Intrinsic precision of aminoacyl-tRNA synthesis enhanced through parallel systems of ligands

Competition of aminoacyl-tRNA synthetases for tRNA ensures the accuracy of aminoacylation

The accuracy of protein biosynthesis rests on the high fidelity with which aminoacyl-tRNA synthetases discriminate between tRNAs. Correct aminoacylation depends not only on identity elements (nucleotides in certain positions) in tRNA (1), but also on competition between different synthetases for a given tRNA (2). Here we describe in vivo and in vitro experiments which demonstrate how variations in the levels of synthetases and tRNA affect the accuracy of aminoacylation. We show in vivo that concurrent overexpression of Escherichia coli tyrosyl-tRNA synthetase abolishes misacylation of supF tRNA(Tyr) with glutamine in vivo by overproduced glutaminyl-tRNA synthetase. In an in vitro competition assay, we have confirmed that the overproduction mischarging phenomenon observed in vivo is due to competition between the synthetases at the level of aminoacylation. Likewise, we have been able to examine the role competition plays in the identity of a non-suppressor tRNA of ambiguous identity, tRNA(Glu). Finally, with this assay, we show that the identity of a tRNA and the accuracy with which it is recognized depend on the relative affinities of the synthetases for the tRNA. The in vitro competition assay represents a general method of obtaining qualitative information on tRNA identity in a competitive environment (usually only found in vivo) during a defined step in protein biosynthesis, aminoacylation. In addition, we show that the discriminator base (position 73) and the first base of the anticodon are important for recognition by E. coli tyrosyl-tRNA synthetase.

Altered expression of plant lysyl tRNA synthetase promotes tRNA misacylation and translational recoding of lysine

The Arabidopsis thaliana lysyl tRNA synthetase (AtKRS) structurally and functionally resembles the well-characterized prokaryotic class IIb KRS, including the propensity to aminoacylate tRNA(Lys) with suboptimal identity elements, as well as non-cognate tRNAs. Transient expression of AtKRS in carrot cells promotes aminoacylation of such tRNAs in vivo and translational recoding of lysine at nonsense codons. Stable expression of AtKRS in Zea mays causes translational recoding of lysine into zeins, significantly enriching the lysine content of grain.

Uniform binding of aminoacyl-tRNAs to elongation factor Tu by thermodynamic compensation

Elongation factor Tu (EF-Tu) binds all elongator aminoacyl-transfer RNAs (aa-tRNAs) for delivery to the ribosome during protein synthesis. Here, we show that EF-Tu binds misacylated tRNAs over a much wider range of affinities than it binds the corresponding correctly acylated tRNAs, suggesting that the protein exhibits considerable specificity for both the amino acid side chain and the tRNA body. The thermodynamic contributions of the amino acid and the tRNA body to the overall binding affinity are independent of each other and compensate for one another when the tRNAs are correctly acylated. Because certain misacylated tRNAs bind EF-Tu significantly more strongly or weakly than cognate aa-tRNAs, EF-Tu may contribute to translational accuracy.

Twenty-first aminoacyl-tRNA synthetase-suppressor tRNA pairs for possible use in site-specific incorporation of amino acid analogues into proteins in eukaryotes and in eubacteria

Two critical requirements for developing methods for the site-specific incorporation of amino acid analogues into proteins in vivo are (i) a suppressor tRNA that is not aminoacylated by any of the endogenous aminoacyl-tRNA synthetases (aaRSs) and (ii) an aminoacyl-tRNA synthetase that aminoacylates the suppressor tRNA but no other tRNA in the cell. Here we describe two such aaRS-suppressor tRNA pairs, one for use in the yeast Saccharomyces cerevisiae and another for use in Escherichia coli. The "21st synthetase-tRNA pairs" include E. coli glutaminyl-tRNA synthetase (GlnRS) along with an amber suppressor derived from human initiator tRNA, for use in yeast, and mutants of the yeast tyrosyl-tRNA synthetase (TyrRS) along with an amber suppressor derived from E. coli initiator tRNA, for use in E. coli. The suppressor tRNAs are aminoacylated in vivo only in the presence of the heterologous aaRSs, and the aminoacylated tRNAs function efficiently in suppression of amber codons. Plasmids carrying the E. coli GlnRS gene can be stably maintained in yeast. However, plasmids carrying the yeast TyrRS gene could not be stably maintained in E. coli. This lack of stability is most likely due to the fact that the wild-type yeast TyrRS misaminoacylates the E. coli proline tRNA. By using error-prone PCR, we have isolated and characterized three mutants of yeast TyrRS, which can be stably expressed in E. coli. These mutants still aminoacylate the suppressor tRNA essentially quantitatively in vivo but show increased discrimination in vitro for the suppressor tRNA over the E. coli proline tRNA by factors of 2.2- to 6.8-fold.

