Molecular modeling of the reductase domain to elucidate the reaction mechanism of reduction of peptidyl thioester into its corresponding alcohol in non-ribosomal peptide synthetases

Abstract

Background: Nonribosomal peptide synthetases (NRPSs) are multienzymatic, multidomain megasynthases involved in the biosynthesis of pharmaceutically important nonribosomal peptides. The peptaibol synthetase from Trichoderma virens (TPS) is an important member of the NRPS family that exhibits antifungal properties. The majority of the NRPSs terminate peptide synthesis with the thioesterase (TE) domain, which either hydrolyzes the thioester linkage, releasing the free peptic acid, or catalyzes the intramolecular macrocyclization to produce a macrolactone product. TPS is an important NRPS that does not encompass a TE domain, but rather a reductase domain (R domain) to release the mature peptide product reductively with the aid of a NADPH cofactor. However, the catalytic mechanism of the reductase domain has not yet been elucidated.

Results: We present here a three-dimensional (3D) model of the reductase domain based on the crystal structure of vestitone reductase (VR). VR belongs to the short-chain dehydrogenase/reductase (SDR) superfamily and is responsible for the nicotinamide dinucleotide phosphate (NADPH)-dependent reduction of the substrate into its corresponding secondary alcohol product. The binding sites of the probable linear substrates, alamethicin, trichotoxin, antiamoebin I, chrysopermin C and gramicidin, were identified within the modeled R domain using multiple docking approaches. The docking results of the ligand in the active site of the R domain showed that reductase side chains have a high affinity towards ligand binding, while the thioester oxygen of each substrate forms a hydrogen bond with the OH group of Tyr176 and the thiol group of the substrate is closer to the Glu220. The modeling and docking studies revealed the reaction mechanism of reduction of thioester into a primary alcohol.

Conclusion: Peptaibol biosynthesis incorporates a single R domain, which appears to catalyze the four-electron reduction reaction of a peptidyl carrier protein (PCP)-bound peptide to its corresponding primary alcohol. Analysis of R domains present in the non-redundant (nr) database of the NCBI showed that the R domain always resides in the last NRPS module and is involved in either a two or four-electron reduction reaction.

Figure 1

BLAST analysis. (A) Domain organization of TPS: Different colors depict discrete modules. Each module contains three catalytic domains shown in the same color, as well as the C-terminal containing 422 amino acids, which were investigated in detail; A: Adenylation domain; C: Condensation domain; T: Thiolation domain; KS: Ketoacyl synthase; AT: Acyl transferase; ACP: Acyl carrier protein. (B) Phylogenetic tree of the characterized NRPS proteins: The neighbor-joining algorithm was used to infer the topology based on multiple sequence alignment and Poisson distances. Bootstrap scores of >50% are presented. The accession numbers of the synthases involved in the biosynthesis of the following products were as follows: saframycin (AAC44129), lyngbyatoxin (AAT12283), myxalamid (AAK57184), myxochelin (AAG31130), gramicidin (Q70LM4), Lys2 (AAA34747), glycopeptidolipid (CAB55600), BT peptide (AAY29583), nostocyclopeptide (AA023334), and TPS (AAM78457). (C) MOE 2008.10-generated 2D ligands of the final four residues were used in reductase model docking; * represents the thioester group involved in the reduction reaction.

Figure 2

Homology modeling of the R domain. (A) Multiple sequence alignment of the R domains within characterized proteins: Multiple sequence alignments of the R domains from the experimentally characterized NRPS/PKS clusters. Sequence information of the synsthases can be found at Figure 1B. The alignment showed that the NADPH binding site and the catalytic site (marked with black and green asterisks) are conserved. (B) Sequence alignment used to build the R domain model (glycopeptidolipid) based on the VR template retrieved by MOE. Gray blocks indicate the level of sequence similarity. Tallest blocks: residues are identical at that position. Intermediate blocks: residues are not identical but relatively similar based on their properties. Small blocks: residues are somewhat conserved with respect to structure or function. The absence of blocks indicates no appreciable structure/function conservation. Gaps in one sequence relative to the other are indicated by dashes. The UCSF chimera visualization system was used to generate this figure [64].

Figure 3

Evaluation of the reductase domain model. (A) The superposition of the modeled R domain (colored in cyan) with the crystal structure of VR (colored in yellow). The reductase characterizes the SYK triad for catalytic function (left). The modeled structure containing Ser142, Tyr176 and Lys180 is highlighted in cyan. The red residues are the corresponding residues of VR (right). (B) 3D structure profile for the modeled R domain. The structures of the template (2p4h) and the R domain are shown in green and blue color, respectively. (C) The Ramachandran plot of modeled reductase refined by MD simulation. A total of 270 residues (84.6%) fall in the most favored regions (cyan), while 38 residues (11.8%) fall in additional allowed regions (orange) and 12 residues (3.7%) fall in the additional allowed regions (pale orange). No residues fall in the additionally allowed region.

Figure 4

Structural refinement. (A) Spatial similarities of the VR and reductase domain binding sites. Vestitone (white line ball-and-stick) is superimposed on the R domain using the VR-vestitone docking coordinates. Energetically optimal conformations for Alamethicin (cyan, ball-and-stick), Trichotoxin (green, ball-and-stick), Antiamoebin I (magenta, ball-and-stick), Chrysopermin (yellow, ball-and-stick), Gramicidin (red, ball-and-stick) and NADPH (orange, ball-and-stick) predicted by MOE-Dock 2008.10 are pictured. For a given ligand, the result of each docking simulation is represented by a single chemical structure. (B) The putative substrate-binding pockets in the modeled structure of the R domain. Some amino acid residues in the binding pockets of the R domain are labeled and shown as a brown colored stick model. (C) RMSD of the R domain backbone atoms during 2-ns MD simulation.

Figure 5

Ligand interaction plot of the ASEDock-generated reductase Chrysopermin (panel A), Antiamoebin (panel B), Alamethicin (panel C), Gramicidin (panel D) and Trichotoxin (panel E). The plot depicts the 2D ("flattened") spatial arrangement of the ligand and the R domain with respect to key interactions. The proximity contour (dashed lines) and solvent exposed areas (solid purple spheres) of the ligand atoms are indicated as the polar (pink), hydrophobic (green), and solvent-exposed (light blue shadow) binding pocket amino acids. Acidic and basic residues are highlighted with red and blue halos, respectively.

Figure 6

Scheme of the reaction mechanism proposed for the R domain. The two steps of the reduction reaction from the peptidyl thioester to the primary alcohol are shown, and the roles proposed for the key catalytic residues Tyr176 and Glu220 are indicated.