Characterization of ligand binding in ribonuclease a computer modelling and simulation approach

Abstract

In recent times, the fields of biochemistry, molecular biology, structural studies, and biomolecular computation have moved closer to each other and are now interconnected. Often, the inputs from one discipline can serve another, promoting the elucidation of structure-function relationships, engineering proteins with desired specificity, rational design of inhibitors and drugs, and more. Through the progress of molecular biology, a wealth of information on an increasing number of enzymes has continued to emerge. To parallel such developments, the demand for structural studies is constantly on the rise. Xray crystallography and its viable alternative, nuclear magnetic resonance (NMR) spectroscopy, have unravelled the threedimensional structures of hundreds of proteins. However, biologically relevant processes such as substrate binding and enzymatic action often defy experimental characterization at the molecular level owing to complexities such as reaction time scales. Computer modelling and simulation methods provide a new dimension to study these issues, even when starting from a lowresolution structure of an enzyme. The main objective of the research work reported in this thesis has been to characterize the modes of ligand binding and the activesite geometry in bovine pancreatic ribonuclease (RNase A - EC 3.1.27.5) using docking methods that involve systematic conformational search with contact criteria, energy minimization, and molecular dynamics. The results are used in conjunction with available experimental data to arrive at suitable models of enzyme-substrate, enzyme-intermediate, and enzyme-product complexes. Such an approach is very helpful for better understanding the specificity and mechanistic aspects of RNase A. RNase A is an extensively studied enzyme. It has served as a “test protein” for various physicochemical investigations owing to its small size (124 aminoacid residues) and ease of availability. It continues to be of considerable interest. It cleaves singlestranded RNA (or even a dinucleotide segment in RNA) through a concerted acid-base mechanism and is specific for a pyrimidine nucleoside at the 3side of the phosphodiester linkage that is hydrolysed. The enzyme shows a small preference for a purine nucleoside on the 5side of the linkage. The catalytic reaction proceeds via the formation of a 2,3cyclic phosphate intermediate. Two activesite histidines, His12 and His119, have been identified as the major promoters of catalysis. Based on Xray crystallography, Richards and Wyckoff classified the active site into subsites known as B1, R1, P1, R2, and B2 (where B stands for base, R for ribose, and P for phosphate; “1” refers to the 3side of the phosphate and “2” refers to the 5side). Hydrogenbonding contacts in the B1 subsite, especially between Thr45 and pyrimidine bases at positions 2 and 3, are vital for enzyme specificity. Mutagenesis, protein engineering, spectroscopic, and kinetic studies have established that several residues other than His12 and His119 contribute to ligand binding and/or catalysis. These include Lys7, Lys41, Asn44, Gln69, Asn71, Glu111, Phe120, Asp121, and Ser123. Gln69 and Glu111 have been identified as forming the secondary basebinding subsite B2. Some conclusions drawn from experimental studies differ regarding the conformation (e.g., His119) or the roles of these residues in binding and catalysis (e.g., Lys41). The descriptions of enzyme-substrate interactions obtained from crystallographic investigations are indirect, since the ligands used are usually substrate analogues, and the structure determination often involves nonphysiological pH conditions. The understanding of the bound dinucleotide conformation and the structural descriptions of interactions at the B2 and R2 (purine nucleoside binding) subsites remain incomplete. Recent mutagenesis and kinetic studies on the B2 subsite suggest that the present understanding of interactions may be inaccurate. Therefore, it is desirable to elucidate the molecular details of enzyme-ligand complexes systematically. Energy minimization and molecular dynamics simulations based on contact criteria can yield specific structural information that complements experimental efforts. Such calculations have been performed on RNase A-ligand complexes; the investigation is presented in seven chapters. Chapter 1 provides an overview of spectroscopic, kinetic, crystallographic, and theoretical studies on RNase A-ligand