tRNA leucine identity and recognition sets

Transfer RNAs (tRNAs) are grouped into two classes based on the structure of their variable loop. In Escherichia coli, tRNAs from three isoaccepting groups are classified as type II. Leucine tRNAs comprise one such group. We used both in vivo and in vitro approaches to determine the nucleotides that are required for tRNA(Leu) function. In addition, to investigate the role of the tRNA fold, we compared the in vivo and in vitro characteristics of type I tRNA(Leu) variants with their type II counterparts.A minimum of six conserved tRNA(Leu) nucleotides were required to change the amino acid identity and recognition of a type II tRNA(Ser) amber suppressor from a serine to a leucine residue. Five of these nucleotides affect tRNA tertiary structure; the G15-C48 tertiary "Levitt base-pair" in tRNA(Ser) was changed to A15-U48; the number of nucleotides in the alpha and beta regions of the D-loop was changed to achieve the positioning of G18 and G19 that is found in all tRNA(Leu); a base was inserted at position 47n between the base-paired extra stem and the T-stem; in addition the G73 "discriminator" base of tRNA(Ser) was changed to A73. This minimally altered tRNA(Ser) exclusively inserted leucine residues and was an excellent in vitro substrate for LeuRS. In a parallel experiment, nucleotide substitutions were made in a glutamine-inserting type I tRNA (RNA(SerDelta); an amber suppressor in which the tRNA(Ser) type II extra-stem-loop is replaced by a consensus type I loop). This "type I" swap experiment was successful both in vivo and in vitro but required more nucleotide substitutions than did the type II swap. The type I and II swaps revealed differences in the contributions of the tRNA(Leu) acceptor stem base-pairs to tRNA(Leu) function: in the type I, but not the type II fold, leucine specificity was contingent on the presence of the tRNA(Leu) acceptor stem sequence. The type I and II tRNAs used in this study differed only in the sequence and structure of the variable loop. By altering this loop, and thereby possibly introducing subtle changes into the overall tRNA fold, it became possible to detect otherwise cryptic contributions of the acceptor stem sequence to recognition by LeuRS. Possible reasons for this effect are discussed.

The reliability of in vivo structure-function analysis of tRNA aminoacylation

The G.U wobble base-pair in the acceptor helix of Escherichia coli tRNAAlais critical for aminoacylation by the alanine synthetase. Previous work by several groups probed the mechanism of enzyme recognition of G.U by a structure-function analysis of mutant tRNAs using either a cell assay (amber suppressor tRNA) or a test tube assay (phage T7 tRNA substrate and purified enzyme). However, the aminoacylation capacity of particular mutant tRNAs was about 10(4)-fold higher in the cell assay. This led us to scrutinize the cell assay to determine if any parameter exaggerates the extent of aminoacylation in mutants forming substantial amounts of alanyl-tRNAAla. In doing so, we have refined and developed experimental designs to analyze tRNA function. We examined the level of aminoacylation of amber suppressor tRNAAlawith respect to the method of isolating aminoacyl-tRNA, the rate of cell growth, the cellular levels of alanine synthetase and elongation factor TU (EF-Tu), the amount of tRNA and the characteristics of EF-Tu binding. Within the precision of our measurements, none of these parameters varied in a way that could significantly amplify cellular alanyl-tRNAAla. A key observation is that the extent of aminoacylation of tRNAAlawas independent of tRNAAlaconcentration over a 75-fold range. Therefore, the cellular assay of tRNAAlareflects the substrate quality of the molecule for formation of alanyl-tRNAAla. These experiments support the authenticity of the cellular assay and imply that a condition or factor present in the cell assay may be absent in the test tube assay.

Specific, rapid synthesis of Phe-RNA by RNA

RNA 77, derived by selection amplification, accelerates its own conversion to Phe-RNA (relative to randomized RNA) more than 6 x 10(7)-fold, by using amino acid adenylates as substrate. A modified assay system allows measurement of slow rates of aa-RNA formation, which for disfavored amino acid substrates can be more than 10(4)-fold slower than phenylalanine. Thus unlike previously characterized self-aminoacylators, RNA 77 catalysis is highly amino acid selective. Remarkably, both rates of aminoacyl transfer and amino acid specificities are greater for RNA 77 than measured for protein PheRS. These data experimentally support the possible existence of an ancestral amino acid-specific translation system relying entirely on RNA catalysis. RNA 77 itself embodies a possible transitional evolutionary state, in which side-chain-specific aa-RNA formation and anticodon-codon pairing were invested in the same molecule.

Functional compensation by particular nucleotide substitutions of a critical G*U wobble base-pair during aminoacylation of transfer RNA

Expression of the genetic code depends on precise tRNA aminoacylation by cognate aminoacyl-tRNA synthetase enzymes. The G.U wobble base-pair in the acceptor helix of Escherichia coli alanine tRNA is the primary aminoacylation determinant of this molecule. Previous work on the process of synthetase recognition of the G.U pair showed that replacing G.U by a G.C Watson-Crick base-pair inactivates alanine acceptance by the tRNA, but that C.A and G.A wobble pair replacements preserve acceptance. Work by another group reported that the effects of a G.C replacement were reversed by a distal wobble base-pair in the anticodon helix. This result is potentially interesting because it suggests that distant regions in alanine tRNA are functionally coupled during synthetase recognition and more generally because recognition determinants of many other tRNAs lie in both the acceptor helix and anticodon helix region. Here, we have conducted an extensive in vivo analysis of the distal wobble pair in alanine tRNA and report that it does not behave like a compensating mutation. Restoration of alanine acceptance was not detected even when the synthetase enzyme was overproduced. We discuss the previous experimental evidence and suggest how the distal wobble pair was incorrectly analyzed. The available data indicate that all principal recognition determinants of alanine tRNA lie in the molecule's acceptor helix.

Universal rules and idiosyncratic features in tRNA identity

Correct expression of the genetic code at translation is directly correlated with tRNA identity. This survey describes the molecular signals in tRNAs that trigger specific aminoacylations. For most tRNAs, determinants are located at the two distal extremities: the anticodon loop and the amino acid accepting stem. In a few tRNAs, however, major identity signals are found in the core of the molecule. Identity elements have different strengths, often depend more on k cat effects than on K m effects and exhibit additive, cooperative or anti-cooperative interplay. Most determinants are in direct contact with cognate synthetases, and chemical groups on bases or ribose moieties that make functional interactions have been identified in several systems. Major determinants are conserved in evolution; however, the mechanisms by which they are expressed are species dependent. Recent studies show that alternate identity sets can be recognized by a single synthetase, and emphasize the importance of tRNA architecture and anti-determinants preventing false recognition. Identity rules apply to tRNA-like molecules and to minimalist tRNAs. Knowledge of these rules allows the manipulation of identity elements and engineering of tRNAs with switched, altered or multiple specificities.

