Analysis of Enzyme Structure and Activity by Protein Engineering

AlphaFold - A Personal Perspective on the Impact of Machine Learning

I outline how over my career as a protein scientist Machine Learning has impacted my area of science and one of my pastimes, chess, where there are some interesting parallels. In 1968, modelling of three-dimensional structures was initiated based on a known structure as a template, the problem of the pathway of protein folding was posed and bets were taken in the emerging field of Machine Learning on whether computers could outplay humans at chess. Half a century later, Machine Learning has progressed from using computational power combined with human knowledge in solving problems to playing chess without human knowledge being used, where it has produced novel strategies. Protein structures are being solved by Machine Learning based on human-derived knowledge but without templates. There is much promise that programs like AlphaFold based on Machine Learning will be powerful tools for designing entirely novel protein folds and new activities. But, will they produce novel ideas on protein folding pathways and provide new insights into the principles that govern folds?

Engineered Protein Scaffolds as Next-Generation Therapeutics

The concept of engineering robust protein scaffolds for novel binding functions emerged 20 years ago, one decade after the advent of recombinant antibody technology. Early examples were the Affibody, Monobody (Adnectin), and Anticalin proteins, which were derived from fragments of streptococcal protein A, from the tenth type III domain of human fibronectin, and from natural lipocalin proteins, respectively. Since then, this concept has expanded considerably, including many other protein templates. In fact, engineered protein scaffolds with useful binding specificities, mostly directed against targets of biomedical relevance, constitute an area of active research today, which has yielded versatile reagents as laboratory tools. However, despite strong interest from basic science, only a handful of those protein scaffolds have undergone biopharmaceutical development up to the clinical stage. This includes the abovementioned pioneering examples as well as designed ankyrin repeat proteins (DARPins). Here we review the current state and clinical validation of these next-generation therapeutics.

The role of NMR in leveraging dynamics and entropy in drug design

Nuclear magnetic resonance (NMR) spectroscopy has contributed to structure-based drug development (SBDD) in a unique way compared to the other biophysical methods. The potency of a ligand binding to a protein is dictated by the binding free energy, which is an intricate interplay between entropy and enthalpy. In addition to providing the atomic resolution structural information, NMR can help to identify protein-ligand interactions that potentially contribute to the enthalpic component of the free energy. NMR can also illuminate dynamic aspects of the interaction, which correspond to the entropic term of the free energy. The ability of NMR to access both terms in the free energy equation stems from the suite of experiments developed to shed light on various aspects that contribute to both entropy and enthalpy, deepening our understanding of the biological function of macromolecules and assisting to target them in physiological conditions. Here we provide a brief account of the contribution of NMR to SBDD, highlighting hallmark examples and discussing the challenges that demand further method development. In the era of integrated biology, the unique ability of NMR to directly ascertain structural and dynamical aspects of macromolecule and monitor changes in these properties upon engaging a ligand can be combined with computational and other structural and biophysical methods to provide a more complete picture of the energetics of drug engagement with the target. Such efforts can be used to engineer better drugs.

Coarse-grained (hybrid) integrative modeling of biomolecular interactions

The computational modeling field has vastly evolved over the past decades. The early developments of simplified protein systems represented a stepping stone towards establishing more efficient approaches to sample intricated conformational landscapes. Downscaling the level of resolution of biomolecules to coarser representations allows for studying protein structure, dynamics and interactions that are not accessible by classical atomistic approaches. The combination of different resolutions, namely hybrid modeling, has also been proved as an alternative when mixed levels of details are required. In this review, we provide an overview of coarse-grained/hybrid models focusing on their applicability in the modeling of biomolecular interactions. We give a detailed list of ready-to-use modeling software for studying biomolecular interactions allowing various levels of coarse-graining and provide examples of complexes determined by integrative coarse-grained/hybrid approaches in combination with experimental information.

