Specificity and mechanism of action of ribonuclease Ti-A computer modelling study

Abstract

Enzymes recognize their substrates with very high specificity and bring about a large increase in the rates of reactions. A knowledge of the specificity and mechanism of action of enzyme-catalyzed reactions at the molecular level is very important, among other things, for the rational design of new inhibitors, drugs, and enzymes of “tailor-made” specificity. Remarkable progress has been made in the recent past in X-ray crystallographic and 2-D NMR spectroscopic techniques, providing information about the three-dimensional structure of a large number of proteins. However, obtaining high-resolution three-dimensional structural information about enzyme–inhibitor complexes by X-ray crystallography is both time-consuming and expensive. It is also not always possible to obtain crystals of enzyme–inhibitor complexes suitable for diffraction studies. This has hindered the progress of X-ray diffraction studies in biology to a large extent. On the other hand, spectroscopic techniques available today are still inadequate to study the interaction between two biomacromolecules.

Using computer modeling techniques, however, even low-resolution X-ray crystallographic data obtained on a native protein can be effectively utilized to understand the modes of binding of inhibitors and substrates to the enzyme and the mechanism of enzyme action.

Ribonuclease T? (RNase T?), an extracellular endonuclease secreted by the fungus Aspergillus oryzae, has been selected as the model system for the present work because of both its high specificity and small size (104 amino acid residues). It specifically recognizes guanine bases in single-stranded ribonucleic acid and catalyzes the hydrolysis of the phosphodiester bonds on the 3?-side of this base in a two-step reaction mechanism involving the intermediate formation of guanosine 2?,3?-cyclic phosphate.

The crystal structure of the enzyme complexed with its specific inhibitor guanosine 2?-monophosphate (2?-GMP) has been determined at a high resolution of 1.9 Å by two groups independently. In addition, high-resolution structures of RNase T? complexed with guanyl 2?,5?-guanosine and guanosine-free RNase T? complexed with vanadate are also known. The solution structures of RNase T? and its complexes with 2?-GMP and guanosine 3?-monophosphate (3?-GMP) have been determined by 2-D NMR spectroscopy combined with distance geometry and restrained molecular dynamics calculations.

Apart from these studies, ^1H and ^15N NMR and site-directed mutagenesis studies have also been carried out. But these studies do not agree with each other in some of their conclusions, such as the puckering of the ribose moiety of the inhibitor when bound to RNase T? and the hydrogen bonding scheme for the RNase T?–2?-GMP complex. Further, these studies did not provide conclusive evidence for the high specificity of RNase T? towards guanine. Although His40, Glu58, Arg77, and His92 have been identified as essential for enzyme activity by chemical modification and other physicochemical studies, no definite role for these amino acid residues in catalysis has been assigned so far. Based on X-ray crystallographic, 2-D NMR spectroscopic, and site-directed mutagenesis studies, three different schemes for the mechanism of action of RNase T? have been proposed, differing significantly in the nature of the amino acid residues involved in catalysis.

The present computer modeling studies were undertaken to seek a plausible explanation for the specificity and mechanism of action of RNase T?. Using semi-empirical potential energy functions, the three-dimensional structures of RNase T? complexed with various inhibitors and substrates have been determined. Information about the modes of binding of inhibitors/substrates to the enzyme, hydrogen bonding, and other nonbonded interactions between the enzyme and the bound ligand has been obtained from these calculations. The results have also been used to successfully explain experimental observations.

The work is presented in six chapters:

Chapter 1: A comprehensive review of the information available in the literature about RNase T?. After a brief note about its characterization by physicochemical techniques such as chemical modification, kinetic, and spectroscopic studies, the X-ray crystallographic studies of RNase T? complexes with various inhibitors are described in detail. The aim and scope of the present work follow a brief description of NMR and site-directed mutagenesis studies.

Chapter 2: Methods of calculation employed in the present work. The calculations were performed in three steps: (i) contact criteria, (ii) energy minimization in torsion angle space, and (iii) energy minimization in Cartesian coordinate space. All three steps are described in detail. Various potential functions available in the literature for calculating energy are mentioned briefly, and the parameters used in the present study are listed.

Chapter 3: Calculations on the complexes of RNase T? with 2?-GMP, 3?-GMP, and guanosine 5?-monophosphate (5?-GMP). The results are in good agreement with the observed inhibitory power of the three nucleotides. An extensive hydrogen bonding scheme has been predicted for all three complexes. The calculations explain the observed high pK? of Glu58 in the presence of 2?-GMP and reconcile discrepancies between crystallographic and spectroscopic conclusions.

Chapter 4: Specificity of RNase T?. Minimum energy conformations for complexes of RNase T? with AMP, CMP, and UMP were determined. The conformation assumed by the dipeptide fragment Asn43–Asn44 differs in complexes with 3?-GMP and AMP, playing an important role in determining specificity.

Chapter 5: Binding of guanosine 3?,5?-bis(phosphate) to RNase T?. NMR studies indicated that this ligand binds like 5?-GMP rather than 3?-GMP, though the latter binds more strongly. A stereochemical explanation is provided. Calculations were extended to dinucleotides ApG, CpG, and UpG. The theoretical binding modes differ from those assumed in spectroscopic interpretations. A possible repuckering of the ribose moiety of pGp in the RNase T?–pGp complex is also discussed.

Chapter 6: Studies on RNase T? complexes with substrate dinucleotides. Binding modes of dinucleotides to native RNase T? and mutants (His40 and Glu58 replaced) were determined. Based on the arrangement of active site residues His40, Glu58, Arg77, and His92 with respect to the substrate, a possible mechanism of action for RNase T? is proposed. Implications of binding modes of pGp, ApG, CpG, UpG, ApGp, CpGp, and UpGp in product release are discussed.

The thesis concludes with a summary of the main findings of the present work.