Major identity determinants for enzymatic formation of ribothymidine and pseudouridine in the T psi-loop of yeast tRNAs

Almost all transfer RNA molecules sequenced so far contain two universal modified nucleosides at positions 54 and 55, respectively: ribothymidine (T54) and pseudouridine (psi 55). To identify the tRNA elements recognized by tRNA:m5uridine-54 methyltransferase and tRNA:pseudouridine-55 synthase from the yeast Saccharomyces cerevisiae, a set of 43 yeast tRNA(Asp) mutants were used. Some variants contained point mutations, while the others included progressive reductions in size down to a tRNA minisubstrate consisting of the T psi-loop with only one G.C base-pair as stem (9-mer). All substrates (full-sized tRNA(Asp) and various minihelices) were produced in vitro by T7 transcription and tested using yeast extract (S100) as a source of enzymatic activities and S-adenosyl-L-methionine as a methyl donor. The results indicate that the minimal substrate for enzymatic formation of psi 55 is a stem/loop structure with only four G.C base-pairs in the stem, while a longer stem is required for efficient T54 formation. None of the conserved nucleotides (G53, C56, A58 and C61) and U54 for psi 55 or U55 for T54 formation can be replaced by any of the other three canonical nucleotides. Yeast tRNA:m5uridine-54 methyltransferase additionally requires the presence of a pyrimidine-60 in the loop. Interestingly, in a tRNA(Asp) variant in which the T psi-loop was permuted with the anticodon-loop, the new U32 and U33 residues derived from the T psi-loop were quantitatively converted to T32 and psi 33, respectively. Structural mapping of this variant with ethylnitrosourea confirmed that the intrinsic characteristic structure of the T psi-loop was conserved upon permutation and that the displaced anticodon-loop did not acquire a T psi-loop structure. These results demonstrate that a local conformation rather than the exact location of the U-U sequence within the tRNA architecture is the important identity determinant for recognition by yeast tRNA:m5uridine-54 methyltransferase and tRNA:pseudouridine-55 synthase.

Important role of the amino acid attached to tRNA in formylation and in initiation of protein synthesis in Escherichia coli

In attempts to convert an elongator tRNA to an initiator tRNA, we previously generated a mutant elongator methionine tRNA carrying an anticodon sequence change from CAU to CUA along with the two features important for activity of Escherichia coli initiator tRNA in initiation. This mutant tRNA (Mi:2 tRNA) was active in initiation in vivo but only when aminoacylated with methionine by overproduction of methionyl-tRNA synthetase. Here we show that the Mi:2 tRNA is normally aminoacylated in vivo with lysine and that the tRNA aminoacylated with lysine is a very poor substrate for formylation compared with the same tRNA aminoacylated with methionine. By introducing further changes at base pairs 4:69 and 5:68 in the acceptor stem of the Mi:2 tRNA to those found in the E. coli initiator tRNA, we show that change of the U4:A69 base pair to G4:C69 and overproduction of lysyl-tRNA synthetase and methionyl-tRNA transformylase results in partial formylation of the mutant tRNA and activity of the formyllysyl-tRNAs in initiation of protein synthesis. Thus, the G4: C69 base pair contributes toward formylation of the tRNA and protein synthesis in E. coli can be initiated with formyllysine. We also discuss the implications of these and other results on recognition of tRNAs by E. coli lysyl-tRNA synthetase and on competition in cells among aminoacyl-tRNA synthetases.

Distinctive acceptor-end structure and other determinants of Escherichia coli tRNAPro identity

The previously uncharacterized determinants of the specificity of tRNAPro for aminoacylation (tRNAPro identity) were defined by a computer comparison of all Escherichia coli tRNA sequences and tested by a functional analysis of amber suppressor tRNAs in vivo. We determined the amino acid specificity of tRNA by sequencing a suppressed protein and the aminoacylation efficiency of tRNA by examining the steady-state level of aminoacyl-tRNA. On substituting nucleotides derived from the acceptor end and variable pocket of tRNAPro for the corresponding nucleotides in a tRNAPhe gene, the identity of the resulting tRNA changed substantially but incompletely to that of tRNAPro. The redesigned tRNAPhe was weakly active and aminoacyl-tRNA was not detected. Ethyl methanesulfonate mutagenesis of the redesigned tRNAPhe gene produced a mutant with a wobble pair in place of a base pair in the end of the acceptor-stem helix of the transcribed tRNA. This mutant exhibited both a tRNAPro identity and substantial aminoacyl-tRNA. The results speak for the importance of a distinctive conformation in the acceptor-stem helix of tRNAPro for aminoacylation by the prolyl-tRNA synthetase. The anticodon also contributes to tRNAPro identity but is not necessary in vivo.

The transfer RNA identity problem: a search for rules

Correct recognition of transfer RNAs (tRNAs) by aminoacyl-tRNA synthetases is central to the maintenance of translational fidelity. The hypothesis that synthetases recognize anticodon nucleotides was proposed in 1964 and had considerable experimental support by the mid-1970s. Nevertheless, the idea was not widely accepted until relatively recently in part because the methodologies initially available for examining tRNA recognition proved hampering for adequately testing alternative hypotheses. Implementation of new technologies has led to a reasonably complete picture of how tRNAs are recognized. The anticodon is indeed important for 17 of the 20 Escherichia coli isoaccepting groups. For many of the isoaccepting groups, the acceptor stem or position 73 (or both) is important as well.