Expanding the Genetic Code

Although chemists can synthesize virtually any small organic molecule, our ability to rationally manipulate the structures of proteins is quite limited, despite their involvement in virtually every life process. For most proteins, modifications are largely restricted to substitutions among the common 20 amino acids. Herein we describe recent advances that make it possible to add new building blocks to the genetic codes of both prokaryotic and eukaryotic organisms. Over 30 novel amino acids have been genetically encoded in response to unique triplet and quadruplet codons including fluorescent, photoreactive, and redox-active amino acids, glycosylated amino acids, and amino acids with keto, azido, acetylenic, and heavy-atom-containing side chains. By removing the limitations imposed by the existing 20 amino acid code, it should be possible to generate proteins and perhaps entire organisms with new or enhanced properties.

Temperature effects on S1- and S'1-enantioselectivity of alpha-chymotrypsin

The temperature dependence of E (enantiomeric ratio or enantioselectivity, a quantitative measure for enzyme stereospecificity) has been studied for the alpha-chymotrypsin catalysed hydrolysis of the enantiomeric N-Boc-L/D-TyrOMe, L/D-TyrOMe, Ac-L/D-PhgOMe, L/D-PhgOMe and for the kinetically controlled synthesis of the diastereomeric dipeptides N-Ac-L-Tyr-L/D-ArgNH2 and N-Ac-L-Tyr-L/D-ValNH2. The results show that the S1- and S'1-enantioselectivity can be modulated by the temperature (3-15 fold for the studied substrates in the range 5-45 degrees C). For L/D-PhgOMe a reversal in stereospecificity was found in this temperature interval. For the studied substrates both an increase or decrease of the enantiomeric ratio with increasing temperature was observed. For these processes the following relation for the temperature dependence of E has been derived ln E = ln(kL/kD) = -(delta deltaH# - delta deltaHb)/RT + (delta deltaS# - delta deltaSb)/R where kL and kD are apparent second order rate constants for the reactions with the L- and D-enantiomers, respectively. Delta delta denotes the differences between the thermodynamic parameters for transformation of the enantiomeric substrates. The subscript b applies for the binding of the substrate or the nucleophile and the superscript # for the formation of the transition state of the enzyme acylation or deacylation. For the studied processes either the enthalpy (delta deltaH# - delta deltaHb) or the entropy (delta deltaS# - delta deltaSb) term was found to control the discrimination. Thus, the enantioselectivity decreases or increases with temperature, respectively. The influence of ground-state interactions and transition-state stabilisation on enzyme enantioselectivity has been discussed.

Molecular evolution of RNA in vitro

Experimental studies of RNA evolution in vitro are reviewed in the context of Eigen's 1971 theory and its subsequent extensions. Current research activity and future prospects for using automated molecular biology techniques for in vitro evolution experiments are surveyed.

Site-directed mutagenesis

The development of genetic engineering, whereby a specific gene or cDNA (c is copy or complementary) can be isolated as part of a minichromosome that can replicate and be expressed (by transcription and translation) in a living cell, has made possible in vitro techniques for micromanipulation (i.e. site-directed mutagenesis) of a cloned gene, to make defined changes in the portion of the gene that encodes its protein product. The methods by which this micromanipulation of a structural gene are effected fall under three broad headings: (i) the production of random single base-pair substitutions by chemical or enzymatic means; (ii) the construction of heteroduplex DNA by annealing single-strand target DNA with a chemically synthesized mutagenic oligonucleotide and (iii) the total or partial synthesis of mutant duplex DNA from chemically synthesized oligonucleotides. As a consequence it is now possible to modify a gene so that any amino acid in its product protein can be replaced by any other amino acid.

