Structure of Mesorhizobium loti arylamine N-acetyltransferase 1

The arylamine N-acetyltransferase (NAT) enzymes have been found in a broad range of both eukaryotic and prokaryotic organisms. The NAT enzymes catalyse the transfer of an acetyl group from acetyl Co-enzyme A onto the terminal nitrogen of a range of arylamine, hydrazine and arylhydrazine compounds. Recently, several NAT structures have been reported from different prokaryotic sources including Salmonella typhimurium, Mycobacterium smegmatis and Pseudomonas aeruginosa. Bioinformatics analysis of the Mesorhizobium loti genome revealed two NAT paralogues, the first example of multiple NAT isoenzymes in a eubacterial organism. The M. loti NAT 1 enzyme was recombinantly expressed and purified for X-ray crystallographic studies. The purified enzyme was crystallized in 0.5 M Ca(OAc)₂, 16% PEG 3350, 0.1 M Tris–HCl pH 8.5 using the sitting-drop vapour-diffusion method. A data set diffracting to 2.0 Å was collected from a single crystal at 100 K. The crystal belongs to the orthorhombic spacegroup P2₁2₁2₁, with unit-cell parameters a = 53.2, b = 97.3, c = 114.3 Å. The structure was refined to a final free-R factor of 24.8%. The structure reveals that despite low sequence homology, M. loti NAT1 shares the common fold as reported in previous NAT structures and exhibits the same catalytic triad of residues (Cys-His-Asp) in the active site.

1. Introduction

The arylamine N-acetyltransferase (NAT) family of phase II enzymes catalyse the acetylation of primary arylamines, arylhydrazines and N-­hydroxylated metabolites (Riddle & Jencks, 1971; Weber & Hein, 1985; Evans, 1989; Delomenie et al., 2001). Initially identified in humans through their ability to inactivate the anti-tuberculosis therapeutic isoniazid, NAT-encoding sequences have now been identified in a broad range of eukaryotic and prokaryotic species (Pompeo et al., 2002). To date, X-ray crystallographic structures of three NAT paralogues have been reported, all of which possess a highly conserved three-domain fold whose active sites contain a triad of residues (Cys-His-Asp) proposed to catalyse substrate acetylation (Sinclair et al., 2000).

Mesorhizobium loti is a nitrogen-fixing rhizobacterium that lives in symbiosis with leguminous plants (Kaneko et al., 2000). Recent bioinformatics analysis of the M. loti genome revealed two NAT isoforms, NAT1 and NAT2, and was the first example of a eubacterial species with multiple NAT isoforms (Rodrigues-Lima & Dupret, 2002). In this paper, we report the 2 Å X-ray crystallographic structure of M. loti NAT1, a comparison with other NAT structures and a discussion of the implications for M. loti NAT2.

2. Experimental

The M. loti NAT1 open-reading frame was cloned, expressed and purified from M. loti MAFF303099 strain. Details of cloning, expression and purification will be published elsewhere.

Pure recombinant protein (in 20 mM Tris–HCl pH 7.5, 1 mM EDTA, 1 mM DTT) was concentrated to 10 mg ml⁻¹ using Amicon ultracentrifugation concentrators (Millipore, Watford, Hertfordshire). The crystals described in this paper were grown at 292 K using the sitting-drop vapour-diffusion technique, with initial conditions formed by mixing equal volumes (1 µl) of the concentrated M. loti NAT1 solution with mother liquor [0.5 M Ca(OAc)₂, 16% PEG 3350, 0.1 M Tris–HCl pH 8.5]. Typically, crystals appeared after 2–5 d.

Crystals cryoprotected in paraffin oil were frozen at 100 K and diffraction data were collected at beamline ID14-2 at the European Synchrotron Radiation Facility, Grenoble, France. Data were indexed, integrated and scaled using the programs MOSFLM and SCALA (Collaborative Computational Project, Number 4, 1994). Initial phases were determined by molecular replacement with the program AMoRe using NAT from Mycobacterium smegmatis (PDB code 1gx3 ) as a search model. Initial model building was carried out using a combination of automatic rebuilding programs (ARP/wARP and FFFEAR; Collaborative Computational Project, Number 4, 1994) and manual rebuilding using the program O (Jones et al., 1991). Subsequent refinement involved iterative manual rebuilding and maximum-likelihood refinement using the program REFMAC5 (Collaborative Computational Project, Number 4, 1994). A final refinement step was to solvate the protein using the program ARP (Perrakis et al., 1999). The stereochemical quality of the final model was verified using the program PROCHECK (Laskowski et al., 1993). Refinement statistics are presented in Table 1.

Table 1
Crystallographic summary

Values in parentheses are for the highest resolution shell.