Mosaic tile model for tRNA-enzyme recognition

An improved algorithm was elaborated to analyse tRNA interaction with aminoacyl-tRNA synthetase based on analysis of tRNA sequences. The fundamental element defining the interaction between the tRNA and the synthetase is not a single nucleotide but a nucleotide combination named a tile which comprises of a given nucleotide and its neighbours as they are defined by the tertiary structure of the molecule. Informational content of each tile is calculated as its probability to occur exclusively in a set of cognate tRNAs. Based on this algorithm the identity sites of E. coli tRNA(Ala) and tRNA(Gln) were determined. The results are in a good agreement with the biochemical data and provide new information about identity sites of these tRNAs.

From elongator tRNA to initiator tRNA

We show that the two most important properties needed for a tRNA to function in initiation in Escherichia coli are its ability to be formylated and its ability to bind to the ribosomal P site. This conclusion is based on conversion of two different elongator tRNAs to ones that can act as initiators in E. coli. We transplanted the features unique to E. coli and eubacterial initiator tRNAs to E. coli elongator methionine tRNA (tRNA(Met)) along with an anticodon sequence change and analyzed their activities in initiation in E. coli. Introduction of a C1.A72 mismatch at the end of the acceptor stem of tRNA(Met), which generates the minimal features necessary for formylation, produces a tRNA with very low activity in initiation. Subsequent introduction of three consecutive G.C base pairs at the bottom of the anticodon stem, which is necessary for ribosomal P site binding, produces a tRNA with significant activity in initiation. Furthermore, introduction of the features necessary for formylation and for ribosomal P site binding into E. coli elongator glutamine tRNA produces a tRNA that initiates protein synthesis in E. coli.

Acceptor end binding domain interactions ensure correct aminoacylation of transfer RNA

The recognition of the acceptor stem of tRNA(Gln) is an important element ensuring the accuracy of aminoacylation by Escherichia coli glutaminyl-tRNA synthetase (GlnRS; EC 6.1.1.18). On the basis of known mutations and the crystal structure of the tRNA(Gln).GlnRS complex, we mutagenized at saturation two motifs in the acceptor end binding domain of GlnRS. Mutants with lowered tRNA specificity were then selected in vivo by suppression of a glutamine-specific amber mutation (lacZ1000) with an amber suppressor tRNA derived from tRNA(1Ser). The mischarging GlnRS mutants obtained in this way retain the ability to charge tRNA(Gln), but in addition, they misacylate a number of noncognate amber suppressor tRNAs. The critical residues responsible for specificity are Arg-130 and Glu-131, located in a part of GlnRS that binds the acceptor stem of tRNA(Gln). On the basis of the spectrum of tRNAs capable of being misacylated by such mutants we propose that, in addition to taking part in productive interactions, the acceptor end binding domain contributes to recognition specificity by rejecting noncognate tRNAs through negative interactions. Analysis of the catalytic properties of one of the mischarging enzymes, GlnRS100 (Arg-130-->Pro, Glu-131-->Asp), indicates that, while the kinetic parameters of the mutant enzyme are not dramatically changed, it binds noncognate tRNA(Glu) more stably than the wild-type enzyme does (Kd is 1/8 that of the wild type). Thus, the stability of the noncognate complex may be the basis for mischarging in vivo.

Association of tRNA(Gln) acceptor identity with phosphate-sugar backbone interactions observed in the crystal structure of the Escherichia coli glutaminyl-tRNA synthetase-tRNA(Gln) complex

We isolated several mutants with nucleotide substitutions in alanine tRNA (tRNA(Ala)) that resulted in glutamine tRNA (tRNA(Gln)) acceptor identity in Escherichia coli. These substitutions were in three regions of tRNA structure not previously associated with tRNA(Gln) acceptor identity. Only the phosphate-sugar backbone moieties of these nucleotides interact with the enzyme in the previously determined X-ray crystal structure of the complex between tRNA(Gln) and glutaminyl-tRNA synthetase. We conclude that these sequence-dependent phosphate-sugar backbone interactions contribute to tRNA(Gln) identity, and argue that the interactions help communicate enzyme recognition of the anticodon to the acceptor end of the tRNA and the catalytic center of the enzyme.

Discrimination between transfer-RNAs by tyrosyl-tRNA synthetase

We have constructed a model of the complex between tyrosyl-tRNA synthetase (TyrRS) from Bacillus stearothermophilus and tRNA(Tyr) by successive cycles of predictions, mutagenesis of TyrRS and molecular modeling. We confront this model with data obtained independently, compare it to the crystal structures of other complexes and review recent data on the discrimination between tRNAs by TyrRS. Comparison of the crystal structures of TyrRS and GlnRS, both of which are class I synthetases, and comparison of the identity elements of tRNA(Tyr) and tRNA(Gln) indicate that the two synthetases bind their cognate tRNAs differently. The mutagenesis data on tRNA(Tyr) confirm the model of the TyrRS:tRNA(Tyr) complex on the following points. TyrRS approaches tRNA(Tyr) on the side of the variable loop. The bases of the first three pairs of the acceptor stem are not recognized. The presence of the NH2 group in position C6 and the absence of a bulky group in position C2 are important for the recognition of the discriminator base A73 by TyrRS, which is fully realized only in the transition state for the acyl transfer. The anticodon is the major identity element of tRNA(Tyr). We have set up an in vivo approach to study the effects of synthetase mutations on the discrimination between tRNAs. Using this approach, we have shown that residue Glu152 of TyrRS acts as a purely negative discriminant towards non-cognate tRNAs, by electrostatic and steric repulsions. The overproductions of the wild type TyrRSs from E coli and B stearothermophilus are toxic to E coli, due to the mischarging or the non-productive binding of tRNAs. The construction of a family of hybrids between the TyrRSs from E coli and B stearothermophilus has shown that their sequences and structures have remained locally compatible through evolution, for folding and function, in particular for the specific recognition and charging of tRNA(Tyr).