Structure and activity of the tyrosy1-tRNA synthetase: the hydrogen bond in catalysis and specificity

The role ofhydrogen bonding in specificity, binding and catalysis by the tyrosyl-tRNA synthetase from Bacillus stearothermophilus has been investigated by systematic mutation of residues which form hydrogen bonds with substrates during the reaction between ATP and tyrosine to form tyrosyl adenylate. Data on hydrogen bonding as a determinant of biological specificity are summarized thus: deletion of an hydrogen-bond donor or acceptor between the enzyme and substrate to leave an unpaired but uncharged acceptor or donor weakens binding by only 2-7 kJ mol -1 ; but deletion to leave an unpaired but charged acceptor or donor weakens binding by some 17 kJ mol -1 or so. Hydrogen bonding is found to have a profound role in catalysis by mediating the differential binding of substrates, transition states and products. The formation of tyrosyl adenylate is not catalysed by classical mechanisms of acid-base or nucleophilic catalysis but the enhancement of rate is solely a result of a combination ofhydrogen bonding and electrostatic interactions which stabilize the transition state of the substrates relative to their ground states. The binding energy of ATP increases by more than 29 kJ mol -1 as it passes through the transition state, enhancing the rate by more than a factor of 10 5 . The residues involved in differential binding are spread over the molecule, away from the seat of reaction. The catalysis is delocalized over the whole binding site and not restricted to one or two specific residues. Some regions of the binding site are complementary in structure to the intermediate, tyrosyl adenylate. The apparent binding energies of certain side chains increase as the reaction proceeds, being weakest in the enzyme—substrate complex, stronger in the enzyme-transition-state complex and strongest in the enzyme-intermediate complex. This converts the unfavourable equilibrium constant for the formation of tyrosyl adenylate in solution to a favourable value for the enzyme-bound reagents and helps sequester the reactive tyrosyl adenylate.

Protein engineering of tyrosyl-tRNA synthetase: the charging of tRNA

Protein engineering has been used to identify residues of the tyrosyl-tRNA synthetase from Bacillus stearothermophilus that are in contact with the tRNA Tyr . By using improved techniques in oligonucleotide-directed mutagenesis, forty lysine, arginine, or histidine residues on the surface of the enzyme were altered to either asparagine or glutamine. With an in vivo genetic complementation test, only thirteen mutants were found that seriously affect the overall activity of the enzyme. Detailed kinetics on the purified enzymes revealed that four of these mutants had a lesion at the level of the activation of tyrosine, and nine at the level of tRNA charging. Three of the mutants in tRNA charging lie in the N-terminal domain of the enzyme which is responsible for tyrosine activation, and the six others in the disordered C-terminal domain which is necessary for tRNA binding. This indicates that the tRNA spans both domains of the enzyme. The construction of heterodimers allows us to suggest a model for tRNA binding in which the acceptor stem of the tRNA binds to the N-terminal domain of one subunit, and other regions of tRNA Tyr such as the anticodon arm or extra loop bind to the C-terminal domain of the other subunit.

Conventional X-ray diffraction approaches to the study of enzyme mechanism: serine proteinases, aminoacyl-tRNA synthetases and xylose isomerase

Techniques that have been used to study enzyme mechanism by conventional steady-state crystallographic techniques are reviewed. Substrates and substrate analogues can often be diffused into crystals, but occasionally co-crystallization is necessary. The poor solubility of substrates and inhibitors may pose a problem. Even if a substrate is present at adequate concentration, it may not be observed by X -ray diffraction. To observe a substrate, special measures may be needed to stop enzyme action, but sometimes this is not necessary because an equilibrium is established. Inhibitors may usefully model a particular reaction state, but one must always question whether the inhibitor provides a correct model. Stabilization of a transition state is often discussed, but rarely achieved. Where practicable, protein engineering can provide a powerful tool to test proposals about the catalytic mechanism. Molecular mechanics calculations can also be useful. These themes are developed in relation to enzymes studied in the authors’ laboratory. Many of the same problems are encountered in the application of time-resolved techniques to the study of enzyme mechanism.

Perspectives for biophysicochemical modifications of enzymes

This article reviews the strategies and successes of modifying enzymes by means of biophysicochemical transformations. By judicious choice of methods, it has been possible to modify enzymes through physical interactions, chemical reactions and/or mutagenesis to alter a very wide range of properties ranging from stability and solubility on the one hand to catalytic activity and selectivity on the other.