complexes, focusing on structural aspects. Anomalies in the interpretations of existing studies are highlighted. Recent proposals on the mechanism of RNase A action are discussed. The scope of computational modelling and simulation methods for studying binding modes, specificity, and mechanisms is outlined. Chapter 2 describes the computational methods adopted. The philosophy behind forcefield methods is briefly explained, followed by a discussion of potentialenergy functions. General protocols of contact criteria, energy minimization, and molecular dynamics simulations are also presented. Chapter 3 reports characterization of the primary binding subsites B1, R1, and P1 through energy minimization of RNase A-mononucleotide complexes. Ligands docked into a highresolution Xray structure (1.26 Å) of phosphatefree RNase A include pyrimidine monophosphates (2UMP, 3UMP, 2CMP, 3CMP) and purine monophosphates (3GMP, 3AMP). Regions of the protein undergoing conformational deviation upon ligand binding are discussed. The possible involvement of Ser123 in pyrimidine recognition is highlighted. Conformers of significance are identified through hydrogen bonding and energetics. Recognition of the aminoacid segments Val43-Thr45 and His119-Asp121 as forming a groove for mononucleotide binding is described, along with the differing consequences of pyrimidine vs. purine binding. Chapter 4 examines RNase A complexes with the dinucleotide substrates UpA and UpG as an extension of the mononucleotide study. RNase A-UpA represents an enzyme-dinucleotide substrate complex not yet successfully studied by crystallography. Substrate analogues used previously involve substitutions (e.g., 2,5linkage, bulky substituents, electronegative groups, removal of hydroxyl groups) that may alter stereochemistry or electrostatics. The present work offers a comprehensive description of secondary binding subsites R2 and B2. A suitable RNase A-UpA conformer is identified for enzyme-substrate modelling. Substrate conformations C3endo/anti (pyrimidine) and O4endo/syn (purine) are detailed. The purine base interacts with Gln69, Asn71, and Glu111 at positions 6 and 7. Possible roles of Asn44 and Asp121 are proposed. Structural complementarity between ligand moieties and protein segments is discussed. Conformational variations of His119 reported in the literature are examined. A structural rationale for RNase A specificity is suggested. Chapter 5 covers the dynamical properties of isolated nucleoside 2,3cyclic phosphates (Y>p), intermediates in RNase A catalysis. Fluctuations and transitions in pseudorotation phase angle and glycosidic torsion are monitored. Effects of hydrogenbond constraints and cyclization on molecular flexibility are analysed, along with simulations of 3phosphate analogues. RNase A-Y>p complexes are explored using energy minimization with appropriate protonation states of activesite histidines. Rearrangements of hydrogen bonding around the cyclic phosphate are described. Comparison with RNase A-uridine vanadate (a transitionstate analogue) shows reasonable agreement in ligand conformation and interactions. Chapter 6 discusses results from activesite solvated MD simulations on the RNase A-UpA complex. A transition of adenosine ribose pucker to C1endo was observed, enabling alternative hydrogenbonding modes. In this transition, Asn71 is the only residue interacting with adenine, consistent with recent experimental findings. Explicitsolvent effects are addressed, including watermediated His119-phosphate interactions. Snapshots from the trajectory were energyminimized. The results indicate that small but concerted changes in ligand conformation can yield substrate geometries suitable for 2,3cyclization via favourable proton transfer between the 2hydroxyl and the phosphate—supporting the intraligand protontransfer mechanism proposed in the literature. Chapter 7 summarizes the major conclusions, including: • Examination of binding modes of inhibitors, substrates, intermediates, and products. • Characterization of subsites B1, R1, P1, R2, and B2, and roles of residues Lys7, Glu111, Asn44, Asn67, Gln69, Asn71, Glu111, Phe120, Asp121, and Ser123, in addition to His12, Thr45, and His119. • Detailed analysis of RNase A-UpA complex: substrate conformation, B2subsite interactions, protein conformational changes, and surface complementarity. • Dynamical behaviour of the complex and solvent effects. • Possible mechanistic explanations for RNase A specificity. The thesis concludes with remarks on future research scope involving RNase A and related systems.