• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Madame Curie Bioscience Database [Internet]. Austin (TX): Landes Bioscience; 2000-.

Cover of Madame Curie Bioscience Database

Madame Curie Bioscience Database [Internet].

Show details

Trytophanyl-tRNA Synthetases


Tryptophan may have been the latest addition to the genetic code, as there are more significant differences be tween eukaryotic and prokaryotic Tryptophanyl-tRNA synthetases than between prokaryotic Tryptophanyl- and Tyrosyl-tRNA synthetases. TrpRS from B. stearothermophilus has proven to be an excellent model system for structural studies of conformational changes during the cycle of amino acid activation by class I aminoacyl-tRNA synthetases. Consistency between the conformational cycle documented in the published crystal structures of TrpRS and the enzymatic analysis of the close relative, Tyrosyl-tRNA synthetase, suggests a detailed model for catalysis. In this model, the binding of ATP induces an unfavorable conformational change that aligns catalytic residues for amino acid activation, graphically illustrating how ATP binding energy is “stored” in an unfavorable protein conformation for subsequent use in catalysis.

Introduction and Summary

Tryptophanyl-tRNA synthetase represents perhaps the most recent addition to the genetic code. It is a class I aminoacyl-tRNA synthetase (aaRS) that seems to break the symmetry of the class-specific hierarchy,1 there being only a single member of sub-class IIc (PheRS), while sub-class Ic includes the aminoacyl tRNA synthetases (aaRS) for both tyrosine and tryptophan. The B. stearothermophilus TrpRS monomer has 328 amino acids; with a Mr of ~ 37,000 it is smaller than the catalytic subunits of all other aaRS. The active enzyme is a dimer. Although the monomeric subunit appears to have an intact catalytic apparatus for amino acid activation such activity has not been demonstrated. The second subunit in the dimer is certainly required to recognize the CCA anticodon of tRNATrp, which falls beyond the dimensions of a monomer bound to the acceptor stem. Thus, TrpRS represents the closest extant approximation to a minimal class I catalytic apparatus for amino acid activation.

TrpRS enzymes from prokaryotic and archaeal/eukaryotic sources form quite distinct groups, even to the extent that highly conserved sequence motifs are somewhat different in the two groups. This distinction has already provoked interest in medically relevant areas. Active sites in eukaryotic and prokaryotic enzymes are distinct enough that bacterial TrpRSs should be attractive targets for anti-infective drug design.2 Perhaps even more interesting, TrpRSs from higher eukaryotes have an additional amino-terminal domain, which in humans apparently assumes an endocrine function.3,4 Moreover, the human TrpRS, for unknown reasons that may relate to the endocrine function of the amino terminal domain, is induced by interferon.5

Analysis of class Ic aaRS has benefited substantially from a detailed synergy between mechanistic studies on TyrRS6-11 and structural studies of conformational differences associated with differently liganded states of TrpRS (Table 1). The consistency of the two complementary approaches has provided considerable insights into the catalytic mechanism of class I aaRS. These pose, in turn, a new set of questions.

Table 1. Published, deposited B. stearothermophilus crystal structures.

Table 1

Published, deposited B. stearothermophilus crystal structures.

Multiple Sequence Alignments

Crystal structures are known for two closely related bacterial species, Bacillus stearothermophilus12-15 and Bacillus subtilis.16 These two species share 85% sequence identity, and there are only minimal three-dimensional structural differences between them. The seventy-odd known amino acid sequences of bacterial TrpRSs can all be aligned with reasonable confidence to that of B. stearothermophilus (HSSP:17). Until quite recently,18 there have been no structures available for either eukaryotic or achaeal TrpRSs, and the sequence homologies between these and the prokaryotic enzymes are sufficiently weak that in the absence of structures they cannot be aligned with confidence, except locally, within regions involved in the active site for amino acid activation. Even within highly conserved regions there are significant variations throughout the active site between the pro- and eukaryotic sequences.

Seventeen amino acids are identical in all known bacterial species (fig. 1). Two of these occur in the adenosine binding site identified with the HIGH signature; six occur in the tryptophan-binding site, one interacts with the ribose, and four occur in the mobile binding site for the pyrophosphate leaving group (KMSKS). Three invariant residues, G69, Q80, and S81, occur within the Rossmann dinucleotide-binding fold in a region remote from the catalytic center and which apparently serves as a bearing between domains in the conformational cycle, and one, K269, occurs in the anticodon-binding domain.

Figure 1. Invariant amino acids in TrpRS sequences.

Figure 1

Invariant amino acids in TrpRS sequences.

Curiously, the highly conserved residues derive predominantly (82%; 14/17) from the ten amino acids encoded by class II aaRS, and 35% of these are glycine or proline and have no obvious roles in either catalysis or specificity.

The relationship between bacterial and eukaryotic TrpRSs has been the subject of considerable comment1,12,19 and may reflect an important anomaly in the evolutionary history of aaRS. Only nine of the invariant residues in prokaryotic TrpRSs appear to be at all conserved in eukaryotic TrpRSs, while twelve are conserved in TrpRSs from archea. Notably missing is any homolog of K195 in the pyrophosphate-binding sequences from TrpRSs on either branch. Thus, the KMSKS signature present in most other class I aaRS from all species is never observed and is predominantly KMSAS in eukaryotic and archeal TrpRSs. An anomalous KMS sequence in which the second lysine is replaced by alanine (in yeast), and more generally by serine, is also characteristic of eukaryotic TyrRSs.