A structure-based analysis of the inhibition of class A beta-lactamases by sulbactam

From the crystal structure of the Bacillus licheniformis 749/C beta-lactamase, energy-minimized structures for the precatalytic, the acyl-enzyme intermediate, and the acylated linear inactivating species for sulbactam--a clinically useful mechanism-based inactivator for class A beta-lactamases--were generated. The effect of individual Ser-235-Ala and Arg244-Ser point mutations on the inactivation and turnover processes was consistent with the existence of hydrogen bonds between the side chains of these residues and the sulbactam species. The departure of the sulfinate leaving group from the acyl-enzyme intermediate of sulbactam is believed to be a prerequisite for the inactivation process. In order to explore the influence of the leaving group, penicillanic acid (2), penicillanic acid alpha-S-oxide (3), and penicillanic acid beta-S-oxide (4) were synthesized and studied in kinetic experiments with the TEM-1 beta-lactamase. Penicillanic acid is only a substrate, but penicillanic acid S-oxides were both substrates and inactivators for the enzyme. An argument is presented to rationalize these observations on the basis of the leaving ability of thiolate, sulfenate, and sulfinate from the acyl-enzyme intermediates of penicillanic acid (2), the penicillanic acid S-oxides (3 and 4), and sulbactam, respectively. The departure of the leaving group does not appear to be rate limiting in the inactivator process, but is an indispensable component of the irreversible inactivation of the enzyme. Molecular dynamics calculations of the putative inactivating species suggest that Lys-73, Lys-234, and Ser-130 are three likely residues that may be modified in the course of the inactivation chemistry. A discussion is presented of the mechanism of formation of the transiently inhibited enzyme species, which comes about as a consequence of the tautomerization of the double bond of the inactivating iminium moiety. In addition, the mechanistic details presented for sulbactam are compared and contrasted with those of clavulanic acid, another clinically used inactivator for class A beta-lactamases.

Role of the TIGN sequence in E. coli tryptophanyl-tRNA synthetase

Tryptophanyl-tRNA synthetase in E. coli does not have the HIGH sequence that is normally characteristic of class I aminoacyl-tRNA synthetases (EC 6.1.1.2), but instead contains a TIGN sequence at residues 17-20, which has been suggested to be equivalent to the HIGH sequence (Jones, M.D. et al. (1986) Biochemistry 25, 1887-1891). We have overexpressed E. coli Trp-tRNA synthetase and have used site-directed mutagenesis to mutate Thr-17 in the TIGN sequence to alanine. The mutant enzyme has the same Km values as the wild-type for tryptophan or tRNA(Trp), and a slightly increased Km for ATP, from 0.37 to 0.64 mM. On the other hand, the kcat for either the first step or the overall reaction is decreased by a factor of 30. In comparing the Thr-17 and Ala-17 enzymes, the delta delta G for the conversion of substrate to transition state is +9.6 kJ/mol (2.3 kcal/mol). Thr-17 is therefore important in binding the substrate in the transition state, thus supporting the suggestion that TIGN may fulfill the role of a HIGH sequence.

Valyl-tRNA synthetase from yeast. Discrimination between 20 amino acids in aminoacylation of tRNA(Val)-C-C-A and tRNA(Val)-C-C-A(3'NH2)

For discrimination between valine and the 19 naturally occurring noncognate amino acids, as well as between valine and 2-amino-isobutyric acid by valyl-tRNA synthetase from baker's yeast, discrimination factors (D) have been determined from kcat and Km values in aminoacylation of tRNA(Val)-C-C-A. The lowest values were found for Trp, Ser, Cys, Lys, Met and Thr (D = 90-870), showing that valine is 90-870 times more frequently attached to tRNA(Val)-C-C-A than the noncognate amino acids at the same amino acid concentrations. The other amino acids exhibit D values between 1,100 and 6200. Generally, valyl-tRNA synthetase is considerably less specific than isoleucyl-tRNA synthetase, but this may be partly compensated in the cell by valine concentrations higher than those of noncognate acids. In aminoacylation of tRNA(Val)-C-C-A(3'NH2) discrimination factors D1 are in the range of 40-1260. From D1 values and AMP formation stoichiometry, pretransfer proof-reading factors pi 1 were determined: post-transfer proof-reading factors II 2 were determined from D values and AMP formation stoichiometry in acylation of tRNA(Val)-C-C-A. II 1 values (7-168) show that pretransfer proof-reading is the main correction step, post-transfer proof-reading (II 2 approximately 1-7) is less effective and in some cases negligible. Initial discrimination factors were calculated from discrimination and proof-reading factors according to a two-step binding process. These factors, due to different Gibbs free energies of binding can be related to hydrophobic interaction forces, and a hypothetical 'stopper' model of the amino-acid-binding site is discussed.

