle was used for sequence alignment, Gblocks was used for additional sequence curation, PhyML using the approximate likelihood-ratio test was used to calculate phylogenetic relationships, and TreeDyn was used for sequence visualization. The sequence alignment in Figure 1B was constructed in Geneious using Muscle to perform the alignment. The region around the insertion loop was manually aligned to reflect structural knowledge.

RESULTS AND DISCUSSION

The final model includes residues 36–112 and 119 to 186, one copper ion, two sulfate ions, and 260 water molecules. 5 amino acids at the N-terminus (excluding the putative signal peptide) and 6 amino acids in the loop between β-strands 3 and 4 were not visible in the electron density map and could not be modeled. Ny_amoB is a monomer in the crystal structure as well as in solution. Comparison of the Ny_amoB sequence with that of the N-terminal cupredoxin domains of the pmoB subunits from Methylococcus capsulatus (Bath) (PDB accession code 3RGB), Methylosinus trichosporium OB3b (PDB accession code 3CHX), and Methylocystis sp. strain M (PDB accession code 3RFR) gives sequence identities of 24.1, 19.5, and 20.1%, respectively. Despite this low sequence identity, superposition of Ny_amoB onto the N-terminal cupredoxin domain of pmoB from M. capsulatus (Bath) yields an rmsd value of 1.5 Å (averaged over the three copies of pmoB in the pMMO trimer). Similar values of 1.6 Å and 1.5 Å are obtained for the M. trichosporium OB3b, and Methylocystis sp. strain M pmoB structures, respectively. The core cupredoxin domains align nearly perfectly (Fig. 2).

Figure 2.

Structure of N-terminal cupredoxin domain of N. yellowstonii amoB subunit (Ny_amoB). Center: Overall structure (pink with copper ions as gray spheres) with analogous domain of M. capsulatus (Bath) pmoB (magenta with copper ions as dark blue spheres) superimposed. Upper-left: Zoomed-in view of insertion region with residues 55–88 shown in gray and residues at interface with core domain shown as sticks. Upper-right: Zoomed-in view of copper binding sites in M. capsulatus (Bath) pmoB and Ny_amoB.

There are two regions of Ny_amoB that differ from the pmoB structures. First, the region between β strand 1 and β strand 2 formed by residues 55–88 is 30 amino acids longer than the analogous loop in pmoB (residues 61–63 in M. capsulatus (Bath) pmoB). This extended region is composed of two helices and two loop regions and interacts extensively with the core cupredoxin domain, with more than 20 hydrogen bonds and a disulfide bond between Cys 81 and Cys 146 (Fig. 2). This structural feature thus tethers all the β strands together and may impart additional stability that accounts for the thermotolerance of Ny_amoB. Interestingly, the insertion is only present in Ny_amoB sequence (Fig. 1C).

The other structural difference between Ny_amoB and pmoB occurs at the N-terminus. In all three pmoB structures, the N-terminal residue, which is a conserved histidine, is adjacent to a conserved HXH motif located in β strand 7 of the cupredoxin fold. These three histidines provide the ligands to the copper active site.^13–16 In Ny_amoB, the 5 N-terminal residues could not be modeled, and a single copper ion, modeled at an occupancy of 0.7 and likely derived from the inclusion of 0.5 mM CuSO₄ in the growth medium, is coordinated by His 156 and His 158 from the HXH motif (Figs. 2). The disorder at this location is due to a crystal contact, which precludes packing of the N-terminus adjacent to the HXH motif and may explain the binding of only one copper ion at less than full occupancy.

Also relevant to the apparent flexibility of the Ny_amoB N-terminus is the absence of a C-terminal cupredoxin domain. In all the pMMO structures, the pmoB N- and C-terminal cupredoxin domains interact extensively, likely stabilizing the N-terminus and positioning the histidine ligands for copper binding. It is not clear how a similar effect could be achieved in Ny_amoB. One possibility is that a conserved hypothetical transmembrane protein found in the AOA AMO gene cluster (amoHYP, Fig. 1B), but not in other CuMO gene clusters, may interact directly with amoB. AMOs, at least those from ammonia oxidizing bacteria, can oxidize methane to methanol,² but Ny_amoB does not exhibit methane oxidation activity in contrast to previously characterized soluble domain constructs of M. capsulatus (Bath) pmoB.¹⁶ The lack of activity may be due to incomplete assembly or lability of the catalytic site as suggested by the structural comparison. Ammonia oxidation was not assayed because the presence of primary amines appeared to increase copper lability. Nevertheless, the presence of a copper binding site coordinated by the HXH motif is consistent with the location of the dicopper active site in pmoB.

In sum, the Ny_amoB structure reveals that the fold of the N-terminal domain of the B subunit in the AMO/pMMO superfamily is very well conserved as is the presence of an N-terminal copper binding site. Additional structural elements in Ny_amoB likely account for its thermotolerance, whereas the absence of a C-terminal domain may impact the stability of the copper binding site. Further studies of the intact N. yellowstonii AMO are needed to assess the physiological implications of these findings.