The second lysine, K195, figures prominently in structural changes related to catalytically productive ATP binding in bacterial class Ic,15,20 class Ib (GluRS(21), GlnRS22), and Ia enzymes (ArgRS23) and is therefore likely a general feature of the class I amino acid activation reaction. Absence of this group from the eukaryotic active sites suggests important differences in catalytic mechanism. Comparison of human and bacterial TyRSs24,25 led to the striking conclusion that the role of the second lysine in KMSKS can be replaced by potassium ion in the human enzyme,26 which is actually inactive in the absence of potassium. Thus, the mechanism of the eukaryotic class Ic aaRS pose significant questions of how and why the second lysine, apparently so important for catalysis in the bacterial class I aaRS, is eliminated in the eukaryotic class Ic aaRS.

Putative Endocrine Functions of Human TrpRS

TrpRSs from eukaryotes and achaea have an N-terminal extension whose function is apparently unrelated to aminoacylation. Eukaryotic N-terminal domains are about twice as long (>N< = 127) than those of archaeal species (>N< = 65), and those of mammalian TrpRSs are ~154 amino acids long. The N-terminal domains in mammals have been expressed separately and implicated in the inhibition of angiogenesis.427 Little is known about the physiological role of this domain as part of the intact TrpRS. It is of interest, however, that the human TyrRS is also augmented by a carboxy-terminal extension with cytokine activity.27-29

Evolutionary Divergence and Definition of Specificity for Aromatic Side Chains in TrpRS and TyrRS

Molecular interactions with the two low-molecular weight substrates in the active site ofBst TrpRS are shown schematically in fig. 2, which also highlights specificity determinants involved in the selection of the indole group of tryptophan. The origins of side-chain specificity in the amino acid binding pockets of the two class Ic aaRS has elicited considerable interest arising from the initial observation that identical sidechains were used in the same locations in both enzymes, raising important questions about how such unexpected homology could be used to achieve such high fidelity in side chain selection.12 The indole moeity is wedged between the sidechain of M129 and the invariant G7; The indole nitrogen donates a hydrogen bond to the carboxylate of D132; and the "bottom" of the pocket is defined by hydrophobic side chains of F5 and V141.

Figure 2. A) Active site comparison showing homologous residues interacting with amino acid and adenosine moiety in B.

Figure 2

A) Active site comparison showing homologous residues interacting with amino acid and adenosine moiety in B. stearothermophilus TrpRS (black) and B. stearothermophilus TyrRS (gray) (adapted from ref. ). Interactions in reversed contrast are not observed (more...)

Detailed mutagenesis of these Bst TrpRS residues surrounding the indole ring, motivated by comparative analysis of the TrpRS and TyrRS substrate complexes revealed that the TrpRS preference for indole is remarkably resistent to variation.30 In particular, the catalytic efficiency of tryptophan activation could be reduced by up to three orders of magnitude with scarcely any increase in specificity for tyrosine. Moreover, one mutant, D132N, while it did increase the relative affinity for tyrosine as measured by inhibition, did not support increased activation of tyrosine. Factors responsible for the relative kcat/Km values for tryptophan and tyrosine in the two enzymes therefore remain very much uncertain.

The absence of the second lysine in both eukaryotic TyrRSs and eukaryotic TrpRSs among other aspects of early multiple sequence alignments, led Ribas de Pouplana to propose that the tyrosyl and tryptophanyl enzymes were actually one and the same enzyme at the separation of the prokaryotic and eukaryotic lineages, diverging independently along each lineage to form two separate enzymes with the same amino acid and tRNA specificities.1 The second lysine of the KMSKS signature is somewhat more highly conserved than the first lysine in most class I enzymes, so its absence from both eukaryotic TrpRSs and TyrRSs argues in favor of such a lineage. Independent divergence of TrpRS and TyrRS was, however, quickly disputed.31 Recent crystal structures of the catalytic domain of human TyrRS and of intact human TrpRS have recently helped to resolve this conundrum, as decribed in a recent landmark paper.18 The best guess based on all of the available sequence data and aided by structural alignment, is that the TyrRS and TrpRS genes actually did separate before the pro-and eukaryote split.

Notwithstanding the canonical evolutionary divergence of the Trp- and TyrRS genes, the amino-acid specificity actually differentiated differently along each of the four different paths, in ways that could only be clarified by the structures of human TyrRS and TrpRS complexes.18 Two discrete locations, one in the initial beta strand (F5 inB. stearothermophilus TrpRS) and the second in the specificity-determining alpha helix (D132 in TrpRS) of the Rossmann fold figure importantly in the reconfiguration of hydrogen-bonding to the single polar moeity of the two aromatic amino acids bound in the four enzyme lineages. TrpRSs invariably have only one hydrogen bond acceptor from among the two sites, while TyrRSs invariably offer both a donor and an acceptor, making use of the bifunctionality of the phenolic oxygen. Thus the site homologous to D132 in TrpRS (D176 in Bst TyrRS) is always aspartate in all TyrRSs, while in human TrpRS it is proline. The site homologous to F5 in Bst TrpRS is hydrophobic in prokaryotic and many archaeal TrpRSs whereas in human TrpRS it is a lysine, which can donate a hydrogen bond to the indole nitrogen atom. Thus, the evolutionary development of specificity now sheds new light on the problem of molecular recognition, though without resolving many of the issues raised by the earlier mutagenesis.30

The close evolutionary relationship between the two class Ic aaRS is important for two reasons. First, it permits us to extrapolate what is known about one of them to the other, and vice-versa. Considerable mechanistic insight developed from the work of Fersht on TyrRS.7,10,11,32-36These insights greatly enhance the interpretation of TrpRS structures.

