Mechanism of substrate specificity in 5'-methylthioadenosine/S-adenosylhomocysteine nucleosidases

Abstract

5'-Methylthioadenosine/S-adenosylhomocysteine (MTA/SAH) nucleosidase (MTAN) plays a key role in the methionine-recycling pathway of bacteria and plants. Despite extensive structural and biochemical studies, the molecular mechanism of substrate specificity for MTAN remains an outstanding question. Bacterial MTANs show comparable efficiency in hydrolyzing MTA and SAH, while the plant enzymes select preferentially for MTA, with either no or significantly reduced activity towards SAH. Bacterial and plant MTANs show significant conservation in the overall structure, and the adenine- and ribose-binding sites. The observation of a more constricted 5'-alkylthio binding site in Arabidopsis thalianaAtMTAN1 and AtMTAN2, two plant MTAN homologues, led to the hypothesis that steric hindrance may play a role in substrate selection in plant MTANs. We show using isothermal titration calorimetry that SAH binds to both Escherichia coli MTAN (EcMTAN) and AtMTAN1 with comparable micromolar affinity. To understand why AtMTAN1 can bind but not hydrolyze SAH, we determined the structure of the protein-SAH complex at 2.2Å resolution. The lack of catalytic activity appears to be related to the enzyme's inability to bind the substrate in a catalytically competent manner. The role of dynamics in substrate selection was also examined by probing the amide proton exchange rates of EcMTAN and AtMTAN1 via deuterium-hydrogen exchange coupled mass spectrometry. These results correlate with the B factors of available structures and the thermodynamic parameters associated with substrate binding, and suggest a higher level of conformational flexibility in the active site of EcMTAN. Our results implicate dynamics as an important factor in substrate selection in MTAN.

Fig. 1

Structures of S-adenosylhomocysteine (SAH) and 5′-methylthioadenosine (MTA).

Fig. 2

Time-course percent deuteration of AtMTAN1 and EcMTAN. Deuteration levels are mapped onto the sequence and structural alignment. The structural topologies at the top and bottom of the alignment correspond to AtMTAN1 and EcMTAN, respectively. Residues that are strictly conserved are displayed in white in a red box, while residues that are conservatively substituted are displayed in red. Residues that are similar are framed in blue. Residues involved in ligand binding are marked with green asterisks. The alignment was generated using ESPript (Gouet et al., 1999). Time-course experiments of 10, 30, 100, 300, 1000 and 3000 s of deuteron exchange are shown from top to bottom. The residue numbers for each peptide are indicated on top and below the AtMTAN1 and EcMTAN peptides, respectively.

Fig. 3

Hydrogen exchange profiles of AtMTAN1 and EcMTAN. Hydrogen exchange profiles mapped onto the crystal structures of (a) AtMTAN1 and (b) EcMTAN. Regions that show ≥50% deuteration after 10, 300, 1000 and 3000 s of exchange are coloured in red, yellow, green and cyan, respectively. Regions that demonstrate <50% deuteration after 3000 s of exchange are illustrated in purple. Residues shown in gray are not covered in the fragmentation maps. Residues 1–22 and 1–21 are not modeled in the top and bottom subunits, respectively, of the AtMTAN1 structure. MTT is modeled in sticks format to depict the active sites. The figure was prepared by PYMOL (DeLano, 2002). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4

Stereodiagram of differences in solvent-accessibility between AtMTAN1 and EcMTAN. At all time points, the deuteration levels of peptides in AtMTAN1 were subtracted by the deuteration levels found in equivalent peptides in EcMTAN. The average differences are represented as follows: Blue and green indicate a decrease in deuteration of ≥20% and 10–20%, respectively; yellow indicates a difference in deuteration of less than 10%; orange and red indicate an increase in deuteration of 10–20% and ≥20%, respectively; gray indicates regions where deuteration data are not available for both enzymes. The stereodiagram was prepared using PYMOL (DeLano, 2002). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

B factor comparison of AtMTAN1 and EcMTAN. The differences in B factors between the open and closed states of (a) AtMTAN1 and (b) EcMTAN. The B factors of all structures were standardized to the average B factor of all protein atoms in AtMTAN1-MTT (PDB: 2QTG). B factors of the open and closed states are plotted in red and blue, respectively. Regions highlighted in grey indicate residues that are involved in substrate binding. The differences in B factors between the open and closed states are presented in the bar graph, where orange and black indicate an increase and decrease in B factors upon the binding of a ligand, respectively. The topologies of the enzymes are illustrated above the residue numbers. Please note that B factor information is not available for residues 202–205 in EcMTAN-ADE (PDB: 1JYS). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 6

Binding of SAH, MTA and ADE to AtMTAN1 and EcMTAN. The binding isotherms for the titrations of (a) AtMTAN1 with SAH, (b) AtMTAN1 with ADE, (c) AtMTAN1-E202Q with SAH, (d) AtMTAN1-E202Q with MTA, (e) EcMTAN-E174Q with SAH and (f) EcMTAN-E174Q with MTA. No interaction is detected between AtMTAN1-Glu202Gln and SAH. The top panel shows the changes in heat over time as the ligand is titrated into the protein solution. The bottom panel shows the normalized change in heat after subtracting reference data of ligand injections into buffer. The single-binding site model was used to fit the binding isotherms and the fit is shown in red on the bottom panel.

Fig. 7

The AtMTAN1-SAH active site. (a) The catalytic residue, Glu202, is found in two conformations in the crystal structure of AtMTAN1 in complex with SAH. In monomers A and C, Glu202 is observed in the catalytic position at 40% occupancy. This orientation allows direct hydrogen-bond interactions with the substrate, which are essential for catalysis. The dominant conformation of Glu202 in all four monomers is too far for direct interaction with the ligand. Monomers A, B, C and D are illustrated in orange, yellow, white and blue, respectively. Hydrogen bonds are represented by dotted lines. (b) The experimental σ_A-weighted 3F_o − 2F_c electron density map, contoured at 0.8σ, superimposed over SAH in the refined structure. Adenine and SAH are each refined to 50% occupancy. The 5′-homocysteinyl substituent of SAH has not been modeled because no clear electron density was observed for this region. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 8

Schematic representation of the ligand-bound AtMTAN1 active sites. Ligand interactions in (a) monomers B and D, and the dominant state of monomers A and C of AtMTAN1-SAH, (b) the minor conformation of monomers A and C of AtMTAN1-SAH, and (c) the structure of AtMTAN1 in complex with the substrate-analogue, MTT (PDB: 2QTG). Hydrogen bonds are represented by dotted lines. Distances are expressed in angstroms and correspond to the average distances observed in all monomers. The distance between WAT3 and SAH exceeds our definition of 2.2–3.2 Å for a hydrogen bond, and is represented in purple. Water molecules are represented by red circles. Water molecules that are present at half occupancy are coloured in orange. Panel (c) was adapted from Siu et al. (2008a).