Protein Engineering by Chemical and Genetic Methods Applications to Protein Recognition and Thrombin Function

In the past three decades, the advent of recombinant DNA technology allowed the site-specific alteration of a given polypeptide chain at a glance, thus much expanding the tools available to study the molecular mechanisms of protein folding, stability, and function. This approach, also known as protein engineering, is based on the possibility of modifying the chemical composition of a protein at a single or multiple sites with the 20 available coded amino acids, thus obtaining recombinant mutant proteins with altered structure, function, or stability properties. Evaluation of the effects of the mutation on the property under investigation (e.g., folding, stability or molecular recognition) will allow quantitative estimation of the contribution of that specific amino acid which has been mutated in the wild-type protein. This information will in turn help in understanding the physico-chemical determinants exploited by natural polypeptide chains to spontaneously acquire a unique, stable and functionally active conformation. A comprehensive view of these fundamental processes at the molecular level is of paramount importance in the successful design of novel proteins, which is the ultimate goal of protein engineering. More recently, the incorporation of noncoded amino acids, possessing unique physico-chemical or spectroscopic properties, into proteins has emerged as a novel and promising approach in protein science. Protein engineering with non-coded amino acids, in fact, allow investigators to finely tune the structure at a protein site, thus much expanding the scope of physical-organic chemistry in the study of proteins. With respect to this, stepwise solid-phase chemical synthesis remains the easiest and fastest approach to site-specifically incorporate in high yields noncoded amino acid into even long (50-80 amino acids) polypeptide chains, approaching the size of real proteins. In this doctoral Thesis, relevant applications of protein engineering experiments by both genetic and chemical methods will be presented. During our studies aimed to dissect the structure-function relationships of human ?-thrombin, we approached the task of devising an efficient system to produce large amounts of recombinant protein, either in native or mutated form. Thrombin is a serine protease that plays a pivotal role in haemostasis. Human thrombin is a glycoprotein consisting of two polypeptides, a 259-residue B-chain and a smaller 36-residue A-chain, connected by a disulfide bond. B-chain contains the catalytic triad of thrombin and three disulfide bonds. Hence, prethrombin-2, the smallest physiological single-chain precursor of ?-thrombin, was expressed in E. coli, a prokaryotic system which is easy to work with, to scale up, and less time-consuming than eukaryotic systems. Using this expression system, we addressed several issues in structure-function studies on thrombin. Firstly, taking advantage of the fact that E. coli lacks the biochemical machinery for conjugating carbohydrate chains to proteins, we have studied the role of glycosylation on the structure, stability and function of the wild-type recombinant protein. Secondly, with the aim to understand the effect of natural mutations (i.e., desLys9a and Gly25Ser) in the thrombin A-chain on the structure and function of the enzyme, we have produced the corresponding recombinant forms of the naturally occurring variants of thrombin. Thirdly, we produced two mutants of human thrombin in which key Arg-residues (i.e., Arg73 in exosite I and Arg101 in exosite II) in the positively charged exosite I and II binding sites were replaced by Ala, in order to abrogate ligand-binding at each exosite. In all cases, the recombinant proteins accumulated in the inclusion bodies, from which disulfide-coupled renaturation was achieved in significantly high yields in almost all cases, yielding about 10 mg per liter of fully active wild-type human ?-thrombin. All mutant proteins were subjected to thorough characterization with respect to their chemical identity, as well as conformational, stability and functional properties. A major finding of our work was that the recombinant wild-type enzyme, lacking the carbohydrate moiety, has conformational and functional properties indistinguishable from those pertaining to the natural thrombin, but it is significantly less stable than the natural counterpart. A major application of protein engineering with noncoded amino acids entails the incorporation of suitable spectroscopic probes for studying ligand-protein and protein-protein interactions. With respect to this, 3-nitrotyrosine (NT) in absorbs radiation in the wavelength range where Tyr and Trp emit fluorescence (300-450 nm) and it is essentially nonfluorescent. Therefore, NT may function as an energy acceptor in resonance energy transfer (FRET) studies for investigating ligand-protein interactions. Here, the potentialities of NT were tested on the hirudin-thrombin system, a well-characterized protease-inhibitor pair of key pharmacological importance. We synthesized two analogues of the N-terminal domain (residues 1-47) of hirudin: Y3NT, in which Tyr3 was replaced by NT, and S2R/Y3NT, containing the substitutions Ser2?Arg and Tyr3?NT. The binding of these analogues to thrombin was investigated at pH 8 by FRET and UV/Vis-absorption spectroscopy. Upon hirudin binding, the fluorescence of thrombin was reduced by ?50%, due to the energy transfer occurring between the Trp-residues of the enzyme (i.e, the donors) and the single NT of the inhibitor (i.e., the acceptor). Our results indicate that the incorporation of NT can be effectively used to detect protein-protein interactions with sensitivity in the low nanomolar range, to uncover subtle structural features at ligand-protein interface, and to obtain reliable Kd values for structure-activity relationships studies. High throughput screening of protein-protein and protein-peptide interactions is of high interest both for biotechnological and pharmacological applications. Here, we propose the use of the non-coded amino acids o-nitrotyrosine and p-iodophenylalanine as spectroscopic probes in combination with circular dichroism and fluorescence quenching techniques (i.e., collisional quenching and resonance energy trasfer) as a mean to determine the peptide orientation in protein-peptide complexes. The method was successfully tested on an SH3 domain from a yeast myosin which is known to recognize specifically class-I peptides. The chemical strategies outlined here highlights the broad applicability of noncoded amino acids in biotechnology and pharmacological screening.