Second, the two dimeric class I enzymes could be appropriate targets for antimicrobial drug development because it should be possible to exploit the greater evolutionary divergence between prokaryotic and eukaryotic enzymes.37Structures of eukaryotic and archaeal TrpRSs, and of the mitochondrial TrpRS which belongs with the bacterial lineage, will thus be of considerable interest in resolving such questions and possibly exploiting these differences. To that end, several groups have deposited structures of the human TrpRS to the protein data bank and these are awaiting publication. These structures confirm and extend the comments made here with regard to the evolutionary differences between eukaryotic and prokaryotic TrpRSs.

The remainder of this chapter will focus on what the crystal structures of B. stearothermophilus TrpRS suggest about the role of conformational change in catalysis of aminoacid activation.

TrpRS Domain Structure and Structural Reaction Profile

The sine qua non for adducing mechanistic evidence from crystal structures is to provide a structural reaction profile consistent with an experimental kinetic reaction profile. Many, if not all of requisite allosteric states suggested by early studies of the ligand-dependent crystal growth polymorphism38,39 have been solved for the B. stearothermophilus TrpRS (Table 1). These crystal structures12-15,40have revealed a conformational polymorphism that is strongly coupled to the binding of adenine nucleotide. An appreciation of TrpRS architecture is necessary to understand this conformational cycle.

The TrpRS monomer has two separate domains (fig. 3). A canonical Rossmann dinucleotide-binding fold forms binding sites opposite each crossover connection, one (Fig. 2B) for amino acid and the other for ATP. The highly twisted β-sheet brings three “signature sequences” characteristic of both class Ic aaRS, TIGN and KMSKS, and GXDQ, together, forming the catalytic apparatus. TrpRS structures fall into three, narrowly defined families based principally on the relative orientation of the two domains that contain, respectively, the amino acid and anticodon binding sites.

Figure 3. A) TrpRS monomer, with determinants of ATP and amino acid binding and the important domain boundaries.

Figure 3

A) TrpRS monomer, with determinants of ATP and amino acid binding and the important domain boundaries. Relative movements of the small domain are shown schematically to the left. LF denotes ligand-free TrpRS, IT denotes the indolmycin:ATP complex representative (more...)

The structural polymorphism involves relative domain motion. A major domain consisting of most of the Rossmann fold (RF-αA) lies at the dimer inferface, a smaller domain formed by the combination of the N-terminal helix, αA of the Rossmann fold and the C-terminal helical bundle behaves as a distinct rigid body in all inter-conformational transitions and has been called the small domain (SD; fig. 3).

Much available data, including flourescence titrations with ATP and detailed analysis of the bimodal interaction of B. stearothermophilus TrpRS with ATP15suggest communication between the two subunits in the dimer. The latter behavior was interpreted to be a reflection of the inherent half-site reactivity observed in TyrRS, albeit with minor differences in experimental details. Despite this compelling evidence, however, none of the class Ic structures solved to date indicate any changes in the dimer interface. For this reason, further discussion will focus on the conformational cycle of the Bst TrpRS monomer. Key to the mechanism are movements of the N-terminal helix, αA the TIGN signature and the KMSKS signature relative to the rest of the RF. Three distinct TrpRS conformations have been observed: an open conformation (LF= ligand-free) in which the RF and SD are rotated fully away from one another,13 and two distinctly different closed conformations involving different orientations of the “closed” SD configuration. These will be referred to hereafter as PreTS (= pre-transition state) and Prod (= Products)

The open state is different enough from either closed state to produce experimentally measurable differences in radius of gyration, so that the conformational transitions between open and closed states have been observed by small angle X-ray solution scattering SAXS.15Radii of gyration from these SAXS measurements correspond quite closely to those calculated on the basis of atomic coordinates, lending credence to their mechanistic relevance (fig. 4). Although quite distinct, the two closed states have essentially the same overall dimensions, and cannot be distinguished by SAXS measurements. The close agreement of the Rg values measured in solution and in the crystals strongly supports the conclusion that the conformational changes are chemically and biologically relevant.

Figure 4. Small angle X-ray scattering analysis of TrpRS conformational changes.

Figure 4

Small angle X-ray scattering analysis of TrpRS conformational changes. Blue spheres denote the experimentally measured radii of gyration, RG_obs. Red spheres denote the RG calculated on the basis of the bivariate model given above as a function of ATP (more...)

The three TrpRS structural families match closely to the kinetic scheme worked out for TyrRS, Figure 5, in which the likely biochemical relevance is suggested by their shape and shading. Use of different substrate analogs and inhibitors, together with noncrystallographic symmetry afford both chemical and crystallographic redundancies giving rise to multiple examples of all thee conformations, from which we can extract several important average properties and their statistical significance.

Figure 5. TrpRS kinetic scheme, adapted from that derived by Fersht for TyrRS.

Figure 5

TrpRS kinetic scheme, adapted from that derived by Fersht for TyrRS.10 Shapes suggest relative conformations of the TrpRS monomer, and shading suggests the position on the reaction path. Distinct structures have been solved representing each microstate (more...)

Fragmentation of the adenosine-binding site in LF TrpRS13 reveals that the domain boundary actually falls between the TIGN and KMSKS signatures, on the one hand, and the GXDQ signature with the rest of the active site on the other hand. The unusual domain boundary is notable because it divides the ribose-binding pocket into two fragments. In the open conformation this binding site is fragmented. On ATP binding, motion of the small domain narrows the active-site pocket, assembling the ribose-binding site concomitantly with the formation of interactions from both halves to the ATP. In this closed state TIGN and GXDQ at the N-termini of the αA and αE helices move relative to one another during induced-fit, clamping the adenosine ribose like the two plates of a notary public's seal.