Selectivity and specificity in the recognition of tRNA by E coli glutaminyl-tRNA synthetase

The specific recognition by Escherichia coli glutaminyl-tRNA synthetase (GlnRS) of tRNA(Gln) is mediated by extensive protein:RNA contacts and changes in the conformation of tRNA(Gln) when complexed with GlnRS. In vivo accuracy of aminoacylation depends on two factors: competition between synthetases, and the context and recognition of identity elements in the tRNA. The structure of the tRNA(Gln):GlnRS complex supports studies from amber and opal suppressor tRNAs, complemented by in vitro aminoacylation of the mutated tRNA transcripts, that the glutamine identity elements are located in the anticodon and acceptor stem of tRNA(Gln). Recognition of individual functional groups in tRNA, for example the 2-amino group of guanosine, is also evident from the result with inosine-substituted tRNAs. Communication between anticodon and acceptor stem recognition is indicated by mutants in GlnRS isolated by genetic selection with opal suppressor tRNAs which are altered in interactions with the inside of the L-shaped tRNA. We have also used genetic selection to obtain mutants of GlnRS altered in acceptor stem recognition with relaxed specificity for amber suppressor tRNAs, and a more extensive mutational analysis shows the importance of the acceptor binding domain to accurate recognition of tRNA.

Synthetase competition and tRNA context determine the in vivo identify of tRNA discriminator mutants

The discriminator nucleotide (position 73) in tRNA has long been thought to play a role in tRNA identity as it is the only variable single-stranded nucleotide that is found near the site of aminoacylation. For this reason, a complete mutagenic analysis of the discriminator in three Escherichia coli amber suppressor tRNA backgrounds was undertaken; supE and supE-G1C72 glutamine tRNAs, gluA glutamate tRNA and supF tyrosine tRNA. The effect of mutation of the discriminator base on the identity of these tRNAs in vivo was assayed by N-terminal protein sequencing of E. coli dihydrofolate reductase, which is the product of suppression by the mutated amber suppressors, and confirmed by amino acid specific suppression experiments. In addition, suppressor efficiency assays were used to estimate the efficiency of aminoacylation in vivo. Our results indicate that the supE glutamine tRNA context can tolerate multiple mutations (including mutation of the discriminator and first base-pair) and still remain predominantly glutamine-accepting. Discriminator mutants of gluA glutamate tRNA exhibit increased and altered specificity probably due to the reduced ability of other synthetases to compete with glutamyl-tRNA synthetase. In the course of these experiments, a glutamate-specific mutant amber suppressor, gluA-A73, was created. Finally, in the case of supF tyrosine tRNA, the discriminator is an important identity element with partial to complete loss of tyrosine specificity resulting from mutation at this position. It is clear from these experiments that it may not be possible to assign a specific role in tRNA identity to the discriminator. The identity of a tRNA in vivo is determined by competition among aminoacyl-tRNA synthetases, which is in turn modulated by the nucleotide substitution as well as the tRNA context.

Role of methionine and formylation of initiator tRNA in initiation of protein synthesis in Escherichia coli

We showed recently that a mutant of Escherichia coli initiator tRNA with a CAU-->CUA anticodon sequence change can initiate protein synthesis from UAG by using formylglutamine instead of formylmethionine. We further showed that coupling of the anticodon sequence change to mutations in the acceptor stem that reduced Vmax/Km(app) in formylation of the tRNAs in vitro significantly reduced their activity in initiation in vivo. In this work, we have screened an E. coli genomic DNA library in a multicopy vector carrying one of the mutant tRNA genes and have found that the gene for E. coli methionyl-tRNA synthetase (MetRS) rescues, partially, the initiation defect of the mutant tRNA. For other mutant tRNAs, we have examined the effect of overproduction of MetRS on their activities in initiation and their aminoacylation and formylation in vivo. Some but not all of the tRNA mutants can be rescued. Those that cannot be rescued are extremely poor substrates for MetRS or the formylating enzyme. Overproduction of MetRS also significantly increases the initiation activity of a tRNA mutant which can otherwise be aminoacylated with glutamine and fully formylated in vivo. We interpret these results as follows. (i) Mutant initiator tRNAs that are poor substrates for MetRS are aminoacylated in part with methionine when MetRS is overproduced. (ii) Mutant tRNAs aminoacylated with methionine are better substrates for the formylating enzyme in vivo than mutant tRNAs aminoacylated with glutamine. (iii) Mutant tRNAs carrying formylmethionine are significantly more active in initiation than those carrying formylglutamine. Consequently, a subset of mutant tRNAs which are defective in formylation and therefore inactive in initiation when they are aminoacylated with glutamine become partially active when MetRS is overproduced.

Competition of aminoacyl-tRNA synthetases for tRNA ensures the accuracy of aminoacylation

The accuracy of protein biosynthesis rests on the high fidelity with which aminoacyl-tRNA synthetases discriminate between tRNAs. Correct aminoacylation depends not only on identity elements (nucleotides in certain positions) in tRNA (1), but also on competition between different synthetases for a given tRNA (2). Here we describe in vivo and in vitro experiments which demonstrate how variations in the levels of synthetases and tRNA affect the accuracy of aminoacylation. We show in vivo that concurrent overexpression of Escherichia coli tyrosyl-tRNA synthetase abolishes misacylation of supF tRNA(Tyr) with glutamine in vivo by overproduced glutaminyl-tRNA synthetase. In an in vitro competition assay, we have confirmed that the overproduction mischarging phenomenon observed in vivo is due to competition between the synthetases at the level of aminoacylation. Likewise, we have been able to examine the role competition plays in the identity of a non-suppressor tRNA of ambiguous identity, tRNA(Glu). Finally, with this assay, we show that the identity of a tRNA and the accuracy with which it is recognized depend on the relative affinities of the synthetases for the tRNA. The in vitro competition assay represents a general method of obtaining qualitative information on tRNA identity in a competitive environment (usually only found in vivo) during a defined step in protein biosynthesis, aminoacylation. In addition, we show that the discriminator base (position 73) and the first base of the anticodon are important for recognition by E. coli tyrosyl-tRNA synthetase.