Enzymatic oxidation of xenobiotic chemicals

Studies with biomimetic models can yield considerable insight into mechanisms of enzymatic catalysis. The discussion above indicates how such information has been important in the cases of flavoproteins, hemoproteins, and, to a lesser extent, the copper protein dopamine beta-hydroxylase. Some of the moieties that we generally accept as intermediates (i.e., high-valent iron oxygen complex in cytochrome P-450 reactions) would be extremely hard to characterize were it not for biomimetic models and more stable analogs such as peroxidase Compound I complexes. Although biomimetic models can be useful, we do need to keep them in perspective. It is possible to alter ligands and aspects of the environment in a way that may not reflect the active site of the protein. Eventually, the model work needs to be carried back to the proteins. We have seen that diagnostic substrates can be of considerable use in understanding enzymes and examples of elucidation of mechanisms through the use of rearrangements, mechanism-based inactivation, isotope labeling, kinetic isotope effects, and free energy relationships have been given. The point should be made that a myriad of approaches need to be applied to the study of each enzyme, for there is potential for misleading information if total reliance is placed on a single approach. The point also needs to be made that in the future we need information concerning the structures of the active sites of enzymes in order to fully understand them. Of the enzymes considered here, only a bacterial form of cytochrome P-450 (P-450cam) has been crystallized. The challenge to determine the three-dimensional structures of these enzymes, particularly the intrinsic membrane proteins, is formidable, yet our further understanding of the mechanisms of enzyme catalysis will remain elusive as long as we have to speak of putative specific residues, domains, and distances in anecdotal terms. The point should be made that there is actually some commonality among many of the catalytic mechanisms of oxidation, even among proteins with different structures and prosthetic groups. Thus, we see that cytochrome P-450 has some elements of a peroxidase and vice versa; indeed, the chemistry at the prosthetic group is probably very similar and the overall chemistry seems to be induced by the protein structure. The copper protein dopamine beta-hydroxylase appears to proceed with chemistry similar to that of the hemoprotein cytochrome P-450 and, although not so thoroughly studied, the non-heme iron protein P. oleovarans omega-hydroxylase.(ABSTRACT TRUNCATED AT 400 WORDS)

Arginyl-tRNA synthetase from yeast. Discrimination between 20 amino acids in aminoacylation of tRNA(Arg)-C-C-A and tRNA(Arg)-C-C-A(3'NH2)

For discrimination between arginine and 19 other amino acids in aminoacylation of tRNA(Arg)-C-C-A by arginyl-tRNA synthetase from baker's yeast, discrimination factors (D) have been determined from kcat and Km values. The lowest values were found for Trp, Cys, Lys (D = 800-8500), showing that arginine is 800-8500 times more often incorporated into tRNA(Arg)-C-C-A than noncognate acids at the same amino acid concentrations. The other noncognate amino acids exhibit D values between 10,000 and 60,000. In aminoacylation of tRNA(Arg)-C-C-A(3'NH2) discrimination factors D1 are in the range 10-600. From these values and AMP formation stoichiometry, pretransfer proof-reading factors II1 were determined; from D values and AMP stoichiometry in aminoacylation of tRNA(Arg)-C-C-A, posttransfer proof-reading factors II2 could be calculated, II1 values between 2 and 120 show that pretransfer proof-reading is the main correction step, posttransfer proof-reading (II2 approximately 1-10) plays a marginal role. Initial discrimination factors due to different Gibbs free energies of binding between arginine and the noncognate amino acids were calculated from discrimination and proof-reading factors. According to a two-step binding process, two factors (I1 and I2) were determined. They can be related to hydrophobic interaction forces and hydrogen bonds that are especially formed by the arginine side chain. A hypothetical 'stopper' model of the amino acid recognition site is discussed.