The two closed states are strongly correlated with different bound nucleotide configurations. One is observed only when the bound adenine nucleotide is the triphosphate,15The other closed conformation is observed only when the pyrophosphate moiety is cleaved from AMP and the nucleotide is an acylated 5' AMP resembling the intermediate “products” complexes of which the naturally occurring example is tryptophanyl-5'AMP.12,15The correlation between the former conformation and binding of intact nucleotide triphosphate implies that it is a closed, pre-transition state complex, resembling the configuration anticipated previously for the pre-transition state that results from induced-fit active-site assembly.10

The chief difference between the two closed states is a relocation of the SD. The transition from the pre-transition state to the Trp-5'AMP complex moves the KMSKS loop forming the PPi binding site by 1.3Å away from the α-phosphate group. This movement, and the strict correlation between formation of this conformation and a bound Trp-5'AMP analog imply that the second closed conformation almost certainly represents a products complex. This moreover implies that significant domain motion accompanies formation of the transition state during a distinct subsequent catalytic step. The transition state therefore likely develops during the transition between the two closed states. Possible mechanistic implications of this movement are discussed further below.

Study of TrpRS:tRNATrp cocrystals imply that acyl transfer restores the SD position to that in unliganded TrpRS. Co-crystals cannot be grown with bound tRNATrp in the presence of ligands stabilizing closed TrpRS states. A partial molecular replacement solution of a crystal structure of the enzyme tRNATrp complex, formed at low pH in the absence of low molecular-weight ligands, fits best to a protein conformation close to that of unliganded TrpRS. This implies that binding of the cognate tRNA to the adenylate complex actually pries the SD and its bound adenosine moiety away from the tryptophan moiety with a motion that could also stabilize a dissociative transition state for acyl transfer analogous to that which implicitly forms during activation. This cycle of domain movements is shown schematically in Figure 6.

Figure 6. Schematic of the B.

Figure 6

Schematic of the B. stearothermophilus TrpRS conformational cycle, showing the important locations of the small domains of the dimer as the activation and acyl transfer reactions proceed. The green background denotes the activation reaction, and is based (more...)

The αA helix is tightly coupled to the 4-helix bundle containing the anticodon binding determinants via the hydrophobic side chains of I16 and M193 from the two signatures which, together with I20, join a substantial hydrophobic core (fig. 3), within the SD,13 thereby coordinating the position of the anticodon-binding site to the active-site chemistry. This core endows the SD with the ability to move as a rigid body. However, it is not statically configured, as movement between states involves significant repacking.

The ATP-dependent conformational excursions of the TrpRS small domain exhibit a larger range of motion than is generally observed in other class I synthetase structures in similar ligation states.41,42However, both the leucyl and tyrosyl enzymes exhibit similar motion, and much remains to be determined before a description can be given of a “unified” class I mechanism, if such exists. The tryptophan binding site is similarly subdivided into two halves by a more subtle opening and closing of both backbone and sidechains and the structural evidence15 suggests coupling between the configuration of Y125 at the entrance to the tryptophan binding pocket and the open/closed configuration of the SD. Curiously, while a similar conformational change is induced in MetRS by methionine binding, the corresponding motion in TrpRS appears to be triggered by ATP binding and not by tryptophan binding. This suggests that a modest and as yet poorly uncharacterized fragmentation of the major RF domain in the unliganded state responds to ATP binding, closing the tryptophan binding pocket as well, and suggesting some linkage between binding of the two low molecular weight substrates.

Consistency of the TrpRS Polymorphism with the TyrRS Kinetic Reaction Profile

Three distinct functional regions within the Rossmann fold, the binding sites for ribose, tryptophan, and PPi, can be distinguished (fig. 7) as suggested in Figures 1, 2, and 3. Functional distinctions between these three structural regions rest on dissecting the corresponding TyrRS sites by mutation and pre-steady state kinetics.10 Mutational alteration of any of the three sites in TyrRS generally results in decreases of affinity. These losses in affinity have different time courses relative to the catalytic process (fig. 5). Alteration of the tyrosine binding site affects amino acid binding throughout the catalytic cycle, with more or less equivalent effects in the Michaelis complex, the transition-state complex, and products complex.

Figure 7. Fersht's investigation of the TyrRS mechanism.

Figure 7

Fersht's investigation of the TyrRS mechanism. a. Spatial delineation by mutation of static (tyrosine, smooth) and dynamic (ATP, rough) subsites in the TyrRS and, by implication TrpRS active sites. TyrRS active-site mutations to tyrosine-(yellow), ribose-(red), (more...)

Mutation of amino acids associated with ribose binding, notably T40, H45, H48, and D194, and with PPi binding, H45 and K230, have no effect on the ground-state affinity for ATP, but they exert a profound impact on the transition state affinity. The second lysine in the KMSKS loop, K233, is a special case. Mutations of K233 do affect ATP affinity in the ground state. However, the implications are far from clear, as this mutation changes the cooperativity between tyrosine and ATP binding.43,44 In the TrpRS cycle, all the corresponding residues behave consistently with their mutational behavior in TyrRS.

The mutational and pre-steady state results suggest an attractive interpretation in light of the TrpRS structural changes. The temporal invariance of a mutational impact on binding energy along the reaction path as measured by pre-steady state kinetics implies a static site that has no need to rearrange to alter interaction energies. Catalytically senstive sites that show differential effects in the transition (and products) states, on the other hand, are likely to be structurally dynamic, requiring internal rearrangments to achieve the differential binding affinity.