Proofreading, NTPases and translation: successful increase in specificity

The discussion of proofreading started in the April issue of TIBS is completed by treating the two branched Michaelis-Menten enzymes that can proofread. The conditions required for proofreading can be seen to determine the expression of proofreading in its biological settings. There are surely instances of proofreading as yet unrecognized.

Role of residue Glu152 in the discrimination between transfer RNAs by tyrosyl-tRNA synthetase from Bacillus stearothermophilus

Residue Glu152 of tyrosyl-tRNA synthetase (TyrTS) from Bacillus stearothermophilus is close to phosphate groups 73 and 74 of tRNATyr in the structural model of their complex. TyrTS(E152A), a mutant synthetase carrying the change of Glu152 to Ala, was toxic when overproduced in Escherichia coli. The toxicity strongly increased with the growth temperature. It was measured by the ratios of the efficiencies with which the producing cells plated in induced or repressed conditions and at 30 degrees C or 37 degrees C. TyrTS(E152Q), TyrTS(E152D) and the wild-type synthetase were not toxic in conditions where TyrTS(E152A) was toxic. The toxicity of TyrTS(E152A) was abolished by additional mutations of the synthetase that prevent the binding of tRNATyr but not by a mutation that prevents the formation of Tyr-AMP. Because TyrTS(E152A) was active for the aminoacylation of tRNATyr, its toxicity could only be due to faulty interactions with non-cognate tRNAs, either their non-productive binding or their mischarging with tyrosine. TyrTS(E152A) and TyrTS(E152Q) mischarged tRNAPhe and tRNAVal in vitro with tyrosine unlike TyrTS(E152D) or the wild-type enzyme. Thus, several features of the side-chain in position 152 of TyrTS, including its negative charge, are important for the rejection of non-cognate tRNAs. TyrTS(E152A), TyrTS(E152D) and TyrTS(E152Q) had similar steady-state kinetics parameters for the charging of tRNATyr with tyrosine in vitro, with kcat/KM ratios improved 2.5 times relative to the wild-type synthetase. We conclude that the side-chain of residue Glu152 weakens the binding of TyrTS to tRNATyr and prevents its interaction with non-cognate tRNAs.

Mutant enzymes and tRNAs as probes of the glutaminyl-tRNA synthetase: tRNA(Gln) interaction

This paper focuses on several aspects of the specificity of mutants of Escherichia coli glutaminyl-tRNA synthetase (GlnRS) and tRNA(Gln). Temperature-sensitive mutants located in glnS, the gene for GlnRS, have been described previously. The mutations responsible for the temperature-sensitive phenotype were analyzed, and pseudorevertants of these mutants isolated and characterized. The nature of these mutations is discussed in terms of their location in the three-dimensional structure of the tRNA(Gln).GlnRS complex. In order to characterize the specificity of the aminoacylation reaction, mutant tRNA(Gln) species were synthesized with either a 2'-deoxy AMP or 3'-deoxy AMP as their 3'-terminal nucleotide. Subsequent assays for aminoacylation and ATP/PPi exchange activity established the esterification of glutamine to the 2'-hydroxyl of the terminal adenosine; there is no glutaminylation of the 3'-OH group. This correlates with the classification of GlnRS as a class I aminoacyl-tRNA synthetase. Mutations in tRNA(Gln) are discussed which affect the recognition of GlnRS and the current concept of glutamine identity in E coli is reviewed.

Four sites in the acceptor helix and one site in the variable pocket of tRNA(Ala) determine the molecule's acceptor identity

The structural features that determine tRNA(Ala) acceptor identity have been studied with amber-suppressor tRNAs in Escherichia coli cells. Previous work established that a wobble pair composed of guanosine at position 3 and uridine at position 70 (G3-U70) in the acceptor helix of tRNA(Ala) is a determinant of the molecule's acceptor identity. We show that additional determinants are located at three other sites in the acceptor helix and at one site in the variable pocket of tRNA(Ala). These latter determinants are less important than G3.U70 since their individual alterations in mutants of tRNA(Ala) have smaller degrading effects on the functions of the molecules, and subsets of the determinants, when combined with G3.U70, are sufficient to switch the identities of several other tRNAs to that of tRNA(Ala). Other workers are using fragments of the tRNA(Ala) acceptor helix to study the molecule's acceptor identity. Our demonstration that the variable pocket contributes to tRNA(Ala) acceptor identity means that such fragments do not faithfully replicate the structure-function relationship of the cellular process.

Rapid determination of nucleotides that define tRNA(Gly) acceptor identity

Expression of the genetic code depends on the recognition of specific tRNAs by the enzymes that aminoacylate them. A computer comparison of tRNA sequences coupled with analysis of mutant nonsense-suppressor tRNAs has revealed the structural features that distinguish the acceptor identity of Escherichia coli tRNA(Gly) from tRNAs that accept phenylalanine, arginine, lysine, and glutamine. On replacement of several nucleotides in the acceptor stem and anticodon of the latter tRNAs with tRNA(Gly)-derived residues, the resulting molecules acquired a tRNA(Gly) identity.