Space group	P212121
Unit-cell parameters (Å)	a = 53.2, b = 97.3, c = 114.3
Maximum resolution (Å)	2.0 (2.00–2.11)
Observed reflections	185511
Unique reflections	40315
Completeness (%)	98.5 (91.3)
I/σ(I)	4.9 (2.9)
Mosaicity (°)	0.62
Redundancy	4.6 (2.7)
Rmerge† (%)	9.8 (24.8)
VM (Å^3 Da^−1)	2.4
Refinement
Protein atoms	4174
Protein residues
Chain A	6–231, 235–275
Chain B	7–230, 235–274
Other atoms	256 water molecules
Resolution range (Å)	25.05–2.00
Rconv‡ (%)	19.5
Rfree§ (%)	24.8
Mean temperature factor (Å^2)
Main chain	26.8
Side chain	30.4
Water atoms	37.3
R.m.s.d. bond lengths (Å)	0.015
R.m.s.d. bond angles (°)	1.457

†R_merge = , where I_h,j is the intensity of the jth observation of unique reflection h.
‡R_conv = , where F_oh and F_ch are the observed and calculated structure-factor amplitudes for reflection h.
§R_free is equivalent to R_conv, but is calculated using a 5% disjoint set of reflections excluded from the maximum-likelihood refinement stages.

3. Results and discussion

The M. loti NAT1 structure adopts the characteristic three-domain NAT fold in which the first two domains, a helical bundle and a β-­barrel, are aligned to allow three residues (Cys73, His112 and Asp127, M. loti NAT1 numbering) to form a catalytic triad (Fig. 1). This three-residue structural motif is well documented in NATs and is similar to that found in the structurally related cysteine protease superfamily (Sinclair et al., 2000). The relative geometry of these catalytic residues is absolutely conserved between M. loti NAT1 and other NAT structures. The sequence and structural homology within the active site of M. loti NAT1 and other NATs is significantly higher than the global sequence identity, which ranges between 34 and 41%, and suggests that the different NAT isoforms share a common enzymatic mechanism. The third domain is linked to the second via an interdomain helix (residues 202–207), whilst the interface formed between domains II and III and the lid region together form a substantial active-site cleft. This structure extends the phylogenetic range of species over which NAT folds have been observed and supports the prediction of a conserved fold for bacterial NAT orthologues (Payton et al., 2001; Westwood et al., 2005).

Figure 1 X-ray crystal structure of M. loti NAT1. (a) and (b) Orthogonal secondary-structure representation of M. loti NAT1 (mauve) superimposed with worm representations of the NAT structures from Salmonella typhimurium (cyan; PDB code 1e2t ), Mycobacterium smegmatis (yellow; PDB code 1gx3 ) and Pseudomonas aeruginosa (green; PDB code 1w4t ) using the program O (Jones et al., 1991). The M. loti NAT1 active-site catalytic triad is also shown in ball-and-stick representation. (c) Surface representation of M. smegmatis NAT1 (yellow) and M. loti NAT1 (mauve) in the region of the active-site entrance. Underlying secondary-structural elements are drawn in a similar colour scheme. For clarity, the α6 helix, corresponding to residues 183–205 in the M. loti structure, is not shown in this figure. The active-site catalytic triad is also shown in ball-and-stick mode to aid orientation. All images were generated using AESOP (M. E. M. Noble, Oxford University, England).

Comparisons between NAT structures have previously revealed minor conformational differences resulting from both single-residue insertions/deletions that are accommodated within loop regions of the structure and from mobile loop regions linking more stable secondary-structural elements (Sinclair et al., 2000; Sandy et al., 2002; Westwood et al., 2005). These result in little perturbation to the protein fold and as such are not considered significant. The superposition of M. loti NAT1 with other NAT structures reveals a similar pattern of structural conservation (Fig. 1). The largest and potentially most significant deviation centres on the loop between β3 and β4 (approximately residue 103). Although the conformation of this loop differs between all published NAT structures, this is the only example where accessibility of the active site is potentially reduced owing to the conformation of the loop (Fig. 1). Little is known about the interactions between NATs and their substrates/cofactors and so the effect, if any, of such a conformation on the enzymatic character of M. loti NAT1 remains to be seen.

Despite sharing moderate to low levels of sequence identity, the high level of structural homology between NAT structures suggests that M. loti NAT2 will adopt a similar fold. Mapping sequence conservation between the two M. loti NAT paralogues onto the NAT1 structure reveals that with the exception of the active-site core, there are no extensive regions of residue conservation (data not shown). This may reflect a differentiation of cellular role and/or location from a common ancestral protein or simply a lack of evolutionary pressure on conservation away from the active site. In human cells, NAT isoforms exhibit overlapping but distinct substrate selectivity and cellular location profiles (Sim et al., 2000). If the M. loti paralogues represent a similar scenario, then the unusually low levels of residue conservation may reflect more divergent enzymatic roles.

In this communication, we have presented the X-ray crystallographic structure of M. loti NAT1. Despite low sequence similarity, M. loti NAT1 shares the well conserved NAT fold and highlights the sensitive nature of the profile-based search methods which first identified the sequence as encoding a NAT (Rodrigues-Lima & Dupret, 2002). The M. loti NAT1 structure adds to the growing number of published NAT structures and will complement current work aimed at identifying bacterial NAT antagonists.