The paradigm defined by mutation of the TyrRS active site is reflected in detail in the domain movements observed in the sequence of TrpRS structures. Although there are side chain motions at the entrance to the tryptophan binding site, none of the conformational transitions alters the basic properties of the indole binding site. The most prominent of the changes at the mouth of the tryptophan pocket involves Y125. Mutation of the corresponding tyrosine in TyrRS, Y169, has no differential binding effects in the transition state, despite the fact that this side chain has been observed to close in the T. thermophilus TyrRS crystal structures.20 This evidence suggests that the structural modifications of the amino acid binding site have little differential binding effect and hence are of little importance in catalysis of amino acid activation. Mutation of Y169 does, however, impact the rate of acyl transfer in B. stearothermophilus TyrRS,45 implicating catalytically relevant, structurally dynamic changes in this region during the second half of the tRNA charging reaction.

Ribose-binding residues in TrpRS corresponding to those mutated in TyrRS lie on opposite sides of the fragmented ATP binding site in the unliganded and open ATP complex structures. S11 and D134 are part of the Rossmann fold and do not interact with ATP in the open ATP complex. T15, N18, and K192 on the other hand form hydrogen bonded and/or Van der Waals contacts with ATP in both open and closed states. The two parts of the ATP binding site come together to interact jointly with ATP only in the closed, pre-transition state and product complexes.15

Similarly, the PPi-binding site formed by the KMSKS loop is incompletely folded in the open ground state. It assumes a fully integrated conformation in which K195 moves around the PPi moiety and into position on the opposite side of the β-phosphate from the K192 side chain (fig. 8) only in the closed, pre-TS state. The KMSKS loop becomes disordered again in the products complex.14 This subtle, transient assembly of the PPi leaving group site has been observed in all class I crystal structures for which complexes with ATP have been reported.21,22,42

Figure 8. Comparison of the open and closed ATP complexes.

Figure 8

Comparison of the open and closed ATP complexes. The red segments are the signature sequences, TIGN and KMSKS. Yellow segments are the two belts linking the RF-α and SD domains. Light blue denotes the RF-α domain and pink the linking peptide (more...)

Thus, the induced-fit process documented in the TrpRS conformational cycle provides a sensible rationale for the pre-steady state and kinetic studies of homologous residues in TyrRS by showing that sidechains whose mutation changes only the transition state affinity essentially coincide with residues observed to move relative to one another during the conformational cycle of active-site assembly and product formation.

Mechanistic Implications of Domain Movement: A Transition State with Dissociative Character?

The fundamental mechanistic question posed by enzymologists concerned with phosphoryl transfer is to describe the structure of the phosphate moiety in the transition state, and how enzymes interact with this transition state. Two limiting scenarios are often distinguished: dissociative and associative (fig. 9). These adjectives describe the distances from the phosphate moiety to the nucleophile and leaving groups. A complementary view is expressed by the distinction between late and early , which describe the order in which bond breaking and bond making occur. Actual transition states usually lie somewhere between the two limiting geometries.

Figure 9. Limiting transition state structures for the phosphoryl transfer reaction in tryptophan activation.

Figure 9

Limiting transition state structures for the phosphoryl transfer reaction in tryptophan activation. A) Dissociative transition state. The bond to the leaving pyrophosphate group is ruptured prior to the attack by the tryptophan carboxyate nucleophile. (more...)

The dissociative mechanism is generally thought to limit the rate of model phosphoryl transfer reactions in aqueous solution including phosphate monoester hydrolysis, in the absence of catalysts. 46 It proceeds by breaking the bond to the leaving group before the approach of the nucleophile, forming a metaphosphate transition state. If the nucleophile is brought close enough to the phosphate group before the leaving group departs, thereby forming a pentavalent phosphoryl group in the transition state, the reaction is said to procede by an associative transition state. The pentavalent phosphoryl spreads additional negative charge onto the equatorial oxygen positions, leading to different stereochemical requirements for the active site.

Most conclusions regarding transition state structures rely on indirect evidence, which is often contradictory. One diagnostic familiar to structural biologists involves using stable structural transition-state analogs. Thus, it has been proposed that compounds like vanadate, AlF3, and AlF4 resemble the pentavalent phosphoryl moiety in an associative transition state. Inhibition by these compounds suggests associative character of the transition states. The AlFn and vanadate compounds inhibit, for example, the signaling GTPases, Ras, Gα, transducin, and so forth, and the structures of inhibited complexes provide evidence for stereochemical interactions from active-site components. These compounds likely mimic the high negative charge on the pentavalent phosphoryl transition state, and so are transition-state analogs for enzymes that use an associative mechanism. Structures of the inhibited AlF3/AlF4 complexes generally reveal cationic side chains like arginine and lysine ligating the equatorial flouride positions around the Aluminum, and we can draw the inference that they bind to and stabilize the associative transition state. In contrast, however, chemical arguments have been advanced that Ras GTPase stabilizes a dissociative transition state,47 and the issue has not yet been settled.

The ultimate arbitor of transition-state geometry is generally considered to be kinetic isotope effects, which are quite precise and unambiguous.48,49 Such studies have not been performed on many of the enzymes that catalyze phosphoryl transfer; consequently, tentative conclusions rest on indirect evidence.

Considerable circumstantial evidence can be adduced regarding the TrpRS transition state. Neither AlFn nor vanadate compounds inhibit TrpRS, suggesting by this argument that TrpRS catalysis does not stabilize an associative transition state. Neither the TrpRS crystal structures nor the mutagenesis and pre-steady-state kinetics of TyrRS provide evidence for stereochemical stabilization of a pentavalent phosphoryl moiety in the transition state. Specifically, no polar moieties in the TrpRS active site appear capable of interacting with the ground-state configuration of the α-phosphate in any of the bound adenine nucleotide ligands, and there is no obvious way to model such interactions to an altered, pentavalent configuration. Nor do any of the TyrRS active-site mutations that affect kcat interact directly with the α-phosphate in the TyrRS complex with tyrosyl-5'adenylate. Nearly all of these residues are identical in the two enzymes, and hence it is unlikely that TrpRS provides sidechain interactions with the α-phosphate. Indeed, it is singularly odd that no catalytic attention is devoted by TrpRS to the alpha phosphate in any of the complexes.