Exploring the aminoacylation function of transfer RNA by macromolecular engineering approaches. Involvement of conformational features in the charging process of yeast tRNA(Asp)

This report presents the conceptual and methodological framework that presently underlies the experiments designed to decipher the structural features in tRNA important for its aminoacylation by aminoacyl-tRNA synthetases. It emphasizes the importance of conformational features in tRNA for an optimized aminoacylation. This is illustrated by selected examples on yeast tRNA(Asp). Using the phage T7 transcriptional system, a series of tRNA(Asp) variants were created in which conformational elements were modified. It is shown that aspartyl-tRNA synthetase tolerates conformational variability in tRNA(Asp) at the level of the D-loop and variable region, of the tertiary Levitt base-pair 15-48 which can be inverted and in the T-arm in which residue 49 can be excised. However, changing the anticodon region completely abolishes the aspartylation capacity of the variants. Transplanting the phenylalanine identity elements into a different tRNA(Asp) variant presenting conformational characteristics of tRNA(Phe) converts this molecule into a phenylalanine acceptor but is less efficient than wild-type tRNA(Phe). This engineered tRNA completely loses its aspartylation capacity, showing that some aspartic acid and phenylalanine identity determinants overlap. The fact that chimeric tRNA(Asp) molecules with altered anticodon regions lose their aspartylation capacity demonstrates that this region is part of the aspartic acid identity of tRNA(Asp).

Accuracy of in vivo aminoacylation requires proper balance of tRNA and aminoacyl-tRNA synthetase

The fidelity of protein biosynthesis in any cell rests on the accuracy of aminoacylation of tRNA. The exquisite specificity of this reaction is critically dependent on the correct recognition of tRNA by aminoacyl-tRNA synthetases. It is shown here that the relative concentrations of a tRNA and its cognate aminoacyl-tRNA synthetase are normally well balanced and crucial for maintenance of accurate aminoacylation. When Escherichia coli Gln-tRNA synthetase is overproduced in vivo, it incorrectly acylates the supF amber suppressor tRNA(Tyr) with Gln. This effect is abolished when the intracellular concentration of the cognate tRNA(Gln2) is also elevate. These data indicate that the presence of aminoacyl-tRNA synthetase and the cognate tRNAs in complexed form, which requires the proper balance of the two macromolecules, is critical in maintaining the fidelity of protein biosynthesis. Thus, limits exist on the relative levels of tRNAs and aminoacyl-tRNA synthetases within a cell.

Changing the acceptor identity of a transfer RNA by altering nucleotides in a "variable pocket"

The specificity of tRNA(Arg) (arginine transfer RNA) for aminoacylation (its acceptor identity) were first identified by computer analysis and then examined with amber suppressor tRNAs in Escherichia coli. On replacing two nucleotides in tRNA(Phe) (phenylalanine transfer RNA) with the corresponding nucleotides from tRNA(Arg), the acceptor identity of the resulting tRNA was changed to that of tRNA(Arg). The nucleotides used in the identity transformation occupy a "variable pocket" structure on the surface of the tRNA molecule where two single-stranded loop segments interact. The middle nucleotide in the anticodon also probably contributes to the interaction, since an amber suppressor of tRNA(Arg) had an acceptor identity for lysine as well as arginine.

Specific incorporation of 5-fluorocytidine into Escherichia coli RNA

RNAs isolated from Escherichia coli B grown in the presence of 5-fluorouracil have high levels of the analog replacing uridine and uridine-derived modified nucleosides. Cytidine has also been shown to be replaced in these RNAs by 5-fluorocytidine, a metabolic product of 5-fluorouracil, but to a considerably lesser extent. When 5-fluorocytidine is added to cultured of E. coli B little 5-fluorocytidine (0.20 mol%) is incorporated into cellular RNAs because of the active cytosine/cytidine deaminase activities. Addition of the cytidine deaminase inhibitor tetrahydrouridine (70 micrograms/ml) increases 5-fluorocytidine incorporation to about 3 mol% in tRNAs, but does not eliminate 5-fluorouridine incorporation. E. coli mutants lacking cytosine/cytidine deaminase activities are able to more than double the extent of 5-fluorocytidine incorporation into their transfer and ribosomal RNAs, replacing cytidine with no detectable 5-fluorouridine incorporation. Levels of 5-methyluridine, pseudouridine and dihydrouridine in tRNAs are not affected. These fluorocytidine-containing tRNAs show amino acid-accepting activities similar to control tRNAs. Fluorocytidine was found to be quite susceptible to deamination under alkaline conditions. Its conversion to primarily 5-fluorouridine follows pseudo-first-order reaction kinetics with a half-life of 10 h in 0.3 M KOH at 37 degrees C. This instability in alkali probably explains why 5-fluorocytidine was not found earlier in RNAs isolated from cells treated with 5-fluorouridine, since most early RNA hydrolyses were carried out in alkali. It may also explain the mild mutagenic properties observed in some systems following 5-fluorouridine treatment. Initial 19F-NMR measurements in fluorocytidine-containing tRNAs indicate that this modified tRNA may be useful in future structural studies of tRNAs and in probing tRNA-protein complexes.

