Active site architecture reveals coordination sphere flexibility and specificity determinants in a group of closely related molybdoenzymes

Abstract

MtsZ is a molybdenum-containing methionine sulfoxide reductase that supports virulence in the human respiratory pathogen Haemophilus influenzae (Hi). HiMtsZ belongs to a group of structurally and spectroscopically uncharacterized S-/N-oxide reductases, all of which are found in bacterial pathogens. Here, we have solved the crystal structure of HiMtsZ, which reveals that the HiMtsZ substrate-binding site encompasses a previously unrecognized part that accommodates the methionine sulfoxide side chain via interaction with His182 and Arg166. Charge and amino acid composition of this side chain-binding region vary and, as indicated by electrochemical, kinetic, and docking studies, could explain the diverse substrate specificity seen in closely related enzymes of this type. The HiMtsZ Mo active site has an underlying structural flexibility, where dissociation of the central Ser187 ligand affected catalysis at low pH. Unexpectedly, the two main HiMtsZ electron paramagnetic resonance (EPR) species resembled not only a related dimethyl sulfoxide reductase but also a structurally unrelated nitrate reductase that possesses an Asp-Mo ligand. This suggests that contrary to current views, the geometry of the Mo center and its primary ligands, rather than the specific amino acid environment, is the main determinant of the EPR properties of mononuclear Mo enzymes. The flexibility in the electronic structure of the Mo centers is also apparent in two of three HiMtsZ EPR-active Mo(V) species being catalytically incompetent off-pathway forms that could not be fully oxidized.

Keywords: Haemophilus influenzae; electrochemistry; electron paramagnetic resonance (EPR); enzyme kinetics; enzyme specificity; enzyme structure; methionine sulfoxide; molybdenum.

Figure 1

HiMtsZ domain and active site structure at pH 7.0.A, general fold of HiMtsZ and the structure of the HiMtsZ active site. The four structural domains are shown in different colors. N-terminal domain I (green), domain II (pink), domain III (light blue), C-terminal domain IV (orange). The ‘call out’ shows the HiMtsZ active site with key residues labeled. The two alternative configurations of Ser187 are shown with the oxygen molecules of the serine and the corresponding Mo aquo-ligand labeled in turquoise (hexacoordinate structure with an approximately trigonal prismatic shape, 80% occupancy) and red (pentacoordinate structure, 20% occupancy). B, HiMtsZ structure rotated by 90°. The coloring of the domains is as set out for panel A. C, HiMtsZ substrate access area. HiMtsZ is shown in yellow. Left panel, the view of the Mo active site from the outside. Right panel, 90° rotated view, presenting a section through HiMtsZ, with the substrate access funnel located at the top of the protein. HiMtsZ, Haemophilus influenzae MtsZ.

Figure 2

Redox and electron paramagnetic resonance properties of the HiMtsZ Mo center. Redox properties: A, calculated spectra of the Mo^VI, Mo^V, and Mo^IV forms of HiMtsZ from optical spectroelectrochemistry. B, single wavelength absorbance plots (579, 653, and 841 nm) versus electrochemical potential. The calculated redox potentials are Mo^VI/V +11 (±10) mV and Mo^V/IV −49 (±10) mV versus NHE (at pH 8). These potentials are in a similar region to those of Rhodobacter sphaeroides DMSO reductase (+83 and +37 mV versus NHE, pH 8.5) (42). Data were modeled with ReactLab Redox (Ver. 1.1, JPlus Consulting Pty Ltd) using a sequential two-electron transfer model. Electron paramagnetic resonance: C–E, 135K X-band EPR spectra of HiMtsZ forms 1 (C) 2 (D) and 3 (E) derived from redox titrations. The spin-Hamiltonian parameters determined from spectra simulations are listed in Table 1. EPR spectra for forms 1 and 2 were successfully simulated with a combination of both sets of spin-Hamiltonian parameters. HiMtsZ forms 1 and 2 display spectra that are very similar to low-pH and high-pH forms of EcNarGH, respectively. Form 1 was collected using an oxidation–reduction potential of +291 mV versus. NHE, form 2 at +34 mV versus NHE, and form 3 at −7 mV versus NHE. Black lines–experimental data; blue lines–data from simulations. DMSO, dimethyl sulfoxide; Ec, Escherichia coli; EPR, electron paramagnetic resonance; HiMtsZ, Haemophilus influenzae MtsZ; NHE, normal hydrogen electrode.