Rather, the TrpRS domain motions seem designed to “pull” the PPi away from the adenosine monophosphate (fig. 10), consistent with their stabilization of a dissociative transition state. Conversion of the pre-transition state structure to the product complex structure results in a relocation of the PPi binding site by about 1.3_he primary change is a rotation of the small domain, which brings the KMSKS loop and PPi binding subsite along with it because methionine M193 is embedded in the intensive hydrophobic core at the center of the small domain. In contrast, the adenosine-binding site is nearly at the center of rotation, and hence remains relatively fixed. As a consequence, the PPi and AMP binding sites, which Fersht showed to be the locus of the most important catalytic contributions, are moving away from one another in the transition state!

Figure 10. Relocation of the PPi-binding site on product formation.

Figure 10

Relocation of the PPi-binding site on product formation. Structures of the Pre-TS complex, the products complex with the sulfoamoyl inhibitor plus PPi, and the original product complex are oriented and aligned vertically such that the alpha phosphate (more...)

Ligand-Binding Affinity and Conformational Free Energy Changes

Use of the pyrophosphate exchange assay affords direct measurements of the affinity of TrpRS for its substrates from the Km values obtained by Michaelis Menten experiments. Measurements of these affinities from several laboratories including our own are in agreement that the affinity of TrpRS for tryptophan is ~2 μM, while that for ATP is ~400 μM, or two hundred fold weaker.

Moreover, factorial analysis of these two affinities on the concentration of the other sbustrate show that affinities for both ligands change as a function of the concentration of the other.15 The kinetic measurements indicate that the interactions are not reciprocal: increasing tryptophan concentration increases the affinity for ATP, while increasing ATP concentration actually reduces the affinity for tryptophan (fig. 11). This effect persists over a range of ATP concentrations well above the Km for ATP, and implies that ATP binding is bimodal: the dependence of this unusual phenomenon goes as the square of the ATP concentration, which implies in turn that the effect of ATP is bimolecular and suggests that it results from binding of ATP to both sites on the dimer.

Figure 11. Variation of substrate affinities induced by different concentrations of the second substrate.

Figure 11

Variation of substrate affinities induced by different concentrations of the second substrate.

Bimodal binding of ATP has also been demonstrated structurally by the solution of two distinct ATP bound states,15 one in which the TrpRS conformation is essentially that of the unliganded enzyme, and another indistinguishable from that of the closed, pre-transition state complex with ATP and an unreactive tryptophan analog (fig.8). The ATP concentration necessary for formation of the more compact pre-transition state complex has been determined by titration of the crystal morphology with ATP39 and by using small angle X-ray scattering in solution.15 Both titrations suggest an ATP concentration of ~5mM for the midpoint of the conformational transition. Curiously, the open conformation in which the ATP binding site is fragmented has higher ATP affinity than does the closed, pre-transition state conformation in which all or nearly all of the potential interactions with the nucleotide are fully developed (fig.12).

Figure 12. Projected binding interactions in the open and closed ATP complexes.

Figure 12

Projected binding interactions in the open and closed ATP complexes. A) The open complex (interactions with Mg^++ are inferred as they have not been identified from the corresponding crystal structure). B) The closed, pre-transition state complex.

The high ATP concentration required for active-site assembly, together with the dramatic contrast between the small number of binding interactions evident in the high-affinity, open state relative to the large number of binding interactions evident in the low-affinity closed state strongly suggests that the reduced binding affinity of ATP in the catalytically competent pre-transition state conformation is actually evidence of significant destabilization of the intrinsic free energy of the closed TrpRS conformation itself. In turn, this suggests that a considerable amount of the total binding energy afforded by ATP is used to induce an unfavorable conformation change in TrpRS itself.

The amount of binding energy that is actually converted in this conformational free energy storage presents an intriguing and as yet unsolved problem. Several different lines of evidence suggest, however, that it is a considerable fraction of the standard free energy change of nucleotide triphosphate hydrolysis. The TrpRS affinity for ATP is remarkably weak, relative to the extremely tight binding observed in the Ras:GTP and Ras:GDP complexes50 which are ~8 orders of magnitude more tightly bound. Much of this loss of affinity is likely to represent destabilization. A complementary observation comes from the fact that the binding of cognate tRNA is required to form the comparable closed forms of ArgRS,51 GluRS,21 and GlnRS.22 Although no attempt has been made to estimate the effect quantitatively, these phenomena clearly illustrate the thermodynamic difficulty of active-site assembly in class I aaRS generally.


Supported by NIGMS 48519. I am indebted to Lluis Ribas de Pouplana for communicating results on the human TrpRS prior to publication.