Isoleucyl-tRNA synthetase from bakers' yeast: variable discrimination between tRNAIle and tRNAVal and different pathways of cognate and noncognate aminoacylation under standard conditions, in the presence of pyrophosphatase, elongation factor Tu-GTP complex, and spermine

Error rates in discrimination between cognate tRNAIle and noncognate tRNAVal in the aminoacylation reaction with isoleucine catalyzed by isoleucyl-tRNA synthetase from yeast have been investigated in three sets of experiments under different assay conditions. The overall discrimination factor was first determined by isoleucylation of tRNAVal/tRNAIle mixtures. In the second set of experiments, the number of AMP molecules formed per Ile-tRNA in the cognate and noncognate reactions was measured. The higher AMP formation in the noncognate aminoacylation is assigned to a proofreading reaction step. The calculated proofreading factors and an estimated initial discrimination factor yield overall discriminations that are consistent with those obtained from the first set of experiments. In the third series of studies, the orders of substrate addition and product release of cognate and noncognate isoleucylation reactions were investigated by initial rate kinetic methods. From kcat and Km values, the overall discrimination factors were calculated and showed again a good coincidence with those observed in the preceding sets of experiments. Besides under standard assay conditions, aminoacylation reactions were studied in the presence of pyrophosphatase or elongation factor Tu-GTP complex, under addition of both these proteins, in presence of these two additional proteins and spermine at high and low magnesium concentrations, and under special conditions that favor misacylations. Furthermore, isoleucylation of tRNAIle was tested at increased and decreased pH in the standard enzyme assay. Variation of the assay conditions results in changing discrimination factors, which differ by a factor of about 10. Substitution of tRNAIle by tRNAVal in the isoleucylation reaction causes changes in substrate addition and product release orders and thus of the whole catalytic cycle. For aminoacylation of tRNAIle, four different orders of substrate addition and product release appear: the sequential ordered ter-ter, the rapid equilibrium sequential random ter-ter, the random bi-uni uni-bi ping-pong, and a bi-bi uni-uni ping-pong mechanism with a rapid equilibrium segment. tRNAVal is aminoacylated in rapid equilibrium random ter-ter order, in a bi-bi uni-uni ping-pong mechanism with a rapid equilibrium segment, and in two bi-uni uni-bi ping-pong mechanisms. It is assumed that the different assay conditions can be regarded as a stepwise approximation to physiological conditions and that considerable changes in error rates may be also possible in vivo up to 1 order of magnitude.

Aminoacylation of undermethylated mammalian transfer RNA

To study the role of 5-methylcytidine in the aminoacylation of mammalian tRNA, bulk tRNA specifically deficient in 5-methylcytidine was isolated from the livers of mice treated with 5-azacytidine (18 mg/kg) for 4 days. For comparison, more extensively altered tRNA was isolated from the livers of mice treated with DL-ethionine (100 mg/kg) plus adenine (48 mg/kg) for 3 days. The amino acid acceptor capacity of these tRNAs was determined by measuring the incorporation of one of eight different 14C-labeled amino acids or a mixture of 14C-labeled amino acids in homologous assays using a crude synthetase preparation isolated from untreated mice. The 5-methylcytidine-deficient tRNA incorporated each amino acid to the same extent as fully methylated tRNA. The tRNA from DL-ethionine-treated livers showed an overall decreased amino-acylation capacity for all amino acids tested. The 5-methylcytidine-deficient tRNA from DL-ethionine-treated mice were further characterized as substrates in homologous rate assays designed to determine the Km and V of the aminoacylation reaction using four individual 14C-labeled amino acids and a mixture of 14C-labeled amino acids. The Km and V of the reactions for all amino acids tested using 5-methylcytidine-deficient tRNA as substrate were essentially the same as for fully methylated tRNA. However, the Km and V were increased when liver tRNA from mice treated with DL-ethionine plus adenine was used as substrate in the rate reaction with [14C]lysine as label. Our results suggest that although extensively altered tRNA is a poorer substrate than control tRNA in both extent and rate of aminoacylation, 5-methylcytidine in mammalian tRNA is not involved in the recognition of the tRNA by the synthetase as measured by aminoacylation activity.

Ribosome editing and the error catastrophe hypothesis of cellular aging

Ribosome editing involves the dissociation during protein synthesis of inappropriate peptidyl-tRNA's, ones whose structure does not correctly complement the codon of the mRNA. This process is one of three stages in protein biosynthesis in which the frequency of errors in cellular proteins is controlled. These stages are reviewed and the implications of ribosome editing are described. A model for stability of the translation apparatus is criticized. Calculations using a revision of the model and experimentally reasonable values for the various parameters show varying time courses for error catastrophes.

Theories of enzyme specificity and their application to proteases and aminoacyl-transfer RNA synthetases

The question of enzyme specificity which is a corollary of the phenomenon of biological recognition is reviewed. The following theories are outlined briefly: non-productive binding, induced fit, transition state binding, the general strain theory and the kinetic proofreading hypothesis. Data on proteolytic enzymes and aminoacyl-tRNA synthetases are discussed in the light of predictions made by the various theories. The specificity of inhibitor and substrate binding to chymotrypsin and subtilisins is revealed at the sub-molecular level as an example of binding specificity. Kinetic specificity is experimentally distinguished from binding specificity. Conformational adaptability of enzyme and substrate, which is crucial in some theories, is documented by data on aminoacyl-tRNA synthetases. Expected and observed specificity of tRNA charging is discussed with regard to a theoretical limit of specificity. Additional means seem necessary beside those contained in the isolated enzyme-substrate system to account for the high specificity of most synthetases. In conclusion, we have arrived at quite good explanations for moderate specificity such as is displayed by many proteases, but there are still ample difficulties in the understanding of highly specific enzyme reactions.

Kinetic amplification of enzyme discrimination

The dependence of the accuracy of enzymatic systems on the mechanism of the catalyzed reaction is investigated, using a probabilistic approach. Certain mechanisms of reaction, involving a delay in one of the steps act as kinetic amplifiers of molecular discriminations. The relationship between our scheme for a delayed reaction [1] and Hopfield's scheme [i] is discussed.