Structure of tyrosyl-tRNA synthetase refined at 2.3 A resolution. Interaction of the enzyme with the tyrosyl adenylate intermediate

The crystal structure of tyrosyl-tRNA synthetase (EC 6.1.1.1) from Bacillus stearothermophilus has been refined to a crystallographic R-factor of 22.6% at 2.3 A resolution using a restrained least-squares procedure. In the final model the root-mean-square deviation from ideality for bond distances is 0.018 A and for angle distances is 0.044 A. Each monomer consists of three domains: an alpha/beta domain (residues 1 to 220) containing a six-stranded beta-sheet, an alpha-helical domain (248 to 318) containing five helices, and a disordered C-terminal domain (319 to 418) for which the electron density is very weak and where it has not been possible to trace the polypeptide chain. Complexes of the enzyme with the catalytic intermediate tyrosyl adenylate and the inhibitor tyrosinyl adenylate have also been refined to R-factors of 23.9% at 2.8 A resolution and 21.0% at 2.7 A resolution, respectively. Formation of the complexes results in some crystal cracking, but there is no significant difference in the conformation of the polypeptide chain of the three structures described here. The relative orientation of the alpha/beta and alpha-helical domains is similar to that previously observed for the "A" subunit of a deletion mutant lacking the C-terminal domain. Differences between these structures are confined to surface loops that are involved in crystal packing. Tyrosyl adenylate and tyrosinyl adenylate bind in similar conformations within a deep cleft in the alpha/beta domain. The tyrosine moiety is in the equivalent position to that occupied by tyrosine in crystals of the truncated mutant and makes similar strong polar interactions with the enzyme. The alpha-phosphate group interacts with the main-chain nitrogen of Asp38. The two hydroxyl groups of the ribose form strong interactions with the protein. The 2'-hydroxyl group interacts with the carboxylate of Asp194 and the main-chain nitrogen of Gly192 while the 3'-hydroxyl interacts with a tightly bound water molecule (Wat326). The adenine moiety appears to make no significant polar interactions with the protein. The results of site-directed mutagenesis studies are examined in the light of these refined structures.

Dissection of the structure and activity of the tyrosyl-tRNA synthetase by site-directed mutagenesis

To understand an enzyme reaction, one has to characterize the bound substrates, intermediates, products, and transition states on the reaction pathway and determine the interaction energies between them and the enzyme as the reaction proceeds. Site-directed mutagenesis is invaluable in this task, enabling the systematic dissection of the active site. Residues involved in catalysis may be detected and the energetics probed. The contributions of each hydrogen-bonding site chain in the active site of the tyrosyl-tRNA synthetase to binding and catalysis are being determined by making sensible mutations, which remove defined interactions with the substrates. The difference in free energy between complexes of wild-type and mutant enzymes gives the apparent binding energy of the relevant side chain in each complex. By this means, the following have been determined: the contributions of different types of hydrogen bonds to specificity; their roles in catalysis; the importance of enzyme--substrate versus enzyme--transition-state versus enzyme--intermediate complementarity; the fine tuning of enzyme catalysis during "evolution"; and the existence of linear free energy relationships between rate and binding constants. Prior to these studies, the mechanism of activation by the aminoacyl-tRNA synthetases was totally unknown. It is now seen that catalysis results solely from the use of binding energy. There are residues that do not bind the substrates in the ground state but stabilize just the transition state, consistent with the classical ideas of Haldane and Pauling of enzyme--transition-state complementarity. There are, however, regions of the protein that bind the ribose ring more tightly in the intermediate tyrosyl adenylate than in the transition state.(ABSTRACT TRUNCATED AT 250 WORDS)