Ribas de Pouplana L, Frugier M, Quinn C. et al. Evidence that two present-day components needed for the genetic code appeared after nucleated cells separated from eubacteria. Proc Natl Acad Sci USA. 1996;93:166–170. [PMC free article: PMC40199] [PubMed: 8552597]
Kanamaru T, Nakano Y, Toyoda Y. et al. In Vitro and In vivo Antibacterial Activities of TQAK-083, an Agent for Treatment of Helicobacter pylori Infection. Antimicrobial Agents and Chemotherapy. 2001;45(9):2455–2459. [PMC free article: PMC90677] [PubMed: 11502514]
Ibba M. Eyeing up tryptophanyl-tRNA synthetase. TIBS. 2002;27(5):227. [PubMed: 12076529]
Otani A, Silka BM, Dorrell MI. et al. A fragment of human TrpRS as a potent antagonist of ocular aniogenesis. Proc Nat Acad Sci USA. 2002;99:178–183. [PMC free article: PMC117535] [PubMed: 11773625]
Xue H, Wong JT-F.Interferon. induction of human tryptophanyl-tRNA synthetase safeguards the synthesis of tryptophan-rich immune-system proteins, a hypothesis. Gene. 1995;165:335–339. [PubMed: 8522205]
Fersht AR, Shi J-P, Knill-Jones J. et al. Hydrogen bonding and biological specificity analysed by protein engineering. Nature. 1985;314(3:21)235–238.
Fersht AR, Leatherbarrow RJ, Wells TNC. Binding energy and catalysis: a lesson from protein engineering of the tyrosyl-tRNA synthetase. TIBS. 1986;11(8):321–325.
Fersht AR, Leatherbarrow RJ, Wells TNC. Structure and activity of the tyrosyl-tRNA synthetase: the hydrogen bond in catalysis and specificity. Phil Trans R Soc Lond. 1986;317(A):305–320.
Fersht AR. The charging of tRNA. New York: Chapman & Hall. 1986
Fersht AR. Dissection of the structure and activity of the tyrosyl-tRNA synthetase by site-directed mutagenesis. Biochem. 1987;26(25):8031–8037. [PubMed: 3442641]
Fersht AR, Knill JonesJW, Bedouelle H. et al. Reconstruction by site-directed mutagenesis of the transition state for the activation of tyrosine by the tyrosyl-tRNA synthetase: a mobile loop envelopes the transition state in an induced-fit mechanism. Biochem. 1988;27(5):1581–7 Issn 0006-2960. [PubMed: 3284584]
Doublié S, Bricogne G, Gilmore CJ. et al. Tryptophanyl-tRNA synthetase crystal structure reveals an unexpected homology to Tyrosyl-tRNA synthetase. Structure. 1995;3(1):17–31. [PubMed: 7743129]
Ilyin VA, Temple B, Hu M. et al. 2.9 _ystal structure of ligand-free tryptophanyl-tRNA synthetase: Domain movements fragment the adenine nucleotide binding site. Prot Sci. 2000;9:218–231. [PMC free article: PMC2144547] [PubMed: 10716174]
Retailleau P, Hu M, Bricogne G. et al. High resolution experimental phases for Tryptophanyl-tRNA synthetase (TrpRS) complexed with Tryptophanyl-5'AMP. Acta Cryst. 2001;D57:1595–1608. [PubMed: 11679724]
Retailleau P, Huang X, Yin Y. et al. Interconversion of ATP binding and conformational free energies by Tryptophanyl-tRNA Synthetase: structures of ATP bound to open and closed, pre-transition conformations. J Mol Biol. 2003;325:39–63. [PubMed: 12473451]
Ala P, Xue H, Leung L. et al. Crystallization of Bacillus subtilis tryptophanyl-tRNA synthetase. J Mol Biol. 1993;230(3):1089–90. [PubMed: 8478922]
Sander C, Schneider R. Database of homology-derived protein structures and the structural meaning of sequence alignment. Proteins. 1991;9:56–68. [PubMed: 2017436]
Yang X-L, Otero FJ, Skene RJ. et al. Crystal Structures that suggest late development of genetic code components for differentiating aromatic side chains. Proc Nat Acad Sci USA. 2003;In Press [PMC free article: PMC307575] [PubMed: 14671330]
Nagel GM, Doolittle RF. Evolution and Relatedness in two aminoacyl-tRNA Synthetase Families. Proc Nat Acad Sci USA. 1991;88(September 1991):8121–8125. [PMC free article: PMC52458] [PubMed: 1896459]
Yaremchuk A, Kriklivyi I, Tukalo M, Cusack S. Class I tyrosyl-tRNA synthetase has a class II mode of cognate tRNA recognition. EMBO J. 2002;21(14):3829–40. [PMC free article: PMC126118] [PubMed: 12110594]
Sekine S-i, Nureki O, Dubois DY. et al. ATP Binding by glutamyl-tRNA synthetase is switched to the productive mode by tRNA binding. EMBO J. 2003;22:676–688. [PMC free article: PMC140737] [PubMed: 12554668]
Sherlin LD, Perona JJ. tRNA-dependent active site assembly in a class I aminoacyl-tRNA synthetase. Structure. 2003;11(5):591–603. [PubMed: 12737824]
Delagoutte B, Moras D, Cavarelli J. tRNA aminoacylation by arginyl-tRNA synthetase: induced conformations during substrates binding. EMBO J. 2000;19(21):5599–5610. [PMC free article: PMC305789] [PubMed: 11060012]
Austin J, First EA. Comparison of the Catalytic Roles Played by the KMSKS Motif in the Human and Bacillus stearothermophilus Tyrosyl-tRNA Synthetases. J Biol Chem. 2002;277(23):28394–28399. [PubMed: 12016229]
Austin J, First EA. Catalysis of Tyrosyl-Adenylate Formation by the Human Tyrosyl-tRNA Synthetase. J Biol Chem. 2002;277(17):14823–24820.