Figure 3

Characterization of HiMtsZ catalysis.A, Left, activity of HiMtsZ with different stereoisomers of MPTSO (methyl p-tolyl sulfoxide) showing a clear preference for the S- stereoisomer. Right, effect of glycerol on HiMtsZ activity with 10 mM DMSO at pH 8.0. Glycerol is a known inhibitor of DorA-type proteins but had no effect on HiMtsZ catalysis up to 30% glycerol. B, exploration of the HiMtsZ substrate range electrochemically driven catalysis at a GC/MtsZ electrode. S- and N-oxide substrates are shown. C, HiMtsZ activity with varying concentrations of S-R/S-MetSO (contains both sulfoxide stereoisomers) at pH 8.0 fit to the Michaelis–Menten equation. The data fit indicates a k_{M L-R/S MetSO_app} of 0.115 ± 0.025 mM with a k_cat of 45.2 ± 2.6 s⁻¹. D, pH dependence of HiMtsZ activity with S_C-RS_S-MetSO, showing peak activity between pH 7 and 9. DMSO, dimethyl sulfoxide; DorA, DorA DMSO reductase; DPSO, dipropyl sulfoxide; GC, glassy carbon; MetSO, methionine sulfoxide; MPSO, methyl phenyl sulfoxide; NMM-NO, N-methyl morpholine N-oxide; PVSO, phenyl vinyl sulfoxide; Py NO, pyridine-N-oxide; S-BSO, S-biotin sulfoxide; TEMEDO, tetramethylethane-1,2-diamine dioxide; TMAO, trimethylamine N-oxide.

Figure 4

CVs obtained for the methyl viologen–mediated HiMtsZ reduction of substrates.A, Sc-methionine-R/SS-sulfoxide with 50 μM MV²⁺, B, trimethylamine N-oxide and 50 μM MV²⁺, C, S(S)-S-biotin sulfoxide with 100 μM MV²⁺, and D, S(R)-S-biotin sulfoxide with 100 μM MV²⁺ at a glassy carbon/HiMtsZ-modified electrode in 50 mM Tris Cl buffer (pH 8.0) at a scan rate of 5 mV s⁻¹. Substrate concentrations are shown in the figure legends. CV, cyclic voltammetry; HiMtsZ, Haemophilus influenzae MtsZ; MV, methyl viologen.

Figure 5

HiMtsZ active site structure and electron density observed at pH 7.0 and pH 5.5. The gray mesh represents a composite omit map (73) contoured at 1 σ. The model is included in stick representation (white/light blue, carbon, main chain; yellow, sulfur; red, oxygen). A, active site at pH 7.0. Ser187 is coordinated to the molybdenum ion (occupancy of ligated conformation = 80%, shown in turquoise, occupancy of dissociated conformation = 20%, shown in red). The Mo-aquo ligand for each conformation is shown in the corresponding color. B, active site at pH 5.5. No electron density supporting Mo coordination by Ser187 is observed. HiMtsZ, Haemophilus influenzae MtsZ.

Figure 6

Comparison of electrostatic surface and substrate binding pockets of Dor-/Tor-type S-/N-oxide reductases.A, electrostatic surface representation of Haemophilus influenzae MtsZ (HiMtsZ), Rhodobacter capsulatus DorA (RcDorA), and Shewanella massilia TorA (SmTorA). The surface potential is visualized by a color gradient ranging from −5 k_bT/e_c (red) over 0 k_bT/e_c (white) to +5 k_bT/e_c (blue). While the overall shape and fold of the three enzymes are remarkably similar, the charged surfaces in the substrate access funnel differ significantly, with predominantly positive charges for HiMtsZ (left) and SmTorA (right), while RcDorA lacks this strong positive charge. In addition, the SmTorA Mo ion is much more solvent-accessible because of the absence of a tyrosine residue (HiMtsZ numbering: Y154). B, alignment of residues important for substrate binding in Dor-/Tor-type enzymes. Bold, residues important for Mo binding (Ser187), stabilization of serine dissociation stabilization (Asp185), and catalysis (Tyr154, Trp156). Pink boxes, residues important for binding of the substrate molecule side chain. HiMtsZ–Haemophilus influenzae MtsZ WP_046067646.1; EcTorZ–Escherichia coli TorZ WP_000176781.1, EcBisC–E. coli BisC WP_000013950.1, RcDorA–Rhodobacter capsulatus DorA AAD13674.1, RsDorA–Rhodobacter sphaeroides DorA WP_011338998.1, EcTorA–E. coli TorA WP_001062091.1, SmTorA–Shewanella massilia TorA CAA06851.1, RsBisC–R. sphaeroides BisC AAA74739.1. C, substrate binding to HiMtsZ and RcDorA. The surface potential is visualized by a color gradient ranging from −5 k_bT/e_c (red) over 0 k_bT/e_c (white) to +5 k_bT/e_c (blue). Interactions of substrate molecules (S_C-S_S-MetSO and DMSO) were modeled as described in the Experimental procedures section; additional data are shown in Fig. S12. BisC, biotin sulfoxide reductases; DMSO, dimethyl sulfoxide; DorA, DorA DMSO reductase; Ec, Escherichia coli; HiMtsZ, Haemophilus influenzae MtsZ; MetSO, methionine sulfoxide; Rc, Rhodobacter capsulatus; Rs, Rhodobacter sphaeroides; Sm, Shewanella massilia; TorA, TorA TMAO reductase.

Figure 7

Model of the position of methionine sulfoxide within the HiMtsZ active site.S_C-S_S-MetSO (green); HiMtsZ residues are colored by the structural domain: domain II (pink) and domain III (light blue). HiMtsZ, Haemophilus influenzae MtsZ; MetSO, methionine sulfoxide.