Structural elements and organization of the ancestral translational machinery

The molecular mechanisms underlying the primitive translational apparatus have been studied in light of present day protein biosynthesis. Using the structural information available from the contemporary system as a key to its function, both the structural necessities for an early adaptor and the multipoint recognition properties of such adaptors have been investigated. This was done by first critically examining the potential feasibility of right- and left-handed hairpin adaptor models. Second, a molecular model of the contemporary transpeptidation complex has been constructed in order to ascertain the structural requirements of the adaptor molecule needed for peptidyl transfer. Third, a model of the tRNATyr-tyrosyl tRNA synthetase complex including the positioning of the disordered region is proposed. This model is used to illustrate those required recognition properties of aminoacyl synthetase which lead to a perspective on the structure of the ancestor synthetase.

Tinkering with enzymes: what are we learning?

It is now possible, by site-directed mutagenesis of the gene, to change any amino acid residue in a protein to any other. In enzymology, application of this technique is leading to exciting new insights both into the mechanism of catalysis by particular enzymes, and into the basis of catalysis itself. The precise and often delicate changes that are being made in and near the active sites of enzymes are illuminating the interdependent roles of catalytic groups, and are allowing the first steps to be taken toward the rational alteration of enzyme specificity and reactivity.

Crystal structure of a deletion mutant of a tyrosyl-tRNA synthetase complexed with tyrosine

The crystal structure of a deletion mutant of tyrosyl-tRNA synthetase from Bacillus stearothermophilus has been determined at 2.5 A resolution using molecular replacement techniques. The genetically engineered molecule catalyses the activation of tyrosine with kinetic properties similar to those of the wild-type enzyme but no longer binds tRNATyr. It contains 319 residues corresponding to the region of the polypeptide chain for which interpretable electron density is present in crystals of the wild-type enzyme. The partly refined model of the wild-type enzyme was used as a starting point in determining the structure of the truncated mutant. The new crystals are of space group P2(1) and contain the molecular dimer within the asymmetric unit. The refined model has a crystallographic R-factor of 18.7% for all reflections between 8 and 2.5 A. Each subunit contains two structural domains: the alpha/beta domain (residues 1 to 220) containing a six-stranded beta-sheet and the alpha-helical domain (residues 248 to 319) containing five helices. The alpha/beta domains are related by a non-crystallographic dyad while the alpha-helical domains are in slightly different orientations in the two subunits. The tyrosine substrate binds in a slot at the bottom of a deep active site cleft in the middle of the alpha/beta domain. It is surrounded by polar side-chains and water molecules that are involved in an intricate hydrogen bonding network. Both the alpha-amino and hydroxyl groups of the substrate make good hydrogen bonds with the protein. The amino group forms hydrogen bonds with Tyr169-OH, Asp78-OD1 and Gln173-OE1. The phenolic hydroxyl group forms hydrogen bonds with Asp76-OD1 and Tyr34-OH. In contrast, the substrate carboxyl group makes no direct interactions with the enzyme. The results of both substrate inhibition studies and site-directed mutagenesis experiments have been examined in the light of the refined structure.

Subunit interface of triosephosphate isomerase: site-directed mutagenesis and characterization of the altered enzyme

We have replaced asparagine residues at the subunit interface of yeast triosephosphate isomerase (TIM) using site-directed mutagenesis in order to elucidate the effects of substitutions on the catalytic activity and conformational stability of the enzyme. The mutant proteins were expressed in a strain of Escherichia coli lacking the bacterial isomerase and purified by ion-exchange and immunoadsorption chromatography. Single replacements of Asn-78 by either Thr or Ile residues had little effect on the enzyme's catalytic efficiency, while the single replacement Asn-78----Asp-78 and the double replacement Asn-14/Asn-78----Thr-14/Ile-78 appreciably lowered kcat for the substrate D-glyceraldehyde 3-phosphate. The isoelectric point of the mutant Asn-78----Asp-78 was equivalent to that of wild-type yeast TIM that had undergone a single, heat-induced deamidation, and this mutant enzyme was less resistant than wild-type TIM to denaturation and inactivation caused by elevated temperature, denaturants, tetrabutylammonium bromide, alkaline pH, and proteases.