Austin J, First EA. Potassium Functionally Replaces the Second Lysine of the KMSKS Signature Sequence in Human Tyrosyl-tRNA Synthetase. J Biol Chem. 2002;277(23):20243–20248. [PubMed: 11927599]
Wakasugi K, Slike BM, Hood J. et al. A human aminoacyl-tRNA syntetase as a regulator of aniogenesis. Proc Nat Acad Sci USA. 2002;99:173–177. [PMC free article: PMC117534] [PubMed: 11773626]
Kleeman TA, Wei D, Simpson KL. et al. Human Tyrosyl-tRNA Synthetase Shares Amino Acid Sequence Homology with a Putative Cytokine. J Biol Chem. 1997;272:14420–14425. [PubMed: 9162081]
Wakasugi K, Schimmel P. Two Distinct Cytokines Released from a Human Aminoacyl-tRNA Synthetase. Science. 1999;284:147–151. [PubMed: 10102815]
Praetorius-Ibba M, Stange-Thomann N, Kitabatake M. et al. Ancient adaptation of the active site of tryptophanyl-tRNA synthetase for tryptophan binding. Biochem. 2000;39(43):13136–43. [PubMed: 11052665]
Brown JR, Robb FT, Weiss R. et al. Evidence for the early divergence of tryptophanyl- and tyrosyl-tRNA synthetases. J Mol Evol. 1997;45(1):9–16. [PubMed: 9211729]
Fersht AR. Basis of biological specificity. TIBS. 1984;9(4):145–147.
Fersht AR, Wilkinson AJ, Carter P. et al. Fine structure-activity analysis of mutations at position 51 of tyrosyl-tRNA synthetase. Biochem. 1985;24(21):5858–61 Issn 0006-2960. [PubMed: 3002425]
Fersht AR, Leatherbarrow RJ, Wells TNC. Quantitative analysis of structure-activity relationships in engineered proteins by linear free-energy relationships. Nature. 1986;322(7:17)284–286.
Fersht AR. Protein engineering of homodimeric tyrosyl-tRNA synthetase to produce heterodimers: Kinetics of dimeric. Biochem. 1986;61:9576–9578. [PubMed: 3733687]
Fersht A. Dissection of the structure and activity of an enzyme. In: Kaiser ET, ed. Design of Enzymes and Enzyme Models. Houston: Robert A. Welch. Foundation;1988:159–182.
Qiu X, Janson CA, Smith WW. et al. Crystal structure of Staphylococcus aureus tyrosyl-tRNA synthetase in complex with a class of potent and specific inhibitors. Prot Sci. 2001;10:2008–2016. [PMC free article: PMC2374228] [PubMed: 11567092]
Carter CWJr, Carter CW. Protein Crystallization Using Incomplete Factorial Experiments. J Biol Chem. 1979;254:12219–12223. [PubMed: 500706]
Carter CWJr, Doublié S, Coleman DE. Quantitative Analysis of Crystal Growth: Tryptophanyl-tRNA Synthetase Polymorphism and its Relationship to Catalysis. J Mol Biol. 1994;238:346–365. [PubMed: 8176729]
Yin Y. Crystallographic Study of Bacillus stearothermophilus Tryptophanyl-tRNA Synthetase in the Catalytic Reaction [Ph.D]: University of North Carolina at Chapel Hill; 1995. [PMC free article: PMC152490]
Cusack S, Yaremchuk A, Tukalo M. The 2 A crystal structure of leucyl-tRNA synthetase and its complex with a leucyl-adenylate analogue. EMBO J. 2000;19(10):2351–61. [PMC free article: PMC384370] [PubMed: 10811626]
Yaremchuk A, Kriklivyi I, Tukalo M. et al. Class I tyrosyl-tRNA synthetase has a class II mode of cognate tRNA recognition. EMBO J. 2002;21(14):3829–3849. [PMC free article: PMC126118] [PubMed: 12110594]
First EA, Fersht AR. Mutation of lysine 233 to alanine introduces positive cooperativity into tyrosyl-tRNA synthetase. Biochem. 1993;32(49):13651–13657. [PubMed: 8257698]
First EA, Fersht AR. Analysis of the role of the KMSKS loop in the catalytic mechanism of the tyrosyl-tRNA synthetase using multimutant cycles. Biochem. 1995;34(15):5030–5043. [PubMed: 7711024]
Xin Y, Li W, First EA. Stabilization of the transition state for the transfer of tyrosine to tRNA(Tyr) by tyrosyl-tRNA synthetase. J Mol Biol. 2000;303(2):299–310. [PubMed: 11023794]
Hollfelder F, Herschlag D. The Nature of the Transition State for Enzyme-Catalyzed Phosphoryl Transfer. Hydrolysis of O-Aryl Phosphorothioates by Alkaline Phosphatase. Biochem. 1995;34:12255–12264. [PubMed: 7547968]
Maegely KA, Admiraal SJ, Herschlag D. Ras-catalyzed hydrolysis of GTP: A new perspective from model studies. Proc Nat Acad Sci USA. 1996;93:8160–8166. [PMC free article: PMC38640] [PubMed: 8710841]
Schramm VL. Enzymatic Transition States and Transition State Analog Design. Annu Rev Biochem. 1998;67:693–720. [PubMed: 9759501]
Schramm VL. Enzymatic Transition-State Analysis and Transition-State Analogs. Meth Enzymol. 1999;308:301–355. [PubMed: 10507010]
Feuerstein J, Goody RS, Wittinghofer A. Preparation and Characterization of Nucleotide-free and Metal Ion-free p21 “Apoprotein” J Biol Chem. 1987;262:8455–8458. [PubMed: 3298232]
Delagoutte B, Moras D, Cavarelli J. tRNA aminoacylation by arginyltRNA synthetase: induced conformations during substrates binding. EMBO J. 2000;19:5599–5610. [PMC free article: PMC305789] [PubMed: 11060012]
Yang XL, Otero FJ, Skene RJ. et al. Crystal structures that suggest late development of genetic code components for differentiating aromatic side chains. PNAS. 2003;100:15376–15380. [PMC free article: PMC307575] [PubMed: 14671330]
Copyright © 2000-2013, Landes Bioscience.
Bookshelf ID: NBK6228